Characterization of Fas (Apo-1, CD95)-Fas Ligand Interaction
1997; Elsevier BV; Volume: 272; Issue: 30 Linguagem: Inglês
10.1074/jbc.272.30.18827
ISSN1083-351X
AutoresPascal Schneider, Jean-Luc Bodmer, Nils Holler, Chantal Mattmann, Patricia Scuderi, Alexey V. Terskikh, Manuel C. Peitsch, Jürg Tschopp,
Tópico(s)Global Peace and Security Dynamics
ResumoThe death-inducing receptor Fas is activated when cross-linked by the type II membrane protein Fas ligand (FasL). When human soluble FasL (sFasL, containing the extracellular portion) was expressed in human embryo kidney 293 cells, the threeN-linked glycans of each FasL monomer were found to be essential for efficient secretion. Based on the structure of the closely related lymphotoxin α-tumor necrosis factor receptor I complex, a molecular model of the FasL homotrimer bound to three Fas molecules was generated using knowledge-based protein modeling methods. Point mutations of amino acid residues predicted to affect the receptor-ligand interaction were introduced at three sites. The F275L mutant, mimicking the loss of function murine gld mutation, exhibited a high propensity for aggregation and was unable to bind to Fas. Mutants P206R, P206D, and P206F displayed reduced cytotoxicity toward Fas-positive cells with a concomitant decrease in the binding affinity for the recombinant Fas-immunoglobulin Fc fusion proteins. Although the cytotoxic activity of mutant Y218D was unaltered, mutant Y218R was inactive, correlating with the prediction that Tyr-218 of FasL interacts with a cluster of three basic amino acid side chains of Fas. Interestingly, mutant Y218F could induce apoptosis in murine, but not human cells. The death-inducing receptor Fas is activated when cross-linked by the type II membrane protein Fas ligand (FasL). When human soluble FasL (sFasL, containing the extracellular portion) was expressed in human embryo kidney 293 cells, the threeN-linked glycans of each FasL monomer were found to be essential for efficient secretion. Based on the structure of the closely related lymphotoxin α-tumor necrosis factor receptor I complex, a molecular model of the FasL homotrimer bound to three Fas molecules was generated using knowledge-based protein modeling methods. Point mutations of amino acid residues predicted to affect the receptor-ligand interaction were introduced at three sites. The F275L mutant, mimicking the loss of function murine gld mutation, exhibited a high propensity for aggregation and was unable to bind to Fas. Mutants P206R, P206D, and P206F displayed reduced cytotoxicity toward Fas-positive cells with a concomitant decrease in the binding affinity for the recombinant Fas-immunoglobulin Fc fusion proteins. Although the cytotoxic activity of mutant Y218D was unaltered, mutant Y218R was inactive, correlating with the prediction that Tyr-218 of FasL interacts with a cluster of three basic amino acid side chains of Fas. Interestingly, mutant Y218F could induce apoptosis in murine, but not human cells. The Fas ligand (CD95 ligand) is a 40-kDa type II membrane protein belonging to the tumor necrosis factor (TNF) 1The abbreviations used are: TNF, tumor necrosis factor; sFasL, recombinant human soluble Fas ligand; DMEM, Dulbecco's modified Eagle's medium; FCS, fetal calf serum; PCR, polymerase chain reaction; hFas-Fc, human Fas-Fc; muFas-Fc, murine Fas-Fc; ELISA, enzyme-linked immunosorbent assay; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; MTS, 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt. family of proteins (1Nagata S. Cell. 1997; 88: 355-365Abstract Full Text Full Text PDF PubMed Scopus (4561) Google Scholar, 2Smith C.A. Farrah T. Goodwin R.G. Cell. 1994; 76: 959-962Abstract Full Text PDF PubMed Scopus (1839) Google Scholar). This family consists of trimeric ligands that induce defined cellular responses upon binding to their respective receptors. Fas and the other members of the TNF receptor family are type I membrane proteins. They are characterized by the presence of cysteine-rich motives conferring an elongated structure to their extracellular domains (1Nagata S. Cell. 1997; 88: 355-365Abstract Full Text Full Text PDF PubMed Scopus (4561) Google Scholar). The Fas ligand is one of the major effectors of CD8+cytotoxic T lymphocytes (3Kägi D. Vignaux F. Lerdermann B. Bürki K. Deraetere V. Nagata S. Hengartner H. Goldstein P. Science. 1994; 265: 528-530Crossref PubMed Scopus (1476) Google Scholar, 4Lowin B. Hahne M. Mattmann C. Tschopp J. Nature. 1994; 370: 650-652Crossref PubMed Scopus (994) Google Scholar) and natural killer cells (5Tsutsui H. Nakanishi K. Matsui K. Higashino K. Okamura H. Miyazawa Y. Kaneda K. J. Immunol. 1996; 157: 3967-3973PubMed Google Scholar). It is also involved in the establishment of peripheral tolerance (6Rathmell J.C. Cooke M.P. Ho W.Y. Grein J. Townsend S.E. Davis M.M. Goodnow C.C. Nature. 1995; 376: 181-184Crossref PubMed Scopus (418) Google Scholar), in the activation-induced cell death of lymphocytes (7Alderson M.R. Tough T.W. Davis-Smith T. Braddy S. Falk B. Schooley K.A. Goodwin R.G. Smith C.A. Ramsdell F. Lynch D.H. J. Exp. Med. 1995; 181: 71-77Crossref PubMed Scopus (871) Google Scholar, 8Brunner T. Mogil R.J. LaFace D. Yoo N.J. Mahboubi A. Echeverri F. Martin S.J. Force W.R. Lynch D.H. Ware C.F. Green D.R. Nature. 1995; 373: 441-444Crossref PubMed Scopus (1272) Google Scholar, 9Dhein J. Walczak H. Bäumler C. Debatin K.-M. Krammer P.H. Nature. 1995; 373: 438-441Crossref PubMed Scopus (1609) Google Scholar, 10Ju S.-T. Panka D.J. Cui H. Ettinger R. El-Khatib M. Sherr D.H. Stanger B.Z. Marshak-Rothstein A. Nature. 1995; 373: 444-448Crossref PubMed Scopus (1456) Google Scholar), and in the delimitation of immunoprivileged regions such as the eye and testis (11Griffith T. Brunner T. Fletcher S. Green D. Ferguson T. Science. 1995; 270: 1189-1192Crossref PubMed Scopus (1879) Google Scholar, 12Bellgrau D. Gold D. Selawry H. Moore J. Franzusoff A. Duke R.C. Nature. 1995; 377: 630-632Crossref PubMed Scopus (1103) Google Scholar). Along the lines of this latter feature, cotransplantation of myoblasts engineered to express FasL can protect an islet allograft from rejection (13Lau H.T. Yu M. Fontana A. Stoeckert C.J. Science. 1996; 273: 109-112Crossref PubMed Scopus (427) Google Scholar). The loss of function due to mutations in murine Fas ligand (gld), murine Fas (lpr), human Fas, or human FasL leads to lymphoproliferation, lymphadenopathy, and autoimmune diseases (14Watanabe-Fukunaga R. Brannan C.I. Copeland N.G. Jenkins N.A. Nagata S. Nature. 1992; 356: 314-317Crossref PubMed Scopus (2737) Google Scholar, 15Takahashi T. Tanaka M. Brannan C.I. Jenkins N.A. Copeland N.G. Suda T. Nagata S. Cell. 1994; 76: 969-976Abstract Full Text PDF PubMed Scopus (1474) Google Scholar, 16Fisher G.H. Rosenberg F.J. Straus S.E. Dale J.K. Middelton L.A. Lin A.Y. Strober W. Lenardo M.J. Puck J.M. Cell. 1995; 81: 935-946Abstract Full Text PDF PubMed Scopus (1308) Google Scholar, 17Rieux-Laucat F. Le Deist F. Hivroz C. Roberts I.A.G. Debatin K.M. Fisher A. de Villartay J.P. Science. 1995; 268: 1347-1349Crossref PubMed Scopus (1185) Google Scholar, 18Wu J. Wilson J. He J. Xiang L.B. Schur P.H. Mountz J.D. J. Clinical. Invest. 1996; 98: 1107-1113Crossref PubMed Scopus (408) Google Scholar). Fas-null mice have a similar but more severe phenotype (19Adachi M. Suematsu S. Suda T. Watanabe D. Fukuyama H. Ogasawara J. Tanaka T. Yoshida N. Nagata S. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 2131-2136Crossref PubMed Scopus (179) Google Scholar). The Fas-FasL system is implicated in a number of pathogenesis. Abnormally elevated levels of soluble Fas ligand are detected in leukemia/lymphomas of T and natural killer cells, and in an aggressive nasal lymphoma (20Tanaka M. Suda T. Haze K. Nakamura N. Sato K. Kimura F. Motoyoshi K. Mizuki M. Tagawa S. Ohga S. Hatake K. Drummond A.H. Nagata S. Nat. Med. 1996; 2: 317-322Crossref PubMed Scopus (652) Google Scholar, 21Sato K. Kimura F. Nakamura Y. Murakami H. Yoshida M. Tanaka M. Nagata S. Kanatani Y. Wakimoto N. Nagata N. Motoyoshi K. Br. J. Haematol. 1996; 94: 379-382Crossref PubMed Scopus (69) Google Scholar). Various tumor cells express FasL, therefore potentially creating their own immunoprivileged sites (22Oconnell J. Osullivan G.C. Collins J.K. Shanahan F. J. Exp. Med. 1996; 184: 1075-1082Crossref PubMed Scopus (871) Google Scholar, 23Strand S. Hofmann W.J. Hug H. Muller M. Otto G. Strand D. Mariani S.M. Stremmel W. Krammer P.H. Galle P.R. Nat. Med. 1996; 2: 1361-1366Crossref PubMed Scopus (881) Google Scholar, 24Hahne M. Rimoldi D. Schröter M. Romero P. Schreier M. French L.E. Schneider P. Bornand T. Fontana A. Lienard D. Cerottini J.-C. Tschopp J. Science. 1996; 274: 1363-1366Crossref PubMed Scopus (1202) Google Scholar). The Fas/Fas ligand system is also involved in the CD4+ T cell deletion observed in human immunodeficiency virus-infected individuals (25Westendorp M.O. Frank R. Ochsenbauer C. Stricker K. Dhein J. Walczak H. Debatin K.-M Krammer P.H. Nature. 1995; 375: 497-500Crossref PubMed Scopus (916) Google Scholar, 26Katsikis P.D. Wunderlich E.S. Smith C.A. Herzenberg L.A. Herzenberg L.A. J. Exp. Med. 1995; 181: 2029-2036Crossref PubMed Scopus (420) Google Scholar, 27Estaquier J. Tanaka M. Suda T. Nagata S. Golstein P. Ameisen J.C. Blood. 1996; 87: 4959-4966Crossref PubMed Google Scholar), in multiple sclerosis (28Dowling P. Shang G.F. Raval S. Menonna J. Cook S. Husar W. J. Exp. Med. 1996; 184: 1513-1518Crossref PubMed Scopus (269) Google Scholar), and in acute graft-versus-host disease (29Via C.S. Nguyen P. Shustov A. Drappa J. Elkon K.B. J. Immunol. 1996; 157: 5387-5393PubMed Google Scholar, 30Braun M.Y. Lowin B. French L. Acha-Orbea H. Tschopp J. J. Exp. Med. 1996; 183: 657-661Crossref PubMed Scopus (281) Google Scholar). Membrane-bound Fas ligand can be processed into a soluble form by a metalloprotease whose inhibitor profile is similar to that of the protease solubilizing TNFα (20Tanaka M. Suda T. Haze K. Nakamura N. Sato K. Kimura F. Motoyoshi K. Mizuki M. Tagawa S. Ohga S. Hatake K. Drummond A.H. Nagata S. Nat. Med. 1996; 2: 317-322Crossref PubMed Scopus (652) Google Scholar, 31Gearing A.J.H. Beckett P. Christodoulou M. Churchill M. Clements J. Davidson A.H. Drummond A.H. Galloway W.A. Gilbert R. Gordon J.L. Leber T.M. Mangan M. Miller K. Nayee P. Owen K. Patel S. Thomas W. Wells G. Wood L.M. Woolley K. Nature. 1994; 370: 555-557Crossref PubMed Scopus (1111) Google Scholar, 32Kayagaki N. Kawasaki A. Ebata T. Ohmoto H. Ikeda S. Inoue S. Yoshino K. Okumura K. Yagita H. J. Exp. Med. 1995; 182: 1777-1783Crossref PubMed Scopus (778) Google Scholar, 33Mariani S.M. Matiba B. Bäumler C. Krammer P.H. Eur. J. Immunol. 1995; 25: 2303-2307Crossref PubMed Scopus (216) Google Scholar, 34Martinez-Lorenzo M.J. Alava M.A. Anel A. Piñeiro A. Naval J. Immunology. 1996; 89: 511-517Crossref PubMed Scopus (90) Google Scholar). The soluble forms of the TNF family members that have been crystallized so far include TNFα (35Jones E.Y. Stuart D.I. Walker N.P.C. Nature. 1989; 338: 225-228Crossref PubMed Scopus (482) Google Scholar), lymphotoxin α (also known as TNFβ) (36Eck M.J. Ultsch M. Rinderknecht E. de Vos A.M. Sprang S.R. J. Biol. Chem. 1992; 267: 2119-2122Abstract Full Text PDF PubMed Google Scholar), and CD40L (37Karpusas M. Hsu Y.