Rad51-mediated replication of damaged templates relies on monoSUMOylated DDK kinase

Abstract

DNA damage tolerance (DDT), activated by replication stress during genome replication, is mediated by translesion synthesis and homologous recombination (HR). Here we uncover that DDK kinase, essential for replication initiation, is critical for replication-associated recombination-mediated DDT. DDK relies on its multi-monoSUMOylation to facilitate HR-mediated DDT and optimal retention of Rad51 recombinase at replication damage sites. Impairment of DDK kinase activity, reduced monoSUMOylation and mutations in the putative SUMO Interacting Motifs (SIMs) of Rad51 impair replication-associated recombination and cause fork uncoupling with accumulation of large single-stranded DNA regions at fork branching points. Notably, genetic activation of salvage recombination rescues the uncoupled fork phenotype but not the recombination-dependent gap-filling defect of DDK mutants, revealing that the salvage recombination pathway operates preferentially proximal to fork junctions at stalled replication forks. Overall, we uncover that monoSUMOylated DDK acts with Rad51 in an axis that prevents replication fork uncoupling and mediates recombination-dependent gap-filling.

Fig. 1. DDK kinase facilitates template switch replication of damaged templates.

a Schematic representation of replication intermediates as visualized by 2D gel electrophoresis and of the ARS305 fragment spanned by NcoI restriction sites. b Exponentially grown Tc-sgs1 and cdc7-4 Tc-sgs1 cells were arrested in G1 phase at 25 °C using α-factor and released into S phase at 28 °C in the presence of 0.033% MMS and tetracycline (1 mM) to induce DNA damage and Sgs1 depletion respectively. Samples were collected at the indicated timepoints for 2D gel electrophoresis, FACS and western blotting analysis in two independent experiments. Collected cells were subjected to in vivo psoralen-mediated DNA inter-strand cross-linking followed by genomic DNA extraction and genome digestion with NcoI. 2D gel filters were hybridized with a probe recognizing the early replication origin, ARS305. HA-Sgs1 levels were detected by western blotting with Pgk1 as loading control. FACS profiles showing the cellular DNA content to monitor cell cycle progression are reported for each strain at the indicated time point. 1 N and 2 N cellular DNA contents below the FACS indicate G1 and G2/M phases of the cell cycle, respectively. The relative enrichment of X-shaped replication intermediates is represented in the quantified plots where signal intensities are normalized to the monomer spot and the highest value for the X molecule obtained during quantification was assigned as 100%. c Schematic representation of the cdc7-as3 allele with respective mutations. sgs1Δ and cdc7-as3 sgs1Δ cells were synchronized in G1 phase with α-factor and released into YPD media containing 0.033% MMS. PP1 drug (20 mM) was added after 20 min from the release of cells from G1 into S phase. Samples for 2D gel analysis of replication-recombination intermediates and FACS analysis to detect the cellular DNA content and cell cycle phase were collected at the indicated time points and processed as described in panel b.

Fig. 2. MonoSUMOylated DDK promotes replication-associated recombination.

a Tc-sgs1 and ddk-KR Tc-sgs1 cells were synchronized in G1 phase at 25 °C and released in YPD medium containing 0.033% MMS and tetracycline (1 mM) at 28 °C. Samples were then analyzed by 2D gel electrophoresis in two independent experiments. Experimental setup is the same as described in Fig. 1a. b (Top) Schematic representation of the 2D gel patterns arising from the modified diploid strain with the expected migration of inter-homolog Junctions (IHJs) in red versus sister chromatid junctions in black. IHJs are indicated by red arrows. (Bottom) Tc-sgs1/Tc-sgs1*, cdc7-4 Tc-sgs1/cdc7-4 Tc-sgs1*, ddk-KR Tc-sgs1/ddk-KR Tc-sgs1* (* indicates mutation at the EcoRV restriction site adjacent to the ARS305 region) were arrested in G2/M phase with nocodazole at 25 °C and released at 28 °C in YPD medium containing 0.033% MMS and tetracycline (1 mM) to induce Sgs1 depletion. Samples were collected at the indicated timepoints which were psoralen crosslinked and genomic DNA was extracted for 2D gel analysis in two independent experiments. Restriction digestion was performed with EcoRV and NcoI and the resulting fragments separated by 2D gel electrophoresis and later probed for ARS305. HA-Sgs1 depletion was visualized using an HA antibody with Pgk1 as loading control. Cell cycle progression of each strain was monitored by FACS analysis. The relative enrichment of X-shaped replication intermediates is represented in the quantified plots where signal intensities from IHJs were normalized with respect to the monomer spot and the highest value for the X molecule obtained during quantification was assigned as 100%.

Fig. 3. SUMOylated DDK kinase prevents replication fork uncoupling.

a Examples of normal and gapped forks visualized by Transmission electron microscopy (TEM) analysis along with schematic representations of DNA molecules with dsDNA (black) and ssDNA (red). Scale bars (black) of 360 nm calculated for dsDNA are reported. b TEM analysis of replication intermediates. Exponential cultures were arrested in G1 phase and released into S-phase in medium containing 0.033% MMS at 28 °C for 60 min. Histogram reports the percentages of gapped replication forks, from three independent experiments. The gapped forks category contains a small fraction of broken forks, derived from the mechanical breakage of the long ssDNA stretches present at the branching point of the gapped forks. The total number of replication intermediates scored in the three experiments is indicated as ‘n’. Error bars represent standard deviation of three independent experiments, the center of the bar represents the average value. P-values were calculated by ordinary one-way ANOVA test. c Scatter dot plot representing the distribution of lengths of ssDNA stretches at the branching points of forks of indicated cells. Values are derived from the analysis of the samples presented in b. Central line in the plot indicates the average length of ssDNA discontinuity (in nucleotides), the bars represent distribution range. P values were calculated by unpaired two-tailed Student’s t test. d TEM analysis of the DNA replication intermediates formed in the same experimental conditions described in b in two independent experiments. ‘n’ represents the total number of DNA replication intermediates analyzed in the indicated sample. Histogram represents the percentage of gapped forks in the indicated cells as described in b. P values were calculated by Fischer’s exact two-sided test. e Scatter dot plot representing the distribution of the lengths of the ssDNA stretches (in nucleotides) at the fork branching points of the indicated yeast strains. ‘n’ is the total number of DNA replication intermediates analyzed. The middle line in the scatter dot plot indicates the average length of the ssDNA discontinuity (in nucleotides), bars show the distribution range. P-values were calculated by unpaired two-tailed Student’s t test.

Fig. 4. SUMOylated DDK promotes Rad18- and Mms2-dependent gap-filling.

a Tc-sgs1, mms2Δ Tc-sgs1, cdc7-4 Tc-sgs1, cdc7-4 mms2Δ Tc-sgs1 cells were synchronously released from G1 phase into S-phase in medium containing 0.033% MMS and 1 mM tetracycline at 28 °C in two independent experiments. Samples were collected at the indicated timepoints and subjected to in vivo psoralen-mediated DNA inter-strand crosslinking followed by genomic DNA extraction, digestion with NcoI, 2D gel electrophoresis and probing of the 2D gels filters with the ARS305 probe. HA-Sgs1 depletion was monitored by western botting using Pgk1 as control. FACS plots show cell cycle progression. The signal of X molecules was quantified with respect to the monomer spot. The highest valued obtained for the signal of X molecules was considered as 100%. b, c Exponentially growing cultures of the indicated strains were adjusted to the same concentration, serially diluted 10-fold and spotted on YPD plates containing the indicated MMS concentrations. Plates were incubated at 28 °C and photographed after 2 days from the spotting in two independent experiments with similar results.

Fig. 5. SUMOylated DDK facilitates salvage recombination that prevents fork uncoupling.

a 2D gel analysis of MMS-induced DNA replication-recombination intermediates accumulated in the indicated strains in two independent experiments. Exponentially growing cells were synchronized in G1 and released in YPD medium containing 0.033% MMS and tetracycline (1 mM) at 28 °C. Samples were collected at the indicated timepoints for FACS, western blotting, and 2D gel analysis. Experimental setup is the same as described in Fig. 1a. b Rad51 recruitment at the ARS305 replication origin was analyzed using ChIP-qPCR. Exponentially growing cells were arrested in G1 phase at 25 °C and released into 0.033% MMS-containing YPD medium at 28 °C with samples for ChIP-qPCR collected at the indicated time points. Each ChIP was repeated three times and each real time PCR was performed in triplicates. Histogram reports the enrichment signals expressed as percentage of the signal in the corresponding input. Enrichment values obtained were then normalized and expressed as fractions of the enrichment obtained in the wild type cells, which was considered as 1. Error bars are represented as SEM (mean value ± standard error of mean) of three independent experiments. The indicated p-values were calculated by two-way ANOVA test. c TEM analysis of the DNA replication intermediates in the indicated strains exposed to MMS. Experimental setup is the same as described in Fig. 3b. Histogram represents the percentage of gapped replication forks generated in cdc7-4 siz1Δ and ddk-KR siz1Δ mutants in comparison with siz1Δ. WT, cdc7-4 and ddk-KR values are indicated for reference (see Fig. 3). Total number (n) of DNA replication-recombination intermediates analyzed for each genotype is indicated. The indicated p-values were calculated by Fischer exact test two-sided. d Scatter dot plot represents the distribution of length of the ssDNA gaps (in nucleotides) at the fork branching points of DDK mutants in comparison with WT. Analysis was conducted on the same samples of the experiment presented in c. Middle line in the scatter dot plot indicates the average value of the length of the ssDNA discontinuity, bars represent distribution range. P-values were calculated by unpaired two-tailed Student’s t test are indicated.

Fig. 6. Rad51 and putative SIMs mediate recombination-mediated damage bypass and prevent fork uncoupling.

a Exponentially growing cultures of the indicated strains were adjusted to the same concentration, serially diluted (1:10), and then spotted on YPD plates with the indicated drug concentrations in two independent experiments. Plates were photographed after 3 days of spotting. b sgs1Δ, rad51-SIM1 sgs1Δ, rad51-SIM2 sgs1Δ, and rad51Δ sgs1Δ cells were synchronized in G1 phase and then released into 0.033% MMS containing YPD medium. Samples were collected for the indicated timepoints for FACS, western blot, and 2D gel analysis in two independent experiments. Collected samples for 2D gel analysis were subjected to in vivo psoralen crosslinking, after which genomic DNA was extracted followed by NcoI digestion and analysis of 2D gel electrophoresis-resolved intermediates for the ARS305 region. The relative enrichment of X-shaped replication intermediates was quantified as described in Fig. 1c. c Electron microscopic analysis of wildtype, rad51-SIM1, rad51-SIM2, and rad51Δ. Cells were arrested in G1 phase using α-factor and then released into medium containing 0.033% MMS. Samples were collected at 60 min and subjected to in vivo psoralen crosslinking. Genomic DNA was extracted and subjected for electron microscopy analysis. The histograms indicate the percentage of gapped replication forks observed for the above strains. The total number of forks analyzed for each strain is reported as ‘n’. The P-values were calculated by Fischer exact test two-sided. d Scatter dot plot representing the distribution of the length of ssDNA at fork branching points of the indicated strains. ‘n’ is the total number of DNA replication intermediates analyzed. Values derive from the analysis of the samples of c. The middle line in the scatter dot plot indicates the average length of the ssDNA gaps at the fork branching points; the bar shows the distribution range. P-values were calculated by unpaired two-tailed Student’s t test.

Fig. 7. Model illustrating SUMOylated DDK kinase as regulator of gap-filling mediated DNA damage tolerance and fork restart by recombination.

The model illustrates monoSUMOylated DDK kinase traveling with replication forks exposed to genotoxic stress. SUMOylated DDK kinase prevents fork uncoupling and stabilizes Rad51 filaments on exposed ssDNA gaps. These filaments may facilitate repriming creating additional substrates for gap-filling in addition to the ones generated on the lagging strand by physiological repriming events. Alternatively, Rad51 assembly on ssDNA stretches at stalled replication fork can be engaged by salvage recombination pathways with potential to generate genome rearrangements. MonoSUMOylated DDK kinase facilitates both recombination-mediated gap-filling and DNA damage tolerance/fork protection in the context of the stalled replication fork.