Subclasses of simian virus 40 large T antigen: differential binding of two subclasses of T antigen from productively infected cells to viral and cellular DNA

Two major subclasses of simian virus 40 (SV40) large T antigen were separated by zone velocity sedimentation of crude extracts from productively infected cells. These subclasses, which have been shown to differ biologically and biochemically ( Fanning et al., 1981), sedimented at 5-6S and 14-16S. The amount of T antigen in each form was estimated by complement fixation and by immunoprecipitation of T antigen from extracts of cells chronically labeled with [35S]methionine. Each form of T antigen was tested for specific binding to end-labeled restriction fragments of SV40 DNA using an immunoprecipitation assay. The 5-6S and 14-16S forms of T antigen both bound specifically to DNA sequences in the SV40 HindIII C fragment. The sequences required for binding both forms were localized in the same 35-bp region of the origin. However, significant differences in binding activity and affinity for specific and nonspecific DNA were demonstrated. These properties suggest that T antigen subclasses may serve different functions in the lytically infected cell.     

Salt-stable binding of the large T protein to DNA in polyoma virus chromatin.

Nucleoprotein complexes extracted from the nuclei of mouse cells lytically infected with polyoma virus contain an ATPase activity which appears to correspond to that of the viral large T protein, as it exhibits the same characteristic properties; in particular, the activity is extensively inhibited by polyclonal antibodies from animals bearing polyoma tumors (anti-T antigen antibodies) and by monoclonal antibodies against large T. Significant amounts of DNA were immunoprecipitated by adding these antibodies to the nucleoprotein complex, suggesting that the protein is tightly bound to DNA in the viral chromatin. Since one of the monoclonal antibodies quantitatively immunoprecipitated the pulse-labeled replicative intermediates, we conclude that some large T protein remains physically associated with the DNA throughout its replication cycle. After exposure to salt concentrations higher than 1 M KCl, about half of the large T-specific ATPase activity was still observed to co-sediment with 21S form I viral DNA. The observations that the sedimentation coefficient of the salt-stable complexes was shifted to 16S after a limited endonucleolytic digestion, and that both the viral DNA and the ATPase activity were co-precipitated in the presence of polyethylene glycol at high ionic strength, further demonstrated that the protein is engaged in an unusually stable complex with DNA in the viral chromatin.     

The distribution of large T antigen in simian virus 40 nucleoprotein complexes

Simian virus 40 large T antigen binds to two types of nucleoprotein complexes from lytically infected cells: those containing replicating virus DNA (100S complexes) and those containing nonreplicating virus DNA (70S complexes). Analysis by agarose gel electrophoresis showed that replicating DNA was found exclusively in 100S complexes, although these complexes also contained large amounts of form I and form II DNA. In contrast, no replicating DNA was found in 70S complexes, and pulse-labeled DNA in these complexes migrated as form I and form II DNA that presumably had recently completed replication. Immunoprecipitation and gel electrophoresis showed that large T antigen was associated with both types of complexes. From 21 to 62% of replicating DNA in 100S complexes was bound to T antigen. Our estimates indicated, however, that more than three-fourths of the DNA molecules in 100S complexes were nonreplicating and unassociated with T antigen. In 70S complexes, 12 to 31% of pulse-labeled DNA was bound to T antigen, but because there were more DNA molecules in the 70S complexes, they contained a greater absolute amount of T antigen.     

A small subclass of SV40 T antigen binds to the viral origin of replication

We examined the affinities of SV40 large T antigen for unique viral DNA sequences by binding SV40 Bst NI DNA fragments in extracts of infected or transformed cells, and then immunoprecipitating the T antigen-DNA complex. The G fragment, which spans the viral origin of replication (ori) was quantitatively bound to T antigen. A T-antigen-specific monoclonal antibody (McI 7), which recognized only 5%-10% of the T antigen from infected or transformed cells, immunoprecipitated the majority of the ori-binding activity. This suggests that only a minor subclass of wild-type T antigen is active in binding to the origin. C6 cells contain a replication-defective mutant T antigen that when tested in the DNA-binding immunoassay, showed no affinity for the ori fragment. McI 7 not only failed to immunoprecipitate ori binding in C6 cells, but also did not detect any labeled C6 T antigen whatever. Thus McI 7 recognizes an immunologically distinct subset of wild-type 7 antigen that comprises the origin-binding form of the viral protein, which is absent in the C6 T antigen population. McI 122, which recognizes a 53 kilodalton host protein that complexes with T antigen, immunoprecipitated ori-binding activity from extracts of infected or transformed cells, but not from C6 cells. Thus wild-type T antigen can bind ori sequences even when complexed to the host protein. These data suggest that T antigen consists of different subpopulations with different functions.     

