Instability at Trinucleotide Repeat

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science

By

RUJUTA YASHODHAN GADGIL

M.Sc. Microbiology, Fergusson College, India, 2013

______

2016 Wright State University

i

WRIGHT STATE UNIVERSITY

GRADUATE SCHOOL

August 1, 2016

I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY Rujuta Yashodhan Gadgil ENTITLED Instability at Trinucleotide Repeat DNAs BE ACCEPTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of Science.

______

Michael Leffak, Ph.D. Thesis Director

______

Madhavi Kadakia, Ph.D. Committee on Final Examination Chair, Department of and Molecular Biology ______Michael Leffak, Ph.D.

______John Paietta, Ph.D.

______Michael Markey, Ph.D.

______Robert E. W. Fyffe, Ph.D. Vice President for Research and Dean of the Graduate School

ABSTRACT

Gadgil, Rujuta Yashodhan. M.S. Department of Biochemistry and Molecular Biology, Wright State University, 2016. Instability at Trinucleotide repeat DNAs.

Trinucleotide repeats (TNRs) are sequences prone to formation of non-B DNA

structures and ; undergo expansions in vivo to cause various inherited

neurodegenerative diseases. Hairpin structures formed during DNA replication or

repair can cause replication fork stalling and if left unrepaired could cause single or

double strand DNA breaks. To test and study this hypothesis we have devised a novel

two color marker gene assay to detect DNA breaks at TNRs. By inducing replication

stress our results show that TNRs are prone to DNA strand breaks and it is dependent

on the repeat tract length. Double strand breaks at structured DNA are repaired

differently than ‘clean’ DSBs. The cells which undergo breaks die off, possibly due to

inability to repair breaks. Translesion polymerases help tolerate DNA damage at TNR

region.

iii

Table of Contents

INTRODUCTION...... 1

Trinucleotide repeats and related disorders...... 3

Importance of number of repeats...... 5

Expansions or contractions...... 8

Translesion polymerases...... 10

MATERIALS AND METHODS...... 13

Cell culture...... 13

DF2 Myc (CTG/CAG)100/23...... 15

DF2 Myc...... 15

DF2...... 17

PCR...... 17

Hydroxyurea, aphidicolin and telomestatin treatment assays...... 18

I-SCE-I transfection...... 18

Flow cytometry...... 19

Flow cytometry to detect cell cycle phase...... 19

siRNA treatment and flow cytometry...... 19

iv

siRNA treatment and western blot...... 20

RESULTS...... 21

I. Experimental DF2 Myc CTG cell line repeat length...... 21

II. Analysis patterns for flow cytometry...... 21

III. CTG100 TNRs are unstable due to hydroxyurea replication stress...... 22

IV. Recovery is required for separation of markers...... 27

V. Hydroxyurea shows considerable effect as a replication stress inducer..32

VI. Chk1 is activated at basal levels in CTG100 cells after HU treatment...... 32

VII. Cells display breaks before has occurred...... 35

VIII. Loss of eGFP+, RFP- cells after stress induced double strand breaks...... 37

IX. eGFP+, RFP- cells are unstable...... 39

X. Short TNRs are stable under replication stress...... 40

XI. DF2 cells are stable under replication stress...... 40

XII. DF2 Myc cells are stable under hydroxyurea replication stress...... 43

XIII. CTG100 TNRs are unstable due to apidicolin replication stress...... 46

XIV. CTG100 TNRs are unstable due to oxidative stress...... 46

XV. DSBs at structured DNA are treated differently than ‘clean’ DSBs…..48

XVI. DF2-TTR cells are unstable due to Telomestatin and HU………….…53

XVII. DF2-TTF cells are stable in Telomestatin and HU……………………54

XVIII. CTG 100 TNRs are unstable after Telomestatin treatment……………..57

XIX. DF2 Myc cells are stable under replication stress…………………….59

XX. Loss of translesion polymerases in DF2 Myc CTG100 cells affects DNA

damage tolerance...... 61

v

XXI. Loss of translesion polymerases in DF2 Myc cells does not affect DNA

damage tolerance...... 68

XXII. Knockdown of translesion polymerases in DF2 Myc CTG100 cells...... 74

DISCUSSION...... 78

Trinucleotide repeats as fragile sites...... 78

Trinucleotide repeat length is critical...... 80

Unrepaired DNA breaks cause cell death or ...... 81

Structure forming sequences influence repair mechanisms at broken

ends of DNA...... 82

Translesion polymerases help overcome DNA damage and DNA breaks

at trinucleotide repeat ...... 83

Translesion polymerases bypass secondary structures at TNR region...... 86

FUTURE DIRECTIONS...... 88

REFERENCES...... 89

vi

List of Figures

Figure 1. Microsatellite non-B DNA structures...... 6

Figure 2. Microsatellite instability related human diseases...... 7

Figure 3. Model showing replication fork arrest due to formation of hairpin structure ...... 9

Figure 4. Chromosomal integration of c-myc origin construct...... 14

Figure 5. Cell line constructs...... 16

Figure 6. CTG repeat length in engineered cell lines...... 24

Figure 7. Analysis pattern for flow cytometry...... 25

Figure 8. CTG100 TNRs are unstable due to HU replication stress...... 26

Figure 9. Recovery is required for separation of markers...... 28

Figure 10. Recovery is required for separation of markers...... 29

Figure 11. Recovery is required for separation of markers...... 30

Figure 12. Recovery is required for separation of markers...... 31

Figure 13. Hydroxyurea shows considerable effect as a replication stress inducer...33

Figure 14. Chk1 is activated at basal levels in CTG100 cells after HU treatment.....34

vii Figure 15. Cells undergo DNA breakage before they go through mitosis...... 36

Figure 16. Loss of eGFP+, RFP- cells after stress induced double strand breaks....38

Figure 17. GFP positive cells are unstable...... 41

Figure 18. Short TNRs are not unstable under replication stress...... 42

Figure 19. DF2 cells are stable under HU replication stress...... 44

Figure 20. DF2 Myc cells are stable under hydroxyurea replication stress...... 45

Figure 21. CTG100 TNRs are unstable due to aphidicolin replication stress...... 47

Figure 22. CTG100 TNRs are unstable due to oxidative stress...... 50

Figure 23. Cartoon of single-sided and double-sided double strand breaks...... 51

Figure 24. DSBs at structured DNA are treated differently than clean DSBs...... 52

Figure 25. DF2-TTR cells are unstable due to Telomestatin and HU...... 55

Figure 26. DF2-TTF cells are stable in Telomestatin and HU...... 56

Figure 27. DF2 CTG100 cells are unstable due to Telomestatin and HU...... 58

Figure 28. DF2 Myc cells are stable under replication stress...... 60

Figure 29. Loss of Rad18 in DF2 Myc CTG100 cells affects DNA damage tolerance...... 64

Figure 30. Loss of Pol eta in DF2 Myc CTG100 cells affects DNA damage tolerance...... 65

viii Figure 31. Loss of Pol kappa in DF2 Myc CTG100 cells affects DNA damage tolerance...... 66

Figure 32. Loss of Rev1 in DF2 Myc CTG100 cells affects DNA damage tolerance...... 67

Figure 33. Loss of Rad18 in DF2 Myc cells does not affect DNA damage tolerance...... 70

Figure 34. Loss of Pol eta in DF2 Myc cells does not affect DNA damage tolerance...... 71

Figure 35. Loss of Pol kappa in DF2 Myc cells does not affect DNA damage tolerance...... 72

Figure 36. Loss of Rev1 in DF2 Myc cells does not affect DNA damage tolerance...... 73

Figure 37. Knockdown of Pol eta using siRNA in DF2 Myc CTG100 cells...... 75

Figure 38. Knockdown of Pol kappa using siRNA in DF2 Myc CTG100 cells...... 76

Figure 39. Knockdown of Rev1 using siRNA in DF2 Myc CTG100 cells...... 77

ix

ACKNOWLEDGEMENTS

I would like to take this opportunity to thank my mentor, Dr Michael Leffak who has

been guiding, teaching and supporting me throughout. He has been encouraging as

well as appreciative. Not only did he appreciate the good work done but was also equally receptive if things did not work out through the course of the study. The excitement he would have while reporting data at the weekly meetings would surely encourage me to work harder and with enthusiasm. I would be forever grateful to him for giving me this opportunity to learn from him. I would also like to thank my committee members Dr John Paietta and Dr Michael Markey for agreeing to be on my committee. Their suggestions, inputs and help have been invaluable. I would also like to thank Dr Madhavi Kadakia for being supportive. Working in the lab you spend most of the day with your lab members and so here I would like to thank present and past members of the Leffak lab; Todd Lewis, Dr Eric Romer, Dr Joanna Barthelemy,

Sumeet Poudel, Caitlin Castagno, Tu Dahn, Dr David Hitch and French Damewood

IV. A special thanks to Todd Lewis who helped me to learn everything in the lab and has always been helpful.

Coming to US for further studies is a big decision and supporting me through this whole journey are my parents. They have been standing firm by me in my good and bad days. They have been there to appreciate my work as well as the difficult days. I could never have done this without you both. Along with them my whole family- the

Kanetkars’, Godbole and Phatak and my grandparents have always been supportive

and loving. I would also like to thank all my friends in India as well as those here.

x

You people made this a fun and crazy journey and life without you all would have

been dull. A special thanks to my professors- Mrs Meghana Kulkarni, Dr Suneeti

Gore, Mrs Saylee Kalekar and all others who helped me all through my undergraduate and post graduate time. They laid the foundation for my understanding of the subject and were supportive and caring always.

xi

Dedicated to my loving parents

Mrs Rashmi Gadgil

Mr Yashodhan Gadgil

For always supporting me and standing by me

You mean the world to me

xii

INTRODUCTION

Faithful replication should take place during cell proliferation to maintain genomic stability and prevent formation and accumulation of mutations which jeopardizes this integrity. However, during DNA replication the replication fork frequently encounters fork stalling agents. These could be intrinsic factors like sequences that form non - classical DNA structures, transcription bubbles, tightly bound histone and other DNA binding proteins or extrinsic factors like UV light, gamma radiation, dNTP pool depleting agents, etc. (Mirkin et al. 2007). If these replication blocks are present for a short duration it causes fork stalling. Prolonged blocks lead to fork collapse. Structurally, this means that one or more replisome components may detach from the replication fork; functionally, a collapsed fork cannot restart once replication stress is removed. Whereas a stalled fork can restart replication once the blocking agent is removed, a collapsed fork is completely deactivated and this proves to be an even greater damage since now the fork can be processed into a double strand break

(Petermann, E. et al. 2010). However the cell undertakes several different mechanisms in order to overcome the DNA damage and continue unperturbed DNA replication with the ultimate goal of cell survival and genomic stability. Some of the mechanisms include DNA repair proteins, fork restart, fork remodelling processes and checkpoint activation. The cell can also resort to homologous recombination, however it may come at the cost of chromosomal rearrangement (Lambert et al. 2005). Some of these mechanisms are discussed below. Stalled forks are protected by the homologous recombination protein Rad51 which helps to stabilize the fork. Here 1

Rad51 functions in a homologous recombination independent manner. It does so by replacing RPA bound to single stranded DNA (ssDNA) at the stalled fork, and initiates the process of fork reversal, explained below (Petermann, E., Orta, M. L et al. 2010). It also protects ssDNA from degradation by Mre11 (Hashimoto Y et al.

2010).

Higgins et al. in 1976 proposed the idea of reversed forks as a mechanism of overcoming DNA damage during replication. Fork reversal is described as the reannealing of parental strands with consequent annealing of nascent strands. This creates a four-way junction also commonly called as ‘chicken foot structure’. This was shown to be a common phenomenon against DNA damage by topoisomerase I

(Top1) inhibitors in yeast cells, mouse and human cells and Xenopus egg extract (Ray

Chaudhuri et al. 2012). Poly (ADP-ribose) polymerase (PARP) and RECQ1 were shown to play a central role in this process. Eventually it was shown that fork reversal is a universal response to any tested source of replication stress (Zellweger et al.

2015).

As mentioned above, prolonged replication stress can turn a stalled fork into a collapsed fork. Loss of the replisome components exposes the ssDNA which can be acted upon by certain proteins to process it into a double strand break. Specialized nucleases like MUS81 endonucelease cause double strand breaks at the stalled or reversed fork (Regairaz M et al. 2011). Such a break is repaired by homology directed recombination. This is where the Rad51 nucleoprotein filament dependent processing of a double strand break occurs.

Trinucleotide repeats are prone to formation of non-classical DNA structures and mutations (Gacy AM et al. 1995). To date, these mutations have been shown to be in 2 the form of trinucleotide repeat expansions or contractions and generally occur during

DNA replication or repair (Liu G et al. 2010, Chatterjee N et al. 2015). This is because during these processes a single strand of DNA is exposed which increases the chances of forming a DNA secondary structure. Figure 1 shows representative images of various possible non-classical DNA structures. Expansions of these trinucleotide repeats are the basis of various inherited genetic diseases. Thus it is a known fact that trinucleotide repeats are unstable regions in the . However the present study puts forth the idea that in addition to length instability of trinucleotide repeat region, the presence of these repeats can also lead to chromosome breakage. The hypothesis is built on the idea that since trinucleotide repeats have the potential to form secondary structures it will lead to stalling of the replication fork. If this secondary structure is left unrepaired it would lead to fork collapse and DNA strand breaks- single strand/double strand breaks (DSBs).

Trinucleotide repeats and related disorders

Trinucleotide repeat expansions are known to cause at least 17 neurological disorders

(Lopez Castel et al., (2010), McMurray et al., (2010), Mirkin et al., (2007)).

