The interaction of SV40 large T antigen with unspecific double-stranded DNA: An electron microscopic study

VIROLOGY

189,

293-303

(19%‘)

The interaction

of SV40 Large T Antigen with Unspecific An Electron Microscopic Study

Double-Stranded

DNA:

RAINER WESSEL, UWE RAMSPERGER, HANS STAHL, AND ROLF KNIPPERS’ Division of Biology, Universitat Received February

Konstanz, D 775 Konstanz, Germany

12, 1992; accepted

March 20, 1992

T antigen, an early protein encoded by simian virus 40 (SV40), is a specific DNA-binding protein with high affinity for elements in the viral origin of replication where it forms a double-hexameric complex as a prerequisite for DNA untwisting and, in the presence of ATP hydrolysis, for DNA unwinding. Like other specific DNA-binding proteins, T antigen also associates with DNA strands of random sequence albeit at reduced affinity. In addition, T antigen is able to unwind unspecific DNA sequences starting from internal binding sites. This property could be a step in the pathway leading to the chromosomal rearrangements that are frequently observed in SV40-transformed cells. This possibility prompted us to investigate the binding of T antigen to unspecific DNA using electron microscopy. We observed that the protein binds randomly to many unspecific DNA sites excluding a preference for particular DNA sequences or structural features. Addition of ATP to the binding buffer induces the formation of oligomeric, possibly hexameric, T antigen complexes that frequently align to form long arrays of DNA-bound protein. Magnesium salts induce the formation of tightly packed T antigen aggregates which bind to DNA to form many DNA branches and loops that emanate from the aggregated protein core. Upon ATP hydrolysis, aggregated T antigen catalyzes the unwinding of DNA duplices. c 1992 Academic

Press,

Inc.

that covers a 70-bp region of DNA (Mastrangelo et al., 1989), inducing an untwisting of the DNA double helix (Borowiec and Hurwitz, 1988). Upon ATP hydrolysis, the DNA-bound T antigen functions as a DNA helicase, separating the hydrogen bonds between complementary DNA strands (Stahl et al., 1986). According to recent experiments, the functional helicase appears to be a hexamer, and the adjacent T antigen hexamers, bound at the origin, remain associated while DNA is reeled through the stationary complex (Wessel et a/., 1992). In addition, biochemical assays have shown that T antigen is also able to initiate DNA unwinding from internal binding sites of any sequence (Wold et al., 1987; Scheffner et al., 1989; Yang et al., 1989). Since T antigen binding to nonspecific DNA is weak, unwinding is best observed at low salt concentrations and at relatively high concentrations of T antigen, thereby increasing the probability for the formation of properly oriented hexamers at adjacent sites (Scheffner et al., 1989). Indeed, due to the high concentration of up to lo6 molecules of T antigen/nucleus of infected cells, origin-independent replication of viral DNA occurs in viva at an efficiency that may be about 1% of that observed for origin-dependent viral DNA replication (Martin and Setlow, 1980). Thus, T antigen could very well initiate the unwinding of cellular DNA sequences. An unwinding of cellular DNA could be an important step in the pathway leading to the various genome rearrange-

INTRODUCTION Simian virus 40 (SV40) encodes a multifunctional protein, large T antigen, which is essential for viral multiplication as well as for cell transformation (reviewed by Rigby and Lane, 1983; Stahl and Knippers, 1987; Livingston and Bradley, 1987). Large T antigen is a DNA-binding protein with high affinity to double-stranded DNA sites containing two or more closely spaced GAGGC pentanucleotides. Two pairs of these recognition elements, oriented as a palindrome in opposite orientations, form a specific T-antigen-binding site in the center of the viral origin of replication (Deb et a/., 1986; Mastrangelo et a/., 1989). Like other specific DNA-binding proteins, T antigen has a weak affinity to double-stranded DNA of random sequence. The first contacts of T antigen with DNA may involve these random sites, followed by dissociations and reassociations causing T antigen to move relative to the DNA strand until it meets the high-affinity origin site. In the absence of ATP, several T antigen molecules bind to the origin site covering a 35-bp stretch of DNA centered about the GAGGC palindrome (Mastrangelo eta/., 1985). However, in the presence of ATP (or of nonhydrolyzable ATP analogues), the affinity of T antigen to the GAGGC site increases about 1O-fold (Deb and Tegtmeyer, 1987) and two closely spaced T antigen hexamers at the origin form a bilobal structure ’ To whom reprint requests should be addressed. 293

0042.6822192

$5.00

CopyrIght 0 1992 by Academic Press, Inc All rights of reproduction I” any form resewed

WESSEL ET AL.

