Intestinal Hyperplasia Induced by Simian Virus 40 ... - Journal of Virology

JOURNAL OF VIROLOGY, Dec. 2007, p. 13191–13199 0022-538X/07/$08.00⫹0 doi:10.1128/JVI.01658-07 Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Vol. 81, No. 23

Intestinal Hyperplasia Induced by Simian Virus 40 Large Tumor Antigen Requires E2F2䌤 M. Teresa Saénz-Robles,1† Jennifer A. Markovics,1†‡ Jean-Leon Chong,2† Rene Opavsky,2 Robert H. Whitehead,3 Gustavo Leone,2 and James M. Pipas1* Department of Biological Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania 152601; Department of Molecular Genetics, Comprehensive Cancer Center, The Ohio State University, Columbus, Ohio 432102; and Cellular and Animal Modeling Core, Vanderbilt Digestive Research Center, Vanderbilt University, Nashville, Tennessee 372323 Received 30 July 2007/Accepted 5 September 2007

The simian virus 40 large T antigen contributes to neoplastic transformation, in part, by targeting the Rb family of tumor suppressors. There are three known Rb proteins, pRb, p130, and p107, all of which block the cell cycle by preventing the transcription of genes regulated by the E2F family of transcription factors. T antigen interacts directly with Rb proteins and disrupts Rb-E2F complexes both in vitro and in cultured cells. Consequently, T antigen is thought to inhibit transcriptional repression by the Rb family proteins by disrupting their interaction with E2F proteins, thus allowing E2F-dependent transcription and the expression of cellular genes needed for entry into S phase. This model predicts that active E2F-dependent transcription is required for T-antigen-induced transformation. To test this hypothesis, we have examined the status of Rb-E2F complexes in murine enterocytes. Previous studies have shown that T antigen drives enterocytes into S phase, resulting in intestinal hyperplasia, and that the induction of enterocyte proliferation requires T-antigen binding to Rb proteins. In this paper, we show that normal growth-arrested enterocytes contain p130–E2F4 complexes and that T-antigen expression destroys these complexes, most likely by stimulating p130 degradation. Furthermore, unlike their normal counterparts, enterocytes expressing T antigen contain abundant levels of E2F2 and E2F3a. Concomitantly, T-antigen-induced intestinal proliferation is reduced in mice lacking either E2F2 alone or both E2F2 and E2F3a, but not in mice lacking E2F1. These studies support a model in which T antigen eliminates Rb-E2F repressive complexes so that specific activator E2Fs can drive S-phase entry. Simian virus 40 (SV40) is among the best-characterized DNA tumor viruses and has been employed widely to probe mechanisms of cellular growth control (1, 10, 31). The oncogenic potential of SV40 is harbored by two virus-encoded proteins, the 708-amino-acid large T antigen and the 174-aminoacid small t antigen. The large T antigen is necessary and often sufficient to induce cellular transformation, while the small t antigen contributes to transformation in some cell types or under certain assay conditions. In some cases, oncogenic signals in addition to those provided by the large and small T antigens are needed for full transformation (13). The small t antigen’s contribution to transformation is effected through its association with the cellular phosphatase pp2A (26, 29, 38). Much of the large T antigen’s transforming activity is explained by its direct interaction with the retinoblastoma (Rb) protein family of tumor suppressors and with the tumor suppressor p53. The inactivation of Rb proteins by T antigen is thought to drive quiescent cells into S phase, while T antigen’s action on p53 is thought to prevent apoptosis. Amino-terminal truncation mutants of the large T antigen that retain the ability to bind and inactivate the Rb proteins induce cell proliferation and neoplasia when expressed in many established cell lines or

* Corresponding author. Mailing address: Department of Biological Sciences, University of Pittsburgh, Pittsburgh, PA 15260. Phone: (412) 624-4691. Fax: (412) 624-4759. E-mail: [email protected]. † M.T.S.-R., J.A.M., and J.-L.C. contributed equally to this work. ‡ Present address: Department of Pathology, SFGH, University of California—San Francisco, San Francisco, CA 94110. 䌤 Published ahead of print on 12 September 2007.

in multiple tissues of transgenic mice (5, 16, 17, 25, 30, 34, 37, 40). The Rb family of tumor suppressors consists of three proteins, pRb, p107, and p130, which are major regulators of the G1/S checkpoint of the cell cycle (reviewed in reference 6). Growth regulation by the Rb family is effected primarily through its interaction with the E2F family of transcription factors. The E2Fs are subcategorized into activators, including E2F1, E2F2, and E2F3a; repressors E2F3b, E2F4, and E2F5; and the less characterized E2F6, E2F7, and E2F8 (8, 22, 23). The E2Fs regulate the transcription of many genes necessary for both the G1/S and G2/M phase transitions (3, 43). Rb proteins induce and maintain growth arrest by binding to E2Fs and thereby repressing E2F-responsive gene transcription. Upon growth-stimulatory signals, the Rb protein becomes hyperphosphorylated by cyclin-dependent kinases, ultimately resulting in the release of E2F and derepression of E2F transcription, thus allowing entry into the cell cycle. The SV40 large T antigen bypasses Rb-dependent repression, even in the absence of external growth signals. T antigen binds to pRb and the related proteins p107 and p130 via an LXCXE motif. T antigen disrupts p130–E2F complexes by recruiting the cellular molecular chaperone hsc70 through the J domain at its amino terminus (36). T antigen then catalyzes the release of E2F from p130 by an energy-dependent mechanism involving ATP hydrolysis by hsc70. As a consequence of T-antigen action, p130 is transported to the proteasome, where it is degraded (35). T antigen is thought to stimulate the release of E2F from pRb and p107 by similar mechanisms, although these processes are less well studied. The disruption of

13191

13192

´ ENZ-ROBLES ET AL. SA

FIG. 1. Domain maps of SV40 T antigens: J domain, Rb proteinbinding motif (LXCXE), nuclear localization signal (NLS), originbinding domain (OBD), Zn domain, ATPase domain, and host range domain (HR). Interactions with cellular components and phenotypic effects of protein expression in intestinal enterocytes are indicated.

