dynamics during DNA repair investigated via chromatin-directed proteomics Gianluca Sigismondo 1, Lavinia Arseni 3, Jeroen Krijgsveld 1,2 1 German Cancer Research Center (DKFZ), Division of Proteomics of Stem Cells and Cancer, INF 581, Heidelberg, Germany 2 Heidelberg University hospital, INF 672, Heidelberg, Germany [email protected] 3 German Cancer Research Center (DKFZ), Division of Molecular Genetics, INF 580, Heidelberg, Germany Website: https://www.dkfz.de/en/proteomik-stammzellen-krebs/index.php

- Abstract DNA lesions predispose to genomic instability, a hallmark of cancer; therefore cells have evolved repair pathways to solve those harmful insults. Double-strand breaks (DSBs) represent the most lethal DNA damage first marked by the phosphorylation of the histone H2A.X (γH2A.X) which triggers the recruitment of sensor belonging to either the error-prone non-homologous end joining (NHEJ) or the efficient homologous recombination (HR) pathway. It is now established that chromatin has an active role also in DNA repair, thus its characterization at DSB repair foci is essential to better understand the coordinate action of the repair mechanisms and to identify novel players participating in tumor-associated apoptotic resistance and cell survival. Here we dissect chromatin changes upon exposure to ionizing radiations through multiple proteomics-based approaches. We applied the Selective Isolation of Chromatin-Associated strategy (ChIP-SICAP; Rafiee, 2016) to investigate the interactors of core NHEJ, HR proteins and γH2A.X while bound to the DNA or in the chromatin soluble fraction. Through a click chemistry-assisted procedure we profiled the configuration of DNA-bound proteins during DSBs repair; finally we analyzed the histone post-translational modifications (hPTMs) cross-talk at mono-nucleosomes marked by γH2A.X. Our integrated analysis identified the dynamics of expected chromatin determinants during the DNA repair and interestingly suggested the role for new candidates specifically enriched- upon DSB formation. Validation experiments based on monitoring of DSB foci formation and resolution in AID-DIvA cells proficient or knock-down cells provided evidence of a role for novel candidates in DNA repair. FACS-based analysis of Traffic-light Reporter (TLR) isogenic- cells upon silencing of proteins identified by MS characterized their functional role in NHEJ, HR or pathway choice. Furthermore, we defined hPTMs associated with γH2A.X-marked mono-nucleosomes and their dynamics during DSB resolution. This analysis corroborated expected enrichments (e.g. H4K20me1/me2) and provided insights on new modifications specifically From Freeman and Monteiro Cell Communication and Signaling 2010 enriched at γH2A.X-nucleosomes.

1. hPTMs cross-talk during DNA damage 2. Functional ON-chromatin interactors of NHEJ, HR and γH2AX identified by SICAP

A) Time-course hPTMs profiling upon IR at mono-nucleosome resolution A) Selective Isolation of Chromatin-Associated proteins in NHEJ and HR B) Dynamic profiling of γH2AX-associated proteins during DSBs repair

Lysis and nuclei Mononucleosome PTM Protein ArgC-like Intensity-based LC-MS/MS analysis SILAC cell labeling, Streptavidin pull-down and Reverse cross-linking, fractionation preparation ChIP separation digestion hPTM quantification Chromatin Antibody release SILAC cell labeling, Streptavidin pull-down and Reverse cross-linking, formaldehyde cross-linking, separation between on- and tryptic digestion and Chromatin Antibody release immunoprecipitation and DNA labeling formaldehyde cross-linking, enrichment of on-chromatin tryptic digestion and chromatin preparation off-chromatin interactors SP3 peptide clean up immunoprecipitation and DNA labeling 8 h.p.i chromatin preparation interactors SP3 peptide clean up M 0 0.5 1 4 8 MarkerInput ChIP Relative

