Isolation and characterization of polyoma uncoating intermediates from the nuclei of infected mouse cells

A cell regulatory agent, CeReS-18, inhibits mouse 3T6 cell proliferation but not polyomavirus replication

We have purified a cell regulatory sialoglycopeptide, CeReS-18, from intact bovine cerebral cortex cells. This is an 18-kDa molecule that reversibly inhibits cellular DNA synthesis and the proliferation of a wide array of target cells. In the present study, the effect of CeReS-18 on mouse 3T6 host cell proliferation and polyomavirus replication was investigated. The results showed that CeReS-18 was able to inhibit 3T6 cell cycling in a concentration-dependent, calcium-sensitive, and reversible manner. Despite the inhibition of cell proliferation, CeReS-18 did not influence polyomavirus infection of 3T6 cells. Indirect immunofluorescent assays revealed that CeReS-18-treated, and cell cycle-arrested, 3T6 cells remained permissive to polyomavirus replication. Electron microscopy and immunogold labeling showed that new viral particles were assembled inside the nuclei of infected cells in the presence of CeReS-18 and during cell cycle arrest. The cellular requirements for the replication of polyomavirus DNA and the synthesis of viral proteins, as well as for the assembly of viral particles, therefore, remained available in CeReS-18-inhibited 3T6 cells. In addition, although polyomavirus infection can be mitogenic, infection of CeReS-18-treated 3T6 cells did not reverse the cell cycle arrest mediated by this cell cycle inhibitor.

Caveolae are involved in the trafficking of mouse polyomavirus virions and artificial VP1 pseudocapsids toward cell nuclei

Electron and confocal microscopy were used to observe the entry and the movement of polyomavirus virions and artificial virus-like particles (VP1 pseudocapsids) in mouse fibroblasts and epithelial cells. No visible differences in adsorption and internalization of virions and VP1 pseudocapsids ("empty" or containing DNA) were observed. Viral particles entered cells internalized in smooth monopinocytic vesicles, often in the proximity of larger, caveola-like invaginations. Both "empty" vesicles derived from caveolae and vesicles containing viral particles were stained with the anti-caveolin-1 antibody, and the two types of vesicles often fused in the cytoplasm. Colocalization of VP1 with caveolin-1 was observed during viral particle movement from the plasma membrane throughout the cytoplasm to the perinuclear area. Empty vesicles and vesicles with viral particles moved predominantly along microfilaments. Particle movement was accompanied by transient disorganization of actin stress fibers. Microfilaments decorated by the VP1 immunofluorescent signal could be seen as concentric curves, apparently along membrane structures that probably represent endoplasmic reticulum. Colocalization of VP1 with tubulin was mostly observed in areas close to the cell nuclei and on mitotic tubulin structures. By 3 h postinfection, a strong signal of the VP1 (but no viral particles) had accumulated in the proximity of nuclei, around the outer nuclear membrane. However, the vast majority of VP1 pseudocapsids did not enter the nuclei.

Use of electron microscopic and immunogold labeling techniques to determine polyomavirus recombinant VP1 capsid-like particles entry into mouse 3T6 cell nucleus

Murine polyomavirus major structural protein VP1 could assemble into capsid-like particles when expressed in the baculovirus system. The recombinant capsid-like particles that were purified by CsCl density gradient centrifugation were capable of packaging host DNA. Electron microscopic and immunogold labeling techniques were used to study the entry of these VP1 recombinant capsid-like particles into mouse 3T6 cells. It was found that these VP1 recombinant capsid-like particles, which lack polyomavirus minor structural proteins (VP2 and VP3), use the same mechanism to enter mouse 3T6 cell cytoplasm and nucleus as that used by native polyomavirus virions.

A mutation in the DE loop of the VP1 protein that prevents polyomavirus transcription and replication

Natural mutants of the DE loop of the Polyomavirus (Py) major coat protein VP1 have been previously shown to display an altered host specificity (L. Ricci, R. Maione, C. Passananti, A. Felsani, and P. Amati, 1992, J. Virol. 66, 7153-7158). To better understand the role of this outfacing loop of the VP1 protein in Py infectivity, we constructed and characterized a Py mutant (Py M17) harboring a deletion of 7 AA within the tip of the DE loop. The mutant virions obtained after DNA transfection were unable to replicate and initiate early transcription in fibroblast cells. Complementation experiments performed to rescue the deficient M17 replication by means of wt functions revealed the cis-dominance of the mutation. In situ cell fractionation experiments demonstrated that the Py mutant, like the Py wt, enters the cells, reaches the nucleus and that both the viral DNA and VP1 protein are found tightly bound to the nuclear matrix. These data suggest that the VP1 protein, associated to the viral DNA, conditions early viral gene expression and that the DE loop of the protein must be involved in this process.

