JOURNAL OF VIROLOGY, May 1986, p. 450-4580022-538X/86/050450-09$02.00/0Copyright C) 1986, American Society for Microbiology

DNA Requirements at the Bacteriophage G4 Origin ofComplementary-Strand DNA Synthesis

PAUL F. LAMBERT,t DAVID A. WARING, ROBERT D. WELLS,t AND WILLIAM S. REZNIKOFF*

Department of Biochemistry, College of Agricultural and Life Sciences, University of Wisconsin, Madison,Wisconsin 53706

Received 11 June 1985/Accepted 7 January 1986

An in vivo assay was used to define the DNA requirements at the bacteriophage G4 origin of complementary-strand DNA synthesis (G4 origin). This assay made use of an origin-cloning vector, mRZ1000, a defective M13recombinant phage deleted for its natural origin of complementary-strand DNA synthesis. The minimal DNAsequence of the G4 genome sufficient for the restoration of normal M13 growth parameters was determined tobe 139 bases long, located between positions 3868 and 4007. This G4-M13 construct was also found to give riseto proper initiation of complementary-strand synthesis in vitro. The cloned DNA sequence contains all theregions of potential secondary structure which have been implicated in primase-dependent replication initiationas well as additional sequence information. To address the role of one region which potentially forms a DNAsecondary structure, the DNA sequence internal to the G4 origin was altered by site-directed mutagenesis. A3-base insertion at the Avall site as well as a 17-base deletion between the AvaI and Avall sites both resultedin loss of origin function. The 17-base deletion was also generated within the G4 genome and found todramatically reduce the infectious growth rate of the resulting phage. These results are discussed with respectto the role of the G4 origin as the recognition site for primase-dependent replication initiation and its possiblerole in stage II replication.

Bacteriophage G4 is a single-stranded DNA phage whichinfects Escherichia coli C (11). Upon infection, the single-stranded G4 genome is converted to a double-strandedreplicative intermediate. This replication event is dependentupon the synthesis of an RNA primer by the host dnaG geneproduct, primase, at a unique position on the G4 genome, theorigin of complementary-strand DNA synthesis (3). Themechanism by which primase interacts on the G4 genome isdifferent from its mode of interaction on other genomes,such as that of bacteriophage (X174, in which a multiproteincomplex termed the primosome is required for primaseactivity (1). The primosome is also thought to be necessaryfor the synthesis of RNA primers by primase during E. colichromosomal replication as well as lambda DNA replication(9, 10). On the other hand, both in vivo and in vitro studiesindicate that primase activity on the bacteriophage G4genome does not require the primosome (2, 4, 27) but isdependent solely upon the addition of stoichiometricamounts of single-stranded DNA-binding protein (SSB). Thefeatures of the G4 origin that cause primase to interactindependently of the primosome are not understood.The DNA sequence of the G4 genome was determined,

and the position of primase-dependent RNA synthesis wasmapped to a location within the gene F-G intercistronicregion (8, 13). Sequence comparison with the primase-dependent (primosome-independent) origins of three closelyrelated, single-stranded DNA bacteriophages, ~k, a3, andst-1 (25), revealed several interesting features. First, thereare two regions of strong sequence conservation within theintercistronic regions of these phages. Second, these regions

* Corresponding author.t Present address: Laboratory of Tumor Virus Biology, National

Cancer Institute, Building 41, Bethesda, MD 20205.Present address: Department of Biochemistry, University of

Alabama at Birmingham School of Medicine and Dentistry, Univer-sity Station, Birmingham, AL 35294.

of conserved sequence can potentially form secondary struc-tures (stemloops I and II; see Fig. 2 and 8). Whether thesesequence homologies are significant with respect to originfunction or are simply a consequence of the relatednessbetween the phages has not been determined; however, thehigh degree of sequence conservation within this regioncompared with other regions of these genomes implicates afunctional role (25).

In this study, the DNA sequence requirements at the G4origin are defined. We used an in vivo assay for originactivity that measured the ability of a DNA fragment torestore normal infectious growth character to a defectiveM13 phage. Normal infectious growth was found to correlatewith proficient rifampin-resistant DNA replication assayedin vitro. The boundaries of the G4 origin as well as therequirements for regions within this DNA sequence weredetermined. In addition, the effect of destroying G4 originactivity within phage G4 was addressed.

MATERIALS AND METHODS

Bacterial strains and phage stocks. E. coli K12 strains JM101[F' traD-36 lacIqZm]5 proA+B+/l(lac-pro)supE thi] andCSH26/Pox38-gen (CSH26 is described below; Pox38-gen isan F' plasmid encoding gentamycin resistance which was

constructed as described by Johnson and Reznikoff [15]) wereused for infectious growth curves and efficiency of infectionexperiments. Strain CSH26 [ara A(lac-pro) thi] was used fortransfection experiments. E. coli C was used for infectiousgrowth of bacteriophage G4 and derivatives constructed inthis study. Bacteriophage M13mp8 was purchased fromBethesda Research Laboratories. M13AE101 was a gift fromD. Ray, University of California at Los Angeles, LosAngeles. Construction of all other phages is described below.

Biochemicals. Restriction enzymes, E. coli DNA polymer-ase I, and DNA nucleases were purchased from New Eng-land BioLabs, Bethesda Research Laboratories, Inc., or

Promega Biotec and used as specified by the supplier. T4

450

Vol. 58, No. 2

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from

G4 ORIGIN OF COMPLEMENTARY-STRAND DNA SYNTHESIS

Bam HI

FIG. 1. Origin cloning vector mRZ1000. The a-complementingregion of lacZ (indicated by arrows), as present in M13mp8, isinserted into the filled-in EcoRI site of M13AE101 at the position ofthe deleted origin of complementary-strand DNA synthesis (indi-cated by broken line). The M13 genes are indicated by Romannumerals.

