-
Proc. Nat. Acad. Sci. USAVol. 69, No. 9, pp. 2518-2522,
September 1972
Isolation of Supercoiled Colicinogenic Factor E1 DNASensitive to
Ribonuclease and Alkali
(chloramphenicol/Escherichia coli/supercoiled
DNA/rifampicin)
D. G. BLAIR*, D. J. SHERRATTt, D. B. CLEWELLT, AND D. R.
HELINSKI
Department of Biology, University of California, San Diego,
Revelle College (La Jolla) 92037
Communicated by S. J. Singer, June 19, 1972
ABSTRACT The synthesis of the covalently-closed,circular DNA
form of colicinogenic factor El (ColEl) con-tinues in Escherichia
coli cells after the addition of chlor-amphenicol. A large portion
of the purified supercoiledColE, DNA molecules made in the presence
of chloram-phenidol are converted to the open circular DNA form
aftertreatment with alkali (pH 13), RNase A, or RNase H.
Thesetreatments do not significantly affect the
covalently-closedform of ColE, DNA isolated from normally growing
E. colicells. The open circular product resulting from treatmentof
supercoiled ColE, DNA with RNase A possesses a singlebreak in one
strand of the circular duplex. The site sensi-tive to RNase A
occurs with equal probability in either ofthe complementary
strands. Both synthesis of ColE, DNAand the formation of
supercoiled ColE, DNA sensitive toRNase A or alkali are prevented
by the inhibitor of RNAsynthesis, rifampicin. These results
indicate that cova-lently-closed ColE, DNA containing one or more
ribo-nucleotides accumulates during ColE, replication in
thepresence of chloramphenicol. It is proposed that this
in-corporated RNA served as a primer during the initiation
ofsynthesis of ColE, DNA and that its removal from thecircular DNA
is inhibited in cells incubated in the presenceof
chloramphenicol.
The colicinogenic factor El (ColEl) is one of a group of
relatedbacterial plasmids found in Escherichia coli that code for
theproduction of extracellular, antibiotic proteins. The
ColElplasmid is a covalently-closed, circular duplex DNA
moleculewith a molecular weight of 4.2 X 106. About 20-30 copies
percell are present under conditions of normal logarithmic
growth(1, 2). A fraction of these molecules can be isolated in
theform of a relaxation complex of supercolled DNA and protein(3).
Synthesis of ColE1 DNA continues when protein syn-thesis is
inhibited by the addition of chloramphenicol (CM)(4). While
synthesis of the bacterial chromosome ceases within1-2 hr after
addition of CM, plasmid synthesis is maintainedfor up to 10-20 hr,
resulting in the accumulation of as many as1000-3000 copies of
supercoiled ColEl DNA per cell (4, 5).Under these conditions,
virtually all of the plasmid DNAcan be isolated as covalently
closed, protein-free molecules(4). In this report the sensitivity
of these molecules to highpH and to certain ribonucleases is
described, properties that
Abbreviations: ColE,, colicinogenic factor El; CM,
chloram-phenicol.* Department of Chemistry, U.C.S.D.t Present
address: Microbial Genetics Group, School of Biology,University of
Sussex, Falmer, Brighton BN1 9QF, England.I Present address:
Departments of Oral Biology and Micro-biology, University of
Michigan, Schools of Dentistry andMedicine, Ann Arbor, Mich.
48104.
2518
indicate that they contain one or more ribonucleotides aspart of
their covalently-closed, double-stranded structure.A model will be
presented for the formation of these ribonu-cleotide-containing DNA
molecules as the result of a re-quirement for an RNA primer for the
initiation of ColE,DNA synthesis. An RNA primer was first suggested
by Brut-lag et al. (6) to explain the requirement for RNA synthesis
inthe conversion of single-stranded M13 bacteriophage DNAto the
RFII form. A similar dependence of DNA synthesis onRNA synthesis
has also been demonstrated in several othersystems including the
plasmid ColEj (5-8).
MATERIALS AND METHODS
Strains and Media. The E. coli K-12 strains, JC411 (ColE,)and
CR34 (ColE,), and the Tris HCl and phosphate-bufferedmedia used in
these experiments have been described (i, 9).The ColE, plasmid in
these strains was derived from E. coliK-30.
