Effects of the Bacteriophage T4 ddu Protein on DNA Synthesis ...

8 downloads 0 Views 2MB Size Report
effect by speeding up the rate of replication fork move- ment; in this respect, .... terminated single-stranded chain at the end of a linear DNA molecule, which will ...
Vol. 259, No. 20, Issue of October 25, pp. 12933-12938.1984 Printed in U.S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1984 by The American Society of Biological Chemists, Inc

Effects of the Bacteriophage T4 ddu Protein on DNA Synthesis Catalyzed by Purified T4 Replication Proteins* (Received for publication, March 24, 1984)

C. Victor Jongeneela, Patricia BedingerG, and Bruce M. Alberts From the Department of Biochemistry and Biophysics, University of California, San Francisco, Sun Francisco, California 94143

The T4 bacteriophage dda protein is a DNA-dependent ATPase and DNA helicase that is theproduct of an apparently nonessential T4 gene. We have examined its effects on in vitro DNA synthesis catalyzed by a purified, multienzyme T4 DNA replication system. When DNA synthesis is catalyzed by the T4 DNA polymerase on a single-stranded DNA template, the addition of the dda protein is without effect whether or not other replication proteins are present. In contrast, on a double-stranded DNA template, where a mixture of the DNA polymerase, its accessory proteins, and the gene 32 protein is required, the dda protein greatly stimulates DNA synthesis. The dda protein exerts this effect by speeding up the rateof replication fork movement; in this respect, it actsidentically with the other DNA helicase in the T4replication system, the T4gene 41 protein. However, whereas a 41 protein molecule remains bound to the same replication fork for a prolonged period, the dda protein seems to be continually dissociating from the replication fork andrebinding to it as the forkmoves. Some gene 32 protein is required to observe DNA synthesis on a double-stranded DNA template, even in the presence of the dda protein. However, there is a direct competition between this helix-destabilizing dda proteinfor binding to singleproteinandthe stranded DNA, causing the rate of replication fork movement to decrease at a high ratio of gene 32 protein to dda protein. As shown elsewhere, the dda protein becomes absolutely required for in vitro DNA synthesis when E. coli RNA polymerase molecules are bound to the DNA template, because these molecules otherwise stop fork movement (Bedinger, P., Hochstrasser, M., Jongeneel, C. V., and Alberts, B. M. (1983)Cell 34, 115-123).

The replication of double-stranded DNA requires that the hydrogen bonds that hold together the strands of the DNA double helix be broken,so that thebases become available for pairing to incoming nucleotides. DNA helicases are enzymes that use the energy of nucleoside triphosphate hydrolysis to catalyze the opening of the DNA helix (1, 2). However, in many cases, their suspected role in DNA replication has yet to be demonstrated conclusively. The Escherichia coli rep helicase was isolated as a required

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 4 Present address, Swiss Institute for Experimental Cancer Research, CH-1066 Epalinges, Switzerland. Present address, Department of Biological Sciences, Stanford University, Stanford, CA 94305.

protein factor for the replication of single-stranded DNA phages (3). In addition, uninfected rep mutant E. coli cells have been reported to have slower than normal rates of replication fork movement (4). The E. coli helicase I1 is the product of the uvrD gene (5); it can stimulate the E. coli polymerase I11 holoenzyme to catalyze strand displacement synthesis in vitro, albeit with a rather low efficiency (6). It has also been shown that antibodies against helicase I1 can inhibit the DNA synthesis that occurs in cells plasmolyzed on cellophane disks (7). Both of these results suggest a role for helicase I1 in E. coli DNA replication. The gene 4 protein of bacteriophage T7 has both a DNA helicase and an RNA primase activity and it is an essential component of the T7 DNA replication system (8). The homologous protein in the bacteriophage T4 system is the gene 41 protein, which requires the presence of another T4 protein (the product of gene 61)for RNA primer synthesis (9,lO). In contrast to the E. coli helicases, both the T7 gene 4 protein and the T4 gene 41 protein are able to denature only short stretches of double-stranded DNA when tested in a standard DNA helicase assay. In addition to the gene 41 protein, bacteriophage T4 codes for a second DNA helicase, the product of the nonessential dda gene (11-14). The dda protein binds strongly to the T4 helix-destabilizing protein (gene 32 protein), aswell as to the T4 UVSXrecombination protein (15,16), and it has enzymatic properties that resemble those of E. coli helicase I1 (14, 17). In order to clarify the possible role of the dda protein in the replication of phage T4 DNA, we have examined its effects on the multienzyme in uitro DNA replication system in use in our laboratory (18, 19). We have shown previously that an RNA polymerase molecule bound to the DNA template can block fork movement completely and that thisblock is overcome by the addition of small amounts of the dda protein (20). In thiscommunication, we document the effects of the dda protein on the DNA synthesis catalyzed by the T4 replication proteins inthe absence of such blocking agents. MATERIALS AND METHODS