100

c e XIC modified peak e n Abundance = x100 Ctr R v Ctr i a IgG IgG γH2AX t 4 d ∑ XIC all peaks 1000bp HN O a (RA)

l -chromatin n 50 OFF e

Ab u 500bp RN B

b 8 N CD3 R ChIP ChIP

150bp A Tdt 1 O 0 Relative RA ChIP Peptides B post irradiation (h) 50bp = Light (K R ) Light (K R ) 0.5 CH3 0 50 100 Enrichment RA Input 0 0 Bait Y 0 0 0 Time (min) B Bait for MS 1:1 B Bait B UT mix H2A H2A B Bait B 1:1 Peptides Peptides Strep B Bait Tdt Y 4 ChIP mix B Bait for MS Medium (K4R6) Strep Bait for MS Medium (K R ) 1:1 Target B 4 6 Tdt Y ON-chromatin mix ChIP ON-chromatin IR γH2AX γH2AX B) Two different scenarios for hPTMs associated with γH2AX 1 DNA for 0.5 ChIP 5 Heavy (K R ) Heavy (K R ) 8h 8 10 qPCR/Seq (p.i.h) 0 8 10 2 4h Ctr= UT:IR 1to1 subjected to IgG ChIP UT= untreated IR = irradiated 5Gy, 1h recovery 1 1h

H2AX -1 γ 30’ RA on UT -2 B2) Identification of γH2AX-specific interactors and their dynamic characterization upon DNA damage B) Chromatin interactors UT XIC ratio (log2) A2) Modulation of ON-chromatin interactors in NHEJ or HR repair pathways upon DNA damage -5 untreated 0.5h recovery 1h recovery γH2AX γH2AX 8h 5 of the DDR sensor Z-score over UT 5 ChIP-SICAP time course 4h RPA_ON RAD50_ON MDC1_ON 4 3 4 hPTM γH2AX were 1.5 UT IR UT IR UT IR 4 0.5h 1h 4h 8h 0.5h 1h 4h 8h

1h 3 2 Input MDC1 3 0.5 profiled in a time-

133 89 215 3 MDC1, ATAD2, Loading... Loading... 30’ 10 1 0 0 17 10 Loading... -0.5 RA on UT TP53BP1 PHIP 2 1 TPX2 2 -1.5 UT SMARCE1 2 course experiment

1 -3

BRG1 1 -Log pvalue -Log pvalue -Log pvalue 1 MDC1 1 MDC1 during DSB repair RAD50 HMGN2 RE UT ATR 0 0 RAD52 HMGA1 MDC1 0 1 elicited by sub-lethal NBN HNRNPA3 -1 0 1 2 3 4 5 -1 0 1 2 3 4 -1 0 1 2 3 4 POLR1E un un un un un un un un un un un un un RPA1 0 Ac Ac Ac Ac Ac Ac Ac Ac Ac BRG1 5 9 29 64 36 95 56 79 20 18 27 77 HMGA2 18 14 14 56 64 36 36 KAc KAc me1 me1 me1 me2 me1 me1 me1 me1 me2 me1 me1 me1 me2 me1 me2 me3 me2 me1 me2 me3 MRE11A YBX1

118 γH2AX - H2A γH2AX - H2A γH2AX - H2A K K 118

K 5 doses of IR. At each 9 4Kun 2KAc 3KAc 4KAc K K K K K K K K K K K K K K R 18 79 20 20 29 95 27 36 36 36 79 77 36 27 56 36 36 36 36 K /K K NUMA1 MATR3 K TOP3A NuMA K K K K K K K K K K K K K R SFPQ XRCC6 4h recovery 8h recovery time point, IgG was RPA2 5 γH2AX-specific 0 XRCC5

5 γH2AX Ac/K me1/K SFRS3 K 9Ac 9 27 me2/K me1/K me1/K me1/K RPA1 MRE11A Histone H2A K Histone H4 proteins per experiment time course

K the technical control 4 27 27 27 27 0 4 Log2 Ratio IR/UT R2 K K K K -1 (over UT) Log2 Ratio IR/UT R2 Log2 Ratio IR/UT R2 0.5h 1h 4h 8h Loading... Loading... 3 MDC1 while histone H2A RAD50 3 Histone H3 n = 133 n = 89 n = 215 N.e xp N. proteins 2 RFC1 recruited recruited recruited 2 was used as internal SAFB NBN NONO untreated upon IR untreated upon IR HCC5 evicted evicted evicted MDC1 1 283