Growth and Labeling Conditions. Cells were routinely grownfrom a
2% inoculum to a cell density of 3 to 5 X 108 cells perml. In the
absence of chloramphenicol, cells were labeled with10-20 /Ci/ml of
['H]thymine (40-60 Ci/mmol) in the pres-ence of 1 /Ag/ml of
unlabeled thymine. Chloramphenicol treat-ment involved the addition
of solid chloramphenicol to aconcentration of 150 MAg/ml to cells
growing logarithmicallyat a density of 3-5 X 108 cells per ml.
Label was added eitherimmediately after addition of CM, or 1-2 hr
later, whenchromosomal DNA synthesis had ceased. For labeling
with[I4C]thymine, the final concentration was 0.3 MCi/ml in atotal
thymine concentration of 2.7 ,ug/ml.For 32p labeling, 200-500 mCi
of carrier-free H3 2PO4 was
added per 30 ml of culture. The media contained about
1mmol/liter of unlabeled inorganic phosphate.
Isolation and Purification of DNA. Labeled cells were lysedand
the ColEj DNA was partially purified from the bulk ofthe
chromosomal DNA by the lysozyme-EDTA-Triton-X100procedure (3, 9).
Supercoiled ColE, DNA was isolated by thedye-bouyant density
procedure of Radloff et al. (10). Theethidium bromide was removed
by 1 or 2 extractions at 40with CsCl-saturated isopropanol (11).
CsCl was then removedby dialysis at 4° against TESP [50 mM Tris HCl
(pH 8)-50mM NaCl-5mM EDTA-50 mM K2HPO4].
Analysis of Alkali-induced Breakdown of Supercoiled ColE,DNA.
ColE, DNA (0.1-1 ,ug) was incubated at pH 13 and370 in a total
volume of 200 Mul consisting of 20 ,ul of
Dow
nloa
ded
by g
uest
on
Apr
il 5,
202
1
-
RNase-Sensitive ColE, DNA 2519
purified DNA in TESP plus 180 Ml of 0.2 M K2HPO4 (pH13). After
incubation, the reaction mixture was neutralizedby addition of 200
jl of 1.0 M Tris*HCl (pH 7.5). 300 ulof the neutralized mixture
were then layered on a 5-ml 5-20% sucrose gradient containing 50 mM
Tris-1.05 M NaCl-5 mM EDTA and centrifuged 110 min at 45,000 rpm
and 150in a Spinco SW50.1 rotor in a Beckman model L2 or IA
prepar-ative ultracentrifuge. Under these conditions,
irreversiblydenatured, covalently-closed, circular DNA sediments to
aposition near the bottom of the tube, while denatured strandsof
ColE1 DNA and any undenatured or renatured ColE1 DNAare found near
the center of the tube as two clearly resolvedpeaks. The denatured
material is found in the leading peak inthis region, ahead of the
position of native supercoiled ColE1DNA.
RNase Treatment of Supercoiled ColE1 DNA. PancreaticRNase A at a
concentration of 3 mg/ml in 9 mM Tris HCl(pH 7:5) was first treated
by heating for 5 min in a boilingH20 bath. The solution was then
quickly cooled on ice andkept at 40 until use. Incubations were
performed at 370 in afinal reaction volume of 300 Al. The reaction
mixtures con-tained 0.1-1ug of ColE1DNA and 300 Ag of heat-treated
RNaseA in the following buffer: 36 mM Tris*HCl (pH 8.0)-33
mMNaCl-3mM EDTA-13 mM K2HPO4. After RNase incubationfor the desired
time, 100 Ml of a solution of 5 mg/ml of Pronasein TES [50 mM Tris
HCl (pH 8.0)-50 mM NaCl-5 mMEDTA] was added to each reaction
mixture and incubationwas continued for 10 min at 37°. The Pronase
had been di-gested for 30 min at 370 to eliminate possible nuclease
con-tamination. After Pronase treatment, a 0.3-ml portion of
eachmixture was analyzed by centrifugation in a 5-ml, 5-20%sucrose
gradient for 165 min in the Spinco SW50.1 rotor. Thesucrose
gradient contained 50 mM Tris. HCl (pH 8)-0.55 MNaCl-5mM EDTA. The
Pronase treatment was necessary toinsure good recoveries of DNA
from the sucrose gradients.Without this treatment, the ColE1 DNA
pelleted to the bottomof the centrifuge tube, apparently as a
result of binding toRNase molecules.RNase T1 and RNase H
incubations were performed in a
similar fashion. RNase T, reaction mixtures (200,gl)
contained0.1-1 Ag ColE1 DNA and 300-3000 units of T, ribonuclease
in55 mM Tris HCl-20 mM EDTA-5 mM NaCl at pH 7.5.Pronase treatment
was as described above. Both heated andunheated T, RNase were
tested. For RNase H, the buffer was2 mM Mn++-20 mM (NH4)2S04-36 mM
Tris * HCl-1 mM 2-mercaptoethanol-6 mM NaCl, at pH 7.9. The
heat-labileRNase H was not heated before use.
