Oncogene (2005) 24, 7686–7696 & 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00 www.nature.com/onc

Inactivating intracellular antiviral responses during adenovirus

Matthew D Weitzman*,1 and David A Ornelles2

1Laboratory of Genetics, The Salk Institute for Biological Studies, La Jolla, CA 92037, USA; 2Department of Microbiology and Immunology, Wake Forest University School of Medicine, Winston-Salem, NC 27157, USA

DNA viruses promote cycle progression, stimulate normal cells, and this may impact the balance between unscheduled DNA synthesis, and present the cell with an virus infection and host defenses. These differences extraordinary amount of exogenous DNA. These insults between normal cells and tumor cells are exploited by elicit vigorous responses mediated by cellular factors that oncolytic viral agents, which are designed to replicate govern cellular homeostasis. To ensure productive infec- selectively for cancer therapy. Understanding how tion, adenovirus has developed means to inactivate these intracellular antiviral defenses, such as the DNA intracellular antiviral responses. Among the challenges to damage response, affect viral growth will direct the host cell is the viral DNA genome, which is viewed as improvements to these viral weapons. DNA damage and elicits a cellular response to inhibit The cell mounts a multifaceted response to adenovirus replication. Adenovirus therefore encodes that infection in an attempt to inhibit viral gene expression, dismantle the cellular DNA damage machinery. Studying disrupt viral DNA replication and promote death of the virus–host interactions has yielded insights into the infected cell. Additional arms of the response serve to molecular functioning of fundamental cellular mechan- signal other cells of the viral encounter and to facilitate isms. In addition, it has suggested ways that viral recognition of the infected cell by the immune system. cytotoxicity can be exploited to offer a selective means This multifaceted response of the cell is met by an of restricted growth in tumor cells as a therapy against equally diverse array of viral activities involving multi- cancer. In this review, we discuss aspects of the ple transcription regions that aim to counter these host intracellular response that are unique to adenovirus responses. Many of the viral proteins that combat the infection and how adenoviral proteins produced from the host immune responses have been described in detail by early region E4 act to neutralize antiviral defenses, with a a number of excellent recent reviews (see Horwitz, 2001; particular focus on DNA damage signaling. Burgert et al., 2002; Fessler et al., 2004; Lichtenstein Oncogene (2005) 24, 7686–7696. doi:10.1038/sj.onc.1209063 et al., 2004; Schaack, 2005). In this review, we have therefore chosen to focus on aspects of the disarming of Keywords: adenovirus; DNA repair; DNA damage the intracellular response that involve the early region 4 response; translation (E4) of the adenoviral genome. Recent studies suggest that the ability of the adenoviral E4 proteins to promote viral gene expression through interactions with cellular targets is key to overcoming some of the intracellular barriers encountered during adenovirus infection. Here Introduction we review the current state of knowledge of E4 functions and their potential implications for tumor therapy. We Virus challenge the host with foreign material will specifically deal with how these proteins counter the in the form of proteins and nucleic acids. These assaults DNA damage response, restrictions and upon the host cell inevitably induce antiviral responses. obstacles to growth through regulation of protein In the dynamic warfare that is established, there will be translation. battles won by either side as the virus attempts to control the cell, and the cell in turn tries to keep the invader at bay. The adenoviruses have provided a powerful model for studying virus–host interactions and Introduction to the adenovirus genome and its replication tools to unravel molecular mechanisms in mammalian cells (reviewed in Ben-Israel and Kleinberger, 2002; The adenovirus genome Burgert et al., 2002; Russell et al., 2004; O’Shea, 2005). The adenovirus genome consists of a linear double- Detailed knowledge of their genetic composition has stranded DNA molecule of approximately 36 kb (Shenk, also facilitated their development into vectors for gene 1996). At either end are inverted terminal repeats delivery and exploitation for tumor cell killing. The (ITRs), which contain sequences that serve as the molecular constitution of tumor cells differs from that of origins for DNA replication. The genes are grouped into transcriptional units: five early (E1a, E1b, E2, E3 *Correspondence: MD Weitzman; E-mail: [email protected] and E4), two delayed early and one major late unit. Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7687 Each region is transcribed to yield multiple mRNAs Transcription of the E4 region is regulated by a single through the use of alternative splicing and differential promoter that is activated early in infection by E1a and polyadenylation. In many cases, the transcription units there are six known E4 proteins that are produced generate products that carry out related functions within temporally through alternative splicing (Dix and Leppard, the temporally regulated viral life cycle. There are also 1993; Leppard, 1997; Tauber and Dobner, 2001b). two viral-associated (VA) RNAs transcribed from a Transcription declines at later times in infection due polymerase III promoter. Early proteins not only to repression by E2a, and by feedback inhibition of promote expression of key viral and cellular genes, but E1a-mediated E4 transactivation by the E4orf4 protein also suppress the ability of the cell to mount an antiviral (Bondesson et al., 1996). A virus lacking the response. E4 region is severely impaired, with defects in viral DNA replication, accumulation of late viral messages Adenovirus gene expression and replication and proteins, virus particle assembly and shut-off of host protein synthesis (Halbert et al., 1985; Weinberg The E1a gene encodes proteins that play a major role in and Ketner, 1986; Yoder and Berget, 1986; Falgout and the transcriptional regulation of viral and cellular genes Ketner, 1987; Sandler and Ketner, 1989). (Shenk and Flint, 1991; Nevins, 1995; Frisch and Mymryk, 2002; discussed in detail by Berk in this issue). The E4 region and its functions E1a is the first gene to be transcribed and is alternatively spliced to generate viral proteins that fulfill a plethora of Genetic analysis of the E4 region (mainly in tumor cells) functions, including transcriptional activation and showed that within individual open reading induction of the host cell into . In addition to frames have minimal or modest effects (Halbert et al., promoting early viral gene expression, the E1a proteins 1985; Bridge and Ketner, 1989; Huang and Hearing, inhibit activation of genes induced by interferon (IFN) 1989). This is partly due to functional redundancy and IL-6 (Anderson and Fennie, 1987; Reich et al., between E4orf3 and E4orf6, while the other E4 gene 1988; Takeda et al., 1994), and may thus blunt the effect products are dispensable for lytic growth of the virus, at of inflammatory cytokines and chemokines. The E1b least in tumor cells analysed in vitro. Expression of an region produces two gene products involved in counter- E4orf1 protein has only been detected for some of the ing cellular responses. The E1b19K protein is a serotypes and mutations in Ad5 E4orf1 have minimal functional homolog of the cellular protein BCL2 that effects on viral growth in most cells. However, deletion inhibits and prevents the premature death of of E4orf1 was deleterious to virus growth in primary the infected cells that would limit virus production cells (Frese et al., 2003; O’Shea et al., 2005). The E4orf4 (Cuconati and White, 2002). The E1b55K protein forms product is not essential for virus growth and its deletion a complex with the adenoviral E4orf6 protein that has minimal effect in tumor cells (Halbert et al., 1985; promotes degradation of cellular proteins, disarms the Branton and Roopchand, 2001). However, a mutant cellular DNA damage response and facilitates the expressing E4orf4 in the absence of all other E4 genes preferential nucleo-cytoplasmic transport of late adeno- was highly defective for DNA synthesis, suggesting an viral mRNA (Dobner and Kzhyshkowska, 2001; Flint inhibitory role for the viral protein in regulating and Gonzalez, 2003). replication (Bridge et al., 1993). The E4orf4 product Proteins encoded by the adenovirus E2 genes are affects protein phosphorylation during infection essential for viral DNA replication (de Jong et al., 2003; through binding to protein phosphatase 2A (PP2A), Liu et al., 2003). Adenoviral DNA replication is one of the major serine/threonine-specific phosphatases initiated by a protein priming mechanism involving the in the cell (Kleinberger and Shenk, 1993). E4orf4 also preterminal protein (pTP), which is covalently attached inhibits colony formation by inducing apoptosis of to the 50 end of the genomic DNA. Viral DNA is transformed cells in a -independent manner (Lavoie synthesized by the E2b DNA polymerase. The E2a gene et al., 1998; Marcellus et al., 1998; Shtrichman and encodes a single-stranded DNA-binding protein (DBP) Kleinberger, 1998) that also requires interaction with that is required for viral DNA replication and also PP2A (Shtrichman et al., 1999; Marcellus et al., 2000). regulates transcription and translation of viral genes The E4orf3 protein is highly conserved, associates during the late stages of infection (Shenk, 1996). with the nuclear matrix and induces the reorganization Many viral proteins counter the host immune of nuclear bodies called PML oncogenic domains response to adenovirus infection (Horwitz, 2001). The (PODs) or ND10 (Doucas et al., 1996). These are large E3 region encodes proteins specifically involved in nuclear structures that contain proteins implicated in immune evasion (Fessler et al., 2004; Lichtenstein multiple cellular functions including transformation, et al., 2004). The E3 proteins affect the functioning of genomic stability, DNA repair, transcriptional control, cell surface receptors, intracellular signaling events and apoptosis and the IFN response (Borden, 2002). E4orf3 secretion of proinflammatory molecules. There is also expression induces rearrangement of PODs into track- an E3 product that promotes cell death to facilitate like structures, when it is ectopically expressed and in release of progeny virions late in the infection cycle the context of viral infection (Carvalho et al., 1995; (Tollefson et al., 1996). Together, these E3 proteins Doucas et al., 1996). The mechanism by which E4orf3 protect infected cells from inflammatory challenges and accomplishes these drastic changes in the architecture of cytolysis. nuclear structures and the cellular consequence of POD

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7688 disruption are not clear. As many viruses target the has developed E4-encoded strategies to inactivate this PODs and in some cases rearrange their components, it intracellular antiviral response. has been suggested that reorganization is advantageous for viral replication (Maul, 1998; Everett, 2001). E4orf3 The DNA damage response to adenovirus infection may play a role in the initiation of viral replication (Evans and Hearing, 2003), although its functions are Mammalian cells have evolved an elaborate network of dispensable at high multiplicities of infection (Doucas cellular checkpoints that monitor the integrity of et al., 1996). E4orf3 has been shown to bind E1b55K, genomic DNA during cell proliferation. DNA damage relocalizing it to the nucleus and recruiting it into the results in activation of signal transduction pathways discrete PML tracks (Konig et al., 1999). E4orf3 may that control cell cycle progression, the DNA repair be situated to counter the multifaceted antiviral response machinery and programmed cell death (Zhou and through its ability to target constituents of the PODs Elledge, 2000; Bartek and Lukas, 2001; Rouse and and because of its ability to regulate late viral gene Jackson, 2002). A central player in the cellular response expression. to DNA double-strand breaks (DSBs) is the Mre11 The E4orf6 gene encodes a 34 kDa protein that is complex, containing Mre11, Rad50 and NBS1 proteins highly conserved across a variety of adenovirus sero- (D’Amours and Jackson, 2002; Stracker et al., 2004). types (Boyer and Ketner, 2000). E4orf6 has a central Protein phosphorylation is the principal method of cysteine-rich section, which forms a functional zinc- signaling used by the damage response, and two central binding domain (Boyer and Ketner, 2000). There is also signal transducers are the related PI3-like protein a conserved arginine-rich region at the C-terminus of kinases, ataxia-telangiectasia mutated (ATM) and E4orf6 that forms an amphipathic a-helix important for ATM-Rad3-related (ATR). It has been proposed that its function (Orlando and Ornelles, 1999; Cathomen and ATM is activated by intermolecular autophosphoryla- Weitzman, 2000). E4orf6 forms a complex with the viral tion on Ser1981 and dimer dissociation, which enables E1b55K protein that is involved in viral DNA replica- phosphorylation of substrate proteins (Bakkenist and tion, RNA processing, nucleo-cytoplasmic transport of Kastan, 2003; Bakkenist and Kastan, 2004). ATR may late viral mRNA and the shut-off of host protein be recruited to sites of DNA damage by its cofactor synthesis. E4orf6 also recruits cellular factors including ATRIP and by RPA bound to single-stranded DNA Cullin-5, Elongin B and Elongin C into a ubiquitin (Zou and Elledge, 2003). Many cellular DNA repair and ligase that together with E1b55K mediates polyubiqui- checkpoint proteins are substrates for ATM and ATR tinylation and degradation of cellular targets (Querido kinase activity (Kastan and Lim, 2000; Shiloh, 2003). et al., 2001; Harada et al., 2002). The targeting of These include the downstream checkpoint kinases Chk1 cellular proteins for proteasome-mediated degradation and Chk2, and other damage response proteins such as may be one of the ways that the E1b55K/E4orf6 53BP1, H2AX, BRCA1, p53 and RPA32. Both Mre11 complex functions to inactivate cellular antiviral de- and NBS1 proteins are also targets of ATM and fenses. This complex also facilitates export of late viral possibly ATR (D’Amours and Jackson, 2002; Lavin, mRNAs at the expense of cellular RNAs (Dobner 2004). This same apparatus that recognizes damage to and Kzhyshkowska, 2001; Flint and Gonzalez, 2003). the host genomic DNA also appears to respond to The complex can shuttle between the nucleus and the adenovirus infection. In the absence of the E4 region, cytoplasm, and nuclear export signals of differing the cellular machinery is activated, ultimately resulting strength have been reported for both E1b55K in joining together of viral genomes in a DNA repair (Kratzer et al., 2000) and E4orf6 (Dobbelstein et al., pathway. 1997). Early after adenovirus infection with a virus that lacks the E4 genes, the signaling pathways characteristic of a cellular response to DNA DSBs are activated (Stracker et al., 2002; Carson et al., 2003). This can be detected using Adenovirus and the cellular DNA damage machinery phospho-specific that recognize modified residues on DNA repair proteins. The ATM kinase Adenovirus may be unique among animal viruses becomes activated, as demonstrated by autophosphor- because its genome exists predominantly as discrete ylation on S1981, and many of the known substrates for monomers of double-stranded DNA throughout the ATM and ATR become phosphorylated. Infections in course of a productive infection. At late times of mutant cell lines showed that there is some redundancy infection, amplification of the linear adenovirus genome for ATM, ATR and possibly other kinases in signaling presents the cell with as many as 100 000 double- (Carson et al., 2003). The signaling events require the stranded DNA ends. The host cell responds to this new Mre11 complex (Carson et al., 2003) and are not insult by activation of the cellular DNA damage observed during infection with wild-type virus due to response (Stracker et al., 2002; Carson et al., 2003), inactivation of the complex by the E4 proteins (see which can be construed as an antiviral response. The below). viral genome elicits a DNA damage response but this is Proteins involved in the cellular response to DNA only detected in the absence of a functional E4 region. damage are recruited to sites of DSB in the host genome The DNA repair machinery therefore functions as an and appear as foci within the nucleus of damaged cells. obstacle to adenovirus infection but the wild-type virus Many of these proteins are found to form foci

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7689 surrounding viral replication centers during infection the Mre11 complex components, which partially colo- with the E4-deleted adenovirus, including the Mre11 calize in nuclear tracks with E4orf3 (Stracker et al., complex and the autophosphorylated form of ATM 2002, 2005; Evans and Hearing, 2005). In addition to (Stracker et al., 2002; Carson et al., 2003). The Mre11 mislocalization, infection is also accompanied by complex accumulates in foci at the virus centers even in proteasome-mediated degradation of the Mre11 com- the absence of ATM and ATR signaling. This suggests plex proteins. The E1b55K/E4orf6 proteins of adeno- that recognition of the virus is an upstream event that virus are responsible for targeting degradation of the occurs prior to kinase activation. It is not yet known Mre11 complex (Stracker et al., 2002), and E1b55K what viral structures are specifically recognized by the binds to the complex (Carson et al., 2003). This suggests cellular repair machinery. Viral replication presents the that members of the Mre11 complex may also be cell with a large amount of foreign DNA, which will substrates for modification by the ubiquitin ligase that is contain unusual DNA structures and replication inter- recruited by E4orf6. mediates. The adenovirus genome also possesses a The E4orf3 and E4orf6 proteins can each rescue an terminal protein that is covalently linked to the 50 end E4-deleted virus by preventing concatemerization, and of the genome to initiate replication. This protein–DNA both affect signaling in the cellular DNA damage complex may represent a substrate for recognition and response. Expression of the E4orf3 protein from Ad5 processing by the Mre11 complex, analogous to the can prevent accumulation of the Mre11 complex at viral situation with Spo11, which requires the Mre11 complex replication centers (Stracker et al., 2005), and can also for its removal from DSBs during meiotic recombina- abrogate ATR-mediated phosphorylation events during tion. infection (MD Weitzman, unpublished observations). The result of the cellular DNA damage response to Inhibition of concatemer formation and ATR activation adenovirus infection is processing of the viral genome correlates with redistribution of the Mre11 complex into concatemers. Infection with viruses deleted of E4 (Stracker et al., 2005; MD Weitzman, unpublished results in the formation of large concatemers of viral observations). E4orf3 also induces the accumulation of DNA that can be detected by pulsed field gel electro- the Mre11 complex components into cytoplasmic phoresis. These concatemers consist of covalently linked aggregates that assemble into a structure that resembles molecules with no specific orientation and heteroge- an aggresome (Araujo et al., 2005), and this could neous junctions (Weiden and Ginsberg, 1994). There contribute to its inactivation. Mislocalization of the appears to be a requirement for the nonhomologous Mre11 protein can be separated from POD disruption end-joining repair pathway, as concatemers are not (Evans and Hearing, 2005; Stracker et al., 2005) and in produced in cells with mutant DNA-PKcs (Boyer et al., the absence of E4orf6, it may also play a role in 1999; Stracker et al., 2002) and ligase IV (Stracker et al., promoting virus replication (Evans and Hearing, 2005). 2002). In addition, concatemerization requires the Degradation of the Mre11 complex by the adenovirus Mre11 complex and is not observed in mutant cell lines E1b55K/E4orf6 proteins blocks joining of viral genomes (Stracker et al., 2002). Although neither ATM nor ATR and also prevents the cell from mounting a DNA is absolutely required for concatemer formation, there may damage response to DSBs (Stracker et al., 2002; Carson be a role for signaling pathways in the processing and et al., 2003). Degradation of the Mre11 complex joining of viral genomes (MD Weitzman, unpublished correlates with prevention of ATM activation and observations). checkpoint induction, consistent with an upstream sensor role for the complex (Carson et al., 2003). The E4orf6 protein may also have some effects that Inactivating the damage response are independent of its association with E1b55K and the Mre11 complex. For example, expression of E4orf6 can In contrast to the E4-deleted adenovirus, during radiosensitize cells (Hart et al., 2005). Although E4orf6 infection with wild-type virus, the genome is observed does not inhibit the kinase activity of DNA-PKcs on an as a linear monomer and there is no activation of ATM in vitro substrate (Boyer et al., 1999; Hart et al., 2005), it and ATR signaling (Stracker et al., 2002; Carson et al., can prolong the autophosphorylation of DNA-PKcs (Hart 2003). This suggests that the E4 region encodes viral et al., 2005), implying an inhibition of dephosphorylation. factors that can prevent both viral concatemer forma- The cellular tumor suppressor protein p53 is also part tion and activation of the cellular DNA damage of the cellular response to DNA damage, and is itself a response. Adenovirus has evolved two mechanisms to phosphorylation target of the kinases activated by DNA dismantle the obstacle of DNA repair during infection, damage (Wahl and Carr, 2001). Infections with mutant thus preventing activation of cellular checkpoints and viruses that do not express either E1b55K or E4orf6 concatemerization. There is redundancy between E4orf3 gene products results in stabilization of p53 protein and E4orf6 proteins for their ability to prevent levels (Grand et al., 1994; Steegenga et al., 1998), which concatemer formation and both achieve this by target- may reflect phosphorylation and effects mediated by the ing the Mre11 complex (Stracker et al., 2002). Early viral E1a polypeptides (Lowe and Ruley, 1993). Even after infection, the Mre11/Rad50/NBS1 proteins are higher levels of p53 protein accumulated in cells infected redistributed from their diffuse localization in the with a double-mutant virus deleted of both E1b55K and nucleoplasm into nuclear speckles and cytoplasmic E4orf3 genes (DA Ornelles, unpublished observations). aggregates. The E4orf3 protein is sufficient to reorganize The E1b55K/E4orf6 complex promotes proteasomal

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7690 degradation of p53 during viral infections (Grand et al., coming cell cycle restrictions. Adenoviruses deleted of 1994; Querido et al., 1997; Steegenga et al., 1998; the E4orf3 and either E1b55K or E4orf6 genes are Cathomen and Weitzman, 2000). The E4orf6 and severely defective for late gene expression. In contrast to E1b55K proteins can also both independently inhibit the E1b55K or E4orf6 single-mutant viruses, the the transcriptional activity of p53 through direct binding double-mutant viruses did not derive any benefit from (Yew and Berk, 1992; Dobner et al., 1996; Cathomen infecting S-phase cells (Goodrum and Ornelles, 1999; and Weitzman, 2000). Therefore, functions of p53 in the Shepard and Ornelles, 2003). The need for E4orf3 to cellular responses to DNA damage are inactivated enhance replication in S-phase-infected cells was un- during infection with wild-type adenovirus. expected because deletion of E4orf3 alone has no effect on virus yield or viral gene expression (Halbert et al., 1985; Huang and Hearing, 1989). Thus, although E4orf3 cannot overcome the G1 restriction, E4orf3 is important The cell cycle restriction to adenovirus infection for enhanced S-phase replication of E1b55K and E4orf6 mutant viruses. In addition to their role in countering the DNA damage The cellular activity that imposes the G1 restriction response to virus infection, the E4orf3, E4orf6 and remains unknown. RNA transport may generally be less E1b55K proteins have also been implicated in over- efficient in G1-infected cells compared to S-phase- coming cell cycle restrictions to virus replication. It is infected cells. If so, late viral mRNA transport may possible that these disparate activities are actually require E1b55K/E4orf6 function in G1 cells, whereas connected to each other. Infections with virus mutant late viral mRNA transport in S-phase cells may require in E1b55K and E4orf6 revealed that virus replication only E4orf3 function. Another possibility is that the can be affected by the stage of the cell cycle at the time E1b55K/E4orf6 complex is needed to suppress a cellular of infection. When individual cells infected with these response that is elicited in G1-infected cells. If this viruses were evaluated by electron microscopy or by cellular response is different or less robust in S-phase- assays for infectious centers, approximately one in five infected cells, the response may be suppressed by E4orf3 cells contained progeny virus (Goodrum and Ornelles, alone. A candidate for such a response that is linked to 1997, 1999; Shepard and Ornelles, 2003). Infections in the cell cycle is the DNA damage response. Because S- synchronized cells showed that cells infected during phase cells experience transient single- and double- S phase produced more progeny mutant virus than cells stranded DNA breaks during cellular DNA synthesis, infected during G1 (Goodrum and Ornelles, 1997, 1999). the DNA damage response during S phase is different By contrast, all cells infected with the wild-type virus from that during G1. Indeed, it has been shown that produced progeny virus irrespective of the cell cycle. cells are least sensitive to the DNA-damaging effects of Therefore, the cell cycle can be considered an obstacle to ionizing radiation at late parts of S phase (Sinclair and productive adenovirus infection that is manifested as a Morton, 1966). Adenovirus infection releases linear G1 restriction. double-stranded molecules of DNA into the nucleus The G1 restriction is overcome by the ability of the and consequently infection immediately presents the cell E1b55K/E4orf6 complex to promote late viral gene with molecules that can trigger the DNA damage expression. Analysis of the growth of mutant viruses response. Therefore, if the DNA damage response is bearing in-frame insertions in the E1b55K gene demon- elicited early after infection, this response could be strated that only those viruses that failed to express late considered to be an antiviral response that is more viral genes efficiently were restricted by G1 (Goodrum robustly expressed in the G1-infected cell than in the and Ornelles, 1999). The ability of the E1b55K protein S-phase-infected cell. It may therefore be significant that to inactivate p53 had no impact on the cell cycle the E1b55K/E4orf6 protein complex and the E4orf3 restriction (Goodrum and Ornelles, 1997, 1998). Rather, protein share the ability both to promote late viral gene the molecular basis for the cell cycle restriction appears expression and to block the DNA damage response. to be linked to decreased efficiency of late viral mRNA export in G1 cells. The ratio of cytoplasmic to nuclear viral mRNA was decreased in G1 cells infected with either the E1b55K or E4orf6 mutant virus compared to Regulation of protein translation during adenovirus the wild-type virus (Goodrum and Ornelles, 1999; infection Gonzalez and Flint, 2002). By contrast, the efficiency of late viral mRNA transport in S-phase-infected cells All viruses require the cellular translation apparatus for was identical among cells infected with mutant or wild- the synthesis of viral polypeptides. Therefore, a common type viruses (Goodrum and Ornelles, 1999). The G1 antiviral strategy is to disable the translation process. restriction may also be exerted at the level of translation One mechanism by which this occurs is through (Goodrum and Ornelles, 1999; DA Ornelles, unpub- phosphorylation of Ser51 of the alpha subunit of the lished observations). It is possible that the regulation of eucaryotic initiation factor (eIF-2a). Protein kinases translation in S-phase cells, like that in tumor cells, activated in response to signals associated with virus differs substantially from the regulation of translation in infection and the early events of virus replication G1 or nontumor cells (see the review by Mohr in this phosphorylate eIF-2a, which leads to inactivation issue). The E4orf3 protein is also important for over- of a limiting factor required for initiation of translation,

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7691 eIF-2B. As a result, the rate of translational initiation on only bind a single PKR molecule, thereby competitively typical messenger RNA is reduced. Since the production preventing activation of PKR (Ghadge et al., 1994). of progeny virus requires adequate synthesis of viral Early studies established that VAI RNA mutant viruses proteins, viruses have evolved means to block inactiva- induced high levels of phosphorylated eIF2a, suffered tion of translation (Schneider and Mohr, 2003). reduced viral and host cell translation, and were Adenovirus both prevents the phosphorylation of eIF- defective for replication in some, although not all cells 2a and overcomes antiviral responses that might (Kitajewski et al., 1986). Subsequent studies have otherwise inhibit translation in the early stages of suggested that cells with activated Ras pathways contain infection (This is also discussed by Shmulevitz and Lee an inhibitor of PKR (Mundschau and Faller, 1994) and and by Barber and Mohr in their reviews in this issue.) are not restrictive for the VAI RNA mutant viruses (Cascallo et al., 2003). These tumor cells are defective in The control of translation at early times of infection their ability to mount the early antiviral response to adenovirus that is mediated by PKR. The type I IFNs induce synthesis of the double-stranded RNA-activated protein kinase (PKR). PKR is one of Translation in the late adenovirus-infected cell four that phosphorylate eIF-2a, and is one of many IFN-inducible antiviral effectors regulated by the After the onset of viral DNA replication, the battle over IFN-regulatory family (IRF) of transcription factors. the control of translation is reversed. At early times of These transcription factors are important for expression infection, the host cell is poised to shut off translation by of the IFN genes as well as for expression of several the action of the IFN-inducible enzymes such as PKR. hundred IFN-activated genes (Samuel, 2001). IRF The PKR-mediated shut off of translation is prevented members such as IRF-3 and IRF-7 are latent transcrip- at the transcriptional level by E1a and at the post- tion factors that are phosphorylated by a virus-activated transcriptional level by VAI RNA. However, at late kinase (Sharma et al., 2003) and become incorporated times of infection, adenovirus itself blocks translation. into a cellular complex termed DRAF1. Like PKR, This reversal in the battle over translation prevents DRAF1 is activated by double-stranded RNA (Weaver antiviral responses that depend on new gene expression. et al., 1998). Additional transcription factors important Protection from intracellular pathogens and the sup- to the IFN response include the signal transduction and pression of aberrant cell growth share the need to activator of transcription (STAT) and nuclear factor- recognize and eliminate wayward cells. This common kappaB (NF-kB). These transcription factors and the goal is evident by the convergence of type I IFN DRAF1 complex require chromatin-remodeling proteins signaling and p53-mediated pathways (Hiscott et al., such as p300/CBP for activity. It is this requirement 2003; Pestka, 2003; Takaoka et al., 2003; Munoz- that enables the adenovirus E1a proteins to counter the Fontela et al., 2005). New transcripts that are produced antiviral effects of IFN at multiple steps. in response to signaling through these pathways must be The E1a proteins directly and indirectly block translated in order to have an impact. Many viruses transcription of both IFN genes and IFN-responsive therefore prevent antiviral responses by blocking new genes. The E1a proteins directly bind STAT and NF-kB host gene expression at several steps (Lyles, 2000). and interfere with their ability to stimulate transcription Elements of the cellular DNA damage response to the (Leonard and Sen, 1996; Look et al., 1998). The amino- virus may also be inactivated through regulation of host terminal domain of the E1a proteins binds transcrip- protein translation by adenovirus. At late times of tional coactivators such as CBP/p300, p400, GCN5, infection, products of the E4 region act directly and P/CAF and TRAPP (Frisch and Mymryk, 2002; Lang indirectly to block new host gene expression. In and Hearing, 2003). While bound to E1a, the transcrip- association with the E1b55K protein, E4orf6 directly tional coactivators are no longer available for transcrip- prevents nuclear export of cellular mRNA (Dobner and tion factors such as NF-kB or for the DRAF1 complex. Kzhyshkowska, 2001; Tauber and Dobner, 2001b; Flint Through these interactions, the E1a proteins can and Gonzalez, 2003). By promoting late viral gene suppress transcription of a wide variety of antiviral expression, the E4orf6/E1b55K and E4orf3 proteins effectors. indirectly block host gene expression by increasing Double-stranded RNA permits dimerization of PKR, expression of the nonstructural adenoviral 100K protein which leads to autophosphorylation and activation of L4. The 100K protein serves the dual function of (Kaempfer, 2003). Double-stranded RNA is believed blocking host translation while promoting late viral to arise in the adenovirus-infected cell from transcripts protein translation. synthesized from opposing promoters at early times of infection, although evidence for significant amounts of The L4 100K protein blocks host translation and promotes this double-stranded RNA is lacking. Adenovirus viral translation encodes two small (160 nucleotide) virus-associated RNAs transcribed by RNA polymerase III that The L4 100K protein product blocks translation by antagonize activation of PKR by double-stranded altering the composition and phosphorylation state of RNA (Burgert et al., 2002). Mutational analysis of the the eIF4F protein complex. This multiprotein complex abundant VAI RNA suggests that this highly structured binds to the 50 terminal cap structure of typical RNA contains a double-stranded RNA domain that can eucaryotic mRNA to initiate translation (Kapp and

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7692 Lorsch, 2004; Richter and Sonenberg, 2005). eIF4F virus reflects an intrinsic property of the (Bell includes three initiation factors: eIF4A (an ATP- et al., 2002). Cancer cells are believed to arise from the dependent helicase important for unwinding terminal dysregulation of multiple pathways that regulate cell secondary structures), eIF4E (the cap-binding protein) growth and survival (Hanahan and Weinberg, 2000). and a large scaffold protein eIF4G. Phosphorylated These changes include proliferation in the absence of eIF4E is associated with active translation, whereas extracellular signals, diminished apoptosis, resistance to conditions of diminished eIF4E phosphorylation are growth-suppressive cytokines including IFN and in- associated with reduced translation. At late times of creased rates of protein synthesis. The machinery that adenovirus infection, eIF4E is underphosphorylated recognizes DNA damage and activates cellular check- (Huang and Schneider, 1991). Work of Schneider and points is also inactivated in cancer cells. Adenovirus associates established that the 100K protein displaces expresses genes that achieve these same changes in the Mnk1 from eIF4F, thus preventing phosphorylation of course of a productive infection. For example, the E1a eIF4E (Cuesta et al., 2001). In this way, adenovirus (Frisch and Mymryk, 2002) and E4 proteins (Tauber compromises the ability of the eIF4F cap-binding and Dobner, 2001a) promote cellular proliferation, the complex to initiate translation on most mRNA. E4 proteins potentiate growth signals transmitted Adenovirus late mRNAs escape this block by using through the PI3-kinase pathway (Frese et al., 2003; the modified eIF4F cap-binding complex to initiate O’Shea et al., 2005), the E1b proteins block apoptosis translation by the unusual mechanism of ribosome (Cuconati and White, 2002) and the E4 proteins shunting (Xi et al., 2004). Ribosome shunting allows dismantle the cellular DNA damage response (Stracker adenovirus late mRNAs to escape the block to transla- et al., 2002; Carson et al., 2003). In addition, the E3 tion even though they are capped and polyadenylated in proteins as well as interactions among E1a, E1b and E2 an identical manner to cellular mRNA (Schneider and proteins blunt the response to proinflammatory and Mohr, 2003). The 100K protein also recruits translation death-inducing signals (McNees and Gooding, 2002; factors to the 50 end of late viral mRNA and promotes Fessler et al., 2004; Schaack, 2005). Consequently, functional circularization of the mRNA independently intrinsic properties of cancer cells may obviate the need of the phosphorylation status of the cap-binding protein for specific viral functions that act to reprogram the cell eIF4E (Cuesta et al., 2000). The 100K protein accom- to favor virus replication. These features have been plishes this as part of the modified eIF4F complex exploited by a number of strategies aimed at creating through its ability to bind specifically to late viral oncolytic viruses with increased selectivity for cancer mRNA and the poly(A)-binding protein. The 100K cells. This is discussed in detail by O’Shea in this issue. protein contains both nonspecific (Adam and Dreyfuss, An important characteristic of a replicating, oncolytic 1987; Riley and Flint, 1993) and sequence-specific RNA- virus is the ability to replicate and cause death more binding domains (Xi et al., 2004, 2005). As part of efficiently in tumor cells compared to normal cells. eIF4F, the 100K protein binds the tripartite leader in a Viruses such as ONYX-015 are attractive oncolyic tyrosine phosphorylation-specific manner (Xi et al., agents (Bischoff et al., 1996) that replicate better in 2004, 2005). Therefore, through interactions with viral tumor cells compared to normal cells, reflecting unique mRNA, eIF4G and the poly(A)-binding protein, the properties of tumor cells including cell cycle differences. 100K protein recruits factors required for translation Additionally, both E4orf6 and E1b55K mutant viruses of viral mRNA in the face of the overall inhibition to killed S-phase-infected cells more rapidly than G1- cap-dependent translation. infected cells (Goodrum and Ornelles, 1998; Shepard and Ornelles, 2003; DA Ornelles, unpublished observa- tions). Since a number of reports showed that the Implications for tumor therapy E1b55K mutant viruses replicated within and killed a variety of tumor cells irrespective of p53 status (Goodrum The lessons derived from studies of the natural biology and Ornelles, 1998; Harada and Berk, 1999; Steegenga of adenovirus infections will also have implications for et al., 1999; Dix et al., 2001), it has been suggested that the potential harnessing of adenovirus and its proteins the oncolytic nature of the E1b55K mutant virus may be for cancer therapy. The interactions of adenovirus more closely linked to the cell cycle-restricted nature of products with key cellular regulators, together with the virus rather than the p53 status of the cell. Recently, knowledge of the molecular differences between normal O’Shea et al. (2004) concluded that late viral RNA and tumor cells can be exploited to target death of transport is more efficient in tumor cells than in normal cancer cells. cells. This may suggest that the efficiency of mRNA transport and translation varies in different stages of the Replicating oncolytic adenoviruses cell cycle. The E4 proteins may also be a contributing factor to Adenovirus replication ultimately results in death of the the observed synergy between and infected cell. Although the myriad of paths that lead to either or radiation (Chen et al., 2001; Li cell death are complex, harnessing this cell killing in a et al., 2001; Dilley et al., 2005). The precise reason for tumor-selective fashion is the goal of oncolytic therapies this synergy is unknown, but inactivation of cellular (Bischoff et al., 1996; Kirn, 2000; Dobbelstein, 2004). DNA damage responses by E4 proteins sensitizes tumor The oncolytic nature of many viruses including adeno- cells to the other treatments. Therefore, retaining the E4

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7693 region may be beneficial for these vectors in some responses activated during adenovirus infection are settings. likely to be similar to those encountered by other viruses. The cellular DNA repair machinery probably E4 gene products as antitumor agents recognizes the DNA genomes of many viruses, but what may serve as a barrier for adenovirus may be exploited The identification of DNA repair proteins as targets for to aid the growth of other viruses. For example, recent the E4 gene products has raised the idea of their studies have demonstrated the accumulation of cellular potential use as a novel means to radiosensitize tumor DNA repair factors at the replications centers formed by cells. Cells with mutations in Mre11 and NBS1 are herpes simplex virus (HSV) infection (Taylor and Knipe, hypersensitive to DNA-damaging agents and patients 2004; Wilkinson and Weller, 2004; Lilley et al., 2005). are predisposed to tumor development (Stewart et al., HSV does not appear to inactivate completely the 1999; Kobayashi et al., 2004; Taylor et al., 2004). In damage machinery and may even exploit repair proteins addition, cells deficient in DNA-PKcs activity are to aid its own replication (Taylor and Knipe, 2004; hypersensitive to radiation and other DNA-damaging Lilley et al., 2005). In this case, the Mre11 complex is agents (Burma and Chen, 2004; Meek et al., 2004; Collis also involved in the cellular response (Lilley et al., 2005). et al., 2005). Therefore, inactivation of either the Mre11 The cellular DNA repair apparatus is probably also complex or DNA-PKcs by the E4 proteins could harnessed during integration into the host genome radiosensitize tumor cells. Expression of E4orf6 alone during infection by retroviruses and parvoviruses (Miller has recently been shown to induce radiosensitization of et al., 2004; Skalka and Katz, 2005). The terminal human glioblastoma and colorectal carcinoma cells repeats of parvoviruses are also recognized by repair (Hart et al., 2005). The precise mechanism for the proteins and this may result in activation of a damage sensitization is not clear but it occurred in the absence of response or recombination between viral ends (Raj other viral proteins. It could be due to an inhibitory et al., 2001; Zentilin et al., 2001; Duan et al., 2003). effect on Mre11, DNA-PKcs or another repair factor. Therefore, what may appear as antiviral responses to Sustained autophosphorylation of DNA-PKcs that was one virus may be manipulated in beneficial ways by observed in the presence of E4orf6 (Hart et al., 2005) other viruses. may signal that repair has been incomplete and this Since the synthesis of viral proteins depends com- would trigger growth arrest or cell death. These results pletely on cellular translational machinery, a predomi- suggest that E4orf6 could be used as an adjuvant gene nant innate defense against viral invaders is the therapy approach for solid tumors. wholesale suppression of translation. Therefore, many The E4orf4 gene product has also been proposed as a viruses including adenovirus have evolved mechanisms potential tool in cancer (Branton and to take control of protein translation within the infected Roopchand, 2001). E4orf4 expression is highly toxic and cell. Elaborate strategies have been adopted to ensure it can induce a novel form of p53-independent apoptosis the translation of viral mRNAs while preventing the in cancer cells but not in normal human cells (Marcellus translation of cellular mRNA in order to diffuse et al., 1998; Shtrichman and Kleinberger, 1998; Shtrichman antiviral responses (Thompson and Sarnow, 2000; et al., 1999; Marcellus et al., 2000; Kleinberger, 2004). Schneider and Mohr, 2003). Inactivation of PKR and The exact mechanism for cell killing in mammalian cells the eIF4F translation initiation complex in conjunction is not known, but is thought to require binding to PP2A. with ribosome shunting as an alternative means of E4orf4 has also been linked to dysregulation of Src initiating translation are strategies used by adenovirus family kinases, which could lead to rearrangement of the that are common to other viruses (Schneider and Mohr, actin cytoskeleton and loss of survival signals (Lavoie 2003). et al., 2000). E4orf4 becomes phosphorylated by Src Viral-induced cell death can be the by-product of a kinase, which regulates its subcellular localization and successful lytic infection. The nature of the cell death aspects of the death signal (Gingras et al., 2002). The reflects the balance between the host antiviral responses E4orf4 protein might therefore be exploited as a tumor- and the methods employed by the virus to inactivate specific killing agent, and study of its mechanism of them. Most studies of the adenovirus life cycle have been action might also identify cellular targets to aid in the performed with infections of transformed cells grown in development of novel small molecule anticancer drugs. tissue culture. These cannot accurately reflect all aspects of the regulation of the infectious pathway and the interactions with host responses. The kinetics of Conclusions transcription, gene expression and replication may be very different during the natural in vivo setting of The recent studies linking adenovirus infection to the infection, and genes identified as nonessential for the cellular DNA damage machinery demonstrate that virus may prove otherwise in the natural setting. viruses can be useful model systems for studying aspects In spite of its inherent limitations, the study of of DNA repair. In the case of adenovirus, the targeting adenovirus–host interactions in vitro has provided of the Mre11 complex by E4 and E1b proteins tremendous insight into the inner workings of the demonstrated that the cellular complex can play a role mammalian cell. Cellular proteins that act as key as a sensor of viruses and DNA damage in general regulators of fundamental cellular processes are often (Stracker et al., 2002; Carson et al., 2003). The antiviral the targets of virus countermeasures. Just as viruses

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7694 have been useful tools in studies of replication, Acknowledgements transcription and cell cycle control, their manipulation We thank the members of our labs for lively discussions and of DNA repair and the control of protein translation their contributions to our work on adenovirus manipulation of will highlight important aspects of cell biology. In cell cycle, protein translation and the DNA damage response. addition, understanding the manner in which antiviral We apologize to our colleagues whose original work may not have been cited due to space constraints. Our work on responses have been dismantled in tumor cells will adenovirus–host cell interactions has been supported in part by identify pathways and approaches that can be exploited grants from the National Institutes of Health (AI43341 and in the fight against cancer. CA97093 to MDW, and AI35589 and CA77342 to DAO).

References

Adam SA and Dreyfuss G. (1987). J. Virol., 61, 3276–3283. de Jong RN, van der Vliet PC and Brenkman AB. (2003). Anderson KP and Fennie EH. (1987). J. Virol., 61, 787–795. Curr. Top. Microbiol. Immunol., 272, 187–211. Araujo FD, Stracker TH, Carson CT, Lee DV and Weitzman Dilley J, Reddy S, Ko D, Nguyen N, Rojas G, Working P and MD. (2005). J. Virol., in press. Yu DC. (2005). Cancer Gene Ther., 12, 715–722. Bakkenist CJ and Kastan MB. (2003). Nature, 421, 499–506. Dix BR, Edwards SJ and Braithwaite AW. (2001). J. Virol., 75, Bakkenist CJ and Kastan MB. (2004). Cell, 118, 9–17. 5443–5447. Bartek J and Lukas J. (2001). Curr. Opin. Cell Biol., 13, Dix I and Leppard KN. (1993). J. Virol., 67, 3226–3231. 738–747. Dobbelstein M. (2004). Curr. Top. Microbiol. Immunol., 273, Bell JC, Garson KA, Lichty BD and Stojdl DF. (2002). Curr. 291–334. Gene Ther., 2, 243–254. Dobbelstein M, Roth J, Kimberly WT, Levine AJ and Shenk Ben-Israel H and Kleinberger T. (2002). Front. Biosci., 7, T. (1997). EMBO J., 16, 4276–4284. d1369–95. Dobner T, Horikoshi N, Rubenwolf S and Shenk T. (1996). Bischoff JR, Kirn DH, Williams A, Heise C, Horn S, Muna Science, 272, 1470–1473. M, Ng L, Nye JA, Sampson-Johannes A, Fattaey A and Dobner T and Kzhyshkowska J. (2001). Curr. Top. Microbiol. McCormick F. (1996). Science, 274, 373–376. Immunol., 259, 25–54. Bondesson M, Ohman K, Manervik M, Fan S and Akusjarvi Doucas V, Ishov AM, Romo A, Juguilon H, Weitzman MD, G. (1996). J. Virol., 70, 3844–3851. Evans RM and Maul GG. (1996). Genes Dev., 10, Borden KL. (2002). Mol. Cell. Biol., 22, 5259–5269. 196–207. Boyer J, Rohleder K and Ketner G. (1999). Virology, 263, Duan D, Yue Y and Engelhardt JF. (2003). J. Virol., 77, 307–312. 4751–4759. Boyer JL and Ketner G. (2000). J. Biol. Chem., 275, Evans JD and Hearing P. (2003). J. Virol., 77, 5295–5304. 14969–14978. Evans JD and Hearing P. (2005). J. Virol., 79, 6207–6215. Branton PE and Roopchand DE. (2001). Oncogene, 20, Everett RD. (2001). Oncogene, 20, 7266–7273. 7855–7865. Falgout B and Ketner G. (1987). J. Virol., 61, 3759–3768. Bridge E and Ketner G. (1989). J. Virol., 63, 631–638. Fessler SP, Delgado-Lopez F and Horwitz MS. (2004). Curr. Bridge E, Medghalchi S, Ubol S, Leesong M and Ketner G. Top. Microbiol. Immunol., 273, 113–135. (1993). Virology, 193, 794–801. Flint SJ and Gonzalez RA. (2003). Curr. Top. Microbiol. Burgert HG, Ruzsics Z, Obermeier S, Hilgendorf A, Wind- Immunol., 272, 287–330. heim M and Elsing A. (2002). Curr. Top. Microbiol. Frese KK, Lee SS, Thomas DL, Latorre IJ, Weiss RS, Immunol., 269, 273–318. Glaunsinger BA and Javier RT. (2003). Oncogene, 22, Burma S and Chen DJ. (2004). DNA Repair (Amst.), 3, 710–721. 909–918. Frisch SM and Mymryk JS. (2002). Nat. Rev. Mol. Cell Biol., Carson CT, Schwartz RA, Stracker TH, Lilley CE, Lee DV 3, 441–452. and Weitzman MD. (2003). EMBO J., 22, 6610–6620. Ghadge GD, Malhotra P, Furtado MR, Dhar R and Carvalho T, Seeler J-S, hman K, Jordan P, Pettersson U, Thimmapaya B. (1994). J. Virol., 68, 4137–4151. Akusja¨ rvi G, Carmo-Fonseca M and Dejean A. (1995). J. Gingras MC, Champagne C, Roy M and Lavoie JN. (2002). Cell Biol., 131, 45–56. Mol. Cell. Biol., 22, 41–56. Cascallo M, Capella G, Mazo A and Alemany R. (2003). Gonzalez RA and Flint SJ. (2002). J. Virol., 76, 4507–4519. Cancer Res., 63, 5544–5550. Goodrum FD and Ornelles DA. (1997). J. Virol., 71, 548–561. Cathomen T and Weitzman MD. (2000). J. Virol., 74, Goodrum FD and Ornelles DA. (1998). J. Virol., 72, 11407–11412. 9479–9490. Chen Y, DeWeese T, Dilley J, Zhang Y, Li Y, Ramesh N, Lee Goodrum FD and Ornelles DA. (1999). J. Virol., 73, J, Pennathur-Das R, Radzyminski J, Wypych J, Brignetti D, 7474–7488. Scott S, Stephens J, Karpf DB, Henderson DR and Yu DC. Grand RJ, Grant ML and Gallimore PH. (1994). Virology, (2001). Cancer Res., 61, 5453–5460. 203, 229–240. Collis SJ, DeWeese TL, Jeggo PA and Parker AR. (2005). Halbert DN, Cutt JR and Shenk T. (1985). J. Virol., 56, Oncogene, 24, 949–961. 250–257. Cuconati A and White E. (2002). Genes Dev., 16, 2465–2478. Hanahan D and Weinberg RA. (2000). Cell, 100, 57–70. Cuesta R, Xi Q and Schneider RJ. (2000). EMBO J., 19, Harada JN and Berk AJ. (1999). J. Virol., 73, 5333–5344. 3465–3474. Harada JN, Shevchenko A, Pallas DC and Berk AJ. (2002). Cuesta R, Xi Q and Schneider RJ. (2001). Cold Spring Harb. J. Virol., 76, 9194–9206. Symp. Quant. Biol., 66, 259–267. Hart LS, Yannone SM, Naczki C, Orlando JS, Waters SB, D’Amours D and Jackson SP. (2002). Nat. Rev. Mol. Cell Akman SA, Chen DJ, Ornelles D and Koumenis C. (2005). Biol., 3, 317–327. J. Biol. Chem., 280, 1474–1481.

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7695 Hiscott J, Grandvaux N, Sharma S, Tenoever BR, Servant MJ Orlando JS and Ornelles DA. (1999). J. Virol., 73, 4600–4610. and Lin R. (2003). Ann. NY Acad. Sci., 1010, 237–248. Pestka S. (2003). Cancer Cell, 4, 85–87. Horwitz MS. (2001). Virology, 279, 1–8. Querido E, Blanchette P, Yan Q, Kamura T, Morrison M, Huang JT and Schneider RJ. (1991). Cell, 65, 271–280. Boivin D, Kaelin WG, Conaway RC, Conaway JW and Huang M-H and Hearing P. (1989). J. Virol., 63, 2605–2615. Branton PE. (2001). Genes Dev., 15, 3104–3117. Kaempfer R. (2003). EMBO Rep., 4, 1043–1047. Querido E, Marcellus RC, Lai A, Charbonneau R, Teodoro Kapp LD and Lorsch JR. (2004). Annu. Rev. Biochem., 73, JG, Ketner G and Branton PE. (1997). J. Virol., 71, 657–704. 3788–3798. Kastan MB and Lim DS. (2000). Nat. Rev. Mol. Cell Biol., 1, Raj K, Ogston P and Beard P. (2001). Nature, 412, 914–917. 179–186. Reich N, Pine R, Levy D and Darnell Jr JE. (1988). J. Virol., Kirn D. (2000). Oncogene, 19, 6660–6669. 62, 114–119. Kitajewski J, Schneider RJ, Safer B and Shenk T. (1986). Mol. Richter JD and Sonenberg N. (2005). Nature, 433, 477–480. Cell. Biol., 6, 4493–4498. Riley D and Flint SJ. (1993). J. Virol., 67, 3586–3595. Kleinberger T. (2004). Prog. Mol. Subcell. Biol., 36, 245–267. Rouse J and Jackson SP. (2002). Science, 297, 547–551. Kleinberger T and Shenk T. (1993). J. Virol., 67, 7556–7560. Russell IA, Royds JA and Braithwaite AW. (2004). Prog. Mol. Kobayashi J, Antoccia A, Tauchi H, Matsuura S and Subcell. Biol., 36, 207–243. Komatsu K. (2004). DNA Repair (Amst.), 3, 855–861. Samuel CE. (2001). Clin. Microbiol. Rev., 14, 778–809table of Konig C, Roth J and Dobbelstein M. (1999). J. Virol., 73, contents. 2253–2262. Sandler AB and Ketner G. (1989). J. Virol., 63, 624–630. Kratzer F, Rosorius O, Heger P, Hirschmann N, Dobner T, Schaack J. (2005). Viral Immunol., 18, 79–88. Hauber J and Stauber RH. (2000). Oncogene, 19, 850–857. Schneider RJ and Mohr I. (2003). Trends Biochem. Sci., 28, Lang SE and Hearing P. (2003). Oncogene, 22, 2836–2841. 130–136. Lavin MF. (2004). DNA Repair (Amst.), 3, 1515–1520. Sharma S, tenOever BR, Grandvaux N, Zhou GP, Lin R and Lavoie JN, Champagne C, Gingras MC and Robert A. (2000). Hiscott J. (2003). Science, 300, 1148–1151. J. Cell Biol., 150, 1037–1056. Shenk T. (1996). Fields Virology Vol. 2. Fields BN, Knipe DM, Lavoie JN, Nguyen M, Marcellus RC, Branton PE and Shore Howeley PM et al. (eds). Lippincott-Raven Publishers: GC. (1998). J. Cell Biol., 140, 637–645. Philadelphia, pp. 2111–2148. Leonard GT and Sen GC. (1996). Virology, 224, 25–33. Shenk T and Flint J. (1991). Adv. Cancer Res., 57, 47–85. Leppard KN. (1997). J. Gen. Virol., 78, 2131–2138. Shepard RN and Ornelles DA. (2003). J. Virol., 77, Li Y, Yu DC, Chen Y, Amin P, Zhang H, Nguyen N and 8593–8595. Henderson DR. (2001). Cancer Res., 61, 6428–6436. Shiloh Y. (2003). Nat. Rev. Cancer, 3, 155–168. Lichtenstein DL, Toth K, Doronin K, Tollefson AE and Wold Shtrichman R and Kleinberger T. (1998). J. Virol., 72, WS. (2004). Int. Rev. Immunol., 23, 75–111. 2975–2982. Lilley CE, Carson CT, Muotri AR, Gage FH and Weitzman Shtrichman R, Sharf R, Barr H, Dobner T and Kleinberger T. MD. (2005). Proc. Natl. Acad. Sci. USA, 102, 5844–5849. (1999). Proc. Natl. Acad. Sci. USA, 96, 10080–10085. Liu H, Naismith JH and Hay RT. (2003). Curr. Top. Sinclair W and Morton R. (1966). Radiat. Res., 29, 450–474. Microbiol. Immunol., 272, 131–164. Skalka AM and Katz RA. (2005). Cell Death Differ.. Look DC, Roswit WT, Frick AG, Gris-Alevy Y, Dickhaus Steegenga WT, Riteco N and Bos JL. (1999). Oncogene, 18, DM, Walter MJ and Holtzman MJ. (1998). Immunity, 9, 5032–5043. 871–880. Steegenga WT, Riteco N, Jochemsen AG, Fallaux FJ and Bos Lowe SW and Ruley HE. (1993). Genes Dev., 7, 535–545. JL. (1998). Oncogene, 16, 349–357. Lyles DS. (2000). Microbiol. Mol. Biol. Rev., 64, 709–724. Stewart GS, Maser RS, Stankovic T, Bressan DA, Kaplan MI, Marcellus RC, Chan H, Paquette D, Thirlwell S, Boivin D and Jaspers NG, Raams A, Byrd PJ, Petrini JH and Taylor AM. Branton PE. (2000). J. Virol., 74, 7869–7877. (1999). Cell, 99, 577–587. Marcellus RC, Lavoie JN, Boivin D, Shore GC, Ketner G and Stracker TH, Carson CT and Weitzman MD. (2002). Nature, Branton PE. (1998). J. Virol., 72, 7144–7153. 418, 348–352. Maul GG. (1998). BioEssays, 20, 660–667. Stracker TH, Lee DV, Carson CT, Araujo FD, Ornelles DA McNees AL and Gooding LR. (2002). Virus Res., 88, 87–101. and Weitzman MD. (2005). J. Virol., 79, 6664–6673. Meek K, Gupta S, Ramsden DA and Lees-Miller SP. (2004). Stracker TH, Theunissen JW, Morales M and Petrini JH. Immunol. Rev., 200, 132–141. (2004). DNA Repair (Amst.), 3, 845–854. Miller DG, Petek LM and Russell DW. (2004). Nat. Genet., Takaoka A, Hayakawa S, Yanai H, Stoiber D, Negishi H, 36, 767–773. Kikuchi H, Sasaki S, Imai K, Shibue T, Honda K and Mundschau LJ and Faller DV. (1994). Biochimie, 76, 792–800. Taniguchi T. (2003). Nature, 424, 516–523. Munoz-Fontela C, Angel Garcia M, Garcia-Cao I, Collado M, Takeda T, Nakajima K, Kojima H and Hirano T. (1994). Arroyo J, Esteban M, Serrano M and Rivas C. (2005). J. Immunol., 153, 4573–4582. Oncogene, in press. Tauber B and Dobner T. (2001a). Oncogene, 20, 7847–7854. Nevins JR. (1995). Curr. Top. Microbiol. Immunol., 199 (Part Tauber B and Dobner T. (2001b). Gene, 278, 1–23. 3), 25–32. Taylor AM, Groom A and Byrd PJ. (2004). DNA Repair O’Shea C, Klupsch K, Choi S, Bagus B, Soria C, Shen J, (Amst.), 3, 1219–1225. McCormick F and Stokoe D. (2005). EMBO J., 24, Taylor TJ and Knipe DM. (2004). J. Virol., 78, 5856–5866. 1211–1221. Thompson SR and Sarnow P. (2000). Curr. Opin. Microbiol., O’Shea CC. (2005). Curr. Opin. Genet. Dev., 15, 18–26. 3, 366–370. O’Shea CC, Johnson L, Bagus B, Choi S, Nicholas C, Shen A, Tollefson AE, Scaria A, Hermiston TW, Ryerse JS, Wold LJ Boyle L, Pandey K, Soria C, Kunich J, Shen Y, Habets G, and Wold WS. (1996). J. Virol., 70, 2296–2306. Ginzinger D and McCormick F. (2004). Cancer Cell, 6, Wahl GM and Carr AM. (2001). Nat. Cell Biol., 3, 611–623. E277–E286.

Oncogene Adenovirus infection and antiviral responses MD Weitzman and DA Ornelles 7696 Weaver BK, Kumar KP and Reich NC. (1998). Mol. Cell. Xi Q, Cuesta R and Schneider RJ. (2005). J. Virol., 79, Biol., 18, 1359–1368. 5676–5683. Weiden MD and Ginsberg HS. (1994). Proc. Natl. Acad. Sci. Yew PR and Berk AJ. (1992). Nature, 357, 82–85. USA, 91, 153–157. Yoder SS and Berget SM. (1986). J. Virol., 60, 779–781. Weinberg DH and Ketner G. (1986). J. Virol., 57, 833–838. Zentilin L, Marcello A and Giacca M. (2001). J. Virol., 75, Wilkinson DE and Weller SK. (2004). J. Virol., 78, 4783–4796. 12279–12287. Xi Q, Cuesta R and Schneider RJ. (2004). Genes Dev., 18, Zhou BB and Elledge SJ. (2000). Nature, 408, 433–439. 1997–2009. Zou L and Elledge SJ. (2003). Science, 300, 1542–1548.

Oncogene