Transforming functions of Simian Virus 40 - Nature

Oncogene (2001) 20, 7899 ± 7907 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc

Transforming functions of Simian Virus 40 M Teresa SaÂenz-Robles1, Chris S Sullivan1 and James M Pipas*,1 1

Department of Biological Sciences. University of Pittsburgh, Pittsburgh, Pennsylvania, PA 15260, USA

Oncogene (2001) 20, 7899 ± 7907.

SV40 and the polyomaviruses

Keywords: SV40; tumor suppressor; chaperone; tumor virus

SV40 is a prototype member of the polyomavirus group (for a complete review, see: Fields, 1996), having a 45 nm diameter virion containing a single molecule of closed circular double-stranded DNA of 5243 bp. There are seven known virus-encoded proteins generated by dierential splicing of two transcription units (Figure 1). The viral T antigens, large T antigen (LT), small T antigen (ST) and 17K T antigen (17KT) are expressed soon after infection from a transcript initiating at the early promoter. Gene expression is accomplished using the cellular transcription apparatus and, in fact, the SV40 early promoter is used to drive expression of transgenes in a number of eucaryotic expression vectors. The functions of the viral T antigens are discussed below. The SV40 transcript initiated from the viral late promoter is dierentially spliced to result in mRNAs encoding the major capsid protein VP1, the minor capsid proteins VP2 and VP3, or the viral agno protein. VP1, VP2, and VP3 are the only virus-encoded proteins found in mature virions. The agno protein seems to function at some stage of virion assembly or virus exit from infected cells. All polyomaviruses share the same basic genome structure, a circular DNA with the T antigens and the capsid proteins expressed from separate transcription units, and an approximately 400 bp noncoding regulatory region containing the origin of viral DNA replication and the early and late promoters. All polyomaviruses encode LT, ST, VP1, VP2, and VP3. Murine polyomavirus and hamster polyomavirus both encode a middle T antigen (MT) as well. Like LT and ST, MT is produced by dierential splicing of the early region primary transcript. SV40 is one of 14 characterized polyomaviruses, of which 12 have been fully sequenced. These include two human viruses, JCV and BKV, and the well-characterized murine polyomavirus (PyV). Some of these viruses such as SV40 and BKV are rarely pathogenic in their natural hosts, while others such as BFDV, PyV, or murine K virus are deadly (Pipas, 1992). The host of SV40 is the Rhesus macaque and many of these animals harbor apparently harmless lifelong persistent infections of SV40. The kidney is thought to be the primary target organ of SV40 infection. The virus undergoes productive infection in terminally dierentiated epithelial cells. This presents a problem for the virus in that because of its small genome size it

Introduction Cancer is thought to progress through multiple stages with each stage giving rise to cells showing increased malignant characteristics. Progression through these stages is driven by genetic events such as mutations in tumor suppressor genes or oncogenes, or the acquisition of speci®c viral transforming sequences. These tumor suppressor genes and oncogenes encode proteins that are key components of regulatory pathways that govern genomic stability, cell proliferation, and apoptosis. Three key questions in cancer research are: (1) what speci®c genetic pathways, and which speci®c proteins within each pathway must be altered at each stage of tumor progression?; (2) how is cellular behavior changed by altering each speci®c pathway? and (3) are the cellular consequences the same for altering a given pathway in every cell type, or does a single pathway govern multiple cellular properties depending on the cell-type? DNA tumor viruses are important both because they provide us with powerful tools to address mechanisms of tumorigenesis, and because, in some cases, they contribute directly to cancer. Rather than alter cellular pathways by mutation, DNA tumor viruses encode dominant acting transforming proteins. Each of these transforming proteins target one or more key regulatory proteins within a cell, thus altering the cell's growth/survival properties. Because the viral transforming proteins must directly associate with their target, they serve as `proteo-divining rods,' guiding investigators through a milieu of cellular pathways and proteins irrelevant to the tumorigenic process and directly to the key players involved in malignant transformation. In addition, insights into how these key pathways are regulated can be gleaned from studies of the biochemical mechanisms by which the viral transforming proteins alter their target. Finally, cell culture and transgenic mouse systems have allowed studies of the role that each regulatory pathway plays in controlling cellular behavior. *Correspondence: JM Pipas; E-mail: [email protected]

Transforming functions of Simian Virus 40 MT SaÂenz-Robles et al

7900

Figure 1 Genomic Organization of SV40 virus (5243 bp). Arrows indicate the mRNAs encoding the capsid proteins VP1, VP2 and VP3 (found in mature virions) originated from the late promoter (PL), and the viral antigens large T antigen (LT), small T antigen (ST) and 17K T antigen (17KT) (involved in cellular transformation and originated from the early promoter PE). Origin of replication is indicated (Ori)

relies primarily on cellular proteins to eect viral DNA replication. Since the expression of most genes encoding replication-related proteins is restricted to late G1 or S phases of the cell cycle, SV40 must ®nd a way to activate the expression of these genes in growtharrested, dierentiated kidney cells. SV40 undergoes ecient productive infection in established cell lines such as CV1, BSC, or VERO, all of which were derived from the kidneys of African green monkeys. In these cell lines productive infection is complete about 96 hours post-infection, resulting in about 300 infectious progeny virions per cell. The program of SV40 productive infection initiates with attachment of a virion to a permissive cell and is followed by penetration, transport to the nucleus, and virion uncoating. Release of the viral chromatin is followed by activation of the early promoter by the cellular transcription machinery and by subsequent synthesis of LT, ST, and 17KT. This results in a change in the pattern of cellular gene expression. The expression of many genes associated with cellular DNA synthesis is enhanced and the cells are in fact driven through G1 and enter S phase. Shortly afterwards, viral DNA replication begins resulting in the production of progeny viral DNA. Concomitant with the initiation of viral DNA replication, the late promoter becomes active resulting in the expression of VP1, VP2, VP3, and agno protein. This results in the assembly of progeny virions and eventually in cell death. SV40 and cellular transformation Shortly after SV40 was discovered it was found to induce tumors when inoculated into newborn baby hamsters. This observation, coupled with the fact that it is relatively easy to grow in cell culture, catalyzed an impressive burst of activity in SV40 research. At about the same time, cell culture assays that mimic or measure aspects of oncogenic transformation were being developed. Although the variations are many, ®ve basic assays are used to assess cellular transformation: (1) growth in low serum; (2) growth to high saturation density; (3) focus formation; (4) anchorage independent growth and (5) tumorigenicity in animals. SV40 induces cellular transformation by all of these criteria. It is important to note that most of these studies have been performed on established rodent ®broblast cell lines. We discuss below more recent studies on SV40 transformation of multiple cell types. Oncogene