CAG/CTG, CCG/CGG, GAA/TTC expand in vivo to cause inherited genetic diseases

(Cummings, C.J et al. 2000).Depending on where these expansions occur in the genome there are different downstream effects. In reference to Figure 2 if expansions occur in the exon it can shift the reading frame. If expansions occur in the it can cause improper splicing and lead to formation of a faulty transcript. Expansions in exons can also lead to protein over-expression (Cummings, C.J et al. 2000).

Secondary structures formed due to CAG or GAA can cause RNA polymerase II to arrest and thus cause transcriptional instability. This initiates transcription coupled

3

DNA repair (Kim N et al. 2012, Hanawalt et al. 2008) and potential breakage in the displaced strand of an R-loop. Most of the disorders are due to expansions of a characteristic sequence. CAG/CTG expansions are a cause of Huntington’s disease, myotonic dystrophy type 1 and several forms of spinocerebellar ataxia. is caused by expansions of CGG repeats. GAA expansions are responsible for Friedreich’s ataxia (Mitas M. 1997). While the diseases are not lethal in immediate effects, the life span of patients suffering from them is 10-20 years after the symptoms first appear. Also their quality of life is affected once the symptoms start appearing. Among all trinucleotide repeats, expansions of CAG/CTG are most common.

CAG/CTG, CGG/CCG sequences are shown to primarily form hairpin structures at the trinucleotide repeats (Lahue RS et al. 2003). CGG repeats can form both hairpin and quadruplex structures which could block replication in vitro and in vivo (Fry M et al. 1994, Samadashwily GM et al. 1997, Gacy AM et al. 1995, Usdin K et al.

1995). GAA repeats that cause Friedreich’s ataxia are known to form stable triplex structures (Rajeswari MR et al. 2012). The stability of these trinucleotide repeats is highly dependent on the length of the repeat (Gacy et al. 1995). Thus greater the length of the repeat, the greater is the stability of the secondary structures formed.

Formation of these secondary structures can be avoided by interruptions in the repeating sequences that will also help regulate disease severity (Jarem DA et al.

2010, Weisman-Shomer P et al. 2000).

4

Importance of number of repeats

All individuals carry a normal number of these repeats (usually n <30-40). However if these repeats fall above a threshold value then the individual starts showcasing symptoms and ultimately suffers from the disease (Fig. 2). In most individuals repeats in the range of 4-30 is considered normal and such individuals remain unaffected

(Benton CS et al. 1998). However beyond a threshold value of 30-50 the repeats can become pathogenic (Goula, A.V et al. 2013)

5

Mirkin, S. (2007) Expandable DNA repeats and human disease. Nature 447; 932-940.

Figure 1. Microsatellite non-B DNA structures. (A) Hairpin formation in one DNA strand.

(B) G quadruplex structure. (C) Hairpin formation in both strands of DNA. (D) H-DNA and

sticky DNA. (E) Unwound DNA.

6

Expandable repeats responsible for human disease

BPES DRPLA CCD HD CCHS SBMA HFG SCA1 HPE5 SCA3 ISSX FRAXA SCA6 MRGH DM1 FRAXE SCA7 OPMD HDL2 EPM1 FXTAS SCA12 SCA17 FRDA DM2 SCA31 SCA10 SPD SCA8

polyglutamine polyalanine

(C4GC4GCG4)n (CGG)n (CAG)n (CAG)n (GAA)n (CCTG)n (TGGAA)n (ATTCT)n (GCN)n (CTG)n

promoter 5’ UTR TR- intron TR- 3’ UTR exon exon

BPES, blepharophimosis, ptosis and epicanthus inversus HD, Huntington’s disease CCD, cleidocranial dysplasia HDL2, Huntington’s-disease-like 2 CCHS, congenital central hypoventilation syndrome HFG, hand–foot–genital syndrome DM, myotonic dystrophy HPE5, holoprosencephaly 5 DRPLA, dentatorubral–pallidoluysian atrophy ISSX, X-linked infantile spasm syndrome EPM1, progressive myoclonic epilepsy 1 MRGH, mental retardation with isolated growth hormone deficiency FRAXA, fragile X syndrome OPMD, oculopharyngeal muscular dystrophy FRAXE, fragile X mental retardation associated with FRAXE site SBMA, spinal and bulbar muscular atrophy FRDA, Friedreich’s ataxia SCA, spinocerebellar ataxia FXTAS, fragile X tremor and ataxia syndrome SPD, synpolydactyly

Figure 2. Microsatellite instability-related human diseases. The figure shows diseases caused by

microsatellite repeats, and location of repeats.

7

Expansion or contraction

Whether a trinucleotide repeat has a tendency to undergo expansions or contractions is based on whether these sequences that can form stable non B DNA structures are present in the nascent strand or template strand of DNA, respectively (Mirkin et al.

2007). If the hairpin forming sequence is in the nascent DNA strand it can lead to over replication and thus an expansion which can get fixed in the genome with the consecutive round of replication. If however the hairpin forming sequence is on the template strand it causes under replication and thus a contraction of the sequence.

CTG/CAG and CGG/CCG repeats within normal range have an inclination to contractions over expansions. Thus it shows that to maintain genetic integrity and stable, short normal alleles, contractions are important (Pelletier et al. 2005).

Polymerase slippage can be responsible for small changes in repeat number resulting in small expansions or contractions (Petruska et al.1998, Hartenstine et al. 2002).

However as the number of repeats increase, their ability to form hairpins increases as well as the stability of these hairpins which can escalate the problem of expansions

(Gacy AM et al. 1995). Environmental stress is also a contributing factor to contractions as well as expansion of these repeats (Chatterjee N et al 2015).

Expansion is associated only with structure forming repeat sequences which suggests that formation of secondary structures is essential for expansion (Gacy AM et al.

1995).

8

Figure 3. Model showing replication fork arrest due to formation of

hairpin structure. The cartoon here shows formation of a hairpin structure

that leads to stalling of replication fork.

9

Translesion polymerases

There are multiple DNA lesions -both intrinsic and extrinsic that can stall the replication fork, including abasic sites, covalent base adducts, and non-canonical

DNA structures. In this situation the usual replicative polymerases that function with high accuracy cannot accommodate unpaired or atypical base pairs in their sterically constrained active sites. This condition will lead to base substitutions, frame shift mutations or sequence loss. Incomplete replication can lead to loss of part of the chromosome and cells with broken chromosomes may not survive. In this case DNA damage is overcome by two pathways – postreplicative DNA damage repair (PRR) or

DNA damage tolerance. Under PRR, the damage to DNA is repaired by means of nucleotide excision repair, base excision repair, mismatch repair, etc. It also includes single and double strand break repair mechanisms like non-homologous end joining

(NHEJ), homologous recombination (HR) and single strand annealing (SSA). Cells also have mechanisms to tolerate DNA damage until DNA repair pathways can act and remove the lesion. The purpose of the damage tolerance pathway is to allow some form of DNA replication across the lesion which can be repaired subsequently. Two mechanisms are involved as part of the tolerance pathway. They include template switching and translesion synthesis. Template switching or fork reversal as mentioned before is one mechanism to overcome DNA damage. Translesion synthesis is described in detail below.

Translesion synthesis (TLS) is a mechanism by which a DNA lesion is bypassed by filling in a nucleotide opposite the lesion (Friedberg E. et al. 2005). This requires special polymerases which can accommodate damaged DNA templates in their active 10 sites and insert nucleotides, most frequently , against a lesion, as opposed to high fidelity polymerases which work to insert required nucleotides accurately and thus cannot function at DNA lesions (Friedberg E. et al. 2005, Goodman M. F et al.

2002, Prakash S et al. 2005).. The polymerases are thus called translesion polymerases. TLS polymerases belong to the Y family of polymerases which include

Rev 1, Pol kappa (κ), Pol eta (η), Pol iota (ι) (Ohmori H et al. 2001). The ability of replicating through DNA lesions however comes at a cost of lower replication fidelity

(Goodman M. F et al. 2002).The process of TLS also requires another initiating factor called Rad18. Rad18 is an E3-ubiquitin ligase that helps monoubiquitinate

PCNA which initiates the TLS pathway by promoting polymerase switching. The process of TLS, although it promotes cell survival, is at least temporarily mutagenic since it involves adding a noncognate nucleotide against the DNA lesion.

Rad18 is an E3 ubiquitin ligase that plays a primary role to monoubiquitinate PCNA that triggers the TLS pathway (Watanabe et al., 2004; Ulrich, 2009). In response to stress lysine residue 164(K164) of PCNA is covalently bound to ubiquitin. DNA lesions cause accumulation of RPA on single-stranded DNA which also signals ubiquitination of PCNA by Rad18.This monoubiquitinated form of PCNA can be bound by Y family DNA polymerases that have ubiquitin binding domains. This initiates the process of translesion synthesis (Bienko et al. 2005).Loss of Rad18 increases rates in cells and thus shows importance of the protein for maintaining genomic integrity of the cell (Tateishi et al. 2003). However overexpression of Rad18 is also deleterious since it can lead to unwanted

11 ubiquitination of PCNA and thus trigger TLS polymerase switching even in absence of any DNA damage.

Pol eta (η) is found only in eukaryotes and is evolutionarily conserved. It acts with high efficiency and accuracy in vitro. Boyer et al. in 2013 showed that Pol eta is required for common fragile site (CFS) stability. In vivo Pol eta is responsible for restarting replication forks that are stalled due to UV damage (Lehmann et al. 2005).

It has also been shown that depletion of Pol eta by siRNAs increases mutation frequency (Choi et al. 2005). Thus this information shows that Pol eta helps to control mutagenesis. Loss of Pol eta causes Xeroderma Pigmentosum Variant (XPV) which is characterized by increased probability of skin cancer and sensitivity to sunlight (Masutani et al. 1999).

Pol kappa(κ) is comparatively more accurate than other TLS polymerases, however it has limited ability to synthesize across several DNA lesions (Bavoux et al. 2005). As opposed to Rev1 ; Pol kappa deletion does not show a profound effect on the mutagenesis caused spontaneously or induced by damage. However overexpression of

Pol kappa leads to inhibition of replication fork progression and can also cause DNA strand breaks (Bavoux et al. 2005; Leopoldino et al. 2005).

Rev1 functions to resolve G quadruplexes. However contrasting reports also suggest that Rev1 induces spontaneous mutations (Gibbs et al. 2000). It has been shown to have a rather unexpected role in preventing hairpin forming trinucleotide repeat expansions (Collins et al. 2007).

12

MATERIALS AND METHODS

Cell culture

Cell lines were made using a HeLa acceptor cell line which has a single FRT (FLP recombinase target) site. A expressing FLP recombinase and a donor plasmid containing CTG/CTG repeats along with c-myc replicator were co-transfected in the acceptor cells (Figure 4). The repeats are placed right next to the c-myc replicator so that we understand the polarity of replication. The CTG repeats are specifically placed on the template for lagging strand synthesis. Thus far we have two different cell lines harbouring CTG in the template for lagging strand synthesis- DF2 Myc (CTG/CAG)

100 and DF2 Myc (CTG/CAG) 23. The threshold frequency of number of trinucleotide repeats beyond which diseases can be caused is between 30 – 40 (Goula,A.V et al.

2013). With this information in hand we incorporated 100 repeats of CTG/CAG since this would fall well beyond the threshold range. Similarly 23 repeats were incorporated so that they work as controls falling below the minimal length of repeats known to cause diseases. The cell lines have been developed by Todd Lewis (Wright

State University). All cell lines were grown and maintained in Dulbeco’s Modified

Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum and 5% CO2.

13

Hygr FRT TK FRT Neor

acceptor cell line

c-myc

Donor plasmid FLP recombinase

FRT Hygr FRT Neor TK

c-myc

Figure 4. Chromosomal integration of c-myc origin constructs. Diagram showing

acceptor cell line with a single FRT site and donor plasmid with c-myc replicator. The CTG/

CAG repeats are right next to c-myc replicator and form basis of DF2 Myc (CTG/CAG)cell

lines.

14

DF2 Myc (CTG/CAG) 23, 100

DF2 denotes dual . The DF2 cell line has two color marker genes red

(Tomato, red fluorescent protein, RFP) and green (eGFP) whose presence can be analyzed by flow cytometry. The ectopic site has three identical Alu sequences – Alu

1, Alu 2, Alu 3 which are known hotspots for recombination. The c-myc region follows Alu 2 and is an origin of replication. CTG/CAG repeats are placed adjacent to the c-myc replicator. Tomato is placed in between Alu 1 and Alu 2 while eGFP is placed in between Alu 2 and Alu 3 (Figure 5). As mentioned before the trinucleotide region is placed besides the c-myc region. So far there are two cell lines one with 100 repeats of CTG/CAG and another with 23 repeats of CTG. The idea is that if the DNA undergoes breaks at the trinucleotide repeat region it would lead to separation of the colour marker genes. Depending on the repair mechanism and recombination patterns undertaken by the cell we can observe different flow cytometry profiles.

DF2 Myc

This is a dual fluorescence cell line too. It has three Alu sequences - Alu 1, Alu 2, Alu

3. The c-myc region follows Alu 2. Tomato is placed between Alu 1 and Alu 2 while

GFP is between Alu 2 and Alu 3. However this cell line does not have a trinucleotide repeat region. It thus acts as a control cell line for our assays.

15

A

B

23

23

C

IVS1 IVS6 IVS6 Hyg Neo Alu Alu Alu TK

DF2 myc only Tomato c-myc origin eGFP FRT I-Sc e1 FRT

D

Figure 5. Cell line constructs. (A) Map of DF2 Myc (CTG/CAG)100 cell line.