Rb-E2F complexes and the consequent release of E2F are thought to lead to the derepression of E2F-responsive genes. In this study, we have taken advantage of the unique architecture of the small intestine, where pluripotent progenitor cells residing in the crypts can be separated from terminally differentiated enterocytes that occupy the villi, to allow exploration of the role of the E2F family in SV40-mediated transformation. Previous studies have shown that expression of wild-type large T antigen in enterocytes induces their ectopic proliferation and that the induction of hyperplasia is dependent on T-antigen binding to Rb proteins (4, 15, 18) (Fig. 1). Furthermore, expression of the first 121 or 13b amino acids of T antigen is capable of inactivating Rb proteins and is sufficient for hyperplasia (17) (Fig. 1). The ability to isolate highly enriched populations of enterocytes from transgenic mice has allowed us to examine the molecular interactions of T antigen with the Rb-E2F pathway and to explore their effects on E2Fmediated gene regulation and enterocyte proliferation. We found that specific activators, E2F2 and E2F3a, are responsible for the proliferative phenotype induced by T antigen. These results favor a model in which each E2F factor has tissuespecific roles and cannot be replaced by increased levels of other members of the family. MATERIALS AND METHODS Generation of T-antigen-transgenic mice. All animal use and manipulation described in this report complied with all relevant federal guidelines and insti-

J. VIROL. tutional policies. The promoter region (1.2 kb) of the rat intestinal fatty acid binding protein (IFABP) was excised from pEP.IFABP (J. Gordon, Washington University, St. Louis, MO) via EcoRI/SalI digestion and inserted into the EcoRI/ SalI-linearized phrGFP vector (Stratagene), generating the pIFABP.GFP plasmid. A BamHI/XhoI fragment (2.7 kb) from pRSVBneo3213 (28), encoding a T-antigen mutant with Glu-to-Lys substitutions at residues 107 and 108, was ligated to the 3.9-kb BclI/XhoI vector portion of pIFABP.GFP. The resulting plasmid (pIFABP.3213) was digested with SacII/MluI to yield a 4.2-kb fragment containing the IFABP promoter, the coding sequences for SV40 large-T-antigen mutant 3213, and the SV40 polyadenylation signals. Pronuclear microinjection of this fragment resulted in transgenic mice (Fabpi-SV40 TAg3213) expressing TAg3213 and small t antigen in the enterocytes of the small intestine. Similarly, a 1.0-kb BamHI/XhoI fragment from pRSVBneoN136 (34) was inserted into the 3.9-kb BclI/XhoI vector portion of pIFABP.GFP, generating pIFABP.N136. The N136 mutant expresses an amino-terminally truncated version of T antigen encoding the first 136 amino acids including the nuclear localization signal. Pronuclear microinjection of a 2.6-kb NsiI/MluI fragment rendered mice FabpiSV40 TAgN136 transgenic. The TAg3213 mutant protein fails to bind pRb and related pocket proteins, while the TAgN136 mutant retains the ability to bind pRb-related proteins but lacks the sequences capable of binding p53. All constructs are able to produce small t antigen and were confirmed by sequencing analysis. TAg3213 and TAgN136 transgenic mice were generated by conventional pronuclear injection at the University of Michigan Transgenic Mouse Facility Service. We obtained 13 transgenic founders for the N136 construct and 5 for the 3213 construct, and we established transgenic lines from 2 and 4 of them, respectively. We used TAg3213 from pedigree A and TAgN136 from pedigree B for this study. Transgenic mice expressing SV40 wild-type T antigen (TAgwt) under the control of the rat intestinal fatty acid binding protein promoter were obtained from Jeff Gordon and were from pedigree 103 (18). For purposes of simplification, animals are referred to below as TAgwt (FabpiSV40 TAgwt), TAg3213 (Fabpi-SV40 TAg3213), and TAgN136 (Fabpi-SV40 TAg N136). All pedigrees were maintained by crosses to nontransgenic FVB mice obtained commercially (Taconic Labs). All routine screenings for several murine pathogens and parasites were negative. The genotype of each mouse was determined using DNA from murine tails and PCR with primers spanning the region from the IFABP promoter (5⬘-CTG TTT GGT TTG GTT TAT CGC C) to the second exon in the T antigen (5⬘-CCA TTC ATC AGT TCC ATA GGT TGG), which yielded a 1.1-kb product after 30 cycles of conventional PCR with annealing at 55°C. Generation of T-antigen-transgenic mice in an E2F knockout (KO) background. E2F1⫺/⫺, E2F2⫺/⫺, and E2F3a⫺/⫺ mice have been generated previously (9, 20; R. Opavsky and G. Leone, unpublished data). Standard genetic crosses were performed to generate E2F2⫺/⫺, E2F3a⫺/⫺, and TAgwt mice, or their various combinations. The genotype of each mouse was determined using DNA isolated from murine tails followed by PCR with protocols detailed elsewhere (20) and in the preceding section. All mice used in this study were adults, with ages ranging from 1.5 to 6 months. Mice used for comparisons between different genotypes were age matched. Intestinal epithelial fractionation. Intestinal fractions enriched for villi were prepared as described previously (24), except that different concentrations of EDTA were used for the first incubation, depending on the genotype of the mouse under analysis: 5 mM for nontransgenic and TAg3213 mice and 3 mM for TAgwt and TAgN136 mice. Imununoblot and immunoprecipitation analysis. Conventional Western blotting and immunoprecipitation techniques have been described previously (24). Appropriate dilutions of the following primary antibodies were used: anti-pRb mouse immunoglobulin G (IgG) (G-3245, 14001A) from BD-Pharmingen and anti-pRb mouse IgG1 (IF8; catalog no. sc-102), anti-p130 rabbit IgG (C-20; catalog no. sc-317), anti-p107 rabbit IgG (C-18; catalog no. sc-318), anti-E2F1 rabbit IgG (C-20; catalog no. sc-193), anti-E2F2-mouse IgG1 (TFE25; catalog no. sc-9967), anti-E2F2 rabbit IgG (C-20; catalog no. sc-633), anti-E2F3 rabbit IgG (C-18; catalog no. sc-878), anti-E2F4 rabbit IgG (A-20; catalog no. sc-1082), and anti-histone H3 rabbit IgG (FL-136; catalog no. sc-10809) from Santa Cruz Biotechnologies. Goat anti-mouse (A2554) and goat anti-rabbit (A0545) antibodies (Sigma) linked to horseradish peroxidase were used as secondary antibodies. The peroxidase reactions were developed with ECL-plus reagents (Amersham Life Sciences). Appropriate positive and negative E2F controls were obtained from nuclear extracts from S-phase-synchronized wild-type or E2F KO mouse embryonic fibroblasts (MEFs), prepared as described previously (21, 24), or from 293 cells infected with an E2F-expressing adenovirus vector. E2F electrophoretic mobility shift assay (EMSA). E2F protein-DNA complexes were analyzed as described previously, with minor modifications (36). Centri 䡠 Spin-20 columns (Princeton Separations) were used to remove unincor-