DNA ligase was a gift from R. Simoni, Stanford University,Stanford, Calif. SSB was a gift from M. Cox. Deoxy- anddideoxy-nucleoside triphosphates were purchased from P-LPharmacia Biochemicals, and radioisotopes were purchasedfrom Amersham Corp.Phage constructions. The origin vector mRZ1000 was

constructed by ligating the 761-base-pair (bp) fragment ofM13mp8 containing the lac region, generated by cleavage atthe AvaII and HgaI sites and purified by polyacrylamide gelelectrophoresis, into the filled-in EcoRI site of M13AE101.The resulting DNA was transformed into competent JM101cells, and minute lac+ plaques were screened for on TYEagar containing 5-bromo-4-chloro-3-indolyl-,-D-galactosideand isopropyl-f3-D-thiogalactopyranoside (19). DNAs fromthese phage were further analyzed by extensive restrictionenzyme cleavage analysis. The resulting phage genomeorganization is illustrated in Fig. 1. The 274-bp AluI restric-tion fragment from duplex G4 DNA was blunt-end ligatedinto the filled-in EcoRI site of mRZ1000 to generatemRZ1200 (Fig. 2). This phage gave rise to normal-sizedplaques which were still lac+, owing to the generation of afusion protein which starts at a G4 translational start codonand continues in frame into the a-complementing region oflacZ. Minute lac- plaques were also found which have theopposite orientation of the G4 fragment inserted into theorigin-cloning vector (mRZ1205). The constructs mRZ1211-mRZ1213 were generated by inserting the respective restric-

tion fragment, with staggered ends filled in by the action ofE. coli DNA polymerase I, into the SmaI site in mRZ1000.Deletion endpoints at the right side of the G4 origin,mRZ1216-mRZ1220 were generated, and phage mRZ1200was linearized with HindIII and treated with S1 nuclease forvarious times as described previously (30). The DNA wassubsequently cleaved with EcoRI, and the fragments re-leased were isolated by elution from a polyacrylamide gel.The DNA fragments were then recloned into EcoRI-SmaI-cleaved mRZ1000. Endpoints were determined by bothrestriction enzyme analysis and dideoxy sequencing (16).Deletion endpoints at the left end of the G4 origin wereconstructed by an identical approach, but with Bal 31nuclease in place of S1 nuclease; the conditions describedpreviously were used (17), and mRZ1200, linearized at theEcoRI site, was used as the substrate.

Site-directed mutagenesis. Single-stranded DNA preparedfrom stocks of mRZ1200 and mRZ1205 were mixed at a 1:2molar ratio in a buffer containing 10 mM Tris (pH 7.9), 6 mMMgCl2, and 6 mM NaCl and incubated at 65°C for 45 min.After the sample was cooled to room temperature, AvaI orAvaII or a mixture of the two was added and digestion wascarried out at 37°C for 60 min. The 5' overhangs generatedby these restriction enzymes were made blunt ended (filledin) by incubation at room temperature for 30 min with DNApolymerase I and 0.1 mM deoxynucleoside triphosphates.Ligation was carried out at 14°C for 10 h in the presence of0.1 mM ATP and 1 U of T4 ligase. This was followed by theaddition of the same restriction enzyme used in the initialcleavage of the partial duplex. This step is important,because it causes the linearization of partial duplexes notaltered in their DNA sequence, thus reducing their efficiencyof transformation. Competent JM101 was transformed withthe DNA sample and plated on TYE agar (19) supplementedwith 5-bromo-4-chloro-3-indolyl-3-D-galactoside andisopropyl-,-D-thiogalactopyranoside. DNA from phageswhich gave rise to blue (lac+) plaques were screened for lossof the appropriate restriction enzyme site and sequenced.

Efficiency of infection assays. Aliquots of host bacterium(0.3 ml of fresh overnight cuiltures grown in LB medium),JM101 and CSH26(Pox38-gen), were infected with 100-,lIportions of serial dilutions of phage stocks and plated onTYE plates (19) with T top agar (19). The concentrations ofphage stocks were made equal by quantitating the concen-tration of viral DNA on agarose gels and correcting fordifferences in the DNA content. The DNA content wasdetermined by incubating 20 ,ul of phage stock containing0.1% sodium dodecyl sulfate and 0.001% bromphenol blue at65°C for 30 min, followed by electrophoresis on a 0.8%agarose gel. The gel was stained in 0.05 ,ug of ethidiumbromide per ml for 30 min and photographed under short-wave UV irradiation on Polaroid type 55 positive/negativefilm. The negative was scanned densitometrically, and theDNA content was approximated by using DNA standardsrun on the same gel. After the concentration of each phagestock was adjusted, the DNA concentrations were measuredagain and found to be equal. On the basis of densitometricscanning of duplicate samples, the error of this quantitationwas found to be less than 5%. The titrating of these phagestocks was done in triplicate; less than 10% difference in thetiters was found.

Infectious growth assays. Strain JM101 was introduced intofresh LB medium (19) and grown at 37°C. At early log-phasegrowth (optical density at 550 nm, 0.2), the cells wereinfected with 104 PFU. The production of phage was moni-tored during the rest of the cell culture growth by taking a

VOL. 58, 1986 451

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from

452 LAMBERT ET AL.

HinfI AvaIl AvolStemloopIl 1 jHinf I StnioopI

,- _-* , W _v I I lI l-l s T

I0 3900 I23800 3820 3840 3860 3880 3900 3920 3940 3960 39804000402040404060

3863

3892

3868

3852

FIG. 2. Phage constructs. The top line indicates the 274-bp AluI restriction fragment from bacteriophage G4, containing the G4 origin. Thenumbers below the line represent nucleotide positions on the G4 genome. The wavy line and arrowhead indicate the position and directionof primase-dependent RNA synthesis. Inverted complementary sequences, stemloops I and II, are indicated above the line by convergingarrows. The boxed region (positions 3992 through 4034) indicates the region homologous to the DNA origin. Relevant restriction enzymesites are indicated above the line. The portion of the 274-bp fragment present in each construct is indicated below the lines. Nucleotidesinserted in mRZ1206 and mRZ1207 are indicated above each line. The internal deletion present in mRZ1208 is indicated by the boxed area.In the far right column is indicated the number of picomoles of [3H]dTMP incorporated into a trichloroacetic acid-insoluble precipitate as

measured by the fraction II replication assay. These assays were done in triplicate in the presence of rifampin (see Materials and Methods).