Enzymes-Bacteriophage fd gene 2 protein was the kind gift of G. P. Dotto and N. Zinder (Rockefeller University); it had been purified from an E. coli strain carrying the pD2 plasmid, which overproduces this protein (21). The products of T4 genes 43,44/62,45,and 32 were purified in our laboratory according to published procedures (22,23). The T4dda helicase was purified according to theprocedure described in the preceding paper (14). Some of the dda protein preparations used in these experiments contained small amounts of gene 32 protein, which were accounted for inthe description of the results. Otherwise, all of our enzyme preparations were essentially homogeneous and free of detectable contaminating nucleases, as judged by the sensitive assays described in Ref. 23. DNAs-The poIy(dA)and oligo(dT),*homopolymers were purchased from P-L Biochemicals. Phage T7 DNA was extracted from purified viral particles by detergent lysis and phenol extraction. The

12933

T4 DNA Helicase and DNA Replication

12934

pJMCllO plasmid contains 393 nucleotides from the phage M13 DNA replication origin inserted into pBR322 (24) and it was originally obtained from D. Ray (UCLA). The plasmid DNA was purified by standard procedures from the supernatant of a cell lysate from which protein and chromosomal DNA had been co-precipitated with detergent under alkaline conditions. To obtain circular double-stranded pJMCll0 DNA that is specifically nicked at thegene 2 protein recognition site, 20 pg ofsupercoiled DNA was mixed with 1 ng of gene 2 protein in 200 pl of 20 mM TrisHC1 (pH 8.5), 80 mM KCl, 2 mM MgC12, 1 mM 2-mercaptoethanol, 5% (v/v) glycerol, and incubated for 30 min at 30 “C. The gene 2 protein was removed by phenol extraction and residual phenol was removed from the DNA with ether. The above conditions were chosen to maximize the formation of nicked DNA rather than covalently closed relaxed DNA (25) and they resulted in more than 90% conversion of the DNA supercoils to these two DNA forms. DNA Replication Reactions-Unless otherwise indicated, in vitro DNA synthesis was carried out in the presence of 33 mM Tris/acetate (pH 7.8), 66 mM potassium acetate, 10 mM magnesium acetate, 0.5 mM dithiothreitol, 0.5 mM ATP, 0.15 mM concentration each of dATP, dGTP, dTTP, and [cY-~’P]~CTP, plus 100 pg/ml nuclease-free human serum albumin as a protein carrier. In our standard core replication reactions, the T4 replication proteins were present at the followingconcentrations: 2 pg/ml gene 43 protein (DNA polymerase), 20 pg/ml gene 44/62 protein, 18 pg/ml gene 45 protein, and gene 32 protein at theconcentration indicated. Synchronization of the replication reactions was achieved by preincubating in the absence of dCTP and then allowing extensive DNA synthesis to begin by the addition of dCTP. The products of the reaction were analyzed by separating them by electrophoresis through a 0.5% agarose gel runin 30 mM NaOH, 1 mM Na3EDTA and subjecting the dried gel to autoradiography (these procedures are described in detail elsewhere (20)).

