Gene Therapy & Molecular Biology Volume 1 Issue A - Part 2

Davie et al: Nuclear matrix in cancer diagnosis and chromatin structure

510

matrix consisting of residual nucleoli, surrounding nuclear

pore-lamina complex, and internal matrix is revealed when

nuclease-digested nuclei are extracted with salt (e.g., 0.25

M ammonium sulfate). The protocol that we use to

isolate nuclear matrices is shown in Fig . 1 . Briefly,

nuclei are digested with DNAase I followed by extraction

with 0.25 M ammonium sulfate, yielding NM1-IF

[nuclear matrices (NM) with attached intermediate

filaments (IF)] (Sun et al., 1994; Chen et al., 1996).

Further extraction of the NM1-IF nuclear matrices with 2

M NaCl yields NM2-IF. The internal matrix of NM1-IF

preparations has a fibrogranular appearance (Chen et al.,

1996). Extraction of the NM1-IF with high salt removes

proteins that decorate core filaments of the internal matrix

(Penman, 1995; Nickerson et al., 1995). The core

filament fiber network is also seen when nuclear DNA is

removed from nuclease-digested cells by electroelution in

solutions of physiological ionic strength (Jackson and

Cook, 1988). Core filaments, composition of which is

currently unknown, have a diameter of 10-13 nm. These

filaments appear to be the underlying structure onto which

other nuclear components are bound.

The nuclear matrix is composed of protein and RNA.

The nuclear pore-lamina consists of lamins and pore

proteins. The internal matrix has a complex protein

composition, with heterogeneous nuclear ribonuclear

proteins (hnRNP) being major components (Mattern et al.,

1996). Most nuclear RNA is associated with the nuclear

matrix and contributes to the structural integrity of the

nuclear matrix (Nickerson et al., 1995). The absence of

nuclear RNA may weaken nuclear matrix internal

structures. For example, nuclear matrices isolated from

chicken mature erythrocytes lack nuclear RNA and internal

structures, while nuclear matrices from immature

erythrocytes of anemic adult birds have internal structures

and nuclear RNA (Chen et al., 1996).

II. Nuclear matrix proteins and thediagnosis of cancer

The protein composition of the nuclear matrix is both

tissue and cell type specific, and undergoes changes with

differentiation and transformation (Fey and Penman, 1988;

Stuurman et al., 1990; Dworetzky et al., 1990; Cupo,

1991). Pathologists have long appreciated that irregular

nuclear appearance is the signature of a malignant cell

(Miller et al., 1992; Nickerson et al., 1995). Changes in

the composition of nuclear matrix proteins in malignant

cells may contribute to alterations in nuclear structure.

Nuclear matrix proteins are informative markers of disease

states (Khanuja et al., 1993; Keesee et al., 1994).

Informative nuclear matrix proteins have been identified for

bladder, breast, colon, prostate, head, and neck cancers

(Getzenberg et al., 1991a;1996; Khanuja et al., 1993;

Keesee et al., 1994; Donat et al., 1996). For example, the

nuclear matrix protein PC-1 is found in the nuclear matrix

proteins from prostate cancer but not in the nuclear matrix

from normal prostate or benign prostatic hyperplasia

(Getzenberg et al., 1991a). Recently, we reported that the

nuclear matrix protein composition was radically altered in

highly metastatic oncogene transformed mouse fibroblasts

(Samuel et al., 1997b). Interestingly, highly metastatic

ras-transformed 10T1/2 cells and highly metastatic fes-

transformed NIH 3T3 cells had a similar set of nuclear

matrix proteins that were not seen in poorly metastatic or

non-tumorigenic parental mouse fibroblast cell lines.

Clearly, this study shows a correlation between the nuclear

matrix protein profile and the metastatic potential of the

cell. Of potential importance is the demonstration that

nuclear matrix proteins can be detected in the serum and

urine of cancer patients, thus suggesting that the detection

of specific nuclear matrix proteins may be of value in

breast cancer diagnosis (Miller et al., 1992; Replogle-

Schwab et al., 1996; Carpinito et al., 1996).

We have identified informative breast cancer nuclear

matrix proteins (Samuel et al., 1997a). Typically we

prepare NM2-IF nuclear matrices from breast cancer cell

lines or breast tumours. To remove IFs from these

preparations we disrupt nuclear matrices and attached IFs

with urea (Fig . 1 ). The IFs are then allowed to

reassemble and are removed from the soluble nuclear

matrix proteins (Fey and Penman, 1988). Over a broad

protein concentration range, this process is independent of

protein concentration, but it is dependent upon temperature

(Fig . 2 ). Performing the reconstitution at room

temperature is recommended. About 8-10% of the nuclear

protein is recovered in the nuclear matrix protein fraction.

In the search for informative breast cancer nuclear

matrix proteins, we used human breast cancer cell lines

T47D, MCF-7 and ZR-75 (ER+/hormone dependent),

MDA MB231, and BT-20 (ER-/hormone independent), and

T5-PRF (ER+/hormone independent). A non-tumorigenic,

spontaneously immortalized human breast epithelial cell

line known as MCF-10A1 (ER-/hormone independent)

obtained from reduction mammoplasty was chosen as the

closest representative of normal breast epithelia. Typically

we isolate proteins from at least three nuclear matrix

preparations of each cell line, and these proteins are

electrophoretically separated on two dimensional gels.

Comparative analysis of the two dimensional gel patterns

identified nuclear matrix proteins of estrogen receptor (ER)

positive breast cancer cells that were not found in ER-

breast cancer cells or normal breast epithelial cells (Samuel

et al., 1997a). Our criteria for designating a nuclear matrix

protein as being informative in breast cancer was that the

protein had to be present in each of the relevant

preparations (either ER+ and/or ER- breast cancer cell

nuclear matrix proteins), but not in the preparations of

Gene Therapy and Molecular Biology, Vol 1, page 511

511

Figure 1 . Method to Isolate Nuclear Matrix and Nuclear

Matrix Proteins.

Figure 2 . Effect of Temperature on the Separation of

Intermediate Filament (IF) Proteins from Nuclear Matrix

Proteins (NMP). NM2-IF from human breast cancer cells was

disrupted in urea and then made to different protein

concentrations prior to removal of IF proteins as described

(Samuel et al., 1997b).

nuclear matrix proteins from "normal" breast epithelial

cells. Using the nomenclature proposed by Khanuja et al.

(1993), we refer to these proteins as NMBCs (nuclear

matrix proteins in breast cancer). Five NMBCs (1-5)

exclusive to the ER+ cell lines and one NMBC (6)

exclusive to the ER- cell lines were identified (Samuel et

al., 1997a).

The extracellular environment can alter the cellular

morphology as well as the protein composition of the

cytoskeletal and nuclear matrix compartments (Getzenberg

et al., 1991b; Pienta et al., 1991; Fallaux et al., 1996).

Thus, it was important to find out whether the changes in

nuclear matrix proteins we observed with cancer cells

grown on plastic were observed with cancer cells present in

a breast tumour. In the preparation of nuclear matrices

from breast tumours, we found that it was necessary to

remove the adipose tissue surrounding the tumour. We

found all NMBCs (1-5) exclusive to ER+ status in the

human breast cancer cell lines as being present in the ER+

breast tumours, while NMBC-6 was not detectable (see

Fig . 3 , tumour 12797). NMBC-6, but not NMBCs 1-5,

were present in ER- tumour nuclear matrix proteins.

The effect of cellular transformation on nuclear matrix

protein composition is illustrated in the following study.

Nuclear matrix proteins were isolated from MCF10A1

breast cancer cells that were transformed with the human

T-24 mutated Ha-ras oncogene (MCF10AneoT) or with

wild type human ER (cell line 139B6). MCF10AneoT

cells are transformed and show anchorage independent

growth (Basolo et al., 1991). The cell line 139B6

expresses ER at a similar level to that of MCF7 breast

cancer cells (Pilat et al., 1996). In the presence of

estradiol, this cell line has a slight inhibition in growth.

This is typical of results of studies in which the ER is

expressed in a ER- breast epithelial cell line or ER- breast

cancer cell line (Pilat et al., 1996; Lundholt et al., 1996).

Estradiol activated ER failed to elevate the expression of

endogenous estrogen responsive genes but did induce the

transient expression of an estrogen responsive element-

regulated reporter gene in the 139B6 cell line (Pilat et al.,

1996). Analysis of the two dimensional gel patterns of

the nuclear matrix proteins from these cell lines revealed

several alterations in nuclear matrix protein composition

when MCF10A1 cells were transformed with ras or

expressing ER. These differences were seen against a

pattern of proteins found in all cell lines, for example

hnRNP K (hk in F i g . 3 ). With the MCF10AneoT (Ha-

ras transformed) cells, nuclear matrix proteins with a

molecular mass of 47 kDa and pI range of 5.8-6.2

(constellation C in Fig. 3) were found to be exclusive to

this cell line. Similarly, nuclear matrix proteins with

molecular masses 50-57 kDa and pIs 5.5-5.7 (constellation

B in Fig . 3 ) and proteins with molecular masses of 30-36


512

Figure 3 . Human Breast Cancer Nuclear Matrix Proteins

Nuclear matrix proteins were isolated from MCF10A1 (parent, ER-, human breast epithelial cells), MCF10A-139B6 (parent

transfected with human wild type ER), MCF10AneoT (parent transformed by T-24 Ha-ras), and human breast tumour 12797 (ER+).

Protein (40 ug) was electrophoretically resolved on two-dimension gels. The gels were stained with silver. The position of the

molecular weight standards (in thousands) is shown on the left side of each gel pattern. LA and LC are lamin A and C, respectively.

The circles in MCF10A1 (parent) show the absence or decreased amount of nuclear matrix proteins highlighted in other gel

patterns.


513

Figure 4 . Sites of Post-Synthetic Modifications on the

Histones.

The structures of H2A-H2B dimers and (H3-H4)2 tetramers

and the sites of modification are shown. Ac, acetylation; Ub,

ubiquitination. The enzymes catalyzing reversible histone

acetylation are shown.

kDa and pIs 4.5 (constellation A in Fig . 3 ) were

determined to be exclusive to the ER expressing cell line

MCF10A-139B6. However, within constellation B, a 48

kDa (pI 5.5) nuclear matrix protein (denoted by * in Fig .

3 ) was observed in the MCF10A parent cell line as well as

in the ras- transformed and ER transfected cell lines.

Relative to the parent cell line, the level of this protein in

MCF10AneoT and MCF10A-139B6 was higher. NMBC1

present in the ER+ breast tumor nuclear matrix proteins

was also detected in the ER expressing cell line (F i g . 3 ).

The presence of NMBC1 in the ER transfected cell line

suggests that ER expression has a role in the association

of NMBC1 with the nuclear matrix. These results

illustrate how nuclear matrix protein profiles reflect

alterations in a cell's physiological state.

III. Nuclear matrix and organization ofnuclear DNA

Nuclear DNA is packaged into nucleosomes, the

repeating structural units in chromatin (Van Holde, 1988).

The nucleosome consists of an histone octamer core

around which DNA is wrapped. The four core histones of

the octamer are arranged as a (H3-H4)2 tetramer and two

H2A-H2B dimers positioned on both sides of the tetramer.

The core histones have a similar structure with a basic N

terminal unstructured domain, a globular domain organized

by the histone fold, and a C terminal unstructured tail

(Arents and Moudrianakis, 1995) (F i g . 4 ). Histone H1

binds to the linker DNA, which joins nucleosomes

together, and to core histones (Boulikas et al., 1980;

Banères et al., 1994). H1 has a tripartite structure with a

basic N terminal domain, a basic C terminal tail domain,

and a central globular core (Ramakrishnan, 1994).

In low ionic strength, chromatin fibers depleted of H1

have a "beads-on-a-string" structure, but with H1, folding

of the fiber is evident (Leuba et al., 1994). At

physiological ionic strength chromatin is folded into a 30

nm fiber. H1 stabilizes the folding of the chromatin fiber

(Shen et al., 1995). The native 30 nm chromatin fiber has

an irregular structure (an irregular three dimension zigzag)

in vitro (Woodcock and Horowitz, 1995). Woodcock and

colleagues show that the irregularities of the 30 nm

chromatin fiber can be accurately reflected in a model that

accounts for variability in linker DNA length and angle of

trajectory that the linker DNA has as it enters and leaves

the nucleosome. Thirty nm fibers are usually not seen

inside nuclei (Woodcock and Horowitz, 1995). The

chromatin is observed as matted patches. It appears that

neighboring zigzags interdigitate, preventing individual

chromatin fibers from being seen in nuclei. The core

histone tails contribute to the condensation of the

chromatin fiber (Garcia-Ramirez et al., 1995; Schwarz et

al., 1996; Krajewski and Ausió, 1996). H3 and H4 tails

are needed for fiber-fiber interactions (Schwarz et al.,

1996).

The chromatin fiber is organized into loop domains,

with an average size of 86 kb (Jackson et al., 1990; Gerdes

et al., 1994) (F i g . 5 ). Transcriptionally active genes are

found in DNAase I-sensitive, presumably decondensed

chromatin loops that are accessible to transcription factors

and transcription machinery (Davie, 1995).

Transcriptionally inactive genes are in higher order,

interdigitated chromatin patches, being essentially

invisible to transcription factors and the transcription

machinery. At the base of the loop there are DNA

sequences called MARs (matrix associated regions) that

bind to nuclear matrix proteins (Bode et al., 1995). MARs

tend to be AT-rich, but do not have a consensus sequence

(Bode et al., 1995; Mielke et al., 1996). MAR-DNA

binds to both internal matrix and nuclear pore-lamina,

suggesting that proteins of the nuclear pore-lamina and

internal matrix are involved in the organization of chroma-


514

Figure 5 . A Model for

Transcriptionally Active and

Repressed Chromatin Domains

At the base of the loop are

nuclear matrix associated

regions (MARs). HET (SAF-B)

is a nuclear matrix protein that

binds MARs. The repressed

chromatin loop has a condensed

chromatin structure. Multiple

dynamic attachment sites

between the transcriptionally

active domain and the internal

nuclear matrix are presented in

the box outline. Histone

acetyltransferase (HAT A),

histone deacetylase (HDAC),

transcription machinery and

transcription factors are shown

associated with the internal

nuclear matrix, mediating a

dynamic attachment between

transcriptionally active

chromatin and the nuclear

matrix. HAT A and HDAC are

shown as multiprotein

complexes.

tin (Zini et al., 1989). MAR-binding proteins include

lamins, which are found in the nuclear pore-lamina and

internal matrix (Hozák et al., 1995), topoisomerase II,

SATB1, HET (SAF-B), and attachment region binding

protein which is an internal matrix protein or nuclear

matrin (Pommier et al., 1990; Nakayasu and Berezney,

1991; von Kries et al., 1991; Luderus et al., 1992;

Nakagomi et al., 1994; Buhrmester et al., 1995;

Oesterreich et al., 1997).

Alterations in MAR-binding proteins have been

reported in cancer cells. In Southwestern blotting

experiments with a radiolabelled mouse IgH MAR

sequence, Yanagisawa et al. (1996) detected a 114-kDa

MAR binding protein expressed in breast carcinomas but

not normal or benign breast tissue. Further, the levels of

this MAR-binding protein were elevated in poorly

differentiated breast ductal carcinomas. A recent study

shows that mutant, but not wild type, p53 binds to MARs

(Müller et al., 1996). Changes in nuclear matrix, MAR-

binding proteins could result in reorganization of nuclear

DNA.

IV. In situ crosslinking with cisplatin

Recent studies suggest that cisplatin (cis-diammine

dichloroplatinum or cis-DDP) preferentially crosslink

MARs to nuclear matrix proteins in situ. Either cells or

nuclei can be incubated with cisplatin to crosslink protein

to DNA. Most proteins crosslinked to DNA with

cisplatin are nuclear matrix proteins, and the DNA

crosslinked to protein is enriched in MAR-DNA sequences

(Wedrychowski et al., 1986; 1989; Ferraro et al., 1992;

1995; Bubley et al., 1996; Olinski et al., 1987). F i g . 6

shows the protocol to isolate proteins crosslinked to DNA

in situ. A comparison of two dimension gel patterns of

nuclear matrix proteins and proteins crosslinked to DNA

with cisplatin in ZR-75 human breast cancer cells shows

that several abundant nuclear matrix proteins are

crosslinked to DNA in the cells (F i g . 7 ). Lamins A and

C, components of the nuclear pore-lamina, are crosslinked

in situ to nuclear DNA consistent with in vitro data

suggesting that these proteins are involved in the

organization of nuclear DNA (Wedrychowski et al., 1986;

1989). Abundant nuclear matrix proteins found


515

Figure 6 . Method to Isolate Proteins Crosslinked to

DNA in Cells or Nuclei with Cisplatin

crosslinked to nuclear DNA in situ with cisplatin are F-

actin and hnRNP K (Miller et al., 1991; Sauman and

Berry, 1994) (Fig . 7 ). HnRNP K is a single-strand

DNA-binding protein that is associated with the nuclear

matrix and has an important role in regulating the

expression of the c-myc gene (Michelotti et al., 1996;

Mattern et al., 1996). Further, hnRNP K interacts with

TATA-binding protein (Michelotti et al., 1996). This

transcription factor is a prominent protein observed in both

the nuclear matrix fraction and proteins crosslinked to

DNA in situ with cisplatin in ZR-75 human breast cancer

cells (Fig . 7 ).

The ability of cisplatin to preferentially crosslink

nuclear matrix proteins to nuclear DNA in situ has great

potential in identifying nuclear matrix proteins involved in

the organization and function of nuclear DNA. Several

transcription factors are nuclear matrix proteins thought to

interact with promoter and enhancer elements of specific

genes. It has been proposed that the interaction of nuclear

matrix bound transcription factors with regulatory DNA

sequences has a role in attaching transcriptionally active

chromatin to nuclear matrix (see below). Crosslinking

with cisplatin may provide a method to find if the nuclear

matrix associated transcription factor is bound to the DNA

sequence of interest in situ. We are currently developing

methods that will identify nuclear matrix associated

transcription factors and their bound DNA sequences.

Further, these methods are being used to find informative

DNA-binding nuclear matrix proteins in the diagnosis of

cancer.

Figure 7 . Analysis of Nuclear Matrix

Proteins and Proteins Crosslinked to

DNA

ZR-75 human breast cancer nuclear

matrix proteins (40 ug) and proteins

crosslinked to DNA by cis-DDP in situ

(40 ug) were electrophoretically

resolved on two dimension gels. The

gels were stained with silver. The

position of the carbamylated forms of

carbonic anhydrase is indicated by ca.

The position of the molecular weight

standards (in thousands) is shown to

the left of the gel patterns. LA and LC

show lamin A and C, respectively.

HnRNP K is shown as hk.


516

V. Transcription factors: activators oftranscription

Current evidence suggest that an interaction between an

enhancer or locus control region and promoter is an

essential step in forming the open chromatin domain

(Reitman et al., 1993). The enhancer/locus control region-

promoter interaction is mediated by protein-protein

associations between transcription factors bound to these

cis-acting regulatory elements. This complex recruits the

transcription initiation machinery and initiates the

transcription cycle. The transcription cycle can be

separated into at least four stages: initiation, promoter

clearance, elongation, and termination. During the

initiation stage, the pre-initiation complex (PIC) is formed

at the promoter of a RNA polymerase II transcribed gene.

The basal transcription factors TFIIA, TFIIB, TFIID,

TFIIE, TFIIF, and TFIIH and RNA polymerase II are

involved in the assembly of the PIC (for review see

Orphanides et al., 1996; Pugh, 1996; Nikolov and Burley,

1997). In vitro studies show that there is a defined order

by which the basal transcription factors are assembled into

the PIC. TFIID, one of the first factors involved, binds to

the TATA-box and consists of the TATA-binding protein

(TBP) and several TBP-associated factors (TAFs).

However, there is evidence that the PIC comes partially

preassembled (Maldonado et al., 1996). In the formation

of the PIC at any given promoter, TFIID binding appears

to be the rate limiting step. Transcription factors bound to

enhancers and upstream promoter elements interact through

their activation domains with the TAFs of TFIID or with

other components of the PIC (e.g., TFIIB), increasing the

rate that the PIC is formed at the promoter (for review see

Tjian and Maniatis, 1994; Chiang et al., 1996; Gupta et

al., 1996). TAFs are referred to as coactivators, proteins

that mediate an interaction between transcription factors

and PIC. Transcription factors bind to different TAFs, and

these multiple contacts between the transcription factors

and TAFs ensure the efficient recruitment of TFIID (Sauer

et al., 1995; Chi et al., 1995). Transcription factors with

multimerization domains also have key roles in

juxtapositioning enhancer and promoter elements. For

example, multimers of the transcription factor Sp1 bound

at DNA sites separated by 1.8 kb will interact, resulting in

a looping out of the intervening sequences (Pascal and

Tjian, 1991). Once these cis-acting elements are

positioned next to each other, there will be a high local

concentration of activators in the vicinity of the promoter

(Fig . 8 ).

Figure 8 . Model for Histone H5

Chromatin in Chicken Immature

Erythrocytes.

The 3' enhancer is positioned next to

the 5' promoter through protein-protein

interactions. NF1 and multiprotein

complexes (HAT As, HDACs,

transcription machinery) are shown

mediating the dynamic attachments of

the histone H5 gene to the nuclear matrix

at sites of transcription.


517

VI. Regulation of transcription

Within the DNAase I sensitive chromatin domains

containing transcriptionally active genes are regions that

are hypersensitive to DNAase I attack. The DNAase I

hypersensitive (DH) regions of chromatin may lack

nucleosomes and often mark chromatin for the presence of

cis-acting regulatory DNA sequences and trans-acting

factors. DH sites in human breast cancer c-myc chromatin

and chicken erythrocyte histone H5 chromatin map with

promoters and enhancers, and using in vitro assays we

identified the transcription factors binding to these

regulatory DNA elements (Penner and Davie, 1992; 1994;

Sun et al., 1992; 1993; Miller et al., 1993; 1996; Murphy

et al., 1996). However, in vitro assays can sometimes be

misleading, and the most rigorous method in finding

transcription factor occupancy is in situ footprinting

(Becker et al., 1987; Mueller and Wold, 1989). We did in

situ footprinting using a procedure called ligation-

mediated PCR to reveal the occupancy of factor binding

sites in the promoter and enhancer of the H5 gene in

chicken erythrocytes (Sun et al., 1996b). Some factor

binding sites in the promoter and enhancer identified in

vitro were not occupied in situ. Based upon our studies on

the chromatin structure and transcription factors associated

with the H5 promoter and enhancer, we put forth a model

for the transcriptionally active H5 gene (Fig . 8 ).

VII. Nuclear matrix and processing ofthe genetic information

The nuclear matrix is involved in the processing of the

genetic information. In recent years we have come to

appreciate that functional components (e.g., transcript

domains, RNA processing sites, sites of replication) of the

nucleus are highly organized (Hendzel and Bazett-Jones,

1995; Penman, 1995; Xing et al., 1995). Transcribed

genes are found in discrete foci (Jackson et al., 1993;

Iborra et al., 1996; Wansink et al., 1996). The nuclear

matrix is the foundation from which this organization is

built, providing a scaffold from which nuclear processes

such as DNA replication and transcription occur (Berezney,

1991; Iborra et al., 1996). It is important to note that

these functional centers in the nucleus are dynamic in their

formation and dissociation. For example, sites of

replication will assemble on the nuclear matrix at or near

transcription foci in early S phase of the cell cycle. Once

replication of these regions of the genome is complete, the

replication machinery will disassemble from its site on the

nuclear matrix and reassemble at other sites continuing

replication of other regions of the genome (Hozák et al.,

1993; Bassim Hassan et al., 1994). The process of

transcription occurs at the nuclear matrix, and it has been

proposed that the chromatin fiber moves through the

nuclear matrix bound transcription apparatus as

transcription proceeds (Cook, 1994; Hendzel and Bazett-

Jones, 1995; Iborra et al., 1996). The transcription

machinery is a massive multiprotein complex (Aso et al.,

1995; Chao et al., 1996; Maldonado et al., 1996). Thus,

it is unlikely that RNA polymerase travels along the DNA

as text book models often show. Further, the nascent

RNA becomes associated with the nuclear matrix. A solid

state process by which DNA is driven through the nuclear

matrix bound machinery and the nascent RNA is processed

at the nuclear matrix would be an efficient way of dealing

with these nuclear activities.

VIII. Nuclear matrix andtranscriptionally active chromatin

Transcribed and nontranscribed sequences are precisely

compartmentalized within the nucleus (Andreeva et al.,

1992; Gerdes et al., 1994; Davie, 1995). Actively

transcribed, but not inactive, chromatin regions are

immobilized on the nuclear matrix by multiple dynamic

attachment sites (F i g . 5 and 8 ). When histones are

removed by high salt, loops of DNA are seen emanating

from a central nuclear skeleton, forming a halo around this

nuclear structure. Transcriptionally inactive genes are

found in the halo, while DNA loops with transcriptionally

active genes remain associated with the nuclear skeleton

(Gerdes et al., 1994). The transcription machinery,

specific transcription factors, and nuclear enzymes (e.g.,

histone acetyltransferase, histone deacetylase, see F i g . 8 )

are thought to mediate the dynamic attachments between

transcribing chromatin and nuclear matrix (van Wijnen et

al., 1993; Cook, 1994; 1995; Bagchi et al., 1995;

Merriman et al., 1995).

The nuclear matrix is selective for which transcription

factors it binds, and this selectivity varies with cell type

(van Wijnen et al., 1993; Sun et al., 1994; 1996a). It has

been postulated that the nuclear matrix has a role in the

expression of genes by concentrating a subset of

transcription factors at specific nuclear sites (Stein et al.,

1991; Merriman et al., 1995). Transcription factors

associated with the nuclear matrix include ER, HET,

GATA-1, YY1, AML-1, Sp1, Oct1, mutant p53, and Rb

(Dworetzky et al., 1992; Isomura et al., 1992; Vassetzky

et al., 1993; van Wijnen et al., 1993; Sun et al., 1994;

Merriman et al., 1995; Guo et al., 1995; Müller et al.,

1996; Mancini et al., 1996; Kim et al., 1996; Oesterreich

et al., 1997). HET is a transcriptional repressor.

Interestingly, sequencing of HET revealed that it was

identical to SAF-B, a protein isolated by its ability to bind

MARs (Renz and Fackelmayer, 1996; Oesterreich et al.,

1997). Thus, HET (alias SAF-B) is a nuclear matrix

protein in breast cancer cells that binds to MARs and acts

as a repressor.


518

Protein domains involved in targeting transcription

factors to the nuclear matrix have been identified.

However, it is too early to know whether a consensus

nuclear matrix localization signal will emerge from these

studies. The N-terminal domains of the androgen and

glucocorticoid receptor are involved in directing these

receptors to the nuclear matrix (van Steensel et al., 1995).

There are examples of where the association of

transcription factors with the nuclear matrix is regulated by

modifications. For example, the association of Rb with

the nuclear matrix appears to be regulated by

phosphorylation and is cell-cycle dependent. When Rb is

in a hypophosphorylated state in G1-phase of the cell

cycle, it is attached to the nuclear matrix. But the highly

phosphorylated Rb of S-phase is not associated with the

matrix (Mancini et al., 1994). The amino terminus of

hypophosphorylated Rb binds to a p84 nuclear matrix

protein (Durfee et al., 1994).

ER is associated with the nuclear matrix of estrogen

responsive tissues (Metzger and Korach, 1990; Metzger et

al., 1991; Thorburn and Knowland, 1993). In vitro

reconstitution studies with nuclear matrices and hormone

receptors (e.g., ER and androgen receptor) show that

nuclear acceptor sites for the hormone receptors are

associated with the nuclear matrix (Barrack, 1987; Metzger

and Korach, 1990; Lauber et al., 1995). The binding of

the ER to the nuclear matrix was saturable, of high

affinity, target tissue specific, and receptor specific

(Metzger and Korach, 1990). Acceptor proteins for ER

have been identified in a variety of estrogen-responsive

tissues (Lauber et al., 1995; Ruh et al., 1996).

Transcription factors associated with the nuclear matrix

can change throughout development and differentiation

(Stein et al., 1994; Davie, 1995; Merriman et al., 1995;

Bagchi et al., 1995; Sun et al., 1996a). For example,

transcription factors associated with the chick erythrocyte

nuclear matrix change throughout development (Sun et al.,

1996a). Primitive red blood cells from 5-day old embryos

have high levels of nuclear matrix-bound transcription

factors, including GATA-1, CACCC-binding proteins, and

NF1; factors that have key roles in erythroid-specific gene

expression. In definitive red blood cells (11-day and 15-day

embryos) the levels of these nuclear matrix bound

transcription factors decline. Erythroid nuclear matrices

preferred to bind CACCC-binding proteins and not Sp1.

Promoters and enhancers of erythroid-specific genes have

Sp1 binding sites that bind both CACCC-binding proteins

and Sp1. It is possible that the selective nuclear matrix

binding of CACCC-binding proteins gives the CACCC-

binding proteins an advantage over Sp1 in binding to a

Sp1/CACCC site.

Although we know that transcription factors are

associated with the nuclear matrix, evidence that nuclear

matrix associated transcription factors are bound to

regulatory DNA elements of specific genes is currently

lacking. For example, NF1 is a nuclear matrix associated

transcription factor that binds to the enhancer of the

chicken histone H5 gene (F i g . 8 ). In vitro footprinting

and electrophoretic mobility shift assays show that NF1

isolated from immature erythrocyte nuclear matrices binds

to the H5 enhancer (Sun et al., 1994). In situ footprinting

shows that the NF1 binding site in the H5 enhancer is

occupied in chicken immature erythrocytes (Sun et al.,

1996b). We have proposed that NF1 recruits the H5

enhancer to the nuclear matrix (Davie, 1996). However,

we have yet to show that nuclear matrix associated NF1 is

the protein occupying the H5 enhancer NF1 binding site in

erythroid cells. Cisplatin crosslinking may provide direct

evidence to test this model (see above).

IX. Dynamic histone acetylation

Transcribed DNA is associated with acetylated histones

(Hebbes et al., 1994; O'Neill and Turner, 1995; Mutskov

et al., 1996). The core histones are reversibly modified by

acetylation of lysines located in their basic N terminal

domains (Fig . 4 ). Reversible histone acetylation is

catalyzed by histone acetyltransferases (HATs) and

deacetylases (HDACs), with the level of acetylation being

decided by the net activities of these two enzymes.

Histone acetylation alters nucleosome and higher order

chromatin structure (Davie, 1995; 1997). For example,

chromatin associated with highly acetylated histones does

not undergo histone H1 mediated aggregation at

physiological ionic strength, while chromatin with

unacetylated histones aggregates when associated with H1

(Ridsdale et al., 1990; Davie, 1997). Besides modulating

nucleosome and higher order chromatin packaging, the core

histone tails bind to regulatory proteins (Ma et al., 1996;

Edmondson et al., 1996). For example, yeast repressor

protein Tup1 binds to the tails of H3 and H4. Acetylation

of H3 and H4 prevents the binding of Tup1 (Edmondson et

al., 1996). In mammalian cells and chicken erythrocytes,

transcriptionally active chromatin regions have core

histones undergoing high rates of acetylation and

deacetylation, while in repressed chromatin regions the rate

of reversible acetylation is slow (Davie, 1996; 1997).

Thus, we expect that the interaction of regulatory proteins

with the histone tails and chromatin structure of

transcriptionally active regions of mammalian and chicken

erythrocytes is in dynamic flux.

The process of reversible histone acetylation is not

dependent upon ongoing transcription (Ruiz-Carrillo et al.,

1976). To date, the only histone modifications dependent

upon ongoing transcription are ubiquitination of H2B (see

F i g . 4 ) and phosphorylation of mouse H1b (Davie and

Murphy, 1990; Chadee et al., 1997). However,

interference of dynamic acetylation by inhibiting


519

deacetylation with histone deacetylase inhibitors (e.g.,

sodium butyrate, trichostatin A or trapoxin) greatly affects

cell cycle progression, arresting cells in G1 or G2, and

may enhance or repress the expression of genes (Yoshida et

al., 1995; Johnston et al., 1992; Girardot et al., 1994;

Miyashita et al., 1994; Laughlin et al., 1995).

Figure 9 . CBP/p300 Cointegrates Diverse Signalling

Pathways

CBP and its functional/protein interaction domains are

shown.

X. Histone acetyltransferase and geneactivation

Histone acetylation is not limited to transcriptionally

active chromatin, but also has a role in DNA replication

(deposition-related acetylation) and DNA repair (for review

see Davie, 1995; 1997). Deposition-related acetylation of

H4 is catalyzed by HAT B, a cytoplasmic enzyme (Kleff et

al., 1995; Brownell and Allis, 1996). HAT A is

responsible for transcriptionally active chromatin-

associated acetylation. Nuclear HAT A is bound to

chromatin and acetylates all core histones when free or

within nucleosomes (Brownell and Allis, 1996). Dr. Allis

and colleagues were the first to purify and clone a HAT A.

There studies showed that Tetrahymena HAT A (p55) is

homologous to yeast Gcn5, a transcriptional adaptor, that

has HAT activity (Brownell et al., 1996). This important

breakthrough provided a direct link between the process of

transcription activation and histone acetylation.

Tetrahymena HAT A (p55) and yeast Gcn5 are

components of large multiprotein complexes, and the

substrate specificity of the catalytic subunit is regulated by

the proteins binding to it (Grant et al., 1997). Yeast Gcn5

and Tetrahymena p55 can acetylate free histone H3 but

these HAT As are unable to acetylate histones in

nucleosomes. Yeast Gcn5 is a component of two high

molecular mass complexes (0.8 and 1.8 megadaltons)

(Grant et al., 1997). These high molecular mass,

multiprotein complexes acetylated histones in

nucleosomes and free histones. Both HAT A complexes

contain Ada2 and Ada3. Gcn5-Ada2-Ada3 is a putative

adaptor complex that connects DNA-bound transcription

factors (activators) to components of the PIC (Candau et

al., 1997). The HAT domain of yeast Gcn5 has been

localized. Gcn5 requires both the HAT domain of Gcn5

and interaction with Ada2 for transcriptional activation

(Candau et al., 1997). Human homologues of Gcn5 and

Ada2 have been identified (Candau et al., 1996).

Tetrahymena HAT A and yeast Gcn5 have a

bromodomain that is lacking in yeast Hat1p. The

bromodomain, which is thought to be a protein-protein

interaction domain, is found in the C-termini of these

proteins (Haynes et al., 1992). Several other recently

identified HAT As have the bromodomain, including

TAFII250 (a 250 kDa protein that binds to TATA-binding

protein), CBP/p300 but not P/CAF (Yang et al., 1996;

Bannister and Kouzarides, 1996; Mizzen et al., 1997;

Ogryzko et al., 1997). It is possible that HAT As interact

with other transcription factors through the bromodomain,

directing HAT A to specific regions in chromatin and in

the nucleus (Brownell and Allis, 1996).

Most HAT As are coactivators (e.g., Gcn5, TAFII250,

CBP/p300). CBP/p300 binds to hormone receptors, AP-

1, c-Myb, SV40 large T antigen, and adenovirus E1a, and

appears to be an integrator of multiple signalling pathways

(Kamei et al., 1996; Avantaggiati et al., 1996;

Oelgeschläger et al., 1996; Hanstein et al., 1996) (Fig .

9 ). Unlike Gcn5, Tetrahymena p55, TAFII250, or

P/CAF, CBP acetylates all four core histones in

nucleosomes (Ogryzko et al., 1997). The discovery that

several transcription modulators or coactivators have HAT

activity provides a mechanism by which chromatin

structure is altered in the vicinity of DNA-bound

transcription activators. A variety of transcription factors

including hormone receptors, CREB, and fos-jun will bind

directly or indirectly to CBP, recruiting a coactivator with

histone acetyltransferase activity (F i g . 1 0 ). The HAT

activity of CBP would then acetylate surrounding histones

in nucleosomes, leading to the destabilization of nucleo-


520

Figure 10 . Role of HAT As and

HDACs in Transcriptional

Activation and Repression

Top panel: Fos-Jun is shown

recruiting the coactivator CBP,

resulting in the acetylation of

nucleosomal histones. Bottom

panel: Mad-Max is shown

recruiting the corepressor HDAC

multiprotein complex, resulting

in the deacetylation of

nucleosomal histones

some and higher order chromatin structure. Such a

chromatin state is thought to be facilitate the binding of

other transcription factors and, in general, aid the

transcription process.

XI. Histone deacetylase and generepression

Histone deacetylases are nuclear enzymes that have been

isolated from a variety of sources. Our studies have

focused on the chicken erythrocyte histone deacetylase, an

enzyme associated with the nuclear matrix (Hendzel et al.,

1991). Chicken erythrocyte histone deacetylase is a

component of a multiprotein complex that has a molecular

mass in excess of 400 kDa (Li et al., 1996). The chicken

histone deacetylase complex extracted from nuclei

dissociates to a 66-kDa form in 1.6 M NaCl or when

applied to an ion-exchange column (e.g., Q-sepharose).

However, the high molecular mass histone deacetylase

complex extracted from chicken erythrocyte nuclear

matrices does not dissociate in 1.6 M NaCl, but this

HDAC complex did dissociate to a 66-kDa form when

applied to a Q-sepharose column (Li et al., 1996). These

observations suggest that the solubilized nuclear matrix

histone deacetylase is associated with proteins that

stabilize the complex from dissociation into the 66-kDa

form in a high concentration of salt.

The high molecular mass chicken erythrocyte histone

deacetylase complex deacetylates the four core histones in

chromatin, but has a preference for H2B (Li et al., 1996).

Dissociation of the multiprotein histone deacetylase

complex resulted in a change in substrate preference. The

66-kDa enzyme could not deacetylate histones in

chromatin and had a preference for free H3. The data

suggest that proteins important in regulating HDAC

activity were lost during enzyme purification (Li et al.,

1996).

Dr. Schreiber and colleagues were the first to clone a

mammalian histone deacetylase (HDAC1, 55 kDa) (Tauton

et al., 1996). They found that mammalian histone

deacetylase was related to the yeast transcription regulator

Rpd3p, providing a link between transcription regulation

and histone deacetylation. At around the same time, Dr.

Grunstein and colleagues purified two yeast histone

deacetylase complexes, HDA (350 kDa) and HDB (600

kDa) (Carmen et al., 1996; Rundlett et al., 1996). The

HDA complex consists of multiple peptides with

molecular masses of approximately 70 kDa. Two peptides

from the HDA complex have been sequenced, and yeast

genes HDA1 (codes for p75) and HDA3 (codes for p71)

isolated. HDA1 shares sequence similarity with Rpd3p, a

yeast histone deacetylase. Gene disruptions of HDA1 or

HDA3 resulted in the loss of the HDA, but not, HDB

complex. Rpd3p is a component of HDB (Rundlett et al.,

1996). Rpd3p binds to Sin3 which in turn associates with


521

Ume6, a DNA-binding protein required for the repression

of several genes including those involved in meiosis

(Kadosh and Struhl, 1997). These observations suggest

that the yeast HDB complex consists of Rpd3p, Sin3 and

Ume6.

Mammalian histone deacetylase HDAC1 is related to

yeast transcriptional regulator Rpd3p (Tauton et al., 1996).

Although HDAC1 has a reported molecular mass of 55

kDa, we have found that it migrates on our SDS

polyacrylamide gels with an apparent molecular mass of

66 kDa. The mammalian homologue of Rpd3p, named

HDAC2, has been cloned (Yang et al., 1996). Chicken

erythrocyte histone deacetylase has been purified, and the

enzyme migrates as a 66-kDa band on SDS gels (J.-M.

Sun, H. Y. Chen, J. R. Davie, unpublished observations).

Thus, chicken erythrocyte histone deacetylase has a

molecular mass similar to that of mammalian HDAC1.

Figure 1 1 . Regions of mSin3A Involved in Protein

Interactions

N-CoR, SMRT and Mad family members interact with

different paired amphipathic helix (PAH) domains in mSin3A.

The HID domain which binds to HDAC 1 and 2 is shown.

As with yeast and chicken histone deacetylases,

mammalian histone deacetylases, HDAC1 and HDAC2,

exist as high molecular mass, multiprotein complexes

(Hassig et al., 1997). HDAC1 and HDAC2 bind to a

variety of proteins, including RbAp48, YY1, and

mammalian (m) Sin3A and mSin3B (Tauton et al., 1996;

Yang et al., 1996; Laherty et al., 1997). These HDAC-

binding proteins may exist in different HDAC

multiprotein complexes. For example, HDAC complexes

with mSin3 do not contain YY1 (Zhang et al., 1997).

HDAC1 was purified as a complex with RbAp48, a 50

kDa Rb-binding protein that binds to the C-terminus of

unphosphorylated or hypophosphorylated Rb (Tauton et

al., 1996; Qian et al., 1993). RbAp48 has several partners

in addition to HDAC1. RbAp48 is a component of

human and Drosophila CAF-1 (chromatin assembly factor

1) (Verreault et al., 1996). A yeast protein similar to

RbAp48, Hat2p, is component of yeast HAT B (Roth and

Allis, 1996; Parthun et al., 1996).

HDAC1 and/or HDAC2 are in large multiprotein

complexes that contain mSin3, N-CoR, and SMRT,

proteins that are corepressors (Nagy et al., 1997; Hassig et

al., 1997; Laherty et al., 1997; Heinzel et al., 1997).

Mammalian Sin3A and mSin3B have four paired

amphipathic helix (PAH) domains thought to be involved

in protein-protein interactions (F i g . 1 1 ). HDAC1 and

HDAC2 bind to the region between PAH3 and PAH4,

referred to as HID [the h istone deacetylase i nteraction

d omain (HID)] (Laherty et al., 1997). The HID region is

conserved in mSin3A, mSin3B and yeast Sin3.

Mammalian Sin3A (150 kDa) interacts with many other

proteins, including SAP18 (m S in3 a ssociated p rotein),

Mad family members (Mad1, Mad3, Mad4, Mxi1) and

Max-binding repressor Mnt, SMRT, and N-CoR (Laherty

et al., 1997; Zhang et al., 1997; Nagy et al., 1997;

Heinzel et al., 1997; Alland et al., 1997). mSin3A does

not have DNA binding ability; however, several of the

proteins associated with mSin3A can direct it to specific

DNA regulatory regions. The N-terminal region (SID,

m S in3 i nteraction d omain) of the Mad family members

and Mnt binds to PAH2 of mSin3 (Fig . 11 ). Mad

family members form a dimer with Max, a DNA-binding

complex that binds to E-box related DNA sequences

(Laherty et al., 1997). Max and Mad proteins are members

of the basic region-helix-loop-helix-leucine zipper (bHLH-

Zip) transcription factors. Myc forms a heterodimer with

Max which binds to the same E-box-related DNA

sequences as does Mad-Max heterodimers. However, Myc-

Max activates genes, while Mad-Max represses their

transcription. The repressive action of Mad-Max is

mediated in part by the interaction of Mad with mSin3

which in turn is associated with HDAC1 and/or HDAC2.

N-CoR and SMRT bind to unliganded retinoid and thyroid

hormone receptors (Nagy et al., 1997; Heinzel et al.,

1997). Thus, like the bHLH-Zip repressor proteins,

unliganded hormone receptors recruit the HDAC

multiprotein complex. HDAC has a principal role in

transcription repression. Several studies show that

tethering HDAC1 or 2 to a promoter by fusing HDAC to

a DNA-binding domain (e.g., Gal4 DNA-binding domain)

results in transcription inhibition (Yang et al., 1996;

Zhang et al., 1997; Kadosh and Struhl, 1997; Nagy et al.,

1997). These studies suggest that repressors recruit

histone deacetylase which would deacetylate histones in

nucleosomes, leading to the condensation of chromatin

(Wolffe, 1997) (Fig . 10 ).


522

Although these studies show that HDAC is involved

in repression, HDAC is associated with transcriptionally

active chromatin. Both HAT As and HDACs are needed to

catalyze dynamic acetylation of histones associated with

transcribed chromatin domains. The presence of both HAT

As and HDACs at transcriptionally active regions allows

the rapid manipulation of nucleosome and chromatin

structure (Wade and Wolffe, 1997).

XII. Histone acetylation and nuclearmatrix

Vertebrate histone acetyltransferase (HAT A) and

histone deacetylase (HDAC) are associated with the nuclear

matrix (Hendzel et al., 1991; 1992; 1994; Li et al., 1996).

Nuclear skeletons from chicken immature erythrocytes

retain 80% of the nuclear HAT A and HDAC activities,

and these enzymes catalyze reversible acetylation using as

substrate the chromatin fragments associated with the

nuclear skeletons (Hendzel et al., 1994). These studies

suggest that HAT A and HDAC are colocalized to specific

sites on the nuclear matrix. However, there is no evidence

that HAT A and HDAC are part of the same large

complex. We proposed a model in which nuclear matrix-

bound HAT A and HDAC mediate dynamic interactions

between the nuclear matrix and transcriptionally active

chromatin (Fig . 8 ) (Davie and Hendzel, 1994; Davie,

1995). We have evidence that HDAC1 is associated with

the matrix, but the identity of the nuclear matrix bound

HAT A is currently unknown. Several transcription

factors binding directly or indirectly with HAT A and

HDAC are nuclear matrix proteins. For example, YY1 is

a nuclear matrix protein that binds to HDACs (Yang et al.,

1996; Guo et al., 1995). Estrogen receptors bound to the

nuclear matrix could recruit CBP, an HAT A (Hanstein et

al., 1996). Determining how HATs and HDACs are

recruited to nuclear matrix sites engaged in transcription

will be an important challenge.

Acknowledgements

This research was supported by grants (MT-9186, PG-

12809) from the Medical Research Council of Canada and

the Cancer Research Society, Inc., and by a Medical

Research Council of Canada Senior Scientist Award to

J.R. Davie.

References

Alland L, Muhle R, Hou Jr. H, Potes J, Chin L, Schreiber-Agus

N, and DePinho RA ( 1 9 9 7 ) Role of N-CoR and histone

deacetylase in Sin3-mediated transcriptional repression.

Nature 387, 49-55.

Andreeva M, Markova D, Loidl P, and Djondjurov L ( 1 9 9 2 )

Intranuclear compartmentalization of transcribed and

nontranscribed c-myc sequences in Namalva-S cells. Eur.

J . B iochem. 207, 887-894.

Arents G and Moudrianakis EN ( 1 9 9 5 ) The histone fold: A

ubiquitous architectural motif utilized in DNA compaction

and protein dimerization. Proc. Nat l . Acad. S c i .

USA 92, 11170-11174.

Aso T, Conaway JW, and Conaway RC ( 1 9 9 5 ) The RNA

polymerase II elongation complex. FASEB J. 9, 1419-

1428.

Avantaggiati ML, Carbone M, Graessmann A, Nakatani Y,

Howard B, and Levine AS ( 1 9 9 6 ) The SV40 large T

antigen and adenovirus E1a oncoproteins interact with

distinct isoforms of the transcriptional co-activator,

p300. EMBO J. 15, 2236-2248.

Bagchi M, Van Wijnen A, Katar M, Merriman H, Lian J, Stein

J, Stein G, and Maisel H ( 1 9 9 5 ) Sequence-specific DNA

binding activities of nuclear matrix proteins of

mammalian lens epithelial cells. J . C e l l . B i o c h e m .

58, 1-5.

Banères J-L, Essalouh L, Jariel-Encontre I, Mesnier D, Garrod

S, and Parello J ( 1 9 9 4 ) Evidence indicating proximity in

the nucleosome between the histone H4 N termini and the

globular domain of histone H1. J . M o l . B i o l . 243, 48-

59.

Bannister AJ and Kouzarides T ( 1 9 9 6 ) The CBP co-activator

is a histone acetyltransferase. Nature 384, 641-643.

Barrack ER ( 1 9 8 7 ) Steroid hormone localization in the

nuclear matrix: interaction with acceptor sites. J .

Steroid Biochem. 27, 115-121.

Basolo F, Elliott J, Tait L, Chen XQ, Maloney T, Russo IH,

Pauley R, Momiki S, Caamano J, Klein-Szanto AJP,

Koszalka M, and Russo J ( 1 9 9 1 ) Transformation of

human breast epithelial cells by c-Ha-ras oncogene. M o l

Carcinogen 4, 25-35.

Bassim Hassan A, Errington RJ, White NS, Jackson DA, and

Cook PR ( 1 9 9 4 ) Replication and transcription sites are

colocalized in human cells. J . Ce l l Sc i . 107, 425-434.

Becker RB, Ruppert S, and Schultz G ( 1 9 8 7 ) Genomic

footprinting reveals cell type-specific DNA binding of

ubiquitous factors. Cel l 51, 435-443.

Berezney R ( 1 9 9 1 ) The nuclear matrix: a heuristic model for

investigating genomic organization and function in the

cell nucleus. J . Ce l l B iochem. 47, 109-123.

Bode J, Schlake T, Ríos-Ramírez M, Mielke C, Stengert M,

Kay V, and Klehr-Wirth D ( 1 9 9 5 ) Scaffold matrix-

attached regions: Structural properties creating

transcriptionally active loci. I n t . R e v . C y t o l . 162A,

389-454.

Boulikas T, Wiseman JM, and Garrard WT ( 1 9 8 0 ) Points of

contact between histone H1 and the histone octamer.

Proc. Nat l . Acad. Sc i . U. S . A. 77, 127-131.


523

Brownell JE, Zhou JX, Ranalli T, Kobayashi R, Edmondson

DG, Roth SY, and Allis CD ( 1 9 9 6 ) Tetrahymena histone

acetyltransferase A: A homolog to yeast Gcn5p linking

histone acetylation to gene activation. Cel l 84, 843-

851.

Brownell JE and Allis CD ( 1 9 9 6 ) Special HATs for special

occasions: Linking histone acetylation to chromatin

assembly and gene activation. Curr. Opin . Genet .

Dev . 6, 176-184.

Bubley GJ, Xu J, Kupiec N, Sanders D, Foss F, O'Brien M, Emi

Y, Teicher BA, and Patierno SR ( 1 9 9 6 ) Effect of DNA

conformation on cisplatin adduct formation. B i o c h e m .

Pharmacol. 51, 717-721.

Buhrmester H, von Kries JP, and Strätling WH ( 1 9 9 5 )

Nuclear matrix protein ARBP recognizes a novel DNA

sequence motif with high affinity. Biochemis try 34,

4108-4117.

Candau R, Moore PA, Wand L, Barlev N, Ying CY, Rosen CA,

and Berger SL ( 1 9 9 6 ) Identification of human proteins

functionally conserved with the yeast putative adaptors

ADA2 and GCN5. M o l . C e l l . B i o l . 16, 19337-19344.

Candau R, Zhou JX, Allis CD, and Berger SL ( 1 9 9 7 ) Histone

acetyltransferase activity and interaction with ADA2 are

critical for GCN5 function in vivo. EMBO J. 16, 555-

565.

Carmen AA, Rundlett SE, and Grunstein M ( 1 9 9 6 ) HDA1 and

HDA3 are components of a yeast histone deacetylase

(HDA) complex. J . B io l . Chem. 271, 15837-15844.

Carpinito GA, Stadler WM, Briggman JV, Chodak GW,

Church PA, Lamm DL, Lange PH, Messing EM, Pasciak

RM, Reservitz GB, Ross RN, Rukstalis DB, Sarosdy MF,

Soloway MS, Thiel RP, Vogelzang N, and Hayden CL

( 1 9 9 6 ) Urinary nuclear matrix protein as a marker for

transitional cell carcinoma of the urinary tract. J . U r o l .

156, 1280-1285.

Chadee DN, Allis CD, Wright JA, and Davie JR ( 1 9 9 7 )

Histone H1b phosphorylation is dependent upon ongoing

transcription and replication in normal and ras-

transformed mouse fibroblasts. J . B i o l . C h e m . 272,

8113-8116.

Chao DM, Gadbois EL, Murray PJ, Anderson SF, Sonu MS,

Parvin JD, and Young RA ( 1 9 9 6 ) A mammalian SRB

protein associated with an RNA polymerase II

holoenzyme. Nature 380, 82-85.

Chen HY, Sun J-M, Hendzel MJ, Rattner JB, and Davie JR

( 1 9 9 6 ) Changes in the nuclear matrix of chicken

erythrocytes that accompany maturation. B i o c h e m . J .

320, 257-265.

Chi TH, Lieberman P, Ellwood K, and Carey M ( 1 9 9 5 ) A

general mechanism for transcriptional synergy by

eukaryotic activators. Nature 377, 254-257.

Chiang YC, Komarnitsky P, Chase D, and Denis CL ( 1 9 9 6 )

ADR1 activation domains contact the histone

acetyltransferase GCN5 and the core transcriptional factor

TFIIB. J . B io l . Chem. 271, 32359-32365.

Cook PR ( 1 9 9 4 ) RNA polymerase: structural determinant of

the chromatin loop and the chromosome. B i o E s s a y s

16:6, 425-430.

Cook PR ( 1 9 9 5 ) A chromomeric model for nuclear and

chromosome structure. J . Ce l l Sc i . 108, 2927-2935.

Cupo JF ( 1 9 9 1 ) Electrophoretic analysis of nuclear matrix

proteins and the potential clinical applications. J .

Chromatogr. 569, 389-406.

Davie JR ( 1 9 9 5 ) The nuclear matrix and the regulation of

chromatin organization and function. I n t . R e v . C y t o l .

162A, 191-250.

Davie JR ( 1 9 9 6 ) Histone modifications, chromatin

structure, and the nuclear matrix. J . C e l l . B i o c h e m .

62, 149-157.

Davie JR ( 1 9 9 7 ) Nuclear matrix, dynamic histone

acetylation and transcriptionally active chromatin. M o l .

B i o l . R e p . 24, 197-207.

Davie JR and Hendzel MJ ( 1 9 9 4 ) Multiple functions of

dynamic acetylation. J . Ce l l . B iochem. 55, 98-105.

Davie JR and Murphy LC ( 1 9 9 0 ) Level of ubiquitinated

histone H2B in chromatin is coupled to ongoing

transcription. Biochemis try 29, 4752-4757.

Donat TL, Sakr W, Lehr JE, and Pienta KJ ( 1 9 9 6 ) Unique

nuclear matrix protein alterations in head and neck

squamous cell carcinomas: Intermediate biomarker

candidates. Otolaryngo l . Head Neck Surg. 114,

387-393.

Durfee T, Mancini MA, Jones D, Elledge SJ, and Lee W-H

( 1 9 9 4 ) The amino-terminal region of the retinoblastoma

gene product binds a novel nuclear matrix protein that co-

localizes to centers for RNA processing. J . C e l l B i o l .

127, 609-622.

Dworetzky SI, Fey EG, Penman S, Lian JB, Stein JL, and Stein

GS ( 1 9 9 0 ) Progressive changes in the protein

composition of the nuclear matrix during rat osteoblast

differentiation. Proc . Nat l . Acad . Sc i . U . S . A . 87,

4605-4609.

Dworetzky SI, Wright KL, Fey EG, Penman S, Lian JB, Stein

JL, and Stein GS ( 1 9 9 2 ) Sequence-specific DNA-binding

proteins are components of a nuclear matrix-attachment

site. P r o c . N a t l . A c a d . S c i . U . S . A . 89, 4178-

4182.

Edmondson DG, Smith MM, and Roth SY ( 1 9 9 6 ) Repression

domain of the yeast global repressor Tup1 interacts

directly with histones H3 and H4. Genes Dev . 10,

1247-1259.

Fallaux FJ, Hoeben RC, Cramer SJ, Van den Wollenberg DJM,

Briët E, Van Ormondt H, and Van der Eb AJ ( 1 9 9 6 ) The

human clotting factor VIII cDNA contains an

autonomously replicating sequence consensus- and matrix

attachment region-like sequence that binds a nuclear

factor, represses heterologous gene expression, and

mediates the transcriptional effects of sodium butyrate.

M o l . C e l l . B i o l . 16, 4264-4272.


524

Ferraro A, Grandi P, Eufemi M, Altieri F, and Turano C

( 1 9 9 2 ) Crosslinking of nuclear proteins to DNA by cis-

diamminedichloroplatinum in intact cells. Involvement of

nuclear matrix proteins. FEBS Lett. 307, 383-385.

Ferraro A, Eufemi M, Cervoni L, Altieri F, and Turano C

( 1 9 9 5 ) DNA-nuclear matrix interactions analyzed by

crosslinking reactions in intact nuclei from liver. Acta

Biochem. Po lon ica 42, 145-152.

Fey EG and Penman S ( 1 9 8 8 ) Nuclear matrix proteins reflect

cell type of origin in cultured human cells. Proc . Nat l .

Acad. Sc i . U. S . A. 85, 121-125.

Garcia-Ramirez M, Rocchini C, and Ausio J ( 1 9 9 5 )

Modulation of chromatin folding by histone acetylation.

J . B io l . Chem. 270, 17923-17928.

Gerdes MG, Carter KC, Moen PT, Jr., and Lawrence JB

( 1 9 9 4 ) Dynamic changes in the higher-level chromatin

organization of specific sequences revealed by in situ

hybridization to nuclear halos. J . C e l l B i o l . 126, 289-

304.

Getzenberg RH, Pienta KJ, Huang EY, and Coffey DS

( 1 9 9 1 a ) Identification of nuclear matrix proteins in the

cancer and normal rat prostate. Cancer Res . 51, 6514-

6520.

Getzenberg RH, Pienta KJ, Huang EY, Murphy BC, and Coffey

DS ( 1 9 9 1 b ) Modifications of the intermediate filament

and nuclear matrix networks by the extracellular matrix.

Biochem. B i o p h y s . Res . Commun. 179, 340-

344.

Getzenberg RH, Konety BR, Oeler TA, Quigley MM, Hakam

A, Becich MJ, and Bahnson RR ( 1 9 9 6 ) Bladder cancer-

associated nuclear matrix proteins. Cancer Res . 56,

1690-1694.

Girardot V, Rabilloud T, Yoshida M, Beppu T, Lawrence J-J,

and Khochbin S ( 1 9 9 4 ) Relationship between core

histone acetylation and histone H10 gene activity. Eur.

J . B iochem. 224, 885-892.

Grant PA, Duggan L, Côté J, Roberts SM, Brownell JE,

Candau R, Ohba R, Owen-Hughes T, Allis CD, Winston F,

Berger SL, and Workman JL ( 1 9 9 7 ) Yeast Gcn5

functions in two multisubunit complexes to acetylate

nucleosomal histones: characterization of an Ada complex

and the SAGA (Spt/Ada) complex. Genes Dev . 11,

1640-1650.

Guo B, Odgren PR, van Wijnen AJ, Last TJ, Nickerson J,

Penman S, Lian JB, Stein JL, and Stein GS ( 1 9 9 5 ) The

nuclear matrix protein NMP-1 is the transcription factor

YY1. Proc. Nat l . Acad. S c i . USA 92, 10526-

10530.

Gupta R, Emili A, Pan GH, Xiao H, Shales M, Greenblatt J,

and Ingles CJ ( 1 9 9 6 ) Characterization of the interaction

between the acidic activation domain of VP16 and the

RNA polymerase II initiation factor TFIIB. Nucleic

Acids Res . 24, 2324-2330.

Halle J-P and Meisterernst M ( 1 9 9 6 ) Gene expression:

increasing evidence for a transcriptosome. Trends

Genet . 12, 161-163.

Hanstein B, Eckner R, DiRenzo J, Halachmi S, Liu H, Searcy

B, Kurokawa R, and Brown M ( 1 9 9 6 ) p300 is a

component of an estrogen receptor coactivator complex.

Proc. Natl . Acad. Sci . USA 93, 11540-11545.

Hassig CA, Fleischer TC, Billin AN, Schreiber SL, and Ayer

DE ( 1 9 9 7 ) Histone deacetylase activity is required for

full transcriptional repression by mSin3A. Cel l 89, 341-

347.

Haynes SR, Dollard C, Winston F, Beck S, Trowsdale J, and

Dawid IB ( 1 9 9 2 ) The bromodomain: a conserved

sequence found in human, Drosophila and yeast proteins.

Nucleic Acids Res . 20, 2603

Hebbes TR, Clayton AL, Thorne AW, and Crane-Robinson C

( 1 9 9 4 ) Core histone hyperacetylation co-maps with

generalized DNase I sensitivity in the chicken -globin

chromosomal domain. EMBO J. 13, 1823-1830.

Heinzel T, Lavinsky RM, Mullen T-M, Soderstrom M, Laherty

CD, Torchia J, Yang W-M, Brard G, Ngo SD, Davie JR,

Seto E, Eisenman RN, Rose DW, Glass CK, and Rosenfeld

MG ( 1 9 9 7 ) A complex containing N-CoR, mSin3A and

histone deacetylase mediates transcriptional repression.

Nature 387, 43-48.

Hendzel MJ, Delcuve GP, and Davie JR ( 1 9 9 1 ) Histone

deacetylase is a component of the internal nuclear matrix.

J . B io l . Chem. 266, 21936-21942.

Hendzel MJ, Sun J-M, Chen HY, Rattner JB, and Davie JR

( 1 9 9 4 ) Histone acetyltransferase is associated with the

nuclear matrix. J . B io l . Chem. 269, 22894-22901.

Hendzel MJ and Bazett-Jones DP ( 1 9 9 5 ) RNA polymerase II

transcription and the functional organization of the

mammalian cell nucleus. Chromosoma 103, 509-516.

Hendzel MJ and Davie JR ( 1 9 9 2 ) Nuclear distribution of

histone deacetylase: a marker enzyme for the internal

nuclear matrix. B i o c h i m . B i o p h y s . A c t a 1130, 307-

313.

Hozák P, Hassan AB, Jackson DA, and Cook PR ( 1 9 9 3 )

Visualization of replication factories attached to a

nucleoskeleton. Cel l 73, 361-373.

Hozák P, Sasseville AM-J, Raymond Y, and Cook PR ( 1 9 9 5 )

Lamin proteins form an internal nucleoskeleton as well as

a peripheral lamina in human cells. J . C e l l S c i . 108,

635-644.

Iborra FJ, Pombo A, Jackson DA, and Cook PR ( 1 9 9 6 )

Active RNA polymerases are localized within discrete

transcription 'factories' in human nuclei. J . C e l l S c i .

109, 1427-1436.

Isomura T, Tamiya Koizumi K, Suzuki M, Yoshida S,

Taniguchi M, Matsuyama M, Ishigaki T, Sakuma S, and

Takahashi M ( 1 9 9 2 ) RFP is a DNA binding protein

associated with the nuclear matrix. Nucle ic Ac ids Res .

20, 5305-5310.


525

Jackson DA, Dickinson P, and Cook PR ( 1 9 9 0 ) The size of

chromatin loops in HeLa cells. EMBO J. 9, 567-571.

Jackson DA, Hassan AB, Errington RJ, and Cook PR ( 1 9 9 3 )

Visualization of focal sites of transcription within human

nuclei. EMBO J. 12, 1059-1065.

Jackson DA and Cook PR ( 1 9 8 8 ) Visualization of a

filamentous nucleoskeleton with a 23 nm axial repeat.

EMBO J. 7, 3667-3677.

Johnston LA, Tapscott SJ, and Eisen H ( 1 9 9 2 ) Sodium

butyrate inhibits myogenesis by interfering with the

transcriptional activation function of MyoD and

myogenin. M o l . C e l l B i o l . 12, 5123-5130.

Kadosh D and Struhl K ( 1 9 9 7 ) Repression by Ume6 involves

recruitment of a complex containing Sin3 corepressor and

Rpd3 histone deacetylase to target promoters. Cel l 89,

365-371.

Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B,

Lin SC, Heyman RA, Rose DW, Glass CK, and Rosenfeld

MG ( 1 9 9 6 ) A CBP integrator complex mediates

transcriptional activation and AP-1 inhibition by nuclear

receptors. Cel l 85, 403-414.

Keesee SK, Meneghini MD, Szaro RP, and Wu Y-J ( 1 9 9 4 )

Nuclear matrix proteins in human colon cancer. Proc.

Natl . Acad. Sci . USA 91, 1913-1916.

Khanuja PS, Lehr JE, Soule HD, Gehani SK, Noto AC,

Choudhury S, Chen R, and Pienta KJ ( 1 9 9 3 ) Nuclear

matrix proteins in normal and breast cancer cells. Cancer

Res . 53, 3394-3398.

Kim MK, Lesoon-Wood LA, Weintraub BD, and Chung JH

( 1 9 9 6 ) A soluble transcription factor, Oct-1, is also

found in the insoluble nuclear matrix and possesses

silencing activity in its alanine-rich domain. M o l .

C e l l . B i o l . 16, 4366-4377.

Kleff S, Andrulis ED, Anderson CW, and Sternglanz R ( 1 9 9 5 )

Identification of a gene encoding a yeast histone H4

acetyltransferase. J . B io l . Chem. 270, 24674-24677.

Krajewski WA and Ausió J ( 1 9 9 6 ) Modulation of the higher-

order folding of chromatin by deletion of histone H3 and

H4 terminal domains. Biochem. J . 316, 395-400.

Laherty CD, Yang W-M, Sun J-M, Davie JR, Seto E, and

Eisenman RN ( 1 9 9 7 ) Histone deacetylases associated

with the mSin3 corepressor mediate Mad transcriptional

repression. Cel l 89, 349-356.

Lauber AH, Sandhu NP, Schuchard M, Subramaniam M, and

Spelsberg TC ( 1 9 9 5 ) Nuclear matrix acceptor binding

sites for steroid hormone receptors: A candidate nuclear

matrix acceptor protein. I n t . R e v . C y t o l . 162B, 337-

376.

Laughlin MA, Chang GY, Oakes JW, Gonzalez-Scarano F, and

Pomerantz RJ ( 1 9 9 5 ) Sodium butyrate stimulation of

HIV-1 gene expression: A novel mechanism of induction

independent of NF-kappaB. J. Acquir. Immune

Defic . Syndr. Hum. Retrovirol . 9, 332-339.

Leuba SH, Yang G, Robert C, Van Holde KE, Zlatanova J, and

Bustamante C ( 1 9 9 4 ) Three-dimensional structure of

extended chromatin fibers as revealed by tapping-mode

scanning force microscopy. P r o c . N a t l . A c a d . S c i .

U . S . A . 91, 11621-11625.

Li W, Chen HY, and Davie JR ( 1 9 9 6 ) Properties of chicken

erythrocyte histone deacetylase associated with the

nuclear matrix. Biochem. J . 314, 631-637.

Luderus ME, de Graaf A, Mattia E, den Blaauwen JL, Grande

MA, de Jong L, and van Driel R ( 1 9 9 2 ) Binding of

matrix attachment regions to lamin B1. Cel l 70, 949-

959.

Lundholt BK, Madsen MW, Lykkesfeldt AE, Petersen OW, and

Briand P ( 1 9 9 6 ) Characterization of a nontumorigenic

human breast epithelial cell line stably transfected with

the human estrogen receptor (ER) cDNA. M o l . C e l l .

Endocrinol . 119, 47-59.

Ma XJ, Lu Q, and Grunstein M ( 1 9 9 6 ) A search for proteins

that interact genetically with histone H3 and H4 amino

termini uncovers novel regulators of the Swe1 kinase in

Saccharomyces cerevisiae. Genes Dev. 10, 1327-1340.

Maldonado E, Shiekhattar R, Sheldon M, Cho H, Drapkin R,

Rickert P, Lees E, Anderson CW, Linn S, and Reinberg D

( 1 9 9 6 ) A human RNA polymerase II complex associated

with SRB and DNA-repair proteins. Nature 381, 86-89.

Mancini MA, Shan B, Nickerson JA, Penman S, and Lee W-H

( 1 9 9 4 ) The retinoblastoma gene product is a cell cycle-

dependent, nuclear matrix-associated protein. Proc.

Natl . Acad. Sci . USA 91, 418-422.

Mancini MA, He D, Ouspenski II, and Brinkley BR ( 1 9 9 6 )

Dynamic continuity of nuclear and mitotic matrix proteins

in the cell cycle. J . Ce l l . B iochem. 62, 158-164.

Mattern KA, Humbel BM, Muijsers AO, de Jong L, and van

Driel R ( 1 9 9 6 ) HnRNP proteins and B23 are the major

proteins of the internal nuclear matrix of HeLa S3 cells. J .

C e l l . B i o c h e m . 62, 275-289.

Merriman HL, van Wijnen AJ, Hiebert S, Bidwell JP, Fey E,

Lian J, Stein J, and Stein GS ( 1 9 9 5 ) The tissue-specific

nuclear matrix protein, NMP-2, is a member of the

AML/CBF/PEBP2/Runt domain transcription factor

family: Interactions with the osteocalcin gene promoter.

Biochemis try 34, 13125-13132.

Metzger DA, Curtis S, and Korach KS ( 1 9 9 1 )

Diethylstilbestrol metabolites and analogs: differential

ligand effects on estrogen receptor interactions with

nuclear matrix sites. Endocr inology 128, 1785-1791.

Metzger DA and Korach KS ( 1 9 9 0 ) Cell-free interaction of

the estrogen receptor with mouse uterine nuclear matrix:

evidence of saturability, specificity, and resistance to KCl

extraction. Endocr inology 126, 2190-2195.

Michelotti EF, Michelotti GA, Aronsohn AI, and Levens D

( 1 9 9 6 ) Heterogeneous nuclear ribonucleoprotein K is a

transcription factor. M o l . C e l l . B i o l . 16, 2350-2360.

Mielke C, Maass K, Tümmler M, and Bode J ( 1 9 9 6 )

Anatomy of highly expressing chromosomal sites

targeted by retroviral vectors. Biochemis try 35, 2239-

2252.


526

Miller CA, Cohen MD, and Costa M ( 1 9 9 1 ) Complexing of

actin and other nuclear proteins to DNA by cis-

diamminedichloroplatinum(II) and chromium compounds.

Carc inogenes i s 12, 269-276.

Miller TE, Beausang LA, Winchell LF, and Lidgard GP

( 1 9 9 2 ) Detection of nuclear matrix proteins in serum

from cancer patients. Cancer Res. 52, 422-427.

Miller TL, Huzel NJ, Davie JR, and Murphy LC ( 1 9 9 3 ) C-

myc gene chromatin of estrogen receptor positive and

negative breast cancer cells. M o l . C e l l . E n d o c r i n o l .

91, 83-89.

Miller TL, Jin Y, Sun J-M, Coutts AS, Murphy LC, and Davie

JR ( 1 9 9 6 ) Analysis of human breast cancer nuclear

proteins binding to the promoter elements of the c-myc

gene. J . Ce l l . B iochem. 60, 560-571.

Miyashita T, Yamamoto H, Nishimune Y, Nozaki M, Morita

T, and Matsushiro A ( 1 9 9 4 ) Activation of the mouse

cytokeratin A (endoA) gene in teratocarcinoma F9 cells by

the histone deacetylase inhibitor Trichostatin A. FEBS

Lett . 353, 225-229.

Mizzen CA, Yang XJ, Kokubo T, Brownell JE, Bannister AJ,

Owen-Hughes T, Workman J, Wang L, Berger SL,

Kouzarides T, Nakatani Y, and Allis CD ( 1 9 9 7 ) The

TAFII250 subunit of TFIID has histone acetyltransferase

activity. Cel l 87, 1261-1270.

Mueller PR and Wold B ( 1 9 8 9 ) In vivo footprinting of a

muscle specific enhancer by ligation mediated PCR.

Science 246, 780-786.

Murphy LC, Huzel N, and Davie JR ( 1 9 9 6 ) Novel DNase I

hypersensitive sites in the 3'-flanking regions of the

human c-myc gene. DNA Cel l B io l . 15, 543-548.

Mutskov VJ, Russanova VR, Dimitrov SI, and Pashev IG

( 1 9 9 6 ) Histones associated with non-nucleosomal rat

ribosomal genes are acetylated while those bound to

nucleosome-organized gene copies are not. J. B i o l .

Chem. 271, 11852-11857.

Müller BF, Paulsen D, and Deppert W ( 1 9 9 6 ) Specific

binding of MAR/SAR DNA-elements by mutant p53.

Oncogene 12, 1941-1952.

Nagy L, Kao H-Y, Chakravarti D, Lin RJ, Hassig CA, Ayer DE,

Schreiber SL, and Evans RM ( 1 9 9 7 ) Nuclear receptor

repression mediated by a complex containing SMRT,

mSin3A, and histone deacetylase. Cel l 89, 373-380.

Nakagomi K, Kohwi Y, Dickinson LA, and Kohwi-Shigematsu

T ( 1 9 9 4 ) A novel DNA-binding motif in the nuclear

matrix attachment DNA-binding protein SATB1. M o l .

C e l l . B i o l . 14, 1852-1860.

Nakayasu H and Berezney R ( 1 9 9 1 ) Nuclear matrins:

identification of the major nuclear matrix proteins. Proc.

Nat l . Acad. Sc i . U. S . A. 88, 10312-10316.

Nickerson JA, Blencowe BJ, and Penman S ( 1 9 9 5 ) The

architectural organization of nuclear metabolism. Int .

R e v . C y t o l . 162A, 67-123.

Nikolov DB and Burley SK ( 1 9 9 7 ) RNA polymerase II

transcription initiation: A structural view. P r o c . N a t l .

Acad. Sci . USA 94, 15-22.

O'Neill LP and Turner BM ( 1 9 9 5 ) Histone H4 acetylation

distinguishes coding regions of the human genome from

heterochromatin in a differentiation-dependent but

transcription-independent manner. EMBO J. 14, 3946-

3957.

Oelgeschläger M, Janknecht R, Krieg J, Schreek S, and

Lüscher B ( 1 9 9 6 ) Interaction of the co-activator CBP

with Myb proteins: Effects on Myb-specific

transactivation and on the cooperativity with NF-M.

EMBO J. 15, 2771-2780.

Oesterreich S, Lee AV, Sullivan TM, Samuel SK, Davie JR, and

Fuqua SAW ( 1 9 9 7 ) A novel nuclear matrix protein HET

binds to and influences activity of the HSP27 promoter.

J . Ce l l . B iochem. 67, 275-286.

Ogryzko VV, Schiltz RL, Russanova V, Howard BH, and

Nakatani Y ( 1 9 9 7 ) The transcriptional coactivators

p300 and CBP are histone acetyltransferases. Cel l 87,

953-959.

Olinski R, Wedrychowski A, Schmidt WN, Briggs RC, and

Hnilica LS ( 1 9 8 7 ) In vivo DNA-protein cross-linking

by cis- and trans-diamminedichloroplatinum(II). Cancer

Res . 47, 201-205.

Orphanides G, Lagrange T, and Reinberg D ( 1 9 9 6 ) The

general transcription factors of RNA polymerase II.

Genes Dev. 10, 2657-2683.

Parthun MR, Widom J, and Gottschling DE ( 1 9 9 6 ) The

major cytoplasmic histone acetyltransferase in yeast:

Links to chromatin replication and histone metabolism.

Cel l 87, 85-94.

Pascal E and Tjian R ( 1 9 9 1 ) Different activation domains of

Sp1 govern formation of multimers and mediate

transcriptional synergism. Genes Dev. 5, 1646-1656.

Penman S ( 1 9 9 5 ) Rethinking cell structure. P r o c . N a t l .

Acad. Sci . USA 92, 5251-5257.

Penner CG and Davie JR ( 1 9 9 2 ) Multisubunit erythroid

complexes binding to the enhancer element of the chicken

histone H5 gene. Biochem. J . 283, 905-911.

Penner CG and Davie JR ( 1 9 9 4 ) Transcription factor GATA-

1-multiprotein complexes and chicken erythroid

development. FEBS Lett. 342, 273-277.

Pienta KJ, Murphy BC, Getzenberg RH, and Coffey DS

( 1 9 9 1 ) The effect of extracellular matrix interactions on

morphologic transformation in vitro. B i o c h e m .

Biophys . Res . Commun. 179, 333-339.

Pilat MJ, Christman JK, and Brooks SC ( 1 9 9 6 )

Characterization of the estrogen receptor transfected

MCF10A breast cacner cell line 139B6. Breast Cancer

Res. Treat. 37, 253-266.

Pommier Y, Cockerill PN, Kohn KW, and Garrard WT ( 1 9 9 0 )

Identification within the simian virus 40 genome of a

chromosomal loop attachment site that contains

topoisomerase II cleavage sites. J . Viro l . 64, 419-423.


527

Pugh BF ( 1 9 9 6 ) Mechanisms of transcription complex

assembly. Curr. Opin. Cel l Biol . 8, 303-311.

Qian Y-W, Wang Y-C, J., Hollingsworth, Jr., Jones D, Ling

N, and Lee EY-HP ( 1 9 9 3 ) A retinoblastoma-binding

protein related to a negative regulator of Ras in yeast.

Nature 364, 648-652.

Ramakrishnan V ( 1 9 9 4 ) Histone structure. Curr. Opin .

Struct . B io l . 4, 44-50.

Reitman M, Lee E, Westphal H, and Felsenfeld G ( 1 9 9 3 ) An

enhancer/locus control region is not sufficient to open

chromatin. M o l . C e l l . B i o l . 13, 3990-3998.

Renz A and Fackelmayer FO ( 1 9 9 6 ) Purification and

molecular cloning of the scaffold attachment factor B

(SAF-B), a novel human nuclear protein that specifically

binds to S/MAR-DNA. N u c l e i c A c i d s R e s . 24, 843-

849.

Replogle-Schwab TS, Pienta KJ, and Getzenberg RH ( 1 9 9 6 )

The utilization of nuclear matrix proteins for cancer

diagnosis. Crit . Rev . Eukaryot. Gene Expr. 6,

103-113.

Ridsdale JA, Hendzel MJ, Delcuve GP, and Davie JR ( 1 9 9 0 )

Histone acetylation alters the capacity of the H1 histones

to condense transcriptionally active/competent

chromatin. J . B io l . Chem. 265, 5150-5156.

Roth SY and Allis CD ( 1 9 9 6 ) Histone acetylation and

chromatin assembly: A single escort, multiple dances.

Cel l 87, 5-8.

Ruh MF, Cox LK, and Ruh TS ( 1 9 9 6 ) Estrogen receptor

interaction with specific histones - Binding to genomic

DNA and an estrogen response element. B i o c h e m .

Pharmacol. 52, 869-878.

Ruiz-Carrillo A, Wangh LJ, and Allfrey VG ( 1 9 7 6 ) Selective

synthesis and modification of nuclear proteins during

maturation of avian erythroid cells. Arch. B i o c h e m .

B i o p h y s . 174, 273-290.

Rundlett SE, Carmen AA, Kobayashi R, Bavykin S, Turner

BM, and Grunstein M ( 1 9 9 6 ) HDA1 and RPD3 are

members of distinct yeast histone deacetylase complexes

that regulate silencing and transcription. Proc. Nat l .

Acad. Sci . USA 93, 14503-14508.

Samuel SK, Minish MM, and Davie JR ( 1 9 9 7 a ) Nuclear

matrix proteins in well and poorly differentiated human

breast cancer cell lines. J . Ce l l . B iochem. 66, 9-15.

Samuel SK, Minish TM, and Davie JR ( 1 9 9 7 b ) Altered

nuclear matrix protein profiles in oncogene transformed

fibroblasts exhibiting high metastatic potential. Cancer

Res . 57, 147-151.

Sauer F, Hansen SK, and Tjian R ( 1 9 9 5 ) Multiple TAFIIs

directing synergistic activation of transcription.

Science 270, 1783-1788.

Sauman I and Berry SJ ( 1 9 9 4 ) An actin infrastructure is

associated with eukaryotic chromosomes: Structural and

functional significance. E u r . J . C e l l B i o l . 64, 348-

356.

Schwarz PM, Felthauser A, Fletcher TM, and Hansen JC

( 1 9 9 6 ) Reversible oligonucleosome self-association:

Dependence on divalent cations and core histone tail

domains. Biochemis try 35, 4009-4015.

Shen X, Yu L, Weir JW, and Gorovsky MA ( 1 9 9 5 ) Linker

histones are not essential and affect chromatin

condensation in vivo. Cel l 82, 47-56.

Stein GS, Lian JB, Dworetzky SI, Owen TA, Bortell R, Bidwell

JP, and van Wijnen AJ ( 1 9 9 1 ) Regulation of

transcription-factor activity during growth and

differentiation: involvement of the nuclear matrix in

concentration and localization of promoter binding

proteins. J . Ce l l B iochem. 47, 300-305.

Stein GS, van Wijnen AJ, Stein JL, Lian JB, Bidwell JP, and

Montecino M ( 1 9 9 4 ) Nuclear architecture supports

integration of physiological regulatory signals for

transcription of cell growth and tissue-specific genes

during osteoblast differentiation. J . C e l l . B i o c h e m .

55, 4-15.

Stuurman N, Meijne AM, van der Pol AJ, de Jong L, van Driel

R, and van Renswoude J ( 1 9 9 0 ) The nuclear matrix from

cells of different origin. Evidence for a common set of

matrix proteins. J . B io l . Chem. 265, 5460-5465.

Sun J-M, Penner CG, and Davie JR ( 1 9 9 2 ) Analysis of

erythroid nuclear proteins binding to the promoter and

enhancer elements of the chicken histone H5 gene.

Nucleic Acids Res . 20, 6385-6392.

Sun J-M, Penner CG, and Davie JR ( 1 9 9 3 ) Repression of

histone H5 gene expression in chicken mature

erythrocytes is correlated with reduced DNA-binding

activites of transcription factors Sp1 and GATA-1. FEBS

Lett . 331, 141-144.

Sun J-M, Chen HY, and Davie JR ( 1 9 9 4 ) Nuclear factor 1 is a

component of the nuclear matrix. J . C e l l . B i o c h e m .

55, 252-263.

Sun J-M, Chen HY, Litchfield DW, and Davie JR ( 1 9 9 6 a )

Developmental changes in transcription factors

associated with the nuclear matrix of chicken

erythrocytes. J . Ce l l . B iochem. 62, 454-466.

Sun J-M, Ferraiuolo R, and Davie JR ( 1 9 9 6 b ) In situ

footprinting of chicken histone H5 gene in mature and

immature erythrocytes reveals common factor binding

sites. Chromosoma 104, 504-510.

Tauton J, Hassig CA, and Schreiber SL ( 1 9 9 6 ) A mammalian

histone deacetylase related to the yeast transcriptional

regulator Rpd3p. Science 272, 408-411.

Thorburn A and Knowland J ( 1 9 9 3 ) Attachment of

vitellogenin genes to the nucleoskeleton accompanies

their activation. Biochem. B i o p h y s . R e s .

Commun. 191, 308-313.

Tjian R and Maniatis T ( 1 9 9 4 ) Transcriptional activation: A

complex puzzle with few easy pieces. Cel l 77, 5-8.

Van Holde, K.E. (1988). Chromatin (New York: Springer-

Verlag).


528

van Steensel B, Jenster G, Damm K, Brinkmann AO, and van

Driel R ( 1 9 9 5 ) Domains of the human androgen receptor

and glucocorticoid receptor involved in binding to the

nuclear matrix. J . Ce l l . B iochem. 57, 465-478.

van Wijnen AJ, Bidwell JP, Fey EG, Penman S, Lian JB, Stein

JL, and Stein GS ( 1 9 9 3 ) Nuclear matrix association of

multiple sequence-specific DNA binding activities related

to SP-1, ATF, CCAAT, C/EBP, OCT-1, and AP-1.


Vassetzky YS, De Moura Gallo CV, Bogdanova AN, Razin SV,

and Scherrer K ( 1 9 9 3 ) The sequence-specific nuclear

matrix binding factor F6 is a chicken GATA-like protein.

Mol. Gen. Genet . 238, 309-314.

Verreault A, Kaufman PD, Kobayashi R, and Stillman B

( 1 9 9 6 ) Nucleosome assembly by a complex of CAF-1

and acetylated histones H3/H4. Cel l 87, 95-104.

von Kries JP, Buhrmester H, and Stratling WH ( 1 9 9 1 ) A

matrix/scaffold attachment region binding protein:

identification, purification, and mode of binding. Cel l

64, 123-135.

Wade PA and Wolffe AP ( 1 9 9 7 ) Histone acetyltransferases in

control. Curr. Biol . 7, R82-R84.

Wansink DG, Sibon OCM, Cremers FFM, van Driel R, and de

Jong L ( 1 9 9 6 ) Ultrastructural localization of active

genes in nuclei of A431 cells. J . C e l l . B i o c h e m . 62,

10-18.

Wedrychowski A, Schmidt WN, and Hnilica LS ( 1 9 8 6 ) The

in vivo cross-linking of proteins and DNA by heavy

metals. J . B io l . Chem. 261, 3370-3376.

Wedrychowski A, Bhorjee JS, and Briggs RC ( 1 9 8 9 ) In vivo

crosslinking of nuclear proteins to DNA by cis-

diamminedichloroplatinum (II) in differentiating rat

myoblasts. Exp. Ce l l Res . 183, 376-387.

Wolffe AP ( 1 9 9 7 ) Sinful repression. Nature 387, 16-17.

Woodcock CL and Horowitz RA ( 1 9 9 5 ) Chromatin

organization re-viewed. Trends Ce l l B i o l . 5, 272-

277.

Xing YG, Johnson CV, Moen PT, Jr., McNeil JA, and

Lawrence JB ( 1 9 9 5 ) Nonrandom gene organization:

Structural arrangements of specific pre-mRNA

transcription and splicing with SC-35 domains. J . C e l l

B i o l . 131, 1635-1647.

Yanagisawa J, Ando J, Nakayama J, Kohwi Y, and Kohwi-

Shigematsu T ( 1 9 9 6 ) A matrix attachment region

(MAR)-binding activity due to a p114 kilodalton protein

is found only in human breast carcinomas and not in

normal and benign breast disease tissues. Cancer Res .

56, 457-462.

Yang WM, Inouye C, Zeng YY, Bearss D, and Seto E ( 1 9 9 6 )

Transcriptional repression by YY1 is mediated by

interaction with a mammalian homolog of the yeast

global regulator RPD3. Proc . Nat l . Acad . Sc i . USA

93, 12845-12850.

Yang XJ, Ogryzko VV, Nishikawa J, Howard BH, and Nakatani

Y ( 1 9 9 6 ) A p300/CBP-associated factor that competes

with the adenoviral oncoprotein E1A. Nature 382, 319-

324.

Yoshida M, Horinouchi S, and Beppu T ( 1 9 9 5 ) Trichostatin

A and trapoxin: novel chemical probes for the role of

histone acetylation in chromatin structure and function.

BioEssays 17, 423-430.

Zhang Y, Iratni R, Erdjument-Bromage H, Tempst P, and

Reinberg D ( 1 9 9 7 ) Histone deacetylases and Sap18, a

novel polypeptide, are components of a human Sin3

complex. Cel l 89, 357-364.

Zini N, Mazzotti G, Santi P, Rizzoli R, Galanzi A, Rana R, and

Maraldi NM ( 1 9 8 9 ) Cytochemical localization of DNA

loop attachment sites to the nuclear lamina and to the

inner nuclear matrix. Histochemistry 91, 199-204.

Gene Therapy and Molecular Biology Vol 1, page 529

529

Gene Ther Mol Biol Vol 1, 529-542. March, 1998.

Structural organization and biological roles of thenuclear lamina

Amnon Harel1, Michal Goldberg, Nirit Ulitzur2 and Yosef Gruenbaum

Department of Genetics, The Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem 91904, Israel.__________________________________________________________________________________________________Correspondence : Yosef Gruenbaum, Tel: +972-2-6585995, Fax: +972-2-5633066 or +972-2-6586975, E-mail:[email protected]. Present address: Department of Biology, University of California, San Diego, San Diego CA.2. Present address: Department of Biochemistry, Stanford Medical School, Stanford CA.

SummaryThe nuclear lamina is a protein meshwork that l i es on the nucleoplasmic side of the nuclearenvelope and is associated with the peripheral chromatin. It i s involved in several biologicalactivities including: the mitotic disassembly and reassembly of the nuclear envelope, determinationof the size and shape of the nucleus, higher order chromatin organization, cell differentiation, andapoptosis . Lamins are the major proteins of the nuclear lamina. They are type V intermediatefilaments and, l ike al l intermediate filaments, they form filamentous structures. Lamins caninteract in vitro with specific DNA sequences, with chromosomal proteins and with severalproteins of the inner nuclear membrane, including otefin, LBR, LAP1 and LAP2. In this paper weshow that Drosophila lamin Dm0 and otefin proteins are required for the assembly of the

Drosophila nuclear envelope. We also demonstrate that the lack of lamin Dm0 activity causes the

dissociation of peripheral chromatin from the nuclear envelope, accumulation of annulate lamellaeand lethality. In addition, we show that the carboxy (tai l) domain of lamin Dm0 can interact in

vitro with chromosomes and the central (rod) domain of lamin Dm0 is essential and sufficient for

the in v i tro assembly of lamin Dm0 into f i lamentous structures. These results are discussed in

relationship to the biological roles of the nuclear lamina.

I. Introduction

In eukaryotic cells, DNA replication and RNAprocessing occur in the nucleus, while protein synthesisoccurs in the cytoplasm. These activities are physicallyseparated by the nuclear envelope. The nuclear envelope isa complex structure composed of outer and inner lipidbilayer membranes. The two membranes are separated by a20-40 nm perinuclear space and are connected at the nuclearpore complexes, which are passageways for transport ofmacromolecules between the nucleoplasm and thecytoplasm (reviewed in Davis, 1995; Gorlich and Mattaj,1996). Underlying the inner nuclear membrane there is aproteinaceous meshwork of intermediate filaments termedthe nuclear lamina (F i g . 1 A ; reviewed in Hutchison etal., 1994; Moir et al., 1995).

A. Proteins of the inner nuclear membraneand nuclear lamina

Several components of the inner nuclear membrane andthe lamina have been identified. These include the integralmembrane proteins (IMPs): LBR (Worman et al., 1990),LAP1, LAP2 (Furukawa et al., 1995; Harris et al., 1994;Martin et al., 1995), p34 (Simos and Georgatos, 1994)and p18 (Simos et al., 1996), and the peripheral proteins:nuclear lamins (Fisher et al., 1986; McKeon, 1991),otefin (Harel et al., 1989; Padan et al., 1990) and YA(Lopez et al., 1994; Lopez and Wolfner, 1997). Theexisting experimental data suggests that lamins caninteract with LBR, LAP1, LAP2, otefin and YA (Foisnerand Gerace, 1993; Goldberg et al., 1997; Worman etal., 1988) . p18 and p34 are associated with LBR and p18is distributed equally between the inner and the outernuclear membranes (Simos and Georgatos, 1994). The dataon the peripheral proteins indicates that otefin is closelyassociated with the inner nuclear membrane, lamin canassociate with both the inner nuclear membrane and theperipheral chromatin, and YA is associated with theperipheral chromatin (F i g . 1 A ; Goldberg et al., 1997).These proteins are present in the insoluble NMPCL (nu-

Harel et al: Structural Organization and Biological Roles of the Nuclear Lamina

530

Figure 1 . (A) Schematic view of the structural organization of nuclear envelope. OMN, outer nuclear membrane; INM, innernuclear membrane; NPC, nuclear pore complex. (B ) Schematic view and the putative roles of different regions in lamin Dm0 and

otefin. The numbers are of amino acids positions in these proteins.


531

clear matrix-pore-complex-lamina) fraction, after salt andTriton X-100 extraction.

LBR (lamin B receptor) was isolated by its ability tobind in a saturable and specific fashion to lamin B.Binding of lamin B to LBR is affected by itsphosphorylation. LBR is a 58 kDa protein containing anucleoplasmic amino-terminal domain of 204 amino acidsfollowed by a hydrophobic domain with eight putativetransmembrane segments (Worman et al., 1990). Itssequence shows high homology to the yeast sterol C-14reductase (Gerace and Foisner, 1994). Both the firsttransmembrane domain (Smith and Blobel, 1993) and theamino-terminal domain of LBR (Soullam and Worman,1993; Soullam and Worman, 1995) mediate the targetingof LBR to the inner nuclear membrane. The highly chargedamino-terminal domain of LBR can also direct cytosolicproteins to the nucleus and type II integral membraneproteins to the inner nuclear membrane in transfectedCOS-7 cells (Smith and Blobel, 1993). LBR isphosphorylated in a cell cycle-dependent manner on serineresidues in interphase and on serine and threonine residuesin mitosis. Its phosphorylation is mediated by p34cdc2-kinase and by an unidentified kinase that resides in thenuclear envelope and associates with LBR in vivo(Nikolakaki et al., 1997; Simos and Georgatos, 1992).LBR can interact with several proteins including p34 andp18 (Simos and Georgatos, 1994), lamin B (Worman etal., 1988) and with the human homologue of theDrosophila heterochromatin associated protein HP1 (Yeand Worman, 1996).

LAP1A-C - (Lamina-associated polypeptides 1A-C)is a group of three related integral membrane proteins ofthe inner nuclear membrane that are recognized bymonoclonal antibody RL13. LAP1 proteins can bind bothtype A and B lamins (Foisner and Gerace, 1993). Cloningof LAP1C revealed that it is a type II integral membraneprotein with a single membrane-spanning region and ahydrophilic amino terminal domain that is exposed to thenucleoplasm (Martin et al., 1995). The different LAP1isotypes are differentially expressed during developmentand appear to bind lamins with different affinities (Martinet al., 1995).

LAP2 (Lamina-associated polypeptide 2 - also namedthymopoietin) is a type II integral membrane protein ofthe inner nuclear membrane. The LAP2 gene isalternatively spliced to give rise to at least 5 differentproducts (Theodor et al., 1997). The most abundantproducts: LAP2!, LAP2", and LAP2# (75 kDa, 51 kDaand 39 kDa, respectively) are present in most cell types.LAP2! is present diffusely throughout the nucleus, whileLAP2" and LAP2# are confined to the inner nuclearmembrane (Harris et al., 1995). LAP2" contains a largehydrophilic domain with several potential cdc2 kinasephosphorylation sites and a single putative membrane-spanning sequence close to its carboxy terminus. Theamino-terminal domain of this protein is hydrophilic andis exposed to the nucleoplasm. LAP2 can bind directly toboth lamin B and chromosomes and associates with

chromosomes at the same time that lamins begin toreassemble around them (Foisner and Gerace, 1993; Yanget al., 1997). The phosphorylation of LAP2 duringmitosis inhibits its binding to both lamin B andchromosomes. (Foisner and Gerace, 1993). Themechanism for inner membrane targeting and retention ofLAP2 probably involves lateral diffusion in theinterconnected membranes of the endoplasmatic reticulumand nuclear envelope, and interaction with components ofthe nuclear lamina and chromatin (Furukawa et al., 1995).

YA (Young Arrest) is an essential Drosophila gene forthe transition from meiosis to the initiation of the rapidmitotic divisions by early embryos (Judd and Young,1973; Lin et al., 1991; Liu et al., 1995). Thechromosome condensation state is abnormal in nuclei inYA-deficient eggs and embryos (Liu et al., 1995). The YAprotein is present during the first two hours of zygoticdevelopment, where it is localized to the nuclear lamina(Lin et al., 1991). Ectopically expressed YA associateswith polytene chromosomes in vivo (Lopez and Wolfner,1997), and YA can associate with both chromosomes andlamin Dm0 (Goldberg et al., 1997; Lopez and Wolfner,

1997).

Otefin is a 45 kDa peripheral nuclear envelopeprotein with no apparent homology to other knownproteins (Padan et al., 1990). It includes a largehydrophilic domain, a single carboxy terminal hydrophobicsequence of 17 amino acids and a high content of serineand threonine residues (Fig . 1B ). With the exception ofsperm cells, otefin is present in the nuclear envelope of allcells examined during the different stages of Drosophiladevelopment. In eggs and young embryos, otefin is alsoassociated with the maternal fraction of membrane vesicles(Ashery-Padan et al., 1997b). The COOH-terminal, 17-aahydrophobic sequence of otefin is essential for thetargeting of otefin to the nuclear periphery. Othersequences of otefin are required for its efficient targeting tothe nuclear envelope and for further stabilizing otefin'sinteraction with the nuclear envelope (Ashery Padan et al.,1997a). Otefin is a phosphoprotein in vivo and a substratefor in vitro phosphorylation by cdc2 kinase and cAMP-dependent protein kinase.

Lamins are the major proteins of the nuclearenvelope. They are classified as type V intermediatefilaments and, like all intermediate filaments, they containan ! helical rod domain flanked by amino (head) andcarboxy (tail) domains (Fig . 1B). Unlike thecytoplasmic intermediate filaments that are 10 nm wide,lamins can make up to 200 nm thick fibers (Belmont etal., 1993; Paddy et al., 1990). The rod domain of laminsis 52 nm long and contains three ! helices, each composedof heptad repeats (reviewed in McKeon, 1987). Thesehelices form coiled-coil interactions between laminmonomers. The lamin dimers associate longitudinally toform polar head-to-tail polymers. These polar head-to-tailpolymers further associate laterally to form the 10 nmthick filaments (Heitlinger et al., 1991). The 10 nmfilaments further associate to form the 50-200 thick


532

nuclear lamina (this study). The head-to-tail binding sitesare at the ends of the rod domain that are highly conservedamong all intermediate filament proteins. Point mutantsthat cause defects in binding were mapped to theseconserved regions (Stuurman et al., 1996; Zhao et al.,1996).

Lamins are divided into types A and B. Type A laminsare mainly expressed in differentiated cells, have a neutralisoelectric point and are soluble during mitosis. Type Blamins are expressed constitutively in all somatic cells,have an acidic isoelectric point and remain associated withmembrane vesicles during mitosis (reviewed in McKeon,1991; Nigg, 1992). Different eukaryotes possess betweenone to six lamin genes. Mammalian lamins A and C arethe result of alternative splicing of the same gene. LaminsB1-B3 and C2 are coded by separate genes (Alsheimer andBenavente, 1996). The two major lamins in chicken arelamins A and B2 (Peter et al., 1989). An additional minorspecies is termed lamin B1. Xenopus laevis has at leastfive different lamin genes (Stick, 1992; Stick, 1994).Drosophila melanogaster has two lamin genes, termedlamin Dm0 and C (Bossie and Sanders, 1993; Gruenbaum

et al., 1988). Caenorhabditis elegans probably has a singlelamin gene, termed CeLam-1 (Riemer et al., 1993).

Lamins undergo specific post translationalmodifications. All nuclear lamins except lamins C containCaaX box at their carboxy terminus. The CaaX boxundergoes proteolytic cleavage of the last three aminoacids, farnesylation of the C-terminal cysteine, andcarboxyl methylation. The isoprenylation is essential butnot sufficient for the association of lamins with thenuclear envelope (Firmbach and Stick, 1995; Firmbach-Kraft and Stick, 1993; Hennekes and Nigg, 1994).Lamins are phosphorylated by several protein kinases invivo and in vitro. These include: cdc2 kinase (Dessev etal., 1991; Heald and McKeon, 1990; Peter et al., 1990;Ward and Kirschner, 1990), Casein kinase II (Li and Roux,1992), PKA (Lamb et al., 1991), "II PKC (Fields et al.,1988; Hennekes et al., 1993; Hocevar et al., 1993;Hocevar and Fields, 1991; Kasahara et al., 1991) andMAP kinase (Peter et al., 1992). The phosphorylationstate of lamins is cell-cycle regulated (Ottaviano andGerace, 1985). It is involved in lamin polymerization anddisassembly, and in importing lamin molecules into thenucleus. The Drosophila lamin Dm0 undergoes post

translational modifications to give rise to at least threedistinct isoforms termed, Dm1, Dm2 and Dmmit which

differ in their phosphorylation pattern. Dm1 and Dm2 are

present in most types of interphase nuclei as a randommixture of homo- and hetero-dimers (Smith et al., 1987;Stuurman et al., 1995). Dmmit is present in the maternal

pool and in mitotic cells (Smith and Fisher, 1989).

3-D in vivo studies in Drosophila and in mammaliancells revealed that lamin fibers are closely associated withchromatin fibers (Belmont et al., 1993; Paddy et al.,1990). Studies in vitro have shown that lamins canspecifically bind chromatin fragments and interphasechromatin (Hoger et al., 1991; Taniura et al., 1995;

Yuan et al., 1991), as well as condensed in vitroassembled chromatin (Ulitzur et al., 1992) or mitoticchromosomes (Glass et al., 1993; Glass and Gerace,1990). Lamins can also bind to specific DNA sequences(Baricheva et al., 1996; Luderus et al., 1992; Luderuset al., 1994; Shoeman and Traub, 1990; Zhao et al.,1996) and to chromosomal proteins (Burke, 1990; Glasset al., 1993; Glass and Gerace, 1990; Hoger et al., 1991;Taniura et al., 1995; Yuan et al., 1991). Binding oflamins to chromatin is specific and depends on theintegrity of the chromosomes. Lamin A binds in vitro topoly-nucleosomes with a dissociation constant of about

1x10-9 M (Yuan et al., 1991). A binding site formammalian lamins A and B was localized at the taildomain (Taniura et al., 1995). In the latter study, thedissociation constant of the tail domain binding tointerphase chromatin was estimated to be in the range of

3x10-7 M and the binding was mediated by histones. Sincelamins form large polymers in vivo, the actual associationbetween the lamin filament and chromatin may bestronger. A specific binding site to mitotic chromosomeswas also found in the rod domain. However, the in vivorelevance of this binding is not yet clear since the roddomain binding occurred only under acidic, non-physiological, conditions (Glass et al., 1993). Chickenlamin B and Drosophila lamin Dm0 polymers also bind

specifically to M/SARs fragments (Luderus et al., 1992;Luderus et al., 1994). These DNA sequences are severalhundred base pairs long with several stretches of AT richsequences and are likely to form an "open" form ofchromatin. Indeed, the binding to these sequences could becompeted to some extent with single strand DNA (Luderuset al., 1994). The binding of Drosophila lamin Dm0 to

M/SARs is mediated by the rod domain and requires itspolymerization (Zhao et al., 1996). Lamin-DNAinteractions can occur, for example, in the centromericregions since the l20p1.4 Drosophila centromeric sequencehas DNA composition similar to M/SAR and it bindsspecifically to polymers of Drosophila lamin Dm0(Baricheva et al., 1996). Lamin polymers can also bindstrongly to telomeric sequences (Shoeman and Traub,1990).

B. Biological roles of the nuclear lamina

Several functions have been ascribed to the nuclearlamina concerning nuclear organization and activity. Thesefunctions include: (i) regulating the size, shape andassembly of the nuclear envelope, (ii) a role in higher orderchromatin organization by providing docking sites forchromatin , (iii) a role in DNA replication, (iv) a possiblerole in differentiation, as indicated by the change in laminacomposition during development. In addition, the nuclearlamina is a major substrate for signals that control the cellcycle and lamins are specifically degraded in apoptosis(Nigg, 1992; Oberhammer et al., 1994).

(i) Nuclear envelope disassembly.


533

During mitosis, the nuclear envelope breaks down inprophase and starts to reassemble at late anaphase. Nuclearlamins and lamina associated proteins are likely to play arole both in the assembly and disassembly of the nuclearenvelope. Disassembly of the nuclear lamina is controlledby phosphorylation of sites outside the rod domain oflamins that prevents the head-to tail association of thelamin molecules. For example, mutations in Ser-22 andSer-392 of human lamin A in transfected COS cellsprevented phosphorylation at these sites and blocked thedisassembly of the nuclear lamina during mitosis (Healdand McKeon, 1990).

( i i ) Nuclear envelope assembly depends onlamins and on lamin-associated proteins. Microinjection oflamin antibodies into cultured PtK2 cells resulted indaughter nuclei that remained arrested in a telophase-likeconfiguration, and telophase-like chromatin that remainedinactive (Benavente and Krohne, 1986). In mammaliancell-free extracts, antibodies directed against type A or Blamins blocked vesicles binding to chromatin, which is thefirst step of nuclear envelope assembly (Burke and Gerace,1986). Similarly, anti-lamin Dm0 antibodies blocked the

interaction between vesicles and chromatin in a Drosophilacell-free system that assembles nuclei from spermchromatin (Ulitzur et al., 1992; Ulitzur et al., 1997). Therole of lamin proteins in the association between nuclearvesicles and chromatin in Xenopus extracts has been thesubject of debate; Depletion of lamin B3 from theassembly extract did not prevent the formation of nuclearenvelopes consisting of membranes and nuclear pores.These lamin B3-depleted nuclei were small, fragile andfailed to replicate their DNA (Jenkins et al., 1995; Meieret al., 1991; Newport et al., 1990). In contrast,Dabauvalle et al. (Dabauvalle et al., 1990) were able toblock the formation of nuclear envelopes by using anantibody directed against both lamins B2 and B3. A majorreason for the discrepancy between the above studies couldbe that Xenopus extracts contain lamins B2 and B1, inaddition to lamin B3 (Lourim et al., 1996; Lourim andKrohne, 1993). In cell-free extracts of Xenopus eggs andDrosophila melanogaster it was shown that trypsinizationof the membrane fraction abolished its ability to binddemembranated sperm chromatin and hence to supportassembly of the nuclear envelope (Ulitzur et al., 1997;Wilson and Newport, 1988). Possible target proteins forthe Trypsin treatment are IMPs. Indeed, several studiessuggest a role for LBR, LAP1 and LAP2 in nuclearassembly. LAP2 associates with chromosomes at the sametime as lamins, which suggests a role for LAP2 in initialevents of nuclear envelope reassembly (Foisner and Gerace,1993). A recent study (Yang et al., 1997) shows thatLAP1 and LAP2 become completely dispersed throughoutER membranes during mitosis and proposes that thereassembly of the nuclear envelope at the end of mitosisinvolves sorting of IMPs to chromosome surfaces bybinding interactions with lamins and chromatin.Pyrpasopoulou et al. (Pyrpasopoulou et al., 1996)analyzed the role of LBR in providing chromatin dockingsites for nuclear vesicles by binding in vitro reconstituted

vesicles of nuclear envelopes to chromatin. The results ofthis study suggest that LBR is involved in providingchromatin anchorage site at the nuclear envelope. It wasalso suggested that the homologue of LBR in sea urchintargets membranes to chromatin and later anchors themembrane to the lamina (Collas et al., 1996). Theessential role of otefin in the assembly of the nuclearenvelope was recently demonstrated in a Drosophila cell-free system (Ashery-Padan et al., 1997b). The similarphenotype obtained when otefin or lamin Dm0 activities

are inhibited (Ashery-Padan et al., 1997b) is probably dueto the fact that otefin and lamin are part of the sameprotein complex in the vesicle fraction (Goldberg et al.,1997). In summary, the above data implies that theassembly of nuclear membranes following mitosis requiresthe function of protein complexes containing bothperipheral and integral membrane proteins including:lamin, otefin, LAP2 and LBR.

Lamin genes are not present in significant homologyin the yeast Saccharomyces cerevisiae (Gruenbaum, Y.,unpublished observations) and in the protozoon Amoebaproteus (Schmidt et al., 1995). In addition, the lamina-associated proteins LAP1, LAP2 and otefin are not presentin significant homology in Saccharomyces cerevisiae(Gruenbaum, Y., unpublished observations), while LBR isthe enzyme sterol C14 reductase (reviewed in Gerace andFoisner, 1994). One possible explanation for theappearance of lamins only in organisms with an openmitosis concerns their roles in nuclear envelope breakdownat the begining of mitosis and nuclear reassembly at theend of mitosis. These activities are not required inorganisms with a closed mitosis. The involvement of thenuclear lamina in nuclear organization, development andDNA replication may have appeared later in evolution.

(iii). Nuclear and chromatin organization .

The nuclear lamina is a major component of thenuclear matrix. It was, therefore, suggested that a laminfilamentous meshwork is involved in nuclear andchromatin organization. An example for a directinvolvement of a lamin protein in nuclear organizationcomes from an ectopic expression of the mouse sperm-specific lamin B3 in cultured somatic cells. This ectopicexpression resulted in transformation of the nuclearmorphology from spherical to hook-shaped (Furukawa andHotta, 1993). Also, depletion of soluble lamin B3 fromXenopus nuclear assembly extracts gave in vitro assemblednuclei that were small and fragile (Meier et al., 1991;Newport et al., 1990). Another evidence for the role oflamin in nuclear organization comes from the analysis offlies mutated in the Drosophila lamin Dm0 gene. Flies

homozygous for a strong mutation in the lamin Dm0 gene

had an aberrant nuclear structure and died following 9-16hours of development. The dissociation of chromatin fromthe nuclear membrane was one of the first phenotypesobserved in these flies (Osman, 1992). A weak mutationin the lamin Dm0 gene (<20% of lamin expression)

resulted in a retarded development, reduced viability,


534

sterility, and impaired locomotion. The nuclei in thesemutant flies are enriched in nuclear pore complexes, incytoplasmic annulate lamellae and contain defective nuclearenvelopes (Lenz-Bohme et al., 1997). In vitro studiessupport the role of nuclear lamin in chromatinorganization. As discussed above, the nuclear laminainteracts in vivo with chromatin, and lamin proteins canbind histones and specific DNA sequences.

The Drosophila YA protein is needed to initiateembryonic cleavage divisions (Lopez et al., 1994). Ya islikely to be involved in mediating the association ofchromosomes with the lamina (Goldberg et al., 1997),thus contributing to the organization of the nucleus in adevelopmental stage-specific manner (Lopez and Wolfner,1997). Nuclei in YA-deficient eggs and embryos haveabnormal chromosome condensation states (Liu et al.,1995), ectopically expressed YA associates with polytenechromosomes in vivo, and YA can associate withchromosomes in vitro (Lopez et al., 1994; Lopez andWolfner, 1997).

(iv). DNA replication requires nuclearlamins .

Several reports demonstrated that, during interphase,lamin B molecules are present in foci in the nucleoplasm,in addition to their presence in the nuclear envelope. Thesefoci coincide with sites of DNA replication (Goldman etal., 1992; Moir et al., 1994; Spann et al., 1997). Inaddition, nuclei assembled in Xenopus egg extracts thatwere depleted of lamin B3 were unable to initiate DNAreplication. These lamin B3-depleted nuclei had continuousnuclear envelopes and nuclear pores and were able toimport proteins required for DNA synthesis such asPCNA, MCM3, ORC2 and DNA polymerase ! (Goldberget al., 1995; Meier et al., 1991; Newport et al., 1990;Spann et al., 1997). Addition of purified lamin B3 to thedepleted extracts could rescue lamina assembly and DNAreplication. Microinjection of a truncated human lamin,that was utilized as a dominant negative mutant to perturblamin organization in mammalian cells, caused a dramaticreduction in DNA replication (Spann et al., 1997). Nuclear

lamins are likely to be required for the elongation phase ofDNA replication since the distribution of MCM3, ORC2,and DNA polymerase ! that are required for the initiationstage of DNA replication was not affected by the depletionof lamin B3 activity (Spann et al., 1997).

II. Results and discussion

A. Mutations in the Drosophila laminDm0 gene reveal that it is an essential gene

that is required for nuclear organization.

During egg chamber development, large amounts oflamin Dm0 are secreted by the nurse cells into the

developing Drosophila oocyte (Ashery-Padan et al., 1997b;Smith and Fisher, 1989; Ulitzur et al., 1992). The

amounts of lamin Dm0 RNA and protein that are

maternally stored in the oocyte are sufficient for theassembly of many thousands of nuclei. In addition, laminDm0 is a very stable protein with an estimated half life of

about 24 hr (Dr. Paul A. Fisher, personalcommunication). Therefore, flies mutated in their laminDm0 gene are expected to show a phenotype only

following the consumption of the large maternal pool oflamin Dm0.

Drosophila melanogaster (canton S) males weremutagenized with ethyl methane sulphanate (ems) andoffspring flies mutated in their second chromosome werecrossed with flies containing the deletion Df(2L) gdh-A(Knipple et al., 1991). This deletion is between 25D7-26A7 bands and contains the 25F1 locus of lamin Dm0(Gruenbaum et al., 1988). One of the complementationgroups was specific for a mutation in lamin Dm0 since it

could be specifically rescued by a P-element mediatedtransformation with a CaspeR vector (Pirrotta, 1988)containing 1.2 kb upstream sequences of lamin Dm0 and

either the complete genomic lamin Dm0 gene (EcoRI-

EcoRI fragment; Osman et al., 1990) or the two firstexons and part of the third exon of the genomic lamingene (EcoRI-HindIII fragment; Osman et al., 1990) ligatedto the HindIII-EcoRI fragment of lamin Dm0 cDNA

(Gruenbaum et al., 1988). A second mutagenesis screenutilized a P-element targeted gene mutation, using theBirm-2/Birm-2; ry/ry line (Ballinger and Benzer, 1989;Kaiser and Goodwin, 1990). Candidate lines for a mutationin lamin Dm0 were crossed with an ems-mutated line in

lamin Dm0 (Osman, 1992). One of the isolated mutations,

termed PM-15, was analyzed in more details. PM-15/PM-15 or PM-15/Df(2L)gdh-A flies showed an abnormalchromatin organization following 9-16 hr of development.The variability in the time of phenotype appearance islikely to be due to differences in the amounts of thematernal pool of lamin Dm0. Flies homozygous or trans-

heterozygous for these mutations eventually die. Therefore,lamin Dm0 is an essential gene of the fruit fly (Osman,

1992). One of the first phenotypes in embryoshomozygous for a mutation in lamin Dm0 gene is the

detachment of the peripheral chromatin from the nuclearenvelope. This detachment occurs in many regions ofaffected nuclei and is followed by condensation ofchromatin (Fig . 2 . compare panels C,D to panels A,B).The later phenotypes of these embryos include nucleiaggregation and formation of cytoplasmic annulatelamellae (F i g . 2 E ). A Drosophila line mutated in itslamin Dm0 gene (Lenz-Bohme et al., 1997), in which the

amounts of lamin Dm0 protein are reduced to less than

20% of their normal levels, also revealed enrichment inannulate lamellae and in nuclear envelope clusters. These


535

Figure 2 . A mutation inlamin Dm0 gene results in

dissociation of chromatin fromthe nuclear envelope andaccumulation of annulate

lamellae (Osman et al., 1990).Embryos mutated in lamin Dm0

showed a visible phenotypefollowing 9-16 hr of

development. Electronmicroscope analysis of PM-

15/PM-15 cells (C,D ) revealedchromatin dissociation from

the nuclear envelope ascompared to normal cells

(A,B ). (E) Annulate lamellae inPM-15/Df(2L)gdh-A embryos

following degradation ofnuclei. The bars in panels

A,C,E represent 1 µm. The barsin panels B,D represent 5 µm.

flies showed reduced viability, retardation in theirdevelopment, sterility, and impaired locomotion. In somecells, defective nuclear envelopes were also observed (Lenz-Bohme et al., 1997). In summary, these studiesdemonstrate the essential role of lamins in nuclear andchromatin organization.

B. Filament assembly properties of laminDm0 and derivative proteins.

The assembly properties of lamin Dm0 were

investigated in vitro using bacterially expressed andpurifies lamin Dm0 and derivatives (Ulitzur et al., 1992).

To test for the ability of filamentous protein to polymerizewe used the sedimentation test (Heitlinger et al, 1991),which is based on the separation of pelletable polymersfrom soluble protein, following incubation under variouschemicals and pH conditions. Reduction of saltconcentration from 0.5 M NaCl to 50-150 mM NaCl in

pH range of 5-9 was sufficient to induce 35-95%polymerization of lamin Dm0 protein. Electron

microscope analysis of negative stained pellets confirmedthe formation of filamentous structures (Fig . 3 ). Theobserved paracrystals were characterized by a distinct stain-excluding pattern with 25 nm axial repeat unit, which ishalf the size of the lamin rod domain (Fig . 3 A,B).Figure 3C shows a relatively rare case which reveals thatthese paracrystals are composed of separate laminfilaments. These filaments are 8-10 nm wide, which is thenormal size of cytoplasmic intermediate filaments.Although there is no evidence for the existence ofparacrystals in vivo, it is noteworthy that the width ofthese paracrystals fits is in the size range of lamin fibersthat were visualized in Drosophila cells in vivo (Paddy etal., 1990).

The ability of the isolated rod domain of lamin Dm0(amino acids 55-413) to polymerize was analyzed utilizingbacterially expressed protein that was purified to near ho-


536

Figure 3 .Supramolecularstructures formed bylamin Dm0 at low

ionic strength.Lamin Dm0 protein

at 2 mg/ml in bufferH (30 mM Tris-HClpH 7.5, 1 mM DTT)containing 0.5 MNaCl was diluted 5times in buffer H andincubated for 80 minon ice. Samples wereplaced on electronmicroscope grid andnegatively stainedwith 1% uranylacetate. Tightlypacked paracrystalsexhibit ~25 nm axialrepeat unit and theirthickness rangedbetween 40-200 nm(A,B ). The thickparacrystallinearrays are composedof a large number ofthin filaments (C).

mogeneity. Unlike the complete lamin Dm0 molecule,

polymerization of the isolated rod domain was salt-independent. However, under acidic conditions (pH 5.5) andin the presence of 25 mM CaCl2, the isolated rod domain

was organized in higher order structures, as judged by thesedimentation test (F i g . 4 A ) and by electron microscopeanalysis (Fig . 4B). The filamentous structure of thepolymerized rod domain resembled that of the completelamin protein, but lacked the 25 nm repeat unit. Underneutral and basic pH conditions the rod domain wasorganized into dimers which were 52 nm long and about0.5 nm in diameter (not shown). In summary, these resultsdemonstrate that the rod domain contains enoughinformation to form the lamin filaments and that sequencesoutside the rod domain are required for the properorganization of the lamin filaments and for their assemblyunder physiological conditions.

C. Interaction between lamin Dm0 and

chromatin.

Our previous analysis demonstrated that lamin Dm0can interact specifically with sperm chromatin (Ulitzur etal., 1992). These experiments also showed that theaddition of bacterially expressed lamin Dm0 to Drosophila

embryonic extracts that can assemble nuclei from spermchromatin resulted in increased amounts of lamin Dm0around the peripheral chromatin (Ulitzur et al., 1992). Themitotic chromosome assay that measures the associationbetween lamin and mitotic CHO chromosomes (Glass etal., 1993; Glass and Gerace, 1990) was used to analyzedomains in lamin Dm0 protein that are capable of

interaction with chromatin. When lamin Dm0 protein was

incubated for 30-60 min at 22oC with isolated mitoticchromosomes, in the presence of excess amounts of either5% BSA or 10% FCS, a strong lamin staining wasobserved following immunofluorescence analysis withanti-lamin antibodies. The staining was mostly peripheralto the chromosomes and included aggregates of lamin (notshown). These aggregations are probably due to theorganization of lamin Dm0 into polymers since the


537

aggregates were absent when mitotic chromosomes wereincubated with lamin Dm0 containing the mutation

R64>H (Zhao et al., 1996), which impairs the ability oflamin Dm0 to form filaments (Fig . 5B). The tail domain

of lamin Dm0 (amino acids 425-622) contains specific

binding site(s) to chromatin since it bound specifically tothe mitotic chromosomes (Fig . 5A ). The R64>H mutantprotein was incubated with mitotic chromosomes in thepresence of hundred fold molar excess of the isolated taildomain in order to find other possible domains in laminthat bind chromatin. As shown in Fig . 5C , staining withaffinity purified polyclonal antibodies against the roddomain of lamin Dm0 gave intensity levels that were close

to background levels. In conclusion, lamin Dm0 binds

specifically to chromatin and its binding site(s) arelocalized to its tail domain. The specific lamin sequencethat binds to chromatin, the affinity of its binding and thetarget chromosomal proteins are currently underinvestigation.

Figure 4 . Polymerization properties of the isolated roddomain of lamin Dm0. (A) Isolated rod domain protein (2

mg/ml) in buffer H was diluted 5 folds in buffer H or in 50 mMsodium citrate pH 5.5, 25 mM CaCl2. Pellet (p) and

supernatant (s) were separated by 30 min centrifugation at15,000xg, boiled in sample loading buffer and subjected toSDS-10% PAGE stained with Comassie Brillant blue. Theposition of the size markers are shown on the left of the panel.The rod domain polymerized at pH 5.5 but not at pH 7.5. (B)The pellet fraction was placed on electron microscope grid andnegatively stained with 0.75% uranyl acetate.

Figure 5 . Binding of lamin Dm0 to chromosomes. Lamin

Dm0 protein mutated in Arginine 64 (R64>H) (Zhao et al.,

1996), which is impaired in its ability to form head-to-tailpolymers (Stuurman et al., 1996; Zhao et al., 1996), boundspecifically to mitotic chromosomes (B). The tail domain oflamin Dm0 (amino acids 425-622) also bound specifically to

mitotic chromosomes (A). The tail domain of lamin Dm0 could

compete for the binding of the complete lamin Dm0 molecule to

mitotic chromosomes since addition of a hundred fold molarexcess of the tail domain could efficiently compete for thebinding of R64>H (C). DAPI staining of DNA, left panels;antibody staining, right panels. Affinity purified polyclonalantibodies against the rod domain of lamin Dm0 B,C;

monoclonal antibody 611A3A6 anti-lamin Dm0, A. This

monoclonal antibody recognize an epitope in the tail domain.The bar represents 6µm and applies to all panels.


