Proc. NatL Acad. Sci. USAVol. 78, No. 10, pp. 6111-6115, October 1981Biochemistry

Escherichia coli factor Y sites of plasmid pBR322 can function asorigins of DNA replication

(site specificity/ATP hydrolysis/initiation of DNA synthesis)

S. L. ZIPURSKY AND K. J. MARIANSDepartment of Developmental Biology and Cancer, Division of Biological Sciences, Albert Einstein College of Medicine, Bronx, New York 10461

Communicated by jerard Hurwitz, July 6, 1981

ABSTRACT The Eacherichia coli replication factor Y, in con-junction with other genetically undefined E. coli replication fac-tors and the gene products of the E. coli dnaB, dnaC, and dnaGloci, is involved in de novo primer formation on the qbX174 (+)single-strand circular DNA template [(+) ss(c)DNA]. The partic-ipation of factor Y in this series of reactions is correlated with its4X174 (+) ss(c)-specific DNA-dependent ATPase activity. Re-cently two factor Y effector DNA segments of the plasmid pBR322have been identified in close proximity to the plasmid origin ofDNA replication. We report here that insertion of these factor Ysites into the filamentous phage flR229 (+) ss(c)DNA confers uponit the ability to be converted to RF DNA in vitro through a rif-ampicin-resistant dnaB, dnaC, and dnaG gene product-dependentpathway. Our data suggest that factor Y effector sites can functionas origins of DNA replication.

Three enzymatic pathways for do novo primer formation on sin-gle-stranded circular DNA [ss(c)DNA] templates in E. coli havebeen described. The initiation of bacteriophage fl and G4ssDNA -- RF DNA synthesis involves the polymerization ofa primer molecule at a unique site on the template DNAs bythe Escherichia coli RNA polymerase (1-3) and the dnaG pro-tein [(4-6, primase], respectively. Primer formation on bacte-riophage 4X174 ss(c)DNA requires the E. coli dnaB, dnaC, anddnaG gene products and other genetically undefined host fac-tors (7, 8). The template specificity of the 4X174 pathway isdetermined by the E. coli factor-Y protein (9) [n' in the no-menclature of Shlomai and Kornberg (10)]. This protein directsthe formation of a multienzyme priming complex [the "pri-mosome" (11)] at a specific site on the DNA template [the originof 4X174 (-) strand synthesis (10)]. Although formation of thiscomplex is DNA-sequence specific, the primosome can migratealong the DNA template and synthesize primers at many dif-ferent sites (11). Thus, the 4X174 factor-Y site is here consid-ered a replication origin insofar as its presence is required forprimosome-catalyzed de novo primer formation at many differ-ent sites on contiguous ssDNA..Factor Y was initially characterized as a 4X174 ss(c)DNA-

dependent ATPase required in the prepriming step of 4X174ssDNA -* RF DNA synthesis (7). The region of the 4X174(+) strand responsible for factor-Y ATPase effector activity hasbeen localized to a DNA sequence (located within the F-G in-tergenic space) which contains the presumed 4X174 (-)-strandorigin of DNA replication (10). This suggests that DNA se-quences that can serve as factor-Y effectors may also functionas templates for the dnaB, dnaC, and dnaG gene product-de-pendent mechanism ofde novo primer formation. We have re-ported the identification oftwo segments ofpBR322 DNA, oneon each DNA strand, which serve as effectors in the factor Y