Alkaline CsCl Centrifugation. DNA was centrifuged in aSpinco
SW56-Ti rotor for 36-40 hr at 40,000 rpm in a 1.5-mlgradient
containing 1.38 g CsCl, 30 Ag bovine-serum albumin,and 0.02%
Sarkosyl, in 0.12 M Na3PO4. The uncorrected pHof this solution was
12.6-12.7 (Radiometer type B electrode-GX2301B standardized with
Beckman pH 12.45 buffer solu-tion no. 3010).
Reagents and Enzymes. The sources of most of the reagentsused
have been described (3, 9). Ribonuclease A (beef pan-creas, 3000
units/ml; code RASE) and ribonuclease T1 (3 X105 units/mg; code
RT,) were obtained from WorthingtonBiochemical Corp., Freehold,
N.J. Pronase CB and chlor-amphenicol were obtained from
Calbiochem., San Diego,Calif. Ribonuclease H was generously
provided by Dr. Gordon
E6
)ICci
FRACTION NUMBERFIG. 1. Alkaline CsCl equilibrium centrifugation
of a mixture
of CM and non-CM ColE1 DNA. Covalently-closed ColE1 DNAwas
purified from logarithmically growing JC411 (ColEi) beforeand after
incubation of the cells in the presence of chloramphen-icol (150
Mg/ml) for 18 hr. Conditions of alkaline CsCl centrifuga-tion,
fractionation, and counting were as described in Methods.(0-0)
3H-labeled non-CM ColE1 DNA. (0- - -0) 32P-labeledCM ColE1 DNA.
Gill. Rifampicin (rifampin) was a gift of the CIBA
Pharma-ceutical Company, Summit, N.J.
RESULTSAlkaline sensitivity of covalently-closed ColE1 DNA
The differential behavior in E. coli strains JC411 (ColE1)
orCR34 (ColEj) of supercoiled ColE1 DNA synthesized in thepresence
and absence of CM was first noted when purifiedmixtures (see
Methods) of the two types of DNA were centri-fuged to equilibrium
in alkaline CsCl (pH 12.8). As shown inFig. 1, a large fraction of
the supercoiled DNA from CM-treated JC411 (ColE1) cells (CM ColEj)
is converted from thecovalently-closed form, banding in the dense
position, to aform banding in the lower density (single-stranded)
positionby the introduction of one or more breaks in the closed
DNAhelix. ColE1 DNA isolated from cells growing logarithmicallyin
the absence of chloramphenicol (non-CM ColE1) is lesssensitive to
conditions of alkaline centrifugation, since a muchsmaller fraction
bands in the single-stranded position. Thishigh pH-induced
conversion was examined in more detail inthe experiment shown in
Fig. 2. Mixtures of CM and non-CM ColE1 supercoils were incubated
for various times at pH13 in phosphate buffer, after which the pH
was reduced to 8and the mixtures were analyzed by sucrose gradient
velocitysedimentation to determine the fraction of
covalently-closedDNA remaining. As shown in Fig. 2, CM ColE1 DNA
loses itscovalently-closed structure during incubation at high
pH,while non-CM ColE1 is only slightly affected by the
alkalineconditions during the experiment. Denaturation of
CM-ColE1DNA at pH 13 at 40 followed immediately by
neutralizationdoes not result in a significant loss of its
covalently-closedstructure. The alkaline-sensitive supercoiled CM
ColE1 DNAwas indistinguishable from supercoiled non-CM ColE1 DNAby
other criteria. Both exhibit identical sedimentationbehaviorin
neutral sucrose density gradients and the same bouyantbehavior
during neutral CsCl and CsCl-ethidium bromideequilibrium
centrifugation, and appear the same when ex-amined in the electron
microscope.
