UHRF1 and the DNA damage response
by
Helena Sumantrai Mistry
A thesis submitted in conformity with the requirements For the Degree of Doctor of Philosophy Graduate Department of Pharmacology & Toxicology University of Toronto
©Copyright by Helena Sumantrai Mistry 2011
UHRF1 and the DNA damage response Helena Sumantrai Mistry Doctor of Philosophy, 2011 Graduate Department of Pharmacology & Toxicology University of Toronto
THESIS ABSTRACT
Our DNA is under constant threat from endogenous and exogenous damaging agents.
Our cells have evolved a network of signaling pathways and repair mechanisms that detect and counteract this threat, collectively referred to as the DNA damage response. Cells that lose the ability to cope with DNA damage risk the acquisition of deleterious changes to DNA sequence or structure.
I initially set out to identify and characterize candidate proteins that interact with Mus81-
Eme1, an endonuclease that processes DNA intermediates that arise from aberrant or stalled
DNA replication. I focused on one interesting candidate protein known as Nuclear protein 95
(Np95) which now is called UHRF1 (Ubiquitin-like, containing PHD and RING finger domains,
1). Although previous studies demonstrate the importance of Mus81-Eme1 enzyme in DNA repair, genome integrity, and tumor suppression, little is known about how the enzyme acts together with other components of signaling pathways that comprise the DNA damage response.
My findings in chapter two characterized this interaction and linked Mus81-Eme1 with UHRF1 in the cellular response to DNA damage. Although UHRF1 levels have been linked with sensitivity to antineoplastic agents, a direct role for UHRF1 in the DNA damage response had not been elucidated or reported. Accordingly, the third chapter of my thesis focuses on investigating the role of UHRF1 in the cellular response to DNA damage caused by exposure to
γ-irradiation. Our findings for chapter three indicate that (i) UHRF1 is crucial for the cellular
ii response to double strand breaks caused by γ-irradiation and that (ii) UHRF1 is critical for maintenance of chromosome integrity.
Recent studies have now implicated UHRF1 in processes required for heterochromatin replication. This protein has been shown to play a role in the replication of heterochromatin by helping to replicate DNA methylation patterns and playing a role in propagating the epigenetic mark known as histone 3 lysine 9 trimethylation (H3K9me3). H3K9me3 has been shown to play a role in a signaling pathway involved in the repair of DNA damage in heterochromatic regions.
In the fourth chapter of my thesis, we hypothesize that UHRF1 is playing a role in a pathway that responds to DSB damage in heterochromatic regions of chromatin. Our results indicate that a loss of UHRF1 results in a loss of heterochromatic H3K9me3 and heterochromatin associated
HP1β. Our findings support the idea that epigenetic alterations maintained by UHRF1 contribute to signals that relax heterochromatin to facilitate access for repair factors. In summary, findings presented in this thesis shed light on processes that protect cells from DNA damage caused by radiation and chemotherapy and safeguard genome integrity.
iii
ACKNOWLEDGEMENTS
I would like to thank my supervisor Dr. Peter McPherson for giving me the opportunity to pursue research in his laboratory. I am grateful for his mentorship and belief in my ability as a scientist. His leadership skills, enthusiasm and encouragement will forever be admired and never be forgotten. His dedication to the laboratory and ability to put down whatever he is working on, and listen to my daily experimental update is greatly appreciated. I would also like to thank Dr.
Patricia Harper and Dr. Jason Matthews for their constructive comments during my committee meetings. I owe a special thanks to Dr. Jane Mitchell and Dr. Denis Grant for the career advice and encouragement along the way.
I am grateful to past and present members of the laboratory: Haya Sarras, Brenda Yun,
Loni Gibson, Solmaz Alizadeh, Meghan Larin, Hussein Butt, Aileen Gracias and Daniel
Sisgoreo. A special thanks to Laura Tamblyn for the help with FACS analysis and mentorship throughout the program.
I would like to also thank my friends in the department for their encouragement and support. A special mention goes to Rosalia Yoon, Aman Mann, Shuang Wang (Sammi), Lick
Lai, Ewa Hoffman and Craig Hayden for all their support.
My hard work and experimental dedication does not compare to the encouragement and consistent support I have received from my mother. During the good times and the tougher times, she has always supported me emotionally and financially. Words cannot express my gratitude towards her never ending support, and guidance in making my dream a reality. I know I have been very difficult at times, and the patience she has displayed as well as her love and
iv support have allowed me to achieve my career goals. Dear mother, I could not have done this without you.
v
TABLE OF CONTENTS
THESIS ABSTRACT ...... ii ACKNOWLEDGEMENTS ...... iv TABLE OF CONTENTS ...... vi LIST OF PUBLICATIONS ...... x LIST OF ORAL AND POSTER PRESENTATIONS ...... x ABBREVIATIONS ...... xii LIST OF TABLES ...... xiv LIST OF FIGURES ...... xv CHAPTER ONE: GENERAL INTRODUCTION ...... 1 I.1 Importance of Genome Integrity...... 1 I.2 DNA damage- characteristic lesions and their sources ...... 3 I.2.1 Endogenous sources ...... 3 I.2.2 Environmental Radiation...... 5 I.2.3 Antineoplastic agents ...... 5 I.3 The DNA damage response ...... 6 I.3.2 Surveillance and signalling in response to DNA double-strand break damage ...... 9 I.4 Cellular events triggered by the DNA damage response ...... 11 I.4.1 Cell cycle arrest and senescence ...... 11 1.4.2 Apoptosis ...... 12 I. 4.3 DNA repair mechanisms ...... 14 I.5 Consequences of impaired DNA damage response and loss of genome integrity ...... 23 I.5.1 Diseases linked with defective responses to DNA damage ...... 24 1.5.2 Defective responses to DNA damage and neoplastic transformation ...... 24 I.5.3 Aging ...... 26 I.6 Role of Mus81- Eme1 endonuclease in the DNA damage response ...... 26 I.6.1 Mus81 and Genome Instability ...... 30
vi
I.7 The DNA damage response and its relationship to chromatin ...... 31 I.7.1 The heterogeneous nature of mammalian chromatin ...... 31 1.7.2 Chromatin dynamics/remodelling ...... 33 I.7.3 Post-translational modifications of histones that demarcate chromatin states ...... 34 I.7.4 Histone Acetylation ...... 34 I.7.5 Histone Lysine Methylation ...... 35 I.7.6 Ubiquitination of Histones ...... 37 I.7.7 Histone Variant Incorporation ...... 37 I.7.8 DNA methylation ...... 39 I.8 Chromatin Accessibility and DNA repair ...... 40 1.8.1 Unmasking of H3K9me3 facilitates activation of Tip60 Histone Acetyltransferase during the DNA damage response ...... 40 I.9 UHRF1 (Ubiquitin-like, containing PHD and RING finger domains 1) ...... 44 I.9.1 UHRF1 is a proliferation-associated ubiquitin ligase ...... 44 I.9.2 UHRF1 and the replication of heterochromatin...... 45 I.9.3 UHRF1 facilitates transfer of DNA methylation during DNA replication ...... 45 I.9.4 UHRF1 and maintenance of the H3K9me3 mark in heterochromatin ...... 46 I.9.5 UHRF1 roles in sensitivity to DNA damage and cancer ...... 47 Aims of my thesis research ...... 49 CHAPTER TWO: Interaction between Eme1 and UHRF1 ...... 52 Introduction ...... 53 Materials and Methods ...... 54 Eme1-Interacting Protein Screen ...... 54 GST Pulldown Assay...... 55 Creation of Eme1 Mutants ...... 56 Generation of vector constructs and cell lines for cell culture ...... 58 Transient Transfections ...... 59 Immunoprecipitation and Western Blotting ...... 59 Indirect Immunofluorescence ...... 60 Results ...... 61 Discussion...... 69
vii
Significance and Impact ...... 74 CHAPTER THREE: UHRF1 facilitates the DNA damage response to γ-Irradiation...... 76 Introduction ...... 77 Methods ...... 78 Cell lines ...... 78 Growth Curve ...... 79 Western Blotting ...... 79 Clonogenic assays...... 80 Flow cytometry (FACS analysis) ...... 80 Karyotype analysis ...... 81 Results ...... 82 Discussion...... 92 Significance and Impact ...... 96 CHAPTER FOUR: UHRF1, heterochromatin and the DNA damage response ...... 97 Introduction: ...... 98 Methods: ...... 100 Indirect Immunofluorescence ...... 100 Western Blotting to detect phospho-ATM ...... 101 Results ...... 101 Discussion...... 108 Significance and Impact: ...... 115 CHAPTER FIVE: GENERAL CONCLUSIONS AND FUTURE DIRECTIONS ...... 117 Eme1 interacts with UHRF1 ...... 118 UHRF1‟s role in double strand breaks caused by gamma-irradiation ...... 122 Does UHRF1‟s have a role in the chromosomal maintenance? ...... 128 UHRF1 is a contributor in a signalling cascade that repairs DNA damage in heterochromatin 129 Clinical Implications in Cancer ...... 136 Conclusion ...... 138 Future Studies ...... 138 Querying the possible role of Mus81-Eme1 in heterochromatin repair ...... 139 Role of chromatin accessibility in relation to γH2AX formation and amplification ...... 141
viii
Does restoring the ability of UHRF1 to maintain H3K9me3 on heterochromatin rescue: (a) chromosomal instability and (b) sensitivity to irradiation? ...... 143 REFERENCES ...... 145
ix
LIST OF PUBLICATIONS
Helena Mistry*, Laura Tamblyn*, Hussein Butt, Daniel Sisgoreo, Aileen Gracias, Meghan Larin, Kalpana Gopalakrishnan, Manoor Prakash Hande and John Peter McPherson. UHRF1 is a genome caretaker that facilitates the DNA damage response to γ-irradiation. Genome Integrity (BMC) 2010 Jun 8;1(1):7.
Helena Mistry*, Lianne Gibson*, Ji Weon Yun, Haya Sarras, Laura Tamblyn, John Peter McPherson. Interplay between Np95 and Eme1 in the DNA damage response. Biochem Biophys Res Commun. 2008 Oct 24;375(3):321-5.
*These authors contributed equally to the work
LIST OF ORAL AND POSTER PRESENTATIONS
Helena Mistry, Laura Tamblyn, Hussein Butt, Daniel Sisgoreo, Aileen Gracias, and John Peter McPherson. Maintenance of H3K9me3 by UHRF1 impacts the DNA damage response. Histone Code: Fact or Fictions? (Keystone symposia) (2011). January 10-15, Midway,UT
Helena Mistry, Laura Tamblyn, Hussein Butt, Daniel Sisgoreo, Aileen Gracias, Meghan Larin and John Peter McPherson. UHRF1 facilitates the DNA damage response to γ-irradiation. Experimental Biology (FASEB) (2010). April 24-28, Anaheim,CA
Helena Mistry, Laura Tamblyn, Hussein Butt, Daniel Sisgoreo, Aileen Gracias, and John Peter McPherson. Np95 is a crucial puzzle piece in genome integrity maintenance. Puzzles in Biology 2009, European Molecular Biology Laboratory (EMBL) Heidelberg, Germany. (Invited Oral Presentation).
Helena Mistry, Laura Tamblyn and John Peter McPherson. A Role for Np95 in Genome Integrity Maintenance. Experimental Biology (FASEB) (2009). April 18-22, New Orleans, LA
Helena Mistry, Lianne Gibson, Ji Weon Yun, Haya Sarras, Laura Tamblyn and Peter McPherson. Interplay between Np95 and Eme1 in the DNA damage response. August 2008. Presented to the Developmental Human Physiology- University of Toronto/Karolinska Institute (Sweden) Exchange.
x
Helena Mistry, Lianne Gibson, Ji Weon Yun, Haya Sarras, Laura Tamblyn and Peter McPherson (2007). Interplay between Mus81-Eme1 Endonuclease and Nuclear Protein 95 (Np95): Role in the DNA damage response. Patterns in Biology: Organization of Life in Space and Time 2007, European Molecular Biology Laboratory (EMBL) Heidelberg, Germany. Programs and Proceedings: (Invited Poster Presentation).
Helena Mistry, Ji Weon Yun, Haya Sarras, Laura Tamblyn and Peter McPherson (2007). A Novel Interaction Between Nuclear Protein 95 and the Mus81/Eme1 Endonuclease. Proc. Am. Assoc. Cancer Res. 2007, April 14-18, Los Angeles, CA . Programs and Proceedings: Abstract 1098.
xi
ABBREVIATIONS
γH2AX histone variant H2AX phosphorylated on serine 139 µg microgram µL microlitre µM micromolar AP apurinic/apyrimidinic site Ara-C cytosine arabinoside AT ataxia telangiectasia ATM ataxia telangiectasia mutated ATR ATM (ataxia telangiectasia mutated)- Rad3-related BER base excision repair BLM Bloom Syndrome BRCA1 breast cancer susceptibility gene 1 BRCA2 breast cancer susceptibility gene 2 CD chromodomain cDNA complementary DNA CHK1 checkpoint kinase 1 CHK2 checkpoint kinase 2 CK2 casein kinase 2 CPT camptothecin CS catalytic subunit CSD chromoshadow domain DAPI 4‟,6‟-diamidino-2-phenylindole DNA Deoxyribonucleic acid DNA-PK DNA-dependent protein kinase DNMT DNA methyltransferase DSB double-strand break EDTA Ethylenediaminetetraacetic acid EGTA Ethylene glycol-bis(β-aminoethyl ether) N,N,N‟,N‟,-tetraacetic acid Eme1 essential meiotic endonuclease 1 ES embryonic stem FA Fanconi anemia FITC Fluorescein isothiocyanante G1 gap 1 G2 gap 2 GST glutathione-S-transferase Gy gray H histone H3K9me3 histone 3 lysine 9 trimethylation H4K20me3 histone 4 lysine 20 trimethylation HAT histone acetyl transferase HDAC histone deacetylase
xii
HP1 heterochromatin protein 1 HR homologous recombination ICL interstrand crosslink ICPB90 Inverted CCAAT box-binding protein of 90 kDa IF immunofluorescence IP immunoprecipitation IR ionizing radiation K lysine KAP-1 KRAB-ZFP–associated protein 1 KRAB Krüppel-associated box M phase mitosis phase mM millimolar MMR mismatch repair Mms4 methyl methanesulfonate-sensitive 4 Mre11 meiotic-recombination protein-11 MRN MRE11 (meiotic recombination 11 homologue), RAD50 and NBS1 (Nijmegen breakage syndrome 1) MSH MutS homologues Mus81 methyl methanesulfonate and UV-sensitive clone 81 NBS1 Nijmegen-breakage-syndrome-1 NER nucleotide excision repair NHEJ non-homologous end joining NP95 nuclear protein 95 NM nanometer ORF open reading frame PAGE polyacrylamide gel electrophoresis PBS phosphate-buffered saline PCNA proliferating cell nuclear antigen PHD plant homeodomain PIK phospho-inositide kinase PTM post-translational modifications ROS reactive oxygen species RPA replication protein A S synthesis SDS Sodium dodecyl sulphate SRA SET and Ring-associated domain SSB single-strand break TBS Tris-buffered saline TRITC tetramethylrhodamine isothiocyanante UBL ubiquitin like domain UV ultraviolet WCL whole cell lysate XP Xeroderma pigmentosum YTH yeast two-hybrid
xiii
LIST OF TABLES
Table 1 Sequences of Primers and Restriction Enzyme Sites used to create Eme1 mutants, page 52.
Table 2 Antibodies used for Immunoprecipitation and Western analysis, page 56.
Table 3 Frequency of chromosomal aberrations in UHRF1-depleted cells, page 75.
xiv
LIST OF FIGURES
Figure 1 DNA damage repair pathways, page 2.
Figure 2 DNA damage response signalling, page 8.
Figure 3 Nuclear focus formation of the phosphorylation of serine 139 on histone H2AX following DNA damage, page 10.
Figure 4 The cell cycle and checkpoints, page 12.
Figure 5 DNA replication fork, page 19.
Figure 6 Interstrand crosslink formation, 20.
Figure 7 Interstrand crosslink and ability to cause replication fork damage, page 21.
Figure 8 The Fanconi Anemia pathway, page 23.
Figure 9 Holliday junction and blocked replication fork, page 28.
Figure 10 Role for Mus81-Eme1 in the restart of replication forks, page 29.
Figure 11 The nucleosome, page 32.
Figure 12 The ubiquitination cascade, page 38.
Figure 13 Model for ATM activation, page 43.
Figure 14 Model for UHRF1, page 49.
Figure 15 Interaction between Eme1 and UHRF1 as shown by yeast two-hybrid, page 57.
Figure 16 Amino acids 40-186 of Eme1 are sufficient for the interaction with UHRF1, page 62.
Figure 17 Confirmation of UHRF1-Eme1 interaction by GST-fusion pull-down assay, page 65.
Figure 18 Interaction between UHRF1 and Eme1 in mammalian cells by co- immunoprecipitation, page 66.
xv
Figure 19 Eme1 co-localization with UHRF1 following CPT exposure, page 68.
Figure 20 UHRF1 reduces steady- state levels of Eme1, page 73.
Figure 21 Characterization of UHRF1-depleted HeLa cells, page 83.
Figure 22 Impact of UHRF1 loss on proliferation, page 84.
Figure 23 Susceptibility of UHRF1-depleted cells to γ-irradiation, page 85.
Figure 24 Susceptibility of UHRF1-depleted cells to Ara-C and CPT, page 86.
Figure 25 Cell cycle analysis of UHRF1-depleted cells following γ-irradiation, page 88.
Figure 26 Decreased irradiation-induced γH2AX in UHRF1-depleted cells, page 90.
Figure 27 Chromosomal instability in UHRF1-depleted cells, page 91.
Figure 28 Model of signal transactions that facilitate chromatin relaxation in heterochromatin following DNA damage, page 99.
Figure 29 H3K9me3 levels remain the same in control and UHRF1-depleted cells following DNA damage, page 103.
Figure 30 Representative indirect immunofluorescence of H3K9me3 and H4K20me3 in control and UHRF1-depleted cells, page 104.
Figure 31 Rescue of H3K9me3 status in UHRF1-depleted cells, page 105.
Figure 32 Chromatin-bound Hp1α and Hp1β in UHRF1-depleted cells exposed to irradiation, page 106.
Figure 33 Chromatin-bound Hp1α and Hp1β in UHRF1-depleted cells exposed to CPT, page 108.
Figure 34 Induction of acetyl-H2AK5 following DNA damage is defective in UHRF1-depleted cells, page 109.
Figure 35 Defective ATM activation following DNA damage in UHRF1-depleted cells, page 110.
Figure 36 UHRF1‟s has multiple roles in heterochromatic DNA replication, page 123.
xvi
Figure 37 UHRF1 and Mus81-Eme1‟s role in DNA damage within heterochromatin, page 141.
xvii
CHAPTER ONE: GENERAL INTRODUCTION
I.1 Importance of Genome Integrity.
DNA is our recipe for life – it contains all of the information needed for the regulation and reproduction of all of the cells in our body. Our ability to maintain accurate long term storage of this information and pass on our genetic heritage from generation to generation depends on the accurate replication and transmission of DNA from parental cell to daughter cells. This accurate transmission is continually threatened by radiation and chemical agents that originate from endogenous and exogenous sources (reviewed in (Kastan and Bartek, 2004)).
Various sources of DNA damage contribute to the generation of significant numbers of
DNA lesions throughout our lifespan (figure 1). During our lifetime, an average of 100 trillion cells, each with 3 x 109 bases of DNA, will go through 1016 cell divisions (Trosko and Chang,
2010). DNA damage from either intrinsic or extrinsic sources occurs at a rate of 1,000 to 1,000
000 lesions per cell per day [3]. Even in the absence of damage stimuli, DNA is inherently susceptible to chemical decomposition in that approximately 18,000 purine residues per genome are lost daily as a result of spontaneous hydrolysis, whereas spontaneous deamination of cytosine to uracil occurs 100-500 times per genome per day (Errol C. Friedberg, 2005). Accordingly, our cells have evolved safeguards that protect DNA and rectify lesions when they occur. These safeguards are regarded as sophisticated “molecular circuitry” that detect, signal and repair DNA damage. This „molecular circuitry‟ is commonly referred to as the DNA Damage Response (van
Gent et al., 2001). The various forms of DNA damage that we encounter can trigger a variety of responses based on the nature of the damage. One such response is the utilization of
“checkpoints”, which are a collection of surveillance and response pathways that halts
1
progression through cell cycle stages in response to DNA damage to prevent the conversion of lesions into permanent mutations. In certain situation, cell with damaged genomes forego DNA repair and conduct „cellular proofreading‟ by eliminating themselves via apoptosis (programmed cell death) or remove themselves from the pool of proliferating cells by entering a state of permanent cell cycle withdrawal known as senescence (Kastan and Bartek, 2004). The surveillance, signalling and repair factors involved in these processes are numerous: in humans, well over 150 DNA repair proteins have been identified (Wood et al., 2005). The proofreading capacity of this response system is remarkable in that it allows on average only three base pairing mistakes when copying three billion base pairs in the human genome (Errol C. Friedberg, 2005).
I.2 DNA damage- characteristic lesions and their sources
I.2.1 Endogenous sources Endogenous sources of DNA damage originate from a variety of metabolic processes that generate by-products that can react with macromolecules such as lipids, proteins or nucleic acids
(Friedberg, 2006). Endogenous DNA damage is typically caused by three classes of chemical reactions: (i) oxidation, (ii) alkylation and (iii) hydrolysis.
Oxidation: Molecular transactions involving oxygen and oxygen-containing compounds
can contribute to DNA damage through the generation of reactive oxygen species (ROS),
compounds with unpaired valence shell electrons that react with DNA and other cellular
molecules (Errol C. Friedberg, 2005). ROS such as superoxide, hydrogen peroxide,
hydroxyl radicals and singlet oxygen are generated as natural by-products of
mitochondrial respiration and cytochrome P450-mediated biotransformation of
xenobiotics such as phenobarbitol and 2,3,7,8-tetrachlorodibenzo-p-dioxin (Choi et al.,
3
2008; Denda et al., 1995). To combat ROS-mediated oxidative damage, defence mechanisms such as maintenance of a sharp gradient in oxygen tension from the environment (20%) to tissues (0.5-5%), as well a battery of antioxidant enzymes have evolved in our cells, tissues, and organs (De Bont and van Larebeke, 2004; Errol C.
Friedberg, 2005). Defects in these protective measures can precipitate ROS accumulation (oxidative stress) and, thus, increase the frequency of mutations and chromosomal aberrations (Brown and Bicknell, 2001).
(ii) Alkylation: Alkylating agents are typically biotransformed into electrophilic compounds that readily form chemical adducts with biological molecules with a nucleophilic center such as DNA. Exposure to alkylating compounds occurs both endogenously and exogenously. Endogenously, DNA can, for instance, react inappropriately with S-adenosylmethionine, a methyl group donor that participates in the methylation of a subset of cytosines. The products of this reaction, 7-methylguanine and
3-methyladenine, are made at a rate of several hundred per day per mammalian haploid genome and must be counteracted given their ability to block replication (De Bont and van Larebeke, 2004). Exogenous sources of damage include compounds found in tobacco smoke (i.e. N-nitrosonornicotine) (Atamna et al., 2000).
(iii) Hydrolysis: The glycosidic bond of DNA, located between the base and deoxyribose sugar, is prone to spontaneous hydrolysis, which leads to the formation of an abasic site
(AP site, where AP stands for “apurinic” or “apyrimidinic”). An AP site is a location in the DNA that does not have a purine or a pyrimidine base. DNA is also naturally susceptible to hydrolytic deamination of cytosines and 5-methylcytosines, with approximately 100 to 500 cytosines undergoing deamination to uracil per cell/per day
4
(Lindahl, 1990). Cells counteract this decomposition through a pathway that senses and
replaces altered bases (uracil DNA-glycosylase and base excision repair (BER)) (De Bont
and van Larebeke, 2004; Friedberg, 2006; Rideout et al., 1990).
I.2.2 Environmental Radiation Exposure to ionizing radiation (IR) or ultraviolet (UV) radiation is an unavoidable and constant environmental risk. Each year, we are exposed to about 1 to 2 milli sieverts of IR from cosmic radiation or radioactive decay of compounds in soil (Errol C. Friedberg, 2005). The predominant form of DNA damage generated by IR is scission of the sugar-phosphate backbone
(strand break damage), with 1 gray (Gy) of IR capable of generating 600 to 1000 strand breaks/cell (Prise et al., 2001; Rydberg, 1996; Vilenchik and Knudson, 2003). UV radiation is typically categorized into three forms: UV-A, -B or –C, where UV-B (295-320 nm) and UV-C
(100-295 nm) are responsible for nearly all UV-induced DNA damage. Unlike IR mediated damage, the most predominant type of damage that occurs is the formation of two major bulky adducts between adjacent pyrimidines known as cyclobutane pyrimidine dimers and pyrimidine- pyrimidine 6-4 photoproducts (Errol C. Friedberg, 2005).
I.2.3 Antineoplastic agents The observation that mustard gas exposure during World War I led to reduced white cell counts led to the eventual utilization of nitrogen mustard and other genotoxic compounds in the treatment of lymphoma and ushered in the modern use of chemotherapy in the treatment of cancer. Eighteen of the twenty most important antineoplastic agents in use today are known to mediate their cytotoxic effect through DNA damage in cancer cells. These agents can be classified into four categories: (i) alkylating agents, (ii) antimetabolites, (iii) topoisomerase inhibitors, (iv) and antimicrotubule agents (Lichtman, 2008).
5
(i) Alkylating agents such as carmustine, cyclophosphamide and melphalan, damage DNA through the chemical transfer of one or more alkyl groups to DNA deoxyribose or phosphate groups leading to either strand scission, or to the production of DNA bases that alter Watson-
Crick base pairing preferences.
(ii) Antimetabolites such as methotrexate, pemetrexed and cytosine arabinoside (Ara-C), generate replication-associated DNA damage by compromising the synthesis of nucleotides needed for DNA replication fork progression.
(iii) Topoisomerase inhibitors such as camptothecin (CPT), topotecan, etoposide and doxorubicin covalently bond and thus inhibit the catalytic action of topoisomerase I or topoisomerase II, which are enzymes that regulate DNA topology by breaking and rejoining the phosphodiester backbone of DNA during DNA replication, transcription, recombination and repair.
(iv) Antimicrotubule agents such as taxanes, vinblastine and vincristine interfere with the synthesis or degradation of microtubules required for the proper segregation of chromosomes during mitosis.
I.3 The DNA damage response
The cellular DNA damage response safeguards genome integrity. It is coded to rectify damage to the organism by either programmed cell death, or by permanent withdrawal from the cell cycle if damage is too extensive for repair (Ljungman, 2010). Proteins that participate in these processes can be considered as sensors, transducers, or effectors. Sensors perform surveillance and signalling duties, in that they initially detect abnormalities in DNA and emit signals that recruit other signalling and repair factors to the damaged lesion. Transducers
6 amplify the signalling cascade initiated by sensor proteins and communicate the DNA damage signal throughout the cells. Many sensors and transducers exert their action as protein kinases and were originally identified as regulators of cell cycle checkpoints. When DNA damage occurs, signalling from sensors and transducers holds the cell in a given phase and inhibits transition to the next phase until the damage is rectified. Mediator proteins, such as MRE11, acquire post-translational modifications generated by sensor proteins. These mediator proteins are then able to amplify and transducer the signals to effector proteins which are downstream.
Sensors and transducers both participate in the activation and recruitment of effector proteins, which serve to conclude the DNA damage response, either by participation in the repair process or through transcriptional regulation of processes involved in cell cycle progression, apoptosis, and senescence (Houtgraaf et al., 2006).
Several response pathways are initiated by phospho-inositide kinase (PIK)-related proteins, which include sensors such as ataxia-telangiectasia mutated (ATM), DNA-dependent protein kinase (DNA-PK) and ataxia-telangiectasia-and-RAD3-related (ATR). Damage leads to activation of these sensors in the presence of Ku, ATRIP and Nijmegen breakage syndrome 1
(NBS1) co-factors, respectively. ATM is primarily activated by DSB damage and associated changes in chromatin density. ATR is activated by damage that impedes DNA replication, particularly replication protein A (RPA)-coated tracts of single-stranded DNA.
Upon activation, PIK-related kinases phosphorylate and activate the transducer kinases
Chk1 and Chk2, where ATM primarily activates Chk2, and ATR primarily activates Chk1 (Zhou and Elledge, 2000) (figure 2). Both Chk1 and Chk2 activate cell cycle checkpoints by phosphorylating members of a protein phosphatase family known as Cdc25 (Sancar et al., 2004; van Gent et al., 2001). Phosphorylation of Cdc25 targets it for degradation and, consequently,
7
cyclin-dependent kinases normally activated by Cdc25-mediated dephosphorylation remain inactive and cell cycle progression is halted (Helt et al., 2005; Sancar et al., 2004). Chk1 and
Chk2, together with PIK-related kinases, also activate and recruit p53, a transcription factor that has predominant roles in the transcriptional activation of genes associated with cell cycle arrest,
DNA repair and apoptosis; and thus, is considered to be a major effector (Batinac et al., 2003;
Errol C. Friedberg, 2005; Levine, 1997).
In addition to p53, Chk1, ChK2, and PIK-related kinases activate and recruit various other proteins to sites of DNA damage (figure 2). Although many of these proteins have only recently been discovered and knowledge of their role in the DNA damage response is only superficial, it is generally believed that these proteins facilitate reorganization of the chromatin or DNA in the vicinity of the lesion to facilitate repair.
