The Cap Region of Topoisomerase I Binds to Sites near Both Ends of Simian Virus 40 T Antigen

Two independent binding sites on simian virus 40 (SV40) T antigenfor topoisomerase I (topo I) were identified. One was mappedto the N-terminal domain (residues 83 to 160) by a combinationof enzyme-linked immunosorbent assays (ELISAs) and glutathioneS-transferase (GST) pull-down assays performed with variousT antigen deletion mutants. The second was mapped to the C-terminaldomain (residues 602 to 708). The region in human topo I thatbinds to both sites in T antigen was identified by ELISAs, GSTpull-down assays, and double-hexamer binding assays with topoI deletion mutants. This region corresponds to a distinct domainon topo I known as the cap region that maps from residues 175to 433. By combining these data with information about the structureof T-antigen double hexamers associated with origin DNA, wepropose that the cap region of topo I associates specificallywith both ends of the double hexamer bound to the SV40 originto initiate DNA replication.

INTRODUCTION

Top

Introduction

The replication of simian virus 40 (SV40) DNA is becoming fairlywell understood. Advances in this field have become possiblebecause one can carry out DNA replication assays, in additionto the biochemistry, in a cell-free system with purified components(12, 13, 20, 24, 45, 49, 52). All of the proteins required forthis in vitro reaction have been isolated. Initiation of replicationbegins when a double hexamer of T antigen assembles over theorigin. The formation of the double hexamer then triggers astructural rearrangement of the DNA whereby the EP region ispartially melted and the AT track is untwisted (1-4, 7, 30).Synthesis of RNA primers takes place after topo I, replicationprotein A, and DNA polymerase ${alpha}$ /primase (pol/prim) assemble overthe origin (11, 27, 51). The RNA primers are extended by polymerase ${alpha}$ . Elongation of DNA chains is catalyzed by DNA polymerase ${delta}$ inassociation with PCNA and RF-C (10, 11, 28).

There is considerable interest concerning the mechanism by whichthe initiation complex assembles. Older evidence indicated thatdouble hexamers of T antigen assemble over the replication originfrom individual monomers (6, 19), but more recent data demonstratethat preformed hexamers can also bind and unwind duplex DNA(46). In either case, it is clear that a complete double hexameris the active form of the helicase (39). The three cellularinitiation factors (topo I, replication protein A [RPA], andpol/prim) are then recruited, but the order in which they bindhas not been determined. Our laboratory has generated a largeamount of data supporting the idea that topo I is an integralcomponent of the initiation complex and that it must be presentfrom the beginning of DNA replication to function (14, 44).The recruitment of topo I is dependent on the presence of doublehexamers in association with DNA (14). The presence of a nucleotideis required, but ATP hydrolysis is not needed (14). topo I canbind to T antigen directly in the absence of DNA (35, 36), aswell as to DNA (26, 40). Likewise, the recruitment of RPA andpol/prim appears to depend on multiple protein-protein and protein-DNAinteractions (see reference 33 for a review).

In order to elucidate the structure and composition of the initiationcomplex, it is necessary to map all protein-protein interactionsites. Previous data from our lab have demonstrated the existenceof two independent topo I binding regions in T antigen (35):one in the N-terminal region (residues 1 to 246) and the otherafter residue 246. The functional significance of each of thesetwo binding sites is not known. Previous mapping studies implicatedthe N-terminal region of topo I in binding to T antigen (16),but data from our lab has shown that topo 70, which is missingthe first 174 amino acids of the enzyme (42), can stimulateT-antigen-dependent DNA replication to the same extent as full-lengthtopo I (P. Trowbridge and D. Simmons, unpublished data), indicatingthat this binding site is not physiologically significant. Inan effort to gain insight into their function, each topo I bindingregion on T antigen was mapped. Likewise, the region in topoI responsible for binding to each of these sites in T antigenwas identified. The results, combined with recent structuredata of T antigen double hexamers (47, 48), strongly suggestthat the cap region of topo I interacts specifically with sequencesnear the tips of the double hexamer in the presence of originDNA.

