Simian virus 40 agnoprotein facilitates normal nuclear location of the major capsid polypeptide and cell-to-cell spread of virus.

The simian virus 40 agnoprotein is a 61-amino-acid, highly basic polypeptide that is coded within the 5' leader of late 16S mRNAs. To better understand agnoprotein function and to more effectively differentiate cis-from trans-acting effects of an agnogene mutation, we constructed a mutant virus that carries a single-base-pair substitution and fails to produce agnoprotein. pm 1493 contains a T/A to A/T transversion at sequence position 335. This mutation converts the agnoprotein initiation codon from ATG to TTG, preventing synthesis of the protein. The mutant displays only a modest growth defect in CV-1P and AGMK cells and no defect in BSC-1 cells. Early-gene expression, DNA replication, synthesis of late viral products, and the kinetics of virion assembly all appear normal in pm 1493-infected CV-1P cells. Immunofluorescent studies, however, indicate that localization of the major capsid polypeptide VP1 is different in mutant- than wild-type virus-infected cells. Furthermore, the lack of agnoprotein led to inefficient release of mature virus from the infected cell. Agnogene mutants could be severely compromised in their ability to propagate in monkeys given their reduced capacity for cell-to-cell spread.     

Simian virus 40 host range/helper function mutations cause multiple defects in viral late gene expression.

Simian virus 40 (SV40) deletion mutants dlA2459 and dlA2475 express T antigens that lack the normal carboxy terminus. These mutants are called host range/helper function (hr/hf) mutants because they form plaques at 37 degrees C on BSC-1 and Vero monkey kidney cell lines but not on CV-1p monkey kidney cells. Wild-type SV40 can provide a helper function to permit growth of human adenoviruses in monkey kidney cells; the hr/hf mutants cannot. Progeny yields of hr/hf mutants are also cold sensitive in all cell lines tested. Patterns of viral macromolecular synthesis in three cell lines (Vero, BSC-1, and CV-1) at three temperatures (40, 37, and 32 degrees C) were examined to determine the nature of the growth defect of hr/hf mutants. Mutant viral DNA replication was similar to that of the wild type in all three cell lines, indicating that the mutations affect late events in the viral lytic cycle. In mutant-infected Vero cells, in which viral yields were highest, late mRNA levels were similar to those observed during wild-type infection. Levels of viral late mRNA from mutant-infected CV-1 and BSC-1 cells at 32 and 37 degrees C were reduced relative to those of wild-type-infected cells. The steady-state level of the major viral capsid protein, VP1, in mutant-infected CV-1 cells was reduced to the same extent as was late mRNA. The synthesis of agnoprotein could not be detected in mutant-infected CV-1 cells but was readily detected in CV-1 cells infected by wild-type SV40. Primer extension analyses indicated that most late mRNAs from mutant-infected CV-1 cells utilize start sites downstream from the major wild-type cap site (nucleotide 325) and the agnoprotein initiation codon (nucleotide 335). These results indicate that deletion of the carboxyl-terminal domain of T antigen affects viral late mRNA production, both quantitatively and qualitatively. The agnoprotein is detected late in the wild-type SV40 lytic cycle and is thought to play a role in the assembly or maturation of virions. Reduced hr/hf progeny yields could result from decreased capsid protein synthesis and, in the absence of detectable levels of agnoprotein, from inefficient use of available capsid proteins.     

Two separable functional domains of simian virus 40 large T antigen: carboxyl-terminal region of simian virus 40 large T antigen is required for efficient capsid protein synthesis.

