Proc. Natl. Acad. Sci. USAVol. 90, pp. 10173-10177, November 1993Biochemistry

Bacteriophage T7 gene 2.5 protein: An essential protein forDNA replication

(DNA binding protein/recombination/T7 DNA polymerase)

YOUNG TAE KIM* AND CHARLES C. RICHARDSONDepartment of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, MA 02115

Contributed by Charles C. Richardson, July 22, 1993

ABSTRACT The product of gene 2.5 of bacteriophage T7,a single-stranded DNA binding protein, physically interacts withthe phage-encoded gene S protein (DNA polymerase) and gene4 proteins (helicase and primase) and stimulates their activities.Genetic analysis ofT7 phage defective in gene 2.5 shows that thegene 2.5 protein is essential for T7 DNA replication and growth.T7 phages that contain null mutants ofgene2.5 were constructedby homologous recombination. These gene 2.5 null mutantscontain either a deletion ofgene2.5 (T7A2.5) or an insertion intogene 2.5 and cannot grow in Escherichia coli (efficiency ofplating, <10-8). After infection of E. coli with T7A2.5, hostDNA synthesis is shut off, and phage DNA synthesis is reducedto <1% ofphage DNA synthesis in wild-type T7-infected E. colicells as measured by incorporation of [3H]thymidine. In con-trast, RNA synthesis is essentially normal in T7A2.5-infectedcells. The defects in growth and DNA replication are overcomeby wild-type gene 2.5 protein expressed from a plasmid har-boring the T7 gene 2.5.

Single-stranded DNA binding proteins (SSBs), such as Esch-erichia coli SSB and T4 gene 32 protein, are essentialcomponents of DNA metabolism in prokaryotic cells (1-3).The gene 2.5 protein of bacteriophage T7, originally isolatedbased on its strong affinity for single-stranded DNA and itsability to stimulate DNA synthesis by T7 DNA polymerase(4, 5), is thought to be analogous to these well characterizedSSBs. Like E. coli SSB and T4 gene 32 protein, gene 2.5protein has been implicated in T7 DNA replication, recom-bination, and repair (6-11). We purified gene 2.5 protein fromcells overexpressing the gene and characterized its physicalproperties and interactions with DNA (6). Gene 2.5 proteinexists as a dimer of two identical subunits of Mr 25,562. Itbinds specifically to single-stranded DNA with a stoichiom-etry of -7 nt bound per monomer of gene 2.5 protein andextends the length of the DNA molecules as measured byelectron microscopy. The binding constant of gene 2.5 pro-tein for single-stranded DNA is -2.5 x 106 M-1, as deter-mined by fluorescence quenching and nitrocellulose filterbinding assays (6). Fluorescence studies suggest that tyrosineresidue(s) on gene 2.5 protein interacts with single-strandedDNA, whereas tryptophan residues do not (6).

In T7 DNA replication, the gene 2.5 protein has thepotential to play one of several essential roles. At the T7replication fork, three proteins, the products of T7 genes 4and 5 and the host trxA gene, account for the fundamentalreactions (12, 13). Gene 5 protein is a DNA polymerase,catalyzing the polymerization of nucleotides with low pro-cessivity (14, 15). E. coli thioredoxin, the product of the trxAgene, binds to gene S protein in a 1:1 stoichiometry andconfers processivity on the polymerization reaction by in-creasing the affinity of the enzyme for a primer/template (14,

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

15). Both direct and indirect evidence support an interactionbetween these essential replication proteins and the T7 gene2.5 protein.T7 gene 2.5 protein stimulates DNA synthesis catalyzed by

the T7 DNA polymerase/thioredoxin complex on single-stranded DNA templates (4, 5, 11, 14) and increases theprocessivity ofthe reaction (11). It has been shown by affinitychromatography and fluorescence emission anisotropy thatT7 DNA polymerase and gene 2.5 protein physically interactwith a dissociation constant of 1.1 ,uM (11). Similarly, inter-action of the T7 helicase/primase with gene 2.5 protein hasbeen inferred from the ability ofgene 2.5 protein to stimulatesynthesis of primers (10, 16). The T7 helicase/primase-gene2.5 protein interaction has also been confirmed by affinitychromatography (11). We have recently found (unpublisheddata) that the C-terminal acidic domain of gene 2.5 protein isrequired for T7 growth in vivo and it participates in gene 2.5dimerization and protein-protein interactions.The role of gene 2.5 protein in recombination is not as well

understood. Sadowski et al. (17) demonstrated that extractsof T7 phage-infected cells contain an activity that promotesrenaturation of complementary single strands and suggestedthat this activity resided in the T7 SSB. Recent studies (S.Tabor and C.C.R., unpublished results) have, in fact, dem-onstrated that the gene 2.5 protein facilitates renaturation ofhomologous single-stranded DNA even more efficiently thandoes E. coli recA protein, E. coli SSB, or T4 gene 32 protein.A determination of the definitive role of the T7 gene 2.5