Subclasses of simian-virus-40 large tumor antigen. Partial purification and DNA-binding properties of two subclasses of tumor antigen from productively infected cells

Two major subclasses of simian virus 40 tumor antigen were prepared from productively infected monkey cells. These subclasses can be distinguished by their sedimentation properties: one tumor antigen form sediments at 5-6S and the other at 14-16S. The DNA-binding properties of these subclasses were investigated by two different experimental procedures. In the first procedure, the DNA binding of subclasses of crude tumor antigen, separated by zone velocity sedimentation, were assayed by immunoprecipitation of the DNA-protein complexes. In the second procedure, the two tumor antigen forms were partially purified by column chromatography and DNA binding was tested in a filter binding assay. Both procedures gave comparable results. (a) The 5-6-S and the 14-16-S tumor antigen bound specifically to a DNA restriction fragment containing the viral genome control regions. (b) At low salt concentrations, both subclasses bound to specific and to nonspecific DNA sequences; competition experiments in the presence of nonspecific DNA showed, however, that the affinity of both tumor antigen forms for the viral genome control region was at least 10-fold higher than their affinity for nonspecific DNA sequences. (c) The binding of the 5-6-S subclass to viral control region DNA was optimal at 60-80 mM NaCl while specific DNA binding of the 14-16-S form was optimal at 150-200 mM NaCl; however, binding of the 14-16-S form to nonspecific DNA sequences was also more resistant to high salt concentrations than that of the 5-6S form. (d) Both tumor antigen forms bound well to specific and to nonspecific DNA at pH 6-6.5; with increasing pH values, binding to nonspecific DNA decreased while binding to specific DNA reached an optimum at pH 7-7.5. Binding of the 14-16-S form to viral origin DNA was more resistant to pH values above 7.5 than binding of the 5-6-S form.     

Self-assembly of simian virus 40 large T antigen oligomers by divalent cations.

In simian virus 40-transformed cells, simian virus 40 large T antigen can be detected in different forms separable by sucrose density gradient centrifugation. In our experiments, light forms sedimented around 5 to 7S, oligomers such as tetramers were detected around 16S, and higher aggregates sedimented in a broad distribution reaching above 23S. The oligomers sedimenting at and above 16S could be disassembled into the slowly sedimenting 5 to 7S forms by chelating agents [EDTA or ethylene bis(oxonitrilo)tetraacetate]. After the addition of divalent cations (CaCl2 or MgCl2) in excess of chelating agents, oligomeric forms reassembled and appeared in a sedimentation pattern resembling that observed before treatment with chelating agents. Time course studies permitted the identification of the 5 to 7S forms as precursors upon pulse-labeling (15 min); the 16S and higher oligomers were identified as the successors after a 14-h chase. Treatment of extracts of pulse-chase-labeled cells with chelating agents again disassembled the oligomers, whereas pulse-labeled precursors did not change their 5 to 7S sedimentation pattern. Adding an excess of divalent cations reassembled the pulse-chase-labeled T antigen to oligomers but did not influence the sedimentation behavior of pulse-labeled 5 to 7S precursors. It is therefore reasonable to assume that a posttranslational modulation induces divalent cation binding, leading finally to the oligomerization of T antigen. Thus, some of the multifunctional activities of T antigen can be dictated by divalent cation binding properties.     

Association of simian virus 40 T antigen with the nuclear matrix of infected and transformed monkey cells.