(B) Map of DF2 Myc (CTG/CAG) 23 cell line.(C) Map of DF2 Myc cell line.

(D) Map of DF2 cell line.

16

DF2

This is also a dual fluorescence cell line. It has the three Alu sequences - Alu 1, Alu 2,

Alu 3. Tomato is in between Alu 1 and Alu 2, GFP is in between Alu 2 and Alu 3.

However this cell line does not have both c-myc as well as the trinucleotide repeat region. Thus this also acts as a control cell line.

PCR

After constructing the cell lines it was imperative to confirm the number of

CTG/CAG repeats. DNA was isolated from DF2 Myc (CTG/CAG) 100and DF2 Myc

(CTG/CAG) 23 cells using E.Z.N.A. ® Tissue DNA Kit from Omega Bio-Tek

(D3396). DNA template used per reaction was 200 ng. Four PCR reactions were performed for each DNA sample. All PCR reactions were run using HotStar Taq polymerase and at following conditions- 95° (15 min), 35 cycles of 94°(30 sec),

55°(30 sec), 72°(45 sec) and final extension 72°(5 min). The sequence of the forward primer used was (5’-AAAAAAGATCCTCTCTCGCTAATCTCC-3’) and reverse primer (5’-GAACTTCCCTATTCTTGGTTTGG-3’). After completion of PCR, the reactions were pooled together and products resolved using 1.5% agarose gels. Gels were imaged using Fuji LAS-3000 and the software ImageReader. The desired band was cut and DNA extracted using E.Z.N.A. ® Gel Extraction Kit from Omega Bio-

Tek (D2500-02). The samples were sent for sequencing to Retrogen.

17

Hydroxyurea, aphidicolin and telomestatin treatment assays

Various replication stress inducers were used in order to induce formation of DNA secondary structures. The reagents were added to the medium and after treatment cells were allowed to recover in normal media. Hydroxyurea concentrations of 0.1-0.25 mM and 2mM were used for a dose response assay and based on this result 0.2 mM concentration for all remaining assays was used. Hydroxyurea is an inhibitor of ribonucleotide reductase and leads to decreased formation of dNTPs. In order to check if results obtained for hydroxyurea treatment are drug specific or mechanism specific, another DNA polymerase inhibitor; aphidicolin was used at a concentration of 0.2 µM. In presence of these reagents replication will slow down and thus give an opportunity for formation of DNA secondary structures. Finally telomestatin (TMS) was used in order to see its effect if any. Telomestatin is a telomerase inhibitor, responsible for stabilizing G quadruplexes (Wu, Y et al 2010). 0.5 µM of telomestatin was used in all TMS assays.

I-SCE-1 transfection

In order to understand repair patterns for cells experiencing double strand breaks, DF2

Myc CTG100 cells were treated with I-SCE-1. I-SCE-1 is a restriction enzyme that does not cleave human genomic DNA unless an exogenous I-SCE1 site is integrated into the genome. I-SCE1 cutting leads to double strand breaks. For this DF2 Myc

CTG100 cells were plated on day 0. 24 hrs later cells are transfected with plasmid expressing I-SCE-1using lipofectamine and OPTI-MEM. During transfection media without antibiotic is used. Eight days after transfection cells were harvested and analyzed by flow cytometry. 18

Flow cytometry

After treatment of cells with any of the reagents inducing replication stress they were allowed to recover in normal media and then harvested to analyze their flow cytometry profiles or were harvested for flow cytometry without a recovery period.

Harvested cells were aliquoted in 1.5 ml Eppendorf tubes and centrifuged at 300g for

3 min. Media was aspirated and cells washed with cold PBS and cells spun down at

300g for 3 min. After a final wash with PBS, cells were analyzed using a C-Flow ®

Plus Accuri Cytometers.

Flow cytometry to detect cell cycle phase

After treatment of cells with replication stress inducing agent and required recovery, cells were harvested 1.5 ml Eppendorf tubes and centrifuged at 500g for 3 min and processed to check their flow cytometry profiles. Media was aspirated and cells washed with cold PBS and cells spun down at 500g for 3 min. Cells were then treated with 70% ethanol and permeabilized at -20°C for 20 mins or overnight. Cells were then centrifuged at 1500 rpm for 5 min and ethanol aspirated. This was followed by washing cells with 1 ml PBS and adding 0.75 ul of 100mg/ml RNAse and incubating at 37°C for 20 mins . Finally DRAQ7 dye was added at a final concentration of 7.5 uM and incubated in the dark for 25 mins after which cells were analyzed using C-

Flow ® Plus Accuri Cytometers. siRNA treatment and flow cytometry

The siRNAs used to knockdown translesion synthesis polymerases were generously provided by Dr Yangzhe Gao, University of North Carolina, Chapel Hill. Depending

19 on the assay DF2 Myc CTG100 or DF2 Myc cells were plated on day 1 in DMEM medium with 10% FBS and antibiotics. 24 hrs later cells were treated with siControl and siRad18/ siPol eta/ siRev1/ siPol kappa. Transfection was done as follows. In a

1.5 ml eppendorf tube 250 µl OPTIMEM and 10µl Lipofectamine2000 was mixed and incubated for 5 min. In another tube 250 µl OPTIMEM was mixed with 10 µl of the respective siRNA. After 5 min the Lipofectamine2000 mix was added to the siRNA mix and incubated for 20 min. In the meantime cells were trypsinized and plated again in DMEM media with 20% FBS but without antibiotics. After 20 min. incubation the siRNA mix was added to the cells and incubated at 37°C. 24 hrs later cells were treated with 0.2 mM hydroxyurea for 4 days after which they are allowed to recover for 4 days and then analyzed by flow cytometry. siRNA treatment and western blot

The protocol mentioned above for siRNA treatment was followed with the respective cell lines and siRNAs until transfection with the siRNA. Cells were harvested 24 hrs,

48 hrs and 96 hrs after transfection to check for protein knockdown by western blot.

20

RESULTS

I. Experimental DF2 Myc CTG cell line repeat length

Once the DF2 Myc CTG cell line was made it was essential to confirm the number of repeats incorporated. In order to do so it was decided to run a PCR using primers which flank the repeat region. The two primers used were Blue-F (forward) and Sffv-

R (reverse) (Figure 6). A no template control was also run along with the samples of interest. After PCR, samples were separated on a 1.5% agarose gel. Separation of

DNA revealed a 620 bp band in the Myc CTG100 cell and a 320 bp band in the Myc

CTG23 cell line. These bands were excised and sent for sequencing. Sequencing results confirmed number of repeats present to be 100 and 23 respectively (Figure 6).

The size of the band on the gel is not exactly the same size as the TNR region due to presence of flanking sequences which add to the actual size of the band. Thus we had two experimental cell lines with differing length of repeats as follows DF2 Myc

CTG100 and DF2 Myc CTG23.

II. Analysis patterns for flow cytometry

If there are breaks in the CTG region which are repaired using different repair mechanisms the dual florescence will help detect them by flow cytometry. Based on whether the cell is able to retain both the fluorescent proteins or loses either or both of

21 its fluorescent proteins they will be found in particular quadrants of a typical flow cytometry profile as shown in Figure 7. If the break is repaired by non homologous end joining or homologous recombination then both the colours- red and green would remain intact and the cell will thus be able to express both proteins. Such a cell would be given a double positive attribute and such cells are detected in the top right quadrant of a flow cytometry profile. In case of recombination between two of the 3

Alu recombination sites, three different outcomes are possible. Recombination between Alu 1 and Alu 2 would lead to loss of Tomato or red color. Thus the cell will have only eGFP and obtain an attribute of a green cell. Another mechanism by which a cell can retain eGFP and thus get represented as a green cell is when DNA breaks in the TNR region and repair cannot occur, such that Tomato, and its neighbouring chromosomal DNA, is lost. These cells will be detected in the bottom right quadrant of a flow cytometry profile. Similarly if recombination is between Alu 2 and Alu 3, the cell would lose eGFP or green and retain Tomato or red. As a result the cell will have an attribute of a red cell and detected in the top left quadrant of the flow cytometry profile. If however recombination is between Alu 1 and Alu 3 the cell would lose both Tomato or red and eGFP or green. Thus this cell will have a double negative attribute and will be detected in the bottom left quadrant. It is necessary to note that once a cell loses either or both of its colors there is no means to acquire that colour again.

III. CTG 100 TNRs are unstable due to hydroxyurea replication stress

After understanding the importance of recovery on treatment of cells with replication stress inducers, an assay to treat cells with hydroxyurea and see its effect on DF2 Myc 22

CTG100 cells after allowing them to recover was performed. From Figure 8 it was seen that for control, 86% cells were double positive and 10% cells were eGFP positive.

However treatment with 0.2 mM hydroxyurea caused only 60% cells to be double positive while a significant number i.e. 35% cells were only eGFP positive. Similar results were obtained on repeating the experiment several times. Thus the assay concluded that under replication stress CTG100 repeats are susceptible to breakage.

The results for this assay are shown in Figure 8. From the flow profile for control

CTG 100cells it can be seen that 10.5% cells are already eGFP positive. This is because the 100 repeats of CTG cause some spontaneous instability in the cells supporting the conclusion that the CTG/CAG repeat causes DSBs under normal growth conditions

23

NT CTG 100 CTG 23

Number of Repeats

700 bp 600 bp 100 500 bp 400 bp 300 bp 23

Blue (forward) Sffv (reverse)

Figure 6. CTG repeat length in engineered cell lines. After making DF2

Myc CTG cell line, a PCR was run across the repeat region using primers

Blue-F (forward), Sffv-R (reverse). PCR samples were run on a gel and the

samples sent for sequencing. Sequencing results confirmed presence of 100

and 23 repeats of CTG. Two working cell lines were thus obtained DF2 Myc

CTG100 and DF2 Myc CTG23.

24

I-Sce1, replication stress

Alu Alu RED! RED, ! GREEN!

Nonhomologous Homologous Recombination (HR) End Joining (NHEJ) NEGATIVE,! Alu Alu Alu Alu GREEN! NEGATIVE!

Alu

Figure 7. Analysis patterns for flow cytometry. Based on different ways of

recombination the cell will be able to retain both red and green color or retain only

one of the colors or lose both the colors. This would result in four different patterns

visible on a flow cytometry profile where each quadrant depicts presence of red or

green protein or absence of both.

25

(CTG) 100 control (CTG) 100 HU treatment

B A

0.1% 86.6% RED R,G 0.0% RED R,G 60.1%

N,N GREEN N,N GREEN

2.8% 10.5% 4.2% 35.7%

Figure 8. CTG 100 TNRs are unstable due to HU replication stress. DF2 Myc CTG100 cells were (A)

untreated, (B) treated with HU.

26

IV. Recovery is required for separation of markers

It is predicted that the separation of colour markers will require mitosis and degradation of pre-existing Tomato and eGFP proteins. This assay was performed by a former lab student, Todd Lewis (Wright State University) on one of the control cell lines DF2 and treating them with I-SCE-1 which is a restriction enzyme and would make a double strand break. Results for this assay are shown in Figure 9 and 10. It can be seen that proper separation of colours is seen starting day 5 (Panel F) after I-

SCE-1 transfection. This showed that a recovery period of 4 days is enough so that the existing proteins degrade and we see roper separation of colors. Based on this assay for all our assays ahead we decided to allow 4 days of recovery.

In order to observe the effect of a no recovery phase on cells after treatment with hydroxyurea or aphidicolin, DF2 Myc CTG100 cells were treated with the replication stress inducers and analyzed by flow cytometry without allowing them to recover. A control without treatment with either hydroxyurea or aphidicolin was also run. The flow cytometry results were as shown in Figure 11. While 86% cells were double positive around 10% cells were eGFP positive. Even after 10 days of hydroxyurea treatment 94% cells remained double positive with only 2% cells remaining eGFP positive. Similarly aphidicolin treatment left about 88% cells double positive and 8% cells were eGFP positive. Thus these results show that in order to allow separation of colours the cells should be allowed to recover. Phenotypically however the treated cells in the absence of a recovery phase were ‘fat’ and it was reflected on the side scatter and forward scatter plot. Figure 12 graphically represents the data shown in

Figure 11. All values are calculated for percentage of double positive cells. 27

A B C Control I-SCE 1- 1 DAY I-SCE 1-2 DAYS

98.8% RED R,G 98.6% RED R,G RED R,G 98.5% 0.0% 0.0% 0.1%

N,N GREEN N,N GREEN N,N GREEN

0.0% 1.4% 0..% 1.1% 1.3% 0.0%

D I-SCE 1-3 DAYS E I-SCE 1-4 DAYS

98.1% RED R,G RED R,G 70.5% 21.2% 7.9%

N,N GREEN N,N GREEN

1.2% 6.0% 2.8% 2.2%

Figure 9 Recovery is required for separation of markers. (A) Flow cytometry profile

of control or untreated DF2 cells. (B) Flow cytometry profile of DF2 cells treated with I-

SCE-1 for 1 day (C) Flow cytometry profile of DF2 cells treated with I-SCE-1 for 2

days (D) Flow cytometry profile of DF2 cells treated with I-SCE-1 for 3 days (E) Flow

cytometry profile of DF2 cells treated with I-SCE-1 for 4 days

28

F I-SCE 1- 5 DAYS G I-SCE 1- 6 DAYS

RED R,G 65.9% 68.8% RED R,G 25.9% 25.4%

N,N GREEN N,N GREEN 2.4% 6.1% 2.2 % 6.3%

H I-SCE 1-7 DAYS I I-SCE 1-8 DAYS

66.9% RED R,G 66.7% RED R,G 25.6% 25.2%

N,N GREEN N,N GREEN

2.0% 5.6% 2.0% 5.8%

Figure 10 Recovery is required for separation of markers. (F) Flow cytometry

profile of DF2 cells treated with I-SCE-1 for 5 days. (G) Flow cytometry profile

of DF2 cells treated with I-SCE-1 for 6 days (H) Flow cytometry profile of DF2

cells treated with I-SCE-1 for 7 days (I) Flow cytometry profile of DF2 cells

treated with I-SCE-1 for 8 days

29

CTG100 control CTG100 HU treatment CTG100 APH treatment

RED R,G 94.5% RED R,G RED R,G 88.6% 86.3% 0.0% 0.1% 0.1%

N,N GREEN N,N GREEN N,N GREEN

9.9 % 2.3 % 3.7% 3.1% 3.0% 8.4 %

Figure 11. Recovery is required for separation of markers. DF2 Myc CTG100 cells were (A) untreated, (B) treated

with HU, (C) treated with APH.

30

120%

100%

Percentage of green 80% cells 60%

40% Double Positive 20%

Normalized 0% to control

CTG-100-CTRL

CTG-100-HU-2D-NOCTG-100-HU-4D-NO REC CTG-100-HU-6D-NO REC CTG-100-HU-8D-NO REC REC

Figure 12. Recovery is required for separation of markers. The graph is a

representation of the percentage of double positive cells which are normalized to

control.

31

V. Hydroxyurea shows considerable effect as a replication stress inducer

In order to understand a particular concentration of hydroxyurea for which a considerable percentage of cells could turn green, a dose response assay was performed. Results for this assay are shown in Figure 13. The percentage of eGFP positive cells in a control sample, which is not treated with hydroxyurea, was calculated. The percentage of eGFP positive cells for different concentrations of hydroxyurea treatment has been normalized to above mentioned control values. Thus the graph represents only the population of eGFP positive cells under different conditions. The assay showed that a concentration of 0.2 mM hydroxyurea works best to show a substantial percentage of eGFP positive cells. Based on these results for all assays performed using hydroxyurea a working concentration of 0.2 mM is used.

VI. Chk1 is activated at basal levels in CTG 100 cells after HU treatment

It was seen that hydroxyurea causes a significant damage to DF2 Myc CTG 100cells acting as a replication stress inducer. This damage would thus mean that the checkpoint pathway should ideally get activated. In order to detect this activation of checkpoint pathway an assay was run by treating DF2 Myc CTG 100cells with 0.2 mM hydroxyurea but the cells were not allowed to recover. The rationale was that during a

4 day recovery, the cells may not find it necessary to have a checkpoint pathway active. A positive control sample included treating cells with 2 mM hydroxyurea overnight which has been shown to induce fork collapse. Negative control sample were untreated DF2 Myc CTG 100cells. For all conditions samples were immediately harvested after treatment. Protein estimation was done by BCA assay and samples

32

250%

Percentage of green 200% cells

150%

Green quadrant Normalized to 100% control

50%

0% NO HU 0.1mM HU 0.15mM HU 0.2mM HU

Figure 13. Hydroxyurea shows considerable effect as a replication stress inducer.

The graph shows that 0.2 mM concentration of hydroxyurea shows an effect to make

cells turn green meaning there are possible breaks in the TNR region leading to

resection such that Tomato is lost or there is recombination between Alu 1 and Alu 2.

This results in cells retaining only eGFP and thus cells are green. The values are

normalized to control.

33

HU treatment (positive control) hrs 6 days HU8 days HU10 days HU 24 No treatment2 days (negative HU4 days control) HU

pChk1 (56 kd) actin (40 kd)

Figure 14. Chk1 is activated at basal levels in CTG 100 cells after HU treatment. Activation

of checkpoint pathway is determined by phospho chk1 band at 56 kd. This band is seen in

positive control and the samples but not in negative control.

34 were run on a 10% SDS gel overnight and transferred using a PVDF membrane.

Primary antibody used was pChk1 S345 and subsequently detected using anti-rabbit

HRP.

From Figure 14 it is seen that Chk1 is activated at a basal level during all days of hydroxyurea treatment. The positive control (2 mM HU) shows a heavier band of activated p-Chk1. However it is also seen that even the negative control shows a weak pChk1 band. We think this represents the small eGFP positive population already present in the DF2 Myc CTG 100cells and have an activated checkpoint due to their unstable nature and the consequent DNA damage. The assay thus shows that Chk1 is only activated to a basal level when cells are treated with 0.2 mM hydroxyurea.

VII. Cells display breaks before mitosis has occurred

Treatment of DF2 Myc CTG100cells showed that DNA breaks are occuring as a result of replication stress. However the phase of cell cycle during which these breaks could be occurring was not known. In order to have an understanding about this we decided to treat DF2 Myc CTG 100cells with hydroxyurea for 4 days to cause ample opportunities to induce DNA breaks however cells were harvested at particular time intervals during the 4 days of recovery in order to gauge the cell cycle phase responsible for DNA breaks. These intervals were as follows- 0 hours recovery or essentially no recovery, 6 hours, 12, 24, 48, 72 and 96 hours. A no treatment sample was also included. Results for this data are shown in Figure 15. From the data it can be seen that the no treatment cells as well as treated and recovered cells do not have a usual G1-S-G2 cell cycle profile. We think this could be because the dye we are

35

Figure 15. Cells undergo DNA breakage before they go through mitosis. (A)

DF2 Myc CTG100 control cells. (B) DF2 Myc CTG100 cells treated with HU for 4

days and 0 hrs of recovery (C) DF2 Myc CTG100 cells treated with HU for 4 days

and 24 hrs of recovery (D) DF2 Myc CTG100 cells treated with HU for 4 days and

48 hrs of recovery(E) DF2 Myc CTG100 cells treated with HU for 4 days and 72

hrs of recovery (F) DF2 Myc CTG100 cells treated with HU for 4 days and960 hrs

of recovery

36 using-DRAQ 7 may be interfering with the emission wavelnegth of red and green fluorescent protein. However focusing on the X-axis we notice that cells are definitely stuck in one of the phases probably late S-early G2 phase. As recovery continues the cells overcome this arrest and slowly start moving into M phase and some cells have an early progress into G1 phase. By 96 hours of recovery the cells have a profile similar to the untreated cells suggesting its complete recovery. The separation of colour markers occurs at roughly the same time after HU treatment and I-SCE1 treatment, suggesting that DSBs occur during the HU treatment. Compared to the flow data of Figure 15, this suggests that DNA breaks occur at or before mitosis.

VIII. Loss of eGFP+, RFP- cells after stress induced double stand breaks

After observing that 35% cells are eGFP positive after 10 days of hydroxyurea treatment and 4 days of recovery, it was decided to allow the cells to recover over an extended period of time in order to understand more about the stability pattern of eGFP positive cells. Accordingly after 8 and 10 days respectively flow cytometry analysis was performed on DF2 Myc CTG100 cells treated with hydroxyurea for 10 days along with recovery. From Figure 16, 8 days of recovery showed only 26% cells that are eGFP positive.

This was accompanied by increase in double positive cells from 60% to 68%.

Similarly after 10 days of recovery there were only 7% eGFP positive cells and a corresponding increase of double positive cells to 89%. Thus over the course of an extended recovery period a loss of eGFP positive or green only cells was observed

37

After 4 days After 8 days After 10 days

A B C

89.7% RED R,G 60.1% RED R,G 68.4% RED R,G 0.0% 0.2% 0.0%

N,N GREEN N,N GREEN N,N GREEN

35.7 % 26.1 % 7.5 % 4.2% 5.3% 2.7%

+ - Figure 16. Loss of eGFP , RFP cells after stress induced double strand breaks. DF2 Myc CTG100 cells

were treated with HU and allowed to recover for (A) 4 days (B) 8 days © 10 days. Notice the loss of eGFP+,

RFP- cells with time.

38 after being subjected to stress induced double strand breaks. Figure 16 shows flow cytometry results of this assay. Note the loss of eGFP positive cells represented by decrease in percentage of cells detected in bottom right quadrant as recovery period increases. As discussed before once a cell loses either or both of its colours it can by no means regain the lost colour. Thus a decrease in percentage of eGFP positive cells with a simultaneous increase in percentage of double positive cells does not mean that the eGFP positive cells have become double positive. Rather is shows that eGFP positive cells could have undergone DNA breaks which could be extensive enough to lead to them undergoing apoptosis. Meanwhile the double positive cells being healthy could continue to grow and reconstitute the population.

IX. eGFP+, RFP-cells are unstable

After seeing the loss of eGFP positive cells over an extended recovery phase we decided to see if there is any loss of eGFP positive cells during the treatment period.

For this it was decided to check the number of eGFP positive cells during the course of the treatment. Results for this assay are shown in Figure 17. The graph represents values for percentage of double positive cells. It can be observed that there is not a definite trend to either increase or decrease of double positive cells over the course of hydroxyurea treatment and recovery. The days where number of double positive cells decreases goes in hand with same days showing an increase in the number of eGFP positive cells. Similarly increase in number of double positive cells simultaneously shows decrease in population of eGFP positive cells. Thus it goes to show that the eGFP positive cells are not stable as a population themselves. These results support data represented in Figure 16 which show loss of eGFP positive cells by flow 39 cytometry and help conclude that eGFP positive cells are unstable. It can also be said that RFP negative cells are unstable. Our interpretation of this data and the conclusion is based on our thinking about the position of the TNR region. The centromere is to the right of the TNR region thus DNA breaks at this region and loss of Tomato region would mean a considerable portion of the chromosome is broken and lost. If this loss is extensive it may cause the now eGFP positive cells to undergo apoptosis and thus show instability of these cells.

X. Short TNRs are stable under replication stress

To test the effect of hydroxyurea treatment on cells with a shorter length of trinucleotide repeat the DF2 MycCTG23 cells were used. The purpose of performing this assay was to understand if replication stress can induce DNA breaks in a short

TNR region. Results for this assay are shown in Figure 18. As can be observed the control or untreated cells have 94% cells that are double positive. However even after

10 days of hydroxyurea treatment and recovery of 4 days it was observed that around

95% cells remain double positive (panel B of figure 18). This assay was done several times with similar results and thus helped to conclude that cells with shorter trinucleotide repeats are stable under replication stress.

XI. DF2 cells are stable under replication stress

Since the DF2 cell line acts as one of the control cell lines it was imperative to run a stress response assay using these cells. Results shown in Figure 19 represent data for hydroxyurea stress response assay, where DF2 cells are treated with 0.2 mM hydroxyurea over a period of 2-10 days and allowed to recover. Panel A graphically

40

120%

Percentage of green 100% cells

80%

Double Positive 60%

Normalized to 40% control 20%

0%

CTG-100-CTRLCTG-100-HU-2D CTG-100-HU-4D CTG-100-HU-6D CTG-100-HU-8D CTG-100-HU-10D

Figure 17. GFP positive cells are unstable. Graph above represents percentage of green

only cells normalized to control. Notice the fluctuation in percentage of cells thus

showing the unstable nature if green cells as a population. .

41

(CTG) 23 control (CTG) 23 HU treated B A

RED R,G 0.1% RED R,G 95.1% 0.7% 94.8%

N,N GREEN N,N GREEN

1.2% 3.3% 0.7% 4.1%

23 Figure 18. Short TNRs are not unstable under replication stress. (A) Flow cytometry

profile of DF2 Myc CTG23 control or untreated cells. (B) Flow cytometry profile of DF2 Myc

CTG23 cells treated for 10 days with hydroxyurea and recovered for 4 days. Since percentage

of cells that are eGFP positive does not fluctuate they are said to be stable.

42

depicts percentage values for double positive cells which are normalized to control.

As can be observed after 2-10 days of HU treatment and recovery also almost all cells

are double positive (eGFP+, RFP+). Flow cytometry profiles for control as well as

hydroxyurea treatment conditions are shown in panel B and C respectively. These

results confirm with the behaviour of a control cell line. Thus in the absence of

trinucleotide repeat region cells are observed to be double positive.

XII. DF2 Myc cells are stable under hydroxyurea replication stress

After seeing little effect of replication stress on one of the control cell lines another

control cell line DF2 Myc was tested for hydroxyurea stress response. This was

necessary because the myc region also has some secondary structure (G quadruplex)

forming sequences. It was essential to know for sure that the effect seen on CTG100

cells in response to a replication stress was only due to the TNR region and not due to

other secondary structure forming sequences. Results for this assay are as shown in

Figure 20. In the control or untreated cells there are 98% cells double positive. Even

after 2, 4 or 6 days of treatment with hydroxyurea this number does not undergo a

significant change. With each increasing treatment the number changes from 98% to

90%. The cell line thus works as a good control. It also goes to show that the

trinucleotide repeats are regions undergoing DNA breaks and not any other region

between eGFP and RFP genes in the ectopic site.

43

Double Positive 120%

Percentage of green cells 100%

80%

60% Normalized Double Positive to control 40%

20%

0% DF2 DF2 day 2 DF2 day 4 DF2 day 6 DF2 day 8 DF2 day control 10

DF2 Control DF2 HU treated B C

RED R,G RED R,G 0.0% 98.4% 0.3% 98.7%

N,N GREEN N,N GREEN

1.2% 0/3% 2.4% 0.6%

Figure 19. DF2 cells are stable under HU replication stress. (A) graph representing

percentage of double positive cells in a control or untreated as well as HU treatment for

2-10 days. All values are normalized to control. (B) Flow cytometry profile of control

or untreated DF2 cells. (C) Flow cytometry profile of DF2 cells treated with

hydroxyurea.

44

DF2 Myc Control DF2 Myc HU treated

0.0% RED R,G 97.8% RED R,G 1.1% 88.1%

N,N GREEN N,N GREEN

0.9% 1.3 % 10.8% 0.2%

Figure 20. DF2 Myc cells are stable under hydroxyurea replication stress. (A) Flow

cytometry profile of DF2 Myc control or untreated cells. (B) Flow cytometry profile of DF2

Myc treated with hydroxyurea for 2 days, (C) 4 days, (D) 6 days.

45

XIII. CTG 100 TNRs are unstable due to aphidicolin replication stress

An aphidicolin stress response assay was performed to check whether hydroxyurea

stress response results were drug specific. The assay was similar to hydroxyurea

replication stress assay but aphidicolin was used in place of hydroxyurea. Results for

this assay are as shown in Figure 21. Panel A shows flow cytometry profile of control

or untreated DF2 Myc CTG100 cells. As can be seen 79% cells are double positive

while 17% cells are eGFP positive. With aphidicolin treatment however only 29%

cells remain double positive while 63% cells turn eGFP positive. This assay was

performed several times and similar results were obtained. Thus it showed that

CTG100 cells are unstable even when another replication stress inducer is used and

thus this response is not drug specific. Hydroxyurea is an inhibitor of ribonucleotide

reductase and causes depletion of dNTP polls while aphidicolin inhibits replicative

DNA polymerases.

XIV. CTG 100 TNRs are unstable due to oxidative stress

Up until now we saw how agents that cause direct replication stress affect stability of

TNR sequences. Next we wanted to induce environmental stress by use of hydrogen

peroxide (H2O2) and see if this has any effect on TNR sequences. This data is a result

of the work done by a former post doc in the lab, Dr Eric Romer (Wright State

University). Results for this assay are shown in Figure 22. Flow cytometry results

shown in Panels A, B, C are representative for DF2 Myc CTG 100 cells treated with

H2O2 3 times at an interval of 2 days while Panels D, E, F are representative for

46

(CTG) 100 Control (CTG) 100 APH treatment

A B

RED R,G RED R,G 29.1% 0.0% 79.1% 0.0%

N,N GREEN N,N GREEN

3.3% 17.5% 7.4% 63.5%

Figure 21. CTG 100 TNRs are unstable due to aphidicolin replication stress. (A)

Untreated cells (B) APH treated cells

47

DF2 Myc CTG 100 cells treated with H2O2 5 times at an interval of 2 days. Panel A are control cells showing only 11.7% eGFP positive cells. Treatment with 20 µM H2O2 gives only 11% eGFP positive cells. However a 200 µM treatment has a considerable effect and gives 25% eGFP positive cells. In Panel D control cells have only 17% eGFP cells and treatment with 20 µM H2O2 only gives 16% eGFP positive cells.

However 200 µM treatment shows 28.2% eGFP positive cells. The 200 µM concentration has been used to mimic in vivo levels of reactive oxygen

(ROS). This shows that a higher concentration of H2O2 is required to cause oxidative cells and this stress can also cause instability at the TNR region. However changing treatment regime from 3 to 5 times at intervals did not aggravate the effect which could suggest that cells could be reaching a saturation point for H2O2 treatment. The data thus shows that TNR sequences are unstable to replication stress as well as environmental stress and suggest that variety of DNA damaging agents can lead to increased instability of a long TNR region.

XV. DSBs at structured DNA are treated differently than ‘clean’ DSBs

Trinucleotide repeats are known to adopt non-canonical DNA structures. As can be seen from the results until now these repeats can also cause DNA breaks. A replication fork that is stalled at the hairpin forming sequences can lead to a one sided break (Figure 23) that must be repaired by homology directed mechanisms.

In contrast double sided breaks can be efficiently repaired by nonhomologous end joining (NHEJ) throughout the cell cycle, or by homologous recombination with the

48 sister chromatid during S/G2. However to understand influence of a clean double strand break on the trinucleotide repeatI-SCE-1, which is a restriction enzyme without an endogenous recognition site in the human genome, was used. Results for this assay are shown in Figure 24. Panel A shows flow cytometry profile of control or untreated

DF2 Myc CTG100 cells. Panel B shows flow cytometry profile of hydroxyurea treated

DF2 Myc CTG100 cells. It is seen that control has 85% double positive cells while in hydroxyurea treated cells there are only 60% double positive cells. Similarly while in control there are 11% eGFP positive, in hydroxyurea treated cells there are 36% eGFP positive. Both control or hydroxyurea treatment does not show presence of any cells which are RFP positive only. This however changes for cells treated by I-SCE-1.

While there are 43% cells double positive and 25% cells that are eGFP positiveonly, there are also significant percent of cell i.e. 25% cells which are double negative and

7% cells that are RFP positive only. It goes to show that depending on whether double strand breaks are due to structure forming sequences or are just ‘clean’ double strand breaks they are repaired by different mechanisms. Thus it gives rise a different sets of recombination products that can be analyzed by flow cytometry.

49

Figure 22. CTG 100 TNRs are unstable due to oxidative stress (A) Untreated

DF2 Myc CTG100 cells. (B) DF2 Myc CTG100 cells treated with 20 uM H2O2 - 3

times (C) DF2 Myc CTG100 cells treated with 200 uM H2O2 – 3 times (D)

Untreated DF2 Myc CTG100 cells. (E) DF2 Myc CTG100 cells treated with 20 uM

H2O2 - 5 times (F) DF2 Myc CTG100 cells treated with 200 uM H2O2 – 5 times

50

Single'sided*double*strand*break* Double'sided*double*strand*break*

Figure 23. Cartoon of single-sided and double –sided break double strand break.

Structure forming sequences that stall replication forks can lead to single sided double

strand break as shwon in the left panel. Factors other that structure forming sequences can

cause double sided double stranded breaks as shown in the right panel.

51

A B C Control HU I-SCE 1

RED R,G 85.5% RED R,G 60.1% RED R,G 0.0% 43.5% 0.0% 7.4%

N,N GREEN N,N GREEN N,N GREEN

11.6% 2.9% 35.7 % 4.2% 24.2% 24.9%

Figure 24. DSBs at structured DNA are treated differently than ‘clean’ DSBs. (A) flow

cytometry profile of control or untreated DF2 Myc CTG100 cells. (B) Flow cytometry profile of

DF2 Myc CTG100 cells treated with hydroxyurea. (C) Flow cytometry profile of DF2 Myc

CTG100 cells treated with I-SCE-1.

52

XVI. DF2-TTR cells are unstable due to Telomestatin and HU

Amongst secondary structure forming sequences another category are the pulypurine- polypyrimidine sequences. These sequences are prone to form triplex DNA and G quadruplexes. Formation of such structures can cause replication forks to stall and unless resolved can lead to DNA breaks. In order to study this further a former student in the laboratory, Joanna Barthelemy (Wright State University) made a cell line called

DF2 Myc TTR/TTF which has the same construct as one mentioned above but the

TNR region is replaced by the TTR/TTF sequence which involves the 88 bp polypurine-polypyrimidine tract from the human PKD1 , and carried out assays discussed in the next two figures. DF2 Myc TTR (Reverse) cell line has the polypurine strand as template for lagging strand synthesis. DF2 Myc TTF (Forward) cell line has polypurine strand as template for leading strand synthesis. The principle of placing purines in either lagging or leading strand template is as mentioned before.

Since lagging strand is single stranded for a longer duration it provides greater chance for formation of secondary structures. These cells were treated with telomestatin

(TMS) which recognizes G quadruplexes and stabilizes them.

Telomestatin showed an effective replication stress on the DF2 Myc TTR as shown in

Figure 25. Hydroxyurea alone gives 22.4% eGFP positive cells. Telomestatin alone shows 65.5% cells that are eGFP positive. A combination of telomestatin with hydroxyurea shows 73.9% eGFP positive cells. Thus it shows that DF2 Myc TTR cells are unstable under effect of telomestatin alone and in combination with hydroxyurea.

53

XVII. DF2-TTF cells are stable in Telomestatin and HU

On seeing effect of telomestatin on DF2 Myc TTR cells it was essential to know how cells with a differing orientation of purine sequences would react to replication stress.

For this DF2 Myc TTF cells were treated with telomestatin, hydroxyurea and a combination of both. Results for this data are represented in Figure 26.Interestingly it was observed that while hydroxyurea alone had no effect on the cells; telomestatin and a combination of both reagents also did not show any significant effect on these cells. Most of the cells remained double positive. These results thus show that DF2

Myc TTF cells are stable under effect of telomestatin and hydroxyurea. The data thus shows that effective of replication stress causing agents greatly depends on whether secondary structure forming sequences are present in the template for lagging strand synthesis or leading strand.

54

A B 75.6% RED R,G 77.7% RED R,G 0.1% 0.2%

Control 0.2 mM HU N,N GREEN N,N GREEN

1.9% 22.4% 1.8% 20.4%

C D

RED R,G RED R,G 23.9% 0.1% 32.6% 0.1% 0.5 µM TMS 0.5 µM TMS + N,N GREEN N,N GREEN 0.2 mM HU

1.9% 65.5% 2.0% 73.9%

Figure 25. DF2-TTR cells are unstable due to Telomestatin and HU. (A) Flow cytometry profile of control or

untreated DF2 Myc TTR cells. (B) Flow cytometry profile of DF2 Myc TTR cells treated with 0.2 mM hydroxyurea. (C)

Flow cytometry profile of DF2 Myc TTR cells treated with0.5µM of telomestatin. (D) Flow cytometry profile of DF2

Myc TTR cells treated with0.5µM of telomestatin and 0.2 mM hydroxyurea.

55

Figure 26. DF2-TTF cells are stable in Telomestatin and HU. (Data from

Barthelemy). (A) Flow cytometry profile of control or untreated DF2 Myc

TTRcells. (B) Flow cytometry profile of DF2 Myc TTRcells treated with0.2 mM

hydroxyurea. (C) Flow cytometry profile of DF2 Myc TTRcells treated with0.5µM

of telomestatin.(D) Flow cytometry profile of DF2 Myc TTRcells treated

with0.5µM of telomestatin and 0.2 mM hydroxyurea.

56

XVIII. CTG 100 TNRs are unstable after Telomestatin treatment

Seeing effect of a replication stress inducer on other structure forming sequences we were interested in seeing how the DF2 Myc CTG100 cells would react under effect of telomestatin. Ideally we thought since CTG is responsible to form hairpin structures telomestatin would not have any effect on the cells.

To our surprise however we saw a significant effect as shown in Figure 27. As can be seen telomestatin alone makes 49.6% cells eGFP positive while in combination with hydroxyurea makes 52.9% cells eGFP positive. We think that since 100 repeats of

CTG have ample scope to form multiple hairpins, these hairpins could be coming together to form a G quadruplex type structure. This can be recognized by telomestatin and which is why we see an effect. Thus it shows that DF2 Myc CTG100 cells are also unstable to telomestatin alone.

57

Control 0.5µM TMS HU + TMS A B C

RED R,G RED R,G 44.1% RED R,G 40.6% 82.8% 0.0% 0.0% 0.0%

N,N GREEN N,N GREEN 49.6% N,N GREEN 52.9% 12.3% 4.9% 6.3% 6.4%

Figure 27. DF2-CTG100cells are unstable due to Telomestatin and HU. (A) Flow cytometry profile of

control or untreated DF2 Myc CTG100cells. (B) Flow cytometry profile of DF2 Myc CTG100cells treated

with0.5µM of telomestatin. (C) Flow cytometry profile of DF2 Myc CTG100cells treated with0.5µM of

telomestatin and 0.2 mM hydroxyurea.

58

XIX. DF2 Myc cells are stable under replication stress

Telomestatin was able to induce instability in the DF2 Myc CTG100 cells. It made sense to run a similar assay on the control cells. For this the DF2 Myc cells were used.

The myc region has sequences which form G quadruplexes. Thus use of this cell line would help to show specificity of telomestatin effect towards either TNR region or to the quadruplex forming sequences in the myc region. Results for this assay are represented in Figure 28. As can be seen from the assay telomestatin alone makes only 10% cells eGFP positive wile a combination with hydroxyurea causes only 8% cells eGFP positive. This shows that effect of telomestatin is specific to the TNR region and not due to other secondary structure contributing sequences. Nor are the results of Figure 27 due to a general effect on the cell.

59

Control 0.5 mM TMS HU + TMS

RED R,G RED R,G 87.0% RED R,G 89.2% 0.0% 97.4% 0.0% 0.1%

N,N GREEN N,N GREEN N,N GREEN

0.8% 1.9 % 10.3 % 2.7% 8.4 % 2.4%

Figure 28. DF2 Myc cells are stable under replication stress. (A) Flow cytometry profile of control or untreated

DF2 Myc cells. (B) Flow cytometry profile of DF2 Myc cells treated with0.2 mM of hydroxyurea. (C) Flow

cytometry profile of DF2 Myc cells treated with0.5µM of telomestatin. (D) Flow cytometry profile of DF2 Myc

cells treated with0.5µM of telomestatin and 0.2 mM hydroxyurea.

60

XX. Loss of translesion polymerases in DF2 Myc CTG100 cells affects DNA

damage tolerance

As mentioned earlier translesion polymerases are specialized polymerases that

function as part of DNA damage tolerance pathway to help overcome DNA damage.

The data earlier showed that in response to replication stress induced by different

agents DF2 Myc CTG100 cells undergo DNA breaks and are thus unstable. Hence it

was decided to understand whether these cells overcome the induced replication stress

by means of translesion synthesis. The first assay performed in this context was a

usual flow cytometry assay where using siRNAs we tried to knockdown Rad18, Pol

eta, Pol kappa and Rev1. Along with knockdown of these TLS polymerases we also

combined knockdown and replication stress induction using hydroxyurea. The set up

was accompanied by use of siControl as well and a combination of siControl and

hydroxyurea. Thus each treatment condition was accompanied by its respective

control. Use of siControl essentially means there is no knockdown of any protein.

Results for this assay are shown in Figures 29-32.

In Figure 29 flow cytometry profiles represent data for knockdown/ no knockdown of

Rad18 in presence or absence of hydroxyurea. 53.6% cells are found to be eGFP

positive under hydroxyurea in the presence of siControl compared to only 6.4% cells

that are eGFP positive in the siControl cells in absence of hydroxyurea. Knocking

down Rad18; minimally increased the percentage of eGFP positive (RFP negative)

cells to 10%. However when knockdown of Rad18 was combined with hydroxyurea

61 replication stress the percentage of eGFP positive, RFP negative cells increases from

53.6% to 74.9%. This shows that knocking down Rad18 causes cells to lose their ability to tolerate the non-canonical CTG/CAG structure. In other words Rad18 is an important protein to the cell to overcome DNA lesions.

Figure 30 represents data for knockdown/ no knockdown of Pol eta in presence or absence of hydroxyurea. 57.4 % cells are found to be eGFP positive under the effect of hydroxyurea in the presence of siControl compared to only 7.8 % cells that are eGFP positive in the siControl cells in absence of hydroxyurea. Knocking down Pol eta, increases the percentage of cells that are eGFP positive to 13 %. However when knockdown of Pol eta is combined with hydroxyurea replication stress it gives a dramatic increase in to the percentage of eGFP positive cells to 67.3 %. These results thus show that Pol eta is an essential protein for survival of cells under replication stress by tolerating the damage occurring and kick starting the process of translesion synthesis.

Figure 31 represents data for knockdown/ no knockdown of Pol kappa in presence or absence of hydroxyurea. 48.4 % cells are found to be eGFP positive under the effect of hydroxyurea in the presence of siControl compared to only 8.2 % cells that are eGFP positive in the siControl cells in absence of hydroxyurea. Knocking down Pol kappa, gives 30 % cells that are eGFP positive. However when knockdown of Pol kappa is combined with hydroxyurea replication stress it does not increase or decrease the effect of HU, i.e. 30 % eGFP positive cells. The absence of a synergistic effect of

Pol kappa siRNA and HU suggests that Pol kappa knockdown mimics the effects of

62 replication stress. Finally Figure 32 represents data for knockdown/ no knockdown of

Rev1 in presence or absence of hydroxyurea. 48.2 % cells are found to be eGFP positive under effect of hydroxyurea in the presence of siControl compared to only

8.8 % cells that are eGFP positive in the siControl cells in absence of hydroxyurea.

Knocking down Rev1 gives 8.5 % eGFP positive cells. However when knockdown of

Rev1is combined with hydroxyurea replication stress it gives only 45.7 % eGFP positive cells. These results thus suggest that Rev1 is not involved in stabilization of

CTG/CAG structures during replication stress. Knockdown of Rev1 also does not exacerbate number of eGFP positive cells but it surely does not protect the cells from damage. Rev1 thus may not be critical for cell survival it but is an essential protein for the cell.

63

A C

0.1% RED R,G 91.8% RED R,G 0.1% 87.2%

N,N GREEN N,N GREEN

1.7% 6.4% 1.9% 10.7%

siControl siRad18

B D

RED R,G RED R,G 0.1% 43.0% 0.2% 17.7%

N,N GREEN N,N GREEN

53.6% 74.9% 3.3% 7.3%

siControl + HU siRad18 + HU

Figure 29. Loss of Rad18 in DF2 Myc CTG100 cells affects DNA damage tolerance. (A) Flow cytometry profile

of DF2 Myc CTG100control or untreated cells. (B) Flow cytometry profile of DF2 Myc CTG100 treated with

siControl and hydroxyurea for 4days (C) Flow cytometry profile of DF2 Myc CTG100 treated with siRad18

(D)Flow cytometry profile of DF2 Myc CTG100 treated with siRad18 and hydroxyurea for 4days.

64

C A RED R,G RED R,G 89.9% 0.2% 0.2% 83.8%

N,N GREEN N,N GREEN

2.1% 7.8% 3.0% 13%

siControl siPol Eta

B D

RED R,G RED R,G 0.1% 38.8% 0.1% 27.6%

N,N GREEN N,N GREEN

57.4% 67.3% 3.7% 5.0%

siControl + HU siPol Eta + HU

Figure 30. Loss of Pol eta in DF2 Myc CTG100 cells affects DNA damage tolerance. (A) Flow

cytometry profile of DF2 Myc CTG100control or untreated cells. (B) Flow cytometry profile of DF2

Myc CTG100 treated with siControl and hydroxyurea for 4days (C) Flow cytometry profile of DF2

Myc CTG100 treated with siPol eta (D)Flow cytometry profile of DF2 Myc CTG100 treated with

siPol eta and hydroxyurea for 4days.

65

A C

RED R,G 89.6% RED R,G 0.1% 0.2% 65.7%

N,N GREEN N,N GREEN

2.1% 8.2% 4.1% 30%

siControl siPol Kappa

B D

RED R,G RED R,G 0.1% 47.8% 1.0% 63.1%

N,N GREEN N,N GREEN

48.4% 30% 2.7% 5.9%

siControl + HU siPol Kappa + HU

Figure 31. Loss of Pol kappa in DF2 Myc CTG100 cells affects DNA damage tolerance. (A) Flow

cytometry profile of DF2 Myc CTG100control or untreated cells. (B) Flow cytometry profile of DF2

Myc CTG100 treated with siControl and hydroxyurea for 4days (C) Flow cytometry profile of DF2

Myc CTG100 treated with siPol Kappa (D)Flow cytometry profile of DF2 Myc CTG100 treated with

siPol Kappa and hydroxyurea for 4days

66

A C

RED R,G 89.5% RED R,G 0.2% 0.2% 89.5%

N,N GREEN N,N GREEN 1.5% 8.8% 1.8% 8.5%

siControl siRev1

B D

RED R,G RED R,G 0.1% 48.4% 0.1% 50.7%

N,N GREEN N,N GREEN 48.2% 45.7% 3.3% 3.5%

siControl + HU siRev1 + HU

Figure 32. Loss of Rev1 in DF2 Myc CTG100 cells minimally affects DNA damage tolerance. (A)

Flow cytometry profile of DF2 Myc CTG100control or untreated cells. (B) Flow cytometry profile of

DF2 Myc CTG100 treated with siControl and hydroxyurea for 4days (C) Flow cytometry profile of DF2

Myc CTG100 treated with siRev1 (D)Flow cytometry profile of DF2 Myc CTG100 treated with siRev1

and hydroxyurea for 4days.

67

XXI. Loss of translesion polymerases in DF2 Myc cells does not affect DNA

damage tolerance

From the experiments above it was seen that certain translesion polymerases are

essential for a cell to tolerate non-canonical structures, and in their absence this ability

to overcome this impediment diminishes. As we had seen before that DF2 Myc cells

are stable to replication stress it was decided to use these cells and knockdown

proteins involved in translesion synthesis to see how they cope up. This also included

treatment/no treatment of cells with hydroxyurea as a replication stress inducer.

Results for this assay are shown in Figures 33-36.

Figure 33 represents data for knockdown/ no knockdown of Rad18 (siControl) in

presence or absence of hydroxyurea. As can be observed hydroxyurea has no effect on

cells treated with siControl. Similarly knockdown of Rad18 alone or a combination of

hydroxyurea treatment also maintains stability of DF2 Myc cells; more than 90% cells

are double positive. This shows that loss of translesion polymerases does not make the

DF2 Myc cells unstable. It indicates that instability seen earlier in DF2 Myc CTG100

cells is due to the TNR region alone. It also shows that G quadruplex forming

sequences present in c-myc region are able to cope up with loss of translesion

polymerases suggesting these structures if formed are resolved by normal repair

mechanisms.

68

Figure 34-36 represents data for knockdown/ no knockdown of Pol Eta, Pol Kappa and Rev1 respectively in presence or absence of hydroxyurea. All of them show that treatment of cells with hydroxyurea alone or combined with knockdown of either Pol eta, Pol kappa or Rev1 does not affect stability of the cells. This shows that translesion polymerases are critical to allow replication to bypass through DNA lesions. Their absence particularly affects the ability of cells to overcome large DNA lesions.

69

A C

RED R,G 0.0% RED R,G 97.8% 0.0% 97.4%

N,N GREEN N,N GREEN

1.4% 0.8% 1.7% 0.8%

siControl siRad18

B D

RED R,G RED R,G 0.0% 96.8% 0.5% 93.0%

N,N GREEN N,N GREEN

2.1% 4.3% 1.1% 2.2%

siControl + HU siRad18 + HU

Figure 33. Loss of Rad18 in DF2 Myc cells does not affect DNA damage tolerance. (A) Flow

cytometry profile of DF2 Myc control or untreated cells. (B) Flow cytometry profile of DF2 Myc

treated with siControl and hydroxyurea for 4days (C) Flow cytometry profile of DF2 Myc treated with

siRad18 (D)Flow cytometry profile of DF2 Myc treated with siRad18 and hydroxyurea for 4days.!

!!

70

A C

RED R,G RED R,G 0.0% 97.7% 0.0% 97.4%

N,N GREEN N,N GREEN

1.9% 0.3% 1.3% 1.3%

siControl siPol Eta

B D

RED R,G 0.0% 97.3% 0.2% RED R,G 95.8%

N,N GREEN N,N GREEN 1.5% 2.0% 1.2% 1.9%

siControl + HU siPol Eta + HU

Figure 34. Loss of Pol Eta in DF2 Myc cells does not affect DNA damage tolerance. (A) Flow

cytometry profile of DF2 Myc control or untreated cells. (B) Flow cytometry profile of DF2 Myc

treated with siControl and hydroxyurea for 4days (C) Flow cytometry profile of DF2 Myc treated with

siPol Eta (D)Flow cytometry profile of DF2 Myc treated with siPol Eta and hydroxyurea for 4days.!

!!

71

A C RED R,G RED R,G 0.0% 95.3% 97.7% 0.0%

N,N GREEN N,N GREEN

4.1% 1.9% 0.3% 0.6%

siControl siRev1

B D

RED R,G RED R,G 0.0% 96.4% 0.2% 95.8%

N,N GREEN N,N GREEN 2.3% 2.1% 1.2% 1.9%

siControl + HU siRev1 + HU

Figure 35. Loss of Pol Kappa in DF2 Myc cells does not affect DNA damage tolerance. (A) Flow

cytometry profile of DF2 Myc control or untreated cells. (B) Flow cytometry profile of DF2 Myc treated

with siControl and hydroxyurea for 4days (C) Flow cytometry profile of DF2 Myc treated with siPol

Kappa(D)Flow cytometry profile of DF2 Myc treated with siPol Kappa and hydroxyurea for 4 days.!

72

A C

0.0% RED R,G 97.6% RED R,G 0.0% 97.6%

N,N GREEN N,N GREEN

1.2% 1.3 % 1.3% 1.0%

siControl siPol Kappa

B D

RED R,G RED R,G 96.8% 92.2% 0.0% 0.2%

N,N GREEN N,N GREEN 1.8% 5.2% 1.3% 7.3%

siControl + HU siPol Kappa + HU

Figure 36. Loss of Rev1 in DF2 Myc cells does not affect DNA damage tolerance. (A) Flow

cytometry profile of DF2 Myc control or untreated cells. (B) Flow cytometry profile of DF2 Myc treated

with siControl and hydroxyurea for 4days (C) Flow cytometry profile of DF2 Myc treated with

Rev1(D)Flow cytometry profile of DF2 Myc treated with Rev1 and hydroxyurea for 4days!

!!

73

XXII. Knockdown of translesion polymerases in DF2 Myc CTG100 cells

In order to confirm that the effects seen on DF2 Myc CTG100 cells by knockdown of respective translesion polymerases was indeed because of their knockdown, we decided to check their knockdown efficiency by Western blotting. Accordingly

Western blots were run for knockdown of Pol eta, Pol kappa and Rev1. Figure 37-39 represents data for the same. Figure 37 shows knockdown of Pol eta. The Pol eta band is seen in the control lane running at 76 kd. However the 24 hours siRNA sample shows significant knockdown of the protein. Actin loading is even; suggesting proper sample loading. Pol kappa knockdown is shown in Figure 38. Pol kappa runs above

95 kd and is seen here as doublet. Although actin loading is lower for 24 hours siRNA sample, knockdown is seen till 48 hours post siRNA treatment. Finally Figure 39 shows knockdown of Rev1. There is a significant Rev1 knockdown at 24 hours post siRNA treatment. Since siRNA functions only for a limited time, Rev1 band can be seen again at 48 hours post treatment. All these blots thus show that siRNAs used in our assays were functioning to knockdown the specific proteins. We are currently working on western blot of knockdwon of Rad18.

74

hrs hrs

Eta -24

siControl-24 siPol

Pol Eta (76 kd)

actin (40 kd)

Figure 37. Knockdown of Pol eta using siRNA in DF2 Myc CTG100 cells.

Pol eta band runs at 76 kd. This band is seen in control at 24 hours but not in

the siRNA against Pol eta lane at 24 hours. Actin loading is even at 40 kd.!

75

hrs hrs hrs hrs

kappa-24 kappa-48 kappa-96 siControl-24 siPol siPol siPol

Pol kappa (95 kd)

actin (40 kd)

Figure 38. Knockdown of Pol kappa using siRNA in DF2 Myc CTG100 cells. Pol

kappa band runs at 95 kd. This band is seen in control at 24 hours but not in the

siRNA against Pol kappa lane at 24 hours. Actin loading is even at 40 kd.

76

hrs hrs hrs hrs hrs hrs

siControl-24siRev1-24 siControl-48 siRev1-48 siControl-96 siRev1-96

Rev1 (138 kd)

actin (40 kd)

Figure 39. Knockdown of Rev1 using siRNA in DF2 Myc CTG100 cells. Rev1

band runs at 138 kd. This band is seen in control at 24 hours but not in the siRNA

against Rev1 lane at 24 hours. Actin loading is even at 40 kd.

77

DISCUSSION

Trinucleotide repeats as fragile sites

Expansions and contractions of trinucleotide repeats are well known and extensively studied. The data presented here shows that expanded trinucleotide repeats and the

PKD1 asymmetric Pu/Py repeat tract can also be considered as fragile sites that can lead to chromosome breaks during replication stress. In this study while making DF2

Myc CTG100/23 cell line the trinucleotide repeats were placed in the template for lagging strand synthesis. This was done in order to increase the chances for formation of DNA secondary structures (Cleary JD et al. 2002, Burrow AA et al. 2010,

Mirkin SM 2006). Since the lagging strand is expected to be single stranded for a slightly longer duration compared to leading strand during replication, the lagging strand template is thus more inclined to form hairpin structures. Once stable cell lines were made it was seen that Pu/Py and CTG/CAG repeats show spontaneous instability, which is represented by the eGFP positive cells. Also in order to induce formation of structures that may contribute to instability, reagents responsible for causing replication stress (hydroxyurea, aphidicolin, telomestatin) were used. Our lab was also able to show that DF2 Myc CTG100 cells are unstable under effect of hydrogen peroxide (H2O2). Multiple such reagents were used to rule out specificity of response of DF2 Myc CTG100/23 cells to any one of the reagents. This also helped to 78 show that these cells are unstable to reagents that can induce stress in different forms and affect replication in some manner.

Treatment of DF2 Myc CTG100 cells with hydroxyurea followed by a recovery phase lead to significant percent of cells turning green. The concentration of hydroxyurea used was enough to stall replication forks (Petermann et al. 2010). As the treatment duration increased percentage of green only cells also increased. Thus it shows that

DNA breaks and recombination between Alu 1 and Alu2 occurs. Similar results were obtained for aphidicolin treatment of DF2 Myc CTG100 cells. Use of aphidicolin showed a greater increase in percentage of green only cells. Thus both assays showed that DF2 Myc CTG100 cells are unstable to replication stress. Also it showed that effect on these cells was not reagent specific. The Pu/Py repeats however showed instability depending on replication polarity when replication stress inducing agents were used. This shows that polarity of structure forming sequences-whether these sequences are in leading or lagging strand can determine their stability.

Telomestatin is a drug that acts as a telomerase inhibitor. It recognizes G quadruplexes (Kim, M.Y. et al. 2003, Wu, Y et al. 2010).It can bind to pre-formed

G4 DNA however it can also induce formation of G4 DNA and then stabilize that structure by attaching to it. Thus it has the potential to induce formation of secondary structures and thus move the equilibrium to formation of non-B DNA over B form of

DNA. CAG/CTG repeats are known to form hairpin structures in vivo ( Liu, G. and

Leffak, M. (2012), Liu, G et al. (2010) ). So the idea of using telomestatin was to show that it cannot induce formation of hairpin structures. In absence of hairpin

79 structures we did not expect to see any effect on the DF2 Myc CTG100 cells. However to our surprise use of telomestatin alone as well as its combination with hydroxyurea showed a significant effect. The results were similar as many number of times as they were repeated. This helped to make a conclusion that possibly the several hairpins that are formed due to a long CTG repeat could be coming together and forming a structure similar to a G quadruplex. If this is so, telomestatin can recognize this structure and attach to it which is why we are seeing an effect to such a dramatic extent. The assay thus concludes that CTG/CAG repeats are unstable after TMS treatment. Stabilization of G quadruplexes by TMS leads to DSBs. This effect is increased by HU, indicating a role for G4 interference with replication.

Thus through this study we were able to show that expanded CTG/CAG repeats are hypersensitive to DSBs as a result of replication fork stalling (HU, APH, peroxide).

Some of these DSBs appear before mitosis.

Trinucleotide repeat length is critical

From the various assays performed using DF2 Myc CTG100 cells and treated with hydroxyurea or aphidicolin or telomestatin it was clear that the region is unstable to any kind of replication stress and that it was undergoing DNA breaks. However treatment of DF2 Myc CTG23 cells with hydroxyurea did not cause instability of cells.

Rather all cells remained double positive suggesting that the cells are not undergoing any DNA breaks. It also meant that the short trinucleotide repeat region is stable to replication stress. This indicated that a short trinucleotide repeat is immune to DNA

80 breaks. Although the possibility of formation of hairpins at the trinucleotide region is not refuted it is not sufficient to cause instability and thus lead to DNA strand breaks.

Another possibility could be, although hairpins are formed at such short sequences and replication forks are stalled, the proteins responsible for resolving stalled forks could function. This would avoid action of endonucelease for eg. MUS81 at the stalled forks and thus not allow the forks to get processed as double strand breaks and resort to homologous recombination. Treatment of DF2 Myc cells with hydroxyurea also did not lead to any instability. Both these results thus help to conclude that length of trinucleotide region is crucial as far as instability and DNA breaks are concerned and that normal length repeats and c-myc origin are not unstable.

Unrepaired DNA breaks cause cell death or senescence

The initial set of assays performed by treating DF2 Myc CTG100 cells with hydroxyurea and aphidicolin helped to conclude that trinucleotide repeats are prone to

DNA breaks. However when the prolonged recovery phase assay was performed it showed some interesting results. From the results it was observed that with increasing recovery period the population of double positive cells was replenishing while the eGFP positive cells were disappearing. As discussed before once the cell loses either or both of its colours there is no way of regaining the lost colour. Thus the probability of eGFP positive cells turning double positive was ruled out. The only explanation available was that since the eGFP positive cells represent a population that has undergone DNA break these cells could be undergoing senescence apoptosis. A

81 different set of experiments helped to further strengthen this interpretation. In the graphical representation for instability of eGFP positive cells, the percentage of double positive cells decreases on day 2, 6, 8 of hydroxyurea treatment. This was simultaneously accompanied by increase in percentage of eGFP positive cells.

However on day 4 when number of double positive cells would increase it was observed that number of eGFP positive cells would decrease. We interpret this as apoptosis of eGFP positive cells and thus a decrease in their number, while the double positive cells or ‘healthy’ cells are able to duplicate. Both results in combination help to conclude that the population which undergoes DNA breaks and are left unrepaired could undergo apoptosis. From this study we found that DSBs at noncanonical DNA structures are not readily repaired and lead to cell death or senescence.

Structure forming sequences influence repair mechanisms at broken ends of

DNA

Throughout the study it was observed that DNA breaks in the trinucleotide repeat region if any would cause loss of Tomato and thus give eGFP positive cells. It meant that repair mechanisms undertaken for repair of these DNA breaks were of a particular kind. To study this in details the assay using I-SCE-1 was performed.

Results of this assay gave some interesting results, where in addition to eGFP positive cells, a significant percent of double negative cells and even RFP positive cells were observed. Thus it showed that use of a restriction enzyme like I-SCE-1 that can generate a ‘clean’ double strand break leads to the cell undertaking different repair

82 mechanisms than those to repair broken DNA due to structure forming sequences. In case of a double strand break the cell has an only choice of going for a homologous recombination repair to salvage the DNA. This is because the only template available to generate lost part of DNA/chromosome is from the sister chromtid. This however is not the only option left for a cell if it undergoes DNA breaks due to non-B DNA structures. Generally only the structure forming region is excised and with the help of the other strand of DNA as a template the excised portion is filled. Thus another conclusion made from this study was single-sided DSBs are repaired differently than double-sided DSBs. Use of the dual florescence – DF system can be used to screen for agents tat cause DSBs.

It is clear thorough several studies that replication forks can stall under certain circumstances and thus damage DNA (Mirkin et al. 2007). High fidelity polymerases that function at the replication fork fall short to overcome this damage which is when it becomes necessary to repair the damage maintaining genetic stability, or bypass the

DNA lesion at the cost of accuracy of genetic information. The DNA damage tolerance pathway which helps to bypass the damage however promotes cell survival

(Waters et al. 2009). We applied this information to study chromosome breaks at

TNR region and gained further insight into it.

Translesion polymerases help overcome DNA damage and DNA breaks at trinucleotide repeats

Although it was established that translesion polymerases help tolerate DNA damage we sought to see if this holds true for DNA damage at trinucleotide repeat structures.

83

The data we obtained does indeed suggest this. Consider the flow cytometry data for

DF2 Myc CTG100 cells treated with siControl or siTLS Polymerase but in absence of hydroxyurea. siControl essentially means there is no knockdown. However use of specific siRNAs would cause knockdown of specific TLS polymerase. Only the knockdown of a particular protein without inducing any DNA damage would not cause cells to turn eGFP positive. This shows that the TLS polymerases are active only when there is some kind of DNA damage induced (Waters et al. 2009). Also loss of translesion polymerase alone does not affect stability of these cells. However, when knockdown is followed by induction of DNA damage it leads to major instability in the cells. Compared to damage induction but no knockdown, loss of translesion polymerases leads to increased DNA breaks at the TNR region.

Both Rad18 and Pol eta knockdown cause major instability in the TNR region as can be observed from their respective flow profiles. Under effect of hydroxyurea but in absence of Rad18, PCNA may not have been ubiquitinated and thus the translesion pathway may not have been activated. This may have lead to increased DNA breaks and thus an increased number of cells turned eGFP positive. Thus it shows that cells undertake translesion synthesis as a mechanism to tolerate and overcome DNA distortions.

Pol Eta is the first polymerase recruited by Rad18 after PCNA gets ubiquitinated. The replicative polymerase delta is replaced by Pol eta with the help of Rad18. It is after this replacement that the process of bypassing the DNA lesion and filling gaps that are formed if any starts. However, if Pol eta is knocked down then polymerase delta

84 cannot be replaced since this is the role played by Pol eta only. Thus translesion synthesis cannot begin. Also, Pol eta is known to act at stalled forks during S phase to replicate across non-B DNA structures (Boyer et al 2013). In the absence of this protein due to a knockdown, translesion synthesis across the TNR region would not have been possible, which led to DNA breaks.

Under hydroxyurea replication stress but without knocking down Pol Kappa the TNR region shows instability. However when hydroxyurea replication stress is coupled with Pol kappa knockdown it does not seem to exacerbate instability at the TNR region. Rather the number of eGFP positive cells decrease suggesting fewer DNA breaks at the TNR region. Pol kappa is shown to be unable to replicate across DNA with numerous lesions (Yang et al. 2007). The TNR region is long enough to form multiple hairpins and thus form several DNA lesions, which may be why loss of Pol kappa under stress does not increase instability of TNR region and thus increase DNA breaks.

Loss of Rev1 alone did not affect stability of TNR region. However when this loss was combined with replication stress by hydroxyurea it increased DNA breaks at

TNR region although not to the same extent as cells which were only treated with hydroxyurea but did not experience knockdown of Rev1. Rev1 has been shown to promote mutagenesis in yeast as well as multi cellular organisms including humans

(Gibbs et al. 2000). Mutations in Rev1 have been shown to decrease the number of spontaneous or induced mutagenesis (Lawrence et al. 2004). Based on this

85 information it may not be surprising to see that even though cells are stressed and

Rev1 has been knocked down it does not increase DNA breaks.

We thus conclude that translesion polymerases contribute to cell survival in case of

DNA damage when the replication machinery fails to do its job. These proteins help takeover the function of replicative polymerases in order to ensure replication continues to promote cell survival. It was shown already that TLS functions when

DNA is damaged. However this data also shows that these polymerases and thus the process of translesion synthesis is necessary to avoid chromosome breaks as a result of DNA damage.

Translesion polymerases bypass secondary structures at TNR regions

It was fascinating to see a significant effect of knockdown of TLS polymerases on the

TNR region. However it was vital to rule out that this could be an effect of other structure forming sequences or to ensure that this was a structure-damage inter dependent phenomenon. This is where the DF2 Myc cells played their part. Results shown above thus show that DNA damage is very much dependent on presence of structure forming sequences and in its absence loss of translesion polymerases does not cause DNA breaks. None of the proteins on being knocked down resulted in DNA breaks represented by eGFP positive cells. This shows specificity of TLS polymerases to function at TNR region specifically.

Knockdown assays confirm that the effect seen by use of siRNA and analyzed by flow cytometry was indeed due to knockdown of these proteins. Pol eta did not show a complete knockdown at 24 hours; however flow cytometry results showed

86 significant effect on DF2 Myc CTG100 cells. This suggests that even a small knockdown of these proteins can cause a considerable effect further strengthening their importance in overcoming instability. The assays thus help show that flow cytometry effects were a result of knockdown of these proteins and not other off-side effects.

In conclusion loss of translesion polymerases hinders with the ability of a cell to cope up with DNA damage. In this study which shows chromosome breaks at TNR region, loss of TLS polymerases not only affects their ability to handle DNA damage but also leads to increased DNA breaks in some cases. This shows how crucial TLS polymerases are essential to a cell for its survival.

