Differential Regulation of c-Jun-dependent Transcription by SUMO-specific Proteases*

c-Jun is a transcription factor that plays an important role in regulating cell growth, apoptosis, differentiation, and transformation. The transcriptional activity of c-Jun can be regulated by both phosphorylation and sumoylation. It has also been shown that c-Jun transcription can be regulated by SuPr-1, an alternatively spliced form of SUMO-specific protease 2 (SENP2). However, the ability of SuPr-1 to enhance c-Jun transcription is dependent on promyelocytic leukemia but is independent of the desumoylation activity of SuPr-1. Here, we show that SUMO-specific protease 1 (SENP1) also markedly enhances the transcription activity of c-Jun. The action of SENP1 on c-Jun transcription is independent of the sumoylation and phosphorylation status of c-Jun but is critically dependent on the desumoylation activity of SENP1. We further show that p300 is essential for SENP1 to enhance c-Jun-dependent transcription because SENP1 can desumoylate the CRD1 domain of p300, thereby releasing the cis-repression of CRD1 on p300. Thus, two SUMO-specific proteases regulate c-Jun-dependent transcription through entirely different mechanisms.

c-Jun is a major component of the heterodimeric transcription factor AP-1 and plays an important role in regulating cell growth, apoptosis, differentiation, and transformation (1)(2)(3)(4)(5). A plethora of physiological stimuli and environmental stresses can induce c-Jun activity through the c-Jun N-terminal kinase (JNK) 1 -mediated pathway (3)(4)(5)(6)(7)(8)(9). Once activated, JNKs translocate to the nucleus, phosphorylate c-Jun at Ser-63 and Ser-73, and thereby enhance c-Jun transcriptional activity (3, 6, 8 -11). Although JNK-mediated c-Jun phosphorylation is a well documented mechanism for activating c-Jun-dependent transcription, studies with knock-in mice indicated that phosphorylation of c-Jun at Ser-63 and Ser-73 is not essential for some of the biological functions of c-Jun (12)(13)(14). These observations suggest that regulation of c-Jun transcriptional activity is complex and may occur at different levels.
It is well known that c-Jun interacts with various co-regulatory proteins; this interaction can regulate the transcriptional activity of c-Jun (7,(15)(16)(17)(18). p300 is a well known co-activator of c-Jun (7,16,19,20) that has been shown to physically interact with c-Jun and activate c-Jun-dependent transcription (16). Because the transcriptional activity of p300 can be modulated by a number of signaling pathways (21)(22)(23)(24)(25), p300 provides an additional level of regulation for c-Jun-dependent transcription.
It has been reported that p21 regulates p300 transcriptional activity (23,(25)(26)(27). p21 not only inhibits p300-bound cyclin E-Cdk2 activity through repression of the histone acetyltransferase activity of p300 (28), it also stimulates p300 transactivation (27). Within p300, a domain named CRD1 has been identified as a domain with strong transcriptional repression (27). CRD1 functions independently of the p300 histone acetyltransferase domains but can repress the transactivational activity of p300 (22,27). p21 de-represses this CRD1 activity and thus selectively activates p300-dependent transcription at specific promoters (27). Recent findings indicate that sumoylation is required for CRD1-dependent transcriptional repression (22). The two SUMO modification sites within the CRD1 domain of p300 have been identified, and mutation at these two sites can reduce the repression of CRD1 domain and p21 inducibility (22). Therefore, SUMO modification provides a new mechanism to control p300 function.
In this study, we show that SENP1 can induce c-Jun-dependent transcription. In contrast to that of SuPr1, the desumoylation activity of SENP1 is absolutely required for its ability to induce c-Jun-dependent transcription. We show that p300 is essential for the activation of c-Jun-dependent transcription by SENP1. Furthermore, SENP1 can desumoylate p300 and release the repression of SUMO-CRD1, leading to an increase in p300 transactivation. Our findings demonstrate a difference between SENP1 and SENP2 in the regulation of c-Jun-dependent transcription and provide a novel mechanism for regulating c-Jun-dependent transcription.