538

Figure 6 . Inhibition of lamin Dm0 and otefin activity prevents the in vitro nuclear envelope assembly in Drosophila

embryonic extracts. Twenty microliters of embryonic extracts were preincubated for 90 min with either 100 µg polyclonal anti-lamin Dm0 antibodies (C), 100 µg polyclonal anti-otefin antibodies (D), or 100 µg of preimmune serum antibodies (IgG fraction),

(A,B ). Sperm chromatin was added and the incubation proceeded for additional 90 min. Samples from the two experimentalsystems were viewed by standard transmission electron microscope. Decondensed chromatin was enveloped with nuclearmembranes in preimmune antibodies-treated extracts (A,B) but not in anti-lamin Dm0 (C) or anti-otefin (D) antibodies-treated

extracts. The bar represents 1 µm.

D. Lamin and otefin are essential fornuclear envelope formation

To analyze lamin Dm0 and otefin function in nuclear

envelope formation, 0-6 hr old Drosophila embryo extracts,in which interphase-like nuclei can be assembled from spermchromatin (Berrios and Avilion, 1990; Crevel and Cotterill,

1991; Ulitzur and Gruenbaum, 1989), were incubated witheither 100-300 µg polyclonal anti- Drosophila lamin Dm0 or

anti-Drosophila otefin antibodies (IgG fractions). Incubationof the extract under the same conditions with 100-300 µg ofpreimmune rabbit sera (IgG fraction) or with normal rabbitIgG served as controls. No membrane assembly wasobserved when lamin Dm0 or otefin activities were inhibited


539

(Ashery Padan et al., 1997; Ulitzur et al., 1992; Ulitzur etal., 1997). Electron microscope (Fig . 6 ), light, andfluorescent microscope analyses (not shown) revealed thatwhile chromatin went through the characteristicdecondensation process, membrane vesicles did not attach toits surface, and nuclear envelope did not assemble around it(F i g . 6 C , D ). Incubation of the extract with preimmunesera, (IgG fraction), or with commercially available normalrabbit IgG fraction had no effect on nuclear assembly, thepresence of membranes around the chromatin was observed(Fig. 6A,B ). Addition of 2 µg of interphase lamin isolatedfrom Drosophila embryos to extracts that were preincubatedwith the anti-lamin Dm0 antibodies restored binding of

vesicles to chromatin (Ulitzur et al., 1997).

Acknowledgment

The isolation of Drosophila lines with mutation inlamin Dm0 gene and the analysis of these lines was

performed in collaboration with Dr. R. Falk and was partof the Ph.D. thesis of Dr. Midhat Osman. This study wassupported by the US-Israel Binational Fund (BSF), and bythe Israel Academy of Sciences.

References

Alsheimer, M., and Benavente, R. (1 9 9 6 ). Change ofkaryoskeleton during mammalian spermatogenesis:expression pattern of nuclear lamin C2 and its regulation.Exp. Ce l l . Res . 228, 181-188.

Ashery Padan, R., Weiss, A. M., Feinstein, N., andGruenbaum, Y. (1 9 9 7 a ). Distinct regions specify thetargeting of otefin to the nucleoplasmic side of the nuclearenvelope. J . B i o l . Chem . 272, 2493-2499.

Ashery-Padan, R., Ulitzur, N., Arbel, A., Goldberg, M.,Weiss, A. M., Maus, N., Fisher, P. A., and Gruenbaum, Y.(1 9 9 7 b ). Localization and posttranslationalmodifications of otefin, a protein required for vesicleattachment to chromatin, during Drosophila melanogasterdevelopment. M o l . C e l l . B i o l . 17, 4114-4123.

Ballinger, D. G. and Benzer, S. (1 9 8 9 ). Targeted genemutations in Drosophila. Proc. Nat l . Acad. S c i .USA. 86, 9402-9406.

Baricheva, E. A., Berrrios, M., Bogachev, S. S., Borisevich,I. V., Lapik, E. R., Sharakhov, I. V., Stuurman, N., and

Fisher, P. A. (1 9 9 6 ). DNA from Drosophila "-

heterochromatin binds specifically to nuclear lamins invitro and the nuclear envelope in situ. Gene . 171, 171-176.

Belmont, A. S., Zhai, Y., and Thilenius, A. (1 9 9 3 ). Lamin Bdistribution and association with peripheral chromatinrevealed by optical sectioning and electron microscopytomography. J . C e l l B i o l . 123, 1671-1685.

Benavente, R., and Krohne, G. (1 9 8 6 ). Involvement ofnuclear lamins in postmitotic reorganization of chromatinas demonstrated by microinjection of lamin antibodies. J .Ce l l B i o l . 103, 1847-1854.

Berrios, M., and Avilion, A. A. (1 9 9 0 ). Nuclear formationin Drosophila cell free system. E x p . C e l l R e s . 191,64-70.

Bossie, C. A., and Sanders, M. M. (1 9 9 3 ). A cDNA fromDrosophila melanogaster encodes a lamin C-likeintermediate filament protein. J . C e l l S c i . 104, 1263-1272.

Burke, B. (1 9 9 0 ). On the cell-free association of lamins Aand C with metaphase chromosomes. E x p . C e l l R e s .186, 169-176.

Burke, B., and Gerace, L. (1 9 8 6 ). A cell free system to studyreassembly of the nuclear envelope at the end of mitosis.Cel l 44, 639-652.

Collas, P., Courvalin, J. C., and Poccia, D. (1 9 9 6 ).Targeting of membranes to sea urchin sperm chromatin ismediated by a lamin B receptor-like integral membraneprotein. J . Ce l l . B i o l . 135, 1715-1725.

Crevel, G., and Cotterill, S. (1 9 9 1 ). DNA replication incell-free extracts from Drosophila melanogaster. EMBOJ. 10, 4361-4369.

Dabauvalle, M. C., Loos, K., and Scheer, U. (1 9 9 0 ).Identification of a soluble precursor complex essential fornuclear pore assembly in vitro. Chromosoma 100, 56-66.

Davis, L. I. (1 9 9 5 ). The nuclear pore complex. Annu.Rev . Biochem . 64, 865-896.

Dessev, G., Iovcheva, D. C., Bischoff, J. R., Beach, D., andGoldman, R. (1 9 9 1 ). A complex containing p34cdc2and cyclin B phosphorylates the nuclear lamin anddisassembles nuclei of clam oocytes in vitro. J. Ce l lB i o l . 112, 523-533.

Fields, A. P., Pettit, G. R., and May, W. S. (1 9 8 8 ).Phosphorylation of lamin B at the nuclear membrane byactivated protein kinase C. J . B i o l . C h e m . 263, 8253-8260.

Firmbach, K. I., and Stick, R. (1 9 9 5 ). Analysis of nuclearlamin isoprenylation in Xenopus oocytes: isoprenylationof lamin B3 precedes its uptake into the nucleus. J . C e l lB i o l . 129, 17-24.

Firmbach-Kraft, I., and Stick, R. (1 9 9 3 ). The role of CaaX-dependent modifications in membrane association ofXenopus nuclear lamin B3 during meiosis and the fate ofB3 in transfected mitotic cells. J. Ce l l B i o l . 123,1661-1670.

Fisher, D. Z., Chaudhary, N., and Blobel, G. (1 9 8 6 ). cDNAsequencing of nuclear lamins A and C reveals primary andsecondary structural homology to intermediate filamentproteins. Proc Natl Acad Sci USA 83, 6450-6454.

Foisner, R., and Gerace, L. (1 9 9 3 ). Integral membraneproteins of the nuclear envelope interact with lamins andchromosomes, and binding is modulated by mitoticphosphorylation. Cel l 73, 1267-1279.

Furukawa, K., and Hotta, Y. (1 9 9 3 ). cDNA cloning of a germcell specific lamin B3 from mouse spermatocytes andanalysis of its function by ectopic expression in somaticcells. EMBO J. 12, 97-106.

Furukawa, K., Pante, N., Aebi, U., and Gerace, L. (1 9 9 5 ).Cloning of a cDNA for lamina-associated polypeptide 2


540

(LAP2) and identification of regions that specify targetingto the nuclear envelope. EMBO J. 14, 1626-1636.

Gerace, L., and Foisner, R. (1 9 9 4 ). Integral membraneproteins and dynamic organization of the nuclearenvelope. Trends Cel l Biol . 4, 127-131.

Glass, C. A., Glass, J. R., Taniura, H., Hasel, K. W., Blevitt,J. M., and Gerace, L. (1 9 9 3 ). The alpha-helical roddomain of human lamins A and C contains a chromatinbinding site. EMBO J. 12, 4413-4424.

Glass, J. R., and Gerace, L. (1 9 9 0 ). Lamins A and C bind andassemble at the surface of mitotic chromosomes. J . C e l lB i o l . 111, 1047-1057.

Goldberg, M., Huihua, L., Ashery-Padan, R., Weiss, A. M.,Fisher, P. A., Gruenbaum, Y., and Wolfner, M. F.(1 9 9 7 ). Interactions between Drosophila nuclearenvelope proteins: lamin otefin and YA. submmited.

Goldberg, M., Jenkins, H., Allen, T., Whitfield, W. G., andHutchison, C. J. (1 9 9 5 ). Xenopus lamin B3 has a directrole in the assembly of a replication competent nucleus:evidence from cell-free egg extracts. J . C e l l . S c i . 108,3451-3461.

Goldman, A. E., Moir, R. D., Montag, L. M., Stewart, M., andGoldman, R. D. (1 9 9 2 ). Pathway of incorporation ofmicroinjected lamin A into the nuclear envelope. J . C e l lB i o l . 119, 725-735.

Gorlich, D., and Mattaj, I. W. (1 9 9 6 ). Nucleocytoplasmictransport. Sc ience . 271, 1513-1518.

Gruenbaum, Y., Landesman, Y., Drees, B., Bare, J. W.,Saumweber, H., Paddy, M. R., Sedat, J. W., Smith, D. E.,Benton, B. M., and Fisher, P. A. (1 9 8 8 ). Drosophilanuclear lamin precursor Dm0 is translated from either oftwo developmentally regulated mRNA species apparentlyencoded by a single gene. J . C e l l B i o l . 106, 585-596.

Harel, A., Zlotkin, E., Nainudel, E. S., Feinstein, N., Fisher,P. A., and Gruenbaum, Y. (1 9 8 9 ). Persistence of majornuclear envelope antigens in an envelope-like structureduring mitosis in Drosophila melanogaster embryos. J .Ce l l Sc i . 94, 463-470.

Harris, C. A., Andryuk, P. J., Cline, S., Chan, H. K.,Natarajan, A., Siekierka, J. J., and Goldstein, G. (1 9 9 4 ).Three distinct human thymopoietins are derived fromalternatively spliced mRNAs. Proc . Nat l . Acad . Sc i .USA 91, 6283-6287.

Harris, C. A., Andryuk, P. J., Cline, S. W., Mathew, S.,Siekierka, J. J., and Goldstein, G. (1 9 9 5 ). Structure andmapping of the human thymopoietin (TMPO) gene andrelationship of human TMPO beta to rat lamin-associatedpolypeptide 2. Genomics 28, 198-205.

Heald, R., and McKeon, F. (1 9 9 0 ). Mutations ofphosphorylation sites in lamin A that prevent nuclearlamina disassembly in mitosis. Cel l 61, 579-589.

Heitlinger, E., Peter, M., Haner, M., Lustig, A., Aebi, U., andNigg, E. A. (1 9 9 1 ). Expression of chicken lamin B2 inEscherichia coli: characterization of its structure,assembly, and molecular interactions. J . Ce l l B i o l .113, 485-95.

Hennekes, H., and Nigg, E. A. (1 9 9 4 ). The role ofisoprenylation in membrane attachment of nuclearlamins. A single point mutation prevents proteolytic

cleavage of the lamin A precursor and confers membranebinding properties. J . Ce l l Sc i . 107, 1019-1029.

Hennekes, H., Peter, M., Weber, K., and Nigg, E. A. (1 9 9 3 ).Phosphorylation on protein kinase C sites inhibitsnuclear import of lamin B2. J . C e l l B i o l . 120, 1293-1304.

Hocevar, B. A., Burns, D. J., and Fields, A. P. (1 9 9 3 ).Identification of protein kinase C (PKC) phosphorylationsites on human lamin B. Potential role of PKC in nuclearlamina structural dynamics. J . B i o l . Chem . 268,7545-7552.

Hocevar, B. A., and Fields, A. P. (1 9 9 1 ). Selectivetranslocation of beta II-protein kinase C to the nucleus ofhuman promyelocytic (HL60) leukemia cells. J . B i o l .Chem . 266, 28-33.

Hoger, T. H., Krohne, G., and Kleinschmidt, J. A. (1 9 9 1 ).Interaction of Xenopus lamins A and LII with chromatin invitro mediated by a sequence element in thecarboxyterminal domain. Exp. Ce l l Res . 197, 280-289.

Hutchison, C. J., Bridger, J. M., Cox, L. S., and Kill, I. R.(1 9 9 4 ). Weaving a pattern from disparate threads: laminfunction in nuclear assembly and DNA replication. J .Ce l l S c i . 107, 3259-3269.

Jenkins, H., Whitfield, W. G., Goldberg, M. W., Allen, T. D.,and Hutchison, C. J. (1 9 9 5 ). Evidence for the directinvolvement of lamins in the assembly of a replicationcompetent nucleus. Acta Biochim. Po l . 42, 133-143.

Judd, B. H., and Young, M. W. (1 9 7 3 ). An examination ofthe one cistron : one chromomere concept. Cold SpringHarbor Symp. Quant . Bio l . 38, 573-579.

Kaiser, K. and Goodwin, S. F. (1 9 9 0 ). "Site-selected"transposon mutagenesis of Drosophila. Proc. Nat l .Acad. Sci . USA 87, 1686-1690.

Kasahara, K., Chida, K., Tsunenaga, M., Kohno, Y., Ikuta, T.,and Kuroki, T. (1 9 9 1 ). Identification of lamin B2 as asubstrate of protein kinase C in BALB/MK-2 mousekeratinocytes. J . Bio l Chem . 266, 20018-20023.

Knipple, D. C., Fuerst, T. R., and MacIntyre, R. J. (1 9 9 1 ).Molecular cloning and analysis of the chromosomalregion 26A of Drosophila melanogaster. M o l . Gen.Genet . 226, 241-249.

Lamb, N. J., Cavadore, J. C., Labbe, J. C., Maurer, R. A., andFernandez, A. (1 9 9 1 ). Inhibition of cAMP-dependentprotein kinase plays a key role in the induction of mitosisand nuclear envelope breakdown in mammalian cells.EMBO J. 10, 1523-1533.

Lenz-Bohme, B., Wismar, J., Fuchs, S., Reifegerste, R.,Buchner, E., Betz, H., and Schmitt, B. (1 9 9 7 ).Insertional mutation of the Drosophila nuclear lamindm(0) gene results in defective nuclear envelopes,clustering of nuclear pore complexes, and accumulation ofannulate lamellae. J . Ce l l B i o l . 137, 1001-1016.

Li, H., and Roux, S. J. (1 9 9 2 ). Casein kinase II proteinkinase is bound to lamina-matrix and phosphorylateslamin-like protein in isolated pea nuclei. P r o c . N a t l .Acad. Sci . USA 89, 8434-8348.

Lin, H., Song, K., Hutcheson, G. A., Goutte, C. E., andWolfner, M. F. (1 9 9 1 ). A maternally encoded nuclearenvelope protein required for embryonic mitosis in


541

Drosophila. Cold Spring Harbor Symp. Quant.B i o l . 56, 719-727.

Liu, J., Song, K., and Wolfner, M. F. (1 9 9 5 ). Mutationalanalyses of fs(1)Ya, an essential, developmentallyregulated, nuclear envelope protein in Drosophila.Genet ics 141, 1473-1481.

Lopez, J. M., Song, K., Hirshfeld, A. B., Lin, H., andWolfner, M. F. (1 9 9 4 ). The Drosophila fs(1)Ya protein,which is needed for the first mitotic division, is in thenuclear lamina and in the envelopes of cleavage nuclei,pronuclei, and nonmitotic nuclei. D e v . B i o l . 163, 202-211.

Lopez, J. M., and Wolfner, M. F. (1 9 9 7 ). Thedevelopmentally regulated drosophila embryonic nuclearlamina protein young arrest (fs(1)ya) is capable ofassociating with chromatin. J . Ce l l Sc i . 5, 643-651.

Lourim, D., Kempf, A., and Krohne, G. (1 9 9 6 ).Characterization and quantitation of three B-type laminsin Xenopus oocytes and eggs: increase of lamin LI proteinsynthesis during meiotic maturation. J . C e l l S c i . 109,1775-1785.

Lourim, D. and Krohne, G. (1 9 9 3 ). Membrane-associatedlamins in Xenopus egg extracts: identification of twovesicle populations. J . C e l l B i o l . 123, 501-512.

Luderus, M. E., de Graaf, A. , Mattia, E., den Blaauwen, J. L.,Grande, M. A., de Jong, L., and van Driel, D. R. (1 9 9 2 ).Binding of matrix attachment regions to lamin B1. Cel l70, 949-959.

Luderus, M. E., den Blaauwen, J. L., de Smit, Oj., Compton,D. A., and van Driel, D. R. (1 9 9 4 ). Binding of matrixattachment regions to lamin polymers involves single-stranded regions and the minor groove. M o l . C e l l .B i o l . 14, 6297-6305.

Martin, L., Crimaudo, C., and Gerace, L. (1 9 9 5 ). cDNAcloning and characterization of lamina-associatedpolypeptide 1C (LAP1C), an integral protein of the innernuclear membrane. J . B io l . Chem . 270, 8822-8828.

McKeon, F. D. (1 9 8 7 ). Nuclear lamin proteins and thestructure of the nuclear envelope: where is the function?B i o e s s a y s . 7, 169-173.

McKeon, F. D. (1 9 9 1 ). Nuclear lamin proteins: domainsrequired for nuclear targeting, assembly, and cell-cycle-regulated dynamics. Curr. Opin. Cel l Biol . 3, 82-86.

Meier, J., Campbell, K. H., Ford, C. C., Stick, R., andHutchison, C. J. (1 9 9 1 ). The role of lamin LIII innuclear assembly and DNA replication, in cell-freeextracts of Xenopus eggs. J . Ce l l Sc i . 98, 271-279.

Moir, R. D., Montag, L. M., and Goldman, R. D. (1 9 9 4 ).Dynamic properties of nuclear lamins: lamin B isassociated with sites of DNA replication. J . C e l l B i o l .125, 1201-1212.

Moir, R. D., Spann, T. P., and Goldman, R. D. (1 9 9 5 ). Thedynamic properties and possible functions of nuclearlamins. In t . Rev . Cyto l . 162b, 141-182.

Newport, J. W., Wilson, K. L., and Dunphy, W. G. (1 9 9 0 ).A lamin-independent pathway for nuclear envelopeassembly. J . C e l l B i o l . 111, 2247-2259.

Nigg, E. A. (1 9 9 2). Assembly and cell cycle dynamics of thenuclear lamina. Semin . Ce l l B io l . 3, 245-253.

Nikolakaki, E., Meier, J., Simos, G., Georgatos, S. D., andGiannakouros, T. (1 9 9 7 ). Mitotic phosphorylation ofthe lamin B receptor by a serine/arginine kinase andp34(cdc2). J . B io l . Chem . 272, 6208-6213.

Oberhammer, F. A., Hochegger, K., Froschl, G.,Tiefenbacher, R., and Pavelka, M. (1 9 9 4 ). Chromatincondensation during apoptosis is accompanied bydegradation of lamin A+B, without enhanced activation ofcdc2 kinase. J . Ce l l B i o l . 126, 827-837.

Osman, M. (1 9 9 2 ). Characterization of the lamin gene andsearch for biological roles of the lamin protein inDrosophila. P h . D . t h e s i s . The Hebrew University ofJerusalem.

Osman, M., Paz, M., Landesman, Y., Fainsod, A., andGruenbaum, Y. (1 9 9 0 ). Molecular analysis of theDrosophila nuclear lamin gene. Genomics 8, 217-224.

Ottaviano, Y., and Gerace, L. (1 9 8 5 ). Phosphorylation ofthe nuclear lamins during interphase and mitosis. J .B io l . Chem . 260, 624-632.

Padan, R., Nainudel, E. S., Goitein, R., Fainsod, A., andGruenbaum, Y. (1 9 9 0 ). Isolation and characterization ofthe Drosophila nuclear envelope otefin cDNA. J . B i o l .Chem . 265, 7808-7813.

Paddy, M. R., Belmont, A. S., Saumweber, H., Agard, D. A.,and Sedat, J. W. (1 9 9 0 ). Interphase nuclear envelopelamins form a discontinuous network that interacts withonly a fraction of the chromatin in the nuclear periphery.Cel l 62, 89-106.

Peter, M., Kitten, G. T., Lehner, C. F., Vorburger, K., Bailer,S. M., Maridor, G., and Nigg, E. A. (1 9 8 9 ). Cloning andsequencing of cDNA clones encoding chicken lamins Aand B1 and comparison of the primary structures ofvertebrate A- and B-type lamins. J . M o l . B i o l . 208,393-404.

Peter, M., Nakagawa, J., Doree, M., Labbe, J. C., and Nigg, E.A. (1 9 9 0 ). In vitro disassembly of the nuclear laminaand M phase-specific phosphorylation of lamins by cdc2kinase. Cel l 61, 591-602.

Peter, M., Sanghera, J. S., Pelech, S. L., and Nigg, E. A.(1 9 9 2 ). Mitogen-activated protein kinasesphosphorylate nuclear lamins and display sequencespecificity overlapping that of mitotic protein kinasep34cdc2. Eur. J. Biochem . 205, 287-94.

Pirrotta, V. (1 9 8 8 ). Vectors for P-mediated transformationin Drosophila. B i o t e c h n o l o g y . 10, 437-456.

Pyrpasopoulou, A., Meier, J., Maison, C., Simos, G., andGeorgatos, S. D. (1 9 9 6 ). The lamin B receptor (LBR)provides essential chromatin docking sites at the nuclearenvelope. EMBO J. 15, 7108-7119.

Riemer, D., Dodemont, H., and Weber, K. (1 9 9 3 ). A nuclearlamin of the nematode Caenorhabditis elegans withunusual structural features; cDNA cloning and geneorganization. Eur. J . C e l l B i o l . 62, 214-223.

Schmidt, M., Grossmann, U., and Krohne, G. (1 9 9 5 ). Thenuclear membrane-associated honeycomb structure of theunicellular organism Amoeba proteus: on the search forhomologies with the nuclear lamina of metazoa. Eur. J.Ce l l B i o l . 67, 199-208.


542

Shoeman, R. L., and Traub, P. (1 9 9 0 ). The in vitro DNA-binding properties of purified nuclear lamin proteins andvimentin. J . B io l . Chem . 265, 9055-9061.

Simos, G., and Georgatos, S. D. (1 9 9 2 ). The inner nuclearmembrane protein p58 associates in vivo with a p58kinase and the nuclear lamins. EMBO J. 11, 4027-4036.

Simos, G., and Georgatos, S. D. (1 9 9 4 ). The lamin Breceptor-associated protein p34 shares sequencehomology and antigenic determinants with the splicingfactor 2-associated protein p32. FEBS Lett . 346, 225-228.

Simos, G., Maison, C., and Georgatos, S. D. (1 9 9 6 ).Characterization of p18, a component of the lamin Breceptor complex and a new integral membrane protein ofthe avian erythrocyte nuclear envelope. J . B i o l . C h e m .271, 12617-12625.

Smith, D. E. and Fisher, P. A. (1 9 8 9 ). Interconversion ofDrosophila nuclear lamin isoforms during oogenesis,early embryogenesis, and upon entry of cultured cells intomitosis. J . C e l l B i o l . 108, 255-265.

Smith, D. E., Gruenbaum, Y., Berrios, M., and Fisher, P. A.(1 9 8 7 ). Biosynthesis and interconversion of Drosophilanuclear lamin isoforms during normal growth and inresponse to heat shock. J . C e l l B i o l . 105, 771-790.

Smith, S., and Blobel, G. (1 9 9 3 ). The first membranespanning region of the lamin B receptor is sufficient forsorting to the inner nuclear membrane. J . C e l l B i o l .120, 631-637.

Soullam, B., and Worman, H. J. (1 9 9 3 ). The amino-terminaldomain of the lamin B receptor is a nuclear envelopetargeting signal. J . Ce l l B io l . 120, 1093-1100.

Soullam, B., and Worman, H. J. (1 9 9 5 ). Signals andstructural features involved in integral membrane proteintargeting to the inner nuclear membrane. J . C e l l B i o l .130, 15-27.

Spann, T. P., Moir, R. D., Goldman, A. E., Stick, R., andGoldman, R. D. (1 9 9 7 ). Disruption of nuclear laminorganization alters the distribution of replication factorsand inhibits DNA synthesis. J . C e l l B i o l . 136, 1201-1212.

Stick, R. (1 9 9 2 ). The gene structure of Xenopus nuclearlamin A: a model for the evolution of A-type from B-typelamins by exon shuffling. Chromosoma 101, 566-574.

Stick, R. (1 9 9 4 ). The gene structure of B-type nuclearlamins of Xenopus laevis: implications for the evolutionof the vertebrate lamin family. C h r o m o s o m e R e s . 2 ,376-382.

Stuurman, N., Maus, N., and Fisher, P. A. (1 9 9 5 ).Interphase phosphorylation of the Drosophila nuclearlamin: site-mapping using a monoclonal antibody. J .Ce l l Sc i . 108, 3137-3144.

Stuurman, N., Sasse, B., and Fisher, P. A. (1 9 9 6 ).Intermediate filament protein polymerization: molecular

analysis of Drosophila nuclear lamin head-to-tail binding.J. Struct . B i o l . 117, 1-15.

Taniura, H., Glass, C., and Gerace, L. (1 9 9 5 ). A chromatinbinding site in the tail domain of nuclear lamins thatinteracts with core histones. J . Ce l l B io l . 131, 33-44.

Theodor, L., Shoham, J., Berger, R., Gokkel, E.,Trachtenbrot, L., Simon, A. J., Brok Simon, F., Nir, U.,Ilan, E., Zevin Sonkin, D., Friedman, E., and Rechavi, G.(1 9 9 7 ). Ubiquitous expression of a cloned murinethymopoietin cDNA. Acta Haematol . 97, 153-163.

Ulitzur, N., and Gruenbaum, Y. (1 9 8 9 ). Nuclear envelopeassembly around sperm chromatin in cell-freepreparations from Drosophila embryos. FEBS Lett .259, 113-116.

Ulitzur, N., Harel, A., Feinstein, N., Goldberg, M., andGruenbaum, Y. (1 9 9 7 ). Nuclear membrane vesicletargeting to chromatin in a Drosophila embryo cell-freesystem. M o l . B i o l . C e l l . 8, 1439-1448.

Ulitzur, N., Harel, A., Feinstein, N., and Gruenbaum, Y.(1 9 9 2 ). Lamin activity is essential for nuclear envelopeassembly in a Drosophila embryo cell-free extract. J .C e l l B i o l . 119, 17-25.

Ward, G. E., and Kirschner, M. W. (1 9 9 0 ). Identification ofcell cycle-regulated phosphorylation sites on nuclearlamin C. Cel l 61, 561-577.

Wilson, K. L., and Newport, J. (1 9 8 8 ). A trypsin-sensitivereceptor on membrane vesicles is required for nuclearenvelope formation in vitro. J . Ce l l B io l . 107, 57-68.

Worman, H. J., Evans, C. D., and Blobel, G. (1 9 9 0 ). Thelamin B receptor of the nuclear envelope inner membrane:a polytopic protein with eight potential transmembranedomains. J . Ce l l B io l . 111, 1535-1542.

Worman, H. J., Yuan, J., Blobel, G., and Georgatos, S. D.(1 9 8 8 ). A lamin B receptor in the nuclear envelope.Proc. Natl . Acad. Sci . USA 85, 8531-8534.

Yang, L., Guan, T. L., and Gerace, L. (1 9 9 7 ). Integralmembrane proteins of the nuclear envelope are dispersedthroughout the endoplasmic reticulum during mitosis. J .Ce l l B i o l . 137, 1199-1210.

Ye, Q., and Worman, H. J. (1 9 9 6 ). Interaction between anintegral protein of the nuclear envelope inner membraneand humanchromodomain proteins homologous toDrosophila Hp1. J . B io l . Chem . 271, 14653-14656.

Yuan, J., Simos, G., Blobel, G., and Georgatos, S. D.(1 9 9 1 ). Binding of lamin A to polynucleosomes. J .B io l . Chem . 266, 9211-9215.

Zhao, K., Harel, A., Stuurman, N., Guedalia, D., andGruenbaum, Y. (1 9 9 6 ). Binding of matrix attachmentregions to nuclear lamin is mediated by the rod domain anddepends on the lamin polymerization state. FEBS Lett .380, 161-164.


543


Analysis of mutant p53 for MAR-DNA binding:determining the dominant-oncogenic function ofmutant p53

Katrin Will and Wolfgang Deppert

Heinrich-Pette-Institut für Experimentelle Virologie und Immunologie an der Universität Hamburg, Martinistr. 52,

D-20251 Hamburg, Germany.

__________________________________________________________________________________________________

Correspondence to: Wolfgang Deppert, Tel: +49-40-48051261, Fax: +49-40-48051117, E-mail: [email protected]

Summary

At least some mutant p53 proteins not s imply have los t the wild-type p53 specific tumor

suppressor function, but exhibit oncogenic functions on their own. Recently we showed that

binding of mutant p53 to MAR/SAR elements is an act ivity specif ic for mutant p53 and clearly

distinguishable from the previously reported DNA-binding activities of p53. Since MAR/SAR

elements are considered to be important regulatory elements for a variety of nuclear processes, the

interaction of mutant p53 with MAR/SAR elements might form the molecular basis for oncogenic

potential of mutant p53. By employing different binding assays (the target-bound DNA binding

assay, the South-western blotting technique and an adapted liquid phase binding assay), we studied

MAR/SAR binding of various p53 proteins to different MAR/SAR elements. Murine mutant p53

bound different MAR/SAR elements with an approximately 1,000-fold higher affinity than murine

wild-type p53. Analysis of MAR/SAR binding of human wild-type and mutant p53 proteins

revealed also high affinity MAR/SAR binding of several human p53 mutant proteins (175

Arg His , 273 Arg Pro) , but not o f human wi ld- type p53 , conf irming that MAR binding i s a

general property of mutant p53. By antibody interference analysis using a panel of different p53-

specific monoclonal antibodies and by deletion mutant analysis the MAR/SAR binding domain on

mutant p53 was mapped, revealing a bipartite domain consisting of the mutated core region and the

C-terminal 60 amino acids.

I. Introduction

Mutations in the p53 gene constitute the most frequent

alteration in a single gene in human cancer (Soussi et al.,

1994). Wild-type (wt) p53 is a tumor suppressor, whose

main function is to preserve the integrity of the genome as

a cell cycle checkpoint protein. Thereby p53 not only

mediates DNA damage response, growth arrest or

apoptosis by modulating cellular transcription, it also

exhibits a variety of other biochemical activities, which

are directly related to its function as major control element

in preserving the integrity of the cells' genetic

information. 90% of all mutations in the p53 gene are

single missense point mutations. These mutations are

localized mostly in the p53 core domain, which mediates

most of the biochemical activities of wild-type p53.

Consequently, these mutations serve to inactivate the

tumor suppressor functions of p53, and the expression of

mutant p53 often is considered being equivalent to a p53

"null" situation.

However, considering the many different activities exerted

by wild-type (wt) p53, it is quite astounding that a single

point mutation in the p53 molecule should totally

eliminate p53 function. Furthermore, point mutations are

a rather unique way for inactivating a tumor suppressor.

All of the other known tumor suppressors are inactivated

mostly by loss of functional gene expression, resulting

either from gene truncations or deletions, or promoter

inactivation ( Soussi et al., 1994). This, and the fact that

there is a strong selection for the maintenance of mutant

p53 expression, provoked the idea that mutant p53 not

simply is an inactivated tumor suppressor, but exerts

oncogenic functions on its own (Deppert et al., 1990,

Dittmer et al., 1993, Michalowitz et al., 1991, Levine et


544

al., 1995). This view very well corresponds to the initial

characterization of p53 as an oncogene, when all of the

available p53 cDNAs directed the expression of mutant

p53 proteins.

The notion that mutant p53 proteins can exhibit

endogenous dominant- oncogenic functions of their own

(Deppert et al., 1990, Dittmer et al., 1993, Michalowitz et

al., 1991, Levine et al., 1995, Zambetti et al., 1993) is

strongly supported by a variety of experimental

observations, perhaps most remarkably by the findings

that mutant p53 not only leads to full transformation of

the weakly Abelson murine leukemia virus transformed

L12 cells (Shaulsky et al., 1991), but also increases the

metastatic capacity of cells of a p53-deficient murine

bladder carcinoma cell line (Pohl et al., 1988). As another

example, expression of mutant p53 in p53-negative human

SAOS-2 or murine BALB/c (10/3) cells resulted in

increased proliferation rates and higher tumorigenicity of

these cells (Dittmer et al., 1993).

The molecular basis for this gain of function of mutant

p53 is still elusive. Mutant p53 has retained some of the

biochemical activities of the wt p53 protein, like non-

sequence-specific RNA or DNA binding and specific

binding to RNA with extensive secondary structures

(Mosner et al., 1995; Steinmeyer et al., 1988).

Furthermore, mutant p53 binds to various cellular proteins

which may lead to deregulation of cellular functions.

Alterations in gene expression by mutant p53, e.g.

upregulation of the mdr1 gene (Strauss et al., 1995), have

been consistently reported; however, upregulation of a

particular gene by a transfected mutant p53 was observed

in one type of cell but was totally absent in another one

(Deppert et al., 1996).

This clearly indicates that stimulation of gene

expression by mutant p53 must be due to a different

mechanism than the wt p53 specific transactivator

function, especially since most mutant p53 proteins have

lost this ability, due to loss of sequence-specific DNA

binding (Deppert et al., 1994). Interestingly, however,

additional mutations in the transactivator domain of

mutant p53 also abolish the ability of mutant p53 to

upregulate the expression of certain reporter genes (Li et

al., 1995). The most likely interpretation of these results

thus seemed that mutant p53 still is able to interact with

the cellular transcription machinery, but interacts with

DNA in a different way than wt p53. Nevertheless this

interaction must be specific, as there obviously is only a

limited number of genes which are regulated by mutant

p53.

II. DNA binding properties of murine wtand mutant p53.

To find a specific interaction of mutant p53 with DNA

which differs from that of wt p53, our laboratory has

analyzed in detail the DNA binding properties of murine

wt and mutant p53. Using !DNA as a model substrate for

a DNA which, due to its length and complexity, contains

abundant sequence elements for sequence specific

interactions, as well as structural elements for more

complex interactions of a protein with DNA, we were able

to demonstrate that highly purified mutant p53 from

MethA cells, a methylcholanthrene induced mouse tumor

cell line (DeLeo et al., 1977), binds to the 1,215 bp AluI

fragment of !DNA using a target-bound DNA-binding

assay (Figure 1).

Computer analysis of this fragment then showed that it

had both sequence and structure similarities with nuclear

matrix attachment region/scaffold attachment regions

(MAR/SAR elements). Both mutant and wild-type p53

previously were shown to interact with nuclear

substructures within the nucleus, the chromatin and the

nuclear matrix, with mutant p53 binding even more

strongly to the nuclear matrix than wt p53. Therefore the

binding of mutant p53 to a DNA fragment with homology

to MAR/SAR elements raised the possibility that mutant

p53 indeed might interact with such DNA, and prompted

us to analyze this interactions in more detail using the

target-bound DNA-binding assay specifically developed to

detect this interaction.

Its main features are that p53 is doubly

immunopurified: p53 is immunoprecipitated with the p53-

specific monoclonal antibody PAb122 and protein A-

Sepharose (PAS), then eluted by PAb122 epitope-specific

peptide (Steinmeyer et al., 1988) and reprecipitated with

another p53-specific monoclonal antibody, recognizing an

N-terminal epitope of p53 (PAb248) and PAS. This

second immune complex then is incubated with DNA

under saturating conditions for both specific and

competitor DNA, followed by rigorous washing to remove

non-bound or non-specifically bound DNA. Specifically

bound DNA and p53 are eluted separately and analysed by

SDS-PAGE (Weißker et al., 1992).

These analyses revealed that binding of mutant p53 to

the 1,215 bp AluI DNA fragment is a complex process,

involving recognition of both structural and sequence

determinants, as the fragment could not be narrowed down

to any small consensus oligonucleotide. Binding was of

high affinity (KD 10-10 M), as shown by Scatchard analysis

(Figure 1 , Weißker et al., 1992). Further studies

provided evidence that this type of complex DNA binding

indeed reflected the affinity of MethA mutant p53 to

MAR-DNA elements, as it could be extended to several

bona fide MAR-DNA elements.

Despite the usefulness of the target-bound DNA

binding assay for quantitatively assessing MAR-DNA

Will and Deppert: Interaction of mutant p53 with MAR DNA

545

Figure 1 Scatchard analysis

of the mutant p53 MAR-DNA

binding using the target-bound

DNA binding assay. (A) Equal

amounts of the doubly

immunopurified MethA p53 (1µg)

bound to PAb248 were incubated in

the target-bound DNA-binding

assay with increasing amounts of

pA1215 restricted with HindIII,

BglI and EcoRI. Lane a, 0.05 µg,

lane b, 0.1 µg, lane c, 0.2 µg, lane

d, 0.4 µg, lane e, 0.6 µg, lane f, 0.8

µg, lane g, 1 µg, lane h, 2 µg, lane

i, 4 µg, lane j, 8 µg. Lanes M,

marker DNA (1215-bp fragment,

1369-bp EcoRI-BglI pUC18

fragment, 1118-bp BglI-BglI

pUC18 fragment): M1, 10 ng, M 2,

50 ng, M3, 100 ng. DNA is marked

with an arrow. (B) Binding curve of

the Scatchard analysis of A. (C)

Linear Scatchard plot of B (from

Weißker et al., 1992)

binding by mutant p53, further comparative analyses of wt

and mutant p53 MAR-DNA binding required the

development of another assay system, as the target-bound

DNA-binding assay had intrinsic limitations. For instance,

this assay did not allow appropriate competition

experiments, as the amount of competitor DNA required

would be out of any experimentally feasible range. Thus it

was difficult to discriminate in MAR-DNA binding by wt

and mutant p53 between binding activities reflecting non-

specific DNA binding by these proteins, and their specific

MAR-DNA binding properties. Furthermore, the target-

bound DNA binding assay also did not allow a direct

comparative analysis of the MAR-DNA binding activity of

deletion fragments of mutant p53 due to the necessity of

binding the p53 proteins to a monoclonal antibody, and,

last not least, sterical interference of the affinity column

material during the binding reaction could not be excluded.

Therefore, we adapted an alternative binding assay, the

South-western blotting technique, for the analysis of

MAR-DNA binding by wt and mutant p53. After

separation of the p53 proteins by SDS polyacrylamide gel

electrophoresis, the proteins were transferred onto a

nitrocellulose membrane, and renatured on this membrane.

The membrane then was incubated with excess

radioactively labeled MAR-DNA in the presence of

unlabeled non-specific competitor DNA, washed

extensively, and the DNA bound by p53 was visualized by

autoradiography.

These analyses, using the XbaI MAR/SAR fragment

of the murine immunoglobulin heavy chain gene enhance

locus (Cockerill et al., 1987), revealed that this binding

was specific for mutant p53. The affinity of MethA p53 to

MAR-DNA was approximately 1,000-fold higher than that

of wt p53 (Müller et al., 1996). By antibody interference

analysis using a panel of different p53-specific monoclonal

antibodies (Figure 2) and deletion mutant binding studies

(Figure 3), we mapped the MAR/SAR binding region on

mutant p53 to a bipartite domain consisting of the mutated

core region and the C-terminal 60 amino acids (Müller et

al., 1996).

Thus both the non-sequence specific DNA binding

domain localized on the C-terminus of p53, as well as the

core domain of p53 mediate MAR/SAR binding

synergistically (Müller et al., 1996), thereby clearly

discriminating this activity from sequence-specific DNA-

binding by wt p53 (mediated by the core domain), and

from non-sequence specific DNA binding of both wt and

mutant p53 (mediated by the C-terminus).

An important question regarding the relevance of the

MAR-DNA binding observed with murine MethA mutant

p53 was, whether this interaction would be limited to this

special mutant p53 or exhibited also by other mutant p53


546

Figure 2 . Mapping of the MAR/SAR binding domain of

mutant p53 by antibody-interference using South-western

blotting. (A) Schematic representation of the epitopes of

various anti-p53 monoclonal antibodies on the p53 molecule.

(B) The influence of the anti-p53 monoclonal antibodies

indicated below each panel upon binding of the IgE-MAR

element by wild-type and mutant p53 was monitored in South-

western analyses. 2µg of purified wild-type and MethA mutant

p53, respectively, were analysed in the presence of a 3 x 104

fold excess of calf thymus genomic DNA. Antibody-

incubation was performed prior to DNA-binding (From Muller

et al, 1996).

proteins. Various murine mutant p53 proteins were

selected, isolated from different cellular and recombinant

sources and subsequently subjected to South-western

binding analysis using the IgE-MAR element. In

accordance with our earlier findings, all murine mutant p53

proteins in repeated experiments clearly showed high

affinity binding to the IgE-MAR also in the presence of a

Figure 3 . Mapping of the MAR/SAR binding domain of

mutant p53 by deletion analysis using South-western

blotting. (A) Schematic representation of the wild-type and

MethA mutant p53 deletion-molecules, constructed according

to the tripartite structure of the p53 molecule. (B) 1µg of the

purified wild-type and MethA mutant p53 deletion fragments

were subjected to SDS-PAGE and stained with Coommasie

blue. (C) South-western analysis of the binding of 2 µg of

each wild-type and MethA mutant p53 deletion fragment to the

IgE MAR element in the presence of a 3 x 104 fold molar

excess of non-labeled calf thymus genomic DNA. (from Müller

et al., 1996).


547

Figure 4 . Analysis of MAR-

DNA binding using the liquid phase

binding assay. Mutant p53 protein

175 (aa175 Arg"His) binds to the

IgE-MAR element with higher

affinity than wild-type p53. Equal

amounts of wild-type p53 and

mutant 175 p53 were added to

binding buffer (SWB-buffer)

including the radioactively labeled

IgE-MAR-DNA fragments and

increasing amounts of unlabeled

competitor DNA and incubated for

30 min at room temperature.

Subsequently, after incubation of

the mixture with antibody

PAb1018 and protein A-Sepharose

(PAS) the proteine-DNA complexes

were washed and the bound IgE-

MAR-DNA eluted. The eluates were

lyophilized, dissolved and

subjected to DNA-SDS-PAGE and

visualised by autoradiography

(manuscript in preparation).

high excess of non-specific competitor DNA, whereas

murine wt p53 failed to bind to this MAR element under

such conditions.

III. Studies with human p53 proteins

A disturbing result was obtained when we subjected

different human p53 proteins to MAR-DNA binding

analysis in South-western experiments. In contrast to

murine mutant p53, which reproducibly bound to MAR-

DNA in repeated experiments, we obtained quite varying

results when human mutant p53 proteins were used,

ranging from weak to no binding at all. Rather than

assuming that MAR-DNA binding is a property specific

for murine mutant p53, we considered the possibility that

the apparent lack of a reproducible MAR-DNA binding by

human mutant p53 reflected technical problems related to

structural differences between human and murine p53.

Although human and murine p53 share extensive

homologies, there are sequence and conformational

differences between these proteins, already reflected by the

fact that there are species-specific monoclonal antibodies

for human and murine p53. The most important step in

the South-western binding assay is a renaturation step,

which is very critical for reconstructing the capability for

DNA binding. Therefore, we suspected that problems in

refolding the human p53 proteins accounted for our

difficulties to unequivocally demonstrate MAR-DNA

binding for human mutant p53 proteins.

This forced us to develop an assay which did not

require renaturation procedures. The liquid-phase binding

assay fulfilled this criterion. In this assay, the desired

MAR-DNA fragments were isolated and end-labeled using

T4 polynucleotide kinase and # (-32-P) ATP by standard

procedures and subjected to the binding assays including

mutant p53 and unlabeled competitor DNAs. To avoid

interference of the column material, p53 and the DNA were

first incubated alone, and an N-terminal p53 specific

monoclonal antibody (PAb248 for murine p53 and

PAb1801 for human p53) and PAS were added later.

Finally, the DNA-Protein-PAS complexes were washed

and the bound DNA quantitatively eluted. The eluates were

lyophilized, resuspended in sample buffer and separated by

gel electrophoresis.

Application of this assay first for murine mutant p53

proteins in accordance with our earlier findings in repeated

experiments clearly showed high affinity binding to the

IgE-MAR also when high excess of non-specific DNA was

added. Murine wt p53 again failed to bind to this MAR

element under these conditions. When this assay then was

applied to human wt and mutant p53, we in repeated

experiments observed high affinity MAR-DNA binding


548

also of human mutant p53 proteins (Figure 4 , 175

Arg"His, not shown 273 Arg"Pro), but not of human

wt p53, thereby confirming the assumption that MAR-

DNA binding is a general property of mutant p53.

IV. Conclusions

Many questions remain to be resolved before MAR-

DNA binding of mutant p53 can be related to its

oncogenic activities, and before the molecular

consequences of such interactions are understood within

tumor cells. Most importantly, we must identify the

structural features within MAR-DNA which mediate the

specific interaction of mutant p53 with these DNA

elements. Although our understanding of the oncogenic

effects of mutant p53 is still at the beginning, the exciting

possibility emerges that by interfering with MAR-DNA

binding of mutant p53 it might be possible to abrogate its

oncogenic functions in the tumor cell. Considering the

strong selection for the maintenance of mutant p53

expression in tumor cells, one can hope that elimination

of mutant p53 function in tumor cells is detrimental to

tumor cell growth and will lead to its destruction.

V. Protocols

A. Target-bound p 5 3 b inding to MAR/SAR

elements

1. Immunoprecipitation of p53 (approximately 1 µg

total) from extracts of MethA cells, or from High five

insect cells infected with recombinant baculoviruses

expressing the respective mutant p53 protein using

PAb122.

2. Elution of p53 from the immune complex with a

100-fold molar excess of a PAb122 epitope-specific

peptide, followed by reprecipitation of p53 with an

antibody recognizing a different epitope on p53.

3. Target-bound DNA binding assay of the doubly

immunopurified p53: The immune complexes were washed

with binding buffer (10 mM MOPS, pH 7, 150 mM NaCl,

1 mM DTT, 0.5 mM MgCl2 ) and incubated with 8 µg of

the respective DNA (restricted plasmid DNA containing

the respective DNA fragment) in a total volume of 200 µl

of binding buffer for 1 hr at 4°C. Immune complexes were

washed three times with high-salt buffer ( 10 mM Tris-

HCl, pH 7.8, 10 mM NaCl) to separate bound and free

DNA.

4. Two-step elution of bound DNA and p53 and SDS-

PAGE: DNA fragments bound to p53 immune complexes

were quantitatively eluted with 500 µl of 100 mM

ammonium hydrogen carbonate, pH 9.5 for 45 min at

35°C. Proteins were eluted with 50 µl SDS sample buffer

and subjected to SDS-PAGE.

5. The DNA eluates were lyophilized and dissolved in

50 µl of gel loading buffer (water, 10% glycerol,

bromophenol blue). Samples of 5 µl were subjected to

SDS-PAGE and visualised by silver staining. Marker

proteins of known concentration, electrophoresed on the

same gel, served as standards.

B. South-western DNA binding assay

1. Purified protein was subjected to SDS-PAGE and

electrophoretically transferred to a nitrocellulose

membrane soaked in transfer buffer (20mM Tris-acetate

pH 8.3, 0.1% SDS, 20% 2-propanol) at 60V for 2 h.

2. Proteins were fixed on the filters with 50% 2-

propanol.

3. Filters were washed with demineralized water and

incubated 2 times for 30 min in renaturation buffer I

(50mM NaCl, 10mM Tris HCl pH 7, 2mM EDTA,

0.1mM DTT, 4M urea, 1% TritonX100) and renaturation

buffer II (as renaturation buffer I, without TritonX 100),

respectively.

4. A 30 min incubation at 30°C in renaturation

buffer III (50mM NaCl, 10mM Tris HCl pH 7.0, 1mM

EDTA, 6mM MgCl2, 0.02% BSA, 0.02% Ficoll, 0.02%

polyvinylpyrolidone, 1 µg/ml DnaK, 0.5mM DTT,

1mM ATP) strongly enhanced renaturation, but was

optional for MAR/SAR binding.

5. After renaturation the membranes were

equilibrated and saturated in DNA binding buffer with

genomic calf thymus DNA and bovine serum albumin

(SWB: 50mM NaCl, 10mM Tris-HCl pH 7.0, 1mM

EDTA, 6mM MgCl2, 0.02% BSA, 0.02% Ficoll, 0.02%

polyvinylpyrolidone, 100 µg/ml, calf thymus DNA

with an average fragment length of 10 3 to 10 4 bp

(Sigma)). Filters were incubated in a total volume of 5

ml binding buffer with 5x 106 cpm of the specific DNA

probe (MAR/SAR DNA elements), which was

radioactively labeled by primer extension.

6. After 4 h the membrane was washed 3 times with

DNA binding buffer and subjected to autoradiography.

C. Liquid phase binding assay

1. p53 was isolated from extracts of MethA cells, from

High five insect cells infected with recombinant

baculoviruses or from bacteria expressing the respective

wild-type or mutant p53 protein using antibody PAb248

columns. PAS-antibody-p53 -complexes were washed with

buffer A (30 mM KPi, pH 8.0, 50 mM KCl, 1 mM EDTA, 2

mM DTT) and eluted with buffer A including 1 M KCl. and

subsequently with buffer B (100 mM KPi, pH 12, 1M KCl,

1 mM EDTA, 2 mM DTT), followed by immediate

neutralisation with KH2PO4. Aliquots of the eluates were

subjected to SDS-PAGE and protein concentrations were

determined after Coomassie blue staining.

Will and Deppert: Interaction of mutant p53 with MAR DNA

549

2. The desired MAR-DNA fragments were isolated by

restriction digest and gel electrophoretic separation,

purified from the gel and end-labeled using T4

polynucleotide kinase and #(-32 -P) ATP by standard

procedures.

3. Afterwards equal amounts of each p53 preparation

were added to the binding buffer (SWB-buffer, see South-

western DNA binding assay) including the desired

radioactively labeled MAR-DNA fragments and unlabeled

competitor DNA and incubated for 30 min at room

temperature.

4. Subsequently, antibodies PAb248 or PAb1810 and

PAS were added and shaken for 30 min at room

temperature.

5. The DNA-protein-antibody complexes were washed

three times with SWB-buffer.

6. DNA fragments bound to p53 immune-complexes

were quantitatively eluted with 500 µl of 100 mM

ammonium hydrogen carbonate, pH 9.5 for 45 min at

35°C.

7. The eluates were lyophilized and dissolved in 20 µl

of gel loading buffer (water, 10% glycerol, bromophenol

blue). Samples are subjected to DNA-SDS-PAGE and

visualized by autoradiography.

References

Cockerill, P.N., Yuen, M.-H.and Garrard, W.T. (1 9 8 7 ). The

enhancer of the immunglubulin heavy chain locus is

flanked by presumptive chromosomal loop anchorage

elements. J . B io l . Chem. 262, 5394-5397.

Deppert, W., Buschhausen-Denker, G., Patschinsky, T.and

Steinmeyer, K. (1 9 9 0 ). Cell cycle control of p53 in

normal (3T3) or chemically transformed (Meth A) mouse

cells. II. Requirement for cell cycle progression.

Oncogene 5, 1701-1706.

Deppert, W. (1 9 9 4 ). The yin and yang of p53 in cellular

proliferation. Cancer Bio l . 5, 187-202.

Deppert, W. (1 9 9 6 ). Binding of MAR-DNA elements by

mutant p53: possible implications for its oncogenic

functions. J . Ce l l . B iochem. 62, 172-180.

DeLeo, A.B., Shiku, H., Takahashi, T., John, M.and Old, L.J.

(1 9 7 7 ). Cell surface antigens of chemically induced

sarcomas of the mouse. I. Murine leukemia virus-related

antigens and alloantigens on cultured fibroblasts and

sarcoma cells: description of a unique antigen on BALB/c

Meth A sarcoma. J. Exp. Med. 146, 720-734.

Dittmer, D., Pati, S., Zambetti, G., Chu, S., Teresky, A.K.,

Moore, M., Finlay, C.and Levine, A.J. (1 9 9 3 ). Gain of

function mutations in p53. Nature Genet. 4, 42-46.

Levine, A.J., Wu, M.C., Chang, A., Silver, A., Attiyeh, E.F.,

Lin, J.and Epstein, C.B. (1 9 9 5 ). The spectrum of

mutations at the p53 locus. Ann. NY Acad . Sc i . 768,

111-128.

Lin, J., Teresky, A.K., Levine, A.J. (1 9 9 5 ). Two critical

hydrophobic amino acids in the N-terminal domain of the

p53 protein are required for the gain of function

phenotypes of human p53 mutants. Oncogene 10,

2387-2390.

Michalovitz, D., Halevy, O.and Oren, M. (1 9 9 1 ). p53

Mutations: Gains or losses? J . C e l l . B i o c h e m . 45 ,

22-29.

Mosner, J., Mummenbrauer, T., Bauer, C., Szakiel, G.,

Grosse, F.and Deppert, W. (1 9 9 5 ). Negative feedback

regulation of wild-type p53 biosynthesis. EMBO J. 14,

4442-4449.

Müller, B.F., Paulsen, D.and Deppert, W. (1 9 9 6 ). Specific

binding of MAR/SAR DNA-elements by mutant p53.

Oncogene 12 , 1941-1952.

Pohl, J., Goldfinger, N., Radler-Pohl, A., Rotter, V.and

Schirrmacher, V. (1 9 8 8 ). p53 increases experimental

metastatic capacity of murine carcinoma cells. M o l .

C e l l . B i o l . 8, 2078-2081.

Shaulsky, G., Goldfinger, N.and Rotter, V. (1 9 9 1 ).

Alterations in tumor development in vivo mediated by

expression of wild type or mutant p53 proteins. Cancer

Res 51 , 5232-5237.

Soussi, T., Legros, Y., Lubin, R., Ory, K.and Schlichtholz, B.

(1 9 9 4 ). Multifactorial analysis of p53 alterations in

human cancer: A review. Int. J. Cancer 57 , 1-9.

Steinmeyer, K.and Deppert, W. (1 9 8 8 ). DNA binding

properties of murine p53. Oncogene 3, 501-507.

Strauss, B.E.and Haas, M. (1 9 9 5 ). The region 3' to the major

transcriptional start site of the MDR1 downstream

promoter mediates activation by a subset of mutant p53

proteins. B i o c h e m . B i o p h y s . R e s . C o m m u n 217 ,

333-340.

Weißker, S., Müller, B., Homfeld, A.and Deppert, W. (1 9 9 2 ).

Specific and complex interactions of murine p53 with

DNA. Oncogene 7, 1921-1932.

Zambetti, G.P.and Levine, A.J. (1 9 9 3 ). A comparison of the

biological activities of wild-type and mutant p53. Faseb

J . 7, 855-865.


551


Transcription-promoting genomic sites inmammalia: their elucidation and architecturalprinciples

Jürgen Bode1, Jörg Bartsch2, Teni Boulikas3, Michaela Iber1, Christian Mielke4,Dirk Schübeler1, Jost Seibler1, and Craig Benham5

1 GBF, National Research Center for Biotechnology, Genregulation und Differenzierung, D-38124 Braunschweig,

Mascheroder Weg 1, Germany.2

Entwicklungsbiologie, Universität Bielefeld, D-33501 Bielefeld, Germany.

3 Institute of Molecular Medical Sciences, 460 Page Mill Road, Palo Alto, California 94306 USA.

4 Department of Molecular Biology, University of Aarhus, DK-8000 Aarhus, Denmark.

5 The Mount Sinai Med. Center, New York/Biomathematical Sciences New York, 10029 USA.

___________________________________________________________________________________________________

Corresponding author: Jürgen Bode Tel./Fax: +49 531 6181 251/262, E-mail: [email protected]

Summary

Scaffold/matrix attached regions (S/MARs) represent a relatively novel addition to the class of cis-

acting DNA sequences in the eukaryotic genome. These elements are thought to operate via

functional contacts to the protein backbone of the nucleus. S/MARs of several kilobases are found

at the putative borders of several chromatin domains, and shorter elements with basically the same

physicochemical properties occur in close association with certain enhancers or in introns.

Accordingly, S/MARs can be situated either in nontranscribed regions or within transcription

units. Biological roles that have been assigned to them include insulating and chromatin domain

opening functions. These activities apparently are not separable, and both are compatible with the

same kind of structure.

In this contribution we present a series of recent results suggesting that S/MARs act as

topological gauges with the potential to adapt their functions to environmental stresses. We also

suggest that previously noted uncertainties regarding their activities may have arisen from

inadequacies in the methods that were used for their characterization. We discuss the application of

new, highly controlled site specific recombination methodologies that integrate single copies into

controlled genomic positions to the study of transgene and S/MAR functions in cell cultures and in

transgenic organisms.

I. Organization of the eukaryoticgenome

The eukaryotic genome is organized on at least four

levels. At the lowest level the double-stranded DNA

molecule combines with octamers of core histones,

wrapping around each in two left-handed superhelical turns.

This produces a string of nucleosomes whose spacing is

largely determined by the presence of a linker histone.

Since these basic features emerged, evidence has

accumulated that the orderly arrangement of nucleosomes

can be affected by transacting factors and structural features

of the DNA. In particular, DNA sequences with an

intrinsic curvature or bendability prefer to be

accommodated within a nucleosome, causing phased arrays

even in the absence of additional proteins (reviewed by

Wolffe, 1994b).

Bode et al: Architectural principles of S/MARs

552

The nucleosome string, also called the 10 nm filament,

shows a propensity to fold into a fiber with a diameter of

30 nm. This fiber in turn is organized into looped

domains. Early evidence suggesting this domain structure

included the observation that neither micrococcal nuclease

nor restriction enzymes are able to release from nuclei

soluble chromatin with a DNA chain length in excess of

75000 base pairs (Igó-Kemenes and Zachau, 1977). Around

the same time the existence of topologically independent

domains was established by microscopic studies of histone-

depleted metaphase chromosomes (Paulson and Laemmli,

1977) and nuclei (Cook and Brazell, 1978). These studies

revealed the presence of a supporting structure, the nuclear

scaffold or matrix, to which DNA was periodically attached

to form superhelical loops. Experiments with intercalating

agents, which at low concentrations cause an expansion

and at higher concentrations a contraction of the halo, were

explained by dye-induced relaxation of negative superhelical

loops and to their subsequent overwinding. These loops

evidently were held in a way that constrained their

topologies by preventing changes in their linking

numbers. Attached regions of DNA were subsequently

characterized by several extraction procedures (Mirkovitch

et al., 1984; Cockerill and Garrard, 1986; review:

Boulikas, 1995) and accordingly they were either termed

scaffold- or matrix-attached regions (S/MARs). Since the

same elements are recovered by various protocols, the

original distinction of SAR- and MAR-elements seems no

longer justified (Kay and Bode, 1994).

Besides the common extraction approach there are

other, supposedly milder methods aimed at detecting

attachment sequences. However, these mostly fail to

establish the existence of functional S/MARs (Jackson et

al., 1990; Eggert and Jack, 1991; Hempel and Strätling,

1996). A critical evaluation of these experiments shows

that the S/MAR elements either were not probed in their

genuine transcriptional context and/or that the topological

state of their domains had been perturbed by restriction

(Bode et al., 1996). A careful study by Ferraro et al. (1995,

1996) used cis-diamminedichloroplatinum to form

reversible crosslinks between matrix proteins and DNA in

intact cells. The use of authentic S/MAR probes for

Southwestern blotting strongly suggested that the separated

matrix proteins in fact are the interacting partners of the

S/MARs.

A. Chromatin domains and boundaryelements

Genes which are committed to transcription are

generally accessible to the action of DNaseI (Weisbrod et

al., 1982). An elevated sensitivity has been demonstrated

to extend several kbp from the transcribed region until an

area of lower accessibility is reached. It is tempting to

speculate that the boundaries between these regions could

be formed by S/MARs, which would prevent the

topological changes within an active domain from

propagating into quiescent ones (Bode et al., 1992).

Examples for which this situation has been documented

include the domains of the chicken lysozyme gene (Phi-

Van and Strätling, 1988), the human apolipoprotein gene

(Levy-Wilson and Fortier, 1989), the human ß-globin

cluster (Dillon and Grosveld, 1993), and the human

interferon-ß gene (Bode et al., 1995). These findings led to

the idea that the DNA loops defined above represent

functional units within the genome, so-called chromatin

domains.

The group of DNA “boundary elements” that have been

implicated in the functional compartmentalization of the

eukaryotic genome share certain common attributes. One

defining property is insulation: a boundary element placed

between two cis-acting elements inhibits their interactions.

When a promoter is separated from an enhancer in this

way, for example, the enhancer is no longer able to interact

with the transcription initiation complex at the promoter

(review: Corces et al., 1995). Early work found certain

sequences that exhibited insulation, although they did not

appear to be S/MARs. Examples are the scs and scs´

sequences flanking the Drosophila heat shock locus

(Kellum and Schedl, 1992, Vazquez et al., 1994).

Although scs and scs' did not behave as S/MARs, at least

in the initial assay, they have a number of properties in

common with them. Each contains a large, nuclease-

resistant core spanning a DNA segment that is very AT

rich and flanked by DNaseI hypersensitive sites. After heat

shock, both elements are primary targets for the action of

topoisomerase II, which is an abundant S/MAR-associated

protein (Laemmli et al., 1992). Other examples are a

sequence within the gypsy transposon of Drosophila

(Wolffe, 1994a), and a flanking element in the ß-globin

locus of chicken (Chung et al., 1993). The latter element

coincides with a constitutive hypersensitive site (HS4), and

blocks the action of enhancers in a way resembling scs and

scs'. Although this GC-rich sequence is similar in many

aspects to the HS5-associated sequence in the human ß-

globin locus, which is a S/MAR (Li and

Stamatoyannopoulos, 1994), it has no S/MAR activity in

vitro. If it were matrix- attached in vivo, it would have to

be by a different, but possibly related, mechanism (see

chapter V).

B. Structural factors affecting transgeneexpression levels

Several factors conspire to make the expression levels

of transgenes highly unpredictable when transfections are

performed using conventional techniques. The two most


553

important of these are positional effects and copy number

effects.

Position effect variegation (PEV) can be defined as a

position-dependent inactivation of gene expression in a

fraction of cells that generate a particular tissue. The first

and best documented instance of PEV is a chromosomal

rearrangement in Drosophila in which an allele of the

white (w+) gene is transferred to a site close to the

centromere. After this translocation its previously uniform

expression becomes “variegated,” producing patches of

pigmented and unpigmented cells in the eye. It is thought,

but still unproven, that a pericentromeric location renders

the gene susceptible to the spreading of heterochromatic

condensation. PEV has also been demonstrated in yeast and

in mammals for gene sequences within centromeres or

close to telomeres (Dobie et al., 1997).

Variegated expression can occur in transgenes as well

as in endogenous genes. As the expression level of a

transgene is highly dependent on its integration site, which

cannot be predetermined with conventional transfection

techniques, the forces leading to PEV can cause large and

unpredictable variations in expression levels. In the case of

mice, it has been reported that transgene integration into

pericentromeric regions is the most frequent inactivating

process (Festenstein et al., 1996).

However, in some remarkable cases transgenes are

expressed in an integration-site independent manner

(Chamberlain et al., 1991; Greaves et al., 1989; Aronow et

al., 1992; Palmiter and al., 1993; Schedl et al., 1993;

Thorey et al., 1993, Neznanov et al., 1996). Attempts to

identify the sequences responsible for this insulating effect

have not yet led to unambiguous candidate elements. One

simple explanation might be that these constructs are

delimited by boundary structures (Eissenberg and Elgin,

1991). Alternatively or additionally, they may contain

elements that prevent mislocalization into heterochromatic

nuclear compartments. The latter class of elements has

originally been termed “dominant control regions” (DCR;

Collis et al., 1990) and now, more commonly, “locus

control regions” (LCR; Epner et al., 1992). They are

thought to form extraordinarily stable complexes with their

coordinated promoters in a way that overcomes external

influences.

In vertebrate transfection experiments the transgenes

frequently insert in large tandem arrays. In these arrays

expression levels are not strictly correlated with copy

number, the extreme case being where expression is

completely absent. This phenomenon is termed “repeat-

induced gene silencing” (RIGS) in animals (Dobie et al.,

1997), and “cosuppression” in plant systems (Matzke and

Matzke, 1995). It is presumed that the close repetition of

sequences leads to the formation of unproductive

multiprotein complexes between transcription factors

and/or to the sequestering of these complexes in a

heterochromatic nuclear compartment.

It is obvious that studies of PEV promise insights into

the basis for heterochromatin formation and the role of

higher order chromatin and chromosome structure in gene

regulation. On the other hand, positional and copy number

effects on transgene expression levels are serious obstacles

to the straightforward application of reverse genetics. The

various ways in which transgene expression patterns are

affected by multiple copy integration events and inadvertent

occupation of certain integration sites will be discussed

further below (chapter III). Chapter IV presents a proposed

transfection strategy that does not have these problems.

C. Locus control elements

A prototype LCR has been defined upstream from the

human ß-globin genes. It contains five DNaseI

hypersensitive sites, each of which is a small region of

200-300 bp containing a high density of transcription

factor binding sites. This LCR is absolutely required for

expression, and it confers an altered chromatin structure on

a region of more than 150 kbp. (Dillon and Grosveld,

1993). The existence of an LCR has also been

demonstrated at the human CD2 locus (Festenstein et al.,

1996) and in the chicken lysozyme gene domain. In the

latter case a group of proximal regulatory sites (between -1

and -3 kb) and two distal sites (at -6.1 and -7.9 kb) are all

required for high-level, position-independent expression. It

follows that the collection of these separated functions

together constitutes the LCR (Sippel et al. 1993, Bonifer

al., 1994, 1996).

Many transgenic studies involving either a complete

LCR or its core sequences have relied on the analysis of

cells possessing more than one copy. More refined studies

are possible using retroviral vectors, which enable a single

copy of the transcription unit per cell to be integrated in a

precise way and in the absence of selection. (This approach

will be described in detail below). In retroviral transfection

experiments performed to date, the LCR and its

components act more like a classical enhancer than as an

element dominating chromatin structure. These unexpected

observations raise the question of whether an LCR can

truly confer position-independent expression when present

in one copy per cell (Novak et al. 1990).

II. The elusive roles of scaffold/matrixattached regions

S/MARs have been observed to have both boundary

functions and to act upon transcriptional rates. Due to their

locations at the putative ends of chromatin domains, they

have been considered as domain borders (Phi-Van and

Strätling, 1988; Bode and Maass, 1988; Levy-Wilson and


554

Fortier, 1989; Dillon and Grosveld, 1993). In some

systems they were found to dampen positional effects, as

expected of insulating boundary elements (Stief et al.,

1989; Phi-Van et al., 1990, 1996; Kalos and Fournier,

1995). Although S/MARs are clearly distinguishable from

enhancers, they augment transcriptional levels by a

distinct, enhancer-independent mechanism (Klehr et al.,

1991, 1992; Dietz et al., 1994; Poljak et al., 1994).

Conversely, in some systems enhancers are fully active

only when associated with S/MAR elements (reviews:

Bode et al., 1995, 1996, and section V).

A. An operational definition

Originally, S/MARs were defined using only the

protocols that led to their detection. These procedures

involve the isolation of interphase nuclei, followed by

extraction of non-matrix proteins to yield a nuclear halo. It

has been shown that the total number of loops per cell can

depend on details of the procedure used. For every

attachment existing in vivo, several new attachments may

be created in vitro as nuclei are prepared, stabilized and

lysed (Jackson et al., 1990). Moreover, the matrix-S/MAR

contacts of degraded, topologically unrestrained halos are

accessible to competing, soluble S/MAR elements (Mielke

et al., 1990; Kay and Bode, 1994, 1995; Bode et al., 1995

and references therein). These observations suggest that not

all S/MARs are constitutively associated with the nuclear

matrix in vivo (Dillon and Grosveld, 1993). Techniques

able to assess the occupation of S/MARs in vivo are

currently emerging (Ferraro et al. 1995, 1996).

Although the central biological effects of S/MAR

elements are clearly compatible with their affinity for the

nuclear matrix, others have been harder to explain. Below

we describe the more prominent properties of these

elements. These diverse properties are reconciled by recent

experiments that have been performed in our laboratory

(section IVA). In Section V we will discuss novel

transfection strategies using S/MAR elements. These have

the capacity to avoid the uncertainties arising from

positional and copy number effects, providing a

comprehensive and unambiguous analysis of the associated

functional aspects of transfected elements, including

S/MARs.

B. Transcriptional augmentation

S/MARs are a relatively new addition to the list of cis-

acting elements known to elevate transcriptional rates. In

our experience the simultaneous presence of an enhancer is

not required for this S/MAR activity (Klehr et al., 1991).

This effect, which we call 'transcriptional augmentation', is

clearly separable from prototypical enhancement since

enhancers, but not S/MARs, are active in transient assays.

But like enhancers, S/MARs act independent of orientation

and independent of distance, provided this is at least several

kilobases (Klehr et al., 1991 and section IVA). This

activity is found for minimal, viral and cellular promoters,

of both inducible and housekeeping genes. If an enhancer is

also present, enhancement and augmentation factors act

roughly multiplicatively (Bode et al., 1995).

Although enhancers are generally believed to interact

with the basal transcription machinery by looping, the

experimental data supporting such a mechanism are not

conclusive. To explain the correct choice of promoter, it

has been suggested that the initial contacts must be

checked by a tracking mechanism (Weintraub, 1993). Both

looping and tracking may be modulated by the occurrence

of S/MARs, although there are divergent views of how

this is accomplished. On one hand, S/MARs may associate

nearby enhancers to the nuclear matrix, and thereby assure

their proximity to transcribed units (Boulikas, 1995). Such

a mechanism could enable the formation of alternative

functional units, activating their respective promoters at

the matrix by a directional transfer of transcription factors.

On the other hand, by acting as domain boundaries (IA),

S/MARs could limit the effect of an enhancer to the

domain in which it occurs. This view is compatible with

the results from enhancer-blocking experiments (Bode et

al., 1995; Li and Stamatoyannopoulos, 1994). Finally,

S/MARs might serve as domain openers, as first proposed

by Zhao et al. (1993).

Although a direct correlation has been demonstrated

between the binding and augmenting activities of S/MARs

(Mielke et al., 1990; Kay and Bode, 1995; Allen et al.

1996), it is not clear whether nuclear matrix binding per se

is sufficient to augment transgene expression in stably

transfected cells (Phi-Van and Strätling 1996). If this were

the case one could use a simple in vitro binding test to

predict in vivo properties. Very recent studies show that

the general augmenting effect of S/MARs can be disrupted

by the overexpression of certain S/MAR binding proteins

(Kohwi-Shigematsu et al, 1997). A simple explanation for

this phenomenon would be the interference of such a

(possibly soluble) factor with genuine S/MAR-matrix

contacts, but more sophisticated explanations also are

possible (see the model by Scheuermann and Chen, 1989;

Zong and Scheuermann, 1995).

The biological effects of S/MAR elements commonly

are studied by transfecting a S/MAR-reporter construct and

relating its expression level to that of a S/MAR-free

control that has been transfected in parallel. In animal

cells, such an approach consistently leads to a significant

elevation of the reporter signal. Although this usually is

intuitively ascribed to cis-actions of the S/MAR, such an

effect could in principle arise in several ways:


555

(i) as an immediate effect of the S/MAR on

transcriptional initiation. This is the proposed explanation

for almost every transcriptional S/MAR effect published to

date.

( i i ) as a copy number effect. Many (but not all)

S/MAR elements tend to increase the number of integrated

copies relative to S/MAR free control (Schlake, 1994).

This has been ascribed to the recombinogenic nature of

these sequences (Sperry et al., 1989; Bode et al., 1995,

1996). It should be noted that, due to PEV-related

phenomena (Section IB), the activity of a single copy can

normally not be derived by just dividing the overall

expression by the number of integrated copies.

( i i i ) as a targeting effect. Owing to the affinity of

S/MARs for the nuclear matrix, which is thought to be the

site of active transcription, a preferential integration of

S/MAR constructs into actively transcribed regions or

active nuclear compartments could occur. S/MAR free

controls would not exhibit this preference.

C. Insulator functions

The presence of S/MARs has been shown to be

required to prevent the ectopic expression of transgenes

(Sippel et al., 1993). In one case a S/MAR is needed to

enable correct, hormonal gene regulation (McKnight et al.,

1992, 1996). These and other results from a range of

insulation experiments strongly implicate S/MARs in

defining the boundaries of autonomously regulated

chromatin domains.

S/MARs have been studied in both animal and plant

systems. As a rule, transcriptional augmentation by

S/MARs is most easily seen in mammalian cells, while in

plants their insulator functions are more frequently

observed (Bode et al., 1995). We ascribe this difference to

the fact that most plant studies are based on T-DNA

vectors which, when integrated into genomic loci, yield

expression levels high enough to cope with applied

selection pressure (Dietz et al, 1994). In this case the

maximum attainable level of gene expression does not

depend on the presence of S/MARs, whose most easily

observed function then becomes to stabilize an elevated

(but not extreme) expression level shielded from negative

and positive influences of the surroundings (Dietz et al.,

1996, Bode et al., 1996). This would agree with the

insulating or bordering function suggested by Mlynarova et

al. (1995).

After the enhancer-blocking assay (IA), described

previously, copy number dependence is by far the most

common principle used to detect and assay insulator

functions. However, there are major difficulties inherent in

this method (demonstrated by Poljak et al., 1994; Schlake,

1994). These are due, at least in part, to the fact that

multiple copies often do not integrate at separate genomic

locations, but rather as a single tandem array at a site

which is unique for each cell. This induces many

unexpected interactions, rearrangements and cellular

'defense mechanisms' (Mehtali et al., 1990; Kricker et al.,

1992; Dorer and Henikoff, 1994; Kalos and Fournier,

1995; Dorer, 1997). For this reason there frequently is not

a linear correlation between copy number and gene

transcription, even in the case of low copy numbers (see IB

and IIIC).

D. Intronic S/MARs

S/MARs cannot simply be considered to represent

static delimiters of functional domains. This is

demonstrated by the detection of S/MARs within introns.

Examples of single genes that are apparently divided over

two domains include the genomic sequences encoding

hamster DHFR (Käs and Chasin, 1987), human topo I

(Romig et al., 1992), human interleukin-2 (Artelt and

Bode, unpublished) the mouse immunoglobulin !- and

µ-chains (Cockerill and Garrard, 1986, Cockerill et al.,

1987) and a light-inducible plant gene (Stockhaus et al.,

1987, Mielke et al., 1990). Studies on MPC-11

plasmacytoma cells have shown that 9% of poly(A)

mRNA arises from the !-locus, whose transcription

must be completed on average once every 3.2 s (Cockerill

and Garrard, 1986).

Since by definition intronic S/MARs are transcribed,

and since they do not impede passage of RNA polymerase

II, their occupation must be regulated. Functional analyses

have recently revealed complex roles for S/MARs in gene

expression, which can only be appreciated after structural

analyses of these composite elements (section VB).

E. Current uses of S/MAR elements

Retroviral delivery systems enable the efficient transfer

and expression of transgenes in primary cells. Sometimes,

a pitfall of these methods has been a continuous

downregulation of the transgene (but see IVA). Because

S/MAR elements augment transcription and stabilize its

level over extended periods of time (Bode et al., 1995),

they could be used to construct a new generation of

expression vectors. The mechanism of retroviral replication

enables one to establish a minidomain by simply

introducing a S/MAR element into the 3´- LTR (Schübeler

et al., 1996 and Figure 3 ). In this minidomain the

expression unit is flanked by two S/MARs, which can

stabilize transgene expression. A number of current

pharmaceutical developments are exploiting this approach.

Although short and apparently functional S/MAR

elements can be constructed by oligomerizing certain sub-

S/MAR motifs (core unwinding elements, see Mielke et


556

al., 1990, Bode et al., 1992 and VB), the practical use of

such synthetic elements for retroviral transfection is

limited by the intrinsic genomic instability within and

between the multimers. One can avoid this problem by

using minidomain bordering elements from different

sources (Bode et al., 1995, Dietz et al., 1994).

A whey acidic protein (WAP) transgene was found to

be active in just 1 out of 17 lines of transgenic mice,

demonstrating copy-number independent (position

dependent) expression. After ligating S/MARs to the

transgene, 7 of 9 lines exhibited transgene activity, and

correct hormonal regulation occurred in the majority of

cases (McKnight et al., 1992, 1996). These experiments

show that S/MARs are able to prevent ectopic expression

(Sippel et al., 1993).

S/MARs have found widespread use in the

manipulation of plant cells, callus cultures and complete

plants. Although increased and position-independent

transcription has been described by various authors, these

phenomena are only linked in some cases (Schöffl et al.,

1993, Mlynarova et al., 1994, van der Geest et al., 1994).

Examples are known of plant cells transformed by

Agrobacterium in which the use of S/MARs improves

copy-number dependent (Breyne et al., 1992) or position-

independent (Dietz et al., 1996) expression without

affecting transcriptional levels (Breyne et al., 1992). In

other cases where microprojectile bombardment was used,

transcription was dramatically raised independent of copy

number (Allen et al., 1993, 1996). These differences may,

in part, be due to the gene transfer technique used. The first

method tends to target transcriptionally competent sites,

while random positions are hit in the second one (Dietz et

al., 1994).

III. Problems associated with reversegenetics

The level of in vivo expression of a transfected gene

may be quite different from that which occurs when the

same gene is in its natural context. Although transient

assays in cultured cells have been used in the past to

characterize tissue-specific enhancers, their potential is

limited because the actions of these and other cis-acting

elements depend sensitively on chromosomal context.

Proximity of enhancers to S/MAR elements provides a

paradigm example of context-specific behavior (reviews:

Bode et al., 1995, 1996). These observations underscore

the difficulty of establishing a dependable approach to the

study of gene function. The analysis of gene expression

must be performed in a native chromatin context.

However, this approach is complicated by a number of

problems arising from the following facts (section IIIA-C).

A. Need for a selection marker

Since only a small percentage of cells incorporate

foreign DNA after transfection, selection for a drug

resistance marker is required to isolate those cells that

harbor the construct. This marker has to be expressed at a

threshold level, which creates a selection bias against low

producers (Blasquez et al., 1992). Positive selection

markers are constitutively expressed, which can seriously

interfere with the transcription even of remote genes. An

extreme example has been reported of a gene that became

completely deregulated by a selection marker which was

placed 50 kbp away (Fiering et al., 1993). Further, some

prokaryotic vector sequences have been found to inhibit

(Palmiter and Brinster, 1986) or to promote gene

expression (Seibler and Bode, 1997).

B. Integration occurs at random

Since there is no evidence for site-directed insertion, the

generation of random chromosomal breaks is thought to be

the rate-limiting step of the prevailing integration

mechanism. This would explain the predominance of

unique integration sites in mammalian cells and the

observation that certain (but not all; see Mielke et al.,

1990) cell types prefer linearized over circular templates for

integration (Palmiter and Brinster, 1986). Random

integration can subject transgene expression to unwanted

local influences (position effects) resulting from their

proximity to regulatory elements or heterochromatic

regions (IB).

C. Multiple copy integration

If DNA is introduced by standard transfection

techniques or by microinjection, multiple gene copies are

usually integrated at a single site. This is probably due to

homologous recombination events occurring among the

transfected molecules (Phi-Van and Strätling, 1996). There

are several ways in which this tandem arrangement can

alter the effect of an element from what it would be as a

single copy. One usually cannot determine how many

copies of a gene are functional templates, or whether

different copies perform different functions (Oancea et al.,

1997). Tandem repetition of promoter elements could

trigger the formation of multiprotein complexes between

transcription factors, with largely unpredictable

consequences (IB). It also may lead to a more than

proportional effect of a cis-acting regulatory sequence, such

as a S/MAR (Stief et al., 1989). More typical are shutoff

processes occurring with time. These have been associated

with methylation followed by mutagenesis (Mehtali et al.,

1990; Dorer, 1997) and/or a genomic instability resulting

from the continuous loss of members of the array

(Palmiter et al., 1982; Weidle et al., 1988). As a


557

consequence, expression levels cannot be expected to be proportional to copy number (Figure 1).

Figure 1. Inadequacy of

convent ional gene transfer

techniques .

Conventional transfection

techniques lead to multi-copy

integration events, usually at a single

site and in tandem, head-to-tail

orientation. Whether all members of a

multigene complex are transcibed at the

same rate is an unsolved question

While a tandem head-to-tail integration at a given site

is considered typical, more complicated integration forms

were recently demonstrated for single S/MAR constructs

(Phi-Van and Strätling, 1996). Most of the copies were

colocalized as usual, but they clearly differed regarding their

relative orientation: whereas head-to-head (hh) and head-to-

tail (ht) were preferred, hh plus tt and ht plus tt tandems

were strongly disfavored and exclusive tt integrations were

lacking altogether. It may be speculated in this case that

the unequal ends, which contained a S/MAR at their 5' end

and a unique vector sequence at their 3' terminus, were

responsible for this effect. This nonrandom distribution

could result from preintegration ligations of two

molecules. Alternatively, the close juxtaposition of two

identical copies could induce inversions. The

recombinogenic nature of S/MAR elements (Sperry et al.

1989) might add to the complications associated with

conventional gene transfer techniques.

In conclusion, with current transformation methods,

insertion of DNA into the genomes occurs at random, and

in the case of plant systems in many instances at multiple

sites. The associated position effects, copy number

differences and multigene interactions can make

conventional gene expression experiments difficult to

interpret.

1. An immediate solution: gene transfer byelectroporation can be optimized

Classical gene transfer experiments are based on the

cotransfer of a reporter and a selector construct which are

coprecipitated by Ca++-phosphate and taken up by

endocytosis (transfection). As an alternative, DNA can be

transferred from solution by transiently permeabilizing the

cells in an electric field (electroporation). This technique is

most efficient if the reporter and the selector gene are


558

physically linked in a single construct. For some cell lines

the electrical parameters and cell survival rates can be

optimized in a way that yields predominantly single copy

integration events (Mielke et al., 1990, 1996; J. Bartsch,

unpublished). This technique eliminates the unpredictable

effects of multicopy integration.

This method has been used to compare expression

levels of S/MAR-constructs to those of S/MAR-free

controls. Representative data are found in Figure 2 . The

average level of expression from a S/MAR-free luciferase

control (Lu: 85 700 light units) was increased eightfold if

an upstream S/MAR element was present (E-Lu: 698 000

light units), and 26-fold if it was transfected as a

minidomain with upstream and downstream S/MAR

elements attached (E-Lu-W: 2 253 000 light units). These

results are similar to those previously reported (Klehr et

al., 1991, 1992; Bode et al., 1995, 1996; Phi-Van et al.,

1990; Poljak et al., 1994).

This analysis has been extended to individual clones to

derive information about the insulator function of S/MARs

that is difficult to obtain by either the enhancer-blocking

assay (IA) or the copy number dependence of expression

(section IIC). Since expression levels vary by three orders

of magnitude, they have been plotted in Figure 2 on a

logarithmic scale. The largest inter-clone variation of

expression was 340 fold, found for the S/MAR-free

control. A single S/MAR element reduces the variation to

4.6 fold. This E-Lu constructs has a single insulator at its

5' end and could, in principle, experience position effects

acting through its “open” 3' terminus. If the domain is

closed by a second S/MAR element (E-Lu-W) the variation

between clones is reduced to a factor of two. These data are

in general agreement with a recent study by Kalos and

Fournier (1995), who demonstrated that in the presence of

both apolipoprotein B and S/MARs the expression of a

test construct was increased from the detection limit to at

least 200 fold higher levels. At the same time the variation

between the clones decreased from >10-fold to <3-fold. For

technical reasons these studies were restricted to a small

number of clones, and apparently they have not been

generally accepted (see Wang et al., 1996).

We believe that the combined results of our studies

strongly support the notion that S/MARs can enable

position independent expression of transfected genes.

Whether this is due to insulator function, as currently

believed, is not clear. Ours is the first study to demonstrate

the dramatic influence even of a single S/MAR, which

would also be compatible with a dominant (domain opener)

effect in case of the E-Lu construct. This information

cannot be derived from any assay based on the presence of

several copies (Figure 1), since in the common case of

tandem integration each individual member but one will be

an effective double-S/MAR construct (...E-Lu-E-Lu...).

Figure 2 . Insulator funct ion of

S/MARs demonstrated at the s ingle

c o p y l e v e l

A luciferase-neomycin resistance

construct was introduced into CHO cells

either as a S/MAR-free control (Lu), flanked

by an upstream S/MAR element (E-Lu) or by

as a minidomain flanked by an upstream and

a downstream S/MAR (E-Lu-W).

Electroporation of these constructs resulted

in 10 clones (Lu), 20 clones (E-Lu) and 90

clones (E-Lu-W), resp.. 9-10 clones with a

single integrated copy of each construct

(solid bars) and 2-3 clones with 2-3 copies

(hatched bars) were selected for

determinations of the luciferase expression

level. Average luciferase levels (light units)

were 86 E3 +/-124% (Lu), 590E3 +/-39% (E-

Lu), 2273E3+/-14% (E-Lu-W).


559

Figure 3 . Use of retroviral vectors

Because those retroviral genes that are necessary

for retrovirus replication have been replaced by a

reporter gene (secretory alkaline phosphatase, SEAP)

and a selection marker (fusion gene from

Hygromycin-B-phosphotransferase and HSV-

Thymidine kinase, HygTK), these functions have to

be provided in trans by a packaging cell line. These

helper cells produce infectious retroviral particles

which can be used to transfer the construct into the

recipient cell. After reverse transcription of viral

RNA, the transgene is stably linked with the genome

(provirus state). Promoter and terminator functions

reside in the long terminal repeats (LTRs). Moderate

manipulation of the LTRs is compatible with their

function and the introduction of S/MAR elements

(Mielke et al., 1996) and Flp-recognition target sites

(FRTs) has been described (Schübeler and Bode,

1997; see also section IVA).

Interestingly, the properties documented in our study

appear to be independent of the source of the S/MAR.

Elements of human origin (E) behave in our system like

their counterparts from plants (W), provided that they both

show a corresponding affinity for the scaffold in the in

vitro test system (cf. also Dietz et al., 1994). To date it

has not been possible to separate the insulation and

augmentation properties of S/MARs (Phi-Van and

Strätling, 1996) which both require a threshold length of

DNA in excess of about 300 bp. Together, these results

suggest that the basic functions of insulation and

transcriptional augmentation are widely conserved, and are

based on structural features which are recognized by

ubiquitous proteins rather than by specialized factors.

We note that five clones in our experimental series

contained more than a single copy (hatched bars in Figure

2 ). Since their expression is in the same range found for

the single copy integrates, a standard copy-number

dependence test (IIC) would have failed to reveal any

evidence for an insulator function in this system. This re-

emphasizes the fact that any test of sophisticated S/MAR

functions must use single copy integrants.

IV. Advanced approaches: A progressreport

While the electroporation method has effectively

eliminated the detrimental effects of multicopy

integrations, the other critical problems (numbered (i) to

(iii) in section IIB) still need to be addressed. The

following studies describe recent developments in this

direction. They allow the integration of single, intact

copies, in some cases without the cotransfer of vector

sequences. They also are aimed at preventing any

integration bias which may arise from the biochemical

properties of S/MAR elements (criterion 3 in IIB).

A. What retroviruses can tell us

We used a retroviral vector infection procedure as it

facilitates the introduction of single copies with strictly

defined ends, the long terminal repeats (LTRs). For these

studies we have constructed an expression cassette

consisting of a reporter gene (SEAP, secretory alkaline

phosphatase) and a selector gene (PAC, puromycin N-

acetyltransferase). Both members of this bi-cistronic

cassette are synchronously expressed owing to the fact that

an internal ribosome entry site (IRES) enables the

translation of the second cistron. To obtain an infectious

retrovirus particle, accessory functions have to be provided

in trans. This is done by first transfecting the construct

into a helper cell line (Psi2). The helper cells contribute

the gag, env and pol functions, which package the viral

mRNA and secrete the virus into the medium. These

viruses can be harvested and used for an infection of the

ultimate target cells. There copies of the viral mRNA are

reverse-transcribed into DNA. During this process the

sequence of the 3'-LTR is copied to form an identical 5'-

LTR (Figure 3 ). Elements cloned into the 3'-LTR

(marked by a half arrow) will then appear also in the 5'-

terminus. This strategy has been used both to introduce a


560

minidomain (S/MARs at both ends, see section IIE), and

to generate two Flp recombinase target sites which are

recognized and recombined by the Flp enzyme (Schübeler

et al, 1997 and section IVB).

Before making extensive use of these possibilities, we

decided to thoroughly characterize the sites of provirus

integration. As an initial step, the identical construct was

introduced into target cell by standard transfection,

electroporation and infection techniques. Expression levels

of the SEAP reporter were compared and related to the

number of integrated copies (Figure 4). While the overall

expression level turned out to be mostly independent of the

procedure used for transferring the vector, the per-copy

level for transfection is clearly inferior to electroporation

and electroporation in turn is inferior to infection. Only

infection guarantees the incorporation of single copies

which remain stable over extended time intervals

(Schübeler et al, 1997 and unpublished). This indicates that

there is no gradual inactivation due to methylation. If such

an event were to occur, selection pressure could be applied

to eliminate it.

1. Anatomy of retroviral integration sites

It is the prevailing view that transcriptionally active

genomic regions and regions associated with DNase I

hypersensitive sites are preferred by the integration

machinery of a retrovirus (Rohdewohld et al., 1987). We

wanted to study the architecture of these integration sites in

order to obtain information about the factors which mediate

a consistently high and stable level of expression.

Until recently, most methods for isolating proviral

flanking sequences involved plasmid rescue. This approach

led to the identification of a small number of highly

preferred integration targets (Shih et al., 1988). Some of

these results could not be confirmed by subsequent PCR-

based techniques that were developed to study integration

without prior selection by molecular cloning. This has led

to a critical re-evaluation of the prevailing views (Withers-

Ward et al., 1994). We have applied retroviral vectors in

conjunction with inverse PCR techniques to reconstruct a

number of these sites for further characterization. As in

many previous studies, the recovered integration sites

conformed to no obvious consensus sequence. This

suggests that the site of integration into the host genome

may be determined by other factors, such as DNA

secondary structure and/or host proteins. While the

retroviral integrase performs the central cutting and joining

steps as part of a 160 S nucleoprotein complex, the final

resolution steps require host functions which may lead to a

further selection among a number of initial target sites.

Figure 4. Transcriptional

propert ies of transgenes

introduced by three different

gene transfer techniques.

A single construct (pM5sepa)

was introduced into NIH3T3 murine

embryo fibroblasts by three gene transfer

techniques. pM5sepa contains a reporter

(SEAP, secretory alkaline phosphatase)

and a selector gene (PAC, puromycin N-

acetyltransferase) linked by an internal

ribosomal binding site (IRES) to enable

coupled expression of both cistrons.

Solid bars represent total expression

from an entire pool of PAC-resistant

clones. Hatched bars refer to expression

levels divided by copy numbers.


561

. . . . S /MAR-type DNA:

-recombinogenic

-supporting transcription

. . . . bent DNA

-no positioned nucleosomes

(Boulikas, unpublished)

-backbone strain retrievable for

unpairing (Ramstein & Lavery, 1988)

Figure 5 . Architecture o f h i g h l y expres s ing genomic s i t e s which are the preferred targets for

retroviral integrat ion . Integration occurs into S/MAR type DNA which is flanked by bent segments. Bending prevents the

association of nucleosomes as to keep these sites accessible. It is therefore different from the type of curvature which

accommodates nucleosomes (Mielke et al., 1996).

Remarkably, all investigated examples conformed to a

unique pattern (Mielke et al., 1996), summarized in

Figure 5 . All integration events occurred into

scaffold/matrix attached regions (S/MAR-elements) which

were flanked by DNA that was bent in a way that

discourages nucleosome assembly (Boulikas et al.,

unpublished). These S/MARs belong to a novel class that

does not conform to the AT-rich prototype. On the other

hand they exhibit most of the in vitro and in vivo

properties of S/MARs. Their binding is competed by

ssDNA, and they induce significant transcriptional

augmentation (as opposed to enhancement) if they are

investigated as parts of mammalian expression vectors. We

have suggested that the entire insertion process might be

guided by the nuclear matrix, which also could provides the

enzymatic functions for the final steps of the integration

process (Mielke et al., 1996). Integration into S/MARs

may also be ascribed to their recombinogenic properties

while the selection of non AT-rich subtypes seems to be

directed by the mechanism of integration. In summary,

retrovirus integration occurs at sites which are available

due to their lack of nucleosomes, and which show

structural and functional features reminiscent of S/MARs

(see analyses under VB/C).

2. S/MAR effects on transcriptional initiationand elongation

At this stage of our investigation it was unclear if

retroviral integration sites were appropriate for the study of

S/MAR constructs, as S/MAR functions also resided in

their immediate vicinity. But these tools permit integration

into a subclass of genomic sites whose selection is dictated

by the retroviral integration machinery, not by the presence

or absence of the S/MARs within the vectors (criterion 3

under IIB).

We inserted an 800 bp S/MAR element from the

upstream border of the human interferon-ß domain into the

vector shown in Figure 4 . This element was cloned into

various positions, both within and outside a transcribed

region of 4.3 kb. Insertion into the 3'-LTR yielded a

minidomain (two flanking S/MAR elements) according to

the scheme in Figure 3 . This study revealed a range of

unexpected S/MAR effects that were obscured when the

same constructs were introduced by transfection (Schübeler

et al., 1996). The most striking observation was that at a

distance of about 4 kb, the S/MAR supported

transcriptional initiation whereas at distances below 2.5 kb

transcription was essentially shut off. Controls proved the

functionality of all constructs in the transient expression

phase, and ruled out any influence of S/MAR position on

transcript stability. Moreover, no pausing or premature

termination was observed within these elements.

One interpretation of these results is presented in

Figure 6 . This assumes that S/MARs are kept in a

single-stranded state by association with ssDNA binding

proteins. This "unwound" structure is able to facilitate the

progression of an approaching polymerase if the buildup of

positive supercoils is sufficient for breaking the contacts

between the single-strands and the binding proteins

(transition D2 -> D2.1). If positive superhelicity is minor,


562

Figure 6 . Scaffold/Matrix-

attached reg ions act upon

transcript ion in a contex t -

dependent manner.

S/MARs can be trapped in vivo as

a single-stranded structure by

chloroacetaldehyde (CAA). It is

hypothesized that single strands are

kept apart by ssDNA binding proteins

(triangles). Unconstrained positive

supercoils arising from a S/MAR-

scaffold association are ultimately

removed by the action of

topoisomerases (B, C). If the S/MAR

is located immediately downstream

from a transcription initiation site,

the polymerase cannot pass the

attachment point as the buildup of

positive superhelicity is insufficient

to rupture S/MAR-scaffold contacts

(D1). If it is situated further upstream

the contacts can be broken by the

approaching polymerase and positive

superhelicity can be relaxed as it is

compensated be the now

unconstrained unwound DNA structure

of the S/MAR segment

(complementarity of plectonemic and

paranemic structures, see Yagil,

1991). Other polymerases can initiate

in the wake of the first one.

binding of the ssDNA is strong enough to prevent a

rotation of the DNA helix about its axis, which inhibits

the progress of the polymerase(state D1). This model

explains the bordering and insulator functions of a

S/MAR, which will prevail if the element is long enough.

It also explains the observation that, in principle, a

polymerase can progress through a short S/MAR, and even

benefit from the fact that S/MARs are repositories of

unwound DNA. This may explain the essential fact that

transcriptional augmentation is only found after anchoring

the construct in the genome of the host cell. It is

unnecessary to postulate the existence of different S/MAR

types to perform bordering and augmenting functions. In

our experiments both functions are provided by an

intermediate-size S/MAR isolated from the center of an

extended putative domain border.

B. A novel concept: genomic referenceintegration sites

The study of cis-acting regulatory elements is often

confounded by the variability of gene expression among

independent transformants. This variability is ascribed to

chromosomal position effects (PEV, see IB) at the sites of

transgene integration. Inserting single copy test constructs

into the same genomic target would control these effects

and facilitate valid comparisons of expression levels.

Targeting a given site via homologous recombination,

though successful in fungal and some animal systems, is

not always practical because it occurs at very low

frequencies compared with the high background of

illegitimate recombination events.


563

Figure 7 . React ions cata lyzed by s i te - spec i f i c recombinases : Exc is ion/ integrat ion and invers ion

Excision is the consequence of removing a stretch of DNA which is flanked by two equally-oriented FRT sites; the reaction is

mediated by a crossover between these sites. In principle, this reaction is reversible in the sense that a circular vector can be

accommodated at a genomic site carrying an FRT tag. The native role of Flp recombinase in yeast is the inversion of the origin of

replication region on the 2 plasmid which is localized between two inversely-oriented FRTs (right-hand part).

A full Flp-recognition target site consists of three 13 bp repeats and an 8 bp spacer within 48 bp of DNA (bottom). Each of the

13 bp repeats represents an Flp binding element (FBE) and the spacer is the region in which single strand cuts (vertical arrows) are

introduced in preparation of the crossover and resolution steps. There is no physical contact of the recombinase to the spacer

which determines the polarity and identity of the site. Hence, spacer mutants will be recombined with an identical FRT site but not

with a wild type FRT (Schlake and Bode, 1994). For convenience, the FRT-site is symbolized by an half arrow.

Flp recombinase from the 2 plasmid of

Saccharomyces cerevisiae can be introduced into

mammalian cells to perform site-specific integration

reactions (O´Gorman et al., 1991). This enzyme excises

any piece of DNA that is flanked by two Flp-recognition

target (FRT-) sites of identical orientation (Walters et al.,

1996). Even more important in the present context, it also

performs the reverse reaction of integrating an FRT-

labelled circular vector into an FRT-tag placed in the

genome (Figure 7 ; cf. Schlake and Bode, 1994 and

references therein). This enables site-specific integration

while avoiding the problems of earlier methods. Related

approaches are being developed on the basis of the

Cre/loxP1system of bacteriophage P1 which is mostly

used for excision-type reactions and with an increasing

number of other members of the Int recombinase family

(reviews: Kilby et al., 1993; Sauer, 1994).

Suitable sites for integration can be prepared by

introducing an FRT-tagged construct into an endogenous

locus with known properties. This is only possible in

systems permissive of homologous recombination, and for

loci that are redundant and constitutively expressed such as

the histone locus in mice (Wigley et al., 1994).

Alternatively, the construct can be transferred by

electroporation (section IIIC1). Among the multiple clones

that result, those integrants will be chosen which mediate a

high and consistent expression in the absence of continued

selection pressure. This will guarantee that the respective


564

site does not become inactivated by heterochromatization,

by DNA methylation, or by being situated in a locus that

is genetically unstable. Both procedures are initially

laborious, but screening has only to be performed once. As

an additional advantage of the approach, if an FRT-tagged

construct tends to form head-to-tail multimers, these

concatemers will be reduced by a continuous action of the

recombinase which (according to Figure 7 ) will excise

pieces of DNA that are flanked by equally-oriented sites

(Lakso et al., 1996). If the individual vector contains just

one of these sites, ideally the excision will continue until a

single, intact copy is left.

For a convenient characterization of expression

parameters, it is advantageous first to introduce a reporter

gene which can be monitored easily. Since integration can

be reversed by a second pulse of Flp activity, the

remaining FRT site is then open for other rounds of

integration during which any gene of interest can be

directed to the pre-defined locus. The general validity of

this concept was demonstrated using the Cre/loxP system

(Fukushige and Sauer, 1992). However, the straightforward

application of the scheme in Figure 7 faces a number of

problems which will be discussed after describing some

molecular features of site specific recombinases,

exemplified by Flp/FRT.

1. Mechanism of recombination: design andfunction of FRT sites

The amino acid sequence of Flp bears no resemblance

to any known DNA binding motif. The full 48 bp FRT

site consists of three individual 13 bp Flp binding

elements (FBE a-c), two of which form an inverted repeat

around an 8 bp spacer. The third 13 bp element (c)

represents a direct repeat and is separated from b by a single

base pair. While most reactions are also possible with a

minimal 34 bp site consisting of the inverted repeat and

the spacer, the integration reaction appears to benefit from

the presence of the extra FBE (Lyznik et al., 1993).

Association of Flp with its site causes bending of two

types. While each Flp monomer introduces an individual

bend of 40o; a larger bend (>140o ) is caused by the

interaction of the two Flp monomers across the spacer.

This is a consequence of strong interprotomer contacts

which occur despite the fact that the protein binding sites

lie on opposite faces of the DNA. The Flp monomer

bound by the third element does not participate in these

strong cooperative interactions, which require flexibility of

the spacer. This spacer becomes single stranded upon Flp

binding (Kimball et al., 1995).

Strand cutting can be initiated when one individual

FRT site is occupied by two Flp monomers. This “trans-

horizontal” cleavage mode does not require the presence of

a synaptic complex. Each monomer of the recombinase has

only a partial active site and contributes to the formation

of a full active center by donating the catalytic tyrosine to

the Arg-His-Arg cleft of the partner that is bound across

from it on the other side of the spacer. This mechanistic

property of Flp guarantees that during the two-step strand

transfer of a complete recombination reaction, the

activation of one pair of active sites is coupled to the

disassembly of the other (Kwon et al., 1997).

Synapsis is mediated by protein-protein interactions

between the bound recombinase molecules. In this way the

paired strand cleavage steps become coordinated (Figure

8 ). This underlines the importance of the synaptic

complex, which channels the chemistry of strand breakages

so that recombination, not self-healing, is the final result.

After the initial strand breaks, strand transfer, and ligation,

a Holliday intermediate is formed. Homology permits the

Holliday junction to undergo branch migration and

isomerization, during which the crossover strands and the

helical strands switch functions. This isomerization

probably results from a preference for one pair of stacking

isomers over another (Li and al., 1997).

2. The homology checkpoints

A central requirement for recombinations catalyzed by

members of the INT family is an absolute homology

between the partner substrates in a strand-exchange

reaction. This implies a DNA-DNA interaction at some

point in the reaction. Recent evidence suggests that

homology is not checked before strand cleavage, so the

first strand transfer can occur in spite of one or more

mismatches. Subsequently there are two homology

checkpoints. First, the strand-joining step requires

complementary base pairing to orient the 5-OH group for

its attack on the phosphotyrosyl bond at the cleavage

point. Second, the branch-migration event and associated

isomerization of the recombination complex require

homology (see insert to Figure 8).

3. Construction and use of reference integrationsites

An early application of the Flp/FRT technology

investigated chromatin domains in situ (Figure 9A). It

was based on the assumption that, in a single-S/MAR

construct, the reporter remains sensitive to influences from

the surrounding chromatin. We transfected the S/MAR-

Luciferase-FRT cassette and subsequently isolated a series


565

Figure 8. Formation and resolution of the Holl iday structure during the act ion of Flp recombinase.

Four molecules of the recombinase (ellipses) participate in site specific recombination which are bound next to the crossover

region (8 bp spacer). After single strand cuts are introduced into the recombining partner, a first strand transfer step occurs,

followed by ligation. During a branch migration process, which depends on absolute homology of the interacting strands, the end

of the spacer is reached. During or following this process an isomerization is thought to produce a structure in which the crossover

and noncrossover strands are switched. The structure is resolved by two more single strand cuts and ligation. Insert: Relative

movement of two (more extended) homologous helices during branch migration.

of clones with widely different expression characteristics.

After closing the domain by Flp-mediated targeting of a

second S/MAR element to the endogenous site, clones

were isolated and the levels of luciferase expression were

compared to the initial ones. Among the recovered clones

there was only a minority showing an augmented luciferase

activity. When a second pulse was applied to these to re-

excise the second S/MAR, the old expression

characteristics were re-established. These clones underwent

the modification-demodification cycle depicted in Figure

9A (Bode et al., 1996). The majority of them did not

permanently acquire the second S/MAR element, but rather

gained hygromycin resistance due to a faulty integration of

the circular FRT-S/MAR-HygTk vector. In other cases

integration of the circular Flp expression vector occurred,

leading to a permanent base level of recombinase activity

(Iber, 1997). These results underline a major problem with

the Flp and Cre recombinase systems: because the

reactions are reversible and excision is highly favored

(being an intramolecular process), they do not allow one to

control the direction of recombination.

This approach did therefore not permit the investigation

of a large number of events to establish an insulation

function of S/MARs, especially at sites with an initially

mediocre expression. Nevertheless, it provided the clear

demonstration of a transcriptional augmentation at a

singular integration reference site. This experiment is

therefore considered additional evidence that augmentation

is - at least in part - due to a cis-effect of S/MARs on

transcriptional initiation.

At present the fastest progress in the field is expected

from a simple reversal of the above strategy (Figure 9B):

A complete minidomain is constructed, and the flanking

S/MAR elements are removed by the successive action of

two different site specific recombinases. For this approach


566

F i g u r e 9 . E n g i n e e r i n g t h e g e n o m e : A d d i t i o n , e x c i s i o n and exchange react ions catalyzed by Flp

recombinase

A . Circular vectors can be integrated at a genomic locus that has been tagged with an identical site. The reaction has to be

driven by an excess of the vector and will easily reverse if Flp activity persists after its dilution or degradation. The depicted

experiment has been used to complete an artificial domain by the addition of a second S/MAR element (Bode et al., 1996)

B . Excision of boundary elements by site-specific recombinases (Flp, Cre) can be used to decompose an artificial domain in a

stepwise fashion. The upstream S/MAR is surrounded by FRT sites and is removed by the action of Flp recombinase. In a related

reaction, the downstream S/MAR can be excised by Cre recombinase acting upon the lox sites.

C. An expression cassette (gene 1) that is surrounded by a wild type FRT site and an FRT linker-mutant can be exchanged for a

cassette (gene 2) with an analogous set of sites. This double-reciprocal crossover reaction will delete the plasmid sequences

contained in the circular gene 2 vector and will hence result in a so called “clean exchange”.

we initially used the same 800 bp S/MAR element in both

the 5´ and 3´ positions so the results could be related to the

position of the element without the need to consider

different S/MAR structures. The upstream S/MAR was

surrounded by two FRT sites, hence could be excised by a

pulse of Flp recombinase. Similarly, the downstream

S/MAR was surrounded by two loxP sites, and could be

excised by Cre recombinase. Selection of cells which had

received the Flp construct was facilitated by a bicistronic

vector encoding Flp (first cistron) and GFP (green-

fluorescent protein, second cistron). By sorting for green

fluorescence a population of cells was obtained in which

excision had occurred with over 90% efficiency. This

underscores the fact that excision is a spontaneous event.

In a separate step, the downstream S/MAR was removed

by Cre recombinase. Here, Cre was expressed stably and

expressors were selected via hygromycin resistance,

generated by a cotransfected selection marker. The results

of a pilot experiment on a cell population, generated by

electroporation as for Figure 2 (thereby mostly

consisting of single copy constituents), are shown in

Figure 10 . These data show a moderate effect (25%


567

decrease of expression) for the removal of the 5´ S/MAR

and a strong one (85%) for the removal of the

corresponding 3´ element. It is noted that these findings

agree with the model of S/MAR action depicted in Figure

6 (transition D2 -> D2.1): the unwound structure stored in

a S/MAR element downstream from the site of

transcriptional initiation can be utilized to release the

positive superhelical strain that accumulates in front of a

transcribing polymerase. Recent experiments by Wang and

Dröge (1996) have demonstrated the persistence of

supercoiling even in the presence of topoisomerase

activities which resolve topological problems on a longer

time scale.

The experiments outlined in Figure 8B will permit

an extended series of tests for the properties of individual

clones, even at the single cell level. This possibility arises

from the use of the LacZneo gene (Walters et al. 1995) the

product of which confers neomycin resistance and at the

same time allows fluorescence-activated cell sorting

(FACS analysis) of clones according to their lacZ

expression level. In another system it was shown that

integration occurs next to the centromere in up to 50% of

clones, which causes position-effect variegation (PEV)

unless the chromatin is either kept open by an enhancer-

Figure 10 . Probing the act iv i ty of S /MARs in s i tu: excis ion of the 5´ -S/MAR and the 3´ -S/MAR from an

art i f ic ia l minidomain. The experiment followed the outline given in Figure 9 and is based on two identical 800 bp S/MAR

elements flanking a LacZ/neo fusion gene. Excision of the 3´-S/MAR by Cre results in a larger effect than excision of the 5´

element.


568

like function (Festenstein et. al., 1996) or flanked by

insulator-type elements. Both the domain-opener (Zhao et

al. 1993) and the insulator functions (Stief et al., 1989;

Phi-Van et al., 1996) that have been ascribed to S/MARs

should prevent PEV. We hope to trace these functions by

successively eliminating the 5´- and the 3´- elements. The

effect of each of these deletions on the decay of expression

will be traced over an extended period of time (cf. Walters

et al., 1996).

4. The equilibrium problem and its solutions

The simple integration/excision system of Figure 9A

has one major drawback, caused by the reversibility if the

recombination reactions. Since intramolecular excision is

kinetically favored over bimolecular integration, insertion

products are inherently unstable in the presence of

recombinase (Seibler and Bode, 1997). As an example, we

have reported the facile excision of retroviral sequences

between FRT sites which were strategically placed into the

long terminal repeats (LTRs) of a provirus (Schübeler et

al, 1997; cf. also Figure 3). Reversal of this step, i.e.

use of the remaining site for re-integration, proved to be

unfeasible. We have recently shown that this goal can be

achieved by applying a very stringent selection system

(integration of a promoter- and ATG-free cassette next to a

preexisting promoter and translation-initiation site) and by

maintaining an open chromatin structure around the target

site (Seibler et al, submitted).

Several measures have been taken to limit the activity

of the recombinase to a time interval where a high

concentration of the circular exchange vector drives the

integration. Conventionally, this is done by generating a

pulse of Flp activity from an appropriate concentration of a

transiently expressed construct. Since in many cases the

transient expression phase is followed by the

integration/stable expression of the vector, recombinase-

mRNA or -protein has been used instead. In another

approach the recombinase gene has been placed under the

control of an inducible vector (Logie and Stewart, 1995).

Alternatively, one could abolish recombinase activity

following integration by directing the integration event so

that it separates the promoter from its coding sequence

(Kilby et al., 1993). Another strategy is to use mutant

target sites. For both loxP and FRT exact 13 bp inverted

repeats are the recombinase binding sites, which implies a

stringent sequence requirement. If a point mutation is

introduced in one of the repeats, a recombination between

the site with a mutation in the left element (LE) and

another site with a mutation in the right element (RE)

would yield two recombination product sites, i.e. a wild-

type one and one with mutations in LE and RE. Since the

LE plus RE site has a dramatically reduced affinity for the

recombinase, a subsequent excisional recombination

between this and the wild-type site becomes less probable

favoring the forward (integration) reaction (Senecoff et al.,

1988; Araki et al., 1997). Unfortunately, when compared

with wild type FRTs, point mutations also lead to a

reduced recombination between the LE and RE sites, resp.,

and there are even cases where the enhanced stability

conferred upon the integrated molecule was outweighed by

the far fewer integration events which resulted from an

inefficient forward reaction.

There are no identified protein-DNA interactions in the

8 bp spacer sequence of an FRT site. At least six (and

possibly all) of these bases can be changed without

destroying Flp binding activity. Such changes produce a

mutant site that will recombine with a second mutant site

of the same composition, but not with one having a

different spacer sequence (Schlake and Bode, 1994). We

have indicated above (section IVB2) that this is an

immediate consequence of the homology check points

occurring during the branch migration step of the

recombination cycle.

Based on these results we have studied the feasibility of

a cassette exchange reaction mediated by Flp (RMCE

concept, see Figure 8C and Seibler and Bode, 1997). We

demonstrated that the double-reciprocal crossover events

occurring between FRT couples of identical composition

enable the efficient substitution of a recombination target

flanked by a wild-type FRT site and an FRT- mutant for

another cassette designed in the same way. Since RMCE is

a true equilibrium reaction (both the forward and the reverse

reaction are bimolecular processes) it proceeds to near

completion if the exchange plasmid can be provided at a

sufficiently high excess (Seibler and Bode, 1997). So far,

the RMCE concept not only provides the most efficient

solution for the equilibrium problem but it also enables a

“clean” replacement of one expression cassette for another

in the sense that prokaryotic vector parts can be deleted

during the exchange step by an appropriate placing of the

wild type FRT site (F) and the FRT mutant (Fn)

5. Outlook

The ability to perform a clean exchange of one

expression cassette for another provides an entirely new

way to gently manipulate the genome. In its most

stringent form this approach requires that expression

patterns remain unperturbed, as assessed by the

simultaneous transcription of a selection marker which has

to be removed in a second step. This goal can be achieved

by a novel two-step strategy called "tag-and-exchange"

(Askew et al., 1993) or double replacement (Stacey et al.,

1994). In our modification of the concept, an expression

cassette carrying the HygTk positive/negative selection

marker is introduced in step 1. In the appropriate cell

types, this can either be achieved by an 3-type homologous


569

recombination which substitutes an endogenous gene for

the HygTk- “tag” or by a random integration followed by

the selection of suitable integrants. The presence of the tag

can be assessed because the HygTk gene product mediates

resistance to hygromycin. Following the outline in

Figure 9C , the HygTk tag is removed in step 2 by an

exchange reaction utilizing the RMCE principle.

Successful events can be screened for the absence of the

HygTk cassette (negative selection). This is done in the

presence of ganciclovir which is converted to a toxic

compound by the thymidine kinase activity of the HygTk

fusion gene product. As a result, the initial HygTk tag

serves as a selection marker in both steps of the procedure,

obviating the need for a marker on the final DNA. This

procedure can be performed in embryonic stem (ES-) cells

with high efficiency since this cell type does not

spontaneously integrate circular DNA. In this way a

circular Flp expression construct can be provided at high

enough concentrations to generate sufficient recombinase

activity. Although it can disappear by dilution or

degradation, it will not be incorporated into the genome by

an unspecific integration event. An excess of the circular

exchange plasmid also can be used whereby the specific

exchange mediated by the two sets of Flp sites becomes

the favored pathway (Seibler et al, submitted).

The combination of targeted gene modification and

production of animals derived from ES cells has established

a powerful method for studying gene function in the

developing animal. Genes can be disrupted, inserted, or

modified in the ES cell genome, and the altered cells can be

used to generate chimaeric animals. If ES cells contribute

to the germ line, chimeras can be outcrossed to produce

progeny that are heterozygous or homozygous for the

genomic modification. Therefore, using the tag-and-

exchange concept in combination with the RMCE

technology the way is open to exchange an endogenous

gene for an analogue carrying gentle mutation(s). Since the

method avoids any further modification at this particular

locus, the effect of the mutation will become immediately

obvious.

V. Can S/MAR functions be derivedfrom sequence information?

S/MARs are polymorphic and appear to be distributed

throughout the eukaryotic genome. They are specific for

eukaryotes, as demonstrated by the observation that

S/MAR-scaffold interactions cannot be disrupted by an up

to 60,000-fold excess of double-stranded bacterial DNA

(Kay and Bode, 1994, 1995). Although prototype S/MARs

are AT-rich, they do not share sufficient sequence

similarity to allow cross-hybridization (Gasser and

Laemmli, 1987; Phi-Van and Strätling, 1990). Biologists

have physically identified S/MARs and tried to correlate

their presence with the occurrence of sequence and

structural motifs which have subsequently been used to

develop algorithms for the prediction of S/MARs from

sequence data.

A. Six prominent rules

The predictive scheme introduced by Krawetz and

colleagues (Kramer et al., 1996; Singh et al., 1997) and

Boulikas (1993a,b) makes use of six features which in

various combinations confer an affinity for the nuclear

matrix or scaffold: (i) DNA replication occurs in

association with the matrix, so sites of matrix attachment

share certain AT-rich tracts with homeotic protein

recognition sites (including several ATTA and ATTTA

tracts) and origins of replication. ( i i ) A number of genes

contain TG-rich sequences in their 3'-UTRs which can be

S/MARs (Boulikas et al, 1996). ( i i i ) Intrinsically curved

DNA occurs within or near several S/MARs, although

curvature or bending is no prerequisite for S/MAR

activities in vitro (von Kries et al., 1990). ( i v ) Certain

dinucleotides, TG, CA or TA, that produce a kink when

separated by 2-4 or 9-12 nucleotides, are prominent features

of some S/MARs (Boulikas, 1993a). ( v ) Topoisomerase

II consensus sequences and cleavage sites are concentrated

at sites of nuclear attachment. These have been used to

excise complete chromatin domains (Targa et al., 1994).

Although this enzyme responds to topology rather than to

a strict consensus, the presence of Drosophila and

vertebrate consensus sequences have served as important

criteria for the prediction of S/MARs. (vi) Many S/MARs

contain significant stretches of AT-rich sequences and both

the occurrence of An runs (Käs et al., 1993) and (AT)n tracts

(Bode et al., 1992) has been implicated in S/MAR

functions. These six patterns have been used to define a set

of decision rules with which DNA sequences can be

searched to find regions having S/MAR potential. Several

examples were published which show a reasonable

correspondence between these predictions and wet-lab

results (Kramer et al., 1996; Singh et al., 1997; Krawetz

and Bode, unpublished).

B. Stress-induced duplex destabilization(SIDD)

The above criteria indicate that S/MAR activity may be

related to structural or topological features which are not

strictly linked to primary sequence. Chemical probing and

2D gel analyses of S/MAR constructs under superhelical

tension revealed that these elements readily relieve strain

by becoming stably base-unpaired. In all cases, unpairing

could be shown to initiate at a nucleation site, the core

unwinding element (CUE), then extend to a wider region

(Figure 11 ). These observations have led to the

suggestion that S/MARs contain efficient base-unpairing


570

Figure 11. Stress-induced structures in particular double stranded DNA sequences.

Increasing negative superhelical densities lead to base unstacking which initiates at core unwinding elements (CUEs). This

process is followed by a more extensive strand separation which finally involves an entire base-unpairing region (BUR). The

energy stored in the open structures may be retrieved by nearby cruciform-, Z-DNA- or triplex-forming sequences. CUEs and BURs

can be trapped by chloroacetaldehyde (CAA) which forms etheno-derivatives with cytosine and adenosine bases that are located in

single-stranded regions. The figure shows a derivatized adenine which is no longer able to base pair.

regions (BURs). Since single-stranded character could also

be found at S/MARs in living cells, it is possible that

duplex destabilization mediates at least some of their

functions (Bode et al, 1992, 1995, 1996). This hypothesis

is supported by observations that base-unpairing properties

correlate with the strength of binding of S/MARs to

nuclear scaffold/matrix preparations in vitro, and to the

potential of these elements to augment transcriptional

initiation rates in vivo (Mielke et al., 1990, Bode et al.,

1992, see also Allen et al., 1996).

We have recently put these hypotheses to a critical test

by calculating the stress-induced duplex destabilization

(SIDD) profiles for prototype S/MARs for which chemical

reactivity data were already available. Sample results are

shown in Figure 12 for several S/MARs integrated into

plasmids. These are an 800 bp fragment from the 7 kb

S/MAR upstream from the human interferon-ß (huIFN-ß)

gene (Bode and Maass, 1988; element IV in Mielke et al.,

1990), an inactive mutant of that sequence, and the

immunoglobulin µ-chain enhancer-associated S/MAR.

For these analyses the S/MAR sequences were placed

in the pTZ-18R plasmid. This is the same plasmid as was

used to experimentally determine the reactivity of the

huIFN-ß S/MAR and ist mutant with the single-strand

specific reagent chloroacetaldehyde (CAA) (Bode et al.,

1992 and Figure 11). In all cases a superhelix density of

-0.05 was used, simulating the conditions existing in a

bacterial plasmid (Benham et al., 1997). In the resulting

destabilization plot a value near zero indicates an


571

essentially completely destabilized base pair, which is

predicted to denature with almost no input of additional free

energy. But partial destabilization, indicated by

intermediate energy values, may also be important, as it

may enable protein binding or other events to occur.

Figure 12. Stress- induced duplex destabi l izat ion (SIDD-)profi les for prototype S/MARs and a mutant.