294

ments that have frequently been observed in SV40transformed cells (for recent references, see Gurney and Gurney, 1989; Stray et al., 1989; Ray et al., 1990; St. Onge et al., 1990; Stewart and Bacchetti, 199 1). Therefore, T antigen binding to unspecific DNA deserves more attention than it has received in recent years. Here, we extend earlier biochemical studies on the interaction of T antigen with unspecific DNA sequences (Giacherio and Hager, 1979; Tegtmeyer et al., 1981; Montenarh and Henning, 1982; Dorn et a/., 1982; Wright et a/., 1984) using electron microscopy for the investigation of complexes formed between T antigen and double-stranded DNA. This approach provides a clearer picture of the mode of interaction between T antigen and unspecific DNA and contributes to an understanding of the mechanism of T-antigencatalyzed DNA unwinding. MATERIALS

AND METHODS

DNA and proteins We used the following DNA constructs (Baur and Knippers, 1988): plasmid pSV-MO1 (or? DNA), a derivative of pUC8, carrying the SV40 origin (the 109-bplong HindIll-Ncol fragment, including SV40 nucleotides 5172 to 41); and plasmid pSV-C4 (ori- DNA), carrying the SV40 Hinfl fragment G (1 1 1 bp, including SV40 nucleotides 4460 to 457 1). Before use, plasmid pSV-MO1 was linearized by Avall restriction: the large fragment carried the SV40 origin at distances of 1 .l kbp from one end and of 1.5 kbp from the other end. Plasmid pSV-C4 was linearized by EcoRl restriction. Single-stranded ends were filled in using the Klenow fragment of bacterial DNA polymerase I. Large T antigen was prepared by immunoaffinity chromatography essentially as described by Simanis and Lane (1985). The Escherichia co/i single-strand binding (SSB) protein was obtained from Pharmacia. DNA binding In different experiments, between 50 and 200 ng DNA were mixed with 0.1-3 pg T antigen in 20 mM triethanolamine-HCI buffer at pH 7.5 or 6.0 as indicated below. If required, the buffer contained MgCI, in the concentrations given below. Nonhydrolyzable ATP(+) (adenosine 5’-o-3-thio-triphosphate; Sigma) was used in concentrations of l-3 mM. After incubation for 15 min at 37”, the resulting

protein-DNA complexes were fixed in 0.1% (final) formaldehyde (15 min at 37”). DNA unwinding Depending on the experiment, between 50 and 400 ng DNA were mixed with 0.1 to 3 pg T antigen and 0.1 to 1 kg bacterial SSB protein in 0.02- to 0.04-ml volumes of 20 mM triethanolamine-HCI, pH 7.5, 0.5 mM dithiothreitol, 7 mM MgCI,, and 4 mMATP plus 10 mM phosphocreatine and 50 pg/ml creatine phosphokinase as an ATP-regenerating system. After 30 min at 37”, the reaction was blocked by addition of 20 mM Protein-DNA EDTA and 0.1 YO (final) glutaraldehyde. complexes were then prepared by centrifugation through Sepharose Cl 4B columns as described by Silhavy et al. (1984) and used for electron microscopy. For biochemical assays, the restricted DNA was labeled by [a-32P]-dATP in the Klenow polymerase fill-in reaction. Double-stranded and unwound singlestranded DNA were separated by gel electrophoresis as described before (Stahl et al., 1986). Electron

microscopy

For our experiments, we used the BAC (alkyl benzyl dimethyl ammonium chloride; Sigma) spreading technique of Vollenweider et al. (1975). A final concentration of 0.001% BAC was added to a ~-PI sample of the protein-DNA mixture and spread on doubly distilled water. The resulting surface film was mounted on carbon-coated grids, which were pretreated for 10 s at 150 V and 3 mA to glow discharge in tripropylamine vapor at 12 Pa (Dubochet and Groom, 1982). The grid was washed with doubly distilled water, and treated for 1 min with 5 mM uranyl acetate, 5 mM HCI in 80% ethanol. After removal of water in 90% ethanol and drying, further contrast was achieved by electron beam evaporation of tungsten at an angle of 8” on the rotating grid (see Wessel et al., 1990). Electron micrographs were taken with a Hitachi H-7000 electron microscope. Length measurements of DNA and protein-DNA complexes were done on a CRP digitizer using magnified positives (Wessel et al., 1992). Diameters of individual T antigen complexes were determined with a measuring lens using magnified (500,000X) photographs rotating the positive image clockwise at 45” angles. RESULTS Conditions

for binding to unspecific

DNA

As previously observed, T antigen binds unspecifically with equal affinities to both ori+ and to ori- DNA at