MATERIALS AND METHODS

Top

Materials and Methods

Purification of T antigen and T-antigen fragments. Wild-type (WT) T antigen was purified from baculovirus-infectedinsect cells by immunoaffinity purification as previously described(35) by using PAb101 antibody (15). Baculovirus transfer vectorsexpressing various T antigen constructs were generated by ligatinga PCR-derived DNA fragment into pVL1393 or pAcG1 for glutathioneS-transferase (GST)-tagged proteins (Pharmingen) as previouslydescribed (35). After DNA sequencing to confirm the correctsequence, the transfer vector was cotransfected with BaculoGold(Pharmingen) into Sf9 cells to generate recombinant baculovirusesaccording to the recommendations of the manufacturer. UntaggedT-antigen fragments were purified by immunoaffinity with theappropriate monoclonal antibody (PAb101 or PAb419 [15] or PAb416[17]). PAb101 was used for all C-terminus containing polypeptides,PAb419 was used for the N-terminal constructs except for 83-246,in which case PAb416 was used. GST-tagged T-antigen fragmentswere purified by binding to GST-Sepharose beads as describedpreviously (53). Proteins were either left attached to beadsand the beads were used directly in binding assays, or theywere first eluted from beads with free glutathione. The concentrationof each protein was estimated by Coomassie blue staining ofsodium dodecyl sulfate-polyacrylamide gels with phosphorylaseb as a marker.

Purification of topo I and fragments. Untagged WT topoisomerase I, topo 70, and topo 58 were purifiedby standard chromatography from extracts of baculovirus-infectedSf9 cells as previously described (42). Protein concentrationswere estimated by reactivity with Bio-Rad protein reagent. GST-taggedproteins were either expressed in bacteria or in insect cells.The bacterially expressed polypeptides (see Fig. 1) were purifiedas described by Yang and Champoux (53) from Escherichia coliTOP₁₀F' (Invitrogen) cells transformed with pGEX-derived plasmids.Some GST-tagged proteins were expressed in insect cells. Thesewere made by first introducing a PCR-derived DNA fragment intopAcG1, followed by cotransfection in Sf9 cells to make the recombinantvirus. Purification of GST-tagged topo I constructs was carriedout as described above for the GST-tagged T-antigen fragments.

View larger version (12K):
[in this window]
[in a new window]
FIG. 1. (Top) Map of topoisomerase I deletion mutants used in this study. Various fragments of human topo I were expressed either in bacteria (t) or in baculovirus-infected insect cells (v). The topo I polypeptides correspond to known topo I domains (40-42).

ELISAs. Enzyme-linked immunosorbent assays (ELISAs) were performed essentiallyas described previously (35). In brief, WT or truncated topoI was allowed to bind to a microtiter plate. The wells wereblocked with phosphate-buffered saline containing 1% gelatin,and various amounts of WT T antigen or T-antigen fragments wereadded. After binding and washing steps, an antibody specificfor the T-antigen peptide was added, followed by the additionof horseradish peroxidase-conjugated anti-mouse immunoglobulinG (IgG). A peroxidase substrate (ortho-phenylenediamine) wasadded and, after a few minutes, H₂SO₄ was used to stop the reactionand the absorbance was read at 490 nm by using a Dynex platereader.

GST-protein binding assays. GST-protein binding assays were performed essentially as describedpreviously (18) with some modifications. GST-tagged proteinsbound to glutathione-Sepharose or -agarose beads were washedtwice with 30 mM HEPES-KOH (pH 7.8)-10 mM KCl-7 mM MgCl₂ andincubated for 3 h at room temperature with either T antigenor topo I intact molecules or fragments in 250 µl of thesame buffer containing 2% nonfat dry milk. The beads were recoveredby centrifugation and washed three times with 1 ml of 30 mMHEPES-KOH (pH 7.8)-25 mM KCl-7 mM MgCl₂-0.25% inositol-0.25mM EDTA-0.1% Nonidet P-40. This washing procedure was not sufficientlystringent for the binding assays with GST-bound C-terminal fragmentsof T antigen (see Fig. 9), and so for these experiments beadswere washed with 0.05 M Tris (pH 8.0)-0.5 M NaCl-0.001 M EDTA-10%glycerol-1% Nonidet P-40. The bound proteins were eluted byboiling in 1 volume of 4x sample buffer, separated by sodiumdodecyl sulfate-polyacrylamide gel electrophoresis, and detectedby Western blot analysis.

View larger version (48K):
[in this window]
[in a new window]
FIG. 9. GST-binding assay between WT topo I and various C-terminal T antigen constructs. A total of 30 µl of Sepharose beads containing 8 µg of bound GST 110-708 or the molar equivalent of various C-terminal constructs or control GST beads was incubated with 1.8 µg of purified WT topo I. After the beads were washed, bound topo I was eluted with electrophoresis sample buffer and applied to an 8% gel. Detection was with 8G6 monoclonal antibody and horseradish peroxidase-conjugated anti-mouse IgG.

Origin DNA-dependent binding assays. The preparation and purification of a small circular DNA containingthe origin of replication was as previously described (14).Binding assays contained 5 ng of circular DNA, 400 ng of immunoaffinity-purifiedT antigen, and 100 ng of GST-topo 70 or equivalent molar amountsof other truncated products. After 1 h at 37°C, glutaraldehydewas added to 0.1%, and incubation was continued for an additional10 min. The reactions were loaded onto a composite polyacrylamide-agarosegel as described previously (14). Protein-DNA complexes weretransferred to nitrocellulose, and the proteins on the membranewere detected by a Western reaction with an anti-GST antibody.

RESULTS

Top

Results

Mapping of T-antigen binding site on topo I. A number of topo I deletion mutants were generated (Fig. 1)and expressed either in bacteria ("t" in Fig. 1) or in baculovirus-infectedinsect cells ("v" in Fig. 1). The untagged proteins were purifiedby standard chromatographic procedures as described by Stewartet al. (42). GST-tagged topo I constructs from bacteria werepurified on glutathione-agarose beads, and those from insectcells were purified by binding to glutathione-Sepharose beads.The binding of purified untagged topo I deletion mutants toimmunoaffinity-purified WT T antigen was monitored by ELISAsas shown in Fig. 2. Both topo I deletion mutants tested (topo70 and topo 58) bound to T antigen, indicating that the T-antigenbinding site was located between residues 175 and 659. topo70 routinely bound better in this assay than full-length topoI, but the reason for this is not known.

View larger version (9K):
[in this window]
[in a new window]
FIG.2. (Bottom) ELISA between WT T antigen and various topo I fragments. A total of 100 ng of purified WT topo I, topo 70, or topo 58 was bound to ELISA plates, followed by incubation with various amounts of WT T antigen as shown. The plates were washed and incubated with PAb101, followed by the addition of horseradish peroxidase-labeled anti-mouse IgG and peroxidase substrate. After the reactions were stopped with sulfuric acid, the absorbance at 490 nm was measured by using a plate reader. Control reactions were also performed without topo I but in the presence of all other components. The absorbance values from these wells were subtracted from the corresponding ones obtained in the presence of WT or mutant topo I.

GST-tagged topo I deletion mutants were tested for their abilityto bind to WT T antigen, which we have previously shown to containtwo separate binding sites for topo I (35) (Fig. 3A). Sincetopo I and T antigen both bind DNA, we performed these bindingreactions in the presence of ethidium bromide, which disruptsDNA-protein interactions (23), to eliminate the possibilitythat binding is mediated through small amounts of contaminatingDNA. WT T antigen bound best to topo 70 and topo 31 (see Fig.1), although there was some background binding with the otherconstructs. We also examined the binding to deletion mutants246-708 and 1-246 of T antigen (Fig. 3B and C), each of whichcontains one binding site for topo I (35). GST-tagged topo 70and topo 31 bound to both T antigen mutants well, whereas GST-taggedtopo 10 and topo 17 bound poorly, if at all. These data indicatethat both the N-terminal and C-terminal binding sites in T antigeninteract with topo I within the region specified by topo 31(amino acids 175 to 433).

View larger version (38K):
[in this window]
[in a new window]
FIG. 3. GST-binding assay between WT T antigen and various topo I fragments. A total of 10 µl of glutathione-Sepharose or -agarose beads containing 18 µg of GST-tagged topo 70 or the molar equivalent of topo 31, topo 17, topo 10, or GST alone was incubated with 500 ng of WT T antigen (A) or the molar equivalent of 246-708 (B) or 1-246 (C) in the presence of 1 mM ethidium bromide. The beads were then washed and resuspended in electrophoresis sample buffer. Samples were loaded onto 10% acrylamide gels. The proteins were transferred to nitrocellulose, and the T-antigen polypeptides were detected with PAb101 (A and B) or PAb419 (C) in a Western blot assay. A small amount of WT T antigen and 246-708 was also added directly to the gels to serve as markers (last lanes).

We have previously demonstrated that the binding between T antigenand topo I under physiological conditions is highly specific(14). topo I bound to T antigen only when the viral proteinwas associated with SV40 origin DNA in the form of double hexamers.No binding was detected with smaller oligomeric forms of T antigensuch as single hexamers. These observations imply that the bindingobserved between these proteins under the in vitro conditionsused here (e.g., ELISAs and GST-binding assays) is not physiological.Therefore, we sought to determine whether the binding site identifiedabove is a true binding site that would function under DNA replicationconditions. To investigate this question, we incubated WT Tantigen, origin DNA, and various GST-topo I mutants under replicationconditions as described by Gai et al. (14). The origin DNA usedwas a 388-bp fragment of SV40 DNA encompassing the origin andcircularized by hybridizing a double-stranded oligonucleotidelinker as described previously (14). The protein-DNA complexesthat formed were cross-linked with glutaraldehyde and appliedto a composite polyacrylamide agarose gel. After electrophoresis,the complexes were transferred to nitrocellulose, and the topoI detected by reactivity with an anti-GST antibody. When topo17, B7, A2, or D26 was used, there was no difference betweenthe protein bands that formed in the presence or absence ofT antigen (Fig. 4). However, a band corresponding to the positionof double hexamers was detected with topo 70 and topo 31 inthe presence of T antigen but not in its absence (the faintband in the topo 31 -T lane is further up on the gel and mustrepresent a background topo 31-DNA complex). Therefore, onlytopo 70 and topo 31 bound to double hexamers. These data supportthe conclusion that the T-antigen-binding site is located withintopo 31. The failure of topo 17, B7, and D26 to bind to WT Tantigen under these conditions indicates that binding is dependenton an intact topo 31 domain structure.

View larger version (38K):
[in this window]
[in a new window]
FIG. 4. Origin DNA-binding assay with various topo I fragments. A total of 100 ng of GST-tagged topo 70 or the molar equivalent of various topo deletion mutants was incubated with 5 ng of circular origin DNA in the presence or absence of 400 ng of WT T antigen under replication conditions. After cross-linking, the samples were subjected to electrophoresis in a composite acrylamide agarose gel, and the protein complexes were transferred to nitrocellulose. topo I polypeptides were detected by reactivity with an anti-GST antibody in a Western blot assay. The positions of double hexamers (DH) are indicated and were determined by probing the same blots with an anti-T antibody (not shown).

Mapping of the N-terminal binding site on T antigen. Our previous data indicated that the N-terminal binding siteon T antigen is located between residues 1 and 246 (35). Tomap this site more closely, we generated a series of N-terminalT-antigen constructs (Fig. 5) and purified the deletion mutantsby immunoaffinity chromatography by using an appropriate monoclonalantibody. ELISAs measuring the binding between full-length topoI and these T antigen mutants are shown in Fig. 6. All T-antigenconstructs bound to topo I with the apparent exception of 1-131.The apparent weaker binding of 83-246 to topo I is probablydue to lower stability of this fragment since only small amountswere purified. Since mutants 83-246 and 1-160 bound topo I,the N-terminal binding site for topo I appears to be locatedbetween residues 83 and 160.

View larger version (10K):
[in this window]
[in a new window]
FIG. 5. (Top) Map of T antigen polypeptides used in this study. All of these proteins were expressed in baculovirus-infected insect cells. Nontagged polypeptides were purified by immunoaffinity chromatography, and GST-tagged proteins were purified by binding to glutathione-Sepharose beads.

View larger version (12K):
[in this window]
[in a new window]
FIG.6. (Bottom) ELISA between topo 70 and various N-terminal T antigen constructs. A total of 100 ng of topo 70 was first bound to microtiter plates. After a blocking step, various amounts of WT T antigen, as shown, or the molar equivalent of various N-terminal constructs of T antigen were added. After the wells were washed, PAb416 (which reacts with all T constructs used here) was added, followed by horseradish peroxidase-conjugated anti-mouse IgG. The absorbance at 490 nm was measured as described in Fig. 2. Controls without topo 70 were also performed for each T antigen construct, and these numbers were subtracted from the corresponding experimental numbers.

GST-binding assays were then performed to corroborate thesefindings (Fig. 7). Again, all N-terminal constructs except for1-131 bound to GST-tagged topo 31 (Fig. 7A). Small T antigen(a gift of Kathy Rundell), which shares the first 82 amino acidswith large T antigen, also did not bind (Fig. 7B). Therefore,the results of the ELISAs and GST-binding assays indicate thatthe N-terminal binding site on T antigen is located betweenresidues 83 and 160 and that this site binds within topo 31(residues 175 to 433).

View larger version (29K):
[in this window]
[in a new window]
FIG. 7. GST-binding assay between topo 31 and N-terminal T-antigen polypeptides and between small T antigen and topo I constructs. (A) A 10-µl portion of beads containing 2.5 µg of topo 31 was incubated with 500 ng of WT T antigen or the molar equivalent of various N-terminal T antigen constructs. The beads were washed and resuspended in electrophoresis sample buffer, and the samples were applied to a 13% acrylamide gel. Detection of the T-antigen polypeptides was performed with PAb419, which binds to the N-terminal region of T antigen. Samples of all polypeptides were loaded onto the gel to serve as markers of the input material (lanes on left). (B) Portions (10 µl) of beads containing 18 µg of GST-tagged topo 70 or the molar equivalent of topo 31, topo 17, topo 10, or GST alone were incubated with 150 ng of small T antigen (a gift of Kathy Rundell). Bound proteins were applied to a 13% acrylamide gel, and small T antigen on the membrane was detected with PAb419. A sample of small T antigen was loaded directly in the last lane.

Mapping of the C-terminal binding site on T antigen. The C-terminal binding site in T antigen exists between residues246 and 708 (35). In an initial attempt to map this region moreclosely, we generated a series of internal deletion mutantsof T antigen similar to those described by Kierstead and Tevethiawith deletions extending up to residue 550 (21). These deletionmutants were made in a background of the C-terminal construct246-708, which contains only one binding site. All of the internaldeletion mutants tested bound in ELISAs and GST-binding assays(data not shown), indicating that this second binding site islocated after residue 550. To probe this further, we made aseries of untagged T-antigen proteins from the C-terminal region(Fig. 5), and these were tested for binding to GST-tagged topo31 (Fig. 8). A peptide containing residues 423 to 708 boundwell, a finding consistent with our initial findings that thetopo I binding site is located between residues 550 and 708.Smaller untagged deletion mutants from the C-terminal regioncould not be made in stable form, and we therefore could notinitially refine this map.

View larger version (60K):
[in this window]
[in a new window]
FIG. 8. GST-binding assay between topo 31 and C-terminal T antigen constructs. A total of 10 µl of agarose beads containing 2.5 µg of GST-topo 31 was incubated with 500 ng of WT T antigen (A) or the molar equivalent of either 371-708 or 423-708 T-antigen polypeptides (B). Bound T-antigen proteins were detected by reactivity with PAb101, which binds to the C-terminal end of T antigen. Samples of the input proteins are shown as indicated.

To map the C-terminal site more closely, we turned to makingGST-tagged T-antigen deletion mutants. Fortunately, these constructswere stably expressed in infected insect cells. GST pull-downassays (Fig. 9) showed that mutant 602-708 bound to WT topoI but that 636-708 did not bind any better than the GST control.Fragment 564-708 appeared to bind better to T antigen than fragment602-708 in this experiment, but in others the binding was onlyslightly elevated. These data therefore indicate that the C-terminalbinding site is present primarily within residues 602 to 708.

DISCUSSION

Top

Discussion

topo I binds to two sites on T antigen. In an effort to more fully understand the molecular interactionsthat take place during the early phases of SV40 DNA replication,we mapped the sites in T antigen and topo I that bind to oneanother. The binding of these proteins to one another is anearly event in DNA replication, and it is very likely that bothproteins are components of the initiation complex (14, 44).We report here that there are two independent topo I bindingsites on T antigen and that both sites interact with a regionof topo I specified by topo 31 (residues 175 to 433). The twobinding sites are near the N-terminal (residues 83 to 160) andC-terminal (residues 602 to 708) ends of T antigen. The N-terminalbinding site contains a number of known features, includingthe nuclear localization signal and the retinoblastoma bindingpocket, as well as the beginning of the origin binding domain(33). This region also contains a number of sites that, whenphosphorylated, activate or inactivate DNA replication (33).We have previously reported that topo I binding does not interferewith the activity of the origin binding domain (36), but itseffect on other activities is not known. The C-terminal bindingsite in T antigen also contains sites of phosphorylation (33),as well as a region whose C-terminal end is at residue 626 or627 and that is required for structural integrity and stabilityof T antigen (H.-J. Lin and D. T. Simmons, unpublished results).Therefore, the C-terminal binding site is likely to extend toat least residue 626 but probably does not extend further thanresidue 670 since deletions from the C terminus to this pointdo not impair DNA replication (31).

Since physiologically relevant binding occurs only with doublehexamers of T antigen, it is important to determine where thesesites are located in the double hexamers. VanLoock et al. (48)have examined T-antigen double hexamers by electron microscopy,followed by three-dimensional reconstruction, and they reportthat the N-terminal and C-terminal ends of T antigen are localizedto each end of individual hexamers. Hexamers are oriented relativeto one another with either N-terminal or C-terminal domainsfacing each other, although it is not known if both are functional.Therefore, regardless of the orientation of the hexamers relativeto each other, there would be a topo I binding site on eachend of a double hexamer and potentially two binding sites inthe middle. However, stoichiometry analyses (D. Gai and D. T.Simmons, unpublished results) shows that only two moleculesof topo I are present in the complex per double hexamer. Toensure that there is a molecule of topo I in close proximityto each replication fork during DNA unwinding, we propose thattopo I is positioned on the outside of the double hexamer (Fig.10A). Accordingly, one molecule of topo I would be associatedclose to each end of double hexamers during the process of unwindingthe DNA, notwithstanding the hexamer-hexamer orientation (Fig.10A).

View larger version (39K):
[in this window]
[in a new window]
FIG. 10. (A) Model of topo I associated with a T-antigen double hexamer. This model is based on the data presented by VanLoock et al. (48) and Valle et al. (47). Each hexamer of T antigen consists of three domains: an N-terminal DNA-binding domain, a central helicase domain, and a C-terminal domain. The hexamers are oriented either with their C-terminal domains facing each other (model on left) or with the N-terminal domains facing one another (model on right). topo I is shown binding to the N-terminal domains on left or to the C-terminal domains on right. (B) topo 31 is a distinct domain of topo I. This illustration is based on the topo 70 crystal structure by Rebindo et al. (32) and Stewart et al. (43) (pdb entry 1a36). The topo 31 domain (the cap region) is shown below the DNA.

topo I domain that interacts with T antigen. Topo 31 is a distinct domain in the three-dimensional structureof topo I (32, 43) (Fig. 10B). Since the regions of T antigenwith which it interacts are completely different in sequenceand presumably in structure, it is likely that different portionsof topo 31, also known as the cap region (53), are involved.The binding of T antigen to this region should not interferewith the catalytic domain that is located in the section ofthe protein shown above the DNA in Fig. 10B. However, the capregion is required for nicking DNA (but not for the religationstep) during catalysis (53) and, therefore, T antigen must notbe blocking the cap region's ability to participate in the nickingreaction needed to relax supercoiled DNA.

Preliminary data from our lab (data not shown) indicate thatDNA length is a factor in the recruitment of topo I in the contextof the double hexamer. This suggests to us that some of thetopo I may be recruited by first binding to the DNA before lockinginto position on T antigen. Therefore, depending on the orientationof the hexamers, topo I would attach to the N-terminal or C-terminaldomain of T antigen. It would make sense if topo I itself facesT antigen differently in each case so that different surfacesof the cap region make contact with T antigen. Interestingly,since partially assembled double hexamers do not stably bindtopo I (14), fully formed double hexamers must undergo a structuralrearrangement at both ends in order to accommodate topo I.

Protein-protein interactions during DNA replication. Our data have further implications for protein-protein interactionsduring SV40 DNA replication. In the VanLoock structure, theregions of the DNA-binding domain that make direct contactswith the DNA are located on the inside of the channel formedby each hexamer. These regions correspond to residues 151 to160 and residues 210 to 215 (34, 37, 38). In order for topoI to bind to the N-terminal site on T antigen, it would mostlikely have to bind to regions exposed on the outside of thehexamer, probably corresponding to residues 83 to 140, basedon the nuclear magnetic resonance structure of the DNA-bindingdomain (25) and on the structure of residues 7 to 117 (22).In the double-hexamer reconstructions (48), this region liesclose to but not right at the end. Information is availableabout the binding sites of the other two cellular proteins (RPAand pol/prim) needed for initiation of DNA replication. TheRPA binding site has been mapped to residues 164 to 249 on Tantigen (50). This turns out to be immediately adjacent to thetopo I binding site and also within the N-terminal domain inthe T-antigen hexamer. These two cellular proteins may thereforebe in close proximity on the surface of the hexamer and maywell interact with one another. The binding site for DNA polymeraseappears to be between residues 195 and 313 (8), a region thatwould overlap with the N-terminal and central helicase domainsof each hexamer (see Fig. 10A) (48). Since DNA polymerase andRPA are known to interact with one another (5, 9, 29), the bindingof these two cellular proteins to T antigen might place themat the correct distance from each other so they could interactwith one another and also permit the association of topo I withthe N-terminal binding region. On the other hand, binding oftopo I to the C-terminal domain of T antigen would place itat some distance from the other cellular proteins. However,it is this interaction that is likely to be the more functionallysignificant one because the arrangement of the two hexamerswith the N-terminal domains facing one another is the more probableorientation given the structural architecture of the SV40 origin.If this is true, it remains to be determined whether the N-terminalbinding site has any function in this particular hexamer-hexamerorientation. One possibility is that it is used to facilitatethe recruitment of topo I to generate initiation complexes.

ACKNOWLEDGMENTS

This work was supported by grant CA36118 to D.T.S. from theNational Cancer Institute and grant GM60330 to J.J.C. from theNational Institutes of Health.

FOOTNOTES

* Corresponding author. Mailing address: Department of Biological Sciences, University of Delaware, Newark, DE 19716-2590. Phone: (302) 831-8547. Fax: (302) 831-2281. E-mail: dsimmons{at}udel.edu.

REFERENCES

Top