Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

SUMOylation of hnRNP‐K is required for p53‐mediated cell‐cycle arrest in response to DNA damage

SUMOylation of hnRNP‐K is required for p53‐mediated cell‐cycle arrest in response to DNA damage The EMBO Journal (2012) 31, 4441–4452 & 2012 European Molecular Biology Organization All Rights Reserved 0261-4189/12 | | THE THE www.embojournal.org EMB EMB EMBO O O JO JOU URN R NAL AL SUMOylation of hnRNP-K is required for p53-mediated cell-cycle arrest in response to DNA damage Seong Won Lee, Moon Hee Lee, downstream targets, such as the p53 transcription factor and the checkpoint CHK1 and CHK2 kinases (Abraham, 2001, Jong Ho Park, Sung Hwan Kang, 2004; Ciccia and Elledge, 2010). This process in turn regulates Hee Min Yoo, Seung Hyun Ka, the functions of downstream effector proteins involved in Young Mi Oh, Young Joo Jeon* cell-cycle arrest, DNA repair, and/or apoptosis. A key and Chin Ha Chung* example is ATM- and ATR-mediated phosphorylation of Department of Biological Sciences, College of Natural Sciences, Seoul both p53 and HDM2, which impairs their interaction and National University, Seoul, Korea thereby prevents HDM2-mediated ubiquitination of p53 for degradation by proteasome, leading to stabilization and activation of p53 (Perry, 2004). Heterogeneous ribonucleoprotein-K (hnRNP-K) is nor- A major consequence of p53 activation in response to DNA mally ubiquitinated by HDM2 for proteasome-mediated damage is the induction of cell-cycle arrest (Vogelstein et al, degradation. Under DNA-damage conditions, hnRNP-K is 2000; Bartek and Lukas, 2001; Vousden and Lu, 2002; Horn transiently stabilized and serves as a transcriptional co- and Vousden, 2007) at the G1/S or G2/M phase. Cell-cycle activator of p53 for cell-cycle arrest. However, how the arrest at the G1/M phase is primarily achieved by expression stability and function of hnRNP-K is regulated remained of p53-downstream genes, such as p21, an inhibitor of cyclin- unknown. Here, we demonstrated that UV-induced dependent kinases (CDKs). Notably, p21 also acts as an anti- SUMOylation of hnRNP-K prevents its ubiquitination for apoptotic protein. This function of p21 is mediated by its stabilization. Using SUMOylation-defective mutant and ability to inhibit caspase-3 (Suzuki et al, 1998), stabilize the purified SUMOylated hnRNP-K, SUMOylation was shown anti-apoptotic cIAP1 (Steinman and Johnson, 2000), or to reduce hnRNP-K’s affinity to HDM2 with an increase in downregulate caspase-2 (Baptiste-Okoh et al, 2008). Thus, that to p53 for p21-mediated cell-cycle arrest. PIAS3 served p21 plays an important role in inhibiting apoptosis as well as as a small ubiquitin-related modifier (SUMO) E3 ligase for in cell-cycle arrest, allowing cells to repair damaged DNA and hnRNP-K in an ATR-dependent manner. During later per- prevent tumorigenesis. iods after UV exposure, however, SENP2 removed SUMO Small ubiquitin-related modifier (SUMO) is an ubiquitin-like from hnRNP-K for its destabilization and in turn for protein that is conjugated to a variety of cellular proteins. Like release from cell-cycle arrest. Consistent with the rise- ubiquitin, SUMO is conjugated to target proteins by a cascade and-fall of both SUMOylation and stability of hnRNP-K, enzyme system consisting of E1 activating enzyme (SAE1/ its ability to interact with PIAS3 was inversely correlated SAE2), E2 conjugating enzyme (Ubc9), and E3 ligases to that with SENP2 during the time course after UV (PIASs) (Kerscher et al, 2006; Capili and Lima, 2007; Rytinki exposure. These findings indicate that SUMO modification et al, 2009). Conjugated SUMO can be removed by a family plays a crucial role in the control of hnRNP-K’s function as of SUMO-specific proteases (SENPs) (Mukhopadhyay and a p53 co-activator in response to DNA damage by UV. Dasso, 2007; Yeh, 2009). This reversible SUMOylation process The EMBO Journal (2012) 31, 4441–4452. doi:10.1038/ participates in the control of diverse cellular processes, emboj.2012.293; Published online 23 October 2012 including transcription, nuclear transport, and signal Subject Categories: proteins; genome stability & dynamics transduction (Kim et al, 2002; Johnson, 2004; Hay, 2005; Keywords: HDM2; p21; PIAS3; SENP2; ubiquitin Geiss-Friedlander and Melchior, 2007; Gareau and Lima, 2010). Significantly, many proteins involved in DNA-damage response are modified by ubiquitin and/or SUMO, implicating the role of ubiquitination, SUMOylation, or both in the control Introduction of checkpoint responses and DNA-repair pathways (Hoege The p53 tumour suppressor plays a pivotal role in mainte- et al, 2002; Lee et al, 2006; Bergink and Jentsch, 2009; nance of genome integrity under cellular stresses, such as Altmannova et al,2010;Dou et al,2010;Polo and Jackson, DNA damage (Lane, 1992; Lakin and Jackson, 1999; Kruse 2011; Cremona et al, 2012). For example, Rad52, a mediator of and Gu, 2009; Levine and Oren, 2009). Upon DNA damage, homologous recombination in yeast, is SUMOylated in ATM, ATR, and DNA-PK are activated for phosphorylation of response to DNA damage, and this modification stabilizes Rad52 for its sustained function (Sacher et al, 2006). *Corresponding authors. YJ Jeon or CH Chung, Department of Heterogeneous ribonucleoprotein-K (hnRNP-K) is an RNA- Biological Sciences, College of Natural Sciences, Seoul National University, 56-1 Shillim-dong, Gwanak-gu, Seoul 151-742, Korea. binding protein that is associated with various cellular pro- Tel.: þ82 2 880 6693; Fax: þ82 2 871 9193; cesses, including chromatin remodelling, transcription, E-mail: [email protected] or [email protected] mRNA splicing, and translation (Matunis et al, 1992; Bomsztyk et al, 1997, 2004). Intriguingly, hnRNP-K was Received: 8 June 2012; accepted: 8 October 2012; published online: 23 October 2012 shown to be transiently stabilized and function as a &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4441 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 1 UV induces SUMOylation of hnRNP-K. (A) UV increases the cellular level of hnRNP-K. After exposure of HeLa cells to UV (10 J/m ), cell lysates were subjected to immunoblot with anti-hnRNP-K or anti-p53 antibody. The resulting gels were scanned using a densitometer, and the intensities of hnRNP-K bands were quantified by using ‘Image J’ program. The intensity of hnRNP-K seen before UV (i.e., 0 h) was expressed as 1.0 and the others as its relative values. (B) UV induces SUMOylation of hnRNP-K. After UV treatment, cell lysates were subjected to immunoprecipitation with anti-hnRNP-K antibody followed by immunoblot with anti-SUMO1 or anti-hnRNP-K antibody. (C) Modification of hnRNP-K by SUMO isoforms. Flag-tagged SUMO isoforms were expressed in HEK293T cells with Flag-Ubc9 and HisMax-hnRNP-K. After incubation with 10 mM MG132 for 4 h, cell lysates were subjected to pull down with NTA beads followed by immunoblot with anti-Flag or anti-Xpress antibody. Figure source data can be found with the Supplementary data. transcriptional co-activator of p53 in response to DNA SUMOylation could be induced under other DNA damage damage (Moumen et al, 2005). However, how the stability conditions. Both SUMOylation and stabilization of hnRNP-K and function of hnRNP-K is regulated remained unknown. were also induced by treatments with ionizing radiation (IR) Here, we showed that UV induces PIAS3-mediated hnRNP-K and doxorubicin, although the timing of their rise-and-fall SUMOylation, which increases hnRNP-K stability, interaction was significantly different from that induced by UV between hnRNP-K and p53, and p21 expression in an ATR- (Supplementary Figure S1). Thus, hnRNP-K SUMOylation dependent manner, leading to cell-cycle arrest. At later appears to be a common response to DNA damage for its periods after UV treatment, however, SENP2 reversed the stabilization. SUMOylation-mediated processes by removing SUMO from When SUMO isoforms were overexpressed with hnRNP-K, hnRNP-K, implicating the role of SENP2 in the release of SUMO1 was more efficiently conjugated to hnRNP-K than cells from cell-cycle arrest to resume normal growth after SUMO2 or SUMO3 (Figure 1C). Thus, further studies were DNA repair. These findings indicate that reversible SUMO performed only with SUMO1. Since two SUMOylated hnRNP- modification of hnRNP-K by PIAS3 and SENP2 plays a crucial K bands appeared under the overexpression conditions, two role in the control of hnRNP-K stability and thereby its Lys residues in the sequences closely matched to the con- function as a p53 co-activator in response to DNA damage sensus motif for SUMOylation (c-K-X-D/E) were substituted by UV. with Arg (Figure 2A). Replacement of Lys422 alone or together with Lys198 by Arg prevented hnRNP-K SUMOylation, whereas that of Lys198 alone did not Results (Figure 2B). Similar results were obtained by in vitro UV-induced SUMOylation increases the stability of SUMOylation assay using purified SAE1/SAE2 (E1), Ubc9 (E2), and SUMO1 (Figure 2C), indicating that Lys422 hnRNP-K serves as the major SUMOylation site of hnRNP-K. hnRNP-K has been identified as a candidate for SUMOylation Henceforth, the SUMOylation-defective mutant was referred by proteomic analysis (Li et al, 2004). Therefore, we first to as K422R. examined whether hnRNP-K could indeed be modified by SUMO and whether this modification is related with DNA We next examined whether UV-induced SUMOylation in- damage-induced stabilization of hnRNP-K. UV treatment led fluences hnRNP-K ubiquitination and in turn its stability. The level of ubiquitinated hnRNP-K was markedly reduced at 6 h to 2- to 3-fold increase in the level of hnRNP-K by 6 h and after UV and returned almost to the initial level at 18 h after declined thereafter (Figure 1A). Moreover, the level of UV (Figure 3A), indicating that the change in the level of SUMOylated hnRNP-K was markedly increased by 6 h and ubiquitinated hnRNP-K is inversely correlated with that of declined by 18 h after UV treatment and this change occurred in parallel with that of hnRNP-K level (Figure 1B), suggesting SUMO1-conjugated hnRNP-K. However, SUMOylation-defec- that UV-induced SUMOylation stabilizes hnRNP-K. Thus, tive K422R, unlike wild-type hnRNP-K, remained ubiquiti- further studies were performed at three time points; prior nated at 6 h after UV (Figure 3B). Consistently, UV treatment increased the stability of hnRNP-K, but not K422R (Figure 3C to, 6 h after, and 18 h after UV treatment, which were and D). In addition, MG132, a proteasome inhibitor, pre- henceforth referred to as before UV, 6 h after UV, and 18 h vented K422R destabilization under the same conditions. after UV, respectively. We also examined whether hnRNP-K 4442 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 2 Lys422 is the major SUMO1 acceptor site in hnRNP-K. (A) Potential SUMOylation sites in hnRNP-K. The Lys residues in the underlined sequences of hnRNP-K were substituted with Arg by site-directed mutagenesis. (B) K422R mutation ablates hnRNP-K SUMOylation in vivo. Flag-tagged hnRNP-K, K198R, K422R, and the double mutant (K198R/K422R) were overexpressed in HEK293T cells with HisMax- SUMO1 and Flag-Ubc9. Cell lysates were subjected to immunoprecipitation with anti-Flag antibody followed by immunoblot with anti-Flag or anti-SUMO1 antibody. (C) K422R mutation ablates hnRNP-K SUMOylation in vitro. SUMOylation was performed using purified proteins followed by immunoblot with anti-His antibody as described under ‘Materials and methods’. Figure source data can be found with the Supplementary data. These results indicate that UV-induced SUMOylation of overexpressed with HDM2, p53, and Ubc9. Co-expression of hnRNP-K is responsible for the increase in its stability. increasing amounts of SUMO1 (i.e., increasing the level of SUMOylated hnRNP-K) led to a gradual increase in the level of hnRNP-K-bound p53 concurrently with a decrease in that SUMOylation of hnRNP-K switches its interaction with of hnRNP-K-bound HDM2 (Figure 4J). On the other hand, the HDM2 to that with p53 level of K422R-bound p53 and HDM2 remained the same To elucidate the mechanism for SUMOylation-mediated sta- bilization of hnRNP-K, we first examined the effect of UV regardless of SUMO1 expression. Although the experiments treatment on the interaction of hnRNP-K with HDM2. The were performed under overexpression conditions, which could be non-physiological, these results strongly suggest level of HDM2 co-immunoprecipitated with hnRNP-K was that SUMOylated hnRNP-K preferentially binds p53 whereas significantly decreased at 6 h after UV and returned to the its unmodified form binds better to HDM2. Thus, UV-induced initial level at 18 h after UV (Figure 4A). Moreover, the ability SUMOylation of hnRNP-K appears to switch its interaction of hnRNP-K to bind HDM2 was markedly reduced at 6 h after UV, whereas that of K422R remained the same regardless of with HDM2 to that with p53. UV treatment (Figure 4B and C). In addition, purified Of note was the finding that without UV treatment, SUMOylated hnRNP-K (Figure 4D) showed a lower affinity hnRNP-K binds p53 better than K422R (see Figure 4G and H), whereas K422R binds HDM2 better than hnRNP-K to HDM2 than unmodified hnRNP-K (Figure 4E). Note that (see Figure 4B and C). However, in vitro binding assays the C-terminal region harbouring the SUMOylation site showed that purified K422R interacts with p53 or HDM2 as Lys422 overlaps with that for HDM2 binding (see below). well as wild-type hnRNP-K (Supplementary Figure S2A These results indicate that UV-induced SUMOylation of hnRNP-K interferes with its interaction with HDM2, leading and B), indicating that the K-to-R mutation itself has no effect to hnRNP-K stabilization. on the binding affinity of hnRNP-K to p53 or HDM2. Since endogenous hnRNP-K can be SUMOylated in the absence of We next examined whether UV-induced SUMOylation also UV although to a basal level (see Figure 1B), it appeared that influences the interaction of hnRNP-K with p53. In contrast overexpression of hnRNP-K (i.e., elevation of the substrate to HDM2, the amount of p53 co-immunoprecipitated with concentration for SUMOylation) increases the level of hnRNP-K was significantly increased at 6 h after UV and returned almost to the initial level at 18 h after UV SUMOylated hnRNP-K and this increase alters the binding (Figure 4F). Moreover, the ability of hnRNP-K to bind p53 affinity of hnRNP-K to p53 and HDM2. Indeed, increased expression of hnRNP-K led to an increase in the level of was markedly increased at 6 h after UV, whereas that of SUMOylated hnRNP-K in the absence of UV treatment K422R remained decreased regardless of UV treatment (Supplementary Figure S2C). Moreover, when hnRNP-K (Figure 4G