HELB Is a Feedback Inhibitor of DNA End Resection

Ján Tkáč

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Graduate Department of Molecular Genetics University of Toronto

ABSTRACT HELB Is a Feedback Inhibitor of DNA End Resection

Ján Tkáč Doctor of Philosophy Department of Molecular Genetics University of Toronto 2016

DNA double-strand breaks are toxic lesions, which jeopardize the genomic integrity and survival of all cells and organisms. Repair of these lesions by homologous recombination requires the formation of 3′ single-stranded DNA (ssDNA) overhangs by a nucleolytic process known as

DNA end resection. Recent studies have significantly expanded our understanding of the initiation of resection, the molecular machinery involved in its execution, and its regulation throughout the cell cycle. However, the mechanisms that control and limit DNA end resection once the process has begun are unknown. I hypothesized that such activities may be coordinated by the ssDNA-binding complex Replication Protein A (RPA), which rapidly coats the 3′ ssDNA overhangs produced by resection. A proteomic analysis of RPA interactions following DNA damage identified the superfamily 1B translocase, DNA helicase B (HELB). Using cellular and biochemical approaches, I found that following RPA-dependent recruitment of HELB to the sites of DNA double-strand breaks, HELB inhibits EXO1 and BLM-DNA2 nucleases, which catalyze long-range resection. This function requires HELB’s catalytic activity and ssDNA binding, suggesting a mechanism where HELB translocates along ssDNA to displace the nucleases.

HELB acts independently of 53BP1 and is exported from the nucleus as cells approach S phase, concomitant with the upregulation of resection. I conclude that mammalian DNA end resection triggers its own inhibition via the recruitment of HELB.

To Roja

iii

ACKNOWLEDGEMENTS

The completion of this thesis would not have been possible without the contribution and support of several great colleagues and friends. My PhD journey and the scientific advancement it has produced is truly the result of a team effort rather than an individual accomplishment.

First, I would like to thank my supervisor and mentor, Daniel Durocher. Dan has been a constant source of motivation and encouragement throughout this process. Rather than merely showing me how to do science, Dan has taught me how to be a scientist. His dedication to research and his excitement towards discovery have been very inspirational to me.

I would like to thank all of the members of the Durocher laboratory whom I had a chance to work with over the past seven years. I am truly lucky to have been surrounded by people who are not only exceptional scientists, but also great friends. I specifically want to acknowledge the scientific contributions of Jordan Young, Cristina Escribano-Díaz, Meagan Munro, Alexandre Orthwein, Rachel Szilard, Sébastien Landry and Abdallah Al-Hakim.

I am thankful to Zhen-Yuan Lin and Anne-Claude Gingras for their mass spectrometry work; Guotai Xu and Sven Rottenberg for their work on PARP inhibitor sensitivity of BRCA1- deficient cells and tumors; Hemanta Adhikary, Jana Krietsch and Jean-Yves Masson for carrying out the HELB biochemical assays; David Gallo and Grant Brown for carrying out the DNA combing analysis. I thank the Centre for Modeling Human Disease Pathology Core (Toronto Centre for Phenogenomics) and the Princess Margaret Hospital Flow Cytometry Facility for their technical services.

I thank my supervisory committee members Marc Meneghini and Jack Greenblatt for their thoughtful comments, suggestions and critiques, all of which were indispensable to the successful completion of my PhD.

I thank my parents, Ivan Tkáč and Ružena Tkáčová, my brother, Micky Tkáč and my sister, Veronika Tkáčová, for their unconditional love and support. Although they are usually far away, they always manage to keep me grounded and ensure I keep a sense of perspective. Finally, thank you to my partner, my teammate, my best friend, my wife, Roja Ghahari – without you, none of this would be possible.

STATEMENT OF RIGHTS, PERMISSIONS AND CONTRIBUTIONS

This thesis was written de novo, but includes data and text which has been reproduced, adapted, and modified from an article I co-authored during my PhD:

Tkáč J*, Xu G*, Adhikary H, Young JTF, Gallo D, Escribano-Díaz C, Krietsch J, Orthwein A, Munro M, Sol W, Al-Hakim A, Lin Z-Y, Jonkers J, Borst P, Brown GW, Gingras A-C, Rottenberg S, Masson J-Y, Durocher D. (2016). HELB is a feedback inhibitor of DNA end resection. Molecular Cell 61, 405–418. * Co-first authors

This article can be found at DOI: http://dx.doi.org/10.1016/j.molcel.2015.12.013

The reproduction was done with permission from Molecular Cell (Elsevier), in compliance with the publisher’s editorial policies.

The above article was the result of collaboration between the laboratory of Dr. Daniel Durocher (my supervisor and corresponding author) and those of Dr. Grant Brown (University of Toronto), Dr. Anne-Claude Gingras (University of Toronto), Dr. Sven Rottenberg (Netherlands Cancer Institute, Amsterdam), and Dr. Jean-Yves Masson (Laval University, Québec). Attribution of the work of specific colleagues on specific experiments is included at the beginning of each chapter and within the text of the thesis.

TABLE OF CONTENTS

ABSTRACT ...... ii ACKNOWLEDGEMENTS ...... iv STATEMENT OF RIGHTS, PERMISSIONS AND CONTRIBUTIONS ...... v TABLE OF CONTENTS ...... vi LIST OF TABLES ...... viii LIST OF FIGURES ...... ix LIST OF APPENDICES ...... xi ABBREVIATIONS ...... xii PROTEIN NAMES ...... xiii

Chapter 1: Introduction ...... 1

1.1 Formation of DNA double-strand breaks ...... 2 1.2 Repair of DNA double-strand breaks ...... 5 1.3 Regulation of DNA end resection by the cell cycle ...... 10 1.4 Replication protein A ...... 13 1.5 Helicases and translocases ...... 17 1.3.1 UvrD ...... 20 1.3.2 Pif1 and Rrm3 ...... 21 1.3.3 HELB ...... 22 1.6 Hypothesis ...... 24

Chapter 2: HELB is an RPA-interacting protein that limits resection in human cells ...... 25

2.1 Searching for regulators of DNA end resection: Exploring the RPA interactome .... 26 2.2 HELB inhibits resection in human cells ...... 27 2.3 HELB is recruited to sites of DSBs in an RPA-dependent manner ...... 32

Chapter 3: HELB limits BLM-DNA2- and EXO1-dependent resection ...... 38

3.1 Generating the Helb-/- mouse ...... 39 3.2 Loss of HELB promotes SSA in mouse cells ...... 42 3.3 HELB inhibits CtIP- and ATM-dependent resection ...... 45 3.4 HELB inhibits BLM-DNA2- and EXO1-mediated long-range resection ...... 49

Chapter 4: HELB participates in the cell cycle regulation of DNA end resection ...... 52

4.1 HELB does not regulate DSB repair pathway choice ...... 53 4.2 HELB mediates the cell cycle regulation of DNA end resection ...... 56

Chapter 5: Discussion and future directions ...... 61

5.1 Conclusions ...... 62 5.2 The molecular mechanism of HELB ...... 62 5.3 Cell cycle regulation of HELB ...... 65 5.4 Is HELB involved in DNA replication? ...... 66 5.5 Phenotypic consequences of unregulated resection ...... 68 5.6 Therapeutic potential of modulating resection ...... 69 5.7 Model ...... 72

Chapter 6: Methods and materials ...... 73

6.1 Cell culture and treatments ...... 74 6.2 Affinity purification and mass spectrometry ...... 75 6.3 Mass spectrometry data extraction ...... 76 6.4 Interaction scoring for Flag AP-MS ...... 77 6.5 RNA interference ...... 78 6.6 Plasmids ...... 79 6.7 Antibodies ...... 79 6.8 HR and SSA DNA repair assays ...... 80 6.9 Phospho-RPA2 S4/S8 focus formation assay using high-content microscopy ...... 81 6.10 Laser microirradiation ...... 81 6.11 Single-strand annealing traffic light reporter (SSA-TLR) ...... 82 6.12 Native BrdU resection assay ...... 83 6.13 Neutral comet assay ...... 83 6.14 Class switch recombination ...... 84 6.15 Extrachromosomal NHEJ assay ...... 84 6.16 Protein purification ...... 85 6.17 RPA-bound ssDNA pull-down ...... 86 6.18 Streptavidin displacement assay ...... 87 6.19 DNA combing analysis ...... 87 6.20 Clonogenic assays ...... 88 6.21 Mice, tumor transplantation and olaparib treatment ...... 89

REFERENCES ...... 90

Appendix 1: Mass spectrometry data ...... 105

Appendix 2: Pathology reports of two Helb+/+ and two Helb-/- mice ...... 110

Appendix 3: Loss of HELB results in PARP inhibitor resistance and partial restoration of HR in BRCA1-deficient cells ...... 118

vii

LIST OF TABLES

Table 1.1 CDK-mediated phosphorylation of factors involved in DNA end resection ...... 13

Table 1.2 Family classification and mechanistic characteristics of SF1 and SF2 helicases/translocases ...... 20

Table 2.1 Quantitation of two orthogonal siRNA-based screens responsive to changes in DNA end resection ...... 30

Table 3.1 There is no neonatal lethality associated with Helb deletion in mice ...... 41

Table 5.1 Helb-/- mutation does not rescue the embryonic lethality of Brca1Δ11/Δ11 mice ...... 71

Table A.1 SAINTexpress (v 3.3) output for IP-MS using RPA1 as bait ...... 106

Table A.2 SAINTexpress (v 3.3) output for IP-MS using RPA2 as bait ...... 107

Table A.3 SAINTexpress (v 3.3) output for IP-MS using RPA3 as bait ...... 108

Table A.4 SAINTexpress (v 3.3) output for IP-MS using HELB as bait ...... 109

viii

LIST OF FIGURES

Figure 1.1 DNA double-strand break repair ...... 6

Figure 1.2 DNA double-strand break repair by the single-strand annealing (SSA) pathway ...... 9

Figure 1.3 Replication protein A (RPA) ...... 15

Figure 1.4 Modeling the multitude of activities of helicases/translocases...... 19

Figure 1.5 DNA helicase B (HELB) ...... 23

Figure 2.1 The RPA interactome following DNA damage ...... 28

Figure 2.2 HELB restricts resection in human cells ...... 31

Figure 2.3 Acidic residues in the helicase domain of HELB mediate its interaction with RPA .. 33

Figure 2.4 HELB is recruited to DSB sites in a CtIP-dependent manner ...... 35

Figure 2.5 HELB is recruited to DSB sites in an RPA-dependent manner ...... 37

Figure 3.1 Generation of the Helb-/- mouse ...... 40

Figure 3.2 Helb-/- mice are viable and do not display any phenotypes in the absence of environmental or experimental challenges ...... 42

Figure 3.3 HELB antagonizes SSA in MEFs ...... 45

Figure 3.4 HELB is a resection inhibitor in MEFs ...... 48

Figure 3.5 HELB inhibits EXO1- and BLM-DNA2-mediated resection ...... 51

Figure 4.1 HELB is not involved in DSB repair pathway choice ...... 55

Figure 4.2 The nuclear export of HELB promotes resection ...... 58

Figure 4.3 The combined loss of HELB and 53BP1 results in an additive increase in DNA end resection ...... 60

Figure 5.1 Exploring the mechanism of HELB ...... 65

Figure 5.2 DNA replication dynamics in Helb-/- cells ...... 67

Figure 5.3 A model of HELB activity ...... 72

Figure A.1 Loss of HELB results in PARP inhibitor resistance in BRCA1-deficient cells ...... 120

Figure A.2 HELB requires its resection-inhibiting activities to mediate the cytotoxicity of PARP inhibitors in BRCA1-deficient cells ...... 121

Figure A.3 Loss of HELB results in PARP inhibitor resistance in BRCA1-deficient mammary tumor allografts ...... 122

Figure A.4 Loss of HELB leads to a partial reactivation of HR in BRCA1-deficient cells ...... 123

LIST OF APPENDICES

Appendix 1 Mass spectrometry data ...... 105

Appendix 2 Pathology reports of two Helb+/+ and two Helb-/- mice ...... 110

Appendix 3 Loss of HELB results in PARP inhibitor resistanc and partial restoration of HR in BRCA1-deficient cells ...... 118

ABBREVIATIONS

BLAST basic local alignment search tool BrdU 5-bromo-2'-deoxyuridine CldU 5-chloro-2′-deoxyuridine CSR class switch recombination dHJ double Holliday junction DNA deoxyribonucleic acid DR-GFP direct repeat-GFP (homologous recombination reporter assay) DSB DNA double-strand break dsDNA double-stranded DNA FACS fluorescence-activated cell sorting hr hour(s) HR homologous recombination HU hydroxyurea IdU 5-iodo-2'-deoxyuridine IP immunoprecipitation IR ionizing radiation IRES internal ribosome entry site MEF mouse embryonic fibroblast min minute(s) MMEJ microhomology-mediated end-joining MS mass spectrometry n number of experimental biological replicates NCS neocarzinostatin NES nuclear export signal NHEJ non-homologous end joining NDP nucleoside diphosphate NTP nucleoside triphosphate OB fold oligonucleotide/oligosaccharide-binding fold domain PI propidium iodide RNA ribonucleic acid ROS reactive oxygen species RT-qPCR reverse transcription - quantitative polymerase chain reaction SA-GFP strand annealing-GFP (single-strand annealing reporter assay) SAINT significance analysis of interactome SD standard deviation SDSA synthesis-dependent strand annealing SEM standard error of the mean SF superfamily SSA single-strand annealing SSA-TLR single-strand annealing - traffic light reporter assay ssDNA single-stranded DNA UV ultraviolet WH winged helix domain WT wild type (genotype) xii

PROTEIN NAMES

53BP1 p53-binding protein 1 AID Activation-induced cytidine deaminase APE Apurinic/apyrimidinic endonuclease ATM Ataxia telangiectasia mutated ATR Ataxia telangiectasia and Rad3-related ATRIP ATR-interacting protein BFP Blue fluorescent protein BLM Bloom syndrome helicase BRCA1/2 Breast cancer type 1 susceptibility protein / type 2 susceptibility protein BRIP1 BRCA1-interacting protein C-terminal helicase 1 CDK Cyclin-dependent kinase CHTF18 Chromosome transmission fidelity protein 18 CtIP CtBP-interacting protein Dda DNA-dependent ATPase (Bacteriophage T4) DNA2 DNA replication ATP-dependent helicase/nuclease DNA2 DNA-PKcs DNA-dependent protein kinase catalytic subunit EME1 Essential meiotic endonuclease 1 ERCC1/2/3/4 Excision repair cross-complementation group 1/2/3/4 ETAA1 Ewing's tumor-associated antigen 1 EXO1 Exonuclease 1 GEN1 Flap endonuclease GEN homolog 1 GFP Green fluorescent protein H2A Histone 2A H2AX Histone 2A.X γ-H2AX Histone 2A.X phosphorylated on Ser139 H4 Histone 4 HELB DNA helicase B Ig Immunoglobulin Ku80/70 X-ray repair cross-complementing protein 5/6 LIG4 DNA ligase 4 MAD2L2 Mitotic spindle assembly checkpoint protein MAD2B MRE11 Meiotic recombination 11 homolog 1 MRN MRE11-RAD50-NBS1 complex MUS81 MMS and UV-sensitive protein 81 NBS1 Nijmegen breakage syndrome protein 1 p53 Tumor suppressor phosphoprotein 53 PALB2 Partner and localizer of BRCA2 PARP Poly (ADP-ribose) polymerase PCNA Proliferating cell nuclear antigen Pif1 Petite integration frequency 1 (S. cerevisiae) PIKK Phosphatidylinositol 3-kinase-like kinase POLA1/2 DNA polymerase α catalytic subunit / subunit B PRIM1/2 DNA primase small subunit / large subunit

xiii

RAD50/51/52 Radiation sensitive homolog 50/51/52 RAG1/2 Recombination activating gene 1/2 RecD RecBCD enzyme subunit RecD (E. coli) RFP Red fluorescent protein RIF1 Rap1-interacting factor 1 RMI1/2 RecQ-mediated genome instability protein 1/2 RNR Ribonucleotide reductase RPA1/2/3 Replication protein A 70 kDa subunit / 32 kDa subunit / 14 kDa subunit Rrm3 rDNA recombination mutation protein 3 (S. cerevisiae) SLX1/4 Structure-specific endonuclease subunit SLX1/4 SMARCAL1 SWI/SNF-related matrix-associated actin-dependent regulator of chromatin subfamily A-like protein 1 SPO11 Meiotic recombination protein SPO11 SSB Single-stranded DNA-binding protein (E. coli) TopoIIIα DNA topoisomerase III alpha TraI Multifunctional conjugation protein TraI (E. coli) UNG Uracil-DNA glycosylase UvrA/B/C/D UV radiation sensitive mutant A/B/C/D (E. coli) WRN Werner syndrome ATP-dependent helicase XLF XRCC4-like factor XPA Xeroderma pigmentosum group A-complementing protein XRCC4 X-ray repair cross-complementing protein 4 ZUFSP Zinc finger with UFM1-specific peptidase domain protein

xiv

Chapter 1: Introduction

1.1 Formation of DNA double-strand breaks

The maintenance of genome integrity and the faithful transmission of genetic information from one generation to the next are fundamental for life of all cells and organisms. However, cellular DNA is continually subjected to various types of damage as a result of both endogenous and exogenous sources. Although DNA is a relatively stable macromolecule (compared to RNA or polypeptides (Lindahl, 1993), based on the size of the human genome it is estimated that more than 10,000 DNA bases become damaged during each cell cycle in a single human cell (Lindahl and Nyberg, 1972; Nakamura et al., 1998). This base damage is due to spontaneous (non- enzymatic) hydrolysis, oxidation, and transmethylation reactions associated with normal cellular metabolism (Lindahl and Barnes, 2000). Enzymatic sources of damage include DNA mismatches introduced during DNA replication (albeit at a very low frequency) and DNA strand breaks caused by abortive topoisomerase activity (Jackson and Bartek, 2009). The most prevalent exogenous DNA-damaging agent is solar ultraviolet (UV) radiation, which induces widespread covalent dimerization of pyrimidine nucleotides in exposed cells (Jackson and Bartek, 2009). In the face of this constant exposure to DNA-damaging agents, cells have evolved multiple mechanisms to detect, signal, and repair the various types of lesions. The focus of this thesis is the repair of a less frequent, but extremely toxic form of damage, the DNA double-strand break

(DSB), which is the disruption of both sugar-phosphate backbones of the DNA double helix.

A replicating human cell encounters approximately 10-50 DSBs every cell cycle, an estimate based on the frequency of chromatid breaks and sister chromatid exchanges observed on metaphase spreads (Lieber, 2010; Vilenchik and Knudson, 2003). DSBs commonly arise as a result of oxidative attacks on the DNA backbone by reactive oxygen species (ROS). ROS arise as by-products of normal cellular metabolism, but can also be induced by exposure to exogenous

agents such as ionizing radiation (IR, including X-rays and γ-rays) and heavy metals. ROS generate mostly DNA single-strand breaks, which are converted to DSBs if they occur in close proximity to each other on opposite strands (Jackson and Bartek, 2009; Ward, 1988).

Furthermore, DSBs can arise when an advancing DNA replication fork encounters a single- strand break or another lesion (such as chemically-induced DNA adducts and crosslinks), which stalls its progression. Failure to restart the stalled replication fork in a timely manner can lead to fork collapse and DSB formation (Cortez, 2015).

Interestingly, there also exist specialized cases where DSBs are generated enzymatically in a programmed manner. For example, DSB formation by the type II topoisomerase protein

SPO11 is critical for recombination between homologous chromosomes during meiotic prophase

I (Keeney, 2008). This crossover recombination ensures accurate chromosome segregation and increases the genetic diversity of gametes. Additionally, the development of adaptive immunity in vertebrates involves programmed genome rearrangements in maturing lymphocytes. V(D)J recombination (which generates a diverse set of antibodies and T-cell receptors in B and T cells, respectively) is initiated when RAG1 and RAG2 proteins induce DSBs at specific sites adjacent to the variable (V), diversity (D) and joining (J) segments of the immunoglobulin (Ig) gene.

Subsequently, the constant region of the Ig gene in B cells is rearranged to generate various antibody isotypes in a process called class switch recombination (CSR). CSR begins when cytosines in the switch regions between the Ig constant region gene segments are deaminated to form uracil by the enzyme activation-induced cytidine deaminase (AID). The uracil bases on opposing strands are processed into DSBs by the sequential activities of base excision repair enzymes. First, uracil bases are removed by uracil-DNA glycosylase (UNG), resulting in abasic sites in the DNA. Subsequently, the abasic sites are recognized and cleaved by apurinic/apyrimidinic endonuclease (APE), producing DSBs. These examples illustrate that

programmed DSB formation is essential for sexual reproduction and for development of the vertebrate immune system.

Finally, some chemical agents have clastogenic properties. For example, inhibitors of topoisomerase I (e.g. camptothecin) and topoisomerase II (e.g. etoposide) lead to the formation of ternary covalent complexes, which include the chemical, the topoisomerase protein, and the

DNA end at the site of the cut. These ternary complexes prevent the re-ligation of the cut strands and require extensive processing by nucleases in order to be successfully repaired. Other types of clastogens include radiomimetic drugs, such as bleomycin and neocarzinostatin (NCS) (Povirk,

1996). NCS was used extensively in my experiments and so warrants a detailed description. It was identified as an antitumor antibiotic obtained from Streptomyces carzinostaticus culture filtrates (Ishida et al., 1965). NCS consists of a single non-protein chromophore tightly (but non- covalently) bound to a single carrier protein, which protects the labile chromophore from degradation (Goldberg, 1987). In aqueous solution at neutral pH, the isolated chromophore is completely degraded in a matter of seconds, whereas it is stable for hours when complexed with its binding protein (Povirk and Goldberg, 1980). All DNA-damaging activity of NCS resides within the chromophore (Goldberg, 1987), which intercalates into the DNA double helix with a preference for A- and T-rich sequences (Poon et al., 1977). The NCS chromophore is a free radical, which mediates oxidative attacks on the deoxyribose sugar backbone of DNA (Kappen and Goldberg, 1983), producing mostly single- and double-strand breaks. In addition, the NCS chromophore can form stable adducts with DNA (Goldberg, 1987). Similarly to the ternary complexes associated with topoisomerase inactivation, the NCS-DNA adducts likely require extensive processing by nucleases in order to be successfully repaired.

1.2 Repair of DNA double-strand breaks

The importance of DSB repair is underscored by the cytotoxicity of this lesion. In budding yeast, a single unrepaired DSB was shown to be sufficient to induce permanent cell cycle arrest and cell death (Bennett et al., 1993). In human cells, inaccurate repair of DSBs can lead to loss of genetic information and gross chromosomal rearrangements, which may result in neoplastic transformation. Genetic defects in DSB recognition, signaling, and repair pathways lead to various syndromes characterized by developmental, immunological and neurological symptoms, as well as cancer susceptibility (Jackson and Bartek, 2009; O'Driscoll, 2012).

DSBs are primarily repaired by one of two major mechanisms: non-homologous end joining (NHEJ) or homologous recombination (HR) (Figure 1.1). NHEJ is a non-templated repair pathway, which involves the direct re-ligation of the broken DNA ends. These ends are almost invariably bound by the Ku heterodimer (consisting of Ku70 and Ku80 subunits), which has a very high affinity for duplex DNA ends and is one of the most abundant nuclear proteins, with an estimated abundance of ~400,000 molecules per cell (Blier et al., 1993; Lieber, 2010; Mimori and Hardin, 1986). Ku stabilizes the ends and acts as a docking site for the recruitment of downstream repair factors. In parallel, phosphorylation- and ubiquitylation-mediated signaling cascades lead to the recruitment of the chromatin binding factor 53BP1 and its downstream effectors RIF1, PTIP, and MAD2L2 (Boersma et al., 2015; Chapman et al., 2013; Escribano-

Díaz et al., 2013; Kolas et al., 2007; Stewart et al., 2009; Xu et al., 2015). The 53BP1 pathway protects the DSB ends from resection nucleases by preventing recruitment of BRCA1, an important mediator of HR repair (Panier and Durocher, 2013). Subsequently, DNA-PKcs and

Artemis are recruited to the Ku-bound DSB ends to prepare the ends for ligation by cleaving excess overhangs and hairpin structures (Ma et al., 2002). Finally, the ligation is catalyzed by

Figure 1.1 DNA double-strand break repair The non-homologous end joining (NHEJ) and homologous recombination (HR) repair pathways are illustrated. See text for details. The cell cycle-mediated regulation of HR is not depicted here in detail, but is described in depth in Section 1.3. 6

DNA ligase 4 (LIG4) in complex with XRCC4 and XLF proteins. Due to the variety of end processing outcomes inherent in this pathway, NHEJ is sometimes referred to as “error-prone” and may result in microdeletions or microduplications at the site of repair. Remarkably, the programmed DSBs created in lymphocytes during V(D)J and class switch recombination are repaired exclusively by NHEJ. Both processes benefit from the repair inaccuracy associated with

NHEJ by further increasing antibody diversity. It is worth noting that both V(D)J and CSR are unfortunately misnamed as “recombination” when in fact they employ non-templated end joining and not HR.

DSB repair by HR is a lengthier process involving DNA synthesis using a homologous template, which is why it is sometimes referred to as “error-free” repair. The initiating and rate- limiting step in HR is DNA end resection, which creates 3’ ssDNA overhangs (Ferretti et al.,

2013). Resection is a two-step mechanism: the first step is resection initiation (i.e. short-range) and the second is extensive resection (i.e. long-range) (Symington, 2016). The initiation of resection is triggered by CtIP in collaboration with the MRE11-RAD50-NBS1 (MRN) complex

(Ferretti et al., 2013; Sartori et al., 2007; Shibata et al., 2014), a process stimulated by the interaction of CtIP with BRCA1 (Escribano-Díaz et al., 2013). The resulting short stretches of resected DNA preclude end protection by Ku (Mimitou and Symington, 2010). Thus, once resection has been initiated, the cell has committed to repair the DSB by HR rather than NHEJ.