T ANTIGEN

AND UNSPECIFIC

pH values < 7 and in the absence of salt (Dorn et a/., 1982). These conditions are certainly unphysiological, and it is likely that, at low pH values, positively charged amino acid side chains become available for an electrostatic interaction with DNA. With higher salt concentrations (0.1 M NaCI) and/or at pH values between 7.5 and 8, the interaction of T antigen with unspecific DNA is largely suppressed; and T antigen shows a much higher affinity for ori+ DNA than for ori- DNA (Dorn et al., 1982). To determine the distribution of T antigen on unspecific DNA, we investigated the nucleoprotein complexes, formed at pH 6 and 7.5, by electron microscopy. At pH 6, T antigen covered the entire DNA double strand, forming a dense nucleoprotein rod; and very little unbound T antigen could be detected (Fig. 1A). At pH 7.5, the affinity of T antigen to DNA was reduced as only a fraction of total T antigen was found in association with DNA, and a considerable fraction of T antigen could be detected as free isometrical particles (Fig. 1 B). Bound T antigen was found at many different sites on the DNA duplex. A comparison of many protein-DNA complexes revealed that no part of the linear DNA was preferred as a binding site over other parts, excluding the possibility that T antigen may recognize a special sequence or structural feature in the oriDNA. For example, we obtained no evidence for a preferential binding of T antigen to the (blunt) DNA ends of the ori- DNA. Thus, our electron microscopic results are in perfect agreement with earlier biochemical data showing that T antigen binds with high affinity to unspecific DNA at low pH, and that the affinity to unspecific DNA decreases with higher pH values. All further experiments were performed at pH 7.5 because a relatively loose binding of T antigen to unspecific DNA is an obvious requirement for a sliding or jumping of T antigen along the DNA strand, and the mobility of T antigen should eventually lead to an association of two closely spaced T antigen complexes as a prerequisite for duplex unwinding from internal sites. As demonstrated in Fig. 1, the protein-DNA complexes, formed at pH 6 as well as at pH 7.5, were quite uniform in appearance and were well spread on the electron microscope grids. They could therefore be used to determine their lengths relative to protein-free DNA (added as an internal standard to preparations of T antigen-DNA complexes). We determined an average length ratio of complexed DNA/free DNA of 0.87 for T antigen-DNA structures formed at pH 6 and a ratio of 0.91 for T antigen-DNA structures at pH 7.5. These numbers exclude a wrapping of DNA about

DNA

295

bound T antigen and suggest that the path of the DNA strand is not much distorted in the T antigen-DNA complexes, formed under the conditions of Fig. 1. This may be different when magnesium salt is added to the binding buffer as earlier work had shown that magnesium ions induce the formation of high molecular weight aggregates of free T antigen (Montenarh and Henning, 1983). To investigate this possibility, we prepared T antigen-DNA complexes at pH 7.5 in triethanolamine buffer, containing final concentrations of 2 mM and of 5 mM MgCI,. In Fig. 2A, we show the control (no magnesium salt): T antigen complexes were randomly distributed on the DNA strands; and a considerable fraction of T antigen remained free in solution as already demonstrated above in Fig. 1 B. At 2 mM MgCI,, the number of T antigen complexes on DNA strands was reduced as individual complexes had aggregated to form larger structures, occasionally including T antigen complexes on distant sites of the same DNA molecules, leading to the formation of DNA loops (Fig. 2B) (Schiedner et a/., 1990). At 5 mM MgCI,, aggregation was much more pronounced; and large and apparently tightly packed aggregates of T antigen had formed containing dense irregular cores from which many DNA strands emanated as loops or branches (Fig. 2C). It appears that free T antigen molecules participated in aggregate formation as the fraction of unbound T antigen was significantly reduced in the presence of magnesium salts (compare Fig. 2A and C). We interpret these data to indicate that the DNA binding sites of many individual T antigen molecules were buried within the aggregates, and that only some of these sites may be exposed and available for DNA binding. Aggregate formation was reversible. Addition of EDTA in concentrations sufficient to chelate-free magnesium ions converted the aggregates of Fig. 2C into extended T antigen-DNA complexes like those shown in Figs. 1 B and 2A (not shown). At this point, it may be useful to emphasize that all experiments were performed with homogeneous preparations of T antigen (see Methods). Therefore, all protein structures detectable on the electron microscope grids as DNA-bound or as free isometric particles corresponded to T antigen. This was confirmed using the T antigen-specific monoclonal antibody Pab 101, which reacts with a C-terminal region of T antigen (Gurney et al., 1980) and which does not interfere with its DNA-binding activity (Wiekowski et al., 1987). All protein complexes, seen in the electron micrographs, had an increased diameter when T antigen had been pre-

