Phosphorylation of Simian Virus 40 T Antigen on Thr 124 Selectively Promotes Double-Hexamer Formation on Subfragments of the Viral Core Origin

Abstract

Cell cycle-dependent phosphorylation of simian virus 40 (SV40) large tumor antigen (T-ag) on threonine 124 is essential for the initiation of viral DNA replication. A T-ag molecule containing a Thr→Ala substitution at this position (T124A) was previously shown to bind to the SV40 core origin but to be defective in DNA unwinding and initiation of DNA replication. However, exactly what step in the initiation process is defective as a result of the T124A mutation has not been established. Therefore, to better understand the control of SV40 replication, we have reinvestigated the assembly of T124A molecules on the SV40 origin. Herein it is demonstrated that hexamer formation is unaffected by the phosphorylation state of Thr 124. In contrast, T124A molecules are defective in double-hexamer assembly on subfragments of the core origin containing single assembly units. We also report that T124A molecules are inhibitors of T-ag double hexamer formation. These and related studies indicate that phosphorylation of T-ag on Thr 124 is a necessary step for completing the assembly of functional double hexamers on the SV40 origin. The implications of these studies for the cell cycle control of SV40 DNA replication are discussed.

Initiation of DNA replication is a complicated and highly regulated process that takes place during the S phase of the cell cycle. Progress in understanding initiation events in eukaryotes includes the identification of many of the factors that catalyze nascent DNA synthesis (reviewed in references 6, 7, 19, 32 and 90). Moreover, the isolation of the origin replication complex (ORC) (2) and related factors (reviewed in references 22 and 90) has provided considerable insight into initiation of eukaryotic DNA replication. However, since origins of replication from higher eukaryotes have not been characterized (8, 19), much remains to be learned about the protein-DNA interactions that are responsible for the initiation of DNA replication in higher organisms.

Experiments conducted with viral model systems have overcome certain of these limitations and aided in efforts to understand the molecular interactions that are necessary to initiate DNA replication in eukaryotes (19). In several instances, the sequences that define viral origins of replication have been established and the protein-DNA interactions that take place at these sequences have been extensively characterized (19). One particularly useful viral model system is based on simian virus 40 (SV40) DNA replication in vitro (43, 83, 95). SV40 encodes an 82-kDa protein, termed T antigen (T-ag) (84), that plays a number of critical roles during initiation of DNA replication. The functions of T-ag during the initiation of viral DNA replication have been the topic of several reviews (4, 7, 26). Briefly, T-ag site specifically binds to the SV40 origin of replication as a monomer and, as a result of a series of additional protein-protein and protein-DNA interactions (reviewed in references 4 and 7), oligomerizes into a double hexamer (13, 15, 51, 71). The double hexamer that assembles on the SV40 origin, via cooperative interactions (59, 66, 89, 93), is a functional helicase (14, 29, 80, 82, 94) that is able to unwind the SV40 origin (14, 21, 96). At the molecular level, T-ag assembly and unwinding events are poorly understood. Progress in understanding these processes includes the determination of the solution structure of the T-ag origin binding domain (OBD) (48) and images of T-ag double hexamers assembled on the SV40 origin (88).

The initiation of SV40 DNA replication is highly regulated. One very important form of regulation is determined by the phosphorylation state of T-ag (for reviews, see references 24 to 26, 67, and 92). Of particular importance is phosphorylation of T-ag on threonine 124 (23, 38, 53–56, 77). Indeed, phosphorylation of T-ag on Thr 124 is the sole posttranslational modification required for origin-dependent unwinding (54) and DNA replication (53, 54, 56, 77). The enzyme that phosphorylates T-ag on Thr 124 has not been unequivocally identified (25); however, in vitro studies suggest that it is a member of the cyclin–cyclin-dependent kinase (CDK) complex (31, 53, 54, 56). In contrast to activation via Thr 124 phosphorylation, phosphorylation of serine residues 120, 123, 677, and 679 inhibits initiation of viral replication (9, 23, 30, 40, 58, 73, 75, 79, 89). Consistent with these findings, newly synthesized T-ag, phosphorylated at Thr 124 and Thr 701, has a higher affinity for SV40 DNA than older T-ag molecules that are also phosphorylated on numerous serine residues (63, 73). Models for the control of SV40 replication, via dephosphorylation of serine residues and phosphorylation of Thr 124, have been proposed (23–25, 67).

A mutant T-ag molecule, containing a threonine-to-alanine substitution at position 124 (T124A), has proven to be a useful reagent for studies designed to understand the role played by Thr 124 phosphorylation during initiation of replication. T124A molecules assemble both hexamers and double hexamers on the SV40 origin, have helicase activity, distort the structure of the core origin, bind to cellular proteins required for initiation, and yet are not able to support origin unwinding or DNA synthesis (23, 54, 56, 77, 89, 93). These studies indicate that the T124A mutant is defective at some point between T-ag binding and DNA unwinding.

Considerable effort has been expended in characterizing the SV40 core origin, the segment of DNA at which T-ag site specifically binds. The core origin contains three subdomains: a central region, termed site II, that is flanked by an AT-rich domain, and a second region termed the early palindrome (EP) (16, 17, 65). Site II consists of four pentanucleotides (GAGGC) arranged as inverted pairs that serve as binding sites for T-ag (18, 48, 85, 86). It has been reported that the entire core origin is not required for T-ag assembly (35, 39, 81a). Indeed, mutant origins containing single pentanucleotides support hexamer formation, while properly arranged pairs of pentanucleotides support double hexamer formation (35). Related studies demonstrated that the 64-bp core origin contains two separate assembly units for double hexamer formation. One consists of pentanucleotides 1 and 3 and the EP (called the penta 1,3 + EP assembly unit) while the second is composed of pentanucleotides 2 and 4 and the AT region (penta 2,4 + AT assembly unit) (39, 81a). Recent experiments indicate that on the penta 1,3 and EP assembly unit, the first hexamer assembles on pentanucleotide 1 and the second hexamer assembles on pentanucleotide 3. On the penta 2,4 + AT assembly unit, the first hexamer forms on pentanucleotide 4 and the second forms on pentanucleotide 2 (81a).

The 64-bp core origin is known to afford alternative modes of T-ag assembly (35, 39; Sreekumar et al., unpublished). In contrast, oligonucleotides containing individual assembly units restrict the positions at which hexamers and double hexamers can assemble. This is an obvious advantage when conducting studies designed to examine how double hexamer assembly is regulated. Therefore, we have used T-ag and the T124A mutant, along with subfragments of the core origin containing single assembly units (39, 81a), to reinvestigate the role of Thr 124 phosphorylation in the regulation of T-ag assembly. Results from these studies are presented herein.

