REGISTER  |  SIGN IN
United States Patent5770196
StudnickaJune 23, 1998
Title
Modified antibody variable domains and therapeutic uses thereof
Abstract
Methods are described for identifying the amino acid residues of an antibody variable domain which may be modified without diminishing the native affinity of the domain for antigen while reducing its immunogenicity with respect to a heterologous species and for preparing so modified antibody variable domains which are useful for administration to heterologous species. Antibody variable regions prepared by the methods of the invention are also described.
Inventors:Studnicka; Gary M. (Santa Monica, CA)
Assignee:XOMA Corporation (Berkeley, CA)
Appl. No.:472788
Filed:June 7, 1995
Current U.S. Class:424/133.1 424/134.1 424/135.1 424/143.1 424/144.1 424/152.1 424/153.1 424/154.1 424/178.1 424/181.1 424/183.1 514/825 514/866 530/387.3 530/388.1 530/388.22 530/388.7 530/391.3 530/391.7 
Field of Search:530/387.3 424/133.17
U.S. Patent Documents
4391904July 1983Litman et al.
4816397March 1989Boss et al.
4816567March 1989Cabilly et al.
4853871August 1989Patoliano et al.
4888415December 1989Lambert et al.
4925673May 1990Steiner et al.
4946778August 1990Ladner et al.
5101025March 1992Piatak, Jr. et al.
5225539July 1993Winter
5585089December 1996Queen et al.
5614192March 1997Vandenbark
Foreign Patent Documents
0 120 694 A2Oct., 1984EP
0 125 023 A1Nov., 1984EP
0 125 023 B1Jun., 1991EP
0 173 494 A2Mar., 1986EP
0 239 400 A2Sep., 1987EP
0 440 351 A2Aug., 1991EP
0 451 216 B1Jan., 1996EP
0 519 596 A1Dec., 1992EP
0 592 106 A1Apr., 1994EP
2177096May., 1989GB
2188638Oct., 1987GB
WO 86/01533Mar., 1986WO
WO 89/00999Feb., 1989WO
WO 89/01783Mar., 1989WO
WO 89/09622Oct., 1989WO
WO 90/07861Jul., 1990WO
WO 92/04380Mar., 1992WO
WO 92/07075Apr., 1992WO
WO 92/15327Sep., 1992WO
WO 92/22324Dec., 1992WO
WO 92/22653Dec., 1992WO
WO 93/05168Mar., 1993WO
Other References
Ahmed, BioEssays, 6(4): 175-177 "Structure and Function of Chimaeric Antibodies". .
Alegre et al., J. Immunol., 148(11): 3461-3468 (1992) "Effect of a Single Amino Acid Mutation on the Activating and Immunosuppressive Properties of a Humanized OKT3 Monoclonal Antibody". .
Antin, et al., Blood, 78(8): 2139-2149 (1991) "Selective Depletion of Bone Marrow T Lymphocytes With Anti-CD5 Monoclonal Antibodies: Effective Prophylaxis for Graft-Versus-Host Disease in Patients With Hematologic Malignancies". .
Barry, Dermatological Formulations p. 181 (1983) "Percutaneous Absorption". .
Better, et al., Science, 240: 1041-1043 (1988) "Escherichia coli Secretion of an Active Chimeric Antibody Fragment". .
Better et al., J. Biol. Chem., 267(23): 16712-16718 (1992) "Activity of Recombinant Mitogillin and Mitogillin Immunoconjugates". .
Better et al., Proc Natl. Acad. Sci. USA, 90: 457-461 (1993) "Potent Anti-CD5 Ricin A Chain Immunoconjugates from Bacterially Produced Fab' and F(ab')2". .
Bird et al., Science, 242: 423-426 (1988) "Single-Chain Antigen-Binding Proteins". .
Bolt et al., Eur. J. Immunol, 23(2): 403-411 (1993) "The Generation of a Humanized, Non-Mitogenic CD3 Monoclonal Antibody Which Retains In Vitro Immunosuppressive Properties". .