296

WESSEL ET AL.

FIG. 1. Effects of pH on the interaction of T antigen with unspecific DNA. Linearized pSVC4 DNA (100 ng) and T antigen (3 pg) were kept for 15 min at 37” in 0.04 ml 1 mM EDTA, 20 mM triethanolamine-HCI at pH 6.0 (A) or at pH 7.5 (B). The resulting complexes were prepared for electron microscopy as described under Methods. The length of the bar corresponds to 800 nm.

treated with Pab 101 (data not shown; for details see Wessel et al., 1992). T antigen binding to DNA in the presence of ATP(+) It is known that ATP or its nonhydrolyzable analogues have profound effects on the mode of interaction between individual T antigen molecules leading to the formation of hexamers which associate as a bilobe at the viral origin (Mastrangelo et a/., 1989; Wessel et a/., 1992). To investigate the effect of ATP on the association of T antigen and unspecific DNA, we used triethanolamine buffer at pH 7.5 with 2 mM MgCI, and 2 mNI ATP(+). Under these conditions, most magnesium ions were probably chelated by ATP; and their concentration should be too low for extensive aggregate formation.

In Fig. 3A, we present a control experiment using ori+ DNA to show that the experimental conditions were appropriate for the formation of bilobal T antigen complexes at the origin of replication. Using the same relative amount of T antigen and unspecific ori- DNA, we detected a few T antigen molecules distributed at random sites on the DNA strands (Fig. 3B). Obviously, the concentration of T antigen was too low for the formation of bilobal T antigen complexes as the ori- DNA lacks the recognition sequences which guide T antigen to adjacent DNA sites. Consequently, we used a lo-fold higher T antigen concentration to increase the probability for T antigen complex formation on unspecific DNA. We found again that T antigen was bound in an apparently random fashion at many sites on the DNA strands but, clearly, a large fraction of the bound T antigens contact each

T ANTIGEN

AND UNSPECIFIC

DNA

297

FIG. 2. Effects of magnesium salts on the IInteractIon of T antigen with DNA. T antigen (1.5 wg) and linearized plasmid pSWC4 DNA (100 ng) were Incubated as in Fig. 1 without MgCI, (/J), with 2 ml\/l MgCI, (6) or with 5 mM MgCI,. Ti ie bar corresponds to 1 pm.

other, forming clusters of aligned complexes including two or more isometric units (Fig. 3C), possibly indicating some limited degree of cooperativity between T antigen molecules, bound to DNA in the presence of ATP(+‘). For a better evaluation of the data, we present in Fig. 4 selected parts of these and additional electron micrographs in a fourfold higher magnification. We have used these electron micrographs to estimate the diameter of individual T antigen molecules. In Fig. 4A, we show a selection of single T antigen units, either in a free, unbound form or in association with internal sections of DNA and with DNA ends. All these T antigen forms had very similar diameters of 14.2 (kO.8) nm, which are clearly too large for T antigen mono- or dimers (monomeric Mr = 82.5 kDa). In fact, according to previous estimates (Wessel et al., 1992), a diameter of ca. 14 nm is most consistent with a hexameric T antigen oligomer. In Fig. 4B, we present a collection of bilobal T antigen complexes. Their diameter, vertical to the axis of DNA, was again determined to be in the range of 14 nm, whereas their diameter, parallel to the axis of DNA, was 26.1 (kO.9) nm. These numbers suggest that the bilobal complexes of Fig. 4B were double hexamers, comparable to those formed at the SV40 origin in the presence of ATP (Mastrangelo et al., 1989; see Fig. 3A).

In Fig. 4C, we show clusters of three, four, and more units of T antigen, which formed on unspecific doublestranded DNA when the binding reaction was carried out in the presence of ATP(+). The diameters of individual units, vertical to the axis of DNA, were again determined to be in the range of 14 nm, suggesting that each unit probably corresponded to a T antigen hexamer. Cluster formation of DNA-bound T antigen was not only detected in the presence of ATP(-+S) but also in the presence of hydrolyzable ATP (Fig. 4D). In addition, the hydrolysis of ATP caused an unwinding of doublestranded DNA (Fig. 4D, lower panel); and T antigen clusters were frequently found in front of and at the unwinding fork. The formation of T antigen clusters, however, is probably not required for the unwinding of unspecific double-stranded DNA since, as previously shown, single T antigen units, most likely hexamers, are quite commonly found at unwinding forks (Wessel et a/., 1992). DNA unwinding Finally, we have investigated the structure of T antigen-DNA complexes formed under the conditions commonly used in biochemical DNA-unwinding assays. These conditions are: low ionic strength buffers (such as Tris-HCI or, in the present case, triethanol-

298

WESSEL ET AL.

amine-H(I) at pH 7.5, containing 7 mlLl MgCI, and 4 mM ATP (see Methods). Based on the results presented in the preceding sections, we expected the formation of complex protein-DNA aggregates as free magnesium ions induce an aggregation of T antigen while ATP causes the clustering of DNA-bound T antigen oligomers. Indeed, electron microscopy revealed large T antigen-DNA aggregates with dense internal protein complexes and numerous DNA branches, carrying long arrays of bound T antigen (Fig. 5A). We excluded the possibility that these structures arose as artifacts, occuring during the preparation of T antigen-DNA complexes for electron microscopy, because they could be pelleted by low-speed centrifugation before glutaraldehyde was added for fixation (data not shown). Even though our biochemical assays clearly indicated that double-stranded DNA was unwound under the buffer conditions used (Scheffner et a/., 1989) (insert in Fig. 5) we detected no evidence for ongoing unwinding reactions in these electron micrographs. However, partially or totally unwound DNA may simply be buried within the large aggregates. To investigate this possibility, we treated the protein-DNA aggregates with 0.25 M NaCl to dissociateT antigen from unspecific binding sites on doublestranded DNA. However, T antigen remains bound to unwinding forks under these conditions (Wiekowski et al., 1988) and magnesium-induced aggregates do not dissociate. The salt-treated complexes were isolated by gel filtration on Sepharose Cl 4B, and prepared for electron microscopy. The structures observed consisted of densely packed cores of T antigen with many branches of protein-free DNA (see Fig. 3C). Some of the emanating DNA branches were single-stranded as visualized by bound SSB protein, which was included in the reaction mixture to prevent the reannealing of separated complementary DNA strands (black arrows in Fig. 5B). In addition, salt treatment induced a partial disintegration of the T antigen-DNA aggregates resulting in released DNA molecules including some partially unwound molecules, carrying T antigen at the unwinding forks (white arrow in Fig. 5B). Elsewhere, we have systematically described the structures of ori- DNA, par-

tially unwound by T antigen (Scheffner et a/., 1989; Wessel et a/., 1992). DISCUSSION The experiments, reported in this communication, were performed to better understand the conditions of interaction between T antigen and unspecific DNA sequences. We demonstrated that T antigen binds with high affinity to unspecific DNA at low pH values. This result is in agreement with earlier biochemical data which had shown that, under these conditions, T antigen is unable to distinguish between specific and unspecific DNA (Dorn et a/., 1982). A likely explanation is that electrostatic interactions between protein and DNA are favored at low pH. At pH 7.5, T antigen has a reduced affinity to unspecific DNA. However, a relatively loose binding to DNA is an essential requirement for a movement of T antigen relative to the DNA strand. On ori+ DNA, T antigen will eventually reach its high affinity binding site at the palindromic pentanucleotide-recognition element where a properly oriented double-hexameric bilobe is formed when ATP is present as an allosteric effector and where DNA strand separation begins with the hydrolysis of ATP. The formation of the double-hexameric bilobe on oriDNA cannot be guided by a recognition sequence and therefore occurs by chance if two T antigen hexamers meet in an appropriate orientation. At least in the in V&O assay system, T antigen catalyzed DNA unwinding from internal sites is complicated by the fact that magnesium ions, which are required for this reaction, induce large T antigen aggregates. These aggregates possess dense cores of tightly packed T antigen whose DNA binding sites are only partially available for interaction with DNA. However, we could show by biochemical procedures as well as by electron microscopy that these conditions do not impede the duplex unwinding reaction which is, in fact, quite active. Only a fraction of the aggregated T antigen molecules may be actively engaged in the unwinding reaction, and treatment with 0.25 M NaCl removes a substantial amount of T antigen, which is only loosely bound to unspecific double strands. The salt-resistant fraction is probably the active component as it is