protein in DNA replication and recombination in vivo isdependent on genetic analysis of gene 2.5 mutants. Previ-ously, T7 phage containing mutations in gene 2.5 have beenisolated based on their inability to grow on E. coli strains thathave a defective SSB (8). These T7 mutant phages contain anamber mutation in gene 2.5 that leads to synthesis of ashortened polypeptide -90% of the length of the wild-typeprotein (9). These gene 2.5 mutant phages can grow onwild-type E. coli strains but not on strains expressing atemperature-sensitive SSB at the nonpermissive tempera-ture. Furthermore, these T7 phages are defective in recom-bination (8). Recently, however, a mutational analysis hasidentified T7 gene 2.5 mutants that cannot grow even in E.coli strains producing wild-type SSB (F. W. Studier, per-sonal communication). We show that T7 phages with adeletion ofgene 2.5 (T7A2.5) do not grow in wild-type E. coliand have no detectable T7 DNA replication; the T7A2.5phages grow normally in E. coli strains expressing wild-typegene 2.5 from a plasmid.

MATERIALS AND METHODSBacterial Strains. E. coli HMS157 (F- recB21 recC22

sbcA5 endA gal thi sup) (laboratory collection), E. coli

Abbreviations: SSB, single-stranded DNA binding protein; moi,multiplicity of infection.*Present address: Department of Microbiology, National FisheriesUniversity of Pusan, Pusan, 608-737, Korea.

10173

Dow

nloa

ded

by g

uest

on

Janu

ary

14, 2

022

10174 Biochemistry: Kim and Richardson

HMS174 (F- hsdR rK,2- mKj2+ recAl) (18), E. coli HB101[F- A(mcrCmrr) leu supE44 aral4 galK2 lacYl proA2rpsL20(Strr) xyl-5 mtl-l recA13] (19), E. coli HMS262 (F-hsdR pro leu-lac-thi-supE tonA-trxA) (laboratory collec-tion), E. coli JH21(F- pcnB80, AtrxA307) (20), and E. coliAN1(F- AtrxA307, metE::TnlO), a derivative of E. coli C600(21) with P1 transduction (22) have been described. Growthand manipulation of bacteriophage T7 and E. coli wereperformed as described (23, 24).

Construction of Gene 2.5 Plasmids. DNA fragments wereprepared and cloned by standard procedures (25). E. coliHB101 was transformed with the designated plasmids (26).Three different plasmids containing a wild-type gene 2.5 wereconstructed by a standard PCR protocol (27). To constructthe first plasmid, two primers, a 5'-end primer with an NdeI site (5'-CGTAGGATCCATATGGCTAAGAAGATTT-TCACCTC-3') and a 3'-end primer with a BamHI site (5'-CGTAGGATCCACTTAGAAGTCTCCGTC-3'), were usedto amplify T7 DNA sequences 9157-9862 containing gene 2.5coding sequence (28). PCR-generated DNA fragments wereincubated with the large fragment ofE. coliDNA polymeraseI in the presence of the four nucleoside 5'-triphosphates, andthe resulting fragments containing blunt ends were thenligated into the EcoRV site of plasmid pBR322 (29). Theresulting plasmid pGP2.5-1 contains the gene 2.5 proteincoding sequence under control of the pBR322 tetracyclinegene promoter.The second plasmid contains gene 2.5 under the control of

a T7 RNA polymerase promoter. The 700-bp Nde I/BamHIfragment containing gene 2.5 was excised from pGP2.5-1 andthen cloned into the Nde I and BamHI sites ofplasmid pT7-7,generating plasmid pGP2.5-2. Plasmid pT7-7, constructed byS. Tabor (Harvard Medical School, Boston), contains the T7RNA polymerase promoter 410 as well as a strong translationinitiation region prior to the polylinker (30).The third plasmid, pGP2.5-3, contains the T7 sequence of

the C-terminal half ofgene 2, the entire sequence ofgene 2.5,and all of gene 2.8 (T7 sequence at nt 8961-10279). PlasmidpGP2.5-3 was constructed as follows. Using the two primers5'-CGTAGGATTCTCGGAGCATTCC-3' and 5'-CGTAG-GATCCACTTAGAAGTCTCCGTC-3', a PCR-generatedfragment was created and inserted by blunt-end ligation intoplasmid pBR322 that had been linearized by BamHI digestionand followed by filing-in using the large fragment of E. coliDNA polymerase I. The resulting pGP2.5-3 clone was se-lected and shown to contain an insert of the correct size andorientation. All the clones generated from PCR-amplifiedDNA were sequenced and found to be free of mutations.

