Pc2-mediated Sumoylation of Smad-interacting Protein 1 Attenuates Transcriptional Repression of E-cadherin
2005; Elsevier BV; Volume: 280; Issue: 42 Linguagem: Inglês
10.1074/jbc.m504477200
ISSN1083-351X
AutoresJianyin Long, Dongmei Zuo, Morag Park,
Tópico(s)Cancer-related gene regulation
ResumoEpithelial-mesenchymal transition (EMT) is important in embryonic development and tumorigenesis. Smad-interacting protein 1 (SIP1) can induce EMT by repressing the transcription of E-cadherin through recruitment of the corepressor C-terminal-binding protein (CtBP). How the activity of SIP1 is regulated still remains unclear. Here we show in vivo and in vitro that SIP1 is covalently modified by sumoylation at two conserved sites, Lys391 and Lys866. The polycomb protein Pc2, but not the PIAS (protein inhibitor of activated STAT) family proteins, acts as a Small ubiquitin-like modifier E3 ligase for SIP1. Sumoylation of SIP1 does not affect its subcellular localization, but regulates its transcriptional activity. Compared with the wild-type, a SIP1 sumoylation null mutant shows more potent repression on E-cadherin transcription but similar repression on two transforming growth factor-β-responsive reporter genes and comparable activation on vitamin D3 receptor transcription. Coexpression of SIP1 with Pc2 can partially relieve E-cadherin repression by SIP1. We further show that SIP1 sumoylation disrupts the recruitment of CtBP. Thus SIP1 sumoylation regulates its transcriptional activity in a promoter context-dependent manner and may represent an important intervention target to modulate EMT in tumorigenesis. Epithelial-mesenchymal transition (EMT) is important in embryonic development and tumorigenesis. Smad-interacting protein 1 (SIP1) can induce EMT by repressing the transcription of E-cadherin through recruitment of the corepressor C-terminal-binding protein (CtBP). How the activity of SIP1 is regulated still remains unclear. Here we show in vivo and in vitro that SIP1 is covalently modified by sumoylation at two conserved sites, Lys391 and Lys866. The polycomb protein Pc2, but not the PIAS (protein inhibitor of activated STAT) family proteins, acts as a Small ubiquitin-like modifier E3 ligase for SIP1. Sumoylation of SIP1 does not affect its subcellular localization, but regulates its transcriptional activity. Compared with the wild-type, a SIP1 sumoylation null mutant shows more potent repression on E-cadherin transcription but similar repression on two transforming growth factor-β-responsive reporter genes and comparable activation on vitamin D3 receptor transcription. Coexpression of SIP1 with Pc2 can partially relieve E-cadherin repression by SIP1. We further show that SIP1 sumoylation disrupts the recruitment of CtBP. Thus SIP1 sumoylation regulates its transcriptional activity in a promoter context-dependent manner and may represent an important intervention target to modulate EMT in tumorigenesis. The epithelial cell-cell junction protein E-cadherin is a potent suppressor of tumor cell invasion and metastasis (1Cavallaro U. Schaffhauser B. Christofori G. Cancer Lett. 2002; 176: 123-128Crossref PubMed Scopus (174) Google Scholar, 2Wheelock M.J. Johnson K.R. Annu. Rev. Cell Dev. Biol. 2003; 19: 207-235Crossref PubMed Scopus (540) Google Scholar, 3Peinado H. Portillo F. Cano A. Int. J. Dev. Biol. 2004; 48: 365-375Crossref PubMed Scopus (489) Google Scholar). Down-regulation of E-cadherin is often accompanied by conversion from well organized epithelial cells into migratory, invasive, and fibroblast-like cells, an event collectively referred to as epithelial-mesenchymal transition (EMT). 