Sensitivity ofCM ColE1 DNA to ribonucleasesThe observation by
Clewell et al. (5) that ColE1 DNA syn-thesis in the presence of
chloramphenicol could be inhibited
Proc. Nat. Acad. Sci. USA 69 (1972)
Dow
nloa
ded
by g
uest
on
Apr
il 5,
202
1
-
Proc. Nat. Acad. Sci. USA 69 (1972)
100 Lnormal supercoiled Cal El
60 - CM supercoiled
40 \
20 CM-supercoiled Col El
0 Jo0 6 12 18 24 30 40
MINUTES OF INCUBATION
e
C=
300 E.X-
FRACTION NUMBER70 100
FIG. 2. Stability of covalently-closed ColE1 DNA duringtreatment
at pH 13 in 0.18 M P04-3* Samples of ColE1 DNA,similar to that
described in Fig. 1 (prepared by Dr. P. Williams),were incubated at
the temperatures and for the times indicated,then neutralized and
analyzed on 5-20% sucrose gradients con-taining 1.05 M NaCl.
Supercoiled DNA is expressed as a per-centage of total ColE1 DNA
recovered from the gradients. Re-coveries of 70% or more of the
counts layered were obtained forboth labels in all cases but one,
the 1-min incubation at 370,where the 3H recovery was 64%, compared
to a 32p recovery of95%. (- * and 0---O) incubations at 370; (A--A
andA- - -A) incubations at 600.
by the inhibitor of RNA synthesis, rifampicin (12),
implicatedRNA involvement in the ColE1 replication process. Since
thepresence of RNA in supercoiled CM ColE1 DNA could accountfor its
alkaline sensitivity, it was decided to test the sensitivityof this
DNA to various ribonucleases. The results of incubat-ing a mixture
of differentially labeled CM and non-CM ColE1supercoiled DNA in the
presence of pancreatic ribonuclease Aare shown in Fig. 3. The
sucrose gradient profiles (a and b)show that a lO-min treatment
with 1 mg/ml of RNase A con-verts over 40% of the CM ColE1 DNA from
a form sedimentingas a covalently-closed molecule to one
sedimenting at the ratecharacteristic of an open circular molecule.
The differentiallylabeled non-CM ColE1 DNA, present as an internal
control,is not affected by this treatment. No further conversion of
thesupercoils could be induced by the subsequent addition offresh
ribonuclease. The RNase A used in all incubations hadbeen treated
by heating for 5 min at 1000 to inactivate anydeoxyribonucleases
that might be present. Prior phenol treat-ment of CM ColE1 DNA does
not alter its sensitivity to RNaseA. The fraction of CM-supercoils
that were resistant to pan-creatic RNase were no more sensitive to
alkali treatment [40hr, alkaline CsCl (pH 12.8)] than non-CM
supercoiled ColE1DNA. The incubations in this experiment were done
with anexcess of RNase over that needed to induce a detectable
con-version, in that 25 ,g/ml of RNase A was sufficient to
producedetectible nicking in 60 min at 370 under these
conditions.For determination of the nature of the products of the
RN-
ase A treatment, the CM CoLE1 DNA, after incubation withRNase A,
was banded in a dye-CsCl gradient and the lower(covalently-closed)
and upper (open) bands, as well as un-treated starting material,
were examined in the electron micro-scope. Essentially 100% of the
untreated CM ColE1 DNA waspresent in the supercoiled form. The
heavy peak isolated fromthe RNase-treated sample was also
predominately super-coiled DNA, while the light peak was
essentially all open cir-cular material, with C->
Be
Dow
nloa
ded
by g
uest
on
Apr
il 5,
202
1
-
RNase-Sensitive ColE1 DNA 2521
from one such experiment is shown in Fig. 4a. The ratio
ofcircular to linear strands indicates that at least 80% of
thestarting open circular DNA duplexes contained only onenicked
strand. In other experiments this fraction has ap-proached 100%.
