Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Structure of the replicative helicase of the oncoprotein SV40 large tumour antigen

A Corrigendum to this article was published on 04 September 2003


The oncoprotein large tumour antigen (LTag) is encoded by the DNA tumour virus simian virus 40. LTag transforms cells and induces tumours in animals by altering the functions of tumour suppressors (including pRB and p53) and other key cellular proteins. LTag is also a molecular machine that distorts/melts the replication origin of the viral genome and unwinds duplex DNA. LTag therefore seems to be a functional homologue of the eukaryotic minichromosome maintenance (MCM) complex. Here we present the X-ray structure of a hexameric LTag with DNA helicase activity. The structure identifies the p53-binding surface and reveals the structural basis of hexamerization. The hexamer contains a long, positively charged channel with an unusually large central chamber that binds both single-stranded and double-stranded DNA. The hexamer organizes into two tiers that can potentially rotate relative to each other through connecting α-helices to expand/constrict the channel, producing an ‘iris’ effect that could be used for distorting or melting the origin and unwinding DNA at the replication fork.


LTag, an oncoprotein encoded by polyomaviruses (reviewed in ref. 1), has diverse biological functions and also performs essential functions in replicating the viral DNA (reviewed in refs 2–4). During viral replication, LTag assembles at the origin of replication as a double hexamer that distorts and melts the origin DNA5,6,7. The LTag double hexamer is also the replicative helicase that unwinds duplex DNA at the replication forks8,9,10,11,12. The simian virus 40 (SV40) LTag is a 708-residue protein with multiple domains. Its helicase domain has been mapped to residues 131–627 (refs 13, 14), which contain a core-origin-binding domain (LTag-obd, residues 131–250)15,16 and three conserved motifs of the helicase superfamily III (SF3; ref. 17), which are also members of the AAA + superfamily18.

Here we report the 2.8-Å crystal structure of a LTag fragment that assembles into a hexamer with helicase activity. This is the first structure for a hexameric helicase that belongs to the helicase SF3 and AAA + superfamilies. The structure defines LTag hexamer assembly interactions and provides insight into LTag's role in cell transformation. Furthermore, multiple hexameric structures of LTag reveal conformational changes leading to the expansion and constriction of the central channel, suggesting a potential mechanism for origin melting and the unwinding of DNA at the replication fork.

Overall structure of LTag hexamer

The SV40 LTag structure contains residues 251–627 (LTag251–627). This LTag251–627 fragment has helicase activity (Fig. 1a), which is unexpected given that the minimal helicase domain was previously assigned to residues 131–627 (refs 13, 14). Hereafter we shall refer to LTag251–627 as the helicase domain, even though the fragment is also capable of transforming cells19,20 and actually contains three distinct structural domains.

Figure 1: The helicase activity and hexamer structure of LTag251–627.
figure 1

a, The helicase assay was performed with a fork DNA substrate containing 20-bp dsDNA, a 16-nucleotide 3′ overhang, and a three-nucleotide 5′ overhang labelled with 32P. All lanes contained 0.1 pmol DNA. Lanes 1 and 2, native and boiled DNA; lanes 3 and 4, 0.5 and 0.7 pmol LTag251–627 hexamer, respectively, showing helicase activity comparable to that of the full-length LTag (data not shown); lanes 5 and 6, 0.35 and 1.0 pmol LTag303–627, respectively, showing no helicase activity. b, c, Side (b) and top (c) views of the LTag251–627 hexameric structure. The pink balls represent Zn atoms. Each monomer is depicted with a distinct colour. Structures were generated with MOLSCRIPT49.

The LTag helicase domain assembles into a hexameric structure with a central channel (Fig. 1b, c). Viewed from the side, the hexamer ring appears as two layers (or tiers) of different diameters, the smaller tier at the top and the larger tier at the bottom (Figs 1b and 4a). Viewed from the top, the hexameric ring has six points radiating from the centre of the larger bottom tier (Fig. 1c). The amino-terminal domain located in the top tier reaches over clockwise to its neighbour, stacking on top of the larger carboxy-terminal domains of the neighbouring molecule. A cleft exists between each pair of monomers in the larger bottom tier (Fig. 1c), which might provide space for the hexamer to undergo an ATP-dependent conformational change, driving DNA remodelling.

Figure 4: The p53-binding surface.
figure 4

a, Side view of LTag hexamer surface, showing a patch formed by the four residues (in cyan) important for binding p53. b, c, The surface of the p53 DNA-binding domain (p53–DBD)39, showing the locations of residues important for binding LTag (magenta, in b), as well as for binding DNA (blue, in c). This shows that the two surfaces on p53–DBD overlap, indicating that LTag might inhibit p53 function by preventing its DNA-binding activity.

A gap separates the top from the bottom tier (Fig. 1b); α-helical structure connects the smaller and the larger tiers. The thickness of the two-tiered hexamer is 80 Å. Previous electron-microscope studies21 have shown that the thickness of a LTag hexamer is 120 Å, indicating that the missing N terminus (residues 1–250) would extend 40 Å (ref. 22) on top of the smaller tier. This dimension of LTag (120 Å per hexamer, or 240 Å per double hexamer) is similar to that of an archaeal MCM double hexamer23 and suggests that a LTag double hexamer at the replication origin should cover 70-base-pair (bp) double-stranded DNA (dsDNA), which is consistent with the observation that LTag protects 74-bp (or about seven turns) dsDNA16. Similarly, Xenopus MCM protects 80-bp DNA24.