Murine polyomavirus infection of 3T6 mouse cells shows evidence of predominant necrosis as well as limited apoptosis

The current study was developed to determine if polyomavirus infected 3T6 mouse cells evoked an apoptotic or a necrotic mechanism during infection. Infected cells were analyzed by flow cytometry, transmission electron microscopy (TEM), DNA electrophoresis and by measuring caspase-3 enzymatic activity. Infected cells that were analyzed at 72 h post-infection showed the following: flow cytometry analysis revealed a 5% increase in apoptotic cells and a 46% increase in necrotic cells when compared to uninfected cells; electron microscopy showed 10% cells with characteristic apoptotic morphology and 40% with necrotic appearance; caspase-3 activity was found to increase two fold when compared to uninfected cells and DNA fragmentation (laddering) was clearly evident late in infection. It was concluded that infected cells predominantly showed necrosis, although some cells showed apoptosis in late infection. Recombinant capsid-like particles composed of the polyomavirus structural proteins were not able to induce cell death.

The polyomavirus major capsid protein VP1 interacts with the nuclear matrix regulatory protein YY1

Polyomavirus reaches the nucleus in a still encapsidated form, and the viral genome is readily found in association with the nuclear matrix. This association is thought to be essential for viral replication. In order to identify the protein(s) involved in the virus-nuclear matrix interaction, we focused on the possible roles exerted by the multifunctional cellular nuclear matrix protein Yin Yang 1 (YY1) and by the viral major capsid protein VP1. In the present work we report on the in vivo association between YY1 and VP1. Using the yeast two-hybrid system we demonstrate that the VP1 and YY1 proteins physically interact through the D-E region of VP1 and the activation domain of YY1.

Mutations in the putative calcium-binding domain of polyomavirus VP1 affect capsid assembly

Calcium ions appear to play a major role in maintaining the structural integrity of the polyomavirus and are likely involved in the processes of viral uncoating and assembly. Previous studies demonstrated that a VP1 fragment extending from Pro-232 to Asp-364 has calcium-binding capabilities. This fragment contains an amino acid stretch from Asp-266 to Glu-277 which is quite similar in sequence to the amino acids that make up the calcium-binding EF hand structures found in many proteins. To assess the contribution of this domain to polyomavirus structural integrity, the effects of mutations in this region were examined by transfecting mutated viral DNA into susceptible cells. Immunofluorescence studies indicated that although viral protein synthesis occurred normally, infective viral progeny were not produced in cells transfected with polyomavirus genomes encoding either a VP1 molecule lacking amino acids Thr-262 through Gly-276 or a VP1 molecule containing a mutation of Asp-266 to Ala. VP1 molecules containing the deletion mutation were unable to bind 45Ca in an in vitro assay. Upon expression in Escherichia coli and purification by immunoaffinity chromatography, wild-type VP1 was isolated as pentameric, capsomere-like structures which could be induced to form capsid-like structures upon addition of CaCl2, consistent with previous studies. However, although VP1 containing the point mutation was isolated as pentamers which were indistinguishable from wild-type VP1 pentamers, addition of CaCl2 did not result in their assembly into capsid-like structures. Immunogold labeling and electron microscopy studies of transfected mammalian cells provided in vivo evidence that a mutation in this region affects the process of viral assembly.

Synthesis, posttranslational modifications, and nuclear transport of polyomavirus major capsid protein VP1

Polyomavirus major capsid protein VP1 synthesis was studied in infected primary baby mouse kidney cells. A standard curve of VP1 protein was used to quantitate VP1 in the cytoplasm and nucleus of infected cells during the time course of infection. Polyomavirus VP1 continued to be accumulated in the cytoplasm of the cells until 27 h postinfection, at which time the synthesis of VP1 leveled off. VP1 continued to accumulate in the nucleus of the infected cells throughout the course of infection. The presence of the six isospecies, A to F, of polyomavirus VP1 was also studied to determine the relative quantity of each species during the time course of infection. All six species were found in the cytoplasm and nucleus of infected cells at various times postinfection. However, the relative quantity of each species was different at early as compared with later times of infection. In addition, phosphorylated VP1 was found in isolated polyribosomes of infected cells, suggesting that phosphorylation of VP1 is a cotranslational modification. Examination of the effect of macromolecular synthesis on the transport of VP1 into the nucleus of infected baby mouse kidney cells as well as the rate of its nuclear accumulation during and after protein synthesis inhibition revealed that the continual transport and accumulation of VP1 in the nucleus required protein synthesis.