A. 800 bp fragment from the huIFN-ß upstream S/MAR shown in a vector backbone. The S/MAR insert is indicated by the

horizontal bar and the CUE is marked by an asterisk. The peaks at map positions 2.2, 3.2 and 3.7 kb are destabilized sites at the

amp terminator, amp promoter and f1ori of the vector backbone

B . same as A but after mutagenesis of the CUE (light asterix)

C. SIDD profile for the murine IgH-enhancer-S/MAR sequence, superimposed on CAA- modification data according to Kohwi-

Shigematsu and Kohwi (1990, 1997). Sites for some S/MAR-binding proteins have been added (from Dickinson et al., 1992).

The calculated destabilization profiles show some

distinct features which recurred in all other analyses

performed to date:

(i) Those parts of the sequence derived from the original

plasmid are generally stabilized by 8-10 kcal/mole.

However, there are three sharp and well separated minima

between 2.0 and 3.7/0 kbp which correspond to the

terminator and promoter of the "-lactamase gene, resp.

and to the f1 origin. All these elements have been the

subject of earlier analyses (Figures 2 and 3 in Benham,

1993). For our purposes they serve the role of well defined

internal standards;

( i i ) in striking contrast to the prokaryotic part, the

S/MAR sequence (present between 0.2 and 1.0 map units)

is chaotically destabilized exhibiting a characteristic

succession of minima with a spacing of 200-400 bp over

its entire length. Such a modular design is thought to be

related to function and may thereby be of diagnostic value

(see the analyses by Okada et al., 1996);

( i i i ) a core unwinding element (AATATATTT in this

case), mapped by chemical labeling techniques to position

0.72 occurs at one of the most destabilized sites on the

molecule (Bode et al., 1992);

The "-lactamase-associated sites mentioned above were

among the first for which local denaturation had been

demonstrated by nuclease digestion on superhelical

pBR322 DNA (Kowalski et al., 1988). The free-energy

parameters governing these transitions have been calculated

from these experimental results (Benham, 1992). The

plasmid-derived peak centered at position 3.7/0 coincides

with the f1 ori which, in the context of this multipurpose

plasmid, enables the generation of single stranded DNA

after superinfection with a helper phage. Although the

f1ori is too short to constitute a S/MAR per se, we have

found that it contributes synergistically to scaffold binding

in the presence of other S/MAR-sequences (Figure 1 in

Mielke et al., 1990).


572

pCL (Figure 12A) and pCLmut (Figure 12B)

contain the unwinding core of the huIFN-ß S/MAR in its

wild type and mutagenized form, respectively. A critical

comparison shows that the core unwinding element which

is stabilized by less than 1 kcal/mol in pCL reaches 8

kcal/mole after mutagenesis, again in perfect agreement

with the chemical reactivity data (Bode et al, 1992).

We also have calculated destabilization profiles for an

extended series of S/MAR elements whose relative binding

strengths have been established by S. Michalowski and S.

Spiker (in preparation). A prediction of the relative binding

strength of these elements was obtained by relating the area

covered by these S/MARs in the SIDD profile to the area

covered by the ampr- related peaks. The results agreed well

with the experimental data (correlation coefficient 0.89).

This was better than predictions based on the occurrence of

other criteria, such as A-boxes (0.58), AT richness (0.77)

and even the occurrence of a motif common to this

particular set of sequences (0.81).

Evidence from several laboratories shows that some

S/MARs cohabit with enhancers (Gasser and Laemmli,

1986). This association is particularly intriguing, as

S/MARs have the capacity to augment transcription via a

non-enhancer mechanism (IIB). The most thoroughly

studied examples are the immunoglobulin !- and µ-

chain intronic enhancers, which are associated with one and

with two distinct S/MAR elements, respectively.

(Cockerill and Garrard, 1986a,b ; Cockerill et al., 1987).

They function in domain opening (Zhao et al., 1993; Bode

et al., 1996), which operates during embryonic

development (Jenuwein et al., 1993, 1994, 1997; Forrester

et al., 1994; Oancea et al., 1997). A regional

demethylation occurs in a process that relies on several cis-

acting modules, including the S/MAR (Lichtenstein et al.,

1994; Kirillov et al., 1996; Jenuwein et al., 1997). While

any S/MAR sequence appears to be able to function in this

reaction, tissue specificity is contributed by the intronic

enhancer (Kirillov et al., 1996).

For the µ-chain intronic enhancer (Figure 12c)

Kohwi-Shigematsu & Kohwi (1990, 1997) have

demonstrated an overlap between specific protein binding

sites and locations that become stably and uniformly

unpaired when this region is subjected to torsional stress.

Prominent destabilized sites coincide with both the 3´- and

the 5´-S/MAR which have been characterized by Cockerill

et al., 1987 and Mielke et al., 1990 (cf. elements XVIl and

XVIr).

We also have evaluated the destabilization properties of

the 992 bp XbaI fragment (Figure 12c) by computation.

A striking tripartite destabilization profile occurs in the

insert region, in which the (stable) enhancer is bounded by

two strongly destabilized flanks. The latter regions coincide

with the S/MARs, and also with the regions accessible to

CAA. The precision of this analysis becomes evident from

the fact that the unwinding feature which initiates at a core

unwinding element, AATATATTT, then spreads in the 5´

direction and is stalled at the enhancer border (Kohwi-

Shigematsu and Kohwi, 1990), is precisely predicted. This

directional preference is not readily explained by the mere

A+T-contents of neighboring sequences which are both

70%. In contrast to this, the unwinding region 5´ of the

enhancer does not influence neighboring regions, i.e. it

shows an all or none reactivity towards CAA. This

property is reflected by the two steep flanks bordering the

CAA-reactive region in the destabilization profile.

C. Occurrence of secondary structures

These results raise the question of whether the presence

of long stretches of base-unpaired or destabilized duplex

DNA is sufficient to account for S/MAR activities, or if

these are modulated by alternative stress-induced structures.

Schroth and Ho (1995) have demonstrated that strong

cruciform forming sequences (inverted repeats, IR) occur at

relatively high frequency in yeast (1/19700 bp) and humans

(1/41800 bp) whereas triple-helix promoting sequences

(mirror repeats, MR) are abundant only in humans

(1/49400 bp). While eukaryotic IRs are very A+T-rich,

prokaryotic ones have a relatively high G+C-content and

occur almost exclusively in transcription termination sites.

Base composition is important because cruciforms form

more easily in AT-rich sequences. Since strong cruciform

and triplex DNA forming sequences are not abundant in the

E. coli genome, these results suggest they may have

specific roles in eukaryotes, where they are concentrated in

S/MARs and in ORIs (Boulikas & Kong, 1993; Boulikas,

1995; Mielke et al., 1996 and below). Still other

observations hint at a regulatory role of direct repeats in

the replication of eukaryotic genomes. Direct repeats are

also common in S/MARs (Opstelten et al., 1989, Mielke

et al., 1996).

Among the possible supercoiled-induced alternate DNA

structures, triplexes are now felt to be the best candidates

for serving a role in gene expression, as their requirements

for specific environmental conditions and negative

supercoiling are the least stringent (Palacek, 1976, 1991).

After strand separation, the experimentally best

characterized structural change in a negatively supercoiled

DNA is the C-type cruciform extrusion. This transition

initiates with a coordinated opening of many base pairs to

form a large bubble that can be trapped by a single-strand

specific reagent like chloroacetaldehyde (Figure 11). As a

prelude to opening, the base pairs must be unstacked. This

partial relaxation already mediates reactivity to osmium

tetroxide, which allows the researcher to visualize

subsequent intermediates on the extrusion pathway

(Furlong et al., 1989).


573

For C-type extrusions, AT-rich BURs may be

positioned at the center of a cruciform as in Figure 11 .

These also could be responsible for a coordinate

destabilization of a large domain in the supercoiled DNA,

and thereby increase the probability that more distant sites

"harvest" the energy stored in these structures. In this case

a BUR could be separated from a stress-induced non-B

structure, and might be recognized and thereby stabilized by

certain scaffold proteins. Clearly, this point needs

clarification and it can only be approached when the

appropriate experiments have been performed to determine

the precise energetics for various alternatives under the

same environmental conditions. For the time being,

analyses have to be restricted to the sequence features

which would be compatible with secondary structure

formation in and around the destabilized regions.

1. Potential cruciforms in S/MAR-typesequences

Several independent lines of evidence support the idea

that cruciform structures might be enriched in S/MARs

(Boulikas, 1993, 1995): First, hnRNA, a component of

the nuclear matrix, is anchored by regions that correspond

to DNA inverted repeats. That these features direct origins

of replications and S/MARs to the nuclear matrix is

suggested by the results of a recent random cloning

experiment which found a number of elements that were

significantly enriched in IRs. In 77 kb of the human ß-

globin locus 22 potential cruciforms have been found,

some of which coincide with S/MAR sequences. The

recognition of these sequences could be due to several

established protein components of the matrix (review: Bode

et al., 1996), among these HMG1 and a number of

transcription factors.

Using a computer program for a search of regions with

the sequence requirements for cruciform formation, we

noted an extended inverted repeat in the human interferon-ß

upstream S/MAR. The corresponding stem-loop structure

(Figure 13 ) would expose the core unwinding element

(ATATTT) on the tip of a loop. We have recently

determined, that retroviruses prefer S/MAR-type sequences

for their integration into the host's genome (Figure 5 ).

Within these sequences we noted a high proportion of

direct and inverted repeats (Mielke et al., 1996). When we

calculated the corresponding SIDD profiles there were again

several coincidences between stabilization minima and

potential stem-loop structures (see Int-19 and Int-26 in

Figure 13).

A much investigated retroviral integration event which

leads to an extensive deregulation of the collagen(I)

expression is localized in the gene's first intron (Breindl et

al., 1984). As in the other reported cases, a 300 bp

sequence around this site behaves as a prototype S/MAR

element of the same extension (H. Rühl, unpublished). An

SIDD analysis of the intron reveals a succession of peaks,

one of which represents the actual site of integration.

While a more extensive in vitro analysis of the S/MAR

potential along the entire sequence is in progress (C.

Mielke, unpublished), the question arises whether this

particular minimum was preferred by incidence or whether

there are signals superimposed to it which might have

attracted the retroviral integration machinery. It is

demonstrated in Figure 13 that integration had again

occurred at the tip of a potential stem loop structure i.e.

between the two guanidine residues shown in bold print.

While none of these analyses can yet be considered proof

for the existence or relevance of these structures (VC), it is

certainly tempting to consider their contribution to explain

the range of properties associated with S/MAR-type

sequences (IIB-E).

VI. Conclusions and perspectives

S/MARs are typically found at the borders of

eukaryotic gene domains. A recent compilation (Boulikas,

1995) covers 50 well characterized S/MAR-elements for

mammalian, rodent, chicken, Drosophila and plant genes

defining domains of 5-400 kb (average size 60 kb). Since

there is an inverse relation between domain size and

potential gene activity and a frequent association of

enhancers with S/MARs, the study of S/MAR

organization yields valuable initial information about the

nature and expression of those genes that are associated

with them.

The human genome project is aimed at the localization

of all 50-100000 human genes. Progress depends, to a

large extent, on the availability of markers that are

polymorphic, very common (1E5 to 1E6 copies per

genome) and evenly dispersed. Recovery of all S/MARs

from human chromosome 19 by Nikolaev et al. (1996) has

shown that these are indeed individual and do not belong to

any family of repeated sequences (IIB-D, VB). These

attributes could make S/MARs valuable genomic markers

in sequencing projects.

Many S/MAR-related functions seem to depend on

particular DNA structures (Boulikas 1995, Singh et al.,

1997) which are recognized by distinct sets of single- or

double-strand specific binding proteins (Bode et al, 1996).

Their pronounced base unpairing character (Bode et al.,

1992) together with a possible propensity subsequently to

form non-B DNA structures under superhelical tension

(Boulikas 1993) may explain the observation that some

S/MARs coincide with recombination hot spots (Sperry et

al., 1989, Kohwi and Panchenko, 1993). Recently

investigated examples include sites of translocation in the

human type I interferon gene cluster (Pomykala et al.,

1994; Diaz 1995) and at the MLL breakpoint cluster region


574

Figure 13. Is base-unpaired DNA stabil ized by secondary structure formation?

The figure shows sequences at the sites of duplex destabilization which have the potential to yield cruciforms. The CUE (ATATTT)

of the human interferon upstream S/MAR is localized at the top of a potential stem-loop structure. Related sites are also the targets

for retrovirus integration which is exemplified by a Moloney murine leukemia integration site (Breindl et al., 1984) and by two

targets for a retroviral vector (Mielke et al., 1996).


575

(Broeker et al., 1996a, b), all of which occur within

S/MARs. Currently several projects address the question of

whether S/MAR elements, besides structuring the human

genome, might also give rise to the extreme instability of

these loci.

Until recently, the identification of genomic segments

associated with the nuclear matrix has essentially relied on

biochemical strategies. Sequence searches for S/MARs

have met with serious difficulties since, although several

characteristic motifs are known, no true consensus is

apparent (Boulikas, 1993, Kramer & Krawetz, 1995,

Kramer et al., 1996). The novel, structure-related approach

that is discussed above (VB-C) suggests that it may be

possible to recognize S/MARs on the basis of subtle

underlying properties of their sequences. This would allow

rapid progress in the localization of functional genes and

some of their associated regulatory features (Benham et al.,

1997). The availability of an increasing number of S/MAR

elements and the characterization of both their common and

individual properties will provide valuable insights

regarding genomic organization and regulation.

In the emerging fields of improving agricultural crops, and

human gene therapy the inclusion S/MARs that regulate

chromatin structure in transgene constructs appears of

immediate use to obtain consistent and authentic

expression patterns. Any of these protocols relies on the

efficiency of DNA delivery as well as on expression

properties. These are profoundly influenced by the nature of

the insertion site and the presence of DNA elements with

the potential to overcome chromosomal position effects.

Site specific recombination systems are being developed

which will ultimately allow successive rounds of

transformation with different genes inserted into the same

locus. This locus could either be an endogenous site which

can be targeted without interrupting central genomic

functions or a site which has been constructed in situ by

the inclusion of elements which define an autonomously

regulated gene domain (IVB-C).

References

Allen G. C., Hall Jr. G. E., Childs L. C., Weissinger A. K.,

Spiker S., and Thompson W. F. (1 9 9 3 ). Scaffold

attachment regions increase reporter gene expression in

stably transformed plant cells. Plant Cel l 5, 603-613.

Allen G. C., Hall Jr. G., Michalowski S., Newman W., Spiker

S., Weissinger A. K., and Thompson W. F. (1 9 9 6 ). High-

level transgene expression in plant cells: Effects of a

strong scaffold attachment region from tobacco. Plant

Ce l l 8, 899- 913.

Araki K., Araki M., and Yamamura K. I. (1 9 9 7 ). Targeted

integration of DNA using mutant lox sites in embryonic

stem cells. Nucleic Acids Res 25 (4), 868-872 (4).

Aronow B. J., Silbiger R. N., Dusing M. R., Stock J. L., Yager

K. L., Potter S. S., Hutton J. J., and Wigington D. A.

(1 9 9 2 ). Functional analysis of the human adenosine

deaminase gene thymic regulatory region and its ability to

generate position- indeperndent transgene expression.

M o l C e l l B i o l 12, 4170- 4185.

Askew G. R., Doetschman T., and Lingrel J. B. (1 9 9 3 ). Site-

directed point mutations in embryonic stem cells: A gene

targeting tag-and-exchange strategy. Mol Ce l l B i o l 13,

4115- 4124.

Benham C. (1 9 9 3 ). Sites of predicted stress-induced DNA

duplex destabilization occur preferentially at regulatory

loci. Proc Natl Acad Sci USA 90, 2999-3003.

Benham C. J. (1 9 9 2 ). Energetics of the strand separation

transition in superhelical DNA. J M o l B i o l 225, 835-

847.

Benham C., Kohwi-Shigematsu T., and Bode J (1 9 9 7 ) Stress-

induced duplex DNA destabilization in scaffold/matrix

attachment regions. J Mol B io l 272, 181-196.

Blasquez V. C., Hale M. A., Trevorrow K. W., and Garrard W.

T. (1 9 9 2 ). Immunoglobulin kappa gene enhancers

synergistically activate gene expression but

independently determine chromatin structure. J B i o l

Chem 267, 23888-23893.

Bode J., and Maass K. (1 9 8 8 ). Chromatin domain

surrounding the human interferon-beta gene as defined by

scaffold-attached regions. Biochemis try 27, 4706-

4711.

Bode J., Kohwi Y., Dickinson L., Joh R. T., Klehr D., Mielke

C., and Kohwi-Shigematsu T. (1 9 9 2 ). Biological

significance of unwinding capability of nuclear matrix-

associating DNAs. Sc ience 255, 195-197.

Bode J., Schlake T., Ríos-Ramírez M., Mielke C., Stengert

M., Kay V., and Klehr-Wirth D. (1 9 9 5 ). Scaffold/Matrix-

Attached Regions (S/MARs): Structural Properties

Creating Transcriptionally Active Loci. In Structural and

Functional Organization of the Nuclear Matrix -

International Review of Cytology, Jeon K. W., and

Berezney R., eds., vol. 162A, chapt. 8 (Orlando Academic

Press), pp. 389-453.

Bode J., Stengert-Iber M., Schlake T., Kay V., and Dietz-

Pfeilstetter A. (1 9 9 6 ). Scaffold/matrix-attached regions:

Topological switches with multiple regulatory functions.

Crit Rev Eukaryot Gene Expr 6, 115-138.

Bonifer C., Yannoutsos N., Krueger G., Grosveld F., and

Sippel A. E. (1 9 9 4 ). Dissection of the locus control

function located on the chicken lysozyme gene domain in

transgenic mice. Nucleic Acids Res 22, 4202-4210.

Bonifer C., Huber M. C., Jaegle U., Faust N., and Sippel A. E.

(1 9 9 6 ). Prerequisites for tissue specific and position

independent expression of a gene locus in transgenic

mice. J Mol Med 74, 663-671.


576

Boulikas, T. (1 9 9 3 a ) Nature of DNA sequences at the

attachment regions of genes to the nuclear matrix. J .

C e l l . B i o c h e m . 52, 14-22.

Boulikas, T (1 9 9 3 b ) Homeodomain protein binding sites,

inverted repeats, and nuclear matrix attachment regions

along the human "-globin gene complex. J . C e l l .

B i o c h e m . 52, 23-36.

Boulikas T. (1 9 9 5 ). Chromatin and prediction of MAR

sequences. In International Review of Cytology, Jeon K.

W., and Berezney R., eds., vol. 162A (Orlando Academic

Press), pp. 279-388.

Boulikas T., and Kong C. F. (1 9 9 3 ). Multitude of inverted

repeats characterizes a class of anchorage sites of

chromatin loops to the nuclear matrix. J Ce l l Biochem

53, 1-12.

Boulikas T, Kong CF, Brooks D, and Hsie L (1 9 9 6 ) The 3'

untranslated region of the human poly(ADP-ribose)

polymerase gene is anchored to the nuclear matrix. Int

J Oncol 9, 1287-1294.

Breindl M., Harbers K., and Jaenisch R. (1 9 8 4 ). Retrovirus-

induced lethal mutation in collagen I gene of mice is

associated with an altered chromatin structure. Cel l 38, 9-

16.

Breyne P., Van Montagu M., Depicker A., and Gheysen G.

(1 9 9 2 ). Characterization of a plant scaffold attachment

region in a DNA fragment that normalizes transgene

expression in tobacco. Plant Cel l 4, 463-471.

Broeker P. L., Harden A., Rowley J. D., and Zeleznik-Le N.

(1 9 9 6 ). The mixed lineage leukemia (MLL) protein

involved in 11q23 translocations contains a domain that

binds cruciform DNA and scaffold attachment region (SAR)

DNA. Curr Top Microbiol Immunol 211, 259-268.

Broeker P. L. S., Super H. G., Thirman M. J., Pomykala H.,

Yonebayashi Y., Tanabe S., Zeleznik-Le N., and Rowley J.

D. (1 9 9 6 ). Distribution of 11q23 breakpoints within the

MLL breakpoint cluster region in de novo acute leukemia

and in treatment-related acute myeloid leukemia:

Correlation with scaffold attachment regions and

topoisomerase II consensus binding sites. Blood 87,

1912-1922.

Chamberlain J. W., Vasavada H. A., Ganguly S., and

Weissmann S. M. (1 9 9 1 ). Identification of cis sequences

controlling efficient position-independent tissue-specific

expression of human major histocompatibility complex

class I genes in transgenic mice. M o l C e l l B i o l 11,

3564-3572.

Chung J. H., Whiteley M., and Felsenfeld G. (1 9 9 3 ). A 5'

element of the chicken Beta-globin domain serves as an

insulator in human erythroid cells and protects against

position effect in Drosophila. Cel l 74, 505-514.

Cockerill P. N., and Garrard W. T. (1 9 8 6 ). Chromosomal

loop anchorage of the kappa immunoglobulin gene occurs

next to the enhancer in a region containing topoisomerase

II sites.. Cel l 44 , 273-282 .

Cockerill P. N., Yuen M. H., and Garrard W. T. (1 9 8 7 ). The

enhancer of the immunoglobulin heavy chain locus is

flanked by presumptive chromosomal loop anchorage

elements. J Biol Chem 262, 5394-5397.

Collis P., Antoniou M., and Grosveld F. (1 9 9 0 ). Definition

of the minimal requirements within the human "-globin

gene and the dominant control region for high level

expression. EMBO J 9, 233-240.

Cook P. R., and Brazell I. A. (1 9 7 7 ). The superhelical density

of nuclear DNA from human cells. Eur J Biochem 74,

527-531.

Corces V. G. (1 9 9 5 ). Chromatin insulators: Keeping

enhancers under control. Nature 376, 462-463.

Diaz M. O. (1 9 9 5 ). The human type I interferon gene cluster.

Seminars in Viro logy 6, 143-149.

Dickinson L., Joh T., Kohwi Y.-, and Kohwi-Shigematsu T.

(1 9 9 2 ). A tissue-specific MAR/SAR binding protein with

unusual binding site recognition. Cel l 70, 631-645.

Dietz A., Kay V., Schlake T., Landsmann J., and Bode J.

(1 9 9 4 ). A plant scaffold attached region detected close to

a T- DNA integration site is active in mammalian cells.

Nucleic Acids Res 22, 2744-2751.

Dietz-Pfeilstetter A., Kay V., Landsmann J., and Bode J.

( 1 9 9 6 ). Characterization of a plant scaffold attached

region (SAR) from a T-DNA integration site /Biosafety of

transgenic organism, horizontal gene transfer and

expression of transgenes., Schmidt E. R., ed., vol. /25

(Heidelberg Springer), pp. 261-268.

Dillon N., and Grosveld F. (1 9 9 3 ). Transcriptional regulation

of multigene loci: Multilevel control. Trends Genet 9 ,

134-137.

Dobie K., Mehtali M., McClenaghan M., and Lathe R.

(1 9 9 7 ). Variegated gene expression in mice. Trends

Genet 13, 127- 130.

Dorer D. R. (1 9 9 7 ). Do transgene arrays form

heterochromatin in vertebrates?. Transgenic Res 6, 3-

10.

Dorer D. R., and Henikoff S. (1 9 9 4 ). Expansions of

transgene repeats cause heterochromatin formation and

gene siencing in Drosophila. Cel l 77, 993-1002.

Eggert H., and Jack R. S. (1 9 9 1 ). An ectopic copy of the

Drosophila ftz associated SAR neither reorganizes local

chromatin structure nor hinders elution of a chromatin

fragment from isolated nuclei. EMBO J 10, 1237-1243.

Eissenberg J. C., and Elgin S. C. R. (1 9 9 1 ). Boundary

functions in the control of gene expression. Trends

Genet 7, 335-340.

Epner E., Kim C. G., and Groudine M. (1 9 9 2 ). What does the

locus control region control?. Curr Biol 2, 262-264.

Ferraro A., Eufemi M., Cervoni L., and Turano C. (1 9 9 5 ). The

binding of DNA to the nuclear matrix studied by

crosslinking reactions. J C e l l B i o c h e m Suppl. 21B,

134 (J7-205).

Ferraro A., Cervoni L., Eufemi M., Altieri F., and Turano C.

(1 9 9 6 ). Comparison of DNA-protein interactions in


577

intact nuclei from avian liver and erythrocytes: A cross-

linking study. J Cel l Biochem 62, 495-505.

Festenstein R., Tolaini M., Corbella P., Mamalaki C.,

Parrington J., Fox M., Miliou A., Jones M., and Kioussis

D. (1 9 9 6 ). Locus control region function and

heterochromatin-induced position effect variegation.

Sc ience 271, 1123-1125.

Fiering S., Kim C. G., Epner E. M., and Groudine M. (1 9 9 3 ).

An "in-out" strategy using gene targeting and Flp

recombinase for the functional dissection of complex DNA

regulatory elements: Analysis of the ß-globin locus

control region. P r o c N a t l A c a d S c i U S A 90, 8469-

8473.

Forrester W. C., Van Genderen C., Jenuwein T., and

Grosschedl R. (1 9 9 4 ). Dependence of enhancer-mediated

transcription of the immunoglobulin Mu gene on nuclear

matrix attachment regions. Sc ience 265, 1221-1225.

Fukushige S., and Sauer B. (1 9 9 2 ). Genomic targeting with a

positive-selection lox integration vector allows highly

reproducible gene expression in mammalian cells. Proc

Natl Acad Sci USA 89, 7905-7909.

Furlog J. C., Sullivan K. M., Murchie A. I. H., Gough G. W.,

and Lilley D. M. J. (1 9 8 9 ). Localized chemical

hyperreactivity in supercoiled DNA: Evidence for base

unpairing in sequences that induce low-salt cruciform

extrusion. Biochemis try 28, 2009-2017.

Gasser S. M., and Laemmli U. K. (1 9 8 6 ). Cohabitation of

scaffold binding regions with upstream/enhancer elements

of three developmentally regulated genes of D.

melanogaster. Cel l 46, 521-530.

Gasser S. M., and Laemmli U. K. (1 9 8 7 ). A glimpse at

chromosomal order. Trends in Genetics 3, 16-22.

Greaves D. R., Wilson F. D., Lang G., and Kioussis D.

(1 9 8 9 ). Human CD2 3'flanking sequences confer high-

level T-cell specific, position-independent gene

expression in transgenic mice. Cel l 56, 979-986.

Hempel K., and Strätling W. H. (1 9 9 6 ). The chicken

lysozyme gene 5'MAR and the Drosophila histone SAR

are electroelutable from encapsulated and digested nuclei. J

Ce l l Sc i . 109, 1459-1469.

Iber, M. (1 9 9 7 ). Sequenzspezifische Rekombinationsysteme

zur Charakterisierung von SAR-Elementen innerhalb

genomischer Referenz-Integrationsstellen. PhD Thes i s ,

Technische Universität Braunschweig

Igó-Kemenes T., and Zachau H. G. (1 9 7 7 ). Domains in

chromatin structure. Cold Spring Harb Symp Quant

B i o l 42, 109-118.

Jackson D. A., Dickinson P., and Cook P. R. (1 9 9 0 ). The

size of chromatin loops in HeLa cells. EMBO J 9, 567-

571.

Jenuwein T., Forrester W., and Grosschedl R. (1 9 9 3 ). Role of

enhancer sequences in regulating accessibility of DNA in

nuclear chromatin. Cold Spring Harbor Symp

Quant Biol 58, 97- 103.

Jenuwein T., Forrester W. C., Qiu R.-G., and Grosschedl R.

(1 9 9 3 ). The immunoglobulin Mu enhancer

coreestablishes local factor access in nuclear chromatin

independent of transcriptional stimulation. G e n e s D e v

7, 2016-2032.

Jenuwein T., Forrester W. C., fernandez-Herrero L. A., Laible

g., Dull M., and Grosschedl R. (1 9 9 7 ). Extension of

chromatin accessibility by nuclear matrix attachment

regions. Nature 385, 269-272.

Kalos M., and Fournier R. E. K. (1 9 9 5 ). Position-

independent transgene expression mediated by boundary

elements from the apolipoprotein B chromatin domain.

M o l C e l l B i o l 15, 198-207.

Käs E., and Chasin L. A. (1 9 8 7 ). Anchorage of the chinese

hamster dihydrofolate reductase gene to the nuclear

scaffold occurs in an intragenic region. J M o l B i o l 198,

677-692.

Käs E., Poljak L., Adachi Y., and Laemmli U. K. (1 9 9 3 ). A

model for chromatin opening: Stimulation of

topoisomerase II and restriction enzyme cleavage of

chromatin by distamycin. EMBO J 12, 115-126.

Kay V., and Bode J. (1 9 9 4 ). Binding specificity of a nuclear

scaffold: Supercoiled, single-stranded, and scaffold-

attached- region DNA. Biochemis try 33, 367-374.

Kay V., and Bode J. (1 9 9 5 ). Detection of scaffold-attached

regions (SARs) by in vitro techniques; activities of these

elements in vivo. I n Methods in Molecular and

C e l l u l a r B i o l o g y : Methods for studying DNA-protein

interactions - an overview, Papavassiliou A. G., and King

S. L., eds., vol. 5 (Inc. Wiley-Liss), pp. 186-194.

Kellum R., and Schedl P. (1 9 9 2 ). A group of scs elements

function as domain boundaries in an enhancer-blocking

assay. M o l C e l l B i o l 12, 2424-2431.

Kimball A. S., Kimball M. L., Jayaram M., and Tullius T. D.

(1 9 9 5 ). Chemical probe and missing nucleoside analysis

of Flp recombinase bound to the recombination target

sequence. Nucleic Acids Res 23, 3009-3017.

Kilby N. J., Snaith M. R., and Murray J. A. H. (1 9 9 3 ). Site-

specific recombinases: Tools for genome engineering.

Trends Genet 9, 413-421.

Kirillov A., Kistler B., Mostoslavsky R., Cedar H., Wirth T.,

and Bergman Y. (1 9 9 6 ). A role for nuclear NF-!B in B-

cell specific demethylation of the Ig kappa locus. Nature

Genet 13, 435-441.

Klehr D., Maass K., and Bode J. (1 9 9 1 ). SAR elements from

the human IFN-ß domain can be used to enhance the stable

expression of genes under the control of various

promoters. Biochemis try 30, 1264-1270.

Klehr D., Schlake T., Maass K., and Bode J. (1 9 9 2 ). Scaffold-

attached regions (SAR elements) mediate transcriptional

effects due to butyrate. Biochemis try 31, 3222-3229.

Kohwi Y., and Panchenko Y. (1 9 9 3 ). Transcription-

dependent recombination induced by triple-helix

formation. Genes Dev 7, 1766-1778.


578

Kohwi-Shigematsu T., and Kohwi Y. (1 9 9 0 ). Torsional stress

stabilizes extended base-unpairing in DNA flanking the

immunoglobulin heavy chain enhancer. Biochemis try

29, 9551- 9560.

Kohwi-Shigematsu T., and Kohwi Y. (1 9 9 7 ). High unwinding

capability of matrix attachment regions and ATC-sequence

context-specific MAR-binding proteins. in Nuclear

Structure and Gene Express ion , (Berezney R and

Stein G, Editors), pp111-114, Academic Press.

Kohwi-Shigematsu, T., Maass K, and Bode J (1 9 9 7 ). A

thymocyte factor, SATB1, suppresses transcription of

stably integrated MAR-linked reporter genes.

Biochemis try , 36, 12005-12010.

Kowalski D., Natale D. A., and Eddy M. J. (1 9 8 8 ). Stable

DNA unwinding, not "breathing" accounts for single-

strand- specific nuclease hypersensitivity of specific A+T-

rich sequences. P r o c N a t l A c a d S c i U S A 85, 9464-

9468.

Kramer J. A., Singh G. B., and Krawetz S. A. (1 9 9 6 ).

Computer- assisted search for sites of nuclear matrix

attachment. Genomics 33, 305-308.

Kricker M. C., Drake J. W., and Radman M. (1 9 9 2 ).

Duplication- targeted DNA methylation and mutagenesis

in the evolution of eukaryotic chromosomes. Proc Natl

Acad Sci USA 89, 1075-1079.

Kwon H. J., Tirumalai R., Landy A., and Ellenberger T.

(1 9 9 7 ). Flexibility in DNA recombination: Structure of

the lambda integrase catalytic core. Sc ience 276, 126-

131.

Laemmli U. K., Kaes E., Poljak L., and Adachi Y. (1 9 9 2 ).

Scaffold-associated regions: cis-Acting determinants of

chromatin structural loops and functional domains. Curr

Opin Genet Dev 2, 275-285.

Lakso M., Pichel J. G., Gorman J. R., Sauer B., Okamoto Y.,

Lee E., Alt F. W., and Westphal H. (1 9 9 6 ). Efficient in

vivo manipulation of mouse genomic sequences at the

zygote stage. Proc Nat l Acad Sc i USA 93, 5860-

5865.

Levy-Wilson B., and Fortier C. (1 9 8 9 ). The limits of the

DNAse I-sensitive domain of the human apolipoprotein B

gene coincide with the location of chromosomal

anchorage loops and define the 5' and 3'-boundaries of the

gene. J Biol Chem 264, 21196- 21204.

Li Q., and Stamatoyannopoulos G. (1 9 9 4 ). Hypersensitive

site 5 of the human Beta locus control region functions as

a chromatin insulator. Blood 84, 1399-1401.

Li X. J., Wang H., and Seeman N. C. (1 9 9 7 ). Direct evidence

for Holliday junction crossover isomerization.


Lichtenstein M., Keini G., Cedar H., and Bergman Y. (1 9 9 4 ).

B cell-specific demethylation: A novel role for the

intronic kappa chain enhancer sequence. Cel l 76, 913-

923.

Logie C., and Stewart A. F. (1 9 9 5 ). Ligand-regulated site-

specific recombination. Proc Nat l Acad Sci USA 92,

5940- 5944.

Lyznik L. A., Mitchell J. C., Hirayama L., and Hodges T. K.

(1 9 9 3 ). Activity of yeast Flp recombinase in maize and

rice protoplasts. Nucleic Acids Res 21, 969-975.

Matzke M. A., and Matzke A. J. M. (1 9 9 5 ). Homology-

dependent gene silencing in transgenic plants: What does

it really tell us?. Trends Genet 11, 1-3.

McKnight R. A., Shamay A., Sankaran L., Wall R. J., and

Hennighausen L. (1 9 9 2 ). Matrix-attachment regions can

impart position-independent regulation of a tissue-

specific gene in transgenic mice. Proc Nat l Acad S c i

USA 89, 6943-6947.

McKnight R. A., Spencer M., Wall R. J., and Hennighausen L.

(1 9 9 6 ). Severe position effects imposed on a 1 kb mouse

whey acidic protein gene promoter are overcome by

heterologous matrix attachment regions. Mol Reprod

Dev 44, 179-184.

Mehtali M., LeMeur M., and Lathe R. (1 9 9 0 ). The

methylation- free status of a housekeeping transgene is

lost at high copy number. Gene 91, 179-184.

Mielke C., Kohwi Y., Kohwi-Shigematsu T., and Bode J.

(1 9 9 0 ). Hierarchical binding of DNA fragments derived

from scaffold- attached regions: Correlations of properties

in vitro and function in vivo. Biochemis try 29, 7475-

7485.

Mielke C., Maass K., Tuemmler M., and Bode J. (1 9 9 6 ).

Anatomy of highly expressing chromosomal sites

targeted by retroviral vectors. Biochemis try 35, 2239-

2252.

Mirkovitch J., Mirault M. E., and Laemmli U. K. (1 9 8 4 ).

Organization of the higher-order chromatin loop: specific

DNA attachment sites on nuclear scaffold. Cel l 39, 223-

232.

Mlynarova L., Loonen A., Heldens J., Jansen R. C., Keizer P.,

Stiekema W. J., and Nap J. P. (1 9 9 4 ). Reduced position

effect in mature transgenic plants conferred by the chicken

lysozyme matrix-associated region. Plant Ce l l 6, 417-

426.

Mlynarova L., Jansen R. C., Conner A. J., Stiekema W. J.,

and Nap J.-P. (1 9 9 5 ). The MAR-mediated reduction in

position effect can be uncoupled from copy number-

dependent expression in transgenic plants. P l a n t C e l l

7, 599-609.

Neznanov N., Kohwi-Shigematsu T., and Oshima R. G.

(1 9 9 6 ). Contrasting effects of the SATB1 core nuclear

matrix attachment region and flanking sequences of the

keratin 18 gene in transgenic mice. M o l B i o l C e l l 7 ,

541-552.

Nikolaev L. G., Tsevegiyn T., Akopov S. B., Ashworth L. K.,

and Sverdlov E. D. (1 9 9 6 ). Construction of a

chromosome specific library of human MARs and

mapping of matrix attachment regions on human

chromosome 19th. Nucleic Acids Res 24, 1330-1336.


579

Novak U., Harris E. A. S., Forrester W., Groudine M., and

Gelinas R. (1 9 9 0 ). High-level ß-globin expression after

retroviral transfer of locus activation region-containing

human ß-globin derivatives into murine erythroleukemia

cells. Proc Natl Acad Sci USA 87, 3386-3390.

O'Gorman S., Fox D. T., and Wahl G. M. (1 9 9 1 ).

Recombinase- mediated gene activation and site-specific

integration in mammalian cells. Sc ience 251, 1351-

1355.

Oancea A. E., Berru M., and Shulman M. J. (1 9 9 7 ).

Expression of the (recombinant) endogenous

immunoglobulin heavy-chain locus requires the intronic

matrix attachment regions. Mol Ce l l B i o l 17 (5),

2658-2668 (5).

Okada S., Tsutsui K., Tsutsui K., Seki S., and Shohmori T.

(1 9 9 6 ). Subdomain structure of the matrix attachment

region located within the mouse immunoglobulin kappa

gene intron. Biochem B i o p h y s Res Comm 222,

472- 477.

Opstelten R. J. G., Clement J. M. E., and Wanka F. (1 9 8 9 ).

Direct repeats at nuclear matrix-associated DNA regions

and their putative control function in the replicating

eukaryotic genome. Chromosoma 98, 422-427.

Palacek E. (1 9 7 6 ). Premelting changes in DNA

conformation. P r o g r N u c l A c i d R e s M o l B i o l 18,

151-183.

Palacek E. (1 9 9 1 ). Local supercoil-stabilized DNA structures.

Crit Rev Biochem Mol Bio l 26, 151-159.

Palmiter R. D., and Brinster R.L. (1 9 8 6 ). Germ-line

transformation of mice. Ann Rev Genet 20, 465-499.

Palmiter R. D., Chen H. Y., and Brinster R. L. (1 9 8 2 ).

Differential regulation of metallothionein-thymidine

kinase fusion genes in transgenic mice and their

offspring. Cel l 29, 701-710.

Palmiter R. D., Sandgren E. P., Koeller D. M., and Brinster R.

(1 9 9 3 ). Distal regulatory elements from the mouse

metallothionein locus stimulate gene expression in

transgenic mice. M o l C e l l B i o l 13, 5266-5275.

Paulson J. R., and Laemmli U. K. (1 9 7 7 ). The structure of

histone-depleted metaphase chromosomes. Cel l 12, 817-

828.

Phi-Van L., and Strätling W. H. (1 9 8 8 ). The matrix

attachment regions of the chicken lysozyme gene co-map

with the boundaries of the chromatin domain. EMBO J 7,

655-664.

Phi-Van L., and Strätling W. H. (1 9 9 6 ). Dissection of the

ability of the chicken lysozyme gene 5' matrix attachment

region to stimulate transgene expression and to dampen

position effects. B iochemis try 35, 10735-10742.

Phi-Van L., von Kries J. P., Ostertag W., and Strätling W. H.

(1 9 9 0 ). The chicken lysozyme 5'matrix attachment

region increases transcription from a heterologous

promotor in heterologous cells and dampens position

effects on the expression of transfected genes. M o l C e l l

B i o l 10, 2302-2307.

Poljak L., Seum C., Mattioni T., and Laemmli U. K. (1 9 9 4 ).

SARs stimulate but do not confer position independent

gene expression. Nucleic Acids Res 22, 4386-4394.

Pomykala H. M., Bohlander S. K., Broeker P. L., Olopade O.

I., and Diaz M. O. (1 9 9 4 ). Breakpoint junctions of

chromosome 9p deletions in two human glioma cell lines.

M o l C e l l B i o l 14, 7604-7610.

Rohdewohld H., Weiher H., Reik W., Jaenisch R., and Breindl

M. (1 9 8 7 ). Retrovirus integration and chromatin

structure: Moloney Murine Leukemia proviral integration

sites map near DNase I hypersensitive sites. J Viro l 61,

336-343.

Ramstein J., and Lavery R. (1 9 8 8 ). Energetic coupling

between DNA bending and base pair opening. Proc Natl

Acad Sci USA 85, 7231-7235.

Romig H., Ruff J., Fackelmayer F. O., Patil M. S., and Richter

A. (1 9 9 4 ). Characterisation of two intronic nuclear-

matrix-attachment regions in the human DNA

topoisomerase I gene. Eur J Biochem 221, 411-419.

Sauer B. (1 9 9 4 ). Site-specific recombination: Developments

and applications. Curr Opin Biotechnol 5, 521-527.

Schedl A., Montoliu L., Kelsey G., and Schuetz G. (1 9 9 3 ). A

yeast artificial chromosome covering the tyrosinase gene

confers copy number-dependent expression in transgenic

mice. Nature 362, 258-261.

Scheuermann R. H., and Chen U. (1 9 8 9 ). A developmental-

specific factor binds to suppressor sites flanking the

immunoglobulin heavy-chain enhancer. G e n e s D e v 3,

1255-1266.

Schlake T. (1 9 9 4 ). Entwicklung einer auf SAR-Elementen und

sequenzspezifischer Rekombination basierenden Strategie

zur stabilen Hochexpression in Mammaliazellen. PhD

t h e s i s , Technische Universität Braunschweig.

Schöffl F., Schröder G., Kliem M., and Rieping M. (1 9 9 3 ).

An SAR sequence containing 395 bp DNA fragment

mediates enhanced, gene-dosage-correlated expression of a

chimaeric heat shock gene in transgenic tobacco plants.

Transgenic Res 2, 93-100.

Schroth G. P., and Ho P. S. (1 9 9 5 ). Occurence of potential

cruciform and H-DNA forming sequences in genomic DNA.

Nucleic Acids Res 23, 1977-1983.

Schübeler D., Mielke C., Maass K. and Bode J. (1 9 9 6 ).

Scaffold/Matrix-attached regions act upon transcription in

a context-dependent manner. Biochemis try 35, 11160-

11169.

Schübeler D., Mielke C. and Bode J. (1 9 9 7 ) Excision of an

integrated provirus by the action of FLP-recombinase. In

Vitro Cel l Dev Bio l Anim 33, 825-830.

Seibler J., and Bode J. (1 9 9 7 ). Double-reciprocal crossover

mediated by FLP-recombinase: A concept and an assay.


Senecoff J. F., Rossmeissl P. J., and Cox M. M. (1 9 8 8 ). DNA

recognition by the Flp recombinase of the yeast 2

plasmid: A mutational analysis of the Flp binding site. J

M o l B i o l 201, 405-421.


580

Shih C.-C., Stoye J. P., and Coffin J. M. (1 9 8 8 ). Highly

preferred targets for retrovirus integration. Cel l 53, 531-

537.

Singh G. B., Kramer J. A., and Krawetz S. A. (1 9 9 7 ).

Mathematical model to predict regions of chromatin

attachment to the nuclear matrix. N u c l e i c A c i d s R e s

25, 1419-1425 .

Sippel A. E., Schaefer G., Faust N., Saueressig H., Hecht A.,

and Bonifer C. (1 9 9 3 ). Chromatin domains constitute

regulatory units for the control of eukaryotic genes. Cold

Spring Harb Symp Quant Biol 58, 37-44.

Sperry A. O., Blasquez V. C., and Garrard W. T. (1 9 8 9 ).

Dysfunction of chromosomal loop attachment sites:

Illegitimate recombination linked to matrix association

regions and topoisomerase II. Proc Natl Acad S c i

USA 86, 5497- 5501.

Stacey A., Schnieke A., McWhir J., Cooper J., Colman A.,

and Melton D. W. (1 9 9 4 ). Use of double-replacement

targeting to replace the murine -lactalbumin gene with

its human counterpart in embryonic stem cells and mice.

M o l C e l l B i o l 14, 1009-1016.

Stief A., Winter D. M., Strätling W. H., and Sippel A. E.

(1 9 8 9 ). A nuclear DNA attachment element mediates

elevatedand position-independent gene activity. Nature

341, 343- 345.

Stockhaus J., Eckes P., Blau A., Schell J., and Willmitzer L.

(1 9 8 7 ). Organ-specific and dosage-dependent expression

of a leaf/stem specific gene from potato after tagging and

transfer into potato and tobacco plants. Nucle ic Ac ids

Res 15, 3479- 3491.

Targa F. R., Razin S. V., Moura Gallo C. V. De, and Scherrer

K. (1 9 9 4 ). Excision close to matrix attachment regions

of the entire chicken Alpha-globin gene domain by

nuclease S1 and characterization of the framing structures.

Proc Natl Acad Sci USA 91, 4422-4426.

Thorey I. S., Cecena G., Reynolds W., and Oshima R. G.

(1 9 9 3 ). Alu Sequence involvement in transcriptional

insulation of the keratin 18 gene in transgenic mice. M o l

C e l l B i o l 13, 6742-6751.

Van der Geest A. H. M., Hall Jr. G. E., Spiker S., and Hall T. C.

(1 9 9 4 ). The Beta-phaseolin gene is flanked by matrix

attachment regions. Plant J 6, 413-423.

Vazquez J., and Schedl P. (1 9 9 4 ). Sequences required for

enhancer blocking activity of scs are located within two

nuclease-hypersensitive regions. EMBO J 13, 5984-

5993.

von Kries J. P., Phi-Van L., Diekmann S., and Strätling W. H.

(1 9 9 0 ). A non-curved chicken lysozyme 5'matrix

attachment site is 3'followed by a strongly curved DNA

sequence. Nucleic Acids Res 18, 3881-3885.

Walters M. C., Fiering S., Eidemiller J., Magis W., Groudine

M., and Martin D. I. K. (1 9 9 5 ). Enhancers increase the

probability but not the level of gene expression. Proc


Walters M. C., Magis W., Fiering S., Eidemiller T., Scalzo D.,

Groudine M., and Martin D. I. K. (1 9 9 6 ). Transcriptional

enhancers act in cis to suppress position-effect

variegation. Genes Dev 10, 185-195.

Wang D. M., Taylor S., and Levy-Wilson B. (1 9 9 6 ).

Evaluation of the function of the human apolipoprotein B

gene nuclear matrix association regions in transgenic mi.

J Lipid Res 37, 2117-2124.

Wang Z. Y., and Dröge P. (1 9 9 6 ). Differential control of

transcription-induced and overall DNA supercoiling by

eukaryotic topoisomerases in vitro. EMBO J 15, 581-

589.

Weidle U. H., Buckel P., and Wienberg J. (1 9 8 8 ). Amplified

expression constructs for human tissue-type plasminogen

activator in Chinese hamster ovary cells: instability in

the absence of selective pressure. Gene 66, 193-203.

Weintraub H. (1 9 9 3 ). Summary: Genetic tinkering--Local

problems, local solutions. Cold Spr ing Harb Symp

Quant Biol 58, 819-836.

Weisbrod S. (1 9 8 2 ). Active chromatin. Nature 297, 289-

295.

Wigley P., Becker C., Beltrame J., Blake T., Crocker L.,

Harrison S., Lyons I., McKenzie Z., Tearle R., Crawford

R., and Robins A. (1 9 9 4 ). Site-specific transgene

insertion: An approach. Reprod Fert i l Dev 6, 585-

588.

Withers-Ward E. S., Kitamura Y., Barnes J. P., and Coffin J.

M. (1 9 9 4 ). Distribution of targets for avian retrovirus

DNA integration in vivo. Genes Dev 8, 1473-1487.

Wolffe A. P. (1 9 9 4 a ). Gene regulation: Insulating chromatin.

Curr Biol 4, 85-87.

Wolffe A. P. (1 9 9 4 b ). The transcription of chromatin

templates. Curr Opin Genet Dev 4, 245-254.

Yagil G. (1 9 9 1 ). Paranemic structures of DNA and their role

in DNA unwinding. CRC Crit Rev Biochem M o l

B i o l 26, 475-559.

Zhao K., Kaes E., Gonzalez E., and Laemmli U. K. (1 9 9 3 ).

SAR- dependent mobilization of histone H1 by HMG-I/Y

in vitro: HMG- I/Y is enriched in H1-depleted chromatin.

EMBO J 12, 3237- 3247.

Zong R. T., and Scheuermann R. H. (1 9 9 5 ). Mutually

exclusive interaction of a novel matrix attachment region

binding protein and the NF-MuNR enhancer repressor -

Implications for regulation of immunoglobulin heavy

chain expression. J Biol Chem 270, 24010-24018.


581


Synthetic concatemers as artificial MAR: importanceof a particular configuration of short AT-tracts forprotein recognition

Ken Tsutsui

Department of Molecular Biology, Institute of Cellular and Molecular Biology, Okayama University Medical School, 2-5-1

Shikatacho, Okayama 700, Japan

_____________________________________________________________________________________Correspondence to: Ken Tsutsui, Tel: 086-235-7386, Fax: 086-224-5150, E-mail: [email protected]

Summary

The matrix attachment region (MAR), a class of sequences involved in organization of genomic DNA

into looped superhelical domains, i s also believed to be important in the regulation of nuclear

functions such as DNA replication, transcription, and recombination. The association of MARs to the

nuclear matrix is probably mediated by a variety of proteins which selectively bind to MAR. Because

of the complex nature of MAR-protein interactions, i t i s s t i l l difficult to te l l , from sequence

information alone, whether or not a particular DNA region is a MAR. An abundant nuclear protein,

termed SP120, is one of the major MAR binding proteins identif ied so far that select ively binds to

AT-rich MARs of different origins in v i t ro . We have recently found that SP120 also binds to a GC-

rich concatemer synthesized by random ligation of a short duplex oligonucleotide. Although the result

seemed rather paradoxical at first, subsequent experiments with sequence-manipulated concatemers

indicated that intactness of short homopolymeric AT-tracts harbored in the concatemer and a particular

pattern of their distribution within it are prerequisites for the binding.

I. Introduction

A. Short history

Presence of a long-range organization in nuclear DNA

was originally proposed in early studies analyzing the

sedimentation behavior of detergent-lysed cells in a solution

containing intercalating agents (Cook and Brazell, 1975; Ide

et al., 1975). The interphase chromatin DNA appeared to be

folded into supercoiled loop domains through its attachment

to a subnuclear skeletal structure, which was later isolated

and characterized as the nuclear matrix (Berezney and Coffey,

1977) or nuclear scaffold (Mirkovitch et al., 1984). This

model acquired further support when characteristic DNA

regions were discovered at the attachment sites (Cockerill and

Garrard, 1986; Mirkovitch et al., 1984). Such DNA regions,

designated SAR or MAR, have been allocated in many

cloned genes and implicated in the regulation of gene

expression and other nuclear processes (reviewed in Gasser et

al., 1989; Zlatanova and van Holde, 1992; Boulikas, 1995).

B. Characteristic features of MAR

Comparison of MAR sequences identified thus far

revealed some characteristic features. 1) Relatively long DNA

region is required for the attachment and very few MARs are

shorter than 300 bp. 2) MARs are usually located in

noncoding regions. 3) With a few exceptions MARs are rich

in adenines (A) and thymines (T); A+T content of a typical

AT-rich type MAR is higher than 65%. 4) Regulatory

sequences are frequently present in the vicinity of MAR or

cohabit with MAR. These include replication origins,

transcriptional enhancers and promoters. 5) Other sequence

motifs like AT-tracts, DNA unwinding elements, clustered

recognition sites for various transcription factors, inverted

repeats or palindromes, cleavage consensus for DNA

topoisomerase II, curved/kinked DNA motifs, left handed and

triple helical structures, and DNase I-hypersensitive sites are

frequently associated with MARs (Boulikas, 1995). 6)

MARs appear to function beyond species barrier. For

instance MARs isolated from yeast can specifically bind to

the nuclear matrix prepared from mouse cells.

Tsutsui: SP120 interacts with short homopolymeric AT-tracts

582

C. SP120 and other MAR binding proteins

Since MAR sequences are heterogeneous in nature,

proteins interacting with MAR can also be heterogeneous.

Indeed, several nuclear proteins including ARBP (von Kries

et al., 1991), SATB1 (Dickinson et al., 1992), lamin B1

(Ludérus et al., 1992), SAF-A (Romig et al., 1992),

nucleolin (Dickinson and Kohwi-Shigematsu, 1995), histone

H1 (Izaurralde et al., 1989), and DNA topoisomerase II

(Adachi et al., 1989) have already been shown to bind MAR

with considerable affinity. We have identified an MAR

binding protein, termed SP120, which is associated with the

nuclear matrix (Tsutsui et al., 1993). Sequence determination

of cDNA for the rat SP120 showed that the protein is a

homologue of human hnRNP U protein (Kiledjian &

Dreyfuss, 1992), one of the major components of the

heterogeneous nuclear RNA-protein complex (hnRNP). U

protein had been characterized as an RNA binding protein by

UV-induced cross-linking experiments in vivo (Dreyfuss et

al., 1984) and suggested to be involved in the processing of

pre-mRNA (Portman and Dreyfuss, 1994). Therefore,

SP120/hnRNP U protein is a multifunctional protein

operating in those nuclear processes. Using partial cDNA

segments expressed in E. coli, a putative MAR binding site

of SP120 has been located within the C-terminal tail region

of 171 amino acids, designated RG domain, that contains 16

repeats of Arg-Gly dipeptide (Tsutsui et al., unpublished).

D. Aims of this study

Molecular interactions involving MAR at the base of

chromatin loops can be quite complex and probably protein-

protein interactions would also be important in the

stabilization of "anchorage complex". What determines the

binding specificity of MARs toward the matrix, therefore, is

a difficult question to answer. However, it would be

reasonable to assume that relatively small numbers of

proteins play a key role to recognize MAR sequences and

serve as a nucleation center or a platform for subsequent

association of additional proteins. To evaluate this model,

sequence motifs recognized by these MAR binding proteins

must be nailed down within individual MARs. The proteins

listed above are abundant nuclear proteins and expressed

ubiquitously except for SATB1 which is specific to thymus.

Although consensus motifs for the binding of ARBP

(Buhrmester et al., 1995) and SATB1 (Dickinson et al.,

1992) have been proposed, those for the other proteins are

less characterized.

Before searching for the sequence motifs recognized by

SP120, we first tried to identify essential regions within a

representative MAR reported previously. In the first part of

this article, relative affinities of subfragments from the

mouse Ig! MAR are analyzed in an in vitro binding reaction

with isolated nuclear matrix enriched with SP120. The

results suggest an interesting possibility that the MAR

activity is elicited through a synergistic interaction of two

subdomains separated no farther than 300 bp (Okada et al.,

1996). The second part describes experiments with purified

SP120 and synthetic concatemers, which suggested that

multiple An/Tn tracts positioned in a particular configuration

are essential for the recognition by SP120.

II. Analysis of subdomain structures innatural MAR

A. Methods for assessment of MAR activity

All the data presented here are based on the in vitro

binding assay in which a preferential binding of end-labeled

MAR fragments to isolated nuclear matrix is measured in the

presence of a large excess of unlabeled competitor DNA

(Tsutsui et al., 1993). The matrix preparation used here is

similar to the "nuclear scaffold" in that histones were

extracted with lithium diiodosalicylate after stabilization of

nuclei with Cu2+ (Tsutsui et al., 1988). SP120 is highly

concentrated in the matrix, being a single major MAR

binding protein detectable by southwestern blotting. The 32P-

labeled DNA fragments bound to the matrix was directly

counted in a scintillation counter or analyzed by

electrophoresis in 5% polyacrylamide gels after digesting the

matrix with proteinase K. The radioactive DNA bands were

quantified by densitometric scanning of autoradiograms.

B. The intronic MAR of mouse Ig gene

The MAR localized within the mouse immunoglobulin k

gene is a typical example of AT-rich MARs except for its

intronic location, consisting of a stretch of DNA (370 bp)

with high A+T content (70%). This MAR is interesting in

functional aspects since, together with the nearby

transcriptional enhancer, it has been shown to be required as

a cis element for the enhanced expression (Xu et al., 1989),

demethylation (Lichtenstein et al., 1994), and somatic

hypermutation (Betz et al., 1994) of Ig! transgenes.

As shown in Figure 1 , numerous short stretches of

consecutive A's and T's are contained in the MAR.

Importance of AT-tracts for protein recognition has been

noticed in previous studies (Adachi et al., 1989; Käs et al.,

1989). Comparison of sequences for known MARs strongly

suggested that these tracts (referred to as A- or T-patches

hereafter) are a characteristic landmark for MARs. Another

interesting observation is that the occurrence of both patches

appears to be essential for the MAR activity since a biased

abundance of T-patches, for instance, shows a poor

correlation.


583

Figure 1 . Short homopolymeric AT-tracts cluster in the intronic MAR of mouse immunoglobulin ! gene.

Homopolymeric stretches (n"3) of adenines (A-patch) and thymines (T-patch) are highlighted by light shade and dark shade,

respectively. Cleavage sites for the restriction enzymes used in this study are indicated.

Figure 2 . Relative binding

affinities of restriction fragments to the

nuclear matrix.

Ig! MAR was cut with the enzymes

shown on the left side (also abbreviated

on the line map. A, AluI; S, SspI; M,

MboII; Dd, DdeI; Dr, DraI). Unlabeled

internal fragments are omitted from the

figure. Fragments with high, low, and

negligible (1%>) affinities are

represented by filled, hatched, and open

bars, respectively. The boxed figures

stand for percentages of input fragments

bound to the matrix.

C. Ig MAR is composed of two elementarysubdomains

As the minimal region required for the binding activity of

this MAR had not been analyzed, we first compared the

binding affinities of several restriction fragments

(summarized in Figure 2 ). Under the experimental

conditions used here, about 20% of the full length MAR

input is retained on the matrix. Both fragments generated by

a cleavage at the unique MboII site show a much reduced, but

still significant binding of 5%. This level of binding can be

easily overlooked but it has a prime importance as a contri-


584

Figure 3 . Effects of dimerization of MAR fragments.

(a) Self-dimerization of the MboII fragments. (b) Dimerization of the MboII-AluI fragment. pMM22 is a control construct with an

extension of nonMAR segment derived from the # lactamase gene.

bution from elementary local structures, because the binding

activity can be largely restored by their self-dimeriza-tion (see

below). Further fragmentation of the MboII 3'-fragment

suggests that the region between the MboII site and the

downstream AluI site is required for the residual binding.

Does the large decrease in the affinity after cleavage with

MboII simply imply that the whole intact MAR is essential

for the high affinity? The experiment shown in Figure 3a

reveals, however, that this is not the case. Self-dimerization

of the 5' half (S) or 3' half (L) of Ig! MAR greatly increases

their affinity to the matrix. This effect is more evident with

the latter half; the dimer fragment (2L) shows even greater

affinity than the original MAR.

The binding experiments with restriction fragments

(Figure 2) suggested that the 3' portion of L fragment is

dispensable. When the DNA segment between the MboII site

and the 3' AluI site is dimerized , a similar level of

enhancement or synergism is observed (Figure 3b). This

effect is not due to the doubled fragment length since a

control experiment with the monomer fragment extended by

a nonMAR sequence (A+T=50%) showed no effect.

D. Synergism between the subdomains

The MboII-AluI segment may thus serve as a minimal

sequence element that confers MAR-like features upon any

DNA, provided that it is duplicated in the same molecule.

This possibility can be tested by inserting nonMAR spacers

of varying length between the monomer units of the dimer

clone (Figure 4). As expected, DNA fragments flanked by

the duplicated elements are strong MARs as long as they are


585

not separated by more than 300 bp. As the spacer length

exceeds 500 bp, however, the binding affinity decreases

abruptly and levels off to the monomer level at about 3 kb.

Therefore, the synergistic effect does not propagate beyond

500 bp.

It appears that each subdomain is complexed with a

subset of matrix proteins and a cooperative interaction

between these complexes contributes to the increased

affinity. Obviously, the complexes should be positioned

within 300 bp for the interaction to occur. This excludes a

simple DNA-looping-out mechanism. The bipartite

subdomain organization could be a general feature of AT-

rich MARs. Similar observations have been reported

previously on the Drosophila histone gene MAR (Gasser and

Laemmli, 1986) and chicken lysozyme 5' MAR (von Kries

et al., 1991). This principle may also be applied to the

Drosophila fushitarazu gene MAR (Tsutsui et al.,

unpublished).

Figure 4 . Attenuation of the synergism between MAR

subdomains as a function of increasing distance.

The dimerized MboII-AluI fragments (fat arrows) were

separated by insertion of nonMAR spacers derived from plasmid

pUC18 (thick lines).

III. Synthetic concatemers as an artificialMAR

A. Initial hints

The frequent occurrence of AT-patches in MARs suggests

that they interact directly with MAR binding proteins

including SP120. Once the elementary subdomain (about

130 bp) has been identified in the Ig! MAR, next approach

would be to compare the affinities of mutated subdomain

sequences in which AT-patches are disrupted one by one.

Before completion of this type of experiments, however,

another effective strategy was suggested from a rather

serendipitous finding.

In a purification protocol for a transcription factor IRF-2

using a DNA-affinity resin, a nuclear protein of 120 kDa

copurifies with IRF-2 (Vaughan et al., 1995). Although this

is actually a contaminating protein, it shows a strong

affinity to the concatemerized recognition sequence for IRF-2

as revealed by southwestern blotting. The sequence shown in

Figure 5a resides in the promoter region of human histone

H4 gene but no MAR has been identified nearby. When the

sequence is mutated by an insertion of A/T-pair (Figure 5b),

the binding of IRF-2 is abolished whereas the 120 kD

protein persists to bind, indicating that this protein has less

stringent sequence requirements.

An interesting feature of the IRF-2 recognition sequence

is the presence of two AT-patches in a GC-rich context (A+T

content of the mutant 26-mer is only 42%). When the

mutant duplex is ligated through its sticky ends, two

alternative orientations are expected for each monomer.

Inverted ligation between neighboring monomer units results

in the generation of both A- and T-patches on the same

strand. This implies a variegated presentation of AT-patches

in the ligation product, which is also a characteristic feature

of natural MARs. Based on these observations we suspected

that the protein might be SP120.

B. The concatemer interacts with SP120

Before the synthesis of concatemer to test its possible

binding with SP120, one modification was made on the

mutant sequence (Figure 5c). This base change introduces

restriction sites only between the monomer units ligated in

inverted orientations (Figure 5d), enabling easy

confirmation of the ligation product and cleavage of

concatemers at specific sites. As expected, the concatemer

synthesized by ligating the oligomer duplex shown in

Figure 5c exhibits a strong binding to SP120 on

southwestern blots. Poly(dA)•poly(dT), a synthetic duplex

DNA frequently used as a competitor for AT-rich MARs,

competes the binding effectively, indicating that the

concatemer behaves like AT-rich MARs despite its low AT

content. When the concatemer is digested with either BglII or

FbaI, the binding decreases to a basal level. This is

consistent with the predicted importance of both AT-patches

in the concatemer binding.

The concatemer immobilized on Sepharose beads

selectively binds SP120 in nuclear extracts and can be used

as an affinity resin to purify the protein. The specific binding

of the concatemer was confirmed by a nitrocellulose filter


586

Figure 5 . Sequences of

oligonucleotide duplexes for concatemer

synthesis.

AT-patches are shaded. (a)

recognition motif of IRF-2 (wild type

sequence). (b ) mutant sequence. (c )

modified mutant. (d) structure of a

representative concatemer containing all

possible orientations of neighboring

monomer units (bars). Dark shaded areas

in the bar represent AT-patches.

Restriction sites created after ligation are

also indicated. The upper strand in (c )

containing T-patches are designated "A",

and the complementary lower strand, "B".

Using this definition, the concatemer

sequence can be described either as

AABBA (upper strand) or as BAABB

(lower strand).

binding assay with purified SP120. The dissociation constant

estimated from these data is in the same order as Ig! MAR

(about 3 nM).

C. Concatemers with defined monomerorientations

Use of synthetic concatemers as an artificial MAR is an

extremely powerful strategy to delineate sequence

requirements for the recognition by SP120. Easiness of

obtaining mutated sequences is an obvious advantage.

Inaddition, the GC-rich background of the concatemer should

provide a highly informative environment to assess the

importance of AT-rich motifs in it. Orientation of monomer

units in the concatemer described above are randomized

because of the palindromic nature of the overhang or sticky

ends. In the following sentences this type of concatemer is

referred to as "RC" for random concatemer. It is possible to

synthesize other types of concatemers simply by

manipulating the overhang sequences; "DC" and "AC"

representing direct and alternating orientations, respectively

(Figure 6). Also here, the terminal sequences were designed

so as to introduce junctional restriction sites.

Comparison of affinities of these concatemers to purified

SP120 ,using a filter binding assay or southwestern blotting,

gives an unequivocal result; only RC shows a significant

binding. It should be emphasized that all the concatemers

have identical AT content and the same number of AT-

patches. Thus, the result strongly suggests that the binding

activity is accounted for by only a subset of RC molecules

with a particular pattern of patch distribution that are

contained in the heterogeneous population.

D. Patch mutants

The possible involvement of AT-patches in the protein

recognition can be tested more directly with RCs with altered

patches. Three categories of mutant sequences were designed

and RCs were synthesized by random ligation of mutant

oligos (Figure 7). In the "patch disruptants", both T3- and

T4-patches are disrupted by replacing T's with G/C or A. In

the "patch convertant", only the T4-patch is converted to an

A4-patch. All these concatemers show a marginal binding to

SP120, indicating that the intactness of AT-patches are

indeed required for the binding.

In the "phase mutants", the two T-patches are either

moved toward the center without changing the spacing, or

the distance between them are elongated. Distribution pattern

of AT-patches in these concatemers is different from that in

the wild type RC. Although a decreased binding is observed

with these mutants, a significant level of binding persists.

An important conclusion from this experiment is that

functional patches are homopolymeric and patches with

mixed A's and T's have no contribution to the binding. The

decreased affinity of the patch convertants, whose A/T patch

ratio is the same as in wild RC, suggests again the

distribution pattern of patches is the determinant factor for

successful binding.


587

Figure 6 . Synthesis of concatemers with defined monomer orientations.

Complementary oligonucleotide pairs shown in the left column were annealed and ligated. Junctional restriction sites created in the

concatemers are shown in shaded boxes and expected restriction fragments are shown in the right column. Structure of the concatemers

were confirmed by fragmentation analysis with these enzymes.

Figure 7 . Oligonucleotides used for the synthesis of patch

mutants. Only the T-patch strands are shown. Positions of T-

patches and altered bases are highlighted.

E. Hexamer is the shortest concatemer withbinding activity

Under the ligation conditions used here, the resulting

RCs are heterogeneous in size, ranging from 50 bp (dimer)

to over 2,000 bp. It is anticipated that longer RC binds more

tenaciously to SP120. To determine the minimal size of RC

capable of interacting with SP120 with significant affinity,

RC was size-fractionated in agarose gels and

oligoconcatemers were extracted from the gel. When

individual oligomers were subjected to the filter binding

assay, a significant binding was observed with oligomers

longer than hexamer but not with shorter ones (dimer to

pentamer).

This result suggests that 156 bp is the minimal length to

harbor all essential AT-patches supporting the MAR

activity. Active hexamer units may be equivalent to the

elementary subdomains of Ig! MAR, and when they are

duplicated in RC, synergistic interaction between them

should largely promote the binding. Interestingly, the size of

the hexamer is comparable to that of the Ig! MAR

subdomain (about 130 bp).

In light of available evidence it should be possible to

determine the configuration of essential AT-patches that

confers MAR activity to RC. One way to do this is to clone

the heterogeneous concatemer in a plasmid vector and

examine the binding activity of each clone. Subsequent

alignment of a battery of sequences for active concatemers

should reveal the pattern of essential patches. Although we

have made several trials with different cloning systems, no

clones with desired inserts have been obtained probably due

to the instability of repetitive sequences in host bacteria.


588

Therefore, we decided to synthesize all possible

hexaconcatemers by a similar strategy as described above.

There exist 36 different hexamers theoretically (Figure 8 ).

To synthesize all the hexamers, 26 different 26-mer

oligonucleotides with complementary overhangs are required.

Six sets of complementary oligos are first annealed pairwise

and then mixed and ligated. The resulting concatemer is

heterogeneous in size but the monomer units are ligated in

desired orientations and BamHI sites are introduced at every 6

units. Digestion with BamHI releases a hexamer which can

be purified by polyacrylamide gel electrophoresis.

Experimental results with hexamers are not shown here since

the work along this line is still under way.

F. MAR organization and mode ofinteraction with proteins

Based on the results presented in this article, I propose

here the following principles for MAR organization. 1)

MAR activity is exerted by a synergistic interaction between

at least two subdomains located no farther than 300 bp apart.

2) Each subdomain contains essential AT-patches positioned

with particular intervals. The latter principle was deduced

from the binding experiments with concatemers and purified

SP120. This could be generalized, however, to any MARs

because a similar mode of MAR-protein interaction appears

to involve other abundant MAR binding proteins, e.g.,

histone H1 and topoisomerase II.

It is possible that essential AT-patches are the sites

interacting directly with MAR binding proteins. The Arg,

Gly-rich C-terminal domain of SP120 may selectively

recognize AT-patches by its insertion into the small groove

of DNA which is significantly narrowed at homopolymeric

AT-tracts. As this type of interaction is usually weak,

multiple contacts between MAR and proteins would be

essential for a stable complex to form. The complex can be

formed either by disordered aggregation of proteins or by

organized interaction between protein molecules (Figure 9).

In the latter case the MAR subdomain is likely to be folded

in a particular way. This model is also consistent with the

ordered positioning of AT-patches in the subdomain.

It should be pointed out finally that other factors like

bent structure or inverted repeats might also be involved

since these motifs are indeed enriched in the concatemers. We

have not done any systematic experiments to test this

possibility.

IV. Prediction of MAR from sequenceinformation

As genome projects on human and other species advance,

sequence databases are increasingly flooded with

uncharacterized genomic sequences. Since MAR is related to

many important nuclear functions, it would greatly facilitate

the prediction of functional organization of unknown

genomic regions if one could predict the precise genomic

location of MARs by using computer-assisted search

methods. This information may also be useful in gene

manipulation technologies including gene therapy. The

common strategy adopted in recent works along this line is

to search for a variety of sequence patterns that are known to

occur frequently in MARs, e.g., AT- or GT-rich short tracts,

curved or kinked DNA motifs, or topoisomerase II cleavage

motifs. The statistical significance calculated for observed

frequencies of these patterns are then combined to derive a

probability function ("MAR potential") which is computed

over the region of interest using a sliding window algorithm

(Singh et al., 1997). This survey protocol successfully

detected known MARs in several genomic regions.

We have applied the MAR organization principle

formulated above to predict MARs in various DNA regions

containing known MARs and obtained satisfactory results

with considerably small incidence of false negatives and false

positives. The method also predicted MARs in some

genomic regions where presence of MAR had not been

assessed experimentally. The MAR activity of candidate

regions was confirmed later by binding assays with isolated

nuclear matrix (Tsutsui et al., unpublished).

Acknowledgments

I thank Drs. Shoshiro Okada and Kimiko Tsutsui for

their contributions to this work, and Mika Ishimaru for

technical assistance.

References

Adachi, Y., Käs, E., and Laemmli, U.K. (1 9 8 9 ) Preferential,

cooperative binding of DNA topoisomerase II to scaffold-

associated regions. EMBO J. 8, 3997-4006.

Berezney, R., and Coffey, D.S. (1 9 7 7 ) Nuclear matrix.

Isolation and characterization of a framework structure from

rat liver nuclei. J . C e l l B i o l . 73, 616-637.

Betz, A.G., Milstein, C., González-Fernández, A., Pannell, R.,

Larson, T., and Neuberger, M.S. (1 9 9 4 ) Elements

regulating somatic hypermutation of an immunoglobulin

kappa gene: critical role for the intron enhancer/matrix

attachment region. Cel l 77, 239-248.

Boulikas, T. (1 9 9 5 ) Chromatin domains and prediction of MAR

sequences. I n t . R e v . C y t o l . 162A, 279-388.

Buhrmester, H., von Kries, J.P., and Strätling, W.H. (1 9 9 5 )

Nuclear matrix protein ARBP recognizes a novel DNA

sequence motif with high affinity. Biochemis try 34,

4108-4117.

Cockerill, P.N., and Garrard, W.T. (1 9 8 6 ) Chromosomal loop

anchorage of the kappa immunoglobulin gene occurs next to

the enhancer in a region containing topoisomerase II sites.

Cel l 44, 273-282.


589


590

Figure 8 . List of all possible hexaconcatemers.

Monomers with the alternative orientations are shown by red and yellow bars. T- and A-patches are painted dark blue and green,

respectively.

Figure 9 . Models for organization of

MAR subdomains by interacting proteins.

(a) Complex formation by disordered

aggregation of proteins. (b ) Formation of

an organized complex aided by specific

protein-protein interactions. AT-patches

are shown by dark disks and proteins by

shadowed spheres.

Cook, P.R., and Brazell, I.A. (1 9 7 5 ) Supercoils in human DNA.

J . Ce l l Sc i . 19, 261-279.

Dickinson, L.A., Joh, T., Kohwi, Y., and Kohwi-Shigematsu, T.

(1 9 9 2 ) A tissue-specific MAR/SAR DNA-binding protein

with unusual binding site recognition. Cel l 70, 631-645.

Dickinson, L.A., and Kohwi-Shigematsu, T. (1 9 9 5 ) Nucleolin

is a matrix attachment region DNA-binding protein that

specifically recognizes a region with high base-unpairing

potential. M o l . C e l l . B i o l . 15, 456-465.

Dreyfuss, G., Choi Y.D., and Adam S.A. (1 9 8 4 )

Characterization of heterogeneous nuclear RNA-protein

complexes in vivo with monoclonal antibodies. M o l .

C e l l . B i o l . , 4, 1104-1114.

Gasser, S.M., and Laemmli, U.K. (1 9 8 6 ) The organization of

chromatin loops: characterization of a scaffold attachment

site. EMBO J. 5, 511-518.

Gasser, S.M., Amati, B.B., Cardenas, M.E., and Hofmann, J.F.-

X. (1 9 8 9 ) Studies on scaffold attachment sites and their

relation to genome function. I n t . R e v . C y t o l . 119, 57-

96.

Ide, T., Nakane, M., Anzai, K., & Andoh, T. (1 9 7 5 ) Supercoiled

DNA folded by non-histone proteins in cultured mammalian

cells. Nature 258, 445-447.

Izaurralde, E., Käs, E., and Laemmli, U.K. (1 9 8 9 ) Highly

preferential nucleation of histone H1 assembly on scaffold-

associated regions. J . M o l . B i o l . 210, 573-585.

Käs, E., Izaurralde, E., and Laemmli, U.K. (1 9 8 9 ) Specific

inhibition of DNA binding to nuclear scaffolds and histone

H1 by distamycin. The role of oligo(dA).oligo(dT) tracts. J .

M o l . B i o l . 210, 587-599.

Kiledjian, M., and Dreyfuss, G. (1 9 9 2 ) Primary structure and

binding activity of the hnRNP U protein: binding RNA

through RGG box. EMBO J. 11, 2655-2664.

Lichtenstein, M., Keini, G., Cedar, H., and Bergman, Y. (1 9 9 4 )

B cell-specific demethylation: a novel role for the intronic

kappa chain enhancer sequence. Cel l 76, 913-923.

Ludérus, M.E., de Graaf, A., Mattia, E., den Blaauwen, J.L.,

Grande, M.A., de Jong, L., and van Driel, R. (1 9 9 2 )

Binding of matrix attachment regions to lamin B1. Cel l 70,

949-959.

Mirkovitch, J., Mirault, M.-E., and Laemmli, U.K. (1 9 8 4 )

Organization of the higher-order chromatin loop: specific

DNA attachment sites on nuclear scaffold. C e l l 39, 223-

232.

Okada, S., Tsutsui, K., Tsutsui, K., Seki, S., and Shohmori, T.

(1 9 9 6 ) Subdomain structure of the matrix attachment region

located within the mouse immunoglobulin ! gene intron.

Biochem. B iophys . Res . Commun. 222, 472-477.

Portman, D.S., and Dreyfuss, G. (1 9 9 4 ) RNA annealing

activities in HeLa nuclei. EMBO J. 13, 213-221.

Romig, H., Fackelmayer, F.O., Renz, A., Ramsperger, U., and

Richter, A. (1 9 9 2 ) Characterization of SAF-A, a novel

nuclear DNA binding protein from HeLa cells with high

affinity for nuclear matrix/scaffold attachment DNA

elements. EMBO J. 11, 3431-3440.

Singh, G.B., Kramer, J.A., and Krawetz, S.A. (1 9 9 7 )

Mathematical model to predict regions of chromatin

attachment to the nuclear matrix. Nucle ic Acids Res . 25,

1419-1425.

Tsutsui, K., Tsutsui, K., & Muller, M.T. (1 9 8 8 ) The nuclear

scaffold exhibits DNA-binding sites selective for

supercoiled DNA. J Biol Chem 263, 7235-7241.

Tsutsui, K., Tsutsui, K., Okada, S., Watarai, S., Seki, S., Yasuda,

T., and Shohmori, T. (1 9 9 3 ) Identification and

characterization of a nuclear scaffold protein that binds the

matrix attachment region DNA. J. B i o l . Chem. 268,

12886-12894.

Vaughan, P.S., Aziz, F., van Wijnen, A.J., Wu, S., Harada, H.,

Taniguchi, T., Soprano, K.J., Stein, J.L., and Stein, G.S.

(1 9 9 5 ) Activation of a cell-cycle-regulated histone gene by

the oncogenic transcription factor IRF-2. Nature, 377,

362-365.

von Kries, J.P., Buhrmester, H., and Strätling, W.H. (1 9 9 1 ) A

matrix/scaffold attachment region binding protein:

identification, purification, and mode of binding. Cel l 64,

123-135.


591

Xu, M., Hammer, R.E., Blasquez, V.C., Jones, S.L., and Garrard,

W.T. (1 9 8 9 ) Immunoglobulin kappa gene expression after

stable integration. II. Role of the intronic MAR and

enhancer in transgenic mice. J. B i o l . Chem. 264,

21190-21195.

Zlatanova, J.S., and van Holde, K.E. (1 9 9 2 ) Chromatin loops

and transcriptional regulation. Crit . Rev . Eukar . Gene

Expr. 2, 211-244.


591


Replicon map of the human dystrophin gene:asymmetric replicons and putative replicationbarriers

Lilia V. Verbovaia1,2 and Sergey V. Razin1,3.

1Institute of Gene Biology RAS, Vavilov St. 34/5, 117334 Moscow, Russia. 2International Centre for Genetic Engineering

and Biotechnology, Padriciano 99, I-34012 Trieste, Italy. 3Institut Jack Monod, 2, place Jussieu-tour 43, 75251 Paris,CEDEX 05, France.

_________________________________________________________________Correspondence to: Sergey V. Razin Tel: +7-095-135 97 87; Fax: +7-095-135 41 05; E-mail: [email protected]

Summary

Using the replication direction assay and oligonucleotide probes designed on the basis of the known exon sequencesof the human dystrophin gene we have made a replicon map of this giant gene. It has been found that dystrophin geneis organized into at least six replicons ranging in size from 170 to more than 500 kb. One of the replicon junctions(sites of replication termination) was mapped in intron 44, i.e. roughly in the same area where the majorrecombination hot spot is located. It is also worth mentioning that the central part of the dystrophin gene (exons 8 -48) is organized into relatively short symmetrical replicons surrounded by two extended regions of apparentlyunidirectional replication (exons 1 - 8 and exons 49 - 64). These observations suggest for the first time that thereshould be certain signals for the termination of replication in euchromatic areas of the genome of higher eukaryotes.Furthermore, it may be concluded that the replication of the central part of dystrophin gene must be completedmuch faster than the replication of its ends. This may induce some topological stresses resulting in an increased rateof chromosomal rearrangements within this gene. The experimental approach used in our study may be helpful forfast analysis of the replication structure of other areas of the human genome provided that these areas are saturatedwith STS markers.

I. Introduction

The human dystrophin gene is the largest gene so faridentified and characterized. It extends over 2 mb on theshort arm of the X-chromosome (Burmeister et al., 1988).This gene frequently undergoes different rearrangementscausing Duchenne or Becker muscular dystrophy(Wapenaar et al., 1988; Den Dunnen et al., 1989;Blonden et al., 1991). Analysis of the replicationstructure of the dystrophin gene may give new insightinto the mechanisms of this gene rearrangement as itseems probable that at least some recombination eventsoccur in connection with DNA replication.

It has long been shown that the genome of highereukaryotes is replicated as a set of quazi-independentreplication units (replicons). Each replicon seems topossess a specific site (or area) where the replicationstarts (for a review see Hamlin, 1992; DePamphilis,1993; Hamlin and Dijkwel, 1995). As far as the sites oftermination of DNA replication (i.e. replicon junctions)are concerned, the situation seems to be less clear.

Although these sites can be mapped using the analysis ofreplication polarity (see below and also Handeli et al.,1989), it is possible that their positions are determinedsimply by a distance from the replication origins and bythe speed of replication forks progression. Such is indeedthe case in the simian virus 40 circular genome, as theinsertion in one arm of the SV-40 replicon of a DNAsequence element retarding the progression of thereplication fork was found to cause a displacement of thereplication termination site in the direction of the moreslowly moving replication fork (Rao et al., 1988; Rao,1994). In yeast cells the termination of replication doesnot occur at specific places determined (at least in non-nucleolar regions) by any specific DNA sequenceelement. It appears to be a consequence of converging ofthe replicating forks within a relatively broad region (Zhuet al., 1992). At the same time, some DNA sequencespausing the replication forks progression (such as thetranscription termination signal for RNA polymerase I)were reported to serve as preferential sites of replication

Verbovaia and Razin: Replicon Map of the Human Dystrophin Gene

592

termination in yeast and mammalian cells (Umek et al.,1989; Kobayashi et al., 1992; Little et al., 1993).

One may be surprised to realise how little we knowabout replication structure of DNA of higher eukaryotes.Even the average size of replicons constitutes a matter ofdiscussion. The common view is based on the results ofDNA fiber radioautography studies carried out more then20 years ago. These studies lead to a conclusion thatDNA of higher eukaryotes is organized in clusters ofsimultaneously working replicons. The size of individualreplicons within a cluster was estimated as 50 to 300 kb(Huberman and Riggs, 1966, 1968; Callan, 1974;Stubblefield, 1974; Edenberg and Huberman 1975;Painter, 1976). This common interpretation of the DNAfiber radioautography data was, however, questioned byLiapunova and coauthors who presented arguments forthe much larger size of replicons (150-900 kb) inmammalian cells and for the absence of replicon clusters(Yurov Yu. B. and Liapunova, 1977; Liapunova, 1994).Several procedures for mapping replication origins in

Figure 1. A scheme illustrating the experimental procedureused to determine the polarity of leading DNA strand synthesis.The nascent DNA chains in a replication loop are shown bythick arrows. Short arrows show ligated Okazaki fragments(synthesised before addition of emetine). The scheme is basedon the data of Burhans et all. (1991) who have demonstratedthat emetine induce imballanced DNA synthesis. Althoughbased on a wrong assumption, the protocol for determining thepolarity of leading DNA synthesis was developed two yearsearlier by Handeli et al. (1989).

mammalian genome have been developed recently (forreview see Hamlin, 1992; Vassilev and DePamhilis,1992; DePamphilis, 1993; Hamlin and Dijkwel, 1995).However, most of these procedures are not suitable forthe analyzis of replication structure of large genomicareas. Only one modern protocol, namely that based onthe determination of the polarity of leading DNA strandsynthesis (Handeli et al., 1989; Burhans et al., 1991) maybe used for this purpose as it is relatively simple andpermits the approximate positions of both replicationorigins and termination sites to be mapped.

Here we are presenting a replicon map of thedystrophin gene constructed using the replicationdirection assay. It has been found that this gene isorganized into at least six replicons ranging in size from170 to more than 500 kb. One of the replicon junctions(sites of replication termination) was mapped in intron44, i.e. roughly in the same area where the majorrecombination hot spot is located (Wapenaar et al., 1988;Den Dunnen et al., 1989; Blonden et al., 1991). Theexperimental approach used in our study (utilization ofoligonucleotide probes in the replication direction assay)may be helpful for fast analysis of the replicationstructure of other areas of the human genome providedthat these areas are saturated with STS markers.

II. Results

A. Mapping approach

Determination of the polarity of leading DNA strandssynthesis became possible due to the demonstration thatthe inhibition of protein synthesis in proliferating cellspreferentially suppresses the synthesis of thediscontinuous (lagging) DNA strand. Hybridization of thenascent DNA synthesised under these condition withstrand-specific probes can thus be used to assay thepolarity of leading DNA strand synthesis (Handely et al.,1989; Burhans et al., 1991). The principle of the above-described mapping protocol is illustrated in Fig. 1.Although the mechanism of imbalanced synthesis ofleading and lagging DNA strands in the presence ofprotein synthesis inhibitors is still not known, the validityof the approach has been verified in experiments withdifferent genomic areas (Handely et al., 1989; Burhans etal., 1991; Kitsberg et al., 1993) and can hardly bequestioned. It was originally proposed to use as strand-specific probes for the replication direction assay theRNA chains transcribed in opposite directions from thesame DNA fragment (Handely et al., 1989). Naturallythese probes could be made only after cloning of thenecessary DNA fragment in an appropriate vector. Inorder to facilitate the mapping protocol we havedeveloped conditions for using 20-mer oligonucleotidesas strand-specific probes. To test the approach we haveanalysed the direction of replication forks movementwithin the domain of chicken alpha-globin genes(Verbovaia and Razin, 1995). The results obtained were


593

in perfect agreement with the previously published dataon mapping the replication origin in this domain.Oligonucleotide probes can be easily washed out from thefilters and the same filters with immobilized nascent andtotal DNA from cells treated with emetine or otherinhibitor of protein synthesis can be used sequentially ina number of hybridization experiments. To study thereplication structure of human dystrophin gene we haveused HEL 92.1.7 cells derived from a male patient as itwas not clear whether the replication structures of theactive and non-active copies of the X-chromosome infemale cells were identical. The cells were cultivated for18 h in presence of emetine and 5-bromo-2'-deoxy-uridine (BrdU) exactly as described by Handeli et al.(1989). (See also Methods section in the end of thispaper). The DNA was then isolated, denatured, sheared toabout 1 kb fragments and nascent DNA chains containingBrdU were separated from the bulk DNA by doubleimmunoprecipitation, as described previously (Vassilevand Russev, 1988). Equal amounts (2 µg) of total DNAand nascent DNA from emetine-treated cells wereimmobilised on nylon filters and hybridized witholigonucleotide probes representing complementary DNAchains.

In order to exclude the possibility of artefacts due tothe uneven sorption of DNA on filters, each filter wassequentially hybridized to probes derived from bothstrands and each pair of probes was hybridized to at leasttwo different filters. In all cases the results of these fourhybridization experiments confirmed each other. Atypical example is shown in Fig. 2. Two similarlyprepared filters with immobilized nascent and total DNAwere hybridized to the "lower chain" and the "upperchain" probes derived from the sequence of the brainpromoter of the dystrophin gene (here and further we use

Figure 2. Reciprocal hybridization of "lower chain" and"upper chain" oligonucleotide probes from the dystrophin genebrain promoter with nascent (nc) and total (tot) DNAimmobilized on two similarly prepared filters. The filter "a" wasfirst hybridized to the "lower chain" probe and then, afterexposure and dehybridization, to the "upper chain" probe. Thefilter "b" was first hybridized to the "upper chain" probe andthen, after exposure and dehybridization, to the "lower chain"probe. Note the preferential hybridization of the nascent DNAwith the "upper chain" probe in both cases.

the designation "upper chain" for the chain which istranscribed into dystrophin pre-mRNA). This experimenthas demonstrated preferential hybridization of the "upperchain" probe to the nascent DNA. After exposure, theprobes were washed off the filters and the "upper chain"probe was hybridized to the filter previously hybridizedto the "lover chain" probe and vise versa. Again,preferential hybridization of the "upper chain" probe withthe nascent DNA was observed. The asymmetry ofhybridization of the "lower chain" and the "upper chain"probes to the nascent DNA remained visible even afterhigh-stringency wash (wash with 0.1X SSC-0.1%SDS for15 min at 420C instead of normally used wash with 1XSSC for 15 min at 420C).

B. Mapping of replication units within thedystrophin gene

To assay the polarity of replication of different partsof the dystrophin gene we prepared 36 pairs ofoligonucleotide probes (Table I). Some of these probeswere made on the basis of the previously describedprimers for STS markers (Coffey, et al., 1992). Theseprobes are referred to by their name in the originalpublication (Coffey, et al., 1992) with the number of acorresponding exon indicated in parentheses. Otheroligonucleotide probes were designed on the basis of theknown primary structure of dystrophin mRNA (Koenig etal., 1987) and the exon-intron structure of the dystrophingene (Roberts et al., 1993). These probes are referred toby the number of a corresponding exon. Approximatepositions of the probes on the physical map of thedystrophin gene are shown in Fig. 3A.

The results of hybridization of the whole set of strand-specific probes with total DNA and nascent DNAsamples enriched in leading strands are shown in Fig. 3B. The polarity of the leading DNA strand synthesis wasfound to switch eleven times within the area under study.Keeping in mind the fact that the replication forks meet atthe termination sites and move in opposite directionsfrom the replication origins one can say that the areaunder study contains 5 replication origins and 6termination sites. The first of the termination sites islocated between the brain and muscle promoters. Indeed,the brain promoter (R24 probes) is replicated in thedirection of dystrophin gene transcription, while themuscle promoter (R22(E1) probes) and exons 2 to 7(probes R12(E2), R13(E3) and R7(E7)) are replicated inthe direction opposite to the direction of transcription.This conclusion follows from preferential hybridizationof the nascent DNA leading strands with the "upperchain" probe of the R24 pair and with the "lower chain"probes of the R22(E1), R12(E2), R13(E3) and R7(E7)pairs, as shown schematically in Fig. 4 . The next switchin replication polarity occurs between exons 7 and 8. Thisis a switch from the minus chain to the plus chain whichis indicative of the presence of a replication originbetween probes R7(E7) and R2(E8) (see the scheme inFig. 4). Similar considerations make it possible to


594

conclude that the replication origins are located betweenexons 28 and 29, between exons 43 and 44, betweenexons 46 and 48 and between exons 64 and 68. Thereplication termination sites are located between probes

87-1 and 87-15, between exons 40 and 43, between exons44 and 45, between exons 48 and 49 and between exons70 and 75.

Table I. Oligonucleotide probes used for determination of the dystrophin gene replication structure.____________________________________________________________________________________________Names of Nucleotide sequence of Nucleotide sequence ofprobes the probe from the "upper" chain the probe from the lower chain____________________________________________________________________________________________R24 CTTTCAGGAAGATGACAGAATC GATTCTGTCATCTTCCTGAAAGR22(E1) CTTTCCCCCTACAGGACTCAG CTGAGTCCTGTAGGGGGAAAGR12(E2) GAAAGAGAAGATGTTCAAAAG CTTTTGAACATCTTCTCTTTCR13(E3) GGCAAGCAGCATATTGAGAAC GTTCTCAATATGCTGCTTGCCR7(E7) CTATTTGACTGGAATAGTGTG CACACTATTCCAGTCAAATAGR2(E8) CCTATCCAGATAAGAAGTCC GGACTTCTTATCTGGATAGGR14(E11) GTACATGATGGATTTGACAGC GCTGTCAAATCCATCATGTAC87-1 CTATCATGCCTTTGACATTCCA TGGAATGTCAAAGGCATGATAG87-15 ATAATTCTGAATAGTCACA TGTGACTATTCAGAATTATR21(E25) CAATTCAGCCCAGTCTAAAC GTTTAGACTGGGCTGAATTGR25(E27) GCTAAAGAAGAGGCCCAAC GTTGGGCCTCTTCTTTAGCE28 GTTTGGGCATGTTGGCATGAG CTCATGCCAACATGCCCAAACE29 TGCGACATTCAGAGGATAACC GGTTATCCTCTGAATGTCGCAE31 GGCTGCCCAAAGAGTCCTGTC GACAGGACTCTTTGGGCAGCCR16(E33) GTCTGAGTGAAGTGAAGTCTG CAGACTTCACTTCACTCAGACE35 GAAGGAGACGTTGGTGGAAGA TCTTCCACCAACGTCTCCTTCR31(E39) CAACTTACAACAAAGAATCACA TGTGATTCTTTGTTGTAAGTTGR8(E40) GGTATCAGTACAAGAGGCAG CTGCCTCTTGTACTGATACCE43 GTCTACAACAAAGCTCAGGTCG CGACCTGAGCTTTGTTGTAGACE44 GACAGATCTGTTGAGAATTGC GCATTTCTCAACAGATCTGTCR18(E45) CTCCAGGATGGCATTGGCAG CTGCCAATGCCATCCTGGAGR4(E46) ATTTGTTTTATGGTTGGAGG CCTCCAACCATAAAACAAATE48 GTTTCCAGAGCTTTACCTGA TCAGGTAAAGCTCTGGAAACE49 ACTGAAATAGCAGTTCAAGC GCTTGAACTGCTATTTCAGTE50 GAAGTTAGAAGATCTGAGCTC GAGCTCAGATCTTCTAACTTCE53 CAGAATCAGTGGGATGAAGTA TACTTCATCCCACTGATTCTGE54 CCAGTGGCAGACAAATGTAG CTACATTTGTCTGCCACTGGE55 TGAGCGAGAGGCTGCTTTGG CCAAAGCAGCCTCTCGCTCAR20(E56) GGTGAAATTGAAGCTCACAC GTGTGAGCTTCAATTTCACCE60 ACTTCGAGGAGAAATTGCGC GCGCAATTTCTCCTCGAAGTE61 GCCGTCGAGGACCGAGTCAG CTGACTCGGTCCTCGACGGCE64 ACTCCGAAGACTGCAGAAGG CCTTCTGCAGTCTTCGGAGTE68 TAAGCCAGAGATTGAAGCGG CCGCTTCGATCTCTGGCTTAE70 ACATCAGGAGAAGATGTTCG CGAACATCTTCTCCTGATGTE75 CTGCAAGCAGAATATGACCG CGGTCATATTCTGCTTGCAGR5(E79) CAGAGTGAGTAATCGGTTGG CCAACCGATTACTCACTCTG

Figure 3 (Following page). Determining replication polarity within the dystrophin gene. (A) A scheme illustrating the exon-intronstructure of the dystrophin gene and the results of determination of replication polarity. On the map of the dystrophin gene the exons areshown by vertical dark bars. Each tenth exon is indicated by the number. Positions of the brain and muscle promoters are shown byarrows above the map. The results of the analysis of replication direction are shown below the map. The vertical bars indicate thepositions of the probe pairs used to assay the replication polarity. The direction of replication determined by hybridization of nascentDNA with each of the probe pairs is shown by horizontal arrows. Approximate positions of the origins ( ori ) and termination sites ( t ) are

indicated above the arrows. (B) Hybridization of strand-specific probes with total DNA (tot) and nascent DNA (nc) from emetine-treatedcells. The names of the probe pairs are indicated above the autoradiographs. "-" and "+" indicate the results of hybridization with probesderived from the lower and the upper chains, respectively.


595


596

Figure 4. A scheme illustrating the interpretation of theresults of hybridization of strand-specific probes with DNAsamples enriched in nascent DNA leading strands. The upperchain and the lower chain probes are designatedcorrespondingly by "+" and "-".

III. Discussion

A. The size of replicons

The present study has demonstrated for the first timethat a single gene may be organized into severalreplicons. The average size of replicons mapped withinthe area under study constitutes 500 kb (with variationsfrom 170 to 1000 kb). This finding contradicts to thecommon view that the average sizes of replicons inmammalian cells are from 50 to 300 kb. However ourobservations are in perfect agreement with theestimations of replicon sizes made by Liapunova andYurov (reviewed by Liapunova, 1994). Furthermore,analysis of the temporal order of DNA replication in theH-2 mouse majour histocompactibility complex alsosuggested that mammalian replicons are larger then 300kb (Spack et al., 1992). Similar conclusion follows fromthe results published by Bickmore and Oghene (1996).

B. Asymmetrical replicons and replicationbarriers.

The results of the present study demonstrate that inthe human genome the replicons may be asymmetrical.Indeed, an extended (500 kb) region including exons 49 -64 seems to be replicated unidirectionally. The oppositearm of the same replicon is relatively small (less than 100kb). It is possible that the left end of the dystrophin gene(500 kb DNA stretch) is also replicated unidirectionally.At least all exons scattered along this region arereplicated in the same direction. Some of the replicationtermination sites mapped in the present study are notlocated at the middle of the distance between twoneighbouring origins. This suggests that there should besome specific signals determining positions oftermination sites. Up to now the replication barriers ofthis kind were observed only in yeast and mammalianribosomal genes clusters (Umek et al., 1989; Kobayashiet al., 1992; Little et al., 1993).

C. The replication structure of thedystrophin gene and recombination hot-spots

It may be of interest that one of the replicationjunctions (termination sites) identified in the presentstudy is located in intron 44, i. e. roughly colocalizes withthe main recombination hot-spot in the dystrophin gene(Wapenaar et al., 1988; Den Dunnen et al., 1989;Blonden et al., 1991). Although the significance of thiscolocalization (if any) is not presently clear, it is worthmentioning that in prokaryotic cells the sites ofreplication termination have long been known toconstitute recombination hotspots (Bierne et al., 1991;Horiuchi et al., 1994; Horiuchi et al., 1995). Accordingto one of the models, the replication fork posed at atermination site is a weak point on DNA where a double-stranded-break may occur with a high probability(Horiuchi et al., 1995; Michel et al., 1997). Some datasuggest that a similar mechanism may account for theformation of recombination hot-spots also in eukaryoticcells (Horiuchi et al. , 1995). In agreement with this ideait was demonstrated that pausing of the replicationmachinery by certain DNA secondary structures, DNAdamage or DNA-protein interaction cause an increase inthe rate of DNA rearrangements (Bierne and Michel,1994). It is known that in eukaryotic cells finalization ofDNA replication (juncture of neighbouring replicons) is arelatively slow process. During this step the replicationforks retain single-stranded regions which can berelatively easy converted into double-stranded breaks.Furthermore, merging of replicons depends on thereactions catalysed by DNA topoisomerases which seemto be able under certain conditions to carry outillegitimate recombination of DNA strands and hence tointroduce deletions and insertions into DNA (Gale andOsheroff, 1992; Shibuya et al., 1994; Henningfeld andHecht, 1995; Bierne et al., 1997).

An interesting feature of the replication structure ofdystrophin gene is that the central part of the gene (exons8 - 48) is organized into relatively short symmetricalreplicons which are surrounded by two extended regionsof apparently unidirectional replication (exons 1 - 8 andexons 49 - 64). Assuming that the rate of replication forksprogression is the same in all replicons, it may beconcluded that the replication of the central part of thegene must be completed much faster than the replicationof its ends. This may cause some topological stressesresulting in an increased rate of chromosomalrearrangements within the dystrophin gene.

IV. Methods

A. Cell culture.

Human erythroleukemia cells HEL 92.1.7 were purchasedfrom the American Type Culture Collection. The cells weregrown in RPMI 1640 medium supplemented with 10% fetalbovine serum.


597

B. Isolation of DNA samples enriched innascent DNA leading strands.

To induce imbalanced synthesis of nascent DNA strands,exponentially growing cells were treated with emetine, asdescribed previously (Handeli et al., 1989; Burhans et al.,1991). Emetine was added to the conditional medium up to a

concentration of 2 µM. This was followed (after 15 min

incubation) by the addition of 5-bromo-2'-deoxy-uridine (10

µg/ml) and 3H deoxy-cytidine (2 µCi/ml). The cells were

cultured in this medium for 16 h. Then they were collected andtheir DNA was isolated. After shearing (to give fragments withan average size of 1 kb) and denaturation of the DNA, theBrdU-labelled nascent DNA chains were separated from thebulk DNA by double immunoprecipitation, as describedpreviously (Vassilev and Russev, 1988).

C. Immobilization of DNA on nylon filtersand hybridization experiments.

Equal amounts (2 µg) of the nascent and bulk DNA were

immobilized on Hybond-N+ nylon filters (Amersham) using aBio-Dot SF microfiltration unit (Bio-Rad). The equivalency ofimmobilization of all probes was verified by hybridization with32P-labelled human repeated sequence of alu type. The

oligonucleotides were labelled with !32P-ATP using T4 phage

polynucleotide kinase, as described previously (Maniatis et al.,1982). Hybridization was carried out in a Rapid Hyb solution

(Amersham) for 1 h at 42 0C. After hybridization, the filterswere washed one time in 5XSSC - 0.1% (w/v) SDS solution for20 min at room temperature and two times (15 min each) in 1X

SSC - 0.1% (w/v) SDS solution at 42 0C. Then the filters were

exposed to the Kodak film at -75 0C with an intensifying screen(Dupont). For dehybridization of the radioactive probes thefilters were incubated in 0.4 M NaOH solution for 30 min at 450C. Then they were neutralized (15 min at room temperature) inthe following solution: 0.1X SSC - 0.1%(w/v)SDS - 0.2M Tris-HCl (pH 7.5).

Acknowledgements

This work was supported by grant N 097 from the RussianState Program "Frontiers in Genetics", by the grant 96-04-49120 from the Russian Foundation for Support of FundamentalScience and by the ICGEB grant CRP/RUS 93-06 to S.V.R.

References

Bickmore WA and Oghene K (1996) Visualizing the spatialrelationships between defined DNA sequences and the axialregion of extracted metaphase chromosomes. Cell 84, 95-104.

Bierne H, Ehrlich SD and Michel B (1991) The replicationtermination signal terB of the Escherichia coli chromosome is adeletion hot spot. EMBO J. 10, 2699-2705.

Bierne H and Michel B (1994) When replication forks stop. Mol.Microbiol. 13, 17-23.

Bierne H, Ehrlich SD and Michel B (1997) Deletions at stalledreplication forks occur by two different pathways. EMBO J.16, 3332-3340.

Blonden LAJ, Grooyscholten PM, Den Dunnen JT, Bakker E.,Abbs S, Bobrow M, Boehm C, van Broeckhoven C, BaumbachL, Chamberlain J, Caskey CT, Denton M, Felicetti L, GallusiG, Fischbeck KH, Francke U, Darras B, Gilgenkrantz H,Kaplan J-C, Hermann FN, Junien C, Boileau C, Liechti-GallatiS, Lindlof M, Matsumoto T, Niikawa N, Muller CR, Poncin J,Malcolm S, Robertson E, Romeo G, Colone AE, Scheffer H,Schroder E, Schwartz M, Verellen C, Walker A, Worton R,Gillard E and Van Ommen GJB (1991) 242 Breakpoints in the200-kb deletion-prone P20 region of the DMD gene are widelyspread. Genomics 10, 631-639.

Burhans WC, Vassilev LT, Wu J, Sogo JM, Nallaseth F andDePamphilis ML (1991) Emetine allows identification oforigins of mammalian DNA replication by imbalanced DNAsynthesis, not through conservative nucleosome segregation.EMBO J. 10, 4351-4360.

Burmeister M, Monaco AP, Gillard EF, Van Ommen G-F, AffaraNA, Ferguson-Smith MA, Kunkel LM and Lehrach H (1988) A10-megabase map of human Xp21 including the Duchennemuscular dystrophy gene. Genomics 2, 189-202.

Callan HG (1974) DNA replication in the chromosomes ofeukaryotes. Cold Spring Harb. Symp. Quant. Biol. 38, 195-204.

Coffey AJ, Roberts RG, Green ED, Cole CG, Butler R, Anand R,Giannelli F and Bentley DR (1992) Construction of a 2.6-mbcontig in yeast artificial chromosomes spanning the humandystrophin gene using an STS-based approach. Genomics 12,474-484.

DePamphilis ML (1993) Eukaryotic DNA replication: anatomy ofan origin. Annu. Rev. Biochem. 62, 29-63.

Den Dunnen JT, Grootscholten PM, Bakker E, Blonden LAJ,Ginjaar HB, Wapenaar MC, van Paassen HMB, vanBroeckhoven C, Pearson PL and Van Ommen GJB (1989)Topography of the Duchenne muscular dystrophy (DMD) gene:FIGE and cDNA analysis of 194 cases reveals 115 deletionsand 13 duplications. Am. J. Hum. Genet. 45, 835-847.

Edenberg HJ and Huberman JA (1975) Eukaryotic chromosomereplication. Annu. Rev. Genet. 9, 245-284.

Gale KC and Osheroff N (1992) Intrinsic intermolecular DNAligation activity of eukaryotic topoisomerase II. Potential rolesin recombination. J. Biol. Chem. 267, 12090-12097.

Gorecki DC, Monako AP, Derry JML, Walker AP, Bernard EA andBernard PJ (1992) Expression of four alternative dystrophintranscripts in brain regions regulated by different promoters.Hum. Mol. Genet. 1, 505-510.

Hamlin JL (1992) Mammalian origins of replication. BioEssays 14,651-659.

Hamlin JL and Dijkwel PA (1995) On the nature of replicationorigins in higher eukaryotes. Current Opin. Genet and Dev.5, 153-161.

Handeli S, Klar A, Meuth M and Cedar H (1989) Mappingreplication units in animal cells. Cell 57, 909-920.

Henningfeld KA and Hecht SM (1995) A model for topoisomeraseI-mediated insertions and deletions with duplex DNAsubstrates containing branches, nicks, and gaps. Biochemistry34, 6120-6129.

Horiuchi T, Fujimura Y, Nishitani H, Kobayashi T and Hidaka M(1994) The DNA replication fork blocked at the Ter site maybe an entrance for the RecBCD enzyme into duplex DNA. J.Bacteriol. 176, 4656-4663.


598

Horiuchi T, Nishitani H and Kobayashi T (1995) A new type of E.coli recombination hotspot which requires for the activity bothDNA replication termination events and the Chi sequence.Adv. Bioph. 31, 133-147.

Huberman JA and Riggs AD (1966) Autoradiography ofchromosomal DNA fibers from Chinese hamster cells. Proc.Natl. Acad. Sci. USA 55, 599-606.

Huberman JA and Riggs AD (1968) On the mechanism of DNAreplication in mammalian chromosomes. J. Mol. Biol. 32, 327-341.

Kitsberg D, Selig S, Keshet I and Cedar H (1993) Replication

structure of the human "-globin gene domain. Nature 366,

588-590.

Kobayashi T, Hidaka M, Nishizawa M and Horiuchi N (1992)Identification of a site required for DNA replication forkblocking activity in the rRNA gene cluster in Saccharomycescerevisiae. Mol. Gen. Genet. 233, 355-362.

Koenig M, Hoffman EP, Bertelson CJ, Monaco AP, Feener C andKunkel LM (1987) Complete cloning of Duchenne musculardystrophy (DMD) cDNA and preliminary genomicorganization of the DMD gene in normal and affectedindividuals. Cell 50, 509-517.

Liapunova NA (1994) Organization of replication units and DNAreplication in mammalian cells as studied by DNA fiberradioautography. Int Rew. Cytol. 154, 261-308.

Little RD, Platt TH, Schildkraut CL (1993) Initiation andtermination of DNA replication in human rRNA genes. Mol.Cell. Biol. 13, 6600-6613.

Michel B, Ehrlich SD and Uzest M (1997) DNA double-strandbreaks caused by replication arrest. EMBO J. 16, 430-438.

Maniatis T, Fritsch EF and Sambrook J (1982) Molecular cloning:a laboratory manual. (Cold Spring Harbor, New York: ColdSpring Harbor Laboratory).

Painter RB (1976). Organization and size of replicons. InHandbook of Genetics, R.C. King, ed. (Plenum, New York).vol. 5, pp. 169-186..

Rao BS, Manor H and Martin RG (1988) Pausing in simian virus

40 DNA replication by a sequence containing (dG-dA)27.(dT-dC)27. Nucl. Acids Res. 16, 8077-8094.

Rao BS (1994). Pausing of simian virus 40 replication fork

movement in vivo by (dG-dA).(dT-dC) tracts. Gene 140, 233-237.

Roberts RG, Coffey AJ, Bobrow M and Bentley DR (1993). Exonstructure of the human dystrophin gene. Genomics 16, 536-538.

Shibuya ML, Ueno AM, Vannais DB, Craven PA and Waldren CA(1994) Megabase pair deletions in mutant mammalian cellsfollowing exposure to amsacrine, an inhibitor of DNAtopoisomerase II. Cancer Res. 54, 1092-1097.

Spack EG, Lewis ED, Paradowski B, Schimke RT and Jones PP(1992) Temporal order of DNA replication in the H-2 majorhistocompactibility Complex of the mouse. Mol. Cell. Biol. 12,5174-5188.

Stubblefield E (1974) The kinetics of DNA replication inchromosomes. In The cell nucleus, H. Busch, ed. (AcademicPress, New York), vol. 2, pp. 149-162. .

Umek RM, Linskens MHK, Kowalski D and Huberman J (1989)New beginnings in the studies of eukaryotic DNA replicationorigins. Biochim. Biophys. Acta 1007, 1-14.

Vassilev L and Russev G (1988). Purification of nascent DNAchains by immunoprecipitation with anti BrdU antibodies.Nucl. Acids Res. 16, 10397.

Vassilev LT and DePamhilis ML (1992) Guide to identification oforigins of DNA replication in eukaryotic cell chromosomes.Crit. Rev. Biochem. Mol. Biol. 27, 445-472.

Verbovaia L and Razin SV (1995) Analysis of the replication

direction through the domain of #-globin-encoding chicken

genes. Gene 166, 255-259.

Wapenaar MC, Kievits T, Hart KA, Abbs S, Blonden LAJ, denDunnen JT, Grootscholten PM, Bakker E, Verellen-DumoulinCh, Bobrow M, van Ommen GJB and Pearson PL (1988). Adeletion hot spot in the Duchenne muscular dystrophy gene.Genomics 2, 101-108.

Yorov YB and Liapunova NA (1977) The units of DNA replicationin the mammalian chromosomes: evidence for a large size ofreplication units. Chromosoma 60, 253-267.

Zhu J, Newlon CS and Huberman J (1992) Localization of a DNAreplication origin and termination zone on chromosome III ofSaccharomyces cerevisiae. Mol. Cell. Biol. 12, 4733-4741.


599


Initiation of DNA replication at the rat aldolase Blocus

—An overlapping set of DNA elements regulates transcription andreplication?—

Ken-ichi Tsutsumi and Yunpeng Zhao

Institute for Cell Biology and Genetics, Faculty of Agriculture, Iwate University, Ueda, Morioka, Iwate 020, Japan

________________________________________________________________________________________________Correspondence to: K. Tsutsumi, Fax: + 81 -19 621 6243, E-mail: [email protected]

Summary

In higher eukaryotes, DNA replication initiates at multiple s i tes on each chromosome.Positioning and firing of the replication origins are not fixed, but different and selected origins mayini t iate at d i f ferent t imes in a s ingle ce l l cyc le of part icular ce l l s through a range of complexmechanisms controll ing, for example, cel l differentiation. The origin region at the rat aldolase Blocus (ori A1) has been found to encompass the promoter which governs liver-specific transcription.Ori A1 is , thus, thought to be a suitable target in investigating causal relationships among thoseunder control of cel l differentiation, i . e . , firing or si lencing of the origin, cel l type-specificregulation of transcription, and posit ioning of nearby origins. In this article , we summarize ourapproach to elucidate such relationships. We describe sequence-dependent replication from ori A1,overlapping of essential regions required for replication and transcription, cel l cycle-regulatedbinding of factors to the essential region, and then chromosomal state of the ori A1 region in thenucleus.

I. Introduction

Initiation of DNA replication is one key control pointin a process of cell cycle progression. In eukaryotic cells,each chromosome contains multiple replication originswhich are activated in a stringently controlled temporal andspatial order during S phase. In addition, not all origins arefired for genome duplication in a single cell cycle, someorigins are generally used while others are not(DePamphilis, 1993a; Coverley and Laskey, 1994; Hamlinet al., 1995; Stillman, 1996). Which origins are selectivelyused and when they are activated are, however, not fullyunderstood. The selection and positioning of origins on thechromosomes might be controlled by multiple regulatoryprocesses such as developmental program, transcriptionactivity of nearby genes, and firing of nearby origins(James and Leffak, 1986; Wolffe and Brown, 1988; Trempeet al., 1988; Leffak and James, 1989; Fangman and Brewer,1992).

Evidences have been accumulating that cis-elements fortranscription promote replication activity as well (forreview see DePamphilis, 1993b). For example,transcription factors AP1 (Guo and DePamphilis, 1992),NF1 (Mul and Van der Vliet, 1992), Oct1 (O'Neill et al.,1988) and c-Jun (Ito, Ko et al., 1996) strongly stimulate

virus DNA replication. In cellular chromosomes, originregion or regulatory region for DNA replication oftenencompasses transcriptional promoter or contains cis-elements for transcription (Vassilev and Johnson, 1990;Ariizumi et al., 1993; Taira et al., 1994; Tasheva andRoufa, 1994; Zhao et al., 1994). Mutant human cellshaving deletions at either near the promoter (Kitsberg et al.,1993) or locus control region (LCR) (Aladjem et al., 1995)of the !-globin locus fail to initiate DNA replication fromthe origin located within the !-globin locus.

On the contrary, several studies suggested thattranscription and DNA replication are antagonistic events.A head-on collision between a replication fork andtranscribing RNA polymerase complex arrests or pausesreplication in yeast (Brewer et al., 1992; Deshpande andNewlon, 1996), in a similar way to E. coli (Liu et al.,1993; Liu and Alberts, 1995). Origin recognition complex(ORC), the eukaryotic replication initiator, repressestranscription of certain genes in yeast (Bell and Stillman,1992; Micklem et al., 1993; Foss et al., 1993; Bell et al.,1995), although the role of ORC in the repression wasrecently shown to be separable from its role in replicationinitiation (Fox et al., 1997).

Tsutsumi and Zhao: Initiation of DNA replication at the rat aldolase B locus

600

A similar inverse correlation occurs in Xenopusribosomal RNA genes, in which replication initiation isspecifically repressed within transcription units and limitedto nontranscribed regions, whereas the replication randomlyinitiates throughout the transcribed and nontranscribedsequences in early embryos (Hyrien et al., 1995).

Taken together, these observations strongly suggest atight link between transcription regulation and positioningor firing of replication origins in the chromosomal context.However, how transcription and replication interact witheach other, and what biological system(s) the interaction isinvolved in mammalian cells are still unknown. In thisreview, we describe our approach to characterize replicationorigin at the rat aldolase B locus. We also discuss on DNAelements required for replication initiation and on thepossible correlation between regulatory systems ofreplication and transcription.

II. The rat aldolase B gene: function,structure and liver-specific expression

Before focusing on replication origin, we start bybriefly describing about the gene coding for aldolase. Aldolase is an enzyme acting on fructose-1,6-bisphosphatemetabolism in a processes of glycolysis andgluconeogenesis. The enzyme is a tetrameric proteincomposed of a combination of three different subunits, A, Band C which are distributed in different tissues and organs inanimals (Horecker et al., 1972). Expression of the geneencoding the aldolase B subunit (AldB) is cell type-specificand is under control of cell differentiation; the gene ispreferentially expressed in the liver, kidney and jejunalmucosa in adult animal, but the expression is repressed atan early fetal stage and in dedifferentiated hepatocellularcarcinomas (Horecker et al., 1972; Numazaki et al., 1984;Tsutsumi et al., 1985; Sato et al., 1987). Studies on themechanisms operating in such a regulated expressionrevealed the importance of at least four cis-elements (sitesA, B, C, and D) on the proximal 200 bp promoter in theliver-specific transcription, to which a number ofregulatory factors interacts cell type-specifically orubiquitously (Fig . 1 ).

For example, bindings of HNF1 to site A, AlF-B orNF-Y to site B, and C/EBP or AlF-C to site C seem toconfer liver-specific transcription (Tsutsumi et al., 1989;Ito, Ki et al., 1990; Raymondjean et al., 1991; Gregori etal., 1993; Tsutsumi et al., 1993; Yabuki et al., 1993;Gregori et al., 1994). Site D acts as a silencer-like elementupon transfection, but its role is still unknown (Gregori etal., 1993). On the other hand, in AldB non-expressingrapidly dividing cells, a different set of factors bind to thesesites though their functions are not fully known, e.g., siteB and site C bind growth-inducible factors Ryb-a (Ito, Ki etal., 1994) and an alternate type of AlF-C (Yabuki et al.,1993), respectively (discussed later). Thus, various set ofregulatory factors interact with the proximal 200 bppromoter, determining cell type-specificity and the level oftranscription.

III. An initiation region of DNA replicationencompasses promoter of the AldB gene inAldB non-expressing rat hepatoma cells

Regulated state of eukaryotic genes is achieved throughthe assembly of specialized, heritable chromatin structurewith confined domains on chromosomes, which mighthave causal relationship with locus control region (LCR),insulator, nuclear matrix-association, and positioning ofreplication origins etc. It is also suggested that regulatorypathways that govern transcription and initiation of DNAreplication affect each other or cross-talk (DePamphilis,1993b; Hamlin et al., 1995; Stillman, 1996). Based onthese considerations, we initially thought thattranscriptional repression of the AldB gene in rapidlydividing cells, such as fetal liver and hepatoma cells, mightsomehow relate to initiation of replication. For a steptoward understanding the functional and positionalrelationship between transcription and replication, we triedto identify the replication initiation region nearest to theAldB gene.

To locate initiation region of replication, newlyreplicated DNA chains were labeled with bromodeoxyuridine(BrdU) using synchronously cultured rat hepatoma dRLh84cells. BrdU-substituted DNA has higher density ascompared to unsubstituted parent DNA and can be separatedby ultracentrifugation through CsCl density-gradient. Cellsarrested at G1/S boundary by double-thymidine-block werereleased from the arrest to enter S phase, and cultured in afresh medium containing BrdU. After various time period,BrdU-labeled DNA was prepared by CsCl isopycniccentrifugation after digestion with an appropriate restrictionenzyme, and hybridized with probes corresponding tovarious regions in and around the AldB gene region. Theseexperiments showed that (i) replication of the AldB generegion starts at mid-S phase, and (ii) the initiation regionlocates near or within the AldB gene region. Furtheranalysis of the newly replicated short DNA fragmentsprepared by alkaline sucrose density-gradient centrifugationrevealed that (iii) the initiation region expands about 1.5 Kbor less, which encompasses transcription promoter of theAldB gene (Fig . 1 ) (Zhao et al., 1994).

IV. Specific sequence is required forreplication of plasmids carrying the AldBorigin fragments

In vivo analyses of newly replicated DNA identified anorigin region of DNA replication which encompassedpromoter of the AldB gene. Since several mammalianorigins have been reported to possess activities to replicateautonomously (for example, Ariga et al., 1987;McWhinney and Leffak, 1990; Wu et al., 1993), we triedto examine whether the AldB origin fragment promotesreplication in a plasmid form upon transfection intomammalian cells. For this purpose, large DNA fragmentsranging from 4 Kb to 6.3 Kb derived from the AldB origin


601

F i g . 1 . Two origin regions A1 and A2 in the vicinity of the AldB locus in rat. Vertical lines represent EcoRI sites and thelengths of the EcoRI fragments are shown in Kb below the map. Lower panel shows structure off the promoter within thereplication origin (A1) region. Transcription factors that interact with the promoter are also shown. Bent arrow indicates positionand direction of the AldB gene transcription.

F i g . 2 . Replication of aplasmid carrying origin fragment(A1) in transfected cells. Plasmidscarrying in vivo origin fragmentsfrom - 5.7 Kb to + 0.625 Kb(pBOR6.3) or from - 0.675 to +0.263 Kb (pBOR0.94) were co-transfected with pUC19 orpBOR6.3, respectively, into Cos-1cells. After the transfected cellswere cultured for 72 hr in thepresence of BrdU, low-molecular-weight DNA was extracted, digestedwith EcoRI, and fractionated byCsCl isopycnic ultracentrifugation.DNA in each fraction was separatedon an 1% agarose gel, transferredonto a nylon membrane, andhybridized with a random-primed,32P-labeled pUC19 DNA fragmentas a probe. LL (light-light ) DNA,HL (heavy-light) DNA, and HH(heavy-heavy) DNA indicateunsubstituted, hybrid, and fullysubstituted DNAs, respectively (seetext).