-M. Wang J.-H. Thompson J. Lederman S. Chess L. Thomas D. Structure. 1995; 3: 1031-1039Abstract Full Text Full Text PDF PubMed Scopus (175) Google Scholar). They all share a similar conformation resulting in a compact, pear-shaped trimeric structure. Other family members (Fas ligand (38Takahashi T. Tanaka M. Inazawa J. Abe T. Suda T. Nagata S. Int. Immunol. 1994; 6: 1567-1574Crossref PubMed Scopus (420) Google Scholar), TRAIL (39Wiley S.R. Schooley K. Smolak P.J. Din W.S. Huang C.-P. Nicholl J.K. Sutherland G.R. Smith T.D. Rauch C. Smith C.A. Goodwin R.G. Immunity. 1995; 3: 673-682Abstract Full Text PDF PubMed Scopus (2664) Google Scholar), CD30L, CD27L, OX40L, and 4–1BBL) show clear sequence homology at the amino acid level, and there is little doubt that they are all trimeric. In this study, we produce biologically active human soluble Fas ligand (sFasL) and amino acid residues essential for the Fas/FasL interaction are identified. We also demonstrate the importance of extensiveN-glycosylation for the efficient secretion of FasL. The anti-flag M2 monoclonal antibody and the anti-flag M2 antibody coupled to agarose were purchased from Integra Biosciences (Wallisellen, Switzerland). Protein A-Sepharose was purchased from Pharmacia (Uppsala, Sweden). Tunicamycin and Protein A were obtained from Sigma (Buchs, Switzerland). Peptide N-glycanase F was purchased from New England Biolabs (Schwalbach, Germany). The PCR-2 TA cloning vector and PCR-3 mammalian expression vector were obtained from InvitroGen (NV Leek, the Netherlands). Cell culture media and antibiotics were obtained from Life Sciences (Basel, Switzerland). The non-radioactive cell proliferation assay was purchased from Promega (Wallisellen, Switzerland). The fusion protein muFas-Fc was kindly provided by Dr. C. A. Smith (Immunex, Seattle, WA). Murine B lymphoma A20 cells were grown into DMEM containing 5% heat-inactivated fetal calf serum (FCS) and the human T lymphoblastoma Jurkat cell line was grown in RPMI supplemented with 10% FCS. Human embryonic kidney 293 cells (ATCC CRL 1573) were cultured in DMEM:nutrient mix F-12 (1:1) supplemented with 2% FCS. Human embryonic kidney 293 cells stably transfected with the large T antigen of SV40 (293T cells, kindly provided by Dr. M. E. Peters, German Cancer Research Center, Heidelberg, Germany) were grown in DMEM supplemented with 10% FCS. All media contained antibiotics (penicillin and streptomycin at 5 μg/ml each and neomycin at 10 μg/ml). A DNA fragment coding for the signal peptide of hemaglutinin, including 6 bases of its 5′-untranslated sequence (40Gething M.-J. Bye J. Skehel J. Waterfield M. Nature. 1980; 287: 301-306Crossref PubMed Scopus (186) Google Scholar), the flag epitope (41Hopp T.P Prickett K.S. Price V.L. Liggy R.T. March C.J. Cerreti D.P. Urdal D.L. Conlon P.J. Biotechnology. 1988; 6: 1204-1210Crossref Scopus (754) Google Scholar), a linker (GPGQVQLQ), and the PstI,SalI, XhoI, and BamHI restriction sites, was cloned between the HindIII and BamHI sites of a modified PCR-3 vector in which nucleotides 720–769 had been deleted. This plasmid was called pHAflag-038. The full-length cDNA of human Fas ligand was amplified by PCR from the cDNA of activated peripheral blood lymphocytes (oligonucleotides: 5′-CCTCTACAGGACTGAGAAGAAG-3′ and 5′-CAACATTCTCGGTGCCTGTAAC-3′), and cloned into PCR-2 TA cloning vector. This plasmid was used as PCR template for the amplification of a portion of the extracellular domain of the FasL (amino acids 139–281) with suitable restriction sites added at each end. The resulting PstI/EcoRI fragment was inserted into pHAflag-038, in frame with