Abstract
Progression of mitotic cell cycle and chromosome condensation and segregation are controlled by posttranslational protein modifications such as phosphorylation and SUMOylation. However, how SUMO isopeptidases (SENP) regulate cell mitotic procession is largely unknown. Here, we demonstrate that precise phosphorylation of SENP3 during mitosis suppresses SENP3 deSUMOylation activity towards chromosome-associated proteins, including topoisomerase IIα (TopoIIα). Cyclin B-dependent kinases 1 and protein phosphatase 1α were identified as the kinase and phosphatase in control of mitotic SENP3 phosphorylation, respectively. SENP3 phosphorylation decreased its interaction with TopoIIα, resulting in reduced SENP3 deSUMOylation activity on TopoIIα. Furthermore, we observed mitotic arrest, increased chromosome instability, and promotion of tumorigenesis in cells expressing a nonphosphorylatable SENP3 mutant. These data show that SENP3 phosphorylation plays a crucial role in regulating the SUMOylation of chromosome-associated proteins and chromosome stability in mitosis.
Significance: Phosphorylation of SENP3 regulates SUMOylation of chromosome-associated proteins to maintain genomic stability during mitosis. Cancer Res; 78(9); 2171–8. ©2018 AACR.
Introduction
Mitosis is the fundamental process by which cellular contents are segregated into two daughter cells. Protein phosphorylation is responsible for the initiation of mitosis including chromosome condensation and segregation (1), which is tightly and accurately controlled by numerous kinases and phosphatases. Among them, cyclin B-dependent kinase 1 (CDK1) and protein phosphatase-1 (PP1) are critical in controlling mitotic protein phosphorylation to maintain precise chromosome condensation and segregation (2, 3). Once cells have duplicated their DNA, CDK1 becomes activated by cyclin B, promoting cellular processes such as centrosome maturation and separation, chromosome condensation, and mitotic entry after nuclear envelope breakdown by phosphorylation of mitotic-related proteins. Upon exit of mitosis, PP1 is one of the phosphatases to dephosphorylate mitotic-related proteins to allow cells to re-enter G1 phase of next cell cycle.
Like phosphorylation, accumulated evidence has shown that the SUMO modification pathway plays a pivotal role in regulation of chromosome segregation and cell-cycle progression (4). Multiple functions of SUMOylation have been proposed at mitotic centromeres and kinetochores, and various SUMOylated proteins within these domains are critical for accurate distribution of chromosomes into two daughter cells after mitosis (5). Many centromeric proteins, such as aurora B, Borealin, topoisomerase IIα (TopoIIα), and CENP-I are subjected to SUMOylation at the G2–M phase, which facilitate their functions in control of mitotic progression (6, 7). DeSUMOylation by SUMO isopeptidases (SENP) is essential to ensure the reversible nature of SUMO conjugation. There are six different isopeptidases (SENP1, SENP2, SENP3, SENP5, SENP6, and SENP7) in human cells, all of them share a common C-terminal catalytic domain but have distinct N-terminal domains, which are critical for their subcellular localizations and substrate specificities (8). Deregulated SENP2 or SENP6 can lead to abnormal centromere or kinetochore protein SUMOylation and irregular mitotic cycle in the cells (6, 9). Silencing SENP3 would either delay cells before mitotic entry or by-pass Taxol-induced mitotic arrest (10). Although SENPs are critical regulators of mitosis, it is still largely unknown how exactly they are regulated during the cell cycle.
Chromosome instability (CIN) is one of the hallmarks of cancer cells. Most solid tumor cells with CIN are aneuploidy, indicating that abnormal mitosis is involved in the CIN phenotype (11). SUMOylation emerges as a regulation mechanism for the proper segregation of chromosomes in mitosis (4). TopoIIα has been shown to be SUMOylated at mitosis by SUMO E3 ligase RanBp2 in mammalian cells, and SUMOylation of TopoIIα is essential for decatenation of chromosome at centromere (12, 13). RanBP2 deficiency downregulates TopoIIα SUMOylation, leading to abnormal mitotic chromosome segregation and chromosome instability (12). However, it is unknown how SENPs engage in this process.
Here we show that SENP3 is phosphorylated by protein kinase CDK1 upon entry into mitosis and is dephosphorylated by protein phosphatase PP1α upon mitotic exit. Mitotic phosphorylation turns down SENP3 deSUMOylation of chromosome-associated proteins including TopoIIα. Expression of nonphosphorylatable SENP3 mutant decreases the SUMOylation of chromosome-associated proteins and TopoIIα, which resulted in abnormal chromosome segregation, mitotic cell cycle, and tumorigenesis. These data reveal that mitotic phosphorylation plays a crucial role in SENP3, regulating the SUMOylation and function of chromosome-associated proteins in mitosis.
Materials and Methods
Cell culture and reagents
U2OS and Hela cells (ATCC) were purchased from ATCC (December 2007) and were detected for mycoplasma infection with LookOut Mycoplasma PCR Detection Kit (Sigma, #MP0035) as the manual. U2OS and Hela cells were authenticated using short tandem repeat (STR) profiling analysis by Suzhou Genetic Testing Biotechnology Co. Ltd. Cells were cultured in DMEM (Gibco) plus 10% FBS (Gibco) supplemented with 1% penicillin–streptomycin (Gibco) at 37°C and 5% CO2. To generate stable transfection cell lines, SENP3(WT) or SENP3(9A) with or without TopoIIα was cloned into the lentivector pCDH (System Biosciences, #CD510B-1). Production of lentivirus and infection were carried out according to user manual. The cells were selected with 2 μg/mL puromycin (Amresco) for 1 week. For synchronization experiments, cells were synchronized either to the G1–S phase using 2 mmol/L thymidine (Sigma) for double block, or to the G2–M phase using 50 ng/mL nocodazole (Sigma) for 24 hours. Ro-3306 and Okadaic acid were purchased from Selleck and Santa Cruz Biotechnology, respectively. λ-PPase was purchased from New England Biolabs.
Construction of plasmids and siRNA
Full-length cDNAs of SENP1, SENP2, and SENP3 were cloned into 3 × FLAG (Sigma) vector, respectively, by using standard procedures. Site-specific mutation of SENP3 was introduced using a QuikChange Site-Directed Mutagenesis Kit (Agilent). All constructs were confirmed by sequencing. The siRNA sequences (GenePharma) were as follows: negative-control (NC) siRNA, 5′-UUCUCCGAAGGUGUCACGUTT-3′; CDK1 siRNA, 5′-GGCACUGAAUCAUCCAUAUTT-3′; PP1α siRNA, 5′-GAGACCCUCAUGUGCUCUUTT-3′. All the siRNAs were transfected with Lipofectamine2000 (Invitrogen).
Immunofluorescence
Cultured U2OS cells were transient transfected with GFP-H2B and Flag-SENP3(WT) or Flag-SENP3 (9A) at ratio of 1:5. After 36 hours, cells were stained with DAPI and observed under microscope with green fluorescence.
Cell fractionation
Cell fractionation was performed as described (14). Cells were resuspended (1 × 107 cells/mL) in buffer A [10 mmol/L HEPES, pH 7.9, 10 mmol/L KCl, 1.5 mmol/L MgCl2, 0.34 M sucrose, 10% glycerol, 1 mmol/L DTT, 1 mmol/L PMSF, and protease inhibitors cocktail; Sigma). Triton X-100 (0.1%) was added, and the cells were incubated for 5 minutes on ice. Cells were centrifuged at 1,300 × g, 4°C for 10 minutes, and then supernatant (S1) was discarded, whereas pellet (P1) was collected and washed once in buffer A. Then pellet was lysed in buffer B (3 mmol/L EDTA, 0.2 mmol/L EGTA, 1 mmol/L DTT, protease inhibitors cocktail). Insoluble chromatin pellet (P3) was collected by centrifugation (5 min, 1,700 × g, 4°C), soluble supernatant (S3) was discarded, and then pellet was washed once in buffer B for subsequent studies.
Western blot analysis, immunoprecipitation, and Talon beads pull-down
U2OS cells was lysed with RIPA buffer [50 mmol/L Tris (pH 7.4), 300 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L EGTA, 1% Triton, 1 mmol/L Na3VO4, 1 mmol/L β-glycerophosphate, 1 mmol/L PMSF, protease inhibitor cocktail). Supernatant was used for routine immunoprecipitation and Western blotting with anti-SENP3(CST, #5591), anti-SENP1(GeneTex, #GTX115889), anti-SENP2 (Abcam, #ab3660), anti-cyclin B1 (Santa Cruz Biotechnology, #sc-595), anti-Flag (Sigma, #F1804), anti-actin (Sigma, #A5441), anti-TopoIIα (Topogen, #TG2011-1), anti-SUMO2 (Abcam, #ab3742) antibodies, anti-MAD2 (Santa Cruz Biotechnology, #SC-47747), anti-Histone H3 (proteintech, #17168-1-AP), anti-CDK1 (Santa Cruz Biotechnology, #SC-54), anti-PP1a (CST, #2582S), anti-Lamin B1 (Santa Cruz Biotechnology, #SC-30264), antiphospho-(Ser/Thr) antibody (Abcam, #ab17464). Talon beads pull-down was performed after cells were transfected with RGS-His-tagged SUMO2 and Flag-SENP3(WT) or Flag-SENP3 (9A) for 48 hours, Cells were lysed, and RGS-His-tagged SUMO2 was pulled down by Talon beads (Clontech Laboratories) according to the manufacturers' protocols.
In vitro assay
U2OS cells were transfected with Flag-SENP3 (WT) or Flag-SENP3 (9A) and then SENP3 was purified by using anti-Flag M2 beads (Sigma). Purified Flag-SENP3 (WT) or Flag-SENP3 (9A) was incubated with cyclin B1/CDK1 (catalog no. BML-SE295-0010; Enzo Life Sciences), with or without PP1α (catalog no. 14-595; Merck Millipore) at 30°C for 45 minutes. For in vitro interaction assay, purified Flag-SENP3 (WT) or Flag-SENP3 (9A) was incubated with cyclin B1/CDK1, with or without PP1α or TopoIIα (Topogen, #TG2000H-2) at 4°C overnight. The reaction samples were subjected to SDS-PAGE and blotted with indicated antibodies.
FACS
FACS analysis for cell cycle was performed using standard protocol and propidium iodide DNA staining (Biolegend). Percentage of cells in each cell-cycle stage was counted and analyzed with FlowJo softwar. For analysis of apoptosis, cells were stained with Annexin V and 7-AAD (Biolegend) according to the manufacturer's instructions. Flow cytometry was performed using FACS caliber (Becton Dickinson), and the data were analyzed with FlowJo softwar.
Animal study approval
Animal study was approved by Shanghai Jiao Tong University School of Medicine Animal Care and Use Committee. All mice were kept in a specific pathogen-free facility in the university.
Each 8-week-old nude mouse was subcutaneously injected with 5 × 106 cells. Three weeks after injection, mice were sacrificed, and the weight and volume of tumors were measured.
Statistical analyses
Statistical analysis was performed with Excel software. In all experiments, two-sided Student t test was used to test for differences between two groups. P values of <0.05 were considered statistically significant.
Results
SENP3 phosphorylation in mitotic phase of the cell cycle
Stegmeier and colleagues utilized shRNA screening to identify SENP3 as a mitotic regulator and found that depletion of SENP3 would either delay cells before mitotic entry or by-pass Taxol-induced mitotic arrest (10). Dephoure and colleagues identified SENP3 as a phosphorylated protein in cells arrested in mitotic phase of cell cycle (15). We thus proposed that the phosphorylation would regulate SENP3 function as a mitotic regulator. Consistent with other reports, we detected the upshifted SENP3 bands in nocodazole-treated cells (Fig. 1A). We also detected the similar slow migration bands of endogenous SENP3 upon mitotic arrest by either nocodazole or taxol (Fig. 1B; Supplementary Fig. S1A). SENP3 upshifted bands disappeared after lambda-PPase treatment, indicating these upshifted bands as the phosphorylated SENP3 forms in mitosis (Fig. 1C; Supplementary Fig. S1B). However, we did not observe any shifted bands in SENP1 or SENP2 in the M phase of synchronized cells (Fig. 1A and B). We further utilized thymidine or nocodazole arrest and release procedure to confirm that these shifted bands only appeared in the G2–M phases of the cell cycle, suggesting that the mitotic phosphorylation might modulate SENP3 function in the G2–M phases (Fig. 1D and E).
Cell-cycle–regulated phosphorylation of SENP3. A, Flag-SENP3 is phosphorylated upon nocodazole (Nocod) arrest. U2OS cells were transfected with indicated constructs for 36 hours and then treated with nocodazole for 24 hours. Cell lysates were collected for Western blot analysis with indicated antibodies. B, Endogenous SENP3 is phosphorylated upon nocodazole arrest. U2OS cells treated with nocodazole for 24 hours. Cell lysates were collected for Western blot analysis with indicated antibodies. C, Dephosphorylation of SENP3 by λ-phosphatase in vitro. Cell lysates from asynchronous or synchronous U2OS cells were subjected to λ-phosphatase treatment, and then Western blot analysis was performed with indicated antibodies. D, SENP3 is phosphorylated upon entry into the G2–M phase after thymidine release. U2OS cells were treated with thymidine for 24 hours and then released into fresh medium for the indicated times. Cell lysates were collected for Western blot analysis with indicated antibodies. E, SENP3 is dephosphorylated upon nocodazole release. U2OS cells were treated with nocodazole for 24 hours and then released into fresh medium for indicated times. Cell lysates were collected for Western blot with the indicated antibodies. F, Mutation of SENP3 phosphorylation sites. U2OS cells were first transfect with Flag–SENP3(WT) and Flag–SENP3(9A) for 36 hours, followed by nocodazole treatment. Cell lysates were collected for Western blot with indicated antibodies. p-SENP3, phosphorylated SENP3.
Cell-cycle–regulated phosphorylation of SENP3. A, Flag-SENP3 is phosphorylated upon nocodazole (Nocod) arrest. U2OS cells were transfected with indicated constructs for 36 hours and then treated with nocodazole for 24 hours. Cell lysates were collected for Western blot analysis with indicated antibodies. B, Endogenous SENP3 is phosphorylated upon nocodazole arrest. U2OS cells treated with nocodazole for 24 hours. Cell lysates were collected for Western blot analysis with indicated antibodies. C, Dephosphorylation of SENP3 by λ-phosphatase in vitro. Cell lysates from asynchronous or synchronous U2OS cells were subjected to λ-phosphatase treatment, and then Western blot analysis was performed with indicated antibodies. D, SENP3 is phosphorylated upon entry into the G2–M phase after thymidine release. U2OS cells were treated with thymidine for 24 hours and then released into fresh medium for the indicated times. Cell lysates were collected for Western blot analysis with indicated antibodies. E, SENP3 is dephosphorylated upon nocodazole release. U2OS cells were treated with nocodazole for 24 hours and then released into fresh medium for indicated times. Cell lysates were collected for Western blot with the indicated antibodies. F, Mutation of SENP3 phosphorylation sites. U2OS cells were first transfect with Flag–SENP3(WT) and Flag–SENP3(9A) for 36 hours, followed by nocodazole treatment. Cell lysates were collected for Western blot with indicated antibodies. p-SENP3, phosphorylated SENP3.
Dephoure and colleagues reported that SENP3 might be phosphorylated at nine Serine/Threonine residues (S169, T176, S181, S188, S212, T229, S232, S242, T353) in mitotic cells by mass spectrum analysis (Supplementary Fig. S1C; ref. 15). We mutated these Ser/Thr residues to alanine individually and observed no changes in shifting of SENP3 band in M-phase synchronized cells (Supplementary Fig. S1D). Shifting of SENP3 bands totally disappeared in the M-phase synchronized cells expressing the mutant containing all nine serine/threonine to alanine mutations (designated as SENP3 9A; Fig. 1F), indicating that these Ser/Thr residues on SENP3 are phosphorylated in the G2–M phases.
CDK1 and PP1α control mitotic SENP3 phosphorylation
Because cyclin B/CDK1 is a major kinase complex for mitotic protein phosphorylation and most of these nine phosphorylation sites in SENP3 are in typical sequences S/TP targeted by CDK1 (2), we speculated that CDK1 would be a kinase to phosphorylate SENP3. To test this possibility, cells were either treated with Ro-3306, a specific CDK1 inhibitor, or transfected with small interfering RNA for CDK1. SENP3 shifting was disappeared after Ro-3306 treatment or depletion of CDK1 in nocodazole-synchronized cells (Fig. 2A and B). We further carried out in vitro kinase assay and demonstrated that when cyclin B/CDK1 was incubated with purified Flag-SENP3, SENP3 wild type (SENP3 WT) but not SENP3 mutant (SENP3 9A), could be upshifted by phosphorylation, which was further confirmed by blotting with antiphospho-S/T antibody (Fig. 2C). These data suggest that CDK1 is a kinase for SENP3 phosphorylation in mitotic cells.
CDK1 and PP1α modulate phosphorylation of SENP3 in mitosis. A, Phosphorylation of SENP3 was abandoned by RO-3306 treatment. U2OS cells were treated with indicated reagents for 24 hours. Cell lysates were subjected to Western blot analysis with indicated antibodies. B, Phosphorylation of SENP3 was abandoned by depletion of CDK1. U2OS cells were transfected with control or CDK1 siRNA, and then cells were treated with nocodazole for 24 hours. Cell lysates were subjected to Western blot with indicated antibodies. C, The in vitro phosphorylation assay was performed with purified Flag-SENP3(WT), Flag-SENP3(9A), and recombinant CDK1/cyclin B1. Protein mixture was subjected to Western blot analysis with anti-Flag or antiphospho-Ser/Thr antibody. D and E, Dephosphorylation of SENP3 was postponed by either Okadaic acid treatment or depletion of PP1α. Synchronous U2OS cells were released into fresh medium with or without Okadaic acid for the indicated times (D). U2OS cells transfected with control siRNA or PP1α siRNA were synchronized by nocodazole, and then cells were released into fresh medium for indicated times (E). Cell lysates were subjected to Western blot analysis with indicated antibodies. F, The in vitro phosphorylation assay was performed with purified Flag-SENP3(WT), recombinant CDK1/cyclin B1, and recombinant PP1α. Protein mixture was subjected to Western blot analysis with anti-Flag antibody. p-SENP3, phosphorylated SENP3.
CDK1 and PP1α modulate phosphorylation of SENP3 in mitosis. A, Phosphorylation of SENP3 was abandoned by RO-3306 treatment. U2OS cells were treated with indicated reagents for 24 hours. Cell lysates were subjected to Western blot analysis with indicated antibodies. B, Phosphorylation of SENP3 was abandoned by depletion of CDK1. U2OS cells were transfected with control or CDK1 siRNA, and then cells were treated with nocodazole for 24 hours. Cell lysates were subjected to Western blot with indicated antibodies. C, The in vitro phosphorylation assay was performed with purified Flag-SENP3(WT), Flag-SENP3(9A), and recombinant CDK1/cyclin B1. Protein mixture was subjected to Western blot analysis with anti-Flag or antiphospho-Ser/Thr antibody. D and E, Dephosphorylation of SENP3 was postponed by either Okadaic acid treatment or depletion of PP1α. Synchronous U2OS cells were released into fresh medium with or without Okadaic acid for the indicated times (D). U2OS cells transfected with control siRNA or PP1α siRNA were synchronized by nocodazole, and then cells were released into fresh medium for indicated times (E). Cell lysates were subjected to Western blot analysis with indicated antibodies. F, The in vitro phosphorylation assay was performed with purified Flag-SENP3(WT), recombinant CDK1/cyclin B1, and recombinant PP1α. Protein mixture was subjected to Western blot analysis with anti-Flag antibody. p-SENP3, phosphorylated SENP3.
SENP3 is dephosphorylated in cells upon mitotic exit (Fig. 1E). We asked which phosphatase would contribute to SENP3 dephosphorylation at mitotic exit. Two classes of phosphatases, PP1 and PP2α, accounts for the majority of dephosphorylation of CDK1 substrates during mitosis. Thus, we utilized okadaic acid, an inhibitor of PP1 and PP2α, to test whether both phosphatases are involved in the dephosphorylation of SENP3 at mitotic exit. As shown in Figure 2D, phosphorylated SENP3 almost disappeared in control cells (DMSO treated) after 2 hours release from nocodazole arrest. However, addition of okadaic acid (200 nmol/L) significantly inhibited SENP3 dephosphorylation (Fig. 2D), suggesting either PP1 or PP2α, or both, were responsible for SENP3 dephosphorylation. We further determined that the lower concentration of okadaic acid (2 nmol/L) could not inhibit SENP3 dephosphorylation, indicating that PP1α but not PP2α could dephosphorylate SENP3 (Supplementary Fig. S2A), as PP2α, but not PP1α, is sensitive to inhibition by lower concentration of okadaic acid (16). To further confirm it, silencing endogenous PP1α could extend SENP3 phosphorylation to the post mitotic phase (Fig. 2E). We also carried out in vitro dephosphorylation assay to confirm PP1α as a SENP3 phosphatase (Fig. 2F). All together, we identify CDK1 as a kinase to catalyze mitotic SENP3 phosphorylation and PP1α as a phosphatase to dephosphorylate SENP3 at mitotic exit.
Mitotic phosphorylation suppresses SENP3 deSUMOylation function during mitosis
All nine phosphorylation sites are located in the noncatalytic N-terminal region of SENP3 (Supplementary Fig. S1C). We asked whether the phosphorylation could modulate SENP3 deSUMOylation function in mitosis. As most of SENP3 associated with chromosome, we isolated the chromosome-associated proteins from the cells expressing either SENP3 WT or SENP3 9A and detected a global SUMO2/3-conjugation levels in these cells (Fig. 3A; Supplementary Fig. S2B). There was no significant difference in the SUMOylation levels of chromosome-associated proteins between SENP3 WT and SENP3 9A cells without nocodazole treatment (Fig. 3A). However, SUMOylation of chromosome-associated proteins was markedly increased in cells expressing SENP3 WT, which was highly-phosphorylated at M phase, but not in SENP3 9A-expressing cells, suggesting that phosphorylation would suppress SENP3 deSUMOylation of chromosome-associated proteins in mitosis (Fig 3A). We further showed that inhibition of SENP3 phosphorylation by depletion of CDK1 or Ro-3306 decreased the SUMOylation level of chromosome-associated proteins in mitotic cells (Fig. 3B; Supplementary Fig. S2C). Moreover, depletion of PP1α or okadaic acid treatment delayed the deSUMOylation of chromosome-associated proteins in the cells released from M phase arrest (Fig. 3C; Supplementary Fig. S2D). These results indicate that modulation of mitotic SENP3 phosphorylation would affect its deSUMOylation function on chromosome-associated proteins.
Phosphorylation suppresses SENP3 deSUMOylation function. A, SENP3 9A showed much stronger deSUMOylation activity than SENP3 WT did. U2OS cells were transfected with Flag-SENP(WT) or Flag-SENP3(9A) for 24 hours, and then cells were treated with or without nocodazole for another 24 hours. Cells were fractionated and chromatin was collected for Western blot analysis with indicated antibodies. B, U2OS cells were transfected with control siRNA (NC), or siRNA for CDK1; 24 hours later, cells were treated with or without nocodazole for another 24 hours. Whole cell lysates (WCL) or chromatin was collected for Western blot analysis with indicated antibodies. C, U2OS cells were transfected with control siRNA (NC), or siRNA for PP1α; 24 hours later, cells were treated with nocodazole for another 24 hours. Cells were then released into fresh medium for indicated time. Whole cell lysates or chromatin was collected for Western blot analysis with indicated antibodies. D, TopoIIα was deSUMOylated by SENP3 at G2–M phase. U2OS cells were transfected with His-SUMO2 and Flag-SENP(WT) or Flag-SENP3(9A) for 24 hours, and then cells were treated with or without nocodazole for 24 hours. Cell lyates were collected for the Talon beads pull-down. Input and precipitates were blotted with indicated antibodies. E, DeSUMOylation of TopoIIα was impaired by Okadaic acid treatment. Nocodazole-synchronized U2OS cells transfected with His-SUMO2 and Flag-SENP3 (WT) were released into fresh medium with Okadaic acid or not for the indicated times. Cell lysates were collected for Talon beads pull down. Input and precipitates were blotted with indicated antibodies. F, TopoIIα interacted with SENP3 9A at the G2–M phase. Cell lysates from nocodazole-synchronized U2OS cells transfected with Flag-SENP(WT) or Flag-SENP3(9A) were collected and subjected to immunoprecipitation and Western blot analysis with indicated antibodies. G, U2OS cells transfected with Flag-SENP(WT) or Flag-SENP3(9A) were treated with indicated reagents and then cells were collected and subjected to immunoprecipitation and Western blot analysis with indicated antibodies. H, U2OS cells transfected with Flag-SENP(WT) or Flag-SENP3(9A), or Flag-SENP3(9E) were treated with or without nocodazole and then cells were collected and subjected to immunoprecipitation and Western blot analysis with indicated antibodies.
Phosphorylation suppresses SENP3 deSUMOylation function. A, SENP3 9A showed much stronger deSUMOylation activity than SENP3 WT did. U2OS cells were transfected with Flag-SENP(WT) or Flag-SENP3(9A) for 24 hours, and then cells were treated with or without nocodazole for another 24 hours. Cells were fractionated and chromatin was collected for Western blot analysis with indicated antibodies. B, U2OS cells were transfected with control siRNA (NC), or siRNA for CDK1; 24 hours later, cells were treated with or without nocodazole for another 24 hours. Whole cell lysates (WCL) or chromatin was collected for Western blot analysis with indicated antibodies. C, U2OS cells were transfected with control siRNA (NC), or siRNA for PP1α; 24 hours later, cells were treated with nocodazole for another 24 hours. Cells were then released into fresh medium for indicated time. Whole cell lysates or chromatin was collected for Western blot analysis with indicated antibodies. D, TopoIIα was deSUMOylated by SENP3 at G2–M phase. U2OS cells were transfected with His-SUMO2 and Flag-SENP(WT) or Flag-SENP3(9A) for 24 hours, and then cells were treated with or without nocodazole for 24 hours. Cell lyates were collected for the Talon beads pull-down. Input and precipitates were blotted with indicated antibodies. E, DeSUMOylation of TopoIIα was impaired by Okadaic acid treatment. Nocodazole-synchronized U2OS cells transfected with His-SUMO2 and Flag-SENP3 (WT) were released into fresh medium with Okadaic acid or not for the indicated times. Cell lysates were collected for Talon beads pull down. Input and precipitates were blotted with indicated antibodies. F, TopoIIα interacted with SENP3 9A at the G2–M phase. Cell lysates from nocodazole-synchronized U2OS cells transfected with Flag-SENP(WT) or Flag-SENP3(9A) were collected and subjected to immunoprecipitation and Western blot analysis with indicated antibodies. G, U2OS cells transfected with Flag-SENP(WT) or Flag-SENP3(9A) were treated with indicated reagents and then cells were collected and subjected to immunoprecipitation and Western blot analysis with indicated antibodies. H, U2OS cells transfected with Flag-SENP(WT) or Flag-SENP3(9A), or Flag-SENP3(9E) were treated with or without nocodazole and then cells were collected and subjected to immunoprecipitation and Western blot analysis with indicated antibodies.
TopoIIα has been shown to be modified by SUMO2/3 during mitosis (12, 13). We also detected SUMO2-conjugated TopoIIα in mitosis-arrested cells (Fig. 3D). SUMOylation level of TopoIIα was markedly reduced in cells expressing SENP3 9A mutant compared with that in cells expressing SENP3 WT during mitotic arrest (Fig. 3D). Okadaic acid inhibited dephosphorylation of SENP3 as well as SENP3 deSUMOylation of TopoIIα in cells released from nocodazole arrest (Fig. 3E). These data suggest that mitotic phosphorylation may suppress SENP3 deSUMOylation of TopoIIα. It is worth to note that TopoIIα only bound with SENP3 9A mutant but not with phosphorylated SENP3 WT protein in mitotic cells (Fig. 3F). Moreover, Ro-3306 treatment reduced SENP3 phosphorylation and increased its interaction with TopoIIα (Fig. 3G). We did in vitro assay to show that phosphorylation inhibited SENP3 interaction with recombinant TopoIIα (Supplementary Fig. S3A). SENP3 9E, which mimics constant phosphorylation of SENP3, could not bind with TopoIIα (Fig. 3H). Put together, these results indicate that mitotic phosphorylation of SENP3 could block SENP3 binding to TopoIIα, which might be a reason for mitotic phosphorylation could suppress SENP3 deSUMOylation of TopoIIα.
Mitotic SENP3 phosphorylation is essential for resolution of sister centromeres
It has been shown that SUMOylation of chromosome-associated proteins regulates the progression of mitosis (4, 5, 12). We thus speculated that mitotic SENP3 phosphorylation would modulate progression of mitotic cell cycle through deSUMOylation. Cell-cycle analysis showed that SENP3 9A cells were temporarily arrested at the G2–M phase when compared with SENP3 WT cells (Fig. 4A; Supplementary Fig. S4A). We used Mad2, an essential component of the mitotic checkpoint, as an indicator of mitotic checkpoint activation (17–19). We observed that the mitosis-caused higher level of Mad2 was maintained for longer time in SENP3 9A cells than that in SENP3 WT cells after cells were released from nocodazole arrest, indicating that expression of SENP3 9A could induce the mitotic checkpoint and cell-cycle arrest at the metaphase-anaphase transition (Fig. 4B). We showed that overexpression of TopoIIα in SENP3 9A cells reduced the cells arrested in the G2–M phase by 50% compared to the cells without TopoIIα overexpression after nocodazole-arrested cells were released for 90 minutes, suggesting that TopoIIα could contribute to the cell-cycle defects caused by SENP3 9A (Supplementary Fig. S4B). Because mitotic checkpoint is related with abnormal chromosome segregation (1, 11), we further analyzed the chromosome segregation in SENP3 9A cells. We observed much more lagging chromosomes in SENP3 9A mitotic cells (36%) than that in SENP3 WT cells (9%; Fig. 4C; Supplementary Fig. S4C). Furthermore, an increase in centrophilic chromosomes (chromosomes that lie near the centrosomes and do not migrate straightaway to the equator) and anaphase bridges was also observed in SENP3 9A mitotic cells compared with that in SENP3 WT cells (Fig. 4C). These observations were similar to the phenotypes shown in cells with TopoIIα SUMOylation deficiency (12), suggesting that the defects in chromosome segregation in SENP3 9A cells might be related to SENP3 9A deSUMOylation of TopoIIα in mitosis.
Phosphorylation of SENP3 is essential for resolution of sister chromosome. A, Cell-cycle profiles of control EV cells, SENP3 (WT) cells, or SENP3 (9A) cells. Nocodazole-synchronized cells were released into fresh medium for indicated times. Cells at each time point were subjected to both flow cytometry analysis (top) and Western blot analysis with indicated antibodies (bottom). B, Spindle assembly checkpoint was activated in SENP3 (9A) cells. Chromatin pellet of above SENP3 (WT) or SENP3 (9A) cells at various time points (A) were subjected to Western blot analysis with indicated antibodies. C, Chromosome segregation defects in SENP3 (9A) cells. Chromosome of mitotic cells expressing GFP-H2B and SENP3 (WT) or SENP3 (9A) was visualized by microscopy. Percentages of cells with different defects are shown on the right. One hundred mitotic cells in WT or 9A cells were analyzed. Results are the mean ±SD of three independent experiments. D, Ectopic expression of SENP3 (9A) increased colony formation. Clone formation of SENP3 (WT) cell and SENP3 (9A) cells were determined by soft agar colony formation assay. A total of 500 cells was seeded into each well and cultured for another 2 weeks. Cells were stained with crystal violet and counted. E, SENP3 (9A) cell was prone to tumorigenesis on nude mice. Each nude mouse was subcutaneously injected with 5 × 106 cells stably expressing SENP3 (WT) or SENP3 (9A) for 3 weeks. n = 10 for nude mice injected with SENP3 (WT) or SENP3 (9A) cells, separately. **, P < 0.01.
Phosphorylation of SENP3 is essential for resolution of sister chromosome. A, Cell-cycle profiles of control EV cells, SENP3 (WT) cells, or SENP3 (9A) cells. Nocodazole-synchronized cells were released into fresh medium for indicated times. Cells at each time point were subjected to both flow cytometry analysis (top) and Western blot analysis with indicated antibodies (bottom). B, Spindle assembly checkpoint was activated in SENP3 (9A) cells. Chromatin pellet of above SENP3 (WT) or SENP3 (9A) cells at various time points (A) were subjected to Western blot analysis with indicated antibodies. C, Chromosome segregation defects in SENP3 (9A) cells. Chromosome of mitotic cells expressing GFP-H2B and SENP3 (WT) or SENP3 (9A) was visualized by microscopy. Percentages of cells with different defects are shown on the right. One hundred mitotic cells in WT or 9A cells were analyzed. Results are the mean ±SD of three independent experiments. D, Ectopic expression of SENP3 (9A) increased colony formation. Clone formation of SENP3 (WT) cell and SENP3 (9A) cells were determined by soft agar colony formation assay. A total of 500 cells was seeded into each well and cultured for another 2 weeks. Cells were stained with crystal violet and counted. E, SENP3 (9A) cell was prone to tumorigenesis on nude mice. Each nude mouse was subcutaneously injected with 5 × 106 cells stably expressing SENP3 (WT) or SENP3 (9A) for 3 weeks. n = 10 for nude mice injected with SENP3 (WT) or SENP3 (9A) cells, separately. **, P < 0.01.
We further detected slower cell growth and more apoptosis in SENP3 9A cells than that in SENP3 WT cells, which was most probably due to mitotic catastrophe (Supplementary Fig. S4D and S4E). However, soft agar assay in U2OS cells showed that expression of SENP3 9A significantly enhanced the cellular anchorage-independent growth (Fig. 4D). Nude mice experiments further confirmed that expression of SENP3 9A markedly promoted U2OS cells tumor growth (Fig. 4E). These data revealed that mitotic SENP3 phosphorylation play a role in control of chromosome stability and tumorigenesis.
Discussion
In this study, we find that mitotic SENP3 phosphorylation play a crucial regulation role in mitosis of the cell cycle. Importantly, the mitotic phosphorylation suppresses SENP3 deSUMOylation of chromosome-associated proteins including TopoIIα during mitosis. Phosphorylation deficiency increases SENP3 deSUMOylation function and leads to mitotic arrest and chromosome instability in mitosis, which are hallmarks of tumorigenesis.
We demonstrate SENP3 as a new mitotic phosphorylated protein catalyzed by CDK1 and PP1α. CDK1-mediated SENP3 phosphorylation turns off its deSUMOylation function and allows chromosome-associated proteins to be SUMOylated, which is essential for progression through mitosis. PP1α then dephosphorylates SENP3 at the mitotic exit to turn on its function in other phases of the cell cycle.
Many centromeric proteins are SUMOylated upon entry into mitosis, and SUMOylation of these proteins is important for chromosome condensation and segregation (5, 7, 12). TopoIIα is important for chromosome decatenation and is critical for sister chromosome segregation before onset of anaphase (20). Inhibition of TopoIIα decatenation activity can lead to abnormal chromosome segregation at mitosis, which eventually leads to chromosome instability and appearance of more aneuploid cells, which is a hallmark of cancer (20). Furthermore, TopoIIα has been shown as a mitotic SUMOylated protein and SUMOylation is essential for TopoIIα localization at centromere for decatenation of chromosome in mitosis (12, 13). Deregulation of TopoIIα SUMOylation can lead to abnormal mitotic chromosome segregation and chromosome instability (12). Our study identified TopoIIα as a target for SENP3 deSUMOylation. In mitosis, mitotic SENP3 phosphorylation suppresses its deSUMOylation of TopoIIα to allow TopoIIα in SUMOylated status for decatenation of chromosome. If mitotic SENP3 phosphorylation is deficiency, deSUMOylation of TopoIIα by unphosphorylated SENP3 would cause abnormal mitotic chromosome segregation and chromosome instability.
We mapped nine mitotic phosphorylation sites on SENP3 protein. Interestingly, all nine sites are located in the noncatalytic N-terminal of SENP3. How the phosphorylation in N-terminal could affect the catalytic activity of SENP3 C-terminal? We find that TopoIIα prefers to bind to unphosphorylated SENP3 or SENP3 phosphorylation mutant, indicating that phosphorylation would reduce SENP3 interaction with its substrates TopoIIα. However, we do not know whether this mechanism could be used to explain how SENP3 phosphorylation modulates other targets except TopoIIα in mitosis. Considering all 9 phosphorylation sites are widely distributed throughout the N-terminal region of SENP3, we propose that each phosphorylation site of SENP3 might act as a specific signaling in regulation of a specific substrate-engaged activity. It is worth to be determined in future study.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Authors' Contributions
Conception and design: B. Wei, C. Huang, J. Cheng
Development of methodology: B. Wei, C. Huang, B. Liu, Y. Wang, N. Xia
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): B. Wei, C. Huang, B. Liu, J. Cheng
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): B. Wei, C. Huang, B. Liu
Writing, review, and/or revision of the manuscript: B. Wei, C. Huang, G.-Q. Chen, J. Cheng
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): B. Wei, C. Huang, Y. Wang, N. Xia, Q. Fan, J. Cheng
Study supervision: C. Huang, J. Cheng
Acknowledgments
This work was supported by grants from the National Natural Science Foundation of China (81430069 to J. Cheng; 81721004 to G.Q. Chen).