Construction of Gene 2.5 Mutants of Bacteriophage T7.Gene 2.5 mutant phages were constructed by homologousrecombination (18). To construct a gene 2.5 null mutant, weused two different plasmids as in vitro recombination sub-strates. The first plasmid, pGP2.5::trxA, inserts the E. colithioredoxin gene (trxA) into the middle of gene 2.5. Thesecond plasmid, pGP2.5T2.8, replaces gene 2.5 ofT7 with theE. coli thioredoxin gene (trxA). Plasmid pGP2.5::trxA wasconstructed as follows. E. coli trxA was amplified by PCRwith two primers, each containing an Mlu I site within itssequence (5'-CAACACGCGTGGCTTATTCCTGTG-3' and5'-CGAACGCGTTTAAGCCAGGTTAG-3'). PCR-ampli-fled DNA fragments of trxA were cloned into pGP2.5, whichhad been made linear by cleavage with Mlu I, generatingpGP2.5::trxA. The resulting plasmid expresses a truncatedgene 2.5 protein but an active thioredoxin. Plasmid pGP2T2.8was constructed as follows. Part of gene 2, 8961-9096 in thenucleotide sequence of T7 DNA, was amplified by using twoprimers containing a BamHI site within both primers (5'-CGTAGGATCCTCGGAGCATTCC-3' and 5'-GAAT-TCGTCGACGGATCCCGGGCGTATTACTTCGGTGCT-3'). Similarly, gene 2.8 (nt 9854-10279) was amplified by

using two primers with an EcoRI site within both primers(5'-CGACGAATTCTAAGTGGAACTGCGGG-3' and 5'-CGACGAATTCCCTTTAGCGCCGTAAC-3'). The PCR-generated fragment of T7 gene 2 was cloned into BamHI-linearized pBR322 to generate pGP2. pGP2 was digested byEcoRI, and the PCR-generated gene 2.8 fragment was clonedinto the EcoRI site of pGP2. The resulting plasmid is pGP2-2.8. Finally, the PCR-produced trxA DNA was cloned intopGP2-2.8 that had been digested with Sma I at the junctionof genes 2 and 2.8. The isolation ofa plasmid containing gene2-trxA-gene 2.8 was determined by PCR with suitable prim-ers. The resulting plasmid, pGP2T2.8, expresses a functionalthioredoxin. pGP2.5::trxA and pGP2T2.8 were transformedtoE. coli HMS157, generating E. coli HMS157/pGP2.5::trxAand E. coli HMS157/pGP2T2.8, respectively. E. coliHMS157/pGP2T2.8 was grown in LB medium at 37°C withvigorous shaking to A600 = 0.5 and infected [multiplicity ofinfection (moi) = 0.1] with wild-type T7 phage. After celllysis, phages were harvested as described (31). To select andamplify the recombinant (T7A2.5::trxA), HMS262 cells con-taining pGP2.5-1 that supplements wild-type gene 2.5 proteinwere grown to midexponential phase and infected with thelysate produced from wild-type T7 infection of E. coliHMS157/pGP2T2.8. After cell lysis, amplified T7A2.5::trxAwas plated onE. coli HMS262/pGP2.5-1, and the presence ofthe trxA gene and the absence of gene 2.5 in the T7 genomewere confirmed by PCR and restriction analysis of individualT7 plaques. The resulting T7 phage (T7A2.5::trxA) substi-tutes trxA for gene 2.5. Isolation of T72.5::trxA phage wascarried out by an analogous procedure with plasmidpGP2.5::trxA instead of pGP2T2.8.Measurements ofDNA and RNA Synthesis. DNA and RNA

synthesis was measured essentially as described (32). E. coliHMS174 cells were grown with shaking at 30°C in M9 CAAmedium. At 3 x 108 cells per ml, the bacteria were infectedwith the indicated T7 phage at moi = 7. At the indicatedtimes, aliquots (0.2 ml) of phage-infected cells were removedand placed in tubes containing 10 ,ul of [3H]thymidine (50,uCi/ml; 1 Ci = 37 GBq) or [3H]uridine (50 ,uCi/ml) tomeasure DNA or RNA synthesis, respectively. After 90 secof incubation at 30°C, growth was terminated by addition of3 ml of cold 5% trichloroacetic acid. The acid-insolublematerial was collected on GF/C filters (Whatman) andwashed three times with 3 ml of ice-cold 1 M HCl and twotimes with 3 ml of 95% ethanol. The acid-insoluble radioac-tivity was measured in a toluene-based solvent in a liquidscintillation counter.In Vivo Labeling of T7 DNA. E. coli HMS174 cells were

grown to 3 x 108 cells per ml in M9CAA medium and infectedwith various T7 phage at moi = 3-5. At the desired times afterinfection, aliquots (2 ml) were removed and [3H]thymidine(80 Ci/mmol) was added to the infected cells (final concen-tration, 40 Ci/ml). Labeling was stopped by addition of anequal volume of freshly made stop solution (75% ethanol/2%phenol/21 mM CH3COONa, pH 5/2 mM EDTA). The ra-dioactively labeled T7 DNA was isolated as described (33),cleaved with Xmn I, and electrophoresed through 0.8%agarose gel in TBE (12.1 g of Tris base per liter/4.1 g of boricacid per liter/0.74 g ofNa2EDTA per liter) at =0.5 V/cm. Thegels were treated with EN3HANCE (NEN), dried, andexposed to XAR Kodak film with an intensifying screen at-70°C; bands were identified by autoradiography.Other Methods. Plating efficiencies of T7 wild-type and

T7A2.5::trxA phage were measured as follows. Bacterialstrains were grown to 2 x 108 cells per ml. Various T7 phagesin LB were diluted (0.1 ml) and mixed with 0.2 ml of bacterialculture and 3 ml of top agar, plated on LB or LB/ampicillinplates, and allowed to incubate at 30°C.

