Biochemical Journal

Research article

MAPK-induced Ser727 phosphorylation promotes SUMOylation of STAT1

Sari Vanhatupa, Daniela Ungureanu, Maija Paakkunainen, Olli Silvennoinen

Abstract

STAT1 (signal transducer and activator of transcription 1) is a critical mediator of IFN-γ (interferon-γ)-induced gene responses, and its function is regulated through phosphorylation of Tyr701 and Ser727. MAPK (mitogen-activated protein kinase) pathways mediate phosphorylation of Ser727 in response to microbial infections, stress stimuli and growth factors. Recently, STAT1 was found to become modified by PIAS (protein inhibitor of activated STAT)-mediated SUMO-1 (small ubiquitin-related modifier-1) conjugation at Lys703, but the regulation of this modification is largely unknown. Here, we have investigated the role of MAPK-induced Ser727 phosphorylation in regulation of STAT1 SUMOylation. Activation of the p38MAPK pathway by upstream activating kinase, MKK6 (MAPK kinase-6) or osmotic stress enhanced the SUMOylation of STAT1, which was counteracted by the p38MAPK inhibitor SB202190 or by dominant-negative p38MAPK. Activation of the ERK1/2 (extracellular-signal-regulated kinase 1/2) pathway by Raf-1 also enhanced Ser727 phosphorylation and SUMOylation of STAT1, and this induction was counteracted by PD98059 inhibitor. Mutation of Ser727 to alanine abolished the p38MAPK-induced SUMOylation. Furthermore, S727D and S727E mutations, which mimic the phosphorylation of Ser727, enhanced the basal SUMOylation of STAT1 and interaction between PIAS1 and STAT1. Taken together, these results identify Ser727 phosphorylation as a regulator of STAT1 SUMOylation and highlight the central role of Ser727 in co-ordination of STAT1 functions in cellular responses.

  • mitogen-activated protein kinase (MAPK)
  • signal transducer and activator of transcription (STAT)
  • protein inhibitor of activated STAT1 (PIAS1)
  • Ser727 phosphorylation
  • small ubiquitin-related modifier (SUMO)
  • transcription factor

INTRODUCTION

IFN-γ (interferon-γ) has an important function in immune responses against microbes by regulating macrophage functions and Th1 differentiation. IFN-γ-induced gene responses are largely dependent on STAT1 (signal transducer and activator of transcription 1). Activation of STAT1 requires phosphorylation on Tyr701 by JAK (Janus kinase) kinases, and the tyrosine-phosphorylated STAT1 forms dimers that are translocated to the nucleus to bind to specific elements termed GAS (γ-activated sequence) on IFN-γ-responsive promoters [1,2]. Regulation of STAT1 activation involves also phosphorylation of Ser727 in the TAD (transactivation domain), and this post-translational modification modulates the interaction of STAT1 with co-activator proteins and selectively affects IFN-induced gene responses [3,4]. STAT1 Ser727 phosphorylation can be induced by various kinases activated by IFN stimulation or through extracellular signals such as mediators of inflammation, stress and growth signals. p38MAPK (where MAPK is mitogen-activated protein kinase) mediates the microbial and stress responses [3], while CaMKII (Ca2+/calmodulin-dependent protein kinase II) and PKCδ (protein kinase C δ) have been shown to phosphorylate STAT1 in response to IFN-α and IFN-γ stimulations respectively [5,6]. During antimicrobial response, the combinatorial input of these pathways secures optimal functional response [7,8].

The function of STAT1 is negatively regulated by several transacting proteins such as protein tyrosine phosphatases, SOCS1 (suppressor of cytokine signalling 1), SLIM1 (STAT-interacting LIM protein 1) [9], as well as by PIAS (protein inhibitor of activated STAT) proteins. The family of PIAS proteins consists of five members, PIAS1, PIAS3, PIASxα, PIASxβ and PIASy, that have been implicated in the regulation of several nuclear signalling pathways. PIAS1 and PIAS3 were identified as interaction partners for STAT1 and STAT3 respectively, and they have been shown to inhibit the DNA-binding activity of activated STATs [1013]. Analysis of PIAS1−/− mice showed that disruption of PIAS1 resulted in enhanced immune responses to viral or bacterial infections. PIAS1 was shown to selectively regulate a subset of IFN-γ- and IFN-β-responsive genes in macrophages by interfering with the recruitment of STAT1 to the promoters that show low affinity to STAT1 [14]. Interestingly, all PIAS proteins exert SUMO (small ubiquitin-related modifier)-E3 ligase activity towards various SUMO targets. SUMO family of proteins (SUMO-1, -2, -3 and -4) are small ubiquitin-related proteins covalently conjugated to the ϵ-amino group of lysine residues in target proteins [9]. The RING (really interesting new gene)-domain-containing E3 ligases facilitate the transfer of SUMO from E2 conjugation enzyme Ubc9 to the substrate by catalysing the conjugation reaction [13,1517].

The functional consequences of SUMOylation vary significantly from protein to protein, and the modification has been implicated in regulation of protein–protein interactions, protein stability, localization and transactivation, capacity [10]. Recently, STAT1 was shown to become SUMOylated by the PIAS proteins. SUMO-1 modification occurs in vivo at a single, evolutionarily conserved C-terminal amino acid residue Lys703. Mutation of Lys703 (STAT1-KR) resulted in increased IFN-γ-mediated transactivation, thus suggesting a negative regulatory function of SUMOylation in STAT1 [18,19]. More detailed analysis of IFN-γ target genes indicated that SUMOylation selectively modulated the induction of genes with low-affinity promoter (GBP-1) to STAT1, while high-affinity promoter [IRF-1 (IFN regulatory factor-1)] was not significantly affected, thus showing a similar pattern of gene regulation to that observed in PIAS1−/− mice [14,20].

Appropriate biological responses to external challenges such as infections require the convergence and cross-talk of different signalling pathways. Phosphorylation is an important mechanism for regulation of other post-translational modifications such as ubiquitination and acetylation [21,22]. Recently, phosphorylation has been also implicated in regulation of SUMO modification. Taking into account the diverse functional roles of SUMOylation in cells, it is not surprising that the effect of phosphorylation on SUMOylation varies between different proteins. Serine phosphorylations of c-Jun, p53, IκBα (inhibitory κBα) and PML (promyelocytic leukaemia) have been reported to inhibit their SUMOylation [2325], and in the case of the transcription factor ELK-1 [ETS (E twenty-six)-like kinase-1], ERK (extracellular-signal-regulated kinase)–MAPK-induced phosphorylation results in deSUMOylation of ELK-1 and increased transcriptional activation [26]. Phosphorylation has also been reported to positively regulate SUMO-1 conjugation, and a well-characterized example is HSF1 (heat-shock factor 1), where stress-induced phosphorylation of a serine residue enhances the SUMOylation of an adjacent lysine residue [27].

In the present study, we have investigated the regulation of SUMO-1 modification in STAT1. Our results indicate that MAPK-induced phosphorylation of Ser727 enhances PIAS1 binding and SUMO-1 conjugation to STAT1. The MAPK-pathway-mediated regulation of STAT1 SUMOylation suggests that the interplay between SUMOylation and Ser727 phosphorylation is involved in fine-tuning the transcriptional responses during immune responses to bacterial infections, stress stimuli and growth signals.

MATERIALS AND METHODS

Cell culture, transfections and plasmids

COS-7 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal bovine serum (Gibco BRL, Life Technology, Gaithersburg, MD, U.S.A.) and 100 units/ml penicillin and 100 μg/ml streptomycin. COS-7 cells were seeded on to 6-well plates and transfected using FuGENE™ reagent according to the manufacturer's instructions (Roche Molecular Biochemicals) with STAT1-WT (wild-type)–HA (haemagglutinin) or different mutants (0.7 μg), SUMO-1 (0.2 μg), MKK6b (MAPK kinase 6b) (E) (0.03–0.2 μg) or 0.1 μg of MKK6b (E), MKK1, Raf-1 (Figures 3A and 3B). His–SUMO-1, STAT1-WT–HA (pEBB-HA) and STAT1-K703R–HA (pEBB-HA) plasmids were previously described [18,28]. STAT1-Y701F and STAT1-S727A mutations and FLAG–PIAS1 have been previously described [28,29]. Dominant-negative p38MAPKα (p38AF) and constitutively active MKK6b [MKK6b (E)] were kindly provided by Dr J. Han (Department of Immunology, The Scripps Research Institute, La Jolla, CA, U.S.A.). Constitutively active MMK1 was provided by Dr Natalie Ahn (Howard Hughes Medical Institute, University of Colorado, Boulder, CO, U.S.A.) and Raf-1 by Dr Ulf Rapp [Institut für Medizinische Strahelnkunde und Zellforschung (MSZ), University of Würzburg, Wurzburg, Germany].