Subsequently, EXO1 and DNA2 nucleases (the latter in complex with BLM helicase) are recruited to produce the long tracts of ssDNA necessary for homology search and invasion of duplex DNA (Gravel et al., 2008; Mimitou and Symington, 2008; Nimonkar et al., 2011; Zhu et al., 2008). As soon as ssDNA arises in the cell, it is recognized and tightly bound by replication protein A (RPA), which stabilizes the ssDNA overhangs (Chen et al., 2013). RPA-ssDNA serves as a critical platform for the recruitment of downstream repair factors and for the initiation of

ATR-mediated checkpoint signalling. The key enzyme in HR is the RAD51 recombinase, which requires BRCA2 (in complex with BRCA1 and PALB2) for the displacement of RPA and loading onto the ssDNA overhangs. The resulting RAD51-ssDNA nucleofilament scans the template DNA (usually on the sister chromatid) for homologous sequences and invades the template duplex at the site of homology. The invading strand then acts as a primer for polymerases to synthesize DNA across the break site. This newly extended strand can dissociate from the template and anneal back to the other resected DSB end, in a process called synthesis- dependent strand annealing (SDSA). Alternately, the newly synthesized strand can remain annealed to the template DNA, which leads to the formation of cross-stranded joint molecule intermediates known as double Holliday junctions (dHJs) (Heyer et al., 2010). dHJs can be

“dissolved” into non-crossover products by the BLM-TopoIIIα-RMI1-RMI2 complex, or

“resolved” into both crossover and non-crossover products by the action of nucleases SLX1-

SLX4, MUS81-EME1 and/or GEN1 (Svendsen and Harper, 2010).

In the absence of RAD51 nucleofilament formation, resected DSBs can be repaired by a relatively less understood pathway called single-strand annealing (SSA) (Figure 1.2). SSA relies on long-range resection to expose regions of homology (generally > 30 bp) within the same

DNA molecule (Pâques and Haber, 1999). The complementary sequences are annealed and the resulting protruding non-homologous flaps are cleaved by the ERCC1-ERCC4 nucleotide excision repair nucleases (Frankenberg-Schwager et al., 2009). Thus, SSA invariably results in the deletion of the intervening DNA sequence, making it a highly mutagenic repair pathway.

Since it requires sufficient resection to expose relatively long complementary repeats, SSA has frequently been used as a readout for extensive resection (Symington, 2016).

Figure 1.2 DNA double-strand break repair by the single-strand annealing (SSA) pathway SSA and HR both rely on extensive resection by the EXO1 and BLM-DNA2 nucleases. The black boxes indicate complementary DNA sequences, which are capable of annealing during SSA. See text for details.

1.3 Regulation of DNA end resection by the cell cycle

The maintenance of genomic integrity necessitates a profound integration of DNA repair with the cell division cycle (Chapman et al., 2012; Symington and Gautier, 2011). Indeed, cyclin-dependent kinases (CDKs), which drive cell cycle progression, have multiple known inputs into the HR pathway of DSB repair. This level of regulation likely evolved to maximize the fidelity of HR by restricting it to the S and G2 phases of the cell cycle, when a sister chromatid produced by DNA replication is present. The sister chromatid provides an ideal template for mitotic HR (Jasin and Rothstein, 2013) by helping to avoid undesirable outcomes such as loss of heterozygosity associated with recombination between homologous chromosomes or unscheduled HR between interspersed repetitive sequences. HR is thus suppressed during G1

(when NHEJ predominates) and is activated as cells enter S phase. Cells remain HR-competent until they enter mitosis, a period in which DSB repair by both canonical pathways is suppressed for the benefit of accurate chromosome segregation (Orthwein et al., 2014).

The regulation of HR by the cell cycle depends in large part on the regulation of DNA end resection, as it is the initiating and rate-limiting step in HR (Ferretti et al., 2013). Resection has been known to require high CDK activity for more than a decade (Aylon et al., 2004; Ira et al., 2004), however the exact targets of CDK-mediated phosphorylation and the effects of these modifications have only begun to emerge more recently. Resection also serves as the critical determinant of DSB repair pathway choice, because its initiation commits the cell to HR and precludes repair by the NHEJ pathway (Figure 1.1). In G1, 53BP1 accumulates on the chromatin surrounding DSB sites by recognizing mononucleosomes containing both dimethylated H4

Lys20 and ubiquitylated H2A Lys15; the latter modification resulting from a ubiquitin signaling cascade triggered by DNA damage (Fradet-Turcotte et al., 2013; Kolas et al., 2007; Stewart et

al., 2009). 53BP1, along with its interacting proteins RIF1, PTIP and MAD2L2, protect DSB ends from resection and promote DSB repair by NHEJ (Boersma et al., 2015; Bothmer et al.,

2010; Callén et al., 2013; Chapman et al., 2013; Di Virgilio et al., 2013; Escribano-Díaz et al.,

2013; Feng et al., 2013; Xu et al., 2015; Zimmermann et al., 2013).

In particular, RIF1 has been shown to block the accumulation of BRCA1, a positive regulator of resection, at sites of DSBs (Figure 1.1) (Escribano-Díaz et al., 2013). However, as cells enter S phase, CDKs phosphorylate the resection initiation protein CtIP at Ser327 and

Thr847 (Huertas and Jackson, 2009; Yu and Chen, 2004). Phosphorylation of CtIP Ser327 promotes its interaction with BRCA1 (Yun and Hiom, 2009), which inhibits the accumulation of

RIF1 at DSB sites once S phase has begun (Chapman et al., 2013; Escribano-Díaz et al., 2013).

In addition, phosphorylation of CtIP Thr847 upregulates its resection activity (Huertas and

Jackson, 2009) and expression of the unphosphorylatable CtIP T847A mutant as the sole source of this protein results in embryonic lethality in mice (Polato et al., 2014), suggesting that CDK- triggered end resection is the essential function of CtIP in mammalian cells.

Another resection initiation protein regulated by CDK activity is the MRN complex component NBS1, which is mutated in the autosomal recessive genetic disorder Nijmegen breakage syndrome, characterized by short stature, microcephaly, immunodeficiency and cancer predisposition (Digweed and Sperling, 2004; Weemaes et al., 1981). NBS1 is phosphorylated on

Ser432 in S/G2/M phases of the cell cycle in a CDK-dependent manner (Falck et al., 2012;

Ferretti et al., 2013). Expression of the wild type NBS1 protein in NBS1-deficient, patient- derived cells, restored normal levels of DNA end resection; however, expression of the unphosphorylatable NBS1 S432A mutant impaired resection initiation (Falck et al., 2012).

Importantly, the NBS1 S432A mutation did not affect its interaction with CtIP, suggesting that the phenotype was specific to (the lack of) CDK-mediated phosphorylation of NBS1.

Long-range resection, which is catalyzed by BLM-DNA2 and EXO1 nucleases is also under the control of CDK activity. In budding yeast, Dna2 was shown to be phosphorylated at

Thr4, Ser17 and Ser237 by CDK (Chen et al., 2011). This Dna2 phosphorylation promotes its nuclear entry, recruitment to DSBs, and extensive resection. Finally, four sites in the C-terminus of human EXO1 are also targets of CDK-mediated phosphorylation in S/G2 phases (Tomimatsu et al., 2014). An unphosphorylatable EXO1 mutant is impaired in its recruitment to DSBs compared to the wild type protein. Expression of this mutant reduces resection, chromosomal integrity, cell survival and HR upon DNA damage, but increases NHEJ. In contrast, expression of a phosphomimetic mutant of EXO1 facilitates resection even when CDK activity is inhibited, and channels cells to repair by HR instead of NHEJ (Tomimatsu et al., 2014).

Together, the above examples (summarized in Table 1.1) illustrate that the influence of the cell cycle in DSB repair pathway choice and DNA end resection in particular has emerged as a critical mechanism of regulation. Future research is likely to identify additional cell cycle- regulated inputs into these pathways. Furthermore, it is important to note that the cell cycle also regulates the HR repair pathway downstream of DNA end resection. One instance is at the level of RAD51-ssDNA nucleofilament formation, which requires a functional BRCA1-PALB2-

BRCA2 complex, the assembly of which is restricted to the S/G2 phases (Orthwein et al., 2015).

Table 1.1 CDK-mediated phosphorylation of factors involved in DNA end resection

Protein CDK targets Functional relevance of phosphorylation References NBS1 Ser432 Promotes resection initiation (Falck et al., 2012) CtIP Ser327 pSer327 promotes interaction with BRCA1 (Yun and Hiom, 2009) Thr847 pThr847 promotes resection initiation (Huertas and Jackson, 2009) Dna2 Thr4 Phosphorylation of all three sites promotes (Chen et al., 2011) (S. cerevisiae) Ser17 nuclear import, recruitment to DSBs, Ser237 long-range resection EXO1 Ser639 Phosphorylation of all four sites promotes (Tomimatsu et al., 2014) Thr732 recruitment to DSBs, long-range resection Ser815 Thr824 RPA2 Ser23 “Priming” modification to facilitate RPA (Maréchal and Zou, 2015) Ser29 hyperphosphorylation by PIKK enzymes

1.4 Replication protein A

When ssDNA arises in a cell, it is recognized and rapidly bound by the heterotrimeric complex RPA. RPA is the eukaryotic analog of the prokaryotic single-stranded DNA-binding protein SSB (Sancar et al., 1981), which plays critical roles in DNA replication and recombination in E. coli (Shereda et al., 2008). Human RPA was initially discovered as a factor essential for the replication of simian virus 40 (SV40) DNA using HeLa cellular extracts in vitro

(Wold and Kelly, 1988). RPA is an obligatory heterotrimer comprised of a 70 kDa subunit

(RPA1), a 32 kDa subunit (RPA2) and a 14 kDa subunit (RPA3), which together contain six oligonucleotide/oligosaccharide-binding (OB) fold domains (Figure 1.3A) (Murzin, 1993). All three RPA subunits are essential in yeast (Brill and Stillman, 1991).

RPA binds ssDNA with a defined polarity (de Laat et al., 1998; Iftode and Borowiec,

2000). This interaction is dynamic and involves at least three different binding modes

(Bochkareva et al., 2001; Chen and Wold, 2014; Fanning, 2006). The major mode is characterized by an occluded binding site of ~30 nucleotides per trimer (Kim et al., 1992) and a

very high affinity (dissociation constant is ~10-10 M) (Kim et al., 1994). This binding mode involves four of the six potentially available OB folds (three within RPA1 and one within RPA2)

(Figure 1.3A). Interestingly, E. coli SSB forms a homotetrameric complex, which also binds ssDNA using four OB folds (one per subunit) (Raghunathan et al., 1997). The heterotrimerization of the RPA complex is mediated by the C-terminal OB fold of RPA1 and the single OB folds of RPA2 and RPA3 (Figure 1.3A). Each of these OB folds is flanked by a helix at its C-terminus. The arrangement of these helices in parallel forms a three-helix bundle structure, which supports the trimerization (Bochkareva et al., 2002).

After binding to ssDNA, RPA provides a critical platform for the recruitment of downstream repair factors and for the initiation of ATR-mediated checkpoint signaling

(Maréchal and Zou, 2015). As a result, RPA physically interacts with a large number of proteins through several interaction domains (Fanning, 2006; Maréchal and Zou, 2015). For example, the basic cleft in the N-terminal OB fold of RPA1 binds to the MRN complex component MRE11, the ATR-mediated checkpoint activators RAD9 and ATRIP, as well as the tumor suppressor p53

(Bochkareva et al., 2005; Xu et al., 2008). Another example is the C-terminal winged helix (WH) domain of RPA2, which interacts directly with the base excision repair protein UNG, the nucleotide excision repair protein XPA, as well as the DNA recombination factor RAD52 (Mer et al., 2000). All three of these proteins contain an α-helix and interact with the same surface of the RPA2 WH domain. It has been proposed that the conserved nature of these interactions indicates that RPA2 serves a common function in these three different DNA repair pathways

(Fanning, 2006). Consistent with this notion, a C-terminally-truncated, temperature-sensitive

RPA2 mutant in budding yeast displays both mutator and hyper-recombination phenotypes

(Santocanale et al., 1995).

Figure 1.3 Replication protein A (RPA) (A) Schematic representation of the RPA subunits. OB fold domains are designated A-F as per (Maréchal and Zou, 2015). The OB folds which directly interact with ssDNA (A-D) are colored dark blue. Protein-protein interactions are designated by double-headed arrows. Only a subset of known RPA-interacting proteins is shown. Phosphorylation target sites are designated by single-headed arrows. WH, winged helix domain. (B) The N-terminal amino acid sequence of RPA2 is shown. Phosphorylated residues are highlighted in red and their respective kinases are indicated.

RPA is subject to various post-translational modifications, of which, phosphorylation is the best characterized. Specifically, the N-terminus of the RPA2 subunit contains seven phosphorylation target sites, five of which are substrates for the phosphatidylinositol 3-kinase- like kinase (PIKK) family enzymes ATM, ATR and DNA-PKcs, and two of which are substrates for CDKs (Figure 1.3B). Bruce Stillman and colleagues first observed cell cycle-dependent phosphorylation of both yeast and human RPA2 more than 25 years ago (Din et al., 1990).

Indeed, RPA2 is phosphorylated by CDKs at Ser23 and Ser29 starting at the G1/S phase transition and these modifications persist until mitosis during an unperturbed cell cycle (Din et al., 1990; Stephan et al., 2009). The phosphorylation of Ser23 and Ser29 appears to be a

“priming” event – the cell cycle-dependent phosphorylation of RPA2 by CDKs is required for the efficient DNA damage-induced hyperphosphorylation by PIKKs (Maréchal and Zou, 2015).

In the RPA2 S23A/S29A mutant, both IR- and camptothecin-induced hyperphosphorylation of the PIKK target sites is strongly impaired (Anantha et al., 2007; Liu and Weaver, 1993).

Furthermore, the phosphorylation of Ser4, Ser8 and Thr21 by DNA-PKcs (and/or ATM) requires the earlier phosphorylation of Ser33 by ATR (Anantha et al., 2007), which is the first PIKK enzyme recruited to RPA-coated ssDNA. Although there is considerable debate regarding the biological significance of the differentially phosphorylated states of RPA, it is clear that the pre- requisite for this RPA2 N-terminal hyperphosphorylation is the binding of the RPA complex to ssDNA exposed by end resection or stalling of DNA polymerases during replication (Maréchal and Zou, 2015). Thus, RPA2 phosphorylation can be used as a marker of ssDNA accumulation in cells.

1.5 Helicases and translocases

Helicases and translocases comprise a diverse class of nucleic acid-based “motor proteins” which function in nearly all biological processes involving DNA or RNA (Lohman et al., 2008). Therefore, all forms of cellular life and most viruses encode helicases/translocases.

The human genome encodes upwards of 120 such proteins (Fairman-Williams et al., 2010).

Helicases are defined as nucleic-acid dependent enzymes that couple nucleoside triphosphate

(NTP, usually ATP) hydrolysis to the separation of DNA or RNA duplex substrates into their component single strands (Singleton et al., 2007). The broader term, translocases, refers to all nucleic-acid dependent “motor proteins” that convert the chemical energy from NTP hydrolysis into mechanical energy for directional movement along single- or double-stranded nucleic acid substrates.

Helicase activities in E. coli were first discovered using biochemical methods in the

1970s (Abdel-Monem and Hoffmann-Berling, 1976; Abdel-Monem et al., 1976; Richet and

Kohiyama, 1976). In 1982, seminal studies of ATPase proteins by Walker and colleagues identified two conserved amino acid motifs responsible for ATP binding and subsequent hydrolysis (Walker et al., 1982). These motifs later became known as Walker A (ATP binding) and Walker B (ATP hydrolysis). Rigorous early bioinformatic analyses by Gorbalenya and

Koonin formed the basis of putative helicase classification (Gorbalenya and Koonin, 1988;

Gorbalenya et al., 1988a; 1988b), originally into 2 superfamilies (SF) (Gorbalenya et al., 1989) and later into 3 larger superfamilies and 2 smaller families (Gorbalenya and Koonin, 1993). The latter classification laid the foundation for most structure-function analyses of this class of proteins for more than a decade (Singleton et al., 2007). During this time it was shown that some putative “helicases” do not actually unwind complementary strands of DNA/RNA, but rather

simply couple NTP hydrolysis to movement (usually directionally biased) along nucleic acids, leading to the introduction of the broader term “translocases”. Finally, Wigley and colleagues used comparisons of helicase/translocase tertiary structures to develop the current classification into 6 superfamilies (Singleton et al., 2007).

Additionally, helicases/translocases can be subdivided on the basis of mechanistic properties. The minimal active state of SF1 and SF2 enzymes is monomeric or dimeric, whereas

SF3-6 enzymes form hexameric (or double-hexameric) ring structures. Translocation along single-stranded nucleic acids may occur in either the 3’-5’ (denoted type A) or the 5’-3’ (type B) direction. SF1, SF2, and SF6 contain examples of both type A and B helicases/translocases, whereas all SF3 enzymes are type A, and all SF4 and SF5 enzymes are type B (Singleton et al.,

2007). Finally, enzymes that translocate along ssDNA are denoted type α, whereas those that translocate along dsDNA are denoted type β. All SF1 enzymes are type α, whereas all other superfamilies contain examples of both type α and β translocases (Singleton et al., 2007). These mechanistic properties and their associated biological activities are illustrated in Figure 1.4.

SF1 and SF2 comprise by far the largest superfamilies of helicases/translocases in human cells. Using protein sequence alignments of the SF1 and SF2 enzymes, Jankowski and colleagues subdivided these superfamilies into 3 and 10 families, respectively (Fairman-Williams et al.,

2010). This classification, along with the mechanistic characteristics of the families are summarized in Table 1.2. Below, I detail some examples of SF1 helicases/translocases, which are relevant to this dissertation.

Figure 1.4 Modeling the multitude of activities of helicases/translocases. Simplified models of ssDNA and dsDNA helicases/translocases and their associated mechanistic activities. See text for details.

Table 1.2 Family classification and mechanistic characteristics of SF1 and SF2 helicases/translocases. Adapted and modified from (Fairman-Williams et al., 2010) and references therein.

1.3.1 UvrD

The UvrD helicase (SF1Aα) participates in the nucleotide excision repair (NER) pathway in E. coli, which removes and replaces covalently modified nucleotides from DNA, including pyrimidine dimers resulting from ultraviolet (UV) radiation. These lesions are detected and excised by the UvrABC endonuclease complex (Boyce and Howard-Flanders, 1964; Sancar and

Rupp, 1983; Setlow and Carrier, 1964). Following UvrC-mediated incision of the phosphodiester backbone surrounding the pyrimidine dimer, UvrD (sometimes referred to as E. coli helicase II) uses its 3’-5’ helicase activity to release the excised damaged oligomer and UvrC protein from the DNA (Orren et al., 1992). The resulting ssDNA gap in the excised region is filled in by DNA polymerase I and the nick is immediately sealed by DNA ligase.

Human cells do not possess a direct ortholog of UvrD. Instead, analogous functions during NER are performed by two SF2 helicase subunits of the multi-protein general transcription factor II (TFIIH). These subunits are ERCC2 (also known as XPD, because its mutation leads to the UV sensitivity and skin cancer predisposition syndrome xeroderma pigmentosum), which acts as a 5’-3’ helicase; and ERCC3 (also known as XPB, another xeroderma pigmentosum-related protein), which acts as a 3’-5’ helicase (Drapkin et al., 1994).

1.3.2 Pif1 and Rrm3

Pif1, the prototypical member of the Pif1-like family of SF1Bα helicases/translocases, was initially identified in a genetic screen for mutations affecting mitochondrial DNA recombination between wild type and cytoplasmic petite mutant strains (i.e., petite integration frequency 1, PIF1) in budding yeast (Foury and Kolodynski, 1983). Subsequently, the same group purified the Pif1 protein and showed that it possesses ssDNA-dependent ATPase and 5’-3’ helicase activities (Lahaye et al., 1993; 1991). The first clue that Pif1 is involved in the maintenance of the nuclear genome (as opposed to the mitochondrial genome) came from a genetic screen of yeast mutants that frequently lost expression of subtelomeric genes (Schulz and

Zakian, 1994). The authors identified Pif1 as an inhibitor of telomere elongation and telomere healing, which refers to de novo telomeric DNA synthesis at DSBs occurring at non-telomeric sites. Since then, Pif1 has been implicated in several other molecular activities on nuclear DNA, including unwinding of G-quadruplex structures (Ribeyre et al., 2009), Okazaki fragment maturation during DNA replication (Budd et al., 2006; Pike et al., 2010; Rossi et al., 2008), and blocking of replication of ribosomal DNA (rDNA) (Ivessa et al., 2000).

In budding yeast, PIF1 has a paralog called RRM3, which was first identified in a screen as a mutant that stimulates mitotic recombination in rDNA (Keil and McWilliams, 1993). Rrm3 has since been shown to be a stable component of the yeast replisome (Azvolinsky et al., 2006), perhaps via an interaction with PCNA (Bochman et al., 2010). Interestingly, Rrm3 appears to have opposing activity to Pif1 at yeast rDNA, where it promotes DNA replication through the replication fork barriers (Ivessa et al., 2000). In addition, there is a considerable amount of data supporting the notion that Rrm3 disrupts stable protein-DNA complexes to promote DNA replication through various genomic sites, including tRNA genes, centromeres, telomeres, inactive replication origins, and transcriptional silencers (Azvolinsky et al., 2009; Ivessa et al.,

2003; 2002; Torres et al., 2004). Thus, Rrm3 provides an interesting model of an SF1Bα enzyme that utilizes ATP-driven translocation to dissociate ssDNA-bound proteins rather than unwind dsDNA.

1.3.3 HELB

The chief focus of this dissertation is DNA helicase B (HELB), which is a SF1Bα ATP- dependent helicase/translocase that belongs to the Pif1-like family. I performed BLAST homology searches and amino acid sequence alignments of HELB across a wide variety of organisms with published genomic sequences. Although highly conserved among vertebrates (as well as some cephalochordates and molluscs), no obvious HELB orthologs are found in the popular genetic model organisms D. melanogaster, C. elegans, S. cerevisiae, or S. pombe.

Interestingly, the amino acid sequence within the helicase domain is most similar to the prokaryotic enzymes RecD and TraI (Fairman-Williams et al., 2010).

HELB, schematically depicted in Figure 1.5, was originally identified biochemically as a major source of DNA-dependent ATPase activity in the mouse mammary carcinoma cell line

FM3A (Seki et al., 1987; Tawaragi et al., 1984). Subsequently, HELB was found to be mutated in a temperature-sensitive mutant of this cell line (Seki et al., 1995; Tada et al., 2001). Fanning and colleagues were the first to clone the human HELB cDNA and to purify the recombinant protein (Taneja, 2002). The same group subsequently demonstrated that the intracellular localization of HELB is regulated by CDK-dependent phosphorylation: HELB is mostly nuclear during G1 phase and predominantly cytoplasmic during S/G2 phases (Gu et al., 2004; Spencer et al., 2013). HELB is also a putative target of ATM/ATR-mediated phosphorylation (Matsuoka et al., 2007) and was shown to bind directly to RPA (Guler et al., 2012). Despite being one of the earliest helicase enzymes to be identified (Seki et al., 1987; Tawaragi et al., 1984), a clear biological function for HELB remains unknown.

Figure 1.5 DNA helicase B (HELB) Schematic representation of HELB. Important amino acid residues are indicated for the human protein (equivalent residues in mouse HELB are shown in brackets). These include residues within the Walker A and Walker B ATPase motifs (K481 and E591, respectively), the RPA- interacting region (E499, D506, D510), the ssDNA-binding residues (N785, N834) and the nuclear export signal (NES; V1061, L1065, L1068, L1070). The putative locations of multiple target sites of ATM/ATR- and CDK-mediated phosphorylation are indicated but not specified in this schematic.

1.6 Hypothesis

Recent studies have significantly expanded our understanding of the initiation of DNA end resection, the molecular machinery involved in its execution, and its regulation throughout the cell cycle. However, as highlighted by Symington and Gautier in their authoritative review of the field (2011), it remains unknown whether there exist mechanisms that curtail resection after it is launched. I hypothesized that such processes would represent an effective way to modulate the extent of DNA end resection and that a negative feedback pathway coordinated by RPA would be an especially attractive means to control the formation of ssDNA overhangs, since the accumulation of RPA at resected DSB sites correlates with the extent of resection.

Chapter 2: HELB is an RPA-interacting protein that limits resection in human cells

2.1 Searching for regulators of DNA end resection: Exploring the RPA interactome

The ssDNA generated by resection is rapidly coated by the RPA complex, which together contains six oligonucleotide/oligosaccharide-binding (OB) fold domains, four of which are directly involved in ssDNA binding (Bochkarev et al., 1997; Bochkareva et al., 2001; Murzin,

1993). Therefore, the accumulation of RPA on ssDNA during HR usually correlates with end- resection.

I postulated that a parsimonious mechanism to regulate the extent of ongoing end resection would be a negative feedback loop, which is triggered by the accumulation of RPA on ssDNA (Figure 2.1A). To identify potential RPA-dependent resection inhibitors, I first sought to identify high-confidence RPA-interacting proteins following DNA damage. Each RPA subunit was tagged with the Flag epitope at its N-terminus and stable human embryonic kidney 293

(HEK293)-derived cell lines were generated using the Flp-In/T-Rex system (O'Donnell et al.,

2010).

