Search

The small DNA tumor viruses have provided a very long-lived source of insights into many aspects of the life cycle of eukaryotic cells. In recent years, the emphasis has been on cancer-related signaling. Here we review murine polyomavirus middle T antigen, its mechanisms, and its downstream pathways of transformation. We concentrate on the MMTV-PyMT transgenic mouse, one of the most studied models of breast cancer, which permits the examination of in situ tumor progression from hyperplasia to metastasis.

Simian Virus 40 (SV40) and Mouse Polyoma Virus (PY) are small DNA tumor viruses that have been used extensively to study cellular transformation. The SV40 early region encodes three tumor antigens, large T (LT), small T (ST) and 17KT that contribute to cellular transformation. While PY also encodes LT and ST, the unique middle T (MT) generates most of the transforming activity. SV40 LT mediated transformation requires binding to the tumor suppressor proteins Rb and p53 in the nucleus and ST binding to the protein phosphatase PP2A in the cytoplasm. SV40 LT also binds to several additional cellular proteins including p300, CBP, Cul7, IRS1, Bub1, Nbs1 and Fbxw7 that contribute to viral transformation. PY MT transformation is dependent on binding to PP2A and the Src family protein tyrosine kinases (PTK) and assembly of a signaling complex on cell membranes that leads to transformation in a manner similar to Her2/neu. Phosphorylation of MT tyrosine residues activates key signaling molecules including Shc/Grb2, PI3K and PLCgamma1. The unique contributions of SV40 LT and ST and PY MT to cellular transformation have provided significant insights into our understanding of tumor suppressors, oncogenes and the process of oncogenesis.

We previously showed that murine polyomavirus mutants that lack both middle T (MT) and small T (ST) functions have a severe pleiotropic defect in early and late viral gene expression as well as genome amplification. The respective contribution of MT and ST to this phenotype was unclear. This work separates the roles of MT and ST in both permissive mouse cells and nonpermissive rat cells. It demonstrates for the first time a role for both proteins. To gain insight into the signaling pathways that might be required, we focused on MT and its mutants. The results show that each of the major MT signaling connections, Shc, phosphatidylinositol 3'-kinase, and phospholipase C gamma1, could contribute in an additive way. Unexpectedly, a mutant lacking all these connections because the three major tyrosines had been converted to phenylalanine retained some activity. A mutant in which all six MT C-terminal tyrosines had been mutated was inactive. This suggests a novel signaling pathway for MT that uses the minor tyrosines. What is common to ST and the individual MT signaling pathways is the ability to signal to the polyomavirus enhancer, in particular to the crucial AP-1 and PEA3/ets binding sites. This connection explains the pleiotropy of MT and ST effects on transcription and DNA replication.

The polyomavirus hr-t class of mutants has served as a major prototype to study the function of middle T + small T in the virus lytic cycle, Biochem. Biophys. Acta 695 (2), 69-95). The properties of these middle T + small T defective mutants were defined by comparisons with "wild-type" strains reconstructed by marker rescue. Similar comparisons in the A2 genetic background have revealed a number of differences, J. Virol. 75, 8380-8389). Here we describe a major divergence in their effects on cell-cycle progression of both permissive mouse NIH3T3 cells and semipermissive Fischer rat FR3T3 cells. Infection of NIH3T3 or FR3T3 cells in serum-rich medium with wild-type A2 (WTA2) or WTA2-derived middle T + small T-defective mutants did not perturb cell cycling, tested up to entry into the third cycle. In contrast, infection with four hr-t mutants analyzed, examined in detail with mutant B2, resulted in an accumulation of cells in G2/M in a dose-dependent and serum-independent manner. The arrest began in the first cell cycle. At multiplicities of infection above 10 PFU/cell, 50-80% of the cell population became arrested by the end of the second cycle. FR3T3 arrested cells detached from the monolayer with a rounded up morphology. Three other hr-t mutants investigated were also found to arrest cells in G2/M. Expression of middle T and/or small T either in trans or in cis did not abrogate this cell-cycle arrest, as demonstrated in the latter case with the middle T + small T expressing strain "wtB2" obtained by repair of the B2 deletion. In FR3T3 cells, the induction of a cell-cycle arrest by wtB2 was accompanied by a severe delay and reduction in neoplastic transformation relative to WTA2 used at equal dose. Mutation(s) in the C-terminal domain of large T antigen, upstream of the site-specific DNA binding activity, is necessary for the cell-cycle block. The possible causes for the cell-cycle block are discussed.