Why rodent cell lines? They were beautifully amenable to these studies because of a weird aspect of arcane virology. Rodent cells are nonpermissive for SV40 productive infection. That is, infection of established rodent ®broblast cell lines such as BalbC/3T3 with SV40 does not result in the production of progeny virions. Careful observation of such nonpermissive infections reveals a remarkable observation. SV40 attaches, penetrates, and uncoats more or less normally in these cell lines and the viral T antigens are expressed on schedule. However, neither viral DNA replication nor transcription from the late promoter occur. Thus, the infection is blocked after T antigen production but before components of the progeny virions are expressed. Like permissive monkey cells expressing the viral T antigen, the infected rodent cells are stimulated to progress through the G1/S transition and cellular DNA replication does occur. Infection of a rodent cell line at a high multiplicity-of-infection (moi) of SV40 results in all of the cells acquiring transformed characteristics. However, as cell proliferation proceeds the viral DNA driving T antigen expression is diluted so fewer and fewer daughter cells are transformed. Eventually the viral DNA is lost and the cell population regains the properties of normal, growthregulated cells. This phenomenon has been termed `abortive transformation.' When abortively transformed cell populations in culture are maintained continuously, a few `mounds' of multilayered, transformed cells overgrow the normal cell monolayer. These are termed foci and represent a true stable transformation event since cells picked and expanded from them continue to exhibit the transformed phenotype through many generations. All such foci-derived cells express the SV40 T antigens and all contain SV40 DNA integrated at a random point within the cellular DNA. In these cases, a portion of the viral DNA integrates into the host genome by nonhomologous recombination at some point in the abortive infection. If such an integration event leaves the viral early promoter and T antigen coding sequences intact, then daughter cells derived from this event will continuously express the T antigens. Since these cells are not growth-arrested by contact inhibition or reduced availability of serum, they continue to proliferate when the normal cells surrounding them have arrested and thus, they form the foci. The isolation of temperature-sensitive mutants in LT (tsA mutants) and deletion mutants that eliminate expression of ST allowed an assessment of the dierent


roles of these proteins in transformation. Foci derived from infections with tsA mutants display a fully transformed phenotype at the permissive temperature, usually 328C, however, they revert to a normal phenotype when maintained at the nonpermissive temperature (418C). Similarly, infection of rodent cells with tsA mutants at 328C leads to foci production, while infection at 418C fails to yield foci (Rassoulzadegan et al., 1978; Seif and Cuzin, 1977). Thus, T antigen function is necessary to initiate and to maintain the transformed phenotype. Some cell lines expressing tsA T antigens maintain the transformed phenotype even at the nonpermissive temperature. This phenomenon has not been fully explained. Studies using constructs that express either a wildtype LT and no ST, or ST and no LT revealed that ST is not sucient to transform cells but is sometimes necessary. In many circumstances, LT is necessary and sucient for transformation. However, in some cell types, or under certain conditions such as limited steady state-levels of LT or dense cell monolayers, both LT and ST are required for transformation (reviewed in Rundell and Parakati, 2001). In summary, a number of dierent types of studies have shown that the full tumorigenic potential of SV40 resides in both LT and ST.

SV40 encoded tumor antigens

7901

There are three tumor antigens encoded by SV40: large T antigen (T antigen), small t antigen (t antigen) and 17K t antigen (see Figure 2, for a complete review, see Fields, 1996). All are expressed early after infection from a common precursor mRNA that is dierentially spliced such that the three proteins share an amino terminal domain but contain dierent carboxy terminal regions. The common amino-terminal domain includes the ®rst 82 amino acids and is a J domain, a domain found in the DnaJ family of co-chaperone regulators of the Hsc70 class of molecular chaperones. T antigen has been shown to function as an authentic DnaJ chaperone both in vivo and in vitro (Campbell et al., 1997; Srinivasan et al., 1997; Sullivan et al., 2001; Zalvide et al., 1998). The 708 amino acid T antigen is a multifunctional protein that is required for several aspects of productive infection and viral tumorigenesis, including both the initiation and elongation steps of viral DNA replication. T antigen binds the viral origin of replication (ori) through a sequence-speci®c DNA binding domain, forming a double-hexamer and inducing structural changes in adjacent sequences. T antigen directly binds DNA polymerase a, DNA

ATPase

Figure 2 Domain map of SV40 tumor antigens. (a) Large T antigen; (b) small t antigen and, (c) 17K T antigen. Known and possible interactions with cellular components are indicated. Structures depicted were obtained from the following references: J domain of murine polyomavirus T antigen (Berjanskii et al., 2000, PDB ID: 1FAF); SV40 T antigen DNA binding domain (Luo et al., 1996, PDB ID: 1TBD); structure of a single zinc ®nger DNA-binding domain (Lee et al., 1989, PDB ID: 1ZNF) Oncogene


7902

Oncogene

primase, RPA, and topoisomerase, thus recruiting the cellular replication apparatus to the SV40 ori and eecting the initiation of viral DNA replication. Subsequently, each T antigen hexamer acts as a DNA helicase, a function evoking multiple domains including the ATP-binding/ATPase region. The role of T antigen in viral DNA replication is regulated by phosphorylation. Phosphorylation of the amino acid T124 is required for the ecient cooperative assembly of double hexamers at ori and T antigen that is not modi®ed at this site is defective for the initiation of replication. In contrast, phosphorylation of several serine residues antagonizes the formation of doublehexamers. T antigen is also a transcriptional regulatory protein. T antigen acts as a transcriptional repressor by binding to speci®c sequences near the viral early promoter thus inhibiting its own expression as well as that of t antigen and 17K t antigen. On the other hand, T antigen activates transcription by binding speci®c components of the cellular transcriptional machinery, including TBP and several transcription factors. Although the exact mechanisms remain to be elucidated, it is clear that these interactions are at least partly responsible for the activation of the SV40 late promoter and several cellular genes. T antigen also plays roles in virion assembly although the precise nature of its contribution is unclear (Spence and Pipas, 1994a,b). In addition to its multiple roles in productive infection, T antigen is also necessary and often sucient for inducing viral tumorigenesis. At least three dierent T antigen domains contribute to cellular transformation. These include a bipartite region within the carboxy-terminal half of the molecule that governs interaction with the cellular tumor suppressor protein, p53; a short domain containing an LXCXE motif responsible for binding the retinoblastoma (Rb) family of tumor suppressors; and, the J domain. The roles of each of these transforming regions in tumorigenesis are discussed below. The 174 aa small t antigen is not essential for productive infection in cell culture. However, t antigen does contribute to viral tumorigenesis, and in some circumstances, the cooperation of large T antigen and small t antigen is necessary for transformation. The unique carboxy-terminal domain of t antigen contains a Zn-®nger that binds the cellular phosphatase pp2A. This interaction is essential for transformation. The direct biochemical target of the t antigen J domain is unclear, although, like large T antigen, it can stimulate the ATPase activity of Hsc70 family members (Srinivasan et al., 1997). The t antigen J domain is not required for binding or action on pp2A but it is essential for t antigen-induced transcription of the cyclin A gene (Rundell and Parakati, 2001). Finally, the 17K t antigen contains the J domain, the adjacent LXCXE domain and nuclear localization signal (NLS), followed by four unique amino acids at the carboxy-terminus. The role of this protein in SV40 infection and transformation is unclear. However, other members of the Polyomavirus family

express similar proteins (Bollag et al., 2000; Riley et al., 1997). T antigens target key cellular regulatory proteins to eect transformation Genetic studies have clearly delineated three regions of T antigen required for transformation. The ®rst region, consisting of the ®rst 82 amino acids, is the J domain. The J domain governs the interaction of T antigen with its DnaK partner, hsc70. Deletion of the J domain results in a transformation-negative phenotype in a number of cultured cell-types as well as in transgenic mice. However, it is not clear if hsc70 is the cellular target relevant to transformation. Amino acid substitution mutations in the conserved HPD motif of the J domain are defective for interaction with hsc70 yet retain the ability to induce dense foci in primary and established cell lines. These mutants do show some transformation-related defects such as the inability to induce primary cells to grow to a high saturation density. One possibility is that hsc70 association is critical for transformation but that T antigen has redundant activities that aect the same cellular pathways targeted by hsc70. Another possibility is that some target other than hsc70 associates with the J domain and that, while deletions of the J domain eliminate action on this target, amino acid substitutions within the HPD motif are only partially defective for this interaction. The observation that constructs in which the SV40 J domain is replaced by the J domain from the E. coli DnaJ protein, or by the yeast protein Ydj1p J domain, are transformation-defective while retaining hsc70 interaction, suggest that additional cellular proteins interact with this region of T antigen (Sullivan et al., 2000b). T antigen binds to the retinoblastoma family of tumor suppressors through an LXCXE motif located between the J domain and the NLS. T antigen binds all three members of this family, pRb, p107, and p130, and in each case, can relieve cellular growth-arrest mediated by these proteins (reviewed in DeCaprio, 1999). One function of the Rb-family is to negatively regulate the transcriptional activation function of the E2F-family of proteins. Release of E2F from Rbmediated control results in the expression of S phase speci®c genes and drives cells into the cycle. In theory, this is necessary for viral productive infection in that SV40, for the most part, uses cellular proteins for viral DNA replication. T antigen mutants that cannot interact with the Rb-family are defective for transformation in most cell-types tested. Clearly, one role of T antigen in this interaction is to inhibit the Rb-protein's ability to interact with and to block E2F-mediated transcriptional activation. Binding of the Rb-proteins by T antigen is insucient to activate E2F. This requires both a functional LXCXE motif and a J domain (Harris et al., 1998; Sheng et al., 1997; Zalvide et al., 1998). One model is that T antigen ®rst binds to Rb/E2F


complexes through its LXCXE motif and then these complexes are dissolved using energy derived from hsc70-mediated ATP hydrolysis (Srinivasan et al., 1997). In fact, puri®ed T antigen has been shown to bind p130/E2F-4 and pRb/E2F-4 complexes in vitro and the dissolution of these complexes requires ATP hydrolysis by hsc70 (Sullivan et al., 2000a). This explains why the J domain is required in addition to the LXCXE motif to eect E2F transcriptional activation. In fact, the shortest fragment of T antigen that induces transformation is the ®rst 121 amino acids, consisting of the J domain followed by 39 amino acids, including the LXCXE motif. Nevertheless, the whole story is far from being completely understood. First, Tevethia et al. (1997a) have shown that a short peptide, including the LXCXE motif but missing the J domain, promotes cellular growth to high density when fused to the carboxy-terminus of T antigen. It is not yet clear whether the construct achieves this eect by disrupting Rb/E2F complexes or whether the LXCXE motif and/or immediately surrounding sequences are able to aect cell growth independent of action on Rb. Second, Kohrman and Imperiale (1992) have identi®ed a protein, termed p185, that binds to the ®rst 121 amino acids of T antigen, and whose binding to T antigen is independent of the J domain and the LXCXE motif. It is not clear if the T antigen/p185 interaction is required for transformation but this observation raises the possibility that the ®rst 121 amino acids of T antigen targets cellular proteins in addition to the Rb-family. Third, the amino-terminus of T antigen has been reported to bind a pro-apoptotic protein, p193 (Daud et al., 1993). p193 induces p53independent apoptosis and binding to T antigen blocks p193-dependent apoptosis (Pasumarthi et al., 2001; Tsai et al., 2000). Fourth, T antigen binds to the transcription factor Tst-1/Oct6/SCIP through its amino terminus and has been shown to synergistically enhance its transactivation activities in a J domaindependent manner (Sock et al., 1999). Again, the eects of T antigen's interactions with p193 or Tst-1/ Oct6/SCIP on its transforming ability have not been determined. Finally, mutations in the amino terminus of T antigen aect its interaction with the transcriptional adaptor protein p300 (Eckner et al., 1996). T antigen binding to p300 has been mapped to both the amino terminal and carboxy terminal portions of the T antigen, and at present it is not clear if this association is direct (Eckner et al., 1996; Lill et al., 1997). T antigen alters the phosphorylation state of p300 and additionally has been shown to down-regulate its transactivation-inducing abilities (Eckner et al., 1996). Again, the role of the LT/p300 interaction in transformation has not been clearly established. A number of cellular stresses, including the inappropriate activation of the E2F-family, activate the p53 tumor suppressor pathway. This leads to elevated steady-state levels of p53 and the expression of p53-dependent genes including p21, a universal cdKinhibitor. Thus, activation of p53 blocks the cell cycle. In addition, p53 induces the expression of pro-