LTag monomer structure

The LTag monomer contains 16 α-helices and 5 β-strands (Fig. 2c). These structural elements fold into three domains: D1 (domain 1), D2 and D3 (Fig. 2a, b). D1 is the Zn domain at the N terminus and contains five α-helices (α1–α5). The Zn atom coordinated by a Zn motif is important in holding α3 (α-helix 3) and α4 together (Fig. 2a), which in turn provide an anchor for α1 and α2. The beginning of α5 packs with α1 and α2 of D1, but its C terminus extends to α6 of D3.

Figure 2: Detailed structure of LTag251–627 monomer.
figure 2

a, b, Two views of the LTag251–627 monomer that are related by a 60° rotation along the horizontal axis. The three domains D1, D2 and D3 are indicated. α-Helices and β-strands are numbered sequentially. c, The secondary structure elements of LTag. Open bars represent β-strands, arrows α-helices, solid lines coiled structure, and the dotted line represents a flexible region.

The D2 domain contains three conserved helicase motifs related to SF3 helicases, namely the modified version of Walker A and B motifs and a motif C (ref. 17). D2 folds into a core β-sheet consisting of five parallel β-strands sandwiched by α-helices (Fig. 2b). The five-β-stranded fold is similar to the AAA + module of other AAA + family members, such as NSF-D2 (membrane fusion25), HslU (polypeptide unfolding and translocation26,27) and RuvB (branch migration28), all of which form hexamers. In addition to the five-β-stranded sheet, the LTag D2 AAA + module has two antiparallel β-strands (β3′ and β3″; Fig. 2a, b) unique to LTag among the known structures of the AAA + family. D2 also contains four helices, α9–α12. Helices α10–α12 are located on one side of the parallel β-sheet, and α9 is on the other (Fig. 2b). The loop between β1 and α9 is the P-loop (Walker A motif), which functions to bind and orient the triphosphate of NTP for hydrolysis.

The third domain, D3, is all α-helical. Its seven α-helices originate from both the N-terminal region (α6–α8) and the C terminus (α13–α16), with D2 inserted between. One unique feature of the structure of D3 is that the three helices from the N-terminal region (α6–α8) circle to form a nearly closed ring that is ‘crosslinked’ with another ‘ring’ composed of four α-helices at the C terminus (α13–α16; Fig. 2a). The two ‘crosslinked’ α-helical rings of D3 pack tightly against D2–α9, the ‘P-loop’ helix important for ATP binding, forming a large globular bulge on one side of the D2 β-sheet (Fig. 2a). In contrast, D1 and D2 or D3 are well separated, connected by long helices (α5 and α6) that could allow motions between domains.


The structure of the LTag hexamer was determined from the P321 crystal form obtained in the absence of ATP, in which one asymmetric unit contains two LTag molecules (Fig. 3a). The two molecules in one asymmetric unit are nearly identical, and the backbone can be superimposed with a root-mean-square deviation of 0.49 Å. Three dimers come together around the crystallographic 3-fold axis to form a hexamer (Fig. 1c), which should reflect the hexamer formed in solution in the absence of ATP (see next section). The monomer–monomer interactions within a hexamer come from two interfaces, the D1–D1 interface in the smaller tier (Fig. 3a, top) and the D2–D2 interface in the larger tier (Fig. 3a, bottom). D3 does not participate directly in the interactions at the interface. Rather, it constitutes the outer layer of the larger tier, giving the appearance of triangular points radiating from the centre (Fig. 1c).

Figure 3: Hexamerization of LTag.
figure 3

a, An overview of the interface between two monomers (in red and green) in a hexamer, showing two separate interfaces (boxed regions): D1–D1 and D2–D2. b, The D1–D1 interface. α5 and α2 from two monomers pack against each other through their leucine-rich hydrophobic surfaces. c, The D2–D2 hydrophobic interface consisting of charged or polar side chains. d, The oligomeric states of LTag251–627 (upper) and LTag303–627 (lower) in solution. Both polypeptides were incubated in buffers without ATP (blue line) or with 1 mM ATP (red line) at 25 °C for 30 min before being loaded on a Superdex-200 column. The apparent molecular masses (kDa) of the peaks are indicated. The calculated molecular mass for LTag251–627 is 43.2 kDa per monomer (259.4 kDa per hexamer) and for LTag303–627 it is 37.2 kDa per monomer. The molecular masses and positions of standards (green arrows) are indicated.

The D1–D1 and D2–D2 interfaces, which bury a total surface area of 2,529 Å2 between two monomers, involve two distinct types of interaction. The D1–D1 interface is uniformly hydrophobic, involving a leucine-zipper-like interaction between α2 of one molecule and α5 of its neighbour (Fig. 3b). Flanking the leucine-rich interface of α2 and α5, charged residues use their hydrophobic carbon side chains to pack against the hydrophobic residues on the neighbouring molecule. In contrast, the D2–D2 interactions are uniformly hydrophilic, involving hydrogen bonds and salt bridges through polar and charged residues (Fig. 3c). A total of 19 charged residues are present at this interface, plus several additional polar residues, including serine, asparagine and glutamine.

