Transcriptional Control of the Factor IX of Five Cis-Acting

Transcriptional Control of the Factor IX Gene: Analysis of Five Cis-Acting Elements and the Deleterious Effects of Naturally Occurring

Hemophilia B Leyden Mutations

By David J. Picketts, Christopher R. Mueller, and David Lillicrap

Hemophilia B Leyden is a rare form of inherited factor IX deficiency in which patients experience spontaneous post- pubertal recovery of factor IX levels. The mutations resulting in this disorder are localized in a 40-nucleotide region en- compassing the major transcriptional start site for factor IX. Here we report the further characterization of five cis-acting elements in the factor IX promoter and the effects on protein binding and transcriptional activation of five Leyden muta- tions (at nucleotides +13, -5, -6, -20, and -26) that occur within the proximal three elements (sites 1 through 3). Band- shift studies using nuclear extracts from four different rat tissues have shown that at least some of the proteins bind- ing to each of the five sites are ubiquitous in nature. The pattern of DNA binding at site 1 suggests that this element plays an important role in mediating the liver-specific ex- pression of factor IX. Additional studies with liver nuclear extracts obtained at several different points in development have shown an increase in DNA binding at sites 1, 4, and 5 between 1 day and 1 week. Using DNase I footprint analysis and competition bandshift studies, we have shown that the binding of nuclear proteins t o each of the mutant sites is

ACTOR IX IS a vitamin K-dependent glycoprotein that is involved in the middle phase of the intrinsic pathway

of coagulation. Inherited absence or dysfunction of factor IX results in the X-linked recessive bleeding disorder hemo- philia B.’ The molecular pathology of hemophilia B is well documented.’ Approximately 95% of the documented factor IX mutations are point mutations and short nucleotide addi- tions and deletions. In addition to the many heterogeneous mutations found throughout the protein coding region, there have been several point mutations described within a 40- nucleotide region encompassing the major transcriptional start site of the factor IX p r ~ m o t e r . ” ~ These mutations give rise to a similar phenotype known as hemophilia B Leyden.

Hemophilia B Leyden is characterized by severe or mod- erately severe factor IX deficiency before puberty.“’ After puberty, factor IX activity and antigen levels gradually in- crease, resulting in resolution of the clinical bleeding disor- der in adulthood as levels rise to greater than 0.30 U/mL

F

From the Departments of Pathology and Biochemistry, Queen’s

Submitted March 5, 1994; accepted July 7, 1994. Supported by grants from the Canadian MRC (to D.P.L. and

C.R.M.). D.J.P. and D.P.L. are recipients of Research Fellowship and Career Scientist Awards, respectively, from the Ontario Ministry of Health and C A M . is a Research Scientist of the National Cancer Institute of Canada.

Address reprint requests to David Lillicrap, MD, FRCPC, Deparf- ment of Pathology, Richardson Laboratory, Queen’s University, Kingston, Ontario, Canada K7L 3N6.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with I8 U.S.C. section 1734 solely to indicate this fact.

University, Kingston, Ontario, Canada.

0 1994 by The American Society of Hematology. 0006-4971/!?4/8409-00$3.00/0

2992

disrupted to a variable extent. There appears to be some, although reduced, protein binding to all of the mutant oligo- nucleotides apart from the -26 mutant. In vitro transcription assays have shown that each of the mutations reduces the global proximal promoter activity by -40%. Two double mutant promoters did not show any additional downregula- tion in the in vitro transcription assay. In experiments de- signed to assess the relative transcriptional activity medi- ated from each of the five sites independently, we have tested artificial homopolymer promoters of each site in the in vitro transcription assay. These studies show that sites 4 and 5 are the strongest activators and that transactivation from site 5 is further enhanced by the albumin D site-binding protein. In summary, these investigations show deleterious effects of each of the Leyden mutations tested on the bind- ing of trans-acting factors and also show disruption of tran- scriptional activation in a functional in vitro transcription assay. Our results also show that cis-acting elements 4 and 5 are the principal activators of this locus. 0 1994 by The American Society of Hematology.

and in some instances, into the lower range of normal (nor- mal, 0.5 to 1.5 U/mL). In contrast with the postpubertal recovery seen in all other cases of this disorder, there is one report of a patient with a mutation at nucleotide -26 who did not recover factor IX activity after puberty.’’

The factor IX proximal promoter is comprised of five cis-acting DNA elements.’’ It lacks an identifiable TATA sequence and has a reverse CCAAT box sequence between nucleotides -92 and -96. Sites 1 and 5 have been shown to bind the liver-enriched transcription factor CCAAT/en- hancer binding protein (Y (CEBPa), whereas site 5 binds an additional unidentified protein.’* Site 4 is a consensus nuclear factor 1 (NF 1) site, and site 3 contains both a consensus androgen response element and a consensus sequence for the liver-enriched factor, hepatocyte nuclear factor 4 (HNF- 4). All of the currently identified Leyden mutations fall within sites I , 2, and 3, which are the three proximal sites in the promoter. Recent work has focused on defining the mechanism responsible for the recovery of factor IX activity after the onset of puberty. Mutations at both + 13 and -6 result in decreased transcriptional activity when a C E B P a expression plasmid is cotransfected in transient transfection experiments.‘j More recently, two reports have suggested that the -20 mutation affects HNF-4 binding and transactivation, but not the activity of the androgen receptor; the -26 muta- tion appears to affect binding and, thus, transactivation by both proteins.”,” The loss of androgen receptor binding has been postulated as a mechanism for the lack of recovery of the -26 mutation after puberty. We have recently shown that synergy between C/EBPa and the albumin D site-bind- ing protein (DBP) results in increased transcriptional activity that compensated for a mutation at the -5 position.”