602

F i g . 3 . The 200 bp AldB gene promoter is essential for initiation of replication in transfected cells. Various deletionconstructs shown in the left panel were transfected together with pBOR6.3 as an internal control and processed as in F i g . 2 . Replication efficiencies were based on the amounts of HH and HL DNAs in total (HH, HL, and LL) DNAs, and values were expressedrelative to the activity of pBOR6.3.

region were inserted into plasmid pUC19, and assayed forautonomous replication in Cos-1 cells based onsemiconservative BrdU-substitution of replicating DNAchain. As briefly mentioned above, incorporation of BrdUinto replicating DNA increases density of the DNA chain(designated as H chain) and causes it to band at a densityhigher than that of unsubstituted DNA chain (L chain) inCsCl density-gradient. The transfected cells were grown inthe presence of BrdU, then low-molecular-weight DNAswere extracted by the procedure described by Hirt (Hirt,1967). Newly-replicated, BrdU-labeled DNA wasfractionated by CsCl isopycnic ultracentrifugation afterdigestion with a restriction enzyme, and then subjected toSouthern blot hybridization. Figure 2 shows a typicalexample of such a replication assay, using a plasmid(designated as pBOR6.3) bearing a 6.3 Kb fragmentextending from - 5.7 Kb to +625 bp in bacterial plasmidvector pUC19. In this case, double-stranded DNA withboth strands being replaced by BrdU (HH DNA) appeared inaddition to HL and LL DNAs. This means, consideringfrom the semiconservative replication, that at least tworounds of replication had occurred. A negative controlplasmid pUC19 co-transfected with pBOR6.3 did notinitiate replication, indicating that the observed replicationdepends on DNA fragment inserted. Plasmid containing the0.94 Kb fragment from - 675 bp to + 263 bp (pBOR0.94)exhibited similar activity as compared to cotransfectedpBOR6.3 (F i g . 2 , lower panel). Thus, the 0.94 Kbfragment extending from -675 bp to +263 bp seems to haveminimum essential components for autonomous

replication. Such replication assays were carried out usingvarious deletion constructs and compared their activities toreplicate. The results indicated that a 200 bp fragmentextending from -200 bp to -1 bp is indispensable toreplication initiation, since deletion of the fragmentabolished replication activity (Fig . 3 ).

The 200 bp fragment alone could not direct replication.But it restored replication activity when ligated to eitherupstream (about 500 bp) or downstream fragment (about300 bp). This observation makes the origin architecturerather complicated. However, since both flanking regionsshare no similar sequence, it is not conceivable that thesame replication elements are present in these two regions.Although entirely unknown at present, one explanation forthis may be the presence of different auxiliary elements ineach flanking region, both of which have similar activitiesfor replication initiation when they cooperate with the 200bp sequence (Fig . 4 ).

So far, controversial observations concerning plasmidreplication in mammalian cells have been reported (reviewedby Coverley and Laskey, 1994). Several of them pointedout, for example, that the length of the DNA template iscrucial, rather than sequence; even bacterial DNA inorigin-depleted mammalian virus vector replicated inmammalian cells (Krysan and Calos, 1991; Heinzel et al.,1991; Krysan et al., 1993). Further, the fact that noconsensus sequences for origins have been found wouldsupport the idea.


603

F i g . 4 . Three important regions inthe predicted origin region. A to D infilled box represent cis-elements fortranscription (see F i g . 1 ). PPu and A/Tindicate purin-rich element having

binding site for a factor Pur " (PUR

consensus sequence) and A/T-richsequence, respectively. Numbersindicate positions in bp relative to thetranscription start site.

In the case of the AldB origin, however, we prefer thatspecific sequence elements rather than length is required forreplication (Zhao et al., 1997). Plasmids used in ourreplication assays do contain none of mammalian virusDNA sequence, origin of virus DNA, and binding sites forvirus T antigen which would activate replication to someextent, so that the observed replication might depend on theAldB origin sequence and is free from initiation machineryfor virus DNA replication. Probably, the possibility thatquite diverse sets of specific sequence elements can promotefiring each of the multiple potential origins onchromosomes might be one of the explanations why noapparently conserved sequence is found in the limitednumbers of mammalian origins so far identified (for review,see Stillman, 1996). Indeed, some eukaryotic origins arereported to require sequence-specific interaction of factors todrive initiation (Caddle et al., 1990, Dimitrova et al.,1996). Several transcription factors have been shown to beactivators of replication initiation (Li and Botchan, 1993,He et al., 1993, DePamphilis, 1993a). In addition, a 28kDa factor found in HeLa cell nuclei binds a purin-richsequence (PUR consensus sequence, discussed later), whichis conserved in several origins from yeast, hamster andhuman to serve initiation of replication as a sequence-specific helix-destabilizing factor (Bergemann and Johnson,1992).

In this view, the AldB origin region has the structuralfeatures for potential origins (DePamphilis, 1993b). The200 bp region at the AldB promoter contains binding sites

for multiple transcription factors, the above mentioned PURconsensus sequence (discussed later), and an A/T-richsequence (Tsutsumi et al., 1989, Zhao et al., 1994).

V. Cell cycle-regulated factors bind to the200 bp region in the AldB origin

We next intended to know whether the 200 bp regionbinds factors from hepatoma cells (dRLh 84), in which thisregion was shown to be centered on an initiation region ofchromosomal DNA replication (Zhao et al., 1994)). Within this 200 bp proximal promoter, at least fourimportant cis-elements ( sites A, B, C and D) have beenshown to confer tissue- and developmentally specifictranscription. Site-A binds both a liver-specific factorHNF-1 and its competitive antagonist HNF-3 (Tsutsumi etal., 1989; Ito, Ki et al., 1990; Raymondjean et al., 1991;Gregori et al., 1993; Gregori et al., 1994), site-B (aCCAAT motif) binds factors AlF-B, NF-Y and a growth-inducible factor Ryb-a (Tsutsumi et al., 1993; Gregori etal., 1994; Ito, Ki et al., 1994). Site-C binds to C/EBP,DBP and a novel helix-loop-helix protein AlF-C (Yabuki etal., 1993; Gregori et al., 1993; Gregori et al., 1994; andYabuki et al., manuscript in preparation). Therefore, it isvery interesting to examine whether or not binding of thefactors to the 200 bp region is under control of cellproliferation and cell cycle progression. For this purpose,growth cycle of rat hepatoma dRLh84 cells wassynchronized by double thymidine block (synchrony wasmonitored by flow cytometry), and their nuclear extracts


604

were prepared every 2 hr after entering S phase. Thenuclear extracts were then subjected to gel electrophoreticmobility shift assay using oligonucleotides correspondingto sites A, B, C, and PPu ( p oly pu rin sequence containing

PUR consensus) as probes. Results showed that inquiescent cells binding activities to these sites wereconsiderably low as compared with those in growing cells.Site A- and site C-binding activities exhibited similarpatterns showing veritable cell cycle regulation. Namely,the activities reach a maximal levels at around G1/Sboundary, then gradually decrease to the lowest level at lateM to early G1 phase. Activity to bind site PPu increasestoward S phase. In contrast, site B-binding activity showedonly weak change throughout the cell cycle. Thus, inAldB non-expressing hepatoma cells, sites A, C and PPubind cell cycle-regulated factors whose activities increaseprior to S phase while site B binds a factor independently ofcell cycle phases. At present, precise characters of thesefactors are unknown. However, considering from tissue-specific or ubiquitous pattern of expression of the factorsthat bind to the promoter, site A might bind a factor HNF3in the hepatoma cells, since another factor HNF1 is notpresent in these cells (Kuo et al., 1991). Similarly, sites Band C are thought to bind Ryb-a and AlF-C, respectively,because these factors are rather enriched in rapidly growingcells such as fetal liver cells, and induced by growth signalssuch as partial hepatectomy or serum-stimulation ofcultured cells (Yabuki et al., 1993; Ito et al., 1994; andYabuki et al., manuscript in preparation).

Since site PPu has PUR consensus sequence, it mightbind a factor similar to that binds to PUR consensus, i.e.,Pur " (Bergemann and Johnson, 1992). The factor Pur" seems to act as a sequence-specific helix-destabilizing

factor, and thus implicating in its involvement in initiationof replication. Recently this factor was shown to associatewith the retinoblastoma protein Rb, and thus Rb mightmodulate binding of Pur " to its recognition site on DNA(Johnson et al., 1995).

These results implied a positive correlation betweenbinding of factors to the 200 bp region and the onset ofDNA replication. In dedifferentiated dRLh84 cells,transcription promoter of the AldB gene is completelyinactivated, and instead, used as a replication origin (seehypothetical model in Fig . 5 ). One interestingspeculation is that preferential binding of the abovementioned factors in AldB non-expressing cells, instead ofthose usually bind in the liver and activate the AldB gene,leads to repress transcription and consequently promotereplication initiation in a cell cycle-dependent manner.Similar observations were reported for Xenopuschromosome where embryo-specific origins are inactivatedwith concomitant activation of nearby transcription units(Hyrien et al., 1995), and for plasmids carrying a humanreplication origin that inhibition of replication depends onthe level of promoter activity (Haase et al., 1994).

VI. Chromosomal state of the AldBorigin/promoter region

We have discussed above on identification, sequencerequirement, and cell cycle-dependent protein-binding of thereplication origin region. In this section, we will focus onthe chromosomal state at the AldB origin/promoter regionin relation to transcription activity, liver cell proliferation,and development.

F i g . 5 . Summary of cell cycle- and growth-dependent binding of factors to the AldB gene origin/promoter region. Details aredescribed in the text.


605

F i g . 6 . Alteration of chromosomal state at the origin/promoter region during fetal liver development. Horizontal and curvedarrows indicate positions and directions of the AldB gene. Vertical arrows marked HS represent DNase I hyper sensitive site. CpGsites which were heavily methylated are shown as Me. Boxes represent promoter regions.

The AldB gene transcription in rat liver is repressed inthe fetal stage until around 16th day (day 16) of gestation.Thereafter, transcription of the gene is drastically activatedin the following two to three days (Numazaki et al., 1984). As expected, the activation during this fetal stageaccompanies the alteration of chromosomal state of theAldB gene (see hypothetical model in Fig . 6 ). Comparison of the chromatin structure among those in thelivers at the stages when the gene is repressed (day 14), justbeing activated (day 16), and fully activated (adult) revealedseveral distinct features for the repressed state (Daimon etal., 1986; Tsutsumi et al., 1987; Ito, Ki et al., 1995;Kikawada et al., unpublished data). Namely, thechromatin had two DNase I-hypersensitive sites II-a and II-bat day 14; the former site disappeared with concomitantactivation of the gene. The region around the transcriptionstart site was considerably resistant against DNase Idigestion as the fragment derived from the DNase I-hypersensitive sites remained almost intact even afterdigestion with higher concentration of DNase I; with sucha concentration of DNase I the chromatin at later stageswas very unstable and was cut into pieces. It was alsoshown that two CpG sites in HhaI and HpaII sequences near

transcription start site are hypermethylated at the repressedstage while those in adult liver are hypo-methylated.

DNase I hypersensitive sites as described above mightbe a reflection of nuclear matrix association, since thematrix contains topoisomerase II whose cleavage sites invivo are often found in DNase I hypersensitive regions(Poljak and Käs, 1995). Indeed, several consensussequences for topoisomerase II cleavage site (Sander andHsieh, 1985) were found around the DNase I-protected andhypersensitive regions, in addition to clusters of A- or T-rich stretches which are often found in matrix-associatedDNA (Gasser and Laemli, 1986) (Fig. 6). With thisrespect, further analyses using cultured cells encapsulatedinto agarose beads were carried out to define matrix-associated region. The results suggested that inproliferating, AldB non-expressing cells the DNase I-protected region in the AldB chromatin, i.e.,promoter/origin region, is within the matrix-associatedregion (MAR) (reviewed by Boulikas, 1995). It is notsurprising to consider that the origin region is attached tonuclear matrix where replication initiates in proliferatingcells, since MAR has been thought to be, for example, anorigin of replication (Amati and Gasser, 1988), a boundaryof DNA loops demarcating a regulatory domain including a


606

set of transcription units (Laemli et al., 1992), andtranscription enhancer elements (Gasser and Laemli, 1986).

Based on these observations, it would be considered thatthe alteration of chromatin state of the AldB genepromoter/origin region during fetal liver developmentreflects a change in the organization of functional domainsin chromosome. If so, positioning of replication originsand chromosomal domain, for example, might differbetween AldB expressing and non-expressing cell nuclei. In this regard, we recently found another origin region atmore than 40 Kb downstream of the promoter/origin regionin rat hepatoma cells. Both origins are fired in AldB non-expressing hepatoma cells, whereas the downstream originwas not used in AldB expressing differentiated hepatomacells, suggesting different organization of chromosomaldomains (Miyagi et al., unpublished observation).

VII. Perspectives

Here, we have described that the transcription promoterof the aldolase B gene is centered on an initiation region ofDNA replication in rat hepatoma cells in vivo. Within theorigin region, the 200 bp promoter fragment extendingfrom -200 bp to -1bp was indispensable to autonomousreplication when assayed by transfection of plasmids bearingvarious origin fragments. Since the 200 bp fragment alonedid not confer replication, the fragment is thought tocooperate with the flanking sequences to play an importantrole in initiation of replication. The 200 bp promoterconsists of multiple cis-elements for liver-specifictranscription. In rat hepatoma cells, in which the AldBgene is completely inactivated, protein factors bound to thecis-elements in a cell cycle- or growth-regulated manner,suggesting the involvement of the 200 bp region inregulation of replication initiation. Thus, the promoter ofthe AldB gene has dual roles in regulation of both liver-specific transcription and initiation of replication. Theresults, however, do not confine actual "start point" ofreplication. The 200 bp region, for example, could notnecessarily be a start point but rather be an auxiliaryelement. Further, whether the start point resides at a singlelocation or distributes throughout the origin/promoterregion is unknown. These points remain to be elucidated.

Firing of replication origin either at or neighboringthe AldB locus, or both, might influence the transcriptionalstate of the gene, since these biological reactions in asingle specialized DNA domain are together regulatedaccording to a DNA loop model, a chromosome model oftopologically independent DNA loop domains which areseparated by periodic association with nuclear matrix(Laemli et al., 1992). Concerning to this, the observationthat the origin/promoter region is attached onto nuclearmatrix in vitro and in vivo implies the importance of theregion in assembly of functional domains in a chromosome,since MAR has been postulated to act as a replication originand, in addition, as an insulator-like element in that itreduces position effect in transgenic mice (for example,McKnight et al,, 1992). Thus, we think that repressionand activation of the AldB gene might reflect positioning of

replication origins and alteration of domain structure inchromosomes. In fact, as mentioned earlier, usage of thetwo origins in the vicinity of the AldB locus differs betweenthe AldB expressing and non-expressing cells. It would bequite important to know how transcription and replicationreflect each other in the chromosomal context, since themechanism that govern such a causal relationship might beinvolved in cell differentiation and development. For thispurpose, the AldB gene promoter/replication origin wouldbe one of the suitable targets.

Acknowledgment

This work was supported in part by a Grant-in-Aid forScientific Research on Priority Areas and a Grant-in-Aid forScientific Research from the Ministry of Education,Science, Sports and Culture of Japan. We greatlyacknowledge Drs. K.Ishikawa, R.Tsutsumi, M.Yamaki,Y.Nagatsuka and K. Ito for their collaboration, help,discussions and encouragement.

References

Aladjem, M.I., Groudine, M., Brody, L.L., Dieken, E.S.,Fournier, R.E.K., Wahl, G.M., and Epner, E.M. (1 9 9 5 )

Participation of the human !-globin locus control region in

initiation of DNA replication. Sc ience 270, 815-819.

Amati,B.B. and Gasser, S.M. (1 9 8 8 ) Chromosomal ARS andCEN elements bind specifically to the yeast nuclearscaffold. Cel l 54, 967-978.

Ariga, H., Itani, T. and Iguchi-Ariga, S.M.M. (1 9 8 7 )Autonomous replicating sequences from mouse cells whichcan replicate in mouse cells in vitro and in vivo. M o l .C e l l . B i o l . 7, 1-6.

Ariizumi, K., Wang, Z., and Tucker, P.W. (1 9 9 3 )Immunoglobulin heavy chain enhancer is located near or inan initiation zone of chromosomal DNA replication. Proc.Nat l . Acad. Sc i . U .S .A. 90, 3695-3699.

Bell, S.P., Mitchell, J., Leber, J., Kobayashi, R., and Stillman,B. (1 9 9 5 ) The multidomain structure of Orc1p revealssimilarity to regulators of DNA replication andtranscriptional silencing. Cel l 83, 563-568.

Bell, S.P., and Stillman, B. (1 9 9 2 ) ATP-dependentrecognition of eukaryotic origins of DNA replication by amultiprotein complex. Nature 357, 128-134.

Bergemann, A.D., and Johnson, E.M. (1 9 9 2 ) The HeLa purfactor binds single-stranded DNA at a specific elementconserved in gene flanking regions and origins of DNAreplication. M o l . C e l l . B i o l . 12, 1257-1265.

Boulikas, T. (1 9 9 5 ) Chromatin domains and prediction ofMAR sequences. In t . Rev . Cyto l . 162, 279-388.

Brewer, B.J., Lockshon, D., and Fangman, W.L. (1 9 9 2 ) Thearrest of replication forks in the rDNA of yeast occursindependently of transcription. Cel l 71, 267-276.

Caddle, M.S., Dailey, L., and Heintz, N.H. (1 9 9 0 ) RIP60, amammalian origin-binding protein, enhances DNA bending


607

near the dihydrolate reductase origin of replication. M o l .C e l l . B i o l . 10, 6236-6243.

Coverley, D. and Laskey, R.A. (1 9 9 4 ) Regulation ofeukaryotic DNA replication. Annu. Rev. Biochem . 63,745-776.

Daimon, M., Tsutsumi, K. and Ishikawa, K. (1 9 8 6 ) DNAmethylation and the regulation of aldolase B geneexpression. J . B iochem . 100, 1279-1286.

DePamphilis, M.L. (1 9 9 3 a ) Eukaryotic DNA replication:anatomy of an origin. Annu. Rev . B iochem . 6 2 , 29-63.

DePamphilis, M.L. (1 9 9 3 b ) How transcription factorsregulate origins of DNA replication in eukaryotic cells.Trends Cel l Biol . 3 , 161-167.

Deshpande, A.M., and Newlon, C.S. (1 9 9 6 ) DNA replicationfork pause sites dependent on transcription. Sc ience 272,1030-1033.

Dimitrova, D.S., Giacca, M., Demarchi, F., Biamonti, G.,Riva, S., and Falaschi, A. (1 9 9 6 ) In vivo protein-DNAinteractions at a human DNA replication origin. Proc.Nat l . Acad. Sc i . U .S .A. 9 3 , 1498-1503.

Fangman, W. L., and Brewer, B.J. (1 9 9 2 ) A question of time:replication origins of eukaryotic chromosomes. Cel l 71,363-366.

Foss, M., McNally, F.J., Laurenson, P., and Rine, J. (1 9 9 3 )Origin recognition complex (ORC) in transcriptionalsilencing and DNA replication in S. cerevisiae. Sc ience 262, 1838-1844.

Fox, C.A., Ehrenhofer-Murray, A.E., Loo, S. and Rine, J.(1 9 9 7 ) The origin recognition complex, SIR1, and the Sphase requirement for silencing. Sc ience , 276, 1547-1551.

Gasser, S.M. and Laemli, U.K. (1 9 8 6 ) Cohabitation ofScaffold binding regions with upstream/enhancer elementsof three developmentally regulated genes of D.melanogaster. Cel l 46, 521-530.

Gregori, C., Kahn, A., and Pichard, A.-L. (1 9 9 3 ) Competitionbetween transcription factors HNF1 and HNF3, andalternative cell-specific activation by DBP and C/EBPcontribute to the regulation of the liver-specific aldolase Bpromoter. Nucl . Acids Res . 2 1 , 897-903.

Gregori, C., Kahn. A., and Pichard, A.-L. (1 9 9 4 ) Activity ofthe rat liver-specific aldolase B promoter is restrained byHNF3. Nucl . Acids Res . 22, 1242-1246.

Guo, Z.-S., and DePamphilis, M.L. (1 9 9 2 ) Specifictranscription factors stimulate simian virus 40 andpolyomavirus origins of DNA replication. M o l . C e l l .B i o l . 12, 2514-2524.

Hamlin, J. L., Mosca, P.J. and Levenson, V.V. (1 9 9 5 ) Defining origins of replication in mammalian cells. B i o c h i m . B i o p h y s . A c t a 1198, 85-111.

Haase, S.B., Heinzel, S.S., and Calos, M.P. (1 9 9 4 )Transcription inhibits the replication of autonomouslyreplicating plasmids in human cells. M o l . C e l l . B i o l .14, 2516-2524.

He, Z., Brinton, B.T., Greenblatt, J., Hassell, J.A., and Ingles,C.J. (1 9 9 3 ) The transactivator proteins VP16 and GAL4bind replication factor A. Cel l 7 3 , 1223-1232.

Heinzel, S.S., Krysan, P.J., Tran, C.T. and Calos, M.P.(1 9 9 1 ) Autonomous DNA replication in human cells isaffected by the size and the source of the DNA. M o l . C e l l .B i o l . 11, 2263-2272.

Hirt, B. (1 9 6 7 ) Selective extraction of polyoma DNA frominfected mouse cell cultures. J . M o l . B i o l . 26, 365-369.

Horecker, B.L., Tsolas, O., and Lai, C.Y. (1 9 7 2 ) Aldolase, inThe Enzymes (P.D.Boyer ed.), Vol.7, pp213-258,Academic Press, Inc., New York .

Hyrien, O., Maric, C., and Méchali, M. (1 9 9 5 ) Transcriptionin specification of embryonic metazoan DNA replicationorigins. Sc ience 270, 994-997.

Ito, Ki., Tanaka, T., Tsutsumi, K., Ishikawa, K. and Tsutsumi,K. (1 9 9 0) Two different HNF1-like transcriptionactivators in the liver bind to the same region of the rataldolase B promoter. Biochem. B i o p h y s . R e s .Commun. 173, 1337-1343.

Ito, Ki., Tsutsumi, K., Kuzumaki, T., Gomez, P.F., Otsu, K. andIshikawa, K. (1 9 9 4 ) A novel growth-inducible gene thatencodes a protein with a conserved cold-shock domain.Nucl . Acids Res . 22, 2036-2041.

Ito, Ki., Tsutsumi, R., Ishikawa, K. and Tsutsumi, K. (1 9 9 5 )Developmental alteration of the chromatin state atpromoter/replication origin region of the aldolase B locusprecedes transcriptional activation in the liver. D e v .Genet . 17, 312-318.

Ito, Ko., Asano, M., Hughes, P., Kohzaki, H., Masutani, C.,Hanaoka, F., Kerppola, T., Curran, T., Murakami, Y., andIto, Y. (1 9 9 6 ) c-jun stimulates origin-dependent DNAunwinding by polyomavirus large T antigen. EMBO J.15, 5636-5646.

James, C.D. and Leffak, M. (1 9 8 6 ) Polarity of DNAreplication through the avian alpha-globin locus. M o l .C e l l . B i o l . 6, 976-984.

Johnson, E.M., Chen, P.-L., Krachmarov, C.P., Barr, S.M.,Kanovsky., M., Ma, Z.-W. and Lee, W.-H. (1 9 9 5 )

Association of human Pur " with the retinoblastoma

protein, Rb, regulates binding to the single stranded DNA

Pur " recognition element. J. B i o l . Chem . 270,

24352-24360.

Kitsberg, D., Selig, S., Keshet, I., and Cedar, H. (1 9 9 3 )

Replication structure of the human !-globin gene domain.

Nature 366, 588-590.

Krysan, P.J., and Calos, M.P. (1 9 9 1 ) Replication initiates atmultiple locations on an autonomously replicating plasmidin human cells. M o l . C e l l B i o l . 11, 1464-1472.

Krysan, P.J., Smith, J.G., and Calos, M.P. (1 9 9 3 )Autonomous replication in human cells of multimers ofspecific human and bacterial DNA sequences. M o l . C e l l .B i o l . 13, 2688-2696.

Kuo, C.J., Mendel, D.B., Hansen, L.P., and Crabtree, G.R.

(1 9 9 1 ) Independent regulation of HNF-1" and HNF-1! by

retinoic acid in F9 teratocarcinoma cells. EMBO J . 10 ,2231-2236.

Laemli, U.K., Käs, E., Poljak, L. and Adachi, Y. (1 9 9 2 )Scaffold-associated regions: cis-acting determinants ofchromatin strucural loops and functional domains.Curr.Opinion Gen.Dev . 2, 275-285.


608

Leffak, M., and James, C.D. (1 9 8 9 ) Opposite replicationpolarity of the germ line c-myc gene in HeLa cellscompared with that of two Burkitt lymphoma cell lines.M o l . C e l l . B i o l . 9, 586-593.

Li, R., and Botchan, M.R. (1 9 9 3 ) The acidic transcriptionalactivation domains of VP16 and p53 bind the cellularreplication protein A and stimulate in vitro BPV-1 DNAreplication. Cel l 73, 1207-1221.

Liu, B., and Alberts, B.M. (1 9 9 5 ) Head-on collision between aDNA replication apparatus and RNA polymerasetranscription complex. Sc ience 267, 1131-1137.

Liu, B., Wong, M.L., Tinker, R.L., Geiduschek, E.P., andAlberts, B.M. (1 9 9 3 ) The DNA replication fork can passRNA polymerase without displacing the nascent transcript.Nature 366, 33-39.

McKnight, R.A., Shamay, A., Sankaran, L., Wall, R.J., andHenninghausen, L. (1 9 9 2 ) Matrix-attachment regions canimpart position-independent regulation of a tissue-specificgene in transgenic mice. Proc . Nat l . Acad . Sc i . USA89, 6943-6947.

McWhinney, C. and Leffak, M. (1 9 9 0 ) Autonomousreplication of a DNA fragment containing thechromosomal replication origin of the human c-myc gene.Nucl . Acids Res . 18, 1233-1242.

Micklem, G., Rowley, A., Harwood, J., Nasmyth, K., andDiffley, J.F.X. (1 9 9 3 ) Yeast origin recognition complexis involved in DNA replication and transcriptionalsilencing. Nature 366, 87-89.

Mul, Y.M., and Van der Vliet, P.C. (1 9 9 2 ) Nuclear factor Ienhances adenovirus DNA replication by increasing thestability of a preinitiation complex. EMBO J. 11, 751-760.

Numazaki, M., Tsutsumi, K., Tsutsumi, R. and Ishikawa, K.(1 9 8 4 ) Expression of aldolase isozyme mRNA species infetal liver. Eur. J. Biochem . 142, 165-170.

O'Neil, E.A., Fletcher, C., Burrow, C.R., Heinz, N., Roeder,R.G. and Kelley, T.J. (1 9 8 8 ) Transcription factor OTF-1is functionally identical to the DNA replication factor NF-III. Sc ience 241, 1210-1213.

Poljak, L. and Käs, E. (1 9 9 5 ) Resolving the role oftopoisomerase II in chromatin structure and function.Trends Cel l Biol . 5, 348-354.

Raymondjean, M., Pichard, A. -L., Gregori, C., Ginot, F., andKhan, A. (1 9 9 1 ) Interplay of an original combination offactors: C/EBP, NFY, HNF3 and HNF1 in the rat aldolase Bgene promoter. Nucl . Acids Res . 19, 6145-6153.

Sander, M. and Hsieh, T. (1 9 8 5 ) Drosophila topoisomeraseII double-strand cleavage: analysis of DNA sequencehomology at the cleavage site. Nucl . Ac ids Res . 13,1057-1071.

Satoh, J., Tustusmi, K., Ishikawa, M. and Ishikawa, K.(1 9 8 6 ) Dietary regulation of aldolase isozymeexpression in rat intestinal mucosa. Arch . B iochem .B i o p h y s . 254, 116-123.

Stillman, B. (1 9 9 6 ) Cell cycle control of DNA replication.Sc ience 274, 1659-1664.

Taira, T., Iguchi-Ariga, S.M.M., and Ariga, H. (1 9 9 4 ) A novelDNA replication origin identified in the human heat shock

protein 70 gene promoter. M o l . C e l l . B i o l . 14, 6386-6397.

Tasheva, E.S., and Roufa, D.J. (1 9 9 4 ) A mammalian origin ofbidirectional DNA replication within the chinese hamsterRPS14 locus. M o l . C e l l . B i o l . 14, 5628-5635.

Trempe, J.P., Lindstrom, Y.I., and Leffak, M. (1 9 8 8 ) Oppositereplication polarities of transcribed and nontranscribedhistone H5 genes. M o l . C e l l . B i o l . 8, 1657-1663.

Tsutsumi, K., Ito, K., and Ishikawa, K. (1 9 8 9 ) Developmentalappearance of transcription factors that regulate liver-specific expression of the aldolase B gene. M o l . C e l l .B i o l . 9, 4923-4931.

Tsutsumi, K., Ito, K., Yabuki, T., and Ishikawa, K. (1 9 9 3 )AlF-B, a novel CCAAT-binding transcription activator thatinteracts with the aldolase B promoter. FEBS Lett . 321,51-54.

Tsutsumi, K., Mukai, T., Tsutsumi, R., Hidaka, S., Arai, Y.,Hori, K. and Ishikawa, K. (1 9 8 5 ) Structure and genomicorganization of the rat aldolase B gene. J . M o l . B i o l .181, 153-160.

Tsutsumi, K., Tsutsumi, R. and Ishikawa, K. (1 9 8 7 ) Tissue-specific changes in chromatin structure of the rat aldolase Blocus. J . B iochem . 102, 1013-1021.

Vassilev, L., and Johnson,E. M. (1 9 9 0 ) An initiation zone ofchromosomal DNA replication located upstream of the c-myc gene in proliferating HeLa cells. M o l . C e l l . B i o l .10, 4899-4904.

Wolffe, A.P., and Brown, D.D. (1 9 8 8 ) Developmentalregulation of two 5S ribosomal RNA genes. Sc ience 241,1626-1632.

Wu, C., Friedlander, P., Lamoureux, C., Zannis-Hadjopoulos,M. and Price, G.B. (1 9 9 3 ) cDNA clones containautonomous replication activity. B i o c h i m . B i o p h y s .Acta 1174, 241-257.

Yabuki, T. Ejiri, S., and Tsutsumi, K. (1 9 9 3 ) Ubiquitousfactors that interact simultaneously with two distinct cis-elements on the rat aldolase B gene promoter. B i o c h i m .Biophys . Ac ta 1216, 15-19.

Zhao, Y., Miyagi, S., Kikawada, T., and Tsutsumi, K. (1 9 9 7 )Sequence requirement for replication initiation at the rataldolase B locus implicated in its functional correlationwith transcriptional regulation. Biochem. B i o p h y s .Res . Commun. 237, 707-713.

Zhao, Y., Tsutsumi, R., Yamaki, M., Nagatsuka, Y., Ejiri, S.,and Tsutsumi, K. (1 9 9 4 ) Initiation zone of DNAreplication at the aldolase B locus encompassestranscription promoter region. N u c l . A c i d s R e s . 22,5385-5390.


609


TARgeting the human genome to make gene isolation easy

Michael A. Resnick, Natalay Kouprina, and Vladimir Larionov

Laboratory of Molecular Genetics, National Institute of Environmental Health Sciences, NIH, P.O. Box 12233,110 Alexander Dr., Research Triangle Park, NC 27709__________________________________________________________________________________________________

Correspondence to: Michael A. Resnick, Tel: 919 541-4480; Fax: 919 541-7593, E-mail: [email protected]

Summary

Considerable information is now available about the human genome and expressed sequences havebeen identi f ied for most genes . Unti l recent ly there was no opportunity to specifically isolategenes or speci f ic chromosomal regions from genomic DNA. We have utilized transformation-associated recombination (TAR) in yeast to isolate genes and specif ic regions from total humanDNA. This has been demonstrated by the direct isolat ion of complete copies of rDNA, BRCA1,BRCA2 and HPRT genes with high fidelity as yeast artificial chromosomes (YACs). We proposethat there are many utilities of TAR cloning including gene therapy and diagnostics.

The Human Genome Project has made great strides inthe decade since its inception including the cloning ofmost of the chromosomal DNA, the identification ofunique sequences (sequence tags sites, STS's)approximately every 150 kb and the sequencing of shortregions of almost all the expressed genes (expressedsequence tags, EST's). The project is ahead of schedule inthat most of the genome will be sequenced within thenext 5 to 10 years. In addition to understandingchromosome organization, this vast amount ofinformation is leading to the isolation of genes thatcorrespond to specific diseases, particularly throughpositional cloning. Furthermore, the genetic makeup ofhumans is better understood because of sequencerelatedness between species.

Until now there has been little opportunity to utilizethe information being generated to isolate specific largeregions (i.e., greater than 10 to 20 kb) or genes directlyfrom total genomic material. Virtually all cloning ofchromosomal DNA from humans, or any organism, hasinvolved the isolation of random DNA fragments intovectors through several steps of enzymatic treatment plusligation and the subsequent transfer into the desiredbacterial or yeast host. The isolation of specific DNAswould provide a variety of opportunities, including studiesof human polymorphisms, clinical diagnosis, gene therapyand the filling-in of gaps in sequenced regions. However,the only available enrichment procedure has been thephysical isolation of entire chromosomes (McCormick etal., 1993). Even then, the subsequent cloning of humanDNAs has involved random DNA fragments.

Over the past year a new approach has emerged that isproviding for the specific isolation of genes and regionsdirectly from total human DNA. The approach draws uponseveral features of the yeast Saccharomyces cerevisiae. The first is that during transformation, yeast can take upseveral small and large molecules (Rudolph et al., 1985;Larionov et al., 1994). Secondly, intermolecular, as wellas intramolecular, recombination is highly efficient duringtransformation between homologous, as well divergedDNAs (Larionov et al., 1994; Ma et al., 1987; Mezard etal., 1992). This includes double-strand breakrecombination between broken molecules. Thirdly, humanDNA contains sequences (about 1 per 20-30 kb) that canfunction as origins of replication (ARS-autonomouslyreplicating sequence) in yeast (Stinchcomb et al., 1980).These features have provided for the development of anovel method based on t ransformation- a ssociated

r ecombination (TAR) to target the isolation of specific

DNAs from total human DNAs.

As described in Figure 1 , genomic DNA is presentedto yeast along with a molar excess of vector containing aselectable marker, a centromere (CEN) to assure productionof a single copy of the cloned material and targetingsequence hooks A and B (the original circular plasmid islinearized at a site between A and B). [The TAR proceduresimply involves the presentation of gently prepared humanDNA, originally isolated in low-melt agarose plugs, tocompetent yeast spheroplasts along with vector DNA.]


610

Figure 1 . Model of TAR cloning togenerate circular YACs. Human DNA is takenup by a yeast cell along with linearizedvector DNA. The vector contains acentromere and a marker for selection. If thehuman DNA contains segmentscorresponding to the segments--hooks--Aand B on the plasmid, recombination willlead to the establishment of a circular YAC.Propagation of the YAC depends on thepresence of a yeast ARS-like sequence in thehuman DNA. The various blocks could bediverged repeats, such as Alu’s or LINES.

The minimum size of the hooks required for TAR cloningappears to be less than 150 bp (Larionov et al., 1996). Ifa human fragment containing a sequence A' and B' windsup in the same cell as the vector, recombination betweenthe cut plasmid and the fragment will generate a circularyeast artificial chromosome (YAC) which can be selectedusing the plasmid marker. Because the plasmid has noyeast replication origin, sequences in human DNAscapable of functioning as ARS 's in yeast provide for thepropagation of the YAC. Thus, the isolation of humanDNA is essentially accomplished by marker rescue throughrecombination. The generation of YACs with largehuman segments was proposed to be due to preferentialdouble-strand break repair at or near ends of moleculesrather than internal regions (Larionov et al., 1996a). [Theoriginal model for double-strand break repair (Resnick,1976) has now had many applications and refinements thatextend from the repair of radiation-induced breaks, naturalbreaks and gap repair of incoming molecules (Orr-Weaveret al., 1983) to gene replacement in mammalian cells(Cappechi, 1988) and the development of knockout miceand now TAR cloning.]

The opportunity to TAR clone human DNA wassuggested from experiments (Larionov et al., 1994) inwhich it was shown that during transformation there wasefficient recombination between an incoming plasmid withan Alu and an incoming human yeast artificialchromosome (YAC) that contained several Alu's . Withthis in mind, transformation-associated recombination wasexplored as an alternative means of generating linear YAC

libraries containing large fragments of chromosomal DNA(Larionov et al., 1996). Subsequently, the originalscheme--which does not involve restricting or ligatingDNA-- was modified to yield circular YACs (Larionov etal., 1996b) as described in Figure 1 .

The efficiency and selectivity of TAR cloning wasinitially demonstrated by the specific isolation of humanDNA from a radiation hybrid rodent cell line containing a5 Mb human chromosome fragment that had the Ku80gene (Larionov et al., 1996b). A circularizing TAR vectorwas used that had the same human Alu for the targeting Aand B hooks (see Figure 1 and Table 1 ).Approximately 25% of the transformants for the vectormarker had YACs containing human DNA and most weregreater than 150 kb. Based on the relative number ofYACs isolated containing rodent DNA, this correspondedto a nearly 5000-fold enrichment (Larionov et al., 1996b)over the 0.1% human DNA present in the hybrid cells.

These results led to the demonstration that TARcloning could be used to isolate a specific human gene(Larionov et al., 1997), the breast cancer gene BRCA2. Although it had been sequenced, no complete BRCA2 genehad been isolated either as a YAC or a BAC (a bacterialartificial chromosome in E. coli). To do this, the TARvector with hooks of approximately 500 bp each of thepromoter sequence and the noncoding region of the lastexon (see Figure 1) was presented to yeast cells alongwith total DNA isolated from human fibroblasts. About 1in 300 transformants (Larionov et al., 1997 andunpublished) selected for the vector marker also contained


611

Table 1. Specific isolation of human DNA by TAR cloning

DNA cloned and source Hook A Hook B

100 Mb Chromosome 16 in a monochromosomal hybrid[6,7]

consensus ALU BLUR13 ALU

5 Mb Ku80 in a radiation hybrid [7] consensus ALU BLUR13 ALU

43 kb rDNA unit in total human DNA [10] non transcribed spacer BLUR13 ALU

90 kb BRCA2 in total human DNA [4]* 5' upstream sequence 3' downstreamsequence

82 kb BRCA1 in total human DNA * (unpublished) 5' upstream sequence 3' downstreamsequence

70-350 kb HPRT in total human DNA * (unpublished) BLUR13 ALU 3' downstreamsequence

* Up to 1% of the yeast transformants had the gene of interest. The genes were identified through pooling of transformants, PCRanalysis, followed by isolation of clones.

Figure 2 . A TAR cloning cycle for the specific isolation of human DNA and its reintroduction into mammalian cells. HumanDNA can be specifically isolated in yeast by TAR cloning, modified for transfer to bacteria and then transferred to mammaliancells. Alternatively, the TAR vector can contain sequences that would enable selection in mammalian cells enabling direct transferfrom yeast.

the BRCA2 gene and these could be easily identified byPCR analysis (see Table 1). Further physical analysisestablished that several independent copies of the completegene had indeed been isolated.

The utility of TAR cloning for the specific isolation ofhuman genes has now been demonstrated further for theBRCA1 and the HPRT genes (unpublished) and the humanribosomal RNA gene family (Kouprina et al., 1997). Asshown in Table 1, specific isolation can be accomplished

with one hook that is unique to the gene(s) being isolatedand the other hook being a common repeat. The numbersof unique genes isolated per µg of total DNA presented toyeast were comparable for the BRCA1 and the HPRTgenes. Thus, direct gene isolation is now possible usinginformation derived from only a small portion of a gene.

Because only a few weeks are required once the vectorsare built, TAR cloning provides new opportunities forinvestigating genes and chromosomal regions directly from

Resnick et al: TARgeting the human genome to make gene isolation easy

612

individuals. Previously, isolation of specific chromosomalregions would have required the development of a libraryfor each person studied followed by extensive analysis tofind the region of interest. These features suggest thatTAR cloning can open the way to clinical investigationsof whole genes or large chromosomal regions since, forexample, only 10 to 20 ml of blood would be needed forthe isolation of a specific gene.

Another novel utility--referred to as radial TARcloning--derives from the isolation of the rDNA and theHPRT genes with a vector that has a unique sequence hookand an Alu repeat hook (A and B, respectively, in Figure1 and Table 1 ). YACs are generated that extend fromthe unique position to various Alu's. By changing theorientation of the unique hook, a radial series of YACs isdeveloped that surround the unique sequence. There aremany applications that include isolating a unique regionsurrounding a particular STS or EST site. In additionchromosomal changes such as amplifications andtranslocations in individuals become directly accessiblewith TAR cloning once a chromosomal sequence isidentified. Radial TAR cloning also provides theopportunity to clone a region lacking an ARS-likesequence since the hook with the common repeat enablesthe isolation of chromosome fragments that aresufficiently large that they are likely to contain such asequence.

The TAR cloning can be used in a cycle that providesfor specific human DNA isolation and reintroduction, asdescribed in Figure 2 . Once DNA is isolated as acircular molecule it can be modified and even retrofittedwith bacterial artificial chromosome sequences andmammalian selectable markers such as neomycin (NEO) orhygromycin resistance (or alternatively the original TARvector could contain these sequences) using recombinationmethods standard to yeast (Larionov et al., 1996b, 1997).The YAC/BAC can than be transferred into E. coli in orderto obtain large amounts of this DNA and it couldsubsequently be introduced into human cells. (Largecircular molecules may be isolated directly from yeast, sothat the step involving transfer to E. coli could beeliminated.) This approach is being applied to theBRCA2, BRCA1 and HPRT genes initially isolated asYACs. The subsequent YAC/BACs are reintroduced intomammalian cells using the NEO marker for selection.Since, as recently shown for HPRT, most of the isolatedgenes are functional when transferred to mammalian cells(in preparation), the cycle of human DNA isolation andreintroduction can be accomplished with high fidelity.

The tremendous success of the human genome projecthas relied on the development of new approaches. Theinformation generated can be applied to many areasincluding functional genomics, investigations of geneticdiseases, gene manipulation and gene therapy. TARcloning is one of the new tools that will make ourchromosomes more accessible.

Acknowledgment

This work was supported in part through a CRADA(Cooperative Research and Development Agreement)between NIEHS and Life Technologies Inc. ofGaithersburg, Maryland.

References

Capecchi, M.R. (1 9 8 9 ). Altering the genome byhomologous recombination, Science 2 3 3 , 1288-1292.

Kouprina, N., Graves, J., Resnick, M.A., Larinov, V. (1 9 9 7 )Specific isolation of rDNA genes by TAR cloning, Gene197, 269-276.

Larionov, V., Kouprina, N., Eldarov, M., Perkins, E., Porter,G., and Resnick, M. (1 9 9 4 ) Transformation-associatedrecombination between diverged and homologous DNArepeats is induced by strand breaks, Yeast 10, 93-104.

Larionov, V., Kouprina, N., Graves, J., Chen, X-N.,Korenberg, J. R. and Resnick, M.A. (1 9 9 6 a ). Specificcloning of human DNA as YACs by transformation-associated recombination. Proc. Nat l . Acad. Sc i . 93,491-496.

Larionov, V., Kouprina, N., Graves, J. and Resnick, M.A.(1 9 9 6 b ) Highly selective isolation of human DNAs fromrodent-human hybrid cells as circular YACs by TARcloning. Proc. Natl Acad. Sc i 93, 13925-13930(1996b).

Larionov, V., Kouprina, N., Nikolaishvili, N. and Resnick,M. A. , (1 9 9 4 ) Recombination during transformation as asource of chimeric mammalian artificial chromosomes inyeast (YACs), Nucle ic Ac ids Research 22, 4154-4162.

Larionov, V., Kouprina, N. Solomon, G., Barrett, J. C. andResnick, M.A. (1 9 9 7 ). Direct isolation of humanBRCA2 gene by transformation-associated recombinationin yeast. Proc. Natl . Acad. Sci . 94, 7384-7387.

Ma, H., Kunes, S., Schatz, P.J., and Botstein, D. (1 9 8 7 ).Plasmid construction by homologous recombination inyeast. Gene 58, 201-216.

McCormick M. K.., Campbell, E., Deaven, L., Moyzis, R.( 1 9 9 3 ) Low-frequency chimeric yeast artificialchromosome libraries from flow-sorted humanchromosomes 16 and 21, P r o c . N a t l A c a d . S c i .90, 1063-1067.

Mezard, C., Pompon, D., and Nicolas, A. (1 9 9 2 ).Recombination between similar but not identical DNAsequences during yeast transformation occurs within shortstretches of identity, Cel l 70, 659-670.

Resnick, M.A. (1 9 7 6 ). The repair of double-strand breaks inDNA: a model involving recombination. J .Theore t i ca l B io logy 59, 97-106.

Rudolph, H., Koenig-Rauseo, I., and Hinnen, A. (1 9 8 5 ).One-step gene replacement in yeast by cotransformation,Gene 36, 87-95.

Stinchcomb, D. T., Thomas, M., Kelly, J., Selker, E., andDavis, R. W. (1 9 8 0 ). Eukaryotic DNA segments capableof autonomous replication in yeast. Proc . Nat l Acad .Sc i 77, 4559-4563 (1980).


613

Szostak, J.W., Orr-Weaver, T.L., Rothstein, R.J., Stahl, F.W.(1 9 8 3 ), The double-strand break model for

recombination. Cel l 33, 25-35.


613


Control of growth and proliferation by theretinoblastoma protein

Robert J. White

Institute of Biomedical and Life Sciences, Division of Biochemistry and Molecular Biology, Davidson Building, Universityof Glasgow, Glasgow, G12 8QQ, U.K.

_______________________________________________________________________________________________Correspondence: Robert J. White, Tel: 0141-330-4628, Fax: 0141-330-4620, E-mail: [email protected]

Keywords: Cancer. Retinoblastoma. Tumour suppression. Transcription. Translation

Summary

The retinoblastoma susceptibility gene Rb is an important tumour suppressor. It wil l inhibitboth growth and proliferation when introduced into many types of ce l l . Furthermore, i t i sfrequently found mutated in a range of human cancers. It is therefore of considerable importancethat we should understand fully how this gene operates. The RB gene product is a 110 kDa nuclearphosphoprotein that regulates the activity of a number of key transcription factors. In turn, itsactivity is control led through phosphorylation by cyclin-dependent kinases in response to theavailability of growth factors. It therefore provides a mechanism for coordinating gene expressionwith growth factor availability. One of the principle targets of RB is a transcription factor calledE2F. E2F contro ls the express ion of a panel o f genes that promote prol i ferat ion . By down-regulating these genes through its inhibitory action on E2F, RB provides a restraining influenceupon ce l l cyc le progress ion . I t has been less clear how RB i s able to suppress the growth( increase in mass) of ce l ls . However, recent s tudies have suggested that i t may achieve this byrepressing the production of rRNA and tRNA. Loss of control over the protein synthetic apparatusmay constitute an important step in tumour development.

I. Introduction

The retinoblastoma susceptibility gene Rb is essentialfor life. Its homozygous inactivation causes mouseembryos to die during the fourteenth day of gestation withdefective neural and erythroid development (Clarke et al.,1992; Jacks et al., 1992; Lee et al., 1992). The Rb geneencodes a 110 kDa nuclear phosphoprotein that isexpressed almost ubiquitously in normal mammalian cells(Weinberg, 1995; Whyte, 1995). It was mapped tochromosome 13q14 by virtue of its association with aninherited predisposition to retinoblastoma, a rare pediatrictumour of the retina (Friend et al., 1986). Inactivatingmutations in this gene also occur in many other types ofhuman malignancy, including small-cell lung cancers,several sarcomas and bladder carcinomas (Weinberg, 1995;Whyte, 1995). These observations suggested that Rb is atumour suppressor and that loss of its function cancontribute to oncogenesis. Support for this idea camefrom experiments in which the wild-type gene wasintroduced into tumour cells that lacked its function(Bookstein et al., 1990; Huang et al., 1988; Qin et al.,

1992). Expression of exogenous Rb was found to inhibitgrowth, proliferation, soft agar colony formation andtumourigenicity in nude mice (Bookstein et al., 1990;Huang et al., 1988; Qin et al., 1992). Further proof of theimportance of Rb in resisting carcinogenesis was providedby the specific mutagenesis of this gene. Althoughhomozygous deletion of Rb is lethal, heterozygous micesurvive and display a strong predisposition to cancer (Hu etal., 1994; Jacks et al., 1992; Lee et al., 1992; Maandag etal., 1994; Nikitin and Lee, 1996; Williams et al., 1994).These observations prove unequivocally that Rb is a bonafide tumour suppressor gene.

Having established the credentials of RB as animportant tumour suppressor, it became a major priorityto determine how it achieves this effect. At a cellular level,RB is involved in constraining both growth (increase incell mass) and proliferation (increase in cell number):without it the ability of cells to shut down these functionsis compromised (Weinberg, 1995; Whyte, 1995).Mammalian cells decide between proliferation andquiescence during the first two thirds of G1 phase: ifgrowth factors are plentiful at this time they continue

White: Control of Growth and Proliferation by RB

614

through the cell cycle, but if conditions are unfavourablethey withdraw from cycle and quiesce (Pardee, 1989).Before reaching the end of G1, cells become committed tocomplete the mitotic cycle regardless of growth conditions(Pardee, 1989). This transition to serum-independence iscalled the R (restriction) point (Pardee, 1989). RB servesan important function in restraining passage through the Rpoint when growth factors are limiting (Sherr, 1994;Weinberg, 1995; Whyte, 1995). When RB function islost, the sensitivity of cells to their normal regulatorysignals is severely compromised (Sherr, 1994; Weinberg,1995; Whyte, 1995). This constitutes a major steptowards uncontrolled proliferation.

Although it is clear that RB regulates passage throughthe R point, many unanswered questions remain as to howthis is achieved in mechanistic terms. To understand fullythe complex biological effects of RB, it will be necessaryto determine how it operates at the molecular level.Although some aspects of this have been characterisedextensively, novel targets for RB are still being discovered(Taya, 1997). The relative contribution of each of thesetargets in inhibiting growth, proliferation and tumourformation will need to be established. A complete pictureof how RB functions will require the careful interlinkingof its various activities.

II. RB and cancer

A. Mutational inactivation of RB

People who inherit a nonfunctional allele of the Rbgene have an approximately 90% chance of developingretinoblastoma at an early age (Whyte, 1995). Inactivationof the remaining allele by somatic mutation seems to be auniversal feature of this cancer and is probably the rate-limiting step in its initiation (Horowitz et al., 1990).Individuals who survive hereditary retinoblastoma show astrong predisposition to osteosarcomas and soft tissuesarcomas later in life: this again is associated with loss ofthe second Rb allele (Whyte, 1995). These osteosarcomasand mesenchymal tumours are less frequent thanretinoblastoma in Rb heterozygotes, and loss of thefunctional copy of Rb may not be rate-limiting for such

tumours (Whyte, 1995). Unlike humans, Rb+/- mice donot develop retinoblastoma: instead over 95% die 300-400days after birth with melanotroph tumours of theintermediate pituitary lobe (Hu et al., 1994; Maandag etal., 1994; Williams et al., 1994). Sequential analyses ofthe initial stages of spontaneous melanotrophcarcinogenesis in heterozygous mice suggest that mutationof the Rb gene is the initiating event of malignanttransformation (Nikitin and Lee, 1996). It is notunderstood why murine and human Rb heterozygotes sufferdifferent types of cancer. Neither is it known whymelanotrophs or retinoblasts are particularly sensitive tothe inactivation of RB.

Many other types of human tumour display somaticmutation of Rb , including osteosarcomas, small cell lungcarcinomas, breast cancers, prostate and bladder

carcinomas. In such cases, the patient inherits two wild-type alleles of Rb , but mutations arise in both copiesduring tumourigenesis. The most striking examples of thisare the small cell lung carcinomas, where Rb changes arefound in nearly all cases (Horowitz et al., 1990). Othertypes of tumour display a lower frequency of Rbmutation. For example, RB was found to be altered orabsent in a third of bladder carcinomas that were surveyed(Horowitz et al., 1990). However, many types of tumourexpress apparently wild-type RB, including melanomas andcolon carcinomas (Horowitz et al., 1990). Thus, mutationof Rb is a tumour-specific phenomenon.

B. Inactivation of RB by viraloncoproteins

A survey of human cervical carcinoma cell lines foundthat two out of seven bear small inactivating mutations inRB (Scheffner et al., 1991). Whereas neither of these lineswere infected by human papillomavirus (HPV), each of theremaining five that expressed normal RB also containedHPV DNA (Scheffner et al., 1991). HPVs play anetiologic role in most cervical neoplasias (Vousden, 1995).The E7 oncoprotein encoded by HPV can transformestablished cell lines and has also been shown to bind toRB (Dyson et al., 1989; Munger et al., 1989). SomeHPVs, such as HPV-16 and -18, are associated withpotentially pre-cancerous genital tract lesions and a largepercentage of anogenital cancers, whereas others, such asHPV-6 and -11, are associated with benign proliferativetumours with a low risk of malignant progression (e.g.condyloma acuminata) (Vousden, 1995). E7 proteins fromthe high risk viruses HPV-16 and -18 have higher bindingaffinity for RB than E7 from the lower risk types HPV-6and -11 (Heck et al., 1992; Munger et al., 1989). Singleresidue substitutions in HPV-6 E7 that cause a substantialincrease in affinity for RB also produce a concomitant gainin transforming activity (Heck et al., 1992; Sang andBarbosa, 1992). It is therefore likely that the ability of E7to bind RB contributes significantly to the oncogeniccapacity of HPVs. Therefore, RB function may be lost inmost if not all cervical cancers; this occurs by genemutation in the minority of HPV-negative cases and bycomplex formation with E7 protein in the remaininginstances (Scheffner et al., 1991).

The transforming proteins of several other DNAtumour viruses can also bind RB and neutralize itsfunction (Vousden, 1995). This property is shown by thelarge T antigen of simian virus 40 (SV40) (DeCaprio etal., 1988; Ewen et al., 1989; Ludlow et al., 1989; Moran,1988) and the E1A protein of adenovirus (Whyte et al.,1988, 1989). Mutagenesis studies have shown that theregions of these oncoproteins that are necessary for bindingRB are also required for their transforming properties(DeCaprio et al., 1988; Ewen et al., 1989; Moran, 1988;Whyte et al., 1989). Furthermore, the parts of RB that areneeded for association with E1A and T antigen are alsocommon sites for mutations (Hu et al., 1990). By bindingto RB, these viral proteins can interfere with its normal


615

cellular functions and thereby mimick the effects of the Rbmutations that occur in many tumours.

C. Inactivation of RB by phosphorylation

RB can be switched off through phosphorylation(Hunter and Pines, 1994; Pines, 1995; Sherr, 1994;Weinberg, 1995). This constitutes a normal controlmechanism that is used to regulate progress through thecell cycle (Hunter and Pines, 1994; Pines, 1995; Sherr,1994; Weinberg, 1995). Thus, RB is underphosphorylatedduring the first two thirds of G1 phase and whilst in thiscondition it helps prevent cells from passing through theR point (Hunter and Pines, 1994; Pines, 1995; Sherr,1994; Weinberg, 1995). Near the end of G1, if conditionsare propitious, RB becomes phosphorylated at multiplesites and loses its ability to inhibit passage into S phase(Hunter and Pines, 1994; Pines, 1995; Sherr, 1994;Weinberg, 1995). Its affinity for the nuclear compartmentis also diminished (Mittnacht and Weinberg, 1991). Thecyclin D- and cyclin E-dependent kinases are responsiblefor controlling RB in this way (Hunter and Pines, 1994;Pines, 1995; Sherr, 1994; Weinberg, 1995).

The activity of cyclin D-dependent kinases isabnormally elevated in a variety of cancers and thisprovides another mechanism whereby RB function is lost(Bates and Peters, 1995; Hunter and Pines, 1994; Pines,1995; Weinberg, 1995). The gene for cyclin D1 isamplified in at least 15% of primary breast cancers and aneven greater proportion of squamous cell carcinomas of theneck, head, oesophagus and lung (Bates and Peters, 1995;Hunter and Pines, 1994). Furthermore, cyclin D1 RNAand protein is overexpressed in 30-40% of primary breasttumours, suggesting that gene amplification is not theonly mechanism contributing to increased levels of theproduct (Bates and Peters, 1995). In some parathyroidadenomas and B cell lymphomas, chromosomaltranslocations cause overproduction of cyclin D1 (Batesand Peters, 1995; Hunter and Pines, 1994). When Epstein-Barr virus immortalizes B-lymphocytes, cyclin D2becomes activated (Sinclair et al., 1994). The gene forcyclin-dependent kinase 4 is amplified in manyglioblastomas and some gliomas (Weinberg, 1995). Inaddition to these diverse situations in which cyclins ortheir associated kinases are activated directly, many othercancers lose the function of p16 and/or p15, which areimportant repressors of the cyclin D-dependent kinases(Hirama and Koeffler, 1995; Hunter and Pines, 1994;Weinberg, 1995). For example, the genes for p16 and p15are deleted in many glioblastomas, oesophageal, bladder,lung and pancreatic carcinomas, and are sometimes mutatedin familial melanomas (Hirama and Koeffler, 1995;Weinberg, 1995). Thus, the cyclin D-dependent kinasesbecome abnormally active in a broad spectrum of cancersthrough a variety of mechanisms. This has the effect ofswitching off RB.

It is therefore certain that RB function is lost in a highproportion of tumours. Indeed, it has been suggested thatthe control pathway involving RB may become deregulated

in all human malignancies (Weinberg, 1995). This can beachieved in a variety of different ways - gene mutation,association with viral oncoproteins, orhyperphosphorylation. A good illustration of theimportance of inactivating RB during tumour progressionwas provided by a survey of small cell lung carcinomas(Otterson et al., 1994). This study tested 55 small celllung cancers and found that 48 lacked normal RBexpression but contained wild-type p16; six out of theremaining seven lacked functional p16 (Otterson et al.,1994).

III. RB targets

A. E2F

As explained above, RB acts as a signal transducerwhich controls gene expression in response to theavailability of growth factors. It does this by targetting anumber of key transcription factors and regulating theirfunctions. Perhaps the best characterised of these is E2F(Adams and Kaelin, 1995; La Thangue, 1994; Lam and LaThangue, 1994; Weinberg, 1996). E2F is a heterodimerictranscription factor composed of an E2F polypeptide and aDP polypeptide. In vertebrates, five E2F genes and threeDP genes have been identified (Adams and Kaelin, 1995).Heterodimerization results in a synergistic increase in boththe DNA-binding and transcriptional activation functionsof these proteins (Bandara et al., 1993; Helin et al., 1993;Krek et al., 1993). It also enhances the ability to recognizeRB (Helin et al., 1993; Krek et al., 1993). Not only doesRB mask the transactivation domain of E2F, but it canexert a dominant silencing activity that repressespromoters with E2F-binding sites (Weintraub et al.,1992). When growth factors are limiting RB isunderphosphorylated and active; it binds to E2F andinhibits it (Figure 1 ). Following serum stimulation,RB becomes phosphorylated at multiple sites by the cyclinD-dependent kinases; this inactivates it and causes it todissociate from E2F, thereby allowing the expression ofE2F-responsive genes (Adams and Kaelin, 1995).

Table 1 lists some of the genes that contain E2Fsites in their promoters. Many of these have been shownto be regulated by E2F, but it has not been proven inevery case. These potential target genes can be dividedinto five categories. One group consists of genes encodingsubunits of E2F, which suggests that autoregulation mayoccur. A second category contains Rb and the related genep107, which implies further opportunities for feedbackcontrol. The next class consists of the oncogenes B-myb ,N-myc and c-myc. Another group contains several genesthat are directly involved in driving the cell cycle,including components of the cyclin-dependent kinases andthe cdc25C phosphatase that activates these. The fifth andlargest group consists of many genes that encodecomponents of the DNA replication apparatus, includingDNA polymerase !, the origin recognition factor HsOrc1,and several enzymes involved in nucleotide biosynthesis.A striking feature of this list is that many of the geneswith E2F sites would be predicted to contribute to cellular


616

proliferation. This is the case for the oncogenes and forcdc2 and the cyclins, which have a positive effect on cell

Figure 1 . When growth factors are limiting, RB binds toE2F and represses its ability to activate transcription.Following serum stimulation, RB becomeshyperphosphorylated at multiple sites through the action ofcyclin-dependent kinases. This inactivates RB and causes itto dissociate from E2F, which allows expression of E2F-responsive genes.

Table 1. Genes regulated by E2F

1. E2F Components- E2F-1, -4, -5- DP-12 . Pocket prote ins- RB- p1073 . Oncogenes- B-myb- c-myc- N-myc4. Genes that dr ive the ce l l cyc le- cdc2- cyclin A- cyclin D- cyclin E- cdc25C5. Genes required for DNA replication

- DNA polymerase !

- HsOrc1- PCNA- topoisomerase I- thymidylate synthase- thymidine kinase- ribonucleotide reductase

cycle progression. Furthermore, DNA replication isclearly a prerequisite of productive cell division. Onewould therefore predict that by inhibiting the expression ofthe batteries of genes listed in Table 1 , through itsrepressive effect on E2F, RB would be able to achieve avery potent block upon proliferation.

As yet it is unclear how many of the genes with E2Fsites are actually regulated by RB. The most stringent test

of this is to look for changes in expression following thespecific deletion of the Rb gene. Very few of the geneslisted in Table 1 pass this test. One study of primarymouse embryonic fibroblasts (MEFs) examined ten geneswith E2F sites and found that only cyclin E and p107synthesis were changed following homozygousinactivation of Rb (Hurford et al., 1997). During G0 andG1 phases, cyclin E and p107 mRNA levels were twofold

higher in Rb-/- MEFs compared to wildtype controls(Hurford et al., 1997). The expression of B-myb, cdc2,E2F-1, TS, RRM2, cyclin A2, DHFR, TK, DNApolymerase , and Cdc25C genes were unaffected by theRB-knockout (Hurford et al., 1997). Another study founda ten-fold increase in cyclin E protein and a two- to

fourfold increase in cyclin D1 when Rb-/- MEFs were

compared to the corresponding Rb+/+cells (Herrera et al.,1996). The surprising lack of effect that deleting Rb hasupon most E2F target genes is probably due to redundancyin the RB family. RB has two close relatives called p107and p130 (Figure 2 ). These three proteins showsubstantial similarity in primary sequence and are thoughtto perform overlapping functions (Whyte, 1995). They aremost highly related in a bipartite domain called the pocket,which is responsible for binding E1A and E2F (Whyte,1995). As a consequence, they are sometimes referred toas the pocket proteins. It may be that when Rb is deleted,its relatives can assume many of its functions. Indeed, adouble knockout of Rb and p107 has a more severephenotype than single knockouts of either (Lee et al.,1996). This is certainly consistent with a functionaloverlap. However, p107 and p130 are much more similarto each other than they are to RB. Indeed, whereas RBspecifically targets E2F-1, -2 and -3, p107 and p130 appear

to bind only E2F-4 and E2F-5 (Weinberg, 1995). A p107-

/-/p130-/- double knockout strongly derepresses B-myb buthas no effect on cyclin E (Hurford et al., 1997). Ittherefore seems that p107 and p130 can only assume someof the functions that are performed by RB. The moststriking difference between the pocket proteins is that p107and p130 have never been found to be mutated in cancers.

Even if many of the genes listed in Table 1 are notsubject to control by RB, repressing the synthesis ofcyclins E and D1 through its action on E2F should initself be sufficient to provide a brake upon cell cycleprogression and hence proliferation. Indeed, under certaincircumstances dominant-negative mutants that abolish E2Factivity can block the cell cycle (Dobrowliski et al.,1994;Wu et al., 1996). However, this is by no means the


617

Figure 2 . Regions of homology between all three pocketproteins are shown as black blocks. Regions that arehomologous between p107 and p130, but are not shared byRB, are shaded.

whole story. In molar terms, RB is two orders ofmagnitude more abundant than E2F within the cell(Weinberg, 1995). This suggests that RB regulatesadditional targets besides E2F. Indeed, one study foundthat the proliferation rate of epithelial cells is not affectedwhen endogenous E2F is inactivated using dominant-negative mutants (Bargou et al., 1996). It is thereforehighly likely that E2F-independent pathways contribute tothe physiological effects of RB. A diverse array of cellularproteins have been shown to bind RB (Taya, 1997; Whyte,1995). Some of these are listed in Table 2 . Theyinclude the tyrosine kinase c-Abl (Welch and Wang, 1993)and the factors BRM and BRG1 that are involved incontrolling nucleosome structure (Dunaief et al., 1994;Singh et al., 1995). Several others are transcriptionfactors. I shall concentrate on two of these, UBF andTFIIIB, which may have key roles in controlling cellgrowth.

Table 2. Some of the cellular proteins that interact with RB

PROTEIN FUNCTIONE2F Transcription factorUBF Transcription factorTFIIIB Transcription factorElf-1 Transcription factorPU.1 Transcription factorBrm Reorganising chromatin structureBRG1 Reorganising chromatin structureMyoD Transcription factorATF-2 Transcription factorMDM2 Oncoproteinc-Abl Tyrosine kinaseD-type cyclins Cdk activatorsPP1 Protein phosphatase

B. UBF

A nucleolar transcription factor called UBF wasidentified as an RB-binding protein by screening a cDNAexpression library with purified RB as probe (Shan et al.,1992). Subsequent studies have confirmed the ability ofrecombinant RB to bind specifically to UBF (Cavanaughet al., 1995; Voit et al., 1997). Furthermore,immunoprecipitation assays with cellular extractsdemonstrated that endogenous RB and UBF associate whenpresent at physiological ratios (Cavanaugh et al., 1995).The identification of UBF as a target for RB wassomewhat unexpected. All previous studies on RB hadconcentrated on genes that are transcribed by RNApolymerase II (pol II), which synthesizes the messengerRNA (mRNA) in cells. However, UBF is only involvedin regulating transcription by pol I, the polymeraseresponsible for synthesizing large ribosomal RNA(rRNA). UBF binds to the promoters of rRNA genes andstimulates transcription in several ways; it helps fold theDNA and it recruits pol I and an essential factor called SL1or TIF-IB (Reeder et al., 1995). In vitro experimentsdemonstrated that recombinant RB can indeed inhibit thesynthesis of large rRNA by pol I (Cavanaugh et al., 1995;Voit et al., 1997). Whereas RB represses rRNAproduction in the presence of UBF, it does not affect thelow level of basal transcription that occurs in a UBF-depleted system (Cavanaugh et al., 1995). RB diminishesspecifically the ability of UBF to bind to DNA (Voit etal., 1997). The physiological relevance of these resultswas confirmed by immunofluorescence analyses of intactcells (Rogalsky et al., 1993; Cavanaugh et al., 1995).This approach allows one to visualise the nucleolus,which is the site of synthesis of large rRNA. It was foundthat RB accumulates in the nucleolus when cells stopgrowing, in parallel with a decrease in pol I activity(Rogalsky et al., 1993; Cavanaugh et al., 1995).Furthermore, immunoprecipitation experiments showedthat the interaction between RB and UBF increases whencells down-regulate pol I transcription as their rate ofgrowth decreases (Cavanaugh et al., 1995). Thus, there is aclear in vivo correlation between growth arrest, theassociation of RB with UBF, and the repression of rRNAsynthesis.

C. TFIIIB

Although 5.8S, 18S and 28S rRNAs are made by pol Ias a single precursor transcript, the 5S rRNA is madeseparately by pol III, the largest and most complex of theeukaryotic RNA polymerases (White, 1994). 5S rRNAsynthesis is independent of E2F and UBF, but isnevertheless repressed by RB (White et al., 1996; Larminieet al., 1997). Indeed, RB appears to be capable ofinhibiting the production of all pol III products, includingtransfer RNA (tRNA), the U6 small nuclear RNA(snRNA) that is required for splicing, and the adenoviralVA RNAs that are involved in subverting the host cell's


618

Figure 3 . RB can regulate transcription by all three nuclear RNA polymerases. Pol I synthesizes large rRNA; pol IIsynthesizes mRNA; and pol III synthesizes tRNA and 5S rRNA. A simplified basal transcription complex is drawn for eachpolymerase; additional basal factors are required that are not shown in the figure. RB represses pol I via the factor UBF. RBrepresses a subset of pol II templates via gene-specific regulatory factors such as E2F. RB represses pol III via the general factorTFIIIB.

translational apparatus (White et al., 1996; Larminie et al.,1997). Initial evidence that this is the case came frombiochemical assays which tested whether RB can regulatepol III in a cell-free system. We found that addingrecombinant RB to a system reconstituted usingfractionated factors repressed expression of every pol IIItemplate tested, whereas control protein had little or noeffect (White et al., 1996). Support for the in vivorelevance of these observations came from transienttransfection experiments, which showed thatoverexpressing RB can repress pol III transcription withoutaffecting a control promoter (White et al., 1996). Theseresults demonstrated for the first time that high levels ofRB can inhibit pol III activity.

Overexpressing proteins at abnormally elevated levelscan sometimes force proteins into artifactual interactions.It was therefore important to determine whether RB plays asignificant role in controlling pol III when present atphysiological concentrations within a cell. To begin toaddress this, we compared two human osteosarcoma celllines; SAOS2, which expresses only a truncatednonfunctional form of RB, and U2OS, which containswild-type RB. SAOS2 cells were shown to express atransfected pol III template 5-fold more actively thanU2OS cells (White et al., 1996). Transcription assayscarried out using extracted proteins confirmed the higheractivity of the pol III factors from the RB-negative SAOS2cells (White et al., 1996). As a more rigorous test of the

function of endogenous RB, we made use of knockoutmice in which the the Rb gene had been inactivated bysite-directed mutagenesis. Nuclear run-on assays were usedto measure directly the transcription of endogenous genesin intact nuclei from primary MEFs of these RB-knockoutmice. We found that tRNA and 5S rRNA synthesis by pol

III is 5-fold more active in the Rb-/- cells than inequivalent fibroblasts from wild-type mice (White et al.,1996). In contrast, the total level of pol II transcription isnot increased when the Rb gene is deleted (White et al.,1996). In vitro assays with extracted factors againestablished that the increased production of tRNA and 5SrRNA in the RB-negative MEFs is due to a more activepol III transcription apparatus (White et al., 1996). Since

the only genetic difference between the Rb+/+ and the Rb-

/- fibroblasts is the presence of the Rb gene, these resultsestablished that endogenous RB plays a very major role insuppressing the level of pol III transcription in vivo.

RB appears to regulate tRNA and rRNA synthesis bytargetting a factor called TFIIIB (Larminie et al., 1997).TFIIIB is a multisubunit complex that contains theTATA-binding protein (TBP) and at least two additionalpolypeptides (Rigby, 1993; White, 1994). Its function isto recruit pol III to the appropriate promoters and positionit at the transcription start site (Kassavetis et al., 1990;White, 1994). We found that recombinant RB interactswith TFIIIB and represses it specifically (Larminie et al.,


619

1997). Furthermore, immunoprecipitation andcofractionation experiments indicated that a population ofendogenous RB molecules associates with TFIIIB atphysiological concentrations (Larminie et al., 1997). Thisinteraction is diminished or abolished in SAOS2osteosarcoma cells, which contain only a truncated mutantform of RB (Larminie et al., 1997). The activity of TFIIIBis elevated specifically in primary fibroblasts from RB-deficient mice (Larminie et al., 1997). These resultsestablished that TFIIIB is a target for repression by RB(Figure 3). This conclusion fits well with previous dataindicating that TFIIIB activity rises as cells progress fromG1 into S phase, the time when RB is silenced throughhyperphosphorylation (White et al., 1995).

A subsequent investigation by Chu et al. (1997)provided independent support for these analyses. Thisstudy confirmed that overexpressing RB represses pol IIItranscription in transfected cells and in vitro (Chu et al.,1997). Consistent with the earlier investigations, RB wasshown to bind to TFIIIB (Chu et al., 1997). Furthermore,clustered substitutions in RB that disrupt the interactionwith TFIIIB also prevent repression (Chu et al., 1997). Inaddition, Chu et al. (1997) reported an interaction betweenoverexpressed RB and another pol III factor called TFIIIC2.A model was proposed in which RB utilises distinctdomains to bind either TFIIIB or TFIIIC2 (Chu et al.,1997). However, there was little correlation betweenTFIIIC2 binding and the ability of RB mutants to represspol III transcription (Chu et al., 1997). Furthermore, thereis no evidence that TFIIIC2 and RB interact when presentat physiological ratios. Chu et al. (1997) concluded thatTFIIIB is the principal target for RB-mediated repression ofpol III, but that a subsidiary interaction with TFIIIC2 mayalso contribute to the effect.

Although it has been shown that RB binds to one ormore of the general pol III factors, it remains to bedetermined how this leads to transcriptional repression.One possibility is that RB blocks interactions withpromoter DNA. Precedent for this is provided by the pol Isystem, where RB interferes with the DNA-bindingproperties of UBF (Voit et al., 1997). An alternative isthat RB disrupts the structure of TFIIIB in some way. Agrowth suppressor called Dr1 has been shown to use thismechanism (White et al., 1994). Dr1 inhibits tRNAsynthesis both in vitro and in vivo (White et al., 1994;Kim et al., 1997). It achieves this by displacing one ofthe essential subunits of TFIIIB from its interaction withTBP (White et al., 1994). Other possible mechanismsmight involve RB disrupting the protein-proteininteractions between TFIIIB and TFIIIC or pol III. Orderof addition experiments showed that the pol III factorsremain susceptible to RB even after they have beenassembled into a stable preinitiation complex on the VAIpromoter (Larminie et al., 1997).

TFIIIB is required for all pol III transcription (White,1994; Willis, 1993). Therefore by repressing TFIIIB, RBcan provide blanket repression of all pol III templates.

This contrasts strongly with the situation for pol II, whereonly a small proportion of promoters, such as those withE2F sites, are controlled by RB. The majority of genesthat are transcribed by pol II are not affected directly by thepresence of RB (White et al., 1996). Therefore, RB is agene-specific regulator of pol II but a general regulator ofpol III. This distinction is meaningless in the pol Isystem, since pol I only transcribes a single highlyreiterated template that encodes the large rRNA.

The number of genes that are controlled by E2F isrelatively small and very few of these become activated in

Rb-/- knockouts (Herrera et al., 1996; Hurford et al.,1997). UBF probably regulates a larger number of genes,since there are ~400 copies of the large rRNA template indiploid human cells (Long and Dawid, 1980). It remainsto be determined whether these are affected by knockingout RB. The number of promoters that require TFIIIBexceeds this by over three orders of magnitude. Thus, adiploid human cell contains around a million Alu genes,2600 tRNA genes, 600 5S rRNA genes, 200 U6 snRNAgenes and a range of other less abundant pol III templates(White, 1994). All of these need TFIIIB to be expressed(White, 1994). The tRNA and 5S genes have been shown

to be activated in Rb-/- knockouts, whereas the otherclasses have yet to be tested in this way (White et al.,1996). Since deleting Rb results in a five-fold increase intRNA and 5S rRNA production (White et al., 1996), it ishighly likely that the majority of these genes are subjectto repression by RB. These observations suggest that thepol III templates constitute by far the largest category ofgenes that are controlled directly by RB.

IV. Control of growth and proliferationby RB

Both the growth (increase in mass) and proliferation(increase in number) of cells are suppressed by RB. It isessential that these two processes are coordinated, becausea significant imbalance can trigger apoptosis (Kung et al.,1993; Qin et al., 1994; Rueckert and Mueller, 1960; Shanand Lee, 1994). In order to maintain a constant size, a cellmust ensure that all its components are duplicated at asimilar rate. Thus, DNA content and protein levelsgenerally increase in parallel (Stanners et al., 1979) andattempts to dissociate them with specific inhibitors canhave lethal consequences (Kung et al., 1993). The controlof proliferation by RB can be largely explained by itsability to regulate E2F. As described above, E2F regulatesa range of pol II-transcribed genes that promote cell cycleprogression (Adams and Kaelin, 1995; La Thangue, 1994;Lam and La Thangue, 1994; Weinberg, 1996). Theseinclude several genes that are required for DNA replication,such as those encoding DNA polymerase ! and thereplication origin-binding protein HsOrc1, as well as genesthat drive the cell cycle, such as cyclin A and cdc2 (Adamsand Kaelin, 1995; Weinberg, 1996). By repressing some ofthese through its inhibitory effect on E2F, RB can often


620

Figure 4 . Mechanisms that may enable RB to restrict cell growth and proliferation. E2F promotes cell cycle progression. Itappears to do this by activating the synthesis of proteins required for DNA replication, such as thymidine kinase (TK),

dihydrofolate reductase (DHFR) and DNA polymerase !, as well as proteins that drive the cell cycle, such as cdc2 and cyclins D and

E. By repressing E2F, RB may limit the production of these products and therefore provide a brake on proliferation. UBF andTFIIIB are required for the synthesis of rRNA and tRNA, essential raw materials for protein synthesis. By repressing thesetranscription factors, RB may be able to limit the rate of translation and therefore provide a brake on cellular growth.

provide a brake on DNA replication and passage throughthe cell cycle. However, the genes regulated by E2F areprimarily involved in controlling proliferation and providefew obvious links to the control of growth. Control ofE2F on its own may therefore be insufficient to achieve abalanced regulation of both growth and proliferation. Onecould imagine that growth is somehow tied to the cellcycle, so that regulating the latter is sufficient to achieveindirect control of the former. However, most of theavailable evidence argues against this (Nasmyth, 1996). Infact, in bacteria and in yeast the dependence worksprimarily the other way round, with cell cycle progressionrequiring attainment of a critical mass (Nasmyth, 1996).The basic principles observed in microrganisms are likelyto be conserved in higher orders. For example, murinefibroblasts must reach a certain mass before they caninitiate DNA synthesis (Killander and Zetterberg, 1965).Furthermore, a survey of mammalian cell types found thatin most cases size continues to increase when DNAsynthesis is inhibited using aphidicolin (Kung et al.,1993). HeLa cells provided a striking exception, and inthis line biosynthesis and growth decrease in response tothe cell cycle block (Kung et al., 1993). HeLa cells arehighly abnormal and it is likely that their anomalousbehaviour does not provide a reliable indicator of thecontrol mechanisms that operate in most animal cells. Ingeneral, the growth of mammalian cells appears not todepend on chromosome duplication, at least in the shortterm. Thus, the control of proliferation seems insufficientto explain fully the ability of RB to inhibit cell growth.

One possibility is that this is achieved by regulating theproduction of tRNA and rRNA, which are majordeterminants of biosynthetic capacity (Nasmyth, 1996;White, 1997). (Figure 4).

A substantial weight of evidence shows that theregulation of protein synthesis is an important aspect ofgrowth control. When cells quiesce, tRNA and rRNAlevels decrease, polysomes disperse into free ribosomes,and the overall rate of protein accumulation is reduced.Following mitogenic stimulation, the production oftRNA, rRNA, ribosomal proteins and translation factorsaccelerates and protein synthesis increases before cellsreach S phase (Clarke et al., 1996; Johnson et al., 1974;Kief and Warner, 1981; Mauck and Green, 1974; Redpathand Proud, 1994; Rosenwald, 1996a,b; Stanners et al.,1979; Tatsuka et al., 1992). Ribosome content isproportional to the rate of growth (Kief and Warner, 1981).Indeed, careful measurements in animal cells havedemonstrated that growth rate is directly proportional tothe rate of protein accumulation (Baxter and Stanners,1978). The main determinant of protein accumulation istranslation, although turnover also makes a significantcontribution (Baxter and Stanners, 1978). A 50% reductionin the rate of protein synthesis is sufficient to causeproliferating cells to withdraw from cycle and quiesce(Brooks, 1977; Ronning et al., 1981). Translation isclearly dependent on the availability of tRNA and rRNA.By limiting the production of these, RB may be able tosuppress the level of protein synthesis, which could inturn provide a brake on cellular growth. As yet, this


621

should still be regarded as speculation, although it must betrue to some extent. However, it remains to be determinedto what degree the activities of pols I and III are everlimiting for growth under physiological conditions. Agood indication that they may be came from a study carriedout in S. cerevisiae. It was found that in this yeast a two-fold reduction in the level of initiator tRNA results in athree-fold increase in doubling time (Francis andRajbhandary, 1990). If the same were true of amammalian cell, then the 5-fold depression in tRNAsynthesis that is imposed by RB must surely have asubstantial impact upon the rate of growth. Duringtumour development, when RB function is compromised,the release of pol III from this major constraint may be animportant step towards neoplasia.

V. RNA polymerase III and cancer

If RB plays an important physiological role inrestraining pol III transcription, then one would expect tofind that pol III activity is elevated in a broad range ofcancers, where RB function is compromised. This isindeed the case. Many studies have observed that theabundance of pol III transcripts is abnormally elevated intransformed and tumour cells. This was first discovered byKramerov et al. (1982), who examined carcinoma andplasmacytoma lines. Subsequent work extended theobservation to include cells that have been transformed byDNA tumour viruses, RNA tumour viruses, or chemicalcarcinogens (Brickell et al., 1983; Carey and Singh, 1988;Carey et al., 1986; Kramerov et al., 1990; Lania et al.,1987; Majello et al., 1985; Ryskov et al., 1985; Scott etal., 1983; Singh et al., 1985; White et al., 1990). Thisactivation is very general, but not universal, there being afew examples of transformed lines that do not display thecharacteristic increase in pol III transcript levels (Ryskovet al., 1985; Scott et al., 1983). A tight causal linkbetween pol III activation and transformation is suggestedby the fact that two fibroblast lines transformed bytemperature-sensitive mutants of the SV40 large T antigendown-regulate pol III transcription at the non-permissivetemperature whilst reverting to normal morphology andphenotype (Scott et al., 1983). The abundance of pol IIItranscripts varies substantially between different SV40-transformed lines and the highest levels correlate withprogression to a more tumorigenic phenotype (Scott et al.,1983; White et al., 1990).

A recent study provided convincing evidence that a polIII product is induced in rodent tumours. This investigationexamined a pol III transcript called BC1, which is unusualbecause it is normally only expressed in neurons(DeChiara and Brosius, 1987). The function of BC1 hasyet to be determined. Northern analysis showed BC1expression in breast carcinomas, colonic adenocarcinomasand skin fibrosarcomas, but not in the correspondinguntransformed tissues (Chen et al., 1997). In situhybridisation studies of theses tumours confirmed thepresence of BC1 RNA in the neoplastic cells, whereas itwas absent from the surrounding tissues (Chen et al.,

1997). Although the fibrosarcomas and adenocarcinomaswere induced by local inoculation with cells that had beentreated with chemical carcinogens, the breast carcinomaanalysed was a primary tumour induced by ras (Chen et al.,1997). Similar studies have shown that BC200 RNA, theprimate analogue of BC1, is expressed in many, but notall, primary human tumours (Chen et al., 1997). LikeBC1, BC200 RNA is found exclusively in the malignantcells and not in the adjacent normal tissue (Chen et al.,1997). Thus, abnormal activation of pol III expression is afrequent feature of tumours in vivo.

As already explained, RB function is lost in manyhuman cancers through a variety of mechanisms. It willbe important to determine to what extent this isresponsible for activating pol III. Deletion andsubstitution analyses have demonstrated that the RBsequences which control pol III correspond to the domainsthat are mutated frequently in tumours (White et al., 1996;Chu et al., 1997). Indeed, the minimal region of RB thatis necessary to regulate cell growth and proliferation isalso sufficient to repress transcription by pol III (White etal., 1996). Several examples have been characterised ofhighly localised mutations that inactivate RB in humancancers. For example, in one small cell lung carcinoma asingle base change in a splice acceptor site gave rise to anRB polypeptide that lacked the 35 amino acids encoded byexon 21 (Horowitz et al., 1990). In another small cell lungcarcinoma, a point mutation created a stop codon and anovel splice donor site within exon 22, therebyeliminating 38 residues from the product (Horowitz et al.,1990). A third inactivating mutation from a small celllung cancer resulted in a single amino acid substitution atcodon 706 (Kaye et al., 1990). We tested the ability ofeach of these three naturally occurring mutants to regulatepol III transcription and found that repression was lost inevery case (White et al., 1996). Although this is clearly alimited survey, it nevertheless demonstrates a correlationbetween the function of RB as a tumour suppressor and itsability to control pol III.

As described above, the E1A oncoprotein of adenovirusand the large T antigen of SV40 bind and neutralize RB, aproperty which is important for their transformingcapabilities (DeCaprio et al., 1988; Ewen et al., 1989;Ludlow et al., 1989; Moran, 1988; Whyte et al., 1988;Whyte et al., 1989). Both E1A and T antigen can alsostimulate the rate of pol III transcription (Loeken et al.,1988; Patel and Jones, 1990; White et al., 1996). Wefound that E1A and T antigen can release pol III fromrepression by RB (White et al., 1996). These viraloncoproteins are believed to regulate gene expressionthrough multiple mechanisms, but one way in which theycan stimulate pol III involves overcoming thephysiological constraint that is normally provided by RB.In cells transformed by E1A or T antigen, the loss of RBfunction is likely to contribute substantially to anactivation of pol III transcription.


622

VI. Components of the translationapparatus are often deregulated in cancercells

There is a multitude of documented examples in whichthe translation machinery has become deregulatedfollowing transformation (Rosenwald, 1996a). This factprovides strong support for the contention that the controlof protein synthesis is an important aspect of growthregulation. For example, fibroblasts transformed bypolyoma virus were found to synthesize protein morerapidly than normal parental cells and have lost controlover ribosome production (Stanners et al., 1979). Inuntransformed revertants of these fibroblasts, correctregulation is recovered (Stanners et al., 1979). One studycompared levels of expression of ribosomal proteins incolorectal tumours from eight different individuals withnormal colonic mucosa from the same patients (Pogue-Geile et al., 1991). In every case, the adenocarcinomasoverexpressed all six ribosomal protein transcripts thatwere tested (Pogue-Geile et al., 1991). The levels of thesemRNAs were also generally elevated in adenomatouspolyps, the presumed precursors of the carcinomas (Pogue-Geile et al., 1991). This implies that increased ribosomalprotein production occurs early during the development ofthese tumours, perhaps concomitant with the onset ofneoplasia. Colorectal tumours and tumour-derived celllines were also reported to produce higher levels of rRNAsthan normal colonic mucosa, consistent with a generalincrease in ribosomal components (Pogue-Geile et al.,1991). Constitutive expression of EF-1!, a translationfactor that catalyses the attachment of aminoacyl-tRNAs tothe ribosome, makes fibroblasts highly susceptible totransformation by 3-methylcholanthrene or ultravioletlight (Tatsuka et al., 1992). These observations suggestthat deregulation of the protein synthesis machinery canpredispose cells to malignant transformation.

Perhaps the most striking demonstration that thetranslation apparatus becomes activated duringtumourigenesis came from a recent study that used serialanalysis of gene expression (SAGE) to document theexpression profiles of 45,000 genes in gastrointestinaltumours (Zhang et al., 1997). Only 108 pol II transcriptswere found to be expressed at higher levels in primarycolon cancers relative to normal colonic epithelium (Zhanget al., 1997). Of these, 48 encode ribosomal proteins and 5encode translation elongation factors (Zhang et al., 1997).Similar results were obtained with pancreatic cancers(Zhang et al., 1997). These observations providecompelling evidence that deregulation of protein synthesisis intimately linked with tumour formation.

Several studies have shown that the abnormalactivation of translation factors is actually sufficient totrigger neoplastic transformation (Rosenwald, 1996a). eIF-4E, the mRNA cap-binding protein, and eIF-2, whichbrings initiator methionine tRNA to the 40S ribosomalsubunit, appear particularly important in this regard. eIF-4E is the least abundant of the translation initiation factorsand is rate limiting for protein synthesis (Duncan et al.,

1987). As such, it is of key importance in controlling therate of translation. Overexpression of eIF-4E in varioustypes of fibroblast stimulates growth and proliferation andinduces morphological transformation (Lazaris-Karatzas etal., 1990; Lazaris-Karatzas and Sonenberg, 1992). Cellswith abnormally high eIF-4E levels also induce tumoursin nude mice (Lazaris-Karatzas et al., 1990; Lazaris-Karatzas and Sonenberg, 1992). Similarly, overexpressionof eIF-4E in HeLa cells accelerates growth and results inthe formation of overcrowded multilayered foci (DeBenedetti and Rhoads, 1990). Conversely, reducing thelevel of eIF-4E with antisense RNA inhibits the growth ofHeLa cells (De Benedetti et al., 1991) and thetumorigenicity of ras-transformed fibroblasts (Rinkerr-Schaffer et al., 1993).

In serum-starved cells, the recycling of eIF-2 isinhibited by phosphorylation of its ! subunit, therebyimpairing translational initiation (Redpath and Proud,1994). Mutation of eIF-2! so that it can no longer bephosphorylated causes malignant transformation of NIH3T3 cells (Donze et al., 1995). Malignancy can also beinduced by dominant negative forms of the eIF-2! kinase,whereas the wild-type kinase inhibits growth whenoverexpressed in mammalian fibroblasts or yeast (Chonget al., 1992; Koromilas et al., 1992; Meurs et al., 1993).The eIF-2! kinase (which is also referred to as PKR) isinducible by interferon and is likely to contribute to theaction of interferons as growth inhibitors and anti-tumouragents (Clemens, 1992; Lengyel, 1993).

The cellular activity of eIF-2! and eIF-4E increases inresponse to various oncogenes (Rosenwald, 1996b). Thelevels of eIF-2! and eIF-4E mRNA and protein areelevated in fibroblasts that overexpress c-myc (Rosenwaldet al., 1993a,b; Rosenwald, 1995; Jones et al., 1996;Rosenwald, 1996b). This is associated with acceleratedrates of protein accumulation and cell growth (Rosenwald,1996b). v-src and v-abl have similar effects on eIF-2! andeIF-4E, but this may reflect the ability of theseoncoproteins to stimulate c-myc production (Rosenwald etal., 1993a,b; Rosenwald, 1995, 1996b). v-src and ras alsoincrease the phosphorylation of eIF-4E, which can activateits function (Frederickson et al., 1991; Rinker-Schaeffer etal., 1992). In addition, ras can deregulate eIF-2 by inducingan inhibitor of eIF-2! kinase (Mundschau and Faller,1992). These many examples provide abundant evidencethat abnormal stimulation of the translation apparatus is afrequent characteristic of transformed cells. This supportsthe idea that elevated rates of protein synthesis arenecessary to sustain the development of many tumours.

VII. c-Myc: a foot in both camps?

The oncogene c-myc may have a foot in both thegrowth and proliferation camps. The c-myc promotercontains an E2F binding site and is subject to repressionby RB (Hiebert et al., 1989; Zou et al., 1997). If c-mycexpression is prevented using antisense technology, cellsstop growing and arrest in G1 phase (Heikkila et al., 1987;Prochownik et al., 1988). Myc has been shown to


623

promote cell cycle progression by stimulating productionof cdc25 and the activation of cyclin D and E-dependentkinases (Galaktionov et al., 1996; Steiner et al., 1995). In

addition, the activation of c-myc in rodent fibroblastsresults in an increase in the abundance of the translation

Figure 5 . Control of c-myc production may provide an additional mechanism for RB to influence both growth andproliferation. Myc can stimulate cell cycle progression through the activation of cyclin-dependent kinases (cdks). Myc may

promote protein synthesis and cellular growth by increasing the production of the translation initiation factors eIF-2! and eIF-4E.

eIF-4E can stimulate cyclin D1 production. This, in turn, might be expected to switch off RB and thereby activate UBF and TFIIIB.The c-myc promoter contains a binding site for E2F and may therefore be subject to repression by RB. Inhibiting the productionof myc may provide a mechanism for RB to control both growth and proliferation.

initiation factors eIF-2! and eIF-4E (Jones et al., 1996;Rosenwald, 1996a,b; Rosenwald et al., 1993a,b). Theelevated concentrations of eIF-2! and eIF-4E thataccompany activation of c-myc correlate with a rise in thenet rate of protein synthesis and accelerated growth(Rosenwald, 1996b). Furthermore, overexpression of eIF-4E results in a selective increase in cyclin D1 production(Rosenwald et al., 1993a). This, in turn, might beexpected to switch off RB and thereby activate UBF andTFIIIB. By silencing the c-myc promoter, RB may beable to suppress both proliferation and growth. (Figure5 ).

VIII. Discussion

There is substantial evidence that the deregulation oftranslation is an important aspect of neoplastictransformation. Rapid growth undoubtedly requires elevatedrates of protein accumulation; without it, a tumour wouldbe unable to maintain its increase in mass. For rapidlydividing cells to sustain a high rate of translation will

require efficient production of tRNA and rRNA. Inaddition to the clear correlation between proteinaccumulation and growth, constitutively elevatedtranslation might drive a population to proliferate. Thiscould work as follows: unbalanced growth in the absenceof cell replication is likely to trigger apoptosis; suchconditions may select for cells that have acquired theability to bypass the apoptotic pathway and multiplycontinuously.

In this review, I have drawn a clear distinction betweengrowth and proliferation. I have also argued that separatemechanisms appear to be involved in controlling theseprocesses. However, the point must be emphasised thatgrowth and proliferation are intimately linked and there isundoubtedly substantial cross-talk between the two. Manypotential examples of this can be envisaged. For example,E2F is involved in regulating the genes for cyclins A, Dand E. Since these cyclins control kinases that caninactivate RB, there is obvious potential for a feedbackloop. Moreover, through its action on cyclins and henceRB, E2F might be expected to influence UBF and TFIIIB


624

activity, and hence translation and growth. As explainedabove, c-myc may also have direct impact upon both thecell cycle machinery and the translation apparatus. It isclearly of benefit to the cell to have growth andproliferation coordinated by unifying control mechanisms.RB may provide such a regulatory switch.

RB is a tumour suppressor of major importance, witha key role in controlling cell growth and proliferation. Byregulating E2F, RB has the potential to inhibit thesynthesis of gene products that are necessary for DNAsynthesis, chromosomal replication and cell cycleprogression. By repressing UBF and TFIIIB, RB can reducethe production of tRNA and rRNA. This may allow it tolimit the rate of protein accumulation, which will providea brake on cellular growth. Coregulating these essentialprocesses may allow RB to achieve the necessary balancebetween growth and proliferation (White, 1997). Manyother molecular targets have been identified for RB andthese provide additional controls over cellular activity(Taya, 1997; Whyte, 1995). Regulating a range of keycomponents may enable RB to coordinate a number ofdisparate processes. The loss of these controls willundoubtedly constitute a major step towards tumourdevelopment.

Acknowledgements

I apologise to any colleagues whose contributions maynot have been mentioned due to limitations of space. Ithank S. Mittnacht, N. La Thangue, T. Hunt and K.Nasmyth for helpful and stimulating discussionsconcerning these ideas. Thanks also to C. Larminie and J.Sutcliffe for comments on the manuscript. Research inmy laboratory is supported by the Cancer ResearchCampaign, the Medical Research Council, theBiotechnology and Biological Sciences Research Counciland the Nuffield Foundation. I am a Jenner ResearchFellow of the Lister Institute of Preventive Medicine.

References

Adams, P. D., and Kaelin, W. G. (1 9 9 5 ). Transcriptionalcontrol by E2F. Seminars in Cancer Bio logy 6, 99-108.

Bandara, L. R., Buck, V. M., Zamanian, M., Johnston, L. H.,and La Thangue, N. B. (1 9 9 3 ). Functional synergybetween DP-1 and E2F-1 in the cell cycle-regulatingtranscription factor DRTF1/E2F. EMBO J. 13, 4317-4324.

Bargou, R. C., Wagener, C., Bommert, K., Arnold, W.,Daniel, P. T., Mapara, M. Y., Grinstein, E., Royer, H. D.,and Dorken, B. (1 9 9 6 ). Blocking of transcription factorE2F/DP by dominant-negative mutants in a normal breastepithelial cell line efficiently inhibits apoptosis andinduces tumour growth in SCID mice. J Exp Med 183,1205-1213.

Bates, S., and Peters, G. (1 9 9 5 ). Cyclin D1 as a cellularproto-oncogene. Seminars in Cancer B i o l o g y 6,73-82.

Baxter, G. C., and Stanners, C. P. (1 9 7 8 ). The effect ofprotein degradation on cellular growth characteristics. JC e l l P h y s i o l 96, 139-146.

Bookstein, R., Shew, J., Chen, P., Scully, P., and Lee, W.(1 9 9 0 ). Suppression of tumorigenicity of human prostatecarcinoma cells by replacing a mutated RB gene. Sc ience247, 712-715.

Brickell, P. M., Latchman, D. S., Murphy, D., Willison, K.,and Rigby, P. W. J. (1 9 8 3 ). Activation of a Qa/Tla class Imajor histocompatibility antigen gene is a general featureof oncogenesis in the mouse. Nature 306, 756-760.

Brooks, R. F. (1 9 7 7 ). Continuous protein synthesis isrequired to maintain the probability of entry into S phase.Cel l 12, 311-317.

Carey, M. F., and Singh, K. (1 9 8 8 ). Enhanced B2transcription in simian virus 40-transformed cells ismediated through the formation of RNA polymerase IIItranscription complexes on previously inactive genes.Proc Natl Acad Sci USA 85, 7059-7063.

Carey, M. F., Singh, K., Botchan, M., and Cozzarelli, N. R.(1 9 8 6 ). Induction of specific transcription by RNApolymerase III in transformed cells. M o l C e l l B i o l 6 ,3068-3076.

Cavanaugh, A. H., Hempel, W. M., Taylor, L. J., Rogalsky,V., Todorov, G., and Rothblum, L. I. (1 9 9 5 ). Activity ofRNA polymerase I transcription factor UBF blocked by Rbgene product. Nature 374, 177-180.

Chen, W., Heierhorst, J., Brosius, J., and Tiedge, H. (1 9 9 7 ).Expression of neural BC1 RNA: induction in murinetumours. Eur. J. Cancer 33, 288-292.

Chong, K. L., Feng, L., Schappert, K., Meurs, E., Donahue, T.F., Friesen, J. D., Hovanessian, A. G., and Williams, B.R. G. (1 9 9 2 ). Human p68 kinase exhibits growthsuppression in yeast and homology to the translationalregulator GCN2. EMBO J 11, 1553-1562.

Chu, W.-M., Wang, Z., Roeder, R. G., and Schmid, C. W.(1 9 9 7 ). RNA polymerase III transcription repressed byRb through its interactions with TFIIIB and TFIIIC2. JBio l Chem 272, 14755-14761.

Clarke, A. R., Maandag, E. R., van Roon, M., van der Lugt, N.M. T., van der Valk, M., Hooper, M. L., Berns, A., and teRiele, H. (1 9 9 2 ). Requirement for a functional Rb-1 genein murine development. Nature 359, 328-330.

Clarke, E. M., Peterson, C. L., Brainard, A. V., and Riggs, D.L. (1 9 9 6 ). Regulation of the RNA polymerase I and IIItranscription systems in response to growth conditions. JBio l Chem 271, 22189-22195.

Clemens, M. (1992). Suppression with a difference. Nature360, 210-211.

De Benedetti, A., Joshi-Barve, S., Rinker-Schaffer, C., andRhoads, R. E. (1 9 9 1 ). Expression of antisense RNAagainst initiation factor eIF-4E mRNA in HeLa cellsresults in lengthened cell division times, diminishedtranslation rates and reduced levels of both eIF-4E andp220 component of eIF-4F. M o l C e l l B i o l 11, 5435-5445.

De Benedetti, A., and Rhoads, R. E. (1 9 9 0 ). Overexpressionof eukaryotic protein synthesis initiation factor 4E inHeLa cells results in abberant growth and morphology.Proc Natl Acad Sci USA 87, 8212-8216.


625

DeCaprio, J. A., Ludlow, J. W., Figge, J., Shew, J.-Y., Huang,C.-M., Lee, W.-H., Marsilio, E., Paucha, E., andLivingston, D. M. (1 9 8 8 ). SV40 large tumour antigenforms a specific complex with the product of theretinoblastoma susceptibility gene. Cel l 54, 275-283.

DeChiara, T. M., and Brosius, J. (1 9 8 7 ). Neural BC1 RNA:cDNA clones reveal nonrepetitive sequence content. ProcNatl Acad Sci USA 84, 2624-2628.

Dobrowliski, S. F., Stacey, D. W., Harter, M. L., Stine, J. T.,and Hiebert, S. W. (1 9 9 4 ). An E2F dominant negativemutant blocks E1A induced cell cycle progression.Oncogene 9, 2605-2612.

Donze, O., Jagus, R., Koromilas, A. E., Hershey, J. W. B.,and Sonenberg, N. (1 9 9 5 ). Abrogation of translationinitiation factor eIF-2 phosphorylation causes malignanttransformation of NIH 3T3 cells. EMBO J 14, 3828-3834.

Dunaief, J. L., Strober, B. E., Guha, S., Khavari, P. A., Alin,K., Luban, J., Begemann, M., Crabtree, G. R., and Goff,S. P. (1 9 9 4 ). The retinoblastoma protein and BRG1 forma complex and cooperate to induce cell cycle arrest. Cel l79, 119-130.

Duncan, R., Milburn, S. C., and Hershey, J. W. (1 9 8 7 ).Regulated phosphorylation and low abundance of HeLacell initiation factor eIF-4F suggests a role intranslational control. Heat shock effects on eIF-4F. JBio l Chem 262, 380-388.

Dyson, N., Howley, P. M., Munger, K., and Harlow, E.(1 9 8 9 ). The human papillomavirus-16 E7 oncoprotein isable to bind to the retinoblastoma gene product. Sc ience243, 934-937.

Ewen, M. E., Ludlow, J. W., Marsilio, E., DeCaprio, J. A.,Millikan, R. C., Cheng, S. H., Paucha, E., andLivingston, D. M. (1 9 8 9 ). An N-terminaltransformation-governing sequence of SV40 large T

antigen contributes to the binding of both p110Rb and asecond cellular protein, p120. Cel l 58, 257-267.

Francis, M. A., and Rajbhandary, U. L. (1 9 9 0 ). Expressionand function of a human initiator tRNA gene in the yeastSaccharomyces cerevisiae. M o l C e l l B i o l 10, 4486-4494.

Frederickson, R. M., Montine, K. S., and Sonenberg, N.(1 9 9 1 ). Phosphorylation of eukaryotic translationinitiation factor 4E is increased in src-transformed celllines. M o l C e l l B i o l 1 1 , 2896-2900.

Friend, S. H., Bernards, R., Rogelj, S., Weinberg, R. A.,Rapaport, J. M., Alberts, D. M., and Dryja, T. P. (1 9 8 6 ).A human DNA segment with properties of the gene thatpredisposes to retinoblastoma and osteosarcoma. Nature323, 643-646.

Galaktionov, K., Chen, X., and Beach, D. (1 9 9 6 ). Cdc25cell-cycle phosphatase as a target of c-myc. Nature 382,511-517.

Hamel, P. A., Gill, R. M., Phillips, R. A., and Gallie, B. L.(1 9 9 2 ). Transcriptional repression of the E2-containingpromoters EIIaE, c-myc, and RB1 by the product of theRB1 gene. M o l C e l l B i o l 12, 3431-3438.

Heck, D. V., Yee, C. L., Howley, P. M., and Munger, K.(1 9 9 2 ). Efficiency of binding the retinoblastoma proteincorrelates with the transforming capacity of the E7

oncoproteins of the human papillomaviruses. Proc NatlAcad Sci USA 89, 4442-4446.

Heikkila, R., Schwab, G., Wickstrom, E., Loke, S. L.,Pluznik, D. H., Watt, R., and Neckers, L. M. (1 9 8 7 ). A c-myc antisense oligodeoxynucleotide inhibits entry into Sphase but not progress from G0 to G1. Nature 328, 445-

449.

Helin, K., Wu, C.-L., Fattaey, A. R., Lees, J. A., Dynlacht, B.D., Ngwu, C., and Harlow, E. (1 9 9 3 ). Heterodimerizationof the transcription factors E2F-1 and DP-1 leads tocooperative transactivation. Genes Dev 7, 1850-1861.

Herrera, R. E., Sah, V. P., Williams, B. O., Makela, T. P.,Weinberg, R. A., and Jacks, T. (1 9 9 6 ). Altered cell cyclekinetics, gene expression, and G1 restriction point

regulation in Rb-deficient fibroblasts. M o l C e l l B i o l16, 2402-2407.

Hiebert, S. W., Lipp, M., and Nevins, J. R. (1 9 8 9 ). E1A-dependent trans-activation of the human MYC promoter ismediated by the E2F factor. Proc Nat l Acad Sci USA86, 3594-3598.

Hirama, T., and Koeffler, H. P. (1 9 9 5 ). Role of the cyclin-dependent kinase inhibitors in the development of cancer.Blood 86, 841-854.

Horowitz, J. M., Park, S.-H., Bogenmann, E., Cheng, J.-C.,Yandell, D. W., Kaye, F. J., Minna, J. D., Dryja, T. P.,and Weinberg, R. A. (1 9 9 0 ). Frequent inactivation of theretinoblastoma anti-oncogene is restricted to a subset ofhuman tumour cells. Proc Nat l Acad Sc i USA 87,2775-2779.

Hu, N., Gutsmann, A., Herbert, D. C., Bradley, A., Lee, W. H.,

and Lee, E. Y.-H. P. (1 9 9 4 ). Heterozygous Rb-1"20/+

mice are predisposed to tumours of the pituitary gland witha nearly complete penetrance. Oncogene 9, 1021-1027.

Hu, Q., Dyson, N., and Harlow, E. (1 9 9 0 ). The regions of theretinoblastoma protein needed for binding to adenovirusE1A or SV40 large T antigen are common sites formutations. EMBO J 9, 1147-1155.

Huang, H.-J. S., Yee, J.-K., Shew, J.-Y., Chen, P.-L.,Bookstein, R., Friedmann, T., Lee, E. Y.-H. P., and Lee,W.-H. (1 9 8 8 ). Suppression of the neoplastic phenotypeby replacement of the RB gene in human cancer cells.Sc ience 242, 1563-1566.

Hunter, T., and Pines, J. (1 9 9 4 ). Cyclins and cancer II: cyclinD and CDK inhibitors come of age. Cel l 79, 573-582.

Hurford, R. K., Cobrinik, D., Lee, M.-H., and Dyson, N.(1 9 9 7 ). pRB and p107/p130 are required for the regulatedexpression of different sets of E2F responsive genes.Genes Dev 11, 1447-1463.

Jacks, T., Fazeli, A., Schmitt, E. M., Bronson, R. T.,Goodell, M. A., and Weinberg, R. A. (1 9 9 2 ). Effects ofan Rb mutation in the mouse. Nature 359, 295-300.

Johnson, L. F., Abelson, H. T., Green, H., and Penman, S.(1 9 7 4 ). Changes in RNA in relation to growth of thefibroblast. I. Amounts of mRNA, rRNA, and tRNA inresting and growing cells. Cel l 1, 95-100.

Jones, R. M., Branda, J., Johnston, K. A., Polymenis, M.,Gadd, M., Rustgi, A., Callanan, L., and Schmidt, E. V.(1 9 9 6 ). An essential E box in the promoter of the geneencoding the mRNA cap-binding protein (eukaryotic


626

initiation factor 4E) is a target for activation by c-myc.M o l C e l l B i o l 16, 4754-4764.

Kassavetis, G. A., Braun, B. R., Nguyen, L. H., andGeiduschek, E. P. (1 9 9 0 ). S. cerevisiae TFIIIB is thetranscription initiation factor proper of RNA polymeraseIII, while TFIIIA and TFIIIC are assembly factors. Cel l60, 235-245.

Kaye, F. J., Kratzke, R. A., Gerster, J. L., and Horowitz, J. M.(1 9 9 0 ). A single amino acid substitution results in aretinoblastoma protein defective in phosphorylation andoncoprotein binding. P r o c N a t l A c a d S c i U S A 87,6922-6926.

Kief, D. R., and Warner, J. R. (1 9 8 1 ). Coordinate control ofsyntheses of ribosomal ribonucleic acid and ribosomalproteins during nutritional shift-up in Saccharomycescerevisiae. M o l C e l l B i o l 1, 1007-1015.

Killander, D., and Zetterberg, A. (1 9 6 5 ). A quantitativecytochemical investigation of the relationship betweencell mass and initiation of DNA synthesis in mousefibroblasts in vitro. Exp Cel l Res 40, 12-20.

Kim, S., Na, J. G., Hampsey, M., and Reinberg, D. (1 9 9 7 ).The Dr1/DRAP1 heterodimer is a global repressor oftranscription in vivo. P r o c N a t l A c a d S c i U S A 94,820-825.

Koromilas, A. E., Roy, S., Barber, G. N., Katze, M. G., andSonenberg, N. (1 9 9 2 ). Malignant transformation by amutant of the IFN-inducible dsRNA-dependent proteinkinase. Sc ience 257, 1685-1689.

Kramerov, D. A., Lekakh, I. V., Samarina, O. P., and Ryskov,A. P. (1 9 8 2 ). The sequences homologous to majorinterspersed repeats B1 and B2 of mouse genome are

present in mRNA and cytoplasmic poly(A)+ RNA.Nucleic Acids Res 10, 7477-7491.

Kramerov, D. A., Tillib, S. V., Shumyatsky, G. P., andGeorgiev, G. P. (1 9 9 0 ). The most abundant nascent

poly(A)+ RNAs are transcribed by RNA polymerase III inmurine tumor cells. Nucleic Acids Res 18, 4499-4506.

Krek, W., Livingston, D. M., and Shirodkar, S. (1 9 9 3 ).Binding to DNA and the retinoblastoma gene productpromoted by complex formation of different E2F familymembers. Sc ience 262, 1557-1560.

Kung, A. L., Sherwood, S. W., and Schimke, R. T. (1 9 9 3 ).Differences in the regulation of protein synthesis, cyclinB accumulation, and cellular growth in response to theinhibition of DNA synthesis in Chinese hamster ovaryand HeLa S3 cells. J Biol Chem 268, 23072-23080.

La Thangue, N. B. (1 9 9 4 ). DRTF1/E2F: an expanding familyof heterodimeric transcription factors implicated in cell-cycle control. Trends Biochem Sci 19, 108-114.

Lam, E. W.-F., and La Thangue, N. B. (1 9 9 4 ). DP and E2Fproteins: coordinating transcription with cell cycleprogression. Current Opin ion in Ce l l B io logy 6 ,859-866.

Lania, L., Pannuti, A., La Mantia, G., and Basilico, C.(1 9 8 7 ). The transcription of B2 repeated sequences isregulated during the transition from quiescent toproliferative state in cultured rodent cells. FEBS Lett219, 400-404.

Larminie, C. G. C., Cairns, C. A., Mital, R., Martin, K.,Kouzarides, T., Jackson, S. P., and White, R. J. (1 9 9 7 ).Mechanistic analysis of RNA polymerase III regulation bythe retinoblastoma protein. EMBO J 16, 2061-2071.

Lazaris-Karatzas, A., Montine, K. S., and Sonenberg, N.(1 9 9 0 ). Malignant transformation by a eukaryoticinitiation factor subunit that binds to mRNA 5' cap.Nature 345, 544-546.

Lazaris-Karatzas, A., and Sonenberg, N. (1 9 9 2 ). The mRNA5' cap-binding protein, eIF-4E, cooperates with v-myc orE1A in the transformation of primary rodent fibroblasts.Mol Ce l l B io l 12, 1234-1238.

Lee, E. Y.-H. P., Chang, C.-Y., Hu, N., Wang, Y.-C. J., Lai,C.-C., Herrup, K., Lee, W.-H., and Bradley, A. (1 9 9 2 ).Mice deficient for Rb are nonviable and show defects inneurogenesis and haematopoiesis. Nature 359, 288-294.

Lee, M.-H., Williams, B. O., Mulligan, G., Mukai, S.,Bronson, R. T., Dyson, N., Harlow, E., and Jacks, T.(1 9 9 6 ). Targeted disruption of p107: functional overlapbetween p107 and Rb. Genes Dev 10, 1621-1632.

Lengyel, P. (1 9 9 3 ). Tumour-suppressor genes: news aboutthe interferon connection. P r o c N a t l A c a d S c i U S A90, 5893-5895.

Loeken, M., Bikel, I., Livingston, D. M., and Brady, J.(1 9 8 8 ). Trans-activation of RNA polymerase II and IIIpromoters by SV40 small t antigen. Cel l 55, 1171-1177.

Long, E. O., and Dawid, I. B. (1 9 8 0 ). Repeated genes ineukaryotes. Annu Rev Biochem 49, 727-764.

Ludlow, J. W., DeCaprio, J. A., Huang, C.-M., Lee, W.-H.,Paucha, E., and Livingston, D. M. (1 9 8 9 ). SV40 large Tantigen binds preferentially to an underphosphorylatedmember of the retinoblastoma susceptibility gene productfamily. Cel l 56, 57-65.

Maandag, E. C. R., van der Valk, M., Vlaar, M., Feltkamp, C.,O'Brien, J., van Roon, M., van der Lugt, N., Berns, A.,and te Riele, H. (1 9 9 4 ). Developmental rescue of anembryonic-lethal mutation in the retinoblastoma gene inchimeric mice. EMBO J 13, 4260-4268.

Majello, B., La Mantia, G., Simeone, A., Boncinelli, E., andLania, L. (1 9 8 5 ). Activation of major histocompatibilitycomplex class I mRNA containing an Alu-like repeat inpolyoma virus-transformed rat cells. Nature 314, 457-459.

Mauck, J. C., and Green, H. (1 9 7 4 ). Regulation of pre-transfer RNA synthesis during transition from resting togrowing state. Cel l 3, 171-177.

Meurs, E. F., Galabru, J., Barber, G. N., Katze, M. G., andHovanessian, A. G. (1 9 9 3 ). Tumour suppressor functionof the interferon-induced double-stranded RNA-activatedprotein kinase. P r o c N a t l A c a d S c i U S A 90, 232-236.

Mittnacht, S., and Weinberg, R. A. (1 9 9 1 ). G1/Sphosphorylation of the retinoblastoma protein isassociated with an altered affinity for the nuclearcompartment. Cel l 65, 381-393.

Moran, E. (1 9 8 8 ). A region of SV40 large T antigen cansubstitute for a transforming domain of the adenovirusE1A products. Nature 334, 168-170.


627

Mundschau, L. J., and Faller, D. V. (1 9 9 2 ). Oncogenic rasinduces an inhibitor of double-stranded RNA-dependent

eukaryotic initiation factor 2!-kinase activation. J B i o l

Chem 267, 23092-23098.

Munger, K., Werness, B. A., Dyson, N., Phelps, W. C.,Harlow, E., and Howley, P. M. (1 9 8 9 ). Complexformation of human papillomavirus E7 proteins with theretinoblastoma tumour suppressor gene product. EMBO J8, 4099-4105.

Nasmyth, K. (1 9 9 6 ). Another role rolls in. Nature 382, 28-29.

Nikitin, A. Y., and Lee, W.-H. (1 9 9 6 ). Early loss of theretinoblastoma gene is associated with impaired growthinhibitory innervation during melanotroph

carcinogenesis in Rb+/- mice. Genes Dev 10, 1870-1879.

Otterson, G. A., Kratzke, R. A., Coxon, A., Kim, Y. W., and

Kaye, F. J. (1 9 9 4 ). Absence of p16Ink4 protein isrestricted to the subset of lung cancer lines that retainswildtype RB. Oncogene 9, 3375-3378.

Pardee, A. B. (1 9 8 9 ). G1 events and regulation of cellproliferation. Sc ience 246, 603-608.

Patel, G., and Jones, N. C. (1 9 9 0 ). Activation in vitro ofRNA polymerase II and III directed transcription bybaculovirus produced E1A protein. Nucle i c Ac ids Res18, 2909-2915.

Pines, J. (1 9 9 5 ). Cyclins, CDKs and cancer. Seminars inCancer Bio logy 6, 63-72.

Pogue-Geile, K., Geiser, J. R., Shu, M., Miller, C., Wool, I.G., Meisler, A. I., and Pipas, J. M. (1 9 9 1 ). Ribosomalprotein genes are overexpressed in colorectal cancer:isolation of a cDNA clone encoding the human S3ribosomal protein. M o l C e l l B i o l 11, 3842-3849.

Prochownik, E. V., Kukowska, J., and Rodgers, C. (1 9 8 8 ). c-myc antisense transcripts accelerate differentiation andinhibit G1 progression in murine erythroleukaemia cells.M o l C e l l B i o l 8, 3683-3695.

Qin, X., Chittenden, T., Livingston, D. M., and Kaelin, W. G.(1 9 9 2 ). Identification of a growth suppression domainwithin the retinoblastoma gene product. G e n e s D e v 6 ,953-964.

Qin, X.-Q., Livingston, D. M., Kaelin, W. G., and Adams, P.D. (1 9 9 4 ). Deregulated transcription factor E2F-1expression leads to S-phase entry and p53-mediatedapoptosis. Proc Natl Acad Sc i USA 91, 10918-10922.

Redpath, N. T., and Proud, C. G. (1 9 9 4 ). Molecularmechanisms in the control of translation by hormonesand growth factors. Bioch im B i o p h y s Acta 1220,147-162.

Reeder, R. H., Pikaard, C. S., and McStay, B. (1 9 9 5 ). UBF,an architectural element for RNA polymerase I promoters.Nucle ic Ac ids and Molecular B i o l o g y , 9, 251-263.

Rigby, P. W. J. (1 9 9 3 ). Three in one and one in three: it alldepends on TBP. Cel l 72, 7-10.

Rinker-Schaeffer, C. W., Austin, V., Zimmer, S., and Rhoads,R. E. (1 9 9 2 ). Ras transformation of cloned rat embryo

fibroblasts results in increased rates of protein synthesisand phosphorylation of eukaryotic initiation factor 4E. JBio l Chem 267, 10659-10664.

Rinkerr-Schaffer, C. W., Graff, J. R., De Benedetti, A.,Zimmer, S. G., and Rhoads, R. E. (1 9 9 3 ). Decreasing thelevel of translation initiation factor 4E with antisenseRNA causes reversal of ras-mediated transformation andtumorigenesis of cloned rat embryo fibroblasts. Int JCancer 55, 841-847.

Rogalsky, V., Todorov, G., and Moran, D. (1 9 9 3 ).Translocation of retinoblastoma protein associated withtumour cell growth inhibition. Biochem B i o p h y sRes Comm 192, 1139-1146.

Ronning, O. W., Lindmo, T., Pettersen, E. O., and Seglen, P.O. (1 9 8 1 ). The role of protein accumulation in the cellcycle control of human NHIK 3035 cells. J Ce l lP h y s i o l 109, 411-418.

Rosenwald, I. B. (1 9 9 5 ). Growth-factor independentexpression of the gene encoding eukaryotic translationinitiation factor 4E in transformed cell lines. CancerLett 98, 77-82.

Rosenwald, I. B. (1 9 9 6 a ). Deregulation of protein synthesisas a mechanism of neoplastic transformation.B i o e s s a y s 18, 243-250.

Rosenwald, I. B. (1 9 9 6 b ). Upregulated expression of thegenes encoding translation initiation factors eIF-4E and

eIF-2! in transformed cells. Cancer Lett 102, 113-123.

Rosenwald, I. B., Lazaris-Karatzas, A., Sonenberg, N., andSchmidt, E. V. (1 9 9 3 a ). Elevated expression of cyclinD1 protein in response to increased expression ofeukaryotic initiation factor 4E. Mol Ce l l B i o l 13,7358-7363.

Rosenwald, I. B., Rhoads, D. B., Callanan, I. D., Isselbacher,K. J., and Schmidt, E. V. (1 9 9 3 b ). Increased expression

of translation initiation factors eIF-4E and eIF-2! in

response to growth induction by c-myc. Proc NatlAcad Sci USA 90, 6175-6178.

Rueckert, R. R., and Mueller, G. C. (1 9 6 0 ). Studies onunbalanced growth in tissue culture. I. Induction andconsequences of thymidine deficiency. Cancer Res 20,1584-1592.

Ryskov, A. P., Ivanov, P. L., Tokarskaya, O. N., Kramerov,D. A., Grigoryan, M. S., and Georgiev, G. P. (1 9 8 5 ).Major transcripts containing B1 and B2 repetitive

sequences in cytoplasmic poly(A)+ RNA from mousetissues. FEBS Letters 182, 73-76.

Sang, B.-C., and Barbosa, M. S. (1 9 9 2 ). Single amino acidsubstitutions in "low-risk" human papillomavirus (HPV)type 6 E7 protein enhance features characteristic of the"high-risk" HPV E7 oncoproteins. Proc Nat l Acad SciUSA 89, 8063-8067.

Scheffner, M., Munger, K., Byrne, J. C., and Howley, P. M.(1 9 9 1 ). The state of the p53 and retinoblastoma genes inhuman cervical carcinoma cell lines. Proc Nat l AcadSci USA 88, 5523-5527.