Borrebaeck et al., Bio/Technology, "Kinetic Analysis of Recombinant Antibody-Antigen Interactions; Relation Between Structural Domains and Antigen Binding" 10(6):697-698 (1992). .
Boulianne et al., Nature, 312: 643-646 (1984) "Production of Functional Chimeric Mouse/Human Antibody". .
Brady et al., J. Mol. Biol., 227: 253-264 (1992) "Crystal Structure of a Chimeric Fab' Fragment of an Antibody Binding Tumour Cells". .
Brown et al., Proc. Natl. Acad. Sci. USA, 88: 2663-2667 (1991) "Anti-Tac-H, a Humanized Antibody To The Interleukin 2 Receptor, Prolongs Primate Cardiac Allograft Survival". .
Bruggemann et al., J. Exp. Med., 170: 2153-2157 (1989) "The Immunogenicity of Chimeric Antibodies". .
Buchner et al., Biotechnology, 9: 157-162 (1991) "Renaturation, Purification and Characterization of Recombinant F.sub.ab -Fragments Produced in Escherichia Coli". .
Byers et al., Blood, 75(7): 1426-1432 (1990) "Use of An Anti-Pan T-Lymphocyte Ricin A Chain Immunotoxin in Steroid-Resistant Acute Graft-Versus-Host Disease". .
Caron et al., Cancer Res., 52(24): 6761-6767 (1992) "Biological and Immunological Features of Humanized M195 (Anti-CD33) Monoclonal Antibodies". .
Carter et al., Proc. Natl. Acad. Sci. USA, 89: 4285-4289 (1992) "Humanization of Anti-p185.sup.HER2 Antibody for Human Cancer Therapy". .
Case et al., Proc. Natl. Acad. Sci. USA, 86: 287-291 (1989) "Chimeric Cytotoxin IL2-PE40 Delays and Mitigates Adjuvant-Induced Arthritis in Rats". .
Cheadle et al., Mol. Immunol., 29: 21-30 (1992) "Cloning and Expression of the Variable Regions of Mouse Myeloma Protein MOPC315 in E. Coli: Recovery of Active F.sub.v Fragments". .
Chothia et al., The EMBO Journal, 7(12): 3745-3755 (1988) "The Outline Structure of the T-Cell .alpha..beta. Receptor". .
Chothia et al., J. Mol. Biol., 196: 901-917 (1987) "Canonical Structures for the Hypervariable Regions of Immunoglobulins". .
Chothia et al., Nature, 342: 877-883 (1989) "Conformations of Immunoglobulin Hypervariable Regions". .
Chothia et al., J. Mol. Biol., 186: 651-663 (1985) "Domain Association in Immunoglobulin Molecules, The Packing of Variable Domains". .
Chothia and Lesk, Cold Spring Harbor Symp. Quant. Biol., 52: 399-405 (1987) "The Evolution of Protein Structures". .
Choy et al., Scandinavian J. Immunol., 36: 291-298 (1992) "Treatment of Rheumatoid Arthritis With Single Dose or Weekly Pulses of Chimaeric Anti-CD4 Monoclonal Antibody". .
Co et al., Nature, 351: 501-502 (1991) "Humanized Antibodies for Therapy". .
Co et al., Proc. Natl. Acad. Sci. USA, 88: 2869-2873 (1991) "Humanized Antibodies for Antiviral Therapy". .
Co et al., J. Immunol., 148: 1149-1154 (1992) "Chimeric and Humanized Antibodies with Specificity for the CD33 Antigen". .
Daugherty, et al., Nucleic Acids Res., 19: 2471-2476 (1991) "Polymerase Chain Reaction Facilitates the Cloning, CDR-grafting, and Rapid Expression of a Murine Monoclonal Antibody Directed Against the CD18 Component of Leukocyte Integrins". .