VOL. 81, 2007

INTESTINAL HYPERPLASIA REQUIRES E2F2

13193

FIG. 2. Expression of T antigen alters specific Rb family protein and transcript levels in villi, and such alteration is dependent on an intact LxCxE motif. (A) Rb family steady-state protein levels in nontransgenic and T-antigen-transgenic villus samples. Protein extracts from villi of nontransgenic and T-antigen-transgenic mice were subjected to immunoblotting for pRb, p130, and p107. Positive (⫹/⫹) and negative (⫺/⫺) controls refer to normal and KO MEFs, respectively. Protein levels for ␤-tubulin were used as loading controls. Arrow indicates the position of p130. (B) Rb family steady-state transcript levels in nontransgenic and T-antigen-transgenic villus samples. Equal amounts of cDNAs reverse transcribed from total RNA extracts from villi of nontransgenic and T-antigen-transgenic mice were subjected to PCR amplification using specific primers for pRb, p130, and p107. Transcript levels of Adh5 were used as loading controls.

porated [32P]␥ATP from the labeled oligonucleotide. Competitor assays were performed with a 100-fold molar excess of unlabeled wild-type and mutant oligonucleotides. In addition to the antibodies used for immunoblot experiments, the following antibodies were used for supershift assays: anti-p107 mouse IgG1 (SD9; catalog no. sc-250), anti-E2F1 mouse IgG2a (KH95; catalog no. sc-251), anti-E2F1 rabbit IgG (C-20; catalog no. sc-193), anti-E2F3a rabbit IgG (N-20; catalog no. sc-879), and anti-E2F5 rabbit IgG (H-111; catalog no. sc-1699) (all from Santa Cruz Biotechnologies, Inc.). Reverse transcription-PCR (RT-PCR) analysis. RNAs were extracted from whole intestines or intestinal fractions (RNeasy; Qiagen). The cDNAs were amplified by PCR using primers specific for Adh (5⬘-ATGACAGATGGGGGCGTGG ATTAC–5⬘-TGGAATGGACGAGTGGAGATTTC), E2F1 (5⬘-TTGCCTGTCTG TTTGCTGAGCC–5⬘-CGGAGATTTTCACACCTTTCCCTG), E2F2 (5⬘-TTCGC TTTACACGCAGACGG–5⬘-AATGAACTTCTTGGTCAGGAGCC), E2F3a (5⬘AGCCTCTACACCACGCCACAA–5⬘-ATCCAGGACCCCATCAGGAGAC), E2F3b (5⬘-TTACAGCAGCAGGCAAAGC–5⬘-GAACTTCTTGGTGAGCAGA CCG), E2F4 (5⬘-CCAAGAATCCTCTCCTCCAAG–5⬘-GCACAGACACCTTCA CTCTCGTCC), Rb (5⬘-TCACACAACCCAGCAGTGC–5⬘-CTATCCGAGCGC TCCTGTTC), p107 (5⬘-TGGATTATTGAAGTTCTC–5⬘-CTGATCCAAATGCC TATC), p130 (5⬘-TTGGGACTCTGTCTCGGTGTCTAAG–5⬘-AATGCGTCAT GCTCCAGAACACCAG), p19ARF (5⬘-TGGACCAGGTGATGATGATGG–5⬘-A AGAAAAAGGCGGGCTGAGG), TK1 (5⬘-GCTTTCGGCAGCATCTTGAAC– 5⬘-CCCTCAGTTGGCAGAGTTGTATTG), or dihydrofolate reductase (DHFR) (5⬘-TGGTTTGGATAGTCGGAGGCAG–5⬘-GGGGAGCAGAGAACTTGAAA GC). Primers for B-myb and cyclin E1 have been described previously (39). Exponential amplifications of PCR products were performed with an initial denaturation step (2 min at 94°C); a series of 18 to 25 cycles between 94°C for 30 s, different annealing temperatures for 30 s, and 72°C for 30 s; and a final extension step for 7 min at 72°C. Annealing temperatures and product sizes for each reaction were as follows: for E2F1, 60°C and 420 bp; for E2F2, 60°C and 289 bp; for E2F3a, 60°C and 309 bp; for E2F3b, 56°C and 189 bp; for E2F4, 59°C and 495 bp; for RB, 56°C and 280 bp; for p130, 57°C and 499 bp; for p107, 50°C (1-min extension) and 936 bp; for p19ARF, 60°C and 415 bp; for TK1, 59°C and 312 bp; for DHFR, 55°C and 261 bp; for B-myb, 58°C and 350 bp; and for cyclin E1, 55°C and 608 bp. Immunohistochemistry. Small intestines were processed for immunohistochemistry as described elsewhere (24). Proliferative status was determined after intraperitoneal injection of 5-bromo-2⬘-deoxyuridine (BrdU; 100 mg/kg of body weight; Sigma) 2 h prior to sacrifice. Intestinal sections were stained with the rat monoclonal antibody BU1/75 (Accurate Scientific, Westbury, NY), and antigenantibody complexes were detected with an ABC Elite kit (rat IgG; Vector Laboratories, Burlingame, CA), followed by development of the peroxidase reaction with diaminobenzidine substrate (Dako). Stained sections of murine intestines were photographed under a Nikon FXA microscope (original magnification, ⫻200).