A comparative analysis of the lytic cycle of wild-type polyomavirus and middle T and small T defective mutants was carried out in the A2 genetic background. The results contrast with those obtained in comparisons between the hr-t type and their middle-T small-T-producing partners as previously described (20). The A2-derived mutants were found to share the maturation defect previously described for the hr-t mutants. However, their defect in DNA replication was more acute, resulting in a 5- to 100-fold decrease in the accumulation of viral genomes. Furthermore, their gene expression pattern was affected. A2-derived mutants displayed an early defect resulting in a 4- to 16-h delay in the expression of large T, and an alteration of the early-to-late transcriptional switch. In wild-type A2 infection, this switch is characterized by a large increase in the accumulation of early transcripts followed by late transcripts after the appearance of middle T and small T proteins and the onset of viral DNA replication (L. Chen and M. M. Fluck, J. Virol. 75: 8368-8379, 2001). In the mutant infection, increases in both classes of transcripts were delayed and reduced, but the effect on early transcripts was more pronounced. As has been described previously for the hr-t mutants (E. Goldman, J. Hattori, and T. Benjamin, Cell 13:505-513, 1979), the magnitude of these defects depended upon experimental conditions. Experiments using cytosine beta-arabinofuranoside to reduce genome amplification suggest that the effect of middle T-small T on the transcriptional switch is not solely mediated by the effect of these protein(s) on increasing the number of templates. These data provide the first direct demonstration of an effect of middle T and/or small T in the viral transcription pattern during viral infection. The results agree with previous results obtained with plasmid reporters and with our understanding that the downstream targets of the middle T signaling pathway include three transcription factors that have binding sites in the enhancer domain that play a key regulatory role in the expression of the viral genes.

BALB/c mice that developed tumors 7 to 8 months following neonatal infection by polyomavirus (PYV) wild-type strain A2 were characterized with respect to the abundance and integrity of the viral genome in the tumors and in 12 nontumorous organs. These patterns were compared to those found in tumor-free mice infected in parallel. Six mice were analyzed in detail including four sibling females with mammary gland tumors. In four of five mammary gland tumors, the viral genome had undergone a unique deletion and/or rearrangement. Three tumor-resident genomes with an apparently intact large T coding region were present in abundant levels in an unintegrated state. Two of these had undergone deletions and rearrangements involving the capsid genes and therefore lacked the capacity to produce live virus. In the comparative organ survey, the tumors harboring replication-competent genomes contained by far the highest levels of genomes of any tissue. However, the levels of PYV genomes in other organs were elevated by up to 1 to 2 orders of magnitude compared to those detected in the same organs of tumor-free mice. The genomes found in the nontumorous organs had the same rearrangements as the genomes residing in the tumors. The original wild-type genome was detected at low levels in a few organs, particularly in the kidneys. The data indicate that a systemic increase in the level of viral genomes occurred in conjunction with the induction of tumors by PYV. The results suggest two novel hypotheses: (i) that genomes may spread from the tumors to the usual PYV target tissues and (ii) that this dissemination may take place in the absence of capsids, providing an important path for a virus to escape from the immune response. This situation may offer a useful model for the spread of HPV accompanying HPV-induced oncogenesis.

The fate of the genome of the polyoma (Py) tumor virus following integration in the chromosomes of transformed rat FR3T3 cells was re-examined. The viral sequences were integrated at a single transformant-specific chromosomal site in each of 22 transformants tested. In situ amplification of the viral sequences was observed in 24 of 34 transformants analyzed. Large T antigen, the unique viral function involved in initiating DNA replication from the viral origin, was essential for the amplification process. There was an absolute requirement for a reiteration of viral sequences and the extent of the reiteration affected the degree of amplification. The reiteration may be important for homologous recombination-mediated resolution of in situ amplified sequences. Among 11 transformants harboring a 1 to 2 kb repeat, the degree of amplification was transformant-specific and varied over a wide range. At the high end of the spectrum, the genome copy number increased 1300-fold at steady state, while at the low end, amplification was below twofold. Some aspect of the host chromatin at the site integration that affected viral gene expression, also directly or indirectly modulated the amplification. Use of high-resolution electrophoresis for the analysis of the integrated amplified sequences revealed a recurring novel pattern, consisting of a ladder with numerous bands separated by a constant distance approximately the size of the Py genome. We suggest that this pattern was generated by conversion of the amplified viral genomes to head to tail linear arrays with cell to cell variations in the number of genome repeats at single, transformant-specific, chromosomal sites. In light of the known "out of schedule" firing of the Py origin, we propose an "onion skin" structure intermediate and present a homologous recombination model for the conversion from onion skins to linear arrays. The relevance of the in situ amplification of the Py genome to cellular gene amplification is discussed. Finally, these results clarify our understanding of the integration of the Py genome in rat cells. They suggest that, in most cases, the multiple bands previously described in Py-transformants are likely to reflect genome amplification rather than multiple independent integration events, as assumed in the past. This interpretation is congruent with the accepted view that the integration of the Py genome is a rare and rate-limiting event in transformation.

A correlation between polyomavirus-induced oncogenesis and viral persistence on the one hand and/or prolonged genome replication potential on the other was established with respect to their respective organ distributions. Prolonged replication potential is defined as the capacity of a genome to replicate in a given organ from the time of infection up to the onset of oncogenesis. This conclusion was derived following intraperitoneal infection of BALB/c mice with wild-type strain A2. Viral genomes were used as parameters of persistence and replication and were detected by Southern blotting and PCR analysis. The major tumor target organs (mammary gland, skin, and bone), which have not been previously analyzed for persistence, were compared with other, non-tumor-prone organs (kidney, liver, lung, spleen, and salivary gland). A progressive loss of viral genomes was observed in all tissues as a function of time postinfection; however, genomes were shown to persist through 20 weeks postinfection in the mammary glands, skin, and bones to an extent similar to that in the previously described kidneys (D. J. McCance, J. Virol. 39:958-962, 1981; W. P. Rowe, J. W. Hartley, J. D. Estes, and R. J. Huebner, Natl. Cancer Inst. Monogr. 4:189-209, 1960). Thus, tumors arise among organs that sustain a persistent infection, but not all such organs develop tumors (e.g., the kidney). The capacity of organs to support de novo replication at various ages, including the age reached when the first tumors are detected, was also determined using a 3-day infection period for ages between 0 and 7 weeks. For all organs tested, a higher level of genomes was observed in organs of mice infected as neonates than in those infected after the age of 3 weeks. However, marked organ-specific differences were seen in the degree and timing of loss of replication. In particular, viral genome replication, although reduced, was maintained in the mammary glands, skin, and bones of adult animals, in contrast to the kidneys. We conclude that organ-specific oncogenesis correlates with two organ-specific parameters: persistence of viral genomes and prolonged viral genome replication potential. This may reflect a requirement for continued viral genome replication and/or gene expression for tumorigenesis. In turn, these parameters may be linked to the tissue-specific continued capacity for cellular division.

The first known member of the Polyomavirus family, murine Polyomavirus (MPyV), was discovered because of its oncogenic properties. The genetic simplicity of MPyV (shared with all members of the Py family), the wide spectrum of tumors induced by MPyV, and the convenient properties of its natural host, the mouse, make it a particularly interesting model system to study oncogenesis. This paper briefly reviews the virus infectious cycle and our current understanding of the viral proteins that are involved in oncogenesis, and focuses on recent studies on oncogenesis of the mammary gland. Mammary gland ductal adenocarcinomas develop at high frequency and with short latency in infected immunoincompetent adult female or normal neonatal mice or in transgenic mice expressing the viral oncogene, middle T. These tumors provide excellent model systems for the study of human breast cancer.

Polyomavirus infection of adolescent athymic female mice causes a high incidence of mammary adenocarcinomas. We have examined the role of ovarian hormones, age and mammary gland developmental stage at infection on subsequent tumor induction, viral replication and gene expression. Ovariectomy (OVX) of adolescent mice 1 week before infection decreased mammary tumor incidence and number, and significantly increased tumor incidence and number, and significantly increased tumor latency. Reduction in tumorigenesis was observed to a lesser degree if mice were OVX at the time of or after infection, indicating that ovarian hormones are mainly required for tumor initiation. Tumor incidence was also reduced with increasing age; OVX prior to infection at older ages drastically reduced tumor development. Treatment of OVX adult mice with estrogen + progesterone for 1-3 weeks prior to infection was unable to restore tumorigenesis to the level observed in intact mice. Thus, in contrast to adolescent mice, the continued presence of ovarian hormones after infection was required for maximal tumorigenesis in adult mice. The decreased tumorigenesis observed in older animals is not likely due to increased differentiation since late pregnant mice with well differentiated mammary glands remained highly susceptible to tumorigenesis. At 10 days post infection, the levels of viral genomes were moderately high and similar in all experimental groups. Early viral protein and middle T-associated kinase levels were undetectable in infected tissues in all experimental conditions. However, high levels were found in tumors, perhaps reflecting a high dosage requirement for oncogenesis.

A major role for polyomavirus middle T/small T antigens in viral DNA synthesis was uncovered by examining the replication of middle T/small T-deficient mutants (hr-t mutants). hr-t mutants in the A2 genetic background showed a 16- to 100-fold defect in genome accumulation relative to the wild type when infections were carried out in exponentially growing NIH 3T3 cells in medium supplemented with low levels of serum (< 2.0%). A proportional decrease in the level of viral early transcripts was also seen. The replication defect of the hr-t mutants was partially overcome in the presence of the phorbol ester 12-O-tetradecanoylphorbol-13-acetate. The defect was also alleviated by a duplication encompassing the alpha core enhancer domain that contains binding sites for the transcriptional activators PEA1/AP-1 and PEA3/c-ets. Such a duplication is present in all naturally occurring hr-t mutants and absent in the A2 strain. The effects of 12-O-tetradecanoylphorbol-13-acetate and alpha core duplication were additive but did not fully complement the absence of middle T/small T. In mixed infection competition experiments with two hr-t mutants, a genome that carried an alpha core duplication had a replication advantage (up to 17-fold) over a genome without duplication. This result demonstrates that one effect of the duplication is exerted directly at the level of DNA replication. The advantage of the duplication-bearing genome was established during the earliest stages of replication and was not further amplified in later rounds of replication. In the presence of middle T/small T, both genomes replicated to high levels and the advantage of the duplication-bearing genome was eliminated. On the basis of these results, we propose that factors that bind the alpha core domain (presumably PEA1 and PEA3) are present in limiting amounts in exponentially growing NIH 3T3 cells and play a crucial role in polyomavirus DNA replication. We further suggest that middle T and/or small T stimulates viral DNA replication by activating these factors. The fact that all middle T-/small T-defective hr-t mutants have evolved to contain enhancer duplications that encompass the PEA1 and PEA3 binding sites in the alpha core domain and partially restore their replication defect (A. Amalfitano, M. C. Chen, and M. Fluck, unpublished data) provides an adequate explanation for the fact that the importance of the role of the middle T and/or small T function in DNA replication has not been recognized previously. Much evidence is available in support of separate elements of this model.(ABSTRACT TRUNCATED AT 400 WORDS)

The cell cycle dependence of polyomavirus transformation was analyzed in infections of nonpermissive Fischer rat (FR3T3) cells released from G0. A 5- to 100-fold (average, ca. 20-fold) difference in relative frequency of transformation was found for cells infected in the early G1 phase of the cell cycle compared with cells infected in G2. Differences in the relative level of early viral gene expression in those two cell populations were equivalent to those obtained for transformation frequencies. The difference in transformation potential was accounted for only in part by a cell cycle control of viral adsorption (2- to 15-fold effect). Furthermore, in cells infected in the early G1 phase, viral gene expression was induced as a big synchronous burst of large transcripts of variable sizes, delayed till the G1 phase of the cell cycle after that in which infection took place. Thus, the results demonstrate that the abortive infection cycle of G0-released FR3T3 cells is cell cycle regulated at least at two steps: adsorption and another early step, nuclear transport, decapsidation, up to or including the transcription of the viral early genes. The cell cycle regulation of these steps results in a similar regulation of the abortive and stable transformation processes, although it is more pronounced for the latter. A model implicating c-fos and c-jun is proposed.

Viral replication in mice infected with murine polyomavirus strains with novel enhancer rearrangements was analyzed by direct in situ hybridization of whole mouse sections and by hybridization of nucleic acids extracted from a specific set of organs. The enhancer rearrangements included a deletion of the B domain as well as duplications within the A domain. Comparisons between enhancer variants demonstrate that the B domain plays an important role in replication in most organs, in particular in the kidney, at the neonatal stage (days 0 to 7 postbirth). In contrast, the B domain is not required in those organs which can sustain replication in the adult, i.e. mammary gland, skin, and bone (class I organs [J. J. Wirth, A. Amalfitano, R. Gross, M. B. A. Oldstone, and M. M. Fluck, J. Virol. 66:3278-3286, 1992]). Altogether, the results suggest that the B and A domains mediate very different functions in infection of mice, controlling the acute and persistent phases of infection, respectively. A model of mouse infection based on the crucial role of differentially expressed host transcription factors is presented.

We have characterized mammary oncogenesis induced after polyomavirus infection of adult female nude mice regarding histopathogenesis, viral replication and viral and cellular oncogene expression. A unique transient generalized epithelial hyperplasia was observed (starting at 2 weeks post infection), preceding the development of dysplasias (onset 6 weeks post infection) and multiple neoplasias (onset 6 weeks post infection) in all glands. The ductal epithelium was the target for neoplastic transformation, and the occurrence of numerous ductal dysplasias coincided with the appearance of frank tumors. Stromal abnormalities were also seen. Tumor growth was not dependent upon ovarian hormones, and new tumors continued to develop in ovariectomized mice. Viral replication, high although variable, preceded but did not correlate with oncogenesis. Most but not all tumors contained high levels of unintegrated viral DNA. Tumors produced very low levels of live virus. Viral gene expression was markedly increased in the tumors compared with the infected but morphologically normal glands. The expression of c-myc was moderately increased (fourfold); changes in c-int-2 and c-Ha-ras expression were slight and inconsistent, while expression of c-neu and c-int-1 was unchanged.

A novel organ- and age-specific pattern of polyomavirus DNA replication in mice is described. Two broadly defined classes of response to polyomavirus infection were observed: class I organs (mammary gland, bone, and skin) responded with high levels of replication in neonate mice and moderate levels in adults; class II organs (kidney, liver, and lung) responded with high levels in neonates and very low levels in adults. Thus, aging affected replication in all organs, and organ specificity was superimposed on this age-related decrease. We argue that the organ- and age-specific pattern likely reflects in part the activities of a multiplicity of general or tissue-specific, age-dependent transcription factors, which modulate viral replication or viral transcription or both. Interestingly, the majority of tumors in mice infected as neonates or as immunoincompetent adults originate in class I organs, suggesting that the ability to replicate in adult tissues is an important factor controlling polyomavirus oncogenesis. From the analysis of the infection process in adult mammary glands, a novel mode of polyomavirus infection emerged which contrasts with that derived from observations of tissue culture systems. A nonproductive infection was seen, characterized by very low levels of live virus (in the range of 10(-4) PFU per cell) and maintenance of the viral genome in an unintegrated, moderately replicating state. Maintenance of the viral genome was accomplished without integration into host cell DNA in all three tumor-prone organs, both prior to as well as beyond oncogenesis.

The uniformly lethal development of mammary tumors in polyomavirus-infected adult female nude mice was prevented by adoptive cell transfer of polyomavirus-immune splenocytes or peritoneal cells. Transferred immune cells also lowered the growth rate of emerging tumors. The induction of other relatively less frequent tumors of the skin and bone was decreased as well. Using in situ hybridization of whole-body sections as well as hybridization of nucleic acids from the mammary glands, we show for the first time that transferred immune cells, but not normal cells, virtually eliminated virus signal in the whole mouse and in the mammary glands. Since infected and tumorous mammary glands produce very little infectious virus, it appears that a major mechanism mediating the prevention of polyomavirus oncogenesis involves the immunological elimination of nonproductively and persistently infected cells.

The fate of polyomavirus genomes in stable transformants was examined in experiments in which rat or hamster cells were infected with mixed viral populations containing either two distinguishable wild types or a wild type plus a transformation-deficient mutant. The results demonstrate that in 90% of the cases in either situation, a single polyomavirus parental genome became integrated into the host genome. Both parents in the mixed infection were present in the same cells and both persisted in the infected cells until transformants appeared, eliminating the possibility that one virus would exclude the other in the early steps of the infection process. We believe that these results can be generalized to transformation events derived from normal single infections. Thus, contrary to previous results for the number of integration sites determined by restriction endonuclease and blot hybridization analysis, it appears that the probability for more than one integration event in transformation by polyomavirus is low (1 in 10). A reinterpretation of previous data is proposed.

Secondary mutations which increase the efficiency of suppression of nonsense mutations in the rIIB cistron of bacteriophage T4 have been isolated. These secondary mutations, called context mutations, map at sites very close to the nonsense codon, possibly on the promotor distal side. In context-nonsense double mutants, the amount of suppressed gene product is increased approximately 10-fold. The context mutations examined can act on the UAA (ochre) nonsense allele as well as on the UAG (amber) nonsense allele at a given site. These context mutations affect all suppression mechanisms analyzed (genetic suppressors, 5-fluorouracil suppression and spontaneous suppression). We suggest that context mutations affect information which is significant to the termination of polypeptide chains. According to our view, context mutations change the immediate neighborhood of nonsense mutations and so reduce the degree of resemblance to the sequences normally used for the termination of translation.

Clones of stably transformed rat cells from the F-111 (Fisher rat embryo) and NRK (normal rat kidney) cell lines have been obtained by complementation using pairs of nontransforming polyoma mutants of the hr-t and ts-a classes. A total of 78 clones were isolated and studied, 21 from the NRK line, and 57 from the F-111 line. These "complementation transformed clones" were then examined for the presence and expression of each of the parental mutant viruses. The expression of viral T (tumor) antigens was analyzed by immunoprecipitation. Every one of 54 clones examined express the 22K (small) and 56K (middle) T antigens which are the products of the hr-t viral gene. This indicates a requirement for the presence of a wild-type hr-t allele contributed by the ts-a mutant parent. Ten clones lack detectable large T antigen, while four show thermolabile large T antigen. These results support the conclusion that middle and small but not large T antigen(s) are essential for the maintenance of the transformed phenotype. Most of the 57 F-111 clones are virogenic when fused to mouse cells under permissive conditions and yield both mutant types. Surprisingly, wild-type recombinants have been recovered from 40 of the F-111 clones. Three clones show evidence of retention of only the ts-a mutant genome. Virus cannot be rescued from the majority of the 21 NRK clones. The few clones which are virogenic, however, yield both hr-t and ts-a mutants. Integration patterns of several clones analyzed by Southern blotting confirm the expectations based on viral T antigen(s) and virus rescue analyses. Hr-t mutant genomes, though not required for the maintenance of the transformed state, are frequently retained in complementation transformed clones. Tandem integration of the two parental mutants is clearly demonstrated in one clone and implicated in the other eight out of nine virogenic F-111 complementation transformed clones examined. This observation provides the basis of a model for the generation of wild-type recombinants following fusion of F-111 clones.

We have investigated the occurrence and role of polyomavirus DNA synthesis in neoplastic transformation by this virus. We show that after infection of Fischer rat F-111 cells at 37 degrees C, there is two- to threefold increase in the level of viral DNA as compared with the input signal, with a peak observed between 5 and 7 days postinfection. Viral DNA synthesis is about 10 times higher at 33 degrees C and increases up to 15 days postinfection. Most of the viral DNA produced is supercoiled (form I DNA). On the basis of in situ hybridization, it appears that viral replication is restricted to a small fraction of the population. At the lower temperature, more cells are permissive for viral DNA synthesis and the level of synthesis per permissive cell is higher. The DNA synthesis observed is large T-antigen dependent, and the increase in viral DNA synthesis at 33 degrees C is paralleled by an increase in the expression of this viral protein. When large T antigen is inactivated, the half-life of de novo-synthesized viral DNA is less than 12 h, suggesting that large T antigen may be responsible for the stability of the viral genomes as well as their synthesis. Surprisingly, at early times postinfection (0 to 48 h), when the essential function of large T antigen in transformation is expressed (as demonstrated in shift-up experiments with tsa mutants), the level of large T antigen is below the detection level and is at least 10-fold lower than the levels observed in permissive infections at the start of viral DNA synthesis. The difference in viral DNA at 37 and 33 degrees C allowed us to study its effect on transformation. Although an increase in transformation frequency is observed in wild-type A2 infections carried at 33 degrees C (frequencies two to three times higher than at 37 degrees C), this increase appears to be unrelated to the increase in viral DNA synthesis. Furthermore, the overall level of viral DNA and large T antigen in F-111 cells may not affect the integration of the viral genome, since the patterns of integration in cells transformed by wild-type A2 at 33 and 37 degrees C appear similar. The results are compatible with a role for large T antigen in integration-transformation which is not simply to amplify the viral genome to enhance the probability of its integration.

A mutation in polyomavirus large T antigen which affects viral DNA synthesis was discovered in strain NG59RA (RA). The effect was most visible in nonpermissive cells. Although a substantial yield in DNA synthesis is normally observed in infections of Fischer rat cells when these are maintained at 33 degrees C (D.L. Hacker, K.H. Friderici, C. Priehs, S. Kalvonjian, and M.M. Fluck, p. 173-181, in R.E. Moses and W.C. Summers, ed., DNA Replication and Mutagenesis, 1988; D.L. Hacker and M.M. Fluck, Mol. Cell. Biol., in press), a 10- to 20-fold decrease in yield was obtained in infections with RA. The yield of free viral DNA in RA transformants was also strongly diminished, whether the transformants were maintained at 37 or 33 degrees C. A large reduction in the apparent number of integration sites, as well as a small reduction in the incidence of tandem integration of the viral genome, was observed in F-111 or FR-3T3 cells transformed by the mutant strain. This appears not to be directly related to the number of integration templates. A DNA fragment was identified which rescues these phenotypes. The fragment is located between the HindIII and NsiI restriction sites (nucleotides 1656 to 1910), a region which encodes only large T antigen. Sequence analysis of this region reveals a C-to-G transition at nucleotide 1791 which causes a proline-to-alanine change in the amino acid sequence of large T antigen. No other mutations have been previously reported in this region of large T antigen.

Transformed Fischer rat cells derived from coinfections with wild-type polyoma virus and transformation-deficient hr-t deletion mutants were studied. The structure of the integrated viral DNA was determined by Southern blot analysis of high-molecular-weight DNA. In several cases the two parental genomes were found cointegrated in a head-to-tail manner at a single site in the host DNA. The results indicate that parental viral genomes recombine in Fischer rat cells in the process of transformation, and that the commonly observed tandem integration of viral genomes in polyoma-transformed cells must be due, in part, to recombination events.

Seven clonal lines of Fischer rat cells transformed with ts-a mutants of polyoma virus were studied. Four clones are characterized by a temperature-sensitive (ts) and three clones by a temperature-insensitive-transformed phenotype. Six clones have retained a functional though temperature-sensitive large T antigen, as judged by a 10- to 20-fold amplification of viral sequences in clones grown at low temperature compared to those grown at high temperature. No amplification is observed in one non-ts clone. As analyzed by Southern blotting, no obvious difference appears in the integration pattern of viral sequences in ts and non-ts clones concerning the number of sites of genome integration, the presence or absence of tandem repeats of the viral genome, or the absence of specific viral sequences. In autoradiograms of gel-electrophoresed immunoprecipitates, no correlation can be drawn between the amounts of either large T antigen or middle T antigen and the type of transformed state of the clones under the conditions tested. In assays of the middle T-antigen-associated kinase, no reproducible difference can be observed between the non-ts and ts clones. Finally, no correlation was observed between a temperature-insensitive phenotype and the production of an N-terminal fragment of large T antigen. Thus the molecular basis for the difference between ts-a transformants with ts or non-ts phenotypes remains elusive.

A series of 10 Fischer rat transformed clonal cell lines were independently obtained in infections with a defective polyomavirus containing a scrambled genome except for an intact middle and small T-antigen-coding region. These cells synthesize middle and small T antigens; no fragment of large T antigen can be detected in any of them. The transformed phenotype of this set of cell lines (designated LT-) has been studied with respect to serum dependence, saturation density, and anchorage independence and compared with the phenotype of a set of six transformants (designated LT+) which synthesize detectable to high levels of shortened or normal-sized large T antigen. Both the LT+ and the LT- groups of polyomavirus transformants display a range of transformed phenotypes. These ranges overlap, and the variations within each group are larger than the variations between the two groups. Thus, the results suggest that, for established Fischer rat fibroblasts, the maintenance of any of the three phenotypes tested and, in particular, of serum independence is not necessarily correlated with the levels of large T antigen or fragments thereof.

An unusually high incidence of interviral recombination was found in the process of integration of the polyomavirus genome concomitant with neoplastic transformation of nonpermissive cells. Transformants were isolated after mixed infections of Fischer rat cells with two mutants lacking restriction endonuclease sites and were analyzed for the presence of unselected integrated recombinant restriction fragments. A large fraction of the transformants isolated (38% of the 64 transformed cell lines studied) contained recombinant viral genomes that had undergone recombination in a 1.3-, 1.7-, or 3.6-kilobase-pair interval. More than 90% of these recombinant transformants showed evidence of crossovers in multiple intervals. To our knowledge, the recombination frequencies observed in these experiments represent the highest frequencies of homologous recombination reported for a mitotic mammalian system that does not involve transfection. In contrast to the elevated level of recombination in the integrated viral genomes, no evidence of recombination was obtained among the replicated unintegrated pool of viral genomes isolated from the same population of infected cells from which the recombinant transformants were derived. Either of two hypotheses can provide an explanation for the segregated recombination: either recombination occurs at elevated levels in a small, recombination-prone fraction of the population destined to become transformed, or recombination occurs only among those viral genomes which are engaged in the process of integration and thus interact with a recombinogenic host machinery (for example, the host scaffold). We favor the latter hypothesis.

We have characterized the viral sequences integrated in a polyomavirus-transformed mouse cell line, Py-3T3 (clone Py-6), and followed their excision and packaging upon superinfection. The polyomavirus sequences contained in Py-6 cells are present as a single insert of nonidentical tandem copies which includes, in addition to a normal middle T-antigen-coding region, some very rearranged sequences. Infection of Py-6 cells with polyomavirus strains encoding a normal large T antigen leads to the reproducible recovery in the resulting viral stock of specific defective viral genomes. The defective genomes contain a wild-type coding region for middle and small T antigens and intact viral origin and enhancer sequences. The remainder of the viral genome is rearranged or lost, so that there is no capacity to code for large T antigen or viral capsid proteins. The recovered defective sequences are also found integrated in Py-6 genomic DNA. Presumably, in infections of Py-6 cells, large T antigen, provided by the superinfecting virus, amplifies and excises the integrated viral sequences. The superinfecting helper virus must also produce viral capsids for packaging of the defective viral DNA and thus provides a means to shuttle the defective sequences from the mouse cells into other hosts, such as rat cells. In the latter host, the defective sequences are able to induce transformation.