The absence of a significant amount of trail-ing from the slower
sedimenting peak of linear material indi-cates that most of the
nicked DNA strands contained onlyone RNase-sensitive site.The
separated linear and circular pools were then analyzed
by poly(UG)-CsCl centrifugation by the method of Szybalski(14)
as described (9). The profiles for the two pools (Fig. 4band c)
indicate that both the circular and linear strands con-sisted of
equal numbers of each of the complementary ColE1DNA strands. The
RNase-sensitive site thus appears to bedistributed randomly with
respect to the two strands in CMColE1 DNA.
Effect of rifampicin on the formation of RNaseA-sensitive ColE,
supercoilsAs has been shown, supercoiled ColEj DNA made in the
ab-sence of chloramphenicol is not sensitive to RNase or
alkali.
to
N
CQ
E
24
18 _
12
6
(a)
(b)
ight
heay
I(c)
heavy light
kI
0 12 24 36 48 0 12 24 36 48
FRACTION NUMBER
FIG. 4. Analysis of the strand specificity of the nick(s)
inducedby RNase A in CM ColE, supercoiled DNA.
Covalently-closedColE, DNA was isolated by equilibrium
centrifugation fromJC411 (ColE,) cells incubated in the presence of
150 Ag/mlchloramphenicol for 7 hr. The [3H]thymine label was added
tothe culture after 2 hr in chloramphenicol. The supercoiled
ColE,DNA was further purified on a 5-20% sucrose gradient.
Thismaterial was treated for 60 min at 370 (see Methods) and
rebandedin a dye-CsCl gradient. The upper band (open circular
DNA)was then pooled and the ethidium bromide and CsCl were
re-moved. A 0.2-ml portion of this material was then layered on a5
ml, 5-20% alkaline sucrose density gradient and centrifugedfor 225
min at 55,000 rpm, 15°, in an SW65 rotor; 4-drop frac-tions were
collected, and 1O-ul aliquots of each fraction werespotted and
counted. The peak fractions were then pooled,neutralized, and
analyzed on CsCl gradients containing poly(UG)as has been described
(9). (a) Alkaline sucrose velocity sedi-mentation; (b)
poly(UG)-CsCl equilibrium centrifugation of thematerial from the
single-stranded linear pool of (a); (c) poly-(UG)-CsCl equilibrium
centrifugation of the material from thesingle-stranded circular
pool of (a).
RNase sensitivity can not be detected until 1-2 hr after
theaddition of CM to a logarithmically growing culture of
JC411(ColEj). The relative percentage of sensitive ColE,
supercoilsthen continues to increase at roughly a linear rate.
WhenDNA synthesis, as detected by the incorporation of radio-active
label into DNA, ceases in the culture, the increase inthe fraction
of sensitive ColE, supercoils stops. When the cul-ture is
incubated, with aeration, at 370 for 10-20 hr afterDNA synthesis
has ceased, there is no change in the level ofRNase- or
alkali-sensitive ColE1 DNA. These results are con-sistent with a
requirement for active ColE, DNA synthesis forthe generation of the
sensitive form of CoLE, DNA.
In support of this hypothesis, rifampicin prevents both
thesynthesis of ColE, DNA and the generation of the RNase-
oralkali-sensitive form of ColE, in the presence of
chlorampheni-col. When chloramphenicol-treated cultures that are
activelysynthesizing ColE, DNA (a fraction of which is RNase-
andalkaline-sensitive) are treated with 3 Ag/ml of rifampicin,DNA
synthesis ceases and no further increase in the percent-age of
sensitive ColE, supercoils is detected over a 4-hr period.Control
cultures that are not treated with rifampicin continueto synthesize
ColE, DNA and to generate sensitive ColE,supercoils.
