Proc. Nati. Acad. Sci. USA Vol. 84, pp. 1834-1838, April 1987 Biochemistry

Replication of simian virus 40 origin-containing DNA in vitro with purified proteins (simian virus 40 tumor antigen/DNA polymerase a-primase complex/HeLa single-stranded DNA binding protein)

C. RICHARD WOBBE, LAWRENCE WEISSBACH, JAMES A. BOROWIEC, FRANK B. DEAN, YASUFUMI MURAKAMI, PETER BULLOCK, AND JERARD HURWITZ Graduate Program in Molecular Biology, Sloan-Kettering Cancer Center, New York, NY 10021

Contributed by Jerard Hurwitz, November 26, 1986

Simian virus 40 (SV40) DNA replication deABSTRACT pendent on the SV40 origin of replication and the SV40 large tumor (T) antigen has been reconstituted in vitro with purified protein components isolated from HeLa cells. In addition to SV40 T antigen, these components included the DNA polymerase G-primase complex, topoisomerase I, and a fraction that contained a single-stranded DNA binding protein. The latter protein, which sediments at 5.1 S on glycerol gradients and copurifles with two major protein species of 72 and 76 kDa, was isolated solely by its ability to support SV40 DNA replication. The purified system retained the species-speciflic DNA polymerase a-primase requirement previously observed with crude fractions; the complex from HeLa cells supported SV40 replication, whereas that from calf thymus and mouse cells did not. DNA containing the polyomavirus origin of replication was replicated in a system containing polyomavirus T antigen, the HeLa single-stranded DNA binding protein-containing fraction, and DNA polymerase primase complex from mouse, but not HeLa, cells. While crude fractions yielded closed circular duplex DNA, none was detected with the purified system. Nevertheless, the addition of a crude fraction to the purified system yielded closed circular monomer products.

mentioned above to increase their efficiency (13). With these studies in mind, we have purified enzymes thought to be required for SV40 (and cellular) DNA replication in vivo and used them, in combination with fractions purified from HeLa cell extracts, to reconstitute SV40 DNA replication in vitro. This report describes the partial reconstitution of SV40 DNA replication in vitro, dependent on a combination of purified enzymes, T antigen, and a fraction derived from HeLa cell extracts that contains a SSB. MATERIALS AND METHODS DNA, T Antigen, and Extract Preparation. SV40 ori+ DNA (plasmid pSV01AEP), T antigen, and cytosolic extracts of HeLa cells (20 liters, 1-2 x 106 cells per ml) were prepared as previously described (7). Purified Enzymes. HeLa pol a-primase was purified to apparent homogeneity by immunoaffinity chromatography (Y.M., unpublished). Mouse and calf thymus pol a-primase preparations were those used previously (11). HeLa topoisomerase (topo) I was purified by a modification of the procedure of Liu and Miller (14, 15). Replication, DNA Binding, and Protein Assays. DNA replication assays (30 ,ul) and product analyses were carried out as described previously (7). In some reaction mixtures (see below), crude fractions were replaced by purified pol a-primase [0.32 unit of pol a, 0.5 unit of primase (11)] and HeLa topo I (1000 units; see ref. 14). Replication was not detected in these reaction mixtures (see below) but was observed when they were supplemented with fractions derived from the HeLa extract. Under these conditions 1 unit of replication activity stimulated incorporation of 1 nmol of dTMP into trichloroacetic acid-insoluble material after 60 mmn at 370C. Binding of single-stranded (ss-) and double-stranded (ds-) DNA was assayed as described for nuclear factor I and adenovirus DNA binding protein (16) except that the incubation temperature was 250C. Reaction mixtures contained 20 ng of native or heat-denatured Acc I-linearized pBR322A&EP DNA (7) that had been 5'-end-labeled with [y32P]ATP and T4 polynucleotide kinase (17). Protein concentrations were determined using the Bio-Rad protein' assay reagent with bovine serum albumin as the standard. NaDodSO4/polyacrylamide gels [4% (wt/vol) acrylamide stacking gel, 10%o (wt/vol) acrylamide resolving gel, 37.5:1 acrylamide to bisacrylamide] were prepared and electrophoresed according' to ref. 18. Protein bands were visualized by using the ICN Rapid Ag Stain kit.