Scott, M. R. D., Westphal, K.-H., and Rigby, P. W. J.(1 9 8 3 ). Activation of mouse genes in transformed cells.Cel l 34, 557-567.


628

Shan, B., and Lee, W.-H. (1 9 9 4 ). Deregulated expression ofE2F-1 induces S-phase entry and leads to apoptosis. M o lC e l l B i o l 14, 8166-8173.

Shan, B., Zhu, X., Chen, P.-L., Durfee, T., Yang, Y., Sharp,D., and Lee, W.-H. (1 9 9 2 ). Molecular cloning of cellulargenes encoding retinoblastoma-asociated proteins:identification of a gene with properties of thetranscription factor E2F. Mol Ce l l B i o l 12, 5620-5631.

Sherr, C. J. (1 9 9 4 ). G1 phase progression: cycling on cue.Cel l 79, 551-555.

Sinclair, A. J., Palmero, I., Peters, G., and Farrell, P. J.(1 9 9 4 ). EBNA-2 and EBNA-LP cooperate to cause G0 toG1 transition during immortalization of resting human B-lymphocytes by Epstein-Barr virus. EMBO J 13, 3321-3328.

Singh, K., Carey, M., Saragosti, S., and Botchan, M.(1 9 8 5 ). Expression of enhanced levels of small RNApolymerase III transcripts encoded by the B2 repeats insimian virus 40-transformed mouse cells. Nature 314,553-556.

Singh, P., Coe, J., and Hong, W. (1 9 9 5 ). A role forretinoblastoma protein in potentiating transcriptionalactivation by the glucocorticoid receptor. Nature 374,562-565.

Stanners, C. P., Adams, M. E., Harkins, J. L., and Pollard, J.W. (1 9 7 9 ). Transformed cells have lost control ofribosome number through their growth cycle. J Ce l lP h y s i o l 100, 127-138.

Steiner, P., Philipp, A., Lukas, J., Godden-Kent, D., Pagano,M., Mittnacht, S., Bartek, J., and Eilers, M. (1 9 9 5 ).Identification of a Myc-dependent step during theformation of active G1 cyclin-cdk complexes. EMBO J

14, 4814-4826.

Tatsuka, M., Mitsui, H., Wada, M., Nagata, A., Nojima, H.,

and Okayama, H. (1 9 9 2 ). Elongation factor-1! gene

determines susceptibility to transformation. Nature 359,333-335.

Taya, Y. (1 9 9 7 ). RB kinases and RB-binding proteins: newpoints of view. Trends Biochem Sci 22, 14-17.

Voit, R., Schafer, K., and Grummt, I. (1 9 9 7 ). Mechanism ofrepression of RNA polymerase I transcription by theretinoblastoma protein. Mol Ce l l B i o l 17, 4230-4237.

Vousden, K. H. (1 9 9 5 ). Regulation of the cell cycle by viraloncoproteins. Seminars in Cancer B io logy 6, 109-116.

Weinberg, R. A. (1 9 9 6 ). E2F and cell proliferation: a worldturned upside down. Cel l 85, 457-459.

Weinberg, R. A. (1 9 9 5 ). The retinoblastoma protein and cellcycle control. Cel l 81, 323-330.

Weintraub, S. J., Prater, C. A., and Dean, D. C. (1 9 9 2 ).Retinoblastoma protein switches the E2F site frompositive to negative element. Nature 358, 259-261.

Welch, P. J., and Wang, J. Y. J. (1 9 9 3 ). A C-terminalprotein-binding domain in the retinoblastoma proteinregulates nuclear c-Abl tyrosine kinase in the cell cycle.Cel l 75, 779-790.

White, R. J. (1 9 9 7 ). Regulation of RNA polymerases I and IIIby the retinoblastoma protein: a mechanism for growthcontrol? Trends Biochem Sci 22, 77-80.

White, R. J. (1 9 9 4 ). RNA polymerase III transcription. R.G. Landes Company, Austin .

White, R. J., Gottlieb, T. M., Downes, C. S., and Jackson, S.P. (1 9 9 5 ). Cell cycle regulation of RNA polymerase IIItranscription. M o l C e l l B i o l 15, 6653-6662.

White, R. J., Khoo, B. C.-E., Inostroza, J. A., Reinberg, D.,and Jackson, S. P. (1 9 9 4 ). The TBP-binding repressorDr1 differentially regulates RNA polymerases I, II and III.Sc ience 266, 448-450.

White, R. J., Stott, D., and Rigby, P. W. J. (1 9 9 0 ).Regulation of RNA polymerase III transcription inresponse to Simian virus 40 transformation. EMBO J 9,3713-3721.

White, R. J., Trouche, D., Martin, K., Jackson, S. P., andKouzarides, T. (1 9 9 6 ). Repression of RNA polymerase IIItranscription by the retinoblastoma protein. Nature 382,88-90.

Whyte, P. (1 9 9 5 ). The retinoblastoma protein and itsrelatives. Seminars in Cancer Bio logy 6, 83-90.

Whyte, P., Buchkovich, K. J., Horowitz, J. M., Friend, S. H.,Raybuck, M., Weinberg, R. A., and Harlow, E. (1 9 8 8 ).Association between an oncogene and an anti-oncogene:the adenovirus E1A proteins bind to the retinoblastomagene product. Nature 334, 124-129.

Whyte, P., Williamson, N. M., and Harlow, E. (1 9 8 9 ).Cellular targets for transformation by the adenovirus E1Aproteins. Cel l 56, 67-75.

Williams, B. O., Schmitt, E. M., Remington, L., Bronson, R.T., Albert, D. M., Weinberg, R. A., and Jacks, T. (1 9 9 4 ).Extensive contribution of Rb-deficient cells to adultchimeric mice with limited histopathologicalconsequences. EMBO J 13, 4251-4259.

Willis, I. M. (1 9 9 3 ). RNA polymerase III. Genes, factors andtranscriptional specificity. Eur J Biochem 212, 1-11.

Wu, C.-L., Classon, M., Dyson, N., and Harlow, E. (1 9 9 6 ).Expression of dominant-negative mutant DP-1 blocks cellcycle progression in G1. Mol Ce l l B i o l 16, 3698-3706.

Zhang, L., Zhou, W., Velculescu, V. E., Kern, S. E., Hruban,R. H., Hamilton, S. R., Vogelstein, B., and Kinzler, K. W.(1 9 9 7 ). Gene expression profiles in normal and cancercells. Sc ience 276, 1268-1272.

Zou, X., Rudchenko, S., Wong, K., and Calame, K. (1 9 9 7 ).Induction of c-myc transcription by the v-Abl tyrosinekinase requires Ras, Raf1, and cyclin-dependent kinases.Genes Dev 11, 654-662.


629


Transcriptional regulation of the H-ras1 proto-oncogene by DNA binding proteins: mechanismsand implications in human tumorigenesis

G. Zachos1,2 and D. A. Spandidos1,2

1 Institute of Biological Research and Biotechnology, National Hellenic Research Foundation, 48 Vas. Constantinou Ave.,

Athens 11635; 2 Medical School, University of Crete, Heraklion, Greece.

_______________________________________________________________________________________________

Correspondence to: Professor D.A. Spandidos, Tel/Fax: +(301)-722 6469

Summary

Altered expression of ras genes is a common event in human tumors. Transcriptional regulation ofthe H-ras1 proto-oncogene occurs through nuclear factors that recognize elements in the promoterregion of the gene, in the f irst and fourth intron and in the VTR unit and involves al ternativespl ic ing and specif ic methylat ion patterns, as wel l . Aberrant levels of the Ras p21 protein aredetected in a variety of human tumors and are often correlated with clinical and prognosticparameters. Thus, understanding the regulation of the expression of ras genes provides a usefultarget for gene therapy treatments.

I. Introduction

ras genes are a ubiquitous eukaryotic gene family.They have been identified in mammals, birds, insects,mollusks, plants, fungi and yeasts. Their sequence ishighly conserved, thus revealing the fundamental role theyplay in cellular proliferation (Spandidos, 1991).

A. The structure of ras genes

Three functional ras genes have been identified andcharacterized in the mammalian genome, H-ras1, K-ras2and N-ras, as well as two pseudogenes, H-ras2 and K-ras1(Barbacid, 1987). All three ras genes have a commonstructure with a 5' non-coding exon (exon -I) and fourcoding exons (exons I-IV). The introns of the genes differwidely in size and sequence, with the coding sequences ofhuman K-ras spanning more than 35 kb, while those of N-ras and H-ras span approximately 7 and 3 kb, respectively.The K-ras gene has two alternative IV coding exons, thusencoding two proteins, K-RasA and K-RasB (McGrath etal, 1983), with the K-RasB form being more abundant.The H-ras gene also has an alternative exon in the fourthintron (Cohen et al, 1989). In addition, H-ras has avariable tandem repeat sequence (VTR), locateddownstream of the polyadenylation signal, which exhibitsenhancer activity (Spandidos and Holmes, 1987, Cohen etal, 1987).

B. Ras proteins: structural characteristicsand function

The H-Ras, N-Ras and K-RasA proteins are 189amino acids long, whereas K-RasB is shorter by oneamino acid. They all have a molecular weight of 21kDaand are termed p21 proteins. The p21 proteins are identicalat the 86 N-terminal amino acid residues, they possess an85% homology in the next 80 amino acid residues anddiverge highly at the rest of the protein molecule, with theexception of the four C-terminal amino acids which sharethe common motif CAAX-COOH (C, Cysteine 186; A,Aliphatic amino acid-Leucine, Isoleucine or Valine; X,Methionine or Serine) (Lowy and Willumsen, 1991). TheRas protein is synthesized as pro-p21, undergoes a seriesof post-translational modifications at the C-terminusincreasing the hydrophobicity of the protein and associateswith the inner face of the plasma membrane. Sequences atthe C-terminus are essential for membrane association andthe conserved Cys 186 is required to initiate the post-translational modifications of pro-p21 (Willumsen andChristensen, 1984).

The superfamily of Ras proteins comprises a group ofsmall GTPases, regulating an astonishing diversity ofcellular functions (Makara et al, 1996). They are located atthe heart of a signal transduction pathway that links cell-surface receptors through a protein kinase cascade to

Zachos and Spandidos: Transcriptional regulation of the H-ras1 proto-oncogene

630

changes in gene expression and cell morphology and tocell division mechanisms. The Ras p21 protein interactsdirectly with the Raf oncoprotein to recruit the MAPkinases and their subordinates, thus converting a mitogenicsignal initiated by membrane receptors with tyrosinekinase activity to a cascade of Serine/ Threonine kinaseswith multiple targets, including cytoskeleton, transcriptionfactors, inflammatory mediators and other kinases (Avruchet al, 1994, Marshall, 1995).

C. ras oncogenes: mechanisms ofactivation

The ras family of proto-oncogenes, is a frequentlydetected family of transformation-inducing genes in humantumors. Implication of ras genes in human tumorigenesisoccurs by four different mechanisms: point mutations(Kiaris and Spandidos, 1995), gene amplification (Pulcianiet al, 1985), insertion of retroviral sequences (Westaway etal, 1986) and alterations in regulation of transcription(Zachos and Spandidos, 1997). With the exception ofmutations, all other mechanisms result in activation of thetransforming properties of ras genes by quantitativemechanisms.

The c-H-ras1 gene is the best studied member of thefamily and provides a good example for understanding themechanisms of gene regulation. The H-ras proto-oncogeneexpression is regulated by elements located in thepromoter region, in intronic sequences and in the 3' end ofthe gene. In addition, H-ras gene expression is regulatedby alternative mechanisms such as DNA methylation andalternative splicing (reviewed by Zachos and Spandidos,1997). Alterations in the H-ras expression levels are acommon mechanism of human tumorigenesis.

II. Transcriptional regulation of the H-ras gene from promoter-like sequences

The H-ras gene promoter contains multiple RNA startsites, multiple GC boxes and has no characteristic TATAbox (Ishii et al, 1985). These features are characteristic ofhousekeeping genes. Most promoter region studies havefocused on the region upstream of the 5' splice site of thefirst intron of the gene (nucleotides 1-577), althoughothers consider the SstI fragment (nucleotides 1-1054) thatencompasses a part of the first intron as well, to be thegene promoter (Spandidos et al, 1988).

Regulation of gene expression depends on a variety ofnuclear factors (Boulikas, 1994). A great number ofregulatory elements in the H-ras promoter has beenreported, but the results were often controversial,depending on the followed experimental procedure.Transcription factors that interplay on the regulatoryregions of the H-ras gene promoter include Sp-1, NF-1,AP-1 and some unknown factors as well.

The Sp-1 is a mammalian DNA binding proteinactivating transcription by interacting through zinc fingerdomains with guanine-rich DNA sequences called GC

boxes (Berg, 1992). The transcription factor AP-1 is thenuclear factor required to mediate transcription induced byphorbol ester tumor promoters and recognizes a shortTGACTCA sequence (Lewin, 1991). Both c-Jun, encodedby members from the jun family (jun, junB, junD), and c-Fos proteins are active components and contribute to theactivity of AP-1 by forming c-Jun homodimers as well asc-Jun-c-Fos heterodimers. In addition, there appears to be amutual antagonism between activation by AP-1 andglucocorticoid receptors at target genes that containrecognition sites for both factors, via protein-proteininteractions (Yang-Yen et al, 1990). Finally, the NF-I(CTF) nuclear factor binds the CCAAT element (CAATbox) and is involved in both gene transcription and DNAreplication. The NF-I C-terminal region is proline rich andactivates transcription through interference with thetranscription machinery (Mermod et al, 1989).

Ishii et al (1986), identified six GC boxes that bindthe Sp-1 transcription factor as the essential regulatoryelements within the H-ras promoter. Using deletionanalysis of the H-ras promoter region by focus formationassay in NIH 3T3 cells, Honkawa et al (1987) reported aminimum promoter region of 51 bp length, which wasGC rich (78%) and contained a GC box. Lowndes et al(1989) located a 47 bp element, distinct of the onereported by Honkawa et al, that upregulated thetranscriptional activity of the promoter region by 20- to40-fold and contained a GC box and a CCAAT box,binding the NF-1 (CTF) factor. Transient expressionassays in which a series of mutants spanning the promoterregion of H-ras were ligated to a promoterlesschloramphenicol acetyl transferase (CAT) vector, wereused in this analysis. Jones et al (1987), also identifiedtwo NF-I binding sites, one strong, also noted byHonkawa et al, and one weak. Trimble and Hozumi(1987), using CAT transfection experiments in CV-1cells, identified a 100 nucleotide region, encompassing theconsensus CCAAT box and two Sp-1 sites. However,Nagase et al (1990), using deletion mutants in CATassays in CV-1 and A-431 cells, suggested that thepresence of Sp-1 binding sites at specific positions maynot be essential for promoter activity, but a number ofSp-1 binding sites in the region could be required. Lee andKeller (1991), transfected recombinant plasmidsencompassing internal deletions and point mutations ofthe promoter region in HeLa cells and performed CATassays. They reported a GC box, an unidentified elementand a new element CCGGAA directly upstream the GCbox, as the most important regulatory elements.Spandidos et al (1988), using recombinant plasmids inCAT activity experiments showed that AP-1-like proteinsparticipate in control of H-ras transcription and identifiedfour TPA responsive-AP-1 binding elements in the H-raspromoter.

A great variety of the transcription initiation sites wasalso identified (Lowndes et al, 1989, Nagase et al, 1990,


631

Figure 1 . A synopsis of the reported nuclear factors participating in c-H-ras1 transcriptional regulation. Factors recognizesequences in the H-ras promoter, in intronic sequences and in the VTR region. Exons, rectangles; coding sequences, filledrectangles; VTR, cross-hatched box; B, BamHI cleavage site; S, SstI cleavage site; arrows, major transcriptional initiation sites;?, unknown regulatory factors; IDX, intron D exon.

Lee et al, 1991) using S1 nuclease analysis. A synopsisof the reported nuclear factors that participate on thepromoter activity of H-ras, and of the major RNA startsites, are shown in Fig . 1 .

III. Regulation of the H-ras geneexpression from intronic sequences

Intronic sequences play an important role in H-rasregulation. The nuclear factor Sp-1, steroid hormonereceptors and the P53 onco-suppressor protein recognizesequences in the first and fourth introns of the H-ras gene.

A. The Sp-1 box

There is evidence that the mutant T24 ras 0.8 kb SstIDNA fragment is a more potent activator of geneexpression, compared to the corresponding normal H-rasfragment (Spandidos and Pintzas, 1988). A structuralbasis for this difference was shown to be a 6 bp elementin the mutant H-ras fragment, that was absent in thenormal H-ras, in the first intron of the gene. This elementwas proved to contain an Sp-1 binding site (Pintzas andSpandidos, 1991) (Fig . 1 ).

B. The Hormone response elements (HREs)

Steroid hormone receptors produce an enormousnumber of biological effects in different tissues ashormone activated transcriptional regulators (Beato, 1989,Beato et al, 1995). Such a process requires the coordinateexpression of multiple genes and likely candidates aresignal transductors, capable of secondarily controlling thetranscription of sets of genes that lack steroid hormoneresponse elements. Proto-oncogenes are theoretically wellsuited for this role, as they exhibit precise temporalpatterns of expression during proliferation anddifferentiation, they participate in signal transduction andregulate the expression of multiple genes in a cascadefashion (Bishop, 1991). Thus, they may integrate signalsfrom steroids and other regulatory factors and amplify thecellular response to the hormone. Evidence for steroidhormone regulation of proto-oncogenes encoding fornuclear transcription factors has already been provided forc-fos, c-jun and c-myc (Hyder et al, 1994).

1. Regulation of c-H-ras by steroid hormonereceptors

Zachos et al (1995), identified sequences in the 3' endof the first intron and in the fourth intron of the H-ras


632

gene, with high similarity with the glucocorticoidresponse element (GRE) and estrogen response element(ERE) consensus oligonucleotides, respectively (F i g . 1 ).Using nuclear extracts from human and murine cell linesin gel retardation assays, it was shown that bothglucocorticoid receptor (GR) and estrogen receptor (ER)specifically recognize the corresponding H-ras elements(Zachos et al, 1995).

H-ras p21 protein, is a G-protein involved in signaltransduction beginning from transmembrane growth factorand peptide receptors with tyrosine kinase activity (Avruchet al, 1994, Marshall et al, 1995). Steroid receptors,participate in dinstinct signalling mechanisms, involvingtranscriptional regulation of genes by ligand activatedreceptors (Beato, 1989, Beato et al, 1995). Thus, hormoneregulation of H-ras provides evidence for a directinteraction of these two pathways, allowing the cell tohave an additional regulatory "switch" by hormonallymodulating the levels of G-proteins at the transcriptionallevel (Zachos et al, 1996a).

Moreover, the first intron of the gene contains wellconserved regions between human and rodents (Hashimoto-Gotoh et al, 1988) and encompasses positive and negativeelements influencing H-ras expression, possibly at post-transcriptional level (Hashimoto-Gotoh et al, 1992). It isnoteworthy that the GRE, as well as the p53 element thatwill be discussed later, are located in the conserved regionof the intron, thus providing evidence for an essential rolein regulating H-ras expression. Moreover, the H-ras GREis located within the first positive element.

2. Interaction of the H-ras and steroidhormone receptors in gynecological cancer

The human endometrium and ovary are major targetsfor action of glucocorticoids. Sex hormones and steroidsact as tumor promoters. The level of receptor binding inH-ras hormone response elements was examined in gelretardation assays, using nuclear extracts from humanendometrial and ovarian lesions and from adjacent normaltissue (Zachos et al, 1996b). Increased binding of theglucocorticoid receptor in H-ras GRE was observed inmore than 90% of endometrial and in all ovarian tumorstested, compared to the adjacent normal tissue. Moreover,elevated binding of the estrogen receptor in H-ras EREwas found in all pairs of ovarian tumor/ normal tissueexamined (Zachos et al, 1996b). Thus, it was proposedthat H-ras is implicated in human gynecological lesionsthrough elevated steroid receptor binding.

In addition, previous data showed cooperation of the H-ras with steroids in cell transformation (Kumar et al, 1990)and overexpression of the Ras p21 in ovarian tumors,compared to normal or benign tumor tissue (Katsaros et al,1995). By combining these data, it was proposed that thehigh levels of steroid and sex hormones in human genitaltract may result in increased amounts of ligand activatedsteroid receptors. Furthermore, receptors bind to the H-rasDNA and induce elevated transcription of the H-rasoncogene, resulting in an increased oncogenic potential.

Thus, endometrial and ovarian epithelial cells may have apredisposition to develop neoplastic abnormalities inaddition to a second tumorigenic event, e.g. viral infection,loss of an onco-suppressor gene, or mutational activationof a proto-oncogene (Zachos et al, 1996a).

C. The p53 element

One of the major roles of wild-type P53 onco-suppressor protein is to trigger cell cycle arrest orapoptosis in response to DNA damage by acting as asequence specific transcription factor that binds to DNAand activates genes involved in the control of the cellcycle, including p21, gadd45, bax, mdm2 and PCNA(Zambetti et al, 1993, Hainaut, 1995). The mutant formsof P53 promote tumorigenesis by dominant-negativeinhibition of wild-type P53 through cross-oligomerization.Moreover, P53 mutants were proved to exert oncogenicfunctions of their own (Dittmer et al, 1993).

1 . Regulation of the c-H-ras by the P53tumor-suppressor protein

H-ras contains within its first intron sequences thatpartially match the p53 consensus binding site (F i g . 1 ).Using gel retardation assays it was shown that wild-typeP53, as well as the "hot spot" mutant His 273 recognizethe H-ras element with high affinity (Spandidos et al,1995, Zoumpourlis et al, 1995). Furthermore, the H-raselement functioned as a P53-dependent transcriptionalenhancer in the context of a reporter plasmid, thussuggesting that P53 is a physiological regulator of H-rasexpression (Spandidos et al, 1995).

Activation of H-ras expression by P53 may seem aparadox, since p53 is a tumor suppressor and H-ras aproto-oncogene. However, precedence has already beenestablished for activation of proto-oncogenes by P53.Wild-type P53 induces expression of mdm-2, whoseprotein product inhibits the tumor suppressor activities ofP53 and Rb (Xiao et al, 1995). Interestingly, there arecertain similarities in the organization of the p53 elementsof the H-ras and mdm-2 genes, which are not shared withthe p53 elements of other genes targeted by P53. In H-rasthere are three half sites: two of them are contiguous,while the third is 8 nucleotides upstream. In mdm-2 thereare again three half-sites: two are contiguous, while thethird one is located 28 nucleotides upstream (Wu et al,1993). In contrast to H-ras and mdm-2, the elements of theother genes regulated by P53 are contiguous. Theorganization of the half-sites affects the ability of the P53protein to recognize these elements. Wild-type P53reversibly switches between two conformations: the"inactive" T state, with dihedral symmetry, which canrecognize only non-contiguous half-sites and the "active"R state, which can recognize even contiguous half-sites(Waterman et al, 1995). Thus, it is suggested that H-rasand mdm-2 genes allow regulation by even the "inactive"T state of P53 protein.


633

The H-ras p53 element is located within the firstintron. Interestingly, the p53 element of the mdm-2 isalso located in the first intron of the gene (Wu et al,1993). The significance of this is not understood at thistime. The mdm-2 has an internal promoter in the firstintron. Transcripts initiating at both promoters contain theentire protein sequence, however they differ in theefficiency with which translation is initiated in codon 1.Thus, transcripts that include the first exon mostly expressan N-terminally truncated Mdm-2 protein, whereastranscripts from the internal promoter express a full-lengthprotein. P53 induces expression only from the internalpromoter and only the full-length form can associate withP53, closing the autoregulatory feedback loop (Barak et al,1994). It remains to be determined whether P53 inducesexpression of transcripts initiating at the first intron of H-ras, as well as the biological significance of suchtranscripts.

The p53 tumor suppressor may therefore exert itscellular effects by coordinate activation of genes thatsuppress and induce cell proliferation.

2. Altered binding of p53 protein to the H-ras element in human tumors

Mutation and overexpression of the p53 tumorsuppressor is a common event in human endometrial andovarian cancer (Berchuck et al, 1994) and is associated withpoor prognosis (Levesque et al, 1995, Kihana et al, 1995).Moreover, aberrant regulation of the H-ras gene expressionalso participates in the development of humangynecological lesions (Zachos and Spandidos, 1997).

Using nuclear extracts from human endometrial andovarian tumors and from the adjacent normal tissue in gelretardation assays, we examined the binding levels of theP53 protein to the H-ras element (our unpublished results).Elevated P53 binding in the tumor tissue was found in5/11 (45%) of endometrial and in 2/5 (40%) of ovariancases. Loss of P53 DNA binding activity was observed in3/11 (27%) of endometrial and in 1/5 (20%) of ovariantumors. In the remaining 3/11 (27%) of endometrial and in2/5 (40%) of ovarian pairs tested, no alteration in the P53binding levels was observed. In order to interpret theresults, all pairs were subsequently tested for mutations inexons 4-9 of the p53 gene using PCR-SSCP analysis. Nomutation was observed in any case showing elevated DNAbinding activity, thus implying for overexpression of thewild-type p53 gene in these tumors. In addition, no p53mutational alteration was observed in the cases showingsimilar DNA binding levels in tumor versus the adjacentnormal tissue. However, a mutated allele was detected inall four endometrial and ovarian cases showing loss of P53DNA binding activity. We therefore suggest that P53could directly modulate the H-ras oncogenic potential inhuman endometrial and ovarian lesions, depending on theexpressed levels of P53 and the status of the protein (wild-

type or mutated forms), thus providing additional evidencefor the role of H-ras in human carcinogenesis.

IV. The role of the VTR

Variable tandem repeats (VTRs, minisatellites) arehighly polymorphic structures characterized by the tandemrepetition of short (up to 100 bp) sequence motifs. Severalobservations on tandemly-repetitive elements within viralgenomes (Yates et al, 1984) have led to the speculationthat some human minisatellites might serve as regulatoryregions for cellular transcription or replication.

A. The H-ras minisatellite sequence astranscriptional enhancer

The human H-ras gene contains a VTR region located1 kb upstream the polyadenylation signal (Fig . 1 ). Itconsists of 30 to 100 copies of a 28 bp consensus repeat.Four common alleles and more than 25 rare alleles havebeen described (Krontiris et al, 1993). It was shown thatthe H-ras VTR sequences possess endogenous enhanceractivity, independently from orientation, however thisactivity is promoter specific (Spandidos and Holmes,1987, Cohen et al, 1987). The 28 bp repeat unit of theminisatellite binds four proteins (p45, p50, p72 and p85)which are members of the rel/ NF-!B family oftranscriptional regulatory factors (Trepicchio and Krontiris,1992).

B. VTR rare alleles of the H-ras andovarian cancer risk

Women who carry a mutation in the BRCA1 genehave an 80% risk of breast cancer and a 40% risk ofovarian cancer by the age of 70 (Easton et al, 1995). Thevarying penetrance of BRCA1 suggests a role for othergenetic and epigenetic factors in tumorigenesis of theseindividuals. H-ras was the first example of a modifyinggene on the penetrance of an inherited cancer syndrome.Rare alleles of the H-ras VTR locus duplicate themagnitude of ovarian cancer risk for BRCA1 carriers, butnot the risk for developing breast cancer (Phelan et al,1996). It was suggested that H-ras VTR alleles showdifferences in modulating gene transcription, that H-rasVTR alleles are in linkage disequilibrium with other genesimportant in tumorigenesis, or that rare alleles provide amarker for genomic instability (Phelan et al, 1996).

V. The role of the DNA methylationstatus

DNA methylation is essential for embryonicdevelopment and alterations in the DNA methylationstatus are common in cancer cells. CpG sites invertebrates are either clustered in 0.5-2 kb regions called

Table I. ras gene overexpression in human tumors and correlation with clinical parameters.


634

Tumor type Frequency(%)

ras gene Stage intumorigenesis

Prognosis of thedisease

Neuroblastoma 50-80 H-, ras early favourableHead and neck 54 H-, K- early favourableEsophagus 40 H- unknown unknownLarynx 57-86 H-, K-, N- unknown unknownThyroid 85 ras early unknownLung 64-85 ras late poorLiver 60 ras unknown unknownSmall intestine 70 ras unknown unknownStomach 35 K-, ras late poorPancreas 42 ras unknown unknownColon 31 H-, K-, ras early poorBreast 65-70 ras unknown unknownBladder 39-58 H-, K-, N- early poorEndometrium 18-95 ras late unknownOvary 45 ras late poorLeukemias 39-67 H-, K-, N- unknown unknown

CpG islands, or are dispersed, in which case they aremostly methylated and constitute mutational hotspots(Jones, 1996). The CpG islands are associated with genepromoters (e.g. H-ras) or coding regions (e.g. p16) and areunmethylated in autosomal genes. 5' Methyl-cytosine canaffect transcription by altering the DNA binding activitiesof transcription factors. This could be done either directly,for example binding of trans-acting proteins at AP-2 sitesis inhibited (Comb and Goodman, 1990), or indirectly, byenhanced binding of methylated DNA binding protein(MDBP) which stereochemically inhibits DNA binding oftranscription factors (Boyes and Bird, 1991).

The promoter region of the H-ras gene ishypomethylated in human tumors compared to thecorresponding normal tissue (Feinberg and Vogelstein,1983). Furthermore, methylation of cis-elements decreasesH-ras promoter activity in vitro (Rachal et al, 1989) andinhibits the transforming activity of the oncogene (Borelloet al, 1987). It is therefore suggested that epigenetic andreversible mechanisms, like DNA methylation, canregulate the expression of proto-oncogenes and silencegenetically activated human oncogenes.

VI. Differential expression of the H-rasgene is controlled by alternative splicing

A proportion of H-ras pre-mRNA is spliced toincorporate an alternative exon, termed IDX (intron Dexon), which contains an in-frame translationaltermination codon that prevents expression of the geneticinformation specified by the exon IV as shown in F i g . 1(Cohen et al, 1989). The abundance of these transcripts islow, apparently due to message instability or defective

processing. The predicted product of the alternatetranscript (p19) lacks transforming potential, since the Cterminal sequence of p21 that is necessary for attachmentof the protein to the inner site of the cellular membrane isabsent. It is suggested that alternative splicing patternsoperate to suppress the H-ras p21 expression. Thisnegative control is abolished by mutations that interferewith this process.

VII. Overexpression of ras genes inhuman tumors

Overexpression of ras genes is a common event inhuman tumors (reviewed by Zachos and Spandidos,1997). Table I summarizes the experimental results byindicating the tumor type where elevated expression of rasgenes was observed, the frequency of the overexpression,the activated member of the ras gene family, the stage intumorigenesis and correlation of altered ras geneexpression with prognosis of the disease. Where noparticular ras gene is mentioned (referred as: ras), nodiscrimination between the ras family members wasperformed, nor was their status (mutated or wild-typealleles) defined.

Elevated ras gene expression was observed in humanneuroblastomas (Spandidos et al, 1992), head and necktumors (Field, 1991), esophageal (Abdelatif et al, 1991),laryngeal (Kiaris et al, 1995), thyroid (Papadimitriou etal, 1988), lung (Miyamoto et al, 1991), liver (Tiniakos etal, 1989), small intestine (Spandidos et al, 1993),stomach (Motojima et al, 1994), pancreatic (Song et al,1996), colorectal (Spandidos and Kerr, 1994), breast (Datiet al, 1991), bladder (Ting-jie et al, 1991), endometrial


635

Figure 2 . A synopsis of molecular therapeutic strategies developed against activated ras oncogenes. Strategies include antisenseoligonucleotides, ribozymes, farnesyltransferase inhibitors and activation of T-cell response after mutant Ras peptidevaccination of patients.

(Long et al, 1988) and ovarian tumors (Scambia et al,1993) and in leukemias (Gougopoulou et al, 1996). Thefrequency of elevated expression of the ras family ofgenes varies from 30% in endometrial and colorectaltumors, to 85-90% of cases in endometrial, lung andlaryngeal tumors. In a number of tumors includingneuroblastomas, head and neck, thyroid, colorectal andbladder cancers, ras overexpression is considered to be anearly genetic event. However, in lung, stomach,endometrial and ovarian lesions, overexpression of rasgenes appears in a later stage in tumorigenesis. Elevatedexpression of the Ras p21 protein is correlated with poorprognosis in lung, stomach, colorectal, bladder andovarian lesions, whereas it is a favourable marker forneuroblastomas and head and neck tumors.

VIII. Molecular therapeutic strategies

The development of effective molecular strategies fortherapy is the aim of tumor biology. Current therapeuticstrategies include ribozymes against mutant ras geneproducts, antisense strategies, inhibitors of Ras proteinpost-translational modifications and Ras peptidevaccination (Fig . 2 ).

A. Ribozymes

Molecular biology applies the site-specific RNAseproperties of ribozymes to gene therapy for cancer. Theanti-ras ribozymes are designed to cleave only activated rasRNA (F i g . 2 ). To develop this strategy into practicalmeans, methods must be developed to accomplish highefficiency delivery of the ribozyme to target neoplastictissue. An adenoviral-mediated delivery was designed(Feng et al, 1995). Using anti-Ras ribozymes, it waspossible to reverse the neoplastic phenotype in mutant H-ras expressing tumor cells with high efficiency (Kashani-Sabet et al, 1994).

B. Antisense strategies

The antisense strategy involves reduction of aparticular gene expression by introduction of a cDNAsegment in antisense orientation, in order to bind thetarget mRNA and prevent its translation (Stein andCheng, 1993) (Fig . 2 ). Critical to the success of such anantisense agent is its ability to enter living cells, tospecifically bind the target mRNA and induce RNAse-Hcleavage of the target RNA. Activated ras genes, bymutation or overexpression, are a common target of these


636

therapeutic trials in cell-free and in vitro systems (Moniaet al, 1992, Schwab et al, 1994).

C. Inhibitors of Ras post-translationalmodifications

Farnesylation of the CAAX motif of Ras protein isessential for the subcellular localization of Ras to theplasma membrane and is critical to Ras cell-transformingactivity. Inhibitors of farnesyltransferase have beendeveloped as potential cancer therapeutic agents (Gibbs etal, 1994) (F i g . 2 ). Requirements for Ras farnesylationinhibitors include specificity for farnesyl proteintransferase, ability to inhibit post-translationalmodifications of the mutant ras specifically, high potency,activity in vivo and lack of toxicity (Kelloff et al, 1997).

D. Ras peptide vaccination

Ras peptide vaccination is a recent, developingmolecular strategy for cancer therapy. Mutant Raspeptides are candidate vaccines for specific immunotherapyin cancer patients. An amount of mutant Ras p21 isdegraded in the cytoplasm and fragments are attached withclass I MHC glycoproteins, in the outer surface of the cellmembrane (F i g . 2 ). When vaccinated with a syntheticRas peptide representing the ras mutation in tumor cells,a transient Ras-specific T-cell response is induced, towardsthe fragments of mutant Ras protein associated withMHC molecules. Ras peptide vaccination was proved tobe effective in 40% of patients with pancreatic cancer(Gjertsen et al, 1995, 1996). However, peptidevaccination of patients, like all other gene therapystrategies previously mentioned, requires considerabledevelopment before useful anti-cancer agents can emerge.

IX. Concluding remarks

Regulation of the c-H-ras1 gene expression is acomplicated procedure, including regulation by a varietyof regulatory proteins (transcription factors, hormonereceptors, tumor-suppressor proteins), alternativemechanisms (methylation, splicing) and by sequenceslocated in the promoter region, in introns and downstreamof the coding sequence. Understanding the molecularmechanisms of the expression of ras genes is of greatsignificance for studying human tumorigenic events anddeveloping effective strategies for gene therapy.

References

Abdelatif OMA, Chandler FW, Mills LR, McGuire BS,Pantazis CG and Barret JM (1 9 9 1 ) Differentialexpression of c-myc and H-ras oncogenes in Barret'sepithelium-a study using colorimetric in situhybridazation. Arch Pathol Lab Med 115, 880-885.

Avruch J, Zhang X and Kyriakis JM (1 9 9 4 ) Raf meets Ras:completing the framework of a signal transductionpathway. TIBS 19, 279-283.

Barak Y, Gottlieb E, Juven-Gershon T and Oren M (1 9 9 4 )Regulation of mdm2 expression by p53: alternativepromoters produce transcripts with nonidenticaltranslation potential. Genes Dev 8, 1739-1749.

Barbacid M (1 9 8 7 ) ras genes. Ann Rev Biochem 56,779-827.

Beato M, Herrlich P and Schutz G (1 9 9 5 ) Steroid hormonereceptors: many actors in search of a plot. Cel l 83, 851-857.

Beato M (1 9 8 9 ) Gene regulation by steroid hormones. Cel l56, 335-344.

Berchuck A, Kohler MF, Marks JR, Wiseman R, Boyd J andBast RC (1 9 9 4 ) The p53 tumor suppressor gene isfrequently altered in gynecologic cancers. Am J ObstetGyneco l 170, 246-252.

Berg JM (1 9 9 2 ) Sp1 and the subfamily of zinc fingerproteins with guanine-rich binding sites. Proc NatlAcad Sci USA 89, 11109-11110.

Bishop JM (1 9 9 1 ) Molecular themes in oncogenesis. Cel l64, 235-248.

Borello MG, Pierotti MA, Bongarzone I, Donghi R,Modellini P and Della Porta G (1 9 8 7 ) DNA methylationaffecting the transforming activity of the human Ha-rasoncogene. Cancer Res 47 , 75-79.

Boyes J and Bird A (1 9 9 1 ) DNA methylation inhibitstranscription indirectly via a methyl-CpG bindingprotein. Cel l 64, 1123-1134.

Boulikas T (1 9 9 4 ) A compilation and classification of DNAbinding sites for protein transcription factors fromvertebrates. Cri t Rev Eukaryot Gene Exp 4, 117-321.

Cohen JB, Broz SD and Levinson AD (1 9 8 9 ) Expression ofthe H-ras proto-oncogene is controlled by alternativesplicing. Cel l 58, 461-472.

Cohen JB, Maureen VW and Levinson AD (1 9 8 7 ) Arepetitive sequence element 3' of the human c-Ha-ras1gene has enhancer activity. J C e l l P h y s i o l (Suppl) 5:75-81.

Comb M and Goodman HM (1 9 9 0 ) CpG methylationinhibits proenkephalin gene expression and binding ofthe transcription factor AP-2. Nucle ic Acids Res 18,3975-3982.

Dati C, Muraca R, Tazartes O, Antoniotti S, Perrotean I, GiaiM, Cortese P, Sismondi P, Saglio G and DeBortoli M(1 9 9 1 ) c-ErbB-2 and Ras expression levels in breastcancer are correlated and show a cooperative associationwith unfavorable clinical outcome. Int J Cancer 47:833-838.

Dittmer D, Pati S, Zambetti G, Chu S, Teresky AK, Moore M,Finlay C and Levine AJ (1 9 9 3 ) Gain of functionmutations in p53. Nat Genet 4, 42-46.

Easton D, Ford D and Bishop D (1 9 9 5 ) Breast and ovariancancer incidence in BRCA1 mutation carriers. Am J HumGenet 56, 265-271.


637

Feinberg AP and Vogelstein B (1 9 8 3 ) Hypomethylation ofras oncogenes in primary human cancers. BiochemBiophys Res Commun 198, 47-54.

Feng M, Cabrera G, Deshane J, Scanlon KJ and Curiel DT(1 9 9 5 ) Neoplastic reversion accomplished by highefficiency adenoviral-mediated delivery of an anti-rasribozyme. Cancer Res 55, 2024-2028.

Field JK (1 9 9 1 ) Prognostic implications of ras oncogeneexpression in head and neck squamous cell carcinoma. In:DA Spandidos, ed. The superfamily of ras related genes:New York, Plenum Press , 213-226.

Gibbs JB, Oliff A and Kohl NE (1 9 9 4 ) Farnesyltransferaseinhibitors: Ras research yields a potential cancertherapeutic. Cel l 77, 175-178.

Gjertsen MK, Bakka A, Breivik J, Saeterdal I, Gadde-Dahl T3rd, Stokke KT, Salheim BG, Egge TS, Soreide O,Thorsby E and Gaudernack G (1 9 9 6 )Ex vivo ras peptidevaccination in patients with advanced pancreatic cancer:results of a phase I/II study. Int J Cancer 65, 450-453.

Gjertsen MK, Bakka A, Breivik J, Saeterdal I, Solheim BG,Soreide O, Thorsby E and Gaudernack G (1 9 9 5 )Vaccination with mutant ras peptides and induction of T-cell responsiveness in pancreatic carcinoma patientscarrying the corresponding RAS mutation. Lancet 346,1399-1400.

Gougopoulou DM, Kiaris H, Ergazaki M, AnagnostopoulosNI, Grigoraki Y and Spandidos DA (1 9 9 6 ) Mutations andexpression of the ras family genes in leukemias. StemC e l l s 14, 725-729.

Hainaut P (1 9 9 5 ) The tumor suppressor protein p53: areceptor to genotoxic stress that controls cell growth andsurvival. Curr Opin Oncol 7, 76-82.

Hashimoto-Gotoh T, Kikuno R, Takahashi M and Honkawa H(1 9 8 8 ) Possible role of the first intron of c-H-ras in geneexpression: anti-cancer elements in oncogenes.Anticancer Res 8, 851-860.

Hashimoto-Gotoh T, Takahashi M, Kikuno R, Ishihara H andTezuka K (1 9 9 2 ) Unusual sequence conservation inintron-1 of c-H-ras oncogene and its effect on p21 proteinsynthesis. In: Spandidos DA, ed. Current Perspectives ofMolecular Cellular Oncology. London: JAI Press , 211-230.

Honkawa H, Masahashi W, Hashimoto S, and Hashimoto-Gotoh T (1 9 8 7 ) Identification of the principal promotersequence of the c-Ha-ras transforming oncogene: deletionanalysis of the 5' flanking region by focus formationassay. M o l C e l l B i o l 7, 2933-2940.

Hyder SM, Stancel GM and Loose-Mitchell DS (1 9 9 4 )Steroid hormone-induced expression of oncogeneencoded nuclear proteins. Cri t Rev Eucaryot GeneExpr 4, 55-116.

Ishii S, Kadonaga JT, Tjian R, Brandy JN, Merlino GT andPastan I (1 9 8 6 ) Binding of the Sp1 transcription factorby the human Harvey ras1 proto-oncogene promoter.Sc ience 232, 1410-1413.

Ishii S, Merlino GT and Pastan I (1 9 8 5 ) Promoter region ofthe human Harvey ras proto-oncogene: Similarity to theEGF receptor proto-oncogene promoter. Sc ience 230,1378-1381.

Jones PA (1 9 9 6 ) DNA methylation errors and cancer.Cancer Res 56, 2463-2466.

Kashani-Sabet M, Funato T, Florenes VA, Fostad O andScanlon KJ (1994) Suppression of the neoplasticphenotype in vivo by an anti-ras ribozyme. Cancer Res54, 900-902.

Katsaros D, Theillet C, Zola P, louason G, Sanfilippo B,Isaia E, Arisio R, Giardina G and Sismondi P (1 9 9 5 )Concurrent abnormal expression of erbB-2, myc and rasgenes is associated with poor outcome of ovarian cancerpatients. Anticancer Res 15, 1501-1510.

Kelloff GJ, Lubet RA, Fay JR, Steele VE, Boone CW, CrowellJA and Sigman CC (1 9 9 7 ) Farnesyl protein transferaseinhibitors as potential cancer chemopreventives. CancerEpidemiol Biomarkers Prev 6, 267-282.

Kiaris H and Spandidos DA (1 9 9 5 ) Analysis of H-ras, K-rasand N-ras genes for expression, mutation andamplification in laryngeal tumours. Int J Oncol 7, 75-80.

Kiaris H and Spandidos DA (1 9 9 5 ) Mutations of ras genes inhuman tumors (Review). Int J Oncol 7, 413-421.

Kihana T, Hamada K, Inoue Y, Yano N, Iketani H, Murao S,Ukita M and Matsuura S (1 9 9 5 ) Mutation and allelic lossof the p53 gene in endometrial carcinoma. Cancer 76,72-78.

Krontiris TG, Devlin B, Karp DD, Robert NJ and Risch N(1 9 9 3 ) An association between the risk of cancer andmutations in the HRAS 1 minisatellite locus. N E n g JMed 329, 517-523.

Kumar R, Sukumar S and Barbacid M (1 9 9 0 ) Activation ofras oncogenes preceding the onset of neoplasia.Sc ience 248, 1101-1104.

Lee W and Keller EB (1 9 9 1 ) Regulatory elements mediatingtranscription of the human Ha-ras gene. J M o l B i o l220, 599-611.

Levesque MA, Katsaros D, Yu H, Zola P, Sismondi P, GiardinaG and Diamantis EP (1 9 9 5 ) Mutant p53 overexpressionis associated with poor outcome in patients with well ormoderately differentiated ovarian carcinoma. Cancer 75,1327-1338.

Lewin B (1 9 9 1 ) Oncogenic conversion by regulatorychanges in transcription factors. Cel l 64, 303-312.

Long CA, O'Brien TJ, Sanders MM, Bard DS and Quirk JG Jr(1 9 8 8 ) ras oncogene is expressed in adenocarcinoma ofthe endometrium. Am J Obste t Gynecol 159: 1512-1516.

Lowndes NF, Paul J, Wu J and Allan M (1 9 8 9 ) c-Ha-ras genebidirectional promoter expressed in vitro. Mol Ce l lB i o l 9: 3758-3770.

Lowy DR and Willumsen BM (1 9 9 3 ) Function andregulation of ras. Annu Rev Biochem 62, 851-891.

Macara IG, Lounsbury KM, Richards SA, McKiernan C andBar-Sagi D (1 9 9 6 ) The Ras superfamily of GTPases.FASEB J 10, 625-630.

Marshall MS (1 9 9 5 ) Ras target proteins in eucaryotic cells.FASEB J 9, 1311-1318.

McGrath JP, Capon DJ, Smith DH, Chen EY, Seeburg PH,Goeddel PV and Levinson AD (1 9 8 3 ) Structure and


638

organization of the human Ki-ras proto-oncogene and arelated processed pseudogene. Nature 304, 501-506.

Mermod N, O'Neill EA, Kelly TJ and Tjian R (1 9 8 9 ) Theproline-rich transcriptional activator of CTF/NF-I isdinstict from the replication and DNA binding domain.Cel l 58, 741-753.

Miyamoto H, Harada M, Isobe H, Akita HD, Haneda H,Yamaguchi E, Kuzumaki N and Kawakami Y (1 9 9 1 )Prognostic value of nuclear DNA content and expressionof the ras oncogene product in lung cancer. Cancer Res51, 6346-6350.

Monia BP, Johnston JF, Eckers DJ, Zounes MA, Lima WF andFreier SM (1 9 9 2 ) Selective inhibition of mutant Ha-rasmRNA expression by antisense oligonucleotides. J B i o lChem 267, 19954-19962.

Motojima K, Furui J, Kohara N, Izawa K, Kanematsu T andShiku H (1 9 9 4 ) Expression of Kirsten-ras p21 in gastric-cancer correlates with tumor progression and isprognostic. Diagn Mol Pathol 3, 184-191.

Nagase T, Ueno Y and Ishii S (1 9 9 0 ) Transcriptional controlof the human Harvey ras proto-oncogene: role of multipleelements in the promoter region. Gene 94, 249-253.

Papadimitriou K, Yiagnisis M, Tolis G and Spandidos DA(1 9 8 8 ) Immunohistochemical analysis of the rasoncogene protein in human thyroid neoplasms.Anticancer Res 8, 1223-1228.

Phelan CM, Rebbeck TR, Weber BL, Devilee P, RuttledgeMH, Lynch HT, Lenoir GM, Stratton MR, Easton DF,Ponder BA, Cannon-Albright C, Larsson C, Goldgar DEand Narod SA (1 9 9 6 ) Ovarian cancer risk in BRCA1carriers is modified by the HRAS1 variable number oftandem repeat (VNTR) locus. Nat Genet 12, 309-311.

Pintzas A and Spandidos DA (1 9 9 1 ) Sp1 specific bindingsites within the human H-ras promoter: potential role ofthe 6 bp deletion sequence in the T24 H-ras1 gene.Anticancer Res 11: 2067-2070.

Pulciani S, Santos E, Long LK, Sorrentino V and Barbacid M(1 9 8 5 )ras gene amplification and malignanttransformation. Mol Ce l l B io l 5, 2836-2841.

Rachal MJ, Yoo H, Becker FF and Lapeyre J-N (1 9 8 9 ) Invitro DNA cytosine methylation of cis-regulatoryelements modulates c-Ha-ras promoter activity in vivo.Nucleic Acids Res 17, 5135-5147.

Scambia G, Catozzi L, Panici PB, Ferrandina G, Coronetta F,Barazzi R, Baiocchi G, Uccelli L, Piffanelli A andMancuso S (1 9 9 3 ) Expression of ras oncogene p21protein in normal and neoplastic ovarian tissues:correlation with histopathologic features and receptorsfor estrogen, progesterone and epidermal growth factor.Am J Obstet Gynecol 168, 71-78.

Schwab G, Chavany C, Duroux I, Goubin G, Lebeau J, HeleneC and Saison-Bechmoaras T (1 9 9 4 ) Antisenseoligonucleotides adsorbed to polyalkylcyanoacrylatenanoparticles specifically inhibit mutated Ha-ras-mediated cell proliferation and tumorigenicity in nudemice. Proc Natl Acad Sci USA 91, 10460-10464.

Song MM, Nio Y, Sato Y, Tamura K and Furuse K (1 9 9 6 )Clinicopathological significance of Ki-ras point mutationand p21 expression in benign and malignant exocrinetumors of the human pancreas. Int J Pancreatol 20, 85-93.

Spandidos DA, Arvanitis D and Field JK (1 9 9 2 ) Ras p21expression in neuroblastomas and ganglio-neuroblastomas: correlation with patients' prognosis.Int J Oncol 1, 53-58.

Spandidos DA, Zoumpourlis V, Zachos G, Toas SH andHalazonetis TD (1 9 9 5 ) Specific recognition of atranscriptional element within the human H-ras proto-oncogene by the p53 tumor suppressor. Int J Oncol 7,1029-1034.

Spandidos DA and Holmes L (1 9 8 7 ) Transcriptionalenhancer activity in the variable tandem repeat DNAsequence downstream of the human Ha-ras1 gene. FEBSLett 218, 41-46.

Spandidos DA and Kerr IB (1 9 9 4 ) Elevated expression of thehuman ras oncogene family in premalignant andmalignant tumours of the colorectum. Br J Cancer 49,681-688.

Spandidos DA, Liloglou T, Arvanitis D and GourtsoyiannisNC (1 9 9 3 ) ras gene activation in human small intestinaltumors. Int J Oncol 2, 513-518.

Spandidos DA, Nichols RAB, Wilkie NM and Pintzas A(1 9 8 8 ) Phorbol ester-responsive H-ras1 gene promotercontains multiple TPA-inducible/ AP-1-bindingconsensus sequence elements. FEBS Lett 240, 191-195.

Spandidos DA and Pintzas A (1 9 8 8 ) Differential potency andtrans-activation of normal and mutant T24 human H-ras1gene promoters. FEBS Lett 232, 269-274.

Spandidos DA (1 9 9 1 ) The superfamily of ras-related genes.New York: Plenum Press , 1-338.

Stein CA and Cheng YC (1 9 9 3 ) Antisense oligonucleotidesas therapeutic agents - is the bullet really magical?Sc ience 261, 1004-1012.

Ting-jie M, Ze W and Nianli S (1 9 9 1 ) Correlation betweenthe expression of the p21 ras oncogene product and thebiological behavior of bladder tumors. Eur Urol 20,307-310.

Tiniakos D, Spandidos DA, Kakkanas A, Pintzas A, Pollice Land Tiniakos G (1 9 8 9 ) Expression of ras and myconcogenes in human hepatocellular carcinoma and non-neoplastic liver tissues. Anticancer Res 9, 715-722.

Trepicchio WL and Krontiris TG (1 9 9 2 ) Members of the rel/

NF-!B family of transcriptional regulatory proteins bindthe HRAS1 minisatellite DNA sequence. Nucle ic AcidsRes 20, 2427-2434.

Trimble WS and Hozumi N (1 9 8 7 ) Deletion analysis of the c-Ha-ras oncogene promoter. FEBS Lett 219, 70-74.

Yang-Yen H-F, Chambard J-C, Sun Y-L, Smeal T, Schmidt TJ,Drouin J and Karin M (1 9 9 0 ) Transcriptional interferencebetween c-Jun and the glucocorticoid receptor: mutualinhibition of DNA binding due to direct protein-proteininteraction. Cel l 62, 1205-1215.

Yates J, Warren, N, Reisman D and Sugden B (1 9 8 4 ) A cis-acting element from the Epstein-Barr viral genome thatpermits stable replication at recombinant plasmids inlatently infected cells. Proc Nat l Acad Sc i USA 81,3806-3810.

Waterman JLF, Shenk JL and Halazonetis TD (1 9 9 5 ) Thedihedral symmetry of the p53 tetramerization domain


639

mandates a conformational switch upon DNA binding.EMBO J 14, 512-519.

Westaway D, Parkoff J, Moscovici C and Varmus HE (1 9 8 6 )Identification of a provirally activated c-Ha-ras oncogenein an avian nephroblastoma via a novel procedure: cDNAcloning of a chimaeric viral-host transcript. EMBO J 5,301-309.

Willumsen BM and Christensen A (1 9 8 4 ) The p21 ras C-terminus is required for transformation and membraneassociation. Nature 310, 583-586.

Wu X, Bayle JH, Olson D and Levine AJ (1 9 9 3 ) The p53-mdm2 autoregulatory feedback loop. Genes Dev 7,1126-1132.

Zachos G and Spandidos DA (1 9 9 7 ) Expression of ras proto-oncogenes: regulation and implications in thedevelopment of human tumors. Crit Rev OncolHematol 26, 65-75.

Zachos G, Zoumpourlis V, Sekeris CE and Spandidos DA(1 9 9 5 ) Binding of the glucocorticoid and estrogenreceptors to the human H-ras oncogene sequences. Int JOncol 6, 595-600.

Zachos G, Varras M, Koffa M, Ergazaki M and Spandidos DA(1 9 9 6 a ) The association of the H-ras oncogene andsteroid hormone receptors in gynecological cancer. JExp Ther Oncol 1, 335-341.

Zachos G, Varras M, Koffa M, Ergazaki M and Spandidos DA(1 9 9 6 b ) Glucocorticoid and estrogen receptors haveelevated activity in human endometrial and ovarian tumorsas compared to the adjacent normal tissues and recognizesequence elements of the H-ras proto-oncogene. Jpn JCancer Res 87, 916-922.

Zambetti GP and Levine AJ (1 9 9 3 ) A comparison of thebiological activities of wild-type and mutant p53.FASEB J 7, 855-865.

Zoumpourlis V, Zachos G, Halazonetis TD, Ergazaki M andSpandidos DA (1 9 9 5 ) Binding of wild-type and mutantforms of P53 protein from human tumors to a specificDNA sequence of the first intron of the H-ras oncogene.Int J Oncol 7, 1035-1041.


641


Periodicity of DNA bend sites in eukaryotic genomes

Ryoiti Kiyama

Institute of Molecular and Cellular Biosciences, University of Tokyo, Yayoi, Bunkyo-ku, Tokyo 113, and NationalInstitute of Bioscience and Human-Technology, Tsukuba-shi, Ibaraki 305, Japan__________________________________________________________________________________________________Correspondence to: Ryoiti Kiyama, Tel: 81-3-3812-2111, ext. 7835, Fax: 81-3-3818-9437, E-mail: [email protected]

Summary

We found that DNA bend sites are distributed regularly and periodically in the genomic DNA of eukaryotes.Their locations were conserved during molecular evolution in otherwise unstable intergenic regions of genomicDNA at intervals of approximately 700 bp, which corresponds to a length of four nucleosomes, suggesting theiractive role in chromatin organization. By further examination of these sites with respect to chromatin structure, weobtained evidence that these sites may act as signals for nucleosome phasing. Here, we summarize our resultsregarding periodic bent DNA in the human -globin, c-myc, and immunoglobulin heavy chain loci and discusstheir biological functions.

I. Bent DNA in biological reactions

Genomic DNA is a source of genetic and functionalinformation in the form of nucleotide sequences. DNA hasa relatively simple composition of purine or pyrimidinebases attached to a common phosphate backbone whichwould not give rise much local structural variation.However, recent studies have revealed that non-B DNA or"unusual" DNA structures are actively involved inbiologically important reactions as functional elements(Crothers et al. , 1990). For example, Z-DNA is known toactivate transcription and recombination presumably byexposing bases on the outside of the phosphate backbone,thereby increasing the chance of interaction with proteinsor other bases (Rich et al., 1984; Blaho and Wells, 1989).Other structures such as triplex DNA and unisomorphicDNA have been discussed in studies of transcriptionalactivation and recombination mechanisms (Crothers et al. ,1990; Soyfer and Potaman, 1996). Although theirmechanisms of action are quite different, these non-BDNA structures seem to act as signals for recognition byprotein factors in a way different from searchingnucleotide bases.

Bent DNA was first discovered as anomalousmigration of DNA fragments in gels, and has beenextensively studied due to its potential involvement as atranscriptional modulator (Travers, 1989; Hagerman,1990; Crothers et al., 1990; Trifonov, 1991; Ioshikhes etal., 1996; Werner et al., 1996). Such structures also playimportant roles in activation of recombination (Nash,1990). Binding of proteins to DNA can cause DNAbending, which would further enhance the recognition byother proteins of the site of the protein-DNA complex(Khan and Crothers, 1992). Therefore, DNA bending

formed by the intrinsic nature of the DNA or by proteinbinding, as well as sequence information, would be a goodsignal for structural recognition. Note that the non-B DNAstructures themselves are the result of nucleotide sequenceinformation, although the sequence-structure relationshipis not simple.

II. Assays for DNA bend sites

The presence and the location of DNA bend sites canbe analyzed by several assays. Among these, the circularpermutation assay (Wu and Crothers, 1984) has beencommonly utilized for mapping the bend sites in DNAfragments of several hundred bp to 1 kb in length. Theassay procedure is schematically illustrated in Figure 1.Plasmids containing the tandem dimers of the fragment ofinterest are cloned, and after digestion of the plasmidDNA with the restriction enzymes that cut the fragmentonce the plasmid DNA samples are resolved byelectrophoresis. We routinely use 8% polyacrylamide(mono: bis = 29: 1) gels, which can resolve bands up to 1kb in size (Wada-Kiyama and Kiyama, 1994).Electrophoresis should be performed at 4˚C twice or threetimes overnight to obtain better resolution of the bands.Cloning of the tandem dimers can be performed by directcloning of two identical fragments into the multiplecloning site of the vector, or cloning them into twodifferent sites one after another. Most of the clones couldbe obtained by the former method under conditions wherethe fragment (0.1 to 1 µg) is present in excess over theamount of vector DNA (ten times or more) in a small-volume reaction mixture (5 to 10 µl). After transformationof E. coli, only direct repeats of the fragments, but not

Kiyama: Periodicity of DNA bend sites in eukaryotic genomes

642

Fig. 1. Assay for bent DNA.

inverted repeats, can be obtained in dimeric form.

The circular permutation assay is a very simple andreliable method to identify and roughly map DNA bendsites. The results of mapping are reproducible underidentical electrophoresis conditions and generallyreproducible among different subclones containing thesame site, and the patterns can be interpreted without

complicate calculations. However, this method does havesome technical limitations. Firstly, the assay is totallydependent upon the availability of suitable restrictionsites. If there are no appropriate restriction sites, the bendsites cannot be localized to a small region of DNA.Secondly, the DNA structure of the other part of the samefragment could influence the mobility. As a result of thiseffect, mapping a bend site to a very small region by thismethod would not always give a precise location.Although rare, we observed a slight difference in thelocation of a site in the !-globin gene region betweenclones of different sizes used for mapping. Therefore, thelower limit of resolution would be 50 to 100 bp.

The more precise location of the bend sites could beachieved by several other methods. For a relatively largeregion, sites can be examined with deletion constructs.When the bend center is completely deleted from theconstruct, all restricted fragments have the same mobility.Meanwhile, mapping the site in regions of approximately100 bp or less would be achieved by using concatenatedoligonucleotides of 20 or 30 bp (Wada-Kiyama andKiyama, 1995). When the oligonucleotide contains a bendsite, the concatemers exhibit retardation onpolyacrylamide gel electrophoresis. The effect of bendingis greater as the length of the oligonucleotide increases.The bend angle can be estimated by comigration ofstandards such as A3N7 (0.63˚/ base) (Calladine et al.,

1988). The bend angle could also be determined by theassay based on ring closure of concatenatedoligonucleotides (Zahn and Blattner, 1987).

III. The human -globin locus

Using the circular permutation assay, we mapped theDNA bend sites in the human "-globin locus which is

located on chromosome 11 and contains five (!-, G#-, A#-,$- and "-) active genes and one (%"-) pseudogene (Figure2). This locus is ideal for mapping the sites because thenucleotide sequence of over 70 kb has been reported.Furthermore, since most of the locus is intergenic, theinfluence of the coding region could be excluded. Thesimilarity of the exon-intron structure and the sequencesof the flanking region would be ideal for evolutional studyof the sites. The chromatin structure in this locus has beenextensively characterized in that switching of globin geneexpression is paralleled by alterations of chromatinstructure as revealed by DNase I-hypersensitivity(reviewed by Stamatoyannopoulos and Nienhuis, 1993;Evans et al., 1990).

The periodicity of the bend sites at intervals of 680 bpon average was first identified in the !-globin region(Wada-Kiyama and Kiyama, 1994). Further studies of thesites in the regions of other globin genes revealed thatrelative locations of the sites to their cap sites wereconserved among most of the members of this familywhich were separated as much as 200 million years ago.


643

Fig. 2. Periodic bent DNA in the human "-globin locus. Mapped DNA bend sites are shown as shadowed boxes. Hatched boxes

indicate putative 150 bp sites aligned at 680 bp intervals as a reference for periodicity.


644

Fig. 3. Conservation of periodic bent DNA in the translocation of the c-myc and Igµ loci. DNA bend sites in the c-myc (bottom) and

Igµ (top) loci are aligned to highlight the conservation of the periodicity of the hypothetical sites (shadowed columns) based on their

universal periodicity, after the rearrangements observed in Manca (A), BL22 (B) and Ramos (C) cell lines. Only three hypothetical sitesnear the junctions are shadowed but they were matched throughout the loci. Reproduced from Ohki et al. (1997).


645

Table 1. Average intervals of periodic bent DNA.________________________________________________________________________________________

Locus Mapped No. of Average S. D. a Ref.region (kb) sites (bp) (bp)

________________________________________________________________________________________"-globin 66 98 679.2 229.6 bc-myc 8 12 694.2 281.4 Ohki et al. 1997Igµ 7 11 654.5 222.7 Ohki et al. 1997Erythropoietin receptor 9 13 651.2 221.0 b

Estrogen receptor c 3 5 688.1 210.9 Kuwabara et al. 1997________________________________________________________________________________________a Standard deviation.b Unpublished results.c 5'-region containing the alternative cap site, P0.

The duplication of the two #-globin genes, which occurredmost recently, was immediately followed bydiversification of the non-coding region by as much as70%, while all of the bend sites were conserved (Slightomet al., 1980). Insertion of an Alu sequence might havedisturbed the periodicity, although as observed in theregion upstream of the !-globin gene, the interval seemedto have returned to the average after a long period ofmolecular evolution. The positions of the bend sites were

conserved even between the human "- and mouse "maj-globin genes (Wada-Kiyama and Kiyama, 1996b).

Mapping of over 90 bend sites in the locus revealedthat the periodicity of the bend sites exists throughout thelocus with an average interval of 680 bp (Wada-Kiyamaand Kiyama, 1994, 1995, 1996b; unpublished results).However, we observed disturbance of the periodicity atseveral locations. Interestingly, all of the locations of thedisturbed periodicity located upstream of the !-globingenes that caused the distances of the adjacent bend sitesto be longer than average were found in or close to theDNase I-hypersensitive sites, which constitute the locuscontrol region ("-LCR). The "-LCR is composed of fouror five developmentally-regulated DNase I-hypersensitivesites (Crossley and Orkin, 1993; Evans et al., 1990;Felsenfeld, 1993). These sites are designated as openchromatin regions and act as the sites of interaction oftranscription factors and the enhancer-binding protein NF-E2. This region interacts with the promoter region of eachmember of the "-like globin gene family and controls theirexpression during development. One of the DNase I-hypersensitive sites, HS2 located 11 kb upstream of thecap site of the !-globin gene, was located in the center oftwo adjacent bend sites separated by a distance of 860 bp,which is longer than average (unpublished results). Itseemed as if HS2 was placed far from the bend sites tominimize the influence of the sites. This is discussed againbelow.

IV. The human c-myc and theimmunoglobulin heavy chain loci

The human c-myc gene has three exons and occupies aregion of approximately 5.5 kb on chromosome 8. Asobserved in the "-globin locus, this locus containedperiodic bent DNA. DNA bend sites were mapped at anaverage interval of 694.2 bp and were present in the 5'-and 3'- non-coding regions, introns and the non-codingexon (exon 1), but not present in the coding region (Ohkiet al., 1997). Interestingly, one of the bend sitescorresponded to the location of TATA box of the P2promoter, suggesting that prebending of the promoterregion can facilitate transcriptional enhancement.

The c-myc gene is involved in the progression ofBurkitt's lymphoma by translocation of the locus into oneof the immunoglobulin genes located on chromosomes 2,14 or 22. These translocation events often result inreshuffling the location of regulatory elements.Deregulation of the expression by juxtaposition of the µenhancer to the c-myc promoter is one of the mechanismsof tumor progression caused by this oncogene. Themechanism of these translocation events has not been welldocumented except that immunoglobulin-specificrecombination is somehow involved (Specer andGroudine, 1991). Translocation junctions were formed atvarious locations in the locus yet no specific sequenceswere commonly found in their immediate proximity.However, when the periodic bent DNA was mapped in thec-myc and the Igµ loci, at least three stable cell linescontaining the translocation junctions within these regionsshowed conservation of the periodicity before and after therearrangements (Figure 3). This would be readilyexplained if we assume that the periodic bent DNA is akey element for chromatin structure. It would be necessaryfor a stable cell line to maintain a similar chromatinstructure as to that before the rearrangement. Otherwise, asecondary rearrangement could alter the sequence until astable structure is eventually formed.


646

V. Other loci in eukaryotic genomes

We have already determined that periodic bent DNA ispresent in the human erythropoietin receptor and thehuman estrogen receptor loci (Kuwabara et al., 1997;unpublished results). The intervals of the sites in these lociwere 651.2 or 688.1 bp, respectively, which are close tothe values for other loci (Table 1). In both cases, itsperiodicity was disturbed by exons. For the estrogenreceptor gene, the alternative cap site locatedapproximately 2 kb upstream of the canonical site causeda shift of the nearby sites. For the erythropoietin receptorgene, the 700 bp periodicity of bend sites was conservedeven within the long introns (1st and 6th introns) althoughthe sites were shifted when the length of introns was notsufficient to accommodate two sites. One of the sites inthe estrogen receptor gene contained motifs of theestrogen response element, the binding site for thehormone-responsive trans-activating factors, and had theaffinity to the nuclear scaffold. This site might play a rolein determining the nuclear localization of this gene as wellas a role in transcriptional regulation.

For other loci of eukaryotes, we examined the potentialbend sites by a computer search (Wada-Kiyama, andKiyama, 1996a). Based on the observation that physicallymapped bend sites often contain A+T-rich sequencesincluding An or Tn tracts at intervals of ten or multiples of

ten nucleotides, we searched A2N8A2N8A2 and the

complementary T2N8T2N8T2 for a periodicity. There was

a statistically significant sequence periodicity at aninterval of roughly 700 bp in eukaryotic genomic DNA.This tendency was absent in prokaryotes and in eukaryoticcDNA, suggesting that the periodicity is universal amongeukaryotic genomes, especially in intergenic regions.

VI. Biological significance of the bend sites

The observations that periodic bent DNA is conservedduring molecular evolution and its intervals aremaintained precisely in otherwise unstable intergenicregions suggested that these sites are biologically relevant.Despite the systematic and organized patterns ofchromatin folding, no specific signals have beendetermined as key elements for the folding mechanism. Acomputer search further revealed the non-randomdistribution of nucleotide sequences on the genomic DNA,while it failed to deduce any specific sequences incommon, suggesting the presence of unidentified codeswhich are not apparent from sequence information alone.Therefore, judging from the periodicity and the length oftheir intervals, periodic bent DNA may be closelyassociated with chromatin structure, presumably with theformation of nucleosomes. We reported that some of thesites were indeed involved in the formation ofnucleosomes by having high affinity to histone coreparticles (Wada-Kiyama and Kiyama, 1996b). Chromatinstructure seems to be extensively stabilized when theoverall periodicity is maintained before and after therearrangement. Meanwhile, open chromatin regions,

revealed by DNase I-hypersensitivity (Gross and Garrard,1988), could be at least partly due to disturbance of theperiodicity. While the nucleosome phasing activity ofthese sites might be effective when the distances of thebend sites are less than or equal to the length of fournucleosomes, open chromatin regions would be moreefficiently formed when their distances are more than thelength of four nucleosomes. Some of the sites seem to beused as multiple sites for chromatin organization, asobserved in the estrogen receptor gene. We are currentlyinvestigating chromatin structure at the replication originbased on the alignment of bend sites to examine therelationship of these sites with DNA replication. Ourresults indicated that periodic bent DNA is a key elementof chromatin structure and plays a role in variousbiological reactions.

References

Blaho, J. A. and Wells, R. D. (1989) Left-handed Z-DNA andgenetic recombination. Prog. Nucl. Acid Res. Mol. Biol. 37,107-126.

Calladine, C. R., Drew, H. R. and McCall, M. J. (1988) Theintrinsic curvature of DNA in solution. J. Mol. Biol. 201,127-137.

Crossley, M. and Orkin, S. (1993) Regulation of the "-globin

locus. Curr. Opin. Genet. Dev. 3, 232-237.

Crothers, D. M., Haran, T. E. and Nadeau, J. G. (1990).Intrinsically bent DNA. J. Biol. Chem. 265, 7093-7096.

Evans, T., Felsenfeld, G. and Reitman, M. (1990) Control ofglobin gene transcription. Ann. Rev. Cell Biol. 6, 95-124.

Felsenfeld, G. (1993) Chromatin structure and the expression ofglobin-encoding genes. Gene 135, 119-124.

Gross, D. S. and Garrard, W. T. (1988) Nuclease hypersensitivesites in chromatin. Ann. Rev. Biochem. 57, 159-197.

Hagerman, P. J. (1990). Sequence-directed curvature of DNA.Ann. Rev. Biochem. 59, 755-781.

Ioshikhes, I., Bolshoy, A., Derenshteyn, K., Borodovsky, M. andTrifonov, E. N. (1996) Nucleosome DNA sequence patternrevealed by multiple alignment of experimentally mappedsequences. J. Mol. Biol. 262, 129-139.

Khan, J. D. and Crothers, D. M. (1992). Protein-induced bendingand DNA cyclization. Proc. Natl. Acad. Sci. USA 89, 6343-6347.

Kuwabara, K., Wada-Kiyama, Y., Sakuma, Y. and Kiyama, R.(1997) Multiple interactions of periodic bent DNA in thepromoter region of the human estrogen receptor gene withthe nuclear scaffold, core histones and nuclear factors.submitted.

Nash, H. A. (1990) Bending and supercoiling of DNA at theattachment site of bacteriophage lambda. Trends Biochem.Sci. 15, 222-227

Ohki, R., Hirota, M., Oishi, M. and Kiyama, M. (1997)Conservation and continuity of periodic bent DNA ingenomic rearrangements between the c-myc and

immunoglobulin heavy chain µ loci. submitted.


647

Rich, A., Nordheim, A. and Wang, A. H.-J. (1984) Thechemistry and biology of left-handed Z-DNA. Ann. Rev.Biochem. 53, 791-846.

Slightom, J. L., Blechl, A. E. and Smithies, O. (1980). Human

fetal G#- and A#-globin genes: Complete nucleotide

sequences suggest that DNA can be exchanged betweenthese duplicated genes. Cell 21, 627-638.

Soyfer, V. N. and Potaman, V. N. (1996) In Triple-HelicalNucleic Acids. Springer Verlag, New York.

Specer, C. A. and Groudine, M. (1991). Control of c-mycregulation in normal and neoplastic cells. Adv. Cancer Res.56, 1-48.

Stamatoyannopoulos, G. and Nienhuis, A. W. (1993)Hemoglobin switching. In The molecular basis of blooddiseases, Stamatoyannopoulos, G., Nienhuis, A. W.,Majerus, P. and Varmus, H. (eds). W. B. Saunders,Philadelphia. pp107-155.

Travers, A. A. (1989) DNA conformation and protein binding.Ann. Rev. Biochem. 58, 427-452.

Trifonov, E. D. (1991) DNA in profile. Trends Biochem. Sci.16, 467-470.

Wada-Kiyama, Y. and Kiyama, R. (1994). Periodicity of DNA

bend sites in the human !-globin gene region: Possibility of

sequence-directed nucleosome phasing. J. Biol. Chem. 269,22238-22244.

Wada-Kiyama, Y. and Kiyama, R. (1995). Conservation and

periodicity of DNA bend sites in the human "-globin gene

locus. J. Biol. Chem. 270, 12439-12445.

Wada-Kiyama, Y. and Kiyama, R. (1996a) Conservation andperiodicity of DNA bend sites in eukaryotic genomes. DNARes. 3, 25-30.

Wada-Kiyama, Y. and Kiyama, R. (1996b) Anintrachromosomal repeating unit based on DNA bending.Mol. Cell. Biol. 16, 5664-5673.

Werner, M. H., Gronenborn, A. M. and Clore, G. M. (1996)Intercalation, DNA kinking, and the control of transcription.Science 271, 778-784.

Wu, H.-M. and Crothers, D. M. (1984). The locus of sequence-directed and protein-induced DNA bending. Nature 308,509-513.

Zahn, K. and Blattner, F. R. (1987). Direct evidence for DNAbending at the lambda replication origin. Science 236, 416-422.


649


DNA methyltransferase: a downstream effector ofoncogenic programs; implications for therapy.

Moshe Szyf

Department of Pharmacology and Therapeutics, McGill University, 3655 Drummond street, Montreal PQ H3G 1Y6,Canada.

__________________________________________________________________________________________________

Correspondence: Moshe Szyf, Tel: 514-398-7107, Fax: 514-398-6690, E-mail: mcms @musica.mcgill.ca

Summary

DNA MeTase is an attractive anticancer target . A molecular analysis of i ts regulation suggeststhat i t is a downstream effector of many oncogenic pathways and that i ts down modulation caninhib i t tumor growth . I t i s poss ib le that DNA MeTase inh ib i tor wi l l be e f fec t ive in a broadspectrum of cancers because i t l i es downstream to nodal cellular checkpoints that could beactivated by multiple ways. An important challenge is to design novel inhibitors of DNA MeTasethat are highly specif ic. Such inhibitors are potentially important pharmacological agents withwide therapeutic applications.

I. Introduction

A. Working hypothesis: DNA MeTase isan anticancer target

This review summarizes recent findings demonstratingthat the cytosine DNA Methyltransferase (DNA MeTase)is a downstream effector of cellular pathways leading eitherto oncogenesis or a change in the state of differentiation ofvertebrate cells (Rouleau et al., 1992; Rouleau et al.,1995; MacLeod et al., 1995; MacLeod and Szyf, 1995;Szyf et al., 1992). It is becoming clear that control ofgene expression by pharmacological means is one of thegreat challenges and hopes of current therapeutics. DNAMeTase is a master regulator of gene expression programs.This review suggests that genomic programs could bespecifically modulated by pharmacological inhibition ofDNA MeTase. Since DNA MeTase is believed to beinvolved in similar processes in a broad group of animalsand plants, my hypothesis is that agents that inhibit DNAMeTase specifically will have broad pharmacologicalapplications (Szyf 1994; Szyf 1996). Recent data fromour laboratory using different classes of DNA MeTaseinhibitors supports this hypothesis.

II. Background

The goal of this section is to illustrate the logicalprogression of concepts and data leading to our workinghypothesis and to the proposal that DNA MeTaseinhibitors could serve as important pharmacologicalagents.

A. What is DNA methylation?

DNA methylation is a postreplicative covalentmodification of DNA that is catalyzed by the DNAmethyltransferase enzyme (DNA MeTase) ( Razin andSzyf, 1984; Bestor et al., 1988). The main concept inDNA methylation is the idea of a pattern of DNAmethylation and its correlation with the state of activity ofgenes (Yisraeli and Szyf, 1984). In vertebrates, thecytosine moiety at a fraction of the CpG sequences ismethylated (60-80%); the non methylated CpGs aredistributed in a nonrandom manner generating a pattern ofmethylation that is gene and tissue specific (Yisraeli andSzyf, 1984). Plant DNA is also methylated at CG as wellas CXG sequences (Gruenbaum et al., 1984). DNAmethylation plays an important role in development ofplant cells and might be a critical element involved insilencing transgene expression in plants (Meyer, 1995).

B. DNA Methylation patterns encodeepigenetic information

1. Correlation of gene expression and DNAmethylation patterns

Does the pattern encode epigenetic information? Alarge number of papers published in the last two decadeshave shown a correlation between the pattern ofmethylation and the state of activity of genes (Yisraeli andSzyf, 1984). That is, some or all of the CpG sites in


650

F i g u r e 1 . D N A m e t h y l a t i o n p a t t e r n s f i x g e n e e x p r e s s i o n p r o g r a m s , D N A M e T a s e i n h i b i t o r s a l t e rgene express ion programs .

The genome of a vertebrate (first line) bears many potential sites for modification by methylation (open circles). However, asubset of these sites is methylated (indicated by an M in the circle) in different tissues. When one looks at the methylation patternof different genes (a to e) in different tissues (for example tissues A and B), one observes that they bear a different pattern ofmethylation. One also observes that in general inactive genes are modified whereas active genes bear sites of methylation thatare not modified. It is proposed that both binding of transcription factors to regulatory sites of the genes (indicated by thetriangles and ovals-activators in green and methylated-DNA binding factors-repressors) as well as the pattern of methylation(methylated sites attract methylated-DNA dependent repressors) define the state of activity of vertebrate genes. The pattern ofmethylation is maintained because of limiting level of DNA MeTase, thus the level of DNA MeTase locks the gene expressionprogram of a tissue. Inhibition of DNA MeTase by DNA MeTase inhibitors results in transient demethylation and unlocking of thegene expression program. Demethylation enables reorganization of the interactions of transcription factors with DNA andresetting of a new program of gene expression. The direction that this reorganization will take is limited by the repertoire oftranscription factors in the cell.

regulatory regions of a specific gene will be methylated inall tissues where the gene is silenced but the same siteswill be nonmethylated in tissues that express the gene(Yisraeli and Szyf, 1984) (Fig . 1 ).

2. DNA methylation and gene expression:cause or effect?

There is a longstanding and unresolved discussionwhether the state of methylation of genes is a cause oreffect of their state of expression. Whereas a series ofexperiments demonstrated that inactivation of a geneprecedes its repression (Lock et al., 1987), other studieshave shown that methylation of genes before they are

introduced into cells can suppress their activity (Stein etal., 1982, for a review of this question see Szyf, 1996).One possible solution to this dilemma is the suggestionthat there is a dynamic interrelationship between DNAmethylation and gene expression (Szyf, 1996). DNAmethylation can play both a primary and secondary role ingene expression. That is, methylation of certain sitesprecipitates gene repression whereas in other instances thechromatin structure of a repressed gene can trigger DNAmethylation. The combination of these processes shouldresult in formation of a stable state of gene repression by acovalent modification of the DNA structure itself. Thus,the pattern of methylation in a cell will stabilize a geneexpression program for this cell (Fig . 1 ).


651

Figure 2 . Twomechanisms for generepress ion b ym e t h y l a t i o n .

DNA methylation caninhibit gene expression by twodifferent mechanisms. Themodel gene described in thisfigure is activated (horizontalarrow indicates mRNAtranscript) by interaction oftranscription factors (triangle)with a cis acting sequence(shaded box) located in front ofthe transcription initiation site.The gene has a number ofmethylatable sites (opencircles), one of which is locatedat the transcription factorrecognition sequence. Mechanism A describes a casewhere the methylatable sitelocated at the transcriptionrecognition site is methylated(M). This methylation inhibitsthe recognition of the cis actingsequence by the transcriptionfactor. Mechanism B describesa case where a regionalmethylation occurred in thebody of the gene. Thismethylation results in bindingof methylated-DNA bindingprotein(s) to the methylatedregion (open oval). Thebinding of this proteinprecipitates the spreading of aninactive chromatin structure,the gene becomes inaccessible

to transcription factors and is not transcribed.

3. What is the mechanism of gene repressionby methylation?

A series of publications suggest that DNAmethylation can repress gene expression directly, byinhibiting binding of transcription factors to regulatorysequences (Becker et al., 1987), or indirectly, by signalingthe binding of methylated-DNA binding factors that repressgene activity or by precipitating an inactive chromatinstructure (Razin and Cedar, 1977; Keshet et al., 1986).Two methylated DNA binding proteins that can represstranscription in a methylation dependent manner have beenrecently characterized MeCP2 and MeCP1(Cross, et al.,1997; Nan et al., 1997). The carboxy terminal half ofMeCP2 contains a repressor domain which can interactwith the transcriptional machinery. MeCP2 is alsosuggested to precipitate or stabilize an inactive chromatinstructure. Kass et al., have shown that methylated DNA isassembled into an inactive chromatin structure (Kass et al.,1997). It is not clear yet whether MeCP2 is generally

involved in the precipitation of an inactive chromatinstructure on methylated DNA or whether othermechanisms are involved in building of inactive chromatinaround methylated DNA (Fig . 2 ).

4. Summary: methylation plays an importantrole in control of genomic functions.

A long list of data supports the hypothesis that DNAmethylation plays an important role in the control ofgenomic functions. It is well established that regulatedchanges in the pattern of DNA methylation occur duringdevelopment (Brandeis et al., 1993) , parental imprinting(Peterson and Sapienza, 1993) and cellular differentiation(Razin et al., 1985) and that aberrant changes in the patternof methylation occur in cellular transformation (Feinberget al., 1983; de Bustros et al., 1988; Baylin et al., 1991).A targeted mutation, by homologous recombination in EScells, of the DNA MeTase gene results in embryonic

Szyf: DNA Methyltransferase in cancer therapy

652

lethality (Li et al., 1992) and inhibition of DNA MeTaseby expression of an antisense to the DNA MeTase resultsin a change in the identity of a cell from a fibroblast to acell of myogenic lineage (Szyf et al., 1992).

Whereas the most studied biological process regulatedby DNA methylation is gene expression, it is possiblethat DNA methylation directly regulates other genomefunctions such as replication and recombination. Theseprocesses might be as important as gene expression in theevents leading to oncogenesis (Szyf, 1996) . If a DNAmethylation pattern locks certain gene expressionprograms, then modulating the pattern should unlock theseprograms and play important therapeutic roles (Szyf, 1994;Szyf, 1996) (F i g . 1 ). To be able to design therapeuticstrategies to unlock a pattern of methylation and geneexpression, one has to understand the mechanisms thatcontrol DNA methylation patterns.

C. The level of DNA MeTase is animportant determinant of DNA methylationpatterns.

1. The semiconservative model of methylationinheritance

The accepted model in the field has been that patternsof methylation are maintained because the DNA MeTase isvery efficient in methylating hemimethylated DNAgenerated in the process of replication (maintenancemethylation) but very inefficient in methylatingnonmethylated DNA (de novo methylation) (Razin andRiggs, 1980). However in spite of the simplicity of thismodel, both the cloned DNA MeTase (Tollefsbol andHutchinson, 1995) and a putative new enzyme (Lei et al.,1996) have been shown to bear de novo methylationactivity.

2. The role of cis acting signals

If de novo methylation is possible, what determines

the specificity of DNA methylation? One importantfactor is cis signals contained in the sequence and putativecellular factors recognizing these signals (Szyf et al.,1989; Szyf, 1991) (Fig . 3 ). These signals possiblydirect the general DNA methylation machinery to specificregions. Several experiments have shown that thepresence of certain cis-acting sequences protect adjacentsequences from methylation (Szyf et al., 1990) while othersequences target adjacent sequences to become methylated(Szyf et al., 1989).

3. DNA methylation patterns are regulated byan interplay between local signals and the level ofDNA MeTase activity. The role of cell legacy

Could the pattern of DNA methylation be determinedalso by central cellular signals? Very limited attention hasbeen given to the role that the cellular level of the enzymecatalyzing DNA methylation might play in determiningand controlling DNA methylation patterns. One obviousreason why this level of regulation was not considered isbecause it had been difficult to explain how a generalchange in the level of the enzyme could lead to discretechanges in DNA methylation. To address that question Ihave previously suggested that the pattern of methylationis determined by an interplay between local signals, assuggested for example by the de novo methylation of theC21 gene in Y1 cells (Szyf et al., 1989), and the level ofDNA MeTase (Szyf et al., 1984) and demethylaseactivities (Szyf 1994; Szyf et al., 1995) in the cell (Fig .3 ). The affinity of each CpG site to DNA MeTase isdetermined by either the properties of the sequence or theDNA binding proteins interacting with it in specific celltypes. Thus, the final pattern of methylation will reflectthe legacy of the cell, its specific repertoire of DNAbinding factors and resulting chromatin structure.

According to this hypothesis we predict that a generalchange in DNA MeTase will result in a predictable changein DNA methylation pattern that is specific per cell type.In accordance with this hypothesis it has been shown thata limited inhibition of DNA MeTase results in specific

Figure 3. What determinesDNA methylation patterns? Amodel .

DNA methylation patterns aredetermined by an interplay between:Transacting factors (triangle-factorsenhancing methylation, oval-factorsenhancing demethylation), Signals inDNA (red-enhancing methylation,green-enhancing demethylation),Levels of DNA MeTase and DNAdemethylase activities. The pattern ofmethylation could be altered bymodulation of any of these factors.The ideal targets for pharmacologicalintervention are the enzymaticactivities.


653

Figure 4: Regulat ion of theDNA MeTase, a model.

The DNA MeTase promoter and5’ upstream region: the filled boxesindicate the first exons. Twotranslation initiation sites wereidentified, potentially resulting ina long and a short form of DNAMeTase (indicated by ATG) whichmight exhibit differenttransforming capabilities. Each ofthese initiation sites is regulatedby a different promoter. The lowertranslation initiation site is aproduct of a transcript(transcription initiation sitesindicated by horizontal arrows)initiating downstream of apromoter that is regulated by theRas signaling pathway. The AP-1recognition sequences located at (~-1.7) upstream of the lowertranscription initiation sites areindicated as ovals. The expressionof DNA MeTase is regulated atmultiple levels: First, by choice ofpromoter resulting in two differentDNA MeTase forms. Second, thebasal promoter which maintainslimiting levels of DNA MeTaseexpression and is possiblyregulated by tumor suppressors.Third, the mRNA is destabilized atG0 phase of the cell cycle. This isregulated by the retinoblastomaprotein. T antigen can remove thisregulation and stabilize the DNAMeTase mRNA. Fourth, a cluster ofAP-1 sites can mediatetransactivation of DNA MeTase bysignal transduction pathways andby oncogenic signals. Theactivation by AP-1 could berepressed by glucocorticoidreceptor.

alterations in DNA methylation patterns and differentiationof the cell to the next stage in differentiation rather than achaotic loss of identity (Szyf et al., 1992). Animalexperiments and clinical trials performed two decades agohave shown that general inhibition of DNA MeTase by theDNA MeTase inhibitor 5-Aza-CdR resulted in specificactivation and demethylation of ! globin gene in animalsand patients (Ley et al., 1982; DeSimone et al., 1982).

4. DNA MeTase is regulated by centralcellular signals at the transcriptional andposttranscriptional level

a. Transcriptional regulation:

During the last five years, we have shown that theDNA MeTase is regulated at both the transcriptional andposttranscriptional level by nodal cellular signalingpathways (Szyf, 1994; Szyf, 1996). Cloning andcharacterizing the promoter of the DNA MeTase enabled usto determine that it bears AP-1 sites which aretransactivated by Jun, a downstream effector of the nodalcellular and oncogenic Ras signaling pathway (Rouleau etal., 1992; Rouleau et al., 1995) (Fig . 4 ). Downregulation of the Ras-Jun pathway in the mouseadrenocarcinoma cell line Y1 leads to inhibition of DNAMeTase expression, inhibition of DNA methylation,alteration of DNA methylation patterns and reversal ofoncogenesis (MacLeod and Szyf, 1995; MacLeod et al.,


654

1995). Recently, an additional start site upstream to theone identified by us which encodes a new translationinitiation site resulting in a larger protein has beenreported (Tucker et al., 1996). This start site is located ina CG island which is characteristic of housekeeping genes.My hypothesis is that differential utilization of these twopromoters, an AP-1-regulated activity versus ahousekeeping basal regulation plays an important role indetermining the cellular level of DNA MeTase and itssubstrate specificity. Our recent data demonstrates thatthe human DNA MeTase promoter region bears as well alarge number of AP-1 sites and is also upregulated by theRas signaling pathway (Ramchandani et al., unpublisheddata).

b. Posttranscriptional regulation:

The DNA MeTase is also regulated with the phase ofthe cell cycle (Szyf et al., 1991). Whereas transcription ofthe message continues throughout the cell cycle, DNAMeTase mRNA is absent in Go cells which is consistentwith posttranscriptional control of DNA MeTase mRNA.It has been recently shown that posttranscriptionalregulation of DNA MeTase is also associated with muscledifferentiation (Liu et al., 1996). Our recent data haslinked the posttranscriptional regulation of DNA MeTaseto basic cellular pathways that are known to play a criticalrole in cellular transformation (Fig . 4 ). We have recentlyshown that ectopic expression of SV40 T antigen innontransformed 3T3 cells results in induction of DNAMeTase mRNA, DNA MeTase protein levels and genomicDNA methylation. A T antigen mutant which has lostthe ability to bind pRb does not induce DNA MeTase. Surprizingly, this upregulation of DNA MeTase by Tantigen occurs mainly at the posttranscriptional level byaltering mRNA stability. Inhibition of DNA MeTase by5-Aza-CdR reverses T antigen induced transformationsuggesting a causal role for increased DNA MeTaseactivity in T antigen triggered transformation (Pinard etal., unpublished observations). This data links the Rbtumor suppressor pathway to the posttranscriptionalregulation of DNA MeTase. It is interesting to note thatRas and T antigen can only transform primary cells whenthey are jointly expressed. It is tempting to speculate thattheir cooperative role in cellular transformation reflects thefact that they regulate DNA MeTase at two different levels.

An alternative mechanism that might be responsiblefor the specificity of DNA MeTase is regulated alternativesplicing of exons encoded by the DNA MeTase gene.Recent data from my laboratory suggests that the humanDNA MeTase is encoded by 40 exons and that there is apotential for in -frame alternative splicing that will resultin different forms of DNA MeTase (Deng et al.,unpublished results). Regulation of this process withdifferent developmental stages might play an importantrole in specifying specific classes of sites for methylation.

In summary, our observations do not only establishthe regulation of DNA MeTase by central cellularsignaling pathways, but also suggest a potential molecular

link between DNA methylation and oncogenic pathways(Szyf, 1994) (Figures 4 , 5 ). The concept that DNAmethylation patterns could be controlled by altering thelevel of the DNA MeTase leads to the idea that partialinhibition of DNA MeTase by pharmacological agentscould result in altering gene expression programs (Fig .4 ). One example of many possible applications of myhypothesis is the use of inhibition of DNA methylation toinhibit cellular transformation (F i g . 5 ). Inhibitors ofDNA methylation could possibly be used to alter andcontrol genetic programs in humans, plants and animalsand might have broad application in clinical medicine,veterinary medicine and agriculture (Meyer, 1995).

III. DNA MeTase is an importanttherapeutic target.

A. DNA methylation and cellulartransformation

1. Is methylation involved in oncogenesis?

The findings that the level of DNA MeTase as well asthe pattern of DNA methylation might be controlled byoncogenic pathways leads to the question of whether DNAmethylation plays an important role in cellulartransformation? An activity that has a widespread impacton the genome such as DNA MeTase is a good candidateto play a critical role in cellular transformation. Thishypothesis is supported by many lines of evidence thathave demonstrated aberrations in the pattern of methylationin transformed cells.

2. Induction of dMTase activity in cancer cellsexplains the hypomethylation observed in thesecells.

Although it is clear that methylation patterns arealtered in cancer cells, the direction that these changes takeis perplexing. While many reports show hypomethylationof both total genomic DNA (Feinberg et al., 1988) andindividual genes in cancer cells (Feinberg et al., 1983),other reports have indicated that hypermethylation ofspecific loci such as "tumor suppressor" genes is animportant characteristic of cancer cells (Makos et al.,1992; Baylin et al., 1988; Baylin et al., 1991). How canone resolve this contradiction? We have recently suggested(Szyf, 1994) that the hypomethylation observed in cancercells is a consequence of increased DNA demethylationactivity induced by oncogenic pathways such as Ras7which acts on a different subset of sites than those that arehypermethylated (Szyf et al, 1995). We have recentlycharacterized a bona fide demethylase activity that isespecially abundant in all cancer cells (Bhattacharya andSzyf unpublished results). We suggest that the dMTaserecognizes CpG sites with different specificity than theDNA MeTase resulting in concomitant hypomethylationof some sites and hypermethylation of other sites (Szyf,1994).


655

Figure 5: DNA MeTase hyperact ivat ion and tumorigenes i s ; reversa l by DNA MeTase antagonis t s , amodel .

Regulated expression of DNA MeTase (standing rectangle indicating MeTase level is partly filled) is critical for maintainingthe pattern of methylation (open circles indicate methylatable sites and M indicates methylation) and locking a somatic cell in inits program. Two facets of genome functions are regulated by DNA methylation. First, the profile of gene expression (themaintenance of the cognate program is indicated by a lock, expressed genes are indicated by horizontal arrows, transcriptionfactors are indicated by ovals and triangles, methylated-DNA binding repressors are indicated by red ovals). Second, control ofDNA replication is regulated by methylation (indicated by a stop sign). The replication control sequences (indicated by the opensquare) are not methylated in a resting somatic cell, signaling arrest of DNA replication. Oncogenic signaling pathways caninduce the DNA MeTase resulting in hypermethylation of certain sequences, both genes and replication control regions. Thisresults in loss of the original gene expression profile of the cell (open lock) and loss of the control over replication. Inhibitors ofDNA MeTase can reduce the level of DNA MeTase, resulting in hypomethylation and activation of the replication control regionsas well as restoration of some of the original gene expression program.

3. Induction of DNA MeTase by oncogenicpathways is a critical component of oncogenicprograms

The critical remaining question is whether the"hypermethylation" observed in cancer cells is aprogrammed or random event? Random events are moredifficult to control pharmacologically. However, ifhypermethylation is a consequence of a programmedincrease in DNA MeTase activity, the probability ofreversing this state by inhibitors of DNA MeTase is high.

A possible explanation for this observed hypermethy-

lation is that it is a consequence of the limited increase inDNA MeTase activity observed in many tumor cells(Kautiainen and Jones, 1986; el-Deiry et al., 1991).Recently Belinsky et al., have shown that increased DNAMeTase activity is an early event in carcinogen inducedlung cancer in mice (Belinsky et al., 1996). Forcedexpression of exogenous DNA MeTase cDNA causestransformation of NIH 3T3 cells supporting the hypothesisthat overexpression of DNA MeTase can cause cellulartransformation (Wu et al., 1993). Our data demonstratingthat the increase in DNA methylation activity in cancercells is an effect of activation of either the oncogenic Ras-


656

Jun signaling pathway (Szyf, 1994; MacLeod et al., 1995)or the oncogenic pathway induced by T antigen (Pinard andSzyf, unpublished data) supports the hypothesis thatincreased DNA MeTase is a critical component of diverseoncogenic programs. Several lines of evidence obtained byus support a causal role for increased DNA MeTase inoncogenesis. First, treatment of Y1 adrenocorticalcarcinoma cells with the DNA MeTase inhibitor 5-azaC-CdR or stably expressing an antisense to DNA MeTase inthese cells results in inhibition of tumorigenesis in vitro(MacLeod and Szyf, 1995). Second, when the DNAMeTase antisense transfected Y1 cells are injected into asyngeneic mouse, tumor formation in vivo issignificantly inhibited (MacLeod and Szyf, 1995). Third,5-Aza-CdR treatment of T antigen transformed 3T3 cellsresults in inhibition of cellular transformation in vitro . Fourth, intra peritoneal administration of phosphorothioatemodified DNA MeTase antisense oligonucleotides inhibitstumorigenesis in vivo in LAF/1 mice (syngeneic strain)bearing Y1 tumors (Ramchandani et al., 1997). Similarly, Laird et al., (1995) have shown that treatingmice bearing the Min mutation with 5-Aza-CdRsignificantly reduces the appearance of intestinal polyps.

Based on these data, our working hypothesis is thatinduction of the enzymatic machinery controlling DNAmethylation is a critical component of oncogenic programsand that oncogenesis could be reversed by inhibiting thisinduction.

B. What is the mechanism by whichoverexpression of the DNA MeTase inducestumorigenesis?

If induction of DNA MeTase is an importantcomponent of an oncogenic program, what is themechanism? Based on what is known about the functionsof DNA methylation in diverse biological systems, threealternative possible modes of actions emerge.Hypermethylation can result in stable mutations, canrepress tumor suppressor genes or possibly directly controlDNA replication.

1. DNA MeTase induces C to T transitions

The first mechanism proposed by Peter Jones is thathypermethylation can increase the probability of reversionof 5mC to T by deamination, resulting in mutagenesis(Jones et al., 1992). However, this is an irreversiblemechanism which is inconsistent with recent data. Lairdet al., have previously shown that treatment of micebearing the Min allele of APC with the DNA MeTaseinhibitor 5-Aza-CdR reduces the frequency of polypformations in these mice suggesting that DNA MeTase iscritical for tumor formation in Min mice. If themechanism by which DNA MeTase induces tumorigenesisis an increase in mutation rate, then treatment with 5-azaCdR should have resulted in reduced mutagenesis.However, recent data by Jackson-Grusby et al., (1997)suggests that incorporation of 5-Aza-CdR into DNA

increases the rate of mutagenesis when DNA MeTase ispresent in the cell. This data strongly suggests that themechanism by which the DNA MeTase inhibitor inhibitspolyp formation in mice does not involve inhibition ofmutagenesis.

2. Inactivation of tumor suppressors

The second proposed mechanism discussed above isthat hypermethylation results in silencing of "tumorsuppressor" genes (Pokora and Schneider, 1992; Ohtani-Fujita et al., 1993; Merlo et al., 1995; Royer- Merlo etal., 1995;Herman et al., 1995). Although there is solidevidence that DNA methylation is an importantmechanism involved in silencing "tumor suppressors", itis not clear whether methylation of tumor suppressorgenes is a consequence of a programmed change in thelevel of DNA MeTase, such as that occurring in the Y1system. An inherent problem in the "tumor suppressor"model is how can a general increase in DNA methylationresult in site-specific methylation of specific genes. Onepossible model is that additional factors such asthe"imprintors", proposed to function in parentalimprinting of genes, might be involved in translating theincrease in DNA methylation into site specificmethylation events (Szyf, 1991). Alternatively, inductionof DNA MeTase might directly activate cellular regulatorypathways that result in inactivation of tumor suppressorgenes. Following inactivation, the tumor suppressorsundergo methylation.

3. Direct control of cell growth

A third hypothesis is that DNA methylation directlycontrols the progression of the cell cycle (Szyf, 1996) .Recent evidence suggests that hypermethylated CG clustersis a marker of active origins (Rein et al., 1997), and thatmethylation of origins of replication occurs concurrentlywith replication (Araujo et al. unpublished). I havetherefore proposed (Szyf, 1996) that hypermethylationcauses firing of normally silent origins, explaining thechromosomal abnormalities observed in cancer cells. Oneinteresting question is what is the kinetics oftransformation induced by methylation. One hypothesis isthat the increased MeTase is required to maintain randomevents of de novo methylation of tumor suppressor genes.Cells that have acquired these methylations are selected. If this model is true, transformation induced bymethylation should be slow and the number of transformedcells should increase with time. On the other hand ifincreased MeTase levels target central controls of cellgrowth , then transformation by methylation should berapid. Future experiments will most probably resolvethis question.

In summary, I will like to suggest this unifyinghypothesis explaining the involvement of DNAmethylation in cancer. The basic oncogenic programs inthe cell trigger an induction of both DNA MeTase anddMTase activities resulting in hypermethylation of certain


657

sites and hypomethylation of others. The specificity ofthis process is determined by the different affinities of thedifferent sites to these respective enzymes. I suggest thatthe milieu of DNA binding factors in the cell directs theDNA MeTase to growth suppressor sites and the dMTaseto growth stimulating sites. Thus, the coordinateinduction of DNA MeTase and dMTase results in asimultaneous repression of all growth suppressionfunctions and induction of growth activation functions(Fig . 5 ). If a general induction in DNA MeTase activityis indeed responsible for launching this program, theninhibition of the enzyme by pharmacological meansshould direct the cell towards the original program of anontransformed cell.

IV. Therapeutic applications of DNAMeTase inhibitors

A. DNA MeTase inhibitors

An essential step in the developing of newpharmacological concepts is identifying novel targets.DNA MeTase was not considered a pharmacological targetof importance because of the prevalent conception thatinhibitors of DNA methylation might be carcinogenic(Platt, 1995). Because of this prevalent conception, noinhibitors to the cytosine DNA MeTase were developedsince the introduction of 5-azacytidine. 5-azacytidine wasoriginally synthesized as a nucleoside analog and was laterfound to inhibit DNA methylation after its incorporationinto cellular DNA by covalently trapping the DNAMeTase (Wu and Santi, 1985; Jones, 1985). Our basicresearch of the mechanisms involved in regulating DNAMeTase activity and DNA methylation reviewed in theprevious sections introduces the DNA MeTase as animportant and very broad pharmacological target.

1. 5-Aza-CdR a DNA MeTase inhibitor withserious side effects

The only specific inhibitor of DNA MeTase that iscurrently available is 5-Aza-CdR which is phosphorylatedby cellular kinases, incorporated into DNA and traps DNAMeTase molecules by forming a covalent bond with thecatalytic site of the protein (Wu and Santi, 1985). Thismechanism of action results in potential toxicities and sideeffects that limit the utility of 5-azadC as a therapeuticagent as well as a research tool (see review in Szyf, 1996).Recent data suggests that the mutagenicity induced by 5-azaCdR is a consequence of the interaction of DNAMeTase with 5-azaCdR incorporated into the DNA(Juttermann et al., 1994; Jackson-Grusby et al., 1997). Itis clear that new DNA MeTase inhibitors that are notincorporated into DNA should be developed (Fig . 6 ).

2. SAM analogs

S-adenosylhomocysteine, an analogue of SAM andone of the products of the methylation reaction is aninhibitor of DNA methylation (Mixon and Dev, 1983).

Inhibitors of SAH hydrolysis such as periodate-oxidizedadenosine or 3-deazaadenosine analogs (Chiang et al.,1992) were used before as inhibitors of DNA methylation.However, SAH, its analogues and inducers will inhibit alarge number of different methylation reactions in the celland must have nonspecific side effects (Papadopoulos etal., 1987) (Fig . 6 ).

3. Antisense oligonucleotides

We have recently shown that a DNA MeTaseantisense mRNA that is expressed in Y1 tumor cellsinhibits tumorigenesis ex vivo and in vitro (MacLeod andSzyf, 1995). The advent of antisenseoligodeoxynucleotides as specific inhibitors of proteinexpression in vivo offers new opportunities to test thetherapeutic value of inhibition of DNA MeTase as well asto use these as novel therapeutic agents. We haverecently shown that a phosphorothioate-modified antisenseoligodeoxynucleotide directed against the DNA MeTaseinhibits DNA MeTase as well as inhibits the growth oftumors in syngeneic mice in vivo (Ramchandani et al.,1997). These results have now been extended toxenografts of human cancer lines in nude mice. ActiveDNA MeTase antisense compounds that can inhibit humanDNA MeTase mRNA as well as growth of human tumorcell lines in vivo have been identified (MacLeod et al.,unpublished). DNA MeTase antisense oligonucleotides arepotential candidates for anticancer agents in humans (Fig .6 ).

4. Direct inhibitors of DNA MeTase

Antisense oligonucleotides only inhibit de novosynthesis but not the existing DNA MeTase protein.There might be certain situations where the turn-over rateof the enzyme will be too slow. In addition, antisensecompounds are species specific and can not be used inanimal models as well as nonanimal models such as plantswhich will most probably be of significant commercialpotential. It is clear that new inhibitors are required tofully realize the research and applied potential of DNAmethylation.

Other approaches that are now tested in my laboratoryis to use analogs of the CG substrate as direct inhibitors ofDNA MeTase. DNA MeTase is an attractive candidate forDNA based antagonists since in distinction from otherDNA binding protein it forms a covalent transition stateintermediate with the DNA substrate (Wu and Santi,1985). An ideal DNA based antagonists would thereforebind the MeTase but would not be an acceptor for methyltransfer. Thus, a stable complex would be formed betweenthe enzyme and the substrate (Fig . 6 ). Recentunpublished data from our laboratory suggests that someanalogs inhibit DNA MeTase activity at the nanomolarrange and inhibit DNA MeTase and tumor growth inliving cells (Bigey et al., unpublished).

B. Potential side effects of DNA MeTaseinhibitors


658

One important issue that might challenge the utility ofDNA MeTase inhibitors as therapeutics is potential sideeffects resulting from activation of unwanted genes. It isobviously impossible to assess the full systemic effect ofDNA MeTase inhibitors at this stage. However, previousdata as well as our understanding of the mechanisms ofaction of DNA methylation suggest that these effects willnot be a major issue. First, demethylation is insufficientper se to activate genes, the presence of the propertranscription machinery is required. Demethylation willonly activate those genes in the cell that have theappropriate transcription factors available. Second, thenew pattern of methylation generated after demethylationwill be dictated by the legacy of the cell. This is probablywhy extensive demethylation of cell lines results in

activation of the next stage in differentiation rather thanchaotic activation of many possible programs (Szyf et al.,1992). Third, DNA MeTase inhibitors will only have aneffect on dividing cells since they passively inhibitmethylation during replication but do not remove methylgroups from DNA. Fourth, chronic treatment of micewith 5-Aza-CdR (Laird et al., 1995) or DNA MeTaseantisense (Ramchandani et al., 1997) does not result inapparent systemic toxicity.

Acknowledgements

The work discussed in this review was supported bythe MRC, NCIC and contracts with MethylGene Inc. andHybridon Inc.

Figure 6: Inhibitors of DNA MeTase

DNA MeTase could be inhibited at different levels. The first line illustrates a scheme (not to scale) of the first exons andintrons of the DNA MeTase gene. The second and third lines are schemes of mRNAs encoded by the DNA MeTase.

A n t i s e n s e o l i g o n u c l e o t i d e s : Antisense oligonucleotides (line under the mRNA) directed against some of the splicejunctions, reduce DNA MeTase mRNA level (MacLeod et al., unpublished). Two different messages are transcribed from the DNAMeTase gene. Antisense oligonucleotides could be directed against the sequence encoding the ATG translation initiation site ofeach protein specifically. Once the protein is synthesized, it could be inhibited by either of three different ways. Hairpininh ib i tors : First, a modified hairpin oligonucleotide substrate (left) bearing a hemimethylated CG sequence will bind the DNAMeTase and form a stable complex with the substrate. As the modification of the hairpin inhibits the transfer of a methyl groupfrom SAM, the enzyme remains bound to the substrate and is unavailable for methylating genomic DNA.

SAM analogs : SAH and its analogs bind the SAM binding pocket of the DNA MeTase and inhibit methylation.

The first line describes the DNA methylation reaction. A double stranded DNA bearing a methylated C in a CG dinucleotide onthe parental strand and a nonmethylated C in the CG dinucleotide on the nascent strand is reacted with S-adenosyl-methionine(SAM) in a reaction catalyzed by the DNA methyltransferase (DNA MeTase). The resulting products of the reactions are a doublestranded methylated DNA and S-adenosyl homocysteine (SAH).


659

References

Baylin SB, Hoppener JW, de Bustros A, Steenbergh PH,Lips CJ and Nelkin BD ( 1 9 8 6 ) DNA methylationpatterns of the calcitonin gene in human lung cancers andlymphomas. Cancer Res. 46, 2917-2922.

Baylin SB, Makos M, Wu JJ, Yen RW, de Bustros A,Vertino P and Nelkin BD ( 1 9 9 1 ) Abnormal patterns ofDNA methylation in human neoplasia: potentialconsequences for tumor progression. Cancer Cel ls 3 ,383-390.

Becker PB, Ruppert S and Schutz G ( 1 9 8 7 ) Genomicfootprinting reveals cell type-specific DNA binding ofubiquitous factors. Cel l 51, 435-443.

Belinsky SA, Nikula KJ, Baylin SB and Issa J-P ( 1 9 9 6 )Increased cytosine DNA-methyltransferase activity istarget-cell-specific and an early event in lung cancer.Proc. Natl . Acad. Sci . USA 93, 4045-4050.

Bestor TH, Laudano A, Mattaliano R and Ingram V ( 1 9 8 8 )Cloning and sequencing of a cDNA encoding DNAmethyltransferase of mouse cells. The carboxyl-terminaldomain of the mammalian enzymes is related to bacterialrestriction methyltransferases. J. M o l . B i o l . 203,971-983.

Brandeis M, Ariel M and Cedar H ( 1 9 9 3 ) Dynamics of DNAmethylation during development. B i o e s s a y s . 15,709-713.

de Bustros A., Nelkin BD, Silverman A., Ehrlich G, PoieszB and Baylin SB ( 1 9 8 8 ) The short arm of chromosome11 is a "hot spot" for hypermethylation in humanneoplasia. Proc. Nat l . Acad. o f S c i . USA 85,5693-5697.

Chiang PK, Burbelo PD, Brugh SA, Gordon RK, Fukuda Kand Yamada Y ( 1 9 9 2 ) Activation of collagen IV geneexpression in F9 teratocarcinoma cells by 3-deazaadenosine analogs. Indirect inhibitors ofmethylation. J . B io l . Chem. 267, 4988-4991.

Cross SH, Meehan RR, Nan X and Bird A. ( 1 9 9 7 ) Acomponent of the transcriptional repressor MeCP1 sharesa motif with DNA methyltransferase and HRX proteins.Nature Gen. 16, 256-259.

DeSimone J, Heller P, Hall L and Zwiers D ( 1 9 8 2 ) 5-Azacytidine stimulates fetal hemoglobin synthesis inanemic baboons. P r o c N a t l A c a d S c i U S A . 79,4428-4431.

el-Deiry, W. S., Nelkin, B. D., Celano, P., Yen, R. W.,Falco, J. P., Hamilton, S. R. and Baylin, S. B. ( 1 9 9 1 )High expression of the DNA methyltransferase genecharacterizes humanneoplastic cells and progressionstages of colon cancer. Proc . Nat l . Acad . Sc i . USA88, 3470-3474.

Feinberg AP and Vogelstein B ( 1 9 8 3 ) Hypomethylationdistinguishes genes of some human cancers fromtheirnormal counterparts. Nature 301, 89-92.

Feinberg AP, Gehrke CW, Kuo KC and Ehrlich M ( 1 9 8 8 )Reduced genomic 5-methylcytosine content in humancolonic neoplasia. Cancer Res. 48, 1159-1161.

Gruenbaum Y, Naveh-Many T, Cedar H and Razin A ( 1 9 8 4 )Sequence specificity of methylation in higher plant DNA.Nature 292, 860-862.

Herman JG, Latif F, Weng Y, Lerman MI, Zbar B, Liu S,Samid D, Duan DS, Gnarra JR, Linehan WM and Baylin,SB ( 1 9 9 4 ) Silencing of the VHL tumor-suppressor geneby DNA methylation in renal carcinoma. Proc NatlAcad Sci U S A . 91:9700-9704.

Jackson-Grusby L, Laird PW, Magge SN, Moeller BJ andJaenisch R ( 1 9 9 7 ) Mutagenicity of 5-aza-2’-deoxycytidine is mediated by the mammalian DNAmethyltransferase. Proc . Nat l . Acad . Sc i . USA 94,4681-4685.

Jones PA. ( 1 9 8 5 ) Altering gene expression with 5-azacytidine. Cel l 40, 485-486.

Jones PA, Rideout WM 3d, Shen JC, Spruck CH and Tsai YC( 1 9 9 2 ) Methylation, mutation and cancer. B i o e s s a y s14, 33-36.

Jutterman R, Li E and Jaenisch R ( 1 9 9 4 ) Toxicity of 5-aza-2’-deoxycytidine to mammalian cells is mediatedprimarily by covalent trapping of DNA methyltransferaserather than DNA demethylation. Proc. Nat l . Acad.Sci . USA 91, 11797-11801.

Kass SU, Landsberger N, Wolffe AP ( 1 9 9 7 ) DNAmethylation directs a time-dependent repression oftranscription initiation. Curr. Biol . 7, 157-165.

Keshet I, Lieman-Hurwitz J and Cedar H ( 1 9 8 6 ) DNAmethylation affects the formation of active chromatin.Cel l 44, 535-543.

Kautiainen TL and Jones PA ( 1 9 8 6 ) DNA methyltransferaselevels in tumorigenic and nontumorigenic cells in culture.J . B io l . Chem. 261, 1594-1598.

Laird PW, Jackson-Grusby L, Fazeli A, Dickinson S L, JungW E, Li E, Weinberg RA and Jaenisch R ( 1 9 9 5 )Suppression of intestinal neoplasia by DNAhypomethylation. Cel l 81, 197-205.

Lei H, Oh SP, Okano M, Juttermann R, Goss KA, JaenischR and Li E ( 1 9 9 6 ) De novo DNA cytosinemethyltransferase activities in mouse embryonic stemcells. Deve lopment . 122, 3195-3205.

Ley TJ, DeSimone J, Anagnou NP, Keller GH, HumphriesRK, Turner PH, Young NS, Keller P and Nienhuis AW( 1 9 8 2 ) 5-azacytidine selectively increases gamma-globin synthesis in a patient with beta+ thalassemia.New. Engl . J . Med. 307, 1469-1475.

Li E, Bestor TH and Jaenisch R ( 1 9 9 2 ) Targeted mutation ofthe DNA methyltransferase gene results in embryoniclethality. Cel l 69, 915-926.

Liu Y, Sun L and Jost JP. ( 1 9 9 6 ) In differentiating mousemyoblasts DNA methyltransferase isposttranscriptionally and posttranslationally regulated.Nucl . Acids Res . 24, 2718-2722.

Lock LF, Takagi N and Martin GR ( 1 9 8 7 ) Methylation ofthe Hprt gene on the inactive X occurs after chromosomeinactivation. Cel l 48, 39-46.

MacLeod AR and Szyf M ( 1 9 9 5 ) Expression of an antisenseto the DNAmethyltransferase mRNA induces DNAdemethylation and inhibits tumorigenesis. J. B i o l .Chem. 270, 8037-8043.

MacLeod AR, Rouleau J and Szyf M ( 1 9 9 5 ) Regulation ofDNA methylation by the Ras signaling pathway. J .B i o l . C h e m . 270, 11327-11337.


660

Makos M, Nelkin BD, Lerman MI, Latif F, Zbar B andBaylin SB ( 1 9 9 2 ) Distinct hypermethylation patternsoccur at altered chromosome loci in human lung and coloncancer. Proc. Natl . Acad. Sci . USA 89, 1929-1933.

Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, Burger P,Baylin SB and Sidransky D (1995) 5’ CpG islandmethylation is associated with transcriptional silencingof the tumor suppressor p16/CDKN2/MTS1 in humancancers. Nature Med. 1, 686-692.

Meyer P. ( 1 9 9 5 ) DNA methylation and transgene silencingin Petunia hybrida. Curr. Topics M i c r o b i o l .Immunol . 197, 15-28.

Mixon JC and Dev VG ( 1 9 8 3 ) Fragile X expression isdecreased by 5-azacytidine and S-adenosylhomocysteine.Am J Hum Genet. 35, 1270-1275.

Nan X, Campoy FJ, Bird A ( 1 9 9 7 ) MeCP2 is atranscriptional repressor with abundant binding sites ingenomic chromatin. Cel l 88, 471-481.

Peterson K and Sapienza C ( 1 9 9 3 ) Imprinting the genome:imprinted genes, imprinting genes, and a hypothesis.Annu. Rev. Genet . 27, 7-31.

Ohtani-Fujita N, Fujita T, Aoike A, Osifchin NE, RobbinsPD and Sakai T ( 1 9 9 3 ) CpG methylation inactivates thepromoter activity of the human retinoblastoma tumor-suppressor gene. Oncogene 8, 1063-1067.

Papadopoulos V, Kamtchouing P, Drosdowsky MA andCarreau S ( 1 9 8 7 ) Effects of the transmethylationinhibitor S-adenosyl-homocysteine and of the methyldonor S-adenosyl-methionine on rat Leydig cell functionin vitro. J. Steroid Biochem. 26, 93-98.

Platt OS ( 1 9 9 5 ) Sickle cell paths converge on hydroxyurea.Nature Med. 1, 307-308.

Ramchandani S, MacLeod AR, Pinard M, von Hofe E and SzyfM ( 1 9 9 7 ) Inhibition of tumorigenesis by a cytosine-DNA methyltransferase antisense oligodeoxynucleotide.Proc. Natl . Acad. Sci . USA 94, 684-689 .

Razin A and Cedar H ( 1 9 7 7 ) Distribution of 5-methylcytosine in chromatin. P r o c . N a t . A c a d . S c iUSA 74, 2725-2728.

Razin A and Riggs AD ( 1 9 8 0 ) DNA methylation and genefunction. Sc ience 210, 604-610.

Razin A and Szyf M. ( 1 9 8 4 ) DNA methylation patterns.Formation and function. B i o c h i m . B i o p h y s . A c t a .782: 331-342.

Razin A, Feldmesser E, Kafri T and Szyf M ( 1 9 8 5 ) Cellspecific DNA methylation patterns; formation and anucleosome locking model for their function. Prog .C l i n i c . B i o l . R e s . 198, 239-253.

Rein T, Zorbas H and DePamphilis M ( 1 9 9 7 ) Activemammalian replication origins are associated with a high-density cluster of mCpG dinucleotides. M o l . C e l l .B i o l . 17, 416-426.

Rouleau J, Tanigawa G and Szyf M. ( 1 9 9 2 ) The mouse DNAmethyltransferase 5'-region. A unique housekeeping genepromoter. J . B io l . Chem. 267, 7368-7377.

Rouleau J, MacLeod AR and Szyf M ( 1 9 9 5 ) Regulation ofthe DNA methyltransferase by the Ras-AP-1 signalingpathway. J . B io l . Chem. 270,1595-1601.

Royer-Pokora B and Schneider S ( 1 9 9 2 ) Wilms' tumor-

specific methylation pattern in 11p13 detected by PFGE.Genes Chromosom. Cancer 5, 132-140.

Stein R, Razin A and Cedar H ( 1 9 8 2 ) In vitro methylation ofthe hamster adenine phosphoribosyltransferase geneinhibits its expression in mouse L cells. Proc NatlAcad Sci USA 79,3418-3422.

Szyf M, Avraham-Haetzni K, Reifman A, Shlomai J, KaplanF, Oppenheim A and Razin A (1 9 8 4 ) DNA methylationpattern is determined by the intracellular level of themethylase. Proc Nat l Acad Sc i USA 81, 3278-3282.

Szyf M, Schimmer BP and Seidman JG ( 1 9 8 9 ) Nucleotide-sequence-specific de novo methylation in a somaticmurine cell line. Proc . Nat l . Acad . Sc i . U S A 86,6853-6857.

Szyf M Tanigawa G and McCarthy PL Jr. ( 1 9 9 0 ) A DNAsignal from the Thy-1 gene defines de novo methylationpatterns in embryonic stem cells. M o l . C e l l . B i o l .10, 4396-4340.

Szyf M, Bozovic V and Tanigawa G ( 1 9 9 1 ) Growthregulation of mouse DNA methyltransferase geneexpression. J . B io l . Chem. 266,10027-10030.

Szyf M ( 1 9 9 1 ) DNA methylation patterns: an additionallevel of information?. B i o c h e m . C e l l B i o l . 69, 764-767.

Szyf M, Rouleau J, Theberge J and Bozovic V ( 1 9 9 2 )Induction of myogenic differentiation by an expressionvector encoding the DNA methyltransferase cDNAsequence in the antisense orientation. J B i o l Chem.267, 12831-12836.