The carboxyl-terminal portion of simian virus 40 large T antigen is essential for productive infection of CV-1 and CV-1p green monkey kidney cells. Mutant dlA2459, lacking 14 base pairs at 0.193 map units, was positive for viral DNA replication, but unable to form plaques in CV-1p cells (J. Tornow and C.N. Cole, J. Virol. 47:487-494, 1983). In this report, the defect of dlA2459 is further defined. Simian virus 40 late mRNAs were transcribed, polyadenylated, spliced, and transported in dlA2459-infected cells, but the level of capsid proteins produced in infected CV-1 green monkey kidney cells was extremely low. dlA2459 large T antigen lacks those residues known to be required for adenovirus helper function, and the block to productive infection by dlA2459 occurs at the same stage of infection as the block to productive adenovirus infection of CV-1 cells. These results suggest that the adenovirus helper function is required for productive infection by simian virus 40. Mutant dlA2459 was able to grow on the Vero and BSC-1 lines of African green monkey kidney cells. Additional mutants affecting the carboxyl-terminal portion of large T were prepared. Mutant inv2408 contains an inversion of the DNA between the BamHI and BclI sites (0.144 to 0.189 map units). This inversion causes transposition of the carboxyl-terminal 26 amino acids of large T antigen and the carboxyl-terminal 18 amino acids of VP1. This mutant was viable, even though the essential information absent from dlA2459 large T antigen has been transferred to the carboxyl terminus of VP1 of inv2408. The VP1 polypeptide carrying this carboxyl-terminal portion of large T could overcome the defect of dlA2459. This indicates that the carboxyl terminus of large T antigen is a separate and separable functional domain.     

Properties of the simian virus 40 (SV40) large T antigens encoded by SV40 mutants with deletions in gene A

The biochemical properties of the large T antigens encoded by simian virus 40 (SV40) mutants with deletions at DdeI sites in the SV40 A gene were determined. Mutant large T antigens containing only the first 138 to 140 amino acids were unable to bind to the SV40 origin of DNA replication as were large T antigens containing at their COOH termini 96 or 97 amino acids encoded by the long open reading frame located between 0.22 and 0.165 map units (m.u.). All other mutant large T antigens were able to bind to the SV40 origin of replication. Mutants with in-phase deletions at 0.288 and 0.243 m.u. lacked ATPase activity, but ATPase activity was normal in mutants lacking origin-binding activity. The 627-amino acid large T antigen encoded by dlA2465, with a deletion at 0.219 m.u., was the smallest large T antigen displaying ATPase activity. Mutant large T antigens with the alternate 96- or 97-amino acid COOH terminus also lacked ATPase activity. All mutant large T antigens were found in the nuclei of infected cells; a small amount of large T with the alternate COOH terminus was also located in the cytoplasm. Mutant dlA2465 belonged to the same class of mutants as dlA2459. It was unable to form plaques on CV-1p cells at 37 or 32 degrees C but could form plaques on BSC-1 monolayers at 37 degrees C but not at 32 degrees C. It was positive for viral DNA replication and showed intracistronic complementation with any group A mutant whose large T antigen contained a normal carboxyl terminus. These findings and those of others suggest that both DNA binding and ATPase activity are required for the viral DNA replication function of large T antigen, that these two activities must be located on the same T antigen monomer, and that these two activities are performed by distinct domains of the polypeptide. These domains are distinct and separable from the domain affected by the mutation of dlA2465 and indicate that SV40 large T antigen is made up of at least three separate functional domains.     

Mechanisms of interference with simian virus 40 (SV40) DNA replication by trans-dominant mutants of SV40 large T antigen.

Mutations at multiple sites within the simian virus 40 (SV40) early region yield large T antigens which interfere trans dominantly with the replicative activities of wild-type T antigen. A series of experiments were conducted to study possible mechanisms of interference with SV40 DNA replication caused by these mutant T antigens. First, the levels of wild-type T antigen expression in cells cotransfected with wild-type and mutant SV40 DNAs were examined; approximately equal levels of wild-type T antigen were seen, regardless of whether the cotransfected mutant was trans dominant or not. Second, double mutants that contained the mutation of inA2827, a strong trans-dominant mutation with a 12-bp linker inserted at the position encoding amino acid 520, and various mutations in other parts of the large-T-antigen coding region were constructed. The trans-dominant interference of inA2827 was not affected by second mutations within the p105Rb binding site or the amino or carboxy terminus of large T antigen. Mutation of the nuclear localization signal partially reduced the trans dominance of inA2827. The large T antigen of mutant inA2815 contains an insertion of 4 amino acids at position 168 of large T; this T antigen fails to bind SV40 DNA but is not trans dominant for DNA replication. The double mutant containing the mutations of both inA2815 and in A2827 was not trans dominant. The large T antigen of dlA2433 lacks amino acids 587 to 589, was unstable, and failed to bind p53. Combining the dlA2433 mutation with the inA2827 mutation also reversed the trans dominance completely, but the effect of the dlA2433 mutation on trans dominance can be explained by the instability of this double mutant protein. In addition, we examined several mutants with conservative point mutations in the DNA binding domain and found that most of them were not trans dominant. The implications of the results of these experiments on possible mechanisms of trans dominance are discussed.     