Replication of simian virus 40 (SV40) DNA requires only one virus-encoded protein, large tumor antigen (T antigen); initiates within a unique, well-defined origin sequence; proceeds bidirectionally; and terminates in a manner thought to be analogous to that utilized by the host chromosome (1, 2). Replication occurs on a template that is associated with nucleosomes in a structure resembling cellular chromatin (3). Thus, the study of SV40 and presumably cellular DNA replication should be facilitated by the recent development of in vitro systems that reproduce many key aspects of SV40 DNA replication in vivo (4-8). By using such systems, the DNA sequences required 'for origin function in vitro have been identified (9, 10), the roles of the complex of DNA polymerase a (pol a) and primase in viral replication and host species specificity have been investigated (11), and a system has been described whereby newly replicated DNA is assembled into a chromatin-like structure (12). Genetic and biochemical analyses of prokaryotic systems have revealed a number of activities directly involved in the enzymatic process of DNA replication: origin-specific binding activity, priming and deoxynucleotide polymerizing activities, helix unwinding activity, single-stranded DNA binding protein (SSB) to maintain the DNA in an unwound configuration, primer removal activity, DNA ligase, activities that relieve torsional strain accumulating ahead of the replication fork and resolve daughter molecules, and factors that modify either the template or one of the activities

Abbreviations: SV40, simian virus 40; T antigen, large tumor antigen; RFI, replicative form I (superhelical circular duplex DNA); ssDNA, single-stranded DNA; dsDNA, double-stranded DNA; SSB, ssDNA-binding protein; topo, topoisomerase; pol a, DNA polymer-

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

ase a.

1834

Biochemistry: Wobbe et al. Fractionation of Crude Extract. Crude cytosolic extract (1055 mg of protein, 110 units) was adjusted to 35% saturation with saturated (40C) ammonium sulfate solution (0.559 ml/ml of extract) at 0C and centrifuged at 10,000 x g for 30 min. The supernatant was increased to 65% saturation (0.857 ml/ml of supernatant) and centrifuged as above (AS65 pellet). The 35% ammonium sulfate pellet was resuspended in a solution containing 30% saturated ammonium sulfate in 50 mM Tris-HCl, pH 8.0/1 mM dithiothreitol/1 mM EDTA (¼5 the starting extract volume), and after stirring at 0C for 30 min, the mixture was centrifuged as above, giving the AS30 pellet. The AS30 and AS65 pellets were dissolved in a volume of buffer A [20 mM Tris-HCl, pH 7.8/1 mM dithiothreitol/0.1 mM EDTA/10% (vol/vol) glycerol] equal to ¼5 the starting extract volume and dialyzed against buffer A containing 50 mM NaCl and 25% glycerol. The AS30 fraction (174 mg of protein, 152 units) was diluted to a protein concentration of 5 mg/ml with buffer A and applied to a Bio-Rex 70 column (Bio-Rad) (15 mg of protein per ml of bed volume) equilibrated with buffer A. The column was washed consecutively with 3 bed volumes each of buffer A containing 0.05, 0.1, and 0.25 M NaCI, and the eluates were assayed for replication activity complementing the AS65 fraction. Activity was detected only in the 0.1 M NaCl eluate. The Bio-Rex fraction (9.1 mg of protein, 194 units) was adjusted to 0.15 M NaCl and applied to a column of dsDNA-cellulose (Sigma, 4.3 mg of DNA/g; 5 mg of protein per ml of bed volume) equilibrated with buffer A containing 0.15 M NaCl. After washing with three bed volumes each of buffer A containing 0.15 and 0.25 M NaCI, replication activity was eluted with 2.0 M NaCl in buffer A. The dsDNAcellulose fraction (0.23 mg of protein, 53 units) was dialyzed against buffer B (20 mM Hepes-NaOH, pH 7.0/1 mM dithiothreitol/0.1 mM EDTA/50 mM NaCi) containing 25% glycerol and stored at -80°C, where activity was stable for at least 3 months. Glycerol Gradient Sedimentation. An aliquot of the ds DNA-cellulose fraction was diluted with an equal volume of buffer B and layered onto a 5.2-ml linear 15-30% (vol/vol) glycerol gradient in buffer B containing 0.25 M NaCl and bovine serum albumin (Miles) at 50 ,g/ml. Gradients were centrifuged at 45,000 x g for 23 hr in a Sorvall AH 650 rotor at 4°C. Cytochrome c (1.7 S), aldolase (7.3 S), and catalase (11.3 S) centrifuged under the same conditions were used as standards for calculating the sedimentation coefficient.

