Proc. Natl. Ad. Sd. USA Vol. 73, No. 5, pp. 1594-1597, May 1976 Biochemistry An enzyme system for replication of duplex circular DNA: The replicative form of phage 4x174 (cistron A proteinlrep protein) SHLOMO EISENBERG, JOHN F. SCOTT, AND ARTHUR KORNBERG Department of Biochemistry. Stanford University Medical Center. Stanford, California 94305 Contributed by Arthur Kmnberg, March 11,1976 ABsliRACT Viral sin le strands (SS) are converted to the duplex form (RF) by a solu%le enzyme fraction from uninfected Escherichia coli [Schekman et al. (1975) J. Biol. Chem. 250, 5859-5865]. When reactions were su plemented with a soluble enzyme fraction from 4~174-iofectJce~s, replication of 4x174 su rhelical RF I DNA was observed. The activity supplied by inrcted cells was absent in cells treated with chloramphenicol or in cells infected with a 6x174 phage mutant in cistron A (cisA). A host function coded by the rep gene, essential in vivo for RF replication (but not for SS -, RF), was su plied by en- zyme fractions from either infected or uninfectef cells. Based on complementation assays, the cisAdependent and the rep- dependent proteins have each been purified about 1000-fold. The synthetic products of the enzymatic reaction were identi- fied as RF I and RF I1 in which viral (+) and complementary (-) strands were newly synthesized. In the life cycles of phages M13 and 4x174, the infecting sin- gle-stranded circles (SS) of the virus are first converted by host proteins to produce the duplex, circular, parental replicative form (RF) (1). Multiplication of RF to produce progeny RF requires participation of a phage-induced protein in addition to that of host proteins (1,2). When the resolution and reconstitution of the enzymatic replication of M13 and 6x174 DNAs were embarked upon (3, 4, the apparent simplicity of the first stage offered an attrac- tive, immediate goal. A soluble enzyme preparation obtained from gently lysed cells was effective in the SS to RF conversion. Such extracts from infected cells were even able to support RF multiplication (5), and it seemed then that the parts of the en- zymatic machinery for the production of progeny RF from SS might be quickly identified and pieced together. However, as the considerable complexity of the SS to RF conversion began to unfold (6-8). it seemed that this system should be understood before the problem of RF multiplication was tackled. Now that some 12 partially purified proteins, presumably nearly all that are required for the conversion of 6x174 SS to RF, are in hand (9, lo), the time seems ripe to explore RF rep lication. Other proteins would be required in this second stage because studies tn vivo have shown the need for the products of the host rep gene (11) and the phage cistron A gene (1). The correct state of the RF template, whether intact or modified, would also have to be determined. In the present report we describe the enzymatic synthesis of 4x174 RF I, and RF I1 from a superhelical RF I template. This reaction requires the soluble enzyme fraction that sustains the SS to RF conversion and, in addition, an extract from &A+ phage-infected cells. It is also essential that one of the extracts be derived from rep+ cells. The extensive purification of the &A-dependent and rep-dependent proteins will be described Abbreviations: SS, singlestranded circles of DNA; RF, replicative form of DNA. RF I is double-stranded, covalently closed circular, and su- perhelical. RF I1 is double-stranded and circular with one or more singlestrand breaks. and the participation of proteins required for SS to RF con- version will be discussed. MATERIALS AND METHODS Bacterial and Phage Strains. Bacterial strains were Esche- rtchta coli H560 (6), HMS 83 (6), C (12), HF4720 (12). and HF4704 rep3 (11). Phages were 4x174 am3, a lysis-defective phage with an amber mutation in cisE (1). and 4x174 am50, an amber mutant with a defect in cisA (obtained from Dr. R. Sinsheimer). Preparation of Fraction I1 from Uninfected Cells. The soluble enzyme fraction (Fraction 11) is an ammonium sulfate precipitate (obtained