2The abbreviations used are: EMTepithelial-mesenchymal transitionSIP1Smad-interacting protein 1SUMOSmall ubiquitin-like modifierTGF-βtransforming growth factor-βCtBPC-terminal-binding proteinHDAChistone deacetylasePIASprotein inhibitor of activated STATPIASyprotein inhibitor of activated STAT yPcGpolycomb groupPMLpromyelocytic leukemia proteinVD3Rvitamin D3 receptorMDCKMadin-Darby canine kidney cellsP/CAFp300/CBP-associated factorHAhemagglutininGSTglutathione S-transferaseE1ubiquitin-activating enzymeE2ubiquitin carrier proteinE3ubiquitin-protein isopeptide ligaseSTATsignal transducers and activators of transcriptionCMVcytomegalovirusN-ZFN-terminal zink-fingerC-ZFC-terminal zinc-fingerCIDCtBP-interacting domainYFPyellow fluorescent proteinHDAC4histone deacetylase 4HDhomeodomain. A typical EMT program comprises dissolution of tight junctions, modulation of adherent junctions, reorganization of the actin cytoskeleton, loss of apical-basal polarity, and induction of mesenchymal gene expression (4Savagner P. BioEssays. 2001; 23: 912-923Crossref PubMed Scopus (610) Google Scholar, 5Shook D. Keller R. Mech. Dev. 2003; 120: 1351-1383Crossref PubMed Scopus (478) Google Scholar). EMT is a highly conserved and fundamental process that governs morphogenesis in multicellular organisms (6Thiery J.P. Nat. Rev. Cancer. 2002; 2: 442-454Crossref PubMed Scopus (5604) Google Scholar). Multiple signaling pathways, such as receptor tyrosine kinases, small GTPases, mitogen-activated protein kinases, integrin-linked kinase, phosphatidylinositol 3-kinase, transforming growth factor-β (TGF-β), matrix-metalloproteinases, and extracellular matrix components, have been implicated in the regulation of EMT and tumor progression, but the cross-talk among these pathways and their downstream targets remains largely unknown (6Thiery J.P. Nat. Rev. Cancer. 2002; 2: 442-454Crossref PubMed Scopus (5604) Google Scholar, 7Grunert S. Jechlinger M. Beug H. Nat. Rev. Mol. Cell. Biol. 2003; 4: 657-665Crossref PubMed Scopus (575) Google Scholar). epithelial-mesenchymal transition Smad-interacting protein 1 Small ubiquitin-like modifier transforming growth factor-β C-terminal-binding protein histone deacetylase protein inhibitor of activated STAT protein inhibitor of activated STAT y polycomb group promyelocytic leukemia protein vitamin D3 receptor Madin-Darby canine kidney cells p300/CBP-associated factor hemagglutinin glutathione S-transferase ubiquitin-activating enzyme ubiquitin carrier protein ubiquitin-protein isopeptide ligase signal transducers and activators of transcription cytomegalovirus N-terminal zink-finger C-terminal zinc-finger CtBP-interacting domain yellow fluorescent protein histone deacetylase 4 homeodomain. One hallmark of EMT, loss of E-cadherin, is induced by various epigenetic mechanisms, including promoter hypermethylation as well as active transcriptional repression by several transcription factors (6Thiery J.P. Nat. Rev. Cancer. 2002; 2: 442-454Crossref PubMed Scopus (5604) Google Scholar), including Snail (8Batlle E. Sancho E. Franci C. Dominguez D. Monfar M. Baulida J. Garcia De Herreros A. Nat. Cell Biol. 2000; 2: 84-89Crossref PubMed Scopus (2215) Google Scholar, 9Cano A. Perez-Moreno M. Rodrigo I. Locascio A. Blanco M.J. del Barrio M.G. Portillo F. Nieto M.A. Nat. Cell Biol. 2000; 2: 76-83Crossref PubMed Scopus (2989) Google Scholar) and its homologue Slug (10Hajra K.M. Chen D.Y.S. Fearon E.R. Cancer Res. 2002; 62: 1613-1618PubMed Google Scholar), Smad-interacting protein 1 (SIP1, also called ZEB2, for zinc-finger E-box-binding protein 2) (11Comijn J. Berx G. Vermassen P. Verschueren K. van Grunsven L. Bruyneel E. Mareel M. Huylebroeck D. van Roy F. Mol. Cell. 2001; 7: 1267-1278Abstract Full Text Full Text PDF PubMed Scopus (1174) Google Scholar) and its homologue δEF1 (also called ZEB1) (12Eger A. Aigner K. Sonderegger S. Dampier B. Oehler S. Schreiber M. Berx G. Cano A. Beug H. Foisner R. Oncogene. 2005; 24: 2375-2385Crossref PubMed Scopus (633) Google Scholar), twist (13Yang J. Mani S.A. Donaher J.L. Ramaswamy S. Itzykson R.A. Come C. Savagner P. Gitelman I. Richardson A. Weinberg R.A. Cell. 2004; 117: 927-939Abstract Full Text Full Text PDF PubMed Scopus (3158) Google Scholar), and the basic helix-loop-helix factor E12/E47 (14Perez-Moreno M.A. Locascio A. Rodrigo I. Dhondt G. Portillo F. Nieto M.A. Cano A. J. Biol. Chem. 2001; 276: 27424-27431Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar) (reviewed in Ref. 3Peinado H. Portillo F. Cano A. Int. J. Dev. Biol. 2004; 48: 365-375Crossref PubMed Scopus (489) Google Scholar). SIP1 was originally identified from a yeast two-hybrid screen through its binding to Smad (15Verschueren K. Remacle J.E. Collart C. Kraft H. Baker B.S. Tylzanowski P. Nelles L. Wuytens G. Su M.-T. Bodmer R. Smith J.C. Huylebroeck D. J. Biol. Chem. 1999; 274: 20489-20498Abstract Full Text Full Text PDF PubMed Scopus (419) Google Scholar). SIP1 plays a crucial role in normal embryonic development of neural structures and the neural crest (16Mowat D.R. Wilson M.J. Goossens M. J. Med. Genet. 2003; 40: 305-310Crossref PubMed Scopus (176) Google Scholar). SIP1 is a member of the zfh-1 family of two-handed zinc-finger transcription factors, which also includes δEF1 (17Remacle J.E. Kraft H. Lerchner W. Wuytens G. Collart C. Verschueren K. Smith J.C. Huylebroeck D. EMBO J. 1999; 18: 5073-5084Crossref PubMed Scopus (232) Google Scholar, 18Postigo A.A. Dean D.C. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 6391-6396Crossref PubMed Scopus (119) Google Scholar). They share the unique structure of two zinc-finger clusters separated by a homeodomain, where each of the zinc-finger clusters binds to an E-box element in the promoter region of target genes (17Remacle J.E. Kraft H. Lerchner W. Wuytens G. Collart C. Verschueren K. Smith J.C. Huylebroeck D. EMBO J. 1999; 18: 5073-5084Crossref PubMed Scopus (232) Google Scholar, 18Postigo A.A. Dean D.C. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 6391-6396Crossref PubMed Scopus (119) Google Scholar). In addition to Smad-binding domains, zfh-1 family members also contain consensus binding motifs for the corepressor CtBP (19Furusawa T. Moribe H. Kondoh H. Higashi Y. Mol. Cell. Biol. 1999; 19: 8581-8590Crossref PubMed Google Scholar, 20Postigo A.A. Dean D.C. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 6683-6688Crossref PubMed Scopus (231) Google Scholar). SIP1 and δEF1 were both shown to be present in a CtBP corepressor core complex, which binds to the E-cadherin promoter and represses its transcription (21Shi Y. Sawada J.-i. Sui G. Affar E.B. Whetstine J.R. Lan F. Ogawa H. Po-Shan L uke M. Nakatani Y. Shi Y. Nature. 2003; 422: 735-738Crossref PubMed Scopus (656) Google Scholar). However, it seems that SIP1 might also be able to repress E-cadherin independent of CtBP binding (22van Grunsven L.A. Michiels C. Van de Putte T. Nelles L. Wuytens G. Verschueren K. Huylebroeck D. J. Biol. Chem. 2003; 278: 26135-26145Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar). Besides E-cadherin, SIP1 and δEF1 have been shown to repress the transcription of many other genes depending on the cell types (23Postigo A.A. EMBO J. 2003; 22: 2443-2452Crossref PubMed Scopus (227) Google Scholar). This includes suppression of interleukin 2, immunoglobulin μ heavy chain, CD4, GATA-3, and α4 integrin in hematopoietic cells; inhibition of p73 gene expression in mesenchymal cells, and repression of type I and type II collagen expression in osteoblasts. δEF1 can also activate the vitamin D3 receptor (VD3R) gene (24Lazarova D.L. Bordonaro M. Sartorelli A.C. Cell Growth & Differ. 2001; 12: 319-326PubMed Google Scholar) and ovalbumin (25Dillner N.B. Sanders M.M. Mol. Cell. Endocrinol. 2002; 192: 85-91Crossref PubMed Scopus (26) Google Scholar), although the detailed mechanism is unknown. Recently, SIP1 and δEF1 were shown to exert opposing effects on TGF-β/Smad signaling through the recruitment of coactivators p300 and/or P/CAF and displacement of CtBP by δEF1 (23Postigo A.A. EMBO J. 2003; 22: 2443-2452Crossref PubMed Scopus (227) Google Scholar, 26Postigo A.A. Depp J.L. Taylor J.J. Kroll K.L. EMBO J. 2003; 22: 2453-2462Crossref PubMed Scopus (292) Google Scholar). Smad is the key signal transducer of TGF-β, which plays a pivotal role in multiple cellular processes, including proliferation, differentiation, adhesion, apoptosis, and importantly tumorigenesis (27Siegel P.M. Massagué J. Nat. Rev. Cancer. 2003; 3: 807-820Crossref PubMed Scopus (1372) Google Scholar). Phosphorylation of Smad2 and Smad3 by the ligand-activated transmembrane serine/threonine kinase receptors results in heteromeric complex formation with Smad4, and translocation into the nucleus, where transcription of target genes is regulated through direct DNA binding and/or the recruitment of transcriptional coactivators or corepressors (28Attisano L. Wrana J.L. Science. 2002; 296: 1646-1647Crossref PubMed Scopus (1148) Google Scholar, 29Dijke P.T. Hill C.S. Trends Biochem. Sci. 2004; 29: 265-273Abstract Full Text Full Text PDF PubMed Scopus (1061) Google Scholar, 30Shi Y. Massague J. Cell. 2003; 113: 685-700Abstract Full Text Full Text PDF PubMed Scopus (4941) Google Scholar). Sumoylation is a covalent modification that adds small ubiquitin-like modifier (SUMO) to lysine residues. It occurs at a consensus motif ΨKX(D/E), where Ψ is a hydrophobic amino acid (31Hay R.T. Mol. Cell. 2005; 18: 1-12Abstract Full Text Full Text PDF PubMed Scopus (1360) Google Scholar, 32Johnson E.S. Annu. Rev. Biochem. 2004; 73: 355-382Crossref PubMed Scopus (1407) Google Scholar). This modification involves the coordinated action of multiple enzymes, in a manner very similar to ubiquitination. These enzymes include E1 SUMO-activating enzyme (heterodimer of SAE1/SAE2), E2-conjugating enzyme Ubc9, and an E3 ligase, which promotes the transfer of SUMO from Ubc9 to specific proteins (31Hay R.T. Mol. Cell. 2005; 18: 1-12Abstract Full Text Full Text PDF PubMed Scopus (1360) Google Scholar, 32Johnson E.S. Annu. Rev. Biochem. 2004; 73: 355-382Crossref PubMed Scopus (1407) Google Scholar). The protein inhibitor of activated STAT (PIAS) family proteins (33Johnson E.S. Gupta A.A. Cell. 2001; 106: 735-744Abstract Full Text Full Text PDF PubMed Scopus (536) Google Scholar, 34Kahyo T. Nishida T. Yasuda H. Mol. Cell. 2001; 8: 713-718Abstract Full Text Full Text PDF PubMed Scopus (396) Google Scholar, 35Sachdev S. Bruhn L. Sieber H. Pichler A. Melchior F. Grosschedl R. Genes Dev. 2001; 15: 3088-3103Crossref PubMed Scopus (470) Google Scholar), nucleoporin protein RanBP2 (36Pichler A. Gast A. Seeler J.S. Dejean A. Melchior F. Cell. 2002; 108: 109-120Abstract Full Text Full Text PDF PubMed Scopus (656) Google Scholar), and polycomb protein Pc2 (37Kagey M.H. Melhuish T.A. Wotton D. Cell. 2003; 113: 127-137Abstract Full Text Full Text PDF PubMed Scopus (462) Google Scholar) have been identified as SUMO E3 ligases. Sumoylation is negatively regulated by multiple proteases, which remove SUMO from its substrates (38Melchior F. Schergaut M. Pichler A. Trends Biochem. Sci. 2003; 28: 612-618Abstract Full Text Full Text PDF PubMed Scopus (340) Google Scholar). Unlike ubiquitination, which signals proteins for degradation by the proteasome or enhances trafficking of transmembrane proteins for degradation in the lysosome, protein sumoylation has been shown to regulate a variety of activities, including protein stability, subcellular localization, transcriptional regulation, DNA repair, as well as genome integrity (31Hay R.T. Mol. Cell. 2005; 18: 1-12Abstract Full Text Full Text PDF PubMed Scopus (1360) Google Scholar, 32Johnson E.S. Annu. Rev. Biochem. 2004; 73: 355-382Crossref PubMed Scopus (1407) Google Scholar, 39Gill G. Genes Dev. 2004; 18: 2046-2059Crossref PubMed Scopus (633) Google Scholar, 40Müller S. Ledl A. Schmidt D. Oncogene. 2004; 23: 1998-2008Crossref PubMed Scopus (247) Google Scholar). In this report, we show that SIP1 and its homologue δEF1 are sumoylated. We have mapped the sumoylation sites to two conserved lysine residues, Lys391 and Lys866, in SIP1, and show that polycomb protein Pc2, acts as a SUMO E3 ligase for SIP1. Importantly, a sumoylation negative mutant of SIP1 shows more potent repression toward E-cadherin. We further show that Pc2-mediated sumoylation can partially relieve the E-cadherin repression by SIP1. Our results provide an example of how sumoylation regulates transcription in a promoter context-dependent manner. Constructs—Human SIP1 cDNA was amplified by PCR from clone pBluescript-KIAA0569 (kindly provided by T. Nagase), human δEF1 cDNA was assembled through PCR from two overlapping expressed sequence tag clones (National Institutes of Health Image clones 4245215 and 6180372, Open Biosystems). Myc (six copies), HA, FLAG, GFP, GAL4-tagged SIP1, and δEF1 were constructed by subcloning the cDNAs into pCS3–6Myc, pcDNA3.1-HA, pCMV5-FLAG, pEGFP-C2 (BD Biosciences), and pSG424, respectively. YFP-SIP1 was constructed by first replacing SUMO1 coding region in YFP-SUMO1 (41Weger S. Hammer E. Engstler M. Exp. Cell Res. 2003; 290: 13-27Crossref PubMed Scopus (43) Google Scholar) and then transferring the YFP-SIP1 fusion region into pCMV5. Myc-tagged SIP1 deletion mutants were constructed by restriction enzymes digestion. SIP1 domains were subcloned into pGEX-4T-1 (Amersham Biosciences) to generate GST fusion proteins for use in GST pull-down assays. One or more different SIP1 and δEF1 sumoylation sites mutants, as well as the CtBP-binding mutant of SIP1 (where four consensus motifs were simultaneously mutated as previously (22van Grunsven L.A. Michiels C. Van de Putte T. Nelles L. Wuytens G. Verschueren K. Huylebroeck D. J. Biol. Chem. 2003; 278: 26135-26145Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar)), were made by using the QuikChange System (Stratagene) or PCR-based mutagenesis. All constructs and mutants were confirmed by sequencing. Myc-Smad4 and Myc-Smad4 (K113/159R), FLAG-Smad2 and FLAG-Smad3, FLAG-SUMO1 and 2FLAG-SUMO1 (34Kahyo T. Nishida T. Yasuda H. Mol. Cell. 2001; 8: 713-718Abstract Full Text Full Text PDF PubMed Scopus (396) Google Scholar), Myc-SUMO1, and FLAG-PIAS3 and FLAG-PIASy (42Liu B. Liao J. Rao X. Kushner S.A. Chung C.D. Chang D.D. Shuai K. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10626-10631Crossref PubMed Scopus (638) Google Scholar) were as previously described (43Long J. Wang G. He D. Liu F. Biochem. J. 2004; 379: 23-29Crossref PubMed Scopus (95) Google Scholar, 44Long J. Wang G. Matsuura I. He D. Liu F. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 99-104Crossref PubMed Scopus (85) Google Scholar). FLAG-Pc2 and CFP-Pc2 (37Kagey M.H. Melhuish T.A. Wotton D. Cell. 2003; 113: 127-137Abstract Full Text Full Text PDF PubMed Scopus (462) Google Scholar), and FLAG-HDAC4 (45Gregoire S. Yang X.-J. Mol. Cell. Biol. 2005; 25: 2273-2287Crossref PubMed Scopus (175) Google Scholar), were as described. FLAG-RanBP2 was constructed by PCR from BP2 ΔFG (36Pichler A. Gast A. Seeler J.S. Dejean A. Melchior F. Cell. 2002; 108: 109-120Abstract Full Text Full Text PDF PubMed Scopus (656) Google Scholar). FLAG-CtBP was kindly provided by J. White (McGill University). Antibodies, Cell Lines, Transfection, and Immunoprecipitation— HDAC4 antisera (46Wang A.H. Kruhlak M.J. Wu J. Bertos N.R. Vezmar M. Posner B.I. Bazett-Jones D.P. Yang X.