Characterization of a simian virus 40-transformed Fanconi anemia fibroblast cell line

We have characterized a SV40-transformed human fibroblast cell line (GM6914) derived from a patient with Fanconi anemia (FA) in order to establish its usefulness for biochemical and genetic experiments, including DNA-mediated gene transfer. GM6914 cells have a growth rate similar to that of SV40-transformed normal human fibroblasts and an indefinite lifespan in culture. As has been established for other FA cell types, GM6914 cells have an increased sensitivity to DNA-crosslinking agents such as mitomycin C (MMC). The D10 for GM6914 cells is 8 times lower than for equivalent controls. GM6914 cells also have an elevated frequency of spontaneous chromosome aberrations and this frequency can be increased by MMC concentrations which show no effect on control cells. Genetic complementation studies with lymphoblasts derived from two affected sibs of the donor of GM6914 cells show that GM6914 belongs to FA complementation group A. In DNA-transfection studies using plasmid pRSVneo, colonies of GM6914 cells resistant to the drug G-418 were observed at frequencies ranging from 1.7 to 16 X 10(-4), values similar to those observed with several other SV40-transformed human cell lines. GM6914 should be a useful recipient cell line in experiments using DNA-mediated gene transfer to clone the normal allele of the gene which is defective in FA complementation group A. GM6914 would also be an excellent cell line for studies on mutagenesis, recombination and repair using plasmid vectors.     

Analysis of a large-T-antigen variant expressed in simian virus 40-transformed mouse cell line mKS-A.

Earlier reports had suggested that the large T antigen expressed in simian virus 40 (SV40)-transformed mKS-A cells may be replication defective. Our experiments support these earlier observations showing that the mKS-A T antigen has a reduced DNA-unwinding activity in vitro. To investigate the molecular basis for this defect, we have isolated from an mKS-A genomic library an EMBL-3 bacteriophage clone carrying in its insert a full-length SV40 DNA element that most likely encodes the expressed T-antigen variant. DNA sequencing revealed only one nonconservative amino acid exchange, Asp to Asn at residue 636. Surprisingly, when a plasmid clone carrying the mKS-A T-antigen-coding sequence was transfected into monkey cells, we found that it replicated quite efficiently, probably suggesting that a high nuclear concentration of the variant T-antigen form compensates for the partial biochemical defect. However, a high nuclear concentration of T antigen was also found in mKS-A T-antigen-transformed mouse cells, yet a fusion of these cells to permissive monkey cells failed to induce in situ replication and excision of integrated SV40 DNA. We discuss possible reasons for the different behavior of T antigen in monkey cells and in mouse cells and suggest that one possibility for the replication-negative phenotype in transformed cells may be related to the fact that T antigen forms a tight complex with the cellular p53 protein in mouse cells but not in monkey cells.     

DNA binding properties of a mutant T antigen from the simian virus 40-transformed human cell line simian virus 80

We investigated whether the T antigen of the simian virus 40-transformed human cell line simian virus 80 ( SV80 ) specifically recognizes DNA sequences of its own template, i.e, the viral sequences integrated in the SV80 cellular genome. In vitro DNA binding experiments clearly indicated that, in contrast to wild-type T antigen, SV80 T antigen does not specifically bind to sites on the integrated viral DNA in SV80 cells.     

Quantification of expression of linked cloned genes in a simian virus 40-transformed xeroderma pigmentosum cell line.

We wished to determine whether simian virus 40 (SV40)-transformed xeroderma pigmentosum cells, despite their defective DNA repair, were suitable for DNA-mediated gene transfer experiments with linked genes. Expression of a nonselectable gene (cat, coding for chloramphenicol acetyltransferase [CAT]) linked to a selectable gene (gpt, coding for xanthine-guanine phosphoribosyltransferase [XPRT]) in the plasmid pSV2catSVgpt was quantified after transfection of SV40-transformed xeroderma pigmentosum [XP20s(SV40)] and normal human [GM0637(SV40)] fibroblast cell lines. A novel autoradiographic assay with [3H]xanthine incorporation showed 0.5 to 0.7% phenotypic expression of XPRT in both cell lines. Without selection, transient CAT activity was 20 times greater in the GM0637(SV40) than in the XP20s(SV40) cells, and transient XPRT activity was 5 times greater. Both of these transient activities were increased and equalized in both cell lines by transfection with pRSVcat or pRSVgpt. Genotypic transformation to gpt+ occurred at a frequency of 2 X 10(-4) to 4 X 10(-4) in both cell lines with pSV2catSVgpt. After 2 to 3 months in selective medium, stable expression of the (nonselected) cat gene was found in 11 (92%) of 12 gpt-containing clones derived from GM0637(SV40) cells and in 13 (81%) of 16 gpt-containing clones from XP20s(SV40) cells. However, the levels of CAT activity did not correlate with those of XPRT activity, and both of these activities varied more than 100-fold among different clones. Copies (1 to 4) of the gpt gene were integrated in four clones of the GM0637(SV40) cells having an XPRT activity of 1 to 5 nmol/min per mg, but 5 to 80 copies were integrated in four XP20s(SV40) clones with an XPRT activity of 0.8 to 1.8 nmol/min per mg. This study shows that XP20s(SV40) is as suitable for gene transfer experiments as the normal human line GM0637(SV40).     

Karyology and tumorigenicity of a simian virus 40-transformed Chinese hamster cell clone.

A pseudodiploid clone of chinese hamster cells tranformed in vitro with Simian virus 40 (SV40) was isolated from soft agar and injected into nude mice through three successive passages with a short in vitro cultivation between each animal inoculation. the original clone and the three subsequent tumor populations were characterized in regard to SV40 T antigen staining, modal chromosome number, and Giemsa-banded karyotype. All tumor cell lines maintained the pseudodiploid mode, as well as the positive sv40 t antigen staining. Nonrandom chromosomal changes included loss of one of the X chromosomes, additions of abnormal variants of chromosomes No. 1 and No. 2, the appearance of unidentified marker chromosomes, and the loss of autosomes No. 5, No. 6, and No. 11. The deletion of one of the X chromosomes occurred with about the same frequency in all cell lines. Additions of abnormal chromosomes No. 1 and No. 2 tended to recur more often in the tumor cell lines than in the original clone. the appearance of marker chromosomes, as well as the loss of autosomes No. 5, No. 6, and No. 11 demonstrated a correlation with tumorigenicity. Yet, the three successive passages of the cells through the animal did not select for a tumor population with a single, homogeneous karyotype.     

Selectivity of interferon action in simian virus 40-transformed cells superinfected with simian virus 40.

Treatment of African green monkey kidney CV-1 cells with human alpha interferons before infection with simian virus 40 (SV40) inhibited the accumulation of SV40 mRNAs and SV40 T-antigen (Tag). This inhibition persisted as long as the interferons were present in the medium. SV40-transformed human SV80 cells and mouse SV3T3-38 cells express Tag, and interferon treatment of these cells did not affect this expression. SV80 and SV3T3-38 cells which had been exposed to interferons were infected with a viable SV40 deletion mutant (SV40 dl1263) that codes for a truncated Tag. Exposure to interferons inhibited the accumulation of the truncated Tag (specified by the infecting virus) but had no significant effect on the accumulation of the endogenous Tag (specified by the SV40 DNA integrated into the cellular genome). The level of Tag in SV40-transformed mouse SV101 cells was not significantly decreased by interferon treatment. SV40 was rescued from SV101 cells and used to infect interferon-treated and control African green monkey kidney Vero cells. Tag accumulation was inhibited in the cells which had been treated with interferons before infection. Our data demonstrate that even within the same cell the interferon system can discriminate between expression of a gene in the SV40 viral genome and expression of the same gene integrated into a host chromosome.     

Abnormal sister-chromatid exchange induction by 3-aminobenzamide in an SV40-transformed Indian muntjac cell line: relationships with DNA maturation and DNA-strand breakage.

In SVM cells, an SV40-transformed line of Indian muntjac fibroblasts, levels of sister-chromatid exchanges are known to be abnormally high after UV-irradiation or alkylation. The SVM line is also known to have a defect in the processing of DNA-strand breaks. Sister-chromatid exchange in other cells is known to be stimulated by the poly(ADP-ribose) transferase inhibitor, 3-aminobenzamide, which also retards DNA-break sealing. Sister-chromatid exchanges in SVM cells are found to be hypersensitive to 3-aminobenzamide, or to nicotinamide deprivation which similarly inhibits poly(ADP-ribosyl)ation; DNA-strand breaks are likewise induced by 3-aminobenzamide. Bromodeoxyuridine, needed to detect sister-chromatid exchanges, is more toxic to SVM cells and itself induces sister-chromatid exchanges to a greater extent than it does in normal muntjac cells. However, in contrast to the situation reported for other cell types prone to sister-chromatid exchange (the Chinese hamster ovary mutant EM9 and human Bloom's Syndrome cells), SVM cells do not show an abnormal delay in DNA maturation when, under the influence of bromodeoxyuridine and 3-aminobenzamide, they show a high level of sister-chromatid exchange. The mechanism by which BrdU exerts its effects can largely be explained in terms of familiar effects on deoxyribonucleotide pools and DNA integrity. 3-Aminobenzamide, however, induces sister-chromatid exchanges in SVM cells by another mechanism.     

Nonlytic simian virus 40-specific 100K phosphoprotein is associated with anchorage-independent growth in simian virus 40-transformed and revertant mouse cell lines.

Normal fibroblasts display two distinct growth controls which can be assayed as requirements for serum or for anchorage. Interaction of mouse 3T3 fibroblasts with simian virus 40 (SV40) thus generates four classes of transformed cells. We have examined viral gene expression in these four classes of cell lines. Immunoprecipitation of [35S]methionine-labeled cell extracts with an antiserum obtained from tumor-bearing hamsters detected the SV40 large T and small t proteins (94,000 molecular weight [94K], 17K) and the nonviral host 54K protein in all cell lines tested. A tumor antigen with an apparent molecular weight of 100,000 was also found in some, but not all, lines. Similar "super T" molecules have been found by others in many rodent transformed lines. We carried out an analysis of the relation of phenotype to relative amounts of these proteins in cell lines of the four classes, using the Spearman rank correlation test. The amount of the 100K T antigen relative to the 94K T antigen or to total viral protein was well correlated with the ability to form colonies in semisolid medium. No significant correlation was found between quantities of labeled 94K T antigen, 54K host antigen, or 17K t antigen and either serum or anchorage independence. Mouse cells transformed with the small t SV40 deletion mutant 884 synthesized a 100K T antigen, suggesting that small t is not required for the production of this protein. The 100K T antigen migrated more slowly than lytic T. Since mixtures of extracts from cells expressing and lacking the 100K T antigen yielded the expected amount of this protein, it is unlikely that the 100K T derives from the 94K protein by a posttranslational modification.     

Simian virus 40 T-antigen-related cell surface antigen: serological demonstration on simian virus 40-transformed monolayer cells in situ.

Simian virus 40 (SV40)-transformed monolayer cells were analyzed in situ by indirect immunofluorescence microscopy for the postulated cell surface location of SV40 T-antigen-related molecules. With antisera prepared against purified, sodium dodecyl sulfate-denatured SV40 T-antigen, positive surface staining was obtained when the cells had been treated with formaldehyde before immunofluorescence analysis. In contrast, living SV40-transformed cells analyzed in monolayer were surface fluorescence negative. The fixation procedure developed in this study combined with a double staining immunofluorescence technique allowed the simultaneous analysis of the same cells for the expression of both SV40 T-antigen-related surface antigen and nuclear T-antigen. The localization of SV40 T-antigen-related surface antigen on the outer surface of the plasma membrane of formaldehyde-fixed SV40-transformed cells was demonstrated directly by the protein A-mediated binding of Staphylococcus aureus bacteria on formaldehyde-fixed SV40-transformed cells precoated with antiserum against sodium dodecyl sulfate-denatured T-antigen. Both cell surface staining and S. aureus binding were found to be highly specific for SV40 T-antigen-related binding sites. These results indicate that T-antigen-related molecules in a cryptic form are located on the surface of SV40-transformed monolayer cells and can be detected in situ after modification of the cell surface architecture.     

Identification and initial characterization of a new low-molecular-weight virus-encoded T antigen in a line of simian virus 40-transformed cells.

SV80 cells, a simian virus 40 (SV40)-transformed derivative of a strain of human fibroblasts, synthesize an 8-kilodalton anti-T reactive polypeptide in addition to large T and small t antigens. Although not observed during lytic infection carried out under a variety of conditions, an anti-T reactive molecule which comigrated with the SV80 8-kilodalton protein during sodium dodecyl sulfate-polyacrylamide gel electrophoresis was synthesized by one of five other SV40-transformed cell lines studied. The SV40 8-kilodalton protein was present in lysates of cells exposed to a brief pulse of radioactive methionine and did not accumulate during an extended chase period. This polypeptide could not by generated by mixing an unlabeled extract of SV80 cells with a labeled extract of infected monkey cells. The 8-kilodalton molecule reacts with antibody raised against homogeneous large T antigen, is present only in the cytoplasm, is not complexed with T, lacks DNA-binding properties, and is not phosphorylated. This protein could be translated in a cell-free system programmed by SV40-specific mRNA. At least two messenger species (approximately 19S and approximately 22S) directed its synthesis. Tryptic peptide analysis of [35S]methionine-labeled proteins demonstrated that the 8-kilodalton protein contains all eight of the common T/t peptides and one additional peptide not present in the maps of t or T. It lacks both of the t-unique peptides. The organization of the integrated viral sequences which encode this molecule was determined by restriction endonuclease analysis. In particular, SV80 cells contain at least two integrated SV40 genomes which are oriented in tandem, with an intervening cellular sequence..     

T antigen banding on chromosomes of simian virus 40 infected muntjac cells.

Chromosomes were prepared from mitotic munjac cells 48 to 72 h after infection with SV40 virus. When stained for SV40 T antigen by indirect immunofluorescence, all chromosomes within an infected cell were fluorescent, indicating the presence of T antigen. Furthermore, the chromosomes were not uniformly stained but appeared to have regions of high and low fluorescence intensity. A variety of controls showed that the banding patterns are specific and highly reproducible and may indeed reflect the binding sites of T antigen. The bright, fluorescent bands T antigen were found to correspond to bands visualized by trypsin-Giesma staining (G-bands) and also by quinacrine staining (Q-bands). Current knowledge of chromosome banding indicates that Q-bands reflect the distribution of AT-rich regions along the chromosome. From the DNA sequence of SV40, it is known that one of the T antigen binding sites contains AT-rich sequences; thus, T antigen banding might be due to the base-specific binding of T antigen to chromatin. In addition, these bands have been implicated as centers for chromosome condensation and units in control of DNA replication. While the functional significance of T antigen binding has yet to be determined, the SV40-muntjac system provides an unusual opportunity to study the interaction of a known regulatory protein with mammalian chromosomes.     

Characterization of a simian virus 40-transformed human podocyte cell line producing type IV collagen and exhibiting polarized response to atrial natriuretic peptide.

Biology of glomerular visceral epithelial cells ("podocytes") and their role in inflammatory process remain obscure, partly because of the lack of well-differentiated podocyte cultures. We have established a human cell line by transfecting with a replication-defective SV40 plasmid (pSVHB1), a primary culture of podocytes derived from an enriched preparation of unencapsulated glomeruli free of tubule and Bowman's capsule contaminants. Podocyte specificity of the primary culture was assessed by a dual immunomorphological and functional approach. The resulting cell line (HGVEC.SV1) was cloned and the clonal cells were adapted to hormonally defined medium supplemented with only 2% newborn bovine serum. Clone A4 has been exhibiting over 35 passages, a combination of markers unique to podocytes, including expression of vimentin, podocalyxin, ectoenzymes (CALLA antigen and mRNA), heparan-sulfate proteoglycans (molecular mass of core protein = 75 kDa), and production of type IV collagen (alpha 1 and alpha 5 chains) established by immunoprecipitation and Northern blot analysis. Cytokeratin was detected in rare cellular foci and the search of Von Willebrand factor was negative. This clonal cell line has been used to demonstrate: (1) that human podocytes are highly sensitive to atrial natriuretic peptide (ANP) which induced a dose-dependent increase in cGMP production (x20 at 0.5 microM ANP), and (2) that secretion of ANP-stimulated cGMP is dramatically polarized as 93% of extracellular cGMP were released in the apical medium when filter-grown HGVEC. SV1A4 cells were stimulated at their basal pole.     

Presence of free viral DNA in simian virus 40-transformed nonproducer cells.

Extracts from several simian virus 40 (SV40)-transformed nonproducer cells were prepared by the hot-phenol procedure normally used to extract cellular RNA. These extracts contained SV40 infectious units. Part of the infectious units were identified as SV40 form I DNA molecules. The results of reconstruction experiments suggest that SV40 form I DNA is extractable by the hot-phenol procedure because of its fast renaturation rate. The significance of the presence of free viral DNA in nonproducer transformed cells is discussed.     

Hormonal regulation of the transformation phenotype in simian virus 40-transformed rat embryonic preadipose cell lines.

The regulation of transformed phenotypes was studied in newly isolated preadipose cell lines which were established after infection with simian virus 40 tsA58 dl2009. The clonal cell lines isolated exhibited most of the characteristics typical of transformed cells. The transformants, however, were able to differentiate into adipocytes in the presence of low calf serum (0.5%) and a combination of several hormones, including hydrocortisone and insulin. Treatment with insulin alone stimulated the growth of these cells but did not induce lipid accumulation without added hydrocortisone. The effect of hydrocortisone was accompanied by a restoration of growth control in the transformants after they reached high cell density. The blot hybridization analysis of cellular DNAs digested by restriction enzymes revealed that simian virus 40 genomes were integrated at multiple separate sites at which a head-to-tail oligomeric insertion took place. Large T antigen was synthesized in growing cells but was regulated at high cell density when cells were committed to differentiate by glucocorticoids. These results suggest that the glucocorticoid hydrocortisone is capable of restoring growth regulation at high cell densities to simian virus 40-transformed preadipose cell lines.     

Surface T-antigen expression in simian virus 40-transformed mouse cells: correlation with cell growth rate.

Cell growth control appears to be drastically altered as a consequence of transformation. Because the cell surface appears to have a role in modulating cell growth and simian virus 40 (SV40)-transformed cells express large T antigen (T-Ag) in the plasma membrane, we investigated whether surface T-Ag expression varies according to cell growth rate. Different growth states were obtained by various combinations of seeding density, serum concentration, and temperature, and cell cycle distributions were determined by flow microcytofluorometry. Actively dividing SV40-transformed mouse cell cultures were consistently found to express higher levels of surface T-Ag and T-Ag/p53 complex than cultures in which cells were mostly resting. In addition, the T-Ag/p53 complex disappeared from the surface of tsA7-transformed cells cultured under restrictive conditions known to induce complete growth arrest (39.5 degrees C), although the surface complex did not disappear from other tsA transformants able to keep cycling at 39.5 degrees C. These results suggest that surface SV40 T-Ag or surface T-Ag/p53 complex, or both, are involved in determining the growth characteristics of SV40-transformed cells.     

Reactivation of herpes simplex virus in a cell line inducible for simian virus 40 synthesis

The reactivation of UV-irradiated herpes simplex virus (HSV) was investigated in irradiated and unirradiated transformed hamster cells in which infectious simian virus 40 (SV40) can be induced. Reactivation was enhanced when the cells were treated with UV light or mitomycin C prior to infection with HSV. The IV dose-response curve of this enhanced reactivation was strikingly similar to that found for induction of SV40 virus synthesis in cells treated under identical condictions. This is the first time that two SOS functions described in bacteria have been demonstrated in a single mammalian cell line.     

Cell density-dependent increase in chromatin-associated ADP-ribosyltransferase activity in simian virus 40-transformed cells.

ADP-ribosyltransferase activity associated with chromatin is two- to tenfold higher in simian virus 40 (SV40)-transformed cells than in untransformed cells. When confluent transformed cells were subcultured, their specific enzyme activity first decreased two- to fourfold and the rapidly increased during the logarithmic phase of growth. This increase ceased or slowed down when the cells entered the stationary phase. In contrast, the activity in the untransformed cells remained low throughout the growth cycle. In SV40tsA-transformed cells (ts = temperature sensitive), this density-dependent increase in the enzyme activity was observed when the cells were cultivated at the permissive temperature, whereas the activity remained low at the restrictive temperature. The enzyme activity did not increase during induction of cellular DNA synthesis in quiescent cells either by addition of fresh medium or by infection with SV40. The chromatin-associated enzyme activity extracted with 1 m NaCl was eluted together with almost all the DNA-binding proteins from a phosphocellulose column with 0.6 m NaCl. The enzyme activity in this fraction from transformed cells, measured with or without added DNA and histones, was higher than that in a similar fraction from untransformed cells, reflecting the difference in the original activities present in the nuclei of these cells. The chain lengths of poly(ADP-ribose) formed by chromatin from SV40-transformed and untransformed cells were not significantly different. These results suggest that the number of initiation sites for ADP-ribosylation is increased in the chromatin of SV40-transformed cells compared to that of untransformed cells.     

Isolation of a simian virus 40 T-antigen-positive, transformation-resistant cell line by indirect selection.

In an attempt to identify cellular genes that might be involved in simian virus 40 (SV40) transformation, we have set out to isolate cells which express T antigen but are not transformed. SV40 DNA and the herpes simplex virus thymidine kinase gene were cotransfected into tk- 3T3 fibroblasts. Of 72 colonies screened that were resistant to hypoxanthine-aminopterin-thymidine, 57 were T antigen positive as judged by immunofluorescence. One of these lines, A27, had a normal growth phenotype in monolayer overgrowth and soft agar assays. It contained intact SV40 sequences that could be rescued by fusion to permissive cells. This rescued virus was fully capable of transforming nonpermissive cells to the same extent as did wild-type virus. The A27 cells, however, were not transformable by infection with SV40 or by transfection of SV40 DNA. It is likely that these cells were altered in a cellular function required for the establishment of the transformed state.     

Species-specific phosphorylation of mouse and rat p53 in simian virus 40-transformed cells.

We have analyzed in detail the phosphorylation of p53 from normal (3T3) and simian virus 40 (SV40)-transformed (SV3T3) BALB/c mouse cells and from normal (F111) and SV40-transformed [FR(wt648)] rat cells by two-dimensional tryptic peptide mapping and phosphoamino acid analyses. To accommodate the different half-lives of p53 in normal (half-life, 15 min) and transformed (half-life, 20 h) cells and possible differences in the rates of turnover of phosphate at specific sites, cells were labeled for 2 h (short-term labeling) or 18 h (long-term labeling). Depending on the labeling conditions, either close similarities or marked differences were observed in the phosphorylation patterns of p53 from normal and transformed cells. After the 2-h labeling, the phosphorylation patterns of p53 from normal and transformed mouse cells were quite similar. In contrast, p53 from normal and transformed rat cells exhibited dramatic quantitative and qualitative differences under these labeling conditions. The reverse was found after an 18-h label leading to steady-state phosphorylation of p53 in transformed cells: while p53 in transformed mouse cells revealed a marked quantitative increase in phosphorylation compared with p53 from normal cells, the corresponding patterns of p53 from normal and transformed rat cells were similar. Our data thus indicate species-specific differences in the phosphorylation of mouse and rat p53 in SV40-transformed cells, reflected by (i) different turnover rates at specific sites in mouse and rat p53 and (ii) phosphorylation of nonhomologous serine and threonine residues in rat p53, as revealed by indirect assignment of phosphorylation sites to the phosphopeptides of rat p53. Analyses of p53 from the SV40 tsA58 mutant-transformed F111 cell lines FR(tsA58)A (N type) and FR(tsA58)57 (A type) yielded no conclusive evidence for a direct correlation between phosphorylation of p53, the metabolic stabilization of p53, and expression of the transformed phenotype.