MATERIALS AND METHODS

Commercial supplies of enzymes, DNA, reagents, and oligonucleotides.

T4 polynucleotide kinase was purchased from Promega. Oligonucleotides were synthesized on an Applied Biosystems 394 DNA synthesizer at the protein chemistry facility at Tufts University. The oligonucleotides were purified by electrophoresis through 10% polyacrylamide-urea gels and isolated using standard methods (70, 81).

Purification of wild-type T-ag and T-ag containing the T124A mutation.

SV40 T-ag was expressed in insect (Sf9) cells using a baculovirus expression vector containing the T-ag-encoding SV40 A gene (62) and purified by immunoaffinity techniques using the monoclonal antibody PAb 419 (20, 78, 95). Purified T-ag was dialyzed against T-ag storage buffer (20 mM Tris-HCl [pH 8.0], 50 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol [DTT], 0.1 mM phenylmethylsulfonyl fluoride, 0.2 μg of leupeptin per ml, 0.2 μg of antipain per ml, 10% glycerol) and frozen at −80°C until use. The T124A mutant was isolated as described above, using a baculovirus expression vector developed by L. Chen, R. Upson, and D. Simmons (unpublished; a similar vector was described by Moarefi et al. [56]). Protein concentrations were determined by the Bradford reagent (Bio-Rad), using bovine serum albumin as the standard.

Band shift assays.

Double-stranded oligonucleotides used as substrates in gel shift assays were formed by annealing complementary pairs of oligonucleotides in hybridization buffer (37). The double-stranded oligonucleotides were labeled at their 5′ ends with ³²P using standard procedures (70). The labeled oligonucleotides were electrophoresed on neutral 15% polyacrylamide gels (run in 1× Tris-borate-EDTA at ∼380 V, 25 mA, and 10 W), subjected to autoradiography, and gel fragments containing DNA of interest were subsequently removed. DNA was then eluted in oligonucleotide extraction buffer (70). After extraction with phenol-chloroform-isoamyl alcohol (25:24:1) and chloroform-isoamyl alcohol (24:1), the labeled oligonucleotides were precipitated with 100% (vol/vol) ethanol, washed with 70% (vol/vol) ethanol, and dissolved in deionized H₂O (∼25 fmol/μl). (Additional details related to oligonucleotide preparation are described by Sreekumar et al. [81].)

Band shift reactions (15, 49, 60) were conducted under replication conditions (95). The reaction mixtures (20 μl) contained 7 mM MgCl₂, 0.5 mM DTT, 4 mM AMP-PNP (a nonhydrolyzable analog of ATP), 40 mM creatine phosphate (pH 7.6), 0.48 μg of creatine phosphate kinase, 5 μg of bovine serum albumin, 0.8 μg of HaeIII-digested pBR322 DNA (∼2.5 pmol; used as a non-sequence-specific competitor), 25 fmol of labeled double-stranded oligonucleotide (∼10⁶ cpm/pmol), and 6 pmol of T-ag, the T124A mutant, or mixtures of the two (T-ag or the T124A mutant was the last component added to the reaction). After a 20-min incubation at 37°C, glutaraldehyde (0.1%, final concentration) was added, and the reaction mixtures were incubated at 37°C for an additional 5 min. Finally, 5 μl of 6× gel loading dye II (15% Ficoll, 0.25% bromophenol blue, 0.25% xylene cyanol [70]) was added to the reaction mixtures. Samples were then loaded on 4 to 12% gradient polyacrylamide gels (19:1 acrylamide-to-bisacrylamide ratio) and electrophoresed in 0.5× Tris-borate-EDTA for ∼95 min (∼500 V, 20 mA, and 10 W). The gels were dried, subjected to autoradiography, and subsequently placed in a PhosphorImager cassette. Products formed in the gel shift reactions were quantitated with a Molecular Dynamics PhosphorImager.

To quantitate the relative ability of T-ag and the T124A mutant to form double hexamers, the fraction of input DNA shifted into double hexamers (DH) was divided by the fraction of input DNA shifted into hexamers (H), generating the DH/H ratio. The relative binding ability of these proteins was also estimated using a modification of the method described by Virshup et al. (89). This method determines the affinities of T-ag, or T124A, for unoccupied and singly shifted DNA. K₁ is the apparent association constant for the reaction T + D ↔ TD (where T is six T-ag monomers, D is free DNA, and TD is the single hexamer-pentanucleotide complex). K₂ is the affinity constant for the reaction TD + T ↔ TDT (where TDT is the double hexamer complex formed on DNA containing a single assembly unit). Thus, K₂/K₁ = (TDT)(D)/(TD)². The effect of Thr 124 phosphorylation on double hexamer formation was estimated by comparing the K₂/K₁ ratios, or cooperativity indices, for T-ag and the T124A mutant.

Nitrocellulose filter binding of SV40 T-ag and T124A complexes.

The nitrocellulose filter assay for T-ag or T124A binding was based on previously published methods (5, 47, 52, 81). Reaction mixtures (20 μl) contained 7 mM MgCl₂, 0.5 mM DTT, 40 mM creating phosphate (di-Tris salt [pH 7.6]), 0.48 μg of creatine phosphate kinase, 0.2 mg of bovine serum albumin per ml, 0.8 μg of HaeIII-digested pBR322 DNA, 25 fmol of a given oligonucleotide (∼10⁶ cpm/pmol), 4.0 mM AMP-PNP, and either T-ag or the T124A mutant. After incubation for 20 min at 37°C, the mixtures were filtered under suction through alkali-treated nitrocellulose filters (Millipore type HAWP; pore size, 0.45 μm; stored in 100 mM Tris-HCl [pH 7.5]). The filters were then washed with 5 ml of 100 mM Tris-HCl [pH 7.5]), dried, and counted in a Beckman LS 3801 scintillation counter.

RESULTS

The extent to which phosphorylation of T-ag, particularly at Thr 124, regulates binding to the core origin is somewhat controversial. This situation reflects, in part, variations in experimental methods and the use of T-ags isolated from different expression vectors (23, 40, 53–57, 77). Therefore, using conditions that support SV40 replication, we analyzed whether Thr 124 phosphorylation affects hexamer formation on oligonucleotides derived from either the penta 1,3 + EP or the penta 2,4 + AT assembly unit (39, 81a).