We now report a comprehensive analysis of the five cis- acting elements in the factor IX promoter and describe the results of experiments with five Leyden mutant promoters.

Blood, Vol 84, No 9 (November l ) , 1994: pp 2992-3000

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

FACTOR IX GENE TRANSCRIPTION 2993

In studies using nuclear protein extracts from a variety of tissues, we have documented the tissue-specific binding of proteins to all five cis-acting sequences. In addition, using liver nuclear extracts, we have examined the developmental changes in the pattern of gel-retarded complexes at each of the five sites. In studies of the mutant promoters, we have shown decreased binding of trans-acting factors to the mu- tant cis-acting sequences and have documented, in a quanti- fiable in vitro transcription assay, that all the mutations result in decreased transcriptional activity. We have also shown that sites 4 and 5 are the strongest activators at this locus and that transactivation from site 5 can be further enhanced by the transcription factor DBP.

MATERIALS AND METHODS

Plasmid constructs. The factor IX wild type (WT) and -5 mu- tant pUC plasmids were described previously.’* They encompass nucleotides -285 to +21 of the factor IX promoter. For in vitro transcription reactions, the WT- and -5-pUC plasmids were digested with HindIII and Sac I and purified on low melting-point agarose. These fragments were then cloned into the G-freenATA vector derived from the Alu-7 plasmid described in Gorski et a1.I4 To prepare this vector, the sequence containing the -35 to +22 region of the albumin gene and the G-free cassette was amplified by poly- merase chain reaction (PCR) and cloned into the Sac I and EcoRI sites of the pBS’ vector (Stratagene, La Jolla, CA). This produces a vector with the pBS+ multilinker from HindIII to Sac I, upstream of the albumin TATA sequence and the G-free cassette. Inclusion of the albumin TATA sequence in this construct was required after initial studies of the native TATA-less factor IX promoter showed products derived from multiple transcriptional start sites. The WT G-free construct was used as a template to construct the other Leyden mutants by site-specific PCR mutagenesis using the method of Jones and Howardi5 with the modifications described below. The mutant oligonucleotides used were: F9 A+ 13G: 5’-ACCACTTTCA- CAGTCTGC; AAAGTGTCAGACGTGGTG-5’; F9 A-5T: 5”GGT- ACAACTAATCGTCCIIT, TTGATTAGCAGGAACCATG-5’; F9 G-6C: 5”GGTACAACTAATCCACCTTA; TTGATTAGGTGG- AAT-5‘; F9 T-20A: 5”AGCTCAGCTTGTACTTAGGT; TCGA- ACATGAATCCA-5’; F9 G-26C: 5”AGCTCAGCTTCTACTT- TGGT; and TCGAAGATGAAACCA-5’.

Fragments spanning the multilinker cloning site were amplified using the mutant primer and the universal reverse-sequencing primer in one reaction and the complementary mutant primer and a primer located at the 3’ end of the G-free cassette in the second reaction. Amplifications were in I X Vent DNA polymerase buffer (New En- gland Biolabs, Beverly, MA) with 200 pmoVL each deoxynucleotide triphosphate (dNTP), 50 pmol of each primer, 42 ng template DNA, and 2 U Vent DNA polymerase. The reaction mix was denatured for 2 minutes at 99°C before 25 amplification cycles (denatured at 99°C for 30 seconds, annealed at 55°C for 30 seconds, and extended at 72°C for 1 minute). One microliter from each reaction was then added to a final PCR reaction mix containing 50 pmol of the reverse sequencing primer and the 3‘ G-free primer. The full-length product was amplified by an initial five-cycle reaction (97°C for 30 seconds, 37°C for 30 seconds, and 72°C for 1 minute) followed by the 25-cycle reaction described above. The PCR products were then digested with EcoRI and HindIII, purified on low-gelling-temperature agarose and cloned into PBS’ (Stratagene). The G-freeRATA clones were then digested with HindIII and Sac I and cloned into PBS+.

Nuclear extract preparations. Nuclear extracts were prepared from adult-rat tissues (liver, spleen, kidney, and brain) at a concen- tration of 5 to 10 pgIpL, according to Gorski et al (1986),j4 with

the modifications described in Maire et a1 (1989).16 Additional liver nuclear extracts were prepared from male rats at l day, 1 week, 2 weeks, and 1 month of life.