RESULTS

*

poly(&) holopolymer primed with oligo(dT) in an attempt to minimize possible effects of template secondary structure. The results of such an experiment are presented in Fig. 1. The stimulatory effects of both the 32 protein and the44/62 and 45 proteins on DNA synthesis are clearly seen (a larger scale has been used in the right-hand panel to plot synthesis in the presence of the accessory proteins). However, the dda protein has no detectable effect on DNA synthesis in any of the reactions. This result (as well as other data not shown) indicates that thedda protein does not affect the DNA polymerase in reactions on a single-stranded DNA template. The dda ProteinStimulates the Exonuclease Activity of the T4 DNAPolymerase-In the absence of deoxyribonucleoside triphosphates, the T4DNA polymerase displays a very potent 3‘-5’ exonuclease activity, which is used for proofreading incorporationerrorsduring DNA synthesis (32, 33). This exonuclease activity is stimulated by both the polymerase accessory proteins and thegene 32 protein (34,35) and itcan be conveniently measured by following the decrease in the size of a linear double-strandedDNA fragment labeled at its 5’ ends. In Fig. 2, we present the data from an experiment in which the dda protein was added to such an exonuclease assay. It can be seen that the dda protein stimulates the exonuclease activity of the DNA polymerase if, and only if, ATP is present. The dda protein has been tested and shown to have no exonuclease activity by itself (data not shown). The data in Fig. 2 are readily explained if we realize that the exonuclease activity of the polymerase will expose a 5‘terminated single-stranded chain at theend of a linear DNA molecule, which will allowthe dda protein to bind and initiate

The dda ProteinDoes Not Affectthe DNA Synthesis Catalyzed by the T4 DNA Polymerase on a Single-stranded DNA Template-The products of T4 genes 44,62, and 45 constitute the “polymerase accessory proteins”; together they increase both the rate and the processivity of the T4 DNA polymerase (the product of gene 43). Their DNA-dependent ATPase activity is required for this stimulatory effect (26 to 28). The accessory proteins are especially helpful in allowing the polymerase to proceed past regions of secondary structure in the DNA template strand (28). Therefore, they were at one time considered as candidates for a possible DNA helicase activity. However, the experimental evidence reveals that theaccessory proteins instead help to hold the polymerase at the templateprimer junction (the growing 3’ end of the newly synthesized strand) andthey are now viewed as a “sliding clamp” that is dependent onthe occasional hydrolysis of ATP for “resetting” (28, 29). The gene 32 protein, the T4 helix-destabilizing protein, binds specifically and cooperatively to single-stranded DNA, thereby lowering the melting temperature of DNA helices. It Reaction t i m e b e e ) also holds DNA single strands in a preferred conformation FIG. 1. DNA synthesis on a single-stranded poly(dA):oligofor the DNA polymerase and thus increases the rate and (dT) template with and without the dda protein. Incorporation processivity of DNA synthesis (28, 30). At sufficiently high of [CY-~’P]~TTP into acid-insoluble an form was monitored under the concentrations, the gene 32 protein will stimulate strand following conditions. Each 20-pl reaction contained standard replidisplacement synthesis by the polymerase and its accessory cation buffer (see “Materials and Methods”) and 10 pg/ml ~~~: (1:lmolar ratio), 1 mM ATP, 166 p~ [oI-~’P] proteins, thus partially bypassing the need for a DNA helicase p o l ~ ( d A )oligo(dT)12 dTTP, 0.1 mg/ml creatine phosphokinase, and 20 mM creatine phos(18,31). This presumably reflects a 32 protein-induced melt- phate. After transfer to 37 “C, the replication proteins in various ing of the double-stranded DNA template at the fork that combinations were added to start the reaction, using the following converts it to 32 protein-bound single strands. final concentrations: 0.1 pg/ml 43 protein, 7.5 pg/ml 44/62 protein, In order to determine whether the dda protein has any 10 Gg/ml 45 protein, 65 pg/ml 32 protein, and 2 pg/ml dda protein. effect on the in vitro DNA synthesis catalyzed by T4 repli- Aliquots of 4 p1 were removed every 30 s for acid precipitation. The cation proteins, we measured rates of DNA synthesis cata- left panel shows the results obtained without polymerase accessory proteins present, while all of the reactions in the rightpanel contained lyzed by the T4 DNA polymerase in the presence of combi- these proteins (note the 5-fold difference in the ordinate scale between nations of the polymerase accessory proteins and the helix- the two panels). Squares, no additions; triangles, plus dda protein; destabilizing protein, with and without the dda protein pres- inverted triangles, plus 32 protein; diamonds, plus dda protein and 32 ent. As a template for the initial reactions, we used a long protein.

T4 DNA Helicase and DNA Replication

-

+

polymerase ddo +ATP dda. ATP (no polymerase) polymerase only polymerase ddo ATP undigeaced

“ “

”””_

-

...............