Antibodies and reagents

Anti-SUMO-1 antibody was purchased from Zymed, and anti-HA (anti-influenza virus HA, clone 16B12) antibody was purchased from Berkeley-Antibody (Richmond, CA, U.S.A.). Anti-FLAG M2 monoclonal antibody was purchased from Sigma. Rabbit polyclonal anti-P-S727-STAT1 antibody was from Cal-biochem (San Diego, CA, U.S.A.), rabbit polyclonal anti-P-p38MAPK (Thr180/Tyr182) antibody and mouse monoclonal anti-P-ERK1/2 (Thr202/Tyr204) were from Cell Signaling Technology. Rabbit polyclonal anti-p38MAPK (N-20) and rabbit polyclonal anti-IgG were from Santa Cruz Biotechnology. Biotinylated anti-mouse antibody for Western blot was from Dako (Glostrup, Denmark), and streptavidin–biotin–horseradish peroxidase conjugate was from Amersham Biosciences (Little Chalfont, Bucks., U.K.). Human IFN-γ and mouse IFN-γ were purchased from R&D Systems. The p38MAPK inhibitor SB202190 was purchased from Calbiochem, and PD98059 (Calbiochem) was kindly provided by Dr J. Westermarck.

Western blotting

Transiently transfected COS-7 cells were starved and left unstimulated or stimulated with IFN-γ (100 ng/ml) for 30 min. To perform osmotic shock, cells were treated with 0.3 M sorbitol and then washed in ice-cold PBS and lysed in Triton X lysis buffer [50 mM Tris/HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 50 mM NaF, 1% Triton X-100 and 10% (v/v) glycerol] supplemented with protease inhibitors and 10 mM NEM (N-ethylmaleimide; Calbiochem). The lysates were cleared by centrifugation at 12000 g for 20 min at 4 °C. The protein amount was determined using the Bio-Rad Dc protein assay kit (Bio-Rad Laboratories).

Immunodetection

After SDS/PAGE, the proteins were transferred on to a nitrocellulose membrane (Protran, Schleicher and Schuell) and blocked with 5% (w/v) non-fat dried milk in TBS (Tris-buffered saline)+0.1% Tween 20. The filters were incubated with the specific antibodies diluted in TBS+0.05% Tween 20. Immunodetection was performed using the ECL® (enhanced chemiluminescence) system (Amersham Biosciences).

Mutagenesis

STAT1-S727D and STAT1-S727E were created from STAT1-WT–HA using the Site-Directed Mutagenesis kit (Stratagene) and the following primers were used in the mutagenesis PCR: 5′-ACAACCTGCTCCCCATGGATCCTGAGGAGTTTGAC-3′, 5′-GTCAAACTCCTCAGGATCCATGGGGAGCAGGTTGT-3′ (S727D mutation) and 5′-ACAACCTGCTCCCCATGGAACCTGAGGAGTTTGAC-3′, 5′-GTCAAACTCCTCAGGTTCCATGGGGAGCAGGTTGT-3′ (S727E mutation).

Co-immunoprecipitation

Transfected COS-7 cells were left unstimulated or stimulated with IFN-γ (50 ng/ml) and sorbitol (0.3 M) for 30 min. Cells were resuspended in NP40 lysis buffer [50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.5% NP40 (Nonidet P40), 10% glycerol and 50 mM NaF] supplemented with PMSF, aprotinin and NEM. FLAG-tagged PIAS1 was immunoprecipitated from lysates with anti-FLAG M2 monoclonal antibody (Sigma) and rabbit polyclonal anti-IgG (Santa Cruz Biotechnology) was used as a control. Immunoprecipitated proteins were separated by SDS/PAGE. PIAS1 was detected by immunoblotting with anti-FLAG M2 antibody, and co-precipitated STAT1 was detected by anti-HA antibody (clone 16B12; BabCO).

