Emerging roles of DNA tumor viruses in cell proliferation: new insights into genomic instability


The small DNA virus proteins E1A and E1B from human Adenovirus, E6 and E7 from human papillomavirus, and large T and small T antigens from SV40, are multifaceted molecular tools that can carry out an impressive number of tasks in the host cell. These viral factors, collectively termed ‘oncoproteins’ for their ability to induce cancer, can be viewed as paradigmatic oncogenic factors which can disrupt checkpoint controls at multiple levels – they interfere with both ‘gatekeeper’ cellular functions, including major control pathways of cell cycle and apoptosis, and with ‘caretaker’ functions, thereby inducing mitotic abnormalities and increasing genomic instability. Both E1A and E7 have been recently found to interact physically with the Ran GTPase. This interaction is key in uncoupling the centrosome cycle from the cell cycle, highlighting a direct link between viral infection and the induction of genomic instability. Further expanding our current knowledge in this field will be crucial to elucidate viral strategies leading to cellular transformation and cancer progression, as well as design novel preventive or therapeutic approaches to human cancer.


Mammalian cells accumulate mutations progressively during their lifespan. This can result in loss of control of major biological processes, including cell life and death; in such a scenario, cancer is essentially depicted as a genetic disease in which the progressive and stepwise acquisition of mutational events evolves in cellular transformation and cancer progression (Kinzler and Vogelstein, 1996).

In the last two decades, the ‘viral hypothesis’ of cancer has received a considerable renewed impulse from the notion that a number of viral proteins, henceforth called oncoproteins, from small DNA viruses, such as E1A and E1B from human adenovirus, E6 and E7 from human papillomavirus (HPV), and large T and small T antigens (Ag's) from SV40, can physically interact with major cellular regulators and alter their function, thereby profoundly interfering with complex cellular processes such as gene expression, cell growth and differentiation. Viral oncoproteins act as effective molecular tools that have evolved to perform very specific tasks. Among known effects of oncoproteins in the host cell, their ability to interact physically and neutralize major tumor suppressor gene products, such as those belonging to the RB and p53 families, is of the highest importance (Paggi et al., 1996; Kaelin, 1999; Thomas et al., 1999). Owing to these features, viral oncoproteins have been widely employed as invaluable experimental tools to identify several key families of regulators, particularly of cell cycle homeostasis (Ali and DeCaprio, 2001; Munger et al., 2001; Sang et al., 2002). For example, members of the retinoblastoma family of tumor suppressor proteins, and other major regulators including Cyclin A, p300 and others, were initially identified via their interaction with E1A (Whyte et al., 1988).

Adenovirus E1A is the most widely investigated small DNA tumor virus oncoprotein. An exhaustive database on Adenovirus 5 E1A, maintained by Dr JS Mymryck (The University of Western Ontario, London, Ontario, Canada), is freely available on the Internet (http://www.geocities.com/jmymryk.geo/). This database contains information on general data on E1A, E1A mutations and their phenotypes, cellular proteins targeted by E1A and much more. In this review, we will mainly focus on E1A, but the related proteins HPV E7 and SV40 large T Ag will also be referred to in the light of cellular processes in which they exert paradigmatic pivotal roles. Considerable interest is raised by the E1A-related E7 protein from transforming HPV strains, which is considered to be a major effector of HPV oncogenicity in humans (Gage et al., 1990; Zwerschke and Jansen-Dürr, 2000). HPV 16 and 18 are defined as ‘high-risk’ viral strains identified as causative agents of at least 90% of cancers of the cervix and are also linked to more than 50% of other anogenital cancers (zur Hausen, 1996). As far as large T Ag is concerned, in spite of the fact that SV40 usually does not infect humans, accidental injection of SV40 has been reported in people taking IPV (the injected form of the polio vaccine) between 1955 and 1963. The issue of the potential role of SV40 in human cancer has been recently rebrought to attention due to its possible involvement in human mesothelioma (Carbone et al., 1994; Carbone, 2000). On the other hand, long-term epidemiology studies have failed to detect an increased cancer risk in those likely to have been exposed to the virus mainly by SV40-contaminated polio vaccines (Mortimer et al., 1981; Strickler et al., 1998). A report recently issued by the Institute of Medicine (IOM) concluded that the evidence is insufficient to prove or disprove the theory that exposure to SV40-contaminated poliovirus vaccine resulted in cancer in humans (http://www.iom.edu/iom/iomhome.nsf/pages/Sv40+Report?OpenDocument).

How do viral oncoproteins induce cancer?

As initially stated by Cairns (1975) and more recently revisited (Kinzler and Vogelstein, 1997; Frank, 2003), two major classes of mutations, compromising ‘gatekeeper’ and ‘caretaker’ activities, are involved in cancer initiation or progression. Gatekeeper mutations accelerate the cell proliferation rate by inducing mutations in the basal cell cycle machinery, either through oncogene activation or through induction of antioncogene. Caretaker mutations, instead, result in an increase in the overall rate of mutation, involving single loss-of-function loci or whole chromosomes, and hence increase genomic instability.

Some viral oncoproteins can be viewed as paradigmatic oncogenic factors that can recapitulate both types of cancer pathways, in that they can interfere with both gatekeeper and caretaker functions. Viral interference with gatekeeper functions yields loss of control of the mechanisms governing cell proliferation and apoptosis, and is essentially exerted by disrupting RB-dependent pathways. More recently, a role of E1A and E7 in the induction of genomic instability is also emerging. In this review, we will first recall the most important concepts in deregulation of cell proliferation by oncogenic viruses, and will then concentrate on their newly discovered function(s) in genomic instability. Finally, we will discuss new perspectives that may be pursued in devising novel therapeutic approaches to virus-induced cell transformation and cancer.

A full understanding of the mechanisms through which these processes are disrupted requires the identification of cellular factors that are targeted by oncoproteins. Recently, development of genome-wide approaches, such as the two-hybrid assay of representative human cDNA libraries in yeast (De Luca et al., 2003), and microarray-based analysis (Alazawi et al., 2002; Berger et al., 2002; Vasseur et al., 2003), is expected to yield novel information on functional interactions that may otherwise be missed.