Szyf M ( 1 9 9 4 ) DNA methylation properties: consequencesfor pharmacology. Trends Pharmacol Sci . 15, 233-238.

Szyf M Theberge J and Bozovic V ( 1 9 9 5 ) Ras induces ageneral DNA demethylation activity in mouse embryonalP19 cells. J . B io l . Chem. 270, 12690-12696

Szyf M ( 1 9 9 6 ) The DNA methylation machinery as a targetfor anticancer therapy. Pharmacol.Ther . 70, 1-37.

Tollefsbol TO and Hutchinson CAIII ( 1 9 9 5 ) MammalianDNA (Cytosine-5-)-methyltransferase expressed inEscherichia coli, purified and characterized J . B i o l .Chem . 270, 18543-18550.

Tucker KL, Talbot D, Lee MA, Leonhardt H and Jaenisch R( 1 9 9 6 ) Complementation of methylation deficiency inembryonic stem cells by a DNA methyltransferaseminigene. P r o c . N a t l A c a d . S c i . U S A 93, 12920-12925.

Wu JC and Santi DV ( 1 9 8 5 ) On the mechanism andinhibition of DNA cytosine methyltransferases. Prog .C l i n i c . B i o l . R e s . 198, 119-129.

Wu J, Issa JP, Herman J, Bassett DE Jr, Nelkin BD andBaylin SB ( 1 9 9 3 ) Expression of an exogenouseukaryotic DNA methyltransferase gene inducestransformation of NIH 3T3 cells. Proc. Nat l . Acad.Sc i . USA 90, 8891-8895

Yisraeli, J., and Szyf, M. ( 1 9 8 4 ) The pattern of methylationof eukaryotic genes. in DNA Methy la t ion:Biochemis try & B i o l o g i c a l S ign i f i cance ,(Razin, A. Cedar, H. & Riggs, A. D., eds) pp 353-378,Springer-Verlag, New York.


661


661


Correlation between DNA methylation andpoly(ADP-ribosyl)ation processes

Giuseppe Zardo°, Stefania Marenzi* and Paola Caiafa°#

°Departments of Biomedical Sciences and Technologies, University of L'Aquila and of *Biochemical Sciences “A.Rossi

Fanelli” University of Rome “La Sapienza", #C.N.R. Centre for Molecular Biology, Rome, Italy.

___________________________________________________________________________________________________

Corresponding author: Dr. Paola Caiafa, Dipartimento di Scienze e Tecnologie Biomediche, Università dell’Aquila; Via Vetoio,

Loc. Coppito, I-67100 L’Aquila, Italy. Tel: (+39) 862-433431; Fax: (+39) 862-433433; E-mail: [email protected]

Summary

The aim of this article is to show the close relationship between DNA methylation and poly(ADP-

ribosyl)ation which are two important nuclear enzymatic mechanisms. An open question i s to

explain how some CpG dinucleotides, in particular those present into GpG islands, can maintain

their unmethylated state in spite of the presence of active DNA methyltransferase in chromatin.

This paper illustrates some data indicating that H1 histone i s a possible trans-acting factor

involved in protecting genomic DNA from full methylation and proposes that the somatic variant

H1e, in its poly(ADP-ribosyl)ated isoform, is the protein capable of undertaking this role.

I. Introduction

A. Inhibitory effect of DNAmethylation on gene expression.

DNA methylation is a specific post-synthetic

modification of DNA that, in eukaryotic cells, appears to

play an important role in the epigenetic modulation of

gene expression. It is the major enzymatic DNA

modification that transfers methyl groups from S-adenosyl

methionine (S-AdoMet) to cytosine (C) and converts these

residues into 5-methylcytosine (5mC) (Bestor and Ingram,

1983). Although in vertebrates the presence of 5mC has

occasionally been reported to be found in dinucleotide

sequences CpC, CpA and CpT (Woodcock et al., 1987,

1988; Toth et al., 1990; Tasheva and Roufa, 1994; Clark

et al., 1995), the best substrate for DNA methyltransferase

is cytosine located in the CpG dinucleotide (Gruembaun et

al., 1981).

As for the distribution of 5mCs, evidence already

existed (Yisraeli and Szyf, 1984) that they are distributed

in a non-random fashion in genomic DNA. Successive

studies have shown that the methylated cytosines are

present in bulk DNA (Bloch and Cedar, 1976) while the

unmethylated ones are essentially located within some

particular DNA regions termed "CpG islands" (Bird et al.,

1985; Bird, 1986,1987), Figure 1 . The specific DNA

methylation pattern results from the combination of

maintenance and de novo methylation and of

demethylation processes, Figure 2 . The maintenance

methylase recognizes and modifies hemimethylated sites

generated during DNA replication thus preserving the

tissue-specific methylation pattern (Razin and Riggs,

1980). In higher eukaryotes this enzymatic process takes

place within a minute or two after replication (Leonhardt

et al., 1992).

The final "correct" methylation pattern is reportedly

obtained, in somatic cells, during the early stages of

embryonic development, through a combination of

demethylation and de novo methylation steps (Brandeis et

al., 1993). Demethylation occurs by an active reaction

(Frank et al., 1991; Brandeis et al., 1993; Jost, 1993;

Weiss et al., 1996) where a 5-methyldeoxycytidine

excision repair system cleaves the DNA strand at 5mCpG

sites, removes the methylcytosine from DNA and replaces

it with cytosine. Subsequently, a burst of de novo

methylation starts the differentiation process leading to a

bimodal pattern of methylation in which the "CpG

islands" at the 5' end of the housekeeping genes remain

constitutively unmethylated, while other genomic

sequences undergo a massive wave of de novo

methylation. Demethylation of individual genes occurs

also during tissue-specific differentiation (Razin et al.,


662

5' XXXCmGXXXCmGXXX 5' XXXCmGXXXCmGXXX

3' XXXGCXXXGCXXX 3' XXXGmCXXXGmCXXX

5' XXXCmGXXXCmGXXX Replication Maintenance Methylation

3' XXXGmCXXXGmCXXX !!!!!" !!!!!!!!!!!!"

5' XXXCGXXXCGXXX 5' XXXCmGXXXCmGXXX

3' XXXGmCXXXGmCXXX 3' XXXGmCXXXGmCXXX

Substrate : hemimethylated DNA

Role : to preserve the tissue specific methylation pattern

When : 1-2 min. after replication.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

5' XXXCGXXXCGXXX De novo methylation 5' XXXCmGXXXCGXXX

3' XXXGCXXXGCXXX !!!!!!!!!" 3' XXXGmCXXXGCXXX

Substrate : unmethylated CpG sequences.

Role : to define the final correct tissue specific methylation pattern involved in the differentiation process,

or repress the active genes in somatic cells.

When : during the early stages of embryonic development, or during carcinogenesis.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

5' XXXCmGXXXCmGXXX Demethylation 5' XXXCmGXXXCmGXXX Demethylation 5' XXXCGXXXCGXXX

3' XXXGmCXXXGmCXXX !!!!!!" 3' XXXGCXXXGCXXX !!!!!!" 3' XXXGCXXXGCXXX

Substrate : fully methylated DNA emimethylated DNA .

Role : to define the final correct methylation pattern, or gene activation in somatic cells.

When : after replication, during the early stages of embryonic development (blastula stage),

or during tissue specific differentiation.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Figure 1 . Processes involved in defining the DNA methylation pattern.

Distribution of CpG and 5'meCpG dinucleotides in eukaryotic DNA.

Bulk CpG island

G+C content

40% 60%

CpG level (CpG/GpC)

0.2 > 0.6

Methylation level

High Unmethylated

Figure 2 . Non-random distribution of CpG and 5mCpG dinucleotides in genomic DNA.

Maintenance DNA methylation

De novo DNA methylation

Active DNA demethylation


663

1986; Brandeis et al., 1993; Jost and Jost, 1994), this

process being probably required for gene activation. To

explain the demethylation process two different

mechanisms have been described.

The first one involves a proteic factor 5-

methylcytosine endonuclease activity that is able to

remove the 5-methylcytosine and to substitute it with

cytosine (Jost, 1993; Jost and Jost, 1994; Jost et al.,

1995). The second one involves the presence of a

ribozyme or maybe a ribozyme associated with a proteic

factor that is able to remove the mCpG dinucleotide and to

substitute it with CpG dinucleotide (Weiss et al., 1996).

The fact that CpG dinucleotides are present in an

unmethylated state in "CpG islands" is of interest since

their frequency in them is five times more than in bulk

DNA, Figure 1 .

As far as the correlation between DNA methylation

and gene expression is concerned, the "CpG islands", that

go from 500-2000 base pairs in size, are usually found in

the 5' promoter region of housekeeping genes and overlap

genes to variable extents (Bird, 1986). There is evidence

that transcription of genes, correlated with "CpG islands",

is inhibited when these regions are methylated (Keshet et

al., 1985).

That the "CpG islands" are not by themselves

unmethylable is demonstrated by in vitro experiments

(Carotti et al., 1989; Bestor et al., 1992). A great deal of

investigation has been and is performed in order to clarify

why the "CpG islands" remain untouched by the action of

DNA methyltransferase (Ysraeli and Szyf, 1984) in spite

of their localization on promoter region of housekeeping

genes which are, in decondensed chromatin, permanently

accessible to the transcriptional factors.

A question yet to be solved is to identify different cis-

acting signals and trans-acting protein factors that may

play a key role in defining the bimodal pattern of

methylation involved in cell differentiation and gene

expression. It has been suggested that the density of CpG

dinucleotide inside "CpG-islands" could be per se a signal

involved in protecting the unmethylated state of these

DNA regions (Frank et al., 1991) but further experiments

suggest that there are some sequence motifs that are

intrinsically protected against de novo methylation (Szyf

et al., 1990; Christman et al., 1995; Tollefsbol and

Hutchinson, 1997) and/or that there are some cis-acting

"centers of methylation" capable of preventing the

methylation pattern of flanking DNA sequences (Szyf et

al., 1990; Szyf, 1991; Mummaneni et al., 1993; Brandeis

et al., 1994; Hasse and Schultz, 1994; MaCleod et al.,

1994; Magewu and Jones, 1994; Mummaneni et al.,

1995). The simple possible explanation that there are

trans-acting protein factors associated with "CpG islands"

which prevent access to those DNA regions, has been

difficult to demonstrate up to now. Research on the

identification of factors able to link methylated DNA has

met with greater success. The first protein identified as

able to bind methylated DNA sequences is the methylated

DNA binding protein (MDBP) a ubiquitous family of

closely related proteins in vertebrates (Huang et al., 1984;

Zhang et al., 1993). Its consensus sequence is composed

of 14 bp and has a substantial degree of degeneration.

MDBP sites can have up to 3 CpG's and generally the

degree of binding increases when more of these are

methylated and when they are methylated on both strands.

In spite of the high level of degeneracy of the consensus

sequence, MDBP can be considered sequence-specific in its

binding because mutations in inopportune positions cause

this protein to lose its ability to link the sequence. This

protein can also link the consensus sequence independently

of its methylated level provided the C is substituted by T

and T is present in TpG or TpA dinucleotides (Zhang et

al., 1986; Khan et al., 1988). It is possible that the

consensus methylation independent sequences could derive

from the spontaneous deamination of 5mC in T. From

transfection experiments a role of down-regulation of gene

expression has been proposed for this protein (Asiedu et

al., 1994; Zhang et al., 1995). The MDBP-2-H1 is

another protein that binds itself preferentially to certain

DNA sequences containing a simple mCpG pair (Pawlak

et al., 1991). Although the protein is not sequence

specific, its affinity for the consensus sequence is highest

in the promoter region (+2 +32) of vitellogenin II gene

where it plays an down-regulatory role of gene expression

(Pawlak et al., 1991). Further investigations have shown

(Jost and Hofsteenge, 1992) that this protein - identified as

H1 histone-like - must undergo phosphorylation before to

its interaction with the methylated DNA sequence (Bruhat

and Jost, 1995). Two other proteins named MeCP1 and

MeCP2, that have the ability to link DNA regions in

which the CpG dinucleotides are methylated to higher or

lower levels, have been proposed as proteins involved in

the silencing of gene expression (Meehan et al., 1989;

Boyes and Bird, 1991; Lewis et al., 1992). In particular

MeCP1, whose molecular weight is of about 800 KDa, is

suggested to be involved in a mechanism through which

its association with methylated DNA could prevent the

linkage of transcription factors in these DNA regions.

This protein binds sequences containing about 12 or more

methylated CpGs and the enrichment in CpG dinucleotides

argues that these DNA regions are "CpG island-like"

(Meehan et al., 1989).

The strength of promoter and the density of mCpGs

(Boyes and Bird, 1992) are two factors which regulate the

association of MeCP1 with DNA. It is clear that low

levels of methylation can repress transcription of a weak

promoter but not of a strong promoter. In fact, sparsely

methylated genes bind MeCP1 weakly and the


664

transcription is partially repressed if the gene promoter is

weak while if the promoter is strong the gene is expressed

(Boyes and Bird, 1992). The MeCP2 factor is able to bind

DNA that contains a single mCpG pair (Lewis et al.,

1992; Meehan et al., 1992). MeCP2, for which a

transcriptional repressor role has been described, is very

abundant in bulk vertebrate genomic DNA - 100 times

more abundant than MeCP1 - where it is in competition

with H1 histone. This result supports the hypothesis that

MeCP2 is involved in condensing chromatin structure

(Nan et al., 1997).

Although these proteins play an important role in

mediating the methylation-dependent repression of genes,

an open question to answer is how the CpG moieties of

the "CpG islands", become vulnerable or resistant to the

action of DNA methyltransferase and can thus lose or

maintain their characteristic pattern of methylation.

This is the goal of our research: our aim is to identify

and pinpoint a nuclear protein trans-acting factor directly

involved in maintaining the unmethylated state of "CpG

islands".

II. H1 histone and DNA methylation.

A. Methylation-dependent binding ofH1 histone to DNA.

An attractive hypothesis to explain the repressive

effects of DNA methylation on gene expression is that H1

histone binds itself preferentially to DNA sequences

containing mCpG dinucleotides.

Although H1 histone is mainly present in highly

methylated condensed chromatin there is ample

disagreement in the scientific literature - at variance from

the other above mentioned proteins - as to whether or not

its presence is dependent on the methylated state of DNA.

A preference of H1 histone for double-stranded DNA with

a relatively high abundance of methylated CpGs has

however been recently shown by McArthur and Thomas

(1996), who have suggested that the condensing ability of

H1 histone could thus be favored by the higher level of

DNA methylation existing in transcriptionally inactive

chromatin. Parallel experiments (Caiafa et al., 1995;

Reale et al., 1996) have been performed in order to

examine whether in oligonucleosomal DNA, purified from

inactive chromatin fraction, an increased methylation of

CpG residues would interfere with the formation of the

appropriate H1-H1 interactions critical for attainment of

folded chromatin structures. Conflicting results respect to

those of McArthur and Thomas (1996) were obtained since

the introduction of new methyl groups into

oligonucleosomal DNA was surprisingly found to decrease

its ability to allow these H1-H1 interactions (Figure 3),

suggesting that, in vivo, the presence of some

unmethylated CpGs in linker DNA is likely to be an

important prerequisite for chromatin compaction. These

differences could be explained by differences in the DNAs

selected for the two experiments as, despite the common

aim of avoiding sequence-specific effects in H1-DNA

binding, there are indeed considerable differences in terms

of CpG frequency and of the overall methylation level of

the DNAs. The DNA sequences used by McArthur and

Thomas (1996), chosen as representative of a large region

of the sea urchin genome, are essentially obtained from

unmethylated CpG-rich DNA regions, while our

oligonucleosomal DNA was extracted from human

placenta inactive chromatin fraction whose relatively

scarce CpG moieties have a rather high basal methylation

level.

The band shift assays did not solve the problem of

methylation dependent binding of H1 histone to DNA. In

fact experiments carried out using DNA fragments with

different amounts of CpGs dinucleotides, failed to show

any effect of CpG methylation on H1 histone binding

since H1 histone has shown an identical affinity for either

methylated or non-methylated DNA (Campoy et al.,

1995). It may be recalled that while Higurashi and Cole

(1991) have also found that the interaction of H1 histone

with CCGG is independent of the methylation level,

Levine et al. (1993) have shown a preferential binding of

total H1 histone to plasmid methylated DNA.

B. Inhibitory effect of H1 histone onin vitro DNA methylation.

In our research on a nuclear proteic factor involved in

DNA methylation process, we focused our attention on

histone proteins since previous papers have reported a

possible inhibitory role played by histones on DNA

methylation (Kautiainen and Jones, 1985; Davis et al.,

1986).

Our experiments (Caiafa et al., 1991) have shown that

the ability of total histones to affect in vitro enzymatic

DNA methylation was essentially due to a single H1

histone that, in the "physiological" range (0.3:1, w/w)

histone:DNA ratio, was the only one able of exerting a

consistent (90%) inhibition on methylation of double

stranded DNA, catalyzed by human placenta DNA

methyltransferase. Neither H1-depleted preparations of

"core" histones nor, separately, any other single histone

(H2a, H2b, H3) were able to affect the methylation

process, Figure 4 .

Since H1 is known to be preferentially associated to

linker DNA (van Holde, 1988) its ability to suppress in

vitro DNA methylation is consistent with previous


665

Figure 3A : SDS-PAGE patterns of H1 histone after treatment, in the presence of native (lanes 1, 2, 3) or of artificially

overmethylated oligonucleosomal DNA (lanes 4, 5, 6), with dithiobis-(succinimidylpropionate) at different H1:DNA ratios -- 0.1,

0.3, 0.5 (w/w) -- in 40 mM NaCl. In lane 7, H1 histone treated with DSP in the absence of DNA; in lane 8, untreated H1 histone.

(B ): Electrophoretic patterns, in 1% agarose stained with ethidium bromide, of glutaraldehyde-fixed H1-DNA complexes, formed

in 40 mM NaCl at H1:DNA ratios ranging from 0.1 to 0.9 (w/w), using native oligonucleosomal DNA (left panel) or artificially

overmethylated DNA (right panel). DNA molecular marker III from Boehringer is in lane III. Naked DNA controls (native in the

left panel, artificially overmethylated in the right one) are in lanes C and C1. "Reprinted from B i o c h e m . B i o p h y s . R e s .

Comm. 227 , Reale et al.. H1-H1 Cross-linking efficiency depends on genomic DNA methylation, 768-774, (1996) with kind

permission of Academic Press, Inc."

Figure 4 . Effect, on the in vitro activity of

human placenta DNA methyltransferase, of

histones H1, H2a, H2b and H3 (from calf

thymus) renaturated by progressive dialysis at

decreasing urea and NaCl concentrations in the

presence (closed triangles) or absence (closed

circles) of 5 mM EDTA. Each point represents

the mean result of at least five different

experiments in triplicate, S.D. "Reprinted from

B i o c h i m . B i o p h y s . A c t a 1 0 9 0 , Caiafa

et al. Histones and DNA methylation in

mammalian chromatin. I° Differential

inhibition by histone H1, 38-42, (1991) with

kind permission of Elsevier Science Publishers

- NL Sara Burgerhartstraat 25, 1055 KV

Amsterdam, The Netherlands".


666

findings of higher 5mC levels in nucleosomal core DNA

as compared to linker DNA (Razin et al., 1977; Solage

and Cedar, 1978; Adams et al., 1984; Caiafa et al., 1986).

Some experiments were carried out to assess whether the

observed hypomethylation of linker DNA sequences reflect

an intrinsic deficiency in CpG dinucleotides or whether the

well-documented association between DNA and H1

histone causes a local inhibition of enzymatic DNA

methylation process (D’Erme et al., 1993).

The net level of methyl-accepting ability of CpG

dinucleotides in linker DNA - defined as the DNA region

which can be hydrolyzed by staphylococcal nuclease

digestion of H1-depleted oligonucleosomes - was evaluated

by making use of a number of distinct experimental

strategies in order to minimize possible artefacts. Since

the removal of H1 histone by two alternative procedures

yielded quite similar results, it is unlikely that artefactual

nucleosome sliding may have significantly altered the

regions of chromatin DNA accessible to methylation. In

the first set of experiments we measured the proportion of

labelled methyl groups remaining in the "100 bp minicore

particles" upon extensive staphylococcal nuclease

digestion of in vitro methylated H1-depleted

oligonucleosomes. As shown in Figure 5a,b - where

H1 had been taken away, respectively from

oligonucleosomes and from nuclei by two alternative

procedures - nuclease treatment removed the majority

(85% in one case, 75% in the other) of the labelled 5-

methylcytosine residues. By contrast, nuclease digestion

removed from native oligonucleosomes (H1-containing)

only a relatively small portion of the 5-methylcytosine

residues which had been inserted by in vitro enzymatic

DNA methylation, Figure 5c.

a- Res idual methyl groups in 100 bp “minicore” part ic les af ter nuclease digest ion of methylated

H1-depleted o l igonucleosomes ( treated with 0 .6 M NaCl):

H1 - depleted

oligonucleosome

s

!!" in vitro

methylation

!!" staphylococcal

nuclease digestion!!" “minicore” particles

(100 bp)

#

methyl - 3H incorporated

(2480 dpm/10µg DNA = 100%)

#


(390 dpm/10µg DNA = 15.6%)

b- Residual methyl groups in the nuclease-resistant fraction from methylated H1-depleted nuclei ( treated

at low pH) :

H1 - depleted

nuclei

in vitro

methylation

staphylococcal

nuclease digestion

nuclease-resistant

fraction


(717 dpm/10µg DNA = 100%)


(174 dpm/10µg DNA =

24.3%)

c - Residual methyl groups in 145 bp “core” part ic les after nuclease digest ion of methylated

n a t i v e o l i g o n u c l e o s o m e s :

native

oligonucleosomes

in vitro

methylation

staphylococcal

nuclease digestion

“core” particles

(145 bp)


(1215 dpm/10µg DNA = 100%)


(716 dpm/10µg DNA = 58.9%)

Figure 5 . Evaluation, by three distinct experimental strategies involving nuclease digestion after in vitro methylation, of

the distribution of methyl-accepting CpGs in the nuclease-sensitive fraction. The data obtained refer to a similar set of

experiments run in parallel, so as to obtain comparable values. Five other similar experiments gave slightly different results in

terms of absolute incorporation of methyl groups, but almost identical as percent radioactivity values remaining in the nuclease-

resistant fractions. "Reprinted from B i o c h i m . B i o p h y s . A c t a 1 1 7 3 , D'Erme et al.. Inhibition of CpG methylation in linker

Zardo et al: DNA methylation and poly(ADP-ribosyl)ation

667

DNA by H1 histone, 209-216, (1993) with kind permission of Elsevier Science Publishers - NL Sara Burgerhartstraat 25, 1055 KV

Amsterdam, The Netherlands".

a - Direct methylat ion of 100 bp “minicore” part ic les vs H1 - depleted o l igonucleosomes :

H1 - depleted

oligonucleosomes

staphylococcal nuclease

digestion

“minicore” particles

(100 bp)

in vitro

methylation

in vitro

methylation


(3380 dpm/10µg DNA = 100%)

methyl -3H incorporated


10.1%)

b - Methylat ion of purif ied DNA’s from 145 bp “core” part ic les and from 100 bp “minicore” part ic les vs

o l igonuc leosomal DNA :

native

oligonucleosomes


digestion


(145 bp)

H1 - depleted

oligonucleosomes


digestion

“minicore” particles

(100 bp)purified DNA from


(145 bp)

purified DNA from

oligonucleosomes

purified DNA from

“minicore” particles in vitro

methylation

in vitro

methylation

in vitro

methylation methyl - 3H incorporated


25.9%)



100%)

methyl - 3H

incorporated


24.3%)

Figure 6 . Evaluation, by two distinct experimental strategies involving nuclease digestion before in vitro methylation, of

the distribution of methyl-accepting CpGs in the nuclease-sensitive fraction. "Reprinted from B i o c h i m . B i o p h y s . A c t a

1 1 7 3 , D'Erme et al.. Inhibition of CpG methylation in linker DNA by H1 histone, 209-216, (1993) with kind permission of

Elsevier Science Publishers - NL Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands".

Histone proteins added

none H1

(0.3 mg/mg DNA

“core” histones

(1.0 mg/mg DNA

H2a

(1.0 mg/mg DNA

Number of experiments: n=6 n=6 n=3 n=3

Native oligonucleosomes

H1-depleted oligonucleosomes

Purified DNA from oligonucleosomes

48.4±0.7

100.0

155.0±2.1

-

58.0±1.3

41.6±0.8

-

101.4±3.5

153.8±5.8

-

102.6±2.7

-

Table 1 . Inhibition by H1 of the methyl-accepting ability of oligonucleosomal DNA. The incorporation of labeled methyl

groups in the DNA of H1-depleted oligonucleosomes is made equal to 100 and all the other results obtained in a same set of

experiments are referred to this value.


668

In a complementary approach, when the "100 bp

minicore particles" obtained by digestion with

staphylococcal nuclease of H1-depleted oligonucleosomes

were used as substrates for subsequent in vitro

methylation, their methyl-accepting ability was found to

be, on a DNA basis, only one-tenth of that of the original

H1-depleted oligonucleosomes, Figure 6a . By assaying

the susceptibility to methylation of the purified DNAs

from the same particles, the methyl-accepting ability of

oligonucleosomal DNA was four times larger than that of

either the "145 bp core particles" or of the "100 bp

minicore particles" Figure 6b.

These data (D’Erme et al., 1993) have shown that the

lower level of DNA methylation in linker regions than in

"core" particles (Razin and Cedar, 1977; Solage and Cedar,

1978; Adams et al., 1984; Caiafa et al., 1986) was not

due to an intrinsic CpG deficiency of linker DNA, which

was, in H1-depleted oligonucleosomes, susceptible to

extensive in vitro methylation, but can rather be ascribed

to the inhibition exerted by H1 histone on the process of

enzymatic DNA methylation (Caiafa et al., 1991), which

would occur in these linker DNA regions because of their

preferential association with H1 histone.

The ability and the specificity of H1 histone to inhibit

CpG methylation in linker DNA were assayed by re-

adding purified H1 to H1-depleted oligonucleosomes or to

the DNA purified from them, Table 1 . H1-depletion

doubled the methyl-accepting ability of oligonucleosomes,

with a further 50% increase as the remaining proteins were

also removed. Addition of H1, in a protein-to-DNA (w/w)

ratio of 0.3, reduced the incorporation of labelled methyl

groups in H1-depleted oligonucleosomes and in the

purified oligonucleosomal DNA to the same level

occurring in native oligonucleosomal particles. This

inhibition was paralleled by a re-condensing effect

occurring upon addition of H1 to H1-depleted

oligonucleosomes, as shown in Figure 7 . Both

phenomena are apparently specific to H1 histone, since

they could not be obtained by addition of other histones or

of serum albumin up to a 1:1 protein/DNA (w/w) ratio.

These experiments and others previously performed by

Davis et al. (1986) have shown that enzymatic DNA

methylation is not entirely suppressed by the intrinsic

presence of H1 histone.

The hypothesis of a competition (Santoro et al., 1993)

between the enzyme and histone H1 for some common

DNA negative control of H1 histone was investigated and

disproved by performing experiments in which increasing

amounts of purified DNA methyltransferase were added to

Micrococcus luteus ds-DNA in the presence of a constant

amount of H1 histone, the H1/DNA ratio being fixed to

its "physiological" value of 0.3. As shown in Figure 8 ,

the enzymatic DNA methylation in vitro was independent

of the H1 to enzyme ratio. It seems therefore unlikely, at

least in our experimental conditions, that competition

between the histone and enzyme for some common DNA

binding site(s) is the main mechanism regulating the

incorporation of methyl groups in the CpG sequences of

chromatin. The methyl-accepting ability of intact

oligonucleosomes was, on the other hand, far from

negligible. Although it underwent a two-fold increase

upon H1 histone depletion (with a further 50% increment

if also all other proteins were removed), it went back,

indeed, to the same level as in native chromatin when

excess H1 histone was added to H1-depleted

oligonucleosome preparations (Table 1).

Other two hypotheses can account for these results: the

presence of some particular variant(s) more or less capable

of inhibiting enzymatic DNA methylation and/or the

presence of DNA regions escaping the negative control of

H1 histone.

Figure 7 . CD spectra, in the region of DNA

chromophores, of native (__) and H1-depleted

oligonucleosomes (---) and re-condensing effect occurring

upon addition to the H1-depleted oligonucleosomes of "core"

histones (protein/DNA ratio, w/w=1) or of H1 histone

(protein/DNA ratio, w/w = 0.1: -.-.- ; w/w = 0.2:-o-o-o-).

Oligonucleosomes were suspended, at a DNA concentration of

60 mg/ml, in a 60 mM NaCl, 5 mM Tris-HCl buffer (pH 7.4).

"Reprinted from B i o c h i m . B i o p h y s . Acta 1 1 7 3 ,

D'Erme et al.. Inhibition of CpG methylation in linker DNA

by H1 histone, 209-216, (1993) with kind permission of

Elsevier Science Publishers - NL Sara Burgerhartstraat 25,

1055 KV Amsterdam, The Netherlands".


669

III. H1 histone somatic variants.

The hypothesis that some particular variant could be

specifically involved in the in vitro inhibition of

enzymatic DNA methylation stems from the fact that H1

histone is composed of a family of different somatic

variants termed H1a, H1b, H1c, H1d and H1e (Cole,

1987). They all have a three domain structure, with a

highly conserved central globular domain (98% identity in

80 aa sequence). The differences between the variants are

located in the N-terminal and C-terminal tails, which

consist of about 40 and 100 amino acids respectively

(Cole, 1987), with the overall variation in molecular mass

being approx 1.0-1.4 kDa.

Figure 8 . Variations in the extent of Micrococcus

luteus ds DNA methylation in vitro, as a function of added

DNA methyltransferase, in absence (open circles) or presence

(open triangles) of H1 histone at a constant histone-to-DNA

ratio equal to 0.3 (w/w). "Reprinted from B i o c h e m .

B i o p h y s R e s . C o m m . 1 9 0 , Santoro et al.. Effect of H1

histone isoforms on the methylation of single- or double-

stranded DNA, 86-91, (1993) with kind permission of

Academic Press, Inc."

The number and relative amounts of these variants

differ in various tissues and species throughout the

development stages of the organism and in neoplastic

systems (Liao and Cole, 1981a,b; Pehrson and Cole,

1982; Lennox and Cohen, 1983; Huang and Cole, 1984;

Lennox, 1984; Cole, 1987; Davie and Delcuve, 1991;

Baubichon-Cortay et al., 1992; Giancotti et al., 1993;

Schulze et al., 1993; De Lucia et al., 1994), so that they

may play different roles in chromatin organization, with a

non-random distribution.

A. Tight correlation between H1evariant and the inhibition of DNAmethylation.Some experiments were performed to verify whether or

not these variants could differ from each other in their

ability to exert a negative control on DNA methylation.

Calf thymus H1 histone somatic variants were purified by

reverse phase HPLC, the protein components in effluent

composition being characterized by SDS/slab gel

electrophoresis in 15% (w/v) polyacrylamide. As shown

in Figure 9 , only a restricted number of fractions,

eluting as a single peak ("p3") was able to cause over 80%

inhibition while the other fractions, namely "p1" and "p2"

were totally ineffective (Santoro et al., 1995). The

SDS/PAGE characterization of the various fractions

indicated, according to Lennox et al. (1984) and to Lindner

et al. (1990) the presence in "p1" of H1a, in "p2" of H1d

and in "p3" of H1e and H1c.

Having not yet achieved a satisfactory separation of H1e

and H1c, we managed to purify H1c and H1e (Zardo et al.,

1996) in order to individuate which variant is really

involved in the inhibition of DNA methylation process. A

good separation in four peaks was obtained when H1

histone from L929 mouse fibroblasts was purified. The

HPLC retention time of each peak, combined with the

electrophoretic mobility of various bands, allowed us to

identify the H1a, H1b, H1e and H1c variants. When the

H1e vs H1c variants were assayed for their effect on in

vitro DNA methyltransferase activity, only H1e was

effective in causing a marked inhibition, at H1:DNA

"physiological" ratio, Figure 10 , so that it can be

concluded that H1e is the unique variant involved in the

inhibition of the DNA methylation process.

B. H1e: the only one variant able tobind the "CpG-rich" sequences.

Gel retardation assays were carried out in order to test

the affinity of the different H1 variants for various

synthetic oligonucleotides which varied in terms of their

sequence and of the relative abundance in methylated or

unmethylated CpGs with respect to NpGs (i.e. to all

dinucleotide sequences having G as their second moiety).

As a representative of genomic DNA we also used a 145

bp DNA prepared by digestion of human placenta

chromatin with Staphylococcus aureus nuclease.

Experiments have shown (Santoro et al., 1995) that

among H1 histone somatic variants, the H1a variant was

able to bind a 145 bp genomic DNA fragment but was

unable to bind 44 bp ds-oligonucleotides containing two

or more CpG dinucleotides. The other variants were

capable of binding sequences containing up to three CpGs,

while the fraction H1e-c was unique in binding CpG rich


670

DNA sequences. Later, using H1e and H1c purified

variants, we assessed that H1e variant binds itself better

than H1c to the 6CpG oligonucleotide, Figure 11 .

Our experimental data underline two important

characteristics of H1e variant: this is the only variant

which suppresses enzymatic DNA methylation and it is

the only variant able to bind itself to CpG-rich sequences.

Figure 9 . Separation and characterization of calf

thymus H1 histone variants and their effect on in vitro DNA

methylation: a) elution profile from the RP-HPLC column; b )

SDS gel electrophoresis of all protein fractions, evidenced by

Coomassie Brilliant Blue; c ) effect of total H1 histone ("t")

and of the various fractions eluted from the RP-HPLC column,

at a protein/DNA ratio equal to 0.2 (w/w), on the in vitro

activity of human placenta DNA methyltransferase. Each

point represents the average results of ten different

separations by RP-HPLC. "Reprinted from Biochem. J .

3 0 5 , Santoro et al.. Binding of histone H1e-c variants to

CpG-rich DNA correlates with the inhibitory effect on

enzymatic DNA methylation, 739-744, (1995) with kind

permission of Portland Press"

Figure 10 . Separation and characterization of H1e and

H1c variants from L929 fibroblasts and their effect on in

vitro DNA methylation: a) HPLC separation of H1 histone

variants and electrophoretic pattern, in 12% SDS-

polyacrylamide gel of the eluted fractions (upon visualization

by silver staining). b ) Inhibition of DNA methyltransferase

activity by H1e (open circles) or H1c (closed circles), at

different protein-to-DNA ratios. "Reprinted from B i o c h e m .

B i o p h y s R e s . C o m m . 2 0 , Zardo et al.. Inhibitory effect

of H1e histone somatic variant on in vitro DNA methylation

process, 102-107, (1996) with kind permission of Academic

Press, Inc."


671

Figure 11 . Binding of H1e (open circles) and H1c (closed

circles) to 44 bp synthetic 6CpG duplex oligonucleotide with

the cytosines in the CpG moieties in unmethylated state. The

binding was evaluated by gel retardation after incubation of

the H1e and the H1c variants with the appropriate

oligonucleotide, the relative amount of free DNA being

measured by densitometric scanning of the autoradiograms.

"Reprinted from B i o c h e m . B i o p h y s R e s . C o m m . 2 0 ,

Zardo et al.. Inhibitory effect of H1e histone somatic variant

on in vitro DNA methylation process, 102-107, (1996), with

kind permission of Academic Press, Inc."

IV. Why poly(ADP-ribosyl)ation wasselected out of all H1 histone post-synthetic modifications.

To explain how H1 histone could be involved in

playing many multiplex very important structural and

functional roles in chromatin it is important to remember

that everyone of the genetic somatic variants can be

dynamically modified by different post-synthetic

enzymatic reactions (Wu et al., 1986; Davie, 1995) and

sometimes the same protein can be substrate for more

than one modification. Only in this way can we consider

H1 histone as a protein characterized by a big

macroheterogeneity that allows different possible

interactions with DNA or with other proteins. Some

experimental data have led us to focus our attention on the

poly(ADP-ribosyl)ation process.

A starting point derived from our results showing that

when H1e is poly(ADP-ribosyl)ated it loses its

condensing effect on chromatin structure even though it

remains associated with linker DNA (D’Erme et al.,

1996), Figure 12 . Taking into account polyADP-ribose

dependent chromatin decondensation (Poirier et al., 1982;

Aubin et al., 1983; D’Erme et al., 1996), we considered

the possibility that this modification may alter the

interaction of H1 histone with linker DNA, causing a

change in the methyl-accepting ability of CpG

dinucleotides present essentially in their unmethylated

form on linker DNA. Our aim was, therefore, to compare

the methyl-accepting ability of native nuclei with that of

nuclei in which chromatin decondensation was induced by

poly(ADP-ribosyl)ation. Figure 13A shows the

incorporation of ADP-ribose polymers into H1 histone

during the experimental time and that at the same time the

methyl-accepting ability was not increased in the

decondensed chromatin structure induced by the poly(ADP-

ribosyl)ation process, Figure 13B . These data suggest

that the poly(ADP-ribosyl)ated H1 histone has not been

removed from linker DNA, despite possible alterations in

the H1-DNA interactions and that, even if poly(ADP-

ribosyl)ation decrease the H1e-H1e interactions that are

essential for the formation of the higher levels of

chromatin structure, the poly(ADP-ribosyl)ated isoform of

H1e could be present in decondensed chromatin structure

where the housekeeping genes are located.

The second starting point was the observation that the

demethylation process utilizes an excision-repair

mechanism to remove 5-methylcytosine. Since it is

known that the poly(ADP-ribosyl)ation of H1 histone

plays a relevant role in the repair mechanism (Boulikas,

1989; Realini and Althaus, 1992; Malanga and Althaus,

1994) H1 histone in its poly(ADP-ribosyl)ated isoform

could indeed, following the demethylation process, remain

bound to demethylated regions and regulate the de novo re-

methylation process that defines the methylation pattern

where the "CpG islands" are in an unmethylated state.

V. Correlation between DNAmethylation and poly(ADP-ribosyl)ationprocesses.

A. Poly(ADP-ribosyl)ation process.

Poly(ADP-ribose) polymerase (EC 2.4.2.30) is a

nuclear enzyme that has been implicated in a number of

important biological processes (Jacobson and Jacobson,

1989; de Murcia et al., 1995). Although poly(ADP-

ribose) polymerase is able to bind undamaged DNA, it

needs DNA strand breaks for its activation. Each monomer

of this enzyme, which is a dimer in its catalytic form

(Mendoza-Alvarez and Alvarez-Gonzales, 1993), has three

domains which play specific roles in the poly(ADP-

ribosyl)ation process. The zinc finger motifs in the N-

terminal domain are responsible for the DNA recognition

site, taking advantage of DNA strand breaks rather than of

specific polynucleotide sequences (Ménissier de Murcia et

al., 1989; Gradwohl et al., 1990; Ikejma et al., 1990; de


672

Figure 12. A ) Cross-linking analysis to investigate the role played by each H1 histone variant on the formation of H1-H1

polymers: SDS-PAGE patterns of H1 histone variants, at 30% (w/w) H1:DNA ratio, incubated with 1.2 kb oligonucleosomal DNA

in 40 mM NaCl for 1 hour at room temperature and then treated with dithiobis(succinimidyl)propionate (DSP 0.2 mg/ml) for 20

min: H1a, H1b, H1e, H1c (lane1-4). In lanes 5 and 6, untreated histone H1 and histone H1 treated with DSP were run as controls in

the absence of DNA and B ) the effect of the "enriched" poly-ADP-ribosylation of H1e variant, vs the native one, on the formation

of H1-H1 polymers: SDS-PAGE patterns of the product of cross-linking of the H1e histone isoforms at different (w/w) H1:DNA

ratio, incubated with 1.2 kb oligonucleosomal DNA, in 40 mM NaCl for 1 hour at room temperature and then treated with

dithiobis(succinimidyl)propionate (DSP 0.2 mg/ml) for 20 min: 30%, 20% and 10% (w/w) of H1e:DNA (lane1-3); 30%, 20% and

10% (w/w) of "enriched" poly(ADP-ribosyl)ated H1e:DNA (lane 4-6). "Reprinted from B i o c h e m . J . 3 1 6 , D'Erme et al.. Co-

operative interactions of oligonucleosomal DNA with the H1e histone variant and its poly(ADP-ribosyl)ated isoform, 475-480,

(1996) with kind permission of Portland Press".

Figure 13 . Methyl-accepting ability as assay to study the interactions of H1 histone to linker DNA in native nuclei vs

poly(ADP-ribosyl)ated ones. A): time course of incorporation of [32 P] ADPribose polymers associated to H1 histone extracted by

10% PCA (w/v) from nuclei incubated with 50 µM [32 P]-NAD; B ): methyl-accepting ability of native nuclei (open circles) vs

poly(ADP-ribosyl)ated ones (closed circles). "Reprinted from B i o c h e m . J . 3 1 6 , D'Erme et al.. Co-operative interactions of


673

oligonucleosomal DNA with the H1e histone variant and its poly(ADP-ribosyl)ated isoform, 475-480, (1996) with kind

permission of Portland Press".

Murcia and Ménissier de Murcia, 1994). The C-terminal

domain contains the catalytic site (de Murcia et al., 1995).

As for the central domain, it undergoes automodification

upon binding of the enzyme on the damaged DNA by

introducing ribose polymers -- up to 200 residues

according to Alvarez-Gonzalez and Jacobson (1987) -- on

28 automodification sites (Kawaichi et al., 1981;

Desmarais et al., 1991) which are essentially localized in

this domain.

The active enzyme can then start a series of

heteromodification reactions that modulate the functions

of chromatin proteins (Ferro et al., 1983; Yoshihara et al.,

1985; Boulikas, 1989; Scovassi et al., 1993).

In vitro experiments have shown that this poly(ADP-

ribosyl)ation mechanism can involve H1 histone binding

polymers both in a covalent and in a non-covalent

manner. The covalent modification introduces in the C and

N-terminal tails of this histone short polymers (8-10

units), whose sizes are specifically defined by the histone

itself (Naegeli and Althaus, 1991), while long branched

polymers of ADP-ribose are able to form non-covalent

interactions with this chromatin protein (Panzeter et al.,

1992).

B. Effect of poly(ADP-ribosyl)atedH1 histone on in vitro DNA methylation.

The aim of these experiments was to examine, in vitro

the possible correlation between DNA methylation and

poly(ADP-ribosyl)ation processes and, in particular,

whether or not the inhibitory effect exerted by H1 histone

on in vitro enzymatic DNA methylation (Caiafa et al.,

1991) could be essentially due to the poly(ADP-

ribosyl)ated isoform of this protein.

In order to verify this hypothesis the poly(ADP-

ribosyl)ated and the poly(ADP-ribose)-free H1 histone

isoforms were purified. The modified protein was purified

by affinity chromatography on an aminophenylboronate

column of H1 histone obtained from permeabilized L929

mouse fibroblasts (Zardo et al., 1997) incubated for 10

min with 500 µM NAD, Figure 14A , while the

unmodified one was obtained from mouse fibroblasts

preincubated for 24 hours with 8 mM 3-aminobenzamide,

a well-known inhibitor of the poly(ADP-ribosyl)ation

process (Griffin et al., 1995). In both preparations the

entire H1 histone fraction was isolated by overnight

extraction in 0.2 M H2SO4 followed by a second

extraction in 10% (w/v) PCA (Johns, 1977). DNA

methyltransferase assays, performed in presence of 5 units

DNA methyltransferase purified from human placenta

nuclei and using as methyl donor 16 µM SAM plus 50

µCi/ml 3H-SAM, have shown that the poly(ADP-ribose)-

free isoform of H1 histone failed to inhibit in vitro DNA

methylation when added up to a protein/DNA ratio of 0.25

(w/w) while the poly(ADP-ribosyl)ated one was, instead,

highly inhibitory under the same condition, Figure

14B .

C. Effect of ADP-ribose polymers onin vitro DNA methylation.

Other experiments were carried out in order to verify

whether ADP-ribose polymers by themselves could play a

direct role in the modulation of DNA methyltransferase

activity. ADP-ribose polymers, isolated from L929

fibroblasts incubated with 50 µM 32P-NAD were

fractionated on Sephadex G-50. These protein-free

polymers caused a clear-cut inhibition of in vitro

methylation of dsDNA but not of ssDNA. The extent of

this inhibition is directly dependent on the size of the

polymers, as compared to a control assay in absence of

polymers considered as 100%, Figure 15 . Since a high

ADP-ribose polymers/DNA ratio did not affect

methylation of ssDNA the polymers can hardly be

visualized as directly interacting with DNA

methyltransferase.

In the close relationship existing between poly(ADP-

ribosyl)ation and DNA methylation processes, the

poly(ADP-ribosyl)ation of H1 histone appears to play a

key role. Since the association of H1 histone with ADP-

ribose polymers can be either covalent (Naegeli and

Althaus, 1991) or non-covalent (Panzeter et al., 1992),

further investigations are needed to ascertain whether also

the latter adduct is effective in maintaining CpG

dinucleotides in their unmethylated state. To go into this

question some in vivo experiments were performed in

which the correlation between DNA methylation and

poly(ADP-ribosyl)ation processes was investigated by

using the methyl-accepting ability assay on isolated nuclei

and/or purified DNA from L929 mouse fibroblasts. The

results shown in Figure 16 , support the working

hypothesis of an in vivo relationship between the two

nuclear processes suggesting a role of poly(ADP-

ribosyl)ation in preserving a number of CpG dinucleotides

from endogenous methylation, maintaining them in an

unmethylated state. By gel retardation assay we could also

show that poly(ADP-ribosyl)ated H1 histone has a high

capacity of linking CpG-rich ds-oligonucleotide, so that it

is possible to suppose that it has a preferential location on

genomic DNA in regions rich in these nucleotides. Since,

on the other hand, only relatively short poly-ADPribose

chain(s) are bound to H1 histone (D'Erme et al., 1996), it

is unlikely that they can be responsible by themselves for


674

the intense inhibitory effect exerted on the methylation of

ds DNA by the poly(ADP-ribosyl)ated isoform of H1

histone. In conclusion our hypothesis is that after DNA

packaging into nucleosomes, the access to the DNA of a

Figure 1 4 . A) Purification of poly(ADPribosyl)ated H1 histone isoform on an aminophenylboronate column

chromatography, monitoring the absorbance at 230 nm (closed circles), or the radioactivity (open circles). B ) Comparison

between poly(ADP-ribose)-free H1 histone (closed squares) and the purified poly(ADP-ribosyl)ated isoform (closed circles) for

their inhibitory effect on in vitro DNA methylation. Each value is the average value of three different experiments. "Reprinted

from Biochemis t ry 36 , Zardo et al.. Does poly(ADP-ribosyl)ation regulate the DNA methylation pattern?, 7937-7943, (1997)

with kind permission of the American Chemical Society".


675

F i g u r e 1 5 . Effect of ADP-ribose polymers of different size (A: striped bars, n>40; B : white bars, n <6n<40; C:

horizontally striped bars, n<20) on in vitro DNA methylation. Control assay, taken as 100%, was performed in absence of

polymers. Different polymers/DNA ratios, ranging from 0.25 to 1.00, are indicated in the abscissa. The assay was carried out for 1

h at 37°C in the presence of 50 units/ml DNA methyltransferase purified from human placenta nuclei, using 30 µg/ml Micrococcus

luteus dsDNA (left panel) and ssDNA (right panel) as substrates and 30 µCi/ml 3H-SAM as donor of methyl groups. The

incorporation of 3H-SAM in control dsDNA was 4.1 ± 0.1 picomoles and in control ssDNA 4.6 ± 0.3 picomoles. Histograms, in

which error bars have been included, represent the average value of three different experiments. "Reprinted from Biochemis try

3 6 , Zardo et al.. Does poly(ADP-ribosyl)ation regulate the DNA methylation pattern?, 7937-7943, (1997) with kind permission

of the American Chemical Society".

Figure 16. - Methyl-accepting ability experiments. In panel A the endogenous methyl accepting ability of native nuclei,

obtained from 6.5x106 L929 fibroblasts preincubated for 24 hrs without (control) and with 8 mM 3ABA, was performed in the

presence of 16 µM 3H-SAM. The level of methyl groups has been evaluated on the total DNA purified from cells. Control DNA,

whose incorporation was 2.8 ± 0.1 picomoles of 3H SAM, was considered as 100%. In panels B and C, DNA samples (3 mg each)

purified from the nuclei -- obtained from 6.5x106 L929 fibroblasts preincubated for 24 hrs without (control) and with 8 mM 3ABA

and where the endogenous methyl accepting ability had previously been saturated with 16 µM “cold” SAM -- were used as

substrates for evaluating their residual methyl accepting ability in the presence either of 50 units/ml human DNA

methyltransferase or of 50 units/ml bacterial SssI methylase. The incorporation of 3H SAM in control DNA was 0.3 ± 0.02

picomoles in panel B and 6 ± 0.2 picomoles in panel C. Histograms, in which error bars have been included, represent the

average value of three different experiments. "Reprinted from B i o c h e m i s t r y 3 6 , Zardo et al.. Does poly(ADP-ribosyl)ation

regulate the DNA methylation pattern?, 7937-7943, (1997) with kind permission of the American Chemical Society".

moving methyltransferase would then be limited by the

presence of poly(ADP-ribosyl)ated H1 and/or by

preferentially long and branched polymers linked in a non-

covalent way to the histone, so as to afford protection of

the unmethylated state of those CpG-rich DNA regions

(Zardo et al., 1997).

Acknowledgment.

This work was supported by the Italian Ministry of

University and Scientific and Technological Research

(60% Progetti di Ateneo and 40% Progetti di Interesse

Nazionale) and by Fondazione "Istituto Pasteur-

Fondazione Cenci Bolognetti".

References

Adams, R.P.L., David, T., Fulton, J., Kirk, D., Qureshi, M.

and Burdon, R.H. (1 9 8 4 ). Eukariotic DNA methylase

properties and action on native DNA and chromatin.

Curr. Top. Microbiol . Immunol . 108, 143-156.


676

Alvarez-Gonzalez, R. and Jacobson, M.K. (1 9 8 7 ).

Characterization of polymers of adenosine diphosphate

ribose generated in vitro and in vivo . Biochemis try

26, 3218-3224.

Asiedu, C.K., Scotto L., Assoian, R.K. and Ehrlich, M.

(1 9 9 4 ). Binding of AP-1/CREB proteins and of MDBP to

contiguos sites downstream of the human TGF-1 gene.

Bioch im. B iophys . Ac ta 1219, 55-63.

Aubin, R.J., Fréchette, A., de Murcia, G., Mandel, P., Lord,

A., Grondin, G. and Poirier, G.G. (1 9 8 3 ). Correlation

between endogenous nucleosomal hyper(ADP-

ribosyl)ation of histone H1 and the induction of

chromatin relaxation. EMBO J. 2, 1685-1693.

Baubichon-Cortay, H., Mallet, L., Denoroy, L. and Roux, B.

(1 9 9 2 ). Histone H1 subtype presents structural

differences compared to other histone H1 subtypes.

Evidence for a specific motif in the C-terminal domain.

B i o c h i m . B i o p h y s . A c t a 1122, 167-171.

Bestor, T.H. and Ingram, V.M. (1 9 8 3 ). Two DNA

methyltransferases from murine erythroleukemia cells:

purification, sequence specificity and mode of interaction

with DNA. P r o c . N a t l . A c a d . S c i . U S A 80, 5559-

5563.

Bestor, T.H., Gundersen, G., Kolsto, A.B., and Pryd H.

(1 9 9 2 ). CpG islands in mammals gene promoter

inherently resistant to de dovo methylation. Genet .

Anal . Tech. Appl . 9, 48-53.

Bird, A.P., Taggart, M., Frommer, M., Miller, O.J. and

Macleod, D. (1 9 8 5 ). A fraction of the mouse genome that

is derived from islands of nonmethylated CpG-rich DNA.

Cel l 40, 91-99.

Bird, A.P. (1 9 8 6 ). CpG-rich islands and the function of DNA

methylation. Nature 321, 209-213.

Bird, A.P. (1 9 8 7 ). CpG islands as gene markers in the

vertebrate nucleus. Trends Genet. 3 , 342-347.

Bloch, S. and Cedar, H. (1 9 7 6 ). Methylation of chromatin

DNA. Nucl . Acids Res . 3, 1507-1519.

Boulikas, T. (1 9 8 9 ). DNA strand breakes alter histone ADP-

ribosylation. Proc. Natl . Acad. Sci . USA 86, 3499-

3503.

Boyes, J. and Bird, A.P. (1 9 9 1 ). DNA methylation inhibits

transcription indirectly via a methyl-CpG binding

protein. Cel l 64, 1123-1134.

Boyes, J. and Bird, A.P. (1 9 9 2 ). Repression of genes by

DNA methylation depends on CpG density and promoter

strenght: evidence for involvment of a methyl-CpG

binding protein. EMBO J. 11, 327-333.

Brandeis, M., Ariel, M. and Cedar, H. (1 9 9 3 ). Dynamics of

DNA methylation during development. B i o e s s a y s 15,

709-713.

Brandeis, M., Frank, D., Keshet, I., Siegfried, Z.,

Mendelsohn, M., Nemes, A., Temper, V., Razin, A. and

Cedar, H. (1 9 9 4 ). Sp1 elements protect a CpG island

from de novo methylation. Nature 371, 435-438.

Bruhat, A. and Jost, J.P. (1 9 9 5 ). In vivo estradiol-dependent

dephosphorilation of the repressor MDBP-2-H1 correlates

with the loss of in vitro preferential binding to

methylated DNA. P r o c . N a t l . A c a d . S c i . U S A 92,

3678-3682.

Caiafa, P., Attinà, M., Cacace, F., Tomassetti, A. and Strom,

R. (1 9 8 6 ). %-methylcitosine levels in nucleosome

subpopulations differently involved in gene expression.

Bioch im. B iophys . Ac ta 867, 195-200.

Caiafa, P., Reale, A., Allegra, P., Rispoli, M., D'Erme M.,

and Strom, R. (1 9 9 1 ). Histones and DNA methylation in

mammalian chromatin. I. Differential inhibition of in

vitro methylation by histone H1. B i o c h i m . B i o p h y s .

Acta 1090, 38-42.

Caiafa, P., Reale, A., Santoro, R., D'Erme M., Marenzi, S.,

Zardo, G. and Strom, R. (1 9 9 5 ). Does hypomethylation

of linker DNA play a role in chromatin condensation?

Gene 157, 247-251.

Campoy, F.J., Meehan, R.R., McKay, S., Nixon, J. and Bird,

A.P. (1 9 9 5 ). Binding of Histone H1 is indifferent to

methylation at CpG sequences. J . B i o l . C h e m . 270,

26473-26481.

Carotti, D., Palitti, F., Lavia, P. and Strom, R. (1 9 8 9 ). In

vitro methylation of CpG islands. N u c l . A c i d s R e s .

17, 9219-9229.

Christman, J.K., Sheikhnejad, J., Marasco, C.J. and Sufrin,

J.R. (1 9 9 5 ). 5-methyl-2’-deoxycytidine in single-

stranded DNA can act in cis to signal de novo DNA

methylation. Proc. Natl . Acad. Sci . USA 92, 7347-

7351.

Clark, S.J., Harrison, J. and Frommer, M. (1 9 9 5 ). CpNpG

methylation in mammalian cells. Nature Genet . 10,

20-27.

Cole, R.D. (1 9 8 7 ). Microheterogeneity in H1 histones and

its consequences. In t . Pept . Pro te in Res . 30, 433-

449.

Davie, J.R. (1 9 9 5 ). The nuclear matrix and the regulation of

chromatin organization and function. Structural and

funct ional organizat ion of the nuclear matr ix ,

191-249 (Berenzey, R. and Jean, K.W. eds.) Academic

Press, Inc.

Davie, J.R. and Delcuve, J.P. (1 9 9 1 ). Characterization and

chromatin distribution of H1 histones and high-mobility-

group non-histone chromosomal proteins of trout liver

and hepato cellular carcinoma. B i o c h e m . J . 280, 491-

497.

Davis, T., Rinaldi, A., Clark, L. and Adams, R.L.P. (1 9 8 6 ).

Methylation of chromatin in vitro. B i o c h i m .

Biophys . Acta 866, 233-241.

De Lucia, F., Faraone-Mennella, M.R., D'Erme, M., Quesada,

P., Caiafa, P. and Farina, B. (1 9 9 4 ). Histone-induced

condensation of rat testis chromatin: testis-specific H1t

versus somatic H1 variants. Biochem. B i o p h y s .

Res. Commun. 198, 32-39.


677

de Murcia, G. and Ménissier de Murcia, J. (1 9 9 4 ). Poly(ADP-

ribose) polymerase: a molecular nick sensor. Trends

B i o c h e m . S c i . 19, 172-176.

de Murcia, G., Jacobson, M. and Shall, S. (1 9 9 5 ).

Regulation by ADP-ribosylation. Trends Cel l B i o l .

5 , 78-81.

D'Erme M., Santoro, R., Allegra, P., Reale, A., Marenzi, S.,

Strom, R. and Caiafa, P. (1 9 9 3 ). Inhibition of CpG

methylation in linker DNA by H1 histone. B i o c h i m .

Biophys . Acta 1173, 209-216.

D'Erme, M., Zardo, G., Reale, A. and Caiafa, P. (1 9 9 6 ). Co-

operative interactions of oligonucleosomal DNA with the

H1e histone variant and its poly(ADP-ribosyl)ated

isoform. Biochem. J . 316, 475-480.

Desmarais, Y., Menard, L., Lagueux, J. and Poirier, G.G.

(1 9 9 1 ). Enzymological properties of poly(ADP-ribose)

polymerase: characterization of automodification sites

and NADase activity. B i o c h i m . B i o p h y s . Acta

1078, 179-186.

Ferro, A.M., Higgins, N.P. and Olivera, B.M. (1 9 8 3 ).

Poly(ADP-ribosylation) of a DNA topoisomerase. J .

B io l . Chem. 258, 6000-6003.

Frank, D., Keshet, I., Shani, M., Levine, A., Razin, A. and

Cedar, H. (1 9 9 1 ). Demethylation of CpG islands in

embrionic cells. Nature 351, 239-241.

Giancotti, V., Bandiera, A., Ciani, L., Santoro, D., Crane

Robinson, C., Goodwin, G.H., Biocchi, M., Dolcetti, R.

and Casetta, B. (1 9 9 3 ). High-mobility-group (HMG)

proteins and histone H1 subtypes expression in normal

and tumor tissues of mouse. E u r . J . B i o c h e m . 213,

825-832.

Gradwohl, G., de Murcia J.M., Molinete M., Simonin F.,

Koken, M, Hoeijmakers, J.H.J. and de Murcia, G.

(1 9 9 0 ). The second zinc-finger domain of poly(ADP-

ribose) polymerase determines specificity for single-

stranded breaks in DNA. Proc . Nat l . Acad. Sc i . USA

87, 2990-2994.

Griffin, R.J., Curtin, N.J., Golding, B.T., Durkacz, B.W. and

Calvert, A.H. (1 9 9 5 ). The role of inhibitors of

poly(ADP-ribose) polymerase as resistance-modifying

agents in cancer therapy. Bioch imie 77, 408-422.

Gruenbaum, Y., Stein, R., Cedar, H. and Razin, A. (1 9 8 1 ).

Methylation of CpG sequences in eukaryotic DNA. FEBS

Lett . 124, 67-71.

Hasse, A. and Schulz, W.A. (1 9 9 4 ). Enhancement of reporter

gene de novo methylation by DNA fragments from the

alpha-fetoprotein control region. J . B io l . Chem. 269,

1821-1826.

Higurashi, M. and Cole, R.D. (1 9 9 1 ). The combination of

DNA methylation and H1 histone binding inhibits the

action of a restriction nuclease on plasmid DNA. J .

B i o l . C h e m . 2 6 6 , 8619-8625.

Huang, H.C. and Cole, R.D. (1 9 8 4 ). The distibution of H1

histone is nonuniform in chromatin and correlates with

different degrees of condensation. J. B i o l . Chem.

259, 14237-14242.

Huang, L., Wang, R., Gama-Sosa, M.A., Shenoy, S. and

Ehrlich, M. (1 9 8 4 ). A protein from human placental

nuclei binds preferentially to 5-methylcytosine-rich

DNA. Nature 308, 293-295.

Ikejma, M., Noguchi, S., Yameshita, R., Ogura, T., Sugimura,

T., Gill, D.M. and Miwa, M. (1 9 9 0 ). The zinc-fingers of

human poly(ADP-ribose) polymerase are differentially

required for the recognition of DNA breaks and nicks and

the consequent enzyme activation. J. B i o l . Chem.

265, 21907-21913.

Jacobson, M.K. and Jacobson, E.L. (1 9 8 9 ). ADP-Ribose

Transfer React ions: Mechanisms and

B i o l o g i c a l S ign i f i cance , Springer Verlag, New

York.

Jost, J.P. and Hofsteenge, J. (1 9 9 2 ). The repressor MDBP-2

is a member of the histone H1 family that binds

preferentially in vitro and in vivo to methylated

nonspecific DNA sequences. P r o c . N a t l . A c a d . S c i .

USA 89, 9499-9503.

Jost, J.P. (1 9 9 3 ). Nuclear extracts of chicken embryos

promote an active demethylation of DNA by excision

repair of 5-methyldeoxicitidine. Proc. Nat l . Acad.

Sci . USA 90, 4684-4688.

Jost, J.P. and Jost, Y.C. (1 9 9 4 ). Transient DNA

demethylation in differentiating mouse myoblasts

correlates with higher activity of 5-methyldeoxycytidine

excision repair. J . B io l . Chem. 269, 10040-10043.

Jost, J.P., Siegmann, M., Sun, L. and Leung, R. (1 9 9 5 ).

Mechanisms of DNA demethylation in chicken embryos.

J . B io l . Chem. 270, 9734-9739.

Kautiainen, T.L. and Jones, P. (1 9 8 5 ). Effects of DNA

binding proteins on DNA methylation in vitro.


Kawaichi, M., Ueda, K. and Hayaishi, O. (1 9 8 1 ). Multiple

autopoly(ADP-ribosyl)ation of rat liver poly(ADP-ribose)

syntetase. J . B io l . Chem. 256, 9483-9489.

Keshet, I., Ysraeli, J. and Cedar, H. (1 9 8 5 ). Effect of

regional DNA methylation on gene expression. Proc.

Natl . Acad. Sci . USA 82, 2560-2564.

Khan, R., Zhang, X.Y., Supakar, P.C., Ehrlich, K.C. and

Ehrlich, M. (1 9 8 8 ). Human methylated DNA-binding

protein. J . B io l . Chem. 263, 14374-14383.

Lennox, R.W. and Cohen, L.H. (1 9 8 3 ). The histone H1

complements of dividing and nondividing cells of the

mouse. J . B io l . Chem. 258, 262-268.

Lennox, R.W. (1 9 8 4 ). H1 differences in evolutionary

stability among mammalian subtipes. J . B i o l . C h e m .

259, 669-672.

Leonhardt, H., Page, A.W., Weier, H.U. and Bestor T.H.

(1 9 9 2 ). A targeting sequence directs DNA

methyltransferase to sites of DNA replication in

mammalian nuclei. Cel l 71, 865-873.


678

Levine, A., Yeivin, A., Ben-Asher, E., Aloni, Y. and Razin,

A. (1 9 9 3 ). Histone-H1 mediated inhibition of

transcription initiation of methylated templates in vitro.

J . B io l . Chem. 268, 21754-21759.

Lewis, J.D., Meehan, R.R., Henzel, W.J., Maurer-Fogy, I.,

Jepsen, P., Klein, F. and Bird, A.P. (1 9 9 2 ). Purification,

sequence, and cellular localizzation of a novel

chromosomal protein that binds to methylated DNA.

Cel l 69, 905-914.

Liao, L.W.L. and Cole, R.D. (1 9 8 1 a ). Condensation of

dinucleosomes by individual subfractions of H1 histone.

J . B io l . Chem. 256, 10124-10128.

Liao, L.W.L. and Cole, R.D. (1 9 8 1 b ). Differences among H1

histone subfractions in binding to linear and superhelical

DNA. J . B io l . Chem. 256, 11145-11150.

Lindner, H., Helliger, W. and Puschendorf, B. (1 9 9 0 ).

Separation of rat tissue histone H1 subtypes by reverse-

phase H.P.L.C. Biochem. J . 269, 359-363.

MaCleod, D., Charlton, J., Mullins, J. and Bird, A.P.

(1 9 9 4 ). Sp1 sites in the mouse aprt gene promoter are

required to prevent methylation of the CpG islands.

Genes and Dev. 8, 2282-2292.

Magewu, A.N. and Jones, P.A. (1 9 9 4 ). Ubiquitous and

tenacious methylation of the CpG sites in codon 248 of

the p53 gene may explain its frequent appearance as a

mutational hot spot in human cancer. M o l . C e l l .

B i o l . 14, 4225-4232.

Malanga, M. and Althaus, F.R. (1 9 9 4 ). Poly(ADP-ribose)

molecules formed during DNA repair in vivo . J . B i o l .

Chem. 269, 17691-17696.

McArthur, M. and Thomas, J.O. (1 9 9 6 ). A preference of

histone H1 for methylated DNA. EMBO J. 15, 1705-

1714.

Meehan, R.R., Lewis, J.D., McKay, S., Kleiner, E.L. and

Bird, A.P. (1 9 8 9 ). Identification of a mammalian protein

that binds specifically to DNA containing methylated

CpGs. Cel l 58, 499-507.

Meehan, R.R., Lewis, J.D. and Bird, A.P. (1 9 9 2 ).

Characterization of MeCP2, a vertebrate DNA binding

protein with affinity for methylated DNA. Nucl . Acids

Res . 20, 5085-5092.

Mendoza-Alvarez, H. and Alvarez-Gonzalez, R. (1 9 9 3 ).

Poly(ADP-ribose) polymerase is a catalytic dimer and the

automodification reaction is intermolecular. J. B i o l .

Chem. 268, 22575-22580.

Ménissier de Murcia, J., Molinete, M., Gradwohl, G.,

Simonin, F. and de Murcia, G. (1 9 8 9 ). Zinc-binding

domain of poly(ADP-ribose) polymerase partecipates in

the recognition of single strand breaks on DNA. J . M o l .

B i o l . 210, 229-233.

Mummaneni, P., Bishop, P.L. and Turker, M.S. (1 9 9 3 ). A

cis-acting element accounts for a conserved methylation

pattern upstream of the mouse adenine

phosphoribosyltransferase gene. J . B i o l . C h e m . 268,

552-558.

Mummaneni, P., Walker, K.A., Bishop, P.L. and Turker, M.S.

(1 9 9 5 ). Epigenetic gene inactivation induced by a cis-

acting methylation centre. J . B i o l . C h e m . 270, 788-

792.

Naegeli, H. and Althaus, F.R. (1 9 9 1 ). Regulation of

poly(ADP-ribose) polymerase. J . B i o l . C h e m . 266,

10596-10601.

Nan, X., Campoy, F.J. and Bird, A.P. (1 9 9 7 ). MeCP2 is a

trascriptional repressor with abundant binding sites in

genomic chromatin. Cel l 88, 471-481.

Panzeter, P.L., Realini, C.A. and Althaus, F.R. (1 9 9 2 ).

Noncovalent interactions of poly(adenosine diphosphate

ribose) with histones. Biochemis try 31, 1379-1385.

Pawlak, A., Bryans, M. and Jost, J.P. (1 9 9 1 ). An avian 40

kDa nucleoprotein binds preferentially to a promoter

sequence containing one single pair of methylated CpG.

Nucl . Acids Res . 19, 1029-1034.

Pehrson, J.R. and Cole, R.D. (1 9 8 2 ). Histone H1

subfractions and H1° turnover at different rates in

nondividing cells. Biochemis try 21, 456-460.

Poirier, G.G., De Murcia, G., Jongstra-Bilen, J., Niedergang,

C. and Mandel, P. (1 9 8 2 ). Poly(ADP-ribosyl)ation of

polynucleosomes causes relaxation of chromatin

structure. P r o c . N a t l . A c a d . S c i . USA 79, 3423-

3427.

Razin, A. and Cedar, H. (1 9 7 7 ). Distribution of 5-

methylcitosine in chromatin. Proc . Nat l . Acad . Sc i .

USA 74, 2725-2728.

Razin, A. and Riggs, A.D. (1 9 8 0 ). DNA methylation and

gene function. Science 210, 604-610.

Razin, A., Szyf, M., Kafri, T., Roll, M., Giloh, H., Scarpa,

S., Carotti, D. and Cantoni, G.L. (1 9 8 6 ). Replacement of

5-methylcytosine by citosine: a possible mechanism for

transient DNA demethylation during differentiation.


Reale, A., Marenzi, S., Santoro, R., D'Erme, M., Zardo, G.,

Strom, R. and Caiafa, P. (1 9 9 6 ). H1-H1 cross-linking

efficiency depends on genomic DNA methylation.

Biochem. B i o p h y s . Res . Commun. 227, 768-

774.

Realini, C.A. and Althaus, F.R. (1 9 9 2 ). Histone shuttling

by poly(ADP-ribosyl)ation. J. B i o l . Chem. 267,

18858-18865.

Santoro, R., D’Erme, M., Reale, A., Strom, R. and Caiafa, P.

(1 9 9 3 ). Effect of H1 histone isoforms on the

methylation of single- or double.stranded DNA.

B i o c h e m . B i o p h y s . R e s . C o m m . 190, 86-91.

Santoro, R., D'Erme, M., Mastrantonio, S., Reale, A.,

Marenzi, S., Saluz, H.P., Strom, R. and Caiafa, P.

(1 9 9 5 ). Binding of histone H1e-c variants to CpG-rich

DNA correlates with the inhibitory effect on enzymic

DNA methylation. Biochem. J . 305, 739-744.

Schulze, E., Trieschemann, L., Schulze, B., Schimdt, E.R.,

Pitzel, S., Zechel, K. and GrossBach, U. (1 9 9 3 ).

Structural and functional differences between histone H1


679

sequence variants with differential intranuclear

distribution. Proc . Nat l . Acad. Sc i . USA 90 , 2481-

2485.

Scovassi, A.I., Mariani, C., Negroni, M., Negri, C. and

Bertazzoni, U. (1993). ADP-ribosylation of non histone

proteins in HeLa cells: modification of topoisomerase II.

Exp . Ce l l Res . 206 , 177-181.

Solage, A. and Cedar, H. (1978). Organization of 5-

methylcytosine in chromosomal DNA. Biochemis try

17, 2934-2938.

Szyf, M., Tanigawa, G. and McCarthy, P.L. Jr. (1 9 9 0 ). A

DNA Signal from the Thy-1 gene defines de novo

methylation patterns in embryonic stem cells. M o l .

C e l l . B i o l . 1 0 , 4396-4400.

Szyf, M. (1 9 9 1 ). DNA methylation patterns: an additional

level of information? B i o c h e m . C e l l B i o l . 69, 764-

767.

Tasheva, E.S. and Roufa, D.J. (1 9 9 4 ). Densely methylated

DNA islands in mammalian chromosomal replication

origins. M o l . C e l l . B i o l . 14, 5636-5644.

Tollefsbol, T.O. and Hutchinson III, C.A. (1 9 9 7 ). Control of

methylation spreading in synthetic DNA sequences by the

murine DNA methyltransferase. J. M o l . B i o l . 269,

494-504.

Toth, M., Muller, U. and Doerfler, W. (1 9 9 0 ). Establishment

of de novo DNA methylation patterns. Transcription

factors binding and deoxycytidine methylation at CpG

and non CpG sequences in an integrated adenovirus

promoter. J . M o l . B i o l . 214, 673-685.

van Holde, K. (1 9 8 8). Chromatin. S p r i n g e r S e r i e s i n

M o l e c u l a r B i o l o g y (Rich, A. ed.) Springer Verlag,

New York.

Weiss, A., Keshet, I., Razin, A., and Cedar, H. (1 9 9 6 ). DNA

demethylation in vitro: involvement of RNA. Cel l 86,

709-718.

Woodcook, D.M., Crowther, P.J. and Diver, W.P. (1 9 8 7 ).

The majority of methylated deoxycytidines in human DNA

are not in the CpG dinucleotide. B i o c h e m . B i o p h y s .

Res. Commun. 145, 888-894.

Woodcook, D.M., Crowther, P.J., Jefferson, S. and Diver,

W.P. (1 9 8 8 ). Methylation at dinucleotide other than

CpG: implications for human maintenance methylation.

Gene 74, 151-160.

Wu, R.S., Panusz, H.T., Hatch, C.L. and Bonner, W.M.

(1 9 8 6 ). Histones and their modification. CRC Crit .

R e v . B i o c h e m . 20, 201-263.

Yisraeli, J. and Szyf, M. (1 9 8 4 ). Gene methylation patterns

and expression. DNA methy la t ion and i t s

b i o l o g i c a l s i g n i f i c a n c e , 353-378 (Razin, A. et al.

eds), Springer-Verlag, N.Y.

Yoshihara, K., Itaya, A., Tanaka, Y., Ohashi, Y., Ito, K.,

Teaoka, H, Tsukada, K., Matsukage, A. and Kamiya, T.

(1 9 8 5 ). Inhibition of DNA polymerase $, DNA

polymerase %, terminal deoxinucleotidyl transferase, and

DNA ligase II by poly(ADP-ribosyl)ation reaction in

vitro. Biochem. Biophys . Res Commun. 128, 61-

67.

Zardo, G., Santoro, R., D’Erme, M., Reale, A., Guidobaldi,

L., Caiafa, P. and Strom, R. (1 9 9 6 ). Specific inhibitory

effect of H1e histone somatic variant on in vitro DNA-

methylation process. Biochem. B i o p h y s . R e s .

Comm. 220, 102-107.

Zardo, G., D’Erme, M., Reale, A., Strom, R., Perilli, M. and

Caiafa, P. (1 9 9 7 ). Does poly(ADP-ribosyl)ation regulate

the DNA methylation pattern? Biochemis try 36,

7937-7943.

Zhang, X.Y., Ehrlich, K.C., Wang, R.Y.H. and Ehrlich, M.

(1 9 8 6 ). Effect of site-specific DNA methylation and

mutagenesis on recognition by methylated DNA-binding

protein from human placenta. N u c l . A c i d s R e s . 14,

8387-8397.

Zhang, X.Y., Jabrane-Ferrat, N., Asiedu, C.K., Samac, S.,

Peterlin, B.M. and Ehrlich, M. (1 9 9 3 ). The major

histocompatibility complex class II promoter-binding

protein RFX (NF-X) is a methylated DNA-binding protein.

M o l . C e l l . B i o l . 1 3 , 6810-6818.

Zhang, X.Y., Ni, Y-S., Saifudeen, Z., Asiedu, C.K., Supakar,

P.C. and Ehrlich, M. (1 9 9 5 ). Increasing binding of a

transcriptor factor immediately downstream of the cap site

of a cytomegalovirus gene represses expression. Nucl .

Acids Res . 23, 3026-3033.


681


Poly (ADP-ribosyl)ation as one of the molecularevents that accompany mammalianspermatogenesis.

Piera Quesada

Department of Organic and Biological Chemistry, University Federico II of Naples; via Mezzocannone 16, 8O134 Napoli,

Italy.

__________________________________________________________________________________________________

Correspondence: Piera Quesada: Phone +39-81-7041235, Fax: +39-81-5521217, E-mail [email protected]

Summary

It is known that mammalian spermatogenesis is a synchronous process of cellular differentiation

during which morphological changes occur, concomitantly with alterations in the complement of

constituent proteins, that reflect differences in the mRNA populations coding for stage-specif ic

proteins. Moreover, the most dramatic changes in chromatin structure observed in eukaryotes, take

place during spermiogenesis, and the main nuclear processes occur in well-defined cell stages. Rat

test is has been used as experimental model in a research project carried out at various levels ,

represented by rat germinal cel ls (primary and secondary spermatocytes, round spermatids),

chromatin fractions (transcriptionally active chromatin, nuclear matrix, MARs) and purified

nuclear proteins (histone and non-histone proteins). Specific experiments have been carried out in

order to determine the poly(ADPR)polymerase content at different stages of germ-cell

differentiation, the poly(ADP-ribose) amount, length and complexity inside the nucleus, and the

poly(ADP-ribose) acceptors among tissue- and stage-specific nuclear proteins. The results indicate

that regulation of the poly(ADPribosyl)ation system accompanies the earlier phases of the

germinal cell differentiation. Indeed, poly(ADPribose)polymerase is particularly active in primary

spermatocytes, being possibly implicated in the recombination events that characterize the

pachytene phase of the meiot ic divis ion. Different c lasses of poly(ADPribose) modify different

chromatin fractions (DNase I-sensitive, DNase I-resistant chromatin, and nuclear matrix)

implicated in DNA replication, repair and transcription. Moreover, the H1 variant H1t, specifically

expressed in pachytene spermatocytes, represents the main poly(ADPribose) acceptor, together with

poly(ADPR)polymerase itself , in rat germ-cells . Its modification can amplify the role of histone

H1 variants as modulators of chromatin structure.

I. Introduction

A. Overview of mammalianspermatogenesis

Mammalian spermatogenesis is a highly specialized

process of differentiation and one of the most dramatic

events that any single cell manifests. It is a continuous

developmental process that occurs in the adult, and it offers

unique opportunity to study the regulation and execution

of cell differentiation programs.

As a result of years of work by many expert biologists

a detailed histological description of the process is

available; the endocrine and paracrine hormonal control has

been assessed. However, very little is still known about

the molecular and cellular events that underlie

spermatogenesis. Such evidences can now be achieved

applying the powerful tools of molecular genetics to this

important area of biology (Meiestrich, 1993).

Spermatogenesis is composed of three major phases:

(i ) mitotic proliferation of spermatogonia, (i i ) two

reductional divisions of meiosis that produce the haploid


682

Figure 1 . Schematic

representation of the different

populations of germinal cells

which characterize the different

stages of spermatogenesis. The

nucleoproteins present during

various stages of

spermatogenesis are correlated

with cell morphology.

spermatids, and (i i i ) extensive remodelling of these cells

during spermiogenesis to form the mature spermatozoa

(Figure 1 ).

Because cytokinesis is incomplete at each of the

mitotic and meiotic cell divisions, descendants of a single

stem cell develop within a syncytium in which cells are

connected by intracellular bridges (Braun et al., 1995). It

has been proposed that the cytoplasmic bridges allow the

passage of various macromolecules between cells, thus

ensuring synchronous development of all cells within a

clone, and gametic equivalence between haploid

spermatids.

The modification of the genome that occurs during

spermatogenesis establishes the pattern of paternal

expression that is essential for successful embryonic

development and the normal phenotype of the adult

(Meiestrich, 1993).

B. Meiosis and recombination

Of the three phases of the spermatogenesis (mitotic

proliferation, meiosis and post-meiotic differentiation)

meiosis most closely defines gametogenesis since no

mammalian cell other than germ cells undergo meiosis.

Moreover, in female gametocytes the meiosis occurs

during the fetal life, whereas in male gametocytes this

process occurs during the adult life and can be more easily

studied.

The unique events of meiosis include the pairing and

recombination of chromosomes during prophase of

meiosis I, and the segregation of homologous

chromosomes during anaphase of meiosis I. In meiotic

prophase chromosomes must find their homologues,

establish and maintain pairing, and they must carry out the

steps of recombination. Later on, chromosomes undergo

condensation, and recombination events are resolved as

crossover, that in the metaphase, stabilize the alignment of

homologues at the spindle equator, thus ensuring regular

segregation of homologous chromosomes from one

another during the ensuing anaphase (Hawley, 1988).

All these events are essential for production of

chromosomally balanced gametes; errors in pairing or

recombination in meiotic prophase lead to errors in

anaphase I segregation, giving rise to aneuploid offspring

by the production of chromosomally unbalanced gametes.

Mutagens, in particular aneuploidogens and clastogens, can

induce such errors. To understand the etiology of

chromosomally abnormal gametes, the knowledge of the

normal mechanisms of chromosome behaviour during

meiosis is required.

Cell-cycle check-points describe the mechanism by

which cells assess the completion of events required for

Quesada: Poly (ADP-ribosyl)ation in mammalian spermatogenesis

683

successful progress through the cell cycle and production

of euploid cell products (Hartwell and Weinert, 1989).

There may be a number of spermatogenesis meiosis

checkpoints and these may assess both completion of

recombination and accumulation of precursors for

spermiogenesis.

For instance, it is not known what happens during

pachytene and why it is so lengthy. The supposition

might previously have been that the process takes days

because this much time is needed to carry out the genetic

events of recombination. However it has been

demonstrated that the recombination events have reached

the point when chiasmata can be formed (Collins and

Newton, 1994). Thus, it may be not the genetic events

that account for the length of time that the spermatocyte

spends in this phase of its differentiation. Instead, the

length of pachytene may be determined by the need to

synthesize precursors that will be used for post-meiotic

spermiogenesis.

Different kinds of pairing occur during meiosis, which

involve base-pairing (DNA-DNA) interactions. Recently

an alternative model has been presented based upon

protein-DNA interactions involved in transcription (Cook,

1997). Interestingly, the condensation of chromatin into

discernible chromosomes usually inhibits transcription. In

contrast the "aligned" chromosomes during meiotic

prophase I are transcriptionally active. Indeed, it is

remarkable the observation that chromosomes only pair

when they are transcriptionally active. A general model has

been described for pairing based upon promoter-polymerase

interactions on the light of the observation that each

chromosome in the haploid set has a unique array of

transcription units. Indeed a correct pairing would be

nucleated when a promoter binds productively to a

homologous site in another transcription factory and is

then the consequence of transcription of partially condensed

chromosomes.

A concept is emerging that multiple components of

nuclear organization contribute to competence for gene

expression. Chromatin structure, nucleosome organization

and gene-nuclear matrix interactions provide a basis for

rendering sequences accessible to transcription factors,

supporting integration of activities at independent

promoter elements of cell cycle and tissue specific genes.

The structure of the eukaryotic nucleus is still an area

of interest in cell biology; it is well established that both

interphase chromatin and mitotic chromosomes are

organized into loops, anchored to a nuclear matrix by

specific DNA sequence landmarks named MARs (matrix

associated regions) or SARs (scaffold associated regions)

(Boulikas, 1993a). Such DNA sequences have been

demonstrated to contain origins of replication and

transcription enhancers (Jackson and Cook, 1985; Jackson

and Cook, 1986 ); then it has been suggested that the

nuclear matrix may play a key role in genome organization

and gene potentiation. For instance, the common structural

features of replication origins in all life forms have been

assessed (Boulikas, 1996); this knowledge is of importance

for the understanding of the mechanisms underlying the

differential expression of genes that colocalize with matrix

associated regions.

As in the somatic nucleus, chromatin within the male

gametes is organized in discrete loops; however, these

loops differ from their somatic counterparts with respect to

the packaging of their DNA and their average size. Loops

within the sperm nucleus are approx. 27Kb in size

(compared to 60Kb in somatic cell nuclei). The sperm

nuclear matrix attachment regions (SMARs) show a

somatic-like organization; furthermore, it has been

demonstrated that a specific subset of haploid specific and

constitutively expressed genes are associated with the

sperm nuclear matrix (Kramer and Krawets, 1996); thus

the mature sperm genome is organized in a specific non-

random manner.

C. Chromatin structure and function indifferentiating germinal cells

The most dramatic changes in chromatin structure and

function observed in eukaryotes take place during

spermatogenesis: (i ) DNA replication occurs in

spermatogonia and prior meiosis in preleptotene

spermatocytes. Spermatocytes undergo meiosis producing

spermatids which no longer divide but differentiate into

mature spermatozoa; (i i ) spermatogonia, spermatocytes,

and early spermatids are active in nuclear transcription,

whereas spermatids undergoing differentiation as well as

spermatozoa are totally inactive; (i i i ) an interval of

regulated DNA nicking followed by repair synthesis occurs

in meiotic cells at pachytene phase (DNA recombination);

(iv ) late spermatids are genetically inactive in DNA

replication and transcription; however, these cells are still

able to repair their genetic damage before the final nuclear

condensation of chromatin occurs; (v ) the chromatin of

spermatids undergoing differentiation to spermatozoa

becomes relaxed in late spermatids, exposing binding sites

on DNA at regions of nucleosome disassembly.

Cell type-specific expression of genes encoding these

germ-line nuclear proteins may be regulated at the

transcriptional level and/or at the post-transcriptional level.

Differences in post-transcriptional processing have

been observed during spermatogenesis. Stabilization of

mRNA via polyadenylation and the increased efficiency of

translational reinitiation associated with this process may

be critical. In this context, the replication-independent

histone mRNAs that are restricted to specific cell types

(e.g. those encoding H5, the H1 variant in nucleated red


684

blood cells) are polyadenylated, while the short-lived

replication dependent histone transcripts do not contain

poly(A) tracts. The polyadenylated histone mRNA lacks

the conserved terminal hairpin structure seen in non-

polyadenylated transcripts and instead contains the (polyA)

addition sequence (AAUAAA) upstream of the poly(A)

tract.

The rat testis-specific TH2b histone gene assumes a

hypomethylated chromatin structure at all stages of

spermatogenesis. The H1t mRNA level rises sharply in

meiotic pachytene spermatocytes, being very low in pre-

meiotic spermatogenic cells as a result of transcriptional

repression of the gene by a pre-meiotic cell-specific

protein. A temporal correlation has been observed between

the appearance of testis specific DNA binding proteins and

the onset of transcription of the testis-specific histone H1t

gene (Grimes et al., 1992).

The appearance of stage-specific mRNA during

spermatogenesis has been demonstrated (Thomas et al.,

1989). Experiments involving in vitro translation of

mRNA from isolated germ cells have suggested that

transcription of the phosphoglycerate kinase-2 (PGK-2)

and lactate dehydrogenase (LDH-C) genes first occurs in

pachytene spermatocytes and continues in round

spermatids. The expression of two DNA repair related

enzymes poly(ADPR)polymerase and DNA polymerase !,

also varies in germinal cells. While the spermatocytes

were shown to contain both enzymes as well as their

transcripts, in other cell types this has not been observed

(Menegazzi et al., 1991).

Screening of testis cDNA libraries have identified

cDNA clones that are transcribed specifically by round

spermatids, such as the protamines (Thomas et al., 1989).

Developmentally-regulated patterns of gene expression

have been explored and the data suggest the following: (i )

stage-specific patterns of transcription occur coincidentally

with the appearance and accumulation of distinct germ cell

types within the seminiferous epithelium; (i i )

transcription of a variety of genes occurs exclusively

within haploid spermatids, while specific transcripts

accumulate in spermatogonia and meiotic germ cells but

not in spermatids; (i i i ) several transcripts are detected

initially during meiotic prophase and continue to

accumulate in round spermatids; (iv ) several multigene

families express germ cell-specific isotypes, including "-

tubulin, !-actin, LDH and PGK; (v ) various oncogenes are

expressed in germ cells.

Collectively, these observations suggest that precise

controls exist for determining stage specific gene

expression during spermiogenesis (Erikson, 1990).

The differentiation of round spermatids into mature

spermatozoa requires the synthesis of hundreds of new

proteins and the assembly of a unique collection of

organelles. During this differentiation process a flagellum

is constructed the acrosomal vesicle is formed and the

nucleus is compacted to approximately one-tenth of its

volume (Hecht, 1986).

Spermiogenesis is characterized by chromatin

condensation and the replacement of histones typical of

earlier spermatogenic cells by the highly basic protamines.

During this period the nucleosomal pattern of somatic

chromatin is lost in round spermatids and replaced by an

highly compacted structure in mature sperm. The

replacement of histones involves the elimination of

somatic and germ cell specific histones. Variants for

histone H1, H2a, H2b and H3 that appear during the

mitotic and meiotic phases of spermatogenesis have been

described in both mouse and rat germinal cells (Meiestrich,

1989).

Several reports suggested that histone synthesis ceases

during spermiogenesis as these proteins are replaced by a

set of transitional proteins (TPs) in both rat and mouse.

Later in spermiogenesis the TPs are replaced by

protamines which persist in mature sperm. While the

organization of sperm chromatin is unknown, the

appearance of germ cell-specific histone variants, TPs and

protamines suggest that multiple proteins are involved in

chromatin remodelling.

The detailed mechanisms by which nucleosomes are

disassembled and the DNA finally compacted during

spermiogenesis remain unknown. Several factors affect

this mechanism including post-translational modification

of histones. A correlation exists between the occurrence of

extensive H4 acetylation in spermatids and the presence of

protamines in spermatozoa (Meiestrich, et al. 1992).

Acetylation of specific lysines, as well as other kind of

post-translational modification, can reduce protein positive

charge which is believed to modify their interaction with

DNA (Oliva et al., 1987)

Besides acetylation, phosphorylation and methylation,

ADPribosylation also plays a fundamental role in gene

regulation as it can dramatically influence DNA/histone

interactions, particularly during cell differentiation (Lautier

et al., 1993).

In recent years the elucidation of the molecular

mechanisms of the poly(ADPribosyl)ation reactions has

advanced rapidly. The nuclear enzyme

poly(ADPR)polymerase (PARP) participates in several

nuclear events (DNA replication, transcription and repair)

by catalyzing the post-translational modification of

chromosomal proteins. After its activation by DNA strand

breaks, this enzyme catalyzes its self-modification and the

modification of other DNA binding proteins with variably

sized ADPribose polymers (pADPR) consuming the

nuclear pool of !-NAD (Althaus and Richter, 1987).


685

pADPR is a homopolymer made up of adenosine

diphosphate ribose units. The nature of this polymer is

now well characterized as structurally similar to a nucleic

acid. Branching points up to a proportion of 3% can be

found within the polymer, contrary to DNA or RNA.

ADPribose polymers vary in complexity and may reach a

length of more than 200 residues. The half-life of pADPR

varies in relation to the length of the polymer chain as

well as to the nature of the acceptor protein: after

stimulation of pADPR synthesis with alkylating agents,

the half-life of pADPR is less than 1 min. The very short

life of pADPR suggests the presence of very high polymer

turnover in intact cells.

Indeed, many enzymes are involved in pADPR

metabolism. The poly(ADPR)polymerase, responsible for

the synthesis of the pADPR, catalyzes the initiation,

elongation and branching steps. PARP has also an

abortive NADase activity, with a Km four times that of

the polymerase activity. The elongation mode is probably

distal and is accomplished in a distributive fashion with

respect to the acceptor (Mendoza-Alvarez and Alvarez-

Gonzales, 1993).

pADPR catabolism in cells is achieved by three

different enzymes the poly(ADPR)glycohydrolase

(Hatakeiama, et al. 1986 ) the ADPribosyl protein lyase

(Okayama et al. 1978) and a phoshodiesterase that breaks

the pyrophosphate moieties of pADPR (Futai and Mizuno,

1967). The action of poly(ADPR)glycohydrolase (PARG)

is the most important and catalyzes polymer degradation

by exoglycosidic hydrolysis following an endoglycosidic

incision. Two different forms have been found in many

tissues: the existence of a nuclear and a cytoplasmic

pADPR glycohydrolase has been suggested. Various

factors modulate pADPR glycohydrolase activity including

the nature of the acceptors and the length of the polymer

(Hatakeiama, et al. 1986 )

D. Poly(ADPribosyl)ation reactionsduring rat germinal cell differentiation

In order to define the role of the poly(ADPribosyl)ation

system during germinal cell differentiation, we have used

the rat testis as an experimental model. Our investigations

have been carried out at various levels: (i ) at the cellular

level using different rat germinal cells (primary and

secondary spermatocytes, haploid spermatids); (ii) at the

nuclear level using isolated chromatin fractions

(transcriptionally active chromatin, nuclear matrix,

MARs); (iii) at the protein level using purified nuclear

proteins (histone and non-histone proteins).

In addition, we have carried out specific experiments in

order to determine: (i ) the content and enzymatic activity

of poly(ADPR)polymerase at different stages of germ-cell

differentiation; (i i ) the amount, length and complexity of

poly(ADPribose); (i i i ) the poly(ADPribose) acceptors

among tissue- and stage-specific nuclear proteins.

1. Cellular level

There is a great deal of evidence advocating for a role of

PARP in the propagation of epigenetic information.

PARP activity correlates with the species-specific life span

among mammalian species (Grube and Burkle, 1992).

Malignant transformation has been correlated to changes in

the regulation and expression of genes caused by

poly(ADPribosyl)ation dysfunction (Borek and Cleaver,

1986). The expression level of the PARP gene varies

during differentiation in several cell lines (Smulson et al.,

1995) and, in general, PARP seems to be more active in

the S and G2 phases of the cell cycle (Leduc et al., 1988).

Moreover, it has been inferred that pADPR

metabolism is involved mainly in base excision repair

(Mathis and Althaus, 1990; Satoh and Lindahl, 1992;

Malanga and Althaus, 1994; Lindahl et al., 1995).

Differences in PARP activity can be interpreted in many

ways: evidence has been obtained for changes in the

pADPR synthesis pattern (Satoh et al., 1994), and in the

nature of the poly(ADPribosyl)ated proteins (Boulikas,

1990), as a function of particular chromatin structure

activities.

A functional correlation was observed between early

phases of spermatogenesis and poly(ADPribosyl)ation in

the rat (Quesada et al., 1996).

2. Nuclear level

It is now possible to have a better insight into the

internal structure of the interphase nucleus. The nuclear

framework is now characterized as a non-histone protein

scaffold supporting the attachment points of DNA loops

(Jack and Eggert, 1992), this structure seems to influence

gene replication and transcriptional activity. It is widely

accepted that DNA replication and transcription (Jackson

and Cook, 1985; Jackson and Cook, 1986) as well as

DNA repair (Mc Cready and Coock, 1984) are actively

coordinated by the nuclear matrix.

It is thought that PARP plays a role in the

maintenance of genetic integrity. As an example, PARP is

defective in the cells from patients suffering from

xeroderma pigmentosum, who are unable to excise

pyrimidine dimers induced by ultraviolet radiation (Wood

et al. 1988).

A possible mechanism for the involvement of the

poly(ADPribosyl)ation reaction, in different chromatin

functions has been proposed (Realini and Althaus, 1992)

that takes into account all the described features of the

poly(ADPribosyl)ation reaction. According to this

mechanism, histone proteins are reversibly detached from

the chromatin by the concerted action of

poly(ADPR)polymerase and poly(ADPR)glycohydrolase;


686

the long ADPribose chains linked to the PARP enzyme

would be responsible for the dissociation of the chromatin

structure, and would then be degraded by the de-

ADPribosylating enzyme to allow reassociation of

histones to DNA.

Boulikas (1993b) also proposed a model showing how

pADPR chains, linked to the automodified form of PARP

might be involved in removing histones from matrix

associated regions of chromatin (MARs). Indeed,

poly(ADPribosyl)ated histone H1 might contribute to the

destabilization of nucleosome structure, unfolding the

DNA around the core histone octamer. The complete

removal of histones from short sequences of DNA (1-4

nucleosomes) seems to be required for repair as well as for

initiation of DNA replication and transcription (Boulikas,

1993b)

In rat testis, a functional form of PARP has been

identified which did not seem to be an intrinsic component

of the nuclear matrix; this form of PARP was rather

indirectly associated to the matrix structure (Quesada et al.,

1994).

Figure 2 . Temporal appearance of germinal cells during

development of rat seminiferous epithelium.

Data are expressed as a percentage of total cells,

solubilized by collagenase digestion from seminiferous

tubules, characterized as haploid spermatids, diploid and

tetraploid spermatocytes, on the basis of their DNA content

determined by cytofluorimetric analysis.

3. Protein level

In all cellular events involving poly(ADPribosyl)ation,

the state of chromatin represents a signal. It has been

shown that poly(ADPribosyl)ation could affect chromatin

structure by direct covalent modification of chromosomal

proteins and by non-covalent interaction with histones

(Lautier et al., 1993; Panzeter et al., 1992 ) due to the

high number of long and branched pADPR chains linked

to the PARP itself. A modulation in chromatin

superstructure, induced by synthesis and degradation of

pADPR, has been visualized by de Murcia et al. (1988)

using electron microscopy.

Moreover, several nuclear proteins have been identified

as ADPribose acceptors in vivo in different systems

(Althaus and Richter, 1987): the list includes both

structural chromosomal proteins (histones, HMGs, nuclear

matrix proteins, etc.) and nuclear enzymes (DNA

polymerase ", DNA ligase, Topoisomerase I and II, etc.).

It has been show that the modification of the linker

histone H1 by pADPR alters drastically chromatin

conformation (de Murcia et al., 1988; Boulikas, 1990). It

has been demonstrated that the stage- and testis-specific

histone H1 variant H1t are the preferential ADP-ribose

acceptors among acid-soluble chromosomal proteins in rat

testis (Quesada et al., 1990).

Poly(ADPR)polymerase has been found preferentially

associated to transcriptionally active chromatin domains

(Hough and Smulson, 1984). The microheterogeneity of

H1 is known to play a role in the compaction of DNA

into the nucleosome fiber. Among the H1 variants, H1t

exerts the lowest condensing effect (De Lucia et al., 1994)

and is mostly associated with transcriptionally active

chromatin regions (De Lucia et al., 1996). Moreover, the

same variants appeared to be ADPribosylated to different

extents. Thus, taken together these findings indicate that

poly(ADPribosyl)ation of histone proteins could

contribute to the structural dynamics characteristic of the

transcriptionally competent chromatin.

Our studies have explored the possible role of ADP-

ribosylation reactions during spermatogenesis; the

spermatogenesis offers a good model to investigate the

relationship between poly(ADPribosyl)ation and structural

and functional changes that chromatin undergoes during

cellular differentiation.

II. Results

The temporal appearance of germinal cells in the rat

seminiferous epithelium has been confirmed by

cytofluorimetric analysis of total germinal cells isolated by

collagenase digestion from testes of rats of different age.

Figure 2 shows that the seminiferous epithelium


687

contains different numbers of three cell types, a function of

animal age. On the basis of their DNA content rat

germinal cells can be discerned as haploid cells

(spermatids), diploid cells (mainly secondary

spermatocytes), and tetraploid cells (primary

spermatocytes). These three cellular populations are

indicative of different stages of germinal cell

differentiation. Indeed, the pachytene stage occurs in the

interval between 25-35 days of age which is the longest

portion of the prophase, characterized by the formation of

the synaptonemal complexes and by a high level of genetic

recombination. A difference is evident in the percentages of

diploid/haploid cells in an interval from 28-32 days of

postnatal age; the amount of tetraploid cells is always less

than 25% and does not change significantly until 60 days

of age.

Figure 3 shows that PARP activity varies in rat

testis in animals of different ages; the major level of

PARP activity was detected in testes of 30 day-old animals

coincident with a crucial phase of the spermatogenesis

(meiosis), when the germ cell content (diploid/haploid

ratio) changes drastically in the seminiferous tubules.

Moreover, the PARP enzyme is specifically present in the

seminiferous epithelium and its activity can be

Fi

gure 3 .

Poly(ADPR)polymerase

activity in testis of rats of

different ages. The specific

activity values of the

enzyme are reported, as

determined in isolated

nuclei, with the enzymatic

assays described in

Materials and Methods. Each

value represents the average

of four experiments done in

duplicate. S.T .:

seminiferous tubules

isolated by collagenase

digestion. Crypt .: artificial

cryptorchid testes obtained

by elevating the temperature

to 37°C and placing them in the body cavity.

Figure 4 : Levels of

PARP mRNA in testis and

prostate of differently aged

rats.

Northern-blot analysis of

total RNA extracted from

testis and prostate of 30-days

and 60-days old rats. Aliquots

of 20 mg (A) and 50 mg (B)

of RNA were hybridized with

[32 P]-labelled pRat cDNA

probe. The migration of 3,6

Kb of PARP mRNA is

reported on the right, and

that of 28S and 18S

ribosomal RNA on the left


688

Figure 5 . Cytofluorimetric analysis of

isolated rat germ cell populations.

Fractions of rat germ cells isolated by

centrifugal elutriation, were analyzed after

propidium staining, by cytofluorimetry. (A)

total germinal cells isolated by collagenase

digestion from testes of 45 days old rats; (B )

Fraction IV containing haploid spermatids;

(C) Fraction VI containing diploid secondary

spermatocytes; (D) Fraction VII containing

tetraploid primary spermatocytes.

functionally related to the ongoing process of

spermatogenesis, since it is drastically reduced in testes in

which artificial cryptorchidism is induced.

Preliminary results obtained by Northern blot analysis

of total RNA extracted from testes of rats of different ages

showed that the expression level of PARP varies

specifically in testis (not in other tissues) in relation with

the animal age (Figure 4). The higher PARP expression

level was observed in testes from 60-day old rats and could

be correlated to the post-meiotic gene expression observed

in mammals for instance to that for oncogenes (Erikson,

1990).

Using the procedure illustrated on Table 1 the three

cellular fractions have been isolated by centrifugal

elutriation. Round spermatids were identified by optic

microscope mainly in fractions V, whereas fraction VI

contains mainly secondary spermatocytes, and fraction VII

primary spermatocytes (data not shown). The enrichment

rate obtained in these fractions was determined by

cytofluorimetric analysis. Figure 5 shows that on

fraction IV, VI and VII approx. 85% enrichment was

obtained for haploid, diploid and tetraploid cells

respectively. Among these fractions, fraction VII

represents a particularly interesting sample since it

contains the highest percentage of tetraploid

spermatocytes, the cells implicated in the pachytene phase

of the meiotic division characterized by a high rate of RNA

synthesis and the expression of stage-specific

chromosomal proteins (histones TH2B and H1t).


689

Figure 6 : Uridine and Thymidine uptake in rat germ cell

populations.

The values indicate the [3H] incorporation observed in the

acid-insoluble material of isolated germ cell fractions.

Figure 7 . Poly(ADPR)polymerase and

poly(ADPR)glycohydrolase activity in rat germ cell

populations.

The specific activity values of the two enzymes are reported,

as determined in total cellular fractions, with the enzymatic

assays described in Materials and Methods. Each value

represents the average of four determinations done in

duplicate.

Germinal cell viability was assessed measuring

[3H]thymidine and [3H]uridine uptake. Figure 6 shows

that the isolated cellular populations are able to incorporate

the two precursors into the TCA precipitable material, and

that the highest incorporation value is associated with

primary spermatocytes.

The three germinal cell populations were analyzed for

poly(ADPR)polymerase and poly(ADPR)glycohydrolase

content. The results of the enzymatic activity assays are

reported on Figure 7 . A different value of PARP specific

activity was detected; the sample enriched on tetraploid

spermatocytes shows an enzymatic activity of 1.5 mU/mg

of DNA higher than the 0.88 mU/mg of DNA detected in

diploid cells, and three times as much as the 0.48 mU/mg

of DNA of haploid spermatids. On the contrary, the PARG

specific activity did not change significantly (0.8-1.0

mU/mg of DNA) in the same three samples.

These findings have been confirmed by Western-blot

analysis of the same samples, using polyclonal antibodies

directed against PARP. Figure 8 shows that, using

immunodetection, the amount of PARP in primary

spermatocytes cellular extract is higher than that in

secondary spermatocytes and haploid spermatids.

These results are in agreement with a previous report

by Corominas and Mezquita (1985) showing different

pADPR levels in successive stages of rooster

spermatogenesis, and support the functional relationship

between spermatogenesis and poly(ADPribosyl)ation we

have postulated (Quesada et al., 1990).

A potential role for PARP in cellular differentiation

has been inferred from studies on enzymatic activity and

protein amounts during differentiation in rat astrocytes

(Chambert et al., 1992) and 3T3LI preadipocytes (Smulson

et al., 1995). Changes in PARP expression accompany

also the differentiation of HL60 cells induced by retinoic

acid (Bhatia et al., 1990) and the proliferation of

lymphocytes with PHA (McKerney et al., 1989). Wein et

al. (1993) reported a cell cycle-related expression of PARP

in proliferating rat thymocytes. In this case, a translational

regulation of PARP occurs in the G1 phase, whereas a

translational as well as transcriptional activation of PARP

occurs in the S phase of the cell cycle.

It seems that different mechanisms of regulation of

PARP also accompany germinal cell differentiation.

Variations in the enzymatic activity have been observed

which can be interpreted as changes in the number of

PARP molecules, their enzymatic stimulation, modulation

of PARP synthesis at the transcriptional and post-

transcriptional levels, as well as in terms of differences in

pADPR turnover rates.

Our evidence indicates that an active

poly(ADPribosyl)ation system is associated mainly with

tetraploid spermatocytes raising the possibility that PARP


690

Figure 8 : Poly(ADPR)polymerase levels in rat germ cells.

Western-blot analysis of PARP levels in crude extracts of (1) haploid spermatids, (2) secondary spermatocytes and (3) primary

spermatocytes. (A) Coomassie-Blue staining; (B ) immunodetection with polyclonal antibody directed against human PARP.

Figure 9 : Isolation procedure of rat testis nuclear matrix.

The distribution of DNA, proteins and newly synthesized32 P-RNA in different chromatin fractions is also reported.

enzyme acts as a modulator of the nuclear processes that

occur in these cells by its automodification or

heteromodification (via covalent bond or non-covalent

interaction) of specific nuclear proteins.

As in other experimental systems we can presume that

poly(ADPribosyl)ation acts by inducing structural changes

of chromatin, thereby facilitating DNA metabolism, which

involves various protein-protein and DNA-protein

interactions. The high rate of replicational activity of

spermatocytes is well known and is accompanied by DNA

recombination, a finely controlled crucial event during

spermatogenesis ensuring integrity of the genetic material.

In order to determine the distribution and activities of

the poly(ADPribosyl)ation molecules inside the nucleus,

we have isolated and characterized different chromatin

fractions, according to the procedure illustrated in Figure

9 . Moreover, these fractions have been analyzed for their

ability to incorporate [32P]CTP in newly synthesized

RNA. A high level of transcriptional activity is present in

the DNase I-sensitive chromatin and also in the nuclear

matrix in agreement with findings already reported in other

systems.

We have also examined the level of ADPribosylation

in the same fractions, taking into account both the extent

of the modification and the size of the ADPribose

oligomers. Figure 10 shows DNA, proteins, PARP and

pADPR content of the three fractions. In particular, the

percentage of pADPR which was found associated with

nuclear matrices prepared from nuclei incubated with 200

µM [ 32P]NAD is approximately 10% of the total nuclear


691

F i g u r e 1 0 : DNA, Proteins, PARP and pADPR content

(%) of different chromatin fractions isolated after incubation

of intact nuclei with 32 P NAD.

1. DNase I-sensitive chromatin; 2. DNase I-resistant/ 2M

NaCl extractable chromatin; 3. Nuclear matrix.

activity. It is noticeable that the pADPR content was not

directly related to the amount of both DNA or protein.

These results seem to indicate that the nuclear matrix is

enriched in ADPribosylated proteins.

To characterize the kind of pADPR synthesized in

nuclei, aliquots of nuclear fractions were subjected to

alkaline hydrolysis (pH 12), thus detaching the intact

polymer, which was then processed for electrophoresis on

sequencing polyacrylamide gels and autoradiography. The

typical ladder of pADPR indicates that a polymer

population is synthesized in isolated nuclear matrix

different from that synthesized in DNase I-sensitive and -

resistant chromatin fractions; the matrix polymer is

enriched in ADPR chains longer than 20 moieties, and in a

fraction to the top of the gel which is known to be

enriched in branched polymers (Figure 11 ). Thus,

nuclear proteins appear to be modified by three classes of

pADPR differing in length and complexity.

To identify ADPribose acceptors, protein components

extracted by 2M NaCl from nuclei were separated by

means of 20% polyacrylamide gel electrophoresis on

acetic-acid-urea (1st dimension) and SDS 15%

polyacrylamide gel electrophoresis (2nd dimension) and

processed for autoradiographic analysis. Autoradiography

of [32P]ADP-ribosylated nuclear proteins revealed that

Figure 11 : Autoradiography of 32 P-pADPR analyzed on a

20% polyacrylamide sequencing slab-gel.

1. Nuclear matrix; 2. DNase I sensitive chromatin; 3.

DNase I resistant/ 2M NaCl extractable chromatin.


692

Figure 1 2 : Electrophoretic pattern of loosely-bound

chromosomal proteins.

2D-gel analysis (see Materials and Methods) of 2M NaCl

extractable chromosomal proteins isolated from intact nuclei

incubated with 32 P NAD.

radioactivity was associated to histone-like proteins and to

component(s) with low electrophoretic mobility and high

molecular weight which can be identified as the

automodified form(s) of PARP.

Our results are in line with recent reports on the

presence, within the nucleus, of multiple classes of PARP

molecules, representing different functional forms of the

enzyme. Others have reported the presence of variable

amounts of PARP in nuclear matrices isolated from

different sources (Wesierska-Gadek and Sauerman, 1985;

Cardenas-Corona et al., 1987; Alvarez-Gonzales and

Ringer, 1988). According to Kaufmann et al. (1991) the

recovery of PARP in rat liver nuclear matrix varies with

the extraction conditions, since the association of the

enzyme to this structure is mediated by the formation of

intermolecular disulfide bonds. Thus, the amount of

pADPR (10%) associated with the rat testis nuclear matrix

is far from being negligible.

An interesting hypothesis can be drawn concerning the

well known automodification mechanism of PARP: the

presence of a portion of ADPR tightly associated to the

nuclear matrix raises the possibility of the occurrence of an

auto-ADPribosylated form of the enzyme anchored to the

nuclear matrix.

In order to identify the ADPribose acceptors among the

rat testis nuclear proteins their ADPribosylation has been

induced in intact nuclei, by incubation with 32P-NAD.

Acid-soluble proteins have been than extracted with 0.2 M

H2SO4 and separated by reverse-phase HPLC as reported in

Materials and Methods. In a one step procedure, several

histone and non-histone proteins have been purified, which

appeared to be ADPribosylated in different extent. The

histogram in Figure 12 shows the specific radioactivity

calculated for each of the purified acid-soluble proteins.

This analysis indicates that it is histone H1 which is

mainly ADPribosylated, compared to core histones and

HMG like chromosomal proteins. Moreover, among the

different histone H1 variants, H1t represents a better

ADPribose acceptor.

Figures 13 and 14 show the reverse phase (RP)-

HPLC pattern of H1 variants. Indeed, the autoradiography

of purified proteins determined by 15% polyacrylamide gel

electrophoresis in the presence of SDS, confirmed the

preferential ADPribosylation of the tissue- and stage-

specific variant of the histone H1, H1t. Moreover, the H1t

histone H1 variant appeared to be modified by ADPribose

oligomers ranging from 4 to 20 residues compared with no

longer than 12 ADPR residues on H1a (data not shown).

The modification of the linker histone H1 by

ADPribosylation is well documented in different systems.

The ADPribosylation sites have been also identified as the

#COOH group of the two glutamic residues in the N-

terminal and C-terminal domain of the protein, besides the

COOH-terminal group itself. All these residues are

conserved in the histone H1 variants and this means that

the different extents of their ADPribosylation can not be

explained on the account of unavailability of

ADPribosylation sites.

The existence of a non-covalent interaction of pADPR

molecules with histone proteins has been described. This

represents a different ADPribosylation mechanism, due to


693

Figure 13 : Purification procedure of 32 P ADPribosylated

nuclear proteins.

The histogram reports the specific radioactivity value that

has been calculated for single component purified by reverse-

phase HPLC.

the long and branched polymers linked to the PARP, able

to compete with DNA for histone binding (Panzeter et al.,

1992). Both the covalent and non-covalent

ADPribosylation of histone H1 variants has to be taken

into account to understand how PARP molecules modulate

chromatin structure.

The differential activation of gene expression between

hetero- and euchromatin is sustained by a non-uniform

distribution of somatic and tissue-specific H1 variants.

However, the specific involvement of different H1 variants

in transcriptional processes needs to be further

investigated; in this scenario, the ADPribosylation could

play an important role.

III. Conclusions

Our data have some important implications with regard

to the role and versatility of the poly(ADPribosyl)ation

reactions in mammalian spermatogenesis.

(i ) Regulation of the poly(ADPribose) turnover,

variations in the number of PARP molecules synthesized,

as well as changes in PARP transcription level, all seem

to accompany the earlier phases of the germinal cell

differentiation.

(i i ) Three classes of poly(ADPribose) molecules of

different length and complexity modify chromosomal

proteins and PARP in the three chromatin fractions which

were isolated (DNase I-sensitive DNase I-resistant

chromatin, and nuclear matrix) is implicated in DNA

replication, repair and transcription.

(i i i ). The testis-specific histone H1t is the main

poly(ADPribose) acceptor, together with PARP itself, in

rat germ-cells. Short oligomers modify covalently histone

components whereas long and branched polymers are

bound to non-histone proteins (PARP, nuclear matrix

proteins, etc.).

Two general conclusions can be drawn: (i ) The

observation of a poly(ADPribosyl)ation system

particularly active in pachytene spermatocytes is in

agreement with the proposed role of PARP as a guardian

of the genome in preventing aberrant recombination. Such

a role is particularly important in primary spermatocytes

undergoing the pachytene phase of the meiotic division,

characterized by DNA recombination events.

(i i ) The observation of the preferential ADP-

ribosylation of the histone H1 variant H1t, among the

chromosomal proteins, can be related to the different roles

attributed to the H1 variants in the compaction of DNA

into the nucleosome fiber. Among the H1 variants H1t

exerts the lower condensing effect and is mostly associated

with transcriptionally active chromatin regions. This can

explain why H1t is specifically expressed in pachytene

spermatocytes. The different ADPribosylation patterns of

histone H1 variants can amplify their role as modulators of

chromatin structure, in relation to the various nuclear

events that require chromatin remodelling.

How to study the molecular and cellular

events of the spermatogenesis using

poly(ADPribosylation) as a marker.

As already stated, to understand the etiology of

chromosomally abnormal gametes, the knowledge of the

normal mechanisms of chromosome behaviour is required.

To obtain this goal it would be advantageous to

understand all the events of meiosis carried out by purified

spermatocytes in culture. Procedures for separation of


694

Figure 14 : Reverse-Phase

HPLC of 32 P ADPribosylated

histone H1 variants.

Chromatographic pattern of

histone H1 variants

separated by reverse-phase

HPLC on a Vydac C4 silica

column (see Materials and

Methods).

The Figure shows also the

electrophoretic pattern on

SDS 15% polyacrylamide

gel and the corresponding

autoradiography of proteins

contained in each peak.

germ-cells, such as the centrifugal elutriation, allow the

isolation of cell populations at defined stages of

differentiation. This method provides large amounts of

meiotic cells and round spermatids suitable for biochemical

studies (Quesada et al., 1996). The purity of the fractions

is about 90% with a cell survival of 99%. Thus gene

expression can be analyzed during the different stages of

spermatogenesis with this approach.

Recently, short-term cultures of pachytene

spermatocytes have been assessed that will be useful for

investigating the defining events of pachytene substage of

meiotic prophase, namely maintenance of chromosome

pairing and recombination (Handel et al., 1995).

Rats of different ages can also be used to study the

progression of germinal cell differentiation; since the

spermatogenesis is a synchronous process, it is possible to

correlate the presence in the seminiferous epithelium of

different amount of each kind of germinal cells to the

animal age.

Synaptonemal complex analysis is a novel approach

for identifying substances hazardous to the germ cells.

This sensitive cytological procedure reveals induction of

structural damages and pairing abnormalities in SCs of

meiotic prophase chromosomes, together with other germ-

line toxic effects, and can allow an estimate of chemical

mutagenicity (Allen et al., 1987).

To study changes in nuclear and chromatin

composition and genomic activity during spermatogenesis

several agents can be used to alter sex-organ development.

Cryptorchidism can be artificially-induced in animals

maintained on a retinol-depleted diet since vitamin A has

long be known to be essential for normal male

reproductive function (La Borde et al., 1995).

Spermatogenesis can be restored by the oral administration

or intratesticular injection of retinoids which have been

demonstrated to activate cell division in type A

spermatogonia and induce their differentiation in vitro.

Cis-DDP is a drug frequently used in the treatment of

testicular neoplasia; Cis-DDP determines changes in

Sertoli cells function and is responsible for spermatid

killing (Chu, 1994). Cis-DDP can be administrated acutely

or chronically in combination with a gonadotrophin

releasing hormone analogue to evaluate the effects of this

agent on spermatogenesis, with the aim to study the

damage induced by chemotherapeutic drugs, and whether

pretreatment with LRA is able to prevent this damage

(Scott et al., 1996) .

Undifferentiated germ cells can offer a rapid, sensitive

and inexpensive alternative to screening for teratogenicity

and genotoxicity of diverse chemicals on live animals. As

an example methyl-methanesulphate (MMS) can be used

which is known to affect mouse sperm chromatin structure

and testicular cell kinetics. MMS is a potent electrophilic,


695

monofunctional alkylating agent which induces clastogenic

damage including dominant lethal mutations. It has been

demonstrated that MMS alkylates cysteine -SH groups in

sperm protamines, thereby destabilizing sperm chromatin

structure and leading to chromosomal breakage and

mutations (Evenson et al., 1993).

Alternatively, physical agents can be used to alter cell

progression. The DNA damage induced after exposure to

ionizing radiation can be measured at different cellular

stages of spermatogenesis as well as the DNA repair rate,

as well as in relation to the apoptotic index (Van Loom et

al., 1991).

The comet assay represents a sensitive a rapid method

for DNA damage detection on individual cells, able to size

highly fragmented DNA and DNA repair efficiency

(Fairbain et al., 1995)

Treatment with okadaic acid (OA) seemingly abrogates

a checkpoint that holds mid-pachytene spermatocytes in

prophase for several days. Since OA is a polyether

monocarboxylic acid able to inhibit several protein

phosphatases, the phosphorylation state of critical proteins

may be an important element of spermatogenic meiosis

cell-cycle control (Wiltshire et al., 1995).

Going from genes to their product, the cellular protein

pattern is also representative of the functional state of the

cell. Moreover post-translational modification of proteins

represents also a signal of the normal or abnormal

progression of the cell cycle. Several post-translational

protein modifications can be directly related to the nuclear

metabolism.

The role of protein modification, could be studied in

vitro and/or in vivo. Reconstitution experiments,

involving addition of specific proteins to DNA in vitro

could help determine how these proteins affect chromatin

structure. A direct but very difficult method would be to

specifically prevent their synthesis in vivo and then

determine the resulting modification on chromatin

structure.

Two dimensional gel electrophoresis of proteins can be

used to detect and purify hundreds of proteins from a single

sample simultaneously. The protein map of rat

spermatocytes and round spermatids has been obtained with

this method, and several proteins were successfully

identified in the SWISS-PROT protein database, in order

to highlight changes in protein expression during meiosis

(Cossio et al., 1995).

On the basis of these considerations and in the light of

the experimental evidence already available, it is now

possible to better understand the functional correlation

between poly(ADPribosyl)ation and spermatogenesis.

Alterations in the poly(ADPribosyl)ation system might be

useful markers in gene therapy of germ-cell line.

IV. Materials and methods

Materials . Wistar rats (28-130 days of age) were used in

all the experiments.

[U14 C]NAD+, nicotinamide [U14 Cadenine dinucleotide

ammonium salt (248 mCi/mmol), [32 P]NAD+, nicotinamide

adenine dinucleotide di(triethylammonium) salt (adenylate32 P), 1000 Ci/mmol, [32 P]CTP (3000 Ci/mmol), [3H]

methylthymidine (40 Ci/mmol) and [3H]Uridine (25-30

Ci/mmol) were supplied by Amersham International plc.

DNase I (EC 3.1.21.1) (2,000 U/mg), collagenase type XI,

phenyl-methyl-sulphonyl-fluoride (PMSF), leupeptin,

spermine, and spermidine were obtained from SIGMA

Chemical Company; guanidine thiocyanate was from FLUKA,

pure grade. Electrophoretic molecular weight markers were

purchased from Pharmacia, X-Omat RP films from Kodak, and

nitro-cellulose filters (0.45 µm pore size, type HA) from

Millipore. Reverse-Phase HPLC silica 300-C4 columns were

from Vydac(ODS 5 mm particles, 0.5 x 25 cm)

I so la t i on o f germinal c e l l s by centrifugal

e lutr iat ion . Testes from two rats were decapsulated,

resuspended in 10 ml of Dulbecco's minimal essential medium

(DMEM) and seminiferous tubules free of the interstitial

tissues were obtained by collagenase treatment (0.25 mg/ml).

The seminiferous tubules were then incubated at 30°C for 60

min in DMEM containing 0.25 mg/ml collagenase, 0.075

mg/ml DNase I, and 0.5% bovine serum albumin (BSA). After

incubation the cell suspension was centrifuged for 10 min at

1,200g. Aliquots of the pellet were complexed with propidium

and subjected to cytofluorimetric analysis in a Becton-

Dickinson cytofluorimeter. Total germinal cells resuspended

in DMEM, in the presence of 0.1 mg/ml DNase I and 0.5%

BSA, were separated into fractions enriched in various cell

types by sedimentation at unit gravity (centrifugal

elutriation), as described by (Quesada et al., 1996).