The subnuclear distribution of simian virus 40 large T antigen within nuclei of transformed Cos and C6 monkey cells was examined. Cos cells express wild-type T antigen but lack viral sequences required for DNA replication, whereas C6 cells contain a functional viral origin but express a replication-defective mutant T antigen which is unable to bind specifically to viral DNA. Discrete subpopulations of T antigen were isolated from the soluble nucleoplasm, chromatin, and nuclear matrix of both cell lines. Although only a small quantity (2 to 12%) of the total nuclear T antigen from Cos cells was associated with the nuclear matrix, a high proportion (25 to 50%) of C6 T antigen was bound to this structure. Results obtained from lytically infected monkey cells showed that early in infection, before viral replication was initiated, a higher proportion (22%) of T antigen was found associated with the nuclear matrix compared with amounts found associated with this structure later in infection (5 to 8%). These results suggest that an increased association of T antigen with this structure is not correlated with viral replication. T antigen isolated from the C6 nuclear matrix was more highly phosphorylated than was soluble C6 T antigen and was capable of binding to the host p53 protein. C6 DNA contains three mutations: two corresponding to N-terminal changes at amino acid positions 30 and 51 and a third located internally at amino acid position 153. By analysis of the subnuclear distribution of T antigen from rat cells transformed by C6 submutant T antigens, it was determined that one or both of the mutations at the NH2 terminus are responsible for the increased quantity of C6 T antigen associated with the nuclear matrix. These results suggest that neither a functional viral DNA replication origin nor the origin binding property of T antigen is required for association of this protein with the nuclear matrix.     

Specific association of simian virus 40 tumor antigen with simian virus 40 chromatin

Simian virus 40 tumor antigen (SV40 T antigen) was bound to both replicating and fully replicated SV40 chromatin extracted with a low-salt buffer from the nuclei of infected cells, and at least a part of the association was tight specific. T antigen cosedimented on sucrose gradients with SV40 chromatin, and T antigen-chromatin complexes could be precipitated from the nuclear extract specifically with anti-T serum. From 10 to 20% of viral DNA labeled to steady state with [3H]thymidine for 12 h late in infection or 40 to 50% of replicating viral DNA pulse-labeled for 5 min was associated with T antigen in such immunoprecipitates. After reaction with antibody, most of the T antigen-chromatin complex was stable to washing with 0.5 M NaCl, but only about 20% of the DNA label remained in the precipitate after washing with 0.5 M NaCl-0.4% Sarkosyl. This tightly bound class of T antigen was associated preferentially with a subfraction of pulse-labeled replicating DNA which comigrated with an SV40 form I marker. A tight binding site for T antigen was identified tentatively by removing the histones with dextran sulfate and heparin from immunoprecipitated chromatin labeled with [32P]phosphate to steady state and then digesting the DNA with restriction endonucleases HinfI and HpaII. The site was within the fragment spanning the origin of replication, 0.641 to 0.725 on the SV40 map.     

Ataxia telangiectasia-mutated damage-signaling kinase- and proteasome-dependent destruction of Mre11-Rad50-Nbs1 subunits in Simian virus 40-infected primate cells

Although the mechanism of simian virus 40 (SV40) DNA replication has been extensively investigated with cell extracts, viral DNA replication in productively infected cells utilizes additional viral and host functions whose interplay remains poorly understood. We show here that in SV40-infected primate cells, the activated ataxia telangiectasia-mutated (ATM) damage-signaling kinase, gamma-H2AX, and Mre11-Rad50-Nbs1 (MRN) assemble with T antigen and other viral DNA replication proteins in large nuclear foci. During infection, steady-state levels of MRN subunits decline, although the corresponding mRNA levels remain unchanged. A proteasome inhibitor stabilizes the MRN complex, suggesting that MRN may undergo proteasome-dependent degradation. Analysis of mutant T antigens with disrupted binding to the ubiquitin ligase CUL7 revealed that MRN subunits are stable in cells infected with mutant virus or transfected with mutant viral DNA, implicating CUL7 association with T antigen in MRN proteolysis. The mutant genomes produce fewer virus progeny than the wild type, suggesting that T antigen-CUL7-directed proteolysis facilitates virus propagation. Use of a specific ATM kinase inhibitor showed that ATM kinase signaling is a prerequisite for proteasome-dependent degradation of MRN subunits as well as for the localization of T antigen and damage-signaling proteins to viral replication foci and optimal viral DNA replication. Taken together, the results indicate that SV40 infection manipulates host DNA damage-signaling to reprogram the cell for viral replication, perhaps through mechanisms related to host recovery from DNA damage.     