apoptotic genes such as bax. DNA tumor viruses abrogate this response by blocking one or more of the p53 functions. SV40 T antigen regulates the function of p53 in at least two ways (Pipas and Levine, 2001). First T antigen directly binds to the DNA binding domain of p53, which results in the inactivation of its transcriptional activation activity. This interaction prevents the growth inhibitory induction of genes downstream of p53, such as the cyclin/cdk inhibitor p21. Secondly, T antigen regulates p53 independent of direct association. The mechanism by which this occurs is unclear but it requires both the J domain and the pRb binding motif of T antigen (Jiang et al., 1993; Quartin et al., 1994). There is also evidence for transforming function(s) in the carboxy terminus of T antigen in addition to the binding of p53. First, Dickmanns et al. (1994) have identi®ed a mitogenic activity aected by mutation of amino acid 189. This mutation does not aect LT binding to p53 but, on the other hand, this mutant is unable to bind to the transcription factor Tef-1, thus supporting a possible role for Tef-1/T antigen interaction in transformation. Second, Cavender et al. (1995) have demonstrated that deletion of the amino acid 400 yields a transformation-defective T antigen, although this mutant binds p53 normally. Finally, coexpression of the ®rst 136 amino acids of LT which inactivates the Rb-family with a dominant-negative allele of p53 does not elicit transformation, again suggesting the presence of T antigen transforming functions in addition to Rb and p53 inactivation (Sachsenmeier and Pipas, 2001).

7903

How does T antigen action on cellular targets contribute to tumorigenesis? Transgenic mouse models have been widely used to study gene expression under physiological conditions in vivo. With tissue-speci®c promoters, it is possible to target the expression of a protein exclusively to a speci®c tissue or to particular cell types. In this context, the DNA coding sequence of the SV40 T antigen has been placed under the control of dierent promoters, thus directing T antigen expression to selected organs, tissues or cell types. The expression of T antigen alters cellular growth control in a broad range of tissues, some of which are listed in Table 1. These experiments illustrate the importance of whole-animal assays in determining the molecular basis of transformation, since the eects of SV40 T antigen have been found to vary considerably in dierent tissues and organs. Furthermore, mutant analysis has revealed distinct roles for each T antigen activity in contributing to neoplasia. The eects of ectopic expression of T antigen varies widely depending on several factors, in particular the tissue where the protein is expressed, the speci®c promoter controlling that expression and the particular cell types involved in the transformation. Consequently, LT expression leads to hyperplasia and Oncogene


7904

Table 1 Proliferation and oncogenesis induced in murine tissues following transgenic expression of SV40 T antigen Tissue/organ

References

Adipose tissue Bone Brain and nervous system

Fox et al., 1989; Zennaro et al., 1998; Le Menuet et al., 2000 Behringer et al., 1988; Knowles et al., 1990; Jensen et al., 1993; Wilkie et al., 1994 Brinster et al., 1984; Reynolds et al., 1988; Chen et al., 1989; Theuring et al., 1990; Chen and Van Dyke, 1991; Perraud et al., 1992; Smith et al., 1992; Jensen et al., 1993; Chiu et al., 2000 Cheah et al., 1995 Li et al., 1995; Kim et al., 1994; Hauft et al., 1992; Asa et al., 1996 Theuring et al., 1990; Penna et al., 1998; Beermann et al., 1999 Behringer et al., 1988 Theuring et al., 1990 Dyer and Messing, 1989; Sandgren et al., 1989; Sepulveda et al., 1989; Butel et al., 1990; Manickan et al., 2001 Sandmoller et al., 1994, 1995; Wikenheiser et al., 1992 Reynolds et al., 1988; Chen et al., 1989 Husler et al., 1998; Green et al., 2000; Schulze-Garg et al., 2000 Hanahan 1985; Murphy et al., 1987; Ornitz et al., 1987; Efrat et al., 1988; Dyer and Messing, 1989; Bell et al., 1990; Ceci et al., 1991; Smith et al., 1992; Asa et al., 1996 Murphy et al., 1987 Garabedian et al., 1998; Green et al., 2000 Pascall et al., 1994; Sandmoller et al., 1994; Ewald et al., 1996 Wilkie et al., 1994; March et al., 1999; Herring et al., 1999 Ceci et al., 1991; Thompson et al., 2000

Cartilage Digestive system Eye Heart Kidney Liver Lung Lymphoid tissue Mammary gland Pancreas Pituitary Prostate Salivary gland Smooth muscle Stomach

carcinoma in some tissue types, but not in others (Lee et al., 1992, Lopez et al., 1995). For instance, SV40 T antigen has been expressed in several dierentiated cell subpopulations of the small intestine and, although all of them originate from common precursor stem cells, the consequences of this expression are extremely dierent. While hyperplasia and dysplasia of the small intestine result from the production of T antigen in enterocytic cells (Hauft et al., 1992; Kim et al., 1994), no apparent pathology is detected in the small bowel in transgenic mice expressing the T antigen in secretory cells (Asa et al., 1996; Lee et al., 1992, Lopez et al., 1995). Expression of T antigen in Paneth cells results in a loss of the mature cell type and leads to an ampli®cation of cells showing an intermediate morphology between Paneth and goblet cells (Garabedian et al., 1997). In contrast, goblet cells expressing T antigen enter S phase but fail to progress through M phase, instead undergoing apoptosis (Gum et al., 2001). Finally, some cell types seem refractory to transformation by T antigen in vivo, as its sole expression does not result in tumor development (Efrat et al., 1988; Reynolds et al., 1988). In cases where T antigen expression is followed by malignant transformation, a general correlation is observed between the production of higher levels of T antigen and the appearance of tumors, in terms of both severity and frequency (Husler et al., 1998; Kim et al., 1994; Van Dyke et al., 1987). In fact, through the use of conditional promoters (e.g. tetracyclineresponsive), it has been shown that continuous expression of LT is required to maintain cellular transformation, as a reversion of the malignant phenotype is observed if LT expression is suppressed (Ewald et al., 1996; Manickan et al., 2001). However, after prolonged LT expression these transformed cells lose their dependence on LT for the maintenance of the transformed state, and remain tumorigenic even after Oncogene

the viral product is no longer expressed (Ewald et al., 1996). As a general rule, cells induced to proliferate in response to LT expression seem to retain their dierentiation markers. For instance, expression of full length SV40 LT in postmitotic, villus-associated enterocytes causes them to reenter the cell cycle without an apparent eect on their state of dierentiation (Hauft et al., 1992) and a similar phenomenon is observed upon expression of LT in secretory cells of the colon (Asa et al., 1996; Lopez et al., 1995), pancreas (Lopez et al., 1995), retinal (Penna et al., 1998) and smooth muscle cells (March et al., 1999; Herring et al., 1999). However, some exceptions have been noticed: in adult animals where brain cells were forced to express LT under the control of the FGF1 promoter, tumors were produced which lacked terminal dierentiation markers for astrocytes or neurons, but expressed high levels of markers for proliferating cells. This phenotype is consistent with the tumor being at an early stage of dierentiation (Chiu et al., 2000). Similarly, pancreatic endocrine cells expressing T antigen only show some low levels of particular dierentiation markers, resembling an intermediate state of dierentiation (Asa et al., 1996). In most cases tested, the amino terminal region of LT (comprising the J domain and Rb binding motif) is necessary and sucient to induce tumors in various murine tissues, in a manner that absolutely requires an intact Rb-interaction domain (Bennoun et al., 1998; Chandrasekaran et al., 1996; Chen et al., 1992; Kim et al., 1994; SaÂenz Robles et al., 1994). The role of the J domain has been less studied in transgenic models, and the reports so far indicate an important -although tissue-speci®c- dependency on the J domain to induce tumors (Chen et al., 1992; Ratineau et al., 2000; Symonds et al., 1993). In each of these cases, the amino-terminal fragment of LT is sucient to drive


dierentiated cells to reenter the cell-cycle resulting in responses varying from hyperplasia to frank tumorigenesis. In contrast, the importance of the contribution of p53 inactivation by T antigen to tumorigenesis varies depending on the cell-type examined. Some tissue types require a functional T antigen/p53 binding domain for oncogenic transformation (McCarthy et al., 1994; SaÂenz-Robles et al., 1994), while in other cases the binding of T antigen to p53 is not required for induction of tumors (Bennoun et al., 1998; Chen et al., 1992; Tevethia et al., 1997b). In the case of choroid plexus epithelium, the J domain coupled with the Rb-binding motif induces hyperplasia as eciently as full length T antigen. However, while tissue expressing wild-type T antigen rapidly expands and kills the animals in a few weeks, the overall expansion of the hyperplastic choroid plexus expressing a truncated T antigen missing the p53 interaction domain is much slower. Inactivation of p53 by T antigen binding, the genetic removal of p53 or the interference with the normal p53 function by expressing a p53-dominant negative mutant causes an accelerated tumor growth which correlates with a decrease in the apoptotic level (Bowman et al., 1996; SaÂenz-Robles et al., 1994; Symonds et al., 1994). On the other hand, T antigen-induced carcinogenesis of pancreatic beta cells is reduced when the expression of p53 is abolished in this tissue, perhaps in response to the lower levels of T antigen found in cells lacking p53 compared to wild-type tumor cells (Herzig et al., 1999). Finally, the loss of p53 function caused by interaction with T antigen only has only a moderate eect on hepatic tumor formation, while it remains completely dependent on the disruption of the Rb pathway (Bennoun et al., 1998). Similarly, the presence or absence of p53 do not seem to aect the T antigen-induced transformation of the intestinal epithelium (Coopersmith et al., 1997; Kim et al., 1993). In this case, one possibility is that the eects of T antigen inhibition of p53-independent apoptosis contribute to tumor progression (Coopersmith et al., 1997; Kim et al., 1993; Li et al., 1995). One complication in trying to determine the roles of TAg in tumorigenesis is that carboxy terminal truncations of TAg produce very unstable products, and attempts to generate transgenic mice expressing only the p53-binding domain of TAg have been dicult to pursue. Nevertheless, only the production of such transgenic mice, either alone or in combination with other mutant TAg transgenics, will indeed address the role of the interaction between TAg and p53 in tumorigenesis, and the speci®city of the pathway in the particular cell types under study. In summary, SV40 T antigen-mediated induction of tumorigenesis in transgenic mice requires an intact Rbbinding function (probably in combination with a functional J domain) to commit cells to further proliferation and exit from the resting state. Further progression towards a more malignant phenotype sometimes requires inactivation of p53, or some other