In vitro transcription. The in vitro transcription assays were performed as described in Gorski et al (1986).14 The individual factor IX promoter sites were ligated unidirectionally to obtain oligomers of 11 copies for site 2, 9 copies for sites 1 and 4, and 5 copies for sites 3 and 5. The fragments were filled in with Klenow and cloned in front of the G-freeRATA cassette at the BamHI and Sal I sites in the forward direction. The oligonucleotides used to make the homopolymer constructs were described previously.’2

DNase Iprotection assays. Probes for DNase I protection assays were prepared as described previously.lz The footprint assays were performed as described in Lichtsteiner et al (1987)” except for the following modifications. The binding reaction was incubated on ice for 15 minutes before DNase I digestion. DNase I digestion was for 2 minutes with the probe alone. The DNase I digestion was stopped by the addition of 200 pL of a solution containing 100 mmoVL TRIS (pH 7.5). 10 mmoVL EDTA, 100 mmol/L NaC1, and 0.1% sodium dodecyl sulfate. Proteinase K (100 pg) and yeast tRNA (10 pp) were added to the samples and digested for 30 minutes at 37°C.

Gel mobility shz$ and supershift assays. Gel mobility shift assays were performed as described in Lichtsteiner et al (1987)” except that the binding reactions were incubated for 15 minutes on ice before the samples were loaded on a 6% nondenaturing poly- acrylamide gel. One microgram of nuclear extract was used for each binding reaction. In the competition assays, the unlabeled competitor oligonucleotide was added to the binding reaction at the same time as the labeled oligonucleotide probe for each site. In the site-l su- pershift experiment, 1 pL of each of the four antitranscription factor antisera was added I O minutes after the addition of nuclear extract and the incubation was continued for 10 minutes on ice (CEBPa antibody was a gift from Dr Steve McKnight [Howard Hughes Re- search Laboratories, Carnegie Institution of Washington, Baltimore]; CEBPP and C/EBP6 antibodies were purchased from Santa Cruz Biotechnology Inc [Santa Cruz, CA]; and the DBP antibody was a gift from Dr Ueli Schibler, [University of Geneva, Geneva, Switzer- land]).

RESULTS

The occurrence of natural mutations in the factor IX pro- moter, which result in a similar developmentally regulated phenotype, provides a unique opportunity to study the func- tion of this clinically significant promoter. We now present new information relating to the five cis-acting elements in the promoter by examining the tissue-specific and develop- mental regulation of the factors interacting with these sites and have assessed the effect of five different Leyden muta- tions on the binding of these proteins and on in vitro tran- scription.

Tissue and developmental bandshifi assays. Previous re- ports assessing the pattern of trans-acting proteins binding to the factor IX promoter have used a variety of purified and recombinant protein preparations as well as liver nuclear extracts. In this study, we have used bandshift assays em- ploying nuclear extracts from four different tissues (liver, spleen, kidney, and brain) and a human hepatoma cell line (HepG2) to assess the tissue specificity of the factors binding to all five of the cis-acting sites (Fig 1A). A very heteroge- neous complex is formed between site l and factors present in the liver, with no discernible difference between extracts prepared in the morning and at night. Several more distinct

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

2994 PICKETS, MUELLER, AND LlLLlCRAP

B Site 1 Site 2 Site 3 Site 4 Site 5

( I d l w 2w l m Ad I Id l w 2w lm Ad I Id l w 2w l m Ad I I d l w 2w l m Ad I Id l w 2w Im Ad I I I I I I I I I I I I I I I I I I I I I I I I I I 7 l

C

Supershift

Bound

2 3 U U1

ANTI-SERA TO cn 0 z a S

C/EBP DBP

Fig 1. (A) Results of bandshift assays using radiolabeled oligonucleotide probes corresponding t o sites 1 through 5 and nuclear extracts (1 pglreaction) prepared from various adult-rat tissues and a human hepatoma cell line. Lane designations are B, brain; K, kidney; S, spleen; HG. HepG2 cells; LA, liver (morning sample); LP, liver (evening sample). Gel-retarded complexes can be seen with each of the five sites and all tissues and the cell line studied. (B) Results of a developmental bandshift assay using male rat liver nuclear extracts (1 pglreactionl obtained at 1 day, 1 week, 2 weeks, 1 month, and adulthood. Apart from the increase in complex formation from l day to 1 week seen at sites 1,4, and 5, similar gel-retarded complexes are seen at all developmental time points with radiolabeled probes corresponding to all five of the promoter elements. (C) An antibody supershift analysis of site 1, with rat-liver nuclear extracts. The left-hand lane shows the pattern of gel- retarded protein-DNA complexes without the addition of antisera, and the four subsequent lanes show the pattern of bound and supershifted complexes after incubation with antibodies t o CIEBPrr, ClEBPp, ClEBP6, and DBP. A densitometric quantification of the bound and supershifted complexes shows that the relative contributions to binding at site 1 are ClEBPcr - 404b. CIEBPP - 20%. and DBP - 5%.

complexes are formed with factors present in other tissues "smear" seen with the liver extracts at site 1 is typical of and in the dedifferentiated HepG2 extracts. In each case, the a site that interacts with members of the CEBP family and complexes are significantly less abundant than the factors in represents binding by a heterogeneous mixture of homodi- the liver (fivefold difference). The pattern of a DNA-protein mers and heterodimers of these factors. This is consistent

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

FACTOR IX GENE TRANSCRIPTION 2995

with the previous characterization of this site as a high- affinity CIEBPa site:'' a finding that is further substantiated by the results of a supershift assay for CEBPa at this site (Fig 1C). This analysis also indicates that site 1 binds C/ EBPp and DBP to a lesser extent.