-J. Mol. Cell. Biol. 2000; 20: 6904-6912Crossref PubMed Scopus (233) Google Scholar) were from X-J. Yang. Other antibodies used are: anti-Myc (BD Biosciences), anti-FLAG and anti-His (Sigma), anti-HA (16B12) (BAbCo), anti-GFP (Molecular Probes), anti-SUMO1 (GMP-1) (Zymed Laboratories Inc.), anti-GAL4 (RK5C1), anti-actin (I-19), anti-Mel-18 (C-20), anti-CtBP (H-440) (Santa Cruz Biotechnology). HeLa, HEK293T, COS-1, Mv1Lu/L17 (mink-lung epithelial cells), and Madin-Darby canine kidney (MDCK) cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (Invitrogen). Transient transfections were performed using Lipofectamine (Invitrogen). Immunoprecipitations were usually carried out as previously described (44Long J. Wang G. Matsuura I. He D. Liu F. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 99-104Crossref PubMed Scopus (85) Google Scholar) from cells lysed in TNTE buffer (10 mm Tris HCl, pH 7.8, 150 mm NaCl, 1 mm EDTA, 1.0% Nonidet P-40) plus protease inhibitors and phosphatase inhibitors. FLAG-SIP1-bound proteins were purified from FLAG-SIP1-transfected 293T cells by anti-FLAG-M2 affinity gel (Sigma) and eluted with 150 μg/ml 3×FLAG peptide (Sigma). In Vivo Sumoylation Assay—In vivo sumoylation assay of SIP1 was carried out the same as previously described (43Long J. Wang G. He D. Liu F. Biochem. J. 2004; 379: 23-29Crossref PubMed Scopus (95) Google Scholar). Briefly, COS-1 cells were transiently transfected and lysed in modified radioimmune precipitation assay/SDS buffer (43Long J. Wang G. He D. Liu F. Biochem. J. 2004; 379: 23-29Crossref PubMed Scopus (95) Google Scholar, 47Desterro J. Rodriguez M. Hay R.T. Mol. Cell. 1998; 2: 233-239Abstract Full Text Full Text PDF PubMed Scopus (929) Google Scholar). The lysis buffer is a 3:1 mixture of radioimmune precipitation assay buffer (25 mm Tris-HCl, pH 8.2, 50 mm NaCl, 0.5% sodium deoxycholate, 0.5% Nonidet P-40, 0.1% SDS, 0.1% sodium azide) and SDS sample buffer (5% SDS, 0.15 m Tris-HCl, pH 6.8, 30% glycerol), plus 10 mm N-ethylmaleimide, protease inhibitors, and phosphatase inhibitors. Cells were then sonicated briefly and boiled for 10 min. The clear lysates from centrifugation were separated on SDS-PAGE and immunoblotted against appropriate primary antibody (anti-Myc or anti-HA). To confirm the in vivo sumoylation, lysates were diluted 10-fold into phosphate-buffered saline/0.5% Nonidet P-40 and immunoprecipitated with anti-Myc antibody followed by immunoblot with anti-FLAG antibody. In Vitro Sumoylation Assay—The 35S-labeled, Myc-tagged wild-type, and mutant SIP1 were synthesized by in vitro translation using the SP6 TnT Quick Coupled Transcription/Translation System (Promega). Purification of GST-SUMO1 (48Tatham M.H. Jaffray E. Vaughan O.A. Desterro J.M.P. Botting C.H. Naismith J.H. Hay R.T. J. Biol. Chem. 2001; 276: 35368-35374Abstract Full Text Full Text PDF PubMed Scopus (666) Google Scholar), Ubc9, E1, and in vitro sumoylation reactions were carried out as previously described (43Long J. Wang G. He D. Liu F. Biochem. J. 2004; 379: 23-29Crossref PubMed Scopus (95) Google Scholar). Alternatively, 35S-labeled Myc-SIP1 was incubated with an in vitro sumoylation control kit (LAE Biotech, Rockville, MD), in the absence or presence of increasing amount of bacterially purified His-Pc2 protein (AmProx, Carlsbad, CA). Reaction products were separated on 4–10% gradient gel and analyzed by autoradiography. GST Pull-down Assay—N-terminal and C-terminal zinc-finger clusters (N-terminal, 90–383; C-terminal, 957–1156), and the middle region (384–956) were expressed as GST fusion protein in Escherichia coli together with empty vector. GST proteins were purified from glutathione-Sepharose 4B (Amersham Biosciences). Beads with 1 μg of GST proteins were mixed with 1 μg of His-Pc2 protein (AmProx) in phosphate-buffered saline/0.5% Nonidet P-40 buffer and subjected to the pull-down assay (44Long J. Wang G. Matsuura I. He D. Liu F. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 99-104Crossref PubMed Scopus (85) Google Scholar). SIP1-bound Pc2 were detected by anti-His antibody. Luciferase Reporter Gene Assay—293T, Mv1Lu/L17, and COS-1 cells in 12 well plates were transfected by Lipofectamine. Human E-cadherin-luc (-1008/+49) (49Tsai C.-N. Tsai C.-L. Tse K.-P. Chang H.-Y. Chang Y.-S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 10084-10089Crossref PubMed Scopus (249) Google Scholar) was used to measure transcriptional repression of E-cadherin by SIP1 constructs in 293T cells. TGF-β-responsive 3TP-lux and SBE4-luc were used to detect repression of TGF-β signaling by SIP1 in Mv1Lu and L17 cells, where cells were treated with or without 250 pm TGF-β for 18–24 h as previously described (44Long J. Wang G. Matsuura I. He D. Liu F. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 99-104Crossref PubMed Scopus (85) Google Scholar). L8G5-luc and LexA-VP16 (50Lemercier C. Verdel A. Galloo B. Curtet S. Brocard M.-P. Khochbin S. J. Biol. Chem. 2000; 275: 15594-15599Abstract Full Text Full Text PDF PubMed Scopus (184) Google Scholar) were used to measure the transcriptional repression activity of GAL4-SIP1 in 293T cells as described before (50Lemercier C. Verdel A. Galloo B. Curtet S. Brocard M.-P. Khochbin S. J. Biol. Chem. 2000; 275: 15594-15599Abstract Full Text Full Text PDF PubMed Scopus (184) Google Scholar). pGL3-VDR-luc (24Lazarova D.L. Bordonaro M. Sartorelli A.C. Cell Growth & Differ. 2001; 12: 319-326PubMed Google Scholar) was used to compare Myc-SIP1 and its sumoylation mutant together with Myc-δEF1, for their transactivation ability on VD3R in COS-1 cells. Luciferase activities were normalized with cotransfected β-galactosidase control driven by pSV-β-gal (Promega). Data presented are means ± S.D. from at least three independent experiments. Confocal Immunofluorescence Microscopy—COS-1, HeLa, 293T, or MDCK cells grown on coverslips in 24-well plates were transfected with plasmids to be tested by Lipofectamine. 40 h post transfection, cells were fixed in 2% paraformaldehyde and analyzed by direct or indirect immunofluorescence microscopy in a LSM510 confocal system (Zeiss). SIP1 Is Modified by Sumoylation—To gain insight into the regulation of the EMT process, we have focused on post-translational regulation of one of the key mediators of EMT, SIP1. Sequence inspection revealed that SIP1 and δEF1 sequences contain multiple ΨKX(D/E) consensus motifs, so we asked whether SIP1 and δEF1 can be modified by sumoylation. To address this, we first generated mammalian expression constructs for SIP1 and δEF1. As previously observed for δEF1 (51Funahashi J. Sekido R. Murai K. Kamachi Y. Kondoh H. Development. 1993; 119: 433-446Crossref PubMed Google Scholar), both SIP1 and δEF1 had apparent molecular masses higher than calculated (170 kDa for Myc-SIP1 and 190 kDa for Myc-δEF1, Fig. 1A, lanes 1 and 3). To test for in vivo sumoylation, COS-1 cells were transfected with Myc-SIP1 and Myc-δEF1, together with FLAG-SUMO1 (Fig. 1A). As a control for sumoylation, we included Smad4, a key mediator of TGF-β signaling, which is modified by sumoylation at Lys-159 and Lys-113 (Fig. 1A, lanes 6 and 7) (43Long J. Wang G. He D. Liu F. Biochem. J. 2004; 379: 23-29Crossref PubMed Scopus (95) Google Scholar, 52Lee P.S.W. Chang C. Liu D. Derynck R. J. Biol. Chem. 2003; 278: 27853-27863Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar, 53Lin X. Liang M. Liang Y.-Y. Brunicardi F.C. Melchior F. Feng X.-H. J. Biol. Chem. 2003; 278: 18714-18719Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar, 54Chang C.-C. Lin D.-Y. Fang H.-I. Chen R.-H. Shih H.-M. J. Biol. Chem. 2005; 280: 10164-10173Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar). Cotransfection of FLAG-SUMO1 with Myc-SIP1 or Myc-δEF1 led to slower migrating bands of 230 and 220 kDa, respectively (Fig. 1A, lanes 1–2 and 4–5). The slower migrating band of SIP1 was further "shifted" when a 2FLAG-SUMO1 construct was transfected (lane 3), confirming that the 230-kDa band of SIP1 is a SUMO conjugate. These results suggest that SIP1, as well as its homologue δEF1, is subjected to sumoylation, the size difference being consistent with the addition of two or three SUMO molecules. SIP1 Sumoylation Sites Localize to the Repression Domain—To systematically map the sumoylation sites in SIP1, we generated several deletion mutants and tested their in vivo sumoylation to delineate domains of SIP1 required for sumoylation. SIP1 is a member of the two-handed zinc-finger transcription factor family, and has N- and C-terminal zinc-finger clusters, N-ZF and C-ZF, separated by a central homeodomain (HD), a Smad-binding domain, and CtBP-interacting domains (CIDs) (see Fig. 1B for details). Deletion of the N-terminal 303 amino acids failed to affect sumoylation (Fig. 1B, dm1, lane 3), whereas a protein with an N-terminal deletion of 413 residues showed a significant decrease in sumoylation (dm2, lane 4), suggesting that the region between N-ZF and the Smad-binding domain (303–413) regulates or contains the major sumoylation sites. Interestingly, we found that the N-terminal 266 amino acids are necessary but not sufficient for efficient sumoylation of deletion mutants. This domain can promote the sumoylation of dm3, which alone showed very little sumoylation potential (lanes 6 versus 5). However the N-terminal 266 amino acids failed to rescue the modification of the C-ZF mutant (975–1214, dm7, lane 8). We speculate that the N-terminal 266 residues may serve as an adaptor to recruit components of the sumoylation machinery (E2-conjugating enzyme Ubc9 and/or E3 ligase) to dm3, which harbors sumoylation site(s). The different sumoylation potential of dm6 and dm7 argues that the region between 810 and 974 is responsible for the sumoylation of dm6. Thus, we have mapped the minimal sumoylation domain of SIP1 to 304–975, which is almost identical to the previously identified "repression domain" (335–998) of SIP1 (18Postigo A.A. Dean D.C. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 6391-6396Crossref PubMed Scopus (119) Google Scholar). This result is consistent with the observation that sumoylation of most transcription factors lies within their repression domains (31Hay R.T. Mol. Cell. 2005; 18: 1-12Abstract Full Text Full Text PDF PubMed Scopus (1360) Google Scholar, 40Müller S. Ledl A. Schmidt D. Oncogene. 2004; 23: 1998-2008Crossref PubMed Scopus (247) Google Scholar, 55Verger A. Perdomo J. Crossley M. EMBO Rep. 2003; 4: 137-142Crossref PubMed Scopus (381) Google Scholar). Identification of SIP1 Sumoylation Sites—The minimal sumoylation domain (304–975) contains three consensus sumoylation motifs: Ile-Lys391-Thr-Glu, Ile-Lys555-Lys-Glu, and Ile-Lys866-Lys-Glu (Fig. 2A, top panel). Single mutants of each of these sites where lysine was substituted with an arginine residue were generated from Myc-SIP1, and their in vivo sumoylation was tested in COS-1 cells. Whereas the sumoylation of Myc-SIP1 (K391R) is significantly decreased when compared with wild-type, sumoylation of K555R or K866R is not affected (Fig. 2A, bottom panel, lanes 8-10). These results indicate that Lys391 is a major sumoylation site for SIP1, and
Referência(s)