ATPase reaction (12). A comparative analysis of the 4X174 andpBR322 effector sites did not reveal extensive regions of se-quence homology. Nevertheless, the close proximity of thesefactor Y sites to the plasmid orn region led us to suggest that theyplay a role in plasmid replication by virtue of their ability tofunction as 40X174-like origins of DNA replication.To determine whether the factor Y sites from pBR322 DNA

could function as origins of DNA replication, we constructedrecombinant phage fl ss(c)DNAs containing these sites andstudied their ability to be converted to duplex DNA in cell-freeextracts (receptor fractions) ofE. coli (13). The conversion of flss(c)DNA to the duplex form (ssDNA -* RF DNA) is sensitiveto the E. coli RNA polymerase inhibitor rifampicin. In contrastto fl, however, 4X174 ssDNA-) RF DNA synthesis is resistantto rifampicin and requires the gene products ofthe dnaB, dnaC,and dnaG loci. We report here that recombinant fl ss(c)DNAscontaining the factor-Y effector sites from pBR322 DNA can beconverted to duplex DNA in the presence ofrifampicin througha dnaB, dnaC, and dnaG gene product-dependent pathway invitro. Our data are consistent with the hypothesis that factor Yeffector DNA sequences function as origins ofDNA replication.

.MATERIALS AND METHODS

Bacteria and Viruses. E. coli strains used and their genotypeswere: K38 (HfiC), CR34 (thr, leu, thi, supE, lac, tonA, thyA,dra, rpsL8), PC22 (thy, his, malA, xyl, arg, thi, polA, dnaC,rpsL8, su), NY73 (thy, leu, metE, dnaG, polA, rpsL8, rifr,su-), BT1029 (thy, endA, polA, dnaB, su-), HF4704 (4X174s,thy), and HMS83 (polA, polB, thyA, lys, lacZ, rpsL8, rha).Phage strains used were: 4X174 am3 (lysis-, gene E) and f1R229(a gift of P. Model, Rockefeller University), which contains asingle EcoRI site within the intergenic space (14).

Preparations of Enzymes, Protein Fractions, and DNAs.Factor Y was purified by the procedure ofWickner and Hurwitz(9). The dnaB (15), and dnaC (unpublished data) gene productswere purified as described from E. coli HMS83. The E. colissDNA binding protein and the dnaG protein were gifts of J.Chase (this institution) and C. McHenry (University of TexasMedical School), respectively. Ammonium sulfate receptorfractions (49-95 mg/ml) were prepared essentially as described(13) except that 45% rather than 40% (wt/vol) ammonium sul-fate was used.

Plasmid pBR322 DNA was purified from E. coli CR34 trans-formed cells as described (16). 4X174 (+) ss(c)DNA was puri-fied by an established procedure (17). fl and recombinant flss(c)DNA and RF DNA were purified as described (18, 19).

Restriction endonucleases were purchased from Bethesda

Abbreviations: RF DNA, double-stranded DNA in the replicative form;RF DNA, negatively supercoiled circular RF DNA; RFII. DNA, cir-cular RF DNA with a discontinuity in one strand; RFIII DNA, linearRF DNA; ss(c)DNA, single-stranded circular DNA.

The publication costs ofthis article were defrayed in part by page chargepayment. This. article must therefore be hereby marked "advertise-ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact.

6111

6112 Biochemistry: Zipursky and Marians

Research Laboratories (Rockville, MD) or New EnglandBioLabs. T4 polynucleotide kinase and T4 DNA ligase werepurchased from P-L Biochemicals and Bethesda Research Lab-oratories, respectively. The EcoRI linker oliginucleotide, 5'-HO-G-G-A-A-T-T-C-C-OH-3' was purchased from Collabora-tive Research (Waltham, MA).ssDNA -* RF DNA Replication. Reaction mixtures (50 pl)

containing 20 mM Tris HCl (pH 7.5); 10 mM MgCl2; 4 mMdithiothreitol; 40 tLM [3H]- or [a-32P]dGTP (200-1500 cpm/pmol); 40 ,tM each ofdCTP, dATP, and dTTP; 2 mM ATP; 0.1mM each of UTP, GTP, and CTP; rifampicin (10 ,ug/ml); 0.1mM NAD+; receptor fraction (0.57-1.38 mg of protein); 0.11pmol (molecules) of ss(c)DNA (4X174, flR229, or recombi-nants); and, where indicated, the dnaB (0.3 unit), dnaC (0.04unit), or dnaG [2400 units (20)] proteins were incubated for 30min at 30°C. Receptor fractions prepared from the E. coli strainsBT1029 and NY73 were incubated at 39°C for 5 min immedi-ately prior to addition to the reaction mixtures.