by adding 0.24 g to 1 ml of bacterial lys- ate) backwashed with an ammonium sulfate solution (0.24 g to 1 ml) (9). Growth of 4X17kInfected Cefls. E. colt HF4720 were grown in mT3XD medium (12) at 37" with aeration in a Fer- macell (New Brunswick Fermentor) to an OD590 of 0.5 (2 X 108 cells per ml). The cells were infected with 6x174 at multiplicity of infection (MOI) of 5. At 5-10 min after infection, chloram- phenicol (Calbiochem) was added (final concentration of 30 pg/ml) to inhibit single-stranded (viral) DNA synthesis and phage production (but not RF multiplication). At 30 min after infection, cells were harvested in a Sharples centrifuge at lo", resuspended (10" cells per ml) in 50 mM Tris-HCl (pH 7.5)- 10% (wt/vol) sucrose, frozen in liquid nitrogen, and stored at -20". Fraction I1 from ~#~X174-Infected Cells. Cells resuspended at 2 X 101O/ml in 130 mM Tris-HC1 (pH 7.5), 10% sucrose, 10 mM spermidine chloride, 4 mM dithiothreitol. 5 mM EDTA, 50 mM NaCl were treated with lysozyme (80 pg/ml), and in- cubated at 0" for 40 min, followed by either heat lysis (3) or Brij lysis (13). NaCl was added to the lysate to a concentration of 1 M. The lysate was centrifuged at 28,OOO X g at 0" for 30-60 min (supernatant is Fraction I). Ammonium sulfate (0.226 g/ml of Fraction I) was added over a period of 10 min; the suspension, at 0". was stirred for 15 min and centrifuged at 29,000 X g for 15 min. The precipitate was dissolved in Buffer A [50 mM Tris-HCI (pH 7.5), 20% (vol/vol) glycerol, 1 mM EDTA, 10 mM dithiothreitol] + 1 M NaCl. C#JX RF I. [3H]Thymidine-labeled +X RF I DNA was isolated from +X am3-infected E. coli C (treated with 30 pg/ml of chloramphenicol at 10 min after infection) and purified by sedimentation in neutral sucrose and ethidium bromide-CsCl gradients (12). The RF preparations were >90% RF I as judged by sedimentation velocity on alkaline sucrose gradients and by electron microscopy. RESULTS A soluble enzyme system for RF replication DNA synthesis by a crude enzyme fraction (Fraction 11), pre- pared from uninfected E. colt, is well supported by the addition 1594 Biochemistry: Eisenberg et al. Proc. Natl. Acad. Sci. USA 73 (1976) 1595 I I 30 2.0 - N N- 2.0 z z & 1.0 1.0 7 I I X 1.5 1.5 ; - m 1 0.5 0.5 8 N Pl 0 0 0 10 20 30 40 I I I 1 8 6 I I 1 1 20 40 60 80 FRACTION NUMBER The product of RF replication is RF I and RF 11. The reaction mixture was as in Table 1, with [3H]thymidine $X RF I template and (LY-~~PI~CTP. (a) The.product was purified by sedi- mentation velocity through a neutral sucrose gradient. (b) The RF peaks were pooled and concentrated with 2-propanol (12). The DNA was dissolved with 50 mM Tris-HC1, pH 7.5,lO mM EDTA denatured with alkali; and centrifuged on a linear, alkaline, 5-20% sucrose gra- dient (0.2 M NaOH, 0.8 M NaCI, 2 mM EDTA, 0.1% Sarkosyl). Cen- trifugation was performed in an SW50.1 rotor for 75 min at 50,OOO rpm and 15' in a Beckman L265B centrifuge. (c) RF I1 DNA was obtained in a reaction as described above, except that unlabeled $X RF I template was used for banding on an alkaline CsCl gradient. The RF I1 DNA, which was purified on neutral sucrose, was concentrated, mixed with 3H-labeled @X single-stranded viral DNA marker, de- natured with alkali, and analyzed on an equilibrium density gradient (16). of 6x174 SS (3) but not by 6x174 RF (Table 1). Fraction 11 from infected cells was inactive in DNA synthesis with either SS or RF as template. However, the combination of Fraction FIG. 1. Table 1. A soluble enzyme system utilizing RF I as a template for replication Source of Fraction I1 DNA synthesis Template Uninfected Infected (pmol) None + - <1 - + <1 + + 1.5 ss + - 100 - + <1 RF I + - 14 - + <1 + + 118 The reaction mixture in a 25 pl final volume contained: 50 mM Tris.HC1 at pH 7.5; 6% sucrose; 10 mM dithiothreitol; 0.1 mg/ml of bovine-serum albumin; 5 mM MgC1,; 50 pM each of dATP, dCTP, and dGTP, and 18 pM of [3H]dTl'P (specific activity 150-700 cpm/pmol), 800 pM ATP, 100 pM each of CTP, UTP, and GTP, and 2 mM spermidine chloride; occasionally [CT-~~PI~CTP was the labeled