High Resolution Mapping of the Binding Site on Human IgG1 for FcγRI, FcγRII, FcγRIII, and FcRn and Design of IgG1 Variants with Improved Binding to the FcγR

Immunoglobulin G (IgG) Fc receptors play a critical role in linking IgG antibody-mediated immune responses with cellular effector functions. A high resolution map of the binding site on human IgG1 for human FcγRI, FcγRIIA, FcγRIIB, FcγRIIIA, and FcRn receptors has been determined. A common set of IgG1 residues is involved in binding to all FcγR; FcγRII and FcγRIII also utilize residues outside this common set. In addition to residues which, when altered, abrogated binding to one or more of the receptors, several residues were found that improved binding only to specific receptors or simultaneously improved binding to one type of receptor and reduced binding to another type. Select IgG1 variants with improved binding to FcγRIIIA exhibited up to 100%enhancement in antibody-dependent cell cytotoxicity using human effector cells; these variants included changes at residues not found at the binding interface in the IgG/FcγRIIIA co-crystal structure (Sondermann, P., Huber, R., Oosthuizen, V., and Jacob, U. (2000)Nature 406, 267–273). These engineered antibodies may have important implications for improving antibody therapeutic efficacy. Immunoglobulin G (IgG) Fc receptors play a critical role in linking IgG antibody-mediated immune responses with cellular effector functions. A high resolution map of the binding site on human IgG1 for human FcγRI, FcγRIIA, FcγRIIB, FcγRIIIA, and FcRn receptors has been determined. A common set of IgG1 residues is involved in binding to all FcγR; FcγRII and FcγRIII also utilize residues outside this common set. In addition to residues which, when altered, abrogated binding to one or more of the receptors, several residues were found that improved binding only to specific receptors or simultaneously improved binding to one type of receptor and reduced binding to another type. Select IgG1 variants with improved binding to FcγRIIIA exhibited up to 100%enhancement in antibody-dependent cell cytotoxicity using human effector cells; these variants included changes at residues not found at the binding interface in the IgG/FcγRIIIA co-crystal structure (Sondermann, P., Huber, R., Oosthuizen, V., and Jacob, U. (2000)Nature 406, 267–273). These engineered antibodies may have important implications for improving antibody therapeutic efficacy. monoclonal antibody antibody-dependent cell cytotoxicity antibody-independent cell cytotoxicity enzyme-linked immunosorbent assay IgG Fc γ-receptor neonatal IgG Fc receptor human glutathione S-transferase lactate dehydrogenase maximal response natural killer cells peripheral blood monocytes (R)-phycoerythrin spontaneous release vascular endothelial growth factor polymerase chain reaction matrix-assisted laser desorption/ionization time-of-flight mass spectrometry Chinese hamster ovary phosphate-buffered saline bovine serum albumin Monoclonal antibodies (mAbs)1 are increasingly being used as therapeutics in human disease (1King D.J. Adair J.R. Curr. Opin. Drug Discovery Dev. 1999; 2: 110-117PubMed Google Scholar, 2Vaswani S.K. Hamilton R.G. Ann. Allergy Asthma Immunol. 1998; 81: 105-119Abstract Full Text PDF PubMed Scopus (49) Google Scholar, 3Holliger P. Hoogenboom H. Nat. Biotechnol. 1998; 16: 1015-1016Crossref PubMed Scopus (36) Google Scholar). Although some of these, e.g. mAbs that bind to a receptor or ligand and thereby block ligand-receptor interaction, may function without utilizing antibody effector mechanisms, other mAbs may need to recruit the immune system to kill the target cell (4Clynes R.A. Towers T.L. Presta L.G. Ravetch J.V. Nat. Med. 2000; 6: 443-446Crossref PubMed Scopus (2340) Google Scholar, 5Clynes R. Takechi Y. Moroi Y. Houghton A. Ravetch J.V. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 652-656Crossref PubMed Scopus (269) Google Scholar, 6Anderson D.R. Grillo-Lopez A. Varns C. Chambers K.S. Hanna N. Biochem. Soc. Trans. 1997; 25: 705-708Crossref PubMed Scopus (240) Google Scholar). If immune system recruitment is desirable for a therapeutic mAb, engineering the IgG Fc portion to improve effector function (via improved binding to IgG receptors and/or complement) could be a valuable enhancement to antibody therapeutics. Currently, immune system recruitment can be abrogated by altering IgG residues in the lower hinge region (7Armour K.L. Clark M.R. Hadley A.G. Williamson L.M. Eur. J. Immunol. 1999; 29: 2613-2624Crossref PubMed Scopus (157) Google Scholar, 8Duncan A.R. Woof J.M. Partridge L.J. Burton D.R. Winter G. Nature. 1988; 332: 563-564Crossref PubMed Scopus (263) Google Scholar), using human IgG2 or IgG4 subclasses that are comparatively inefficient in effector function or using antibody F(ab) or F(ab′)2fragments (although these may have undesirable rapid clearance rates). There are few methods that improve immune system recruitment; these include bispecific antibodies, in which one arm of the antibody binds to an IgG receptor (9Weiner L.M. Alpaugh R.K. von Mehren M. Cancer Immunol. Immunother. 1997; 45: 190-192Crossref PubMed Scopus (11) Google Scholar), cytokine-IgG fusion proteins (10Peng L.S. Penichet M.L. Morrison S.L. J. Immunol. 1999; 163: 250-258PubMed Google Scholar), and optimization of the Asn297-linked carbohydrate (11Uma-a P. Jean-Mairet J. Moudry R. Amstutz H. Bailey J.E. Nat. Biotechnol. 1999; 17: 176-180Crossref PubMed Scopus (650) Google Scholar, 12Lifely M.R. Hale C. Royce S. Keen M.J. Phillips J. Glycobiology. 1995; 5: 813-822Crossref PubMed Scopus (156) Google Scholar). Alteration of clearance rate is also being investigated (13Ghetie V. Popov S. Borvak J. Radu C. Matesol D. Medesan C. Ober R.J. Ward E.S. Nat. Biotechnol. 1997; 15: 637-640Crossref PubMed Scopus (223) Google Scholar). IgG Fc receptors play a critical role in linking IgG antibody-mediated immune responses with cellular effector functions. The latter include release of inflammatory mediators, endocytosis of immune complexes, phagocytosis of microorganisms, antibody-dependent cellular cytotoxicity (ADCC), and regulation of immune system cell activation (14Gessner J.E. Heiken H. Tamm A. Schmidt R.E. Ann. Hematol. 1998; 76: 231-248Crossref PubMed Scopus (345) Google Scholar, 15Gavin A. Hulett M. Hogarth P.M. van de Winkel J.G.J. Hogarth P.M. The Immunoglobulin Receptors and Their Physiological and Pathological Roles in Immunity. Kluwer Academic Publishers Group, Dordrecht, The Netherlands1998: 11-35Crossref Google Scholar, 16Sautes C. Fridman W.H. Sautes C. Cell-mediated Effects of Immunoglobulins. R. G. Landes Co., Austin, TX1997: 29-66Crossref Google Scholar, 17Da'ron M. Annu. Rev. Immunol. 1997; 15: 203-234Crossref PubMed Scopus (1050) Google Scholar). One group of IgG Fc receptors, FcγR, are expressed on leukocytes and are composed of three distinct classes as follows: FcγRI (CD64), FcγRII (CD32), and FcγRIII (CD16). In humans, the latter two classes can be further divided into FcγRIIA and FcγRIIB, FcγRIIIA and FcγRIIIB. Structurally, the FcγR are all members of the immunoglobulin superfamily, having an IgG-binding α-chain with an extracellular portion composed of either two (FcγRII and FcγRIII) or three (FcγRI) Ig-like domains. In addition, FcγRI and FcγRIII have accessory protein chains (γ and ζ) associated with the α-chain that function in signal transduction. The receptors are also distinguished by their affinity for IgG. FcγRI exhibits a high affinity for IgG, Ka = 108–109m−1 (14Gessner J.E. Heiken H. Tamm A. Schmidt R.E. Ann. Hematol. 