-Log pvalue SFPQ 1 1 TPR -Log pvalue 2 198 control in ChIP to stable -1 stable stable SUMO2 -2 ILF3 Input ChIP Input ChIP Input ChIP Input ChIP -1 3 181 0 M WDHD1 Bait Bait Bait 0 THRAP3 characterize γH2AX- -1 0 1 2 3 4 5 -1 0 1 2 3 4 5 4 144 H4 H4 H4 H4 H4 A H4 H4 A H4 M H3 H3 H3 H3 H3 H3 A H3 H3 -2 -1 0 1 2 -1 0 1 -3 -2 -1 0 1 2 3 Log2 Ratio IR/UT R1 Log2 Ratio IR/UT R1 Log2 Ratio IR/UT R1 γH2AX - H2A γH2AX - H2A specific interactors H2A H2AX H2A H2AX H2A H2AX H2A H2AX H2B H2B H2B H2B H2B H2B H2B H2B P P P P B3) Validation of the DSB-promoted interaction between endogenous THRAP3 and γH2AX M M A3) HR or NHEJ canditate selection A) ON-chromatin interactors of core Hoechst THRAP3 γH2AX Colocalised A A A A) NHEJ B2) Volcano plots show the fold-change HR components of the MRN (RAD50) IF P phosphorylation A acetylation M methylation complex, HR (RPA) or NHEJ (MDC1, difference between γH2AX and H2A, NHEJ TP53BP1) DSB repair pathways MDC1 is depicted as positive control. A) Native preparation of mono-nucleosomes from untreated as well as RPA were characterized by an optimized untreated γH2AX-specific interactors identified at NuMA MDC1 TRA2b THRAP3 Hoechst THRAP3 γH2AX Colocalised all the time points are reported in the cells during the DNA damage response was used as input for γH2AX TP53BP1 SMARCE1 ChIP-SICAP protocol in untreated PRPF8 ChIP followed by Native chromatin proteomics (N-ChroP2) MS-based BRG1 and cell exposed to sub-lethal IR γH2AX heat-map and the corresponding line BRG1 HMGA2 identification of co-associated histone post-translational modifica- doses. plots show the time-course behaviour A2) DNA-bound interactors were quantified by means of SILAC protein of those interactors. Of interest, the tions. 5Gy, 1h recovery B) The relative abundance of histone PTMs were profiled over time in labeling and chromatin determinants were classified into recruited protein THRAP3 is enriched at γH2AX HR Hoechst PLA Colocalised Scale bar: 10um input as well as in γH2AX-enriched mono-nucleosomes. hPTMs (red), evicted (blue) or proteins stably associated (black) with RPA, PLA in particular after 1h upon IR. Two scenarios were identified: RAD50 or MDC1. Based on these results, it seems that the MRN B3) Immunofluorescence (IF, top) and proximity-ligation - modifications showing similar dynamics in both γH2AX-containing complex interacts more closely with HR rather than NHEJ, where a assay (PLA3, bottom) validation of the interaction be- mono-nucleosomes (local) and in input chromatin (global). higher number of chromatin regulators were identified. untreated tween γH2AX and endogenous THRAP3. As validation Hoechst PLA Colocalised - hPTMs specifically associated with γH2AX nucleosome (e.g. A3) Through integrated analysis, some interesting candidates specific of proteomic results, the number of interactions spots THRAP3 H4K20me1/me2, H3K56me1. H3K64Ac). for either HR or NHEJ were selected for further functional validation increases upon IR-mediated DNA damage. γH2AX experiments (see box.4). 5Gy, 1h recovery DNA-bound proteome 3. DNA-bound proteome changes 4. Functional validation with site-specific DSBs cellular systems

during DSB repair A) Auxin-Inducible Degron DSBs Induced Validation of NHEJ-, HR- or γH2AX-associated candidates via AsiSI (AID-DIvA6) cells A) isolation of Proteins On Chromatin (iPOC) during DNA repair in comparison with chromatin input 50 53BP1 HMGA2 RPA RAD51 BRG1 NuMA THRAP3 ATAD2 TPX2 PHF14

40 Biotin azide click Streptavidin pull-down Peptide clean up Number of identified proteins and chromatin shearing washed and tryptic digestion and MS analysis 30 (h) 0 1 4 8 Chromatin input B B

foci /cell 20 EdU iPOC

e 19

100 - B c B B n 3124 B v e 77 10 i Light a t d a l

1:1:1 n prS B 50 e u

B b + mix R

A 0 Medium B B B B 0 0 50 100 siCtr + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - Time (min) + siRNA - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + Heavy Heavy Heavy +4OHT - - + + + + - - + + + + - - + + + + - - + + + + - - + + + + - - + + + + - - + + + + - - + + + + - - + + + + - - + + + + B B Auxin - - - - + + - - - - + + - - - - + + - - - - + + - - - - + + - - - - + + - - - - + + - - - - + + - - - - + + - - - - + + A) The protocol for isolation of proteins on nascent DNA B) DNA-bound proteins identified with iPOC D) GO terms enriched in the iPOC during DDR (iPOND ) was optimized for the characterization of 7,8 TLR assay normalised on siRNA Ctr the chromatin composition in terms of proteins stably B) Traffic light reporter assay - log10_p-value iPOC Input iPOC bound to the DNA (iPOC). EdU treated (or not as NHEJ HR Pathway choice over UT over UT over Ctr 2 0 2 4 6 8 HR 1 4 8 0.5 1 4 8 1 4 8 negative control) cells were labeled in SILAC and the NHEJ

ZNF316 ( protein 316) 6 NUFIP2 (Nuclear fragile X mental retardation-interacting protein 2) Log2 SILAC ratio organization DDX54 (ATP-dependent RNA helicase DDX54) HMGB3 (High mobility group protein B3) heavy channel corresponds to similar time points hCG_31253 (Far upstream element-binding protein 3) HMCES (Embryonic stem cell-specific 5-hydroxymethylcytosine-binding protein) DBR1 (Lariat debranching enzyme) 3 histone modification TALDO1 (Transaldolase) HMGA1 ADNP2 (ADNP protein 2) XPC (DNA repair protein complementing XP-C cells) upon IR as in box 1 and 2B. Upon on-beads tryptic UNC80 (Protein unc-80 homolog) chromatin organization SRCAP (Helicase SRCAP) r REPIN1 (Replication initiator 1)

0.5 t PUS7 (Pseudouridylate synthase 7 homolog) HEL-S-77p (Casein kinase I isoform alpha) MECP2 NCAPH2 (Condensin-2 complex subunit H2) DNA metabolic process  C

digestion (on protease-resistant beads, or prS ) i PHF14 (PHD finger protein 14) -0.5 FBXO28 (F-box only protein 28) EHMT2 (Histone-lysine N-methyltransferase EHMT2) CDCA2 (Cell division cycle-associated protein 2) SUV39H1 (Histone-lysine N-methyltransferase SUV39H1) DNA repair AURKB (Aurora kinase B) SCML2 (Sex comb on midleg-like protein 2) -3 DNA-bound proteins were identified and compared NUBP2 (Cytosolic Fe-S cluster assembly factor NUBP2) KMT2B (Histone-lysine N-methyltransferase 2B) 1 ERCC5-201 (DNA repair protein complementing XP-G cells) organelle organization DIS3 (Exosome complex exonuclease RRP44) CDKN2AIP (CDKN2A-interacting protein) CEBPB (CCAAT/enhancer-binding protein beta) C1orf174 (UPF0688 protein C1orf174) -6 with determinants identified in in-depth chromatin BRD1 (Bromodomain-containing protein 1) nucleic acid metabolic process TUFM (Elongation factor Tu, mitochondrial) ORC3L (Origin recognition complex subunit 3) C over s MPG (DNA-3-methyladenine glycosylase) ZNF384 (Zinc finger protein 384) SP110 (Sp110 nuclear body protein) telomere maintenance in response to DNA damage F MORF4L1 (Mortality factor 4-like protein 1) input PUS1 (tRNA pseudouridine synthase) UBA2 (SUMO-activating enzyme subunit 2) LRRFIP1 (Leucine-rich repeat flightless-interacting protein 1) HMGN3 (High mobility group nucleosome-binding domain-containing protein 3) SF3A3 (Splicing factor 3A subunit 3) regulation of chromosome organization DNMT3A (DNA (cytosine-5)-methyltransferase 3A) ZNF512 (Zinc finger protein 512) RACGAP1 (Rac GTPase-activating protein 1) KDM2A (Lysine-specific demethylase 2A) B) Proteins identified in iPOC are reported in the C11orf31 (Selenoprotein H) peptidyl-lysine modification CDCA7L (Cell division cycle-associated 7-like protein) CUL4B (Cullin-4B) NUP153 (Nuclear pore complex protein Nup153) WAPAL (Wings apart-like protein homolog) SEC61B (Protein transport protein Sec61 subunit beta) nucleobase-containing compound metabolic process PDS5A (Sister chromatid cohesion protein PDS5 homolog A) heat-map. Strikingly, the majority of the determinants FLJ00169 (ADP-ribosylation factor-like protein 6-interacting protein 4) KAT7 (Histone acetyltransferase) TOP2B (DNA topoisomerase 2) TBPL1 (TATA box-binding protein-like protein 1) regulation of single strand break repair HIST1H2AC (Histone H2A) hCG_1994130 (40S ribosomal protein S15a) FYTTD1 (UAP56-interacting factor) show a modulation over the untreated sample only in PNO1 (RNA-binding protein PNO1) 0 TERF2 (Telomeric repeat-binding factor 2) macromolecule metabolic process KNOP1 (Lysine-rich nucleolar protein 1) WHSC1 (Histone-lysine N-methyltransferase NSD2) 1h/UT DKFZp666G145 (Protein Dr1) DHX8 (ATP-dependent RNA helicase DHX8) XRCC1 (DNA repair protein XRCC1) negative regulation of cellular biosynthetic process iPOC while in the input, if identified, don’t change. All UBQLN1 (Ubiquilin-1) A RAB35 (Ras-related protein Rab-35) MEN1 (Menin) RIF1 (Telomere-associated protein RIF1) RPL35 (60S ribosomal protein L35) 4h/UT FAU (40S ribosomal protein S30) heterocycle metabolic process siCtr G9Ai CBX1 (Chromobox protein homolog 1) the reported proteins are enriched over the techinical siRP ZSCAN21 (Putative zinc finger protein 702) EZH2i SPANXA2 (Sperm protein associated with the nucleus on the X chromosome C) siEZH2 ATAD2i siTPX2 CDCA8 (Borealin) siADNP siBRG1 siNuMA CBX3 (Chromobox protein homolog 3) regulation of telomere capping si53BP1 siRAD51 siATAD2 siTRA2b siPHF14 siPRPF8 INCENP (Inner centromere protein) siHMGA2 POGZ (Pogo transposable element with ZNF domain) 8h/UT ZMYM3 (Zinc finger MYM-type protein 3) control (- EdU). H2AFY2 (Core histone macro-H2A) siSMARCE H2AFY (Core histone macro-H2A.1) cellular aromatic compound metabolic process siSMARCA1 GTF2I (General II-I) SSBP1 (Single-stranded DNA-binding protein) RBM3 (RNA-binding protein 3) ZNF22 (Zinc finger protein 22) NUCKS1 (Nuclear ubiquitous casein and cyclin-dependent kinase substrate 1) SMTN (Smoothelin) C) Subset of determinants quantified only in the iPOC PSMD4 (26S proteasome non-ATPase regulatory subunit 4) RPL13A (60S ribosomal protein L13a) CFL2 (Cofilin-2) SUMO2 (Small -related modifier) CHAF1A (Chromatin assembly factor 1 subunit A) experiment. All these proteins are highly enriched C) Colony formation assay to screen for synthetic lethality D) Mode of action of the novel chromatin determinants nucleic acid metabolic process chromosome organization over the untreated sample at at least one time point, cellular macromolecule biosynthetic process thus representing proteins specifically recruited at C) iPOC-specific chromatin proteins 1 End 2 RNA metabolic process DNA upon DSBs elicited by sub-lethal doses of IR. Ctr RPA BRG1 TRA2b ATAD2 TPX2 G9Ai PHF14 NuMA SMARCE1 heterocycle metabolic process protection iPOC iPOC s NHEJ NHEJ chromatin organization ontology (GO) analysis of these deteriminats over UT over Ctr i e nucleic acid-templated transcription n 1 4 8 1 4 8 shows a over-representation for proteins involved in SPANXA2 (Sperm protein associated with the nucleus on the X chromosome C) l o SELH (Selenoprotein H)