+

-

12935

I lJ B 4

0

L

0

a L

0 0

?.

a x U

lJ 0

a DNA size (base pairs)

FIG.2. Effect of the dda protein on the 3’-5’ exonuclease activity of theT4 DNA polymerase. Exonuclease reactions were carried out under the same conditions used for replication reactions (see “Materials and Methods”) in the presence of 2 pg/ml DNA polymerase, except that the four deoxyribonucleoside triphosphates were omitted. The DNA substrate (2 pg/ml) was plasmid pBR322 DNA linearized by treatment with the restriction enzyme EcoRI and treated with polynucleotide kinase to label the 5’ phosphate ends with 32P.In indicated reactions, ddu protein was present at 3 pg/ml. Reactions were started by the addition of DNA, incubated at 37 “C for 8 min, and stopped by the addition of sodium dodecyl sulfate to a final concentration of 1%.The DNA product size was then analyzed by 1%agarose gel electrophoresis in TEAbuffer (40mM Tris/acetate, pH 8.1, 20 mM sodium acetate, 2 m M Na,EDTA). After electrophoresis, the gel was dried onto filter paper and exposed for autoradiography. Microdensitometer tracings of the resulting autoradiogram are shown, demonstrating an ATP-dependent stimulation of the exonuclease activity of the DNA polymerase by the ddn protein.

10

20

10

20

Reactlon t l m e ( m l n )

FIG.3. Effect of the dda protein on DNA synthesis on a double-stranded DNA template. DNA synthesis reactions were carried out in the core replication system, as described under “Materials and Methods,” using 5 rg/ml of native phage T7 DNA as the template. Gene 32 protein was used in the reactions at theindicated concentrations. In each panel, the squures represent the reaction without ddu protein added, while triangles represent the same reaction with the addition of 2 pg/ml ddu protein. In this system, DNA synthesis starts by the addition of nucleotides to the 3’-OH end at a randomly placed nick in the DNA double helix.

protein concentrationswill support extensive strand displacement DNA synthesis by the core replication system even with helix unwinding (17). Unpaired single-stranded 3’-OH ends no DNA helicase present (36). are then produced by this helicase activity that are the preOther experiments have demonstrated that theddu protein ferred templates for the polymerase exonuclease (32) and the will not support strand displacement synthesis by the T4 rate of degradation increases. An alternative hypothesis is DNA polymerase in the absence of gene 32 protein or in the that the dda protein directly interacts with the DNA polym- absence of the polymerase accessory proteins (data not erase to stimulate its exonuclease activity. Although this is shown). The absolute requirement for both the accessory an interesting possibility, there is no evidence at thispoint to proteins and the 32 protein eliminatesthe possibility that the support such an interpretation. effect of the dda protein is due to its ability to denature the On a Double-stranded D N A Template, the ddaProtein DNA template into separate single strands; moreover, the Stimulates the Strand DisplacementD N A Synthesis Catalyzed amount of the dda protein added in these assays is not nearly by theT4 Replication Proteins-The effects of a DNA helicase enough to cause extensive DNA denaturation (13, 14, 17). are most likely to be detected during strand displacement The effect of increasing the concentration of the dda protein DNA synthesis on a double-helical template, because this on the total rate of strand displacement DNA synthesis is DNA synthesis-which is equivalent to that on the leading presented in Fig. 4. The two curves shown represent data at side of a replication fork-requires rapid helix unwinding. We two different 32 protein concentrations. Each curve has been therefore examined the effect of the dda protein on the DNA normalized, because at 50 pg/ml32 protein the maximum rate synthesis observed on a double-strandedphage T7 DNA tem- of DNA synthesis reached (1240 nM/min) is more than 5 plate in thepresence of the T4 “core” replication system (gene times higher than at 10 pg/ml32 protein (218 mM/min). 43,44/62,45, and 32 proteins). In this case, synthesis begins Two important facts should be noted with regard to Fig. 4. by the covalent addition of nucleotides onto the 3’-OH end First, instead of showing a simple saturation behavior, the at a random nick in the template,with the DNA strand dda protein begins to inhibit DNA synthesis when very high containing the 5‘ end at thenick being displaced as a single- concentrations are tested. A possible explanation is that at stranded tailwhen new DNA is synthesized (18).The results such concentrations, the dda protein can bind to the DNA obtained at several different concentrationsof 32 protein are template strand ahead of the polymerase, invading the DNA shown in Fig. 3. It is apparent that the addition of a small helix in the direction opposite to fork movement and thereby amount of the dda protein greatly stimulates strand displace- separating the growing 3’ end from its template. A second ment DNA synthesis at low concentrations of 32 protein, the fact to note is that the concentration of the dda protein at effect being most dramatic at a concentration of 32 protein (2 which inhibition occurs depends on the 32 protein concentrapg/ml) that does not normally support significant synthesis tion used in thereaction: at higher 32 protein concentrations, in the core replication system. The magnitude of the stimu- more dda protein can be added before any inhibition is seen. lation decreases at higher concentrations of 32 protein and it This effect is likely to be due to the competition observed nearly disappears at the highest concentration of 32 protein between the two proteins for binding to regions of singleused (180 pg/ml). This observation is in keeping with the stranded DNA (14). inhibitory effect of the 32protein on theDNA helicase activity These resultsindicate that, for optimalstimulation of of the ddu protein (13, 14) and with the fact that high 32 strand displacement DNA synthesis by the ddu protein, one