Linker insertion mutants of simian virus 40 large T antigen that show trans-dominant interference with wild-type large T antigen map to multiple sites within the T-antigen gene.

Linker insertion mutants affecting the simian virus 40 (SV40) large tumor (T) antigen were constructed by inserting a 12-base-pair oligonucleotide linker into restriction endonuclease cleavage sites located within the early region of SV40. One mutant, with the insertion at amino acid 5, was viable in CV-1p and BSC-1 cells, indicating that sequences very close to the amino terminus of large T could be altered without affecting the lytic infection cycle of SV40. All other mutants affecting large T were not viable. In complementation assays between the linker insertion mutants and either a late-gene mutant, dlBC865, or a host range/helper function (hr/hf) mutant, dlA2475, delayed complementation was seen with the 6 of the 10 nonviable mutants. Of these 10 mutants, 5 formed plaques 3 to 4 days later than in control complementations, while complementation by one of the mutants, inA2827, with an insertion at amino acid 520, was delayed more than 1 week. Most mutants which showed delayed complementation replicated less well in Cos-1 cells than did a control mutant, dlA1209, which produced no T antigen. The replication of inA2827(aa520) was reduced by more than 90%. Similar interference with viral DNA replication was seen when CV-1, HeLa, or 293 cells were cotransfected with an origin-defective plasmid encoding wild-type large T antigen and with inA2827(aa520). Only one of the mutant T antigens, inA2807(aa303), was unstable. These results indicate that some of the mutant T antigens interfered with functions of wild-type T required for viral DNA replication. However, not all of the mutants which showed delayed complementation also showed interference with viral DNA replication. This indicates that mutant large T antigens may interfere trans dominantly with multiple activities of wild-type large T antigen.     

SV40-transformed simian cells support the replication of early SV40 mutants

CV-1, an established line of simian cells permissive for lytic growth of SV40, were transformed by an origin-defective mutant of SV40 which codes for wild-type T antigen. Three transformed lines (COS-1, -3, -7) were established and found to contain T antigen; retain complete permissiveness for lytic growth of SV40; support the replication of tsA209 virus at 40 degrees C; and support the replication of pure populations of SV40 mutants with deletions in the early region. One of the lines (COS-1) contains a single integrated copy of the complete early region of SV40 DNA. These cells are possible hosts for the propagation of pure populations of recombinant SV40 viruses.     

Simian virus 40 agnoprotein facilitates perinuclear-nuclear localization of VP1, the major capsid protein.

The agnoprotein of simian virus 40 (SV40) is a 61-amino-acid protein encoded in the leader of some late mRNAs. In indirect immunofluorescence studies with antisera against SV40 capsid proteins, we show that mutants which make no agnoprotein display abnormal perinuclear-nuclear localization of VP1, the major capsid protein, but not VP2 or VP3, the minor capsid proteins. In wild-type (WT) SV40-infected CV-1P cells, VP1 was found predominantly in the cytoplasm until 36 h postinfection (p.i.), approximately the time that high levels of agnoprotein became detectable under our infection conditions. Thereafter, VP1 localized rapidly to the perinuclear region and to the nucleus. In contrast, in agnoprotein-minus mutant-infected CV-1P cells, perinuclear-nuclear accumulation of VP1 occurred much less efficiently; a significantly greater fraction of cells with predominantly cytoplasmic fluorescence was observed up to 48 h p.i. At 48 and 60 h p.i., more cells with largely perinuclear and little nuclear staining were seen than in WT-infected controls. In similar analyses with stably transfected cell lines constitutively expressing the agnoprotein, VP1 localized to the nucleus before 30 h p.i., regardless of the infecting virus. Delayed nuclear entry of VP1 in a mutant which makes no agnoprotein was also overcome in a revertant which has a second site point mutation in VP1. This suggests that an alteration of VP1 can partially overcome the defect of the agnogene mutation by enhancement of the rate of its own nuclear localization. Taken together, these results indicate that at least one function of the agnoprotein is to enhance the efficiency of perinuclear-nuclear localization of VP1.     