Product Analysis. DNA synthesis reactions were stopped bythe addition of EDTA, sodium dodecyl sulfate, and proteinaseK to 10 mM, 0.1%, and 0.1 mg/ml respectively, followed byincubation at 37°C for 90 min. The mixture was extracted threetimes with phenoVchloroform, 1:1 (vol/vol), and one time withchloroform; adjusted to 0.5 M Tris'HCl (pH 7.5); and precipi-tated by the addition of 2.5 vol of ethanol. The products weresubjected to electrophoresis under native conditions in 1% agar-ose gels in 50 mM Tris-HCl/40 mM sodium acetate/i mMEDTA, pH 7.9 at 10 v/cm. Gels were stained with ethidiumbromide and RF values of markers and products were deter-mined by ethidium bromide fluorescence and radioautography,respectively.

RESULTSConstruction and Characterization of Recombinant Bacte-

riophage DNAs. Recombinant DNA molecules containing thefactor Y effector sites from pBR322 DNA and the f1R229 vectorDNA were constructed as described in the legend to Fig. 1. TwoDNA fragments were cloned. One fragment [nucleotides1950-2282 (23)] isolated from an Hae III-Rsa I restriction en-donuclease digest of pBR322 DNA contained the factor Y ef-fector DNA sequence from the H strand. * The other fragment[nucleotides 2283-2521 (23)]t containing the L-strand* factorY effector site was generated by cleavage ofpBR322 DNA withthe restriction endonucleases FnudII and Rsa I. These DNAfragments, each processed separately, were modified withEcoRI linkers (24), inserted into the EcoRI site ofbacteriophagef1R229, and cloned. Recombinant bacteriophage were identi-fied by plaque hybridization (21, 22).Ten recombinant phage DNAs containing the Hae III-Rsa

I fragment (flYBl-10) and ten containing the FnudII-Rsa I frag-ment (flYEl-10) were purified and assayed for their ability toserve as effector DNAs in the factor Y ATPase reaction. FiveDNAs from the flYB series and two from the f1YE series wereactive as effector DNAs. Therefore, these phage DNAs con-tained the factor Y effector site covalently attached to the viral(+) strand DNA. Conversely, those phage DNAs that hybrid-

* We refer to the pBR322 DNA leading- and lagging-strand templatesas H and L, respectively. In ColE1 replication, the leading-strandtemplate is the H strand and the lagging strand, the L strand. TheH and L designations of the two strands of CoIE1 reflect heavy andlight density, respectively.

t Extensive analysis of the sequences required for L-strand effectoractivity has resulted in the reassignment of the limits of the activefragment to nucleotides 2352-2416. The previous limits, which hadbeen reported to be between nucleotides 2418 and 2574, were sub-sequently shown to be incorrect (unpublished data).

Structures of recombinant ss(c)DNA templates

2600 2400 2200 2000 1800I

Fnudli HgiAI Rsa I Pvu II Hae Ill4 A I I

Icnsert/ loned ii

-C)Go0)

4)