To induce DNA damage, the cells were treated with neocarzinostatin (NCS), a radiomimetic drug, for 3 hours prior to harvest. The cells were harvested by passive lysis assisted by two freeze-thaw cycles and immunoprecipitation (IP) of the three RPA subunits was carried out using mouse anti-Flag-M2-coupled magnetic beads (Kean et al., 2012). The precipitated proteins were analyzed by tandem mass spectrometry (MS) using the LTQ-Orbitrap Velos instrument by Zhen-Yuan Lin and Anne-Claude Gingras. Twelve IP-MS experiments were carried out (three biological replicates for each RPA subunit and control Flag-empty vector- expressing cell lines). The proteins were identified using Mascot and high-confidence interactions of each RPA subunit were determined using the Significance Analysis of

INTeractome (SAINT) method (Choi et al., 2011).

A total of 26 proteins were found to interact with two or all three of the RPA subunits

(Figure 2.1B). The complete results of the SAINT analysis for all samples reported here can be found in Appendix 1. This list comprised many known RPA interactors such as DNA polymerase α (POLA1/2, PRIM1/2) (Dornreiter et al., 1992), the Bloom syndrome and Werner syndrome helicases (BLM and WRN, respectively) (Brosh et al., 2000), the SMARCAL1 helicase (Bansbach et al., 2009; Ciccia et al., 2009; Yuan et al., 2009; Yusufzai et al., 2009) and the UNG glycosylase (Nagelhus et al., 1997). Previously unrecognized RPA interacting proteins, such as ETAA1 and ZUFSP, were also identified (Figure 2.1B).

2.2 HELB inhibits resection in human cells

To mine the RPA interactome for novel resection inhibitors, siRNA pools targeting 20 of the high-confidence RPA interactors identified by IP-MS were screened in two orthogonal assays that are responsive to changes in DNA end resection. First, Jordan Young (Durocher lab) monitored the formation of foci of RPA2 phosphorylated at Ser4 and Ser8 (pRPA2 S4/S8) following NCS treatment, using high-content microscopy (Figure 2.2A). RPA2 S4/S8 phosphorylation occurs when RPA is bound to ssDNA at DNA damage sites (Maréchal and Zou,

2015). In order to restrict our analysis to cells in S/G2 phases, pRPA2 S4/S8 foci were assessed in a U2OS cell line carrying the FUCCI cell cycle reporter Geminin-mAG (Sakaue-Sawano et al., 2008). The stability of the Geminin protein is regulated by the opposing functions of the

SCFSKP2 and APCCDH1 ubiquitin ligases, which lead to Geminin accumulation in the nucleus in early S phase and its complete degradation by early G1 phase of the subsequent cell cycle

(Bashir et al., 2004; Sakaue-Sawano et al., 2008). Thus, pRPA2 S4/S8 foci were assessed specifically in Geminin-mAG-positive cells, to hone in on the S/G2 phases during which

Figure 2.1 The RPA interactome following DNA damage (A) A simplified model of a putative RPA- dependent negative feedback mechanism that restrains DNA end resection. I, inhibitor. (B) Visualization of the interactions of each of the RPA subunits and HELB. The combined interaction list for each of the baits was filtered by SAINTexpress (version 3.3) and displayed as a dot plot summarizing the data (Knight et al., 2015). The relative abundance is determined by the size of the circle, the averaged spectral counts (capped to a maximum of 50) by the shade of blue within the circle, and the false discovery rate (FDR) for each individual bait-prey interaction is represented by the outer line (high confidence is FDR ≤ 1%, medium confidence is FDR ≤ 5%, low confidence is FDR >5%).

resection is most active. In a second assay, I assessed DSB repair by HR using the direct repeat

(DR)-GFP reporter in HeLa cells, which uses GFP fluorescence as a readout of the frequency of

HR-mediated repair of a single, exogenously-induced DSB (Pierce et al., 1999). As controls, along with a non-targeting siRNA, I used siRNAs against CtIP and RIF1 whose knockdown cause lower and higher levels of resection, respectively. The primary data for both assays is shown in Table 2.1. When the average of two biological replicates for each assay was analyzed on a 2D scatter plot, three siRNA pools led to increases in both assays that were equal or superior to those seen with the depletion of RIF1 (Figure 2.2B). Those three siRNA pools targeted BRIP1/FANCJ, the alternative RFC subunit CHTF18, and HELB.

Since BRIP1 is known to promote HR (Litman et al., 2005), the phenotype observed was likely due to an off-target effect, thus BRIP1 was not followed further. Of the remaining two candidates, I focused on characterizing HELB because its function in genome maintenance is poorly understood. To further examine its involvement in resection, I assessed whether HELB depletion affected DSB repair by single-strand annealing (SSA) using the strand annealing (SA)-

GFP reporter, which uses GFP fluorescence as a readout of the frequency of SSA-mediated repair of a single, exogenously-induced DSB (Gunn and Stark, 2012; Stark et al., 2004). SSA occurs when extensive end resection reveals complementary DNA sequences that are annealed and processed to generate interstitial deletions (Aparicio et al., 2014). Because it is a RAD51- independent process, conditions that increase both SSA and HR (as in the DR-GFP assay) are strongly indicative of increased resection (Stark et al., 2004). I found that the siRNA pool and two out of four individual siRNAs targeting HELB efficiently depleted the protein and resulted in increased SSA by 2-fold or more (Figure 2.2C).

Table 2.1 Quantitation of two orthogonal siRNA-based screens responsive to changes in DNA end resection Controls are shaded in blue. Twenty of the high-confidence RPA-interacting proteins were screened.

DR-GFP (% GFP-positive cells) Mean pRPA2 S4/S8 foci per nucleus siRNA Rep 1 Rep 2 Average Rep 1 Rep 2 Average CTRL 2.19 1.53 1.86 17.48 26.98 22.23 CTIP 0.37 0.25 0.31 1.49 2.99 2.24 RIF1 5.7 3.01 4.36 42.08 51.69 46.89 BLM 5.53 2.85 4.19 15.16 22.37 18.76 BRIP1 5.75 2.56 4.16 74.93 96.75 85.84 CCDC111 3.49 2 2.75 25.98 26.98 26.48 CHTF18 5.78 3.43 4.61 55.83 69.74 62.79 ETAA1 7.61 4.45 6.03 25.27 22.15 23.71 HELB 8.01 5.49 6.75 51.81 73.48 62.64 HERC2 4.91 3.08 4.00 18.55 20.37 19.46 MSH3 4.42 2.05 3.24 28.15 22.63 25.39 PARP1 4.9 3.57 4.24 17.43 30.38 23.90 POLA1 2.86 2.08 2.47 20.09 31.37 25.73 PRIM1 3.68 2.43 3.06 22.16 25.18 23.67 RAD52 3.17 1.74 2.46 28.29 39.79 34.04 RFWD3 2.88 1.92 2.40 42.84 58.86 50.85 RMI1 4.22 2.33 3.28 9.69 8.17 8.93 RPAIN 1.46 0.36 0.91 62.94 70.15 66.54 SMARCAL1 3.31 1.32 2.32 37.19 42.60 39.90 TOP3A 3.78 2.63 3.21 20.36 20.65 20.51 UNG 1.39 0.82 1.11 24.02 19.68 21.85 WRN 3.78 2.71 3.25 9.58 5.98 7.78 ZUFSP 4.77 3.2 3.99 18.91 25.69 22.30

To ensure that the hyper-recombination phenotype observed upon HELB depletion was not a result of siRNA off-target effects, I generated a Flag-tagged HELB expression vector, which produces an mRNA resistant to siRNA#2. Expression of this siRNA-resistant HELB construct in HELB-depleted HeLa DR-GFP cells confirmed that the increase in HR imparted by

HELB depletion is not due to siRNA off target effects and can be rescued by the reintroduction of HELB (Figure 2.2D). Together, the results of these assays strongly suggest that HELB depletion causes an increase in DNA end resection in human cells. 30

Figure 2.2 HELB restricts resection in human cells (A) Representative micrographs of the high-content microscopy screen to measure pRPA2 S4/S8 focus formation in U2OS cells following the siRNA-mediated depletion of RPA-interacting proteins. DAPI was used to segment nuclei and Geminin-AG was used to segment cells in S/G2 phases of the cell cycle. The mean number of pRPA2 S4/S8 foci per nucleus was determined using a spot segmentation algorithm. (B) High confidence RPA interactors found by IP-MS (see Table 1 for primary data) were screened in assays that monitored pRPA2 S4/S8 focus formation (y axis) and HR via the DR-GFP reporter (x axis). The results of each assay were normalized to the non-targeting siRNA (siCTRL). The positive and negative controls were siRNAs against RIF1 and CtIP,

respectively. The shaded area highlights three siRNAs that scored positive in both assays. Each point represents the mean of two biological replicates for each assay. (C) SSA frequency, as determined by the SA-GFP reporter in U2OS cells transfected with a non-targeting siRNA (CTRL) or siRNAs targeting the indicated proteins. Data are presented as mean +/- SEM, n = 3. (p) represents the pooled HELB siRNAs; the individual HELB siRNAs are denoted by a number. The dashed grey line indicates the SSA frequency of the siCTRL sample. Immunoblots monitoring the efficiency of HELB depletion using the indicated siRNAs are shown below the graph. (D) HR frequency, as determined by the DR-GFP reporter in HeLa cells transfected with non- targeting (siCTRL) or HELB-targeting (siHELB) siRNAs in combination with empty (EV), or siRNA-resistant Flag-HELB (H) vector. Data are presented as mean +/- SEM, n = 3. The dashed grey line indicates the SSA frequency of the siCTRL+EV+I-SceI sample. Immunoblots of a representative experiment are shown below the graph.

2.3 HELB is recruited to sites of DSBs in an RPA-dependent manner

A basic cleft in the N-terminal OB fold of RPA1 mediates the physical interaction between RPA and acidic peptides on ATRIP, RAD9, and MRE11 (Xu et al., 2008). Using elegant biochemical and NMR chemical shift experiments, Fanning and colleagues identified a conserved acidic patch downstream of the Walker A motif of the HELB protein (Figure 2.3A), which directly interacts with the RPA1 N-terminal basic cleft (Guler et al., 2012). To assess the reported interaction domain, I expressed Flag-HELB constructs in HEK293T cells and carried out IP using an anti-Flag antibody. Wild-type HELB efficiently coprecipitated with endogenous

RPA, whereas this interaction was significantly reduced when the 3xA (E499A, D506A, D510A) acidic patch mutant was expressed (Figure 2.3AB). In addition, our collaborators Hemanta

Adhikary, Jana Krietsch and Jean-Yves Masson (Laval University) further validated this data using pull-down experiments with purified recombinant HELB (WT and 3xA) and RPA proteins

(Figure 2.3C-E).

Figure 2.3 Acidic residues in the helicase domain of HELB mediate its interaction with RPA (A) HELB protein sequences from the indicated species were aligned using the MAFFT algorithm. The amino acid sequence similarity between species is represented by the shade of blue. The red asterisks indicate the conserved acidic residues that were mutated to alanine to generate the 3xA mutant. Human, Homo sapiens; Chimpanzee, Pan troglodytes; Mouse, Mus musculus; Chicken, Gallus gallus; Zebrafish, Danio rerio. (B) Flag-HELB constructs or an empty vector (EV) control were transiently overexpressed in HEK293T cells. IP was performed using anti-Flag antibodies. Deletion of the entire helicase domain (ΔHel) completely abrogated the RPA interaction. The 3xA mutation greatly reduced the RPA interaction. (C) Recombinant HELB protein variants purified from SF9 insect cells were separated by SDS- PAGE and visualized using silver staining. (D) Purified recombinant human RPA complex. 600 ng of purified protein was separated by SDS-PAGE and visualized using Coomassie blue staining. (E) HELB binds to RPA-bound ssDNA beads. Streptavidin beads loaded with 100 ng of biotinylated ssDNA were incubated with RPA (20 nM) or with buffer prior to incubation with the indicated HELB proteins (20 nM). The beads were then washed and the bound proteins eluted and separated by SDS-PAGE followed by immunoblotting to detect HELB and RPA. 20 ng of HELB was added in the input lane. 33

HELB accumulates on chromatin in response to various types of genotoxic stress (Guler et al., 2012). This observation, coupled with the finding that HELB interacts with RPA, suggests that HELB may localize to DSB sites by contacting the RPA-ssDNA filament. To test this possibility directly, I performed laser microirradiation experiments in U2OS-derived cell lines that were stably transfected with N-terminal GFP-tagged HELB. In these assays, the nuclei of cells were irradiated with 8 mW power from a 355 nm (UVA) laser source to induce DSBs. I found that GFP-HELB accumulates at laser microirradiation-induced DSBs with a staining pattern that is completely coincident with that of RPA in the central region of the γH2AX stripe

(Figure 2.4A). This sub-compartment of laser microirradiation induced DNA damage has been previously shown to correspond to ssDNA, where RAD51, RAD52, BRCA2 and ATR also accumulate (Bekker-Jensen et al., 2006).

To determine whether this localization is dependent on DNA end resection, I used siRNA-mediated knockdown to deplete CtIP, which is critical for the initiation of resection. CtIP depletion was efficient (Figure 2.4B) and resulted in a significant reduction in the recruitment of both RPA and GFP-HELB to laser microirradiation-induced DSBs (Figure 2.4CD). Finally,

Cristina-Escribano Díaz (Durocher lab) was also able to detect the recruitment of endogenous murine HELB to microirradiation sites in a portion of mouse embryonic fibroblasts (MEFs)

(Figure 2.4EF). Together, these results suggest that HELB is recruited to DSBs in a resection- dependent manner.

Figure 2.4 HELB is recruited to DSB sites in a CtIP-dependent manner (A) U2OS cells expressing GFP-HELB were microirradiated with a UVA laser. Four hours after irradiation, cells were fixed and processed for RPA2 immunofluorescence and counterstained with DAPI to delineate the outline of the nucleus. Scale bar = 10 µm. (B) U2OS cells expressing GFP-HELB were transfected with non-targeting (siCTRL) or CtIP- targeting (siCtIP) siRNA for 72 hr. Whole cell extracts were immunoblotted using an anti- CtIP antibody to confirm efficient depletion of the protein. Tubulin was used as a loading control. (C) U2OS cells expressing GFP-HELB were transfected with non-targeting (siCTRL) or CtIP- targeting (siCtIP) siRNA for 72 hr. The cells were laser microirradiated and processed for immunofluorescence as in (A). Scale bar = 10 µm. (D) Quantitation of the fluorescence intensities of RPA2 (left) and GFP-HELB (right) stripes in U2OS cells transfected with the indicated siRNA. Each dot represents a nucleus analyzed. Stripe intensity was normalized to the mean nuclear background intensity in each cell (represented by the dashed gray line). The distributions were compared using the Mann- Whitney U test.

(E) Validation of the anti-mouse HELB polyclonal antibody (181A) by immunofluorescence analysis of wild-type and Helb-/- MEFs. Scale bar = 10 µm. (F) MEFs were laser microirradiated as in (A), then processed for endogenous HELB and γH2AX immunofluorescence. Scale bar = 10 µm.

To determine the requirements for the recruitment of HELB to DSB sites, I generated several HELB mutants using site-directed mutagenesis. Helicase activity was disrupted by mutating either the Walker A (K481A) or Walker B (E591Q) ATPase motifs, which are responsible for ATP binding and hydrolysis, respectively (Taneja, 2002). Abolishing helicase activity had no effect on HELB recruitment to sites of DNA damage (Figure 2.5A). In contrast, disruption of the RPA interaction using the 3xA mutant significantly decreased the recruitment of HELB to sites of DNA damage (Figure 2.5AB). Importantly, all of the HELB variants in this assay were expressed at similar levels (Figure 2.5C). These results show that HELB is recruited to sites of DNA damage in an RPA-dependent manner, but independent of its catalytic activity.

Figure 2.5 HELB is recruited to DSB sites in an RPA-dependent manner (A) U2OS cells expressing the indicated GFP-HELB derivatives were laser microirradiated as in Figure 2.4A and processed for RPA2 and γH2AX immunofluorescence. Scale bar = 10 µm. (B) Fluorescence intensities of GFP-HELB (WT or 3xA mutant) stripes in U2OS cells were quantitated as in Figure 2.4D. Each dot represents a nucleus analyzed. The dashed gray line represents the background nuclear fluorescence intensity. The distributions were compared using the Mann-Whitney U test. (C) Whole cell extracts from U2OS cells expressing the indicated GFP-HELB variants or empty vector (EV) control were immunoblotted using antibodies against GFP and HELB to assess the levels of protein expression. Tubulin was used as a loading control.

Chapter 3: HELB limits BLM-DNA2- and EXO1-dependent resection

3.1 Generating the Helb-/- mouse

Due to the inherent limitations in studying endogenous HELB functions and genetic interactions using siRNA-mediated knockdown in human cell lines, I decided to develop a mammalian model system for loss of HELB function, by knocking out the Helb gene in the house mouse, Mus musculus. Mouse embryonic stem (ES) cells were obtained from the

Knockout Mouse Project Repository at University of California, Davis. These ES cells carry a heterozygous deletion of the entire Helb gene sequence, which is replaced with a cassette containing the β-galactosidase (lacZ) and neomycin resistance (neor) genes (Figure 3.2AB). The knockout allele is officially designated Helbtm1(KOMP)Vlcg, hereafter referred to simply as Helb -.

The Helb+/- ES cells were aggregated with wild-type (WT) albino morulas (strain CD-1) to make chimeric animals. The chimeras were tested for germline transmission of the Helb deletion allele

(Figure 3.1A). The germline transmitting-chimeras were then backcrossed to the parental strain

WT animals (strain C57BL/6N) to produce heterozygous Helb+/- mice in an isogenic background

(Figure 3.1B). Finally, Helb+/- mice were crossed to generate homozygous knockout Helb-/- mice.

Helb-/- mice are viable and produced at the expected Mendelian frequencies (Table 3.1).

Furthermore, Helb-/- animals are fertile and phenotypically normal under unchallenged conditions (Figure 3.2C). Dr. Hibret Adissu of the Centre for Modeling Human Disease

Pathology Core (Toronto Centre for Phenogenomics) performed a full histopathological evaluation of two WT and two Helb-/- littermates euthanized at 15 weeks of age. All of the histological findings were similar between the WT and Helb-/- animals, and were considered incidental or attributable to the C57BL/6N strain background. The full pathology report can be found in Appendix 2.

Figure 3.1 Generation of the Helb-/- mouse (A) Breeding strategy for germline transmission testing of chimeric mice. The offspring of germline-transmitting chimeras can be identified at birth (prior to coat fur growth) by the presence of colored eyes (arrows). (B) Breeding strategy for generating the knockout Helb-/- mouseline in an isogenic background strain (C57BL/6N). The percentages indicate the expected genotype frequencies based on Mendelian inheritance. P, parental; F1, first filial; F2, second filial.

Table 3.1 There is no neonatal lethality associated with Helb deletion in mice

Cross Helb+/- x Helb+/- Total offspring 66 Offspring genotypes Helb+/+ Helb+/- Helb-/- Expected (Mendelian) 16.5 33 16.5 Observed 17 33 16

In order to sensitize the mice for tumor development and facilitate survival studies, I generated Helb-/- animals in a p53-heterozygous background. To determine whether loss of Helb has any effect on mouse survival, I set up a cohort consisting of 14 Helb+/+ p53+/- and 13 Helb-/- p53+/- animals for aging without any environmental or experimental perturbations. The Helb deletion did not result in a significant difference in the tumour-free survival under these conditions (Figure 3.2D). Please see the Discussion, Chapter 5.5 for potential reasons for our inability to observe a phenotype at the organismal level in Helb-null mice, as well as future directions to address these possibilities.

Figure 3.2 Helb-/- mice are viable and do not display any phenotypes in the absence of environmental or experimental challenges (A) Schematic of the Helbtm1(KOMP)Vlcg allele. (B) Confirmation of the genotypes by allele-specific PCR. (C) The Helb-/- mice are viable and phenotypically normal under unchallenged conditions. (D) Tumor-free survival of unchallenged Helb+/+ p53+/- (n = 14) and Helb-/- p53+/- (n = 13) mice displayed using a Kaplan-Meier plot. No significant difference between the survival distributions was detected using the log-rank test.

3.2 Loss of HELB promotes SSA in mouse cells

Mouse embryonic fibroblasts (MEFs) were derived from E13.5 embryos from a Helb+/- x

Helb+/- cross. Immunoblotting of whole cell extracts confirmed that the knockout allele leads to a complete loss of HELB (Figure 3.3A).

To determine whether HELB is also an antagonist of SSA in mouse cells, I utilized the single-strand annealing traffic light reporter (SSA-TLR) (Kuhar et al., 2014). A simplified schematic of this assay is shown in Figure 3.3B. First, MEFs were transduced with the reporter plasmid pCVL.SSA-TLR using lentiviral infection with the VSV-G viral envelope and the psPAX2 viral packaging vector. Lentiviral particles were packaged in HEK 293T cells.

Following puromycin-mediated selection of the transductants, MEFs carrying the SSA-TLR reporter cassette were electroporated with an expression plasmid carrying BFP-tagged I-SceI meganuclease, which induces a single DSB within the SSA-TLR cassette in each cell. 48 hr after electroporation, the samples were analyzed by flow cytometry. BFP-positive cells were gated in order to focus the analysis only on I-SceI-expresssing cells. This BFP-positive cell population was analyzed for iRFP fluorescence, which is indicative of a successful SSA-mediated repair of the I-SceI-induced break (Figure 3.3B).

Analysis of SSA in WT and Helb-/- MEFs electroporated with two different I-SceI expression constructs (named “Donor-Sce-BFP” and “Sce-IRES-BFP” for short) showed that

HELB is also an SSA antagonist in mouse cells, consistent with a role as an inhibitor of resection

(Figure 3.3C). To assess whether this SSA phenotype is specifically due to the loss of HELB protein in Helb-/- MEFs, I transduced these cells with a pMSCV-GFP-HELB plasmid using retroviral infection. Ecotropic retroviral particles were packaged in the HEK 293T-derived cell line Platinum-E (Morita et al., 2000). Following puromycin-mediated selection of the transductants, GFP-HELB was found to be expressed at near-physiological levels in Helb-/-

MEFs (Figure 3.3D), which resulted in a rescue of the associated hyper-SSA phenotype (Figure

3.3E). Importantly, there was no variation in the cell cycle distribution (as determined by propidium iodide (PI) staining), which could account for the observed differences in SSA frequencies between these samples (Figure 3.3F).

Taken together, these results indicate that HELB is an antagonist of SSA-mediated repair of DSBs in both human and mouse cells. This is consistent with its proposed role as a negative regulator of DNA end resection. However, when examining pRPA2 S4/S8 by immunoblotting, I did not observe an increase in RPA2 phosphorylation following NCS treatment in Helb-/- MEFs, despite its being significantly increased in 53bp1-/- cells, which were used as a positive control because they have a documented hyper-resection phenotype (Figure 3.3G) (Bunting et al., 2010).

Similar results were obtained when the cells were treated with the topoisomerase I poison camptothecin instead of NCS. The potential reasons for this observation are examined in the

Discussion, Chapter 5.2.

Figure 3.3 HELB antagonizes SSA in MEFs (A) Whole cell extracts from Helb+/+, Helb+/-, and Helb-/- MEFs were immunoblotted using antibodies against HELB and tubulin. (B) A simplified schematic of the single-strand annealing traffic light reporter (SSA-TLR). (C) SSA frequency in Helb+/+ and Helb-/- MEFs carrying an integrated SSA-TLR cassette and electroporated with one of two different I-SceI meganuclease expression plasmids. Data are presented as mean +/- SEM, n = 3. (D) Whole cell extracts from Helb+/+, Helb-/-, or GFP-HELB-complemented Helb-/- MEFs were analyzed by immunoblotting with antibodies against HELB and tubulin. (E) SSA frequency in Helb+/+, Helb-/-, or GFP-HELB-complemented Helb-/- MEFs was quantitated using the SSA-TLR reporter. Data are presented as mean +/- SEM, n = 3. (F) To control for variation in cell cycle distribution, Helb+/+, Helb-/-, or GFP-HELB- complemented Helb-/- MEFs (all with an integrated SSA-TLR cassette) were processed for propidium iodide (PI) staining and analyzed by flow cytometry (bottom panel). At least 10000 cells were analyzed for each sample. (G) MEFs derived from 53bp1+/+ and 53bp1-/- littermates or Helb+/+ and Helb-/- littermates were treated with the indicated doses of NCS for 3 hr. Whole cell extracts were prepared and analyzed by immunoblotting with antibodies against pRPA2 S4/S8 and tubulin, which was used as a loading control.

3.3 HELB inhibits CtIP- and ATM-dependent resection

Since the pRPA2 S4/S8 by immunoblotting results in Helb-/- MEFs (Figure 3.3G) were at odds with our SSA data and our results in human cells, I turned to monitoring the formation of ssDNA by quantitating 5-bromo-2’-deoxyuridine (BrdU) levels under native conditions (Figure

3.4A) to have a more definitive view of resection in Helb-/- cells. BrdU quantitation is the most direct measure of ssDNA formation by resection, since it does not rely on the binding and post- translational modification of downstream factors (such as RPA binding and phosphorylation), nor does it require a specific DSB repair outcome (such as SSA or HR), which are influenced by additional variables. Therefore, I adapted and optimized for use in MEFs a previously reported procedure for native BrdU quantitation by flow cytometry (Nishi et al., 2014).

Using this more direct readout of ssDNA formation, I detected a robust increase in resection after NCS treatment in Helb-/- cells that was similar to that observed in 53bp1-/- cells

(Figure 3.4BG). This increase was not caused by higher BrdU incorporation in Helb-/- MEFs

(Figure 3.4B), nor was it due to a greater induction of DNA damage as assessed by quantitation of γH2AX levels (Figure 3.4CD) and the neutral comet assay (Figure 3.4E). Cell cycle distribution was not altered in the knockout MEFs (Figure 3.4F), indicating that the observed hyper-resection was not due to aberrant accumulation/enrichment of cells in S/G2 phases, when resection is normally activated. Critically, I found that the increase in ssDNA observed in Helb-/- cells following NCS treatment was entirely dependent on CtIP using lentiviral transduction of a

CtIP-targeting short hairpin RNA (shRNA) (Figure 3.4H). This finding confirms that the ssDNA signal detected using the native BrdU assay was in fact due to end resection.