T4 DNA Helicase and DNA Replication

12936

\

I

d Id a l (ug/ml)

FIG.4. The effect of dda protein concentrationon total rate of strand displacement DNA synthesis observed in the core replication system. DNA synthesis reactions were carried out in the core replication system as described under “Materials and Methods,” using as template 2pg/ml double-stranded pJMCll0 DNA that had been nicked at a unique site by fd gene 2 protein. The two curues represent the results of two different 32protein concentrations. Rates of DNA synthesis were measured by acid precipitation of 4-pl aliquots every 30 s over a 2-min period. The maximum rates attained (normalized to 100 on the ordinate for each of the two curves) were 218 nM/min with 10 pg/ml32 protein present (diamonds) and 1240 nM/ min with 50 pglml32 protein present (triangles).

FIG. 5. The effect of different concentrations of the dda protein on the rateof replication fork movement at two concentrationsof 32 protein. Reaction conditions were identical with those in Fig. 4. After 0.5,1, and 2 min a t 37 “C, aliquots were removed from the reaction and quenched with an equal volume of 2% sodium dodecyl sulfate and 20 mM N*EDTA. After removing the unincorporated nucleotides by rapid gel filtration, the samples were electrophoresed through a denaturing 0.5% agarose gel in alkali and the radioactive DNA product size distribution was visualized by autoradiography of the dried gel. The indicated positions of the primer strand before elongation (4.75 kilobase pairs (kb))and after one round of rolling circle replication (9.5 kilobase pairs) were determined from the migration rates of DNA markers.

requires a balance between the helicase concentration and the 32 protein concentration. In the infected cell, one of the roles of the 32 protein may thus be to regulate the activity of the dda protein and/or other DNA helicases with a similar function-a view supported by genetic evidence (37, 38), as discussed in Ref. 14. The DNA Helicme Activity of the dda Protein Accelerates the Rate of Replication Fork Movement-All the results presented so far pertain to the effects of the dda protein on the total amount of DNA synthesized per unit time, without distinguishing between an increase in the number of template molecules used in the reaction and an increase in the rate of fork movement. The rate of fork movement can be conveniently measured during synthesis on a singly nicked, circular double-stranded DNA template, as described previously (20). In theassay used, synchronized elongation from a unique nick leads to the formation of “rolling circles” and the length of the growing DNAstrand increases in a continuous and essentially unlimited fashion. The rate of fork movement can be very simply measured by electrophoresing the radioactively labeled newly synthesized strands through a denaturing agarose gel, which allowsthe increase in the strandlength of the product to be measured as a function of reaction time. An autoradiogram of a gel from such an experiment is shown in Fig. 5. The addition of the dda protein to thecore replication system can be seen to result in a striking increase in the length of the product strands found at each time point, reflecting a major increase in the rateof fork movement. The rate attaineddepends on the ratio of dda protein to 32 protein; moreover, whenthe 32 protein concentration is increased, the maximum rate of fork movement is reached at a higher level of dda protein (Fig. 5). While some 32 protein is required to start new forks at a nick (see below), the maximum value of the fork rate attained