Proc. Natl. Acad. Sci. USA 90 (1993)

Dow

nloa

ded

by g

uest

on

Janu

ary

14, 2

022

Proc. Natl. Acad. Sci. USA 90 (1993) 10175

RESULTSConstruction of Bacteriophage T7 Gene 2.5 Mutants and

Gene 2.5 Expression Vectors. T7 phage defective in gene 2.5were constructed by homologous recombination. As shownin Fig. 1, the plasmid pGP2T2.8 provides the flanking se-quences (gene 2 and gene 2.8) of gene 2.5 for geneticcrossover but with the replacement ofgene 2.5 by the E. colithioredoxin gene trxA. Recombinant T7 phage having gene2.5 replaced by trxA (T7A2.5::trxA) by the resulting cross-over were selected for growth on an E. coli trxA- strain (E.coli HMS262) in which wild-type gene 2.5 protein is providedfrom one ofthe three plasmids containing gene 2.5 (pGP2.5-1,pGP2.5-2, pGP2.5-3). pGP2.5-1 and pGP2.5-2 contain gene2.5 under the control of the tetracycline gene promoter ofpBR322 and the T7 RNA polymerase promoter of pT7-7,respectively. pGP2.5-3 contains gene 2.5 flanked by theproximal half of gene 2 on one side and by all of gene 2.8 onthe other; expression of gene 2.5 is under the control of itsown promoter. T7A2.5::trxA lacks the DNA sequence at nt9097-9853 on the T7 DNA molecule (28). This region containsthe entire gene 2.5 sequence as well as the promoter for gene2.5. A second T7 gene 2.5 mutant, T72.5::trxA, preparedsimilarly to T7A2.5::trxA, has the E. coli trxA gene insertedinto the coding sequence of gene 2.5.When T7A2.5::trxA phages are grown in E. coli HMS262

containing either plasmid pGP2.5-2 or pGP2.5-3, wild-typeT7 phage appears in the resulting lysate at a frequency of 10o-or 10-2, respectively (Table 1). Presumably, these wild-typeT7 phages are produced by recombination betweenT7A2.5::trxA and the complementing plasmid (pGP2.5-2 orpGP2.5-3). However, when complementation is provided bypGP2.5-1, no wild-type T7 revertants (<10-8) are detectedsince no sequence homology exists through the flankingregions of gene 2.5 or the promoter regions of T7. In the caseof T72.5::trxA, wild-type revertants arise at a frequency of%2% via homologous recombination during propagation onthe E. coli strains used for complementation.Gene 2.5 Is Essential for T7 Growth. Biochemical evidence

suggests that the gene 2.5 protein plays crucial roles inseveral aspects of T7 DNA metabolism. Consequently, wehave used the T7 gene 2.5 mutants T7A2.5::trxA andT72.5: :trxA to examine the role ofgene 2.5 protein in vivo. Asshown in Table 1, T7A2.5::trxA phage selected and amplified

pGP2T2.8 ori

BamHI EcoRI

T7WT H ~~~Gene 2 Gene 2. 8

T7WT [ Gene Gee. ee.

I Select A2.5::tx,4

T 7A2.5::trxA [::1 1 }

FIG. 1. Construction of T7A&2.5::trxA. Plasmid pGP2T2.8 is aderivative of pBR322 containing 17 gene 2-E. coli thioredoxin gene(trxA)-T7 gene 2.8 sequences between the BamHI and EcoRI sites.DNA fragments of gene 2, gene 2.8, and the E. coli trxA gene wereamplified by PCR, and pGP2T2.8 was constructed as described.During wild-type T7 infection of HMS157/pGP2T2.8, pGP2T2.8recombined homologously with 17 wild-type (WT) phage to generatethe T7A2.5::trxA phage. The presence of the trxA gene and theabsence of gene 2.5 in 17A2.5::trxA phage was confirmed by PCRand restriction analysis.