RESULTS SV40 T antigen increases p107 and decreases p130 protein levels in enterocytes. We assessed the steady-state levels of pRb, p107, and p130 by performing immunoblot experiments on extracts prepared from intestinal fractions enriched for enterocytes from T-antigen-expressing mice or their nontransgenic littermates (Fig. 2A). The pRb and p130 proteins were readily detectable in the growth-arrested enterocytes of nontransgenic mice. T-antigen expression did not alter the levels or phosphorylation state of pRb. In contrast, little or no p130 could be detected in the enterocytes of mice expressing wildtype T antigen, while T-antigen expression induced higher levels of p107. Semiquantitative RT-PCR experiments indicated that the increased p107 levels are due, at least in part, to higher levels of p107 mRNA (Fig. 2B). T-antigen expression had little or no effect on the abundance of pRb or p130 mRNAs. Previous cell culture experiments showing that T antigen induces the proteasome-dependent degradation of p130 (35) are consistent with these results. We also examined the levels of the pRb, p107, and p130 proteins and mRNAs in enterocytes expressing the T-antigen mutant with the E107K and E108K mutations in the LXCXE motif (TAg3213), which is unable to associate with the Rb family of proteins. Both the T-antigeninduced reduction in p130 protein levels and the increase in p107 protein and mRNA levels depend on an intact LXCXE motif (Fig. 2). Furthermore, a truncated T antigen, TAgN136, that consists of the first 136 amino acids of T antigen and includes the J domain and LXCXE motif was sufficient to reduce p130 protein levels and to increase p107 mRNA and protein levels. These results indicate that inactivation of the Rb family by T antigen results in p130 degradation and an increase in p107 transcription. Confirmation that T antigen binds Rb proteins in intestinal epithelial cells was obtained by coimmunoprecipitation experiments. As expected, TAgwt, but

13194


J. VIROL.

FIG. 3. T antigen alters Rb-E2F DNA complexes in villus enterocytes. Protein extracts from villi of nontransgenic and T-antigen-transgenic mice were subjected to E2F EMSA analysis and antibody supershifts. (A) Three major E2F complexes are present in nontransgenic mice, and three new bands appear in TAgwt intestines. (B) Identification of E2F-containing complexes in control and T-antigen-expressing samples. Antibodies specific for E2F1, E2F2, or E2F3 were added to the EMSA reaction mixtures.

not TAg3213, was found associated with pRb, p107, and p130 in extracts prepared from enterocytes (data not shown). T antigen disrupts p130–E2F DNA complexes. Rb proteins exert their effects in part by binding and regulating the E2F family of transcription factors. To examine the effects of Tantigen expression on Rb-E2F DNA binding complexes, we performed a series of EMSA experiments. When a radiolabeled DNA consensus E2F probe is incubated with extracts from normal growth-arrested enterocytes and resolved on a nondenaturing gel, three main shifted bands are observed (Fig. 3A). Competition experiments with an unlabeled wild-type or mutant probe indicated that all three bands represent specific protein-DNA complexes (data not shown). To determine which Rb proteins were present in these complexes, we performed antibody supershift experiments (Fig. 3B). The specificity of the antibodies used in these experiments was confirmed using insect cell lysates overexpressing pRbE2F complexes or intestinal extracts obtained from p130- or p107-null mice (data not shown). The slowest-migrating species consisted of p130 bound to E2F. The two faster-migrating

species contained E2F-DNA complexes without any Rb proteins. The same three E2F-DNA complexes were seen in enterocytes expressing the mutant TAg3213, defective in binding of Rb proteins (Fig. 3A). In contrast, no p130–E2F complexes were detected in enterocytes expressing TAgwt or the truncation mutant TAgN136. In addition to the Rb-free E2F complexes found in control mice, these cells contained minor amounts of complexes containing p107 or pRb and one additional Rb-free E2F complex. None of these complexes were present in nontransgenic or TAg3213-expressing enterocytes. T-antigen expression in enterocytes induces a new pattern of Rb protein-free E2F complexes. EMSA and antibody supershift experiments revealed the presence of two Rb-free E2FDNA complexes in the villi of nontransgenic mice and three species in TAgwt-transgenic animals (Fig. 3A). A detailed characterization of the complexes was performed (Fig. 3B). The slowest-migrating species corresponded to E2F4, while the fastest species contained E2F5. Levels of both of these species were greatly enhanced in extracts prepared from transgenic mice, suggesting that T-antigen expression results in increased

VOL. 81, 2007


13195

FIG. 4. (A) Semiquantitative RT-PCR analysis of RNA samples obtained from villus enterocytes from control or T-antigen-expressing mice. Adh was used as a loading control. (B) Steady-state levels of E2F proteins in transgenic intestines. Protein extracts from whole intestines were used to analyze the levels of E2F2 and E2F3a by immunoblotting; ␤-tubulin was included as a loading control. E2F1 levels were very low and were detected reproducibly only in nuclear extracts; histone H3 was used as a loading control. Lanes labeled (⫹) and (⫺) refer, respectively, to normal and KO MEFs used as controls. Levels of the activators E2F2 and E2F3a are higher in intestinal villi expressing TAgwt or TAgN136 than in control or TAg3213-expressing intestinal villi.