Sites 2 and 3 are characterized by what appears to be a discrete complex present in all of the extracts. Further analy- sis of these complexes by fractionation of the proteins on heparin agarose indicates that the site 2 protein is likely a single complex that elutes at 400 mmol/L NaCl. In contrast, the factors binding to site 3 can be separated into two distinct fractions eluting at 200 and 500 mmol/L NaCl. This observa- tion is consistent with the fact that site 3 has previously been characterized as binding both HNF-4 and the androgen receptor.''.'3

Site 4 is known to bind NF 1, which is present in all tissues, and the more abundant complex formed in the liver is caused by the presence of more liver-specific forms of this factor.I7 Experiments with recombinant CIEBPa indicate that this factor binds to this site to a limited degree." Site 5, in our previous studies, was shown to bind an unknown factor in addition to CIEBPa. This unknown factor, which forms a discrete complex, appears to be present in all of the other tissues examined, suggesting that it may be a ubiqui- tous protein.

Examining the developmental regulation of all of these sites in the liver indicates that a significant induction of DNA binding activity (threefold increase) occurs at sites l, 4, and 5 between day 1 and 1 to 2 weeks of age (Fig 1B). This is the period when terminal differentiation occurs and is associated with the induction of factors such as CIEBPP, which binds to site 1. A second complex with site 3 was also observed from 1 to 2 weeks of age. This is before puberty in the rat and, thus, is unlikely to represent induction of the androgen receptor. No specific changes are observed between 1 month of age (pen-pubertal) and the adult, which corresponds to the period in humans when phenotypic recov- ery occurs. However, given the difficulty of resolving multi- ple complexes at a given site, this finding is not unexpected.

Effect of Leyden mutations on protein binding. We re- cently reported that an A to T transversion at nucleotide -58 greatly decreased the affinity for protein binding to site 2 of the factor D( promoter in DNase I protection assays.'* To determine if other Leyden mutations at each of the three proximal sites disrupted protein binding to the corresponding mutant sites and to assess the possible disruption of protein binding to adjacent sites, we synthesized promoter templates corresponding to mutations at + l 3 (A to G; site 1) , -6 (G to C; site 2), -20 (T to A; site 3), and -26 (G to C; site 3). Each of these templates was studied in DNase I footprint assays using rat-liver nuclear extracts. All of the mutations, with the exception of the substitution at + 13, resulted in the loss of DNase I protection at the site in which the mutation was located (Fig 2). The + l 3 mutation located in site 1 has previously been shown to disrupt the binding of CIEBPCX.~ Repeating the footprints of the + l 3 mutant template with both heat-treated extracts, which preferentially retain C/ EBPa activity, and with recombinant CEBPa showed that CEBPa could not occupy site 1 when the mutation was

present (data not shown). The footprint obtained with non- heat-treated extracts is presumably the result of other factors such as CIEBPP binding to this site.

Oligonucleotides for sites 1,2 ,3 , and each of the Leyden mutant constructs were radiolabeled and examined in gel mobility shift assays for protein binding (Fig 3, A through C). The +l3 and -5 mutant probes showed slightly reduced complex formation compared with the WT probes (70% and 86% binding, respectively). In contrast, the -6, -20, and -26 mutant oligonucleotides showed very poor complex formation (5%, 3 8 , and 0%, respectively). To assess the relative affinities of the mutant sites, protein binding to radio- labeled WT oligonucleotides was competed with either WT or mutant unlabeled oligonucleotides at 50 or 100 times the WT probe concentration. The + l 3 oligonucleotide showed poor competition, with a discrete retarded complex re- maining even with a 100-fold excess of the unlabeled mutant competitor. This presumably represents the preferential com- petition of all of the other site-l binding factors with the exception of CEBPcu, which forms the remaining complex. The -5 oligonucleotide competed poorly for the site-2 com- plex and the -6 oligonucleotide showed no competition, indicating that the -6 mutation has a more profound effect on binding of the currently unknown protein that interacts with site 2. Studies with the -20 and -26 oligonucleotides showed markedly reduced competition compared with the WT site-3 probe. Two other groups have shown that oligonu- cleotides containing the -20 or -26 nucleotide change com- pete weakly (the -20 mutation) or not at all (the -26 muta- tion) for HNF-4 binding to this site.11313

In vitro transcription analysis of full-length mutant pro- moters. Having documented the disruption of protein bind- ing at each of the mutant promoter sites in footprint and bandshift experiments, we have gone on to perform studies to assess the functional consequences of this effect. An in vitro transcription assay was used" in which the WT and mutant promoters were cloned upstream of a G-free vector that contained the -35 to +22 region of the albumin gene including the TATA box (G-Free/TATA cas~ette).'~ As de- scribed above, the inclusion of the TATA sequence in this construct was required to produce a single discrete transcript that could be readily quantified. The activity of the different mutants was compared with the transcriptional activity of a G-free cassette under the control of the adenovirus major late promoter. Although this assay may have limitations in exactly reflecting the magnitude of the in vivo effect of these mutations, it is an excellent in vitro method for comparing the relative effect on transactivation of each of the five pro- moter mutations. Figure 4 shows the results of a single exper- iment in panel A and the mean values obtained from five experiments in panel B. All of the mutations resulted in a decrease in transcriptional activity to 60% or less of the WT. An additional examination of the activity of two double mutants (+l3 and -5 [sites 1 and 21, and -5 and -20 [sites 2 and 31) that retain the function of sites 3 and 1, respectively, showed that they were not significantly more debilitated than the lowest of the two single mutations in each combination (Fig 4B).