CHAF1A (Chromatin assembly factor 1 subunit A) o UBQLN1 (Ubiquilin-1) 53BP1, ADNP, HMGA2 NuMA, PRPF8, SMARCE1 BRD1 (Bromodomain-containing protein 1) methyl-transferase activity to histone lysine residues. c FBXO28 (F-box only protein 28) protein-DNA complex subunit organization CDCA2 (Cell division cycle-associated protein 2) TBPL1 (TATA box-binding protein-like protein 1)

e 1 UNC80 (Protein unc-80 homolog) PHF14 (PHD finger protein 14) EHMT2 (Histone-lysine N-methyltransferase EHMT2) transcription from RNA polymerase II promoter Among those, of particular interest are methylations h ADNP2 (ADNP homeobox protein 2) SUV39H1 (Histone-lysine N-methyltransferase SUV39H1) t

SCML2 (Sex comb on midleg-like protein 2)

NUBP2 (Cytosolic Fe-S cluster assembly factor NUBP2) histone lysine methylation f DBR1 (Lariat debranching enzyme) on H3K4 and H3K9 enriched in mono-nucleosomes KMT2B (Histone-lysine N-methyltransferase 2B) End HR o HR ERCC5 (DNA repair protein complementing XP-G cells) DNMT3A (DNA (cytosine-5)-methyltransferase 3A) chromosome segregation

a resection negative regulation of gene expression, epigenetic marked by γH2AX (see box.1)

GO of iPOC-specific proteins D) GO classification of proteins enriched in iPOC at 1h (green), 4h A r e . 3 NHEJ - log10_p-value nucleic acid metabolic process U 0 1 2 3 chromatin organization (orange) or 8h (red) upon IR-induced DSB. From these results, . A nucleobase-containing compound metabolic process we can speculate that DNA repair proteins are recruited at DNA TRA2b, SMARCA1, PHF14 histone-lysine N-metT activity RNA metabolic process within 1h from DSB, followed by determinants involved in gene 0 SAM-dependent metT activity chromosome organization RPA, RAD51, BRG1, UT PARPi Etoposide Etoposide + PARPi EHMT2, TPX2, ATAD2 transferase activity cellular nitrogen compound metabolic process expression regulation (at 4h) and restoration of the chromatin protein metT activity heterocycle metabolic process HR N-metT activity structure upon damage repair (8h) cellular aromatic compound metabolic process

Conclusion and remarks At mono-nucleosome resolution the analysis of γH2AX-associated hPTMs suggests local (DDR mono-nuclesomes) cross-talks and dynamics compared to global chromatin (input). The integrated analysis between chromatin-associated proteins specifically characterize the functional interactors of either HR or NHEJ core components while on the DNA. The time course ChIP-SICAP experiment of γH2AX identifies the time-dependent dynamics of determinants upon IR and shed light on novel targets in DSB repair e.g. THRAP3 DNA-bound proteome (iPOC) outperforms deep chromatin proteome by profiling both determinants with functional role in DSB repair pathways and chromatin modifiers. Orthogonal functional validation determines the role in DDR for proteomic candidates and identifies novel players of HR repair, thus opening therapeutic opportunities with PARPi.

References: 1 Rafiee Mol Cell, 2016; 2 Soldi MCP 2013; 3 Soederberg Nat.Meth. 2006; 4 Sirbu Gen&Dev 2011; 5 Rafiee Sigismondo Mol Syst Biol 2020, 6 Aymard NSMB 2014; 7 Certo Nat.Meth. 2011; 8 Abu-Zhayia SciRep 2017