Early events in polyomavirus infection: fusion of monopinocytotic vesicles containing virions with mouse kidney cell nuclei

The entry of polyomavirus enclosed in monopinocytotic vesicles into mouse kidney cell nuclei was studied and evidence for a fusion mechanism was obtained. In vivo studies using the fluorescent lipophilic dye diI-C16(3) as a plasma membrane label showed that polyomavirus-infected nuclei accumulate plasma membrane, while uninfected or polyoma capsid-infected nuclei do not. Further evidence for fusion was obtained with electron microscopy of thin sections of infected mouse kidney cells. These specimens showed accumulation of plasma membrane in the outer nuclear membrane as well as evidence of recent fusion events. The polyoma virions (capsid proteins) were seen to accumulate on the inner nuclear membrane and in the nucleus and were identified by immunogold staining of the thin sections. The combined results of the in vivo dye studies and thin section immunoelectron microscopy studies provide evidence for a fusion mechanism for polyomavirus entry into mouse kidney cell nuclei.

Anti-idiotypic antibodies to a polyomavirus monoclonal antibody recognize cell surface components of mouse kidney cells and prevent polyomavirus infection

Anti-idiotypic antibodies have been successfully used to identify and isolate the receptor for several cell ligands. To prepare an immunologic probe for identification of the polyomavirus receptor on mouse kidney cells, polyclonal antisera against antipolyomavirus antibodies were prepared in rabbits. Fab fragments of the previously characterized monoclonal antibody E7, which neutralizes polyomavirus infection, were used for immunization (S. J. Marriott and R. A. Consigli, J. Virol. 56:365-372, 1985). Sera containing the greatest anti-idiotype activity were identified by enzyme-linked immunosorbent assay (ELISA) and purified by a series of affinity columns. The anti-idiotypic antibodies recognized the E7 idiotope in an ELISA, and anti-idiotype binding could be inhibited by preincubation of E7 monoclonal antibody with polyomavirus virions. When mixed with anti-idiotype immunoglobulin G (IgG), E7 was no longer capable of binding or immunoprecipitating polyomavirus virions or neutralizing polyomavirus infection. Direct immunofluorescence showed anti-idiotype IgG reactivity with a cell surface component of mouse kidney cells. Anti-idiotype F(ab')2 effectively competed with polyomavirus for binding to mouse kidney cells and displayed binding kinetics similar to those of polyomavirus. Virus infection of mouse kidney cells was blocked in a dose-dependent manner following treatment of the cells with anti-idiotype IgG. The anti-idiotype identified several proteins (95, 50, and 24 to 31 kilodaltons) in an immunoblot of mouse kidney cell membrane proteins but reacted predominantly with a single 50-kilodalton protein in a radioimmunoassay. The anti-idiotype failed to react with virus proteins in three assays, including ELISA, immunoprecipitation, and immunoblotting. The implications of this work for future identification and characterization of the polyomavirus receptor on mouse kidney cells are discussed.

Localization of calcium on the polyomavirus VP1 capsid protein

Our laboratory has previously shown that the divalent cation Ca2+ is an integral part of the polyomavirus and plays a major role in stabilizing the intact virion structure. In this report, we show that calcium is sequestered on the major capsid protein VP1 of polyomavirus. The virion structural proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis before being transferred to nitrocellulose and probed with 45CaCl2. Autoradiography revealed 45Ca binding exclusively to VP1. Increasing the amount of VP1 transferred to the nitrocellulose resulted in a concomitant increase in 45Ca binding. 45Ca binding to VP1 could be reduced by competition with an excess of unlabeled CaCl2. Separation of the species of VP1 by two-dimensional gel electrophoresis before electroblotting and probing with 45CaCl2 revealed that all six species (A to F) bind the radiolabeled calcium. Formic acid cleavage of the 43-kilodalton (kDa) VP1 protein into 29-, 18-, and 16-kDa fragments before 45Ca-binding analysis revealed that only the 18- and 16-kDa carboxyl-terminal fragments of this protein bind 45Ca.

