CDC-48/p97 coordinates CDT-1 degradation with GINS chromatin dissociation to ensure faithful DNA replication

Abstract

Faithful transmission of genomic information requires tight spatiotemporal regulation of DNA replication factors. In the licensing step of DNA replication, CDT-1 is loaded onto chromatin to subsequently promote the recruitment of additional replication factors, including CDC-45 and GINS. During the elongation step, the CDC-45/GINS complex moves with the replication fork; however, it is largely unknown how its chromatin association is regulated. Here, we show that the chaperone-like ATPase CDC-48/p97 coordinates degradation of CDT-1 with release of the CDC-45/GINS complex. C. elegans embryos lacking CDC-48 or its cofactors UFD-1/NPL-4 accumulate CDT-1 on mitotic chromatin, indicating a critical role of CDC-48 in CDT-1 turnover. Strikingly, CDC-48(UFD-1/NPL-4)-deficient embryos show persistent chromatin association of CDC-45/GINS, which is a consequence of CDT-1 stabilization. Moreover, our data confirmed a similar regulation in Xenopus egg extracts, emphasizing a conserved coordination of licensing and elongation events during eukaryotic DNA replication by CDC-48/p97.

Figure 1. CDC-45 and subunits of the GINS complex persist on mitotic chromatin in embryos depleted for the CDC-48^UFD-1/NPL-4 complex

(A, B) Selected pictures of time-lapse recordings of embryos expressing GFP::CDC-45 or GFP::SLD-5 (green) and mCherry::H2B (red) that are depleted for empty control, cdc-48 and/or ufd-1, npl-4, div-1, or codepleted for ufd-1/atl-1 by RNAi. Each image series shows representative cell-cycle phases (Mitosis or S phase) at distinct times of embryonic development (2 to 4 cell stage) of one single C. elegans embryo. Empty arrows indicate wild-type like mitotic localization, filled arrows indicate persistent association of the indicated proteins with mitotic chromatin. Anterior is to the left. Scale bars represent 5 μm.

Figure 2. Suppression of the cell-cycle progression delay of embryos lacking UFD-1 or NPL-4 by cdc-45, sld-5 or psf-3 (RNAi) depletion

(A) Schematic illustration of the RNAi feeding procedure to achieve a sequential depletion of empty (light grey), ufd-1 or npl-4 (dark grey) for the first 48 h and cdc-45, sld-5, psf-3 or cdc-45+sld-5 (2^nd (RNAi)) for the last 24 h. Subsequent time-lapse analysis was performed to visualize the previously described delay of ufd-1 and npl-4 RNAi embryos in cell-cycle progression of the P1 cell (Mouysset et al., 2008). (B, C) Quantification of the time between division of AB and P1 cell (P1 division delay) of embryos depleted first for empty (light grey) and ufd-1 or npl-4 (dark grey) and sequentially for empty, cdc-45, sld-5, psf-3 or codepleted for cdc-45+sld-5 (2^nd (RNAi)). (D) Two-hybrid assay for the interaction of C. elegans UFD-1 with CDC-45. Yeast cells expressing the indicated proteins were streaked out on medium plates lacking histidine to test for interaction dependent activation of the HIS3 gene. (E) Western-blot analysis of GFP fusions of CDC-45 and SLD-5 in C. elegans embryonic extracts depleted for empty, ufd-1 and cdc-45, or sld-5. Time is shown in hours (A) or minutes (B and C). Data are mean values. Error bars show standard error of the mean (s.e.m.). Statistical significance between cell-division timings are indicated by asterisks in B and C. The single asterisk indicates P ≤ 0.05 and the double asterisk indicates P ≤ 0.001.

Figure 3. CDC-48^UFD-1/NPL-4 depleted embryos show elevated levels of CDT-1 protein

(A) Quantification of the cell division delay between AB and P1 cell (P1 division delay) of embryos depleted first for empty (light grey) or ufd-1(RNAi) (dark grey) and sequentially for empty or cdt-1(RNAi). (B, C, D) Western-blot analysis of CDT-1 protein levels in embryonic extracts that are depleted for the indicated gene products by RNAi. In (C) embryos were depleted for empty or cdc-48(RNAi) in wild-type or cdc-48.1(tm544) mutant background. In (D) ufd-1 and rbx-1(RNAi) bacteria were equally mixed either with empty control bacteria or together. Quantification of the signal intensity was calculated relative to the Tubulin level and normalized to the protein levels of the empty(RNAi) control.

Figure 4. CDT-1 accumulates on mitotic chromatin in cdc-48, ufd-1 and npl-4 RNAi embryos

(A, B) Immunostainings of early C. elegans embryos treated with empty, cdc-48, ufd-1, npl-4, rbx-1 or pcn-1 (RNAi). CDT-1 (green), Tubulin (red) and DAPI (blue) staining is shown as merge images and in separate channels. Distinct cell cycle phases are indicated as Mitosis or S phase. Empty arrowheads indicate wild-type CDT-1 levels, whereas filled arrowheads indicate enhanced signal intensity on mitotic chromatin. In B cdc-48 (RNAi) was performed on cdc-48.1(tm544) mutant background. (C) Selected pictures of time-lapse recordings of C. elegans embryos expressing GFP::SLD-5 (green) and mCherry::H2B (red) that are depleted for empty control, rbx-1, and pcn-1. Each image series shows representative cell-cycle phases of the first mitotic division of one single embryo. Empty arrowheads point to wild-type like SLD-5 localization. Scale bar represents 5 μm.

Figure 5. Depletion of CDT-1 suppresses persistent SLD-5 chromatin association in ufd-1(RNAi) embryos

(A) Selected pictures of time-lapse recordings of embryos expressing GFP::SLD-5 (green) and mCherry::H2B (red) that are depleted first for empty or ufd-1 followed by empty or cdt-1 (seq(RNAi)). Representative pictures of indicated cell-cycle phases (Mitosis or S phase) at distinct time points of embryonic development (1 to 4 cell stage) of one single C. elegans embryo are shown. Empty arrows indicate wild-type like mitotic localization, filled arrows indicate persistent association with mitotic chromatin, shaded arrowheads indicate partial mislocalization. Percentage values represent the number of mitotic divisions where SLD-5 chromatin association was monitored under indicated experimental conditions. (B) Quantification of the GFP signal intensity on mitotic chromatin in embryos treated with empty, cdt-1, ufd-1 or ufd-1/cdt-1 seq(RNAi) shown in (A). GFP::SLD-5 signal intensity is shown relative to the intensity for mCherry::H2B in the same area. Anterior is to the left. Data are mean values. Error bars show standard error of the mean (s.e.m.). Statistical significance relative signal intensities are indicated by asterisks. The single asterisk indicates P ≤ 0.05. Scale bar represents 5 μm.

Figure 6. Regulation of chromatin association of GINS and CDT-1 is conserved in Xenopus laevis egg extracts

(A) Schematic illustration of the experimental procedure to re-isolate sperm chromatin from S phase/interphase or mitotic Xenopus egg extracts. (B) In vivo interaction between Ufd1 and Cdt1. Xenopus egg extracts were incubated with anti-Ufd1 antibodies to coupled to Protein A Dynabeads for immunoprecipitation experiments. (C) Two sequential pulldowns against unspecific IgG/Ufd1 and His-tagged Ubiquitin were performed from Xenopus egg extracts arrested in Metaphase (M) or Interphase (I). Eluates were analyzed for the presence of indicated proteins by western blot. Cdt1 eluted from Ufd1 pulldowns can be precipitated by Ni-NTA (highlighted by <) indicating interaction of Ufd1 with ubiquitylated Cdt1. Asterisk indicates unspecific signal. (D) Western blot analysis showing efficient immunodepletion of Ufd1, Npl4, and p47 after the first (1^st IPΔ) and second (2^nd IPΔ) round of incubation of egg extracts with the respective antibodies. Tubulin was used as loading control. (E) Tubulin and DAPI staining of sperm chromatin incubated for 85 min or 175 min in Mock control, Ufd1, Npl4, and p47 immunodepleted extracts show successive cycling through S phase/Interphase and Mitosis. (F) Immunostaining of mitotic sperm chromatin that was incubated for 175 min in Mock, Ufd1, Npl4, and p47 depleted egg extracts. SLD-5 (green) and DAPI (blue) staining is shown as merge images and in separated channels. (G) Western blot analysis of reisolated S phase/interphasic or mitotic sperm chromatin from Mock, Ufd1, Npl4, and p47 depleted egg extracts. Cdt1 and Sld5 levels are shown. Phosphorylated Histone 3 was used as loading control.

Figure 7. Hypothetical model for the coordination of CDT-1 turnover and GINS chromatin extraction

CDC-48/p97 together with the cofactors UFD-1/NPL-4 coordinates the turnover of ubiquitylated CDT-1 during the licensing phase with chromatin dissociation of CDC-45/GINS. Failure in CDT-1 turnover keeps CDC-45/GINS tightly associated with chromatin, becoming visible with condensing chromosomes at the end of S phase. CDC-45/GINS misregulation may interfere with dynamic progression of the replication fork in S phase. The CDC-48/p97 complex seems to coordinate both events since UFD-1/Ufd1 binds both CDC-45 and Cdt1.