The extent of transcriptional disruption resulting from the

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

PICKETS, MUELLER, AND LILLICRAP 2996

5 1

2

1

WT -5 -6 -20 -26 + 13 0 S 1 0 3 0 0 S l0300 S 10300 S 1 0 3 0 0 S 1 0 3 0 0 S 1 0 3 0

i

I

.

Leyden mutations in this assay, compared with their in vivo effects, suggests that other regulatory elements that are not being tested in this system may also play a role in generating the Leyden phenotype.

In vitro transcription analysis cfhomopolymeric elements. To assess the independent transactivation potential of each of the five cis-acting sequences in the proximal promoter, oligonucleotides corresponding to each of the five sites were concdtamerized and cloned into the G-freeRATA cassette. This strategy has been shown previously to be an effective method to characterize the role of individual cis-acting ele- ments in a complex

The number of monomeric sites contained in each of the constructs ranged from 5 to 1 1 . In the case of the albumin promoter, previous studies have indicated that no significant difference in transcriptional activation is evident in homo- polymers containing between S and 12 copies of the mono- meric site (unpublished data). The homopolymeric con- structs were examined for promoter activity, standardized with the adenovirus major late-promoter activity and com- pared with the activity of the G-FreeRATA plasmid that did not contain an insert (Fig 5 , A and B). Sites 4 and 5 were the strongest transactivators with site I also showing some activity. The liver specificity of site-S-mediated transactiva-

Fig 2. DNase I footprint analysis of the W and five mutant promoter constructs. Increasing amounts of protein have been incubated with each promoter template (0 to 30 pg from left to right). The position of the five DNase l-protected sequences is shown at the left of the figure. Binding of rat-liver nuclear proteins is shown to be variably disrupted at the sites of the mutations. In the -5, -6, and -20 mutant studies, obvious disruption of the footprints at sites 2 and 3, respectively, is documented. In the -26 mutant study, the footprint disruption is subtle and is best observed in the 5-pg nuclear-extract lane. The picture of the +l3 mutant study shows persis- tence of the footprint in the first half of this site (the distal part of site 1 in the +l3 construct has run off the bottom of the gel because of the smaller size of this construct; it is 5 nucleotides shorter than the other templates). The continued protection at this site is presumably caused by proteins other than C/ EBPtr that bind site 1 (eg, CIEBPPI.

tion is also supported by the observation that an in vitro transcription assay performed with kidney nuclear extracts shows only - 18% of the activity seen with liver extracts (data not shown). In this assay, sites 2 and 3 do not appear to mediate independent transcriptional activation. These results suggest that transactivation of the factor IX promoter is mainly caused by the binding of factors to sites 4 and S , and that the three proximal sites may play a different role in mediating transcription from this locus.

Finally, in light of our previous observation concerning a transcriptional recovery mechanism involving the peripuber- tal onset of expression of DBP, we studied the effect of this trans-acting factor on each of the five sites in the in vitro transcription assay. Recombinant DBP was added to the in vitro transcription reactions of the homopolymer promoter sites and transcriptional activation was quantified with refer- ence to the adenovirus control transcript (Fig 5 , A and B). Only site S showed an increase in transcriptional activity with the addition of DBP.

DISCUSSION

In this study we have used a quantifiable in vitro transcrip- tion assay to examine the effects of various Leyden muta- tions on binding of cognate factors and on the efficiency of

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

FACTOR IX GENE TRANSCRIPTION 2997

Fig 3. Competition bandshift assays using oligonucleotides from sites 1 through 3 of the fac- tor IX promoter and rat-liver nu- clear extracts. Gel-retarded com- plexes are indicated by an arrow to the left of the figure in each instance. (A) Lanes 1 and 2 show the binding of rat-liver nuclear protein to WT and + l 3 mutant probes for site 1. Lanes 3 through 6 show the results of competition studies using a ra- diolabeled WT probe and molar excesses, as indicated, of unla- beled WT and + l 3 mutant com- petitor oligonucleotides. (B1 Lanes 1 through 3 show binding of radiolabeled WT, -5, and -6 oligonucleotides from site 2 of the factor IX promoter. Competi- tion bandshifts are shown in lanes 4 through 9 with a radio- labeled WT probe and molar excesses, as indicated, of unla- beled WT. -5, and -6 oligonu- cleotides. (Cl Lanes 1 through 3 show bandshift assays with a WT site-3 probe and -20 and -26 mutant probes, respec- tively. Lanes 4 through 9 show competition bandshifts using a radiolabeled WT probe and mo- lar excesses, as indicated, of WT. -20, and -26 mutant oligonu- cleotides.

