Viewing My Dissertation

MIAMI UNIVERSITY The Graduate School Certificate for Approving the Dissertation

We hereby approve the Dissertation of Dipendra Gautam

Candidate for the Degree: Doctor of Philosophy

Dr. Eileen Bridge, Mentor

Dr. Gary R. Janssen, Reader

Dr. Joseph M. Carlin, Reader

Dr. Xiao-Wen Cheng

Dr. David G. Pennock Graduate School Representative

ABSTRACT

ATAXIA-TELANGIECTASIA MUTATED INTERFERES WITH ADENOVIRUS EARLY REGION 4 MUTANT DNA REPLICATION

by Dipendra Gautam The integrity of eukaryotic genomes is maintained by sophisticated networks of DNA damage mediator and effector proteins. Upon DNA damage, a well-orchestrated DNA damage response (DDR) is initiated by proteins that block cell cycle progression and repair the DNA or induce apoptosis as a last resort. DNA damage is sensed by a sensor complex composed of the Mre11/Rad50/Nbs1 (MRN) proteins that relay the damage signal to activate ataxia-telangiectasia mutated (ATM) and ATM-Rad3- related (ATR) kinases. These central kinases phosphorylate numerous DNA repair proteins and checkpoint kinases. In addition, DNA repair proteins are recruited to the site of DNA damage resulting in the formation of distinct nuclear foci. These signaling networks also respond to Adenovirus (Ad) infection and could potentially interfere with a productive life cycle. In wild type Ad infections, early region 4 gene products target Mre11 for degradation thereby interfering with its ability to sense DNA damage and activate phosphorylation cascades important during DNA repair. E4 mutants that lack these regulatory proteins activate a DNA damage response (DDR), which in turn interferes with a productive viral infection. E4 mutant phenotypes include genome concatemer formation and defective viral DNA and late protein synthesis. E4 mutant growth defects are substantially rescued in cells lacking a functional MRN complex; however the mechanism by which the MRN complex interferes with E4 mutant DNA replication is unknown. We investigated the role of ATM and ATR in interfering with a productive E4 mutant infection. We did not identify a role for ATM and ATR in E4 mutant genome concatenation. In contrast, E4 mutant DNA replication centers were well developed and DNA replication was rescued in cells that lack ATM or in cells treated with ATM kinase inhibitors. Our data show that the MRN complex inhibits E4 mutant DNA replication through its ability to activate ATM, which interferes with E4 mutant DNA replication by localizing to viral replication centers. Further we found that ATM-dependent recruitment of late DDR proteins including RNF8 and 53BP1 correlates with the poor development of replication centers. In conclusion, ATM plays an important role in activating DNA damage responses that interfere with viral DNA replication.

ATAXIA-TELANGIECTASIA MUTATED INTERFERES WITH ADENOVIRUS EARLY REGION 4 MUTANT DNA REPLICATION

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

Dipendra Gautam Miami University Oxford, Ohio 2013

Dissertation Director: Eileen Bridge, Ph.D.

Table of Contents

List of Figures iii

Acknowledgements v

General Introduction 1

Chapter I. The Kinase Activity of Ataxia-Telangiectasia 17

Mutated Interferes with Adenovirus E4 Mutant DNA Replication

Chapter II. Ataxia telangiectasia mutated kinase is important 53 for recruiting late DNA damage repair proteins to Adenovirus

E4 mutant DNA replication centers.

Concluding Remarks 83

References 93

List of Figures

Figure 1. Diagrammatic representation of Adenovirus genome and early 3 region 4 (E4).

Figure 2: Model of Adenovirus DNA replication. 6

Figure 3: Schematic representation of double strand break repair 10 pathways.

Figure 4: Model for IRIF formation. 15

Figure 5. ATM and ATR are not required for genome concatenation. 27

ATM and ATR are not required for genome concatenation.

Figure 6. ATM interferes with E4 mutant late protein production while 30

ATR does not.

Figure 7. ATM interferes with E4 mutant DNA replication. 32

Figure 8. The kinase activity of ATM is important for inhibiting E4 35 mutant DNA replication.

Figure 9. Time course showing pATM localization and BrdU 38 incorporation relative to DNA replication centers in Ad5 and E4- infections.

Figure 10. ATM inhibition facilitates progression of E4- DNA 41 replication centers from small early replication foci to larger intermediate- and late-stage replication centers.

Figure 11. Localization of Mre11 or ATM to viral replication centers 45 does not interfere with E4- DNA replication when the ATM kinase is inactivated.

Figure 12. An Nbs1 mutant lacking the ATM binding motif partially 61 rescues E4- DNA replication.

iii

Figure 13. Reduced pATM localization to E4- replication centers in 64 Nbs1∆ATM cells. Figure 14. pATM fails to localize to 1010 replication centers. 66

Figure 15. Mdc1, γH2AX colocalize with Mre11 at E4- DNA 71 replication centers independent of ATM. Figure 16. RNF8 and 53BP1 recruitment to pATM containing E4- 73

DNA replication centers.

Figure 17. Localization of RNF8 and 53BP1 to E4- replication centers 75 requires ATM.

Figure 18. Failure to localize pATM to replication centers results in 87 inefficient recruitment of 53BP1.

Figure 19. Models for ATM mediated inhibition of E4 mutant 90

DNA replication.

Acknowledgements

I would like to express my sincere gratitude to my advisor Dr. Eileen Bridge for her guidance and constant support to shape my scientific upbringing. You had been there as a supportive mentor at times when I was despaired and helped me find my way. You have always encouraged me to improve my abilities to become a better scientist and a better writer. Your exceptional support during my dissertation writing was amazing. Without you, graduate school would have never been possible.

I am equally indebted to my committee members who have shown me the path to success. Dr. Gary Janssen, you were very supportive as well as encourager while reviewing my dissertation. Dr. Joseph Carlin, you have been very helpful not only with my scientific endeavor but also while teaching MBI405. Dr. Xiao-Wen Cheng, I have always enjoyed chatting with you and I also appreciate your help in understanding the experimental procedure and also the opportunity to collaborate with you. Dr. David Pennock, I have always admired your scientific thoughts and directions during my committee meetings.

Microbiology department gave me the opportunity to meet great people. Aanand, it was wonderful working with you. I shall never forget our never ending philosophical conversation. And I always hope that I have a friend like you everywhere I go. Bill and Steve you have been very helpful especially while writing my dissertation. Dr. Shomita Mathew, I will always remember you as an exemplary scientist and a good person. Tyler, I wish you good luck for your future. I am also thankful to the students and staffs of Microbiology department who made my stay memorable. Barb and Darlene, thank you for all you help and solution to my questions and problems.

General Introduction

The family Adenoviridae consists of a large group of DNA viruses that infect mammals (Mastadenovirus) and avian species (Aviadenovirus). There are 47 serotypes of adenovirus (Ad) that infect humans and have been classified into six sub-groups, A-F (Davison et al., 2003). Ad causes respiratory tract infections, gastroenteritis and conjunctivitis with little clinical impact (Echavarria, 2008). In patients with reduced immune function, Ad can cause debilitating inflammation including pneumonia, cystitis and colitis. Although a few types of Ad can transform rodent cells (Trentin et al., 1962), there has been no evidence to suggest that Ad causes malignant disease in humans (Graham et al., 1977). While its clinical impact is minor, Ad has been extensively studied as a model system for DNA replication and gene expression and for its ability to transform cells as a result of viral oncoprotein expression.

Historical perspective

Ad was first isolated from human adenoids by Wallace Rowe and his colleagues in 1950s and hence the family was named Adenoviridae (Shenk, 1996). Ad was the first virus to demonstrate oncogenic potential in baby hamsters (Trentin et al., 1962). Since then Ad has served the scientific community as a tool to study mechanisms of molecular biology and has provided substantial insights. This includes the discovery of mRNA splicing, and insights into basic cellular processes such as alternative polyadenylation, transcription enhancers, and inactivation of tumor suppressor proteins (Berget et al., 1977; Graham et al., 1977, Whyte et al., 1988). In recent years much of the focus of Ad research has turned to the development of vectors for gene expression and gene therapy (Evans and Hearing, 2002), but Ad continues to be an important tool for probing cellular processes and investigating virus host cell interactions.

Morphology and Genome

Members of the Adenoviridae family have non-enveloped virions that exhibit icosahedral symmetry with a diameter ranging from 70-90 nm. The capsid is comprised of 720 hexon subunits arranged as 240 trimers and 12 penton capsomeres with a projecting fiber from the surface (Rux and Burnett, 1999). During an infection, the Ad fiber protein interacts with the Coxsackie-Adenovirus receptor (CAR) on the surface of the host cell (Russell, 2000). The genome of Ad comprises linear double-strand (ds) DNA with a genome length of 36kb (Shenk, 1996). The 5’ end of each DNA strand is covalently linked to a terminal protein. At each end of

the genome are inverted terminal repeats (ITRs) of about 100 nucleotides that contain the viral origin of replication. Ad genes are transcribed in the nucleus of the host cell. The genome is organized in nucleosomes similar to the host DNA, and is subjected to epigenetic changes (Giberson et al., 2011). Viral genes are expressed in a temporal fashion, and are termed early and late genes, whose expression is separated by the onset of DNA replication (Russell, 2000). The early genes are E1A, E1B, E2, E3, and E4 and two delayed early genes encoding proteins IX and IVa2. A major late transcription unit (MLTU) encodes for five families of late mRNAs. The genome is transcribed from both DNA strands; the top strand directs transcription of E1A, E1B, E3, IX, and the MLTU whereas the lower strand directs synthesis of E2, E4, and IVa2 (Figure 1) (Evans and Hearing, 2002). The early region-encoded proteins play a crucial role in regulating host cell processes to create an environment suitable for virus replication and gene expression. Viral early genes regulate cell cycle progression and apoptosis inhibition, viral and cellular gene expression, immune suppression, and degradation of DNA repair proteins (Branton, 1999).

Life cycle

The life cycle of Ad begins with the attachment of the Ad fiber protein to the CAR on the host cell followed by binding of the viral penton to host αvβ3 or αvβ5 integrins. The virion enters host cell via clathrin-mediated endocytosis (Medina-Kauwe, 2003; Meier and Greber, 2003) and the disassembly initiates en route to the nucleus. A decrease in pH during maturation of endosomes brings about conformational changes in the penton protein that facilitates virion release into the cytoplasm. The partially uncoated virion is then directed toward the nucleus by the aid of microtubules that load the genome onto the nuclear pore complex. The genome is then released into the host nucleus along with the core genome-associated proteins, including protein V, protein VII, mu, and the covalently linked terminal proteins (Brown et al., 1975; Hosokawa and Sung, 1976; Maizel et al., 1968; Prage and Pettersson, 1971; Rekosh et al., 1977; Weber et al., 1983; Greber et al., 1993). Once inside the nucleus, the host RNA polymerase (RNAP) initiates transcription of the immediate early E1a gene. The E1a protein acts as a transcriptional activator of the other early genes as well as some cellular genes. The early gene products are responsible for making the host nucleus a productive environment for virion replication. The viral genome is then amplified by Ad polymerase (Ad pol), a gene product of

Figure 1. Diagrammatic representation of Adenovirus genome and early region 4 (E4). The Ad genome is shown with 5’ terminal proteins on either side. Early region E1a, E1b, E2a and E2b, E3 and E4 are shown in blue. Late genes expressed from the major late major late transcription unit (MLTU) and are shown in red. Arrows indicate the transcription start site and the direction of transcription. E4 region encodes seven polypeptides. ORFS 1, 2, 3, 4, and 6 are expressed directly from the template while ORFS 3/4 and 6/7 are multiply spliced to express the polypeptide.

E2b, and the newly synthesized genomes serve as a template for late gene expression. Late proteins are translated from multiply processed pre-mRNAs transcribed from the MLTU and primarily consist of structural proteins. Virions are assembled and released by lysis of the host cell.

The Ad origin of replication is located within the ITRs (Evans and Hearing, 2002; Shenk, 1996). DNA replication is initiated by a protein priming mechanism where a covalent bond is formed between the α-phosphoryl group of deoxycytidine monophosphate (dCMP), which is the first nucleotide of the genomic DNA, and the β-hydroxyl group of a serine in the pre-terminal protein (pTp). This priming reaction is catalyzed by Ad DNA polymerase. Nucleotides are then added to the 3’ hydroxyl group of pTp-dCMP complex and subsequently the strand elongates. DNA synthesis proceeds via a strand displacement mechanism where the non-template strand is displaced. The displaced strand circularizes and forms a partial duplex by base pairing at the ITRs giving a characteristic “pan-handle” appearance. A second round of DNA synthesis is initiated at the double-stranded pan-handle structure, ultimately yielding two ds genomic DNAs (Challberg and Kelly, 1989) (Figure 2).

The de novo synthesis of genomic DNA marks a switch to the synthesis of late mRNA transcripts. Viral late genes are transcribed from the major late promoter (MLP) as a single long precursor mRNA. This transcript is then differentially spliced and polyadenylated to yield multiple (over 20) distinct mRNAs. These late mRNAs are grouped into five families termed L1 through L5 depending upon the site of polyandenylation (Leppard, 1998). The 5’ end of most late mRNAs originating from the MLTU contains a 200 nucleotide untranslated leader called the tripartite leader. Mature viral mRNAs are then preferentially exported to the cytoplasm with the help of E1b-55kDa and E4-34K proteins (Blackford and Grand, 2009), where Ad late mRNA is preferentially translated. Ad infections shut down aspects of host gene expression including cellular mRNA export and translation. With the accumulation of genomic DNA and late structural proteins, Ad assembly proceeds with the formation of empty capsids into which the viral DNA is packaged. The virion is released by cell lysis with the help of E3 11.6kDa adenovirus death protein (ADP) (Tollefson et al., 1996).

Adenovirus Early Region 4

Ad early region E4 is required for making the host cell a conducive environment for

Figure 2: Model of Adenovirus DNA replication. DNA replication initiates from the end of the genome and displaces the non-template strand. The displaced strand forms a pan handle by self- base pairing at inverted terminal repeats. DNA replication initiates at the handle and forms a double strand. This two-step replication process yields two semi-conserved genomic DNAs.

replication. Deletion of the entire E4 region (E4 mutant) renders the virus defective for a productive life cycle. During E4 mutant infections, early gene expression is unaffected; however the virus is defective for DNA replication as well as late gene expression (Halbert et al., 1985; Weinberg and Ketner, 1986; Bridge and Ketner, 1989; Sandler and Ketner, 1989; Huang and Hearing, 1989). In addition, E4 mutant genomes are found as concatemers of multiple genome length in infected cells (Weiden & Ginsberg, 1994). Consequently, E4 mutant infections result in severely reduced viral yields (Huang and Hearing, 1989; Evans and Hearing, 2003).

The E4 region is situated at the extreme right of the genome and the primary transcript yields at least 18 mRNAs by alternative splicing, which encode seven distinct proteins. Among these proteins, the products of E4 orf3 and E4 orf6 play a critical role in virus life cycle. A single mutation in just one of these two genes does not result in a severe phenotype, but mutations in both genes result in a phenotype similar to a complete E4 deletion suggesting that E4 orf3 and orf6 have redundant functions in promoting an efficient viral life cycle (Bridge and Ketner, 1989; Huang and Hearing, 1989).

The E4 orf3 gene expresses the E4-11kDa protein that reorganizes many cellular proteins including promyelocytic leukemia protein (PML), p53 and Mre11 into track like structures (Evans and Hearing, 2003; Stracker et al., 2002). PML is normally found in distinct nuclear structures termed PML oncogeneic domains (PODs) that are centers for modulating various cellular processes such as transcription, DNA damage and apoptosis and are targeted by multiple viruses, including Ad, to promote efficient viral replication (Borden, 2002). E4-11kDa is responsible for inactivation of DNA damage repair (DDR) responses, promoting viral mRNA splicing, and suppression of interferon proteins (reviewed in Weitzman, 2005). Recently E4- 11kDa has also been implicated in inactivating p53 function by binding to a p53 responsive promoter and forming a repressive chromatin structure (Ou et al., 2012). E4 orf6 encodes a 34kDa (E4-34kDa) protein that forms a complex with a 55kDa protein expressed from E1b (E1b- 55kDa). These viral proteins interact with cellular cullin-based E3 ubiquitin ligase complexes, and are responsible for ubiquitination and proteasome-mediated degradation of several cellular DNA damage response proteins (Harada et al., 2002; Querido et al., 2001; Baker et al., 2007; Dallaire et al., 2009; Orazio et al., 2011; Gupta et al., 2012; Schreiner et al., 2013). E4-34kDa is important for efficient DNA replication (Bridge and Ketner, 1989, 1990), stabilization of late mRNAs (Sandler and Ketner, 1989, 1991) and selection of alternative splice sites during mRNA

processing (Nordqvist et al., 1994; Ohman et al., 1993). In addition, a complex of E1b-55kDa and E4-34kDa stimulates late gene expression and export of late viral mRNAs while preventing cellular mRNAs export (Blackford and Grand, 2009).

Double Strand Break Repair (DSBR)

DSBR becomes relevant during Ad infection because the linear double stranded nature of the genome present during infection can be perceived as a DNA break by the host cell. Integrity of genomic DNA is crucial for all living organism. Eukaryotic cells are constantly challenged by endogenous or exogenous genotoxic agents that can potentially damage DNA. Double strand breaks (DSBs) are a specific type of DNA damage that are highly toxic to eukaryotic cells (Hopfner et al., 2002; Jackson, 2002; Bradbury and Jackson, 2003). DSBs are sensed by the Mre11/Rad50/Nbs1 (MRN) complex that activates double strand break repair (DSBR) kinases including ATM (ataxia telangiectasia mutated), ATR (ATM and Rad3-related) and DNA dependent protein kinase (DNAPK). These kinases phosphorylate numerous proteins that mediate cell cycle arrest, repair the damaged DNA, and trigger apoptosis if the damage is too severe to repair (Figure 3). ATM primarily responds to DSBs, while ATR responds to the presence of single strand (ss) DNA generated by various mechanisms including stalled replication forks (Shiloh, 2003). DSBs are repaired in cells by well-orchestrated actions of proteins involved in homologous recombination and non-homologous end joining (NHEJ) (Hopfner et al., 2002; Jackson, 2002). Homologous recombination repair is precise as it uses the homologous template to repair; in contrast, NHEJ is rapid, error-prone, and does not require extensive sequence homology. NHEJ is mediated by the DNAPKtogether with the Ku70/80 heterodimer, ligase IV and its cofactor XRCC4 (Takata et al., 1998).

The MRN complex proteins are some of the earliest damage response proteins that respond to DSBs (Celeste et al., 2003; Lisby et al., 2004). Once the damage is sensed, the MRN complex recruits ATM to the DNA damage site through physical interaction of ATM with Nbs1. Once at the DNA damage site, ATM is activated by auto-phosphorylation and subsequently phosphorylates numerous target proteins (Dupre et al., 2006; Bakkenist and Kastan, 2003). A hallmark of DSBR during ionizing radiation-induced DNA damage is the formation of distinct nuclear foci of proteins involved in the DSBR pathway called ionizing radiation-induced foci (IRIF) (Figure 4). Several DDR proteins including BRCA1, 53BP1, Mdc1 and Nbs1, accumulate

Figure 3: Schematic representation of double strand break repair pathways. DSBs are generated by various agents (lightning bolt) and are sensed by the MRN complex. The MRN complex relays the signal to ATM and ATR kinases. ATM activates by autophosphorylation. ATM and ATR then initiate signal cascades that phosphorylate numerous substrates responsible for cell cycle arrest including Checkpoint kinase 1 (Chk1) that will provide time to repair the broken ends by the MRN complex and if the damage is too severe then apoptosis will be triggered.

at these foci (Stucki and Jackson, 2006). Recent studies have shown that these proteins are recruited to IRIF in a temporal fashion, in which the presence of one protein depends upon the recruitment of a previous protein. DDR proteins are termed either early and include the MRN complex, Mdc1 and γH2AX, or late and includes RNF8, 53BP1 and BRCA1 depending the time of recruitment after induction of DNA damage. Proteins in the IRIF are subjected to various post translational modifications such as phosphorylation, sumolyation and ubiqitination. For example, ATM dependent phosphorylation of Mdc1 forms a docking station for RNF8, an E3 ubiquitin ligase. Once RNF8 is recruited to the IRIF, it adds lysine 63-linked polyubiquitin (K63Ub) to histones including H2A and H2AX, which in turn forms a docking site for other DDR proteins such as 53BP1 that are required for repair of damaged DNA. The MRN complex is also important for ATR activity although the mechanism is less well understood. It is possible that the MRN complex generates ssDNA during processing of broken ends during DSBR that is important for activating ATR (Myers and Cortez, 2006; Carson et al., 2009). ATR forms a complex with ATR interacting partner (ATRIP) (Zou and Elledge, 2003), which is recruited to the damage site via replication protein A (RPA) that coats the ssDNA (Bakkenist and Kastan, 2003; Bakkenist and Kastan, 2004).

DNA damage is associated with multiple histone modifications at the site of the DSB; include phosphorylation, acetylation, methylation and ubiquitination (Dinant et al., 2008). The first key step during the initiation of the repair process is loosening the histone interactions with DNA (Tsukuda et al., 2005; Shroff et al., 2004; Berkovich et al., 2007; Xu and Price, 2011) to form relaxed DNA. Consequently, chromatin is more accessible to repair proteins such as 53BP1 and BRCA1. This process is facilitated by hNuA4, a multi-subunit complex that consists of p400 (DNA dependent ATPase), Tip60 (histone acetyl transferase) and Ruvbl1 and Ruvbl2 (helicase) that is recruited to DSBs (Xu and Price, 2011). Subunits of the hNuA4 complex work together to locally relax chromatin structure, alter DNA-histone interaction, and promote ubiquitination of γH2AX by RNF8. Besides recruiting BRCA1, ubiquitination of histones by RNF8 increases accessibility of methylated histones required for loading of 53BP1 on to the chromatin at DSBs (Botuyan et al., 2006; Huyen et al., 2004).

Ad E4 mediated inactivation of DDR pathways

Ad delivers linear double-stranded genomic DNA to the nucleus. The linear Ad genome

Figure 4: Model for IRIF formation. The MRN complex senses the DNA damage and recruits and activates ATM at DSBs. Activated ATM (pATM) phosphorylates γH2AX at the vicinity of DSBs. Mdc1 binds to γH2AX and is retained at DSB. pATM phosphorylates Mdc1 that creates a docking site for RNF8. At DSBs, RNF8 add ubiquitin molecules to γH2AX that becomes a recruitment site for 53BP1 and BRCA1. Recruitment of these DDR proteins spreads to either side of DSBs resulting into the formation of IRIF.

has features similar to a DNA DSB and can potentially activate DSBR pathways (Stracker et al., 2002; Carson et al., 2003). Indeed, infection with an E4 mutant virus that fails to inactivate the DNA damage sensor Mre11 results in activation of DSBR signaling pathways mediated by both ATM and ATR kinases (Stracker et al., 2002; Carson et al., 2003). E4 mutant virus genomes are concatenated (Weiden and Ginsberg, 1994) and viral late protein accumulation is markedly reduced (Bridge and Ketner, 1990; Halbert et al., 1985; Huang and Hearing, 1989; Jayaram and Bridge, 2005). Moreover, E4 mutants also display severe DNA replication defects (Weinberg and Ketner, 1986, Halbert et al., 1985 and Jayaram and Bridge, 2005). In contrast to the E4 mutant infection, the wild type Ad5 genome is maintained as a linear monomer throughout the infection and ATM and ATR dependent signaling pathways are not activated (Stracker et al.,2002; Carson et al., 2003). The E4-11kDa and E4-34kDa proteins are both able to inactivate the MRN complex. The E4-11kDa protein sequesters the MRN complex into nuclear tracks and aggresomes and prevents its localization at the virus DNA replication centers (Stracker et al., 2002; Evans and Hearing, 2005). E4-34kDa and E1b-55kDa target Mre11 for proteosome- mediated degradation. While mislocalization and degradation of Mre11 during E4 mutant infection inactivates ATM and ATR mediated signaling pathways, mislocalization of Mre11 alone by E4-11kDa is not sufficient to inactivate ATM pathways (Stracker et al., 2002; Carson et al., 2009). In addition, E4-11kDa and E4-34kDa interfere with the activity of DNAPK (Boyer et al., 1999) and degrade ligase IV (Baker et al., 2007), thereby inhibiting DNA repair by NHEJ.

E4-mediated regulation of DSBR proteins is critical for a normal viral life cycle. The MRN complex inhibits viral DNA synthesis by an unknown mechanism (Mathew and Bridge 2007; Karen et al., 2009). Localization of the MRN complex with E4 mutant DNA replication centers correlates with a defective DNA synthesis (Mathew and Bridge, 2007). Mathew and Bridge (2007) have shown that cells that either lack Mre11 or have components of MRN complex knocked down by RNA interference rescue E4 mutant DNA replication. Furthermore, it has been demonstrated by chromatin immunoprecipitation experiments that Mre11 binds to the viral genome (Mathew and Bridge, 2008). The MRN complex may induce formation of a DDR complex that blocks the ends of the viral genome, rendering the origin of replication inaccessible to viral replication proteins (Mathew and Bridge 2007; Karen and Hearing 2011). Lakdawala et al., (Lakdawala et al., 2008) have shown that a mutation in the C terminus of Nbs1 that abrogates its interaction with ATM or Mre11 does not inhibit E4 mutant DNA replication. This

suggests that interaction of Nbs1 with ATM and the MRN complex may be important for inhibiting E4 mutant DNA replication. The MRN complex has also been shown to be important for activating ATR-dependent DDR responses during Ad infection (Carson et al., 2009). The MRN complex could inhibit E4 mutant DNA replication directly or by its ability to activate the ATM and ATR kinases. Previous studies have focused on the role of the MRN complex in E4 mutant DNA replication. The specific role(s) of the downstream ATM and ATR kinases in mediating E4 mutant phenotypes has not been well characterized.

In this research, the role of ATM and ATR in inhibiting E4 mutant growth has been characterized. In addition, a role for the ATM kinase in interfering with E4 mutant DNA replication has been identified. Furthermore, the ability of ATM kinase to interfere with Ad mutant DNA replication has been correlated with its ability to localize late DDR proteins including RNF8 and 53BP1 to viral replication centers.

The Kinase Activity of Ataxia-Telangiectasia Mutated Interferes with Adenovirus E4

Mutant DNA Replication

Dipendra Gautama and Eileen Bridgeab#

aDepartment of Microbiology

32 Pearson Hall

Miami University

Oxford OH 45056 bCell Molecular and Structural Biology Program

#Corresponding author

Tel: 513-529-7264

Fax: 513-529-2431

Email: [email protected]

Running Title: ATM inhibits Ad E4- DNA replication

Manuscript submitted-

Gautam, D., & Bridge, E. (2013). The kinase activity of ATM interferes with adenovirus E4 mutant DNA replication. Journal of virology, 87(15), 8687-96.

Abstract

Adenovirus (Ad) mutants that lack early region 4 (E4) are unable to produce the early regulatory proteins that normally inactivate the Mre11/Rad50/Nbs1 (MRN) sensor complex, which is a critical component for the ability of cells to respond to DNA damage. E4 mutant infection therefore activates a DNA damage response, which in turn interferes with a productive viral infection. MRN complex proteins localize to viral DNA replication centers in E4 mutant infected cells, and this complex is critical for activating the kinases ataxia-telangiectasia mutated (ATM) and ATM and Rad3-related (ATR), which phosphorylate numerous substrates important for DNA repair, cell cycle checkpoint activation, and apoptosis. E4 mutant growth defects are substantially rescued in cells lacking an intact MRN complex. We have assessed the role of the downstream ATM and ATR kinases in several MRN-dependent E4 mutant phenotypes. We did not identify a role for either ATM or ATR in "repair" of E4 mutant genomes to form concatemers. ATR was also not observed to contribute to E4 mutant defects in late protein production. In contrast, the kinase activity of ATM was important for preventing efficient E4 mutant DNA replication and late gene expression. Our results suggest that the MRN complex interferes with E4 mutant DNA replication at least in part, through its ability to activate ATM.

Keywords: Adenovirus, E4 ORF 3, E4 11kDa, E4 ORF 6, E4 34kDa, Mre11, MRN complex, ATM, ATR, DNA damage response, double strand break repair.

Introduction

Adenovirus (Ad) infection delivers a linear double stranded DNA genome to the nucleus of infected cells. This exogenous DNA has the potential to activate cellular DNA damage responses (DDRs) (reviewed in Weitzman and Ornelles, 2005), which can impede a productive viral infection (Evans and Hearing, 2005; Lakdawala et al., 2008; Mathew and Bridge, 2007; Mathew and Bridge, 2008). Consequently, Ad produces early gene products that interfere with the activity of several cellular DDR proteins. Proteins from early region 1b (E1b-55kDa) and E4 orf6 (E4-34kDa) form a complex that redirects a cellular CUL5-containing E3 ubiquitin ligase to target specific proteins for ubiquitination and proteasome mediated degradation (Blanchette et al., 2004; Cheng et al., 2007; Querido et al., 2001). Mre11 of the MRN complex (Carson et al., 2003; Stracker et al., 2002), the nonhomologous end-joining (NHEJ) enzyme DNA Ligase IV (Baker et al., 2007), and the tumor suppressor p53 (Querido et al., 2001), are examples of DDR proteins targeted for degradation by this complex. E4 orf3 produces an 11kDa protein (E4- 11kDa) that redistributes Mre11 to nuclear filaments (Evans and Hearing, 2005; Karen et al., 2009; Stracker et al., 2002), and both E4-34kDa and E4-11kDa bind and inactivate DNA- dependent protein kinase (DNAPK), which is an essential kinase for NHEJ (Boyer et al., 1999). Mutants deleted for the E4 transcription unit are unable to produce these viral proteins and consequently activate a cellular DDR in infected cells (Carson et al., 2002). This response includes activation of the kinases ATM and ATR, which phosphorylate numerous downstream substrates important for repair, cell cycle arrest and apoptosis (reviewed in Harper and Elledge, 2007). Ad infection also induces the reorganization of DDR proteins to nuclear foci that can be viewed by immunofluorescence staining. Mre11 and mediator of DNA damage checkpoint protein 1 (Mdc1) are redistributed to early foci that appear prior to the onset of viral DNA replication (Mathew and Bridge, 2007; Mathew and Bridge, 2008). Phosphorylated ATM (pATM) is found in foci that contain input E4 mutant DNA (Karen and Hearing, 2011). In E4 mutant infected cells, Mre11 and pATM are present in viral replication centers that contain the viral 72kDa DNA binding protein produced from E2 (E2-72kDa) (Carson et al., 2003; Stracker et al., 2002). ATR is also found to localize to viral DNA replication centers in both Ad5 and E4 mutant infections (Carson et al., 2003).

Activation of cellular DDRs severely reduces productive growth of E4 mutants in cells. Viral genomes are concatenated by DNAPK-mediated NHEJ in E4 mutant infections (Boyer et

al., 1999), potentially affecting virus yields because concatemers are too large to be packaged in assembling virions. However, cells that lack DNAPK and fail to concatenate viral genomes still do not rescue E4 mutant defects in viral DNA replication following low multiplicity infections (Jayaram and Bridge, 2005; Mathew and Bridge, 2007), indicating that genome concatenation does not account for all E4 mutant growth defects. Work from several groups has shown that the MRN complex interferes with E4 mutant growth by inhibiting viral DNA replication (Evans and Hearing, 2005; Lakdawala et al., 2008; Mathew and Bridge, 2007; Mathew and Bridge, 2008). Cells lacking either Mre11 or Nbs1 support efficient E4 mutant replication and growth (Evans and Hearing, 2005; Mathew and Bridge, 2008). Knockdown of Mre11, Rad50, or Nbs1 by RNA interference also dramatically rescues the DNA replication phenotype of an E4 mutant in HeLa cells (Mathew and Bridge, 2007). The mechanism used by the MRN complex to interfere with E4 mutant DNA replication is currently being investigated. Recent results indicate that the nuclease activity of Mre11 is not critical for the DNA replication defect, suggesting that nuclease mediated destruction of the viral origin of replication is unlikely to be the primary mechanism involved (Karen and Hearing, 2011). However, Nbs1 dependent binding of Mre11 to viral DNA is important for inhibiting E4 mutant DNA replication (Mathew and Bridge, 2008). These observations raise the possibility that the MRN complex may be able to inhibit E4 mutant DNA replication by physically interacting with the genome and perhaps preventing viral DNA replication proteins from being able to access the origin of replication located at the termini of the linear DNA genome (Karen and Hearing, 2011, Mathew and Bridge, 2008). The MRN complex acts as a sensor to detect DNA damage, but it is also critical for activating signaling cascades mediated by the ATM and ATR kinases in response to DNA damage. The MRN complex could interfere with E4 mutant DNA replication as a consequence of either its DNA damage sensing activity or its ability to stimulate DDR kinases. We have sought to more specifically address the role of the ATM and ATR kinases in limiting a productive E4 mutant infection. We did not identify a role for either ATM or ATR in E4 mutant genome concatenation; however the kinase activity of ATM is important for inhibiting E4 mutant DNA replication.

Materials and methods

Cells and viruses. HeLa and E4 mutant-complementing W162 cells (Weinberg and Ketner, 1983) were grown in Dulbecco's modified Eagle medium (DMEM) (Fisher) supplemented with 10% fetal bovine serum (FBS), 10 U/ml penicillin, and 10 μg/ml streptomycin. GM16666 (referred to here as ATM−) and GM16667 (referred to here as ATM+) (GM16667 cells are derived from GM16666 cells and have been complemented with a construct producing ATM) were purchased from the Coriell cell repository and maintained in DMEM supplemented with 10% FBS, 10 U/ml penicillin, 10 μg/ml streptomycin, and 100 μg/ml Hygromycin B (Invitrogen). NBS-ILB1 cells stably transduced with pLXIN retroviral vector alone (referred to here as Nbs1− cells) and NBS-ILB1 cells that were stably transduced with the pLXIN retroviral vector expressing the wild-type NBS1 protein (referred to here as Nbs1+ cells) were obtained from Pat Concannon (Cerosaletti et al., 2006) and maintained in DMEM supplemented with 10% FBS, 10 U/ml penicillin, 10 μg/ml streptomycin, and 500 μg/ml G418 (Invitrogen). Wild-type Ad5 and E4 mutant H5dl1007 (E4-) (Bridge and Ketner, 1989) were propagated on W162 cells for stock preparation. Titers were determined on W162 cells as described previously (Phillipson, 1961) and were expressed as fluorescence-forming units/ml (FFU/ml).

Immunofluorescence microscopy. Cells were grown on 12-mm-diameter cover glasses placed in culture dishes and were either left uninfected or infected with E4- or Ad5 and fixed for immunofluorescence staining at the desired time point (Mathew and Bridge, 2007). Briefly, cells were rinsed twice with phosphate-buffered saline (PBS) (0.058 M Na2HPO4, 0.017 M

NaH2HPO4, 0.069 M NaCl), prefixed with 1% paraformaldehyde–PBS for 3 min, extracted with 1% Triton X-100-PBS for 15 min, fixed with 4% paraformaldehyde in PBS for 15 min followed by three 5-min washes with PBS. Cover glasses were then stored at 4°C in PBS until use. For immunofluorescence staining, the cover glasses were rinsed in TBST (Tris-buffered saline Tween; 100 mM Tris-Cl [pH 7.5], 150 mM NaCl, 0.1% Tween 20) followed by incubation with 15 μl of blocking reagent (100 mM Tris [pH 7.5], 150 mM NaCl, 0.5% blocking reagent powder; GE Amersham) for 45 min. Cover glasses were then rinsed with TBST, and 15 μl of primary antibodies diluted in blocking reagent was added for 45 min to 1 h. Cover glasses were then washed 3 times for 10 min each time with TBS (Tris-buffered saline; 100 mM Tris-Cl [pH 7.5], 150 mM NaCl) and rinsed once with TBST. Finally, 10 μl of secondary antibodies diluted in

blocking reagent was added to each cover glass. After 45 min of incubation, the cover glasses were washed 3 times for 10 min each time with TBS and mounted on a glass slide with 4 μl of Vectashield (Vector Laboratories) as the mounting medium.

The following primary antibodies were used for immunofluorescence microscopy at the dilutions specified: goat polyclonal anti-Mre11 (Santa Cruz) (1:100), mouse monoclonal B6-8 anti-72K (a gift from A. Levine) (1:100), rabbit polyclonal anti-72K (a gift from T. Linné) (1:1,500), mouse monoclonal anti-phospho-ATM (p-S1981) (Abcam) (1:20), rat monoclonal anti-bromodeoxyuridine (BrdU) (Abcam) (1:100), and rabbit monoclonal anti-ATM (Abcam) (1:30). The following secondary antibodies from Invitrogen were used for immunofluorescence at the dilutions specified: donkey anti-rabbit Alexa Fluor 594 IgG (1:250), donkey anti-mouse Alexa Fluor 488 IgG (1:250), donkey anti-goat Alexa Fluor 594 IgG (1:250), and goat anti-rat fluorescein isothiocyanate (FITC) from Southern Biotechnology (1:70).

Immunostained cells were scored for specific phenotypes by observation with a Nikon Eclipse E-400 epifluorescence microscope using a 100× oil immersion objective. Confocal images were captured with a confocal laser scanning microscope (Olympus FV500) using a 100× oil immersion objective and Fluoview software. Sequential scans of the Alexa 488 and Alexa 594 channels were performed to prevent bleed through between the channels. Composite images were assembled using Adobe Photoshop CS5 software.

Western blot analysis. Cells cultured on 35-mm-diameter plates were either left uninfected or infected with Ad5 or E4-. At the desired time, cells were washed twice with ice- cold PBS and scraped into a microcentrifuge tube and pelleted by centrifugation. The supernatant was aspirated, and the cell pellet was suspended in lysis buffer (Cell Signaling) (20 mM Tris-

HCl [pH 7.5], 150 mM NaCl, 1 mM Na2EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium pyrophosphate, 1 mM β-glycerolphosphate, 1 mM Na3VO4, 1 μg/ml leupeptin). The extracts were then sonicated and centrifuged at top speed in a microcentrifuge for 1 min to remove insoluble material. The protein concentration of the extracts was quantified using the Bradford assay with Coomassie Plus (Pierce) according to the manufacturer's instructions and analyzed with a Nano-Drop ND-1000 spectrophotometer. Samples containing equal amounts of protein were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) using 4% stacking and 5 to 10% resolving polyacrylamide gels depending upon the size of the protein. The separated

proteins were then transferred to an enhanced chemiluminescence (ECL) nitrocellulose membrane (Amersham) overnight at 20 V and 4°C. The membranes were blocked with nonfat dry milk in TBST for 2 h at room temperature and incubated with specific primary antibody in TBST with 5% bovine serum albumin with shaking overnight at 4°C. Primary antibodies and the dilutions used for Western blotting were mouse monoclonal anti-phospho-ATM (p-S1981) (Abcam) (1:1,000), rabbit polyclonal anti-penton (a gift from U. Pettersson) (1:1,000), rabbit monoclonal anti-ATM (Abcam) (1:3,000), rabbit polyclonal anti-ATM (Novus) (1:1,000), and goat polyclonal anti-ATR (Santa Cruz) (1:200). The membranes were then subjected to multiple washes with TBST, after which they were incubated with the appropriate secondary antibody: either horseradish peroxidase-conjugated anti-rabbit or anti-mouse (Amersham) (1:1,500) or alkaline phosphatase-conjugated anti-rabbit (Santa Cruz) (1:1,500). Secondary antibodies were diluted in 5% nonfat dry milk for 1 h at room temperature with shaking. After multiple washes, the membranes were incubated with ECL reagent (Amersham) to generate chemiluminescence signals, which were subsequently captured on ECL hyperfilm (Amersham). For protein quantification, alkaline phosphatase-conjugated secondary antibodies were used with enhanced chemifluorescence (ECF) substrate (Amersham) and analyzed with a STORM 860 phosphorimager (Molecular Dynamics). ImageQuant 5.2 (Molecular Dynamics) software was used to quantify the amount of protein.

Viral DNA analysis. Viral DNA was prepared as previously described (Prakash et al., 2012). Briefly, cells grown on 35-mm-diameter plates were either left uninfected or infected with E4- or Ad5. Cells were lysed at 24 h postinfection (hpi) in proteinase K buffer (0.05 M Tris [pH 7.8], 0.0025 M EDTA, 0.25% SDS) and incubated with 0.4 mg/ml proteinase K for 4 h at 37°C, followed by phenol:chloroform extraction and ethanol-salt precipitation of nucleic acids. Total nucleic acid was subjected to RNase A (Fisher) digestion in Buffer H (Promega) followed by phenol:chloroform extraction and ethanol-salt precipitation. Dot blotting was performed using the RNA-free DNA. Samples containing 1.0 to 0.1 μg of DNA were denatured by boiling at 100°C for 10 min and immediately cooled on ice. Samples were adjusted to 6× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate). The samples were blotted on Hybond-N nylon membrane (Amersham) that was presoaked with 6× SSC using a dot blot manifold and applying vacuum pressure. The DNA was cross-linked to the membrane using a UV transilluminator. The membranes were then probed with Ad-specific radiolabeled probes generated using a Megaprime

DNA labeling system (GE Healthcare) following the manufacturer's instructions. The hybridization reaction was performed with 5 × 106 cpm/ml probe at 65°C overnight. The amount of labeled DNA bound was detected using a STORM 860 phosphorimager (Molecular Dynamics) and quantified using ImageQuant 5.2 (Molecular Dynamics) software.

RNAi analysis. Cells were grown on 35-mm-diameter tissue culture dishes to 50% confluence and then transfected with ON-Targetplus small interfering RNA (siRNA) for target proteins ATM and ATR (Dharmacon Technologies) or a nontargeting siRNA (negative control) at 25 pmol/plate with 2 μl of lipofection reagent according to the manufacturer's instructions. Knockdown of target protein expression was analyzed at 72 and 96 h posttransfection (hpt) by Western blotting. Cells transfected with the respective siRNAs were infected with viruses at 72 hpt, and the samples were prepared for analysis at 24 h postinfection (hpi).

Pulsed-field gel electrophoresis (PFGE). Total cellular DNA was prepared as described previously (Boyer et al., 1999). Briefly, cells were grown in 35-mm-diameter plates and were either left uninfected or infected with E4- and Ad5 viruses and at 24 hpi were treated with trypsin to remove them from the plate and embedded in low-melting-point agarose plugs. The agarose-embedded cells were subjected to proteinase K digestion overnight at 50°C. The plugs were then incubated with 50 mM EDTA for 8 h and loaded on 1.2% agarose gels for electrophoresis using a CHEF-DR II pulsed-field electrophoresis system (Bio-Rad) in hexagonal mode for 10 h at 6 V/cm with a switch time of 15 s. Gels were stained with ethidium bromide and photographed using a UV photodocumentation system to visualize viral DNA concatemers.

DSBR kinase inhibitors. Caffeine (Sigma) was dissolved in water to make a stock solution of 100 mM and added 2 hpi to a final concentration of 10 mM until the desired time. KU60019 (Tocris) was dissolved in dimethyl sulfoxide (DMSO) to make a 10 mM stock solution and added to the culture at 2 hpi to a final concentration of 5 μM for the desired time.

BrdU labeling. Bromodeoxyuridine (BrdU) labeling was done as described previously (Pombo et al., 1994). Briefly, cells were grown on cover glasses and infected with E4- or Ad5. BrdU was dissolved in medium to a final concentration of 150 mM and added to the cells at the desired time for 1 h of incubation. Cells were then fixed with paraformaldehyde as described above. Prior to immunostaining, the cells were treated with 4 N HCl for 30 min to denature the DNA, washed with PBS twice for 5 min each time, neutralized with 0.1 M sodium borate for 5

min, and washed twice for 5 min each time with PBS.

Results

ATM and ATR are not required for E4 mutant genome concatenation. E4 mutant infection results in genome concatenation by NHEJ. We have previously found that caffeine, which inhibits both ATM and ATR, also inhibited E4 mutant genome concatenation (Jayaram and Bridge, 2011), suggesting a possible role for these kinases. We performed siRNA knockdown experiments to assess the role of ATM and ATR in E4 mutant genome concatenation. HeLa cells were transfected with siRNAs targeted against ATM (siATM) or ATR (siATR) and cultured for 72 or 96 h. The level of ATM and ATR proteins was substantially reduced at both 72 and 96 hpt compared to the level seen with untransfected or control siRNA- treated (siC) cells (Figure 5A). To determine the effect of ATM or ATR knockdown on E4 mutant genome concatenation, HeLa cells were transfected with siATM, siATR, or siC for 72 h and subsequently left uninfected or infected with Ad5 or E4- at a multiplicity of infection (MOI) of 30 FFU/cell. Samples were prepared at 24 hpi for PFGE to measure genome concatenation. Ad5 infection showed no genome concatenation under either set of treatment conditions as expected since Ad5 prevents activation of DNA repair pathways. Concatemers of multiple genome lengths were observed following E4- infection of siATM-, siATR-, or siC-transfected cells (Figure 5B). These results indicate that neither ATM nor ATR is individually critical for inducing viral genome concatenation and confirm previous observations (Stracker et al., 2003). We next wished to address the possibility that ATM and ATR might provide redundant functions in promoting E4 mutant genome concatenation. We performed siRNA knockdown of ATR in ATM− cells to assess the role of the combined lack of ATM and ATR on genome concatenation. ATM− cells are derived from a patient with ataxia telangiectasia (A-T). They carry a mutation in the ATM gene and fail to produce the ATM protein (Ziv et al., 1997). After siATR transfection, ATM− cells showed a marked reduction of ATR compared to siC-treated cells at 72 and 96 hpt (Figure 5C, left panels). ATM− cells transfected with siC or siATR for 72 h were either left uninfected or infected with E4- or Ad5 and assayed for genome concatenation by PFGE at 24 hpi (Figure 5C, right panel). E4 mutant genome concatemers were still observed in both siC- and siATR-treated cells. Similar results were seen at 32 hpi (data not shown). Taken together, these results indicate that ATM and ATR are not required either individually or together for E4 mutant genome concatenation.

Figure 5. ATM and ATR are not required for genome concatenation. ATM and ATR are not required for genome concatenation. HeLa cells were either untreated (UT), transfected with control siRNAs (siC), or specific siRNAs targeting either ATM (siATM) or ATR (siATR) mRNAs. (A) Transfected cells were processed for Western blotting to determine the levels of ATM or ATR at 72 and 96 hpt, or analyzed for the presence of β-actin as a control. (B) Cells transfected with the indicated siRNAs were either left uninfected (UI) or infected for 24 h with Ad5 or E4- at 30 FFU/cell from 72–96 hpt, and total DNA was analyzed by PFGE to measure the level of viral DNA concatemers. (C) siRNA knockdown of ATR was performed in ATM− cells. ATM− cells were transfected with control or ATR specific siRNA and the levels of ATR and β-actin proteins were determined by Western blotting (left panel). Total DNA from transfected cells infected with Ad5 or E4- was analyzed by PFGE to measure the level of viral DNA concatemers (right panel) as described for B.

ATM inhibits E4 mutant late gene expression and DNA replication. E4 mutant phenotypes include defects in late gene expression and DNA replication, which are more severe at low MOI (Bridge and Ketner, 1989; Halbert et al., 1985; Weinberg and Ketner, 1986). The MRN complex is an important mediator of E4 mutant DNA replication and late gene expression phenotypes (Evans and Hearing, 2005; Lakdawala et al., 2008; Mathew and Bridge, 2007; Mathew and Bridge, 2008). MRN could regulate E4 mutant phenotypes directly, or by its activation of the ATM and/or ATR kinases. We performed siRNA knockdowns of ATM and ATR to assess a possible role for these downstream kinases. HeLa cells were transfected with siATR or siATM for 72 h and then either uninfected or infected with Ad5 or E4- at an MOI of 30 FFU/cell. At 24 hpi the cells were harvested and total cell lysates were analyzed by western blotting using antiserum that detects the viral late protein penton, and the results are presented in Figure 6. In control siRNA treated cells E4- showed a substantial defect in penton expression compared to Ad5, as expected. Transfection with siATR had little effect on penton levels in either Ad5 or E4- infected cells, suggesting that ATR is not involved in this phenotype. In contrast, siATM transfected HeLa cells showed a 3-fold increase in E4- penton levels as compared with siC transfected controls, suggesting that lack of ATM significantly rescues E4- late gene expression. Similar results were seen using an antibody against the viral late protein hexon (data not shown). Taken together these results identify a role for ATM, but not ATR, in interfering with E4 mutant late gene expression.

We performed infections in ATM- cells, and in isogenic ATM+ cells complemented with a copy of the ATM gene, to further assess the role of the ATM kinase in E4 mutant phenotypes. ATM+ and ATM- cells were infected with Ad5 or E4- at an MOI of 30 FFU/cell. Whole cell lysates were prepared at 24 hpi and western blot assays were performed using antiserum that detects the viral late protein penton. Late proteins were quantified by phosphorimaging, and the results are presented in Figure 7A. E4- penton levels were about 3-fold higher in ATM− cells than they were in ATM+ cells. The results seen with ATM− cells are similar to the results we obtained via siRNA-mediated knockdown of ATM in HeLa cells (Figure 6) and confirm that the presence of ATM interferes with E4 mutant late gene expression.

Ad DNA replication defines the onset of the late phase and is required for the switch from early to late gene expression (Thomas and Mathews, 1980). We next measured levels of viral DNA in Ad5 and E4- infected ATM+ and ATM- cells to determine if DNA replication was

Figure 6. ATM interferes with E4 mutant late protein production while ATR does not. HeLa cells were either transfected with control siRNAs (siC), or specific siRNAs targeting either ATM (siATM) or ATR (siATR) mRNAs. Cells were then either left uninfected (UI) or infected with Ad5 or E4- as described for Figure 5, and at 24 hpi analyzed for the presence of the late protein penton by Western blotting. Representative Western blots of penton levels achieved in cells treated with siATM (A) or siATR (B) are shown in the top panels. Levels of the control protein β-actin are also shown. E4- penton protein levels were quantified by phosphorimaging analysis of Western blots from 3 independent experiments and expressed as the fraction of the level achieved by Ad5, which was set at 1 (A and B, bottom graphs). Error bars show the standard error of the mean. A one-tailed Student'st test analysis was performed on the data used to generate the graphs. Statistically significant (P < 0.05) differences between the columns are indicated. N.S. = not statistically significantly different.

Figure 7. ATM interferes with E4 mutant DNA replication. ATM− and isogenic ATM+ cells complemented for ATM, were infected with Ad5 or E4- at an MOI of 30 FFU/cell unless otherwise indicated. At 24 hpi cells were harvested and processed for Western blot analysis to study viral late protein levels, or total DNA was prepared and analyzed by dot blotting to quantify viral DNA levels. (A) Top panel shows a representative Western blot using an antibody that detects the viral late protein penton. Levels of the control protein β-actin are also shown. E4- penton protein levels were quantified by phosphorimaging analysis of Western blots from 3 independent experiments and expressed as the fraction of the level achieved by Ad5, which was set at 1 (bottom graph). Error bars show the standard error of the mean. Statistically significant differences between the columns are indicated, as determined by one-tailed Student's t test. (B) Graphs showing the results of dot blot quantitation of viral DNA levels in experiments performed at 30 (top graph) and 3 (bottom graph) FFU/cell. E4- DNA levels were quantified from 3 independent experiments and expressed as the fraction of the level achieved by Ad5, which was set at 1. Error bars show the standard error of the mean. Statistically significant differences between the columns are indicated, as determined by one-tailed Student's t test.

impacted by the presence or absence of ATM. The E4 mutant demonstrated a 3-fold increase in DNA in ATM- compared to ATM+ cells (Figure 7B top graph). These data suggest that the presence of ATM inhibits E4 mutant DNA replication. Furthermore, the 3-fold increase in E4 mutant DNA levels in ATM- cells (Figure 7B top graph) is sufficient to account for the 3-fold increase in E4 mutant late protein levels observed in ATM- cells relative to ATM+ cells (Figure 7A bottom graph), suggesting that the primary effect of ATM is inhibition of E4 mutant DNA replication.

The DNA replication defect of E4 mutants is more severe at low MOI (Halbert et al., 1985; Weinberg and Ketner, 1986). We further assessed the ability of ATM- cells to support E4 mutant DNA replication by assessing viral DNA levels following lower MOI infections. ATM- and ATM+ cells were uninfected or infected with Ad5 or E4- at an MOI of 3 FFU/cell. Total DNA was prepared at 24 hpi and analyzed by dot blotting and the results are shown in Figure 7B (bottom graph). The ratio of E4- to Ad5 DNA levels increased about 5-fold in ATM- relative to ATM+ cells when infections were done at the lower MOI. E4- DNA levels were only about 2- fold reduced relative to Ad5 in ATM- cells indicating that the DNA replication defect of the mutant was substantially rescued in the absence of ATM.

The kinase activity of ATM is important for inhibiting E4 mutant DNA replication. The kinase activity of ATM is critical for generating the ATM-mediated signaling cascade and can be inhibited by caffeine or the more specific ATM inhibitor KU60019 (Golding et al., 2009). We performed experiments in the presence and absence of these drugs to assess the importance of ATM kinase activity for inhibiting E4 mutant DNA replication. HeLa cells were infected with Ad5 or E4- at an MOI of 3 FFU/cell, and at 2 hpi cells were either left untreated or treated with medium containing 10 mM caffeine or 5 μM KU60019. Cells were harvested at 24 hpi, and dot blotting was performed to measure viral DNA levels. The results are presented in Figure 8. Caffeine and KU60019 treatment of HeLa cells increased E4- DNA levels by 10-fold and 30- fold, respectively, compared to the levels seen with control cells that did not receive the inhibitor. The increased magnitude of the effect of KU60019 treatment likely reflects the observation that in these experiments, E4- DNA levels were more severely reduced relative to Ad5 DNA levels in untreated cells than was the case with the untreated controls in the caffeine experiments. Variation in the DNA replication defect of E4 mutants has been reported previously (Bridge and Ketner, 1989). In the presence of both inhibitors, E4 mutant DNA levels rose to

Figure 8. The kinase activity of ATM is important for inhibiting E4 mutant DNA replication. HeLa cells were infected with Ad5 or E4- at an MOI of 3 FFU/cell and subsequently incubated with the ATM kinase inhibitors caffeine (A) or KU60019 (B) from 2–24 hpi. Control infected cells were incubated in medium without the inhibitors but containing the same amount of DMSO solvent used for the drug treatment where indicated. Total DNA samples from 3 independent experiments were analyzed by dot blotting to measure viral DNA levels. E4- DNA levels were expressed as the fraction of the level achieved by Ad5, which was set at 1. Error bars show the standard error of the mean. Statistically significant differences between the columns are indicated, as determined by one-tailed Student's t test.

about one-third of Ad5 levels by 24 hpi. By 32 hpi, E4- DNA levels were very close to Ad5 levels in both the caffeine- and KU60019-treated infected cells (data not shown). Similar increases in E4- DNA levels were also observed in infected cells treated with the ATM inhibitor KU55933 (data not shown). When we used KU60019 to treat ATM− cells infected with E4- and Ad5, we saw very little effect on E4- DNA levels, suggesting that the observed increases in E4- DNA levels in KU60019-treated HeLa cells depend on the presence of ATM (data not shown). Taken together, these results indicate that the kinase activity of ATM is important for ATM- mediated inhibition of E4 mutant DNA replication.

In E4 mutant infections, pATM localizes at viral replication centers that contain the viral DNA binding protein E2-72kDa (Carson et al., 2003) and to foci that colocalize with input viral genomes (Karen and Hearing, 2011). We performed immunofluorescence experiments with pATM antibody to determine if its localization to either of these sites was correlated with inhibition of DNA replication. HeLa cells were infected with Ad5 or E4- at an MOI of 3 FFU/cell, fixed at 5, 7, 12, and 17 hpi, and immunostained with antibodies against pATM or E2- 72kDa. Immunofluorescence micrographs of the staining patterns are presented in Figure 9. At 5 hpi, focal accumulation of pATM was observed in cells infected with both Ad5 and E4- (Figure 9A). These cells were not yet expressing significant levels of viral early proteins as evidenced by the lack of detectable E2-72kDa staining. Cells with pATM foci were equally prevalent at 5 hpi with either Ad5 or E4- and were about six times more prevalent than the background level seen in the uninfected cell population (Figure 9B). At 7 hpi with Ad5, we detected pATM in prominent foci in cells that were not yet expressing E2-72kDa (Figure 9C, panels a to c) but pATM staining was either not detectable or much weaker in cells that had begun to express E2- 72kDa and sometimes contained small E2-72kDa foci (Figure 9C, panels d to f). This suggests that ATM activation is not sustained in Ad5 infections. At 7 hpi with E4-, pATM was still detected in foci in cells with primarily diffusely localized E2-72kDa (Figure 9C, panels g to i). In cells that contained early replication centers, pATM and E2-72kDa were nearly always colocalized or closely juxtaposed (Figure 9C, panels j to l). At 12 hpi, we could detect larger, more developed DNA replication centers in Ad5 infections. In these cells, pATM staining was usually not detected (Figure 9C, panels m to o), although some cells contained residual pATM

Figure 9. Time course showing pATM localization and BrdU incorporation relative to DNA replication centers in Ad5 and E4- infections. HeLa cells were uninfected (UI) or infected with Ad5 or E4- at 3 FFU/cell for the times indicated and then fixed and immunostained with antibodies against pATM (phosphoepitope S1981), BrdU, or the viral DNA binding protein E2- 72kDa (72K) to detect viral DNA replication centers. (A) Immunofluorescence confocal micrographs showing pATM (green) and E2-72kDa (red) staining, or phase contrast images of uninfected cells or cells infected with Ad5 and E4- at 5 hpi. (B) Uninfected cells and cells infected with Ad5 or E4- were blindly scored for the presence of pATM foci at 5 hpi, and the results presented as a graph showing the percentage of cells in the culture with pATM foci. (C) Immunofluorescence confocal micrographs showing pATM (green), E2-72kDa (red), and merged staining patterns observed at 7, 12 and 17 hpi with Ad5 or E4-. (D) Ad5 or E4- infected HeLa cells were incubated with 150 mM BrdU for 1 h at either 12 or 17 hpi. Immunofluorescence confocal micrographs showing BrdU (green), E2-72kDa (red), and merged staining patterns are shown. Cells with early-, intermediate-, or late-stage replication centers were identified based on the size and shape of the E2-72kDa foci. Similarly sized replication centers in Ad5 and E4- infections are marked with white arrowheads to facilitate comparison of BrdU incorporation at these centers. Cells with intermediate- and late-stage DNA replication centers were the most prevalent phenotype in Ad5 infections while cells with early stage replication foci were the most prevalent phenotype in E4- infections at these time points. Cells marked with an asterix are examples of cells with widespread incorporation of BrdU in cellular DNA.

staining that localized to the periphery of the E2-72kDa centers (data not shown). At 12 hpi with E4-, E2-72kDa foci showed little increase in size compared to 7 hpi and were again nearly always observed colocalized or tightly juxtaposed to a pATM focus (Figure 9C, panels p to r). At 17 hpi, Ad5 replication centers were larger and in most of the infected cells very little pATM was present (Figure 9C, panels s to u). At 17 hpi with E4-, E2-72kDa foci had still not increased much in size and remained strongly colocalized with pATM (Figure 9C, panels v to x).

Since E4 mutant E2-72kDa foci did not increase much in size during the time course, we were interested in determining if DNA synthesis was still occurring in those replication foci that had developed in the presence of pATM. We addressed this by comparing BrdU incorporation in Ad5- and E4 mutant-infected cells. HeLa cells infected with Ad5 or E4- were incubated with 150 mM BrdU for 1 h starting at 12 or 17 hpi. The cells were then fixed, denatured, and immunostained with antibodies detecting BrdU or E2-72kDa, and the results are presented in Figure 9D. These experiments were complicated by strong labeling of cellular DNA in S- phase cells (see Figure 9D cells marked with an asterisk), which made it difficult to determine the level of BrdU incorporation at E2-72kDa foci in these cells. We therefore focused on cells that had E2-72kDa foci but lacked widespread BrdU incorporation and were likely to be cells that had initiated viral DNA synthesis outside the S phase. Ad5-infected cells with early E2- 72kDa foci showed detectable BrdU incorporation at the foci (Figure 9D, panels a to c). Ad5- infected cells with intermediate-stage (Figure 9D, panels g to i and p to r) and late-stage (Figure 9D, panels m to o and v to x) replication centers labeled strongly with BrdU, indicating that they are very active for DNA synthesis. Cells infected with E4- also showed detectable BrdU incorporation at many of the early (Figure 9D, panels d to f) and intermediate (Figure 9D, panels j to l and s to u) E2-72kDa foci, indicating that DNA synthesis was not completely blocked at E4- replication centers. However, we did notice that incorporation of BrdU at E4 mutant E2- 72kDa foci was often reduced compared to similarly sized Ad5 E2-72kDa foci (Figure 9D; compare Ad5 and E4- DNA replication foci marked with arrowheads).

We also studied the effect of inhibiting ATM kinase activity with KU60019 on the development of E2-72kDa containing replication centers in time course experiments. Early replication foci were detected in relatively few of the 7 hpi cells (1% or less), but they initiated with roughly similar frequencies in Ad5 and E4- infections in the absence of KU60019 treatment (Figure 10A). KU60019 treatment significantly reduced the level of pATM associated with E4-

Figure 10. ATM inhibition facilitates progression of E4- DNA replication centers from small early replication foci to larger intermediate- and late-stage replication centers. HeLa cells were infected with Ad5 or E4- and then either incubated with medium containing DMSO only, or 5 μM KU60019 in DMSO to inhibit ATM kinase activity from 2 hpi until the times indicated. (A) The percentage of cells infected with Ad5 and E4- with E2-72kDa (72K) replication foci was scored at 7 hpi. (B) The top panel shows immunofluorescence micrographs of pATM and E2-72kDa staining at 17 hpi with Ad5 and E4- incubated in the presence or absence of KU60019. The bottom panel shows Western blot analysis of extracts prepared from uninfected (UI) cells or from cells infected with Ad5 or E4- at 17 hpi that were or were not treated with KU60019, and probed with antibodies against pATM, ATM, or β-actin. (C) The percentage of cells containing small (black bars) or large (gray bars) E2-72kDa containing DNA replication centers was determined in Ad5 and E4- infections that were and were not incubated with KU60019 from 2 hpi until the times indicated.

DNA replication centers, which were dramatically increased in size (Figure 10B, top panel). KU60019 treatment also significantly reduced the level of pATM detected in cells infected with E4- by Western blotting (Figure 10B, bottom panel). A prominent pATM band was detected in extracts from untreated cells infected with E4-. This band was absent in extracts prepared from KU60019-treated E4- infections and from Ad5-infected and uninfected cells regardless of KU60019 treatment. Cells infected with Ad5 and E4- were then scored for the presence of small early E2-72kDa foci and large replication centers in cultures that were either left untreated or treated with KU60019 from 2 to 12 or 2 to 17 hpi. The results are presented in Figure 10C. At 12 hpi in the absence of inhibitor, about 50% of cells infected with both Ad5 and E4- contained focal concentrations of E2-72kDa. However, in E4- infections most of the E2-72kDa was in small early replication foci, while in Ad5 infections the majority of the cells contained larger intermediate- and late-stage replication centers. This indicates that in Ad5 infections, the small replication foci are transient and rapidly progress to larger replication centers. By 17 hpi in the absence of the ATM kinase inhibitor, 70% to 85% of both Ad5- and E4 mutant-infected cells had focal concentrations of E2-72kDa, and again, nearly all of the cells infected with E4- contained small early replication foci, while nearly all of the Ad5-infected cells contained larger intermediate- or late-stage replication centers. We found that KU60019 treatment allowed E4 mutant-infected cells to convert from early replication foci to large replication centers more efficiently. Large replication centers were present in about 10% of the cell population infected with E4- at 12 hpi and about 50% at 17 hpi (Figure 10C). Inhibiting ATM kinase activity had little effect on the conversion of early replication foci to large replication centers in Ad5 infections. Taken together, these results indicate that although pATM does not prevent the formation of early replication foci in E4- infections, it substantially inhibits their development into larger DNA replication centers.

When we studied the localization of pATM relative to late-stage replication centers containing E2-72kDa at 17 hpi with wild-type Ad5, we observed that about 30% of the cells demonstrated unexpectedly high levels of staining with the pATM antibody (data not shown). In these cells, the pATM antibody staining was dispersed throughout the nucleus and did not colocalize with E2-72kDa. The widespread pATM antibody staining was still observed in Ad5- infected cells treated with KU60019, even though this drug effectively inhibited the appearance of pATM foci in cells infected with E4- (Figure 10B, top panel). We also saw no evidence of

pATM in Western blot analyses of extracts from Ad5 infections, although a clear band corresponding to pATM was seen in extracts from cells infected with E4- (Figure 10B, bottom panel). These results suggest the possible presence of a cross-reacting epitope that is upregulated during the late phase of Ad5 infections. We did not see evidence of this cross-reacting epitope being upregulated in E4- infections, even when DNA replication was rescued by KU60019 treatment.

Mre11 and ATM localization to E4 mutant replication centers is not sufficient to prevent E4 mutant DNA replication when ATM kinase activity is inhibited. Mre11 and pATM localize to viral DNA replication centers in E4 mutant infections (Carson et al., 2003; Stracker et al., 2002), and Mre11 is known to bind E4 mutant DNA (Mathew and Bridge, 2008). Our previous results indicate that the interaction of MRN complex proteins with E4 mutant genomes is important for inhibiting DNA replication (Mathew and Bridge, 2008). We studied the localization of Mre11 at E4 mutant DNA replication centers following treatment with KU60019 to determine if inhibition of ATM kinase activity affected recruitment of Mre11 to replication centers. HeLa cells were infected with E4- at an MOI of 3 FFU/cell, and at 2 hpi, cells were either left untreated or incubated with medium containing 5 μM KU60019 until 17 hpi. Infected cells were then fixed and immunostained with antibodies against Mre11 and E2-72kDa. The results are presented in Figure 11A. In E4- infections, Mre11 colocalized with most E2-72kDa centers in both the presence and absence of KU60019. E4 mutant replication centers were significantly larger in cells treated with KU60019. These results indicate that ATM kinase inhibition does not prevent the association of Mre11 with E4 mutant replication centers. We also studied the localization of Mre11 following E4- infection of ATM− cells at 3 FFU/cell. At 17 hpi, cells were fixed and immunostained with antibodies against Mre11 and E2-72kDa. Mre11 colocalized with E2-72kDa in ATM− cells (Figure 11A). Taken together, these data indicate that neither ATM nor its kinase activity is required for Mre11 localization at E4 mutant DNA replication centers. Since E4- DNA replication is substantially rescued in KU60019-treated infections (Figure 8 and 10) and in ATM− cells (Figure 11), the results also suggest that localization of Mre11 at replication centers is not sufficient to inhibit E4- DNA replication in the absence of ATM or its kinase activity.

We next investigated pATM and ATM localization following E4- infection of Nbs1− cells that fail to make a functional MRN complex (Cerosaletti et al., 2006) and ATM localization

Figure 11. Localization of Mre11 or ATM to viral replication centers does not interfere with E4- DNA replication when the ATM kinase is inactivated. (A) HeLa or ATM− cells were infected with E4- at 3 FFU/cell and fixed for immunofluorescence at 17 hpi. The indicated infections were incubated with 5 μM KU60019 from 2–17 hpi to inhibit ATM kinase activity. Immunofluorescence micrographs showing E2-72kDa (green) and Mre11 (red) and merged staining patterns are shown. (B) Nbs1+, Nbs1−, and HeLa cells were infected with E4- at 3 FFU/cell and fixed for immunofluorescence at 17 hpi. E4- infected HeLa cells were incubated with 5 μM KU60019 from 2–17 hpi to inhibit ATM kinase activity. Immunofluorescence micrographs showing E2-72kDa (72K), pATM, ATM, and merged staining patterns are shown.

in HeLa cells that were treated with KU60019 to inhibit ATM kinase activity. Cells were infected with E4- at an MOI of 3 FFU/cell, and at 17 hpi, the cells were fixed and immunostained with the indicated antibodies. The results are presented in Figure 11B. We saw very little pATM in Nbs1− cells compared to Nbs1+ cells. E4 mutant replication foci were significantly larger in Nbs1− cells, which is consistent with previous observations that E4 mutant DNA replication is rescued in these cells (Figure 11B, left panels) (Mathew and Bridge, 2008). Unphosphorylated ATM did not localize to E4- replication centers in Nbs1− cells, as expected, since Nbs1 is required to recruit ATM to sites of DNA damage (Figure 11B, right panels) (Cerosaletti et al., 2006; You et al., 2005). In HeLa cells treated with KU60019 to inhibit ATM kinase activity, unphosphorylated ATM was still recruited to viral DNA replication centers. Since KU60019 significantly rescues E4 mutant DNA replication (Figure 8 and 10), these results indicate that ATM recruitment to viral replication centers does not itself interfere with viral DNA replication if the ATM kinase is inactive. Taken together, these results indicate that a functional MRN complex is needed for activation of pATM, and for the association of ATM with E4- DNA replication centers, but this is not sufficient to interfere with viral DNA replication if the ATM kinase is not active.

DISCUSSION

We have sought to clarify the role of the ATM and ATR kinases in several MRN- dependent E4 mutant phenotypes. Cells lacking either ATM or ATR still concatenate E4 mutant genomes (Figure 5) (Stracker et al., 2002). However, previous reports indicated that genome concatenation is sensitive to the ATM/ATR inhibitor caffeine (Jayaram and Bridge, 2005; Lakdawala et al., 2008), suggesting a possible role for these kinases. We performed siRNA knockdown of ATR in ATM− cells to address the possibility that these kinases might functionally substitute for each other, and to our surprise, we found that E4 mutant genomes were still concatenated (Figure 5). ATR knockdown effectively prevented Chk1 phosphorylation in E4- infections (data not shown), suggesting that knockdown was sufficient for at least this ATR-mediated activity. Although we cannot rule out the possibility that ATR knockdown was insufficient to prevent genome concatenation, we think it is likely that caffeine affects genome concatenation by a mechanism that does not involve ATM and ATR kinase inhibition. DNAPK exhibits sensitivity to caffeine in vitro (Block et al., 2004), although DNAPK-dependent NHEJ in vivo is resistant (Asaad et al., 2000; Block et al., 2004). Nevertheless, it is possible that high concentrations of caffeine or prolonged incubation times could affect DNAPK, which is a critical enzyme for E4 mutant genome concatenation (Boyer et al., 1999). Alternatively, caffeine could affect an as-yet-unidentified substrate critical for genome concatenation.

The MRN complex inhibits E4 mutant DNA replication (Evans and Hearing, 2005; Lakdawala et al., 2008; Mathew and Bridge, 2007; Mathew and Bridge, 2008), but the mechanism involved is not understood. ATR does not inhibit E4 mutant late gene expression (Figure 6) or DNA replication (Lakdawala et al., 2008). In contrast, when E4 mutant infections were analyzed in cells that lack ATM, we saw similar increases in both viral DNA and late protein penton levels in experiments performed at an MOI of 30 FFU/cell (Figures 2 and 3), suggesting that the primary effect of ATM is inhibition of viral DNA replication, which is required for late gene expression. Previous work (Bridge and Ketner 1989) indicated that E4 mutants still display a measurable defect in late gene expression following high MOI infections that show a slight or no DNA replication defect. This may explain why E4- DNA levels were within 70% of Ad5 values in ATM− cells (Figure 7B, top graph) whereas E4- penton levels still demonstrated a significant 3-fold reduction relative to Ad5 in ATM− cells (Figure 7A). Our data suggest that the increase in E4- late gene expression in cells lacking ATM most likely reflects an

increase in viral DNA levels that occurs in these cells, but ATM may not fully complement the component of the late gene expression defect that is independent of viral DNA replication. The E4- DNA replication phenotype was more severe when cells were infected at a lower MOI of 3 FFU/cell. We also observed a more dramatic rescue of E4- DNA levels in the absence of ATM (Figure 7B, lower graph) and following treatment with ATM kinase inhibitors (Figure 8). These results confirm the ability of ATM and its kinase activity to interfere with E4 mutant DNA replication and are in contrast to those of Lakdawala et al. (Lakdawala et al., 2008), who reported that E4 mutant DNA replication was still severely defective in A-T cells that lack ATM. Possible differences between our experiments and theirs include the MOI and the source of the A-T cells used. Although we do not know the reason for the discrepancy between our results and those of Lakdawala et al. (Lakdawala et al., 2008), we have observed an increase in E4 mutant DNA replication in ATM− cells compared with an isogenic ATM+ cell line complemented with ATM (Figure 7B), in HeLa cells treated with ATM kinase inhibitors (Figure 8), and following siRNA knockdown of ATM in HeLa cells (data not shown). Thus, multiple lines of evidence support our conclusion that ATM inhibits E4 mutant DNA replication, particularly following low-MOI infections. Interestingly, E4 mutant DNA replication was rescued to similar extents (about 50% of Ad5 levels) following low-MOI infection in ATM− cells (Figure 7B, lower graph), in Nbs1− cells that lack a functional MRN complex (data not shown), and in siATM- treated HeLa cells (data not shown). This supports the idea that the ability of the MRN complex to inhibit E4 mutant DNA replication is likely due to its ability to activate ATM. This does not rule out ATM-independent roles for the MRN complex in other E4 mutant phenotypes. For example, ATM is not required for the ability of MRN to promote E4 mutant genome concatenation (Figure 5).

Karen and Hearing (2011) have previously shown that pATM colocalizes with foci containing incoming viral DNA in E4 mutant infections, and pATM is associated with E4 mutant replication centers that contain the E2-72kDa DNA replication protein at later times (Stracker et al., 2002). Localization of pATM with either incoming or replicating viral genomes could potentially interfere with E4 mutant DNA replication. Interestingly, we observed focal accumulation of pATM in Ad5-infected cells similar to the pATM foci seen in E4 mutant infections. Ad5 pATM foci were clearly visible at 5 hpi before a detectable level of the early protein E2-72kDa was expressed (Figure 9A and B). However, once early proteins are detected,

the intensity of pATM staining was greatly reduced in Ad5-infected cells (Figure 9C, panels d to f). We think it is likely that either E4- or Ad5 infection can trigger focal concentration of pATM at locations containing input viral DNA at early times. Early gene expression, including that of the Ad5 E4 proteins, would then be expected to inactivate the MRN complex, and this likely interferes with sustained pATM activation in Ad5 infections. Although pATM was observed in early foci before the onset of DNA replication in E4- infections (Figure 9B), this did not prevent the timely development of small replication foci containing the E2-72kDa DNA binding protein (Figure 9A and C). In contrast, we observed a dramatic effect on the conversion of small replication foci to larger E2-72kDa-containing centers in E4- infections with active pATM (Figure 9C and 9C). E4- replication centers also showed less intense incorporation of BrdU (Figure 9D) compared with Ad5 results. ATM kinase inhibition did not affect the overall percentage of E4 mutant-infected cells with E2-72kDa foci, but it significantly increased the ability of the E4 mutant to convert small DNA replication foci into larger DNA replication centers (Figure 10C). Our data suggest that pATM does not completely block E4 mutant DNA replication but significantly reduces its efficiency.

The ability of MRN to localize at viral DNA replication centers correlates with its ability to inhibit E4 mutant DNA replication (Mathew and Bridge, 2007). The Nbs1 C-terminal domain is important for binding Mre11 (You et al., 2005) and for inhibiting E4 mutant DNA replication (Lakdawala et al., 2008). Furthermore, we have found that Nbs1 is important for the ability of Mre11 to both bind E4- DNA and inhibit DNA replication (Mathew and Bridge, 2008). These results are consistent with a model in which Nbs1-dependent recruitment of Mre11 to viral DNA is important for inhibiting E4 mutant DNA replication. Nevertheless, recruitment of Mre11 to viral DNA may not be sufficient for this effect. We found that Mre11 is still recruited to viral replication centers in cells that lack ATM and when the ATM kinase activity has been inhibited (Figure 11A), and yet this is not sufficient to block E4 mutant DNA replication (Figures 7 to 10). Lakdawala et al. (2008) found that a C-terminal fragment of Nbs1 that binds Mre11 but fails to bind ATM was still able to recruit the MRN complex protein Rad50 to viral replication centers but was unable to interfere with E4 mutant DNA replication (Lakdawala et al., 2008). Although not interpreted as such, these results are completely consistent with our observations that the MRN complex is needed to activate pATM, allow it to localize at E4 mutant DNA replication centers (Figure 11B), and inhibit DNA replication (Figures 7 to 10). Taken together, these results

suggest a model in which Ad genomes are delivered to the nucleus and initially induce the focal accumulation of pATM at locations containing viral genomes (Karen and Hearing, 2011) (Figure 9A). We propose that inactivation of Mre11 by E4 proteins prevents sustained activation of pATM in Ad5 infections, which are subsequently able to efficiently develop large late-stage DNA replication centers (Figures 9 and 10). In E4 mutant infections, pATM activation is sustained due to the failure to inactivate Mre11, and pATM remains tightly associated with early DNA replication foci and the larger intermediate-stage DNA replication centers that develop inefficiently in the presence of pATM (Figures 9 and 10). Activated ATM could thus inhibit DNA replication when it is associated with centers containing viral DNA.

The mechanism by which ATM interferes with E4 mutant DNA replication is unknown. Repair of double-strand breaks following ionizing radiation involves extensive chromatin changes in the vicinity of the break. Many DDR proteins are recruited to the ends of the break and form foci that can be observed by microscopy; these include MRN complex proteins, pATM, which can phosphorylate the histone variant H2AX (γH2AX), Mdc1, 53BP1, and BRCA1 (Bekker-Jensen and Mailand, 2010). If a large complex of proteins assembles at the ends of E4 mutant genomes during activation of DDRs, this could possibly interfere with the ability of the viral DNA replication proteins to identify the terminally located Ad origin of replication and initiate viral DNA synthesis (Mathew and Bridge, 2008; Karen and Hearing, 2011). In view of this model, it is interesting that in E4- infections, early replication foci initiated in cells with a frequency similar to that seen with Ad5 (Figures 9C and 10), although they failed to develop into the large late-stage replication centers typically seen in Ad5 infections (Figures 9C and 10). These results suggest that the ability of viral replication proteins to initiate E4 mutant DNA replication is not completely blocked by DDR activation but that E4 mutant DNA replication progresses inefficiently. Possible roles for activated ATM in interfering with E4 mutant DNA replication could include altering the chromatin configuration of viral DNA and thus making replication inefficient, phosphorylating specific substrates that interfere with viral DNA replication, or reducing DNA replication efficiency by affecting cell cycle checkpoints. Further experiments to elucidate these mechanisms are in progress and should help us to better understand how cellular DDRs interfere with E4 mutant DNA replication.

ACKNOWLEDGMENTS

We are very grateful to Ulf Pettersson, Tommy Linné, and Arnold Levine for providing antibodies and Pat Concannon for providing the Nbs1− and Nbs1+ cell lines used in this study. We also thank Joseph Carlin for critically reading the manuscript and all the members of our laboratory for their suggestions and support.

This research was supported by the National Cancer Institute (grant CA82111) and by awards from Miami University.

Ataxia telangiectasia mutated kinase is important for recruiting late DNA damage repair proteins to Adenovirus E4 mutant DNA replication centers.

Dipendra Gautama and Eileen Bridgeab#

aDepartment of Microbiology

32 Pearson Hall

Miami University

Oxford OH 45056 bCell Molecular and Structural Biology Program

#Corresponding author

Tel: 513-529-7264

Fax: 513-529-2431

Email: [email protected]

Manuscript in preparation

Abstract

Adenovirus (Ad) type 5 (Ad5) produces early regulatory proteins from E4 and E1b that prevent activation of the DNA damage response kinase ataxia-telangiectasia mutated (ATM). Early region 4 (E4) mutants lack these regulatory proteins and induce activation of ATM, which prevents efficient viral DNA replication. During E4 mutant infections, activated ATM is phosphorylated (pATM), and colocalizes with viral DNA replication centers. E4 mutant DNA replication is rescued in cells that lack ATM or its kinase activity. Here, we find that Nbs1- dependent recruitment of ATM to E4 mutant replication centers is important for inhibiting DNA replication. Viral replication centers also develop normally in E4 mutants that activate pATM, but fail to maintain pATM localization at viral DNA replication centers. Localization of early DNA damage response (DDR) proteins, including Mre11, Mdc1 and γH2AX, to E4 mutant DNA replication centers is not affected by the absence of ATM. In contrast, recruitment of late DDR proteins RNF8 and 53BP1 to E4 mutant replication centers is substantially impaired when ATM is absent or when pATM fails to localize to the replication centers.

Keywords: Adenovirus, E4 ORF 3, E4 11kDa, E4 ORF 6, E4 34kDa, Mre11, MRN complex, ATM, RNF8, 53BP1, DNA damage response, double-strand break repair, chromatin remodeling

Introduction

Double-strand breaks (DSBs) are sensed by the Mre11/Rad50/Nbs1 (MRN) complex, which activates the DNA repair kinases ATM, and ATM and Rad3-related (ATR). These kinases subsequently phosphorylate effector proteins involved in regulating DNA repair, cell cycle checkpoints, and apoptosis (Kastan and Lim, 2000; Shiloh, 2003; DiTullio et al., 2002; Paull et al., 2000; Fernandez-Capetillo et al., 2002). Activation of the DNA damage response (DDR) following ionizing radiation (IR) results in rapid accumulation of many DNA repair proteins at the site of DNA damage in distinct nuclear foci called IR-induced foci (IRIF). These foci can be readily visualized by immunofluorescence microscopy (Lisby et al., 2004). MRN, pATM (Bakkenist and Kastan, 2003), and mediator of DNA damage checkpoint 1 (Mdc1) are examples of proteins that are found in IRIF (Iijima et al., 2008). The MRN complex binds to ends of the broken DNA and recruits ATM. ATM is activated by auto-phosphorylation and subsequently phosphorylates the histone variant H2AX (γH2AX) (Paull et al., 2000) to create a binding site for Mdc1 (Stucki et al., 2005), which in turn binds more molecules of the MRN complex and ATM, allowing for the spread of γH2AX modified chromatin hundreds of kilobases away from the from the initial site of the break (Stucki and Jackson, 2006; Kim et al., 2006). At DSBs, ATM also phosphorylates Mdc1 (Stewart et al., 2003; Goldberg et al., 2003) to create a docking site for RNF8, a ring finger-containing E3 ubiquitin ligase (Mailand et al., 2007; Xu and Price, 2011). RNF8 recruits RNF168 (Stewart et al., 2009) and together they add lysine 63-linked polyubiquitin (K63Ub) chains to H2A and H2AX on damaged chromatin (Doil et al., 2009; Huen et al., 2007). The ubiquitinated histones create a recruitment site for repair proteins such as 53BP1 and BRCA1 (Mailand et al., 2007). Recruitment of DDR proteins to IRIF is required for checkpoint activation and repair of the damaged DNA (Lavin, 2008), resulting in a fully activated DDR.

Ad mutants that lack E4 activate DDR responses (Carson et al., 2003). Mre11, Mdc1, and pATM are reorganized into nuclear foci similar to IRIF at sites containing viral genomes (Karen and Hearing, 2011), and many of these DDR proteins are recruited to viral DNA replication centers that develop in infected cells (Stracker et al., 2002; Carson et al., 2003; Gautam and Bridge, 2013; Karen and Hearing, 2011). Activation of the DDR requires transcription-mediated remodeling of the viral genome during early gene expression (Karen and Hearing, 2011). Ad5 and E4 mutant infections can both induce autophosphorylation of pATM and the accumulation

of DDR proteins in nuclear IRIF-like foci at early times during infection (Karen and Hearing, 2011; Gautam and Bridge, 2013). However, activation of the DDR is not sustained in Ad5 infection, and many DDR proteins that are present in E4 mutant DNA replication centers are not detected in Ad5 DNA replication centers (Stracker et al., 2002; Carson et al., 2003; Gautam and Bridge, 2013). The failure of Ad5 to sustain the DDR results from the action of early regulatory proteins. A 34kDa protein synthesized from E4 orf6 (E4-34kDa) forms a complex with a 55kDa viral protein expressed from E1b (E1b-55kDa) that targets several DNA repair proteins including Mre11 for proteosomal degradation (Stracker et al., 2002; Carson et al., 2003). An 11kDa protein from E4 orf3 (E4-11kDa) interacts with Mre11 and redistributes it to nuclear tracks preventing Mre11 association with viral DNA replication centers.

Activation of the DDR severely impacts E4 mutant DNA replication (Stracker et al., 2002; Lakdawala et al., 2008; Evans and Hearing, 2003; 2005; Mathew and Bridge, 2007; 2008; Gautam and Bridge, 2013). Cells that lack a functional MRN complex substantially rescue E4 mutant growth defects (Evans and Hearing, 2005; Mathew and Bridge, 2008). Furthermore, we have recently shown that activated ATM kinase interferes with E4 mutant DNA replication, suggesting that MRN may inhibit viral DNA replication through its ability to activate ATM (Gautam and Bridge, 2013). The mechanism of ATM-mediated inhibition of E4 mutant DNA replication is currently unknown. ATM is activated by the MRN complex that binds and recruits the kinase to the site of DSBs. Cells that lack Nbs1 are defective for ATM activation (Cerosaletti et al., 2006). Recruitment of ATM to the damage site and its activation requires its association with the C-terminal ATM binding domain in the Nbs1 protein (Flack et al., 2005; You et al., 2005). Indeed, cells lacking Nbs1 derived from a patient with Nijmegen breakage syndrome (NBS), tranfected with a construct that expressed the C-terminus of Nbs1 (Nb1FR5), were used to show that the C-terminal ATM binding domain of Nbs1 was sufficient to activate ATM in response to ionizing radiation (Cerosaletti et al., 2006). In similar expression experiments, the Nbs1 C-terminus promoted ATM activation and inhibited DNA replication in E4 mutant infections (Lakdawala et al., 2008). These observations raise the possibility that Nbs1-mediated recruitment of ATM to viral DNA is required to inhibit E4 mutant DNA replication. It is also known that E4-11kDa-mediated redistribution of Mre11 is sufficient to allow close to normal levels of Ad DNA replication in mutants that lack a functional E4-34kDa/E1b-55kDa complex (Evans and Hearing, 2005), despite the observation that such mutants still activate ATM (Carson

et al., 2009). Since a functional MRN complex is required to recruit ATM to DSBs, this raises the possibility that E4 11kDa promotes viral DNA replication by preventing activated pATM from localizing to viral DNA replication centers.

Previous studies have indicated that recruitment of DDR proteins such as the components of the MRN complex correlates with the defects in E4 mutant DNA replication (Mathew and Bridge, 2007; 2008). However recruitment of Mre11 to the viral replication centers is not sufficient to inhibit E4 mutant DNA replication in the absence of activated ATM (Gautam and Bridge, 2013). ATM could inhibit E4 mutant DNA replication by recruiting additional DDR factors. For instance, Mdc1 is recruited to DSBs via ATM-mediated phosphorylation of γH2AX (Stucki et al., 2005). Recent studies also indicate that ATM activation is required for recruitment of downstream DDR proteins and chromatin remodeling factors to DSBs (reviewed in Price and D’Andrea, 2013). ATM-mediated recruitment of additional factors to the viral genome or replication centers could possibly obstruct the viral DNA polymerase from accessing the origin of replication thereby inhibiting viral DNA replication. Here we have investigated the importance of localization of ATM and other DDR proteins to Ad replication centers for inhibition of E4 mutant DNA replication. Our results indicate that failure to localize activated pATM to E4 mutant DNA replication centers is correlated with increased DNA replication, and does not prevent the localization of early DDR proteins such as γH2AX, Mre11, and Mdc1 to DNA replication centers. In contrast, late DDR proteins such as RNF8 and 53BP1 were not found at E4 mutant DNA replication centers that lacked pATM.

Materials and methods

Cells and viruses. HeLa and E4 mutant-complementing W162 cells (Weinberg and Ketner, 1983) were maintained in Dulbecco’s Modified Eagle Medium (DMEM) (Fisher) supplemented with 10% fetal bovine serum (FBS), 10 U/ml penicillin and 10 µg/ml streptomycin. Fibroblast cell lines that lack the ATM gene (GM16666, referred to here as ATM-) and that have normal ATM expression (GM00637, referred to here as ATM+) were purchased from the Coriell cell repository. These cells were maintained in DMEM supplemented with 10% FBS, 10 U/ml penicillin, 10 µg/ml streptomycin. Cells lacking the Nbs1 gene (NBS-ILB1) that were stably transfected with empty pLXIN retroviral vector (referred to here as Nbs1-), NBS- ILB1 cells that were stably transfected with the pLXIN retroviral vector expressing the wild-type Nbs1 protein (referred to here as Nbs1+) and NBS-ILB1 cells that were stably transfected with a pLXIN retroviral vector expressing the Nbs1 protein with deletion in C-terminal 19 amino acids containing ATM binding motif (referred to here as Nbs1∆ATM) were obtained from Pat Concannon (Cerosaletti et al., 2006), and maintained in DMEM supplemented with 10% FBS, 10 U/ml penicillin, 10 µg/ml streptomycin and 500 µg/ml G418 (Invitrogen). Wild type Ad5, E4 mutant H5dl1007 (referred to here as E4-) and E4 mutant H5dl1010 (referred to here as 1010) (Bridge and Ketner, 1989) viral stocks were prepared on W162 cells. Virus titers were determined on W162 cells as described (Phillipson, 1961) and were expressed as fluorescence forming units/ml (FFU/ml).

Immunofluorescence microscopy. Immunofluorescence microscopy was performed as previously described (Gautam and Bridge, 2013) for all double immunostainings using two primary antibodies. For triple immunostaining, three primary antibodies generated in different species were incubated with the fixed cells for 1 h. After washing 3 times for 5 min each with Tris-buffered saline (TBS) (Tris-buffered saline; 100 mM Tris-Cl [pH 7.5], 150 mM NaCl), and rinsing once with TBST (Tris-buffered saline Tween; 100 mM Tris-Cl [pH 7.5], 150 mM NaCl, 0.1% Tween 20), donkey anti-mouse and donkey anti-goat secondary antibodies were incubated for 45 min. The cover glasses were then washed 3 times for 5 min each in TBS and cross-linked with 4% PFA in PBS (0.058 M Na2HPO4, 0.017 M NaH2HPO4, 0.069 M NaCl) to prevent cross reaction between secondary antibodies. Subsequently goat anti-rabbit AlexaFluor 405 IgG was added and washed with TSB before mounting the cover glass.

The following primary antibodies were used for immunofluorescence microscopy at the dilutions specified: goat polyclonal anti-Mre11 (Santa Cruz) (1:100), rabbit polyclonal anti- Mdc1 (Betyl) (1:100), mouse monoclonal anti-γH2AX (Abcam) (1:200), mouse monoclonal Bcl6 anti-72K (a gift from A. Levine) (1:100), rabbit polyclonal anti-72K (a gift from T. Linné) (1:1500), mouse monoclonal anti-pATM kinase ( phospho epitope S1981, Novus) (1:200), rabbit polyclonal anti-RNF8 (Abcam) (1:30), rabbit polyclonal anti-53BP1 (Abcam) (1:100). The following secondary antibodies from Invitrogen were used for immunofluorescence at the dilutions specified: goat anti-rabbit AlexaFluor 405 IgG (1:250), donkey anti-rabbit AlexaFluor 594 IgG (1:250), donkey anti-mouse AlexaFluor 488 IgG (1:250), donkey anti-goat AlexaFluor 594 IgG (1:250).

A confocal laser scanning microscope (Olympus FV500) was used to capture images with a 100X oil immersion objective and Fluoview software. Sequential scans of the Alexa 405, Alexa 488 and Alexa 594 channels were performed to prevent bleed through between the channels. Images were assembled using Adobe Photoshop CS5 software.

Viral DNA analysis. Viral DNA was prepared and analyzed as previously described (Gautam and Bridge; 2013). Dot blotting and slot blotting analysis were performed using the RNA-free DNA. Samples containing 1.0, 0.3 and 0.1 µg of DNA were denatured by boiling at 100°C for 10 min and immediately cooled on ice. Samples were adjusted to 6X SSC (1M NaCl, 100mM sodium citrate). The samples were blotted on Hybond-N nylon membrane (Amersham) that was pre-soaked with 6X SSC using a dot blot (Whatman) or a slot blot (Bio-rad) manifold and applying vacuum pressure. The DNA was cross-linked to the membrane using a UV- transilluminator. The membranes were then probed with Ad-specific radiolabeled probes generated using the Megaprime DNA Labeling System (GE Healthcare) following the manufacturer’s instructions. The hybridization reaction was performed with 5 X 106 cpm/ml probe at 65°C overnight. The amount of labeled DNA bound was detected using a STORM 860 phosphorimager (Molecular Dynamics) and quantified using ImageQuant® 5.2 (Molecular Dynamics) software or exposing them to ECL hyperfilm (Amersham).

DSBR kinase inhibitors. Caffeine (Sigma) was dissolved in water and added 2 hpi at final concentration of 10 mM until 22 hpi. KU60019 (Tocris) was dissolved in DMSO and added to the culture at 2 hpi to a concentration of 5 µM for the desired time.

Results

Failure to localize activated pATM to viral replication centers rescues E4- DNA replication. ATM is recruited to DSBs by binding to the C-terminal ATM binding motif of Nbs1. A cartoon showing the motifs present in the Nbs1 protein is shown in Figure 12A. This interaction is important for efficient activation of ATM and for its localization at DSBs (Flacks et al., 2005; You et al., 2005; Lee and Paull, 2005). Deletion of the Nbs1 ATM binding motif dramatically reduced pATM levels following IR (Cerosaletti et al., 2006). To test the dependence on Nbs1/ATM interaction for the ability of ATM to inhibit E4 mutant DNA replication, we performed experiments in cells lacking a functional Nbs1 gene that had been stably transfected with full length Nbs1 (Nbs1+), with vector only (Nbs1-), or transfected with an Nbs1 mutant gene that lacks the ATM binding motif (Nbs1∆ATM) (Cerosaletti et al., 2006). Each of these cell lines were infected with E4- or Ad5 at a multiplicity of infection (MOI) of 3 FFU/cell. At 24 hpi cells were harvested and dot blotting analysis was performed to measure the levels of viral DNA. In Nbs1- cells E4- DNA levels were substantially rescued and not significantly different from Ad5 (p=0.09) (Figure 12B). During infections in Nbs1+ cells, E4- demonstrated a 77-fold defect in DNA compared to Ad5 (Figure 12B). Interestingly, in Nbs1∆ATM cells E4- DNA levels increased 6-fold compared to Nbs1+ cells, although they were still significantly less (8-fold) than E4 mutant DNA levels in Nbs1- cells (Figure 12B). These results suggest that E4 mutant DNA replication in Nbs1∆ATM cells is only partially rescued. Cerosaletti et al., (2006) have shown that after IR treatment, Nbs1∆ATM cells have reduced pATM levels compared to Nbs1+ but still higher than in Nbs1- cells when pATM was measured by western blot. It is possible that residual pATM accounts for the slight increase in E4- DNA replication in Nbs1 ∆ATM cells. We tested if activated pATM was interfering with E4- DNA replication in Nbs1∆ATM cells by treating the cells with ATM kinase inhibitor, KU60019. Nbs1∆ATM cells were infected with either E4- or Ad5 and at 2 hpi cells were treated with 10 µM KU60019 for 22 h. A slot blot analysis of total DNA prepared from the different cultures is shown in Figure 12C. E4- DNA replication in Nbs1∆ATM cells treated with KU60019 increased when compared with DMSO treatment (Figure 12C) suggesting that ATM kinase activity is still inhibiting E4- DNA replication in Nbs1∆ATM cells. Since treatment of Nbs1∆ATM cells with KU60019 increased E4- DNA replication, we determined if Nbs1∆ATM cells had detectable ATM activation during E4- infection by looking for the presence of pATM in DNA replication centers by

Figure 12. An Nbs1 mutant lacking the ATM binding motif partially rescues E4- DNA replication. (A) Cartoon depicting Nbs1 protein with the ATM binding motif. (B) Nbs1+, Nbs1- and Nbs1∆ATM cells were either uninfected or infected with E4- or Ad5 at an MOI of 3 FFU/cell. At 24 hpi cells were harvested and total DNA from 3 independent experiments was prepared for dot blot analysis to measure viral DNA levels. E4- DNA levels were expressed as the fraction of the level achieved by Ad5, which was set as 1. Error bars represent standard error of mean. Statistically significant (p<0.05) differences between the columns are indicated as determined by Student's t-test. (C) Nbs1∆ATM cells were infected with E4- and Ad5 at 3 FFU/cell with either 5 µM KU60019 or the equivalent volume of the solvent DMSO (no drug). At 24 hpi cells were harvested and slot blot analysis was performed. A representative slot blot is shown.

immunofluorescence. Nbs1+, Nbs1- and Nbs1∆ATM cells were infected with E4- or Ad5 for 17 h and subsequently fixed and immunostained with anti-pATM and anti-E2-72kDa antibodies to identify viral replication centers. Confocal micrographs of the immunostained cells are shown in Figure 13. Ad5 infected cells showed undetectable pATM staining in well-developed E2-72kDa containing replication centers in all cell types (Figure 13A, B and C, panels e to g). E4 mutant infection of Nbs1+ cells showed pATM staining associated with poorly developed DNA replication centers, while no such pATM staining was observed in Nbs1- cells (Figure 13A and C, panels i to k), as we have previously reported (Gautam and Bridge, 2013). However, in Nbs1∆ATM cells, we observed reduced (compared to Nbs1+ cells, panels i to k) but visible pATM staining associated with E4- replication centers (Figure 13B, panels i to k). The reduced pATM staining in Nbs1∆ATM was not due to failure to recruit Mre11 to replication centers (data not shown). These results confirm that ATM is partially activated during E4- infection of Nbs1∆ATM cells and this likely contributes to the partial inhibition of E4- DNA replication in these cells. Nevertheless, reduced levels of pATM in E4- DNA replication centers in Nbs1∆ATM cells were correlated with an increase in E4- DNA replication, when compared to control Nbs1+ cells that express full length Nbs1 and fully activate pATM (Figure 12B). Our results indicate that Nbs1 interaction with ATM is important for efficient activation and localization of pATM at E4- DNA replication centers, and for efficient inhibition of E4- DNA replication.

E4 mutant 1010 carries a deletion in the gene that expresses E4-34kDa (Bridge and Ketner, 1989), which is important for degradation of multiple cellular factors including proteins involved in the DDR such as Mre11 (Carson et al., 2003). However, 1010 expresses E4-11kDa and can redistribute Mre11 into nuclear tracks and cytoplasmic aggresomes and prevent its localization to viral replication centers (Carson et al., 2003; Prakash et al., 2012; Araujo et al., 2005). The status of the genes for E4-34kDa and E4-11kDa in the viruses studied in this work is shown in Figure 14A. Previously it was shown that E4-11kDa is sufficient to promote efficient viral replication in mutants that fail to express E4-34kDa (Evans and Hearing, 2005), despite the observation that 1010 and a similar mutant efficiently activate ATM (Carson et al., 2009; Prakash and Bridge, unpublished data). Since activated ATM inhibits DNA replication in E4 mutants that lack both E4-34kDa and E4-11kDa, we hypothesized that E4-11kDa mediated Mre11 redistribution might prevent pATM localization to replication centers in 1010 infections,

Figure 13. Reduced pATM localization to E4- replication centers in Nbs1∆ATM cells. Nbs1+ (A), Nbs1∆ATM (B) and Nbs1- (C) cells were infected at 3 FFU/cell and at 17 hpi cells were fixed and immunostained with antibodies against pATM (phosphoepitope S1981) and the viral DNA binding protein E2-72kDa (72K) to detect replication centers. Immunofluorescence confocal micrographs are presented with pATM (green) and E2-72kDa (red) staining, or phase contrast images of uninfected, Ad5 or E4- infected cells.

Figure 14. pATM fails to localize to 1010 replication centers. (A) Table showing the status of E4-11kDa and E4-34kDa expressing genes in the viruses studies. (B) HeLa cells were infected with E4-, 1010 or Ad5. At 17 hpi cells were fixed and stained with anti-pATM, anti-Mre11 and anti-E2-72kDa antibodies. Immunofluorescence confocal micrographs showing E2-72kDa (72K) in blue, pATM in green, Mre11 in red and merged staining patterns are presented. Cells with early and late stage replication centers were identified based on size of E2-72kDa foci.

thereby preventing the inhibitory effect of pATM on viral DNA replication. We analyzed the localization of pATM relative to the viral replication centers during 1010 infections to address this hypothesis. HeLa cells were infected with E4-, 1010 and Ad5 for 17 h and fixed and immunostained with anti-72K, anti-pATM and anti-Mre11 (Figure 14B). During infections with Ad5, pATM staining was not detected and the DNA replication centers were well developed with no detectable Mre11 staining (Figure 14B, panels a to d). In contrast, pATM staining was always observed in E4- infections and which co-localized with poorly developed DNA replication centers along with Mre11 (Figure 14B, panels e to h). During 1010 infections we observed that in cells where Mre11 was completely redistributed into track like structures, replication centers were well developed. Interestingly, pATM showed diffuse nuclear staining (Figure 14B, panels m to p). In a small population of 1010 infected cells, we also observed small replication foci that co-localized with both pATM and Mre11 suggesting that when Mre11 is localized to the replication centers pATM is also present there (Figure 14B, panels i to l). The data indicate that E4-11kDa-mediated redistribution of Mre11 can prevent pATM localization at replication centers at late stages, and this is correlated with a robust increase in the size of replication centers. Taken together, these results suggest that activated ATM may only interfere with viral DNA replication when it is associated with replication centers.

Early DDR proteins are present in E4- DNA replication centers in the absence of ATM. Several DDR proteins are recruited to the site of DNA damage (Kim et al., 2006; Celeste et al., 2003; Lisby et al., 2004) as well as to E4 mutant replication centers (Carson et al., 2003; Stracker et al., 2002). Recruitment of these DDR proteins occurs in a temporal manner with early DDR proteins, including Mre11, Mdc1 and γH2AX recruited rapidly to the site. Previously it has been suggested that binding of Mre11 to the viral genome might block the access of the viral replication machinery to the origin of replication and hence inhibit DNA replication (Mathew and Bridge, 2008; Karen and Hearing, 2011). However, we have reported that recruitment of Mre11 to E4- replication centers is not sufficient to inhibit E4- DNA replication in absence of ATM or its kinase activity (Gautam and Bridge, 2013). Since Mre11 recruitment alone was not adequate to inhibit E4- DNA replication, we determined if other early DDR proteins such as Mdc1 and γH2AX were localized with Mre11 during E4- infection in the presence and absence of ATM. ATM- and ATM+ cells were either uninfected or infected with E4- at an MOI of 3 FFU/cell and fixed at 17 hpi. Cells were then immunostained with anti-Mre11, anti-Mdc1 and

anti-γH2AX antibodies. Confocal immunofluorescence micrographs are presented in Figure 15. Since Mre11 is colocalized with E4 mutant DNA replication centers throughout the infection (Gautam and Bridge, 2013), we used localization of Mre11 as a marker to represent sites of DNA replication. Mre11 co-localized with γH2AX and Mdc1 in viral DNA replication centers (Figure 15, panels f to i and p to s) in both E4- infected ATM+ and ATM- cells. These results suggest that the early DDR proteins Mre11, γH2AX and Mdc1 are recruited to E4- DNA replication centers irrespective of the presence or absence of ATM. Since E4- DNA replication is substantially rescued in ATM- cells (Gautam and Bridge, 2013), our data indicate that the recruitment of these early DDR proteins is not sufficient to inhibit E4- DNA replication in the absence of ATM.

Late DDR proteins RNF8 and 53BP1 are recruited to the E4- DNA replication centers in an ATM dependent manner. Late DDR proteins such as RNF8 and 53BP1 are recruited to DSBs in an ATM dependent manner following IR (Lavin, 2008; Kolas et al., 2007). Activated pATM phosphorylates Mdc1 which in turn recruits RNF8, an E3 ubiquitin ligase, which adds polyubiquitin to γH2AX. This event then recruits late DDR proteins such as 53BP1 and BRCA1 to the DSB (Mailand et al., 2007). We determined if late DDR proteins such as RNF8 and 53BP1 were present at viral DNA replication sites containing pATM in E4- infections (Gautam and Bridge, 2013). HeLa cells were uninfected or infected with E4-, fixed and immunostained at 7 (early) and 17 (late) hpi with anti-pATM and anti-RNF8 or anti-53BP1. Confocal micrographs of representative staining patters are presented in Figure 16. We found that RNF8 and 53BP1 were present with pATM at DNA replication centers during E4- infections at both 7 hpi (Figure 16A and B, panels d to f) and 17 hpi (Figure 16A and B, panels g to i). RNF8 and 53BP did not localize to DNA replication centers in Ad5 infections (data not shown) as expected since Ad5 replication centers lack pATM (Gautam and Bridge, 2013). Since RNF8 and 53BP1 are colocalized with pATM in E4- infections, we tested if RNF8 and 53BP1 were localized to DNA replication centers in an ATM dependent manner. ATM- and ATM+ cells were infected with E4- or Ad5 for 17 h and cell were fixed and immunostained with anti-E2- 72kDa to label viral DNA replication centers, and either anti-RNF8 or anti-53BP1 antibodies. In ATM+ cells RNF8 and 53BP1 colocalized with poorly developed replication centers in E4- infections (Figure 17A and B, panels u to w), while in Ad5 infected cells RNF8 and 53BP1 were not detected in DNA replication centers (Figure 17A and B, panels m to o). Interestingly, during

E4- infections in ATM- cells both RNF8 and 53BP1 failed to localized with DNA replication centers (Figure 17A and B, panels q to s), similar to wild type Ad5 (Figure 17A and B, panels i to k). Our results indicate that while ATM is not required for recruitment of several early DDR proteins to E4- DNA replication centers, it is required to recruit late DDR proteins RNF8 and 53BP to these sites.

During E4- infections in ATM+ cells RNF8 staining was observed to be brighter than in UI or Ad5 infected cells. Although there have been no reports of elevated RNF8 levels following other forms of DNA damage, we do not yet know if E4- infections upregulate the amount of RNF8 or whether it is concentrated in foci following DDR activation and therefore more visible. We also observed 53BP1 foci in both Ad5 and E4- infected cells with robust replication centers. However, these foci were not colocalized with the replication centers. Although the reason for these foci is not clear, they may involve other functions of 53BP1 independent of virus infections as we also see 53BP1 foci in uninfected cells.

Failure to localize activated pATM to E4- replication centers results in loss of the recruitment of 53BP1 to DNA replication centers. 53BP1 localization to DSBs depends upon ATM kinase activity following IR (Mainland et al., 2007). However, it is not known if localization of activated ATM to DSBs is required for recruitment of 53BP1 to these sites. Treatment of HeLa cells with KU60016 results in the loss of pATM from E4- replication centers (Gautam and Bridge, 2013). We tested if 53BP1 localization to E4- replication centers was affected by KU60019 treatment. HeLa cells were uninfected or infected with E4- and Ad5 at an MOI of 3 FFU/cell and at 2 hpi cells were either treated with DMSO or with KU60019. At 17 hpi cells were fixed and immunostained with anti-72K and anti-53BP1 antibodies. Confocal immunofluorescence micrographs are presented in Figure 18A. Ad5 infected cells treated with or without KU60019 showed no 53BP1 colocalization with DNA replication centers (Figure 18A, panels g to l). During E4- infection in absence of the drug, 53BP1 was colocalized with viral DNA replication centers (Figure 18A, panels m to o). However in presence of KU60019, 53BP1 was not detected in E4- DNA replication centers (Figure 18A, panels p to r). The result suggests that localization of 53BP1 to viral replication centers requires activated ATM kinase. We next asked if 1010 infections that activate pATM but fail to recruit it to replication centers, might also fail to localize 53BP1 to these sites. HeLa cells infected with 1010 for 17 h were immunostained with anti-53BP1 and anti-72K antibodies. Confocal immunofluorescence micrographs in Figure

Figure 15. Mdc1, γH2AX colocalize with Mre11 at E4- DNA replication centers independent of ATM. ATM+ and ATM- cells were infected with E4- and Ad5 viruses at an MOI of 3 FFU/cell. At 17 hpi cells were fixed and immunostained with antibodies against Mdc1, γH2AX and Mre11. Immunofluorescence confocal micrographs shown here are Mdc1 (blue), γH2AX (green) and Mre11 (red) staining, or phase contrast images of uninfected or Ad5 and E4- infected cells.

Figure 16. RNF8 and 53BP1 recruitment to pATM containing E4- DNA replication centers. HeLa cells were uninfected or infected with E4- and at 7 and 17 hpi cells were fixed and immunostained with anti-pATM (phosphoepitope S1981) and anti-RNF8 (A) or 53BP1 (B) antibodies. Immunofluorescence micrographs showing pATM (green) and RNF8 or 53BP1 (red), merged staining patterns and phase contrast images are presented.

Figure 17. Localization of RNF8 and 53BP1 to E4- replication centers requires ATM. ATM+ and ATM- cells were infected with E4- and Ad5 viruses at an MOI of 3 FFU/cell for 17 h. Cells were then fixed and immunostained with antibodies against the viral DNA binding protein E2-72kDa (72K) and RNF8 (A) or 53BP1 (B). Immunofluorescence confocal micrographs shown here show individual E2-72kDa (green) and RNF8 or 53BP1 (red) staining, merged staining patterns, and phase contrast images of uninfected or Ad5 and E4- infected cells.

Figure 18. Failure to localize pATM to replication centers results in inefficient recruitment of 53BP1. (A) HeLa cells were uninfected, or infected with Ad or E4- and KU60019 (5µM) or DMSO was added to the medium after 2 h. Immunofluorescence confocal micrographs showing E2-72kDa (green) and 53BP1 (red) of cells uninfected or infected with Ad5 and E4- at 17 hpi. (B) HeLa cells were uninfected or infected with 1010 for 17 h and cells were prepared for immunostaining with anti-72K and anti-53BP1 antibodies. Immunofluorescence confocal micrographs showing E2-72kDa (green) and 53BP1 (red) of cells uninfected or infected with 1010 at 17 hpi. (C) Nbs1+, Nbs1- and Nbs1∆ATM cells were either uninfected or infected with E4- or Ad5 viruses at 3 FFU/cell and fixed for immunofluorescence at 17 hpi. Immunofluorescence micrographs showing E2-72kDa (green) and 53BP1 (red) in uninfected cells or cells infected with either Ad5 or E4-.

18B panels d to f show that 53BP1 is not localized with the well-developed DNA replication centers that were previously shown to lack pATM (Figure 14B, panels m to p). Since pATM staining was markedly reduced in E4- DNA replication centers in Nbs1∆ATM cells infected with E4-, we next assayed the localization of 53BP1 relative to E4- DNA replication centers in these cells. Nbs1+, Nbs1- and Nbs1∆ATM cells were infected with Ad5 and E4-, and at 17 hpi cells were fixed and immunostained with anti-72K and anti-53BP1 antibodies. Confocal micrographs are shown in Figure 18C. 53BP1 was not localized with replication centers in Ad5 infected cells (Figure 18C, all cells, panels d to f). In Nbs1+ cells infected with E4-, we observed robust localization of 53BP1 to poorly developed replication centers (Figure 18C Nbs1+, panels g to i); however, in Nbs1- cells 53BP1 was not detected with replication centers (Figure 18C Nbs1-, panels g to i). During E4- infections in Nbs1∆ATM cells, localization of 53BP1 to replication centers were markedly reduced compared to Nbs1+ cells (Figure 18C Nbs1∆ATM, panels g to i). This is consistent with intermediate phenotype of these cells in ATM activation and recruitment to DNA replication centers (Figure 13). The data indicate that 53BP1 recruitment to viral DNA replication centers requires active pATM that is localized to replication centers. Since efficient localization of activated pATM to viral DNA replication centers is important for inhibiting viral DNA replication (Figures 12-14) and for recruiting late DDR to viral DNA replication centers, our data imply that pATM-mediated recruitment of late DDR proteins may contribute to the inhibition of E4- DNA replication.

Discussion

We have previously reported that activated pATM interferes with Ad E4 mutant DNA replication (Gautam and Bridge, 2013). However the mechanism by which ATM inhibits viral DNA replication is still unclear. Activated pATM could inhibit E4 mutant DNA replication by physically localizing to the viral DNA. Indeed, previous work has shown that pATM is localized to E4 mutant viral genomes (Karen and Hearing, 2011) as well as viral DNA replication centers (Carson et al., 2003). In response to IR, ATM is activated and recruited to DSBs by interacting with the C-terminal ATM binding motif in the Nbs1 protein (Cerosaletti et al., 2006). We found that E4 mutant DNA levels were significantly rescued and pATM localization to viral replication centers was reduced in cells expressing Nbs1 with a deletion of the C-terminal ATM binding motif (Figures 12 and 13). This suggests that the Nbs1/pATM interaction and efficient recruitment of pATM to replication centers is important for inhibiting E4- DNA replication. In 1010 infections the E4 11kDa protein is produced and can redistribute Mre11 to nuclear tracks such that it fails to localize with viral replication centers (Figure 14, panels m to p). Mislocalization of Mre11 does not prevent ATM activation in response to an Ad infection (Carson et al., 2009; Prakash and Bridge, unpublished data) but interestingly, we find that it does prevent the sustained localization of activated pATM at 1010 DNA replication centers (Figure 14). Furthermore, loss of pATM localization correlates with robust growth of 1010 replication centers (Figure 14B). Our data are consistent with a model in which the MRN complex is important both for activating pATM and localizing it to viral DNA replication centers. These roles are critical for the ability of pATM to inhibit E4- DNA replication, and support a model where the ATM kinase inhibits E4- DNA replication when it is physically localized to DNA replication centers.

Although E4- DNA replication in Nbs1∆ATM was significantly rescued compared to Nbs1+ cells, DNA levels were still less than observed in Nbs1- cells (Figure 12B). We think it is likely that other Nbs1 domains are involved in leaky activation of ATM when the C-terminal ATM binding motif is missing. Previous work has indicated lower levels of pATM activation in Nbs1∆ATM cells in response to IR (Cerosaletti and Concannon, 2003; Zhao et al., 2002). This is supported by the observation that pATM can be detected at E4- DNA replication centers (Figure 13B) and that the E4- DNA replication increased further when Nbs1∆ATM cells were treated with ATM kinase inhibitor, KU60019 (Figure 12C). These data suggest that in Nbs1∆ATM cells,

E4- DNA replication is only partially inhibited due to leaky activation of ATM and its localization to viral DNA replication centers. This likely explains why Nbs1∆ATM cells did not rescue E4- DNA replication to the same extent as Nbs1- cells.

We and others have proposed that recruitment of DDR proteins to the E4 mutant genomes could inhibit viral DNA replication by masking the origin of replication (Mathew and Bridge, 2007; Karen and Hearing 2011). The MRN complex along with Mdc1, γH2AX, and pATM are among the early DDR proteins that localize to sites of DNA damage (Celeste et al., 2003; Lisby et al., 2004). Mre11, Mdc1 and γH2AX are still efficiently recruited to E4 mutant DNA replication centers in the absence of ATM (Figure 15). This result is somewhat surprising since ATM is thought to be important for recruiting γH2AX and Mdc1 to DSBs following IR, but previous work has shown some redundancy during the DDR and other DDR kinases may be able to phosphorylate H2AX and recruit it and Mdc1 to DNA damage sites in the absence of ATM (Giunta et al., 2010). Mre11 and unphosphorylated ATM are still recruited to viral replication centers in the presence of the ATM kinase inhibitor KU60019 (Gautam and Bridge, 2013). Since E4 mutant DNA replication is rescued in ATM- cells and in the presence of ATM kinase inhibitors (Gautam and Bridge, 2013), our data indicate that recruitment of these early DDR proteins is not sufficient to inhibit viral DNA replication. In addition, siRNA-mediated knockdown of Mdc1 did not rescue E4 mutant DNA replication (Mathew and Bridge, 2007), providing further support for the idea that the presence of early DDR proteins at viral replication centers is not sufficient to prevent DNA replication.

Activated pATM could inhibit E4 mutant DNA replication by recruiting additional downstream DDR factors to viral replication centers. ATM-mediated phosphorylation of Mdc1 recruits RNF8 to DSBs, which in turn adds ubiquitin molecules to γH2AX to facilitate the recruitment of 53BP1 (Mailand et al., 2007). We found that RNF8 and 53BP1 are co-localized with pATM throughout E4- infection (Figure 16), consistent with the idea that pATM is important for localization of these proteins at DNA damage sites. Indeed, RNF8 and 53BP1 failed to localize with E4- DNA replication centers in the absence of ATM (Figure 17) or its kinase activity (Figure 18A). Infection conditions that rescued E4 mutant DNA replication also resulted in decreased 53BP1 at DNA replication centers. In 1010 infected cells neither pATM (Figure 14B) nor 53BP1 (Figure 18B) were recruited to well-developed DNA replication centers that form when E4-11kDa is expressed (Figure 18B). Similarly, we also observed reduced

53BP1 localization to E4- DNA replication centers during Nbs1∆ATM cells and following the treatment of HeLa cells with the ATM kinase inhibitor KU60019 (Figure 18A). Our data suggest that failure to recruit pATM to replication centers results in inefficient recruitment of 53BP1 to the replication centers. It is possible that ATM kinase-dependent recruitment of 53BP1 or other late DDR factors could interfere with viral DNA replication, and thereby prevent the normal development of viral replication centers.

The mechanism through which ATM interferes with E4- DNA replication is still unclear. Recent investigations have highlighted the importance of ATM in a multitude of chromatin changes that occur at DSBs including ubiquitination, acetylation and methylation of histones (reviewed in Price and D’Andrea, 2013; Dinant et al., 2008). These changes allow local relaxation of the chromatin at DSBs thereby allowing DDR factors to access the broken DNA. Our data show that ATM is required for recruitment of RNF8 and 53BP1 to E4- replication centers (Figures 17 and 18A). RNF8 is a bona fide histone modifier that adds ubiquitin to histones. Moreover, 53BP1 retention at DSBs requires DDR-dependent methylation of histones (Botuyan et al., 2006; Huyen et al., 2004). Ad genomes interact with cellular histones during infection and are organized in nucleosome like structures (Sergeant et al., 1979). It is possible that the E4- genome is subjected to chromatin modification during DDR induction that could make it a less favorable substrate for the viral DNA polymerase and subsequently inhibit DNA replication. This model is supported by the observation that RNF8 and 53BP1 are detected with pATM only at poorly developed E4- replication centers. When we interfere with activation or localization of pATM, this prevents RNF8 and 53BP1 localization to DNA replication centers, and correlates with their robust development. In conclusion, we have found that the presence of several early DDR proteins at DNA replication centers is not sufficient to interfere with viral DNA replication. In contrast, localization of late DDR proteins that either modify chromatin (RNF8) or are dependent on the presence of ATM-mediated chromatin modifications (53BP1) (Murr et al., 2006; Xu et al., 2010) in viral DNA replication centers, is correlated with poor development of replication centers and reduced levels of viral DNA. Our data raise the possibility that ATM-mediated chromatin changes contribute to inhibition of E4- DNA replication. Further investigations will be needed to determine if specific ATM dependent changes in viral chromatin are responsible for reduced E4 mutant DNA replication following induction of the cellular DDR.

Concluding remarks

DNA Tumor Viruses

DNA tumor viruses (DTVs) serve as a powerful yet simple tool to study carcinogenesis in animal cells. DTVs have evolved strategies to replicate their genomes independent of the normal restrictions imposed on the cell cycle. Therefore, DTVs are capable of overriding many of the intracellular as well as extracellular mechanisms that control cellular DNA replication, and in doing so they often target common regulatory pathways. For instance the tumor suppressor Retinoblastoma (Rb) negatively regulates the cell cycle (Felsani et al., 2006), and is targeted by several DTVs to drive infected cells into S-phase thereby allowing viruses to utilize factors which would otherwise be unavailable in other phases of cell cycle. p53 is a tumor suppressor involved in promoting apoptosis, or programmed cell death (Zuckerman et al., 2009). Virus infection often triggers apoptotic pathways, and such an event could be detrimental for a productive virus life cycle. DTVs inactivate p53 and benefit by preventing premature cell death in response to infection (Lilley et al., 2007). Like many DTVs, Ad expresses oncoproteins from early region 1 (E1) and E4 that misregulate Rb and p53 and can transform host cells. Rb and p53 are mutated in many cancers allowing cells to undergo uncontrolled cellular proliferation that results in tumor development. Understanding the function of these cellular tumor suppressor proteins has been greatly facilitated by studies with DTV oncoproteins that can transform cells.

Virus interactions with the cellular DNA damage response

The DDR is another critical tumor suppressing pathway that responds to replicative stress and DNA damage arising from endogenous as well as exogenous sources. DTVs have the potential to activate DDRs by the delivery of exogenous nucleic acid, the production of replication intermediates, or during viral gene expression. Viral genomes and replicative intermediates can be perceived by the cell as aberrant structures, thereby triggering DDR activation. In addition, expression of viral oncoproteins can activate the cell cycle in quiescent cells resulting in cellular replicative stress that can lead to stalled replication forks that induce DSBs and subsequently activate DDRs. DDR activation can have beneficial or catastrophic consequences for the virus, depending on the specific virus-host interaction. For instance DDR activation during Ad infection repairs the viral genomes and forms genome concatemers that cannot be packaged into the virions. On the other hand DDR activation is shown to be beneficial

for DNA replication in certain DTVs. In the case of HPV, activation of ATM dependent pathways promotes viral genome amplification (Moody & Laimins, 2009). Therefore, DTVs have evolved strategies to activate or inactivate DDR pathways (Weitzman et al., 2010).

DDR activation can act as an innate antiviral response against viruses such as Ad, which activates DDRs that have a detrimental effect on their life cycle (Lilley et al., 2007). These viruses produce regulatory proteins that interfere with the DDR to prevent its activation. Ad E4 proteins inactivate DDR by degrading or mislocalizing critical DDR proteins (Stracker et al., 2002). HPV-16 inactivates ATR signaling by targeting its activating factor TopBP1, for mislocalization (Boner et al., 2002). The HSV-1 early protein ICP0 inactivates ATR pathways by mislocalizing ATR interacting partner (ATRIP) (Wilkinson and Weller, 2006). John Cunningham virus (JCV) inhibits NHEJ by targeting the sensor protein Ku70 for mislocalization (Darbinyan et al., 2004). Such viral regulatory proteins can potentially be developed as tools for manipulating the DDR and be used as therapeutics. Virus infection and specific viral oncoproteins have shown promise for targeted killing of cancer cells. For example, telomerase- dependent oncolytic adenovirus inactivates the MRN complex and ATM, thereby radiosensitizing cancer cells such that they are killed more readily by radiation treatment (Kuroda et al., 2010). Herpes simplex virus (HSV) in combination with radio- or chemo-therapy is extensively used as an oncolytic agent for glioma therapy (Advani et al., 2006). Therefore, the knowledge of virus interactions with the DDR will be useful in developing treatments against both virus infection as well as various cancers.

Activation of ATM during Adenovirus infection

The interaction of the MRN complex protein Nbs1 with ATM via its C-terminal motif has been implicated in the efficient activation of ATM (Cerosaletti and Concannon, 2004). We have also found that the Nbs1 C-terminal motif is important for efficient ATM activation following E4 mutant infection (Chapter II). However, the specific mechanism involved in the activation is still debated. Various models have been proposed for ATM activation following induction of DSBs, including structural changes in the DNA at DSBs (Bakkenist and Kastan, 2003), as well as the direct contact of ATM with damaged DNA (Shiotani and Zou, 2009). Similarly, the mechanism of ATM activation in Ad infection is not clear. During Ad infection, the viral genome initially delivered to the nucleus is coated with core DNA binding proteins that

interact with the viral genome inside the virion. Although injection of linear plasmid in the nucleus has been shown to activate DDRs (Huang et al., 1996), the Ad ds linear genome does not immediately activate DDRs (Karen and Hearing, 2011). Expression of the E1A gene from Ad was shown to be sufficient to activate ATM (Cuconati et al., 2003). Recent work from the Hearing lab suggests that transcription-mediated remodeling of the E4 mutant genome during early gene expression is needed to trigger activation of pATM (Karen and Hearing, 2011). These results suggest a model in which chromatin remodeling could possibly evict viral core proteins from the DNA and subsequently expose free DNA ends that are targeted by the MRN complex. Activation of ATM at early times after infections is also not limited to E4 mutants; we see pATM is localized in foci in infections with Ad5 as well as other Ad mutants (Gautam and Bridge, 2013, Prakash and Bridge, unpublished data). Since these mutants undergo early gene expression, these observations support the idea that chromatin remodeling of the genome during early gene expression could be an important trigger for ATM activation during Ad infection. Other viruses can activate ATM by direct interactions. For example, the HPV E7 protein binds ATM and this is sufficient for ATM-mediated Chk2 activation (Moody and Laimins, 2009). Understanding how a variety of viruses activate and utilize ATM will continue to provide key insights into our understanding of ATM dependent DDR pathways.

Impact of localization of DNA damage response proteins on viral DNA replication

The MRN complex and γH2AX are localized to several viral replication centers, including HPV and HSV-1, and are thought to stimulate DNA replication of these viruses (reviewed in Turnell and Grand, 2012). Furthermore, γH2AX was found to be associated with the HPV genome (Gillespie et al., 2012). Although the mechanism by which MRN and γH2AX contribute to replication is not fully understood, localization of these proteins to viral replication centers was not detrimental. Furthermore, localization of these proteins to replication centers is seen in viruses that activate ATM suggesting that recruitment of these DDR proteins to viral DNA may be a mechanism of ATM activation.

In contrast, Ad5 degrades Mre11 and prevents recruitment of γH2AX to DNA replication centers, and the MRN complex recruitment to DNA replication centers has been implicated in inhibiting DNA replication of E4 mutants (Evans and Hearing, 2005; Mathew and Bridge, 2007; 2008; Lakdawala et al., 2008), as well as ATM activation (Gautam and Bridge, 2013). Although

localization of the MRN complex in E4 mutant DNA replication centers is important for inhibiting viral DNA replication, we found that localization of Mre11, Mdc1 and γH2AX at E4 mutant DNA replication centers was not sufficient to inhibit DNA replication (Chapter II). Rather, we find that a functional MRN complex is required for ATM activation and its localization to DNA replication centers (Gautam and Bridge, 2013). These observations suggest that the MRN complex plays a critical role in inhibiting E4 mutant DNA replication through its ability to activate ATM and recruit it to DNA replication centers. E4 mutant DNA replication is inhibited as a consequence of pATM activation and localization to E2-72kDa containing replication centers (Gautam and Bridge, 2013). Furthermore, in the presence of ATM kinase inhibitors we found ATM was still recruited to E4 mutant DNA replication centers even though DNA replication was rescued (Gautam and Bridge, 2013). This observation suggests that recruitment of unphosphorylated ATM did not inhibit DNA replication. The ability of pATM to inhibit DNA replication also depends on its localization to E4 mutant DNA replication centers (Chapter II). Taken together, these results suggest that localization of Mre11, Mdc1 and γH2AX is not itself detrimental to Ad DNA replication. Rather the MRN complex is required to recruit and activate ATM, which then inhibits Ad DNA replication.

ATM is known to phosphorylate viral proteins including SV40 Large T antigen at Ser120 (Shi et al., 2005; Dahl et al., 2005) and in this case ATM-mediated phosphorylation stimulates DNA replication and promotes a lytic life cycle. Ad replication proteins, including E2-72kDa, Ad DNA polymerase and terminal protein, are also phosphoproteins and their functions are regulated by phosphorylation. It is possible that ATM or another kinase activated by ATM could phosphorylate one or more of these viral proteins rendering them less effective for replication. Bioinformatics analysis (GPS 2.1; Netphos 2.0) show that these proteins are potentially subjected to phosphorylation at serine/threonine residues by ATM; for instance, the Ad pre-terminal protein shows predicted ATM phosphorylation sites at Ser113 and 446. ATM could phosphorylate these proteins either at replication centers or elsewhere in the nucleus when the proteins are not actively engaged in DNA replication. However, in infections with E4 mutant 1010, pATM is active (Prakash and Bridge, unpublished data) but not localized to DNA replication centers (Chapter II). Since 1010 DNA replication centers are well developed (Chapter II), this suggests that pATM only interferes with viral DNA replication when it is present at replication centers. If ATM interferes with viral DNA replication by phosphorylating DNA

replication proteins, it is likely that these phosphorylation events occur at the sites of DNA replication, where the viral proteins would be in a close proximity to pATM.

Several viruses activate ATM and recruit pATM to their replication centers. For instance, ATM is required for HSV-1 DNA replication; however, HSV-1 interferes with activation of downstream ATM-dependent DDR pathways by targeting the histone modifying ubiquitin ligase proteins, RNF8 and RNF168, for degradation via the early regulatory protein ICP0 (Chaurushiya et al., 2012). Although ICP0 is implicated in many functions during HSV-1 infection, RNF8 knockdown alone promoted ICP0-null HSV-1 growth, highlighting its importance for promoting efficient HSV-1 infection. Recruitment of RNF8 and RNF168 correlates with ubiquitination of histones and formation of a repressive chromatin structure that silences transcription at DSBs (Shanbhag et al., 2010). Formation of repressive chromatin on foreign DNA such as viral genomes could be an anti-viral response. We observe that RNF8 is localized to Ad E4 mutant DNA replication centers in an ATM-dependent manner and this localization was correlated with the poor development of DNA replication centers (Chapter II). Therefore it is possible that pATM inhibits Ad DNA replication by recruiting the RNF8 ubiquitin ligase.

The Ad genome is subjected to chromatin modification during DNA damage response activation

The first indication that the Ad genome is subjected to chromatin changes during DDR activation comes from the observation that γH2AX localizes to E4 mutant viral replication centers (Carson et al., 2003) and likely at Ad genomes prior to the onset of DNA replication (Prakash and Bridge, unpublished work). We have shown that RNF8 is recruited to E4 mutant DNA replication in an ATM-dependent manner (Chapter II). RNF8 is a ring finger-containing E3 ubiquitin ligase that modifies histones in the vicinity of DSBs, and has been implicated in recruitment of other DDR proteins including 53BP1 and BRCA1 (reviewed in Xu and Price, 2011). HSV-1 degrades RNF8 and RNF168 and although the implication of this degradation is not yet understood, it raises the possibility that recruitment of these chromatin modifying enzymes to viral replication centers may have undesirable effect on virus life cycle. ATM activation is also important for recruiting Tip60 to DSBs, where it is involved in coordinating DDRs (reviewed in Sun et al., 2010). Tip60 is an acetyltransferase that plays a critical role in several signaling pathways including the DDR, and is a part of chromatin remodeling complex

that acetylates histones at DSBs (Sun et al., 2010). Tip60 acetylates H2AX present at DSBs (Kusch et al., 2004). Gupta et al (2013) have recently shown that Tip60 is targeted for proteasome-mediated degradation during Ad infection. Knockdown of Tip60 resulted in an increase in wild type Ad5 DNA replication (Gupta et al., 2013) suggesting what that Tip60 plays a critical role in inhibiting DNA replication. Additionally, death associated protein (Daxx) is targeted for degradation during Ad infection and knockdown of Daxx stimulated DNA replication (Schreiner et al., 2010). Daxx is a histone chaperone that is part of a chromatin remodeling complex (Xue et al., 2003). Recently it has been shown that Daxx is an ATM substrate and is phosphorylated at Ser564 in DDR dependent manner (Tang et al., 2013). Although, the role of Tip60 and Daxx in E4 mutant DNA replication is not yet known, the ability of these proteins to affect viral DNA replication and their regulation by ATM, supports the idea that activated ATM might interfere with viral DNA replication through its effects on chromatin modifying enzymes.

Models for pATM-mediated inhibition of Ad E4 mutant DNA replication

My interest in the role of DSBR pathways in interfering with E4 mutant DNA replication started from the observation that cells that lack functional MRN complexes rescued E4 mutant DNA replication (Mathew and Bridge, 2007; 2008; Lakdawala et al., 2008). Chapters I and II identify a role for activated pATM and its localization in viral DNA replication centers in this process. I propose the following model to explain possible roles of ATM in interfering with viral DNA replication, based on work presented in this dissertation as well as observations from other studies.

Ad genomes are delivered to the nucleus of the host cell in the form of a linear ds DNA coated by the core proteins pVII and pV. At this stage the genome does not trigger DDR activation. The incoming viral genome undergoes transcription-dependent remodeling during the expression of viral early genes resulting in a more cell-like chromatin configuration containing histones. It is possible that this process exposes the viral genomic DNA ends to the MRN sensor complex. The MRN complex senses these ends as potential DNA breaks and activates ATM. Therefore, we see pATM in foci containing viral genomes (Karen and Hearing, 2011) at early times after infection (Karen and Hearing, 2011; Gautam and Bridge, 2013). The early pATM foci are seen in both Ad5 and E4 mutant infections and do not depend on the absence of E4

proteins. However, once early genes including E1 and E4 are expressed during Ad infections, Mre11 is inactivated. Inactivation of the MRN complex occurs prior to the onset of DNA replication (Karen et al., 2009) and correlates with loss of ATM activation (Carson et al., 2003). Ad5 can thus proceed normally with DNA replication. In contrast, an E4 deletion mutant is not able to inactivate the MRN complex, resulting in sustained pATM activation. pATM could inhibit E4 mutant DNA replication in one of the three possible ways (Figure 19). In pathway I, pATM recruitment to E4 mutant genomes could be important for recruitment of DDR proteins that are involved in chromatin modification and typically cause relaxation of DNA at sites of DSBs (e.g. RNF8). E4 mutant genomes could be subject to similar chromatin modifications and altered chromatin structure could affect initiation and/or synthesis by the viral DNA replication machinery, resulting in a slower rate of DNA synthesis and a non-productive infection. Alternatively (pathway II), ATM-dependent chromatin changes at E4 mutant genome recruits late DDR proteins such as RNF8 and 53BP1. These DDR proteins may obscure the viral origin of replication and hinder its access to viral replication machinery resulting in reduced DNA replication and a non-productive infection. Finally (pathway III), ATM may phosphorylate viral or cellular proteins involved in the DNA replication process (e.g. pre-terminal protein), making it less efficient for DNA replication and blocks viral replication. Since the MRN complex is at the apex of the ATM-mediated DDR pathway, it is a reasonable strategy for Ad to target Mre11 for inactivation, which in turn interferes with sustained activation of ATM. Furthermore, Ad5 targets Tip60 for degradation via the same machinery that targets Mre11. In addition to the activation pf ATM, Tip60 possibly contributes to chromatin remodeling of the viral genome, thus repressing viral DNA replication.

Adenovirus as a model system to study DNA damage response

Our lab uses Ad as a model system to understand DDR pathways. My work using Ad indicates that activated ATM inhibits E4 mutant DNA replication (Chapter I). One can envision a similar scenario where ATM could inhibit cellular DNA replication in the vicinity of DSBs. Upon induction of DSB, cell cycle check points are activated to prevent progression to S phase. However, these events may take some time, and it could be detrimental to risk replicating cellular DNA through the DSB. If pATM inhibits cellular DNA replication at DSBs by a similar mechanism through which it inhibits E4 mutant DNA replication, our work can potentially provide significant insights into this process. Recent work suggests that pATM can inhibit

Figure 19. Models for ATM mediated inhibition of E4 mutant DNA replication. See text for explanation.

transcription at DSBs and this involves histone ubiquitination by RNF8 and RNF168 (Shanbhag et al., 2010). It is possible that a similar mechanism may be involved in pATM mediated inhibition of DNA replication. My work suggests that pATM is involved in chromatin changes (Chapter II). This opens up new possibilities of using Ad as a model system to understand the impact of ATM mediated chromatin changes on viral and possibly cellular DNA replication, and investigating the specific roles of DDR proteins in this process.

References

1) Advani, S. J., Mezhir, J. J., Roizman, B., & Weichselbaum, R. R. (2006). ReVOLT: radiation-enhanced viral oncolytic therapy. International Journal of Radiation Oncology Biology Physics, 66(3), 637-646.

2) Asaad, N. A., Zeng, Z. C., Guan, J., Thacker, J., & Iliakis, G. (2000). Homologous recombination as a potential target for caffeine radiosensitization in mammalian cells: Reduced caffeine radiosensitization in XRCC2 and XRCC3 mutants. Oncogene, 19(50), 5788. 3) Baker, A., Rohleder, K. J., Hanakahi, L. A., & Ketner, G. (2007). Adenovirus E4 34k and E1b 55k oncoproteins target host DNA ligase IV for proteasomal degradation. Journal of Virology, 81(13), 7034-7040. 4) Bakkenist, C. J., & Kastan, M. B. (2003). DNA damage activates ATM through intermolecular autophosphorylation and dimer dissociation. Nature, 421(6922), 499-506. 5) Bakkenist, C. J., & Kastan, M. B. (2004). Initiating cellular stress responses. Cell, 118(1), 9-17. 6) Bekker-Jensen, S., & Mailand, N. (2010). Assembly and function of DNA double-strand break repair foci in mammalian cells. DNA repair, 9(12), 1219-1228. 7) Berget, S. M., Moore, C., & Sharp, P. A. (1977). Spliced segments at the 5' terminus of adenovirus 2 late mRNA. Proceedings of the National Academy of Sciences, 74(8), 3171- 3175. 8) Berkovich, E., Monnat, R. J., & Kastan, M. B. (2007). Roles of ATM and NBS1 in chromatin structure modulation and DNA double-strand break repair. Nature Cell Biology, 9(6), 683-690. 9) Blackford, A. N., & Grand, R. J. (2009). Adenovirus E1B 55-kilodalton protein: multiple roles in viral infection and cell transformation. Journal of virology, 83(9), 4000-4012. 10) Blanchette, P., Cheng, C. Y., Yan, Q., Ketner, G., Ornelles, D. A., Dobner, T., Conaway, R. C., Conaway, J. W., & Branton, P. E. (2004). Both BC-box motifs of adenovirus protein E4orf6 are required to efficiently assemble an E3 ligase complex that degrades p53. Molecular and Cellular Biology, 24(21), 9619-9629.

11) Block, W. D., Merkle, D., Meek, K., & Lees‐Miller, S. P. (2004). Selective inhibition of the DNA‐dependent protein kinase (DNA‐PK) by the radiosensitizing agent caffeine. Nucleic Acids Research, 32(6), 1967-1972. 12) Boner, W., & Morgan, I. M. (2002). Novel cellular interacting partners of the human papillomavirus 16 transcription/replication factor E2. Virus Research,90 (1), 113-118. 13) Botuyan, M. V., Lee, J., Ward, I. M., Kim, J. E., Thompson, J. R., Chen, J., & Mer, G. (2006). Structural basis for the methylation state-specific recognition of histone H4-K20 by 53BP1 and Crb2 in DNA repair. Cell, 127(7), 1361-1373. 14) Boyer, J., Rohleder, K., & Ketner, G. (1999). Adenovirus E4 34k and E4 11k inhibit double strand break repair and are physically associated with the cellular DNA-dependent protein kinase. Virology, 263(2), 307-312. 15) Bradbury, J. M., & Jackson, S. P. (2003). The complex matter of DNA double-strand break detection. Biochemical Society Transactions, 31(1), 40-44. 16) Branton, P. E. (1999). Early Gene Expression. Adenoviruses: from basic research to gene therapy applications, Medical Intelligence Unit, 15, 39-58. 17) Bridge, E., & Ketner, G. (1989). Redundant control of adenovirus late gene expression by early region 4. Journal of virology, 63(2), 631-638. 18) Brown, D. T., Westphal, M., Burlingham, B. T., Winterhoff, U., & Doerfler, W. (1975). Structure and composition of the adenovirus type 2 core. Journal of Virology, 16(2), 366- 387. 19) Carson, C. T., Orazio, N. I., Lee, D. V., Suh, J., Bekker-Jensen, S., Araujo, F. D., Lakdawala, S. S., Lilley, C. E., Bartek, J., Lukas, J., & Weitzman, M. D. (2009). Mislocalization of the MRN complex prevents ATR signaling during adenovirus infection. The EMBO Journal, 28(6), 652-662. 20) Carson, C. T., Schwartz, R. A., Stracker, T. H., Lilley, C. E., Lee, D. V., & Weitzman, M. D. (2003). The Mre11 complex is required for ATM activation and the G2/M checkpoint. The EMBO Journal, 22(24), 6610-6620. 21) Celeste, A., Fernandez-Capetillo, O., Kruhlak, M.J., Pilch, D.R., Staudt, D.W., & Lee, A. (2003). Histone H2AX phosphorylation is dispensable for the initial recognition of DNA breaks. Nature Cell Biology, 5(7), 675-679.

22) Cerosaletti, K. M., & Concannon, P. (2003). Nibrin forkhead-associated domain and breast cancer C-terminal domain are both required for nuclear focus formation and phosphorylation. Journal of Biological Chemistry, 278(24), 21944-21951.

23) Cerosaletti, K., & Concannon, P. (2004). Independent roles for nibrin and Mre11-Rad50 in the activation and function of ATM. Journal of Biological Chemistry, 279(37), 38813- 38819.

24) Cerosaletti, K., Wright, J., & Concannon, P. (2006). Active role for nibrin in the kinetics of atm activation. Molecular and Cellular Biology, 26(5), 1691-1699. 25) Challberg, M. D., Kelly, T. J. (1989). Animal virus DNA replication. Annual Review of Biochemistry, 58(1), 671-713.

26) Chaurushiya, M. S., Lilley, C. E., Aslanian, A., Meisenhelder, J., Scott, D. C., Landry, S., Ticau, S., Boutell, C., Yates, J. R., Schulman, B. A., Hunter, T & Weitzman, M. D. (2012). Viral E3 ubiquitin ligase-mediated degradation of a cellular E3: viral mimicry of a cellular phosphorylation mark targets the RNF8 FHA domain. Molecular Cell, 46(1), 79-90.

27) Cheng, C.Y., Blanchette, P., & Branton, P. E. (2007). The adenovirus E4orf6 E3 ubiquitin ligase complex assembles in a novel fashion. Virology, 364(1), 36-44.

28) Cuconati, A., Mukherjee, C., Perez, D., & White, E. (2003). DNA damage response and MCL-1 destruction initiate apoptosis in adenovirus-infected cells. Genes & Development, 17(23), 2922-2932.

29) Dahl, J., You, J., & Benjamin, T. L. (2005). Induction and utilization of an ATM signaling pathway by polyomavirus. Journal of Virology, 79(20), 13007-13017.

30) Dallaire, F., Blanchette, P., Groitl, P., Dobner, T., & Branton, P. E. (2009). Identification of integrin α3 as a new substrate of the adenovirus E4orf6/E1B 55-kilodalton E3 ubiquitin ligase complex. Journal of Virology, 83(11), 5329-5338. 31) Darbinyan, A., White, M. K., Akan, S., Radhakrishnan, S., Del Valle, L., Amini, S., & Khalili, K. (2007). Alterations of DNA damage repair pathways resulting from JCV infection. Virology, 364(1), 73-86.

32) Davison, A. J., Benkő, M., & Harrach, B. (2003). Genetic content and evolution of adenoviruses. Journal of General Virology, 84(11), 2895-2908. 33) Dinant, C., Houtsmuller, A., & Vermeulen, W. (2008). Chromatin structure and DNA damage repair. Epigenetics & Chromatin, 1(1), 9. 34) DiTullio, R. A., Mochan, T. A., Venere, M., Bartkova, J., Sehested, M., Bartek, J., & Halazonetis, T. D. (2002). 53BP1 functions in an ATM-dependent checkpoint pathway that is constitutively activated in human cancer. Nature Cell Biology, 4(12), 998-1002. 35) Doil, C., Mailand, N., Bekker-Jensen, S., Menard, P., Larsen, D. H., Pepperkok, R., Ellenberg J., Panier S., Durocher D., Bartek J., Lukas J., & Lukas, C. (2009). RNF168 binds and amplifies ubiquitin conjugates on damaged chromosomes to allow accumulation of repair proteins. Cell, 136(3), 435-446. 36) Dupré, A., Boyer-Chatenet, L., & Gautier, J. (2006). Two-step activation of ATM by DNA and the Mre11-Rad50-Nbs1 complex. Nature Structural & Molecular Biology, 13(5), 451-457. 37) Echavarria, M. (2008). Adenoviruses in immunocompromised hosts. Clinical Microbiology Reviews, 21(4), 704-715. 38) Evans, J. D., & Hearing, P. (2002). Adenovirus replication. Adenoviral Vectors for Gene Therapy, 39-70. 39) Evans, J. D., & Hearing, P. (2003). Distinct roles of the adenovirus E4 ORF3 protein in viral DNA replication and inhibition of genome concatenation. Journal of Virology, 77(9), 5295-5304. 40) Evans, J.D., & Hearing, P., (2005). Relocalization of the Mre11-Rad50-Nbs1 complex by the adenovirus E4 ORF3 protein is required for viral replication. Journal of Virology, 79(10), 6207-6215. 41) Falck, J., Coates, J., & Jackson, S. P. (2005). Conserved modes of recruitment of ATM, ATR and DNA-PKcs to sites of DNA damage. Nature, 434(7033), 605-611. 42) Felsani, A., Mileo, A. M., & Paggi, M. G. (2006). Retinoblastoma family proteins as key targets of the small DNA virus oncoproteins. Oncogene, 25(38), 5277-5285. 43) Fernandez-Capetillo, O., Chen, H. T., Celeste, A., Ward, I., Romanienko, P. J., Morales, J. C., Naka K., Xia Z., Camerini-Otero R. D., Motoyama N., Carpenter P. B., Bonner W.

M., Chen J., & Nussenzweig, A. (2002). DNA damage-induced G2–M checkpoint activation by histone H2AX and 53BP1. Nature Cell Biology, 4(12), 993-997. 44) Gautam, D., & Bridge, E. (2013). The kinase activity of ATM interferes with adenovirus E4 mutant DNA replication. Journal of Virology, 87(15), 8687-96 45) Giberson, A. N., Davidson, A. R., & Parks, R. J. (2012). Chromatin structure of adenovirus DNA throughout infection. Nucleic Acids Research, 40(6), 2369-2376.

46) Gillespie, K. A., Mehta, K. P., Laimins, L. A., & Moody, C. A. (2012). Human papillomaviruses recruit cellular DNA repair and homologous recombination factors to viral replication centers. Journal of Virology, 86(17), 9520-9526.

47) Giunta, S., Belotserkovskaya, R., & Jackson, S. P. (2010). DNA damage signaling in response to double-strand breaks during mitosis. The Journal of Cell biology, 190(2), 197-207. 48) Golding, S. E., Rosenberg, E., Valerie, N., Hussaini, I., Frigerio, M., Hussain,i I., Frigerio, M., Cockcroft, X. F., Chong, W. Y., Hummersone, M., Rigoreau, L., Menear, K. A., O'Connor, M. J., Povirk, L. F., van Meter, T., & Valerie, K. (2009). Improved ATM kinase inhibitor KU-60019 radiosensitizes glioma cells, compromises insulin, AKT and ERK prosurvival signaling, and inhibits migration and invasion. Molecular Cancer Therapeutics, 8(10), 2894-2902. 49) Graham, F. L., Smiley, J., Russell, W. C., & Nairn, R. (1977). Characteristics of a human cell line transformed by DNA from human adenovirus type 5. Journal of General Virology, 36(1), 59-72. 50) Greber, U. F., Willetts, M., Webster, P., & Helenius, A. (1993). Stepwise dismantling of adenovirus 2 during entry into cells. Cell, 75(3), 477-486. 51) Gupta, A., Jha, S., Engel, D. A., Ornelles, D. A., & Dutta, A. (2012). Tip60 degradation by adenovirus relieves transcriptional repression of viral transcriptional activator EIA. Oncogene, Epub ahead of print. 52) Halbert, D. N., Cutt, J.R., & Shenk, T. (1985). Adenovirus early region 4 encodes functions required for efficient DNA replication, late gene expression, and host cell shutoff. Journal of Virology, 56(1), 250-257.

53) Harada, J. N., Shevchenko, A., Shevchenko, A., Pallas, D. C., & Berk, A. J. (2002). Analysis of the adenovirus E1B-55K-anchored proteome reveals its link to ubiquitination machinery. Journal of Virology, 76(18), 9194-9206. 54) Harper, J. W., & Elledge, S. J. (2007). The DNA damage response: ten years after. Molecular Cell, 28(5), 739-745. 55) Hopfner, K. P., & Tainer, J. A. (2003). Rad50/SMC proteins and ABC transporters: unifying concepts from high-resolution structures. Current Opinion in Structural Biology, 13(2), 249-255. 56) Hopfner, K. P., Craig, L., Moncalian, G., Zinkel, R. A., Usui, T., Owen, B. A., Karcher A., Henderson B., Bodmer J. L., McMurray C. T., Carney J. P., Petrini J. H., & Tainer, J. A. (2002). The Rad50 zinc-hook is a structure joining Mre11 complexes in DNA recombination and repair. Nature, 418(6897), 562-566. 57) Hopfner, K. P., Putnam, C. D., & Tainer, J. A. (2002). DNA double-strand break repair from head to tail. Current Opinion in Structural Biology, 12(1), 115-122. 58) Horridge, J. J., & Leppard, K. N. (1998). RNA-binding activity of the E1B 55-kilodalton protein from human adenovirus type 5. Journal of Virology, 72(11), 9374-9379. 59) Hosokawa, K., & Sung, M. T. (1976). Isolation and characterization of an extremely basic protein from adenovirus type 5. Journal of Virology, 17(3), 924-934.

60) Huang, L. C., Clarkin, K. C., & Wahl, G. M. (1996). Sensitivity and selectivity of the DNA damage sensor responsible for activating p53-dependent G1 arrest. Proceedings of the National Academy of Sciences, 93(10), 4827-4832.

61) Huang, M. M., & Hearing, P. (1989). Adenovirus early region 4 encodes two gene products with redundant effects in lytic infection. Journal of virology, 63(6), 2605-2615. 62) Huen, M. S., Grant, R., Manke, I., Minn, K., Yu, X., Yaffe, M. B., & Chen, J. (2007). RNF8 transduces the DNA-damage signal via histone ubiquitylation and checkpoint protein assembly. Cell, 131(5), 901-914. 63) Huyen, Y., Zgheib, O., DiTullio Jr, R. A., Gorgoulis, V. G., Zacharatos, P., Petty, T. J., Sheston E. A., Mellert H. S., Stavridi E. S., & Halazonetis, T. D. (2004). Methylated lysine 79 of histone H3 targets 53BP1 to DNA double-strand breaks. Nature, 432(7015), 406-411.

64) Iijima, K., Ohara, M., Seki, R., & Tauchi, H. (2008). Dancing on damaged chromatin: functions of ATM and the RAD50/MRE11/NBS1 complex in cellular responses to DNA damage. Journal of Radiation Research, 49(5), 451-464. 65) Jackson, S. P. (2002). Sensing and repairing DNA double-strand breaks. Carcinogenesis, 23(5), 687-696. 66) Jayaram S., & Bridge E. (2005). Genome concatenation contributes to the late gene expression defect of an adenovirus E4 mutant. Virology, 342(2), 286-296. 67) Karen, K. A., & Hearing, P. (2011). Adenovirus core protein VII protects the viral genome from a DNA damage response at early times after infection. Journal of Virology, 85(9), 4135-4142. 68) Karen, K. A., Hoey, P. J., Young, C. S., Hearing, P. (2009). Temporal regulation of the Mre11-Rad50-Nbs1 complex during adenovirus infection. Journal of Virology, 83(9), 4565-4573. 69) Kastan, M. B., & Lim, D. S. (2000). The many substrates and functions of ATM. Nature Reviews Molecular Cell Biology, 1(3), 179-186. 70) Kim, J. E., Minter‐Dykhouse, K., & Chen, J. (2006). Signaling networks controlled by the MRN complex and Mdc1 during early DNA damage responses. Molecular Carcinogenesis, 45(6), 403-408. 71) Kuroda, S., Fujiwara, T., Shirakawa, Y., Yamasaki, Y., Yano, S., Uno, F., Tazawa, H., Hashimoto, Y., Watanabe, Y., Noma, K., Urata, Y., Kagawa, S., & Fujiwara, T. (2010). Telomerase-dependent oncolytic adenovirus sensitizes human cancer cells to ionizing radiation via inhibition of DNA repair machinery. Cancer Research, 70(22), 9339-9348.

72) Kusch, T., Florens, L., MacDonald, W. H., Swanson, S. K., Glaser, R. L., Yates III, J. R., Abmayr, S. M., Washburn, M. P., & Workman, J. L. (2004). Acetylation by Tip60 is required for selective histone variant exchange at DNA lesions. Science Signaling, 306(5704), 2084.

73) Lakdawala, S. S., Schwartz, R. A., Ferenchak, K., Carson, C. T., McSharry, B. P., Wilkinson, G. W., & Weitzman, M. D. (2008). Differential requirements of the C terminus of Nbs1 in suppressing adenovirus DNA replication and promoting concatemer formation. Journal of Virology, 82(17), 8362-8372.

74) Lavin, M. F. (2008). Ataxia-telangiectasia: from a rare disorder to a paradigm for cell signaling and cancer. Nature Reviews Molecular Cell Biology, 9(10), 759-769. 75) Lee, J. H., & Paull, T. T. (2005). ATM activation by DNA double-strand breaks through the Mre11-Rad50-Nbs1 complex. Science, 308(5721), 551-554. 76) Lilley, C. E., Schwartz, R. A., & Weitzman, M. D. (2007). Using or abusing: viruses and the cellular DNA damage response. Trends in Microbiology, 15(3), 119-126. 77) Lisby, M., Barlow, J. H., Burgess, R. C., & Rothstein, R. (2004). Choreography of the DNA damage response: spatiotemporal relationships among checkpoint and repair proteins. Cell, 118(6), 699-713. 78) Mailand, N., Bekker-Jensen, S., Faustrup, H., Melander, F., Bartek, J., Lukas, C., & Lukas, J. (2007). RNF8 ubiquitylates histones at DNA double-strand breaks and promotes assembly of repair proteins. Cell, 131(5), 887-900. 79) Maizel Jr, J. V., White, D. O., & Scharff, M. D. (1968). The polypeptides of adenovirus: I. Evidence for multiple protein components in the virion and a comparison of types 2, 7A, and 12. Virology, 36(1), 115-125. 80) Mathew, S. S., & Bridge, E. (2007). The cellular Mre11 protein interferes with adenovirus E4 mutant DNA replication. Virology, 365(2), 346-355. 81) Mathew, S. S., & Bridge, E. (2008). Nbs1-dependent binding of Mre11 to adenovirus E4 mutant viral DNA is important for inhibiting DNA replication. Virology, 374(1), 11-22. 82) Medina-Kauwe, L. K. (2003). Endocytosis of adenovirus and adenovirus capsid proteins. Advanced Drug Delivery Reviews, 55(11), 1485-1496. 83) Meier, O., & Greber, U. F. (2004). Adenovirus endocytosis. The Journal of Gene Medicine, 6(S1), S152-S163. 84) Moody, C. A., & Laimins, L. A. (2009). Human papillomaviruses activate the ATM DNA damage pathway for viral genome amplification upon differentiation. PLoS Pathogens, 5(10), e1000605. 85) Murr, R., Loizou, J. I., Yang, Y. G., Cuenin, C., Li, H., Wang, Z. Q., & Herceg, Z. (2005). Histone acetylation by Trrap–Tip60 modulates loading of repair proteins and repair of DNA double-strand breaks. Nature Cell Biology, 8(1), 91-99 86) Myers, J. S., & Cortez, D. (2006). Rapid activation of ATR by ionizing radiation requires ATM and Mre11. Journal of Biological Chemistry, 281(14), 9346-9350.

100

87) Nordqvist, K., Ohman, K., & Akusjärvi, G. (1994). Human adenovirus encodes two proteins which have opposite effects on accumulation of alternatively spliced mRNAs. Molecular and Cellular Biology, 14(1), 437-445. 88) Ohman, K., Nordqvist, K., & Akusjärvi, G. (1993). Two adenovirus proteins with redundant activities in virus growth facilitates tripartite leader mRNA accumulation. Virology, 194(1), 50-58. 89) Orazio, N. I., Naeger, C. M., Karlseder, J., & Weitzman, M. D. (2011). The adenovirus E1b55K/E4orf6 complex induces degradation of the Bloom helicase during infection. Journal of Virology, 85(4), 1887-1892. 90) Ou, H. D., Kwiatkowski, W., Deerinck, T. J., Noske, A., Blain, K. Y., Land, H. S., Soria C., Powers C. J., May A. P., Shu X., Tsien R. Y., Fitzpatrick J. A., Long J. A., Ellisman M. H., Choe S., & O'Shea, C. C. (2012). A Structural Basis for the Assembly and Functions of a Viral Polymer that Inactivates Multiple Tumor Suppressors. Cell,151(2), 304-319. 91) Paull, T. T., Rogakou, E. P., Yamazaki, V., Kirchgessner, C. U., Gellert, M., & Bonner, W. M. (2000). A critical role for histone H2AX in recruitment of repair factors to nuclear foci after DNA damage. Current Biology, 10(15), 886-895. 92) Philipson, L. (1961). Adenovirus assay by the fluorescent cell-counting procedure. Virology, 15(3), 263-268. 93) Pombo, A., Ferreira, J., Bridge, E., & Carmo-Fonseca, M. (1994). Adenovirus replication and transcription sites are spatially separated in the nucleus of infected cells. The EMBO Journal, 13(21), 5075. 94) Prage, L., & Pettersson, U. (1971). Structural proteins of adenoviruses: VII. Purification and properties of an arginine-rich core protein from adenovirus type 2 and type 3. Virology, 45(2), 364-373. 95) Prakash, A., Jayaram, S., & Bridge, E. (2012). Differential Activation of Cellular DNA Damage Responses by Replication-Defective and Replication-Competent Adenovirus Mutants. Journal of Virology, 86(24), 13324-13333. 96) Price, B. D., & D’Andrea, A. D. (2013). Chromatin remodeling at DNA double-strand breaks. Cell, 152(6), 1344-1354.

101

97) Querido, E., Blanchette, P., Yan, Q., Kamura, T., Morrison, M., Boivin, D., Kaelin, W. G., Conaway, R. C., Conaway, J. W., & Branton P. E. (2001). Degradation of p53 by adenovirus E4orf6 and E1B55K proteins occurs via a novel mechanism involving a Cullin-containing complex. Genes & Development, 15(23), 3104-3117. 98) Rekosh, D. M. K., Russell, W. C., Bellet, A. J. D., & Robinson, A. J. (1977). Identification of a protein linked to the ends of adenovirus DNA. Cell, 11(2), 283-295. 99) Russell, W. (2000). Update on adenovirus and its vectors. Journal of General Virology, 81(11), 2573-2604. 100) Rux, J. J., & Burnett, R. M. (1999). Adenovirus capsid proteins. Adenoviruses: from basic research to gene therapy applications, Medical Intelligence Unit, 15, 5-16. 101) Sandler, A. B., & Ketner, G. (1989). Adenovirus early region 4 is essential for normal stability of late nuclear RNAs. Journal of virology, 63(2), 624-630. 102) Sandler, A. B., & Ketner, G. (1991). The metabolism of host RNAs in cells infected by an adenovirus E4 mutant. Virology, 181(1), 319-326. 103) Schreiner, S., Bürck, C., Glass, M., Groitl, P., Wimmer, P., Kinkley, S., Mund A., Everett R. D., & Dobner, T. (2013). Control of human adenovirus type 5 gene expression by cellular Daxx/ATRX chromatin-associated complexes. Nucleic Acids Research,41(6), 3532-3550. 104) Schreiner, S., Wimmer, P., Groitl, P., Chen, S. Y., Blanchette, P., Branton, P. E., & Dobner, T. (2011). Adenovirus type 5 early region 1B 55K oncoprotein-dependent degradation of cellular factor Daxx is required for efficient transformation of primary rodent cells. Journal of Virology, 85(17), 8752-8765. 105) Sergeant, A. L. A. I. N., Tigges, M. A., & Raskas, H. J. (1979). Nucleosome-like structural subunits of intranuclear parental adenovirus type 2 DNA. Journal of Virology, 29(3), 888-898. 106) Shanbhag, N. M., Rafalska-Metcalf, I. U., Balane-Bolivar, C., Janicki, S. M., & Greenberg, R. A. (2010). ATM-Dependent chromatin changes silence transcription in cis to DNA double-strand sreaks. Cell, 141(6), 970-981. 107) Shenk, T. E. (1996). Adenoviridae: the viruses and their replication. Fields virology. Raven Press, New York, N.Y.

102

108) Shi, Y., Dodson, G. E., Shaikh, S., Rundell, K., & Tibbetts, R. S. (2005). Ataxia- telangiectasia-mutated (ATM) is a T-antigen kinase that controls SV40 viral replication in vivo. Journal of Biological Chemistry, 280(48), 40195-40200.

109) Shiloh, Y. (2000). ATM: Sounding the double-strand break alarm. In Cold Spring Harbor Symposia on Quantitative Biology (Vol. 65, pp. 527-534). Cold Spring Harbor Laboratory Press. 110) Shiloh, Y. (2003). ATM and related protein kinases: safeguarding genome integrity. Nature Reviews Cancer, 3(3), 155-168. 111) Shiotani, B., & Zou, L. (2009). Single-stranded DNA orchestrates an ATM-to-ATR switch at DNA breaks. Molecular Cell, 33(5), 547-558. 112) Shroff, R., Arbel-Eden, A., Pilch, D., Ira, G., Bonner, W. M., Petrini, J. H., Haber J. E., & Lichten, M. (2004). Distribution and dynamics of chromatin modification induced by a defined DNA double-strand break. Current Biology, 14(19), 1703-1711. 113) Stewart, G. S., Panier, S., Townsend, K., Al-Hakim, A. K., Kolas, N. K., Miller, E. S., Nakada S., Ylanko J., Olivarius S., Mendez M., Oldreive C., Wildenhain J., Tagliaferro A., Pelletier L., Taubenheim N., Durandy A., Byrd P. J., Stankovic T., Taylor A. M., & Durocher, D. (2009). The RIDDLE syndrome protein mediates a ubiquitin-dependent signaling cascade at sites of DNA damage. Cell, 136(3), 420-434. 114) Stracker, T. H., Carson, C. T., & Weitzman. M. D. (2002). Adenovirus oncoproteins inactivate the Mre11–Rad50–NBS1 DNA repair complex. Nature, 418(6895), 348-352. 115) Stucki, M., & Jackson S. P. (2003). γH2AX and MDC1: anchoring the DNA-damage- response machinery to broken chromosomes. DNA Repair, 5(5), 534-543. 116) Stucki, M., Clapperton, J. A., Mohammad, D., Yaffe, M. B., Smerdon, S. J., & Jackson, S. P. (2005). Mdc1 directly binds phosphorylated histone H2AX to regulate cellular responses to DNA double-strand breaks. Cell, 123(7), 1213-1226. 117) Sun, Y., Jiang, X., & Price, B. D. (2010). Tip60: connecting chromatin to DNA damage signaling. Cell Cycle, 9(5), 930-936. 118) Takata, M., Sasaki, M. S., Sonoda, E., Morrison, C., Hashimoto, M., Utsumi, H., Yamaguchi-Iwai, Y., Shinohara, A., & Takeda, S. (1998). Homologous recombination and non-homologous end-joining pathways of DNA double-strand break repair have

103

overlapping roles in the maintenance of chromosomal integrity in vertebrate cells. The EMBO Journal, 17(18), 5497-5508.

119) Tang, J., Agrawal, T., Cheng, Q., Qu, L., Brewer, M. D., Chen, J., & Yang, X. (2013). Phosphorylation of Daxx by ATM Contributes to DNA Damage-Induced p53 Activation. PloS One, 8(2), e55813.

120) Thomas, G. P., & Mathews, M. B. (1980). DNA replication and the early to late transition in adenovirus infection. Cell, 22(2), 523-533.

121) Tollefson, A. E., Ryerse, J. S., Scaria A., Hermiston, T. W., & Wold, W. S. (1996). The E3-11.6-kDa Adenovirus Death Protein (ADP) Is Required for Efficient Cell Death: Characterization of Cells Infected with adp Mutants. Virology, 220(1), 152-162. 122) Trentin, J. J., Yabe, Y., & Taylor, G. (1962). The Quest for Human Cancer Viruses. A new approach to an old problem reveals cancer induction in hamsters by human adenovirus. Science, 137(3533), 835-841. 123) Tsukuda, T., Fleming, A. B., Nickoloff, J. A., & Osley, M. A. (2005). Chromatin remodelling at a DNA double-strand break site in Saccharomyces cerevisiae. Nature, 438(7066), 379-383. 124) Turnell, A. S., & Grand, R. J. (2012). DNA viruses and the cellular DNA-damage response. Journal of General Virology, 93(Pt 10), 2076-2097. 125) Weber, J. M., Khittoo, G., & Bhatti, A. R. (1983). Adenovirus core proteins. Canadian Journal of Microbiology, 29(2), 235-241. 126) Weiden, M., & Ginsberg, H. (1994). Deletion of the E4 region of the genome produces adenovirus DNA concatemers. Proceedings of the National Academy of Sciences, 91(1), 153-157. 127) Weinberg, D. H., & Ketner, G. (1983). A cell line that supports the growth of a defective early region 4 deletion mutant of human adenovirus type 2. Proceedings of the National Academy of Sciences, 80(17), 5383-5386. 128) Weinberg, D. H., & Ketner, G. (1986). Adenoviral early region 4 is required for efficient viral DNA replication and for late gene expression. Journal of virology, 57(3), 833-838. 129) Hopfner, K. P., Putnam, C. D., & Tainer, J. A. (2002). DNA double-strand break repair from head to tail. Current Opinion in Structural Biology, 12(1), 115-122.

104

130) Weitzman, M. D. (2005). Functions of the adenovirus E4 proteins and their impact on viral vectors. Frontiers in Bioscience, 10, 1106-1117. 131) Weitzman, M. D., & Ornelles, D. A. (2005). Inactivating intracellular antiviral responses during adenovirus infection. Oncogene, 24(52), 7686-7696. 132) Weitzman, M. D., Lilley, C. E., & Chaurushiya, M. S. (2010). Genomes in conflict: maintaining genome integrity during virus infection. Annual Review of Microbiology, 64, 61-81. 133) Whyte, P., Buchkovich, K. J., Horowitz, J. M., Friend, S. H., Raybuck, M., Weinberg, R. & A., Harlow, E. (1988). Association between an oncogene and an anti-oncogene: the adenovirus E 1 A proteins bind to the retinoblastoma gene product. Nature, 334(6178), 124-129. 134) Wilkinson, D. E., & Weller, S. K. (2006). Herpes simplex virus type I disrupts the ATR- dependent DNA-damage response during lytic infection. Journal of Cell Science, 119(13), 2695-2703. 135) Xu, C., Bian, C., Yang, W., Galka, M., Ouyang, H., Chen, C., Qiu W., Liu H., Jones A. E., MacKenzie F., Pan P., Li S. S., Wang H., & Min, J. (2010). Binding of different histone marks differentially regulates the activity and specificity of polycomb repressive complex 2 (PRC2). Proceedings of the National Academy of Sciences, 107(45), 19266- 19271. 136) Xu, Y., & Price, B. D. (2011). Chromatin dynamics and the repair of DNA double strand breaks. Cell Cycle, 10(2), 261-267.

137) Xue, Y., Gibbons, R., Yan, Z., Yang, D., McDowell, T. L., Sechi, S., Qin, J., Zhou, S., Higgs, D., & Wang, W. (2003). The ATRX syndrome protein forms a chromatin- remodeling complex with Daxx and localizes in promyelocytic leukemia nuclear bodies. Proceedings of the National Academy of Sciences, 100(19), 10635-10640.

138) You, Z., Chahwan, C., Bailis, J., Hunter, T., & Russell, P. (2005). ATM activation and its recruitment to damaged DNA require binding to the C terminus of Nbs1. Molecular and Cellular Biology, 25(13), 5363-5379. 139) Zhao, S., Renthal, W., & Eva, Y. H. L. (2002). Functional analysis of FHA and BRCT domains of NBS1 in chromatin association and DNA damage responses. Nucleic acids research, 30(22), 4815-4822.

105

140) Ziv, Y., Bar-Shira, A., Pecker, I., Russell, P., Jorgensen, T. J., Tsarfati, I., & Shiloh, Y. (1997). Recombinant ATM protein complements the cellular A-T phenotype. Oncogene, 15(2), 159-167. 141) Zou, L., & Elledge, S. J. (2003). Sensing DNA damage through ATRIP recognition of RPA-ssDNA complexes. Science, 300(5625), 1542-1548. 142) Zuckerman, V., Wolyniec, K., Sionov, R. V., Haupt, S., & Haupt, Y. (2009). Tumour suppression by p53: the importance of apoptosis and cellular senescence. The Journal of Pathology, 219(1), 3-15.

106