The role of the Zn motif and Zn domain

The Zn motif of LTag was proposed to form a canonical zinc-finger structure for DNA binding2,29. However, the LTag251–627 structure reveals that the Zn domain (D1) has a globular fold (Fig. 2a) stabilized by the coordination of a Zn atom through the Zn motif (C302·C305·H313·H317) that deviates from the previously predicted motif (C302·C305·H317·H320; ref. 29), and no classical zinc-finger structure specialized for DNA binding is present. Furthermore, the helices coordinated by the Zn motif are located away from the central channel of the hexamer (Fig. 1c), which probably accommodates DNA. We therefore conclude that the Zn motif is not directly involved in binding DNA but is instead important for stabilizing the Zn-domain structure.

The structure of the LTag251–627 hexamer suggests that the Zn domain might be important for LTag hexamerization. To test this prediction, a deletion mutant containing residues 303–627 (LTag303–627) was constructed. LT303–627 lacks α1 and α2 (important for D1–D1 interactions; Figs 2a and 3b) and C302 of the Zn motif, therefore lacking an intact Zn domain. The oligomeric states of LTag303–627 and LTag251–627 were examined by gel-filtration chromatography. In the absence of ATP, LTag251–627 eluted in two peaks with an apparent molecular mass corresponding to hexamers and monomers in a 2:1 ratio (Fig. 3d). When the protein isolated from either the hexamer or the monomer peak was subjected to another round of chromatography, both the hexamer and monomer peaks reappeared in the same 2:1 ratio, indicating that LTag251–627 equilibrates between hexamers and monomer. However, if ATP was present, all protein was eluted in the hexamer peak (Fig. 3d). The full-length LTag behaved similarly to LTag251–627 (data not shown). This demonstrates that LTag251–627 contains all the elements needed for hexamerization and hexamerizes in the absence of ATP, even though such hexamers readily dissociate into monomers. ATP stimulates hexamerization and stabilizes the hexamers.

In contrast, LTag303–627, which contains D2 but lacks D1 (Zn domain), was eluted as one peak with an apparent molecular mass close to a monomer in gel-filtration chromatography, even in the presence of ATP (Fig. 3d). The monomeric state is consistent with its lack of ATPase activity (data not shown). Because ATP hydrolysis by LTag probably requires residues supplied in trans by a neighbouring molecule (see next section), only dimers or higher oligomers of LT303–630 should possess ATPase activity. LTag303–627 therefore seems to be monomeric, indicating that the D2–D2 interactions in LTag303–627 are insufficient for dimer or hexamer formation with or without ATP, despite the large D2–D2 interface seen in the structure.

ATP binding by LTag hexamer

The hexamer structure in the p321 crystal form does not contain ATP. At the D2–D2 interface within a hexamer, three charged residues, Lys 418, Arg 498 and Asp 499, reside on one side of the interface with their side chains pointing towards the P-loop of the neighbouring monomer without making bonding contacts with any residues. The binding of ATP/ADP and Mg2+ at the P-loop of the neighbouring monomer should allow the three residues to make contact with either the ATP/ADP or the associated Mg2+ ion, which should strengthen the interactions between monomers to form the hexamer. These additional interactions augmented by ATP binding could enhance the cooperative process in LTag hexamerization and might explain why LTag251–627 equilibrates between hexamers and monomers in the absence of ATP but exists as stable hexamers in the presence of ATP (Fig. 3d).

Besides interacting with the ATP bound by the neighbouring molecule, the three residues (Lys 418, Arg 498 and Asp 499) might also have a role in ATP hydrolysis. The importance of in trans contribution of residues from adjacent monomers for hydrolysing NTPs has also been observed for other NTPases, such as T7 gp4 hexameric helicases and GTPase30,31,32. In T7 gp4, an arginine residue (termed Arg ‘finger’) from a neighbouring molecule interacts with the γ-phosphate of ATP, permitting the efficient hydrolysis of ATP. Arg 498 of LTag occupies a position that might allow it to have an equivalent role to the Arg ‘finger’ for ATP hydrolysis.

Mutational inactivation in p53 binding and replication

The LTag structure provides a basis for understanding how particular mutations affect certain LTag functions. Mutations can be divided into three classes depending on whether they are defective in replication, in transformation or in both. The first class has mutations of residues in either D1 (including the Zn motif), affecting hexamerization, or in D2 and D3, disrupting ATPase activity and DNA replication29,33. Most temperature-sensitive mutants belong to this class34,35. The second class includes mutations of residues on the D3 surface that are important for the binding of p53 (including the four mutations P399L, D402E/N/H, C411R and P584L; Fig. 4a), disrupting cell transformation but not DNA replication19,36,37,38. The four residues form a patch on the outer surface of a hexamer, suggesting that they either contact p53 directly or are positioned at the LTag–p53 interface. On the p53 side, mapping the residues important for binding LTag to the known p53 structure39 shows an overlap within the DNA-binding surface of p53 (Fig. 4b, c), which is consistent with the observation that p53 binding by LTag prevents p53 from binding to its promoter, thus inhibiting p53 transactivation activity. The third class of mutations disrupts the D3 fold. D3 not only interacts with p53 and polymerase-α on its surface but also borders α9 and the P-loop (ATP-binding) of D2. Examples of this class of mutations are those with small deletions or insertions on α14 of D3 (Fig. 2a), which failed to replicate the viral DNA and lost their transformation activity40.