ATM kinase activity has been shown to be required for the unscheduled resection observed in H2AX-/- murine pre-B cells (Helmink et al., 2011) and in 53bp1-/- murine B lymphocytes (Yamane et al., 2013). In addition, ATM activity is critical for the suppression of poly[ADP-ribose] polymerase (PARP) inhibitor sensitivity, a measure that reflects proficiency of

DSB repair by HR, in BRCA1-deficient 53bp1-/- cells (Bunting et al., 2010). To determine whether the increased resection in Helb-/- MEFs was dependent on ATM activity, I titrated the

ATM inhibitor KU55933 at increasing concentrations into the cell growth media 1 hr before addition of NCS, and performed the native BrdU assay. I found that the hyper-resection phenotype of Helb-/- cells was ATM-dependent (Figure 3.4G).

Taken together, these results demonstrate that flow cytometry-based detection of incorporated BrdU under native conditions provides a robust means for quantitating DNA end resection and that HELB is an antagonist of resection in human and mouse cells alike.

Figure 3.4 HELB is a resection inhibitor in MEFs (A) A simplified schematic depicting the native BrdU resection assay. Note that the antibody can only detect the incorporated BrdU epitope when it is exposed in the context of ssDNA. (B) MEFs of the indicated genotype were grown in the presence of 30 µM BrdU for 24 hr prior to incubation in the absence (-NCS) or presence of 200 ng/mL NCS (+NCS) for 3 hr. Cells were then fixed and processed for detection of BrdU by flow cytometry without denaturation. A parallel sample was also processed with acid denaturation (2N HCl for 30 min) to detect total BrdU incorporation. Shown are representative histogram plots of 10000 cells analyzed for each condition. (C) WT and Helb-/- MEFs were treated with NCS (200 ng/mL) for 1 hr and then processed for γH2AX immunofluorescence. DNA was counterstained with DAPI. (D) WT and Helb-/- MEFs were treated with the indicated concentrations of NCS for 3 hr and then processed for γH2AX immunofluorescence. DNA was counterstained with DAPI. The mean nuclear intensity of γH2AX was quantitated in 8000 nuclei per sample using high content microscopy. (E) WT and Helb-/- MEFs were treated with the indicated concentrations of NCS for 3 hr and then processed for analysis by the neutral comet assay. The tail moment was quantified using the OpenComet plugin in ImageJ software. The tail moment value consists of the product of tail length (measured in pixels) and the percentage of total DNA in the tail. (F) WT and Helb-/- MEFs were pulse-labeled with BrdU for 2 hr prior to harvest and fixation. The samples were processed for PI staining coupled with BrdU immunofluorescence after acid-denaturation and analyzed by flow cytometry. Cells in S phase were gated and expressed as a percentage of all cells in the PI versus BrdU plots. At least 10000 cells were analyzed per sample. (G) Quantitation of the flow cytometry data shown in (B), presented as the mean BrdU fluorescence intensity following NCS treatment, normalized to the untreated control cells for each sample (+/- SEM, n = 5). Wild type MEFs corresponding to each mutant were derived from their respective littermates. (H) Quantitation of resection by the native BrdU assay of Helb+/+ and Helb-/- MEFs transduced with lentiviruses expressing non-targeting (CTRL) or CtIP-targeting shRNAs and treated with NCS. Data were normalized as in (G) and presented as mean +/- SEM, n = 3. (I) Quantitation of resection by the native BrdU assay of Helb+/+ and Helb-/- MEFs treated with NCS. The indicated concentration of the ATM inhibitor KU55933 was added to the growth medium 1 hr prior to the addition of NCS and maintained in the medium during the NCS incubation. The mean fold increase in BrdU fluorescence intensity following treatment with NCS was calculated for each sample as in (G). Data were normalized to the KU55933- untreated Helb-/- sample and presented as mean +/- SEM, n = 3.

3.4 HELB inhibits BLM-DNA2- and EXO1-mediated long-range resection

I re-introduced untagged murine HELB (and several mutated variants described below) in

Helb-/- MEFs by retroviral transduction of pMXs-IRES-GFP plasmids. Ecotropic retroviral particles were packaged in the HEK 293T-derived cell line Platinum-E (Morita et al., 2000).

After transduction, the MEF populations were expanded and GFP-expressing transductants were sorted using fluorescence-activated cell sorting (FACS) on the Beckman Coulter MoFlo XDP platform at the Princess Margaret Hospital Flow Cytometry Facility. Following the FACS, all transduced HELB proteins were expressed, albeit at somewhat different levels (Figure 3.5A).

Using the native BrdU assay, I found that the WT protein restored DNA end resection to the levels seen in Helb+/+ cells, confirming that loss of HELB increases resection (Figure 3.5B).

In contrast, expression of the catalytically inactive HELB K462A and E571Q mutants

(equivalent to human K481A and E591Q, respectively) or the RPA binding-defective 3xA mutant did not rescue the hyper-resection phenotype (Figure 3.5B). I also tested a mutant, which impairs the interaction between the helicase domain and ssDNA (N768A/N809A or 2NA) based on the structure of the closely related bacterial SF1B helicase RecD2 (Saikrishnan et al., 2008;

2009). The HELB 2NA mutant could not restore resection to WT levels (Figure 3.5B). Note that the variance in the level of expression of the untagged HELB mutants (Figure 3.5A) (resulting from the fact that the MEF populations were FACS-sorted based on GFP expressed from an internal ribosome entry site (IRES) cannot explain the resection phenotypes observed (Figure

3.5B). Therefore, HELB requires its catalytic activity as well as RPA- and ssDNA-binding to suppress end resection (Figure 3.5B).

Long-range resection is catalyzed by two redundant nuclease machineries: EXO1 and

BLM-DNA2 (Gravel et al., 2008; Nimonkar et al., 2011). To determine whether the increased resection observed in the absence of HELB was due to the action of either of these nucleases, I transduced Exo1- and Blm-targeting shRNAs into Helb-/- MEFs using three rounds of retroviral infection immediately before each iteration of the native BrdU resection experiment. Efficient depletion of Exo1 and Blm mRNA levels alone or in combination was confirmed using RT-qPCR

(Figure 3.5C). I found that depletion of EXO1 or BLM alone partially decreased resection in

Helb-/- cells but the combined depletion of both EXO1 and BLM completely suppressed resection in Helb-/- cells (Figure 3.5D). I conclude from these results that HELB antagonizes, directly or indirectly, the activity of EXO1 and BLM-DNA2 nucleases during resection.

Figure 3.5 HELB inhibits EXO1- and BLM-DNA2-mediated resection (A) Whole cell extracts from Helb+/+ and Helb-/- MEFs transduced with either an empty retrovirus (EV) or a retrovirus expressing the indicated HELB protein variant were analyzed by immunoblotting with the indicated antibodies. Tubulin was used as a loading control. (B) Quantitation of resection by the native BrdU assay of Helb+/+ and Helb-/- MEFs transduced with either an empty retrovirus (EV) or a retrovirus expressing the indicated HELB protein variant and treated with NCS. Data were normalized as in Figure 3.4G and presented as mean +/- SEM, n ≥ 4. (C) RT-qPCR quantitation of Exo1 and Blm mRNA levels in Helb+/+ and Helb-/- MEFs transduced with a retrovirus expressing a non-targeting shRNA (shCTRL) or shRNAs targeting Exo1 and/or Blm. Expression of Exo1 and Blm were normalized to that of Gapdh. (D) Quantitation of resection by the native BrdU assay of Helb+/+ and Helb-/- MEFs transduced with a retrovirus expressing a non-targeting shRNA (shCTRL) or shRNAs targeting Exo1 and/or Blm, and treated with NCS. Data were normalized as in Figure 3.4G and presented as mean +/- SEM, n = 4.

Chapter 4: HELB participates in the cell cycle regulation of DNA end resection

4.1 HELB does not regulate DSB repair pathway choice

The inhibition of the initiation of DNA end resection by the DSB end protection factors

Ku, 53BP1 and RIF1 is linked to their ability to promote repair of the break by the NHEJ pathway instead of HR (Aparicio et al., 2014; Chapman et al., 2012). Therefore, three separate methods were used to assess whether loss of HELB similarly impacts NHEJ. First, with the help of Alexandre Orthwein (Durocher lab), I tested whether Helb-/- B cells undergo class switch recombination (CSR), the process by which the constant region of the immunoglobulin (Ig) gene is rearranged to switch antibody isotype (Boboila et al., 2012). B lymphocytes were isolated from the spleens of isogenic WT, Helb-/- and 53bp1-/- mice by immunodepletion of CD43- positive cells using magnetic anti-CD43 microbeads. The CD43 antigen is expressed on the surface of most leukocytes (including T cells, natural killer (NK) cells, granulocytes, monocytes and macrophages), and also on hematopoietic stem cells and platelets. CD43 is likewise expressed on activated B cells and plasma cells but is absent from the surface of resting B cells, such as splenic naive B cells.

Isolated naive B cells were stimulated with interleukin-4 (IL-4) and lipopolysaccharides

(LPS) to induce Ig class switching from IgM to IgG1. Four days following stimulation, the percentage of IgG1-positive cells was analyzed by flow cytometry after the cells were gated for

CD45R/B220 (a marker of B cells) and exclusion of propidium iodide (to eliminate dead cells).

This experiment revealed that unlike the case of cells isolated from 53bp1-/- animals, which are severely impaired in CSR (Manis et al., 2004; Ward et al., 2004) HELB deletion did not result in a statistically significant change in CSR (Figure 4.1AB), suggesting that HELB may not influence DSB repair pathway choice between NHEJ and HR. Corroborating this observation,

Cristina Escribano-Díaz (Durocher lab) found that Helb-/- MEFs can form normal levels of ionizing radiation-induced RIF1 foci (Figure 4.1CD).

To assess NHEJ efficiency directly, Cristina Escribano-Díaz used a recently developed end-joining assay, which measures the recircularization of a 545 bp extrachromosomal linear

DNA substrate (Waters et al., 2014). Lig4-/- MEFs were used as a positive control in this assay, since the final step of classical NHEJ (direct re-ligation of broken ends) is catalyzed by LIG4 in complex with XRCC4 and XLF. Lig4-/- MEFs showed a markedly reduced NHEJ efficiency, whereas Helb-/- cells were as competent as their WT counterparts in this assay (Figure 4.1E).

Therefore I conclude that HELB does not influence DSB repair pathway choice, unlike the end protection factors 53BP1, RIF1 and Ku. Although HELB regulates HR efficiency by antagonizing resection, it does not promote NHEJ. These conclusions are consistent with the observation that HELB is recruited to RPA-coated ssDNA overhangs downstream of the initiation of resection (see Chapter 2).

Figure 4.1 HELB is not involved in DSB repair pathway choice (A) A representative CSR experiment. B cells were isolated from the spleens of adult WT, Helb-/- and 53bp1-/- mice, and treated with LPS and IL-4 to induce immunoglobulin class switching from IgM to IgG1. Cells were analyzed 4 days later by flow cytometry using an anti-mouse CD45R/B220 antibody (marker of B cells) and an anti-mouse IgG1 antibody, in conjunction with propidium iodide staining (to exclude dead cells). Each dot plot represents at least 10000 B cells analyzed per sample.

(B) Quantification of the CSR data shown in (A). The data were normalized to the CSR levels of the WT sample in each biological replicate and presented as the mean +/- SEM, n=3. Significance was determined using the Mann-Whitney U test (n.s., not significant). (C) Representative micrographs of Helb-/-, 53bp1-/- MEFs and their WT counterparts fixed 1 hr after irradiation with a 3 Gy dose of IR, and processed for γH2AX and RIF1 immunofluorescence. DNA was counterstained with DAPI and used to delineate the nuclei. Scale bar = 5 µm. (D) Quantitation of RIF1 IR-induced focus formation shown in (C). The data are presented as mean +/- SEM, n = 3. (E) Determination of extrachromosomal NHEJ in Helb-/- (blue), Lig4-/- (red) MEFs and their WT counterparts. The data are presented as mean +/- SEM, n = 3.

4.2 HELB mediates the cell cycle regulation of DNA end resection

The subcellular localization of HELB is controlled by the cell cycle (Gu et al., 2004;

Spencer et al., 2013). In early G1 cells, HELB is predominantly nuclear. As cells approach the

G1/S transition, HELB is phosphorylated at its C-terminus by CDK2, which results in its nuclear export, leaving only a minor fraction of HELB in the nucleus as S phase progresses (Guler et al.,

2012). We observed this phenomenon using immunofluorescence of endogenous HELB in MEFs using two methods. First, HELB loses its nuclear enrichment in Cyclin A-positive cells (Figure

4.2A), which designates cells in S/G2 phases. Second, HELB loses its nuclear enrichment in cells that are undergoing DNA replication (Figure 4.2BC), as assessed following a short pulse of

BrdU incorporation.

At first, the cell cycle-dependent subcellular shuttling of HELB was puzzling, since DNA end resection is restricted to the S/G2 phase of the cell cycle, i.e. at a time when HELB becomes predominantly cytoplasmic. To determine whether forcing HELB to remain in the nucleus in

S/G2 phases could restore normal resection levels in Helb-/- MEFs, I introduced mutations that disrupt the HELB nuclear export signal (V1029A, F1033A, M1036A, L1038A; hereafter referred

to as NESm). We observed a partial nuclear enrichment of HELB-NESm in cells undergoing

DNA replication, i.e. in S phase (Figure 4.2C). The enrichment potently suppressed DNA end resection in Helb-/- cells (Figure 4.2D), despite the NESm mutant being expressed at lower levels than the wild type protein (Figure 3.5A). Together, these results indicate that HELB acts in the nucleus to limit end resection.

Figure 4.2 The nuclear export of HELB promotes resection (A) Representative micrographs of WT MEFs processed for HELB and Cyclin A (labeling S/G2 cells) immunofluorescence. DNA was counterstained with DAPI and used to delineate the nuclei. Scale bar = 10 µm. (B) Representative micrographs of WT MEFs pulsed-labeled with BrdU for 3 hr prior to fixation and processing for HELB and BrdU immunofluorescence after the DNA was denatured using 2N HCl. Cells with positive nuclear BrdU staining were considered to be in S or G2 phases. DNA was counterstained with DAPI and used to delineate the nuclei. Scale bar = 10 µm. (C) WT or Helb-/- MEFs transduced with an empty retrovirus (EV) or retroviruses expressing the indicated variant of HELB were pulse-labeled with BrdU and processed for HELB and BrdU immunofluorescence as in (B). The nuclear intensity of the HELB signal in G1 (BrdU-) and S/G2 (BrdU+) cells was determined and plotted. Each point represents a nucleus analyzed. The red bar represents the mean intensity value. (D) Helb-/- MEFs transduced with either an empty retrovirus (EV) or a retrovirus expressing the indicated variant of HELB were grown in the presence of BrdU for 24 hr prior to the addition of 200 ng/mL of NCS for 3 hr. Cells were fixed and processed for the native BrdU resection assay. Shown are representative histogram plots of 10,000 cells analyzed for each condition. The values within each plot represent the mean fold increase in BrdU fluorescence intensity after NCS treatment (+/-SEM, n = 3). (E) 53bp1-/- MEFs transduced with either an empty retrovirus (EV) or a retrovirus expressing the indicated variant of HELB were processed for native BrdU detection as in (D). Shown are representative histogram plots of 10,000 cells analyzed for each condition. The values within each plot represent the mean fold increase in BrdU fluorescence intensity after NCS treatment (+/-SEM, n = 3).

The above results suggest that CDK2-mediated nuclear export of HELB might be critical for the characteristic upregulation of DNA end resection seen in S phase cells. If this model is correct, expression of the HELB-NESm protein variant should curtail the normal upregulation of resection upon S phase entry. Since the dynamic range of the native BrdU resection assay in wild type cells is small (Figure 3.4B), I assessed whether expression of HELB-NESm in 53bp1-/- cells suppressed their associated hyper-resection phenotype. Indeed, the HELB-NESm protein, but not the catalytically inactive HELB-NESm-K462A mutant, potently blocks the formation of ssDNA in 53bp1-/- MEFs following NCS treatment (Figure 4.2E). Furthermore, overexpression of the wild type protein was sufficient to dominantly suppress resection in 53bp1-/- cells (Figure 4.2E, right panel). These results suggest that DNA end resection is exquisitely sensitive to HELB dosage and further validate the notion that HELB and 53BP1 act in distinct pathways to curtail resection.

To test this possibility directly, I generated Helb-/- 53bp1-/- mice and derived MEFs to assess resection following NCS treatment. I observed that the combined loss of HELB and

53BP1 resulted in an additive increase in DNA end resection (Figure 4.3A) suggesting a non- epistatic relationship between the two genes. Immunoblotting confirmed that each deletion allele resulted in a complete loss of its associated protein (Figure 4.3B) and DNA content analysis (by propidium iodide staining) was used to rule out the possibility that cell cycle variation between the MEF cell lines could account for the observed differences in resection (Figure 4.3C). Taken together, the data indicate that HELB acts independently of 53BP1 to suppress resection and suggest that the reduction in the nuclear abundance of HELB prior to S phase entry is critical for the activation of end resection during this stage of the cell cycle.

Finally, it is important to note that a minor fraction of HELB does remain in the nucleus as S phase progresses, when resection is fully active. This small nuclear pool of HELB is likely

sufficient (potentially in concert with other, yet to be identified resection-limiting factors) to prevent excessive “run away” DNA end resection from occurring during S/G2 phases.

Figure 4.3 The combined loss of HELB and 53BP1 results in an additive increase in DNA end resection (A) WT, Helb-/-, 53bp1-/-, and 53bp1-/- Helb-/- were processed for the native BrdU resection assay as in previous figures. Shown are representative histogram plots of 10,000 cells analyzed for each condition. The values within each plot represent the mean fold increase in BrdU fluorescence intensity after NCS treatment (+/-SEM, n = 3). (B) Whole cell extracts from MEFs of the indicated genotype were analyzed by immunoblotting with antibodies against 53BP1, HELB and tubulin (used as a loading control). (C) MEFs of the indicated genotype were stained with propidium iodide (PI) and analyzed by flow cytometry to determine their cell cycle distribution. At least 10000 cells were analyzed for each sample.

Chapter 5: Discussion and future directions

5.1 Conclusions

I conclude that HELB mediates a negative feedback loop initiated by RPA-coated ssDNA, which antagonizes the activity of the EXO1 and BLM-DNA2 nucleases. Interestingly, this feedback mechanism is itself finely tuned by CDK activity through the modulation of HELB concentration within the nucleus, such that resection is suppressed in G1 and activated as cells enter S phase. As this pathway operates independently of the known end-protection systems that target the initiation of resection, vertebrates may have evolved at least two distinct regulatory systems that limit ssDNA overhang formation at DSBs: the first system is embodied by the Ku heterodimer and the 53BP1 pathway, which are critical for DSB repair pathway choice, and the second, reported here and mediated by HELB, which limits DNA end resection in an RPA- dependent manner.

5.2 The molecular mechanism of HELB

Our results suggest that HELB is an ATP-driven motor that translocates on ssDNA in the

5’ to 3’ direction to inhibit the action of the BLM-DNA2 and EXO1 nucleases. However, it remains unclear whether HELB directly displaces the resection machinery from ssDNA, in a manner reminiscent of other SF1Bα helicases/translocases. Some translocases are known to act as molecular “cowcatchers” (also referred to as “sweepases”) that displace DNA-bound proteins using their ATP-dependent motor activity. The prototypical cowcatcher helicase/translocase is

DNA-dependent ATPase (Dda) from bacteriophage T4 (Byrd and Raney, 2004). Dda has multiple activities on T4 DNA, being involved in homologous recombination (Formosa et al.,

1983; Kodadek and Alberts, 1987) and initiation of origin-dependent replication (Barry and

Alberts, 1994; Gauss et al., 1994). Dda can displace DNA-bound proteins in the path of the replication complex (Bedinger et al., 1983). Intriguingly, Dda binds tightly to Gp32, the T4 ssDNA-binding protein (analogous to eukaryotic RPA) (Byrd and Raney, 2004), an interaction which modifies the activity of Dda on various recombination and replication intermediate structures (Jordan and Morrical, 2015).

HELB displays significant cowcatcher activity in vitro, as detected by its ability to displace streptavidin from a 3’-biotinylated ssDNA oligonucleotide (Figure 5.1A), an experiment that was carried out by our collaborators Hemanta Adhikary, Jana Krietsch, and Jean-Yves

Masson (Laval University). The exact mechanism by which HELB acts remains to be determined and will likely require single-molecule studies that track resection enzymes and HELB in parallel. Elegant single-molecule fluorescence resonance energy transfer (FRET) studies with budding yeast Pif1, another SF1Bα helicase/translocase, have demonstrated that Pif1 is anchored at 3’-tailed ssDNA-dsDNA junctions and employs the energy of ATP to periodically “reel in” the ssDNA tail in order to displace the enzyme telomerase (Zhou et al., 2014). Coupled with the observation that Pif1 preferentially removes telomerase from longer 3’ ssDNA telomeric overhangs (Li et al., 2014), Pif1 represents an attractive model for HELB action. An analogous patrolling activity would endow HELB with the ability to remove nuclease complexes selectively at the site of resection, at the ssDNA-dsDNA junction.

However, my early attempts to test whether HELB overexpression can suppress EXO1 or

BLM protein accumulation at DSBs have so far failed to gain support for this model. Analysis of

GFP-EXO1 or endogenous BLM recruitment to laser microirradiation sites did not reveal significant differences between samples where the HELB protein and the empty vector were expressed (Figure 5.1BC). Therefore, the cowcatcher (Dda-like) and patrolling (Pif1-like) mechanistic models should be considered speculative and certainly do not exclude alternative

models. For example, HELB could directly modulate the RPA-ssDNA complex in a manner that inhibits resection and that confounded our attempts to monitor end resection through assaying

RPA phosphorylation in MEFs (Figure 3.3G). To begin answering these questions, we have initiated a collaboration with Taekjip Ha (Howard Hughes Medical Institute), a leading expert in single-molecule analysis. We hope to elucidate the mechanism of HELB using FRET coupled with total internal reflection fluorescence microscopy.

Figure 5.1 Exploring the mechanism of HELB (A) HELB displays cowcatcher activity. 5’-[32P]-labeled, 3’-biotinylated ssDNA substrate was incubated with or without streptavidin. The substrates were then incubated with purified recombinant HELB at increasing concentrations and the reaction products were separated by PAGE and visualized using autoradiography. (B) U2OS cells expressing GFP-NLS or GFP-HELB (WT or 3xA mutant) were laser microirradiated and prepared for BLM and γH2AX immunofluorescence. DAPI staining was used to delineate the outline of the nuclei. Scale bar = 10 µm. (C) Quantitation of fluorescence intensities of BLM stripes in U2OS cells expressing the indicated construct. Each dot represents a nucleus analyzed. Stripe intensity was normalized to the mean nuclear background intensity in each cell (represented by the dashed gray line). The distributions were compared using the Mann-Whitney U test.

5.3 Cell cycle regulation of HELB

Maintenance of a high HELB concentration in the nucleus during S phase, either by impairing its nuclear export or through overexpression, results in potent inhibition of resection that is even sufficient to overcome the hyper-resection phenotype of 53bp1-/- cells. As cells enter

S phase, the decrease in nuclear HELB may be an important feature in the upregulation of end resection. While nuclear export is clearly an important mechanism of regulation, we observed that even in the NESm mutant, there was still some cell cycle regulation of HELB (Figure 4.2C).

The C-terminal region of the human HELB protein contains seven potential CDK target sites

(S/TP), whereas the mouse HELB protein contains five. Two of these sites appear to be precisely positionally conserved between the two species based on sequence alignments using the MAFFT algorithm. Future work will focus on elucidating the minimal CDK sites (and/or additional

HELB C-terminal elements) that are sufficient to completely inhibit nuclear-cytoplasmic shuttling when mutated.

Additionally, the remaining cell cycle regulation of the HELB-NESm mutant might be due to the modulation of nuclear import or might be an entirely different mode of regulation. For example, the interaction of HELB with the F-box protein SKP2, which forms an important cell

cycle-regulated E3 ubiquitin ligase along with SKP1, CUL1 and RBX1 (Yu et al., 1998), hints that HELB nuclear levels might additionally be modulated through ubiquitin-mediated proteolytic degradation.

5.4 Is HELB involved in DNA replication?

Fanning and colleagues have previously proposed that HELB participates in various aspects of DNA replication (Gerhardt et al., 2015; Guler et al., 2012; Taneja, 2002). However, both HELB-depleted human cells and Helb-/- MEFs cycle normally and progress through S phase indistinguishably from their WT counterparts (Figure 3.4F). To examine DNA replication directly, I teamed up with David Gallo and Grant Brown (Donnelly Centre) to carry out molecular combing of nascent DNA fibers (Yang et al., 2012). To monitor unchallenged fork progression, exponentially growing WT and Helb-/- MEFs were sequentially pulse-labeled with the thymidine analogs 5-chloro-2′-deoxyuridine (CldU) and 5-iodo-2'-deoxyuridine (IdU), then embedded in agarose plugs and processed for DNA combing, followed by CldU and IdU immunofluorescence (Figure 5.2A). To analyze replication fork dynamics after recovery from replication stress, the replication complex was stalled by incubation with hydroxyurea (HU) between the CldU and IdU treatments. HU interferes with the activity of ribonucleotide reductase

(RNR), which leads to the depletion of the cellular deoxyribonucleotide pool and subsequent stalling of polymerases (Sinha and Snustad, 1972).