L

FACIDR M PROMOTER SITE 1

FAC'IUK IX PROMOTER SITE 3

transcription of this gene. This assay uses rat-liver nuclear extracts as a source of trans-acting factors to regulate tran- scription, and as such, may represent a more physiologically relevant test of transcriptional activation than other models based on cultured cells that may be deficient in critical trans- acting factors. First, the basis for the liver-specific expression of factor IX was examined using bandshift assays with nu- clear extracts from four different tissues in conjunction with oligonucleotides corresponding to all of the factor IX ele- ments. Only site 1 exhibited significantly more protein bind- ing to it in the liver extracts. This is in keeping with the

previous identification of site I as a high-affinity CEBPa binding site, a finding that we have confirmed in a supershift assay in this study. Site S has also been identified as a C/ EBPa binding site" but in this study, the tissue-specific differences in C/EBP binding are masked by the presence of a non-tissue-specific protein whose identity is unknown. Thus, from the results of this study and our previous report, the primary tissue-specific elements for this promoter appear to be sites I and S, with C/EBPa being the main tissue- specific regulator of this gene through these sites. Consistent with this observation is the strong transcriptional activity of

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

PICKETS, MUELLER, AND LILLICRAP 2998

A

B

+ l 3 -5 W+ 13 -5 -6 -20 -26 -5 -20 - Alb

Experimental

Internal Control

Transcriptional Activity of the Leyden Promoters

I I Average Activity I Standard Deviation 1 I ( n = 5) I

Wild Type I 1 0 0

n- 1

I 1 7 I 10 1 9.4

-5 I 64 1 2.1

.a 49 8.0

-20

5.4 31 -F, -90

9.0 38 + 13,-5

8.5 49 -26

7.0 50

Fig 4. (A) Autoradiogram showing the results of a single in vitro transcription assay. The studies were performed with a WT factor IX promoter (W), five mutant promoters (+13, -5, -6, -20, and -26). two double mutant promoters (+13/-5 and -5/-20), a promoterless G-free cassette, and the albumin promoter using rat-liver nuclear extracts. In each case, the experimental transcript was quantified after normalization with the corresponding transcript under the con- trol of the AdML promoter. (B) Tabulated results of five in vitro tran- scription assays using the WT and various Leyden mutant promoters. The average transcriptional activity and standard deviation are de- tailed for each construct.

site S and, to a lesser extent, site I . Site 4, which binds a liver-specific form of NF 1 as well as having a low affinity for C/EBPa, also shows some activator function. Thus, it appears that this site also contributes to a modest degree to the overall tissue-specific activity of the factor IX promoter.

Levels of circulating factor IX show a changing develop- mental profile in normal subjects. Normal childhood levels of factor IX arc about 7S% of adult levels.” These levels increase by =2S% into adulthood and continue to increase by = I % per year throughout life.”.” C/EBPa, which ap- pears to be the prime regulator of this gene, is expressed in the fetal liver and its level appears to stay relatively constant after this point.”’ This likely establishes the basal transcrip- tional activity of the factor IX gene. The analysis of develop- mental liver nuclear extracts has shown a significant increase in DNA binding activity between 1 day and 1 week, at sites I , 4, and 5. Most of the increase at site 1 is caused by the induction of CEBPP, which occurs at this time and is known to bind to this site. Some of the increase in site 5 activity is also caused by C/EBPP, but in addition, appears to involve the induction of the unknown ubiquitous factor that also binds to this site. These changes may be responsible for the gradual increase in levels of factor IX that occur in normal

children. However, we have observed no apparent changes in the pattern of DNA binding activity at any of the five sites between the peripubertal period and adulthood, corre- sponding to the time when phenotypic recovery occurs in hemophilia B Leyden.

Our previous report of a potential mechanism for the spon- taneous postpubertal improvement of the Leyden phenotype, suggests that the liver-enriched transcription factor, DBP, plays an important role in mediating this recovery.” In the rat, the onset of DBP expression is in the peripubertal period, and DBP levels are higher in older rats.I9 DBP appears to mediate its effect through an enhancement of binding of C/ EBPa to the two distal promoter elements, with an especially marked effect at site 5.’’ This effect may be masked in the bandshift assays by the presence of other factors; however, the addition of DBP to the in vitro transcription assay clearly results in significant transcriptional enhancement solely from site 5. In the context of the full-length promoter, we can

A +m

1 2 3 4 5 - 4 2 3 4 s -

B Transcriptional Activity of the Factor IX

Proximal Promoter Skes

Average Activity

GF construct

Site 1 (9 copies)

Site 2 (1 1 copies)

Fdd Increase + DBP In =31

Site 3 (5 copies)

site 4 (9 copies) I 6.82 1 .o

1.19 1 1.9

Site 5 (5 copies) 3.1 6.08

Fig 5. (A) An in vimo transcription assay with rat-liver nuclear extracts using homopolymers of the five cisacting sequences in the factor IX promoter and a promoterless G-free cassette (-1. The results on the right of the autoradiogram have been performed with recom- binant DBP added t o the liver nuclear extracts. In each instance, the experimental transcript has been quantified in relation t o the corresponding control transcript directed by the AdML promoter. (B) Results of five in vitro transcription assays using the homopolymer- ized sites 1 through 5 of the factor IX promoter with In = 3) and without (n = 6 ) added DBP.