I.3.2 Surveillance and signalling in response to DNA double-strand break damage DNA damage in the form of double stranded breaks (DSBs) is one of the most cytotoxic forms of damage. DSBs are the hallmark lesion that results from exposure to IR and radiomimetic drugs. They are also caused by agents that damage DNA through impediment of
DNA replication fork progression (CPT, Ara-C) and DNA cross-linking agents (mitomycin C, cisplatin) (Dery and Masson, 2007). DSBs are recognized by DNA damage sensor proteins that activate the recruitment of DNA repair proteins into chromatin at the site of damage. Repair proteins can be visualized as subcellular nuclear dots known as „foci‟ using indirect immunofluorescence. The accumulation of repair factors as discrete foci is believed to represent discrete repair complexes assembled on sites of DNA DSBs (figure 3) (Dery and Masson,
2007). In the presence of a DSB, a heterotrimer protein complex known as the MRN complex
(meiotic recombination 11 homologue (MRE11), RAD50 and NBS1), recognizes the damaged
9
DNA and rapidly locates to DSB sites and forms nuclear foci. It binds to the end of the DSB and activates a protein kinase known as ATM. ATM kinase then auto-phosphorylates and becomes activated. This results in phosphorylation of serine 139 on H2AX , a histone H2A variant that constitutes 10% of nuclear H2A in mammals (Bonner et al., 2008; Mah et al., 2010).
Phosphorylation of H2AX (γH2AX) is an early and rapid event that occurs in less than 5 minutes of a DSB; and it is extensive, covering large regions of the chromosome. Amplification of
γH2AX is believed to facilitate further recruitment of more DNA repair proteins to the site of damage (Dery and Masson, 2007).
I.4 Cellular events triggered by the DNA damage response
I.4.1 Cell cycle arrest and senescence The cell cycle is the ordered sequence of events needed to generate two daughter cells.
The stages of the cell cycle are: gap (G)1, synthesis (S), G2 and mitosis (M). The G1 phase of the cell cycle is the main period when the cell experiences growth. It is during this phase that new cellular organelles are synthesized, resulting in high protein synthesis and increased an cellular metabolic rate (Boye et al., 2009). During this phase, external signals instruct the cell whether or not to prepare for DNA synthesis which occurs in the S phase. By the end of the S phase, DNA is duplicated in preparation for subsequent chromosome segregation and cell division. During the G2 phase, the cell prepares for cell division which occurs in M phase. In the
M phase, the cell segregates newly duplicated chromosomes into two identical sets. There also exists one phase outside of the cell cycle, known as G0, during which the cell neither divides, nor prepares to divide (Santoro and Blandino, 2010).
Cell cycle checkpoints aim to ensure fidelity of cell division and reduce the number of mutations that are passed onto daughter cells. The cell has three major cell cycle checkpoints that
11 can be triggered in response to DNA damage: (i) G1/S, (ii) intra-S, and (iii) G2/M checkpoints
(figure 4)
(Errol C. Friedberg, 2005; Flatt and Pietenpol, 2000; Malumbres and Barbacid, 2009).
Activation of these checkpoints halts further events from occurring in the cell cycle stage to facilitate repair of the DNA lesion or abnormality. Upon completion of lesion repair, progression through the cell cycle resumes, thereby ensuring that DNA lesions are not duplicated and segregated during S and M phases. Failure of the cell to complete repair before replication or chromosomal segregation can lead to permanent, irretrievable chromosomal damage. Under certain conditions, cell cycle checkpoints and DNA damage signalling result in an irreversible withdrawal from the cell cycle known as senescence. Senescence is considered to be a growth- arrest mechanism that limits the replicative lifespan of diploid cells (Kerr et al., 1994; Ljungman,
2010).
1.4.2 Apoptosis Apoptosis is also referred to as programmed cell death. Programmed cell death is used to describe processes that metazoan organisms use to eliminate unwanted or potentially deleterious cells during various stages of development or lifespan. Such processes can involve numerous signalling pathways, ultimately leading to the activation of a class of proteases known as caspases that kill and dismantle the cell. This process of cellular demolition facilitates the effective clearance of apoptotic cells by phagocytic cells and manifests as shrinkage in cell volume, DNA fragmentation and loss of plasma membrane integrity (Elmore, 2007). Under certain conditions, DNA damage signalling can initiate cells to undergo apoptosis, usually via activation of p53 which can direct transcriptional activation of pro-apoptotic proteins. This p53- dependent cell death pathway is believed to contribute to the cytocidal effects of many
12
antineoplastic agents that cause DNA damage. In some cases, protection from apoptosis has been linked to resistance to antineoplastic agents by cancer cells (Robertson et al., 2000;
Sinkovics, 1991).
I. 4.3 DNA repair mechanisms Five major DNA repair pathways exist within the DNA damage response: (i) direct repair, (ii) BER, (iii) nucleotide excision repair (NER), (iv) mismatch repair (MMR), and (v) strand break repair, which is comprised of two repair pathways: non-homologous end joining
(NHEJ) and homologous recombination (HR). Although each of these repair pathways target specific types of damage, in certain cases, cells use a combination of pathways to fix complex forms of damage such as crosslinking damage that can impede replication forks.
Direct Repair: A class of cancer drugs known as alkylating agents causes DNA damage by forming covalent bonds, or “adducts" with DNA bases. Examples of these types of drugs include: temozolomide, streptozotocin, procarbazine and dacarbazine, which are commonly used to treat gliomas, melanoma and Hodgkin‟s lymphoma (Verbeek et al., 2008). One of the principal mutagenic forms of damage caused by alkylating agents is the formation of an O6- methylguanine adduct. The modified base can pair with thymine in place of cytosine, thus giving rise to mutation of the cell‟s genetic code. This lesion is repaired by O6-alkylguanine DNA alkyltransferase also known as O-methylguanine-DNA-methyltranferase. This dealkylation reaction step then leads to an irreversible inactivation and degradation of O-methylguanine-
DNA-methyltranferase (Errol C. Friedberg, 2005).
Base Excision Repair: BER is considered to be the primary repair pathway that resolves DNA damage due to cellular metabolism or spontaneous degradation of DNA. It removes bases that
14 have undergone various inappropriate chemical modifications such as methylation, oxidation and deamination.
BER acts as a „cut and patch‟ type of repair system, where damaged DNA is removed from one strand of DNA and the resulting gap is filled in using the complimentary strand as a template
(Errol C. Friedberg, 2005). Specifically, BER is initiated when DNA glycosylases recognize and remove DNA damage by catalyzing hydrolytic cleavage of the C1‟-N-glycosyl group off the base, which results in an AP site. This AP site is then replaced with the correct nucleotide through the coordinated action of APE1 endonuclease, DNA polymerases and ligases (Chan et al., 2006; Errol C. Friedberg, 2005; Hoeijmakers, 2001) .
Nucleotide Excision Repair: NER is a very versatile pathway that removes various chemical adducts as well as lesions caused by UV irradiation and various xenobiotics. This pathway is particularly important in the removal of bulky helix distorting adducts. The system involves two major pathways: global genome repair and transcription-coupled repair. In global genome repair
–NER, DNA lesions are recognized by a protein complex known as XPC-HR23B that initiates repair events. In transcription-coupled repair -NER pathway, RNA polymerase blocking lesions are removed from the transcribed strand of active genes (de Boer and Hoeijmakers, 2000). The transcription-coupled repair -NER pathway is activated when damaged DNA blocks RNA polymerase II. This is followed by the removal or displacement of the polymerase for damage repair. This arrest is followed by the recruitment of a complex that contains two proteins called
CSA and CSB.
The subsequent steps in both global genome repair-NER and transcription-coupled repair
-NER pathways involve XPB and XPD, two helicases which are part of the multi-subunit
15 transcription factor TFIIH. A single-stranded binding protein known as RPA, coats the undamaged strand and protects single-stranded DNA, whereas the endonucleases XPG and
ERCC1/XPF act as molecular scissors that cleave the 3‟ and 5‟ ends of the damaged strand, respectively. Following removal of the damaged strand, the newly formed gap is filled by either
DNA polymerase ε or δ, and the resulting nick is sealed by DNA ligase I (Benhamou and
Sarasin, 2000; Gillet and Scharer, 2006; Wood, 2010).
Mismatch Repair: The MMR pathway repairs insertions, deletions and base mismatches caused by DNA polymerase errors during DNA replication that escape proofreading (Iyer et al.,
2006). Loss of MMR activity leads to an increase in genome-wide point mutations and problems with replication in areas within the genome that commonly represent repeated-sequence motifs referred to as microsatellite DNA. The mechanism by which MMR proteins cooperate to facilitate repair is unclear in mammals.
In Escherichia coli, mismatch repair genes are known as MutS and MutL have been conserved throughout evolution (Acharya et al., 2003). Five MutS homologues (MSH) have been identified in eukaryotes that function as two major heterodimers: MutSα (MSH2 and MSH6 heterodimer) and MutSβ (MSH2 and MSH3 heterodimer). It has been proposed that MutSα recruits MutSβ and than releases it from the mismatch site, as well as facilitates loading of EXO1 nuclease, which degrades the mismatched DNA strand. Polymerase δ then fills the gap and the nick is sealed by DNA ligase 1 (Errol C. Friedberg, 2005; Jascur and Boland, 2006; Jiricny,
2006).
Strand Break Repair: The sugar-phosphate backbone of DNA can be cleaved to form a single strand break (SSB) or DSBs. SSBs can arise from chemical modifications to DNA phosphate
16 groups that disrupts phosphodiester bonds, such as it occurs with alkylating agents, or directly from damage by ROS (Caldecott, 2007). DSBs can occur from naturally occurring DNA transactions, from DNA backbone damage due to exposure to radiation or chemicals (alkylation or oxidative damage), from perturbed DNA replication forks, or from damage occurring during
DNA segregation at the anaphase. Estimated to occur at a rate of 10 breaks per cell/per day
(Lieber, 2010), DSBs are believed to be one of the most severe forms of DNA damage as they have the potential to alter coding information in the genome. There are two major pathways that resolve DSBs: NHEJ and HR.
NHEJ is referred to as „non-homologous‟ because unlike HR, where the sister chromatid is used as a template to resolve DSBs, the ends of the DNA strands where a DSB occurred are directly ligated back together without the use of a template to guide repair. This process is considered to be error prone (Lieber and Wilson, 2010; Valerie and Povirk, 2003).
The NHEJ pathway is activated in the presence of DSBs, with the tethering of MRN and
Ku/DNA-PK complexes to each end of the break. Recruit of the Ku/DNA–PK complex to the site of the DSB is followed by the recruitment of a ligase complex resulting in alignment and joining of DSB ends (Lieber and Wilson, 2010).
HR repair is preferable when a sister chromatid is available to be used as a template to facilitate repair. This usually occurs during the S and G2 phases of the cell cycle. During the HR pathway, strand invasion into homologous neighbouring sequences that can be used as templates occurs. When a DSB occurs, MRN complex is recruited to the 3‟ ends of the damaged sites and the 5‟ ends undergo degradation for HR repair to occur. This is followed by recruitment of RPA, which in turn facilitates the assembly of the Rad51 nucleoprotein filament. Rad51 has a role in
17 protecting the ends of the strand from becoming frayed and it interacts with Rad52, which has roles in the preparation for strand exchange and in the assembly of the filament. In the next step, an identical sister chromatid for use as a template and strand invasion occurs for DNA synthesis and consequent healing of the broken ends. The last step involves the resolution of the recombination intermediates (i.e. mobile junctions that occur between strands of DNA known as
Holliday junctions) (Camerini-Otero and Hsieh, 1995; Ray and Langer, 2002; San Filippo et al.,
2008)
Repair of replication-associated DNA damage DNA is replicated through the formation of a structure known as a „replication fork.‟
Replication forks are formed when complementary strands of DNA unwind and synthesis occurs in a 5‟ to 3‟ direction in the leading strand, and by means of Okasaki fragments in the lagging strand (figure 5). Exposure to various exogenous and endogenous agents can result in DNA damage that may stall the progression of replication forks. Unresolved (stalled) replication forks can lead to the acquisition of mutations and result in chromosomal damage (Kastan and Bartek,
2004). One form of damage that impedes replication forks is a lesion associated with the genotoxic action of many cancer drugs known as a crosslink. Crosslinks damage DNA when agents react with, and covalently bond to two different positions in the DNA. Crosslinks that occur on the same strand are referred to as intrastrand crosslinks, while those that occur in opposite strands of DNA are referred to as interstrand crosslinks (ICL). Crosslinks can also occur between DNA and proteins (figure 6). ICLs prevent the separation of the opposing strands of
DNA during replication and thus block replication and can also block DNA transcription (figure
7). Agents that cause interstrand crosslink include mitomycin C, platinum compounds, psoralens and bifunctional alkylating agents (i.e. Nitrogen mustards ) (Noll et al., 2006).
18
Repair of interstrand cross-links involves the Fanconi anemia pathway The Fanconi anemia (FA) pathway is a pathway that the cell uses to resolve crosslinks.
Defects in this pathway can result a rare genetic disorder that is characterized by congenital abnormalities, progressive bone marrow failure and cancer predisposition. Cells from patients with FA are hypersensitive to DNA crosslinking agents (Kennedy and D'Andrea, 2005). FA has thirteen complementation groups that have been identified, and the corresponding proteins have been shown to cooperate in a common DNA damage response pathway activated upon DNA crosslink formation or other lesions that compromise DNA replication (Kennedy and D'Andrea,
2005). Connections between other DNA damage response pathways such as proteins encoded by the familial breast cancer susceptibility genes BRCA1 (breast cancer type 1 susceptibility protein) and BRCA2 (breast cancer type 2 susceptibility protein) and the FA pathway have been discovered. Following replication-associated damage, a complex consisting of several of the FA proteins (FANC A/B/C/E/F/G/L/M) known as the core complex re-localizes from the cytoplasm to the nucleus of the cell and assembles onto chromatin (Mi and Kupfer, 2005). Once the core complex is assembled, it activates the E3 ubiquitin ligase activity of FANCL, resulting in monoubiquitination of FANCD2 which then interacts with the BRCA1/BRCA2 complex at the sites of replication fork damage (figure 8) (Kobayashi et al., 2008). Another enzyme that participates in repair of replication fork damage is the endonuclease heterodimer Mus81 (methyl methanesulfonate and UV sensitive clone 81), together with its binding partner, Eme1 (essential meiotic endonuclease 1; also known as Mms4 in budding yeast), helps resolve DNA intermediates that occur following replication fork damage. Given that both Mus81-Eme1 and the FA pathway have been shown to have a role in the repair of ICL‟s, Mus81-Eme1 may function with members of the FA pathway in the resolution of this type of damage. Whether there is relationship between Mus81-Eme1 and the FA pathway, however, is not clear. 22
I.5 Consequences of impaired DNA damage response and loss of genome integrity
I.5.1 Diseases linked with defective responses to DNA damage In 1968, James Cleaver found that cells from patients with the hereditary disease xeroderma pigmentosum (XP) had mutations in proteins that inactivated NER. He was the first to link defective DNA repair and cancer (Cleaver, 1968). Deficiency in the NER pathway can result in two rare autosomal-recessive syndromes: XP and Cockayne syndrome (Leibeling et al.,
2006). XP patients develop skin cancers during childhood, while Cockayne syndrome patients exhibit dwarfism and neurological abnormalities (Leibeling et al., 2006). Deficiency in MMR pathways can lead to hereditary nonpolyposis colon cancer (Turcot‟s syndrome). Deficiency in
ICL repair leads to FA, which is an autosomal recessive cancer susceptibility syndrome characterized by multiple congenital abnormalities, progressive bone marrow failure and hypersensitivity to DNA crosslinking agents. Understanding the defective mechanisms responsible for rare chromosome instability disorders such as those mentioned above provide valuable insight into understanding the function of DNA repair pathways, and their roles in cancer in the general population (Kennedy and D'Andrea, 2006). For example, heterozygous mutations of FA genes have been associated with breast and ovarian cancer susceptibility
(Kennedy and D'Andrea, 2006).
1.5.2 Defective responses to DNA damage and neoplastic transformation Cancer is a term used to refer to a collection of diseases sharing a deregulation of cellular proliferation together with inappropriate invasion of tumour cells into neighbouring tissues and colonization of other locations within the body (metastasis). The initiation and progression of various stages of the neoplastic transformation process is now known to be based on the sequential acquisition of key mutations and/or chromosomal alterations that convert proto-
24 oncogenes to oncogenes and inactivate tumour suppressor genes (Hanahan and Weinberg, 2000;
Kalant, 2007). Most human tumours display some form of genome instability. Signs of genome instability may include an increased inclination to accumulate point mutations, chromosomal rearrangements, gene amplification or changes in chromosomal number (aneuploidy). These changes can result in loss of control over normal cellular signalling that govern proliferation and various metabolic processes by affecting genes that have roles in growth regulations (Hanahan and Weinberg, 2000).
Tumour suppressor genes are considered protective genes that govern proliferation, differentiation and cell death by regulating cell cycle checkpoints and control apoptosis. In the presence of DNA damage, some tumour suppressor proteins promote genome stability by repairing or eliminating cells with genomes that are compromised. Accordingly, mutations of tumour suppressor genes can lead to damaged cells escaping cell cycle checkpoints and DNA repair. Such tumour suppressor proteins can be categorized as either gatekeepers or caretakers
(Kinzler and Vogelstein, 1997). Gatekeepers directly monitor cell cycle division processes and cell death. Examples of gatekeepers include p53 and retinoblastoma protein. Retinoblastoma protein acts as a gatekeeper by repressing the expression of genes that are required for DNA replication and cell cycle division (Harbour and Dean, 2000). Retinoblastoma protein inactivation in early life caused by mutations in both retinoblastoma alleles results in a form of eye tumours seen in children known as retinoblastoma (Lai et al., 1997). Unlike gatekeepers, caretaker tumour suppressors decrease the likelihood of neoplastic transformation indirectly, by ensuring maintenance of overall genome integrity. For example, ATM functions as a caretaker tumour suppressor. Individuals afflicted with ataxia-telangiectasia are defective in ATM activity and all have compromised DSB repair. The likelihood of afflicted individuals to develop cancer
25 is markedly higher than the general population, but not all people suffering from ataxia- telangiectasia will develop cancer (Jeggo et al., 1998). The loss of caretaker function during the process of neoplastic transformation is believed to increase the chances of incurring other mutations in other tumour suppressor genes and proto-oncogenes.
I.5.3 Aging Progeroid syndromes are a class of genetic disorders characterized by clinical features that imitate physiological aging at an early age (Navarro et al., 2006). In fact, the word progeria originates from the Greek word progeros and means „prematurely old‟ (Burtner and Kennedy,
2010). The incidence of progeroid syndromes are low and has been estimated to be 1 per 4-8 million births (Capell and Collins, 2006; Coutinho et al., 2009). Certain mutations in various proteins that participate in the cellular response to DNA damage are now known to be responsible for progeroid syndromes. For example, defects in the RecQ helicases WRN, BLM, and RECQ4 have also been identified in patients who have progeroid syndromes known as
Werner, Bloom and Rothmund-Thomson syndromes, respectively. RecQ helicases are proteins that have a role in the restart of DNA replication in yeast (Kaliraman et al., 2001).
I.6 Role of Mus81- Eme1 endonuclease in the DNA damage response
Mus81 was identified in two independent studies through the use of two-hybrid analysis.
Boddy and colleagues discovered in S. cerevisiae that Mus81 interacts with Cds1 (homolog of
Chk2 in mammals) (Boddy et al., 2000), whereas Interthal and colleagues using S. pombe found an interaction between Mus81 and Rad54, when Rad54 was used as bait (Interthal and Heyer,
2000). Cds1 has a role in the stabilization of stalled replication forks and regulation of several
DNA repair proteins (Boddy et al., 2000), while Rad54 belongs to the Rad52 epistasis group and has a role in the several repair pathways including the repair of DSBs by HR (Schmuckli-Maurer
26 et al., 2003). The interaction of Mus81 with Cds1 and Rad54 suggested a role for Mus81 in the
DNA damage response. A role for Mus81 in the restart of stalled replication was suggested in a subsequent study using S. cerevisiae which found that Mus81 was required for viability in the absence of Sgs1, a RecQ helicase (Kaliraman et al., 2001). This suggestion was confirmed when it was found that cells with Mus81 depletion were more sensitive to methyl methanesulfonate, hydroxyurea, mitomycin C and CPT (agents that stall replication) but not irradiation (which causes DSBs), compared to control cells (Boddy et al., 2000; Interthal and Heyer, 2000).
Examination of Mus81 domains revealed that Mus81 contains two helix-hairpin-helix domains
(which bind DNA) at opposing ends of the protein and an XPF endonuclease homology domain.
XPF endonuclease plays a role in HR-repair of DNA crosslink damage and NER (Ciccia et al.,
2008; Interthal and Heyer, 2000).
The endonuclease activity of Mus81 requires a protein partner known as Eme1 (Boddy et al., 2001; Ciccia et al., 2003). Studies in yeast defective for either Mus81 or Eme1 led to claims that Mus81-Eme1 endonuclease might be responsible for resolving Holliday junctions, which are branched intermediate structures believed to represent DNA transition states that arise during HR
(Boddy et al., 2001). However, subsequent in vitro cleavage assays demonstrated that Mus81-
Eme1‟s preferred substrates are in fact Holliday junctions but rather substrates that mimic replication fork structures (Boddy et al., 2001; Doe et al., 2002; Whitby et al., 2003).
Accordingly, the endonuclease activity of Mus81-Eme1 might be recruited during repair of replication-associated DNA (figure 9 and 10) (Heyer, 2004; Whitby, 2004).
Mus81 activity has also been linked to the RecQ helicase BLM. BLM helicase plays an important role in the repair of stalled replication forks by unwinding DNA for repair protein access and to facilitate dynamic changes in DNA structure that occur during repair steps
27
(Chakraverty and Hickson, 1999; Hickson, 2003; Zhang et al., 2005). Although Mus81 deficiency does not affect BLM helicase activity, loss of BLM leads to a decrease in Mus81
DNA binding to 3‟ flaps and Holliday junctions. Therefore, BLM might increase the endonuclease activity of Mus81 by enhancing the ability of this endonuclease to bind to DNA substrate in vitro (Zhang et al., 2005). This suggests that BLM facilitates Mus81 activity during processing and restoration of stalled replication forks (Zhang et al., 2005).
I.6.1 Mus81 and Genome Instability Cells deficient in Mus81 or Eme1 display an increased predisposition to chromosomal instability, with cells showing increased incidences of chromosomal fusions, dicentrics, and aneuploidy. Eme1-/- and Mus81-/- embryonic stem (ES) cells were shown to be particularly hypersensitive to antineoplastic agents; cisplatin and mitomycin C that cause crosslinks, and
DNA lesions that block DNA replication (Abraham et al., 2003) (McPherson et al., 2004).
These findings demonstrated a role for mammalian Mus81-Eme1 in repair of replication- associated damage and as a genome caretaker (Abraham et al., 2003; Dendouga et al., 2005;
Hiyama et al., 2006). In order to investigate the role of mammalian Mus81 in vivo, knock out studies were done by generating Mus81-deficient mice (McPherson et al., 2004). Mus81 deficient mice were viable, indicating that Mus81 has a nonessential role in growth and development (McPherson et al., 2004). Mus81-/- mice were also found to be highly sensitive to mitomycin C compared to wild-type mice. (McPherson et al., 2004). Interestingly, Mus81 was found to act as a tumour suppressor in mice, in that Mus81+/- and Mus81-/- mice showed a marked predisposition to cancer, particularly T and B cell lymphomas and various other cancers
(McPherson et al., 2004). These findings have been disputed by another group that created
Mus81-/- mice using a different strategy and found that although their mice were more sensitive
30 to DNA cross-linking agents, these Mus81-/- mice did not show a predisposition to cancer
(Dendouga et al., 2005). Hiyama and colleagues subsequently demonstrated that gene deletion of MUS81 in a human cell line promoted spontaneous induction of chromosomal abnormalities which led to the activation of intra-S phase and G2-M phase checkpoints (Hiyama et al., 2006).
A precise role for Mus81-Eme1 in replication-associated repair was recently shown by
Hanada and colleagues who determined that Mus81 cleaves replication fork intermediates. This cleavage event creates a free DSB that facilitates repair by HR. They found that in the absence of Mus81, replication fork recovery was inhibited and this led to the presence of chromosomal aberrations. These findings were particularly important as they suggest that Mus81-Eme1 has a role in maintaining genome integrity in mammals by converting DNA structures that arise due to stalled replication fork‟s into structures that are easier to repair (figure 10) (Hanada et al., 2007;
Hanada et al., 2006).
I.7 The DNA damage response and its relationship to chromatin
I.7.1 The heterogeneous nature of mammalian chromatin Eukaryotic cells store and organize DNA in the form of chromatin. The fundamental unit of chromatin is the nucleosome, which is comprised of 147 base pairs of DNA wrapped in a left- handed 1.67 superhelical turn around a histone (H) octamer. The histone octamer consists of 2 copies of each of the following H proteins: H2A, H2B, H3 and H4 (figure 11). Nucleosomes are joined by a linker histone known as H1, and further organized into higher levels of packing, known as solenoid. Also, a variety of non-histone proteins are known to function in regulation of higher-order chromatin organization. The degree of chromatin compaction varies throughout the genome. In regions of the genome that are gene-rich, continual or frequent access to DNA may be required and chromatin is maintained in a relatively accessible form known as
31
euchromatin. Other regions of the genome that are gene-poor or transcriptionally silent are maintained in a highly compacted form of chromatin known as heterochromatin (Grewal and
Rice, 2004). Maintenance of highly condensed, tightly packed nature of heterochromatin is essential for its role in the organization of centromeres and telomeres.
1.7.2 Chromatin dynamics/remodelling Chromatin not only organizes genetic information, but also coordinates its accessibility by the cell in a cell-lineage specific manner (Santos-Rosa and Caldas, 2005). In order to manage the desired control of genetic information the cell has developed four mechanisms to maintain its organization: 1) ATP-dependent chromatin remodelling, 2) post-translational histone modifications, 3) histone variant incorporation and 4) DNA methylation. These mechanisms work together to coordinate a network that controls gene expression (Kouzarides, 2007).
The purpose of chromatin remodelling activity is to expose DNA within the nucleosome that is not normally accessible or deeply embedded. Remodelling can also unfold chromatin in the vicinity of damaged DNA that requires access to repair factors. Chromatin remodelling is a collection of ATP-dependent processes performed by multi-subunit protein complexes that facilitate: 1) nucleosomal sliding, which changes the position of DNA in relation to the nucleosome, 2) nucleosomal displacement, which results in areas that are free of nucleosomes, and 3) spacing adjustment of nucleosomes, which adjust the position of histone‟s in relation to
DNA (Downs et al., 2007). Examples of chromatin remodelling complexes include
SWItch/Sucrose NonFermentable and remodel the structure of chromatin (Bernstein and Hake,
2006).
33
I.7.3 Post-translational modifications of histones that demarcate chromatin states Histones have an amino-terminal tail domain that protrudes from the nucleosomal core.
Histone tails provide a site for post-translational modifications that are covalent and reversible
(Mendez-Acuna et al., 2010). These modifications or chromatin „marks‟ include methylation, acetylation and phosphorylation. These marks act as types of landing pads for other proteins that are in turn able to regulate a variety of downstream effects such as chromatin structure, transcription and DNA repair (Grant, 2001; Mersfelder and Parthun, 2006). Our focus here is not limited to but will be on histone acetylation, lysine methylation and ubiquitination.
I.7.4 Histone Acetylation Histone acetylation has multiple roles in the chromatin regulation. In terms of transcription, this mark is typically associated with active transcription and open chromatin, whereas non-acetylated histones are linked with gene silencing and heterochromatin formation
(Vaquero et al., 2003). The introduction of acetyl groups onto histones is performed by a class of enzymes known as histone acetyl transferases (HATs). Histone acetylation causes relaxation of compact chromatin structure through the neutralization of the positive charge on the amino group of lysine, thereby reducing the affinity between the negatively charged DNA and the histone (Yang and Seto, 2008). Histone deacetylases (HDAC‟s) functionally counteract HAT activity by removing acetyl groups from lysine residues on histone tails, thereby promoting compaction of chromatin which can shield DNA from transcription or repair factors (Kurdistani and Grunstein, 2003). An equilibrium between HAT and HDAC activity is thought to form the basis for proper gene expression and genome integrity maintenance. Accessibility of DNA lesions in chromatin by DNA repair factors is facilitated by histone acetylation which relaxes chromatin, followed by histone deacetylation which restores chromatin structure (Kurdistani and
34
Grunstein, 2003). Of note, acetylation defects have been shown to lead to increased sensitivity to DNA damaging agents (Sun et al., 2010; Sun et al., 2009).
I.7.5 Histone Lysine Methylation Histone methylation usually occurs on a lysine (K) or arginine. Lysine can be methylated by the addition of one, two, or three methyl groups/residue. Histone lysine methylation is generally associated with gene repression but there are cases of histone methylation that are associated with activation of gene expression. Histone trimethylation, especially trimethylation on lysine 9 of H3 and lysine 20 of H4, is a signature modification associated with heterochromatin (Guenatri et al., 2004; Maison and Almouzni, 2004; Santos-Rosa and Caldas,
2005). These trimethylation marks act as docking sites for heterochromatin associated proteins - for example, heterochromatin protein 1 (HP1) proteins bind to H3K9me3 via a specific protein domain known as a chromodomain (CD). Other methylation marks associated with heterochromatin include methylation at H4K16 and H2K27 and H3K64 trimethylation (Daujat et al., 2009; Fraga et al., 2005; Jones et al., 2008; Vaquero et al., 2003).
H3K9 trimethylation is generally associated with large-scale repression of gene expression, with highly condensed heterochromatin regions showing a high degree of H3K9 methylation (Kouzarides, 2007). One study, however, has found H3K9 di- and tri- methylation enrichment in genes that are transcriptionally active in mammalian chromatin (Bannister et al.,
2001). H3K9 trimethylation is also linked with X-inactivation, the process by which one of the two copies of the X chromosome in females is transcriptionally silenced, allowing for the equalized expression of X-linked genes in females (XX) and males (XY) (Barakat et al., 2010;
Heard et al., 2001; Peters et al., 2002).
35
Lysine methylation by histone methyltransferases (HKMTs) was originally believed to be an irreversible modification before the discovery of histone demethylases (Hakimi et al.,
2002; Hakimi et al., 2003; You et al., 2001). Of all the enzymes that modify histones, methyltransferases are the class which appear to exhibit the greatest substrate specificity. As mentioned previously, lysine residues can be mono-, di- or tri-methylated with different methyl transferases responsible for each additional methylation step on a given residue. The existence of these three methylation states suggests that each transition is regulated and that the regulation affects downstream events and may have distinct outcomes (Kouzarides, 2007). The presence of certain methylation modifications is also known to impact the propensity for subsequent modification of histones at certain additional amino acid residues (Kouzarides, 2007).