We found that unchallenged DNA replication in Helb-/- cells is normal (Figure 5.2B).

However, we did detect a mild slowdown of the DNA replication fork rate during recovery from

HU treatment (Figure 5.2B). The nature of this slowdown is intriguing since fork asymmetry, an indicator of fork stalling, is not affected in Helb-/- cells (Figure 5.2C), and since HELB was not

detected at unperturbed, HU-stalled, or HU and ATR inhibitor-collapsed replication forks using immunoprecipitation of nascent DNA (Sirbu et al., 2011; 2013). This suggests that the proposed function for HELB in the recovery from DNA replication stress may be indirect. It is important to note that is likely that in addition to its role in resection, HELB will play a role in other processes by virtue of its strong interaction with RPA.

Figure 5.2 DNA replication dynamics in Helb-/- cells (A) Timelines of incubations for pulse labeling of nascent DNA. (B) DNA fibers from WT and Helb-/- MEFs treated as in (A) were combed onto glass slides and adjoining CldU and IdU tracks were imaged. The length of the IdU tracks was measured in pixels and converted into kb/min by comparison to control DNA fibers of known lengths. Over 200 replication forks (each represented by a grey dot) were analyzed per sample and the Mann Whitney U test was used to compare the distributions of track lengths. The red bar indicates the median value. *** Recovery from HU-induced fork stalling; Helb+/+ vs Helb-/-, p < 0.0001. (C) Fork asymmetry was determined for bidirectional replication forks (a CldU track flanked by IdU tracks on both sides) using the following equation: Asymmetry % = ((Long IdU / Short IdU) – 1) * 100. Over 90 bidirectional replication forks (each represented by a grey dot) were analyzed per sample and the Mann Whitney U test was used to compare the asymmetry distributions. The red bar indicates the median value.

5.5 Phenotypic consequences of unregulated resection

A remaining theoretical question about HELB is why would negative regulators of long- range resection evolve? In other words, what is the selective advantage of cells where resection does not go on unrestrained? Some pathological consequences of extensive resection have been observed (Symington, 2016). 3’ overhangs are subject to degradation over time and clustered mutations can be generated due to the inherent lability of ssDNA compared to dsDNA (Roberts et al., 2012; Zierhut and Diffley, 2008). The binding of RPA to ssDNA normally stabilizes the 3’ tails (Chen et al., 2013). However, the nuclear pool of RPA proteins can become limiting, especially during S phase (when resection is also activated) or in some cancer cells with intrinsically high levels of replication stress (Toledo et al., 2013). Symington and colleagues have demonstrated that RPA deletion leads to the folding of ssDNA into secondary structures and hairpins, which are attractive substrates for cleavage by endonucleases, including Mre11 and

Sae2 (the budding yeast homolog of CtIP) (Chen et al., 2013). In the absence of Mre11 or Sae2 activity, RPA dysfunction can lead to palindromic gene amplification and more complex chromosome rearrangements (Deng et al., 2015). The authors concluded that secondary structures within ssDNA are potent instigators of genome instability and that RPA and Mre11-

Sae2 evolved the ability to prevent their formation and propagation, respectively. Perhaps HELB evolved in vertebrates as a complementary mechanism to restrain excessive ssDNA formation, especially in situations where RPA is limiting.

It is also important to note that SSA and inappropriate HR (both of which require extensive DNA end resection) are mutagenic processes. SSA invariably results in deletions of intervening regions between the sequences of microhomology. Unscheduled mitotic HR between homologous chromosomes (instead of sister chromatids) can lead to loss of heterozygosity for

multiple alleles or entire segments of chromosomes. Similarly, inappropriate HR between repetitive regions can lead to chromosomal translocations and genomic rearrangements, which are a hallmark of cancer cells. It is possible that regulators of long-range resection, including

HELB, evolved under pressure to restrain the frequency of these outcomes.

These conjectures imply that HELB-deficient cells and organisms may not display survival phenotypes unless they are challenged with DNA damaging agents, which induce DSBs, cause replication stress, or both. Indeed, I did not observe any difference in the survival of unchallenged WT and Helb-/- mice (Figure 3.2). Future studies will determine whether Helb-/- animals and cells display hypersensitivity (or perhaps resistance) to various sources of DNA damage. Potential tissues of interest may be the testes and thymus, both of which express high amounts of HELB mRNA (Taneja, 2002). Finally, some genetic disorders display anticipation, where the phenotypic consequences of a mutation become progressively more apparent with each new generation (Armanios et al., 2005; Ridley et al., 1988). It is possible that phenotypes associated with the loss of HELB may present in future generations of Helb-/- mice as a result of the ongoing accumulation of mutations associated with unrestrained DNA resection.

5.6 Therapeutic potential of modulating resection

Independently from my work, Guotai Xu and Sven Rottenberg (Netherlands Cancer

Institute) carried out a screen to identify factors that promote sensitivity to poly (ADP-ribose) polymerase (PARP) inhibitors in BRCA1-deficient mammary tumors. The screen identified several factors, including 53BP1 and MAD2L2 (Jaspers et al., 2013; Xu et al., 2015), as well as

HELB. shRNA-mediated depletion of HELB lead to the development of resistance to PARP inhibitor treatment of Brca1-/- p53-/- cells and allografted tumors (Tkáč et al., 2016). After

initiating a collaboration with Xu and Rottenberg, we found that this phenotype was dependent on HELB’s catalytic, ssDNA- and RPA-binding activities, closely mirroring its requirements in suppressing DNA end resection (Figure 3.5). Furthermore, the resistance of HELB-depleted cells to PARP inhibitors was dependent on ATM activity and expression of HELB-NESm (which dominantly suppresses resection) hyper-sensitized Brca1-/- p53-/- cells to PARP inhibitor treatment (Tkáč et al., 2016).

In addition, Brca1-/- p53-/- tumor cell lines that carried either the control shRNA vector or a vector expressing an shRNA against HELB were injected into the fat pads of female mice. In this allograft experiment, the mice were either left untreated or were treated with the PARP inhibitor olaparib for 28 days, after which the animals were monitored for tumor growth and survival. In the absence of olaparib treatment, tumor growth and the median time of survival was similar for control and HELB-depleted cells (7.5 days versus 11.5 days, respectively). However, in mice grafted with HELB-depleted Brca1-/- p53-/- tumor cells, olaparib treatment resulted in only a minor reduction in tumor growth causing a striking decrease in the median survival time

(28 days versus 49 days for control). These findings indicate that HELB mediates the cytotoxicity of PARP inhibitors in both cellular and allograft tumor models of BRCA1 deficiency (Tkáč et al., 2016). Importantly, the loss of HELB also partially restored the focal accumulation of RAD51 recombinase at sites of DSBs in both Brca1-null mouse tumor cells and

BRCA1-depleted human cells, confirming that HELB loss leads to a partial restoration of the HR pathway. Since these experiments were chiefly carried out by our collaborators, I did not include them in the main body of the thesis. However, the relevant figures can be found in Appendix 3

(Figures A1-A4).

Biallelic deletion of Brca1 exon 11 in mice (denoted Brca1Δ11/Δ11) results in embryonic death around day 13.5 of development (Xu et al., 2001). Deletion of 53BP1 rescues this lethality

through the restoration of end resection and HR (Bouwman et al., 2010; Bunting et al., 2010;

Cao et al., 2009). To test whether loss of HELB similarly affects the development of BRCA1-

deficient embryos, I generated mice double-heterozygous for Helb and Brca1 alleles (Helb+/-

Brca1+/Δ11). These mice were crossed and the resulting offspring were genotyped at birth. I found

that HELB deletion did not rescue the embryonic lethality of BRCA1 deficiency (Table 5.1),

indicating that the reactivation of HR in BRCA1-deficient cells that have lost HELB activity is

not as complete as what is observed following 53BP1 deletion.

Table 5.1 Helb-/- mutation does not rescue the embryonic lethality of Brca1Δ11/Δ11 mice

Cross Helb+/-Brca1+/Δ11 x Helb+/-Brca1+/Δ11

Total offspring (live births) 103 Helb+/+ Helb+/+ Helb+/+ Helb+/- Helb+/- Helb+/- Helb-/- Helb-/- Helb-/- Offspring genotypes Brca1+/+ Brca1+/Δ11 Brca1Δ11/Δ11 Brca1+/+ Brca1+/Δ11 Brca1Δ11/Δ11 Brca1+/+ Brca1+/Δ11 Brca1Δ11/Δ11 Expected (Mendelian) 6.44 12.88 6.44 12.88 25.75 12.88 6.44 12.88 6.44 Expected (Brca1Δ11/Δ11 lethal) 8.58 17.17 0 17.17 34.33 0 8.58 17.17 0 Observed 11 22 0 21 30 0 7 12 0

These findings have important implications for our understanding of HR and the

development of PARP inhibitors. At first, since 53BP1 promotes both NHEJ and end protection,

it had been difficult to untangle which activity of 53BP1 is required to promote PARP inhibitor

sensitivity in BRCA1-deficient cells. Because HELB is not involved in DSB repair pathway

choice but rather limits ATM-dependent end resection, our finding that HELB depletion

promotes resistance of Brca1-/- p53-/- cells to PARP inhibition suggests that increasing the extent

of resection alone, without impacting NHEJ, is sufficient to activate BRCA1-independent HR.

However, as the PARP inhibitor resistance in HELB-depleted cells is not as pronounced as in

53bp1-/- or MAD2L2-depleted cells (Jaspers et al., 2013; Xu et al., 2015), our results also

indicate that the proportion of ends engaging in long-range resection may be another important

factor in the ability of cells to survive PARP inhibitor treatment.

5.7 Model

I propose a simple model of HELB activity, which incorporates the totality of the experimental data presented in this thesis (Figure 5.3).

Figure 5.3 A model of HELB activity (A,B) HELB is recruited to partially resected DNA DSBs in an RPA-dependent manner, where it limits long-range resection by the BLM-DNA2 and EXO1 nuclease machineries. Upon loss of HELB, this constraint on resection is absent leading to a partial restoration of HR, even in the absence of BRCA1. (C,D) Resection is under control of the cell cycle: it is suppressed in G1 phase and active in S and G2 phases. The CDK2-mediated nuclear export of HELB participates in the activation of resection in S phase of the cell cycle.

Chapter 6: Methods and materials

6.1 Cell culture and treatments

HEK293 and HeLa cells were grown in DMEM medium supplemented with 10% fetal bovine serum (FBS) and 2 mM L-alanyl-L-glutamine. U2OS cells were grown in McCoy’s medium supplemented with 10% FBS. The inducible HELB expression cell lines HEK293 Flag-

HELB (used for immunoprecipitation-mass spectrometry) and U2OS GFP-HELB (used for laser microirradiation) were generated using the Flp-In/T-REx system (Invitrogen). Mouse embryonic fibroblasts (MEFs) were derived from embryos at 13.5 days post-coitum. MEFs were grown in

DMEM medium supplemented with 10% FBS, 2 mM L-alanyl-L-glutamine, 1% MEM non- essential amino acids, 1 mM sodium pyruvate, 100 U/ml penicillin, and 100 µg/ml streptomycin.

For expression of pMX-HELB-IRES-GFP constructs and shRNA knockdown experiments,

MEFs were first immortalized using the 3T3 protocol. Murine primary B lymphocytes were grown in RPMI medium supplemented with 10% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin. The G3 (KB1P-G3) and B11 (KB1P-B11) cell lines used in Figures 7, S6 and S7 were derived from a Brca1-/- p53-/- mouse mammary tumor, as described (Jaspers et al., 2013).

These cell lines were cultured in DMEM/F-12 medium supplemented with 10% FCS, 50 units/mL penicillin, 50 ng/ml streptomycin, 5µg/mL insulin (Sigma), 5ng/mL epidermal growth factor and 5 ng/mL cholera toxin (Gentaur) under low oxygen conditions (3% O2, 5% CO2,

37°C). 293T cells were cultured in DMEM supplemented with 10% FCS, 50 units/mL penicillin,

50 ng/ml streptomycin under normal oxygen conditions (21% O2, 5% CO2, 37°C). The PARP inhibitors olaparib and AZD2461 were synthesized by Syncom (Groningen, The Netherlands).

6.2 Affinity purification and mass spectrometry

For each FLAG-tagged bait and FLAG control, cell pellets from two 150 mm plates were lysed in 50 mM HEPES-KOH (pH 8.0), 100 mM KCl, 2 mM EDTA, 0.1% NP-40, 10% glycerol and affinity-purified with Flag-M2 magnetic beads (Sigma), followed by on-bead trypsin digest as described (Kean et al., 2012). A spray tip was formed on a fused silica capillary column (0.75

µm ID, 350 µm OD) using a laser puller (program = 4; heat = 280, FIL = 0, VEL = 18, DEL =

200). 10 cm (+/- 1 cm) of C18 reversed-phase material (Reprosil-Pur 120 C18-AQ, 3 µm) was packed in the column by pressure bomb (in MeOH). The column was then pre-equilibrated in buffer A (see below) (6 µl) before being connected in-line to a NanoLC-Ultra 2D plus HPLC system (Eksigent) coupled to a LTQ-Orbitrap Velos (Thermo Electron) equipped with a nanoelectrospray ion source (Proxeon Biosystems). The LTQ-Orbitrap Velos instrument under

Xcalibur 2.0 was operated in the data dependent mode to automatically switch between MS and up to 10 subsequent MS/MS acquisitions. The HPLC gradient program delivered an acetonitrile gradient over 125 minutes. Buffer A was 99.9% H2O, 0.1% formic acid; buffer B was 99.9%

ACN, 0.1% formic acid. For the first twenty minutes, the flow rate was 400µL/min at 2% B. The flow rate was then reduced to 200 µL/min and the fraction of solvent B increased in a linear fashion to 35% until 95.5 minutes. Solvent B was then increased to 80% over 5 minutes and maintained at that level until 107 minutes. The mobile phase was then reduced to 2% B until the end of the run (125min). The parameters for data dependent acquisition on the mass spectrometer were: 1 centroid MS (mass range 400-2000) followed by MS/MS on the 10 most abundant ions.

General parameters were: activation type = CID, isolation width = 1 m/z, normalized collision energy = 35, activation Q = 0.25, activation time = 10 msec. For data dependent acquisition, the

minimum threshold was 500, repeat count = 1, repeat duration = 30 sec, exclusion size list = 500, exclusion duration = 30 sec, exclusion mass width (by mass) = low 0.03, high 0.03.

6.3 Mass spectrometry data extraction

Raw mass spectrometry files were converted to mzXML using ProteoWizard (3.0.4468)

(Kessner et al., 2008) and analyzed using the iProphet pipeline (Shteynberg et al., 2011) implemented within ProHits (Liu et al., 2010) as follows. The database consisted of the human and adenovirus complements of the RefSeq protein database (version 57) supplemented with

“common contaminants” from the Max Planck Institute (http://maxquant.org/downloads.htm) and the Global Proteome Machine (GPM; http://www.thegpm.org/crap/index.html). The search database consisted of forward and reversed sequences (labeled “DECOY”); in total, 72226 entries were searched. The search engines used were Mascot (2.3.02; Matrix Science) and Comet

(2012.01 rev.3), with trypsin specificity (2 missed cleavages were allowed) and deamidation

(NQ) and oxidation (M) as variable modifications (Eng et al., 2013). Charges +2, +3 and +4 were allowed, and the parent mass tolerance was set at 15 ppm while the fragment bin tolerance was set at 0.6 amu. The resulting Comet and Mascot search results were individually processed by PeptideProphet (Keller et al., 2002), and peptides were assembled into proteins using parsimony rules first described in ProteinProphet (Nesvizhskii et al., 2003) into a final iProphet protein output using the Trans-Proteomic Pipeline (TPP; Linux version, v0.0 Development trunk rev 0, Build 201303061711). TPP options were as follows: general options are -p0.05 -x20 -

PPM -d"DECOY", iProphet options are –ipPRIME and PeptideProphet options are –pP. All proteins with a minimal iProphet protein probability of 0.05 were parsed to the relational module of ProHits. Note that for analysis with SAINT (see below), only proteins with iProphet protein

probability ≥ 0.95 are considered, corresponding to an estimated protein level false-discovery rate (FDR) of ~0.5%.

6.4 Interaction scoring for Flag AP-MS

For each FLAG-tagged bait, biological triplicates were grown and harvested at different times to maximize the variability and increase the robustness in the detection of true interactors.

Negative control purifications (consisting of cells expressing the unfused FLAG tag) were processed in parallel. The interactions were analyzed with SAINTexpress (v3.3), a computationally efficient reimplementation of the Significance Analysis of INTeractome method described previously (Choi et al., 2011; Teo et al., 2014). SAINT probabilities are computed independently for each bait replicate and the average probability (AvgP) of the best two out of three biological replicates is reported as the final SAINT score. Preys with AvgP ≥ 0.85 were considered “true” interactors (estimated FDR ≤ 1%). Keratins (KRT1 and KRT10), the adenovirus E1B protein and trypsin were manually removed from the list of interactions. A link to the SAINTexpress data can be found here: (http://prohits- web.lunenfeld.ca/prohits_report/gingras_36_t29j6bbdr3dbv65vbnrdntj0i6_0f_1448914609.html)

The data is also available in a searchable format at prohits-web.lunenfeld.ca; Durocher lab.

Downloadable files and all raw mass spectrometry files are deposited in the MassIVE repository housed at the Center for Computational Mass Spectrometry at UCSD

(http://proteomics.ucsd.edu/ProteoSAFe/datasets.jsp). The dataset has been assigned the

MassIVE ID (MSV000079395) and is available for FTP download at: ftp://[email protected]. The ProteomeXchange accession is PXD003259.

Visualization of the results was performed using an updated version of the interaction proteomics

dotplot tools (Knight et al., 2015), where the color intensity of the circles maps to the averaged spectral counts detected for each prey in a given purification (capped to a maximum of 50); the color coding on the edge maps to the FDR for each individual bait-prey interaction (high confidence is ≤ 1% FDR; medium confidence is ≤ 5% FDR; low confidence is > 5% FDR). The size of the circles is proportional to the maximal intensity of each given prey across all bait purifications; the smaller circles are sized down linearly in relation to the maximal intensity.

6.5 RNA interference

For all siRNA-mediated knockdowns in human cells, siGENOME SMARTpools and deconvolved siRNA duplexes were obtained from Dharmacon. In all samples, 10 nM siRNA was transfected using Lipofectamine RNAiMAX (Invitrogen). For shRNA-mediated knockdown of

CtIP described in Figure 3, a pFLRu-based shCtIP lentiviral construct (gift from A.

Nussenzweig, National Institutes of Health) was packaged in 293T cells and MEFs were infected using polybrene (8 µg/ml). For shRNA-mediated knockdown of EXO1 and BLM described in

Figures 3 and S3, pSUPERIOR-based retroviral constructs (gifts from T. de Lange, Rockefeller

University) were packaged in Phoenix-Eco cells and MEFs were infected 4 times during a 48 hr period using polybrene (4 µg/ml).

For the knockdown experiments performed in Amsterdam (see Appendix 3), shRNA hairpins were obtained from the Sigma Mission library (TRC Mm 1.0; shHelb-1:

TRCN0000182657_GCCCAGTATTGAACCTGGTAA and shHelb-2:

TRCN0000176637_CCTGATCCTAAACATAACTTT). 293T cells were transfected by the calcium phosphate method using 8µg of plasmid DNA of interest, 3µg of VSVg envelope vector

(pMD.G), 3µg of RSV-Rev, and 3µg of packaging vector pCMVDR8.2 per 10 cm dish. The

medium was refreshed 6 and 24 h after transfection, and the lentivirus-containing supernatant was collected after a further 24 h. Brca1-/- p53-/- cells were infected using polybrene (6 µg/ml).

The medium was refreshed after 12 h and transduced cells were selected using puromycin.

6.6 Plasmids

Human RPA1, RPA2, RPA3 and HELB cDNAs were tagged at the N-terminus with the

Flag or GFP epitopes and cloned into the pcDNA5/FRT/TO vector (Invitrogen) for the generation of stable HEK293- or U2OS-derived cell lines using the Flp-In/T-REx system as described previously (O’Donnell et al., 2010). Mouse Helb cDNA was cloned into the pMX-

IRES-GFP retroviral vector (gift from A. Nussenzweig, National Institutes of Health) for expression in MEFs (Figures 3, 6 and S3) and into the pMSCV-GFP retroviral vector (gift from

J. Jacobs, Netherlands Cancer Institute) for expression in Brca1-/- p53-/- mouse mammary-tumor derived cell lines (Figures 7 and S6).

6.7 Antibodies

Polyclonal antibodies directed against human HELB (designated 159B and 160A) were generated by immunizing rabbits with a peptide containing amino acids 1-199 of the human

HELB protein. Polyclonal antibodies directed against mouse HELB (designated 181A) were generated by immunizing rabbits with a peptide containing amino acids 848-1074 of the mouse

HELB protein. The rabbit anti-cyclin A antibody was a gift from M. Pagano (New York

University) and the goat anti-GFP antibody was a gift from L. Pelletier (Lunenfeld-Tanenbaum

Research Institute). The following antibodies were obtained from commercial sources: mouse

anti-BrdU (RPN202, GE Healthcare), rabbit anti-pRPA2 S4/S8 (A300-245A, Bethyl), mouse anti-γ-H2AX (clone JBW301, Millipore), rabbit anti-γ-H2AX (#2577, Cell Signaling

Technologies), rabbit anti-53BP1 (A300-272A-3 and A300-273A-4, Bethyl), rabbit anti-RIF1

(A300-569A-2, Bethyl), rabbit anti-RAD51 serum (70-001, lot 1, BioAcademia), rabbit anti-CtIP

(ab70163, Abcam), mouse anti-Flag (clone M2, Sigma), biotin-conjugated rat anti-mouse IgG1

(#553441, BD Biosciences), APC-conjugated mouse anti-biotin (#130-090-856, Miltenyi

Biotech), rabbit anti-GAPDH (G9545, Sigma), mouse anti-actin (clone JLA20, Millipore), mouse anti-α-tubulin (clone DM1A, Millipore).

6.8 HR and SSA DNA repair assays

The direct repeat (DR)-GFP assay to measure the frequency of HR and the strand annealing (SA)-GFP assay to measure the frequency of SSA were performed as previously described (Gunn and Stark, 2012). Briefly, HeLa DR-GFP cells (gift from R. Greenberg,

University of Pennsylvania) or U2OS SA-GFP cells (gift from J. Stark, Beckman Research

Institute of City of Hope) were transfected with 10 nM siRNA (Dharmacon) using

Lipofectamine RNAiMAX (Invitrogen). 24 hr later, the cells were transfected with the I-SceI expression plasmid pCBASceI (Addgene #26477) using polyethylenimine. 48 hr post-plasmid transfection, the cells were collected by trypsinization and the percentage of GFP-expressing cells was immediately analyzed by flow cytometry using the FACSCalibur platform (BD

Biosciences).

6.9 Phospho-RPA2 S4/S8 focus formation assay using high- content microscopy

SMARTpool siRNAs (Dharmacon) for 20 RPA interacting proteins were reverse transfected into U2OS FUCCI (Sakaue-Sawano et al., 2008) cells in a clear-bottom 96-well plate

(Corning Life Sciences; No. 3603) using Lipofectamine RNAiMAX (Invitrogen). 48 h-post transfection, cells were incubated in 50 ng/mL of neocarzinostatin (NCS) for 2 hr. Following

NCS treatment, cells were fixed with 4% PFA for 10 min followed by two washes with PBS.

Cells were permeabilized in 0.3% Triton X-100 for 30 min and then blocked in in 1X PBS buffer buffer containing 10% goat serum, 0.5% NP-40 and 0.5% saponin for 30 min. Next, an antibody against pRPA2 S4/S8 (Bethyl) diluted in blocking buffer at 1:1,000 was added to the cells for 1 hr at room temperature. After three washes in 1X PBS, cells were incubated in secondary antibody conjugated to Alexa Fluor 647 (at 1:1,000 in blocking buffer containing 0.5 µg/mL of

DAPI) for 1 hr at room temperature. Lastly, cells were washed three times in 1X PBS and imaged at 60X magnification on the IN Cell Analyzer 6000 (GE Healthcare). Images were imported to the image analysis software package Columbus (PerkinElmer). The average number of pRPA2 S4/S8 foci per nucleus of cells in S/G2 (positive for Geminin-mAG) was determined by spot segmentation and intensity thresholds.

6.10 Laser microirradiation

Cells were sensitized to UV radiation by incubation with Hoechst 33342 for 30 min at 37

°C. The samples were irradiated using a 355 nm (UVA) laser source (Coherent) with 8 mW power, then incubated for 4 hr at 37 °C before fixation with 2 % paraformaldehyde (U2OS cells) or 95% ethanol / 5% acetic acid (MEFs). Fluorescence intensities were quantitated using ImageJ

software. To control for intercellular variability of GFP-HELB expression in U2OS cells, the mean laser microirradiation stripe intensity was normalized to the mean nuclear background intensity in each cell.

6.11 Single-strand annealing traffic light reporter (SSA-TLR)

MEFs were transduced with the reporter plasmid pCVL.SSA TLR (Addgene #45576)

(Kuhar et al., 2014) using lentiviral infection with the VSV-G viral envelope and the psPAX2 viral packaging vector. Transductants were selected using puromycin (2 ng/mL) for three days.