RESULTS Replication of SV40 Origin-Containing DNA with Purified Fractions. Ammonium sulfate fractionation was used to generate two fractions, AS30 and AS65, both of which are absolutely required for activity. The replication activity present in the AS30 fraction, measured by complementation with the AS65 fraction, was further purified by column chromatography on Bio-Rex 70 and dsDNA-cellulose, eluting in a single step in each case. As shown in Fig. 1, replication activity was efficiently reconstituted after each step. Note also that each fraction, assayed individually, was inactive. Assays of the ammonium sulfate fractions for activities likely involved in cellular DNA replication indicated that topo I and II, DNA ligase, and RNase H were found predominantly in the AS65 fraction and pol a-primase was distributed equally between the two fractions (data not shown). Chromatography on dsDNA-ceilulose separated residual pol a-primase from the replication activity in the AS30-derived fractions. Substitution experiments with these purified enzymes indicated that a combination of HeLa pol a-primase and topo I effectively replaced the AS65 fraction

Proc. Natl. Acad. Sci. USA 84 (1987)

1835

E 80

70

0

60

&3 50 40

0

~3020-

10 0

1

2

3

4

Froction ( ±I) FIG. 1. Titration of fractions required for SV40 DNA replication in vitro. The AS30 (12.4 mg of protein per ml) (o, n), or dsDNAcellulose (0.1 mg of protein per ml) (o, *) fractions were added to reaction mixtures containing either 85 jig of AS65 fraction protein (closed symbols) or HeLa pol a-primase and topo I as indicated in Materials and Methods (open symbols) in the presence ( ) or absence (--- - -) of SV40 T antigen. After incubation at 37°C for 90 min, trichloroacetic acid-insoluble radioactive material was determined.

in complementing the AS30, Bio-Rex, or dsDNA-cellulose fractions (Fig. 1). Since the optimal amount of AS65 fraction for these reaction mixtures contained 0.1-0.2 unit of pol a and 1000-2000 units of topo I, close to the amounts of these enzymes added in the purified system (Materials and Methods), the substitution is efficient. Requirements for Replication with Purified Components. DNA synthesis catalyzed by the purified components was absolutely dependent on T antigen, SV40 origin-containing DNA, pol a-primase, and the dsDNA-cellulose fraction (Table 1). Removal of topo I reduced nucleotide incorporation by a factor of 2-3. Linear ori+ DNA was replicated almost as efficiently as RFI. In contrast, this substrate is utilized 1/10th to 1/20th as efficiently by crude fractions (5-7). Like the RFI substrate, replication of linear DNA was dependent on the SV40 origin and T antigen; unlike RFI, but as expected, it was independent of topo I (J.A.B., unpublished observations). An additional requirement for replication with the purified fractions is that the pol a-primase must be from a source (HeLa cells) permissive for SV40 DNA replication; pol a-primase from nonpermissive mouse or calf thymus sources was inactive (Table 1). Similar observations have been made with crude extracts: SV40 DNA is replicated only in monkey or human cell extracts, polyomavirus DNA is replicated only in mouse cell extracts (11, 19). However, nonpermissive extracts could be made permissive for replication simply by supplementation with pol a-primase from a permissive source, suggesting that all other replicative functions could be performed by factors from either permissive or nonpermissive cells. Thus, it should be possible to replicate polyomavirus origin-containing DNA in vitro by using polyomavirus T antigen, the HeLa dsDNA-cellulose fraction, and mouse, but not HeLa, pol a-primase. As shown in Table 1, polyomavirus DNA was replicated in such a system almost as efficiently as in crude extract. Since none of the other reaction components contain detectable topo I, it is not clear why omission of this enzyme has little effect on replication. From these observations we conclude that the

1836

Proc. Natl. Acad. Sci. USA 84 (1987)

Biochemistry: Wobbe et al.

Table 1. Requirements for replication with the purified system dTMP incorporated, pmol/2 hr Component omitted or added 108 (i) Complete 112 - pol a-primase, topo I, + AS65 - T antigen, or dsDNA-cellulose