DCAF14 regulates CDT2 to promote SET8-dependent replication fork protection

Abstract

DDB1- and CUL4-associated factors (DCAFs) CDT2 and DCAF14 are substrate receptors for Cullin4-RING E3 ubiquitin ligase (CRL4) complexes. CDT2 is responsible for PCNA-coupled proteolysis of substrates CDT1, p21, and SET8 during S-phase of cell cycle. DCAF14 functions at stalled replication forks to promote genome stability, but the mechanism is unknown. We find that DCAF14 mediates replication fork protection by regulating CRL4^CDT2 activity. Absence of DCAF14 causes increased proteasomal degradation of CDT2 substrates. When forks are challenged with replication stress, increased CDT2 function causes stalled fork collapse and impairs fork recovery in DCAF14-deficient conditions. We further show that stalled fork protection is dependent on CDT2 substrate SET8 and does not involve p21 and CDT1. Like DCAF14, SET8 blocks nuclease-mediated digestion of nascent DNA at remodeled replication forks. Thus, unregulated CDT2-mediated turnover of SET8 triggers nascent strand degradation when DCAF14 is absent. We propose that DCAF14 controls CDT2 activity at stalled replication forks to facilitate SET8 function in safeguarding genomic integrity.

Figure 1.. DCAF14 regulates monomethylation of H4K20.

(A) Whole-cell lysates were extracted from U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. PCNA serves as a loading control. (B) Whole-cell lysates were extracted from siNT- and siDCAF14-transfected U2OS cells. Immunoblots were probed with the several histone methylation antibodies as shown. PCNA serves as a loading control. (C) U2OS cells were either transfected with the indicated siRNAs or treated with SET8-inhibitor UNC0379 for 4 h. Immunoblots were probed with the antibodies as shown. KU70 serves as a loading control. (D) siNT- and siDCAF14-transfected U2OS or HeLa cells were subjected to immunofluorescence analysis. Cells were immunostained for H4K20me1. Mean nuclei intensity was measured by quantitative imaging using DAPI-stained nuclei. Graphs represent mean ± SEM using at least 450 nuclei. (E) Whole-cell lysates were extracted from siNT- and siDCAF14 (5′UTR)-transfected U2OS cells that were either mock transfected or overexpressing DCAF14. Immunoblots were probed with the antibodies as shown. KU70 serves as a loading control. Graph represents normalized H4K20me1 intensities to histone H4 from three biological replicates. Source data are available for this figure.

Figure 2.. DCAF14 prevents increased turnover of SET8.

(A) Whole-cell lysates were extracted from U2OS, HeLa or hTERT-RPE1 cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. KU70 serves as a loading control. (B) Whole cell lysates were extracted from siNT- and siDCAF14-transfected U2OS cells to analyze changes in SET8. Graph represents normalized SET8 intensities from four biological replicates. (C) Parental U2OS, DCAF14 KO, and DCAF14 cDNA-transfected KO cells were immunostained for SET8. Mean nuclei intensity was measured by quantitative imaging using DAPI-stained nuclei. Graphs represent mean ± SEM using at least 3,500 nuclei. (D) Representative immunofluorescence images of siNT- and siDCAF14-transfected U2OS cells stained for DAPI, EdU, and SET8 are shown with overlay images. Scale bar = 10 μm. (E) siNT- and siDCAF14-transfected U2OS cells were pulsed with EdU for 30 min before immunofluorescence analyses. Mean nuclei intensity of SET8 was measured by quantitative imaging after preselecting EdU+ and EdU− nuclei. Graphs represent mean ± SEM using at least 250 nuclei. (F) siNT- and siDCAF14-transfected U2OS cells were pretreated with either DMSO or MG132 for 2 h and pulsed with EdU during the last 30 min of treatment. Cells were immunostained for SET8 and mean nuclei intensity was measured by quantitative imaging after preselecting EdU+ and EdU− nuclei. Graphs represent mean ± SEM using at least 250 nuclei. Source data are available for this figure.

Figure S1.. SET8 stability is regulated by DCAF14.

(A) Representative cell cycle distribution profiles for the indicated samples are shown. (B) siNT- and siDCAF14-transfected U2OS cells were either untreated or treated with proteasomal inhibitor MG132 for 30, 60, and 120 min. Whole-cell lysates were probed for SET8, DCAF14 and KU70 by immunoblotting. Representative immunoblot from three biological repeats is shown. (C) SET8 levels for each sample were normalized to loading control KU70. Mean values from three biological repeats are plotted. (D) siNT- and siDCAF14-transfected U2OS cells were pretreated with either DMSO or MLN4924 for 4 h and pulsed with EdU during the last 30 min of treatment. Cells were immunostained for SET8 and mean nuclei intensity was measured by quantitative imaging after preselecting EdU+ and EdU– nuclei. Graphs represent mean ± SEM using at least 250 nuclei. Source data are available for this figure.

Figure 3.. CDT2 function is aberrantly elevated when DCAF14 is absent.

(A) siNT- and siDCAF14-transfected U2OS cells were either left untreated or treated with HU for 4 h. Immunoblots were probed with the antibodies as shown. (B) Whole-cell lysates were extracted from siNT- and siDCAF14-transfected U2OS cells to analyze changes in p21 and CDT1. Graph represents normalized p21 and CDT1 intensities from four biological replicates. (C) siNT- and siDCAF14-transfected U2OS cells were pulsed with EdU for 30 min and immunostained for either CDT1 or p21 after preselecting EdU− nuclei. Graphs represent mean ± SEM using at least 200 nuclei. (D) siNT- and siDCAF14-transfected U2OS cells were pretreated with either DMSO or MG132 for 2 h and pulsed with EdU during the last 30 min of treatment. Cells were immunostained for either CDT1 or p21 and mean nuclei intensity was measured by quantitative imaging after preselecting EdU+ nuclei. Graphs represent mean ± SEM using at least 500 nuclei. (E) Parental and DCAF14 KO U2OS cells were pulsed with EdU and immunostained for SET8, CDT1 or p21. Mean nuclei intensity was measured by quantitative imaging after preselecting EdU+ nuclei. Graphs represent mean ± SEM using at least 250 nuclei. (F) Whole-cell lysates were extracted from U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. (G) U2OS cells transfected with the indicated siRNAs were pulsed with EdU and immunostained for SET8, CDT1 or p21. Mean nuclei intensity was measured by quantitative imaging after preselecting EdU+ nuclei. Graphs represent mean ± SEM using at least 125 nuclei. (H) Mock or DCAF14 cDNA-transfected U2OS cells were left either untreated or treated with 4 mM HU or 1 μM CPT for 4 h before immunostaining for SET8. Mean nuclei intensity was measured by quantitative imaging DAPI-stained nuclei. Graphs represent mean ± SEM using at least 200 nuclei. (I) Whole-cell lysates were extracted from siNT- and siDCAF14-transfected U2OS cells to analyze changes in CDT2. Graph represents normalized CDT2 intensities from five biological replicates. (J) U2OS cells transfected with the indicated siRNAs were pulsed with EdU for 30 min and subjected to PLA analyses using antibodies targeting CDT2 and PCNA. Number of PLA foci was measured after preselecting EdU+ and EdU− nuclei. Box and whisker plots represent 10–90 percentile using at least 100 nuclei. Source data are available for this figure.

Figure S2.. Loss of DCAF14 increases turnover of CDT2 substrates.

(A) siNT- and siDCAF14-transfected U2OS cells were pulsed with EdU followed by treatment with HU for 4 h, concomitant with DMSO or MG132 as indicated, and subjected to immunofluorescence analyses. Mean nuclei intensities of SET8, CDT1 or p21 were measured by quantitative imaging after preselecting EdU+ nuclei. Graphs represent mean ± SEM using at least 250 nuclei. (B) Parental and DCAF14 KO U2OS cells were pulsed with EdU followed by treatment with 4mM HU for 4 h. Cells were immunostained for SET8, CDT1 or p21. Mean nuclei intensities were measured by quantitative imaging after preselecting EdU+ nuclei. Graphs represent mean ± SEM using at least 200 nuclei. (C) U2OS cells transfected with the indicated siRNAs were pulsed with EdU followed by treatment with 4 mM HU for 4 h. Cells were immunostained for SET8, CDT1 or p21. Mean nuclei intensities were measured by quantitative imaging after preselecting EdU+ nuclei. Graphs represent mean ± SEM using at least 100 nuclei. (D) Mock or DCAF14 cDNA-transfected U2OS cells were left either untreated or treated with 4 mM HU or 1 μM CPT for 4 h. Whole-cell lysates were probed with the indicated antibodies by immunoblotting. (E) Whole-cell lysates were extracted from siNT- and siDCAF14-transfected U2OS cells treated with either DMSO or MLN4924 for 4 h. Whole-cell lysates were probed with the indicated antibodies by immunoblotting. Immunoblots were probed with the antibodies as shown. (F) siNT- and siDCAF14-transfected U2OS cells were pulsed with EdU and either treated with DMSO or MLN4924 for 4 h before immunostaining for CDT1 and p21. Mean nuclei intensities were measured by quantitative imaging after preselecting EdU+ nuclei. Graphs represent mean ± SEM using at least 350 nuclei. (G) siNT- and siDCAF14-transfected U2OS cells were pulsed with EdU followed by treatment with HU for 4 h, concomitant with DMSO or MLN4924 as indicated, and subjected to immunofluorescence analyses. Mean nuclei intensities of SET8, CDT1, and p21 were measured by quantitative imaging after preselecting EdU+ nuclei. Graphs represent mean ± SEM using at least 300 nuclei. Source data are available for this figure.

Figure S3.. Increased CDT2 activity collapses stalled forks in DCAF14-deficient cells.

(A) Whole-cell lysates were extracted from U2OS, HeLa or hTERT-RPE1 cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. * denotes nonspecific band. (B) siNT- and siCDT2-transfected U2OS, HeLa or hTERT-RPE1 cells were subjected to fork elongation assays in the presence of 100 nM CPT as shown. IdU lengths are plotted. (C) Whole-cell lysates were extracted from parental and DCAF14 KO U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. (D) siNT- and siCDT2-transfected parental or DCAF14 KO U2OS cells were subjected to fork protection analysis. (E) Whole-cell lysates were extracted from U2OS cells transfected with the indicated siRNAs. CDT2#2 represents alternate siRNA to CDT2#1 used in Figs 4 and 6. Immunoblots were probed with the antibodies as shown. (F) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. Fiber labeling experiments depict mean ± SEM and at least 100 fibers were quantified. Source data are available for this figure.

Figure 4.. CDT2 impairs stalled fork stability in DCAF14-deficient cells.

(A) U2OS or DCAF14 KO cells were pulse-labeled with CldU for 30 min followed by IdU for 30 min before DNA combing analysis to measure replication fork speed and replication fork asymmetry. (B) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. (C) siNT- and siDCAF14-transfected U2OS cells were subjected to fork protection analysis. DMSO or MLN4924 were added concomitantly with HU for 4 h. (D) Tail moments were measured for parental and DCAF14 KO U2OS cells transfected with the indicated siRNAs using neutral comet assay. Cells were either untreated or treated with CPT for 1 h. Box and whisker plots represent 10–90 percentile using at least 100 nuclei. (E) Stalled fork recovery post 4 h of HU treatment was measured for the U2OS cells transfected with the indicated siRNAs using DNA fiber analysis. Scale bar = 10 μm. Graphs represent mean ± SD from three biological replicates. Fiber labeling experiments depict mean ± SEM and at least 100 fibers were quantified. Source data are available for this figure.

Figure 5.. SET8 mediates replication fork protection.

(A) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. (B) U2OS cells either transfected with the indicated siRNAs or concomitantly treated with HU and UNC0379 for 4 h were subjected to fork protection analysis. (C) Untreated or HU-treated siNT- and siSET8-transfected U2OS cells were subjected to SIRF analyses using antibodies targeting SET8 and biotin. Box and whisker plots represent 10–90 percentile using at least 450 nuclei. (D) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. (E) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. MRE11 inhibitor mirin or DNA2 inhibitor C5 were added concomitantly with HU for 4 h where indicated. Fiber-labeling experiments depict mean ± SEM and at least 100 fibers were quantified. Source data are available for this figure.

Figure S4.. SET8 deficiency causes nascent strand degradation.

(A) Whole cell lysates were extracted from U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. (B) U2OS cells transfected with two independent siRNAs (#1 and #2) targeting CDT1 or p21 were subjected to fork protection analysis. (C, D, E) Whole-cell lysates were extracted from U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. (F) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. Graph depicts mean ± SEM and at least 100 fibers were quantified. (G) Whole-cell lysates were extracted from U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. KU70 and PCNA serve as loading controls in the immunoblots. Source data are available for this figure.

Figure 6.. CDT2-dependent down-regulation of SET8 causes nascent strand degradation in DCAF14-deficient cells.

(A) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. MG132 was added concomitantly with HU for 4 h where indicated. (B) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. (C) U2OS cells transfected with the indicated siRNAs were subjected to fork elongation assays in the presence of 100 nM CPT as shown. MG132 was added concomitantly with CPT for 30 min where indicated. (D) siNT- and siDCAF14-transfected U2OS cells were pulsed with EdU followed by treatment with HU for 4 h, concomitant with DMSO or p97i as indicated, and subjected to immunofluorescence analyses. Mean nuclei intensity of SET8, CDT1 or p21 was measured by quantitative imaging after preselecting EdU+ nuclei. Graph depicts mean ± SEM using at least 300 nuclei. (E) U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. p97i was added concomitantly with HU for 4 h where indicated. (F) Parental and DCAF14 KO U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. UNC0379 was added concomitantly with HU for 4 h where indicated. (G) Parental and DCAF14 KO U2OS cells transfected with the indicated siRNAs were subjected to fork protection analysis. MG132 was added concomitantly with HU for 4 h where indicated. Fiber-labeling experiments depict mean ± SEM and at least 100 fibers were quantified. Source data are available for this figure.

Figure S5.. Nascent strand degradation in the absence of DCAF14 occurs due to excessive SET8 turnover by CDT2.

(A, B) Whole-cell lysates were extracted from U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. (C) U2OS cells transfected with the indicated siRNAs were subjected to fork elongation assays in the presence of 100 nM CPT as shown. MG132 was added concomitantly with CPT for 30 min where indicated. CldU lengths are plotted. Panel depicts mean ± SEM and at least 100 fibers were quantified. (D) siNT- and siDCAF14-transfected U2OS cells were co-transfected with plasmid expressing HA-tagged ubiquitin. Cells were either untreated or treated with p97i for 4 h. Chromatin fractions were isolated and immunoblotting analysis was performed with the indicated antibodies. (E) U2OS cells transfected with the indicated siRNAs were subjected to SIRF analyses using antibodies targeting SET8 and biotin. Box and whisker plots represent 10–90 percentile using at least 200 nuclei. (F) Whole-cell lysates were extracted from parental and DCAF14 KO U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. (G) Whole-cell lysates were extracted from parental and DCAF14 KO U2OS cells transfected with the indicated siRNAs. Immunoblots were probed with the antibodies as shown. (H) At unperturbed replication forks, SET8 is recruited to PCNA and subsequently proteolyzed by CRL4^CDT2 (CDT2 substrates CDT1 and p21 are not depicted). Challenges to DNA replication triggers enrichment of DCAF14 at stalled replication forks to counteract CDT2 activity. Restrained CDT2 function enables SET8 to mediate replication fork protection before proteolytic degradation by CRL4^CDT2. In the absence of DCAF14, accelerated CDT2 activity disfavors SET8 engagement at stalled forks resulting in nascent strand degradation. Source data are available for this figure.