Comparison of the abilities of T-ag and the T124A mutant to form hexamers on substrates containing single pentanucleotides.

In an initial series of experiments, we analyzed whether baculovirus-expressed T-ag, which is nearly quantitatively phosphorylated on Thr 124 (31), and baculovirus-expressed T124A are equally adept at forming hexamers on core origin subfragments containing single pentanucleotides. To conduct these experiments, band shift reactions were performed with oligonucleotides derived from the penta 1,3 and EP assembly unit: the 48-bp penta 1 + EP and 48-bp penta 1 + EPm (mutant) oligonucleotides (Fig. 1B, diagrams 1 and 2), the 48-bp penta 3 + EP and 48-bp penta 3 + EPm oligonucleotides (diagrams not shown), as well as the 47-bp control oligonucleotide (Fig. 1B, diagram 5). Reactions conducted with an oligonucleotide containing the core origin (Fig. 1A) served as a positive controls. To separate initial binding steps from subsequent remodeling and unwinding steps, which require ATP hydrolysis (14, 83, 96), all of the experiments reported here were conducted with the nonhydrolyzable ATP analog AMP-PNP.

FIG. 1.

Sequences of representative oligonucleotides used to characterize the ability of T-ag and the T124A to form hexamers on subfragments of the core origin. Names of the oligonucleotides are given at the right. (A) Sequences present in the 64-bp core oligonucleotide. Locations of the AT-rich regions, site II, and the EP regions are depicted. SV40 sequences are numbered as described elsewhere (87). Arrows depict the four GAGGC pentanucleotides within site II that serve as binding sites for T-ag, numbered as previously described (41). (B) Diagram D1 provides the sequence of the 48-bp penta 1 + EP oligonucleotide, a derivative of the right-side assembly unit that supports only hexamer formation (81a). Although not depicted, we also synthesized the 48-bp penta 3 + EP oligonucleotide. Diagram D2 presents the sequence of the 48-bp penta 1 + EPm (mutant) oligonucleotide; in this molecule, the wild-type EP sequence was replaced by transition mutations. Although not shown, we also synthesized the 48-bp penta 3 + EPm oligonucleotide. Diagram D3 provides the sequence of the 47-bp penta 4 + AT oligonucleotide, a derivative of the left-side assembly unit that supports hexamer formation (81a). An additional member of this class of molecules, the 47-bp penta 2 + AT oligonucleotide, was also synthesized but is not depicted. Diagram D4 presents the sequence of the 47-bp penta 4 + ATm (mutant) oligonucleotide; in this molecule, transition mutations were used in place of the wild-type AT-rich region. We synthesized, but do not depict, an additional molecule in this class, the 47-bp penta 2 + ATm oligonucleotide. Diagram D5 depicts the sequence of the 47-bp control oligonucleotide, a molecule used to measure non-sequence-specific binding to DNA. Finally, lowercase boldface letters represent transition mutations in particular pentanucleotides or the AT or EP flanking regions.

A band shift reaction conducted with T-ag and the 64-bp core origin oligonucleotide is shown in Fig. 2, lane 2; the products formed in this reaction, hexamers and double hexamers, have been described elsewhere (13, 66, 89). An identical reaction performed with the T124A mutant and the core origin oligonucleotide is presented in lane 3. As previously reported (56, 93), the T124A mutant forms hexamers and double hexamers on the core origin. Reactions performed with the 48-bp penta 1 + EP oligonucleotide plus T-ag and the T124A mutant are presented in lanes 5 and 6, respectively. Inspection of these lanes demonstrates that T-ag and the T124A mutant form hexamers at roughly equal levels (quantitated in Fig. 2B; see below). Evidence that sequence-specific, or perhaps conformation-dependent, interactions with the EP are important for hexamer formation, for both T-ag and the T124A mutant, is demonstrated by experiments conducted with the 48-bp penta 1 + EPm oligonucleotide (lanes 8 and 9, respectively). For both proteins, binding to this oligonucleotide was significantly reduced relative to the 48-bp penta 1 + EP oligonucleotide (compare lanes 5 and 6 with lanes 8 and 9; quantitated in Fig. 2B). Results from experiments conducted with the 48-bp penta 3 + EP oligonucleotide, and either T-ag or the T124A mutant, are presented in lanes 11 and 12, respectively. It is apparent that this oligonucleotide did not support significant levels of hexamer formation when incubated with either T-ag or the T124A mutant. The possibility that the low level of binding to the 48-bp penta 3 + EP oligonucleotide, by either T-ag or the T124A mutant, was due in part to interactions with the EP is supported by experiments conducted with the 48-bp penta 3 + EPm oligonucleotide (lanes 14 and 15, respectively). This molecule supported hexamer assembly at levels as low as the control oligonucleotide (see Fig. 2B for quantitation). Reactions in lanes 1, 4, 7, 10, and 13 were conducted in the absence of protein. PhosphorImager-based quantitation of three identical experiments (Fig. 2B) demonstrates that both T-ag and the T124A mutant bind at least 10-fold more efficiently to the 48-bp penta 1 + EP oligonucleotide than to the 48-bp penta 3 + EP oligonucleotide. The results also show that relative to the 48-bp penta 1 + EP oligonucleotide, binding to the 48-bp penta 1 + EPm oligonucleotide is reduced ∼7-fold for T-ag and ∼9-fold for the T124A mutant.

FIG. 2.

Representative gel mobility shift assay used to assess the ability of T-ag or the T124A mutant to bind to single pentanucleotides in oligonucleotides derived from the penta 1,3 + EP assembly unit. (A) As positive controls, band shift reactions were conducted with the 64-bp core oligonucleotide and either T-ag (lane 2) or the T124A mutant (lane 3). Reaction products formed with the 48-bp penta 1 + EP oligonucleotide and either T-ag or the T124A mutation are shown in lane 5 or 6, respectively. The products formed with the 48-bp penta 1 + EPm oligonucleotide and either T-ag or the T124A mutant are shown in lane 8 or 9, respectively. Additional experiments were conducted with the 48-bp penta 3 + EP oligonucleotide and either T-ag (lane 11) or the T124A mutant (lane 12). Related experiments were conducted with the 48-bp penta 3 + EPm oligonucleotide and T-ag (lane 14) or the T124A mutant (lane 15). An additional set of reactions was performed with 47-bp control oligonucleotide and both T-ag and the T124A mutant (data not shown). Reactions in lanes 1, 4, 7, 10, and 13 were conducted in the absence of protein. All experiments were performed with AMP-PNP and 6 pmol of either T-ag or the T124A mutant. (B) The data presented in panel A and data from two additional experiments (data not shown) were quantitated with a Molecular Dynamics PhosphorImager, and the results are presented in a histogram. The percentage of input DNA shifted into hexamer, with 6 pmol of either T-ag or the T124A mutant, and AMP-PNP is shown on the ordinate.

Figure 3 presents similar studies conducted with oligonucleotides derived from the 47-bp penta 2,4 and AT assembly unit (i.e., the 47-bp penta 4 + AT and 47-bp penta 4 + ATm oligonucleotides (Fig. 1B, diagrams 3 and 4), the 47-bp penta 2 + AT and 47-bp penta 2 + ATm oligonucleotides (diagrams not shown), and the 47-bp control oligonucleotide (Fig. 1B, diagram 5). As a positive control, T-ag was incubated with an oligonucleotide containing the core origin (lane 2). An identical reaction, conducted with the T124A mutant, is presented in lane 3. Reactions performed with the 47-bp penta 4 + AT oligonucleotide plus T-ag and the T124A mutant are presented in lanes 5 and 6, respectively. These experiments demonstrate that T-ag and the T124A mutant form hexamers on pentanucleotide 4 at roughly equal levels (quantitated in Fig. 3B). Evidence that sequence-specific or conformation-dependent interactions with the AT-rich region are important for hexamer formation is demonstrated by studies conducted with the 47-bp penta 4 + ATm oligonucleotide plus T-ag and the T124A mutant (lanes 8 and 9, respectively). For both proteins, binding to this oligonucleotide was significantly reduced relative to the 47-bp penta 4 + AT oligonucleotide (compare lanes 5 and 6 with lanes 8 and 9 [quantitated in Fig. 3B]). Results from experiments conducted with the 47-bp penta 2 + AT oligonucleotide plus T-ag and the T124A mutant are presented in lanes 11 and 12, respectively. Both T-ag and the T124A mutant bound to the 47-bp penta 2 + AT oligonucleotide, but at lower levels than to the 47-bp penta 4 + AT oligonucleotide (compare lanes 11 and 12 with lanes 5 and 6). Additional experiments were conducted with the 47-bp penta 2 + ATm oligonucleotide plus T-ag and the T124A mutant (lanes 14 and 15, respectively). Both T-ag and the T124A mutant bound to pentanucleotide 2 in a manner that is promoted by either the wild-type sequence, or conformation, of the AT region (compare lanes 11 and 12 with lanes 14 and 15). Reactions in lanes 1, 4, 7, 10, and 13 were performed in the absence of protein. PhosphorImager-based quantitation of three identical experiments (Fig. 3B) reveals that both T-ag and the T124A mutant bind the 47-bp penta 4 + AT oligonucleotide approximately threefold more than the 47-bp penta 2 + AT oligonucleotide. It is also clear from this histogram that for both T-ag and the T124A mutant, binding to the 47-bp penta 4 + ATm oligonucleotide is reduced approximately fivefold relative to the same oligonucleotide containing the wild-type AT sequence.

FIG. 3.

Representative gel mobility shift assay used to assess the ability of T-ag or the T124A mutant to bind to single pentanucleotides in oligonucleotides derived from the penta 2,4 + AT assembly unit. (A) Control experiments were conducted with the 64-bp core oligonucleotide and either T-ag (lane 2) or the T124A mutant (lane 3). Reaction products formed with the 47-bp penta 4 + AT oligonucleotide and either T-ag or the T124A mutation are shown in lane 5 or 6, respectively. Lanes 8 and 9 present the products formed in reactions containing the 47-bp penta 4 + ATm oligonucleotide plus T-ag and the T124A mutant, respectively. An additional set of experiments were conducted with the 47-bp penta 2 + AT oligonucleotide and either T-ag (lane 11) or the T124A mutation (lane 12). To assess the contribution of the AT-rich region to binding, additional reactions were conducted with the 47-bp penta 2 + ATm oligonucleotide and either T-ag (lane 14) or the T124A mutation (lane 15). An additional set of experiments was performed with the 47-bp control oligonucleotide and either T-ag or the T124A mutant (data not shown). Reactions in lanes 1, 4, 7, 10, and 13 were conducted in the absence of protein. All experiments were performed with AMP-PNP and 6 pmol of either T-ag or the T124A mutant. (B) The data presented in panel A and data from two additional sets of experiments (data not shown) were quantitated with a Molecular Dynamics PhosphorImager. The percentage of input DNA shifted into hexamers, with 6 pmol of either T-ag or the T124A mutant, in the presence of AMP-PNP is shown on the ordinate. The percentage of input DNA shifted into hexamers is higher in panel B than in Fig. 2B. For unknown reasons, a relatively high percentage of the oligonucleotides derived from the penta 1,3 + EP assembly unit were trapped in the wells (Fig. 2A). Therefore, we are reluctant to make quantitative comparisons between the experiments in Fig. 2 and 3. However, the filter binding assays presented in Fig. 5 indicate that patterns of hexamer formation on the 48-bp penta 1 + EP and 47-bp penta 4 + AT oligonucleotides are roughly equivalent.

The band shift reactions presented in Fig. 2 and 3 require cross-linking with glutaraldehyde (see Materials and Methods). Therefore, to confirm the results obtained in these studies, cross-linking-independent nitrocellulose filter binding assays were conducted with T-ag, the T124A mutant, and oligonucleotides derived from the penta 1,3 and EP and penta 2,4 and AT assembly units (i.e., the 48-bp penta 1 + EP and 47-bp penta 4 + AT oligonucleotides [Fig. 1B, diagrams 1 and 3, respectively]) along with the 48-bp penta 3 + EP and 47-bp penta 2 + AT oligonucleotides (diagrams not shown). Positive control experiments were conducted with the 64-bp core (Fig. 1A) and the 48-bp penta 1,3 + EP and 47-bp penta 2,4 + AT oligonucleotides (Fig. 4, diagrams 2 and 4). The 47-bp control oligonucleotide served as a negative control (Fig. 1B, diagram 5).

FIG. 4.