Differences in biological activity and structural protein VP1 phosphorylation of polyomavirus progeny resulting from infection of primary mouse kidney and primary mouse embryo cell cultures

Both primary mouse kidney and primary mouse embryo cells in culture were used for polyomavirus progeny production. Examination of polyomavirus virion structural integrity revealed that mouse embryo cell progeny contained a threefold greater population of unstable particles when compared with mouse kidney cell progeny. Differences in biological activity between these two progeny virion types were also shown. Mouse kidney cell progeny compared with mouse embryo cell progeny exhibited a 10-fold greater ability to agglutinate guinea pig erythrocytes, a 3-fold lower ability to become internalized into monopinocytotic vesicles, and a 2-fold lower ability to initiate a productive infection based on positive nuclear immunofluorescence when mouse embryo host cell cultures were used. The mouse kidney progeny were also found to bind to host cells less specifically than the mouse embryo cell progeny. When these two progeny virion types were labeled in vivo with 32P and subjected to isoelectric focusing followed by sodium dodecyl sulfate-polyacrylamide gel electrophroesis in the second dimension, differences in the phosphorylation pattern of the major virus-encoded structural protein VP1 species were observed. It was revealed that species D and E of mouse kidney cell progeny were phosphorylated to the same degree, while mouse embryo cell progeny species E and F were phosphorylated equally. These data suggest that the host cells play a role in modulating the biological activity of the virus by affecting the degree and site-specific phosphorylation of the major capsid protein VP1 which may influence the recognition of virus attachment proteins for specific cellular receptors.

Octyl-beta-D-glucopyranoside extracts polyomavirus receptor moieties from the surfaces of mouse kidney cells

Polyomavirus receptor moieties were extracted from the surfaces of mouse kidney cells with the nonionic detergent octyl-beta-D-glucopyranoside. Following extraction with this detergent, mouse kidney cells were refractory to polyomavirus infection. Binding studies demonstrated that this loss of susceptibility resulted from extraction of a peripheral membrane protein or proteins required for proper virus attachment to and infection of mouse kidney cells. Infection of extracted mouse kidney cells returned following a 2-h recovery period. However, the presence of cycloheximide or tunicamycin in the recovery media interfered with recovery from infection. Cells could be infected immediately after extraction by supplying them with the extracted moieties prior to or concomitant with infection. A complex of polyomavirus and the extracted receptor protein was formed by in vitro incubation and was stable in sucrose gradient analysis. Functional receptor moieties were prepared in the form of liposomes from the detergent extract. The virus-receptor complex was immunoprecipitated with anti-polyomavirus immunoglobulin G, and the portion of the complex contributed by the cell was identified. Immunoblot analysis of the mouse kidney cell detergent extract with a receptor-specific 125I-labeled anti-idiotypic antibody or 125I-labeled polyomavirus demonstrated several reactive proteins. Attachment of polyomavirus to mouse kidney cells, followed by extraction of the virus-receptor complex, identified polyomavirus-binding proteins similar to those observed in in vitro binding. Proteins with molecular weights of approximately 95,000, 50,000 and 25,000 to 30,000 were consistently observed in all receptor assays. The relationship between these proteins and their possible involvement as the cell receptor for polyomavirus are discussed.

Production and characterization of monoclonal antibodies to polyomavirus major capsid protein VP1

Four hybridoma cell lines producing monoclonal antibodies against intact polyoma virions were produced and characterized. These antibodies were selected for their ability to react with polyoma virions in an enzyme-linked immunosorbent assay. The antibodies immunoprecipitated polyoma virions and specifically recognized the major capsid protein VP1 on an immunoblot. Distinct VP1 isoelectric species were immunoprecipitated from dissociated virion capsomere preparations. Two-dimensional gel electrophoresis demonstrated antibody reactivity with specific VP1 species. Monoclonal antibodies E7 and G9 recognized capsomeres containing VP1 species D, E, and F, while monoclonal antibodies C10 and D3 recognized capsomeres containing species B and C. Two of the monoclonal antibodies, E7 and G9, were capable of neutralizing viral infection and inhibiting hemagglutination. The biological activity of the monoclonal antibodies correlated well with the biological function of the species with which they reacted.