An important link between histone lysine methylation and neoplastic transformation was described in studies of histone lysine methyltransferases by Jenuwein and colleagues. They discovered that trimethylation of lysine 9 on histone H3 (H3K9me3) by Suv39H1 and
Suv39H2 HKMTs played an important role in maintenance of centromeric heterochromatin, known as pericentric heterochromatin (Rea et al., 2000). When mice deficient in both enzymes were created, surviving mice showed deficits in chromosomal integrity and succumbed to B cell lymphomas. Cells from these mice showed a lack of this trimethylation modification within pericentric regions of chromatin. Taken together, these findings demonstrated an important role of epigenetic modifications of histones for maintenance of chromosomal integrity (Rea et al.,
2000).
36
I.7.6 Ubiquitination of Histones
Ubiquitin is a highly conserved polypeptide comprised of 76 amino acids that is distributed throughout the eukaryotic cell. Ubiquitin is covalently attached by an isopeptide bond between the c-terminal glycine of ubiquitin and the ε-amino group of a lysine residue on the ubiquitin recipient substrate termed the „acceptor‟. This process of ubiquitinationis accomplished in three stages via E1 activating, E2 conjugating and E3 ligase enzymes respectively, with the E3 enzyme having the capacity for specific recognition and ligation of ubiquitin to its substrates
(figure 12). Multiple attachments of ubiquitin molecules (polyubiquitination) are typically a signal for degradation of the protein substrate via the 26S proteosome in the cytoplasm, whereas a single attachment of ubiquitin (monoubiquitination) is typically associated with signal transduction events linked to protein trafficking, cell-cycle regulation, DNA repair, endocytosis and transcriptional regulation (figure 12) (Cuervo et al., 2010). The covalent addition of the
76 amino acid ubiquitin molecule onto a histone is considered large modification that may serve to force chromatin conformation open by forming a „wedge‟ to allow access for DNA transactions (Kouzarides, 2007).
I.7.7 Histone Variant Incorporation Histones H2A, H2B and H3 altogether have approximately 150 known variants that when incorporated into nucleosomes, appear to confer specific attributes to chromatin structure that in turn impact the dynamics of chromosome segregation, transcriptional regulation and DNA repair. Histone variants can also provide unique sites for certain post-translational modifications that impact cellular processes such as DNA repair. One example of such a unique site is the serine 139 residue within the H2A variant H2AX, which is phosphorylated following strand
37
break damage, converting this histone into a binding site for the DNA damage response protein
MDC1 (Talbert and Henikoff, 2010).
I.7.8 DNA methylation DNA methylation is the process by which a methyl group is added to the fifth carbon position of cytosine. Methylation typically occurs at CpG (cytosine-phosphate-guanine sites) in the human genome, and it is estimated that 60-90% of CpG sites are methylated (Levenson,
2010). This modification is frequently observed in transcriptionally silent genomic tracts of CpG dinucleotides known as CpG islands that are commonly found in regions of heterochromatin.
Changes in the extent of DNA methylation within certain genes is known to guide various biological transitions, such as during various stages of development. For example, DNA methylation of the X-inactive specific transcript gene is crucial for embryonic X chromosome inactivation in mammals (Avner and Heard, 2001; Boumil and Lee, 2001) (Leeb and Wutz,
2010) (Ariel et al., 1995) . The extent of global DNA methylation has also been shown to change during the process of neoplastic transformation. Human tumours have been shown to have hypermethylation of certain tumor suppressor genes and hypomethylation of certain oncogenes and regions of heterochromatin (Cheung et al., 2009). Abnormal hypermethylation has been described as an early event in cancer formation and has been linked with deregulation over activity of DNA methyltransferases (DNMTs) (Cheung et al., 2009). Currently three
DNMT‟s have been identified in mammals: DNMT3A, DNMT3B and DNMT1. Both DNMT3A and DNMT3B have roles in de novo methylation processes in response to various cues, whereas
DNMT1 is a maintenance methyltransferase that is responsible for copying parental DNA methylation patterns to the daughter strands during replication (Cheung et al., 2009; Jenkins et al., 2005).
39
I.8 Chromatin Accessibility and DNA repair
The ability of DNA repair proteins to efficiently repair DNA damage is dependent on the ability of compact chromatin to relax following damage (Dery and Masson, 2007). Accordingly, special accommodations need to be made to ensure that compacted chromatin in heterochromatin is relaxed to facilitate access by DNA repair factors. Otherwise, the high degree of chromatin compaction could act as an obstruction for signalling and recruitment of proteins needed for
DNA repair. For example, γH2AX focus intensity is decreased in heterochromatin compared to euchromatin of irradiated cells (Downs et al., 2007; Rogakou et al., 1998).
Recent studies have revealed ATM facilitates access of damaged DNA to repair factors in chromatin by relieving chromatin compaction locally and globally. The ATM kinase pathway has been shown to relieve global heterochromatin compaction for the repair of DSBs by phosphorylating the KRAB-ZFP–associated protein 1 (KAP-1), a protein that helps to maintain the compacted state of heterochromatin. ATM activation following DNA damage results in rapid spreading of KAP-1 phosphorylation and exclusion of this protein from chromatin, resulting in global chromatin relaxation (Goodarzi et al., 2008; Ziv et al., 2006). Following
DNA damage, HATs are also recruited for acetylation of nearby histones, facilitating further chromatin relaxation and allowing access for DNA-repair proteins to the site of damage. For example, ATM recruits a HAT known as Tip60 to sites of DSB damage (Sun et al., 2005). Upon completion of DNA repair, histones are then deacetylated by HDACs and the chromatin is reassembled to its original state.
1.8.1 Unmasking of H3K9me3 facilitates activation of Tip60 Histone Acetyltransferase during the DNA damage response
40
Members of the HP1 family impact various DNA transactions through their ability to bind H3K9me3 (Maison and Almouzni, 2004). Three closely related isoforms Hp1α, Hp1β and
Hp1γ all share the presence of two conserved domains, which are the N-terminal chromodomain
(CD) and C-terminal chromoshadow (CSD) domain (Luijsterburg et al., 2009). Recently, the
HP1β isoform has received considerable interest because of its recently identified role in the
DNA damage response. HP1β specifically binds to H3K9me3 via hydrogen bonds by folding its
CD around this chromatin mark (Sun et al., 2009). Ayoub and colleagues found that during
DNA damage, HP1β is phosphorylated on amino acid Thr51 by casein kinase 2 (CK2) which causes a disruption of the hydrogen bonds that allow HP1β to bind to H3K9me3, causing its release from chromatin (Allende-Vega et al., 2005; Ayoub et al., 2008; Ayoub et al., 2009;
Loizou et al., 2004). Experimental inhibition of this phosphorylation event was sufficient to prevent HP1β ejection from chromatin (Ayoub et al., 2008). Interestingly, the ejection of HP1β from chromatin was found to be required for optimal γH2AX phosphorylation at sites of DNA damage (Ayoub et al., 2008; Ayoub et al., 2009). These findings however were not supported by a subsequent study which reported a distinct role for HP1 proteins in UV-induced DNA damage. In this study, all three isoforms of HP1 were shown to be recruited to UV-damaged
DNA and that this recruitment is especially dependent on its CSD domain (Dinant and
Luijsterburg, 2009; Luijsterburg et al., 2009).
From these findings the question arose: if HP1β ejection leads to increased γH2AX, then what is the signalling mechanism responsible for this event? Recently, exciting findings have been able to fill in these gaps of knowledge and identify the HAT Tip60 as playing a role in this pathway. Tip60 is considered to be both a chromatin remodelling protein and a tumour suppressor that has been shown to have increased expression in human cancers (Gorrini et al.,
41
2007; Halkidou et al., 2003; Kim et al., 2005; ME et al., 2006). Similar to HP1 proteins, Tip60 also has a CD that allows it to specifically bind to H3K9me3. It has been shown that in the presence of DSBs, the overall protein level of Tip60 does not change; however, there is an increase in Tip60 acetyltransferase activity (Sun et al., 2009). In the presence of damage, Tip60 is recruited to sites of DNA lesions as part of a complex with ATM. The CD domain of Tip60 binds to H3K9me3 unmasked by the ejection of HP1β which initiates Tip60 HAT activation.
Once activated, Tip60 then acetylates ATM, a modification which then facilitates activation of
ATM kinase activity through an unknown mechanism (Sun et al., 2009). The CD domain of
Tip60 is crucial for initiating Tip60 activity as ATM activation is attenuated when mutations are created in this domain. Furthermore, reductions in the extent of chromatin-associated H3K9me3 have also been shown to impair Tip60 and ATM activity (Sun et al., 2009). Sun and colleagues have proposed the following model for that illustrates the role of Tip60 and Hp1β in the repair of
DSBs :
1) Upon DNA damage, the CD of Hp1β is phosphorylated by CK2. The origin of the signal that directs this phosphorylation event is unclear.
2) Hp1β is ejected from H3K9me3, unmasking this histone modification for Tip60 association.
3) The association of MRN at the ends of the DSB serves as a beacon for ATM recruitment to the DSB site.
4) Tip60 is then recruited to MRN at DSBs as part of an inactive complex with ATM.
5) Tip60 binds to H3K9me3 through an interaction with its CD which leads to Tip60 HAT activation, ATM acetylation and ATM kinase activation (figure 13) (Sun et al., 2009).
H3K9me3 is found abundantly in regions of heterochromatin where it interacts with HP1 proteins, implicating Tip60 as the HAT required for facilitating access of DNA damage found
42
primarily in heterochromatic regions (Sun et al., 2009). Tip60 may also play additional roles in facilitating access to other forms of DNA damage such as DNA crosslinks, since Tip60 was also identified in a molecular complex containing the FA pathway protein FANCD2 (Hejna et al.,
2008).
I.9 UHRF1 (Ubiquitin-like, containing PHD and RING finger domains 1)
I.9.1 UHRF1 is a proliferation-associated ubiquitin ligase UHRF1 was originally detected as a 95-kDa mouse nuclear protein (Np95) by a monoclonal antibody (Th-10a) specifically in the S phase of normal mouse thymocytes (Muto,
1995). Cloning of the corresponding cDNA revealed that the 782 amino acid open reading frame of UHRF1 consisted of an unusual N-terminal domain with a striking resemblance to ubiquitin, and a RING finger domain located at the C-terminus (Fujimori, 1998). Further examination of cells by indirect immunofluorescence revealed that UHRF1 was localized in a fraction of S- phase nuclei as dot-like foci together with proliferating cell nuclear antigen (Uemura, 2000).
UHRF1 was not found localized with replication centers throughout S phase, suggesting that this protein was not a component of the DNA replication machinery but instead might be involved in other DNA replication-linked nuclear events (Miura, 2001)
Another independent study identified the human counterpart of UHRF1 (Np95- nuclear protein 95 or ICBP90 – Inverted CCAAT box binding protein, 90 kDa) as a transacting factor that activated transcription of DNA topoisomerase IIα (Hopfner, 2000). This study also noted the existence of a PHD (plant homeodomain) within ICBP90 which is shared by Np95. PHD domains are one class of „epigenetic readers‟ as they facilitate binding of proteins to methylated histones, especially tri-methylated lysine 4 or tri-methylated lysine 9 of histone H3 (Aasland R,
44
1995). A subsequent study, however, demonstrated conclusively that UHRF1 did not bind to inverted CCAAT box motifs or regulate transcription of DNA topoisomerase IIα (Muto, 2006).
I.9.2 UHRF1 and the replication of heterochromatin Although early studies demonstrated a link between UHRF1 expression and proliferation, the exact function of this protein remained elusive. The first attempts at inhibiting the function of UHRF1 (via microinjection of antibodies) concluded that UHRF1 was required for progression through S phase of normal cells (Bonapace, 2002). A subsequent study demonstrated that UHRF1 bound to histones via a SRA (SET and RING finger-associated) domain and that the C-terminal RING finger could ubiquitinate histones (Citterio, 2004). The role of UHRF1 in the progression through S phase was also found to be due to a role for this protein in the replication of heterochromatin, an event that occurs in the middle to late period of
S phase (Papait, 2007). Depletion of UHRF1 by siRNA was found to reduce pericentric heterochromatin replication and result in the hyperacetylation of lysines 8, 12, and 16 of heterochromatin histone H4 (Papait, 2007). This hyperacetylation might be due to loss of
HDAC1 which was also found to be associated with UHRF1 (Unoki, 2004).
I.9.3 UHRF1 facilitates transfer of DNA methylation during DNA replication The first clue that UHRF1 could be important for mediating DNA methylation was found in studies of the SRA domain of UHRF1 which was found to methyl-CpG tracts (Unoki, 2004).
Subsequent studies demonstrated that the SRA domain showed strong preferential binding to hemi-methylated CpG sites which are the physiological substrates for DNMT1, the DNA methyltransferase that copies pre-existing methylation patterns onto newly synthesized DNA after replication. UHRF1 was found to co-localize and tether DNMT1 to chromatin. ES cells deficient in UHRF1 displayed deficits in genome methylation patterns, demonstrating a role for
UHRF1 in proper inheritance of global genome methylation (Bostick, 2007; Sharif, 2007).
45
Elegant crystallography studies conclusively demonstrated that the SRA domain selectively binds hemimethylated DNA (Arita, 2008; Avvakumov, 2008; Hashimoto H, 2008). UHRF1 was also found to be associated with DNMT3a and DNMT3b methyltranferases and participate in de novo methylation events that contribute to repression of gene expression (Meilenger, 2009).
I.9.4 UHRF1 and maintenance of the H3K9me3 mark in heterochromatin Human UHRF1 was isolated in a screen for proteins that specifically bind H3 methylated at lysine 9, with the binding specificity determined by the PHD. Cells depleted of UHRF1 were deficient in H3K9me3 localized to heterochromatin, although total H3K9me3 levels remained constant. Over expression of a dominant-negative UHRF1 RING finger mutant led to disruption of heterochromatin structure as evidenced by the loss of dense heterochromatic staining visualized by indirect immunofluorescence of DNA staining by DAPI. Taken together, these findings suggested a new role for UHRF1 in the maintenance of the H3K9me3 mark in heterochromatin, although the exact mechanism behind this role is unclear (Karagianni, 2008).
These conclusions have been challenged by a subsequent study which demonstrated H3K9me3 binding occurs via a tandem tudor domain and not the PHD domain, as well as another study which did not observe a loss of heterochromatic H3K9me3 following depletion of UHRF1
(Papait, 2008; Rottach, 2009). It is worth noting, that the status of H3K9me3 with UHRF1 depletion is disputed. One study by Karagianni and colleagues observed that stable depletion of
UHRF1 resulted in H3K9me3 staining becomes diffuse, whereas Papait and colleagues utilized transient depletion of UHRF1 in their studies and report no change in the staining pattern of
H3K9me3 using indirect immunofluorescence studies (Karagianni, 2008; Papait, 2008). The newly described role for UHRF1 in heterochromatic histone methylation maintenance was further substantiated by a report of UHRF1 binding a HMKT known as G9a via its RING finger.
Embryonic stem cells deficient in UHRF1 were found to exhibit decreased G9a and dimethyl 46
H3K9 on chromatin (Kim, 2008). As H3K9 dimethyl groups serve as substrates for the generation of H3K9me3 by Suv39H methyltransferases, it is possible that UHRF1 facilitates heterochromatic H3K9me3 by tethering HMKT‟s such as G9a or Suv39H to heterochromatin
(Rottach et al., 2010). Further studies will be required to clarify the exact sequence of events for this process.
I.9.5 UHRF1 roles in sensitivity to DNA damage and cancer Early studies pointed to a potential role for UHRF1 in the DNA damage response as ES cells lacking UHRF1 were found to be more sensitive to the cytocidal effect of X-rays, UV, hydroxyurea (an agent that impedes DNA replication) and alkylating agents (Muto, 2002).
Interestingly, UHRF1-deficient cells also exhibited slightly increased incidences of spontaneous sister chromatid exchange compared to normal controls which suggests that genome stability might be threatened or lost with the loss of UHRF1 (Muto et al., 2002; Muto, 2002). Subsequent studies validated that depletion of UHRF1 sensitized cells to genotoxic chemotherapy whereas over expression of UHRF1 could render cells resistant (Jenkins et al., 2005; Muto, 2006). Taken together, these findings suggest a possible role for UHRF1 in the cellular response to DNA damage; however, a model for its role in the DNA damage response is lacking.
Numerous studies have reported over expression of UHRF1 in various forms of cancer, including breast cancer (Bronner, 2007; Hopfner, 2002; Jenkins, 2005; Mousli, 2003), cervical cancer (Lorenzato, 2005), prostate cancer (Jenkins, 2005), pancreatic adenocarcinoma
(Crnogorac-Jurcevic, 2005), rhabdomyosarcomas (Schaaf, 2005), thyroid tumours and various gliomas (Oba-Shinjo, 2005; Pita, 2009). Elevated levels of UHRF1 may confer a selective advantage to cancer cells by deregulating chromatin structure and gene expression that could facilitate silencing of tumour suppressor genes or may possibly render cancer cells resistant to antineoplastic agents that damage DNA. Further exploration of the relationship between UHRF1 47 activity and neoplastic transformation will be required to understand the role, if any, of UHRF1 in these processes.
In summary, UHRF1 is a protein that appears to act as an interface for processes of DNA methylation and histone methylation events linked with heterochromatin replication. Figure 14 summarizes these relationships and the role of UHRF1 protein domains in these events (Unoki et al., 2009).
48
Aims of my thesis research
Mus81-Eme1 is a heterodimeric DNA binding and cleaving endonuclease that was originally shown to be important for the cellular response to replication-associated DNA damage. In particular, this enzyme has been shown to readily cleave forked DNA structures that have been proposed to resemble stalled sites of DNA replication. Cells deficient in either
Mus81 or Eme1 display increased sensitivity to DNA damaging agents, such as hydroxyurea and mitomycin C that stall replication by creating suboptimal conditions for DNA replication progression. In the absence of DNA damaging agents, cells deficient in Mus81-Eme1, display defects in genome integrity and maintenance. The ability of Mus81 to contribute to genome integrity as a „caretaker‟ protein may explain its role as a tumour suppressor in mammals.
Although recent studies have demonstrated the importance of Mus81-Eme1 endonuclease in DNA repair, maintenance of genome integrity and tumour suppression, relatively little is known about how the enzyme acts together with other components of the signalling pathways that comprise the DNA damage response. The original focus of my thesis research was to identify proteins that interact with this endonuclease in order to discover the DNA repair pathway in which Mus81-Eme1 participates. In order to gain insight into these processes, our laboratory performed protein interaction screens using Eme1 as „bait‟ in yeast-two hybrid screens. UHRF1 was identified as an interacting protein partner. UHRF1 is a chromatin- associated ubiquitin ligase that plays a potential role in the DNA damage response as cells deficient in UHRF1 are more sensitive to various DNA damaging agents. Although these
50 findings, point to a role for UHRF1 in protecting cells from the cytocidal action of these agents, a role for this protein in the DNA damage response has not formally been identified.
The focus my thesis research is to:
1) Characterize the interaction and significance of two proteins found in the DNA damage response; Eme1 and UHRF1, which forms the basis for studies described in Chapter Two.
2) To determine the role that UHRF1 plays in response to γ-irradiation, which forms the basis for studies described in Chapter Three.
3) Investigate a role for UHRF1 in the DNA damage response through its previously described role in maintenance of heterochromatic H3K9me3, which forms the basis for studies described in
Chapter Four.
51
CHAPTER TWO: Interaction between Eme1 and UHRF1
Modified versions of the material from this chapter have been published in the following report:
Helena Mistry*, Lianne Gibson*, Ji Weon Yun, Haya Sarras, Laura Tamblyn, John Peter McPherson. Interplay between Np95 and Eme1 in the DNA damage response. Biochem Biophys Res Commun. 2008 Oct 24;375(3):321-5.
52
Introduction
Rationale: Mus81-Eme1 is a heterodimeric DNA binding and cleaving enzyme that participates in the repair of replication-associated DNA damage and the maintenance of genome integrity
(Boddy et al., 2000; Interthal and Heyer, 2000), (Schmuckli-Maurer et al., 2003). The activity and substrate specificity of this endonuclease has been extensively studied; however, relatively little is known about how this enzyme is coordinated with other proteins in the cellular response to DNA damage. In particular we know little about the repair pathway this enzyme participates in. Identification and characterization of proteins that interact with Mus81-Eme1 may shed light on the nature of the role of this enzyme in DNA repair.
Objective: In order to search for proteins that interact with Mus81-Eme1, we originally screened for candidate Eme1 protein partners using expression libraries and yeast two-hybrid technology.
The focus of this chapter is on the identification of an interaction between Eme1 and a protein known as UHRF1, a chromatin-associated ubiquitin ligase (Muto et al., 2002).
Outline: An Eme1-UHRF1 interaction was identified and confirmed using three different methodologies: yeast two-hybrid, GST pull-down, and co-immunoprecipitation studies from mammalian protein lysates. Analysis of UHRF1 interactions with a series of Eme1 deletion mutants revealed that amino acids 40-186 were sufficient for an interaction with UHRF1.
Co-localization between UHRF1 and Eme1 was found to occur on nuclear chromatin following exposure to the replication fork damaging agent CPT. This co-localization was dependent on an intact UHRF1 RING finger domain. We also discovered that transient co-transfection of Eme1 and UHRF1 revealed that forced expression of UHRF1 was found to reduce steady-state Eme1
53 protein levels. Taken together, these findings link Mus81-Eme1 with the chromatin modifier functions of UHRF1 in the cellular response to DNA damage.
Materials and Methods
Eme1-Interacting Protein Screen cDNA encoding the open reading frame of murine Eme1 was originally amplified by
PCR from a Marathon cDNA library (Clontech, Palo Alto, CA) with the following primers:
5‟-GTCATATGGCTCTAAGAAGGTTATCCCT-3‟ 5‟-AAGTCGACTCAGTCAACACTGTCTAAGATGAG-3‟ and cloned into pcDNA3. A two-stage subcloning approach was used to transfer Eme1 cDNA into pGBKT7 (Clontech) in-frame with the open reading frame of GAL4 DBD. A BamH1-Sal1 fragment representing the 3‟ portion of the Eme1 cDNA was initially cloned into the corresponding restriction sites in pGBKT7 (Clontech). To create a DNA fragment of the 5‟ portion of Eme1 to be in-frame with GAL4 vector sequence, PCR was conducted with Eme1 cDNA template and primers 5‟-CG TTA GATC GAA TTC ATG GCT CTA AGA AGG TTA
TCC-3‟ and 3‟-CCC CAC CTT CCA TCT GT-5‟. The resulting product was cloned as an
EcoR1-BamH1 fragment into the pGBKT7 vector containing the BamH1-Sal1 fragment, to create pGBKT7 (performed by Haya Sarras). Yeast Saccharomyces cerevisiae strain AH109 was transformed with pGBKT7-Eme1 and strain Y187 (MAT) was transformed with a murine embryonic day 11 cDNA library cloned into pACT2 (Clontech). Colonies that grew on media lacking leucine, tryptophan, histidine and adenine were picked, isolated and sequenced
(performed by Dr. Peter McPherson). From 2 × 105 clones screened, 23 were found to grow on leucine-, tryptophan-, histidine- and adenine-deficient plates. Eleven of the 23 „prey plasmids‟
54 corresponding to a library plasmid were recovered from yeast, transformed into E. coli DH5α and sequenced. Interestingly, Mus81 was not recovered in the Eme1 screen. The sequence of nine of the eleven sequenced plasmids was identical to known murine genes: AF9, UHRF1
(Np95), Hmgn2, fibrillin, eukTEF1, SAP30 and 2G4. One of these plasmids was found to contain sequence corresponding to residues 502-782 of UHRF1 and is hereafter referred to as
∆502.
GST Pulldown Assay The ∆502 insert from pACT2 was subcloned into pGEX4T3 to create ∆502-GST.
Escherichia coli BL21pLysS cells (Stratagene) transformed with either pGEX4T3 or ∆502-GST
DNA were cultured at 30ºC to mid-log phase. Induction of either GST or GST-fusion proteins was performed by the addition of isopropyl-β-thigalactopyranoside (to 0.5 mM final). After 4 hr, harvested cells were lysed and lysates were purified onto glutathione-agarose beads (Sigma).
In vitro generated mRNAs generated from plasmids encoding Eme1, Mus81 or Luciferase were used to program rabbit reticulocyte lysates and synthesized proteins were detected as species with [35S] Methionine incorporation (TNT-coupled reticulocyte lysate systems, Promega). GST and GST-∆502 proteins were incubated with [35S] Methionine-labelled Mus81, Eme1 or
Luciferase for two hours at 4ºC in binding buffer. Bound proteins were recovered on GST or
GST-∆502 immobilized on glutathione-agarose by boiling in Laemmli buffer, separated on 10%
SDS-PAGE, stained with Coomassie blue to verify equal loading of fusion proteins and 35S activity was detected using a Storm PhosphorImager (Molecular Dynamics) to detect radiolabeled in vitro translated proteins.
55
Creation of Eme1 Mutants Eme1 deletion mutants were created to identify sites of UHRF1 interaction on Eme1.
Four independent deletion mutants (I-IV, figure 15) spanning the open reading frame of Eme1 cDNA were amplified by PCR using the following primer pairs described in Table 1 (performed by Brenda Yun).
Table 1
Deletion mutant I – cloned as an EcoR1-Cla1 fragment corresponding to residues 40-187, 16 kDa: 5‟-CGTTAGATCGAATTCAGGGCGAAGAACATAGTGG-3‟ 5‟-CGTTAGATCGATGGTAGAGGTCGGTCGGAATTT-3' Deletion mutant II – cloned as an EcoR1-Cla1 fragment corresponding to residues 40-341, 33 kDa: 5'-CGTTAAGATCGAATTCAGGGCGAAGAACATAGTGG-3‟ 5‟-CGTTAGATCGATGTTACAAAGCTCCGAAGGGTC-3‟ Deletion mutant III – cloned as an EcoR1-Cla1 fragment corresponding to residues 188-498, 34 kDa: 5‟-GTTAGATCGAATTCACAAATTCCGACCGACCTCTAC-3‟ 5‟-CGTTAGATCGATAGGCGTCCATAGCAC-3‟
Deletion mutant IV – cloned as a Pst I fragment corresponding to residues 342-528, 20 kDa: 5‟-CGTTAGATCGATCCACGTCGCACTTGTATGTCA-3‟ 5‟-CGTTAGATCGATCCACGTCGCACTTGTATGTCA-3‟
Table 1. Sequence of Primers and Restriction Enzyme Sites used to create Eme1 mutants. Restriction enzymes were used to generate Eme1 mutants into the vector of interest (pGBKT7). Italicized sequence does not correspond to Eme1 sequence, underlined sequence indicates engineered restriction sites for cloning.
56
PCR products were gel-purified from agarose gels following electrophoresis using a QIAquick
Gel Extraction kit (QIAGEN, Mississauga, ON) and cloned into pCR 2.1-TOPO vectors (TOPO
TA Cloning Kit, Invitrogen) and sequenced with T7 and M13Reverse primers (ACGT corporation, Toronto, Canada). Upon sequence validation, fragments were subcloned into pGBKT7 and manipulated in AH109 as for pGBKT7-Eme1.
Generation of vector constructs and cell lines for cell culture To express UHRF1 with an N-terminal FLAG epitope, a cDNA containing the entire open reading frame of murine UHRF1 (MGC Clone ID# 4017286, Acc# BCO22167 in pCMV6-
Sport) was inserted into the BamH1 and Xho1 sites of pCMV-TAG 2C (Stratagene). Two
UHRF1-FLAG mutants were then created. The first mutant has a replacement of the histidine at position 730 with alanine (FLAG-UHRF1 H730A). The second mutant has a deletion at 713
(FLAG-UHRF1 C713-762). The mutant constructs were generated by site-directed mutagenesis with the following primer pairs:
H730A: CGT > GGC
5‟-CAC CAC CGT GTG TCA GGC CAA CGT CTG TAA GGA C-3‟ 5‟-GTC CTT ACA GAC GTT GGC CTG ACA CAC GGT GGT G-3‟
C713*: ACG > TAG 5‟-GAAAGAGGCTTTCCAGTAGATCTGCTGCCAGGAGCT-3‟ 5‟-AGCTCCTGGCAGCAGATCTACTGGAAAGCCTCTTTC-3‟
and further verified by DNA sequencing. To create an Eme1 expression vector with a V5 epitope in-frame with the C-terminus, murine Eme1 cDNA was inserted into vector pcDNA3.1V5HisA (Invitrogen) in two stages. A BamHI-XbaI C-terminal PCR fragment corresponding to the 3‟ end of the Eme1 cDNA was amplified by PCR with the following primers:
58
5‟-C GTG GTG GTG CTG GAT CCA (BamHI site underlined) 5‟-CGT TAG TCT AGA GTC AAC ACT GTC TAA GAT GAG-3‟ (Xba I site underlined)
and cloned into pcDNA3.1 V5HisA via BamH1 and Xba1 sites to fuse the coding sequence with the V5 epitope sequence in the vector. A subsequent cloning step transferred a BamH1-BamH1 fragment representing the 5‟ end of the Eme1 cDNA to create Eme1-V5.
Transient Transfections Mouse NIH 3T3 cells or human HeLa cells were transiently transfected with FLAG-
UHRF1 or Eme1-V5 using Lipofectamine 2000 (Invitrogen) according to the instructions of the manufacturer. Cells were collected for experiments 24-36 hours post transfection and harvested for analysis.