Sequences of representative oligonucleotides used to characterize the ability of T-ag and the T124A mutant to form double hexamers on subfragments of the core origin. Names of the oligonucleotides are presented at the right. Diagram D1 depicts sequences present in the 48-bp site II + EP oligonucleotide. Locations of the four GAGGC pentanucleotides in site II are depicted by arrows; the location of the EP is also indicated. Diagram D2 presents sequences present in the 48-bp penta 1,3 + EP oligonucleotide, a molecule containing an assembly unit for double hexamer formation that is located on the right side of the core origin (81a). Diagram D3 presents sequences comprising the 47-bp site II + AT oligonucleotide. Locations of the four pentanucleotides in site II and the AT-rich region are indicated. Diagram D4 presents sequences present in the 47-bp penta 2,4 + AT oligonucleotide, a molecule containing an assembly unit for double hexamer formation that is located on the left side of the core origin (81a). Lowercase boldface letters indicate transition mutations introduced at the indicated locations.

Comparison of the results from these studies for T-ag (Fig. 5A) and the T124A mutant (Fig. 5B) indicates that the two proteins bind very similarly to this set of oligonucleotides. Moreover, those oligonucleotides containing pentanucleotides proximal to the flanking sequences (i.e., the 48-bp penta 1 + EP and 47-bp penta 4 + AT oligonucleotides) bound T-ag and the T124A mutant as readily as those containing complete assembly units for double hexamers (i.e., the 48-bp penta 1,3 + EP and 47-bp penta 2,4 + AT oligonucleotides). In contrast, both T-ag and the T124A mutant bound relatively poorly to oligonucleotides in which the pentanucleotides were distal to the flanking sequences (i.e., the 48-bp penta 3 + EP and 47-bp penta 2 + AT oligonucleotides).

FIG. 5.

Filter binding assays to measure the relative abilities of T-ag and the T124A mutant to interact with subfragments of the core origin under replication conditions. The interaction of T-ag or the T124A mutant (0, 3, or 6 pmol) with 25 fmol of the indicated oligonucleotide was measured by nitrocellulose filter binding assays in the presence of AMP-PNP. The percentage of input DNA bound to a given filter was determined by scintillation counting. As a positive control, the interaction of T-ag and the T124A (0, 3, or 6 pmol) with the 64-bp core oligonucleotide was determined. As a negative control, identical reactions were conducted with the 47-bp control oligonucleotide.

Collectively, the studies in Fig. 2 to 5 indicate that on a given assembly unit, T-ag and the T124A mutant preferentially form hexamers on the pentanucleotide proximal to the flanking sequence, a result consistent with previous studies of T-ag assembly (39, 81a). Our studies also demonstrate that all events required for hexamer formation on single assembly units, including binding of T-ag monomers to individual pentanucleotides and subsequent oligomerization steps, are independent of the phosphorylation status of Thr 124. This observation confirms previous reports indicating that phosphorylation of Thr 124 is not required for assembly of single hexamers on the core origin (54, 56).

Comparison of the ability of T-ag and the T124A mutant to form double hexamers on substrates containing single assembly units.

Based on the results presented in Fig. 2 to 5 and certain earlier studies (23, 54, 56), we concluded that the T124A mutant is defective in its ability to initiate viral replication at a point subsequent to hexamer formation. To test this hypothesis, we used oligonucleotides containing single assembly units for double hexamers (39, 81a) in an additional series of band shift experiments.

Initial experiments (Fig. 6A) were conducted with the 48-bp site II + EP and 47-bp site II + AT oligonucleotides (Fig. 4, diagrams 1 and 3). As a positive control, reactions were conducted with an oligonucleotide containing the 64-bp core origin plus T-ag and the T124A mutant (lanes 2 and 3, respectively). Products of reactions performed with the 48-bp site II + EP oligonucleotide plus T-ag and T124A are shown in lanes 5 and 6, respectively. As reported elsewhere (39, 81a), T-ag forms both hexamers and double hexamers on this substrate (lane 5). In contrast, the T124 mutant accumulates hexamers and is obviously defective in the ability to form double hexamers (lane 6). Similar reactions, conducted with the 47-bp site II + AT oligonucleotide plus T-ag and the T124A mutant, are presented in lanes 8 and 9, respectively. Inspection of lane 8 confirms that upon incubation with T-ag, this core origin subfragment supported both hexamer and double hexamer formation (39, 81a). In contrast, inspection of lane 9 demonstrates that assembly of the T124 mutant is essentially limited to hexamer formation. To measure non-sequence-specific assembly events, an additional set of reactions was conducted with the 47-bp control oligonucleotide plus T-ag and the T124A mutant (lanes 11 and 12, respectively). Oligomerization on this substrate, by either T-ag or the T124A mutant, was negligible (quantitated in Fig. 6B). Reactions in lanes 1, 4, 7, and 10 were conducted in the absence of protein. PhosphorImager-based quantitation of three identical reactions is presented in Fig. 6B. On the 48-bp site II + EP oligonucleotide, the DH/H ratio (see Materials and Methods) for T-ag was ∼12-fold greater than the ratio for the T124A mutant. On the 47-bp site II + AT oligonucleotide, the DH/H ratio was also ∼12-fold greater for T-ag than for the T124A mutant. Comparison of the cooperativity indices (see Materials and Methods) (89) indicates that T-ag binds the 48-bp site II + EP oligonucleotide ∼18-fold more efficiently than the T124A mutant, and it binds the 47-bp site II + AT oligonucleotide ∼11.5-fold more efficiently than the T124A mutant. We conclude that the T124A mutant is defective in the ability to form double hexamers on duplex DNA molecules containing single assembly units. Finally, on the 64-bp core origin oligonucleotide, the T124A mutant had a slightly (∼3-fold) lower DH/H ratio than T-ag. This result, based on five identical experiments, is consistent with similar conclusions reported by Weisshart et al. (93).

FIG. 6.

Representative gel mobility shift assay used to assess the ability of T-ag or the T124A mutant to bind to oligonucleotides that support double hexamer formation. (A) Positive controls were conducted with the 64-bp core oligonucleotide and either T-ag (lane 2) or the T124A mutant (lane 3). Reaction products formed with the 48-bp site II + EP oligonucleotide plus T-ag and the T124A mutant are shown in lanes 5 and 6, respectively. Reactions conducted with the 47-bp site II + AT oligonucleotide plus T-ag and the T124A mutant are shown in lanes 8 and 9, respectively. To measure non-sequence-specific binding, additional reactions were conducted with the 47-bp control oligonucleotide (Ctrl.) and either T-ag (lane 11) or the T124A mutant (lane 12). Reactions were conducted with 6 pmol of either T-ag or the T124A mutant, in the presence of AMP-PNP; the reactions in lanes 1, 4, 7, and 10 were conducted in the absence of protein. (B) The data presented in panel A, and data from two additional sets of experiments (data not shown), were quantitated with a Molecular Dynamics PhosphorImager, and the results are presented in a histogram. The DH/H ratio was calculated as described in Materials and Methods.