Table 1. Plating efficiencies of T7 phages on various E.coli strains

Efficiency of platingStrain/plasmid 17 (WT) T7A2.5::trxA T72.5::trxA

HMS262 trxA- 0 <10-9 <10-9HMS262/pGP2.5-1 <10-9 0.95 0.92JH21/pGP2.5-1 <10-9 0.93 0.95AN1/pGP2.5-1 <10-9 0.94 0.92AN1/pGP2.5-2* <10-9 0.91 0.93AN1/pGP2.5-3* <10-9 0.93 0.95HMS157 1 <10-8t 10-2

Efficiency of plating was calculated by dividing the number ofplaque-forming units on a given strain by the number of wild-type(WT) 17 plaque-forming units on E. coli HMS157. 17A2.5::trxA and172.5::trxA phages were selected and propagated on E. coliHMS262/pGP2.5-1 cells.*E. coli HMS262/pGP2.5-2, JH21/pGP2.5-2, HMS262/pGP2.5-3,and JH21/pGP2.5-3 cells show similar values ofefficiency ofplatingof 17A2.5::trxA and 172.5::trxA phages.tWhen T7A2.5::trxA was selected and propagated on E. coliHMS262/pGP2.5-2 and HMS262/pGP2.5-3 cells, wild-type 17phages arose at frequencies of 10-5 and 10-2, respectively.

on E. coli HMS262/pGP2.5-1 cannot grow in E. coli HMS262trxA- and E. coli HMS157; the plating efficiencies are <10-9and <10-8, respectively. Similar results were obtained whenT7A2.5: :trxA phage was plated on other trxA mutantstrains-E. coli JH21 and AN1 (data not shown). T7 phagewith an insertion of the E. coli thioredoxin gene in gene 2.5(T72.5::trxA) also cannot grow on these strains, althoughwild-type T7 revertants arise at a frequency of -2% (titeredon E. coli HMS157) as a result of homologous recombinationduring propagation (Table 1). In control experiments, bothgene 2.5 mutant T7 phages grow normally on E. coli strainsexpressing gene 2.5 from a plasmid harboring the cloned gene(Table 1). The results show that gene 2.5 is essential forgrowth of T7.DNA and RNA Synthesis in T7A2.5-Infected Cells. In an

attempt to define the biochemical lesions responsible for theinability of T7 gene 2.5 mutants to grow in E. coli cells, weexamined DNA and RNA synthesis in E. coli cells infectedwith the mutant phage. The kinetics ofDNA synthesis afterinfection of E. coli HMS174 with T7A2.5::trxA phage werecompared (Fig. 2) with those obtained with wild-type phageunder identical conditions. In wild-type T7-infected cells therate ofDNA synthesis increased rapidly %7 min after infec-tion and reached a maximum rate 20 min after infection, afterwhich the cells begin to lyse. In striking contrast, there is adecrease inDNA synthesis after infection with 17A2.5::trxA,presumably due to the shut-off of host DNA synthesis. At 20min after infection, a time when wild-type T7 DNA replica-tion is maximal, there is no detectable DNA synthesis in themutant-infected cells (Fig. 2). The defect in DNA synthesis

60

T7 WT

T7A2.5::trxA /40 pGP2.5 (WTl

O I \ X FIG. 2. Time course of DNAsynthesis in wild-type (WT) 17-

c&. ] 2\ and 17A&2.5-infected cells. Rates20 ofDNA synthesis were measured20

l / / bas described at various intervalsafter infection of E. coli HMS174cells with T7 wild-type (o) or

1T7A2.5::trx4 1)>>T7A2.5::trxA (o) phages and afterinfection of E. coli HMS174/

0 20 40 pGP2.5-lwithT7A2.5::trxAphageTimeAfterInfection (min.) (v) at a moi of 7 at 30°C.

Biochemistry: Kim and Richardson

Dow

nloa

ded

by g

uest

on

Janu

ary

14, 2

022

10176 Biochemistry: Kim and Richardson

T7A2.5/

T7 WT T7A2.5 pGP2.5WT

Min.: 5 10 15 5 10 15 5 10 15

I..

after infection with T7 wild-type phage under identical con-ditions (Fig. 4).

1J:~2-

-J- 5-6-7

_8-9

-10

- 12

1 2 3 4 5 6 7 8 9

FIG. 3. Replication of wild-type 17 and T7A2.5 phage DNA. E.coli HMS174 cells were infected with wild-type T7 phage (T7 WT) orT7A2.5::trxA (M7A2.5). E. coli HMS174/pGP2.5-1 cells were in-fected with T7A2.5::trxA phage [T7A2.5/pGP2.5(WT)]. At 5 (lanes 1,4, and 7), 10 (lanes 2, 5, and 8), and 15 (lanes 3, 6, and 9) minpostinfection, cells were labeled with [3H]thymidine for 35 sec. Thelabeled 17 DNA was isolated, digested with Xmn I, and treated asdescribed. Restriction fragments are numbered 1-12 and the con-catameric joint fragment is labeled J.