Isolation and characterization of monopinocytotic vesicles containing polyomavirus from the cytoplasm of infected mouse kidney cells

Monopinocytotic vesicles containing polyomavirus were isolated from the cytoplasm of mouse kidney cells infected with polyomavirus using sucrose density gradients. Nonenclosed, membrane-associated virions released by the action of neuraminidase separated from vesicle-enclosed virions in the sucrose gradient. Marker enzyme assays indicated the derivation of the vesicle membrane from the plasma membrane of the cell. The 125I-labeled virus enclosed in the vesicle sedimented more slowly in the gradient and was not observed unless infection and endocytosis had occurred. Detergent treatment of virion-containing vesicles caused the release of polyomavirus with sedimentation properties similar to those of purified polyoma virions. In addition, sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of virion proteins from vesicles containing virions demonstrated patterns of proteins similar to those of purified intact virions. Electron microscopy confirmed the presence of single intact virions inside vesicles. The study of these monopinocytotic virion-containing vesicles represents a further step in elucidating the early events of polyomavirus infection.

Isolation and characterization of polyoma nucleoprotein complexes

A method for the isolation of polyoma nucleoprotein complexes has been developed using neuraminidase treatment of infected cell lysates. At least three distinct forms of polyoma virion intermediates were identified by their [3H]thymidine labeling kinetics and sedimentation coefficients: a rapidly labeled 95 S "replicating complex" which chases to a 75 S minichromosome and then to a 240 S virion structure. The general properties of these distinct intermediates were similar to those found for SV40. In contrast to SV40, however, a continuum of labeled polyoma viral DNA sedimented between 240 S and 95 S. These complexes were characterized by their release from cell debris with neuraminidase, precipitation with antivirion antibody, complete disruption in 1 M NaCl, and association with hemagglutinating (HA) activity. These intermediates may represent incremental capsid protein additions to the 75 S minichromosome, hypothesized in the current models for SV40 assembly. The ability to isolate a complete complement of polyoma subviral complexes provides a basis for studying the growth defect of polyoma host-range mutants, and the properties of neuraminidase release, hemagglutination, and specific immunoprecipitation suggest purification steps for further characterization of these virion assembly intermediates.

Adenovirus DNA is associated with the nuclear matrix of infected cells

Viral DNA was found to be tightly associated with the nuclear matrix from HeLa cells lytically infected with human adenovirus type 5. The bound viral DNA, like cell DNA, was resistant to nonionic detergent and to extraction with high-salt (2 M NaCl) solution. However, whereas over 95% of the cell DNA was recovered in the matrix fraction, the amount of associated viral DNA varied during infection. Throughout the lytic cycle, the amount of matrix-associated adenovirus type 5 DNA increased until it reached a plateau level at 20 to 24 h after infection. At this stage, the matrix-bound DNA represented 87% of the total viral DNA; after this stage, additional newly synthesized viral DNA accumulated as non-matrix-associated DNA. DNase digestion studies revealed that all viral DNA sequences were equally represented in the matrix-bound DNA both early and late in infection; thus, unlike cell DNA, there seem to be no preferred attachment sites on the viral genome. An enrichment of viral DNA relative to cell DNA was found in the matrix-associated DNA after extensive DNase I digestion. This finding, together with an in situ hybridization study, suggests that the viral DNA is more intimately associated with the nuclear matrix than is cell DNA and probably does not exist in extended loops.

Simian virus 40 maturation: chromatin modifications increase the accessibility of viral DNA to nuclease and RNA polymerase

