Human Combinatorial Fab Library Yielding Specific and Functional Antibodies against the Human Fibroblast Growth Factor Receptor 3
2003; Elsevier BV; Volume: 278; Issue: 40 Linguagem: Inglês
10.1074/jbc.m303164200
ISSN1083-351X
AutoresRobert Rauchenberger, Eric Borges, Elisabeth Thomassen-Wolf, Eran Rom, Rivka Adar, Yael Yaniv, Michael Malka, Irina Chumakov, Sarit Kotzer, Dalia Resnitzky, Achim Knappik, Silke U. Reiffert, Josef Prassler, Karin M. Jury, Dirk Waldherr, Susanne Bauer, Titus Kretzschmar, Avner Yayon, Christine Rothe,
Tópico(s)HER2/EGFR in Cancer Research
ResumoThe human combinatorial antibody library Fab 1 (HuCAL®-Fab 1) was generated by transferring the heavy and light chain variable regions from the previously constructed single-chain Fv library (Knappik, A., Ge, L., Honegger, A., Pack, P., Fischer, M., Wellnhofer, G., Hoess, A., Wölle, J., Plückthun, A., and Virnekäs, B. (2000) J. Mol. Biol. 296, 57–86), diversified in both complementarity-determining regions 3 into a novel Fab display vector, yielding 2.1 × 1010 different antibody fragments. The modularity has been retained in the Fab display and screening plasmids, ensuring rapid conversion into various antibody formats as well as antibody optimization using prebuilt maturation cassettes. HuCAL®-Fab 1 was challenged against the human fibroblast growth factor receptor 3, a potential therapeutic antibody target, against which, to the best of our knowledge, no functional antibodies could be generated so far. A unique screening mode was designed utilizing recombinant functional proteins and cell lines differentially expressing fibroblast growth factor receptor isoforms diversified in expression and receptor dependence. Specific Fab fragments with subnanomolar affinities were isolated by selection without any maturation steps as determined by fluorescence flow cytometry. Some of the selected Fab fragments completely inhibit target-mediated cell proliferation, rendering them the first monoclonal antibodies against fibroblast growth factor receptors having significant function blocking activity. This study validates HuCAL®-Fab 1 as a valuable source for the generation of target-specific antibodies for therapeutic applications. The human combinatorial antibody library Fab 1 (HuCAL®-Fab 1) was generated by transferring the heavy and light chain variable regions from the previously constructed single-chain Fv library (Knappik, A., Ge, L., Honegger, A., Pack, P., Fischer, M., Wellnhofer, G., Hoess, A., Wölle, J., Plückthun, A., and Virnekäs, B. (2000) J. Mol. Biol. 296, 57–86), diversified in both complementarity-determining regions 3 into a novel Fab display vector, yielding 2.1 × 1010 different antibody fragments. The modularity has been retained in the Fab display and screening plasmids, ensuring rapid conversion into various antibody formats as well as antibody optimization using prebuilt maturation cassettes. HuCAL®-Fab 1 was challenged against the human fibroblast growth factor receptor 3, a potential therapeutic antibody target, against which, to the best of our knowledge, no functional antibodies could be generated so far. A unique screening mode was designed utilizing recombinant functional proteins and cell lines differentially expressing fibroblast growth factor receptor isoforms diversified in expression and receptor dependence. Specific Fab fragments with subnanomolar affinities were isolated by selection without any maturation steps as determined by fluorescence flow cytometry. Some of the selected Fab fragments completely inhibit target-mediated cell proliferation, rendering them the first monoclonal antibodies against fibroblast growth factor receptors having significant function blocking activity. This study validates HuCAL®-Fab 1 as a valuable source for the generation of target-specific antibodies for therapeutic applications. The development of the hybridoma technology opened the application of monoclonal antibodies for research and human therapy (2Köhler G. Milstein C. Nature. 