0.2-ml portion at each time point, diluting it into 1.8 ml ofice-cold M9 salts (19), and filtering the phage stock througha 0.2-,um-pore-size filter to remove cells. These phage stockswere titrated. Rates of infectious growth were determined byperforming linear regression analysis on the initial timepoints (usually five points, representing the first 2.5 h post-infection).

Transfection assay. Competent F- cells (CSH26) weretransfected for 30 min on ice with 50 ng of phage DNAprepared by the method of Sanger et al. (23) and quantitatedon agarose gels by using appropriate standards. After heatshock at 42°C for 2 min, a small portion was plated in topagar containing 0.3 ml of a saturated culture of F+ cells[CSH26/Pox38-gen] and underlaid with gentamycin 1 h afterplating. This allowed the determination of the number oftransfectants for each type of phage DNA. The rest of thetransfection stock was outgrown in 4 ml of LB medium for 6h, during which time portions were removed and the phageproduced were titrated as described for the infectious growthassay. The maximal rate of progeny phage production (Table1) was determined by maximizing the value for the quotient(Yn - Yn - )/(Xn - Xn - 1) where Yn is the number of phage pertransfectant at time point n, y, 1 is the number of phage pertransfectant at the previous time point, n - 1, Xn is the time(in minutes) at time point n, and xn _ 1 is the time at theprevious time point, n - 1.

Bacteriophage G4 infectious growth assays. The burst size

for wild-type phage G4 was determined by infecting a 5-mlculture of E. coli C with 1 PFU at early log phase. Becausethe rate of progeny phage production for the G4Aori mutantwas extremely low, cells were infected with 10 PFU in thiscase. Portions were taken every 10 minutes, and the phage

TABLE 1. In vivo growth characteristics of origin-cloning vectormRZ1000Efficiency of infectiona of: Transfection

Construct JM101 CSH26(Pox38-gen) (Lmg) vRalue

M13 2.2 x 1012 3.1 x 1012Single-stranded DNA 24 2.3Double-stranded DNA 24 2.4

mRZ1000 3.1 x 1011 3.5 x 1011Single-stranded DNA 45 0.005Double-stranded DNA 38 0.030

mRZ1200 2.0 x 1012 2.3 x 1012Single-stranded DNA 24 1.6Double-stranded DNA 24 1.9a Number of PFU for an equivalent amount of phage, as indicated in

Materials and Methods.b Time in minutes at which 0.01 PFU per transfectant occurs.c Rate of progeny phage production (maximal rate is given in phage per

minute per transfectant).

(AluI)

3792

(AluI)

3792

TCGA

3792

3955GAC

4070mRZ 1200

mRZI206

mRZ1207

mRZ 1208

mRZ121

mRZ 1212

mRZ 1213

mRZ 1216

mRZ 1217

mRZ I219

mRZ 1220

mRZ 1221

mRZ 1222

mRZ 1223

3935

J. VIROL.

in vitro DNAreplication;pmoles dTMPincorporated

86

34

69

56

573792

3935 3955

4070

4070

4070

3792

3792

3935

3792

3958

3955

3792

3989

4007

4016

32

254070

47

27

n.d

n d

n d

n-d

n d

4027

4027

4027

4027

3792

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from

G4 ORIGIN OF COMPLEMENTARY-STRAND DNA SYNTHESIS

-0- a-mRZ1216~~~~~~~~~~~~~21~,~mZ1000//*-ZI6

7.0 Q51./5;-2mRZ121.0Q850L5Sa .01.

7.Fg2a tmRZ1213 7/er -Metho80

6.0 -

5.0 -

4.0-A

3.0-~~~~~~~~~~~~~~~~

2.0 ffI I A I I I I I I0.0 0.5 1.015 2.0 2. 3.0 4.0 11.0 0.0 Q5 101L5 2. 2.5 3.0 4.0 11.0

Time of infection (hours)FIG. 3. Infectious growth curves. The yield of phage as a function of time (ini hours) afte'r infection is given for the phage constructs listed

in Fig. 2. In addition, the control phage, mRZ1000, mRZ1200, and M13 are represented in panel A. See Materials and Methods for furtherdetails.

titers were determined. Phage G4 was found to produce aburst of approximately 200 to 400 phage at a time between 20and 30 min after infection, while the G4Aori mutant gave riseto only 15 progeny phage after approximately 60 min.

In vitro DNA replication assay. DNA synthesis was mea-sured by using the fraction II protein extract described byFuller et al. (9). The fraction II protein extract was preparedfrom E. coli E177 (CGSC strain no. 5029; thr-1 leuB6 thi-JthyA6 deoCI dnaAJ77 lacYl strA67 tonA21 supE44) exactlyas described, with a 28% ammonium sulfate cut. The con-centration of fraction II protein used in the DNA synthesisreactions described below was optimized by using single-stranded M13 DNA as template. The DNA synthesis reac-tions were carried out in a final volume of 20 p1l andcontained the following components: 40 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES)buffer (pH 7.6), 10 mM magnesium acetate, 40 mnM creatinephosphate, 0.5 mM CTP, GTP, and UTP, 2 mM ATP, 0.1mM dATP, dCTP, and dGTP, 0.1 mM [methyl-3H]dTTP(specific activity, 2,000, dpm/pmol of dTTP), 0.1 mg ofcreatine kinase per ml, 1 ,xg of SSB, 7% (wt/vol) polyvinylalcohol 24,000 (Sigma), and 200 ng of single-stranded DNAtemplate. Rifampin was added, when indicated, at a finalconcentration of 300 ,ug/ml. After the components werecombined at 0°C, 325 ,ug of fraction II protein was added,and the reaction mixtures were incubated at 30°C for 20 min.Incorporation of nucleotide into trichloroacetic acid-precipitable material was measured by liquid scintillationquantitation.