and H). In addition, purified SUMOylated SUMOylation was prevented by knockdown of Ubc9 by hnRNP-K showed a much higher affinity to p53 than unmodified hnRNP-K (Figure 4I). These results indicate that using Ubc9-specific shRNA (shUbc9), both hnRNP-K and UV-induced SUMOylation of hnRNP-K promotes its interac- K422R bound to p53 or HDM2 to similar extents tion with p53. (Supplementary Figure S3). These results indicate that changes in the binding affinity of hnRNP-K to p53 or HDM2 To confirm whether SUMOylation of hnRNP-K is in the absence of UV treatment are due to an increase in the responsible for the alterations in its affinity to HDM2 and level of SUMOylated hnRNP-K upon its overexpression. p53 under in vivo conditions, hnRNP-K and K422R were &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4443 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 3 UV-induced SUMOylation increases the stability of hnRNP-K. (A) UV blocks hnRNP-K ubiquitination. After exposure to UV, HeLa cells were incubated with 10 mM MG132 for 4 h. Cell lysates were subjected to immunoprecipitation with anti-ubiquitin, anti-hnRNP-K, or anti-SUMO1 antibody followed by immunoblot analysis. (B) SUMOylation prevents hnRNP-K ubiquitination. After exposure to UV, cells overexpressing HisMax-tagged hnRNP-K (Wt) or K422R (KR) were incubated for 2 h and then treated with MG132 for the next 4 h. Cell lysates were subjected to pull down with NTA beads followed by immunoblot analysis. (C) SUMOylation increases the hnRNP-K stability. Cells overexpressing Flag-tagged hnRNP-K (Wt) or K422R (KR) were treated with 200 mg/ml of cycloheximide. After exposure to UV, they were incubated with and without MG132 followed by immunoblot with anti-Flag antibody. (D) Band intensities in (C) were quantified by using a densitometer. The data represent the mean s.d. of three independent experiments. Figure source data can be found with the Supplementary data. SUMOylation of hnRNP-K is required for its function as p53 to the p21 promoter site and this increase could be further a p53 co-activator enhanced by hnRNP-K overexpression, but not by that of To determine whether UV-induced SUMOylation of hnRNP-K K422R (Figure 5E). These results indicate that UV-induced influences its co-activator function, p53 transactivity was hnRNP-K SUMOylation promotes p53 transactivity and measured by using two reporter vectors, PG13-Luc and thereby p21 expression. p21-Luc. In both cases, UV treatment increased the luciferase Of note was the finding that hnRNP-K overexpression leads activity and this increase was further enhanced by over- to an increase in the level of endogenous p53 in the absence expression of hnRNP-K, but not by that of K422R of UV treatment (see Figure 5D), raising a possibility that (Figure 5A and B). Under the same conditions, both mRNA overexpressed hnRNP-K may stabilize p53 although it has and protein levels of p21 were increased and this increase was been shown that hnRNP-K knockdown does not affect p53 further enhanced by overexpression of hnRNP-K, but not by stability (Moumen et al, 2005). However, expression of that of K422R (Figure 5C and D). hnRNP-K overexpression increasing amounts of hnRNP-K showed little or no effect without UV treatment also increased p53 transactivity, as it on HDM2-mediated p53 ubiquitination or HDM2 auto- could increase the SUMOylated hnRNP-K level. Moreover, ubiquitination, indicating that hnRNP-K has no effect on chromatin immunoprecipitation (ChIP) analysis revealed that the stability of p53 (Supplementary Figure S4). Since UV treatment increased recruitment of both hnRNP-K and hnRNP-K overexpression causes an increase in the level of 4444 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 4 hnRNP-K SUMOylation switches its interaction with HDM2 to that with p53. (A) UV inhibits the interaction of hnRNP-K with HDM2. After UV treatment, HeLa cells were incubated for the indicated periods. Cell lysates were subjected to immunoprecipitation with anti-hnRNP-K antibody followed by immunoblot with anti-HDM2 and anti-hnRNP-K antibodies. (B, C) SUMOylation inhibits the interaction of hnRNP-K with HDM2. HDM2 was overexpressed in cells with Flag-tagged hnRNP-K or K422R. After exposure to UV, cells were incubated for 6 h. Cell lysates were subjected to immunoprecipitation with anti-HDM2 (B) or anti-Flag antibody (C). (D) Purification of His- SUMO1-conjugated GST-hnRNP-K. SUMOylated hnRNP-K proteins eluted from NTA-agarose column were subjected to SDS–PAGE followed by staining with Coomassie blue R-250. Fractions under the bar were pooled for further use. (E) SUMOylation reduces the affinity of hnRNP-K to HDM2. Purified His-HDM2 was incubated with GST-hnRNP-K-His-SUMO1 or GST-hnRNP-K followed by immunoprecipitation with anti- hnRNP-K antibody. (F) UV promotes the interaction of hnRNP-K with p53. Experiments were performed as in (A), except that anti-p53 antibody was used in place of anti-HDM2 antibody. (G, H) SUMOylation increases the affinity of hnRNP-K to p53. Myc-p53 was overexpressed in cells with Flag-tagged hnRNP-K or K422R. After exposure to UV, cells were incubated for 6 h. Cell lysates were subjected to immunoprecipitation with anti-Myc (G) or anti-Flag antibody (H). (I) SUMOylated hnRNP-K shows higher affinity to p53. Experiments were done as in (E), except that His-p53 was used in place of His-HDM2. (J) SUMOylation inversely affects the binding of hnRNP-K to HDM2 and p53. HisMax-tagged hnRNP-K (Wt) and K422R (KR) were overexpressed in cells with Myc-Ubc9, HA-p53, HDM2, and increasing amounts of Flag-SUMO1. Cell lysates were subjected to pull down with NTA beads followed by immunoblot analysis. Note that 10 mM MG132 was treated 4 h before cell lysis in (A–C), (F–H), and (J). Figure source data can be found with the Supplementary data. SUMOylated hnRNP-K even in the absence of UV (see overexpressed hnRNP-K-mediated increase in endogenous Supplementary Figure S2) and since p53 is known to posi- p53 level without UV treatment is due to the ability of tively regulate its own expression, it appears likely that the SUMOylated hnRNP-K in promotion of p53 expression. &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4445 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 5 SUMOylation of hnRNP-K is required for its function as a p53 co-activator. (A, B) SUMOylation of hnRNP-K promotes p53 transactivity. HeLa cells overexpressing Flag-tagged hnRNP-K or K422R were transfected with PG13-Luc (A)or P21-Luc (B). After exposure to UV, cells were incubated for 6 h. Cell lysates were assayed for the luciferase activity. The activity seen without hnRNP-K overexpression and UV treatment was expressed as 1.0 and the others were as its relative values. The data represent the mean s.d. of three experiments. (C) SUMOylation of hnRNP-K increases the level of p21 transcripts. Total RNAs prepared from the same cells used in (A) were subjected to RT–PCR to determine p21 mRNA levels. (D) SUMOylation of hnRNP-K promotes p21 expression. Cell lysates prepared as in (A) were subjected to immunoblot with anti-p53, anti-p21, or anti-hnRNP-K antibody. (E) SUMOylation of hnRNP-K promotes recruitment of both hnRNP-K and p53 to the p21 promoter. Cells prepared as in (A) were subjected to ChIP assay by using anti-hnRNP-K or anti-p53 antibody. Precipitated DNAs were subjected to PCR with primers covering the p53-response element in the p21 gene. Figure source data can be found with the Supplementary data. PIAS3 and SENP2 counteract on SUMO modification of were generated and subjected to pull-down analysis hnRNP-K (Supplementary Figure S6A–D). Both p53 and SENP2 To identify hnRNP-K-specific SUMO E3 ligase, each of PIAS1-4 bound to the same N-terminal region of hnRNP-K was overexpressed with hnRNP-K. Among them, PIAS3 spe- (Supplementary Figure S6E), suggesting that p53 and cifically interacted with hnRNP-K (Supplementary Figure S5A) SENP2 could compete with each other for binding to and promoted its SUMOylation (Figure 6A). Furthermore, hnRNP-K. While PIAS3 interacted with the middle region of PIAS3 knockdown by shPIAS3 prevented not only hnRNP-K hnRNP-K, HDM2 bound to its C-terminal region, which SUMOylation but also p21 expression (Figure 6B), suggesting includes the SUMO-conjugation site Lys422. The latter data that PIAS3-mediated SUMOylation of hnRNP-K is required for are consistent with the finding that SUMOylated hnRNP-K its function as a p53 co-activator. Interestingly, the amount of shows a lower affinity to HDM2 than its unmodified form (see PIAS3 co-immunoprecipitated with hnRNP-K was significantly Figure 4C and E). To identify hnRNP-K-binding regions within increased at 6 h after UVand returned almost to the initial level HDM2, p53, SENP2, and PIAS3, deletions of each protein at 18 h-after-UV (Figure 6C). Thus, it appears that UV-mediated were generated (Supplementary Figure S7). hnRNP-K bound increase in hnRNP-K SUMOylation is due to an increase in the to the C-terminal regions of p53, SENP2, and PIAS3, while it affinity of hnRNP-K to PIAS3. is interacted with the middle region of HDM2. A map for the We next attempted to identify hnRNP-K-specific interaction between hnRNP-K and HDM2, p53, SENP2, or deSUMOylating enzyme. Among the enzymes tested, over- PIAS3 was shown in Figure 7. expressed SENP1, SENP2, SENP6, and mouse SUSP4 inter- acted with hnRNP-K (Supplementary Figure S5B). Without Effect of hnRNP-K SUMOylation on its subcellular overexpression, however, only SENP2 interacted with localization hnRNP-K and this interaction was markedly decreased at Previous studies have suggested that SUMOylation of hnRNP- 6 h after UV and recovered at 18 h after UV (Figure 6D K is involved in its nucleocytoplasmic transport (Vassileva and E). Moreover, SENP2, but not its catalytically inactive and Matunis, 2004). Therefore, we examined whether UV- mutant (in which the active site Cys548 was replaced by Ser), induced SUMOylation influences the subcellular localization removed SUMO from hnRNP-K (Figure 6F), whereas SENP2 of hnRNP-K. Immunocytochemical analysis showed that in knockdown by shSENP2 promoted hnRNP-K SUMOylation the absence of UV treatment, overexpressed hnRNP-K resided (Figure 6G). Notably, without UV treatment SENP2 in both the nucleus and the cytoplasm in B30% of cells as knockdown significantly increased the level of SUMOylated well as exclusively in the nucleus in the remaining cells hnRNP-K, suggesting that endogenous SENP2 rapidly (Supplementary Figure S8A). In its presence, however, the deSUMOylates hnRNP-K under unstressed conditions. entire hnRNP-K proteins were localized exclusively in the Collectively, these results demonstrate that PIAS3 and SENP2 nucleus. In contrast, K422R was localized in both the nucleus antagonistically regulate SUMO modification and stability of and the cytoplasm in B40% of cells regardless of UV treat- hnRNP-K during the time course after exposure to UV. ment, suggesting that SUMOylation is involved in the nuclear To map the regions within hnRNP-K for binding of localization of hnRNP-K. Therefore, we next examined HDM2, p53, SENP2, and PIAS3, deletions of hnRNP-K whether the nuclear localization of hnRNP-K could be 4446 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 6 PIAS3 and SENP2 antagonistically regulate hnRNP-K SUMOylation. (A) PIAS3 promotes hnRNP-K SUMOylation. HisMax-hnRNP-K was overexpressed in HEK293T cells with Flag-SUMO1, Flag-Ubc9, and Myc-PIAS3. Cell lysates were subjected to pull down with NTA beads followed by immunoblot with anti-SUMO1 or anti-Xpress antibody. (B) PIAS3 knockdown blocks hnRNP-K SUMOylation. HeLa cells transfected with shNS or shPIAS3 were exposed to UV and then incubated for 6 h. Cell lysates were subjected to immunoprecipitation with anti-hnRNP-K antibody followed by immunoblot with anti-SUMO1 or anti-hnRNP-K antibody. (C) UV promotes the interaction of hnRNP-K with PIAS3. After exposure to UV, cells were incubated for the indicated periods. Cell lysates were subjected to immunoprecipitation with anti-hnRNP-K antibody followed by immunoblot with anti-PIAS3 or anti-hnRNP-K antibody. (D, E) UV inhibits the interaction of hnRNP-K with SENP2. Cells treated with UV were subjected to immunoprecipitation with anti-hnRNP-K (D)or anti-SENP2 antibody (E). Note that MG132 was treated 4 h before cell lysis in (C–E). (F) SENP2 deSUMOylates hnRNP-K. HisMax-hnRNP-K was overexpressed in HEK293T cells with Flag-SUMO1, Flag-Ubc9, and Myc-tagged SENP2 (Wt) or its catalytically inactive form (CS). Cell lysates were subjected to pull down with NTA beads followed by immunoblot with anti-SUMO1 or anti-Xpress antibody. (G) SENP2 knockdown promotes hnRNP-K SUMOylation. HeLa cells transfected with shNS or shSENP2 were exposed to UV and incubated for 6 h. Cell lysates were then treated as in (B). Figure source data can be found with the Supplementary data. Figure 7 Map for the interaction between hnRNP-K and p53, HDM2, SENP2, or PIAS3. The data obtained from Supplementary Figures S6 and S7 were summarized. In p53: TAD, transcription activation domain; DBD, DNA-binding domain; TET, tetramerization domain; REG, regulatory domain. In HDM2: p53BP, p53 binding domain; AD, acidic domain; ZF, zinc finger; RING, really interesting gene. In SENP2: NLD, nuclear localization domain; NES, nuclear export signal; SIM, SUMO-interacting motif; CD, catalytic domain. In PIAS3: SAP, SAF-A/B, Acinus and PIAS; PINIT, Pro-Ile-Asn-Ile-Thr; S/T, Ser/Thr-rich. blocked by knockdown of PIAS3. Depletion of PIAS3 hnRNP-K could also be prevented by overexpression of prevented UV-induced nuclear localization of hnRNP-K wild-type SENP2, but not by its catalytically inactive mutant (Supplementary Figure S8B). The nuclear localization of in which the active site Cys548 was replaced by Ser &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4447 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al (Supplementary Figure S8C). These results again suggest that by supplement of shhnRNP-K-insensitive hnRNP-K, but not SUMOylation is involved in the nuclear localization of by that of shhnRNP-K-insensitive K422R (Figure 8F). These hnRNP-K. results indicate that hnRNP-K SUMOylation is required for To confirm this finding, we examined whether the localiza- UV-induced cell-cycle arrest. Knockdown of SENP2 enhanced tion of endogenous hnRNP-K could be influenced by PIAS3 UV-mediated increase in cell fractions in G1 phase, whereas knockdown in the presence and absence of UV. In contrast to that of PIAS3 ablated it (Figure 8G). Knockdown of SENP2 the data obtained by hnRNP-K overexpression, neither UV together with hnRNP-K also prevented UV-induced cell-cycle treatment nor PIAS3 depletion showed any effect on the arrest. Figure 8H shows that PIAS3 knockdown decreases the nuclear localization of hnRNP-K (Supplementary Figure S9). expression of p21 as well as of p53 whereas SENP2 knock- Noteworthy, however, were the findings that in cells having down increases it, and this increase could be ablated by overexpressed hnRNP-K in both the nucleus and the cyto- simultaneous knockdown of hnRNP-K. Collectively, these plasm, the portion of hnRNP-K located in the cytoplasm is results indicate that reversible SUMO modification of much lower than that in the nucleus and that B60% of hnRNP-K by SENP2 and PIAS3 plays a key role in the control SUMOylation-defective K422R mutant resided exclusively in of p21-mediated cell-cycle arrest in response to UV damage. the nucleus regardless of UV treatment (see Supplementary UV-induced DNA damage response is at least in part Figure S8A). Furthermore, overexpression of SENP2 did not mediated by ATR kinase, which phosphorylates downstream alter the population of cells that have hnRNP-K exclusively in targets, such as p53 and CHK1 (Durocher and Jackson, 2001; the nucleus, although it can prevent the UV-induced nuclear Cimprich and Cortez, 2008). Treatment with caffeine, an localization of hnRNP-K in the cytoplasm (see Supplementary inhibitor of ATR (in addition to ATM), and knockdown of Figure S8C). In addition, endogenous hnRNP-K is known to ATR by shATR abrogated not only UV-induced hnRNP-K predominantly localize in the nucleus, and this nuclear SUMOylation and but also the increased interaction of localization is mediated mainly by the nuclear shuttling PIAS3 with hnRNP-K at 6 h after UV (Figure 9A and B). sequence (KNS) and in part by the NLS sequence in They also resulted in sustained interaction of hnRNP-K with hnRNP-K (Michael et al, 1977). Thus, it appears likely that SENP2. These results indicate that UV-induced hnRNP-K SUMOylation plays only a minor, auxiliary role in the nuclear SUMOylation, which is essential for hnRNP-K’s function as localization of hnRNP-K. a p53 co-activator, is ATR dependent. UV-induced hnRNP-K SUMOylation is required for Discussion cell-cycle arrest To determine whether SENP2 is involved in the control of Based on the findings in this study, we propose a model for the hnRNP-K’s co-activator function by altering its SUMOylation role of hnRNP-K SUMOylation in UV-induced cell-cycle arrest state, PG13-Luc and p21-Luc were again used for assaying p53 (Figure 9C). Under unstressed conditions, SENP2 removes transactivity. In both cases, UV treatment increased p53 trans- SUMO from SUMO-conjugated hnRNP-K, if there is any, allow- activity and this increase was further enhanced by SENP2 ing hnRNP-K to bind HDM2 with high affinity for its ubiquiti- knockdown (Figure 8A and B; Supplementary Figure S10A). nation and subsequent degradation by proteasome. Although Without UV treatment, SENP2 knockdown moderately pro- not shown, HDM2 also promotes proteasome-mediated degra- moted p53 transactivity, since it increases the level of dation of p53. Upon exposure to UV, PIAS3 binds and ligates SUMOylated hnRNP-K (see Figure 6G). These stimulatory SUMO to hnRNP-K, allowing p53 to bind SUMOylated hnRNP- effects of SENP2 knockdown on p53 transactivity were ablated K with high affinity and recruitment of their complex to the p21 promoter. Thus, hnRNP-K SUMOylation by PIAS3 serves as a by simultaneous knockdown of hnRNP-K by shhnRNP-K, indicating that the observed effects are specific to hnRNP-K. critical switch for shifting the interaction of hnRNP-K with These results indicate that SENP2 negatively regulates the HDM2 to that with p53 for its function as a transcriptional function of hnRNP-K as a p53 co-activator. On the other co-activator of p53 and in turn for p21 expression and cell-cycle hand, knockdown of PIAS3 alone or together with hnRNP-K arrest in response to DNA damage by UV. completely abrogated UV-induced p53 transactivity (Figure 8C Of interest was the finding that the level of SUMOylated and D; Supplementary Figure S10B), indicating that PIAS3 hnRNP-K increases and then declines during the time course positively regulates the co-activator function of hnRNP-K. after exposure of cells to UV. This phenomenon is apparently Collectively, these results indicate that SENP2 and PIAS3 antag- mediated by a rise-and-fall of hnRNP-K’s ability to interact onistically regulate the role of hnRNP-K as a p53 co-activator. with PIAS3 and its inversed ability to bind SENP2 during the We next examined whether hnRNP-K SUMOylation is re- same time course. This reversible SUMOylation process that quired for UV-induced cell-cycle arrest upon flow cytometry. should occur in conjunction with the p53-HDM2 feedback UV treatment increased cell fractions in G1 phase and this loop is of importance for cells to escape from cell-cycle arrest increase was blocked by hnRNP-K knockdown (Figure 8E; and to resume normal growth after the repair of damaged Supplementary Figure S11). Supplement of shhnRNP-K-insen- DNA. However, it remains unknown how the binding ability sitive hnRNP-K to cells that had been depleted of endogenous of hnRNP-K is shifted to PIAS3 and then to SENP2 after UV hnRNP-K, but not that of shhnRNP-K-insensitive K422R, treatment. Since ATR knockdown prevents the interaction of restored accumulation of G1-phase cells. To determine hnRNP-K with PIAS3 but promotes that with SENP2 and whether the cell-cycle arrest is mediated by p53-induced since UV does not affect the expression of either PIAS3 or expression of p21, the same cells used for flow cytometry SENP2, it seems possible that ATR-mediated phosphorylation were subjected to immunoblot analysis. Knockdown of of hnRNP-K followed by dephosphorylation by an unknown hnRNP-K led to a marked decrease in the expression of p21 protein phosphatase(s) might change the affinity of hnRNP-K as well as in that of p53, and this decrease could be reversed to PIAS3 and SENP2. Notably, hnRNP-K has the SQ and TQ 4448 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 8 PIAS3 and SENP2 antagonistically regulate cell-cycle arrest. (A, B) SENP2 knockdown promotes p53 transactivity. HeLa cells transfected with shNS or shSENP2 alone or together with shhnRNP-K were incubated for 48 h. They were then transfected with PG13-Luc (A)or P21-Luc (B) and further incubated for the next 24 h. After exposure to UV, cells were incubated for 6 h. Cell lysates were assayed for luciferase. The enzyme activity seen in cells transfected with shNS only but without UV treatment was expressed as 1.0 and the others were as its relative values. (C, D) PIAS3 knockdown ablates p53 transactivity. Experiments were performed as above, except that cells were transfected with shPIAS3 in place of shSENP2. (E, F) SUMOylation of hnRNP-K is required for p21-mediated cell-cycle arrest. Cells transfected with shNS or shhnRNP-K were complemented with shhnRNP-K-insensitive Flag-tagged hnRNP-K or K422R. After exposure to UV, they were incubated for 6 h followed by flow cytometry (E) or immunoblot analysis (F). (G, H) PIAS3 and SENP2 inversely regulate cell-cycle arrest. Cells were transfected with shNS, shSENP2, or shPIAS3 alone or together with shhnRNP-K. After exposure to UV, cells were incubated for 6 h followed by flow cytometry (G) or immunoblot analysis (H). The data in (A–E) and (G) represent the mean s.d. of four experiments. Figure source data can be found with the Supplementary data. motifs that can be phosphorylated by ATR (Kim et al, 1999). PIAS3 serves as an endogenous protein inhibitor of acti- However, replacement of the Ser and Thr residues by Ala or vated signal transducers and activators of transcription 3 Glu showed little or no effect on UV-induced SUMOylation of (STAT3), in addition to its role as a SUMO E3 ligase (Chung hnRNP-K. Nonetheless, we could not exclude a possibility et al, 1997; Jackson, 2001; Jang et al, 2004). The STAT3 that kinases downstream of ATR, such as CHK1, or other protein, which promotes cell-cycle progression and inhibits kinase(s) and unknown phosphatases may be involved in apoptosis, has been implicated in the pathogenesis of reversible phosphorylation of hnRNP-K and in turn in the various human cancers (Niu et al, 2002; Wei et al, 2003; control of the affinity of hnRNP-K to PIAS3 and SENP2. Levy and Inghirami, 2006). Interestingly, PIAS3 expression is &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4449 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 9 SUMOylation of hnRNP-K is ATR dependent. (A) Caffeine inhibits hnRNP-K SUMOylation. After exposure of HeLa cells to UV, they were incubated with and without 5 mM caffeine for the indicated periods. Cell lysates were subjected to immunoprecipitation with anti-hnRNP- K antibody followed by immunoblot with anti-SUMO1, anti-SENP2, anti-PIAS3, or anti-hnRNP-K antibody. (B) ATR knockdown prevents hnRNP-K SUMOylation. Cells transfected with shNS or shATR were incubated for 48 h. After exposure to UV, cells were incubated for the indicated periods. Cell lysates were subjected to immunoprecipitation as in (A). Note that MG132 was treated 4 h before cell lysis in (A, B). (C) A model for the role of hnRNP-K SUMOylation in UV-induced cell-cycle arrest. Figure source data can be found with the Supplementary data. downregulated in several cancers, such as human gastric p53 and induction of tumour cell apoptosis) releases HDM2 carcinoma, glioblastoma, and squamous cell carcinoma of from p53 and the freed HDM2 molecules promote protea- the lung (Brantley et al, 2008; Kluge et al, 2011; Liu et al, some-mediated degradation of hnRNP-K, which impairs its 2011). Therefore, it has been suggested that loss or reduction co-activator function in p53-mediated p21 expression (Enge of PIAS3 expression contributes to enhanced STAT3 et al, 2009). HDM2 also directly promotes p21 degradation by transcriptional activity, leading to aberrant cell proliferation proteasome, eliminating anti-apoptotic function of p21 and and tumorigenesis. Here, we showed that PIAS3 promotes thus switching towards induction of apoptosis. Deregulated hnRNP-K SUMOylation and thereby p53-mediated cell-cycle cell-cycle progression such as by preventing p21-mediated arrest. Thus, PIAS3 might exert its anti-tumorigenic function cell-cycle arrest could also evoke tumorigenesis. Therefore, in both E3 ligase activity-dependent and -independent small molecules that specifically inhibit SENP2 could be used manners by promoting cell-cycle arrest. as therapeutic drugs against cancers, as they should prevent Targeted disruption of SENP2 in mice was shown to impair hnRNP-K deSUMOylation and in turn induce p53 activation cell-cycle progression at the G1/S phase, leading to abnorm- for p21 expression, leading to cell-cycle arrest. alities in trophoblast proliferation and differentiation (Chiu While our work was under revision, another study reported et al, 2008). During trophoblast development, SENP2 removes that DNA damage induces hnRNP-K SUMOylation, which in SUMO from Mdm2 and thereby promotes Mdm2-mediated turn enhances the transcriptional activity of p53 (Pelisch et al, ubiquitination of p53 and its subsequent degradation by 2012). However, the major differences between their and our proteasome, allowing cell-cycle progression. Disruption of works are the effect of SUMOylation on the stability of hnRNP-K the SENP2 gene, however, results in cytoplasmic localization and the identity of hnRNP-K-specific SUMO E3 ligase. While they of Mdm2 and in turn in p53 stabilization in the nucleus, concluded that SUMOylation does not alter hnRNP-K stability leading to p53-mediated cell-cycle arrest. On the other hand, and that Pc2 acts as an E3 ligase, we found that the same overexpression of SENP2 makes cells resistant to apoptosis modification leads to hnRNP-K stabilization and that PIAS3 induced by genotoxic stress, such as doxorubicin treatment, serves as an hnRNP-K-specific ligase. These differences might indicating that SENP2 plays a critical role in the control of cell- be due to the use of a single time point in analysing the effect of cycle progression (Jiang et al, 2011). Here, we showed that DNA damage on alterations in the level of hnRNP-K and to the SENP2 knockdown increases hnRNP-K SUMOylation, its use of overexpression system in analysing the role of Pc2 in interaction with p53, and consequently its co-activator hnRNP-K SUMOylation rather than that of RNA interference function in expression of p53-downstream genes, such as system. However, we could not exclude a possibility that diffe- p21, for cell-cycle arrest. In addition, it has been shown that rent SUMO E3 ligases may act on hnRNP-K under different DNA under unstressed conditions hnRNP-K is ubiquitinated by damage conditions, since their work mainly used doxorubicin as HDM2 for degradation by proteasome (Moumen et al, a DNA damaging agent whereas ours utilized UV. 2005). Thus, it appears that SENP2 could regulate cell-cycle progression by targeting two different substrates: one by deSUMOylating HDM2 for HDM2-mediated degradation of Materials and methods p53 and the other by deSUMOylating hnRNP-K for HDM2- Plasmids and antibodies mediated degradation of hnRNP-K, which ablates its function hnRNP-K cDNA was isolated from a cDNA library of HeLa cells, and cloned into pcDNA-HisMax and pCMV2-Flag. It was also cloned as a p53 co-activator. into pET-32b and pGEX-4T3 for bacterial expression. shRNAs were The p21 protein can function as an anti-apoptotic protein as purchased from Open Biosystems. Target sequences for shRNAs are well as an inhibitor of CDKs for cell-cycle arrest. Recently, it 0 0 0 as follows: shhnRNP-K, 5 -ACGATGAAACCTATGATTA-3 ; shSENP2, 5 - 0 0 was shown that an anti-cancer drug, RITA (reactivation of CCCACAGGATGAAATCCTA-3;shPIAS3, 5 -GCTGTCGGTCAGACAT 4450 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al 0 0 CATTT-3 . Antibodies against Myc (9E10), p53 (DO-1), p21 (C-19), instructions. Primers used in PCR for p21 were 5 -CTTTGTCACCGA 0 0 0 hnRNP-K (D-6), GAPDH (2D4A7), HDM2 (SMP14), Ub (A-5), Ubc9 GACACCAC-3 and 5 -GGCGTTTGGAGTGGTAGAAA-3 . (N-15), and GST (Z-5) were purchased from Santa Cruz (Santa Cruz, Luciferase assays CA, USA). Anti-Flag M2 (Sigma-Aldrich), anti-Xpress, anti-SUMO-1 HeLa cells transfected with pcDNA-b-Gal and PG13-Luc or p21-Luc (Invitrogen), anti-His (BD Biosciences), anti-SENP1, anti-SENP2, were incubated for 48 h. After UV treatment, cells were cultured for anti-SENP6 (Abgent), and anti-PIAS3 (Cell Signaling) antibodies 6 h, harvested, and assayed for luciferase. The enzyme activity was were also used. measured in a luminometer and normalized by b-galactosidase Cell culture and transfection expression with a luciferase system (Promega). HEK293T and HeLa cells were grown at 371C in DMEM supplemen- ted with 100 units/ml penicillin, 1 mg/ml streptomycin, and 10% ChIP assay FBS. MRC5 cells were cultured as above except that the use of Assays were conducted with an average size of sheared fragments MEM in place of DMEM. All transfections were carried out using of about 300–1000 bps as described (Jepsen et al, 2000; Shang et al, Metafectene reagent (Biontex) and jetPEIt DNA Transfection 2000). For PCR, 1 ml from 50 ml DNA extraction and 25–30 cycles of Reagent (Polyplus-transfection). amplification were used. Primers used in PCR of p21 promoter 0 0 0 sequence were 5 -GTGGCTCTGATTGGCTTTCTG-3 and 5 -CTGAAA ACAGGCAGCCCAAGG-3 (Zeng et al, 2002). Assays for SUMO modification HisMax-hnRNP-K, Flag-SUMO1, and Flag-Ubc9 were overexpressed in HEK293T cells with or without Myc-tagged SENP2 or PIAS3. After Purification of recombinant SUMOylated hnRNP-K culturing for 36 h, cells were lysed by boiling for 10 min in 150 mM For production of SUMOylated hnRNP-K, BL21(DE3) cells were Tris–HCl (pH 8), 5% SDS, and 30% glycerol. Cell lysates were diluted transformed with pGEX4T3-hnRNP-K and pT-E1/E2/His-SUMO1. 20-fold with buffer A consisting of 20 mM Tris–HCl (pH 8.0), 150 mM BL21 colonies carrying both plasmids were selected as described NaCl, 10 mM imidazole, 1% Triton X-100, 1 protease inhibitor (Zeng et al, 2002). Extracts (10 mg) from the cells were loaded onto 2þ cocktail (Roche), and 2 mM NEM. After incubating them with Ni - a glutathione-Sepharose 4B column, and proteins bound to the NTA-agarose for 2 h at 41C, the resins were collected, washed with column were eluted with PBS containing 50 mM glutathione. After buffer A containing 20 mM imidazole, and boiled in SDS-sampling dialysis against buffer C consisting of NaH PO /Na HPO (pH 8), 2 4 2 4 buffer. Supernatants were subjected to SDS–PAGE followed by immu- 0.5 M NaCl, 50 mM imidazole, 1% Triton X-100, and 2 mM noblot analysis. For assaying SUMOylation of endogenous hnRNP-K, 2-mercaptoethanol, proteins were loaded onto a NTA-agarose HeLa cells without any overexpression were lysed as above. Cell column. Bound proteins (i.e., GST-hnRNP-K-His-SUMO1) were lysates were diluted 20-fold with buffer B consisting of 20 mM Tris– eluted with buffer C containing 200 mM imidazole. HCl (pH 8.0), 150 mM NaCl, 0.1 mM EDTA, 0.2% Triton X-100, 1 protease inhibitor cocktail, and 2 mM NEM. The samples were incu- Immunocytochemistry bated with anti-hnRNP-K antibody for 2 h at 41C and then with protein- HeLa cells were grown on coverslips. After transfection, they were A-Sepharose for the next 2 h. The resins were collected, washed with fixed by incubation with 3.7% paraformaldehyde in PBS for 10 min. buffer B containing 1% Triton X-100, and boiled. Supernatants were Cells were washed three times with PBS containing 0.1% Triton subjectedtoSDS–PAGEfollowedbyimmunoblotanalysis. X-100, and permeabilized with 0.5% Triton X-100 in PBS for 5 min. For in vitro SUMOylation assay, purified His-hnRNP-K (2 mg), After blocking with 3% BSA in PBS for 30 min, cells were incubated SUMO1 (5 mg), SAE1/SAE2 (1.5 mg), and Ubc9 (5 mg) were incubated for 1 h with appropriate antibodies. After washing with PBS con- with an ATP-regenerating system consisting of 50 mM Tris–HCl (pH taining 0.1% Triton X-100, cells were incubated for 1 h with FITC- or 7.5), 5 mM MgCl , 5 mM ATP, 10 mM creatine phosphate, 5 units/ml TRITC-conjugated secondary antibody in PBS containing 3% BSA. of phosphocreatine kinase, and 1  protease inhibitor cocktail in a Cells were then observed using a confocal laser scanning total volume of 30 ml. After incubating the mixtures for 2 h at 371C, microscope (Carl Zeiss-LSM700). they were subjected to SDS–PAGE followed by immunoblot. Supplementary data Flow cytometry Supplementary data are available at The EMBO Journal Online Cells were washed with PBS, trypsinized, and fixed at 41C with 70% (http://www.embojournal.org). ethanol. They were washed with PBS and incubated in PBS contain- ing 0.1% Triton X-100, 200 mg/ml of RNase A, and 20 mg/ml of propidium iodide for 30 min at room temperature in the dark. DNA Acknowledgements contents were then determined by flow cytometry using FACSCalibur (Becton Dickinson). This work was supported by grants from the National Research Found- ation of Korea (NRF-2005-084-C00025 and M10533010001-05N3301). SWL, HMY, and JHP were the recipients of the BK21 fellowship. UV and c irradiation Author contributions: SWL, YJJ, and CHC analysed the data, Cells cultured to 50–70% confluence were washed with PBS and conceived the study, and wrote the paper. SWL and MHL initiated irradiated at 254 nm (UV-C) by using TUV lamp (Philips) or at 5 Gy the project and SWL performed all the experiments except purifica- by using Gammacell Low Dose-rate Research Irradiator (GC 3000 tion of p53, hnRNP-K, SUMO1, Sae1/2, and Ubc9, which was done Elan). UV dose (10 J/m ) was determined by using a UVX radio- by SHK, JHP, and HMY. Cloning of hnRNP-K cDNA and other meter (UVP Inc.). They were then incubated for various periods in plasmids was performed by SHK and SWL, and immunocyto- DMEM supplemented with 100 units/ml of penicillin, 1 mg/ml of chemistry was done by SWL and YMO. streptomycin, and 10% FBS. RT–PCR Total RNAs were isolated from cells by using TRIzol (Invitrogen). Conflict of interest RT–PCR was performed using RevertAid M–MuLV reverse transcrip- tase (Invitrogen) and oligo (dT) primer, according to manufacturer’s The authors declare that they have no conflict of interest. References Abraham RT (2001) Cell cycle checkpoint signaling through the SUMOylation affects the efficiency of the DNA repair. Nucleic ATM and ATR kinases. Genes Dev 15: 2177–2196 Acids Res 38: 4708–4721 Abraham RT (2004) PI 3-kinase related kinases: ’big’ players Baptiste-Okoh N, Barsotti AM, Prives C (2008) Caspase 2 is both in stress-induced signaling pathways. DNA Repair (Amst) 3: required for p53-mediated apoptosis and downregulated by p53 883–887 in a p21-dependent manner. Cell Cycle 7: 1133–1138 Altmannova V, Eckert-Boulet N, Arneric M, Kolesar P, Chaloupkova Bartek J, Lukas J (2001) Mammalian G1- and S-phase checkpoints R, Damborsky J, Sung P, Zhao X, Lisby M, Krejci L (2010) Rad52 in response to DNA damage. Curr Opin Cell Biol 13: 738–747 &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4451 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Bergink S, Jentsch S (2009) Principles of ubiquitin and SUMO Kruse JP, Gu W (2009) Modes of p53 regulation. Cell 137: 609–622 modifications in DNA repair. Nature 458: 461–467 Lakin ND, Jackson SP (1999) Regulation of p53 in response to DNA Bomsztyk K, Denisenko O, Ostrowski J (2004) hnRNP K: one damage. Oncogene 18: 7644–7655 protein multiple processes. Bioessays 26: 629–638 Lane DP (1992) Cancer. p53, guardian of the genome. Nature 358: 15–16 Bomsztyk K, Van Seuningen I, Suzuki H, Denisenko O, Ostrowski J Lee MH, Lee SW, Lee EJ, Choi SJ, Chung SS, Lee JI, Cho JM, Seol JH, (1997) Diverse molecular interactions of the hnRNP K protein. Baek SH, Kim KI, Chiba T, Tanaka K, Bang OS, Chung CH (2006) FEBS Lett 403: 113–115 SUMO-specific protease SUSP4 positively regulates p53 by pro- Brantley EC, Nabors LB, Gillespie GY, Choi YH, Palmer CA, Harrison moting Mdm2 self-ubiquitination. Nat Cell Biol 8: 1424–1431 K, Roarty K, Benveniste EN (2008) Loss of protein inhibitors of Levine AJ, Oren M (2009) The first 30 years of p53: growing ever activated STAT-3 expression in glioblastoma multiforme tumors: more complex. Nat Rev Cancer 9: 749–758 implications for STAT-3 activation and gene expression. Clin Levy DE, Inghirami G (2006) STAT3: a multifaceted oncogene. Proc Cancer Res 14: 4694–4704 Natl Acad Sci USA 103: 10151–10152 Capili AD, Lima CD (2007) Taking it step by step: mechanistic Li T, Evdokimov E, Shen RF, Chao CC, Tekle E, Wang T, Stadtman insights from structural studies of ubiquitin/ubiquitin-like protein ER, Yang DC, Chock PB (2004) Sumoylation of heterogeneous modification pathways. Curr Opin Struct Biol 17: 726–735 nuclear ribonucleoproteins, zinc finger proteins, and nuclear pore Chiu SY, Asai N, Costantini F, Hsu W (2008) SUMO-specific protease complex proteins: a proteomic analysis. Proc Natl Acad Sci USA 2 is essential for modulating p53-Mdm2 in development of 101: 8551–8556 trophoblast stem cell niches and lineages. PLoS Biol 6: e310 Liu LM, Yan MG, Yang DH, Sun WW, Zhang JX (2011) PIAS3 Chung CD, Liao J, Liu B, Rao X, Jay P, Berta P, Shuai K (1997) expression in human gastric carcinoma and its adjacent non- Specific inhibition of Stat3 signal transduction by PIAS3. Science tumor tissues. Clin Res Hepatol Gastroenterol 35: 393–398 278: 1803–1805 Matunis MJ, Michael WM, Dreyfuss G (1992) Characterization and Ciccia A, Elledge SJ (2010) The DNA damage response: making it primary structure of the poly(C)-binding heterogeneous nuclear safe to play with knives. Mol Cell 40: 179–204 ribonucleoprotein complex K protein. Mol Cell Biol 12: 164–171 Cimprich KA, Cortez D (2008) ATR: an essential regulator of Michael WM, Eder PS, Dreyfuss G (1977) The K nuclear shuttling genome integrity. Nat Rev Mol Cell Biol 9: 616–627 domain: a novel signal for nuclear import and nuclear export in Cremona CA, Sarangi P, Yang Y, Hang LE, Rahman S, Zhao X (2012) the hnRNP K protein. EMBO J 16: 3587–3598 Extensive DNA damage-induced sumoylation contributes to re- Moumen A, Masterson P, O’Connor MJ, Jackson SP (2005) hnRNP plication and repair and acts in addition to the mec1 checkpoint. K: an HDM2 target and transcriptional coactivator of p53 in Mol Cell 45: 422–432 response to DNA damage. Cell 123: 1065–1078 Dou H, Huang C, Singh M, Carpenter PB, Yeh ET (2010) Regulation Mukhopadhyay D, Dasso M (2007) Modification in reverse: the of DNA repair through deSUMOylation and SUMOylation of SUMO proteases. Trends Biochem Sci 32: 286–295 replication protein A complex. Mol Cell 39: 333–345 Niu G, Wright KL, Huang M, Song L, Haura E, Turkson J, Zhang S, Durocher D, Jackson SP (2001) DNA-PK, ATM and ATR as sensors of Wang T, Sinibaldi D, Coppola D, Heller R, Ellis LM, Karras J, DNA damage: variations on a theme? Curr Opin Cell Biol 13: 225–231 Bromberg J, Pardoll D, Jove R, Yu H (2002) Constitutive Stat3 Enge M, Bao W, Hedstrom E, Jackson SP, Moumen A, Selivanova G activity up-regulates VEGF expression and tumor angiogenesis. (2009) MDM2-dependent downregulation of p21 and hnRNP K Oncogene 21: 2000–2008 provides a switch between apoptosis and growth arrest induced Pelisch F, Pozzi B, Risso G, Munoz MJ, Srebrow A (2012) DNA by pharmacologically activated p53. Cancer Cell 15: 171–183 damage-induced heterogeneous nuclear ribonucleoprotein K Gareau JR, Lima CD (2010) The SUMO pathway: emerging mechan- SUMOylation regulates p53 transcriptional activation. J Biol isms that shape specificity, conjugation and recognition. Nat Rev Chem 287: 30789–30799 Mol Cell Biol 11: 861–871 Perry ME (2004) Mdm2 in the response to radiation. Mol Cancer Res Geiss-Friedlander R, Melchior F (2007) Concepts in sumoylation: a 2: 9–19 decade on. Nat Rev Mol Cell Biol 8: 947–956 Polo SE, Jackson SP (2011) Dynamics of DNA damage response Hay RT (2005) SUMO: a history of modification. Mol Cell 18: 1–12 proteins at DNA breaks: a focus on protein modifications. Genes Hoege C, Pfander B, Moldovan GL, Pyrowolakis G, Jentsch S (2002) Dev 25: 409–433 RAD6-dependent DNA repair is linked to modification of PCNA Rytinki MM, Kaikkonen S, Pehkonen P, Jaaskelainen T, Palvimo JJ by ubiquitin and SUMO. Nature 419: 135–141 (2009) PIAS proteins: pleiotropic interactors associated with Horn HF, Vousden KH (2007) Coping with stress: multiple ways to SUMO. Cell Mol Life Sci 66: 3029–3041 activate p53. Oncogene 26: 1306–1316 Sacher M, Pfander B, Hoege C, Jentsch S (2006) Control of Rad52 Jackson PK (2001) A new RING for SUMO: wrestling transcriptional recombination activity by double-strand break-induced SUMO responses into nuclear bodies with PIAS family E3 SUMO ligases. modification. Nat Cell Biol 8: 1284–1290 Genes Dev 15: 3053–3058 Shang Y, Hu X, DiRenzo J, Lazar MA, Brown M (2000) Cofactor Jang HD, Yoon K, Shin YJ, Kim J, Lee SY (2004) PIAS3 suppresses dynamics and sufficiency in estrogen receptor-regulated tran- NF-kappaB-mediated transcription by interacting with the p65/ scription. Cell 103: 843–852 RelA subunit. J Biol Chem 279: 24873–24880 Steinman RA, Johnson DE (2000) p21WAF1 prevents down-mod- Jepsen K, Hermanson O, Onami TM, Gleiberman AS, Lunyak V, ulation of the apoptotic inhibitor protein c-IAP1 and inhibits McEvilly RJ, Kurokawa R, Kumar V, Liu F, Seto E, Hedrick SM, leukemic apoptosis. Mol Med 6: 736–749 Mandel G, Glass CK, Rose DW, Rosenfeld MG (2000) Suzuki A, Tsutomi Y, Akahane K, Araki T, Miura M (1998) Combinatorial roles of the nuclear receptor corepressor in tran- Resistance to Fas-mediated apoptosis: activation of caspase 3 is scription and development. Cell 102: 753–763 regulated by cell cycle regulator p21WAF1 and IAP gene family Jiang M, Chiu SY, Hsu W (2011) SUMO-specific protease 2 in Mdm2- ILP. Oncogene 17: 931–939 mediated regulation of p53. Cell Death Differ 18: 1005–1015 Vassileva MT, Matunis MJ (2004) SUMO modification of hetero- Johnson ES (2004) Protein modification by SUMO. Annu Rev geneous nuclear ribonucleoproteins. Mol Cell Biol 24: 3623–3632 Biochem 73: 355–382 Vogelstein B, Lane D, Levine AJ (2000) Surfing the p53 network. Kerscher O, Felberbaum R, Hochstrasser M (2006) Modification of Nature 408: 307–310 proteins by ubiquitin and ubiquitin-like proteins. Annu Rev Cell Vousden KH, Lu X (2002) Live or let die: the cell’s response to p53. Dev Biol 22: 159–180 Nat Rev Cancer 2: 594–604 Kim KI, Baek SH, Chung CH (2002) Versatile protein tag, SUMO: its WeiD,LeX, Zheng L, Wang L, Frey JA,Gao AC,PengZ,Huang S, enzymology and biological function. J Cell Physiol 191: 257–268 Xiong HQ, Abbruzzese JL, Xie K (2003) Stat3 activation regulates the Kim ST, Lim DS, Canman CE, Kastan MB (1999) Substrate specifi- expression of vascular endothelial growth factor and human pan- cities and identification of putative substrates of ATM kinase creatic cancer angiogenesis and metastasis. Oncogene 22: 319–329 family members. J Biol Chem 274: 37538–37543 Yeh ET (2009) SUMOylation and De-SUMOylation: wrestling with Kluge A, Dabir S, Vlassenbroeck I, Eisenberg R, Dowlati A (2011) life’s processes. J Biol Chem 284: 8223–8227 Protein inhibitor of activated STAT3 expression in lung cancer. Zeng SX, Dai MS, Keller DM, Lu H (2002) SSRP1 functions as a co- Mol Oncol 5: 256–264 activator of the transcriptional activator p63. EMBO J 21: 5487–5497 4452 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

SUMOylation of hnRNP‐K is required for p53‐mediated cell‐cycle arrest in response to DNA damage

Download PDF
12 pages

Loading next page...
 
/lp/springer-journals/sumoylation-of-hnrnp-k-is-required-for-p53-mediated-cell-cycle-arrest-kqklFigNNG

References (61)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 2012
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1038/emboj.2012.293
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal (2012) 31, 4441–4452 & 2012 European Molecular Biology Organization All Rights Reserved 0261-4189/12 | | THE THE www.embojournal.org EMB EMB EMBO O O JO JOU URN R NAL AL SUMOylation of hnRNP-K is required for p53-mediated cell-cycle arrest in response to DNA damage Seong Won Lee, Moon Hee Lee, downstream targets, such as the p53 transcription factor and the checkpoint CHK1 and CHK2 kinases (Abraham, 2001, Jong Ho Park, Sung Hwan Kang, 2004; Ciccia and Elledge, 2010). This process in turn regulates Hee Min Yoo, Seung Hyun Ka, the functions of downstream effector proteins involved in Young Mi Oh, Young Joo Jeon* cell-cycle arrest, DNA repair, and/or apoptosis. A key and Chin Ha Chung* example is ATM- and ATR-mediated phosphorylation of Department of Biological Sciences, College of Natural Sciences, Seoul both p53 and HDM2, which impairs their interaction and National University, Seoul, Korea thereby prevents HDM2-mediated ubiquitination of p53 for degradation by proteasome, leading to stabilization and activation of p53 (Perry, 2004). Heterogeneous ribonucleoprotein-K (hnRNP-K) is nor- A major consequence of p53 activation in response to DNA mally ubiquitinated by HDM2 for proteasome-mediated damage is the induction of cell-cycle arrest (Vogelstein et al, degradation. Under DNA-damage conditions, hnRNP-K is 2000; Bartek and Lukas, 2001; Vousden and Lu, 2002; Horn transiently stabilized and serves as a transcriptional co- and Vousden, 2007) at the G1/S or G2/M phase. Cell-cycle activator of p53 for cell-cycle arrest. However, how the arrest at the G1/M phase is primarily achieved by expression stability and function of hnRNP-K is regulated remained of p53-downstream genes, such as p21, an inhibitor of cyclin- unknown. Here, we demonstrated that UV-induced dependent kinases (CDKs). Notably, p21 also acts as an anti- SUMOylation of hnRNP-K prevents its ubiquitination for apoptotic protein. This function of p21 is mediated by its stabilization. Using SUMOylation-defective mutant and ability to inhibit caspase-3 (Suzuki et al, 1998), stabilize the purified SUMOylated hnRNP-K, SUMOylation was shown anti-apoptotic cIAP1 (Steinman and Johnson, 2000), or to reduce hnRNP-K’s affinity to HDM2 with an increase in downregulate caspase-2 (Baptiste-Okoh et al, 2008). Thus, that to p53 for p21-mediated cell-cycle arrest. PIAS3 served p21 plays an important role in inhibiting apoptosis as well as as a small ubiquitin-related modifier (SUMO) E3 ligase for in cell-cycle arrest, allowing cells to repair damaged DNA and hnRNP-K in an ATR-dependent manner. During later per- prevent tumorigenesis. iods after UV exposure, however, SENP2 removed SUMO Small ubiquitin-related modifier (SUMO) is an ubiquitin-like from hnRNP-K for its destabilization and in turn for protein that is conjugated to a variety of cellular proteins. Like release from cell-cycle arrest. Consistent with the rise- ubiquitin, SUMO is conjugated to target proteins by a cascade and-fall of both SUMOylation and stability of hnRNP-K, enzyme system consisting of E1 activating enzyme (SAE1/ its ability to interact with PIAS3 was inversely correlated SAE2), E2 conjugating enzyme (Ubc9), and E3 ligases to that with SENP2 during the time course after UV (PIASs) (Kerscher et al, 2006; Capili and Lima, 2007; Rytinki exposure. These findings indicate that SUMO modification et al, 2009). Conjugated SUMO can be removed by a family plays a crucial role in the control of hnRNP-K’s function as of SUMO-specific proteases (SENPs) (Mukhopadhyay and a p53 co-activator in response to DNA damage by UV. Dasso, 2007; Yeh, 2009). This reversible SUMOylation process The EMBO Journal (2012) 31, 4441–4452. doi:10.1038/ participates in the control of diverse cellular processes, emboj.2012.293; Published online 23 October 2012 including transcription, nuclear transport, and signal Subject Categories: proteins; genome stability & dynamics transduction (Kim et al, 2002; Johnson, 2004; Hay, 2005; Keywords: HDM2; p21; PIAS3; SENP2; ubiquitin Geiss-Friedlander and Melchior, 2007; Gareau and Lima, 2010). Significantly, many proteins involved in DNA-damage response are modified by ubiquitin and/or SUMO, implicating the role of ubiquitination, SUMOylation, or both in the control Introduction of checkpoint responses and DNA-repair pathways (Hoege The p53 tumour suppressor plays a pivotal role in mainte- et al, 2002; Lee et al, 2006; Bergink and Jentsch, 2009; nance of genome integrity under cellular stresses, such as Altmannova et al,2010;Dou et al,2010;Polo and Jackson, DNA damage (Lane, 1992; Lakin and Jackson, 1999; Kruse 2011; Cremona et al, 2012). For example, Rad52, a mediator of and Gu, 2009; Levine and Oren, 2009). Upon DNA damage, homologous recombination in yeast, is SUMOylated in ATM, ATR, and DNA-PK are activated for phosphorylation of response to DNA damage, and this modification stabilizes Rad52 for its sustained function (Sacher et al, 2006). *Corresponding authors. YJ Jeon or CH Chung, Department of Heterogeneous ribonucleoprotein-K (hnRNP-K) is an RNA- Biological Sciences, College of Natural Sciences, Seoul National University, 56-1 Shillim-dong, Gwanak-gu, Seoul 151-742, Korea. binding protein that is associated with various cellular pro- Tel.: þ82 2 880 6693; Fax: þ82 2 871 9193; cesses, including chromatin remodelling, transcription, E-mail: [email protected] or [email protected] mRNA splicing, and translation (Matunis et al, 1992; Bomsztyk et al, 1997, 2004). Intriguingly, hnRNP-K was Received: 8 June 2012; accepted: 8 October 2012; published online: 23 October 2012 shown to be transiently stabilized and function as a &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4441 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 1 UV induces SUMOylation of hnRNP-K. (A) UV increases the cellular level of hnRNP-K. After exposure of HeLa cells to UV (10 J/m ), cell lysates were subjected to immunoblot with anti-hnRNP-K or anti-p53 antibody. The resulting gels were scanned using a densitometer, and the intensities of hnRNP-K bands were quantified by using ‘Image J’ program. The intensity of hnRNP-K seen before UV (i.e., 0 h) was expressed as 1.0 and the others as its relative values. (B) UV induces SUMOylation of hnRNP-K. After UV treatment, cell lysates were subjected to immunoprecipitation with anti-hnRNP-K antibody followed by immunoblot with anti-SUMO1 or anti-hnRNP-K antibody. (C) Modification of hnRNP-K by SUMO isoforms. Flag-tagged SUMO isoforms were expressed in HEK293T cells with Flag-Ubc9 and HisMax-hnRNP-K. After incubation with 10 mM MG132 for 4 h, cell lysates were subjected to pull down with NTA beads followed by immunoblot with anti-Flag or anti-Xpress antibody. Figure source data can be found with the Supplementary data. transcriptional co-activator of p53 in response to DNA SUMOylation could be induced under other DNA damage damage (Moumen et al, 2005). However, how the stability conditions. Both SUMOylation and stabilization of hnRNP-K and function of hnRNP-K is regulated remained unknown. were also induced by treatments with ionizing radiation (IR) Here, we showed that UV induces PIAS3-mediated hnRNP-K and doxorubicin, although the timing of their rise-and-fall SUMOylation, which increases hnRNP-K stability, interaction was significantly different from that induced by UV between hnRNP-K and p53, and p21 expression in an ATR- (Supplementary Figure S1). Thus, hnRNP-K SUMOylation dependent manner, leading to cell-cycle arrest. At later appears to be a common response to DNA damage for its periods after UV treatment, however, SENP2 reversed the stabilization. SUMOylation-mediated processes by removing SUMO from When SUMO isoforms were overexpressed with hnRNP-K, hnRNP-K, implicating the role of SENP2 in the release of SUMO1 was more efficiently conjugated to hnRNP-K than cells from cell-cycle arrest to resume normal growth after SUMO2 or SUMO3 (Figure 1C). Thus, further studies were DNA repair. These findings indicate that reversible SUMO performed only with SUMO1. Since two SUMOylated hnRNP- modification of hnRNP-K by PIAS3 and SENP2 plays a crucial K bands appeared under the overexpression conditions, two role in the control of hnRNP-K stability and thereby its Lys residues in the sequences closely matched to the con- function as a p53 co-activator in response to DNA damage sensus motif for SUMOylation (c-K-X-D/E) were substituted by UV. with Arg (Figure 2A). Replacement of Lys422 alone or together with Lys198 by Arg prevented hnRNP-K SUMOylation, whereas that of Lys198 alone did not Results (Figure 2B). Similar results were obtained by in vitro UV-induced SUMOylation increases the stability of SUMOylation assay using purified SAE1/SAE2 (E1), Ubc9 (E2), and SUMO1 (Figure 2C), indicating that Lys422 hnRNP-K serves as the major SUMOylation site of hnRNP-K. hnRNP-K has been identified as a candidate for SUMOylation Henceforth, the SUMOylation-defective mutant was referred by proteomic analysis (Li et al, 2004). Therefore, we first to as K422R. examined whether hnRNP-K could indeed be modified by SUMO and whether this modification is related with DNA We next examined whether UV-induced SUMOylation in- damage-induced stabilization of hnRNP-K. UV treatment led fluences hnRNP-K ubiquitination and in turn its stability. The level of ubiquitinated hnRNP-K was markedly reduced at 6 h to 2- to 3-fold increase in the level of hnRNP-K by 6 h and after UV and returned almost to the initial level at 18 h after declined thereafter (Figure 1A). Moreover, the level of UV (Figure 3A), indicating that the change in the level of SUMOylated hnRNP-K was markedly increased by 6 h and ubiquitinated hnRNP-K is inversely correlated with that of declined by 18 h after UV treatment and this change occurred in parallel with that of hnRNP-K level (Figure 1B), suggesting SUMO1-conjugated hnRNP-K. However, SUMOylation-defec- that UV-induced SUMOylation stabilizes hnRNP-K. Thus, tive K422R, unlike wild-type hnRNP-K, remained ubiquiti- further studies were performed at three time points; prior nated at 6 h after UV (Figure 3B). Consistently, UV treatment increased the stability of hnRNP-K, but not K422R (Figure 3C to, 6 h after, and 18 h after UV treatment, which were and D). In addition, MG132, a proteasome inhibitor, pre- henceforth referred to as before UV, 6 h after UV, and 18 h vented K422R destabilization under the same conditions. after UV, respectively. We also examined whether hnRNP-K 4442 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 2 Lys422 is the major SUMO1 acceptor site in hnRNP-K. (A) Potential SUMOylation sites in hnRNP-K. The Lys residues in the underlined sequences of hnRNP-K were substituted with Arg by site-directed mutagenesis. (B) K422R mutation ablates hnRNP-K SUMOylation in vivo. Flag-tagged hnRNP-K, K198R, K422R, and the double mutant (K198R/K422R) were overexpressed in HEK293T cells with HisMax- SUMO1 and Flag-Ubc9. Cell lysates were subjected to immunoprecipitation with anti-Flag antibody followed by immunoblot with anti-Flag or anti-SUMO1 antibody. (C) K422R mutation ablates hnRNP-K SUMOylation in vitro. SUMOylation was performed using purified proteins followed by immunoblot with anti-His antibody as described under ‘Materials and methods’. Figure source data can be found with the Supplementary data. These results indicate that UV-induced SUMOylation of overexpressed with HDM2, p53, and Ubc9. Co-expression of hnRNP-K is responsible for the increase in its stability. increasing amounts of SUMO1 (i.e., increasing the level of SUMOylated hnRNP-K) led to a gradual increase in the level of hnRNP-K-bound p53 concurrently with a decrease in that SUMOylation of hnRNP-K switches its interaction with of hnRNP-K-bound HDM2 (Figure 4J). On the other hand, the HDM2 to that with p53 level of K422R-bound p53 and HDM2 remained the same To elucidate the mechanism for SUMOylation-mediated sta- bilization of hnRNP-K, we first examined the effect of UV regardless of SUMO1 expression. Although the experiments treatment on the interaction of hnRNP-K with HDM2. The were performed under overexpression conditions, which could be non-physiological, these results strongly suggest level of HDM2 co-immunoprecipitated with hnRNP-K was that SUMOylated hnRNP-K preferentially binds p53 whereas significantly decreased at 6 h after UV and returned to the its unmodified form binds better to HDM2. Thus, UV-induced initial level at 18 h after UV (Figure 4A). Moreover, the ability SUMOylation of hnRNP-K appears to switch its interaction of hnRNP-K to bind HDM2 was markedly reduced at 6 h after UV, whereas that of K422R remained the same regardless of with HDM2 to that with p53. UV treatment (Figure 4B and C). In addition, purified Of note was the finding that without UV treatment, SUMOylated hnRNP-K (Figure 4D) showed a lower affinity hnRNP-K binds p53 better than K422R (see Figure 4G and H), whereas K422R binds HDM2 better than hnRNP-K to HDM2 than unmodified hnRNP-K (Figure 4E). Note that (see Figure 4B and C). However, in vitro binding assays the C-terminal region harbouring the SUMOylation site showed that purified K422R interacts with p53 or HDM2 as Lys422 overlaps with that for HDM2 binding (see below). well as wild-type hnRNP-K (Supplementary Figure S2A These results indicate that UV-induced SUMOylation of hnRNP-K interferes with its interaction with HDM2, leading and B), indicating that the K-to-R mutation itself has no effect to hnRNP-K stabilization. on the binding affinity of hnRNP-K to p53 or HDM2. Since endogenous hnRNP-K can be SUMOylated in the absence of We next examined whether UV-induced SUMOylation also UV although to a basal level (see Figure 1B), it appeared that influences the interaction of hnRNP-K with p53. In contrast overexpression of hnRNP-K (i.e., elevation of the substrate to HDM2, the amount of p53 co-immunoprecipitated with concentration for SUMOylation) increases the level of hnRNP-K was significantly increased at 6 h after UV and returned almost to the initial level at 18 h after UV SUMOylated hnRNP-K and this increase alters the binding (Figure 4F). Moreover, the ability of hnRNP-K to bind p53 affinity of hnRNP-K to p53 and HDM2. Indeed, increased expression of hnRNP-K led to an increase in the level of was markedly increased at 6 h after UV, whereas that of SUMOylated hnRNP-K in the absence of UV treatment K422R remained decreased regardless of UV treatment (Supplementary Figure S2C). Moreover, when hnRNP-K (Figure 4G and H). In addition, purified SUMOylated SUMOylation was prevented by knockdown of Ubc9 by hnRNP-K showed a much higher affinity to p53 than unmodified hnRNP-K (Figure 4I). These results indicate that using Ubc9-specific shRNA (shUbc9), both hnRNP-K and UV-induced SUMOylation of hnRNP-K promotes its interac- K422R bound to p53 or HDM2 to similar extents tion with p53. (Supplementary Figure S3). These results indicate that changes in the binding affinity of hnRNP-K to p53 or HDM2 To confirm whether SUMOylation of hnRNP-K is in the absence of UV treatment are due to an increase in the responsible for the alterations in its affinity to HDM2 and level of SUMOylated hnRNP-K upon its overexpression. p53 under in vivo conditions, hnRNP-K and K422R were &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4443 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 3 UV-induced SUMOylation increases the stability of hnRNP-K. (A) UV blocks hnRNP-K ubiquitination. After exposure to UV, HeLa cells were incubated with 10 mM MG132 for 4 h. Cell lysates were subjected to immunoprecipitation with anti-ubiquitin, anti-hnRNP-K, or anti-SUMO1 antibody followed by immunoblot analysis. (B) SUMOylation prevents hnRNP-K ubiquitination. After exposure to UV, cells overexpressing HisMax-tagged hnRNP-K (Wt) or K422R (KR) were incubated for 2 h and then treated with MG132 for the next 4 h. Cell lysates were subjected to pull down with NTA beads followed by immunoblot analysis. (C) SUMOylation increases the hnRNP-K stability. Cells overexpressing Flag-tagged hnRNP-K (Wt) or K422R (KR) were treated with 200 mg/ml of cycloheximide. After exposure to UV, they were incubated with and without MG132 followed by immunoblot with anti-Flag antibody. (D) Band intensities in (C) were quantified by using a densitometer. The data represent the mean s.d. of three independent experiments. Figure source data can be found with the Supplementary data. SUMOylation of hnRNP-K is required for its function as p53 to the p21 promoter site and this increase could be further a p53 co-activator enhanced by hnRNP-K overexpression, but not by that of To determine whether UV-induced SUMOylation of hnRNP-K K422R (Figure 5E). These results indicate that UV-induced influences its co-activator function, p53 transactivity was hnRNP-K SUMOylation promotes p53 transactivity and measured by using two reporter vectors, PG13-Luc and thereby p21 expression. p21-Luc. In both cases, UV treatment increased the luciferase Of note was the finding that hnRNP-K overexpression leads activity and this increase was further enhanced by over- to an increase in the level of endogenous p53 in the absence expression of hnRNP-K, but not by that of K422R of UV treatment (see Figure 5D), raising a possibility that (Figure 5A and B). Under the same conditions, both mRNA overexpressed hnRNP-K may stabilize p53 although it has and protein levels of p21 were increased and this increase was been shown that hnRNP-K knockdown does not affect p53 further enhanced by overexpression of hnRNP-K, but not by stability (Moumen et al, 2005). However, expression of that of K422R (Figure 5C and D). hnRNP-K overexpression increasing amounts of hnRNP-K showed little or no effect without UV treatment also increased p53 transactivity, as it on HDM2-mediated p53 ubiquitination or HDM2 auto- could increase the SUMOylated hnRNP-K level. Moreover, ubiquitination, indicating that hnRNP-K has no effect on chromatin immunoprecipitation (ChIP) analysis revealed that the stability of p53 (Supplementary Figure S4). Since UV treatment increased recruitment of both hnRNP-K and hnRNP-K overexpression causes an increase in the level of 4444 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 4 hnRNP-K SUMOylation switches its interaction with HDM2 to that with p53. (A) UV inhibits the interaction of hnRNP-K with HDM2. After UV treatment, HeLa cells were incubated for the indicated periods. Cell lysates were subjected to immunoprecipitation with anti-hnRNP-K antibody followed by immunoblot with anti-HDM2 and anti-hnRNP-K antibodies. (B, C) SUMOylation inhibits the interaction of hnRNP-K with HDM2. HDM2 was overexpressed in cells with Flag-tagged hnRNP-K or K422R. After exposure to UV, cells were incubated for 6 h. Cell lysates were subjected to immunoprecipitation with anti-HDM2 (B) or anti-Flag antibody (C). (D) Purification of His- SUMO1-conjugated GST-hnRNP-K. SUMOylated hnRNP-K proteins eluted from NTA-agarose column were subjected to SDS–PAGE followed by staining with Coomassie blue R-250. Fractions under the bar were pooled for further use. (E) SUMOylation reduces the affinity of hnRNP-K to HDM2. Purified His-HDM2 was incubated with GST-hnRNP-K-His-SUMO1 or GST-hnRNP-K followed by immunoprecipitation with anti- hnRNP-K antibody. (F) UV promotes the interaction of hnRNP-K with p53. Experiments were performed as in (A), except that anti-p53 antibody was used in place of anti-HDM2 antibody. (G, H) SUMOylation increases the affinity of hnRNP-K to p53. Myc-p53 was overexpressed in cells with Flag-tagged hnRNP-K or K422R. After exposure to UV, cells were incubated for 6 h. Cell lysates were subjected to immunoprecipitation with anti-Myc (G) or anti-Flag antibody (H). (I) SUMOylated hnRNP-K shows higher affinity to p53. Experiments were done as in (E), except that His-p53 was used in place of His-HDM2. (J) SUMOylation inversely affects the binding of hnRNP-K to HDM2 and p53. HisMax-tagged hnRNP-K (Wt) and K422R (KR) were overexpressed in cells with Myc-Ubc9, HA-p53, HDM2, and increasing amounts of Flag-SUMO1. Cell lysates were subjected to pull down with NTA beads followed by immunoblot analysis. Note that 10 mM MG132 was treated 4 h before cell lysis in (A–C), (F–H), and (J). Figure source data can be found with the Supplementary data. SUMOylated hnRNP-K even in the absence of UV (see overexpressed hnRNP-K-mediated increase in endogenous Supplementary Figure S2) and since p53 is known to posi- p53 level without UV treatment is due to the ability of tively regulate its own expression, it appears likely that the SUMOylated hnRNP-K in promotion of p53 expression. &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4445 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 5 SUMOylation of hnRNP-K is required for its function as a p53 co-activator. (A, B) SUMOylation of hnRNP-K promotes p53 transactivity. HeLa cells overexpressing Flag-tagged hnRNP-K or K422R were transfected with PG13-Luc (A)or P21-Luc (B). After exposure to UV, cells were incubated for 6 h. Cell lysates were assayed for the luciferase activity. The activity seen without hnRNP-K overexpression and UV treatment was expressed as 1.0 and the others were as its relative values. The data represent the mean s.d. of three experiments. (C) SUMOylation of hnRNP-K increases the level of p21 transcripts. Total RNAs prepared from the same cells used in (A) were subjected to RT–PCR to determine p21 mRNA levels. (D) SUMOylation of hnRNP-K promotes p21 expression. Cell lysates prepared as in (A) were subjected to immunoblot with anti-p53, anti-p21, or anti-hnRNP-K antibody. (E) SUMOylation of hnRNP-K promotes recruitment of both hnRNP-K and p53 to the p21 promoter. Cells prepared as in (A) were subjected to ChIP assay by using anti-hnRNP-K or anti-p53 antibody. Precipitated DNAs were subjected to PCR with primers covering the p53-response element in the p21 gene. Figure source data can be found with the Supplementary data. PIAS3 and SENP2 counteract on SUMO modification of were generated and subjected to pull-down analysis hnRNP-K (Supplementary Figure S6A–D). Both p53 and SENP2 To identify hnRNP-K-specific SUMO E3 ligase, each of PIAS1-4 bound to the same N-terminal region of hnRNP-K was overexpressed with hnRNP-K. Among them, PIAS3 spe- (Supplementary Figure S6E), suggesting that p53 and cifically interacted with hnRNP-K (Supplementary Figure S5A) SENP2 could compete with each other for binding to and promoted its SUMOylation (Figure 6A). Furthermore, hnRNP-K. While PIAS3 interacted with the middle region of PIAS3 knockdown by shPIAS3 prevented not only hnRNP-K hnRNP-K, HDM2 bound to its C-terminal region, which SUMOylation but also p21 expression (Figure 6B), suggesting includes the SUMO-conjugation site Lys422. The latter data that PIAS3-mediated SUMOylation of hnRNP-K is required for are consistent with the finding that SUMOylated hnRNP-K its function as a p53 co-activator. Interestingly, the amount of shows a lower affinity to HDM2 than its unmodified form (see PIAS3 co-immunoprecipitated with hnRNP-K was significantly Figure 4C and E). To identify hnRNP-K-binding regions within increased at 6 h after UVand returned almost to the initial level HDM2, p53, SENP2, and PIAS3, deletions of each protein at 18 h-after-UV (Figure 6C). Thus, it appears that UV-mediated were generated (Supplementary Figure S7). hnRNP-K bound increase in hnRNP-K SUMOylation is due to an increase in the to the C-terminal regions of p53, SENP2, and PIAS3, while it affinity of hnRNP-K to PIAS3. is interacted with the middle region of HDM2. A map for the We next attempted to identify hnRNP-K-specific interaction between hnRNP-K and HDM2, p53, SENP2, or deSUMOylating enzyme. Among the enzymes tested, over- PIAS3 was shown in Figure 7. expressed SENP1, SENP2, SENP6, and mouse SUSP4 inter- acted with hnRNP-K (Supplementary Figure S5B). Without Effect of hnRNP-K SUMOylation on its subcellular overexpression, however, only SENP2 interacted with localization hnRNP-K and this interaction was markedly decreased at Previous studies have suggested that SUMOylation of hnRNP- 6 h after UV and recovered at 18 h after UV (Figure 6D K is involved in its nucleocytoplasmic transport (Vassileva and E). Moreover, SENP2, but not its catalytically inactive and Matunis, 2004). Therefore, we examined whether UV- mutant (in which the active site Cys548 was replaced by Ser), induced SUMOylation influences the subcellular localization removed SUMO from hnRNP-K (Figure 6F), whereas SENP2 of hnRNP-K. Immunocytochemical analysis showed that in knockdown by shSENP2 promoted hnRNP-K SUMOylation the absence of UV treatment, overexpressed hnRNP-K resided (Figure 6G). Notably, without UV treatment SENP2 in both the nucleus and the cytoplasm in B30% of cells as knockdown significantly increased the level of SUMOylated well as exclusively in the nucleus in the remaining cells hnRNP-K, suggesting that endogenous SENP2 rapidly (Supplementary Figure S8A). In its presence, however, the deSUMOylates hnRNP-K under unstressed conditions. entire hnRNP-K proteins were localized exclusively in the Collectively, these results demonstrate that PIAS3 and SENP2 nucleus. In contrast, K422R was localized in both the nucleus antagonistically regulate SUMO modification and stability of and the cytoplasm in B40% of cells regardless of UV treat- hnRNP-K during the time course after exposure to UV. ment, suggesting that SUMOylation is involved in the nuclear To map the regions within hnRNP-K for binding of localization of hnRNP-K. Therefore, we next examined HDM2, p53, SENP2, and PIAS3, deletions of hnRNP-K whether the nuclear localization of hnRNP-K could be 4446 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 6 PIAS3 and SENP2 antagonistically regulate hnRNP-K SUMOylation. (A) PIAS3 promotes hnRNP-K SUMOylation. HisMax-hnRNP-K was overexpressed in HEK293T cells with Flag-SUMO1, Flag-Ubc9, and Myc-PIAS3. Cell lysates were subjected to pull down with NTA beads followed by immunoblot with anti-SUMO1 or anti-Xpress antibody. (B) PIAS3 knockdown blocks hnRNP-K SUMOylation. HeLa cells transfected with shNS or shPIAS3 were exposed to UV and then incubated for 6 h. Cell lysates were subjected to immunoprecipitation with anti-hnRNP-K antibody followed by immunoblot with anti-SUMO1 or anti-hnRNP-K antibody. (C) UV promotes the interaction of hnRNP-K with PIAS3. After exposure to UV, cells were incubated for the indicated periods. Cell lysates were subjected to immunoprecipitation with anti-hnRNP-K antibody followed by immunoblot with anti-PIAS3 or anti-hnRNP-K antibody. (D, E) UV inhibits the interaction of hnRNP-K with SENP2. Cells treated with UV were subjected to immunoprecipitation with anti-hnRNP-K (D)or anti-SENP2 antibody (E). Note that MG132 was treated 4 h before cell lysis in (C–E). (F) SENP2 deSUMOylates hnRNP-K. HisMax-hnRNP-K was overexpressed in HEK293T cells with Flag-SUMO1, Flag-Ubc9, and Myc-tagged SENP2 (Wt) or its catalytically inactive form (CS). Cell lysates were subjected to pull down with NTA beads followed by immunoblot with anti-SUMO1 or anti-Xpress antibody. (G) SENP2 knockdown promotes hnRNP-K SUMOylation. HeLa cells transfected with shNS or shSENP2 were exposed to UV and incubated for 6 h. Cell lysates were then treated as in (B). Figure source data can be found with the Supplementary data. Figure 7 Map for the interaction between hnRNP-K and p53, HDM2, SENP2, or PIAS3. The data obtained from Supplementary Figures S6 and S7 were summarized. In p53: TAD, transcription activation domain; DBD, DNA-binding domain; TET, tetramerization domain; REG, regulatory domain. In HDM2: p53BP, p53 binding domain; AD, acidic domain; ZF, zinc finger; RING, really interesting gene. In SENP2: NLD, nuclear localization domain; NES, nuclear export signal; SIM, SUMO-interacting motif; CD, catalytic domain. In PIAS3: SAP, SAF-A/B, Acinus and PIAS; PINIT, Pro-Ile-Asn-Ile-Thr; S/T, Ser/Thr-rich. blocked by knockdown of PIAS3. Depletion of PIAS3 hnRNP-K could also be prevented by overexpression of prevented UV-induced nuclear localization of hnRNP-K wild-type SENP2, but not by its catalytically inactive mutant (Supplementary Figure S8B). The nuclear localization of in which the active site Cys548 was replaced by Ser &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4447 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al (Supplementary Figure S8C). These results again suggest that by supplement of shhnRNP-K-insensitive hnRNP-K, but not SUMOylation is involved in the nuclear localization of by that of shhnRNP-K-insensitive K422R (Figure 8F). These hnRNP-K. results indicate that hnRNP-K SUMOylation is required for To confirm this finding, we examined whether the localiza- UV-induced cell-cycle arrest. Knockdown of SENP2 enhanced tion of endogenous hnRNP-K could be influenced by PIAS3 UV-mediated increase in cell fractions in G1 phase, whereas knockdown in the presence and absence of UV. In contrast to that of PIAS3 ablated it (Figure 8G). Knockdown of SENP2 the data obtained by hnRNP-K overexpression, neither UV together with hnRNP-K also prevented UV-induced cell-cycle treatment nor PIAS3 depletion showed any effect on the arrest. Figure 8H shows that PIAS3 knockdown decreases the nuclear localization of hnRNP-K (Supplementary Figure S9). expression of p21 as well as of p53 whereas SENP2 knock- Noteworthy, however, were the findings that in cells having down increases it, and this increase could be ablated by overexpressed hnRNP-K in both the nucleus and the cyto- simultaneous knockdown of hnRNP-K. Collectively, these plasm, the portion of hnRNP-K located in the cytoplasm is results indicate that reversible SUMO modification of much lower than that in the nucleus and that B60% of hnRNP-K by SENP2 and PIAS3 plays a key role in the control SUMOylation-defective K422R mutant resided exclusively in of p21-mediated cell-cycle arrest in response to UV damage. the nucleus regardless of UV treatment (see Supplementary UV-induced DNA damage response is at least in part Figure S8A). Furthermore, overexpression of SENP2 did not mediated by ATR kinase, which phosphorylates downstream alter the population of cells that have hnRNP-K exclusively in targets, such as p53 and CHK1 (Durocher and Jackson, 2001; the nucleus, although it can prevent the UV-induced nuclear Cimprich and Cortez, 2008). Treatment with caffeine, an localization of hnRNP-K in the cytoplasm (see Supplementary inhibitor of ATR (in addition to ATM), and knockdown of Figure S8C). In addition, endogenous hnRNP-K is known to ATR by shATR abrogated not only UV-induced hnRNP-K predominantly localize in the nucleus, and this nuclear SUMOylation and but also the increased interaction of localization is mediated mainly by the nuclear shuttling PIAS3 with hnRNP-K at 6 h after UV (Figure 9A and B). sequence (KNS) and in part by the NLS sequence in They also resulted in sustained interaction of hnRNP-K with hnRNP-K (Michael et al, 1977). Thus, it appears likely that SENP2. These results indicate that UV-induced hnRNP-K SUMOylation plays only a minor, auxiliary role in the nuclear SUMOylation, which is essential for hnRNP-K’s function as localization of hnRNP-K. a p53 co-activator, is ATR dependent. UV-induced hnRNP-K SUMOylation is required for Discussion cell-cycle arrest To determine whether SENP2 is involved in the control of Based on the findings in this study, we propose a model for the hnRNP-K’s co-activator function by altering its SUMOylation role of hnRNP-K SUMOylation in UV-induced cell-cycle arrest state, PG13-Luc and p21-Luc were again used for assaying p53 (Figure 9C). Under unstressed conditions, SENP2 removes transactivity. In both cases, UV treatment increased p53 trans- SUMO from SUMO-conjugated hnRNP-K, if there is any, allow- activity and this increase was further enhanced by SENP2 ing hnRNP-K to bind HDM2 with high affinity for its ubiquiti- knockdown (Figure 8A and B; Supplementary Figure S10A). nation and subsequent degradation by proteasome. Although Without UV treatment, SENP2 knockdown moderately pro- not shown, HDM2 also promotes proteasome-mediated degra- moted p53 transactivity, since it increases the level of dation of p53. Upon exposure to UV, PIAS3 binds and ligates SUMOylated hnRNP-K (see Figure 6G). These stimulatory SUMO to hnRNP-K, allowing p53 to bind SUMOylated hnRNP- effects of SENP2 knockdown on p53 transactivity were ablated K with high affinity and recruitment of their complex to the p21 promoter. Thus, hnRNP-K SUMOylation by PIAS3 serves as a by simultaneous knockdown of hnRNP-K by shhnRNP-K, indicating that the observed effects are specific to hnRNP-K. critical switch for shifting the interaction of hnRNP-K with These results indicate that SENP2 negatively regulates the HDM2 to that with p53 for its function as a transcriptional function of hnRNP-K as a p53 co-activator. On the other co-activator of p53 and in turn for p21 expression and cell-cycle hand, knockdown of PIAS3 alone or together with hnRNP-K arrest in response to DNA damage by UV. completely abrogated UV-induced p53 transactivity (Figure 8C Of interest was the finding that the level of SUMOylated and D; Supplementary Figure S10B), indicating that PIAS3 hnRNP-K increases and then declines during the time course positively regulates the co-activator function of hnRNP-K. after exposure of cells to UV. This phenomenon is apparently Collectively, these results indicate that SENP2 and PIAS3 antag- mediated by a rise-and-fall of hnRNP-K’s ability to interact onistically regulate the role of hnRNP-K as a p53 co-activator. with PIAS3 and its inversed ability to bind SENP2 during the We next examined whether hnRNP-K SUMOylation is re- same time course. This reversible SUMOylation process that quired for UV-induced cell-cycle arrest upon flow cytometry. should occur in conjunction with the p53-HDM2 feedback UV treatment increased cell fractions in G1 phase and this loop is of importance for cells to escape from cell-cycle arrest increase was blocked by hnRNP-K knockdown (Figure 8E; and to resume normal growth after the repair of damaged Supplementary Figure S11). Supplement of shhnRNP-K-insen- DNA. However, it remains unknown how the binding ability sitive hnRNP-K to cells that had been depleted of endogenous of hnRNP-K is shifted to PIAS3 and then to SENP2 after UV hnRNP-K, but not that of shhnRNP-K-insensitive K422R, treatment. Since ATR knockdown prevents the interaction of restored accumulation of G1-phase cells. To determine hnRNP-K with PIAS3 but promotes that with SENP2 and whether the cell-cycle arrest is mediated by p53-induced since UV does not affect the expression of either PIAS3 or expression of p21, the same cells used for flow cytometry SENP2, it seems possible that ATR-mediated phosphorylation were subjected to immunoblot analysis. Knockdown of of hnRNP-K followed by dephosphorylation by an unknown hnRNP-K led to a marked decrease in the expression of p21 protein phosphatase(s) might change the affinity of hnRNP-K as well as in that of p53, and this decrease could be reversed to PIAS3 and SENP2. Notably, hnRNP-K has the SQ and TQ 4448 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 8 PIAS3 and SENP2 antagonistically regulate cell-cycle arrest. (A, B) SENP2 knockdown promotes p53 transactivity. HeLa cells transfected with shNS or shSENP2 alone or together with shhnRNP-K were incubated for 48 h. They were then transfected with PG13-Luc (A)or P21-Luc (B) and further incubated for the next 24 h. After exposure to UV, cells were incubated for 6 h. Cell lysates were assayed for luciferase. The enzyme activity seen in cells transfected with shNS only but without UV treatment was expressed as 1.0 and the others were as its relative values. (C, D) PIAS3 knockdown ablates p53 transactivity. Experiments were performed as above, except that cells were transfected with shPIAS3 in place of shSENP2. (E, F) SUMOylation of hnRNP-K is required for p21-mediated cell-cycle arrest. Cells transfected with shNS or shhnRNP-K were complemented with shhnRNP-K-insensitive Flag-tagged hnRNP-K or K422R. After exposure to UV, they were incubated for 6 h followed by flow cytometry (E) or immunoblot analysis (F). (G, H) PIAS3 and SENP2 inversely regulate cell-cycle arrest. Cells were transfected with shNS, shSENP2, or shPIAS3 alone or together with shhnRNP-K. After exposure to UV, cells were incubated for 6 h followed by flow cytometry (G) or immunoblot analysis (H). The data in (A–E) and (G) represent the mean s.d. of four experiments. Figure source data can be found with the Supplementary data. motifs that can be phosphorylated by ATR (Kim et al, 1999). PIAS3 serves as an endogenous protein inhibitor of acti- However, replacement of the Ser and Thr residues by Ala or vated signal transducers and activators of transcription 3 Glu showed little or no effect on UV-induced SUMOylation of (STAT3), in addition to its role as a SUMO E3 ligase (Chung hnRNP-K. Nonetheless, we could not exclude a possibility et al, 1997; Jackson, 2001; Jang et al, 2004). The STAT3 that kinases downstream of ATR, such as CHK1, or other protein, which promotes cell-cycle progression and inhibits kinase(s) and unknown phosphatases may be involved in apoptosis, has been implicated in the pathogenesis of reversible phosphorylation of hnRNP-K and in turn in the various human cancers (Niu et al, 2002; Wei et al, 2003; control of the affinity of hnRNP-K to PIAS3 and SENP2. Levy and Inghirami, 2006). Interestingly, PIAS3 expression is &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4449 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Figure 9 SUMOylation of hnRNP-K is ATR dependent. (A) Caffeine inhibits hnRNP-K SUMOylation. After exposure of HeLa cells to UV, they were incubated with and without 5 mM caffeine for the indicated periods. Cell lysates were subjected to immunoprecipitation with anti-hnRNP- K antibody followed by immunoblot with anti-SUMO1, anti-SENP2, anti-PIAS3, or anti-hnRNP-K antibody. (B) ATR knockdown prevents hnRNP-K SUMOylation. Cells transfected with shNS or shATR were incubated for 48 h. After exposure to UV, cells were incubated for the indicated periods. Cell lysates were subjected to immunoprecipitation as in (A). Note that MG132 was treated 4 h before cell lysis in (A, B). (C) A model for the role of hnRNP-K SUMOylation in UV-induced cell-cycle arrest. Figure source data can be found with the Supplementary data. downregulated in several cancers, such as human gastric p53 and induction of tumour cell apoptosis) releases HDM2 carcinoma, glioblastoma, and squamous cell carcinoma of from p53 and the freed HDM2 molecules promote protea- the lung (Brantley et al, 2008; Kluge et al, 2011; Liu et al, some-mediated degradation of hnRNP-K, which impairs its 2011). Therefore, it has been suggested that loss or reduction co-activator function in p53-mediated p21 expression (Enge of PIAS3 expression contributes to enhanced STAT3 et al, 2009). HDM2 also directly promotes p21 degradation by transcriptional activity, leading to aberrant cell proliferation proteasome, eliminating anti-apoptotic function of p21 and and tumorigenesis. Here, we showed that PIAS3 promotes thus switching towards induction of apoptosis. Deregulated hnRNP-K SUMOylation and thereby p53-mediated cell-cycle cell-cycle progression such as by preventing p21-mediated arrest. Thus, PIAS3 might exert its anti-tumorigenic function cell-cycle arrest could also evoke tumorigenesis. Therefore, in both E3 ligase activity-dependent and -independent small molecules that specifically inhibit SENP2 could be used manners by promoting cell-cycle arrest. as therapeutic drugs against cancers, as they should prevent Targeted disruption of SENP2 in mice was shown to impair hnRNP-K deSUMOylation and in turn induce p53 activation cell-cycle progression at the G1/S phase, leading to abnorm- for p21 expression, leading to cell-cycle arrest. alities in trophoblast proliferation and differentiation (Chiu While our work was under revision, another study reported et al, 2008). During trophoblast development, SENP2 removes that DNA damage induces hnRNP-K SUMOylation, which in SUMO from Mdm2 and thereby promotes Mdm2-mediated turn enhances the transcriptional activity of p53 (Pelisch et al, ubiquitination of p53 and its subsequent degradation by 2012). However, the major differences between their and our proteasome, allowing cell-cycle progression. Disruption of works are the effect of SUMOylation on the stability of hnRNP-K the SENP2 gene, however, results in cytoplasmic localization and the identity of hnRNP-K-specific SUMO E3 ligase. While they of Mdm2 and in turn in p53 stabilization in the nucleus, concluded that SUMOylation does not alter hnRNP-K stability leading to p53-mediated cell-cycle arrest. On the other hand, and that Pc2 acts as an E3 ligase, we found that the same overexpression of SENP2 makes cells resistant to apoptosis modification leads to hnRNP-K stabilization and that PIAS3 induced by genotoxic stress, such as doxorubicin treatment, serves as an hnRNP-K-specific ligase. These differences might indicating that SENP2 plays a critical role in the control of cell- be due to the use of a single time point in analysing the effect of cycle progression (Jiang et al, 2011). Here, we showed that DNA damage on alterations in the level of hnRNP-K and to the SENP2 knockdown increases hnRNP-K SUMOylation, its use of overexpression system in analysing the role of Pc2 in interaction with p53, and consequently its co-activator hnRNP-K SUMOylation rather than that of RNA interference function in expression of p53-downstream genes, such as system. However, we could not exclude a possibility that diffe- p21, for cell-cycle arrest. In addition, it has been shown that rent SUMO E3 ligases may act on hnRNP-K under different DNA under unstressed conditions hnRNP-K is ubiquitinated by damage conditions, since their work mainly used doxorubicin as HDM2 for degradation by proteasome (Moumen et al, a DNA damaging agent whereas ours utilized UV. 2005). Thus, it appears that SENP2 could regulate cell-cycle progression by targeting two different substrates: one by deSUMOylating HDM2 for HDM2-mediated degradation of Materials and methods p53 and the other by deSUMOylating hnRNP-K for HDM2- Plasmids and antibodies mediated degradation of hnRNP-K, which ablates its function hnRNP-K cDNA was isolated from a cDNA library of HeLa cells, and cloned into pcDNA-HisMax and pCMV2-Flag. It was also cloned as a p53 co-activator. into pET-32b and pGEX-4T3 for bacterial expression. shRNAs were The p21 protein can function as an anti-apoptotic protein as purchased from Open Biosystems. Target sequences for shRNAs are well as an inhibitor of CDKs for cell-cycle arrest. Recently, it 0 0 0 as follows: shhnRNP-K, 5 -ACGATGAAACCTATGATTA-3 ; shSENP2, 5 - 0 0 was shown that an anti-cancer drug, RITA (reactivation of CCCACAGGATGAAATCCTA-3;shPIAS3, 5 -GCTGTCGGTCAGACAT 4450 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al 0 0 CATTT-3 . Antibodies against Myc (9E10), p53 (DO-1), p21 (C-19), instructions. Primers used in PCR for p21 were 5 -CTTTGTCACCGA 0 0 0 hnRNP-K (D-6), GAPDH (2D4A7), HDM2 (SMP14), Ub (A-5), Ubc9 GACACCAC-3 and 5 -GGCGTTTGGAGTGGTAGAAA-3 . (N-15), and GST (Z-5) were purchased from Santa Cruz (Santa Cruz, Luciferase assays CA, USA). Anti-Flag M2 (Sigma-Aldrich), anti-Xpress, anti-SUMO-1 HeLa cells transfected with pcDNA-b-Gal and PG13-Luc or p21-Luc (Invitrogen), anti-His (BD Biosciences), anti-SENP1, anti-SENP2, were incubated for 48 h. After UV treatment, cells were cultured for anti-SENP6 (Abgent), and anti-PIAS3 (Cell Signaling) antibodies 6 h, harvested, and assayed for luciferase. The enzyme activity was were also used. measured in a luminometer and normalized by b-galactosidase Cell culture and transfection expression with a luciferase system (Promega). HEK293T and HeLa cells were grown at 371C in DMEM supplemen- ted with 100 units/ml penicillin, 1 mg/ml streptomycin, and 10% ChIP assay FBS. MRC5 cells were cultured as above except that the use of Assays were conducted with an average size of sheared fragments MEM in place of DMEM. All transfections were carried out using of about 300–1000 bps as described (Jepsen et al, 2000; Shang et al, Metafectene reagent (Biontex) and jetPEIt DNA Transfection 2000). For PCR, 1 ml from 50 ml DNA extraction and 25–30 cycles of Reagent (Polyplus-transfection). amplification were used. Primers used in PCR of p21 promoter 0 0 0 sequence were 5 -GTGGCTCTGATTGGCTTTCTG-3 and 5 -CTGAAA ACAGGCAGCCCAAGG-3 (Zeng et al, 2002). Assays for SUMO modification HisMax-hnRNP-K, Flag-SUMO1, and Flag-Ubc9 were overexpressed in HEK293T cells with or without Myc-tagged SENP2 or PIAS3. After Purification of recombinant SUMOylated hnRNP-K culturing for 36 h, cells were lysed by boiling for 10 min in 150 mM For production of SUMOylated hnRNP-K, BL21(DE3) cells were Tris–HCl (pH 8), 5% SDS, and 30% glycerol. Cell lysates were diluted transformed with pGEX4T3-hnRNP-K and pT-E1/E2/His-SUMO1. 20-fold with buffer A consisting of 20 mM Tris–HCl (pH 8.0), 150 mM BL21 colonies carrying both plasmids were selected as described NaCl, 10 mM imidazole, 1% Triton X-100, 1 protease inhibitor (Zeng et al, 2002). Extracts (10 mg) from the cells were loaded onto 2þ cocktail (Roche), and 2 mM NEM. After incubating them with Ni - a glutathione-Sepharose 4B column, and proteins bound to the NTA-agarose for 2 h at 41C, the resins were collected, washed with column were eluted with PBS containing 50 mM glutathione. After buffer A containing 20 mM imidazole, and boiled in SDS-sampling dialysis against buffer C consisting of NaH PO /Na HPO (pH 8), 2 4 2 4 buffer. Supernatants were subjected to SDS–PAGE followed by immu- 0.5 M NaCl, 50 mM imidazole, 1% Triton X-100, and 2 mM noblot analysis. For assaying SUMOylation of endogenous hnRNP-K, 2-mercaptoethanol, proteins were loaded onto a NTA-agarose HeLa cells without any overexpression were lysed as above. Cell column. Bound proteins (i.e., GST-hnRNP-K-His-SUMO1) were lysates were diluted 20-fold with buffer B consisting of 20 mM Tris– eluted with buffer C containing 200 mM imidazole. HCl (pH 8.0), 150 mM NaCl, 0.1 mM EDTA, 0.2% Triton X-100, 1 protease inhibitor cocktail, and 2 mM NEM. The samples were incu- Immunocytochemistry bated with anti-hnRNP-K antibody for 2 h at 41C and then with protein- HeLa cells were grown on coverslips. After transfection, they were A-Sepharose for the next 2 h. The resins were collected, washed with fixed by incubation with 3.7% paraformaldehyde in PBS for 10 min. buffer B containing 1% Triton X-100, and boiled. Supernatants were Cells were washed three times with PBS containing 0.1% Triton subjectedtoSDS–PAGEfollowedbyimmunoblotanalysis. X-100, and permeabilized with 0.5% Triton X-100 in PBS for 5 min. For in vitro SUMOylation assay, purified His-hnRNP-K (2 mg), After blocking with 3% BSA in PBS for 30 min, cells were incubated SUMO1 (5 mg), SAE1/SAE2 (1.5 mg), and Ubc9 (5 mg) were incubated for 1 h with appropriate antibodies. After washing with PBS con- with an ATP-regenerating system consisting of 50 mM Tris–HCl (pH taining 0.1% Triton X-100, cells were incubated for 1 h with FITC- or 7.5), 5 mM MgCl , 5 mM ATP, 10 mM creatine phosphate, 5 units/ml TRITC-conjugated secondary antibody in PBS containing 3% BSA. of phosphocreatine kinase, and 1  protease inhibitor cocktail in a Cells were then observed using a confocal laser scanning total volume of 30 ml. After incubating the mixtures for 2 h at 371C, microscope (Carl Zeiss-LSM700). they were subjected to SDS–PAGE followed by immunoblot. Supplementary data Flow cytometry Supplementary data are available at The EMBO Journal Online Cells were washed with PBS, trypsinized, and fixed at 41C with 70% (http://www.embojournal.org). ethanol. They were washed with PBS and incubated in PBS contain- ing 0.1% Triton X-100, 200 mg/ml of RNase A, and 20 mg/ml of propidium iodide for 30 min at room temperature in the dark. DNA Acknowledgements contents were then determined by flow cytometry using FACSCalibur (Becton Dickinson). This work was supported by grants from the National Research Found- ation of Korea (NRF-2005-084-C00025 and M10533010001-05N3301). SWL, HMY, and JHP were the recipients of the BK21 fellowship. UV and c irradiation Author contributions: SWL, YJJ, and CHC analysed the data, Cells cultured to 50–70% confluence were washed with PBS and conceived the study, and wrote the paper. SWL and MHL initiated irradiated at 254 nm (UV-C) by using TUV lamp (Philips) or at 5 Gy the project and SWL performed all the experiments except purifica- by using Gammacell Low Dose-rate Research Irradiator (GC 3000 tion of p53, hnRNP-K, SUMO1, Sae1/2, and Ubc9, which was done Elan). UV dose (10 J/m ) was determined by using a UVX radio- by SHK, JHP, and HMY. Cloning of hnRNP-K cDNA and other meter (UVP Inc.). They were then incubated for various periods in plasmids was performed by SHK and SWL, and immunocyto- DMEM supplemented with 100 units/ml of penicillin, 1 mg/ml of chemistry was done by SWL and YMO. streptomycin, and 10% FBS. RT–PCR Total RNAs were isolated from cells by using TRIzol (Invitrogen). Conflict of interest RT–PCR was performed using RevertAid M–MuLV reverse transcrip- tase (Invitrogen) and oligo (dT) primer, according to manufacturer’s The authors declare that they have no conflict of interest. References Abraham RT (2001) Cell cycle checkpoint signaling through the SUMOylation affects the efficiency of the DNA repair. Nucleic ATM and ATR kinases. Genes Dev 15: 2177–2196 Acids Res 38: 4708–4721 Abraham RT (2004) PI 3-kinase related kinases: ’big’ players Baptiste-Okoh N, Barsotti AM, Prives C (2008) Caspase 2 is both in stress-induced signaling pathways. DNA Repair (Amst) 3: required for p53-mediated apoptosis and downregulated by p53 883–887 in a p21-dependent manner. Cell Cycle 7: 1133–1138 Altmannova V, Eckert-Boulet N, Arneric M, Kolesar P, Chaloupkova Bartek J, Lukas J (2001) Mammalian G1- and S-phase checkpoints R, Damborsky J, Sung P, Zhao X, Lisby M, Krejci L (2010) Rad52 in response to DNA damage. Curr Opin Cell Biol 13: 738–747 &2012 European Molecular Biology Organization The EMBO Journal VOL 31 NO 23 2012 4451 | | hnRNP-K SUMOylation for cell-cycle arrest SW Lee et al Bergink S, Jentsch S (2009) Principles of ubiquitin and SUMO Kruse JP, Gu W (2009) Modes of p53 regulation. Cell 137: 609–622 modifications in DNA repair. Nature 458: 461–467 Lakin ND, Jackson SP (1999) Regulation of p53 in response to DNA Bomsztyk K, Denisenko O, Ostrowski J (2004) hnRNP K: one damage. Oncogene 18: 7644–7655 protein multiple processes. Bioessays 26: 629–638 Lane DP (1992) Cancer. p53, guardian of the genome. Nature 358: 15–16 Bomsztyk K, Van Seuningen I, Suzuki H, Denisenko O, Ostrowski J Lee MH, Lee SW, Lee EJ, Choi SJ, Chung SS, Lee JI, Cho JM, Seol JH, (1997) Diverse molecular interactions of the hnRNP K protein. Baek SH, Kim KI, Chiba T, Tanaka K, Bang OS, Chung CH (2006) FEBS Lett 403: 113–115 SUMO-specific protease SUSP4 positively regulates p53 by pro- Brantley EC, Nabors LB, Gillespie GY, Choi YH, Palmer CA, Harrison moting Mdm2 self-ubiquitination. Nat Cell Biol 8: 1424–1431 K, Roarty K, Benveniste EN (2008) Loss of protein inhibitors of Levine AJ, Oren M (2009) The first 30 years of p53: growing ever activated STAT-3 expression in glioblastoma multiforme tumors: more complex. Nat Rev Cancer 9: 749–758 implications for STAT-3 activation and gene expression. Clin Levy DE, Inghirami G (2006) STAT3: a multifaceted oncogene. Proc Cancer Res 14: 4694–4704 Natl Acad Sci USA 103: 10151–10152 Capili AD, Lima CD (2007) Taking it step by step: mechanistic Li T, Evdokimov E, Shen RF, Chao CC, Tekle E, Wang T, Stadtman insights from structural studies of ubiquitin/ubiquitin-like protein ER, Yang DC, Chock PB (2004) Sumoylation of heterogeneous modification pathways. Curr Opin Struct Biol 17: 726–735 nuclear ribonucleoproteins, zinc finger proteins, and nuclear pore Chiu SY, Asai N, Costantini F, Hsu W (2008) SUMO-specific protease complex proteins: a proteomic analysis. Proc Natl Acad Sci USA 2 is essential for modulating p53-Mdm2 in development of 101: 8551–8556 trophoblast stem cell niches and lineages. PLoS Biol 6: e310 Liu LM, Yan MG, Yang DH, Sun WW, Zhang JX (2011) PIAS3 Chung CD, Liao J, Liu B, Rao X, Jay P, Berta P, Shuai K (1997) expression in human gastric carcinoma and its adjacent non- Specific inhibition of Stat3 signal transduction by PIAS3. Science tumor tissues. Clin Res Hepatol Gastroenterol 35: 393–398 278: 1803–1805 Matunis MJ, Michael WM, Dreyfuss G (1992) Characterization and Ciccia A, Elledge SJ (2010) The DNA damage response: making it primary structure of the poly(C)-binding heterogeneous nuclear safe to play with knives. Mol Cell 40: 179–204 ribonucleoprotein complex K protein. Mol Cell Biol 12: 164–171 Cimprich KA, Cortez D (2008) ATR: an essential regulator of Michael WM, Eder PS, Dreyfuss G (1977) The K nuclear shuttling genome integrity. Nat Rev Mol Cell Biol 9: 616–627 domain: a novel signal for nuclear import and nuclear export in Cremona CA, Sarangi P, Yang Y, Hang LE, Rahman S, Zhao X (2012) the hnRNP K protein. EMBO J 16: 3587–3598 Extensive DNA damage-induced sumoylation contributes to re- Moumen A, Masterson P, O’Connor MJ, Jackson SP (2005) hnRNP plication and repair and acts in addition to the mec1 checkpoint. K: an HDM2 target and transcriptional coactivator of p53 in Mol Cell 45: 422–432 response to DNA damage. Cell 123: 1065–1078 Dou H, Huang C, Singh M, Carpenter PB, Yeh ET (2010) Regulation Mukhopadhyay D, Dasso M (2007) Modification in reverse: the of DNA repair through deSUMOylation and SUMOylation of SUMO proteases. Trends Biochem Sci 32: 286–295 replication protein A complex. Mol Cell 39: 333–345 Niu G, Wright KL, Huang M, Song L, Haura E, Turkson J, Zhang S, Durocher D, Jackson SP (2001) DNA-PK, ATM and ATR as sensors of Wang T, Sinibaldi D, Coppola D, Heller R, Ellis LM, Karras J, DNA damage: variations on a theme? Curr Opin Cell Biol 13: 225–231 Bromberg J, Pardoll D, Jove R, Yu H (2002) Constitutive Stat3 Enge M, Bao W, Hedstrom E, Jackson SP, Moumen A, Selivanova G activity up-regulates VEGF expression and tumor angiogenesis. (2009) MDM2-dependent downregulation of p21 and hnRNP K Oncogene 21: 2000–2008 provides a switch between apoptosis and growth arrest induced Pelisch F, Pozzi B, Risso G, Munoz MJ, Srebrow A (2012) DNA by pharmacologically activated p53. Cancer Cell 15: 171–183 damage-induced heterogeneous nuclear ribonucleoprotein K Gareau JR, Lima CD (2010) The SUMO pathway: emerging mechan- SUMOylation regulates p53 transcriptional activation. J Biol isms that shape specificity, conjugation and recognition. Nat Rev Chem 287: 30789–30799 Mol Cell Biol 11: 861–871 Perry ME (2004) Mdm2 in the response to radiation. Mol Cancer Res Geiss-Friedlander R, Melchior F (2007) Concepts in sumoylation: a 2: 9–19 decade on. Nat Rev Mol Cell Biol 8: 947–956 Polo SE, Jackson SP (2011) Dynamics of DNA damage response Hay RT (2005) SUMO: a history of modification. Mol Cell 18: 1–12 proteins at DNA breaks: a focus on protein modifications. Genes Hoege C, Pfander B, Moldovan GL, Pyrowolakis G, Jentsch S (2002) Dev 25: 409–433 RAD6-dependent DNA repair is linked to modification of PCNA Rytinki MM, Kaikkonen S, Pehkonen P, Jaaskelainen T, Palvimo JJ by ubiquitin and SUMO. Nature 419: 135–141 (2009) PIAS proteins: pleiotropic interactors associated with Horn HF, Vousden KH (2007) Coping with stress: multiple ways to SUMO. Cell Mol Life Sci 66: 3029–3041 activate p53. Oncogene 26: 1306–1316 Sacher M, Pfander B, Hoege C, Jentsch S (2006) Control of Rad52 Jackson PK (2001) A new RING for SUMO: wrestling transcriptional recombination activity by double-strand break-induced SUMO responses into nuclear bodies with PIAS family E3 SUMO ligases. modification. Nat Cell Biol 8: 1284–1290 Genes Dev 15: 3053–3058 Shang Y, Hu X, DiRenzo J, Lazar MA, Brown M (2000) Cofactor Jang HD, Yoon K, Shin YJ, Kim J, Lee SY (2004) PIAS3 suppresses dynamics and sufficiency in estrogen receptor-regulated tran- NF-kappaB-mediated transcription by interacting with the p65/ scription. Cell 103: 843–852 RelA subunit. J Biol Chem 279: 24873–24880 Steinman RA, Johnson DE (2000) p21WAF1 prevents down-mod- Jepsen K, Hermanson O, Onami TM, Gleiberman AS, Lunyak V, ulation of the apoptotic inhibitor protein c-IAP1 and inhibits McEvilly RJ, Kurokawa R, Kumar V, Liu F, Seto E, Hedrick SM, leukemic apoptosis. Mol Med 6: 736–749 Mandel G, Glass CK, Rose DW, Rosenfeld MG (2000) Suzuki A, Tsutomi Y, Akahane K, Araki T, Miura M (1998) Combinatorial roles of the nuclear receptor corepressor in tran- Resistance to Fas-mediated apoptosis: activation of caspase 3 is scription and development. Cell 102: 753–763 regulated by cell cycle regulator p21WAF1 and IAP gene family Jiang M, Chiu SY, Hsu W (2011) SUMO-specific protease 2 in Mdm2- ILP. Oncogene 17: 931–939 mediated regulation of p53. Cell Death Differ 18: 1005–1015 Vassileva MT, Matunis MJ (2004) SUMO modification of hetero- Johnson ES (2004) Protein modification by SUMO. Annu Rev geneous nuclear ribonucleoproteins. Mol Cell Biol 24: 3623–3632 Biochem 73: 355–382 Vogelstein B, Lane D, Levine AJ (2000) Surfing the p53 network. Kerscher O, Felberbaum R, Hochstrasser M (2006) Modification of Nature 408: 307–310 proteins by ubiquitin and ubiquitin-like proteins. Annu Rev Cell Vousden KH, Lu X (2002) Live or let die: the cell’s response to p53. Dev Biol 22: 159–180 Nat Rev Cancer 2: 594–604 Kim KI, Baek SH, Chung CH (2002) Versatile protein tag, SUMO: its WeiD,LeX, Zheng L, Wang L, Frey JA,Gao AC,PengZ,Huang S, enzymology and biological function. J Cell Physiol 191: 257–268 Xiong HQ, Abbruzzese JL, Xie K (2003) Stat3 activation regulates the Kim ST, Lim DS, Canman CE, Kastan MB (1999) Substrate specifi- expression of vascular endothelial growth factor and human pan- cities and identification of putative substrates of ATM kinase creatic cancer angiogenesis and metastasis. Oncogene 22: 319–329 family members. J Biol Chem 274: 37538–37543 Yeh ET (2009) SUMOylation and De-SUMOylation: wrestling with Kluge A, Dabir S, Vlassenbroeck I, Eisenberg R, Dowlati A (2011) life’s processes. J Biol Chem 284: 8223–8227 Protein inhibitor of activated STAT3 expression in lung cancer. Zeng SX, Dai MS, Keller DM, Lu H (2002) SSRP1 functions as a co- Mol Oncol 5: 256–264 activator of the transcriptional activator p63. EMBO J 21: 5487–5497 4452 The EMBO Journal VOL 31 NO 23 2012 &2012 European Molecular Biology Organization | |

Journal

The EMBO Journal – Springer Journals

Published: Nov 28, 2012

Keywords: HDM2; p21; PIAS3; SENP2; ubiquitin

There are no references for this article.