in the T7 gene 2.5 mutant-infected cells could be overcome bycomplementation with gene 2.5 protein provided from aplasmid (Fig. 2). However, the pattern ofDNA synthesis inE. coli HMS174/pGP2.5-1 infected with T7A2.5: :trxAshowed a pattern different from that of wild-type T7 phageinfection in that the onset ofDNA synthesis was delayed andreached a maximal rate at =40 min postinfection.To examine T7 DNA synthesis specifically and in more

detail, we radioactively labeled the newly synthesized DNAby pulse-labeling with [3H]thymidine after phage infection,cleaved the DNA with the restriction enzyme Xmn I, andexamined the pattern of labeling of the DNA fragments byelectrophoreses through an agarose gel (Fig. 3). E. coliHMS174 or HMS174/pGP2.5-1 cells were infected with T7wild-type or T7A&2.5::trxA phages during logarithmic-phasegrowth at a moi 5. At 10 and 15 min postinfection, all the

fragments of T7A2.5::trxA DNA derived from T7A2.5-infected E. coli HMS174/pGP2.5-1 cells (lanes 8 and 9) arelabeled as are those obtained from E. coli HMS174 infectedwith wild-type T7 phage (lanes 2 and 3). However, noradioactively labeled T7 DNA fragments were detected in E.coli HMS174 cells infected with T7A2.5 phage (lanes 5 and 6).In addition, the shut-off of host DNA synthesis in T7A2.5-infected cells visualized in this experiment accounts for thedecrease in DNA synthesis after infection. From these ex-periments, we conclude that gene 2.5 is essential for T7 DNAreplication.Although noDNA synthesis is observed in T7A2.5-infected

cells, RNA synthesis is grossly normal (Fig. 4). The timecourse ofthe rate ofRNA synthesis in T7A2.5 phage-infectedcells did not differ significantly from that observed in cells

60

c T7Wr

X~ 40

20X T7A2.5::tixA620 °

0 10 20 30 40 50

TimeAfter Infection (min.)

FIG. 4. Time course of RNAsynthesis. The rate of RNA synthe-sis was measured by pulse labelingwith [3H]uridine for 90 sec at 4-minintervals after infection of E. coliHMS174 cells at 30°C with T7 wild-type (o) or T7A2.5 phage (e) at a moi= 7.

DISCUSSIONWe have used a genetic approach to examine the role of thegene 2.5 protein of bacteriophage T7. T7 lacking gene 2.5does not grow in E. coli and is defective in DNA synthesis.Both defects are overcome by gene 2.5 protein producedfrom a plasmid containing the cloned gene 2.5. Earlier studieshad suggested that E. coli SSB could substitute for the T7gene 2.5 protein in vivo (8). However, the gene 2.5 mutantused in those studies produced a shortened polypeptide thatcontained a significant portion of the intact gene 2.5 protein(9). Our results show clearly that E. coli SSB cannot substi-tute for gene 2.5 protein in vivo.The multiple roles of SSBs in DNA metabolism have made

it difficult to define precisely their essential nature in vivo. InE. coli, for example, SSB plays important roles in DNAreplication as well as in DNA recombination and repair (1-3,34). Strains carrying mutations in the structural gene for SSB,ssb, are temperature sensitive for growth andDNA synthesis(35) and are defective inDNA repair and recombination (36).T7 gene 2.5 protein and E. coli SSB have extensive sequencehomology; 24% of the amino acids are identical and another47% represent conserved changes (37). In bacteriophage T4,the roles ofgene 32 protein have also been well characterizedboth in vitro and in vivo. Genetic studies have demonstratedthat the product of gene 32 is essential for DNA replication,recombination, and repair (38-42). It binds cooperatively tosingle-stranded DNA, affects denaturation and renaturationofDNA, and interacts specifically with T4 DNA polymeraseand the recombination proteins uvsX and uvsY (43-46).Gene 2.5 protein shares some of structural features with

other SSBs (1). These proteins possess DNA binding do-mains and acidic C-terminal domains that are presumed to beinvolved in protein-protein interactions (47). Studies of thestructural features and binding properties ofgene 2.5 proteinindicate that gene 2.5 protein contains two distinctive func-tional domains: an N-terminal single-stranded DNA bindingdomain and a C-terminal acidic domain (11). Tyrosine resi-dues in the DNA binding domain are required for gene 2.5protein binding to single-stranded DNA (11). Our recentstudies (unpublished data) show that the C-terminal acidicdomain of gene 2.5 protein is essential for gene 2.5 proteinfunction in vivo. A truncated form of the gene 2.5 proteinlacking the C-terminal 21-amino acid residues cannot supportthe growth of T7 phage lacking gene 2.5. In addition, thepurified truncated gene 2.5 protein can no longer form dimersnor can it physically interact with T7 DNA polymerase;however, its ability to bind to single-stranded DNA is notaffected.Why is gene 2.5 protein essential for T7 growth and

replication? The essential nature ofthe protein does not arisesolely from its ability to bind to single-stranded DNA since E.coli SSB, a protein that has an even higher affinity forsingle-stranded DNA, cannot substitute for gene 2.5 proteinin vivo. Rather, we postulate that gene 2.5 protein is essentialfor two reasons: (i) its specific interaction with other phage-encoded replication proteins, and (ii) its role in recombina-tion. Biochemical studies on the purified gene 2.5 proteinhave demonstrated its physical interaction with both T7 DNApolymerase and T7 helicase/primase (11) and a stimulation ofthe activities of these proteins (4, 5, 10, 11, 14, 16). Preciselywhy these specific interactions occur in 17, however, is notknown since other SSBs can often provide similar degrees ofstimulation. One exception is the rather specific stimulationof primer synthesis by the gene 2.5 protein (10, 16). Presum-ably, the specific interaction of these proteins reflects arequirement for a highly ordered structure at the replication