deoxynucleoside triphosphate in place of d'ITP. To this reaction mixture, 40 pg of Fraction 11 from uninfected E. coli H560, 1 pg of Fraction 11 from $X am3-infected E. coli HF4720 cells, and 1 pg of $X RF I DNA were added. The reaction was carried out at 30" for 20 min and stopped by the addition of 0.2 ml of 0.1 M sodium pyrophosphate and 1 ml of 10% trichloroacetic acid. The precipitate was collected on glass fiber filters (Whatman GF/C), washed three times with a 1 M HCl, 0.2 M sodium pyrophosphate solution, dried, and counted in 5 ml of a toluene-based scintillation fluid in a Nuclear Chicago liquid scintillation counter. DNA syn- thesis is expressed as total nucleotide incorporation. I1 from uninfected cells with one from infected cells supported DNA synthesis in response to the addition of RF I (Table 1). When RF I was replaced by RF 11, synthesis was only 2045% as great. The product of RF replication is RF I and RF I1 The DNA synthesized by the crude enzyme fractions from uninfected and infected cells upon addition of RF I as template (as in Table 1) was first purified by sedimentation on a neutral sucrose gradient (Fig. la) and then analyzed by sedimentation velocity on alkaline sucrose (Fig. Ib). About !E% of the newly synthesized RF was in the form of RF I, indicating that a complete round of replication had been supported by the en- zyme preparation. Analysis of the synthetic RF I1 by banding in alkaline CsCl showed that both the viral (+) and comple- mentary (-) strands were synthesized (Fig. IC). Table 2. A protein dependent on @X174 cistron A required for RF replication DNA synthesis rg (pmol) Fr. I1 uninfected only 40.0 14 Plus Fr. I1 infected, wild type 0.5 51 Plus Fr. I1 infected, wild type 1.0 114 Plus Fr. I1 infected (chloramphenicol) 1.0 13 Plus Fr. I1 infected, am50 (cisA) 1.0 13 Fraction II from uninfected E. coli H560 and @X174-infected HF4720 were each prepared as described in Materials and Methods. $X am3 (cis& lysis-defective) and @X am50 (&A) were used for infections. Chloramphenicol treatment of @X am3-infected E. coli HF4720 was by addition of chloramphenicol (to a concentration of 200 pgjml) 1 min before infection with phage. Synthesis was measured as described in Table 1. 1596 Biochemistry: Eisenberg et al. Proc. Natl. Ad. SG. USA 73 (1976) Table 3. A protein dependent on E. coli rep+ required for RF replication Fraction I1 DNA Complementing synthesis Uninfected Infected rep,+ fraction (pmol) rep rep 0.3 rep + rep+ 24 rep rep+ 11 rep rep FR 11, 0.4 pg 3.7 rep rep FR 11, 0.8 pg 7.0 ~~ Fraction 11 is from uninfected and infected HF4704 rep3 and HF4704 rep+ cells prepared as described in Materials and Methods. Complementing rep+ Fraction II was prepared from HMS83. Syn- thesis was measured as described in Table 1, except that 10 pg of Fraction 11 from infected cells and 0.5 pg of DNA were used in each reaction. A 4X174-induced protein is required for RF multiplication In the absence of cisA protein, in uiw. parental RF accumulates as RF I, but no further replication takes place (14). This con- dition can be achieved either by treating the cells with chlor- amphenicol (at UK) pg/ml) before infection or by using a phage mutant in cistron A (4X arn50) (14). Fraction I1 prepared from infected cells treated with chloramphenicol 1 min before in- fection failed to complement Fraction I1 from uninfected cells in RF replication (Table 2). This result suggests that a protein synthesized after infection is required. That synthesis of this protein depends on an intact cistron A gene in the phage is indicated by the inactivity of Fraction I1 prepared from 4X am504nfected cells in the complementation assay (Table 2). The E. coli rep protein is required for RF replication The infected E. coli rep3 mutant, in uiw. sustains the synthesis of parental RF but fails to promote its replication (11). Fraction I1 prepared from an uninfected rep mutant converted SS to RF (data not shown) but failed to support RF replication when complemented with a Fraction I1 obtained from infected cells also mutant in the rep gene (Table 