RESULTSConstruction of the origin-cloning vector mRZ1000. To

study G4 origin function in vivo, we constructed an origin-

cloning vector, mRZ1000, the growth of which stronglyrelies upon the introduction of a functioning origin of com-plementary-strand DNA synthesis. The design of this vectorand its use in screening DNA inserts for origin function isbased on studies of M13 mutants deleted for their origin ofcomplementary-strand DNA synthesis (16). The phage growvery poorly, giving rise to minute, turbid plaques; however,normal growth can be restored by the introduction of a DNAfragment carrying the G4 origin (21). This observation hasled to the use of these M13 mutants as vectors for theisolation of single-stranded DNA replication determinantsfrom bacterial episomes (20). For our study, we modified oneof these mutant phages, M13AE101, by introducing theregion of lacZ encoding the a chain, as present in M13mp8(28). The resulting construct, mRZ1000 (Fig. 1), has severaladvantages over the parent phage M13AE101 as a cloningvector. The presence of unique restriction enzyme siteswithin the coding region for lacZ permits the insertion ofDNA fragments and the screening of resulting clones owingto the insertional inactivation of lacZ. In addition, comple-mentarity to the universal primer permits rapid DNA se-quence determination by using the dideoxy sequencing pro-tocol (23).

Characterization ofmRZ1000 and cloned G4 origin activity.The viability of the phage was measured by infectiousgrowth assays, in which early-log-phase cultures of hostbacteria (F+ strain, JM101) were infected with phage at a lowmultiplicity of infection and the production of phage wasmonitored over the remaining period of log-phase growth(Fig. 3). The origin-cloning vector mRZ1000 grew verypoorly in comparison with wild-type M13; however, intro-duction of the 274-bp AluI restriction fragment from phageG4, containing the G4 origin of complementary-strand DNA

VOL. 58, 1986 453

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from

454 LAMBERT ET AL.

synthesis (see mRZ1200, Fig. 2), restored normal growthparameters (Fig. 3A). In addition, the efficiency of infectionfor mRZ1000, as measured in two different strains, was only11 to 14% that of wild-type M13, while the efficiency ofinfection for the G4 origin clone mRZ1200 was between 75and 90% that of wild-type M13 (Table 1).The production of progeny phage was also measured by a

single-cycle transfection assay. This assay, in which compe-tent F- cells are transfected with viral DNA and the produc-tion of progeny phage is monitored during outgrowth, pro-vides several advantages over the standard infectious growthassay. First, by introducing the viral DNA directly into thecells, potential differences in the accessibility of packagedDNA templates to the cellular replication apparatus areeliminated. This may be of importance, as it is known thatthe M13 genome is oriented specifically within the virion,such that the M13 origin of complementary-strand DNAsynthesis is located at the end of the filament which isattached to the cell surface upon adsorption (29). Also, theviral DNA does not finish entering the cell until aftercomplementary-strand DNA synthesis has begun (5). There-fore, the successful establishment of an infection could beaffected by these packaging constraints. Second, introduc-tion of the DNAs into competent cells is synchronized by theheat shock step followed by the addition of media at the startof outgrowth. This permits the determination of the lagperiod before which progeny phage appear. Third, becausethe competent cells are F-, they cannot be infected byprogeny phage (filamentous phage require the F plasmid-encoded pili for adsorption to the cell surface). Conse-quently, by determination of the frequency of transfection,the rate of progeny phage production per transfectant maybe determined.The results of such an experiment are presented in Fig. 4

(the lag times and rates of progeny phage production derivedfrom this experiment are presented in Table 1). The rate ofmRZ1000 progeny phage production was less than 1% that ofwild-type M13, and the lag time for the detection ofmRZ1000 progeny phage was 45 min, in comparison with 24min for wild-type M13. In contrast, the G4 origin constructmRZ1200 behaved very similarly to wild-type M13, having arate of progeny phage production 90% that of wild-type M13and a lag time equivalent to that of wild-type M13. In aparallel experiment, double-stranded replicative form DNAswere transfected and the production of progeny phage wasmonitored (Fig. 4). There was little difference in the behaviorof single-stranded versus double-stranded DNAs for M13and mRZ1200, whereas double-stranded mRZ1000 DNA hada shorter lag time, 38 versus 45 min, and a rate of progenyphage synthesis 6 times higher than that of single-strandedmRZ1000 DNA (Table 1). The values obtained for double-stranded mRZ1000 were, however, much lower than thevalues for the origin-competent phages M13 and mRZ1200.This observation suggests a role for the origin of comple-mentary-strand replication in stage II replication (see Dis-cussion).To measure DNA replication on the mRZ1000 template in

comparison with wild-type M13 and the G4 origin clonemRZ1200, in vitro replication assays were performed byusing the Kornberg fraction II system (9). DNA synthesisprogratnmed by single-stranded mRZ1200 DNA, as mea-sured by the incorporation of [methyl-3H]dTTP into acid-precipitable DNA (220 and 186 pmol ofdTTP incorporated inthe absence and presence of rifampin, respectively) wasmore than 5 times greater than synthesis programmed bymRZ1000 (39 and 21 pmol of dTTP incorporated in the

l000

100

I0i

04.-

4.-

-0

0.1

.0110000 6 7 809 0 -10 210 20 30 40 50 60 70 80 90 100 110 120

Period of outgrowth (minutes)FIG. 4. Transfection assay. Yields of progeny phage per

transfectant are given as a function of outgrowth period. StrainCSH26 was transformed with 20 ng of each phage DNA, and theproduction of phage was monitored by titrating portions from eachtime point on JM101 cells. Curves are labeled. See Materials andMethods for further details.