Davies & Metzger, Ann. Rev. Inmunol., 1: 87-117 (1983) "Structural Basis of Antibody Function". .
Davies and Padlan, Ann. Rev. Biochem., 59: 439-473 (1990) "Antibody-Antigen Complexes". .
Derocq et al., Transplantation, 44(6): 763-769 (1987) "Rationale for the Selection of Ricin A-Chain Anti-T Immunotoxins for Mature T Cell Depletion". .
Eigenbrot et al., J. Mol. Biol., 229: 969-995 (1993) "X-ray Structures of the Antigen-binding Domains from Three Variants of Humanized Anti-p185.sup.HER3 Antibody 4D5 and Comparison with Molecular Modeling". .
Ey et al., Immunochem., 15: 429-436 (1978) "Isolation of pure IgG.sub.1, IgG.sub.2a and IgG.sub.2b Immunoglobulins from Mouse Serum Using Protein A-Sepharose". .
Fishwild et al., Clin. Exp. Immunol., 86: 506-513 (1991) "Cytotoxicity Against Human Peripheral Blood Mononuclear Cells and T Cell Lines Mediated By Anti-T Cell Immunotoxins in the Absence of Added Potentiator". .
Fishwild et al., Clin. Exp. Immunol., 97: 10-18 (1994) "Characterization of the Increased Cytotoxicity of Gelonin Anti-T Cell Immunoconjugates Compared with Ricin A Chain Immunoconjugates". .
Foote et al., J. Mol. Biol., 224: 487-499 (1992) "Antibody Framework Residues Affecting the Conformation of Hypervariable Loops". .
Galfre et al., Nature, 266: 550-552 (1977) "Antibodies to Major Histocompatibility Antigens Produced by Hybrid Cell Lines". .
Glaser et al., J. Immunol., 149(8): 2607-2614 (1992) "Dissection of the Combining Site in a Humanized Anti-Tac Antibody". .
Glockshuber et al., Biochemistry, 29: 1362-1367 (1990) "A Comparison of Strategies to Stabilize Immunoglobulin F.sub.V -Fragments". .
Goff et al., Bioconjugate Chem., 1: 381-386 (1990) "Substituted 2-Iminothiolanes: Reagents for the Preparation of Disulfide Cross-Linked Conjugates With Increased Stability". .
Goldberg et al., J. Autoimmunity, 4: 617-630 (1991) "Immunological Effects of High Dose Administration of Anti-CD4 Antibody In Rheumatoid Arthritis Patients". .
Goldberg et al., Arthritis and Rheumatism, 33: S153, Abstract D115 (1990) "Preliminary Trial of an Anti-CD4 Monoclonal Antibody (MoAb) in Rheumatoid Arthritis (RA)". .
Gorman et al., Proc. Natl. Acad. Sci. USA, 88: 4181-4185 (1991) "Reshaping a Therapeutic CD4 Antibody". .
Hafler et al., Neurology, 36: 777-784 (1986) "Immunologic Responses of Progressive Multiple Sclerosis Patients Treated With An Anti-T-Cell Monoclonal Antibody, Anti-T12". .
Hakimi et al., J. Immunol., 147: 1352-1359 (1991) "Reduced Immunogenicity and Improved Pharmacokinetics Of Humanized Anti-Tac in Cynomolgus Monkeys". .
Hakimi et al., J. Immunol., 151: 1075 (1993) "Humanized Mik.beta.1, A Humanized Antibody to the IL-2 Receptor .beta.-chain That Acts Synergistically With Humanized Anti-TAC". .
Hale et al., The Lancet, 11: 1394-1399 (1988) "Remission Induction in Non-Hodgkin Lymphoma With Reshaped Human Monoclonal Antibody Campath-1H". .
Hara et al., Clinical Immunology and Immunopathology, 49: 223-230 (1988) "Stimulatory Effect of CD5 Antibody on B Cells from Patients With Rheumatoid Arthritis". .