3). Replication of RF was observed with Fraction I1 prepared from either uninfected or infected cells of the rep+ genotype (Table 3). suggesting that the product of this gene is essential. Partial purification of the 4X cisA- and E. coli rep- dependent proteins The dependence of the RF replication system upon 4X cisA and E. colt rep functions provided an assay for purification of these activities. Table 4 records current progress in purification of the 4X &A- and E. coli rep-dependent proteins; further purification is needed to obtain homogeneous preparations for characterization. DISCUSSION Understanding the molecular mechanisms of duplex DNA replication requires the resolution of the enzymatic machinery responsible for this operation. An excellent probe for identifying and resolving the parts of this cellular machinery is replication of the single-stranded (SS) and the duplex circular replicative form (RF) of phage 4x174 in E. coli (7). Table 4. Partial purification of ex174 cisA- and E. coli rep-dependent proteins cis A rep Specific activity Specific activity (unitslmg of Recovery (unitslmg of Recovery protein x IO3) (%I protein x lo') (%) ~~~~ ~ I Extract (0.53) (100) I Extract (0.032) (100) 68 III Bio-Gel, A-0.5m 10.4 93 I11 Bio-Rex 70 5.0 46 IV Bio-Rex 70 67 68 IV DNA-cellulose 38 8 V DNA-cellulose 410 19 I1 Ammonium sulfate 4.8 100 11 Ammonium sulfate 0.45 Purification of $X cisA-dependent protein: Fraction I (lysate supernatant, 190 ml) and Fraction LI (6.5 ml) were prepared from $X am3- infected E. coli HF4720 (see Materials and Methods). Fraction II was passed through a Bio-Gel A-0.5m column (150 ml) (Fraction III, 38 ml). Fraction III was dialyzed against Buffer B (50 mM imidazole.HC1 at pH 6.7, 20% glycerol, 1 mM EDTA, 10 mM dithiothreitol) + 0.3 M NaCl, and applied to a Bio-Rex 70 column (18 ml) equilibrated with Buffer B + 0.3 M NaCl. The column was washed with 2 column volumes of Buffer B + 0.3 M NaCl, and the +X174 cisA-dependent activity was eluted with Buffer B + 1 M NaCl (Fraction N, 11.5 ml). Fraction IV (1 ml) was diluted with Buffer B to a conductivity value of Buffer B + 0.4 M NaCl and applied to a single-stranded DNA-cellulose column (0.5 ml). The column was washed with 2 column volumes of Buffer B + 0.4 M NaCl and the activity was eluted with Buffer B + 1 M NaCl. Of the input activity, 50% was recovered in the 1 M NaCl eluate. Assays were performed as in Table 1. One unit is defined as 1 pmol of total nucleotide incorporated/min. Purification of E. coli rep-dependent protein: Fraction I (4400 ml) was prepared from E. colt HMS83 (900 g) as described previously for the purification of dmC protein (9). The Fraction II is an ammonium sulfate precipitate (obtained by the addition of 0.24 g ammonium sulfate to 1 ml of Fraction I) which was backwashed first with 440 ml of an ammonium sulfate solution (0.24 gadded per ml of buffer) and then 440 ml of a more dilute solution (0.20 gadded per ml of buffer) and three successive times with 170 ml, 100 ml, and 90 ml of a still more dilute ammonium sulfate solution (0.18 g added to 1 ml of buffer) essentially as previously described (9). Fraction II(50 ml) in Buffer C (50 mM imidazole.HC1 at pH 6.9, 20% glycerol, 1 mM EDTA, 1 mM dithiothreitol) was diluted to 350 ml and applied to a Bio-Rex 70 (200-400 mesh) column (110 ml) equilibrated with Buffer C + 0.2 M NaCl. The column was washed with 75 ml of Buffer C + 0.2 M NaCl followed by 200 ml Buffer C + 0.4 M NaCl. The rep-dependent activity was then eluted with Buffer C + 2.0 M NaCl(200 ml), diluted to 600 ml in Buffer C, and precipitated by ammonium sulfate (0.31 g for each ml). The precipitate was dissolved in Buffer D (50 mM Tris-HCI at pH 7.5, 20% glycerol, 1 mM EDTA, 1 mM dithiothreitol) (Fraction 111, 9.8 ml), diluted to 190 ml, and applied to a single-stranded DNA- cellulose column (25 ml) equilibrated with Buffer D + 0.1 M NaCl. The column was washed with Buffer D + 0.1 M NaC1, followed by a 0.125- 0.550 M NaCl gradient. The rep-dependent activity was eluted with Buffer D + 2.0 M NaCl (50 ml) and precipitated