absence and presence of rifampin, respectively). DNA syn-thesis programmed by mRZ1200 was not due to RNApolymerase-tnediated primer formation, since rifampin hadlittle effect on the extent of DNA synthesis, in contrast to a6-fold reduction in DNA synthesis seen with the wild-typeM13 template (114 and 18 pmol of dTTP incorporated in theabsence and presence of rifampin, respectively). The dataindicate that the origin-cloning vector is indeed deficient inDNA replication initiation and that the efficient rate ofDNAsynthesis measured on mRZ1200 is due to rifampin-resistantreplication initiation, as expected from the primase-dependent G4 origin.

Defining the DNA requirements at the G4 origin. To define,at a gross level, the DNA requirements at the G4 origin,subfragments of the 274-bp AluI restriction fragment fromG4 were cloned into mRZ1000 and screened for originfunction. Three subclones were chosen for study: mRZ1211,which contains the left half of the 274-bp region up to theAvall sites; mRZ1212, which contains the central regionbetween the two Hinfl sites; and mRZ1213, which containsthe right half of the 274-bp region starting from the AvaI site(Fig. 2). All three subclones were deficient in origin activity,as measured by infectious growth assays (Fig. 3A) and invitro replication assays (Fig. 2).The second approach taken was to determine the bound-

aries of the G4 origin. The right boundary was defined by

J. VIROL.

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from

G4 ORIGIN OF COMPLEMENTARY-STRAND DNA SYNTHESIS

1I 0 °G4

C: 08

(D

~0

cp2 -AG4Aori9J

UA_-A- I ~~~I I I I I I __

F5510 0 1.5 20 2.5 3.0 3.5 4.0Time of infection (hours)

FIG. 5. Bacteriophage G4 production in liquid culture. E. coli Cwas infected at early log phase of growth, and the production of G4phage was monitored for wild-type G4 and G4Aori. See Materialsand Methods for details.

deleting the G4 DNA sequence from the right end of the274-bp AluI restriction fragment by using S1 nuclease as adouble-stranded DNA exonuclease (deletion constructs areshown in Fig. 2). The boundary was found to lie betweenbases 3990 and 4007 as defined by infectious growth charac-teristics (Fig. 3B) and in vitro DNA synthesis properties(Fig. 2) of the deletion constructs mRZ1216 and mRZ1217.Phage mRZ1216 was completely deficient in origin activity,whereas mRZ1217, and the deletion constructs mRZ1219and mRZ1220 are functional in G4 origin activity, equivalentto that of mRZ1200. The left boundary, defined by a similarapproach with mRZ1220 as the template for deletion forma-tion and Bal 31 nuclease for removing the G4 DNA sequencefrom the left end, was found to lie between positions 3892(deletion construct mRZ1221) and 3868 (deletion constructmRZ1222) (Fig. 2). The in vivo infectious growth assay wasused to show that there was complete loss of origin activityin mRZ1221, whereas 100% origin activity was observed inthe next deletion constructs, mRZ1222 and mRZ1223 (Fig.3B). The in vitro DNA synthesis assays were not performedon these constructs. These results indicate that the G4 originlies within the 139 bases defined by the deletion endpoints atpositions 3868 and 4007.

Site-directed mutagenesis within the G4 origin. The DNAsequence within the G4 origin was altered by site-directedmutagenesis. A partial duplex was formed between the viralDNAs of two recombinant M13 phages, mRZ1200, contain-ing the viral strand of the 274-bp AluI restriction fragment,and mRZ1205, containing the complementary strand fromthe same fragment. The partial duplex was linearized by therestriction enzyme AvaI or AvalL. Both of these enzymeshave more than one restriction site in the two M13 tem-plates; however, only one site for each enzyme is foundwithin the double-stranded region of the partial duplex. Thestaggered ends generated by cleavage with either enzymewere made double stranded by being filled in with DNApolymerase I; the linear partial duplex with blunt ends was

then ligated closed with T4 ligase. Before transformation,the partial duplexes were again subjected to cleavage by therestriction enzyme used originally. This step enriches for thedesired product altered in its DNA sequence at the restric-tion enzyme site.

This technique was very efficient in generating smallduplications at the AvaI (mRZ1206) and Avall (mRZ1207)sites within the G4 DNA sequence. Likewise, a deletionbetween these restriction sites was generated (mRZ1208;Fig. 2). These constructs were analyzed for origin activity bythe infectious growth assay (Fig. 3B) and in vitro replicationassay (Fig. 2). The AvaI duplication mRZ1206 retained someorigin activity, while the Avall duplication mRZ1207 and the17-base deletion construct mRZ1208 were completely defi-cient for origin activity.To determine whether the 17-base deletion present in

mRZ1208 results in the loss of G4 origin function in theparent phage, we repeated the site-specific mutagei esisprocedure described above, with bacteriophage G4 viralDNA in place of mRZ1200 viral DNA. The resulting G4phage DNA was transformed into E. coli C and found to giverise to small plaques. These small-plaque variants weregrown up, and replicative-form DNA was isolated. Thedestruction of the Avall site as a result of the deletion wasconfirmed by restriction analysis, indicating that the small-plaque variant is the result of the alteration of the DNAsequence within the G4 origin. In infectious growth assays,the small-plaque variant (G4Aori) grew poorly in comparisonwith wild-type bacteriophage G4 (Fig. 5), having a burst sizeof approximately 15 phage in comparison with 200 to 400phage for wild-type G4.

