DDX47 and MECP2, two novel human functions controlling R-loop-mediated genome integrity

DDX47 and MECP2, two novel human functions controlling R-loop-mediated genome integrity Esther Marchena Cruz Tesis doctoral Universidad de Sevilla 2021 2 DDX47 and MECP2, two novel human functions controlling R-loop-mediated genome integrity Trabajo realizado en el Departamento de Genética, Facultad de Biología y en el departamento de Biología del Genoma, CABIMER, de la Universidad de Sevilla, con el fin de optar al grado de Doctora en Biología Molecular, Biomedicina e Investigación Clínica por la graduada Esther Marchena Cruz. Sevilla, 2021 La doctoranda: Los directores de tesis: 3 4 TABLE OF CONTENTS RESUMEN ..................................................................................... 14 INTRODUCTION ............................................................................ 18 1. GENOME INSTABILITY ................................................................................................... 20 1.1. DNA damage response ................................................................................................................. 21 1.2. Replication .................................................................................................................................... 21 1.3. Transcription-associated genome instability................................................................................ 22 1.4. Transcription-replication conflicts ................................................................................................ 24 2. R LOOPS AS A SOURCE OF GENOME INSTABILITY ........................................................... 27 2.1. Physiological roles of R loops ....................................................................................................... 28 2.2. R loops as genomic threats .......................................................................................................... 30 2.2.1. R loop-mediated genome instability .................................................................................... 30 2.2.2. R loops and replication stress .............................................................................................. 31 2.2.3. R loops and chromatin ......................................................................................................... 32 3. MECHANISMS AND FACTORS INVOLVED IN PREVENTING R LOOP ACCUMULATION ........ 34 3.1. Topoisomerases ........................................................................................................................... 34 3.2. mRNP biogenesis .......................................................................................................................... 34 3.3. Ribonucleases ............................................................................................................................... 35 3.4. Helicases ....................................................................................................................................... 36 3.5. Other factors ................................................................................................................................ 38 4. TOOLS AND TECHNIQUES TO DETECT R-LOOP ACCUMULATION ...................................... 39 4.1. AID as an indirect tool for R loop-detection ................................................................................. 40 5. rDNA LOCUS ................................................................................................................. 41 OBJECTIVES ................................................................................ 46 RESULTS ...................................................................................... 50 1. SCREENING FOR FACTORS INVOLVED IN R LOOP HOMEOSTASIS ..................................... 52 1.1. A high-throughput screening for factors involved in R loop-mediated genome instability ......... 54 1.2. Network analysis of hits from high-throughput screening ........................................................... 59 1.3. Analysis of spontaneous DNA damage in the selected candidates .............................................. 62 1.4. Analysis of R loop accumulation in the selected hits ................................................................... 63 2. VALIDATION OF DDX47 AND MECP2 AS FACTORS RELATED TO R LOOP METABOLISM AND GENOME INSTABILITY ........................................................................................................ 66 2.1. R loop accumulation in DDX47and MECP2depleted cells ........................................................ 68 2.2. R loop-dependent genome instability phenotype in DDX47 and MECP2 depleted cells ............. 71 3. ROLE OF DDX47 IN R-LOOP METABOLISM AND GENOME INSTABILITY ............................ 74 3.1. DDX47 depletion impairs transcription ........................................................................................ 76 3.2. Transcription-replication conflicts in DDX47-depleted cells ........................................................ 79 3.3. DDX47 physically associates to R loops ........................................................................................ 80 3.4. DEAD/box RNA helicase DDX47 is an RNA-DNA helicase ............................................................. 81 3.4.1. Generation of DDX47 mutants for functional in vitro analysis ............................................ 83 3.4.2. In vitro unwinding analysis with DDX47 proteins ................................................................ 84 3.4.3. DDX47 in vivo activity........................................................................................................... 87 DISCUSSION ................................................................................. 92 5 1. A NEW SCREENING TO INDENTIFY NEW FACTORS RELATED TO R-LOOP MEDIATED GENOME INSTABILITY ........................................................................................................ 94 2. ROLE OF MECP2 IN THE MAINTENANCE OF GENOME INTEGRITY .................................... 98 3. ROLE OF DDX47 IN THE MAINTENANCE OF GENOME INTEGRITY ................................... 102 CONCLUSIONS/CONCLUSIONES ............................................... 110 APPENDIX ................................................................................... 116 MATERIALS AND METHODS ...................................................... 122 1. GROWTH MEDIA AND CONDITIONS ............................................................................. 124 1.1. Bacteria cell culture .................................................................................................................... 124 1.2. Human cell culture ..................................................................................................................... 124 2. ANTIBIOTICS, DRUGS, INHIBITORS, ENZYMES AND ANTIBODIES ................................... 124 2.1. Antibiotics................................................................................................................................... 124 2.2. Inhibitors .................................................................................................................................... 125 2.3. Enzymes and antibodies ............................................................................................................. 125 3. BACTERIA AND HUMAN CELL LINES ............................................................................. 128 3.1. Escherichia coli strains................................................................................................................ 128 3.2. Human cell lines ......................................................................................................................... 128 4. PLASMIDS ................................................................................................................... 129 5. BACTERIAL TRANSFORMATION AND HUMAN CELLS TRANSFECTION ............................. 130 5.1. Bacterial transformation ............................................................................................................ 130 5.2. Human cells transfection ............................................................................................................ 130 5.2.1. siRNA transfection ............................................................................................................. 130 5.2.2. Plasmid transfection using Lipofectamine 2000 or Lipofectamine 3000 ........................... 132 6. siRNA SCREENING METHODS ....................................................................................... 132 6.1. siRNA library preparation ........................................................................................................... 133 6.2. Transfection protocol ................................................................................................................. 133 6.3. Fixing and staining ...................................................................................................................... 134 6.4. Imaging ....................................................................................................................................... 135 6.5. Data analysis and statistical analysis .......................................................................................... 136 7. SITE-DIRECT MUTAGENESIS ......................................................................................... 137 8. DEVELOPMENT OF TOOLS FOR IN VITRO ANALYSIS AND IN VIVO OVEREXPRESSION OF DDX47 ............................................................................................................................. 137 8.1. Tools for in vitro analysis ............................................................................................................ 137 8.2. Tools for in vivo overexpression ................................................................................................. 138 9. IN VITRO ANALYSIS ..................................................................................................... 138 9.1. Purification of DDX47 wild-type and mutant proteins ............................................................... 138 9.2. Nucleic acid unwinding assays ................................................................................................... 139 10. PROTEIN-PROTEIN INTERACTION METHODS ................................................................ 140 10.1. Proximity Ligation Assay (PLA) ................................................................................................... 140 11. TRANSCRIPTION ANALYSIS IN HUMAN CELLS ............................................................... 141 11.1. EU incorporation ........................................................................................................................ 141 12. IMMUNOFLUORESCENCE ............................................................................................ 142 6 13. GENOME INSTABILITY ANALYSIS .................................................................................. 143 13.1. Analysis of γH2AX foci ................................................................................................................ 143 13.2. Alkaline single cell gel electrophoresis (Alkaline comet assay) .................................................. 143 13.3. RNA-DNA hybrids detection ....................................................................................................... 144 13.3.1. RNA-DNA hybrids immunoprecipitation (DRIP-qPCR) ....................................................... 144 13.3.2. S9.6 immunofluorescence ................................................................................................. 145 14. MICROSCOPY IMAGES ACQUISITION, DATA ANALYSIS AND STADISTICAL ANALYSIS ...... 145 14.1. Fluorescence microscopy ........................................................................................................... 145 14.2. Data analysis ............................................................................................................................... 146 14.3. Statistical analysis ....................................................................................................................... 146 15. CHROMATIN IMMUNOPRECIPITATION (ChIP) ASSAY.................................................... 147 16. POLYMERASE CHAIN REACTION (PCR) ANALYSIS .......................................................... 149 16.1. Non-quantitative PCR ................................................................................................................. 149 16.2. Quantitative PCR analysis ........................................................................................................... 149 16.2.1. Reverse Transcription quantitative PCR (RT-PCR) analysis ................................................ 149 16.2.2. Primer pairs used for amplification ................................................................................... 150 17. PROTEIN ANALYSIS ..................................................................................................... 151 17.1. Bacteria cell protein extraction .................................................................................................. 151 17.2. Human cell protein extraction .................................................................................................... 151 17.3. SDS-PAGE .................................................................................................................................... 152 17.4. Western Blot analysis ................................................................................................................. 152 17.5. Non-fluorescence WB ................................................................................................................. 152 17.6. Expression of His6-DDX47-WT and mutant forms ..................................................................... 152 REFERENCES ............................................................................. 156 7 INDEX OF FIGURES INTRODUCTION Figure I1. Transcription-associated genome instability .............................................................. 25 Figure I2. Schematic representation of an R loop structure ....................................................... 27 Figure I3. Physiological role of R loops ....................................................................................... 29 Figure I4. R loop-mediated genome instability ........................................................................... 32 Figure I5. Mechanisms and factors to prevent and resolve R loop accumulation ..................... 37 Figure I6. Direct and indirect tools for R-loop recognition ......................................................... 40 Figure I7. Schematic representation of locus motifs at the human rDNA loci ........................... 43 RESULTS Figure R1. Validation of the U2OS-TR-AID cell line for inducible AID-based R-loop detection .. 55 Figure R2. Candidates of high-throughput screening validation ................................................ 56 Figure R3. Gene Ontology analysis for putative R-loop forming factors from high-throughput screening ..................................................................................................................................... 59 Figure R4. Functional network interaction analysis and selection of hits from high-throughput screening ..................................................................................................................................... 61 Figure R5. Spontaneous damage upon siRNA silencing of selected candidate genes. ............... 62 Figure R6. RNA-DNA hybrids analysis in candidate-siRNA depleted cells by S9.6 immunofluorescence microscopy ............................................................................................... 63 Figure R7. RNA-DNA hybrids analysis in candidate-siRNA depleted cells by DRIP assays .......... 64 Figure R8. siRNA silencing of DDX47 and MECP2 ....................................................................... 68 Figure R9. RNA-DNA hybrids accumulation in siDDX47 and siMECP2 HeLa cells by S9.6 immunofluorescence microscopy ............................................................................................... 69 Figure R10. R loop accumulation in siDDX47 and siMECP2 Hela cells ........................................ 70 Figure R11. DNA damage analysis in siDDX47 and siMECP2 HeLa cells ..................................... 72 Figure R12. DDX47 colocalizes with nucleolin and fibrillarin and is recruited at rDNA .............. 76 Figure R13. Analysis of nucleolus area and RNAPI in DDX47-depleted cells .............................. 78 Figure R14. DDX47 depletion delays 5-ethynyluridine (EU) incorporation in the nucleolus ...... 79 Figure R15. Analysis of conflicts between RNAPI and DNA polymerase in DDX47-depleted cells ..................................................................................................................................................... 80 Figure R16. In situ proximity ligation assay between endogenous DDX47 and RNA-DNA hybrids ..................................................................................................................................................... 80 Figure R17. DDX47 conserved domains and mutations in the helicase core analyzed in this study ............................................................................................................................................ 82 8 Figure R18. Development of an in vitro tool for DDX47-WT and DDX47-mutant protein purification .................................................................................................................................. 83 Figure R19. DDX47 RNA helicase in vitro analysis....................................................................... 85 Figure R20. DDX47 RNA-DNA helicase in vitro analysis .............................................................. 86 Figure R21. Tools for in vivo DDX47 overexpression .................................................................. 87 Figure R22. Tools for specific siRNA depletion of the overexpressed DDX47 ............................ 88 Figure R23. DDX47 overexpression effects in RNA-DNA hybrids ................................................ 89 Figure R24. Analysis of R loop accumulation in SETX-, DDX23-, UAP56-, FANCD2-depleted cells upon DDX47 overexpression ....................................................................................................... 91 DISCUSSION Figure D1. A model to explain the possible role of MeCP2 in genome integrity ...................... 100 Figure D2. Role of DDX47 in rRNA transcription and transcription-replication conflicts ......... 103 Figure D3. A dual role of DDX47 in R loop metabolism ............................................................ 106 MATERIALS AND METHODS Figure M1. Workflow for high-throughput screening of factors involved in R-loop metabolism ................................................................................................................................................... 135 INDEX OF TABLES Table R1. γH2AX foci/cell analysis of validated candidate genes ............................................... 57 Table R2. List of validated hits by γH2AX foci screening. ........................................................... 58 Table M1. Primary antibodies used in this study ...................................................................... 126 Table M2. Secondary antibodies used in this study ................................................................. 128 Table M3. Human cell lines used in this study .......................................................................... 128 Table M4. Plasmids used in this study ...................................................................................... 129 Table M5. siRNAs used in this study ......................................................................................... 131 Table M6. Oligonucleotides for unwinding assays used in this study ...................................... 140 Table M7. DNA primers used in this study ............................................................................... 150 9 APPENDIX Appendix 1. Development of an inducible AID-based R-loop detection cell line ..................... 118 Appendix 2. List of 3205 siRNAs of Dharmacon-ON TARGET Plus-Druggable genome siRNA library ........................................................................................................................................ 118 Appendix 3. γH2AX foci/cell analysis of 3205 siRNA library (First round) ................................ 119 Appendix 4. Identification of high-throughput screening candidates (First round) ................. 119 Appendix 5. γH2AX foci/cell analysis of validated candidates (Validation) .............................. 120 RESUMEN 16 loops in vitro. En paralelo hemos demostrado que la sobreexpresión de DDX47 suprime fenotipos de acumulación de híbridos en células con niveles reducidos de factores previamente implicados en la prevención/resolución de estas estructuras. Nuestros resultados confirman que DDX47 es una helicasa conservada con capacidad para resolver híbridos ARN-DNA in vivo. Esta tesis nos ha permitido identificar a MECP2 y DDX47 como nuevos factores implicados en la homeostasis de los R loops y necesarios para prevenir la inestabilidad genómica dependiente de híbridos de ARN-DNA. Tesis doctoral-Esther Marchena Cruz 17 18 INTRODUCTION 19 INTRODUCTION 20 1. GENOME INSTABILITY The complete instructions required for life are encompassed in the genome and stored as DNA (or deoxyuribonucleic acid) molecule in the cells. To maintain genome integrity and propagate faithfully this information during cell divisions to the off-spring, cells have developed numerous processes that act in a closely coordinated manner. Genome changes may occur due to environmental genotoxic agents and endogenous metabolites action, leading to genetic variability, which although can be harmful for our cells and organism, are crucial for evolution (Aguilera & Gómez-González, 2008). Indeed, genome instability plays a key role in the mechanism of diversification of immunoglobulin genes, providing a large battery of molecules that recognize the antigens. Nevertheless, high levels of genome instability are deleterious for the cell and arise upon genotoxic stress or can be the result of certain pathologies that affect a proper DNA repair and/or replication. Importantly, genome instability is considered one of the hallmarks of cancer, aging, and other genetic diseases and disorders (Aguilera & García-Muse, 2013; Gaillard & Aguilera, 2016; Niedernhofer et al., 2018). DNA integrity is compromised by the action of ultraviolet (UV) radiation, ionizing radiations, numerous chemicals and endogenous metabolites and as consequence of biological processes such as transcription and replication (Hoeijmakers, 2009). Depending on the mechanisms involved, DNA can undergo different type of lesions such as abasic sites, bases mismatch, DNA adducts, interand intra-strand crosslinks, single-stranded DNA (ssDNA) gaps and double strand breaks (DSBs). Consequently, these lesions could be translated into a wide range of genetic alteration: point mutations, microand minisatellite instability (MIN) expansion and contractions, chromosomal instability (CIN), gross chromosomal rearrangements (GCRs), copy number variants (CNVs) of a particular DNA fragment, loss of heterozygosity (LOH) and hyper-recombination (Aguilera & García-Muse, 2013). To minimize detrimental consequences of DNA injury, the stability of genome is supported by multiple pathways focused on detection of DNA lesions, their signalling and subsequently DNA repair to counteract DNA damage and ensure cell proliferation or apoptosis. This set of mechanisms form a complex Tesis doctoral-Esther Marchena Cruz 21 signal transduction pathway known as DNA Damage Response (DDR). When such system fails increase cancer susceptibility (Gaillard & Aguilera, 2016; Jeggo et al., 2016). 1.1. DNA damage response The DNA Damage response (DDR) is a complex signal transduction pathway in which a signal (DNA damage) is detected by a protein sensor that triggers the activation of a transduction system that amplifies and diversifies the signal to protein effectors for an efficient DNA repair (Ciccia & Elledge, 2010). Depending on the specific type of DNA damage, DDR signalling pathway can differ in a wide variety of DNA repair mechanisms. DDR signalling pathways are initiated by the recognition of DNA lesions by the complex MRN (Mre 11 Rad50 Nsb1), that recognizes DSBs and RPA (replication protein A). Then, the upstream kinases, Mec1/ATR (the ataxia telangiectasia and RAD3-related) and Tel1/ATM (ataxia telangiectase mutated), are activated and phosphorylate different substrates. As a consequence of ATR/ATM phosphorylation, a qualitycontrol mechanism that blocks cell cycle to guarantee genome integrity, known as checkpoint, is activated preventing progression from G1 to S phase and from G2 to M phase (Sulli et al., 2012). DDR signalling is spread around the DSB by the phosphorylation of the histone variant γH2AX on Ser139. This mark is required to recruit MDC1, (mediator DNA damage checkpoint 1), that maintains DDR signalling by enforcing accumulation of the MRN complex and ATM activation at DSBs sites (van Attikum & Gasser, 2009). Other factors as BRAC1 and 53BP1 (p53-binding protein 1) are also recruited to DSB sites. Downstream kinases CHK2 and CHK1, mainly phosphorylated by ATM and ATR respectively, are activated and the signalling pathway converges on downstream effectors to coordinate different outcomes: transient cell cycle arrest to repair DNA damage and cell death by apoptosis or cellular senescence (Jackson & Bartek, 2009; Sulli et al., 2012). Otherwise, a long-term state of global genome instability is a characteristic of almost all human cancers (Niedernhofer et al., 2018). 1.2. Replication Propagation of the genetic information to the off-spring requires timely and accurate duplication of DNA prior to cell division. The replication fork (RF) must travel along the DNA to accomplish its function. During replication the RF has to INTRODUCTION 22 pass through the chromatinized DNA template and must overcome frequent obstacles such as DNA damage, protein barriers, torsional stress, heterochromatin, non-B DNA structures or transcription machinery itself, that will hamper replication fork progression (Gómez-González & Aguilera, 2019). Indeed, these obstacles are an important source of replication stress and genome instability since encountering an obstacle can cause RF stalling, which turns into an irreversible collapse of the synthesis, causing replisome disassembly and generating hazardous DSBs and ssDNA (Aguilera & García-Muse, 2013; Aguilera & Gómez-González, 2008). In general, any condition that compromises replication it is referred to as replication stress. The surveillance mechanisms that prevent genome instability upon replication stress are mitotic and S phase checkpoint pathways to guarantee replication completion and to prevent replication fork breakage (Gaillard et al., 2015). 1.3. Transcription-associated genome instability Transcription is an essential cellular process in all living cells for proliferation and survival, in which the genetic information encompassed in form of DNA is copied into new RNA molecules that can either play their function directly or be translated into proteins. During transcription, the two complementary DNA strands are separated to allow the RNA polymerase (RNAP) to copy one of the strands as a template into RNA. Specifically, in the transcription bubble, the transcribed strand (template) generate a complementary RNA chain, forming an RNA-DNA hybrid, whereas the non-transcribed strand remains unpaired as ssDNA (Gaillard et al., 2013). During RNAP journey across a gene, the organization and integrity of the genome are transitory affected by the positive and negative DNA supercoiling and chromatin remodelling changes, to allow the progression of the RNAP through the DNA template (Selth et al., 2010). Due to the fact that DNA molecule is the substrate of transcription, this process creates the optimal conditions for high levels of mutations (transcription associated mutation, TAM), recombination (transcription associated recombination, TAR) as well as DSBs, chromosome rearrangements and chromosome loss by making chromatin more susceptible to DNA insulting agents. Both TAM and TAR are conserved processes from prokaryotes to higher eukaryotes (Aguilera, 2002). Tesis doctoral-Esther Marchena Cruz 23 In the 70s by two different laboratories, transcription-associated mutagenesis (TAM) was first reported in Escherichia coli. Both studies showed that transcription activation in presence of mutagenic agents increases reversion rates of point mutations in transcribed genes (Brock, 1971; Herman & Dworkin, 1971). Later works suggested that high levels of transcription impact differently mutagenesis of the non-transcribed strand (NTS) versus the transcribed strand (TS). The analysis of mutation of both DNA strands (NTS and TS) showed that mutations along transcribed regions occur preferentially in NTS (Beletskii & Bhagwat, 1996, 1998). TAM has been evidenced in a wide variety of assays and systems to sense mutations from prokaryotes to eukaryotes (Jinks-Robertson & Bhagwat, 2014). One of the most frequent mutations are C-to-T transitions, caused by spontaneous cytosine deamination to uracil conversion. If uracil is not removed, it is paired with adenine, resulting into C-to-T mutations after replication (Jinks-Robertson & Bhagwat, 2014). In this thesis we have used the overexpression of AID an human cytosine deaminase to increase genome instability as we will explain bellow. In the case of TAR, numerous works based on systems to sense ectopic recombination have shown the transcription can lead to hyper-recombination and chromosomal rearrangements (Gaillard et al., 2013). Studies in bacteria, yeast and mammals indicate that as a consequence of transcription, TAR is enhanced by making DNA more susceptible to undergo recombinogenic DNA damages and more accessible to the enzymatic activities required for recombination directed by site-specific recombinases. Moreover, TAR is necessarily linked to replication. Several studies showed that when RF (replication fork) progression was opposed to transcription, TAR occurred at high levels (Gaillard & Aguilera, 2016; Hamperl & Cimprich, 2016). Genome instability associated with transcription can be due to several factors. During transcription, the chromatin is subjected to topological changes to allow the movement of the transcription machinery, which favoured negative and positive supercoiling accumulated behind and ahead of the elongating RNAP, respectively. Such changes can facilitate the transient formation of ssDNA that is more vulnerable to damage (Figure I1A). According that, mutants of topoisomerases accumulate torsional stress at the transcribed genes and favour INTRODUCTION 24 transient underwound DNA that might lead to stretches of ssDNA behind RNAP resulting in transcription-associated genome instability in different ways. These stretches could be damaged due to their susceptibility to spontaneous base modifications, genotoxic agents or enzymatic activities that could cause ssDNA breaks and base damage. This particular scenario, when DNA unwinding occurs during replication, transcriptional or recombination, also encourages non-B DNA structures formation in different forms as hairpins, G-quadruplexes (G4) DNA or RNA-DNA hybrids (known as R loops when formed outside the transcription bubble) (Figure I1B). It is worth to notice that genome instability associated with these structures is due to their relationship with replication, by blocking replisome (Aguilera & García-Muse, 2013), or impeding transcription by the accumulation of stalled RNAPs (Belotserkovskii et al., 2013) which in turn can serve as a signal for the machinery of repair called transcription-coupled excision repair (TC-NER) in the absence of DNA lesions (Hanawalt & Spivak, 2008), undergoing genome instability at risk in all cases (Gaillard & Aguilera, 2016). 1.4. Transcription-replication conflicts Within the numerous obstacles to replication fork progression, transcription is likely the main source of replicative impairments since the complexity of transcription machinery, together multiple mRNA processing steps and mRNP assembly factors are a challenge to advancing replication fork. Although there is a temporal or spatial separation between transcription and replication (for most of the genes) both processes use the same DNA as template, thus unavoidable encounters on the same DNA region at the same time can be occur, causing transcription-replication conflicts (Figure I1C) (Gómez-González & Aguilera, 2019; Hamperl & Cimprich, 2016; Oestergaard & Lisby, 2017). Tesis doctoral-Esther Marchena Cruz 25 Figure I1. Transcription-associated genome instability. (A) The progression of transcription promotes the formation of local negative supercoiling behind the RNAP and positive supercoiling ahead the RNAP that make the DNA more vulnerable to damage. (B) In the absence of certain factors, non-B structures such as G4-quadruplexes, RNA-DNA hybrids or R loops are produced. The nascent RNA transcript hybridizes with its DNA template strand and displaces NTS remaining as ssDNA, that is more susceptible to the action of genotoxic agents, including reactive oxygen species (ROS), nucleases and other modifying enzymes such as AID. (C) Encounters between replication and transcription machineries cause transcription-replication collisions. Depending on the orientation to the replication origin can occur in a head-on orientation (if both machineries progress in the opposite direction) which are more deleterious than codirectional orientation (when both machineries advance in the same direction). MCM, minichromosome maintenance complex; DNA pol, DNA polymerase; RF, Replication Fork; RNAP, RNA polymerase; NTS, nontranscribed strand. Figure adapted from (Gaillard et al., 2013). DNA replication and transcription exhibit different rates favouring these conflicts to occur (Bermejo et al., 2012; Helmrich et al., 2013). Depending on the orientation to the replication origin of a particular relative gene, collision between transcription and replication machineries can occur in a head-on orientation (genes transcribed from the lagging strand template) or a codirectional orientation (genes transcribed from the leading strand template). Experimental data suggest that head-on collisions seem to be more deleterious and constitute a stronger INTRODUCTION 32 regulator of R loops levels, whereas in the co-directional orientation reduce Rloops levels, in head-on orientation promote their formation (Hamperl et al., 2017; Lang et al., 2017). However, it is an unsolved question whether an R loop is sufficient to provoke RF stalling or if a more complex chromatin structure is involved (Castellano-Pozo et al., 2013; García-Pichardo et al., 2017; Rondón & Aguilera, 2019). Figure I4. R loop-mediated genome instability. (A) As a consequence of R loop formation, the non-transcribed strand that remains unpaired as displaced ssDNA molecule can be more susceptible to the attack of different genotoxic agents or enzymatic activities such as AID. This could lead to DNA lesions such as base damage (red star) or ssDNA gaps, which can block the progression of replication forks, generating genome instability that can ultimately lead to cell death or cancer. (B) Directly, R loops by themselves could act as an obstacle of replication forks or indirectly when a stable R loop force the RNAP to remain attached at the transcription site promoting transcription-replication conflicts. (C) Chromatin compaction as a consequence of R loop formation could also interfere with replication. In all this cases, R loop could hamper RF progression, leading to RF stalling, potential collapse and breakage generating genome instability. Adapted from (Gaillard & Aguilera, 2016). 2.2.3. R loops and chromatin A nascent connection between chromatin changes and R loops homeostasis have recently emerged as possible cause of R-loop associated genome instability (Chédin, 2016). Histone hyper-acetylation has been proposed to facilitate R loop accumulation due to the fact that the high chromatin accessibility facilitates the probability of R loop generation. In fact, it has been shown that depletion of the Tesis doctoral-Esther Marchena Cruz 33 histone deacetylase mSin3a complex or the acetyltranferase MOF or deacetylation inhibition by chemical compounds (TSA and SAGA) causes R loop accumulation and genome instability (Salas‐Armenteros et al., 2017; Singh et al., 2018; Wahba et al., 2011). In contrast, aberrant R loop accumulation have been associated with a mark of chromatin condensation (H3S10-P), in yeast, C. elegans and human cells (Castellano-Pozo et al., 2013; García-Pichardo et al., 2017). In particular, depletion of THO complex or the helicase SETX/Sen1 leads to an accumulation of H3S10-P in yeast and human cells. These observations open the possibility that local chromatin compaction triggered by R loops contributes to replication fork impairment or to delayed or less-efficient replication initiation as possible causes of R-loop dependent genome instability (Figure I4C) (Castellano-Pozo et al., 2013). Moreover, mutants of the chromatin-reorganizing complex FACT (facilitates chromatin transcription) show high levels of R loops and R-loop mediated transcription-replication conflicts in yeast and human cells. This phenotype is likely linked with a specific chromatin organization since this complex is involved in nucleosome disassembly around the RNA polymerase during transcription elongation (Herrera-Moyano et al., 2014). To understand the biological meaning of R-loop chromatin signatures, recent screening studies in yeast have been performed where it has been identified specific histone mutants that accumulate R loops but do not cause genome instability. Thus, R loops do not cause DNA damage by themselves, but requiring a two-step mechanism in which, first, an altered chromatin prone to R loops, and second, a modified chromatin that including H3S10-P for compromising genome integrity (GarcíaPichardo et al., 2017). Histone H3 acetylation, Histone 3 lysine 4 (H3K4) monoand di-methylation and tri-methylation of H3K36 are also significantly enriched over R-loops prone sites, linking R-loop-forming regions with open chromatin and high RNA polymerase occupancy (P. B. Chen et al., 2015; Sanz et al., 2016). Another piece of evidence between R loops and chromatin is a recent work that demonstrates a key role of the conserved chromatin remodeler complex SWI/SNF on the maintenance of genome instability by helping resolve R-loopmediated T-R conflicts (Bayona-Feliu et al., 2021). INTRODUCTION 34 3. MECHANISMS AND FACTORS INVOLVED IN PREVENTING R LOOP ACCUMULATION Cells have developed mechanisms to control R loop homeostasis and mitigate their harmful effect if deregulated (Chédin, 2016; Santos-Pereira & Aguilera, 2015; Skourti-Stathaki & Proudfoot, 2014). In general, two types of mechanisms or factors have been reported: those that help prevent R loop formation (mRNP biogenesis and topoisomerases) and those which remove R loops (ribonucleases, helicases and others). 3.1. Topoisomerases Transcription induces local topological changes, whereas positive supercoiling (overwinding) ahead, negative supercoiling (underwinding) is located behind the advancing RNAP. The advance of replication machinery also leads to transient negative supercoiling accumulation. This more open DNA can favour the hybridization of the nascent RNA with the DNA template and the formation of R loop structures. Supporting this idea topoisomerases, enzymes required to resolve or relax negative and positive supercoils respectively, also prevent R loop accumulation (Gaillard & Aguilera, 2016; Santos-Pereira & Aguilera, 2015). Indeed, yeast topoisomerases mutants show R loop accumulation at rDNA locus (El Hage et al., 2010) and human TOP1-deficient cells accumulate stalled replication forks and chromosome breaks in an R-loop dependent manner (Tuduri et al., 2009). Recently, genome-wide analysis indicate that toposiomerase I prevents transcription-replication conflicts at R-loop-enriched transcription termination sites (TTS) (Promonet et al., 2020). Therefore, topoisomerases are considered as crucial factors involved in preventing R-loop formation during transcription (Figure I5A). 3.2. mRNP biogenesis During eukaryotic gene expression, nascent mRNA need to be processed and correctly packaged to export a properly assembled and integrated messenger ribonucleoparticle (mRNP) to the cytoplasm. The different steps of mRNP biogenesis are linked from transcription, mRNA processing to mRNP export and can influence each other (Bentley, 2014; Proudfoot et al., 2002). mRNP biogenesis is not only important for gene expression, but also for the maintenance of genome integrity. Thus, a wide variety of RNA-binding proteins (RBPs) are Tesis doctoral-Esther Marchena Cruz 35 associated with the nascent mRNA, promoting completely its package and protection and also reducing the capacity to hybridize back with the DNA template (Figure I5A) (Rondón et al., 2010). The connection between mRNA biogenesis and R loops was first shown in yeast mutant of THO/TREX, a conserved complex involved in the coupling of transcription with mRNP biogenesis and export (Huertas & Aguilera, 2003). Cells lacking a functional THO complex accumulate RNA-DNA hybrids and show hyper-recombination and DNA breaks that are R loop dependent (Castellano-Pozo et al., 2012; Chávez & Aguilera, 1997; Domínguez-Sánchez et al., 2011; Gómez-González et al., 2011; Huertas & Aguilera, 2003). Importantly, genome instability of THO mutants is associated with an increase in transcription-replication conflicts (Gómez-González et al., 2011; Salas‐Armenteros et al., 2017; Wellinger et al., 2006). Furthermore, the connection between transcripcion/mRNP biogenesis/export and genome instability is extended to other factors involved in splicing, SRF1 (serine/arginine splicing factor 1) (X. Li & Manley, 2005); in mRNA 3´ end processing and degradation (Figure I5A) (Gavaldá et al., 2013; Luna et al., 2005; Pefanis et al., 2015; Stirling et al., 2012); mRNP biogenesis and export such as THSC/TREX2 among others (González-Aguilera et al., 2008). Nowadays, several global and specific studies have evidenced more factors involved in RNA metabolism that have an important association with R loops accumulation (Chan et al., 2014; Paulsen et al., 2009; Stirling et al., 2012; Wahba et al., 2011), thus highlighting that the structure and fate of the nascent mRNA are crucial for R-loop prevention. Nevertheless, there are too few studies about molecular mechanisms underlying R loop homeostasis, and not all factors protect the genome in an R-loop-dependent manner. Thus for example, certain splicing factors could contribute to genome integrity indirectly by regulating gene expression (Salas-Armenteros et al., 2019). 3.3. Ribonucleases Apart from the active prevention of R loops, cells also possess diverse mechanisms actively involved in removing these structures once formed. First of all, the RNase H enzymes specifically degrade the RNA moiety of RNA-DNA hybrids (Figure I5B). There are two types of RNase H enzymes: RNase H1, which is monomeric, removes RNA-DNA hybrids like primers in Okazaki fragments and INTRODUCTION 36 RNase H2, which is monomeric in bacteria but is composed of three subunits in eukaryotes, is involved in ribonucleotide excision repair (Cerritelli & Crouch, 2019). Despite, their different structure and substrate specificity, both type are able to resolve RNA-DNA hybrids (Cerritelli & Crouch, 2009; Skourti-Stathaki & Proudfoot, 2014). RNase H1 is the main player in removing co-transcriptional R loops (Chon et al., 2013). Indeed, RNase H1 overexpression is widely used to remove R loops and to supress R-loop dependent genome instability phenotypes. 3.4. Helicases R loops can be also removed by RNA-DNA helicases through unwinding these structures (Figure I5B). Helicases are essential enzymes that use ATPdriven motor force to unwind DNA or RNA duplex substrates and are involved in different aspects of nucleic acid metabolism: DNA replication, repair, transcription and recombination (Singleton et al., 2007; Valentini & Linder, 2021). The yeast protein Sen1 and its human homologue senataxin (SETX) are extensively studied examples of proteins implicated in R-loop homeostasis (Figure I5B). In S. cerevisiae, Sen1 helicase-inactive sen1-1 mutants show R loops accumulation and hyper-recombination phenotype (Mischo et al., 2011). In human cell, SETX depletion causes R-loop accumulation at transcription termination pause sites, suggesting that this protein is required to resolve R loops, in particular during transcription termination process (Skourti-Stathaki et al., 2011). Other welldescribed examples of helicases related to R loops homeostasis are DHX9 (Cristini et al., 2018), DDX1 (L. Li et al., 2016), DDX23 (Sreerama Chaitanya Sridhara et al., 2017), DDX21 (Song et al., 2017), DDX19 (Hodroj et al., 2017), UAP56 (Pérez-Calero et al., 2020), DDX5 (Sessa et al., 2021) among others. However, in most of the cases, it is unclear whether these RNA helicases have a direct role in unwinding RNA-DNA hybrids or whether they have the ability to reduce R-loop accumulation based on their activity as RNA chaperones or RNPs disassemblers. Tesis doctoral-Esther Marchena Cruz 37 Figure I5. Mechanisms and factors to prevent and resolve R loop accumulation. (A) Topoisomerase 1 (TOP1) avoid the local negative supercoiling accumulation behind the elongating RNAP. Specific RNA binding proteins involved in different steps of RNA metabolism from RNA biogenesis (THO complex, UAP56, SRSF1 and Pcf11) to RNA surveillance (exosome, TRAMP) prevent the formation of R loops. (B) RNase H enzymes can remove R loops, degrading RNA moiety. Moreover, helicases as SETX and UAP56 could unwind the RNA-DNA hybrids within the R loop. (C) DNA repair factors as BRAC1, BRAC2, TC-NER factors (XPF, XPG) and potentially other Fanconi Anemia (FA) proteins can help counteract this situation. (D) Chromatinassociated factors such as the FACT, the histone deacetylase mSin3a and SWI/SNF complexes have a role in prevention of R loop-mediated genome instability. INTRODUCTION 38 3.5. Other factors In addition to direct players in R loop homeostasis, there are many factors that resolve other forms of stresses that may indirectly contribute to resolve R loops. Importantly, most of them have a role in DNA repair and/or in mediation of transcription-replication conflicts. For instance, this is the case of the transcription-coupled nucleotide excision repair (TC-NER) factors that may contribute to mitigate R-loop-induced DNA damage and instability. Specifically, it is proposed that TC-NER nucleases XPG and XPF could excise R loops, being actively processed into DSBs (Figure I5C) (Sollier et al., 2014; Yasuhara et al., 2018). The DSB repair factors and tumor suppressors BRAC1 and BRAC2 have been shown to play a role regulating transcription elongation by RNA Polymerase II to prevent R-loop accumulation and help to suppress R-loop-mediated transcriptional stress (Figure I5C) (Bhatia et al., 2014; Shivji et al., 2018; X. Zhang et al., 2017). Furthermore, it is proposed that BRAC1 recruits SETX to remove R loops at termination sites (Hatchi et al., 2015). Cells lacking Fanconi Anemia factors show an R loop accumulation and genome instability phenotype (M. L. García-Rubio et al., 2015; Hatchi et al., 2015; Schwab et al., 2015). Chromatin related factors are also relevant in R loop homeostasis. Thus the histone deacetylase Sin3A histone deacetylase complex interacts with THO/UAP56 and suppresses co‐transcriptional R‐loops (Luna et al., 2019; Pérez-Calero et al., 2020; Salas‐Armenteros et al., 2017). Different works have shown that other chromatin-associated factors as FACT, and SWI/SNF complexes play a role in the resolution of R-loop-mediated transcriptionreplication (Figure I5D) (Bayona-Feliu et al., 2021; Herrera-Moyano et al., 2014). Tesis doctoral-Esther Marchena Cruz 39 4. TOOLS AND TECHNIQUES TO DETECT R-LOOP ACCUMULATION A main issue of R loop biology is the development of strategies to detect these structures in vivo (Figure I6A). Electron microscopy (EM) and physical analysis of nucleic acid resistant to RNase A and sensitive to RNase H are direct methods used for R loop detection (Drolet et al., 1995; Duquette et al., 2004; Huertas & Aguilera, 2003). Moreover, the most widely direct tool used is the S9.6 monoclonal antibody that detects RNA-DNA hybrids, which originally was isolated by Carrico and colleagues in 1985. Specifically, Mouse (BALB/c) B-cell was fused with Mouse (BALB/c) Sp2/0-Ag14 myeloma to produce monoclonal antibodies against RNA-DNA hybrids. Animals were immunized with RNA-DNA hybrids formed by using single stranded PhiX174 DNA as template for E. coli DNA dependent RNA polymerase (Boguslawski et al., 1986). RNA-DNA hybrids immunoprecipitation from digested DNA (DRIP) and immunofluorescence (IF) using the monoclonal antibody are methods commonly used to capture RNADNA hybrids (M. García-Rubio et al., 2018). However, recent reports have revealed that S9.6 antibody has ability to bind also to dsRNA (Hartono et al., 2018; Silva et al., 2018). As a consequence, indirect strategies have been developed to overcome S9.6 limitations and detect the existence of R loop accumulation including the use of bisulfite which converts specifically cytosines into uracils of the ssDNA displaced by the RNA-DNA hybrid (Ginno et al., 2012; Yu et al., 2003), the use of the hybrid-binding domain of the RNase H enzyme fused either to GFP (HBD-GFP) (Bhatia et al., 2014) or to MNase (Yan et al., 2019) and the use of an catalytically inactive RNase H in chromatin immunoprecipitation (ChIP) which is bound to the RNA-DNA hybrids but without resolving them (Liang Chen et al., 2017). It worth to notice that all these techniques are applied for genome-wide detection of R loops (Liang Chen et al., 2017; Dumelie & Jaffrey, 2017; Ginno et al., 2012; Sanz et al., 2016; Sanz & Chédin, 2019). Complementarily, the suppression of R loop-dependent phenotypes by in vivo RNase H overexpression or in vitro RNase H and RNase III treatment are essential tools to decipher if the signal measured comes from RNA-DNA hybrids or not. INTRODUCTION 40 Figure I6. Direct and indirect tools for R-loop recognition. (A) RNA-DNA hybrids can be detected with the S9.6 antibody, GFP fused to an inactive RNaseH or the RNA binding domains (HB) of the RNase H. Induced-mutagenesis with bisulfite in vitro or with activation-induced cytidine deaminase (AID) in vivo are used to detect the displaced ssDNA within the R loop. (B) Schematic illustration of the AID-based principle for detecting R-loop dependent DNA damage. AID deaminate cytosines of ssDNA converting them into uracil, whose repair processing by base excision repair (BER) and mismatch repair (MMR) leads to DSBs. Figure adapted from (GarcíaMuse & Aguilera, 2019). 4.1. AID as an indirect tool for R loop-detection There is another tool to indirectly detect R loops based on the expression of the human cytosine deaminase AID (activation-induced cytidine deaminase) (Figure I6B). This enzyme acts on ssDNA and deaminate cytosines of ssDNA turning them into uracils. AID is required to initiate the processes of somatic hypermutation (SHM) and class switch recombination (CSR) in the immunoglobulin genes required for antibody diversification (Chaudhuri & Alt, 2004). The R loop formation behind RNP provides the substrate to the AID enzyme (Yu et al., 2005). Nevertheless, beyond its physiological activity, AID can erroneously act on a subset of off-target genes in a transcription-dependent manner, particularly in R loop prone accumulating genes, promoting chromosome translocations and mutations prone to activate oncogenes (Hendriks et al., 2010; Ruiz et al., 2011). Tesis doctoral-Esther Marchena Cruz 41 Heterologous expression of human AID was first used in an study in yeast with mutants of the THO complex (Gómez-González & Aguilera, 2007). A strong and transcription-dependent hypermutation overall in the non-transcribed strand and hyperrecombination was induced by AID in these mutants validating the accumulation of R loop structures in the absence of these mRNP factors. Later on, the ectopic expression of AID has been reported in yeast and human cells as a tool to increase R loop-dependent genome instability (García-Benítez et al., 2017; García-Pichardo et al., 2017; Gómez-González & Aguilera, 2009; Mischo et al., 2011). These approaches are based on the ability to AID to act on the ssDNA of an R-loop structure deaminating Cs turning them into uracils that finally lead to DSB by the action of other enzymes and as a consequence, causing AIDmediated genomic instability (Figure I6B) (Aguilera & Gómez-González, 2008; Basu et al., 2011; Gómez-González & Aguilera, 2007; Ruiz et al., 2011). Taken advantage of the previous experience in our laboratory of the use of AID to exacerbate R-loop genome instability associated phenotypes, in this thesis, we have used a system based on AID expression to identify new factors involved in R loop-homeostasis in human cells. 5. rDNA LOCUS Genome-wide sequencing of hybrid-harboring loci have revealed a higher R loop distribution in retrotransposons, telomeres and highly expressed genes, such as the ribosomal RNA and tRNA loci. Specifically, in Saccharomyces cerevisiae it has been described that 50% of all mapped R loop are in the rDNA locus in wildtype cells (Wahba et al., 2016). Similar to yeast, R-loops accumulate at repetitive sequences such as transposable elements, ribosomal DNA, centromeres and telomeres in human cells (Ginno et al., 2012). In this section we will summarize some characteristics of the transcription, processing and transcription associated-genome instability of the human rDNA locus. rDNA genes are arranged in clusters of multiple tandem repeats (150-200 in yeast and up to 350 in humans) and symbolize the most highly transcribed genome locus. The chromosomal landmarks that contain rRNA repeats are OBJECTIVES 48 The main goal of this thesis is to further explore the wide variety of factors that could contribute to induce R loop-dependent genome instability in human cells. For this purpose, we addressed the following specific objectives: 1. To carry out a high-throughput screening to identify new factors involved in R loop-dependent genome instability in human cells. 2. To validate and characterize specific hits: MECP2 and DDX47 as novel Rloop related factors. 3. To analyze the role of DDX47 in the maintenance of genome integrity, with the aim of getting new insights into the molecular mechanism of this helicase to prevent genome instability. 49 50 RESULTS 51 RESULTS 52 1. SCREENING FOR FACTORS INVOLVED IN R LOOP HOMEOSTASIS Tesis doctoral-Esther Marchena Cruz 53 RESULTS 54 With the aim at screening for factors involved in R loop-homeostasis we have developed in our laboratory a system based on the action of the human cytidine deaminase AID on single strand DNA (ssDNA), particularly on the displaced DNA strand of an R loop, deaminating Cs that finally leads to DNA double strand breaks (DSBs) by the action of other enzymes (Figure R1A) (Aguilera & GómezGonzález, 2008; Basu et al., 2011). The system, that we referred from now on, AIRD (AID inducible R loop detection cell line), consists of a stable U2OS cell line carrying a tetracycline-regulated human cytidine deaminase AID gene cassette (Figure R1B). We expected that cells accumulating high-levels of R loops would generate high levels of DNA breaks that could be detected by immunofluorescence as γH2AX foci. In order to get further insight into the connection between RNA-DNA hybrids and genome integrity, in this thesis we have used the AIRD system as a tool to perform an siRNA high-throughput microscopy screening by H2AX immunofluorescence and look for new factors related to R-loop-mediated genome instability. 1.1. A high-throughput screening for factors involved in R loop-mediated genome instability To validate AIRD system, we measured DNA damage by γH2AX immunofluorescence in cells depleted of FANCD2, a factor involved in R loop metabolism, with and without AID expression (M. L. García-Rubio et al., 2015; Schwab et al., 2015). As previously reported, transient depletion of FANCD2 in HeLa cells leads to a higher number of γH2AX foci per cell in comparison with siC control cells. Importantly, a significant increase of DNA damage could be observed in siFANCD2 after AID induction (+DOX), and to less extent in siC, compared to their respective non-induced conditions (-DOX) (Figure R1C). Furthermore, in collaboration with Jose Calderón in Aguilera’s lab, the AIRD system was validated for FANCD2 and other factors such as THOC1, an mRNP protein that prevents R loop formation (Domínguez-Sánchez et al., 2011), in 96well plates and automated microscope in order to set up the conditions for a highthroughput screening (Appendix 1). Tesis doctoral-Esther Marchena Cruz 55 Figure R1. Validation of the U2OS-TR-AID cell line for inducible AID-based R-loop detection system (AIRD). (A) Schematic illustration of the AID-based principle for detecting factors involved in R-loop metabolism. (B) U2OS-TR-AID stable cell line carrying a tetracycline-regulated AID gene cassette is used as a tool to induce an increase of DNA double-strand breaks (DSBs) in an R-loop dependent manner. (C) Immunofluorescence of γH2AX in siC and siFACND2transfected U2OS-TR-AID cells (clone #8) expressing (+DOX) or not expressing AID (-DOX) as indicated. Nuclei were stained with DAPI. Scale bars, 25 μm. Quantification of the number of γH2AX foci per cell is shown. The red line indicates the median (n=3). The statistical significance of the difference was calculated with Mann-Whitney U-test; ***P<0.001. In collaboration with Sonia Silva, Lola P. Camino and Jose Javier Marqueta-Gracia, we performed an siRNA high-throughput screening with the AIRD system and analyzed a collection of siRNAs targeting 3205 human genes involved in different processes, as apoptosis, nucleic acid binding, autophagy and others, that are considered potential targets for therapeutic drugs (3205 out of 4796 siRNAs of Dharmacon-ON TARGET Plus-Druggable genome siRNA library) (Appendix 2). We prepared 96-well plates containing a duplicate of each four-siRNA pool for every targeted gene, and siRNA against FANCD2 and a nontargeting siRNA (siC) were included as positive and negative control respectively. RESULTS 56 For the screening, U2OS-TR-AID cells were transfected, treated or not with doxycycline (DOX) to induce AID expression and γH2AX detection was acquired by automated microscopy of cells with IMAGE-Express (137239) equipment available at Genomic Unit at CABIMER (Materials and Methods 6; Figure M1; Figure R2A). We reasoned that comparison of γH2AX signal before and after expression of AID in siRNA transfected cells would allow for the identification of factors involved in R-loop metabolism. The screening was performed in duplicate, and siRNAs whose depletion lead to an increase (≥1.2) in the percentage of cells with γH2AX foci upon AID induction (siRNA +DOX), versus non-induced conditions (siRNA -DOX), that was higher (≥1.2) to that observed in control cells (siC +DOX) were selected as candidates. With these criteria, 156 out of 3205 siRNAs tested were chosen for further γH2AX immunofluorescence analysis (Figure R2B; Appendix 3 and 4). These candidates were analyzed in a second round and 46 hits were confirmed, taking into account the increase in the percentage of cells with γH2AX foci and the reproducibility between duplicates of the two different experiments (Figure R2C; Table R1; Appendix 5). Figure R2. Candidates of high-throughput screening validation. (A) Schematic workflow for testing the human druggable siRNA library (3205 siRNAs). (B) Scheme of hits obtained in the different steps of the screening. (C) Results of the second round of γH2AX immunofluorescence screening analysis. The percentages of cells with ≥5 or ≥10 γH2AX foci was used to calculate the indicated ratios, represented in both axis: Y, siRNA with AID versus itself without AID expression; X, siRNA with AID versus the median of siC with AID expression. The dotted lines show the cutoff (1.2) used to designate a siRNA as positive. Two ratios for each siRNA are represented by a dot: black dots: candidates with ≥10 γH2AX foci per cell; red dots: candidates with ≥5 γH2AX foci per cell; grey dots: others. Tesis doctoral-Esther Marchena Cruz 57 Table R1. γH2AX foci/cell analysis of validated candidate genes. Data show the ratios between percentages of cells with ≥5 or ≥10 γH2AX foci with and without AID expression upon siRNA depletion of the 46 candidate genes. The two ratios for each siRNA consist on: one versus itself with and without AID expression (siX+DOX/siX), and the other one versus the median of siC with AID expression in its plate (siX+DOX/siC+DOX). The selected hits (46 genes) defined as those with both ratios ≥ 1.2, at least in two duplicates of one replicate. RESULTS 64 To confirm further the accumulation of RNA-DNA hybrids, we performed RNA-DNA immunoprecipitation (DRIP) in two human genes (APOE and RPL13A) that have been previously validated for R-loop detection (Ginno et al., 2013; Herrera-Moyano et al., 2014). In vitro treatment with RNAse H was used to confirm the specificity of immunoprecipitation of DNA-RNA hybrids. Depletion of most of the candidates (6 out 7) increased the DRIP signal respect to the siC levels, in at least one of the studied genes (Figure R7). Then, we considered as the top-hit candidates those genes in which an accumulation of RNA-DNA hybrids was detected by both, S9.6 IF and DRIP analysis (relative values above 1.5 of siC levels) (MECP2, DDX42, DDX47, MYOG) (Figure R6 and R7). Figure R7. RNA-DNA hybrids analysis in candidate-siRNA depleted cells by DRIP assays. Relative DRIP-qPCR signal values at APOE and RPL13A genes in U2OS cells transfected with the indicated siRNAs and samples were treated in vitro with RNase H prior immunoprecipitation where indicated. Red lines indicate the regions where the primers used for PCR amplification were located. Data are plotted as mean ± SEM (n = 3). *P < 0.05; **P < 0.01; ***P < 0.001 (onetailed paired t-test). We decided to focus our study in two out of the top-hits R-loop accumulating candidates validated by S9.6 IF and DRIP that showed different activities or properties on the nucleic acids metabolism: DDX47, a nucleolar helicase, that showed the highest values in the DRIP analysis (Figure R7) and a DNA binding protein, MECP2, the methyl-CpG binding protein 2, being the candidate with the highest reproducibility (Figure R4B). Tesis doctoral-Esther Marchena Cruz 65 RESULTS 66 2. VALIDATION OF DDX47 AND MECP2 AS FACTORS RELATED TO R LOOP METABOLISM AND GENOME INSTABILITY Tesis doctoral-Esther Marchena Cruz 67 RESULTS 68 2.1. R loop accumulation in DDX47and MECP2depleted cells To validate the selected candidates, DDX47 and MECP2, we determined the levels of R loops after siRNA depletion in another cell line different to the U2OS cells used in the screening. HeLa cell were transfected with specific siRNAs and first a clear reduction in DDX47 and MECP2 mRNA levels and proteins was detected by RT-qPCR and western, respectively after 72 hours of transfection (Figure R8A and R8B). Figure R8. siRNA silencing of DDX47 and MECP2. (A) Relative DDX47 mRNA levels as measured by RT-qPCR in siRNA transfected HeLa cells after 72 hours of transfection (left panel). HPRT housekeeping gene was used to normalize mRNA expression values. Error bars represent relative target quantity (RQ) minimum and maximum from three technical replicates. Western blot analysis of siRNA-treated HeLa cells with siC or siDDX47 (right panel). Vinculin protein is used as a loading control. (B) Analysis of MECP2-siRNA silencing in HeLa cells. Other details as in figure A. Tesis doctoral-Esther Marchena Cruz 69 Then, S9.6 IF assays were performed in siRNA depleted cells including an in vitro treatment with RNase H, which degrades the RNA from the RNA-DNA hybrid. A significant enrichment of the antibody S9.6 nuclear signal was observed for both siDDX47 and siMECP2 cells. Importantly, this increase was suppressed by RNaseH1 treatment, confirming that they correspond to RNA-DNA hybrids (Figure R9). Figure R9. RNA-DNA hybrids accumulation in siDDX47 and siMECP2 HeLa cells by S9.6 immunofluorescence microscopy. Representative images of immunostaining with S9.6 (red) and anti-nucleolin (green) antibodies in HeLa cells upon DDX47 and MECP2 depletion. The median of S9.6 signal intensity per nucleus after nucleolar signal removal in siC, siDDX47 and siMECP2 HeLa cells and treated in vitro with RNase H where indicated. Data from more than 250 total cells from at least three independent experiments is shown. Boxes and whiskers indicate 595 percentiles. ***P < 0.001 (Mann-Whitney U test, two-tailed). A.U., Arbitrary units. R loop accumulation at molecular level was also validated by DRIP-qPCR analysis in the standard human genes (APOE and RPL13A) and in 18S and 28S rDNA genes, given the nucleolar localization of the helicase DDX47 (Sekiguchi et al., 2006) (Figure R10A). In vitro RNase H treatment was used in order to confirm the RNA-DNA hybrids specific immunoprecipitation. Results clearly show a higher accumulation of RNA–DNA hybrids in siDDX47 cells in rDNA regions, but also in RNAPII transcribed (Figure R10B, left panel). In the case of siMECP2, we extended the analysis including PHLDA2, a gene that is regulated by MECP2 (Meng et al., 2014), and YIF1A as another example of MECP2 ChIP-enriched region (determined by ChIP seq in HCT116 cell typeGSE47677). A significant RNA-DNA hybrid accumulation was observed in APOE, RPL13, and PHLDA2 genes, and high levels but no significant were detected at YIF1A and 28S regions (Figure R10B, right panel). RESULTS 70 Figure R10. R loop accumulation in siDDX47 and siMECP2 Hela cells. (A) Schematic diagrams of APOE, RPL13, PHLDA2, YIF1A, 18S and 28S genes. Exons are depicted as open boxes, the arrows indicate the start of transcription and red lines indicate the regions where the primer pairs used for amplification were located. (B) Relative DRIP-qPCR signal values in siDDX47 (red) and siMECP2 (blue) HeLa cells at the indicated regions. Samples were treated in vitro with RNase H prior immunoprecipitation where indicated. Signal values normalized with respect to the siC control are plotted (n=3) as mean and SEM. *P < 0.05; **P < 0.01; ***P < 0.001 (one-tailed paired t-test). (C) ChIP analysis of DDX47 and MECP2 proteins at APOE, RPL13, 18S, 28S, PHLDA2 and YIF1A genes where indicated. Values represent the percentage of the precipitated DNA (IP) to input DNA (INPUT). Immunoprecipitation in siDDX47 and siMECP2 cells were included as control of specificity. IgG was used as negative control. Data are plotted as mean ± SEM (n = 3). *P < 0.05 (one-tailed paired t-test). (D) mRNA levels of APOE and RPL13 in siDDX47 and siMECP2 cells as determined by RT-qPCR. Error bars represent relative target quantity (RQ) minimum and maximum from three technical replicates. Tesis doctoral-Esther Marchena Cruz 71 In parallel, we performed ChIP analysis to study the recruitment of DDX47 and MECP2 to the selected genes analysed in DRIP assays, and found that both proteins were bound to these chromatin regions consistently with a possible role of these factors in the prevention of R loop (Figure R10C). Importantly, these high levels of RNA-DNA hybrids were not due to an increase in transcription at least in APOE and RPL13 genes, since no significant differences were observed in mRNA levels, as detected by RT-qPCR (Figure R10D). Altogether data indicate that DDX47 and MECP2 were necessary to prevent R loop accumulation. 2.2. R loop-dependent genome instability phenotype in DDX47 and MECP2 depleted cells Once we validated R loop accumulation and AID dependent genome instability upon DDX47 and MECP2 depletion, we asked whether these factors play a role in genome instability that could be mediated by RNA-DNA hybrids. First, we analyzed the spontaneous DNA damage in HeLa siRNA transfected cells by H2AX immunofluorescence. A significant increase in H2AX foci number per cell was detected in siMECP2 and a mild increase was also found in siDDX47 cells (Figure R11A). Single-cell electrophoresis analysis revealed that depletions of both proteins increased DNA breaks, that were partially supressed by RNH1 overexpression (Figure R11B) in agreement with the AID-dependent genome instability phenotype of these genes in the siRNA screening. Altogether data confirm and validate that DDX47 and MECP2 are genes involved in R loop homeostasis and that are necessary to prevent R-loop dependent genome instability. In spite of the interesting role of MECP2 as a critical regulator of the maintenance of chromatin architecture (Ragione et al., 2016) and its connection with genome instability (Enikanolaiye et al., 2020), we decided to focus on a protein closed related to RNA metabolism, the helicase DDX47. In this thesis, previous experience and available tools were used to explore further the role of this helicase in transcription and RNA-DNA hybrid metabolism. RESULTS 72 Figure R11. DNA damage analysis in siDDX47 and siMECP2 HeLa cells. (A) Representative images of immunostaining with γH2AX (green) antibody in HeLa cells upon DDX47 and MECP2 depletion. Quantification of the number of γH2AX foci per cell in HeLa cells transfected with the indicated siRNAs is represented (n=3). The red line in the violin plot indicates the median. ***P < 0.001; *P < 0.05 (Mann-Whitney U-test, two-tailed). (B) Representative images of single-cell alkaline gel electrophoresis (comet assay) of HeLa cells upon DDX47 and MECP2 depletion with and without RNaseH1 overexpression. Medians of comet-tail moments are plotted as mean ± SEM (n=4) (one-tailed paired t-test). Tesis doctoral-Esther Marchena Cruz 73 RESULTS 80 Figure R15. Analysis of conflicts between RNAPI and DNA polymerase in DDX47-depleted cells. Proximity ligation assay (PLA) showing interactions of RNAPI and PCNA endogenous proteins. Red spots are indicative of a positive PLA signal. Negative control with only one of the antibodies are also shown. DNA was stained with DAPI. The median of PLA signal intensity per nucleus in siC and siDDX47 is plotted. More than 250 total cells from at least three independent experiments is shown. ***P < 0.001 (Mann-Whitney U test, two-tailed). 3.3. DDX47 physically associates to R loops Since R loops have been proposed to be an obstacle for transcription, and considering that these structures are increased upon DDX47 silencing we asked whether DDX47 could associate to RNA-DNA hybrids in vivo. For this purpose, we performed proximity ligation assay (PLA). A strong PLA signal was detected with DDX47 and S9.6 antibodies, supporting a close association between this factor and RNA-DNA hybrids in human cells (Figure R16). As expected, such association is observed mainly in the nucleolus, because DDX47 is a nucleolar protein. Figure R16. In situ proximity ligation assay between endogenous DDX47 and RNADNA hybrids. Representative images of proximity ligation assay (PLA) showing interactions of DDX47 and RNADNA hybrids in HeLa cells treated in vitro with RNase III. Red spots are indicative of a positive PLA signal. siDDX47 HeLa cells were included as control of specificity. Negative control with only one of the antibodies are also shown (right panel). DNA was stained with DAPI. Tesis doctoral-Esther Marchena Cruz 81 3.4. DEAD/box RNA helicase DDX47 is an RNA-DNA helicase The DEAD/H box helicases belong to helicase superfamily 2 (SF2), which consist of two groups: 44 members in the DEAD-box group and other 15 in the DEAHbox family. Mostly, these enzymes use ATP to bind or remodel RNA and RNAprotein complexes (RNP complexes), with roles in processes including ribosome biogenesis, RNA processing and folding, and different steps of ribonucleoprotein remodelling and mRNA biogenesis. DEAD/H box proteins are built around a highly conserved helicase core of two virtually identical RecA-like domains (Helicase domain I and Helicase domain II) which contribute to the binding site for RNA substrates and ATP hydrolysis. Specifically, at least 12 characteristic sequence motifs are located at conserved position in the helicase core. The highly conserved Motif II inspire the name of the family: D-E-A-D (asp-glu-alaasp) or D-E-A-H (asp-glu-ala-his) (Linder & Jankowsky, 2011; Schütz et al., 2010). Specifically, whereas the conserved motifs I, II, VI and the Q-motif are crucial for ATP binding and hydrolysis, the motif Ia, Ib, IV, IVa and V are important to RNA binding. Conserved motifs III and Va are essential for communication between ATP-binding and RNA-binding sites (Cordin et al., 2006; Linder & Jankowsky, 2011; Schütz et al., 2010). We compared the amino acid sequence of DDX47 protein (Human) with its ortholog in Saccharomyces cerevisiae Rrp3, the putative ortholog in Drosophila Melanogaster DmeI and Oriza sativa TOGR1 using the Clustal Omega web service (McWilliam et al., 2013) and the BLAST analysis (Basic Local Alignment Search Tool). These proteins have a percent identity with DDX47 about 56.72% (S. cerevisiae), 72.75% (D. melanogaster) and 56.94% (O. sativa) and they share 12 conserved domains in the helicase core as determined and highlighted with black boxes in Figure R17A. This data support that DDX47 is a conserved helicase from yeast to humans. Since siDDX47 cells show an accumulation of RNA-DNA hybrids (Figure R9 and R10), we were interested in to know whether this conserved DEAD-box RNA helicase DDX47 possesses RNA-DNA unwinding activity. RESULTS 82 Figure R17. DDX47 conserved domains and mutations in the helicase core analyzed in this study. (A) Multiple sequence alignment of H. sapiens DDX47 (NP_057439.2), S. cerevisiae RRP3 (NP_011932.2), D. melanogaster DmeI (NP_610090.1) and O. sativa TOGR1 (A2XKG2.1) was conducted using Clustal Omega web service (McWilliam et., 2013). Conserved domains are highlighted with orange boxes. (*) conserved residues (:) conservation of aas with strongly similar properties (.) or aas with weakly similar properties. (B) Schematic representation of mutations in the helicase core of DDX47 analyzed in this study. Tesis doctoral-Esther Marchena Cruz 83 3.4.1. Generation of DDX47 mutants for functional in vitro analysis We carried out site-direct mutagenesis to generate mutants in the conserved motif I (DDX47-K74A) and in the motif II (DDX47-E175A) involved in ATP binding and ATP hydrolysis respectively in other helicases (Cordin et al., 2006; Linder & Jankowsky, 2011). For a schematic representation of mutations in the helicase core of DDX47, see Figure R17B. The desired mutations were introduced into a plasmid containing the full DDX47-WT open reading frame sequence (see Materials and Methods 7). Then, DDX47-WT and DDX47 mutant cDNAs were cloned into the vector pET300/NT-DEST that allows high-level prokaryotic expression controlled by the strong bacteriophage T7 promoter and N-terminal 6xHis tag for protein purification (Figure R18, upper scheme; Table M4; Materials and Methods 8.1). Once we checked the induction of the protein at different conditions as determined by coomassie staining and western blot (Figure R18), a scientific collaboration was established with the Professor Xue Xiaoyu in Yale University to perform the biochemical analysis. Figure R18. Development of an in vitro tool for DDX47-WT and DDX47-mutant protein purification. Schematic representation of vector pET300/NT-DEST containing the full DDX47 cDNA with a N-terminal 6xHis tag controlled by T7 promoter (Table M4). His6-DDX47 expression was induced by adding IPTG to the cultured at the indicated conditions. For protein staining, acrylamide gels were incubated with Coomassie Brilliant Blue. The arrows indicate the band at the DDX47 expected molecular weight (50 kDa). Western blot analysis with anti-Histidine and anti-DDX47 antibodies was performed to validate the in vitro tool (Material and Methods 15). RESULTS 84 3.4.2. In vitro unwinding analysis with DDX47 proteins DDX47-WT, and the two mutants DDX47-K74A and DDX47-E175A were purified to homogeneity (Figure R19A). RNA-RNA unwinding activity was tested using a blunt-ended RNA duplex (dsRNA) of 13 base pairs and different amounts of DDX47 proteins. The results indicated that DDX47-WT unwinds this substrate in an ATP and protein concentration-dependent manner (Figure R19B) as well as other dsRNA substrates with either a 5’ or 3’ overhangs (Figure R19C). In parallel, blunt-ended and either a 5’ or 3’ overhangs dsRNA unwinding activity of DDX47 mutants were tested. As can be seen in Figures R19B and R19C, DDX47 mutants unwind dsRNA substrates with a lower efficiency to those of DDX47WT. Next we tested unwinding of RNA-DNA hybrid substrates. DDX47 was able to unwind RNA-DNA hybrids in a protein concentration and ATP-dependent manner. Indeed, our data suggested that DDX47 could have a preference for RNA-DNA hybrids, since the percentage of unwound hybrid product was up to fourfold of that obtained for dsRNA (Figure R20A). Therefore, DDX47 is more adept at unwinding RNA-DNA hybrids than dsRNA. To study the molecular identity of RNA-DNA unwinding activity, DDX47-K74A and DDX47-E175A mutants were tested indicating that an intact DDX47 is crucial for maintaining full RNA-DNA unwinding activity (Figure R20B). These mutants were not able to unwind blunt RNA-DNA hybrids (Figure R20B), but a basal unwinding activity was detected when 5’ overhang RNA-DNA hybrids were assayed (Figure R20C). Tesis doctoral-Esther Marchena Cruz 85 Figure R19. DDX47 RNA helicase in vitro analysis. (A) DDX47-WT and DDX47 mutants protein purification. Purified wild type and DDX47 mutants were analyzed on SDS-polyacrilamide gel and stained with Coomassie Blue. First four lines representing DDX47-WT, the next four lines DDX47K74A mutant and last line DDX47-E175A mutant. (B) DDX47 RNA-RNA unwinding activity. RNAunwinding assay was performed using a blunt-ended RNA duplex (dsRNA) as substrate with different amounts of DDX47-WT, DDX47-K74A and DDX47-E175A. (C) DDX47 RNA-RNA unwinding activity. RNA-unwinding assay was performed using a dsRNA with a 5’ or 3’ overhang as substrates with different amounts of DDX47-WT, DDX47 K74A and DDX47-E175A. The positions of duplex substrate and unwound products are indicated at the left, where the stars show the position of the radiolabel. Gels were dried and subject to phosphorimaging analysis. Performed in collaboration with Xue Xiaoyu’s group. Finally, DDX47 activity was tested with a 5’ RNA-DNA flap structure that resembles a branch migratable R-loop (Schwab et al., 2015). As shown in Figure R20D, DDX47-WT protein could dissociate the flap structure to yield a dsDNA product. DDX47 mutants showed some residual unwinding activity of these structures. Therefore, our results indicate that DDX47 resolves RNA-DNA hybrids as well as R-loop mimicking structures, and that the mutants DDX47-K74A and DDX47-E175A showed a residual activity, so it is possible that other residues could contribute to this activity. RESULTS 86 Figure R20. DDX47 RNA-DNA helicase in vitro analysis. (A) Comparison between DDX47 RNA helicase and RNA-DNA helicase activity using same amount of dsRNA or RNA-DNA duplex with a serial dilution of DDX47-WT protein. Graph shows the percentage of unwound product respect to the DDX47 concentration-dependent manner. (B) DDX47 RNA-DNA unwinding activity. RNA-DNA unwinding assay was performed using a blunt-ended RNA-DNA duplex as substrate with different amounts of DDX47-WT, DDX47-K74A and DDX47-E175A. (C) DDX47 RNA-DNA unwinding activity. RNA-DNA unwinding assay was performed using a RNA-DNA with a 5’ or 3’ overhang as substrates with different amounts of DDX47-WT, DDX47 K74A and DDX47E175A. The positions of duplex substrate and unwound products are indicated at the left, where the stars show the position of the radiolabel. Gels were dried and subject to phosphorimaging analysis. (D) DDX47 unwinds RNA-DNA flap structures mimicking R-loops. RNA-DNA unwinding assay with DDX47-WT, DDX47-K74A and DDX47-E175A using RNA-DNA flap structures mimicking R-loops as substrates. Other details as in (C). Graph shows the percentage of dsDNA product recovered after the reaction. Performed in collaboration with Xue Xiaoyu’s group. Tesis doctoral-Esther Marchena Cruz 87 3.4.3. DDX47 in vivo activity Concerning the mechanism in which DEAD-box proteins unwind duplexes (Yang et al., 2007) could be expected that any RNA helicase could have RNA-DNA unwinding activity in vitro, so we wondered whether DDX47 shows RNA-DNA hybrids unwinding activity in vivo. We expressed DDX47-WT and DDX47-K74A and DDX47-E175A under the control of the strong promoter CMV (Figure R21A). To confirm DDX47 overexpression we performed immunofluorescence assays with antiDDX47 and anti-Flag antibodies to detect endogenous and tagged-DDX47 proteins respectively (Figure R21A). A strong nuclear signal was detected in cells transfected with DDX47-Flag proteins, whereas a weaker and nucleolar signal was detected in cells transfect with the empty plasmid (endogenous DDX47). Overexpression was also confirmed by western blot analyses (Figure R21B). Figure R21. Tools for in vivo DDX47 overexpression. (A) Schematic representation of expression vector containing fulllength human DDX47-WT or DDX47 mutants with C-terminal Myc-DDK-tag under the control of CMV promoter. Representative images of immunostaining with anti-FLAG (green) in Hela cells transfected with the empty plasmid, pDDX47-WT, pDDX47 K74A and pDDX47 E175A after 24 h of overexpression. Nuclei was stained with DAPI. (B) Western blot analysis with antiDDX47 and anti-Flag of HeLa cells transfected with the indicated plasmids. siC and siDDX47 cells were included as control of endogenous DDX47.Vinculin protein is used as a loading control. RESULTS 88 Then, we assayed whether overexpression of WT and helicase-dead mutants of DDX47 could supress or not the R-loop accumulation phenotype caused by DDX47 depletion itself. For overexpression experiments, we used a different pool of two siRNAs against the 3’ UTR, just to ensure DDX47 genomic silencing but not those of DDX47 expressed from the plasmid. A similar silencing was achieved with the 3’ UTR siRNAs, as determined by western blot (Figure R22A) and importantly these siRNAs conferred a RNA-DNA hybrid accumulation similar to those of siRNA pool (Figure R22B) without affecting transcription levels (Figure R22C). Figure R22. Tools for specific siRNA depletion and overexpression of DDX47. (A) Western blot analysis with anti-DDX47 of siRNA-treated HeLa cells with siC, siDDX47 or different pools of two siRNAs against the 3’UTR after 72 h. (B) The median of S9.6 signal intensity per nucleus after nucleolar signal removal in the indicated siRNA-treated HeLa cells is plotted. At least 200 cells were scored. A.U., Arbitrary units. (C) mRNA levels of DDX47, APOE and RPL13 in the indicated siRNA-treated cells as determined by RT-qPCR. Error bars represent relative target quantity (RQ) minimum and maximum from three technical replicates. We measured R-loop accumulation upon DDX47 overexpression by S9.6 IF assays. To measure exclusively R-loop accumulation in nucleoplasm and nucleolus, S9.6 IF experiments were carried out including an in vitro treatment with RNase III, which degrades dsRNA structures (Silva et al., 2018) and nucleolin staining was used to determine nucleolar area. First, a significant enrichment of the antibody S9.6 nuclear signal was observed in siDDX47 cells in both analysis, confirming the RNA-DNA hybrids accumulation phenotype in the nucleoplasm as well as in the nucleolus upon depletion of the helicase DDX47. Concerning the antibody S9.6 nucleoplasm signal, DDX47-WT overexpression rescued the RNA-DNA hybrid accumulation associated to DDX47 depletion (Figure R23). Furthermore, in the nucleoplasm (left panel) the two helicases mutant forms reduced the R-loop accumulation in siDDX47 depleted cells, Tesis doctoral-Esther Marchena Cruz 89 although to a lesser extent than DDX47-WT protein. These results could be explained by the residual unwinding activity of mutants on hybrids structures (Figure R20C and R20D). Nevertheless, we cannot discard that these mutant proteins could stick to the hybrids reducing the accessibility of S9.6 antibody to these structures and the immunofluorescence signal as can be observed not only in siDDX47 cells, but also in siC cells. Figure R23. DDX47 overexpression effects in RNA-DNA hybrids. Representative images and quantification of S9.6 immunofluorescence signal in HeLa transfected cells with siDDX47 and with pFLAG (-), pFLAG-DDX47 (+DDX47) for DDX47-WT overexpression, and pFLAG-DDX47-K74A (+DDX47-K94A) or pFLAG-DDX47-E175A (+DDX47-E175A) for helicase-dead mutant DDX47 overexpression (blue). All samples were treated with RNase III. The graph shows the median of S9.6 signal intensity in the nucleoplasm after nucleolar signal removal (left panel) or the S9.6 signal in the nucleolus (right panel) (n=4). More than 200 cells per condition were counted in each experiment. ***P < 0.001 (Mann-Whitney U test, two-tailed). A.U., Arbitrary units. Black stars denote significant increases, whereas red stars denote significant decreases. DISCUSSION 96 (histone deacetylase 8) and CHRAC1 (chromatin accessibility complex protein 1) (Table R2), were identified in agreement with the nascent connection between chromatin changes and R loop homeostasis (Chédin, 2016). We have validated HDAC8 and MECP2 candidates with the use of S9.6 antibody (Figure R6). HDAC8 is a histone deacetylase necessary for cohesin recycling during the cell cycle and mutations in these gene are associated to Cornelia de Lange syndrome (Deardorff et al., 2012; Kaiser et al., 2014). In addition to control the acetylation of this and other protein substrates, HDAC8 is a class I HDAC that deacetylates histone H3 and H4 at nonspecific lysines, and recently has been shown to be a novel cofactor of SMAD3/4 complex playing a role in transcriptionally suppression (Tang et al., 2020). Thus, the increase in R loop in siHDAC8 cells could be related with a more relaxed and transcribed chromatin. Indeed, high chromatin accessibility caused by histone hyper-acetylation have been proposed to facilitate R loop generation. Depletion of the histone deacetylase mSin3a complex, both in yeast and human cells, or deacetylation inhibition by chemical compounds (TSA and SAGA) causes R loop accumulation and genome instability (Salas‐ Armenteros et al., 2017; Wahba et al., 2011). CHRAC1, another candidate of our screening, is a subunit of a complex that interacts with the remodelling factor ACF to facilitate nucleosome sliding (Eberharter et al., 2001). In the same line, the FACT complex (facilitates chromatin transcription) that works in transcription elongation and chromatin remodelling, has a role in the prevention of R-loopmediated transcription-replication conflicts, likely associated with a specific chromatin organization (Herrera-Moyano et al., 2014). Moreover, recent results have opened new insights to understand the role of chromatin remodelers in the control of R loop homeostasis and its impact on genome integrity (Bayona-Feliu et al., 2021; Prendergast et al., 2020). Further analysis will be needed to know whether CHRAC1 plays a similar role in R loop and genome instability. We found that some genes involved in transcription and RNA biogenesis as well as translation caused high levels of R loop dependent DNA damage highlighting transcription factors as MYOG, PHOX2A, ARNTL2 and NR2E3; some helicases as DDX42, DDX47 and DHX9; and translation initiation and elongation factors as EIF3A, EIF3C and EEF1D. The identification of transcription and RNA processing related factors is consistent with the importance of mRNP Tesis doctoral-Esther Marchena Cruz 97 biogenesis factors in the prevention of R loop accumulation (Gavaldá et al., 2013; Huertas & Aguilera, 2003; X. Li & Manley, 2005; Paulsen et al., 2009; Stirling et al., 2012). Interestingly, cells depleted of genes involved in protein degradation (PSMC5, PSMD7, STAMBP, SPINK5) exhibited high level of DNA damage upon AID induction (Ferdous et al., 2001; Geng et al., 2012). Accordingly, other proteasome subunits were identified in the RNA-DNA hybrid interactome (Cristini et al., 2018). Nevertheless, since the proteasome participates in numerous cellular processes, including transcription regulation, cell cycle progression, apoptosis, or DNA damage repair it is not clear whether these proteins could have a direct role in R loop homeostasis. Finally, it should be notice that some factors previously related with R-loop genome instability were not identified by the AIRD system, as it would be expected. This could be due to high basal levels of DNA damage in the absence of these factors that did not allow to detect a significant increase upon AID induction; the double selection criteria for the screening; cell death associated to know-down of some factors; or technical issues derived from the high-throughput analysis protocol. Our collection of siRNAs includes targeting 3205 human genes that locate both in nucleus and cytoplasm and are involved in a wide variety of processes, as apoptosis, nucleic acid binding, autophagy and others (Appendix 2). Some of the putative candidates and the process in which they participate have not been previously linked to R-loop dependent DNA damage, such as lipid metabolism, transporters, G-protein related and signalling, metabolic pathways among others. In spite to confirm these candidates, our data suggest a wide range of factors that could contribute may be indirectly to R-loop dependent DNA damage. Further analysis will be required to define the molecular mechanisms and biological meaning behind these factors. In this thesis, we decided to focus our study in two candidates that showed different activities or properties on the nucleic acid metabolism: the methyl-CpG binding protein MECP2 and the helicase DDX47. DISCUSSION 98 2. ROLE OF MECP2 IN THE MAINTENANCE OF GENOME INTEGRITY Mutations in methyl-CpG binding protein 2 (MECP2) are the major cause of Rett syndrome which is a progressive neurologic developmental disorder that primarily affects females, since it is X-linked and subject to X inactivation (Amir et al., 1999; Good et al., 2021; Ip et al., 2018). MECP2 encodes a protein that binds DNA methylated and is highly expressed in neurons (near histone-levels) (Lewis et al., 1992; Meehan et al., 1992; Skene et al., 2010). Genome-wide analysis indicate that MECP2 binds widely across the genome and is enriched at mCG and mCA sites (Lin Chen et al., 2015). Classically MECP2 was proposed to participate in transcription repression by its ability to bind to methylated DNA and co-repressor complexes (Ip et al., 2018). However, MECP2 is a multitalented protein that performs many functions, as a modulator of gene expression, transcriptional repression or activation, depending on its partners (Boxer et al., 2020); and some other crucial functions in chromatin architecture such as binding to pericentric heterochromatin regulating its condensation (Ip et al., 2018; C. H. Li et al., 2020). In addition, other mechanisms of gene expression regulation conferred by MeCP2 have been described. MECP2 is able to bind methylated miRNA loci controlling miRNA biogenesis and gene expression directly by interacting with DGCR8, a main component of the nuclear microRNAprocessing machinery known as DiGeorge syndrome critical region 8, and interfering with Drosha (Cheng et al., 2014; Glaich et al., 2019). Another indirect MECP2 role in gene expression is linked with splicing, favouring splice junctions recognition and intron excision (Glaich et al., 2019; Osenberg et al., 2018). The activity of MECP2 in this process is mediated via interactions with spliceosome components in an RNA-dependent manner, indeed MECP2 protein has been shown to binds RNA in vitro (Jeffery & Nakielny, 2004; Young et al., 2005). In this thesis we have provided evidence of a putative role of MECP2 in suppressing R-loop accumulation and R-loop dependent genome instability by different approaches (γH2AX IF, comet assay, DRIPs and S9.6 IF) in standard cancer cells such U2OS and HeLa. Here we will discuss different possible ways by which MECP2 could prevent R loop accumulation, and R-loop dependent genome instability phenotype (Figure D1): Tesis doctoral-Esther Marchena Cruz 99 One possible mechanism is linked to the capacity of MeCP2 to bind broadly across the genome acting as a modulator of transcription of specific genes through its interaction with DNA and with different transcription-repressor factors promoting a repressive chromatin environment (Figure D1A) (AW et al., 2020; Boxer et al., 2020; Ragione et al., 2016). Thus, we show that siMECP2 cells accumulate R loops in MeCP2-regulated genes such as PHLDA2 in addition to standard R loop accumulating genes as RLP13 and APOE (Figure R10B, right panel) (Meng et al., 2014). The hypothesis is that MECP2 depletion could favour R-loop formation because of an open chromatin state more accessible and prone to the formation of DNA-RNA hybrids. Among the interacting partners of MECP2 we highlight the histone deacetylase complex Sin3A, that is involved in cotranscriptional histone deacetylation and prevention of R-loop accumulation and genome instability (Salas‐Armenteros et al., 2017). In addition, MECP2 also interacts with the chromatin remodelling complexes Brahma (Brm), a catalytic component of SWI/SNFcomplex, and ATRX, both of them with a reported role in R-loop homeostasis and genome instability (Bayona-Feliu et al., 2021; D. T. Nguyen et al., 2017). Another interesting partner is MBD2, the methylated DNA binding component of the methyl-CpG-binding protein 1 (MeCP1) transcriptional repressor complex, identified as an R loop interactor in a biochemical screening (Cristini et al., 2018). Regarding the role of MECP2 on modulating gene expression is interesting to notice that genome-wide analysis support the view of MeCP2 as a transcriptional regulator of transposons and repetitive elements by repressing inappropriate transcription (Muotri et al., 2010). Different studies in yeast and human cells indicate that these repetitive elements are R loop-prone sites (El Hage et al., 2014; Nadel et al., 2015; Zeng et al., 2021), thus it would be interesting to extend our DRIP analysis in siMECP2 cells to repetitive elements. Interestingly, MECP2 interacts with ATRX, a chromatin remodeler (Nan et al., 2007) that plays a role suppressing deleterious DNA secondary and R loops at telomeric repeats regions (D. T. Nguyen et al., 2017). Given that depletion of ATRX and MECP2 confer an increase in genome instability, it would be worth studying whether both proteins could act together to prevent R-loop-mediated genome instability (Huh et al., 2016; Leung et al., 2013). DISCUSSION 100 Figure D1. A model to explain the possible role of MeCP2 in genome integrity. (A) MeCP2 through its interaction with DNA and different transcription-repressor factors represses the expression of certain genes. MeCP2 depletion could lead to altered expression patterns and transcription deregulation which in turn, could lead to R loop formation. (B) MeCP2 through its capacity to interact with both DNA and nucleosomes and its crucial role in chromatin compaction promotes a repressive chromatin environment. MeCP2 depletion could leads to the unbalance between condensed and decondensed chromatin, favouring R-loop formation because of an open chromatin state more accessible and prone to the formation of DNA-RNA hybrids. Genome wide data show that MeCP2 has an affinity to GC-rich chromatin, with a preferential binding profile to nucleosomal DNA (Rube et al., 2016). Furthermore, most of the MeCP2-bound promoters genome-wide are expressed genes being only 6% highly methylated, suggesting other roles of MeCP2 in addition to the silencing of methylated promoters (Yasui et al., 2007). Curiously, in human genome, the majority of unmethylated CGI promoters show significant strand asymmetry in the distribution of guanine and cytosines, known as GC skew, where long, stable, three-stranded nucleic acid structures are prone to be formed upon transcription (Liang Chen et al., 2017; Dumelie & Jaffrey, 2017; Ginno et al., 2012; Nadel et al., 2015; Sanz et al., 2016). For this, MeCP2 could have a role in regulating chromatin state at GC skew rich unmethylated promoters, so, when lost, an altered chromatin could be the responsible of R loop formation. Tesis doctoral-Esther Marchena Cruz 101 On the other hand, it is important to notice that the majority of MeCP2binding sites are intergenic or intronic (Yasui et al., 2007). For this, another possible mechanism could be related to the function of MECP2 in alternative splicing regulation, though interaction with both splicing factors and epigenetic marks (Long et al., 2011; Maunakea et al., 2013; Young et al., 2005). It has been shown that MeCP2 binds to methylated DNA regions slowing down RNAPII elongation to enhance intron recognition and processing (Glaich et al., 2019; Wong et al., 2017). So, it is tempt to speculate that upon MECP2 depletion, the nascent RNA could be stickier, favouring the formation of RNA-DNA hybrids structures, according with the idea that a defective mRNP or RNAPII stalling would increase the possibility of hybridization of the mRNA with its DNA template leading R loop accumulation (García-Muse & Aguilera, 2019). Finally, a third mechanism could be based on the capacity of MeCP2 to interact with both DNA and nucleosomes and its crucial role in chromatin compaction independent of DNA methylation (Figure D1B) (Georgel et al., 2003; Nikitina et al., 2007). MECP2 promotes condensation of pericentric heterocromatin in a RNA-dependent manner and it has been shown to compete with H1 for nucleosome binding (Brero et al., 2005; Ghosh et al., 2010; Good et al., 2021). Recently, it has been shown that MeCP2 and the chromatin remodeler ATRX are reciprocally dependent both for their expression and targeting to pericentric heterochromatin organization as chromocenters (Marano et al., 2019). It is tempt to speculate that the unbalance between condensed and decondensed chromatin could alter expression patterns upon MECP2 depletion and indirectly could cause R-loop accumulation, and genome instability (Figure D1). The R loop accumulation and R-loop dependent DNA damage in siMECP2 cells reveals a novel connection between R loops and a factor that acts at the interface between transcription regulation and chromatin. The multi-faceted characteristic of this protein, its particular genome distribution and its described crosstalk with some proteins involved in the maintenance of genome stability open new perspectives on its biological role that would need to be explored in the future. Genome-wide analysis and the use of other cells lines with higher MECP2 expression such as the human SH-SY5Y neuronal cell line would contribute to expand the knowledge of this protein in a more physiological context. DISCUSSION 102 3. ROLE OF DDX47 IN THE MAINTENANCE OF GENOME INTEGRITY DDX47 is a conserved DEAD box RNA helicase that has been shown to be associated to pre-RNAs suggesting a role of this protein at early steps of prerRNA processing. This protein is mainly located in the nucleolus by its interaction with the nucleolar protein NOP132, but it has also been associated to ribosome biogenesis proteins and mRNA splicing factors, among others, as determined by pulldown analysis (Sekiguchi et al., 2006). The function in rRNA processing is conserved since, Rrp3 the yeast ortholog (identity 57%) has been reported to be an essential protein required for 18S rRNA production (Granneman et al., 2006; O’Day et al., 1996). In this thesis, we identified DDX47 in our screening as a gene that when depleted causes DNA damage that is increased by AID induction. Moreover, DDX47 is a conserved helicase that acts mainly in the nucleolus, in transcription and rRNA processing and preventing R loop accumulation and genome instability (Figure D2). Our results indicated that DDX47 in addition to rRNA processing is an helicase involved in rDNA transcription, as determined by ChIP of RNAPI and DDX47 and EU labelling (Figure D2) (Figure R13D, R12B and R14). DDX47 depletion causes a reduction of nucleolar total area (Figure R13A), that could be a consequence of the defect in transcription. Indeed, it has been described that the state of transcription and the amount of rRNA modulate the kinetics of nucleoli formation (Berry et al., 2015). This phenotype was accompanied by an increase in the total amount of RNAPI and UBF, a factor necessary to active rDNA genes (Potapova & Gerton, 2019). This could be explained as a compensation mechanism of living cells to mitigate the observed transcription defect. Altogether data support that DDX47 is necessary for transcription and homeostasis at the nucleolus. Since DDX47 leads to transcription defects, and genome instability we thought that this helicase could be necessary for a correct progression of polymerases and that its absence could cause transcription-associated genome instability. Indeed, we have observed in siDDX47 cells an increase of rDNA Tesis doctoral-Esther Marchena Cruz 103 transcription-replication collision events (Figure D2) (Figure R15). In agreement with these data, another study propose that DDX47 is recruited by FANCD2 and has a role preventing RNAPII transcription-replication conflicts (Okamoto et al., 2019). It would be necessary to study whether these collisions are suppressed by RNAse H1 overexpression, in order to determine whether this genome instability phenotype is mediated by R loop. Figure D2. Role of DDX47 in rRNA transcription and transcription-replication conflicts. DDX47 could have a possible role in RNAPI-driven transcription since its depletion causes a reduction of nucleolar total area and rDNA transcription impairment at the nucleolus. DDX47 is also important to maintenance of genome integrity since its depletion leads to R loop accumulation and R-loop-mediated DNA damage as well as an increase of rDNA transcriptionreplication collision events in the nucleolus. One interesting question is whether DDX47 has an unique and specific role, or there are helicases with redundant functions. Concerning other nucleolar helicases with similar functions, DHX33 is a nucleolar RNA helicase that is also important for the nucleolar integrity and is involved in mRNA translation initiation by promoting elongation-competent 80S ribosome assembly (C. Zhang et al., 2015). Nevertheless, there is any reported relation of this helicase with genome instability neither with R loops. It is worth to notice the multifaceted nucleolar RNA helicase DDX21 that is required for rRNA processing also plays an important role in resolving R loops and preserving genome instability both in rDNA locus and RNAPII-transcribed regions (Song et al., 2017). Similarly, the RNA-DNA hybrid accumulation phenotype in siDDX47 cells was observed in the nucleoplasm as well as in the nucleolus (Figure R23). Nevertheless, DDX21 has a wider role than DDX47, since in addition to RNA modification also promotes RNAPII transcription elongation (Calo et al., 2015). Interestingly, DDX21 interacts with DDX47 as DISCUSSION 104 determined by pulldown analysis (Sekiguchi et al., 2006), whether these helicases could act in a coordinated manner is something to be explored. Numerous RNA helicases have been reported to be involved in R-loop dynamics, following the classical SETX, RNA helicases highlight as UAP56, DDX23, DDX5, DHX9, DDX19, DDX21, AQR, DDX1 among others (García-Muse & Aguilera, 2019). However, for few cases, it has been tested a comparative analysis of the efficiency removing DNA-RNA hybrids versus RNA-RNA duplexes and the mechanism of action by which these helicases control R-loop homeostasis in vivo is still uncertain. It remains unclear whether these RNA helicases unwind DNA-RNA hybrids directly or whether their ability of remodelling or releasing mRNP could indirectly cause a reduction of R-loop accumulation. We have shown that DDX47 is able to unwind RNA-RNA duplex in a ATP dependent manner (Figure R19) as well as its described yeast ortholog, Rrp3 (Garcia et al., 2012). We show that DDX47 binds substrates with a preferred orientation, in particular dsRNA with a 3’ overhangs and similarly Rrp3 show an unwinding activity stimulated by either 5´ or 3´ single-strand extension. Our data demonstrate a novel role for the highly conserved human DDX47 in unwinding of RNA-DNA hybrids and structures that resemble an R loop in vitro (Figure D3) (Figure R20). In fact, DDX47 showed a preference for RNA-DNA hybrids rather than dsRNA duplexes. In order to test the helicase activity we assayed two different mutants, in the conserved motif I and II involved in ATP binding and ATP hydrolysis respectively (Cordin et al., 2006; Linder & Jankowsky, 2011). These mutations were previously assayed for the helicase UAP56 in vitro and in vivo analysis confirming the helicase activity of the protein (Pérez-Calero et al., 2020). Nevertheless, in the case of DDX47 both mutants showed a residual activity, so it is possible that other residues could contribute to this activity. A comparative structural analysis of some human DEAD-box RNA helicases highlight that the 6th residue upstream of the conserved glutamine of the ATP binding sites is usually an aromatic residue, however DDX47 has a tryptophan in the corresponding position (Schütz et al., 2010), maybe this residue could have a biological meaning that may differentiate DDX47 function from other helicases. It would be interesting to study the effect of tryptophan mutation in DDX47 unwinding activity. Tesis doctoral-Esther Marchena Cruz 105 In parallel to the in vitro analysis, we also show that DDX47 overexpression in different genetic backgrounds, with high levels of R loops, supresses R loops accumulation, regardless the origin of these RNA-DNA hybrid structures. Thus, overexpression of DDX47 was able to suppress the phenotype of cells depleted of helicases involved in termination and splicing (SETX, DDX23) as well as that of cells depleted of components of Fanconi Anemia (FANCD2) confirming the relevance on its RNA-DNA unwinding activity in vivo (Figure D3) (Figure R24). Strikingly, DDX47 was not able to supress UAP56 depletion phenotype. Indeed, the viability of siUAP56 was severely affected upon overexpression of DDX47, suggesting a negative genetic interaction. In addition, the mutants DDX47-K74A and DDX47-E175A show a residual RNA-RNA and RNA-DNA unwinding activity unlike mutants in the same respective residues of UAP56, suggesting that other residues are involved (Figure R20) (Pérez-Calero et al., 2020). Probably DDX47, as RNA helicase with affinity for RNA and RNA-DNA hybrid, could bind to these structures and compete with UAP56 binding interfering with the function of this key helicase. This data, together the different efficiency of RNA-DNA unwinding activities of UAP56 and DDX47 (Figure R20) (Pérez-Calero et al., 2020) suggest that these helicase have different roles. UAP56 will be a general helicase with a global and co-transcriptional unwinding activity over the genome, whereas DDX47 will have a lower efficiency and impact, being more restricted to the nucleolus. DDX47 dsRNA unwinding activity seems to be higher than that of UAP56 protein (Figure R19) (Pérez-Calero et al., 2020). Interestingly a comparative analysis of UAP56 and DDX49, the DDX47 ortholog, reveals that this helicase, that shares 46% of identity with DDX47, has a robust ATPase and RNA helicase activities of dsRNAs, significantly much higher than those of UAP56 (Awasthi et al., 2018). This activity could account for the different roles of DDX49, as an RNA helicase, that modulates translation by regulating mRNA export and the levels of pre-ribosomal (Awasthi et al., 2018). DDX47 and DDX49 could be more promiscuous performing their functions. In the case of DDX47, for which we have demonstrated an RNA-DNA hybrid unwinding activity, the ability to resolve R loops could be a consequence of its action as an RNA chaperone that resolves CONCLUSIONS 112 CONCLUSIONS 146 human genes have been identified in an siRNA high-throughput screening covering 3205 different genes with a potential role in R loop and DNA damage homeostasis. Such genes are functionally related with chromatin remodelling and modification, transcription and RNA biogenesis, protein degradation, cell cycle control, and other cellular processes. 2The two top hits identified were MeCP2, a Methyl-CpG Binding Protein, and DDX47, a nucleolar RNA helicase, both of which were validated and characterized further in this study. 3Depletion of MECP2 leads to DNA damage and R-loop accumulation, identifying a novel connection between R loops and a factor that acts at the interface between transcription regulation and chromatin. 4DDX47 depletion leads to R loop accumulation and R-loop-mediated DNA damage. Depletion of DDX47 increases the levels of rDNA transcriptionreplication collision events in the nucleolus, indicating that this helicase is important to the maintenance of genome integrity. 5DDX47 depletion causes a reduction of nucleolar total area, transcription impairment at the nucleolus and a slight decrease of RNAPI occupancy at rDNA, suggesting a possible role of this helicase in RNAPI-driven transcription. 6DDX47 is an RNA-DNA helicase and R-loop resolvase able to remove R loops regardless of their origin, as shown in vitro and in vivo. Tesis doctoral-Esther Marchena Cruz 113 CONCLUSIONS 114 CONCLUSIONES 1Se han identificado 46 genes humanos en un escrutinio de alto rendimiento de siRNA que abarca 3205 genes diferentes con un papel potencial en la homeostasis de los bucles R (R loops) y el daño en el ADN. Estos genes están funcionalmente relacionados con la remodelación y modificación de la cromatina, la transcripción y la biogénesis del ARN, la degradación de proteínas, el control del ciclo celular y otros procesos celulares. 2Los dos candidatos principales identificados fueron MeCP2, una proteína de unión a metil-CpG, y DDX47, una helicasa de ARN nucleolar, que fueron validados y caracterizados en este estudio. 3La depleción de MECP2 causa un aumento de daño en el ADN y acumulación de R loops, identificándose así una nueva conexión entre los R loops y un factor que actúa en la interfaz entre la regulación de la transcripción y la cromatina. 4La depleción de DDX47 causa acumulación de R loops y daño del ADN mediado por R loops. La depleción de DDX47 aumenta los niveles de eventos de colisión transcripción-replicación en el rDNA, lo que indica que esta helicasa es importante para el mantenimiento de la integridad del genoma. 5La depleción de DDX47 provoca una reducción del área total nucleolar, un deterioro de la transcripción en el nucléolo y una ligera disminución del reclutamiento de RNAPI en el rDNA, lo que sugiere un posible papel de esta helicasa en la transcripción llevada a cabo por la RNAPI. 6DDX47 es una helicasa de ARN-ADN capaz de eliminar R loops independientemente de su origen, como se ha demostrado in vitro e in vivo. Tesis doctoral-Esther Marchena Cruz 115 116 APPENDIX 117 APPENDIX 118 Appendix 1. Development of an inducible AID-based R-loop detection cell line. (A) Analysis of selected U2OS-TR-AID clone (#8) after 48h AID induction. Quantification of nuclear AID immunofluorescence signal is shown (left panel). AID inducible expression in the selected clone, as determined by western blot from total protein extracts (right panel). (B) Quantification of the number of γH2AX foci per cell in control (siC) and FANCD2-depleted (siFANCD2) U2OS-TR-AID cells, in the absence or presence of AID and with or without RNaseH1 overexpression, as indicated. The red line indicates the median (n=3). (A-B) The statistical significance of the difference was calculated with Mann-Whitney U-test; *P <0.05; ***P<0.001. (C) Validation of the U2OS-TR-AID cell system in high-throughput format with automated workflow and analysis. Percentages of cells with ≥5 γH2AX foci per cell in control cells (siC), cells depleted for known Rloop metabolism related factors (siFANCD2 and siTHOC1) in the absence or presence of AID. Data are plotted as mean ± SEM (n=2). *P <0.05; **P <0.01; ***P<0.001 as determined by paired two-tailed t-test. Appendix 2. List of 3205 siRNAs of Dharmacon-ON TARGET Plus-Druggable genome siRNA library. Note: the complete library comprises 4796 siRNAs. In order to improve the screening several cytoplasmic genes were excluded. https://drive.google.com/file/d/1J48wbdHniDLEzFIgf_HDFy4WPRIihr7Z/view?usp=sharing Tesis doctoral-Esther Marchena Cruz 119 Appendix 3. γH2AX foci/cell analysis of 3205 siRNA library (First round). Data show the percentages of cells with ≥5 or ≥10 γH2AX foci with and without AID expression upon siRNA depletion. Two ratios for each siRNA are shown: one versus itself with and without AID expression (siX+DOX/siX), and the other one versus the median of siC with AID expression in its plate (siX+DOX/siC+DOX). The candidates selected (156 out of 3205) in this round of the screening are highlighted. https://drive.google.com/file/d/1L09bRDq2x5d_j5bBwNvfj_nRxEXP73yf/view?usp=sharing Appendix 4. Identification of high-throughput screening candidates (First round). Analysis of γH2AX immunofluorescence data (first round: 3205 siRNA in one replicate with a duplicate). The percentages of cells with ≥5 or ≥10 γH2AX foci when AID was expressed were used to calculate the indicated ratios, represented in both axis: Y, siRNA versus itself without AID expression; X, siRNA versus the median of siC with AID expression in its plate. Scatter plot of the ratios for each siRNA is shown. The green dotted lines show the cutoff (1.2) used to designate a siRNA as positive. Two ratios for each siRNA are represented by a dot: black dots: candidates; grey dots: others. APPENDIX 120 Appendix 5. γH2AX foci/cell analysis of validated candidates (Validation). Data show the percentages of cells with ≥5 or ≥10 γH2AX foci with and without AID expression upon siRNA depletion of the 144 candidate genes and their comparison ratios as explained in Appendix 3. The selected hits (46 genes) defined as those with both ratios ≥ 1.2, at least in two duplicates of one replicate, are highlighted. https://drive.google.com/file/d/1H-_dRcFeA93MNjD5oOe3B7kckI4U7GmU/view?usp=sharing Tesis doctoral-Esther Marchena Cruz 121 MATERIALS AND METHODS 128 Table M2. Secondary antibodies used in this study. Specificity Conjugation Reference Use Rabbit Horseradish peroxidase A0545 (Sigma) WB (1:5000) Mouse Horseradish peroxidase A4416 (Sigma) WB (1:5000) Mouse PLA probe (MINUS oligonucleotide) DUO92004 (Olink Biosciences) PLA (1:5) Rabbit PLA probe (PLUS oligonucleotide) DUO92002 (Olink Biosciences) PLA (1:5) Mouse Alexa Fluor 488 A11029 (Molecular Probes) IF (1:1000) Mouse Alexa Fluor 488 A21200 (Molecular Probes) IF (1:1000) Mouse Alexa Fluor 594 A21201 (Molecular Probes) IF (1:1000) Mouse Alexa Fluor 647 A21463 (Molecular Probes) IF (1:1000) Rabbit Alexa Fluor 488 A11008 (Molecular Probes) IF (1:1000) Rabbit Alexa Fluor 555 A31572 (Molecular Probes) IF (1:1000) Rabbit Alexa Fluor 568 A11011 (Molecular Probes) IF (1:1000) Goat Alexa Fluor 647 A21447 (Molecular Probes) IF (1:1000) WB: Western blot; IF: immunofluorescence; PLA: Proximity Ligation Assay. 3. BACTERIA AND HUMAN CELL LINES 3.1. Escherichia coli strains The DH5α strain (FendA1 gyr96 hsdR17 ΔlacU169(f80lacZΔM15) recA1 relA1 supE44 thi-1) (Hanahan, 1983) was used to maintain and amplify DNA plasmids. Rosetta™ host strain (BL21 derivative) (FompT hsdSB(rBmB-) gal dcm (DE3) pRARE (CamR)) was used for the expression of eukaryotic proteins (His6DDX47) (BLB21 Rosetta Electrocompetent Cells -70954-3, Merck). 3.2. Human cell lines Human cells used in this study are listed in the Table M3. Table M3. Human cell lines used in this study. Cell line Description Medium Source U2OS-TR-AID Human Bone Osteosarcoma Epithelial Cells carrying a tetracycline regulated human cytidine deaminase AID gene cassette DMEM José Manuel Calderón Montaño. Aguilera’s lab U2OS Human Bone Osteosarcoma Epithelial Cells DMEM ATCC HeLa Human Cervical Adenocarcinoma Epitelial Cells DMEM Sigma (ECACC) (12022001-1VL) ATCC: American Type Culture Collection; ECACC: European Collection of Authenticated Cell Cultures. Tesis doctoral-Esther Marchena Cruz 129 U2OS-TR-AID stable cell line was used for siRNA library screening. Functional analyses were performed in both U2OS and HeLa cells. 4. PLASMIDS Plasmids used are shown in Table M4. Table M4. Plasmids used in this study. Plasmid Description Resistance Reference/Source pCDNA3 Vector containing a PCMV for expression in mammalian cells Ampicillin (ten Asbroek et al., 2002) pCDNA3RNaseH1 pcDNA3 containing the human RNase H1 gene under the PCMV Ampicillin (ten Asbroek et al., 2002) pFLAG-CMV6A Expression vector derivative of pCMV5 used to establish transient intracellular expression of FLAGtagged proteins in mammalian cells Ampicillin E1900 (Sigma) pCMV-DDX47FLAG Expression vector derivative of pCMV6-Entry containing full-length human DDX47-WT with C-terminal Myc-DDK-tag. Transient expression is driven by the CMV promoterregulatory region Kanamycin RC209448 (Origene) pCMV-DDX47K74A-FLAG Expression vector derivative of pCMV6-Entry containing DDX47-K74A mutation with C-terminal Myc-DDK-tag Kanamycin This study pCMV-DDX47E175A-FLAG Expression vector derivative of pCMV6-Entry containing DDX47-E175A mutation with C-terminal Myc-DDK-tag Kanamycin This study pDONR221DDX47-WT Vector derivative of pDONOR221 for use in Gateway® Technology. Vector containing a pUC origin for high plasmid yields and the DDX47-WT ORF flanked by Gateway att sites Kanamycin This study pDONR221DDX47-K74A Vector derivative of pDONR221 containing DDX47-K74A mutation flanked by Gateway att sites Kanamycin This study pDONR221DDX47-E175A Vector derivative of pDONR221 containing DDX47-E175A mutation flanked by Gateway att sites Kanamycin This study MATERIALS AND METHODS 130 Plasmid Description Resistance Reference/Source pT7-His6DDX47-WT Vector derivative of pET300/NT-DEST for rapid cloning with a Gateway® entry clone and subsequent high-level prokaryotic expression controlled by the strong bacteriophage T7 promoter. Vector containing DDX47-WT with an N-terminal 6xHis tag for protein purification Ampicillin Chloramphenicol This study pT7-His6DDX47-K74A Vector derivative of pET300/NT-DEST containing DDX47-K74A mutation with an N-terminal 6xHis tag for protein purification Ampicillin Chloramphenicol This study pT7-His6DDX47-E175A Vector derivative of pET300/NT-DEST containing DDX47-E175A mutation with an N-terminal 6xHis tag for protein purification Ampicillin Chloramphenicol This study PCMV: Cytomegalovirus promoter; ORF: Open reading frame. 5. BACTERIAL TRANSFORMATION AND HUMAN CELLS TRANSFECTION 5.1. Bacterial transformation Transformation of bacteria with exogenous DNA was carried out according to standard heat shock transformation protocol (Sambrook et al., 1989). 5.2. Human cells transfection All assays were performed 72 hours after small interfering RNA (siRNA) transfection and 24 or 48 hours after plasmid transfection. 5.2.1. siRNA transfection Concerning the high-throughput microscopy screening, we analyzed a collection of siRNAs targeting 3205 human genes (3205 out of 4796 siRNAs of DharmaconON TARGET Plus-Druggable genome siRNA library), in which four-siRNA pool for every targeted gene was used (Appendix 2). Other siRNA used are shown in Table M5. Tesis doctoral-Esther Marchena Cruz 131 Table M5. siRNAs used in this study. siRNA Time Source ON-TARGETplus Non-targeting Pool (D-001810) 72 h Dharmacon ON-TARGETplus SMARTpool human FANCD2 (L-016376-00) 72 h Dharmacon HDAC8 siRNAs: 5’-GAC GGA AAU UUG AGC GUA U dTdT-3’ 5’-GGA AUU GGC AAG UGU CUU A dTdT-3’ 72 h Dharmacon MECP2 siRNAs: 5’-GGA AAG GAC UGA AGA CCU G dTdT-3’ 5’-ACA CAU CCC UGG ACC CUA A dTdT-3’ 72 h Dharmacon DDX42 siRNAs: 5’-CGU AAA GGG UAU UCG AGA U dTdT-3’ 5’-GUU AAU AGA UCU CCG GCA U dTdT-3’ 72 h Dharmacon DDX47 siRNAs: 5’-GGA UGA AGC CGA CCG AAU A dTdT-3’ 5’-AGA AGA AAC GCU CGC GAG A dTdT-3’ 72 h Dharmacon MYOG siRNAs: 5’-GGA UGA AGC CGA CCG AAU A dTdT-3’ 5’-AGA AGA AAC GCU CGC GAG A dTdT-3’ 72 h Dharmacon PSMC5 siRNAs: 5’-CCA AGA ACA UCA AGG UUA U dTdT-3’ 5’-CAU ACG GAC UGU ACC UUU A dTdT-3’ 72 h Dharmacon PSMD7 siRNAs: 5’-AGA CGA UUC UGU AUG GUU U dTdT-3’ 5’-GAA AGU ACU UGA UGU AUC G dTdT-3’ 72 h Dharmacon DDX47 UTR siRNA: 5’-CUU CGA CUU UGA UUC CUU G dTdT-3’ 72 h Sigma ON-TARGETplus SMARTpool human SETX (L-021420-00) 72 h Dharmacon ON-TARGETplus SMARTpool human DDX23 (L-19861-01) 72 h Dharmacon ON-TARGETplus SMARTpool human UAP56 (L-003805-00) 72 h Dharmacon Cells were transfected with siRNA using DharmaFECT 1 (Dharmacon) at 40-50% confluence. Transfection in a well of 6-well plate was performed using the following protocol: - Mixture A (final volume 100 μl): 50 μl serum free-medium (medium without Fetal Bovine Serum (FBS)), 36 μl H2O, 9 μl 5X siRNA buffer (300 mM KCl, 30 mM HEPES-pH 7.5, 1.0 mM MgCl2) (Dharmacon) and 5 μl siRNA 20 μM (100 nM). - Mixture B (final volume 100 μl): 95 μl serum free-medium and 5 μl DharmaFECT1. MATERIALS AND METHODS 132 Each mixture was incubated at room temperature (RT) for 5 min. Then, Mixture A is added over Mixture B, mixed and incubated for 20 min. Meanwhile, medium was replaced by 800 μl antibiotic-free complete medium (medium with FBS but without antibiotics). Transfection solution was added carefully drop by drop to the cell culture and incubated for 2 hours. Afterwards, 1ml of complete medium was added. 5.2.2. Plasmid transfection using Lipofectamine 2000 or Lipofectamine 3000  For plasmid transfection using Lipofectamine 2000 (Invitrogen), cells were transfected at 80% confluence and were cultured in complete medium (2ml for 6-well plates). Transfection in a well of a 6-well plate was performed using the following protocol: - Mixture A (final volume 200 μl): 2 μg DNA in Opti-MEM (Gibco). - Mixture B (final volume 300 μl): 4 μl Lipofectamine 2000 in Opti-MEM. Each mixture was incubated at RT for 5 min, mixed and incubated for 5 min at RT. Transfection solution was added carefully drop by drop to the cell culture.  For plasmid transfection using Lipofectamine 3000 (Invitrogen), cells were transfected at 80% confluence and cell were cultured in complete medium (2 ml for 6-well plates). Transfection in a well of a 6-well plate was performed using the following protocol: - Mixture A (final volume 200 μl): 2 μg DNA and 4 μl of Enhancer Reagent in Opti-MEM (Gibco). - Mixture B (final volume 300 μl): 4 μl Lipofectamine 3000 in Opti-MEM. Each mixture was incubated at RT for 5 min, mixed and incubated for 5 min at RT. Transfection solution was added carefully drop by drop to the cell culture. 6. siRNA SCREENING METHODS ON TARGET Plus-Druggable genome siRNA library (Dharmacon, 77G-10465505) containing four-siRNA pools targeting human genes considered potential targets for therapeutic drugs was used for the screening of genes controlling Rloops. 3205 out of 4796 siRNAs of the library were analyzed. For a general schematic workflow of high-throughput screening, see Figure M1. Tesis doctoral-Esther Marchena Cruz 133 siRNA library preparation, transfection and immunofluorescence protocol were carry out using Hamilton Microlab Star available at Genomic Unit at CABIMER. This high-throughput screening was performed in collaboration with Sonia Silva, Lola P. Camino and Jose Javier Marqueta-Gracia. 6.1. siRNA library preparation siRNA pools were provided lyophilized (0.5 nmol of each siRNA pool) and distributed in 96-well plates. First, siRNAs were reconstituted to 20 μM with 25 μl of 1x siRNA buffer (60 mM KCl, 6 mM HEPES-pH 7.5, 0.2 mM MgCl2) using the Hamilton Microlab STAR with an 8 tip disposable tip head. Then, siRNAs were diluted to 1μM in siRNA buffer. An aliquot of 5 μl of siRNAs was transferred from the collection plates to dilution plates with 95 μl of 1x siRNA buffer. 30 μl of diluted siRNAs (1 μM) were dispensed to new 96-well V-bottom plates containing 30 μl of Opti-MEM (Gibco) to prepare siRNA master mix-plates for transfection protocol. Collection and dilution plates were immediately sealed using Aluminum Sealing Film (Axygen) and then frozen at -80 ºC so remaining siRNA could be used in future experiments. 6.2. Transfection protocol A reverse transfection protocol was followed by plating the cells onto pre-plated siRNA-transfection mixtures. First, 47 μl of each siRNA pool (0.5 μM) in Opti-MEM were transferred from siRNA master mix plates to a new single-use 96-well V-bottom plates, where 47 μl of the transfection reagent Lipofectamine 2000 (0.21 μl/well) in Opti-MEM was previously added. Then, the siRNA-transfection mixtures were gently mixed and incubated for 20 minutes at room temperature to allow siRNA-lipofectamine complex formation to occur. Meanwhile, transfection mixtures were prepared for positive control FANCD2 and UAP56 siRNA pools and the negative non-targeting control siRNA pool. Next, using Hamilton Microlab Star, two transfection-plates (duplicate-plates) were prepared. 20 μl of each siRNA-transfection mixtures/well were transferred to 96-well tissue culture-treated (TC-treated) polystyrene U-bottom plates (Perkin Elmer) (20 μl/well). In parallel, positive and negative control siRNA -transfection MATERIALS AND METHODS 134 mixtures were added to 6 wells and 4 wells each of duplicate-plates. These transfection plates were immediately sealed and placed at -80 ºC until use. For reverse transfection, U2OS-TR-AID cells in optimal growth phase are previously prepared. A final concentration of 33.3 nM siRNA was used to transfect 6000 cells/well. On the day of transfection, the transfection plates were thawed and 80 μl of U2OS-TR-AID cells per well (75,000 cells/ml in DMEM) were added to the assay plates. 4 hours after incubation, 50 μl/well of high-concentrate (3x) complete medium was added to adjust the normal concentration of the medium. For growth media and conditions, see Materials and Methods 1.1. AID induction U2OS-TR-AID is a stable cell line carrying a tetracycline-regulated AID gene cassette. Transcription was activated with doxycycline. This clone expresses high-levels of AID upon doxycycline addition, as previously validated by western blot and immunofluorescence analysis in our laboratory (Calderon, Jose). Thus, 24 hours post-transfection, for AID induction, U2OS-TR-AID cells were treated or not with doxycycline (6 μg/ml) and 48 hours later cells were harvested. 6.3. Fixing and staining 72 hours post-transfection, the reaction was stopped by manually tossing the media and fixing the cells in 4% formaldehyde in PBS for 10 min at RT. Cells were permeabilized with 70% ethanol for 5 min at -20 ºC, 5 min at 4 ºC and washed three times in PBS. The immunofluorescence protocol was performed using Hamilton Microlab Star. The plates were blocked by adding 100 μl blocking solution (3% bovine serum albumin (BSA) in PBS). After 1 h, the blocking solution was removed and 50 μl of anti-H2A.X Phospho(Ser 139) antibody (1:1000 in blocking solution; Biolegend 613402) was added and plates were incubated at RT for 1h. The plates were then washed three times. 50 μl of the secondary antibody goat-anti-mouse antibody Alexa-488 conjugated (1:1000 in blocking solution; Molecular Probes, A11029) was added, and the plates were incubated at RT for 1 h. The plates were then washed three times and 100 μl of Hoechst (AnaSpec Inc. Cat# 83219) was added in each well to stain the nucleus. Plates were sealed and stored at 4 ºC until imaging. As indicated before, the different Tesis doctoral-Esther Marchena Cruz 135 steps in the immunofluorescence protocol were performed using Hamilton Microlab Star. 6.4. Imaging Data acquisition was performed with ImageXpress Micro Electron (Molecular Devices, 137239) using a 40x objective. The entire 96-well plate (Perkin Elmer) was scanned with a 20 nW 690 nm laser and set up for two channels of acquisition. γH2AX-Alexa-488 fluorescence was acquired using a 472/30 nm excitation filter, and Hoechst fluorescence was acquired using 377/50 nm excitation filter. Randomly fields were acquired and analyzed (≃ 200 cells/well). H2AX foci were quantified by automated scoring using MetaXpress software (version 4.0.0.24 Molecular Devices-granularity application). Figure M1. Workflow for high-throughput screening of factors involved in R-loop metabolism. Schematic workflow for testing the human druggable siRNA library (3205 siRNAs) for γH2AX foci formation in the absence or presence of AID. U2OS-TR-AID cell line are transfected and treated with doxycycline for 48h to induce AID expression. An example of γH2AX detection (green) by automated microscopy of cells with or without AID expression is shown. Nuclei were detected by DAPI (blue). MATERIALS AND METHODS 136 6.5. Data analysis and statistical analysis Here we describe the procedure used to analyze the raw data and to determine which siRNA transfected cells conferred a H2AX increase after AID induction and were considered positive candidates. We used MetaXpress in order to measure Granule Count that characterizes γH2AX foci associated to each siRNA and condition in each well. We calculated the percentage of cells with ≥5 γH2AX foci before and after AID induction of each siRNA depletion. The screening was performed in duplicate, and siRNAs whose depletion lead to an increase (≥1.2) in the percentage of cells with γH2AX foci after AID induction (siRNA +DOX), versus non-induced conditions (siRNA -DOX), that was higher (≥1.2) to that observed in control cells (siRNA +DOX versus siC +DOX) were selected as candidates. To take into account the variability of the data, per-plate and per-date, each condition was always compared versus noninduced conditions and control cells located in the same plate. The same procedure was performed to calculate the percentage of cells with ≥10 γH2AX foci after AID induction. The Z-score and the reproducibility score were calculated to take into consideration when candidate selection. Z − 𝑠𝑐𝑜𝑟𝑒 = % − 𝜇 σ %= percentage of cells with ≥5 γH2AX foci / percentage of cells with ≥10 γH2AX foci. μ= Mean of each plate. σ= Standard deviation of each plate. 𝑅𝑒𝑝𝑟𝑜𝑑𝑢𝑐𝑖𝑏𝑖𝑙𝑖𝑡𝑦 = 2𝑛 n= Number of times that a candidate meets the criteria described above. For positive control ratios in the first round of the high-throughput screening, see Appendix 3. Tesis doctoral-Esther Marchena Cruz 137 7. SITE-DIRECT MUTAGENESIS To generate the DDX47 mutant forms, we used the Q5 Site-Directed Mutagenesis Kit (New England Biolabs) according manufacturer’s recommendations. The protocol consists of a first exponential amplification (PCR) with the mutagenic primer pairs (Table M7), using the pCMV-DDX47FLAG plasmid as template DNA. The primers were designed with the NEBaseChanger tool to optimize the sequence and the annealing temperature. Next, a Kinase, Ligase and DpnI (KLD) treatment for 5 min at RT was performed in order to phosphorylate, ligate the PCR product and remove the template. Competent cells were transformed with this reaction, transformants were checked by restriction enzymes digestion and finally plasmidic DNA was sequenced to confirm the presence of the mutation. For a schematic representation of mutations in the helicase core of DDX47, see Figure R17A. 8. DEVELOPMENT OF TOOLS FOR IN VITRO ANALYSIS AND IN VIVO OVEREXPRESSION OF DDX47 8.1. Tools for in vitro analysis We used the Gateway cloning System (Invitrogen) to generate a vector for expression of DDX47 WT and DDX47 mutants in E. coli. Gateway technology takes advantage of the site-specific recombination properties of bacteriophage lambda to provide a highly efficient way to move the gene of interest into multiple vector systems. First, we amplified DDX47 ORF with specific primers for addition of attB sites. Then, we cloned DDX47 ORF sequences into a Gateway® entry vector pDONR221 to create an entry clone, and after we generated an expression clone by a recombination reaction between the entry clone and the destination vector. In this case, our destination vector was pET300/NT-DEST (Table M4) designed to allow high-level inducible expression from T7 promoter of recombinant proteins in E. coli (N-terminal 6xHis tag). Expression was induced in BL21 E. coli transformant exponential cultures by adding IPTG (0.2 mM) at exponential cultures for 3h (37ºC) or overnight (16ºC). The induction was tested by Coomassie blue staining and western blot (Materials and Methods 17). MATERIALS AND METHODS 144 Lysis solution: Kit CometAssay ES II (Trevigen 4250-050-ES). Alkaline unwinding solution: 0.4 g of NaOH, 250 μl 200 mM/pH10 EDTA were diluted up to 50 ml with H2O. Alkaline electrophoresis solution: 8g of NaOH, 2 ml of 500 mM/pH8 EDTA were diluted up to 1 l with H2O. For comet assays analyses at least three independent experiments were performed. More than 200 cells were scored in each experiment (see Materials and Methods 14). 13.3. RNA-DNA hybrids detection 13.3.1. RNA-DNA hybrids immunoprecipitation (DRIP-qPCR) DRIP assays were performed by immunoprecipitating DNA–RNA hybrids using the S9.6 antibody from gently extracted and enzymatically digested DNA, treated or not with RNase H (New England Biolabs, USA) in vitro as described (M. García-Rubio et al., 2018; M. L. García-Rubio et al., 2015; Herrera-Moyano et al., 2014), with minor modifications. After 72 h of siRNA transfection, pellet from a confluent 10-cm plate of cells was collected using accutase, washed in PBS and resuspended in 800 μl of 1X TE. Then, 20.75 μl SDS 20% and 2.5 μl proteinase K (20 mg/ml) were added and pellet was incubated at 37 °C overnight. DNA was extracted gently with phenol-chloroform in phase lock tubes (VWR, USA). Precipitated DNA was spooled on a glass rod, washed 2 times with 70% EtOH, resuspended gently in 1X TE and digested overnight with 50 U of HindIII, EcoRI, BsrGI, XbaI and SspI. For the negative control, half of the DNA (16 μg) was treated with 5 μl RNase H overnight. In parallel, S9.6 antibody (3 μg/sample) was incubated overnight at 4ºC with Dynabeads Protein A (Invitrogen) (30 μl/sample) in 1X binding buffer (final volume: 60 μl/sample). 5 μg of the digested DNA, untreated or treated with RNase H, were bound to S9.6 antibody-dynabeads complexes during 2h at 4ºC for immunoprecipitation. Next, the beads were washed 3 times with 1X binding buffer. DNA was eluted in 180 μl elution buffer, treated 45 min with 7 μl proteinase K at 55ºC and cleaned with the NucleoSpin Gel and PCR Clean-up (Macherey-Nagel, USA). 1 μg of the digested DNA, untreated or treated with RNase H, was used for input DNA of each condition, in which similar proteinase K treatment and purification were performed. Tesis doctoral-Esther Marchena Cruz 145 Quantitative PCR (qPCR) of immunoprecipitated DNA (IP) fragments and input DNA was performed on a 7500 Fast Real-Time PCR System (Applied Biosystems, Carlsbad,CA). Primers used are listed in Table M7. 10X Binding buffer: 100 mM NaPO4 pH 7.0, 1.4 M NaCl, 0.5% triton X-100. Elution buffer: 50 mM Tris pH 8.0, 10 mM EDTA, 0.5% SDS. DRIP quantification and normalization: Input and immunoprecipitated (IP) were eluted in 150 μl of 1X TE. 2 μl of input and IP were used for qPCR. Changes in the abundance of RNA-DNA hybrids in each region were determined as the percentage of input recovered for each immunoprecipitated sample using the equation 2[ CtIP - (CtINPUT - log2 DF)]. Notes: (Ct -threshold cycle); (DF - dilution factor). 13.3.2. S9.6 immunofluorescence Cells were cultured on glass coverslips and fixed in methanol (see Materials and Methods 12). When required, for RNase treatments cells were incubated in their respective commercial buffers at 1x containing 40 U/ml RNase III (AM2290, Ambion) or 60 U/ml RNase H (M0297S, NEB), for 30 min at 37°C. Next, for S9.6 immunofluorescences, coverslips were blocked in 3% BSA in PBS for 5 h at 4ºC and incubated with anti-S9.6 (Hybridoma cell Line HB-8730, 1:500) and antinucleolin (ab50279, 1:1000) or anti-FLAG (ab1257, 1:1000) primary antibodies diluted in 3% BSA in PBS o/n at 4 °C, washed in PBS three times and incubated with the subsequent secondary antibodies conjugated with Alexa Fluor diluted in 3% BSA in PBS (1:1000) for 1 h at RT in darkness. Washed twice for 5 min, DAPI staining and mounting as described above. More than 200 cells from each experiment were scored (see Materials and Methods 14). 14. MICROSCOPY IMAGES ACQUISITION, DATA ANALYSIS AND STADISTICAL ANALYSIS 14.1. Fluorescence microscopy Data acquisition was performed with LAS AX (Leica) equipped with a DFC390 camera. A 63x objective was used for immunofluorescence (γH2AX, AID, RNAPI, Fibrillarin, Nucleolin, DDX47 IF, PLA, S9.6, FLAG and Nucleolin IF) and a 10x objective was used for comet assays. MATERIALS AND METHODS 146 14.2. Data analysis - γH2AX foci and RNAPI intensity measurements were analyzed and processed with the MetaMorph v7.5.1.0. software using the granularity application. - S9.6 signal intensity per nucleus was analyzed and processed with the MetaMorph v7.5.1.0. software using the multi wavelength cell scoring application. The S9.6 signal corresponding to the nucleolus area was previously removed using the nucleolin signal and granularity application. The S9.6 signal intensity corresponding to nucleolus per cell was analyzed using the nucleolus area and granularity application. - FLAG and PLA signal intensity per nucleus were analyzed and processed with the MetaMorph v7.5.1.0. software using the multi wavelength cell scoring application. - Nucleolar total area measurements were analyzed and processed with the MetaMorph v7.5.1.0. software using the granularity application and nucleolin immunofluorescence signal. - Comet assays tail moments were analyzed using TriTek CometScore Professional (version 1.0.1.60) software. Tail moment (TM) reflects both the tail length (TL) and the fraction of DNA in the comet tail (TM=%DNA in tail x TL/100). - EU signal intensity corresponding to nucleolus per cell was analyzed using the nucleolus area and granularity application with the MetaMorph v7.5.1.0. software. EU signal intensity per nucleus was analyzed and processed using the multi wavelength cell scoring application, where EU signal corresponding to the nucleolus area was previously determined using the nucleolin signal and granularity application. For all experiments, at least three biological repeats (n) were performed. More than 200 cells were scored in each repeat. 14.3. Statistical analysis - γH2AX foci number per cell, PLA signal intensity and S9.6 intensity per nucleus: Graphs show the median of the measurements from at least three biological repeats. Data were analyzed with GraphPad Prism software. For statistical analysis, Mann-Whitney U-test, two tailed was performed and P Tesis doctoral-Esther Marchena Cruz 147 value < 0.05 was considered as statistically significant. (***, P < 0.001; **, P < 0.01; *, P < 0.05). - DRIP: Graphs show changes in the abundance of RNA-DNA hybrids in each region were determined as the percentage of input recovered for each immunoprecipitated sample normalized with respect to the siC control from at least three biological repeats. Data were analyzed with EXCEL program and GraphPad Prism software. For statistical analysis, Student’s t-test, one tailed was performed and a P value < 0.05 was considered as statistically significant. - ChIP: Graphs show values represent the percentage of the precipitated DNA (IP) to input DNA (Input) from at least three biological repeats. Data were analyzed with EXCEL program and GraphPad Prism software. For statistical analysis, Student’s t-test, one tailed was performed and a P value < 0.05 was considered as statistically significant. - Comet assay: Graphs shows the median of tail moment from at least three biological repeats. Data were analyzed with EXCEL program and GraphPad Prism software. For statistical analysis, Mann-Whitney U-test, two tailed was performed. - Nucleolar total area per nucleus, RNAPI intensity and EU experiments. Graph shows the mean of medians of measurements from at least three biological repeats. Data were analyzed with GraphPad Prism software. For statistical analysis, Student’s t-test, two tailed was performed and a P value < 0.05 was considered as statistically significant. (***, P < 0.001; **, P < 0.01; *, P < 0.05). 15. CHROMATIN IMMUNOPRECIPITATION (ChIP) ASSAY After 72 h of siRNA transfection, HeLa cells were crosslinked for 10 min with 11% formaldehyde crosslinking solution, resuspended in 2.5 ml of cell lysis buffer 1, then centrifuged and 1 ml of nuclei lysis buffer 2 was added. Chromatin was sonicated on the maximum intensity setting, with eight pulses of 30 s on and 30 s off in Bioruptor (Diagenode), to obtain approx. 500 bp fragments. For each immunoprecipitation, 30 μg of chromatin were diluted up to 1100 μl with IP buffer. 100 μl and 1000 μl of diluted chromatin were used for input and MATERIALS AND METHODS 148 immunoprecipitation, respectively. Chromatin was incubated overnight at 4 ºC with ChIP-grade antibodies (Table M1). A negative control with IgG rabbit or mouse antibody was used to calculate the background signal. Chromatinantibody complexed were immunoprecipitated for 2 h with 30 μl of Dynabeads Protein A (rabbit)/G (mouse) (Invitrogen) at 4 ºC and washed once with wash buffer 1, once with wash buffer 2, once with wash buffer 3 and twice with 1X TE. Input and immunoprecipitate (IP) were then un-crosslinked in TE - 1% SDS at 65 ºC overnight, treated with proteinase K for 1 h at 37 ºC and phenol-chloroform purified. Finally, DNA was resuspended in 50 μl of MQ H2O. Primer pairs used for amplification are listed Table M7. Samples were run in 7500 Fast Real-time PCR system (Applied Biosystem). Results were analyzed with 7500 System Software V2.0.6. Crosslinking solution - formaldehyde: 50 mM HEPES pH 8, 0.1 M NaCl, 1 mM EDTA pH 8, 0.5 mM EGTA. Formaldehyde was added fresh, in the appropriate concentration to obtain a final concentration of 1% formaldehyde after the addition to the medium. Cell lysis buffer 1: 5 mM PIPES pH 8, 85 mM KCl, 0.5% NP-40, 1 mM PMSF and protease inhibitor cocktail. Nuclei lysis buffer 2: 1% SDS, 10 mM EDTA, 50 mM Tris-HCl pH 8.1, 1 mM PMSF and protease inhibitor cocktail. IP buffer: 0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl pH 8.1, 167 mM NaCl. Wash buffer 1: 0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 150 mM NaCl. Wash buffer 2: 0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 500 mM NaCl. Wash buffer 3: 0.25 M LiCl, 1% NP-40, 1% Sodium deoxycholate, 1 mM EDTA, 10 mM Tris-HCl pH 8.1. ChIP quantification and normalization: Input and immunoprecipitate (IP) were eluted in 50 μl of MQ H2O. 2 μl of 1:10 dilutions of the Input and 2 μl of IP were used for qPCR. A calibration curve with five 5-fold serial dilutions of a standard Tesis doctoral-Esther Marchena Cruz 149 DNA sample was calculated for taking into account the amplification efficiency (based on the slope of the standard curve) of each qPCR reaction. Considering amplification (AF) and dilution (DF) factor (AF(40 – Ct))DF, we calculate the absolute quantification of each sample. IP/Input ratios in the different regions were calculated and multiplied by 100. The IP signal of the IgG (negative control) was considered as background. 16. POLYMERASE CHAIN REACTION (PCR) ANALYSIS 16.1. Non-quantitative PCR PCRs carried out to amplify DNA fragment to be cloned were performed following standard and manufacture’s protocols with the polymerases described in Materials and Methods 2.3. DNA primers used are listed in Table M7. 16.2. Quantitative PCR analysis Real-time quantitative PCRs (qPCRs) were performed on a 7500 Fast Real-Time PCR system (Applied Biosystems, Carlsbad, CA). For PCRs, 6 μl H2O, 2 μl primer mixture (each 10 μM), 2 μl DNA and 10 μl SYBR® green qPCR Mix (Bio-rad) were used. The following PCR reaction was used: 1 cycle (10 minutes 95 °C), 40 cycles (15 s 95 °C and 1 minute 65 °C) and 1 dissociation cycle (15 s 95 °C, 1 minute 65 °C, 15 s 95 °C and 15 s 60 °C). DNA primers were designed using Primer express 3.0 Software (Applied Biosystems) and are listed in Table M7. qPCR primers were validated by qPCR by establishing that each pair of primers had the same amplification efficiency (the slope of the 10-fold serial dilutions of a calibration curve was between -3.3 and -3.4). 16.2.1. Reverse Transcription quantitative PCR (RT-PCR) analysis Relative qPCRs were used to determine the relative mRNA levels in human cells. cDNA was obtained from total RNA extracted using RNeasy Mini Kit (Qiagen) (1 μg) by reverse transcription using QuantiTect Reverse transcription (Qiagen), where an optimized blend of oligo-dT and random primers is used. mRNA expression values were normalized to mRNA expression of the Hypozanthine phosphoRibosylTransferase (HPRT) or Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) housekeeping genes. MATERIALS AND METHODS 150 16.2.2. Primer pairs used for amplification Primer pairs used for amplification are listed Table M7. Table M7. DNA primers used in this study. Non-quantitative PCR primers Primer Sequence 5’ to 3’ Use VP 1.5 Fwd GGACTTTCCAAAATGTCG Sequencing of prey inserts XL39 Rv ATTAGGACAAGGCTGGTGGG Sequencing of prey inserts DDX47 Fwd GAGACAGAGGTTGACAAGATCC Sequencing of prey inserts DDX47 Rv AGTAGCAAAGCTGTTCTCTGG Sequencing of prey inserts T7 promoter TAATACGACTCACTATAGGG Sequencing of prey inserts T7 terminator GCTAGTTATTGCTCAGCGG Sequencing of prey inserts M13 Fwd GTAAAACGACGGCCAGT Sequencing of prey inserts M13 Rv CAGGAAACAGCTATGACC Sequencing of prey inserts K74A Fwd K74A Rv TGGCTCTGGAGCGACAGGCGCC GTTTCTGCAAGCCCAATG Plasmid construction: pCMVDDX47-K74A-FLAG E175A Fwd E175A Rv GGTCATGGATGCAGCCGACCGA AAGTATTTGAGAGCTCTCAAGTTGAAAC Plasmid construction: pCMVDDX47-E175AFLAG DDX47 gateway Fwd GGGGACAAGTTTGTACAAAAAAGCAGGCTTGGC GGCACCCGAGGAACACGATTCTCCG Plasmid construction: pT7His6DDX47WT/K74A/175A DDX47 gateway Rv GGGGACCACTTTGTACAAGAAAGCTGGGTTTTAC TAACGGCCTTTCCGCTTCTTCATTTTTCC Quantitative PCR primers Primer Sequence 5’ to 3’ Use HDAC8 Fwd HDAC8 Rv TGGGAGGAGGAGGCTATAACC CCGGTCAAGTATGTCCAGCAT Relative mRNA expression MECP2 Fwd MECP2 Rv CGCTCTGCTGGGAAGTATGAT CCACTTTAGAGCGAAAGGCTTTT Relative mRNA expression DDX42 Fwd DDX42 Rv CAGCAGATCCATGCAGAATGTAA ACGGCCACTGATCGAAGATT Relative mRNA expression DDX47 Fwd DDX47 Rv TCCTACCCATTCCAAGGATTACA CGGAGCGCCCAGCTCTA Relative mRNA expression PSMC5 Fwd PSMC5 Rv CGAGAACGGCGAGTCCAT TGACCTTGGCTACTGCCATCT Relative mRNA expression PSMD7 Fwd PSMD7 Rv CAACCGAATCGGCAAGGTT TGCCATGACCCCAAAAGC Relative mRNA expression Tesis doctoral-Esther Marchena Cruz 151 Quantitative PCR primers Primer Sequence 5’ to 3’ Use APOE Fwd APOE Rv AAGCTGGAGGAGCAGGCC ACTGGCGCTGCATGTCTTC Relative mRNA expression RPL13 Fwd RPL13 Rv GGGAGCAAGGAAAGGGTCTTA ACAATTCTCCGAGTGCTTTCAAG Relative mRNA expression HPRT Fwd HPRT Rv GGACTAATTATGGACAGGACTG TCCAGCAGGTCAGCAAAGAA Relative mRNA expression GAPDH Fwd GAPDH Rv TGCACCACCAACTGCTTAGC GGCATGGACTGTGGTCATGAG Relative mRNA expression APOE Fwd APOE Rv GGGAGCCCTATAATTGGACAAGT CCCGACTGCGCTTCTCA DRIP ChIP RPL13 Fwd RPL13 Rv GCTTCCAGCACAGGACAGGTAT CACCCACTACCCGAGTTCAAG DRIP ChIP 18S Fwd 18S Rv GGCGTCCCCCAACTTCTTA GGGCATCACAGACCTGTTATTG DRIP ChIP 28S Fws 28S Rv GAATCCGCTAAGGAGTGTGTAACA CTCCAGCGCCATCCATTT DRIP ChIP 5’ETS Fws 5’ETS Rv CAGGCGTTCTCGTCTCCG CACCACATCGATCGAAGAGC ChIP IGS Fws IGS RV GTTGACGTACAGGGTGGACTG GGAAGTTGTCTTCACGCCTGA ChIP PHLDA2 Fwd PHLDA2 Rv CAGCGGAAGTCGATCTCCTT GAGCGCACGGGCAAGTAC DRIP ChIP YIF1A Fws YIF1A Rv TCTCCCAATTGGCCACTGA TGTTGCCTTCTGCCGAATC DRIP ChIP 17. PROTEIN ANALYSIS 17.1. Bacteria cell protein extraction Pellet of bacteria was collected at 12 000 g for 5 min at 4ºC and resuspended using 1X Loading Buffer. Prior to gel loading, samples were boiled for 5 min and centrifuged 5 min at 3000 rpm at RT. 17.2. Human cell protein extraction Pellet of HeLa cells was collected using accutase, washed in cold PBS and kept in ice. Proteins were extracted by adding 1X Lammeli buffer (10 μl/75 000 cells) with gently pipetting up and down. Samples were centrifuged 5 min at 3000 rpm. The lysate was sonicated on the maximum intensity setting, with pulses of 30 s on and 30 s off for 5 min in Bioruptor (Diagenode). Prior to gel loading samples were boiled for 5 min and centrifuged 3 min at 3000 rpm at RT. 4X Lammeli buffer: 200 mM Tris-HCl, 40% glycerol, 8% SDS, 0.4% Bromophenol Blue, 400 mM β-mercaptoethanol. MATERIALS AND METHODS 152 17.3. SDS-PAGE Proteins were separated in 29:1 acrylamide:bis-acrylamide gels with concentrations appropriate to the molecular size of the proteins of interest or in 4-20% gradient SDSPAGE CriterionTM TGX TM Precast Gels (BioRad) and SDS-PAGE was performed according to previously described method (Laemmli, 1970). Electrophoresis were performed in a Mini-PROTEAN 3 Cell in Running Buffer at 100 V. Page Ruler Prestained Protein Ladder (26617, ThermoFisher) and PageRuler Plus Prestained Protein Ladder (26620, ThermoFisher) were used as protein markers. Running buffer: 25 mM Tris base pH 8.3, 194 mM glycine, 0.1% SDS. 17.4. Western Blot analysis For Western blot, proteins were wet-transferred using Trans-Blot system (Biorad) for 1 h at 300 mA in 1X Transfer Buffer with 20% methanol or o/n at 30 V at 4 °C. Membranes were stained with Ponceau S (0.1% w/v Ponceau -SIGMAin 5% acetic acid) to check protein loading and correct transference. 5X Transfer buffer: 6 g/L Tris base, 28.8 g/L glycine, 0.5% SDS. 17.5. Non-fluorescence WB Proteins were transferred to a nitrocellulose membrane (Hybond-ECL, GE Healthcare). Membranes were blocked with 1X TBS - 0.05% Tween 20 - 5% milk for 1 h. Primary antibodies were incubated o/n at 4ºC at the indicated concentrations (Table M1) in 1x TBS - 0.05% Tween 20 - 5% milk. After three washes of 5 min each one in 1x TBS – 0.05% Tween 20, membranes were incubated with the corresponding secondary antibodies conjugated with the horseradish peroxidase at the indicated concentrations (Table M2) in 1x TBS, 0.05% Tween-20 with 5% milk for 1 h at room temperature and washed again. Finally, SuperSignal West Pico Plus (Thermo) was used for chemiluminescence detection. Protein levels were calculated with Chemidoc Imaging system using as loading controls: vinculin or actin antibodies. 17.6. Expression of His6-DDX47-WT and mutant forms BLB21 Rosetta Electrocompetent Cells (70954-3, Merck) were electroporated with 0.5 µg of pT7-His6-DDX47-WT, K94A and E175A plasmids, selected with ampicillin (75 µg/ml) and chloramphenicol (34 µg/ml). Plates were incubated at Tesis doctoral-Esther Marchena Cruz 153 37 ºC overnight. A single colony was inoculated into 5 ml LB + ampicillin (75 µg/ml) + chloramphenicol (34 µg/ml) culture and incubated at 37 ºC with shaking at 200 rpm overnight. Overnight cultures were diluted to an optical density (OD600nm) of 0.4. At this moment, we induced His6-DDX47 expression by adding a final concentration of 0.2 mM IPTG to the culture. His6-DDX47 was expressed at 37 ºC for 3h or at 16ºC overnight with 200 rpm shaking. For bacteria cells protein extraction and western blot analysis, see Materials and Methods 17.1. For protein staining, acrylamide gels were incubated with Coomassie Brilliant Blue for 1h with gentle shaking at RT and then destained with Coomassie destaining solution for 15 min. Coomassie solution: 2.5 g Coomassie Brilliant Blue, 400 ml methanol, 100 ml glacial acetic acid and 500 ml MQ H2O. Coomassie destaining solution: 40 ml methanol, 100 ml glacial acetic acid and 500 ml MQ H2O.