Middle T Antigen Activation of Signal Transduction Pathways Does Not Overcome p53-Mediated Growth Arrest

Abstract

Polyomavirus middle T antigen does not overcome p53-mediated G₁ arrest in mouse embryo fibroblasts. Middle T antigen still associates with the signaling molecules phosphatidylinositol 3-kinase and SHC and activates the transcriptional activity of c-Myc and AP1 in p53-arrested cells. Examination of cell cycle regulatory proteins indicated that p53 does not interfere with these mitogenic signals but acts later in the G₁ phase of the cell cycle.

Middle T antigen (MT) is the major oncoprotein of polyomavirus (36). It is necessary and in many cases sufficient for transformation of established cells and for induction of tumors in mice (12, 28, 29, 37). MT interacts with several cellular proteins including phosphatidylinositol 3-kinase (PI3K) (34, 39), SHC (3, 5), and phospholipase C-γ (33). These associations activate distinct, though interactive, signaling cascades and result in induction of transcription of the early-response genes jun, fos, and myc (13, 27, 30, 35, 41). Overall, MT activates multiple mitogenic signals, and it is through the constitutive activation of these signals that MT transforms cells (2, 17).

The p53 tumor suppressor protein plays a critical role in suppressing cell proliferation and neoplastic transformation (20). Inactivation of the p53 gene frequently occurs during the development of mouse and human neoplasia (8, 16). Wild-type p53 is thought to suppress oncogene-mediated transformation through its ability to induce growth arrest and/or activate apoptosis (1). p53 arrests the growth of cells in the G₁ phase of the cell cycle by inducing the expression of p21/WAF1, which inhibits G₁ cdk activity and prevents the phosphorylation of RB proteins (4, 9, 14, 40).

In a previous study, we demonstrated that MT expression does not overcome p53-mediated growth arrest (7). Cell lines for that study were generated by transfecting a temperature-sensitive p53 gene (10, 23) into mouse embryo fibroblasts derived from a p53-null animal (15) (M/tsp53). Other laboratories have characterized the temperature-sensitive p53 gene (pLTRp53cGval135) and have shown that at 37°C the protein is expressed in a mutant conformation and that at 32°C it is expressed in a functionally wild-type conformation that arrests cell growth in the G₁/G₀ phase (22, 23). When MT was introduced into M/tsp53 cells (M/tsp53/MT) and cultured at the permissive temperature of 32°C, M/tsp53/MT cells did not proliferate. Cell cycle analysis of M/tsp53/MT cells cultured at 32°C indicated that approximately 70% of the cells were arrested in G₁/G₀ phase and only 10% were in the S phase of the cell cycle (7). Additionally, there was no indication that expression of MT in the p53-arrested cells induced apoptosis. Sub-G₁/G₀ DNA fragments were not detected by fluorescence-activated cell sorting analysis, nor were floating, dead cells observed in the cultures incubated at 32°C (7). These results indicated that expression of MT does not override the p53-mediated growth arrest or induce an apoptotic response.

In this report, we examined the mitogenic signals activated by MT in these cells to assess whether p53 was interfering directly with MT function. To determine if p53 interferes with the initial events of MT signaling, we examined whether MT associated with pp60^c-src and was phosphorylated on tyrosine residues in p53 growth-arrested cells. Cell extracts were prepared from M/puro (expressing only drug resistance), M/tsp53, and M/tsp53/MT cells after culture at 37 or 32°C for 24 h. Protein kinase assays were carried out by incubation of immunoprecipitated MT with [γ-³²P]ATP. The top panel of Fig. 1A shows proteins in the immune complex that were phosphorylated. MT-associated kinase activity was not detected in cells lacking MT (M/puro and M/tsp53). In extracts from M/tsp53/MT cells cultured at either 37 or 32°C, proteins of 55, 70, and 85 kDa were phosphorylated. The molecular masses of 55 and 85 kDa correspond to those of MT and the regulatory subunit of PI3K. To further substantiate that MT is associated with tyrosine kinase activity and becomes phosphorylated, blots of total-cell lysates were probed with antibodies to MT and antiphosphotyrosine. We found that MT comigrates with a phosphotyrosine-containing protein in cells cultured at either temperature (bottom panel of Fig. 1A). These results suggest that MT associates with and activates pp60^c-src tyrosine kinase in cells expressing functionally wild-type p53. In addition, MT is phosphorylated on tyrosine in cells expressing both wild-type and mutant p53.

FIG. 1.

MT associates with a tyrosine kinase and interacts with PI3K and SHC in p53 growth-arrested cells. (A) Extracts were prepared from M/puro, M/tsp53, and M/tsp53/MT cells cultured at 37 and 32°C. MT was immunoprecipitated and incubated with [γ-³²P]ATP to detect associated kinase activity. The autoradiographs displayed in the upper panel show the MT-associated kinase from designated cells cultured at indicated temperatures (°C). The lower panel shows Western blots of total-cell lysates probed with antibodies to MT and antibodies to phosphotyrosine (Y-P). (B) Extracts were prepared from M/tsp53 and M/tsp53/MT cells cultured at 37 or 32°C. Immunoblots of MT immune complexes (IP MT) were probed with antibodies to PI3K or SHC (left panels). As a control, Western blots of total-cell lysate were probed to detect cellular levels of PI3K and SHC at each temperature (°C) (right panels). Numbers at right of each panel indicate molecular masses in kilodaltons.

To determine whether MT associates with the signaling molecules PI3K and SHC through its tyrosine-phosphorylated residues, we immunoprecipitated MT from cells cultured at 37 and 32°C and analyzed the associated proteins by Western blotting. Western blots were probed with antibodies to SHC and the 85-kDa subunit of PI3K (Fig. 1B). Approximately equivalent amounts of PI3K and SHC coprecipitated with MT from cells cultured at both temperatures (Fig. 1B, left). Similar amounts of PI3K and SHC are present in cells expressing mutant or wild-type p53 with and without MT (Fig. 1B, right). These results indicate that the association of signaling molecules with MT is not disrupted or diminished by expression of functionally wild-type p53.

The activities of several transcription factors are stimulated by the signaling pathways activated by MT (25, 27, 30, 32, 38, 41). To determine if the MT-activated transcription still occurs in the presence of wild-type p53, we measured the activities of several transcription factors in the p53-arrested cells. Luciferase reporter plasmids that contain multiple Myc, TPA response element (TRE), or cyclic AMP response element (CRE) sites upstream of a minimal promoter were transiently transfected into each cell line. AP1 transcription factors, composed of Jun-Fos heterodimers or jun homodimers, bind to the TRE site, while CREB/ATF transcription factors bind to CRE sites (18). Although TRE and CRE sites are similar, TRE-containing promoters are activated by tyrosine kinase pathways and CRE-containing promoters respond to activation of the protein kinase A pathway (18). The vector with only the minimal promoter and luciferase gene was also transfected into cells to determine the activity of the promoter without the addition of transcription factor binding sites. In addition, a plasmid with the β-actin promoter driving β-galactosidase was cotransfected to control for transfection efficiency.

Reporter plasmids were transfected into M/tsp53 and M/tsp53/MT cells that had been cultured in 10% serum at 37°C. After transfection, the cells were cultured in low serum (0.5%) at 37°C overnight. Wild-type p53 was induced and cell growth was arrested by incubation at 32°C for 48 h before harvesting the cell extract. Figure 2 represents the luciferase activity in cells expressing MT (M/tsp53/MT) compared to that in cells not expressing MT (M/tsp53). The average of four independent transfections with duplicate samples in each transfection is shown. The reporter plasmids containing the c-Myc and AP1 binding sites were approximately sixfold more active in p53-mediated growth-arrested cells expressing MT than in cells without MT. Luciferase expression from the minimal promoter (vector) and from the plasmid containing ATF binding sites was twice as high in MT-expressing cells. These results indicate that MT activates c-Myc and AP1 transcription factors even in wild-type p53-mediated growth-arrested cells. The increase in transcriptional activity detected with the minimal promoter and the plasmid containing the ATF binding sites suggests that MT may also increase the overall activity of the basal transcription machinery. Furthermore, these results indicate that p53 does not suppress MT activation of the signal transduction pathways, which leads to the stimulation of these transcription factors, or suppress the transcriptional activity of these factors.

FIG. 2.

Transcriptional activity driven by c-Myc and AP1 is greater in p53 growth-arrested cells expressing MT. M/tsp53 and M/tsp53/MT cells were transfected with luciferase reporter vectors containing multiple ATF, c-myc, or AP1 binding sites upstream of a minimal promoter. The luciferase reporter plasmid with the minimal promoter was also transfected (Vector). A reporter plasmid containing the rat β-actin promoter driving expression of β-galactosidase was cotransfected and used as a control for transfection efficiency. Following transfection, cells were cultured in Dulbecco modified Eagle medium plus 0.5% serum at 37°C for 24 h and then at 32°C for 48 h. Cell extracts were harvested, and luciferase and β-galactosidase activities were measured. The graph compares luciferase activity in M/tsp53/MT cells to that in M/tsp53 cells. Each transfection was done in duplicate. The data represent the averages of four independent experiments. The error bars represent standard deviations between experiments.

Overexpression of wild-type p53 blocks cell cycle progression in the G₁ phase of the cell cycle (22, 23). We examined the level of cell cycle regulatory proteins expressed in p53 growth-arrested cells and after release from this growth arrest in order to estimate the point in G₁/G₀ where wild-type p53 induces the block in cell cycle progression (Fig. 3). Both M/tsp53 and M/tsp53/MT cells were cultured at 32°C for 24 h to arrest growth and then returned to 37°C to release the cells from the G₁ block. Cell extracts were collected at various times after release. Both cultures of p53-arrested cells (time zero) expressed high levels of p21/WAF1 and cyclin D and underphosphorylated pRB protein. In both M/tsp53 and M/tsp53/MT cells, the amount of p21/WAF1 was reduced after 4 h at 37°C and was undetectable by 8 h. Coincident with the decrease in p21, pRB became phosphorylated by 4 h in M/tsp53/MT cells and by 8 h in M/tsp53 cells. By 12 h after release from growth arrest, cyclin D levels had decreased and the cells had entered S phase (data not shown). These results indicate that p53 arrests cells after expression of cyclin D, a point late in the G₁ phase of the cell cycle (31). Furthermore, expression of MT does not alter the levels of cyclin D or p21 protein or result in phosphorylation of pRB in cells whose growth is arrested by p53.

FIG. 3.

p53 arrests cells in late G₁ phase, after expression of cyclin D. M/tsp53 and M/tsp53/MT cells were grown at 32°C for 24 h to arrest growth (0), followed by culture at 37°C for the times (hours) indicated above the panels. Cell extract was made, and protein levels were analyzed by Western blotting. Blots were probed with antibodies to cyclin D, p21/WAF1, and pRB.

We conclude from this study that the mitogenic pathways activated by middle T antigen are insufficient to overcome the growth-suppressive function of wild-type p53. The data suggest that p53 does not act directly on any of the signal transduction pathways activated by middle T antigen but rather acts later in the G₁ phase of the cell cycle and can block the MT-activated mitogenic signals. These signal transduction pathways are similar to those activated by serum factors and tyrosine kinase receptors (17). Constitutive activation of these receptors overcomes a major control point which appears to operate early in G₁ and which allows cells to exit the cell cycle and enter a nonproliferative state, G₀ (26). Wild-type p53 appears to act later in G₁ at a second control point immediately prior to RB phosphorylation, referred to as the restriction or R point. Previous studies have suggested that growth suppression by wild-type p53 occurs near the R point in late G₁ (21). Our results further support this conclusion and suggest that stimulation of early events in G₁ is unable to overcome p53-mediated growth arrest. Thus, we speculate that the p53/RB-mediated G₁ checkpoint is dominant over oncogene-activated G₀ to G₁ mitogenic pathways.

Transforming viruses have evolved common strategies to stimulate host cell proliferation; simian virus 40, adenovirus, and human papillomavirus inactivate the tumor suppressor genes p53 and RB (24), while bovine papillomavirus and Epstein-Barr virus activate cellular growth factor signaling pathways (6). Polyomavirus is unique in that it combines two different strategies to overcome G₁ checkpoints. MT activates mitogenic signaling pathways, and large T antigen inactivates tumor suppressor proteins, and the cooperation of both proteins is required to transform primary cells (19, 29) and induce a full tumor profile in mice (11, 12).