A cell suspension of 10 ml (180-220x106 cells obtained

from two testes of 40-45 day aged rats) was loaded into a JE-6

Beckman elutriator rotor and separation was performed with

speeds of 3,000-2,000 rpm and flow rates of 13-40 ml/min

(Table 1 ). The buffer employed was phosphate buffered

saline (PBS) containing 0.5% BSA; several fractions of 50 ml

were collected. Aliquots of the pellet of single fractions were

complexed with propidium and subjected to cytofluorimetric

analysis in a Becton-Dickinson cytofluorimeter.

Isolation of sub-cellular fractions was carried out

essentially as described by (Quesada et al., 1996). Testes were

homogenized in a medium containing 0.32 M sucrose, 2 mM

MgCl 2, 5 mM !-mercaptoethanol, centrifuged at 2,000g for

10 min and the nuclei recovered in the sediment. The

supernatant was centrifuged at 20,000 g for 60 min in a SW40

rotor, to obtain in the pellet the microsomal fraction separate

from the soluble fraction.

[ 3H] Thymidine and [3H] Uridine incorporat ion .

Seminiferous tubules were labelled for 1 hr at 37°C in DMEM

in the presence of 30 mCi/ml of [3H] methylthymidine or

[3H]Uridine. The tissue was washed three times with the

medium and the cells isolated as described above. The cellular


696

fractions were washed with 5% trichloracetic acid, ethanol,

ethanol/ether and ether to remove unbound radioactivity. The

residue was suspended in 0.2 M HClO4 and portions of the

solution were used for the determination of radioactivity on a

Beckman LS8100 liquid scintillation spectrometer.

[ 32 P] RNA synthes is . RNA synthesis was followed by

incubation of nuclei in the presence of [32 P]CTP according to

(Greengerg and Zif, 1984) and [32 P]RNA was measured as 25%

thricloroacetic acid-insoluble radioactivity.

Iso lat ion of chromat in fract ions . Rat testis nuclei

were isolated by homogenization and differential

centrifugation, as previously described (Quesada et al., 1990).

Proteases were irreversibly inhibited by 1 mM PMSF and 0.1

mM leupeptin. The nuclear matrix isolation procedure was

essentially as described by (Tubo and Berezney, 1987) and

consisted in the endogenous digestion of isolated nuclei 45

mins at 37°C, followed by a three times repeated extraction

with an high salt buffer (2 M NaCl, 0.2 mM MgCl2, 1 mM

PMSF in 10 mM Tris-HCl pH 7.4) followed by tree washes

with low-salt buffer (0.2 mM MgCl2, 1 mM PMSF in 10 mM

Tris-HCl pH 7.4). The nuclear matrix, was then re-suspended

in one third volume of low-salt buffer.

Alternatively, nuclei were suspended in 10 mM TRIS-HCl

pH 7.5, 1 mM EDTA, 1 mM PMSF, digested with DNase I

(600U/mg of DNA) and lysed for 1 hr at 0°C. Lysed nuclei were

centrifuged for 20 min at 12,000 g, the supernatant collected

and the pellet re-extracted twice as above (DNase I resistant

chromatin). The insoluble fraction was digested with DNase I

(1,000 U/mg DNA) in 60 mM TRIS-HCl pH 7,5, 60 mM NaCl,

20 mM MgCl2 at 37°C for 1 hr, to obtain after centrifugation

12,000g 20 min, the nuclear scaffold.

Nuclear protein extract ion . Acid-soluble proteins

contained in nuclei, and nuclear fractions were extracted three

times by 1 h stirring at 4°C in 0.2 M H2SO4, and centrifuged at

10,000g for 15 min. The extracts were pooled and proteins

precipitated with 6 vols. of ice-cold acetone at -20°C

overnight. Precipitates were collected by centrifugation at

18,000g for 20 min at -10°C

Poly(ADPribose)polymerase and poly(ADPR)-

g lycohydro lase a c t i v i t y a s say . In a typical PARP

activity assay, the reaction mixture (final volume 250 µl)

contained: 100 mM Tris-HCl pH 8, 14 mM !-

mercaptoethanol, 10 mM MgCl 2, 4 mM NaF, 200 µM

[14 C]NAD (10,000 cpm/nmol), 12 µg DNase I, and, as enzyme

source, an amount of cells or sub-cellular fractions

corresponding to 30 µg of proteins. After 10 min incubation

at 20°C, the reaction was stopped with ice-cold trichloracetic

acid and the radioactivity present in the acid-insoluble

material, collected on a HAWP (0.45µm) filter, determined on

a Beckman LS8100 liquid scintillation spectrometer. One

enzymatic unit was defined as the enzyme activity catalyzing

the incorporation, per minute at 20°C, of one µmole of

ADPribose into acid-insoluble material.

The PARG activity assay was performed in a standard

reaction mixture containing 50 mM potassium phosphate (pH

7.2), 10 mM !-mercaptoethanol, 100 µg/ml BSA, 10 mM

[14 C] poly(ADPribose) (20 residues long on average) and, as

enzyme source an aliquot of cells or sub-cellular fractions

corresponding to 30 µg of proteins in a total volume of 100

µl. After 10 minutes of incubation at 37°C, the reaction was

stopped with ice-cold 20% trichloracetic acid and the

radioactivity present in the acid-soluble material determined

on a Beckman LS8100 liquid scintillation spectrometer. One

enzymatic unit was defined as that liberating 1 µmole of

ADPribose, per minute at 37°C.

B l o t t i n g e x p e r i m e n t s . Activity-blots and immuno-

blots were performed as described by (Simonin et al., 1991).

Aliquots of 107 cells were suspended in 100 µl of 50 mM

glucose, 25 mM Tris-HCl pH 8, 10 mM EDTA, 1 mM PMSF

and sonicated with 60-s pulses at 180V. The crude extract,

after incubation at 45°C for 15 min in 50 µl of 50 mM Tris-

HCl pH 6.8, 6 M urea, 6% ß-mercaptoethanol, 3% SDS, was

separated on 10% polyacrylamide slab-gel in the presence of

1% SDS and electrotransferred onto nitro-cellulose sheets at

4°C for 2 hrs at 200 mA. For activity-blot experiments

proteins transferred onto nitrocellulose sheets were incubated

for 1 hr at room temperature in renaturation buffer (50 mM

Tris-HCl pH 8, 100 mM NaCl, 1 mM DTT, 0.3% (v/v) Tween

20, 20 µM Zn(II)acetate, 2 mM MgCl 2) containing DNase I

activated DNA (2 µg/ml) and, for 2 hr more, in the same buffer

containing [32 P]NAD 1 µCi/ml. The blots were then washed

several times with the renaturation buffer, dried and analyzed

by autoradiography. For immuno-blot experiments

nitrocellulose sheets were treated for 1 hr with the blocking

solution (50 mM Tris-HCl pH 8, 150 mM NaCl, 0.5% (v/v)

Tween 20 and 3% (w/v) gelatin). Incubation with anti-human

PARP antibodies was performed, for 2 hrs at room temperature

in the same solution supplemented with 0.3% gelatin. The

blots were then washed twice with TBS-Tween and antibody

binding was detected by using alkaline phosphatase

conjugated goat anti-rabbit IgG second antibody with

BCIP/NBT (SIGMA) tablets as substrate for the color reaction.

For Northern blot experiments total cellular RNA (20-50

µg per sample) was first separated by electrophoresis on a 1%

agarose gel containing 6% formaldehyde, then blotted onto a

Hybond N membrane and hybridized with a fragment of rat

PARP cDNA (pRatC) [32 P]-labelled using a Multiprime DNA

labelling kit (Amersham International plc).

Prote in and DNA assay . Protein concentration was

determined following the method described by (Burton, 1968)

using bovine serum albumin as standard. DNA content was

determined on the basis of the absorbance at 260 nm (1.0

OD260nm = 50µg /ml DNA) or by the diphenylamine method

described by Burton (1968).

Electrophoret ic a n a l y s i s . Acid-soluble nuclear

proteins were analysed by electrophoresis on urea-acetic acid

20% polyacrylamide slab-gels (pH 2.9) and, SDS 7-15%

polyacrylamide slab-gels, as described by (Quesada et al.,

1996). Each labelled sample was run in duplicate and either

stained with Coomassie Brilliant Blue R-250 or

autoradiographed.

A n a l y s i s o f react ion products . Intact

[32 P]poly(ADP-ribose) moieties incorporated into the proteins

were detached by incubation at 60°C for 3 h with 10 mM Tris,

NaOH pH 12, 1 mM EDTA. Samples were extracted with


697

phenol/CHCl3/isoamyl alcohol (49:49:2), dried in a Speed-

Vac and dissolved in 50% urea, 25 mM NaCl and 4 mM EDTA,

pH 7.5, to be analyzed on 20% polyacrylamide slab-gel

(Panzeter and Althaus, 1990)

Reverse-Phase HPLC analysis . Acid soluble nuclear

proteins were purified by reverse-phase HPLC using a 300-5

C4 silica column and a gradient elution system formed by

0.1% trifluoracetic acid (solvent A) and 95% CH3CN in 0.1%

trifluoracetic acid (solvent B). The gradient elution was from

15 to 65% of solvent B in 70 mins (slope 0.65%/min).

Acknowledgements

I thank all my colleagues Prof. Benedetta Farina, Dr.

M.Rosaria Faraone-Mennella, Dr. Maria Malanga, Dr.

Filomena De Lucia, Dr. Luigia Atorino and Dr. Filomena

Tramontano for their contribution to the work summarized

in this review. I thank also Dr. Roy Jones, Principal

Scientific Officer at the Babrahan Institute of the BBSRC

in Cambridge, U.K., for his constant support and for

hosting me in his laboratory. This work was supported by

funding (40%) and 60% M.U.R.S.T..

References

Allen, J.W., De Weese, G.K., Gibson, J.B., Poorman, P.A.,

Moses, M.J., ( 1 9 8 7 ) "Synaptonemal complex damage

as a measure of chemical mutagen effects on mammalian

germ cells" Mutation Res .,190,19-24.

Althaus, F.R., and Richter, C.R., ( 1 9 8 7 ) in

"ADPribosy lat ion o f Prote ins: E n z y m o l o g y

and B i o l o g i c a l S ign i f i cance " Part I vol.37

Springer-Verlag Berlin.

Alvarez-Gonzales, R., and Ringer, D.P., ( 1 9 8 8 ) "Nuclear

matrix associated poly(ADPribose) methabolism in

regenerating rat liver" FEBS Letters , 236, 362-366.

Bhatia, K., Pommier, Y., Giri, C., Fornace, A.J., Imaizumi,

M., Breitman, T.R., Cherney, B.W., Smulson, M.E.,

( 1 9 9 0 ) "Expression of the poly(ADPR)polymerase gene

following natural and induced DNA strand breakage and

efect of hyperexpression on DNA repair"

Carc inogenes i s 11, 123-128.

Borek, C., and Cleaver, J.E., ( 1 9 8 6 ) "Antagonistic action of

a tumor promoter and poly(adenosine diphosphoribose)

synthesis inhibitor in radiation induced transformation"

Biochem. B i o p h y s . Res . Commun. ,134,1334-

1341.

Boulikas, T., ( 1 9 9 0 ) "Poly(ADPribosyl)ated histones in

chromatin replication" J . B i o l . Chem . 265, 14638-

14647.

Boulikas, T., ( 1 9 9 3 a ) "Nature of DNA sequences at the

attachment regions of genes to the nuclear matrix" J .

Ce l l . B iochem . 52, 14-22.

Boulikas T., ( 1 9 9 3 b ) "Relation between carcinogenesis

chromatin structure and poly(ADPribosylation)"

Anticancer Res ., 14, 885-894.

Boulikas, T., ( 1 9 9 6 ) "Common structural features of

replication origins in all life forms" J. C e l l .

B iochem ., 60, 297-316.

Braun, R.E., Lee, K., Schumaker, J.M., and Fajardo, M.A.,

( 1 9 9 5 ) “Molecular genetic analysis of mammalian

spermatids differentiation” in “Recent Progress i n

Hormone Research” 50, 275-286, Acad. Press New

York.

Burton, K., (1 9 6 8 ) "A study of the conditions and

mechanism of diphenylammine reaction for the

colorimetric estimation of DNA" Cel l , 23, 585-593.

Cardenas-Corona, M.E., Jacobson, E.L., Jacobson, M.K.,

( 1 9 8 7 ) "Endogeous polymers of ADPribose are

associated with the nuclear matrix" J. B i o l . Chem.

262, 14863-14866.

Chambert, M.G., Niedergang, C.P., Hog, F., Partisani, M.,

Mandel, P., ( 1 9 9 2 ) "Poly(ADPR)polymerase expression

and activity during proliferation and differentiation of rat

astrocytes and neuronal cultures" B i o c h i m . B i o p h y s .

Acta , 1136, 196-202.

Chu, G., ( 1 9 9 4 ) , "Cellular response to cisplatin" J . B i o l .

Chem. 269, 787-790.

Collins, I., and Newton, C.S., ( 1 9 9 4 ) "Meiosis specific

formation of joint DNA molecules containing sequences

from homologous chromosomes" C e l l , 76, 65-75.

Cook, P.R., ( 1 9 9 7 ) "The transcriptional basis of

chromosomes pairing" J . C e l l S c i e n c e , 110, 1033-

1040.

Corominas, M., and Mezquita, C., (1 9 8 5 ) "Poly(ADP-

ribosylation) at successive stages of rooster

spermatogenesis" J . B io l . Chem. 260, 16269-15273 .

Cossio, G., Sanchez, J.C., Golaz, O., Wettstein, R.,

Hochstrasser, D.F., ( 1 9 9 5 ) "Spermatocytes and round

spermatids of rat testis protein patterns"

Electrophores is , 16, 1225-1230, .

De Lucia, F., Faraone-Mennella, M.R., d'Erme, M., Quesada,

P., Caiafa, P., Farina, B., ( 1 9 9 4 ) "Histone induced

condensation of rat testis chromatin: testis apecific H1t

versus somatic H1 variants" Biochem. B i o p h y s .

Res . Common . 198, 32-39.

De Lucia, F., Faraone-Mennella, Quesada, P., Farina, B.,

( 1 9 9 6 ) "Poly(ADPribosyl)ation system in

transcriptonally active rat testis chromatin fractions" J .

C e l l . B i o c h e m . , 63, 334-341, .

de Murcia, G., Huletsky, A., Lamarre, D., Gaudreau, A.,

Pouyet, J., Daune, M., Poirier, G.G., ( 1 9 8 8 )

"Modulation of chromatin superstructure by

poly(ADPribose) synthesis and degradation" J. B i o l .

Chem .,261, 7011-7017.

de Murcia, G., and Menissier de Murcia, J., ( 1 9 9 4 )

"Poly(ADP-ribose)polymerase: a molecular nick-sensor"

TIBS 19, 172-176.


698

Erikson, R.P., ( 1 9 9 0 ) "Post meiotic gene expression"

Trends Genet. , 6N°8, 264-269.

Evenson, D.P., Jost, L.K., Baer, R.K., ( 1 9 9 3 ) "Effect of

methyl methane sulfonate on mouse sperm chromatin

structure and testicular cell kinetics" Env. & M o l .

Mutagenes is , 21, 144-153.

Fairbain, D.W., Olive, P.L., O'Neill, K.L., ( 1 9 9 5 ) "The

comet assay: a comprehensive review" Mutat ion Res .

369, 37-59.

Futai, M., and Mizuno, D. (1 9 6 7 ) "A new phosphodiesterase

forming nucleoside 5' monophosphate from rat liver" J .

B io l . Chem . 263, 5301-5307.

Greengerg, M.E., and Ziff, E.B., ( 1 9 8 4 ) "Stimulation of 3T3

cells induces transcription of the c-fos proto-oncogene"

Nature, 311, 433-438.

Grimes, S.R., Wolfe, S.A., and Koppel, D.A., ( 1 9 9 2 )

"Temporal correlation between the appearance of testis

specific DNA binding proteins and the onset of

transcription of the testis-specific histone H1t gene"

Exp. Ce l l Res . , 201, 216-224.

Grube, K., and Burkle, A., ( 1 9 9 2 ) "Poly(ADP-

ribose)polymerase activity in mononucleated leukocytes

of 13 mammalian species correlates with species-specific

life span" P r o c . N a t l . A c a d . S c i . U S A 89, 11759-

11763.

Hatakeiama, K., Nemoto, Y., Ueda, K., Haiashi, O., ( 1 9 8 6 )

"Purification and characterization of poly(ADP-

ribose)glycohydrolase: different mode of action on large

and small poly(ADPribose)" J . B i o l . C h e m ., 261,

14902-14911.

Handel, M.A., Caldwell, K.A., & Wiltshire, T., ( 1 9 9 5 )

"Culture of pachytene spermatocytes for analysis of

meiosis" Deve l . Gen . , 16, 128-139.

Hartwell, L.H., and Weinert, T.A., ( 1 9 8 9 ) “Checkpoints

controls that ensure the order of cell cycle events”

Sc ience , 246, 629-634.

Hawley, R.S., ( 1 9 8 8 ) in " Genet ic Recombinat ion" (R.

Kucherlapati & R Smith eds) AAAS Press Washington DC.

Hecht N.B., ( 1 9 8 6 ) in "Exper imenta l approaches to

mammal ian spermatogenes i s " (J. Rossant & R.A.

Pedersen eds) Cambridge Univ. Press, New York.

Hough, C.J., and Smulson, M.E., (1 9 8 4 ) "Association of

poly(ADP-ribosyl)ated nucleosomes with

transcriptionally active and inactive regions of

chromatin" Biochemis try , 23, 5016-5023.

Jack, R.S., and Eggert, H., ( 1 9 9 2 ) "The elusive nuclear

matrix" Eur. J. Biochem . 209, 503-509.

Jackson, D.A., and Cook, P.R., (1 9 8 5 ) "Replication occurs

at a nuclear skeleton" EMBO J., 4, 919-925.

Jackson, D.A., and Cook, P.R., ( 1 9 8 6 ) "Transcription

occurs at a nuclear skeleton" EMBO J., 5, 1403-1410.

Kaufmann, S.H., Brunet, G., Talbot, B., Lamarre, D., Dumas,

C., Shaper, J..H., Poirier, G.G., ( 1 9 9 1 ) "Association of

poly(ADPribose)polymerase with the nuclear matriix: the

role of intermolecular disulfide bond formation" Exp.

Cel l Res ., 192, 524-535.

Kramer, J.A., and Krawets, S.A., ( 1 9 9 6 ) "Nuclear matrix

interactions within sperm genome" J. B i o l . Chem . ,

271,11619-1162.

La Borde, J.B., Pipkin, J.L., Hinson, W.G., Anson, J.F.,

Sheenan, D.M., Young, J.F., Hansen, D.K., ( 1 9 9 5 )

"Retinoic acid induced stress protein synthesis in the

mouse" Life Sc ience , 56 N°21, 1767-1778.

Lautier, R., Lageaux, J., Thibodeau, J., Menard, L., & Poirier,

G.G., ( 1 9 9 3 ) "Molecular and Biochemical features of

poly(ADPribose) metabolism" M o l . C e l l . B i o c h e m .

122, 171-193.

Leduc, Y., Lawrence, J.J., de Murcia, G., Poirier, G.G.,

( 1 9 8 8 ) "Cell cycle regulation of

poly(ADPribose)synthetase in FR3T3 cells" B i o c h i m .

Biophys . Ac ta , 968, 275-282.

Lindahl, T.,Satoh, M.S., and Poirier, G.G., ( 1 9 9 5 ) "Post-

translational modification of poly(ADPribose)polymerase

induced by DNA strand breaks" TIBS 20(10), 405-411.

Malanga, M., and Althaus, F.R., ( 1 9 9 4 ) "Poly(ADPribose)

molecules formed during DNA repair in vivo" J .

B i o l . C h e m ., 269, 17691-17696.

Mathis, G and Althaus, F.R., ( 1 9 9 0 ) "Uncoupling DNA

excision repair and nucleosomal unfolding in

poly(ADPribose) depleted mammalian cells"

Carc inogenes i s , 11, 1237-1239, .

Mc Cready, S.J., and Coock, P.R., ( 1 9 8 4 ) "Lesions induced

in DNA by ultraviolet light are repaired" J. Ce l l

Sc ience , 70, 189-196.

McKerney, R., Tavassoli, M., Shall, S., Bazinsky, A.,

Jonstone, A., ( 1 9 8 9 ) "Changes in mRNA levels of

poly(ADPribose)polymerase during activation of human

lymphocytes" B i o c h i m . B i o p h y s . A c t a , 1009, 185-

187.

Meiestrich, M.L., ( 1 9 8 9 ) in "His tones and other

bas ic nuc lear prote ins" (L.S. Hnilica, G.S. Stein &

J.L. Stein eds), CRC Press, Orlando.

Meiestrich, M.L., ( 1 9 9 3 ) , in "Molecular b i o l o g y o f

the male reproductive system" (D.M. De Kretser ed)

Acad.Press, Orlando.

Meiestrich, M.L., Trostle-Wege, P.K., Lin, R., Bhatnagar,

Y.M., & Allis, C.D., ( 1 9 9 2 ) "Highly acetylated H4 is

associated with histone displacement in rat spermatids"

Mol. Reprod. and Develop . 31, 170-181.

Mendoza-Alvarez, H., and Alvarez-Gonzales, R., ( 1 9 9 3 )

"Poly(ADPribose) polymerase is a catalytic dimer and the

automodification reaction is intermolecular" J. B i o l .

Chem. , 268, 22575-22580.

Menegazzi, M., Grassi-Zuccotti, G., Carcereri-De Prati, A.,

Poltronieri, P., Nyunova, H., Shiratori-Nyunova, Y.,

Miwa, M., Suzuki, H., ( 1 9 9 1 ) "Differential expression

of poly(ADPribose)polymerase and DNA polymerase ß in

rat tissues" Exp . Ce l l . Res . 197, 66-74.


699

Okayama, H., Honda, M., Hayaishi, O., (1 9 7 8 ) "Novel

enzyme from rat liver that cleaves an ADP-ribosyl histone

linkage" Proc. Nat l . Acad. S c i . USA 75, 2254-

2257.

Oliva, R., Bazett-Jones, D., Mezquita, P., Dixon, G.H.,

( 1 9 8 7 ) "Factors affecting nucleosome disassembly by

protamines in vitro" J. B i o l . Chem ., 35, 17016-

17025.

Panzeter, P.L., and Althaus, F.R., ( 1 9 9 0 ) "High resolution

size analysis of ADPribose polymers using modified DNA

sequencing gels" Nucl. Acid Res .,18, 2194.

Panzeter, P., L., Realini, C., and Althaus, F.,R. ( 1 9 9 2 ) "Non

covalent interactions of poly(adenosine-diphosphate-

ribose) with histones" Biochemis try , 31, 1379-1385.

Quesada, P., Farina, B., Jones, R., ( 1 9 9 0 ) "Poly(ADP-

ribosyl)ation of nuclear proteins in rat testis correlates

with active spermatogenesis" B i o c h i m . B i o p y s .

Acta 1007, 167-175.

Quesada, P., d'Erme, M., Parise, G., Farone-Mennella, M.R.,

Caiafa, P., Farina, B., ( 1 9 9 4 ) "Nuclear matrix associated

poly(ADPribosyl)ation system in rat testis chromatin"

Exp. Cel l Res ., 214, 351-357 .

Quesada, P., Atorino, L., Cardone, A., Ciarcia, G., Farina, B.,

( 1 9 9 6 ) "Poly(ADPribosyl)ation system in rat germinal

cells at different stages of differentiation" Exp. Ce l l

Res . 226, 183-190.

Realini, C., and Althaus, F.,R., ( 1 9 9 2 ) " Histone shuttling

by poly(ADPriboslation)" J. B i o l . Chem ., 267,26,

18858-18865.

Satoh, M.S., and Lindal, T., ( 1 9 9 2 ) "The role of

poly(ADPribose) formation in DNA repair" Nature, 356,

356-359.

Satoh, M.S., Poirier, G.G., Lindal, T., ( 1 9 9 4 ) "Dual function

for poly(ADPribose) synthesis in response to DNA strand

breakage" Biochemis try , 33, 7099-7106, .

Scott, C.A., Desinan, L., Maffezini, M., Simonato, A.,

Avellini, C., De Stefani, S., Rizzi, V., Carmignani, G.,

Beltrami, G.A., ( 1 9 9 6 ) "Effects of cisplatin and

luteinizing hormone releasing analogues on rat

spermatogenesis" Anal . & Quant. C y t o l . &

H i s t o l . 18, 361-373.

Simonin, F., Briand, J.P., Muller, S., de Murcia, G., ( 1 9 9 1 )

"Detection of poly(ADPR) polymerase in crude extracts by

activity-blot" Anal . B iochem ., 195, 226-231.

Smulson, M.E., Kang, V.H., Ntambi, J.M., Rosenthal, D.S.,

Ding, R., and Simbulan, M.G., ( 1 9 9 5 ) "Requirement for

the expression of poly(ADPribose)polymerase during the

early steges of diffrentiation of 3T3-L1preadipocytes, as

studied by antisense RNA induction" J . B i o l . C h e m . ,

270, 119-127.

Thomas, K.H., Wilkie,T.M., Tomasherfsky, P., Bellvè, A.R.,

Simon, M.I., ( 1 9 8 9 ) "Differential gene expression

during mouse spermiogenesis" B i o l . o f R e p r o d . 41,

729-739.

Tubo, R.A., and Berezney, R., ( 1 9 8 7 ) "Identification of 100

and 150S DNA polymerase " megacomplexes solubilized

from the nuclear matrix of regenerating rat liver" J .

B io l . Chem. 262, 5857-5865.

Van Loom, A.A.W.M., Den Boer, P.J., Van der Schans,

Mackenbach, P., Grootegoed J.A., Baan, R.A., Lohman,

P.H.M., ( 1 9 9 1 ) "Immunochemical detection of DNA

damage induction and repair at different cellular stages of

spermatogenesis of the hamster after in vitro or in vivo

exposure to ionizing radiation" E x p . C e l l R e s . , 193,

303-309.

Wein, K.H., Netzker, R., Brand, K., ( 1 9 9 3 ) "Cell cycle

related expression of poly(ADPribosysl)transferase in

prolifersting rat thymocytes" B i o c h i m . B i o p h y s .

Acta ,1176, 69-76.

Wesierska-Gadek, J., and Sauerman, G., (1 9 8 5 )

"Modification of nuclear matrix proteins by

ADPribosylation" Eur. J. Biochem ., 153,421-428.

Wiltshire, T., Park, C., Caldwell, K.A., Handel, M.A.,

( 1 9 9 5 ) "Induced premature G2/M-phase transition in

pachytene spermatocytes includes events unique to

meiosis" D e v e l o p . B i o l . 169, 557-567.

Wood, R.D., Robins, P., and Lindahl, T., ( 1 9 8 8 )

"Complementation of the Xeroderma Pigmentosus DNA

repair defect in cell-free extracts" Cel l , 53,97-106.

Haucke: Membrane biogenesis: from mechanism to disease

701


Membrane biogenesis: from mechanism to disease

Volker Haucke

Department of Cell Biology & Howard Hughes Medical Institute, Yale University School of Medicine, 295 CongressAvenue, New Haven, CT 06510, USA

__________________________________________________________________________________________________

Correspondence to: Volker Haucke, phone: +1-203-737 4469, fax: +1-203-737 1762, E-mail: [email protected]

Keywords: membrane biogenesis-peroxisomal disorders-nerve terminal-endocytosis-autoimmune diseases-cancer

Summary

The biogenesis of membranes involves the continous f low of proteins and lipids which areselectively targeted to or retrieved from specific compartments within eukaryotic cells. While somediseases are caused by the impairment of particular protein transport pathways or mislocalizationof a certain protein others may be related to altered signal transduction cascades resulting fromdefective endocytosis of plasma membrane receptors or other membrane trafficking defects. Theimplications of this hypothesis for our understanding of the proper functioning of a eukaryotic celland for the treatment of human diseases are being discussed.

I. Overview

Unlike bacteria, eukaryotic cells are elaboratelysubdivided into membrane-bounded, structurally andfunctionally distinct compartments. Each of these organellescontains a specific set of proteins, lipids and othermolecules which enables them to fulfill characteristicfunctions within the cell. On average, the membrane-bounded compartments together occupy nearly half thevolume of a cell, and about one third of all proteins within aeukaryotic cell are membrane proteins. Thus, membranebiogenesis and organelle maintenance are major tasks whichare essential for all eukaryotic cells (Palade, 1975).

During the past two decades it has become clear that anumber of inherited metabolic and neurological disordersresult from the mistargeting of particular proteins to anincorrect destination within the cell. As an example for thisclass of diseases I will describe a number of disordersresulting from defective peroxisome biogenesis.

Other pathological states pertaining to membranetrafficking however may arise from autoimmune impairmentof cells expressing a particular antigen or from geneticdefects resulting in the generation of an abnormal protein asexemplified by the deposition of !-amyloid protein in brainsof patients suffering from Alzheimer's disease. In the secondpart of this chapter I will, therefore, focus my discussion onthe trafficking of membranes at the nerve terminal in normaland certain pathological states.

II. Peroxisome biogenesis & dysfunction

A. How peroxisomes are formed

Peroxisomes are ubiquitous eukaryotic organelles whichare involved in a variety of metabolic processes such as thescavenging and destruction of peroxides, the !-oxidation offatty acids and the biosynthesis of ether lipids. However,unlike mitochondria and chloroplasts they do not containtheir own DNA and like most other intracellular membranescannot be formed de novo. Biologists and physicians alikehave become increasingly interested in the biogenesis ofthese organelles since Goldfischer reported in 1973 thatpatients with the cerebro-hepato-renal syndrome Zellweger'sdisease lacked demonstrable peroxisomes. Until now thenumber of peroxisomal biogenesis disorders (PBD) hasgrown to sixteen which fall into eleven differentcomplementation groups. In order to learn more about themolecular basis of these diseases investigators have studiedthe way by which peroxisomes import their constituentproteins from the cytosol using both mammalian cellcultures and yeast as model systems.

Protein targeting to the peroxisomal matrix is mediatedby evolutionary conserved peroxisomal targeting signals(PTSs) which bind to specific PTS receptors as depicted inFigure 1 . The majority of peroxisomal matrix proteinscarries a C-terminal tripeptide (SKL or closely similar)termed PTS1. PTS2 is a conserved N-terminal nonapeptide(R/K) (L/V/I) (X5) (H/Q) (L/A) and is used by a smaller

subset of matrix proteins.

Other internally located PTSs have been identified but,as with the targeting signals of peroxisomal membraneproteins, no consensus sequence has been found(Rachubinski & Subramani, 1995).


702

Figure 1: Hypothetical modelfor how proteins get imported intoperoxisomes.

(a) Import of proteins containingthe PTS1 signal for targeting to theperoxisomal matrix. PTS1R, receptorfor PTS1 containing proteins.

(b ) Import of proteins containingthe PTS2 signal for targeting to theperoxisomal matrix. PTS2R, receptorfor PTS2 containing proteins.

Components of a putative importchannel across the peroxisomalmembrane (gray) are indicated bypurple rectangles.

Yeast and human cells selectively deficient in the PTS1or PTS2 import pathway have been used to identify the PTSreceptors (PTSR). PTS1 signals are recognized by thePTS1R now collectively referred to as Pex5p, atetratricopeptide repeat (TPR) protein of 64-69 kDa. It isstill unclear whether this protein is localized to thecytoplasm, the outer face of the peroxisomal membrane, orthe peroxisomal matrix (van der Leij et al., 1993; Dodt etal., 1995; Szilard et al., 1995; Wiemer et al., 1995).Independent reports from three different laboratories nowsuggest that the src-homology domain 3 (SH3 domain) ofthe peroxisomal membrane protein Pex13p functions as adocking site for the mobile cytosolic PTS1R Pex5p tofacilitate the delivery of PTS1 containing proteins(Elgersma et al., 1996; Erdmann & Blobel, 1996). PTS2signals are recognized by the PTS2R termed Pex7p. Againit is unclear whether this protein resides in the cytosol orthe peroxisomal matrix (Zhang & Lazarow, 1995). Theclinical documentation of a series of similar humanperoxisomal disorders (i.e. Zellweger syndrome, neonataladrenoleukodystrohy, rhizomelic chondrodysplasia punctata(RCDP) etc.) has led to the identification of the humanhomolog of PTS1R (Dodt et al., 1995; Wiemer et al.,1995). The PTS1 and PTS2 pathways may be linkedthrough a direct interaction between the tetratricopeptiderepeat (TPR) region (a recently identified protein-protein

interaction motif) of PTS1R and the WD40 repeats (anotherdistinct protein-protein interaction motif) of PTS2Ralthough rigorous biochemical evidence for such aninteraction has not yet been reported (Rachubinski &Subramani, 1995).

Unlike most other protein translocation systemsperoxisomes are capable of importing stably folded (Waltonet al., 1995) or even oligomeric proteins (Glover et al.,1994; McNew & Goodman, 1994). How these proteinsactually cross the membrane is unknown. One possibility isthat peroxisomes contain very large pores, but noexperimental evidence for the existence of such pores hasbeen reported. Alternatively, some form of pino- orendocytosis at the peroxisomal membrane might beinvolved in the protein transport process. It is also possiblethat most peroxisomal proteins are imported into as yetunidentified peroxisomal precursors and that peroxisomes arederived from these precursors by maturation. We also knowvery little about the energetics of protein transport intoperoxisomes although ATP hydrolysis is required for theimport of proteins into the matrix (Subramani, 1996).Thus, protein translocation into peroxisomes turns out toobey somewhat different rules than the protein translocationsystems of mitochondria (Haucke & Schatz, 1997) or theendoplasmic reticulum (Rapoport et al., 1996).


703

B. Human peroxisomal disorders

Human peroxisomal biogenesis disorders occur with arelatively high frequency of about 1/50 000 live births andare a genetically heterogenous group of autosomal,recessive, lethal diseases that fall into at least eleven distinctcomplementation groups as identified by cell fusioncomplementation analysis. Twelve out of the sixteen PBDsknown to date are associated with severe neurologicaldisability while even patients suffering from the remainingPBDs show some sort of neurological defect. Thesedisorders can be grouped into three different classes, A,B,and C, according to the molecular defect leading to thedisease. In group C the subcellular localization or activity ofa single peroxisomal protein or enzyme is compromised.These disorders usually show the least severe phenotype.Patients with group A or B disorders often exhibit thepresence of non-functional peroxisome ghosts which miss afew or many peroxisomal matrix proteins due to deficienciesin either the PTS1 or PTS2 or both protein importpathways (Rachubinski and Subramani, 1995). We will nowturn to a more detailed analysis of the defects associated withthese classes of disorders.

1. Zellweger Syndrome

Zellweger syndrome (ZS), a rare fatal disorder innewborn infants was originally described by Goldfischer etal. in 1973. It is an inherited metabolic disease associatedwith a number of cerebral, hepatic and renal defects andbelongs to group A of the peroxisomal biogenesis disorders.Cells isolated from ZS patients have peroxisome ghostslacking many peroxisomal matrix proteins and thesepatients show elevated levels of very long-chain fatty acidsand are deficient in plasmalogens (ether lipids). ZS is themost severe PBD known to date and is invariably fatal. Anumber of similar diseases such as neonataladrenoleukodystrophy and infantile refsum disease have beendescribed all of which show a related but less severephenotype compared to ZS.

It appears that a number of mutations can lead to ZS andcells belonging to these various complementation groupsshow differences in their capability of importing proteinsvia either the PTS1 or PTS2 pathways. Cells from patientsin complementation group2 with ZS have mutations intheir PTS1 receptor gene. Elegant studies in vitro haveshown that the human PTS1 receptor can complement theprotein import defect in these cells suggesting that themutated PTS1 receptor is indeed the cause for the disease(Dodt et al., 1995; Wiemer et al., 1995). Thus, genetherapeutic approaches may soon provide means of treatingthis horrible disease.

2. Rhizomelic chondrodysplasia punctata

Rhizomelic chondrodysplasia punctata (RCDP) is a rareautosomal recessive phenotype associated withcomplementation group 11 of the peroxisome biogenesisdisorders and is characterized by severe growth failure,

profound developmental delay, cataracts, rhizomelia, and asevere deficiency in plasmalogens (Braverman et al., 1997).Cells from RCDP patients are unable to importperoxisomal thiolase, an enzyme targeted to peroxisomesvia the PTS2 pathway.

Recently, the molecular defects leading to RCDP havebeen elucidated. Analysis of cells from RCDP patients haverevealed a number of mutations within a single gene withhomology to the yeast PTS2 receptor (Baverman et al.,1997; Motley et al., 1997; Purdue et al., 1997). Subsequentcloning identified this gene as the human PTS2R, Pex7(Braverman et al., 1997; Motley et al., 1997; Purdue et al.,1997). Expression of human Pex7 in RCDP cells rescuesPTS2 targeting and restores the activity ofdihydroxyacetone-phosphate acyltransferase, a peroxisomalenzyme of plasmalogen synthesis (Purdue et al., 1997).

The two pathways of protein import into peroxisomesmay however not be completely separate since several ZSpatients with defective PTS1 receptors also show reducedamounts of PTS2 targeted enzymes. Moreover, an isoformof the human PTS1 receptor Pex5 is required for theefficient import of PTS2 targeted proteins and thetetratricopeptide repeats (TPR) of Pex5 directly interact withthe WD40 domain of Pex7 in the two-hybrid system(Braverman et al., 1997). It is therefore possible that themolecular defects associated with some complementationgroups of PBDs may result from a mutant receptor whichnot only is unable to bind its import substrates but mayadditionally fail to associate or cooperate with the other PTSreceptor protein resulting in multiple import defects.

3. Peroxisome-to-mitochondrion mistargeting

An interesting example of a group C peroxisomebiogenesis disorder which is caused by the mistargeting of asingle peroxisomal protein is represented by primaryhyperoxaluria type 1 (PH1). PH1 is an autosomal recessivedisease associated with a normally occuring P11Lpolymorphism and a PH1-specific G170R mutation in thegene encoding for the homodimeric enzymealanine:glyoxylate aminotransferase 1 (AGT) (Leiper et al.,1996). The P11L substitution creates an amino-terminalmitochondrial targeting signal which competes with itscarboxy-terminal peroxisomal import signal and in vitro issufficient to direct the protein into mitochondria. Thismutation alone does not interfere with the peroxisomaltargeting of AGT in living cells. AGT containing bothmutations, however, is mistargeted to mitochondria both invitro and in vivo. Recent work has now shed light on thisphenomenon: the G170R mutation abolishes the ability ofthe protein to form homodimers in the cytosol and therebyprevent its mistargeting to mitochondria which are unable toimport fully folded or dimeric proteins (Haucke and Schatz,1997). Thus, mistargeting is due to the unlikely occuringpolymorphism that generates a functionally weakmitochondrial targeting signal and a disease-specificmutation which, in combination with the polymorphism,inhibits AGT dimerization and therefore allows the proteinto cross the mitochondrial membranes.


704

Figure 2: Components involved inthe formation of a clathrin-coated bud

that mediates synaptic vesicleendocytosis. The individual proteinsare indicated by differentially colored

symbols.

III. Membrane trafficking at the nerveterminal & disease: a putative link betweensynaptic vesicle endocytosis and thebiology of cancer

I will now turn to the description of the mechanism bywhich synaptic vesicles are retrieved and recycled at theplasma membrane and discuss a number of recentobservations which suggest that endocytosis, and byextrapolation also other membrane trafficking events mayplay an important role in regulating signal transductionpathways which in turn are intimately linked to the biologyof cancer, Alzheimer's disease and other major humandiseases.

A. The synaptic vesicle cycle

Synaptic vesicles (SV) are specialized secretoryorganelles involved in synaptic transmission in the nervoussystem. Upon stimulation SVs dock and fuse with theplasma membrane and release their content into the synapticcleft. Membrane fusion occurs by a closely similarmechanism from yeast to neurons, and is mediated byspecific pairing of SNARE proteins on the two membranesundergoing fusion (Ferro-Novick and Jahn, 1994).Following exocytosis, SV membranes are retrieved andreused for the generation of new SVs. This entire cycleoccurs with high specificity and can be very rapid (less thanone minute) (Ryan, 1996).

The most widely accepted model for how SVs are beingregenerated proposes that SVs are retrieved through clathrin-mediated endocytosis (Cremona and De Camilli, 1997)involving a coat complex consisting of the heavy and light

chains of clathrin, the plasma membrane-specific adaptorcomplex AP2 (a heterotetramer composed of ",!, µ and #subunits) and the accessory protein AP180.

The importance of clathrin coats in SV endocytosis hasrecently been corroborated by genetic studies in Drosophilaand C. elegans. It is still unclear what recruits this complexto the membrane, but one possibility is that synaptotagmin(Zhang et al., 1994), an abundant protein of SVs mayfacilitate this process by interacting with the AP2 adaptor.However, both AP2 and AP180 have been found to interactdirectly with membrane phosphoinositides (PIs) indicatingthat both lipid and protein may participate in anchoring thecoat to membranes.

Vesicle fission of the mature coated bud is then effectedby the recruitment and oligomerization of the GTPasedynamin to the stalk of endocytic pits. Upon hydrolysis ofGTP dynamin disassembles and the clathrin-coated vesiclepinches off to eventually re-enter the pool of SVs awaiting astimulus for another round of exocytosis. This step may beaided or regulated by the inositol 5-phosphatasesynaptojanin (Mc Pherson et al, 1996), which is selectivelyconcentrated in nerve terminals in association with endocyticintermediates of SV membranes.

Recent evidence suggests that the SH3 domain of thenerve terminal phosphoprotein amphiphysin I (Bauerfeind etal, 1997), together with its partner protein amphiphysin IIplays an important role in recruiting dynamin to theinvaginated endocytic pit (Shupliakov et al., 1997; Wigge etal., 1997). Through its affinity for both, dynamin and theadaptor AP2, amphiphysin may link the assembly of theclathrin coat to the formation of dynamin rings, therebycoordinating these two events leading to the generation ofcalthrin coated vesicles (David et al., 1996; Ramjaun et al.,1997; Wigge et al., 1997).


705

B. A putative link between endocytosis andthe biology of cancer

Amphiphysin I is a neuron-specific protein which wasoriginally found as a component associated with SVs(Lichte et al., 1992) and as the autoantigen in a subgroup ofpatients suffering from Stiff-man syndrome (SMS) (DeCamilli et al., 1993). Stiff-man syndrome is a rare diseaseof the central nervous system characterized by painfulspasms of limbs, trunk and abdominal muscles (Layzer,1988). Group II patients are characterized by the presence ofautoantibodies against amphiphysin I and all suffer frombreast cancer (Folli et al., 1993). In an effort to elucidate theconnection between amphiphysin I autoimmunity and cancerFloyd et al. (submitted for publication) have analyzed theexpression of amphiphysin I in breast cancer tissues.Amphiphysin I was present as an alternatively spliced,overexpressed 108 kDa isoform in several breast cancertissues and as two 128 and 108 kDa forms in the breastcancer of a SMS patient. Although it is not yet clearwhether the high amphiphysin expression level is directlylinked to the enhanced proliferation of the malignant cells,the observation that amphiphysin I is overexpressed in someforms of cancer supports the idea that amphiphysin familymembers play a role in the biology of cancer cells. It is wellconceivable that overexpression of a mutant proteininvolved in endocytosis could alter signaling cascadesinitiated by endocytosed plasma membrane receptors andcould thereby lead to tumorigenesis as described below.

Another link between endocytosis and signal-dependentcell proliferation has recently emerged from studies intransfected mammalian cells. First, inactivation of theclathrin- and dynamin-dependent uptake of the receptor forepidermal growth factor (EGFR) by overexpressing a mutantform of dynamin leads to enhanced proliferation of theseendocytosis-defective cells (Vieira et al., 1996). The alteredproliferative response is presumably due to thehyperphosphorylation of a subset of EGF-dependent signaltransducing molecules suggesting an important role forEGFR signaling in establishing and controlling specificsignaling pathways.

Second, Grb2, an SH3-SH2-SH3 domain containingprotein involved in transducing signals from growth factorreceptors (i.e. EGFR) to the Ras pathway upon stimulationwith EGF transiently associates with dynamin, a GTPaseinvolved in vesicle fission from the plasma membrane (asdescribed above) (Wang and Moran, 1996). The transientinteraction between dynamin and Grb2 is required for theinternalization of the EGFR as microinjection of a peptidecorresponding to the Grb2 SH3 domain blocks endocytosis.Thus, activation and termination of EGF signaling appear tobe regulated by the diverse interactions of Grb2 with eithersignal transducing or endocytic components providinganother link between endocytosis and the attenuation ofsignal transduction events from the plasma membrane.

IV. Perspectives

The examples described in this article are just some outof a growing number of studies on how mislocalization ofcertain proteins due to genetic alterations either in theprotein itself or in its targeting machinery or perturbation ofmembrane trafficking pathways may lead to disease.Although many of the described connections betweenmembrane traffic, complex inherited disorders, signal-mediated growth control, and pathogenesis remainmechanistically poorly understood accumulating evidencesuggests that the biogenesis of membranes and thetrafficking of organelles and molecules within the cell maybe intimately linked to the regulatory and signaltransduction networks governing the physiological state of acell. A better understanding of this crosstalk mighteventually lead to improved treatments for today's diseasesincluding cancer, Alzheimer's disease, diabetes and others.

Acknowledgements

The author was supported by a long-term fellowshipfrom the European Molecular Biology Organization(EMBO) and currently holds a long-term fellowship fromthe Human Frontier Science Program (HFSP).

References

Braverman, N., Steel, G., Obie, C., Moser, A., Moser, H.,Gould, S. J., and Valle, D. (1 9 9 7 ). Human PEX7 encodesthe peroxisomal PTS2 receptor and is responsible forrhizomelic chondrodysplasia punctata. Nature Genet . 15,369-376.

Bauerfeind, R., Takei, K., and De Camilli, P. (1 9 9 7 )Amphiphysin I is associated with coated endocyticintermediates and undergoes stimulation-dependentdephosphorylation in nerve terminals. J B i o l C h e m , InPress.

Cremona, O., and De Camilli, P. (1 9 9 7 ). Synaptic vesicleendocytosis. Curr. Opin. Neurobiol . 7, 323-330.

David, C., Mc Pherson, P. S., Mundigl, O., and De Camilli, P.(1 9 9 6 ). A role of amphiphysin in synaptic vesicleendocytosis suggested by its binding to dynamin in nerveterminals. Proc. Natl . Acad. Sci . USA 93, 331-335.

De Camilli, P., Thomas, A., Cofiell, R., Folli, F., Lichte, B.,Piccolo, G., Meinck, H.-M., Austoni, M., Fassetta, G.,Bottazzo, G., Bates, D., Cartlidge, N., Solimena, M., andKilimann, M. W. (1 9 9 3 ). The synaptic vesicle-associatedprotein amphiphysin is the 128-kD autoantigen of Stiff-Man Syndrome with breast cancer. J . E x p . Med. 178,2219-2223.

Dodt, G., Braverman, N., Wong, C., Moser, A., Moser, H. W.,Watkins, P., Valle, D., and Gould, S. J. (1 9 9 5 ). Mutationsin the PTS1 receptor gene, PXR1, define complementationgroup 2 of the peroxisome biogenesis disorders. Nat.Genet . 9, 115-125.

Elgersma, Y., Kwast, L., Klein, A., Voorn-Brouwer, T., van denBerg, M., Metzig, B., America, T., Tabak, H., and Distel, B.(1 9 9 6 ). The SH3 domain of the Saccharomyces cerevisiaeperoxisomal membrane protein Pex13p functions as adocking site for Pex5p, a mobile receptor for the PTS1-containing proteins. J . C e l l B i o l . 135, 97-109.


706

Erdmann, R., and Blobel, G. (1 9 9 6 ). Identification of Pex13p,a peroxisomal membrane receptor for the PTS1 recognitionfactor. J . C e l l B i o l . 135, 111-121.

Ferro-Novick, S., and Jahn, R. (1 9 9 4 ). Vesicle fusion fromyeast to man. Nature 370, 191-193.

Folli, F., Solimena, M., Cofiell, R., Austoni, M., Tallini, G.,Fassetta, G., Bates, D., Cartlidge, N., Bottazzo, G. F.,Piccolo, G., and et al. (1 9 9 3 ). Autoantibodies to a 128-kdsynaptic protein in three women with the stiff-mansyndrome and breast cancer. N. Engl . J . Med . 328, 546-51.

Glover, J. R., Andrews, D. W., and Rachubinski, R. A. (1 9 9 4 ).Sacharomyces cerevisiae peroxisomal thiolase is importedas a dimer. P r o c . N a t l . A c a d . S c i . USA 91, 10541-10545.

Haucke, V., and Schatz, G. (1 9 9 7 ). Protein import intomitochondria and chloroplasts. Trends Ce l l B i o l . 7,103-106.

Layzer, R. B. (1 9 8 8 ). Stiff-man syndrome: an autoimmunedisease? N. Engl . J . Med . 318, 1060-1063.

Lichte, B., Veh, R. W., Meyer, H. E., and Kilimann, M. W.(1 9 9 2 ). Amphiphysin, a novel protein associated withsynaptic vesicles. EMBO J. 11, 2521-2530.

Leiper, J. M., Oatley, P. B., and Danpure, C. J. (1 9 9 6 ).Inhibition of alanine:glyoxylate aminotransferase 1dimerization is a prerequisite for its peroxisome-to-mitochondrion mistargeting in primary hydroxaluria type1. J . C e l l B i o l . 135, 939-951.

McNew, J. A., and Goodman, J. M. (1 9 9 4 ). An oligomericprotein is imported into peroxisomes in vivo. J. Ce l lB i o l . 127, 1245-1257.

McPherson, P. S., Garcia, E. P., Slepnev, V. I., David, C.,Zhang, X., Grabs, D., Sossin, W., Bauerfeind, R., Nemoto,Y., and De Camilli, P. (1 9 9 6 ). A presynaptic inositol-5-phosphatase. Nature 379, 353-357.

Motley, A. M., Hettema, E. H., Hogenhout, E. M., Brites, P.,A.L.M.A., t. A., F.A., W., Baas, F., Heijmans, H. S.,Tabak, H. F., Wanders, R. J. A., and Distel, B. (1 9 9 7 ).Rhizomelic chondrodysplasia punctata is a peroxisomalprotein targeting disease caused by a non-functional PTS2receptor. Nature Genet. 15, 377-380.

Palade, G. E. (1 9 7 5 ). Intracellular aspects of the process ofprotein synthesis. Sc ience 189, 347-358.

Purdue, P. E., Zhang, J. W., Skoneczny, M., and Lazarow, P. B.(1 9 9 7 ). Rhizomelic chondrodysplasia punctata is causedby deficiency of human PEX7, a homologue of the yeastPTS2 receptor. Nature Genet. 15, 381-384.

Rachubinski, R. A., and Subramani, S. (1 9 9 5 ). How proteinspenetrate peroxisomes. Cel l 83, 525-528.

Ramjaun, A. R., Micheva, K. D., Bouchelet, I., and McPherson,P. S. (1 9 9 7 ). Identification and characterization of a nerveterminal-enriched amphiphysin isoform. J . B i o l . C h e m .272, 16700-16706.

Rapoport, T. A., Rolls, M. M., and Jungnickel, B. (1 9 9 6 ).Approaching the mechanism of protein transport across theER membrane. Curr. Op. Cell B i o l . 8, 499-504.

Ryan, T. A., Smith, S. J., and Reuter, H. (1 9 9 6 ). The timing ofsynaptic vesicle endocytosis. P r o c . N a t l . A c a d . S c i .USA 93, 5567-5571.

Shupliakov, O., Low, P., Grabs, D., Gad, H., Chen, H., David,C., Takei, K., De Camilli, P., and Brodin, L. (1 9 9 7 ).Synaptic vesicle endocytosis impaired by disruption ofdynamin-SH3 domain interactions. Sc ience 276, 259-63.

Subramani, S. (1 9 9 6 ). Convergence of model systems forperoxisome biogenesis. C u r r . O p . C e l l B i o l . 8, 513-518.

Szilard, R. K., Titorenko, V. I., Veenhuis, M., andRachubinski, R. A. (1 9 9 5 ). Pay32p of the yeast Yarrowialipolytica is an intraperoxisomal component of the matrixprotein translocation machinery. J . Ce l l B i o l . 131,1453-1469.

van der Leij., I., Franse, M. M., Elgersma, Y., Distel, B., andTabak, H. F. (1 9 9 5 ). PAS10 is a tetratricopeptide repeatprotein that is essential for the import of most matrixproteins into peroxisomes of Saccharomyces cerevisiae.Proc. Natl . Acad. Sci . USA 90, 11782-11786.

Vieira, A. V., Lamaze, C., and Schmid, S. L. (1 9 9 6 ). Controlof EGF receptor signaling by clathrin-mediatedendocytosis. Sc ience 274, 2086-2089.

Walton, P. A., Hill, P. E., and Subramani, S. (1 9 9 5 ). Import ofstably folded proteins into peroxisomes. M o l . B i o l .Ce l l 6, 675-683.

Wang, Z. X., and Moran, M. F. (1 9 9 6 ). Requirement for theadapter protein GRB2 in EGF receptor endocytosis.Sc ience 272, 1935-1939.

Whelan, J., Omahony, P., and Harmey, M. A. (1 9 9 0 ).Processing of Precursor Proteins by Plant Mitochondria.Archiv B iochem Biophys 279, 281-285.

Wigge, P., Kohler, K., Vallis, Y., Doyle, C. a., Owen, D., Hunt,S. P., and McMahon, H. T. (1 9 9 7 ). Amphiphysinheterodimers:potential role in clathrin-mediatedendocytosis. M o l . B i o l . C e l l 8, 2003-2015.

Wiemer, E.A., Nuttley, W.M., Bertolaet, B.L., Li, X., Francke,U., Wheelock, M.J., Anne, U.K., Johnson, K.R. andSubramani, S. (1 9 9 5 ) Human peroxisomal targetingsignal-1 receptor restores peroxisomal protein import incells from patients with fatal peroxisomal disorders. J .C e l l B i o l ., 130, 51-65.

Zhang, J. Z., Davletov, B. A., Südhof, T. C., and Anderson, R.G. W. (1 9 9 4 ). Synaptotagmin I is a high-affinity receptorfor clathrin AP-2: implications for membrane recycling.Cel l 78, 751-760.

Zhang, J.W. and Lazarow, P.B. (1 9 9 5 ) PEB1 (PAS7) inSaccharomyces cerevisiae encodes a hydrophilic, intra-peroxisomal protein that is a member of the WD repeatfamily and is essential for the import of thiolase intoperoxisomes. J . C e l l B i o l ., 129, 65-80.


707


Cdc25 protein phosphatase: regulation and its rolein cancer

Jens W. Eckstein

Mitotix, Inc., One Kendall Square, Cambridge, MA 02139 USA

________________________________________________________________________________________________

Correspondence: Jens W. Eckstein, Tel: (617) 225-0001 x255, Fax: (617) 225-0005, E-mail: [email protected]

Summary

The family of the Cdc25 dual-specific protein tyrosine/threonine phosphatases i s criticallyinvolved in cel l cycle control . The substrates of Cdc25 are cycl in-dependent kinases, which areregulated by the phosphorylation of threonine and tyrosine residues. Cdc25 regulation and activityreveals a complex network of counter-balancing mechanisms and puts i t on the crossroads offundamental cel lular events l ike cel l proliferation, cel l cycle arrest and apoptosis. Our presentknowledge of the biology and biochemistry of Cdc25 phosphatases makes them attractive targetsfor drug discovery efforts: (a) they phase critical, non-redundant cell cycle regulatory functions; (b)they are bona fide checkpoint genes; (c) they have tight substrate specificities and a well-definedmechanism of catalysis; (d) they are potential targets of at least two oncogenes (Raf1 and c-Myc)that are frequently altered in human cancers; (e) they co-operate with other oncogenes in celltransformation and thus are bona fide proto-oncogenes; and lastly (f) their expression is altered intumors.

I. Introduction: Cdc25 and the cellcycle

The role of Cdc25 as an inducer of mitosis firstemerged from studies of yeast genetics that linked thephosphorylation state of the Cdc2 cyclin-dependent kinaseto the activity of a protein phosphatase (Russell andNurse, 1986). Later, the gene product of the cdc25 genewas identified to be a dual-specificity phosphatase thatremoves inhibitory phosphorylations of Cdc2, both from ahighly conserved tyrosine residue (Tyr15) and a lessconserved threonine residue (Thr14, Figure 1 ).Homologs of the yeast gene were identified in a widevariety of organisms. This functional conservation ofCdc25 throughout evolution illustrates its fundamentalrole in controlling the cell cycle.

The regulation of proteins of the cyclin-dependentkinase (Cdk) family has been studied in great detail, andseveral cdc25 genes have been identified in mammals. Inhumans the three homologs that were isolated are Cdc25A,B and C. Cdc25 C is the mitotic inducer; its substrate isthe hyperphosphorylated complex of Cdc2/cyclin B. Thefunctions of Cdc25A and B are less clear, with theirpossible substrates ranging from Cdk4/cyclin D (Terada etal., 1995), Cdk2/cyclin E and cyclin A complexes(Hoffmann et al., 1994) to Cdc2/cyclin A and cyclin B

complexes (for a recent review on Cdc25 cell biology andbiochemistry, see Draetta and Eckstein, 1997).

Recent investigations into the regulation of Cdc25itself are beginning to shed light on an intruiging andcomplex network of players (please refer to Figure 2throughout the text). They place Cdc25 squarely on thecrossroads between cell proliferation, apoptosis, mitogenicsignal transduction, and cancer. This chapter reviewsbriefly the emerging understanding of Cdc25 regulation andits implications for human cancer.

II. Cdc25 is a phosphoprotein

The Cdc25 protein undergoes phosphorylation duringthe cell cycle (Izumi et al., 1992), a step that triggers itsphosphatase activity. The phosphorylation of all threehuman versions of Cdc25 is essential for cell cycleprogression. Several phosphorylation sites have beenmapped, suggesting the possibility that more than onekinase is involved in this regulation of Cdc25 by post-translational modification.

Cdc25 can be phosphorylated by its own substrate,cyclin-dependent kinases (Cdks). Cdc25C isphosphorylated and activated by Cdc2/cyclin B in vitro. Invivo, this activation occurs at the G2/M transition, whichinitiates mitosis (Hoffmann et al., 1993; Izumi andMaller, 1993; Strausfeld et al., 1994). Cdc25A was later

Eckstein: Cdc25 protein phosphatase in cancer

708

Figure 1 . Cdc25 is a dual-specificity protein

phosphatase activating cyclin-dependent kinases (Cdks)

Cdks bind to cyclins and arephosphorylated on three

residues. Thr160phosphorylation (Pa, shown

in green) activates the kinase.The phosphorylations on

Thr14 and Tyr15 (Pi, shown inred) are inhibitory and removed

by Cdc25, resulting in anactive kinase. The kinases

Wee1 and Myt1 are counter-acting Cdc25 and

phosphorylate Tyr15 andThr14, respectively.

Figure 2 Cdc25 and the cell cycle.

This scheme summarizes the regulation and function of human Cdc25A, B and C in the cell cyle, as described in the text. Greenarrows indicate induction, (de)phosphorylation or activation, red lines inhibition. Proteins with a Ubi tag are degraded via theubiquitin-dependent pathway. Approximate timing of events and activity of proteins is indicated by the brown dashed linessubdividing the cell cycle into G0/G1, S, and G2/M phases.

shown to be phosphorylated by Cdk2/cyclin E in vitro(Hoffmann et al., 1994). In vivo, hyperphosphorylation of

Cdc25A occurs during the S-phase (Jinno et al., 1994).These results suggest regulation of Cdc25 via a self-


709

amplifying feedback loop. Such a cooperative phenomenonhas been cited to explain the sharp rise of Cdc2 kinaseactivity at the G2/M transition (Hoffmann et al., 1993;Izumi and Maller, 1993; Strausfeld et al., 1994).

In addition to Cdks, other kinases are implicated inCdc25 phosphorylation, as well.

For example, the Raf1 kinase turned out to associatewith Cdc25. Using double immunofluorescencemicroscopy, Cdc25A and B were found to co-localize withRaf1 and Ras at the cell membrane (Galaktionov et al.,1995), a process that is dependent on serum stimulation.Raf1 kinase phosphorylates Cdc25A and B in vitro,leading to an increase in phosphatase activity (Galaktionovet al., 1995).

In a two-hybrid screen experiment, Raf1 also wasfound to be associated with members of the 14-3-3 proteinfamily, which in turn associated with Cdc25A and B(Conklin et al., 1995). 14-3-3 proteins have beenimplicated in a number of mitogenic signaling pathways,including the kinase cascade that contains Raf1 (Fantl etal., 1994; Freed et al., 1994).

Moreover, a role for 14-3-3 proteins in the regulationof Cdc25 was reported in connection with DNA damagesensing in cells. The response of cells to UV-induced DNAdamage is multifaceted. It involves induction of cyclin-

dependent kinase inhibitors, such as p21Cip1/Waf1, aswell as hyperphosphorylation of the Cdks, and ultimatelyleads to cell cycle arrest (Poon et al., 1996). Recently, anew pathway has been proposed that links the genesensing DNA damage in the yeast S. pombe—Rad3—toCdc25 activity (Furnari et al., 1997; Sanchez et al., 1997).Rad3 is related to the human ATM protein that is defectivein ataxia telangiectasia patients, a rare genetic disorderwhose varied symptoms include possibly a high risk ofdeveloping tumors (Xu and Baltimore, 1996).

DNA damage induces increased phosphorylation of theChk1 kinase by a Rad3-dependent process. Cdc25 ispotentially a direct target of Chk1, and Chk1'sphosphorylation of a specific serine residue (Ser216 inhuman Cdc25C) results in binding of Cdc25 to 14-3-3protein (Peng et al., 1997). It was proposed that 14-3-3binding sequesters Cdc25C from functionally interactingwith Cdc2, leading to a G2 arrest in the cell cycle.Regulation of Cdc25 by spatial sequestering rather thaninhibition of the phosphatase activity seems to be themain effect of the phosphorylation of Cdc25 via the Chk1kinase. The Chk1 phosphorylation site is conserved inCdc25A and B, as well, suggesting that a similarregulatory mechanism is involved in other DNA damagecheckpoints earlier in the cell cycle.

The role of the 14-3-3 proteins in connection with theRaf1 kinase is still unclear. One can speculate that 14-3-3proteins act as docking sites—or adaptors—for both Cdc25and Raf1, and that subsequent phosphorylation of Cdc25by Raf1 leads to the release and activation of Cdc25. Thisexample nicely illustrates the fine balance of counter-acting processes in cell cycle regulation. Furthermore, it

identifies Cdc25C, and possibly Cdc25A and B, as bonafide checkpoint genes.

Yet other kinases have been reported to phosphorylateCdc25 protein, suggesting that there are additionalmechanisms for coordinating the regulation of cyclin-dependent kinases with various mitotic processes, such aschromosome segregation (Kumagai and Dunphy, 1996).

III. Other regulatory mechanisms

The level of Cdc25 protein is tightly regulated by bothtranscriptional and post-translational mechanisms(Ducommun et al., 1990; Moreno et al., 1990). Inhumans, Cdc25A is expressed early in the G1 phase of thecell cycle following serum stimulation of quiescentfibroblasts (Jinno et al., 1994). Cdc25 B is expressedcloser to the G1/S transition, and Cdc25C is activated inG2 (Sadhu et al., 1990).

Recently, Galaktionov and colleagues observed thatCdc25 mRNA became more abundant following activationof the Myc proto-oncogene. They were able to show thatCdc25A, and possibly Cdc25B, are physiologicallyrelevant and direct targets of c-Myc (Galaktionov et al.,1996). Their studies suggest furthermore that Cdc25 is ageneral mediator of Myc function. Therefore, Cdc25 is notonly essential to normal cell proliferation but also forinducing Myc-dependent apoptosis.

Downregulation of Cdc25 was reported to be achievedby at least two different mechanisms in the cell: repressionand ubiquitin-dependent degradation.

As an example for repression, consider TGF-ß. Itseffect on cyclin-dependent kinase activity has beenextensively studied as a model anti-mitogenic response, inparticular in connection with cyclin-dependent kinaseinhibitors (CKIs). In a recent report, Iavarone et al.conclude that induction of the cyclin-dependent kinase

inhibitor p15Ink4B and downregulation of Cdc25A byTGF-ß constitute two complementary mechanisms ofinhibition of the cyclin D-dependent kinase (Iavarone andMassague, 1997). Their experiments indicate that Cdc25Adownregulation by TGF-ß occurs at transcription; itremains to be determined whether Myc participates in thisprocess.

Ubiquitin-dependent degradation of proteins is animportant regulatory mechanism for all sorts of cellularprocesses (reviewed in Ciechanover, 1994) and has beenfound to play a key role in the degradation of the mitoticcyclins (Glotzer et al., 1991). In a study on Cdc25degradation in S.pombe, Nefsky and Beach isolated a genenamed Pub1, which encodes an E6-AP like protein(Nefsky and Beach, 1996). E6-AP belongs to a family ofubiquitin ligases, or E3s, which assist in transferring aubiquitin molecule or a polyubiquitin chain to a targetprotein. Once the target protein is tagged with ubiquitin, itis rapidly degraded by the 26S proteasome. Cdc25 wasubiquitinated in a Pub1-dependent fashion, and loss of

Eckstein: Cdc25 protein phosphatase in cancer

710

Pub1 function lead to elevated levels of Cdc25 protein andincreased Cdc25 activity in vivo.

IV. Cooperation of Ras and Myc

The regulation of Cdk activity involves inhibitorysmall proteins (cyclin-dependent kinase inhibitors, orCKIs) from the Ink and the Waf1/Kip1/Cip1 families.Recent findings suggest, firstly, that the regulation ofCdc25 and the CKI proteins through Ras and Myc istightly interconnected and, secondly, that the cooperationof active Ras and Myc leads to accumulation of G1 Cdkactivity (Leone et al., 1997). Expression of Myc and Ras

results in a loss of p27Kip1 protein (probably throughubiquitin-dependent proteolysis) and leads to increasedCdk2/cyclin E activity. At the same time, Cdc25A isinduced by c-Myc and activated by Raf1, a downstreamtarget of Ras. This leads to a synergistic effect inremoving an inhibitory protein and inhibitoryphosphorylations on Cdk2/cyclin E, culminating ininduction of S-phase.

Interestingly, the competition between p21 and Cdc25can be demonstrated directly in binding experiments. Sahaet al. identified a consensus sequence in p21 and Cdc25that is important for their binding to Cdk complexes (Sahaet al., 1997). p21 protein directly competes with Cdc25Aand vice versa, suggesting that the two proteins utilisesimilar docking sites on the Cdk/cyclin complexes.

V. Cdc25 and cancer

Cdc25A and B have oncogenic properties. In rodentcells, human Cdc25A and Cdc25B, but not Cdc25C,phosphatases cooperate with either an activated Ras alleleor loss of Rb1 in oncogenic focus formation (Galaktionovet al., 1995). Such transformants are highly aneuploid,grow in soft agar, and form high-grade tumours in nudemice. Based upon these criteria, Cdc25A and B are bonafide cellular proto-oncogenes.

Indeed, Cdc25B mRNA is expressed at high levels in32 percent of human primary breast cancers tested(Galaktionov et al., 1995). Similar findings have comefrom breast cancer studies on Cdc25 A (M. Loda et al.,unpublished). Overexpression of Cdc25A and Cdc25B, butnot Cdc25C, has also been reported in more than 50percent of tested squamous cell carcinomas of the head andthe neck (Gasparotto et al., 1997).

Given the tight connection between Cdc25 and thewell-known oncogenes Ras and Myc, overexpression andactivation of Cdc25 might be an important feature incancer development, making Cdc25 an attractive target forfuture cancer therapy.

Acknowledgement

I thank my wife, Gabrielle Strobel, for editorialassistance.

References

Ciechanover, A. (1 9 9 4 ). The ubiquitin-proteasomeproteolytic pathway. Cel l 79, 13-21.

Conklin, D. S., Galaktionov, K., and Beach, D. (1 9 9 5 ). 14-3-3 proteins associate with cdc25 phosphatases. ProcNatl Acad Sci U S A 92, 7892-7896.

Draetta, G., and Eckstein, J. (1 9 9 7 ). Cdc25 proteinphosphatases in cell proliferation. B i o c h i m .B i o p h y s . Acta 1332, M53-M63.

Ducommun, B., Draetta, G., Young, P., and Beach, D.(1 9 9 0 ). Fission yeast cdc25 is a cell-cycle regulatedprotein. Biochem Biophys Res Commun 167, 301-309.

Fantl, W. J., Muslin, A. J., Kikuchi, A., Martin, J. A.,MacNicol, A. M., Gross, R. W., and Williams, L. T.(1 9 9 4 ). Activation of Raf-1 by 14-3-3 proteins. Nature371, 612-4.

Freed, E., Symons, M., Macdonald, S. G., McCormick, F., andRuggieri, R. (1 9 9 4 ). Binding of 14-3-3 proteins to theprotein kinase Raf and effects on its activation. Sc ience265, 1713-6.

Furnari, B., Rhind, N., and Russell, P. (1 9 9 7 ). Cdc25 mitoticinducer targeted by chk1 DNA damage checkpoint kinase[In Process Citation]. Sc ience 277, 1495-7.

Galaktionov, K., Chen, X., and Beach, D. (1 9 9 6 ). Cdc25cell-cycle phosphatase as a target of c-myc. Nature 382,511-517.

Galaktionov, K., Jessus, C., and Beach, D. (1 9 9 5 ). Raf1interaction with Cdc25 phosphatase ties mitogenic signaltransduction to cell cycle activation. Genes & Dev 9 ,1046-1058.

Galaktionov, K., Lee, A. K., Eckstein, J., Draetta, G.,Meckler, J., Loda, M., and Beach, D. (1 9 9 5 ). CDC25phosphatases as potential human oncogenes. Sc ience269, 1575-1577.

Gasparotto, D., Maestro, R., Piccinin, S., Vukosavljevic, T.,Barzan, L., Sulfaro, S., and Boiocchi, M. (1 9 9 7 ).Overexpression of CDC25A and CDC25B in head and neckcancers. Cancer Res 57, 2366-8.

Glotzer, M., Murray, A. W., and Kirschner, M. W. (1 9 9 1 ).Cyclin is degraded by the ubiquitin pathway. Nature 349,132-138.

Hoffmann, I., Clarke, P. R., Marcote, M. J., Karsenti, E., andDraetta, G. (1 9 9 3 ). Phosphorylation and activation ofhuman cdc25-C by cdc2--cyclin B and its involvement inthe self-amplification of MPF at mitosis. EMBO J 12,53-63.

Hoffmann, I., Draetta, G., and Karsenti, E. (1 9 9 4 ).Activation of the phosphatase activity of human cdc25Aby a cdk2-cyclin E dependent phosphorylation at the G1/Stransition. EMBO J 13, 4302-4310.

Iavarone, A., and Massague, J. (1 9 9 7 ). Repression of theCDK activator Cdc25A and cell-cycle arrest by cytokineTGF-beta in cells lacking the CDK inhibitor p15. Nature387, 417-22.


711

Izumi, T., and Maller, J. L. (1 9 9 3 ). Elimination of cdc2phosphorylation sites in the cdc25 phosphatase blocksinitiation of M-phase. M o l B i o l C e l l 4, 1337-50.

Izumi, T., Walker, D. H., and Maller, J. L. (1 9 9 2 ). Periodicchanges in phosphorylation of the Xenopus cdc25phosphatase regulate its activity. Mol B i o l Ce l l 3,927-939.

Jinno, S., Suto, K., Nagata, A., Igarashi, M., Kanaoka, Y.,Nojima, H., and Okayama, H. (1 9 9 4 ). Cdc25A is a novelphosphatase functioning early in the cell cycle. EMBO J13, 1549-56.

Kumagai, A., and Dunphy, W. G. (1 9 9 6 ). Purification andmolecular cloning of Plx1, a Cdc25-regulatory kinasefrom Xenopus egg extracts. Science 273, 1377-80.

Leone, G., DeGregori, J., Sears, R., Jakoi, L., and Nevins, J.R. (1 9 9 7 ). Myc and Ras collaborate in inducingaccumulation of active cyclin E/Cdk2 and E2F. Nature387, 422-6.

Moreno, S., Nurse, P., and Russell, P. (1 9 9 0 ). Regulation ofmitosis by cyclic accumulation of p80cdc25 mitoticinducer in fission yeast. Nature 344, 549-52.

Nefsky, B., and Beach, D. (1 9 9 6 ). Pub1 acts as an E6-AP-likeprotein ubiquitiin ligase in the degradation of cdc25.EMBO J. 15, 1301-1312.

Peng, C. Y., Graves, P. R., Thoma, R. S., Wu, Z., Shaw, A. S.,and Piwnica-Worms, H. (1 9 9 7 ). Mitotic and G2checkpoint control: regulation of 14-3-3 protein bindingby phosphorylation of Cdc25C on serine-216 [In ProcessCitation]. Sc ience 277, 1501-5.

Poon, R. Y. C., Jiang, W., Toyoshima, H., and Hunter, T.(1 9 9 6 ). Cyclin-dependent kinases are inactivated by acombination of p21 and Thr-14/Tyr-15 phosphorylationafter UV-induced DNA damage. J B i o l Chem 271,13283-91.

Russell, P., and Nurse, P. (1 9 8 6 ). cdc25+ functions as aninducer in the mitotic control of fission yeast. Cell 45,145-53.

Sadhu, K., Reed, S. I., Richardson, H., and Russell, P.(1 9 9 0 ). Human homolog of fission yeast cdc25 mitoticinducer is predominantly expressed in G2. Proc NatlAcad Sci U S A 87, 5139-43.

Saha, P., Eichbaum, Q., Silberman, E. D., Mayer, B. J., andDutta, A. (1 9 9 7 ). p21CIP1 and Cdc25A: competitionbetween an inhibitor and an activator of cyclin-dependentkinases. M o l C e l l B i o l 17, 4338-45.

Sanchez, Y., Wong, C., Thoma, R. S., Richman, R., Wu, Z.,Piwnica-Worms, H., and Elledge, S. J. (1 9 9 7 ).Conservation of the chk1 checkpoint pathway inmammals: linkage of DNA damage to cdk regulationthrough cdc25 [In Process Citation]. Sc ience 277, 1497-501.

Strausfeld, U., Fernandez, A., Capony, J. P., Girard, F.,Lautredou, N., Derancourt, J., Labbe, J. C., and Lamb, N.J. (1 9 9 4 ). Activation of p34cdc2 protein kinase bymicroinjection of human cdc25C into mammalian cells.Requirement for prior phosphorylation of cdc25C byp34cdc2 on sites phosphorylated at mitosis. J B i o lChem 269, 5989-6000.

Terada, Y., Tatsuka, M., Jinno, S., and Okayama, H. (1 9 9 5 ).Requirement for tyrosine phosphorylation of Cdk4 in G1

arrest induced by ultraviolet irradiation. Nature 376,358-362.

Xu, Y. and Baltimore, D. (1 9 9 6 ) Dual roles of ATM in thecellular response to radiation and in cell growth control.Genes Dev 10, 2401-2410.


713


Nucleocytoplasmic trafficking: implications for thenuclear import of plasmid DNA during gene therapy

Teni Boulikas

Institute of Molecular Medical Sciences, 460 Page Mill Road, Palo Alto, California 94306

and Regulon Inc., 249 Matadero Avenue, Palo Alto, CA 94306

__________________________________________________________________________________Correspondence : Teni Boulikas, Regulon Inc., 249 Matadero Avenue, Palo Alto, CA 94306, Tel (650) 813-9264, Fax: (650)424-9594, E-mail: [email protected]

Key words: pore complex, nucleoporins, nuclear localization signals, karyopherin, nuclear export signals, mRNA export,

Summary

Traff icking of nuclear proteins from the s i te of their synthesis in the cytoplasm to the s i tes of

function in the nucleus through pore complexes is mediated by nuclear localization signals (NLSs)

on proteins to be imported into nuclei . Protein translocation from the cytoplasm to the

nuc leoplasm involves ( i ) the format ion of a complex o f karyopherin with NLS-protein, ( i i )

subsequent binding of karyopherin , ( i i i ) b inding o f the complex to FXFG pept ide repeats on

nucleoporins, (iv) docking of Ran-GDP to nucleoporin and to karyopherin heterodimer by p10, (v)

a number of association-dissociation reactions on nucleoporins which dock the import substrate

toward the nucleoplasmic side with a concomitant GDP-GTP exchange reaction transforming Ran-

GDP into Ran-GTP and catalyzed by karyopherin , and finally (vi) dissociation from karyopherin

and release of the karyopherin /NLS-protein by Ran-GTP to the nucleoplasm. A number of

processes have been found to be regulated by nuclear import including nuclear translocation of the

transcription factors NF- B, rNFIL-6, ISGF3, SRF, c-Fos, GR as well as human cyclins A and B1,

casein kinase II, cAMP-dependent protein kinase II, protein kinase C, ERK1 and ERK2. Failure of

cel ls to import specif ic proteins into nuclei can lead to carcinogenesis . For example, BRCA1 is

mainly localized in the cytoplasm in breast and ovarian cancer cells whereas in normal cells the

protein is nuclear. mRNA is exported through the same route as a complex with nuclear proteins

possess ing nuclear export s ignals (NES) . The majori ty of prote ins with NES are RNA-binding

prote ins which bind to and escort RNAs to the cytoplasm. However , other proteins with NES

function in the export of proteins; CRM1, which binds to the NES sequence on other proteins and

interacts with the nuclear pore complex, is an essential mediator of the NES-dependent nuclear

export of proteins in eukaryotic cells. Nuclear localization and export signals (NLS and NES) are

found on a number of important molecules including p53, v-Rel, the transcription factor NF-ATc,

the c-Abl nonreceptor tyrosine kinase, and the fragile X syndrome mental retardation gene product;

the deregulation of their normal import/export trafficking has important implications for human

disease. Both nuclear import and export processes can be manipulated by conjugation of proteins

with NLS or NES pept ides . During gene therapy the fore ign DNA needs to enter nucle i for i t s

transcription; a pathway is proposed involving the complexation of plasmids and oligonucleotides

with nascent nuclear proteins possessing NLSs as a prerequisite for their nuclear import. Covalent

linkage of NLS peptides to oligonucleotides and plasmids or formation of complexes of plasmids

with proteins possessing multiple NLS peptides is proposed to increase their import rates and the

efficiency of gene expression. Cancer cells are predicted to import more efficiently foreign DNA

into nuclei compared with terminally differentiated cel ls because of their increased rates of

proliferation and protein import.


714

I. Introduction

Evolution has effectively secluded nuclear functionsfrom cytoplasmic activities by the nuclear envelope barrierin order to circumvent the increased organizational andregulatory problems associated with the advent of the largegenomes in more complex organisms. The nuclearenvelope has evolved to allow nuclear and cytoplasmicenvironments to be developed and to selectively keepcertain regulatory proteins, such as replication factors, out,allowing cells to regulate their cell cycle and level ofploidy. This double-membrane structure effectivelyseparates transcription of genes from translation of theirmRNA into proteins and allows proteins destined tofunction in the nucleus to pass selectively through thelumen of the nuclear pores (reviewed by Burke, 1990;Boulikas, 1993, 1994; Laskey et al, 1996).

The nuclear membrane prevents reinitiation of DNAreplication in Xenopus eggs, by excluding a "licensingfactor" that is essential for DNA replication; replicationlicensing may involve the MCM (minichromosomemaintenance) complex and ORC, the origin recognitioncomplex (Laskey et al, 1996).

The interior of the nuclear envelope is lined by apolymer of lamins intimately attached to chromatin DNA.Phosphorylation of lamins at four serines by S6 kinase IIand other kinases, all of which appear to be controlled bycdc2 kinase, specifically occur as cells traverse the G2 toM checkpoint of the cell cycle; this process result in lamindepolymerization and in nuclear envelope breakdown. Aftercompletion of mitosis these processes are reversed and newnuclear envelopes are assembled around daughter cellnuclei.

Selective transport through pores creates a uniquebiochemical environment within the nucleus. All proteinsare synthesized in the cytoplasm; the selective import ofproteins through the pore complexes that straddle the innerand outer nuclear membranes is a sophisticated processdependent on energy and upon the presence of shortkaryophilic peptides, termed nuclear localization signals(NLS), only on nuclear proteins (Dingwall et al, 1982;Kalderon et al, 1984).

The ultimate target of gene therapy is the cell nucleus.The knowledge on nucleocytoplasmic trafficking could beused for enhancing the nuclear import of plasmids or smalloligonucleotides designed to act in the nucleus. Since onlya small portion (less than 1%) of the plasmid moleculesthat reach the cytoplasm might ultimately enter thenucleus, and only 15% of water soluble oligonucleotideswhich reach the cytoplasm might ultimately diffusethrough pore complexes (Boutorine and Kostina, 1993),covalent linkage of NLS peptides via random-coil peptidearms to oligonucleotides or to plasmids is expected toincrease their import rates and the efficiency of expression

of their therapeutic gene loads. In addition, complexationof the plasmid into colloidal particles with proteinspossessing multiple NLS peptide motifs is proposed tofacilitate nuclear import.

II. Morphology of the pore complexes

The nuclear envelope is often studded across both sideswith transcisternal "holes." These hollow cylindricalorganelles spanning the two nuclear membranes are calledpore complexes. Pore complexes have a width (distancefrom cytoplasm to nucleoplasm) of ~70 nm and a diameterof 133 nm (Hinshaw et al 1992). Their frequency greatlydepends on the cell type ranging from 1 to 60 pores/mm2.

Their refined model structure (Hinshaw et al, 1992) israther complex (Figures 1-4). They appear as tripartitestructures composed of two concave rings and a centralgranule. The concave rings are called pore annuli (onecytoplasmic and one nucleoplasmic), each containing eightgranules arranged in an 8-fold rotational symmetry, lyingon top of the pore rims. A three-dimensional electronmicroscopy analysis coupled with image analysis tocalculate 2D and 3D maps of detergent released porecomplexes revealed that this highly symmetric frameworkis built from many distinct and interconnected subunitsarranged in such a way so as to construct a large centralchannel (Hinshaw et al, 1992, Figures 1-4).

Electron microscopy of nuclear pore complexesisolated from Xenopus laevis oocytes spread on a carbon-coated film has shown that each of the eight spokes seenin en face views of pore complexes is built from fourmorphological features: the annular, the column, the ring,and the lumenal subunits (Hirshaw et al, 1992). Eachspoke holds two copies of each subunit. Furthermore, anintricate network connects these subunits to one another.In addition to the large central channel, the spokes areresponsible for the construction of eight peripheralchannels of unknown function (Hinshaw et al, 1992).Image analysis of spokes on en face electron micrographsof pore complexes (Figure 2a ) show that they arecomposed of (i ) bilobed regions that form an inner annulusencircling a 42 nm diameter hole; (i i ) a central region of aradius of 41 nm and (i i i ) an outer region of 52.5 nm . Themaps obtained by Hinshaw and coworkers (1992) providecompelling evidence for a highly symmetric structure ofpore complexes in accordance with previous studies(Unwin and Milligan 1982; Akey, 1989). In the oocytes ofthe frog Xenopus laevis the pore annuli have an insidediameter close to 80 nm and an outer diameter of 120 nm.The inner diameter of the pore complexes is highlyconstant within a certain cell type.

A total of up to 100 distinct proteins (nucleoporins)have been estimated to participate in the structure of the


715

Figure 1 . Electron micrographs of nuclear pore complexes, (NPCs), released from the nuclear envelope by detergent. (a) Enface views of NPCs (lower right) and rings (upper left) in the same field of view. (b ) Image of a deep pool of stain; the two porecomplexes indicated by arrows represent edge views whereas the arrowheads show oblique views of NPCs (see also the model inFigure 3); a number of en face views are also present in the same field of view. Scale bar = 500 nm. Image by courtesy of RonMilligan and Jenny Hinshaw, The Scripps Research Institute, La Jolla, California. From Hinshaw JE, Carragher BO, Milligan RA(1 9 9 2 ) Architecture and design of the nuclear pore complex. Cel l 69, 1133-1141. Reproduced with kind permission from CellPress and the authors.

Figure 2 . Projection maps obtained by averaging images of each of the four structures identified in Figure 1. Regions wherebiological material is concentrated are darker and are enclosed by contours; regions where the negative stain is concentrated arelighter. (a) Average of 168 (n=168) en face images of detergent-released NPCs. (b ) Edge v iew of detergent released NPCs,n=48. (c ) Ring v i ews (n=400). (d) Intermediate structures (n=23). Scale barr = 50 nm. Image by courtesy of Ron Milliganand Jenny Hinshaw, The Scripps Research Institute, La Jolla, California. From Hinshaw JE, Carragher BO, Milligan RA (1 9 9 2 )Architecture and design of the nuclear pore complex. Cel l 69, 1133-1141. Reproduced with kind permission from Cell Press andthe authors.


717

Figure 3 Renderings of the 822-symmetrized nuclearpore complex map. En face (a), oblique (b ), edge (c ), and fronthalf view (d) of the 3D map. A slight ridge indicates thecentral plane of the 3D map and divides the assembly into twosymmetrical halves. Manipulation and display of the mapswere done with specific programs. Annular subunits are green,rings are yellow, and lumenal subunits are blue. The remainingtan colored parts of the 3D map enclose the column subunits.Image by courtesy of Ron Milligan and Jenny Hinshaw, TheScripps Research Institute, La Jolla, California. FromHinshaw JE, Carragher BO, Milligan RA (1 9 9 2 ) Architectureand design of the nuclear pore complex. Cel l 69, 1133-1141.Reproduced with kind permission from Cell Press and theauthors.

pore complex (Iovine et al, 1995). A pore-specifictransmembrane glycoprotein gp210 (mw 204 kD) withtwo transmembrane domains and 13 glycosylation sites isthought to anchor the pore complex to the membrane ofthe nuclear envelope (Wozniak et al, 1989).

The central granule, exclusive route for import ofnuclear protein, occupies the pore center. Its diametervaries from 2.5 to 35 nm and its appearance ranges fromcompact spherical to thin rod shaped (Akey, 1989).Fibrils, which protrude deeply into the nuclear interiorforming a central ring of spokes, emanate from the centralgranule. The fibrils, ~3nm in diameter and extending~200nm to the interior of the nucleus, are involved innuclear transport: nucleoplasmin-coated gold particlesassociate with these tentacles (Richardson et al, 1988) andare docking sites for the karyopherin ! NLS-proteincomplex (Rexach and Blobel, 1995). Antibodies to the O-linked glycoproteins seem to bind close to the 8-fold axisand away from the central plane of the NPC (Snow et al,1987); thus, it is unlikely for those glycoproteins,involved in nuclear import, are integral parts of the spokesubunits (Hinshaw et al, 1992).

III. Nuclear localization signals (NLSs)

A. Historical background

The selective import of proteins that have a function inthe nucleus through the pore complexes that straddle theinner and outer nuclear membranes is a sophisticatedprocess dependent on the presence of short karyophilicpeptides, termed nuclear localization signals (NLS)(reviewed by Boulikas, 1993, 1994, 1996, 1997b). A veryshort sequence of seven amino acids (Pro-Lys-Lys-Lys-Arg-Lys-Val or PKKKRKV), first recognized by Kalderonand coworkers (1984) in the SV40 large T antigen, isrequired for its normal nuclear localization.

Yoneda et al. (1988) have raised antibodies against thepeptide DDDED supposed to be present in nuclear pore orcytoplasmic receptor (transporter) protein molecules and tobe involved in ionic interactions with the NLS (KKKRK)of SV40 large T protein. Indirect immunofluorescencewith these antibodies against the acidic peptide has shownpunctuate staining at the nuclear rim or the nuclear surfacein rat, human, bovine and murine cell lines; in addition,the antibody blocked nuclear import.

A single protein may possess more than one signalsfor nuclear import (Standiford and Richter, 1992). The rateof nuclear import is directly related to the number of NLSit possesses (Dworetzky et al, 1988), as was firstsuggested by Dingwall and coworkers (1982). A smallernumber of nuclear proteins contain "bipartite" NLShypothesized to be reconstituted by two moieties broughttogether by protein folding or conformational change as for

Boulikas: Nucleocytoplasmic trafficking

718

example on the cytoplasmic glucocorticoid receptor byhormone binding (Welsh et al, 1986; Picard andYamamoto, 1987).

Three approaches have been used for NLSidentification: (i ) gene fusion experiments between a NLS-coding DNA segment and the gene coding for puruvatekinase, "-galactosidase, or other cytoplasmic proteins; (i i )nuclear import of non-nuclear proteins conjugated tosynthetic NLS peptides; (i i i ) site-directed mutagenesis ofthe NLS of a nuclear protein resulting in its cytoplasmicretention (Boulikas, 1993; Tables 1-4).

B. Rules to predict nuclear localization ofan unknown protein

Several simple rules have been proposed for theprediction of the nuclear localization of a protein of anunknown function from its amino acid sequence:

(i ). An NLS is defined as four arginines (R) pluslysines (K) within an hexapeptide; the presence of one ormore histidines (H) in the tetrad of the karyophilichexapeptide, often found in protein kinases that have a

Figure 4 oblique map of the pore complex (Figure 3b) superimposed over an electron micrograph of nuclear pore complexesshowing the central plug of the structures. Image by courtesy of Ron Milligan and Jenny Hinshaw, The Scripps Research Institute,La Jolla, California. From Hinshaw JE, Carragher BO, Milligan RA (1 9 9 2 ) Architecture and design of the nuclear pore complex.

Cel l 69, 1133-1141. Reproduced with kind permission from Cell Press and the authors.


718

cytoplasmic and a nuclear function, may specify a weakNLS whose function might be regulated byphosphorylation or may specify proteins that function inboth the cytoplasm and the nucleus (Boulikas, 1996).

(i i ). The K/R clusters are flanked by the !-helixbreakers G and P thus placing the NLS at a helix-turn-helix or end of an !-helix. Negatively-charged amino acids(D, E) are often found at the flank of the NLS and on someoccasions may interrupt the positively-charged NLScluster.

(i i i ). Bulky amino acids (W, F, Y) are not presentwithin the NLS hexapeptide.

(iv ). NLS signals may not be flanked by longstretches of hydrophobic amino acids (e.g. five); a mixtureof charged and hydrophobic amino acids serves as amitochondrial targeting signal.

(v ). The higher the number of NLSs the more readily amolecule is imported to the nucleus (Dworetzky et al,1988). Even small proteins, for example histones (10-22kDa), need to be actively imported to increase their importrates compared with the slow rate of diffusion of smallmolecules through pores.

(vi ). Signal peptides are stronger determinants thanNLSs for protein trafficking; signal peptides direct proteinsto the lumen of the endoplasmic reticulum for theirsecretion or insertion into cellular membranes (presence oftransmembrane domains) (Boulikas, 1994).

(vi i ). Signals for the mitochondrial import of proteins(a mixture of hydrophobic and karyophilic amino acids)may antagonize nuclear import signals and proteinspossessing both type of signals may be translocated toboth mitochondria and nuclei (Beasley and Schatz, 1991;Neupert and Lill, 1995).

(v i i i ). Strong association of a protein with largecytoplasmic structures (membrane proteins, intermediatefilaments) make such proteins unavailable for import eventhough they posses NLS-like peptides (Boulikas, 1994).

(ix ). Transcription factors and other nuclear proteinsposses a great different number of putative NLS stretches;of the sixteen possible forms of putative NLS structuresthe most abundant types are the ##x##, ###x#, ####, and##x#x# where # is R or K, together accounting for about70% of all karyophilic clusters on transcription factors(Boulikas, 1994).

(x ). A small number of nuclear proteins seem to bevoid of a typical karyophilic NLS; in this case either nonkaryophilic peptides function for their nuclear import, suchmolecules possess bipartite NLSs, or these NLS-lessproteins depend absolutely for import on their strongcomplexation in the cytoplasm with a nuclear proteinpartner able to be imported (Boulikas, 1994); this

mechanism might ensure a certain stoichiometric ratio ofthe two molecules in the nucleus and might be ofphysiological significance.

(xi ) A number of proteins may be imported via othermechanisms not dependent on classical NLS (se below).

C. NLS on adenovirus proteins

The pentapeptide KRPRP of Adenovirus E1a whenlinked to the C-terminus of E. coli galactokinase, wassufficient to direct its nuclear accumulation aftermicroinjection into Vero monkey cells (Lyons et al.,1987). The synthetic peptide CGGLSSKRPRP fromadenovirus type 2/5 E1a crosslinked to chicken bovinealbumin and microinjected into HeLa cells caused nuclearlocalization (Chelsky et al., 1989).

Two NLS, PPKKRMRRRIE and PKKKKKRPwere found on adenovirus 5 DBP (DNA-binding protein)which is expressed in nuclei of infected cells and isinvolved in virus replication and early and late geneexpression. Both NLS are needed, and disruption of eithersite impaired nuclear localization of the 529 amino acidprotein (Morin et al., 1989).

The NLS RLPVRRRRRRVP was determined onadenovirus pTP1 and pTP2 (preterminal proteins, 80kD) between amino acid residues 362-373. The 140 kDaDNA polymerase of adenovirus when it had lost its ownNLS could enter the nucleus via its interaction with pTP.This NLS, fused to the N-terminus of E. coli "-galactosidase, was functional in nuclear targeting (Zhaoand Padmanabhan, 1988).

A "tripartite" or "doubly bipartite" NLS was found onadenovirus DNA polymerase (AdPol) having thesequences: signal I: AHRARRLH (amino acids 6-13);signal II: PPRRRVRQQPP (amino acids 23-33); andsignal III: PARARRRRAP (amino acids 39-48). SignalsI and II functioned interdependently as an NLS for thenuclear targeting of AdPol, for which signal III wasdispensable. The combined signal II-III was more efficientNLS than signal I-II (Zhao and Padmanabhan, 1991).

IV. Nucleoporins

A number of nucleoporins (proteins of the porecomplex) posses FXFG motifs and display modification ofSer/Thr by single N-acetyl-glucosamine residues; theseinclude Nup98, p62, Nup153, and Nup214 in vertebratesand NUP1, NUP2, and NSP1 in S. cerevisiae. A differentsubset of pore complex proteins including p270, Nup214,Nup153, and Nup98 contain FXFG and GLFG repetitivepeptide motifs and are able to bind specifically to NLS-containing protein models; a single motif may be a lowaffinity binding site and the affinity of binding could be


719

Table 1 Simple NLS

Signal oligopeptide Protein and features

PKKKRKV Wild-type SV40 large T proteinA point mutation converting lysine-128 (double underlined) to threonine results in the retention of largeT in the cytoplasm. Transfer of this peptide to the N-terminus of "-galactosidase or pyruvate kinase atthe gene level and microinjection of plasmids into Vero cells showed nuclear location of chimericproteins.

PKKKR M V SV40 large T with a K$M change. Site-directed mutagenesis only slightly impaired nuclear import oflarge T.

PKKKRKVEDP Synthetic NLS peptide from SV40 large T antigen crosslinked to BSA or IgG mediated their nuclearlocalization after microinjection in Xenopus oocytes. The PKKGSKKA from Xenopus H2B wasineffective and PKTKRKV was less effective.

CGYGPKKKRKVGG Synthetic peptide from SV40 large T antigen conjugated to various proteins and microinjected into thecytoplasm of TC-7 cells. Specified nuclear localization up to protein sizes of 465 kD (ferritin). IgM of970 kD and with an estimated radius of 25-40 nm was retained in the cytoplasm.

CYDDEA T AD S QH ST PPKKKRKVEDPK

DFESELLSSV40 large T protein long NLS. The long NLS but not the short NLS, was able to localize the bulkyIgM (970 kD) into the nucleus. Mutagenesis at the four possible sites of phosphorylation (doubleunderlined) impaired nuclear import.

CGGPKKKRKVG SV40 large T protein. This synthetic peptide crosslinked to chicken serum albumin and microinjectedinto HeLa cells caused nuclear localization.

PKKKIKV A mutated (R$I) version of SV40 large T NLS. Effective NLS.

MKx11CR L KK L KCSKEKPKCAKCLKx5Rx3KTKR

74 N-terminal amino acid

Yeast GAL4 (99 kD). Fusions of the GAL4 gene portion encoding the 74 N-terminal amino acid withE. Coli "-galactosidase introduced into yeast cells specify nuclear localization.

MKx11CRLKKLKCSKEKPKCA29 N-terminal amino acid

Yeast GAL4. Acted as an efficient nuclear localization sequence when fused to invertase but not to "-galactosidase introduced by transformation into yeast cells.

PKKAREDVSRKRPR

Polyoma large T protein. Identified by fusion with puruvate kinase cDNA and microinjection of VeroAfrican green monkey cells. Mutually independent NLS. Can exert cooperative effects.

CGYGVSRKRPRPG Polyoma virus large T protein. This synthetic peptide crosslinked to chicken serum albumin andmicroinjected into HeLa cells caused nuclear localization.

APTKRKGS SV40 VP1 capsid polypeptide (46 kD). NLS (N terminus) determined by infection of monkey kidneycells with a fusion construct containing the 5' terminal portion of SV40 VP1 gene and the completecDNA sequence of poliovirus capsid VP1 replacing the VP1 gene of SV40.

APKRKSGVSKC(1-11)

Polyoma virus major capsid protein VP1 (11 N-terminal amino acid). Yeast expression vectors codingfor 17 N-terminal amino acid of VP1 fused to "-galactosidase gave a protein that was transported tothe nucleus in yeast cells. Subtractive constructs of VP1 lacking A1 to C11 were cytoplasmic. This,FITC-labeled, synthetic peptide crosslinked to BSA or IgG, caused nuclear import after microinjectioninto 3T6 cells. Replacement of K3 with T did not.

PNKKKRK(amino acid position 317-323)

SV40 VP2 capsid protein (39 kD). The 3' end of the SV40 VP2-VP3 genes containing this peptidewhen fused to poliovirus VP1 capsid protein at the gene level resulted in nuclear import of the hybridVP1 in simian cells infected with the hybrid SV40.

EEDGPQKKKRRL(307-318)

Polyoma virus capsid protein VP2. A construct having truncated VP2 lacking the 307-318 peptidetransfected into COS-7 cells showed cytoplasmic retention of VP2. The 307-318 peptide crosslinked toBSA or IgG specified nuclear import following their microinjection into NIH 3T6 cells.

GKKRSKA Yeast histone H2B. This peptide specified nuclear import when fused to "-galactosidase.

KRPRP Adenovirus E1a. This pentapeptide, when linked to the C-terminus of E. coli galactokinase, wassufficient to direct its nuclear accumulation after microinjection in Vero monkey cells.

CGGLSSKRPRP Adenovirus type 2/5 E1a. This synthetic peptide crosslinked to chicken bovine albumin andmicroinjected into HeLa cells caused nuclear localization.

LVRKKRKTE3SP(NLS 1)LKDKDAKKSKQE (NLS2)

Xenopus N1 (590 amino acid). Abundant in X. laevis oocytes, forming complexes with histones H3, H4via two acidic domains each containing 21 and 9 (D+E), respectively. The NLS1 is required but notsufficient for nuclear accumulation of protein N1. NLS 1 and 2 are contiguous at the C-terminus.

GNKAKRQRST v-Rel or p59 v-rel the transforming protein, product of the v-rel oncogene of the avianreticuloendotheliosis retrovirus strain T (Rev-T). v-Rel NLS added to the normally cytoplasmic "-galactosidase directed that protein to the nucleus.


720

PFLDRLRRDQKPKQKRKMAR

NS1 protein of influenza A virus, that accumulates in nuclei of virus-infected cells. Determined to bean NLS by deletion mutagenesis of NS1 in recombinant SV40. The 1st NLS is conserved among allNS1 proteins of influenza A viruses.

SVTKKRKLE Human lamin A. Dimerization of lamin A was proposed to give a complex with two NLSs that wastransported more efficiently.

SASKRRRLE Xenopus lamin A. NLS inferred from its similarity to human lamin A NLS.

TKGKRKRID Xenopus lamin LI . NLS inferred from its sequence similarity to human lamin A NLS.

CVRTTKGKRKRIDV Xenopus lamin LI. This synthetic peptide crosslinked to chicken bovine albumin and microinjected intoHeLa cells caused nuclear localization.

CGGAMINO ACIDKRVKLD Human c-myc oncoprotein. This synthetic peptide crosslinked to chicken bovine albumin andmicroinjected into HeLa cells caused nuclear localization.

PAMINO ACIDKRVKLD(M1, fully potent NLS)

RQRRNELKRSP(M2, medium potency NLS)

Human c-myc oncoprotein. Conjugation of the M1 peptide to human serum albumin andmicroinjection of Vero cells gives complete nuclear accumulation. M2 gave slower and only partialnuclear localization.

SALIKKKKKMAP Murine c-abl (IV) gene product. The p160gag/v-abl has a cytoplasmic and plasma membranelocalization, whereas the mouse type IV c-abl protein is largely nuclear.

PPKKRMRRRIEPKKKKKRP

Adenovirus 5 DBP (DNA-binding protein) found in nuclei of infected cells and involved in virusreplication and early and late gene expression. Both NLS are needed, and disruption of either siteimpaired nuclear localization of the 529 amino acid protein.

YRKCLQAGMNLEARKTKKKIKGIQQATA (497-524 amino acid)

Rat GR, glucocorticoid receptor (795 amino acid) NLS1 determined by fusion with "-galactosidase(116 kD). NLS1 is 100% conserved between human, mouse and rat GR. Whereas the 407-615 aminoacid fragment of GR specifies nuclear location, the 407-740 amino acid fragment was cytoplasmic inthe absence of hormone, indicating that sequence 615-740 may inhibit the nuclear location activity. Asecond (NLS2) is localized in an extensive 256 amino acid C-terminal domain. NLS 2 requireshormone binding for activity.

RK D RRGGRMLK H KR Q RDDGEGRGEVGSAGDMRAMINOACIDNLWPSPLMIKR S KK.(amino acid 256-303)

Human ER (estrogen receptor, 595 amino acid) NLS. NLS is between the hormone-binding and DNA-binding regions; ER, in contrast with GR, lacks a second NLS. Can direct a fusion product with "-galactosidase to the nucleus.

RKFKKFNK Rabbit PG (progesterone receptor). 100% homology in humans; F$L change in chickens. When thissequence was deleted, the receptor became cytoplasmic but could be shifted into the nucleus byaddition of hormone; in this case the hormone mediated the dimerization of a mutant PG with a wildtype PG molecule.

GKRKNKPK Chicken Ets1 core NLS. Within a 77 amino acid C-terminal segment 90% homologous to Ets2. Whendeleted by deletion mutagenesis at the gene level the mutant Ets1 became cytoplasmic.

PLLKKIKQ c-myb gene product; directs puruvate kinase to the nucleus.

PPQKKIKS N-myc gene product; directs puruvate kinase to the nucleus.

PQPKKKP p53; directs puruvate kinase to the nucleus.

SKRVAKRKL c-erb-A gene product; directs puruvate kinase to the nucleus.

CGGLSSKRPRP Adenovirus type2/5 E1a. This synthetic peptide conjugated with a bifunctional crosslinker to chickenserum albumin (CSA) and microinjected into HeLa cells directed CSA to the nucleus.

MTGSK T RKHR G SGAMTGSKHRKH PG SGA

Yeast ribosomal protein L29. Double-stranded oligonucleotides encoding the 7 amino acid peptides(underlined) and inserted at the N-terminus of the "-galactosidase gene resulted in nuclear import.

RHRKHPKRRKHPKYRKHPKHRRHPKHKKHPRHLKHPKHRKYPKHRQHP

Mutated peptides derived from yeast L29 ribosomal protein NLS, found to be efficient NLS. The lasttwo are less effective NLS, resulting in both nuclear and cytoplasmic location of "-galactosidasefusion protein.

PETTVVRRR G RSPRRRTPSPRRRRSPR

RRRSQS(One sequence, C-terminus)

Double NLS of hepatitis B virus core antigen. The two underlined arginine clusters represent distinctand independent NLS. Mutagenesis showed that the antigen fails to accumulate in the nucleus onlywhen both NLS are simultaneously deleted or mutated.


721

ASKSRKRK L Viral Jun, a transcription factor of the AP-1 complex. Accumulates in nuclei most rapidly during G2and slowly during G1 and S. The cell cycle dependence of viral but not of cellular Jun is due to a C$Smutation in NLS of viral Jun. This NLS conjugated to rabbit IgG can mediate cell cycle-dependenttranslocation.

GGLCSARLHRHALLAT Human T-cell leukemia virus Tax trans-activator protein. The most basic region within the 48 N-terminal segment. Missense mutations in this domain result in its cytoplasmic retention.

DTREKKKFLKRRLLRLDE (604-620) Mouse nuclear Mx1 protein (72 kD), Induced by interferons (among 20 other proteins) . Selectivelyinhibits influenza virus mRNA synthesis in the nucleus and virus multiplication. The cytoplasmic Mx2has R$S and R$E changes in this region.

CGYGPKKKRKV(SV40 large T)CGYGDRNKKKKE(human retinoic acid receptor)CGYGARKTKKKIK(human glucocorticoid receptor)CGYGIRKDRRGGR(human estrogen receptor)CGYGARKLKKLGN(human androgen receptor)

Synthetic peptides crosslinked to bovine serum albumin (BSA) and introduced into MCF 7 or HeLa S3cells with viral co-internalization method using adenovirus serotype 3B induced nuclear import ofBSA.

RKRQRALMLRQAR30-42

Human XPAC (xeroderma pigmentosum group A complementing protein) involved in DNA excisionrepair. By site-directed mutagenesis and immunofluorescence. NLS is encoded by exon 1 which is notessential for DNA repair function.

EYLSRKGKLEL(at the N-terminus)

T-DNA -linked VirD2 endonuclease of the Agrobacterium tumefaciens tumor-inducing (Ti) plasmid.A fusion protein with "-galactosidase is targeted to the nucleus. The T-plasmid integrates into plantnuclear DNA; VirD2 produces a site-specific nick for T integration. VirD2 also contains a bipartiteNLS at its C-terminus (see Table 2).

KKSKKKRC(95-102)

Putative core NLS of yeast TRM1 (63 kD) that encodes the tRNA modification enzyme N2, N2-dimethylguanosine-specific tRNA methyltransferase. Localizes at the nuclear periphery. The 70-213amino acid segment of TRM1 causes nuclear localization of "-galactosidase fusion protein in yeastcells. Site-directed mutagenesis of the 95-102 peptide resulted in its cytoplasmic retention. TRM1 isboth nuclear and mitochondrial. The 1-48 amino acid segment specifies mitochondrial import.

PQSRKKLR Max protein; specifically interacts with c-Myc protein. Fusion of 126-151 segment of Max to chickenpyruvate kinase (PK) gene, including this putative NLS, followed by transfection of COS-1 cells andindirect immunofluorescence with anti-PK showed nuclear targeting.

QPQRYGGGRGRRW Gag protein of human foamy retrovirus; a mutant that completely lacks this box exhibits very littlenuclear localization; binds DNA and RNA in vitro.

proportional to the number of peptide motifs (see Radu etal, 1995 and the references cited therein).

The nucleoporin Nup98, containing 16 perfect andimperfect GLFG repeats and 3 FXFG repeats, is locatedasymmetrically at the nucleoplasmic site of the porecomplex in rat cells, and, along with Nup153, is aconstituent of the nuclear pore basket structure and/ornucleoplasmic ring (Radu et al, 1995). Karyopherins !/"bind cooperatively to FXFG but not GLFG repeat regions;binding of the NLS-protein/ karyopherin !/" heterodimerto FXFGs stimulated dissociation of the NLS-protein fromthe karyopherin !/" (Rexach and Blobel, 1995). Nup98functions as a docking protein via its N-terminal halfwhich contains all of the peptide repeats forming, withpeptide repeats of other nucleoporins, an array of sites to

mediate docking of nuclear proteins across the pore; thesemultiple docking sites were suggested to extend over adistance of 250 nm from the cytoplasmically exposedfibers to the nucleoplasmic baskets (Radu et al, 1995).

Nup133 and Nup145 in S. cerevisiae are involved inmaintaining the architecture of the nuclear envelope andthe position of the pore complex in the nuclear envelope;disruption of their genes leads to clustering of porecomplexes. Nup116 in yeast (Nup98 in rat, p97 inXenopus) interacts with Kap95 (karyopherin " inmammals); Nup116 has a number of GLFG repeats whichare required for pore function whereas the repeats inNup49, Nup57, Nup100, and Nup145 are not.Overexpression of Nup116 blocked export of mRNA(Iovine et al, 1995).


722

Table 2 "Bipartite" or "split" NLS


C-terminus Xenopus nucleoplasmin. Deletion analysis demonstrated the presence of a signal responsible fornuclear location.

AVKRPAMINO ACIDT KKAGQAKKK Xenopus nucleoplasmin

RPAMINO ACIDT KKAGQAKKKKLD Xenopus nucleoplasmin. Whereas these 17 amino acids had NLS activity, shorter versions of the 17amino acid sequences were unable to locate pyruvate kinase to the nucleus.

(AVK)RPAMINOACIDTKKAGQAKKK(KLD)

Xenopus nucleoplasmin. This 14 amino acid segment was identified as a minimal nuclear locationsequence but was unable to locate puruvate kinase to the nucleus; three more amino acids at either end(shown in parenthesis) were needed.

CGQAKKKKLD Xenopus nucleoplasmin-derived synthetic peptide; crosslinked to chicken serum albumin andmicroinjected to HeLa cells specified nuclear localization. This suggests that nucleoplasmin maypossess a simple NLS.

KRPAMINO ACIDT KKAGQAKKKK Xenopus nucleoplasmin bipartite NLS. Two clusters of basic amino acids (underlined) separated by 10amino acid are half NLS components.

HRKYEAPRHx6PRKR Yeast L3 ribosomal protein (387 amino acid) N-terminal 21 amino acid. Possible bipartite NLS.(Ribosomal proteins are transported to the nucleus to assemble with nascent rRNA). Fusion genes with"-galactosidase were used to transform yeast cells followed by fluorescence staining with b-galantibody. The 373 amino acid of L3 fused to "-gal failed to localize to the nucleus, unless a 8 aminoacid bridge containing a proline was inserted between L3 and "-gal.

NKKKRKLSRGSSQKTKGTSASAKARHKRRNRSSRS (one sequence)

SV40 Vp3 structural protein . (35 amino acid C-terminus). By DEAE-dextran-mediated transfectionof TC7 cells with mutated constructs.

RVTIRTVRVRRPPKGKHRK Simian sarcoma virus v-sis gene product (p28sis). The cellular counterpart c-sis gene encodes aprecursor of the PDGF B-chain (platelet-derived growth factor). The NLS is 100% conservedbetween v-sis gene product and PDGF. This protein is normally transported across the ER; introductionof a charged amino acid within the hydrophobic signal peptide results in a mutant protein that istranslocated into the nucleus. Puruvate kinase-NLS fusion product is transported less efficiently thancytoplasmic v-sis mutant proteins to the nucleus.

KRKIEEPEPEPKKAK Putative bipartite NLS of Xenopus laevis protein factor xnf7. Inferred by similarity to the bipartiteNLS of nucleoplasmin. During oocyte maturation xnf7 is cytoplasmic until mid-blastula—gastrula stagedue to high phosphorylation. Partial dephosphorylation results in nuclear accumulation.

KKYENVVIKR S PRKRGRPRKD Yeast SWI5 gene product, a transcription factor. Underlined basic amino acid show similarity tobipartite NLS of Xenopus nucleoplasmin. The SWI5 gene is transcribed during S, G2 and M phases,during which the SWI5 protein remains cytoplasmic due to phosphorylation by CDC28-dependenthistone H1 kinase at three serine residues two near and one (double underlined) in the NLS.Translocated at the end of anaphase/G1 due to dephosphorylation of NLS. NLS confers cell cycle-regulated nuclear import of SWI5—"-galactosidase fusion protein.

MKRKRNS735-741GIESIDNVMGMIGILPDMTPSTEMSMRGVRISKMGVDETSSAEKIV449-495

Bipartite NLS of influenza virus polymerase basic protein 2 (PB2). Mutational analysis of PB2 andtransfection of BHK cells showed that both regions are involved in nuclear import. Deletion of 449-495 region gives perinuclear localization to the cytoplasmic side.

AHRARRLH6-13 (BSI)PPRRRVRQQPP23-33 (BSII)PARARRRRAP39-48 (BSIII)

"Tripartite" or "doubly bipartite" NLS of adenovirus DNA polymerase (AdPol). BSI and II functionedinterdependently as an NLS for the nuclear targeting of AdPol, for which BSIII was dispensable. BSII-III was more efficient NLS than BSI-II.

KRKx11KKKSKK

207-226Human poly(ADP-ribose) polymerase (116 kD). The linear distance between the two basic clusters isnot crucial for NLS activity in this bipartite NLS. Lysine 222 (double underlined) is an essential NLScomponent. DNA binding and poly(ADP-ribosyl)ating active site are independent of NLS.

(GRKRAFHGDDPFGEGPPDKKGD) Herpes simplex virus ICP8 protein (infected-cell protein). This C-terminal portion of ICP8 introducedinto pyruvate kinase (PK) caused nuclear targeting in transfected Vero cells. Inclusion of additionalICP8 regions to PK led to inhibition of nuclear localization.

KR P REDDDGEPSERKR A RDDR Bipartite NLS of VirD2 endonuclease of rhizogenes strains of Agrobacterium tumefaciens. Within theC-terminal 34 amino acid. Each region (underlined) independently directs "-glucuronidase to thenucleus, but both motifs are necessary for maximum efficiency. VirD2 is tightly bound to the 5' end ofthe single stranded DNA transfer intermediate T-strand transferred from Agrobacterium to the plantcell genome.


723

Table 3 . "Nonpositive NLS" lacking clusters of arginines/lysines


QLVWMACNSAMINOACIDFEDL R VLS FIRGTKVSPRG327-356

Influenza virus nucleoprotein (NP). The underlined region (327-345) when fused to chimpanzee a1-globin at the cDNA level and microinjected into Xenopus oocytes specifies nuclear localization.

MNK IPI KDLLNPQ(NLS1 at N-terminus)VRILESWFAKNIENPYLDT (NLS2 atamino acid 141-159, part of thehomeodomain)

Yeast MAT a2 repressor protein, containing a homeodomain. The two NLS are distinct, eachcapable of targeting "-galactosidase to the nucleus. However, deletion of NLS2 results in a2accumulation at the pores. NLS1 and 2 may act at different steps in a localization pathway. Part of thehomeodomain mediates nuclear localization in addition to DNA binding. The core pentapeptidecontaining proline and two other hydrophobic amino acids flanked by lysines or arginines (underlined)was suggested as one type of NLS core.

Rx7Kx15KIPRx3HFYEERLSWYSDNED152-206 (C-terminal segment).

Drosophila HP1 (206 amino acids) that binds to heterochromatin and is involved in gene silencing.NLS identified by "-galactosidase/HP1 fusion proteins introduced by P-element mediatedtransformation into Drosophila embryos.

FVx7-20MxSLxYMx4MF Adenovirus type 5 E1A internal, developmentally-regulated NLS. This NLS functions in Xenopusoocytes but not in somatic cells. This NLS can be utilized up to the early neurula stage.

Table 4 . Nucleolar localization signals (NoLS)


M P K T RRR P RRS Q RKR PPT P Nucleolus localization signal in amino terminus of human p27x-III protein (also called Rex) of T cell

leukemia virus type I (HTLV-I). When this peptide is fused to N-terminus of "-galactosidase, directs itto the nucleolus. Deletion of residues 2-8 (underlined), 12-18 (double-underline) or substitution of thecentral RR (dotted-underlined) with TT abolish nucleolar localization. Other amino acids betweenpositions 20-80 increase nucleolar localization efficiency.

RLPVRRRRRRVP Adenovirus pTP1 and pTP2 (preterminal proteins, 80 kD) between amino acid residues 362-373. The140 kD DNA polymerase of adenovirus when it has lost its own NLS can enter the nucleus via itsinteraction with pTP. The staining was nuclear and nucleolar with some perinuclear staining as well.The NLS fused to the N-terminus of E. coli "-galactosidase was functional in nuclear targeting.

GRKKRRQRRRP HIV (human immunodeficiency virus) Tat protein; localizes pyruvate kinase to the nucleolus. Tat isconstitutively nucleolar.

RKKRRQRRR(AHQ)Nucleolar localization signal

Tat positive trans-activator protein of HIV-1 (human immunodeficiency virus type 1). The 3 aminoacids shown in parenthesis are essential for the localization of the "-galactosidase to the nucleolus.The 9 amino acid basic region is able to localize "-gal to the nucleus but not to the nucleolus.

PAMINO ACIDKRVKLDQRRRP Artificial sequence from c-Myc and HIV Tat NLSs that effectively localizes pyruvate kinase to thenucleolus.

FKRKHKKDISQNKRAVRR Human HSP70 (heat shock protein of 70 kD); localizes pyruvate kinase to the nucleus and nucleolus.HSP70 is physiologically cytoplasmic but with heat-shock HSP70 redistributes to the nucleoli,suggesting that the nucleolar targeting sequence is cryptic at physiological temperature and is revealedunder heat-shock.

RQARRNRRRRWRERQR (35-50) HIV-1 Rev protein (116 amino acid; nucleolar). Mutations in either of the two regions of arginineclusters severely impair nuclear localization. "-galactosidase fused to R4W was targeted to thenucleus, and fused to the entire 35-50 region, was targeted to the nucleolus.

RQA RR NRRRRW RERQRQ (35-51) HIV-1 Rev protein. A fusion of this Rev peptide with "-galactosidase became nuclear but notnucleolar. The 1-59 amino acid segment of Rev fused to "-galactosidase localized entirely within thenucleolus. Whereas the NRRRRW (bold) is responsible for nuclear targeting, the RR and WRERQRQ(double underlined) specify nucleolar localization. Rev may function to export HIV structural mRNAsfrom the nucleus to the cytoplasm.

V. Mechanism of nuclear import andtranslocation across the pore complex

A. Molecular mechanisms

Transport across the pore complex is a two-stepprocess involving binding at a site toward the periphery ofthe pore, docking of the molecules over the centraltransporter channels across the lumen of the pore complex,and release to the nucleoplasm (Akey and Goldfarb, 1989;


724

reviewed by Nigg et al, 1991). The guided diffusioninvolves docking to and undocking from multiple sitesacross the pore (Radu et al, 1995); the movement ofproteins across the pore is a stochastic process operatingby means of repeated association-dissociation reactions ofthe NLS-protein with FXFG repeats on nucleoporins(Rexach and Blobel, 1995). Only the translocation steprequires ATP (Richardson et al, 1988). The mechanism ofexport involves mRNA complexed with proteinscontaining NLSs forming a large globular mRNP (Mehlinet al, 1992); export might be specified by the presence of alarge polyanion (RNA) in the complex whereas importmight be specified by the NLS in the protein.

A single transport gate is located in the central domainof the transporter located within the pore complex thatrestricts passive diffusion; this was shown using smallgold particles coated with polyethylene glycol (PEG; totalparticle diameter 40-70 Å) or large PEG-particles (totaldiameter 110-270 Å) which were microinjected into thecytoplasm or nucleoplasm of Xenopus oocytes;cytoplasmic injections of small gold particles showed thatthe particles were approximately 11 times moreconcentrated in the cytoplasmic half of the transporterstructure whereas the particles were approximately 7 timesmore concentrated in the nuclear half after nuclearinjections. Larger particles were less mobile and aftercytoplasmic injection migrated to the surface of the porecomplex, but entered the transporter less frequently(Feldherr and Akin, 1997).

Macromolecules are poor electrical charge carriers andthis can be exploited to detect their movement alongelectrolyte-filled pores: translocating macromoleculesreduce the net conductivity of the medium inside the pore;lesser values of ion conductance indicate greatermacromolecular translocation (in size and/or number). Thisis the principle used in Coulter counter, an instrument forcounting and sizing particles. The principle that ion flowis restricted during translocation of macromoleculescontaining nuclear targeting signals was demonstrated byBustamante et al (1995). At least four soluble transportproteins have been identified recently: the ! subunit ofkaryopherin involved in NLS binding (a number of othercandidate NLS-binding proteins are known or might bediscovered in the future), the " subunit of karyopherin, theRan, and p10 (Rexach and Blobel, 1995; Radu et al, 1995;Nehrbass and Blobel, 1996).