(about 200 nucleotides/s) is independent of the 32 protein concentration and in all cases only one class of replicating molecule is detected (Fig. 5). The dataare in agreement with two of the known properties of this system. First, the dda protein acts in a distributive fashion as a DNAhelicase, requiring new DNA binding events at a high frequency;thus, when the dda protein is present in suboptimal concentrations, all of the replication forks are accelerated to anintermediate extent. Second, the dda protein must compete with the 32 protein for single-stranded DNA binding in order to express its helicase activity; thus, high 32 protein concentrations inhibit the helicase action, but this inhibition can be completely overcome at a sufficiently high dda protein concentration. An examination of Fig. 5 reveals that the fork rates are identical at concentrations of 32 protein of 10 pg/ml and 50 pg/ml when the concentration of the dda protein is 10 pg/ml. However, under exactly the same conditions, the total amount of DNA synthesis is about 5-fold greater at the higher 32 protein concentration (Fig. 4 above; recallthat thetwo curves shown there have been normalized to about 5-fold different maximum values). Thus, we conclude that, with 10 pg/ml of the dda protein present, a &fold increase in the concentration of 32 protein must increase the number of DNA templates used by about a factor of 5. Therefore, the probability of starting a strand displacement event at a nick in the DNA double helix appears to be directly proportional to the 32 protein concentration in these experiments. The dda protein acts differently than the gene 41 protein, even though both proteins are DNA helicases that move along the lagging strand at the T4 replication fork. First of all, only the gene 41 protein will function with the T4gene 61 protein to make RNAprimers on the lagging strand. Moreover, when

T4 DNA Helicase and DNA Replication limiting amounts of the 41 protein are added to core replication reactions, two distinct populations of replication forks appear: one that contains the 41 protein and moves rapidly and another that moves at a slower rate that is identical with the rate of fork movement in the absence of gene 41 protein (10,36). (Only the firsttype of fork is capable of RNA-primed Okazaki fragment synthesis in the presence of gene 61 protein.) We have never observed twoclasses of replication forks in reactions to which the dda protein is added. Instead, all of the forks inthe reaction seem to move at about the same rate, even under conditions where the dda protein is limiting (Fig. 5 ) . These observations support the view that only the gene 41 protein acts in a highly processive manner, remaining bound to a single replication fork for a prolonged period once the protein has been assembled onto the DNA. Another major difference between the dda protein and the 42 protein as DNA helicases is their very different effect on the pausing of replication forks. Replication complexes containing only the gene 43,44/62,45, and 32proteins normally pause at certain sequence-determined sites on a doublestranded DNA template. These pause sites are detected as bands of DNA of distinct size whenthe product DNA strands are sized by electrophoresis through denaturing agarose gels. The addition of the 41 protein to the core replication system not only speeds up the fork rates, but it completely eliminates pausing at these sites. In contrast, the addition of the dda protein speeds up fork rates by about the same extent without eliminating the pausing.’ A very different result is obtained for the absolute block to fork movement created by a tightly bound RNApolymerase molecule. This block can be overcome by the dda protein, but it is not affected by the 41 protein (20). Thus, these two types of barriers to replication fork movement must differ from each other in afundamental way. DISCUSSION

The effects of the dda protein on in vitro DNA replication reactions catalyzed by purified T4 replication proteins are those expected for a DNA helicase. The dda protein does not stimulate the rate of DNA synthesis by the T4 DNA polymerase on a single-stranded DNA template, whether or not other replication proteins are present. However, when synthesis occurs on a double helical template and stranddisplacement is required, the addition of the d& protein can greatly increase the amount of DNA synthesis observed by increasing the rate of replication fork movement. This stimulation, which presumably requires ATP hydrolysis by the dda protein, allows DNAsynthesis to occur at unusually low concentrations of 32 protein. However, some32protein is still needed and the absolute requirement for the polymerase accessory proteins also remains. The mode of action of the dda protein as a DNA helicase is a distributive one, in that all of the replication complexes in areaction are stimulatedto thesame extent, and the observed fork rate is highly dependent on the dda protein concentration. The distributive mode of DNA unwinding by the dda protein is in sharp contrast to the very processive mechanism that has been demonstrated for the gene 41 protein. There would seem to be a fundamental difference in the functions of these two DNA helicases. The 41 protein is known to be required as anintegral component of the replication complex, both as a subunitof the RNA primase and as aDNA helicase that allows the replication fork to move rapidly and processively through long regions of double-stranded DNA (9, 10, 36, 39). The dda protein, on the other hand, may only be P. Bedinger, M. Munn, and B. Alberts, manuscript in preparation.