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

FACTOR IX GENE TRANSCRIPTION 2999

Site Site Site Site Site

5 4 3 2 1

C/EBP a CiEBP a C/EBP R C/EBP a AR ClEBP R DBP NF 1 HNF-4 ? DBP

5' U 3' Leyden Mutations

Fig 6. Schematic diagram of the factor IX promoter showing the five cis-acting sites examined in this study and indicating the trans- acting factors, documented either in this study or in other re- pOrts6.9.11.12.13 t o bind to these sites.

postulate that the increased activity mediated from site 5 would likely lead to the general activation of transcription. This observation further supports the importance of this site in the recovery of the mutant Leyden promoters and may also play a role in the continued increase in factor IX levels throughout adulthood seen in normal subjects.

We have shown that all five of the Leyden mutations that we have studied result in decreased protein binding by both DNase I protection assays and competition bandshift assays. Using a quantifiable in vitro transcription assay we have shown that these same mutations decrease the efficiency of transcription of this gene by at least 40%. The ability of the + l 3 mutation to interfere with transcription, despite allowing other factors to still bind to site 1, indicates that C/EBPa is the primary effector for this site. Both the -5 and -6 mutations have similar effects on in vitro transcrip- tion despite the -6 mutation being much more debilitated in its binding of the site-2 factor. The situation with the mutations in site 3 is more complicated. As expected, the -20 mutation still binds some protein, presumably the an- drogen receptor, whereas the -26 mutation looses all bind- ing activity. However, despite their differential binding capa- bilities, they are equally defective in promoter activity. Double mutants, which incorporate two different mutations, are no more defective than the single mutants.

In light of the peripubertal timing of recovery of the Leyden phenotype, the role of androgens in this disorder has been explored by several previous studies.'1,2s,26 The results of our experiments do not exclude an effect of the androgen receptor, which may further increase transcription from the mutant promoters. However, we propose that this effect may not be through a direct enhancement of transactivation, but rather by increasing protein binding to site 3, thereby im- proving the stability of the transcriptional initiation complex. A similar role has been proposed for the progesterone recep- tor, for which experimental evidence has been obtained showing that binding to a progesterone response element facilitates the formation of a stable preinitiation complex, thus augmenting transcriptional i n i t i a t i ~ n . ~ ~ In further sup- port of this proposal, site 3 was not an activator of transcrip- tion with nuclear extracts in our homopolymer in vitro tran- scription assay. Factor IX is a TATA-less and details of the cis-acting elements that optimize the assembly and stability of the general transcriptional apparatus in this

situation are lacking. A consensus initiator element'" does not appear to be present, and one possibility is that elements 1 through 3 act together as a site of assembly of the preinitia- tion complex. Understanding the role that androgens play in controlling the expression of factor IX is further complicated by the fact that no one has documented strong transactivation from the native promoter after cotransfection of an androgen- receptor expression plasmid," and finally, if the androgen receptor had a strong effect on transcription of the factor IX gene, one might expect to see lower levels of factor IX in normal prepubertal males and differences in factor IX activ- ity between males and females, neither of which is the case."

In conclusion, our analysis of the independent contribu- tions of each of the five cis-acting elements in the factor 1X promoter shows that the two distal elements, sites 4 and 5 , are the strongest mediators of transcriptional activation, whereas the three proximal elements, within which the Leyden mutations have been described, appear to act by another mechanism. Postpubertal recovery of the Leyden phenotype may be mediated by the combination of increased activation from site 5 because of the onset of DBP expression and binding at site 3 of the androgen receptor (Fig 6). Our data from the homopolymer in vitro transcription assays sug- gests that the role of androgen receptor binding at this site may not be to directly transactivate this locus.

REFERENCES

I . Thompson AR: Structure, function, and molecular defects of factor IX. Blood 67565, 1986

2. Giannelli F, Green PM, High KA, Sommer S, Lillicrap DP, Ludwig M, Olek K, Reitsma PH, Goossens M, Yoshioka A, Brownlee GG: Haemophilia B: Databese of point mutations and short additions and deletions--Third edition, 1992. Nucleic Acids Res 20:2027, 1992 (suppl)

3. Reitsma PH, Bertine RM, Ploos van Amstel JK, Riemans A, Briet E: The putative factor IX gene promotor in hemophilia B Leyden. Blood 72:1074, 1988

4. Reitsma PH, Mandalaki T, Kasper CK, Bertina RM, Briet E: Two novel point mutations correlate with an altered developmental expression of blood coagulation factor IX (hemophilia B Leyden phenotype). Blood 73:743, 1989

S. Crossley M, Winship PR, Austen DEG, Rizza CR, Brownlee GC: A less severe form of Haemophilia B Leyden. Nucleic Acids Res 18:4633, 1990

6. Crossley M, Brownlee GG: Disruption of a C E B P binding site in the factor 1X promoter is associated with haemophilia B. Nature 34S:444, 1990

7. Royle G, Van De Water NS, Berry E, Ockelford PA, Browett PJ: Haemophilia B Leyden arising de novo by point mutation in the putative factor TX promoter region. Br J Haematol 77: 19 I , I991

8. Picketts DJ, D'Souza C, Bridge PJ, Lillicrap D: An A to T transversion at position -S of the factor IX promoter results in hemo- philia B. Genomics 12:161, I992

