Support of Mitochondrial DNA Replication by Human Rad51: a Dissertation

University of Massachusetts Medical School eScholarship@UMMS

GSBS Dissertations and Theses Graduate School of Biomedical Sciences

2011-12-13

Support of Mitochondrial DNA Replication by Human Rad51: A Dissertation

Jay M. Sage University of Massachusetts Medical School

Let us know how access to this document benefits ou.y Follow this and additional works at: https://escholarship.umassmed.edu/gsbs_diss

Part of the Amino Acids, Peptides, and Proteins Commons, Biochemical Phenomena, Metabolism, and Nutrition Commons, Biochemistry, Biophysics, and Structural Biology Commons, Cells Commons, Enzymes and Coenzymes Commons, and the Genetic Phenomena Commons

Repository Citation Sage JM. (2011). Support of Mitochondrial DNA Replication by Human Rad51: A Dissertation. GSBS Dissertations and Theses. https://doi.org/10.13028/mj6a-mr98. Retrieved from https://escholarship.umassmed.edu/gsbs_diss/574

This material is brought to you by eScholarship@UMMS. It has been accepted for inclusion in GSBS Dissertations and Theses by an authorized administrator of eScholarship@UMMS. For more information, please contact [email protected]. Support of Mitochondrial DNA Replication by Human Rad51

A Dissertation Presented

Jay M. Sage

Submitted to the Faculty of the University of Massachusetts Graduate School of Biomedical Sciences, Worcester in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

December 13, 2011

Biochemistry and Molecular Pharmacology

Support of Mitochondrial DNA Replication by Human Rad51 A Dissertation Presented By Jay M. Sage

The signatures of the Dissertation Defense Committee signifies completion and approval as to style and content of the Dissertation

______Dr. Kendall Knight, Thesis Advisor

______Dr. Sharon Cantor, Member of Committee

______Dr. Martin Marinus, Member of Committee

______Dr. Bennett Van Houten, Member of Committee

______Dr. Michael Volkert, Member of Committee

The signature of the Chair of the Committee signifies that the written dissertation meets the requirements of the Dissertation Committee

______Dr. Nicholas Rhind, Chair of Committee

The signature of the Dean of the Graduate School of Biomedical Sciences signifies that the student has met all graduation requirements of the school.

______Anthony Carruthers, Ph.D., Dean of the Graduate School of Biomedical Sciences

Biochemistry and Molecular Pharmacology December 13, 2011

This thesis is dedicated to my incredibly supportive parents Tom and Eileen, brother Bill, and my wonderful boyfriend Nathan. Your love and steadfast support have given me the strength to realize a lifelong dream.

iii

Parts of this dissertation have appeared in the following:

Sage J.M., Gildemeister, O.S., and Knight K.L. (2010) Discovery of a Novel Function for Human Rad51: Maintenance of the Mitochondrial Genome. J Biol Chem, 285 (25), 18984-90.

Figure Number Publisher License Number

Figure 1.1 Elsevier 2810960218462

Figure 1.2 Nature Publishing Group 2810961077487

Figure 1.4 Elsevier 2810970124156

Figure 1.6 Elsevier 2810961248189

ACKNOWLEDGEMENTS

I would like to give sincere thanks to my mentor, Kendall Knight. In addition to support and guidance, he has allowed me the independence to make choices and see them through. I have learned and grown much over the past six years, and I will always be grateful. I must also thank the members of my thesis advisory committee, Dr. Sharon Cantor, Dr. Martin Marinus, Dr.

Michael Volkert, and my chair Dr. Nicholas Rhind. The depth and range of their experiences provided a reassuring foundation throughout the course of this work. I would like to give special thanks to Dr. Bennett Van Houten for taking time from his busy schedule to review my work and sit on my defense committee.

I would like to thank several people I have had the pleasure of working with and getting to know over the past several years. Dr. Elisabet Mandon has been a great help in designing experiments and learning the ropes in the department. I am also indebted to several members of the Carruthers lab who have been there for me as a sounding board when perspective was needed to think objectively about “The Work”. In particular, I would like to thank Anthony and Sabrina for keeping me sane through the ups and downs of graduate school, and for their extensive help in many rounds of editing of this thesis.

I must once again thank my family for their support and love throughout the years. They have given me the confidence to do things in life that are not easy and the drive to finish the things that I start. I promise I will get a real job soon…

ABSTRACT

The function of homologous DNA recombination in human mitochondria has been a topic of ongoing debate for many years, with implications for fields ranging from DNA repair and mitochondrial disease to population genetics. While genetic and biochemical evidence supports the presence of a mitochondrial recombination activity, the purpose for this activity and the proteins involved have remained elusive. The work presented in this thesis was designed to evaluate the mitochondrial localization of the major recombinase protein in human cells, Rad51, as well as determine what function it plays in the maintenance of mitochondrial DNA (mtDNA) copy number that is critical for production of chemical energy through aerobic respiration. The combination of subcellular fractionation with immunoblotting and immunoprecipitation approaches used in this study clearly demonstrates that Rad51 is a bona fide mitochondrial protein that localizes to the matrix compartment following oxidative stress, where it physically interacts with mtDNA. Rad51 was found to be critical for mtDNA copy number maintenance under stress conditions. This requirement for Rad51 was found to be completely dependent on ongoing mtDNA replication, as treatment with the DNA polymerase gamma (Pol ϒ) inhibitor, ddC, suppresses both recruitment of Rad51 to the mitochondria following the addition of stress, as well as the mtDNA degradation observed when Rad51 has been depleted from the cell.

The data presented here support a model in which oxidative stress induces a three-part response:

(1) The recruitment of repair factors including Rad51 to the mitochondrial matrix, (2) the activation of mtDNA degradation systems to eliminate extensively or persistently damaged mtDNA, and (3) the increase in mtDNA replication in order to maintain copy number. The stress-induced decrease in mtDNA copy number observed when Rad51 is depleted is likely the

result of failure to stabilize or repair replication forks that encounter blocking lesions resulting in further damaged to the mtDNA and its eventual degradation.

vii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS……………………………………………………………………….iv

ABSTRACT………………………………………………………………………………………..v

TABLE OF CONTENTS………………………………………………………………………...vii

List of Figures………………………………………………………………………..…...ix

List of Abbreviations……………………………………………………………………...x

CHAPTER I………………………………………………………………………………………..1

INTRODUCTION AND LITERATURE REVIEW………………………………………………1

The Repair of DNA Double Strand Breaks……………………………………………….3

The Human Rad51 Protein………………………………………………………………..4

Rad51-mediated Homologous Recombination……………………………………………5

Recombination in Budding Yeast and E. coli……………………………………………..8

The Nuclear DNA Replication Fork……………………………………..………………11

Replication Restart and Repair of Collapsed Forks……………………………………...12

The Mitochondrion………………………………………………………………………15

mtDNA…………………………………………………………………………………...18

mtDNA Replication…………………………………………………………………...…20

Yeast mtDNA Replication……………………………………………………………….26

Maintenance of Mammalian mtDNA Copy Number…………………………………….27

mtDNA Repair and Degradation………………………………………………………...30

viii

Remaining Questions…………………………………………………………………….34

CHAPTER II……………………………………………………………………………………...37

Abstract…………………………………………………………………………………..37

Introduction………………………………………………………………………………38

Experimental Procedures………………………………………………………………...39

Results and Discussion…………………………………………………………………..43

CHAPTER III…………………………………………………………………………………….58

Abstract…………………………………………………………………………………..58

Introduction………………………………………………………………………………59

Experimental Procedures………………………………………………………………...60

Results…………………………………………………………………………….……...63

Discussion……………………………………………………………………….……….77

CHAPTER IV…………………………………………………………………………….………81

CONCLUSIONS AND FUTURE DIRECTIONS…………….………………………………….81

APPENDIX…………………………………….…………………………………………………90

BIBLIOGRAPHY………………………………………………………………………………...95

List of Figures Figure 1.1: Double-strand break repair by Rad51-mediated homologous recombination……………………………………………………………………………..6 Figure 1.2: Models of mammalian replication fork restart………………………………14 Figure 1.3: Electron transport chain of the mitochondria ……………………………….17 Figure 1.4: The human mitochondrial genome…………………………………………..19 Figure 1.5: BrdU-labeled mtDNA in human cell………………………………………...21 Figure 1.6: Replication modes of mammalian mtDNA………………………………….24 Figure 2.1: Purification of mitochondrial proteins from cultured human cells………….44 Figure 2.2: DNA damage-induced increases in mitochondrial Rad51, Rad51C, and Xrcc3…………………………………………………………………….…………..46

Figure 2.3: Production of H2O2 by glucose oxidase………………………….…………..48 Figure 2.4: Rad51 physically interacts with mtDNA in an oxidative stress- dependent manner………………………………………..………………………………50 Figure 2.5: Depleting cells of Rad51 leads to cell cycle arrest and a characteristic arrest-induced increase in mtDNA copy number………………………………………..52 Figure 2.6: Depleting cells of Rad51, Rad51C, or Xrcc3 leads to a stress-induced reduction in mtDNA copy number………………………………………………………54 Figure 3.1: Replication inhibitors selectively prevent incorporation of BrdU into newly synthesized DNA……………………………………….………………………...64 Figure 3.2: Recruitment of Rad51 to the mitochondria in response to stress is suppressed by inhibition of mtDNA replication…..……………………………………..66 Figure 3.3: Stress-induced mtDNA depletion in the absence of Rad51 requires ongoing mtDNA replication…………..…………………………………………………68 Figure 3.4: Ethidium bromide selectively inhibits incorporation of BrdU into mitochondrial DNA……………………….……………………………………………..69 Figure 3.5: Recovery of mtDNA copy number following mtDNA depletion recruits Rad51 to the mitochondria……………………………………………………………….71

Figure 3.6: Transfection with Rad51-specific siRNAs depletes cells of protein………...72

Figure 3.7: Depleting cells of Rad51 does not affect the rate of mtDNA copy number recovery following mtDNA depletion…….…………………………………….74 Figure 3.8: Effect of cell division on mtDNA replication demand during recovery from mtDNA depletion…………..………………..……………………………………..75 Figure 3.9: Effect of Rad51 depletion on total mtDNA synthesized following removal of ethidium bromide……………………...……………………………………..76 Figure 3.10: Depletion of mtDNA copy number through dilution following ddC treatment………………………………………...…………………………………..78 Figure A1: DNA damage-induced increases in mitochondrial Rad51 and Rad51C in HeLa cells……………………………………………………………………………..91 Figure A2: Southern blot analysis of changes in mtDNA copy number following long-term Rad51 depletion………………………………………………………………92 Figure A3: Treatment of cells with 20 mU/mL GO does not disturb mtDNA conformatio distribution…………………………………………………………………93 Figure A4: Recruitment of Rad51 to the mitochondria in response to stress is suppressed by inhibition of mtDNA replication in HCT116 cells…………………...….94

List of Abbreviations 2D-AGE 2-dimensional agarose gel electrophoresis

ATP adenosine-5′-triphosphate AZT azidothymidine BCA bicinchonic acid BER base excision repair BIR break-induced replication BrdU 5-bromo-2'-deoxyuridine DAPI 4',6-diamidino-2-phenylindole ddC 2'-3'-dideoxycytidine DMSO dimethyl sulfoxide dNTP deoxyribonucleotide DSB double strand break EDTA ethylenediaminetetraacetic acid EtBr ethidium bromide FBS fetal bovine serum G1 gap phase 1 G2 gap phase 2 GO glucose oxidase Gy gray HJ Holliday junction HR homologous recombination IR ionizing radiation M mitosis MDS mitochondrial DNA depletion syndrome MRN Mre11-Rad50-Nbs1

xii

mtDNA mitochondrial DNA NHEJ non-homologous end joining PCR polymerase chain reaction PVDF polyvinylidene fluoride qPCR quantitative polymerase chain reaction ROS reactive oxygen species rRNA ribosomal ribonucleic acid S synthesis phase SDS sodium dodecyl sulfate siRNA small interfering ribonucleic acid ssDNA single-stranded DNA

TAS termination-associated sequence

TE tris EDTA tris 2-amino-2-hydroxymethyl-propane-1,3-diol tRNA transfer RNA

UV ultraviolet

CHAPTER I:

Introduction and Literature Review

Stable inheritance of genetic material is an essential process in living things from single- celled bacteria and yeast to complex multicellular organisms. With the exception of a subset of viruses, all species on earth encode their genetic material in molecules called deoxyribonucleic acid (DNA). These molecules are composed of nucleotides containing a nitrogenous base connected to a five-carbon 2′-deoxyribose sugar. The nucleotides are linked together into strands through phosphodiester linkages. Two such strands are hydrogen bonded between bases in an antiparallel fashion to yield the double helix identified by Watson and Crick in 1953 (Watson and Crick, 1953). In humans, the majority of the more than 3 billion base pairs of DNA is contained in the cell nucleus and replicated during a specific phase of the cell cycle. The eukaryotic cell cycle is subdivided into four phases: gap phase 1 (G1) when the cell grows in size and prepares for DNA replication, synthesis (S) when the cell replicates all of its nuclear DNA, gap phase 2 (G2) when the cell prepares for division, and mitosis (M) when cytokinesis results in the production of two daughter cells, each with a complete set of chromosomes.

In addition to chromosomal DNA, a small genome is also maintained in the mitochondria of mammalian cells. Encoded in this subset of genes are proteins required for much of the energy production in the cell.

The maintenance of DNA requires faithful duplication as well as reasonably error-free repair of damaged genetic material. These processes utilize a staggering array of protein 2

factors to properly regulate when and how DNA is replicated or repaired. A central pathway linking DNA replication and repair is homologous recombination (HR). The existence of genetic recombination systems in nearly every living organism is a testament to its fundamental role in DNA metabolism. While the protein machinery varies between species, the catalytic core of HR is the recombinase protein. The recombinase is able to bind to DNA ends and catalyze a variety of DNA transactions in order to repair double strand breaks (DSBs) and allow for efficient and faithful replication of damaged DNA.

In humans, the Rad51 protein is the major recombinase. Clear roles for Rad51 have been identified in the replication and repair of chromosomal DNA in mammalian cells; however, the importance for Rad51-mediated HR mtDNA metabolism remains to be evaluated.

Given the importance of Rad51 in nuclear DNA maintenance, this thesis focuses on evaluating the role of Rad51 in the maintenance of mtDNA. Particular attention has been placed on investigating the importance of Rad51 under conditions of elevated mtDNA damage and/or replication stress. A clear understanding of the full range of mtDNA repair mechanisms is essential for dissecting the root causes of defects in mitochondrial function that lead to a variety of human diseases including Parkinson’s, Alzheimer’s, and mtDNA Depletion Syndrome.

The Repair of DNA Double Strand Breaks

The faithful repair of DNA damage is centrally important for the maintenance of genomic stability in most organisms. This damage may be the result of a variety of exogenous stresses including ultraviolet light (UV) or ionizing radiation (IR), chemotheraputic drugs, or oxidizing agents. Damage also results from normal cellular processes including exposure to the reactive oxygen species (ROS) that are byproducts of metabolism, as well as the collapse of DNA replication forks following an encounter with a preexisting DNA lesion. These varied sources of stress result in a spectrum of damage including abasic sites, strand crosslinks, strand breaks, and chemical modifications to bases or the sugar backbone. Among the most toxic of these types of damage are DSBs. If left unrepaired,

DSBs are a potent inducer of apoptosis in metazoans, while improper repair may result in genetic instability in the form of insertions, deletions, or mutations as well as gross chromosomal rearrangements. However, DSBs are also a component of programmed events in some cells, including immunoglobulin gene rearrangement and meiotic recombination. When dealing with DSBs, complex organisms must strike a delicate balance between preventing potentially catastrophic genomic instability and allowing the generation of diversity necessary for continued evolution.

In mammalian cells, DSBs are repaired primarily by two pathways: ligation of the broken ends by non-homologous end joining (NHEJ) and HR. The choice of repair mechanism depends on several factors including cell type and the phase of the cell cycle in which the break is generated (Rothkamm et al., 2003). 4

Active throughout the cell cycle, NHEJ is the primary mode of DSB repair in mammalian cells during G1. Rejoining of the broken DNA begins by tethering the pieces together through the binding of the Ku70/80 heterodimer to the two DNA ends. This in turn leads to the recruitment of DNA-dependent protein kinase (DNA-PK) that phosphorylates a variety of targets, and is thought to facilitate the access of end processing enzymes. The nucleases involved in NHEJ end processing are not well understood, but may be responsible for removing any damaged bases in preparation for religation by DNA ligase

IV.

In contrast, HR is critical during S and G2 when it is required for the repair of stalled or collapsed replication forks, and sister chromatids are present (Rothkamm et al., 2003;

Hartlerode et al., 2011). In yeast, the choice of DSB repair pathway is determined at the step of end resection. The resection step commits the break to repair by HR as the removal of even a few nucleotides inhibits NHEJ (Daley et al., 2005).

The Human Rad51 Protein

The central catalytic component of HR in mammalian cells is the Rad51 recombinase protein. Rad51 protein is a 339-amino acid, 37 kDa member of the AAA+ superfamily of

ATPases, that binds and hydrolyzes ATP. Like its bacterial homolog RecA, Rad51 is capable of forming long polymers that wrap around single or double-stranded DNA

(Ristic et al., 2005). Rad51 is expressed mainly in S and G2 phases of the cell cycle and is localized in both the cytosol and nucleus (Flygare et al., 1996; Davies et al., 2001; 5

Forget et al., 2004; Gildemeister et al., 2009). While Rad51 expression appears not to change in response to DNA damage (Chen et al., 1997; Henson et al., 2006), part of the

DNA damage response involves movement of Rad51 from the cytosol into the nucleus

(Gildemeister et al., 2009; Sage et al., 2010). DNA damage also results in the phosphorylation of S. cerevisiae Rad51 at Serine 192 by the kinase Mec1 (Flott et al.,

2011). This modification stimulates ATP hydrolysis and is critical for Rad51 function in budding yeast.

Rad51-mediated Homologous Recombination

Following the generation of a DSB in mammalian cells, the multisubunit complex of

Mre11-Rad50-Nbs1 (MRN) localizes to the break site along with the nuclease CtIP, resulting in limited resection of the 5′ DNA end (Sartori et al., 2007). MRN then aids in the recruitment and activation of the ExoI/Dna2 exonucleases along with the RecQ –like

Bloom’s syndrome helicase (BLM) in order to perform extensive 5′-3′ resection

(Nimonkar et al., 2011; Nimonkar et al., 2008) (Figure 1.1a). This resection leads to a 3′ single-stranded overhang that is rapidly bound by replication protein A (RPA), the human analog of bacterial single-strand DNA binding protein (SSB). The trimeric RPA protein eliminates secondary structure in the single-stranded DNA (ssDNA) and leads to the recruitment of the ATR-ATRIP kinase complex, which signals through Chk1 to achieve

S phase check point activation and cell cycle arrest (Zou and Elledge, 2003). The stabilization of the 3′ single-stranded region allows loading of Rad51 (Figure 1.1b).

Rad51 requires the activity of additional factors to displace RPA and allow formation of 6

Figure 1.1: Double strand break repair by Rad51-mediated homologous recombination (a) Detection of a double strand breaks leads to the recruitment of the MRN/CtIP complex, leading to limited processing of the 5’ end. Extensive resection is performed by BLM helicase in conjunction with Exo1 and Dna2. (b) Single-stranded DNA is bound initially by RPA and subsequently displaced by the recombination mediator BRCA2 to allow loading of the Rad51 recombinase. (c) The Rad51 nucleoprotein filament locates and invades a homologous sequence within an intact chromosome, producing a D-loop structure. (d) The Rad51-coated second end anneals to the displaced strand, and repair synthesis is performed by DNA polymerase δ. (e) Ligation of the resulting nicks results in the formation of two Holliday junction structures. (f) These junctions may then be cleaved by endonucleases, including Gen1 and Slx1/4, to yield crossover or non-crossover products. Alternatively, (g,h) through the combined activities of helicases and topoisomerases, junctional molecules may be dissolved without crossover. Adapted from Suwaki et al. Semin Cell Dev Biol. 2011 with permission.

the nucleoprotein filament. The best characterized of these recombination “mediators” is the breast cancer susceptibility protein BRCA2. By virtue of eight highly conserved

BRC repeats as well as a carboxy-terminal binding site, BRCA2 is able to bind multiple monomers of Rad51 and stimulate their loading onto DNA while displacing ssDNA- bound RPA (Figure 1.1b) (Jensen et al., 2010). Additional factors have also been shown to be required for the recruitment of Rad51 to the sites of DNA breaks, including the

Rad51 paralog proteins Rad51B, Rad51C, Rad51D, Xrcc2, and Xrcc3 (Bishop et al.,

1998; Takata et al., 2000; Gasior et al., 2001; Sonoda et al., 2001). In HeLa cells, these paralogs have been shown biochemically to form two distinct complexes: Rad51C-Xrcc3

(CX3) and Rad51B-Rad51C-Rad51D-Xrcc2 (BCDX2) (Masson et al., 2001). While the functional importance of these complexes is still largely unclear, deficiency in any of the paralog proteins results in increased sensitivity to DNA damaging agents. In addition, all five proteins are required for the accumulation of Rad51 at the sites of DSBs as observed in the formation of DNA damage-induced Rad51 nuclear foci (Takata et al., 2001).

The binding of Rad51 to the resected end extends the DNA by ~50% (Sheridan et al.,

2008), and enables the search for a homologous sequence on a sister chromatid or homologous chromosome. Once a homologous sequence is located, the 3′ end of the

Rad51 filament invades the intact duplex, displacing one strand to form a “D-loop” structure (Figure 1.1c). This 3′ end then serves as a primer that is extended by DNA polymerase δ (Maloisel et al., 2008) (Figure 1d). The use of an undamaged strand as a guide for the repair synthesis at this step is the key to the error-free nature of Rad51- 8

mediated recombination (Figure 1.1d). The four-way DNA structures, known as

Holliday junctions (HJ), that form as a result of strand invasion must be resolved in order restore the intact duplexes. This resolution may be achieved without exchange of sequence between duplexes through direct dissolution of the HJ prior to ligation by the activities of BLM helicase and topoisomerase III (Top3) (Figure 1.1g). Conversely, the four-stranded structures may be cleaved symmetrically by the structure-specific endonucleases Gen1 (Ip et al., 2008) or Slx1/4 (Figure 1.1f,h) (Andersen et al., 2009;

Fekairi et al., 2009; Munoz et al., 2009). Depending on which strands are cleaved and religated, the end result of the repair event is either a patch product, where small stretches of sequence are exchanged between the duplexes, or a crossover product, where entire arms of the chromosomes are exchanged. The stage at which the Rad51 filament disassembles is not clearly understood; however, this process is essential for HR completion and is associated with hydrolysis of the ATP bound to Rad51 (van Mameren et al., 2009).

Recombination in Budding Yeast and E. coli

While NHEJ is the major pathway for DSB repair in mammals, HR is the predominant pathway in budding yeast (Argueso et al., 2008). Recombination in S. cerevisiae utilizes homologs of many of the factors found in mammalian cells, and generally proceeds by the same mechanism. One major difference between the two species is the role of the

Rad52 protein. In yeast, Rad52 expression is essential for HR and is the defining 9

member of an epistasis group that includes Rad51. Rad52 interacts with Rad51 and facilitates the displacement of RPA and the loading of Rad51, similar to the function of the HR mediator BRCA2 in mammals (Figure 1.1b) (Krejci et al., 2002; Seong et al.,

2008). In addition, Rad52 also stimulates ssDNA annealing and may play a role in capturing the second end of a DSB (Figure 1.1d) (Mortensen et al., 1996). In sharp contrast to its critical role in yeast, the mammalian homolog of Rad52 appears to be largely redundant. The mouse model in which the RAD52 gene has been knocked out displays little to no defect in DSB repair, HR efficiency, or meiotic recombination

(Rijkers et al., 1998). The significance of this difference and what function mammalian

Rad52 serves remain to be determined.

HR in bacteria has been most studied in E. coli, where the major method of DSB repair is the RecBCD pathway (Dillingham and Kowalczykowski, 2008). The RecBCD enzyme is a helicase-nuclease complex that binds to broken DNA ends. It then unwinds the

DNA, degrading both strands as it progresses until a specific sequence (5′GCTGGTGG-

3′) is reached. The sequence, referred to as a Chi site, is found ~1000 times throughout the E. coli genome and is a hotspot for HR (Anderson and Kowalczykowski, 1997).

Once a Chi site is encountered, the activity of RecBCD is altered such that degradation of the 3′ end is reduced, generating a ssDNA overhang (Spies et al., 2007). This end is then a substrate for the E. coli homolog of Rad51, the RecA protein. As with Rad51, RecA binds to ssDNA, forming an extended nucleoprotein filament. Interestingly, RecA does not require a mediator function, such as BRCA2 or Rad52, to facilitate its loading onto 10

ssDNA. Indeed, the presence of SSB, the bacterial homolog of RPA, actually facilitates loading of RecA and formation of the recombinagenic filament (Kowalczykowski and

Krupp, 1987). The filament then performs a homology search, invading the intact homologous duplex to form the D-loop structure described above and forming two HJ structures. In contrast to mammalian cells, far more is known about the process of HJ resolution in bacteria. The junction structures are bound by the RuvAB complex of proteins that serves as a molecular motor, spooling the DNA through and migrating the

HJ (Amit et al., 2004). The RuvC endonuclease then binds the complex and cleaves two of the strands, resolving the duplexes. As with mammalian HJ resolution, the directionality of the strand cleavage determines if crossover or non-crossover products result (Shah et al., 1997). There is no known mammalian homolog of this HJ resolution system, therefore alternative helicases and structure-specific nuclease systems described above are utilized (Figure 1.1e,f).

While the general mechanism of HR repair in different species is similar, mammalian cells differ in that a larger number of accessory mediator proteins (such as the Rad51 paralogs) is required. It is likely that these additional factors allow HR to be utilized and regulated under specific conditions such as meiotic recombination or replication- associated DNA damage.

The Nuclear DNA Replication Fork

In mammalian cells, the faithful completion of DNA replication is essential for cell viability. Nuclear DNA replication initiates during S phase of the cell cycle and proceeds in a semi-discontinuous, strand-coupled manner. Synthesis of the leading strand is primed by the production of a short RNA primer by RNA polymerase. This is then extended by 10-20 deoxynucleotides in a 5′-3′ direction by DNA polymerase α (Pol α).

At this point, a polymerase switch occurs with Pol α being replaced by the main replicative polymerase of the leading strand, believed to be DNA polymerase ε (Pol ε)

(Pursell et al., 2007). This switch in replication machinery is thought to be mediated in part by the loading of the processivity factor, proliferating cell nuclear antigen (PCNA).

In the absence of disruptions on the template DNA, leading strand synthesis proceeds largely uninterrupted. Concurrently, DNA replication proceeds in the 5′-3′ direction on the lagging strand. Because of the anti-parallel arrangement of the DNA duplex, the lagging strand must be synthesized in a discontinuous manner, with each segment requiring a primer produced by the combined functions of RNA polymerase and Pol α, as with the leading strand. Once again, a polymerase switch occurs with Pol α being replaced by DNA polymerase δ (Pol δ) (Pursell et al., 2007). In mammalian cells, the lagging strand segments, called Okazaki fragments, are ~200 nucleotides in length and are joined together by DNA ligase in a maturation step that also involves removal of the

RNA component of the primer (Zheng and Shen, 2011).

Replication Restart and Repair of Collapsed Forks

Processivity of nuclear DNA replication depends on a number of factors. One such factor is the innate DNA binding affinity of DNA polymerases. The PCNA protein, mentioned above, is a processivity factor that functions as a sliding clamp during bulk

DNA replication. The replicative polymerases Pol δ and Pol ε interact with PCNA, greatly stabilizing their association with DNA (Langston and O'Donnell, 2008; Chilkova et al., 2007). Another important determinant in replication processivity is the quality of the template DNA. As previously described, DNA incurs a variety of different types of damage (e.g. strand breaks, single-strand gaps, base adducts, and abasic sites). When these DNA lesions are left unrepaired or are in the process of being repaired, they represent roadblocks to ongoing DNA replication that cause forks to pause or collapse to form a double strand break. In order for replication to resume, forks must either be stabilized long enough for the DNA lesion to be repaired or reformed following dissociation of core replication factors. In bacteria, reloading of the replication machinery is facilitated in part by the RecA recombinase and the PriA protein, a helicase with affinity for D-loop DNA structures. As mentioned above, Rad51 is the human homolog of RecA; however, no known homolog of PriA has been identified. Clear evidence for replication restart in mammalian cells has been provided in recent years using single molecule DNA combing to monitor progression of individual replication forks (Petermann et al., 2010). This technique has revealed that after release from long periods of replication inhibition, stalled mammalian replication forks largely become inactivated and new origins fire in order to complete replication in the region. The 13

inactivated forks are cleaved through the activity of the Mus81 structure-specific endonuclease to form double strand breaks that are then repaired by Rad51-mediated HR

(Hanada et al., 2007). This repair of collapsed forks involves the formation of Rad51 foci as determined by immunofluorescent microscopy, as well as an increase in the rate of gene conversion, an indication of HR activation (Petermann et al., 2010).

In contrast to extended fork stalling, following short periods of replication inhibition, most mammalian replication forks resume progression. This restart depends on expression of Rad51, but does not involve formation of Rad51 foci, suggesting that

Rad51-mediated replication restart is mechanistically distinct from classic DSB repair

(Petermann et al., 2010). To date, details of the restart mechanism remain elusive, and several pathways have been suggested. The first employs factors such as BLM and

Werner (WRN) syndrome helicases that may regress and stabilize the fork, allowing time for the preexisting DNA lesion to be repaired. These helicases then migrate this 4-way junction forward to affect replication restart (Karow et al., 2000; Ralf et al., 2006) (Figure

1.2a). A deficiency in either helicase results in reduced fork restart following short periods of replication inhibition (Rassool et al., 2003; Constantinou et al., 2000). In this model, Rad51 may stimulate annealing of the nascent DNA strands during fork regression and/or stabilize the strands during repair of the lesion. Models of fork remodeling involve the formation of a four-way DNA intermediate termed a “chicken foot,” which is structurally identical to a HJ. While the formation of chicken feet at stalled replication forks has been shown in bacteria (Postow et al., 2001), direct evidence 14

Figure 1.2: Models of mammalian replication fork restart (a) A stalled replication fork may result in unwinding of the template to genearate ssDNA that may be reannealed, or the fork may be regressed to form a 4-way “chicken foot” structure. (b) ssDNA at the regressed fork may be bound by Rad51 to form a recombinagenic end that invades the intact duplex. The 3’ invading end could prime DNA synthesis and lead to reestablishment of the fork. Two Holliday junctions are formed in the process and must be resolved. (c) Conversely, stalled fork DNA may be cleaved by a structure-specific nuclease such as Mus81 to yield a one-ended DSB. Rad51-mediated repair of this break could allow priming of DNA synthesis and fork restart. Adapted from Petermann et al. Nat Rev Mol Cell Biol. 2010 with permission.

in mammals has yet to be reported. A second potential mechanism of fork restart involves Rad51 binding the double-stranded end generated by fork regression. This end may then invade the template, forming a D-loop structure that could allow reconstitution of the replication machinery, as well as the formation of a HJ that must then be resolved as described above (Figure 1.2b). In a third mammalian replication restart model, cleavage of the stalled fork by Mus81 generates a single-ended DSB. Rad51-mediated

HR then repairs the break, with the invading 3′ DNA end priming the restart of replication (Figure 1.2c). This method of restart is analogous to break-induced replication (BIR), which has been observed in yeast (Voelkel-Meiman and Roeder, 1990) and likely involves the generation of a single HJ. The defining feature of each model of replication restart is the indispensable role of the Rad51 recombinase. Indeed, the requirement of HR for efficient replication restart in addition to the specific activation of the pathway during S phase has led to the suggestion that the main purpose of recombination is to support productive DNA replication.

The Mitochondrion

The cytosol of every mammalian cell contains a network of organelles called mitochondria. Mitochondria are responsible for the vast majority of cellular energy production in the form of adenosine triphosphate (ATP) through the aerobic metabolism of glucose (Figure 1.3). In addition to its well-characterized role as the cell’s “power plant,” the mitochondrion is responsible for a variety of cellular activities including Ca2+ signaling, heme and steroid synthesis, and regulating the induction of apoptosis. The 16

mitochondrion is composed of a double membrane system, which encloses the inner membrane space and matrix compartments. The outer membrane is similar in composition to the cell membrane and contains pores that allow small molecules (<5 kDa) to freely diffuse. In contrast, the inner membrane is highly invaginated into structures called cristae and has a relatively high protein to lipid ratio. In addition, the inner membrane contains high levels of cardiolipin. This phospholipid may serve to generate a tighter membrane barrier in order to prevent the leakage of ions and disruption of the proton gradient that is essential for ATP production. The inner membrane is embedded with more than 150 different proteins, including all of the membrane-bound components of the electron transport chain, and the ATP synthase complex responsible for ATP generation. Of the hundreds of proteins that localize to the various mitochondrial compartments, all but 13 are encoded in the nuclear genome. The transcripts of these nuclear genes are translated by cytosolic ribosomes and subsequently translocated to their targeted mitochondrial compartment. Movement of proteins into the mitochondrial membranes and the compartments they enclose is mediated by the translocase complexes of the outer (TOM) and inner (TIM) membranes. These complexes recognize signal sequences in the proteins they transport to ensure proper compartment localization and topological orientation of integral membrane proteins. The targeting of many mitochondrial proteins is directed by a signal sequence of hydrophobic and acidic amino acid residues (Pfanner, 2000). N-terminal targeting signals frequently target proteins to the mitochondrial matrix, where the signal sequence is subsequently cleaved off, while internal targeting signals often direct the imbedding of proteins into the 17

Figure 1.3: Electron transport chain of the mitochondria The enzyme complexes of the electron transport chain are embedded into the inner membrane of the mitochondria. Complexes I and II accept electrons from donors NADH and succinate respectively, transporting 4 protons into the intermembrane space in the process. Electrons from both complexes are donated to coenzyme Q (labeled Q), which transferes them to complex III. Cytochrome C (labeled C) accepts electrons from complex III, resulting in the transport of 4 protons. Finally, complex IV reduces molecular oxygen to water, transporting 2 protons in the process. Flow of protons down the electrochemical gradient from the intermembrane space into the matrix is coupled to ATP synthesis through the ATP synthase complex (complex V). Premature leakage of protons leads to the generation of superoxide, which is converted to a variety of reactive intermediates (ROS).

inner or outer membrane (Dukanovic and Rapaport, 2011). There are, however, numerous examples of mitochondrial protein localization without any discernable targeting signal. Studies of the molecular details of the full range of mitochondrial protein import machinery and the peptides they transport are an active area of research (Schmidt et al., 2010).

mtDNA

Housed within the matrix of the mitochondria is the 16.568 kb human mitochondrial genome (Figure 1.4). In mammals, mtDNA is inherited exclusively from the mother, as the sperm mtDNA is selectively degraded soon after fertilization (Sutovsky et al., 2000).

Encoded in this genome are 13 polypeptides that are essential components of the electron transport chain, responsible for supplying the majority of the cell’s energy (Figure 1.3).

These genes are transcribed and translated within the mitochondrial matrix. The 22 tRNAs and 2 rRNAs that are also encoded in the mtDNA are required for expression of mtDNA genes, in addition to a number of nuclear-encoded factors. Unlike its nuclear counterpart, the mitochondrial genome is circular and present in many copies per cell. In the human mitochondria, 8-10 mtDNA molecules are packaged with a number of proteins in a structure referred to as a nucleoid. These nucleoids are found throughout the mitochondrial network, and localize in close proximity to the inner surface of the inner membrane (Figure 1.5) (Wang and Bogenhagen, 2006; Albring et al., 1977).

Figure 1.4: The human mitochondrial genome The circular human mitochondrial genome is approximately 16.5 Kb in length and encodes 13 polypeptides, including 7 subunits of NADH dehydrogenase (complex I), 1 subunit of cytochrome c reductase (complex III), and two subunits of ATP synthase. Twenty-two transfer RNAs, and 2 ribosomal RNAs, that are required for mtDNA transcription and translation are also encoded. Adapted from Wanrooij et al. Biochem Biophys Act. 2010 with permission.

mtDNA Replication

Replication of mtDNA is entirely dependent on nuclear-encoded genes that are synthesized in the cytoplasm and transported into the mitochondrial matrix. Relative to its nuclear counterpart, mtDNA replication appears to involve a much smaller number of factors. The disruption of any of these core replication components leads to severe mtDNA depletion in cells and a variety of neuro-muscular disorders in animals (Hudson and Chinnery, 2006).

In vitro, mtDNA replication can be reconstituted with only five purified proteins:

TWINKLE, mtSSB, POLRMT, Pol γα, and Pol γβ (Korhonen et al., 2004). The T7 gp4- like protein with intramitochondrial nucleoid localization (TWINKLE) is the 5′-3′ helicase responsible for unwinding double-stranded DNA (dsDNA) at the replication fork. As its name indicates, TWINKLE shares sequence similarity to the T7 helicase- primase protein, although no primase activity has been detected. In addition to its critical role in replication, TWINKLE may also play a role in HJ migration in mtDNA.

Mitochondrial single-stranded DNA binding protein (mtSSB), like its nuclear counterpart

RPA, binds and stabilizes ssDNA to prevent the formation of secondary structure that would inhibit the binding or activity of other replication factors. MtSSB also has been shown to stimulate the processivity of DNA polymerase γ (Pol γ) and the unwinding activity of TWINKLE, functions that cannot be replicated by substitution with the E. coli homolog, SSB (Korhonen et al., 2003). This suggests that in addition to preventing 21

Figure 1.5: Bromodeoxyuridine-labeled mtDNA in human cell The osteosarcoma cell line U2OS has been labeled with the nuclear stain DAPI (blue) and the mitochondrion-specific dye MitoTracker (red). Dispersed throughout the mitochondrial network are punctate structures of nascent mtDNA labeled with bromodeoxyuridine (BrdU, green). Nuclear DNA replication has been inhibited with aphidicolin treatment in order to prevent incorporation of BrdU into chromosomal DNA.

ssDNA folding, mtSSB may interact with other mitochondrial replication factors to support efficient DNA synthesis. The mitochondrial RNA polymerase (POLRMT) is required for the priming of DNA synthesis, in addition to its central role in mitochondrial gene transcription. Pol γ is the only known polymerase in the mitochondria and is therefore believed to be responsible for all mtDNA metabolic reactions requiring DNA synthesis. The Pol γ holoenzyme is composed of three subunits. The α subunit forms the catalytic core of the enzyme and also contains the 3′-5′ proofreading exonuclease activity. Two identical β subunits interact with the α subunit at distinct binding sites with one site stimulating holoenzyme processivity and the other increasing the rate of polymerization (Lee et al., 2010; Carrodeguas et al., 2002). Unlike DNA polymerases in the nucleus, Pol γ is able to bind, and in some cases, incorporate a number of nucleotide analogs such as azidothymidine (AZT) and dideoxycytidine (ddC). Treatment of cells with these types of analogs has been shown to inhibit mtDNA replication through either competition with natural nucleotides for binding to Pol γ, or chain termination following incorporation into nascent DNA (Martin et al., 1994). Many of these analogs have been shown to potently inhibit HIV replication, and long-term treatment with nucleoside analog reverse transcriptase inhibitors has been linked to a variety of side effects associated with mitochondrial dysfunction (Chen et al., 1991; Dalakas et al., 1994).

While the majority of critical factors for efficient mtDNA replication may have been identified, the exact mode of mtDNA replication is an area of active research with evidence for several potential models that have yet to be reconciled. Based on buoyant 23

density differences, the two strands of mtDNA have been termed “heavy” and “light.”

According to the “orthodox” model of mtDNA replication described by Clayton and colleagues, synthesis of the two strands proceeds asynchronously from two origins termed the origin of heave strand repication (OH) and the origin of light strand replication (OL) (Figure 1.6a) (Berk and Clayton, 1974). Heavy strand replication is initiated at OH by synthesis of a ~50 bp segment of RNA by POLRMT from the promoter for light strand transcription. This abortive transcript then serves as a primer for Pol γ-mediated DNA synthesis that proceeds continuously around the circular genome. This mode of replication generates a significant amount of ssDNA through template unwinding that is bound and stabilized by mtSSB. Once ~60% of the leading strand has been synthesized, the passing fork unwinds and reveals the origin of light strand replication, OL. When single-stranded, OL forms a stem-loop structure that recruits POLRMT, which then synthesizes the RNA primer required for light strand synthesis. Once again, replication of the light strand proceeds in a continuous manner.

The net result is the complete duplication of the genome, with light strand synthesis concluding in a delayed fashion. A characteristic of this model for DNA replication is the absence of lagging strand Okazaki fragments.

In 2000, Holt and coworkers presented evidence for an alternate mode of mtDNA replication based on the analysis of replication intermediates by 2-dimensional agarose gel electrophoresis (2D-AGE). They reported the presence of mtDNA replication intermediates from mouse and human cells that are characteristic of fully double-stranded 24

Figure 1.6: Replication modes of mammalian mtDNA. (a) In the strand- asynchronous model of mtDNA replication, heavy strand synthesis initiates from its origin, OH, within the non-coding control region and proceeds continuously, displacing the light strand. Once ~60% of the genome has been replicated, the origin of light strand replication (OL) is exposed as a single stranded region (left pathway). Alternatively, an alternate light strand replication may initiate at an undefined region

(OLalt) (right pathway). Secondary structure in this region recruits POLRMT, which primes light strand replication that then proceeds continuously, completing duplication of the genome. (b) RNA incorporated throughout the lagging strand (RITOLS, left pathway) initiates strand-coupled, unidirectional replication at a discreet origin (OR). Leading strand DNA synthesis of the heavy strand occurs continuously, while lagging stand synthesis is discontinuous and largely made up of RNA. Conventional, strand-coupled DNA synthesis (right pathway) initiates from a broad region downstream of OH. Bidirectional DNA replication occurs with continuous DNA replication of the heavy strand and through Okazaki fragment generation of the light strand. Adapted from Wanrooij et al. Biochem Biophys Act. 2010 with permission.

replication forks, as well as large regions of ssDNA indicative of the “orthodox” replication model (Holt et al., 2000). Importantly, they observed extensive incorporation of RNA, primarily in the lagging (light) strand. This conclusion was based mainly on the observation that a subset of replication intermediates was insensitive to the single-strand nuclease S1, but was sensitive to RNase H, which digests RNA in an RNA-DNA duplex.

Interestingly, this coupled leading and lagging strand replication appears to proceed unidirectionally from the OH region of the genome (Figure 1.6b, left pathway). This model was later refined to include RNA incorporation throughout much of the lagging strand, suggesting that mtDNA replication may involve a maturation stage where RNA is removed and replaced with DNA, presumably by Pol γ (Figure 1.6b, right pathway)

(Yang et al., 2002; Yasukawa et al., 2006). Holt and colleagues concluded that this RNA incorporated throughout the lagging strand (RITOLS) model is the main mode of mtDNA replication, and that evidence of large stretches of ssDNA associated with the “orthodox”

(strand-asynchronous) model was due to RNase contamination during isolation. In addition, a difference in the degree of RNA incorporation was reported between normally cultured cells and those recovering from mtDNA depletion (Yasukawa et al., 2005). This finding might indicate that multiple modes of replication may be employed in a given cell. Shifting modes of mtDNA replication could be a novel strategy for altering rates of mtDNA replication initiation, and ultimately copy number in response to changes in growth conditions. Clarification of these models of mtDNA replication awaits standardization of the sources and preparation methods of mtDNA replication intermediates. Nevertheless, regulation of mtDNA replication is critical for ensuring that 26

genome copy number is stable and sufficient to provide for the ATP requirements of the cell.

Yeast mtDNA Replication mtDNA replication in yeast appears to differ greatly from that of mammalian cells described above. In contrast to the circular human mitochondrial genome, mtDNA isolated from budding and fission yeast appears to be largely linear or branched in arrangement and varying size in addition to a small population of circular genomes

(Bendich, 1996). The purpose of this kind of arrangement is not clearly understood. One model of S. cerevisiae mtDNA replication, proposed by Hori et al., utilizes the small population of circular genomes as a template for rolling circle replication (Hori et al.,

2009). Their model involves the binding of a homologous pairing protein, Mhr1, to the end of one of the linear mtDNA molecules. The Mhr1 protein has biochemical properties similar to E. coli RecA (Masuda et al., 2010) and forms a recombinogenic filament capable of invading a circular genome. The invading 3′ DNA end then primes DNA replication that proceeds around the circular template, spooling off nascent genomes.

While the details of this mechanism remain to be clarified, the biochemical properties of

Mhr1 and its importance in yeast mtDNA maintenance strongly suggest that recombination-dependent replication may be an important pathway for mtDNA stability in budding yeast.

Maintenance of Mammalian mtDNA Copy Number

MtDNA copy number is cell-type specific and dependent on the energy needs under a given set of growth conditions, ranging from several hundred to many thousands of genomes (Moraes, 2001). In humans, pathogenic alterations in mtDNA copy number have been linked to a variety of disorders including infertility, tumor growth, as well as a variety of neuromuscular disorders including Alpers’ disease and mtDNA Depletion

Syndrome (Wai et al., 2010; Naviaux and Nguyen, 2004). While the exact mechanisms involved in regulating mtDNA copy number are not known, there is evidence that it is the mass of mtDNA that is regulated, and not the copy number per se. Evidence for regulation at this level was provided through elegant experiments by Tang et al. in which mtDNA harboring partial deletions or partial duplications was transferred into otherwise isogenic cells that had been depleted of endogenous wild type mtDNA (Tang et al.,

2000). Analysis of these mammalian cell lines showed that while copy number varied between wild type, partial deletion, and partial duplication cells, the amount of mtDNA by mass per cell was nearly identical. Furthermore, the duplicated region of mtDNA included both heavy and light strand replication origins, suggesting that replication initiation sites do not themselves limit copy number. These results mirror what has been shown in budding yeast, where a variety of rearranged mtDNA genomes are maintained in the cell at a constant mass rather than constant gene or genome copy number (Nagley and Linnane, 1972).

Regulation of mtDNA copy number and/or mass is not well understood in mammalian cells; however, there is almost certainly an important link between mtDNA replication and copy number. Several groups have shown that in mammalian cells, the majority of replication initiation events from OH are extended for several hundred nucleotides before being terminated in a region termed the termination associated sequence (TAS). This small stretch of DNA (called 7S DNA) remains hybridized to the circular genome, forming a triple-stranded D-loop structure. In 2002, Clayton and coworkers reported that recovery of mtDNA copy number following depletion is associated with a decrease in termination at TAS, and therefore more frequent productive replication (either through extension of preexisting 7S DNA, or through less frequent termination at TAS) (Brown and Clayton, 2002). In addition to the release of replication termination controls, the alteration of mtDNA replication factors described above may be involved in modulating the frequency of mtDNA replication and therefore copy number. Indeed, mutations in

Pol γ (Naviaux and Nguyen, 2004), the replicative helicase TWINKLE (Sarzi et al.,

2007), and mtSSB are associated with mtDNA depletion syndrome (MDS) in humans

(Szczesny et al., 2008). A subset of patients diagnosed with MDS were found to have mutated alleles of either deoxyguanosine kinase (dGK) (Mandel et al., 2001) or thymidine kinase (TK) (Saada et al., 2001). Both of these proteins contribute to the mitochondrial dNTP levels through nucleotide salvage pathways. In addition, the transcription factor A, mitochondrial protein (TFAM), originally characterized by its role in mitochondrial transcription regulation (Fisher and Clayton, 1988), also plays a critical role in regulating mtDNA replication. It has since been shown that TFAM binds 29

extensively (and rather nonspecifically) throughout the mitochondrial genome. There is evidence that TFAM is only stable when bound to mtDNA and therefore may serve to titrate mtDNA levels in the cell based on binding capacity (Larsson et al., 1994).

While the mechanisms that directly regulate the frequency of mtDNA replication remain to be determined, several factors have been reported to influence changes in copy number. One major signal found to increase mtDNA content in human cells is oxidative stress. Treatment of cells directly with H2O2 or with drugs (such as buthionine sulphoximine) to deplete their antioxidant capacity has been shown to increase both mitochondrial mass and mtDNA copy number (Lee et al., 2000; Wei et al., 2001).

Interestingly, acute stress-induced increases in mtDNA copy number appear to be dependent on the synthesis and import of nuclear-encoded proteins (Lee et al., 2002). In addition, aging-associated increases in oxidative stress have been correlated with increases in mtDNA copy number in human lung tissue (Lee et al., 1998). This aging- related phenomenon can be mimicked in cell culture through extensive serial subcultivation of cells, resulting in increased ROS production and mtDNA copy number at high passage number (Lee et al., 2002). The mechanism by which ROS induces increases in mitochondrial biogenesis and mtDNA copy number remains to be shown.

One prominent theory is that production of ROS over time results in increased mtDNA mutations. The general phenotype of these mutations is an impairment of the electron transport chain, which leads to a decrease in the efficiency of aerobic respiration in addition to an increase in ROS production. To compensate for the lack of ATP 30

production, more mitochondria are produced by the cell. This in turn leads to even greater production of ROS and extensive damage to proteins, lipids, and nucleic acids.

This vicious cycle of ROS production and mtDNA damage may be an underlying cause of aging-associated disease (Van Houten et al., 2006; Yakes and Van Houten, 1997).

mtDNA Repair and Degradation

Although DNA repair in mitochondria was once thought to be minimal, recent years have seen an ever-expanding appreciation for the repair of a variety of types of mtDNA damage. One of the first mtDNA repair proteins identified in the mitochondria is the mismatch repair protein MSH1 in yeast (Reenan and Kolodner, 1992b). Disruption of

MSH1 is associated with increases in mutation rate, as well as gross mtDNA rearrangements (Reenan and Kolodner, 1992a). In mammalian cells, a similar protein,

YB-1 has been proposed to fulfill a similar role in preventing mtDNA mutations as a result of replication errors or non-homologous recombination events (Mason et al., 2003).

Given its close proximity to the major site of ROS production in the cell, mtDNA is routinely subjected to high levels of oxidative stress, causing a variety of types of DNA damage that can potentially lead to mutations. Studies using quantitative long range PCR to amplify large regions of mtDNA have indicated that low levels of oxidative stress preferentially damage mtDNA (Santos et al., 2002; Ayala-Torres et al., 2000). In this assay, mtDNA damage is detected as lesions in the DNA template that block polymerase 31

progression. By extension, these studies also indicate that ROS-induced mtDNA damage represents potential blocks to replication fork progression.

One of the most critical mtDNA repair pathways is base excision repair (BER). This system utilizes a series of glycosylases to detect and remove damaged bases followed by endonuclease cleavage of the resulting abasic site to leave a small single-stranded gap.

This gap is then filled by Pol γ and the nicks resealed to restore the intact duplex. So- called short-patch BER is able to remove small lesions such as modified bases. Large modifications such as bulky alkylation adducts to bases or protein-DNA crosslinks require removal of a larger segment of the damaged strand to affect repair through a process called long-patch BER (LP-BER). Recently, several groups have reported long- patch BER activity in mitochondrial extracts from both human cells and mouse tissue

(Liu et al., 2008; Akbari et al., 2008; Szczesny et al., 2008). This process of removing a segment of DNA requires flap endonucleases such as Fen1 and Dna2, both of which localize to the mitochondria as well as the nucleus (Zheng et al., 2008; Kalifa et al.,

2009). Interestingly, these two proteins have been shown previously to remove RNA primers in nuclear DNA during Okazaki fragment maturation (Bae et al., 2001). It is possible that they also serve a similar function in the mitochondria, lending support to the strand-coupled mode of leading and lagging strand mtDNA replication described by Holt and coworkers (Yang et al., 2002).

As mentioned above, the faithful repair of DNA double strand breaks in the nucleus is critical for viability and genetic stability. In 1999, DNA end rejoining was demonstrated for the first time in mammalian mitochondrial extracts (Coffey et al., 1999). It has been shown since that a C-terminally truncated form of the nuclear NHEJ protein Ku80 localizes to the mammalian mitochondria and is able to bind linearized plasmid DNA ends like its nuclear counterpart (Coffey and Campbell, 2000). In similar extracts, the repair of both blunt and cohesive ends has also been demonstrated (Coffey et al., 1999), suggesting that mammalian mitochondria possess a NHEJ-like repair pathway.

The first biochemical evidence of HR in mammalian mitochondria was the demonstration of a strand exchange activity in mitochondrial lysates (Thyagarajan et al., 1996). In this study, mitochondrial extracts from either rat liver tissue or cultured human fibroblasts stimulated conservative recombination events between plasmid substrates, resulting in the reconstitution of a selectable marker when transformed into recombination-deficient E. coli. In other reports, physical evidence of mtDNA recombination has been observed in the form of 4-way structures in mtDNA isolated from human muscle tissue analyzed by

2D-AGE (Kajander et al., 2001). These structures were shown to be fully duplex, resistant to topoisomerase I treatment, and could be resolved by treatment with the bacterial RuvC HJ resolvase. These characteristics suggest that the structures observed are not simply topologically linked concatenated genomes, but in fact bona fide recombination intermediates.

Genetic evidence of recombinant mitochondrial genomes was provided in a study where two cell lines with genetically distinct mitochondrial genomes were fused, mixing their cytosolic contents. The majority of the resulting hybrid clones were found to contain intermolecular recombinant genomes (D'Aurelio et al., 2004). Bacman and coworkers observed similar intermolecular recombination events in mice harboring multiple mtDNA genotypes, and found the frequency of recombination-mediated deletions was increased by the generation of DSBs (Bacman et al., 2009). In contrast, however, a study by

Gilkerson et al. utilized a similar cell fusion system to compliment two cell lines, each harboring different mtDNA deletions (Gilkerson et al., 2008). The resulting hybrid cells maintained both mtDNA genotypes without evidence of recombination between them.

These differing results have yet to be reconciled.

Evidence for the existence of mitochondrial HR in humans came in 2004 with the analysis of mtDNA isolated from a rare case of biparental mtDNA inheritance.

Recombinant mtDNA molecules were identified, bearing sequence from both parental genomes (Kraytsberg et al., 2004). These genetic analyses confirmed what an abundance of biochemical data had strongly suggested: human mitochondria possess and utilize the capacity to carry out HR.

Failure to repair damaged DNA results in either incorporation of misreplicated sequence as a mutation, or the degradation of part or all of the damaged genome. Indirect evidence of mtDNA degradation has been reported by Mita et al., showing that treatment of HeLa 34

cells with a variety of carcinogens resulted in few mutations in mtDNA, indicating that damaged genomes are not frequently replicated (Mita et al., 1988). Recently, the resistance of mtDNA to mutations following prolonged exposure to oxidative stress has been reported (Shokolenko et al., 2009). This report also directly showed the breakage and degradation of mtDNA following acute oxidative stress. Shortly after stress, these breaks are predominantly single-stranded, but are then converted to DSBs several hours later. It is possible that an overabundance of single-strand lesions overwhelm the BER capacity and lead to perpetually damaged genomes that are acted upon by nucleases.

Additionally, ongoing replication of the damaged mtDNA likely leads to the collapse of replication forks and the generation of DSBs as described above. More work is needed to identify the nucleases responsible for mtDNA degradation as well as their role in maintaining mitochondrial genomic stability.

Remaining Questions

It is clear from studies over the last several decades that mitochondria possess robust mechanisms for dealing with a variety of mtDNA damage. These systems are critical for production of sufficient ATP for a range of cellular functions, but also for ensuring that the mitochondrial genome does not acquire mutations that render aerobic metabolism inefficient and prone to ROS production. While it is known that recombination plays an active role in the repair and replication of yeast mtDNA, the importance of this activity in mammalian cells has yet to be shown. As stated above, strand exchange activity has been shown in mitochondrial extracts from human cells (Thyagarajan et al., 1996); however, 35

the proteins responsible have yet to be identified. Decades of research have highlighted two critical roles for Rad51-mediated recombination: the restart of stalled replication forks, and the repair of DNA DSBs. Given the high level of exposure of mtDNA to oxidative stress, dealing with DNA lesions that block replication is likely to be a frequent occurrence in mammalian mitochondria.

Acceptance that human mtDNA does recombine has been slow to take hold, largely as a result of the rarity of genetic evidence of gene conversion in the population. In addition, studies of recombination between genomes observed under rare or artificially-induced conditions where differing mtDNA genotypes are present in the same cell is conflicting

(Kraytsberg et al., 2004; D'Aurelio et al., 2004; Gilkerson et al., 2008). Given the critical role of HR in facilitating replication restart and repairing damaged DNA in the nucleus, it is likely that the recombination activity observed in mammalian mitochondria serves to support mtDNA replication and repair, rather than to generate genetic diversity.

Therefore, it is critical to evaluate the role of HR in the processes of replication and repair and the proteins involved in order to truly understand its importance in human mtDNA maintenance. Mammalian cells lack the homolog for the yeast mitochondrial recombinase Mhr1. The Rad51 protein, as the central catalyst of HR in the nucleus, is therefore a likely candidate for the strand exchange activity observed in mammalian mitochondrial extracts (Thyagarajan et al., 1996).

Given the biochemical and genetic evidence for mtDNA recombination desctibed above, we investigated the mitochondrial localization and function of the major mammalian recombinase, Rad51. The experiments presented in this thesis were designed to explore the role of Rad51 in the maintenance of human mtDNA copy number. Using a combination of biochemical and cell based approaches, we have evaluated the mitochondrial localization of Rad51 in cultured human cells under normal conditions and following DNA-damaging stress. The physical interaction between Rad51 and mtDNA has been assessed as well as the role of HR in the maintenance of mtDNA copy number under oxidative stress conditions. We further investigated the contribution of mtDNA replication in the oxidative stress response and the important role of Rad51 in dealing with replication stress. An understanding of how Rad51-mediated recombination contributes to mitochondrial genome stability is essential in order to better appreciate how mutations in HR proteins may lead to mitochondrial-associated diseases.

CHAPTER II

Discovery of a Novel Function for Human Rad51:

Maintenance of the Mitochondrial Genome

Abstract

Homologous recombination (HR) plays a critical role in facilitating replication fork progression when the polymerase complex encounters a blocking DNA lesion, and also serves as the primary mechanism for error-free repair of DNA double strand breaks. Rad51 is the central catalyst of HR in all eukaryotes and to this point, studies of human Rad51 have focused exclusively on events occurring within the nucleus. However, substantial amounts of

HR proteins exist in the cytoplasm, yet the function of these protein pools has not been addressed. The results presented here provide the first demonstration that Rad51 and the related HR proteins Rad51C and Xrcc3 exist in human mitochondria. We show stress- induced increases in both the mitochondrial levels of each protein and, importantly, the physical interaction between Rad51 and mtDNA. Depletion of Rad51, Rad51C, or Xrcc3 results in a dramatic decrease in mtDNA copy number as well as the complete suppression of a characteristic oxidative stress-induced copy number increase. These results identify human mtDNA as a novel Rad51 substrate and reveal an important role for HR proteins in the maintenance of the human mitochondrial genome.

Introduction

In mammalian cells, mitochondria contain 2-10 copies of a 16.5 kbp genome that encodes rRNAs and tRNAs involved in mitochondrial translation, as well as several proteins that contribute to ATP generation via the oxidative phosphorylation pathway (Graziewicz et al.,

2006). mtDNA is located within the mitochondrial matrix in close proximity to the inner membrane, and therefore suffers significant damage resulting from the production of ROS via the oxidative phosphorylation pathway (Stuart and Brown, 2006). Currently, BER is the only mitochondrial DNA repair pathway for which all required enzymes have been identified in mammalian cells (Graziewicz et al., 2006; Stuart and Brown, 2006).

The occurrence of genetic recombination in the mitochondrial genomes of fungi and plants has been clearly established (Barr et al., 2005; Shedge et al., 2007), but evidence of mtDNA recombination in animal cells has long been elusive as inheritance is almost exclusively maternal, as recombination between identical mtDNA genomes would be difficult to detect

(Barr et al., 2005). An earlier study suggested a DNA recombination activity in mammalian mitochondrial extracts (Thyagarajan et al., 1996), and several studies indicated the rare occurrence of genetic recombination events in human mitochondria (Kajander et al., 2001;

D'Aurelio et al., 2004; Kraytsberg et al., 2004). However, it is not clear whether these activities contribute in any way to mitochondrial genome integrity, nor have the proteins responsible been identified. The significant level of cytoplasmic Rad51 observed in numerous studies (Davies et al., 2001; Essers et al., 2002; Kraakman-van der Zwet et al.,

2002; Tarsounas et al., 2003; Liu and Lim, 2005; Bennett and Knight, 2005; Mladenov et al., 39

2006; Henson et al., 2006) has been shown recently to contribute to a DNA damage-induced increase in nuclear Rad51 levels (Gildemeister et al., 2009), but given previous suggestions of a recombination activity in vertebrate mitochondria, the question remains as to whether some amount of cytoplasmic Rad51 may be present and function in mitochondria.

In this study, we report for the first time the presence of the Rad51 recombinase as well as the

Rad51C and Xrcc3 paralog proteins in mitochondria of cultured human cells. We observe a

DNA damage-induced increase in mitochondrial levels of each protein as well as an oxidative stress-induced increase in the physical association of Rad51 with mtDNA. Strikingly, depletion of Rad51, Rad51C, or Xrcc3 from cells results in a significant decrease in mtDNA copy number following oxidative stress, supporting a novel role for Rad51-mediated recombination processes in the maintenance of mtDNA genome stability.

Experimental Procedures

Cell Culture, Exposure of Cells to DNA Damage and siRNA Knockdowns - HCT116 cells

(ACT #CCL-247) were grown at 37 °C 5% CO2 in McCoy’s 5A medium (Invitrogen) with, and HeLa (ATCC #CCL-2) and U20S (ATCC #HTB-96) cells in DMEM (Invitrogen), both supplemented with 10% fetal bovine serum and antibiotics. Cells were treated with IR using a Gammacell 40 137Cs irradiator (Atomic Energy of Canada, Ottawa, Ontario). For oxidative stress, glucose oxidase (GO; Sigma) stock solutions were prepared fresh prior to each experiment and added at the indicated concentrations. Cells were washed once with non- supplemented medium that was then replaced with non-supplemented medium containing the 40

indicated concentration of GO. The resulting concentrations of hydrogen peroxide produced were determined using a Hydrogen Peroxide Assay Kit (National Diagnostics), which is based upon formation of a complex between Xylenol Orange and ferric iron, which is produced by the peroxide dependent oxidation of ferrous iron. Protein levels were depleted by treating cells with pools of siRNAs targeting either Rad51, Rad51C, Xrcc3 or CDK9 (10 nM final, Dharmacon L-003530-00, M-010534-01, M-012067-00, or M-003243-03, respectively) delivered using Dharmafect1. A pool of non-silencing siRNAs was used as an additional negative control.

Subcellular Fractionation, Isolation of Mitochondria, and Proteinase K Protection Assay -

Cells at 75-85% confluency were harvested and resuspended in mitochondria isolation buffer

[10 mM Tris-HCl pH 7.8, 0.2 mM EDTA, 250 mM sucrose, 150 mM KCl, 0.15% digitonin

(Sigma), protease inhibitors (Roche)], incubated on ice for 20 min then centrifuged at 1,000 x g at 4 °C for 10 min. The supernatant was removed to a fresh tube and centrifuged at 12,000 x g at 4 °C for 15 min to pellet a crude mitochondrial fraction. The new supernatant was removed and saved as the cytosolic fraction. Pellets containing mitochondria were combined and washed multiple times with isolation buffer containing 1 M KCl to yield the enriched mitochondrial fraction. For the Proteinase K protection assay, equal aliquots of the resuspended mitochondrial fraction were treated with 0.8 mg/mL Proteinase K (Qiagen), in the presence or absence of digitonin (0.2 mg/mL), or SDS (1%) for 20 min at room temperature. Proteinase K activity was halted with the addition of two volumes of 20% trichloroacetic acid and incubation on ice for 20 min. Precipitated protein pellets were 41

washed once with ice-cold acetone, resuspended in 2 x Laemmli sample buffer and evaluated by Western blots as described below.

Immunoblots - Total protein in each mitochondrial fraction was determined using the BCA

Protein Assay Kit (Pierce). Samples were prepared by adding Laemmli sample buffer

(Sigma) to 1x and heating to 95 °C for 5 min. These samples were run on 4-12% Bis-Tris acrylamide gels (Invitrogen) and transferred to PVDF membranes using a semidry transfer system (Bio-Rad). Membranes were incubated with blocking buffer (10 mM Tris-HCl pH

8.0, 300 mM NaCl, 0.25% Tween 20, 15% nonfat dry milk) and then with blocking buffer containing 2% nonfat dry milk and primary antibodies [mouse anti-Rad51 (clone 14B4), mouse anti-Rad51C (clone 2H11/6), mouse anti-Xrcc3 (clone 10F1/6) (Novus Biologicals), mouse anti-ATP synthase (BD Biosciences), goat anti-lamin A/C (Santa Cruz

Biotechnologies), mouse anti-glyceraldehyde-3-phosphate dehydrogenase (Millipore), rabbit anti-TFAM (Abcam), rabbit anti-PCNA (Abcam), and rabbit anti-OPA1 (Abcam)]. Blots were then washed with blocking buffer (without milk) followed by incubation with horseradish peroxidase-conjugated secondary antibody [goat anti-mouse (Millipore) or rabbit anti-goat (Jackson Labs)]. Following additional washing, blots were incubated with chemiluminescent visualizer (Denville) and developed using an LAS-4000 imaging instrument (Fuji).

DNA Isolation and Determination of mtDNA Copy Number - Total cellular DNA was isolated using the QIAamp DNA Mini Kit (Qiagen), and DNA concentrations were determined 42

spectrophotometrically. Equal amounts of total DNA were assayed by quantitative PCR

(MJR Research) using QuantiFast SYBR Green mix (Qiagen) with primers designed to amplify a 100 bp segment of the mtDNA genome. Amplification of a 100 bp segment of the

18S ribosomal RNA gene was used as a normalization factor for the determination of changes in mtDNA copy number. Primer pair sequences for qPCR (quantitative PCR) experiments are as follows: 18S rRNA gene (5′-AGCCATGCATGTCTAAGTACGCACG-

3′, 5′-CAAGTAGGAGAGGAGCGAGCGACCA-3′) and mtDNA (5′-

CAGGAGTAGGAGAGAGGGAGGTAAG-3′, 5′-TACCCATCATAATCGGAGG

CTTTGG-3′)

mtDNA Immunoprecipitation - Cells were crosslinked by adding formaldehyde to 1% and incubating at 37 °C for 15 min. The reaction was terminated by adding glycine to 125 mM and incubating at 37 °C for a further 15 min. Cells were harvested by scraping and pelleted at 1,000 x g for 5 min. Pellets were resuspended in lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl pH 8.1, protease inhibitors) and incubated on ice for 20 min. Following another centrifugation at 1,000 x g for 10 min, the soluble fraction was removed to a fresh tube and protein concentrations were determined as described above. For each IP reaction,

100 µg of chromatin was diluted 10-fold with dilution buffer (0.01% SDS, 1.1% Triton X-

100, 1.2 mM EDTA, 16.7 mM Tris-HCl pH 8.1, 167 mM NaCl) and precleared with 10 µL of Protein G Dynabeads (Invitrogen) for 1 h at 4 °C. One µg of anti-Rad51, anti-TFAM, anti-PCNA, or anti-mouse IgG (mock) antibody was added and tubes were incubated overnight at 4 °C with rocking. The following day, immune complexes were precipitated by 43

adding 10 µL of Protein G Dynabeads with rocking at 4 °C for 1 h. Beads were washed once with buffer 1 (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 150 mM

NaCl), once with buffer 2 (0.1% SDS, 1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl pH 8.1, 500 mM NaCl), once with buffer 3 (250 mM LiCl, 1% NP40, 1% sodium deoxycholate, 1 mM EDTA, 10 mM Tris-HCl pH 8.1), and finally twice with TE (10 mM

Tris-HCl pH 7.5, 1 mM EDTA). Crosslinking was reversed by adding TE buffer + 1% SDS and rocking overnight at 65 °C. DNA was purified and analyzed by qPCR as described above. Fold enrichment of mtDNA signal was determined relative to mock IP and normalized to the input signal.

Results and Discussion

High levels of cytosolic Rad51 have been observed in previous studies from our group

(Forget et al., 2004; Bennett and Knight, 2005; Gildemeister et al., 2009) and others (Davies et al., 2001; Kraakman-van der Zwet et al., 2002; Essers et al., 2002; Tarsounas et al., 2003;

Liu and Lim, 2005; Mladenov et al., 2006; Henson et al., 2006). Therefore, we explored the possibility that part of the cytoplasmic Rad51 pool may be present in mitochondria. We developed a stringent fractionation protocol to ensure that mitochondria are free of contaminating nuclear and cytosolic Rad51 (Figure 2.1a). The resulting mitochondrial fraction shows clear separation of the mitochondrial markers ATP synthase and TFAM from cytosolic (GAPDH) and nuclear markers (lamin and PCNA) (Figure 2.1b, lane Mito). After verification by Western blot that detectable nuclear and cytosolic contaminants had been removed, Western blot analyses revealed the presence of Rad51, Rad51C and Xrcc3 in 44

Figure 2.1: Purification of mitochondrial proteins from cultured human cells. (a) Following either IR or GO treatment, cells were subjected to a gentle lysis procedure followed by a series of high salt washes and centrifugation steps (see Experimental Procedures for complete description). (b) Western blot analysis shows separation of mitochondrial proteins from cytosolic and nuclear contaminants. Markers are ATP synthase and TFAM (mitochondria), GAPDH (cytosolic), and lamin A/C and PCNA (nuclear).

mitochondria isolated from normal cycling cells that had not been exposed to exogenous

DNA damage (Figure 2.2a,b). To investigate the mitochondrial localization of Rad51 in greater depth, isolated mitochondria were subjected to Proteinase K treatment in the presence or absence of digitonin to remove the outer mitochondrial membrane, or SDS to solublize both membranes. As shown in Figure 2.2a, the inner membrane space protein OPA1 is sensitive to Proteinase K digestion when mitochondria are treated with digitonin, and while a portion of Rad51 (30%) was sensitive to this treatment, the majority (70%) was resistant to

Proteinase K treatment unless treated with SDS. This indicates that mitochondrial Rad51 is not present at the external surface of the mitochondria and resides predominantly in the inner matrix compartment.

We also note that of these proteins, only Rad51C is predicted to localize to the mitochondria with a probability of 30% or 20% when analyzed using either MitoProt II 1.0a4 (Claros and

Vincens, 1996) or Predotar 1.03, respectively. These computational prediction programs evaluate protein N-terminal domains for the presence of mitochondrial targeting signals without defining a specific peptide sequence. Thus, further work will define the possible presence and function of such a signal in the Rad51C protein.

Given that exposure of cells to exogenous DNA damage induces a response that includes an increase in the nuclear levels of Rad51 (Mladenov et al., 2006; Gildemeister et al., 2009), we asked whether a similar increase occurs in mitochondria following DNA damage. In fact, cells treated with low levels of ionizing radiation (2 Gy IR) show increases in the level of 46

Figure 2.2: DNA damage-induced increases in mitochondrial Rad51, Rad51C, and Xrcc3. (a) Isolated mitochondria were treated with Proteinase K in the presence or absence of digitonin or SDS and blots were developed using antibodies against OPA1, an inner membrane space marker, TFAM, a matrix marker, and Rad51. (b) HCT116 cells were treated with 2 Gy IR, grown for 30 min or 2 hr and fractionated as described in Figure 2.1a. Equal amounts of total mitochondrial protein were immunoblotted for Rad51, Rad51C, and Xrcc3, as well as the loading control, ATP synthase. (c) HCT116 cells were treated with 15 mU/mL GO and fractionated at 15 and 30 min. Equal amounts of total mitochondrial protein were immunoblotted for Rad51, Rad51C, and Xrcc3, as well as the loading control, ATP synthase. All blots are representative of at least 3 independent experiments.

each protein at 30 and 120 min post-IR (Figure 2.2b). We next asked what effect oxidative stress, the major endogenous source of DNA damage encountered by mtDNA, would have on

HR protein localization. To generate this stress, GO was added to the culture media using amounts that result in chronic exposure of cells to a low level of H202 (Salazar and Van

Houten, 1997) (Figure 2.3). HCT116 cells treated with 15 mU/mL GO consistently showed an increase in the mitochondrial levels of Rad51, Rad51C, and Xrcc3 over a 30 min time course (Figure 2.2c). Similar results were obtained using HeLa cells (Figure A1).

Interestingly, increases in mitochondrial levels of the human apurinic/apyrimidinic endonuclease APE/Ref-1, a component of the BER pathway, have been observed in response to DNA damage (Frossi et al., 2002). Together with this and other studies (Tembe and

Henderson, 2007; Gildemeister et al., 2009), our results support the idea that cellular redistribution of HR and other signaling and repair proteins is part of a general cellular response to DNA damage.

To further support a role for HR in the oxidative stress response, we evaluated the physical interaction between Rad51 and mtDNA using mtDNA immunoprecipitation (mIP). We observed no significant interaction between Rad51 and mtDNA in unstressed U2OS cells relative to a mock mIP (Figure 2.4a). However, following exposure of cells to oxidative stress there is a > 2-fold increase in the association of Rad51 with mtDNA at 15 min of GO treatment, with a further increase at 30 min (Figure 2.4a). As controls for the immunoprecipitation conditions, we used TFAM, a known mtDNA binding protein, and

PCNA, a nuclear DNA binding protein not present in the mitochondria. As expected, there 48

Figure 2.3: Production of H2O2 by glucose oxidase. Varying concentrations of GO were added to U20S cells grown in DMEM and incubated at 37∘C for the indicated times.

H2O2 concentrations were assayed using a colorimetric-based kit and detected using a plate reader. Each data point represents the average of 8 experimental replicates. Error bars indicate standard error of the mean.

was no enrichment of mtDNA observed when precipitating with PCNA antibody (Figure

2.4a,b), and we observe a ~3.5 fold enrichment with TFAM (Figure 2.4b). We also find that oxidative stress increases the association of TFAM with mtDNA up to 7-fold over mock after

30 min of GO treatment, a result consistent with previous studies by Yoshida et al. (Yoshida et al., 2002) in which TFAM was shown to preferentially bind oxidatively damaged DNA.

These results are consistent with Rad51 playing a significant role in regulating mtDNA copy number in response to stress relative to non-stress conditions.

What is the mitochondrial function of human Rad51 and other HR proteins? Hundreds of mitochondria can exist in a given cell with each individual organelle containing multiple copies of the 16.5 kbp genome. In fact, the mtDNA copy number of a given cell type has been shown to be tissue specific and linked to cell-specific metabolic requirements, as perturbations in copy number are associated with a variety of human diseases (Longley et al.,

2006). In light of its mitochondrial localization (Figure 2.2) and its interaction with mtDNA

(Figure 2.4), we sought to evaluate the role that Rad51 may play in the maintenance of mtDNA copy number under both non-stress and stress conditions. Rad51 protein was depleted from cycling cells and mtDNA copy number was monitored as a function of time by qPCR. At 48 h post-transfection with Rad51-specific siRNAs, no significant change in copy number was observed relative to cells transfected with a pool of non-silencing siRNAs as determined by quantitative PCR (Figure 2.5a). However, mtDNA copy number increased by

~70% following 5 days of Rad51 depletion. Similar results were observed when changes in mtDNA copy number were evaluated by Southern blot (Figure A2). This increase 50

Figure 2.4: Rad51 physically interacts with mtDNA in an oxidative stress-dependent manner. (a) At 0, 15, and 30 min following treatment with 15 mU/mL GO, U2OS cells were incubated with formaldehyde and mtDNA was immunoprecipitated using anti-Rad51 or anti-PCNA antibody. Resulting DNA samples were evaluated by SYBR Green qPCR, and changes in fold enrichment of mtDNA over mock IP were calculated and normalized to input DNA (n=12). Similar results were obtained using HeLa and HCT116 cells (not shown). (b) Experiments similar to those in (a) were performed but immunoprecipitations were done using anti-TFAM and anti PCNA antibodies.

undoubtedly results from prolonged cell cycle arrest induced by Rad51 depletion (Figure

2.5b), as similar arrest-induced increases in mtDNA copy number have been reported previously (Lee and Wei, 2005). Therefore, changes in mtDNA copy number under these conditions appear to result from a general inhibition of cell cycle progression rather loss of a specific Rad51 function. However, given the important role of Rad51 in the nuclear DNA damage response, and the fact that we observed increases in the mitochondrial levels of

Rad51, Rad51C and Xrcc3 in response to DNA damage, we next assessed the effect of Rad51 depletion on mtDNA copy number in cells exposed to low levels of oxidative stress.

The production of ATP by enzymes of the electron transport chain located in the inner mitochondrial membrane is the primary source of ROS in the cell. Due to its close proximity to this membrane, the mtDNA is routinely subjected to oxidative stress leading to a variety of

DNA lesions including oxidized bases, protein-DNA crosslinks, abasic sites, and strand breaks. While BER is the predominant DNA repair pathway in mammalian mitochondria

(Santos et al., 2002; Longley et al., 2006; Stuart and Brown, 2006), these lesions also represent replication blocks that can lead to fork stalling or collapse. Cells transfected with non-silencing siRNAs, followed by exposure to 5 mU/mL GO showed a characteristic, transient increase in mtDNA copy number at 30 min (Figure 2.6a), consistent with previous reports of stress-induced increases observed in various cell types (Lee and Wei, 2005). This stress-induced increase was also observed when cells were treated with a control siRNA targeting CDK9 (Figures 2.6a,c), a protein not associated with Rad51-mediated recombination that does not localize to the mitochondria. In both cases, copy number 52

Figure 2.5: Depleting cells of Rad51 leads to cell cycle arrest and a characteristic arrest-induced increase in mtDNA copy number. (a) U20S cells were transfected with either nonsilencing or Rad51-specific siRNAs (10 nM final concentration). Total cellular DNA was isolated at 2 and 5 days, and mtDNA copy number was evaluated by SYBR Green qPCR relative to the 18S rRNA gene internal control. Quantification incorporates three independent experiments (error bars indicate standard error of the mean). (b) Representative cell count of control or Rad51 siRNA treated cells harvested prior to processing for total DNA. Error bars indicate the standard error of the mean of 5 independent experiments.

returned to pre-stress levels at 1 h. One possible explanation for this observed increase in copy number following stress treatment, is that breaks that are generated release supercoiling of the mtDNA, rendering it a better substrate for PCR-based amplification. Under these levels of oxidative stress, however, Southern blot analysis of undigested mtDNA reveals no detectable shift in mtDNA conformation distribution (Figure A3). In contrast, cells depleted of Rad51 failed to display this characteristic initial rise in copy number (Figures 2.6a,c).

Strikingly, 4 h of treatment with GO led to an approximate 40% decrease in copy number in cells treated with Rad51 siRNAs relative to the non-silencing and CDK9 controls (Figure

2.6a). Analysis of mtDNA copy number at 8 h of GO treatment showed an overall 60% decrease in cells depleted of Rad51 relative to an approximate 10% decrease in control knock down cells [Figures 2.6a; cells analyzed at 6 and 8 h were exposed to 2.5 mU/ml continuous

GO, rather than 5 mU/ml for all earlier time points, to ensure that H2O2 concentration remained at a level that imposed damage specifically on mitochondrial and not nuclear DNA

(Figure 2.3)] (Salazar and Van Houten, 1997). These data show that depletion of Rad51 leads to a biphasic decrease in mtDNA copy number following treatment of cells with low levels of GO: 1. an initial rapid decrease in response to ROS instead of an early ROS-induced increase (0.5 h), and; 2. a continuous decrease over extended periods of exposure to GO. We observe a related biphasic decrease in mtDNA copy number following GO exposure with depletion of Rad51C or Xrcc3 (Figures 2.6b,c). Together, these results strongly suggest that a Rad51-mediated activity is critical for regulating mtDNA copy number under conditions of oxidative stress, and that this activity requires the functions of Rad51C and Xrcc3.

Figure 2.6: Depleting cells of Rad51, Rad51C, or Xrcc3 leads to a stress-induced reduction in mtDNA copy number. (a) U20S cells were transfected with either non- silencing, CDK9-specific siRNAs or Rad51-specific siRNAs (10 nM final concentration). 48 hr following transfection, cells were treated with 5 mU/mL GO, and DNA was isolated at 0, 0.5, 1, and 4 hr, or cells were treated with 2.5 mU/ml GO and isolated at 6 and 8 hr. mtDNA copy number was determined by qPCR, with changes calculated relative to unstressed controls and normalized to changes in the 18S rRNA gene. Error bars indicate standard error of the mean (non-silencing n=6, CDK9 n=4, Rad51 n=4). (b) U20S cells were transfected as above with siRNAs targeting Rad51C or Xrcc3 transcripts and mtDNA copy number determined by qPCR (Rad51C n=3, Xrcc3 n=3). (c) Equal amounts of total cellular protein were evaluated by Western blot. 55

It is well established that DNA replication and recombination processes are intimately linked, and that recombination proteins carry out functions essential to the restoration of the replication fork following its encounter with a blocking lesion (Flores-Rozas and Kolodner,

2000; Kreuzer, 2005; Heller and Marians, 2006; Li and Heyer, 2008). Particularly relevant to a role in mtDNA replication, human Rad51 and Xrcc3 have been shown to cooperate in regulating replication fork progression on damaged chromosomes (Henry-Mowatt et al.,

2003; Petermann et al., 2010; Petermann and Helleday, 2010). In a related finding, Shedge et al. (Shedge et al., 2007) found that the mitochondrially targeted plant RECA3 and MSH1 proteins appear to be components of a recombination-based genome surveillance mechanism that directs recombination-dependent replication to long DNA repeats. Therefore, a likely role for mitochondrial Rad51, Rad51C and Xrcc3 is to ensure faithful completion of mtDNA replication as the fork encounters blocking lesions.

Studies in yeast have led to the finding that a homologous DNA pairing protein, Mhr1 associates with an oxidative stress-induced double stranded DNA break in the control region of the mtDNA (Hori et al., 2009). The authors present a model in which this recombinagenic

DNA end is used to initiate rolling circle replication, presumably using an undamaged homologous mtDNA molecule as an error-free template (Hori et al., 2009). This ROS- induced use of a rapid, amplification mode of DNA replication could explain the transient increase in mtDNA copy number observed in this (Figure 2.6a,b) and other studies (Lee and

Wei, 2005). Therefore, it is possible that an additional function of human Rad51 in 56

regulating mtDNA copy number involves facilitating the initiation of an alternative replication mechanism when the cell senses an increase in the oxidative stress load. In fact, all potential components required for oxidative stress-induced initiation and catalysis of rolling circle replication are present in human mitochondria. The human Nth1 endonuclease, the homolog of yeast Ntg1, partitions to both the nucleus and mitochondria of human cells

(Trzeciak et al., 2004). Its endonuclease activity targets oxidized bases and, therefore, is appropriate for a role in initiating this replication mechanism (Ikeda et al., 1998). Potential mitochondrial candidates for DSB end processing include Dna2 (Zheng et al., 2008), Xrn2

(Pagliarini et al., 2008), and Exo1 (Pagliarini et al., 2008), with Exo1 being the most likely due to its processive 5´-3´ exonuclease activity (Pagliarini et al., 2008). Rad51 would be expected to function as the required Mhr1-like homologous pairing activity, and lastly, the complex containing mitochondrial Pol γ, the helicase TWINKLE, and mtSSB has been shown to carry out efficient rolling circle replication in vitro (Korhonen et al., 2004). In further studies we will investigate whether a rolling circle mechanism contributes to mtDNA replication and if any of these components play roles in this process.

This work highlights an important new role for Rad51 in the oxidative stress response of human cells. Specifically, the copy number and mIP data identify mtDNA as a novel Rad51 substrate in a pathway that likely facilitates the completion of mtDNA replication in the presence of DNA lesions. This conclusion is in keeping with what is currently known about mitochondrial DNA depletion syndrome, a clinically heterogeneous group of disorders in which all known disease-associated mutations are found in proteins involved in DNA 57

replication (Longley et al., 2006). Our studies demonstrate a specific requirement for Rad51,

Rad51C and Xrcc3 function following an increased oxidative load, and given the excessive amount of oxidative damage occurring during periods of metabolic stress, there is a clear rationale for the need for recombinase function. It is possible that mitochondrial Rad51 may be contributing to the repair of mtDNA DSBs and/or facilitating mtDNA replication fork repair or restart. Future work will focus on clarifying the mechanisms by which these HR proteins facilitate mtDNA replication in a stress-dependent manner.

CHAPTER III

Human Rad51 Supports Mitochondrial DNA Synthesis

Under Conditions of Increased Replication Stress

Abstract

Homologous recombination is an essential pathway in mammalian cells for the error-free repair of DNA double strand breaks. In addition, productive DNA replication of chromosomal

DNA, particularly under stress conditions, depends on Rad51-mediated recombination to repair collapsed forks and to facilitate replication restart. Our previous findings demonstrated for the first time that human Rad51 is recruited to mitochondria following treatment with oxidative stress, and plays a critical role in the maintenance of mtDNA copy number (Sage et al. 2010 J.

Biol. Chem. 285, 18984-18990). In the following study, we show that the recruitment of

Rad51 activity to the mitochondria under stress conditions is the result of ongoing mtDNA replication. Furthermore, the oxidative stress-induced degradation of mtDNA in the absence of

Rad51 depends entirely on ongoing mtDNA replication. In addition, Rad51 levels in the mitochondria also increase in cells recovering from mtDNA depletion. These findings highlight an important new role for Rad51 in supporting productive mtDNA replication, and further solidify recombinase activity as an indispensable tool for dealing with DNA replication stress.

Introduction

The recombinase activity of the Rad51 protein forms the catalytic core of the HR pathway for the repair of DNA DSBs. In contrast to more error-prone pathways such as NHEJ, HR is relatively error-free owing to the use of an undamaged DNA strand as a template for the activity of repair DNA polymerases. In addition to its traditional role in DSB repair, a number of studies have highlighted a critical role for Rad51-mediated HR in facilitating efficient genomic DNA replication in mammalian cells, particularly under DNA-damaging stress conditions (Petermann et al., 2010; Lundin et al., 2003; Lundin et al., 2002; Arnaudeau et al.,

2001). Indeed, in vertebrate cells, Rad51 is regulated such that its activity is highest during the

S and G2 phases of the cell cycle when the bulk of DNA replication occurs and sister chromatids are present (Rothkamm et al., 2003; Hartlerode et al., 2011).

Although the vast majority of cellular DNA is located in the nucleus, a small 16.5 kb cytosolic genome (mtDNA) is maintained in the mitochondrial matrix. While small in size, this circular genome is present in thousands of copies per cell. This copy number is tissue specific and highly regulated to provide for the energetic needs of a given cell (Jeng et al., 2008; Mineri et al., 2009; Pohjoismaki et al., 2010). The mitochondrial genome encodes 13 proteins essential to the oxidative phosphorylation pathway for the production of ATP in addition to transfer and ribosomal RNAs needed for their expression. Given its close proximity to the site of oxidative phosphorylation, mtDNA is routinely subjected to oxidative stress leading to numerous types of DNA damage. Recent years have seen an ever-expanding appreciation for the ability of the mitochondrion to repair its DNA using various repair pathways including long and short patch 60

base excision repair as well as mismatch repair (Liu and Demple, 2010). We have recently shown that the Rad51 protein localizes to human mitochondria, and levels increase significantly in response to oxidative stress (Sage et al., 2010). Additionally, we found that

Rad51 interacts with mtDNA and supports maintenance of copy number under these stress conditions. Oxidative stress of this type largely generates mtDNA damage in the form of single strand breaks and abasic sites (Shokolenko et al., 2009). These lesions block the progression of DNA polymerases leading to stalled or collapsed replication forks, a classic substrate for Rad51 directed repair and replication restart (Petermann and Helleday, 2010). In the following study, we explore what role mtDNA replication plays in the need for Rad51 under oxidative stress conditions.

Experimental Procedures

Cell Culture, Treatment of Cells with Oxidative Stress, and siRNA Transfection- HCT116 cells

(ATCC #CCL-247) and U20S cells (ATCC #HTB-96) were grown at 37 °C in McCoy’s 5A medium (Invitrogen) with 5% CO2, in DMEM (Invitrogen) supplemented with 10% fetal bovine serum and antibiotics. Cell number was routinely determined by staining with trypan blue and counting on a haemocytometer. For oxidative stress, glucose oxidase (GO; Sigma

G7141) stock solutions were prepared prior to each experiment and added at the indicated concentrations. Cells were washed once with non-supplemented medium that was then replaced with non-supplemented medium containing the indicated concentration of GO. The resulting concentrations of hydrogen peroxide produced were determined using a Hydrogen

Peroxide Assay Kit (National Diagnostics). To inhibit DNA replication, aphidicolin (Sigma) 61

was prepared in DMSO and added directly to the culture medium to a final concentration of 3

µM, dideoxycytidine (ddC; Sigma) was prepared in water and added to a final concentration of

20 µM, and ethidium bromide (EtBr; Sigma) was prepared in water and added to a final concentration of 50 ng/mL. Rad51 protein was depleted by treating cells with a pool of siRNAs

(10 nM final, Dharmacon L-003530-00) delivered using Dharmafect1. A pool of non-silencing siRNAs was used as an additional negative control.

Bromodeoxyuridine Labeling of Cells for Immunofluorescent Detection of Nascent DNA-U20S cells were seeded onto glass cover slips prior to the addition of ddC, aphidicolin, or EtBr.

Cells were treated for 2 h with inhibitor prior to the addition of 5-bromo-2'-deoxyuridine

(BrdU; BD Biosciences) to 10 µg/mL for 2 h. Thirty min prior to the end of labeling, cells were treated with 50 nM MitoTracker Deep Red (Invitrogen). Cells were crosslinked by the addition of 4% formaldehyde at room temperature for 10 min. After washing cells twice with

PBS, DNA was denatured by incubating cells in 2N HCL for 10 min, followed by pH neutralization by washing twice with 200 mM Tris-HCl pH 8.0. Cells were permeabilized by incubation in ice-cold methanol for 10 min. Cells were incubated for 1 h in blocking buffer

[3% FBS and 0.5% Triton X100 in PBS] with rocking. Cells were incubated with a 1:500 dilution of mouse anti-BrdU antibody (Roche) in wash buffer [1% FBS and 0.5% Triton X100 in PBS] for 2 h with rocking, followed by 3 x 5 min washes. Cells were then incubated with a

1:500 dilution of Alexa Fluor 488 goat anti-mouse antibody (Invitrogen) in wash buffer for 1 h with rocking followed by 6 x 5 min washes. Finally cover slips were mounted onto glass slides with VECTASHIELD mounting medium containing DAPI (Vector Labs). Images were 62

captured with the Axioskope fluorescence microscope fitted with the Axiocam MRm camera and accompanying software (Zeiss)

Subcellular Fractionation and Isolation of Mitochondrial Protein Fraction, and

Immunoblotting- Intact mitochondria were isolated from cells as previously described (Sage et al., 2010). Briefly, following a gentle lysis using a hypotonic buffer containing digitonin

(Sigma), cells were subjected to a series of centrifugation steps to isolate a crude mitochondrial pellet. This pellet was washed four times with high salt buffer prior to solubilization of the mitochondrial membrane to yield the mitochondrial protein fraction. Protein concentrations were determined using the BCA Protein Assay Kit (Pierce), and samples were prepared by boiling in 2x Laemmli sample buffer (Sigma). Samples were run on precast 4-12% Bis-Tris gels (Invitrogen), and immunoblotted as previously described (Sage et al., 2010) using mouse anti-Rad51 (Millipore), mouse anti-ATP synthase (BD Transduction Laboratories), rabbit anti-

TFAM (Abcam), mouse anti-GAPDH (Millipore), and rabbit anti-PCNA (Abcam) antibodies.

DNA Isolation and Determination of mtDNA Copy Number - Total cellular DNA was isolated using the QIAamp DNA Mini Kit (Qiagen), and DNA concentrations were determined spectrophotometrically. Equal amounts of total cellular DNA were assayed by qPCR as previously described (Sage et al., 2010), with changes in mtDNA copy number quantified relative to the 18S rRNA gene.

Results

Mitochondria show a distinct increase in levels of Rad51 protein following exposure to oxidative stress (Sage et al., 2010), suggesting that Rad51 may be limiting for a required recombination or replication restart function under these conditions. As this type of stress has been shown to induce an increase in mtDNA content (Lee et al., 2000; Lee et al., 2002; Lee and Wei, 2005; Wei et al., 2001), it is possible that Rad51 functions to support productive mtDNA replication following DNA damage. To explore this possibility, we developed a series of experiments to investigate the relationship between Rad51 and mtDNA replication. Using

BrdU incorporation as a readout for DNA replication, we show that a 2 h pulse results in clear incorporation of the label into both nuclear and mtDNA in U20S cells (Figure 3.1a). The nucleoside analogue ddC has been shown to be a potent and specific inhibitor of mtDNA replication at low concentrations (Feng et al., 2001), and pretreatment of cells with 20 µM ddC for 2 h prior to the BrdU pulse prevents detectable incorporation into mtDNA, while allowing nuclear DNA replication to proceed (Figure 3.1b). Conversely, pretreatment of U20S cells with the DNA Pol α inhibitor, aphidicolin, greatly reduces incorporation of BrdU into nuclear

DNA, but not mtDNA (Figure 3.1c). Therefore, we can specifically and selectively inhibit either mitochondrial or nuclear DNA replication.

To evaluate the role of mtDNA replication in the recruitment of Rad51 to the mitochondria, we utilized a subcellular fractionation procedure to isolate a purified mitochondrial protein fraction free of nuclear and cytosolic contaminants from U20S cells (Sage et al., 2010). This procedure results in a protein fraction enriched for the mitochondrial markers TFAM and ATP synthase, 64

Figure 3.1: Replication inhibitors selectively prevent incorporation of BrdU into newly synthesized DNA. (a) U20S cells were treated for 2 hours with 10 ug/mL BrdU and MitoTracker Deep Red (red), then stained with an anti-BrdU antibody (green). (b) Inhibition of mtDNA labeling was achieved with a 2 h pretreatment with 20 M ddC prior to the BrdU pulse. (c) Nuclear incorporation of BrdU was inhibited by pre-treating cells with 3 uM aphidicolin prior to BrdU pulse. Nuclear compartment was stained with DAPI (blue). Images are representative of 250 cells examined.

and free of detectible levels of the nuclear marker PCNA and the cytosolic marker GAPDH

(Fig 3.2a). Increased oxidative stress was generated by adding the enzyme glucose oxidase

(GO) to the culture medium (Sage et al., 2010). This allows the consistent production of H2O2 over an extended time course at concentrations that have previously been shown to preferentially damage mtDNA (Salazar and Van Houten, 1997). Using this fractionation procedure, we show that treatment of cells with oxidative stress induces a recruitment of Rad51 to the mitochondria (Figure 3.2b). Strikingly, inhibiting mtDNA replication by pre-treating cells with ddC for 2 h prior to stress completely abrogates this recruitment. In contrast, inhibiting nuclear DNA replication with aphidicolin treatment has no effect on the recruitment of Rad51 to the mitochondria (Figure 3.2b). Similar results were observed using HCT116 cells

(Figure A4). As ddC pretreatment is unlikely to prevent oxidative damage to DNA due to GO treatment, these results may suggest that recruitment of Rad51 to the mitochondria in response to oxidative stress requires ongoing mtDNA replication, and that oxidatively damaged DNA in itself is not the signal for recruitment.

An important measure of efficient mtDNA replication is the maintenance of the number of copies of the genome per cell. mtDNA copy number is highly regulated and cell type specific, and defects have been linked to a variety of human diseases (Graziewicz et al., 2006). We have shown that the maintenance of mtDNA levels under oxidative stress conditions requires the presence of Rad51 (Sage et al., 2010). To evaluate the potential interplay between mtDNA replication and Rad51 in this maintenance process, we first monitored changes in mtDNA copy number by qPCR in the presence or absence of endogenous levels of Rad51. In cells 66

Figure 3.2: Recruitment of Rad51 to the mitochondria in response to stress is suppressed by inhibition of mtDNA replication. Purified mitochondrial protein fractions were obtained by differential centrifugation followed by sequential wash steps. (a) Purity of fractions was analyzed by Western blot for marker proteins TFAM and ATP synthase (mitochondrial), PCNA (nuclear), and GAPDH (cytosolic). (b) Western blot of purified mitochondrial fractions for Rad51 following treatment of cells with 10 mU/mL GO +/- a 2 hour pretreatment with either 20 M ddC or 3 uM aphidicolin. ATP synthase serves as the loading control.

expressing normal levels of Rad51 (Figure 3.3a, non-silencing siRNA), GO treatment induces a transient boost in mtDNA replication leading to an increase in copy number that returns rapidly to pre-stress levels. However, when cells are depleted of Rad51, copy number rapidly drops in response to stress (Figure 3.3a, Rad51 siRNA). We then repeated this experiment but included a pre-treatment step with ddC to block mtDNA replication. Interestingly, we find that this completely prevents both the boost in copy number seen when Rad51 is present at normal levels as well as the drop observed with depletion of Rad51 (Figure 3.3b). In contrast, pretreatment with aphidicolin fails to suppress the drop in copy number when Rad51 has been depleted (Figure 3.3c). These results suggest that successful mtDNA replication under stress conditions requires the activity of Rad51. Furthermore, the stress-induced reduction in copy number in the absence of Rad51 suggests degradation of mtDNA, which we also find to be dependent on ongoing mtDNA replication.

Previously, we showed that unstressed cells depleted of Rad51 for several days display an increase in mtDNA copy number (Sage et al., 2010), which is characteristic of cells that have been arrested (Lee et al., 2000; Pagliacci et al., 1993). Although this finding may suggest that

Rad51 is not required to maintain mtDNA copy number under non-stress culture conditions, an involvement of Rad51 under these conditions is likely to be masked by the severely reduced rate of cell division when it is RNAi-depleted. To investigate the role of Rad51 in mtDNA copy number maintenance in the absence of DNA damage, we treated cells with ddC or the intercalating dye ethidium bromide (EtBr) to force mtDNA copy number to drop as a result of dilution through cell division. We then washed the drug away and allowed mtDNA copy 68

Figure 3.3: Stress-induced mtDNA depletion in the absence of Rad51 requires ongoing mtDNA replication. (a) U20S cells were transfected with nonsilencing or Rad51- specific siRNAs (10 nm final concentration). 48 h following transfection, cells were treated with 5 mU/ml GO or (b) treated with GO following a 2 h pretreatment with 20 M ddC or (c) 3 uM aphidicolin. DNA was isolated at 0, 0.5, 1, and 2 h. mtDNA copy number was determined by qPCR, with changes calculated relative to unstressed controls and normalized to changes in the 18 S rRNA gene. Error bars indicate standard error of the mean (n = 4).

Figure 3.4: Ethidium bromide selectively inhibits incorporation of BrdU into mitochondrial but not nuclear DNA. (a) U20S cells were treated for 2 h with 10 ug/mL BrdU and MitoTracker Deep Red (red), then stained with an anti-BrdU antibody (green). Nuclear compartment was stained with DAPI (blue) (b) Inhibition of mtDNA labeling was achieved with a 2 h pretreatment with 50 ng/mL EtBr prior to the BrdU pulse.

number to recover. As with ddC treatment, pretreatment with 50 ng/mL EtBr for 2 h prevents detectible incorporation of BrdU into mtDNA while leaving nuclear DNA incorporation unaffected (Figure 3.4a,b). While both inhibitors effectively depleted cells of mtDNA, ddC led to a more rapid decrease in copy number than was seen with EtBr (Figure 3.5a). In addition, a lag period was observed between the removal of ddC and the recovery of copy number, where as removal of EtBr led to a rapid reversal of inhibition. This system generates two distinct phases, “depletion” where mtDNA replication is inhibited and “recovery” where the requirement for mtDNA replication is significantly higher than under normal culture conditions

(hyper replication). Western blot analysis of mitochondrial protein fractions reveals low levels of Rad51 during the “depletion” phase when mtDNA replication has been inhibited with either ddC or EtBr for two days, similar to the untreated control (Figure 3.5b). Following drug removal, however, when mitochondria have entered a hyper replication mode, levels of Rad51 in the mitochondria rise dramatically (Figure 3.5b), suggesting that a recombination activity is a component of the mitochondrial response to an increased demand for mtDNA replication in the absence of exogenous DNA damage.

To further evaluate the importance of Rad51 during conditions of hyper replication, we transfected cells with Rad51 siRNAs (Figure 3.6) and treated them with EtBr for two days

(Figure 3.7a). After removal of the drug, it appears that recovery of mtDNA copy number proceeds at generally the same rate as cells with endogenous levels of Rad51 (Figure 3.7a).

However, it is important to note again that cells expressing normal levels of Rad51 divide 71

Figure 3.5: Recovery of mtDNA copy number following mtDNA depletion recruits Rad51 to the mitochondria. (a) U20S cells were treated with either 20 uM ddC or 50 ng/mL EtBr. Following 2 days of treatment, cells were washed and allowed to recover for 4 days. DNA was isolated every 24 hours, and mtDNA copy number was determined by qPCR, with changes calculated relative to untreated controls and normalized to changes in the 18 S rRNA gene. Error bars indicate standard error of the mean (n = 4). (b) Western blot of purified mitochondrial fractions isolated from untreated cells (Unt), cells treated with either inhibitor for 24 h (Dep) and cells 24 h post drug removal (Rec) for Rad51 as well as ATP synthase loading control. The blot is representative of four independent experiments.

Figure 3.6: Transfection of cells with Rads51 siRNAs depletes cells of protein. (a) U20S cells were transfected with a pool of nonsilencing siRNAs (10 nM) or (b) Rad51 siRNAs (10 nM) and cultured for six days. Every 24 h, cells were harvested and processed for total protein. Equal amounts of total protein were analyzed by Western blot. ATP synthase is a loading control, and blots are representative of three independent experiments.

about 3 times more rapidly than those depleted of Rad51 (Figure 3.7b). Therefore, under

Rad51 knock down conditions, the number of successful rounds of mtDNA replication is far fewer than for actively dividing cells. This difference in replication demand is illustrated in

Figure 3.8 where the mitochondrial genome copy number in two hypothetical cells (presumed to start at 10 copies) increases by 30%. The arrested cell (similar to Rad51 knock down) needs only to replicate 3 new genomes to achieve this increase, while an actively dividing cell must replicate 16 new genomes to account for cell doubling and a 30% increase in each daughter cell (Figure 3.8). Differences in the total amount of mtDNA synthesized by the two populations of cells mirrors the differences in division rate, with the control population synthesizing ~4-fold more mtDNA relative to Rad51 siRNA-transfected cells (Figure 3.9).

The lower amount of total mtDNA synthesized over the 4 day recovery period when Rad51 has been depleted could potentially indicate that mtDNA replication proceeds more slowly under these conditions. Our demonstration that under normal culture conditions Rad51 is recruited to the mitochondria during “recovery” from mtDNA depletion (Figure 3.5b) supports the idea that recombination is playing an active role in mtDNA synthesis under conditions of increased replication demand. Clarification of the role of Rad51 during hyper mtDNA replication will require the generation of conditions where cell growth rates are equivalent +/- Rad51 expression.

Figure 3.7: Depleting cells of Rad51 does not affect the rate of mtDNA copy number recovery following mtDNA depletion. U20S cells were transfected with nonsilencing, or Rad51-specific siRNAs (10 nm final concentration), and simultaneously treated with 50 ng/mL EtBr. Two days following treatment, cells were washed and allowed to recover for four days. (a) Total cellular DNA was isolated from untransfected cells (dotted line) or Rad51 siRNA transfected cells (solid line) every 24 h and assayed for mtDNA copy number by qPCR, with changes calculated relative to untreated controls and normalized to changes in the 18 S rRNA gene. Error bars indicate standard error of the mean (n = 3). (b) Cells were trypsinized and counted every 24 h prior to processing for DNA. Cell numbers displayed relative to initial seeding.

Figure 3.8: Effect of cell division on mtDNA repication demand during recovery from depletion. Model illustrating the difference in DNA replication demand between cycling and non-cycling cells. During the course of one population doubling time, an arrested cell must only replicate a number of genomes equivalent to the observed copy number increase (left pathway). An actively dividing cell must duplicate its mtDNA compliment in addition to replication required to account for the increase in copy number observed (right pathway).

Figure 3.9: Effect of Rad51 depletion on total mtDNA synthesized following removal of ethidium bromide. U20S cells were transfected with either nonsilencing or Rad51- specific siRNAs. The two population of cells were then treated with 50 ng/mL EtBr for two days to inhibit mtDNA replication, then washed and allowed to recover for four days. Each day, total DNA was isolated for qPCR determination of copy number per cell. Fold change in total mtDNA in each population was determined by multiplying copy number per cell by relative cell count for each day.

Discussion

In a previous study, we identified a critical need for Rad51 in the mitochondrial response to an increase in oxidative stress (Sage et al., 2010). Although the majority of damage generated by this stress (single strand breaks, abasic sites, base adducts, etc.) is repaired by base excision repair, all of these DNA lesions represent potent blocks to replication. The restart of stalled or collapsed replication forks is critical for the maintenance of genomic DNA integrity and is a process in which Rad51 recombinase activity is intimately involved (Helleday, 2003;

Petermann and Helleday, 2010). In the present study, we show that oxidatively damaged DNA by itself is insufficient to recruit Rad51 to the mitochondria. Rather, we find that ongoing mtDNA replication in the presence of this damage is required for the recruitment of Rad51. In addition, the stress-induced drop in mtDNA copy number seen when Rad51 has been depleted can be completely prevented by halting mtDNA replication prior to the application of stress

(Figure 3.3b). This copy number decrease in the absence of Rad51 is rapid (Figure 3.3a) and indicative of degradation of extensively or persistently damaged molecules. This rapid drop is in contrast to the gradual decrease in copy number observed when Rad51 has not been depleted and replication has been inhibited (Figure 3.10). In this case, the decline in mtDNA levels can be accounted for by simple dilution due to cell division. Therefore, our data suggest that ongoing replication contributes significantly to degradation of mtDNA following oxidative stress in the absence of Rad51. Additionally, our studies support the idea that Rad51 likely promotes mtDNA replication under stress conditions by assisting in the repair of collapsed 78

Figure 3.10: Depletion of mtDNA copy number through dilution following ddC treatment. U20S cells were seeded and treated with 20 uM ddC. Cells were harvested and processed for total genomic DNA every 24 hours. DNA was assayed for mtDNA content by qPCR and percent mtDNA copy number is displayed relative to initial levels. Error bars indicate srandard error of the mean (n = 3).

forks or stabilizing stalled forks, thus preventing the accumulation of replication-induced mtDNA damage that would lead to its rapid degradation.

The recovery of mtDNA copy number following its depletion using ddC or EtBr requires that more mtDNA molecules are replicated than the cell would otherwise need to produce. As described in Figure 3.8, the magnitude of the increase in replication is far greater in cells that are actively proliferating, as half of the mtDNA is passed onto the daughter cells. Cells that are arrested need to replicate only a fraction of the new genomes relative to their cycling counterparts to achieve an equivalent percent increase in copy number. The recruitment of

Rad51 we observe under maximal replication demand (when cells are cycling during recovery) likely results from an increase in the frequency of fork stalling or collapse, which would require recombinase activity for replication to proceed.

In bacteria, it has been reported that as many as 50% of cells require fork restart during a single round of DNA replication (Cox et al., 2000). While the frequency of fork collapse in mtDNA is unknown, the constant exposure to oxidative damage likely makes replication restart a common event. In a variety of organisms including bacteria (McGlynn and Lloyd, 2002;

Kreuzer, 2005; Heller and Marians, 2006), yeast (Flores-Rozas and Kolodner, 2000; Davis and

Symington, 2004), and humans (Petermann and Helleday, 2010) recombination forms the basis of most models for dealing with collapsed forks during genomic DNA replication. There is also evidence that recombination events occur in the mitochondria of various organisms. For 80

example, both intra- and inter-molecular recombination can be detected following restriction enzyme-mediated introduction of DSBs into mouse mtDNA (Bacman et al., 2009). Four-way junction structures are prevalent in human heart mtDNA (Kajander et al., 2001), and it was proposed that these derive from recombination activity associated with replication initiation and processivity in a tissue with such high metabolic demand (Pohjoismaki et al., 2009;

Pohjoismaki and Goffart, 2011). Our demonstration that recruitment of human Rad51 activity to mitochondria when replication demand is high or when DNA lesions are abundant highlights the central role of HR in supporting productive mtDNA replication and maintenance of the mitochondrial genome. Future studies will shed further light on the mechanism by which

Rad51 activity supports mtDNA replication processivity and fidelity.

CHAPTER IV

Discussion and Future Directions

The function of recombination in mammalian mtDNA maintenance has been an issue of ongoing debate for several decades. Motivating this controversy is the important role mtDNA sequence stability plays in the study of population genetics. A mitochondrial genome that does not recombine is the ideal tool for tracking the origin and age of ancient peoples. For many years, geneticists have concluded that human mtDNA does not recombine, citing the lack of abundant evidence of recombination in mtDNA sequence databases (Macaulay et al., 1999). However, the current lack of evidence for homologous recombination does not preclude the possibility that it exists, as recombination between identical genomes within a mitochondrion cannot be easily detected. This may, in fact, be the very reason for uniparental inheritance of mtDNA; to prevent the generation of genetic variation. The lack of significant noncoding sequence in mtDNA combined with the essential role that mtDNA-encoded genes play in ATP

(and ROS) production means that even small amounts of mitochondrial genome instability could be catastrophic to the cell. If genetic variation is not the desired outcome of mtDNA recombination, what then is the reason for the recombination activity that has been observed in a number of mammalian systems (Thyagarajan et al., 1996;

Kraytsberg et al., 2004)? The answer likely lies in the abundance of evidence linking recombination activity to the support of DNA replication. In particular, recombination appears to be required for efficient replication under DNA-damaging stress conditions in 82

a variety of genetic systems including bacteria, yeast, and mammalian chromosomal

DNA (Cox et al., 2000; Haber, 1999; McGlynn and Lloyd, 2002). It is clear from decades of research in these various systems that DNA recombination is a requirement for reasons beyond the generation of genetic diversity.

In this study, we show for the first time that the major recombinase protein in mammalian cells, Rad51, localizes to the human mitochondrial matrix where it physically interacts with mtDNA. Importantly, Rad51 is recruited to the mitochondria in response to an increase in oxidative stress. Therefore, it appears that Rad51 is part of an adaptive stress response. The recruitment of DNA repair proteins to mitochondria in response to oxidative stress has been previously observed with the multifunctional BER endonuclease

APE/Ref-1 (Frossi et al., 2002). Indeed, relocalization of DNA repair and signaling factors in response to stress is a phenomenon that has been observed in numerous reports

(Tembe and Henderson, 2007; Griffiths et al., 2009). Recently, we have shown that treatment of cells with IR leads to a movement of cytosolic Rad51 into the nucleus

(Gildemeister et al., 2009), similar to the movement of Rad51 into mitochondria shown here (Figure 2.2b) (Sage et al., 2010). Taken together, these findings suggest that redistribution of Rad51 is a critical method of regulating the nuclear and mtDNA damage responses. Although the purpose of this type of regulation is not well understood, the movement of preexisting pools of proteins is, in general, a more rapid mechanism of response to stress relative to de novo synthesis of limiting protein factors. This quicker response time may be critical for mitigating the extent and complexity of damaged DNA 83

structures following exposure to DNA-damaging stresses, particularly in the context of ongoing DNA replication.

An understanding of the signal origins that lead to Rad51 recruitment provides important insight into the function of Rad51-mediated HR in mtDNA maintenance. Interestingly, we show here that neither oxidative stress nor oxidatively damaged DNA per se are sufficient for Rad51 recruitment. The combination of ongoing mtDNA replication and

DNA damage is needed to generate the signal leading to Rad51 movement into mitochondria (Figure 3.3). The stalling of a replication fork presents several possible structures that may elicit the signal(s) for the recruitment of Rad51, including extended tracks of ssDNA (or mtSSB-coated ssDNA) as well as the forked DNA itself, exposed by the dissociation of replication factors. In addition, the double-stranded DNA end resulting from the collapse of a replication fork is likely a substrate for end binding factors (such as the MRN complex) that then activate restart or repair mechanisms. An understanding of the signaling mechanisms that activate mitochondrial replication fork restart awaits a clearer picture of the signaling factors involved, and the nature of their substrates. In addition to the origin of signals resulting from replication fork collapse, the nature of the signal(s) themselves is also unclear. Although a number of different mitochondrially transmitted signals have been identified, including export of Ca2+, iron- sulfur protein complexes, and ATP, it remains to be understood how mtDNA replication fork stress could interface with these signaling mechanisms. Future work will be 84

necessary for understanding how cytosolic recombination proteins detect mtDNA replication fork problems and respond by translocation into the mitochondria.

A further link between HR and mtDNA replication is the recruitment of Rad51 to mitochondria of cycling cells that are recovering from mtDNA depletion (Figure 3.5).

While these conditions do not involve exogenous DNA-damaging stress, inevitably, a basal level of DNA lesions routinely lead to replication fork collapse. The increased replication demand induced by recovery from depletion likely increases the number of fork problems. Taken together, these findings suggest that replication fork damage is the signal for the recruitment of Rad51. This idea is in keeping with what is known about the role of Rad51 in nuclear DNA maintenance. As discussed earlier, the HR pathway is regulated such that it is the primary DSB repair mechanism functioning during S and G2 phases of the cell cycle (Rothkamm et al., 2003; Hartlerode et al., 2011), where it is critical for the restart and repair of stalled and collapsed replication forks (Petermann and

Helleday, 2010). As the mtDNA structures that result from replication fork collapse are likely similar to their nuclear counterparts, it is reasonable to predict that similar pathways are required for repair. As discussed earlier, a detailed mechanistic understanding of how Rad51 functions in replication fork repair has yet to be reported

(Figure 1.2). Future study of fork restart and repair in mitochondria may provide important insight into the larger and more complex problem of DNA replication and repair in the nucleus.

In addition to the recruitment of repair factors, the data presented in this thesis suggest several other cellular responses to an increase in oxidative stress. We show that stress consistently induces an increase in mtDNA copy number within the first hour of treatment (Figures 2.6, 3.3). This increase is suppressed by treatment with a Pol γ inhibitor, suggesting that mitochondria respond to increases in oxidative stress by increasing mtDNA replication. A second response to oxidative stress is manifested as a drop in mtDNA copy number when Rad51 expression has been knocked down. The rapid stress-induced loss of mtDNA is indicative of the active degradation of damaged mtDNA molecules. This is in contrast to the slow depletion of mtDNA when cells are cultured in the presence of a Pol γ inhibitor (Figure 3.10). Under these conditions, copy number per cell drops over the course of several days; however this drop is due to an overall increase in cell number and dilution of existing genomes rather than their degradation (Figure 3.9).

Taken together, our findings suggest a model in which an increase in oxidatively damaged mtDNA results in three major responses: (1) the recruitment of DNA repair pathways to the mitochondrial matrix, (2) the recruitment or activation of degradation machinery and/or mitophagy that eliminates extensively or persistently damaged mtDNA or organelles, and (3) activation of mtDNA replication, presumably to compensate for the degradation of damaged genomes. When Rad51 is depleted, mtDNA replication may be inefficient due to failure to restart or repair stalled or collapsed forks. This replication defect would negatively impact mtDNA copy number in two ways: failure to produce 86

new genomes efficiently, and contribution to mtDNA damage and eventual degradation.

The net effect of this is a decrease in mtDNA copy number in response to stress when

Rad51 is absent (Figures 2.6, 3.3).

Although the importance of Rad51 during mtDNA replication stress is clear from the findings presented here, much remains to be understood. An important question is, how

Rad51 is targeted to mitochondria. While a great many proteins have experimental evidence supporting their mitochondrial localization, there is no computationally detectable mitochondrial targeting signal in the primary sequence of Rad51 (Pagliarini et al., 2008; Sage et al., 2010). It is therefore likely that an unidentified accessory factor or post-translational modification is involved in trafficking Rad51 into the mitochondria in response to stress. As noted above, computational analysis indicates a possible mitochondrial targeting signal in the Rad51 paralog, Rad51C. How this potential signal contributes to HR protein mitochondrial localization remains to be investigated. This issue will be important for experimentally separating the mitochondrial vs. nuclear roles for Rad51, as well as possibly linking disease-associated mutations with failure to recruit

Rad51 to mitochondria in response to stress.

Remaining Questions

Although the results presented here provide a strong functional link between Rad51 and mtDNA replication, it remains to be shown directly if indeed replication restart occurs in mtDNA, and what role Rad51 plays in this process. Evidence for the restart of stalled 87

replication forks in the nucleus has been provided largely by single molecule DNA fiber analysis (Petermann et al., 2010). This approach involves the introduction of halogenated nucleotides into nascent DNA that are then detected indirectly by immunofluorescence microscopy. The use of multiple analogs (such as bromo- and chlorodeoxyuridine) allows quantification of the restart of stalled replication forks (Tuduri et al., 2010). This approach may be useful in addressing the issue of the restart of stalled forks in the mitochondria and evaluating its Rad51 dependence.

As indicated above, the mechanisms by which Rad51 supports replication fork restart and repair are not clearly understood (Figure 1.2). Helleday et al. have provided evidence for at least two mechanistically distinct roles for Rad51 in dealing with nuclear DNA replication fork stress (Petermann et al., 2010). One role, the restart of stalled forks, is not associated with DNA break generation or increased recombination events, therefore Rad51 may promote stabilization of the forked DNA through its single and double strand DNA binding capacity. In addition, Rad51 may facilitate annealing of nascent DNA strands during fork regression to form the “chicken foot” structure. A second role for Rad51 function involves the repair of collapsed replication forks. These repair events are associated with Rad51 foci formation, DSB generation, as well as increase in recombination events. This suggests a mechanism similar to Rad51-mediated DSB repair. While both restart and fork repair have been shown to be Rad51-dependent processes (Petermann et al., 2010), there are likely significant differences in the accessory proteins involved as well as the potential for recombination events. The findings presented here could potentially support either type of 88

Rad51 mechanism in supporting mtDNA copy number maintenance under stress conditions.

Finally, the repair of 2-ended DSBs (as described in Figure 1.1) must be evaluated in the mitochondria in order to clarify if Rad51 is playing a role in the classically studied HR

DSB repair pathway (Figure 1.1). In the nucleus, a reporter-based gene conversion assay has been developed to quantify the repair of DSBs by HR (Pierce et al., 1999). This reporter system has yet to be adapted to monitor mtDNA recombination, however, because methods to genetically modify mtDNA to introduce the reporter construct have yet to be developed. One possible solution could be to utilize previously developed mitochondrially targeted, site-specific endonucleases (Bayona-Bafaluy et al., 2005) to generate discrete breaks whose repair may be monitored by PCR-based approaches. This system would allow the generation of breaks with either staggered or blunt ends. In addition, repair of these breaks may be evaluated both with and without ongoing mtDNA replication.

The failure to maintain appropriate levels of mtDNA in a cell has serious metabolic implications for a variety of tissues in the body. For example, Alpers’ Syndrome, a progressive, early onset neurodegerative disorder, has been linked to decreases in mtDNA copy number (Naviaux and Nguyen, 2004). The low copy number characteristic of this disorder is thought to be associated with mutations in Pol γ that lead to mtDNA replication defects. Alpers syndrome is just one of a variety of disorders associated with 89

tissues that contain <30% mtDNA per cell than age-matched controls (Vu et al., 1998;

Morten et al., 2007). These disorders are clinically heterogenous and are collectively referred to as mitochondrial DNA depletion syndrome (MDS) (Rahman and Poulton,

2009). All known mutations associated with MDS (including Pol γ) are found within nuclear-encoded proteins that are associated with mtDNA replication (Szczesny et al.,

2008). Given the connection we have established here between Rad51 and mtDNA replication, it is likely that mutations associated with Rad51 function or localization may be identified in cases of MDS for which no causative factor has yet been identified.

Therefore, a clear understanding of how Rad51 functions in maintaining the mtDNA copy number is essential for an appreciation of how defects in the function of mitochondrial HR contribute to disease. Recent years have seen an expanding appreciation for the role of HR in DNA replication, particularly under conditions of increased DNA-damaging stress. The findings presented here extend this role to include support of mitochondrial DNA replication, further solidifying Rad51 as a critical adaptive component of DNA replication machinery.

Appendix

Figure A1: DNA damage-induced increases in mitochondrial Rad51 and Rad51C in HeLa cells. HeLa cells were treated with 2 Gy IR, grown for 30 min or 1 hr and fractionated as described in Figure 2.1a. Equal amounts of total mitochondrial protein were immunoblotted for Rad51 and Rad51C.

Figure A2: Southern blot analysis of changes in mtDNA copy number following long-term Rad51 depletion. (a) HCT116 cells were transfected with non- silencing or Rad51-specific siRNAs (10 nM) and cultured for the indicated times. Total cellular DNA was isolated every 24 h, digested with AccI, and resolved on a 0.7% agarose gel. Gel was blotted onto nylon and probed for mtDNA using randomly primed , radioactive oligos corresponding to the 1525-6339 region of the mitochondrial genome. (b) Densitometery was used to quantitate band intensities shown in (a) using the Multigauge software and are displayed relative to initial.

Figure A3: Treatment of cells with 20 mU/mL GO does not disturb mtDNA conformation distribution. HCT116 cells were treated or not with 20 mU/mL GO for 30 min prior to isolation of total cellular DNA. Equal amounts of total DNA were run on a 0.5% agarose gel and Southern blotted for mtDNA. 94

Figure A4: Recruitment of Rad51 to the mitochondria in response to stress is suppressed by inhibition of mtDNA replication in HCT116 cells. Purified mitochondrial protein fractions were obtained by differential centrifugation followed by sequential wash steps. Western blot of purified mitochondrial fractions for Rad51 following treatment of cells with 10 mU/mL GO +/- a 2 hour pretreatment with either 20 M ddC. ATP synthase serves as the loading control.

Bibliography

Akbari, M., Visnes, T., Krokan, H. E., and Otterlei, M. (2008). Mitochondrial base excision repair of uracil and AP sites takes place by single-nucleotide insertion and long- patch DNA synthesis. DNA Repair (Amst) 7, 605-616.

Albring, M., Griffith, J., and Attardi, G. (1977). Association of a protein structure of probable membrane derivation with HeLa cell mitochondrial DNA near its origin of replication. Proc Natl Acad Sci U S A 74, 1348-1352.

Amit, R., Gileadi, O., and Stavans, J. (2004). Direct observation of RuvAB-catalyzed branch migration of single Holliday junctions. Proc Natl Acad Sci U S A 101, 11605- 11610.

Andersen, S. L., Bergstralh, D. T., Kohl, K. P., LaRocque, J. R., Moore, C. B., and Sekelsky, J. (2009). Drosophila MUS312 and the vertebrate ortholog BTBD12 interact with DNA structure-specific endonucleases in DNA repair and recombination. Mol Cell 35, 128-135.

Anderson, D. G., and Kowalczykowski, S. C. (1997). The recombination hot spot chi is a regulatory element that switches the polarity of DNA degradation by the RecBCD enzyme. Genes Dev 11, 571-581.

Argueso, J. L., Westmoreland, J., Mieczkowski, P. A., Gawel, M., Petes, T. D., and Resnick, M. A. (2008). Double-strand breaks associated with repetitive DNA can reshape the genome. Proc Natl Acad Sci U S A 105, 11845-11850.

Arnaudeau, C., Lundin, C., and Helleday, T. (2001). DNA double-strand breaks associated with replication forks are predominantly repaired by homologous recombination involving an exchange mechanism in mammalian cells. J Mol Biol 307, 1235-1245.

Ayala-Torres, S., Chen, Y., Svoboda, T., Rosenblatt, J., and Van Houten, B. (2000). Analysis of gene-specific DNA damage and repair using quantitative polymerase chain reaction. Methods 22, 135-147.

Bacman, S. R., Williams, S. L., and Moraes, C. T. (2009). Intra- and inter-molecular recombination of mitochondrial DNA after in vivo induction of multiple double-strand breaks. Nucleic Acids Res 37, 4218-4226.

Bae, S. H., Bae, K. H., Kim, J. A., and Seo, Y. S. (2001). RPA governs endonuclease switching during processing of Okazaki fragments in eukaryotes. Nature 412, 456-461.

Barr, C. M., Neiman, M., and Taylor, D. R. (2005). Inheritance and recombination of mitochondrial genomes in plants, fungi and animals. New Phytol 168, 39-50.

Bayona-Bafaluy, M. P., Blits, B., Battersby, B. J., Shoubridge, E. A., and Moraes, C. T. (2005). Rapid directional shift of mitochondrial DNA heteroplasmy in animal tissues by a mitochondrially targeted restriction endonuclease. Proc Natl Acad Sci U S A 102, 14392- 14397.

Bendich, A. J. (1996). Structural analysis of mitochondrial DNA molecules from fungi and plants using moving pictures and pulsed-field gel electrophoresis. J Mol Biol 255, 564-588.

Bennett, B. T., and Knight, K. L. (2005). Cellular localization of human Rad51C and regulation of ubiquitin-mediated proteolysis of Rad51. J Cell Biochem 96, 1095-1109.

Berk, A. J., and Clayton, D. A. (1974). Mechanism of mitochondrial DNA replication in mouse L-cells: asynchronous replication of strands, segregation of circular daughter molecules, aspects of topology and turnover of an initiation sequence. J Mol Biol 86, 801-824.

Bishop, D. K., Ear, U., Bhattacharyya, A., Calderone, C., Beckett, M., Weichselbaum, R. R., and Shinohara, A. (1998). Xrcc3 is required for assembly of Rad51 complexes in vivo. J Biol Chem 273, 21482-21488.

Brown, T. A., and Clayton, D. A. (2002). Release of replication termination controls mitochondrial DNA copy number after depletion with 2',3'-dideoxycytidine. Nucleic Acids Res 30, 2004-2010.

Carrodeguas, J. A., Pinz, K. G., and Bogenhagen, D. F. (2002). DNA binding properties of human pol gammaB. J Biol Chem 277, 50008-50014.

Chen, C. H., Vazquez-Padua, M., and Cheng, Y. C. (1991). Effect of anti-human immunodeficiency virus nucleoside analogs on mitochondrial DNA and its implication for delayed toxicity. Mol Pharmacol 39, 625-628.

Chen, F., Nastasi, A., Shen, Z., Brenneman, M., Crissman, H., and Chen, D. J. (1997). Cell cycle-dependent protein expression of mammalian homologs of yeast DNA double- strand break repair genes Rad51 and Rad52. Mutat Res 384, 205-211.

Chilkova, O., Stenlund, P., Isoz, I., Stith, C. M., Grabowski, P., Lundstrom, E. B., Burgers, P. M., and Johansson, E. (2007). The eukaryotic leading and lagging strand 97

DNA polymerases are loaded onto primer-ends via separate mechanisms but have comparable processivity in the presence of PCNA. Nucleic Acids Res 35, 6588-6597.

Claros, M. G., and Vincens, P. (1996). Computational method to predict mitochondrially imported proteins and their targeting sequences. Eur J Biochem 241, 779-786.

Coffey, G., and Campbell, C. (2000). An alternate form of Ku80 is required for DNA end-binding activity in mammalian mitochondria. Nucleic Acids Res 28, 3793-3800.

Coffey, G., Lakshmipathy, U., and Campbell, C. (1999). Mammalian mitochondrial extracts possess DNA end-binding activity. Nucleic Acids Res 27, 3348-3354.

Constantinou, A., Tarsounas, M., Karow, J. K., Brosh, R. M., Bohr, V. A., Hickson, I. D., and West, S. C. (2000). Werner's syndrome protein (WRN) migrates Holliday junctions and co-localizes with RPA upon replication arrest. EMBO Rep 1, 80-84.

Cox, M. M., Goodman, M. F., Kreuzer, K. N., Sherratt, D. J., Sandler, S. J., and Marians, K. J. (2000). The importance of repairing stalled replication forks. Nature 404, 37-41.

D'Aurelio, M., Gajewski, C. D., Lin, M. T., Mauck, W. M., Shao, L. Z., Lenaz, G., Moraes, C. T., and Manfredi, G. (2004). Heterologous mitochondrial DNA recombination in human cells. Hum Mol Genet 13, 3171-3179.

Dalakas, M. C., Leon-Monzon, M. E., Bernardini, I., Gahl, W. A., and Jay, C. A. (1994). Zidovudine-induced mitochondrial myopathy is associated with muscle carnitine deficiency and lipid storage. Ann Neurol 35, 482-487.

Daley, J. M., Laan, R. L., Suresh, A., and Wilson, T. E. (2005). DNA joint dependence of pol X family polymerase action in nonhomologous end joining. J Biol Chem 280, 29030- 29037.

Davies, A. A., Masson, J. Y., McIlwraith, M. J., Stasiak, A. Z., Stasiak, A., Venkitaraman, A. R., and West, S. C. (2001). Role of BRCA2 in control of the RAD51 recombination and DNA repair protein. Mol Cell 7, 273-282.

Davis, A. P., and Symington, L. S. (2004). RAD51-dependent break-induced replication in yeast. Mol Cell Biol 24, 2344-2351.

Dillingham, M. S., and Kowalczykowski, S. C. (2008). RecBCD enzyme and the repair of double-stranded DNA breaks. Microbiol Mol Biol Rev 72, 642-71, Table of Contents.

Dukanovic, J., and Rapaport, D. (2011). Multiple pathways in the integration of proteins into the mitochondrial outer membrane. Biochim Biophys Acta 1808, 971-980.

Essers, J., Houtsmuller, A. B., van Veelen, L., Paulusma, C., Nigg, A. L., Pastink, A., Vermeulen, W., Hoeijmakers, J. H., and Kanaar, R. (2002). Nuclear dynamics of RAD52 group homologous recombination proteins in response to DNA damage. EMBO J 21, 2030-2037.

Fekairi, S., Scaglione, S., Chahwan, C., Taylor, E. R., Tissier, A., Coulon, S., Dong, M. Q., Ruse, C., Yates, J. R. r., Russell, P., Fuchs, R. P., McGowan, C. H., and Gaillard, P. H. (2009). Human SLX4 is a Holliday junction resolvase subunit that binds multiple DNA repair/recombination endonucleases. Cell 138, 78-89.

Feng, J. Y., Johnson, A. A., Johnson, K. A., and Anderson, K. S. (2001). Insights into the molecular mechanism of mitochondrial toxicity by AIDS drugs. J Biol Chem 276, 23832- 23837.

Fisher, R. P., and Clayton, D. A. (1988). Purification and characterization of human mitochondrial transcription factor 1. Mol Cell Biol 8, 3496-3509.

Flores-Rozas, H., and Kolodner, R. D. (2000). Links between replication, recombination and genome instability in eukaryotes. Trends Biochem Sci 25, 196-200.

Flott, S., Kwon, Y., Pigli, Y. Z., Rice, P. A., Sung, P., and Jackson, S. P. (2011). Regulation of Rad51 function by phosphorylation. EMBO Rep 12, 833-839.

Flygare, J., Benson, F., and Hellgren, D. (1996). Expression of the human RAD51 gene during the cell cycle in primary human peripheral blood lymphocytes. Biochim Biophys Acta 1312, 231-236.

Forget, A. L., Bennett, B. T., and Knight, K. L. (2004). Xrcc3 is recruited to DNA double strand breaks early and independent of Rad51. J Cell Biochem 93, 429-436.

Frossi, B., Tell, G., Spessotto, P., Colombatti, A., Vitale, G., and Pucillo, C. (2002). H(2)O(2) induces translocation of APE/Ref-1 to mitochondria in the Raji B-cell line. J Cell Physiol 193, 180-186.

Gasior, S. L., Olivares, H., Ear, U., Hari, D. M., Weichselbaum, R., and Bishop, D. K. (2001). Assembly of RecA-like recombinases: distinct roles for mediator proteins in mitosis and meiosis. Proc Natl Acad Sci U S A 98, 8411-8418.

Gildemeister, O. S., Sage, J. M., and Knight, K. L. (2009). Cellular redistribution of Rad51 in response to DNA damage: novel role for Rad51C. J Biol Chem 284, 31945- 31952.

Gilkerson, R. W., Schon, E. A., Hernandez, E., and Davidson, M. M. (2008). Mitochondrial nucleoids maintain genetic autonomy but allow for functional complementation. J Cell Biol 181, 1117-1128.

Graziewicz, M. A., Longley, M. J., and Copeland, W. C. (2006). DNA polymerase gamma in mitochondrial DNA replication and repair. Chem Rev 106, 383-405.

Griffiths, L. M., Swartzlander, D., Meadows, K. L., Wilkinson, K. D., Corbett, A. H., and Doetsch, P. W. (2009). Dynamic compartmentalization of base excision repair proteins in response to nuclear and mitochondrial oxidative stress. Mol Cell Biol 29, 794-807.

Haber, J. E. (1999). DNA recombination: the replication connection. Trends Biochem Sci 24, 271-275.

Hanada, K., Budzowska, M., Davies, S. L., van Drunen, E., Onizawa, H., Beverloo, H. B., Maas, A., Essers, J., Hickson, I. D., and Kanaar, R. (2007). The structure-specific endonuclease Mus81 contributes to replication restart by generating double-strand DNA breaks. Nat Struct Mol Biol 14, 1096-1104.

Hartlerode, A., Odate, S., Shim, I., Brown, J., and Scully, R. (2011). Cell cycle- dependent induction of homologous recombination by a tightly regulated I-SceI fusion protein. PLoS One 6, e16501.

Helleday, T. (2003). Pathways for mitotic homologous recombination in mammalian cells. Mutat Res 532, 103-115.

Heller, R. C., and Marians, K. J. (2006). Replisome assembly and the direct restart of stalled replication forks. Nat Rev Mol Cell Biol 7, 932-943.

Henry-Mowatt, J., Jackson, D., Masson, J. Y., Johnson, P. A., Clements, P. M., Benson, F. E., Thompson, L. H., Takeda, S., West, S. C., and Caldecott, K. W. (2003). XRCC3 and Rad51 modulate replication fork progression on damaged vertebrate chromosomes. Mol Cell 11, 1109-1117.

Henson, S. E., Tsai, S. C., Malone, C. S., Soghomonian, S. V., Ouyang, Y., Wall, R., Marahrens, Y., and Teitell, M. A. (2006). Pir51, a Rad51-interacting protein with high expression in aggressive lymphoma, controls mitomycin C sensitivity and prevents chromosomal breaks. Mutat Res 601, 113-124.

Holt, I. J., Lorimer, H. E., and Jacobs, H. T. (2000). Coupled leading- and lagging-strand synthesis of mammalian mitochondrial DNA. Cell 100, 515-524.

100

Hori, A., Yoshida, M., Shibata, T., and Ling, F. (2009). Reactive oxygen species regulate DNA copy number in isolated yeast mitochondria by triggering recombination-mediated replication. Nucleic Acids Res 37, 749-761.

Hudson, G., and Chinnery, P. F. (2006). Mitochondrial DNA polymerase-gamma and human disease. Hum Mol Genet 15 Spec No 2, R244-52.

Ikeda, S., Biswas, T., Roy, R., Izumi, T., Boldogh, I., Kurosky, A., Sarker, A. H., Seki, S., and Mitra, S. (1998). Purification and characterization of human NTH1, a homolog of Escherichia coli endonuclease III. Direct identification of Lys-212 as the active nucleophilic residue. J Biol Chem 273, 21585-21593.

Ip, S. C., Rass, U., Blanco, M. G., Flynn, H. R., Skehel, J. M., and West, S. C. (2008). Identification of Holliday junction resolvases from humans and yeast. Nature 456, 357- 361.

Jeng, J. Y., Yeh, T. S., Lee, J. W., Lin, S. H., Fong, T. H., and Hsieh, R. H. (2008). Maintenance of mitochondrial DNA copy number and expression are essential for preservation of mitochondrial function and cell growth. J Cell Biochem 103, 347-357.

Jensen, R. B., Carreira, A., and Kowalczykowski, S. C. (2010). Purified human BRCA2 stimulates RAD51-mediated recombination. Nature 467, 678-683.

Kajander, O. A., Karhunen, P. J., Holt, I. J., and Jacobs, H. T. (2001). Prominent mitochondrial DNA recombination intermediates in human heart muscle. EMBO Rep 2, 1007-1012.

Kalifa, L., Beutner, G., Phadnis, N., Sheu, S. S., and Sia, E. A. (2009). Evidence for a role of FEN1 in maintaining mitochondrial DNA integrity. DNA Repair (Amst) 8, 1242- 1249.

Karow, J. K., Constantinou, A., Li, J. L., West, S. C., and Hickson, I. D. (2000). The Bloom's syndrome gene product promotes branch migration of holliday junctions. Proc Natl Acad Sci U S A 97, 6504-6508.

Korhonen, J. A., Gaspari, M., and Falkenberg, M. (2003). TWINKLE Has 5' -> 3' DNA helicase activity and is specifically stimulated by mitochondrial single-stranded DNA- binding protein. J Biol Chem 278, 48627-48632.

Korhonen, J. A., Pham, X. H., Pellegrini, M., and Falkenberg, M. (2004). Reconstitution of a minimal mtDNA replisome in vitro. EMBO J 23, 2423-2429.

Kowalczykowski, S. C., and Krupp, R. A. (1987). Effects of Escherichia coli SSB protein on the single-stranded DNA-dependent ATPase activity of Escherichia coli RecA protein. 101

Evidence that SSB protein facilitates the binding of RecA protein to regions of secondary structure within single-stranded DNA. J Mol Biol 193, 97-113.

Kraakman-van der Zwet, M., Overkamp, W. J., van Lange, R. E., Essers, J., van Duijn- Goedhart, A., Wiggers, I., Swaminathan, S., van Buul, P. P., Errami, A., Tan, R. T., Jaspers, N. G., Sharan, S. K., Kanaar, R., and Zdzienicka, M. Z. (2002). Brca2 (XRCC11) deficiency results in radioresistant DNA synthesis and a higher frequency of spontaneous deletions. Mol Cell Biol 22, 669-679.

Kraytsberg, Y., Schwartz, M., Brown, T. A., Ebralidse, K., Kunz, W. S., Clayton, D. A., Vissing, J., and Khrapko, K. (2004). Recombination of human mitochondrial DNA. Science 304, 981.

Krejci, L., Song, B., Bussen, W., Rothstein, R., Mortensen, U. H., and Sung, P. (2002). Interaction with Rad51 is indispensable for recombination mediator function of Rad52. J Biol Chem 277, 40132-40141.

Kreuzer, K. N. (2005). Interplay between DNA replication and recombination in prokaryotes. Annu Rev Microbiol 59, 43-67.

Langston, L. D., and O'Donnell, M. (2008). DNA polymerase delta is highly processive with proliferating cell nuclear antigen and undergoes collision release upon completing DNA. J Biol Chem 283, 29522-29531.

Larsson, N. G., Oldfors, A., Holme, E., and Clayton, D. A. (1994). Low levels of mitochondrial transcription factor A in mitochondrial DNA depletion. Biochem Biophys Res Commun 200, 1374-1381.

Lee, H. C., Lu, C. Y., Fahn, H. J., and Wei, Y. H. (1998). Aging- and smoking-associated alteration in the relative content of mitochondrial DNA in human lung. FEBS Lett 441, 292-296.

Lee, H. C., and Wei, Y. H. (2005). Mitochondrial biogenesis and mitochondrial DNA maintenance of mammalian cells under oxidative stress. Int J Biochem Cell Biol 37, 822- 834.

Lee, H. C., Yin, P. H., Chi, C. W., and Wei, Y. H. (2002). Increase in mitochondrial mass in human fibroblasts under oxidative stress and during replicative cell senescence. J Biomed Sci 9, 517-526.

Lee, H. C., Yin, P. H., Lu, C. Y., Chi, C. W., and Wei, Y. H. (2000). Increase of mitochondria and mitochondrial DNA in response to oxidative stress in human cells. Biochem J 348 Pt 2, 425-432.

102

Lee, Y. S., Lee, S., Demeler, B., Molineux, I. J., Johnson, K. A., and Yin, Y. W. (2010). Each monomer of the dimeric accessory protein for human mitochondrial DNA polymerase has a distinct role in conferring processivity. J Biol Chem 285, 1490-1499.

Li, X., and Heyer, W. D. (2008). Homologous recombination in DNA repair and DNA damage tolerance. Cell Res 18, 99-113.

Liu, N., and Lim, C. S. (2005). Differential roles of XRCC2 in homologous recombinational repair of stalled replication forks. J Cell Biochem 95, 942-954.

Liu, P., and Demple, B. (2010). DNA repair in mammalian mitochondria: Much more than we thought? Environ Mol Mutagen 51, 417-426.

Liu, P., Qian, L., Sung, J. S., de Souza-Pinto, N. C., Zheng, L., Bogenhagen, D. F., Bohr, V. A., Wilson, D. M. r., Shen, B., and Demple, B. (2008). Removal of oxidative DNA damage via FEN1-dependent long-patch base excision repair in human cell mitochondria. Mol Cell Biol 28, 4975-4987.

Longley, M. J., Clark, S., Yu Wai Man, C., Hudson, G., Durham, S. E., Taylor, R. W., Nightingale, S., Turnbull, D. M., Copeland, W. C., and Chinnery, P. F. (2006). Mutant POLG2 disrupts DNA polymerase gamma subunits and causes progressive external ophthalmoplegia. Am J Hum Genet 78, 1026-1034.

Lundin, C., Erixon, K., Arnaudeau, C., Schultz, N., Jenssen, D., Meuth, M., and Helleday, T. (2002). Different roles for nonhomologous end joining and homologous recombination following replication arrest in mammalian cells. Mol Cell Biol 22, 5869- 5878.

Lundin, C., Schultz, N., Arnaudeau, C., Mohindra, A., Hansen, L. T., and Helleday, T. (2003). RAD51 is involved in repair of damage associated with DNA replication in mammalian cells. J Mol Biol 328, 521-535.

Macaulay, V., Richards, M., and Sykes, B. (1999). Mitochondrial DNA recombination- no need to panic. Proc Biol Sci 266, 2037-9; discussion 2041-2.

Maloisel, L., Fabre, F., and Gangloff, S. (2008). DNA polymerase delta is preferentially recruited during homologous recombination to promote heteroduplex DNA extension. Mol Cell Biol 28, 1373-1382.

Mandel, H., Szargel, R., Labay, V., Elpeleg, O., Saada, A., Shalata, A., Anbinder, Y., Berkowitz, D., Hartman, C., Barak, M., Eriksson, S., and Cohen, N. (2001). The deoxyguanosine kinase gene is mutated in individuals with depleted hepatocerebral mitochondrial DNA. Nat Genet 29, 337-341.

103

Martin, J. L., Brown, C. E., Matthews-Davis, N., and Reardon, J. E. (1994). Effects of antiviral nucleoside analogs on human DNA polymerases and mitochondrial DNA synthesis. Antimicrob Agents Chemother 38, 2743-2749.

Mason, P. A., Matheson, E. C., Hall, A. G., and Lightowlers, R. N. (2003). Mismatch repair activity in mammalian mitochondria. Nucleic Acids Res 31, 1052-1058.

Masson, J. Y., Tarsounas, M. C., Stasiak, A. Z., Stasiak, A., Shah, R., McIlwraith, M. J., Benson, F. E., and West, S. C. (2001). Identification and purification of two distinct complexes containing the five RAD51 paralogs. Genes Dev 15, 3296-3307.

Masuda, T., Ling, F., Shibata, T., and Mikawa, T. (2010). Analysis of DNA-binding sites on Mhr1, a yeast mitochondrial ATP-independent homologous pairing protein. FEBS J 277, 1440-1452.

McGlynn, P., and Lloyd, R. G. (2002). Recombinational repair and restart of damaged replication forks. Nat Rev Mol Cell Biol 3, 859-870.

Mineri, R., Pavelka, N., Fernandez-Vizarra, E., Ricciardi-Castagnoli, P., Zeviani, M., and Tiranti, V. (2009). How do human cells react to the absence of mitochondrial DNA? PLoS One 4, e5713.

Mita, S., Monnat, R. J. J., and Loeb, L. A. (1988). Resistance of HeLa cell mitochondrial DNA to mutagenesis by chemical carcinogens. Cancer Res 48, 4578-4583.

Mladenov, E., Anachkova, B., and Tsaneva, I. (2006). Sub-nuclear localization of Rad51 in response to DNA damage. Genes Cells 11, 513-524.

Moraes, C. T. (2001). What regulates mitochondrial DNA copy number in animal cells? Trends Genet 17, 199-205.

Morten, K. J., Ashley, N., Wijburg, F., Hadzic, N., Parr, J., Jayawant, S., Adams, S., Bindoff, L., Bakker, H. D., Mieli-Vergani, G., Zeviani, M., and Poulton, J. (2007). Liver mtDNA content increases during development: a comparison of methods and the importance of age- and tissue-specific controls for the diagnosis of mtDNA depletion. Mitochondrion 7, 386-395.

Mortensen, U. H., Bendixen, C., Sunjevaric, I., and Rothstein, R. (1996). DNA strand annealing is promoted by the yeast Rad52 protein. Proc Natl Acad Sci U S A 93, 10729- 10734.

Munoz, I. M., Hain, K., Declais, A. C., Gardiner, M., Toh, G. W., Sanchez-Pulido, L., Heuckmann, J. M., Toth, R., Macartney, T., Eppink, B., Kanaar, R., Ponting, C. P., 104

Lilley, D. M., and Rouse, J. (2009). Coordination of structure-specific nucleases by human SLX4/BTBD12 is required for DNA repair. Mol Cell 35, 116-127.

Nagley, P., and Linnane, A. W. (1972). Cellular regulation of mitochondrial DNA synthesis in Saccharomyces cerevisiae. Cell Differ 1, 143-148.

Naviaux, R. K., and Nguyen, K. V. (2004). POLG mutations associated with Alpers' syndrome and mitochondrial DNA depletion. Ann Neurol 55, 706-712.

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 Dev 25, 350-362.

Nimonkar, A. V., Ozsoy, A. Z., Genschel, J., Modrich, P., and Kowalczykowski, S. C. (2008). Human exonuclease 1 and BLM helicase interact to resect DNA and initiate DNA repair. Proc Natl Acad Sci U S A 105, 16906-16911.

Pagliacci, M. C., Spinozzi, F., Migliorati, G., Fumi, G., Smacchia, M., Grignani, F., Riccardi, C., and Nicoletti, I. (1993). Genistein inhibits tumour cell growth in vitro but enhances mitochondrial reduction of tetrazolium salts: a further pitfall in the use of the MTT assay for evaluating cell growth and survival. Eur J Cancer 29A, 1573-1577.

Pagliarini, D. J., Calvo, S. E., Chang, B., Sheth, S. A., Vafai, S. B., Ong, S. E., Walford, G. A., Sugiana, C., Boneh, A., Chen, W. K., Hill, D. E., Vidal, M., Evans, J. G., Thorburn, D. R., Carr, S. A., and Mootha, V. K. (2008). A mitochondrial protein compendium elucidates complex I disease biology. Cell 134, 112-123.

Petermann, E., and Helleday, T. (2010). Pathways of mammalian replication fork restart. Nat Rev Mol Cell Biol 11, 683-687.

Petermann, E., Orta, M. L., Issaeva, N., Schultz, N., and Helleday, T. (2010). Hydroxyurea-stalled replication forks become progressively inactivated and require two different RAD51-mediated pathways for restart and repair. Mol Cell 37, 492-502. Pfanner, N. (2000). Protein sorting: recognizing mitochondrial presequences. Curr Biol 10, R412-5.

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 Dev 13, 2633- 2638.

Pohjoismaki, J. L., and Goffart, S. (2011). Of circles, forks and humanity: Topological organisation and replication of mammalian mitochondrial DNA. Bioessays 33, 290-299.

105

Pohjoismaki, J. L., Goffart, S., Taylor, R. W., Turnbull, D. M., Suomalainen, A., Jacobs, H. T., and Karhunen, P. J. (2010). Developmental and pathological changes in the human cardiac muscle mitochondrial DNA organization, replication and copy number. PLoS One 5, e10426.

Pohjoismaki, J. L., Goffart, S., Tyynismaa, H., Willcox, S., Ide, T., Kang, D., Suomalainen, A., Karhunen, P. J., Griffith, J. D., Holt, I. J., and Jacobs, H. T. (2009). Human heart mitochondrial DNA is organized in complex catenated networks containing abundant four-way junctions and replication forks. J Biol Chem 284, 21446-21457.

Postow, L., Ullsperger, C., Keller, R. W., Bustamante, C., Vologodskii, A. V., and Cozzarelli, N. R. (2001). Positive torsional strain causes the formation of a four-way junction at replication forks. J Biol Chem 276, 2790-2796.

Pursell, Z. F., Isoz, I., Lundstrom, E. B., Johansson, E., and Kunkel, T. A. (2007). Yeast DNA polymerase epsilon participates in leading-strand DNA replication. Science 317, 127-130.

Rahman, S., and Poulton, J. (2009). Diagnosis of mitochondrial DNA depletion syndromes. Arch Dis Child 94, 3-5.

Ralf, C., Hickson, I. D., and Wu, L. (2006). The Bloom's syndrome helicase can promote the regression of a model replication fork. J Biol Chem 281, 22839-22846.

Rassool, F. V., North, P. S., Mufti, G. J., and Hickson, I. D. (2003). Constitutive DNA damage is linked to DNA replication abnormalities in Bloom's syndrome cells. Oncogene 22, 8749-8757.

Reenan, R. A., and Kolodner, R. D. (1992a). Characterization of insertion mutations in the Saccharomyces cerevisiae MSH1 and MSH2 genes: evidence for separate mitochondrial and nuclear functions. Genetics 132, 975-985.

Reenan, R. A., and Kolodner, R. D. (1992b). Isolation and characterization of two Saccharomyces cerevisiae genes encoding homologs of the bacterial HexA and MutS mismatch repair proteins. Genetics 132, 963-973.

Rijkers, T., Van Den Ouweland, J., Morolli, B., Rolink, A. G., Baarends, W. M., Van Sloun, P. P., Lohman, P. H., and Pastink, A. (1998). Targeted inactivation of mouse RAD52 reduces homologous recombination but not resistance to ionizing radiation. Mol Cell Biol 18, 6423-6429.

Ristic, D., Modesti, M., van der Heijden, T., van Noort, J., Dekker, C., Kanaar, R., and Wyman, C. (2005). Human Rad51 filaments on double- and single-stranded DNA: 106

correlating regular and irregular forms with recombination function. Nucleic Acids Res 33, 3292-3302.

Rothkamm, K., Kruger, I., Thompson, L. H., and Lobrich, M. (2003). Pathways of DNA double-strand break repair during the mammalian cell cycle. Mol Cell Biol 23, 5706- 5715.

Saada, A., Shaag, A., Mandel, H., Nevo, Y., Eriksson, S., and Elpeleg, O. (2001). Mutant mitochondrial thymidine kinase in mitochondrial DNA depletion myopathy. Nat Genet 29, 342-344.

Sage, J. M., Gildemeister, O. S., and Knight, K. L. (2010). Discovery of a novel function for human Rad51: maintenance of the mitochondrial genome. J Biol Chem 285, 18984- 18990.

Salazar, J. J., and Van Houten, B. (1997). Preferential mitochondrial DNA injury caused by glucose oxidase as a steady generator of hydrogen peroxide in human fibroblasts. Mutat Res 385, 139-149.

Santos, J. H., Mandavilli, B. S., and Van Houten, B. (2002). Measuring oxidative mtDNA damage and repair using quantitative PCR. Methods Mol Biol 197, 159-176.

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.

Sarzi, E., Goffart, S., Serre, V., Chretien, D., Slama, A., Munnich, A., Spelbrink, J. N., and Rotig, A. (2007). Twinkle helicase (PEO1) gene mutation causes mitochondrial DNA depletion. Ann Neurol 62, 579-587.

Schmidt, O., Pfanner, N., and Meisinger, C. (2010). Mitochondrial protein import: from proteomics to functional mechanisms. Nat Rev Mol Cell Biol 11, 655-667.

Seong, C., Sehorn, M. G., Plate, I., Shi, I., Song, B., Chi, P., Mortensen, U., Sung, P., and Krejci, L. (2008). Molecular anatomy of the recombination mediator function of Saccharomyces cerevisiae Rad52. J Biol Chem 283, 12166-12174.

Shah, R., Cosstick, R., and West, S. C. (1997). The RuvC protein dimer resolves Holliday junctions by a dual incision mechanism that involves base-specific contacts. EMBO J 16, 1464-1472.

Shedge, V., Arrieta-Montiel, M., Christensen, A. C., and Mackenzie, S. A. (2007). Plant mitochondrial recombination surveillance requires unusual RecA and MutS homologs. Plant Cell 19, 1251-1264. 107

Sheridan, S. D., Yu, X., Roth, R., Heuser, J. E., Sehorn, M. G., Sung, P., Egelman, E. H., and Bishop, D. K. (2008). A comparative analysis of Dmc1 and Rad51 nucleoprotein filaments. Nucleic Acids Res 36, 4057-4066.

Shokolenko, I., Venediktova, N., Bochkareva, A., Wilson, G. L., and Alexeyev, M. F. (2009). Oxidative stress induces degradation of mitochondrial DNA. Nucleic Acids Res 37, 2539-2548.

Sonoda, E., Morrison, C., Yamashita, Y. M., Takata, M., and Takeda, S. (2001). Reverse genetic studies of homologous DNA recombination using the chicken B-lymphocyte line, DT40. Philos Trans R Soc Lond B Biol Sci 356, 111-117.

Spies, M., Amitani, I., Baskin, R. J., and Kowalczykowski, S. C. (2007). RecBCD enzyme switches lead motor subunits in response to chi recognition. Cell 131, 694-705.

Stuart, J. A., and Brown, M. F. (2006). Mitochondrial DNA maintenance and bioenergetics. Biochim Biophys Acta 1757, 79-89.

Sutovsky, P., Moreno, R. D., Ramalho-Santos, J., Dominko, T., Simerly, C., and Schatten, G. (2000). Ubiquitinated sperm mitochondria, selective proteolysis, and the regulation of mitochondrial inheritance in mammalian embryos. Biol Reprod 63, 582- 590.

Szczesny, B., Tann, A. W., Longley, M. J., Copeland, W. C., and Mitra, S. (2008). Long patch base excision repair in mammalian mitochondrial genomes. J Biol Chem 283, 26349-26356.

Takata, M., Sasaki, M. S., Sonoda, E., Fukushima, T., Morrison, C., Albala, J. S., Swagemakers, S. M., Kanaar, R., Thompson, L. H., and Takeda, S. (2000). The Rad51 paralog Rad51B promotes homologous recombinational repair. Mol Cell Biol 20, 6476- 6482.

Takata, M., Sasaki, M. S., Tachiiri, S., Fukushima, T., Sonoda, E., Schild, D., Thompson, L. H., and Takeda, S. (2001). Chromosome instability and defective recombinational repair in knockout mutants of the five Rad51 paralogs. Mol Cell Biol 21, 2858-2866.

Tang, Y., Schon, E. A., Wilichowski, E., Vazquez-Memije, M. E., Davidson, E., and King, M. P. (2000). Rearrangements of human mitochondrial DNA (mtDNA): new insights into the regulation of mtDNA copy number and gene expression. Mol Biol Cell 11, 1471-1485.

Tarsounas, M., Davies, D., and West, S. C. (2003). BRCA2-dependent and independent formation of RAD51 nuclear foci. Oncogene 22, 1115-1123. 108

Tembe, V., and Henderson, B. R. (2007). Protein trafficking in response to DNA damage. Cell Signal 19, 1113-1120.

Thyagarajan, B., Padua, R. A., and Campbell, C. (1996). Mammalian mitochondria possess homologous DNA recombination activity. J Biol Chem 271, 27536-27543.

Trzeciak, A. R., Nyaga, S. G., Jaruga, P., Lohani, A., Dizdaroglu, M., and Evans, M. K. (2004). Cellular repair of oxidatively induced DNA base lesions is defective in prostate cancer cell lines, PC-3 and DU-145. Carcinogenesis 25, 1359-1370.

Tuduri, S., Tourriere, H., and Pasero, P. (2010). Defining replication origin efficiency using DNA fiber assays. Chromosome Res 18, 91-102.

Van Houten, B., Woshner, V., and Santos, J. H. (2006). Role of mitochondrial DNA in toxic responses to oxidative stress. DNA Repair (Amst) 5, 145-152. van Mameren, J., Modesti, M., Kanaar, R., Wyman, C., Peterman, E. J., and Wuite, G. J. (2009). Counting RAD51 proteins disassembling from nucleoprotein filaments under tension. Nature 457, 745-748.

Voelkel-Meiman, K., and Roeder, G. S. (1990). Gene conversion tracts stimulated by HOT1-promoted transcription are long and continuous. Genetics 126, 851-867.

Vu, T. H., Sciacco, M., Tanji, K., Nichter, C., Bonilla, E., Chatkupt, S., Maertens, P., Shanske, S., Mendell, J., Koenigsberger, M. R., Sharer, L., Schon, E. A., DiMauro, S., and DeVivo, D. C. (1998). Clinical manifestations of mitochondrial DNA depletion. Neurology 50, 1783-1790.

Wai, T., Ao, A., Zhang, X., Cyr, D., Dufort, D., and Shoubridge, E. A. (2010). The role of mitochondrial DNA copy number in mammalian fertility. Biol Reprod 83, 52-62.

Wang, Y., and Bogenhagen, D. F. (2006). Human mitochondrial DNA nucleoids are linked to protein folding machinery and metabolic enzymes at the mitochondrial inner membrane. J Biol Chem 281, 25791-25802.

WATSON, J. D., and CRICK, F. H. (1953). Molecular structure of nucleic acids; a structure for deoxyribose nucleic acid. Nature 171, 737-738.

Wei, Y. H., Lee, C. F., Lee, H. C., Ma, Y. S., Wang, C. W., Lu, C. Y., and Pang, C. Y. (2001). Increases of mitochondrial mass and mitochondrial genome in association with enhanced oxidative stress in human cells harboring 4,977 BP-deleted mitochondrial DNA. Ann N Y Acad Sci 928, 97-112.

109

Yakes, F. M., and Van Houten, B. (1997). Mitochondrial DNA damage is more extensive and persists longer than nuclear DNA damage in human cells following oxidative stress. Proc Natl Acad Sci U S A 94, 514-519.

Yang, M. Y., Bowmaker, M., Reyes, A., Vergani, L., Angeli, P., Gringeri, E., Jacobs, H. T., and Holt, I. J. (2002). Biased incorporation of ribonucleotides on the mitochondrial L- strand accounts for apparent strand-asymmetric DNA replication. Cell 111, 495-505.

Yasukawa, T., Reyes, A., Cluett, T. J., Yang, M. Y., Bowmaker, M., Jacobs, H. T., and Holt, I. J. (2006). Replication of vertebrate mitochondrial DNA entails transient ribonucleotide incorporation throughout the lagging strand. EMBO J 25, 5358-5371.

Yasukawa, T., Yang, M. Y., Jacobs, H. T., and Holt, I. J. (2005). A bidirectional origin of replication maps to the major noncoding region of human mitochondrial DNA. Mol Cell 18, 651-662.

Yoshida, Y., Izumi, H., Ise, T., Uramoto, H., Torigoe, T., Ishiguchi, H., Murakami, T., Tanabe, M., Nakayama, Y., Itoh, H., Kasai, H., and Kohno, K. (2002). Human mitochondrial transcription factor A binds preferentially to oxidatively damaged DNA. Biochem Biophys Res Commun 295, 945-951.

Zheng, L., and Shen, B. (2011). Okazaki fragment maturation: nucleases take centre stage. J Mol Cell Biol 3, 23-30.

Zheng, L., Zhou, M., Guo, Z., Lu, H., Qian, L., Dai, H., Qiu, J., Yakubovskaya, E., Bogenhagen, D. F., Demple, B., and Shen, B. (2008). Human DNA2 is a mitochondrial nuclease/helicase for efficient processing of DNA replication and repair intermediates. Mol Cell 32, 325-336.

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