RESULTS

p38MAPK activation enhances SUMO-1 conjugation to STAT1

MAPKs and serine phosphorylation have been implicated in both positive and negative regulation of SUMOylation [23,24,26,27]. The role of the p38MAPK pathway in phosphorylation of Ser727 in STAT1 prompted us to examine whether p38MAPK is involved in regulation of SUMO-1 modification of STAT1. To investigate the effect of p38MAPK activation on STAT1 SUMOylation, COS-7 cells were transiently transfected with increasing amounts of constitutively active mutant of MKK6 [MKK6b (E)] plasmid together with HA-tagged STAT1-WT and SUMO-1 plasmids. After 36 h, the cells were left unstimulated or stimulated with IFN-γ for 30 min. Total cell lysates were resolved by SDS/PAGE and immunoblotted with anti-HA, anti-P-S727-STAT1, anti-P-p38MAPK, anti-p38MAPK and anti-SUMO-1 antibodies. Expression of MKK6b (E) resulted in activation of p38MAPK and induction of Ser727 phosphorylation of STAT1 (Figure 1). In addition, expression of MKK6b (E) also enhanced SUMO-1 conjugation to STAT1 (Figure 1). The levels of SUMOylation as well as Ser727 phosphorylation of STAT1 were further increased after IFN-γ stimulation, thus supporting the concept of combinatorial input of different signalling pathways in regulation of STAT1. Importantly, immunoblotting with anti-P-S727-STAT1 antibody detected STAT1–SUMO-1 moiety, indicating that the SUMOylated form of STAT1 is also Ser727-phosphorylated (Figure 1; anti-P-S727-STAT1 blot).

Figure 1 MKK6 co-transfection enhances SUMO-1 conjugation to STAT1

COS-7 cells were transiently transfected with STAT1-WT–HA together with SUMO-1 and increasing amounts of MKK6 (0.05, 0.1 and 0.3 μg). After 36 h, the cells were treated with IFN-γ and lysed as indicated in the Materials and methods section. Total cell lysates were immunoblotted with anti-SUMO-1, stripped and blotted with anti-P-S727-STAT1, anti-HA, anti-P-p38MAPK and anti-p38MAPK antibodies.

Inhibition of p38MAPK activity abrogates the enhancement of STAT1 SUMOylation

To confirm that p38MAPK is mediating Ser727 phosphorylation and the enhanced SUMOylation of STAT1, we made use of dominant-negative form of p38MAPK (p38MAPK-DN) and the p38MAPK-specific inhibitor SB202190. COS-7 cells were transiently transfected with HA-tagged STAT1-WT, SUMO-1 and MKK6b (E) plasmids as indicated, and activation of the p38MAPK pathway, SUMOylation of STAT1 as well as Ser727 phosphorylation were analysed by immunoblotting with anti-HA, anti-P-S727-STAT1, anti-P-p38MAPK and anti-p38MAPK antibodies respectively. As shown in Figures 2(A) and 2(B), expression of MKK6b (E) resulted in activation of p38MAPK (anti-P-p38MAPK blot; Figures 2A, lane 2, and 2B, lanes 2–4) and enhanced Ser727 phosphorylation (anti-P-S727-STAT1 blot; Figures 2A and 2B, lane 2) and SUMOylation of STAT1 (anti-SUMO-1 blot; Figures 2A and 2B, lane 2). Furthermore, expression of increasing amounts of p38MAPK-DN inhibited activation of p38MAPK and inhibited both Ser727 phosphorylation and SUMO-1 conjugation to STAT1 (Figure 2A, lanes 3 and 4). Similar results were obtained using the p38MAPK inhibitor SB202190. As shown in Figure 2(B), enhanced Ser727 phosphorylation and SUMO-1 conjugation to STAT1 were reduced by the treatment with SB202190.

Figure 2 Inhibition of p38MAPK activation decreases SUMO-1 conjugation to STAT1

(A) p38MAPK dominant negative (DN) decreases SUMOylation of STAT1. COS-7 cells were transiently transfected with STAT1-WT–HA, SUMO-1 and MKK6b (E) plasmids together with increasing amounts of p38MAPKDN (0.1 and 0.3 μg) as indicated. After 36 h, the cells were lysed as indicated in the Materials and methods section. Total cell lysates were analysed by immunoblotting using anti-HA, anti-P-S727-STAT1, anti-P-p38MAPK and anti-p38MAPK antibodies. (B) p38MAPK-specific inhibitor SB202190 blocks the enhancement of SUMO-1 conjugation to STAT1. COS-7 cells were transiently transfected with STAT1-WT–HA together with SUMO-1 and MKK6b (E) plasmids. After 36 h, the cells were treated as indicated. Total cell lysates were analysed as in (A). (C) Osmotic shock enhances SUMO-1 conjugation to STAT1 through p38MAPK activation. COS-7 cells were transiently transfected with STAT1-WT–HA and SUMO-1. After 36 h, the cells were treated as indicated. Cells were lysed and analysed by immunoblotting as described in (A).

To investigate whether stress-induced activation of p38MAPK affects SUMO-1 conjugation to STAT1, COS-7 cells were transiently transfected with HA-tagged STAT1-WT and SUMO-1 plasmids, and subjected to osmotic shock treatment using 0.3 M sorbitol (Figure 2C) for the indicated time points. Total cell lysates were resolved by SDS/PAGE and immunoblotted with anti-HA, anti-P-S727-STAT1, anti-P-p38MAPK and anti-p38MAPK antibodies. Sorbitol treatment enhanced Ser727 phosphorylation of STAT1 (anti-P-S727-STAT1 blot; Figure 2C, lanes 2 and 3) and activation of p38MAPK (anti-P-p38MAPK blot; Figure 2C, lanes 2 and 3), which correlated with the increase in SUMO-1 conjugation to STAT1. Furthermore, SB202190 treatment inhibited the enhanced SUMOylation of STAT1 after sorbitol treatment (anti-SUMO-1 blot; Figure 2C, lane 4). Collectively, these results demonstrate that activation of the p38MAPK pathway enhances SUMOylation of STAT1 in stress responses.

ERK1/2 activation induces Ser727 phosphorylation and SUMOylation of STAT1

The reported phosphorylation of STAT1 Ser727 by ERK1/2 [30,31] prompted us to investigate whether SUMOylation of STAT1 could also be modulated by ERK1/2. To investigate the effect of different MAPK pathways on Ser727 phosphorylation and SUMOylation of STAT1, COS-7 cells were transiently transfected with constitutively active upstream kinases MKK6b (E), Raf-1 or MKK1 together with HA-tagged STAT1-WT and SUMO-1 plasmids. SUMOylation of STAT1, Ser727 phosphorylation and p38MAPK activation were analysed by immunoblotting with anti-HA, anti-P-Ser727-STAT1, anti-P-p38MAPK and anti-p38MAPK antibodies respectively. As shown in Figure 3(A), expression of MKK6 and Raf-1, but not MKK1, resulted in enhanced Ser727 phosphorylation (anti-P-S727-STAT1 blot; lanes 2 and 4) and SUMOylation of STAT1 (anti-SUMO-1 blot; lanes 2 and 4). As expected, phosphorylation of p38MAPK was induced only in response to MKK6 activation (anti-P-p38MAPK blot; lane 2). Ectopic expression of Raf-1 enhanced ERK1/2 phosphorylation, Ser727 phosphorylation and SUMOylation of STAT1 (Figure 3A, lane 4). Next, we aimed to investigate the effect of the MEK (MAPK/ERK kinase) inhibitor PD98059 on Raf-1-induced phosphorylation and SUMOylation of STAT1. For this purpose, COS-7 cells were transiently transfected with HA-tagged STAT1-WT, SUMO-1 and Raf-1 or MKK6b (E) plasmids as indicated. Treatment of cells with PD98059 or SB202190 inhibitors (Figure 3B, lanes 3 and 6) reduced phosphorylation of Ser727 and ERK1/2 as well as SUMOylation of STAT1. Furthermore, SB202190 or PD98059 treatments also inhibited the sorbitol-induced enhancement of SUMOylation and Ser727 phosphorylation of STAT1 (Figure 3C, lanes 5 and 6). These findings demonstrate that Ser727 phosphorylation and SUMOylation of STAT1 are also modulated by activation of the ERK1/2 pathway.

Figure 3 p38MAPK and ERK1/2 activation induces Ser727 phosphorylation and SUMOylation of STAT1

(A) Expression of MKK6 (E) and Raf-1 induces Ser727 phosphorylation and enhanced SUMOylation of STAT1. COS-7 cells were transiently transfected with STAT1-WT–HA together with SUMO-1. Constitutively active forms of MKK6b (E), MKK1 and Raf-1 plasmids were transfected as indicated. After 36 h, the cells were lysed as indicated in the Materials and methods section, and total cell lysates were immunoblotted with anti-P-S727-STAT1, anti-HA, anti-P-p38MAPK and anti-P-ERK1/2 antibodies. (B) Enhancement of SUMOylation by Raf-1 is counteracted by PD98059 treatment. COS-7 cells were transiently transfected with STAT1-WT–HA together with SUMO-1 and Raf-1 or MKK6b (E) plasmids. After 36 h, the cells were left untreated or treated for 60 min with PD98059 or SB202190 as indicated. Cells were lysed and total cell lysates were analysed by immunoblotting as described in (A). (C) Inhibition of p38MAPK and ERK1/2 pathways reduces the stress-induced SUMOylation of STAT1. COS-7 cells were transiently transfected with STAT1-WT–HA together with SUMO-1 plasmid. After 36 h, the cells were subjected to sorbitol treatment as indicated. Samples in lanes 5 and 6 were treated for 60 min with PD98059 or SB202190 inhibitors as indicated. Cells were lysed and total cell lysates were analysed as described in (A).

Phosphorylation of Ser727 in STAT1 is required for p38MAPK-induced SUMO-1 conjugation

Next, we wished to determine whether the observed effect of MAPK-mediated enhancement of SUMO-1 conjugation to STAT1 was directly regulated by phosphorylation of Ser727. For this purpose, we made use of different STAT1 mutants, SUMOylation-deficient mutant STAT1-K703R–HA, tyrosine-phosphorylation mutant STAT1-Y701F–HA and serine-phosphorylation mutant STAT1-S727A–HA, and analysed the level of SUMO-1 conjugation and phosphorylation of Ser727 by Western blotting. The results in Figure 4(A) show that STAT1-WT as well as the STAT1-S727A and STAT1-Y701F mutants were all SUMOylated at low basal level. Expression of MKK6b (E) induced SUMOylation of STAT1-Y701F–HA at similar levels (lanes 7–9) as STAT1-WT–HA (lanes 1–3), indicating that tyrosine phosphorylation itself was not required for SUMOylation. However, expression of MKK6b (E) failed to enhance SUMOylation of the S727A mutant (lanes 10–12). The critical role of Ser727 in MAPK-induced SUMOylation was confirmed in experiments using sorbitol treatment. Conjugation of SUMO-1 to STAT1-WT–HA was increased after sorbitol treatment, and this effect was counteracted by treatment with SB202190 inhibitor (Figure 4B, lanes 1–4). In contrast, sorbitol treatment failed to enhance SUMOylation of STAT1-S727A–HA (Figure 4B, lanes 5–8). Taken together, these results show that the effect of p38MAPK activation on SUMO-1 modification is mediated through Ser727 phosphorylation of STAT1.

Figure 4 Ser727 is required for MKK6- and stress-induced SUMOylation of STAT1

(A) Ser727 is required for MKK6b (E)-induced SUMO-1 conjugation to STAT1. COS-7 cells were transiently transfected with plasmids encoding STAT1-WT–HA (lanes 1–3), STAT1-KR–HA (lanes 4–6), STAT1-YF–HA (lanes 7–9) and STAT1-SA–HA (lanes 10–12) together with SUMO-1 and increasing amounts of MKK6b (E) plasmid. After 36 h, the cells were lysed as described in the Materials and methods section, and immunoblotting was performed using anti-HA, anti-P-S727-STAT1, anti-P-p38MAPK and anti-p38MAPK antibodies. (B) Stress-induced activation of MAPKs modulates SUMO-1 conjugation differently between STAT1-WT and STAT1-Ser727 mutant. COS-7 cells were transiently transfected with plasmids encoding STAT1-WT–HA (lanes 1–4) or STAT1-S727A–HA (lanes 5–8) and SUMO-1. After 36 h, the cells were treated with sorbitol for the indicated time points and left untreated or treated with SB202190 as indicated (lanes 4 and 8). Cells were lysed and immunoblotted as described in (A).

Ser727 phosphorylation promotes PIAS1 binding to STAT1

To validate the role of Ser727 phosphorylation in regulation of SUMOylation of STAT1, we constructed two mutations, S727D and S727E, that mimic phosphorylation of the serine residue. COS-7 cells were transiently transfected with HA-tagged STAT1-WT, STAT1-S727E, STAT1-S727D or STAT1-S727A together with SUMO-1 plasmid and subjected to treatment with osmotic shock using 0.3 M sorbitol or left untreated as indicated. SUMOylation of STAT1, Ser727 phosphorylation, p38MAPK and p-p38MAPK levels were analysed by immunoblotting. As expected, osmotic shock enhanced SUMOylation of STAT1–WT (Figure 5A, lanes 1 and 2), but did not affect SUMOylation of the S727A mutant (lanes 7 and 8). S727D and S727E mutants showed clearly enhanced SUMOylation in untreated samples, which was not markedly affected by osmotic shock (lanes 3–6). These findings provide strong evidence for a direct function of Ser727 phosphorylation as a regulator of STAT1 SUMOylation.

Figure 5 Phosphorylation of Ser727 enhances SUMOylation of STAT1 and interaction with PIAS1

(A) Phospho-Ser727 mimic mutants STAT1-S727E and STAT1-S727D show enhanced SUMOylation. COS-7 cells were transiently transfected with plasmids encoding STAT1-WT–HA (lanes 1 and 2), STAT1-S727E–HA (lanes 3 and 4), STAT1-S727D–HA (lanes 5 and 6) or STAT1-S727A–HA (lanes 7 and 8) and SUMO-1. After 36 h, the cells were treated with sorbitol as indicated. Cells were lysed and immunoblotted as described in Figure 4(A). (B) PIAS1 readily binds to serine-phosphorylated STAT1. COS-7 cells were transiently transfected with plasmids encoding STAT1-WT–HA (lanes 1, 2 and 7), STAT1-S727D–HA (lanes 2–4) or STAT1-S727A–HA (lanes 5 and 6) together with PIAS1 and SUMO-1 (lanes 1–7). After 36 h, the cells were treated with sorbitol and IFN-γ as indicated, and cells were lysed as described in the Materials and methods section. PIAS1 was immunoprecipitated using anti-FLAG antibody. Co-precipitated STAT1 was analysed using anti-HA blotting, and total cell lysates were immunoblotted using anti-HA and anti-P-S727-STAT1 antibodies.

Next we wanted to investigate the underlying mechanism of Ser727 phosphorylation in regulation of SUMOylation, and a plausible mechanism could involve modulation of interaction between STAT1 and the SUMO-E3-ligase PIAS1. For this purpose, COS-7 cells were transiently transfected with HA-tagged STAT1–WT, STAT1–S727D or STAT1–S727A together with SUMO-1 and FLAG-tagged PIAS1. Cells were either left untreated or treated with IFN-γ (50 ng/ml) and sorbitol (0.3 M) for 30 min to ensure maximal phosphorylation of STAT1. PIAS1 was immunoprecipitated from the lysates with anti-FLAG antibody. SUMOylation of STAT1, Ser727 phosphorylation and PIAS1 levels were analysed by immunoblotting with anti-HA, anti-P-Ser727-STAT1 and anti-FLAG antibodies. The results in Figure 5(B) show that, in untreated samples, PIAS1 and STAT1 did not co-immunoprecipitate, but after stimulation STAT1 and PIAS1 are detected in the same complex (lanes 1 and 2). The phosphomimetic S727D mutant, on the other hand, co-immunoprecipitated with PIAS1 already in unstimulated cells and stimulation further increased the protein complex formation between PIAS1 and STAT1-S727D (lanes 3 and 4). In contrast, S727A mutant of STAT1 failed to significantly co-immunoprecipitate with PIAS1 even after stimulation (lanes 5 and 6). Taken together, these results indicate that phosphorylation of Ser727 enhances the interaction between STAT1 and PIAS1, which provides a likely explanation for the role of Ser727 phosphorylation in enhancing the SUMOylation of STAT1.

DISCUSSION

The function of STAT1 is regulated by various post-translational modifications that act in a highly co-ordinated fashion to combine signals from secreted cytokines as well as from microbial agents. IFN-γ stimulation induces Tyr701 phosphorylation and DNA binding of STAT1, but the transcriptional response is also regulated by phosphorylation of Ser727 in the TAD. Phosphorylation of Tyr701 and Ser727 are independent events, but they act in concert, as tyrosine phosphorylation and presumably dimer formation of STAT1 enhance the serine phosphorylation [3,4,7]. Recently, SUMOylation has emerged as another post-translational modification that targets STAT1, but understanding of the function and regulation of SUMO modification of STAT1 has remained incomplete [18,19,32,33]. The present study was aimed to investigate the regulation of SUMO-1 modification of STAT1, and our results demonstrate that MAPK-induced phosphorylation of Ser727 modulates SUMO-1 conjugation and PIAS1 binding to STAT1.

Sumoylation is emerging as a relatively common post-translational modification for several proteins implicated mainly in transcriptional regulation. A rational approach to obtain insight into the mechanism of SUMOylation is to analyse the regulatory mechanisms involved in this modification. Here, we show that activation of the p38MAPK pathway, either through expression of activating upstream kinase (MKK6) or through external stimulations, results in enhanced SUMOylation of STAT1. The target for this regulation was Ser727 in the TAD of STAT1. Also expression of ERK1/2 upstream activator resulted in increased Ser727 phosphorylation and SUMOylation of STAT1, thus suggesting that activation of ERK1/2 might modulate SUMOylation of STAT1 in growth factor responses.

In several proteins, SUMOylation has been found to be regulated by phosphorylation. This regulation is well exemplified by HSF1, where phosphorylation of Ser303 stimulates SUMOylation of an adjacent Lys298 [34], presumably by inducing a conformation change that favours the conjugation reaction. The SUMOylation site in HSF1 was shown to form phosphorylation-dependent SUMOylation motif (ΦKXEXXSP, where Φ is a hydrophobic amino acid), which is also found in may other SUMO substrates [27]. Also STAT1 SUMOylation site (I702KTELISV709) resembles the phosphorylation-dependent SUMOylation motif. Recently, Ser708 in STAT1 was found to become phosphorylated by IKKϵ (IκB kinase ϵ) kinase that was implicated in IFN-inducible antiviral transcriptional response [35]. However, mutation of Ser708 to alanine did not affect the SUMOylation of STAT1 (S. Vanhatupa and O. Silvennoinen, unpublished work), indicating that Ser727 is the only phosphorylated residues in STAT1 involved in regulation of SUMOylation.

Phosphorylation of Ser727 is generally considered to have a stimulatory effect on STAT1-mediated gene responses. However, phosphorylation of Ser727 in STAT1 has been shown to differentially affect basal and induced expression levels of target genes [3], but the underlying mechanism for the selective regulatory role of Ser727 in IFN-γ-induced gene responses is currently not known. Thus, although Ser727 phosphorylation promotes SUMOylation of STAT1, these modifications have distinct functional roles during immune response. Ser727 phosphorylation precedes the relatively slowly emerging SUMOylation of STAT1, suggesting that Ser727 phosphorylation might induce negative feedback loop through SUMO conjugation. Recently, the functional connection between SUMOylation and PIAS1 received further strengthening, when IKKα-induced phosphorylation of PIAS1 was shown to repress activation of STAT1-regulated genes in a SUMOylation-dependent manner [36]. Our present results indicate that phosphorylation of Ser727 modulates the interaction between PIAS1 and STAT1, either directly or indirectly, and thus provides an explanation for the Ser727-phosphorylation-regulated SUMOylation of STAT1. The Ser727-phosphorylation-promoted recruitment of PIAS1 and SUMOylation of STAT1 is likely to be involved in securing the tight regulation of STAT1 activity at the transcriptional level and mediates the fine-tuning of STAT1 signalling in immune responses.

STAT1 functions as a critical regulator of the innate immunity and immune responses during microbial infections. Microbial infections induce Ser727 phosphorylation through the p38MAPK pathway, positioning Ser727 as a converging point in immune surveillance. STAT1 is an essential protein in combining the signals from the infectious agents and from the immune modulator cytokines. Our present results showing that PIAS1 interaction and SUMO-1 modification of STAT1 are regulated in a Ser727-phosphorylation-dependent manner identify another mechanism for co-ordinated post-translational regulation of STAT1.

Acknowledgments

We thank Dr N. Ahn, Dr I. Kerr (Biochemical Regulatory Mechanism Laboratory, Cancer Research UK London Research Institute, London, U.K.), Dr U. Rapp and J. Westermarck for reagents and Merja Lehtinen and Paula Kosonen for technical assistance. This work was supported by the Medical Research Council of Academy of Finland, Medical Research Fund of Tampere University Hospital, the Finnish Foundation for Cancer Research, the Sigrid Juselius Foundation, the Finnish Cancer Foundation and the Tuberculosis Foundation of Tampere.

Abbreviations: ELK, ETS (E twenty-six)-like kinase; ERK, extracellular-signal-regulated kinase; HA, haemagglutinin; HSF1, heat-shock factor 1; IFN, interferon; IκB, inhibitory κB; IKKϵ, IκB kinase ϵ; MAPK, mitogen-activated protein kinase; MKK, MAPK kinase; NEM, N-ethylmaleimide; STAT1, signal transducer and activator of transcription 1; PIAS, protein inhibitor of activated STAT; SUMO, small ubiquitin-related modifier; TAD, transactivation domain; TBS, Tris-buffered saline; WT, wild-type

References

View Abstract