.Uca¢

1-

.u

HgiAI iA

e2/ HgiAI 4

flYE3

HgiAl ~~~BamHIHgiAI 0 --M10

ol) HgiAI4

flYE5

HgiA\

BamHI

ed and,n flR229

HgW4~ - t -

4~~~~flYB5

HgiAI ~~~BamHI_HgAWJ-G$o

flYBl

HgiAI

Fio. 1. CloningofpBR322factorYeffectorsitesintoflR229 DNA.A FnudII-Rsa I fragment (nucleotides 2282-2521) and an Haemfl-Rsa I fragment (nucleotides 1950-2282) containing the L-strandand H-strand factorY sites, respectively, were isolated from polyacryl-amide gels by electroelution. These fragments were modified by blunt-ended ligation to hybridizedEcoRI linker oligonuclotides, cleavedwiththe EcoRI restriction endonuclease, and separated from free EcoRIlinker by G-50 column chromatography. flR229 RFI DNA, linearizedbydigestionwiththeEcoRlrestrictionendonuclease (flR229RFIDNAcontains one EcoRI site in the gene II-IV intergenic space), was in-cubated separately with the modified FnudH-Rsa I and Hae -RsaI fragments in the presence of T4 DNA ligase for 48 hr at 15°C. Thisligation mixture was used to transform E. coli K38 and was subse-quently plated on E. coli K38 indicator bacteria. Recombinant bacte-riophage containing the pBR322 factor Y sites were identified byplaque hybridization (21, 22). Characterization of recombinant bac-teriophage is described in Results. The scale at the top of Fig. 1 indi-cates the positions of the nucleotide residues in pBR322 DNA as de-scribed by Sutcliffe (23). r, Origin and direction of pBR322 DNAreplication; *-, Direction of DNA synthesis of the recombinant phageDNAs containing the pBR322.factorY effector sites; IGS, gene H-geneIV intergenic space in flR229 DNA; active ss(c) and inactive ss(c), abil-ity of the recombinant DNAs to serve as factor Y effector DNAs; ,the pBR322 L strand; ----, the pBR322 H strand; -, the flR229(+) strand.

ized to the probe in the initial screen but were inactive as factorY effectors were deduced to contain the sequences that werecomplementary to the Y-site DNA. An inactive (flYBl andf1YE5) and active isolate (flYB5 and f1YE3) from each serieswas plaque-purified. All subsequent experiments were per-formed with DNAs prepared from plaque-purified phagestocks. Restriction endonuclease cleavage of recombinant RFIDNAs confirmed the structure of these recombinants (Fig. 2).

FactorYATPaseEffector Activity ofRecombinant ss(c)DNAs.The ATPase effector activities of cfX174 ss(c)DNA and the re-combinant bacteriophage DNA were compared (Table 1). Therate ofATP hydrolysis catalyzed by factor Y in the presence offlYB5 and f1YE3 ss(c)DNA was 49% and 42%, respectively, of

Proc. Nad Acad. Sci. USA 78 (1981)

Proc. Natd Acad. Sci. USA 78 (1981) 6113

FIG. 2. Orientation of factorY sites in recombinant DNAs. f1YE3(lane B) and f1YE5 (lane C) RFI DNAs were digested with a combi-nation of restriction endonucleasesHgiAI and BamHI. BacteriophageflR229 (lane D), flYBl (lane E), and flYB5 (lane F) RFI DNAs weredigested with a combination of restriction endonucleases HgiAI,BamHI, andPvu II. AHpa II restriction digest offR229RFDNA (laneA) and an Hae Im restriction endonuclease digest of qbX174 RFIDNA(lane G) served as markers. The appearance of new bands of 382 (laneB), 158 (lane 0), 368 (lane E), and 266 (lane F) base pairs is consistentwith the orientation inferred from factorY effector activity and plaquehybridization data. Digests were run on a 6% polyacrylamide gel(acrylamide:bisacrylamide, 30:1 wt/wt) for 60 min at 10 V/cm in 50mM Tlis/50 mM boric acid/0.5 mM EDTA, pH 8.3.

that observed with 4X174 ss(c)DNA as effector. BacteriophageflR229 and the recombinant phages fIYBl and flYE5 ss(c)DNAsdid not serve as effectors in this reaction. Mixing of active andinactive effector DNAs demonstrated that the inability of fl,f1YE5, and flYBI ss(c)DNAs to function as effector DNAs was

not due to inhibitors in these DNA preparations (data notshown).From these data, the restriction enzyme analysis, and plaque

hybridization data (not shown), we concluded that the flYB5and f1YE3 ss(c)DNAs contain the pBR322 H- and L-strand fac-