B. Components of the soluble importmachinery

1. The subunit of karyopherin

The ! subunit of karyopherin, equivalent to the 60kDa importin (Görlich et al, 1994, 1995a,b) and to SRP1and SRP1a recognizes and binds the NLS peptide of the

protein to be imported in the cytoplasm. Karyopherin !accompanies the proteins to be imported from their site ofsynthesis through the pores to the sites of their function inthe nucleus (Görlich et al, 1995b). Two other cytosolicproteins with molecular weights of 56 and 66 kDa havebeen identified, along with the 66 and 90 kDakaryopherins, to form with NLS-protein a five proteincomplex (Imamoto et al, 1995). Karyopherins ! and "cooperate to bind to FXFG but not GLFG repeats onnucleoporins (Rexach and Blobel, 1995).

Görlich and coworkers (Görlich et al, 1994, 1995b)have identified the 60 and 90 kDa importin subunits inboth Xenopus and human cells corresponding tokaryopherins ! and " (Moroianu et al, 1996); togetherthey constitute a cytosolic receptor for NLS binding; bothsubunits appear bound to the pore complex but only thelarger subunit enters the nuclear interior. The 60 kDasubunit shows homology to S. cerevisiae SRP1 , a porecomplex protein that contributes to the maintenance of thenucleolar structure (Yano et al, 1992, 1994). Importin-!mediates nuclear protein import by binding nuclearlocalization signals and importin-". A role for the !subunit of importin in RNP export has been considered(Laskey et al, 1996).

The second human homolog of yeast SRP1, hSRP1,was identified using the yeast two-hybrid system (Zervoset al, 1993) because of its interaction with a RAG-1activator of V(D)J recombination in immunoglobulingenes (Cortes et al, 1994); the domain of RAG-1interacting with hSRP1 was not required for recombination(Cortes et al, 1994). hSRP1 contains eight degeneraterepeats of 40-45 amino acids four of which (repeats 4-7between amino acids 245 and 437 not including the acidicstretches of the molecule) are involved in interaction withRAG-1 (Cortes et al, 1994); these repeats are known asarm motifs, have been found in other proteins, and areinvolved in specific protein-protein interactions (Peifer etal, 1994). For example, arm motifs participate in theinteraction between the tumor suppressor adenomatouspolyposis coli and "-catenin (Rubinfeld et al, 1993; Su etal, 1993). The human SRP1, interacting with RAG-1 andperhaps also with RNA polymerases, was proposed tolocalize recombination and transcription near the porecomplex providing the anchoring activity and assemblingcomponents essential for these processes (Cortes et al,1994).

SRP1 interacts with the Nup1p and Nup2p nuclearpore proteins, is also implicated in the correct orientationof mitotic tubulin spindles, and has been proposed toanchor structural cytoplasmic components to pores therebyorganizing proper nuclear matrix structures (Yano et al,1994). SRP1 and importin 60 are also homologous toRch1, a protein that interacts with the immunoglobulin


725

gene recombinase RAG-1 perhaps via the NLS of RAG-1(Cuomo et al, 1994).

2. The subunit of karyopherin

The " subunit of karyopherin, equivalent to the 90 kDaimportin (Görlich et al, 1994, 1995a,b) and to Kap95 inyeast binds to the karyopherin !/NLS-protein complex inthe cytoplasm and mediates docking of the complex tonucleoporins with repetitive tetrapeptide motifs (Iovine etal, 1995; Radu et al, 1995; Weis et al, 1995; Moroianu etal, 1996). Karyopherin " enhances binding of karyopherin! to NLS-protein (Rexach and Blobel, 1995). Only the "subunit is able to bind pores and binding of the a subunitto the pore depends on karyopherin "; karyopherin "moves only to a distance of 100 nm from its initialcytoplasmic docking site but remains associated with poresand does not appear in the nucleoplasm (Görlich et al,1994, 1995a,b). The FXFG repeats on nucleoporins appearto stimulate the dissociation of the NLS-protein from thekaryopherin !/" heterodimer (Rexach and Blobel, 1995).

3. RanGTP

Ran (in complex with p10) release the docked complexby displacing karyopherin !/NLS-protein; RanGTP andkaryopherin ! bind to overlapping sites on karyopherin ";a cluster of basic residues on karyopherin " are the bindingsites for RanGTP and karyopherin ! (Moroianu et al,1996). The small GTPase Ran executes the energy-dependent step of translocation across the pore complex,results in accumulation of import substrate andkaryopherin ! in the nucleus, and in the retention ofkaryopherin " in the pore complex on both sides of thenuclear pore; in the absence of Ran or energy, karyopherin! accumulates in the pore but not in the nucleoplasm inpermeabilized HeLa cells (Görlich et al, 1995a,b). Rancauses the dissociation of the NLS-protein/ karyopherin !from the karyopherin " (Rexach and Blobel, 1995; Paschaland Gerace, 1995); incubation of RanGTP withkaryopherin !/" heterodimer led to the dissociation of the! subunit and to the association of the " subunit withRan; RanGDP had no effect (Rexach and Blobel, 1995).Ran/TC4 is absolutely required for the efficient transport(Moore and Blobel, 1993; Görlich et al, 1995a,b). Ranrequires the p10 protein as an active component for itsefficient functioning (Nehrbass and Blobel, 1996).

The binding determinants of karyopherin " for Ran-GTP are similar to Ran BP1, a cytoplasmic Ran-GTP-binding protein, (Coutavas et al, 1993) and to similardomains on nucleoporin Nup 358 (Yokoyama et al, 1995).Displacement of Ran-GTP from karyopherin " may be arequisite for GTP hydrolysis by Ran-GAP (Floer andBlobel, 1996) and may serve to recycle karyopherin ".

4. The p10 protein

The p10 protein can associate with Ran-GDP (but notto Ran-GTP) and to karyopherin ". p10 binds tonucleoporins possessing peptide repeats (Nehrbass andBlobel, 1996). Addition of GTP to the p10/nucleoporin/Ran-GDP/karyopherin !/" complex resulted in formationof Ran-GTP causing dissociation of karyopherin ! leavingthe karyopherin " bound to nucleoporin (Nehrbass andBlobel, 1996). Release of karyopherin !/NLS-protein thenallows the protein to be imported and karyopherin ! todiffuse into the nucleus across the central plug (Görlich etal, 1995a,b).

The entrance of karyopherin ! in the nucleus isconsistent with the model of a shuttling nuclear importreceptor (Adam et al, 1989). Dissociation of the NLS-protein from karyopherin ! in the nucleoplasm might bemediated by a difference in the ionic environment betweenthe nucleoplasm and the cytoplasm (Boulikas, 1994), byassociation of karyopherin ! with other nuclear factors(Görlich et al, 1995b), by association of karyopherin !with shuttling proteins during their exit from the nucleus(Schmidt-Zachmann et al, 1993), or by phosphorylation ofkaryopherin ! by a mammalian homolog of the yeastSRP1 kinase (see Radu et al, 1995; Moroianu et al, 1996).

The entrance of karyopherin ! in the nucleus isconsistent with the model of a shuttling nuclear importreceptor (Adam et al, 1989). Dissociation of the NLS-protein from karyopherin ! in the nucleoplasm might bemediated by a difference in the ionic environment betweenthe nucleoplasm and the cytoplasm (Boulikas, 1993,1994), by association of karyopherin ! with other nuclearfactors (Görlich et al, 1995a), by association ofkaryopherin ! with shuttling proteins during their exitfrom the nucleus (Schmidt-Zachmann et al, 1993), or byphosphorylation of karyopherin ! by a mammalianhomolog of the yeast SRP1 kinase (see Radu et al, 1995;Moroianu et al, 1996).

C. A summary on the translocationprocess

In summary, protein translocation from the cytoplasmto the nucleoplasm involves the following steps (Nehrbassand Blobel, 1996; Moroianu et al, 1996) (Figure 5):

(i ). A weak complex of karyopherin !/NLS-protein isformed in the cytoplasm.

(i i ). Karyopherin " interacts with karyopherin !forming a strong karyopherin "/!/NLS-protein complex.Additional proteins may participate to the formation of alarger cytoplasmic complex (Imamoto et al, 1995).

(i i i ). The complex binds to FXFG peptide repeats onnucleoporins at the cytoplasmic side of the pore complex


726

via karyopherin ". The FXFG repeats may dissociate theNLS-protein from the karyopherin !/".

(iv ). p10 docks Ran-GDP to nucleoporin and to thekaryopherin heterodimer.

(v ). A number of association-dissociation reactions onnucleoporins dock the import substrate toward thenucleoplasmic side; this process requires the GTPase Ranand p10.

(vi ). A GDP-GTP exchange reaction takes placetransforming Ran-GDP into Ran-GTP catalyzed bykaryopherin ! which shares sequence homology with aGDP-GTP exchange factor of Ras (Görlich et al, 1994;Peifer et al, 1994). A cytosolic Ran-GTPase activatingprotein (Ran-GAP) in yeast has been found keeping Ranprimarily in the GDP-bound form. Ran-GTP is asecondary product found locally at the pore.

(vi i ). Ran-GTP (but not Ran-GDP) causesdissociation of the heterodimeric !/" complex by bindingto karyopherin " thus releasing the karyopherin !/NLS-protein.

(v i i i ). A complex of karyopherin !/NLS-proteindiffuses into the nucleoplasm whereas karyopherin "remains bound to the pore (because of its affinity to FXFGrepeats and p10).

Shuttling proteins might contribute to coordinatingnucleocytoplasmic import/export; these proteins includenucleolin and NO38 (Borer et al 1989), two hsp70-relatedproteins in Xenopus oocytes (Mandell and Feldherr 1990),the A1 pre-mRNA binding protein (Piñol-Roma andDreyfuss 1992), Nopp 140 (Meier and Blobel, 1992),progesterone receptor (Guiochon-Mantel et al, 1991), Laantigen, and several protein kinases (reviewed in Boulikas1993, 1996). A cap-binding protein has been identified thatmight mediate export of RNA polymerase II transcripts(Izaurralde et al, 1992).

In conclusion, several independent import/exportpathways seem to operate in the same cell. Nonboundnucleolin is exported from nuclei and the rate of export isdetermined by structural domains involved in interactionsin the nucleolus; a fusion construct between thecytoplasmic pyruvate kinase and the NLS of T antigen isable to shuttle between nucleus and cytoplasm; lamin B2,

a normally nuclear protein, can be converted into ashuttling protein by introducing mutations on its nuclearsignal (Schmidt-Zachmann et al, 1993).

D. Import of influenza virusribonucleoproteins

Influenza virus is unusual among RNA viruses in thatit replicates in the nucleus. When infecting cells it firstbinds to receptors containing sialic acid and is then

internalized by receptor-mediated endocytosis into lateendosomes. A conformational change of peptides onhemagglutinin (HA) spikes at the lower pH of theendosome (pH 5.5) causes disruption of the endosomalmembrane and release of the virus into the cytoplasm(reviewed by Bui et al, 1996). The fusogenic peptides ofHA protein of influenza virus have been exploited in genetherapy for the efficient release of DNA-cationic polymercomplexes from endosomes (see Fusogenic peptides inBoulikas, 1998, pages 1-172, this volume).

Since the genome is segmented, eight separate helicalviral RNPs are formed containing the antisense viral RNAand numerous copies of the 56 kDa (one every 20nucleotides); other viral proteins in the complex includethe three subunits of the polymerase and the 27 kDa M1viral matrix protein which is released from the vRNPs,presumably in the acid pH of the endosome (see Bui et al,1996 and the references cited therein). This dissociationstep is essential for nuclear import of the vRNPs; twoanti-influenza virus drugs, amantadine and rimantadine,inhibit the dissociation of M1 protein from vRNPs. M1assumes a master regulatory role for the transport ofvRNPs across the cell membrane; however, the associationof vRNPs with M1 inhibits their nuclear transportationacross the pore complex; acidification of the cytosoliccompartment caused dissociation of M1 from vRNPs andeliminated the inhibition in import.

M1 appears to prevent reimport of vRNPs into thenucleus of the infected cell and thus commits them to anassembly pathway leading to the budding of the virusparticles at the cell membrane (Bui et al, 1996). When M1was dissociated from vRNPs at late times during infectionthe vRNPs failed to be reimported into nuclei; cell fusiontechniques, however, have shown that vRNPs which weredissociated from M1 in acid pH were import competent inthe uninfected nucleus; for some unknown reason, infectednuclei, although capable of general nuclear import were nolonger able to import vRNPs (Bui et al, 1996). TheHepatitis B virus that, like influenza virus, replicates inthe nucleus can be reimported into the nucleus in infectedcells, a fact that may explain the chronic nature of HepatitsB infection.

VI. Regulated protein import

A number of processes have been found to be regulatedby nuclear import. These include: the NF- B

translocation; the import of Dorsal protein in dorsal butnot ventral compartments of Drosophila embryos, aprocess playing a decisive role in cell type establishmentduring morphogenesis; the nuclear import of the factorsrNFIL-6 and ISGF3 after their phosphorylation in thecytoplasm; the Xenopus nuclear factor 7 which isretained in the cytoplasm from fertilization through the


727

mid-blastula stage (Li et al, 1994); and the import of anumber of protein kinases including casein kinase II ,the catalytic subunit of cAMP-dependent protein

kinase II following stimulation of cells with cAMP, theRSK after its phosphorylation by MAP kinase, protein

kinase C after stimulation of cells with the phorbol esterTPA, and the Mitogen-Activated Protein kinases (MAPK)ERK1 and ERK2 following stimulation of cells withgrowth factors or mitogens, (reviewed by Boulikas, 1995).

MAPK is activated in cytoplasm by MAPK kinase(MAPKK) in response to extracellular signals; whereasMAPK then is translocated to nucleus, MAPKK remainscytoplasmic because of a NES in the N-terminal region(residues 32-44) rich in leucine residues; the NES peptideof MAPKK, inhibited the nuclear export of ovalbumin(Fukuda et al, 1996).

The NLS of the serum response factor (SRF ) is inclose proximity to potential phosphorylation sites for thecAMP-dependent protein kinase (A-kinase) and nucleartransport of SRF proteins requires basal A-kinase activity(Gauthier-Rouviere et al, 1995).

The NLS needs to be exposed on the surface of theprotein and available for binding to nuclear transporterprotein molecules. The SV40 large T protein NLS isnonfunctional when it is located in a region of pyruvatekinase predicted not to be exposed to the surface (Robertset al., 1987). In addition, nuclear proteins have beenidentified which are synthesized in the form of precursormolecules which remain in the cytoplasm, presumablybecause their NLS is hidden. Cleavage of such proteinsinto mature molecules by specific proteases exposes theirNLS, and they are then rapidly transported to the nucleus.As an example, the p50 subunit of the transcription factorNF- B (50 kD) is synthesized in the form of a 110 kDprecursor with the NLS buried in the protein; proteolyticcleavage giving the 50 kD transcription factor exposes theNLS (Henkel et al., 1992) and facilitates nuclearlocalization.

Nuclear translocation of protein factors, presumably byexposure of their hidden NLS or by reconstitution of afunctional NLS from two remote half NLS, can betriggered by phosphorylation (Shelton and Wasserman,1993), dephosphorylation (Moll et al., 1991; Nasmyth etal., 1990), subunit association (Levy et al., 1989), or bydissociation of an inhibitory subunit (Baeuerle andBaltimore, 1988a,b). Interferon-! regulates nucleartranslocation of the transcription factor ISGF3 (Kessler etal., 1990). Thus, distant peptide regions can either "mask"the NLS or anchor the protein in the cytoplasm; examplesof proteins regulated in this manner include protein

kinase Ca (James and Olson, 1992), human cycl ins A

and B1 (Pines and Hunter, 1991), and c-Fos (Roux etal., 1990). In other cases, binding of hormone to a

cytoplasmic protein such as glucocorticoid receptor

(GR) induces a conformational change that exposes theNLS, or reconstitutes a functional NLS from remote halfNLSs, and the protein is rapidly transported to the nucleus(Picard and Yamamoto, 1987). The ability of GRs to bindDNA is an important determinant for localization and tightbinding of GR to the nucleus; mutant GRs localized to thenucleus were only weakly associated with the nuclearcompartment (Sackey et al, 1996). However, the relatedestrogen receptor and retinoic acid receptor, contrary to theglucocorticoid receptor, are nuclear even in the absence oftheir receptor hormone/morphogen (Picard et al., 1990).An unusual case has been described by Lutz andcollaborators (1992) where prenylation of the C-terminusof prelamin A by addition of a 15-carbon or 20-carbonisoprenoid, a posttranslational modification that functionsin the proteolytic processing of prelamin A to lamin A inthe nucleus, is also required for its nuclear import.

VII. Deregulation in nuclear importand molecular carcinogenesis

Several studies have provided a link between thederegulation in nuclear import mechanisms of specificproteins and cancer. The BRCA1 breast cancer markerprotein was found to be mainly localized in the cytoplasmin 16 of 17 breast and ovarian cancer lines and in 17 of 17samples of cells from malignant effusions whereas innormal cells the protein was nuclear (Chen et al, 1995).

The transforming oncoprotein v-Abl of Abelsonmurine leukemia virus, a mutated form of the c-Ablnonreceptor tyrosine kinase, is a fusion protein in whichportions of the retroviral Gag protein replace the N-terminal SH3 domain of c-Abl; this results in loss ofphosphorylation sites in v-Abl that down-regulate itsactivity and render the tyrosine kinase activity of v-Ablconstitutive; in addition, the viral Gag sequence provides amyristoylation site on v-Abl which confers apredominantly inner plasma membrane anchorage whereasc-Abl is predominantly located in the nucleus (Wong et al,1995). Both of these properties of v-Abl, not found in c-Abl, contribute to its ability to transform cells.

VIII. RNA and protein export

A. An historical perspective

The concepts that (i ) transcription occurs preferentiallyat the nuclear periphery near pores versus (i i ) transcriptionoccurring at any location within nuclei both have beenentertained and data to support either model are available.Transcripts from the interior of nuclei have been visualizedpassing through channels originating at the sites oftranscription within the interior of the nucleus andemanating to the pore complex (Huang and Spector,


728

1991). RNA transcripts seem to be exported vectoriallytoward a single pore or a small subset of nuclear pores(Lawrence et al., 1989). This vectorial export of RNAmight contribute to an asymmetric mRNA distributionwithin the cytoplasm consistent with experimentalevidence (Weeks and Melton, 1987). Foci where splicingof pre-mRNA takes place are often seen associated withpore complexes via morphologically distinct channels; U3snRNA was exclusively detected in the nucleolus and U2,U5 and U6 snRNAs were in discrete nucleoplasmic foci(Carmo-Fonseca et al, 1991).

RNA export is facilitated by proteins that shuttlebetween nucleus and cytoplasm. Assays based oninterspecies heterokaryons and microinjection of Xenopus

oocytes using nucleolin, mutant lamins, differing in theirabilities to be incorporated into the lamina and pyruvatekinase-NLS artificial reporter protein, have shown thatproteins unable to interact with large nuclear structures canbe exported from nuclei; protein export was suggested notto require any export signals (Schmidt-Zachmann et al,1993). However, recently a nuclear export sequence (NES)has been identified in proteins that are exported activelyfrom the nucleus such as Rev and PKI (Fischer et al,1995; Wen et al, 1995; reviewed by Gerace, 1995;Izaurralde and Mattaj, 1995). Several RNA-bindingproteins have been suggested to be involved in RNAexport. Export of 5SRNA requires interactions withribosomal protein L5 or TFIIIA (Guddat et al, 1990).Export of influenza virus RNA-protein complex requiresthe viral protein M1 which also prevents the nuclear re-import of the viral RNA (Martin and Helenius, 1991).

Export of tRNA is a translocation process mediated byprotein carrier(s) (Zasloff, 1983). tRNA molecules undergoa complex maturation processing involving trimming ofthe 5' and 3' ends, addition of three terminal CCA residues,and base modification; removal of introns (only 20% oftRNAs contain introns) is a highly regulated processoccurring in association with the inner side of the nuclearenvelope close to the pores (for references see Simos et al,1996). Studies in yeast have shown that Los1 (required forpre-tRNA splicing) and Pus1 (involved in tRNAbiogenesis) interact with the pore complex protein Nsp1;this involves pore proteins in the splicing of pre-tRNA(Simos et al, 1996).

Ribosomal subunits are assembled in the nucleus fromimported ribosomal proteins; export of ribosomal subunitsto the site of their function (cytoplasm) is energy-dependent (Khanna-Gupta and Ware 1989; Bataillé et al1990). The nucleoside triphosphatase of nuclear envelopemight be involved in the nucleocytoplasmic translocationof ribonucleoprotein (Agutter et al 1976).

The mechanism of export of mRNA deduced bymicroinjection of Xenopus oocytes is also an energy-

dependent process requiring the 5' cap structure (Hamm andMattaj, 1990; Dargemont and Kuhn, 1992). RNP particlesbecome attached to filaments which project into thenucleoplasm and which guide the particles to the pores.The central channel can expand to permit export of largecomplexes such as ribosomal subunits and mRNPs andimport of large nuclear proteins.

RNA export comprises (i) initial binding to thetentacles of the central plug (ii) energy-dependenttranslocation toward the cytoplasmic side of the porecomplex (Mehlin et al, 1992). Some asymmetries betweenthe cytoplasmic and nucleoplasmic rings have beendescribed; for example, the cytoplasmic ring appears largerand the nucleoplasmic filaments of the pore are longer,forming a basket structure. One of the nucleoporins isattached to the nucleoplasmic side (Snow et al, 1987).These asymmetries are likely to contribute to the RNPexport distinct from protein import into nuclei (Mehlin etal, 1992).

Electron microscope tomography has examined theexport of large RNPs in the salivary glands of the dipteranChironomus. A 75S pre-mRNA is transcribed from theBalbiani ring granules in Chironomus tentans and ispacked into RNP ribbon particles, 30-60 nm broad and 10-15 nm thick, bent into a ring conformation; during exportthe particle is first oriented in a specific manner by specificrecognition signals and subsequently the bent ribbon isgradually straightened and exported through the pore withthe 5' end of RNA in the lead (Mehlin et al, 1992). Uponpassage through pores, the proteins of the pre-mRNPdissociate; the protein composition of the particle in thenucleus and cytoplasm is different. The particle thenunfolds and becomes associated with ribosomes (Mehlin etal, 1992).

Some early studies showed that no specific nuclearexport signals are required. Schmidt-Zachmann et al (1993)arrived to the conclusion that a protein does not requirepositively acting export signals to be transported fromnucleus to cytoplasm but instead, its shuttling ability islimited primarily by intranuclear interactions.

Expression of some proteins can inhibit mRNAexport; the mechanism could be exerted at the splicing steprather than on the actual translocation process. Expressionof NS1 protein (which is encoded by the influenza virusRNA segment 8 along with NS2 produced from the sametranscript as NS1 by differential splicing) in influenzavirus-infected cells induced a generalized block of mRNAexport from the nucleus; NS1 mRNA, NS2 mRNA andother mRNAs were retained in the nucleus of cellsexpressing NS1 protein, but no effect was observed whenonly NS2 protein was expressed (Fortes et al, 1994).


729

B. Identification of NLS-independentimport pathways

Proteins lacking basic type of NLS are thought topiggy-back into the nucleus in association with NLS-containing molecules. However, a new type of receptormolecule mediates nuclear import in a basic NLS-independent manner (Pollard et al, 1996; Aitchison et al,1996; reviewed by Dingwall, 1996). This model applies tothe import of hnRNPA1 molecules (Pollard et al, 1996)and to the import nuclear mRNA-binding proteins inSaccharomyces cerevisiae which have been isolated as acomplex with Kap104 (Aitchison et al, 1996).

The cytosolic yeast karyopherin, Kap104p, acts forreturning mRNA binding proteins to the nucleus aftermRNA export. Indeed, Kap104p binds directly to repeat-containing nucleoporins and to the mRNA bindingproteins, Nab2p and Nab4p, and functions for their nuclearimport; depletion of Kap104p resulted in a rapid shift ofNab2p from the nucleus to the cytoplasm withoutaffecting the localization of other nuclear proteins(Aitchison et al, 1996).

C. Nuclear export signals (NES) andexport of mRNA: recent studies

According to Guiochon-Mantel et al (1994) the NLS ofthe progesterone receptor or the NLS of T antigen wereshown to impart to "-galactosidase the ability to shuttlebetween the nucleus and the cytoplasm; microinjectedproteins devoid of a nuclear localization signal were unableto exit from the nucleus. The authors thought that thenuclear import requires energy whereas the nuclear exportdoes not and this determines whether the NLS willfunction as an import or export signal.

The discovery of nuclear export signals (NESs) in anumber of proteins revealed the occurrence of signal-dependent transport of proteins from the nucleus to thecytoplasm. The consensus motif of the NESs is a leucine-rich, short amino-acid sequence. The NES is defined by itsability to translocate a protein from the nucleus to thecytoplasm when the two are tethered by a membrane-permeable ligand (Klemm et al, 1997). The majority ofproteins with NES are RNA-binding proteins which bindto and escort RNAs to the cytoplasm; nuclear export ofRNA molecules is likely to be driven by protein-basednuclear export signals (reviewed by Nakielny and Dreyfuss,1997; Lee and Silver, 1997). Nascent pre-mRNAsassociate with the abundant hnRNPs and remain associatedwith them throughout the time they are in the nucleus.One group of HnRNPs is strictly nuclear in interphasecells (for example hnRNP C proteins), whereas the othergroup, although primarily nuclear at steady state, shuttlesbetween the nucleus and the cytoplasm via NES; NES-

bearing hnRNP proteins are mediators of mRNA export(Nakielny and Dreyfuss, 1996).

Microinjection into the nucleoplasm of Xenopusoocytes of PEG-coated gold particles showed that thesewere coated with protein containing nuclear export signals(NES) suggesting that the NES is not only required fortranslocation, but also for migration within thenucleoplasm (Feldherr and Akin, 1997).

Nuclear trafficking of the catalytic (C) subunit ofcAMP-dependent protein kinase (cAPK) is regulated by theheat-stable inhibitor (Pkl) of cAPK which contains anuclear export signal (NES) (residues 35-49). Pkl has noobvious association with RNA. The core NES of Pklcomprises only residues 37-46, LALKLAGLDI and isable to trigger rapid, active net extrusion of the C-PKlcomplex from the nucleus.

The identification of NES have established a novelmechanism for regulation of gene expression: nuclearexport of pre-mRNA can contribute to the regulation ofgene expression. The processing of transcripts of TNF-"and "-globin was found to be regulated by the signaltransduction pathway that includes the Src protein; Srcseems to act on a general mechanism of splicing and/ormRNA transport. This regulation could involve RNA-binding proteins, which interact with Src (Neel et al,1995).

D. CRM1 or exportin 1 binds NES

A protein of 110 kDa (CRM1 or exportin 1 or XPO1)was identified in Xenopus oocyte extracts that binds to theintact NES but not to the mutated, non-functional NES.CRM1 is an essential mediator of the NES-dependentnuclear export of proteins in eukaryotic cells (Fukuda et al,1997). CRM1 is an evolutionarily conserved protein,shown to , originally found as an essential nuclear proteinin fission yeast; S. cerevisiae CRM1 shows homology toimportin "-like transport factors and was shown to be anessential mediator of nuclear protein export in S. cerevisiae

(Stade et al, 1997). The cytotoxin leptomycin B whichinhibits the NES-mediated transport of Rev proteininhibited the binding of Xenopus CRM1 to NES (Fukudaet al, 1997).

Overexpression of CRM1 in Xenopus oocytesstimulated Rev and U snRNA export from the nucleus andthis process was inhibited by leptomycin B, a cytotoxinthat was shown to bind to CRM1 protein; CRM1 wasable to form a complex involving cooperative binding ofboth RanGTP and the nuclear export signal (NES) fromeither the Rev or PKI proteins implicating RanGTP innuclear export (Stade et al, 1997; Fornerod et al, 1997). Amutation in the shuttling protein Crm1p affects not onlyprotein export, but also mRNA export, indicating that


730

these pathways are tightly coupled in S. cerevisiae (Stadeet al, 1997).

Retroviruses export unspliced, intron-containing RNAto the cytoplasm of infected cells despite the fact thatintron-containing cellular RNAs cannot be exported; inHIV-1 this is accomplished by Rev which binds toelements in the viral RNA; in the absence of Rev, theseintron-containing HIV-1 RNAs are retained in the nucleus(Zhang et al, 1996). The NES on Rev is the sequenceLQLPPLERLTL (Wen et al, 1995). Visualization ofviral transcripts using oligonucleotide probes specific forthe unspliced or spliced forms of a particular HIV-1 viralRNA showed that in the absence of Rev, the unsplicedHIV-1 viral RNAs were predominantly nuclear and weredistributed (i) as approximately 10-20 intranuclear punctatesignals of nascent transcripts and (ii) as a stable populationof viral transcripts dispersed throughout the nucleoplasmexcluding nucleoli (Zhang et al, 1996).

Kim and coworkers (1996) have pinpointed the NESon HTLV-1 Rex that fully complements HIV-1 Rev as astretch of 17 amino acids; four leucines within theminimal region were essential for NES function; this NESpeptide could serve as nuclear export signal whenconjugated with bovine serum albumin.

The genome of the simpler retrovirus Mason-Pfizermonkey virus (MPMV) contains an element that serves asan autonomous nuclear export signal for intron-containingviral and cellular RNA through interaction withendogenous cellular factors; the same element is alsoessential for MPMV replication (Ernst et al, 1997).

E. Transportin is distinct fromkaryopherin (importin)

A novel 38 amino acid transport signal was identifiedby Pollard and coworkers (1996) in the hnRNP A1 protein(which shuttles rapidly between the nucleus and thecytoplasm), termed M9, which confers bidirectionaltransport across the nuclear envelope. Furthermore, aspecific M9-interacting protein, termed transportin, bindsto wild-type M9. Transportin is a 90 kDa protein,distantly related to karyopherin " which also participates inmRNA export in a complex with hnRNP A1 and mRNA.Thus, it appears that there are at least two receptor-mediated nuclear protein import pathways.

Transportin mediates the nuclear import of additionalhnRNP proteins, including hnRNP F. A novel transportinhomolog, transportin 2, which may differ from transportin1 in its substrate specificity has also been identified andsequenced (Siomi et al, 1997). Because transportin 1 islocalized both in the cytoplasm and the nucleoplasm and apyruvate kinase-M9 fusion, which normally localizes inthe nucleus, accumulates in the cytoplasm when RNA

polymerase II is inhibited, it seems that the M9 signal is aspecific sensor for transcription-dependent nucleartransport. Consistent with in vitro data A1 dissociatesfrom transportin 1 by RanGTP after nuclear import andbecomes incorporated into hnRNP complexes, where A1functions in pre-mRNA processing (Siomi et al, 1997).

A novel human protein, termed MIP (101 kDa) whichbears significant homology to human karyopherin/importin-", binds M9 specifically; cytoplasmicmicroinjection of a truncated form of MIP that retains theM9 binding site blocked the in vivo nuclear import of asubstrate containing the M9 without affecting the importof basic NLS-bearing substrates (Fridell et al, 1997).

The shuttling hnRNP K protein contains also anovel shuttling domain (termed KNS) which has many ofthe characteristics of M9, in that it confers bi-directionaltransport across the nuclear envelope. KNS-mediatednuclear import is dependent on RNA polymerase IItranscription, and a classical NLS can override this effect.Furthermore, KNS accesses a separate import pathwayfrom either classical NLSs or M9 demonstrating theexistence of a third protein import pathway into thenucleus (Michael et al, 1997).

IX. Regulated nuclear import andexport

A. Proteins with NLS and NES

The transcription factor NF-ATc plays a key role inthe activation of many early immune response genes and isregulated by subcellular localization. NF-ATc translocatesfrom the cytoplasm to the nucleus in response to a rise inintracellular calcium. Calcineurin dephosphorylatesconserved serine residues in the amino terminus of NF-AT,resulting in nuclear import (Beals et al, 1997). NF-ATcimmediately returns to the cytoplasm when intracellularcalcium levels fall a process mediated by a NES; glycogensynthase kinase-3 (GSK-3) phosphorylates conservedserines necessary for nuclear export and opposing Ca2+

/calcineurin signaling (Klemm et al, 1997; Beals et al,1997).

The distribution of the v-Rel oncoprotein between thenucleus and the cytoplasm was experimentally manipulatedusing NLS and NES; the respective abilities of the v-Relto localize to the nucleus in chicken embryo fibroblasts, toactivate %B-dependent transcription in yeast, and totransform avian lymphoid cells were each markedly reducedby the fusion of a cis-acting NES onto v-Rel; theoncogenic properties of v-Rel were manifested only after athreshold of this protein in the nucleus was attained(Sachdev et al, 1997).

Fluorescein iodoacetamide-labeled human p53 , injectedinto the cytoplasm or nuclei of 3T3 cells, was imported


731

into or exported from nuclei within minutes. Import wasinhibited by co-injection of the lectin wheat germagglutinine (WGA). In contrast, the protein HSAconjugated with the T antigen NLS was only imported butnot exported. These studies demonstrate the presence ofNES on p53 in addition to import signals (NLS) andprovide new views for its implication in carcinogenesis(Middeler et al, 1997).

The subcellular localization and activity of c-Abl

nonreceptor tyrosine kinase is regulated by cell adhesion.Upon adhesion of fibroblasts to fibronectin (anextracellular matrix protein) there is a coincident export ofc-Abl from the nucleus to the cytoplasm. The cytoplasmicpool of c-Abl is reactivated within 5 min of adhesion andthe activated cytoplasmic c-Abl becomes nuclear after 30min. Thus, c-Abl can transmit integrin signals to thenucleus where it may integrate these to cell cycle signals(Lewis et al, 1996).

Fragile X syndrome, a leading cause of inheritedmental retardation, is attributable to the unstableexpansion of a CGG-repeat within the FMR1 (Fragile Xsyndrome Mental Retardation) gene; the encoded protein(FMRP) is a ribosome-associated RNA-binding proteinthat contains both NLS and NES. Immuno-gold studiesprovided evidence of nucleocytoplasmic shuttling ofFMRP, which was localized in neuronal nucleoplasm andwithin nuclear pores. FMRP was highly expressed inneurons but not glia throughout the rat brain; the dendriticlocalization of FMRP implicated this ribosomal protein inthe translation of proteins involved in dendritic structure orfunction that could relate to the mental retardationoccurring in fragile X syndrome (Feng et al, 1997).

B. Import/export of HnRNPs

Heterogeneous nuclear ribonucleoproteins (HnRNPs) isa group of 20 different hnRNP proteins designated RNPAto RNPU. Among these the C1, C2, and U moleculespossess the basic NLS. However, the group A moleculesdo not. In spite of this, a major group of hnRNP proteinsconstantly shuttle between the nucleus and the cytoplasm(Michael et al, 1995; Pollard et al, 1996). HnRNPs can bedivided into those that remain always nuclear and thosethat shuttle between the cytoplasm and the nucleus; theassociation of mRNA with those that posses NES andshuttle is believed to be largely responsible for mRNAexport from the nucleus. hnRNP C proteins are restrictedto the nucleus not because they lack an NES, but becausethey bear a nuclear retention sequence (NRS ) that iscapable of overriding NESs (Nakielny and Dreyfuss,1996). The NRS in hnRNP C1 is a stretch of 78 aminoacids; it was proposed that removal of NRS-containinghnRNP proteins from pre-mRNA is an essential step formRNA export (Nakielny and Dreyfuss, 1996).

The total number of hnRNPA1 and hnRNPA2molecules in each HeLa cell nucleus is in the order of 70-90 millions; during mitosis these proteins are released intothe cytoplasm and are reimported after biogenesis of thenew nuclear envelopes around the daughter cell nuclei. Theimport rate for these molecules could be 500 molecules perpore per minute assuming one hour for re-accumulation ofthe hnRNPA molecules in the nucleus (Dingwall, 1996).HnRNPA1 is one of a set of hnRNP proteins that do notposses an NLS. A stretch of 38 amino acids in A1, whichinteracts with a human protein called transportin, is bothnecessary and sufficient for nuclear import. Yeast Kap104seems to be the analog of transportin and both display aregion with homology to a domain in importin-" (Görlichand Mattaj, 1996) which might interact with similardomain in nucleoporins to mediate docking of thetransportin-hnRNPA1 or Kap104-protein complex throughthe pore (Aitchison et al, 1996).

X. Import and export of U snRNPs

In contrast to the concept of export of mRNPs thereare cases where RNA-protein complexes are imported intonuclei. Small nuclear ribonucleoprotein particles (snRNPs)in particular U1 snRNPs, are assembled in the cytoplasmand are then imported into nuclei to facilitate splicing.Import of snRNPS and proteins may involve distinctpathways (Fischer et al 1991; Michaud and Goldfarb,1992). The import of U1 snRNPs requires a trimethyl-Gcap structure as well as protein binding to the Sm domainof U1 snRNA (Hamm et al, 1990; Fischer and Lührmann,1990).

Recently a role of the yeast importin-! (SRP1p) innuclear export of capped U snRNAs has been unraveled ina remarkable series of events. Approximately 30% ofSRP1p were found in a nuclear complex with theSaccharomyces cerevisiae nuclear cap-binding proteincomplex (CBC) which promotes nuclear export of cappedU snRNAs and shuttles between nucleus and cytoplasm.Xenopus CBC is associated with importin-! in thenucleus and CBC might shuttle while bound to importin-!. Binding of importin-" in the cytoplasm, a bindingwhich displaces the RNA from the CBC-importin-!complex, and the commitment of CBC for nuclear reentrytrigger and promote the release of capped U snRNAs intothe cytoplasm (Görlich et al, 1996).

XI. Observations on nucleocytoplasmictrafficking pertaining to plasmid import

A model was proposed (Boulikas, 1997a) for theimport of plasmid DNA by taking into consideration thefollowing observations or ideas:


732

(i ) DNA microinjected into Xenopus eggs or nakedDNA incubated with frog oocyte extracts is spontaneouslycondensed into nucleosomes and chromatin and is thenassembled into nuclei in vitro by the formation of a doublenuclear envelope around the condensed DNA; the egg oregg extracts contain all the essential components for thisprocess (Newport, 1987). We expect plasmid DNA to becomplexed with a number of DNA-binding proteins in thecytoplasm and after nuclear import to be converted intochromatin structures; attachment to the nuclear matrix is aprerequisite for transcription and replication.

(i i ) RNA is exported from nuclei in the form of acomplex with proteins (Mehlin et al, 1992); ribosomalsubunits are preformed in the nucleus from importedribosomal proteins and are then exported as large RNA-protein complexes. Some U snRNPs are imported intonuclei (Hamm et al, 1990). We expect expulsion ofplasmid DNA through pore complexes to the cytoplasmafter its nuclear import to be negligible.

(i i i ) Condensation of plasmid DNA with histonesincreases several-fold the efficiency of expression offoreign genes (Wagner et al, 1991; Fritz et al, 1996).

(iv ) Complexation of the DNA with HMG proteinsshortens significantly the time required for gene expressionafter transfection (Kaneda et al, 1989). According to thisprocedure Sendai virus was used to fuse DNA-loadedganglioside liposomes with protein-containing membranevesicles purified from red blood cells; cointroduction ofHMG-1 protein showed rapid uptake of plasmids bynuclei; replacement of HMG-1 by BSA resulted inlocalization of the grains of the in situ hybridization in thecytoplasm after 6 h reaching the nucleus only after about24h (Kaneda et al, 1989).

(v ) Plasmid DNA condensation with polylysine alsoenhances transfection of cell cultures; however, polylysine(18-24 kDa), microinjected into the cytoplasm ofTetrahymena, remained cytoplasmic; polylysine (5-9 kDa)was evenly distributed between the cytoplasm and themicro- and macronuclei of Tetrahymena by diffusionfollowing microinjection; thus, large polylysine moleculescannot be imported into nuclei (White et al, 1989).Polylysine-plasmid complexes are proposed to beuncomplexed in the cytoplasm followed by binding ofnascent nuclear proteins before plasmid import can takeplace. Plasmid complexation with polylysine may onlyhelp internalization through the cell membrane but notnuclear import.

(vi ) Oligonucleotides tagged with NLS target thenucleus more efficiently than free oligos (Seibel et al,1995); the same oligos tagged with mitochondrial signalsenter mitochondria (Seibel et al, 1995).

(vi i ) Agrobacterium tumefaciens elicits tumors onplant hosts by transporting a single-stranded (ss) copy of

transferred DNA (T-DNA) portion of Ti (tumor-inducing)plasmid which enters infected plant cells and integratesinto plant nuclear DNA (direct repeats define the T-DNAends on Ti plasmid). Transfer begins when the VirD2endonuclease produces a site-specific nick. TwoAgrobacterium proteins, VirD2 and VirE2 containing NLSassociate directly with T plasmid and mediate its nuclearimport. VirE2 alone, which has been shown to activelytransport ssDNA into the plant cell nucleus, packagesssDNA into semi-rigid, hollow cylindrical filaments witha telephone cord-like coiled structure as was shown byscanning transmission electron microscopy (STEM); thesecomplexes were proposed to be actively imported throughpore complexes (Citovsky et al, 1997). This is a clearexample of plasmid import mediated by plasmid-associatedproteins possessing NLS.

(v i i i ) Binding NLSs from SV40 T antigen toluciferase plasmid DNA promoted transgene expressionfollowing injection of DNA-NLS complexes into thecytoplasm of zebra fish eggs; NLS peptides, but notnuclear-import-deficient peptides, mediated import of DNAfrom the cytoplasm into embryo nuclei, under conditionsin which naked DNA was not imported. Thus, use of NLSmay reduce the need for elevated DNA copy numbers insome gene transfer applications (Collas et al, 1996; Collasand Alestrom, 1997).

(ix ) Intact, protein-free SV40 DNA was localized tothe nucleus after it was cytoplasmically injected into cellsin a process which was inhibited (i) by wheat germagglutinin (ii) by an anti-nucleoporin antibody whichblock the nuclear pore complex and (iii) by energydepletion. During this process the DNA accumulated at thenuclear periphery before its import and, as opposed toprotein import, DNA import required transcription;furthermore, imported DNA colocalized with the SC-35splicing complex antigen, suggesting localization to areasof transcription or message processing. The SV40 originof replication and the early and late promoters supportedimport, whereas bacterial sequences alone and other SV40-derived sequences did not (Dean, 1997).

(x ) Fusion of liposomes with the cell membrane willrelease the encapsulated DNA into the cytoplasm; thismechanism is rather rare and liposomes seem to beinternalized by receptor mediated endocytosis if appropriateligands are exposed on its surface, by poration, especiallywhen cationic lipids are present, or via phagocytosisending into endosomes and lysosomes (Martin andBoulikas, 1997). Cationic lipids destabilize the biologicalmembranes (both cytoplasmic and lysosomal) and mediaterapid delivery of plasmid to the cytoplasm (reviewed byBoulikas, 1998 page 1-172, this volume). Lysis of theliposome in the endosome or caveolae will release DNA


733

inside vesicles; fusion of the liposome with the endosomeor caveolae membranes will release the DNA into thecytoplasm; the presence of fusogenic peptides on theliposome will promote lysis of the endosomal membraneand release of the liposome-plasmid complex into thecytoplasm.

(xi ) Viruses have evolved different mechanisms forentry into cells (and into nuclei). For example, after entryinto the cytoplasm the adenoviral particle is attached to thecytoplasmic side of pore complexes and the DNA isreleased to the interior of pore annuli entering thenucleoplasm. These highly ordered processes areaccompanied by losses or protease degradation of specificproteins on the viral particles; a viral protease, L3/p23,located inside the capsid at 10 copies per virion, plays akey role in the stepwise dismantling and in the weakeningof the capsid structure culminating with the release of theadenovirus DNA by degrading of the viral capsid proteinVI (Greber et al, 1996).

(xi i ) Spliced mRNAs are exported from nuclei viainteraction with RNA-binding proteins (mainly from theHnRNP family) which contain nuclear export signals.Since these interactions are specific we expect the exportof plasmid DNA, once it is imported to the nucleus, to benegligible.

(x i i i ) Fluorescently labeled oligonucleotides, afterdelivery using DOTAP liposomes, entered the cell usingan endocytic pathway and redistributed from punctatecytoplasmic regions into the nucleus; nuclear uptake tookplace only with positively charged complexes; DOTAPincreased over 100 fold the antisense activity of a specificanti-luciferase oligonucleotide (Zelphati and Szoka, 1997).The nuclear membrane was found to pose a barrier againstnuclear import of oligonucleotides which accumulated inthe perinuclear area; although DOSPA/DOPE liposomescould deliver ODNs into the cytosol, these liposomes wereunable to mediate nuclear import of ODNs; on the contraryoligonucleotide-DDAB/DOPE complexes with a netpositive charge were released from vesicles into thecytoplasm and mediated nuclear import of the oligos(Lappalainen et al, 1997). Labeled oligonucleotidesdelivered to animals by tail vein injection in complexeswith DC-Chol:DOPE liposomes were localized primarilyto phagocytic vacuoles of Kupffer cells at 24 h post-injection; nuclear delivery of oligonucleotide in vivo wasnot observed (Litzinger et al, 1996).

XII. A model for the nuclear import ofplasmid DNA

Taking into account these observations we haveproposed a plausible model for the nuclear import ofplasmid DNA after its cytoplasmic localization (Boulikas,1997a; Figure 5 ). Plasmid is complexed with nuclear

proteins in the cytoplasm as a prerequisite for its import.The binding efficiency and the type of proteins that arecomplexed in the cytoplasm with the plasmid DNA arematters of speculation. A number of studies support themodel that nascent cytoplasmic proteins containing nuclearlocalization signals are complexed with the transfectedDNA and mediate its nuclear import. What type of nascentnuclear proteins might be responsible for mediatingplasmid translocation into nuclei in vivo? Certainlyhistones are abundantly synthesized in dividing cells andhistone H1 has been shown to display an affinity forsupercoiled over relaxed DNA plasmids (Singer and Singer,1976). A number of transcription factors (TFs) and othernuclear proteins are synthesized de novo in activelyproliferating cells (but at lower rates in terminallydifferentiated cells); these proteins could bind to theplasmid, especially to promoters and enhancers in asequence-specific manner, and mediate the import of theplasmid-TF complex.

XIII. Perspectives

Antisense and triplex-forming oligonucleotides, intheir single- or double-stranded form, as well as RNA orDNA oligonucleotides have been extensively used intargeting nuclear DNA. The chemistry for covalentcoupling of oligonucleotides to peptides has beenestablished. Linkage of oligonucleotides to NLS peptidesor to mitochondrial import peptides resulted in nuclear ormitochondrial targeting, respectively (Seibel et al, 1995).Oligo-nucleotides may enter nuclei after theircrosscomplexation with nuclear proteins in the cytoplasm.Studies with fluorescent-labeled single-strandedoligonucleotides show binding to RPA in vitro (CostasKoumenis, Stanford, Personal communication). RPA isthe main single-stranded DNA-binding activity present inmammalian cells.

Understanding the rules that govern trafficking throughthe pore complex is instructive to our comprehension ofplasmid uptake by nuclei during somatic gene transfer andfor developing strategies to overcome obstacles for foreigngene expression by enhancing the nuclear import.

Because of their increase rates of proliferation andprotein import, cancer cells are expected to be moresusceptible to nuclear import of plasmid and to uptaketransfected plasmid at higher rates compared withterminally differentiated cells. However, cancer cellsespecially solid tumors of epithelial origin (lung, colon,head & neck, brain tumors) do not readily internalizeparticles such as liposomes (Martin and Boulikas, 1997);the step of translocation across the cell membrane, and notthe step of nuclear import, is expected to be the ratelimiting step in the overall gene transfer procedure in thesecancer cells.


734

Figure 5 . A model for the import of proteins into nuclei. The pore complex is shown with its octagonal symmetry. S t e p 1 .A complex of the plasmid DNA with one or more proteins possessing NLS (NLS-protein) is formed in the cytoplasm; NLS proteinsmight include histones, HMGs, transcription factors, or other DNA-binding proteins after their de novo synthesis onpolyribosomes; the NLS-protein then binds to karyopherins !/". 2 . The complex is docked by binding to multiple sites on

nucleoporins (structural proteins of the pore complex). 3 . The p10 and Ran-GTP dissociate karyopherin !/NLS-protein-plasmidcomplex which is expelled to the nucleoplasm resulting in plasmid DNA nuclear import. Adapted from Boulikas T (1 9 9 7 a )Nuclear localization signal peptides for the import of plasmid DNA in gene therapy. Int J Oncol 10, 301-309. Reproduced withkind permission from the International Journal of Oncology.


736

Cancer is a disease of the control of the cell cycle andcell signaling involving mutations in a number ofoncogenes and tumor suppressor genes (Spandidos, 1985).Our prediction that tumor cells will import plasmid-protein complexes across the nuclear envelope moreefficiently than nondividing cells provides a basis for thepreferential targeting of cancer cells and might haveimportant implications in human gene therapy.

Acknowledgments

Special thanks to Emile Zuckerkandl for stimulatingdiscussions.

References

Adam SA, Lobl, TJ, Mitchell MA, Gerace L (1 9 8 9 )Identification of specific binding proteins for a nuclearlocation sequence. Nature 337, 276-279.

Agutter PS, McArdle HJ, McCaldin B (1 9 7 6 ) Evidence forinvolvement of nuclear envelope nucleosidetriphosphatase in nucleocytoplasmic translocation ofribonucleoprotein. Nature 263, 165-166.

Aitchison JD, Blobel G, Rout MP (1 9 9 6 ) Kap104p: akaryopherin involved in the nuclear transport ofmessenger RNA binding proteins. Sc ience 274, 624-627.

Akey CW (1 9 8 9 ) Interactions and structure of the nuclearpore complex revealed by cryo-electron microscopy. J

C e l l B i o l 109:,955-970.

Akey CW (1 9 8 9 ) Interactions and structure of the nuclearpore complex revealed by cryo-electron microscopy. J

C e l l B i o l 109, 955-970.

Akey CW, Goldfarb DS (1 9 8 9 ) Protein import through thenuclear pore complex is a multistep process. J Ce l l

B i o l 109, 971-982.

Baeuerle, P.A. and Baltimore, D. (1 9 8 8 a ) Activation of DNA-binding activity in an apparently cytoplasmic precursorof the NF-%B transcription factor. Cel l 53, 211-217.

Baeuerle, P.A. and Baltimore, D. (1 9 8 8 b ) I%B: a specific

inhibitor of the NF-%B transcription factor. Sc ience

242, 540-546.

Bataillé N, Helser T, Fried HM (1 9 9 0 ) Cytoplasmic transportof ribosomal subunits microinjected into the Xenopus

laevis oocyte nucleus: A generalized, facilitated process. JC e l l B i o l 111, 1571-1582.

Beals CR, Sheridan CM, Turck CW, Gardner P, Crabtree GR(1 9 9 7 ) Nuclear export of NF-ATc enhanced by glycogesynthase kinase-3. Sc ience 275, 1930-1934.

Beasley EM, Schatz G (1 9 9 1 ) Import of proteins intomitochondria. Chemtracts. Biochem Mol B io l 2, 305-319.

Borer RA, Lehner CF, Eppenberger HM, Nigg EA (1 9 8 9 )Major nucleolar proteins shuttle between nucleus andcytoplasm. Cel l 56, 379-390.

Boulikas T (1 9 9 3 ) Nuclear localization signals (NLS). Crit

Rev Eukar Gene Expression 3, 193-227.

Boulikas T (1 9 9 4 ) Putative nuclear localization signals(NLS) in protein transcription factors. J Ce l l Biochem

55, 32-58.

Boulikas T (1 9 9 5 ) Phosphorylation of transcription factorsand the control of the cell cycle. Cr i t Rev Eukar Gene

Express ion 5, 1-77.

Boulikas T (1 9 9 6 ) Nuclear import of protein kinases andcyclins. J Cel l Biochem 60, 61-82.

Boulikas T (1 9 9 7 a ) Nuclear localization signal peptides forthe import of plasmid DNA in gene therapy. Int J Oncol

10, 301-309.

Boulikas T (1 9 9 7 b ) Nuclear import of DNA repair proteins.Anticancer Res 17, 843-864.

Boutorine AS, Kostina EV (1 9 9 3 ) Reversible covalentattachment of cholesterol to oligodeoxyribonucleotidesfor studies of the mechanisms of their penetration intoeucaryotic cells. B i o c h i m i e 75, 35-41.

Bui M, Whittaker G, and Helenius A (1 9 9 6 ) Effect of M1protein and low pH on nuclear transport of influenza virusribonucleoprteins. J Virol 70, 8391-8401.

Burke B (1 9 9 0 ) The nuclear envelope and nuclear transport.Curr Opin Cell Biol 2, 514-520, 1990.

Bustamante JO, Hanover JA, Liepins A (1 9 9 5 ) The ionchannel behavior of the nuclear pore complex. J Membr

B i o l 146, 239-251.

Carmo-Fonseca M, Tollervey D, Pepperkok R, Barabino SML,Merdes A, Brunner C, Zamore PD, Green M, Hurt E, andLamond AI (1 9 9 1 ) Mammalian nuclei contain foci whichare highly enriched in components of the pre-mRNAsplicing machinery. EMBO J 10, 195-206.

Carter KC and Lawrence JB (1 9 9 1 ) DNA and RNA within thenucleus; how much sequence-specific spatialorganization? J Cel l Biochem 47, 124-129.

Chelsky, D., Ralph, R. and Jonak, G. (1 9 8 9 ) Sequencerequirements for synthetic peptide-mediated translocationto the nucleus. M o l . C e l l . B i o l . 9, 2487-2492.

Chen Y, Chen C-F, Riley DJ, Allred DC, Chen P-L, Von HoffD, Osborne CK, and Lee W-H (1 9 9 5 ) Aberrant subcellularlocalization of BRCA1 in breast cancer. Sc ience 270,789-791.

Citovsky V, Guralnick B, Simon MN, Wall JS (1 9 9 7 ) Themolecular structure of Agrobacterium VirE2-singlestranded DNA complexes involved in nuclear import. JM o l B i o l 271, 718-727.

Collas P and Alestrom P (1 9 9 7 ) Rapid targeting of plasmidDNA to zebrafish embryo nuclei by the nuclearlocalization signal of SV40 T antigen. Mol Mar Biol

B i o t e c h n o l 6, 48-58.


737

Collas P, Husebye H, Alestrom P (1 9 9 6 ) The nuclearlocalization sequence of the SV40 T antigen promotestransgene uptake and expression in zebrafish embryonuclei. Transgenic Res 5, 451-458.

Cortes P, Ye Z-S, Baltimore D (1 9 9 4 ) RAG-1 interacts withthe repeated amino acid motif of the human homologue ofthe yeast protein SRP1. Proc Nat l Acad Sci USA 91,7763-7637.

Coutavas E, Ren M, Oppenheim JD, D'Eustachio P, RushMG(1 9 9 3 ) Characterization of proteins that interact withthe cell-cycle regulatory protein Ran/TC4. Nature 366,585-587.

Cuomo CA, Kirch SA, Gyuris J, Brent R, Oettinger MA(1 9 9 4 ) Rch1, a protein that specifically interacts withthe RAG-1 recombination-activating protein. Proc Natl

Acad Sci USA 91, 6156-6160.

Dargemont C, Kühn LC (1 9 9 2 ) Export of mRNA frommicroinjected nuclei of Xenopus laevis oocytes. J C e l l

B i o l 118, 1-9.

Dean DA (1 9 9 7 ) Import of plasmid DNA into the nucleus issequence specific. Exp Cel l Res 230, 293-302.

Dingwall C (1 9 9 6 ) Transportin nuclear proteins. Nature

384, 210-211.

Dingwall C, Sharnick SV, Laskey RA (1 9 8 2 ) A polypeptidedomain that specifies migration of nucleoplasmin into thenucleus. Cel l 30, 449-458.

Dworetzky SI, Lanford RE, Feldherr CM (1 9 8 8 ) The effects ofvariations in the number and sequence of targeting signalson nuclear uptake. J Ce l l B io l 107, 1279-1287.

Ernst RK, Bray M, Rekosh D, Hammarskjold ML (1 9 9 7 ) Astructured retroviral RNA element that mediatesnucleocytoplasmic export of intron-containing RNA.M o l C e l l B i o l 17, 135-144.

Feldherr C and Akin D (1 9 9 7 ) The location of the transportgate in the nuclear pore complex. J C e l l S c i 110, 3065-3070.

Feng Y, Gutekunst CA, Eberhart DE, Yi H, Warren ST, HerschSM (1 9 9 7 ) Fragile X mental retardation protein:nucleocytoplasmic shuttling and association withsomatodendritic ribosomes. J Neurosci 17, 1539-1547.

Fischer U, Darzynkiewicz E, Tahara SM, Dathan NA,Lührmann R, Mattaj IW (1 9 9 1 ) Diversity in the signalsrequired for nuclear accumulation of U snRNPs and varietyin the pathways of nuclear transport. J C e l l B i o l 113,705-714.

Fischer U, Huber J, Boelens WC, Mattaj IW, Lührmann R(1 9 9 5 ) The HIV-1 REV activation domain is a nuclearexport signal that accesses an export pathway used byspecific cellular RNAs. Cel l 82, 475-483.

Fischer U, Lührmann R (1 9 9 0 ) An essential signaling rolefor the m3G cap in the transport of U1 snRNP to the

nucleus. Sc ience 249, 786-790.

Floer M, Blobel G (1 9 9 6 ) The nuclear transport factorkaryopherin " binds stoichiometrically to Ran-GTP and

inhibits the Ran GTPase activating protein. J B i o l

Chem 271, 5313-5316.

Fornerod M, Ohno M, Yoshida M, Mattaj IW (1 9 9 7 ) CRM1is an export receptor for leucine-rich nuclear exportsignals. Cel l 90, 1051-1060.

Fortes P, Beloso A, Ortin J (1 9 9 4 ) Influenza virus NS1protein inhibits pre-mRNA splicing and blocks mRNAnucleocytoplasmic transport. EMBO J 13, 704-712.

Fridell RA, Truant R, Thorne L, Benson RE, Cullen BR(1 9 9 7 ) Nuclear import of hnRNP A1 is mediated by anovel cellular cofactor related to karyopherin-". J C e l l

S c i 110, 1325-1331.

Fritz JD, Herweijer H, Zhang G, and Wolff JA (1 9 9 6 ) Genetransfer into mammalian cells using histone-condensedplasmid DNA. Hum Gene Ther 7, 1395-1404.

Fukuda M, Asano S, Nakamura T, Adachi M, Yoshida M,Yanagida M, Nishida E (1 9 9 7 ) CRM1 is responsible forintracellular transport mediated by the nuclear exportsignal. Nature 390, 308-311.

Fukuda M, Gotoh I, Gotoh Y, Nishida E (1 9 9 6 ) Cytoplasmiclocalization of mitogen-activated protein kinase kinasedirected by its NH2-terminal, leucine-rich short aminoacid sequence, which acts as a nuclear export signal. J

Bio l Chem 271, 20024-20028.

Gauthier-Rouviere C, Vandromme M, Lautredou N, Cai QQ,Girard F, Fernandez A, Lamb N (1 9 9 5 ) The serumresponse factor nuclear localization signal: generalimplications for cyclic AMP-dependent protein kinaseactivity in control of nuclear translocation. M o l C e l l

B i o l 15, 433-444.

Gerace L (1 9 9 5 ) Nuclear export signals and the fast track tothe cytoplasm. Cel l 82, 341-344.

Görlich D and Mattaj IW (1 9 9 6 ) Nucleocytoplasmictransport. Sc ience 271, 1513-1518.

Görlich D, Prehn S, Laskey RA, Hartmann E (1 9 9 4 ) Isolationof a protein that is essential for the first step of nuclearprotein import. Cel l 79, 767-778.

Görlich D, Kostka S, Kraft R, Dingwall C, Laskey RA,Hartmann E, Prehn S (1 9 9 5 a ) Two different subunits ofimportin cooperate to recognize nuclear localizationsignals and bind them to the nuclear envelope. Curr

B i o l 5, 383-392.

Görlich D, Vogel F, Mills AD, Hartmann E, Laskey RA(1 9 9 5 b ) Distinct functions for the two importin subunitsin nuclear protein import. Nature 377: 246-248.

Görlich D, Kraft R, Kostka S, Vogel F, Hartmann E, LaskeyRA, Mattaj IW, Izaurraide E (1 9 9 6 ) Importin provides alink between nuclear protein import and U snRNA export.Cel l 87, 21-32.

Greber UF, Webster P, Weber J, and Helenius A (1 9 9 6 ) Therole of the adenovirus protease in virus entry into cells.EMBO J 15, 1766-1777.


738

Guddat U, Bakken AH, Pieler T (1 9 9 0 ) Protein-mediatednuclear export of RNA: 5S rRNA containing small RNPsin Xenopus oocytes. Cel l 60, 619-628.

Guiochon-Mantel A, Delabre K, Lescop P, Milgrom E (1 9 9 4 )Nuclear localization signals also mediate the outwardmovement of proteins from the nucleus. Proc Natl

Acad Sci USA 91, 7179-7183.

Guiochon-Mantel A, Lescop P, Christin-Maitre S, Loosfelt H,Perrot-Applanat M, Milgrom E (1 9 9 1 )Nucleocytoplasmic shuttling of the progesteronereceptor. EMBO J 10, 3851-3859.

Hamm J and Mattaj IW (1 9 9 0 ) Monomethylated capstructures facilitate RNA export from the nucleus. Cel l

63, 109-118.

Hamm J, Darzynkiewicz, Tahara SM, Mattaj IW (1 9 9 0 ) Thetrimethylguanosine cap structure of U1 snRNA is acomponent of a bipartite nuclear targeting signal. Cel l

62, 569-577.

Henkel, T., Zabel, U., van Zee, K., Müller, J.M., Fanning, E.and Baeuerle, P.A. (1 9 9 2 ) Intramolecular masking of thenuclear location signal and dimerization domain in theprecursor for the p50 NF-%B subunit. Cel l 68, 1121-1133.

Hinshaw JE, Carragher BO, Milligan RA (1 9 9 2 ) Architectureand design of the nuclear pore complex. Cel l 69, 1133-1141.

Huang, S and Spector DL (1 9 9 1 ) Nascent pre-mRNAtranscripts are associated with nuclear regions enriched insplicing factors. Genes Dev 5, 2288-2302.

Imamoto N, Tachibana T, Matsubae M, Yoneda Y (1 9 9 5 ) Akaryophilic protein forms a stable complex withcytoplasmic components prior to nuclear pore binding. J

Bio l Chem 270, 8559-8565.

Iovine MK, Watkins JL, Wente SR (1 9 9 5 ) The GLFGrepetitive region of the nucleoporin Nup116p interactswith Kap95p, an essential yeast nuclear import factor. J

C e l l B i o l 131, 1699-1713.

Izaurralde E, Mattaj IW (1 9 9 5 ) RNA export. Cel l 81, 153-159.

Izaurralde E, Stepinski J, Darzynkiewicz E, Mattaj IW (1 9 9 2 )A cap binding protein that may mediated nuclear export ofRNA polymerase II-transcribed RNAs. J C e l l B i o l 118,1287-1295.

James, G. and Olson, E. (1 9 9 2 ) Deletion of the regulatorydomain of protein kinase Ca exposes regions in the hingeand catalytic domains that mediate nuclear targeting. J .

C e l l . B i o l . 116, 863-874.

Kalderon D, Roberts BL, Richardson WD, Smith AE (1 9 8 4 ) Ashort amino acid sequence able to specify nuclearlocation. Cel l 39, 499-509.

Kaneda Y, Iwai K, and Uchida T (1 9 8 9 ) Increased expressionof DNA cointroduced with nuclear protein in adult rat liver.Sc ience 243, 375-378.

Kessler, D.S., Veals, S.A., Fu, X.Y. and Levy, D.E. (1 9 9 0 )Interferon-a regulates nuclear translocation of ISGF3, amultimeric transcription activator. Genes D e v . 4,1753-1765.

Khanna-Gupta A, Ware VC (1 9 8 9 ) Nucleocytoplasmictransport of ribosomes in a eukaryotic system: Is there afacilitated transport process? Proc Natl Acad S c i

USA 86, 1791-1795.

Kim FJ, Beeche AMINO ACID, Hunter JJ, Chin DJ, Hope TJ(1 9 9 6 ) Characterization of the nuclear export signal ofhuman T-cell lymphotropic virus type 1 Rex reveals thatnuclear export is mediated by position-variablehydrophobic interactions. M o l C e l l B i o l 16, 5147-5155.

Klemm JD, Beals CR, Crabtree GR (1 9 9 7 ) Rapid targeting ofnuclear proteins to the cytoplasm. Curr Biol 7, 638-644.

Lappalainen K, Miettinen R, Kellokoski J, Jaminoacidskelainen I, Syrjanen S (1 9 9 7 ) Intracellulardistribution of oligonucleotides delivered by cationicliposomes: light and electron microscopic study.Histochem Cytochem 45, 265-274.

Laskey RA, Gorlich D, Madine MA, Makkerh JPS,Romanowski P (1 9 9 6 ) Regulatory roles of the nuclearenvelope. Exp Cel l Res 229, 204-211.

Lee MS and Silver PA (1 9 9 7 ) RNA movement between thenucleus and the cytoplasm. Curr Opin Genet Dev 7 ,212-219.

Levy, D.E., Kessler, D.S., Pine, R. and Darnell, J.E., Jr.(1989) Cytoplasmic activation of ISGF3, the positiveregulator of interferon-a-stimulated transcription,reconstituted in vitro. Genes Dev 3, 1362-1371.

Lewis JM, Baskaran R, Tamino acidgepera S, Schwartz MA,Wang JY (1 9 9 6 ) Integrin regulation of c-Abl tyrosinekinase activity and cytoplasmic-nuclear transport. Proc


Li X, Shou W, Kloc M, Reddy B, Etkin LD (1 9 9 4 ) Theassociation of Xenopus nuclear factor 7 with subcellularstructures is dependent upon phosphorylation and specificdomains. Exp Cel l Res 213: 473-481.

Litzinger DC, Brown JM, Wala I, Kaufman SA, Van GY,Farrell CL, Collins D (1 9 9 6 ) Fate of cationic liposomesand their complex with oligonucleotide in vivo.Bioch im Biophys Acta 1281, 139-149.

Lutz, R.J., Trujillo, M.A., Denham, K.S., Wenger, L. andSinensky, M. (1 9 9 2 ) Nucleoplasmic localization ofprelamin A: implications for prenylation-dependent laminA assembly into the nuclear lamina. Proc . Nat l . Acad.

Sci . USA 89, 3000-3004.

Lyons, R.H., Ferguson, B.Q. and Rosenberg, M. (1 9 8 7 )Pentapeptide nuclear localization signal in adenovirusE1a. M o l C e l l B i o l 7, 2451-2456.

Mandell RB, Feldherr CM (1 9 9 0 ) Identification of twoHSP70-related Xenopus oocyte proteins that are capable


739

of recycling across the nuclear envelope. J C e l l B i o l

111, 1775-1783.

Martin F and Boulikas T (1 9 9 7 ) The challenge of liposomesin gene therapy. Gene Ther Mol Biol (this volume).

Martin K, Helenius A (1 9 9 1 ) Nuclear transport of influenzavirus ribonucleoproteins: The viral matrix protein (M1)promotes export and inhibits import. Cel l 67, 117-130.

Mehlin H, Daneholt B, Skoglund U (1 9 9 2 ) Translocation of aspecific premessenger ribonucleoprotein particle throughthe nuclear pore studied with electron microscopetomography. Cel l 69, 605-613.

Michael WM, Choi M, Dreyfuss G (1 9 9 5 ) A nuclear exportsignal in hnRNP A1: a signal-mediated, temperature-dependent nuclear protein export pathway. Cel l 83, 415-422.

Michael WM, Eder PS, Dreyfuss G (1 9 9 7 ) The K nuclearshuttling domain: a novel signal for nuclear import andnuclear export in the hnRNP K protein. EMBO J 16,3587-3598.

Michaud N, Goldfarb D (1 9 9 2 ) Microinjected U snRNAs areimported to oocyte nuclei via the nuclear pore complex bythree distinguishable targeting pathways. J C e l l B i o l

116, 851-861.

Middeler G, Zerf K, Jenovai S, Thulig A, Tschodrich-Rotter M,Kubitscheck U, Peters R (1 9 9 7 ) The tumor suppressorp53 is subject to both nuclear import and export, and bothare fast, energy-dependent and lectin-inhibited.Oncogene 14, 1407-1417.

Moll, T., , G., Surana, U., Robitsch, H. and Nasmyth, K.(1 9 9 1 ) The role of phosphorylation and the CDC28protein kinase in cell cycle-regulated nuclear import of theS. cerevisiae transcription factor SWI5. Cel l 66, 743-758.

Moore MS, Blobel G (1 9 9 3 ) The GTP-binding proteinRan/TC4 is required for protein import into the nucleus.Nature 365, 661-663.

Morin, N., Delsert, C. and Klessig, D.F. (1 9 8 9 ) Nuclearlocalization of the adenovirus DNA-binding protein:requirement for two signals and complementation duringviral infection. M o l . C e l l . B i o l . 9, 4372-4380.

Moroianu J, Blobel G, Radu A (1 9 9 6 ) Nuclear protein import:Ran-GTP dissociates the karyopherin !" heterodimer bydisplacing a from an overlapping binding site on b. Proc


Nakielny S and Dreyfuss G (1 9 9 6 ) The hnRNP C proteinscontain a nuclear retention sequence that can overridenuclear export signals. J Ce l l B io l 134, 1365-1373.

Nakielny S and Dreyfuss G (1 9 9 7 ) Nuclear export of proteinsand RNAs. Curr Opin Cell Biol 9, 420-429.

Nasmyth, K., Adolf, G., Lydall, D. and Seddon, A. (1 9 9 0 )The identification of a second cell cycle control on the HO

promoter in yeast: cell cycle regulation of SWI5 nuclearentry. Cel l 62, 631-647.

Neel H, Gondran P, Weil D, Dautry F (1 9 9 5 ) Regulation ofpre-mRNA processing by src. Curr Biol 5, 413-422.

Nehrbass U, Blobel G (1 9 9 6 ) Role of the nuclear transportfactor p10 in nuclear import. Science 272: 120-122.

Neupert W, Lill R (1 9 9 5 ) Mitochondrial protein import:Reversible binding of the presequence at the trans side ofthe outer membrane drives partial translocation andunfolding. Cel l 80, 127-137.

Newport J (1 9 8 7 ) Nuclear reconstitution in vitro: Stages ofassembly around protein-free DNA. Cel l 48, 205-217.

Nigg EA, Baeuerle PA, Lührmann R (1 9 9 1 ) Nuclear import-export: In search of signals and mechanisms. Cel l 66,15-22.

Paschal BM, Gerace L (1 9 9 5 ) Identification of NTF2, acytosolic factor for nuclear import that interacts withnuclear pore complex protein p62. J Ce l l B i o l

129:,925-937.

Peifer M, Berg S, Reynolds AB (1 9 9 4 ) Letter to the Editor. Arepeating amino acid motif shared by proteins withdiverse cellular roles. Cel l 76, 789-791.

Picard D and Yamamoto KR (1 9 8 7 ) Two signals mediatehormone-dependent nuclear localization of theglucocorticoid receptor. EMBO J 6, 3333-3340.

Picard D, Yamamoto KR (1 9 8 7 ) Two signals mediatehormone-dependent nuclear localization of theglucocorticoid receptor. EMBO J 6, 3333-3340.

Picard, D., Kumar, V., Chambon, P. and Yamamoto, K.R.(1 9 9 0 ) Signal transduction by steroid hormones: nuclearlocalization is differentially regulated in estrogen andglucocorticoid receptors. Cel l Reg . 1, 291-299

Pines, J. and Hunter, T. (1 9 9 1 ) Cyclin-dependent kinases: anew cell cycle motif? Trends Cel l Biol 1, 117-121.

Piñol-Roma S, Dreyfuss G (1 9 9 2 ) Shuttling of pre-mRNAbinding proteins between nucleus and cytoplasm. Nature

355, 730-732.

Pollard VW, Michael WM, Nakielny S, Siomi MC, Wang F,Dreyfuss G (1 9 9 6 ) A novel receptor-mediated nuclearprotein import pathway. Cel l 86, 985-994.

Radu A, Moore MS, Blobel G (1 9 9 5 ) The peptide repeatdomain of nucleoporin Nup98 functions as a docking sitein transport across the nuclear pore complex. Cel l 81,215-222.

Rexach M and Blobel G (1 9 9 5 ) Protein import into nuclei:association and dissociation reactions involvingtransport substrate, transport factors, and nucleoporins.Cel l 83, 683-692.

Richardson WD, Mills AD, Dilworth SM, Laskey RA,Dingwall C (1 9 8 8 ) Nuclear protein migration involvestwo steps: Rapid binding at the nuclear envelope followedby slower translocation through nuclear pores. Cel l 52,655-664.

Roberts, B.L., Richardson, W. D. and Smith, A.E. (1 9 8 7 )The effect of protein context on nuclear location signalfunction. Cel l 50, 465-475.


740

Roux, P., Blanchard, J.-M., Fernandez, A., Lamb, N.,Jeanteur, P. and Peichaczyk, M. (1 9 9 0 ) Nuclearlocalization of c-Fos, but not v-Fos proteins, iscontrolled by extracellular signals. Cel l 63, 341-351.

Rubinfeld B, Souza B, Albert I, Muller O, Chamberlain SH,Masiarz FR, Munemitsu S, Polakis P (1 9 9 3 ) Associationof the APC gene product with b-catenin. Sc ience 262:1731-1734.

Sachdev S, Diehl JA, McKinsey TA, Hans A, Hannink M(1 9 9 7 ) A threshold nuclear level of the v-Reloncoprotein is required for transformation of avianlymphocytes. Oncogene 14, 2585-2594.

Sackey FN, Hache RJ, Reich T, Kwast-Welfeld J, Lefebvre YA(1 9 9 6 ) Determinants of subcellular distribution of theglucocorticoid receptor. Mol Endocrinol 10, 1191-1205.

Schmidt-Zachmann MS, Dargemont C, Kuhn LC, Nigg EA(1 9 9 3 ) Nuclear export of proteins: The role of nuclearretention. Cel l 74, 493-504.

Seibel P, Trappe J, Villani G, Klopstock T, Papa S, andReichman H (1 9 9 5 ) Transfection of mitochondria:strategy towards a gene therapy of mitochondrial DNAdiseases. Nucleic Acids Res 23, 10-17.

Shelton, C.A. and Wasserman, S.A. (1 9 9 3 ) pelle encodes aprotein kinase required to establish dorsoventral polarityin the Drosophila embryo. Cel l 72, 515-525.

Simos G, Tekotte H, Grosjean H, Segref A, Sharma K,Tollervey D, Hurt EC (1 9 9 6 ) Nuclear pore proteins areinvolved in the biogenesis of functional tRNA. EMBO J

15: 2270-2284.

Singer DS and Singer MF (1 9 7 6 ) Studies on the interaction ofH1 histone with superhelical DNA: Characterization of therecognition and binding regions of H1 histone. Nucleic

Acids Res . 3, 2531-2547.

Siomi MC, Eder PS, Kataoka N, Wan L, Liu Q, Dreyfuss G(1 9 9 7 ) Transportin-mediated nuclear import ofheterogeneous nuclear RNP proteins. J C e l l B i o l 138,1181-1192.

Snow CM, Senior A, Gerace L (1 9 8 7 ) Monoclonal antibodiesidentify a group of nuclear pore complex glycoproteins. J

C e l l B i o l 104, 1143-1156.

Spandidos, DA (1 9 8 5 ) Mechanism of carcinogenesis: Therole of oncogenes, transcriptional enhancers and growthfactors. Anticancer Res 5, 485-498.

Stade K, Ford CS, Guthrie C, Weis K (1 9 9 7 ) Exportin 1(Crm1p) is an essential nuclear export factor. Cel l 90,1041-1050.

Standiford DM, Richter JD (1 9 9 2 ) Analysis of adevelopmentally regulated nuclear localization signal inXenopus. J Ce l l B io l 118: 991-1002.

Su L-K, Vogelstein B, Kinzler KW (1 9 9 3 ) Association of theAPC tumor suppressor protein with catenins. Sc ience

262, 1734-1737.

Unwin, P. N. T., and Milligan, R. A. (1 9 8 2 ) A large particleassociated with the perimeter of the nuclear pore complex.J . C e l l B i o l . 93, 63-75.

Wagner E, Cotten M, Foisner R, Birnstiel ML (1 9 9 1 )Transferrin-polycation-DNA complexes: The effect ofpolycations on the structure of the complex and DNAdelivery to cells. Proc Nat l Acad Sci USA 88, 4255-4259.

Weeks DL and Melton DA (1 9 8 7 ) A maternal mRNA localizedto the vegetal hemisphere in Xenopus eggs codes for agrowth factor related to TGF-b. Cel l 51, 861-867,.

Weis K, Mattaj IW, Lamond AI (1 9 9 5 ) Identification ofhSRP1a as a functional receptor for nuclear localizationsequences. Sc ience 268, 1049-1053.

Welsh JD, Swimmer C, Cocke T, Shenk T (1 9 8 6 ) A seconddomain of simian virus 40 T antigen in which mutationscan alter the cellular localization of the antigen. M o l

C e l l B i o l 6, 2207-2212.

Wen W, Meinkoth JL, Tsien RY, Taylor SS (1 9 9 5 )Identification of a signal for rapid export of proteins fromthe nucleus. Cel l 82, 463-473.

White EM, Allis CD, Goldfarb DS, Srivastva A, Weir JW, andGorovsky MA (1 9 8 9 ) Nucleus-specific and temporallyrestricted localization of proteins in Tetrahymena

macronuclei and micronuclei. J C e l l B i o l 109, 1983-1992.

Wong K-K, Zou X, Merrell KT, Patel AJ, Marcu KB,Chellappan S, and Calame K (1 9 9 5 ) v-Abl activates c-myc transcription through the E2F site. M o l C e l l B i o l

15, 6535-6544.

Wozniak RW, Bartnik E, Blobel G (1 9 8 9 ) Primary structureanalysis of an integral membrane glycoprotein of thenuclear pore. J Cell Biol 108, 2083-2092.

Yano R, Oakes M, Yamaghishi M, Dodd JA, Nomura M(1 9 9 2 ) Cloning and characterization of SRP1, asuppressor of temperature-sensitive RNA polymerase Imutations, in Saccharomyces cerevisiae. Mol Ce l l

B i o l 12, 5640-5651.

Yano R, Oakes ML, Tabb MM, Nomura M (1 9 9 4 ) YeastSrp1p has homology to armadillo/plakoglobin/b-cateninand participates in apparently multiple nuclear functionsincluding the maintenance of the nucleolar structure. Proc


Yokoyama N, Hayashi N, Seki T, Panté N, Ohba T, Nishii K,Kuma K, Hayashida T, Miyata T, Aebi U, Fukui M,Nishimoto T (1 9 9 5 ) A giant nucleopore protein thatbinds Ran/TC4. Nature 376, 184-188.

Yoneda Y, Imamoto-Sonobe N, Matsuoka Y, Iwamoto R, KihoY, Uchida T (1 9 8 8 ) Antibodies to Asp-Asp-Glu-Asp caninhibit transport of nuclear proteins into the nucleus.Sc ience 242, 275-278.

Zasloff M (1 9 8 3 ) tRNA transport from the nucleus in aeukaryotic cell: Carrier-mediated translocation process.Proc Natl Sci Acad USA 80, 6436-6440.


741

Zelphati O, Szoka FC Jr (1 9 9 7 ) Intracellular distribution andmechanism of delivery of oligonucleotides mediated bycationic lipids. Pharm Res 13, 1367-1372.

Zervos AS, Byuris J, Brent R (1 9 9 3 ) Mxi1, a protein thatspecifically interacts with Max to bind Myc-Maxrecognition sites. Cel l 72, 223-232.

Zhang G, Zapp ML, Yan G, Green MR (1 9 9 6 ) Localization ofHIV-1 RNA in mammalian nuclei. J C e l l B i o l 135, 9-18.

Zhao, L.-J. and Padmanabhan, R. (1 9 8 8 ) Nuclear transport ofadenovirus DNA polymerase is facilitated by interactionwith preterminal protein. Cel l 55, 1005-1015.

Zhao, L.-J. and Padmanabhan, R. (1 9 9 1 ) Three basic regionsin adenovirus DNA polymerase interact differentiallydepending on the protein context to function as bipartitenuclear localization signals. New Bio l . 3, 1074-1088.


741


The ATP-driven protein translocation-motor ofmitochondria

Chitkala Satyanarayana and Martin Horst*

Zentrum für Biochemie und Molekulare Zellbiologie der Universität Göttingen, Abtl. Biochemie 2, Gosslerstr. 12d, D-

37073 Göttingen, Germany

__________________________________________________________________________________________________* Corresponding author, tel: (49) 551 39 5949, fax: (49) 551 39 5979; e-mail: [email protected]

Summary

The majority of mitochondrial proteins are encoded in the nucleus. In the cytosol they are

synthesized as precursor proteins which are transported into mitochondria. Protein import into

mitochondria requires a concerted action of a variety of different proteins. For the transport of

precursor proteins across the mitochondrial inner membrane and their folding in the matrix the

mitochondrial hsp70 (mhsp70) chaperone plays an essential role. Mhsp70 is found in at least two

protein complexes within mitochondria: together with Tim44 and mGrpE, mhsp70 forms the

import-complex and together with Mdj1 and mGrpE i t forms a folding-complex. This review

focuses on the function of the import-complex. It is believed that mhsp70 can act as a mechano-

chemical enzyme that actively pulls precursor proteins across the inner membrane.

I. Introduction

Mitochondria do contain their own DNA. However,

this DNA encodes only about a dozen proteins. The

majority of the mitochondrial proteins are encoded in the

nucleus and are imported from the cytosol. These proteins

are synthesized in the cytosol as precursor proteins

containing usually an N-terminal mitochondrial targeting

signal. They are co- or post-translationally transported into

mitochondria. Cytosolic members of the hsp70 family as

well as a mitochondrial import stimulating factor facilitate

the transfer of the precursor to the receptor complex

(Hachiya et al., 1995). Different precursor proteins bind to

different subcomplexes of the heterotetrameric receptor.

These precursors are then delivered to the insertion pore

(Hachiya et al., 1995; Komiya et al., 1997). The

translocation of precursor proteins across both

mitochondrial membranes is facilitated by the components

of the translocase found in the outer membrane (Tom) and

the translocase found in the inner membrane (Tim).

Translocation requires a loosely folded conformation of the

precursor protein during import, an electrochemical

potential across the mitochondrial inner membrane and

ATP in the matrix space. Following translocation, the

presequence is removed by the matrix localized processing

peptidase. Finally, the precursor protein folds with or

without the help of the matrix localized chaperones.

Mitochondrial hsp70 (mhsp70) is essential for the

translocation across the inner membrane for all matrix

targeted precusor proteins as well as for the folding of

some precursor proteins in the matrix space.

One of the most fascinating questions about protein

import into mitochondria is how a matrix localized

mhsp70 chaperone can facilitate import of precursor

proteins from the cytosol into mitochondria. In this review

we will focus on this aspect of mhsp70 function.

II. The HSP70 chaperone family

Heat shock proteins of the Hsp70 family are found in

nearly all organisms. These proteins play an essential role

in protein folding, transport into different cellular

compartments, and regulation of the heat-shock response

(reviewed in: Hightower et al., 1994; McKay et al., 1994).

The members of this protein family are highly conserved

from bacteria to man (Lindquist and Craig 1988; Boorstein

et al., 1994). Some members of the hsp70 family are

constitutively expressed whereas others are expressed only

under stress conditions (Lindquist and Craig, 1988;

Boorstein et al., 1994). Hsp70 proteins bind to unfolded,

Satyanarayana and Horst: The ATP-driven protein translocation-motor of mitochondria

742

hydrophobic surface-exposed segments of polypeptide

chains (Pelham, 1986; Blond Elguindi et al., 1993; Flynn

et al., 1991; Landry et al., 1992; Zhu et al., 1996).

However, the actual mechanism of hsp70 function is still

not completely understood. Binding and release of

substrates (polypeptides or peptides) by hsp70 is regulated

by adenine nucleotides. The conformation of hsp70

changes during the cyle of ATP-binding, ATP-hydrolysis,

and release of ADP and Pi (Flaherty et al., 1990;

Hightower et al., 1994; von Ahsen et al., 1996). Studies

on DnaK, the hsp70 protein of E. coli, demonstrated that

optimal functioning of this chaperone requires the co-

chaperones DnaJ and GrpE (reviewed in Georgopoulos and

Welch, 1993). DnaJ accelerates ATP hydrolysis by DnaK,

whereas GrpE acts as an adenine nucleotide-exchange factor

for DnaK (Liberek et al., 1991; McCarty et al., 1995).

Homologs of DnaJ and/or GrpE cooperate with hsp70

proteins in the eukaryotic cytosol and in the lumenal

spaces of mitochondria and the endoplasmic reticulum

(Bolliger et al., 1994; Caplan et al., 1993; Ikeda et al.,

1994; Laloraya et al., 1994; Schlenstedt et al., 1995;

Horst et al., 1997a).

A. HSP70 chaperones involved in proteintranslocation

a. BiP

The lumen of the endoplasmic reticulum (ER) contains

an abundant hsp70 chaperone: BiP (heavy chain binding

protein) in mammals and Kar2p in yeast. The protein was

initially identified because of its key role in the folding of

newly-imported ER proteins (Brodsky and Schekman,

1993). Later it was found that this protein also mediates

the ATP-dependent translocation of proteins into the yeast

ER, and that this function involves the ATP-regulated

interaction of BiP/Kar2p with the membrane protein

Sec63p (reviewed in: Brodsky and Schekman, 1994;

Brodsky, 1996). Kar2p is needed for co- as well as post-

translational translocation into the yeast ER (Brodsky et

al., 1995). When ER membranes are solubilized and

fractionated in the absence of ATP, Kar2p copurifies with

Sec63p (Brodsky and Schekman, 1993). Genetic studies

have confirmed the functional importance of the Kar2p-

Sec63p interaction (Scidmore et al., 1993). Kar2p appears

to recognize a region in Sec63p that is homologous to the

conserved "J domain" in members of the DnaJ protein

family (Sadler et al., 1989; Ang et al., 1991).

b. Mitochondrial HSP70

In Saccharomyces cerevisiae the product of the SSC1

gene (mitochondrial hsp70) is located within mitochondria.

Genetic and biochemical evidence implicate mhsp70 as a

component of the mitochondrial protein import machinery.

Mhsp70 is essential for growth (Craig et al., 1987). In

vitro import studies of mitochondria isolated from a

temperature sensitive mhsp70 mutant showed defects in

protein import and an accumulation of precursor proteins

at contact sites (Kang et al., 1990). Furthermore, mhsp70

can be crosslinked to or co-immunoprecipitated with

precursor proteins on their way to the matrix (Kang et al.,

1990; Scherer et al., 1990). Subsequent experiments have

shown that mhsp70 forms a transient, ATP-dependent

interaction with newly imported precursor proteins

(Manning-Krieg et al., 1991). Interestingly, it has been

shown that a temperature sensitive mhsp70 allele fails to

bind and to complete import of a partially translocated

precursor protein (Gambill et al., 1993). These data and

others suggest that mhsp70 is playing a major role in

mitochondrial protein import and protein folding (reviewed

in: Langer and Neupert, 1994; Rassow et al., 1996).

Mhsp70 performs these different functions together with

several different partner proteins.

The first identified mhsp70 complex was the site-

specific endonuclease Endo.SceI, which is a dimer

consisting of mhsp70 and a 50-kDa nuclease subunit

(Morishima et al., 1990). In the heterotetrameric protein

folding complex, mhsp70 works with two partner proteins

(Horst et al., 1997a): Mdj1p, a homolog of bacterial DnaJ

(Rowley et al., 1994); and a mitochondrial GrpE (mGrpE)

dimer, a homolog of bacterial GrpE (bGrpE) (Bolliger et

al., 1994; Laloraya et al., 1994; Nakai et al., 1994). Mdj1

is itself a chaperone (Prip-Buus et al., 1996). However, it

also stimulates the ATPase-activity of mhsp70 suggesting

that it supports mhsp70 function during protein folding

(Horst et al., 1997a). MGrpE functions as a mitochondrial

ADP-ATP exchange factor for mhsp70 (Azem et al.,

1997). In E. coli the hsp70 chaperone system comprises

of three proteins: DnaK, DnaJ, and bGrpE (Georgopoulos

and Welch, 1993; Szabo et al., 1994). The mitochondrial

hsp70 folding system is the only one in eukaryotes that

contains a DnaJ as well as a bGrpE homolog, suggesting

that the mitochondrial system is the closest eukaryotic

homolog of the bacterial system (Bolliger et al., 1994;

Laloraya et al., 1994; Nakai et al., 1994; Rowley et

al.,1994; Westermann et al., 1995; Prip-Buus et al., 1996;

Horst et al. 1997a). In the import complex, mhsp70 forms

a heterotetrameric complex with Tim44 and a mGrpE

dimer (Kronidou et al., 1994; Rassow et al., 1994;

Schneider et al., 1994). Tim44 functions in this complex

as a membrane anchor for mhsp70. Tim44 spans the

mitochondrial inner membrane, appears to be intimately

associated with the translocation channel (Maarse et al.,

1992; Scherer et al., 1992) and contains a domain that is

weakly homologous to the J domain of Sec63p. It was

suggested that the J-like domain of Tim44 is the binding

site for mhsp70 (Rassow et al., 1994). If so, the


743

translocation systems of mitochondria and the ER would

be rather similar (see however, Brodsky, 1996).

III. Protein translocation models

What is the mechanism for the translocation of

precursor proteins across the mitochondrial membranes?

Two translocation mechanisms were suggested: a

"Brownian-Ratchet" and "Translocation-Motor" (Neupert et

al., 1990; Simon et al., 1992; Glick, 1995; Pfanner and

Meijer, 1995; Horst et al., 1997b).

A. The "Brownian-Ratchet" model

In the "Brownian-Ratchet" model, precursor

polypeptides randomly oscillate within the translocation

channel. In the first version of the "Brownian-Ratchet"

model this random oscillation was suggested to be only

due to Brownian-molecular motion. When the precursor

protein moves inward, a mhsp70 molecule binds to the

emerging segment of the polypeptide chain in the matrix,

thereby preventing its reverse movement. Repeating such a

binding event will finally lead to the translocation of the

entire precursor protein across the inner membrane into the

matrix space (Figure 1). Tim44 can have two different

functions: a more passive function just as a membrane

anchor for mhsp70 or a more active one as a promotor for

binding of mhsp70 molecules to the precursor chain.

Tim44 could perform the latter function by acting in a

DnaJ-like fashion to catalyze ATP hydrolysis by mhsp70.

The observation that precursor chains can slide

bidirectionally in the mitochondrial translocation channel

strongly supports the "Brownian-Ratchet" model

(Ungermann et al., 1994). This "sliding" was suggested to

be due to random thermal motion of the polypeptide chain

in the translocation channel (Neupert et al., 1990; Simon

et al., 1992; Glick, 1995; Pfanner and Meijer, 1995; Horst

et al., 1997b).

If Brownian motion was the sole factor involved in

protein translocation a precursor protein which can not be

fully imported due to a tightly folded domain should slide

back out of the translocation channel in a few

milliseconds. In reality this process takes some minutes

(Ungermann et al., 1996). This suggests that some sort of

interaction occurs between the translocating polypeptide

and the components of the translocation channel, hindering

free diffusion of the polypeptide chain in the translocation

channel. The created "friction" between the polypeptide

chain in the translocation channel could partly be

responsible for the unidirectionality of the translocation

process. This would assume that the "friction" for the

inward movement of the polypeptide chain is smaller than

the one for the opposite movement of the polypeptide

chain.

An extended version of the "Brownian-Ratchet" model

has been proposed to explain the import of precursor

proteins containing a tightly folded domain. In this model

the folded domain spontaneously unfolds at the

mitochondrial surface due to random thermal movements

("breathing" of the molecule). This would allow an inward

movement of the precursor chain (Stuart et al., 1994).

Thus, the rate of import of such a folded precursor protein

is limited by its spontaneous unfolding rate. For example,

fusion proteins containing mouse dihydrofolate-reductase

(DHFR) are imported into mitochondria within 10-20

minutes at room temperature (Endo and Schatz, 1988); and

DHFR spontaneously unfolds, on average, once every few

minutes at this temperature (Viitanen et al., 1991). Such

import kinetics can easily be explained by the "Brownian-

Ratchet" model. However, the DHFR unfolds

spontaneously very fast, much faster than the majority of

other proteins. A single unfolding transition usually

requires hours or even days (Creighton, 1993). For

example, this is true for the heme-binding domain of

cytochrome b2, which is already tightly folded after release

of the nascent chain from the ribosome. Nevertheless, the

precursor is imported in vitro within only a few minutes

(Glick et al., 1993). Interestingly, in contrast to DHFR

fusion proteins, cytochrome b2 absolutely requires both

matrix ATP and functional mhsp70 for import (Glick et

al., 1993; Voos et al., 1993; Stuart et al., 1994),

suggesting that mhsp70 uses the energy of ATP

hydrolysis to accelerate unfolding of the heme-binding

domain at the mitochondrial surface.

B. The "ATP-dependent translocation-motor" model

The above mentioned results and others have lead to

another model for the translocation across the

mitochondrial inner membrane, the so called "ATP-

dependent Translocation-Motor" model (Glick, 1995;

Pfanner and Meijer, 1995, Horst et al., 1997). In this

model the precursor bound mhsp70 undergoes a

conformational change upon ATP hydrolysis, thereby

generating an inward force on the precursor chain. As a

result the polypeptide is pulled into the matrix space

(Figure 1 ). The role of Tim44 in both translocation

models would be different: in the "Brownian-Ratchet"

model Tim44 positions mhsp70 just close to the import

site, whereas in the "Translocation-Motor" model Tim44

would not only position mhsp70 close to the import pore,

but would furthermore allow for force generation by

serving as a membrane anchor for mhsp70 during its

"powerstroke". In both models, Tim44 could also promote


744

F i g . 1 Comparison of the

"Brownian-Ratchet" and the

"Translocation-Motor" model for

the translocation of polypeptides

across the mitochondrial inner

membrane (adapted from: Glick et

al, 1995). Dark grey: translocation

channel; light grey: mhsp70; IM:

Mitochondrial inner membrane.

the binding of mhsp70 to the precursor chain. In the

"ATP-dependent Translocation-Motor", the action of

mhsp70 during protein import is analogous to the ATP-

dependent translocation of actin filaments by myosin

(Spudich, 1994).

To test the prediction that mhsp70 can accelerate

unfolding of a precursor protein the following experiments

using a precursor protein containing the DHFR-domain

fused to the cytochrome b2 presequence of variable length

were performed. Precursor whose presequence is not long

enough to span both mitochondrial membranes, and

having a folded DHFR-domain, are imported slowly.

Precursor proteins with a longer presequence are imported

orders of magnitude faster (Matouschek et al., 1997). It

seems therefore that if the presequence of these precursor

proteins is long enough to span both mitochondrial

membranes, mhsp70 can actively unfold the tightly folded

heme-binding domain on the mitochondrial surface.

Precursor proteins with a short presequence can not be

unfolded as the presequence can not interact with mhsp70.

Mhsp70 may have an unfolding activity linked to a

conformational change which generats a power stroke on

the precursor. If the "Translocation-Motor" model should

hold true it has to be ruled out that the environment at the

entrance of the mitochondrial import channel can unfold

precursor proteins. There is some evidence against that:

import of DHFR fusion proteins with short presequences

is not faster than the spontaneous unfolding of DHFR in

solution. Furthermore, in ATP-depleted mitochondria the

cytochrome b2 presequence inserts proper into the outer

membrane close to the import site and the heme-binding

domain remains stably folded (Glick et al., 1993).

Another approach to test the two different models is to

characterize the nucleotide-dependent interactions (ATPase-

cycle) of mhsp70 with Tim44 and a precursor protein.

According to the "Translocation-Motor" model the

generation of a force on the translocating precursor requires

that at some point hsp70 must be bound simultaneously

to its membrane anchor and to the translocating

polypeptide chain. In contrast the "Brownian-Ratchet"

model requires that mhsp70 dissociates from the membrane

anchor after ATP-hydrolysis so that the translocating

polypeptide chain can continue to oscillate in the channel.

ATP promotes the dissociation of hsp70 proteins from

unfolded polypeptides due to nucleotide dependant

conformational changes (Hightower et al., 1994). ATP

also disrupts the Tim44-mhsp70 and Sec63p-BiP/Kar2p

complexes (Brodsky and Schekmann, 1993; Kronidou et

al., 1994; Schneider et al., 1994; Rassow et al., 1994).

Recent kinetic studies have suggested that ATP binding

rather than ATP hydrolysis leads to the release of hsp70

proteins from peptide substrates (Palleros et al., 1993;

Prasad et al., 1994; Schmid et al., 1994; Azem et al.,

1997). Therfore the effect of poorly hydrolysable ATP

analogs on the interactions between mhsp70 and Tim44

was investigated (Horst et al., 1996). These analogs

promote dissociation of mhsp70 from Tim44, whereas

mhsp70 remains bound to Tim44 in the presence of ADP.

Thus, the current evidence implies that the ATP-bound

form of mhsp70 reversibly associates with Tim44 and

incoming precursor proteins, whereas ATP hydrolysis


745

generates an ADP-bound form of mhsp70 that associates

tightly with Tim44 and the precursor. A conformational

change of the ADP-bound mhsp70 molecule would then

exert an inward pulling force on the precursor chain

The presequence of a mitochondrial precursor is first

translocated across both membranes into the matrix where

it interacts with mhsp70. Experiments with different

length precursors have shown that at least 50 residues are

needed to span the two mitochondrial membranes (Rassow

et al., 1990). Therefore some process other than mhsp70-

dependent pulling must initiate the translocation process.

It was suggested that the electrochemical potential across

the inner membrane is essential for insertion of the

mitochondrial presequence across the inner membrane

(Schleyer and Neupert, 1985; Cyr et al., 1993). Many

mitochondrial precursor proteins like DHFR fusion

proteins with presequences as short as 12 amino acids can

be imported, even if the DHFR moieties are initially

folded (Hurt et al., 1985). In these cases the DHFR-domain

is most likely spontaneously unfolded on the

mitochondrial surface ("breathing" of the DHFR domain)

followed by the translocation of the presequence into the

matrix where it comes into contact with mhsp70. The

F i g . 2 Model of the reaction cycle of the ATP-dependent

"Translocation-Motor" (modified from: Horst et al., 1996).

For details see paragraph IV. Dark grey: translocation

channel; light grey: mhsp70 and mGrpE; IM: Mitochondrial

inner membrane.

"Brownian Ratchet" model may therefore account for this

initial step in the import process.

Taken together these results have lead us to propose

the following schematic representation of the mhsp70

action (Glick, 1995; Pfanner and Meijer, 1995, Horst et

al., 1997b; Figure 2).

IV. The function of the "translocation-motor"

Based on the observation that ATP is initially needed

for the interaction of mhsp70 with the incoming precursor

chain (Schneider et al., 1994) in stage 1, mhsp70-ATP

associates transiently with Tim44 and the precursor. In the

next stage mhsp70 hydrolyzes ATP and mhsp70-ADP

associates stably with the precursor and Tim44. This is

consistent with the observation that in the absence of

ATP, mhsp70 copurifies with Tim44 (Kronidou et al.,

1994; Schneider et al., 1994; Rassow et al., 1994) and that

mhsp70, the precursor and Tim44 form a stable complex

under these conditions (Horst et al., 1996). In stage 3,

mhsp70 undergoes a conformational change thereby

pulling a stretch of the precursor inwards. Indeed, it was

recently shown that upon ATP-binding mhsp70 undergoes

a conformational change (von Ahsen et al., 1996). In stage

4, an ADP-ATP exchange reaction, which is facilitated by

mGrpE (Azem et al., 1997) takes place. Following the

conformational change and the ATP-ADP exchange,

mhsp70-ATP dissociates from Tim44 and the precursor

(stage 5). This cycle (Figure 2 ) is repeated until the

precursor protein is completely imported into the matrix.

According to this model mhsp70 together with its

associated proteins function remarkably similar to the

proposed ATP-dependent mechanism of other hsp70

chaperones (Greene et al., 1995; McCarty et al., 1995) and

to the conventional force generating systems such as the

interaction found between myosin and actin cables

(Spudich et al., 1994).

V. Outlook

Future experiments on the structure of this

translocation complex and the interaction of its subunits at

the molecular level should allow us to distinguish between

the two translocation models. Most interestingly, in the

intermembrane space of chloroplasts there is an hsp70 like

protein anchored to the inner face of the outer membrane

(Marshall et al., 1990; Schnell et al., 1994). This protein

may also function by pulling precursor proteins across the

chloroplast outer membrane. The "Translocation-Motor"

may therefore be an universal mechanism for translocating

proteins across organellar membranes.

It could also be possible to use techniques originally

invented for studies in the motor protein field (Walker and

Sheetz, 1993): mhsp70 could be coupled to a plane solid

support so that all molecules are aligned. If mhsp70


746

functions as a molecular motor a precursor protein may

move along the mhsp70 molecules.

Acknowledgements

We thank W. Oppliger and S. Fedkenhauer for

excellent technical assistence. We are indebted to members

of our laboratories for helpful discussions. We would

especially like to thank Drs. N.G. Kronidou and P.V.

Schu for comments on the manuscript. M. Horst was

supported by a Heisenberg fellowship and by grants from

the German Research Society (DFG).

References

Ang, D., Liberek, K., Skowyra, D., Zylicz, M., and

Georgopoulos, C. ( 1 9 9 1 ) . Biological role and

regulation of the universally conserved heat shock

proteins. J . B io l . Chem. 266, 24233-24236.

Azem, A., Oppliger, W., Lustig, A., Jenö, P., Feifel, B.,

Schatz, G., and Horst, M. ( 1 9 9 7 ) . The mitochondrial

hsp70 chaperone system. Effect of adenine nucleotides,

peptide substrate, and mGrpE on the oligomeric state of

mhsp70. J . B io l . Chem. 272, 20901-20906.

Blond-Elguindi, S., Cwirla, S.E., Dower, W.J., Lipshutz, R.J.,

Sprang, S.R., Sambrook, J.F., and Gething, M.-J.

( 1 9 9 3 ) . Affinity panning of a library of peptides

displayed on bacteriophages reveals the binding

specificity of BiP. Cel l 75, 717-728.

Bolliger, L., Deloche, O., Glick, B.S., Georgopoulos, C.,

Jenö, P., Kronidou, N.G., Horst, M., Morishima, N., and

Schatz, G. ( 1 9 9 4 ) . A mitochondrial homolog of

bacterial GrpE interacts with mitochondrial hsp70 and is

essential for viability. EMBO J. 13, 1998-2006.

Boorstein, W.R., Ziegelhoffer, T., Craig, E.A. ( 1 9 9 4 ) .

Molecular evolution of the HSP70 multigene family. J .

M o l . E v o l . 38, 1-17.

Brodsky, J.L. ( 1 9 9 6 ) . Post-translational protein

translocation: not all hsc70s are created equal. Trends.

B i o c h e m . S c i . 21, 122- 126.

Brodsky, J.L., Goeckeler, J., and Schekman, R. ( 1 9 9 5 ) . BiP

and Sec63p are required for both co- and posttranslational

protein translocation into the yeast endoplasmic

reticulum. P r o c . N a t l . A c a d . S c i . U S A 92, 9643-

9646.

Brodsky, J.L., and Schekman, R. ( 1 9 9 4 ) . Heat shock

cognate proteins and polypeptide translocation across the

endoplasmic reticulum membrane. In: The Biology of Heat

Shock Proteins and Molecular Chaperones, R.I.

Morimoto, A. Tissières and C. Georgopoulos, eds.: Cold

Spring Harbor Laboratory Press), pp. 85-109.

Brodsky, J.L., and Schekman, R. ( 1 9 9 3 ) . A Sec63p-BiP

complex from yeast is required for protein translocation in

a reconstituted proteoliposome. J. Ce l l B i o l . 123,

1355-1363.

Caplan, A.J., Cyr, D.M., and Douglas, M.G. ( 1 9 9 3 ) .

Eukaryotic homologues of Escherichia coli dnaJ: a diverse

protein family that functions with hsp70 stress proteins.

M o l . B i o l . C e l l 4, 555-563.

Craig, E.A., Kramer, J., Kosic-Smithers, J. ( 1 9 8 7 ) . SSC1, a

member of the 70-kDa heat shock protein multigene

family of Saccharomyces cerevisiae, is essential for

growth. Proc. Natl . Acad. Sci . USA 84, 4156-4160.

Creighton, T.E. ( 1 9 9 3 ) . Proteins. New York, Freeman.

Cyr, D.M., Stuart, R.A., and Neupert, W. ( 1 9 9 3 ) . A matrix

ATP requirement for presequence translocation across the

inner membrane of mitochondria. J . B i o l . C h e m . 268,

23751-23754.

Endo, T., and Schatz, G. ( 1 9 8 8 ) . Latent membrane

perturbation activity of a mitochondrial precursor protein

is exposed by unfolding. EMBO J. 7, 1153-1158.

Flaherty, K.M., DeLuca-Flaherty, C., and McKay, D.B.

( 1 9 9 0 ) . Three-dimensional structure of the ATPase

fragment of a 70k heat-shock cognate protein. Nature

346, 623-628.

Flynn, G.C., Pohl, J., Flocco, M.T., and Rothman J.E.

( 1 9 9 1 ) . Peptide-binding specificity of the molecular

chaperone BiP. Nature 353, 726-730.

Gambill, B.D., Voos, W., Kang, P.J., Miao, B., Langer, T.,

Craig E.A., and Pfanner N. ( 1 9 9 3 ) . A dual role for

mitochondrial heat shock protein 70 in membrane

translocation of preproteins. J . C e l l B i o l . 123, 109-

117.

Georgopoulos, C., and Welch, W.J. ( 1 9 9 3 ) . Role of the

major heat shock proteins as molecular chaperones.

Annu. Rev . Ce l l B io l . 9, 601-634.

Glick, B.S. ( 1 9 9 5 ) . Can Hsp70 proteins act as force-

generating motors? Cel l 80, 11-14.

Glick, B.S., Wachter, C., Reid, G.A., and Schatz, G. ( 1 9 9 3 ) .

Import of cytochrome b2

to the mitochondrial

intermembrane space: the tightly folded heme-binding

domain makes import dependent upon matrix ATP.

Prote in Sc i . 2, 1901-1917.

Greene, L.E., Zinner, R., Naficy, S., and Eisenberg, E.

( 1 9 9 5 ) . Effect of nucleotide on the binding of peptides

to 70-kDa heat shock protein. J . B i o l . C h e m . 270,

2967-2973.

Hachiya, N., Mihara, K., Suda, K., Horst, M., Schatz, G., and

Lithgow, T. ( 1 9 9 5 ) . Reconstitution of the initial steps

of mitochondrial protein import. Nature 376, 705-709.

Hightower, L.E., Sadis, S.E., and Takenaka, I.M. ( 1 9 9 4 ) .

Interactions of vertebrate hsc70 and hsp70 with unfolded

proteins and peptides. In: The Biology of Heat Shock

Proteins and Molecular Chaperones, R. I. Morimoto, A.

Tissières and C. Georgopoulos, eds.: Cold Spring Harbor

Laboratory Press), pp. 179-207.

Horst, M., Oppliger, W., Rospert, S., Schönfeld, H.J.,

Schatz, G., and Azem, A. ( 1 9 9 7 a ) . Sequential action of


747

two hsp70 complexes during protein import into

mitochondria. EMBO J. 16, 1842-1849.

Horst, M., Azem, A., Schatz, G., and Glick, B.S. ( 1 9 9 7 b ) .

What is the driving force for protein import into

mitochondria? B i o c h e m . B i o p h y s . A c t a 1318, 71-

78.

Horst, M., Oppliger, W., Feifel, B., Schatz, G., and Glick,

B.S. ( 1 9 9 6 ) . The mitochondrial protein import motor:

dissociation of mitochondrial hsp70 from its membrane

anchor requires ATP binding rather than ATP hydrolysis.

Prote in Sc i . 5, 759-767.

Hurt, E.C., Pesold-Hurt, B., Suda, K., Oppliger, W., and

Schatz, G. ( 1 9 8 5 ) . The first twelve amino acids (less

than half of the presequence) of an imported mitochondrial

protein can direct mouse cytosolic dihydrofolate reductase

into the yeast mitochondrial matrix. EMBO J. 4, 2061-

2068.

Ikeda, E., Yoshida, S., Mitsuzawa, H., Uno, I., and Toh-e, A.

( 1 9 9 4 ) . YGE1 is a yeast homologue of Escherichia coli

grpE and is required for maintenance of mitochondrial

functions. FEBS Lett. 339, 265-268.

Kang, P.J., Ostermann, J., Shilling, J., Neupert, W., Craig,

E.A., and Pfanner, N. ( 1 9 9 0 ) . Requirement for hsp70 in

the mitochondrial matrix for translocation and folding of

precursor proteins. Nature 348, 137-143.

Komiya, T., Rospert, S., Schatz, G., and Mihara, K. ( 1 9 9 7 ) .

Binding of mitochondrial precursor proteins to the

cytoplasmic domains of the import receptors Tom70 and

Tom20 is determined by cytoplasmic chaperones. EMBO

J. 16, 4267-4275.

Kronidou, N.G., Oppliger, W., Bolliger, L., Hannavy, K.,

Glick, B.S., Schatz, G., and Horst, M. ( 1 9 9 4 ) . Dynamic

interaction between Isp45 and mitochondrial hsp70 in the

protein import system of the yeast mitochondrial inner

membrane. Proc . Nat l . Acad. Sc i . USA 91, 12818-

12822.

Laloraya, S., Gambill, B.D., and Craig, E.A. ( 1 9 9 4 ) . A role

for a eukaryotic GrpE-related protein, Mge1p, in protein

translocation. Proc. Nat l . Acad. S c i . USA 91,

6481-6485.

Landry, S.J., Jordan, R., McMacken, R., and Gierasch, L.M.

( 1 9 9 2 ) . Different conformations for the same

polypeptide bound to chaperones DnaK and GroEL.

Nature 355, 455-457.

Langer, T., and Neupert, W. ( 1 9 9 4 ) . In: The Biology of Heat

Shock Proteins and Molecular Chaperones. R.I.,

Morimoto, A., Tissieres, and C.,Georgopouolos, eds.:

Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

53-83.

Liberek, K., Marszalek, J., Ang, D., Georgopoulos, C., and

Zylicz, M. ( 1 9 9 1 ) . Escherichia coli DnaJ and GrpE heat

shock proteins jointly stimulate ATPase activity of DnaK.


Lindquist, S., Craig, E.A. ( 1 9 8 8 ) . The heat shock proteins.

Annu. Rev. Genet . 22, 631-677.

Maarse, A.C., Blom, J., Grivell L.A., and Meijer, M.

( 1 9 9 2 ) . MPI1, an essential gene encoding a

mitochondrial membrane protein, is possibly involved in

protein import into yeast mitochondria. EMBO J. 11,

3619-3628.

Manning-Krieg, U.C., Scherer, P.E., and Schatz, G. ( 1 9 9 1 ) .

Sequential action of mitochondrial chaperones in protein

import into the matrix. EMBO J. 10, 3273-3280.

Marshall, J.S., DeRocher, A.E., Keegstra, K., and Vierling, E.

( 1 9 9 0 ) . Identification of heat shock protein hsp70

homologues in chloroplasts. P r o c . N a t l . A c a d . S c i .

USA 87, 374-378.

Matouschek, A., Azem, A., Glick, B.S., Schmid, K., and

Schatz, G. ( 1 9 9 7 ) . Protein unfolding by mhsp70

suggests that it can act as an import motor. EMBO J. (in

press).

McCarty, J.S., Buchberger, A., Reinstein, J., and Bukau, B.

( 1 9 9 5 ) . The role of ATP in the functional cycle of the

DnaK chaperone system. J . M o l . B i o l . 249, 126-137.

McKay, D.B., Wilbanks, S.M., Flaherty, K.M., Ha, J.-H.,

O'Brien, M.C., and Shirvanee, L.L. ( 1 9 9 4 ) . Stress-70

proteins and their interaction with nucleotides. In: The

Biology of Heat Shock Proteins and Molecular

Chaperones, R.I. Morimoto, A. Tissières and C.

Georgopoulos, eds.: Cold Spring Harbor Laboratory

Press), pp. 153-177.

Morishima, N., Nakagawa, K., Yamamoto, E., and Shibata, T.

( 1 9 9 0 ) . A subunit of yeast site-specific endonuclease

SceI is a mitochondrial version of the 70-kDa heat shock

protein. J . B io l . Chem. 265, 15189-15197.

Nakai, M., Kato, Y., Ikeda, E., Toh-e, A., and Endo, T.

( 1 9 9 4 ) . Yge1p, a eukaryotic GrpE homolog, is localized

in the mitochondrial matrix and interacts with

mitochondrial hsp70. Biochem. B i o p h y s . R e s .

Commun. 200, 435-442.

Neupert, W., Hartl, F.-U., Craig, E.A., and Pfanner, N.

( 1 9 9 0 ) . How do polypeptides cross the mitochondrial

membranes? Cel l 63, 447-450.

Palleros, D.R., Reid, K.L., Shi, L., Welch, W.J., and Fink,

A.L. ( 1 9 9 3 ) . ATP-induced protein-Hsp70 complex

dissociation requires K+ but not ATP hydrolysis. Nature

365, 664-666.

Pelham, H.R. ( 1 9 8 6 ) . Spectulations on the functions of the

major heat shock and glucose regulated proteins. Cel l

46, 959-961.

Pfanner, N., and Meijer, M. ( 1 9 9 5 ) . Pulling in the proteins.

Curr. Biol . 5, 132-135.

Prip-Buus, C., Westermann, B., Schmitt, M., Langer, T.,

Neupert,W., and Schwarz,E. ( 1 9 9 6 ) . Role of the

mitochondrial DnaJ homologue, Mdj1p, in the prevention

of heat-induced protein aggregation. FEBS Lett . 380,

142-146.

Prasad, K., Heuser, J., Eisenberg, E., and Greene, L. ( 1 9 9 4 ) .

Complex formation between clathrin and uncoating

ATPase. J . B io l . Chem. 269, 6931-6939.


748

Rassow, J., and Pfanner, N. ( 1 9 9 6 ) . Protein biogenesis:

chaperones for nascent polypeptides. Curr. B i o l . 6,

115-118.

Rassow, J., Maarse, A.C., Krainer, E., Kübrich, M., Müller,

H., Meijer, M., Craig, E.A., and Pfanner, N. ( 1 9 9 4 ) .

Mitochondrial protein import: biochemical and genetic

evidence for interaction of matrix hsp70 and the inner

membrane protein MIM44. J . C e l l B i o l . 127, 1547-

1556.

Rassow, J., Hartl, F.-U., Guiard, B., Pfanner, N., and Neupert,

W. ( 1 9 9 0 ) . Polypeptides traverse the mitochondrial

envelope in an extended state. FEBS Lett . 275, 190-

194.

Rowley, N., Prip-Buus, C., Westermann, B., Brown, C.,

Schwarz, E., Barrell, B., and Neupert, W. ( 1 9 9 4 ) .

Mdj1p, a novel chaperone of the DnaJ family, is involved

in mitochondrial biogenesis and protein folding. Cel l

77, 249-259.

Sadler, I., Chiang, A., Kurihara, T., Rothblatt, J., Way, J.,

and Silver, P. ( 1 9 8 9 ) . A yeast gene important for

protein assembly into the endoplasmic reticulum and the

nucleus has homology to DnaJ, an Escherichia coli heat

shock protein. J . C e l l B i o l . 109, 2665-2675.

Scherer, P.E., Manning-Krieg, U.C., Jenö, P., Schatz, G.,

Horst, M. ( 1 9 9 2 ) . Identification of a 45 kDa protein at

the protein import site of the yeast mitochondrial inner

membrane. Proc . Nat l . Acad . Sc i . USA 89, 11930-

11934.

Scherer, P.E., Krieg, U.C., Hwang, S.T., Vestweber, D.,

Schatz, G. ( 1 9 9 0 ) . A precursor protein partly

translocated into yeast mitochondria is bound to a 70 kD

mitochondrial stress protein. EMBO J. 9, 4315-4322.

Schnell, D.J., Kessler, F., and Blobel, G. ( 1 9 9 4 ) . Isolation

of components of the chloroplast protein import

machinery. Sc ience 266, 1007-1012.

Schlenstedt, G., Harris, S., Risse, B., Lill, R., and Silver,

P.A. ( 1 9 9 5 ) . A yeast DnaJ homologue, Scj1p, can

function in the endoplasmic reticulum with BiP/Kar2p via

a conserved domain that specifies interactions with

Hsp70s. J . C e l l B i o l . 129, 979-988.

Schleyer, M., and Neupert, W. ( 1 9 8 5 ) . Transport of proteins

into mitochondria: translocational intermediates

spanning contact sites between outer and inner

membranes. Cel l 43, 339-350.

Schmid, D., Baici, A., Gehring, H., and Christen, P. ( 1 9 9 4 ) .

Kinetics of molecular chaperone action. S c i e n c e 263,

971-973.

Schneider, H.C., Berthold, J., Bauer, M.F., Dietmeier, K.,

Guiard, B., Brunner, M., and Neupert, W. ( 1 9 9 4 ) .

Mitochondrial Hsp70/MIM44 complex facilitates protein

import. Nature 371, 768-774.

Scidmore, M.A., Okamura, H.H., and Rose, M.D. ( 1 9 9 3 ) .

Genetic interactions between KAR2 and SEC63 , encoding

eukaryotic homologues of DnaK and DnaJ in the

endoplasmic reticulum. M o l . B i o l . Ce l l 4, 1145-

1159.

Simon, S.M., Peskin, C.S., and Oster, G.F. ( 1 9 9 2 ) . What

drives the translocation of proteins? Proc . Nat l . Acad.

Sci . USA 89, 3770-3774.

Spudich, J.A. ( 1 9 9 4 ) . How molecular motors work. Nature

372, 515-518.

Stuart, R.A., Gruhler, A., van der Klei, I., Guiard, B., Koll, H.,

and Neupert, W. ( 1 9 9 4 ) . The requirement for matrix ATP

for the import of precursor proteins into the

mitochondrial matrix and intermembrane space. Eur. J .

B iochem. 220, 9-18.

Szabo, A., Langer, T., Schröder, H., Flanagan, J., Bukau, B.,

and Hartl, F.-U. ( 1 9 9 4 ) . The ATP hydrolysis-dependent

reaction cycle of the Escherichia coli Hsp70 system

DnaK, DnaJ, and GrpE. Proc . Nat l . Acad . Sc i . USA

91, 10345-10349.

Ungermann, C., Neupert, W., and Cyr, D.M. ( 1 9 9 4 ) . The

role of Hsp70 in conferring unidirectionality on protein

translocation into mitochondria. Science 266, 1250-

1253.

Ungermann, C., Guiard, B., Neupert, W., and Cyr, D.M.

( 1 9 9 6 ) . The !/"- and Hsp70/MIM44-dependent reaction

cycle driving early steps of protein import into

mitochondria. EMBO J. 15, 735-744.

Viitanen, P.V., Donaldson, G.K., Lorimer, G.H., Lubben,

T.H., and Gatenby, A.A. ( 1 9 9 1 ) . Complex interactions

between the chaperonin 60 molecular chaperone and

dihydrofolate reductase. Biochemis try 30, 9716-9723.

von Ahsen, O., Voos, W., Henninger, H., and Pfanner, N.

( 1 9 9 6 ) . The mitochondrial protein import machinery.

Role of ATP on dissociation of the HSP70.MIM44

complex J . B io l . Chem. 270, 29848-29853.

Voos, W., Gambill, B.D., Guiard, B., Pfanner, N,. and Craig,

E.A. ( 1 9 9 3 ) . Presequence and mature part of preproteins

strongly influence the dependence of mitochondrial

protein import on heat shock protein 70 in the matrix. J .

C e l l B i o l . 123, 119-126.

Walker R.A., and Sheetz M.P. ( 1 9 9 3 ) . Cytoplasmic

microtubule-associated motors. Annu. R e v .

B iochem. 62, 429-451.

Westermann, B., Prip-Buus, C., Neupert, W., and Schwarz, E.

( 1 9 9 5 ) . The role of the GrpE homologue, Mge1p, in

mediating protein import and protein folding in

mitochondria. EMBO J. 14 3452-3460.

Zhu, X., Zhao, X., Burkholder, W.F., Gragerov, A., Ogata,

C.M., Gottesman, M.E., and Hendrickson, W.A. ( 1 9 9 6 ) .

Structural analysis of substrate binding by the molecular

chaperone DnaK. Science 272, 1606-1614.

Gene Therapy & Molecular Biology Volume 1 Issue A - Part 2

Documents

nuclear structure

nuclear dna

nuclear matrixproteins

toisolate nuclear matrices

nuclear porelamina

whichother nuclear components

internal matrix of nm1

thenuclear matrix protein