by the addition of an equal volume of saturated neutralized ammonium sulfate solution. The precipitate was collected after standing overnight on ice, and was dissolved in Buffer C (Fraction N, 1.7 ml). Calculation of yield was based on the activity recovered from Fraction I in an ammonium sulfate pellet (prepared by adding 0.24 g for each ml) and backwashed with an ammonium sulfate solution (0.24 g added to 1 ml of buffer). Assays of Fraction I were not reliable. Biochemistry: Eisenberg et al. Proc. Natl. Ad. Sci. USA 73 (1976) 1597 The present investigation is an extension from a series of earlier studies in which examination of the first stage of 4x174 viral DNA teplication, conversion of S to RF, identified and provided some 12 cellular components. The complexity of this multienzyme system, comprising the proteins required for replication of the host chromosome, suggests that this system would, to some extent, be employed for duplex DNA replica- tion. Not included in the roster of proteins that serve in S to RF conversion are two additional proteins, a need for which may be anticipated from earlier investigations (1.11). One of these, the rep protein, whose function is still unknown, is furnished by the host (17). The other, the cistron A protein, is induced by the phage (1) and is thought to nick the RF I at a specific loca- tion (13,14). We have purified each of these gene-dependent proteins by complementation assays The extensive purification, approximately 1000-fold at this point, suggests that these pro- teins are in fact the products coded by these genes. Charac- terization of their functions awaits their complete purification and coordination with other proteins required in RF replication. Among the proteins, one which is surely needed is a DNA polymerase for chain elongation. We expect that the DNA polymerase I11 holoenzyme would be the choice. Evidence supporting this conjecture is the specific and total inhibition of RF replication by antiserum to copolymerase III* (data not shown), one of the components of the holoenzyme (15). We may also expect that DNA polymerase I serves in gap filling and DNA ligase, in sealing of chains. The need for additional pro- teins, which may include some of those required for conversion of SS to RF, remains to be determined. This work was supported in part by grants from the National Insti- tutes of Hdth and tbe National Science Foundation A grant from the Damon Runyon-W. A. Walter Winchell Cancer Fund provided postdoctoral fellowship support for S. E. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Sinsheimer. R. L. (1968) Prog. Nuc. Acid Rm. Mol. Bbl. 8, 115-169. Denhardt, D. T., Eisenberg, S., Harben, B., Lane, H. E. D. & McFadden, G. (1975) in DNA Synthesis and Its Regulation, eds Goulian, M. & Hanawalt, P. (W. A. Benjamin, Inc., Menlopark, California), pp. 39-22. Wickner, W., Brutlag, D., Schekman, R. & Kornberg. A. (1972) Proc. Natl. Ad. Sd. USA 69,965-969. Schekman, R., Wickner, W., Westergaard, O., Bruthg, D., Geider, K., Bertrch, L. L. & Kornberg, A. (1972) Pm. Natl. Ad. Sd. USA 69,2691-2695. Schekman, R. (1972) Fed. Proc. 31,442 abstr. Geider, K. & Kornberg, A. (1974) 1. B&l. Chem. 249,39994005. Schekrnan. R., Weiner, A. & Kornberg, A. (1974) Sdence 188, 987-993. Wickner, S. & Hurwitz, J. (1974) Proc. Natl. Ad. Sd. USA 71, 4120-4124. Schekmnn, R., Weiner, J. H., Weiner, A. & Komberg, A. (1975) J. Bbl. Chem. e50,5859-5865. Wickner, S. & Hurwitz, J. (1975) in DNA Synthesfs and Its Regulatfon, eds. Godian, M. & Hanawak, P. (W. A. Benjamip, Inc., Menlo Park, Calif.), pp. 227-238. Denhardt, D. T., Iwaya, M. & Larison, L. L. (1972) Virology 49, 486-496. Eisenberg, S., Harben. B., Hours, C. & Denhardt, D. T. (1975) 1. Mol. Bbl. 99,107-123. Henry, J T. & KNppen, R. (1974) Proc. Natl. Ad. Sd. USA 71,1549-1553. Francke, B. & Ray, D. S. (1971) 1. Mol. Bbl. 61,565-586. Wickner, W., Schekman, R., Geider. K. & Komberg, A. (1973) Proc. Natl. Ad. Sd. USA 70,1764-1767. Brunk, C. (1969) Bloclrfm. Bbphys. Acta 179,136-144. Lane, H. E. D. & Denhardt, D. T. (1975)J. Mol. Bid. 97,99-113.