DTL/CDT2 is essential for both CDT1 regulation and the early G2/M checkpoint

Abstract

Checkpoint genes maintain genomic stability by arresting cells after DNA damage. Many of these genes also control cell cycle events in unperturbed cells. By conducting a screen for checkpoint genes in zebrafish, we found that dtl/cdt2 is an essential component of the early, radiation-induced G2/M checkpoint. We subsequently found that dtl/cdt2 is required for normal cell cycle control, primarily to prevent rereplication. Both the checkpoint and replication roles are conserved in human DTL. Our data indicate that the rereplication reflects a requirement for DTL in regulating CDT1, a protein required for prereplication complex formation. CDT1 is degraded in S phase to prevent rereplication, and following DNA damage to prevent origin firing. We show that DTL associates with the CUL4-DDB1 E3 ubiquitin ligase and is required for CDT1 down-regulation in unperturbed cells and following DNA damage. The cell cycle defects of Dtl-deficient zebrafish are suppressed by reducing Cdt1 levels. In contrast, the early G2/M checkpoint defect appears to be Cdt1-independent. Thus, DTL promotes genomic stability through two distinct mechanisms. First, it is an essential component of the CUL4-DDB1 complex that controls CDT1 levels, thereby preventing rereplication. Second, it is required for the early G2/M checkpoint.

Figure 1.

Insertional mutations in the zebrafish dtl gene cause a G2 DNA damage checkpoint defect. (A) An early G2 checkpoint response was verified by labeling mitotic cells using an anti-pH3 antibody in untreated (−IR), IR-treated (+IR), and caffeine-pretreated/IR-treated (+IR; +Caffeine) 32-hpf wild-type zebrafish embryos. (B) Scheme of the shelf screen used to identify recessive lethal insertional mutants with G2 checkpoint defects. (C) Analysis of embryos derived from hi3627 and hi447 heterozygous crosses showed a presence of a robust early G2 checkpoint response in wild-type (+/+; +/−) but not mutant (−/−) clutchmates. (D) The positions of the hi447 and hi3627 insertions relative to the first three exons of the zebrafish dtl gene. Analysis of total mRNA by RT–PCR showed decreased dtl mRNA, relative to the mrpL13 loading control, in 28-hpf hi3627 and hi447 mutant embryos. (E) dtl morpholinos reduce dtl mRNA levels and impair the IR-induced G2 arrest in 28-hpf embryos. (F) Nonirradiated (−IR) and irradiated (+IR) wild-type (+/+; +/−) and hi3627 mutant (−/−) clutchmates were maintained in the absence or presence of nocodazole for 2 h, and the percentage of pH3-positive cells was quantified by FACS (mean ± SD; n = 20,000 counts in each of three embryos).

Figure 2.

Dtl is required for normal cell cycle progression in zebrafish embryos. (A) Cell cycle profiles of cells from 32-hpf hi3627 mutant (−/−) and wild-type clutchmate (+/+; +/−) zebrafish embryos showing an increase in 4N and >4N cells in the mutants. (B) The S-to-M transition was slower in hi3627 mutant (−/−) versus wild-type (+/+; +/−) clutchmates, as assessed by pulse labeling with BrdU, incubating for the indicated times, and determining the fraction of pH3-positive (mitotic) cells that had incorporated BrdU (mean ± SD; n = 5 embryos). (C) Nuclear diameter was increased in hi3627 mutant (−/−) versus wild-type (+/+; +/−) 32-hpf clutchmates (n = 40 cells in each of five embryos). (D) Mitotic cells in 32-hpf hi3627 mutants stained with anti-pH3 (green) and anti-α-tubulin (red) have supernumerary spindle poles. (E) Quantitation of prophase, metaphase, or anaphase cells (mean ± SD; n = 3 embryos), based on chromatin and spindle morphology, revealed a significant loss of anaphase cells in hi3627 mutants (−/−).

Figure 3.

Human DTL interacts with the CUL4A–DDB1 complex. (A) TAP alone or TAP–DTL expressed in HeLa-S3 cells was purified from whole-cell lysates using a TAP procedure. Immunoblotting shows DTL, DDB1, and CUL4A in the input lysate copurified with TAP–DTL but not the TAP control. The asterisk indicates endogenous DTL. TEV protease cleavage caused a slight change in TAP–DTL’s size. (B,C) Whole-cell lysates were prepared from either unirradiated (−IR) or irradiated (+IR) HeLa cells, and endogenous DTL (B) or CUL4A (C) was immunoprecipitated. Input lysate and the immunoprecipitates were then screened for DTL, DDB1 and CUL4A proteins by immunoblotting.

Figure 4.

DTL is required for regulating CDT1 and activating the G2 checkpoint in human cells. (A) HeLa cells were transfected with single siRNAs, siGFP, or siDTL(#1), or a pool of four distinct siRNAs [siDTL(#2–#5)], and after 72 h, the levels of DTL and CDT1, relative to the β-actin (ACTB) loading control, were determined by immunoblotting of whole-cell lysates. (B) Cell cycle profiles of siDTL(#1)- or siDTL(#2–#5)-transfected HeLa cells (72 h post-transfection) or HCT-116 cells (96 h post-transfection). (C,D) HeLa cells were transfected with siRNAs as described in A, and half of the cells were exposed to 10 Gy of ionizing radiation (+IR) and then incubated for 1 h before labeling with anti-pH3 and PI. (C) The percentages of pH3-positive cells were calculating using FloJo software. (D) Averages from four separate experiments are shown (mean ± SD; n = 20,000 counts in each experiment).

Figure 5.

Knockdown of cdt1 rescues 4N accumulation in dtl morphants but not the G2 checkpoint defect. (A) Zebrafish embryos were injected with vehicle (Control), dtl morpholino, cdt1 morpholino, or both (dtl + cdt1). Cell cycle profiles of cells dissociated from those morpholino-injected embryos at 28 hpf are shown. (B) Mitotic cells in 28-hpf zebrafish embryos, which were morpholino-injected as in A, were labeled with an anti-pH3 antibody 1 h after the embryos were exposed to 15 Gy of IR. (C,D) Total RNA was prepared from morpholino-injected embryos at 28 hpf and, using β-actin as a control, were analyzed for cdt1 mRNA by real-time PCR (C) (mean ± SD; n = 4) or dtl mRNA by RT–PCR (D).

Figure 6.

DTL is required both for the down-regulation of CDT1 during S phase and for the activation of the early G2/M checkpoint. The CUL4–DDB1–DTL complex inhibits CDT1 during S phase. Loss of DTL causes inappropriate expression of CDT1 during S phase, leading to rereplication, prolonged G2 checkpoint activation, multipolar spindles, and mitotic catastrophe. DTL loss also prevents the activation of the IR-induced early G2/M checkpoint in a CDT1-independent manner.