Subsequently, 2.0 x 106 MEF cells carrying the SSA-TLR reporter cassette were electroporated with 5 µg of an expression plasmid carrying BFP-tagged I-SceI nuclease

(pCVL.SFFV.d14GFP.EF1a.HA.NLS.SceOpt.T2A.TagBFP, Addgene # 32627, named “Donor-

Sce-BFP”; or pCVL.SFFV.Kozak.HA.NLS.SceOpt.IRES.BFP, Addgene # 45574, named Sce-

IRES-BFP), using the Amaxa Nucleofector and the MEF1 Nucleofector Kit (Lonza #1004). The cells were harvested 48 hr after electroporation and analyzed using the BD LSRFortessa flow cytometer and FlowJo software. At least 1.0 x 105 cells were analyzed for each biological replicate. BFP+ cells were gated in order to focus the analysis only on I-SceI-expressing cells

(405 nm laser for excitation, 450/50 filter for detection). To determine the frequency of SSA- mediated repair of the I-SceI-induced DNA double-strand breaks, the gated BFP+ cell population was analyzed for iRFP fluorescence (640 nm laser for excitation, 710/50 filter for detection).

6.12 Native BrdU resection assay

Analysis of end resection using BrdU detection under native conditions was carried out as previously described (Nishi et al., 2014) with some modifications. Briefly, MEFs were incubated with 30 µM BrdU for 24 h, followed by treatment with 200 ng/mL NCS for 3 hr. For experiments examining the ATM-dependence of end resection, the ATM inhibitor KU55933

(Selleck Chemicals #S1092) was added to the growth medium 1 hr prior to the addition of NCS and kept in the medium during the NCS incubation. Cells were collected by trypsinization, washed with PBS, and fixed with 75% ethanol in PBS for 16 hr at -20 °C. Cells were washed with 0.1% Tween 20 in PBS following fixation and each subsequent incubation. One half of each sample of cells was acid-denatured using 2 N HCl for 30 min at room temperature, in order to test the levels of total BrdU incorporation. Blocking was performed in 5% FBS in PBS for 30 min at room temperature. BrdU in resected ssDNA was detected under native conditions by incubation with mouse anti-BrdU antibody in 5% FBS in PBS for 2 hr at room temperature. The cells were then incubated with fluorophore-conjugated anti-mouse IgG antibody in 5% FBS in PBS for 1 hr at room temperature. Subsequently, BrdU fluorescence intensity was analyzed using the

BD FACSCalibur flow cytometer and FlowJo software. The amount of resected ssDNA was quantitated by normalizing the mean BrdU fluorescence intensity following NCS treatment to that of the untreated control cells for each sample.

6.13 Neutral comet assay

MEFs were treated with the indicated concentrations of NCS for 3 hr and processed according to the manufacturer’s recommendations (Trevigen). The neutral comet assay was

carried out as described previously (Orthwein et al., 2014). Images were captured using a Zeiss

LSM 780 laser-scanning confocal microscope and the tail moment was quantified using the

OpenComet plugin in ImageJ. The tail moment value consists of the product of tail length

(measured in pixels) and the percentage of total DNA in the tail.

6.14 Class switch recombination

Immunoglobulin class switching to IgG1 was assayed in murine primary B cells, as previously described (Escribano-Diaz et al., 2013). Briefly, mature B lymphocytes were isolated from the spleens of adult wild-type, 53bp1-/-, and Helb-/- mice by depletion of CD43+ cells, using

CD43 microbeads (Miltenyi Biotec) according to the manufacturer's instructions. Purified B cells were resuspended at a concentration of 1 x 106 cells/mL in the presence of 50 ng/mL IL-4

(Preprotech) and 25 µg/mL lipopolysaccharides (Sigma-Aldrich) to induce class switching to

IgG1. The cells were analyzed after 4 days by flow cytometry using a biotin-conjugated rat anti- mouse IgG1 antibody (#553441, BD Biosciences), followed by an APC-conjugated anti-biotin antibody (#130-090-856, Miltenyi Biotech) in conjunction with propidium iodide staining to exclude dead cells.

6.15 Extrachromosomal NHEJ assay

NHEJ efficiency was determined using a linear extrachromosomal NHEJ substrate as previously described (Waters et al., 2014) with some modifications. The extrachromosomal

NHEJ substrate was generated by ligating short linkers to the head and tail of a 545 bp linear double-stranded DNA fragment generated by PCR (5’-

GCTATGAATTCCCTCGTGACCACCCTGACC-3’ and 5’-

GCTATGAATTCCCTTGTACAGCTCGTCCATGC-3’). The linkers used have 6 nt 3´single- stranded overhangs and were made by annealing 5’-CTCACACCCATCTCA-3’ to 5’-

AATTTGAGATGGGTGTGAGATCTGC-3’ (“head” linker) and 5’-

AATTTATACAGCTAAGCGATGATGCAG-3’ to 5’-CATCGCTTAGCTGTATA-3’ (“tail” linker). Excess linker was removed by QIAquick (Qiagen) purification and substrate purity validated by polyacrylamide gel electrophoresis. The substrate was mixed at a 1:11 ratio with a plasmid carrier and used to transfect the indicated MEFs cell lines with Lipofectamine 3000

(Invitrogen), following the manufacturer’s instructions. Cells were harvested after 4 hr incubation at 37oC, washed and resuspended in Hank’s buffered saline solution supplemented with 5 mM MgCl2. Extracellular DNA was digested by incubation with 6.25 U Benzonase for 15 min at 37oC. Cells were pelleted and DNA purified with the Qiamp kit (Qiagen). Joining efficiency was determined by quantification of head-to-tail junctions by qPCR using primers that anneal within double-stranded flanks (5’-GAGCTGTACAAGGgaattTATACAG-3’ and 5’-

GTGGTCACGAGGGaattTGAGATG- 3’). To correct for differences in transfection efficiency, qPCR was performed using primers specific for the double-stranded DNA used to ligate the linkers (5’- CGACCACATGAAGCAGCACG-3’ and 5’-CGTCCTTGAAGAAGATGGTGC-

3’).

6.16 Protein purification

RPA was purified as previously described (Henricksen et al., 1994). For recombinant

HELB protein purification, SF9 insect cells (1 L at 106 cells/ml) were infected with GST-HELB-

His10 baculoviruses at a 1:100 ratio. At 60 hr post-infection, cells were harvested by

centrifugation and the pellet was frozen on dry ice. Cells were lysed in Buffer 1 (PBS containing

150 mM KCl, 1% Triton X-100, 0.5 mM DTT and protease inhibitors) and homogenized by 20 passes through a Dounce homogenizer (pestle A). The cell lysate was incubated with 1 mM

MgCl2 and 2.5 U/ml benzonase nuclease at 4°C for 1 hr followed by centrifugation at 35000 rpm for 30 min. The soluble fraction was incubated with 400 µl of glutathione-sepharose beads in 10 ml of Buffer 1 and incubated for 1 hr at 4°C with gentle rotation. The beads were washed twice with Buffer 1 containing 250 mM NaCl followed by incubation with Buffer 2 (Buffer 1 with 5 mM ATP, 15 mM MgCl2) for 1 h. The beads were then washed twice with Buffer 1 containing

350 mM NaCl and once with P5 buffer (50 mM NaHPO4 pH 7.0, 500 mM NaCl, 10% glycerol,

0.05% Triton X-100, 5 mM imidazole), followed by cleavage with PreScission protease (60

U/ml, GE Healthcare Life Sciences) overnight at 4°C in P5 buffer. The cleaved, His-tagged protein was subjected to Talon affinity column purification and eluted with 500 mM imidazole.

The purified protein was dialysed in storage buffer (20 mM Tris-HCl pH 7.5, 200 mM NaCl,

10% glycerol, 1 mM DTT) and stored in small aliquots at -80°C.

6.17 RPA-bound ssDNA pull-down

Streptavidin-bound magnetic beads containing 100 ng of biotinylated ssDNA were blocked in reaction buffer (25 mM MOPS pH 7, 60 mM KCl, 1% Tween 20, 2 mM DTT, 5 mM

MgCl2, 2 mM ATP) containing BSA (0.75 µg/ul). 20 nM RPA was first incubated with the ssDNA beads in a volume of 20 µl for 10 min at 37°C, followed by washing three times with reaction buffer. Pre-bound, saturated RPA-ssDNA beads were incubated with 20 nM HELB proteins for 15 min at 37°C followed by three washes with reaction buffer. Proteins were eluted in SDS-loading buffer and analyzed by immunoblotting.

6.18 Streptavidin displacement assay

5’-[32P]-labeled, 3’-biotinylated oligonucleotide JYM3319 (10 nM) was incubated for 3 min with 1 µM streptavidin on ice in buffer containing 25 mM MOPS pH 7.0, 60 mM KCl, 1%

Tween 20 and 2 mM DTT. The oligonucleotide consisted of only thymidines to eliminate variations due to sequence effect. HELB was added at concentrations of 3 nM, 6 nM and 9 nM and the reactions were pre-incubated at 37°C for 3 min. The reactions were initiated (at 37°C) by addition of 5mM MnCl2 and 2mM ATP. Free biotin (10 uM) was added as a trap to prevent the displaced streptavidin from rebinding to the DNA. Reactions were quenched after 30 minutes with 350 mM EDTA and treated with proteinase K. The samples were analyzed by 8% polyacrylamide gel electophoresis and visualized by autoradiography.

6.19 DNA combing analysis

To analyze replication fork progression, exponentially growing MEFs were first incubated with 25 µM CldU for 30 min, rinsed with PBS, incubated with 125 µM IdU for 30 min, rinsed with PBS, and then trypsinized immediately. An additional incubation with 2 mM hydroxyurea for 30 min (between the CldU and IdU) was included in the experiments to analyze fork dynamics after recovery from replication fork stalling. The trypsinized cells were resuspended in 0.5% low melting point agarose (BioShop #AGA101) at a density of 5.0 x 106 cells/ml and 100 µl gel plugs were cast. DNA extraction, immunofluorescence, image acquisition and data analysis were carried out as previously described (Yang et al., 2012) with the following modifications. Following Proteinase K digestion, plugs were washed 5 x 10 min in 10 ml TE50 buffer (10 mM Tris-HCl pH 7.0, 50 mM EDTA). One plug was transferred to a round bottom

polycarbonate tube with 100 µl of 6.7 µM YOYO-1 (Y3601, Invitrogen) in TE50 for 30 min at room temperature in the dark to stain genomic DNA. Plugs were washed 3 x 5 min in 10 ml TE and incubated for 5 min in 5 ml of 50 mM MES buffer at pH 5.7. The MES buffer was replaced with 2 ml of fresh buffer and heated to 72 ˚C for 15-20 min to melt the agarose plugs. Following staining and coverslip mounting, images were taken of over 200 replication forks and 90 bidirectional replication forks at 63X magnification using an AxioImager ZI epifluorescence microscope and AxioVision software (Zeiss). Replication fork velocities and bidirectional asymmetries from two independent experiments were pooled for the final distributions. The formula used to calculate bidirectional fork asymmetry was:

Asymmetry % = ((Long IdU / Short IdU) – 1) * 100.

6.20 Clonogenic assays

The colony formation assays with PARP inhibitors (olaparib or AZD2461) were performed as described previously (Xu et al., 2015). Briefly, on day 0, 1.5x104 (B11) or 1x104

(G3) cells were seeded per well of a 6-well plate in PARP inhibitor-containing (or drug-free control) media. The medium of the PARP inhibitor treatment groups was refreshed with the drug every 4 days. The untreated control groups were stopped on day 5, whereas the PARP inhibitor treatment groups were stopped after another 2-3 weeks and stained with 0.1% crystal violet.

For ATM inhibitor (KU55933 or KU60019) and PARP inhibitor (olaparib or AZD2461) combination studies, on day 0, 1.5 x104 (B11) cells were seeded per well into 6-well plates and then ATM inhibitor, PARP inhibitor, or both were added. The medium was refreshed every 3 days with the indicated treatments. For the combination therapy groups, ATM inhibitor was applied for 6 days. On day 6, the ATM inhibitor alone and untreated control groups were stopped

and the other treatment groups were stopped on day 12 and stained with 0.1% crystal violet.

Colony growth was quantitated by measuring the absorbance of crystal violet at 590 nm.

6.21 Mice, tumor transplantation and olaparib treatment

Research in Toronto involving animals was performed in accordance with protocols approved by the animal facility at the Toronto Centre for Phenogenomics, in compliance with

Canadian legislation. Helbtm1(KOMP)Vlcg heterozygous mouse embryonic stem (ES) cells (strain

C57BL/6N) were obtained from the Knockout Mouse Project Repository (University of

California, Davis). In these ES cells, the entire coding sequence of one of the Helb alleles is replaced with a DNA cassette containing the LacZ gene and the neomycin resistance marker

(Figure S3A). Chimeric animals were generated by aggregation of Helbtm1(KOMP)Vlcg ES cells with albino morula-stage embryos (strain CD-1), 2.5 days post-coitum (dpc). The chimeras were tested for germline transmission of the Helb deletion allele. The germline-transmitting chimeras were then backcrossed to wild-type C57BL/6N animals to produce heterozygous Helb+/- mice, which were further crossed to produce Helb+/+ and Helb-/- mice in an isogenic background.

All mouse experiments carried out by Guotai Xu and Sven Rottenberg in Amsterdam were approved by the Animal Experiments Review Board of the Netherlands Cancer Institute, complying with Dutch legislation. To generate mouse mammary tumors from cell lines, shHelb- transduced or empty vector-transduced G3 cells (2 x 106) were orthotopically transplanted into female wild type FVB/N-Ola129 mice as reported previously (Jaspers et al., 2013). When tumors reached a volume of 200 mm3, mice were either left untreated or were treated with one regimen of 50 mg olaparib per kg daily for 28 days, after which the mice were followed until the tumor relapsed to a volume of about 1,500 mm3.

REFERENCES

Abdel-Monem, M., and Hoffmann-Berling, H. (1976). Enzymic unwinding of DNA. 1. Purification and characterization of a DNA-dependent ATPase from Escherichia coli. Eur. J. Biochem. 65, 431–440.

Abdel-Monem, M., Dürwald, H., and Hoffmann-Berling, H. (1976). Enzymic unwinding of DNA. 2. Chain separation by an ATP-dependent DNA unwinding enzyme. Eur. J. Biochem. 65, 441–449.

Anantha, R.W., Vassin, V.M., and Borowiec, J.A. (2007). Sequential and synergistic modification of human RPA stimulates chromosomal DNA repair. J. Biol. Chem. 282, 35910– 35923.

Aparicio, T., Baer, R., and Gautier, J. (2014). DNA double-strand break repair pathway choice and cancer. DNA Repair 19, 169–175.

Armanios, M., Chen, J.-L., Chang, Y.-P.C., Brodsky, R.A., Hawkins, A., Griffin, C.A., Eshleman, J.R., Cohen, A.R., Chakravarti, A., Hamosh, A., et al. (2005). Haploinsufficiency of telomerase reverse transcriptase leads to anticipation in autosomal dominant dyskeratosis congenita. Proc. Natl. Acad. Sci. U.S.a. 102, 15960–15964.

Aylon, Y., Liefshitz, B., and Kupiec, M. (2004). The CDK regulates repair of double-strand breaks by homologous recombination during the cell cycle. The EMBO Journal 23, 4868–4875.

Azvolinsky, A., Dunaway, S., Torres, J.Z., Bessler, J.B., and Zakian, V.A. (2006). The S. cerevisiae Rrm3p DNA helicase moves with the replication fork and affects replication of all yeast chromosomes. Genes & Development 20, 3104–3116.

Azvolinsky, A., Giresi, P.G., Lieb, J.D., and Zakian, V.A. (2009). Highly transcribed RNA polymerase II genes are impediments to replication fork progression in Saccharomyces cerevisiae. Molecular Cell 34, 722–734.

Bansbach, C.E., Bétous, R., Lovejoy, C.A., Glick, G.G., and Cortez, D. (2009). The annealing helicase SMARCAL1 maintains genome integrity at stalled replication forks. Genes & Development 23, 2405–2414.

Barry, J., and Alberts, B. (1994). A role for two DNA helicases in the replication of T4 bacteriophage DNA. J. Biol. Chem. 269, 33063–33068.

Bashir, T., Dorrello, N.V., Amador, V., Guardavaccaro, D., and Pagano, M. (2004). Control of the SCF(Skp2-Cks1) ubiquitin ligase by the APC/C(Cdh1) ubiquitin ligase. Nature 428, 190– 193.

Bedinger, P., Hochstrasser, M., Jongeneel, C.V., and Alberts, B.M. (1983). Properties of the T4 bacteriophage DNA replication apparatus: the T4 dda DNA helicase is required to pass a bound RNA polymerase molecule. Cell 34, 115–123.

Bekker-Jensen, S., Lukas, C., Kitagawa, R., Melander, F., Kastan, M.B., Bartek, J., and Lukas, J. (2006). Spatial organization of the mammalian genome surveillance machinery in response to DNA strand breaks. The Journal of Cell Biology 173, 195–206.

Bennett, C.B., Lewis, A.L., Baldwin, K.K., and Resnick, M.A. (1993). Lethality induced by a single site-specific double-strand break in a dispensable yeast plasmid. Proc. Natl. Acad. Sci. U.S.a. 90, 5613–5617.

Blier, P.R., Griffith, A.J., Craft, J., and Hardin, J.A. (1993). Binding of Ku protein to DNA. Measurement of affinity for ends and demonstration of binding to nicks. J. Biol. Chem. 268, 7594–7601.

Boboila, C., Alt, F.W., and Schwer, B. (2012). Classical and alternative end-joining pathways for repair of lymphocyte-specific and general DNA double-strand breaks. Adv. Immunol. 116, 1–49.

Bochkarev, A., Pfuetzner, R.A., Edwards, A.M., and Frappier, L. (1997). Structure of the single- stranded-DNA-binding domain of replication protein A bound to DNA. Nature 385, 176–181.

Bochkareva, E., Belegu, V., Korolev, S., and Bochkarev, A. (2001). Structure of the major single-stranded DNA-binding domain of replication protein A suggests a dynamic mechanism for DNA binding. The EMBO Journal 20, 612–618.

Bochkareva, E., Kaustov, L., Ayed, A., Yi, G.-S., Lu, Y., Pineda-Lucena, A., Liao, J.C.C., Okorokov, A.L., Milner, J., Arrowsmith, C.H., et al. (2005). Single-stranded DNA mimicry in the p53 transactivation domain interaction with replication protein A. Proc. Natl. Acad. Sci. U.S.a. 102, 15412–15417.

Bochkareva, E., Korolev, S., Lees-Miller, S.P., and Bochkarev, A. (2002). Structure of the RPA trimerization core and its role in the multistep DNA-binding mechanism of RPA. The EMBO Journal 21, 1855–1863.

Bochman, M.L., Sabouri, N., and Zakian, V.A. (2010). Unwinding the functions of the Pif1 family helicases. DNA Repair 9, 237–249.

Boersma, V., Moatti, N., Segura-Bayona, S., Peuscher, M.H., van der Torre, J., Wevers, B.A., Orthwein, A., Durocher, D., and Jacobs, J.J.L. (2015). MAD2L2 controls DNA repair at telomeres and DNA breaks by inhibiting 5' end resection. Nature 521, 537–540.

Bothmer, A., Robbiani, D.F., Feldhahn, N., Gazumyan, A., Nussenzweig, A., and Nussenzweig, M.C. (2010). 53BP1 regulates DNA resection and the choice between classical and alternative end joining during class switch recombination. Journal of Experimental Medicine 207, 855–865.

Bouwman, P., Aly, A., Escandell, J.M., Pieterse, M., Bartkova, J., van der Gulden, H., Hiddingh, S., Thanasoula, M., Kulkarni, A., Yang, Q., et al. (2010). 53BP1 loss rescues BRCA1 deficiency and is associated with triple-negative and BRCA-mutated breast cancers. Nat. Struct. Mol. Biol. 17, 688–695.

Boyce, R.P., and Howard-Flanders, P. (1964). Release of ultraviolet light-induced thymine dimers from DNA in E. coli K-12. Proc. Natl. Acad. Sci. U.S.a. 51, 293–300.

Brill, S.J., and Stillman, B. (1991). Replication factor-A from Saccharomyces cerevisiae is encoded by three essential genes coordinately expressed at S phase. Genes & Development 5, 1589–1600.

Brosh, R.M., Li, J.L., Kenny, M.K., Karow, J.K., Cooper, M.P., Kureekattil, R.P., Hickson, I.D., and Bohr, V.A. (2000). Replication protein A physically interacts with the Bloom's syndrome protein and stimulates its helicase activity. J. Biol. Chem. 275, 23500–23508.

Budd, M.E., Reis, C.C., Smith, S., Myung, K., and Campbell, J.L. (2006). Evidence suggesting that Pif1 helicase functions in DNA replication with the Dna2 helicase/nuclease and DNA polymerase delta. Molecular and Cellular Biology 26, 2490–2500.

Bunting, S.F., Callén, E., Wong, N., Chen, H.-T., Polato, F., Gunn, A., Bothmer, A., Feldhahn, N., Fernandez-Capetillo, O., Cao, L., et al. (2010). 53BP1 Inhibits Homologous Recombination in Brca1-Deficient Cells by Blocking Resection of DNA Breaks. Cell 141, 243–254.

Byrd, A.K., and Raney, K.D. (2004). Protein displacement by an assembly of helicase molecules aligned along single-stranded DNA. Nat. Struct. Mol. Biol. 11, 531–538.

Callén, E., Di Virgilio, M., Kruhlak, M.J., Nieto-Soler, M., Wong, N., Chen, H.-T., Faryabi, R.B., Polato, F., Santos, M., Starnes, L.M., et al. (2013). 53BP1 mediates productive and mutagenic DNA repair through distinct phosphoprotein interactions. Cell 153, 1266–1280.

Cao, L., Xu, X., Bunting, S.F., Liu, J., Wang, R.-H., Cao, L.L., Wu, J.J., Peng, T.-N., Chen, J., Nussenzweig, A., et al. (2009). A Selective Requirement for 53BP1 in the Biological Response to Genomic Instability Induced by Brca1 Deficiency. Molecular Cell 35, 534–541.

Chapman, J.R., Barral, P., Vannier, J.-B., Borel, V., Steger, M., Tomas-Loba, A., Sartori, A.A., Adams, I.R., Batista, F.D., and Boulton, S.J. (2013). RIF1 is essential for 53BP1-dependent nonhomologous end joining and suppression of DNA double-strand break resection. Molecular Cell 49, 858–871.

Chapman, J.R., Taylor, M.R.G., and Boulton, S.J. (2012). Playing the End Game: DNA Double- Strand Break Repair Pathway Choice. Molecular Cell 47, 497–510.

Chen, H., Lisby, M., and Symington, L.S. (2013). RPA Coordinates DNA End Resection and Prevents Formation of DNA Hairpins. Molecular Cell 50, 589–600.

Chen, R., and Wold, M.S. (2014). Replication protein A: single-stranded DNA's first responder: dynamic DNA-interactions allow replication protein A to direct single-strand DNA intermediates into different pathways for synthesis or repair. Bioessays 36, 1156–1161.

Chen, X., Niu, H., Chung, W.-H., Zhu, Z., Papusha, A., Shim, E.Y., Lee, S.E., Sung, P., and Ira, G. (2011). Cell cycle regulation of DNA double-strand break end resection by Cdk1-dependent Dna2 phosphorylation. Nat. Struct. Mol. Biol. 18, 1015–1019.

Choi, H., Larsen, B., Lin, Z.-Y., Breitkreutz, A., Mellacheruvu, D., Fermin, D., Qin, Z.S., Tyers, M., Gingras, A.-C., and Nesvizhskii, A.I. (2011). SAINT: probabilistic scoring of affinity purification-mass spectrometry data. Nat. Methods 8, 70–73.

Ciccia, A., Bredemeyer, A.L., Sowa, M.E., Terret, M.-E., Jallepalli, P.V., Harper, J.W., and Elledge, S.J. (2009). The SIOD disorder protein SMARCAL1 is an RPA-interacting protein involved in replication fork restart. Genes & Development 23, 2415–2425.

Cortez, D. (2015). Preventing replication fork collapse to maintain genome integrity. DNA Repair 32, 149–157. de Laat, W.L., Appeldoorn, E., Sugasawa, K., Weterings, E., Jaspers, N.G., and Hoeijmakers, J.H. (1998). DNA-binding polarity of human replication protein A positions nucleases in nucleotide excision repair. Genes & Development 12, 2598–2609.

Deng, S.K., Yin, Y., Petes, T.D., and Symington, L.S. (2015). Mre11-Sae2 and RPA Collaborate to Prevent Palindromic Gene Amplification. Molecular Cell 60, 500–508.

Di Virgilio, M., Callén, E., Yamane, A., Zhang, W., Jankovic, M., Gitlin, A.D., Feldhahn, N., Resch, W., Oliveira, T.Y., Chait, B.T., et al. (2013). Rif1 prevents resection of DNA breaks and promotes immunoglobulin class switching. Science 339, 711–715.

Digweed, M., and Sperling, K. (2004). Nijmegen breakage syndrome: clinical manifestation of defective response to DNA double-strand breaks. DNA Repair 3, 1207–1217.

Din, S., Brill, S.J., Fairman, M.P., and Stillman, B. (1990). Cell-cycle-regulated phosphorylation of DNA replication factor A from human and yeast cells. Genes & Development 4, 968–977.

Dornreiter, I., Erdile, L.F., Gilbert, I.U., Winkler, von, D., Kelly, T.J., and Fanning, E. (1992). Interaction of DNA polymerase alpha-primase with cellular replication protein A and SV40 T antigen. The EMBO Journal 11, 769–776.

Drapkin, R., Reardon, J.T., Ansari, A., Huang, J.C., Zawel, L., Ahn, K., Sancar, A., and Reinberg, D. (1994). Dual role of TFIIH in DNA excision repair and in transcription by RNA polymerase II. Nature 368, 769–772.

Eng, J.K., Jahan, T.A., and Hoopmann, M.R. (2013). Comet: an open-source MS/MS sequence database search tool. Proteomics 13, 22–24.

Escribano-Díaz, C., Orthwein, A., Fradet-Turcotte, A., Xing, M., Young, J.T.F., Tkáč, J., Cook, M.A., Rosebrock, A.P., Munro, M., Canny, M.D., et al. (2013). A Cell Cycle-Dependent Regulatory Circuit Composed of 53BP1-RIF1 and BRCA1-CtIP Controls DNA Repair Pathway Choice. Molecular Cell 1–12.

Fairman-Williams, M.E., Guenther, U.-P., and Jankowsky, E. (2010). SF1 and SF2 helicases: family matters. Current Opinion in Structural Biology 20, 313–324.

Falck, J., Forment, J.V., Coates, J., Mistrik, M., Lukas, J., Bartek, J., and Jackson, S.P. (2012). CDK targeting of NBS1 promotes DNA-end resection, replication restart and homologous recombination. EMBO Rep. 13, 561–568.

Fanning, E. (2006). A dynamic model for replication protein A (RPA) function in DNA processing pathways. Nucleic Acids Research 34, 4126–4137.

Feng, L., Fong, K.-W., Wang, J., Wang, W., and Chen, J. (2013). RIF1 counteracts BRCA1- mediated end resection during DNA repair. Journal of Biological Chemistry 288, 11135–11143.

Ferretti, L.P., Lafranchi, L., and Sartori, A.A. (2013). Controlling DNA-end resection: a new task for CDKs. Front Genet 4, 99.

Formosa, T., Burke, R.L., and Alberts, B.M. (1983). Affinity purification of bacteriophage T4 proteins essential for DNA replication and genetic recombination. Proc. Natl. Acad. Sci. U.S.a. 80, 2442–2446.

Foury, F., and Kolodynski, J. (1983). pif mutation blocks recombination between mitochondrial rho+ and rho- genomes having tandemly arrayed repeat units in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. U.S.a. 80, 5345–5349.

Fradet-Turcotte, A., Canny, M.D., Escribano-Díaz, C., Orthwein, A., Leung, C.C.Y., Huang, H., Landry, M.-C., Kitevski-LeBlanc, J., Noordermeer, S.M., Sicheri, F., et al. (2013). 53BP1 is a reader of the DNA-damage-induced H2A Lys 15 ubiquitin mark. Nature 499, 50–54.

Frankenberg-Schwager, M., Gebauer, A., Koppe, C., Wolf, H., Pralle, E., and Frankenberg, D. (2009). Single-strand annealing, conservative homologous recombination, nonhomologous DNA end joining, and the cell cycle-dependent repair of DNA double-strand breaks induced by sparsely or densely ionizing radiation. Radiat. Res. 171, 265–273.

Gauss, P., Park, K., Spencer, T.E., and Hacker, K.J. (1994). DNA helicase requirements for DNA replication during bacteriophage T4 infection. J. Bacteriol. 176, 1667–1672.

Gerhardt, J., Guler, G.D., and Fanning, E. (2015). Human DNA helicase B interacts with the replication initiation protein Cdc45 and facilitates Cdc45 binding onto chromatin. Exp. Cell Res. 334, 283–293.

Goldberg, I.H. (1987). Free radical mechanisms in neocarzinostatin-induced DNA damage. Free Radic. Biol. Med. 3, 41–54.

Gorbalenya, A.E., and Koonin, E.V. (1988). One more conserved sequence motif in helicases. Nucleic Acids Research 16, 7734.

Gorbalenya, A.E., Koonin, E.V., Donchenko, A.P., and Blinov, V.M. (1988a). A conserved NTP-motif in putative helicases. Nature 333, 22–22.

Gorbalenya, A.E., Koonin, E.V., Donchenko, A.P., and Blinov, V.M. (1988b). A novel superfamily of nucleoside triphosphate-binding motif containing proteins which are probably involved in duplex unwinding in DNA and RNA replication and recombination. FEBS Letters 235, 16–24.

Gorbalenya, A.E., Koonin, E.V., Donchenko, A.P., and Blinov, V.M. (1989). Two related superfamilies of putative helicases involved in replication, recombination, repair and expression of DNA and RNA genomes. Nucleic Acids Research 17, 4713–4730.

Gorbalenya, A.E., and Koonin, E.V. (1993). Helicases: amino acid sequence comparisons and structure-function relationships. Current Opinion in Structural Biology 3, 419–429.

Gravel, S., Chapman, J.R., Magill, C., and Jackson, S.P. (2008). DNA helicases Sgs1 and BLM promote DNA double-strand break resection. Genes & Development 22, 2767–2772.

Gu, J., Xia, X., Yan, P., Liu, H., Podust, V.N., Reynolds, A.B., and Fanning, E. (2004). Cell cycle-dependent regulation of a human DNA helicase that localizes in DNA damage foci. Molecular Biology of the Cell 15, 3320–3332.

Guler, G.D., Liu, H., Vaithiyalingam, S., Arnett, D.R., Kremmer, E., Chazin, W.J., and Fanning, E. (2012). Human DNA Helicase B (HDHB) Binds to Replication Protein A and Facilitates Cellular Recovery from Replication Stress. Journal of Biological Chemistry 287, 6469–6481.

Gunn, A., and Stark, J.M. (2012). I-SceI-based assays to examine distinct repair outcomes of mammalian chromosomal double strand breaks. Methods Mol. Biol. 920, 379–391.

Helmink, B.A., Tubbs, A.T., Dorsett, Y., Bednarski, J.J., Walker, L.M., Feng, Z., Sharma, G.G., McKinnon, P.J., Zhang, J., Bassing, C.H., et al. (2011). H2AX prevents CtIP-mediated DNA end resection and aberrant repair in G1-phase lymphocytes. Nature 469, 245–249.

Henricksen, L.A., Umbricht, C.B., and Wold, M.S. (1994). Recombinant replication protein A: expression, complex formation, and functional characterization. J. Biol. Chem. 269, 11121– 11132.

Heyer, W.-D., Ehmsen, K.T., and Liu, J. (2010). Regulation of homologous recombination in eukaryotes. Annu. Rev. Genet. 44, 113–139.

Huertas, P., and Jackson, S.P. (2009). Human CtIP mediates cell cycle control of DNA end resection and double strand break repair. J. Biol. Chem. 284, 9558–9565.

Iftode, C., and Borowiec, J.A. (2000). 5“ --> 3” molecular polarity of human replication protein A (hRPA) binding to pseudo-origin DNA substrates. Biochemistry 39, 11970–11981.

Ira, G., Pellicioli, A., Balijja, A., Wang, X., Fiorani, S., Carotenuto, W., Liberi, G., Bressan, D., Wan, L., Hollingsworth, N.M., et al. (2004). DNA end resection, homologous recombination and DNA damage checkpoint activation require CDK1. Nature 431, 1011–1017.

Ishida, N., Miyazaki, K., Kumagai, K., and Rikimaru, M. (1965). Neocarzinostatin, an antitumor antibiotic of high molecular weight. J. Antibiot. 18, 68–76.

Ivessa, A.S., Zhou, J.Q., and Zakian, V.A. (2000). The Saccharomyces Pif1p DNA helicase and the highly related Rrm3p have opposite effects on replication fork progression in ribosomal DNA. Cell 100, 479–489.

Ivessa, A.S., Lenzmeier, B.A., Bessler, J.B., Goudsouzian, L.K., Schnakenberg, S.L., and Zakian, V.A. (2003). The Saccharomyces cerevisiae helicase Rrm3p facilitates replication past nonhistone protein-DNA complexes. Molecular Cell 12, 1525–1536.

Ivessa, A.S., Zhou, J.-Q., Schulz, V.P., Monson, E.K., and Zakian, V.A. (2002). Saccharomyces Rrm3p, a 5“ to 3” DNA helicase that promotes replication fork progression through telomeric and subtelomeric DNA. Genes & Development 16, 1383–1396.

Jackson, S.P., and Bartek, J. (2009). The DNA-damage response in human biology and disease. Nature 461, 1071–1078.

Jasin, M., and Rothstein, R. (2013). Repair of strand breaks by homologous recombination. Cold Spring Harbor Perspectives in Biology 5, a012740–a012740.

Jaspers, J.E., Kersbergen, A., Boon, U., Sol, W., van Deemter, L., Zander, S.A., Drost, R., Wientjens, E., Ji, J., Aly, A., et al. (2013). Loss of 53BP1 Causes PARP Inhibitor Resistance in Brca1-Mutated Mouse Mammary Tumors. Cancer Discovery 3, 68–81.

Jordan, C.S., and Morrical, S.W. (2015). Regulation of the bacteriophage T4 Dda helicase by Gp32 single-stranded DNA-binding protein. DNA Repair 25, 41–53.

Kappen, L.S., and Goldberg, I.H. (1983). Deoxyribonucleic acid damage by neocarzinostatin chromophore: strand breaks generated by selective oxidation of C-5' of deoxyribose. Biochemistry 22, 4872–4878.

Kean, M.J., Couzens, A.L., and Gingras, A.-C. (2012). Mass spectrometry approaches to study mammalian kinase and phosphatase associated proteins. Methods 57, 400–408.

Keeney, S. (2008). Spo11 and the Formation of DNA Double-Strand Breaks in Meiosis. Genome Dyn Stab 2, 81–123.

Keil, R.L., and McWilliams, A.D. (1993). A gene with specific and global effects on recombination of sequences from tandemly repeated genes in Saccharomyces cerevisiae. Genetics 135, 711–718.

Keller, A., Nesvizhskii, A.I., Kolker, E., and Aebersold, R. (2002). Empirical statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search. Anal. Chem. 74, 5383–5392.

Kessner, D., Chambers, M., Burke, R., Agus, D., and Mallick, P. (2008). ProteoWizard: open source software for rapid proteomics tools development. Bioinformatics 24, 2534–2536.

Kim, C., Paulus, B.F., and Wold, M.S. (1994). Interactions of human replication protein A with oligonucleotides. Biochemistry 33, 14197–14206.

Kim, C., Snyder, R.O., and Wold, M.S. (1992). Binding properties of replication protein A from human and yeast cells. Molecular and Cellular Biology 12, 3050–3059.

Knight, J.D.R., Liu, G., Zhang, J.P., Pasculescu, A., Choi, H., and Gingras, A.-C. (2015). A web- tool for visualizing quantitative protein-protein interaction data. Proteomics 15, 1432–1436.

Kodadek, T., and Alberts, B.M. (1987). Stimulation of protein-directed strand exchange by a DNA helicase. Nature 326, 312–314.

Kolas, N.K., Chapman, J.R., Nakada, S., Ylanko, J., Chahwan, R., Sweeney, F.D., Panier, S., Mendez, M., Wildenhain, J., Thomson, T.M., et al. (2007). Orchestration of the DNA-damage response by the RNF8 ubiquitin ligase. Science 318, 1637–1640.

Kuhar, R., Gwiazda, K.S., Humbert, O., Mandt, T., Pangallo, J., Brault, M., Khan, I., Maizels, N., Rawlings, D.J., Scharenberg, A.M., et al. (2014). Novel fluorescent genome editing reporters for monitoring DNA repair pathway utilization at endonuclease-induced breaks. Nucleic Acids Research 42, e4–e4.

Lahaye, A., Leterme, S., and Foury, F. (1993). PIF1 DNA helicase from Saccharomyces cerevisiae. Biochemical characterization of the enzyme. J. Biol. Chem. 268, 26155–26161.

Lahaye, A., Stahl, H., Thines-Sempoux, D., and Foury, F. (1991). PIF1: a DNA helicase in yeast mitochondria. The EMBO Journal 10, 997–1007.

Li, J.-R., Yu, T.-Y., Chien, I.-C., Lu, C.-Y., Lin, J.-J., and Li, H.-W. (2014). Pif1 regulates telomere length by preferentially removing telomerase from long telomere ends. Nucleic Acids Research 42, 8527–8536.

Lieber, M.R. (2010). The mechanism of double-strand DNA break repair by the nonhomologous DNA end-joining pathway. Annu. Rev. Biochem. 79, 181–211.

Lindahl, T. (1993). Instability and decay of the primary structure of DNA. Nature 362, 709–715.

Lindahl, T., and Barnes, D.E. (2000). Repair of endogenous DNA damage. Cold Spring Harb. Symp. Quant. Biol. 65, 127–133.

Lindahl, T., and Nyberg, B. (1972). Rate of depurination of native deoxyribonucleic acid. Biochemistry 11, 3610–3618.

Litman, R., Peng, M., Jin, Z., Zhang, F., Zhang, J., Powell, S., Andreassen, P.R., and Cantor, S.B. (2005). BACH1 is critical for homologous recombination and appears to be the Fanconi anemia gene product FANCJ. Cancer Cell 8, 255–265.

Liu, G., Zhang, J., Larsen, B., Stark, C., Breitkreutz, A., Lin, Z.-Y., Breitkreutz, B.-J., Ding, Y., Colwill, K., Pasculescu, A., et al. (2010). ProHits: integrated software for mass spectrometry- based interaction proteomics. Nat Biotechnol 28, 1015–1017.

Liu, V.F., and Weaver, D.T. (1993). The ionizing radiation-induced replication protein A phosphorylation response differs between ataxia telangiectasia and normal human cells. Molecular and Cellular Biology 13, 7222–7231.

Lohman, T.M., Tomko, E.J., and Wu, C.G. (2008). Non-hexameric DNA helicases and translocases: mechanisms and regulation. Nat Rev Mol Cell Biol 9, 391–401.

Ma, Y., Pannicke, U., Schwarz, K., and Lieber, M.R. (2002). Hairpin opening and overhang processing by an Artemis/DNA-dependent protein kinase complex in nonhomologous end joining and V(D)J recombination. Cell 108, 781–794.

Manis, J.P., Morales, J.C., Xia, Z., Kutok, J.L., Alt, F.W., and Carpenter, P.B. (2004). 53BP1 links DNA damage-response pathways to immunoglobulin heavy chain class-switch recombination. Nat. Immunol. 5, 481–487.

Maréchal, A., and Zou, L. (2015). RPA-coated single-stranded DNA as a platform for post- translational modifications in the DNA damage response. Cell Research 25, 9–23.

Matsuoka, S., Ballif, B.A., Smogorzewska, A., McDonald, E.R., Hurov, K.E., Luo, J., Bakalarski, C.E., Zhao, Z., Solimini, N., Lerenthal, Y., et al. (2007). ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science 316, 1160– 1166.

Mer, G., Bochkarev, A., Gupta, R., Bochkareva, E., Frappier, L., Ingles, C.J., Edwards, A.M., and Chazin, W.J. (2000). Structural basis for the recognition of DNA repair proteins UNG2, XPA, and RAD52 by replication factor RPA. Cell 103, 449–456.

Mimitou, E.P., and Symington, L.S. (2008). Sae2, Exo1 and Sgs1 collaborate in DNA double- strand break processing. Nature 455, 770–774.

Mimitou, E.P., and Symington, L.S. (2010). Ku prevents Exo1 and Sgs1-dependent resection of DNA ends in the absence of a functional MRX complex or Sae2. The EMBO Journal 29, 3358– 3369.

Mimori, T., and Hardin, J.A. (1986). Mechanism of interaction between Ku protein and DNA. J. Biol. Chem. 261, 10375–10379.

Morita, S., Kojima, T., and Kitamura, T. (2000). Plat-E: an efficient and stable system for transient packaging of retroviruses. Gene Ther. 7, 1063–1066.

Murzin, A.G. (1993). OB(oligonucleotide/oligosaccharide binding)-fold: common structural and functional solution for non-homologous sequences. The EMBO Journal 12, 861–867.

Nagelhus, T.A., Haug, T., Singh, K.K., Keshav, K.F., Skorpen, F., Otterlei, M., Bharati, S., Lindmo, T., Benichou, S., Benarous, R., et al. (1997). A sequence in the N-terminal region of human uracil-DNA glycosylase with homology to XPA interacts with the C-terminal part of the 34-kDa subunit of replication protein A. J. Biol. Chem. 272, 6561–6566.

Nakamura, J., Walker, V.E., Upton, P.B., Chiang, S.Y., Kow, Y.W., and Swenberg, J.A. (1998). Highly sensitive apurinic/apyrimidinic site assay can detect spontaneous and chemically induced depurination under physiological conditions. Cancer Res. 58, 222–225.

Nesvizhskii, A.I., Keller, A., Kolker, E., and Aebersold, R. (2003). A statistical model for identifying proteins by tandem mass spectrometry. Anal. Chem. 75, 4646–4658.

Nimonkar, A.V., Genschel, J., Kinoshita, E., Polaczek, P., Campbell, J.L., Wyman, C., Modrich, P., and Kowalczykowski, S.C. (2011). BLM-DNA2-RPA-MRN and EXO1-BLM-RPA-MRN constitute two DNA end resection machineries for human DNA break repair. Genes & Development 25, 350–362.

Nishi, R., Wijnhoven, P., le Sage, C., Tjeertes, J., Galanty, Y., Forment, J.V., Clague, M.J., Urbé, S., and Jackson, S.P. (2014). Systematic characterization of deubiquitylating enzymes for roles in maintaining genome integrity. Nat. Cell Biol. 16, 1016–26–1–8.

O'Donnell, L., Panier, S., Wildenhain, J., Tkach, J.M., Al-Hakim, A., Landry, M.-C., Escribano- Díaz, C., Szilard, R.K., Young, J.T.F., Munro, M., et al. (2010). The MMS22L-TONSL Complex Mediates Recovery from Replication Stress and Homologous Recombination. Molecular Cell 40, 619–631.

O'Driscoll, M. (2012). Diseases associated with defective responses to DNA damage. Cold Spring Harbor Perspectives in Biology 4, a012773–a012773.

Orren, D.K., Selby, C.P., Hearst, J.E., and Sancar, A. (1992). Post-incision steps of nucleotide excision repair in Escherichia coli. Disassembly of the UvrBC-DNA complex by helicase II and DNA polymerase I. J. Biol. Chem. 267, 780–788.

Orthwein, A., Fradet-Turcotte, A., Noordermeer, S.M., Canny, M.D., Brun, C.M., Strecker, J., Escribano-Díaz, C., and Durocher, D. (2014). Mitosis inhibits DNA double-strand break repair to guard against telomere fusions. Science 344, 189–193.

Orthwein, A., Noordermeer, S.M., Wilson, M.D., Landry, S., Enchev, R.I., Sherker, A., Munro, M., Pinder, J., Salsman, J., Dellaire, G., et al. (2015). A mechanism for the suppression of homologous recombination in G1 cells. Nature 528, 422–426.

Panier, S., and Durocher, D. (2013). Push back to respond better: regulatory inhibition of the DNA double-strand break response. Nat Rev Mol Cell Biol 14, 661–672.

Pâques, F., and Haber, J.E. (1999). Multiple pathways of recombination induced by double- strand breaks in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 63, 349–404.

Pierce, A.J., Johnson, R.D., Thompson, L.H., and Jasin, M. (1999). XRCC3 promotes homology- directed repair of DNA damage in mammalian cells. Genes & Development 13, 2633–2638.

Pike, J.E., Henry, R.A., Burgers, P.M.J., Campbell, J.L., and Bambara, R.A. (2010). An Alternative Pathway for Okazaki Fragment Processing: RESOLUTION OF FOLD-BACK FLAPS BY Pif1 HELICASE. Journal of Biological Chemistry 285, 41712–41723.

Polato, F., Callén, E., Wong, N., Faryabi, R., Bunting, S., Chen, H.-T., Kozak, M., Kruhlak, M.J., Reczek, C.R., Lee, W.-H., et al. (2014). CtIP-mediated resection is essential for viability and can operate independently of BRCA1. Journal of Experimental Medicine 211, 1027–1036.

Poon, R., Beerman, T.A., and Goldberg, I.H. (1977). Characterization of DNA strand breakage in vitro by the antitumor protein neocarzinostatin. Biochemistry 16, 486–493.

Povirk, L.F. (1996). DNA damage and mutagenesis by radiomimetic DNA-cleaving agents: bleomycin, neocarzinostatin and other enediynes. Mutat. Res. 355, 71–89.

Povirk, L.F., and Goldberg, I.H. (1980). Binding of the nonprotein chromophore of neocarzinostatin to deoxyribonucleic acid. Biochemistry 19, 4773–4780.

Raghunathan, S., Ricard, C.S., Lohman, T.M., and Waksman, G. (1997). Crystal structure of the homo-tetrameric DNA binding domain of Escherichia coli single-stranded DNA-binding protein determined by multiwavelength x-ray diffraction on the selenomethionyl protein at 2.9-A resolution. Proc. Natl. Acad. Sci. U.S.a. 94, 6652–6657.

Ribeyre, C., Lopes, J., Boulé, J.-B., Piazza, A., Guédin, A., Zakian, V.A., Mergny, J.-L., and Nicolas, A. (2009). The yeast Pif1 helicase prevents genomic instability caused by G- quadruplex-forming CEB1 sequences in vivo. PLoS Genet 5, e1000475.

Richet, E., and Kohiyama, M. (1976). Purification and characterization of a DNA-dependent ATPase from Escherichia coli. J. Biol. Chem. 251, 808–812.

Ridley, R.M., Frith, C.D., Crow, T.J., and Conneally, P.M. (1988). Anticipation in Huntington's disease is inherited through the male line but may originate in the female. J. Med. Genet. 25, 589–595.

Roberts, S.A., Sterling, J., Thompson, C., Harris, S., Mav, D., Shah, R., Klimczak, L.J., Kryukov, G.V., Malc, E., Mieczkowski, P.A., et al. (2012). Clustered mutations in yeast and in human cancers can arise from damaged long single-strand DNA regions. Molecular Cell 46, 424–435.

Rossi, M.L., Pike, J.E., Wang, W., Burgers, P.M.J., Campbell, J.L., and Bambara, R.A. (2008). Pif1 helicase directs eukaryotic Okazaki fragments toward the two-nuclease cleavage pathway for primer removal. J. Biol. Chem. 283, 27483–27493.

Saikrishnan, K., Griffiths, S.P., Cook, N., Court, R., and Wigley, D.B. (2008). DNA binding to RecD: role of the 1B domain in SF1B helicase activity. The EMBO Journal 27, 2222–2229.

Saikrishnan, K., Powell, B., Cook, N.J., Webb, M.R., and Wigley, D.B. (2009). Mechanistic Basis of 5′-3′ Translocation in SF1B Helicases. Cell 137, 849–859.

Sakaue-Sawano, A., Kurokawa, H., Morimura, T., Hanyu, A., Hama, H., Osawa, H., Kashiwagi, S., Fukami, K., Miyata, T., Miyoshi, H., et al. (2008). Visualizing Spatiotemporal Dynamics of Multicellular Cell-Cycle Progression. Cell 132, 487–498.

Sancar, A., and Rupp, W.D. (1983). A novel repair enzyme: UvrABC excision nuclease of Escherichia coli cuts a DNA strand on both sides of the damaged region. Cell 33, 249–260.

Sancar, A., Williams, K.R., Chase, J.W., and Rupp, W.D. (1981). Sequences of the ssb gene and protein. Proc. Natl. Acad. Sci. U.S.a. 78, 4274–4278.

Santocanale, C., Neecke, H., Longhese, M.P., Lucchini, G., and Plevani, P. (1995). Mutations in the gene encoding the 34 kDa subunit of yeast replication protein A cause defective S phase progression. J. Mol. Biol. 254, 595–607.

Sartori, A.A., Lukas, C., Coates, J., Mistrik, M., Fu, S., Bartek, J., Baer, R., Lukas, J., and Jackson, S.P. (2007). Human CtIP promotes DNA end resection. Nature 450, 509–514.

100

Schulz, V.P., and Zakian, V.A. (1994). The saccharomyces PIF1 DNA helicase inhibits telomere elongation and de novo telomere formation. Cell 76, 145–155.

Seki, M., Enomoto, T., Hanaoka, F., and Yamada, M. (1987). DNA-dependent adenosinetriphosphatase B from mouse FM3A cells has DNA helicase activity. Biochemistry 26, 2924–2928.

Seki, M., Kohda, T., Yano, T., Tada, S., Yanagisawa, J., Eki, T., Ui, M., and Enomoto, T. (1995). Characterization of DNA synthesis and DNA-dependent ATPase activity at a restrictive temperature in temperature-sensitive tsFT848 cells with thermolabile DNA helicase B. Molecular and Cellular Biology 15, 165–172.

Setlow, R.B., and Carrier, W.L. (1964). The disappearance of thymine dimers from DNA: an error-correcting mechanism. Proc. Natl. Acad. Sci. U.S.a. 51, 226–231.

Shereda, R.D., Kozlov, A.G., Lohman, T.M., Cox, M.M., and Keck, J.L. (2008). SSB as an organizer/mobilizer of genome maintenance complexes. Crit. Rev. Biochem. Mol. Biol. 43, 289– 318.

Shibata, A., Moiani, D., Arvai, A.S., Perry, J., Harding, S.M., Genois, M.-M., Maity, R., van Rossum-Fikkert, S., Kertokalio, A., Romoli, F., et al. (2014). DNA double-strand break repair pathway choice is directed by distinct MRE11 nuclease activities. Molecular Cell 53, 7–18.

Shteynberg, D., Deutsch, E.W., Lam, H., Eng, J.K., Sun, Z., Tasman, N., Mendoza, L., Moritz, R.L., Aebersold, R., and Nesvizhskii, A.I. (2011). iProphet: multi-level integrative analysis of shotgun proteomic data improves peptide and protein identification rates and error estimates. Mol. Cell Proteomics 10, M111.007690–M111.007690.

Singleton, M.R., Dillingham, M.S., and Wigley, D.B. (2007). Structure and Mechanism of Helicases and Nucleic Acid Translocases. Annu. Rev. Biochem. 76, 23–50.

Sinha, N.K., and Snustad, D.P. (1972). Mechanism of inhibition of deoxyribonucleic acid synthesis in Escherichia coli by hydroxyurea. J. Bacteriol. 112, 1321–1324.