DISCUSSION
The direct dependence of DNA replication on RNA synthesishas
been proposed on the basis of scattered observations fromseveral
systems (15, 16). Recently, the number of cases wheresuch A
relationship appears to exist has increased rapidly withthe finding
that DNA synthesis in several systems is sensitiveto inhibition of
RNA synthesis by inhibitors of RNA poly-merase such as rifampicin
and streptolydigin (5-8). Thissensitivity does not appear to be a
universal property, how-ever, since DNA synthesis of the
replicative form of phage,X174 is resistant to rifampicin (17).
Covalently-closed
Insertion of DNA(W) RNA Primer Replication -
Completionof
Replication
Removal g ( )Repar or
/~Z blocked ord, I+('(f\~ reduced by\ /CM addition)
FIG. 5. Primer model for the formation of RNA-containingColE,
supercoiled DNA. We propose that the initiation of DNAsynthesis
involves the rifampicin-sensitive synthesis of RNA thatis
complementary to a single site on either DNA strand. DNAsynthesis
involves the addition of deoxyribonucleotides to thisprimer RNA.
Either a single RNA primer, complementary toeither strand, is
involved in the duplication of any one circularDNA molecule and
synthesis proceeds unidirectionally, or syn-thesis proceeds
bidirectionally from two RNA primer moleculespresent in a single
replicating DNA molecule. Normally, theRNA primer is removed upon
completion of replication of thecircular DNA, however, in the
presence of chloramphenicol theRNA-containing, supercoiled DNA
molecules accumulate.
Proc. Nat. Acad. Sci. USA 69 (1972)
Dow
nloa
ded
by g
uest
on
Apr
il 5,
202
1
-
Proc. Nat. Acad. Sci. USA 69 (1972)
Where it appears to be required, RNA synthesis has
beenpostulated to function either directly or indirectly in
theinitiation of DNA synthesis. For conversion of M13
single-stranded phage DNA to the double-stranded RFII form, ithas
been demonstrated that, at least in vitro, RNA acts as aprimer and
is covalently joined to the newly synthesizedDNA strand (18). The
idea that RNA might serve as a primerfor DNA synthesis is
attractive, since it offers a way of initiat-ing DNA synthesis
using existing DNA polymerizing enzymes,none of which are capable
of initiating de novo synthesis invitro in their purified forms
(19, 20).The evidence we have presented here suggests that when
protein synthesis is inhibited by chloramphenicol,
covalently-closed ColE1 DNA containing one or more ribonucleotides,
ineither one strand or the other, accumulates in colicinogenicE.
coli cells during the period that these cells are
activelysynthesizing ColE, DNA. Such structures, containing a
shortsegment of DNA-RNA hybrid, would clearly be expected tobe
sensitive to alkaline and RNase H hydrolysis, and evidencesupports
the notion that these molecules would also be sensi-tive to RNase A
(21, 22). The lack of sensitivity to RNase T,may imply the absence
of ribosylguanine in the molecules orperhaps some unknown
structural limitations to T, action inthe case of a hybrid
substrate. The possibility that an unusualdeoxyribonucleotide might
be responsible for these unusualproperties of the ColE1 DNA cannot
be rigorously excluded,but it seems unlikely that such a base would
be sensitive toalkali and to two different ribonucleases.The
existence of a covalently-closed DNA duplex containing
RNA raises the question of its origin and function in
DNAmetabolism. One possibility is that it arises as an artifact
ofthe long-term exposure of the colicinogenic cells to
chlor-amphenicol, which perhaps induces infrequently the
erroneousinsertion of a ribonucleotide during DNA
polymerization.Polymerase I will make such errors in the presence
of Mn++(22), and ColE1 requires polymerase I for normal
plasmidmaintenance (23). This model cannot be ruled out at this
time,but in view of the rifampicin sensitivity of CotE1 synthesis,
anessential role of RNA in the process is suggested. Evidence
hasbeen obtained for a role for RNA as a primer in the initiationof
DNA synthesis in the conversion of M13 viral single strandsto the
double-stranded replicative form (6, 18). Fig. 5 illus-trates a
possible model for the formation of RNA-containingCotE1 supercoils
as a result of such an initiation step. Theobservation that the
RNase-sensitive segment of ColE1 DNAis not strand-specific
requires, according to this model, thateither initiation is
unidirectional and random with respect tothe DNA strand, or
synthesis is bidirectional and initiated
with short RNA segments from both strands of the
replicatingmolecule. One might expect, however, that the site of
initia-tion on a particular strand is specific.We thank Mr. Bernard
Ashcraft for his excellent technical
assistance and Dr. Gordon Gill for his generous provision ofthe
enzyme RNase H. This work was supported by U.S. PublicHealth
Service Research Grant AI-07194 and National ScienceFoundation
research grant GB-29492. D. G. B. was supportedby a U.S. Public
Health Service Predoctoral Traineeship (2-T01-6M-1045). D. R. H. is
a U.S. Public Health Service ResearchCareer Development Awardee
(K04-6M07821).