Features and functions of LTag channel

Positively charged channels and the large ‘chamber’. Two essential functions of LTag in DNA replication are melting the origin and unwinding the duplex DNA at the fork. Hexameric LTag has a central channel whose inner surface is strongly positively charged (Fig. 5a), indicating a possible function in binding DNA, which is reminiscent of the DNA-binding channel of the archaeal MCM double hexamer23. A new feature is the presence of a large ‘chamber’ in the middle of the channel (Fig. 5b), which is sufficiently wide (67 Å) to accommodate strand separation during origin melting and fork DNA unwinding (see model in Fig. 6). Additionally, there are six positively charged holes formed between monomers on the side wall of the large ‘chamber’ (Fig. 5b). These holes (side channels) could potentially be used as outlets for the unwound single-stranded DNA (ssDNA) (see Fig. 6e).

Figure 5: The channel features and DNA-binding activity of LTag hexamer.
figure 5

a, Surface representation of LTag hexamer with the smaller tier in front, showing the positively charged (blue) central channel. Red, negatively charged surface; white, uncharged surface (GRASP50). b, The interior of the central channel, showing the large ‘chamber’ and the side channel. Two monomers have been taken away from the front to show the vertical path of the central channel, which widens inside the larger tier, forming a ‘chamber’. c, d, Autoradiographs of gels showing LTag (LT) hexamer binding to ssDNA (c) or dsDNA (d). Lanes 1–5 and lanes 6–10 contained 0, 0.1, 0.4, 1.6 and 6.4 pmol, respectively, LTag251–627 hexamer.

Figure 6: Models for origin distortion and melting, and DNA unwinding, by LTag.
figure 6

a, LTag double hexamer assembled at origin dsDNA. The helicase domain (LTag251–627) is labelled, with the larger ring representing the larger tier. b, The model for origin distortion and melting by the LTag helicase domain through the ‘iris’ mechanism. The distorted AT sequence and melted early palindrome (EP) region are indicated by arrows (green). ce, A looping model for bidirectional DNA unwinding by LTag double hexamer. The colour coding of LTag domains is the same as in a. After LTag double hexamer assembly (c), the dsDNA is distorted and melted (d) by the ‘iris’ mechanism. One separated ssDNA strand is extruded through a side channel of the helicase domain to unwind the two forks bidirectionally (e).

DNA binding. The central-channel openings of the LTag251–627 hexamer differ for the two tiers (Fig. 5a). The smaller top tier has an opening of 23 Å between side chains at the narrowest point, a dimension sufficient to allow the passage of dsDNA. The larger bottom tier has an opening of 15.0 Å (Fig. 5a), which is wide enough for ssDNA, but not dsDNA despite the large ‘chamber’ in the middle of this tier, which could accommodate two separated ssDNA strands simultaneously (Fig. 5b). DNA-binding assays show that LTag251–627 binds ssDNA (Fig. 5c), which is consistent with previous reports14,41. The LTag251–627 hexamer also binds dsDNA (Fig. 5d), indicating that the LTag hexamer might undergo a structural rearrangement so that the narrowest openings of the central channel expand to accommodate dsDNA.

Observed changes of channel size. To examine whether the LTag hexamer can change channel openings, we determined the structures of two other LTag crystal forms obtained under different conditions. The channel openings of the two crystal forms as well as of the first crystal form (P321) differ significantly at the narrowest point. One structure has a narrower opening in the larger tier, 12.5 Å (compared with 15.0 Å in P321), but no change in the smaller tier, whereas the other structure has smaller channel openings for both tiers, with 10.1 Å (compared with 15.0 Å) for the larger tier and 20.1 Å (compared with 23 Å in P321) for the smaller tier at the narrowest point (Supplementary Figure). Significantly, the structures determined from individual crystals of the same crystal form grown under identical conditions, but soaked in different buffers, vary in channel openings by as much as 3.6 Å (data not shown). These observations suggest that the LTag hexamer is intrinsically dynamic, which is consistent with its role as a molecular machine. Examination of these structures reveals that the changes in channel dimensions are achieved partly through a slight untwist or twist between the smaller top tier and the larger bottom tier (Supplementary Figure). The untwist or twist motion between layers is also observed in the electron-microscope images of p97 hexamer42. Although binding to dsDNA can conceivably cause LTag to expand its central channel for dsDNA, these results show that the channel openings can expand or constrict even in the absence of dsDNA.