E2F4 and E2F5 DNA binding. In addition, DNA complexes containing E2F2 and E2F3a were detected in the proliferating enterocytes from TAgwt- and TAgN136-expressing intestinal villi. Protein extracts from gene KO mice were used to confirm the specificity of the antibodies used in the supershift experiments (data not shown). T antigen upregulates E2F2, E2F3a, and E2F-responsive genes. We explored the molecular mechanisms leading to increased E2F-DNA binding activity in T-antigen-expressing enterocytes. Semiquantitative RT-PCR experiments showed higher E2F2 and E2F3a mRNA levels in T-antigen-expressing intestines (Fig. 4A). In each case, the increase in the mRNA level required the LXCXE motif but not the carboxy-terminal region of T antigen. The levels of E2F3b and E2F4 mRNAs did not change in response to T antigen. Concomitantly with the increases in E2F2 and E2F3a mRNA levels, Western blot analysis showed higher levels of E2F2 and E2F3a proteins in TAgwt-expressing intestinal extracts (Fig. 4B). Similarly, increased levels of E2F2 and E2F3a were observed in mice expressing TAgN136. Normal levels of E2F2 and E2F3a were seen in TAg3123 mice, indicating that their upregulation requires an intact LXCXE motif. Furthermore, this increase correlated with a change in the protein migration pattern, but at this time the cause of the altered mobility of E2F3a in transgenic intestines has not been elucidated. E2F1 protein levels were very low and were detectable only in nuclear extracts from murine villi. Under those conditions, we also observed a modest increase in the steady-state levels of E2F1 (Fig. 4B). T-antigen expression did not significantly alter the steady-state levels of E2F3b or E2F4 (data not shown).

The appearance of new DNA complexes containing the activating E2Fs, E2F2 and E2F3a, should lead to increased E2Fdependent transcription. To test this, we used semiquantitative RT-PCR to assess the mRNA levels of five genes known to be regulated by E2F: thymidine kinase, B-myb, DHFR, cyclin E1, and p19ARF (Fig. 4A). Each of these showed increased mRNA levels in the presence of wild-type T antigen, and in each case the elevation in mRNA levels depended on a functional interaction through the LXCXE domain. T-antigen-induced hyperplasia is reduced in the absence of E2F2 and E2F3a. To determine if the induction of E2F2 and E2F3a contributes to T-antigen-mediated hyperplasia, we generated T-antigen-expressing transgenic mice in an E2F1-, E2F2-, or E2F3a-null background and assessed their morphology and proliferative status (Fig. 5). E2F1- and E2F2-null mice have been generated previously and have no intestinal defects (9, 20, 41). E2F3a KO mice have been generated in our group and show no signs of intestinal anomalies (R. Opavsky et al., unpublished data). The partial (n ⫽ 6) or complete (n ⫽ 9) removal of E2F1 had no effect on the ability of T antigen (n ⫽ 17) to induce intestinal hyperplasia (Fig. 5B and C). Concomitantly, ectopic proliferation induced by T antigen was not diminished by the absence of E2F1 (Fig. 6A). In contrast, the loss of E2F2 in T-antigen-expressing intestines (n ⫽ 5) resulted in an intermediate phenotype, comprising areas of normal and hyperplastic tissue, where the proliferative cells extended to the differentiation zone (Fig. 5B and D) but the total numbers of proliferative cells in both the villi and the crypts were drastically diminished in comparison to those for the mice expressing T antigen in the presence of E2F2 (Fig. 6A). Further removal of E2F3a in TAg⫹; E2F2 KO mice (n ⫽ 3) did not

13196


J. VIROL.

FIG. 5. Intermediate proliferative phenotype in the absence of functional E2F2 and E2F3a. Duodenal sections from adult mice were stained for BrdU and counterstained with hematoxylin. In contrast with control mice (A), the expression of T antigen induces ectopic proliferation and intestinal hyperplasia (B), which is ameliorated in the absence of E2F2 alone (D) or in the absence of both E2F2 and E2F3a (E) but not in T-antigen-expressing mice lacking E2F1 (C).

accentuate the amelioration of the T-antigen-induced phenotype already observed in the absence of E2F2 (Fig. 5B and E and Fig. 6A), suggesting that E2F2 is the primary mediator of T-antigen-induced proliferation. The residual ectopic proliferation observed in TAgwt; E2F2 KO or TAgwt; E2F2 KO; E2F3a KO mice did not appear to be caused by upregulation of the other main activator (E2F1), because the steady-state levels of E2F1 remained the same in all intestines expressing T antigen, regardless of the genetic background (Fig. 6B). Despite having reduced numbers of cycling enterocytes, mice expressing T antigen in the absence of E2F2 and/or E2F3a still showed histological abnormalities of the villi and a partial hyperplastic phenotype (Fig. 5D and E; compare with Fig. 5A). Presumably this altered tissue structure is due either to enterocytes that are stimulated to proliferate even in the absence of the activating E2Fs or to some T-antigen activity that acts independently of the E2Fs. Finally, we note that this is the first description of E2F2 playing a direct role in oncogenesis. The results reported here firmly establish the requirement for activating E2Fs in SV40-mediated transformation. DISCUSSION In this paper, we show that a specific activating E2F transcription factor, namely, E2F2, is required for the induction of enterocyte proliferation and intestinal hyperplasia by SV40 large T antigen. Our results support a model in which Rbmediated repression of E2F-dependent transcription is the major mechanism that maintains enterocyte growth arrest. T antigen stimulates enterocyte proliferation by destroying p130–

E2F4 repressive complexes and increasing the levels of activating E2Fs, most notably E2F2. The increased levels of E2F2 result in the expression of E2F-regulated genes and the induction of enterocyte proliferation. The number of proliferating enterocytes is greatly reduced in T-antigen-transgenic mice lacking E2F2. These results demonstrate that E2F-dependent gene expression is required for the transformation of enterocytes by SV40. The mouse small intestine provides an in vivo model for studies of cell growth regulation. This tissue consists of stem cells and pluripotent progenitor cells that lie in the crypts, invaginations located below the villi within the stromal tissue. These structures harbor a population of about 300 cycling cells that differentiate to give rise to the villi, finger-like projections that project into the intestinal lumen. The villi are composed of three epithelial cell types—enterocytes, goblet cells, and enteroendocrine cells—the vast majority of which are enterocytes (12, 32). Expression of the SV40 large T antigen (TAgwt) in the enterocytes of transgenic mice induces ectopic cell proliferation and results in intestinal hyperplasia (15, 18). TAgwt induces cell proliferation, in part by binding the Rb family of tumor suppressor proteins via its LXCXE motif, and, in cooperation with its J domain, is able to eliminate the repression of E2F-dependent transcription mediated by pRb’s (36; reviewed in reference 31). Expression of truncated T-antigen mutants containing the J domain and LXCXE motif is also able to induce enterocyte proliferation and hyperplasia (17; M. T. Saénz-Robles et al., unpublished data). In contrast, enterocytes expressing a mutant T antigen that is unable to interact

VOL. 81, 2007


13197

FIG. 6. (A) The increase in proliferation following expression of T antigen in enterocytes is reduced in the absence of E2F2 alone or of both E2F2 and E2F3a but not in the absence of E2F1. Total numbers of positive BrdU-labeled cells in either the villi or the crypt were counted and averaged for each group of mice and are represented in the graph. Error bars indicate the standard deviation in each group. (B) Ectopic proliferation in TAgwt; E2F2 KO mice does not correlate with increased E2F1 protein levels. E2F1 expression was measured by Western blotting against a positive control, nontransgenic intestinal samples, or intestinal samples expressing TAgwt in the presence or absence of functional products of E2F2 alone or of both E2F2 and E2F3a. Depletion of E2F2 was confirmed for the corresponding samples.

with Rb proteins remain growth arrested, indicating that the inhibition of the Rb family is required for entry of the enterocytes into the cell cycle and for intestinal hyperplasia (4). These results are consistent with gene KO studies indicating that inactivation of pRb in embryonic intestines induces hyperproliferation and abnormal gut formation (42) and that the ablation of both pRb and either p107 or p130 in adult villi is required to induce enterocyte proliferation (14, 19). We have taken advantage of the morphology of the mouse intestine to physically separate the villus epithelial cells from the crypts and underlying mesenchyme. This allowed us to obtain extracts of highly enriched enterocytes for molecular analysis (24). We find that normal growth-arrested enterocytes contain abundant p130–E2F4 complexes. As expected, enterocytes express very low levels of E2F-dependent genes and of genes involved in DNA replication or cell

cycle progression. Enterocytes expressing the large T antigen contain little or no p130, consistent with cell culture experiments demonstrating that T antigen induces the proteasome-dependent degradation of p130 (35).We also observe that T-antigen expression results in a change in the phosphorylation state and subcellular localization of E2F4 (data not shown). In addition, T-antigen expression results in a significant increase in E2F4–DNA complexes, even though E2F4 protein levels remain unchanged. Previous reports (7, 11) have suggested a correlation between the phosphorylation status of E2F4 and its role in cell proliferation, but at present we do not know if the change in E2F4 phosphorylation is related to changes in its DNA binding activity or to its redistribution from the cytoplasm to the nucleus. Since the repressive complexes consist of E2F4 in association with p130, we do not think E2F4 acts as a re-

13198


pressor in the presence of T antigen, despite the changes observed in DNA binding activity and phosphorylation state. Additionally, T-antigen expression results in marked increases in the levels of E2F2 and E2F3a proteins, as well as in the appearance of E2F2– and E2F3a–DNA complexes. A modest increase in E2F1 levels is also observed, but we do not detect any E2F1–DNA complexes. Consistent with this observation, E2F-dependent genes, as well as DNA replication genes and genes associated with cell cycle progression, are upregulated in T-antigen-expressing enterocytes. The increased levels of E2F2, E2F3a, and E2F-dependent genes require T-antigen binding to Rb proteins. Furthermore, a truncated T antigen capable of inactivating all the Rb proteins also induces expression of E2F2 and E2F3a and of E2F-dependent genes. We tested the hypothesis that specific E2Fs mediate the ectopic enterocyte proliferation induced by T antigen by generating mice expressing the oncogene in a genetic background void of E2F1, E2F2, or both E2F2 and E2F3a. We observed that elimination of E2F2 by itself or in conjunction with E2F3a reduces the hyperplastic and proliferative phenotype observed in TAgwt-expressing mice. In contrast, the absence of E2F1 has no effect on T-antigen-induced enterocyte proliferation. The residual ectopic proliferation observed in transgenic mice without functional E2F2 or without functional E2F2 and E2F3a cannot be attributed to compensation through upregulation of E2F1, since no increase in E2F1 steady-state levels is observed in TAgwt-expressing mice lacking those activator E2Fs. In agreement with our results, expression of E2F1 in the enterocytes of transgenic mice does not induce ectopic proliferation (4). Other transgenic systems have shown involvement of E2F1 in either the apoptosis (2) or both the proliferation and the apoptosis induced by T antigen (27). Thus, the specific E2F family members that drive proliferation may depend on the cell type and context used. Although Rb proteins are able to interact with and presumably regulate the activity of a number of proteins in addition to E2F, our results suggest that the Rb/E2F pathway is a main mediator of ectopic cell proliferation in enterocytes. Nevertheless, our experiments address only the roles of E2F2 and E2F3a in the reentry of growth-arrested enterocytes into the cell cycle. Thus, the events leading to the exit of pluripotent progenitor cells from the cell cycle during differentiation, or those that govern the continuous cycling of progenitor or stem cells, may differ from those observed here.

ACKNOWLEDGMENTS We thank Liz Laposata, Chevaunne Edwards, and Danielle Oliver for expert technical assistance and Simon Watkins for allowing us use of the microscopes and equipment at the Center for Biologic Imaging of the University of Pittsburgh. We also thank James DeGregori for E2F2 expression vectors and KO cell lines. This work was supported by NIH grant CA098956 to J.M.P., NIH/ NIDDK grant DK61218 to R.H.W., NIH grants R01CA85619, R01CA82259, R01HD047470, and P01CA097189 to G.L., and NIH grant DK058404 to the Vanderbilt University Medical Center’s Digestive Disease Research Center. J.-L.C. is the recipient of DoD award BC061730.

J. VIROL. REFERENCES 1. Ahuja, D., M. T. Saenz-Robles, and J. M. Pipas. 2005. SV40 large T antigen targets multiple cellular pathways to elicit cellular transformation. Oncogene 24:7729–7745. 2. Athanasiou, M. C., W. Yunis, N. Coleman, R. Ehlenfeldt, H. B. Clark, H. T. Orr, and R. M. Feddersen. 1998. The transcription factor E2F-1 in SV40 T antigen-induced cerebellar Purkinje cell degeneration. Mol. Cell. Neurosci. 12:16–28. 3. Blais, A., and B. D. Dynlacht. 2004. Hitting their targets: an emerging picture of E2F and cell cycle control. Curr. Opin. Genet. Dev. 14:527–532. 4. Chandrasekaran, C., C. M. Coopersmith, and J. I. Gordon. 1996. Use of normal and transgenic mice to examine the relationship between terminal differentiation of intestinal epithelial cells and accumulation of their cell cycle regulators. J. Biol. Chem. 271:28414–28421. 5. Chen, J., G. J. Tobin, J. M. Pipas, and T. Van Dyke. 1992. T-antigen mutant activities in vivo: roles of p53 and pRB binding in tumorigenesis of the choroid plexus. Oncogene 7:1167–1175. 6. Cobrinik, D. 2005. Pocket proteins and cell cycle control. Oncogene 24: 2796–2809. 7. Descheˆnes, C., L. Alvarez, M. E. Lizotte, A. Vezina, and N. Rivard. 2004. The nucleocytoplasmic shuttling of E2F4 is involved in the regulation of human intestinal epithelial cell proliferation and differentiation. J. Cell. Physiol. 199:262–273. 8. Dimova, D. K., and N. J. Dyson. 2005. The E2F transcriptional network: old acquaintances with new faces. Oncogene 24:2810–2826. 9. Field, S. J., F. Y. Tsai, F. Kuo, A. M. Zubiaga, W. G. Kaelin, Jr., D. M. Livingston, S. H. Orkin, and M. E. Greenberg. 1996. E2F-1 functions in mice to promote apoptosis and suppress proliferation. Cell 85:549–561. 10. Fields, B. N. (ed.). 1985. Virology. Raven, New York, NY. 11. Gaubatz, S., G. J. Lindeman, S. Ishida, L. Jakoi, J. R. Nevins, D. M. Livingston, and R. E. Rempel. 2000. E2F4 and E2F5 play an essential role in pocket protein-mediated G1 control. Mol. Cell 6:729–735. 12. Gordon, J. I., and M. L. Hermiston. 1994. Differentiation and self-renewal in the mouse gastrointestinal epithelium. Curr. Opin. Cell Biol. 6:795–803. 13. Hahn, W. C., S. K. Dessain, M. W. Brooks, J. E. King, B. Elenbaas, D. M. Sabatini, J. A. DeCaprio, and R. A. Weinberg. 2002. Enumeration of the simian virus 40 early region elements necessary for human cell transformation. Mol. Cell. Biol. 22:2111–2123. 14. Haigis, K., J. Sage, J. Glickman, S. Shafer, and T. Jacks. 2006. The related retinoblastoma (pRb) and p130 proteins cooperate to regulate homeostasis in the intestinal epithelium. J. Biol. Chem. 281:638–647. 15. Hauft, S. M., S. H. Kim, G. H. Schmidt, S. Pease, S. Rees, S. Harris, K. A. Roth, J. R. Hansbrough, S. M. Cohn, D. J. Ahnen, N. A. Wright, R. A. Goodland, and J. I. Gordon. 1992. Expression of SV-40 T antigen in the small intestinal epithelium of transgenic mice results in proliferative changes in the crypt and reentry of villus-associated enterocytes into the cell cycle but has no apparent effect on cellular differentiation programs and does not cause neoplastic transformation. J. Cell Biol. 117:825–839. 16. Hill, R., Y. Song, R. D. Cardiff, and T. Van Dyke. 2005. Heterogeneous tumor evolution initiated by loss of pRb function in a preclinical prostate cancer model. Cancer Res. 65:10243–10254. 17. Kim, S. H., K. A. Roth, C. M. Coopersmith, J. M. Pipas, and J. I. Gordon. 1994. Expression of wild-type and mutant simian virus 40 large tumor antigens in villus-associated enterocytes of transgenic mice. Proc. Natl. Acad. Sci. USA 91:6914–6918. 18. Kim, S. H., K. A. Roth, A. R. Moser, and J. I. Gordon. 1993. Transgenic mouse models that explore the multistep hypothesis of intestinal neoplasia. J. Cell Biol. 123:877–893. 19. Kucherlapati, M. H., A. A. Nguyen, R. T. Bronson, and R. S. Kucherlapati. 2006. Inactivation of conditional Rb by Villin-Cre leads to aggressive tumors outside the gastrointestinal tract. Cancer Res. 66:3576–3583. 20. Leone, G., R. Sears, E. Huang, R. Rempel, F. Nuckolls, C. H. Park, P. Giangrande, L. Wu, H. I. Saavedra, S. J. Field, M. A. Thompson, H. Yang, Y. Fujiwara, M. E. Greenberg, S. Orkin, C. Smith, and J. R. Nevins. 2001. Myc requires distinct E2F activities to induce S phase and apoptosis. Mol. Cell 8:105–113. 21. Lindeman, G. J., L. Dagnino, S. Gaubatz, Y. Xu, R. T. Bronson, H. B. Warren, and D. M. Livingston. 1998. A specific, nonproliferative role for E2F-5 in choroid plexus function revealed by gene targeting. Genes Dev. 12:1092–1098. 22. Logan, N., A. Graham, X. Zhao, R. Fisher, B. Maiti, G. Leone, and N. B. La Thangue. 2005. E2F-8: an E2F family member with a similar organization of DNA-binding domains to E2F-7. Oncogene 24:5000–5004. 23. Maiti, B., J. Li, A. de Bruin, F. Gordon, C. Timmers, R. Opavsky, K. Patil, J. Tuttle, W. Cleghorn, and G. Leone. 2005. Cloning and characterization of mouse E2F8, a novel mammalian E2F family member capable of blocking cellular proliferation. J. Biol. Chem. 280:18211–18220. 24. Markovics, J. A., P. A. Carroll, M. T. Robles, H. Pope, C. M. Coopersmith, and J. M. Pipas. 2005. Intestinal dysplasia induced by simian virus 40 T antigen is independent of p53. J. Virol. 79:7492–7502. 25. McLear, J. A., G. Garcia-Fresco, M. A. Bhat, and T. A. Van Dyke. 2006. In

VOL. 81, 2007

26.

27. 28. 29.

30.

31. 32. 33. 34.

35.

vivo inactivation of pRb, p107 and p130 in murine neuroprogenitor cells leads to major CNS developmental defects and high seizure rates. Mol. Cell. Neurosci. 33:260–273. Pallas, D. C., L. K. Shahrik, B. L. Martin, S. Jaspers, T. B. Miller, D. L. Brautigan, and T. M. Roberts. 1990. Polyoma small and middle T antigens and SV40 small t antigen form stable complexes with protein phosphatase 2A. Cell 60:167–176. Pan, H., C. Yin, N. J. Dyson, E. Harlow, L. Yamasaki, and T. Van Dyke. 1998. Key roles for E2F1 in signaling p53-dependent apoptosis and in cell division within developing tumors. Mol. Cell 2:283–292. Peden, K. W., and J. M. Pipas. 1992. Simian virus 40 mutants with aminoacid substitutions near the amino terminus of large T antigen. Virus Genes 6:107–118. Rodriguez-Viciana, P., C. Collins, and M. Fried. 2006. Polyoma and SV40 proteins differentially regulate PP2A to activate distinct cellular signaling pathways involved in growth control. Proc. Natl. Acad. Sci. USA 103:19290– 19295. Sa énz Robles, M. T., H. Symonds, J. Chen, and T. Van Dyke. 1994. Induction versus progression of brain tumor development: differential functions for the pRB- and p53-targeting domains of simian virus 40 T antigen. Mol. Cell. Biol. 14:2686–2698. Sa énz-Robles, M. T., C. S. Sullivan, and J. M. Pipas. 2001. Transforming functions of simian virus 40. Oncogene 20:7899–7907. Sancho, E., E. Batlle, and H. Clevers. 2004. Signaling pathways in intestinal development and cancer. Annu. Rev. Cell Dev. Biol. 20:695–723. Reference deleted. Srinivasan, A., A. J. McClellan, J. Vartikar, I. Marks, P. Cantalupo, Y. Li, P. Whyte, K. Rundell, J. L. Brodsky, and J. M. Pipas. 1997. The aminoterminal transforming region of simian virus 40 large T and small t antigens functions as a J domain. Mol. Cell. Biol. 17:4761–4773. Stubdal, H., J. Zalvide, K. S. Campbell, C. Schweitzer, T. M. Roberts, and


36.

37.

38. 39. 40.

41. 42. 43.

13199

J. A. DeCaprio. 1997. Inactivation of pRB-related proteins p130 and p107 mediated by the J domain of simian virus 40 large T antigen. Mol. Cell. Biol. 17:4979–4990. Sullivan, C. S., P. Cantalupo, and J. M. Pipas. 2000. The molecular chaperone activity of simian virus 40 large T antigen is required to disrupt Rb-E2F family complexes by an ATP-dependent mechanism. Mol. Cell. Biol. 20:6233–6243. Tevethia, M. J., R. H. Bonneau, J. W. Griffith, and L. Mylin. 1997. A simian virus 40 large T-antigen segment containing amino acids 1 to 127 and expressed under the control of the rat elastase-1 promoter produces pancreatic acinar carcinomas in transgenic mice. J. Virol. 71:8157–8166. Walter, G., R. Ruediger, C. Slaughter, and M. Mumby. 1990. Association of protein phosphatase 2A with polyoma virus medium tumor antigen. Proc. Natl. Acad. Sci. USA 87:2521–2525. Wells, J., K. E. Boyd, C. J. Fry, S. M. Bartley, and P. J. Farnham. 2000. Target gene specificity of E2F and pocket protein family members in living cells. Mol. Cell. Biol. 20:5797–5807. Xiao, A., H. Wu, P. P. Pandolfi, D. N. Louis, and T. Van Dyke. 2002. Astrocyte inactivation of the pRb pathway predisposes mice to malignant astrocytoma development that is accelerated by PTEN mutation. Cancer Cell 1:157–168. Yamasaki, L., T. Jacks, R. Bronson, E. Goillot, E. Harlow, and N. J. Dyson. 1996. Tumor induction and tissue atrophy in mice lacking E2F-1. Cell 85: 537–548. Yang, H. S., and P. W. Hinds. 2007. pRb-mediated control of epithelial cell proliferation and Indian hedgehog expression in mouse intestinal development. BMC Dev. Biol. 7:6. Zhu, J. W., S. J. Field, L. Gore, M. Thompson, H. Yang, Y. Fujiwara, R. D. Cardiff, M. Greenberg, S. H. Orkin, and J. DeGregori. 2001. E2F1 and E2F2 determine thresholds for antigen-induced T-cell proliferation and suppress tumorigenesis. Mol. Cell. Biol. 21:8547–8564.