1998; 76: 231-248Crossref PubMed Scopus (345) Google Scholar), and can bind monomeric IgG. In contrast, FcγRII and FcγRIII show a weaker affinity for monomeric IgG, Ka ≤ 107m−1 (14Gessner J.E. Heiken H. Tamm A. Schmidt R.E. Ann. Hematol. 1998; 76: 231-248Crossref PubMed Scopus (345) Google Scholar), and hence can only interact effectively with multimeric immune complexes. Given the interest in and increasing use of antibody therapeutics, a comprehensive mapping of the binding site on human IgG for the different FcγR could provide for alternative methods of either abrogating or enhancing immune recruitment via FcγR. Previous studies mapped the binding site on human and murine IgG for FcγR primarily to the lower hinge region composed of IgG residues 233–239 (Eu numbering, see Ref. 18Kabat E.A. Wu T.T. Perry H.M. Gottesman K.S. Foeller C. Sequences of Proteins of Immunological Interest. 5th Ed. United States Public Health Service, National Institutes of Health, Bethesda1991Google Scholar) (8Duncan A.R. Woof J.M. Partridge L.J. Burton D.R. Winter G. Nature. 1988; 332: 563-564Crossref PubMed Scopus (263) Google Scholar, 14Gessner J.E. Heiken H. Tamm A. Schmidt R.E. Ann. Hematol. 1998; 76: 231-248Crossref PubMed Scopus (345) Google Scholar, 15Gavin A. Hulett M. Hogarth P.M. van de Winkel J.G.J. Hogarth P.M. The Immunoglobulin Receptors and Their Physiological and Pathological Roles in Immunity. Kluwer Academic Publishers Group, Dordrecht, The Netherlands1998: 11-35Crossref Google Scholar, 16Sautes C. Fridman W.H. Sautes C. Cell-mediated Effects of Immunoglobulins. R. G. Landes Co., Austin, TX1997: 29-66Crossref Google Scholar, 17Da'ron M. Annu. Rev. Immunol. 1997; 15: 203-234Crossref PubMed Scopus (1050) Google Scholar, 19Canfield S.M. Morrison S.L. J. Exp. Med. 1991; 173: 1483-1491Crossref PubMed Scopus (252) Google Scholar, 20Chappel M.S. Isenman D.E. Everett M. Xu Y.-Y. Dorrington K.J. Klein M.H. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 9036-9040Crossref PubMed Scopus (119) Google Scholar, 21Woof J.M. Partridge L.J. Jefferis R. Burton D.R. Mol. Immunol. 1986; 23: 319-330Crossref PubMed Scopus (103) Google Scholar, 22Wines B.D. Powell M.S. Parren P.W.H.I. Barnes N. Hogarth P.M. J. Immunol. 2000; 164: 5313-5318Crossref PubMed Scopus (114) Google Scholar). Other studies proposed additional broad segments, e.g.Gly316–Lys338 for human FcγRI (21Woof J.M. Partridge L.J. Jefferis R. Burton D.R. Mol. Immunol. 1986; 23: 319-330Crossref PubMed Scopus (103) Google Scholar), Lys274–Arg301 and Tyr407–Arg416 for human FcγRIII (23Sarmay G. Benzcur M. Petranyi G. Klein E. Kahn M. Stanworth D.R. Gergely J. Mol. Immunol. 1984; 21: 43-51Crossref PubMed Scopus (21) Google Scholar, 24Gergely J. Sandor M. Sarmay G. Uher F. Biochem. Soc. Trans. 1984; 12: 739-743Crossref PubMed Scopus (4) Google Scholar), or found few specific residues outside the lower hinge, e.g.Asn297 and Glu318 for murine IgG2b interacting with murine FcγRII (25Lund J. Pound J.D. Jones P.T. Duncan A.R. Bentley T. Goodall M. Levine B.A. Jefferis R. Winter G. Mol. Immunol. 1992; 29: 53-59Crossref PubMed Scopus (48) Google Scholar). The very recent report of the 3.2-Å crystal structure of the human IgG1 Fc fragment with human FcγRIIIA delineated IgG1 residues Leu234–Ser239, Asp265–Glu269, Asn297–Thr299, and Ala327–Ile332 as involved in binding to FcγRIIIA (26Sondermann P. Huber R. Oosthuizen V. Jacob U. Nature. 2000; 406: 267-273Crossref PubMed Scopus (606) Google Scholar). The current study provides a complete, high resolution mapping of human IgG1 for human FcγR receptors (FcγRI, FcγRIIA, FcγRIIB, and FcγRIIIA) as well as for human FcRn, an Fc receptor belonging to the major histocompatability complex structural class, which is involved in IgG transport and clearance (27Raghavan M. Bjorkman P.J. Annu. Rev. Cell Dev. Biol. 1996; 12: 181-220Crossref PubMed Scopus (268) Google Scholar, 28Ward E.S. Ghetie V. Ther. Immunol. 1995; 2: 77-94PubMed Google Scholar). The binding site on human IgG1 for the various receptors was determined by individually changing all solvent-exposed amino acids in human IgG1 CH2 and CH3 domains, based on the crystal structure of human IgG1 Fc (30Deisenhofer J. Biochemistry. 1981; 20: 2361-2370Crossref PubMed Scopus (1365) Google Scholar), to Ala. A common set of IgG1 residues is involved in binding to all FcγR; FcγRII and FcγRIII also utilize distinct residues in addition to this common set. As well as residues that abrogated binding to one or more Fc receptors when changed to Ala, several positions were found that improved binding only to specific receptors or simultaneously improved binding to one type of FcγR and reduced binding to another type. Notably, for both FcγRIIIA and FcRn, which have crystal structures of complexes with IgG available (26Sondermann P. Huber R. Oosthuizen V. Jacob U. Nature. 2000; 406: 267-273Crossref PubMed Scopus (606) Google Scholar, 29Burmeister W.P. Huber A.H. Bjorkman P.J. Nature. 1994; 372: 379-383Crossref PubMed Scopus (415) Google Scholar), several IgG residues not found at the IgG:receptor interface had a profound effect on binding and biological activity. Select IgG1 variants with improved binding to FcγRIIIA showed an enhancement in ADCC when either peripheral blood monocyte cells (PBMC) or natural killer cells (NK) were used. These variants may have important implications for using Fc-engineered antibodies for improved therapeutic efficacy. The cDNAs encoding extracellular and transmembrane domains of human FcγRIIA (CD32A; His131 allotype), FcγRIIB (CD32B), and FcγRIIIA (CD16A; Val158 allotype) were provided by Dr. J. Ravetch (Rockefeller University, New York). FcγRIIA-Arg131 allotype and FcγRIIIA-Phe158allotype were generated by site-directed mutagenesis (31Kunkel T.P. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 488-492Crossref PubMed Scopus (4903) Google Scholar). The cDNA for FcγRI (CD64) was isolated by reverse transcriptase-PCR (GeneAmp, PerkinElmer Life Sciences) of oligo(dT)-primed RNA from U937 cells using primers that generated a fragment encoding the α-chain extracellular domain. The cDNAs encoding human neonatal Fc receptor (FcRn) α-chain, β2-microglobulin subunit, and human FcγR γ-chain were obtained from the I.M.A.G.E. Consortium (32Lennon G.G. Auffray C. Polymeropoulos M. Soares M.B. Genomics. 1996; 33: 151-152Crossref PubMed Scopus (1089) Google Scholar). The coding regions of all receptors were subcloned into previously described pRK mammalian cell expression vectors (33Eaton D.L. Wood W.I. Eaton D. Hass P.E. Hollingshead P. Wion K. Mather J. Lawn R.M. Vehar G.A. Gorman C. Biochemistry. 1986; 25: 8343-8347Crossref PubMed Scopus (205) Google Scholar). For all FcγR and the FcRn α-chain pRK plasmids, the transmembrane and intracellular domains were replaced by DNA encoding a Gly-His6 tag and human glutathione S-transferase (GST). The 234-amino acid GST sequence was obtained by PCR from the pGEX-4T2 plasmid (Amersham Pharmacia Biotech) with NheI andXbaI restriction sites at the 5′ and 3′ ends, respectively. Thus, the expressed proteins contained the extracellular domains of the α-chain fused at their carboxyl termini to Gly/His6/GST at amino acid positions as follows: FcγRI, His292; FcγRIIA, Met216; FcγRIIB, Met195; FcγRIIIA, Gln191; FcRn, Ser297 (residue numbers include signal peptides). Plasmids were transfected into the adenovirus-transformed human embryonic kidney cell line 293 by calcium phosphate precipitation (34Gorman C.M. Gies D.R. McCray G. DNA Protein Eng. Tech. 1990; 2: 3-10Google Scholar). Supernatants were collected 72 h after conversion to serum-free PSO4 medium supplemented with 10 mg/liter recombinant bovine insulin, 1 mg/liter human transferrin, and trace elements. Proteins were purified by nickel-nitrilotriacetic acid chromatography (Qiagen, Valencia, CA) and buffer exchanged into phosphate-buffered saline (PBS) using Centriprep-30 concentrators (Millipore, Bedford, MA). Proteins were analyzed on 4–20%SDS-polyacrylamide gels (NOVEX, San Diego, CA), transferred to polyvinylidene difluoride membranes (NOVEX), and their amino termini sequenced to ensure proper signal sequence cleavage. Receptor conformation was evaluated by ELISA using murine monoclonals 32.2 (anti-FcγRI), IV.3 (anti-FcγRII), 3G8 (anti-FcγRIII) (Medarex, Annandale, NJ), and B1G6 (anti-β2-microglobulin) (Beckman Coulter, Palo Alto, CA). Receptor concentrations were determined by amino acid analysis. The humanized IgG1 anti-IgE E27, an affinity-matured variant of anti-IgE E25, binds to the Fcε3 domain of human IgE (35Presta L.G. Lahr S.J. Shields R.L. Porter J.P. Gorman C.M. Jardieu P.M. J. Immunol. 1993; 151: 2623-2632PubMed Google Scholar). When mixed with human IgE in a 1:1 molar ratio, the IgE and anti-IgE form a hexameric complex composed of three IgE and three anti-IgE (36Liu J. Lester P. Builder S. Shire S.J. Biochemistry. 1995; 34: 10474-10482Crossref PubMed Scopus (140) Google Scholar). Site-directed mutagenesis (31Kunkel T.P. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 488-492Crossref PubMed Scopus (4903) Google Scholar) on E27 IgG1 was used to generate IgG1 variants in which all solvent-exposed residues in the CH2 and CH3 domains were individually altered to Ala; selection of solvent-exposed residues was based on the crystal structure of human IgG1 Fc (30Deisenhofer J. Biochemistry. 1981; 20: 2361-2370Crossref PubMed Scopus (1365) Google Scholar). Human IgG2, IgG3, and IgG4 isotypes of E27 were constructed by subcloning the appropriate heavy chain Fc cDNAs from a human spleen cDNA library into a pRK vector containing the E27 variable heavy domain. All IgG isotypes and variants were expressed using the same E27 κ light chain. Following cotransfection of heavy and light chain plasmids into 293 cells, IgG1, IgG2, IgG4, and variants were purified by protein A chromatography (Amersham Pharmacia Biotech). IgG3 isotype was purified using protein G chromatography (Amersham Pharmacia Biotech). All proteins were analyzed by SDS-polyacrylamide gel electrophoresis. Protein concentrations were determined usingA280 and verified using amino acid composition analysis and a human IgG Fc ELISA. IgGs were also tested for their binding to human IgE in an ELISA format to ensure that they bound IgE as well as native E27. Structural integrity of the variants was also evident by their ability to be purified using protein A (which binds at the CH2:CH3 domain interface (30Deisenhofer J. Biochemistry. 1981; 20: 2361-2370Crossref PubMed Scopus (1365) Google Scholar)) as well as all variants, except P329A, binding similar to native IgG1 to at least one of the five receptors. FcγRIIA, FcγRIIB, and FcγRIIIA fusion proteins at 1 μg/ml in PBS, pH 7.4, were coated onto ELISA plates (Nalge-Nunc, Naperville, IL) for 48 h at 4 °C. Plates were blocked with Tris-buffered saline, 0.5%bovine serum albumin, 0.05%polysorbate-20, 2 mm EDTA, pH 7.45 (assay buffer), at 25 °C for 1 h. E27-IgE hexameric complexes were prepared in assay buffer by mixing equimolar amounts of E27 and human myeloma IgE (37Nilsson K. Bennich H. Johansson S.G.O. Ponten J. Clin. Exp. Immunol. 1970; 7: 477-489PubMed Google Scholar) at 25 °C for 1 h. Serial 3-fold dilutions of native E27 standard or variant complexes (10.0–0.0045 μg/ml) were added to plates and incubated for 2 h. After washing plates with assay buffer, bound complexes to FcγRIIA and FcγRIIB were detected with peroxidase-conjugated F(ab′)2 fragment of goat anti-human F(ab′)2-specific IgG (Jackson ImmunoResearch, West Grove, PA). Binding of complexes to FcγRIIIA was detected with peroxidase-conjugated protein G (Bio-Rad). The substrate used was o-phenylenediamine dihydrochloride (Sigma). Absorbance at 490 nm was read using a Vmax plate reader (Molecular Devices, Mountain View, CA). Any contribution to binding via interaction of the IgE in the E27-IgE complexes with the human FcγRII and FcγRIIIA was not apparent based on the lack of binding of several Ala variants (Class 1, Table I).Table IBinding of human IgG1 variants to human FcRn and FcγRVariant1-aResidue numbers are according to the Eu numbering system (18). Variants that had no effect on binding (i.e. did not reduce binding by more than 60%or improve binding by more than 20%) to FcγR or FcRn were as follows: Lys246, Lys248, Asp249, Met252, Thr260, Lys274, Tyr278, Val282, Glu283, Thr289, Glu294, Y300F, Glu318, Lys320, Ser324, A330Q, Thr335, Lys340, Gln342, Arg344, Glu345, Gln347, Arg355, Glu356, Met358, Thr359, Lys360, Asn361, Tyr373, Ser375, Ser383, Asn384, Gln386, Glu388, Asn389, Asn390, Y391F, Lys392, Leu398, Ser400, Asp401, Asp413, Arg416, Gln418, Gln419, Asn421, Val422, Thr437, Gln438, Lys439, Ser440, Ser442, Ser444, and Lys447.FcRn1-bValues are the ratio of binding of the variant to that of native IgG1 at 0.33 or 1 μg/ml. A value greater than 1 denotes binding of the variant was improved compared with native IgG1, whereas a ratio less than 1 denotes reduced binding compared with native IgG1. Reduced binding to any given receptor was defined as a reduction of ≥40%compared to native IgG; better binding was defined as an improvement of ≥25%compared with native IgG1.mean(S.D.)nFcγRI mean(S.D.)nFcγRIIA mean(S.D.)FcγRIIB mean(S.D.)FcγRIIIA mean(S.D.)n1-cNumber of independent assays for FcγRIIA, FcγRIIB and FcγRIIIA. At least two separately expressed and purified lots of each variant were assayed.Class 1, reduced binding to all FcγRE233P0.54 (0.20) 30.12 (0.06) 60.08 (0.01)0.12 (0.01)0.04 (0.02)2L234VL235AG236 deletedP238A1.49 (0.17) 30.60 (0.05) 50.38 (0.14)0.36 (0.15)0.07 (0.05)4D265A1.23 (0.14) 40.16 (0.05) 90.07 (0.01)0.13 (0.05)0.09 (0.06)4N297A0.80 (0.18) 80.15 (0.06) 70.05 (0.00)0.10 (0.02)0.03 (0.01)3A327Q0.970.60 (0.12) 90.13 (0.03)0.14 (0.03)0.06 (0.01)4P329A0.800.48 (0.10) 60.08 (0.02)0.12 (0.08)0.21 (0.03)4Class 2, reduced binding to FcγRII and FcγRIIIAD270A1.050.76 (0.12) 60.06 (0.02)0.10 (0.06)0.14 (0.04)6Q295A0.791.00 (0.11) 40.62 (0.20)0.50 (0.24)0.25 (0.09)5A327S0.86 (0.03) 40.23 (0.06)0.22 (0.05)0.06 (0.01)4Class 3, improved binding to FcγRII and FcγRIIIAT256A1.91 (0.43) 61.01 (0.07) 51.41 (0.27)2.06 (0.66)1.32 (0.18)9K290A0.79 (0.14) 31.01 (0.06) 111.30 (0.21)1.38 (0.17)1.31 (0.19)9Class 4, improved binding to FcγRII and no effect on FcγRIIIAR255A0.59 (0.19) 40.99 (0.12) 71.30 (0.20)1.59 (0.42)0.98 (0.18)5E258A1.181.18 (0.13) 41.33 (0.22)1.65 (0.38)1.12 (0.12)5S267A1.081.09 (0.08) 101.52 (0.22)1.84 (0.43)1.05 (0.24)11E272A1.34 (0.24) 41.05 (0.06) 71.23 (0.12)1.53 (0.22)0.80 (0.18)6N276A1.15 (0.21) 31.05 (0.14) 41.29 (0.20)1.34 (0.40)0.95 (0.04)4D280A0.821.04 (0.08) 101.34 (0.14)1.60 (0.31)1.09 (0.20)10H285A0.850.96 (0.07) 41.26 (0.12)1.23 (0.15)0.87 (0.04)4N286A1.24 (0.04) 20.95 (0.18) 161.24 (0.23)1.36 (0.15)1.05 (0.04)6T307A1.81 (0.32) 60.99 (0.14) 41.07 (0.15)1.27 (0.24)1.09 (0.18)10L309A0.63 (0.18) 40.93 (0.18) 61.13 (0.08)1.26 (0.12)1.07 (0.20)3N315A0.76 (0.14) 30.99 (0.16) 61.15 (0.06)1.30 (0.17)1.07 (0.21)5K326A1.031.03 (0.05) 101.23 (0.20)1.41 (0.27)1.23 (0.23)7P331A0.851.01 (0.09) 71.29 (0.14)1.34 (0.35)1.08 (0.19)4S337A1.031.17 (0.23) 31.22 (0.30)1.26 (0.06)0.94 (0.18)4A378Q1.32 (0.13) 31.06 (0.05) 31.40 (0.17)1.45 (0.17)1.19 (0.17)5E430A0.93 (0.03) 21.05 (0.02) 31.24 (0.11)1.28 (0.10)1.20 (0.18)5Class 5, improved binding to FcγRII and reduced binding to FcγRIIIAH268A1.02 (0.22) 31.09 (0.11) 81.21 (0.14)1.44 (0.22)0.54 (0.12)13R301A0.861.06 (0.10) 41.14 (0.13)1.29 (0.16)0.22 (0.08)7K322A0.980.94 (0.04) 91.17 (0.11)1.28 (0.21)0.62 (0.12)6Class 6, reduced binding to FcγRII and no effect on FcγRIIIAR292A0.81 (0.18) 40.95 (0.05) 80.27 (0.13)0.17 (0.07)0.89 (0.17)10K414A1.021.00 (0.04) 30.64 (0.15)0.58 (0.18)0.82 (0.27)3Class 7, reduced binding to FcγRII and improved binding to FcγRIIIAS298A0.801.11 (0.03) 90.40 (0.15)0.23 (0.13)1.34 (0.20)16Class 8, no effect on FcγRII and reduced binding to FcγRIIIAS239A1.060.81 (0.09) 70.73 (0.25)0.76 (0.36)0.26 (0.08)3E269A1.050.61 (0.14) 90.65 (0.18)0.75 (0.29)0.45 (0.13)5E293A0.851.11 (0.07) 41.08 (0.19)1.07 (0.20)0.31 (0.13)6Y296F0.791.03 (0.09) 80.97 (0.23)0.86 (0.17)0.55 (0.12)6V303A1.26 (0.21) 30.91 (0.11) 50.86 (0.10)0.65 (0.17)0.33 (0.09)8A327G0.96 (0.01) 30.92 (0.09)0.83 (0.10)0.36 (0.05)3K338A1.140.90 (0.05) 30.78 (0.09)0.63 (0.08)0.15 (0.01)2D376A1.45 (0.36) 41.00 (0.05) 30.80 (0.16)0.68 (0.14)0.55 (0.10)5Class 9, no effect on FcγRII and improved binding to FcγRIIIAE333A1.03 (0.01) 20.98 (0.15) 50.92 (0.12)0.76 (0.11)1.27 (0.17)10K334A1.05 (0.03) 21.06 (0.06) 111.01 (0.15)0.90 (0.12)1.39 (0.19)16A339T1.06 (0.04) 61.09 (0.03)1.20 (0.03)1.34 (0.09)2Class 10, affect only FcRnI253A<0.100.96 (0.05) 41.14 (0.02)1.18 (0.06)1.08 (0.14)3S254A<0.100.96 (0.08) 40.97 (0.24)1.15 (0.38)0.73 (0.14)3K288A0.38 (0.12) 50.88 (0.15) 151.15 (0.26)1.14 (0.20)1.06 (0.04)4V305A1.46 (0.48) 61.04 (0.19) 101.12 (0.12)1.23 (0.22)0.84 (0.15)4Q311A1.62 (0.25) 40.93 (0.05) 41.11 (0.06)1.19 (0.13)0.93 (0.17)3D312A1.50 (0.06) 41.01 (0.12) 51.20 (0.24)1.19 (0.07)1.23 (0.14)3K317A1.44 (0.18) 40.92 (0.17) 61.13 (0.05)1.18 (0.27)1.10 (0.23)4K360A1.30 (0.08) 41.02 (0.04) 31.12 (0.10)1.12 (0.08)1.23 (0.16)6Q362A1.25 (0.24) 31.00 (0.04) 31.03 (0.10)1.02 (0.03)1.03 (0.16)4E380A2.19 (0.29) 61.04 (0.06) 31.18 (0.01)1.07 (0.05)0.92 (0.12)2E382A1.51 (0.18) 41.06 (0.03) 30.95 (0.11)0.84 (0.04)0.76 (0.17)3S415A0.441.04 (0.03) 30.90 (0.11)0.88 (0.05)0.86 (0.18)2S424A1.41 (0.14) 30.98 (0.03) 31.04 (0.06)1.02 (0.02)0.88 (0.09)2H433A0.41 (0.14) 20.98 (0.03) 30.92 (0.18)0.79 (0.18)1.02 (0.15)3N434A3.46 (0.37) 71.00 (0.04) 30.96 (0.06)0.97 (0.12)0.77 (0.13)6H435A<0.10 41.25 (0.09) 30.77 (0.05)0.72 (0.05)0.78 (0.03)3Y436A<0.10 20.99 (0.02) 20.93 (0.05)0.91 (0.06)0.91 (0.15)31-a Residue numbers are according to the Eu numbering system (18Kabat E.A. Wu T.T. Perry H.M. Gottesman K.S. Foeller C. Sequences of Proteins of Immunological Interest. 5th Ed. United States Public Health Service, National Institutes of Health, Bethesda1991Google Scholar). Variants that had no effect on binding (i.e. did not reduce binding by more than 60%or improve binding by more than 20%) to FcγR or FcRn were as follows: Lys246, Lys248, Asp249, Met252, Thr260, Lys274, Tyr278, Val282, Glu283, Thr289, Glu294, Y300F, Glu318, Lys320, Ser324, A330Q, Thr335, Lys340, Gln342, Arg344, Glu345, Gln347, Arg355, Glu356, Met358, Thr359, Lys360, Asn361, Tyr373, Ser375, Ser383, Asn384, Gln386, Glu388, Asn389, Asn390, Y391F, Lys392, Leu398, Ser400, Asp401, Asp413, Arg416, Gln418, Gln419, Asn421, Val422, Thr437, Gln438, Lys439, Ser440, Ser442, Ser444, and Lys447.1-b Values are the ratio of binding of the variant to that of native IgG1 at 0.33 or 1 μg/ml. A value greater than 1 denotes binding of the variant was improved compared with native IgG1, whereas a ratio less than 1 denotes reduced binding compared with native IgG1. Reduced binding to any given receptor was defined as a reduction of ≥40%compared to native IgG; better binding was defined as an improvement of ≥25%compared with native IgG1.1-c Number of independent assays for FcγRIIA, FcγRIIB and FcγRIIIA. At least two separately expressed and purified lots of each variant were assayed. Open table in a new tab For the high affinity FcγRI, the receptor fusion protein at 1.5 μg/ml in PBS, pH 7.4, was coated onto ELISA plates (Nunc) for 18 h at 4 °C. Plates were blocked with assay buffer at 25 °C for 1 h. Serial 3-fold dilutions of monomeric E27 and variants (10.0–0.0045 μg/ml) were added to plates and incubated for 2 h. After washing plates with assay buffer, IgG bound to FcγRI was detected with peroxidase-conjugated F(ab′)2 fragment of goat anti-human F(ab′)2-specific IgG (Jackson ImmunoResearch) or with peroxidase-conjugated protein G (Bio-Rad). The substrate used was o-phenylenediamine dihydrochloride (Sigma). Absorbance at 490 nm was read using aVmax plate reader (Molecular Devices). For all FcγR, binding values reported are the binding of each E27 variant relative to native E27, taken as (A490 nm(variant)/A490 nm(native IgG1)) at 0.33 or 1 μg/ml for FcγRII and FcγRIIIA and 2 μg/ml for FcγRI. A value greater than 1 denotes binding of the variant was improved compared with native IgG1, whereas a ratio less than 1 denotes reduced binding compared with native IgG1. Reduced binding to any given receptor was defined as a reduction of ≥40%compared with native IgG; better binding was defined as an improvement of ≥25%compared with native IgG1. The latter was chosen based on the observation that variants with ≥25%improved binding in the ELISA format assay, such as E333A, K334A, and S298A, also showed improved efficacy in the cell-based binding and ADCC assays. ELISA plates (Nunc) were coated with 2 μg/ml NeutrAvidin (Pierce) or streptavidin (Zymed Laboratories Inc., South San Francisco, CA) in 50 mmcarbonate buffer, pH 9.6, at 4 °C overnight (the same results were obtained with either molecule). Plates were blocked with PBS, 0.5%BSA, 10 ppm Proclin 300 (Supelco, Bellefonte, PA), pH 7.2, at 25 °C for 1 h. FcRn-Gly-His6-GST was biotinylated using a standard protocol with biotin-X-NHS (Research Organics, Cleveland, OH) and bound to NeutrAvidin-coated plates at 2 μg/ml in PBS, 0.5%BSA, 0.05%polysorbate-20 (sample buffer), pH 7.2, at 25 °C for 1 h. Plates were then rinsed with sample buffer, pH 6.0. Eight serial 2-fold dilutions of E27 standard or variants (1.6–200 ng/ml) in sample buffer at pH 6.0 were incubated for 2 h. Plates were rinsed with sample buffer, pH 6.0, and bound IgG was detected with peroxidase-conjugated goat F(ab′)2 anti-human IgG F(ab′)2 (Jackson ImmunoResearch) in pH 6.0 sample buffer using 3,3′,5,5′-tetramethylbenzidine (Kirkegaard & Perry Laboratories, Gaithersburg, MD) as substrate. Absorbance at 450 nm was read on aVmax plate reader (Molecular Devices). Titration curves were analyzed by a four-parameter nonlinear regression fit (KaleidaGraph, Synergy Software, Reading, PA). For each ELISA plate assay, a full titration curve of E27 standard was done. The absorbance at the midpoint of the titration curve (mid-OD) and its corresponding E27 concentration were determined. Then the concentration of each variant at this mid-OD was determined, and the concentration of E27 was divided by the concentration of each variant. Hence, the values are a ratio of the binding of each variant relative to native IgG1 standard. A control human IgG1 was run on each ELISA plate as a control and had a ratio of 1.12 ± 0.07 (n = 92). A second format was