FIG. 3. T antigen-DNA interaction in the presence of ATP($S). (A) Linearized plasmid pSV-MO1 (ori’) DNA(100 ng) and T antigen (100 ng) were incubated for 15 min at 37” in the presence of 2 mM ATP-(r-S) and processed for electron microscopy as described under Methods. Insert: the position of the two adjacent T antigen complexes was determined relative to the DNA ends in 18 protein-DNA complexes. (B) As in (A) except that linearized plasmid pSV-C4 (ori-) DNA was used in this experiment. The insert shows the location of DNA-bound T antigen in 18 molecules. (C) Linearized ori- DNA (100 ng) and T antigen (1 pg) were incubated and investigated by electron microscopy as in (A) and (B). The horizontal bar corresponds to 500 nm.

T ANTIGEN

A

2607

AND UNSPECIFIC

bp

5

.-

10

number of molecules

‘w 3000 -j

15

DNA

299

300

WESSEL ET AL.

FIG. 4. T antigen in the presence of ATP. The structures, presented in (A-C), are from an experiment performed with ATP-(-/S) as in Fig. 3C. The data in (D) were obtained using hydrolyzable ATP as described below in Fig. 5. (A) Single T antigen complexes either in a free, unbound form (upper left) or bound to unspecific DNA. (B) Bilobal T antigen complexes on unspecific ori- DNA. (C)Aligned T antigen complexes on unspecific ori- DNA. (D) T antigen, bound to ori- DNA, in the presence of ATP. Lower panel: partially unwound DNA. Note that the emerging single strands are covered by bacterial SSB protein, which can be distinguished from bound T antigen by its smaller diameter. The horizontal bar corresponds to 200 nm.

known that T antigen is salt-stably bound at the singleto-double strand junctions of unwinding forks (Stahl and Knippers, 1983; Wiekowski et al., 1988). It is an interesting possibility that the unwinding of double-stranded DNA could be catalyzed by aggregated and immobilized T antigen. Yang et al. (1989) found that T antigen induces positive superhelical turns into relaxed circularly closed DNA substrates when the reaction mixture contained bacterial DNA topoisomerase I, which removes negative but not positive superhelices. These results were interpreted to indicate a transient unwinding of the

DNA double helix by a fixed T antigen complex, producing positive supercoils ahead of and negative supercoils behind the unwinding complex according to the twin domain model of Liu and Wang (1987). Yang et al. (1989) also observed that the reaction was greatly stimulated by polyethylene glycol (PEG 20,000) at concentrations that are expected to aggregate T antigen. When discussing the data of Yang et al. (1989) in the context of a model of matrix-bound DNA replication, Cook (1991) proposed that T antigen as a replicative helicase does not act by moving along the double helix, and that T antigen remains at a fixed site, most

T ANTIGEN

AND UNSPECIFIC

DNA

FIG. 5. Complexes of T antigen and unspecific DNA under unwinding condrtrons (7 mM MgCI, and 4 mM ATP). (A) Lrneanzed plasmid pSVC4 (100 ng) and T antigen (3 fig) were incubated under DNA-unwinding condrtrons as described In the text and under Methods. (6) As In (A), except that the protern-DNA complexes were treated with 0.25 M NaCl before they were processed for electron microscopy. Black arrows: single stranded DNA branches covered by bacterial SSB protein. White arrow: partially unwound unit length DNA. The horizontal bar corresponds to 500 nm. Insert: Results of a biochemical unwinding assay performed under buffer condrtions rdentrcal to those used for the electron mrcroscopy of protein-DNA complexes. After incubation, the DNA was deproternized before separation on a 1% agarose gel (Scheffner et a/., 1989). Lane 1, heat denatured 32P-labeled lrneanzed pSVC4 (ori-) DNA; lane 2, native labeled substrate; lane 3, incubation under unwinding assay condrtrons In the absence of T antigen; lane 4, incubation with T antigen. ss, single-stranded DNA; ds, double-stranded DNA.

302

WESSEL ET AL.

probably the nuclear matrix (see Schirmbeck and Deppert, 1989), while the DNA rotates through the stationary complex. This model is supported by the recent finding that the T antigen dodecamer complex, assembled at the origin, is subsequently converted to a processive helicase with two reaction centers (Wessel et a/., 1992). Consequently, the helicase complex does not track along the DNA, rather the DNA is translocated relative to the complex. ACKNOWLEDGMENTS R.W. was a recipient of a Boehringer lngelheim fellowship. work was supported by Deutsche Forschungsgemeinschaft Fonds der Chemischen Industrie.

This and

REFERENCES BAUR, C. P., and KNIPPERS,R. (1988). Protein-induced bending of the simian virus 40 origin of replication. f. Mol. Biol. 203, 1009-l 019. BOROWIEC,J. A., and HURWI~~,J. (1988). Localized melting and structural changes in the SV40 origin of replication induced by T antigen. EMBO J. 7, 3149-3158. COOK, P. R. (1991). The nucleoskeleton and the topology of replication. Cell 66, 627-635. DEB, S., DELUCIA, A. L.. BAUR, C. P., KOFF, A., and TEGTMEYER, P. (1986). Domain structure of the simian virus 40 core origin of replication. Mol. Cell. Biol. 6, 1663-1670. DEB, S. P., and TEGTMEYER,P. (1987). ATP enhances the binding of simian virus 40 large T antigen to the origin of replication. J. Viral. 61,3649-3654. DORN, A., BRAUER,D., Oreo, B., FANNING, E., and KNIPPERS,R. (1982). Subclasses of simian-virus-40 large T antigen: Partial purification and DNA binding properties of two subclasses of tumor antigen from productively infected cells. Eur. J. Biochem. 128, 53-62. DUBOCHET, J., and GROOM, M. (1982). The mounting of macromolecules for electron microscopy with particular reference to surface phenomena and the treatment of support films by glow discharge. Adv. Opt. Elect. Microsc. 8, 107-135. GIACHERIO, D., and HAGER, P. (1979). A poly(dT)-stimulated ATPase activity associated with simian virus 40 large T antigen. /. Biol. Chem. 254, 8113-8116. GURNEY, E. G., HARRISON, R. O., and FENNO, J. (1980). Monoclonal antibodies against simian virus 40 T antigens: Evidence for distinct subclasses of large T antigen and for similarities among nonviral T antigens. 1. Viral. 34, 752-763. GURNEY, T., and GURNEY, E. G. (1989). Spontaneous rearrangement of integrated simian virus 40 DNA in nine transformed rodent cell lines. /. Viral. 63, 165-l 74. LIU, L. F., and WANG, J. C. (1987). Supercoiling of the DNA template during transcription. Proc. Nat/. Acad. Sci. USA 84, 7024-7027. LIVINGSTON,D., and BRADLEY, M. (1987). The simian virus 40 large T antigen. A lot packaged into a little. Mol. Biol. Med. 4, 63-80. MARTIN, R. G., and SETLOW, V. P. (1980). The initiation of SV40 DNA synthesis is not unique to the replication origin. Cell20, 381-391. MASTRANGELO, I. A., HOUGH, P. V. C., WILSON, V. G., WALL, J. S.,