1. Bazaral, M. & Helinski, D. R. (1968) J. Mol. Biol. 36,
185-194.
2. Clewell, D. B. & Helinski, D. R. (1970) Biochemistry 9,
4428-4440.
3. Clewell, D. B. & Helinski, D. R. (1969) Proc. Nat. Acad.
Sci.USA 62, 1159-1166.
4. Clewell, D. B. & Helinski, D. R., J. Bacteriol. (1972)
110,1135-1146.
5. Clewell, D. B., Evenchick, B. & Cranston, J. W.
(1972)Nature New Biol. 237, 29-31.
6. Brutlag, D., Schekman, R. & Kornberg, A. (1971) Proc.Nat.
Acad. Sci. USA 68, 2826-2829.
7. Lark, K. G. (1972) J. Mol. Biol. 64, 47-60.8. Kline, B. C.
(1972) Biochem. Biophys. Res. Commun. 46,
2019-2025.9. Blair, D. G., Clewell, D. B., Sherratt, D. J. &
Helinski, D. R.
(1971) Proc. Nat. Acad. Sci. USA 68, 210-214.10. Radloff, R.,
Bauer, W. & Vinograd, J. (1967) Proc. Nat.
Acad. Sci. USA 57, 1514-1521.11. Sebring, E. D., Kelley, T. J.,
Jr., Thoren, M. M. & Salzman,
N. P. (1971) J. Virol. 8, 478-490.12. Sippel, A. & Hartmann,
G. (1968) Biochim. Biophys. Acta.
157, 218-219.13. Hausen, P. & Stein, H. (1970) Eur. J.
Biochem. 14, 278-283.14. Szybalski, W., Kubinski, H., Hradecna, Z.
& Summers, W.
C. (1971) in Methods in Enzymology eds. Grossman, L.
&Moldave, K. (Academic Press, New York), Vol. XXI, partD, pp.
383-413.
15. Doudney, C. 0. (1966) Nature 211, 39-41.16. Dove, W. F.,
Inokuchi, H. & Stevens, W. F. (1971) in The
Bacteriophage Lambda, ed. Hershey, A. D. (Cold SpringHarbor
Laboratory, New York), pp. 747-771.
17. Silverstein, S. & Billen, D. (1971) Biochim. Biophys.
Acta.247, 383-390.
18. Wickner, W., Brutlag, D., Schekman, R. & Kornberg,
A.(1972) Proc. Nat. Acad. Sci. USA 69, 965-969.
19. Goulian. M. (1968) Cold Spring Harbor Symp. Quant. Biol.33,
11-20.
20. Wickner, R., Ginsberg, B. & Hurwitz, J.. (1972) J.
Biol.Chem. 247, 498-504.
21. Van de Sande, J. H., Loewen, P. & Khorana, H. G.
(1972)Fed. Proc. 31, 426Abstr.
22. Berg, P., Fancher, H. & Chamberlin, M. (1963) in
Informa-tional Macromolecules (Academic Press, New York),
pp467-483.
23. Kingsbury, D. T. & Helinski, D. R. (1970) Biochem.
Biophys.Res. Commun. 41, 1538-1544.
2522 Biochemistry: Blair et al.
Dow
nloa
ded
by g
uest
on
Apr
il 5,
202
1