E1A in disruption of ‘gatekeeper’ functions: proproliferative and apoptotic functions

Adenovirus E1A includes a set of viral oncoproteins able to induce a dramatic reprogramming of gene expression in the host cell (Gallimore and Turnell, 2001). All of the E1A forms are synthesized immediately after infection. Of those, two forms, 289R and 243R, also known as 13S and 12S E1A, respectively, are the most abundant products (Sang et al., 2002). Numerous, sometimes opposing, functions have been attributed to E1A, such as induction of cellular proliferation and transformation, inhibition of differentiation (Shenk and Flint, 1991; Nevins, 1995), induction of apoptosis (Debbas and White, 1993; Teodoro et al., 1995; White, 2001) and even tumor suppression (Chinnadurai, 1992; Mymryk, 1996; Frisch, 2001; Frisch and Mymryk, 2002). These viral oncoproteins disrupt the coordination of cell cycle events mainly by modifying the activity of endogenous cellular factors. The first and most investigated task of E1A is the interaction with the RB family of gene products. The RB oncosuppressor gene is the founder of the RB family, which comprises two additional structurally and functionally related genes, namely p107 (Ewen et al., 1991; Zhu et al., 1993) and Rb2/p130 (Hannon et al., 1993; Li et al., 1993; Mayol et al., 1993); the latter is also thought to act as a bona fide tumor suppressor gene (Paggi and Giordano, 2001). All three RB family gene products act in the control of cell cycle progression by negatively regulating the transition between the G1 and S phases (Goodrich et al., 1991; Zhu et al., 1993; Claudio et al., 1994; Starostik et al., 1996). This task is largely performed by functionally inactivating endogenous transcription factors that normally ensure control of the gene expression program during the cell cycle. Among those, a prominent role is played by factors of the E2F family, which act in complementary, specific ways to promote entry into the S phase (Mulligan and Jacks, 1998; Lavia and Jansen-Dürr, 1999; Nevins, 2001). Repression of E2F-dependent transcription by the RB proteins is also mediated via their interaction with histone deacetylases (Brehm et al., 1998; Ferreira et al., 1998; Magnaghi-Jaulin et al., 1998; Stiegler et al., 1998a). In addition, the RB gene family regulates a wide spectrum of complex biological processes, including differentiation, embryonic development and apoptosis (Riley et al., 1994; Sidle et al., 1996; Herwig and Strauss, 1997; Stiegler et al., 1998b and references therein).

E1A, E7 and large T Ag oncoproteins share the LXCXE motif, through which they physically interact with all of the members of the RB family proteins (Dahiya et al., 2000), which in turn bind the viral factors through the ‘pocket’ region (Helt and Galloway, 2003). The ‘pocket’ name refers to the unique tridimensional structure characterizing the three members of the RB family, responsible for most of the specific and functionally relevant interactions in which these molecules are involved (Paggi et al., 1996). Binding of E1A, E7 or large T Ag to the pocket region of the RB proteins results in the physical displacement of physiologically interacting partners, regardless of the presence (e.g. HDAC1 and 2) or absence (e.g. E2F family) of an LXCXE motif. In this way, the viral oncoproteins strongly compromise the regulatory properties of the RB proteins, thereby allowing the cell to converge irreversibly to an anticipated, or unscheduled, progression to the S phase of the cell cycle.

The multifocal strategy used by DNA tumor viruses to infect/transform the host cell also provides the opportunity to neutralize p53, another key tumor suppressor gene. Actually, the p53 protein was first identified as a T-Ag-binding protein (Lane and Crawford, 1979; Linzer and Levine, 1979), and physically interacts with adenovirus E1B (Yew and Berk, 1992) and E6 from high-risk HPV strains (Scheffner et al., 1992). The E6/p53 interaction promotes p53 degradation via a ubiquitin-dependent pathway (Rapp and Chen, 1998), leading to the abrogation of p53-dependent control of growth arrest and apoptosis. E6-induced p53 degradation is regarded as the functional equivalent of an inactivating mutation of p53. As a matter of fact, the majority of human cervical cancers harbor wild-type p53 protein as revealed by sequence analysis, yet it is functionally neutralized by E6 (Thomas et al., 1999).

Somewhat unexpectedly in the light of the properties described thus far, the expression of E1A can also induce apoptosis in some mammalian cells (Mymryk et al., 1994), via both p53-mediated and -independent mechanisms (Teodoro et al., 1995). In addition, E1A confers a higher sensitivity to DNA-damaging agents to tumor cells, which is mediated by p53 (Sanchez-Prieto et al., 1996). In yeast and lower eukaryotes, which do not undergo apoptosis, E1A can strongly inhibit the cell cycle, thus producing slow-growing phenotypes. Recently, we have sought to identify cellular factors able to counteract E1A-dependent growth repression using a genetic screening based on a yeast growth rescue strategy. This led to the identification of human RACK1, a receptor for activated C kinase, as an E1A-antagonizing factor, which permits cell growth in the presence of E1A. Actually, RACK1 is also able to antagonize E1A-induced apoptosis in human cancer cells (Sang et al., 2001).

E1A in disruption of ‘caretaker’ functions: induction of mitotic abnormalities and genetic instability

Many, if not all, types of solid tumors develop genetic instability, frequently associated with gain or loss of an entire chromosome or group of chromosomes (chromosomal instability, CIN). Unraveling the underlying mechanisms of genetic instability has become a central theme in cancer research.

Genomic instability can arise in consequence of cell transformation, or, on the contrary, can favor it, for example by inducing the loss of specific tumor suppressor genes linked to the lost chromosome: this can confer a growth advantage, possibly giving rise to a clone of genetically imbalanced cells. In either scenario – consequence or cause – the appearance of clones of aneuploid cells can be diagnostic of the loss of checkpoint controls. These controls can normally trigger cell cycle arrest or elimination of cells harboring mitotic defects and prone to become genetically imbalanced. In contrast, cells that can proliferate with an abnormal chromosome complement are unable to detect aneuploidy as a pathological condition: therefore, their ability to trigger growth arrest or apoptosis in response to chemotherapeutic agents targeting the mitotic apparatus (i.e. taxanes or vinca alkaloids) can be expected to be less effective than in euploid cells.

Role of centrosomes in mitosis and genomic stability

Chromosome mis-segregation often reflects failure, or inaccuracy, in the activity of the mitotic apparatus. An ever-increasing importance of this process is attributed to the spindle-organizing functions of centrosomes (Pihan et al., 1998; Brinkley, 2001; Hinchcliffe and Sluder, 2001; D'Assoro et al., 2002; Nigg, 2002). In mammalian somatic cells, centrosomes are the major microtubule-organizing centers and form mitotic spindle poles. Structurally they consist of two orthogonal centrioles embedded in a ‘pericentriolar matrix’. The correct reproduction and structural organization of centrosomes are critically important for the assembly of a functional bipolar mitotic spindle. Since the pioneering observations by Boveri, over a century ago, that normal cells contain 1 or 2 centrosomes, different from tumor cells, which typically exhibit supernumerary centrosomes, abnormalities in the size (hypertrophy) or number (amplification) of centrosomes have been repeatedly noticed in transformed cells (reviewed by Balmain, 2001). Supernumerary centrosomes can underlie the formation of multipolar spindles that direct chromosome mis-segregation to more than two poles, hence facilitating the onset of aneuploidy (reviewed by Lingle and Salisbury, 2000; Brinkley, 2001; Nigg, 2002).

Reproduction of centrosomes must be strictly limited to once per cell cycle for the establishment of the spindle bipolarity. First, the two centrioles composing a centrosome undergo a partial separation, termed ‘disorientation’. During S phase, each parental centriole then acts as the template for duplication of a daughter centriole in a semiconservative manner (Figure 1). The activity of Cdk2 (Matsumoto et al., 1999), complexed to G1 Cyclins (Meraldi et al., 1999; Okuda et al., 2000) is crucial to this process. Cdk2 phosphorylates key factor(s) in the duplication of parental centrioles, including: (i) nucleophosmin (known as NPM/B23), which associates with unduplicated centrioles and is detached upon Cdk2-dependent phosphorylation, a requisite for the onset of duplication of parental centrioles (Okuda et al., 2000); (ii) Mps1, a protein kinase acting as a positive inducer of centrosome duplication (Fisk and Winey, 2001). Centrosome amplification occurs when the activity of G1 Cyclin/Cdk2 complexes is deregulated, for example in p21−/− (Mantel et al., 1999; Fisk and Winey, 2001) and p27−/− (Nakayama et al., 2000; Fisk and Winey, 2001) cells, or in the absence of efficient proteasome activity, under which conditions cyclin turnover is altered (Nakayama et al., 2000).

Figure 1

The centrosome cycle and major regulators. G1 Cyclin/Cdk 2 activity is the major link between centrosome duplication and the cell cycle machinery. The conditions (i.e. HU) or regulatory factors that can modulate this link are indicated (see text for details). Images depict the configuration of centrioles (visualized by centrin fluorescence, in green, and schematized below) in G1, after normal duplication (S) and after rounds of abnormal duplication (amplification)

One approach to study the link between the centrosome duplication cycle and the cell cycle has made use of drugs inhibiting DNA replication, such as hydroxyurea (HU). In certain cell lines, this uncouples the centrosome cycle from the cell cycle and centrosomes undergo reduplication. The RB protein, pRb plays a major role in the negative control of reduplication and can block it in HU-arrested cells (Meraldi et al., 1999). Overexpression of either Cyclin A or E2F1 can override the pRb-dependent reduplication block (Meraldi et al., 1999): thus, pRb may act through transcriptional repression of the Cyclin A gene, a transcriptional target of E2F factors (Lavia and Jansen-Dürr, 1999). Recent experiments, however, have shown that deregulation of the pRb pathway, by either overexpressing Cdk4 (hence, forcing pRb inactivation by phosphorylation) or inducing loss-of-function of the p16 kinase inhibitor (hence, preventing the inhibition of pRb phosphorylation), are not sufficient to disrupt control of centrosome duplication (Piboonniyom et al., 2003). In G2, a variety of regulatory and structural factors are recruited to centrosomes. Among those, prominent roles are played by two families of protein kinases, PLK and Aurora, implying that mitotic activities of centrosomes are largely dependent on phosphorylation. Of relevance, Aurora-A overexpression occurs in breast, ovarian, bladder and colorectal cancer (reviewed by Nigg, 2002; Schmeichel, 2002), often associated with centrosome amplification (Goepfert et al., 2002).

Mitotic regulators, including Cyclin B (Jackman et al., 2003) and the p34/Cdc2 kinase (Bailly et al., 1989; Pockwinse et al., 1997), are also recruited to centrosomes in G2/early M. By analogy with DNA replication, a level of control of centrosome homeostasis must rely on negatively acting factor(s) operating after duplication, that is in G2/M phases, such that newly duplicated centrosomes will not reduplicate again within the same cell cycle. Such factor(s) may be inactivated or eliminated after centrosome segregation to daughter cells, thus ensuring centrosome ‘licensing’ for a novel round of duplication in the next cell cycle. A licensing mechanism for centrosome duplication has been formally demonstrated in yeast, and is controlled by M-specific cyclins (Haase et al., 2001). In mammalian cells, the NPM/B23 protein is recruited to spindle poles in mitosis (Zatsepina et al., 1999) and may act in the inhibition of reduplication. Cdk2 drives phosphorylation-mediated detachment of NPM/B23 from the centrosome prior to duplication and may be a pivotal factor in licensing.

Finally, relevant to the mechanism(s) of genomic instability, major tumor suppressor proteins (p53, BRCA1, TACC) are recruited at centrosomes during mitosis (reviewed by Fisk et al., 2002; Raff, 2002). In the case of p53, the transient association with mitotic centrosomes can be viewed as a regulatory step in the control of genomic stability (Ciciarello et al., 2001). p53 localizes to centrosomes during a normal mitosis, which is associated with low intracellular levels of p53. When mitosis is abnormal, p53 is delocalized from centrosomes; this triggers p53 stabilization and accumulation, eventually dictating G1 arrest in the following cell cycle. Thus, p53 can monitor mitotic abnormalities through its ability to associate with mitotic centrosomes. When centrosomal defects occur in cells lacking functional p53, no cell cycle arrest follows (Borel et al., 2002). Thus, although p53 failure does not directly contribute to centrosome amplification, it is important in genomic instability, as it prevents the elimination of cells carrying centrosomal abnormalities and allows their clonal propagation (Carroll et al., 1999; Borel et al., 2002). Errors in cytokinesis are another source of centrosomal abnormalities in p53-defective cells, as abnormal centrosome numbers can mis-segregate to daughter cells after a transient tetraploid state (Meraldi et al., 2002).

Interference of viral oncoprotein with the centrosome cycle

It has long been known that both HPV E6 and E7, as well as Adenovirus E1A, induce mitotic abnormalities, aneuploidy and genomic instability (Braithwaite et al., 1983; Caporossi and Bacchetti, 1990; Duensing and Munger, 2002). SV40 large T Ag also disrupts the effectiveness of mitotic checkpoint controls (Chang et al., 1997). Increasing evidence suggests that some of these effects reflect the ability of viral oncoproteins to disrupt centrosome control.

E6 interferes with p53-dependent elimination of abnormal cells, yielding the accumulation of cells with centrosomal and mitotic abnormalities (Duensing and Munger, 2002). A similar role in genomic instability through targeting of the p53 function may apply to adenovirus E1B. Although fewer data are available on the function of E1B in centrosome control, its ability to associate with mitotic centrosomes (Brown et al., 1994), as does p53, is intriguing, and reinforces the idea that the centrosomal localization of p53 is central to some of its monitoring functions. Thus, part of the E6 oncogenic function may be exerted by locally inactivating the p53 fraction that associates with mitotic centrosomes.

A different, specific pathway has recently been proposed for SV40 small T Ag. Small T interacts with another mitotic regulator, PP2A, and induces centrosomal abnormalities through its PP2A-interacting domain (Gaillard et al., 2001). The role of PP2A in centrosome control is not fully clarified, yet Aurora-A is one of its substrates (Eyers et al., 2003). PP2A mutations in Drosophila cellularized embryos cause the loss of coordination between the nuclear and centrosome cycles, with centrosomes undergoing continuous duplication cycles (Snaith et al., 1996). In mammalian cells expressing wild-type small T, or its PP2A-binding domain, centrosome maturation seems to be affected: the recruitment of two fundamental components, that is γ-tubulin and centrin, is impaired, the mitotic spindle cannot assemble, and therefore cells arrest in early prophase (Gaillard et al., 2001).

By and large, the most complete studies on the disruption of centrosome control have been carried out with HPV-derived E7. The expression of E7 from highly transforming strains induces centrosome amplification (Munger et al., 2001; Duensing and Munger, 2002); supernumerary centrosomes assemble mutipolar spindles that direct chromosome mis-segregation into daughter cells, and hence can initiate a genetically imbalanced cellular clone (Duensing et al., 2000, 2001). We have recently shown that adenovirus E1A shares this function. Indeed, E1A-expressing cells develop supernumerary centrosomes and abnormal mitotic spindles similar to those induced by E7 (Figure 2).

Figure 2

Centrosome amplification in E1A-expressing cells. Each supernumerary centrosome (visualized by γ-tubulin staining in panel a) gives rise to a spindle pole (depicted by α-tubulin staining in panel b). Therefore, chromosomes (in blue) fail to align correctly and will mis-segregate to multiple poles

Although pRb is an obvious target of both E1A and E7, it is not the sole targeted element in loss of centrosome control: indeed, neither altered expression of pRb regulators (Piboonniyom et al., 2003) nor pRb functional knockout (Borel et al., 2002; Lentini et al., 2002) induce centrosome or spindle abnormalities in a way comparable to E7/E1A in normally cycling cells, although pRb function is required to prevent centrosome reduplication in S phase-arrested cells. Thus, E7/E1A induction of centrosome overduplication may proceed through more complex pathways than the sole antagonization of Rb-dependent-pathways, suggesting that new players must be involved (Munger et al., 2001; Duensing and Munger, 2002).

A novel target of viral oncoproteins in genomic instability: the Ran GTPase

We have recently identified the Ran GTPase as a novel target of viral oncoproteins. E1A, E7 and SV40 large T Ag all interact with Ran in vitro; in the case of E1A, the interaction has also been demonstrated in vivo (De Luca et al., 2003). Ran controls a variety of cellular processes (Clarke and Zhang, 2001; Dasso, 2002; Hetzer et al., 2002; Weis, 2003). Like all GTPases, Ran exists in the GTP- or GDP-bound forms, and its nucleotide-bound state is crucial to its activity. The latter is controlled by regulatory partners of Ran: the guanine exchange factor RCC1, which generates RanGTP; RanGAP1, which activates GTP hydrolysis on Ran; and RanBP1, a Ran effector that modulates nucleotide turnover by increasing the rate of hydrolysis. A major function of Ran is the control of nucleocytoplasmic transport of RNA and proteins in interphase cells. Ran has affinity for transport vectors (importins and exportins) and regulates their association or dissociation with transported cargoes. The compartmentalization of RanGDP (predominantly in the cytoplasm, where the concentration of hydrolysing factors is high) and RanGTP (in the nucleus, where RCC1 resides) determines the directionality of transport in and out of the nucleus.

After nuclear envelope breakdown (NEB), RanGTP regulates the delivery of active ‘spindle activating factors’ (SAFs), which were complexed with transport vectors in the preceding interphase, and thus promotes mitotic spindle assembly (Dasso, 2002; Hetzer et al., 2002; Weis, 2003). Mechanistic insight into the mitotic function of Ran is largely derived from model systems (in particular, Xenopus oocyte-derived systems, which have no centrosomes). Growing evidence is also emerging, linking the Ran system to mitotic centrosomes and spindle poles in mammalian cells (Guarguaglini et al., 2000; Garrett et al., 2002; Gruss et al., 2002; Moore et al., 2002; Di Fiore et al., 2003).

Recent experiments with E1A have elicited a requirement for Ran function in control of centrosome duplication. We have recently found that E1A expression induces S phase entry and, concomitantly, centrosome amplification, in cells in which the Ran network is functional (Figure 2). However, in the tsBN2 cell line, in which the RCC1 exchange factor can be conditionally inactivated, E1A expression stimulates cells to enter the S phase, indicating that the proproliferative functions of E1A are not absolutely dependent on Ran network integrity, but no supernumerary centrosomes are induced: on the contrary, the majority of cells display 1 or 2 centrosomes. These findings suggest that E1A interferes with centrosome control through at least two mechanisms, that is deregulation of pRb-dependent pathways, and through the Ran GTPase network. A similar mechanism may apply to E7, which also interacts with Ran.

E1A downregulates RCC1-dependent nucleotide exchange on Ran in an in vitro assay. Interestingly, an E1A mutant lacking the Ran-interacting domain does not interfere with RCC1-mediated exchange on Ran and fails to induce supernumerary centrosomes (De Luca et al., 2003). Thus E1A induces centrosome amplification only in conditions under which it can produce deregulation of RCC1/Ran activities, but not in the absence of functional RCC1 factor, nor when the Ran-interacting region of E1A is deleted. These findings may indicate several, non mutually exclusive clues worthy of further investigation. First, E1A may deregulate one or more duplication-activating factor(s); Cyclin A/Cdk2 would be an obvious candidate (see above). In the absence of RCC1, E1A may fail to deregulate the activity of Cyclin/Cdk2, or a Cdk2 substrate effectively. It is possible that E1A expression in the absence of RCC1, although stimulating S phase entry, is not sufficient to restore a proficient cell cycle program as would be required for induction of centrosome duplication. Second, subcellular localization of key factors may be crucial to centrosome duplication. For example, NPM/B23 has a nucleolar localization and a small fraction is associated to centrosomes; as recalled above, Cdk2-dependent phosphorylation of NPM/B23 detaches it from centrosomes and is a prerequisite for duplication. Cyclin A/Cdk2 complexes do themselves shuttle between the nucleus and the cytoplasm (Jackman et al., 2002) and a small fraction localizes at centrosomes (Den Elzen and Pines, 2001). Ran/RCC1 function may be required for the localization of at least some of the factors that associate with centrosomes during the cell cycle. Centrosomal regulatory factors may be correctly localized in E1A-expressing cells, but not in RCC1-deficient cells, such that E1A cannot target them. Third, both E1A (De Luca et al., 2003) and E7 (Duensing et al., 2000) induce centrosome amplification in a length of time that does not accommodate more than one or two rounds of DNA replication, and hence may act by over-riding a centrosome licensing mechanism. In yeast, M-specific cyclins control centrosome licensing (Haase et al., 2001). In RCC1-defective cells, M-specific factors are deregulated, and, under certain conditions, prematurely activated (reviewed by Nishimoto, 2000; Moore, 2001). This may result in enforced negative control at the centrosome level, such that a licensing step cannot be targeted by E1A when RCC1 is not functional.

The failure of the Ran-binding defective E1A mutant to induce centrosome amplification is consistent with the idea that E1A deregulates a local Ran-dependent pathway in centrosome control. More generally, the use of viral oncoproteins and mutant derivatives that specifically lack the ability to perform certain functions (for example, lacking specific protein–binding domains) can provide an important experimental tool to unravel pathways of control of genome stability that are altered in tumorigenesis.

Therapeutic perspectives

It is worth recalling that E1A itself is regarded – at least in certain cell types – as a tumor suppressor, essentially for its ability to induce apoptosis in cancer cells (Frisch, 1991, 1996, 2001; Chinnadurai, 1992; Mymryk, 1996; Frisch and Mymryk, 2002). For this reason, its use has been proposed as a specific device to hit cancer cells (Deng et al., 1998; Shao et al., 1999) and productively employed in phase I trials in human cancer gene therapy (Hortobagyi et al., 2001; Yoo et al., 2001; Hubberstey et al., 2002).

Our ever-increasing knowledge of the molecular events underlying the effects of viral oncoproteins in the host cell is currently providing novel useful information for the engineering of a new category of specific pharmacological tools. A remarkable number of amino-acid sequences involved in protein–protein interactions are now identified and their functional role is the object of intensive investigation. Some of these sequences, reproduced as synthetic oligopeptides, may be of interest due to their ability to interfere with specific protein–protein interactions and, consequently, for their potential effect. As a novel molecular tool, the fine tuning of pathophysiological processes that depend upon interactions between proteins is a realistic and appealing challenge. The molecular mapping of the interactions between the viral oncoprotein and key cellular regulators is of great help to devise specifically engineered synthetic oligopeptides, a tool often proposed for the artificial modulation of protein–protein interactions (Severino et al., 2003). Other competitive approaches can also be envisaged, involving the direct screening of libraries of small molecules to identify candidates that block certain receptors or pathways, and new methodologies to find molecules able to modulate protein activity selectively (Leung et al., 2003). Small oligopeptides or small molecules can became promising candidates to inhibit pivotal viral processes inside the infected cell. Obviously, some cells would escape the effect of the peptide/drug – for mechanistic or genetic reasons – and would outgrow the rest of the sensitive population. This phenomenon, widely described for example in conventional cancer chemotherapy, is among the main reasons for the choice of polychemotherapeutic regimens, which guarantee better efficacy and fewer side effects. These new molecules should be simply considered as a further potential therapeutic tool, with the likely advantage of a high selectivity in identifying their own target.


  1. Alazawi W, Pett M, Arch B, Scott L, Freeman T, Stanley MA and Coleman N . (2002). Cancer Res., 62, 6959–6965.

  2. Ali SH and DeCaprio JA . (2001). Semin.Cancer Biol., 11, 15–23.

  3. Bailly E, Doree M, Nurse P and Bornens M . (1989). EMBO J., 8, 3985–3995.

  4. Balmain A . (2001). Nat. Rev. Cancer, 1, 77–82.

  5. Berger AJ, Baege A, Guillemette T, Deeds J, Meyer R, Disbrow G, Schlegel R and Schlegel R . (2002). Am. J. Pathol., 161, 603–610.

  6. Borel F, Lohez OD, Lacroix FB and Margolis RL . (2002). Proc. Natl. Acad. Sci. USA, 99, 9819–9824.

  7. Braithwaite AW, Cheetham BF, Li P, Parish CR, Waldron Stevens LK and Bellett AJ . (1983). J. Virol., 45, 192–199.

  8. Brehm A, Miska EA, McCance DJ, Reid JL, Bannister AJ and Kouzarides T . (1998). Nature, 391, 597–601.

  9. Brinkley BR . (2001). Trends Cell Biol., 11, 18–21.

  10. Brown CR, Doxsey SJ, White E and Welch WJ . (1994). J. Cell Physiol., 160, 47–60.

  11. Cairns J . (1975). Nature, 255, 197–200.

  12. Caporossi D and Bacchetti S . (1990). J. Gen. Virol., 71 (Pt 4), 801–808.

  13. Carbone M . (2000). J. Cell. Biochem., 76, 189–193.

  14. Carbone M, Pass HI, Rizzo P, Marinetti M, Di Muzio M, Mew DJ, Levine AS and Procopio A . (1994). Oncogene, 9, 1781–1790.

  15. Carroll PE, Okuda M, Horn HF, Biddinger P, Stambrook PJ, Gleich LL, Li YQ, Tarapore P and Fukasawa K . (1999). Oncogene, 18, 1935–1944.

  16. Chang TH, Ray FA, Thompson DA and Schlegel R . (1997). Oncogene, 14, 2383–2393.

  17. Chinnadurai G . (1992). Oncogene, 7, 1255–1258.

  18. Ciciarello M, Mangiacasale R, Casenghi M, Zaira LM, D'Angelo M, Soddu S, Lavia P and Cundari E . (2001). J. Biol. Chem., 276, 19205–19213.

  19. Clarke PR and Zhang C . (2001). Trends Cell Biol., 11, 366–371.

  20. Claudio PP, Howard CM, Baldi A, De Luca A, Fu Y, Condorelli G, Sun Y, Colburn N, Calabretta B and Giordano A . (1994). Cancer Res., 54, 5556–5560.

  21. D'Assoro AB, Lingle WL and Salisbury JL . (2002). Oncogene, 21, 6146–6153.

  22. Dahiya A, Gavin MR, Luo RX and Dean DC . (2000). Mol. Cell. Biol., 20, 6799–6805.

  23. Dasso M . (2002). Curr. Biol., 12, R502.

  24. De Luca A, Mangiacasale R, Severino A, Malquori L, Baldi A, Palena A, Mileo AM, Lavia P and Paggi MG . (2003). Cancer Res., 63, 1430–1437.

  25. Debbas M and White E . (1993). Genes Dev., 7, 546–554.

  26. Den Elzen N and Pines J . (2001). J. Cell Biol., 153, 121–136.

  27. Deng J, Xia W and Hung MC . (1998). Oncogene, 17, 2167–2175.

  28. Di Fiore B, Ciciarello M, Mangiacasale R, Palena A, Tassin AM, Cundari E and Lavia P . (2003). J. Cell Sci., in press.

  29. Duensing S, Duensing A, Crum CP and Munger K . (2001). Cancer Res., 61, 2356–2360.

  30. Duensing S, Lee LY, Duensing A, Basile J, Piboonniyom S, Gonzalez S, Crum CP and Munger K . (2000). Proc. Natl. Acad. Sci. USA, 97, 10002–10007.

  31. Duensing S and Munger K . (2002). Oncogene, 21, 6241–6248.

  32. Ewen ME, Xing YG, Lawrence JB and Livingston DM . (1991). Cell, 66, 1155–1164.

  33. Eyers PA, Erikson E, Chen LG and Maller JL . (2003). Curr. Biol., 13, 691–697.

  34. Ferreira R, Magnaghi-Jaulin L, Robin P, Harel-Bellan A and Trouche D . (1998). Proc. Natl. Acad. Sci. USA, 95, 10493–10498.

  35. Fisk HA, Mattison CP and Winey M . (2002). Curr. Opin. Cell Biol., 14, 700–705.

  36. Fisk HA and Winey M . (2001). Cell, 106, 95–104.

  37. Frank SA . (2003). Curr. Biol., 13, R261–R263.

  38. Frisch SM . (1991). Proc. Natl. Acad. Sci. USA, 88, 9077–9081.

  39. Frisch SM . (1996). Mutat. Res., 350, 261–266.

  40. Frisch SM . (2001). Adv. Cancer Res., 80, 39–49.

  41. Frisch SM and Mymryk JS . (2002). Nat. Rev. Mol. Cell Biol., 3, 441–452.

  42. Gage JR, Meyers C and Wettstein FO . (1990). J. Virol., 64, 723–730.

  43. Gaillard S, Fahrbach KM, Parkati R and Rundell K . (2001). J. Virol., 75, 9799–9807.

  44. Gallimore PH and Turnell AS . (2001). Oncogene, 20, 7824–7835.

  45. Garrett S, Auer K, Compton DA and Kapoor TM . (2002). Curr. Biol., 12, 2055–2059.

  46. Goepfert TM, Adigun YE, Zhong L, Gay J, Medina D and Brinkley WR . (2002). Cancer Res., 62, 4115–4122.

  47. Goodrich DW, Wang NP, Qian YW, Lee EY and Lee WH . (1991). Cell, 67, 293–302.

  48. Gruss OJ, Wittmann M, Yokoyama H, Pepperkok R, Kufer T, Sillje H, Karsenti E, Mattaj IW and Vernos I . (2002). Nat. Cell Biol., 4, 871–879.

  49. Guarguaglini G, Renzi L, D'Ottavio F, Di Fiore B, Casenghi M, Cundari E and Lavia P . (2000). Cell Growth Differ., 11, 455–465.

  50. Haase SB, Winey M and Reed SI . (2001). Nat. Cell Biol., 3, 38–42.

  51. Hannon GJ, Demetrick D and Beach D . (1993). Genes Dev., 7, 2378–2391.

  52. Helt AM and Galloway DA . (2003). Carcinogenesis, 24, 159–169.

  53. Herwig S and Strauss M . (1997). Eur. J. Biochem., 246, 581–601.

  54. Hetzer M, Gruss OJ and Mattaj IW . (2002). Nat. Cell Biol., 4, E177–E184.

  55. Hinchcliffe EH and Sluder G . (2001). Genes Dev., 15, 1167–1181.

  56. Hortobagyi GN, Ueno NT, Xia W, Zhang S, Wolf JK, Putnam JB, Weiden PL, Willey JS, Carey M, Branham DL, Payne JY, Tucker SD, Bartholomeusz C, Kilbourn RG, De Jager RL, Sneige N, Katz RL, Anklesaria P, Ibrahim NK, Murray JL, Theriault RL, Valero V, Gershenson DM, Bevers MW, Huang L, Lopez-Berestein G and Hung MC . (2001). J. Clin Oncol., 19, 3422–3433.

  57. Hubberstey AV, Pavliv M and Parks RJ . (2002). Cancer Gene Ther., 9, 321–329.

  58. Jackman M, Kubota Y, Den Elzen N, Hagting A and Pines J . (2002). Mol. Biol. Cell, 13, 1030–1045.

  59. Jackman M, Lindon C, Nigg EA and Pines J . (2003). Nat. Cell Biol., 5, 143–148.

  60. Kaelin Jr WG . (1999). J. Natl. Cancer Inst., 91, 594–598.

  61. Kinzler KW and Vogelstein B . (1996). Cell, 87, 159–170.

  62. Kinzler KW and Vogelstein B . (1997). Nature, 386, 761–763.

  63. Lane DP and Crawford LV . (1979). Nature, 278, 261–263.

  64. Lavia P and Jansen-Dürr P . (1999). Bioessays, 21, 221–230.

  65. Lentini L, Pipitone L and Di Leonardo A . (2002). Neoplasia, 4, 380–387.

  66. Leung D, Hardouin C, Boger DL and Cravatt BF . (2003). Nat. Biotechnol. 21, 687–691.

  67. Li Y, Graham C, Lacy S, Duncan AM and Whyte P . (1993). Genes Dev., 7, 2366–2377.

  68. Lingle WL and Salisbury JL . (2000). Curr. Top. Dev. Biol., 49, 313–329.

  69. Linzer DI and Levine AJ . (1979). Cell, 17, 43–52.

  70. Magnaghi-Jaulin L, Groisman R, Naguibneva I, Robin P, Lorain S, Le Villain JP, Troalen F, Trouche D and Harel-Bellan A . (1998). Nature, 391, 601–605.

  71. Mantel C, Braun SE, Reid S, Henegariu O, Liu L, Hangoc G and Broxmeyer HE . (1999). Blood, 93, 1390–1398.

  72. Matsumoto Y, Hayashi K and Nishida E . (1999). Curr. Biol., 9, 429–432.

  73. Mayol X, Grana X, Baldi A, Sang N, Hu Q and Giordano A . (1993). Oncogene, 8, 2561–2566.

  74. Meraldi P, Honda R and Nigg EA . (2002). EMBO J., 21, 483–492.

  75. Meraldi P, Lukas J, Fry AM, Bartek J and Nigg EA . (1999). Nat. Cell Biol., 1, 88–93.

  76. Moore JD . (2001). Bioessays, 23, 77–85.

  77. Moore W, Zhang C and Clarke P . (2002). Curr. Biol., 12, 1442.

  78. Mortimer Jr, EA, Lepow ML, Gold E, Robbins FC, Burton GJ and Fraumeni Jr JF . (1981). N. Engl. J. Med., 305, 1517–1518.

  79. Mulligan G and Jacks T . (1998). Trends Genet., 14, 223–229.

  80. Munger K, Basile JR, Duensing S, Eichten A, Gonzalez SL, Grace M and Zacny VL . (2001). Oncogene, 20, 7888–7898.

  81. Mymryk JS . (1996). Oncogene, 13, 1581–1589.

  82. Mymryk JS, Shire K and Bayley ST . (1994). Oncogene, 9, 1187–1193.

  83. Nakayama K, Nagahama H, Minamishima YA, Matsumoto M, Nakamichi I, Kitagawa K, Shirane M, Tsunematsu R, Tsukiyama T, Ishida N, Kitagawa M and Hatakeyama S . (2000). EMBO J., 19, 2069–2081.

  84. Nevins JR . (1995). Curr. Top. Microbiol. Immunol., 199 (Pt 3), 25–32.

  85. Nevins JR . (2001). Hum. Mol. Genet., 10, 699–703.

  86. Nigg EA . (2002). Nat. Rev. Cancer, 2, 815–825.

  87. Nishimoto T . (2000). Biol. Chem., 381, 397–405.

  88. Okuda M, Horn HF, Tarapore P, Tokuyama Y, Smulian AG, Chan PK, Knudsen ES, Hofmann IA, Snyder JD, Bove KE and Fukasawa K . (2000). Cell, 103, 127–140.

  89. Paggi MG, Baldi A, Bonetto F and Giordano A . (1996). J. Cell. Biochem., 62, 418–430.

  90. Paggi MG and Giordano A . (2001). Cancer Res., 61, 4651–4654.

  91. Piboonniyom SO, Duensing S, Swilling NW, Hasskarl J, Hinds PW and Munger K . (2003). Cancer Res., 63, 476–483.

  92. Pihan GA, Purohit A, Wallace J, Knecht H, Woda B, Quesenberry P and Doxsey SJ . (1998). Cancer Res., 58, 3974–3985.

  93. Pockwinse SM, Krockmalnic G, Doxsey SJ, Nickerson J, Lian JB, Van Wijnen AJ, Stein JL, Stein GS and Penman S . (1997). Proc. Natl. Acad. Sci. USA, 94, 3022–3027.

  94. Raff JW . (2002). Trends Cell Biol., 12, 222–225.

  95. Rapp L and Chen JJ . (1998). Biochim. Biophys. Acta Rev. Cancer, 1378, F1–F19.

  96. Riley DJ, Lee EYHP and Lee W-H . (1994). Annu. Rev. Cell Biol., 10, 1–29.

  97. Sanchez-Prieto R, Quintanilla M, Cano A, Leonart ML, Martin P, Anaya A and Cajal S . (1996). Oncogene, 13, 1083–1092.

  98. Sang N, Caro J and Giordano A . (2002). Front. Biosci., 7, D407–D413.

  99. Sang N, Severino A, Russo P, Baldi A, Giordano A, Mileo AM, Paggi MG and De Luca A . (2001). J. Biol. Chem., 276, 27026–27033.

  100. Scheffner M, Takahashi T, Huibregtse JM, Minna JD and Howley PM . (1992). J. Virol., 66, 5100–5105.

  101. Schmeichel KL . (2002). Breast Cancer Res., 5, 59.

  102. Severino A, Felsani A, Giordano A and Paggi MG . (2003). Cell Cycle Inhibitors in Cancer Therapy. Giordano A, Soprano KJ (eds). Humana Press Inc: Totowa, NJ, pp. 297–310.

    Google Scholar 

  103. Shao R, Hu MC, Zhou BP, Lin SY, Chiao PJ, von Lindern RH, Spohn B and Hung MC . (1999). J. Biol. Chem., 274, 21495–21498.

  104. Shenk T and Flint J . (1991). Adv. Cancer Res., 57, 47–85.

  105. Sidle A, Palaty C, Dirks P, Wiggan O, Kiess M, Gill RM, Wong AK and Hamel PA . (1996). Crit. Rev. Biochem. Mol. Biol., 31, 237–271.

  106. Snaith HA, Armstrong CG, Guo Y, Kaiser K and Cohen PT . (1996). J. Cell Sci., 109 (Pt 13), 3001–3012.

  107. Starostik P, Chow KNB and Dean DC . (1996). Mol. Cell. Biol., 16, 3606–3614.

  108. Stiegler P, De Luca A, Bagella L and Giordano A . (1998a). Cancer Res., 58, 5049–5052.

  109. Stiegler P, Kasten M and Giordano A . (1998b). J. Cell. Biochem., 30–36.

  110. Strickler HD, Rosenberg PS, Devesa SS, Hertel J, Fraumeni Jr JF and Goedert JJ . (1998). JAMA, 279, 292–295.

  111. Teodoro JG, Shore GC and Branton PE . (1995). Oncogene, 11, 467–474.

  112. Thomas M, Pim D and Banks L . (1999). Oncogene, 18, 7690–7700.

  113. Vasseur S, Malicet C, Calvo EL, Labrie C, Berthezene P, Dagorn JC and Iovanna JL . (2003). Mol. Cancer, 2, 19.

  114. Weis K . (2003). Cell, 112, 441–451.

  115. White E . (2001). Oncogene, 20, 7836–7846.

  116. Whyte P, Buchkovich KJ, Horowitz JM, Friend SH, Raybuck M, Weinberg RA and Harlow E . (1988). Nature, 334, 124–129.

  117. Yew PR and Berk AJ . (1992). Nature, 357, 82–85.

  118. Yoo GH, Hung MC, Lopez-Berestein G, LaFollette S, Ensley JF, Carey M, Batson E, Reynolds TC and Murray JL . (2001). Clin. Cancer Res., 7, 1237–1245.

  119. Zatsepina OV, Rousselet A, Chan PK, Olson MO, Jordan EG and Bornens M . (1999). J. Cell Sci., 112 (Part 4), 455–466.

  120. Zhu L, van den Heuvel S, Helin K, Fattaey A, Ewen ME, Livingston DM, Dyson N and Harlow E . (1993). Genes Dev., 7, 1111–1125.

  121. zur Hausen H . (1996). Biochim. Biophys. Acta, 1288, F55–F78.

  122. Zwerschke W and Jansen-Dürr P . (2000). Adv. Cancer Res., 78, 1–29.

Download references


Research in our laboratories was supported by Associazione Italiana per la Ricerca sul Cancro (AIRC), Agenzia Spaziale Italiana (ASI) and Ministero della Salute grants to MGP and PL.

Author information



Corresponding author

Correspondence to Marco G Paggi.

Additional information

Note added in proof

While this review was in preparation, it has been shown that a fraction of the Ran GTPase itself is tightly associated to centrosomes (Keryer G, Di Fiore B, Celati C, Lechtreck KF, Mogensen M, Delouvée M, Lavia P, Bornens M and Tassin AM (2003) Mol. Biol. Cell, in press).

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lavia, P., Mileo, A., Giordano, A. et al. Emerging roles of DNA tumor viruses in cell proliferation: new insights into genomic instability. Oncogene 22, 6508–6516 (2003). https://doi.org/10.1038/sj.onc.1206861

Download citation


  • E1A
  • E7
  • mitosis
  • genomic instability
  • cellular transformation

Further reading