Channel expansion/constriction: the ‘iris'motion. Three structural features of LTag hexamer might provide a basis for the expansion and constriction of the central channel in a similar manner to the ‘iris’ (diaphragm) of a camera. First, each molecule has the D1 and the D2/D3 domains spanning the smaller and larger tiers in a twisted fashion, with D1 of one monomer extending into its neighbour (Fig. 1c), an arrangement potentially permitting an untwist or twist motion between domains embedded in the two separate tiers. Second, the connector between D1 and D2/D3, or between the smaller and large tiers, is a long α-helical structure (α5; Figs 1c and 2a), which might act like a spring allowing the two tiers to untwist and twist against each other. This untwist and twist motion between the tiers might cause the channel to expand and constrict like an ‘iris’. Third, the uniformly hydrophobic interactions at the D1–D1 interface and hydrophilic interactions at the D2–D2 interface (Fig. 3b, c) might allow the neighbouring monomers to slide against each other during the ‘iris’ motion in changing the hexameric channel openings.


On the basis of the ‘iris’ hypothesis, we propose a model for the untwisting or melting of the origin dsDNA (Fig. 6a, b). It is known that a LTag double hexamer untwists or melts the AT-rich sequence on one side of the core origin and the early palindrome sequence on the other side, a process requiring ATP binding but not hydrolysis6. The dimensions of the LTag double hexamer binding to the core origin DNA15 imply that the two helicase domains at both ends must capture both the AT-rich and early palindrome sequences (Fig. 6a). We propose that the two tiers of the helicase domain untwist and twist relative to each other through the ‘iris’ mechanism, distorting the duplex DNA captured within the channel (Fig. 6b). Accompanying the untwisting of the two tiers is the expansion of the channel opening, which might facilitate the separation of the two strands of the distorted dsDNA (Fig. 6b), leading to the eventual melting of the origin for replication initiation. The wide ‘chamber’ (67 Å in width) inside the channel (Fig. 5b) can provide sufficient space for strand separation of dsDNA.

Similarly, a repeated ‘iris’ motion, probably driven by ATP hydrolysis, could be employed for destabilizing dsDNA ahead of the fork for unwinding. Here we propose a model for the bidirectional DNA unwinding by a LTag double hexamer bound to the origin (Fig. 6c–e). First, LTag assembles at the origin by means of sequence-specific interactions mediated through LTag-obd (residues 131–250), forming a double hexamer that covers about seven turns of dsDNA (Fig. 6a, c). The helicase domain then uses the ‘iris’ mechanism to distort and melt the origin to initiate unwinding (Fig. 6d). Next, the LTag double hexamer unwinds dsDNA by means of the ‘iris’ mechanism at the two forks held together by the double hexamer (Fig. 6e), with dsDNA being pulled into the LTag double hexamer and the unwound ssDNA loop being extruded from the side wall of the double hexamer, a process probably powered by ATP hydrolysis.

The positively charged side channel (Fig. 5b) might provide an avenue for the exit of the unwound ssDNA loop (Fig. 6e). The exit of ssDNA through the side channel might have important implications. First, each LTag hexamer of a double hexamer would cover both ssDNA and dsDNA at a fork (Fig. 6e), which is consistent with previous DNase protection results10,43. Second, the contact with dsDNA ahead of the fork, as well as the holding of a ssDNA strand in one of the six side channels behind the fork, would allow the relative rotation between the two tiers (‘iris’ mechanism) within each LTag hexamer to operate in untwisting the dsDNA. This configuration permits each hexamer in a functional double-hexamer unit5,6,7 to unwind two growing forks bidirectionally and eliminates the need for the two hexamers of a double-hexamer unit to track separately along (or rotate around) the helical path of dsDNA in opposite directions. Third, the exit of ssDNA from the side channel suggests that a complete opening of the hexamer ring is not required for the release of ssDNA, although a partial opening between two N-terminal obd domains would be needed to allow the unwound ssDNA to pass and reach to the side channel of the helicase domain.

The ‘iris’ model presented here describes a molecular event for origin distortion and melting. Furthermore, our model proposes that the helicase domain of a LTag double hexamer distorts and unwinds the duplex DNA bidirectionally using the ‘iris’ mechanism, and that the unwound ssDNA loops are released from the side channel of the helicase domains, a hypothesis consistent with the ‘rabbit ear’ observed by electron microscopy43. The evidence that cellular MCM might also function as a double hexamer23 and that the cellular DNA replication occurs at defined foci on the nuclear matrix where replication proteins are assembled44 suggests that a mechanism similar to that proposed in the LTag models might exist for DNA replication in cellular chromosomes.


Protein purification, characterization and crystallization

Crystallized LTag251–627 was produced in an Escherichia coli expression system as a glutathione S-transferase (GST) fusion protein. The protein was first purified on a glutathione affinity column. After cleavage of the GST fusion with thrombin, LTag was further purified by Superdex-200 gel-filtration chromatography. The protein was concentrated to 20 mg ml-1 in a buffer containing 25 mM Tris-HCl, pH 8.0, 250 mM NaCl, 1 mM EDTA, 10 mM dithiothreitol. Crystals were grown at 4 °C by the hanging-drop vapour diffusion method. The P321 crystal grew from 20 mM dithiothreitol, 50 mM HEPES, pH 7.6, 105 mM NaCl, 5% ethanol, 1 mM EDTA. The cell constants for the P321 crystal are a = b = 120.3 Å, c = 133.2 Å, α = β = 90°, γ = 120°. Two other crystal forms were obtained at different pH values: pH 7.25 to give a C2 crystal form (a = 194.3 Å, b = 85.9 Å, c = 127.8 Å, α = γ = 90°, β = 128.8°) and pH 7.0 to give an R3 crystal form (a = b = 168.3 Å, c = 129.0 Å, α = β = 90°, γ = 120°).