The experiments presented in Fig. 6 were conducted at a protein-to-DNA ratio of 240:1 and with DNA substrates that contained four pentanucleotides. Therefore, to further analyze the apparent defect in double hexamer formation exhibited by the T124A mutant, additional experiments were conducted with oligonucleotides containing the 48-bp penta 1,3 + EP and 47-bp penta 2,4 + AT assembly units (Fig. 4, diagrams 2 and 4, respectively) over a range of protein concentrations.

Experiments conducted with the 48-bp penta 1,3 + EP oligonucleotide are shown in Fig. 7A. As a positive control, band shift reactions were conducted with the 64-bp core oligonucleotide and either T-ag (lane 2) or the T124A mutant (lane 3) at a protein-to-DNA ratio of 240:1. Reactions in lanes 5, 7, 9, and 11 were performed with T-ag at protein-to-DNA ratios of 240:1, 120:1, 60:1, and 30:1, respectively. It is obvious that as the protein-to-DNA ratio decreased, this assembly unit supported reduced levels of hexamers and double hexamers. Similar experiments, conducted with the T124A mutant at identical protein-to-DNA ratios, are presented in lanes 6, 8, 10, and 12. The data in Fig. 7 demonstrate that relative to T-ag, the T124A mutant is defective in double hexamer formation at all protein-to-DNA ratios. The reactions in lanes 1 and 4 were conducted with the indicated oligonucleotides in the absence of protein. When these results and those in Fig. 6 are viewed in terms of the experiments presented in Fig. 2 and 5, it seems likely that the T124A mutant accumulates hexamers on pentanucleotide 1 but is defective in forming the second hexamer on pentanucleotide 3. The experiments in Fig. 7A, and two additional sets of experiments, were quantitated with a PhosphorImager, and the DH/H ratios were determined. Results of these analyses (Fig. 7B) reveal that on the penta 1,3 + EP assembly unit, the T124A mutant is defective in double hexamer formation over a range of protein-to-DNA ratios.

FIG. 7.

Comparison of the abilities of T-ag and the T124A mutant to assemble on the penta 1,3 + EP assembly unit. (A) Results of experiments conducted with 25 fmol of the 48-bp penta 1,3 + EP oligonucleotide and different amounts of either T-ag or the T124A mutant (from 6 pmol of protein (240:1 protein-to-DNA ratio) to 0.75 pmol of protein (30:1 protein-to-DNA ratio) (lanes 5 to 12). As positive controls, reactions were conducted with the 64-bp core origin and either T-ag (lane 2) or the T124A mutant (lane 3) at a protein-to-DNA ratio of 240:1. The reactions in lanes 1 and 4 were conducted in the absence of protein. All reactions were performed in the presence of AMP-PNP. (B) The data in panel A were quantitated using a Molecular Dynamics PhosphorImager and used to calculate the DH/H ratio for both T-ag and the T124A mutant.

Similar titration experiments, conducted with the 47-bp penta 2,4 + AT oligonucleotide, are shown in Fig. 8A. As in previous experiments, a positive control was provided by conducting band shift reactions with the 64-bp core oligonucleotide and either T-ag (lane 2) or T124A (lane 3) at a protein-to-DNA ratio of 240:1. Reactions in lanes 5, 7, 9, and 11 were performed with T-ag at protein to DNA ratios of 240:1, 120:1, 60:1, and 30:1, respectively. The results show that at all protein-to-DNA ratios, this assembly unit supported formation of T-ag hexamers and double hexamers over a range of protein-to-DNA ratios (on a darker exposure, double hexamers could be detected at the 30:1 ratio). The reaction products presented in lanes 6, 8, 10, and 12 were formed in the presence of the T124A mutant, using the protein-to-DNA ratios described above. As in Fig. 7A, it is clear that relative to T-ag, the T124A mutant is defective in double hexamer formation at all protein-to-DNA ratios. The reactions in lanes 1 and 4 were conducted in the absence of protein with the indicated oligonucleotides. In light of the experiments presented in Fig. 3 and 5, it is probable that on this assembly unit, the T124A mutant accumulates hexamers on pentanucleotide 4 and is impaired in its ability to form the second hexamer on pentanucleotide 2. The experiments in Fig. 8A, and two additional sets of experiments, were quantitated with a PhosphorImager, and the DH/H ratios were determined. Results of these analyses (Fig. 8B) reveal that on the penta 2,4 + AT assembly unit, the T124A mutant is defective in double hexamer formation over a range of protein-to-DNA ratios. Based on the analyses presented in Fig. 6 to 8, we conclude that a phosphate on Thr 124 promotes protein-protein, or perhaps protein-DNA, interactions necessary for assembly of the second hexamer on either of the assembly units.

FIG. 8.

Comparison of the abilities of T-ag and the T124A mutant to assemble on the penta 2,4 + AT assembly unit. The experiments were conducted in the presence of AMP-PNP, with 25 fmol of the 47-bp penta 2,4 + AT oligonucleotide and different amounts of either T-ag or the T124A mutant (from 6 pmol of protein (240:1 protein to DNA ratio) to 0.75 pmol of protein (30:1 protein-to-DNA ratio) (lanes 5 to 12). As positive controls, reactions were conducted with the 64-bp core origin and either T-ag (lane 2) or the T124A mutant (lane 3) at a protein-to-DNA ratio of 240:1. The reactions in lanes 1 and 4 were conducted in the absence of protein. (B) The data in panel A were quantitated using a Molecular Dynamics PhosphorImager and used to calculate the DH/H ratio for both T-ag and the T124A mutant. Lower levels of double hexamer assembly on the penta 2,4 + AT assembly unit, relative to the penta 1,3 + EP assembly unit, are reflected in lower DH/H ratios (compare panel B with Fig. 7B).

T124A molecules are inhibitors of double hexamer formation on the penta 1,3 + EP assembly unit.

Previous mixing experiments of T-ag and the T124A mutant showed that increasing amounts of the T124A mutant suppressed origin-specific unwinding in a dose-dependent manner (93). Mixing experiments also demonstrated that the T124A mutant is a dominant-negative inhibitor of SV40 DNA replication in vitro (93). However, at the molecular level, it is not clear how the T124A mutant is able to inhibit these processes.