12937

required when the replication fork passes through regions where other proteins are tightly bound to the DNA template (20). In this view, the 41 protein would be an integral component of the replication complex that is required at all times, while the dda protein might be mainly found in a free pool, being drawn from that pool when needed. This hypothesis is consistent with the observation that the41 and dda proteins do not seem to cooperate in vitro, either for DNA unwinding (14) or for the stimulation of the strand displacement DNA synthesis catalyzed by purified T4 replication proteins? Our demonstration that thedda protein can strongly stimulate DNA synthesis in vitro does not constitute proof of its involvement in DNA replication inside the cell. It should be recalled that the core-41-61 system reconstituted from seven purified T4 replication proteins can perform most of the reactions expected for an in vivo replication fork, including mimicking the in vivo fork rate and replication fidelity (18, 19,36). However, the dda protein is required in vitrofor DNA synthesis in the presence of template-bound RNA polymerase molecules, which otherwise prevent fork movement past the site where they are bound (20). Whether this reflects the true in vivo function of the dda protein remains to be determined. To date, there is no evidence that the absence of the dda protein (during infections by T4 dda- mutants) prevents normal T4 DNA replication. It is possible that the function of the dda protein is vital to the virus, but that it is complemented by a second host or bacteriophage protein with an equivalent function. In the former case, one would expect to be able to find host mutants that are specifically nonpermissive for the growth of T4 dda mutants. Although E. coli optA mutants were initially thought to have this phenotype (40), a further analysis has recently revealed that theobserved failure of the T4 dda- mutants to grow on an E. coli optA- host is due to a second independent mutation in the dda mutant stocks tested? It has recently been suggested that the helicase I1 and the rep helicase of E. coli perform verysimilar functions at the E. coli replication fork, because the mutational loss of either enzyme is relatively harmless to the cell, while the loss of both enzymes seemsto be lethal ( 5 ) .This raises the possibility that more than one DNA helicase is able to provide the missing dda function during infection by T4 dda- mutants, which could greatly complicate determination of the in vivo role of the dda protein through the use of genetic methods. REFERENCES 1. Geider, K., and Hoffmann-Berling, H. (1981) Annu. Rev. Bwchem. 50,233-260 2. Abdel-Monem, M., and Hoffmann-Berling, H. (1980) Trends Biochem. Sci. 5,12%130 3. Scott, J. F., Eisenberg, S., Bertsch, L., and Kornberg, A. (1977) Proc. Natl. Acad. Sci. U. S. A. 7 4 , 193-197 4. Lane, H. E. D., and Denhardt, D. T. (1975) J. Mol. Bid. 97,99112 5. Taucher-Scholz, G., Abdel-Monem, M., and Hoffmann-Berling, H. (1983)in Mechanisms of DNA Replication and Recombimtwn: UCLA Symposium on Molecular Biology (Cozzarelli, N., ed) Vol. 10, pp. 65-76, Academic Press, New York 6. Kuhn, B., and Abdel-Monem, M. (1982) Eur. J. Biochern. 125, 63-68 7. Klinkert, M.-Q., Klein, A., and Abdel-Monem, M. (1980) J. Bwl. Chem. 265,9746-9752 8. Fuller, C. W., Beauchamp, B. N., Engler, M. J., Lechner, R. L., Matson, S. W., Tabor, S., White, J. H., and Richardson, C. C. (1983) Cold Spring Harbor Symp. Quant. BwZ. 47,669-679