Data collection, structure determination and refinement

All the data sets were collected at the SBC 19ID beamline at the Advanced Photon Source (Chicago, Illinois). Crystals were flash-frozen in liquid nitrogen in crystallization buffer supplemented with 25% glycerol. Data were processed with HKL2000 (Table 1; ref. 45). The structure of LTag was determined by the multiple anomalous dispersion (MAD) method with the program SOLVE46 using a Zn-MAD data set. A solvent flattening step with SOLVE based on the Zn-MAD phases yielded an electron density map containing regions of well-featured α-helices, which allowed the initial model building with the program O. Cyclic refinement and model building led to a final model that, after an individual B-factor refinement, has an Rfree of 27.9% for the 30.0–2.8-Å resolution data (Table 1). For the determination of the structures of the other two crystal forms, initial attempts of molecular replacement (AMoRe47, CCP4 program suite48) using the complete LTag251–627 model did not give obvious solutions, probably owing to the different domain orientations of LTag in these crystal forms. However, the correct solutions were obtained by the molecular replacement method when the LTag model minus the Zn domain (D1) was used as the search model.

DNA-binding assay

A 33-nucleotide oligomer labelled with 32P was used as a ssDNA substrate. To generate a dsDNA substrate, this labelled ssDNA was annealed to its complementary oligonucleotide. LTag was preincubated with 1 mM ATP so that LTag was in its hexameric form (see Fig. 3d). To perform the binding assay, a fixed amount (0.1 pmol) of ssDNA or dsDNA substrate was added to a solution containing 1 mM ATP, 50 mM Tris-HCl, pH 8.0, 200 mM NaCl and LTag251–627 protein in varied quantities (0.0, 0.1, 0.4, 1.6 and 6.4 pmol hexamer). After a 10-min incubation at 25 °C, the reaction mixture was loaded on a 6% native polyacrylamide gel for electrophoresis analysis. The gel was dried for autoradiography.

Helicase assay

The substrate for helicase assay was a short-fork DNA annealed from two oligonucleotides that had been purified with MonoQ column chromatography. The substrate DNA was incubated with LTag251–627 at 25 °C for 30 min in a buffer containing 50 mM HEPES, pH 7.5, 1 mM ATP, 3 mM MgCl2, 1 mM dithiothreitol and 50 mM NaCl. The reaction was terminated by adding a stop solution containing 100 mM EDTA, 0.5% SDS and 50% glycerol. Samples were analysed on a 12% native polyacrylamide gel in 1 × Tris/borate/EDTA running buffer. The unwinding of the substrate DNA was detected by autoradiography.


  1. Pipas, J. M. Common and unique features of T antigens encoded by the polyomavirus group. J. Virol. 66, 3979–3985 (1992)

    CAS  PubMed  PubMed Central  Google Scholar 

  2. Simmons, D. T. SV40 large T antigen functions in DNA replication and transformation. Adv. Virus Res. 55, 75–134 (2000)

    CAS  Article  PubMed  Google Scholar 

  3. Stillman, B. Smart machines at the DNA replication fork. Cell 78, 725–728 (1994)

    CAS  Article  PubMed  Google Scholar 

  4. Bullock, P. A. The initiation of simian virus 40 DNA replication in vitro. Crit. Rev. Biochem. Mol. Biol. 32, 503–568 (1997)

    CAS  Article  PubMed  Google Scholar 

  5. Mastrangelo, I. A. et al. ATP-dependent assembly of double hexamers of SV40 T antigen at the viral origin of DNA replication. Nature 338, 658–662 (1989)

    ADS  CAS  Article  PubMed  Google Scholar 

  6. Borowiec, J. A. & Hurwitz, J. Localized melting and structural changes in the SV40 origin of replication induced by T-antigen. EMBO J. 7, 3149–3158 (1988)

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  7. Joo, W. S., Kim, H. Y., Purviance, J. D., Sreekumar, K. R. & Bullock, P. A. Assembly of T-antigen double hexamers on the simian virus 40 core origin requires only a subset of the available binding sites. Mol. Cell. Biol. 18, 2677–2687 (1998)

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  8. Dean, F. B. et al. Simian virus 40 (SV40) DNA replication: SV40 large T antigen unwinds DNA containing the SV40 origin of replication. Proc. Natl Acad. Sci. USA 84, 16–20 (1987)

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  9. Wold, M. S., Li, J. J. & Kelly, T. J. Initiation of simian virus 40 DNA replication in vitro: Large-tumor-antigen- and origin-dependent unwinding of the template. Proc. Natl Acad. Sci. USA 84, 3643–3647 (1987)

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  10. Smelkova, N. V. & Borowiec, J. A. Synthetic DNA replication bubbles bound and unwound with twofold symmetry by a simian virus 40 T-antigen double hexamer. J. Virol. 72, 8676–8681 (1998)

    CAS  PubMed  PubMed Central  Google Scholar 

  11. Tsurimoto, T., Melendy, T. & Stillman, B. Sequential initiation of lagging and leading strand synthesis by two different polymerase complexes at the SV40 DNA replication origin. Nature 346, 534–539 (1990)