HAINFELD, J. F., and TEGTMEYER, P. (1985). Monomers through trimers of large T antigen in region I and monomers through tetramers in region II bind to SV 40 origin DNA as stable structures in solution. Proc. Nat/. Acad. Sci. USA 82, 3626-3630. MASTRANGELO,I. A., HOUGH, P. V. C., WALL, J. S., DODSON, M., DEAN, F. B., and HURWITZ,J. (1989). ATP-dependent assembly of double hexamers of SV40 T antigen at the viral origin of replication. Nature 338, 658-662. MONTENARH. M., and HENNING, R. (1982). The binding of simian virus 40 T antigen to the polyphosphate backbone of nucleic acids. Biochim. Biophys. Acta 697, 322-329. MONTENARH, M., and HENNING, R. (1983). Self-assembly of simian virus 40 large T antigen oligomers by divalent cations. 1. Viral. 45, 531-538. RAY, F. A., PEABODY,D. S., COOPER,J. L., and KRAEMER,L. M. (1990). SV40 T antigen drives karyotype instability that precedes neoplastic transformation of human diploid fibroblasts. J. Cell. Biochem. 42, 13-31, RIGBY, P., and LANE, D. (1983). The structure and function of SV40 large T antigen. Adv. Viral Oncol. 3, 31-57. SCHEFFNER,M.. WESSEL, R., and STAHL, H. (1989). Sequence independent duplex DNA opening reaction catalyzed by SV40 large tumor antigen. Nucleic Acids Res. 17, 93-l 06. SCHIEDNER, G., WESSEL, R., SCHEFFNER, M., and STAHL, H. (1990). Renaturation and DNA looping promoted by the SV40 large tumor antigen. EMBO J. 9, 2937-2943. SCHIRMBECK, R., and DEPPERT,W. (1989). Nuclear subcompartmentalization of simian virus 40 large T antigen: Evidence for in vivo regulation of biochemical activities. /. Viral. 63, 2308-2316. SILHAW, T. J., BERMAN, M., and ENQUIST, L. (1984). Preparation of mini-gel filtration columns. In “Experiments with Gene Fusions” Cr. J. Silhavy, Ed.), pp. 203-204. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. SIMANIS, V., and LANE, D. P. (1985). An immunoaffinity purification procedure for SV40 large T antigen. tirology 144, 88-l 00. STAHL, H., and KNIPPERS,R. (1983). SV40 large T antigen on replicating viral chromatin. Tight binding and localization on the viral genome. J. Viral. 47, 65-76. STAHL, H., DR~GE, P., and KNIPPERS,R. (1986). DNA helicase activity of SV40 large tumor antigen. EMBO J. 5, 1939-l 944. STAHL, H., and KNIPPERS,R. (1987). The simian virus 40 large tumor antigen. Biochim. Biophys. Acta 910, l-l 0. ST. ONGE, L., BOUCHARD, L., LAURENT, S., and BASIN, M. (1990). lntrachromosomal recombination mediated by papova virus large T antigens. J. Viral. 64, 2958-2966. STRAY, A., JAMES, M. R., and SARASIN,A. (1989). High recombination rate of an Epstein-Barr virus-simian virus 40 hybrid shuttle vector in human cells. /. Viral. 63, 3837-3843. STEWART, N.. and BACCHE~I, S. (1991). Expression of SV40 large T antigen, but not small t antigen, is required for the induction of chromosomal aberrations in transformed human cell. Virology 180,49-57. TEGTMEYER,P., ANDERSEN, B., SHAW, S. B.. and WILSON, V. G. (1981). Alternative interactions of the SV40 A protein with DNA. virology 115, 75-87. VOLLENWEIDER,H., SOGO, J.. and KOLLER,T. (1975). A routine method for protein-free spreading of double- and single-strand nucleic acid molecules. Proc. Nat/. Acad. Sci. USA 72, 83-87. WESSEL, R., MOLLER, H.. and HOFFMANN-BERLING,H. (1990). Electron microscopy of DNA helicase I complexes in the act of strand separation. Eur. J. Biochem. 189, 277-285.

T ANTIGEN

AND UNSPECIFIC

WESSEL, R., SCHWEIZER,J., and STAHL, H. (1992). Simian virus 40 T-antigen DNA helicase is a hexamer which forms a binary complex during bidirectional unwinding from the viral origin of DNA replication. /. viral. 66, 804-815. P., and STAHL, H. (1987). Monoclonal antibodWIEKOWSKI,M., DR~IGE, ies as probes for a function of large T antigen during the elongation process of simian virus 40 DNA replication. /. viral. 61, 41 l418. WIEKOWSKI,M., SCHWARZ, M., and STAHL. H. (1988). Simian virus 40 large T antigen DNA helicase. Characterization of the ATPase dependent unwinding activity and its substrate requirements. /. Biol. Chem. 263, 436-442.

DNA

303

WOLD, M. S., LI, J. J., and KELLY,T. 1. (1987). Initiation of simian virus 40 DNA replication in vitro: Large-tumor-antigenand origin-dependent unwinding of the template. Proc. /Vaf/. Acad. Sci. USA 84, 3643-3647. WRIGHT, P. J., &LUCIA, A. L., and TEGTMEYER,P. (1984). Sequencespecific binding of simian virus 40 A protein to nonorigin and cellular DNA. Mol. Cell. Biol. 4, 2631-2638. YANG, L., JESSEE,C. B., LAu, K., ZHANG, H., and LIU, L. F. (1989). Template supercolling during ATP-dependent DNA helix tracking: Studies with simian virus 40 large tumor antigen. Proc. Nat/. Acad. SC;. USA 86, 6121-6125.