Harlow et al., Eds., Antibodies: A Laboratory Manual, Chapter 14, "Immunoassays", Cold Spring Harbor Laboratory, Cold Spring Harbor (1988) pp. 553-612. .
Hodgson, Biotechnology, 8: 1245-1247 (1990) "Protein Design: Rules, Empiricism, & Nature". .
Horneff et al., Arthritis and Rheumatism, 34(2): 129-140 (1991) "Treatment of Rheumatoid Arthritis With An Anti-CD4 Monoclonal Antibody". .
Hsiao et al., Antibody Engineering Meeting, Dec. 14-16, 1992, Abstract "Humanization of Anti-CD18 mAb 60.3". .
Huse et al., Science, 246: 1275-1281 (1989) "Generation of a Large Combinatorial Library of the Immunoglobulin Repertoire in Phage Lambda". .
Huston et al., Proc. Natl. Acad. Sci. USA, 85: 5879-5883 (1988) "Protein Engineering of Antibody Binding Sites: Recovery of Specific Activity in an Anti-Digoxin Single-Chain Fv Analogue Produced in Escherichia coli". .
Janin et al., J. Biol. Chem., 265: 16027-16030 (1990) "The Structure of Protein-Protein Recognition Sites". .
Jones et al., Biotechnology, 9: 88-89 (1991) "Rapid PCR Cloning of Full-length Mouse Immunoglobulin Variable Regions". .
Jones et al., Nature, 321: 522-525 (1986) "Replacing the Complementarity-Determining Regions in a Human Antibody with those from a Mouse". .
Junghans et al., Cancer Res., 50 (5): 1495-1502 (1990) "Anti-Tac-H, A Humanized Antibody to the Interleukin 2 Receptor with New Features for Immunotherapy in Malignant and Immune Disorders". .
Kabat et al., J. Biol. Chem., 252 (19): 6609-6616 (1977) "Unusual Distributions of Amino Acids in Complementarity-Determining (Hypervariable) Segments of Heavy and Light Chains of Immunoglobulins and their Possible Roles in Specificity of Antibody-Combining Sites". .
Kelley et al., Biochem., 32: 6828-6835 (1993) "Thermodynamic Analysis of an Antibody Functional Epitope". .
Kelley et al., Biochem., 31: 5434-5441 (1992) "Antigen Binding Thermodynamics and Antiproliferative Effects of Chimeric and Humanized Anti-p185.sup.HER2 Antibody Fab Fragments". .
Kernan et al., J. Immnunol., 133 (1): 137-146 (1984) Specific Inhibition of in vitro Lymphocyte Transformation by an Anti-Pan T Cell (gp67) Ricin A Chain Immunotoxin. .
Kettleborough et al., Protein Engineering, 4: 773-783 (1991) "Humanization of a Mouse Monoclonal Antibody by CDR-Grafting: The Importance of Framework Residues on Loop Conformation". .
Kirkham et al., Brit. J. Rheumatology, 30: 88 Abstract 16 (1991) "Chimaeric (Human/Mouse) CD7 Monoclonal Antibody Treatment in Rheumatoid Arthritis". .
Kirkham et al., Brit. J. Rheumatology, 30: 459-463 (1991) "Monoclonal Antibody Treatment in Rheumatoid Arthritis: The Clinical and Immunological Effects of a CD7 Monoclonal Antibody". .
Kirkham et al., J. Rheumatology, 19: 1348-1352 (1992) "Chimeric CD7 Monoclonal Antibody Therapy in Rheumatoid Arthritis". .
Knowles, Chapter 22 in Reinherz et al., Leukocyte Typing II, 1: 259-288 (Springer-Verlag, 1986) "Immunochemical Analysis of the T-Cell Specific Antigens". .
Koda et al., Hum. Antibody Hybridomas 1(1): 15-22 (1990) Review "In Vitro Immunization for the Production of Human Monoclonal Antibody". .
Kohler et al., Eur. J. Immunol., 6: 292-295 (1976) "Fusion Between Immunoglobulin-Secreting and Nonsecreting Myeloma Cell Lines". .
Kyle et al., J. Rheumatol., 18: 1737-1738 "Humanized Monoclonal Antibody Treatment in Rheumatoid Arthritis". .
Lambert et al., J. Biol. Chem., 246: 12035-12041 (1985) "Purified Immunotoxins that are Reactive with Human Lymphoid Cells". .
Laurent et al., Bone Marrow Transplantation. 4: 367-371 (1989) "Donor Bone Marrow Treatment With T101 Fab Fragment-Ricin A-Chain Immunotoxin Prevents Graft-Versus-Host Disease". .
Lazarovits et al., J. Immunol., 150(11): 5163-5174 (1993) "Human Mouse Chimeric CD7 Monoclonal Antibody (SDZCHH380) for the Prophylaxis of Kidney Transplant Rejection". .
Lesk et al., Nature, 335: 188-190 (1988) "Elbow Motion in the Immunoglobulins Involves a Molecular Ball-and-socket Joint". .
Liu et al., Gene, 54: 33-40 (1987) "Expression of Mouse: Human Immunoglobulin Heavy-Chain cDNA in Lymphoid Cells". .
LoBuglio et al., Proc. Natl. Acad. Sci. USA, 86: 4220-4224 (1989) "Mouse/Human Chimeric Monoclonal Antibody in Man: Kinetics and Immune Response". .
Marks et al., J. Mol. Biol., 222: 581-597 (1991) "By-Passing Immunization: Human Antibodies from V-Gene Libraries Displayed on Phage". .
Marks et al., J. Biol. Chem., 267(23): 16007-16010 (1992) "Molecular Evolution of Proteins on Filamentous Phage". .
Martin et al., Proc. Natl. Acad. Sci. USA, 86: 9268-9272 (1989) "Modeling Antibody Hypervariable Loops: A Combined Algorithm". .
Mathieson et al., New England J. Med., 323 (4): 250-254 (1990) "Monoclonal-Antibody Therapy in Systemic Vasculitis". .
McCafferty et al., Nature, 348: 552-554 (1990) "Phage Antibodies: Fitamentous Phage Displaying Antibody Variable Domains". .
Miglietta, et al., Antibody Engineering Meeting, Dec. 14-16, 1992 Abstract "Alteration of Framework Residues Modulate Binding of a CDR-Grafted Anti-Human ICAM-1". .
Morrison, Science, 229: 1202-1207 (1985) "Transfectomas Provide Novel Chimeric Antibodies". .
Morrison et al., Adv. in Imunol., 44: 65-92 (1989) "Genetically Engineered Antibody Molecules". .
Morrison et al., Proc. Natl. Acad. Sci., 81: 6851-6855 (1984) "Chimeric Human Antibody Molecules: Mouse Antigen-Binding Domains With Human Constant Region Domains". .
Munson et al., Anal. Biochem., 107: 220-239 (1980) "LIGAND: A Versatile Computerized Approach for Characterization of Ligand-Binding Systems". .
Near et al., J. Immunol., 146 (2): 627-633 (1991) "The Specificity Properties that Distinguish Members of a Set of Homologous Anti-Digoxin Antibodies are Controlled by H Chain Mutations". .
Nishimura et al., Eur. J. Immunol., 18: 747-753 (1988) "Expression and Function of a CD5 cDNA in Human and Murine T Cells". .
Nisonoff et al., Archives of Biochem., 93 460-462 (1961) "Recombination of a Mixture of Univalent Antibody Fragments of Different Specificity". .
Novotny et al., Proc. Natl. Acad. Sci. USA, 82: 4592-4596 (1985) "Structural Invariants of Antigen Binding: Comparison of Immunoglobulin V.sub.L -V.sub.H and V.sub.L -V.sub.L Domain Dimers". .
Novotny et al., J. Mol. Biol., 189: 715-721 (1986) "Location of Antigenic Epitopes on Antibody Molecules". .
Novotny et al., Proc. Natl. Acad. Sci. USA, 83: 226-230 (1986) "Antigenic Determinants in Proteins Coincide with Surface Regions Accessible to Large Probes (Antibody Domains)". .
Novotny et al., Proc. Natl. Acad. Sci. USA, 83: 742-746 (1986) "Secondary, Tertiary, and Quaternary structure of T-cell-specific Immunoglobulin-like Polypeptide Chains". .
Padlan et al., Proc. Natl. Acad. Sci. USA, 86: 5938-5942 (1989) "Structure of an Antibody-Antigen Complex: Crystal Structure of the HyHEL-10 Fab-Lysozyme Complex". .
Padlan, E.A. Molecular Immunology, 28(4/5): 489-498 (1991) "A Possible Procedure For Reducing The Immunogenicity of Antibody Variable Domains While Preserving Their Ligand-Binding Properties". .
Peacock et al., Arthritis and Rheumatology, 35 (Suppl.) Abstract No. B141 (1992) "An Angiogenesis Inhibitor in Combination with Anti-CD5 Mab Suppresses Established Collagen Induced Arthritis Significantly More Than Single Agent Therapy". .
Pluckthun, Biotechnology, 9: 545-551 (1991) "Antibody Engineering: Advances From the Use of Escherichia coli Expression Systems". .
Potter et al., Proc. Natl. Acad. Sci. USA, 81: 7161-7165 (1984) "Enhancer-dependent Expression of Human K Immunoglobulin Genes Introduced into Mouse pre-B Lymphocytes by Electroporation". .
Presta et al., J. Immunol, 151: 2623-2632 (1993) "Humanization of an Antibody Directed Against IgE". .
Queen et al., Proc. Natl. Acad. Sci. USA, 86: 10029-10033 (1989) "A Humanized Antibody that Binds to the Interleukin 2 Receptor". .
Racadot et al., Brit. J. Rheumatology, 30: 88 (1991) Abstract "Immunological Follow Up of 13 Patients With Rheumatoid Arthritis Treated by Anti-T CD4+ Monoclonal Antibodies". .
Riechmann et al., Nature, 332: 323-327 (1988) "Reshaping Human Antibodies for Therapy". .
Roberts et al., Nature, 328: 731-734 (1987) "Generation of an Antibody with Enhanced Affinity and Specificity for Its Antigen by Protein Engineering". .
Robinson et al., Hum. Antib. Hybridomas, 2: 84-93 (1991) "Chimeric Mouse-Human Anti-Carcinoma Antibodies that Mediate Different Anti-Tumor Cell Biological Activities". .
Rodwell, Nature, 342: 99-100 (1989) "Engineering Monoclonal Antibodies". .
Roguska et al., Proc. Natl. Acad. Sci. USA, 91: 969-973 (1994) "Humanization of Murine Monoclonal Antibodies Through Variable Domain Resurfacing". .
Rostaing-Capaillon et al., Cancer Immunol. Immunother., 34: 24-30 (1991) "In Vivo Cytotoxic Efficacy of Immunotoxins Prepared From Anti-CD5 Antibody Linked to Ricin A-Chain". .
Routledge et al., Eur. J. Immunol., 21: 2717-2725 (1991) "A Humanized Monovalent CD3 Antibody which can Activate Homologous Complement". .
Royston et al., J. Immunol. 125(2): 725-731 (1980) "Human T Cell Antigens Defined By Monoclonal Antibodies: The 65,000-Dalton Antigen of T Cells (T65) Is Also Found On Chronic Lymphocytic Leukemia Cells Bearing Surface Immunoglobulin". .
Schlom, J. Molecular Foundations of Oncology, Samuel Broder (Ed.), Williams and Wilkins, Baltimore, MD Chapter 6 pp. 95-134 "Monoclonal Antibodies: They're More And Less Than You Think". .
Sharon et al., Nature, 309: 364-367 (1984) "Expression of a V.sub.H C.sub.K Chimaeric Protein in Mouse Myeloma Cells". .
Shearman et al., Antibody Engineering Meeting, Dec. 10-11, 1990 Abstract "Humanized Antibodies with Specificity for the Human .alpha./.beta. T Cell Receptor". .
Shearman et al., J. Immunol., 147(12): 4366-4373 (1991) "Construction, Expression and Characterization of Humanized Antibodies Directed Against the Human .alpha./.beta. T Cell Receptor". .
Sims et al., J. Immunol., 151(4): 2296-2308 (1993) "A Humanized CD18 Antibody can Block Function Without Cell Destruction". .
Singer et al., J. Immunol., 150: 2844-2857 (1993) "Optimal Humanization of 1B4, An Anti-CD18 Murine Monoclonal Antibody, Is Achieved by Correct Choice of Human V-Region Framework Sequences". .
Skerra et al., Science, 1038-1041 (1988) "Assembly of a Functional Immunoglobulin F.sub.v Fragment in Escherichia coli". .
Strand et al., Arthritis and Rheumatism, 33 (9 Suppl.)(1990) p. S25 "Treatment of Rheumatoid Arthritis With An Anti-CD5 Immunoconjugate: Clinical And Immunologic Findings and Preliminary Results of ReTreatment". .
Studnicka, G.M., Biochem J., 252: 825-831 (1988) "Escherichia coli Promoter -10 and -35 Region Homologies Correlate with Binding and Isomerization Kinetics". .
Takeda et al., Nature, 314: 452-454 (1985) "Construction of Chimaeric Processed Immunoglobulin Genes Containing Mouse Variable and Human Constant Region Sequences". .
Tempest et al., Biotechnology, 9: 266-271 (1991) "Reshaping a Human Monoclonal Antibody to Inhibit Human Respiratory Syncytial Virus Infection in vivo". .
Thornton, Nature, 343: 411-412 (1990) "Tackling a Loopy Problem". .
Tramontano et al., J Mol. Biol., 215: 175-182 (1990) "Framework Residue 71 is a Major Determinant of the Position and Conformation of the Second Hypervariable Region in the V.sub.H Domains of Immunoglobulins". .
Tramontano et al., Proteins: Structure, Function and Genetics, 6: 382-394 (1989) "Structural Determinants of the Conformations of Medium-Sized Loops in Proteins". .
Verhoeyen et al., Science, 239: 1534-1536 (1988) "Reshaping Human Antibodies: Grafting an Antilysozyme Activity". .
Verhoeyen et al., BioEssays, 8(2): 74-78 (1988) "Engineering of Antibodies". .
Vitetta et al., Science, 238: 1098-1104 (1987) "Redesigning Nature's Poisons to Create Anti-Tumor Reagents". .
Ward et al., Nature, 341: 544-546 (1989) "Binding Activities of a Repertoire of Single Immunoglobulin Variable Domains Secreted from Escherichia coli". .
Winter et al., Nature, 349: 293-299 (1991) "Man-Made Antibodies". .
Wofsky et al., J. Imnun., 134(2): 852-857 (1985) "Treatment of Murine Lupus with Monoclonal Anti-T Cell Antibody". .
Woodle et al., J. Immunol., 148: 2756-2763 (1992) "Humanized OKT3 Antibodies: Successful Transfer of Immune Modulating Properties and Idiotype Expression". .
Wu et al., J. Exp. Med., 132: 211-250 (1970) "An Analysis of the Sequences of the Variable Regions of Bence Jones Proteins and Myeloma Light Chains and Their Implications for Antibody Complementarity". .
Wu and Kabat., Mol. