MIAMI UNIVERSITY The Graduate School

CERTIFICATE FOR APPROVING THE DISSERTATION

We hereby approve the Dissertation

of

Sumithra Jayaram

Candidate for the Degree:

Doctor of Philosophy

______Dr. Eileen Bridge, Director

______Dr. Gary Janssen, Reader

______Dr. Anne Morris Hooke

______Dr. Mary E. Woodworth

______Dr. Qingshun Quinn Li, Graduate School Representative ABSTRACT

INVESTIGATING ADENOVIRUS INTERACTIONS WITH HOST DOUBLE- STRAND BREAK REPAIR DEFENSES By Sumithra Jayaram

The goal of this study was to investigate the role of host double-strand break repair (DSBR) on the life cycle of Adenovirus (Ad). Ad mutants that lack the entire E4 region activate a cellular DNA damage response accompanied by phosphorylation of several host DSBR and DNA damage response proteins. We find that aspects of the E4 mutant induced DNA damage response occurs at the onset of viral DNA replication and may be activated by physical replication of viral genomes. Genetic analysis of the E4 mutants revealed that the E4-34kDa protein was required to prevent the activation of DNA damage response. Redistribution of MRN complex proteins away from viral replication centers by the E4-11kDa protein was not sufficient to prevent the DNA damage response. Activation of the DNA damage response does not interfere with viral DNA replication in the presence of E4-11kDa protein. E4 mutants are severely defective for late gene expression following concatenation of their genomes by host DSBR proteins. We find that E4 mutant late gene expression improves in MO59J cells that fail to form genome concatemers. DSBR kinase inhibitors interfere with genome concatenation and also stimulate late gene expression. Concatenation of E4 mutant genomes interferes with cytoplasmic accumulation of viral late messages and leads to reduced late protein levels and poor viral yields following high multiplicity infection. However, failure to concatenate viral genomes did not rescue either the DNA replication defect or virus yield following low multiplicity E4 mutant infection. Our results indicate that if the E4 mutant DNA replication defect is overcome by high multiplicity infection, concatenation of the replicated genomes by host DSBR interferes with viral late gene expression. These studies provide insight into the role of host DSBR as an obstacle to a productive Ad infection, and how the virus dismantles this barrier. INVESTIGATING ADENOVIRUS INTERACTIONS WITH HOST DOUBLE-STRAND BREAK REPAIR DEFENSES

A DISSERTATION

Submitted to the faculty of Miami University in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Microbiology

by

Sumithra Jayaram Miami University Oxford, Ohio 2005

Dissertation Director: Dr. Eileen Bridge Table of Contents Page INTRODUCTION…………………………………………………………….. 1 Adenovirus life cycle…………………………………………………………… 1 Role of E4 region in Ad life cycle……………………………………………… 5 Role of host DSBR proteins in the Ad life cycle……………………………….. 7 E4 proteins interfere with cellular DNA damage response and prevent genome concatenation…………………………………………………………………… 16

CHAPTER 1: Investigating the induction of the cellular DNA damage response by an Adenovirus E4 mutant……………………………………………………… 18 Abstract………………………………………………………………………….. 19 Introduction……………………………………………………………………… 20 Materials and Methods…………………………………………………………… 23 Results…………………………………………………………………………… 25 Discussion……………………………………………………………………….. 45

CHAPTER 2: Genome concatenation contributes to the late gene expression defect of an Adenovirus E4 mutant……………………………………………….. 49 Abstract………………………………………………………………………. 50 Introduction………………………………………………………………….. 51 Materials and Methods………………………………………………………. 54 Results……………………………………………………………………….. 57 Discussion…………………………………………………………………… 77 CONCLUDING REMARK……………………………………………….. 82 Activation of the DNA damage response affects viral DNA replication and gene expression………………………………………………………………….. 86

REFERENCES………………………………………………………………. 95

ii List of Figures Page Figure 1 Adenovirus genome organization and detailed map of the E4 2 transcription unit.

Figure 2 Schematic representation of the cellular response to DNA DSBs. 9

Figure 3 Schematic representation of the pathway of DNA NHEJ in 14 eukaryotes.

Figure 4 E4 mutant does not activate a DNA damage response at low 26 multiplicities.

Figure 5 E1a- Ad β-gal does not induce a DNA damage response. 29

Figure 6 E4 mutant-induced DNA damage response correlates with the 32 onset of viral DNA replication.

Figure 7 Redistribution of the MRN complex members by E4 11kDa 37 protein does not prevent the activation of a DNA damage response.

Figure 8 The activation of the DNA damage response does not affect 40 H5dl1010 DNA replication.

Figure 9 DNA-PKcs is not required for the activation of DNA damage 43 signaling.

Figure 10 E4 mutant late protein synthesis improves in MO59J cells that 58 lack DNA-PKcs.

iii Figure 11 Inhibition of DSBR kinases interferes with E4 mutant genome 61 concatenation.

Figure 12 Inhibition of DSBR kinase activity partially rescues the late 64 protein defect of an E4 mutant virus.

Figure 13 The effect of DSBR kinase inhibitors on the DNA damage 67 response induced by E4 mutant infections.

Figure 14 E4 mutant genome concatenation interferes with viral late 70 mRNA accumulation.

Figure 15 The DNA replication defect of E4 mutants at low MOI is 74 not rescued by the failure to concatenate viral genomes.

Figure 16 A model for the role of host DSBR proteins on the life 93 cycle of an Adenovirus E4 mutant.

iv List of Tables Page

Table 1 Summary of the genotype of E4 mutants used in this study 36

Table 2 Virus yield following high and low multiplicity infection 76 of MO59J and MO59K cells.

v Acknowledgements I would like to express my deepest gratitude to my advisor, Dr. Eileen Bridge, for her excellent guidance, and support over the past years. Your are an amazing role model for women in science. Throughout my doctoral work you have encouraged me to develop independent thinking and research skills. Thank you for teaching me to be good scientist and greatly assisting me with scientific writing. I am also very grateful for having an exceptional doctoral committee and wish to thank Dr. Gary Janssen, Dr. Anne Morris Hooke, Dr. Mary E. Woodworth and Dr. Qinshun Li for their continual support and encouragement. I would specially like to thank Dr. Janssen for his words of wisdom during bleaker times when I needed to hear them. Thank you for always believing in me and encouraging me to do well in my scientific career. Thank you to my current lab mate Shomita Mathew. I have enjoyed working with you and wish you all the very best in your scientific pursuits. I also wish to thank my past lab mates Arunima and Kara Corbin-Lickfette, for their support during my initial phases of doctoral studies. I would like to thank the people of the department of microbiology for all your help in the past years. I will always be eternally be grateful for the opportunity given to me to come all the way from India to pursue my doctoral degree. Finally, I would like to thank my mom and dad, to whom I dedicate this dissertation. It was their dream and I am very happy to be able to fulfill this for them. Last but definitely not the least, my heartfelt thanks to my loving husband, Pradeep Dinakar, without whose support I would have not be able to complete my doctoral studies successfully. I hope I can always be equally supportive of your goals in life.

vi INTRODUCTION Adenovirus (Ad) is used extensively as a model system for the study of DNA replication, gene expression and cell transformation. Ad is a double-stranded DNA virus whose replication cycle is largely dependent on the activity of cellular factors. The virus life cycle is divided into early and late phases separated by the onset of viral DNA replication. A set of early proteins initiates DNA replication, which is required for viral late gene expression. During the late phase, host cell machinery is used for the production of viral late messages and the structural proteins that associate with the viral DNA to form mature viral particles in the nucleus of the infected cells (reviewed in Pettersson and Robert, 1986). A. Adenovirus life cycle: Early phase. Adenovirus attaches with high efficiency to cellular receptors via the fiber protein (Chroboczek et al., 1995). The primary receptor for the human Ad is identical to that for coxsackie B virus and has therefore been termed the coxsackie-adenovirus receptor (CAR) (Tomko et al., 2000). The early phase of infection is characterized by the production of more than 20 regulatory proteins encoded by five early transcription units: early region 1a (E1a), E1b, E2, E3 and E4 (Fig. 1). Each unit produces multiple differentially spliced mRNAs encoding a variety of distinct polypeptides that are required to establish an optimal environment for efficient viral DNA replication and subsequent expression of viral late genes. After the virus uncoats and its genome enters the nucleus, the E1a promoter is immediately recognized by the cellular transcription machinery. The E1a gene products are required for efficient transactivation from the other early promoters (E1b, E2, E3 and E4). The E1a products also deregulate normal cell cycle control allowing quiescent cells to enter the synthesis phase (S-phase) of cell cycle, thereby creating an environment conducive to viral DNA replication. Gene products from early region 1b (E1b) prevent apoptosis and promote viral late gene expression. The E2 products are required for viral DNA replication and play a role in both replication initiation and DNA chain elongation. E3 products help the virus evade the immune system of the host by counteracting early host cell defense mechanisms but are expendable in cell culture. E4 products are required for normal DNA and late mRNA

1 Figure 1. Adenovirus genome organization and detailed map of the E4 transcription unit. The Adenovirus double-stranded DNA genome is shown with the terminal protein attached to the 5’end of each strand. Early transcription units are designated E1a, E1b, E2a and E2b, E3 and E4. The major late transcription unit (MLTU) is indicated by the thicker black arrow and extends the entire length of the genome. A detailed map of the E4 open reading frame is shown. The use of multiple splice acceptor sites within the primary transcript determines which open reading frame is translated. E4orf3/4 and E4orf6/7 utilize an additional splice donor site and have sequences corresponding to both open reading frames.

2 3 Ad DNA replication. As E2 gene products accumulate and the infected cell enters the S- phase of the cell cycle, the stage is set for viral DNA replication. The Ad double-stranded DNA genome is 35-Kbp with a covalently attached terminal protein on the 5’ end of each strand. Ad replicates its DNA by strand displacement and a protein priming mechanism, using its own DNA polymerase (DNA pol) and pre-terminal protein (pTP) encoded by the E2 transcription unit. Cellular transcription factors are recruited to Ad replication origins and facilitate the separation of the DNA strands. Ad DNA pol associated with the pTP recognizes the origin of replication and covalently attaches a cytosine to a serine residue on the pTP. The cytosine is complementary to the first guanine at the 3’ end of the Ad genome. Synthesis of a complementary strand in the 5’ to 3’ direction initiates from the 3’OH of the cytosine and the original 5’ DNA strand is displaced as DNA pol continues polymerizing the complementary strand. The displaced single-stranded DNA strand can self-associate through inverted terminal repeats to create an origin of replication recognized in the same fashion as described above (reviewed in Challberg and Kelly, 1989). Late gene expression. The Ad late coding regions are organized into a single large transcription unit referred to as major late transcription unit (MLTU) whose primary transcript is about 29,000 nucleotides. This transcript is processed by differential poly(A) site utilization and splicing to generate at least 18 different mRNAs. These mRNAs have been grouped into five families termed L1 to L5, based on the utilization of common poly(A) addition sites. Once a polyadenylation site is chosen, there are multiple 3’ splice acceptor sites within that transcript that can be used to make the mRNAs for the structural proteins needed for virion assembly (reviewed in Leppard, 1998). Expression of this large family of late mRNAs is controlled by the major late promoter (MLP). This promoter exhibits a low level activity early after infection, and becomes several hundredfold more active at late times. All the mRNAs transcribed from the MLP contain a common 5’ non-coding region of 201 nucleotides termed the tripartite leader (TPL), which is derived by the splicing of three small exons. Within a few hours of the onset of the late phase, several complex metabolic changes occur that ensure preferential synthesis of viral late proteins and efficient assembly of progeny virions. Late mRNAs are preferentially exported to the cytoplasm where they are translated to the exclusion of the

4 host mRNAs. In addition cytoplasmic accumulation of most cellular mRNAs is blocked. Viral mRNAs are preferentially translated due to protein kinase R (PKR) induced phosphorylation of eIF-4F. As a result, large quantities of viral structural polypeptides are produced while cellular protein synthesis is shutoff almost completely (reviewed in Dobner and Kzhyshkowska, 2001). Virus assembly and release from the cell. The replication of viral DNA coupled with the production of large quantities of the Ad structural polypeptides sets the stage for virus assembly. The assembly of the trimeric hexon capsomeres occurs rapidly from monomers after their synthesis in the cytoplasm. Assembly of the hexon protein requires the participation of a second viral late protein, termed L4-100kDa. The L4-100kDa protein acts as a scaffold to facilitate the assembly of hexon trimers by an unknown mechanism. Viral late penton capsomeres consisting of a pentameric penton base and trimeric fiber assemble more slowly in the cytoplasm. After their production the hexon and the penton capsomeres accumulate in the nucleus where the assembly of the virion occurs. Assembly begins with the formation of an empty capsid and subsequently the viral DNA enters the capsid by means of the cis-acting packaging sequence located in the left end of the viral chromosome. Finally, the escape and the spread of the progeny virus occur by the disruption of the host cytoskeleton (reviewed in Shenk, 1996). B. The role of E4 region in Ad life cycle: Ad E4 region: organization and protein products. E4 produces a diverse set of proteins involved in a variety of processes crucial to virus growth. The Ad serotype 5 (Ad5) E4 transcription unit is located between map units 91.3 to 99.1 at the right-hand end of the Ad genome and is transcribed in the leftward direction. The E4 transcription unit is controlled by the E4 promoter and generates a primary transcript approximately 2800 nucleotides. This transcript is subject to a complex pattern of differential splicing producing at least 18 distinct mRNAs that share common 5’ and 3’ terminal sequences. These mRNAs potentially encode seven different polypeptides: the products of open reading frame (ORF) 1, ORF2, ORF3, ORF3/4, ORF4, ORF6 and ORF6/7 (Fig. 1). Ad5 E4 ORF1 mRNA accumulates late after infection but the expression of an E4 ORF1 protein during infection has only been described in Ad9 and Ad26. The Ad9 E4 ORF1 protein alone causes estrogen-dependent mammary tumors in rats. The E4 ORF2 protein

5 is produced at early times post-infection but has no known function. E4 ORF3/4 mRNA is low in abundance during infection and the protein has not been detected in infected cells. The E4 ORF4 protein product interacts with the cellular protein phosphatase 2A thereby inhibiting signal transduction, viral and cellular gene expression, and inducing p53-dependent apoptosis. The E4 ORF4 protein product has also been implicated in regulating late mRNA splicing by dephosphorylating serine and arginine rich (SR) family of splicing factors. The E4 ORF6/7 protein interacts with the cellular E2F transcription factor and promotes expression of other viral transcription units. The E4 ORF6-encoded 34kDa and E4 ORF3-encoded 11kDa protein products have redundant functions in promoting efficient viral late gene expression and each is sufficient to substitute for the whole E4 region during Ad lytic infection of HeLa cells (reviewed in Tauber and Dobner, 2001). The phenotype of E4 deletion mutants is discussed in detail in the next section. Phenotypes of E4 deletion mutants. Ad mutants that lack E4 show a very complex phenotype, which includes defects in viral DNA accumulation, late viral mRNA accumulation and protein synthesis, as well as a failure to shut off host protein synthesis (Halbert et al., 1985; Weinberg and Ketner, 1986; Bridge and Ketner, 1989; Sandler and Ketner, 1989; Huang and Hearing, 1989, Jayaram and Bridge, 2005). Ad mutants that are unable to express either E4-34kDa or E4-11kDa proteins exhibit reduced levels of viral late RNA in the nucleus indicating that these proteins provide functions that improve the stability of viral late RNAs (Bridge and Ketner, 1989; Sandler and Ketner, 1989). The reduction in the stability of the viral late RNAs in the nuclei of E4 mutant infected cells is responsible for the failure of late messages to accumulate in the cytoplasm leading to the late gene expression defect. This deficiency is not due to reduced transcriptional activity of the major late promoter (MLP) in E4 mutant infected cells (Sandler and Ketner, 1989). More likely the E4 products are required for stable nuclear accumulation of the major late pre-mRNA (reviewed in Imperiale, 1995). Both the E4-34kDa and the E1b-55kDa proteins have leucine-rich nuclear export signals (NES) of the HIV-1 Rev type and shuttle between the nucleus and cytoplasm during Ad infection (Dobbelstein et al., 1997; Dosch et al., 2001; Kratzer et al., 2000; Rabino et al., 2000). However, the ability to shuttle between the nucleus and the cytoplasm is not critical for either protein to promote Ad late gene expression in complementation assays (Rabino et al., 2000; Carter et al.,

6 2003). This raises the possibility that other functions of these proteins could contribute to the effects of these early proteins in viral late gene expression. E4 mutants are delayed for viral replication particularly at low multiplicity of infection (MOI) and early in the viral growth cycle (Halbert et al., 1985; Weinberg and Ketner, 1983). This E4-associated defect disappears at high MOI’s and at late times (24hrs) after infection. Both the E4-11kDa and E4-34kDa proteins are involved in promoting efficient viral DNA replication in Ad infections (Bridge and Ketner, 1989; Evan and Hearing, 2003; Evan and Hearing, 2005). The late gene expression defect of E4 deletion mutants is independent of a DNA replication defect in infections done at high multiplicities (Bridge and Ketner, 1989; Jayaram and Bridge, 2005). E4 mutants induce a DNA damage response in host cells similar to that caused by ionizing radiation or genotoxic stress (Carson et al., 2003). The genomes of E4 mutants are also concatenated by the activity of the host double-strand break repair (DSBR) proteins. Recent observations show that the E4-34kDa and E4-11kDa proteins can individually prevent genome concatenation by interfering with the activity of the host DSBR proteins (Weiden and Ginsberg, 1994; Boyer and Ketner, 1999; Stracker et al., 2002). It is possible that many activities attributed to the E4 products in viral DNA, late RNA and protein accumulation are the result of their role in inactivating the host DSBR proteins. In the next section, I will discuss in detail the role of E4 proteins in preventing DNA damage response and genome concatenation and its consequence for virus gene expression. C. Role of host DSBR proteins in the Ad life cycle: Host DSBR proteins participate in the DNA damage response. DNA double strand breaks (DSBs) pose a continuous threat to genomic integrity of mammalian cells. Cells respond to DNA DSBs through systems that detect the DNA lesion and then trigger various downstream events collectively known as the “DNA damage response.” These systems can be viewed as classical signal-transduction cascades in which a “signal” (DNA damage) is detected by a “sensor” (DNA-damage binding protein) that then triggers the activation of a “transducer” system (protein kinase cascade), which amplifies and diversifies the signal by targeting a series of downstream “effectors” of the DNA damage response. Clearly, such systems need to be exquisitely sensitive and selective, as

7 they must be triggered rapidly and efficiently by chromosomal DNA double strand breaks (DSB’s). The cellular response to DSBs is to activate the host DSBR proteins, which are then physically recruited to the site of the DNA lesion to bring about its repair. In addition, dividing cells respond to DNA DSB’s by slowing down progression through the cell cycle. This presumably provides time to allow DNA repair to occur before a replicative DNA polymerase encounters the lesions. Such pauses in cell-cycle progression are termed “cell-cycle checkpoints” (reviewed in Jackson, 2002). The importance of sensing and responding to DSB’s can be seen through the profound phenotypes associated with mutations in genes encoding host double strand break repair (DSBR) proteins that participate in the cellular DNA damage response. For example, in humans, mutations inactivating the genes encoding the host DSBR proteins causes genetic disorders associated with genomic instability and immunodeficiency, all leading to cancer predisposition (reviewed in Khanna and Jackson, 2001). A cartoon describing a model for the cellular DNA damage response is shown in Fig. 2. The accumulating evidence indicates that DNA breaks are “sensed” by the Mre11- Nbs1-Rad50 (MRN) complex, which binds DNA, unwinds the ends, and recruits the protein kinase, ataxia telangiectasia mutated (ATM) to the site of DSB (reviewed in Jackson, 2002). A crucial component of the DNA DSB signaling cascade in mammalian cells is ATM. Autophosphorylation and activation of ATM kinase is one of the earliest characterized events in the response to DSBs. Autophosphorylation of ATM at serine- 1981 leads to dimer dissociation and it has been proposed that this leads directly to the

8 Figure 2. Schematic representation of cellular response to DNA DSB’s. DNA breaks are “sensed” by the MRN complex, which binds DNA, unwinds the ends, and recruits ATM to the site of the DSB. Autophosphorylation of ATM at serine-1981 leads to dimer dissociation. This leads directly to the release of active ATM monomers that modify downstream effector molecules like Chk-1 and Nbs1. ATR is another DNA-damage surveillance protein that phosphorylates a set of target proteins overlapping with ATM. A cascade of phosphorylation events finally leads to DNA repair and cell cycle checkpoint induction. If the DNA damage is too severe to repair, cells may undergo apoptosis.

9 10 release of active ATM monomers that modify downstream effector molecules such as Nbs1, 53BP1, Chk1, H2AX, NFBD1/MDC1, and BRCA1 leading to a variety of effects on DNA repair, cell-cycle progression and apoptosis (Bakkenist and Kastan, 2003; Kurz and Lees-Miller, 2004). Nbs1 is part of the Mre11-Rad50-Nbs1 (MRN) complex, which is essential for DNA DSBR and genomic stability. In eukaryotic cells the MRN complex is required for both ATM activation and the ability of ATM to phosphorylate and activate a number of proteins involved in DNA repair and checkpoint signaling (Petrini and Stracker, 2003). Another DNA damage surveillance protein that is related to ATM is ataxia telangiectasia mutated and Rad3 related (ATR) (Tibbetts et al., 2000). Disruption of the gene for ATR leads to early embryonic lethality in the mouse. The reason for this lethality is not yet clear but is likely to reflect a role for ATR in the recognition and repair of DNA replication complexes that have stalled at sites of DNA damage. The available evidence indicates that ATR and ATM phosphorylate an overlapping set of target proteins (Zou and Elledge, 2003). E4 deletion mutants activate the cellular DNA damage response. E4 deletion mutant infections induce a cellular DNA damage response. Analysis of phosphorylation events characteristic of a damage response during E4 mutant Ad infections has revealed that both ATM and ATR signaling cascades are activated during E4 mutant infection. Wild- type Ad targets Mre11 for proteasome-mediated degradation. Degradation of the Mre11 protein by E4-34kDa and E1b-55kDa complex prevents ATM and ATR signaling. Hence, these data demonstrate a requirement for the MRN complex in ATM and ATR signaling in the damage response. Further, the DNA damage response is abrogated in A-TLD1 cells that lack a functional Mre11 protein (Carson et al., 2003). It is unclear as to what DNA structures are required for the activation of the MRN complex and damage response signaling in E4 mutant infections. Replication of a double- stranded DNA virus like Ad produces numerous linear DNA templates with double- stranded ends, which could be recognized as DSBs in mammalian cells. The presence of unusual replication intermediates and single-stranded DNA could also activate the MRN complex (Wang et al., 2000; Weitzman et al., 2004). The eukaryotic cell responds to DSBs by recruiting host DSBR proteins to the sites of the DNA lesion in a dynamically organized and timely manner. These proteins form a multi-component assembly, known

11 as a focus, in response to damage induced by ionizing radiation (Lisby and Rothstein, 2005). Analogous foci containing the host DSBR proteins are formed at the sites of viral DNA replication in E4 mutant infections (Stracker et al., 2002; Carson et al., 2003). During infection with wild-type Ad (when the MRN complex is degraded) many host DSBR proteins do not accumulate at viral replication centers. These data suggests that the MRN complex can function as an upstream “sensor” of Ad infections leading to induction of DNA damage response and genome concatenation in the absence of the E4 proteins (Carson et al., 2003; Stracker et al., 2002). E4 mutant genomes are concatenated by the activity of host DSBR proteins. Genomes of the Ad E4 mutants that are unable to express either E4-34kDa protein or E4- 11kDa protein are concatenated by the host DSBR proteins (Weiden and Ginsberg, 1994; Boyer and Ketner, 1999; Stracker et al., 2002). These concatemers can be detected by pulsed field gel electrophoresis (PFGE) of DNA from infected cells. The concatemers that arise in E4 mutant infections do not have any specific orientation of monomer joining. The junctions are covalently linked molecules that are not precisely joined as revealed by the heterogeneous size distribution of junction fragments. In contrast, DNA extracted from cells infected with wild type Ad5 shows viral genomes exclusively in a linear monomeric form. The E4-34kDa and E4-11kDa protein products can individually prevent the concatenation of viral genomes in infected cells. Recent studies have addressed the role of host DSBR proteins in formation of viral genome concatemers during Ad infections. Concatemer formation has been assessed in mutant human cell lines defective for host DSBR proteins. These studies have identified two sets of host DSBR proteins needed for concatenating the E4 mutant genomes. The members of the non-homologous end-joining (NHEJ) pathway, including the catalytic subunit of the enzyme DNA-dependent protein kinase (DNA-PKcs) and DNA ligase IV are needed for preventing genome concatenation. In addition, the conserved multi-protein MRN complex is also required for concatenating E4 mutant genomes. The MRN complex is a central player sensing DNA damage including DSB’s (Fig. 2), and also functions directly in repair by NHEJ (Boyer et al., 1999; Stracker et al., 2002; Carson et al., 2003; Petrini and Stracker, 2003). In the next section I will discuss in detail the mechanism of NHEJ in eukaryotes.

12 Mechanism of NHEJ repair of DSB’s. NHEJ is the primary mechanism of DSB’s repair in mammalian cells (Jeggo et al., 1998). The basic mechanism and factor requirements of NHEJ are described in Fig. 3. Central to NHEJ in organisms from yeast to humans is the DNA-dependent protein kinase (DNA-PK) holoenzyme consisting of the regulatory Ku proteins and the catalytic subunit known as DNA-PKcs. The Ku protein is a heterodimer of two subunits, Ku70 and Ku80 (Critchlow and Jackson, 1998). Biochemical studies of mammalian Ku have revealed that it is activated in the presence of DSBs and binds to DNA in a sequence-independent fashion (Dynan and Yoo, 1998). The crystal structure of Ku has revealed that it forms an open ring-type structure that can be threaded onto a DNA end. One side of the ring forms a cradle that protects one surface of the DNA double helix, whereas the other side is much more open, presumably to allow other NHEJ factors to access the DSB. Ku serves as the DNA targeting subunit of the DNA-PK holoenzyme (Burma and Chen, 2004). DNA-PKcs is a ~465kDa polypeptide, the C-terminal region of which has homology to the catalytic domains of proteins of the phosphatidyl inositol 3- kinase-like family. Strikingly, this family includes two other mammalian proteins that participate in DNA damage response, ATM and ATR. DNA-PKcs itself has affinity for DNA ends and its activation appears to be triggered by its interaction with a single- stranded DNA region derived from a DSB (Hammarsten et al., 2000). Once bound to DNA DSBs, DNA-PK displays protein Serine/Threonine kinase activity. One of the in vivo substrates for DNA- PK includes the XRCC4 protein that is a part of the DNA ligase IV complex. Biochemical studies in the mammalian cells have shown that Ku can load the XRCC4/ligase IV complex onto DNA ends and stimulate DNA end-ligation. DNA- PKcs phosphorylates XRCC4 and removes or relocates the DNA ligase IV/XRCC4 complex from Ku-bound DNA ends, thus allowing necessary processing steps to occur (Chan et al., 1996; Calsou et al., 1999; Chan et al., 1999; McElhinny et al., 2000). DNA ligase IV finally brings about DNA end joining in NHEJ pathway. Most DNA DSBs that are generated by mutagenic agents cannot be directly religated, instead some limited processing of the ends must take place before NHEJ can occur.

13 Figure 3. Schematic representation of the pathway of DNA NHEJ in eukaryotes. The repair of two broken DNA ends by NHEJ is shown. It requires the regulatory Ku proteins, which are shown binding to the free ends. Ku proteins then recruit the catalytic subunit, DNA-PKcs. Ku also recruits XRCC4/DNA ligase IV complex and DNA-PKcs- mediated phosphorylation of XRCC4 may influence its activity. The Mre11–Rad50–Nbs1 complex, which contains exonuclease, endonuclease and helicase activities is required for processing the ends before ligation. NHEJ typically results in removal of sequence from the ends of the double strand break to expose regions of micro homology prior to ligation of the ends. Ligase IV finally brings about the physical religation of the DNA ends.

14 15 Consequently, NHEJ is rarely error-free and sequence deletions of various lengths are usually introduced. One candidate for an enzyme involved in the nucleolytic processing stages of NHEJ is the mammalian MRN complex. This complex possesses exonuclease, endonuclease and DNA unwinding activities in vitro and has been shown by immunofluoresence studies to localize to sites of DNA DSBs in mammalian cells (Maser et al., 1997; Paull and Gellert, 1999) D. E4 proteins interfere with the cellular DNA damage response and prevent genome concatenation by NHEJ. The MRN complex is also targeted for inactivation by the E4 11kDa and E4 34kDa proteins during Ad infection. Immunofluoresence studies have indicated two ways in which the MRN complex is altered during wild-type Ad infections. In uninfected cells, the MRN complex is diffusely localized throughout the nucleus. In E4 mutant infections the MRN complex is localized to the periphery of the viral replication centers. Expression of the E4-11kDa protein prevents genome concatenation by redistribution of the members of the MRN complex into track-like formation away from viral replication centers. Additionally, the complex formed between the E4-34kDa and the E1b-55kDa proteins prevents genome concatenation by targeting the MRN complex for proteasome-mediated degradation. Either relocalization (E4-11kDa) or degradation (E4-34kDa/ E1b- 55kDa) is sufficient to prevent the host DSBR proteins from producing concatemers of Ad DNA (Stracker et al., 2002). The physical binding of E4-11kDa and E4-34kDa proteins to DNA-PKcs may also prevent genome concatenation in wild type Ad infections (Boyer and Ketner, 1999). Inactivation of the DNA-PKcs enzyme has additional benefits in the virus life cycle. DNA-PKcs is required for activation of p53 DNA-binding activity following ionizing radiation (Woo et al., 1998). p53 is a tumor suppressor protein that commonly responds to DNA damage signals by inducing apoptosis. Therefore, by interfering with DNA-PKcs function, E4 products may prevent activation of p53, and thus apoptosis. E4-34kDa and E1b-55kDa also prevent apoptosis by targeting p53 for proteasome-mediated degradation (White, 1995) Prevention of early apoptotic cell death contributes to the efficiency of viral infections (Hardwick, 1998) The formation of genome concatemers may interfere with a productive Ad infection at several steps. The concatemers that arise in E4 mutant infections are

16 covalently linked DNA molecules that are imprecisely ligated in random head-to-head, head-to-tail and tail-to-tail orientation (Weiden and Ginsberg, 1994). Extensive genome concatenation could destroy the viral origin of replication located at the ends of the linear genomes. The Ad packaging sequences do not function when located more than a few hundred base pairs from the genomic termini (Hearing et al., 1987). Concatemers would not be good packaging substrates since they are too large to be packaged into viral capsids. The importance of inactivating the host DSBR proteins during Ad infection is highlighted by evolution of redundant strategies to prevent the concatenation of viral DNA. It is also possible that E4 proteins promote efficient viral late gene expression by inhibiting the host cell’s ability to concatenate viral DNA genomes. Compensatory roles for the E4-11kDa and E4-34kDa proteins in both processes are consistent with this hypothesis (Bridge and Ketner, 1989; Stracker et al., 2002). In summary, the E4-34kDa and E4-11kDa proteins have a diverse range of activities within the host cell, but either is sufficient to substitute for the whole E4 region during Ad lytic infection of HeLa cells. This redundancy suggests there might be a common mechanism by which the E4 proteins promote productive infection. In this study, I have characterized the role of the host DSBR proteins in Ad life cycle. My results indicate that viral replication is required to induce at least some aspects of the host DNA damage response (Chapter 1), and that E4 mutant genome concatenation by host DNA repair proteins interferes with accumulation of viral late RNA’s, leading to a defect in late gene expression and virus yield (Chapter 2). These studies provide insight into the role of host DSBR as an obstacle to a productive Ad infection, and how the virus dismantles this barrier.

17 CHAPTER 1: Investigating the induction of the cellular DNA damage response by an Adenovirus E4 mutant

Sumithra Jayaram and Eileen Bridge Department of Microbiology 32 Pearson Hall Miami University Oxford, Ohio 45056 USA

Corresponding author Tel: 513-529-7264 Fax: 513-529-2431 Email: [email protected]

Running title: E4 mutant genome replication generates a DNA damage response

Manuscript in preparation for submission to “Virology”

18 Abstract Adenovirus mutants that lack the entire E4 region activate a cellular DNA damage response accompanied by phosphorylation of several host double-strand break repair (DSBR) and DNA damage response proteins. DSBR proteins, including the Mre11- Rad50-Nbs1 (MRN) complex concatenate E4 mutant genomes. We find that aspects of the E4 mutant-induced DNA damage response correlate with the onset of viral DNA replication and may be activated by the physical replication of viral genomes. Genetic analysis of the E4 mutants revealed that the E4-34kDa protein was required to prevent the activation of DNA damage response. Redistribution of MRN complex proteins away from viral replication centers by the E4-11kDa protein was not sufficient to prevent the DNA damage response even though it prevents genome concatenation. Activation of the Ataxia telangiectasia mutated (ATM)-mediated DNA damage response does not interfere with viral DNA replication in the presence of E4-11kDa protein. Our results indicate that replication of the incoming Ad genomes is important for triggering aspects of the DNA damage response mediated by the MRN complex and can be prevented by the E4-34kDa protein.

Keywords: Adenovirus, E4, MRN complex, ATM, DNA damage response, DSBR

19 Introduction Adenovirus (Ad) contains a 36kbp double-stranded linear DNA genome. The protein products of early region 4 (E4) are required for modulating splicing, apoptosis, transcription, DNA replication, and DNA repair pathways (reviewed in Tauber and Dobner, 2001). Host double-strand break-repair (DSBR) proteins concatenate the genomes of Ad E4 mutants (Boyer et al., 1999; Stracker et al., 2002; Weiden and Ginsberg, 1994). E4 mutant genome concatemers consist of covalently linked DNA molecules with heterogeneity in the junction fragments and no specific orientation of monomer joining (Weiden and Ginsberg, 1994). Host DSBR proteins that participate in non-homologous end joining (NHEJ), including DNA-PKcs, DNA ligase and the MRN complex, are also important for genome concatenation (Boyer et al., 1999; Stracker et al., 2002; Carson et al., 2003). E4 proteins prevent genome concatenation in wild-type Ad by interfering with the activity of DSBR proteins (Boyer et al., 1999; Stracker et al., 2002). The E4-11kDa and E4-34kDa proteins can form a physical complex with the catalytic subunit of the enzyme DNA-PK (Boyer et al., 1999). The E4-34kDa protein forms a complex with the viral early region 1b-encoded 55kDa protein (E1b-55kDa); this complex targets the members of the MRN complex for proteosome-mediated degradation. The E4-11kDa protein causes relocalization of the MRN complex away from sites of active viral replication (Stracker et al., 2002). Infection with an E4 deletion mutant triggers a DNA damage response in host cells (Carson et al., 2003). DNA damage poses a continuous threat to the genomic integrity of mammalian cells. To cope with this problem the cells have evolved an elaborate network of sensor, transducer and effector proteins that coordinate cell-cycle progression with the repair of the DNA double-strand break (reviewed in Khanna and Jackson, 2001). Autophosphorylation and activation of ATM kinase activity is one of the earliest characterized events in the response to double strand breaks (DSBs). Autophosphorylation of ATM at serine-1981 leads to dimer dissociation and it has been proposed that this leads directly to the release of active ATM monomers that modify

20 downstream effector molecules such as Nbs1, 53BP1, Chk1, H2AX, NFBD1/MDC1, and BRCA1 (Bakkenist and Kastan, 2003; Kurz and Lees-Miller, 2004). In eukaryotic cells, the MRN complex appears to be required both for ATM activation and the ability of ATM to phosphorylate and activate a number of proteins involved in DNA repair and checkpoint signaling. Early association of the MRN complex with DSBs in the DNA damage response is well established (Petrini and Stracker, 2003). This property, together with the observation that the DSB association of the MRN complex correlates with checkpoint activation, establishes that the MRN complex functions as a DSB sensor that collaborates with transducing kinases to impose checkpoint regulation in response to DSBs. ATM autophosphorylation and downstream signaling is profoundly impaired in infections with wild-type Ad due to degradation of the Mre11 complex (Carson et al., 2003). Hence, the MRN complex acts as both a sensor and effector of ATM activation and signaling in response to E4 mutant infections and DNA DSBs (Carson et al., 2003; Lee and Paull, 2005). Ad replication requires three viral proteins that participate directly in viral DNA synthesis. These are the pre-terminal protein (pTP), the DNA polymerase (pol), and the DNA-binding protein (DBP) encoded by the E2 transcription unit. Ad replicates its DNA by strand displacement and a protein priming mechanism using its own DNA polymerase (Challberg and Kelly, 1989; Van der Vliet, 1995). It is unclear as to what DNA structures are required for the activation of the MRN complex and the damage response in E4 mutant infections. Ad replication produces numerous linear DNA templates with double- stranded ends, which could be recognized as DSBs in mammalian cells. The presence of unusual replication intermediates and single-stranded DNA could also activate the MRN complex (Wang et al., 2000; Weitzman et al., 2004). The cellular response to DSBs includes the recruitment of host DSBR proteins to the sites of the lesion in a dynamically organized and timely manner. These proteins form a multi-component assembly known as a focus in response to damage induced by ionizing radiation (Lisby and Rothstein, 2005). Analogous foci containing the host DSBR proteins are formed at the sites of viral DNA replication in E4 mutant infections (Stracker et al., 2002; Carson et al., 2003). We have investigated the role of Ad DNA replication as a trigger for the activation of the DNA damage response induced by the MRN complex. We find that the

21 activation of a DNA damage response does not occur in infections with a replication- defective Ad mutant, or when E4 mutants are used to infect cells at such low multiplicities that they fail to replicate their DNA. Time course analysis indicates that phosphorylation of several host damage response proteins correlates with the onset of viral DNA replication. Genetic analysis of E4 mutants revealed that the E4-34kDa protein was required to prevent activation of the DNA damage response. Redistribution of MRN complex members by E4-11kDa was not sufficient to prevent phosphorylation of downstream targets during damage response signaling. This suggests that degradation of the MRN complex by the E4-34kDa and E1b-55kDa plays a crucial role in preventing signal transduction during DNA damage response. In summary, our results suggest that E4 mutant replication is needed to induce phosphorylation of at least some damage response proteins. Although both E4-34kDa and E4-11kDa can interfere with genome concatenation, only E4-34kDa is able to prevent the induction of the damage response.

22 Materials and Methods Cells and viruses. HeLa cells and W162 (Weinberg and Ketner, 1983) and HEK-293 monolayer cultures were maintained in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2mM L-glutamine, 10U/mL penicillin and 10µg/mL streptomycin. MO59J and MO59K cells were provided by Gary Ketner (Boyer et al., 1999). MO59J and MO59K cells were maintained in DMEM with Ham’s F12 (1:1 mix) supplemented with 10U/mL penicillin and 10µg/mL streptomycin, 0.5% sodium pyruvate and 1% non-essential amino acids and 10% fetal bovine serum (FBS). Wild-type Ad5 and E4 mutant H5dl1007, H5dl1006, H5dl1010, H5dl1017 (Bridge and Ketner, 1989; Bridge and Ketner, 1990) were propagated and titered on W162 cells. Ad E1a-(B-gal) was propagated and titered on HEK-293 cell lines. Virus titers were determined as previously described (Philipson, 1961) and expressed as fluorescence forming units (FFU)/mL. Virus titers for Ad5 and E4 mutants H5dl1007, H5dl1006, H5dl1010, H5dl1017 were determined in parallel on W162 cells. Western blotting analysis. Infected or uninfected cells (5 Χ 105) seeded in 35-mm dishes were washed twice with ice-cold phosphate-buffered saline (PBS) and then lysed in 300µL of lysis buffer (150mM NaCl, 50mM Tris pH 8.0, 5mM EDTA, 0.15% (vol/vol) Nonidet P-40, 0.1mM DTT and 5µg/mL aprotinin and leupeptin). Cell lysates were sonicated and total protein levels were measured by Bradford assay using Coomassie Plus protein reagent (Pierce) according to manufacturer’s specification. Equal amounts of total protein were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) using 8% or 12% polyacrylamide gels. Proteins were transferred to enhanced chemiluminescence (ECL) nitrocellulose (Amersham Pharmacia) overnight and the membranes were probed with primary antibodies diluted in 5% nonfat dry milk. Rabbit polyclonal antibody to NBS1 (phospho S343 from Abcam) was used at 1:1000 dilution. For detection of rabbit polyclonal antibody to phosphorylated Chk-1 (ser345) antibody (Cell signalling) was diluted 1:1000 in primary antibody dilution buffer made with 5% bovine serum albumin and incubated over night. Rabbit polyclonal antibody to phosphorylated catalytic subunit of DNA-PK enzyme (Phospo T-2609 provided by B. Chen) was used at 1:1000 dilution in primary antibody dilution buffer

23 made with 5% nonfat dry milk and incubated overnight. Rabbit polyclonal antibody to phosphorylated γH2AX (phospho serine 139 from Bethyl Laboratory) was used at 1:1000 dilution in primary antibody dilution buffer made with 5% nonfat dry milk for 2 hours. B610 (mouse monoclonal from A. Levine) was diluted 1:100 to detect the early DNA binding protein E2-72kDa. Protein blots were incubated with horseradish peroxidase- conjugated goat anti-rabbit or anti-mouse IgG (1:1500) secondary antibodies diluted in 5% non-fat dry milk. Proteins were visualized by incubating blots with ECL reagent (Amersham Pharmacia) and the chemiluminscent signals were captured using hyperfilm ECL (Amersham Pharmacia). For phosphoimager analysis of proteins, an alkaline phosphatase-conjugated goat anti-rabbit or anti-mouse IgG (1:2000) secondary antibody (Sigma) was used and enhanced chemifluorescence (ECF) substrate (Amersham) was used for detection on a STORM 860 phosphorimager (Molecular Dynamics). Scanned images were analyzed by ImageQuant® 5.2 (Molecular Dynamics) software to quantify the amount of protein. Agarose gel electrophoresis. Infected or uninfected cells (5 Χ 105) seeded in 35mm dishes were washed twice with ice-cold phosphate buffered saline (PBS). Total DNA was isolated as described previously (Bridge and Ketner, 1989). 10µg of total DNA from each sample was digested with EcoR1 and subjected to electrophoresis in a 1% agarose gel for 20 hrs at 20 Volts. DNA was transferred to Hybond-N nylon membrane (Amersham) according to manufacturer’s specifications. 32P-labeled Ad2 DNA was synthesized using the multiprime DNA labeling system (Amersham). Hybridization with 5 x 106 cpm/ml was performed at 65°C for 20 hours as described (Sambrook et al., 1989). Levels of viral DNA were quantified by phosphorimaging analysis. Scanned images were analyzed using ImageQuant® 5.2 (Molecular Dynamics) software to quantify the amount of DNA.

24 Results Replication of E4 mutant genomes activates a cellular DNA damage response. E4 mutant H5dl1007 is derived from Ad5 and lacks all the E4 open reading frames. This mutant induces a cellular DNA damage response accompanied by phosphorylation of host proteins involved in repair, cell cycle control and apoptosis (Carson et al., 2003; Jayaram and Bridge, 2005). Low multiplicity H5dl1007 infections are associated with a substantial DNA replication defect (Halbert et al., 1985; Weinberg and Ketner, 1983). We have investigated the role of E4 mutant DNA replication in generating a DNA damage response following low multiplicity infections that result in a DNA replication defect. HeLa cells were infected with wild-type Ad5 or H5dl1007 at 30, 3, 1, 0.3 FFU/cell. Even at the lower multiplicity infections with H5dl1007, greater than 70% of cells were infected as determined by immunofluoresence detection of the early DNA- binding protein 72kDa at 24 hpi. At lower multiplicities the viral 72kDa protein staining was primarily diffuse and replication foci containing 72kDa protein were rarely observed (data not shown). Total DNA was isolated from infected HeLa cells at 22-24 hpi for Southern blot analysis. Results in Fig. 4A show phosphorimager analysis of H5dl1007 DNA levels expressed as the percent of wild-type Ad5. Wild-type Ad5 DNA levels were similar at all multiplicities tested (data not shown). DNA replication of H5dl1007 at an MOI of 30 FFU/cell is close to wild-type Ad5 levels, as expected (Jayaram and Bridge, 2005). H5dl1007 DNA replication is reduced 30-fold at an MOI of 3 FFU/cell, and was not detectable above the background signal of uninfected HeLa cells at an MOI of 1 or 0.3 (Fig. 4A and B.). Results in Fig. 4b. show the EcoR1 C-fragment that is not affected by E4 deletion used for comparison of viral DNA levels by Southern blotting. In a parallel experiment we measured the levels of phosphorylated Nbs1 to assess activation of the cellular damage response in H5dl1007 infections. Total cellular extracts were prepared at 22 hpi for Western blot analysis using antibodies against phosphorylated Nbs1. The results are shown in Fig. 4C and D. H5dl1007 infection done at an MOI of 30 FFU/cell resulted in 8- to 10-fold increase in the levels of phosphorylated Nbs1 compared to uninfected HeLa cell control. We find a similar 8-fold increase in levels of phosphorylated Nbs1 protein in H5dl1007 infections done at 3

25 Figure 4. E4 mutant does not activate a DNA damage response at low multiplicities. HeLa were either uninfected (UI) or infected with wild-type Ad serotype 5 (Ad5), E4 mutant H5dl1014 (1014) at 30 FFU/cell or E4 deletion mutant H5dl1007 (1007) at 30, 3, 1 and 0.3 FFU/cell for 22-24 Hrs. The multiplicities of infection are indicated in parentheses. Southern blot was performed with 10µg of EcoR1-digested DNA prepared from each infection. The E4 mutant DNA levels are expressed as the percent of Ad5 levels (set at 100%). Viral DNA levels were quantified by phosphorimaging analysis (A). The error bars indicate the standard deviation (SD) for three independent experiments. The EcoR1 C fragment used for comparison of viral DNA levels is shown (B). Western blot analysis was performed to measure levels of phosphorylated Nbs1 (p-Nbs1) as an indication of DNA damage signaling. 50µg of total protein was subjected to SDS-PAGE analysis using a rabbit polyclonal antibody against cellular Nbs1 protein. Protein levels were quantified by phosphorimaging analysis (see Material and Methods). The level of p- Nbs1 expressed was estimated as fold increase compared to uninfected HeLa cells (C). The error bars indicate the standard deviation (SD) for three independent experiments performed. A representative Western blot of phosphorylated Nbs1 levels is also shown (D).

26 A

120 100 80 60 40

% DNA levels 20 0 Ad5 1014 1007 1007 1007 1007 (30) (30) (30) (3) (1) (0.3)

B

-36kb

C (30) (3) (1) (0.3) Ad5 UI H5dl1007

12 10 8 6 4 2 0 UI Ad5 1007 1007 1007 1007

Fold increase in levels of p-Nbs1 (30) (30) (3) (1) (0.3)

D P-Nbs1 -95kDa

H5dl1014 Ad5 1 0.3 3 30 UI H5dl1007

27 FFU/cell. In contrast, H5dl1007 infections carried out at multiplicities of 1 and 0.3 FFU/cell did not activate a DNA damage response. We also observed no phosphorylation of another damage response protein, Chk-1, in H5dl1007 infections carried out at multiplicities of 1 and 0.3 FFU/cell (data not shown). We next studied the induction of the damage response by E4 mutant H5dl1014, which has a substantial DNA replication defect even at high multiplicities (Bridge et al., 1993). H5dl1014 is derived from wild type Ad5 and carries two deletions that destroy all the E4 open reading frames except ORF4 (Bridge and Ketner, 1989). HeLa cells were infected with H5dl1014 at a high MOI of 30 FFU/cell. Total cell extracts were prepared at 22 hpi for Western blot analysis using an antibody against phosphorylated Nbs1. These results are presented in Fig. 4D. H5dl1014 infected HeLa cells did not activate the phosphorylation of Nbs1 above the background signal of uninfected HeLa cells. We have also analyzed the DNA prepared from H5dl1014 and Ad5 infected HeLa cells at 22 hpi. H5dl1014 DNA levels were not above background signal of uninfected HeLa cells (Fig. 4A). Taken together, these results suggest that replication of Ad DNA templates may be required to trigger the phosphorylation of damage response proteins in E4 mutant infected cells. Adenovirus DNA synthesis depends on viral replication factors that are encoded by the early region 2 (E2) genes. The activity of the E2 early promoter is controlled by the protein product of early region 1a (E1a) and host transcription factors. Ad E1a deletion mutants are replication defective and fail to express viral early proteins including the E2 proteins needed for viral replication (Challberg and Kelly, 1989; Van der Vliet, 1995). We next investigated the damage response induced by a replication defective Ad E1a mutant in which the gene for beta-galactosidase was used to replace the E1 region (E1a- Ad β-gal) (provided by M. Perricaudet). HeLa cells were infected with E1a- Ad β- gal at 300 FFU/cell. These titers were measured in E1a-complementing HEK 293 cells and correspond to an approximate W162 titer of 30 FFU/cell. Total cell extracts were prepared at 22 hpi for Western blot analysis using antibodies against phosphorylated Nbs1 and Chk-1. These results are presented in Fig. 5A and B. E1a- Ad β-gal did not

28 Figure 5: E1a- Ad β-gal does not induce a DNA damage response. HeLa were either uninfected (UI) or infected with wild-type Ad serotype 5 (Ad5) at 30 FFU/cell, replication defective E1a- Ad β-gal (E1a-) virus at 300 FFU/cell or E4 deletion mutant H5dl1007 (07) at 30, 3, 1 and 0.3 FFU/cell for 22-24 hrs. The multiplicities of infection are indicated in parentheses. 50µg of total protein was subjected to SDS-PAGE analysis using a rabbit polyclonal antibody against cellular Nbs1 (A) or Chk1 (B) protein as an indication of DNA damage signaling. A representative ethidium bromide-stained agarose gel used for comparison of viral DNA levels in a parallel experiment is shown (C).

29 A

P-Nbs1 -95kDa

UI 07(30) 07(0.3) E1a− (300) 07(3) Ad5(30)

B

P-Chk1 -56kDa

UI Ad5 E1a− 0.3 3 30 (30) (300) H5dl1007

C

-C fragment

UI Ad5 E1a− (.3) (1) (3) (30) (30) (300) H5dl1007

30 activate the phosphorylation of Nbs1 or Chk1 above the background signal observed in uninfected HeLa cells. We also assayed the DNA prepared from E1a- Ad β-gal at 22 hpi and found that this mutant did not replicate DNA above the background signal of uninfected HeLa cells (Fig. 5C). These results further support the idea that DNA replication may be required for activating a DNA damage response. The activation of damage response proteins in E4 mutant infections coincides with the onset of viral DNA replication. Our results indicate the activation of the damage response by E4 mutant Ad requires the replication of incoming viral genomes. Therefore, we next wanted to investigate whether the onset of H5dl1007 DNA replication correlates with the activation of host DSBR proteins. HeLa cells were infected with wild type Ad5 or E4 mutant virus H5dl1007 at 30 FFU/cell. Total DNA was isolated from infected HeLa cells at 4, 8, 16 and 22 hpi for Southern blot analysis. Results in Fig. 6A show phosphorimaging analysis of E4 mutant viral DNA levels expressed as a fold increase compared to wild-type Ad5 at 16 hpi. We find that H5dl1007 DNA levels increase by 10-fold between 12 and 16 hpi indicating the onset of viral DNA replication. In a parallel experiment, total cell extracts were prepared at 4, 8, 16 and 22 hpi for western blot analysis using antibodies against phosphorylated Nbs1, Chk1, H2AX and DNA-PKcs (Fig. 6B-G). At 4 and 8 hpi the phosphorylation of the damage response proteins Nbs1, Chk1 and DNA-PKcs was comparable to uninfected HeLa cells. Between 12 and 16 hpi levels of phosphorylated Nbs1 (Fig. 6B), Chk1 (Fig. 6C) and DNA-PKcs (Fig. 6D) proteins increase dramatically by 8-to 10-fold correlating with the onset of viral DNA replication. This result indicates that viral replication correlated with activation of DNA damage signaling. In eukaryotic cells, γH2AX is an early sensor of DNA DSB’s that is recruited early to sites of DNA double-strand breaks (Friesner et al., 2005). Interestingly, we detect a 2 to 3-fold increase in the phosphorylation of γH2AX by 5 hpi in H5dl1007 infections (Fig. 6G). This result suggests that γH2AX may also act as an early sensor of E4 mutant infections and its activation occurs before viral DNA replication.

31 Figure 6. E4 mutant induced DNA damage response correlates with the onset of viral DNA replication. HeLa cells were either uninfected (UI) or infected with wild type Ad serotype 5 (Ad5) or E4 deletion mutant H5dl1007 at 30 FFU/cell. Southern blot analysis was performed with 10µg of EcoR1-digested DNA prepared from infected cells at 0, 4, 8, 12, 16 and 22 hpi. E4 mutant viral DNA levels were quantified by phosphorimaging analysis (A). The E4 mutant DNA levels measured as fold increase compared to Ad5 at 16hpi. Western blot analysis was performed using 50µg of total protein prepared from cells infected with H5dl1007 (1007) for the indicated times. Blots were probed with rabbit polyclonal antibodies to detect phosphorylated Nbs1 (B), Chk1 (C), DNA-PKcs (T-2609) (D), or γH2AX (G). Protein levels were quantified by phosphorimaging analysis (see Material and Methods). The levels of activated damage response proteins in E4 mutant infection are expressed as fold increase compared to uninfected HeLa cells (Graph B-D). The error bars indicate the standard deviation (SD) for three independent experiments performed. A representative western blot of activated Chk1 (E) and DNA-PKcs (F) is shown. The multiplicities of infections (MOI) are indicated in parentheses. Phosphorimaging data showing γH2AX levels at 5 hpi with Ad5 or H5dl1007 (1007) shown in G.

32 A B

14 14

12 12

10 10

8 8

6 6

4 4

2 2 Fold increase in levels of p-Nbs1 Fold increase in levels of DNA 0 0

0hpi 0hpi 4hpi 8hpi 12hpi 16hpi 22hpi 4hpi 8hpi 12hpi 16hpi 22hpi

C D

12 12

10 10

8 8

6 6

4 4

2 2

0 0 Fold increase in levels of p-CHK-1 0hpi 12hpi 16hpi 22hpi 0 hpi 4hpi 8hpi 12hpi 16hpi Fold increase in levels of pDNA-PKcs

33 E

22hpi 16hpi 12hpi P-Chk1 -56kDa (MOI) (30) (3) (30) (3) (30) (3) UI H5dl1007

F

P-DNA-PKcs -460kDa UI 1007 Ad5 1007 Ad5 MOI: 30 MOI: 3 16hpi

G

4 3.5 3 2.5 2 1.5 1 0.5 0 Fold increase in levels of H2AX UI Ad5(30) 5hpi 1007(30) 5hpi

34 Redistribution of the MRN complex proteins by E4 11kDa does not prevent induction of the damage response. Infection with an Ad E4 mutant results in concatenation of the viral genomic DNA by cellular DNA repair proteins including the MRN complex. Concatemer formation can be prevented by the E4-34kDa and E1b- 55kDa complex, which target the MRN complex for proteosome-mediated degradation (Stracker et al., 2002). Degradation of the MRN complex prevents the activation of DNA damage signaling involving both the ATM and ATR kinases (Carson et al., 2003). E4- 11kDa protein also prevents genome concatenation by causing the redistribution of the MRN complex away from sites of active viral DNA replication (Stracker et al., 2002). We were interested in investigating the effect of E4-11kDa mediated redistribution of the members of the MRN complex on the induction of DNA damage response. HeLa cells were infected with wild type Ad5 or E4 mutants H5dl1007, H5dl1006, H5dl1010, H5dl1017 at 30 FFU/cell. The genotype of the E4 mutants used in the study is described (Bridge and Ketner, 1989); the status of E4-11kDa, E4-34kDa and E1b-55kDa in each mutant in summarized in Table 1. Total cell extracts were prepared at 22 hpi for Western blot analysis using antibodies against phosphorylated Nbs1. Results in Fig. 7A show levels of phosphorylated Nbs1, expressed as fold increase relative to uninfected HeLa cells in Ad5, H5dl1007, H5dl1006 and H5dl1010. Infection of Ad5 and H5dl1006 did not induce phosphorylation of Nbs1 (Fig. 7A), due to the proteosome-mediated degradation of the MRN complex by E4-34kDa and E1b-55kDa proteins (Stracker et al., 2002; Carson et al., 2003). Infection with E4 mutant H5dl1007 induced an 8- to 10-fold increase in phosphorylation of Nbs1 compared to uninfected HeLa cells. The E4 mutant H5dl1010 also showed a similar 8- to 10- fold increase in the levels of phosphorylated Nbs1. These results suggest that E4-34kDa is required to prevent the induction of DNA damage response. In a parallel experiment, we characterized the activation of Chk1 in E4 mutant infections. We find a similar increase in levels of phosphorylated Chk1 in E4 mutant H5dl1017 and H5dl1010 infections (Fig. 7B). The results of this mutational analysis indicates that E4-34kDa is required for preventing the cellular DNA damage

35 Table 1. Summary of the genotype of E4 mutants used in this study

Virus E4 11kDa E4 34kDa E1b 55kDa

Ad5 + + +

H5dl1007 - - +

H5dl1010 + - +

H5dl1006 - + +

H5dl1017 + - -

36 Figure 7. Redistribution of the MRN complex members by E4-11kDa protein does not prevent the activation of a DNA damage response. HeLa cells were either uninfected (UI) or infected with wild-type Ad5, E4 deletion mutant H5dl1007 (1007), or E4 mutant H5dl1010, H5dl1017 or H5dl1006 at 30 FFU/cell for 22-24 Hrs. 50µg of total protein was subjected to SDS-PAGE analysis using a rabbit polyclonal antibody against cellular Nbs1 (A) or Chk1 protein (B). Protein levels were quantified by phosphorimaging analysis (see Material and Methods). The levels of activated damage response proteins were measured as fold increase compared to uninfected HeLa cells. The error bars indicate the standard deviation (SD) for three independent experiments. A representative western blot is also shown. The status of E4-34kDa, E4-11kDa and E1b- 55kDa proteins is shown in parentheses.

37 A

14

12

10

8

6

4

2

Fold increase in levels of p-Nbs1 0 UI Ad5 H5dl1007 H5dl1010 H5dl1006

B

P-Chk1 -56kD

1017 1010 1007 Ad5 UI (55K-/34K-) (34K-) (34-/11K-)

38 response and the redistribution of the MRN complex away from sites of active viral replication by E4-11kDa protein present in H5dl1010 and H5dl1017 is not sufficient to prevent the induction of ATM-mediated DNA damage signaling, even though it prevents genome concatenation (Stracker et al., 2002). ATM-mediated DNA damage signaling does not affect H5dl1010 DNA replication. We have investigated the effect of activation of ATM-mediated DNA damage signaling on viral DNA replication, by comparing viral DNA levels in H5dl1007, H5dl1006 and H5dl1010 infections. HeLa cells were infected with wild-type Ad5 or the E4 mutant viruses at multiplicities of 30 and 1. Total DNA was isolated from infected HeLa cells at 22 hpi for Southern blot analysis. Results in Fig. 8 show phosphorimaging analysis of viral DNA levels expressed as the percent of wild-type Ad5. E4 mutants H5dl1007, H5dl1006 and H5dl1010 accumulated normal levels of viral DNA comparable to wild type Ad5 at a MOI of 30. H5dl1007 at an MOI of 1 FFU/cell does not replicate DNA above background of uninfected cells. E4 mutants H5dl1006 and H5dl1010 accumulate levels of viral DNA close to wild-type Ad5 even at an MOI of 1 FFU/cell. These results suggest that even at low multiplicities, the presence of the E4 ORF 3 encoded 11kDa protein in H5dl1010 infections could promote DNA replication despite the activation of DNA damage response. DNA-PKcs is not required for the activation of Nbs1 and Chk1. The role of the MRN complex has been well documented in ATM-mediated DNA damage signaling (Carson et al., 2003). We next investigated the role of DNA-dependent protein kinase catalytic subunit (DNA-PKcs) in inducing ATM-mediated DNA damage response signaling in response to E4 mutant virus infections by comparing the activation of damage response proteins Nbs1 and Chk1 in MO59J and MO59K cells. MO59J and MO59K cells are derived from a human glioma. MO59J cells lack the DNA-PKcs enzyme, and fail to concatenate the E4 mutant genomes, where as MO59K cells have the DNA-PKcs enzyme and concatenate the E4 mutant genomes (Boyer et al., 1999). MO59J cells also have lowered expression of ATM kinase (Gately et al., 1998). MO59J and MO59K cells were infected with wild-type Ad5 or the E4 deletion mutant H5dl1007 at 30 FFU/cell. Total

39 Figure 8. Activation of the DNA damage response does not affect H5dl1010 DNA replication. HeLa cells were infected with wild-type Ad serotype 5 (Ad5) or E4 deletion mutant H5dl1007 (1007), or E4 mutants H5dl1010 (1010), or H5dl1006 (1006) at 30 or 1 FFU/cell for 22-24 Hrs. The multiplicities of infection used are indicated in parentheses. Southern blot was performed with 10µg of EcoR1 digested DNA prepared from each infection. The Ad EcoR1 C fragment used for comparison of viral DNA levels by phosphorimaging analysis. The E4 mutant DNA levels are expressed as the percent of Ad5 levels (set at 100%). The error bars indicate the standard deviation (SD) for three independent experiments.

40 140

120

100

80

60

% DNA levels 40

20

0 Ad5 Ad5 1006 1006 1007 1007 1010 1010 (30) (1) (30) (1) (30) (1) (30) (1)

41 cell extracts were prepared at 22 hpi for Western blot analysis using antibody specific for phosphorylated Nbs1 (Fig. 9A) and phosphorylated Chk1 (Fig. 9B). Infection with H5dl1007 activated phosphorylation of Nbs1 and Chk1 in both MO59J and MO59K cell lines (Fig. 9A and B), suggesting that the ATM-mediated activation of damage response proteins Nbs1 and Chk1 could occur in the absence of DNA-PKcs. Ad5 did not activate the phosphorylation of Nbs1 and Chk1 in either cell line consistent with its failure to stimulate a damage response (Carson et al., 2003). These results suggest the involvement of other PI-3 kinases like ATM and ATR in phosphorylating damage response proteins in MO59J cells. We also find that the Mre11 protein still localizes to H5dl1007 replication centers in MO59J cells indicating that the DNA damage signaling apparatus is able to “sense” the replicating viral DNA even in the absence of DNA-PKcs (data not shown).

42 Fig. 9. DNA-PKcs is not required for the activation of DNA damage signaling. MO59J and MO59K cells were either uninfected (UI) or infected with wild-type Ad5 or H5dl1007 (1007) at 30 FFU/cell for 22-24 Hrs. 50µg of total protein was subjected to SDS-PAGE analysis using a rabbit polyclonal antibody against phosphorylated Nbs1 (A) or Chk1 (B).

43 A

P-Nbs1 -95kD

UI 1007 Ad5 Ad5 1007 UI MO59K MO59J

B

P-Chk1 -56kD

UI 1007 Ad5 1007 Ad5 UI MO59K MO59J

44 Discussion: Ad E4 deletion mutants display a number of severe phenotypes that include defects in viral late gene expression and DNA replication (reviewed in Tauber and Dobner, 2001). E4 deletion mutant genomes are also concatenated (Boyer et al., 1999; Stracker et al., 2002; Weiden and Ginsberg, 1994). Our previous work has shown that genome concatenation interferes with efficient expression of viral late genes (Jayaram and Bridge, 2005). E4 mutant genome concatenation requires the activity of host DSBR proteins important for non-homologous end-joining (NHEJ) (Boyer et al., 1999; Stracker et al., 2002). Ad E4-34kDa and 11kDa proteins have compensatory functions in preventing genome concatenation; E4-34kDa targets Mre11 for proteasome-mediated degradation, while E4-11kDa redistributes Mre11 away from viral replication centers and into nuclear tracks. E4 mutant infections induce a DNA damage response mediated by the MRN complex with the activation of the ATM and ATR kinases (Boyer et al., 1999; Stracker et al., 2002; Carson et al., 2003). These host DSBR proteins are normally activated by chromosomal DNA damage and by the introduction of exogenous DNA (Khanna and Jackson, 2001). Ad genomes are also substrate for repair by host DSBR proteins. The nature of Ad DNA structures that are recognized by the DSBR proteins to activate repair and checkpoint responses remains obscure. Adenovirus DNA replication produces large quantities of linear, double stranded viral DNA and single stranded replication intermediates that could be perceived by host DSBR proteins as damaged DNA (Van der vliet, 1995; Weitzman et al., 2004). The linear Ad genome also has double-stranded DNA termini that could potentially be sufficient to activate cellular damage response pathways. We have investigated the requirements for induction of the damage response in E4 mutant infected cells. We find that E4 mutant infections that do not result in viral DNA replication fail to induce a DNA damage response with the phosphorylation of Nbs1 and Chk1 proteins (Fig. 4). E1a deletion mutants fail to activate viral early gene transcription including the transcription of the E2 region needed for viral DNA replication. Our data shows that this replication defective Ad mutant fails to induce a DNA damage response (Fig. 5). Phosphorylation of several damage response proteins is coincident with the onset of viral DNA replication in E4 mutant infected cells (Fig. 6).

45 These results suggest that DNA replication is important for activating at least some aspects of the host DNA damage response. Ad replicates its DNA by strand displacement and a protein priming mechanism with the aid of E2 proteins and cellular factors (Van der Vliet, 1995). Inverted terminal repeats that contain the origin of replication are present at the ends of the Ad genome. During Ad replication there is formation of unusual panhandle structures by association of the inverted terminal repeats of the displaced single-stranded DNA (ssDNA) and DNA binding protein (DBP) (Zijderveld et al., 1993). An intriguing feature of the MRN complex and ATM kinase is that they each possess the ability to bind abnormal DNA structures, such as double-strand breaks, base-pair mismatches, Holliday junctions, cruciform DNA, template-primer junctions, and telomere repeat sequences (Uchiumi et al., 1996; Alani et al., 1997; Bennett et al., 1999; Marsischky et al., 1999). It is possible that the panhandle intermediates formed during viral replication could serve as trigger to activate damage response signaling. Alternatively, single-stranded DNA (ssDNA) formed during Ad replication could trigger the damage response. ssDNA is a common DNA structure generated by interference with DNA replication and DNA repair (Abraham and Tibbetts, 2005). Ad replication generates numerous ssDNA templates that could trigger the induction of the damage response (Van der vliet, 1995). Repair pathways usually recognize abnormal DNA structures through specific protein-DNA interactions. During the Ad replication the displaced ssDNA is released as ssDNA-DBP complex. In mammalian cells the replication protein A (RPA)-coated ssDNA acts as a key structure for the activation of the ATR kinase in response to DSBs and replication interference (Ball et al., 2005). ATR, ATR interacting protein (ATRIP) and RPA accumulate at viral replication centers in both Ad5 and E4 mutant infections. Although the wild type Ad does not activate the ATM-mediated DNA damage response signaling, since it degrades the MRN complex, E4 mutant infections lead to activation of cellular DNA damage response (Carson et al., 2003). The Ad double stranded DNA genome has a covalently attached terminal protein on the 5’ end of each strand. The Mre11 protein is a nuclease with broad range substrate recognition that is required for the nucleolytic processing of DNA double-strand breaks. The Mre11 nuclease is essential for releasing DNA ends for repair, when 5' ends are

46 blocked, either by the attachment of a protein or an unusual structure (Moreau et al., 1999). It is possible that the terminal protein has a protective role in preventing recognition of the double stranded DNA ends of the incoming viral genomes, but that following replication Mre11 is activated to cleave the terminal protein away from viral genome ends. The failure of incoming viral genomes to induce Mre11-mediated DNA damage response signaling supports this hypothesis (Fig. 4). The role of E4-34kDa in preventing the genome concatenation and the induction of DNA damage response signaling has been characterized (Stracker et al., 2002; Carson et al., 2003). The E4-11kDa protein can interfere with genome concatenation by redistributing the MRN complex away from sites of active viral DNA replication, however its role in DNA damage response signaling has not yet been investigated. Genetic analysis of E4 mutants has revealed that E4-34kDa is required to prevent the induction of a DNA damage response (Fig. 7C). We find that E4-11kDa mediated redistribution of the MRN complex members away from viral replication centers was not sufficient to prevent the activation of the damage response proteins (Fig. 7), even though it prevents genome concatenation (Stracker et al., 2002). It is likely that E4-11kDa prevents genome concatenation by redistributing the MRN complex members or targeting the complex to cytoplasmic aggresomes (Evans and Hearing, 2003; Araujo et al., 2005). This would prevent the activated MRN complex proteins from accessing viral DNA. Additionally, E4-11kDa binding to other host DSBR proteins like DNA-PKcs could also prevent genome concatenation despite the activation of other aspects of the DNA damage response. E4-11kDa is a multifunctional protein that possesses genetically distinct roles in promoting viral DNA replication and preventing genome concatenation (Evans and Hearing, 2003). We have investigated the affect of ATM-mediated DNA damage response signaling on viral DNA replication in E4 mutant infections. An E4 deletion mutant that lacks both the E4-34kDa and 11kDa displays a severe defect in viral DNA replication at low multiplicities (Fig. 8). In contrast, the presence of E4-11kDa protein promotes viral DNA replication despite the presence of ATM-mediated DNA damage response signaling. If E4-11kDa binds and redistributes “activated” repair proteins away

47 from viral genomes, then the activation of damage response may not interfere with genome replication. DNA-PKcs is a serine/threonine protein kinase that is activated by association with DNA. Biochemical analysis has revealed that the activation of DNA-PKcs requires interactions with both double and single-stranded DNA (Jackson, 1997). The role of DNA-PKcs has been well documented in NHEJ (Wetering and van Gent, 2004). Initial analyses of cells defective in DNA-PK components have, however, shown them to have intact DNA damage checkpoints and to be capable of activating p53 in response to ionizing radiation (IR) (Jackson, 1997). We find that E4 mutant infections of both DNA- PKcs proficient MO59K and DNA-PKcs deficient MO59J cells resulted in activation of DNA damage response signaling (Fig. 9). This suggests that DNA-PKcs is not required for DNA damage signaling, but instead plays a more important role in actual repair of the ends. Following virus infections and induction of DSBs, the MRN complex appears to function as a damage sensor upstream of ATM kinase. The mechanism by which the MRN complex activates ATM-mediated signaling pathways in response to virus infections and double strand breaks is not known. We find infections with replication defective Ad mutants do not induce a DNA damage response in spite of the presence of the Mre11 protein in these cells. This raises the interesting point that viral replication may be required for the MRN complex to sense viral genomes as “damage” and to initiate subsequent repair. The DNA repair apparatus forms a part of the innate intracellular defense against the foreign invaders. The activity of DSBR proteins serves as a major obstacle to a productive Ad infection, hence the virus has evolved multiple ways to inactivate it. Understanding the relationship between the DNA damage response and virus infections is likely to provide important insights into how our cells recognize and repair their own damaged DNA.

48 CHAPTER 2: Jayaram, S., and Bridge, E. 2005. Genome concatenation contributes to the late gene expression defect of an adenovirus E4 mutant. Virology. 342(2): 286-96.

49 Abstract Adenovirus mutants that lack the entire E4 region are severely defective for late gene expression. E4 mutant genomes are also concatenated by host double strand break repair (DSBR) proteins. We find that E4 mutant late gene expression improves in MO59J cells that fail to form genome concatemers. DSBR kinase inhibitors interfere with genome concatenation and also stimulate late gene expression. Concatenation of E4 mutant genomes interferes with cytoplasmic accumulation of viral late messages and leads to reduced late protein levels and poor viral yields following high multiplicity infection. However, failure to concatenate viral genomes did not rescue either the DNA replication defect or virus yield following low multiplicity E4 mutant infection. Our results indicate that if the E4 mutant DNA replication defect is overcome by high multiplicity infection, concatenation of the replicated genomes by host DSBR interferes with viral late gene expression.

Keywords: Adenovirus, E4, concatemer, DNA-PKcs, damage response, double strand break repair

50 Introduction Adenovirus (Ad) early region 4 (E4) encodes a diverse set of proteins that are crucial to a productive lytic infection. The protein products of E4 open reading frames 3 (E4-11kDa) and 6 (E4-34kDa) possess complementary roles in the efficient onset of viral DNA replication, late mRNA splicing, and in the accumulation of late mRNAs and proteins (reviewed in Tauber and Dobner, 2001). Single mutations affecting either the E4-11kDa or E4-34kDa genes have only a modest effect on virus late gene expression, while mutations affecting both genes result in more severe late gene expression defects. These results suggest that the E4-11kDa and E4-34kDa proteins are able to partially compensate for each other's function (Bridge and Ketner, 1989; Huang and Hearing, 1989). Ad mutants that lack the entire E4 region are severely defective for late protein synthesis; this defect is attributed to decreased levels of nuclear and cytoplasmic late RNA (Bridge and Ketner, 1990; Halbert et al., 1985; Huang and Hearing, 1989). Although late message levels are reduced, the transcription rate of viral late genes is close to normal, suggesting that E4 products might directly or indirectly stabilize newly synthesized viral late RNA (Sandler and Ketner, 1989). Cells infected with E4 mutant viruses produce concatenated viral DNA genomes (Boyer et al., 1999; Stracker et al., 2002; Weiden and Ginsberg, 1994). E4 protein products prevent genome concatenation by interfering with the activity of host double strand break repair (DSBR) proteins. The E4-11kDa and E4-34kDa proteins each form a physical complex with the DSBR enzyme DNA-dependent protein kinase catalytic subunit (DNA-PKcs) (Boyer et al., 1999). The E4-34kDa protein forms a complex with a viral 55kDa protein encoded by early region 1b (E1b-55kDa); this complex targets DSBR proteins of the Mre11-Rad50-Nbs1 (MRN) complex for proteasome-mediated degradation (Stracker et al., 2002). The E4-11kDa protein redistributes MRN complex proteins, preventing them from localizing to sites of viral replication (Evans and Hearing, 2005; Stracker et al., 2002). It is possible that E4 proteins promote efficient viral late gene expression by inhibiting the host cell's ability to concatenate viral DNA genomes.

51 Compensatory roles for the E4-11kDa and E4-34kDa proteins in both processes (Bridge and Ketner, 1989; Stracker et al., 2002) are consistent with this hypothesis. Ad genomes are double-stranded linear DNA. Infection by E4 mutants triggers a DNA damage response in host cells similar to that induced by DNA breaks caused by ionizing radiation or genotoxic stress (Boyer et al., 1999; Carson et al., 2003; Stracker et al., 2002). The DNA damage response leads to activation of DSBR proteins that are normally involved in repairing double strand breaks (reviewed in Khanna and Jackson, 2001). DSBR proteins repair double strand breaks by at least two distinct mechanisms: homologous recombination and non-homologous end joining (NHEJ). Members of the Rad52 epistasis group carry out homologous recombination. The MRN complex is involved in both homologous recombination and NHEJ. NHEJ is the predominant mode of DSBR in higher eukaryotes (Allen et al., 2003). A key component of the NHEJ apparatus is the DNA-PKcs enzyme. DNA-PKcs belongs to the phosphatidylinositol-3 (PI-3) family of protein kinases that includes the cell cycle regulators ataxia- telangiectasia mutated (ATM) and ATM-Rad3-related (ATR) (Smith and Jackson, 1999). Ad E4 mutant infections induce a cellular DNA damage response with the activation of ATM and ATR (Carson et al., 2003). These activities affect cell cycle regulation and apoptosis as well as DNA repair. Thus, preventing the DNA damage response and the concatenation of viral DNA may play important roles in a productive Ad infection. The concatemers that arise in E4 mutant infections are covalently linked DNA molecules that are imprecisely ligated in random head-to-head, head-to-tail and tail-to- tail orientation (Weiden and Ginsberg, 1994). Extensive genome concatenation could destroy viral origins of replication located at the ends of the linear genome, or produce DNA molecules too large to be packaged into the viral capsids. Here we have tested the hypothesis that genome concatenation may also interfere with efficient viral late gene expression. Our results indicate that an E4 mutant virus expresses late proteins better in a cell line that lacks the DSBR enzyme, DNA-PKcs, and hence does not form genome concatemers. We show that interfering with the activity of DSBR kinases inhibits E4 mutant genome concatenation and substantially rescues its late gene expression defect. Preventing genome concatenation with DSBR kinase inhibitors stimulates the accumulation of viral late mRNA. These results suggest that E4 proteins promote

52 efficient Ad late gene expression, at least in part, by interfering with genome concatenation.

53 Materials and Methods Cells and viruses. HeLa and E4 mutant-complementing W162 (Weinberg and Ketner, 1983) monolayer cell cultures were maintained in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2mM L-glutamine, 10U/mL penicillin and 10µg/mL streptomycin. MO59J and MO59K cells were provided by Gary Ketner (Boyer et al., 1999), and were maintained in DMEM with Ham’s F12 (1:1 mix) supplemented with 10U/mL penicillin, 10µg/mL streptomycin, 0.5mM sodium pyruvate, 0.1mM non-essential amino acids (GIBCO) and 10% FBS. Wild-type Ad5 and E4 mutant H5dl1007 (Bridge and Ketner, 1989) were propagated on W162 cells. Ad5 and H5dl1007 virus titers were determined in W162 cells, and expressed as fluorescence forming units (FFU)/mL (Philipson, 1961). Cells were infected with 30 FFU/cell unless otherwise indicated. DSBR kinase inhibitors. Wortmannin (Sigma-Aldrich) was dissolved in DMSO and added 2 hpi to a final concentration of 15µM, for the times indicated. Caffeine (Sigma- Aldrich) was dissolved in water and added to the culture 2 hpi to a final concentration of 10mM, for the times indicated. Western blotting analysis. Infected or uninfected cells (~5 Χ 105) seeded in 35mm dishes were washed twice with ice cold phosphate buffered saline (PBS) and then lysed in 300µL of lysis buffer (150mM NaCl, 50mM Tris pH 8.0, 5mM EDTA, 0.15% (vol/vol) Nonidet P-40, 0.1mM DTT and 5µg/mL aprotinin and leupeptin). Cell lysates were sonicated and total protein levels were measured by Bradford assay using Coomassie plus protein reagent (Pierce) according to the manufacturer's specifications. Equal amounts of total protein were subjected to sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 8% or 12% polyacrylamide gels. Proteins were transferred to enhanced chemiluminescence (ECL) nitrocellulose (Amersham Pharmacia) overnight and the membranes were probed with primary antibodies diluted in 5% nonfat dry milk. Rabbit polyclonal antiserum to the late proteins penton and fiber (from U. Pettersson) was used at a 1:1000 dilution. Monoclonal antibody B610 (A. Levine) was diluted 1:100 to detect the early DNA binding protein E2-72kDa. Rabbit polyclonal antibody to phosphorylated Nbs1 (ab15088 from Abcam) was used at a 1:1000 dilution. Rabbit polyclonal antibody to phosphorylated Chk-1 (a phospho-Chk1(S345) from Cell

54 Signaling) was diluted 1:1000 into 5% bovine serum albumin and incubated over night. Protein blots were incubated with a 1:1500 dilution of horseradish-peroxidase conjugated goat anti-rabbit or anti-mouse IgG secondary antibodies in 5% non-fat dry milk. Proteins were visualized by incubating blots with ECL reagent and the chemiluminescent signals were captured using hyperfilm ECL (Amersham Pharmacia). For phosphorimaging analysis of proteins, a 1:2000 dilution of alkaline phosphatase conjugated goat anti-rabbit or anti-mouse IgG secondary antibody (Sigma-Aldrich) was used with enhanced chemifluorescence (ECF) substrate (Amersham) for detection on a STORM 860 phosphorimager (Molecular Dynamics). Scanned images were analyzed using ImageQuant® 5.2 (Molecular Dynamics) software to quantify the amount of protein. Viral DNA analysis. Total DNA was isolated from ~ 5 x 105 infected or uninfected cells as described (Bridge and Ketner, 1989). 10µg of total DNA from each sample was digested with EcoRI and subjected to electrophoresis through a 1% agarose gel. DNA was transferred to Hybond-N nylon membrane (Amersham) according to the manufacturer's specifications and used for Southern blotting. A 32P-labeled probe was synthesized from Ad genomic DNA using the multiprime DNA labeling system (Amersham). Hybridization with 5 x 106 cpm/ml was performed at 65°C for 20 hours as described (Sambrook et al., 1989). Levels of viral DNA were quantified by phosphorimaging analysis. Scanned images were analyzed using ImageQuant® 5.2 (Molecular Dynamics) software to quantify the amount of DNA. Pulsed-field gel electrophoresis. Intracellular DNA was prepared and analyzed for the presence of genome concatemers as previously described (Boyer et al., 1999). Agarose plugs containing DNA from ~1.25 x 105 cells were loaded onto 1.2% agarose gels for electrophoresis using a CHEF-DR® II pulsed field electrophoresis system in hexagonal field mode for 10 hours at 6V/cm with a switch time of 15 seconds. Gels were stained with ethidium bromide for visualization of viral DNA concatemers. In some experiments gels were further analyzed by Southern blotting as described above. Viral RNA analysis. Total cytoplasmic RNA was isolated from ~5 x 105 cells and used for northern blot analysis (Sambrook et al., 1989). 10µg of total RNA from each sample was fractionated on a 2.2% formaldehyde-1% agarose gel. The RNA was then transferred to Hybond-N nylon membrane (Amersham) according to the manufacturer's

55 specifications and probed with 32P-labeled pVI DNA. The pVI probe was prepared by PCR amplification of the pVI gene from Ad2 DNA using pVI forward primer corresponding to 5′CAA CTT TGC GTC TCT GGC3′, and a reverse primer corresponding to 5′ CTG TTC AGT GTG CTT TGC 3′ (Gen Bank accession number BK000408; Davidson et al., 2003). The pVI PCR product was pooled, purified and used for random priming with the multiprime DNA labeling system (Amersham). Hybridization was performed at 65°C for 20 hours as described (Sambrook et al., 1989). The amount of cytoplasmic RNA was measured by phosphorimaging analysis. Scanned images were analyzed using ImageQuant® 5.2 (Molecular Dynamics) software to quantify the amount of RNA.

56 Results Late protein synthesis of an E4 mutant improves in cells that fail to concatenate viral genomes. We have investigated the role of genome concatenation in the late protein defect of E4 mutant H5dl1007 by comparing viral late gene expression in cell lines that do and do not form genome concatemers. MO59K and MO59J cells are derived from a human glioma (Allalunis-Turner, et al., 1993). MO59J cells lack DNA-PKcs and fail to concatenate E4 mutant genomes, while MO59K cells make normal levels of DNA-PKcs and efficiently concatenate E4 mutant viral genomes (Boyer et al., 1999). We assayed viral late protein production following infection of MO59J and MO59K cells to assess the role of DNA-PKcs and genome concatenation in E4 mutant late gene expression. HeLa, MO59K and MO59J cells were infected with wild-type Ad serotype 5 (Ad5) or the E4 deletion mutant H5dl1007 at 30 FFU/cell. H5dl1007 is derived from Ad5 and lacks all E4 open reading frames (Bridge and Ketner, 1989). Total cell extracts were prepared 20- 22 hours post infection (hpi) for western blot analysis using antiserum against the viral late proteins penton and fiber (Aspegren et al., 1998). Late protein levels were quantified by phosphorimaging and the results are shown in Fig. 10A. H5dl1007 late protein levels were reduced 10-fold compared to Ad5 in HeLa and MO59K cells; this result is consistent with previous reports (Bridge and Ketner, 1989), and demonstrates that H5dl1007 has a similar late gene expression defect in HeLa and MO59K cells. In contrast, the late protein defect of H5dl1007 was only 2-fold in MO59J cells (Fig. 10A). The levels of DNA produced by H5dl1007 were similar to Ad5 in all three cell lines, indicating that a defect in DNA replication did not contribute to reduced late protein levels at this multiplicity of infection (data not shown). We analyzed DNA prepared from Ad5 and H5dl1007 infected HeLa, MO59K and MO59J cells at 22 hpi by pulsed field gel electrophoresis (PFGE) to detect the presence of genome concatemers. H5dl1007 infections of HeLa and MO59K cell lines resulted in the production of concatenated genomes that were not detected in MO59J cells (Fig. 10B), as reported previously (Boyer et al., 1999). Taken together, these results indicate that late gene expression of H5dl1007 improves following infection of a cell line that lacks DNA-PKcs and is unable to concatenate viral genomes.

57 Fig. 10. E4 mutant late protein synthesis improves in MO59J cells that lack DNA- PKcs. HeLa, MO59K and MO59J cells were infected with wild-type Ad5 or E4 deletion mutant H5dl1007 (1007) for 20-24 h. 30µg of total protein was analyzed by western blotting using a rabbit polyclonal antiserum against the viral late proteins penton and fiber. Late protein levels were quantified by phosphorimager analysis (A). E4 mutant late protein levels are expressed as the percent of Ad5 levels (set at 100%). Error bars indicate the standard deviation from three independent experiments performed in each cell line. Intracellular DNA was prepared from infected cells and analyzed by PFGE (see Materials and Methods). Ethidium bromide staining was used to visualize unit length (36kbp) and concatenated Ad genomes (B).

58 59 Inhibition of DSBR kinases interferes with E4 mutant genome concatenation and stimulates late gene expression. DNA-PKcs is required for concatenating E4 mutant virus genomes (Boyer et al., 1999). Interfering with DNA-PKcs kinase activity is therefore expected to inhibit genome concatenation. Wortmannin is a fungal metabolite that covalently binds to a lysine residue in the kinase domain and irreversibly inhibits DNA-PKcs, thereby preventing its kinase activity (Izzard et al., 1999). We studied E4 mutant genome concatenation in infected cultures treated with wortmannin. HeLa cells were infected with Ad5 or H5dl1007, and cultured in the presence or absence of 15µM wortmannin for 32 h to measure the effect of this inhibitor on genome concatenation. The amount of viral DNA concatenation was measured by PFGE and Southern blotting. As expected, H5dl1007 infection of HeLa cells produced concatemers of at least six genome lengths. Treatment with wortmannin significantly reduced concatenation of H5dl1007 genomes (Fig. 11A). Ad5 produced exclusively monomeric viral DNA regardless of wortmannin treatment. H5dl1007 genomes were not concatenated in MO59J cells, as expected (Fig. 11B). Our results suggest that wortmannin interferes with concatenation of E4 mutant genomes in cell lines with active DNA-PKcs. Wortmannin did not have a significant effect on levels of viral DNA in any cell line tested (Fig. 11 and data not shown), indicating that it did not interfere with Ad DNA replication. Wortmannin affects DNA-PKcs and at high concentrations can also interfere with other PI-3 kinases involved in DNA damage response signaling, including ATM and ATR (Sarkaria et al., 1998). ATM and ATR play an important role in DNA damage- induced arrest at G1/S, S, and G2/M checkpoints (Kurz and Lees-Miller, 2004; Shechter et al., 2004). ATM participates in signal transduction following DNA damage but has not been implicated in genome concatenation of E4 mutants (Stracker et al., 2002). Caffeine inhibits both ATM- and ATR-mediated downstream signaling, abolishing the ionizing radiation-induced checkpoint response (Sarkaria et al., 1999). Caffeine inhibits DNA- PKcs induced phosphorylation of the 32kDa subunit of replication protein A (RPA) in response to DNA damage, but does not affect DNA-PKcs-mediated NHEJ following

60 Fig. 11. Inhibition of DSBR kinases interferes with E4 mutant genome concatenation. HeLa cells (A) or MO59J cells (B) were infected with Ad5 or H5dl1007 and cultured with or without 15µM wortmannin (wort.) for 32 h. Intracellular DNA was prepared from infected cells for analysis of genome concatemers by PFGE and Southern blotting. HeLa cells were infected with Ad5 and H5dl1007 and cultured for 20-22 h with or without 10mM caffeine (C). Intracellular DNA was analyzed by PFGE. Ethidium bromide staining was used to visualize unit length (36kbp) and concatenated Ad genomes.

61 62 ionizing radiation (Block et al., 2004). We next studied the effect of caffeine on E4 mutant genome concatenation. HeLa cells were infected with Ad5 or H5dl1007 and cultured in the presence or absence of 10mM caffeine for 20-22 h. Intracellular DNA was then analyzed by PFGE. We found that caffeine treatment significantly reduced genome concatenation in H5dl1007 infections (Fig. 11C). Ad5 produced exclusively monomeric viral DNA regardless of caffeine treatment. This result suggests a possible role for ATR, ATM or another caffeine-sensitive kinase in E4 mutant genome concatenation. Inhibition of DSBR kinase activity substantially rescues the late protein defect of an E4 mutant virus. Since inhibiting DSBR kinases with wortmannin or caffeine affects E4 mutant genome concatenation, we next investigated the effect of these inhibitors on late gene expression by H5dl1007. Ad5 and H5dl1007 infected HeLa cells were cultured with or without 15µM wortmannin, or 10mM caffeine. Total cell extracts were prepared at 20- 22 hpi for western blot analysis to compare viral late protein levels. The results are presented in Fig. 12. H5dl1007 infected HeLa and MO59K cells showed a 10-fold reduction in late protein levels compared to Ad5. Treatment with 15µM wortmannin (Fig. 12A) or 10mM caffeine (Fig. 12B) increased E4 mutant late protein levels by 5- to 7- fold. Late protein levels of Ad5 were not affected by wortmannin or caffeine treatment indicating that these kinase inhibitors do not have a general stimulatory effect on Ad late gene expression. We performed a dose response experiment to further characterize the relationship between genome concatenation and late protein levels in the presence of increasing amounts of wortmannin. Intracellular DNA from HeLa cells infected with H5dl1007 was assayed for the presence of genome concatemers by PFGE. Total cell extracts from parallel experiments were assayed for the presence of viral late proteins by western blotting. PFGE of H5dl1007 infected HeLa cells revealed a decrease in genome concatenation in the presence of increasing concentration of wortmannin that correlated with a simultaneous increase in late protein levels (Fig. 12C). A similar response was observed to an increasing dose of caffeine (data not shown). Wortmannin and caffeine also stimulated H5dl1007 late protein levels following infection of MO59K cells (data not shown). These results show that DSBR kinase inhibitors that interfere with E4 mutant

63 Fig. 12. Inhibition of DSBR kinase activity partially rescues the late protein defect of an E4 mutant virus. HeLa cells were either uninfected (UI), infected with Ad5, or H5dl1007 and cultured with or without 15µM wortmannin (wort.) (A), or 10mM caffeine (B). DMSO (D) was added to some cultures as a control for the solvent used to dissolve wortmannin (see western blot shown in part A). Western blot analysis was performed to detect levels of the late proteins penton and fiber (top panels). Late protein levels were quantified as described in Fig. 10 and the results are graphed in the lower panels. W and C indicate experiments performed with the inhibitors wortmannin and caffeine, respectively. The error bars indicate the standard deviation from three independent experiments. We performed a dose-response experiment by treating H5dl1007 infected HeLa cells with increasing concentrations of wortmannin for 20-22 h. Genome concatemers and levels of late proteins penton and fiber were assessed by PFGE (top panel) or western blotting (lower panel), respectively (C). MO59J cells were infected with Ad5 or H5dl1007 and cultured with or without 15µM wortmannin (W) or 10mM caffeine (C) for 20-22 h (D). Levels of the late proteins penton and fiber were assayed by western blotting and quantified as described in Fig. 10. The error bars indicate the standard deviation from three independent experiments.

64 65 genome concatenation substantially rescue its late gene expression defect. We next assayed the ability of wortmannin and caffeine to induce late gene xpression in E4 mutant infections of MO59J cells. MO59J cells lack functional DNA- PKcs, have substantially reduced ATM kinase activity (Chan et al., 1998), and fail to concatenate E4 mutant genomes (Boyer et al., 1999; Fig.10). If wortmannin and caffeine promote E4 mutant late gene expression by inhibiting genome concatenation they should not affect E4 mutant late gene expression in MO59J cells. MO59J cells were infected with Ad5 and H5dl1007 and cultured in the presence or absence of 15µM wortmannin or 10mM caffeine. Total cell extracts from Ad5 and H5dl1007 infected MO59J cells were prepared for western blot analysis at 20-22 hpi to compare viral late protein levels. Both DSBR kinase inhibitors failed to stimulate H5dl1007 late protein levels in MO59J cells (Fig. 12D) indicating that the increase in E4 mutant late gene expression seen in HeLa cells (Fig. 12A-C) requires that cells are able to concatenate viral genomes, or the presence of functional DNA-PKcs and ATM. E4 mutants induce a DNA damage response that leads to the phosphorylation of several cellular proteins (Carson et al., 2003). Thus, it is possible that treatment with wortmannin or caffeine would interfere with cellular damage response signaling in addition to inhibiting E4 mutant genome concatenation. We studied the effect of wortmannin and caffeine treatment on phosphorylation of two cellular DSBR proteins involved in the damage response, Nbs1 and Chk1, to assess this possibility. Ad5 and H5dl1007 infected HeLa cells were cultured in the presence or absence of 15µM wortmannin, or 10mM caffeine. Total cell extracts were prepared at 20-22 hpi for western blot analysis using an antibody specific for the phosphorylated form of Nbs1. Infection with H5dl1007 activated phosphorylation of Nbs1 even following wortmannin and caffeine treatment (Fig. 13A), suggesting that these inhibitors did not prevent this aspect of the damage response despite their ability to inhibit Ad genome concatenation. Levels of total Nbs1 were not affected by E4 mutant infection or treatment with DSBR kinase inhibitors (data not shown). Ad5 did not activate phosphorylation of Nbs1 consistent with its failure to stimulate a damage response (Carson et al., 2003). When we

66 Fig. 13. The affect of DSBR kinase inhibitors on the DNA damage response induced by E4 mutant infections. HeLa cells were either uninfected (UI) or infected with Ad5 or H5dl1007 and cultured with or without 15µM wortmannin (W) or 10mM caffeine (C) for 20-22 h. Cell extracts were prepared and 50µg of total protein was used for western blot analysis of A) phosphorylated Nbs1 and B) phosphorylated Chk1.

67 68 examined H5dl1007-induced phosphorylation of Chk1, we found that wortmannin treatment did not interfere with Chk1 phosphorylation in E4 mutant infected cells, while treatment with caffeine reduced the phosphorylation of Chk1 2-3 fold, as shown previously (Carson et al., 2003). Thus, although both kinase inhibitors interfered with E4 mutant genome concatenation and stimulated E4 mutant late gene expression, only caffeine interfered with phosphorylation of Chk1 during the damage response. These data suggest that the increase in E4 mutant late gene expression observed following the inhibition of DSBR kinases is more tightly correlated with their inhibition of genome concatenation, rather than the ability to interfere with damage response signaling. E4 mutant genome concatenation interferes with viral late mRNA accumulation. Previous investigations on the late gene expression defect of E4 mutants suggested that the defect was the result of a posttranscriptional effect on nuclear RNA stability. A reduction in nuclear and cytoplasmic steady state mRNA levels was responsible for the late protein synthesis defect of E4 mutants (Bridge and Ketner, 1989; Sandler and Ketner, 1989). We have characterized the molecular step at which late gene expression is stimulated by DSBR kinase inhibitors. We first determined whether the stimulation of E4 mutant late gene expression was due to an effect on early gene expression and/or DNA replication. Ad5 and H5dl1007 infected HeLa cells were cultured in the presence or absence of 15µM wortmannin or 10mM caffeine to interfere with genome concatenation and stimulate late gene expression. Cell extracts were processed at 20-22 hpi for western blot analysis with antibodies raised against the early 72kDa viral DNA binding protein, or for analysis of viral DNA. Levels of 72kDa were similar in H5dl1007 or Ad5 infected cells and did not change following treatment with wortmannin (Fig. 14A); identical results were obtained with caffeine (data not shown). This indicates that the effect of DSBR kinase inhibitors on late gene expression was at a step following early gene expression. Southern blotting and phosphorimaging analysis of EcoRI-digested total DNA, revealed that H5dl1007 accumulated normal levels of viral DNA compared to Ad5 and the levels did not change following treatment with DSBR kinase inhibitors (Fig. 14B). This shows that differences in viral DNA levels were not responsible for the increase in late protein levels induced by DSBR kinase inhibitors.

69 Fig. 14. E4 mutant genome concatenation interferes with viral late mRNA accumulation. HeLa cells either uninfected (UI) or infected with Ad5 or H5dl1007 were cultured with or without 15µM wortmannin (W) or 10mM caffeine (C) for 20-22 h to compare viral early protein levels (A), viral DNA levels (B) and viral late RNA levels (C). Levels of the early 72kDa DNA binding protein were measured by western blotting (A). Southern blot analysis was performed with 10µg of EcoRI digested DNA prepared from each infection (B). The Ad EcoRI C fragment used for comparison of viral DNA levels is shown (top panel). Viral DNA levels were quantified by phosphorimaging analysis (bottom panel). The E4 mutant DNA levels are expressed as the percent of Ad5 (set at 100%). The error bars indicate the average standard deviation from three independent experiments. Northern analysis was carried out to measure the cytoplasmic levels of the viral late message for pVI (C). A representative northern blot is shown in the top panel. Viral late RNA levels were quantified by phosphorimaging analysis. E4 mutant RNA levels are expressed as the percent of Ad5 (set at 100%). The error bars indicate the average standard deviation from three independent experiments.

70

71 We next determined if treatments that interfere with genome concatenation rescue E4 mutant late message levels. Cytoplasmic mRNA prepared from H5dl1007 or Ad5 infected HeLa cells 20-22 hpi was analyzed by northern blotting using random primed DNA from the pVI gene as a probe. The results are shown in Fig. 14C. H5dl1007 pVI mRNA levels were reduced about 13-fold compared with Ad5, and inhibition of DSBR kinases with 15µM wortmannin or 10mM caffeine increased pVI late message levels about 7-fold, such that they were within 2-fold of Ad5. We saw a similar increase in levels of hexon mRNA following treatment with DSBR kinase inhibitors (data not shown). These data indicate that inhibitors of genome concatenation can substantially rescue the late mRNA accumulation defect of an E4 mutant. E4 mutant DNA replication and virus yield are severely reduced at lower multiplicities of infection and are not rescued by preventing genome concatenation. The experiments described above were performed at an MOI of 30FFU/cell; under these infection conditions E4 mutants replicate their genomes at close to wild type Ad5 levels. However, E4 mutants have a substantial DNA replication defect at lower MOIs (Halbert et al., 1985; Weinberg and Ketner, 1986). We next assessed the role of genome concatenation for E4 mutant DNA replication following low multiplicity infection. MO59K and MO59J cells were infected with Ad5 and H5dl1007 at 3 or 30 FFU/cell and total DNA was prepared at 24 hpi for Southern blot analysis. The results in Fig. 15 show that although H5dl1007 DNA levels were comparable to Ad5 at high MOI, they were substantially reduced at low MOI. Furthermore, the DNA defect observed following low multiplicity infection with H5dl1007 was severe in both MO59K and MO59J cells, despite the inability of the latter to concatenate viral genomes. This result suggests that the inability of MO59J cells to concatenate E4 mutant genomes does not rescue the E4 mutant DNA replication defect observed in low multiplicity infections. We also measured virus yield following low and high multiplicity infections of MO59K and MO59J cells to assess the role of genome concatenation on virus production in the presence and absence of a DNA replication defect. Cells were infected with Ad5 or H5dl1007 at 3 or 30 FFU/cell. A freeze-thaw lysate of the infected cells was prepared at 48 hpi, and the amount of infectious virus in each sample was measured by FFU assay (Phillipson, 1961) in W162 cells that complement E4 mutants (Weinberg and Ketner,

72 1983). The results are presented in Table 2. Following high multiplicity infection of MO59K cells, H5dl1007 virus yield was reduced about 300-fold. In contrast, H5dl1007 yield was only reduced about 2-fold in MO59J cells infected at high MOI. After low multiplicity infection Ad5 achieved titers of ~8x107 FFU/ml in both MO59K and MO59J cells. In sharp contrast, we failed to detect infectious E4 mutant virus in lysates prepared from either cell line infected at low multiplicity. About 80% of MO59J and MO59K cells express the early 72kDa replication protein by 24 hpi following low MOI with H5dl1007 (data not shown), suggesting that reasonable levels of infection were achieved. In most of these cells the 72kDa protein is diffusely distributed in the nucleus. We find that less than 5% of the cells have well-developed replication foci characteristic of viral replication (data not shown). The resulting severe defect in DNA replication (Fig. 15) resulted in a profound defect in new virus production in both cell lines at 48 hpi (Table 2). Previous work (Bridge and Ketner, 1989) has shown that H5dl1007 eventually produces greatly reduced levels of infectious virus in plaquing assays, indicating that the block to replication and new virus production at low MOI is not absolute. Taken together, our results indicate that failure to concatenate viral genomes can substantially increase E4 mutant virus yields when the viral replication defect is alleviated by high multiplicity infection. However, failure to concatenate viral genomes is insufficient to rescue the E4 mutant DNA replication defect or virus yield following low MOI.

73 Fig. 15. The DNA replication defect of E4 mutants at low MOI is not rescued by the failure to concatenate viral genomes. MO59J and MO59K cells were infected with Ad5 and H5dl1007 at multiplicities of 3 or 30 FFU/cell. At 22 hpi total intracellular DNA was prepared, digested with EcoRI, and analyzed by Southern blotting. The EcoRI C fragment is shown.

74 75 Table 2

76 Discussion E4 mutants are defective for DNA replication and late gene expression (Halbert et al., 1985; Weinberg and Ketner, 1985). Genetic studies have shown that two E4 proteins, E4-34kDa and E4-11kDa, can independently promote efficient viral DNA replication and late gene expression (Bridge and Ketner, 1989; Huang and Hearing 1989). Activities of the E4-34kDa and E4-11kDa proteins include the ability to bind, degrade or redistribute several host DSBR proteins (Boyer et al., 1999; Stracker et al., 2002) thereby preventing the concatenation of Ad genomes. The significance of these E4-mediated activities on host DSBR for viral growth is under investigation. The E4-34kDa protein promotes protein degradation by forming a complex with protein constituents of an E3 ubiquitin ligase, which then targets selected proteins for ubiquitination and degradation by the proteasome (Querido et al., 2001). The ability to form this complex is critical for the function of E4-34kDa in complementing E4 mutant growth (Blanchette et al., 2004). Our previous results indicate that E4-34kDa requires active proteasomes for its ability to promote late gene expression (Corbin-Lickfett and Bridge, 2003). Recently, Evans and Hearing (2005) have shown that the function of the E4-11kDa protein in promoting viral DNA replication depends on its ability to redistribute Mre11, and that cells lacking Nbs1 or Mre11 of the MRN complex complement the growth defect of an E4 mutant. Taken together, these results support a model in which inactivation of host DSBR proteins by either degradation or redistribution is a critical aspect of the ability of E4 proteins to promote a productive lytic infection. E4 mutant genome concatenation requires proteins of the MRN complex since cell lines lacking either Mre11 or Nbs1 fail to concatenate E4 mutant genomes (Stracker et al., 2002). However, the precise effect of genome concatenation in the virus life cycle is not yet clear. Shepard and Ornelles (2004) find that growth of an E1b-55kDa/E4- 11kDa double mutant is defective relative to Ad5 in both MO59J and MO59K cells, suggesting that failure to concatenate viral genomes was not sufficient to rescue mutant growth. Furthermore, Evans and Hearing (2003) have found that the ability of E4-11kDa to promote viral replication can be genetically uncoupled from its ability to prevent viral DNA concatenation. We find that failure to concatenate E4 mutant genomes substantially rescues its late gene expression defect following high multiplicity infections that do not

77 result in a DNA replication defect (Figs. 10-12). However, E4 mutants have a significant DNA replication defect relative to Ad5 at low MOI, irrespective of the cell's ability to concatenate viral genomes (Fig. 15). MO59J cells were unable to rescue E4 mutant virus yield at low MOI (Table 2), and we also find that inhibition of DSBR with wortmannin did not rescue either the DNA replication defect or the late gene expression defect of E4 mutants in cells infected at low multiplicity (data not shown). These results indicate that the DNA replication defect of E4 mutants is not rescued by the inability of cells to concatenate viral genomes, and suggest that E4 proteins play an important role in stimulating viral DNA replication irrespective of genome concatenation. We do not know the precise role of E4 proteins in promoting DNA replication at low MOIs. It is possible that in the absence of E4, DSBR proteins such as the MRN complex could inactivate viral genomes even in cells that are unable to complete genome concatenation. For example, the nuclease activity of Mre11 could still destroy the viral origin of replication in MO59J cells that are unable to complete "repair" and concatenate the viral genome. Alternatively, it is possible that the expression of other viral early proteins may compensate for an E4- mediated function in DNA replication at high MOI. The DNA replication defect of E4 mutants is alleviated by high multiplicity infection. However, E4 mutants display a severe reduction in late gene expression despite their ability to achieve normal levels of viral DNA (Bridge and Ketner 1989). This replication-independent late gene expression defect is substantially rescued by conditions that interfere with genome concatenation. E4 mutant late protein levels and virus yield are within 2-fold of Ad5 in MO59J cells; this is significantly improved from the respective 10- and 300-fold defects observed in MO59K cells that concatenate E4 mutant genomes (Fig. 10 and Table 2). We also find that the DSBR kinase inhibitors, wortmannin and caffeine, interfere with genome concatenation (Fig. 11) and significantly improve E4 mutant late protein levels (Fig. 12). DSBR kinase inhibitors could affect damage response signaling in addition to genome concatenation. However, Carson et al., (2003) did not see any effect of wortmannin treatment on E4-mutant induced phosphorylation of a battery of damage response proteins, and we did not find that wortmannin affected phosphorylation of either Nbs1 or Chk1 (Fig. 13). Caffeine is a known inhibitor of damage response signaling by ATM and ATR (Sarkaria et al., 1999);

78 it inhibits phosphorylation of Chk1 (Fig. 13B) and was previously shown to inhibit phosphorylation of several damage response proteins in E4 mutant infected cells (Carson et al., 2003). Thus, although wortmannin and caffeine have different effects on damage response signaling, they have similar effects on genome concatenation and late gene expression, suggesting that their ability to stimulate E4 mutant late gene expression is more tightly associated with inhibition of genome concatenation than with damage response signaling. Although our data do not rule out the possibility that wortmannin could stimulate E4 mutant late gene expression through an as yet untested effect on damage response signaling, to date our data are most consistent with a model in which inhibition of genome concatenation results in increased late gene expression. Wortmannin and caffeine had no effect on E4 mutant early gene expression (Fig. 14A) or DNA replication (Fig. 14B), but did stimulate the accumulation of late mRNA in the cytoplasm (Fig. 14C). Taken together, these data support the hypothesis that the DNA- independent defect in late mRNA accumulation exhibited by E4 mutants (Bridge and Ketner, 1989) can be substantially rescued by preventing genome concatenation. E4 mutant late gene transcription has been assayed by nuclear run-on analysis, and when normalized for the number of available DNA templates, the transcription rate is close to that of wild type Ad (Sandler and Ketner, 1989; Halbert et al., 1985). This suggests that newly synthesized transcripts from concatenated E4 mutant genomes are less stable than those produced in cells containing unconcatenated genomes. Why would genome concatenation interfere with the stability of newly synthesized transcripts? Gene expression is highly organized in the nuclear compartment of eukaryotic cells. Transcription and posttranscriptional processing of mRNA is tightly coordinated, and both mRNA processing and RNA export factors have been found to associate with elongating RNA transcripts (reviewed in Vinciguerra and Stutz, 2004). Concatenation of viral genomes might interfere with the normal coordination of transcription, processing and export and result in decreased nuclear stability of viral mRNA. We have previously reported that DNA replication centers that form in cells infected with E4 mutants are often abnormally large, and contain dense concentrations of the 72kDa DNA binding protein (Bridge et al., 2003). Aberrant organization of concatenated genome templates

79 might prevent proper access of newly synthesized RNA to processing and export factors, thereby explaining the reduced stability of E4 mutant late mRNA. Ad infection exposes a cell to exogenous genetic material and leads to the activation of cellular DSBR proteins that usually respond to abnormal and damaged chromosomal DNA (reviewed by Weitzman et al., 2004). The activity of DSBR proteins and their effect on the viral genome may pose a major obstacle to productive Ad infection that is overcome by viral regulatory proteins capable of inactivating host DSBR. DSBR could potentially interfere with viral activities at several steps in the virus life cycle. When Mre11 localizes to sites of E4 mutant DNA replication (Stracker et al., 2002), its nuclease activity could render the ends of Ad genomes incapable of initiating replication. We have found that Mre11 still localizes to E4 mutant replication centers in MO59J cells that lack DNA-PKcs and fail to form genome concatemers (data not shown). Thus, it is possible that the severe DNA replication defect we see following low multiplicity infection of either MO59J or MO59K cells (Fig. 15) stems from the activity of Mre11 on E4 mutant genomes. We show here that even if E4 mutant viral DNA is replicated, genome concatenation by host DSBR still interferes with the subsequent accumulation of viral late mRNA. Genome concatenation could also interfere with virion assembly and packaging. The ability of Ad to inactivate host DSBR is thus a critical aspect of establishing a productive infection, and provides an important new avenue for investigating virus host interactions.

80 Acknowledgments We are very grateful to Gary Ketner for providing MO59J and MO59K cells , and Arnold Levine and Ulf Pettersson for providing antibodies used in this study. We wish to thank Gary Janssen for critically reading the manuscript, Julie Shapiro for her assistance as an undergraduate researcher on this project, and all the members of our laboratory for their suggestions and support. This research was supported by the National Cancer Institute (grant CA82111), and awards from Miami University.

81 CONCLUDING REMARKS The integrity of genomic information is critical to the survival and propagation of all cellular organisms. Cells have evolved complex surveillance mechanisms that monitor genomic integrity during normal cell-cycle progression and in response to DNA damage, and they orchestrate a multifaceted response to DNA damage to ensure accurate transmission of genetic information. These responses are mediated by host DSBR proteins, which appear to have been conserved throughout eukaryotic evolution (Hartwell and Weinert, 1989; Hartwell et al., 1994; reviewed in Elledge, 1996). Infection with linear double-stranded DNA virus such as Ad, presents an assault on the host cell. Ad tries to use the host cell as a factory for producing more infectious virions, and the host cell responds with defense systems that create obstacles for the virus. It appears that this battle extends into the field of DNA damage recognition and repair. Recent developments in the field of DNA repair have provided insights into the role of the DSBR proteins as an innate intracellular defense against foreign invaders. The host DSBR proteins that have evolved to detect damaged chromosomal DNA can also recognize viral genomes (M. D. Weitzman, et al., 2004). The response of eukaryotic cells to double strand breaks (DSBs) in genomic DNA includes the sequestration of many factors into large nuclear domains called “foci”. Several host proteins involved in DNA repair or DNA damage response signaling accumulate in these nuclear foci, which represent the actual site of the DSB after exposure of cells to DNA damage such as ionizing radiation (Petrini and Stracker, 2003). The member of the histone H2A family protein, H2AX, becomes extensively phosphorylated (γH2AX) within 1–3 minutes of DNA damage and forms foci at the site of DNA damage. Introduction of DSBs elicits this response, but other types of damage, for instance UV irradiation, do not (Rogakou et al., 1999). γH2AX is thought to be the sensor for DSBs due to its early activation and subsequent localization at the sites of DNA damage (Paull et al., 2000). Such early activation and localization of γH2AX is required to signal the cell to activate and recruit other host DSBR proteins, which participate in DNA repair and checkpoint activation (Fernandez-Capetillo et al., 2002). Interestingly, E4 mutant infections induce early activation of γH2AX in HeLa cells suggesting its role as possible sensor for incoming viral genomes (Fig. 6). Mediator of

82 DNA damage checkpoint protein 1 (MDC1) is another DNA repair protein that is activated early during induction of DNA damage responses (Shang et al., 2003). MDC1 foci are observed as early as 1 min post ionizing radiation (IR) treatment and colocalize with γH2AX (Petrini and Stracker, 2003). There is also a significant increase in MDC1 foci early after E4 mutant infection, indicating its possible role as an additional sensor for incoming viral genomes (Mathew and Bridge, unpublished data). The Ad genome has a terminal protein attached to the 5’ ends; this could be a signal that triggers the activation and recruitment of both γH2AX and MDC1 to foci. These observations open up the possibility of understanding how the mammalian cells are able to “sense” the presence of viral genomes. It further suggests that activation of some aspects of the DNA damage response may respond to incoming genomes prior to the onset of viral DNA replication. The Ad replication occurs at discrete sites in nuclei, known as the replication centers, where the viral replication proteins cluster and synthesize a large amount of viral DNA. Ad DNA replication produces large quantities of linear double stranded viral DNA, ssDNA bound to viral 72kDa DNA binding-protein (DBP), and also unusual panhandle structures; these are all potential substrates for host DSBR (Van der vliet, 1995, Weitzman et al., 2004; Zijderveld et al., 1993). My work has suggested that replication of incoming viral genomes is required to activate repair and cell cycle checkpoint aspects of the DNA damage response (Fig. 4 and 5). The nature of Ad DNA structures that are recognized by the DSBR proteins to activate repair and checkpoint responses remains obscure; I will discuss the known triggers for activation of the DNA damage response in other model systems and discuss their relationship to Ad replication intermediates. Single-stranded DNA (ssDNA) mediated activation of the DNA damage response. The ability to recognize and repair abnormal DNA structures is common to all forms of life. The activation of DNA repair through recognition of DNA-protein interactions (nucleoprotein complexes) occurs in both prokaryotes and eukaryotes (Abraham and Tibbetts, 2005). The Escherichia coli “SOS” response is induced after the cell is exposed to DNA damaging agents, resulting in expression of many gene products that are involved in DNA damage tolerance and repair. The presence of ssDNA acts as the cell's internal signal that it has suffered DNA damage. When ssDNA is present, the

83 RecA protein polymerizes on ssDNA to form a helical, multimeric nucleoprotein complex that turns on the “SOS” response (Sutton et al., 2000). The use of ssDNA as a signal of damaged DNA is evolutionarily conserved. In mammalian cells, replication protein A (RPA) coated ssDNA is the critical structure at sites of DNA damage that recruits host DSBR proteins, like ATR and ATR-interacting protein (ATRIP) and facilitates the recognition of substrates for phosphorylation and the initiation of DNA repair and checkpoint activation (Ball et al., 2005). In yeast, the production of ssDNA at telomeres correlates with activation of checkpoint signaling (Vaze et al., 2002). Also in yeast, the activation of checkpoint signal intensity following an enzymatically induced DSB was shown to parallel the quantity of ssDNA generated by nuclease resection (Lydall and Weinert, 2004). Studies done in the frog Xenopus also found that ssDNA was required to activate RPA and ATR, which turn on the signaling cascade for DNA repair (Jullien, et al., 1999). Ad replication also generates numerous ssDNA templates bound to viral DNA binding-protein (DBP) (Van der vliet, 1995). The host DSBR proteins could recognize these nucleoprotein complexes as substrates for repair and activate the damage response signaling. In both the wild type Ad5 and E4 mutant infections there is accumulation of foci of ATR, ATRIP and RPA at viral replication centers, supporting the above hypothesis (Carson et al., 2003). Abnormal DNA structures activate the DNA damage response In eukaryotes, the MRN complex is involved in activating the DNA damage response in the presence of abnormal DNA structures. The MRN complex interacts with another group of proteins called the BASC complex (BRCA1-associated genome surveillance complex) that act as sensors of abnormal DNA structures caused both by distortions of the helix due to mismatches, and chemical alterations of the DNA structure by cisplatin, DNA-methylating agents, and other chemicals (Wang et al., 2000). The recruitment of MRN complex proteins to nuclear foci has attracted particular attention in recent years (Petrini and Stracker, 2003; Stracker et al., 2002; Carson et al., 2003). Mre11 is homogeneously distributed throughout the nucleus in resting, undamaged cells; however, within 30 min of treatment in vivo with agents that cause DNA DSBs, Mre11 relocates to nuclear foci. (Nelms et al., 1998). These properties of the Mre11 protein together with the observation the DSB association of MRN complex is correlated with checkpoint

84 activation and DNA repair, supports its role as a DNA damage sensor (Petrini and Stracker, 2003). Mre11 plays an upstream role in activating the DNA damage response mediated by the ATM and ATR kinases in response to virus infections (Carson et al., 2003; Lilley et al., 2004). The MRN complex also forms foci in response to E4 mutant infections surrounding the viral replication centers (Stracker et al., 2002). The presence of the unusual panhandle intermediates at viral replication centers could trigger the activation of MRN complex proteins in E4 mutant infections (Weitzman et al., 2004; Zijderveld et al., 1993). The activation of DNA damage response by unusual DNA structures is not unique to Ad. Adeno-associated virus (AAV) is a human Parvovirus whose genome consists of a ssDNA molecule with inverted terminal repeats (ITR) at either end. The ITR’s form double hairpin secondary structures that provide priming for replication and are also required for integration and packaging (Muzyczka and Berns, 2001). Microinjection of viral ITR’s into cells activates a cellular DNA damage response (Raj et al., 2001). Recently it was shown that the infection with Herpes simplex virus (HSV-1) also results in activation of the cellular DNA damage response mediated by the MRN complex. HSV- 1 has a double-stranded linear genome and its replication generates a large amount of unusual branched replication intermediates that trigger the MRN complex mediated activation of DNA damage response (Shirata et al., 2005; Lilley et al., 2004). Linear double stranded DNA activates a DNA damage response. Nuclear injection of linearized plasmid DNA can activate DNA damage response suggesting the role of DNA ends in activating a DNA damage response (Huang et al., 1996). Since Ad genomes are linear double stranded DNA, this could also be a potential trigger (Zijderveld et al., 1993). However, we find that replication-defective Ad mutants do not activate phosphorylation of damage response proteins Nbs1 and Chk1, suggesting that the presence of incoming viral genomes alone was an insufficient trigger to activate at least some aspects of DNA damage response (Fig. 4 and 5. Chapter 1). This suggests that the terminal protein bound to the 5’ ends of the viral DNA could have a protective role against the recognition of viral genomes as a substrate for repair. Further, an Mre11 mutant that is defective for nuclease activity is unable to concatenate the E4 mutant genomes, supporting the above hypothesis (Stracker et al., 2002). We speculate that

85 unusual replication intermediates activate the MRN complex, which then triggers the DNA damage response in the host cells, cleaving the terminal protein bound ends and “freeing” them for ligation by nonhomologous end joining. Activation of DNA damage response affects viral DNA replication and gene expression. The activation of the DNA damage response could pose an obstacle to productive Ad infection. Here I will discuss the potential consequences of activation of host DSBR proteins during Ad infections. During wild type Ad infections, the MRN complex is targeted for proteasome-mediated degradation by the E4-34kDa and E1b-55kDa complex; this prevents the activation of ATM-mediated DNA damage signaling and viral genome concatenation (Stracker et al., 2002; Carson et al., 2003). Additionally, the wild type Ad infections also lead to the redistribution of the MRN complex members by the E4-11kDa protein away from the viral replication centers into nuclear “track” like formations. The redistribution of the MRN complex proteins prevents concatenation of the viral genomes (Stracker et al., 2002; Carson et al., 2003). These results indicate that inactivating the host DSBR proteins could be critical step in productive Ad infection. The severe DNA replication defect exhibited by the E4 mutants at low multiplicities (Fig. 4 and Fig. 15) could be due to interactions of the host DSBR proteins with viral DNA ends. The origin of replication of Ad is situated in the ends of the DNA (Rawlins, 1984). The MRN complex possesses nuclease activity that could chew up the ends (Paull and Gellert, 1998); hence such templates could be poor substrates for DNA replication. Concatenation of viral genomes could also destroy Ad origins of replication. Experiments are currently in progress to identify the physical interaction of host DSBR proteins, including Mre11, with viral DNA (Mathew and Bridge, unpublished data). Interestingly, when cells are infected with an E4 mutant at high MOI, the virus accumulates normal levels of viral DNA. Therefore, at high multiplicities, an excess of viral DNA genomes may saturate the host DSBR response. It is also likely that at high MOI, viral replication proteins could compete more efficiently with the host DSBR proteins for binding to viral origin so that DNA replication remains unaffected. In support of this hypothesis, there is evidence that suggests that the viral DBP expressed from the E2 region is over produced in E4 mutant infections done at high multiplicities

86 (Medghalchi et al., 1997). These results support a model in which there is competition between the viral early proteins and host DSBR proteins to bind to the viral DNA following Ad infections. The balance is likely to shift in favor of the virus when there are high levels of viral genomes and early proteins. E4 mutant genomes replicate at normal levels during high multiplicity infections but are substantially concatenated (Weiden and Ginsberg, 1993; Boyer et al., 1999; Stracker et al., 2002; Jayaram and Bridge, 2005). E4 mutants are also severely defective for late message accumulation. This late RNA accumulation defect is not due to a reduction in late mRNA transcription rate (Bridge and Ketner, 1989; Sandler and Ketner, 1989). The viral replication centers formed during E4 mutant infections become abnormally large during late times post infection (Bridge et al., 2003). It is possible that these replication centers could have an aberrant organization of concatenated genome templates, which might prevent proper access of newly synthesized RNA to processing and export factors, thereby explaining the reduced stability of E4 mutant late mRNA. Another possibility is that genome concatemers that are formed from ligation of tail-tail joints could give rise to messages that are complementary to one another inducing an RNAi response against viral late RNA. The RNAi response is an intrinsic defense mechanism for combating virus infections in plants (Soosaar et al., 2005). The importance of RNAi in antiviral defense is also suggested by the fact that plant viruses have developed counter defensive strategies, encoding proteins that suppress different steps in RNA silencing (Pooggin and Hohn, 2004) The E4 proteins might play a similar role preventing induction of RNAi response in mammalian cells by preventing genome concatenation by host DSBR. My work indicates that concatenation of E4 mutant genomes inhibits viral late RNA accumulation. An E4 mutant expresses late genes more efficiently in a cell line that lacks the host DSBR protein, DNA-PKcs, and hence does not form genome concatemers (Fig. 10). Preventing genome concatenation improves late gene expression by stimulating the accumulation of viral late messages (Fig. 12 and Fig. 14). The E4 mutant virus yield is also higher in the absence of genome concatemers (Table 2). Taken together, these results indicate that a critical function of the E4 proteins in promoting efficient Ad lytic infections is inactivating host DSBR proteins which otherwise interfere with viral replication and gene expression at multiple steps.

87 Interfering with the activity of host DSBR proteins is also beneficial to other viruses including HSV-1. The HSV-1 immediate early protein, ICP0 induces the degradation of the host DSBR protein, DNA-PKcs by the ubiquitin mediated proteasome degradation pathway. DNA-PKcs is a nuclear enzyme which plays a role in non- homologous end joining (NHEJ) of DNA, in V(D)J recombination, regulation of transcription and apoptosis (Burma and Chen, 2004). HSV-1 replicates more efficiently in DNA-PKcs deficient cells (MO59J) than in DNA-PKcs proficient cells (MO59K), suggesting that the ICPO-mediated elimination of DNA-PKcs enhances virus replication. One hypothesis is that since DNA-PKcs is required for NHEJ of DNA, it may interfere with the proper replication and dynamics of progeny viral genomes (Parkinson et al., 1999). Another possibility is that elimination of DNA-PKcs activity alters the modification of key cellular regulatory factors that promote transcription of the viral DNA or translation or the functions of the viral protein products. DNA-PKcs is known to inhibit transcription by RNAP II so inhibition of DNA-PK activity may provide a mechanism for increasing the levels of RNAP II transcription on viral DNA (Lees-Miller et al., 1996). Single-stranded DNA viruses such as Adeno-associated virus (AAV) also encounter the potential blocks to productive infection by host DSBR proteins. The AAV genome consists of a single-stranded DNA molecule with inverted terminal repeats (ITR) at either end. The ITRs form double hairpin secondary structures that provide priming for replication and are also required for integration and packaging. During early stages of parvovirus infection the single-stranded DNA genome is translocated into the nucleus where it is converted into a double-stranded DNA genome that is the replicative form prior to expression of viral genes (Muzyczka and Berns, 2001). AAV is a defective virus in the sense that it generally needs co-infection with a helper virus, like Ad or HSV to replicate viral DNA efficiently (Muzyczka and Berns, 2001). The E4 34kDa and E1b 55kDa viral proteins provide some of the helper functions from Ad. These proteins also enhance expression of a transgene from a recombinant AAV vector by overcoming barriers to second-strand synthesis of the viral genome. The mechanism by which this occurs is undefined, but because E1b-55kDa and E4-34kDa proteins degrade the MRN complex, it seems plausible that this could contribute to the helper function for AAV.

88 Indeed, there is evidence to suggest that degradation of the MRN complex correlates with increased recombinant AAV transduction and replication (Weitzman et al., 2004). This suggests that one of the helper functions of Ad is to inactivate the host repair machinery that may otherwise present a barrier to AAV. This is further supported by a study with another parvovirus, where replication was more efficient in cells that were defective for DSBR. Its speculated that the DNA-PKcs and Ku proteins together interfere with the conversion of the ssDNA to dsDNA in AAV infections (Tauer et al., 1996). Therefore the helper functions provided by Ad could also involve inactivation of the DNA-PKcs by physical binding of E4-34kDa and E4-11kDa proteins (Boyer et al., 1999). It is evident that in the case of AAV the DNA repair machinery poses a block to productive virus infection, which the virus is able to overcome with help from coinfection with Ad or HSV. Viruses with RNA genomes that pass through a DNA intermediate during their infectious program also encounter the DNA repair machinery. Retroviruses are a family of RNA viruses that use virally encoded reverse transcriptase to generate a cDNA that is then integrated into the host genome (Goff, 2001). The pre-integration complex contains the cDNA copy as well as the viral-encoded integrase enzyme. The integrase enzyme mediates the integration of the viral DNA with the host chromosome in a non site- specific fashion (Yoder and Bushman, 2001). The host DSBR proteins of the NHEJ pathway have been implicated in sensing and processing of the linear cDNA molecule (Li et al., 2001). It is thought that the unintegrated viral cDNA is a substrate for NHEJ repair, circularizing the viral DNA by end ligation. This circularization of viral genomes causes a block to integration of the virus in the host chromosome. On the other hand, circularization of unintegrated DNA by host DSBR proteins may actually protect the cell from apoptosis (Baekelandt et al., 2000). In support of this result, retrovirus infection of cells with defects in the NHEJ pathway results in increased apoptosis, and this has led to the suggestion that DNA-PKcs and Ku provide a protective role against apoptosis induced by high levels of retrovirus cDNAs. Inactivating DSBR proteins is important for preventing apoptosis during many virus infections. DNA-PKcs is an important regulator of one of the main factors in the establishment of the antiviral state, interferon regulatory factor (IRF) 3. Interferons

89 (IFNs) are well-characterized components of the innate host defense, and rapid induction of IFN expression in response to viral infection requires posttranslational modification of transcription factors, like IRF3. DNA-PKcs is required to phosphorylate IRF3 in response to viral infections (Karpova et al., 2001). These studies highlight the functional link between the activation of host DSBR proteins and the innate immune response. DNA- PKcs is also required for the activation of tumor suppressor protein p53, that usually responds to DNA damage signals by inducing apoptosis in eukaryotic cells (Woo et al., 1998). The DNA repair machinery acts as an obstacle to productive Ad infection, so this virus has evolved multiple ways to inactivate the cellular DNA damage response. On the other hand, there are viruses that actually recruit the host DSBR proteins for crucial steps in their life cycle. For example, the retroviruses harness host DSBR proteins for integrating their genetic material into the host chromosome (Daniel et al., 1999). Intermediate products of HSV DNA replication are thought to be large head-to-tail concatemers with highly branched structures that might be generated by homologous recombination coupled with replication events. It is speculated that the MRN complex might actually promote the formation of stable replication structures through its role in homologous recombination mediated repair. In support of this, HSV-1 replication was found to be defective in ATLD cells that lack functional Mre11 protein, and hence do not activate a DNA damage response (Lilley et al., 2004). In this way the HSV-1 can manipulate the host DSBR proteins for its own benefit. Inactivation of the cellular DNA repair pathways during virus infections has obvious implications for the integrity of the host cell genome. As Ad targets the MRN complex to prevent concatenation of the viral genome, it is likely that it also affects DSBR processes in mammalian cell in general. It has been shown that interaction of E4- 34kDa with the DNA-PKcs can inhibit V(D)J recombination in a plasmid transfection assay (Boyer et al., 1999). Analysis of chromosomal DNA following ionizing radiation also suggests that the E1b-55kDa and E4-34kDa proteins can inhibit repair of cellular DSBs (Weitzman et al., 2004). Such interference with DNA repair can lead to genome instability and contribute to cellular transformation. In the case of Ad infections, targeting the MRN complex for inactivation provides a possible mechanism for cellular

90 transformation by the E4-34kDa and E4-11kDa oncoproteins by the “hit-and-run” mechanism. Even if the E4 gene products are not maintained in the transformed cell, damage done to cellular genome while E4 proteins were present could establish the transformed phenotype (Stracker et al., 2002). Understanding the relationship between the DNA damage response machinery and virus infections will provide insights into viral pathology and persistence, as well as possibly generating ideas for the development of antiviral drugs. For example, the ability of HSV-1 to cause persistent infections in neurons is related to its interactions with the host DSBR proteins. It may also guide the effective design of viral vectors for gene therapy. In case of Ad and AAV, designing gene therapy vectors could involve gutting the E4 proteins due to their potential to inactivate the host DSBR proteins. In addition, knowledge of virus–host interactions can illuminate key cellular regulators and aid in defining signaling pathways. For example, the studies with Ad and DNA damage signaling, in which viral degradation of the MRN complex suggested that it has an upstream role in damage detection (Carson et al., 2003). The goal of this study was to investigate the role of host DSBR proteins on the life cycle of an E4 mutant adenovirus. We propose the following model to describe the effect of host DSBR on E4 mutant replication and late gene expression (Fig. 16). Ad has a linear double-stranded DNA genome with terminal proteins attached to the 5’ ends. The viral early regulatory proteins produced during the early gene expression compete with the host DSBR proteins for binding to viral genomes. At high multiplicity infections when the incoming viral genomes saturate the host cell, the early proteins are able to bind to viral origin of replication and synthesize viral DNA to normal levels (Fig.4 and 15). On the other hand, when the infections are carried out at low multiplicities the host DSBR proteins titrate out the viral early proteins for binding to viral genomes leading to a viral replication defect (see Fig. 4). If the early regulatory proteins win the battle, viral DNA replication leads to the production of intermediates including ssDNA bound to early DNA binding protein (DBP) and panhandle structures formed by the association of inverted terminal repeats. These replication products may be “sensed” by the MRN complex, which is required for activating ATM and ATR kinases, which in turn phosphorylate the downstream damage response proteins, leading to either induction of

91 cell cycle checkpoint or apoptosis. Host DSBR proteins ligate the E4 mutant genomes end-to-end by non-homologous end joining, leading to concatenation (see Fig. 10). The ATR kinase might also play a role in repairing the E4 mutant genomes leading to concatenation (see Fig. 11). The late messages produced from the concatenated genomes are unstable (see Fig. 14). This late message accumulation defect translates into late protein synthetic defect leading to reduced virus yields (Fig. 14 and Table 2). In summary, I have shown that the E4 mutant DNA replication was required to induce phosphorylation of several damage response proteins (Chapter 1). The activated DSBR proteins pose a potential block to replication and expression of viral late genes by Ad. I have shown that E4 mutant genome concatenation by host DSBR proteins interferes with the accumulation of viral late messages contributing to the mutants late protein synthetic defect (Chapter 2). These data indicate that interfering with the activity of the host DSBR proteins is critical to establishing a productive infection. Understanding how the host cell senses and responds to viral genomes and replication intermediates will provide important insights into how cells recognize and repair their own damaged DNA.

92 Fig. 16. A model for the role of host DSBR proteins on the life cycle of an Adenovirus E4 mutant. See text for explanation of the model.

93 94 REFERENCES

Abraham, R. T., and R. S. Tibetts. 2005. Cell biology: Guiding ATM to broken DNA Science. 308: 510-511.

Allen, C., J. Halbrook, and J. A. Nickoloff. 2003. Interactive competition between homologous recombination and non-homologous end joining. Mol. Cancer Res. 1: 913- 920.

Allalunis-Turner, M. J., Barron, G. M, Day, R. S. 3rd, Dobler, K. D, and R. Mirzayans. 1993. Isolation of two cell lines from a human malignant glioma specimen differing in sensitivity to radiation and chemotherapeutic drugs. Radiat. Res. 134: 349- 354.

Alani, E., S. Lee, M. F. Kane, J. Griffith, and R. D. Kolodner. 1997. Saccharomyces cerevisiae MSH2, a mispaired base recognition protein, also recognizes Holliday junctions in DNA. J. Mol. Biol. 265: 289-301.

Araujo, F. D., T. H. Stracker, C. T. Carson, D. V. Lee, and M. D. Weitzman. 2005. Adenovirus type 5 E4orf3 protein targets the Mre11 complex to cytoplasmic aggresomes. J Virol. 79: 11382-11391.

Aspegren, A., C. Rabino and E. Bridge. 1998. Organization of splicing factors in adenovirus-infected cells reflects changes in gene expression during the early to late phase transition. Exp. Cell Res. 245: 203-213.

Babiss, L. E., H. S. Ginsberg, and J. E. Darnell, Jr. 1985. Adenovirus E1B proteins are required for accumulation of late viral mRNA and for effects on cellular mRNA translation and transport. Mol Cell Biol. 5: 2552-2558.

95 Baekelandt, V., A. Claeys, P. Cherepanov, E. De Clercq, B. De Strooper, B. Nuttin, and Z. Debyser. 2000. DNA-Dependent protein kinase is not required for efficient lentivirus integration. J Virol. 74(23): 11278-11285.

Bakkenist, C. J., and M. B. Kastan. 2003. DNA damage activates ATM through intermolecular autophosphorylation and dimer dissociation. Nature. 421: 499-506.

Ball, H. L., J. S. Myers, and D. Cortez. 2005. ATRIP binding to replication protein A- single-stranded DNA promotes ATR-ATRIP localization but is dispensable for Chk1 phosphorylation. Mol Biol Cell. 16: 2372-2381.

Bennett, R. J., J. L. Keck, and J. C. Wang. 1999. Binding specificity determines polarity of DNA unwinding by the Sgs1 protein of S. cerevisiae. J. Mol. Biol. 289: 235- 248.

Blanchette, P., C. Y. Cheng, Q. Yan, G. Ketner, D. A. Ornelles, T. Dobner, R. C. Conaway, J. W. Conaway, and P. E. Branton. 2004. Both BC-box motifs of adenovirus protein E4orf6 are required to efficiently assemble an E3 ligase complex that degrades p53. Mol. Cell. Biol. 24: 9619-9629.

Block, W.D., D. Merkle, K. Meek, and S. P. Lees-Miller. 2004. Selective inhibition of the DNA-dependent protein kinase (DNA-PK) by the radiosensitizing agent caffeine. Nucleic Acids Res. 32:1967-1972.

Boyer, J., K. Rohleder, and G. Ketner. 1999. Adenovirus E4 34k and 11k inhibit double strand break repair and are physically associated with the cellular DNA-dependent protein kinase. Virology. 263: 307-312.

Bridge, E., and G. Ketner. 1989. Redundant control of Adenovirus late gene expression by early region 4. J. Virol. 63: 631-638.

96 Bridge, E., and G. Ketner. 1990. Interaction of adenoviral E4 and E1b products in late gene expression. Virology 174: 345-353.

Bridge, E., S. Medghalchi , S. Ubol, M. Leesong, and G. Ketner. 1993. Adenovirus early region 4 and viral DNA synthesis. Virology.193: 794-801.

Bridge, E., K. Mattsson, A. Aspegren, and A. Sengupta. 2003. Adenovirus early region 4 promotes the localization of splicing factors and viral RNA in late-phase interchromatin granule clusters. Virology. 311: 40-50.

Burma S., and D. J. Chen. 2004. Role of DNA-PK in the cellular response to DNA double-strand breaks. DNA Repair (Amst). (8-9): 909-918.

Carter C. C., R. Izadpanah, E. Bridge. 2003. Evaluating the role of CRM1-mediated export for adenovirus gene expression. Virology. 315(1): 224-233.

Carson, C.T., R. A. Schwartz, T. H. Stracker, C. E. Lilley, D. V. Lee, and M. D. Weitzman. 2003. The Mre11 complex is required for ATM activation and the G2/M checkpoint. EMBO J. 22: 6610-6620.

Calsou, P., P. Frit, O. Humbert, C. Muller, D. J. Chen, and B. Salles. 1999. The DNA-dependent protein kinase catalytic activity regulates DNA end processing by means of Ku entry into DNA. J Biol Chem. 274(12):7848-7856.

Challberg, M. D., and T. J. Kelly. 1989. Animal virus DNA replication. Annu. Rev. Biochem. 58: 671-717.

97 Chan, D.W., D.P. Gately, S. Urban, A. M. Galloway, S. P. Lees-Miller, T. Yen, and J. Allalunis-Turner. 1998. Lack of correlation between ATM protein expression and tumour cell radiosensitivity. Int. J. Radiat. Biol. 74: 217-224.

Chan, D. W., C. H. Mody, N. S. Ting, and S. P. Lees-Miller.1996. Purification and characterization of the double-stranded DNA-activated protein kinase, DNA-PK, from human placenta. Biochem Cell Biol. 74(1): 67-73.

Chan, D. W., R. Ye, C. J. Veillette, and S. P. Lees-Miller. 1999. DNA-dependent protein kinase phosphorylation sites in Ku 70/80 heterodimer. Biochemistry. 38(6): 1819- 1828.

Chroboczek, J., R. W. Ruigrok, and S. Cusack. 1995. Adenovirus fiber. Curr Top Microbiol Immunol. 199:163-200.

Corbin-Lickfett, K. A., and E. Bridge. 2003. Adenovirus E4-34kDa requires active proteosomes to promote late gene expression. Virology. 315: 234-244.

Critchlow, S. E., and S. P. Jackson. 1998. DNA end-joining: from yeast to man. Trends Biochem Sci. 23(10): 394-398.

Daniel, R., R. A. Katz, and A. M. Skalka.1999. A role for DNA-PK in retroviral DNA integration. Science. 284(5414): 644-647.

Davison, A. J., M. Benko, and B. Harrach. 2003. Genetic content and evolution of adenoviruses. J. Gen. Virol. 84: 2895-2908.

Dobner, T., and J. Kzhyshkowska. 2001. Nuclear export of adenovirus RNA. Curr Top Microbiol Immunol. 259: 25-54.

98 Dobbelstein, M., J., W. T. Roth, A. J. Kimberly, and T. Shenk. 1997. Nuclear export of the E1B 55-kDa and E4 34-kDa adenoviral oncoproteins mediated by a rev-like signal sequence. EMBO J 16:4276-4284.

Dosch, T., F. Horn, G. Schneider, F. Kratzer, T. Dobner, J. Hauber, and R. H. Stauber. 2001. The adenovirus type 5 E1B-55K oncoprotein actively shuttles in virusinfected cells, whereas transport of E4orf6 is mediated by a CRM1-independent mechanism. J Virol. 75: 5677-5683.

Dynan, W.S., and S. Yoo. 1998. Interaction of Ku protein and DNA-dependent protein kinase catalytic subunit with nucleic acids. Nucleic Acids Res. 26(7): 1551-1559.

Elledge, S. J. 1996. Cell cycle checkpoints: preventing an identity crisis. Science. 274(5293): 1664-1672.

Evans, J. D., and P. Hearing. 2003. Distinct roles of the Adenovirus E4 ORF3 protein in viral DNA replication and inhibition of genome. J Virol. 77: 5295-5304.

Evans, J. D., and P. Hearing. 2005. Relocalization of the Mre11-Rad50-Nbs1 complex by the adenovirus E4 ORF3 protein is required for viral replication. J. Virol. 79: 6207- 6215.

Fernandez-Capetillo, O., H. T. Chen, A. Celeste, I. Ward, P. J. Romanienko, J. C. Morales, K. Naka, Z. Xia, R. D. Camerini-Otero, N. Motoyama, P. B. Carpenter, W. M. Bonner, J. Chen, and A. Nussenzweig. 2002. DNA damage-induced G2-M checkpoint activation by histone H2AX and 53BP1. Nat Cell Biol. 4(12): 993-997.

Friesner, J. D., B. Liu, K. Culligan, and A. B. Britt. 2005. Ionizing radiation- dependent gamma-H2AX focus formation requires ataxia telangiectasia mutated and ataxia telangiectasia mutated and Rad3-related. Mol Biol Cell. 16: 2566-2576.

99 Gately, D.P., J. C. Hittle, G. K. Chan, and T. J. Yen. 1998. Characterization of ATM expression, localization, and associated DNA-dependent protein kinase activity. Mol Biol Cell. 9(9): 2361-2374.

Goff, S. P. 2001. Intracellular trafficking of retroviral genomes during the early phase of infection: viral exploitation of cellular pathways. J Gene Med. 3(6): 517-528.

Halbert, D.N., J. R. Cutt, and T. Shenk, T. 1985. Adenovirus early region 4 encodes functions required for efficient DNA replication, late gene expression, and host cell shut off. J. Virol. 73: 5333-5344.

Hammarsten, O., L. G. DeFazio, and G. Chu. 2000. Activation of DNA-dependent protein kinase by single-stranded DNA ends. J Biol Chem. 275(3): 1541-1550.

Hardwick, J. M. 1998. Viral interference with apoptosis. Semin Cell Dev Biol. 9(3): 339-349.

Hartwell, L. H., and T. A. Weinert. 1989. Checkpoints: controls that ensure the order of cell cycle events. Science. 246(4930): 629-634.

Hartwell, L. H., and M. B. Kastan. 1994. Cell cycle control and cancer. Science. 266(5192): 1821-1828.

Hearing, P., R. J. Samulski, W. L. Wishart, and T. Shenk. 1987. Identification of a repeated sequence element required for efficient encapsidation of the adenovirus type 5 chromosome. J Virol. 61(8): 2555-2558.

Huang, M.M., and P. Hearing. 1989. Adenovirus early region 4 encodes two gene products with redundant effects in lytic infection. J. Virol. 63 (6), 2605-2615.

100 Huang, L. C., K. C. Clarkin, and G. M. Wahl. 1996. Sensitivity and selectivity of the DNA damage sensor responsible for activating p53-dependent G1 arrest Proc Natl Acad Sci U S A. 93(10): 4827-4832.

Imperiale, M. J., G. Akusjnarvi, and K. N. Leppard. 1995. Post-transcriptional control of adenovirus gene expression. Curr Top Microbiol Immunol. 199: 139-171.

Izzard, R.A., S. P. Jackson, and G. C. Smith. 1999. Competitive and noncompetitive inhibition of the DNA-dependent protein kinase. Cancer Res. 59: 2581-2586.

Jackson, S. P. 1997. DNA-dependent protein kinase. Int J Biochem Cell Biol. 7: 935- 938.

Jackson, S. P. 2002. Sensing and repairing DNA double-strand breaks Carcinogenesis, 23: 687-696.

Jayaram, S., and B. Bridge. 2005. Genome concatenation contributes to the late gene expression defect of an adenovirus E4 mutant. Virology. 342(2): 286-296.

Jullien, D., D. Gorlich, U. K. Laemmli, and Y. Adachi. 1999 Nuclear import of RPA in Xenopus egg extracts requires a novel protein XRIP alpha but not importin alpha. EMBO J. 18(15): 4348-4358.

Jeggo, P. A., A. M. Carr, and A. R. Lehmann. 1998. Splitting the ATM: distinct repair and checkpoint defects in ataxia-telangiectasia. Trends Genet. 14(8): 312-316.

Khanna, K. K., and S. P. Jackson. 2001. DNA double strand breaks: signaling, repair and the cancer connection. Nature genetics. 27: 247-254

101 Karpova, A.Y., L. V. Ronco, and P. M. Howley. 2001. Functional characterization of interferon regulatory factor 3a (IRF-3a), an alternative splice isoform of IRF-3. Mol Cell Biol. 21(13): 4169-4176.

Kratzer, F., O. Rosorius, P. Heger, N. Hirschmann, T. Dobner, J. Hauber, and R. H. Stauber. 2000. The adenovirus type 5 E1B-55K oncoprotein is a highly active shuttle protein and shuttling is independent of E4orf6, p53 and Mdm2. Oncogene 19:850-857.

Kurz, E.U., and S. P. Lees-Miller. 2004. DNA damage-induced activation of ATM and ATM-dependent signaling pathways. DNA Repair. 3: 889-900.

Lilley, C.E., C. T. Carson, A. R. Muotri, F. H. Gage, and M. D. Weitzman. 2004. DNA repair proteins affect the lifecycle of herpes simplex virus 1. Proc Natl Acad Sci U S A. 102(16): 5844-5849.

Lisby, M., and R. Rothstein. 2005. Localization of checkpoint and repair proteins in eukaryotes. Biochimie. 87: 579-589.

Li, L., J. M. Olvera, K. E. Yoder, R. S. Mitchell, S. L. Butler, M. Lieber, S. L. Martin, and F. D. Bushman. 2001. Role of the non-homologous DNA end joining pathway in the early steps of retroviral infection. EMBO J. 20(12): 3272-3281.

Leppard, K. N. 1998. Regulated RNA processing and RNA transport during adenovirus infection. Semin Virol 8:301-307.

Lee, J. H., and T. T. Paull. 2005. ATM activation by DNA double-strand breaks through the Mre11-Rad50-Nbs1 complex. Science. 308(5721): 551-554.

102 Lees-Miller, S. P., M. C. Long, M. A. Kilvert, V. Lam, S. A. Rice, and C. A. Spencer. 1996. Attenuation of DNA-dependent protein kinase activity and its catalytic subunit by the herpes simplex virus type 1 transactivator ICP0. J Virol. 70(11): 7471- 7477.

Lydall, D., and T. Weinert. 2004 From DNA damage to cell cycle arrest and suicide: a budding yeast perspective. Curr Opin Genet Dev. 6(1):4-11.

Maser, R. S., K. J. Monsen, B. E. Nelms, and J. H. Petrini.1997 hMre11 and hRad50 nuclear foci are induced during the normal cellular response to DNA double-strand breaks Mol Cell Biol.17(10):6087-6096.

Marsischky, G.T., S. Lee, J. Griffith, and R.D. Kolodner. 1999. Saccharomyces cerevisiae MSH2/6 complex interacts with Holliday junctions and facilitates their cleavage by phage resolution enzymes. J. Biol. Chem. 274: 7200-7206.

Muzyczka, N. 1994. Adeno-associated virus (AAV) vectors: will they work? J Clin Invest. 94(4): 1351

Medghalchi, S., R. Padmanabhan, and G. Ketner. 1997. Early region 4 modulates adenovirus DNA replication by two genetically separable mechanisms. Virology. 236: 8- 17

Moreau, S., J. R. Ferguson, and L. S. Symington. 1999. The nuclease activity of Mre11 is required for meiosis but not for mating type switching, end joining, or telomere maintenance. Mol Cell Biol. 19: 556-566

Nelms, B. E., R. S. Maser, J. F. MacKay, M. G. Lagally, and J. H. Petrini. 1998. In situ visualization of DNA double-strand break repair in human fibroblasts. Science. 280(5363): 590-592.

103 Nick McElhinny, S. A., C. M. Snowden, J. McCarville, and D. A. Ramsden. 2000. Ku recruits the XRCC4-ligase IV complex to DNA ends. Mol Cell Biol. 20(9):2996- 3003.

Paull, T. T., and M. Gellert. 1999 Nbs1 potentiates ATP-driven DNA unwinding and endonuclease cleavage by the Mre11/Rad50 complex. Genes Dev. 13(10): 1276-1288.

Parkinson, J., S. P. Lees-Miller, and R. D. Everett. 1999. Herpes simplex virus type 1 immediate-early protein vmw110 induces the proteasome-dependent degradation of the catalytic subunit of DNA-dependent protein kinase. J Virol. 73(1): 650-657.

Paull, T. T., E. P. Rogakou, V. Yamazaki, C. U. Kirchgessner, M. Gellert, and W. M. Bonner. 2000 A critical role for histone H2AX in recruitment of repair factors to nuclear foci after DNA damage. Curr Biol. 10(15): 886-895.

Phillipson, L. 1961. Adenovirus assay by fluorescent cell counting procedure. Virology 15: 263-268.

Pettersson, U., and R.J. Roberts. 1986. Adenovirus gene expression and replication: A historical review, p. 37-57, In Cancer Cells 4/DNA Tumor Viruses, Cold Spring Harbor Laboratory, New York, N.Y.

Petrini, J. H., and T. H. Stracker. 2003. The cellular response to DNA double-strand breaks: defining the sensors and mediators. Trends Cell Biol. 13(9): 458-462.

Pooggin, M., and T. Hohn. 2004 Fighting geminiviruses by RNAi and vice versa. Plant Mol Biol. 55(2): 149-152.

Querido, E., P. Blanchette, Q. Yan, T. Kamura, M. Morrison, D. Boivin, W. G. Kaelin, R. C. Conaway, J. W. Conaway, and P.E. Branton. 2001. Degradation of p53

104 by adenovirus E4orf6 and E1B55K proteins occurs via a novel mechanism involving a Cullin-containing complex. Genes Dev. 15: 3104-3117.

Rabino, C., A. Aspegren, K. Corbin-Lickfett, and E. Bridge. 2000. Adenovirus late gene expression does not require a Rev-like nuclear RNA export pathway. J Virol 74: 6684-6688.

Rawlins, D. R., P. J. Rosenfeld, R. J. Wides, M. D. Challberg, and T. J. Kelly. 1984 Structure and function of the adenovirus origin of replication. Cell. 37(1): 309-319

Rogakou, E. P., C. Boon, C. Redon, and W. M. Bonner. 1999. Megabase chromatin domains involved in DNA double-strand breaks in vivo. J Cell Biol. 146(5): 905-916.

Raj, K., P. Ogston, and P. Beard. 2001. Virus-mediated killing of cells that lack p53 activity. Nature. 412(6850): 914-917.

Sandler, A., and G. Ketner. 1989. Adenovirus early region 4 is essential for normal stability of late nuclear RNAs. J. Virol. 63: 624-630.

Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. "Molecular Cloning: A Laboratory Manual" second ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

Sarkaria, J.N., E. C. Busby, R. S. Tibbetts, P. Roos, Y. Taya, L. M. Karnitz, and R. T. Abraham. 1999. Inhibition of ATM and ATR kinase activities by the radiosensitizing agent, caffeine. Cancer Res. 59: 4375-4382.

Shirata, N., A. Kudoh, T. Daikoku, Y. Tatsumi, M. Fujita, T. Kiyono, Y. Sugaya, H. Isomura, K. Ishizaki, and T. Tsurumi. 2005 Activation of ataxia telangiectasia-mutated DNA damage checkpoint signal transduction elicited by herpes simplex virus infection. J Biol Chem. 280(34): 30336-30341.

105 Stracker, T.H., C. T. Carson, and M. D. Weitzman. 2002. Adenovirus oncoproteins inactivate the Mre11-Rad50-NBS1 DNA repair complex. Nature. 418: 348-352.

Shang, Y.L. , A. J. Bodero, and P. L. Chen. 2003. NFBD1, a novel nuclear protein with signature motifs of FHA and BRCT, and an internal 41-amino acid repeat sequence, is an early participant in DNA damage response J Biol Chem. 278(8): 6323-6329.

Shechter, D., V. Costanzo, and J. Gautier. 2004. Regulation of DNA replication by ATR: signaling in response to DNA intermediates. DNA Repair. 3: 901-908.

Shenk, T. 1996. Adenoviridae: the viruses and their replication., p. 2111-2148. In B. N. Fields, Knipe, D.M., Howley, P.M. (ed.), Virology. Lippincott-Raven, New York.

Shepard, R.N., and D. A. Ornelles. 2004. Diverse roles for E4orf3 at late times of infection revealed in an E1B 55-kilodalton protein mutant background. J. Virol. 78: 9924-9935.

Smith, G.C., and S. P. Jackson. 1999. The DNA-dependent protein kinase. Genes Dev. 13: 916-934.

Soosaar, J. L., T. M. Burch-Smith, and S. P. Dinesh-Kumar. 2005. Mechanisms of plant resistance to viruses. Nat Rev Microbiol. 3(10): 789-798.

Sutton, M. D., B. T. Smith, V. G. Godoy, and G. C. Walker. 2000. The SOS response: recent insights into umuDC-dependent mutagenesis and DNA damage tolerance. Annu Rev Genet. 34: 479-497.

Tauber, B., and T. Dobner. 2001. Molecular regulation and biological functions of Adenovirus early genes: the E4 ORFs. Gene. 278: 1-23.

106 Tibbetts, R.S., D. Cortez, K. M. Brumbaugh, R. Scully, D. Livingston, S. J. Elledge, and R. T. Abraham. 2000. Functional interactions between BRCA1 and the checkpoint kinase ATR during genotoxic stress. Genes Dev. 2000 14(23): 2989-3002.

Tomko, R. P., C. B. Johansson, M. Totrov, R. Abagyan, J. Frisen, and L. Philipson. 2000. Expression of the adenovirus receptor and its interaction with the fiber knob. Exp Cell Res. 55 (1): 47-55

Tauer, T. J., M. H. Schneiderman, J. K. Vishwanatha, and S. L. Rhode. 1996. DNA double-strand break repair functions defend against parvovirus infection. Genes Dev. 14(23): 2989-3002.

Uchiumi, F., T. Ohta, and S. Tanuma. 1996. Replication factor C recognizes 5'- phosphate ends of telomeres. Biochem. Biophys. Res. Commun. 229: 310-315.

Van der vliet, P. C. 1995. Adenovirus DNA replication. Curr. Top. Microbiol. Immunol. 199: 1-30.

Vaze, M.B., A. Pellicioli, S. E. Lee, G. Ira, G. Liberi, A. Arbel-Eden, M. Foiani, and J. E. Haber. 2002. Recovery from checkpoint-mediated arrest after repair of a double- strand break requires Srs2 helicase Mol Cell. 10(2): 373-385.

Vinciguerra, P., and F. Stutz. 2004. mRNA export: an assembly line from genes to nuclear pores. Curr. Opin. Cell Biol. 16: 285-292.

Wang, Y., D. Cortez, P. Yazdi, N. Neff, S. J. Elledge, and J. Qin. 2000. BASC, a super complex of BRCA1-associated proteins involved in the recognition and repair of aberrant DNA structures. Genes Dev. 14: 927-939.

107 Weterings, E., and D. C. van Gent. 2004. The mechanism of non-homologous end- joining: a synopsis of synapsis. DNA Repair 3: 1425-1435.

Weitzman, M.D., C. T. Carson, R. A. Schwartz, and C. E. Lilley. 2004. Interactions of viruses with the cellular DNA repair machinery. DNA repair. 3: 1165-1173.

Weinberg, D.H., and G. Ketner. 1983. A cell line that supports the growth of a defective early region 4 deletion mutant of human adenovirus type 2. Proc. Natl. Acad. Sci. USA. 80: 5383-5386.

Weinberg, D.H., and G. Ketner. 1986. Adenoviral early region 4 is required for efficient viral DNA replication and for late gene expression. J. Virol. 57: 833-838.

Weiden, M.D., and H. S. Ginsberg. 1994. Deletion of the E4 region of the genome produces adenovirus DNA concatemers. Proc. Natl. Acad. Sci. USA. 91: 153-157.

White, E. 1995. Regulation of p53-dependent apoptosis by E1A and E1B. Curr Top Microbiol Immunol. 199:34-58.

Woo, R. A., K. G. McLure, S. P. Lees-Miller, D. E. Rancourt , and P. W. Lee. 1998. DNA-dependent protein kinase acts upstream of p53 in response to DNA damage. Nature. 394(6694): 700-704.

Yoder, K. E., and F. D. Bushman. 2001. Repair of gaps in retroviral DNA integration intermediates. J Virol. 74(23): 11191-111200.

Zou, L., and S. J. Elledge. 2003. Sensing DNA damage through ATRIP recognition of RPA-ssDNA complexes. Science 300(5625): 1542-1548.

108 Zijderveld, D. C., M. H. Stuiver, and P. C. van der Vliet. 1993. The adenovirus DNA binding protein enhances intermolecular DNA renaturation but inhibits intramolecular DNA renaturation. Nucleic Acids Res. 21: 2591-2598.

109