1975; 256: 495-497Crossref PubMed Scopus (12802) Google Scholar). A major drawback of these first generation monoclonal antibodies, especially for clinical application, was their murine origin, which often caused immune response in human and led to the generation of human anti-mouse antibodies (human anti-mouse antibody reaction), limiting efficacy in long term and repeated administration (3Jaffers G. Fuller T.C. Cosimi A.B. Russell P.S. Winn H.J. Colvin R.B. Transplantation. 1986; 41: 572-578Crossref PubMed Scopus (187) Google Scholar). Phage display and antibody library technologies have evolved as a powerful alternative for the generation of human antibody fragments for research, clinical, and therapeutic applications (for a review, see Ref. 4Kretzschmar T. von Rüden T. Curr. Opin. Biotechnol. 2002; 13: 598-602Crossref PubMed Scopus (110) Google Scholar), since it was shown that peptides (5Smith G.P. Science. 1985; 228: 1315-1317Crossref PubMed Scopus (3151) Google Scholar) and antibody fragments (6McCafferty J. Griffiths A.D. Winter G. Chiswell D.J. Nature. 1990; 348: 552-554Crossref PubMed Scopus (1949) Google Scholar) can be displayed on the surface of filamentous bacteriophage, and functional antibody fragments can be expressed in the periplasm of Escherichia coli cells (7Skerra A. Plückthun A. Science. 1988; 240: 1038-1041Crossref PubMed Scopus (836) Google Scholar, 8Better M. Chang C.P. Robinson R.R. Horwitz A.H. Science. 1988; 240: 1041-1043Crossref PubMed Scopus (486) Google Scholar). To date, a variety of different antibody libraries have been generated, which range from immune to naive and even synthetic antibody libraries. Immune libraries derived from IgG genes of immunized donors (9Sanna P.P. Williamson R.A. De Logu A. Bloom F.E. Burton D.R. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6439-6443Crossref PubMed Scopus (82) Google Scholar) are useful if immunized patients are available but have the disadvantage that antibodies can only be made against the antigens used for immunization. In contrast, antibodies against virtually any antigen, including self-, nonimmunogenic, or toxic antigens, can be isolated from naive or synthetic libraries. Naive libraries from nonimmunized donors have been generated by PCR-cloning Ig repertoires from various B-cell sources (10Marks J.D. Hoogenboom H.R. Bonnert T.P. McCafferty J. Griffiths A.D. Winter G. J. Mol. Biol. 1991; 222: 581-597Crossref PubMed Scopus (1437) Google Scholar, 11Vaughan T.J. Williams A.J. Pritchard K. Osbourn J.K. Pope A.R. Earnshaw J.C. McCafferty J. Hodits R.A. Wilton J. Johnson K.S. Nature Biotechnol. 1996; 14: 309-314Crossref PubMed Scopus (865) Google Scholar, 12Sheets M.D. Amersdorfer P. Finnern R. Sargent P. Lindqvist E. Schier R. Hemingsen G. Wong C. Gerhardt J.C. Marks J.D. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 6157-6162Crossref PubMed Scopus (381) Google Scholar, 13de Haard H.J. van Neer N. Reurs A. Hufton S.E. Roovers R.C. Henderikx P. de Bruine A.P. Arends J.-W. Hoogenboom H. J. Biol. Chem. 1999; 274: 18218-18230Abstract Full Text Full Text PDF PubMed Scopus (450) Google Scholar). Semisynthetic libraries were built by in vitro assembly of PCR-amplified antibody genes derived from human germ line genes and randomized only in the CDR3 1The abbreviations used are: CDR, complementarity-determining region; FCS, fetal calf serum; FGF, fibroblast growth factor; FGFR, fibroblast growth factor receptor; HuCAL®, human combinatorial antibody library; IL, interleukin; POD, peroxidase; scFv, single-chain Fv; TBS, Tris-buffered