Polyubiquitinated PCNA recruits the ZRANB3 translocase to maintain genomic integrity after replication stress

Abstract

Completion of DNA replication after replication stress depends on PCNA, which undergoes monoubiquitination to stimulate direct bypass of DNA lesions by specialized DNA polymerases or is polyubiquitinated to promote recombination-dependent DNA synthesis across DNA lesions by template switching mechanisms. Here we report that the ZRANB3 translocase, a SNF2 family member related to the SIOD disorder SMARCAL1 protein, is recruited by polyubiquitinated PCNA to promote fork restart following replication arrest. ZRANB3 depletion in mammalian cells results in an increased frequency of sister chromatid exchange and DNA damage sensitivity after treatment with agents that cause replication stress. Using in vitro biochemical assays, we show that recombinant ZRANB3 remodels DNA structures mimicking stalled replication forks and disassembles recombination intermediates. We therefore propose that ZRANB3 maintains genomic stability at stalled or collapsed replication forks by facilitating fork restart and limiting inappropriate recombination that could occur during template switching events.

Figure 1. PCNA-dependent recruitment of ZRANB3 to DNA damage sites

(A) Schematic representation of the protein domains of SMARCAL1 and ZRANB3. The helicase domains are indicated in red. (B) Sequence alignments of the PIP box (left) and APIM (right) motifs of ZRANB3 with known PIP box and APIM motifs of other PCNA interacting proteins. The amino acids of the PIP box and APIM motifs that have been mutated are indicated by arrows. (C) Localization of WT or mutant GFP-ZRANB3 to DNA damage sites generated by UV-laser microirradiation. U2OS cells expressing WT GFP-ZRANB3 are shown with or without PCNA siRNA treatment prior to UV microirradiation. (D) Graphical representation of the percentage of U2OS cells that display co-localization of GFP-ZRANB3 with γH2AX at laser-generated stripes. The data represent the average and standard deviation of three independent experiments performed on cells expressing the GFP-constructs shown in (C). (E) Co-localization of ZRANB3 with PCNA. U2OS cells expressing HA-ZRANB3 were left untreated or subjected to hydroxyurea (2 mM) or UV radiation (25 J/m²) treatment. Cells subjected to UV radiation were also treated with caffeine (2 mM). Cells were stained with anti-HA (green) and anti-PCNA (red) antibodies. (F) Immunoprecipitation of FLAG-ZRANB3 with PCNA. FLAG-ZRANB3 either WT or mutated in the PIP and APIM motifs was purified from insect cells with anti-FLAG beads and then incubated with recombinant PCNA. Immunoprecipitated complexes were detected by western blotting.

Figure 2. Interaction of ZRANB3 with ubiquitinated PCNA in vitro

(A) Sequence alignment of the NZF motifs of ZRANB3, TAB2, TAB3, ZRANB1 and RANBP2. Residues involved in the binding of proximal and distal ubiquitin binding sites, as described in (Kulathu et al., 2009), are within green and blue boxes, respectively. NZF motif residues that have been mutated are indicated. (B) Association of K63-linked ubiquitin chains with the ZRANB3 NZF motif. GST-ZRANB3 containing a WT or mutant NZF motif were affinity purified from bacteria and incubated with mono-ubiquitin or ubiquitin chains linked on lysine 48 (K48) or lysine 63 (K63). Complexes were detected with anti-ubiquitin and anti-GST antibodies after western blotting. (C) Association of ZRANB3 with ubiquitinated PCNA. WT or mutant FLAG-ZRANB3 was immunoprecipitated with anti-FLAG beads from insect cells and incubated with a mixture of unmodified, mono- and poly-ubiquitinated PCNA. Immunoprecipitated PCNA and FLAG-ZRANB3 were detected by western blotting.

Figure 3. Association of ZRANB3 with ubiquitinated PCNA in mammalian cells

(A) Association of ZRANB3 with PCNA after DNA damage. Protein complexes from U2OS cells expressing WT or mutant HA-ZRANB3 were crosslinked and immunoprecipitated with an anti-PCNA antibody +/− UV radiation (30 J/m²) and caffeine (2 mM). ZRANB3 and PCNA were detected with anti-HA and anti-PCNA antibodies. The fold change of immunoprecipitated HA-ZRANB3 is indicated. (B) Depletion of USP1 increases the association of ZRANB3 with PCNA. 293T-Rex cells expressing FLAG-ZRANB3 were treated with control or USP1 siRNA prior to UV (30 J/m²) and caffeine (2 mM) treatment. Protein complexes were crosslinked, immunoprecipitated with an anti-PCNA antibody and immunoblotted with anti-FLAG and anti-PCNA antibodies. The fold change of immunoprecipitated FLAG-ZRANB3 is indicated. (C) Interaction of ZRANB3 with ubiquitinated PCNA in mammalian cells. U2OS cells expressing FLAG-ZRANB3 were treated with USP1 siRNA prior to UV (30 J/m²) and caffeine (2 mM) treatment. Protein complexes were subjected to anti-FLAG immunoprecipitation after crosslinking and then detected by western blotting. (D) Graphical representation of the percentage of cells that display co-localization of HA-ZRANB3, either WT or NZF mutant, with γH2AX at laser-generated stripes following treatment with siRNAs targeting RAD18, UBC13 or USP1. The time after microirradiation in which the samples were fixed is indicated. The data represent the average and standard deviation of three independent experiments. (E) Formation of GFP-ZRANB3 foci after depletion of USP1. U2OS cells were treated with control or USP1 siRNA and stained with antibodies recognizing GFP (green) and PCNA, ubiquitinated proteins (FK2), WRNIP1 or RAD18 (all in red), as indicated. (F) Foci formation after USP1 siRNA treatment of U2OS cells expressing either WT or NZF mutant HA-ZRANB3. Cells were stained with antibodies recognizing HA (green) and PCNA (red).

Figure 4. Effects of ZRANB3 depletion in mammalian cells

(A) Cell competition assay in U2OS cells treated with control siRNAs or three independent ZRANB3 siRNAs following treatment with camptothecin (CPT, 5 nM), hydroxyurea (HU, 2 mM) or cisplatin (CIS, 0.5 μM). The data represent the average and standard deviation of three independent experiments. (B) Cell competition assay in U2OS cells expressing either control or SMARCAL1 shRNAs treated with control or ZRANB3 siRNAs following treatment with camptothecin (CPT, 5 nM). Error bars have been calculated as in (A). (C) Graphical representation of the frequencies of sister chromatid exchanges (SCEs) of mitotic chromosomes isolated from U2OS cells transfected with control or ZRANB3 siRNA with or without mitomycin C (MMC, 20 nM) or camptothecin (CPT, 2.5 nM) treatment. The SCE frequencies of U2OS cells expressing a ZRANB3 cDNA clone resistant to ZRANB3 siRNA treatment are indicated. The average frequencies of SCEs and the standard deviation are indicated. Statistically significant p-values calculated using the Mann-Whitney test are indicated by asterisks (^*p<0.05, ^***p<0.001). n.s., not significant. (D) Chromosome spreads from U2OS cells transfected with control or ZRANB3 siRNA after camptothecin (CPT, 2.5 nM) treatment. SCEs are indicated. (E) Percentage of U2OS cells transfected with control or ZRANB3 siRNAs displaying more than 10 RAD51 foci. Cells were fixed 6 h or 12 h following 1 h camptothecin (CPT, 10 nM) treatment. The percentage of U2OS cells with more than 10 RAD51 foci that expressed siRNA resistant cDNA clones coding for either WT, PIP and APIM or NZF mutant ZRANB3 is also indicated. The data represent the average and standard deviation of three independent experiments in which 100 or more cells were counted.

Figure 5. DNA fiber analysis after ZRANB3 depletion in mammalian cells

(A) Schematics of the pulse-labeling experiment performed for DNA fiber analysis. (B) Images of DNA fibers isolated from cells treated with control or ZRANB3 siRNAs following the expression of siRNA resistant WT, PIP and APIM or NZF mutant ZRANB3. DNA fibers were stained with antibodies that recognize IdU (red) and CldU (green). (C) Detection by western blot of the ZRANB3 protein levels in the cells subjected to DNA fiber analysis shown in (B). (D) Graphical representation of the percentage of stalled forks (red only tracts) from DNA fiber analyses of the samples shown in (B). The data represent the average and standard deviation of three independent experiments.

Figure 6. Fork regression activities of ZRANB3 and SMARCAL1

(A) Purification of FLAG-ZRANB3 and FLAG-SMARCAL1 from insect cells. WT and helicase-dead (HD) mutant proteins were subjected to CM sepharose chromatography, affinity purified with anti-FLAG beads and eluted with FLAG peptide. The eluates were analyzed by SDS-PAGE and visualized by Coomassie. (B) Regression of synthetic fork substrates using ZRANB3 and SMARCAL1. ³²P-labeled fork substrates were incubated with WT and mutant proteins in a time course reaction. DNA products were then analyzed by gel electrophoresis and visualized by autoradiography. The position of the ³²P-labels on the DNA substrates is indicated by circles. (C) Schematic representation of the plasmid based assay employed in (D). ³²P-labels are indicated by circles. (D) Regression of plasmid based replication forks by ZRANB3 and SMARCAL1. Following incubation with ZRANB3 or SMARCAL1, fork structures were digested with restriction enzymes that cleave the dsDNA generated by annealing of the regressed DNA strands. DNA products of the restriction digests were analyzed by gel electrophoresis and visualized by autoradiography.

Figure 7. Disruption of D-loop structures by ZRANB3 and SMARCAL1

(A) Dissociation of D-loops (400 nM) by ZRANB3, either WT or helicase-dead (HD) (100 nM) as a function of time. Experiments were performed in buffer containing either magnesium or calcium acetate (5 mM). (B) D-loop structures generated as described in (A) were incubated in a time course reaction with either WT or helicase-dead (HD) SMARCAL1 (100 nM) in the presence of either magnesium or calcium acetate (5 mM). (C) Schematics of the dissociation of preformed RAD51-containing D-loop structures by ZRANB3 and SMARCAL1. (D) Increasing amounts of ZRANB3 and SMARCAL1 (25, 50 and 200 nM), either WT or mutant, were incubated with preformed RAD51-containing D-loops as depicted in (C). (E) Quantification of the disruption of preformed RAD51-containing D-loop structures following addition of increasing (upper panel) or fixed amounts of ZRANB3 and SMARCAL1 proteins (100 nM) in a time course reaction (lower panel). The activity of BLM (100 nM) in the time course reaction is indicated. Points with error bars represent the average and standard deviation of three or more independent experiments. (F) Schematics of the formation of RAD51-containing D-loop structures in the presence of ZRANB3 and SMARCAL1. (G) Formation of RAD51-containing D-loop structures following addition of increasing amounts of ZRANB3 or SMARCAL1 proteins (50, 100 and 200 nM) as represented in the schematics in (F). (H) Formation of RAD51-containing D-loop structures following addition of increasing amounts of either WT or mutant ZRANB3 (50, 100 and 200 nM). (I) Quantification of the formation of RAD51-containing D-loops in the presence of ZRANB3 and SMARCAL1 proteins. Points with error bars represent the average and standard deviation of three or more independent experiments.