Interaction of DNA polymerase alpha-primase with cellular replication protein A and SV40 T antigen.

The purified human single-stranded DNA binding protein, replication protein A (RP-A), forms specific complexes with purified SV40 large T antigen and with purified DNA polymerase alpha-primase, as shown by ELISA and a modified immunoblotting technique. RP-A associated efficiently with the isolated primase, as well as with intact polymerase alpha-primase. The 70 kDa subunit of RP-A was sufficient for association with polymerase alpha-primase. Purified SV40 large T antigen bound to intact RP-A and to polymerase-primase, but not to any of the separated subunits of RP-A or to the isolated primase. These results suggest that the specific protein-protein interactions between RP-A, polymerase-primase and T antigen may play a role in the initiating of SV40 DNA replication.     

Chemical crosslinking of different forms of the simian virus 40 large T antigen using bifunctional reagents

Chemical crosslinking was used for a direct analysis of the different forms of large tumor (T) antigen, the simian virus 40 A gene product. The first subclass, sedimenting at 14-16S, is composed of monomeric to tetrameric units, whereas the second, sedimenting at 5-6S, only contains dimers and monomers of T. The occurrence of oligomeric structures of T in solution which are higher than dimers suggests the possibility of direct binding of such trimers or tetramers to the origin of replication of the viral DNA as an alternative to the formation of these structures by aggregation of bound dimers or monomers after their sliding along the DNA.     

Binding of the rat liver 7-8 S dexamethasone receptor to deoxyribonucleic acid

The 7-8 S form of the [3H]dexamethasone (9 alpha-fluoro-11 beta,17,21-trihydroxy-16 alpha-methylpregna-1,4-diene-3, 20-dione) receptor from rat liver cytosol can be converted to the 3-4 S form by RNase treatment or high salt, suggesting a salt-sensitive association between the receptor protein and RNA. In DNA-cellulose column assays, the gradient-purified 3-4 S form bound DNA more efficiently than the 7-8 S form, though the 7-8 S form was also capable of binding to DNA-cellulose to a significant extent. Activated 7-8 S dexamethasone receptor could be released from its association with soluble DNA by treatment with DNase I. Sucrose gradient analysis showed that the released receptor sedimented as the 7-8 S form and was sensitive to RNase treatment, which induced a conversion to the 3-4 S form. Activated RNase-generated 3-4 S receptor again displayed a higher degree of binding to soluble DNA and was recovered in the 3-4 S form following DNase extraction. The fact that the 3-4 S form bound immobilized or soluble DNA more efficiently suggests that the associated RNA of the 7-8 S form interferes directly or indirectly with the receptor association with DNA. The observation that the receptor binds to DNA in its 7-8 S form suggests that the receptor complex is capable of binding RNA and DNA concurrently.     

DNA-binding activity of simian virus 40 large T antigen correlates with a distinct phosphorylation state

