Ubiquitin-specific protease 22 is a deubiquitinase of CCNB1

Abstract

The elevated level of CCNB1 indicates more aggressive cancer and poor prognosis. However, the factors that cause CCNB1 upregulation remain enigmatic. Herein, we identify USP22 as a CCNB1 interactor and discover that both USP22 and CCNB1 are dramatically elevated with a strong positive correlation in colon cancer tissues. USP22 stabilizes CCNB1 by antagonizing proteasome-mediated degradation in a cell cycle-specific manner. Phosphorylation of USP22 by CDK1 enhances its activity in deubiquitinating CCNB1. The ubiquitin ligase anaphase-promoting complex (APC/C) targets USP22 for degradation by using the substrate adapter CDC20 during cell exit from M phase, presumably allowing CCNB1 degradation. Finally, we discover that USP22 knockdown leads to slower cell growth and reduced tumor size. Our study demonstrates that USP22 is a CCNB1 deubiquitinase, suggesting that targeting USP22 might be an effective approach to treat cancers with elevated CCNB1 expression.

Keywords: APC8; CCNB1; USP22; cell cycle; tumorigenesis.

Figure 1

Both ubiquitin-specific protease 22 (USP22) and cyclin B1 (CCNB1) are upregulated in human colon cancers. (a) Flag-CCNB1 pull-down products from affinity purification using HCT116 cells were separated by SDS–PAGE and visualized by coomassie brilliant blue staining. (b) Interaction network of CCNB1-associated proteins. (c) The lysates of frozen colon tumor tissues (T) and their adjacent normal colon controls (N) were subjected to immunoblotting analysis with antibodies to USP22, CCNB1 and tubulin. Samples from 10 patients were analyzed. (d) The expression levels of USP22 and CCNB1 in normal human colon tissues (top) and colon cancer tissues (bottom) were determined by immunohistochemistry (IHC) staining with specific antibodies. Representative images are shown. (e) The expression levels of USP22 and CCNB1 in micro tissue array (MTA) slides of paraffin-fixed normal human colon or colon tumor tissue were determined by IHC with specific antibodies, and then scored and analyzed; *P<0.05; **P<0.01; ***P<0.001. NS, no significant difference. (f) The correlation of USP22 and CCNB1 protein expression was analyzed by IHC staining in normal and human colon cancer tissues.

Figure 2

USP22 interacts with CCNB1. (a) HCT116 cells were transfected with Flag-tagged CCNB1 expression plasmid. After 48 h, cells were lysed and incubated with GST-USP22 or GST-USP10 and GSH-Sepharose as well. Proteins retained on Sepharose were then blotted with the indicated antibodies. (b) HCT116 cells were lysed, and the interaction between endogenous CCNB1 and USP22 was determined by immunoprecipitation of CCNB1 using normal rabbit IgG as control and immunoblotting with anti-USP22 antibody (top panel). (c) Domain structures of USP22 and its truncated mutants. USP22 contains an N-terminal zinc-finger domain and a C-terminal C19 peptidase catalytic domain. (d) CCNB1 expression plasmids were co-transfected with USP22 or each of the truncated mutants shown in c into HCT116 cells. The interaction between USP22 or its mutants and CCNB1 was examined. (e) Domain structures of CCNB1 and its truncated mutants. CCNB1 contains an N-terminal destruction box (DB), followed by a cytoplasmic retention sequence (CRS) and a cyclin box domain. (f) Myc-USP22 expression plasmids were co-transfected with CCNB1 or each of the truncated mutants shown in e into HCT116 cells. The interaction of CCNB1 or its mutants was examined. CCNB1, cyclin B1; IgG, immunoglobulin G; USP2, ubiquitin-specific protease 22.

Figure 3

USP22 deubiquitinates and stabilizes CCNB1. (a and b) The mRNA levels of USP22, CCNB1, CCNA and CCNE in transfected HCT116 cells (a) or MEF cells (b) were analyzed by real-time PCR. Error bars represent data from three independent experiments. (c) Flag-CCNB1 and HA-Ub expression plasmids were co-transfected with empty vector, myc-USP22 or myc-USP10 into HCT116 cells. CCNB1 ubiquitination was determined by immunoprecipitation of CCNB1 with anti-Flag antibodies and immunoblotting with anti-HA antibody. (d) Deubiquitination of CCNB1 by USP22 in vitro. Ubiquitinated CCNB1 was purified from HCT116 cells transiently transfected with Flag-CCNB1 and HA-Ub expression plasmids, and the cell lysates were incubated with indicated GST or GST-fusion proteins at 37 °C for 2 h and then subjected to SDS–PAGE analysis. CCNB1 ubiquitination levels were determined by immunoblotting with anti-HA antibody. (e) Knockdown of USP22 by siRNA in HCT116 cells promotes CCNB1 ubiquitination. HCT116 cells were transfected with indicated siRNA and ubiquitination assay was analyzed as in c. (f) HCT116 cells stably expressing USP22, its C185A mutant, or USP22 shRNA were synchronized by treatment of nocodazole for 12 h, then mitotic cells were collected by shake-off approach and released into fresh medium for the indicated times. Levels of indicated proteins were detected by corresponding antibodies. Tubulin was used as loading control. (g) The band densities of CCNB1 in f were analyzed using a Bio-Rad imaging software. (h) The mRNA levels of USP22, CCNB1, CCNA and CCNE in the established HCT116 stable cell lines were analyzed as in a). (i) Loss of USP22 facilitates CCNB1 degradation in MEF cells. The indicated proteins in wild-type or USP22 knockout MEF cells were analyzed as in f. (j) HCT116 cells stably expressing control or USP22-specific shRNA were treated with or without MG132 (20 μm) for 2 h before harvested. The expression levels of CCNB1, USP22 and tubulin were analyzed by immunoblotting. CCNA, cyclin A; CCNB1, cyclin B1; CCNE, cyclin E; MEF, mouse embryonic fibroblast; shRNA, short hairpin RNA; siRNA, small interfering RNA; USP2, ubiquitin-specific protease 22.

Figure 4

Cyclin-dependent kinase 1 (CDK1) phosphorylates ubiquitin-specific protease 22 (USP22) to promote its deubiquitinase activity. (a) The endogenous interaction between CDK1 and USP22 was analyzed as in Figure 2b. (b) Myc-USP22 plasmids were co-expressed with wild-type (WT), the constitutively active form (AF) of CDK1, or the kinase-inactive D146N mutant of CDK1. After 24 h, cells were treated without or with nocodazole for 12 h. USP22 phosphorylation in the lysates of transfected cells was analyzed. (c) HCT116 cells were treated with double thymidine (Thy–Thy) or with thymidine followed by nocodazole (Thy–Noc). The cell lysates were treated with or without calf intestinal alkaline phosphatase (CIP) for 1 h as indicated. USP22 phosphorylation in the lysates of treated cells was analyzed as described in b. (d) Conserved T147 and S237 amino acids of USP22 in each indicated species are shown. (e) WT USP22 or its phosphorylation-defective mutant USP22/AA was co-expressed in HCT116 cells. Their phosphorylation was determined as in b. (f) HCT116 cells stably expressing WT USP22 or its phosphomimetic mutant (USP22/DD) or phosphorylation-defective mutant USP22/AA were synchronized in prometaphase with Thy and Noc treatment, then released into fresh medium for the indicated times (Noc-re). (g) HA-ubiquitin-conjugated Flag-CCNB1 was affinity purified from co-transfected HCT116 cells. Flag-USP22 was purified from HCT116 cells and treated with or without CIP for 1 h. Ubiquitinated CCNB1 was then mixed with USP22 for indicated time and analyzed by immunoblotting.

Figure 5

APC^CDC20 destructs USP22 protein during cell exit from M phase. (a) HCT116 cells were arrested at the G1/S boundary with a double thymidine treatment and then released into fresh medium. Cells were collected every 2 h and lysates were analyzed by immunoblotting with indicated antibodies. (b) HCT116 cells were synchronized in prometaphase with thymidine and nocodazole, then released into fresh medium without or with 20 μm MG132 treatment for the indicated durations. (c) USP22 specifically interacts with CDC20 but not with FBW7 or CDH1. HCT116 cells were transfected with indicated plasmids and USP22 interactions with each of them were analyzed as in Supplementary Figure S1B. (d) CDH1, CDC20 or FBW7 expression plasmids were transfected into HCT116 cells. The endogenous levels of USP22 protein in transfected cells were analyzed by immunoblotting (top panel). (e) CDC20 is required for USP22 destruction in mitosis. Cdc20 was depleted from HCT116 cells using siRNA, cells were synchronized by treatment of nocodazole for 12 h, then mitotic cells were collected by the shake-off approach and released into fresh medium for the indicated times. Levels of indicated proteins were detected by corresponding antibodies. Tubulin was used as a loading control. (f) USP22 interacts with APC8. USP22 plasmids were co-transfected without or with each of the indicated APC proteins. Their interactions were determined as described in c. (g) HCT116 cells were transfected with control or CDC20-specific siRNA. The interaction of USP22 with APC8 was examined as in c. (h) APC8 is required for USP22 destruction in mitosis. APC8 was depleted from HCT116 cells using siRNA, then cells were synchronized and analyzed as in e. (i) Sequence alignment of putative USP22 D-box motif compared with those from cyclin A, cyclin B1 and securin. The two residues in USP22 mutated to alanine to yield the USP22 D-box mutant (R98A/L101A) are indicated (upper panel). The conserved D-box sequence of USP22 is indicated (lower panel). (j) The USP22 D-box mutant is stable in cells arrested in mitosis. HCT116 cells stably expressing either wild-type USP22 (WT) or the D-box mutant were synchronized and analyzed as in e. APC, anaphase-promoting complex; siRNA, small interfering RNA; USP2, ubiquitin-specific protease 22.

Figure 6

Knockdown of USP22 by shRNA inhibits colorectal tumorigenesis. (a) 2×10⁵ cells were seeded in six-well plates. The growth of cells was examined by counting cell numbers every 24 h after plating. Error bars represent data from three independent experiments; P<0.05. (b) The growth of stably expressing HCT116 cells indicated plasmids was examined by WST-1 assay as described in materials and methods. (c) Wild-type or usp22-null MEF cells were subjected to WST-1 assay as in b. (d) Stably expressing HCT116 cells indicated plasmids were seeded in soft agar and cultivated for 3–4 weeks. Colony formation was assayed by light microscopy and representative images are shown. (e) The number of colonies in each plate was counted. Error bars represent data from three independent experiments with a total of three plates per group. P<0.05. (f–h) 1×10⁶ HCT116 or stable USP22 knockdown (KD) cells were injected subcutaneously into nude mice (n=8 per group). Twenty-two days after injection, two pairs of representative mice were photographed (f, top panel). Tumors were isolated and photographed. 7–8 represents no tumors (f, bottom panel). The tumor sizes were measured every other day, and their growth curves are shown (g). Mice were euthanized at day 22, and the tumors were excised and weighed at the end of the experiment (h). P<0.05. (i) Our working model of CCNB1 regulation by USP22. CCNB1, cyclin B1; shRNA, short hairpin RNA; siRNA, small interfering RNA; USP2, ubiquitin-specific protease 22.