SENP1 Is a Crucial Regulator for Cell Senescence through DeSUMOylation of Bmi1

Cell senescence can limit proliferative potential and prevent tumorigenesis. Bmi1 is a key regulator in cell senescence by suppressing the Ink4a/Arf locus. However, how to regulate Bmi1 activity in cell senescence is largely unknown. Here, we show that SENP1 plays an important role in cell senescence by regulating Bmi1 SUMOylation. Senp1−/− primary MEF cells show resistance to cell senescence induced by passaging or other senescence inducing signals. SENP1 deficiency also reduces oncogene H-RasV12-induced senescence, and enhances H-RasV12-induced cell transformation. We further show that in Senp1−/− MEFs the expression of p19Arf, an important regulator in p53/p21-mediated cell senescence, is markedly reduced. Meanwhile, we demonstrate that SENP1 can specifically de-SUMOylate Bmi1 and thereby decreases the occupancy of Bmi1 on p19Arf promoter leading to decrease of H2AK119 mono-ubiquitination and up-expression of p19Arf. These data reveal a crucial role of SENP1 in regulation of cell senescence as well as cell transformation.


SENP1 deficiency reduces the senescence of MEF cells. Our previous study has demonstrated that
Senp1 −/− MEF cells were easily immortalized by using a 3T3 protocol compared to wild-type (WT) cells, indicating that SENP1 deficiency might keep MEFs away from cell senescence. To test it, we freshly isolated MEF cells from Senp1 +/+ and Senp1 −/− embryos at E12.5 and cultured them by using a 3T3 protocol. As shown in Fig. 1a, Senp1 +/+ MEF cells entered growth plateau phase at 6 to 8 passages, whereas Senp1 −/− MEF cells proliferated steadily up to 20 passages. Consistent with this observation, Senp1 −/− MEF cells exhibited a higher BrdU incorporation rate than Senp1 +/+ MEF cells did (Fig. 1b), indicating higher proliferation rate in Senp1 −/− MEF cells. We further stained SA-β -gal and found that SA-β -gal positive cells were markedly less in Senp1 −/− but not in Senp2 −/− MEF cells when compared with control WT MEF cells (Fig. 1c). Furthermore, we determined whether SENP1 deficiency could affect the senescence induced by serum deprivation or oxidative stress. Senp1 +/+ and Senp1 −/− MEF cells at passage 5 were cultured in medium containing 50 μ M H 2 O 2 or 3% serum. Both conditions reduced the proliferation of Senp1 +/+ MEFs but not Senp1 −/− cells (Fig. 1d). We also confirmed that SENP1 deficiency could decrease the senescence in both conditions (Fig. 1e).

Downregulation of p19 Arf in Senp1 −/− MEF cells.
To understand the mechanism underlying SENP1 regulating senescence, we analysed the activity of the two major senescence-related signal pathways p19 Arf /p53/ p21 and p16 Ink4a /Rb in Senp1 +/+ and Senp1 −/− MEF cells. As shown in Fig. 3a, the expressions of p19 Arf , p53, and p21 proteins were markedly decreased in Senp1 −/− MEF cells when compared to Senp1 +/+ cells. However, p16 Ink4a and Rb proteins showed only mild reduction in Senp1 −/− MEF cells in comparison to Senp1 +/+ cells. We also detected the expressions of p19 Arf and p16 Ink4a in H-Ras V12 -transduced Senp1 +/+ or Senp1 −/− MEF cells and found that H-Ras V12 significantly induced p19 Arf but not p16 Ink4a expression in both cells (Fig. 3b). The expression of p19 Arf was less increased in H-Ras V12 -transduced Senp1 −/− MEFs when compared to Senp1 +/+ cells. We further transfected SENP1 wild-type (SENP1 WT) or SENP1 catalytic mutant (SENP1m) in Senp1 −/− MEF cells and found that re-expression of SENP1 WT but not SENP1m could rescue p19 Arf expression in Senp1 −/− MEF cells (Fig. 3c). To further confirm the role of SENP1 in regulating p19 Arf expression, we constructed p19 Arf promoter-driven luciferase reporter gene and showed that SENP1 WT but not SENP1m could induce p19 Arf transcription (Fig. 3d). Taken together, these data reveal that SENP1 promotes p19 Arf expression. SENP1 promotes p19 Arf expression via de-SUMOylation of Bmi1. Bmi1 is a critical negative regulator for p19 Arf and p16 Ink4a expression 23,29,30 . It has also been reported as a SUMOylated protein 27 . Therefore, we postulated that SENP1 might regulate p19 Arf expression via de-SUMOylation of Bmi1. To test whether SENP1 directly de-SUMOylates Bmi1, Flag-Bmi1 and HA-SUMO1 were co-expressed with RGS-tagged SENP1 or SENP1m in 293T cell lines. As a result, the co-expression of SENP1, not SENP1m, could de-conjugate SUMOylated Bmi1, while the SENP1m increase the SUMOylated Bmi1 (Fig. 4a,b) through a dominant negative activity to block the de-SUMOylation activity of endogenous SENP1 31 . More importantly, we observed that the SUMOylated Bmi1 proteins were accumulated in Senp1 −/− MEF cells (Fig. 4c). To exclude the possibility that Bmi1 expression was affected by SENP1, we checked the expression of Bmi1, as well as the expressions of other PRC1 components CBX4 and Ring1B, and found no changes in Senp1 −/− MEF cells ( Figure S2A). These results indicate that SENP1 is a specific protease to de-SUMOylate Bmi1.
We further assessed whether SENP1 could affect Bmi1 ubiquitin ligase activity in mono-ubiquitination of histone 2A at K119. As shown in Fig. 4d, mono-ubiquitination of H2AK119 was increased in Senp1 −/− MEFs when compared to Senp1 +/+ cells, suggesting that SENP1 mediated de-SUMOylation decreases Bmi1 ubiquitin ligase activity. We also measured the suppressive activity of Bmi1 WT and Bmi1 SUMOylation mutant (K88R) using p19 Arf promoter-driven luciferase assay. Bmi1 WT significantly reduced p19 Arf transcription, whereas Bmi1 K88R mutant showed a weaker repressive activity than Bmi1 WT (Fig. 4e). Meanwhile we found that SENP1, but not SENP1m could counteract the suppressive activity of Bmi1 WT on p19 Arf transcription (Fig. 4f).
We reasoned that SUMOylation might promote Bmi1 binding to p19 Arf promoter. To test it, we used chromatin immunoprecipitation (ChIP) assay to demonstrate that the occupancy of Bmi1 WT on p19 Arf promoter was much higher than that of Bmi1 K88R mutant (Fig. 4g). We also illustrated that Bmi1 occupancy on p19 Arf promoter was more profound in Senp1 −/− MEFs than that in WT cells, and that the ubiquitination of H2A at K119 Scientific RepoRts | 6:34099 | DOI: 10.1038/srep34099 of p19 Arf promoter was also increased in Senp1 −/− MEFs when compared to WT cells (Fig. 4h). Taken together, these data suggest that SENP1 promotes p19 Arf expression via de-SUMOylation of Bmi1.  However, more SA-β -gal positive cells were observed in Bmi1 K88R-transduced cells than that in Bmi1 WT control. We further found that both Bmi1 WT and Bmi1 K88R expressions increased colony formation in H-Ras V12 -Senp1 −/− MEF cells compared to vector control. Less colony were found in Bmi1 K88R-transduced H-Ras V12 -Senp1 −/− MEF cells in comparison to Bmi-1 WT control (Fig. 5b). These data suggest that SUMOylation is essential for Bmi1 repression of cell senescence.

SENP1 expression is associated with cell senescence in prostate PIN lesion. Previously, we have
shown that SENP1 was overexpressed in prostrate intraepithelial neoplasia (PIN) lesion 32 and tumor cells 33,34 . Overexpression of SENP1 in mouse prostate induced PIN lesions, but not tumor formation 33 . We speculated that overexpression of SENP1 could promote the growth of prostate epithelial cell as well as cell senescence in PIN lesion. We thus collected 44 human prostate PIN samples and analysed the expression of SENP1, senescence markers heterochromatin proteins 1 γ (HP1γ ) 35,36 , p16 INK4a , p14 ARF and Bmi1 by immunohistochemistry staining in these samples. A score indicating SENP1, HP1γ , p16 INK4a , p14 ARF or Bmi1 was given ranging from 0 to 3 based on the percentage of the stained area and immunostaining intensity in these samples. As shown, SENP1 expression was positively related to HP1γ (rs = 0.553, P < 0.0001), p14 ARF (rs = 0.6578, P < 0.0001), and p16 INK4a (rs = 0.6755, P < 0.0001) expressions in human PIN samples (Fig. 6a,b) ( Figure S3a,b). However, it was not related to Bmi1 expression (rs = − 1501, P = 3309) ( Figure S3a,b). These data indicate that SENP1 expression is associated with cell senescence in prostate PIN lesion.

Discussion
In this study, we demonstrate a role of SENP1 in cell senescence. SENP1 de-SUMOylates Bmi1 and reduces the occupancy of Bmi1 on p19 Arf promoter, leading to p19 Arf expression and cell senescence (Fig. 6c). SUMOylated Bmi1 was accumulated in Senp1 −/− MEFs (Fig. 4c) and SUMOylation promoted Bmi1 occupancy on p19 Arf promoter to suppress p19 Arf expression (Fig. 4f-h). Interestingly, Bmi1 −/− cells showed increased Ink4a/Arf expression, especially p16 Ink4a expression 23 , whereas p16 Ink4a expression is only mildly decreased in Senp1 −/− MEFs compared to p19 Arf . These results suggest that SUMOylation might selectively modulate Bmi1 effect on p19 Arf by promoting its binding to the promoter locus, although currently it is unknown how SUMOylation modulates Bmi1 selection.  We showed the deficiency of SENP1 delays MEF cell senescence, which seems to contradict with the previously report on SENP1 represses cell senescence in human fibroblasts 37 . Yates et al. showed that shSENP1 in human fibroblasts induced senescence. They further showed that shSENP1 increased p53 expression via unknown mechanism. In the present study, we find that SENP1 deficiency reduces p19 Arf /p16 Ink4a /p53 expression via Bmi1. As mouse cells have very long telomeres (40-60 kb) when compared to human cells (5-15 kb) 38 , we currently don't know whether this difference would contribute to their different responses to senescence. We also note another difference in the approaches that we have utilized for SENP1 deficiency from Yates et al. 37 . We deleted Senp1 gene in genome of Senp1 −/− MEF cells, whereas SENP1 was silenced at mRNA level by shSENP1in HFF cells, which might cause off target effect.
The SUMOylated Bmi1 was first reported to accumulate at the DNA damage sites 27 . Our data showed the SUMOylation increased Bmi1 binding to DNA. It is possible the increased Bmi1 may recruit some DNA repairing proteins to repair oncogene induced DNA damage and to reduce the γ H2Ax foci. Bmi1 is proposed to bind and increase Ring1B's E3 ligase activity 39 . Bmi1/Ring1b, an autoinhibited RING E3 ubiquitin ligase, was also reported to promote SENP1 ubiquitination and degradation 40 , suggesting that this regulation might be a negative feedback mechanism to control SENP1 action on Bmi1, which would promote cell to be transformed during tumorigenesis. However, we don't know whether SUMOylated Bmi1 further increases the E3 ligase activity of Ring1B and thereby affects ubiquitination of SENP1.
It is well-known that ROS-induced DNA damage can increase p53 activation, which in turn leads to cellular senescence 38,[41][42][43] . Previously, we have shown that ROS production decreases in Senp1 −/− MEF cells because of the defect in mitochondrial biogenesis 44 . Thus, the lower ROS production might partially contribute to the reduction of senescence in Senp1 −/− MEF cells. It is evident that we observe a decrease in DNA damage marker γ H2AX foci in Senp1 −/− MEF cells.
We previously found SENP1 was overexpressed in PIN lesion 32 and tumor cells 33,34 . We also showed that overexpression of SENP1 in mouse prostate induced PIN lesions, but no tumor formation in the transgenic model 33 . We show herein that SENP1 expression is associated with the senescence in PIN lesion. We expect to see if bypassing senescence would promote PIN to cancer in the SENP1 prostate transgenic model. This is supported by our recent observations that tumor suppressor PTEN deficiency promote cancer development in SENP1 prostate transgenic mice partially through the decrease of p53 expression and cell senescence (unpublished data), suggesting the role of SENP1-mediated senescence in tumorigenesis, which need to be further explored in future.

Material and Methods
Mouse embryo fibroblasts isolation and cell culture. All animal procedures were approved by the Institutional Animal Care and Use Committee of Shanghai Jiao Tong University School of Medicine, which were carried out in accordance with the guidelines (IACUC No. 2013027). The generation of Senp1 +/+ or Senp1 −/− MEF cells was previously described 45 . MEF cells were cultured in completed DMEM containing 10% FBS at 37 °C under 5% CO 2 . For replicative senescence assay, cells were passaged following a 3T3 protocol 46,47 . Cells were split every 3 days and replanted at a density of 1 × 10 6 cells per 10 cm dish for 20 passages. For colony forming assay, MEF cells were seeded at a density of 5 × 10 4 cells per 10 cm dish and cultured for 2 weeks. Then cells were washed with phosphate-buffered saline, fixed in 70% ethanol for 15 min and stained with crystal violet. Colonies were observed under microscope. The cell mass with more than 50 cells was counted as a colony.
Senescence associated β-galactosidase staining. MEF cells were plated at a density of 1 × 10 5 cells in 35 mm dishes. Cells were fixed and stained according to the instructions supplied by the senescence β -galactosidase staining kit (Beyotime biotechnology, China). Briefly, cells were washed with PBS and fixed in SA-β -Gal fixing solution for 15 min at room temperature. Cells then were washed 3 times with PBS and stained with working solution (10 μ l buffer A, 10 μ l buffer B, 930 μ l buffer C, and 50 μ l X-Gal solution) overnight at 37 °C. Cells were counterstained with nuclear fast red. The population of SA-β -Gal positive cells was determined by counting 400 cells in at least 5 fields per dish and images were taken using a phase-contrast microscope at 400× analyzed in MEFs transfected with p19 Arf luciferase plasmid, Flag-Bmi1 WT or K88R plasmids. Data represent the mean ± S.E.M. of three independent experiments (*P < 0.05, Student's t-test). (f) P19 Arf transcription was analyzed in MEFs transfected with p19 Arf luciferase plasmid, Flag-Bmi1 WT or K88R, plus SENP1 or SENP1m plasmid. Data represent the mean ± S.E.M. of three independent experiments (*P < 0.05, Student's t-test; n.s., non-significant). (g) Senp1 −/− MEFs were transfected with Flag-Bmi1 WT, Flag-Bmi1 K88R, or vector as indicated. Chromatins prepared from these cells were subjected to IP with anti-Flag beads followed by Real-time PCR using primers for p19 Arf promoter. Data represent the mean ± S.E.M. of three independent experiments (*P < 0.05, Student's t-test). (h) Chromatins from Senp1 +/+ and Senp1 −/− MEFs were subjected to IP with anti-Bmi1, -H2AK119ub or IgG followed by Real-time PCR using primers for p19 Arf promoter. Data represents the mean ± S.E.M. of three independent experiments (*P < 0.05, Student's t-test).
Scientific RepoRts | 6:34099 | DOI: 10.1038/srep34099 magnification (Olympus, Japan). The proportions of cells positive for SA-β -Gal activity are shown as the percentage of the total number of cells counted in each dish.
BrdU incorporation assay. Sub-confluent Senp1 +/+ and Senp1 −/− MEF cells at passage 5 (P5) and passage 12 (P12) were labelled for 4 h with 10 μ M 5-bromo-2′ -deoxyuridine (BrdU; Amersham). Labelled MEF cells were planted on cover slips and fixed in 5% acetic acid and 95% ethanol for 15 min at − 20 °C. After fixation, cells were washed with PBS and treated with 0.1 mg/ml RNase A for 30 min at 37 °C. After washed with PBS, cells were washed with 2 N HCl, 0.5% Triton X-100 for 30 min at room temperature. Thereafter the cells were washed with PBS and incubated with BrdU antibody and FITC conjugated secondary antibody for 1 h. After washing with PBS, cells were counterstained with DAPI and analyzed under fluorescence microscope.
Immunoprecipitation and immunoblotting. Cells were washed once and lysed in the presence of 10 mM N-ethylmaleimide (NEM) using ice-cold RIPA buffer (50 mM Tris-HCl (pH 7.4), 400 mM NaCl, 1% Triton X-100, 0.1% SDS, 1 mM PMSF, and protease inhibitors) for 30 min on ice. Cell lysates were centrifugated at 14,000 × g for 10 min at 4 °C. The supernatants were collected and incubated with appropriate antibodies overnight at 4 °C, followed by incubation with protein A/G-Sepharose beads (Amersham Biosciences) for another 2 h at 4 °C. After washing three times with RIPA buffer, the immunoprecipitates were eluted with Laemmli sample buffer, elute boiled, and subjected to western blot analysis. Immunofluorescent staining. Paraformaldehyde-fixed cells were treated with 0.3% Triton X-100 in PBS for 15 min. After washing with PBS, cells were blocked with 1% BSA in PBS for 1 h, and then incubated with anti-γ H2AX(1:200) antibody overnight at 4 °C, followed by incubation with APC conjugated secondary antibody(1:200) for 1 h. Cells were then counterstained with DAPI. Samples were visualized with an Olympus fluorescence microscopy.
Chromatin immunoprecipitation (ChIP) assays. Formaldehyde-crosslinked chromatin was prepared from MEF cells and immunoprecipitations were performed using the ChIP Assay Kit according to the manufacturer's recommended protocol (Upstate Biotechnology). Antibodies used for ChIP assays were anti-Flag-M2, anti-Bmi1, anti-H2AK119ub or IgG. A pair of real-time PCR primers was used for amplification of the promoter segments of the mouse p19 Arf gene as previously 49 : forward, 5′ -AAAACCCTCTCTTGGAGTGGG-3′ ; reverse, 5′ -GCAGGTTCTTGGTCACTGTGAG-3′ . Luciferase assays. P19 Arf promoter containing 1202 nucleotides upstream the ATG start codon was cloned into the pGL3-basic luciferase reporter vector (Promega) using standard cloning procedures. For reporter assays, MEF cells were transfected with X-tremeGENE HP DNA Transfection Reagent (Roche Applied Science). 36 h after transfection, the luciferase was assayed using the Dual-Luciferase reporter assay system (Promega). Luciferase activity of the reporter construct was normalized on the basis of the Renilla luciferase activity.
Statistical analyses. All data are represented by the mean ± S.E.M. of at least three independent experiments. GraphPad Prim 5 (GraphPad Software, La Jolla, CA) was used for statistical analysis. Data was analysed using two-tailed student's t-test. A p value of < 0.05 was considered significant. Immunohistochemistry was performed to evaluate SENP1, p14 ARF and HP1γ expression on PIN TMA slides from patients. SENP1, p14 ARF and HP1γ protein level in these samples was scored from 0 to 3 based on the stained area percentage and immunostaining intensity. Spearman correlation coefficient was performed to evaluate the association of SENP1 with p14 ARF or HP1γ .
(c) Working model shows the mechanism of SENP1 in regulating cell senescence.