The state of phosphorylation and the relationship of various subclasses of simian virus 40 large T antigen (large T) differing in DNA-binding activity, degree of oligomerization, age, and subcellular distribution were investigated. Young large T (continuously labeled for 4 h late in infection) comprised about 20% of the total cellular large T. It was phosphorylated to a low degree and existed primarily in a monomeric form, sedimenting at 5S. More than 50% of this fraction bound to simian virus 40 DNA, preferentially to origin-containing sequences. Old large T (continuously labeled for 17 h, followed by a 4-h chase) represented the majority of the population. It was highly phosphorylated and predominantly in an oligomeric form, sedimenting at 15S to 23S. Only 10 to 20% of this fraction bound to simian virus 40 DNA. Another subclass of large T which was extracted from nuclei with 0.5 M salt resembled newly synthesized molecules in all properties tested; it was phosphorylated to a low degree, sedimented at 5S, and bound to viral DNA with high efficiency (greater than 70%). Two-dimensional phosphopeptide analysis of the individual subclasses revealed two distinct phosphorylation patterns, one characteristic for young, monomeric, and DNA-binding large T, the other for old, oligomeric, and non-DNA-binding large T. All sites previously identified in unfractionated large T (K.H. Scheidtmann et al., J. Virol. 44:116-133, 1982) were also phosphorylated in the various subclasses, but to different degrees. Peptide maps of the DNA-binding fraction, the 5S form, and the nuclear high-salt fraction showed two prominent phosphopeptides not previously characterized. Both peptides were derived from the amino-terminal region of large T, presumably involved in origin binding, and probably represent partially phosphorylated intermediates of known phosphopeptides. Our data show that the DNA-binding activity, age, and oligomerization of large T correlate with distinct states of phosphorylation. We propose that differential phosphorylation might play a role in the interaction of large T with DNA.     

Polyomavirus Large T Antigen Binds Cooperatively to Its Multiple Binding Sites in the Viral Origin of DNA Replication

Polyomavirus large T antigen binds to multiple 5′-G(A/G)GGC-3′ pentanucleotide sequences in sites 1/2, A, B, and C within and adjacent to the origin of viral DNA replication on the polyomavirus genome. We asked whether the binding of large T antigen to one of these sites could influence binding to other sites. We discovered that binding to origin DNA is substantially stronger at pH 6 to 7 than at pH 7.4 to 7.8, a range often used in DNA binding assays. Large T antigen-DNA complexes formed at pH 6 to 7 were stable, but a fraction of these complexes dissociated at pH 7.6 and above upon dilution or during electrophoresis. Increased binding at low pH is therefore due at least in part to increased stability of protein-DNA complexes, and binding at higher pH values is reversible. Binding to fragments of origin DNA in which one or more sites were deleted or inactivated by point mutations was measured by nitrocellulose filter binding and DNase I footprinting. The results showed that large T antigen binds cooperatively to its four binding sites in viral DNA, suggesting that the binding of this protein to one of these sites stabilizes its binding to other sites via protein-protein contacts. Sites A, B, and C may therefore augment DNA replication by facilitating the binding of large T antigen to site 1/2 at the replication origin. ATP stabilized large T antigen-DNA complexes against dissociation in the presence, but not the absence, of site 1/2, and ATP specifically enhanced protection against DNase I digestion in the central 10 to 12 bp of site 1/2, at which hexamers are believed to form and begin unwinding DNA. We propose that large T antigen molecules bound to these multiple sites on origin DNA interact with each other to form a compact protein-DNA complex and, furthermore, that ATP stimulates their assembly into hexamers at site 1/2 by a “handover” mechanism mediated by these protein-protein contacts.     

Regulation of simian virus 40 gene expression in Xenopus laevis oocytes.

Expression of the simian virus 40 (SV40) early and late regions was examined in Xenopus laevis oocytes microinjected with viral DNA. In contrast to the situation in monkey cells, both late-strand-specific (L-strand) RNA and early-strand-specific (E-strand) RNA could be detected as early as 2 h after injection. At all time points tested thereafter, L-strand RNA was synthesized in excess over E-strand RNA. Significantly greater quantities of L-strand, relative to E-strand, RNA were detected over a 100-fold range of DNA concentrations injected. Analysis of the subcellular distribution of [35S]methionine-labeled viral proteins revealed that while the majority of the VP-1 and all detectable small t antigen were found in the oocyte cytoplasm, most of the large T antigen was located in the oocyte nucleus. The presence of the large T antigen in the nucleus led us to investigate whether this viral product influences the relative synthesis of late or early RNA in the oocyte as it does in infected monkey cells. Microinjection of either mutant C6 SV40 DNA, which encodes a large T antigen unable to bind specifically to viral regulatory sequences, or deleted viral DNA lacking part of the large T antigen coding sequences yielded ratios of L-strand to E-strand RNA that were similar to those observed with wild-type SV40 DNA. Taken together, these observations suggest that the regulation of SV40 RNA synthesis in X. laevis oocytes occurs by a fundamentally different mechanism than that observed in infected monkey cells. This notion was further supported by the observation that the major 5' ends of L-strand RNA synthesized in oocytes were different from those detected in infected cells. Furthermore, only a subset of those L-strand RNAs were polyadenylated.     

Mutations in the phosphorylation sites of simian virus 40 (SV40) T antigen alter its origin DNA-binding specificity for sites I or II and affect SV40 DNA replication activity

A series of mutants of simian virus 40 was constructed by oligonucleotide-directed mutagenesis to study the role of phosphorylation in the functions of large T antigen. Each of the previously mapped phosphorylated serine and threonine residues in large T antigen was replaced by an alanine or cysteine residue or, in one case, by glutamic acid. Mutant DNAs were assayed for plaque-forming activity, viral DNA replication, expression of T antigen, and morphological transformation of rat cells. Viable mutants were isolated, suggesting that modification of some residues is not essential for the biological functions of T antigen. Two of these mutants replicated more efficiently than did the wild type. Seven mutants were partially or completely deficient in viral DNA replication but retained cell transformation activity comparable with that of the wild-type protein. Biochemical analysis of the mutant T antigens demonstrated novel origin DNA-binding properties of several mutant proteins. The results are consistent with the idea that differential phosphorylation defines several functional subclasses of T-antigen molecules.     

Simian virus 40 and polyomavirus large tumor antigens have different requirements for high-affinity sequence-specific DNA binding.

By using a DNA fragment immunoassay, the binding of simian virus 40 (SV40) and polyomavirus (Py) large tumor (T) antigens to regulatory regions at both viral origins of replication was examined. Although both Py T antigen and SV40 T antigen bind to multiple discrete regions on their proper origins and the reciprocal origin, several striking differences were observed. Py T antigen bound efficiently to three regions on Py DNA centered around an MboII site at nucleotide 45 (region A), a BglI site at nucleotide 92 (region B), and another MboII site at nucleotide 132 (region C). Region A is adjacent to the viral replication origin, and region C coincides with the major early mRNA cap site. Weak binding by Py T antigen to the origin palindrome centered at nucleotide 3 also was observed. SV40 T antigen binds strongly to Py regions A and B but only weakly to region C. This weak binding on region C was surprising because this region contains four tandem repeats of GPuGGC, the canonical pentanucleotide sequence thought to be involved in specific binding by T antigens. On SV40 DNA, SV40 T antigen displayed its characteristic hierarchy of affinities, binding most efficiently to site 1 and less efficiently to site 2. Binding to site 3 was undetectable under these conditions. In contrast, Py T antigen, despite an overall relative reduction of affinity for SV40 DNA, binds equally to fragments containing each of the three SV40 binding sites. Py T antigen, but not SV40 T antigen, also bound specifically to a region of human Alu DNA which bears a remarkable homology to SV40 site 1. However, both tumor antigens fail to precipitate DNA from the same region which has two direct repeats of GAGGC. These results indicate that despite similarities in protein structure and DNA sequence, requirements of the two T antigens for pentanucleotide configuration and neighboring sequence environment are different.     

SV40 T antigen binds directly to the large subunit of purified DNA polymerase alpha.

Purified SV40 large T antigen and purified DNA polymerase alpha-primase form a complex detectable by ELISA and by a modified immunoblotting technique. The interaction is specific for the large catalytic subunit of polymerase alpha. The amino terminal 83 amino acids of T antigen are both necessary and sufficient for binding to the polymerase. However, antibody epitopes located in the carboxy terminal ATPase domain of T antigen are masked in the polymerase-T antigen complex, and complex formation is inhibited by an antibody directed against the carboxy terminus of T antigen, suggesting that this region of T antigen, though not required for binding, is in close proximity to the bound polymerase. The affinity of human DNA polymerase alpha for T antigen is approximately 10-fold greater than that of polymerase alpha from calf thymus, consistent with the interpretation that polymerase alpha is at least in part responsible for the primate-specific replication of SV40 DNA in vivo and in vitro. The results suggest that specific protein-protein interaction between DNA polymerase alpha and T antigen plays an important role in viral DNA replication.