Table 1. Factor Y ATPase activity on different ss(c)DNAs

DNA effector 32Pi formed,added nmol/10 min

gbX174 3.52f1R229 0.01f1YE3 1.48f1YE5 <0.01flYBl <0.01f1YB5 1.75

Reaction mixtures were as described by Zipursky and Marians (12)and included 0.012 unit of factor Y and 18 fmol of ss(c)DNA(molecules).

tor Y effector sites, respectively, and that the flYB1 and flYE5ss(c)DNAs contain their complementary DNA sequences.DNA Synthesis in Receptor Fractions Prepared from E. coli

Thermosensitive DNA Replication Mutants. Receptor frac-tions of E. coli efficiently convert exogenous 4bX174 ss(c)DNAbut not flR229 ss(c)DNA to the RFII form in the presence ofthe RNA polymerase inhibitor rifampicin (13). In order to sup-

port 4X174 ssDNA -+ RF DNA synthesis, receptor fractions

prepared from E. coli strains carrying a thermosensitive dnaB(BT1029), dnaC (PC22), or dnaG (NY73) gene product requiresupplementation with their respective wild-type proteins. Wehave investigated the template specificity ofthese receptor frac-tions and the gene product requirements for the conversion ofthe recombinant ss(c)DNAs to the duplex form. All DNA syn-thesis reactions were performed at optimal receptor concentra-tions for each particular DNA template and equimolar [mole-cules of ss(c)DNA] concentrations of ss(c)DNA.

Recombinant ss(c)DNAs that contained the pBR322 factor Ysites (f1YB5 and flYE3) served as templates for rifampicin-re-sistant, dnaB-, dnaC-, and dnaG-dependent DNA synthesis(Table 2). The rate and extent of DNA synthesis in reactionssupplemented with f1YE3 or f1YB5 ss(c)DNA was 53-112% and27-53%, respectively, of that observed in reactions containing46X174 ss(c)DNA. In contrast, flR229 and the recombinantss(c)DNAs carrying the DNA sequences complementary to thepBR322 factor Y effector sites (f1YE5 and flYBl) supportedDNA synthesis at a level only 1-7% ofthat observed in reactionssupplemented with 4X174 ss(c)DNA. The nature of the lowlevel of rifampicin-resistant DNA synthesis observed in some

reactions supplemented with fl, flYBi, and f1YE5 ss(c)DNAsand incubated either with or without exogenous wild-type pro-tein is unclear.DNA synthesis on all recombinant ss(c)DNA templates in-

creased upon the addition of wild-type gene products. Thisstimulation was at least 10-fold greater in reactions supple-mented with ss(c)DNAs containing factor Y effector sites thanwith ss(c) DNAs that contained their complementary DNA se-

quences. We consistently observed that DNA synthesis in theabsence ofwild-type gene products in the dnaB and dnaG ther-mosensitive receptor fractions supplemented with ss(c)DNAscontaining factor Y effector sites was higher than that observedon ss(c)DNAs that do not contain factor Y sites. We attributethis difference to incomplete inactivation of the temperature-sensitive gene products in the receptor fractions.

Table 2. Requirement for dnaB, dnaC, and dnaG in 4X174 andrecombinant ssDNA -- RF DNA replication

dGMP incorporation in receptor fractions,pmol/30 min

BT1029(dnaBta) PC22 (dnaC') NY73 (dnaGt)

Template DNA No With No With No Withadded dnaB dnaB dnaC dnaC dnaG dnaG4X174 8.5 77.3 0.6 28.5 3.9 38.0fi 3.5 8.1 0.7 2.0 0.9 1.3f1YE3 5.9 83.5 0.6 15.4 5.8 37.6flYE5 3.5 5.7 0.5 2.0 0.9 1.3flYBi 2.6 5.1 0.6 1.4 0.9 1.4flYB5 4.2 40.8 0.6 8.2 2.1 16.4

Reaction mixtures were as described and included, where indicated,0.74 mg, 0.57 mg, and 1.38 mg of protein from receptor fractions pre-pared from E. coli strains BT1029, PC22, and NY73, respectively. Theamount of wild-type dnaB, dnaC, and dnaG gene products added toeach receptor fraction was as described.

Biochemistry: Zipursky and Marians

6114 Biochemistry: Zipursky and Marians

A B C D

dX174markers

RF II

RF III -

RF I -

Recombinant

markers

411-

.- -

s.. .:

*IE m

- RFII

_ RF III

-- RF I

FIG. 3. Agarose gel electrophoretic analysis of DNA productsformed in receptor fractions. Products were purified from reactionscontaining 4X174 (lane A), f1R229 (lane B), F1YE3 (lane C), andf1YB5 (lane D) ss(c)DNA templates. Reaction and electrophoretic con-

ditions were as described, with 0.74 mg of E. coli BT1029 receptor frac-tions and 0.3 unit of dnaB protein added to each reaction.

Products Formed in Receptor Fractions with Exogenousss(c)DNA Templates. DNA products were isolated from DNAreplication reactions containing E. coli BT1029 receptor frac-tions, exogenous ss(c)DNA template and wild-type dnaB pro-

tein. These 32P-labeled replication products were analyzed bygel electrophoresis under neutral conditions, followed by ra-

dioautography (Fig. 3). The predominent DNA species formedin these reactions comigrated with RFII. Denaturing methyl-mercury/agarose gel electrophoretic analysis (data not shown)indicated that the majority of the DNA products formed in re-actions supplemented with 4X174, flR229, and flYE3 wereunit length linear DNA molecules (5386 pb, 6415 bp, and 6654bp, respectively). These data suggest, therefore, that the lowlevel ofDNA synthesis in reactions supplemented with flR229ss(c)DNA reflected an inability of the receptor fraction to formprimers efficiently on the DNA template rather than an inef-ficient mechanism of primer elongation. In contrast to f1YE3,flR229, and OX174, the flYB5 products migrated as linearDNA molecules ranging in length from 1500 to 6500 bases. Thesignificance of this difference is unclear.

DISCUSSIONThe identification of two factor Y ATPase effector sites locatedin close proximity to the plasmid pBR322 and ColEl ori DNAregions led us to propose a role.for these sequences in plasmidDNA replication. The effector site on the L strand (the lagging-strand template) was postulated to function as a lagging-strandorigin ofDNA replication and that on the H strand (the leading-strand template) as a sequence that converted leading-strandDNA synthesis from a continuous to a discontinuous mode. Thedemonstration in this report that small DNA restriction frag-ments that contain pBR322 factorY effector sequences can func-tion as origins ofDNA replication in vitro is consistent with theroles we have postulated for these sites.The in vitro replication. properties of the pBR322 L-strand

DNA factorY effector site reported here are in agreement withboth in vivo and in vitro replication studies on the closely re-

lated plasmid DNA ColEl. Nomura and Ray (25) have shownthat in the presence of rifampicin, M13 phage DNA containingthe L strand of the ColEl Hae II-E fragment is converted invivo by a dnaB- and dnaG-dependent mechanism to the duplex

replicative form. In contrast, the cloned H strand ofthe ColElHaeII-E fragment was unable. to rescue M13 ssDNA -- RFDNA synthesis when incubated in the presence of rifampicin.These data led Nomura and Ray to propose that the L strandof the ColEl HaeII-E fragment contained a lagging-!strand or-igin of DNA replication that signaled a #X174-like mechanismof de novo primer formation. Their data are consistent with thedemonstration by Staudenbauer et aL (26) that ColEl lagging-strand DNA synthesis in vitro requires the dnaB and dnaG geneproducts. We have reported thatthe ColEl HaeII-E fragment,when denatured, can serve as an effector in the factor Y ATPasereaction (12). Although we have not established on which strandthis site resides, studies on the ColEl-like plasmid DNApBR322 strongly suggests its location on the L strand. BecauseColEl leading- and lagging-strand DNA synthesis occur nearlysimultaneously, a 4X1744ike primosome on the lagging-strandDNA template must be capable of efficiently moving in the 5'-+3' direction along the template strand. Recent studies ofAraiand Kornberg (11) have shown that the primosome assembledat the 4X174 factor Y site moves processively along the templateDNA in precisely this direction.Our proposal that the pBR322 L-strand factor-Y effector site

is required for pBR322 lagging-strand DNA synthesis is not inaccordance with the experiments of Backman et at (27). Theyhave shown that a 580-base-pair fragment generated by FnudIIrestriction endonuclease cleavage ofpBR345 DNA (identical insequence to the corresponding region in pBR322 DNA with theexception of two base substitutions), which contains the originof leading-strand DNA synthesis but not the L-strand factor Yeffector site, is capable ofautonomous replication when ligatedto nonreplicating tetracycline-resistance genes. This suggeststhat either the FnudII fragment itself or the DNA strand of thetetracycline-resistance genes ligated to the L strand of theFnudII fragment contains a lagging-strand origin of DNA rep-lication. Further analysis is required to determine whether thefactor Y effector site on the L strand ofpBR322 is required forlagging-strand DNA synthesis.The demonstration that the factor Y effector site on the H

strand can function as a dnaB-, dnaC-, and dnaG-dependentorigin of DNA replication suggests that this site might be ca-pable of converting leading-strand DNA synthesis from a con-tinuous to a discontinuous mode as suggested. This is consistentwith the observation of Sakakibara (28) that the leading strandof ColEl DNA is synthesized in a discontinuous fashion but isdifficult to reconcile with the data of Staudenbauer et at (26),which shows that leading-strand DNA synthesis does not re-quire the dnaG protein. Furthermore, for factor Y-site-me-diated leading-strand DNA synthesis to occur, the primosomeassembled at the H-strand factorY site must be capable of mov-ing in the 3' -+ 5' direction along the template DNA, which isthe opposite of its reported directionality (11). This would sug-gest that in order for the H-strAnd factor Y site to be functionalin pBR322 DNA replication, either additional factors or featuresof the DNA sequences around this factor Y site are responsiblefor changing the directionality ofthe primosome. Alternatively,as in vivo evidence suggests, this factor Y effector site may notbe required for vegetative replication ofthe plasmids ColEl andpBR322 DNAs. Viable deletion mutants of ColEl and pBR322DNAs have been constructed in which the H-strand factor Ysites have been removed. These deletion mutants are unableto undergo conjugal transfer and have an increased copy num-

ber (29, 30). Perhaps the factor Y effector site on the H strandfunctions as an origin of DNA replication for the conversion ofthe conjugatively transferred H-strand DNA to the duplex formin the recipient cell or is involved in some way with control ofplasmid copy number, or both.

Proc. Nad Acad. Sci. USA 78 (1981)

Proc. Natd Acad. Sci. USA 78 (1981) 6115

We thank Dr. J. Hurwitz for helpful discussion and a critical readingof the manuscript. In addition, we would like to thank Ms. Pearl Nurseand Mr. David Valentin for technical assistance and Dr. Peter Modelfor advice on cloning in flR229. This work was supported by Grant 5ROl GM26410 from the National Institutes of Health and Grant JFRA-14 from the American Cancer Society to K.J.M., by Grant 2 R01GM13344 from the National Institutes of Health to J. Hurwitz, and byGrant INT 8018468 from the National Science Foundation to K.J. M.and J. Hurwitz. S.L.Z. was supported by National Institutes of HealthTraining Grant GM07491.

1. Wickner, W., Brutlag, D., Schekman, R. & Kornberg, A. (1972)Proc. NatL Acad. Sci. USA 69, 965-969.

2. Vicuna, R., Hurwitz, J., Wallace, S. & Girard, M. (1977)J. BioLChean. 252, 2524-2533.

3. Vicuna, R., Ikeda, J.-E. & Hurwitz, J. (1977)1. BiOL Chem. 252,2534-2544.

4. Bouch6, J.-P., Zechel, K. & Kornberg, A., (1975)J. BioL Chean.250, 5995-6001.

5. Bouche, J.-P., Rowen, L. & Kornberg, A., (1978)J. BioL Chean.253, 765-769.

6. Benz, E., Jr., Reinberg, D., Vicuna, R. & Hurwitz, J. (1980) J.BioL Chean. 255, 1096-1106.

7. Wickner, S. & Hurwitz, J. (1974) Proc. NatL Acad. Sci. USA 71,4120-4124.

8. Schekman, R., Weiner, J., Weiner, A. & Kornberg, A. (1975)J.BioL Chean. 250, 5859-5865.

9. Wickner, S. & Hurwitz, J. (1975) Proc. NatL Acad. Sci. USA 72,3342-3346.

10. Shlomai, J. & Kornberg, A. (1980) Proc. NatL Acad. Sci. USA 77,799-803.

11. Arai, K. & Kornberg, A. (1981) Proc. Natl Acad. Sci. USA 78,69-73.

12. Zipursky, S. L. & Marians, K. J. (1980) Proc. Natl Acad. Sci. USA77, 6521-6525.

13. Wickner, S., Wright, M., Berkower, I. & Hurwitz, J. (1974) inMethods in Molecular Biology, ed. Wicker, R. (Academic, NewYork), Vol. 7, pp. 195-219.

14. Boeke, J. E. (1981) Mol, Gen. Genet. 181, 288-291.15. Reha-Krantz, L. J. & Hurwitz, J. (1978) J. Biol Chem. 253,

4043 4049.16. Kupersztoch-Portnoy, Y., Lovett, M. & Helinski, D. (1974) Bio-

chemistry 13, 5484-5490.17. Franke, B. & Ray, D., (1980) Virology 44, 168-187.18. Zinder, N. D. & Boeke, J. E. (1981) Single-Stranded DNA

Phage, (Cold Spring Harbor Laboratories, Cold Spring Harbor,NY), 2nd Ed., in press.

19. Model, P. & Zinder, N. D. (1974) J. MoL Biol 83, 237-251.20. Rowen, L. & Kornberg, A. (1978)J. Biol Chean. 253, 758-764.21. Benton, W. D. & Davis, R. W. (1977) Science, 196, 180-182.22. Wahl, J. M., Stern, M. & Stark, G. R., (1979) Proc. NatL Acad.

Sci. USA 76, 3683-3687.23. Sutcliffe, J. G. (1978) Cold Spring Harbor Symp. Quant. BioL 43,

77-90.24. Rothstein, R. J., Lau, L. F., Bahl, C. P., Narang, S. A. & Wu,

R. (1979) Methods Enzymol, 68, 98-109.25. Nomura, N. & Ray, D. S. (1980) Proc. Nati Acad. Sci. USA 77,

6566-6570.26. Staudenbauer, W., Scherzinger, E. & Lanka, E. (1979) Mol. Gen.

Genet. 177, 113-120.27. Backman, K., Betlach, M., Boyer, H. W. & Yanofsky, S. (1979)

Cold Spring Harbor Symp. Quant. Biol 43, 69-75.28. Sakakibara, Y. (1978) 1. MoL BioL 124, 373-389.29. Twigg, A. & Sherrat, D. (1980) Nature (London) 283, 216-218.30. Warren, G., Twigg, A. & Sherratt, D. (1978) Nature (London)

274, 259-261.

Biochemistry: Zipursky and Marians