Immunoprecipitation and Western Blotting Immunoprecipitation experiments were carried out by lysis of seeded cells in immunoprecipitation lysis buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, 5% glycerol, 0.1%
NP-40, 0.1% -mercaptoethanol, 0.5 mM PMSF, 5 µg/ml aprotinin,5 µg/ml leupeptin, 5 µg/ml pepstatin). Lysates were incubated overnight at 4 ºC in 100 µl of either anti-FLAG M2 affinity beads or anti-V5 agarose affinity beads (Sigma) (partial endogenous co-immunoprecipitation performed by Loni Gibson). Washes were performed with immunoprecipitation lysis buffer and then lysates were boiled in 6x Laemmli sample buffer for 5 minutes to release protein from beads. For Western analysis, cells were lysed in RIPA buffer (50 mM Tris-HCl pH 8.0, 150 mM
NaCl, 1 mM PMSF, 1 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 5
µg/ml aprotinin, 5 µg/ml leupeptin, and 5 µg/ml pepstatin). Protein quantification was done by
Bradford analysis to load equal amounts of protein. Lysates were then resolved on either 6, 10 or
59
12 % SDS-PAGE gels and transferred onto PVDF membranes overnight at 4 ºC. Membranes were then blocked in Tris-buffered saline/5% milk/0.1% Tween, and then immunoblotted with the following primary antibodies indicated in Table 2, followed by incubation with the appropriate HRP-linked secondary antibody: anti-rabbit HRP-linked IgG or anti-mouse HRP- linked IgG (GE Healthcare). Protein detection was performed using the ECL Western blotting detection system (GE healthcare).
Antigen Supplier Monoclonal Dilution
/Polyclonal used
Eme1 Santa Cruz Monoclonal 1:333
FLAG Sigma Monoclonal 1:400
UHRF1 Abnova Monoclonal 1:200
Actin Sigma Polyclonal 1:1000
V5 Sigma Polyclonal 1:1000
Table 2. Antibodies used for Immunoprecipitation and Western experiments.
Antibodies were incubated overnight at 4 ˚C to detect the corresponding protein of interest.
Indirect Immunofluorescence Indirect immunofluorescence for NIH 3T3 cells in 60 mm dishes was performed 48 hours following transient transfection with either Eme1-V5 and/or UHRF1-FLAG. Forty-eight hours following transfection, cells were seeded onto chamber slides (Nunc) and cultured in the absence or presence of camptothecin (1 µM, Sigma) for 12 h. Cells were then fixed in 4% paraformaldehyde and incubated with anti-V5 and anti-FLAG antibodies overnight at 4 ˚C. The
60 following day, cells transfected with Eme1-V5 were incubated with rabbit FITC-conjugated secondary antibody, whereas cells transfected with FLAG-UHRF1 cells were incubated with mouse TRITC-conjugated secondary antibody. Cells were transfected with both constructs were incubated with both secondary antibodies (Jackson ImmunoResearch Labs). Cell were then counterstained with 4,6-diamidino-2-phenylindole (DAPI) and mounted with Vectorshield
(Vector laboratories). Images were acquired using an Imager.ZI fluorescence microscope and
Axiovision software (Zeiss) following deconvolution using Autolinear and Clip algorithms
(performed by Loni Gibson).
Results
In order to find interacting proteins with Eme1, a yeast two hybrid screen of mouse embryonic day 11 cDNA library (Matchmaker 3, Clontech, Palo Alto, CA) was performed with the use of Eme1-GAL4 DNA binding domain fusion protein as a „bait‟. Of eleven positive interactions sequenced, one of the positive interacting yeast clones was found to contain 280 amino acids (amino acids 502 to 782) corresponding to the carboxy-terminus of UHRF1. This partial fragment contained some of the sequence corresponding to the SRA domain and all of the
RING finger domain (figure 16). The yeast-two hybrid assay was performed on this isolated
UHRF1 prey plasmid (hereafter referred to as ∆502) to confirm the observed interaction between
UHRF1 and Eme1 sequences and rule out the presence of a contaminant (figure 16B and C).
S. cerevisiae Y187 was transformed with ∆502 and mated with AH109 transformed with either
GAL4BD-Eme1 or GAL4BD-Mus81 (both in vector pGBKT7). Diploids that were transformed with both pGBKT7-p53 and pGADT7-TD1-1, encoding the GAL4-AD fused to SV40 large T antigen sequence served as a positive interaction control. GAL4AD-UHRF1 mated with
61
pGBKT7 vector alone and GAL4BD-Eme1 mated with vector alone served as negative controls to confirm that UHRF1 sequence in that plasmid and Eme1 are not capable of non-specific reporter gene transcriptional activation. GAL4BD-Eme1 mated with GAL4AD-T-antigen and
GAL4BD-p53 mated with GAL4AD-Δ502 also serve as negative controls showing that the interaction between the positive control is specific and further verifying the specific protein- protein interaction between UHRF1-Eme1. When streaked on media deficient in leucine, tryptophan, adenine and histidine, growth was seen in p53-SV40 large T containing diploids
(positive control), Mus81-Eme1 containing diploids (positive control) and Eme1-∆502 containing diploids (positive control) but not with other combinations (figure 16C).
In order to determine the region of Eme1 that is responsible for the interaction with
UHRF1, we constructed four deletion mutants of Eme1 fused to the DNA-BD of GAL4 and investigated their ability to interact with ∆502 (figure 15A). Yeast cells expressing Eme1 mutants I and II containing amino acid regions 40-187 and 40-341, together with ∆502 grew on media lacking tryptophan, leucine, histidine and adenine (figure 15B and C). Selective growth was not seen for yeast cell expressing Eme1 deletion mutants III and IV, containing amino acids
188-498 and 342-528 respectively, together with ∆502.
To validate the interaction between Eme1 and UHRF1 using an independent methodology we queried interactions between Eme1 and ∆502 using GST (Glutathione S- transferase)-pull-down assays. In this assay, a protein-protein interaction is considered positive if radiolabeled protein is selectively recovered from glutathione agarose beads containing a GST- fusion protein of the interacting protein of interest. ∆502-GST fusion protein was created and expressed in E. coli BL21(DE3)pLysS following induction of expression by isopropyl-β-
63 thigalactopyranoside and purification with glutathione-agarose beads (Sigma). As a control, bacteria expressing the GST vector alone was also induced and immobilized to glutathione- agarose beads for use as a negative control. Once ∆502-GST and GST fusion proteins were immobilized on beads, they were incubated with [35S] Methionine-labelled, in vitro translated
Eme1, Mus81 or Luciferase (negative control). Radiolabeled Eme1 was selectively recovered on
∆502-GST beads following centrifugation but not on beads with GST alone (figure 17).
Furthermore, Mus81 and Luciferase were not recovered on either GST-beads or beads with
∆502-GST. All samples and recovered fractions were recovered on an SDS-PAGE gel and
Coomassie stained to detect GST and ∆502-GS fusion protein (figure 17, lower panel), and scanned by Phosphoimager to detect radiolabeled in vitro translated proteins Eme1, Mus81 and
Luciferase (figure 17, upper panel). Radiolabeled Eme1, but not Mus81 or lucificerase was selectively recovered from ∆502-GST beads but not GST beads, confirming the interaction findings observed with the yeast two-hybrid approach (figure 17, upper panel).
To determine whether UHRF1 and Eme1 interact in mammalian cells, NIH 3T3 cells were transiently transfected with FLAG-UHRF1 (114 kDa), Eme1-V5 (72 kDa), or both FLAG-
UHRF1 and Eme1-V5 and subjected to immunoprecipitation using anti-FLAG conjugated beads.
Proteins pulled down were separated by SDS–PAGE gels and blotted with anti-FLAG and anti-
V5. Co-immunoprecipitation of Eme1-V5 was detected in cell lysates transiently transfected with both FLAG-UHRF1 and Eme1-V5, suggestive of an interaction between Eme1 and UHRF1
(figure 18A). We were also able to co-immunoprecipitate endogenous UHRF1 and Eme1-V5.
Eme1-V5 was precipitated using anti-V5 coupled to Protein A-agarose beads. Co- immunoprecipitation of endogenous UHRF1 was detected in cell lysates transiently transfected with Eme1-V5 and not vector control (figure 18B).
64
To determine whether the candidate interaction can be detected in cells, NIH3T3 cells were transiently transfected with FLAG-UHRF1 and Eme1-V5, and treated 24 hrs post transfection with either DMSO (figure 19 A-D) or 1 μM CPT (figure 19 E-H) for 12 hrs.
UHRF1 and Eme1 were visualized by indirect immunofluorescence using anti-FLAG and anti-
V5 antibodies followed by TRITC- (red) and FITC- (green) conjugated secondary antibodies, respectively. NIH 3T3 cells were also transfected with RING finger mutants; FLAG-UHRF1
H730A (figure 19 I-P) or FLAG-UHRF1 C713* (figure 19 Q-X) and treated as described above. Cells transfected with FLAG-UHRF1 and Eme1-V5 and treated with DMSO display an accumulation of UHRF1 in DAPI dense areas that are representative of heterochromatin. In contrast, Eme1 is shown to accumulate in DAPI low areas, representative of areas of euchromatin. A change in the nuclear staining pattern for both proteins resulting in co- localization between Eme1 and UHRF1 was observed in cells that treated with 1 µM CPT
(figure 19 F-H, representative of multiple experiments examining >100 transfected cells/experiment). Further analysis (tracing/histogram) revealed that UHRF1 and Eme1 selectively accumulated in DAPI-dense nuclear regions that are normally associated with the presence of heterochromatin but was not noted in DMSO-treated cells or cells expressing FLAG-
UHRF1 H730A (figure 19 I-P) or FLAG-UHRF1 C713* (figure 19 Q-X) either in the presence or absence of CPT. These findings suggest that an intact RING finger domain is required for UHRF1-Eme1 co-localization in the presence of CPT treatment.
67
Discussion
Mus81-Eme1 is a heterodimer endonuclease that has been shown to have roles in tumour suppression, the maintenance of genome integrity and DNA repair. Given its roles, relatively little is actually know about how the enzyme acts together other components of the DNA damage response to facilitate repair. Accordingly, we conducted a cDNA expression library screen looking for interacting proteins with Eme1. One of our interacting clones encoded the carboxy terminal domain for a protein known as UHRF1.
The interaction between Eme1 and UHRF1 was confirmed by transforming bait and prey plasmids into yeast strains and then we mated the transformed strains to confirm an interaction between Eme1 and UHRF1. This finding verified that the Eme1-UHRF1 interaction was not due to a contaminant or artefact found in the original cDNA library screen. In order to rule out a
„false positive result‟ from the yeast two-hybrid analysis, we confirmed our interactions using the
GST-pulldown and co-immunoprecipitation assays. GST-UHRF1 (containing the partial length) fusion protein was able to selectively pull-down Eme1 confirming results obtained by the yeast- two hybrid assay. The full length construct for UHRF1 was not used as we were not able to successfully express UHRF1-GST fusion proteins. This could occur due to the size of the protein which may not be able to be expressed efficiently in our E. coli BL21(DE3)pLysS cells. We created an GST-Eme1 construct and were not able to express the construct in our
BL21(DE3)pLysS cells. Given the above challenges, we created a partial length GST-UHRF1 construct that was able to be successfully expressed. In order to overcome these challenges, we decided to conduct co-immunoprecipitation assays using the full length epitope tagged UHRF1.
69
In order to validate the Eme-UHRF1 interaction in mammalian cells, co- immunoprecipitation experiments were conducted. As an antibody appropriate for immunoprecipitation or Western analysis of endogenous Eme1 in protein lysates was not available and given that attempts to raise an antibody against Eme1 in the laboratory were unsuccessful, we sought to explore whether Eme1-V5 transfected into NIH 3T3 cells would exist in complexes with transfected FLAG-UHRF1. We were able to show co-immunoprecipitation of transfected Eme1 with a V5 epitope tag (Eme1-V5) with UHRF1 with a FLAG epitope tag
(FLAG-UHRF1). We further validated this interaction with successful co-immunoprecipitation of endogenous UHRF1 with transfected Eme1-V5. This strongly suggests that these two proteins interact in mammalian cells.
To determine the region of Eme1 that interacts with UHRF1, four overlapping deletion mutants of Eme1 fused to the DNA-binding domain of GAL4 were created and their ability to interact with UHRF1 were assessed using the yeast-two hybrid assay. Eme1 deletion mutants I and II were able to show an interaction with UHRF1. This indicates that amino acids 40-187 of
Eme1 are sufficient for an interaction with UHRF1. No functional domain has been indicated in this region of Eme1. This experiment is the first to suggest a functional role in this region for
Eme1. It would be worthwhile further investigating this interaction by creating more deletion mutants to further dissect the exact amino acids needed for the interaction.
Given that both Eme1 and UHRF1 play a role in the DNA damage response, and that we see an interaction between the two proteins using three independent methodologies, we next sought to determine if the interaction of these two proteins was impacted by the presence of
DNA damage. Mus81-Eme1 has been shown to play a role the resolution of DNA replication fork intermediates in the presence of DNA damage, whereas cells deficient of UHRF1 have been
70 shown to be more sensitive to DNA damaging agents that include agents that damage replication forks (Muto et al., 2002). Taken together, we expected to see co-localization of the two proteins in the presence of replication fork damaging. We used NIH 3T3 cells as UHRF1 has been shown to localize to DAPI-dense chromocenters that are representative of heterochromatin (Citterio et al., 2004; Karagianni et al., 2008; Papait et al., 2007). We transiently transfected NIH3T3 cells with FLAG-UHRF1 and Eme1-V5 and observed UHRF1 staining is predominant at DAPI dense areas (representative of areas of heterochromatin), while Eme1 staining is strongest in DAPI light areas (representative of euchromatin), in untreated cells. Since Mus81-Eme1 has been shown to play a role in replication associated damage, and UHRF-deficient cells display increased sensitivity to agents that perturb replication forks, we expected to see co-localization between the two proteins when our cells were exposed to replication associated damaging agents.
We observed, following CPT treatment co-localization between Eme1 and UHRF1 in a nuclear speckle pattern. UHRF1 ring finger mutants failed to co-localize with Eme1 following CPT treatment. This indicates that the ring finger domain is needed for Eme1-UHRF1 co-localization following DNA damage by CPT. This finding points to other roles for the ring finger domain outside of its known role as a domain responsible for endowing UHRF1 with E3 ubiquitin ligase activity. Future studies should aim to repeat this experiment using UHRF1 constructs that are lacking its other protein domains. This would help us understand if the Eme1-UHRF1 interaction is contingent on its other protein domains. Cells deficient of UHRF1 have been shown to be more sensitive to DNA damaging agents, and in particular cells that are over expressing UHRF1 ring finger mutants that are deficient in ubiquitin ligase activation are more sensitive to damaging agents (Jenkins et al., 2005; Muto et al., 2006; Muto et al., 2002). It is tempting to speculate that given the chromatin modification properties of UHRF1 that are heterochromatin
71 associated may also have a role in DNA repair. UHRF1‟s ring finger domain may play a recruitment role for Eme1 following DNA damage to replication centers in the nucleus.
Although UHRF1 and Eme1 were found to co-localize following replication-associated damage, the reasons for this co-localization remain unclear. One possible explanation for this interaction is that UHRF1 interacts with Eme1 during the DNA damage response in order to remove Eme1 following completion of DNA repair by ubiquitin-mediated proteolysis. In support of this finding, we noted that Western analysis of NIH 3T3 cells transfected with Eme1-
V5 and FLAG-UHRF1 revealed that transfected steady-state Eme1-V5 levels always appeared lower when cells were transfected with FLAG-UHRF1 (figure 20). Subsequent studies demonstrated that the protein half-life of Eme1-V5 was substantially reduced in the presence of
FLAG-UHRF1 (Lianne Gibson and J.P. McPherson, personal communication). In order to determine whether Eme1 was ubiquitinated by UHRF1, Eme1-V5 together with Ub-HA and
FLAG-UHRF1 were expressed in NIH 3T3 cells and Eme1-V5 immunoprecipitates were evaluated for incorporated Ub-HA by Western analysis using an anti-HA antibody. Eme1-V5 was found to be ubiquitinated in cells; however, co-transfection with FLAG-UHRF1 and Ub-HA resulted in decreased levels of ubiquitinated Eme1-V5 (Lianne Gibson and J.P. McPherson, personal communication). Reduced ubiquitinated Eme1-V5 was also observed when cells were transfected with FLAG-UHRF1 RING finger mutants. Taken together, these findings do not support a role for UHRF1 mediating proteolytic degradation via ubiquitin conjugation.
Our experiments suggest that UHRF1 and Eme1 might cooperate together to facilitate repair to certain forms of DNA damage. Whether these proteins are playing a role in the same pathway or different pathways remain to be formally explored. The creation of cell lines that are depleted in either protein, along with a cell line that is depleted in both proteins would give us
72
the tools needed to answer this question. We could then conduct clonogenic assays comparing survival following replication-associated damage in wild-type, Eme1 deficient cells, UHRF1 deficient cells, and cells deficient in both proteins (Pamidi et al., 2007). If both proteins are playing separate roles to repair DNA damage, then we would expect to see increased sensitivity in the Eme1-UHRF1 deficient cells compared to the other cell lines deficient for either Eme1 or
UHRF1. Currently, we are unable to reconcile the observed loss of Eme1-V5 half-life in the presence of FLAG-UHRF1 with the observed co-localization and interaction of these proteins. It is not clear how or why UHRF1 facilitates degradation of Eme1-V5; however, this observation might help to explain why there is very little Eme1-V5 found together with FLAG-UHRF1 in our co-immunoprecipitation studies. Our attempts to construct a mechanistic explanation for the observed interaction between UHRF1 and Eme1 are also limited by our understanding of
UHRF1 and its role in DNA metabolism in the absence and presence of DNA damage. My thesis research detailed in Chapters 3 and 4 were conducted to assess what role, if any, UHRF1 plays in the cellular response to DNA damage.
Significance and Impact
Our findings were the first to identify a novel protein-protein interaction involving Eme1 in addition to its previously described interaction with Mus81. We confirm this interaction using three independent methodologies: yeast-two hybrid, GST pulldown and co-immunoprecipitation.
We determined that the N-terminal domain of Eme1 that is necessary for the interaction between the two proteins. Lastly, we observe co-localization between the two proteins that occurs upon
CPT treatment, and that is dependent on the RING finger domain. These findings suggest that
DNA damage repaired by Mus81-Eme1 may require chromatin modifications mediated by
74
UHRF1 and that Mus81-Eme1 may participate in repair of replication-associated DNA damage in heterochromatin. Our findings also support a potential role for UHRF1 in the DNA damage response, a possibility that we explore further in Chapters 3 and 4. Findings from this chapter were published in Mistry et al. BBRC 375 (2008) 321-325 (Mistry et al., 2008).
75
CHAPTER THREE: UHRF1 facilitates the DNA damage response to γ-Irradiation.
Modified version of material from this chapter has been published in the following report:
Helena Mistry*, Laura Tamblyn*, Hussein Butt, Daniel Sisgoreo, Aileen Gracias, Meghan Larin, Kalpana Gopalakrishnan, Manoor Prakash Hande and John Peter McPherson. UHRF1 is a genome caretaker that facilitates the DNA damage response to γ-irradiation. Genome Integrity (BMC) 2010 Jun 8;1(1):7.
76
Introduction
DNA double strand breaks (DSBs) caused by ionizing radiation and radiomimetic agents are considered to be one of the most deleterious forms of DNA damage (Coupier et al., 2004). A network of DNA surveillance and signalling factors detects the presence of DSBs and directs the cell to halt cell cycle progression in order to rectify damage. Under certain conditions, DSBs are not repaired and cells are instructed to undergo apoptosis or senescence. Cells deficient in these surveillance, signalling and repair factors show defects in directing the DNA damage response, rendering them more sensitive to ionizing radiation and radiomimetic drugs that create DSBs.
Earlier studies have demonstrated that loss of UHRF1 in mouse embryonic stem cells can render cells more sensitive to radiation; however, the cause for this sensitization remained unclear
(Muto et al., 2002).
Objective: The objective of chapter 3 of my thesis focuses on identifying the role (if any) that
UHRF1 plays in the DNA damage response to γ-irradiation.
Outline: In order to investigate the role of UHRF1 within the DNA damage response when cells are exposed to γ-irradiation, we created three HeLa cell lines stably depleted of UHRF1. Three clonally-derived control lines stably expressing nonspecific shRNA (directed against luciferase) and three clonally-derived lines expressing UHRF1-specific shRNA (with concomitant depletion of UHRF1) were generated and confirmed by Western analysis. Clonogenic assays revealed an increased sensitivity of UHRF1-depleted cells to γ-irradiation compared to control cells. We also demonstrated that UHRF1 protects cells from agents that cause replication-associated DNA damage, in that UHRF1-depleted cells were more sensitive to replication fork inhibitors Ara-C and CPT compared to control cells. Examination of cell cycle parameters indicated that UHRF1- depleted cells show an attenuated G2/M arrest in response to γ-irradiation, indicative of a
77 compromised cellular response to DNA damage. This compromised response was further validated by quantification of cells with histone H2AX phosphorylation on serine 139 (γH2AX).
UHRF1-depletion showed a marked decrease in cells with elevated γH2AX compared to control cells following γ-irradiation. To determine whether an innate defect in genome integrity accompanied the compromised damage response in UHRF1-depleted cells, untreated control and depleted cell lines were evaluated for chromosome integrity. Three independent methodologies demonstrated that depletion of UHRF1 increases the incidence of chromosome aberrations.
Taken together, these findings point to a role for UHRF1 in maintenance of an optimal cellular response to DNA damage and a caretaker role for this protein in genome integrity.
Methods
Cell lines Human HeLa cells (ATCC) were either transiently or stably transfected with one of four short hairpin RNA shRNA expression plasmids against UHRF1 (HuSH 29mer shRNA constructs against UHRF1, #TR308492, Origene, Rockville MD). The shRNA sequences of the four plasmids are as follows:
ID Sequence
TI333961 5‟-GAT TCT CTG AAC GAC TGT CGG ATC ATC TT-3‟ TI333962 5‟-AGG AGA CGT TCC AGT GTA TCT GCT GTC AG-3‟ TI333963 5‟-TTC GTG GAC GAA GTC TTC AAG ATT GAG CG-3‟ TI333964 5‟-CTG TGG AAT GAG GTC CTG GCG TCA CTC AA-3‟
Control HeLa clonal cell lines were also established following stable transfection of a control shRNA plasmid (pRS- shGFP, non-effective, Origene) which expresses a non-effective luciferase shRNA. All stable transfectants were selected for resistance to puromycin conferred by co-transfection of shRNA plasmids with pPUR. Cell colonies were cloned, expanded and
78 examined for reduced steady-state levels of UHRF1 by Western analysis (see below). Three
HeLa cell lines with reduced UHRF1 (two lines from TI333962, one line from TI333963 transfections) were chosen for further analysis together with three HeLa cell lines stably transfected with the control shRNA plasmid.
Growth Curve Growth curves were generated by seeding cells at 3.5x 105 cells per 10 cm dish. Cells were then counted using a hemocytometer at three-day intervals for thirteen days. The growth curves represent the cumulative mean cell number ± standard deviation of five counts
(performed by Meghan Larin).
Western Blotting For Western analysis, cells were lysed in RIPA buffer (50 mM Tris-HCL pH 8.0, 150 mM NaCL, 1 mM PMSF, 1 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS,
5 µg/ml aprotinin, 5 µg/ml leupeptin, and 5 µg/ml pepstatin). Protein quantification was done by
Bradford analysis to load equal amounts of protein. Lysates were then resolved on 12 % SDS-
PAGE gels and transferred onto PVDF membranes overnight at 4 ºC. Membranes were then blocked in Tris-buffered saline / 5% milk / 0.1% Tween, and then immunoblotted with the following primary antibodies indicated in Table 2, followed by incubation with the appropriate
HRP-linked secondary antibody (anti-rabbit or anti-mouse HRP-linked IgG, GE Healthcare).
Protein detection was performed using the ECL Western blotting detection system (GE
Healthcare) (assisted by Hussein Butt).
79
Antigen Company Source Dilution
H2A.X (phosphoS139) Millipore mouse monoclonal 1:1000
H3K9me3 Upstate (Millipore) rabbit polyclonal 1:500
H4K20me3 Upstate (Millipore) rabbit polyclonal 1:500
α-Tubulin Calbiochem mouse monoclonal 1:1000
UHRF1 BD transduction rabbit polyclonal 1:500
Table 3. Antibodies used for Immunoprecipitation and Western experiments.
Antibodies were incubated overnight at 4 ˚C to detect the proteins of interest.
Clonogenic assays 1×103 cells were seeded in 60mm dishes in Dulbecco‟s Minimum Essential Media supplemented with 10% Fetal Bovine Serum (Invitrogen) and exposed to varying doses of either
CPT, Ara-C (Sigma), mitomycin C (Sigma) or γ-irradiation (Nordion Gamma-cell, Ontario
Cancer Institute) and left to grow for seven days. Colonies were fixed and stained with methylene blue in methanol. All survival curves were produced from the means ±standard deviation of four to six determinations and are presented as a percent of control (non-treated) cells (assisted by Hussein Butt and Aileen Gracias).
Flow cytometry (FACS analysis) Cells were exposed to either 1 Gy or 5 Gy of irradiation and collected at 0, 1, 3, 6, 12 or
24 h after exposure, fixed in 70% ice cold ethanol and stored at -20 ºC prior to staining in triplicate. Cells were then washed in PBS, gently vortexed in PBS + 0.4% Triton X and then washed again in PBS. Cells were incubated in the dark for 30 min at room temperature with anti-γH2AX (1:200, Millipore) in PBS + 0.2% Triton X / 1% donkey serum. Cells were then
80 washed again with PBS and incubated in the appropriate secondary antibody (Jackson
ImmunoResearch Labs) for 30 min at room temperature. Following one final wash, cells were stained with Propidium Iodide (PI) (50 µg/ml in PBS) for 30 min at room temperature. Flow cytometry data was acquired using a FACSCalibur flow cytometer (BD Biosciences, located in the Department of Immunology, University of Toronto) and analysis was performed using
FLOJO software (Treestar) (performed by Laura Tamblyn).
Karyotype analysis Slides were hydrated in PBS and fixed in 4% v/v formaldehyde: PBS, washed 3 times for
5 min each wash. Slides were then treated with 0.1 mg/mL pepsin (Sigma) at pH 2 for 3 min, followed by repeated formaldehyde fixation and PBS washes before dehydration in an ethanol series (70%, 90% and 100%). Slides were then air-dried, then denatured at 80 º C for 3 min with hybridization solution containing deionised formaldehyde, 0.5 μg/mL Cy-3-conjugated-
(CCCTAA)3 peptide nucleic acid probe complementary to telomeric sequence and 3 μg/mL fluorescein isothiocyanate (FITC)-conjugated-centromic peptide nucleic acid probe (Applied
Biosystems) in 10mM Tris (pH 7). Hybridization was performed in the dark for 2 h at room temperature. Slides were then washed 2 × 15 min in 70% v/v formamide (Merck) / 1% w/v BSA
/ 10 mM Tris(pH 7.4) and 3 × 5 min in 100 mM Tris / 150 mM NaCl (pH 7.2) / 0.08% v/v
Tween 20 (Sigma). Slides were dehydrated in an ethanol series, air-dried in the dark and counterstained with 0.0375 μg/mL DAPI in mounting media (Vectashield; Vector Laboratories).
Images from approximately 50 metaphases were captured using an Axioplan 2 imaging fluorescence microscope (Zeiss) and analyzed for chromosomal aberrations such as breaks or fusions with Isis Imaging Software (Metasystems, Germany) (Performed Kalpana
Gopalakrishnan and Dr. Manoor Prakash Hande).
81
Results
In order to generate cell lines with permanent depletion of UHRF1, HeLa cells were stably transfected with plasmids expressing one of four shRNAs against UHRF1. Following screening of clonal isolates, three HeLa cell lines were obtained with reduced expression of
UHRF1 protein, when compared to UHRF1 expression in control lines (figure 21). HeLa cell lines identified lines derived from two different shRNA plasmids.
Previous studies reported that loss of UHRF1 can impede cellular proliferation in certain cell lines. To assess if UHRF1 depleted HeLa lines showed impaired cellular proliferation, we compared proliferative capacity of two UHRF1-depleted cell lines (UHRF1 shRNA 1 and 2) to two control-shRNA transfected cell lines (control shRNA 1 and 2). UHRF1-depleted cell lines were shown to proliferate at a slightly slower rate than control cells, an effect that was only discernible following several days of analysis (figure 22).
Previous studies have shown that UHRF1 protects cells from the cytotoxic effects of radiation (Muto et al., 2006). We validated these findings with our UHRF1-depleted cell lines by clonogenic assay. As predicted, both UHRF1-depleted cell lines displayed an increased sensitivity to γ-irradiation compared to control lines, indicating that UHRF1 is playing some role in protecting cells from the cytotoxic impact of irradiation exposure (figure 23). We have also found that UHRF1-depleted cells are more sensitive to varying doses of replication fork damaging agents (Ara-C and CPT) (figure 24) when compared to control cells.
We wished to determine whether the increased radiosensitivity of UHRF1-depleted cells could be due to a compromised response to DNA damage caused by radiation exposure. Given that HeLa cells typically accumulate in the G2/M phase of the cell cycle in response to radiation, we quantified the percentage of cells in various stages of the cell cycle (G1, S, G2/M) before and
82
at various times following exposure of control or UHRF1-depleted cells to 1 or 5 Gy of γ- irradiation. Following irradiation, cells were collected at 0, 3, 6, 12 and 24 hr time points. Cells were then prepared for analysis of BrdU incorporation and propidum iodide (PI) staining. Our results indicated that in untreated cells, the UHRF1-depleted cells had a higher percentage of cells in the G1 phase and decreased percentage in the S phase of the cell cycle when compared to the control cells (figure 25). In un irradiated control cells, 34.1 ±0.7% of cells (mean ±s.d.) were in G1 phase, 46.5 ±0.8% of cells were in S phase and 11.9 ±0.3% of cells were in G2/M phase. In contrast, 51.2 ±1.4% of un irradiated UHRF1-depleted cells were in G1 phase, 30.9
±1.5% were in S phase and 7.2 ±0.7% were in G2/M phase. As expected, when control cells were exposed to 1 Gy of irradiation, cells accumulated in the G2/M phase of the cell cycle that was maximal at 12 hours following irradiation (41% in the control compared to 18% in the
UHRF1-depleted cell line, figure 25A). There was an approximate three and a half increase in the number of cells arrested in G2/M phase compared to untreated cells for control cells. In contrast, UHRF1-depleted cells showed a marked decrease in their ability to arrest in the G2/M phase of the cell cycle, in that there was only a two and a half fold increase in the G2/M phase fraction with no obvious change in the number of cells in S phase 12 hours after irradiation when compared to untreated cells (29% in untreated and 37% 12 hours after IR). Twenty four hours after irradiation, the cell cycle resumed in both cell lines (figure 25). When both cell lines were exposed to 5 Gy of irradiation, the cell cycle arrest experienced was more pronounced with the accumulation of cells in the G2/M phase of the cell cycle. For the control cells, 12 hours following 5 Gy of irradiation resulted in the accumulation of cells in the control cells in the
G2/M phase of the cell cycle. This accumulation was delayed in the UHRF1-depleted cell line
(28% in the control as opposed to 8% in the UHRF1-depleted cells). Failure of the UHRF1-
87
depleted cells to arrest following irradiation treatment suggests that these cells have a defect in the G2/M cell cycle checkpoint (figure 25C-E).
Given the defects seen in checkpoint activation, we decided to determine whether the activation was linked to changes in signal transduction pathways activated following irradiation.
Using FACS analysis, we measured the kinetics of γH2AX induction and disappearance following exposure of control and UHRF1-depleted cell lines to either 1 Gy or 5 Gy at 0, 1, 3, 6,
12 and 24 hour time points. As expected, 1 Gy exposure resulted in an increased number of control cells with elevated γH2AX after irradiation that was maximal at 6h (figure 26A).
Following 5 Gy, a rapid increase in control cells with elevated γH2AX was apparent after 1 hour
(figure 26B). In contrast, UHRF1-depleted cells exposed to 1 Gy or 5 Gy showed a drastically reduced number of cells with elevated γH2AX compared to control cells (figure 26A and B).
These results were confirmed by detecting γH2AX foci formation by indirect immunofluorescence (representative pictograph is shown for cells 12 h after 5 Gy in figure
26C).
Next, we sought to determine if our UHRF1-depleted cells displayed signs of innate chromosomal instability. We examined three biomarkers of chromosomal instability: the presence of micronuclei, supernumerary centrosomes and aberrations in metaphase chromosomes. We scored the presence of micronuclei in our control and UHRF1-depleted cell lines by indirect immunofluorescence following staining of cells with DAPI. Micronuclei are extranuclear chromosomal fragments in daughter cells that result from improper chromosomal segregation during mitosis (Hoffelder, 2004; Therman, 1993). Following DAPI staining, micronuclei present in interphase cells as small circular staining species closely juxtaposed to the nucleus (figure 27A-D). For these experiments, 3 control and 3 UHRF1-depleted cell lines
89
were DAPI stained and scored for the incidence of micronuclei. UHRF1-depleted cells displayed a 3-fold higher incidence of micronuclei compared to control cells (figure 27E).
As changes in centrosome copy number often appear in cells with chromosomal instability, we queried whether UHRF1-depleted cells would show changes in centrosome number compared to control cells (Thompson et al., 2010). We immunostained control and
UHRF1-depleted cells for the centrosome protein γ-tubulin and scored centrosome copy number following examination of cells by microscopy. Quantification of centrosome copy number following immunostaining with γ-tubulin revealed that the number of UHRF1-depleted cells with 2 or more centrosomes was increased compared to control cells (figure 27F).
We conducted karyotypic analysis of control and UHRF1-depleted cells to qualitate and further quantitate the compromised chromosomal stability of UHRF1-depleted cells.
Karyotyping of metaphase chromosomes from three independently-derived clonal control and
UHRF1-depleted cell lines was conducted following labelling the ends of chromosomes
(telomeric red Cy3-derived signals and centromeric green FITC-derived signals in figure 27G-
I). As anticipated, UHRF1-depleted cells showed elevated levels of chromosomal aberrations compared to control cells, especially chromosome fragments (figure 27G-I and table 3).
Discussion
Our findings in Chapter Two describe an interaction between UHRF1 and Eme1 that led us to postulate that both proteins may participate in a common pathway in response to DNA damage. However, the role if any of UHRF1 in such a pathway remained to be determined.
In order to investigate this possibility, we have created and characterized cell lines
92 depleted for UHRF1 using a siRNA approach mediated by stable transfection and expression of shRNAs against UHRF1.
TABLE 3. Increased chromosomal instability in UHRF1-depleted cellsa.
Cell line Chromosomal Chromatid Fusions Fragments Total Breaks Breaks
Control 0 0 1 8 9 shRNA-1 (0.02/cell) (0.16/cell) (0.18/cell)
Control 0 0 2 12 14 shRNA-2 (0.04/cell) (0.24/cell) (0.28/cell)
Control 2 0 3 10 15 shRNA-3 (0.04/cell) (0.06/cell) (0.2/cell) (0.3/cell)
UHRF1 1 0 5 47 53 shRNA-1 (0.02/cell) (0.1/cell) (0.94/cell) (1.06/cell)
UHRF1 1 0 5 51 57 shRNA-2 (0.02/cell) (0.1/cell) (1.02/cell) (1.14/cell)
UHRF1 9 1 1 54 65 shRNA-3 (0.18/cell) (0.02/cell) (0.02/cell) (1.08/cell) (1.3/cell)
aResults shown are number of chromosomal abnormalities detected in fifty metaphase spreads from three control shRNA and three UHRF1 shRNA (UHRF1-depleted) cell lines. The frequency of a given abnormality/metaphase is shown below in parentheses
93
Previous studies utilizing strategies to deplete UHRF1 in cells reported wide-ranging effects of UHRF1 on proliferation. Compared to previous studies, we noted only a mild decrease in proliferation that is discernable after several days of culture. There is no clear explanation for the wide-ranging impact of UHRF1 on proliferation; however, it is interesting to note that studies describing a profound block in proliferation typically are utilizing primary or simple-step immortalized cells with a strategy to transiently deplete UHRF1 using antibodies or siRNA
(Papait, 2008). In contrast, our studies and others where UHRF1 is stably depleted, do not show as drastic an impact on proliferation (Karagianni, 2008) (Mistry, 2010). Differences in proliferation may occur due to differences in the packing properties of heterochromatin between different cell lines. Chromatin architecture changes may also affect the cell cycle, as well as chromatin accessibility has been shown to be a determinant of DNA replication timing (Black and Whetstine, 2011; Hansen et al., 2010; Karnani et al., 2010; MacAlpine et al., 2004;
MacAlpine et al., 2010). The impediment to proliferate may also occur due to disrupted heterochromatin dynamics which may impact S phase progression during the cell cycle (Black and Whetstine, 2011). Finally, heterochromatin is known to regulate diverse chromosomal processes such as transcriptional silencing. Altered heterochromatin dynamics in our UHRF1- depleted cells could result in altered transcription that modifies the cell cycle.
Next we monitored our cells for their ability to arrest in the cell cycle. Our untreated
UHRF1- depleted cells displayed an increase in the number of cells in G1 phase and a decrease in the S phase of the cell cycle. A previous study noted an increase in G1 cells following exposure to doxorubicin which they attributed to an increased G1 checkpoint response to DNA damage that was revealed with UHRF1 depletion (Arima et al., 2004). When we challenged our
94 cells with irradiation, we also saw an increase in the percentage of cells in the G1 phase. We attribute this change to the profile of our cells prior to damage and not to a novel checkpoint. A loss of UHRF1 has been shown to decrease the DNA methylation status in cells, and it could result in a loss of the propagation of epigenetic cues (e.g. H3K9me3) (Sharif et al., 2007). A disruption in these important pathways could lead to the cell not progressing effectively through the cell cycle and remaining in the G1 phase of the cell.
UHRF1 has been postulated to protect cells from DNA damage as murine UHRF1- depleted embryonic stem cells display increased sensitivity to DNA damaging agents that include irradiation (Muto et al., 2002). Our observations confirmed these findings as UHRF1- depleted cells displayed increased sensitivity to irradiation and replication inhibitors. The reason for this sensitivity has previously not been explored. We discovered that following irradiation, our UHRF1-depleted cells displayed a defect in the ability to accumulate γH2AX compared to control cells.
Given UHRF1‟s role in heterochromatin replication, it seems plausible that this protein might participate in the DNA damage response specific for heterochromatin. Previous studies have shown that formation of γH2AX foci occurs preferentially in euchromatin and not heterochromatin (Cowell et al., 2007). γH2AX formation may be blocked in heterochromatin due to its compressed structure, and this can be relieved or at least partially relieved, when the cell is undergoing replication when the chromatin structure is less compact (Cowell et al., 2007).
An „open‟ chromatin conformation by a depletion in histone H1 or treatment with HDAC inhibitors has been shown to render cells resistant to DNA damage (Clausell et al., 2009). In both situations, cells have a “super-sensitive” DNA damage response when compared to control cells. We propose that UHRF1 plays a role in regulating changes in the packing properties of
95 heterochromatin needed to facilitate access to DNA damage. Loss of UHRF1 might hinder the cell‟s ability to correctly detect and repair DNA damage located within heterochromatin. Our findings indicate that UHRF1 is a genome caretaker that facilitates the DNA damage response to irradiation. This work was published in Mistry et al. (Helena Mistry, 2010). Thesis research in the following chapter seeks to understand whether the role of UHRF1 in heterochromatin is linked with the response to DNA damage.
Significance and Impact Our findings for chapter 3 indicate for the first time definitive data demonstrating a role for UHRF1 in (i) the maintenance of chromosome integrity and (ii) an optimal response to DSB damage. We also show for the first time that depletion of UHRF1 confers sensitivity to CPT and
Ara-C, replication inhibitors known to result in DSBs. Given that UHRF1 has recently been shown to play a critical role in the maintenance of heterochromatin, we propose that the disruption of heterochromatin following the loss of UHRF1 may compromise the DNA damage response.
96
CHAPTER FOUR: UHRF1, heterochromatin and the DNA damage response
97
Introduction:
Rationale: The highly ordered nature of heterochromatin serves as a structural barrier to efficient DNA repair. Heterochromatin is composed of DNA with a higher frequency of methylated CpG dinucleotides and histones with signature modifications, such as histone
H3K9me3 and H4 trimethylated specifically at lysine 20 (H4K20me3) that serve as docking sites for heterochromatin binding proteins such as HP1 proteins that recognize and bind the epigenetic code of heterochromatic histones (Grewal and Rice, 2004; Kouzarides, 2007). It has been shown that maintenance of H3K9me3 is important for maintenance of genome integrity as the activity of H3K9 specific HMKT‟s (Suv39H1 and Suv39H2) have been shown to be critical for maintenance of genomic integrity (Peters et al., 2001). Until recently, the mechanistic connection between H3K9me3 status and signalling pathways that respond to DNA damage and maintain genome integrity remained obscure. Important recent studies have now shed light on this connection; by demonstrating that unmasking of H3K9me3 sites in heterochromatin are required for activating Tip60 which then facilitates the activation of ATM (Sun et al., 2005).
Ejection of HP1β from its binding sites containing H3K9me3 adjacent to DSBs are exposed following DNA damage allowing activation of Tip60 complexed with ATM, which then acetylates ATM to activate this kinase together with acetylation of histones that facilitate a chromatin state amenable to repair (Figure 28). We propose that this model provides a possible mechanistic explanation as to how UHRF1 impacts the DNA damage response, given the recently described roles for this protein in maintenance of heterochromatic H3K9me3
(Karagianni, 2008).
98
Objective: To determine if UHRF1 plays a role in a DNA damage response pathway that is mediated by H3K9me3 and results in Tip60 activation within heterochromatin.
Outline: Since H3K9me3 is a docking site for HP1β, and HP1β ejection from heterochromatin has been shown to lead to Tip60 activation we decided to investigate the status of HP1β and compare it to the status of its isoform HP1α in our UHRF1-depleted and control cells. Using indirect immunofluorescence we show that isoform HP1β and not HP1α is lost from heterochromatin in our UHRF1-depleted cells when exposed to γ-irradiation as well as the replication fork inhibitor CPT. Tip60 is known to acetylate various histones including histone
H2A at lysine 5 (acetyl-H2AK5). In order to assess whether acetyl-H2AK5 levels increased in cells with DNA damage we investigated the levels of this modification in control and UHRF1- depleted cells following exposure to γ-radiation. Our findings showed an increase in acetyl-
H2AK5 staining in cells following γ-irradiation that was conspicuously absent in UHRF1- depleted cells. We also observed a decreased activation of ATM in UHRF1-depleted cells following γ-irradiation compared to control cells under the same conditions. Taken together, our findings suggest that the loss of heterochromatic H3K9me3 in UHRF1-depleted cells disrupts a
DNA damage response pathway required for histone acetylation and ATM kinase activation.
Methods:
Indirect Immunofluorescence UHRF1-depleted and control cells generated from HeLa cells were used for immunofluorescence by seeding 5 × 104 cells onto coverslips (pre-coated with 1% gelatin for 15 min followed by 1% BSA for 1 h). Cells were fixed in 1:1 methanol/acetone for ten min and permeabilized with 0.4 % Triton-X 100 in PBS for 20 minutes, following by blocking (1%
100 donkey serum/0.2% Triton-X 100) for 20 minutes. Cells were then incubated overnight with a
1:500 dilution of either; anti-H3K9me3, anti-H4K20me3, anti-Hp1α, or anti-Hp1β (all from
Millipore), or anti-acetyl H2AK5 (Cell Signalling) overnight. The following day, cells were incubated with FITC- or TRITC-conjugated antibody (Jackson ImmunoResearch Labs), counterstained with DAPI and mounted with Vectashield (Vector laboratories). Images were acquired using an Imager.ZI fluorescence microscope and Axiovision software (Zeiss) following deconvolution using Autolinear and Clip algorithms (assisted by Daniel Sisgoreo).
Western Blotting to detect phospho-ATM Cells were seeded onto 10 cm plates and suspended in lysis buffer (20 mM Hepes, pH
7.4, 150 mM NaCl, 5% glycerol, 0.1% NP-40, 0.1% -mercaptoethanol, 0.5 mM PMSF, 5
µg/ml aprotinin,5 µg/ml leupeptin, 5 µg/ml pepstatin, 1mM NaF, 1mM NaVO4). Lysates were separated on either 10% or 6% SDS-PAGE gels and transferred onto PVDF membrane overnight at 4 ºC. Membranes were blocked in Tris-buffered saline / 1% Bovine Serum Albumin / 0.1%
Tween for 1 h. Membranes were then incubated overnight at 4˚C with anti-phosphoserine1981
ATM (Cell signalling pS1981, 1:1000), followed by incubation with anti mouse HRP-linked secondary antibody (GE Healthcare). Protein detection was performed using the ECL Western blotting detection system (GE Healthcare).
Results
As UHRF1 was previously reported to contribute to the maintenance of the H3K9me3 mark on heterochromatin, we decided to investigate heterochromatin markers H3K9me3 and
H4K20me3 in our UHRF1-depleted and control cells with the use of western analysis and indirect immunofluorescence. UHRF1-depleted and control cells exhibited equal protein levels
101 when untreated and exposed to 5 Gy of γ-irradiation (figure 29). As expected, control HeLa cell lines exhibited H3K9me3 and H4K20me3 superimposed onto DAPI-dense areas that represent areas of heterochromatin (figure 30). In contrast, UHRF1-depleted cell lines showed a drastic redistribution of chromatin-bound H3K9me3 staining to the peripheral area of the nucleus, whereas H4K20me3 remained superimposed on DAPI-dense areas but H4K20me3 foci seemed to be fewer in number/nucleus but larger in size (figure 30B). Our laboratory is currently attempting to quantify this change. The observed redistribution of heterochromatic
H3K9me3 did not appear to be a coincidental secondary event that occurred following generation of UHRF1-depleted cells as we were able to successfully rescue the defective
H3K9me3 observed in UHRF1-depleted cells (figure 31). We observed that focal H3K9me3 staining was restored in UHRF1-depleted cells following transient transfection with a murine
FLAG epitope tagged UHRF1 expression vector (FLAG-UHRF1). Rescue experiments were performed with murine UHRF1 cDNA as expression from this plasmid was predicted to be refractory to the effect of the stably-transfected shRNA directed against human UHRF1.
As Hp1β was previously reported to be released from H3K9me3-containing chromatin following DNA damage, we investigated the status of Hp1α and Hp1β in control and UHRF1- depleted cells before and after 1 h of 1Gy or 5 Gy of γ-irradiation (figure 32). Using indirect immunofluorescence, we were able to visualize focal staining for both HP1 isoforms in control untreated cells which represents the chromatin-bound fraction of Hp1α and Hp1β. One hour following irradiation, control cells were observed to retain focal staining of Hp1α but lost the chromatin-bound Hp1β fraction. Focal staining of Hp1α was similar in UHRF1-depleted cells before and after irradiation, similar to control cells. In contrast to control cells, focal Hp1β staining was absent in untreated UHRF1-depleted cells before and after irradiation (figure 32).
102
Figure 31. Rescue of H3K9me3 status in UHRF1-depleted cells
A. Rescue of focal histone H3K9me3 in UHRF1-depleted cells transfected with murine UHRF1 cDNA as shown by indirect immunofluoresence. B. Expression of murine UHRF1-FLAG was confirmed in these cells by Western analysis (tubulin loading control shown).
105
Similar results were obtained when cells were examined following 3 h exposure to 500 nM CPT
(figure 33).
Tip60 HAT activity has been reported to be activated following its recruitment to DNA damage sites and binding of the Tip60 CD to H3K9me3 unmasked by H3K9me3 (Sun, 2009).
As reagents to measure Tip60 HAT activity were not available, we sought to determine whether acetylation of Tip60 substrates was altered in UHRF1-depleted cells following DNA damage.
Tip60 has been reported to be the enzyme solely responsible for acetylation of histone 2A lysine
5 (acetyl-H2AK5) (Kimura and Horikoshi, 1998). Accordingly, we investigated the presence of acetyl-H2AK5 in control and UHRF1-depleted cells by indirect immunofluorescence (figure
34). Chromatin-bound H2AK5 was poorly detected in control and UHRF1-depleted cells before
γ-irradiation. Levels of acetyl-H2AK5 were observed to markedly increase in control cells 1 hour after 5 Gy irradiation but a corresponding increase was absent in UHRF1-depleted cells.
As compromised Tip60 activation has been reported to attenuate ATM activation via lack of Tip60-dependent ATM acetylation, we investigated ATM activation in control and UHRF1- depleted cells by measuring levels of autophosphorylated ATM (phosphoSer1981-ATM) by
Western analysis following irradiation (Sun et al., 2009) (figure 35). Western analysis of protein lysates from cells before and after 1 h of 5 Gy irradiation revealed increased levels of phosphoSer1981-ATM in control cells following 5 Gy compared to untreated cells, whereas
UHRF1 depleted cells displayed low levels of phosphoSer1981-ATM before and after irradiation.
107
Discussion
Tightly packed heterochromatin can act as a barrier to DNA repair in the presence of damage. In order for mammalian cells to fix this damage, chromatin must be relaxed in order for repair factors to access to the damage. An important mediator of this relaxation is HAT activity directed at histones within the vicinity of the DNA lesion (Mahlknecht and Hoelzer, 2000).
Upon completion of repair, the chromatin structure is restored to its original state via deacetylation of histones by HDACs (Kristensen et al., 2009). Tip60 HAT activation and subsequent ATM activation was shown to be initiated by the unmasking of H3K9me3 in heterochromatin by ejection of Hp1β. Accordingly we reasoned that loss of heterochromatic
H3K9me3 following UHRF1 depletion might compromise the DNA damage response in these cells through inhibition of this pathway.
We observed a specific depletion of H3K9me3 associated with heterochromatin as has been reported previously in HeLa cell lines stably-depleted of UHRF1 using an shRNA strategy
(Karagianni, 2008). In the report by Karagianni and colleagues, total levels of H3K9me3 are equivalent yet accumulation of this modification is lost in heterochromatin. These findings contradict another study which concluded that UHRF1 depletion did not alter the distribution of heterochromatin-associated histone methylation (Papait, 2008). These differences might be explained by the different experimental strategies used. Our study and those of Karagianni and colleagues used HeLa cell lines with UHRF1 stably-depleted by shRNA whereas the studies of
Papait and colleagues utilized transient depletion of UHRF1 by siRNA. It is conceivable that transient depletion of UHRF1 over a short period of time might not alter heterochromatic
111 distribution of H3K9me3. Although the PHD and/or the tandem tudor domains of UHRF1 have been shown to bind H3K9me3, the mechanism whereby UHRF1 retains H3K9me3 on heterochromatin remains to be determined (Karagianni, 2008; Rottach, 2009). One possibility is that UHRF1 recruits a histone trimethyltransferase such as Suv39H to heterochromatin in a manner similar to the recruitment of G9a histone H3 dimethyltransferase (Kim, 2008).
HP1 proteins specifically bind to H3K9me3 via their chromodomains; however, upon
DNA damage it has been reported that CK2 phosphorylates Hp1β specifically on Thr51, which inactivates the ability of its chromodomain to bind to H3K9me3. Ejection of Hp1β was found to precede the rapid increase in γH2AX minutes after DNA damage (Luijsterburg et al., 2009).
This finding contrasts another study which reported that HP1 proteins are instead recruited to sites of DNA damage (Ayoub et al., 2008). It should be noted; however, that the second study based their findings largely on observations with transfected HP1 proteins which may not reflect the actions of their endogenous counterparts (Ayoub et al., 2008). Recently, Sun and colleagues validated the initial findings that support a role for ejection of Hp1β following DNA damage and further concluded that exposing H3K9me3 sites masked by this protein facilitates Tip60 HAT activation (Sun et al., 2009). Our findings support this notion, as we too have observed displacement of HP1β but retention of Hp1α following exposure of cells to irradiation or CPT.
Importantly, untreated UHRF1-depleted cells showed an absence of chromatin-bound HP1β, which likely is due to the absence of heterochromatic H3K9me3 in these cells. What is not clear is why HP1α but not HP1β is retained on chromatin in UHRF1-depleted cells if both HP1 proteins bind to the same histone trimethylation mark that is absent in the absence of UHRF1. It seems likely that the requirement of HP1α for H3K9me3 might be lower than that of HP1β.
HP1α binding to heterochromatin might be regulated by additional factors to H3K9me3. For
112 example, HP1α has also been shown to bind to lamin receptor B, a component of the nuclear envelope that interacts with chromatin (Fischle et al., 2005). Hp1α can also associate with CAF-
1, a component of the replication fork, and this interaction is crucial in replication of pericentric heterochromatin which occurs during late during the S-phase of the cell cycle (Quivy et al.,
2008).
Our findings suggest that HP1β bound to chromatic H3K9me3 is lost following depletion of UHRF1, but we have not established the cause of this loss. Previous studies demonstrated that following DNA damage, HP1β is ejected from chromatin where after it accumulates in the soluble nuclear fraction as shown by protein fractionation experiments (Ayoub et al., 2008; Sun et al., 2009). In order to exclude other possible explanations for the loss of chromatin-bound
HP1β in UHRF1-depleted cells, future studies will need to be performed to show that redistribution of this protein is occurs in a similar manner and is not removed by another means such as increased protein turnover.
Mislocalization of H3K9me3 from heterochromatin would be expected to jeopardize the ability of Tip60 to be activated in the vicinity of heterochromatin that has encountered DNA damage. Reagents to properly investigate this possibility are currently not available. In order to establish that Tip60 HAT activity is attenuated in UHRF1-depleted cells, HAT activity would need to be measured from Tip60 immunoprecipitates. The commercially available antibodies against Tip60 are notoriously poor reagents for conducting such studies and are not suitable even for Western analysis to measure Tip60 levels in protein lysates, let alone Tip60-associated HAT activity (as has been communicated to our laboratory from Brendan Price- an expert in the field).
An alternative strategy would be to transfect epitope-tagged Tip60 and measure HAT activity by
ELISA from immunoprecipitates obtained with the appropriate antibody against the epitope
113
(Sun, 2009). Such a study is currently beyond the scope of this project in this thesis chapter but is a key component of future experimental strategies in the McPherson laboratory. In lieu of the absence of appropriate Tip60 reagents, we investigated acetyl-H2AK5 by indirect immunofluorescence in control and UHRF1-depleted cells following irradiation. The limitation of such an approach is that we cannot rule out the possibility that acetylation of this histone is mediated by HATs other than Tip60.
A recent study reported that UHRF1 associated with Tip60 and that UHRF1 depletion resulted in a decrease in acetyl-H2AK5 in these cells (Achour et al., 2009). Interestingly, we made the novel observation that acetyl-H2AK5 which was barely detectable in untreated control and UHRF1-depleted cells, was dramatically increased 1 h after exposure to irradiation. This increase was noticeably absent in UHRF1-depleted cells. If Tip60 was the sole or predominant
HAT responsible for acetylating this histone, then these findings would support the hypothesis that Tip60 activation is attenuated in UHRF1-depleted cells. Further experiments will be needed to support this finding by examining whether H2AK5 acetylation was also deficient in Tip60 immunoprecipitates from these cells and whether or not other known histone acetylation marks associated with Tip60 are affected, such as acetyl-H4K16 and acetyl-H3K9.
Sun and colleagues proposed that exposed H3K9me3 is required for the activation of
Tip60 HAT activity but not the recruitment of Tip60 to sites of DNA damage. Tip60 has been shown to form foci that co-localize with γH2AX when challenged with DNA damage (Sun et al.,
2009). To confirm this finding in UHRF1 cells, future studies might employ epitope-tagged
Tip60 transfected into control and UHRF1-depleted cells and co-localization of Tip60 foci with
γH2AX in both our control and UHRF1-depleted cell lines could be performed in the absence and presence of DNA damage.
114
Given that lack of Tip60 activation impairs the activation of ATM kinase activity, we sought to examine ATM activation by measuring ATM autophosphorylation in control and
UHRF1-depleted cells as measured by Western analysis. Our findings suggest that ATM activation is impaired and importantly provide a plausible reason why the percentage of cells with elevated levels of γH2AX are reduced in UHRF1-depleted cells following irradiation. A defect in ATM activation would be sufficient to explain the radiosensitivity and chromosomal instability inherent in these cells. It is important to note; however, that we have not rigorously linked the lack of ATM activation with the apparent lack of Tip60 activation in UHRF1-depleted cells. In order to make this link, we would need to demonstrate that Tip60 acetylation of ATM is impaired, as the acetylation of this kinase facilitates its activation, as measured by autophosphoryation. Future studies would need to determine whether immunoprecipitated ATM shows a loss of acetylation, which could be measured with an anti-acetyl lysine antibody by
Western analysis of ATM immunoprecipitates.
Significance and Impact:
We found that a loss of UHRF1 specifically resulted in loss of heterochromatic
H3K9me3 but not H4K20me3. We showed for the first time that UHRF1 loss specifically ablates chromatin binding of HP1 β which is normally ejected following irradiation of normal cells. We also show for the first time that acetylation of H2AK5 is specifically induced following DNA damage and that UHRF1 depletion counteracts this acetylation event.
Furthermore, we show for the first time that ATM kinase activation is impaired in UHRF1- depleted cells following DNA damage which provides a plausible explanation for the impaired
γH2AX levels observed, given that ATM phosphorylates H2AX to form γH2AX as part of the
115
DNA damage response to DSBs. Our findings suggest a link between maintenance of H3K9me3 on heterochromatin by UHRF1 and the DNA damage response to DSBs. Our findings in this chapter help us better understand the role that UHRF1 is playing in areas of the genome that may be a barrier to repair, such as heterochromatin.
116
CHAPTER FIVE: GENERAL CONCLUSIONS AND FUTURE DIRECTIONS
Mus81-Eme1 is an endonuclease known to play a role in processing DNA intermediates that result from aberrant or stalled DNA replication (Hanada et al., 2007). Mus81-Eme1 is a heterodimer endonuclease that is important for the repair of replication-associated DNA damage.
This endonuclease has been proposed to cleave replication fork structures into a DNA intermediate that contains a free end of double-stranded DNA which can be repaired by homologous recombination (Hanada et al., 2007). Several DNA repair pathways are known to participate in the repair of damaged replication forks, including NER, HR, translesion synthesis and FA signalling (Allen et al., 2011; Moldovan and D'Andrea, 2009). Our understanding of how Mus81-Eme1 participates in DNA repair would be greatly aided by determining whether this enzyme participates in one of the above-mentioned pathways or whether it acts in a novel repair process. In order to gain insight into the pathway Mus81-Eme1 participates in, I participated in Mus81-Eme1 interaction screens designed to identify candidate proteins that interact with either Mus81 or Eme1. My project commenced by identifying interacting proteins with Eme1 and focused on one candidate interacting protein known as UHRF1. Chapter Two of my thesis described experimentation that confirmed and characterized this interaction, whereas
Chapter Three demonstrated that UHRF1 is required for a maximal cellular response to DNA damage caused by exposure to γ-irradiation. In Chapter Four, I provide experimental evidence that loss of heterochromatic H3K9me3 in the absence of UHRF1 disrupts HP1β binding to chromatin. UHRF1-depleted cells were found to exhibit impaired DNA damage-induced H2A acetylation on lysine 5, which may reflect a loss of Tip60-mediated chromatin relaxation and
117
ATM activation. These findings are in agreement with previous studies that point to a role for
H3K9me3 in activating Tip60 HAT following DNA damage (Sun et al., 2005; Sun et al., 2010;
Sun et al., 2009) and will pave the way for future studies to investigate whether the ability of
UHRF1 to maintain heterochromatic H3K9me3 facilitates an optimal DNA damage response in heterochromatic regions of the genome.
Eme1 interacts with UHRF1 Prior to this investigation, relatively little was known about how Mus81-Eme1 coordinated DNA repair with other proteins. When we originally set out to identify proteins that interacted with
Eme1, it was previously determined that Mus81-Eme1 interacted with Chk2 and BLM helicase.
In fact, Mus81 (prior to its association with Eme1) was identified in yeast as a protein partner for
Cds1 (homolog of Chk2 in mammals) (Boddy et al., 2001; Ciccia et al., 2003). After the discovery of Eme1 as an essential cofactor for Mus81, the enzyme heterodimer was also found to associate with the BLM helicase. BLM helicase activity was found to enhance the endonuclease activity of Mus81-Eme1 (Zhang et al., 2005). Recent studies now have demonstrated that
Mus81-Eme1 interacts with SLX4, which together with SLX1 comprises another endonuclease.
SLX4 also associates with ERCC4(XPF)-ERCC1, an endonuclease originally associated with removal of DNA adducts during the process of NER. Both XPF-ERCC1 and SLX1-SLX4 endonuclease activities have also been found capable of cleaving replication fork structures in vitro, similar to Mus81-Eme1. The association of all three structure-specific endonucleases together in one complex might help to spatially and temporally coordinate the distinct activities of each enzyme. For example, one endonuclease might initiate the repair of a replication fork by cleaving one DNA strand or molecule, leading to the conversion of this replication fork into a different DNA intermediate such as a Holliday junction, which is then the optimal substrate for
118 another endonuclease and so on. In support of this possibility, the DNA cleavage product of
SLX4-SLX1 appears to be a preferred substrate for Mus81-Eme1. Mus81-Eme1, SLX1-SLX4 and XPF-ERCC1 are all likely to coordinate the resolution of replication fork damage through the stepwise manipulation of distinct DNA intermediates (Svendsen et al., 2009).
We initiated our search for new interacting proteins of Mus81-Eme1 by conducting a cDNA library screen using Eme1 as bait. Our initial screen did not identify any previously described Eme1-interacting proteins (such as Mus81) which might be explained by a limitation in the screening library used for the studies. The cDNA library was over-represented in coverage of 3‟ sequence yet under-represented in coverage of 5‟ sequence, meaning that one was more likely to identify protein partner interactions if the missing partner cDNA contained the interaction domain in the 3‟ end of the open reading frame as opposed to the 5‟ end. The region of Mus81 that interacts with Eme1 is located in the 5‟ end of the Mus81 open reading frame, making it less likely to be represented in the library that was used for screening. Seven of the interacting candidates identified in the initial screen include Af9, Uhrf1, Hm2, fibrillin, eukTEF1, Sap30 and 2G4. AF9, Uhrf1, Hmgn2, and 2G4 have been associated with modification of chromatin structure, suggesting that optimal activity of Mus81-Eme1 might be intimately tied to changes in chromatin accessibility. Further investigation of Uhrf1 as a candidate interacting protein with Eme1 was prioritized given that previous studies had linked
Uhrf1 to cellular resistance to DNA damaging agents. The spectrum of agents that cells were found to be more sensitive to when depleted of Uhrf1 overlapped with the spectrum of agents that Mus81-deficient cells were sensitive to. In particular, both Uhrf1-deficient cells and Mus81- or Eme1-deficient cells were noted to share sensitivity to hydroxyurea, a replication fork inhibitor, suggesting that both Uhrf1 and Mus81-Eme1 might be operating in the same pathway
119
(Abraham et al., 2003; Boddy et al., 2000; Interthal and Heyer, 2000). The Yeast Two Hybrid assay is a powerful technique as it allows a high number of interacting candidate proteins to be assayed in a simple experiment. Another advantage of this procedure is that neither protein purification of the prey proteins nor antibody to the prey protein is required (Knudsen et al.,
2002). Drawbacks to this technique include the occurrence of false positives, which may occur if cells exhibit an active reporter gene in the absence of a bona fide interaction between products of the bait and prey plasmid (Knudsen et al., 2002; Serebriiskii and Golemis, 2001). Other in vitro alternatives to this assay include the GST pull-down assay, which allows a quick and reliable method to isolate tagged proteins and analyse proteins that co-purify with the GST tag. GST pull-down assay drawbacks include the generation of a fusion protein with a large (22 kDa) GST tag, which may result in the steric hindrance or improper folding of proteins (John M. Walker,
2008). Other protein interaction assays include co-immunoprecipitation, which allows for appropriate protein posttranslational modifications to take place (John M. Walker, 2008).
Given that these three protein interaction assays mentioned above have different advantages and disadvantages, we decided to test the Eme1-UHRF1 interaction using all three assays, and confirm the Eme1-UHRF1 interaction following the initial Yeast Two Hybrid cDNA library screen. The confirmation of this interaction using these three independent methodologies built the foundation for the further investigation between the two proteins.
We hypothesized that Eme1 and UHRF1 may work together in a common pathway that the cell uses to fix replication fork damage for two reasons. First, Mus81-Eme1 has been shown to be directly involved in making replication fork damage structures more amenable to repair
(Muto et al., 2002) and secondly, UHRF1-deficient ES cells are more sensitive to hydroxyurea, an agent that also results in replication fork damage. (Muto et al., 2002). This hypothesis was
120 further supported by subsequent experiments that showed our UHRF1-depleted HeLa cells are more sensitive to replication fork damaging agents (CPT and Ara-C) compared to control cell lines (figure 24). The association of Mus81-Eme1 and UHRF1 are likely dependent on the presence of DNA damage. We found that our untreated cells displayed an accumulation of
Eme1 in euchromatin regions, and an accumulation of UHRF1 in heterochromatic regions. In striking contrast, CPT treatment resulted in co-localization between the two proteins into a nuclear speckle formation. This co-localization was lost upon transfection of the UHRF1 ring finger mutants (figure 19). The ring finger domain of UHRF1 is responsible for endowing
UHRF1 with E3 ubiquitin ligase activity (Citterio et al., 2004). A possible explanation for observing a loss of this interaction upon transfection of UHRF1 ring finger mutants is that Eme1 could possibly be a substrate for UHRF1, and may mediate its ubiquitination and enzyme turnover just as it has been shown for its role with mediating DNMT1 turnover (Bronner, 2011).
It is presently unclear whether Eme1 is a bona fide substrate for the E3 ligase activity of UHRF1.
Further studies will be required using an in vitro reconstituted ubiquitination system to determine whether UHRF1 transfers a ubiquitin peptide to Eme1. Even if Eme1 indeed accepts a ubiquitin peptide from UHRF1, the biological significance is presently unclear. In many cases, ubiquitination (especially polyubiquitination) is a beacon that signifies protein turnover. Mus81-
Eme1 turnover might be a required step for the effective repair of replication forks. On the other hand, ubiquitination may serve a role in signalling independent of protein turnover. The E3 ubiquitination activity of UHRF1 may be important for DNA repair by causing the turnover and removal of histones in the vicinity of DNA damage via ubiquitination (Citterio et al., 2004), and therefore this ring finger dependent interaction between the two proteins may be representative of UHRF1 ubiquitinating histones and targeting them for removal from the nucleosome to
121 facilitate repair (summarized in figure 36). Further experiments will be required in order to determine the exact role and substrates of UHRF1 ubiquitination following DNA damage.
UHRF1’s role in double strand breaks caused by gamma-irradiation Before investigating further the relationship between Eme1 and UHRF1, it was deemed necessary to determine what role UHRF1 had in the DNA damage response. UHRF1 likely has multiple roles in the DNA damage response outside of its possible role with Eme1, as UHRF1- deficient cells exhibit sensitivity to DNA damaging stimuli such as UV and X-rays that are not associated with replication-associated DNA damage. Increased sensitivity to X-rays usually signifies a defect in the repair of DSB‟s. Given that our understanding of DSB repair processes is better developed than repair of other lesions, we decided to focus on determining why UHRF1 confers resistance to radiation. Many cell models with a DNA repair defect also exhibit a coincidental impediment in proliferation. Provided that UHRF1 has previously been shown to be important for proliferation in some cell systems, we decided to investigate whether UHRF1- depleted cells exhibited a defect in proliferation. Previous studies have reported cells depleted of
UHRF1 divided at a slower rate (Bonapace et al., 2002; Jenkins et al., 2005; Kim et al., 2009;
Unoki et al., 2004). In keeping with these observations, our UHRF1-depleted cells display a mild proliferation impediment that was only found to be significantly different from control cells after nine days of monitoring. This mild effect contrasts with previous reports which stated
UHRF1-depletion results in a more drastic proliferation defect (Bonapace et al., 2002; Jenkins et al., 2005; Kim et al., 2009; Unoki et al., 2004). The reason for the proliferation impediment is unclear. The impediment could be due to (i) a delay in transit through one particular stage of the cell cycle (ii) an overall delay in transit through the cell cycle
122
(iii) a delay in reinitiating subsequent rounds of cell cycling or (iv) a loss of cells due to apoptosis or other forms of cell death following each round of cell division. Such a delay in cell cycle transit might be anticipated if processes regulating replication of heterochromatic regions are impaired in UHRF1-depleted cells as previously described. The proliferation impediment might also be caused by the increased chromosomal instability seen in UHRF1-depleted cells that might compromise efficient chromosome segregation during mitosis, leading to cell death caused by mitotic catastrophes. A loss of cells due to such events might manifest as decreased cell numbers following repeated passage of cells. Another reason that could explain this slight decrease in proliferation could be due to increases in the cyclin-dependent kinase inhibitor (p21) levels. One recent study suggested that the proliferation defects observed in UHRF1-deficient
ES cells could be ascribed to an increase in p21 protein levels in UHRF1-deficient cells (Kim et al., 2009). Our findings to date do not show a link between p21 status and UHRF1 however
(Laura Tamblyn, unpublished observations).
As ES cells missing UHRF1 are more sensitive to radiation, we checked the sensitivity of our UHRF1-depleted cells to γ-irradiation with the use of clonogenic assays. We decided to use clonogenic assays to measure cell survival as this methodology can determine log differences in cell number between cell lines and assess survival over a long period of time (over a period of several days). An alternative to this assay could have been the use of an (3-(4,5-
Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay. Drawbacks to this assay include; a short term assessment resulting in a less robust assay, when compared to the clonogenic assay
(Enomoto et al., 1997). Similar to UHRF1-deficient ES cells, we discovered that UHRF1- depleted HeLa cells also exhibited an increased sensitivity to IR, in that they also displayed increased sensitivity to varying doses of γ-irradiation (figure 23).
124
After establishing that our UHRF1-depleted cells were more sensitive to γ-irradiation compared to control cells, we decided to further explore the reason behind the observed radio sensitivity by querying the role of UHRF1 as a protein involved in the repair of DNA damage caused by radiation. Many previously classified defects in DSB repair have been attributed to defects in checkpoints, accordingly we examined the status of cell cycle checkpoints in UHRF1- depleted cells following irradiation (Schmitt et al., 2007).
In general, the cellular response to γ-irradiation can involve the activation of cell cycle checkpoints that lead to the halt of cell cycle progression. These checkpoints can include the
G1-S, intra-S and the G2-M checkpoint. In the presence of DNA damage such as DSBs, the G1-
S checkpoint is activated following the phosphorylation of a phosphatase known as cdc25a. This phosphorylation step signals the degradation of cdc25a which leaves the cyclin dependent kinase cdk2 in a phosphorylated state. Phosphorylated cdk2 is inactive, therefore its substrate cdc45 remains unphosphorylated and origins of replication fail to initiate. The intra-S phase checkpoint is activated by DSBs generated within S phase that occur independent of active replication forks. The G2-M checkpoint proceeds in a manner similar to the G1/S checkpoint, with cdc25c degradation resulting in phosphorylated inactive Cdk1/cyclin B and the activation of the G2-M checkpoint. ATM is required for all three checkpoints. ATM-induced phosphorylation of Chk2 results in phosphorylation of cdc25 members, leading to checkpoint activation. Activation of p53 following irradiation also causes checkpoint activation, most notably by the increased production of p21, a cdk inhibitor.
125
The HeLa cells used to generate our UHRF1-depleted cell lines typically undergo a G2-
M cell cycle arrest following irradiation. The G2-M checkpoint predominates over the G1-S checkpoint in these cells as HeLa cells are impaired in the production of p53, which essentially renders them „p53-null‟ and incapable of mounting a G1-S checkpoint response that involves increased production of p21. Accordingly, we assessed the G2-M checkpoint by BrdU-PI incorporation analysis using FACS.
We challenged our control and UHRF1-depleted cells with 1 and 5 Gy of γ-irradiation and examined the ability of cells to accumulate in the G2-M phase as a function of time
(Warenius et al., 1998). These doses were chosen based on the results in our clonogenic assay which show at 1 and 5 Gy a differential sensitivity of UHRF1-depleted cells compared to control cells. After exposure to 1 Gy of irradiation, control cells accumulated in the G2-M phases of the cell cycle as a function of time that was maximal at 12 hours following irradiation. At this time point, the percentage of cells in the G2-M fraction was almost fourfold compared to untreated cells. In comparison, UHRF1-depleted cells showed only a two and half fold increase in the G2-
M fraction 12 hours following irradiation. In general, UHRF1-depleted cells displayed a decreased tendency to arrest at the various time points, which was most prominent at 12 hours.
Taken together, these results signify a defect in the G2-M checkpoint for UHRF1-depleted cells.
Twenty-four hours after exposure to 1 Gy, the observed arrest was diminished and the cells resumed progression through the cell cycle. When the cells were irradiated with 5 Gy, a more drastic arrest in the cell cycle was seen at twelve hours with an accumulation in the G2-M phase of the cell cycle for the control cells. In contrast, the UHRF1-depleted cells did not display an accumulation of cells at this time point in the G2/M phase. This study demonstrated that a loss of
UHRF1 results in a defective G2-M checkpoint in the presence of γ-irradiation. We cannot at
126 this point rule out the possibility that the impaired checkpoint response might be specific to our cell model in question and our conclusions could be strengthened by repeating these experiments in another cell type. Our study is the first to investigate the checkpoint status of the G2-M checkpoint in cells depleted of UHRF1 following exposure to DNA damage and definitely shows that UHRF1-depleted cells have an impediment in their ability to mount an optimal response to
DNA damage caused by γ-irradiation. A recent study that has emerged following our published results found that a transient depletion of UHRF1 results in the activation of the G2-M checkpoint (Tien et al., 2011). This study fails to query the status of the G2-M checkpoint in the presence of DNA damage, although this finding might explain the cause of the observed proliferation defect observed in UHRF1-depleted cells (figure 22).
One of the initial steps in the signalling cascade that results in checkpoint activation is the autophosphorylation of ATM protein kinase, which leads to γH2AX. Given that our UHRF1- depleted cells exhibited defective checkpoint activation, we postulated that a loss of DNA damage response signalling such as the signal emanating from ATM could be the underlying cause. If this was the case, then we might expect to see a defect in the accumulation of γH2AX in our UHRF1-depleted cells following irradiation. As expected, irradiated UHRF1-depleted cells indeed showed reduced accumulation of elevated levels of γH2AX compared to control cells. This defect could be explained by a potential role for UHRF1 in providing access for DNA damage repair factors in heterochromatin. We later propose (in chapter Four) that UHRF1 is a significant contributor in a signalling cascade that relaxes heterochromatin and gives access to
DNA repair proteins in the presence of heterochromatic DSB‟s. Therefore, a loss of UHRF1 may render these areas inaccessible due to a lack of heterochromatin relaxation and render the cell unable to propagate the γH2AX signal. Chromatin can pose a structural barrier to the DNA
127 damage response and is transiently removed or remodelled to give access to repair proteins to fix
DNA damage. γH2AX has been shown to occur less efficiently in areas of heterochromatin in the cell, when compared to euchromatin (Murga et al., 2007). This suggests that the strength of
γH2AX signal could be dependent on chromatin compaction properties. Our findings of decreased γH2AX in UHRF1-depleted cells strongly suggest that a loss of UHRF1 could be leading to heterochromatin remaining inaccessible and therefore compromising the ability of
DNA repair proteins to access and repair DNA damage. UHRF1-deficient cells, show less cells with high levels of γ H2AX foci staining than controls. It remains to be determined if the lower level of γH2AX is due to fewer γH2AX foci/lesion generated in UHRF1-deficient cells
(corresponding to fewer DNA lesions generated by γ-irradiation), or whether there is less phosphorylation of H2AX per focus (corresponding to less γH2AX generated per DNA lesion).
Does UHRF1 have a role in the chromosomal maintenance? Our studies show that UHRF1 has a role in optimizing the DNA damage response for
DSB repair. Many cell models deficient in optimal DSB repair exhibit a general defect in chromosomal stability. Accordingly, we examined whether UHRF1 contributed to chromosomal integrity in the absence of DNA damaging stimulus. Interestingly, we noted that loss of UHRF1 caused cells to exhibit an increased propensity for chromosomal instability, as measured by scoring micronuclei incidence, scoring of supernumerary centrosomes and karyotyping of metaphase chromosomes. A reason for this could be that UHRF1-depletion results in disruption of heterochromatin structure, particularly in heterochromatin surrounding the centromeres
(pericentric heterochromatin). UHRF1 has been linked to the replication of pericentric heterochromatin (Papait et al., 2007), and major satellite silencing. Maintenance of this heterochromatin has been shown to be important for genome integrity and stability. A loss of
128
UHRF1 could affect the integrity of pericentric heterochromatin and result in chromosome missegregation ultimately leading to chromosomal instability (Peters et al., 2001; Taddei et al.,
2001).
Another explanation could be that a loss of H3K9me3 from pericentric heterochromatin is sufficient to result in chromosomal segregation problems that occur in our UHRF1-depleted cells (Peters et al., 2001). We purpose that UHRF1 is playing a role upstream of Suv39h1 and
Suv39h2 by binding to and propagating H3K9me3 (summarized in figure 14). UHRF1 depletion results in H3K9me3 disruption and would likely disrupt pericentric heterochromatin dynamics and stability, resulting in chromosomal aberrations.
UHRF1 is a contributor in a signalling cascade that repairs DNA damage in heterochromatin An in vitro study has shown that UHRF1 can help give access to modifying enzymes to
DNA found in nucleosomal arrays (Papait et al., 2008). Given our previous findings of
UHRF1‟s involvement in DSB repair in the presence of γ-irradiation, we propose that UHRF1 may be involved in a signalling cascade that opens heterochromatin and permits access to sensor and DNA repair proteins. Therefore, a loss of UHRF1 may result in a loss of the cell‟s ability to open heterochromatin for repair in a manner analogous to its previously described role in facilitating access of heterochromatin to replication enzymes during DNA synthesis. This may explain the impaired DNA damage response observed in our UHRF1-depleted cells.
UHRF1‟s role in this signalling cascade could be due to its ability to bind and facilitate the retention of the heterochromatin mark, H3K9me3. H3K9me3 levels have not been shown to fluctuate in response to DNA damage (Ayoub et al., 2008; Karagianni et al., 2008), but it serves as a docking site for heterochromatin proteins, such as the HP1 family that has recently been
129 shown to respond to DNA damage. Hp1β has recently been found to disassociate from
H3K9me3 following DSB damage in areas of heterochromatin, and facilitates accumulation of
γH2AX (Ayoub et al., 2008). The mechanism behind the release of Hp1β and subsequent accumulation of γH2AX formation remained unclear until recently. Sun and colleagues found that Hp1β‟s release from H3K9me3 resulted in Tip60-ATM recruitment. Binding of Tip60 to
H3K9me3, resulted in the activation of the HAT activity of Tip60, leading to the acetylation of
ATM and activation of ATM kinase (Sun et al., 2005). Accordingly if UHRF1 is playing a role in maintaining and propagating H3K9me3, it may be a contributor to this repair pathway that was proposed in figure 28. In order to test this model, we decided to investigate the status of
H3K9me3 in our cell lines. Western analysis of H3K9me3 indicated that levels of H3K9me3 remained the same in both control and UHRF1-depleted cells; however, with the use of indirect immunofluorescence, we were able to observe loss of chromatin-bound H3K9me3 in our
UHRF1-depleted cells. This redistribution for H3K9me3 was rescued upon transfection of murine UHRF1 cDNA (refractory to human UHRF1 shRNA) (figure 31). Our findings are similar to those reported by Karagianni and colleagues but contrast those of Papait and colleagues who report that a transient knock down of UHRF1 results in the H3K9me3 staining pattern remaining the same as the control cells (Karagianni et al., 2008; Papait et al., 2008). An explanation for the differences could be due to experimental differences in generating UHRF1 depleted cell lines. Our findings as well at Karagianni‟s use stable depletion of UHRF1 to investigate H3K9me3, while Papait uses (figure 27) transient depletion. Since UHRF1 plays a role in the propagation of H3K9me3, and that the doubling time of our HeLa cells is one day, a longer-term depletion of UHRF1 might be required in order to see any impact on chromatin composition in our UHRF1-depleted cells.
130
The redistribution of the H3K9me3 staining pattern observed in our UHRF1 depleted cells prompted us to investigate if this change in our H3K9me3 staining pattern was specific for this particular mark or was reflective of a more general effect on heterochromatin organization.
Investigation of another heterochromatin mark, H4K20me3, revealed a more subtle change in staining pattern of H4K20me3 in our UHRF1-depleted cells. Control cell lines exhibited focal nuclear staining of H4K20me3 that superimposed onto DAPI dense regions of nuclear chromatin that represent heterochromatin. H4K20me staining appeared to have equivalent focal staining to control cells, but appeared to be localized in larger focal regions in the UHRF1-depleted cells.
This change was not as drastic as the re-distribution observed for H3K9me3 (Figure 27). It is possible that the loss of UHRF1 results in a re-distribution of H3K9me3, and this redistribution could indirectly impact the spatial distribution of other heterochromatin marks and possibly other areas within heterochromatin.
Our model for the role of UHRF1 in a signalling pathway that repairs heterochromatic
DSB‟s (Figure 28) was further tested by conducting indirect immunofluorescence studies on chromatin bound Hp1β in control and UHRF1 depleted cells in the absence or presence of irradiation. According to our model, chromatin-bound Hp1β would be expected to be (i) present in untreated control cells but absent in irradiated control cells and (ii) UHRF1-depleted cells would be expected to exhibit a lack of chromatin-bound Hp1β owing to a disruption of
H3K9me3. These assumptions are based on findings from previous studies which show that
HP1β is ejected from chromatin following exposure to either γ-irradiation or treatment with the topoisomerase II inhibitor, etoposide (Ayoub et al., 2008; Montecucco et al., 2001; Sun et al.,
2005). Our findings are consistent with these studies; however, it is worth noting that these findings are hotly debated as Luijsterburg and colleagues show the opposite effect, in that all
131 three isoforms of the HP1 proteins; Hp1α, Hp1β, and Hp1γ are recruited to damaged sites
(Luijsterburg et al., 2009). Although both studies have limitations in their approach, the study by Ayoub and colleagues seems to be based on more sophisticated biochemical data that conjures a more satisfying mechanism of action. (Ayoub et al., 2008; Ayoub et al., 2009;
Luijsterburg et al., 2009). Our findings strongly support those of Ayoub and colleagues, as we observe a loss of chromatin bound Hp1β following irradiation in our control cells, and a chromatic loss of Hp1β in our untreated and irradiated UHRF1-depleted cells. Our findings are consistent with Ayoub and colleagues‟ findings; however, we cannot rule out other possibilities such as the accumulation of Hp1 proteins over time. A time course would help clarify if perhaps
Hp1β is subsequently recruited to sites of damage after its initial dispersal, and if this accumulation is affected in UHRF1-depleted cells due to altered H3K9me3.
Loss of Hp1β in our UHRF1-depleted cells was expected given the re-distributed
H3K9me3 staining pattern observed in our cells. We also conducted these experiments treating our cells with the replication fork damaging agent, CPT. UHRF1 is believed to play a role in maintaining the compacted nature of heterochromatin, while relaxing heterochromatin to facilitate replication. These findings are important, as they indicate that our proposed model of
UHRF1 being a significant contributor in a signalling pathway that repairs DSB in heterochromatic regions (Figure 28) may also be crucial in resolving DNA intermediates
(DSBs) that arise due to replication fork damage. It is worth considering that we are only able to observe what is bound to chromatin in our indirect immunofluorescence studies. If Hp1β is indeed ejected from H3K9me3 in UHRF1-depleted cells, we would expect to see an accumulation of Hp1β in the nuclear solution fraction of our cells. It would be worthwhile
132 conducting future sub cellular fractionation experiments to confirm that Hp1β is ejected from heterochromatin and found in the nuclear solution fraction (and not being degraded).
Hp1α but not Hp1β is retained on chromatin in untreated and γ-irradiated and CPT treated cells, as well as in control and UHRF1-depleted cells (figure 32 and 33). Perhaps Hp1α is less dependent on its H3K9me3-binding chromodomain for binding to chromatin than Hp1β.
This retention might be explained by Hp1α‟s requirement for H3K9me3‟s binding being lower than that of Hp1β. Also, Hp1α has also been shown to bind to a component of the nuclear envelopment, lamin receptor B, so it may be found in other places in the nucleus. Hp1α has also been shown to be found in a complex with CAF-1 (Loyola et al., 2009). This complex has an important role in pericentric heterochromatin replication. Thus additional multiple roles of Hp1α may help explain why focal staining of this Hp1 isoform is not lost following DNA damage, and
UHRF1 depletion. According to our proposed model (Figure 28), if UHRF1 is depleted, heterochromatic H3K9me3 would become lost, and as a result Tip60 HAT activity would not be capable of being activated following DNA damage. Tip60 was suggested to be the sole HAT known to acetylate H2AK5 based on in vitro experiments (Bhaumik et al., 2007). Our findings indicate that following exposure to γ-irradiation, the levels of this histone modification in our cells dramatically increased when we examine cells by indirect immunofluorescence.
Importantly, the DNA damage-dependent increase in acetyl-H2AK5 levels is totally eradicated in UHRF1-depleted cells following DNA damage. These findings strongly suggest that UHRF1 eradicates the ability of Tip60 to activate chromatin relaxation following DNA damage. A drawback of this study is that we are relying on measuring Tip60 activity by investigating levels of acetyl-H2AK5. Our model would be strengthened if we ensured acetyl-H2AK5 levels were dependent on Tip60 following exposure to irradiation and other DNA damaging agents (CPT and
133
Ara-C) that UHRF1-depleted cells are sensitive to. One way to ensure this is to investigate if
Tip60 depletion by siRNA blocked increased acetyl-H2AK5 following DNA damage. If Tip60 depletion attenuates increased acetyl-H2AK5 levels induced by DNA damage, then we can directly attribute the increased acetyl-H2AK5 to increased activation of Tip60 acetyltransferase activity. Alternatively, Tip60-dependent acetyltransferase activity could be investigated by measuring acetyltransferase activity in Tip60 immunoprecipitates from protein lysates transfected with epitope tagged Tip60, together with a commercially available measurement of acetylation of biotinylated histone H4 peptide by ELISA, as has previously been reported (Sun et al., 2005; Sun et al., 2009).
Following DNA damage, Tip60 has been shown to bind to and acetylate ATM. This acetylation step has been shown to be critical in activating the kinase activity of ATM by an unknown mechanism (Jiang et al., 2006; Sun et al., 2005). If Tip60 activation following DNA damage is defective in UHRF1 depleted cells, we would expect that acetylation of ATM following DNA damage would likewise be reduced or absent. Our western analysis show that
UHRF1-depleted cells do indeed have lower levels of phosphoSer1981-ATM following irradiation compared to control cells. Our experiments performed thus far show this to be the case but require further studies to determine that the lack of ATM acetylation is responsible for this effect.
Based on our findings that UHRF1-depletion results in a re-distribution of H3K9me3, ejection of Hp1β and defective Tip60 acetylation and ATM activation, we propose that maintenance of H3K9me3 by UHRF1 facilitates an optimal response to DNA damage in heterochromatin through the activation of Tip60 and ATM (Sun et al., 2005). Tip60 could also
134 be acetylating other histones (e.g. acetyl-H4K20), facilitating heterochromatin relaxation
(Vidanes et al., 2005).
We propose a model for UHRF1‟s role in DSB repair by binding to and propagation of
H3K9me3. The steps in this model are: 1) ejection of Hp1β from H3K9me3, 2) binding of the
MRN complex to the ends of the DSB, 3) recruitment of Tip60 together with ATM, 4) binding of
Tip60 to H3K9me3 and activation of Tip60, 5) acetylation of ATM by Tip60, 6) auto phosphorylation of ATM, 7) phosphorylation of KAP-1 by ATM resulting in heterochromatin relaxation, and 8) activation of the ATM pathway by phosphorylation of γH2AX and downstream substrates, such as Chk2 (summarized in figure 13). Accordingly, a loss of
UHRF1 would result in: 1) altered distribution of H3K9me3, 2) ejection of Hp1β, 3) reduced or loss of Tip60 activity, 4) reduced or loss of ATM autophosphorylation resulting in, 5) reduced or loss of γH2AX, 6) reduced or defective KAP-1 phosphorylation and, 7) defect in the activation of ATM and its downstream substrates (e.g. Chk2). This would be expected to result in defective
DNA repair.
During the replication of heterochromatin, UHRF1 may be facilitating access of replication machinery to replication forks through its interaction with factors such as Tip60 that serve to facilitate chromatin access. When replication forks are damaged or stalled, we propose that UHRF1 might also play a role in a pathway that the cells use to fix this damage as follows:
1) UHRF1 ubiquitinates histones facilitating their remodelling/removal, or 2) this results in heterochromatin relaxation, then 3) Mus81-Eme1 now has access to the damaged replication fork and can mediate repair, and 4) UHRF1 could ubiquitinate Eme1, leading to Eme1 turnover following completion of replication. Accordingly, a loss of UHRF1 in this model may result in:
1) compacted heterochromatin, resulting in replication fork damage being inaccessible, 2)
135
Mus81-Eme1 and other repair factors not having access to the damaged replication fork, 4) loss of turnover of Eme1 levels due to inability of UHRF1 to ubiquitinate Eme1.
Clinical Implications in Cancer Our research presents important findings that could help identify important cancer diagnostic markers, pathways that contribute towards chemotherapeutic resistance and therapeutic strategies used to treat a number of human cancers.
Eme1 and UHRF1 have both been proposed as candidate diagnostic markers for chemotherapy resistance (Tomoda et al., 2009), which means that both proteins have a role in helping make cancer cells resistant to DNA damage induced by therapy. UHRF1 is found to be highly expressed in human tumours originating from the lung and bladder (Unoki et al., 2010; Unoki et al., 2009b), while Eme1 has been found to be highly expressed in lung cancer. If Eme1 and
UHRF1 are playing a role in the same pathway in DNA repair, it would be worth detecting levels of Eme1 in human tumours that have elevated levels of UHRF1. Such tumour cells might exhibit a heightened resistance to replication-associated DNA damage, making them resistant to agents such as Ara-C. Knowing the levels of these proteins might have predictive value in determining the ability of tumour cells to respond to such therapy.
Diagnostic development includes screening a number of markers simultaneously in malignancies, which helps establish a diagnostic profile and may give us clues to what DNA repair pathways are being up or down regulated in cancer cells (Fabian, 2010). Future studies are necessary to validate whether Eme1 and UHRF1 are playing a role in the same pathway to mediate replication fork repair. Confirmation of this could contribute towards future development for cancer diagnostics that will help with the early detection of some human cancers and prevent malignancy to go undetected until long past the time of effective treatment.
136
Eme1 and UHRF1 could be particularly useful because if proven that they function together in replication repair, it will give us insight into the repair pathways that cancer cells are using to confer resistance to radiation and antineoplastics. It is also possible that a significant percentage of the general population might have lower levels of Eme1 and/or UHRF1 activity due to single nucleotide polymorphisms. Lower levels of the activity of these proteins might sensitize their responsiveness to chemotherapy. Advance knowledge of their state might allow us to tailor regimens appropriate for these individuals.
This research also presents important findings that may aid in the development of future therapeutics that potentiate DNA damage signalling responses by chromatin decompaction, such as agents referred to as HDAC inhibitors. Examples of HDAC inhibitors that are currently used in cancer treatment are suberoylanilide hydroxamic acid and MS-275 (a benzamide derivate)
(Munster et al., 2001). Specific deregulation of HDACS has been associated with leukemia, and breast and ovarian cancer (Grignani et al., 1998; Yarden and Brody, 1999). The mechanism of action of HDAC inhibitors is not entirely understood, however the rationale behind their development is that HDAC inhibition leads to hyperacetylation, which results in chromatin relaxation and the re-activation of genes that are silenced in tumorigenesis (Gui et al., 2004).
HDAC inhibitors have been shown to have synergistic effects by reducing cellular proliferation when combined with other cancer therapeutics or radiotherapy (La Thangue, 2004). If HDAC inhibitors relax heterochromatin, a possible combination of this agent with the depletion in Eme1 and UHRF1, followed by treatment with replication fork damaging agents (e.g. Ara-C) could be effective at rendering cancer cells more susceptible to damage.
137
Conclusion My thesis overall has initially demonstrated an interaction between two proteins found in the DNA damage response and proposes that they may function together to repair replication fork associated damage. Limitations include a lack of data that is needed to characterize this interaction and rigorously test our proposed model of how Eme1 and UHRF1 may cooperate together to resolve replication fork damage. My subsequent studies identify UHRF1 a genome caretaker, and a critical need for this protein in DSB repair. We build upon these findings by proposing a mechanism whereby UHRF1 is a significant contributor in a signalling pathway that resolves heterochromatic DSB‟s. We cannot exclude the possibility that this may be one of the many mechanisms that the cell uses to mediate DSB repair. Given that Eme1 and UHRF1 has been found to be over expressed in cancer cells compared to its normal tissue counterparts, it will be of special interest to determine whether the over expression of Eme1 and UHRF1 confers special advantages to neoplastic cells. As down-regulation of Eme1 and UHRF1 was found to sensitize cells to DNA-damaging chemotherapy, it would be of interest to know if Eme1 and
UHRF1 facilitates resistance to radiotherapy or chemotherapy in these tumours, and will the
H3K9me3 status of tumour cells have predictive value in determining whether tumour cells with an elevated level of UHRF1 have a increased H3K9me3 staining, and whether increased levels of this histone modification confer a more efficient response to DNA damage or impact neoplastic transformation.
Future Studies We discovered and validated an interaction between Eme1 and UHRF1 and propose that they could be working together in one pathway that the cell uses to repair replication fork associated damage. Subsequent studies identified a role for UHRF1 in DSB repair, as well as a crucial role for this protein in genome integrity maintenance without the presence of any damaging agents.
138
Finally, we proposed and tested a model whereby UHRF1 is contributing to maintaining and propagating heterochromatin mark, H3K9me3. Its role in doing so, seems necessary for the cells ability to repair DSBs located in heterochromatin (summarized in figure 36).
Querying the possible role of Mus81-Eme1 in heterochromatin repair Given our findings on the role of UHRF1 in maintenance of heterochromatic H3K9me3 and the role of this modification in facilitating repair of DNA damage in heterochromatin, it will be of considerable interest to re-evaluate the role of Mus81-Eme1 in the repair of replication- associated damage. In particular, it will be of interest to determine whether Mus81-Eme1 has an important role in repair of replication-associated damage located within heterochromatin. One possible way to assess this would be to evaluate whether replication-associated DNA damage foci (i.e. γH2AX foci) that persist in the absence of Mus81 overlap or juxtapose with regions of heterochromatin. Using wild-type and Mus81-deficient murine embryonic fibroblasts, it should be relatively straight-forward to examine whether γH2AX foci in these cells (in the absence or presence of replication inhibitors) co-localize with the clustered regions of heterochromatic satellite DNA in these cells that is apparent following DAPI staining during indirect immunofluorescence procedures. Such an approach has been used previously to demonstrate that ATM facilitates DSB repair within heterochromatin (Goodarzi, 2008).
Interestingly, in untreated cells, transfected Eme1-V5 and FLAG-UHRF1 were found to exist in mutually exclusive areas, with Eme1 found in euchromatic DAPI-light regions whereas UHRF1 found in heterochromatic DAPI-dense regions. Future studies should strive to further validate this finding using detection reagents specific for the endogenous counterparts. Unfortunately, a suitable antibody that recognizes endogenous Eme1 in fixed cells or protein lysates is not currently available for mouse cells; however it should be possible to consider performing such
139 experiments using antibodies against endogenous UHRF1 and Mus81 (summarized in figure
37).
140
Role of chromatin accessibility in relation to γH2AX formation and amplification In UHRF1-depleted cells, we described a decrease in the percentage of cells with elevated γH2AX compared to control cells following irradiation, yet observed an increase in the number of UHRF1-depleted cells that were γH2AX-positive compared to control cells without irradiation. We have yet to conclusively determine whether or not loss of UHRF1 compromises repair of DNA damage specifically in heterochromatin. Given our observations of excess
γH2AX and increased incidence of chromosomal aberrations in UHRF1-depleted cells, we might expect that loss of UHRF1 could result in an increase in DNA lesions occurring during normal replication of heterochromatin, as well as a defect in the cell‟s ability to repair these lesions.
Future studies could address this possibility by examining whether γH2AX foci in UHRF1- depleted cells show a higher tendency to be located in or juxtaposed with heterochromatin
(which appear as DAPI-dense regions of DNA) compared to normal cells. Such an approach could also be utilized to examine whether the DSB damage caused by radiation accumulates preferentially in heterochromatin, in UHRF1-depleted cells.
We presently do not know at this time whether the attenuated γH2AX following DNA damage in these cells is due to a lower number of DSB lesions with γH2AX per cell or whether the amount of γH2AX/focus is reduced. We favour the latter possibility, as it is unlikely that the initial scission of the sugar-phosphate backbone following exposure of cells with γ-irradiation would not be altered by changes in chromatin structure. If a loss of UHRF1 renders the cell unable to relax heterochromatin in order to facilitate repair, we believe that this would contribute towards impairment in the amplification of γH2AX. The tight packing properties, as well as histone modifications such as H3K9me3 and Hp1 proteins are heterochromatin specific features
142 that may act at a barrier to DNA repair proteins, such as ATM. A blockage of ATM would likely limit the extent of γH2AX signal. Another reason why γH2AX is more prominent in areas of more open chromatin (euchromatin) is that differences in free radicals scavenging capabilities exist between euchromatin (more open) and heterochromatin (closed). Heterochromatin has more protection from free radical damage due to it being highly condensed and increased amount of histones (Cowell et al., 2007). Therefore it is possible that a compromised ability of cells to relax chromatin following DNA damage would be expected to impact the propagation of the
γH2AX within the damage foci. Closed heterochromatin is known to be refractory to γH2AX foci formation, thus if cells lacking H3K9me3 with a compromised ability to relax chromatin when exposed to γ-irradiation, we might expect that the γH2AX generated would be of smaller volume/lesion that those of normal cells. This would then result in a decreased γH2AX signal as was measured in our experiments by flow cytometry. Further experiments could investigate this possibility by performing densitometry on γH2AX foci or westerns by western analysis in normal vs. UHRF1-depleted cells exposed to equivalent doses of irradiation.
Does restoring the ability of UHRF1 to maintain H3K9me3 on heterochromatin rescue: (a) chromosomal instability and (b) sensitivity to irradiation? We have recently been able to complement the H3K9me3 defect observed in our
UHRF1-depleted cells by transfection of a vector that expresses murine UHRF1. Given that this rescue can be performed opens up a wide range of approaches for determining more precisely what activity of UHRF1 is required for maintenance of heterochromatic histone trimethylation and its relationship to the DNA damage response. In particular, this complementation approach can be utilized to query the specific domain or domains required for UHRF1 in heterochromatic
DSB repair. E.g. is the same domain required for H3K9me3 retention the same domain needed
143 for resistance to DNA damage? Our laboratory has already constructed vectors that are able to express murine UHRF1 RING finger domain mutants. Further mutants could be constructed with inactivating mutations in the UBL, tandem tudor, SRA and PHD domains. We would expect that these mutants would help us tease out the domain or domains required for UHRF1‟s role in DSB repair. We could then deduce which domains of UHRF1 are required for restoring the DNA damage response by evaluating the status of HP1β retention on chromatin, the formation of γH2AX, and acetyl-H2AK5 levels. A better understanding of how Eme1 and
URHF1 protects genome integrity will help us understand how these proteins mediate the sensitivity of cancer cells to γ-irradiation and chemotherapeutics that damage DNA and may aid in the rational design of improved therapeutic strategies.
144
REFERENCES
Aasland R, G.T., Stewart AF (1995). The PHD finger: implications for chromatin-mediated transcriptional regulation. Trends Biochem Sci 20, 56-59.
Abraham, J., Lemmers, B., Hande, M.P., Moynahan, M.E., Chahwan, C., Ciccia, A., Essers, J., Hanada, K., Chahwan, R., Khaw, A.K., et al. (2003). Eme1 is involved in DNA damage processing and maintenance of genomic stability in mammalian cells. Embo J 22, 6137-6147.
Acharya, S., Foster, P.L., Brooks, P., and Fishel, R. (2003). The coordinated functions of the E. coli MutS and MutL proteins in mismatch repair. Mol Cell 12, 233-246.
Achour, M., Fuhrmann, G., Alhosin, M., Ronde, P., Chataigneau, T., Mousli, M., Schini-Kerth, V.B., and Bronner, C. (2009). UHRF1 recruits the histone acetyltransferase Tip60 and controls its expression and activity. Biochem Biophys Res Commun 390, 523-528.
Allen, C., Ashley, A.K., Hromas, R., and Nickoloff, J.A. (2011). More forks on the road to replication stress recovery. J Mol Cell Biol 3, 4-12.
Allende-Vega, N., Dias, S., Milne, D., and Meek, D. (2005). Phosphorylation of the acidic domain of Mdm2 by protein kinase CK2. Mol Cell Biochem 274, 85-90.
Ariel, M., Robinson, E., McCarrey, J.R., and Cedar, H. (1995). Gamete-specific methylation correlates with imprinting of the murine Xist gene. Nat Genet 9, 312-315.
Arima, Y., Hirota, T., Bronner, C., Mousli, M., Fujiwara, T., Niwa, S., Ishikawa, H., and Saya, H. (2004). Down-regulation of nuclear protein ICBP90 by p53/p21Cip1/WAF1-dependent DNA- damage checkpoint signals contributes to cell cycle arrest at G1/S transition. Genes Cells 9, 131- 142.
Arita, K., Ariyoshi, M., Tochio, H., Nakamura, Y., Shirakawa, M. (2008). Recognition of hemi- methylated DNA by the SRA protein UHRF1 by a base-flipping mechanism. Nature 455, 818- 821.
Atamna, H., Cheung, I., and Ames, B.N. (2000). A method for detecting abasic sites in living cells: age-dependent changes in base excision repair. Proc Natl Acad Sci U S A 97, 686-691. Avner, P., and Heard, E. (2001). X-chromosome inactivation: counting, choice and initiation. Nat Rev Genet 2, 59-67.
Avvakumov, G.V., Walker, J.R., Xue, S., Li, Y., Duan, S., Bronner, C., Arrowsmith, C.H., Dhe- Paganon, S. (2008). Structural basis for recognition of hemi-methylated DNA by the SRA domains of human UHRF1. Nature 455, 822-825.
145
Ayoub, N., Jeyasekharan, A.D., Bernal, J.A., and Venkitaraman, A.R. (2008). HP1-beta mobilization promotes chromatin changes that initiate the DNA damage response. Nature 453, 682-686.
Ayoub, N., Jeyasekharan, A.D., and Venkitaraman, A.R. (2009). Mobilization and recruitment of HP1: a bimodal response to DNA breakage. Cell Cycle 8, 2945-2950.
Bannister, A.J., Zegerman, P., Partridge, J.F., Miska, E.A., Thomas, J.O., Allshire, R.C., and Kouzarides, T. (2001). Selective recognition of methylated lysine 9 on histone H3 by the HP1 chromo domain. Nature 410, 120-124.
Barakat, T.S., Jonkers, I., Monkhorst, K., and Gribnau, J. (2010). X-changing information on X inactivation. Exp Cell Res 316, 679-687.
Batinac, T., Gruber, F., Lipozencic, J., Zamolo-Koncar, G., Stasic, A., and Brajac, I. (2003). Protein p53--structure, function, and possible therapeutic implications. Acta Dermatovenerol Croat 11, 225-230.
Benhamou, S., and Sarasin, A. (2000). Variability in nucleotide excision repair and cancer risk: a review. Mutat Res 462, 149-158.
Bernstein, E., and Hake, S.B. (2006). The nucleosome: a little variation goes a long way. Biochem Cell Biol 84, 505-517.
Bhaumik, S.R., Smith, E., and Shilatifard, A. (2007). Covalent modifications of histones during development and disease pathogenesis. Nat Struct Mol Biol 14, 1008-1016.
Black, J.C., and Whetstine, J.R. (2011). Chromatin landscape: Methylation beyond transcription. Epigenetics 6.
Boddy, M.N., Gaillard, P.H., McDonald, W.H., Shanahan, P., Yates, J.R., 3rd, and Russell, P. (2001). Mus81-Eme1 are essential components of a Holliday junction resolvase. Cell 107, 537- 548.
Boddy, M.N., Lopez-Girona, A., Shanahan, P., Interthal, H., Heyer, W.D., and Russell, P. (2000). Damage tolerance protein Mus81 associates with the FHA1 domain of checkpoint kinase Cds1. Mol Cell Biol 20, 8758-8766.
Bonapace, I.M., Latella, L., Papait, R., Nicassio, F., Sacco, A., Muto, M., Crescenzi, M., and Di Fiore, P.P. (2002). Np95 is regulated by E1A during mitotic reactivation of terminally differentiated cells and is essential for S phase entry. J Cell Biol 157, 909-914.
Bonapace, I.M., Latella, L.,Papait, R., Nicassio, F., Sacco, A., Muto, M., Crescenzi, M., and Di Fiore, P.P. (2002). Np95 is regulated by E1A during mitotic reactivation of terminally differentiated cells and is essential for S phase entry. J Cell Biol 157, 909-914.
146
Bonner, W.M., Redon, C.E., Dickey, J.S., Nakamura, A.J., Sedelnikova, O.A., Solier, S., and Pommier, Y. (2008). GammaH2AX and cancer. Nat Rev Cancer 8, 957-967.
Bostick, M., Kim, J.K., Esteve, P.-O., Clark, A., Pradham, S., Jacobsen, S.E. (2007). UHFR1 plays a role in maintaining DNA methylation in mammalian cells. Science 317, 1760-1764. Boumil, R.M., and Lee, J.T. (2001). Forty years of decoding the silence in X-chromosome inactivation. Hum Mol Genet 10, 2225-2232.
Boye, E., Skjolberg, H.C., and Grallert, B. (2009). Checkpoint regulation of DNA replication. Methods Mol Biol 521, 55-70.
Bronner, C. (2011). Control of DNMT1 Abundance in Epigenetic Inheritance by Acetylation, Ubiquitylation, and the Histone Code. Sci Signal 4, pe3.
Bronner, C., Achour, M., Arima, Y., Chataigneau, T., Saya, H., Schini-Kerth, V.B. (2007). Pharmacol Ther 115, 419-434.
Brown, N.S., and Bicknell, R. (2001). Hypoxia and oxidative stress in breast cancer. Oxidative stress: its effects on the growth, metastatic potential and response to therapy of breast cancer. Breast Cancer Res 3, 323-327.
Burtner, C.R., and Kennedy, B.K. (2010). Progeria syndromes and ageing: what is the connection? Nat Rev Mol Cell Biol 11, 567-578.
Caldecott, K.W. (2007). Mammalian single-strand break repair: mechanisms and links with chromatin. DNA Repair (Amst) 6, 443-453.
Camerini-Otero, R.D., and Hsieh, P. (1995). Homologous recombination proteins in prokaryotes and eukaryotes. Annu Rev Genet 29, 509-552.
Capell, B.C., and Collins, F.S. (2006). Human laminopathies: nuclei gone genetically awry. Nat Rev Genet 7, 940-952.
Chakraverty, R.K., and Hickson, I.D. (1999). Defending genome integrity during DNA replication: a proposed role for RecQ family helicases. Bioessays 21, 286-294.
Chan, K.K., Zhang, Q.M., and Dianov, G.L. (2006). Base excision repair fidelity in normal and cancer cells. Mutagenesis 21, 173-178.
Cheung, H.H., Lee, T.L., Rennert, O.M., and Chan, W.Y. (2009). DNA methylation of cancer genome. Birth Defects Res C Embryo Today 87, 335-350.
Choi, H., Chun, Y.S., Shin, Y.J., Ye, S.K., Kim, M.S., and Park, J.W. (2008). Curcumin attenuates cytochrome P450 induction in response to 2,3,7,8-tetrachlorodibenzo-p-dioxin by ROS-dependently degrading AhR and ARNT. Cancer Sci 99, 2518-2524.
147
Ciccia, A., Constantinou, A., and West, S.C. (2003). Identification and characterization of the human mus81-eme1 endonuclease. J Biol Chem 278, 25172-25178.
Ciccia, A., McDonald, N., and West, S.C. (2008). Structural and functional relationships of the XPF/MUS81 family of proteins. Annu Rev Biochem 77, 259-287.
Citterio, E., Papait, R., Nicassio, F., Vecchi, M., Gomiero, P., Mantovani, R., Di Fiore, P.P., and Bonapace, I.M. (2004). Np95 is a histone-binding protein endowed with ubiquitin ligase activity. Mol Cell Biol 24, 2526-2535.
Citterio, E., Papait, R., Nicassio, F., Vecchi, M., Gomiero, P., Mantovani, R., Di Fiore, P.P., and Bonapace, I.M. (2004). Np95 is a histone-binding protein endowed with ubiquitin ligase activity. Mol Cell Biol 24, 2526-2535.
Clausell, J., Happel, N., Hale, T.K., Doenecke, D., and Beato, M. (2009). Histone H1 subtypes differentially modulate chromatin condensation without preventing ATP-dependent remodeling by SWI/SNF or NURF. PLoS One 4, e0007243.
Cleaver, J.E. (1968). Defective repair replication of DNA in xeroderma pigmentosum. Nature 218, 652-656.
Coupier, I., Baldeyron, C., Rousseau, A., Mosseri, V., Pages-Berhouet, S., Caux-Moncoutier, V., Papadopoulo, D., and Stoppa-Lyonnet, D. (2004). Fidelity of DNA double-strand break repair in heterozygous cell lines harbouring BRCA1 missense mutations. Oncogene 23, 914-919.
Coutinho, H.D., Falcao-Silva, V.S., Goncalves, G.F., and da Nobrega, R.B. (2009). Molecular ageing in progeroid syndromes: Hutchinson-Gilford progeria syndrome as a model. Immun Ageing 6, 4.
Cowell, I.G., Sunter, N.J., Singh, P.B., Austin, C.A., Durkacz, B.W., and Tilby, M.J. (2007). gammaH2AX foci form preferentially in euchromatin after ionising-radiation. PLoS One 2, e1057.
Crnogorac-Jurcevic, T., Gangeswaran, R., Bhakta, V., Capurso, G., Lattimore, S., Akada, M., Sunamura, M., Prime, W., Campbell, F., Brentnall, T.A., Costello, E., Neoptolemos, J., Lemoine, N.R. (2005). Gastroenterology. 129.
Cuervo, A.M., Wong, E.S., and Martinez-Vicente, M. (2010). Protein degradation, aggregation, and misfolding. Mov Disord 25 Suppl 1, S49-54.
Daujat, S., Weiss, T., Mohn, F., Lange, U.C., Ziegler-Birling, C., Zeissler, U., Lappe, M., Schubeler, D., Torres-Padilla, M.E., and Schneider, R. (2009). H3K64 trimethylation marks heterochromatin and is dynamically remodeled during developmental reprogramming. Nat Struct Mol Biol 16, 777-781.
148 de Boer, J., and Hoeijmakers, J.H. (2000). Nucleotide excision repair and human syndromes. Carcinogenesis 21, 453-460.
De Bont, R., and van Larebeke, N. (2004). Endogenous DNA damage in humans: a review of quantitative data. Mutagenesis 19, 169-185.
Denda, A., Endoh, T., Nakae, D., and Konishi, Y. (1995). Effects of oxidative stress induced by redox-enzyme modulation on rat hepatocarcinogenesis. Toxicol Lett 82-83, 413-417.
Dendouga, N., Gao, H., Moechars, D., Janicot, M., Vialard, J., and McGowan, C.H. (2005). Disruption of murine Mus81 increases genomic instability and DNA damage sensitivity but does not promote tumorigenesis. Mol Cell Biol 25, 7569-7579.
Dery, U., and Masson, J.Y. (2007). Twists and turns in the function of DNA damage signaling and repair proteins by post-translational modifications. DNA Repair (Amst) 6, 561-577. Dinant, C., and Luijsterburg, M.S. (2009). The emerging role of HP1 in the DNA damage response. Mol Cell Biol 29, 6335-6340.
Doe, C.L., Ahn, J.S., Dixon, J., and Whitby, M.C. (2002). Mus81-Eme1 and Rqh1 involvement in processing stalled and collapsed replication forks. J Biol Chem 277, 32753-32759. Downs, J.A., Nussenzweig, M.C., and Nussenzweig, A. (2007). Chromatin dynamics and the preservation of genetic information. Nature 447, 951-958.
Elmore, S. (2007). Apoptosis: a review of programmed cell death. Toxicol Pathol 35, 495-516. Enomoto, T., Takao, S., Mure, H., Baba, M., and Aikou, T. (1997). MTT-hybrid assay incorporates the advantages of both clonogenic and MTT assay radiosensitivity testing for fresh tumor samples. J Exp Clin Cancer Res 16, 273-280.
Errol C. Friedberg, G.C.W., Wolfram Siede (2005). DNA repair and mutagenesis. 1-663. Fabian, P. (2010). [The role of pathology in the predictive oncology of solid tumours]. Cas Lek Cesk 149, 462-463.
Fischle, W., Tseng, B.S., Dormann, H.L., Ueberheide, B.M., Garcia, B.A., Shabanowitz, J., Hunt, D.F., Funabiki, H., and Allis, C.D. (2005). Regulation of HP1-chromatin binding by histone H3 methylation and phosphorylation. Nature 438, 1116-1122.
Flatt, P.M., and Pietenpol, J.A. (2000). Mechanisms of cell-cycle checkpoints: at the crossroads of carcinogenesis and drug discovery. Drug Metab Rev 32, 283-305.
Fraga, M.F., Ballestar, E., Villar-Garea, A., Boix-Chornet, M., Espada, J., Schotta, G., Bonaldi, T., Haydon, C., Ropero, S., Petrie, K., et al. (2005). Loss of acetylation at Lys16 and trimethylation at Lys20 of histone H4 is a common hallmark of human cancer. Nat Genet 37, 391-400.
Friedberg, E.C. (2006). DNA repair and mutagenesis, 2nd edn (Washington, D.C., ASM Press).
149
Fujimori, A., Matsuda, Y., Takemoto, Y., Hashimoto, Y., Kubo, E., Araki, R., Fukumura, R., Mita, K., Tatsumi, K., and Muto, M. (1998). Cloning and mapping of NP95 gene which encodes a novel nuclear protein associated with cellular proliferation. Mamm Genome 9, 1032-1035.
Gillet, L.C., and Scharer, O.D. (2006). Molecular mechanisms of mammalian global genome nucleotide excision repair. Chem Rev 106, 253-276.
Goodarzi, A.A., Noon, A.T., Deckbar, D., Ziv, Y., Shiloh, Y., Lobrich, M., and Jeggo, P.A. (2008). ATM signaling facilitates repair of DNA double-strand breaks associated with heterochromatin. Mol Cell 31, 167-177.
Goodarzi, A.A., Noon, A.T., Deckbar, D., Ziv, Y., Shiloh, Y., Lobrich, M., Jeggo, P.A. (2008). ATM signaling facilitates repair of DNA double-strand breaks associated with heterochromatin. Mol Cell 31, 167-177.
Gorrini, C., Squatrito, M., Luise, C., Syed, N., Perna, D., Wark, L., Martinato, F., Sardella, D., Verrecchia, A., Bennett, S., et al. (2007). Tip60 is a haplo-insufficient tumour suppressor required for an oncogene-induced DNA damage response. Nature 448, 1063-1067.
Grant, P.A. (2001). A tale of histone modifications. Genome Biol 2, REVIEWS0003.
Grewal, S.I., and Rice, J.C. (2004). Regulation of heterochromatin by histone methylation and small RNAs. Curr Opin Cell Biol 16, 230-238.
Grignani, F., De Matteis, S., Nervi, C., Tomassoni, L., Gelmetti, V., Cioce, M., Fanelli, M., Ruthardt, M., Ferrara, F.F., Zamir, I., et al. (1998). Fusion proteins of the retinoic acid receptor- alpha recruit histone deacetylase in promyelocytic leukaemia. Nature 391, 815-818.
Guenatri, M., Bailly, D., Maison, C., and Almouzni, G. (2004). Mouse centric and pericentric satellite repeats form distinct functional heterochromatin. J Cell Biol 166, 493-505.
Gui, C.Y., Ngo, L., Xu, W.S., Richon, V.M., and Marks, P.A. (2004). Histone deacetylase (HDAC) inhibitor activation of p21WAF1 involves changes in promoter-associated proteins, including HDAC1. Proc Natl Acad Sci U S A 101, 1241-1246.
Hakimi, M.A., Bochar, D.A., Chenoweth, J., Lane, W.S., Mandel, G., and Shiekhattar, R. (2002). A core-BRAF35 complex containing histone deacetylase mediates repression of neuronal- specific genes. Proc Natl Acad Sci U S A 99, 7420-7425.
Hakimi, M.A., Dong, Y., Lane, W.S., Speicher, D.W., and Shiekhattar, R. (2003). A candidate X-linked mental retardation gene is a component of a new family of histone deacetylase- containing complexes. J Biol Chem 278, 7234-7239.
Halkidou, K., Gnanapragasam, V.J., Mehta, P.B., Logan, I.R., Brady, M.E., Cook, S., Leung, H.Y., Neal, D.E., and Robson, C.N. (2003). Expression of Tip60, an androgen receptor coactivator, and its role in prostate cancer development. Oncogene 22, 2466-2477.
150
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.
Hanada, K., Budzowska, M., Modesti, M., Maas, A., Wyman, C., Essers, J., and Kanaar, R. (2006). The structure-specific endonuclease Mus81-Eme1 promotes conversion of interstrand DNA crosslinks into double-strands breaks. EMBO J 25, 4921-4932.
Hanahan, D., and Weinberg, R.A. (2000). The hallmarks of cancer. Cell 100, 57-70.
Hansen, R.S., Thomas, S., Sandstrom, R., Canfield, T.K., Thurman, R.E., Weaver, M., Dorschner, M.O., Gartler, S.M., and Stamatoyannopoulos, J.A. (2010). Sequencing newly replicated DNA reveals widespread plasticity in human replication timing. Proc Natl Acad Sci U S A 107, 139-144.
Harbour, J.W., and Dean, D.C. (2000). Rb function in cell-cycle regulation and apoptosis. Nat Cell Biol 2, E65-67.
Hashimoto H, H.J., Zhang X, Bostick M, Jacobsen SE, Cheng X. (2008). The SRA domain of UHRF1 flips 5-methylcytosine out of the DNA helix. . Nature 455, 826-829.
Heard, E., Rougeulle, C., Arnaud, D., Avner, P., Allis, C.D., and Spector, D.L. (2001). Methylation of histone H3 at Lys-9 is an early mark on the X chromosome during X inactivation. Cell 107, 727-738.
Hejna, J., Holtorf, M., Hines, J., Mathewson, L., Hemphill, A., Al-Dhalimy, M., Olson, S.B., and Moses, R.E. (2008). Tip60 is required for DNA interstrand cross-link repair in the Fanconi anemia pathway. J Biol Chem 283, 9844-9851.
Helt, C.E., Cliby, W.A., Keng, P.C., Bambara, R.A., and O'Reilly, M.A. (2005). Ataxia telangiectasia mutated (ATM) and ATM and Rad3-related protein exhibit selective target specificities in response to different forms of DNA damage. J Biol Chem 280, 1186-1192. Heyer, W.D. (2004). Recombination: Holliday junction resolution and crossover formation. Curr Biol 14, R56-58.
Hickson, I.D. (2003). RecQ helicases: caretakers of the genome. Nat Rev Cancer 3, 169-178.