In view of the results presented above, it seemed possible that the T124A mutant would inhibit T-ag's ability to form double hexamers on individual assembly units. To test this hypothesis, an additional band shift experiment was conducted with the 48-bp penta 1,3 + EP oligonucleotide and mixtures of T-ag and the T124A mutant. The products formed in these experiments were quantitated with a PhosphorImager, and the levels of hexamers and double hexamers were determined. The results (Fig. 9) reveal that addition of increasing amounts of T124A blocked the formation of T-ag double hexamers. The block in T-ag double hexamer formation was accompanied by a concomitant increase in the level of hexamers. Both the inhibition of double hexamer formation and rise in hexamer assembly were linear with respect to the amount of T124A added to the reaction. These studies indicate that hexamers containing T124A molecules inhibit the formation of the second hexamer on the penta 1,3 + EP assembly unit. The ability of T124A molecules to act as inhibitors of the formation of T-ag double hexamers may explain previous results indicating that the T124A mutant is a dominant-negative inhibitor of SV40 origin unwinding and DNA replication.

FIG. 9.

Double hexamer formation on the penta 1,3 + EP assembly unit is suppressed by T124A. Band shift reactions were conducted with 25 fmol of the 48-bp penta 1,3 + EP oligonucleotide, AMP-PNP, and 6 pmol of T-ag (left side of graph), 6 pmol of T124A (right side of graph), or 6 pmol of T-ag and T124A combined in the indicated ratios. The products of these reactions were loaded on 4 to 12% gradient polyacrylamide gels, and the percentages of input DNA shifted into hexamers (open squares) and double hexamers (filled diamonds) were determined by a PhosphorImager.

DISCUSSION

Biochemical studies have demonstrated that pairs of pentanucleotides arranged in a precise head-to-head orientation are a prerequisite for double hexamer formation (35, 36, 81a). Moreover, recent transmission electron microscopy studies have shown that double hexamers assembled on properly arranged pairs of pentanucleotides also have a very precise spatial arrangement (88). Based on these considerations, we elected to reexamine how double hexamer formation is regulated on the SV40 core origin.

In one series of experiments using core origin subfragments, we determined that like T-ag, the T124A mutant preferentially binds to pentanucleotides proximal to the flanking sequences. Based on these studies, we conclude that Thr 124 phosphorylation is not necessary for binding of T-ag monomers to individual pentanucleotides or for the assembly of hexamers at these sites. These and related studies (54, 56, 93) indicated that the T124A mutant is defective in the initiation process at a step following hexamer assembly. Consistent with this hypothesis, we have demonstrated that the T124A mutant is impaired in its ability to form double hexamers on individual assembly units. Indeed, at a protein-to-DNA ratio of 240:1, T-ag formed double hexamers at least 12-fold more efficiently than the T124A mutant on the site II + EP and site II + AT oligonucleotides.

Furthermore, the relative abilities of T-ag and T124A to form double hexamers on subfragments of the core origin may be underestimated in our studies. This conjecture is based on the presence of hexamers in all of the reactions conducted with T-ag and substrates containing individual assembly units. This result is readily explained if our T-ag preparations contain low levels of T-ag molecules that lack phosphate at Thr 124 and are therefore inactive for double hexamer formation. It is noted that isolates of T-ag purified from baculovirus have been reported to be nearly quantitatively phosphorylated on Thr 124 (31). Nevertheless, we do not know the percentage of T-ag molecules that are phosphorylated at Thr 124 in our preparations. Moreover, the presence of phosphates on particular serines can also block double hexamer formation on the SV40 core origin (40, 58, 79, 89). Therefore, the failure of all of the hexamers in the T-ag containing reactions to mature into double hexamers (Fig. 6A, 7A, and 8A) may reflect the presence of molecules that are not phosphorylated at Thr 124 or are phosphorylated on inhibitory serines.

It was previously reported that the T124A mutant is slightly (∼2-fold) impaired in terms of its ability to form double hexamers on the core origin (56, 93). In keeping with these reports, our experiments with the 64-bp core origin oligonucleotide also indicate that the DH/H ratio is slightly (∼3-fold) lower for the T124A mutant than for T-ag. Given that we have detected a significant defect in the ability of T124A molecules to form double hexamers on oligonucleotides containing individual assembly units, it is of interest to consider why equally significant defects are not detected on the full-length core origin. Moarefi et al. (56) proposed that when T124A molecules are assembled on the 64-bp core origin, the two origin-bound hexamers make weak or inappropriate protein-protein interactions and consequently are unable to initiate origin unwinding. Our studies indicate, however, that in order for these inappropriate T124A hexamer-hexamer interactions to occur on the core origin, additional protein-DNA interactions must take place between T-ag and the additional DNA sequences present in the larger oligonucleotides. An alternative possibility is that the additional sequences present in the full-length, 64-bp oligonucleotides may allow the T124A mutant to oligomerize inappropriately on pairs of pentanucleotides that are not components of individual assembly units. For example, the T124A mutant may form double hexamers owing to simultaneous, though noncooperative, interactions with pentanucleotides 1 and 4. Indeed, it was previously demonstrated that a 64-bp mutant core origin containing pentanucleotides 1 and 4 supported T-ag double hexamer formation (35). It is also known that T-ag hexamers formed on any pentanucleotide distort the proximal flanking sequence (K. R. Sreekumar and P. A. Bullock, unpublished data). Therefore, T124A hexamers assembled on pentanucleotides 1 and 4 would engender the previously described structural alterations of the AT-rich and EP regions (54, 56). These and related possibilities need to be tested. Nevertheless, since they do not support DNA unwinding or replication, it is clear that T124A molecules form defective double hexamers on the full-length core origin (54, 56, 93).

The results from the in vitro experiments presented herein are summarized in Fig. 10. For simplicity, only the penta 1,3 + EP assembly unit is used to contrast the ability of T-ag or the T124A mutant to complete oligomerization. The sequence and protein-protein interactions necessary for hexamer formation are complex (11, 39, 81a; reviewed in reference 7). Our present results demonstrate, however, that all interactions leading to hexamer formation, initially on pentanucleotide 1 or less frequently on pentanucleotide 4, are independent of the phosphorylation status of Thr 124 (Fig. 10, line 1). Previous studies, conducted in the presence of the core origin (54, 56) and in the absence of DNA (68), also concluded that phosphorylation and ATP hydrolysis are not required for hexamer formation. In contrast to initial hexamer formation, our data indicate that phosphorylation of Thr 124 is a major determinant for subsequent assembly of the second hexamer (Fig. 10, line 2). Upon phosphorylation of Thr 124, the second hexamer appears to form on the penta 1,3 + EP assembly unit at pentanucleotide 3. On the penta 2,4 + AT assembly unit, the second hexamer forms preferentially on pentanucleotide 2. The number of phosphates required for double hexamer formation, and their spatial distribution on the two hexamers, has yet to be determined. However, the mixing experiments in Fig. 9 indicate that double hexamers can form at substoichiometric levels of Thr 124 phosphorylation.