87

FUTURE DIRECTIONS

In future we would like to study the stability pattern of DF2 Myc CAG 100 cell line under replication stress. Here too we plan to use the same stress inducing agents we used in this study like hydroxyurea, aphidicolin, hydrogen peroxide and telomestatin.

In this study we saw that the trinucleotide region is what could be causing DNA breaks. So in future we would like to map the location of these breaks. Finally we would like to knockdown additional fork remodeling proteins and screen for protein knockdowns that restore processing to structured breaks. This includes knockdown of proteins like RECQ1, MUS81 etc. and other translesion polymerase proteins like Pol zeta and see effect of loss of these proteins on the stability of cells containing trinucleotide repeat region.

88

REFERENCES

1. Bavoux, C., J. S. Hoffmann, and C. Cazaux. 2005. Adaptation to DNA

damage and stimulation of genetic instability: the double-edged sword

mammalian DNA polymerase kappa. Biochimie87:637–646.

2. Bavoux, C., A. M. Leopoldino, V. Bergoglio, O. W. J., T. Ogi, A. Bieth, J. G.

Judde, S. D. Pena, M. F. Poupon, T. Helleday, M. Tagawa, C. Machado, J. S.

Hoffmann, and C. Cazaux. 2005. Up-regulation of the error-prone DNA

polymerase _ promotes pleiotropic genetic alterations and tumorigenesis.

Cancer Res. 65:325–330Benton CS, de Silva R, Rutledge SL, Bohlega S,

Ashizawa T, Zoghbi HY.Molecular and clinical studies in SCA-7 define a

broad clinical spectrum and the infantile phenotype. Neurology.

1998;51(4):1081-1086

3. Bienko,M., Green,C.M., Crosetto,N., Rudolf,F., Zapart,G., Coull,B.,

Kannouche, P.,Wider,G., Peter,M., Lehmann,A.R. et al. (2005) Ubiquitin-

binding domains in Y-family polymerases regulate translesion synthesis.

Science, 310, 1821–1824.

4. Boyer AS, Grgurevic S, Cazaux C, Hoffmann JS (2013) The human

specialized DNA polymerases and non-B DNA: vital relationships to preserve

genome integrity. J MolBiol 425(23):4767–4781

89

5. Burrow AA, Marullo A, Holder LR, Wang YH (2010) Secondary structure

formation and DNA instability at fragile site FRA16B. Nucleic Acids Res 38:

2865-2877.

6. Chatterjee N, Lin Y, Santillan BA, Yotnda P, Wilson JH: Environmental stress

induces trinucleotide repeat mutagenesis in human cells. Proc Natl Acad Sci U

S A 2015, 112:3764e3769

7. Choi, J. H., and G. P. Pfeifer. 2005. The role of DNA polymerase eta in UV

mutational spectra. DNA Repair (Amsterdam) 4:211–220

8. Cleary JD, Nichol K, Wang YH, Pearson CE (2002) Evidence of cis-acting

factors in replication-mediated trinucleotide repeat instability in primate cells.

Nat Genet 31: 37-46.

9. Collins, N. S., S. Bhattacharyya, and R. S. Lahue. 2007. Rev1 enhances CAG.

CTG repeat stability in Saccharomyces cerevisiae. DNA Repair (Amsterdam)

6:38–44.

10. Cummings, C.J. and Zoghbi, H.Y. (2000) Fourteen and counting: unravelling

trinucleotide repeat diseases. Hum.Mol.Genet., 9 , 909-916.

11. Friedberg, E. C., G. C. Walker, W. Siede, R. D. Wood, R. A. Schultz, and T.

Ellenberger.2005. DNA repair and mutagenesis, 2nd ed. ASM Press,

Washington, DC

12. Fry M, Loeb LA (1994) The fragile X syndrome d(CGG)n nucleotide repeats

form a stable tetrahelical structure. Proc Natl Acad Sci U S A 91: 4950-4954.

13. Fugger, K., Chu, W.K., Haahr, P., Kousholt, A.N., Beck, H., Payne, M.J.,

Hanada, K., Hickson, I.D., and Sørensen, C.S. (2013). FBH1 co-operates with

90

MUS81 in inducing DNA double-strand breaks and cell death following

replication stress. Nat. Commun. 4, 1423.

14. Gacy AM, Goellner G, Juranic N, Macura S, McMurray CT. Trinucleotide

repeats that expand in human disease form hairpin structures in

vitro. Cell. 1995;81:533–540. doi: 10.1016/0092-8674(95)90074-8.

15. Gibbs, P. E., X. D. Wang, Z. Li, T. P. McManus, W. G. McGregor, C. W.

Lawrence, and V. M. Maher. 2000. The function of the human homolog of

Saccharomyces cerevisiae REV1 is required for mutagenesis induced by UV

light. Proc. Natl. Acad. Sci. USA 97:4186–4191.

16. Goodman, M. F. 2002. Error-prone repair DNA polymerases in prokaryotes

and eukaryotes. Annu. Rev. Biochem. 71:17–50.

17. Goula,A.V. and Merienne,K. (2013) Abnormal base excision repair at

trinucleotide repeats associated with diseases: a tissue-selective mechanism.

Genes, 4, 375–387.

18. Hanawalt PC, Spivak G (2008) Transcription-coupled DNArepair: two

decades of progress and surprises. Nat Rev Mol CellBiol 9:958–970

19. Hartenstine MJ, Goodman MF, Petruska J. 2002. Weak strand displacement

activity enables human DNA polymerase beta to expand CAG/CTG triplet

repeats at strand breaks. J. Biol. Chem. 277:41379-41389.

20. Hashimoto, Y., Ray Chaudhuri, A., Lopes, M. & Costanzo, V. Rad51 protects

nascent DNA from Mre11-dependent degradation and promotes

continuousDNA synthesis. Nat. Struct. Mol. Biol. 17, 1305–1311 (2010).

91

21. Jarem DA, Huckaby LV, Delaney S (2010) AGG interruptions in (CGG) (n)

DNA repeat tracts modulate the structure and thermodynamics of non-B

conformations in vitro. Biochemistry 49: 6826-6837.

22. Kee Y, D’Andrea A. 2010. Expanded roles of the Fanconi anemia pathway in

preserving genomic stability. Genes Dev 24:1680–1694.

23. Kim,M.Y., Gleason-Guzman,M., Izbicka,E., Nishioka,D. and Hurley,L.H.

(2003) The different biological effects of telomestatin and TMPyP4 can be

attributed to their selectivity for interaction with intramolecular or

intermolecular G-quadruplex structures. Cancer Res., 63, 3247–3256.

24. Kim N, Jinks-Robertson S (2012) Transcription as a source of genome

instability. Nat Rev Genet 13:204–214

25. Lahue RS, Slater DL (2003) DNA repair and trinucleotide repeat

instability. Front Biosci 8: s653-665

26. Lambert, S., Watson, A., Sheedy, D. M., Martin, B. & Carr, A. M. Gross

chromosomal rearrangements and elevated recombination at an inducible site

specific replication fork barrier. Cell 121, 689–702 (2005).

27. Lawrence, C. W. 2004. Cellular functions of DNA polymerase zeta and Rev1

protein. Adv. Protein Chem. 69:167–203

28. Lehmann, A. R. 2005. Replication of damaged DNA by translesion synthesis

in human cells. FEBS Lett. 579:873–876.

29. Liu, G. and Leffak, M. (2012) Instability of (CTG)n∗(CAG)n trinucleotide

repeats and DNA synthesis. Cell Biosci., 2, 7

92

30. Liu G, Chen X, Bissler JJ, Sinden RR, Leffak M. Replication-dependent

instability at (CTG) × (CAG) repeat hairpins in human cells. Nat Chem

Biol. 2010; 6:652–659.

31. Lopez Castel, A., Cleary, J.D., and Pearson, C.E. (2010). Repeat instability as

the basis for human diseases and as a potential target for therapy. Nat. Rev.

Mol. Cell Biol. 11, 165–170,

32. Masutani, C., R. Kusumoto, A. Yamada, N. Dohmae, M. Yokoi, M. Yuasa, M.

Araki, S. Iwai, K. Takio, and F. Hanaoka. 1999. The XPV

(xerodermapigmentosum variant) gene encodes human DNA polymerase eta.

Nature 399:700–704.

33. McMurray, C.T. (2010). Mechanisms of trinucleotide repeat instability during

human development. Nat. Rev. Genet. 11, 786–799.

34. Mirkin SM (2006) DNA structures, repeat expansions and human hereditary

disorders. Curr Opin Struct Biol 16: 351-358.

35. Mirkin, S.M. (2007).Expandable DNA repeats and human disease. Nature

447, 932–940.

36. Mirkin, E. V., and S. M. Mirkin. 2007. Replication fork stalling at natural

impediments. Microbiol. Mol. Biol. Rev. 71:13–35

37. Mitas, M. 1997. Trinucleotide repeats associated with human diseases.

Nucleic Acids Res. 25:2245–2253

38. Ohmori, H., E. C. Friedberg, R. P. Fuchs, M. F. Goodman, F. Hanaoka, D.

Hinkle, T. A. Kunkel, C. W. Lawrence, Z. Livneh, T. Nohmi, L. Prakash, S.

93

Prakash, T. Todo, G. C. Walker, Z. Wang, and R. Woodgate. 2001. The Y-

family of DNA polymerases. Mol. Cell 8:7–8.

39. Pelletier, R., Farrell, B. T., Miret, J. J., and Lahue, R. S. (2005) Mechanistic

features of CAG*CTG repeat contractions in cultured cells revealed by a

novel genetic assay. Nucleic Acids Res. 33, 5667–5676

40. Petermann, E. & Helleday, T. Pathways of mammalian replication fork restart.

Nat. Rev. Mol. Cell. Biol. 11, 683–687 (2010)

41. Petermann, E., Orta, M. L., Issaeva, N., Schultz, N. & Helleday, T.

Hydroxyurea-stalled replication forks become progressively inactivated and

require two different RAD51-mediated pathways for restart and repair. Mol.

Cell 37, 492–502 (2010).

42. Petruska J, Hartenstine MJ, Goodman MF. 1998. Analysis of strand slippage

in DNA polymerase expansions of CAG/CTG triplet repeats associated with

neurodegenerative disease. J. Biol. Chem. 273:5204-5210.

43. Prakash, S., R. E. Johnson, and L. Prakash.2005. Eukaryotic translesion

synthesis DNA polymerases: specificity of structure and function. Annu. Rev.

Biochem. 74:317–353.

44. Rajeswari MR (2012) DNA triplex structures in neurodegenerative

disorder, Friedreich's ataxia. J Biosci 37: 519-532.

45. Ray Chaudhuri, A., Y. Hashimoto, R. Herrador, K.J. Neelsen, D. Fachinetti,

R. Bermejo, A. Cocito, V. Costanzo, and M. Lopes. 2012. Topoisomerase I

94

poisoning results in PARP-mediated replication fork reversal. Nat. Struct.

Mol. Biol. 19:417–423.

46. Regairaz, M., Zhang, Y.W., Fu, H., Agama, K.K., Tata, N., Agrawal, S.,

Aladjem, M.I., and Pommier, Y. (2011). Mus81-mediated DNA cleavage

resolves replication forks stalled by topoisomerase I-DNA complexes. J. Cell

Biol. 195, 739–749.

47. Samadashwily GM, Raca G, Mirkin SM (1997) Trinucleotide repeats

affect DNA replication in vivo. Nat Genet 17: 298-304.

48. Sarkies P, Reams C, Simpson LJ, Sale JE. Epigeneticinstability due to

defective replication of structured DNA. MolCell 2010;40:703–13.

49. Tateishi, S., H. Niwa, J. Miyazaki, S. Fujimoto, H. Inoue, and M. Yamaizumi.

2003. Enhanced genomic instability and defective postreplication repair in

RAD18 knockout mouse embryonic stem cells. Mol. Cell. Biol. 23:474–481.

50. Ulrich, H.D. 2009. Regulating post-translational modifications of the

eukaryotic replication clamp PCNA. DNA Repair (Amst.). 8:461–469.

51. Usdin K, Woodford KJ (1995) CGG repeats associated with DNA instability

and chromosome fragility form structures that block DNA

synthesis in vitro. Nucleic Acids Res 23: 4202-4209.

52. Watanabe, K., S. Tateishi, M. Kawasuji, T. Tsurimoto, H. Inoue, and M.

Yamaizumi. 2004. Rad18 guides poleta to replication stalling sites through

physical interaction and PCNA monoubiquitination. EMBO J. 23:3886–3896

95

53. Waters LS, Minesinger BK, Wiltrout ME, D’Souza S, Woodruff RV, Walker

GC. 2009. Eukaryotic translesion polymerases and their roles and regulation in

DNA damage tolerance. Microbiol Mol Biol Rev 73: 134–154.

54. Weisman-Shomer P, Cohen E, Fry M (2000) Interruption of the fragile X

syndrome expanded sequence d(CGG)(n) by interspersed d(AGG)

trinucleotides diminishes the formation and stability of d(CGG)(n)

tetrahelical structures. Nucleic Acids Res 28: 1535-1541.

55. Wu, Y., and Brosh, R. M., Jr. (2010) G-quadruplex nucleic acids and human

disease. FEBS J. 277, 3470–3488

56. Yang, W., and R. Woodgate. 2007. What a difference a decade makes:

insights into translesion DNA synthesis. Proc. Natl. Acad. Sci. USA 104:

15591–15598.

57. Zellweger, R., D. Dalcher, K. Mutreja, R. Herrador, M. Berti, A. Vindigni,

and M. Lopes. 2015. Rad51-mediated replication fork reversal is a global

response to genotoxic treatments in human cells. J. Cell Biol. 208:563–579

96