the flag sequence. For each point mutation, a set of complementary oligonucleotides containing the target mutation was used. In the first round of the PCR, two products were produced with pHAflag-FasL as template using: (a) the forward oligonucleotide and Sp6 primer, and (b) the reverse oligonucleotide and T7 primer. Purified PCR products, containing the 3′ and 5′ portions of FasL, respectively, were mixed and allowed to undergo three cycles of PCR before amplification with T7 and Sp6 primers. The PstI/EcoRI fragment of the resulting PCR product was cloned into pHAflag-038. The extracellular domain of hFas (GenBank X63717, nucleotides −24 to 510, the A of ATG being nucleotide 1) with 5′HindIII and 3′ SmaI sites was amplified by PCR from a full-length cDNA clone (kindly provided by Prof. P. H. Krammer, German Cancer Research Center, Heidelberg). TheHindIII-SmaI fragment was cloned between theHindIII and EcoRV sites of a modified PCR-III vector containing an added SalI site after the existingEcoRV site. A SalI/NotI cDNA cassette encoding the hinge, CH2, and CH3 domains (amino acid residues 231–447) of human IgG1 (42Peppel K. Crawford D. Beutler B. J. Exp. Med. 1991; 174: 1483-1489Crossref PubMed Scopus (298) Google Scholar) was cloned in frame at the 3′ end of the extracellular domain of Fas. Both strands of each construct were checked by sequencing. Plasmids were either expressed transiently in 293T cells or stably in 293 cells. Plasmids (10 μg) were transfected by the calcium phosphate method (3 × 105 cells/28-cm2plate) in HEPES buffer (43Kingston R.E. Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Current Protocols in Molecular Biology. John Wiley & Sons, New York1995: 9.1.1-9.1.11Google Scholar). After transfection, cells were grown for 48–72 h in serum-free Opti-MEM medium, and supernatants were harvested. Stably transfected 293 cells were obtained by selection in 800 μg/ml G418 (70% active) for 2 weeks and cloned at that stage. Supernatants of stably transfected clones were harvested after 10–12 days in culture and screened by Western blotting or receptor binding ELISA (see below) for expression levels. 293 cells (2 × 105) transiently transfected with sFasL and their corresponding supernatants (20 × concentrated, 15 μl) were heated in 20 μl of 0.5% SDS, 1% 2-mercaptoethanol for 3 min at 95 °C. Samples were cooled and supplemented with 10% Nonidet P-40 (2 μl) and 0.5 msodium phosphate, pH 7.5 (2 μl). Peptide N-glycanase F (125 units/μl, 1 μl) was added (or omitted in controls), and samples were incubated for 3 h at 37 °C prior to analysis by Western blotting. Supernatants of stably transfected cells were filtered using a 0.22-μm membrane and loaded as 40-ml aliquots onto 1-ml columns of anti-flag M2 agarose (for sFasL) or Protein A-Sepharose (for hFas-Fc) equilibrated in PBS. The columns were washed with 10 ml of PBS and eluted with 2.5 ml of 50 mm citric acid. The eluate was neutralized with 1 m Tris base, concentrated, and exchanged into PBS using Centriprep-30 concentrators. Protein concentration was determined by the bicinchoninic acid method (Pierce) using bovine serum albumin as the standard, and the purity of the samples was assessed by SDS-PAGE and Coomassie Blue staining. SDS-PAGE and Western blotting were performed on 12% mini gels according to previously published methods (44Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (207538) Google Scholar, 45Towbin H. Staehelin T. Gordon J. Proc. Natl. Acad. Sci. U. S. A. 1979; 76: 4350-4354Crossref PubMed Scopus (44939) Google Scholar). Blots were incubated with anti-flag M2 monoclonal antibodies (5 μg/ml, 0.02% NaN3, in blocking buffer: PBS, 0.5% Tween 20, 4% skim milk), followed by rabbit anti-mouse immunoglobulins coupled to horseradish peroxidase (diluted 1:2000 in blocking buffer). Peroxidase activity was detected by enhanced chemiluminescence. Tunicamycin was stored at −70 °C at a concentration of 1 mg/ml in 10 mm Tris-HCl, pH 9. Stably transfected cells secreting sFasL were grown for 10 days in the presence of 1, 100, or 1000 ng/ml tunicamycin. Cells and supernatants were harvested and analyzed by Western blotting. A20 or Jurkat cells (100 μl, 50,000 cells, in 96-well plates) were incubated at 37 °C in the presence of sFasL at the indicated concentrations and 1 μg/ml M2 monoclonal antibody. In some experiments, hFas-Fc or muFas-Fc was added at the indicated concentrations in the presence of 1 μg/ml Protein A. Four to 8 h after the addition of FasL, 20 μl of a solution containing 2 mg/ml 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) reagent (Promega) and 50 μg/ml phenazine methosulfate was added to the cells. Following color development (2–4 h), absorbance at 490 nm was taken with an ELISA reader. 96-well ELISA plates (Nunc Maxisorp) were coated with either hFas-Fc or muFas-Fc (1 μg/ml in PBS, 100 μl, 2 h, 37 °C). The following incubation and washing steps were performed: (a) saturation in block buffer (PBS, 5% FCS, 300 μl, 1 h, 37 °C), (b) three washes (PBS, 0.05% Tween-20), (c) incubation with sFasL (10–1000 ng/ml in PBS containing 50 μg/ml bovine serum albumin, 100 μl, 1 h, 37 °C), (d) three washes, (e) incubation with M2 monoclonal antibody (1 μg/ml in block buffer, 100 μl, 37 °C, 30 min), (f) three washes, (g) incubation with rabbit anti-mouse IgG coupled to peroxidase (1/1000 dilution in block buffer, 100 μl, 30 min, 37 °C), (h) three washes, (i) detection (0.3 mg/ml o-phenylenediamine hydrochloride, 0.01% H2O2 in 50 mmcitric acid, 100 mm Na2HPO4, 200 μl, as necessary (1–5 min), 25 °C), and (j) termination (2 n HCl, 50 μl). Absorbance was taken at 490 nm with an ELISA reader. sFasL samples (5 μg in 100 μl) were mixed with the internal standards catalase and ovalbumin, then loaded onto a Superdex-200 HR10/30 column, and the proteins were eluted in PBS at 0.5 ml/min. Fractions (0.25 ml) were analyzed using the receptor binding ELISA (using 5 μl of fractions for active ligands), and Western blotting was carried out after trichloroacetic acid precipitation of the entire fraction. The column was calibrated with standard proteins: thyroglobulin (669 kDa), ferritin (440 kDa), catalase (232 kDa), aldolase (158 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa), chymotrypsinogen A (25 kDa), and ribonuclease A (13.7 kDa). Molecular models for both Fas and FasL were generated using knowledge-based protein modeling methods as implemented in the Swiss-Model server (46Peitsch M.C. Biotechnology. 1995; 13: 658-660Crossref Scopus (116) Google Scholar, 47Peitsch M.C. Biochem. Soc. Trans. 1996; 24: 274-279Crossref PubMed Scopus (900) Google Scholar). A molecular model for FasL was built using the known tridimensional structures of both TNFα (48Eck M.J. Sprang S.R. J. Biol. Chem. 1989; 264: 17595-17605Abstract Full Text PDF PubMed Google Scholar) and lymphotoxin α (TNFβ) (49Banner D.W. D'Arcy A. Janes W. Gentz R. Schoenfeld H.-J. Broger C. Loetscher H. Lesslauer W. Cell. 1993; 73: 431-445Abstract Full Text PDF PubMed Scopus (990) Google Scholar) (Protein Data Bank entries 1TNF and 1TNR chains A, B, and C) (50Peitsch M.C. Tschopp J. Mol. Immunol. 1995; 32: 761-772Crossref PubMed Scopus (65) Google Scholar). The known structure of the 55-kDa tumor necrosis factor receptor (51Peitsch M.C. Jongeneel C.V. Int. Immunol. 1993; 5: 233-238Crossref PubMed Scopus (153) Google Scholar) was used to produce a molecular model for Fas. In both cases, the modeling procedure was started by submitting the respective protein sequences to Swiss-Model via the First Approach Mode. The resulting models were then used to correct the automated multiple sequence alignments generated by the server in several loop regions. These corrected alignments were resubmitted to Swiss-Model through the Optimize Mode. The resulting models were structurally sound, and did not show obvious sequence or structure inconsistencies (51Peitsch M.C. Jongeneel C.V. Int. Immunol. 1993; 5: 233-238Crossref PubMed Scopus (153) Google Scholar) according to three-dimensional/one-dimensional profiles (52Luethy R. Bowie J.U. Eisenberg D. Nature. 1992; 356: 83-85Crossref PubMed Scopus (2616) Google Scholar) and ProsaII (53Sippl M.J. Proteins. 1993; 17: 355-362Crossref PubMed Scopus (1792) Google Scholar). The quaternary structure of the Fas-FasL complex was generated using the x-ray structure of the human lymphotoxin α-TNF receptor I complex (54Brooks B.R. Bruccoleri R.E. Olafson B.D. States D.J. Swaminathan S. Karplus M. J. Comput. Chem. 1983; 4: 187-217Crossref Scopus (14019) Google Scholar). Three copies of the ligand model and three copies of the receptor model were superimposed onto the corresponding subunits of the experimental structure. The hexameric protein complex was further refined by 200 cycles of energy minimization with CHARMM (53Sippl M.J. Proteins. 1993; 17: 355-362Crossref PubMed Scopus (1792) Google Scholar). The three N-glycosylation sites are located at Asn-184, Asn-250, and Asn-260 in the FasL model. A short branched N-linked glycan structure (GlcnAcβ1–4Manα1–6[Manα1–3]Manβ1–4GlcNAcβ1 –4GlcNAc) was extracted from Brookhaven Protein Data Bank entry 9API, and was linked to the respective asparagines using interactive graphics. A plasmid encoding the signal peptide of hemaglutinin, in frame with a flag epitope and the COOH-terminal portion of the extracellular domain of human Fas ligand (amino acids 139–281), was transfected into the human embryonic kidney 293 cell line. Secreted sFasL was affinity-purified using immobilized anti-flag antibodies. The theoretical molecular mass of the encoded recombinant protein is 18.2 kDa. Purified sFasL migrated as a doublet on SDS-PAGE with deduced molecular masses of 29 and 25.5 kDa (Fig.1 A). Taken together with previous data (55Tanaka M. Suda T. Takahashi T. Nagata S. EMBO J. 1995; 14: 1129-1135Crossref PubMed Scopus (608) Google Scholar), this heterogeneity and the discrepancy between predicted and observed molecular masses suggest that carbohydrates are present on the sFasL. Indeed, sFasL present in both cell extracts and cell supernatants could be digested with peptide N-glycanase F to a single band with the predicted molecular mass of 18 kDa (Fig. 1 B), indicating that the various species of sFasL differed by their degree ofN-glycosylation. This result was confirmed when cells were treated with the N-glycosylation inhibitor tunicamycin; a dose-dependent accumulation of cellular, unglycosylated 18-kDa sFasL was observed with concomitant loss of sFasL secretion (Fig. 1 C). A total of four evenly spaced bands of sFasL could be detected, which probably correspond to the unglycosylated, mono-, di-, and tri-N-glycosylated sFasL monomers. Thus, all three potential N-glycosylation sites of human FasL (Asn-184, Asn-250, and Asn-260) appear to be used. Interestingly, secreted sFasL is consistently found in its highly glycosylated form, even at intermediate tunicamycin concentrations where unglycosylated sFasL is by far the predominant cellular species (Fig. 1 C). This strongly suggests that N-linked oligosaccharides are required for efficient secretion of sFasL.Figure 1Glycosylation of sFasL. Panel A, Coomassie Blue staining of sFasL. The flag-tagged sFasL in conditioned supernatants of stably transfected 293 cells was affinity-purified on anti-flag M2 agarose, and 1.5 μg was analyzed by SDS-PAGE (12%) followed by Coomassie Blue staining. The molecular mass markers (in kDa) are indicated. Panel B, peptide N-glycanase F digestion of sFasL. 293 cells transiently transfected with flag-tagged sFasL were grown for 72 h in serum-free medium. Cell extracts (Cells, 2.5 × 105 cells/lane) and 20 × concentrated supernatant (S/N, 15 μl/lane) were treated with or without peptide N-glycanase F (PNGase F) and analyzed by Western blotting using anti-flag M2 antibodies.Panel C, the effect of tunicamycin on the glycosylation and secretion of sFasL. Stably transfected 293 cells were grown for 10 days in serum-free medium and in the presence of the indicated concentrations of tunicamycin. Cells (2.5 × 105/lane) and supernatants (20 μl) were analyzed by Western blotting using anti-flag M2 antibodies.View Large Image Figure ViewerDownload Hi-res image Download (PPT) We next generated a molecular model of the FasL-Fas complex (Fig.2) using knowledge-based protein modeling methods and the known tridimensional structures of lymphotoxin α (TNFβ) and the 55-kDa tumor necrosis factor receptor (49Banner D.W. D'Arcy A. Janes W. Gentz R. Schoenfeld H.-J. Broger C. Loetscher H. Lesslauer W. Cell. 1993; 73: 431-445Abstract Full Text PDF PubMed Scopus (990) Google Scholar). It can be seen that theN-linked oligosaccharides are concentrated on the lateral edge of each FasL monomer, leaving vertical clefts, which would allow the receptor to reach the interaction site of the ligand. Two of the three glycosylation sites of the FasL have counterparts in other family members. Residue Asn-260 of FasL is also found at the corresponding position in lymphotoxin β, CD27L, and CD40L, whereas residue Asn-250 of FasL corresponds to a site in CD30L. In some TNF family members, N-glycosylation appears to be important for the biological activity of the protein. For example, FasL, which can be expressed in a variety of eukaryotic systems (20Tanaka M. Suda T. Haze K. Nakamura N. Sato K. Kimura F. Motoyoshi K. Mizuki M. Tagawa S. Ohga S. Hatake K. Drummond A.H. Nagata S. Nat. Med. 1996; 2: 317-322Crossref PubMed Scopus (652) Google Scholar, 56Mariani S.M. Matiba B. Sparna T. Krammer P.H. J. Immunol. Methods. 1996; 193: 63-70Crossref PubMed Scopus (20) Google Scholar, 57Jumper M.D. Nishioka Y. Davis L.S. Lipsky P.E. Meek K. J. Immunol. 1995; 155: 2369-2378PubMed Google Scholar), forms inactive inclusion bodies when expressed in bacteria. 2P. Schneider, unpublished observation. The extensively glycosylated CD30L (four putative N-glycosylation sites) is also best produced in a recombinant form in eukaryotic systems (57Jumper M.D. Nishioka Y. Davis L.S. Lipsky P.E. Meek K. J. Immunol. 1995; 155: 2369-2378PubMed Google Scholar). Glycosylated recombinant CD40L is readily expressed at the surface of eukaryotic cells, but transport to the cell surface is blocked in the presence of the N-glycosylation inhibitor tunicamycin (57Jumper M.D. Nishioka Y. Davis L.S. Lipsky P.E. Meek K. J. Immunol. 1995; 155: 2369-2378PubMed Google Scholar). In contrast, TNFα (58Van Ostade X. Vandenabeele P. Tavernier J. Fiers W. Eur. J. Biochem. 1994; 220: 771-779Crossref PubMed Scopus (49) Google Scholar) and lymphotoxin α (59Goh C.R. Loh C.-S. Porter A.G. Protein Eng. 1991; 4: 785-792Crossref PubMed Scopus (27) Google Scholar) can be produced in a soluble, non-glycosylated and active form in prokaryotic expression systems. When these two latter ligands are expressed in eukaryotic systems,N-glycosylation can result in minor effects such as a 10-fold decrease in specific activity of TNFα (60Koyama Y. Hayashi T. Fujii N. Yoshida T. Bio
Referência(s)