Identification of the Simian Virus 40 Which Replicates When Simian Virus 40-Transformed Human Cells Are Fused with Simian Virus 40-Transformed Mouse Cells or Superinfected with Simian Virus 40 Deoxyribonucleic Acid

Simian virus 40 (SV40) was rescued from heterokaryons of transformed mouse and transformed human cells. To determine whether the rescued SV40 was progeny of the SV40 genome resident in the transformed mouse cells, the transformed human cells, or both, rescue experiments were performed with mouse lines transformed by plaque morphology mutants of SV40. The transformed mouse lines that were used yielded fuzzy, small-clear, or large-clear plaques after fusion with CV-1 (African green monkey kidney) cells. The transformed human lines that were used did not release SV40 spontaneously or after fusion with CV-1 cells. From each mouse-human fusion mixture, only the SV40 resident in the transformed mouse cells was recovered. Fusion mixtures of CV-1 and transformed mouse cells yielded much more SV40 than those from transformed human and transformed mouse cells. The rate of SV40 formation was also greater from monkey-mouse than from human-mouse heterokaryons. Deoxyribonucleic acid (DNA) from SV40 strains which form fuzzy, largeclear, or small-clear plaques on CV-1 cells was also used to infect monkey (CV-1 and Vero), normal human, and transformed human cell lines. The rate of virion formation and the final SV40 yields were much higher from monkey than from normal or transformed human cells. Only virus with the plaque type of the infecting DNA was found in extracts from the infected cells. Two uncloned sublines of transformed human cells [W18 Va2(P363) and WI38 Va13A] released SV40 spontaneously. Virus yields were not appreciably enhanced by fusion with CV-1 cells. However, clonal lines of W18 Va2(P363) did not release SV40 spontaneously or after fusion with CV-1 cells. In contrast, several clonal lines of WI38 Va13A cells did continue to shed SV40 spontaneously.     

Attenuation of late simian virus 40 mRNA synthesis is enhanced by the agnoprotein and is temporally regulated in isolated nuclear systems.

Studies were performed to verify the physiological significance of attenuation in the life cycle of simian virus 40 and the role of agnoprotein in this process. For these purposes, nuclei were isolated at various times after infection and incubated in vitro in the presence of [alpha-32P]UTP under the standard conditions which lead to attenuation. Attenuation was evident by the production of a 94-nucleotide attenuator RNA, revealed by gel electrophoresis. In parallel, the synthesis of agnoprotein was studied at various times after infection by labeling the cells for 3 h with [14C]arginine, lysing them, and analyzing the labeled proteins by gel electrophoresis. Both attenuation and the synthesis of agnoprotein were predominant towards the end of the infectious cycle. At earlier times, there was almost no attenuation and no synthesis of agnoprotein. Moreover, there was almost no attenuation even at the latest times after infection in nuclei isolated from cells infected with simian virus 40 deletion mutants that do not synthesize agnoprotein. Finally, analysis by dot blot hybridization showed higher amounts of cytoplasmic viral RNA in cells infected with an agnoprotein gene insertion mutant, delta 79, that does not produce agnoprotein, compared with cells infected with wild-type virus. The present studies indicate that attenuation is temporally regulated and suggest that agnoprotein enhances attenuation in isolated nuclei and that may also enhance it in vivo.     

Simian virus 40 maturation in cells harboring mutants deleted in the agnogene

The predominant leader region of the late 16 S mRNAs of simian virus 40 encodes a histone-like, 61-amino acid, DNA-binding protein called the agnoprotein or LP1. To test the hypothesis that this protein facilitates assembly of viral minichromosomes into virions, we have studied the synthesis of virions in cells infected with mutants deleted in this region of the SV40 genome. We found that 220 S mature virions, indistinguishable from those of wild type, were produced in cells infected with these mutants. As in wild-type-infected cells, no assembly intermediates other than 75 S chromatin were observed. However, data obtained from both steady-state and pulse-chase labeling experiments indicated that cells infected with agnogene deletion mutants produced virions more slowly than cells infected with wild-type virus. Taken together with data showing that similar levels of virion proteins were present in the wild-type- and mutant-infected cells, these findings strongly suggest that LP1 plays a role in expediting virion assembly.     