DISCUSSIONUse of mRZ1000 to assay origin function. We used the

origin-cloning vector mRZ1000 to define the DNA require-ments at the G4 origin. This vector was shown to be deficientin DNA replication in vitro and to grow poorly in vivo.Introduction of the 274-bp restriction fragment frombacteriophage G4, containing the G4 origin, resulted inefficient rifampin-resistant DNA replication in vitro andrestored near-normal infectious growth in vivo. The G4origin did not restore 100% of the wild-type M13 growth rate,as measured by both the infectious growth and transfectionassays, which may simply indicate that the primase-dependent G4 origin is not perfectly equipped to replace theRNA polymerase-dependent M13 origin of complementary-strand DNA synthesis. The vector mRZ1000 has been usedfor the isolation of single-stranded DNA replication initiationdeterminants from a mini-F plasmid (referred to asM13AElac in this study) (14) as well as studies of thebacteriophage G4 origin of complementary-strand DNA syn-thesis (this study; P. F. Lambert, E. Kawashima, and W. S.Reznikoff, submitted for publication).The characterization of this vector suggested an interest-

ing insight concerning the mechanism of M13 replicative-form DNA synthesis. In the transfection experiment (Fig. 4),the rate of progeny phage production seen for double-stranded mRZ1000 DNA was extremely low, similar to thatof the single-stranded form. This result strongly supports thehypothesis that the M13 RNA polymerase-dependent originof complementary-strand DNA synthesis is necessary notonly for the initial conversion of viral DNA to a double-stranded replicative intermediate (stage I replication), butalso for the subsequent amplification of this replicativeintermediate (stage II replication). This proposal is consis-tent with those made by Staudenbauer et al. (26), but is in

VOL. 58, 1986 455

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from

456 LAMBERT ET AL.

StemloopIIT A

A CG=CA-TC-G

T C - 3920G=CG-CT=A

AC AT StemloopUC TG 3980 I,

TI-/Aa 39OAAT 40 423900 C A C

C=G GEC7G CE-G

=

CEG GECG C AvailIIolGlCC=G Av- C=G mRZC=G ~~~~~A=T 1216 4000 4020

GACTCAATCATCATGACCTCGTAACGCAACAAAG ,CGTGCCTACGGAGATACTCGAGTCTCCGATACATG=EC,TACTGCAAAGCCAAAAGGACTAACATATGTTCCAGAAATIC~AT3863 mZ3880 mR 903960 ImRZ 4036

1222 122 1 1217

,1 1

\ /

FIG. 6. G4 origin region sequence. The sequence is drawn with secondary structures, stemloops I and II. Nucleotide positions on the G4genome are indicated, as is the position of the primase-dependent RNA primer (broken line). Relevant deletion endpoints and restrictionenzyme sites are shown. The boxed area is the region partially homologous to the DNA sequence within the X DNA origin.

contrast with studies on the thermosensitivity of M13 repli-cation in dnaG and dnaB strains. These results led to analternative proposal for the mechanism of complementary-strand DNA synthesis during stage II replication, involvingdiscontinuous synthesis (22). The proposed utilization of theM13 origin of complementary-strand synthesis in stage IIreplication is also consistent with the models for DNAreplication for bacteriophage Fl, 4)X174, and G4 (6, 7, 12).One caveat to this interpretation of the results of the

transfection assay (Fig. 3) is that, while it permits thedetermination of sequences that are essential for normalgrowth of mRZ1000, it cannot distinguish whether the func-tion being rescued is merely a replicative function (ourhypothesis), or whether there is some other additional func-tion which may have been impaired by the original deletionin this M13 mutant. Although we have no evidence for theexistence of an impaired function in addition to the origin ofcomplementary-strand synthesis, we cannot preclude thispossibility. Several facts, however, are consistent with theorigin of complementary strand synthesis as the only defec-tive function in this M13 mutant. First, the defective infec-tious-growth phenotype found with mRZ1000 and its parentdeletion mutant M13AE101 is rescued by the introduction ofDNA sequences from heterologous sources, including plas-mids as well as bacteriophages (14, 20, 21). In each case, theDNA sequence was found to contain a single-stranded DNAreplication initiation determinant. It seems highly implausi-ble that in every case a second function, hypotheticallyimpaired in the M13 mutant, has been fortuitously suppliedin addition to the replication functions. Second, in our study,in all cases in which in vitro rifampin-resistant DNA repli-cation assays were performed (Fig. 2), mRZ1200 derivativesdefective in vivo were also found to be deficient in DNAreplication in vitro. We did not find mutations which wereproficient in DNA replication but defective in growth, nor,vice versa, a phenotype which could be expected were a

second function encoded on the cloned G4 DNA sequence.These arguments are consistent with there being one func-tion impaired in the M13 mutant, mRZ1000, and one functionencoded for by the G4 DNA sequence; that is, the origin ofcomplementary-strand synthesis.DNA requirements at the G4 origin. In this study, the

minimal region of the bacteriophage G4 genome sufficient fororigin activity was determined to be 139 bp long. The DNAsequence includes all regions conserved among the origins ofcomplementary-strand synthesis in the single-stranded DNAbacteriophages 4k, x3, st-1, and G4.

Just to the right of stemloop I, there is a 42-base-long A+Trich region which shares 70% homology with the origin ofreplication for bacteriophage lambda (Fig. 6). Since thebacteriophage lambda origin is also primase dependent (16),it was postulated that this region of homology may be the sitefor primase binding. In addition, the location of this homol-ogous sequence just upstream of the RNA primer site in theG4 origin also supports the concept of its being a bindingsite. The deletion endpoint in mRZ1217, however, containsless than half of this region of homology, and yet themRZ1217 origin is fully functional in vivo. It is not clearwhether the remaining portion of homology is important, asthe next deletion endpoint, in mRZ1216, removes not onlythis region but also part of stemloop I. That at least part ofthis region homologous to the bacteriophage lambda origin isnot necessary for origin function may not be surprising,since the mechanism of primase action at these origins isquite different. McMacken et al. (18) have suggested thatprimase binding to the lambda origin is primosome depen-dent at positions where the assembly of the primosome isdirected by lambda 0 and P proteins. This is quite differentfrom the primosome-independent binding by primase at theG4 origin. A significant aspect of this region of homologymay be its A + T richness (21 of 26 bases conserved betweenthese origins were A or T).

J. VIROL.

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from

G4 ORIGIN OF COMPLEMENTARY-STRAND DNA SYNTHESIS

The minimal origin sequence includes regions of potentialsecondary structure, stemloops I and II, present in the geneF-G intercistronic region (Fig. 6). Neither region of second-ary structure alone, however, was sufficient for origin func-tion. Of particular interest are the subclones mRZ1212 andmRZ1213, which contain stemloop II and stemloop I, re-spectively. The infectious growth related of these constructswere no higher than that of the vector mRZ1000. DNA inaddition to these regions of potential secondary structuremust be necessary, based upon the origin-deficient characterof deletion endpoint mRZ1221. This deletion contains G4DNA sequence information up to and including the last baseinvolved in stemloop II structure. The additional G4 DNAsequence present to the left of stemloop II in deletionendpoint mRZ1222 includes part of the direct repeat indi-cated in Fig. 6. This direct repeat is highly conserved in theorigin sequences of the other primase-dependent single-stranded DNA bacteriophage (25). Furthermore, both copiesof the direct repeat are located within the stemloop regionsof the 4~K origin protected by primase from nuclease diges-tion (24). Another characteristic of this region is its highcontent of A residues, as is true of the region to the right ofstemloop I.

Site-directed mutagenesis. The formation of partial du-plexes in vitro between recombinant M13 viral DNAs allowsfor the alteration of DNA sequences at nonunique restrictionenzyme sites. In this study, small insertions and deletionswere generated within the G4 origin. The 3-base duplicationat the Avall site (mRZ1207), as well as the 17-base deletionbetween the AvaI and Avall sites, resulted in the total loss oforigin function. Both of these mutations destroy 1 bp at thebase of stemloop II and affect the spacer region betweenstemloop II and stemloop I. In contrast, the 4-base duplica-tion at the AvaI site (mRZ1206), which only changes thespacing between stemloop I and II, resulted in a slightdecrease in origin activity as measured by the in vivo and invitro assays. Therefore, the strong effect of the mutations inmRZ1207 and mRZ1208 may be due to the disruption ofsecondary structure.

Site-specific mutagenesis was also used to alter the G4origin DNA sequence within the G4 genome. This resulted ina small-plaque phenotype, which may be similar to thephenotype of M13 mutants defective in their origin ofcomplementary-strand DNA synthesis (16). That this 17-base deletion resulted in a defective phenotype in bacterio-phage G4 is consistent with the effect it had on the G4 originwhen assayed in the origin cloning vector mRZ1000. We alsonoticed that this G4 mutant reverted at a high frequency,giving rise to medium-sized plaques (data not shown). Theserevertants appear to be quite stable, although their rate ofinfectious growth is still quite poor in comparison with thatof wild-type G4. It would be interesting to determinewhether these revertants are due to a compensating changein the origin region or perhaps to the generation of a morefunctional secondary origin.

In conclusion, we have determined that the minimalsequence sufficient for G4 origin activity is no longer than139 bases. Within this minimal origin are the two segments ofDNA sequence highly conserved within the origins of thesingle-stranded DNA bacteriophages 4k, a3, st-1, and G4.Not contained within the sequence is over 50% of the regionof homology to the bacteriophage lambda origin of DNAreplication. The role of secondary structure at primase-dependent origins of complementary-strand DNA synthesishas been suggested both by in vitro studies on primaseinteraction at these origins and by the sequence conservation

within the regions of potential secondary structure. Morerecently, single-base substitutions which disrupt intrastrandbasepairing at stemloop I have been found to affect G4 originfunction in vivo (Lambert et al., submitted). The resultspresented in this study confirm the requirement of theseregions of potential secondary structure for G4 origin func-tion in vivo and indicate the need for additional G4 DNAsequence outside of these regions.

ACKNOWLEDGMENTS

We acknowledge Barbara Funnell and Ross Inman for theirassistance in the electron microscopy and Anne Griep for editorialassistance. We thank Dan Ray for supplying us bacteriophageM13AE101.

This study was supported by Public Health Service grants (num-ber GM30822 to R.D.W. and number GM19670 to W.S.R.) from theNational Institutes of Health and National Science Foundation grantnumber 08644 to R.D.W. P.F.L. was supported in part by aUniversity of Wisconsin Biochemistry Department Wharton fellow-ship. D.A.W. was supported in part by a University of Wisconsinundergraduate research grant.

LITERATURE CITED1. Arai, K., and A. Kornberg. 1981. Unique primed start of phage

4X174 DNA replication and mobility of the primosome in adirection opposite chain synthesis. Proc. Natl. Acad. Sci. USA78:69-73.

2. Benz, E., D. Reinberg, R. Vicuna, and J. Hurwitz. 1980. Initia-tion of DNA replication by the dna G protein. J. Biol. Chem.255:1096-1106.

3. Bouche, J.-P., L. Rowen, and A. Kornberg. 1978. The RNAprimer synthesized by primase to initiate G4 DNA synthesis. J.Biol. Chem. 253:765-769.