Proc. Natl. Acad Sci. USA 90 (1993)

Dow

nloa

ded

by g

uest

on

Janu

ary

14, 2

022

Proc. Natl. Acad. Sci. USA 90 (1993) 10177

fork to coordinate leading- and lagging-strand DNA synthe-sis.Gene 2.5 is required not only for DNA synthesis but also

for T7 recombination (8, 9). In this process, it undoubtedlymust also interact with other recombination and replicationproteins, such as T7 DNA polymerase and primase/helicase.However, biochemical studies support a more direct role of17 gene 2.5 protein in recombination. Upon infection of E.coli with 17 phage, a single-stranded DNA renaturationactivity is induced, suggesting that 17 gene 2.5 protein isinvolved in this activity (17). It has, in fact, recently beenshown that gene 2.5 protein facilitates the renaturation ofsingle-stranded DNA much more efficiently that does E. coliSSB, recA, or T4 gene 32 protein (S. Tabor and C.C.R.,unpublished results). Such an activity is likely to be impor-tant for catalysis of homologous pairing and may provide anexplanation for the essential role of the protein in vivo inrecombination (8, 9).By what mechanism would a defect in recombination lead

to the severe defect in replication observed in cells infectedwith the gene 2.S 17 mutant? One possibility is that, in the

absence of gene 2.5 protein and the catalysis of homologouspairing, recombination intermediates accumulate and aresubsequently degraded. On the other hand, it is quite likelythat recombination plays an essential role in T7 DNA repli-cation. The genome of bacteriophage T7 is replicated throughexclusively linear intermediates (concatemers) (48, 49). For-mation of concatemers provides an intermediate by which the3' ends of the 17 chromosome can be replicated. Concate-mers can then be processed to yield fully replicated duplexchromosomes (50, 51). All evidence indicates that concate-mers are formed in vivo by annealing single-stranded termi-nally redundant ends of replication intermediates (28). Wepropose that the gene 2.5 protein is required for rapid andefficient homologous pairing of the exposed terminal redun-dancy in phage-infected cells. In the absence of concatemerformation, subsequent DNA synthesis may well be defective,terminating after replication of the parental DNA. This oneround of DNA synthesis would represent only 1% or less ofthe total DNA synthesis usually observed in phage-infectedcells. In this regard, concatemer formation in phage T7-infected cells requires recombination (52, 53).

We thank Stanley Tabor for helpful discussions throughout thisproject and Qingyun Liu for assisting with the design ofT7A2.5::trxAand for his helpful advice. This work was supported by grants fromthe Department of Energy (DEG02-88ER60688), the U.S. PublicHealth Service (AI-06045), and the American Cancer Society (NP-1U).

1. Chase, J. W. & Williams, K. R. (1986) Annu. Rev. Biochem. 55,103-136.

2. Lohman, T. M. & Bujalowski, W. (1990) in The Biology of Non-spec rc DNA-Protein Interactions, ed. Revzen, A. (CRC, BocaRaton, FL), pp. 131-171.

3. Kowalczykowski, S. C. (1991) Annu. Rev. Biophys. Chem. 20,539-575.

4. Scherzinger, E., Litfin, F. & Jost, E. (1973) Mol. Gen. Genet. 123,247-262.

5. Reuben, R. C. & Gefter, M. L. (1974) J. Biol. Chem. 249, 3843-3850.

6. Kim, Y. T., Tabor, S., Bortner, C., Griffith, J. D. & Richardson,C. C. (1992) J. Biol. Chem. 267, 15022-15031.

7. Reuben, R. C. & Gefter, M. L. (1973) Proc. Natl. Acad. Sci. USA70, 1846-1850.

8. Araki, H. & Ogawa, H. (1981) Mol. Gen. Genet. 183, 66-73.9. Araki, H. & Ogawa, H. (1981) Virology 111, 509-515.

10. Nakai, H. & Richardson, C. C. (1988) J. Biol. Chem. 263, 9831-9839.

11. Kim, Y. T., Tabor, S., Churchich, J. E. & Richardson, C. C. (1992)J. Biol. Chem. 267, 15032-15040.

12. Richardson, C. C. (1983) Cell 33, 315-317.

13. Huber, H. E., Bernstein, J., Nakai, H., Tabor, S. & Richardson,C. C. (1988) Cancer Cells 6, 11-17.

14. Tabor, S., Huber, H. E. & Richardson, C. C. (1987) J. Biol. Chem.262, 16212-16223.

15. Huber, H. E., Tabor, S. & Richardson, C. C. (1987) J. Biol. Chem.262, 16224-16232.

16. Mendelman, L. V. & Richardson, C. C. (1991) J. Biol. Chem. 266,23240-23250.

17. Sadowski, P. D., Bradley, W., Lee, D. & Roberts, L. (1980) inMolecular Mechanisms of Recombination and Genetic Recombi-nation, eds. Alberts, B. & Cox, C. F. (Academic, New York), pp.941-952.