    ADS  CAS  Article  PubMed  Google Scholar 

  12. Tsurimoto, T. & Stillman, B. Replication factors required for SV40 DNA replication in vitro. II. Switching of DNA polymerase α and δ during initiation of leading and lagging strand synthesis. J. Biol. Chem. 266, 1961–1968 (1991)

    CAS  PubMed  Google Scholar 

  13. Wun-Kim, K. & Simmons, D. T. Mapping of helicase and helicase substrate-binding domains on simian virus 40 large T antigen. J. Virol. 64, 2014–2020 (1990)

    CAS  PubMed  PubMed Central  Google Scholar 

  14. Wu, C., Roy, R. & Simmons, D. T. Role of single-stranded DNA binding activity of T antigen in simian virus 40 DNA replication. J. Virol. 75, 2839–2847 (2001)

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  15. Borowiec, J. A., Dean, F. B., Bullock, P. A. & Hurwitz, J. Binding and unwinding—how T antigen engages the SV40 origin of DNA replication. Cell 60, 181–184 (1990)

    CAS  Article  PubMed  Google Scholar 

  16. Fanning, E. & Knippers, R. Structure and function of simian virus 40 large tumor antigen. Annu. Rev. Biochem. 61, 55–85 (1992)

    CAS  Article  PubMed  Google Scholar 

  17. Koonin, E. V. A common set of conserved motifs in a vast variety of putative nucleic acid-dependent ATPases including MCM proteins involved in the initiation of eukaryotic DNA replication. Nucleic Acids Res. 21, 2541–2547 (1993)

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  18. Neuwald, A. F., Aravind, L., Spouge, J. L. & Koonin, E. V. AAA + : A class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res. 9, 27–43 (1999)

    CAS  PubMed  Google Scholar 

  19. Beachy, T. M., Cole, S. L., Cavender, J. F. & Tevethia, M. J. Regions and activities of simian virus 40 T antigen that cooperate with an activated ras oncogene in transforming primary rat embryo fibroblasts. J. Virol. 76, 3145–3157 (2002)

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  20. Cavender, J. F., Conn, A., Epler, M., Lacko, H. & Tevethia, M. J. Simian virus 40 large T antigen contains two independent activities that cooperate with a ras oncogene to transform rat embryo fibroblasts. J. Virol. 69, 923–934 (1995)

    CAS  PubMed  PubMed Central  Google Scholar 

  21. Valle, M., Gruss, C., Halmer, L., Carazo, J. M. & Donate, L. E. Large T-antigen double hexamers imaged at the simian virus 40 origin of replication. Mol. Cell. Biol. 20, 34–41 (2000)

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  22. Luo, X., Sanford, D. G., Bullock, P. A. & Bachovchin, W. W. Solution structure of the origin DNA-binding domain of SV40 T-antigen. Nature Struct. Biol. 3, 1034–1039 (1996)

    CAS  Article  PubMed  Google Scholar 

  23. Fletcher, R., Bishop, B., Sclafani, R., Ogata, G. & Chen, X. The structure and function of MCM dodecamer from Archaeal M. thermoautotrophicum. Nature Struct. Biol. 10, 160–167 (2003)

    CAS  Article  PubMed  Google Scholar 

  24. Edwards, M. C. et al. MCM2–7 complexes bind chromatin in a distributed pattern surrounding ORC in Xenopus egg extracts. J. Biol. Chem. 277, 33049–33059 (2002)

    CAS  Article  PubMed  Google Scholar 

  25. Lenzen, C. U., Steinmann, D., Whiteheart, S. W. & Weis, W. I. Crystal structure of the hexamerization domain of N-ethylmaleimide-sensitive fusion protein. Cell 94, 525–536 (1998)

    CAS  Article  PubMed  Google Scholar 

  26. Sousa, M. C. et al. Crystal and solution structures of an HslUV protease–chaperone complex. Cell 103, 633–643 (2000)

    CAS  Article  PubMed  Google Scholar 

  27. Bochtler, M. et al. The structures of HsIU and the ATP-dependent protease HsIU–HsIV. Nature 403, 800–805 (2000)

    ADS  CAS  Article  PubMed  Google Scholar 

  28. Putnam, C. D. et al. Structure and mechanism of the RuvB Holliday junction branch migration motor. J. Mol. Biol. 311, 297–310 (2001)

    CAS  Article  PubMed  Google Scholar 

  29. Loeber, G., Parsons, R. & Tegtmeyer, P. The zinc finger region of simian virus 40 large T antigen. J. Virol. 63, 94–100 (1989)

    CAS  PubMed  PubMed Central  Google Scholar 

  30. Sawaya, M. R., Guo, S., Tabor, S., Richardson, C. C. & Ellenberger, T. Crystal structure of the helicase domain from the replicative helicase–primase of bacteriophage T7. Cell 99, 167–177 (1999)

    CAS  Article  PubMed  Google Scholar 

  31. Singleton, M. R., Sawaya, M. R., Ellenberger, T. & Wigley, D. B. Crystal structure of T7 gene 4 ring helicase indicates a mechanism for sequential hydrolysis of nucleotides. Cell 101, 589–600 (2000)

