The ubiquitin ligase CRL2ZYG11 targets cyclin B1 for degradation in a conserved pathway that facilitates mitotic slippage

Abstract

The anaphase-promoting complex/cyclosome (APC/C) ubiquitin ligase is known to target the degradation of cyclin B1, which is crucial for mitotic progression in animal cells. In this study, we show that the ubiquitin ligase CRL2^ZYG-11 redundantly targets the degradation of cyclin B1 in Caenorhabditis elegans and human cells. In C. elegans, both CRL2^ZYG-11 and APC/C are required for proper progression through meiotic divisions. In human cells, inactivation of CRL2^ZYG11A/B has minimal effects on mitotic progression when APC/C is active. However, when APC/C is inactivated or cyclin B1 is overexpressed, CRL2^ZYG11A/B-mediated degradation of cyclin B1 is required for normal progression through metaphase. Mitotic cells arrested by the spindle assembly checkpoint, which inactivates APC/C, often exit mitosis in a process termed "mitotic slippage," which generates tetraploid cells and limits the effectiveness of antimitotic chemotherapy drugs. We show that ZYG11A/B subunit knockdown, or broad cullin-RING ubiquitin ligase inactivation with the small molecule MLN4924, inhibits mitotic slippage in human cells, suggesting the potential for antimitotic combination therapy.

Figure 1.

C. elegansCYB-1 is a direct substrate of the CRL2^ZYG-11 complex. (A) Egg hatching percentage ± SEM for zyg-11(ts) at the semipermissive temperature of 20°C with the listed RNAi treatments; from at least three independent counts, with the total number of embryos analyzed (from left to right): 667, 1,163, 448, and 222, respectively. (B) ZYG-11 negatively regulates CYB-1 levels in a proteasome-dependent manner. VSV-G–CYB-1 was coexpressed with FLAG–ZYG-11 or control FLAG–CYE-1 in HEK293T cells treated with human ZYG11A/B siRNA. Cells were treated for 8 h with nocodazole and (where indicated) the proteasome inhibitor MG132 for the last 4 h. Western blots of whole cell lysate were probed with the indicated antibodies. The relative VSV-G–CYB-1 to actin ratio is given below the top blot. Similar results were obtained in six independent experiments with varied cell treatments (± ZYG11A/B knockdown, ± nocodazole, and ± nocodazole and proTAME). (C) ZYG-11 and CYB-1 physically interact. VSV-G–CYB-1 was coexpressed in HEK293T cells with either FLAG–ZYG-11 or FLAG–CUL-4 in the presence of the proteasome inhibitor LLnL. Anti-FLAG antibody was used for IP, with analysis by Western blot (top two panels). Similar results were obtained in two independent experiments. (D) Endogenous ZYG-11 physically interacts with GFP::CYB-1 in C. elegans. Wild-type and animals expressing CYB-1::GFP were grown ± cul-2 RNAi. GFP::CYB-1 was immunoprecipitated and analyzed by Western blot (top two panels). (E) ZYG-11 physically interacts with the CBOX1 domain of CYB-1. FLAG–ZYG-11 was coexpressed with VSV-G–tagged CYB-1 truncations, NTER, CBOX1, or CBOX2 in HEK293T cells in the presence of MG132. FLAG–ZYG-11 was immunoprecipitated and analyzed by Western blot (top two panels). A diagram of the full-length and truncated CYB-1 proteins is shown at bottom. The asterisk marks the location of the residue homologous to the cyb-2.1 E120K zyg-11 mutant suppressor allele. (F) CBOX1 of CYB-1 is degraded during meiosis in the C. elegans embryo in a ZYG-11–dependent manner. Bar graph showing the levels ± SEM of GFP::CYB-1 truncations in two-cell stage embryos relative to the level in meiosis I–stage embryos (set to 100) in the RNAi treatments listed; from left to right, n = 5, 4, 12, 4, and 7, respectively. P-values are calculated for GFP levels in two-cell stage embryos versus the corresponding meiotic embryos. For all figures, asterisks above bars represent p-values relative to the control sample; asterisks above lines denote comparisons under the lines. ns, not significant; *, P < 0.05; **, P < 0.01.

Figure 2.

Human CRL2^ZYG11A/B physically interacts with cyclin B1. (A) IP of endogenous cyclin B1 coprecipitates FLAG-ZYG11B in HEK293T cells treated with MG132. Left panels are for cyclin B1 IP and right panels are whole cell lysate. Similar results were obtained in two independent experiments. (B) In vitro ubiquitylation of cyclin B1 by the CRL2^ZYG11B complex. T7-tagged cyclin B1 and FLAG-ZYG11B were coexpressed in HEK293T cells in the presence of MG132. The CRL2^ZYG11B complex was immunoprecipitated with anti-FLAG antibody (1st IP). The immunocomplex (with associated T7–cyclin B1) was used in an in vitro ubiquitylation reaction with HA-tagged ubiquitin (Ub). T7–cyclin B1 was subsequently immunoprecipitated in the presence of 0.1% SDS (2nd IP), followed by immunoblotting for HA-ubiquitin and T7–cyclin B1. Similar results were obtained in two independent experiments. (C) Western blot showing specificity of anti-ZYG11A and anti-ZYG11B antibodies on whole cell lysate from HEK293T cells overexpressing FLAG-CDT2, FLAG-ZYG11A, or FLAG-ZYG11B. Similar results were obtained in three independent experiments. (D) IP/Western blot with control, anti-ZYG11A, and anti-ZYG11B antibodies in U2OS cells. Gels were cut above the chicken IgY antibody heavy chain band. Similar results were obtained in two independent experiments. (E) IP of endogenous cyclin B1 coprecipitates endogenous ZYG11B in HeLa and HEK293T cells treated with MG132. Control (anti–VSV-G) and anti–cyclin B1 IPs were performed from the same initial cell lysate. Similar results were obtained in three to four independent experiments.

Figure 3.

Human CRL2^ZYG11A/B negatively regulates the level of cyclin B1 in mitosis. (A) Western blot demonstrating the effectiveness of ZYG11A and ZYG11B siRNA knockdown of FLAG-ZYG11A and FLAG-ZYG11B expressed in U2OS cells. The relative FLAG-ZYG11A/B to actin ratio is given below the blots. Similar results were obtained in two independent experiments. (B) IP/Western blot with anti-ZYG11A and anti-ZYG11B antibodies demonstrating the effectiveness of the siRNA knockdown of the endogenous proteins in U2OS cells. The relative ZYG11A/B to actin ratio is given below the blots. Comparable RNAi knockdowns were also observed in HEK293T cells (not depicted). (C) ZYG11A, ZYG11B, or ZYG11A/B siRNA knockdowns increase the level of endogenous cyclin B1 in cells treated with proTAME at 2 h after release from RO-3306 G₂ arrest. Epifluorescence images of representative β-tubulin, DNA, and endogenous cyclin B1 staining of metaphase-arrested proTAME-treated cells. Bar, 10 µm. (D) Bar graph shows the relative level of cyclin B1 ± SEM in the indicated siRNA treatments; n = 20 cells for each condition. Bar graphs presenting the ratios of the chromosome-to-cytoplasm (E) and chromosome-to-spindle (F) immunofluorescence signals ± SEM for endogenous cyclin B1 for the experiment described in C; n = 25 cells for each condition. For C–F, similar results were obtained in two independent experiments. ***, P < 0.001; ****, P <0.0001.

Figure 4.

Human CRL2^ZYG11B interacts with cyclin B1 independently of Cdk1. (A) FLAG-ZYG11B interacts with Venus–cyclin B1 mutant proteins that are incapable of binding Cdk1. FLAG-ZYG11B was coexpressed with wild-type or mutated versions of Venus–cyclin B1 with R202A or R202A R42A substitutions in HEK293T cells that were treated with MG132. The Venus tag was immunoprecipitated and immunoblotted for Venus–cyclin B1, FLAG-ZYG11B, or Cdk1 (left). Similar results were obtained in two independent experiments. (B) Endogenous Cdk1 coprecipitates HA-ZYG11B but not control HA–leucine-rich repeat (LRR) 1 in HEK293T cells treated with MG132. Similar results were obtained in two independent experiments. (C) HA-ZYG11B predominantly interacts with the CBOX1 domain of cyclin B1. HA-ZYG11B was coexpressed with Venus–cyclin B1, Venus-tagged truncations of cyclin B1 (N terminus, CBOX1, or CBOX2), or pEGFP-N1 control vector in HEK293T cells treated with MG132. Note that the truncations were cleaved into smaller isoforms in vivo (the full-length size is marked by an arrow on the right side). The Venus tag was immunoprecipitated and immunoblotted for HA or Venus tags (left panels). The results are representative of five independent experiments.

Figure 5.

Human CRL2^ZYG11A/B is required for mitotic progression in cells overexpressing cyclin B1. (A) DIC micrographs of U2OS cells treated with the indicated siRNAs as they progress through mitosis. Timing in minutes is shown in the top right corner of images; the 0 time point denotes nuclear envelope breakdown. The last time point reflects mitotic exit. Bars, 20 µm. (B) Bar graph of mitotic timing ± SEM for siRNA-treated U2OS cells. Timing data from two to four independent experiments are summarized in the graph; for the conditions listed from top to bottom, n = 19, 18, 22, and 29, respectively. (C) Matching DIC and epifluorescence micrographs of U2OS cells expressing Venus–cyclin B1 treated with control and ZYG11A/B siRNA. (D) Bar graph of mitotic timing ± SEM for siRNA-treated U2OS cells expressing Venus–cyclin B1. Timing from three independent experiments are summarized in the graph; for conditions listed from top to bottom, n = 6 and 8, respectively. Bars, 20 µm. (E) Line graph of the level of Venus–cyclin B1 in cells as they progress through mitosis (times are after nuclear envelope breakdown); n = 6 and 8, as described in D. **, P < 0.01; ***, P < 0.001.

Figure 6.

Human CRL2^ZYG11A/B functions redundantly with APC/C to promote mitotic progression. (A) DIC micrographs of U2OS cells treated with the indicated siRNA and proTAME as they progress through mitosis. Bars, 20 µm. Bar graph of mitotic timing ± SEM for U2OS cells treated with the indicated siRNA in the presence of proTAME. The timing of control siGFP without proTAME is from Fig. 4 B and presented in the figure for comparison. The timings are from four independent experiments: control siRNA + proTAME, n = 23; and ZYG11A/B siRNA + proTAME, n = 25. (B) Inhibition of Cdk1 activity by the addition of the Cdk-inhibitor RO-3306 (arrows) allows mitotic exit. Data from three independent experiments are shown. (C) Bar graph of mitotic timing ± SEM for IMR90-hTERT cells treated with the indicated siRNA in the presence of proTAME. The timings are from two to three independent experiments; for conditions listed from top to bottom, n = 12, 13, and 10, respectively. *, P < 0.05; ***, P < 0.001; ****, P < 0.0001.

Figure 7.

Inactivation of ZYG11A/B blocks mitotic slippage in nocodazole-treated cells. U2OS (A) and IMR90-hTERT (B) cells were treated with nocodazole under the conditions listed. Bar graphs represent the percentage of cells that underwent mitotic slippage or cell death. The χ² significance for the occurrence of mitotic slippage is relative to nocodazole + control siRNA. The graphs summarize the data from two to four independent experiments for each condition. From left to right: for A, n = 33, 39, 27, and 22; and for B, n = 17 and 25, respectively. *, P < 0.05; ***, P < 0.001; ****, P < 0.0001.