Spontaneous Virus Production by Clonal Lines of Simian Virus 40-Transformed Cells and Effects of Superinfection by Deoxyribonucleic Acid from Mutant Simian Virus 40 Strains

Small amounts of infectious simian virus 40 (SV40) were recovered from parental cultures of SV40-transformed human embryonic lung (WI38 Va13A) cells, from 12 primary clones, from 17 secondary clones, and from 18 tertiary clones. The cloning experiments demonstrated that the capacity for spontaneous virus production is a hereditary property of WI38 Va13A cells. Infectious virus was not recovered from every clone at every passage. Repeated trials at different passage levels were necessary to detect virus production. Approximately one in 10(5) to 10(6) of the cells of the clonal lines initiated plaque formation when plated on the CV-1 line of African green monkey kidney cells. No increase in infectious center formation was observed after the clonal lines were treated with bromodeoxyuridine, iododeoxyuridine, or mitomycin C or after heterokaryon formation of treated cells with CV-1 cells. The clonal lines of WI38 Va13A cells were susceptible to superinfection by SV40 deoxyribonucleic acid (DNA). To determine whether only those cells which spontaneously produced virus supported the replication of superinfecting SV40 DNA, cultures were infected with DNA from a plaque morphology mutant and a temperature-sensitive mutant of SV40. After infection by SV40 DNA, approximately 100 to 4,400 times more transformed cells formed infectious centers than were spontaneously producing virus. To determine whether the resident SV40 genome or the superinfecting SV40 genome was replicating, infectious centers produced by SV40 DNA-infected WI38 Va13A cells on CV-1 monolayers were picked and the progeny virus was analyzed. Only the superinfecting SV40 was recovered from the infectious centers, indicating that in the majority of superinfected cells the resident SV40 was not induced to replicate.     

Infection of CV1 cells expressing the polyoma virus middle T antigen or the SV40 agnogene product with simian virus 40 host-range mutants

Summary SV40 viruses bearing mutations at the carboxy-terminus of large T antigen exhibit a host-range phenotype: such viruses are able to grow in BSC monkey kidney cells at 37° C, but give at least 10 000-fold lower yields than wild type virus in BSC cells at 32° C or in CV1 monkey kidney cells at either temperature. The block to infection in the nonpermissive cell type occurs after the onset of viral DNA replication. Infectious progeny virions are produced at very low efficiency. Although capsid proteins are synthesized at decreased levels, this does not account for the magnitude of the defect. Presumably some step of virion assembly or maturation is affected in these mutants. We have previously reported that the viral agnogene product, a protein throught to be involved in viral assembly or release, fails to accumulate in CV1 cells infected with host-range mutants. In polyoma virus the middle T antigen plays a role in virion maturation by influencing the phosphorylation of capsid proteins. In this communication we show that host-range mutants fail to undergo productive infection of CV1 cells expressing middle T antigen. These mutants do form plaques on an agnoprotein-expressing cell line. However, the agnoprotein does not seem to act by correcting the mutational block but rather increases the efficiency of plaque formation. This work was supported by grants CA40586 and BRSG 2S07RR07084-23 to J. M. P. and grant CA33079 to L. T., from the National Institutes of Health, Bethesda, MD.     

SV40 small T antigen enhances progression to >G2 during lytic infection.

The infection of monkey kidney (CV-1) cells with simian virus 40 (SV40) stimulates the cells into successive rounds of DNA synthesis without an intervening mitosis, leading to the acquisition of a >G2 DNA content. To elucidate the role of small t antigen in cell cycle progression and in viral replication during infection, studies were performed using an SV40 mutant (dl888) that lacks the ability to produce small t. Initially dl888-infected cells move through the first S phase at roughly the same rate as wild-type infected cells. Upon reaching G2, however, the dl888-infected cells progressed to >G2 at a reduced rate relative to wild-type. The slower rate of entry into >G2 of dl888-infected cells is associated with a decrease in total pRb and an increase in the ratio of hypophosphorylated to hyperphosphorylated pRb. The expression of cyclin D1 and p27(kip1) were elevated in dl888-infected cells compared to wild-type-infected CV-1 cells. Taken together, these results indicate that small t antigen plays a role in stimulating entry into >G2 in SV40-infected CV-1 cells, possibly by affecting the regulation of key cell cycle proteins.     

Simian virus 40 large T-antigen function is required for induction of tetraploid DNA content during lytic infection.

Infection of quiescent CV-1 cells with simian virus 40 mutant tsA30 at 37 degrees C resulted in the induction of two rounds of cellular DNA synthesis in T-antigen-positive cells, as previously described for wild-type simian virus 40. Following infection with tsA30 at 40.5 degrees C, T-antigen-positive cells were induced into S phase and reached a diploid G2 DNA content; however, a second S phase was not initiated. The failure of tsA30-infected CV-1 cells to enter tetraploid S phase at 40.5 degrees C identifies a T-antigen function, distinct from T-antigen functions responsible for stimulation of cell DNA synthesis, which is required for initiation of a second round of DNA synthesis without mitosis.     

Simian virus 40 large T antigen host range domain functions in virion assembly.

The simian virus 40 (SV40) T antigen host range mutants dl1066 and dl1140 display a postreplicative block to plaque formation which suggests a novel role for T antigen late in the viral life cycle. The host range mutants dl1066 and dl1140 are able to grow in and plaque on BSC but not on CV1 monkey kidney cells, a normally permissive host. Previous work showed that in CV1 cells infected with dl1066 and dl1140, levels of viral DNA replication and of late capsid protein accumulation were only slightly reduced and the failure to accumulate agnoprotein was not likely to be the major factor responsible for the mutants' growth defect. Here we show that the host range mutants are defective in the assembly of viral particles. SV40 assembly proceeds as the progressive conversion of 75S viral chromatin complexes to 200S-240S assembled virions. When virus-infected cell extracts are separated on 5 to 40% sucrose gradients, wild-type extracts show the greatest accumulation of viral late protein in the 200S-240S fractions corresponding to the assembled virus peak and lesser amounts in the 75S-150S fractions corresponding to immature assembly intermediates. The host range mutants dl1066 and dl1140 grown in nonpermissive CV1 cells, however, failed to assemble any appreciable amounts of mature 200S-240S virions and accumulate 75S intermediates, whereas in permissive BSC cells, levels of assembly were more slightly reduced than those of the wild type. Analysis of the protein composition of gradient fractions suggests that SV40 assembly proceeds by a mechanism similar to that proposed for polyomavirus and suggests that the host range blockage may result from a failure of such mutants to add VP1 to 75S assembly intermediates.     

Selective translation initiation on bicistronic simian virus 40 late mRNA.

We described previously a simian virus 40 (SV40) mutant, pSVAdL, that was defective in synthesis of the late viral protein VP1. This mutant, which contains a 100-base-pair fragment of adenovirus DNA encompassing the major late promoter inserted in the SV40 late promoter region (SV40 nucleotide 294), efficiently synthesizes agnoprotein, a protein encoded by the leader region of the same mRNA that encodes VP1. When the agnoprotein AUG initiation codon in pSVAdL was mutated to UUG, agnoprotein synthesis was abolished, and VP1 synthesis was elevated to wild-type levels. Because levels of late mRNA synthesis were not affected by this mutation, these results support a scanning model of translation initiation and suggest that internal translational reinitiation does not occur efficiently in this situation.     

trans-dominant defective mutants of simian virus 40 T antigen.

We constructed a collection of linker insertion mutants in the simian virus 40 (SV40) genome and studied several of these with changes limited to a part of the large T antigen gene corresponding to an amino acid sequence shared with other ATPases. Two of these mutants were found to have a novel phenotype in that they could not be complemented for plaque formation by a late-region deletion mutant. These two mutants, in contrast to other mutants in this region, were able to transform rat cells in culture at a frequency close to that of the wild-type gene. The noncomplementing mutants were found to be potent inhibitors of SV40 DNA replication despite the presence of wild-type T antigen in the transfected cells. This inhibition was shown to be the result of the introduced mutations in the large T antigen gene. We conclude that the large T antigens of the noncomplementing mutants can act as inhibitors of SV40 DNA replication.     

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.