Cell Transfection and Luciferase Assays-PC-3 cells were grown in RPMI 1640 medium (Invitrogen) supplemented with 10% fetal bovine serum. COS-7 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. After 24 h of cultivation, these cells were transiently transfected with expression plasmids by Lipofectamine (Invitrogen) according to the manufacturer's instructions. The cells were starved for 18 h before luciferase was assayed as described previously (50). ␤-Galactosidase activity was used as an internal control.
TALON Resin Precipitation-TALON resin (Clontech) precipitation of His-p300 was carried out as described in a previous publication (51). Briefly, total cell lysates were prepared in lysis buffer (6 M guanidine hydrochloride, 20 mM sodium phosphate, 500 mM sodium chloride, pH 7.8). DNA in the sample was sheared with a 25-gauge needle, and the lysate was centrifuged at 100,000 ϫ g at 15°C for 30 min. The supernatant was incubated with TALON resin beads for 1 h at room temperature. The beads were washed twice with washing buffer (8 M urea, 20 mM sodium phosphate, 500 mM sodium chloride, pH 7.8) and then twice more with another washing buffer (8 M urea, 20 mM sodium phosphate, 500 mM sodium chloride, pH 6.0). Subsequently, the beads were washed with phosphate-buffered saline twice and treated in 2% sodium dodecyl sulfate treating solution for SDS-polyacrylamide gel electrophoretic analysis.
Western Blotting-Western blotting was carried out as described in our previous publication (52).
RNA Interference-The 21-nucleotide SENP1 small interfering RNA (siRNA) (GTGAACCACAACTCCGTATTC) was synthesized by Dharmacon (39). The same sequence in the inverted orientation was used as a nonspecific siRNA control. The SENP1 and nonspecific siRNA oligonucleotides were inserted into the pSuppressorNeo vector (IMGENEX Corporation) according to the manufacturer's instructions. PC-3 cells were grown in 6-well plates and transfected with the siRNA plasmid (1 g) three times at 12-h intervals using Lipofectamine 2000 (Invitrogen). The cells were harvested 72 h after transfection. Expression of SENP1 and c-Jun was detected by using real-time PCR (for SENP1) and Western blot (for c-Jun).

SENP1 Is a Stronger
Activator of c-Jun than SENP2-Best et al. (32) reported that SuPr-1, an alternatively spliced form of SENP2, could induce c-Jun-dependent transcription. Because both SENP1 and SENP2 belong to the SUMO-specific protease family with broad substrate specificity (41), we speculated that SENP1 might also be an activator of c-Jun. To test this hypothesis, we performed a luciferase reporter gene assay by using Gal4 fused to the transactivation domain (1-223) of c-Jun (G4-c-Jun) and the Gal4-luciferase reporter plasmid. When expressed in PC-3 cells, SENP1 markedly induced G4-c-Jundependent transcription (Fig. 1A). SENP1 exhibited stronger activation of c-Jun-dependent transcription than SENP2. Ti-tration of SENP1 showed a dose-dependent effect of SENP1 on c-Jun-dependent transcription (Fig. 1B). The effects of SENP1 in different cell lines, such as 293, MCF-7, HeLa, and U-2OS cells, were tested, and cell type specificity was not observed (data not shown). These results suggest that SENP1 can function as a strong activator of c-Jun-dependent transcription.
A previous study indicated that desumoylation activity was not required for SuPr-1, a splice variant of SENP2, to induce c-Jun activity (32). We similarly observed that the catalytic mutant of SENP2 actually induces more c-Jun activity than wild-type SENP2 (Fig. 1A). However, these two proteases SENP1 and SENP2 diverged in their effects on c-Jun-dependent transcription. In contrast to SENP2, the action of SENP1 on c-Jun is dependent on its catalytic activity, as the catalytic inactive mutation markedly reduced the effect of SENP1 on c-Jun-dependent transcription (Fig. 1A). We confirmed this result by increasing the amount of SENP1 mutant transfected, which did not prompt any significant change in c-Jun-dependent transcription (Fig. 1B). Western blot analysis insured that the wild-type and mutant SENP1 were expressed at similar levels and did not alter G4-c-Jun expression (Fig. 1C). Collectively, these data suggest that the action of SENP1 on c-Jundependent transcription, unlike that of SENP2, is mediated through a desumoylation mechanism.
To determine whether SENP1 could affect transcription of an endogenous promoter, we examined the effect of SENP1 on the c-Jun promoter (Ϫ79/ϩ170), which contains AP-1 binding sites in the Ϫ72 position (53,54). As shown in Fig. 1D, SENP1 induced c-Jun promoter activity in a dose-dependent manner. The catalytic activity of SENP1 is also required for this effect. We also examined whether the effect of SENP1 on the c-Jun promoter is through the AP-1 binding site. The mutation of the AP-1 binding site markedly abolished the activity of the c-Jun promoter by SENP1 (Fig. 1D). To further confirm the effect of SENP1 on c-Jun-dependent transcription, we used siRNA to silence endogenous SENP1 and then examined whether the expression of endogenous c-Jun, a target dependent on c-Jun transactivation, was affected. The transfection of the SENP1specific siRNA plasmid into PC-3 cells decreased endogenous SENP1 expression by 53% (real-time PCR analysis, data not shown), whereas expression of SENP1-siRNA reduced endogenous c-Jun expression by 60% (Fig. 1E). Collectively, these data indicate that SENP1 can strongly activate c-Jun-dependent transcription through its desumoylation activity.
SENP1 Activation of c-Jun Is Independent of Phosphorylation-c-Jun is a transcription factor involved in the JNK signaling pathway (2,3,7,8,(55)(56)(57). JNK modulates c-Jun-dependent transcription through phosphorylation of c-Jun at Ser-63 and Ser-73 (8). To investigate the mechanism underlying c-Jun activation by SENP1, we first examined whether SENP1 could indirectly induce phosphorylation of c-Jun. Anti-phosphorylated c-Jun antibody was used to identify G4-c-Jun phosphorylation. As shown in Fig. 2A, phorbol 12-myristate 13-acetate (PMA), a stimulator of the JNK pathway, strongly induced c-Jun phosphorylation; however, co-expression of SENP1 and the SENP1 mutant could not induce the phosphorylation of G4-c-Jun ( Fig. 2A). We also used the G4-c-Jun S63A,S73A construct in which Ser-63 and Ser-73 are mutated to Ala to perform the luciferase reporter assay. Mutation of both Ser-63 and Ser-73 to Ala did not affect the ability of SENP1 to induce G4-c-Jun activity (Fig. 2B), whereas PMA induction of c-Jun was abolished by the mutation (Fig. 2C). These results suggest that the action of SENP1 on c-Jun-dependent transcription is not mediated through a phosphorylation mechanism.
SENP1-Inducing c-Jun Transcriptional Activity Occurs Independently of c-Jun Desumoylation-c-Jun could be conju-  (50 ng) in the presence of SENP1 or SENP1 mutant plasmids (100 ng). Luciferase activity was measured as described in Fig. 1A. The c-Jun expression level was measured by using Western blotting with anti-Gal4-DBD antibody. mut, mutant; FL, full-length.

FIG. 1. SENP1 induces c-Jundependent transcription.
A, SENP1 is a stronger activator than SENP2 in G4-c-Jun-dependent transcription. PC-3 cells were transfected with Gal4-luciferase (G4-Luc) (100 ng) and Gal4-DBD or G4-c-Jun plasmid (50 ng) in the absence or presence of SENP1 or SENP2 wild-type and mutant plasmids (100 ng). Luciferase activity was measured. Transfection efficiency was normalized by using a ␤-galactosidase expression construct. B, dose response of SENP1 action. PC-3 cells were transfected with Gal4-luciferase (100 ng) and Gal4-DBD or G4-c-Jun plasmid (50 ng) in the absence or presence of increasing amounts of SENP1 wild-type or mutant plasmids (10, 50, 200 ng). Luciferase activity was measured as described in A. C, Western blots of cell extracts from B. D, SENP1 induces c-Jun promoter (Ϫ79/ ϩ170) activity. PC-3 cells were transfected with Jun (Ϫ79/ϩ170)-luciferase or Jun (Ϫ79/ϩ170)-luciferase mutant reporter plasmid (100 ng) in the absence or presence of increasing amounts of SENP1 wild-type or mutant plasmids (10, 50, 200 ng). Luciferase activity was measured as described in A. mut, mutant. E, silencing endogenous SENP1 reduces endogenous c-Jun expression. PC-3 cells were transfected with nonspecific siRNA (NS-siRNA) and SENP1-siRNA expression plasmids. The c-Jun expression was measured by using Western blotting with an anti-c-Jun antibody. gated by SUMO at amino acid 229 (40). Although the G4-c-Jun plasmid used in the above experiments only contained amino acids 1-223 of the c-Jun transactivation domain and thus could not be sumoylated, it is still possible that c-Jun transcriptional activity induced by SENP1 is dependent on its sumoylation status in vivo. To test this possibility, we generated G4-c-Jun full-length and G4-c-Jun full-length sumoylation mutant (K229R) plasmids. The mutant was then compared with the wild-type protein in the Gal4-luciferase reporter system. Consistent with the previous study (40), the mutant exhibited higher transcriptional activity than wild-type c-Jun (2-fold) (Fig. 2D, lane 2 versus lane 5). However, co-expression of SENP1 still markedly enhanced the c-Jun mutant's transcriptional activity by 28-fold when compared with vector control (Fig. 2D, lane 5 versus lane 6). The effect of SENP1 on the c-Jun mutant was similar to that of the c-Jun wild type and was also dependent on its catalytic activity (Fig. 2D). These data suggest that most of the enhancement of c-Jun-dependent transcription by SENP1 is independent of the sumoylation status of c-Jun.
p300 Is Essential for Induction of c-Jun Activity by SENP1-Because most of the c-Jun-mediated transcription regulated by SENP1 is not through desumoylation of c-Jun itself, the coregulator might be the mediator in the effect of SENP1 on c-Jun transcriptional activity.
HDACs act as repressors to down-regulate transcription (58). Among the HDAC family, HDAC1 and HDAC4 can be sumoylated (33,39,59,60). HDAC1 has been shown to be desumoylated by SENP1 in regulating androgen receptor-dependent transcription (39). It is possible that desumoylation of HDAC by SENP1 plays a role in inducing c-Jun activity. We first examined whether HDACs are required for SENP1 action on G4-c-Jun activity. Cells were treated with the HDAC inhibitor trichostatin A (TSA) after co-transfection with the G4-c-Jun reporter system and SENP1 or empty vector. TSA could not augment G4-c-Jun activity (Fig. 3A), suggesting that HDACs were not involved in the regulation of G4-c-Jun in our reporter system. Because HDACs had no effect on G4-c-Jun activity, TSA could not alter the effect of SENP1 on G4-c-Jun either (Fig. 3A).
Another important co-regulator for c-Jun activity is p300, which can interact with c-Jun and enhance c-Jun-dependent transcription (16). Importantly, p300 could be modified by SUMO, and the sumoylation of p300 is essential for the cisrepression function of CRD1 on p300 transactivation (22). Therefore, we tested whether p300 is involved in the activation of c-Jun by SENP1. Because adenovirus E1A specifically interacts with p300 and inhibits p300 activity (16, 61-63), we tested whether E1A could inhibit SENP1 activity on G4-c-Jun. As shown in Fig. 3B, overexpression of E1A strongly repressed c-Jun-dependent transcription (lane 2 versus lane 1), whereas E1A⌬2-36, a p300 binding-defective mutant (64), did not. This suggests that p300 is a crucial cofactor for G4-c-Jun transcriptional activity. When SENP1 was co-expressed with E1A, the enhancement of G4-c-Jun transcription by SENP1 was almost completely inhibited (Fig. 3B). We further confirmed that the effect of E1A is dependent on its ability to bind to p300 by using E1A⌬2-36. The effect of E1A on SENP1 was severely impaired by mutation of the p300 binding region of E1A (Fig. 3B). These results suggest that SENP1 action on G4-c-Jun activity is mediated by p300.
In addition, we overexpressed p300 to evaluate its contribution to SENP1-mediated activation of c-Jun. As shown in Fig.  3C, p300 synergized with SENP1 in enhancing G4-c-Jun-mediated transcription. Overexpression of p300 alone could induce G4-c-Jun activity by 3.5-fold. In the presence of SENP1, however, the transactivation activity of p300 was enhanced 6.1-fold. The effect of SENP1 is dependent on its catalytic activity because the p300 could not enhance transcription in the presence of the SENP1 catalytic mutant (Fig. 3C). These data indicate that the ability of SENP1 to enhance G4-c-Jun transcription is mediated through the desumoylation of p300.
SENP1 Action on c-Jun Is Mediated through the CRD1 Domain of p300 -Because the two sumoylation sites are located exclusively in the CRD1 repressive domain (22), we speculated that SENP1 desumoylates the CRD1 domain and thereby releases the repression of G4-c-Jun activity by p300. To test this hypothesis, we first determined whether G4-c-Jun activity could be affected by the CRD1 domain of p300. For this purpose, we used a p300 mutant lacking the CRD1 domain to evaluate the transcription activity of G4-c-Jun. The CRD1-deleted mutant elevated the activity of G4-c-Jun to a greater extant than wildtype p300 (Fig. 4A), indicating that the CRD1 domain represses p300 transactivation on G4-c-Jun-dependent transcription.
We wanted to further ensure that the CRD1 domain is required for SENP1 action on c-Jun activity. G4-c-Jun reporter plasmids were co-transfected into PC-3 cells with wild-type p300 or the CRD1-deleted mutant. As shown in Fig. 4B, the p300 mutant transactivated G4-c-Jun more than the wild type. SENP1 alone induced G4-c-Jun activity by 29.4-fold, whereas overexpression of wild-type p300 increased the activity of SENP1 on G4-c-Jun by as much as 58.5-fold. Moreover, deletion of CRD1 decreased SENP1 induction to 26.8-fold. These results suggest that the CRD1 domain of p300 mediates the action of SENP1 on G4-c-Jun. SENP1 Increases p300 Transactivation by Desumoylating the CRD1 Domain-Although previous reports suggest that the CRD1 domain of p300 is sumoylated, it was unknown whether SENP1 could remove SUMO from the conjugated CRD1 domain. We first determined whether p300 could associate with SENP1 in vivo. SENP1 was co-precipitated with p300 (1-1045) in cell extracts (Fig. 5A). We then performed a co-transfection experiment to examine whether the CRD1 domain is the target of the desumoylation activity of SENP1. SENP1 was co-expressed with His-p300 (1-1045), a fragment containing the CRD1 domain, and HA-SUMO-1 in COS-7 cells. As shown in Fig. 5B, p300 (1-1045) was conjugated by SUMO-1 (Fig. 5B, lane 4). SENP1 removed all SUMO-conjugated bands (Fig. 5B, lane 5), but the SENP1 mutant did not (Fig. 5B, lane 6). Consistent with another study (22), a CRD1 domain-deleted form of p300 (1-1004) cannot be conjugated by SUMO (Fig. 5B, lane 2). These results clearly show that SENP1 desumoylates p300.
To determine a direct effect of SENP1 on p300 activity, we used the Gal4-DBD reporter system in which p300 is fused to the Gal4 DNA-binding domain (Fig. 5C). When expressed in PC-3 cells, SENP1 induced the G4-p300 activity by 30.8-fold, but SENP1 mutant did not (Fig. 5D). Deletion of the CRD1 domain increased p300 transactivation by 8-fold (Fig. 5D, lane  4 versus lane 7), whereas it severely impaired the effect of SENP1 on p300 transactivation (Fig. 5D). These data suggest that the CRD1 domain is required for the action of SENP1 on p300 transactivation.
To further determine whether SENP1 action is mediated through desumoylating the CRD1 domain of p300, we used a Gal4-p300Nϩminimal CRD1 domain (mCRD1) fusion construct. The mCRD1 is required for CRD1-mediated repression and contains two sites for SUMO modification (22) (Fig. 5C). The p300NϩmCRD1 fusion protein exhibited only 4% of the activity of p300N (Fig. 5E, lane 1 versus lane 4). However, expression of SENP1, but not the SENP1 mutant, completely reversed mCRD1 repression up to the level of p300N (Fig. 5E). Substitution of both lysines of mCRD1 by Arg (Fig. 5C) completely relieved its repression and reduced the effect of SENP1 on p300 transactivation (Fig. 5E). These data suggest that the action of SENP1 on p300 transactivation is mediated through desumoylation of the CRD1 domain. DISCUSSION c-Jun, a major component of AP-1, plays an important role in cell growth, apoptosis, differentiation, and transformation (1-5, 10, 57, 65-67). Numerous reports from biochemical and functional studies have established that activation of JNK leads to increased c-Jun Ser-63 and Ser-73 phosphorylation and transcriptional activity (1, 6 -8, 17, 68). Results from a knock-in mouse model showed, however, that some biological functions of c-Jun did not require its N-terminal phosphorylation (12)(13)(14), indicating that JNK-c-Jun signaling transduction is not the only mechanism for regulating c-Jun activity. SuPr-1, a spliced form of SENP2, was recently identified as an activator of c-Jun independent of c-Jun phosphorylation (32). The mechanism underlying SuPr-1 action on c-Jun activity is through SuPr-1 binding of SUMO-modified PML, altering the distribution of promyelocytic leukemia and promyelocytic leukemia oncogenic domain-associated proteins (32). The effect of SuPr-1 on c-Jun activity is independent of its desumoylation activity (32).
In the present study, we have found that SENP1, another member of the SENP family, strongly induces c-Jun-dependent transcription. We observed similar results in the regulation of c-Jun activity by SENP2 as observed for SuPr-1 (32); neither protease requires its catalytic activity to regulate c-Jun. In contrast to SENP2, the desumoylation activity of SENP1 is required for enhancement of c-Jun activity. In our study, the modulation of c-Jun by SENP1 was not via desumoylation of c-Jun. Instead we show that the action of SENP1 on c-Jun transcription is mediated through desumoylation of p300.
co-activator for c-Jun-dependent transcription (7, 16 -18, 69). Expression of p300 enhances c-Jun activity (16). A previous study identified a transcriptional repression domain, CRD1, within p300 (27), and Girdwood et al. (22) showed that this domain represses p300 transactivation in a sumoylation-dependent manner. We hypothesized that SENP1 induces c-Jundependent transcription via desumoylation of the CRD1 domain of p300. This hypothesis was based on the function of the CRD1 domain of p300 on the c-Jun-dependent transcription and on the finding that the sumoylation sites are located exclusively in the repressive domain (32). The hypothesis was confirmed by the following results. First, p300 is essential for SENP1 action on c-Jun-dependent transcription (Fig. 3, B and  C). Second, the CRD1 domain represses p300 transactivation on c-Jun-dependent transcription (Fig. 4A) and is required for the action of SENP1 on c-Jun (Fig. 4B). Third, the CRD1 domain is a direct target of SENP1, and SENP1 could induce p300 transactivation through the desumoylation of this domain (Fig. 5). We proposed a SENP1 action model (Fig. 6). In this model, SENP1 deconjugates SUMO from CRD1, relieving cisrepression of SUMO-CRD1 on p300 transactivation and thereby increasing c-Jun-dependent transcription.
SUMO modification could repress the activity of several transcription factors (22,31). HDAC has been shown to be recruited to sumoylated Elk1 and p300, repressing the transactivation activity of these proteins (22,31). In the present study, we did not observe any effect of TSA on G4-c-Jun on SENP1 action on G4-c-Jun (Fig. 3A). This suggests that at least the members of class I and II histone deacetylase families were not involved in either the regulation of G4-c-Jun activity or the enhancement of G4-c-Jun activity by SENP1. This is in contrast to the profound TSA effect on the SENP1 effect on androgen receptor-dependent transcription that we reported earlier (39). The reason for the lack of TSA effect is not precisely known at present. Recently Pestell and co-workers (70) reported that in SirT1, a member of the class III histone deacetylase family, repression of p300 required sumoylation on the CRD1 domain of p300, indicating that sumoylated p300 could also recruit SirT1 to repress p300 transactivation independent of TSA.