    CAS  Article  PubMed  Google Scholar 

  32. Scheffzek, K. et al. The Ras–RasGAP complex: structural basis for GTPase activation and its loss in oncogenic Ras mutants. Science 277, 333–338 (1997)

    CAS  Article  PubMed  Google Scholar 

  33. Farber, J. M., Peden, K. W. & Nathans, D. Trans-dominant defective mutants of simian virus 40 T antigen. J. Virol. 61, 436–445 (1987)

    CAS  PubMed  PubMed Central  Google Scholar 

  34. Loeber, G., Tevethia, M. J., Schwedes, J. F. & Tegtmeyer, P. Temperature-sensitive mutants identify crucial structural regions of simian virus 40 large T antigen. J. Virol. 63, 4426–4430 (1989)

    CAS  PubMed  PubMed Central  Google Scholar 

  35. Ray, S., Anderson, M. E., Loeber, G., McVey, D. & Tegtmeyer, P. Functional characterization of temperature-sensitive mutants of simian virus 40 large T antigen. J. Virol. 66, 6509–6516 (1992)

    CAS  PubMed  PubMed Central  Google Scholar 

  36. Lin, J. Y. & Simmons, D. T. The ability of large T antigen to complex with p53 is necessary for the increased life span and partial transformation of human cells by simian virus 40. J. Virol. 65, 6447–6453 (1991)

    CAS  PubMed  PubMed Central  Google Scholar 

  37. Lin, J. Y. & Simmons, D. T. Stable T–p53 complexes are not required for replication of simian virus 40 in culture or for enhanced phosphorylation of T antigen and p53. J. Virol. 65, 2066–2072 (1991)

    CAS  PubMed  PubMed Central  Google Scholar 

  38. Kierstead, T. D. & Tevethia, M. J. Association of p53 binding and immortalization of primary C57BL/6 mouse embryo fibroblasts by using simian virus 40 T-antigen mutants bearing internal overlapping deletion mutations. J. Virol. 67, 1817–1829 (1993)

    CAS  PubMed  PubMed Central  Google Scholar 

  39. Cho, Y., Gorina, S., Jeffrey, P. D. & Pavletich, N. P. Crystal structure of a p53 tumor suppressor–DNA complex: understanding tumorigenic mutations. Science 265, 346–355 (1994)

    ADS  CAS  Article  PubMed  Google Scholar 

  40. Peden, K. W., Srinivasan, A., Farber, J. M. & Pipas, J. M. Mutants with changes within or near a hydrophobic region of simian virus 40 large tumor antigen are defective for binding cellular protein p53. Virology 168, 13–21 (1989)

    CAS  Article  PubMed  Google Scholar 

  41. Wu, C., Edgil, D. & Simmons, D. T. The origin DNA-binding and single-stranded DNA-binding domains of simian virus 40 large T antigen are distinct. J. Virol. 72, 10256–10259 (1998)

    CAS  PubMed  PubMed Central  Google Scholar 

  42. Rouiller, I. et al. Conformational changes of the multifunction p97 AAA ATPase during its ATPase cycle. Nature Struct. Biol. 9, 950–957 (2002)

    CAS  Article  PubMed  Google Scholar 

  43. Wessel, R., Schweizer, J. & Stahl, H. 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. J. Virol. 66, 804–815 (1992)

    CAS  PubMed  PubMed Central  Google Scholar 

  44. Cook, P. R. The organization of replication and transcription. Science 284, 1790–1795 (1999)

    CAS  Article  PubMed  Google Scholar 

  45. Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, 307–325 (1997)

    CAS  Article  PubMed  Google Scholar 

  46. Terwilliger, T. C. & Berendzen, J. Automated structure solution for MIR and MAD. Acta Crystallogr. D 55, 849–861 (1999)

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  47. Navaza, J. AMoRe: an automated package for molecular replacement. Acta Crystallogr. A 50, 157–163 (1994)

    Article  Google Scholar 

  48. CCP4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D 50, 760–763 (1994)

    Article  Google Scholar 

  49. Kraulis, P. E. MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, 946–950 (1991)

    Article  Google Scholar 

  50. Nicholls, A., Sharp, K. A. & Honig, B. Protein folding and association: insights from the interfacial and thermodynamic properties of hydrocarbons. Proteins 11, 281–296 (1991)

    CAS  Article  PubMed  Google Scholar 

Download references


We thank L.G. Chen for her assistance in the artwork, R. Garcea and L. Chen for comments on the manuscript, other members of the X. Chen laboratory for help and input, staff at 19id and 14bmc in Argonne National Laboratory and at X25 in Brookhaven National Laboratory for assistance in data collection, and the UCHSC X-ray centre in the Biomolecular Structure Program for support. This work is supported in part by start-up and cancer-centre funds from UCHSC to X.C. and NIH-R01 to X.C., J.A.D. and E.F., and a DOE grant to R.Z. and A.J.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Xiaojiang S. Chen.

Ethics declarations

Competing interests

The authors declare that they have no competing financial interests.

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Li, D., Zhao, R., Lilyestrom, W. et al. Structure of the replicative helicase of the oncoprotein SV40 large tumour antigen. Nature 423, 512–518 (2003).

Download citation

  • Received:

  • Accepted:

  • Issue Date:

  • DOI:

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing