Inactivation of cellular p53 is a crucial step in carcinogenesis. Accordingly, p53 is inactivated in most human cancers by different mechanisms. In cells infected with DNA tumor viruses, p53 is bound to the viral tumor antigens (Tag). The current “dogma” views the Tag-p53 complexes as a way of sequestering and inactivating p53. Using primary human cells and SV40-transformed human cells, we show that in addition to inactivating p53 tumor suppressor activities, the Tag-p53 complex has growth stimulatory activities that are required for malignant cell growth. We found that in human cells, Tag-p53 complexes regulate transcription of the insulin-like growth factor I (IGF-I) gene by binding to the IGF-I promoter together with pRb and p300. Depletion of p53 leads to structural rearrangements of this multiprotein complex, resulting in IGF-I promoter transcriptional repression and growth arrest. Our data provide a novel mechanistic and biological interpretation of the p53-Tag complexes and of DNA tumor virus transformation in general. In the model we propose, p53 is not a passive inactive partner of Tag. Instead the p53-Tag complex promotes malignant cell growth through its ability to activate the IGF-I signaling pathway. [Cancer Res 2008;68(4):1022–9]
- SV40 Large T antigen
- DNA tumor viruses
- cell transformation
The p53 protein plays a critical role in carcinogenesis. Malignant transformation requires that p53 becomes inactivated to prevent p53-mediated apoptosis or cell growth arrest in cells that have genetic damage. Inactivation of p53 impairs DNA repair, thus favoring the early steps of carcinogenesis. The latter effect is mediated mostly by the ability of p53 to induce p21 expression, a cyclin-dependent kinase inhibitor that in turn causes cell cycle arrest, allowing DNA repair to take place. By inducing p21 expression, p53 prevents cells that have accumulated genetic damage from undergoing mitosis and propagating the damaged DNA to the descendants. Because of its critical role in regulating proper cell growth, “normal” wild-type p53 activity is undesirable for cancer cells. Accordingly, p53 must be inactivated to “create” human tumor cells in the laboratory ( 1). Moreover, the p53 pathway is found to be inactivated in most human tumors either by direct mutation or by alterations of p53 partners. DNA tumor viruses target p53 through the activities of their tumor antigens (the large T antigen, or Tag, of human polyomaviruses JCV and BKV, the SV40 Tag, the E1b of adenoviruses), which bind to and inactivate p53 tumor suppressor functions ( 2). Some viruses have developed additional strategies for inactivating p53: for example, the human papillomavirus 16 (HPV16) oncoprotein E6 binds to cellular p53, promoting its ubiquitylation and degradation ( 3, 4).
Unexpectedly, some studies have shown that p53 complexed to Tag can still bind DNA at p53 binding sites ( 8– 10). Moreover, Tag-bound p53 extracted from monkey and human cells was able to stimulate transcription of a p53-regulated promoter in cell-free extracts ( 9). SV40-mediated transformation of fibroblasts was enhanced by wild-type mouse p53 ( 11). Similarly, transformation of rat fibroblasts required both Tag and a metabolically stabilized p53 ( 12). Animal studies have shown that SV40 is more efficient in promoting tumor growth in the presence of wild-type p53 ( 13). These data did not fit with the generally accepted hypothesis that polyomavirus Tags bind to and inhibit cellular p53. Here we investigated the possible biological effects of Tag-p53 complexes on cell growth and transformation. We focused our studies on the insulin-like growth factor-I (IGF-I) pathway because this pathway has a critical role in regulating normal and malignant cell growth and because SV40-mediated transformation is dependent on the IGF-I signaling pathway ( 14– 16).
Materials and Methods
Plasmids, oligonucleotides, and antibodies are described in Supplementary data.
Cells and gene transfer procedures. Primary human mesothelial cell cultures were obtained from noncancerous donors, cultured, and characterized as described ( 17). These cells contain wild-type p53 ( 17). SV40-transformed human mesothelial cells (S-HML) were obtained through human mesothelial cell infection with SV40 virions (10 plaque-forming units/cell). Six to eight weeks after infection, tridimensional foci of transformed cells were handpicked and cultured into cell lines. The latter were grown in DMEM supplemented with 5% fetal bovine serum (FBS). In this study, we confirmed critical results in three independent S-HML. Primary human astrocytes were from Lonza and were cultured as recommended. SV40-transformed astrocytes were obtained and characterized after SV40 infection of primary astrocyte cultures essentially as previously described for S-HML ( 17).
Retroviral packaging was done using 293 human kidney cells following standard procedures. S-HML transduced with the retrovirus expressing the tetracycline regulator (TR; TET-ON system, Clontech Laboratories, Inc.) were selected with 600 μg/mL G418. After selection, the functionality of the system was assayed as recommended by the manufacturer. These cells were transduced with the retrovirus expressing HPV16 E6 (in the presence of doxycycline). Transduced cells were then selected with 600 μg/mL hygromycin and the resulting clone was named S-HML/E6. We also expressed E6 transiently through electroporation. Transfection was done with an electroporator (Gene Pulser II, Bio-Rad) using the following parameters: 300 kV and 975-μF capacity; 1 μg of plasmid DNA per 106 cells. Efficiency of transfection was >95%. The level of p53 down-regulation 48 h after transfection was very high and reproducible ( Fig. 1C ).
35S pulse-chase experiments. Western blot and immunostaining experiments were done as described ( 17). Pulse-chase experiments were done as follows: S-HML/E6 and control cells were grown in methionine-free medium containing 2 μg/mL doxycycline for 48 h. Cells were then pulsed with 500 μCi of [35S]methionine (Amersham) for 30 min, washed twice, and then cultured for 3 h in medium containing 100 μg/mL methionine. Cells were then handled as described, 3 with the exception that electrophoresis gels were transferred onto nitrocellulose membranes and analyzed by Western blot analysis to identify the band corresponding to p53. After Western blot, the membranes were exposed to both X-ray films and to a phosphoimager FLA-2000 (FujiFilm).
DNA synthesis analyses. We used the 5-bromo-2′-deoxyuridine (BrdUrd) flow kit (BD PharMingen) according to the manufacturer's instructions. Briefly, 1 × 106 cells were stimulated with IGF-I (5 nmol/L) in medium. A total of 100 μL of BrdUrd solution (1 mmol/L BrdUrd in PBS) were added to each 10-mL medium–containing dish and incubated for 8 h at 37°C. The cells were then fixed and permeabilized with Cytofix/Cytoperm buffer (BD PharMingen). The cellular DNA was digested with DNase for 1 h at 37°C. The cells were stained with an anti-BrdUrd FITC-conjugated antibody and analyzed by FACSCanto Flow Cytometer (BD Bioscience).
Apoptosis was assayed by annexin V/7-amino-actinomycin D staining according to standard procedures.
RNA studies. Nuclear run-on assays were done as described ( 18). Each slot contained 5 μg of alkali-denatured probes, which corresponded to a 133-bp PCR fragment of the human 18S rRNA gene; a 654-bp PCR fragment of the human p21 cDNA; a 378-bp PCR fragment of the human Bax cDNA; a 593-bp PCR fragment of the human IGF-I cDNA; and λ phage DNA digested with HindIII. After hybridization to 32P-labeled nuclear transcripts, membranes were washed and exposed to both X-ray film and a phosphoimager.
Real-time reverse transcription-PCR was done using standard protocols. Briefly, cells were dissociated with trypsin-EDTA, harvested, and snap frozen. Total cellular mRNA was obtained using the RNeasy kit (Qiagen) in the presence of RNase-free DNase I. The concentration of RNA in each sample was measured by using a spectrophotometer (GE Healthcare), and the quality of the mRNA was assayed in 1% formaldehyde agarose gels. Two micrograms of total RNA from each sample were reverse transcribed using a first-strand synthesis kit (MBI-Fermentas) in the presence of 10 pmol of random primers. Real-time PCR was done as follows: 1/5 of the reverse transcription reaction from each sample was diluted serially in H2O to determine the optimal range of dilution for the samples (CT between 15 and 25) using a Gene Amp 5700 (PE-Applied Biosystems). Oligo combinations (see Supplementary data) were chosen using the Primer Express 1.0 software (Applied Biosystems). Reactions were done with the SYBR Green PCR Master Mix (Applied Biosystems). After estimating the sample with higher levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA, 1:2 serial dilutions were made in H2O (range, 1–128) to construct a calibration curve for each mRNA. Similar calibration curves were run along with each experiment. “No reverse transcription” of each sample were the negative controls. mRNA values were normalized for GAPDH amounts and plotted as a percentage of the control sample.
Reporter assays were done using the Dual-Glow Luciferase Assay system (Promega) and measured with a luminometer (Veritas, Turner Biosystems).
Tag-p53 in vitro binding to the rat IGF-I promoter. Plasmid pSmaBgl-LUC was cut with SmaI and FspI, giving rise to three bands, which were extracted from gel with the QIAquick gel extraction kit (Qiagen). S-HML were lysed in buffer [20 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 0.5% NP40, 0.1% SDS, 1 μg/mL of aprotinin, leupeptin, pepstatin, and 1 mmol/L phenylmethylsulfonyl fluoride] on ice for 30 min. Lysates were then centrifuged and the supernatants were collected. Total cell lysates (50 μg) were incubated with 100 ng of top, middle, and bottom bands (Supplementary Fig. S7) supplemented with 0.5 μg of λ DNA digested with HindIII. Reactions were incubated at room temperature for 15 min, then loaded onto 1% agarose gels poured in 0.25× Tris-borate EDTA buffer. Electrophoresis was done for 1 h at 4°C (3 V/cm). The gel was then electroblotted onto Hybond-C membranes (Amersham), with membranes facing both electrodes (therefore the agarose gels were sandwiched between two sheets of Hybond C). Western blot analysis was done following standard procedures and visualized by enhanced chemiluminescence.
Chromatin immunoprecipitation assays. Cells (1 × 107) were cross-linked for 10 min at room temperature using 1% formaldehyde in PBS. The reaction was stopped by adding 0.250 mol/L glycine in PBS. The cells were harvested and the DNA was sheared by sonication in lysis buffer [20 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 1% NP40] supplemented with 1% SDS, using a Branson Sonifier 250 (Wolf Laboratories Ltd.) to generate DNA fragments ≥600 bp. Cell lysates were diluted 10 times with lysis buffer to reduce the SDS final concentration to 0.1%. Lysates were precleared with protein A/G agarose beads (Pierce) for 4 h at 4°C. Cell lysates were then incubated overnight at 4°C with the respective primary antibodies. Immune complexes were collected on Protein A/G agarose beads overnight at 4°C. Samples were dialyzed twice against dialysis buffer [2 mmol/L EDTA (pH 8.0), 50 mmol/L Tris-HCl (pH 8.0)]. After four washes in 1-mL wash buffer [100 mmol/L Tris-HCl (pH 8.0), 500 mmol/L LiCl, 1% NP40], immune complexes were eluted from the beads by vigorously shaking twice in 150 μL of 50 mmol/L NaHCO3, 1% SDS. Cross-linking was reversed by incubation in 0.15 mol/L NaCl overnight at 65°C. Each sample DNA was ethanol precipitated and purified with QIAprep spin miniprep kit (Qiagen). DNA was resuspended in 100 μL of Tris-EDTA buffer. Finally, each sample was PCR amplified using appropriate oligonucleotides as primers (see Supplementary data) and the results were visualized on 2% agarose gels.
S-HML/E6 cells express functionally active recombinant HPV16 E6. We wanted to study the effects of lowering p53 in cells expressing the SV40 Tag. To achieve this goal, we expressed HPV16 E6 in S-HML. We used a stable, tetracycline-inducible transduced cell clone expressing a fusion protein consisting of six histidines at its NH2-terminal portion (for conjugation to Ni-charged carriers), an Xpress peptide flag, and the full-length HPV16 E6 (see Supplementary data). These cells (named S-HML/E6) express the SV40 Tag and, on doxycycline treatment, also express E6 that bound and degraded p53 ( Fig. 1; Supplementary Fig. S1).
Decreased amounts of p53 in S-HML/E6 correspond to decreased expression of proteins transcriptionally regulated by p53 and cause cell growth arrest. We measured p53 and Tag expression levels at different time points after doxycycline induction in S-HML/E6 cells. Forty-eight hours after doxycycline-mediated induction of HPV16 E6, S-HML/E6 had reduced expression of p53 and p21 and mdm2 (proteins regulated by p53; reviewed in ref. 19); Tag expression was not influenced ( Fig. 1A and B). We hypothesized that these effects could have caused apoptosis/mitotic catastrophe or a proliferative advantage. Instead, annexin V/propidium iodide staining followed by fluorescence-activated cell sorting (FACS) analysis did not show evidence of apoptosis or the appearance of aberrant DNA peaks (an indication of mitotic catastrophe), and S-HML/E6 showed a doxycycline dose–dependent reduction in DNA synthesis compared with controls ( Fig. 2A ). Because doxycycline has pleiotropic effects that might have influenced these findings, we expressed recombinant E6 protein in transiently transfected S-HML. We obtained identical results: E6-transfected S-HML had undetectable p53 and were growth arrested compared with controls ( Figs. 1C and 2B). The reciprocal experiments (growth curves for the inducible system and DNA incorporation assay for S-HML transiently transfected with E6) gave identical results (Supplementary Fig. S2). To study why depletion of cellular p53 in S-HML resulted in growth arrest, we investigated genes that are transactivated by Tag ( 20, 21). We found that doxycycline treatment of S-HML/E6 abolished IGF-I precursor expression and the IGF-I receptor (IGF-IR; Fig. 2C). The effects observed on E6 transfection seemed to be dependent on the presence of Tag because E6 transfection of primary human mesothelial cells (which do not contain SV40) caused the opposite effect: a 4.3-fold increase of IGF-I expression ( Fig. 2D, top). Because E6 did not influence Tag expression but influenced p53 expression in S-HML ( Fig. 1A), we speculated that the E6 activities in Tag-containing cells were mediated through the degradation of p53. To test our hypothesis that the decreased expression of p21, mdm2, IGF-I, and IGF-IR on E6 induction was related to p53 down-regulation, we deregulated p53 in S-HML using a short hairpin RNA against p53, a dominant negative p53 (p53mt135, which interferes with proper p53 complex formation; Supplementary Fig. S3A and B, respectively), or by overexpressing mdm2 in S-HML ( Fig. 2D, bottom); all these approaches yielded reproducible p21, IGF-I, and IGF-IR down-regulation. The most effective and reproducible way to down-regulate p53 expression in S-HML was expressing HPV16 E6 in these cells ( Fig. 1C). These results together suggested that the decreased expression of p21, IGF-I, and IGF-IR was related to p53 depletion independently of how that was achieved.
E6-mediated p53 down-regulation causes S-HML cell growth arrest through the IGF-I/IGF-IR signaling pathway. To confirm that p53 depletion in S-HML causes cell growth arrest through IGF-I/IGF-IR signaling, we designed the experiment summarized in Fig. 3A . S-HML were transfected with a plasmid expressing the IGF-IR under the control of a cytomegalovirus (CMV) promoter (1R cells). S-HML transfected with the empty plasmid served as control (1C cells). Both 1R and 1C cells were transfected in parallel either with the E6-expressing plasmid (yielding either 1R6 or 1C6 cells) or with the control plasmid for E6 (yielding either 1Rc or 1Cc cells). Twenty-four hours after transient transfection, cells were cultured in 1% FBS or 1% FBS supplemented with 5 nmol/L purified IGF-I. Forty-eight hours after transient transfection, all cells were analyzed for DNA synthesis by BrdUrd incorporation assay/FACS analyses ( Fig. 3B–D). Both 1Rc and 1Cc (these cells do not express E6 and have normal p53 amounts) were able to resume DNA synthesis after IGF-I treatment ( Fig. 3B). Instead, 1C6 cells (which have down-regulated p53 and the IGF-IR is under the control of its own promoter) could not resume DNA synthesis ( Fig. 3C). However, 1R6 cells (which have down-regulated p53 but an IGF-IR under the control of a CMV promoter; Supplementary Fig. S4C) resumed DNA synthesis on IGF-I stimulation ( Fig. 3D). These data confirmed that E6-mediated down-regulation of p53 in S-HML impaired DNA synthesis and that the growth impairment was mediated through effects of the Tag-p53 complexes on the IGF-I signaling pathway. This interpretation was supported by the finding that E6 expression in primary mesothelial cells that do not contain Tag did not affect the rate of DNA synthesis in these cells (Supplementary Fig. S5). We found that down-regulation of p53 with short hairpin RNA in SV40 transformed primary human astrocytes caused down-regulation of the IGF-I precursor and of the IGF-IR (Supplementary Fig. S6), suggesting that the data we observed in human mesothelial cells are of general relevance to human cells.
E6-mediated down-regulation of p53 causes transcriptional repression of p53-regulated promoters and of the IGF-I promoter in S-HML. We hypothesized that the decreased expression levels of p21, mdm2, and IGF-I protein detected after E6-mediated p53 depletion in S-HML could have been mediated at the transcriptional level. Treatment of S-HML with IGF-I caused increased expression of the IGF-IR, whereas a small interfering RNA directed against IGF-I caused decreased IGF-IR expression in S-HML (Supplementary Fig. S7). Therefore, we concluded that the study of the IGF-I promoter regulation would have provided major insights into the positive IGF-I/IGF-IR autocrine feedback loop in S-HML. We used real-time PCR to study the expression of IGF-I and of genes under p53 regulation. We found that E6-mediated p53 down-regulation in S-HML almost abolished p21 and IGF-I mRNA expression and caused 50% and 80% reductions in the expression of Bax (a protein also regulated by p53) and mdm2 mRNAs, respectively ( Fig. 4A ). Nuclear run-on experiments confirmed that the transcriptional activity of these promoters is suppressed in S-HML expressing E6 ( Fig. 4B). To further verify that E6 induction in S-HML caused repressed transcription at the IGF-I promoter, we carried out reporter assays using a plasmid kindly provided by Dr. R. Baserga (Thomas Jefferson University, Philadelphia, PA), in which firefly luciferase was under the control of the rat IGF-I promoter (ref. 21; see Supplementary Fig. S8 for a map of this plasmid). Induction of E6 in S-HML/E6 caused a 70% reduction of firefly luciferase activity compared with controls ( Fig. 4C). We hypothesized that either Tag or p53 (or both) could have directly bound the IGF-I promoter. We first tested this hypothesis using the rat IGF-I promoter because this promoter has previously been characterized ( 21) and the results supported our hypothesis (Supplementary Fig. S9A). The results in the rat model suggested that Tag and cellular p53 could have regulated the human IGF-I promoter in S-HML through direct/indirect association with the promoter.
Tag and cellular p53 associate with the IGF-I promoter within a multiprotein complex in vivo. To test whether Tag and p53 were associated with the cellular IGF-I promoter in S-HML, we resorted to chromatin immunoprecipitation assays. We aligned the sequence of the rat promoter with the genomic sequence of the human IGF-I region. The fragment in the human IGF-I promoter that showed the highest homology with the rat promoter (72% nucleotide identity) was the one spanning positions −1,952 to −322 of the human IGF-I promoter. This promoter region was analyzed by chromatin immunoprecipitation assay dividing the 1,630-bp DNA promoter region in seven partially overlapping amplicons ( Fig. 5A ). As a negative control, we used a 266-bp region on chromosome 12q12 (see Supplementary data for a detailed description of this amplicon). We compared nontransfected S-HML with p53-depleted S-HML via E6 transfection. Cells were cross-linked, then cell lysates were mechanically sheared and immunoprecipitated with antibodies specific for either Tag or p53 ( Fig. 5B, lanes 3 and 4, respectively). The cross-linking in immunoprecipitated materials was reversed and PCR amplified with primers that yielded the amplicons outlined in Fig. 5A. In nontransfected S-HML (cells with an active IGF-I promoter), Tag was associated with the “B” region, and p53 was associated with both the “A” and “B” regions ( Fig. 5B). Instead, in p53-depleted S-HML, p53 was associated only with the “B” region ( Fig. 5C, lanes 7 and 8) and Tag was associated prevalently with the “A” region ( Fig. 5C, top, compare lanes 5 and 6), although the signal corresponding to the “B” region persisted ( Fig. 5C, lanes 5 and 6, bottom). This indicated that on E6-mediated p53 down-regulation, the Tag-p53 complex on the IGF-I promoter underwent a conformational change, with Tag moving upstream (i.e., to the A region) from the IGF-I starting codon and p53 losing occupancy of the same region. To further investigate the Tag-p53 complex on the IGF-I promoter, we tested the “A” through “G” region of the IGF-I promoter by chromatin immunoprecipitation for two major binding partners of Tag: pRb and p300 (reviewed in ref. 2). When we immunoprecipitated S-HML extracts with a pRb-specific antibody, we PCR amplified region “B” ( Fig. 5D, lane 3). When we immunoprecipitated with a p300-specific antibody, we amplified both “A” and “B” regions ( Fig. 5B, lane 8). No other region of the portion of the IGF-I promoter was amplified. These data indicated that within the 538 bp encompassing the “A” and “B” regions of the IGF-I promoter, there is a multiprotein complex that includes Tag, p53, pRb, and p300. When we decreased p53 expression through E6 induction, we were still able to amplify region “B” by immunoprecipitating with the pRb-specific antibody ( Fig. 5D, lane 7); however, immunoprecipitations conducted with the p300 antibody failed to yield both the “A” (not shown) and “B” amplicons ( Fig. 5D, lane 8). This indicated that on p53 depletion, the multiprotein complex at regions −1,952 to −1,414 of the IGF-I promoter underwent modifications that included structural rearrangements of individual complex partners and exit of p300, a transcriptional coactivator ( 22). The result of these modifications was paralleled by transcriptional inhibition at the IGF-I promoter. We detected no quantitative differences in the amount of pRb bound to the “B” amplicon before and after E6-mediated p53 depletion in S-HML by quantitative PCR (Supplementary Fig. S9B). No E6 association with the IGF-I promoter was detected. Furthermore, no association of either Tag or p53 was detected on the IGF-IR promoter by chromatin immunoprecipitation analysis (not shown).
More than 30,000 articles examining the role and function of p53 have appeared in the literature since its discovery in 1979. This extraordinary amount of work on a single gene reflects the central role that p53 has in governing normal cell growth and, consequently, in determining the outcome of genetic insults that may lead to malignant cell transformation and tumor cell growth. In spite of the enormous amount of scientific literature on p53, research on this protein continues to turn up new surprises.
One of the current “dogmas” about p53 is that its transcriptional and biological activities are impaired by its binding to the Tags of DNA tumor viruses. A similar dogma in the DNA tumor virus field assumes that the oncogenes of these viruses bind to and “inactivate” cellular p53, a process that is required for viral replication and also for virus-mediated cellular transformation.
Here, we present evidence indicating that this current dogma is only partially correct because newly formed Tag-p53 complexes acquire new transcriptional and biological activities. We found that this complex binds to the IGF-I promoter, stimulating IGF-I expression and the IGF-I signaling pathway, an effect that leads to cell growth. Specifically, we found that the Tag-p53 complexes interact with the IGF-I promoter as part of a complex that consists of several partners, including pRb and p300. When we depleted p53, we observed a loss of p300 on the IGF-I promoter.
Our results provide a rationale to a number of studies that had found that the presence of wild-type p53 was required to stimulate transcription and for polyomaviruses-mediated malignant cell transformation ( 8– 13).
These findings support recent studies showing that Tag requires p53 to interact with p300, and RNAi-mediated p53 depletion disrupts Tag-p300 interactions ( 23). Therefore, we propose the following model ( Fig. 6 ). In normal S-HML, a multiprotein complex that includes Tag, p53, pRb, and p300 occupies positions −1,952 to −1,414 (approximately) of the IGF-I promoter. On p53 depletion (obtained using different techniques), this complex undergoes structural rearrangements, probably as a result of the exit of some critical components (such as p300) that ultimately regulate the transcription of the IGF-I gene.
The biological effects observed on SV40 infection are species specific ( 24) because rodent cells are nonpermissive to SV40 replication whereas human cells are. Therefore, animal models could not be used to test in vivo the possible oncogenic effects of the findings reported here.
In our experimental model, we used mesothelial cells and SV40. Specifically, we used different primary (normal) human mesothelial cells obtained from separate donors with nonmalignant pleural effusions ( 17). We used human mesothelial cells because these cells allow SV40 replication and are rarely lysed by the virus; thus, human mesothelial cells are ideal to study the biological interactions between Tag and p53 ( 17). Human mesothelial cells are also frequently transformed by SV40 ( 17, 24), an effect that allowed us to compare the Tag-p53 biological activities in both primary and malignant human cells. However, the results presented here seem to be of general relevance because we observed similar effects in SV40-transformed primary human astrocytes. It should be noted that p53 depletion caused decreased IGF-IR expression in both S-HML and SV40-transformed human astrocytes. Although we did not detect a direct association of either p53 or Tag on the IGF-IR promoter, Tag-p53 complexes may indirectly regulate IGF-IR expression. Alternatively, the IGF-IR may be mainly under a positive feedback loop regulated by autocrine IGF-I. The results presented in Supplementary Fig. S7 seem to support such interpretation.
In summary, our data provide a novel mechanistic and biological interpretation of the p53-Tag complexes and possibly of DNA tumor virus transformation in general. In the model, we propose that the p53-Tag complex in human cells promotes malignant cell growth through its ability to activate the IGF-I signaling pathway. The results presented here suggest that there is a “threshold” effect for p53 when complexed to Tag, and that a correct stoichiometry between these two components influences the biological effects of Tag-p53 complexes.
Grant support: National Cancer Institute grant R21-CA91122 and American Cancer Society grant RSG-05-077 (M. Bocchetta) and grants RO1CA 092657-01 and PO1CA114047-01A1 (M. Carbone).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Drs. Antonio Pannuti, Haning Yang, John Jenkins, Premkumar Reddy, and James Pipas for critical reading of the manuscript; Drs. Renato Baserga and Martin W. Kast for providing critical reagents; and Patricia Simms for her invaluable help with FACS analysis.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
- Received September 6, 2007.
- Revision received November 29, 2007.
- Accepted December 5, 2007.
- ©2008 American Association for Cancer Research.