The accessibility of extracellular and nuclear simian virus 40 (SV40-M and SV40-I, respectively) virion chromatin DNAs to micrococcal nuclease, DNase I, BglI, EcoRI, and RNA polymerase was examined. Our results support the following conclusions: (i) the intranucleosomal DNA of SV40-I chromatin, similar to the precursor 75S chromatin complex, is resistant to enzymatic activity; and (ii) SV40-M virion chromatin is modified in a manner which increases the accessibility of viral DNA to enzymes, and the distinction between nucleosomal DNA and linker DNA is absent. Micrococcal nuclease digestion of SV40-I virion chromatin gave a typical nucleosomal DNA ladder pattern with a repeat unit of 205 base pairs of DNA. SV40-I chromatin was sensitive to cleavage with endonuclease BglI, but not with EcoRI. When SV40-I virion chromatin was used as a template, the rate of incorporation of ribonucleoside triphosphates into RNA was 5% of that obtained with naked form SV40 form I DNA. Micrococcal nuclease digestion of SV40-M virion chromatin resulted in submonomeric DNA fragments of approximately 55 base pairs, but no larger repeating unit of DNA was observed. SV40-M virion chromatin was sensitive to cleavage with either BglI or EcoRI and was approximately 20% more susceptible to digestion with DNase I than was SV40-I virion chromatin. The transcriptional efficiency of the extracellular virion chromatin was almost equivalent to that of naked SV40 form I DNA and was 16-fold higher than the rate observed with nuclear virion chromatin. The increased transcriptional activity was dependent upon the presence of nonhistone viral protein VP1 or VP2 or both.

Distribution of DNase I-sensitive sites in simian virus 40 nucleoprotein complexes from disrupted virus particles

Nucleoprotein complexes (core particles) released from simian virus 40 (SV40) virions were compared with similar complexes (SV40 chromatin) extracted from nuclei of infected cells. Core particles were sensitive to cleavage by DNase I at about the same enzyme concentration required to cleave SV40 chromatin. DNase I preferentially cleaved SV40 chromatin adjacent to the viral origin of replication; however, cleavage of core particles occurred with much less selectivity. The difference between these nucleoproteins was not due to a structural alteration induced by the virion disruption procedure, since SV40 chromatin retained its pattern of DNase I-sensitive sites when subjected top this treatment. On the other hand, core particles did not acquire the nuclease-sensitive feature typical of SV40 chromatin when they were exposed to infected cell nuclei and the Triton X-100-EDTA extraction procedure. Hence, the nuclease-sensitive feature was lost or altered during the normal process of virion assembly and maturation.

Differences in the subpopulations of the structural proteins of polyoma virions and capsids: biological functions of the multiple VP1 species

The structural proteins of polyoma virions and capsids were analyzed by isoelectric focusing and sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Polyoma virion VP1 was found to be composed of six distinct species which had pI's between pH 6.75 and 5.75. Polyoma capsid VP1 was found to contain four species with pI's between pH 6.60 and 5.75. The different forms of virion and capsid VP1 appeared to be generated by modifications (phosphorylation and acetylation) of the initial translation product. The most basic of the virion VP1 species (pI, pH 6.75) was absent in capsids and was found to be exclusively associated with the viral nucleoprotein complex. Three of the virion VP1 species and three of the capsid VP1 species were found in capsomere preparations enriched for hexon subunits. Two VP1 species were specifically immune precipitated from virions with hemagglutination-inhibiting antibodies. These two VP1 species were common to both virions and capsids. Polyoma virions, but not capsids, possessed a single VP1 species which was immune precipitated with neutralizing antibodies. Both virion and capsid VP2 were found to have pI's of approximately pH 5.50. Virion VP3 had a pI of approximately pH 7.00, whereas capsid VP3 had a pI of approximately pH 6.50.

Association of polyoma T antigen and DNA with the nuclear matrix from lytically infected 3T6 cells

Nuclear matrix prepared from mouse 3T6 cells lytically infected with polyoma virus retained significant amounts of the 100K T antigen and intact viral genomes. Bound T antigen was resistant to the extraction by high salt (2 M NaCl), detergent (1% Triton X-100) and exhaustive DNAase treatment. Only conditions sufficient to disrupt the integrity of the matrix itself solubilized the matrix T antigen. During the time period of 16-30 hr after infection, both the accumulation (in microgram) and the incorporation of 35S-methionine into T antigen increased steadily in cell extracts to a peak at 26 hr and then declined. In contrast, the amount of labeled T antigen retained by the matrix was relatively constant over the same time period. Matrix-bound T antigen was more highly phosphorylated and newly synthesized compared with the extractable T antigen. Viral DNA steadily accumulates in nuclei and on the matrix from 18 to 30 hr after infection. The fraction of viral DNA retained by the matrix was greatest early in infection (25% at 16 hr), declining to less than 10% by 24 hr. These data are consistent with the existence of a fixed (and limited) number of sites for T antigen (more highly phosphorylated) on the matrix and implicate the nuclear matrix as a site of viral DNA replication and possibly encapsidation.