4. Bouche, J.-P., K. Zechel, and A. Kornberg. 1975. Dna G geneproduct, a rifampicin resistant RNA polymerase, initiates con-version of a single stranded coliphage DNA to its duplexreplicative form. J. Biol. Chem. 250:5995-6001.

5. Brutlag, D., R. Schekman, and A. Kornberg. 1971. A possiblerole ofRNA polymerase in the initiation of M13 DNA synthesis.Proc. Natl. Acad. Sci. USA 68:2826-2829.

6. Dressier, D., D. Howicade, K. Koths, and J. Sims. 1979. TheDNA replication cycle of the isometric phages, p. 187-214. InD. Denhardt, D. Dressler, and D. Ray (ed.), The single strandedDNA phages. Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.

7. Dumas, L., S. Goltz, and D. Benesh. 1974. Host cell DNA chaininitiation protein requirements for replication of bacteriophageG4 replicative form DNA. J. Virol. 31:370-375.

8. Fiddes, J., B. Barrell, and N. Godson. 1978. Nucleotide se-quences of the separate origins of synthesis of bacteriophage G4viral and complementary DNA strands. Proc. Natl. Acad. Sci.USA 75:1081-1085.

9. Fuller, R. S., J. M. Kaguni, and A. Kornberg. 1981. Enzymaticreplication of the origin of the E. coli chromosome. Proc. Natl.Acad. Sci. USA 78:7370-7374.

10. Furth, M. E., and S. H. Wickner. 1983. Lambda DNA replica-tion, p. 145-173. In R. Hendrix, J. Roberts, F. Stahl, and R.Weisberg (ed.), Lambda II, 2nd ed. Cold Spring Harbor Labo-ratory, Cold Spring Harbor, N.Y.

11. Godson, N. 1974. Evolution of 4~X174: isolation of four new4X-like phages and comparison with 4X174. Virology 58:272-289.

12. Horiuchi, K., and N. D. Zinder. 1976. Origin and direction ofsynthesis of bacteriophage Fl DNA. Proc. Natl. Acad. Sci.USA 73:2341-2345.

13. Hourcade, D., and D. Dressler. 1978. The site specific initiationof a DNA fragment. Proc. Natl. Acad. Sci. USA 75:1652-1656.

14. Imber, R., R. Low, and D. Ray. 1983. Identification of aprimosome assembly site in the region of the ori 2 replicationorigin of the E. coli mini-F plasmid. Proc. Natl. Acad. Sci. USA80:7132-7136.

VOL. 58, 1986 457

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from

458 LAMBERT ET AL.

15. Johnson, R. C., and W. S. Reznikoff. 1984. Copy number controlof TnS transposition. Genetics 107:9-18.

16. Kim, M. H., J. C. Hines, and D. Ray. 1981. Viable deletions ofthe M13 complementary strand origin. Proc. Natl. Acad. Sci.USA 78:6784-6788.

17. Maniatis, T., E. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

18. McMacken, R., M. Wold, J. H. LeBowitz, J. D. Roberts, J.Mallory, J. A. Wilkinson, and C. Loehrlin. 1983. Initiation ofDNA replication in vitro promoted by the bacteriophage X 0 andP replication proteins, p. 819-848. In N. Cozzarelli (ed.),Mechanisms of DNA replication and recombination. Alan R.Liss, Inc., New York.

19. Miller, J. 1972. Experiments in molecular genetics. Cold SpringHarbor Laboratory, Cold Spring Harbor, N.Y.

20. Nomura, N., R. L. Low, and D. Ray. 1982. Identification of ColEl DNA sequences that direct single strand to double strandconversion by a fX174 type mechanism. Proc. Natl. Acad. Sci.USA 79:3153-3157.

21. Ray, D., J. Cleary, J. Hines, M. H. Kim, M. Strathearn, L.Kaguni, and M. Roark. 1981. DNA initiation determinants ofbacteriophage M13 and of chimeric derivatives carrying foreignreplication determinants, p. 169-193. In D. Ray (ed.), Theinitiation ofDNA replication. Academic Press, Inc., New York.

22. Ray, D. S. 1978. In vivo replication of filamentous phage DNA,p. 325-340. In D. Denhardt, D. Dressler, and D. Ray (ed.), The

single strand DNA phages. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

23. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

24. Sims, J., and E. Benz. 1980. Initiation of DNA replication by thedna G protein: evidence that tertiary structure is involved. Proc.Natl. Acad. Sci. USA 77:900-904.

25. Sims, J., D. Capon, and D. Dressler. 1979. DNA G dependentorigins of DNA replication. J. Biol. Chem. 254:12615-12628.

26. Staudenbauer, W., B. Kessler-Liebscher, D. Schneck, B. vanDorp, and P. Hofshneider. 1978. Replication of M13 duplexDNA in vitro, p. 369-378. In D. Denhardt, D. Dressler, and D.Ray (ed.), The single strand DNA phages. Cold Spring HarborLaboratory, Cold Spring Harbor, N.Y.

27. Taketo, A. 1977. Conversion of bacteriophage G4 singlestranded DNA to double stranded replicative form in dnamutants of E. coli. Biochim. Biophys. Acta 476:149-155.

28. Vieira, J., and J. Messing. 1982. The pUC plasmids, an M13mp7derived system for insertion mutagenesis and sequencing withsynthetic universal primers. Gene 19:259-268.

29. Webster, R. E., R. A. Grant, and L. A. W. Hamilton. 1981.Orientation of the DNA in filamentous bacteriophage Fl. J.Mol. Biol. 152:357-374.

30. Yu, X.-M., and W. S. Reznikoff. 1984. Deletion analysis of theCAP-cAMP binding site of the E. coli lac promoter. NucleicAcids Res. 12:5449-5464.

J. VIROL.

at Univ of W

isconsin - Mad on M

arch 29, 2007 jvi.asm

.orgD

ownloaded from