Hiyama, T., Katsura, M., Yoshihara, T., Ishida, M., Kinomura, A., Tonda, T., Asahara, T., and Miyagawa, K. (2006). Haploinsufficiency of the Mus81-Eme1 endonuclease activates the intra- S-phase and G2/M checkpoints and promotes rereplication in human cells. Nucleic Acids Res 34, 880-892.
Hoeijmakers, J.H. (2001). DNA repair mechanisms. Maturitas 38, 17-22; discussion 22-13.
151
Hoffelder, D.R., Luo, L., Burke, N.A., Watkins, S.C., Gollin, S.M., Saunders, W.S. (2004). Resolution of anaphase bridges in cancer cells. Chromosoma 112, 389-397.
Hopfner, R., Mousli, M., Jeltsch, J.-M., Voulgaris, A., Lutz, Y., Marin, C., Bellocq, J.-P., Oudet, P., and Bronner, C. (2000). ICBP90, a novel human CCAAT binding protein, involved in the regulation of Topoisomerase IIa expression. Cancer Res 60, 121-128.
Hopfner, R., Mousli, M., Oudet, R., Bronner, C. (2002). Anticancer Res 22, 3165-3170.
Houtgraaf, J.H., Versmissen, J., and van der Giessen, W.J. (2006). A concise review of DNA damage checkpoints and repair in mammalian cells. Cardiovasc Revasc Med 7, 165-172.
Interthal, H., and Heyer, W.D. (2000). MUS81 encodes a novel helix-hairpin-helix protein involved in the response to UV- and methylation-induced DNA damage in Saccharomyces cerevisiae. Mol Gen Genet 263, 812-827.
Iyer, R.R., Pluciennik, A., Burdett, V., and Modrich, P.L. (2006). DNA mismatch repair: functions and mechanisms. Chem Rev 106, 302-323.
Jascur, T., and Boland, C.R. (2006). Structure and function of the components of the human DNA mismatch repair system. Int J Cancer 119, 2030-2035.
Jeggo, P.A., Carr, A.M., and Lehmann, A.R. (1998). Splitting the ATM: distinct repair and checkpoint defects in ataxia-telangiectasia. Trends Genet 14, 312-316.
Jenkins, Y., Markovtsov, V., Lang, W., Sharma, P., Pearsall, D., Warner, J., Franci, C., Huang, B., Huang, J., Yam, G.C., et al. (2005). Critical role of the ubiquitin ligase activity of UHRF1, a nuclear RING finger protein, in tumor cell growth. Mol Biol Cell 16, 5621-5629.
Jenkins, Y., Markovtsov, V., Lang, W., Sharma, P., Pearsall, D., Warner, J., Franci, C.,. Huang, B., Huang, J., Yam, G.C., Vistan, J.P., Pali, E., Vialard, J., Janicot, M., Lorens, J.B., Payan, D.G., Hitoshi, Y. (2005). Critical role of the ubiquitin ligase activity of UHFR1, a nuclear RING finger protein, in tumour cell growth. Mol Biol Cell 16, 5621-5629.
Jiang, X., Sun, Y., Chen, S., Roy, K., and Price, B.D. (2006). The FATC domains of PIKK proteins are functionally equivalent and participate in the Tip60-dependent activation of DNA- PKcs and ATM. J Biol Chem 281, 15741-15746.
Jiricny, J. (2006). The multifaceted mismatch-repair system. Nat Rev Mol Cell Biol 7, 335-346. John M. Walker, R.R. (2008). Molecular Biomethods Handbook. In (Humana Press).
Jones, B., Su, H., Bhat, A., Lei, H., Bajko, J., Hevi, S., Baltus, G.A., Kadam, S., Zhai, H., Valdez, R., et al. (2008). The histone H3K79 methyltransferase Dot1L is essential for mammalian development and heterochromatin structure. PLoS Genet 4, e1000190.
Kalant, H. (2007). Principles of Medical Pharmacology, 7 edn.
152
Kaliraman, V., Mullen, J.R., Fricke, W.M., Bastin-Shanower, S.A., and Brill, S.J. (2001). Functional overlap between Sgs1-Top3 and the Mms4-Mus81 endonuclease. Genes Dev 15, 2730-2740.
Karagianni, P., Amazit, L., Qin, J., and Wong, J. (2008). ICBP90, a novel methyl K9 H3 binding protein linking protein ubiquitination with heterochromatin formation. Mol Cell Biol 28, 705- 717.
Karnani, N., Taylor, C.M., Malhotra, A., and Dutta, A. (2010). Genomic study of replication initiation in human chromosomes reveals the influence of transcription regulation and chromatin structure on origin selection. Mol Biol Cell 21, 393-404.
Kastan, M.B., and Bartek, J. (2004). Cell-cycle checkpoints and cancer. Nature 432, 316-323.
Kennedy, R.D., and D'Andrea, A.D. (2005). The Fanconi Anemia/BRCA pathway: new faces in the crowd. Genes Dev 19, 2925-2940.
Kerr, J.F., Winterford, C.M., and Harmon, B.V. (1994). Apoptosis. Its significance in cancer and cancer therapy. Cancer 73, 2013-2026.
Kim, J.H., Kim, B., Cai, L., Choi, H.J., Ohgi, K.A., Tran, C., Chen, C., Chung, C.H., Huber, O., Rose, D.W., et al. (2005). Transcriptional regulation of a metastasis suppressor gene by Tip60 and beta-catenin complexes. Nature 434, 921-926.
Kim, J.K., Esteve, P.O., Jacobsen, S.E., and Pradhan, S. (2009). UHRF1 binds G9a and participates in p21 transcriptional regulation in mammalian cells. Nucleic Acids Res 37, 493- 505.
Kimura, A., and Horikoshi, M. (1998). Tip60 acetylates six lysines of a specific class in core histones in vitro. Genes Cells 3, 789-800.
Kinzler, K.W., and Vogelstein, B. (1997). Cancer-susceptibility genes. Gatekeepers and caretakers. Nature 386, 761, 763.
Knudsen, C.R., Jadidi, M., Friis, I., and Mansilla, F. (2002). Application of the yeast two-hybrid system in molecular gerontology. Biogerontology 3, 243-256.
Kobayashi, J., Iwabuchi, K., Miyagawa, K., Sonoda, E., Suzuki, K., Takata, M., and Tauchi, H. (2008). Current topics in DNA double-strand break repair. J Radiat Res (Tokyo) 49, 93-103.
Kouzarides, T. (2007). Chromatin modifications and their function. Cell 128, 693-705.
Kristensen, L.S., Nielsen, H.M., and Hansen, L.L. (2009). Epigenetics and cancer treatment. Eur J Pharmacol 625, 131-142.
153
Kurdistani, S.K., and Grunstein, M. (2003). Histone acetylation and deacetylation in yeast. Nat Rev Mol Cell Biol 4, 276-284.
La Thangue, N.B. (2004). Histone deacetylase inhibitors and cancer therapy. J Chemother 16 Suppl 4, 64-67.
Lai, S., Benedict, W.F., Silver, S.A., and El-Naggar, A.K. (1997). Loss of retinoblastoma gene function and heterozygosity at the RB locus in renal cortical neoplasms. Hum Pathol 28, 693- 697.
Leeb, M., and Wutz, A. (2010). Mechanistic concepts in X inactivation underlying dosage compensation in mammals. Heredity 105, 64-70.
Leibeling, D., Laspe, P., and Emmert, S. (2006). Nucleotide excision repair and cancer. J Mol Histol 37, 225-238.
Levenson, V.V. (2010). DNA methylation as a universal biomarker. Expert Rev Mol Diagn 10, 481-488.
Levine, A.J. (1997). p53, the cellular gatekeeper for growth and division. Cell 88, 323-331. Lichtman, M.A. (2008). Battling the hematological malignancies: the 200 years' war. Oncologist 13, 126-138.
Lieber, M.R. (2010). The mechanism of double-strand DNA break repair by the nonhomologous DNA end-joining pathway. Annu Rev Biochem 79, 181-211.
Lieber, M.R., and Wilson, T.E. (2010). SnapShot: Nonhomologous DNA end joining (NHEJ). Cell 142, 496-496 e491.
Lindahl, T. (1990). Repair of intrinsic DNA lesions. Mutat Res 238, 305-311. Ljungman, M. (2010). The DNA damage response--repair or despair? Environ Mol Mutagen 51, 879-889.
Loizou, J.I., El-Khamisy, S.F., Zlatanou, A., Moore, D.J., Chan, D.W., Qin, J., Sarno, S., Meggio, F., Pinna, L.A., and Caldecott, K.W. (2004). The protein kinase CK2 facilitates repair of chromosomal DNA single-strand breaks. Cell 117, 17-28.
Lorenzato, M., Caudroy, S., Bronner, C., Evrard, G., Simon, M., Durlach, A., Birembaut, P., Clavel., C. (2005). Human Pathol 36, 1101-1107.
Loyola, A., Tagami, H., Bonaldi, T., Roche, D., Quivy, J.P., Imhof, A., Nakatani, Y., Dent, S.Y., and Almouzni, G. (2009). The HP1alpha-CAF1-SetDB1-containing complex provides H3K9me1 for Suv39-mediated K9me3 in pericentric heterochromatin. EMBO Rep 10, 769-775.
154
Luijsterburg, M.S., Dinant, C., Lans, H., Stap, J., Wiernasz, E., Lagerwerf, S., Warmerdam, D.O., Lindh, M., Brink, M.C., Dobrucki, J.W., et al. (2009). Heterochromatin protein 1 is recruited to various types of DNA damage. J Cell Biol 185, 577-586.
MacAlpine, D.M., Rodriguez, H.K., and Bell, S.P. (2004). Coordination of replication and transcription along a Drosophila chromosome. Genes Dev 18, 3094-3105.
MacAlpine, H.K., Gordan, R., Powell, S.K., Hartemink, A.J., and MacAlpine, D.M. (2010). Drosophila ORC localizes to open chromatin and marks sites of cohesin complex loading. Genome Res 20, 201-211.
Mah, L.J., Vasireddy, R.S., Tang, M.M., Georgiadis, G.T., El-Osta, A., and Karagiannis, T.C. (2010). Quantification of gammaH2AX foci in response to ionising radiation. J Vis Exp. Mahlknecht, U., and Hoelzer, D. (2000). Histone acetylation modifiers in the pathogenesis of malignant disease. Mol Med 6, 623-644.
Maison, C., and Almouzni, G. (2004). HP1 and the dynamics of heterochromatin maintenance. Nat Rev Mol Cell Biol 5, 296-304.
Malumbres, M., and Barbacid, M. (2009). Cell cycle, CDKs and cancer: a changing paradigm. Nat Rev Cancer 9, 153-166.
McPherson, J.P., Lemmers, B., Chahwan, R., Pamidi, A., Migon, E., Matysiak-Zablocki, E., Moynahan, M.E., Essers, J., Hanada, K., Poonepalli, A., et al. (2004). Involvement of mammalian Mus81 in genome integrity and tumor suppression. Science 304, 1822-1826.
ME, L.L., Vidal, F., Gallardo, D., Diaz-Fuertes, M., Rojo, F., Cuatrecasas, M., Lopez-Vicente, L., Kondoh, H., Blanco, C., Carnero, A., et al. (2006). New p53 related genes in human tumors: significant downregulation in colon and lung carcinomas. Oncol Rep 16, 603-608.
Meilenger, D., Fellinger, K., Bultmann, S., Rothbauer, U., Bonapace, I.,M., Klinkert, W.E.F., Spada, F., Leonhardt, H. (2009). Np95 interacts with de novo DNA methyltransferases, Dnmt3a and Dnmt3b, and mediates epigenetic silencing of the viral CMV promoter in embryonic stem cells. EMBO reports 10, 1259-1264.
Mendez-Acuna, L., Di Tomaso, M.V., Palitti, F., and Martinez-Lopez, W. (2010). Histone post- translational modifications in DNA damage response. Cytogenet Genome Res 128, 28-36. Mersfelder, E.L., and Parthun, M.R. (2006). The tale beyond the tail: histone core domain modifications and the regulation of chromatin structure. Nucleic Acids Res 34, 2653-2662.
Mi, J., and Kupfer, G.M. (2005). The Fanconi anemia core complex associates with chromatin during S phase. Blood 105, 759-766. Mistry, H., Gibson, L., Yun, J.W., Sarras, H., Tamblyn, L., and McPherson, J.P. (2008). Interplay between Np95 and Eme1 in the DNA damage response. Biochem Biophys Res Commun 375, 321-325.
155
Mistry, H., Tamblyn, L., Butt, H., Sisgoreo, D., Gracias, A., Larin, M., Gopalakrishnan, K., Hande, M.P., McPherson, J.P. (2010). Uhrf1 is a genome caretaker that facilitates the DNA damage response to y-Irradiation. Gemome Integrity.
Miura, M., Watanabe, H., Sasaki, T., Tatsumi, K.,Muto, M. (2001). Dynamic changes in subnuclear NP95 location during the cell cycle and its spatial relationship with DNA replication foci. Exp Cell Res 263, 202-208.
Moldovan, G.L., and D'Andrea, A.D. (2009). How the fanconi anemia pathway guards the genome. Annu Rev Genet 43, 223-249.
Montecucco, A., Rossi, R., Ferrari, G., Scovassi, A.I., Prosperi, E., and Biamonti, G. (2001). Etoposide induces the dispersal of DNA ligase I from replication factories. Mol Biol Cell 12, 2109-2118.
Mousli, M., Hopfner, R., Abbady, A.Q., Monte, D., Jeanblanc, M., Oudet, P., Louis, B., Bronner, C. (2003). Br J Cancer 89, 120-127.
Munster, P.N., Troso-Sandoval, T., Rosen, N., Rifkind, R., Marks, P.A., and Richon, V.M. (2001). The histone deacetylase inhibitor suberoylanilide hydroxamic acid induces differentiation of human breast cancer cells. Cancer Res 61, 8492-8497.
Murga, M., Jaco, I., Fan, Y., Soria, R., Martinez-Pastor, B., Cuadrado, M., Yang, S.M., Blasco, M.A., Skoultchi, A.I., and Fernandez-Capetillo, O. (2007). Global chromatin compaction limits the strength of the DNA damage response. J Cell Biol 178, 1101-1108.
Muto, M., Fujimori, A., Nenoi, M., Daino, K., Matsuda, Y., Kuroiwa, A., Kubo, E., Kanari, Y., Utsuno, M., Tsuji, H., et al. (2006). Isolation and Characterization of a Novel Human Radiosusceptibility Gene, NP95. Radiat Res 166, 723-733.
Muto, M., Fujimori, A,. Nenoi, M., Daino, K., Matsuda, Y., Kuroiwa, A., Kubo, E., Kanari, Y., Utsuno, M., Tsuji, H., Ukai, H., Mita, K, Takahagi, M., Tatsumi, K. (2006). Isolation and characterization of a novel human radiosusceptibility gene, NP95. Radiation Research 723-733.
Muto, M., Kanari, Y., Kubo, E., Takabe, T., Kurihara, T., Fujimori, A., and Tatsumi, K. (2002). Targeted disruption of Np95 gene renders murine embryonic stem cells hypersensitive to DNA damaging agents and DNA replication blocks. J Biol Chem 277, 34549-34555.
Muto, M., Utsuyama, M., Horiguchi, T., Kubo, E., Sado, T., Hirokawa, K. (1995). The characterization of the monoclonal antibody Th-10a, specific for a nuclear protein appearing in the S phase of the cell cycle in normal thymocytes and its upregulated expression in lymphoma cell lines. Cell Proliferation 28, 645-657. Navarro, C.L., Cau, P., and Levy, N. (2006). Molecular bases of progeroid syndromes. Hum Mol Genet 15 Spec No 2, R151-161.
156
Noll, D.M., Mason, T.M., and Miller, P.S. (2006). Formation and repair of interstrand cross-links in DNA. Chem Rev 106, 277-301.
Oba-Shinjo, S.M., Bengtson, M.H., Winnischofer, S.M., Colin, C., Vedoy, C.G., de Mendonca, Z., Marie, S.K., Sogayar, M.C. (2005). Mol Brain Res 140, 25-33.
Pamidi, A., Cardoso, R., Hakem, A., Matysiak-Zablocki, E., Poonepalli, A., Tamblyn, L., Perez- Ordonez, B., Hande, M.P., Sanchez, O., and Hakem, R. (2007). Functional interplay of p53 and Mus81 in DNA damage responses and cancer. Cancer Res 67, 8527-8535.
Papait, R., Pistore, C., Grazini, U., Babbio, F., Cogliati, S., Pecoraro, D., Brino, L., Morand, A.L., Dechampesme, A.M., Spada, F., et al. (2008). The PHD domain of Np95 (mUHRF1) is involved in large-scale reorganization of pericentromeric heterochromatin. Mol Biol Cell 19, 3554-3563.
Papait, R., Pistore, C., Negri, D., Pecoraro, D., Cantarini, L., and Bonapace, I.M. (2007). Np95 is implicated in pericentromeric heterochromatin replication and in major satellite silencing. Mol Biol Cell 18, 1098-1106.
Papait, R., Pistore, C., Grazini, U., Babbio, F., Cogliati, S., Pecoraro, D., Brino, L., Morand, A.- L., Dechampesme, A.-M., Spada, F., Leonhardt, H., McBlane, F., Oudet, P., Bonapace, I.M. (2008). The PHD domain of Np95 (mUHRF1) is involved in large-scale reorganization of pericentromeric heterochromatin. Mol Biol Cell 19, 3554-3563.
Papait, R., Pistore, C., Negri, D., Pecoraro, D., Cantarini, L., and Bonapace, I.M. (2007). Np95 is implicated in pericentromeric Heterochromatin replication and in major satellite silencing. Mol Biol Cell 18, 1098-1106.
Peters, A.H., Mermoud, J.E., O'Carroll, D., Pagani, M., Schweizer, D., Brockdorff, N., and Jenuwein, T. (2002). Histone H3 lysine 9 methylation is an epigenetic imprint of facultative heterochromatin. Nat Genet 30, 77-80.
Peters, A.H., O'Carroll, D., Scherthan, H., Mechtler, K., Sauer, S., Schofer, C., Weipoltshammer, K., Pagani, M., Lachner, M., Kohlmaier, A., et al. (2001). Loss of the Suv39h histone methyltransferases impairs mammalian heterochromatin and genome stability. Cell 107, 323- 337.
Pita, J.M., Banito, A., Cavaco, B.M., Leite, V. (2009). Gene expression profiling associated with the progression to poorly differentiated thyroid carcinoma. Br J Cancer DOI:10.1038/sj.bjc.66045340.
Prise, K.M., Pinto, M., Newman, H.C., and Michael, B.D. (2001). A review of studies of ionizing radiation-induced double-strand break clustering. Radiat Res 156, 572-576.
157
Quivy, J.P., Gerard, A., Cook, A.J., Roche, D., and Almouzni, G. (2008). The HP1-p150/CAF-1 interaction is required for pericentric heterochromatin replication and S-phase progression in mouse cells. Nat Struct Mol Biol 15, 972-979.
Ray, A., and Langer, M. (2002). Homologous recombination: ends as the means. Trends Plant Sci 7, 435-440.
Rea, S., Eisenhaber, F., O'Carroll, D., Strahl, B.D., Sun, Z.W., Schmid, M., Opravil, S., Mechtler, K., Ponting, C.P., Allis, C.D., et al. (2000). Regulation of chromatin structure by site- specific histone H3 methyltransferases. Nature 406, 593-599.
Rideout, W.M., 3rd, Coetzee, G.A., Olumi, A.F., and Jones, P.A. (1990). 5-Methylcytosine as an endogenous mutagen in the human LDL receptor and p53 genes. Science 249, 1288-1290. Robertson, J.D., Orrenius, S., and Zhivotovsky, B. (2000). Review: nuclear events in apoptosis. J Struct Biol 129, 346-358.
Rogakou, E.P., Pilch, D.R., Orr, A.H., Ivanova, V.S., and Bonner, W.M. (1998). DNA double- stranded breaks induce histone H2AX phosphorylation on serine 139. J Biol Chem 273, 5858- 5868.
Rottach, A., Frauer, C., Pichler, G., Bonapace, I.M., Spada, F., and Leonhardt, H. (2010). The multi-domain protein Np95 connects DNA methylation and histone modification. Nucleic Acids Res 38, 1796-1804.
Rydberg, B. (1996). Clusters of DNA damage induced by ionizing radiation: formation of short DNA fragments. II. Experimental detection. Radiat Res 145, 200-209.
San Filippo, J., Sung, P., and Klein, H. (2008). Mechanism of eukaryotic homologous recombination. Annu Rev Biochem 77, 229-257.
Sancar, A., Lindsey-Boltz, L.A., Unsal-Kacmaz, K., and Linn, S. (2004). Molecular mechanisms of mammalian DNA repair and the DNA damage checkpoints. Annu Rev Biochem 73, 39-85. Santoro, R., and Blandino, G. (2010). p53: The pivot between cell cycle arrest and senescence. Cell Cycle 9, 4262-4263.
Santos-Rosa, H., and Caldas, C. (2005). Chromatin modifier enzymes, the histone code and cancer. Eur J Cancer 41, 2381-2402.
Schaaf, G.J., Ruijter, J.M., van Ruissen, F., Zwinjnenburg, D.A., Waaiejer, R., Valentijn, L.J., Benit-Deekman, J., van Kampen, A.H., Baas, F., and Kool, M. (2005). Faseb J 19, 404-406.
Schmitt, E., Paquet, C., Beauchemin, M., and Bertrand, R. (2007). DNA-damage response network at the crossroads of cell-cycle checkpoints, cellular senescence and apoptosis. J Zhejiang Univ Sci B 8, 377-397.
158
Schmuckli-Maurer, J., Rolfsmeier, M., Nguyen, H., and Heyer, W.D. (2003). Genome instability in rad54 mutants of Saccharomyces cerevisiae. Nucleic Acids Res 31, 1013-1023.
Serebriiskii, I.G., and Golemis, E.A. (2001). Two-hybrid system and false positives. Approaches to detection and elimination. Methods Mol Biol 177, 123-134.
Sharif, J., Muto, M., Takebayashi, S., Suetake, I., Iwamatsu, A., Endo, T.A., Shinga, J., Mizutani-Koseki, Y., Toyoda, T., Okamura, K., et al. (2007). The SRA protein Np95 mediates epigenetic inheritance by recruiting Dnmt1 to methylated DNA. Nature 450, 908-912.
Sharif, J., Muto, M., Takebayashi, S.-I., Suetake, I., Iwamatsu, A., Endo, T.A., Shinga, J., Mizutani-Koseki, Y., Toyoda, T., Okamura, K., Tajima, S, Mitsuya, K., Okano, M., Koseki, H. (2007). The SRA protein Np95 mediates epigenetic inheritance by recruiting Dnmt1 to methylated DNA. Nature 450, 908-912.
Sinkovics, J.G. (1991). Programmed cell death (apoptosis): its virological and immunological connections (a review). Acta Microbiol Hung 38, 321-334.
Sun, Y., Jiang, X., Chen, S., Fernandes, N., and Price, B.D. (2005). A role for the Tip60 histone acetyltransferase in the acetylation and activation of ATM. Proc Natl Acad Sci U S A 102, 13182-13187.
Sun, Y., Jiang, X., and Price, B.D. (2010). Tip60: connecting chromatin to DNA damage signaling. Cell Cycle 9, 930-936.
Sun, Y., Jiang, X., Xu, Y., Ayrapetov, M.K., Moreau, L.A., Whetstine, J.R., and Price, B.D. (2009). Histone H3 methylation links DNA damage detection to activation of the tumour suppressor Tip60. Nat Cell Biol 11, 1376-1382.
Svendsen, J.M., Smogorzewska, A., Sowa, M.E., O'Connell, B.C., Gygi, S.P., Elledge, S.J., and Harper, J.W. (2009). Mammalian BTBD12/SLX4 assembles a Holliday junction resolvase and is required for DNA repair. Cell 138, 63-77.
Taddei, A., Maison, C., Roche, D., and Almouzni, G. (2001). Reversible disruption of pericentric heterochromatin and centromere function by inhibiting deacetylases. Nat Cell Biol 3, 114-120. Talbert, P.B., and Henikoff, S. (2010). Histone variants--ancient wrap artists of the epigenome. Nat Rev Mol Cell Biol 11, 264-275.
Therman, E., Susman, M. (1993). Human Chromosomes, Structure, Behaviour and Effects (New York, NY, Springer-Verlag).
Thompson, S.L., Bakhoum, S.F., and Compton, D.A. (2010). Mechanisms of chromosomal instability. Curr Biol 20, R285-295.
159
Tien, A.L., Senbanerjee, S., Kulkarni, A., Mudbhary, R., Goudreau, B., Ganesan, S., Sadler, K.C., and Ukomadu, C. (2011). UHRF1 depletion causes a G2/M arrest, activation of DNA damage response and apoptosis. Biochem J.
Tomoda, Y., Katsura, M., Okajima, M., Hosoya, N., Kohno, N., and Miyagawa, K. (2009). Functional evidence for Eme1 as a marker of cisplatin resistance. Int J Cancer 124, 2997-3001.
Trosko, J.E., and Chang, C.C. (2010). Factors to consider in the use of stem cells for pharmaceutic drug development and for chemical safety assessment. Toxicology 270, 18-34.
Uemura, T., Kubo, E., Kanaari, Y., Ikemura, T., Tatsumi, K., and Muto, M. (2000). Temporal and spatial localization of novel nuclear protein NP95 in mitotic and meiotic cells. Cell Structure and Function 25, 149-159.
Unoki, M., Brunet, J., and Mousli, M. (2009a). Drug discovery targeting epigenetic codes: the great potential of UHRF1, which links DNA methylation and histone modifications, as a drug target in cancers and toxoplasmosis. Biochem Pharmacol 78, 1279-1288.
Unoki, M., Daigo, Y., Koinuma, J., Tsuchiya, E., Hamamoto, R., and Nakamura, Y. (2010). UHRF1 is a novel diagnostic marker of lung cancer. Br J Cancer 103, 217-222.
Unoki, M., Kelly, J.D., Neal, D.E., Ponder, B.A., Nakamura, Y., and Hamamoto, R. (2009b). UHRF1 is a novel molecular marker for diagnosis and the prognosis of bladder cancer. Br J Cancer 101, 98-105.
Unoki, M., Nishidate, T., and Nakamura, Y. (2004). ICBP90, an E2F-1 target, recruits HDAC1 and binds to methyl-CpG through its SRA domain. Oncogene 23, 7601-7610.
Unoki, M., Nishidate, T., and Nakamura, Y. (2004). ICBP90, an E2F-1 target, recruits HDAC1 and binds to methyl-CpG through its SRA domain. Oncogene 23, 7601-7610. Valerie, K., and Povirk, L.F. (2003). Regulation and mechanisms of mammalian double-strand break repair. Oncogene 22, 5792-5812.
Gent, D.C., Hoeijmakers, J.H., and Kanaar, R. (2001). Chromosomal stability and the DNA double-stranded break connection. Nat Rev Genet 2, 196-206. Vaquero, A., Loyola, A., and Reinberg, D. (2003). The constantly changing face of chromatin. Sci Aging Knowledge Environ 2003, RE4.
Verbeek, B., Southgate, T.D., Gilham, D.E., and Margison, G.P. (2008). O6-Methylguanine- DNA methyltransferase inactivation and chemotherapy. Br Med Bull 85, 17-33.
Vidanes, G.M., Bonilla, C.Y., and Toczyski, D.P. (2005). Complicated tails: histone modifications and the DNA damage response. Cell 121, 973-976.
160
Vilenchik, M.M., and Knudson, A.G. (2003). Endogenous DNA double-strand breaks: production, fidelity of repair, and induction of cancer. Proc Natl Acad Sci U S A 100, 12871- 12876.
Warenius, H.M., Jones, M., Jones, M.D., Browning, P.G., Seabra, L.A., and Thompson, C.C. (1998). Late G1 accumulation after 2 Gy of gamma-irradiation is related to endogenous Raf-1 protein expression and intrinsic radiosensitivity in human cells. Br J Cancer 77, 1220-1228.
Whitby, M.C. (2004). Junctions on the road to cancer. Nat Struct Mol Biol 11, 693-695.
Whitby, M.C., Osman, F., and Dixon, J. (2003). Cleavage of model replication forks by fission yeast Mus81-Eme1 and budding yeast Mus81-Mms4. J Biol Chem 278, 6928-6935.
Wood, R.D. (2010). Mammalian nucleotide excision repair proteins and interstrand crosslink repair. Environ Mol Mutagen 51, 520-526.
Wood, R.D., Mitchell, M., and Lindahl, T. (2005). Human DNA repair genes, 2005. Mutat Res 577, 275-283.
Yang, X.J., and Seto, E. (2008). Lysine acetylation: codified crosstalk with other posttranslational modifications. Mol Cell 31, 449-461.
Yarden, R.I., and Brody, L.C. (1999). BRCA1 interacts with components of the histone deacetylase complex. Proc Natl Acad Sci U S A 96, 4983-4988.
You, A., Tong, J.K., Grozinger, C.M., and Schreiber, S.L. (2001). CoREST is an integral component of the CoREST- human histone deacetylase complex. Proc Natl Acad Sci U S A 98, 1454-1458.
Zhang, R., Sengupta, S., Yang, Q., Linke, S.P., Yanaihara, N., Bradsher, J., Blais, V., McGowan, C.H., and Harris, C.C. (2005). BLM helicase facilitates Mus81 endonuclease activity in human cells. Cancer Res 65, 2526-2531.
Zhou, B.B., and Elledge, S.J. (2000). The DNA damage response: putting checkpoints in perspective. Nature 408, 433-439.
Ziv, Y., Bielopolski, D., Galanty, Y., Lukas, C., Taya, Y., Schultz, D.C., Lukas, J., Bekker- Jensen, S., Bartek, J., and Shiloh, Y. (2006). Chromatin relaxation in response to DNA double- strand breaks is modulated by a novel ATM- and KAP-1 dependent pathway. Nat Cell Biol 8, 870-876.
161