saline; FACS, fluorescence-activated cell sorting; PBS, phosphate-buffered saline; ELISA, enzyme-linked immunosorbent assay. regions (14Hoogenboom H.R. Winter G. J. Mol. Biol. 1992; 227: 381-388Crossref PubMed Scopus (391) Google Scholar, 15Nissim A. Hoogenboom H.R. Tomlinson I.M. Flynn G. Midgley C. Lane D. Winter G. EMBO J. 1994; 13: 692-698Crossref PubMed Scopus (528) Google Scholar, 16de Kruif J. Boel E. Logtenberg T. J. Mol. Biol. 1995; 248: 97-105Crossref PubMed Scopus (161) Google Scholar) or derived from a single V-gene with complete randomization of all CDRs (17Jirholt P. Ohlin M. Borrebaeck C.A.K. Söderlind E. Gene (Amst.). 1998; 215: 471-476Crossref PubMed Scopus (68) Google Scholar, 18Söderlind E. Strandberg L. Jirholt P. Kobayashi N. Alexeiva V. Aberg A.-M. Nilsson A. Jansson B. Ohlin M. Wingren C. Danielson L. Carlsson R. Borrebaeck C.A.K. Nature Biotechnol. 2000; 18: 852-856Crossref PubMed Scopus (276) Google Scholar). The affinities of the antibodies derived from these libraries are assumed to be dependent on the size of the library (19Perelson A.S. Immunol. Rev. 1989; 110: 5-36Crossref PubMed Scopus (292) Google Scholar). Therefore, many attempts have been undertaken to make the library size as big as possible, and site-specific recombination systems have been created to overcome the library size limitations given by the conventional cloning strategies (20Griffiths A.D. Willimands S.C. Hartley O. Tomlinson I.M. Waterhouse P. Crosby W.L. Kontermann R.E. Jones P.T. Low N.M. Allison T.J. Prospero T.D. Hoogenboom H.R. Nissim A. Cox J.P.L. Harrison J.L. Zaccolo M. Gherardi E. Winter G. EMBO J. 1994; 13: 3245-3260Crossref PubMed Scopus (879) Google Scholar, 21Sblattero D. Bradbury A. Nature Biotechnol. 2000; 18: 75-80Crossref PubMed Scopus (281) Google Scholar). Besides library generation, also the panning process itself limits the library size that can be handled conveniently. Therefore, it is important to generate libraries with a high quality of the displayed antibodies, thus emphasizing the functional library size and not only the apparent library size. With the recently described human combinatorial antibody library (HuCAL®) concept, a synthetic library in the scFv format was created, focusing on a high number of correct antibody fragments (1Knappik A. Ge L. Honegger A. Pack P. Fischer M. Wellnhofer G. Hoess A. Wölle J. Plückthun A. Virnekäs B. J. Mol. Biol. 2000; 296: 57-86Crossref PubMed Scopus (622) Google Scholar, 22Krebs B. Rauchenberger R. Reiffert S. Rothe C. Tesar M. Thomassen E. Cao M. Dreier T. Fischer D. Höss A. Inge L. Knappik A. Marget M. Pack P. Meng X.-Q. Schier R. Söhlemann P. Winter J. Wölle J. Kretzschmar T. J. Immunol. Methods. 2001; 254: 67-84Crossref PubMed Scopus (136) Google Scholar). HuCAL® is a fully human antibody library wherein each VH and VL subfamily frequently used in human is represented by a consensus framework, resulting in seven VH and seven VL master genes giving 49 different combinations. The master genes were optimized for expression and folding; furthermore, a high functional quality of the library is guaranteed by diversifying both CDR3 regions with trinucleotide mixtures (23Virnekaes B. Ge L. Plückthun A. Schneider K.C. Wellnhofer G. Moroney S. Nucleic Acids Res. 1994; 22: 5600-5607Crossref PubMed Scopus (183) Google Scholar), reflecting the natural amino acid composition of CDR3. The modular design of the library with unique restriction sites flanking the CDR and framework regions as well as compatible vector modules facilitate (i) conversion into different antibody formats, (ii) the addition of effector functions, and (iii) further antibody optimization by exchanging the CDR regions of selected binders by prebuilt CDR libraries (24Nagy Z.A. Hubner B. Loehning C. Rauchenberger R. Reiffert S. Thomassen-Wolf E. Zahn S. Leyer S. Schier E.