9. Reijnen MJ, Peerlinck K, Maasdam D, Bertina RM, Reitsma PH: Hemophilia B Leyden: Substitution o f thymine for guanine at position -21 results in a disruption of a hepatocyte nuclear factor 4 binding site in the factor IX promoter. Blood 82: 1 S I , I993

10. Briet E, Bertina RM, Van Tilburg NH, Veltkamp JJ: Hemo- philia B Leyden: A sex linked hereditary disorder that improves after puberty. N Engl J Med 306:788, 1982

I 1. Reijnen MJ, Sladek FM, Bertina RM, Reitsma PH: Disruption

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

3000 PICKETS, MUELLER, AND LlLLlCRAP

of a binding site for hepatocyte nuclear factor 4 results in hemophilia B Leyden. Proc Natl Acad Sci USA 89:6300, 1992

12. Picketts DJ, Lillicrap DP, Mueller CR: Synergy between tran- scription factors DBP and C/EBP compensates for a haemophilia B Leyden factor IX mutation. Nature Genet 3: 175, 1993

13. Crossley M, Ludwig M, Stowell KM, De Vos P, Olek K, Brownlee GG: Recovery from hemophilia B Leyden: An androgen- responsive element in the factor IX promoter. Science 257:377, 1992

14. Gorski K, Carneiro M, Shibler U: Tissue-specific in vitro transcription from the mouse albumin promoter. Cell 47:767, 1986

15. Jones DH, Howard BH: A rapid method for site-specific muta- genesis and directional subcloning by using the polymerase chain reaction to generate recombinant circles. Biotechniques 8: 178, 1990

16. Maire P, Wuarin J, Schibler U: The role of cis-acting promoter elements in tissue-specific albumin gene expression. Science 244:343, 1989

17. Lichtsteiner S, Wuarin J, Schibler U: The interplay of DNA- binding proteins on the promoter of the mouse albumin gene. Cell 5 1 :963, 1987

18. Sawadogo M, Roeder RG: Factors involved in specific tran- scription by human RNA poymerase 11: Analysis by a rapid and quantitative in vitro assay. Proc Natl Acad Sci USA 824394, 1985

19. Mueller CR, Maire P, Schibler U: DBP, a liver-enriched tran- scriptional activator, is expressed late in ontogeny and its tissue specificity is determined posttranslationally. Cell 61:279, 1990

20. van der Hoorn FA, Tamasky HA: Factors involved in regula- tion of the RT7 promoter in a male germ cell-derived transcription system. Proc Natl Acad Sci USA 89:704, 1992

21. Andrew M, Vegh P, Johnston M, Bowker J, Ofosu F, Mitchell

L: Maturation of the hemostatic system during childhood. Blood 80: 1998, 1992

22. Simpson NE, Biggs R: The inheritance of Christmas factor. Br J Haematol 8: 191, l962

23. Sweeney JD, Hoernig LA: Age-dependent effect on the level of factor IX. Am J Clin Pathol 99:687, 1993

24. Birkenmeier EH, Gwynn B, Howard S, Jerry J, Gordon JI, Landshulz WH, McKnight SL: Tissue-specific expression, develop- mental regulation, and genetic mapping of the gene encoding CCAAT/enhancer binding protein. Genes Dev 3: 1146, 1989

25. Hirosawa S, Fahner JB, Salier J-P, Wu C-T, Lovrien EW, Kurachi K: Structural and functional basis of the developmental regulation of human coagulation factor IX gene: Factor 1X Leyden. Proc Natl Acad Sci USA 87:4421, 1990

26. Yao S-N, DeSilva AH, Kurachi S, Samuelson LC, Kurachi K: Characterization of a mouse factor IX cDNA and developmental regulation of the factor IX gene expression in liver. Thromb Haemost 65:52, 1991

27. Klein-Hitpass L, Tsai SY, Weigel NL, Allan GF, Riley D, Rodriguez R, Schrader WT. Tsai M-J, O’Malley BW: The progester- one receptor stimulates cell-free transcription by enhancing the for- mation of a stable preinitiation complex. Cell 60:2427, 1990

28. Yoshitake S, Schach BG, Foster DC, Davie EW, Kurachi W: Nucleotide sequence of the gene for human factor IX (antihaemo- philic factor B). Biochem J 24:3736, 1985

29. Anson DS, Choo KH, Rees DJG, Giannelli F, Gould K, Hud- dleston JA, Brownlee GG: The gene stucture of human anti-haemo- philic factor IX. EMBO J 3:1053, 1984

30. Smale ST, Baltimore D: The “Initiator” as a transcription control element. Cell 57:103, 1989

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom

1994 84: 2992-3000  

DJ Picketts, CR Mueller and D Lillicrap hemophilia B Leyden mutationsacting elements and the deleterious effects of naturally occurring Transcriptional control of the factor IX gene: analysis of five cis- 

http://www.bloodjournal.org/content/84/9/2992.full.htmlUpdated information and services can be found at:

Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

  Copyright 2011 by The American Society of Hematology; all rights reserved.Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American

For personal use only.on December 27, 2018. by guest www.bloodjournal.orgFrom