Sirbu, B.M., Couch, F.B., Feigerle, J.T., Bhaskara, S., Hiebert, S.W., and Cortez, D. (2011). Analysis of protein dynamics at active, stalled, and collapsed replication forks. Genes & Development 25, 1320–1327.

Sirbu, B.M., McDonald, W.H., Dungrawala, H., Badu-Nkansah, A., Kavanaugh, G.M., Chen, Y., Tabb, D.L., and Cortez, D. (2013). Identification of proteins at active, stalled, and collapsed replication forks using isolation of proteins on nascent DNA (iPOND) coupled with mass spectrometry. Journal of Biological Chemistry 288, 31458–31467.

Spencer, S.L., Cappell, S.D., Tsai, F.-C., Overton, K.W., Wang, C.L., and Meyer, T. (2013). The Proliferation-Quiescence Decision Is Controlled by a Bifurcation in CDK2 Activity at Mitotic Exit. Cell 155, 369–383.

101

Stark, J.M., Pierce, A.J., Oh, J., Pastink, A., and Jasin, M. (2004). Genetic steps of mammalian homologous repair with distinct mutagenic consequences. Molecular and Cellular Biology 24, 9305–9316.

Stephan, H., Concannon, C., Kremmer, E., Carty, M.P., and Nasheuer, H.-P. (2009). Ionizing radiation-dependent and independent phosphorylation of the 32-kDa subunit of replication protein A during mitosis. Nucleic Acids Research 37, 6028–6041.

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., et al. (2009). The RIDDLE syndrome protein mediates a ubiquitin-dependent signaling cascade at sites of DNA damage. Cell 136, 420–434.

Svendsen, J.M., and Harper, J.W. (2010). GEN1/Yen1 and the SLX4 complex: Solutions to the problem of Holliday junction resolution. Genes & Development 24, 521–536.

Symington, L.S. (2016). Mechanism and regulation of DNA end resection in eukaryotes. Crit. Rev. Biochem. Mol. Biol. 1–18.

Symington, L.S., and Gautier, J. (2011). Double-Strand Break End Resection and Repair Pathway Choice. Annu. Rev. Genet. 45, 247–271.

Tada, S., Kobayashi, T., Omori, A., Kusa, Y., Okumura, N., Kodaira, H., Ishimi, Y., Seki, M., and Enomoto, T. (2001). Molecular cloning of a cDNA encoding mouse DNA helicase B, which has homology to Escherichia coli RecD protein, and identification of a mutation in the DNA helicase B from tsFT848 temperature-sensitive DNA replication mutant cells. Nucleic Acids Research 29, 3835–3840.

Taneja, P. (2002). A Dominant-negative Mutant of Human DNA Helicase B Blocks the Onset of Chromosomal DNA Replication. Journal of Biological Chemistry 277, 40853–40861.

Tawaragi, Y., Enomoto, T., Watanabe, Y., Hanaoka, F., and Yamada, M. (1984). Multiple deoxyribonucleic acid dependent adenosinetriphosphatases in FM3A cells. Characterization of an adenosinetriphosphatase that prefers poly [d(A-T)] as cofactor. Biochemistry 23, 529–533.

Teo, G., Liu, G., Zhang, J., Nesvizhskii, A.I., Gingras, A.-C., and Choi, H. (2014). SAINTexpress: improvements and additional features in Significance Analysis of INTeractome software. J Proteomics 100, 37–43.

Tkáč, J., Xu, G., Adhikary, H., Young, J.T.F., Gallo, D., Escribano-Díaz, C., Krietsch, J., Orthwein, A., Munro, M., Sol, W., et al. (2016). HELB Is a Feedback Inhibitor of DNA End Resection. Molecular Cell 61, 405–418.

Toledo, L.I., Altmeyer, M., Rask, M.-B., Lukas, C., Larsen, D.H., Lou Klitgaard Povlsen, Bekker-Jensen, S., Mailand, N., Bartek, J., and Lukas, J. (2013). ATR Prohibits Replication Catastrophe by Preventing Global Exhaustion of RPA. Cell 155, 1088–1103.

102

Tomimatsu, N., Mukherjee, B., Catherine Hardebeck, M., Ilcheva, M., Vanessa Camacho, C., Louise Harris, J., Porteus, M., Llorente, B., Khanna, K.K., and Burma, S. (2014). Phosphorylation of EXO1 by CDKs 1 and 2 regulates DNA end resection and repair pathway choice. Nature Communications 5, 3561.

Torres, J.Z., Schnakenberg, S.L., and Zakian, V.A. (2004). Saccharomyces cerevisiae Rrm3p DNA helicase promotes genome integrity by preventing replication fork stalling: viability of rrm3 cells requires the intra-S-phase checkpoint and fork restart activities. Molecular and Cellular Biology 24, 3198–3212.

Vilenchik, M.M., and Knudson, A.G. (2003). Endogenous DNA double-strand breaks: production, fidelity of repair, and induction of cancer. Proc. Natl. Acad. Sci. U.S.a. 100, 12871– 12876.

Walker, J.E., Saraste, M., Runswick, M.J., and Gay, N.J. (1982). Distantly related sequences in the alpha- and beta-subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold. The EMBO Journal 1, 945–951.

Ward, I.M., Reina-San-Martin, B., Olaru, A., Minn, K., Tamada, K., Lau, J.S., Cascalho, M., Chen, L., Nussenzweig, A., Livak, F., et al. (2004). 53BP1 is required for class switch recombination. The Journal of Cell Biology 165, 459–464.

Ward, J.F. (1988). DNA damage produced by ionizing radiation in mammalian cells: identities, mechanisms of formation, and reparability. Prog. Nucleic Acid Res. Mol. Biol. 35, 95–125.

Waters, C.A., Strande, N.T., Pryor, J.M., Strom, C.N., Mieczkowski, P., Burkhalter, M.D., Oh, S., Qaqish, B.F., Moore, D.T., Hendrickson, E.A., et al. (2014). The fidelity of the ligation step determines how ends are resolved during nonhomologous end joining. Nature Communications 5, 4286.

Weemaes, C.M., Hustinx, T.W., Scheres, J.M., van Munster, P.J., Bakkeren, J.A., and Taalman, R.D. (1981). A new chromosomal instability disorder: the Nijmegen breakage syndrome. Acta Paediatr Scand 70, 557–564.

Wold, M.S., and Kelly, T. (1988). Purification and characterization of replication protein A, a cellular protein required for in vitro replication of simian virus 40 DNA. Proc. Natl. Acad. Sci. U.S.a. 85, 2523–2527.

Xu, G., Chapman, J.R., Brandsma, I., Yuan, J., Mistrik, M., Bouwman, P., Bartkova, J., Gogola, E., Warmerdam, D., Barazas, M., et al. (2015). REV7 counteracts DNA double-strand break resection and affects PARP inhibition. Nature 521, 541–544.

Xu, X., Qiao, W., Linke, S.P., Cao, L., Li, W.M., Furth, P.A., Harris, C.C., and Deng, C.X. (2001). Genetic interactions between tumor suppressors Brca1 and p53 in apoptosis, cell cycle and tumorigenesis. Nat. Genet. 28, 266–271.

Xu, X., Vaithiyalingam, S., Glick, G.G., Mordes, D.A., Chazin, W.J., and Cortez, D. (2008). The basic cleft of RPA70N binds multiple checkpoint proteins, including RAD9, to regulate ATR signaling. Molecular and Cellular Biology 28, 7345–7353. 103

Yamane, A., Robbiani, D.F., Resch, W., Bothmer, A., Nakahashi, H., Oliveira, T., Rommel, P.C., Brown, E.J., Nussenzweig, A., Nussenzweig, M.C., et al. (2013). RPA Accumulation during Class Switch Recombination Represents 5′-3′ DNA-End Resection during the S-G2/M Phase of the Cell Cycle. CellReports 3, 138–147.

Yang, J., O'Donnell, L., Durocher, D., and Brown, G.W. (2012). RMI1 promotes DNA replication fork progression and recovery from replication fork stress. Molecular and Cellular Biology 32, 3054–3064.

Yu, X., and Chen, J. (2004). DNA damage-induced cell cycle checkpoint control requires CtIP, a phosphorylation-dependent binding partner of BRCA1 C-terminal domains. Molecular and Cellular Biology 24, 9478–9486.

Yu, Z.K., Gervais, J.L., and Zhang, H. (1998). Human CUL-1 associates with the SKP1/SKP2 complex and regulates p21(CIP1/WAF1) and cyclin D proteins. Proc. Natl. Acad. Sci. U.S.a. 95, 11324–11329.

Yuan, J., Ghosal, G., and Chen, J. (2009). The annealing helicase HARP protects stalled replication forks. Genes & Development 23, 2394–2399.

Yun, M.H., and Hiom, K. (2009). CtIP-BRCA1 modulates the choice of DNA double-strand- break repair pathway throughout the cell cycle. Nature 459, 460–463.

Yusufzai, T., Kong, X., Yokomori, K., and Kadonaga, J.T. (2009). The annealing helicase HARP is recruited to DNA repair sites via an interaction with RPA. Genes & Development 23, 2400– 2404.

Zhou, R., Zhang, J., Bochman, M.L., Zakian, V.A., and Ha, T. (2014). Periodic DNA patrolling underlies diverse functions of Pif1 on R-loops and G-rich DNA. Elife 3, e02190.

Zhu, Z., Chung, W.-H., Shim, E.Y., Lee, S.E., and Ira, G. (2008). Sgs1 helicase and two nucleases Dna2 and Exo1 resect DNA double-strand break ends. Cell 134, 981–994.

Zierhut, C., and Diffley, J.F.X. (2008). Break dosage, cell cycle stage and DNA replication influence DNA double strand break response. The EMBO Journal 27, 1875–1885.

Zimmermann, M., Lottersberger, F., Buonomo, S.B., Sfeir, A., and de Lange, T. (2013). 53BP1 regulates DSB repair using Rif1 to control 5' end resection. Science 339, 700–704.

104

Appendix 1: Mass spectrometry data

105

Table A.1 SAINTexpress (v 3.3) output for IP-MS using RPA1 as bait Only prey genes with a SAINT score ≥ 0.7 are included.

Prey Gene: per NCBI Entrez Gene Prey Accession: NCBI protein accession number Spectra: spectral counts for the prey (replicates separated by "I" delimiter) Control spectra: spectral counts for the prey across the negative controls (replicates separated by "I" delimiter) SAINT: final Significance Analysis of INTeractome score (Choi et al., 2011) AvgP: averaged probability across all replicates MaxP: maximal probability BFDR: Bayesian false discovery rate

Prey Gene Prey Accession Spectra Control Spectra SAINT AvgP MaxP BFDR RPA2 4506585 359|558|489 3|4|5 1 1 1 0 RPA3 4506587 75|118|93 1|3|1 1 1 1 0 PARP1 156523968 54|19|18 3|6|5 1 1 1 0 HELB 32455236 29|21|20 0|0|0 1 1 1 0 SMARCAL1 187761314 28|18|16 0|0|0 1 1 1 0 POLA1 106507301 8|18|23 0|3|2 1 1 1 0 RFWD3 71143112 11|11|18 0|0|0 1 1 1 0 CHTF18 27501458 14|13|9 0|0|0 1 1 1 0 TOP3A 10835218 20|6|9 0|0|0 1 1 1 0 BLM 4557365 24|4|5 0|0|0 1 1 1 0 UNG 19718751 8|10|8 0|0|0 1 1 1 0 RMI1 153791761 10|7|8 0|0|0 1 1 1 0 POLA2 20127448 4|10|9 0|1|1 1 1 1 0 CCDC111 22749373 8|8|6 0|0|0 1 1 1 0 ZUFSP 292494919 8|8|5 0|0|0 1 1 1 0 SDF4 18699732 5|5|4 0|0|0 1 1 1 0 C17orf53 284005249 2|6|5 0|0|0 1 1 1 0 IPO7 5453998 11|19|17 0|2|4 0.99 0.99 1 0 PRIM2 41349495 8|10|11 1|1|3 0.99 0.98 0.99 0 MSH3 284813531 16|4|3 0|0|0 0.99 0.99 1 0 RFC5 194306567 11|4|3 0|0|0 0.99 0.99 1 0 HIST2H2AA4 106775678 0|4|7 0|0|0 0.99 0.99 1 0 BRIP1 301897118 2|4|4 0|0|0 0.99 0.99 0.99 0 SRSF7 306482694 5|4|0 0|0|0 0.99 0.99 1 0 DNA2 320461728 0|4|4 0|0|0 0.99 0.99 0.99 0 PRIM1 4506051 7|10|12 0|0|2 0.98 0.98 0.99 0 RFC4 31881687 11|8|3 0|3|2 0.88 0.82 0.97 0.03 WRN 110735439 14|0|2 0|0|0 0.85 0.85 1 0.01 ETAA1 37059814 4|2|2 0|0|0 0.85 0.85 0.99 0.02 HERC2 126032348 5|2|0 0|0|0 0.85 0.85 1 0.01 TECR 24475816 0|4|3 0|0|1 0.84 0.84 0.97 0.02 RAD52 109637798 3|2|2 0|0|0 0.82 0.82 0.94 0.02 RPAIN 237858695 3|0|2 0|0|0 0.82 0.82 0.94 0.02 ADD1 29826319 1|2|2 0|0|0 0.7 0.7 0.7 0.04 PGM1 21361621 0|2|2 0|0|0 0.7 0.7 0.7 0.04 CTSB 22538431 0|2|2 0|0|0 0.7 0.7 0.7 0.04

106

Table A.2 SAINTexpress (v 3.3) output for IP-MS using RPA2 as bait Only prey genes with a SAINT score ≥ 0.7 are included.

Prey Gene Prey Accession Spectra Control Spectra SAINT AvgP MaxP BFDR RPA1 4506583 783|826|655 4|14|15 1 1 1 0 RPA3 4506587 121|201|157 1|3|1 1 1 1 0 HELB 32455236 34|24|28 0|0|0 1 1 1 0 SMARCAL1 187761314 29|18|14 0|0|0 1 1 1 0 RFWD3 71143112 12|22|24 0|0|0 1 1 1 0 UNG 19718751 15|15|15 0|0|0 1 1 1 0 TOP3A 10835218 20|7|13 0|0|0 1 1 1 0 BLM 4557365 18|9|12 0|0|0 1 1 1 0 MEN1 18860839 12|13|11 0|0|0 1 1 1 0 POLA1 106507301 1|15|16 0|3|2 1 1 1 0 RMI1 153791761 10|9|11 0|0|0 1 1 1 0 FAM115A 332635045 7|10|5 0|0|0 1 1 1 0 CHTF18 27501458 5|6|10 0|0|0 1 1 1 0 MSH3 284813531 10|2|6 0|0|0 1 1 1 0 HIST2H2AA4 106775678 0|9|9 0|0|0 1 1 1 0 ZUFSP 292494919 7|4|6 0|0|0 1 1 1 0 PCSK5 299523015 8|0|9 0|0|0 1 1 1 0 PARP1 156523968 27|14|17 3|6|5 0.99 0.98 1 0 BRIP1 301897118 1|7|4 0|0|0 0.99 0.99 1 0 TIPIN 157388910 3|4|4 0|0|0 0.99 0.99 0.99 0 RPAIN 237858695 3|3|4 0|0|0 0.96 0.96 0.99 0 PRIM2 41349495 5|9|10 1|1|3 0.95 0.94 0.97 0 PRIM1 4506051 5|8|9 0|0|2 0.91 0.91 0.94 0.01 DNA2 320461728 2|4|0 0|0|0 0.85 0.85 0.99 0.02 DSP 58530840 0|3|2 0|0|0 0.82 0.82 0.94 0.02 RAD52 109637798 2|2|2 0|0|0 0.7 0.7 0.7 0.04 HIST1H2BB 10800140 1|2|2 0|0|0 0.7 0.7 0.7 0.04 THOC3 14150171 2|2|0 0|0|0 0.7 0.7 0.7 0.04 ETAA1 37059814 2|0|2 0|0|0 0.7 0.7 0.7 0.04

107

Table A.3 SAINTexpress (v 3.3) output for IP-MS using RPA3 as bait Only prey genes with a SAINT score ≥ 0.7 are included.

Prey Gene Prey Accession Spectra Control Spectra SAINT AvgP MaxP BFDR RPA1 4506583 740|535|484 4|14|15 1 1 1 0 RPA2 4506585 320|191|162 3|4|5 1 1 1 0 SMARCAL1 187761314 24|26|20 0|0|0 1 1 1 0 PARP1 156523968 25|19|18 3|6|5 1 1 1 0 BLM 4557365 23|19|10 0|0|0 1 1 1 0 TOP3A 10835218 20|18|12 0|0|0 1 1 1 0 HELB 32455236 21|18|10 0|0|0 1 1 1 0 POLA1 106507301 9|15|12 0|3|2 1 0.99 1 0 RMI1 153791761 11|14|9 0|0|0 1 1 1 0 RFWD3 71143112 8|9|8 0|0|0 1 1 1 0 MSH3 284813531 16|3|5 0|0|0 1 1 1 0 UNG 19718751 9|8|6 0|0|0 1 1 1 0 WRN 110735439 13|7|2 0|0|0 1 1 1 0 ZUFSP 292494919 9|5|4 0|0|0 1 1 1 0 CHTF18 27501458 7|7|3 0|0|0 1 1 1 0 AIFM2 14318424 7|5|4 0|0|0 1 1 1 0 RAD52 109637798 5|2|5 0|0|0 1 1 1 0 POLA2 20127448 5|7|8 0|1|1 0.99 0.99 0.99 0 HERC2 126032348 10|4|1 0|0|0 0.99 0.99 1 0 CCDC111 22749373 3|5|4 0|0|0 0.99 0.99 1 0 HIST2H2AA4 106775678 0|5|4 0|0|0 0.99 0.99 1 0 IPO7 5453998 12|16|16 0|2|4 0.98 0.98 0.98 0 RFC5 194306567 5|3|1 0|0|0 0.97 0.97 1 0 LCP1 167614506 0|3|4 0|0|0 0.96 0.96 0.99 0 TPT1 4507669 0|4|3 0|0|0 0.96 0.96 0.99 0 KRT1 119395750 14|0|20 0|0|4 0.9 0.89 0.96 0.01 PRIM1 4506051 4|8|8 0|0|2 0.88 0.88 0.88 0.01 ETAA1 37059814 5|2|2 0|0|0 0.85 0.85 1 0.01 BRIP1 301897118 2|2|4 0|0|0 0.85 0.85 0.99 0.02 SRSF7 306482694 5|2|0 0|0|0 0.85 0.85 1 0.01 KRT10 195972866 2|0|4 0|0|0 0.85 0.85 0.99 0.02 C10orf2 39725942 3|2|0 0|0|0 0.82 0.82 0.94 0.02 PRIM2 41349495 7|7|8 1|1|3 0.76 0.66 0.78 0.06 RPAIN 237858695 2|2|2 0|0|0 0.7 0.7 0.7 0.04 NAMPT 5031977 2|2|1 0|0|0 0.7 0.7 0.7 0.04 H2AFZ 4504255 0|2|2 0|0|0 0.7 0.7 0.7 0.04

108

Table A.4 SAINTexpress (v 3.3) output for IP-MS using HELB as bait Only prey genes with a SAINT score ≥ 0.7 are included.

Prey Gene Prey Accession Spectra Control Spectra SAINT AvgP MaxP BFDR RPA1 4506583 424|817|776 4|14|15 1 1 1 0 RPA2 4506585 237|310|321 3|4|5 1 1 1 0 RPA3 4506587 46|49|53 1|3|1 1 1 1 0 CCNA2 4502613 34|35|45 0|0|0 1 1 1 0 CDK2 16936528 21|30|34 0|0|0 1 1 1 0 CDK1 16306492 21|28|36 4|5|7 1 1 1 0 SKP2 16306595 31|20|22 0|0|0 1 1 1 0 GSK3A 49574532 15|25|26 0|1|0 1 1 1 0 GSK3B 21361340 13|19|25 0|0|0 1 1 1 0 SKP1 25777713 9|19|17 2|3|3 1 1 1 0 SMARCAL1 187761314 12|7|6 0|0|0 1 1 1 0 UNG 19718751 8|8|9 0|0|0 1 1 1 0 CDC20 118402582 4|4|12 0|0|0 1 0.99 1 0 TOP3A 10835218 8|5|3 0|0|0 1 1 1 0 BLM 4557365 6|5|2 0|0|0 1 1 1 0 RMI1 153791761 7|4|3 0|0|0 0.99 0.99 1 0 SCAI 116256473 5|4|2 0|0|0 0.99 0.99 1 0 MSH3 284813531 12|3|3 0|0|0 0.97 0.97 1 0 CCNB1 14327896 4|6|8 1|0|2 0.94 0.85 0.97 0.01 CKS2 4502859 1|3|3 0|0|0 0.94 0.94 0.94 0.01 CKS1B 4502857 2|2|3 0|0|0 0.88 0.82 0.94 0.02 ARGLU1 134152708 15|7|13 4|0|0 0.82 0.82 0.86 0.03 SRSF7 306482694 0|3|2 0|0|0 0.82 0.82 0.94 0.02 CNBP 187608726 0|3|2 0|0|0 0.82 0.82 0.94 0.02 SH3GL1 317108191 0|3|3 0|0|1 0.71 0.71 0.71 0.03

109

Appendix 2: Pathology reports of two Helb+/+ and two Helb-/- mice

110

111

112

113

114

115

116

117

Appendix 3: Loss of HELB results in PARP inhibitor resistance and partial restoration of HR in BRCA1-deficient cells

118

119

Figure A.1 Loss of HELB results in PARP inhibitor resistance in BRCA1-deficient cells (A) Brca1-/- p53-/- murine breast tumor cells were transduced with a virus expressing either a non-targeting shRNA (shCTRL) or two independent shRNAs targeting Helb. Whole cell extracts were obtained and analyzed by immunoblotting using antibodies against HELB and tubulin (loading control). (B) HELB depletion in BRCA1-deficient cells results in resistance to PARP inhibitors. Representative plates from a clonogenic experiment are shown. At day 0, 10000 G3 cells expressing the indicated shRNAs were seeded and left untreated or were treated with the PARP inhibitors olaparib or AZD2461 at the indicated concentrations. Medium was refreshed with olaparib or AZD2461 every 4 days. On day 5, the untreated control group was stopped and the other groups were stopped together after another 2-3 weeks and stained with crystal violet. (C) Quantitation of the clonogenic assays in (B). Data is presented as the mean +/- SD, n = 3. (D) The PARP inhibitor resistance of HELB-depleted Brca1-/- p53-/- cells is dependent on ATM activity. Clonogenic experiments were performed as in (B), but in addition to treatment with PARP inhibitors, some samples were also treated with the ATM inhibitors KU55933 or KU60019.

120

Figure A.2 HELB requires its resection-inhibiting activities to mediate the cytotoxicity of PARP inhibitors in BRCA1-deficient cells (A) Whole cell extracts from Brca1-/- p53-/- mouse mammary tumor cells expressing a non- targeting shRNA (CTRL) or shRNAs against 53bp1, Mad2l2, or Helb, and stably transduced with an empty lentivirus (EV) or a virus expressing the indicated GFP-HELB protein were analyzed by immunoblotting with HELB and tubulin (loading control) antibodies. (B) Long-term clonogenic assay using Brca1-/- p53-/- cells transduced with the indicated shRNA and GFP-HELB expression constructs and treated with the indicated concentration of olaparib, AZD2461, or left untreated. Shown here are representative images of the culture dishes following staining of the cells with crystal violet.

121

(C) Quantitation of the clonogenic assay shown in (B) by determining the absorbance of extracted crystal violet. The data were normalized to the absorbance of the shHelb + empty vector (EV) control for each treatment. Data is presented as the mean +/- SD, n = 3. (D) Brca1-/- p53-/- cells were first rendered resistant to PARP inhibitor treatment by transduction of shRNAs targeting 53bp1 or Mad2l2. Subsequently, these cells were transduced with a retroviral GFP-HELB expression construct or empty control vector (EV) and treated with olaparib, AZD2461, or left untreated. The clonogenic experiment was performed as in (B). (E) Quantitation of clonogenic assays shown in (D). Data is presented as the mean +/- SD, n = 3.

Figure A.3 Loss of HELB results in PARP inhibitor resistance in BRCA1-deficient mammary tumor allografts (A) Relative tumor volume in allografted mice injected with Brca1-/- p53-/- cells expressing either a control shRNA (shCTRL) or an shRNA targeting Helb (shHelb) and treated with one regimen of 50 mg olaparib per kg daily for 28 days or treated with vehicle. Tumor volume data is expressed as a percentage relative to day 0. (B) Survival of allografted mice injected with Brca1-/- p53-/- cells expressing either a control shRNA (shCTRL) or an shRNA targeting Helb (shHelb) and treated with one regimen of 50 mg olaparib per kg daily for 28 days or treated with vehicle. The indicated p value was calculated using a log-rank test.

122

Figure A.4 Loss of HELB leads to a partial reactivation of HR in BRCA1-deficient cells A) HELB depletion restores RAD51 focus formation in BRCA1-deficient cells. Brca1-/- p53-/- cells depleted of HELB using two independent shRNAs or transduced with a non-targeting shRNA (shCTRL) were fixed 5 h following a 10 Gy dose of IR and then processed for RAD51 immunofluorescence. DNA was counterstained with DAPI. Scale bar = 10 µm. (B) Quantitation of RAD51 focus formation shown in (A). Data is presented as the mean +/- SD, n=3. (C) U2OS cells transfected with the indicated combination of siRNAs were fixed 3 h following a 10 Gy dose of IR and then processed for RAD51 immunofluorescence. Data is presented as the mean +/- SEM, n = 3. (D) Whole cell extracts from U2OS cells transfected with the indicated combination of siRNAs were analyzed by immunoblotting using antibodies against BRCA1, 53BP1, HELB, RAD51 and actin (loading control).

123