18. Campbell, J. L., Richardson, C. C. & Studier, F. W. (1978) Proc.Natl. Acad. Sci. USA 75, 2276-2280.

19. Boyer, H. W. & Roulland-Dussoix, D. (1969) J. Mol. Biol. 104,459-465.

20. Himawan, J. & Richardson, C. C. (1992) Proc. Natl. Acad. Sci.USA 89, 9774-9778.

21. Young, R. A. & Davies, R. W. (1983) Proc. Natl. Acad. Sci. USA80, 1194-1198.

22. Miller, J. H. (1992) Experiments in Molecular Genetics (Cold SpringHarbor Lab. Press, Plainview, NY), 2nd Ed.

23. Studier, F. W. (1969) Virology 39, 562-574.24. Studier, F. W. (1975) J. Mol. Biol. 94, 283-295.25. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular

Cloning: A Laboratory Manual (Cold Spring Harbor Lab. Press,Plainview, NY).

26. Hanahan, D. (1985) in DNA Cloning, ed. Glover, D. M. (IRL,Oxford), pp. 109-135.

27. Innis, M. A. & Gelfand, D. H. (1990) in PCR Protocols:A Guide toMethods and Applications, eds. Innis, M. A., Gelfand, D. H.,Sninsky, J. J. & White, T. J. (Academic, San Diego), pp. 3-12.

28. Dunn, J. J. & Studier, F. W. (1983) J. Mol. Biol. 166, 477-535.29. Bolivar, F., Rodriguez, R. L., Greene, P. J., Betlach, M. C., Hey-

neker, H. L., Boyer, H. W., Crosa, J. H. & Falkow, S. (1977) Gene2, 95-113.

30. Tabor, S. (1990) in Current Protocols in Molecular Biology, eds.Ausubel, F. A., Brent, R., Kingston, R. E., Moore, D. D., Sei-dman, J. G., Smith, J. A. & Struhl, K. (Wiley, New York), pp.16.2.1-16.2.11.

31. Richardson, C. C. (1966) J. Mol. Biol. 15, 49-61.32. Saito, H. & Richardson, C. C. (1981) J. Virol. 37, 343-351.33. Rabkin, S. & Richardson, C. C. (1990) J. Mol. Biol. 204, 903-916.34. Meyer, R. R. & Laime, P. S. (1990) Microbiol. Rev. 54, 342-380.35. Meyer, R. R., Glassberg, J. & Kornberg, A. (1979) Proc. Natl.

Acad. Sci. USA 76,1702-1705.36. Glassberg, J., Meyer, R. R. & Kornberg, A. (1979)J. Bacteriol. 140,

14-19.37. Argos, P., Tucker, T. A. & Philipson, L. (1986) Virology 149,

208-216.38. Alberts, B. M. & Frey, L. (1970) Nature (London) 227, 1313-1318.39. Kozinski, A. & Felgenhauer, Z. Z. (1967) J. Virol. 1, 1193-1202.40. Epstein, R. H., Bolle, A., Steinberg, C., Kellenberger, C., Boy de

la Tour, E., Chevalley, R., Edgar, R. S., Susman, M., Denhardt,G. H. & Lielaulis, A. (1963) Cold Spring Harbor Symp. Quant. Biol.28, 375-394.

41. Tomizawa, J., Araku, N. & Iwama, Y. (1966) J. Mol. Biol. 21,247-255.

42. Wu, J.-R. & Yeh, Y.-C. (1973) J. Virol. 12, 758-765.43. Formosa, T., Burke, R. L. & Alberts, B. M. (1983) Proc. Natl.

Acad. Sci. USA 80, 2442-2446.44. Huberman, J. A., Kornberg, A. & Alberts, B. M. (1971) J. Mol.

Biol. 62, 39-52.45. Delius, H., Mantel, N. & Alberts, B. (1972) J. Mol. Biol. 67,

341-350.46. Wackernagel, W. & Radding, C. M. (1974) Proc. Natl. Acad. Sci.

USA 71, 431-435.47. Burke, R. L., Alberts, B. M. & Hosoda, J. (1980) J. Biol. Chem.

255, 11484-11493.48. Kelly, T. G., Jr., & Thomas, C. A., Jr. (1969) J. Mol. Biol. 44,

459-475.49. Wolfson, J.. Dressler, D. & Magasin, M. (1972) Proc. Natl. Acad.

Sci. USA 69, 499-594.50. Watson, J. D. (1972) Nat. (New Biol.) 239, 197-201.51. Serwer, P., Greenhaw, G. A. & Allen, J. L. (1982) Virology 76,

263-285.52. Gilbert, W. & Dressler, D. (1968) Cold Spring Harbor Symp. Quant.

Biol. 33, 173-184.53. Mosig, G. (1970) Advan. Genet. 15, 1-53.

Biochemistry: Kim and Richardson

Dow

nloa

ded

by g

uest

on

Janu

ary

14, 2

022