CRL4(Wdr70) regulates H2B monoubiquitination and facilitates Exo1-dependent resection

Abstract

Double-strand breaks repaired by homologous recombination (HR) are first resected to form single-stranded DNA, which binds replication protein A (RPA). RPA attracts mediators that load the Rad51 filament to promote strand invasion, the defining feature of HR. How the resection machinery navigates nucleosome-packaged DNA is poorly understood. Here we report that in Schizosaccharomyces pombe a conserved DDB1-CUL4-associated factor (DCAF), Wdr70, is recruited to DSBs as part of the Cullin4-DDB1 ubiquitin ligase (CRL4(Wdr70)) and stimulates distal H2B lysine 119 mono-ubiquitination (uH2B). Wdr70 deletion, or uH2B loss, results in increased loading of the checkpoint adaptor and resection inhibitor Crb2(53BP1), decreased Exo1 association and delayed resection. Wdr70 is dispensable for resection upon Crb2(53BP1) loss, or when the Set9 methyltransferase that creates docking sites for Crb2 is deleted. Finally, we establish that this histone regulatory cascade similarly controls DSB resection in human cells.

Figure 1. CRL4^Wdr70 influences DSB repair pathway selection.

(a) The putative DWD motif of Wdr70. (b) Co-immunoprecipitation (co-IP) of tagged Ddb1 and Wdr70. MYC, TAP and HA-tagged proteins were expressed from their genomic loci. (c) Co-IP as in b following ultraviolet (UV; 500 J m⁻²) or ionising radiation (IR; 80 Gy) treatment. (d) Clonogenic survival analysis of wdr70⁺ and wdr70Δ cells following MMS or UV exposure. rad3Δ was used as a hypersensitive control. n=3 biological repeats. Error bars=s.d. (e) Pulse field electrophoresis of genomic DNA extracted from wdr70⁺ and wdr70Δ cells at the indicated time points after MMS or IR treatment. (f) Comparative analysis of HR efficiency between wdr70⁺ and wdr70Δ cells (spd1Δ background) using a heteroallelic recombination assay. Ratios represents the proportion of conversion (ade⁺ his⁺) to deletion (ade⁺ his⁻) type repair products. n=3 biological repeats. HR, homologous recombination. Error bars=s.d. (g) Comparative analysis of NHEJ efficiency using a plasmid re-joining assay (spd1Δ background). n=3 biological repeats. NHEJ, non homologous end joining. Error bars=s.d.

Figure 2. wdr70 mutants are defective in DSB resection.

All genetic backgrounds are spd1Δ. (a) Formation of Rad52 and Rpa1 foci after 80 Gy ionizing radiation in wdr70⁺ cells and wdr70Δ cells. n=3 biological repeats. Error bars=s.d. (b) Top: schematic representation of the HO-DSB locus: vertical lines: ApoI sites (35 bp and 3 kb from break site). HO cs: HO cut site; Kan^R; kanamycin resistance gene; black bars, PCR amplicons. Bottom: resection was assessed by PCR across the ApoI sites following digestion in the indicated strains. Samples were taken at time points shown following thiamine removal. HO induction from the nmt41 promoter requires ∼14 h to initiate. Loading was normalized by amplification of an uncut region (total). (c) Rpa1 recruitment was assayed by chromatin immunoprecipitation (ChIP) at 0.2, 3 and 9.4 kb from the DSB in wdr70⁺, wdr70Δ and mre11Δ backgrounds. Vertical axis: fold enrichment relative to T=0. n=3 biological repeats. Error bars=s.d. (d) Ddb1 recruitment measured by ChIP in wdr70Δ cells transformed with empty Rep41 vector (V), wdr70⁺ or wdr70-W325A/D326A (wdr70-wd) plasmids. (e) Recruitment of Wdr70 was measured by ChIP in the indicated strains 0.2 and 3 kb from the DSB site.

Figure 3. CRL4^Wdr70 stimulates spreading of uH2B at a DSB. All genetic backgrounds are spd1Δ.

(a) Western blot of uH2B levels in the indicated strains following DNA damage (NT; no treatment). (b) uH2B enrichment at the 0.2, 3 and 9.4 kb from a DSB assayed by chromatin immunoprecipitation (ChIP) in wdr70⁺ and wdr70Δ backgrounds. n=3 biological repeats. Error bars=s.d. (c) Resection assessed at the indicated times by PCR across the ApoI sites (see Fig. 2b) following ApoI digestion in wdr70⁺ (spd1Δ), wdr70Δ, ddb1Δ, rhp6Δ and H2B-K119R backgrounds. (d) Recruitment of Rhp6 in wdr70⁺ (WT) and wdr70Δ cells 0.2 and 3 kb from the break site assayed by ChIP. n=3 biological repeats. Error bars=s.d. (e) Recruitment Wdr70 assayed by ChIP at 0.2 and 3 kb from the break site for wild-type and H2B-K119R mutant cells. n=3 biological repeats. Error bars=s.d. (f) co-IP of EGFP-tagged Wdr70 and Flag-tagged H2B (K119 WT or K119R). Cell lysates were obtained from MMS-challenged cells, treated with ethidium bromide for 10 min and precipitated by anti-Flag antibody. * Indicates Ub-H2B. (g) Chromatin association of Wdr70, Rhp6 and uH2B at 0.2 and 3 kb from the break site assayed by ChIP for ctp1⁺ and ctp1Δ. n=2 biological repeats each with three technical repeats. *P<0.05; **P<0.01 (t-test). (h) co-IP of Rpa1-GFP with Rhp6-TAP. Method as described in f.

Figure 4. CRL4^Wdr70 promotes recruitment of Exo1 to DSB. All genetic backgrounds are spd1Δ.

(a) A two-step model for H2B mono-ubiquitination after DSB induction. CRL4^Wdr70 facilitates the loading of HULC^Rhp6 and the spreading of uH2B distal from the break site after formation of the HULC^Rhp6-catalysed uH2B core proximal to the DSB. (b) An experiment showing the recruitment of Exo1, Rqh1, Ctp1 and Mre11 when assayed by chromatin immunoprecipitation (ChIP) in wdr70⁺ and wdr70Δ backgrounds 0.2 and 3 kb from the DSB at the indicated times after thiamine removal. HO induction from the nmt41 promoter requires ∼14 h to initiate. WT, wild type; n=3. Error bars not shown. (c) Resection was subsequently assessed by quantitative PCR across the ApoI sites (see Fig. 2b) following ApoI digestion of DNA isolated at 0 and 18 h in the indicated strains. n=3 biological repeats. Error bars=s.d. *P<0.05; **P<0.01 (t-test). (d) Heteroallelic recombination assay for wdr70Δ, exo1Δ and the double wdr70Δ exo1Δ background. n=3 biological repeats. Error bars=s.d. **P<0.01 (t-test). NS, no statistical significance between strains. (e) Exo1 enrichment measured by ChIP 0.2 and 3 kb from an induced DSB in wdr70Δ, H2B-K119R and wdr70Δ H2B-K119R backgrounds to assess epistasis.

Figure 5. CRL4^Wdr70 suppresses Crb2^53BP1 loading at DSBs. All genetic backgrounds are spd1Δ.

(a) Recruitment of Crb2^53BP1 assayed by ChIP at 0.2 and 3 kb from the DSB following induction in wdr70⁺ (WT) and wdr70Δ backgrounds. (b) Recruitment of Crb2^53BP1 assayed by ChIP at 0.2 and 3 kb from the DSB in set9Δ, wdr70Δ and set9Δ wdr70Δ double mutant backgrounds. (c) ChIP assay of H4K20me2 enrichment 0.2 and 3 kb from the DSB in wdr70⁺, wdr70Δ and H2B-K119R backgrounds. (d) Recruitment of Exo1 assayed by ChIP (left) and resection analysed by quantitative PCR (right) 3 kb distal to an HO break in the crb2-F400A, wdr70Δ and crb2-F400A wdr70Δ double mutant backgrounds. (e) An equivalent experiment as d for set9Δ, wdr70Δ and set9Δ wdr70Δ double mutant backgrounds. All panels: n=3 biological repeats. Error bars=s.d. *P<0.05, **P<0.01 (t-test).

Figure 6. Prolonged occupancy of 53BP1 upon ablation of mammalian CRL4^WDR70.

(a) Quantification of Rad51 and phospho-Rpa32 (pS33) foci in 293T cells transfected with DDB1 or WDR70 siRNA. n=3 biological repeats. Error bars=s.d. *P<0.05, t-test. (b) CRISPR knockout (KO) WDR70 cells are defective for Rad51 IRIF and ionising radiation (IR)-induced RPA32-S33 phospho foci. n=3 biological repeats. Error bars=s.d. **P<0.01, t-test. (c) Immuno-detection of damage-induced H2B ubiquitination in 293T cells either pre-treated with siRNA specific for human CRL4^WDR70 genes or a scrambled control. (d) Western blot of two independent WDR70 KO cell lines generated by CRISPR. W-b was used for subsequent experiments. (e) Quantification of repair efficiency for I-SceI-induced DSB using quantitative PCR following transfection of linearized plasmids into parental and WDR70 KO 293T cells. n=3 biological repeats. Error bars=s.d. **P<0.01, t-test. (f) Quantification of 53BP1 foci with obvious cavities 6 h following IR treatment. n=3 biological repeats. Error bars=s.d. (g) Immunoblot for CPT-induced Rpa32 phosphorylation in 293T and WDR70 KO cells treated with either control or 53BP1 siRNA.