T antigen function located in the carboxy terminal portion of the molecule. Most of the transgenic constructs driving expression of large T also encode the small t antigen from SV40. In most of these cases, the possible role of small t antigen in cooperating with large T antigen or in producing tumors by itself has not been addressed. One recent report has evaluated the possible role of small t antigen in inducing proliferation of several secretory cell types (Ratineau et al., 2000). This study revealed that small t antigen is not necessary to produce insulinomas or lymphomas, but that a functional cooperation between intact large T and small t oncoproteins might be required to produce neoplasms of intestinal secretory cells. Clearly, this is an area that requires more intensive study. Many animal models expressing T antigen in mouse, hamsters and rats closely resemble speci®c human cancers (e.g. Green et al., 2000 for mammary gland; Thompson et al., 2000 for stomach) or particular aspects of the disease, such as hormone dependency and response (e.g. androgen dependency of prostate carcinomas, Asamoto et al., 2001). These models are therefore incredibly useful tools to ascertain molecular markers as indicators of tumor progression, as well as to test preventive or corrective therapies, such as the use of chemotherapy or cytokines (e.g. Baratin et al., 2001 gamma interferon treatment of hepatomas). Although the consequences of T antigen expression are normally restricted to the tissues in which it is expressed, there have been reports of eects aecting other organs and/or cell-types. For instance, a generalized peripheral neuropathy was observed in transgenic mice even though the SV40 early genes were expressed only in pancreas and liver, but not in the peripheral or central nervous systems (Dyer and Messing, 1989). In this case the secondary consequences of T antigen in the nervous system could be related to alterations in the circulating glucose levels resulting from the direct damage to the islet cells. Similarly, expression of T antigen in neuroendocrine cells of the prostate leads to the development of prostatic neoplasia which progresses rapidly to local invasion and metastases (Garabedian et al., 1998). In this case, the endocrine cells from the prostate secrete a variety of growth factors that may aect development and maintenance of this tissue, and their alteration by tumorigenesis could aect other contiguous and/or distant tissues. Other cases have been reported where expression of T antigen results in apparent metastases to other tissues (Asa et al., 1996; Fox et al., 1989; Garabedian et al., 1998). However, the expression of T antigen in these putative metastatic nodes was not evaluated. Thus, it remains unclear whether these distant tumors represent true clonal descendents of the original neoplasia, or if they are the consequence of secondary events derived from the eects of T antigen in cells from the primary tumor. Thus, T antigen expression can result in both cell autonomous and non-autonomous eects. The non-autonomous eects could result either from an expansion of a cell-type that naturally produces

7905

Oncogene


7906

hormones or growth-factors that act on distant tissues and/or alter systemic metabolism, or from the T antigen-induced expression of such factors in cells where they are normally not transcribed. Summary SV40 has provided investigators with very powerful tools for probing and understanding molecular mechanisms contributing to tumorigenesis. The large and small T antigens of SV40 exert their eects by binding to and by altering the activity of key cellular targets. Thus, small t antigen targeting of phosphatase pp2A, and large T antigen action on

the Rb-family and on p53 clearly contribute to neoplasia. However, it is also clear that both of these molecules possess additional, as yet uncharacterized functions that contribute to cellular transformation. Among the challenges facing investigators studying these proteins is to: (1) identify the additional cellular targets relevant to transformation; (2) elucidate the biochemical basis for large and small T antigen action on these targets, and ®nally (3) understand how altering each of these targets changes cellular behavior and contributes to the tumorigenic phenotype. Thus, the DNA tumor viruses continue to play a central role in uncovering the mysteries of tumorigenesis, and SV40 continues to be an important contributor to these eorts.

References Asa SL, Lee YC and Drucker DJ. (1996). Virchows Arch., 427, 595 ± 606. Asamoto M, Hokaiwado N, Cho YM, Takahashi S, Ikeda Y, Imaida K and Shirai T. (2001). Cancer Res., 61, 4693 ± 4700. Baratin M, Ziol M, Romieu R, Kayibanda M, Gouilleux F, Briand P, Leroy P, Haddada H, Renia L, Viguier M and Guillet JG. (2001). Cancer Gene Ther., 8, 193 ± 202. Beermann F, Hunziker A and Foletti A. (1999). Pigment. Cell Res., 12, 71 ± 80. Behringer RR, Peschon JJ, Messing A, Gartside CL, Hauschka SD, Palmiter RD and Brinster RL. (1988). Proc. Natl. Acad. Sci. USA, 85, 2648 ± 2652. Bell Jr RH, Memoli VA and Longnecker DS. (1990). Carcinogenesis, 11, 1393 ± 1398. Bennoun M, Grimber G, Couton D, Seye A, Molina T, Briand P and Joulin V. (1998). Oncogene, 17, 1253 ± 1259. Berjanskii MV, Riley MI, Xie A, Semenchenko V, Folk WR and Van Doren SR. (2000). J. Biol. Chem., 275, 36094 ± 36103. Bollag B, Prins C, Snyder EL and Frisque RJ. (2000). Virology, 274, 165 ± 178. Bowman T, Symonds H, Gu L, Yin C, Oren M and Van Dyke T. (1996). Genes Dev., 10, 826 ± 835. Brinster RL, Chen HY, Messing A, van Dyke T, Levine AJ and Palmiter RD. (1984). Cell, 37, 367 ± 379. Butel JS, Sepulveda AR, Finegold MJ and Woo SL. (1990). Intervirology, 31, 85 ± 100. Campbell KS, Mullane KP, Aksoy IA, Stubdal H, Zalvide J, Pipas JM, Silver PA, Roberts TM, Schahausen BS and DeCaprio JA. (1997). Genes Dev., 11, 1098 ± 1110. Cavenden JF, Conn A, Eply M and Terethia, MJ (1995). J. Virol., 69 923 ± 934. Ceci JD, Kovatch RM, Swing DA, Jones JM, Snow CM, Rosenberg MP, Jenkins NA, Copeland NG and Meisler MH. (1991). Oncogene, 6, 323 ± 332. Chandrasekaran C, Coopersmith CM and Gordon JI. (1996). J. Biol. Chem., 271, 28414 ± 28421. Cheah KS, Levy A, Trainor PA, Wai AW, Kuner T, So CL, Leung KK, Lovell-Badge RH and Tam PP. (1995). J. Cell. Biol., 128, 223 ± 237. Chen JD, Neilson K and Van Dyke T. (1989). J. Virol., 63, 2204 ± 2214. Chen JD and Van Dyke T. (1991). Mol. Cell. Biol., 11, 5968 ± 5976.

Oncogene

Chen J, Tobin GJ, Pipas JM and Van Dyke T. (1992). Oncogene, 7, 1167 ± 1175. Chiu IM, Touhalisky K, Liu Y, Yates A and Frostholm A. (2000). Oncogene, 19, 6229 ± 6239. Coopersmith CM, Chandrasekaran C, McNevin MS and Gordon JI. (1997). J. Cell. Biol., 138, 167 ± 179. Daud AI, Lanson Jr NA, Claycomb WC and Field LJ. (1993). Am. J. Physiol., 264, H1693 ± H1700. DeCaprio JA. (1999). Biologicals, 27, 23 ± 28. Dickmanns A, Zeitvogel A, Simmersbach F, Weber R, Arthur AK, Dehde S, Wildeman AG and Fanning E. (1994). J. Virol., 68, 5496 ± 5508. Dyer KR and Messing A. (1989). Am. J. Pathol., 135, 401 ± 410. Eckner R, Ludlow JW, Lill NL, Oldread E, Arany Z, Modjtahedi N, DeCaprio JA, Livingston DM and Morgan JA. (1996). Mol. Cell. Biol., 16, 3454 ± 3464. Harris KF, Christensen JB, Radany EH and Imperiale MJ. (1998). Mol. Cell. Biol., 18, 1746 ± 1756. Efrat S, Teitelman G, Anwar M, Ruggiero D and Hanahan D. (1988). Neuron, 1, 605 ± 613. Ewald D, Li M, Efrat S, Auer G, Wall RJ, Furth PA and Hennighausen L. (1996). Science, 273, 1384 ± 1386. Fields BN. (1996). Virology, third edition. Lippincott-Raven Publishers. Philadelphia-New York. Fox N, Crooke R, Hwang LH, Schibler U, Knowles BB and Solter D. (1989). Science, 244, 460 ± 463. Garabedian EM, Roberts LJ, McNevin MS and Gordon JI. (1997). J. Biol. Chem., 272, 23729 ± 23740. Garabedian EM, Humphrey PA and Gordon JI. (1998). Proc. Natl. Acad. Sci. USA, 95, 15382 ± 15387. Green JE, Shibata MA, Yoshidome K, Liu ML, Jorcyk C, Anver MR, Wigginton J, Wiltrout R, Shibata E, Kaczmarczyk S, Wang W, Liu ZY, Calvo A and Couldrey C. (2000). Oncogene, 19, 1020 ± 1027. Gum Jr JR, Hicks JW, Gillespie AM, Rius JL, Treseler PA, Kogan SC, Carlson EJ, Epstein CJ and Kim YS. (2001). Cancer Res., 61, 3472 ± 3479. Hanahan D. (1985). Nature, 315, 115 ± 122. Hauft SM, Kim SH, Schmidt GH, Pease S, Rees S, Harris S, Roth KA, Hansbrough JR, Cohn SM and Ahnen DJ, Wright NA, Goodland RA and Gordon JI. (1992). J. Cell. Biol., 117, 825 ± 539. Herring BP, Hoggatt AM, Smith AF and Gallagher PJ. (1999). J. Biol. Chem., 274, 17725 ± 17732.


Herzig M, Novatchkova M and Christofori G. (1999). Biol. Chem., 380, 203 ± 211. Husler MR, Kotopoulis KA, Sundberg JP, Tennent BJ, Kunig SV and Knowles BB. (1998). Transgenic Res., 7, 253 ± 263. Jensen NA, Rodriguez ML, Garvey JS, Miller CA and Hood L. (1993). Proc. Natl. Acad. Sci. USA, 90, 3192 ± 3196. Jiang D, Srinivasan A, Lozano G and Robbins PD. (1993). Oncogene, 8, 2805 ± 2812. Kohrman DC and Imperiale MJ. (1992). J. Virol., 66, 1752 ± 1760. Kim SH, Roth KA, Moser AR and Gordon JI. (1993). J. Cell. Biol., 123, 877 ± 893. Kim SH, Roth KA, Coopersmith CM, Pipas JM and Gordon JI. (1994). Proc. Natl. Acad. Sci. USA, 91, 6914 ± 6918. Knowles BB, McCarrick J, Fox N, Solter D and Damjanov I. (1990). Am. J. Pathol., 137, 259 ± 262. Le Menuet D, Zennaro MC, Viengchareun S and Lombes M. (2000). Kidney Int., 57, 1299 ± 1306. Lee MS, Gippert GP, Soman KV, Case DA and Wright PE. (1989). Science, 245, 635 ± 637. Lee YC, Asa SL and Drucker DJ. (1992). J. Biol. Chem., 267, 10705 ± 10708. Li Q, Karam SM and Gordon JI. (1995). J. Biol. Chem., 270, 15777 ± 15788. Lill NL, Tevethia MJ, Eckner R, Livingston DM and Modjtahedi N. (1997). J. Virol., 71, 129 ± 137. Lopez MJ, Upchurch BH, Rindi G and Leiter AB. (1995). J. Biol. Chem., 270, 885 ± 891. Luo X, Sanford DG, Bullock PA and Bachovchin WW. (1996). Nat. Struct. Biol., 3, 1034 ± 1039. Manickan E, Satoi J, Wang TC and Liang TJ. (2001). J. Biol. Chem., 276, 13989 ± 13994. March KL, Sandusky G and Fan L. (1999). Oncogene, 18, 3773 ± 3782. McCarthy SA, Symonds HS and Van Dyke T. (1994). Proc. Natl. Acad. Sci. USA, 91, 3979 ± 3983. Murphy D, Bishop A, Rindi G, Murphy MN, Stamp GW, Hanson J, Polak JM and Hogan B. (1987). Am. J. Pathol., 129, 552 ± 566. Ornitz DM, Hammer RE, Messing A, Palmiter RD and Brinster RL. (1987). Science, 238, 188 ± 193. Pascall IC, Surani MA, Barton SC, Vaughan TJ and Brown KD. (1994). J. Mol. Endocrinol., 12, 313 ± 325. Pasumarthi KB, Tsai SC and Field LJ. (2001). Circ. Res., 88, 1004 ± 1011. Penna D, Schmidt A and Beermann F. (1998). Oncogene, 17, 2601 ± 2607. Perraud F, Yoshimura K, Louis B, Dalemans W, Ali-Hadji D, Schultz H, Claudepierre MC, Chartier C, Danel C and Bellocq JP, et al. (1992). Oncogene, 7, 993 ± 997. Pipas JM. (1992). J. Virol., 66, 3979 ± 3985. Pipas JM and Levine AJ. (2001). Semin. Cancer Biol., 11, 23 ± 30. Quartin RS, Cole CN, Pipas JM and Levine AJ. (1994). J. Virol., 68, 1334 ± 1341. Rassoulzadegan M, Perbal B and Cuzin F. (1978). J. Virol., 28, 1 ± 5. Ratineau C, Ronco A and Leiter AB. (2000). Gastroenterology, 119, 1305 ± 1311. Reynolds RK, Hoekzema GS, Vogel J, Hinrichs SH and Jay G. (1988). Proc. Natl. Acad. Sci. USA, 85, 3135 ± 3139. Riley MI, Yoo W, Mda NY and Folk WR. (1997). J. Virol., 71, 6068 ± 6074.

Rundell K and Parakati R. (2001). Semin. Cancer Biol., 11, 5 ± 13. Sachsenmeier KF and Pipas JM (2001). Virology, 283, 40 ± 48 SaÂenz-Robles MT, Symonds H, Chen J and Van Dyke T. (1994). Mol. Cell. Biol., 14, 2686 ± 2898. Sandmoller A, Halter R, Gomez-La-Hoz E, Grone HJ, Suske G, Paul D and Beato M. (1994). Oncogene, 9, 2805 ± 2815. Sandmoller A, Halter R, Suske G, Paul D and Beato M. (1995). Cell. Growth. Dier., 6, 97 ± 103. Sandgren EP, Quaife CJ, Pinkert CA, Palmiter RD and Brinster RL. (1989). Oncogene, 4, 715 ± 724. Schulze-Garg C, Lohler J, Gocht A and Deppert W. (2000). Oncogene, 19, 1028 ± 1037. Seif R and Cuzin F. (1977). J. Virol., 24, 721 ± 728. Sepulveda AR, Finegold MJ, Smith B, Slagle BL, DeMayo JL, Shen RF, Woo SL and Butel JS. (1989). Cancer Res., 49, 6108 ± 6117. Sheng Q, Denis D, Ratnofsky M, Roberts TM, DeCaprio JA and Schahausen B. (1997). J. Virol., 71, 9410 ± 9416. Smith MS, Lechago J, Wines DR, MacDonald RJ and Hammer RE. (1992). DNA Cell. Biol., 11, 345 ± 358. Sock E, Enderich J and Wegner M. (1999). Mol. Cell. Biol., 19, 2455 ± 2464. Spence SL and Pipas JM. (1994a). J. Virol., 68, 4227 ± 4240. Spence SL and Pipas JM. (1994b). Virology, 204, 200 ± 209. Srinivasan A, McClellan AJ, Vartikar J, Marks I, Cantalupo P, Li Y, Whyte P, Rundell K, Brodsky JL and Pipas JM. (1997). Mol. Cell. Biol., 17, 4761 ± 4773. Sullivan CS, Cantalupo P and Pipas JM. (2000a). Mol. Cell. Biol., 20, 6233 ± 6243. Sullivan CS, Gilbert SP and Pipas JM. (2001). J. Virol., 75, 1601 ± 1610. Sullivan CS, Tremblay JD, Fewell SW, Lewis JA, Brodsky JL and Pipas JM. (2000b). Mol. Cell. Biol., 20, 5749 ± 5757. Symonds HS, McCarthy SA, Chen J, Pipas JM and Van Dyke T. (1993). Mol. Cell. Biol., 13, 3255 ± 3265. Symonds H, Krall L, Remington L, Saenz-Robles M, Lowe S, Jacks T and Van Dyke T. (1994). Cell., 78, 703 ± 711. Tevethia MJ, Lacko HA, Kierstead TD and Thompson DL. (1997a). J. Virol., 71, 1888 ± 1896. Tevethia MJ, Bonneau RH, Grith JW and Mylin L. (1997b). J. Virol., 71, 8157 ± 8166. Theuring F, Gotz W, Balling R, Korf HW, Schulze F, Herken R and Gruss P. (1990). Oncogene, 5, 225 ± 232. Thompson J, Epting T, Schwarzkopf G, Singhofen A, EadesPerner AM, van Der Putten H and Zimmermann W. (2000). Int. J. Cancer, 86, 863 ± 869. Tsai SC, Pasumarthi KB, Pajak L, Franklin M, Patton B, Wang H, Henzel WJ, Stults JT and Field LJ. (2000). J. Biol. Chem., 275, 3239 ± 3246. Van Dyke TA, Finlay C, Miller D, Marks J, Lozano G and Levine AJ. (1987). J. Virol., 61, 2029 ± 2032. Wikenheiser KA, Clark JC, Linnoila RI, Stahlman MT and Whitsett JA. (1992). Cancer Res., 52, 5342 ± 5352. Wilkie TM, Schmidt RA, Baetscher M and Messing A. (1994). Oncogene, 9, 2889 ± 2895. Zalvide J, Stubdal H and DeCaprio JA. (1998). Mol. Cell. Biol., 18, 1408 ± 1415. Zennaro MC, Le Menuet D, Viengchareun S, Walker F, Ricquier D and Lombes M. (1998). J. Clin. Invest., 101, 1254 ± 1260.

7907

Oncogene