C. V. Jongeneel, P. Bedinger, and B. M. Alberts, unpublished observations. P. Gauss, personal communication.

12938

T4 DNA Helicase and DNA Replication

9. Liu, C.-C., and Alberts, B. M. (1981) J. Biol. Chem. 2 5 6 , 28212829 10. Venkatesan, M., Silver, L. L., and Nossal, N.G. (1982) J. Biol. Chem. 257,12426-12434 11. Purkey, R.M., and Ebisuzaki, K. (1977) Eur. J. Biochem. 7 5 , 303-310 12. Behme, M. T., and Ebisuzaki, K. (1975) J. Virol. 15,50-54 13. Krell, H., Durwald, H., and Hoffmann-Berling, H. (1979) Eur. J. Biochem. 93,387-395 14. Jongeneel, C. V., Formosa, T., and Alberts, B. M. (1984) J . Biol. Chem. 269,12933-12938 15. Formosa, T., Burke, R. L., and Alberts, B. M. (1983) Proc. Nutl. Acad. Sci. U. S. A. 80,2442-2446 16. Formosa, T., and Alberts, B. M. (1984) Cold Spring Harbor Symp. Quant. Biol. 4 9 , in press 17. Kuhn, B., Abdel-Monem, M., Krell, H., and Hoffmann-Berling, H. (1979) J. Biol. Chern. 254,11343-11350 18. Alberts, B. M., Barry, J., Bedinger, P., Formosa, T., Jongeneel, C.V., and Kreuzer, K. N. (1983) Cold Spring Harbor Symp. Quant. Biol. 47,655-679 19. Nossal, N., and Alberts, B. (1983) in Bacteriophage T4 (Matbews, C., Kutter, E.,Mosig, E., and Berget, P., eds) pp. 71-81, American Society for Microbiology, Washington, D. C. 20. Bedinger, P., Hochstrasser, M., Jongeneel, C. V., and Alberts, B. M. (1983) Cell 3 4 , 115-123 21. Dotto, G. P., Enea, V., and Zinder, N. (1981) Proc. Nutl. Acad. Sci. U. S. A. 78,5421-5424 22. Bittner, M., Burke, R. L., and Alberts, B. M. (1979) J. Biol. Chem. 264.9565-9572 23. Morris, C.F., Hama-Inaba, H., Mace, D., Sinha, N.K., and Alberts, B. M. (1979) J. Biol. Chem. 254,6787-6796 24. Cleary, J. M., and Ray, D. S. (1980) Proc. Natl. Acad. Sci.U. S. A. 77,4638-4642 25. Meyer, T. F., and Geider, K. (1979) J. Biol. Chem. 2 5 4 , 12642-

12646 26. Mace, D.C., and Alberts, B. M. (1984) J. Mol. Biol. 1 7 7 , 313327 27. Piperno, J. R., and Alberts, B.M. (1978) J. Biol. Chem. 2 5 3 , 5174-5179 28. Huang, C.-C., Hearst, J. E., and Alberts, B.M. (1981) J. Biol. Chem. 256,4087-4094 29. Newport, J. H.,Kowalczykowski, S. C., Lonberg, N., Paul, L. S., and von Hippel, P. H. (1980) in Mechanistic Studies of DNA Replication and Genetic Recombination (Alberts, B. M., ed) pp. 485-505, Academic Press, New York 30. Huberman, J. S., Kornberg, A., and Alberts, B. M. (1971) J. Mol. Biol. 62,39-52 31. Nossal, N.G., and Peterlin, B.M. (1979) J. Biol. Chem. 264, 6032-6037 32. Huang, W.M., and Lehman, I. R. (1972) J. Biol. Chem. 2 4 7 , 3139-3146 33. Brutlag, D., and Kornberg, A. (1972) J. Biol. Chem. 2 4 7 , 241248 34. Venkatesan, M.,and Nossal, N. C. (1982) J. Biol. Chem. 257, 12435-12443 35. Bedinger, P., and Alberts, B. M. (1983) J. Biol. Chem. 268,96499656 36. Alberts, B. M., Barry, J., Bedinger, P., Burke, R. L., Hibner U., Liu, C.-C., and Sheridan, R. (1980) in Mechanistic Studies of DNA Replication and Genetic Recombination (Alberts, B. M., ed) pp. 449-473, Academic Press, New York 37. Little, J. W. (1973) Virology 53, 47-59 38. Dohertv. D. H.. Gauss., P... and Gold., L. (1982) . . Mol. Gen. Genet. 188,-77-90 ' 39. Liu. C.-C.. and Alberts. B. M. (1981) . , J. Biol. Chem. 256. 28132820 . 40. Gauss, P., Doherty, D. H., and Gold, L. (1983) Proc. Nutl. Acad. Sci. U. S. A. 80, 1669-1673