-M. Zahradnik A. Brunner C. Lobenwein K. Rattel B. Stanglmaier M. Hallek M. Wing M. Anderson S. Dunn M. Kretzschmar T. Tesar M. Nat. Med. 2002; 8: 801-807Crossref PubMed Scopus (132) Google Scholar). Here, we describe the generation of a second version of the HuCAL® library (HuCAL®-Fab 1), wherein we combined all of the characteristics of the HuCAL® concept with the Fab format. Whereas scFv fragments have a high tendency to form multimers (25Weidner K.M. Denzin L.K. Voss Jr., E.W. J. Biol. Chem. 1992; 267: 10281-10288Abstract Full Text PDF PubMed Google Scholar, 26Holliger P. Prospero T. Winter G. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 6444-6448Crossref PubMed Scopus (795) Google Scholar, 27Marks J.D. Ouwehand W.H. Bye J.M. Finnern R. Gorick B.D. Voak D. Thorpe S.J. Hughes-Jones N.C. Winter G. Biotechnology. 1993; 11: 1145-1149Crossref PubMed Scopus (205) Google Scholar), Fab fragments tend to stay completely monomeric, allowing selection for affinity in contrast to selection for avidity. The monomeric behavior even in crude periplasmic extracts allows a rapid off-rate screening for affinity ranking of the antibody fragments without time-consuming purifications. Furthermore, the Fab format ensures conversion into complete immunoglobulins without loss of function, thus being optimally suited for therapeutic applications. HuCAL®-Fab 1 was challenged against FGFR3, so far resistant to generation of function-blocking antibodies. FGFR3 belongs to a family of highly homologous cell surface-expressed receptor tyrosine kinases, currently including five members (FGFR1 to 5) (reviewed in Ref. 28Klint P. Claesson-Welsh L. Front. Biosci. 1999; 4: D165-D177Crossref PubMed Google Scholar; Ref. 29Sleeman M. Fraser J. McDonald M. Yuan S. White D. Grandison P. Watson J.D. Murison J.G. Gene (Amst.). 2001; 271: 171-182Crossref PubMed Scopus (212) Google Scholar for FGFR5). The FGFRs are glycoproteins comprising two or three Ig-like extracellular domains, a hydrophobic transmembrane domain, and a cytoplasmic region that contains the tyrosine kinase domain (30Jaye M. Schlessinger J. Dionne C.A. Biochim. Biophys. Acta. 1992; 1135: 185-199Crossref PubMed Scopus (597) Google Scholar, 31Givol D. Yayon A. FASEB J. 1992; 6: 3362-3369Crossref PubMed Scopus (400) Google Scholar, 32Johnson D.E. Williams L.T. Adv. Cancer Res. 1993; 60: 1-41Crossref PubMed Scopus (1178) Google Scholar). Binding of a complex of fibroblast growth factor (FGF) ligand and heparin to FGFR3 leads to its dimerization, resulting in autophosphorylation of the intracellular kinase domain and a downstream activation of intracellular signaling cascades (33Kanai M. Goke M. Tsunekawa S. Podolsky D.K. J. Biol. Chem. 1997; 272: 6621-6628Abstract Full Text Full Text PDF PubMed Scopus (123) Google Scholar). So far, more than 20 ligands (34Wang Q. McEven D.G. Ornitz D.M. Mech. Dev. 2000; 90: 283-287Crossref PubMed Scopus (77) Google Scholar) have been identified, varying in their specificity for individual FGFRs (e.g. FGF2 shows a very promiscuous binding pattern, whereas FGF9 shows a high preference for FGFR2 and FGFR3) (35Hecht D. Zimmerman N. Bedford M. Avivi A. Yayon A. Growth Factors. 1995; 12: 223-233Crossref PubMed Scopus (100) Google Scholar). FGFs and their receptors regulate a multitude of cellular processes, including cell growth, differentiation, migration, and survival (reviewed in Refs. 36Basilico C. Moscatelli D. Adv. Cancer Res. 1992; 59: 115-165Crossref PubMed Scopus (1057) Google Scholar and 37Naski M.C. Ornitz D.M. Front. Biosci. 1998; 3: D781-D794Crossref PubMed Google Scholar). FGFR3 is mainly expressed during embryogenesis and development of bones, brain, lungs, and spinal cord (38Avivi A. Zimmer J. Yayon A. Yarden Y. Givol D. Oncogene. 1991; 6: 1089-1092PubMed Google Scholar, 39Peters K. Ornitz D.M. Werner S. Williams L. Dev. Biol. 1993; 155: 423-430Crossref PubMed Scopus (410) Google Scholar). Different pathological disorders such as dwarfism or tumor genesis were linked to increased FGFR3 activity (reviewed in Refs. 40Webster M.K. Donoghue D.J. Trends Genet. 1997; 13: 178-182Abstract Full Text PDF PubMed Scopus (270) Google Scholar and 41Burke D. Wilkes D. Blundell T.L. Malcolm S. Trends Biochem. Sci. 1998; 23: 59-62Abstract Full Text PDF PubMed Scopus (155) Google Scholar). The transforming potential of FGFR3 harboring an activating mutation was especially shown for tumor progression of ectopically expressed FGFR3 in NIH3T3 or mouse bone marrow cells (42Chesi M. Brents L.A. Ely S.A. Bais C. Robbiani D.F. Mesri E.A. Kuehl W.M. Bergsagel P.L. Blood. 2001; 97: 729-736Crossref PubMed Scopus (254) Google Scholar, 43Li Z. Zhu Y.X. Plowright E.E. Bergsagel P.L. Chesi M. Patterson B. Hawley T.S. Hawley R.G. Stewart A.K. Blood. 2001; 97: 2413-2419Crossref PubMed Scopus (84) Google Scholar). Correlation of FGFR3 expression and tumor genesis was demonstrated for multiple myeloma patients with the t(4;14) translocation (42Chesi M. Brents L.A. Ely S.A. Bais C. Robbiani D.F. Mesri E.A. Kuehl W.M. Bergsagel P.L. Blood. 2001; 97: 729-736Crossref PubMed Scopus (254) Google Scholar, 44Sibley K. Fenton J.A. Dring A.M. Ashcroft A.J. Rawstron A.C. Morgan G.J. Br. J. Haematol. 2002; 118: 514-520Crossref PubMed Scopus (30) Google Scholar), and in bladder carcinoma and cervical cancer (45Cappellen D. De Oliveira C. Ricol D. Gil Diez de Medina S. Bourdin J. Sastre-Garau X. Chopin D. Thiery J.P. Radvanyi F. Nat. Genet. 1999; 23: 18-20Crossref PubMed Scopus (489) Google Scholar, 46Billerey C. Chopin D. Aubrio-Lorton M.-H. Ricol D. Gil Diez de Medina S. Van Rhijn B. Bralet M.-P. Lefrere-Belda M.-A. Lahaye J.-B. Abbou C.C. Bonaventure J. Zafrani E.S. Van der Qast T. Thiery J.P. Radvanyi F. Am. J. Pathol. 2001; 158: 1955-1959Abstract Full Text Full Text PDF PubMed Scopus (403) Google Scholar, 47van Rhijn B.W. Montironi R. Zwarthoff E.C. Jobsis A.C. van der Kwast T.H. J. Pathol. 2002; 198: 245-251Crossref PubMed Scopus (151) Google Scholar). A prominent mutation is the exchange of glycine to arginine at position 380 (FGFR3G380R). The increased receptor activity caused by this amino acid change manifests itself in retarded bone growth, causing the most common form of achondroplasia and dwarfism (48Shiang R. Thompson L.M. Zhu Y.-Z. Church D.M. Filder T.J. Bocian M. Winokur S.T. Wasmuth J.J. Cell. 1994; 787: 335-342Abstract Full Text PDF Scopus (1104) Google Scholar, 49Naski M.C. Wang Q. Xu J. Ornitz D.M. Nat. Genet. 1996; 13: 233-237Crossref PubMed Scopus (422) Google Scholar). Accordingly, high affinity human antibodies that block FGFR3 activity could be of great therapeutic benefit in treating FGFR3-mediated skeletal disorders and tumor genesis. Here, we report the generation of the high quality, large size (2.1 × 1010) HuCAL®-Fab 1 library based on the already described HuCAL®-scFv library. HuCAL®-Fab 1 display and screening vectors were optimized for protein expression to overcome the problem that Fab fragments are often less optimally produced than scFv proteins (11Vaughan T.J. Williams A.J. Pritchard K. Osbourn J.K. Pope A.R. Earnshaw J.C. McCafferty J. Hodits R.A. Wilton J. Johnson K.S. Nature Biotechnol. 1996; 14: 309-314Crossref PubMed Scopus (865) Google Scholar). We show the generation of high affinity HuCAL®-Fab fragments with specificity for FGFR3, using a differential whole cell panning approach. Several selected antibodies completely inhibit FGFR3-mediated cell proliferation, some of which possess subnanomolar affinities without any maturation step. To the best of our knowledge, these are the first monoclonal antibodies to FGFRs that have function-blocking activity, which makes them promising candidates for therapeutic application. Enzymes, Antibodies, and Growth Factors—DNA restriction and modification enzymes as well as polymerases were purchased from New England Biolabs (Beverly, MA) and Roche Applied Science. Goat anti-human IgG (Fcγ-specific) (109-005-098), R-phycoerythrin-conjugated F(ab′)2 fragment of goat anti-human IgG (109-116-088), fluorescein isothiocyanate-conjugated goat anti-rabbit IgG (111-095-144), and POD-conjugated goat anti-human IgG (F(ab′)2 fragment-specific) (109-035-097) were supplied by Jackson Laboratories (West Grove, PA). POD-conjugated goat anti-human IgG (Fcγ fragment-specific) (A-0170), POD-conjugated goat anti-rabbit IgG (A-0545), POD-conjugated sheep anti-mouse IgG (whole molecule) (A-6782), and mouse anti-FLAG M2 antibody (F-3165) were purchased from Sigma. Recombinant IL-3 was supplied by PeproTech (London, UK). Recombinant FGF2 was expressed in E. coli. Cells were sonicated in PBS supplemented with a mixture of proteinase inhibitors (Complete; purchased from Roche Applied Science), and cleared lysate was batch-incubated with heparin-Sepharose (Amersham Biosciences). The resin was washed with PBS, and bound protein was eluted with PBS, 2 m NaCl. The partially purified material was diluted 10 times with PBS and subjected to a second purification step by fast protein liquid chromatography using a prepacked heparin-Sepharose column (Amersham Biosciences). NaCl gradient was employed to elute the FGF2, typically released from the heparin at 1 m NaCl. Recombinant FGF9 was a kind gift from Herbert Weich (GBF Braunschweig, Germany). Construction and Production of Recombinant FGFR Proteins—For cloning of an FGFR1-Fc fusion protein (FGFR11–371hFc), the primers 5′-CGGGGATCCGAGGTCATCACTGCCGGC-3′ and 5′-CGGGGTACCGGGATGTGGAGCTGGAAGTG-3′ were used to amplify the FGFR1 sequence from amino acid 1 to 371 from human cDNA by PCR. Fragments were cloned via the KpnI and BamHI restriction sites into the pCXFc vector, which is a pcDNA3.1/Zeo(+) (Invitrogen) vector with the sequence of human Fc between BamHI and XbaI sites. The FGFR3-Fc fusion protein (FGFR31–369hFc) was cloned by amplifying FGFR3 from human cDNA by PCR with the primers 5′-CGGGGATCCCCCGCCTCGTCAGCCTCC-3′ and 5′-CGGGGTACCGCGCGCTGCCTGAGG-3′ and cloning into pCXFc via the KpnI and BamHI restriction sites. cDNAs of both FGFR-Fc fusions were used to transfect 293T cells, and conditioned medium was harvested after 2 and 4 days. A soluble His-tagged FGFR3 (FGFR323–374TDhis) was constructed by amplifying FGFR3 by PCR with the primers 5′-ACGTGCTAGCTGAGTCCTTGGGGACGGAGCAG-3′ and 5′-ACGTCTCGAGTTAATGGTGATGGTGATGGTGTGCATACACACAGCCCGCCTCGTC-3′ from human cDNA. The fragment was cut with XhoI and cloned into pBlueScript (Stratagene, La Jolla, CA) digested with XhoI and EcoRV to obtain the plasmid (pBsFR323–374TDhis). Digesting this plasmid with NheI and XhoI resulted in a fragment encoding the extracellular domain of FGFR3, which was placed at the same sites in pCEP-Pu (50Kohfeldt E. Maurer P. Vannahme C. Timpl R. FEBS Lett. 1997; 414: 557-561Crossref PubMed Scopus (203) Google Scholar), generating pCEP-FR323–374TDhis. FGFR323–374TDhis was expressed transiently in 293E cells by transfection with pCEP-FR323–374TDhis and purified from supernatants with Ni2+-nitrilotriacetic acid beads (Qiagen, Hilden, Germany) followed by elution with a step gradient ranging from 20 mm to 0.5 m imidazole. SDS-PAGE and immunoblot analyses demonstrated peak amounts of purified FGFR323–374TDhis at 50 mm imidazole. Cell Lines and Medium—293 cells were cultured in Dulbecco's modified Eagle's medium, 10% FCS, penicillin/streptomycin, and glutamine (all from Invitrogen). Nontransformed rat chondrocytes (RCJ3.1C5.18) were transfected with full-length FGFR1 (RCJ-FGFR1), FGFR3 (RCJ-FGFR3), and FGFR3G380R (RCJ-FGFR3G380R) as described elsewhere (51Monsonego-Ornan E. Adar R. Feferman T. Segev O. Yayon A. Mol. Cell Biol. 2000; 20: 516-522Crossref PubMed Scopus (120) Google Scholar). Expression of FGFRs was regulated by a tetracycline suppression system (Invitrogen) expressing the receptor in the absence of tetracycline (–tet) and lacking the receptor if tetracycline was added to the culture medium (+tet). The mouse myeloid progenitor cell line FDCP-1 was cultured in Iscove's medium (Invitrogen) supplemented with 10% FCS, penicillin, streptomycin, glutamine, and 0.1 ng/ml IL-3. FDCP-1 cells transfected with full-length human FGFR1 (FGFR1 was a generous gift from Josef Schlessinger (New York University School of Medicine)) or human FGFR3 were grown in the same medium, but IL-3 was substituted by 10 ng/ml FGF2 or FGF9, respectively. Cloning of Fab Expression Vectors Containing Cysteines for Covalent Linkage of Heavy and Light Chains—Genes encoding the human CH1 (subtype IgG1; GenBank™ accession number A49444), Cκ, and Cλ domains (GenBank™ accession numbers P01834 and P01842, respectively) had been previously constructed by gene synthesis (1Knappik A. Ge L. Honegger A. Pack P. Fischer M. Wellnhofer G. Hoess A. Wölle J. Plückthun A. Virnekäs B. J. Mol. Biol. 2000; 296: 57-86Crossref PubMed Scopus (622) Google Scholar) without the region responsible for formation of the intermolecular disulfide bond. Gene fragments containing the cysteine codon at the 3′-end of CH1, Cκ, or Cλ, respectively, were constructed by adding additional nucleotides to a κ- and λ-Fab in the expression vector pMORPH®X7_Fab_FS (22Krebs B. Rauchenberger R. Reiffert S. Rothe C. Tesar M. Thomassen E. Cao M. Dreier T. Fischer D. Höss A. Inge L. Knappik A. Marget M. Pack P. Meng X.-Q. Schier R. Söhlemann P. Winter J. Wölle J. Kretzschmar T. J. Immunol. Methods. 2001; 254: 67-84Crossref PubMed Scopus (136) Google Scholar) containing a FLAG (52Prickett K.S. Amberg D.C. Hopp T.P. BioTechniques. 1989; 7: 580-589PubMed Google Scholar) and a Strep tag (53Schmidt T.G.M. Koepke J. Frank R. Skerra A. J. Mol. Biol. 1996; 255: 753-766Crossref PubMed Scopus (273) Google Scholar) by PCR using Pwo polymerase. 5′-GTGACGGTTAGCTCAGCGTC-3′ and 5′-GATATCTGCAGAATTCGCAGCTTTTCGGTTCCAC-3′ were used as primers for CH1, 5′-GAAATTAAACGTACGGTGGCTGC-3′ and 5′-TCCTCTAGATGCATGCTTATCAGCACTCGCCAC-3′ for Cκ, and 5′-GCGGCACGAAGTTAACCGTTC-3′ and 5′-TCCTCTAGATGCATGCTTATCAGCTGCACTCAGTCGG-3′ for Cλ. PCR products were gel-purified and digested with BlpI/EcoRI (CH1), BsiWI/SphI (Cκ), and HpaI/SphI (Cλ). Digests were separated on agarose gels and bands corresponding to the final products purified with the QIAquick gel extraction kit (Qiagen). FabCys vectors containing the cysteine codon at the 3′-end of CH1, Cκ, or Cλ, respectively, were constructed as follows. In the case of κ-Fabs, the BsiWI/EcoRI-digested vector fragment and in the case of λ-Fabs the HpaI/EcoRI vector fragment of pMORPH®X7_FS was ligated with the fragments encoding either Cκ (BsiWI/SphI) or Cλ (HpaI/SphI), CH1 (BlpI/EcoRI), and a fragment corresponding to the intergenic region, the PhoA signal sequence, and VH (SphI/BlpI) in a four-fragment ligation. E. coli JM83 (54Yanisch-Perron C. Vieira J. Messing J. Gene (Amst.). 1985; 33: 103-119Crossref PubMed Scopus (11472) Google Scholar) was subsequently transformed with the ligation products. The final expression vectors were designated pMORPH®X7_FabCys_FS. Cloning and Expression of Disulfide-linked and Noncovalently Linked Fab Fragments—Control κ-scFv and λ-scFvs were converted into the Fab format by cloning the VL and VH fragments into the expression vectors for disulfide-linked Fabs pMORPH®X7_FabCys_FS and the vectors for noncovalently linked Fabs pMORPH®X7_Fab_FS. The cloning procedure and expression in JM83 (54Yanisch-Perron C. Vieira J. Messing J. Gene (Amst.). 1985; 33: 103-119Crossref PubMed Scopus (11472) Google Scholar) were performed as described (22Krebs B. Rauchenberger R. Reiffert S. Rothe C. Tesar M. Thomassen E. Cao M. Dreier T. Fischer D. Höss A. Inge L. Knappik A. Marget M. Pack P. Meng X.-Q. Schier R. Söhlemann P. Winter J. Wölle J. Kretzschmar T. J. Immunol. Methods. 2001; 254: 67-84Crossref PubMed Scopus (136) Google Scholar). Purification from crude periplasmic extracts generated by osmotic shock (55Ausubel 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 and Sons, Inc., New York1998Google Scholar) was performed by StrepTactin affinity chromatography (IBA, Göttingen, Germany) (56Voss S. Skerra A. Protein Eng. 1997; 10: 975-982Crossref PubMed Scopus (239) Google Scholar). Thermal Stability of Disulfide-linked and Noncovalently Linked Fab Fragments—To analyze the heat stability of disulfide-linked and unlinked Fab-fragments, 1 μm solutions of purified proteins in PBS were incubated at temperatures of 4, 30, 40, 50, 60, and 70 °C for 30 min. Samples were centrifuged, and supernatants were tested for binding activity using the Biacore instrument. Biacore analysis was performed at 22 °C as described (22Krebs B. Rauchenberger R. Reiffert S. Rothe C. Tesar M. Thomassen E. Cao M. Dreier T. Fischer D. Höss A. Inge L. Knappik A. Marget M. Pack P. Meng X.-Q. Schier R. Söhlemann P. Winter J. Wölle J. Kretzschmar T. J. Immunol. Methods. 2001; 254: 67-84Crossref PubMed Scopus (136) Google Scholar) under mass transport-limited conditions. The binding signal at the end of the injection was measured. The binding signal of the sample incubated at 4 °C was set to 100%, and the binding signals of the samples incubated at the other temperatures were normalized to that value. Comparison of the OmpA and StII Signal Sequence for Light Chain Expression—The StII signal sequence for light chain expression in pMORPH®X7_Fab_FS (22Krebs B. Rauchenberger R. Reiffert S. Rothe C. Tesar M. Thomassen E. Cao M. Dreier T. Fischer D. Höss A. Inge L. Knappik A. Marget M. Pack P. Meng X.-Q. Schier R. Söhlemann P. Winter J. Wölle J. Kretzschmar T. J. Immunol. Methods. 2001; 254: 67-84Crossref PubMed Scopus (136) Google Scholar) was replaced by the OmpA signal sequence from a derivative of pIG10 (57Ge L. Knappik A. Pack P. Freund C. Plückthun A. Borrebaeck C.A.K. Antibody Engineering. Oxford University Press, Oxford, UK1995: 229-266Google Scholar) via XbaI/EcoRV. The new expression vector was designated pMORPH®X9_Fab_FS. Heavy and light chains of the control κ-Fab and λ-Fab were excised from pMORPH®X7_Fab_FS and cloned into pMORPH®X9_Fab_FS by EcoRV/EcoRI. Expression yields of these Fabs containing either the StII or the OmpA signal sequence were measured as follows. After expression in JM83 (54Yanisch-Perron C. Vieira J. Messing J. Gene (Amst.). 1985; 33: 103-119Crossref PubMed Scopus (11472) Google Scholar, 22Krebs B. Rauchenberger R. Reiffert S. Rothe C. Tesar M. Thomassen E. Cao M. Dreier T. Fischer D. Höss A. Inge L. Knappik A. Marget M. Pack P. Meng X.-Q. Schier R. Söhlemann P. Winter J. Wölle J. Kretzschmar T. J. Immunol. Methods. 2001; 254: 67-84Crossref PubMed Scopus (136) Google Scholar), crude periplasmic extracts were generated by osmotic shock (55Ausubel 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 and Sons, Inc., New York1998Google Scholar), and Fab protein was analyzed by ELISA. 500 ng of the corresponding antigens were coated on Maxisorp plates (Nunc, Rochester, NY), and after blocking with 5% nonfat milk powder in TBS, 0.05% Tween 20 (Sigma), periplasmic extracts were added in
Referência(s)