FIG. 10.

Model illustrating the role of Thr 124 phosphorylation during T-ag oligomerization on the penta 1,3 + EP assembly unit. The T-ag OBD is symbolized by the smaller ovoids, while the remaining regions of T-ag are symbolized by the larger ovoids (88). Pathways for T-ag and T124A oligomerization are shown on the left and right. The drawings on line 1 indicate that on a given assembly unit, T-ag and the T124A mutant prefer to bind to the pentanucleotides proximal to the flanking sequences (pentanucleotide 1 or 4; binding to pentanucleotide 1 is depicted). Upon monomer binding, a complicated set of protein-protein interactions (reviewed in reference 7) gives rise to hexamer formation; however, all steps required for hexamer formation are independent of the phosphorylation status of Thr 124. Phosphate residues on Thr 124 (P) are indicated. Depicted on line 2 is the observation that double hexamer formation on the penta 1,3 + EP assembly unit is highly dependent on phosphorylation of Thr 124. As a result, T-ag phosphorylated on Thr 124 is able to assemble a double hexamer, but the T124A molecule is blocked at the level of hexamer formation. Similar results were obtained with the penta 2,4 + AT assembly unit (see text). Although monomers of T-ag are depicted as being phosphorylated on Thr 124, it is possible that other oligomeric forms of T-ag are the actual substrates for phosphorylation. Finally, double hexamer formation is inhibited by phosphorylation of certain serine residues (reviewed in references 24 to 26) (not illustrated).

Since T124 phosphorylation plays an important role in the formation of the second hexamer on individual assembly units, it is of interest to consider the possible interactions that this phosphorylation event regulates. Experiments employing class 4 T-ag mutants, which are located in the T-ag OBD and display properties similar to those of the T124A mutant (97), indicated that the T-ag OBD is involved in the protein-protein interactions necessary for double hexamer formation (93). Indeed, it was suggested that one surface of the T-ag OBD, in each of the six subunits of one hexamer, interacts with neighboring T-ag OBD subunits in the second hexamer (93). This conclusion is supported by transmission electron microscopy studies of T-ag double hexamers assembled on the SV40 core origin (88). That phosphorylation governs these interactions is supported by our studies and experiments implicating Thr 124 phosphorylation in cooperative assembly of T-ag double hexamers (56, 93). It is noted, however, that bacterially expressed T-ag OBD_131-260 and OBD_112-260, lacking phosphates and other posttranslational modifications, formed dimers on the SV40 core origin (36). Thus, phosphorylation of Thr 124 may not play a direct role in the T-ag OBD–T-ag OBD interactions that appear to be necessary for double hexamer formation. Alternatively, Thr 124 phosphorylation may be necessary for conformational changes in T-ag that enable interactions between T-ag OBDs domains on neighboring hexamers. In summary, progress has been made in defining the hexamer-hexamer interface, but the precise interactions, or conformation changes, governed by Thr 124 phosphorylation have yet to be determined.

We have considered our results, and those of previous studies (for reviews, see references 7, 24, 25, and 26), in terms of the cell cycle control of SV40 replication in vivo. Upon synthesis, T-ag is rapidly transported as monomers from the cytoplasm to the nucleus (76). Since hexamer formation is independent of the phosphorylation status of Thr 124, it is proposed that monomers form hexamers, and perhaps nonfunctional double hexamers, on the SV40 core origin at all stages of the cell cycle. It is noted that SV40 replication takes place only during the S phase of the cell cycle (64). This temporal control of DNA replication depends, in part, on the sequential activation of particular cyclin-CDK complexes (for recent reviews, see references 33, 61, and 69). While the kinase responsible for Thr 124 phosphorylation in vivo has yet to be rigorously established, experiments suggest that it may be the CDK2-cyclin A complex (1, 27). Collectively, these observations raise an important question: In what cellular location does Thr 124 phosphorylation takes place? It has been reported that Thr 124 is phosphorylated in the cytoplasm (74), relatively soon after its synthesis (72, 73). However, it has also been suggested that the bulk of the T-ag population enters the nucleus prior to being phosphorylated on Thr 124 (34, 53). The finding that the bulk of the G₁/S-phase cyclin-CDK complexes are located in the nucleus (for a review, see reference 98) is consistent the hypothesis that T-ag is phosphorylated at Thr 124 within this cellular compartment. Thus, it can be argued that as infected cells approach S phase, a significant percentage of the T-ag molecules are phosphorylated on Thr 124 in the nucleus, utilizing as substrates either T-ag monomers (73) or possibly hexameric rings. One consequence of Thr 124 phosphorylation may be the formation of functional double hexamer forms on pentanucleotides 1 and 3 (35). Dephosphorylation of inhibitory serine residues, by protein phosphatase 2A, is also likely to play an important role in the regulation of double hexamer formation (23, 89). Formation of a functional double hexamer would enable subsequent events in the initiation process, such as protein remodeling and DNA unwinding, to ensue (4, 7, 26).

The architectural features of the SV40 core origin are similar to those in the replication origins of other DNA viruses (e.g., BK and JC viruses [16, 44], polyomavirus [3], and bovine papillomavirus [10]). Moreover, related viral initiator proteins, including polyomavirus large T-ag (91) and papillomavirus E1 (28, 46), are known to form hexamers and double hexamers on their origins. Furthermore, viral initiator proteins such as polyomavirus large T-ag (42) and papillomavirus E1 (12, 45, 50) have been reported to be substrates for cyclin-CDK complexes. Therefore, it will be interesting to determine if initiation events at other viral origins of replication are regulated at the level of the assembly of functional double hexamers. It will also be interesting to determine if the completion of assembly of other key initiator proteins on genomic origins of replication, such as the Mcm complex (99), are regulated in a similar manner.

ADDENDUM IN PROOF

Additional evidence that cyclin A and cdk2 are components of the SV40 replication initiation complex is provided by Cannella et al. (D. Cannella, J. M. Roberts, and R. Fotedar, Chromosoma 105:349–359, 1997).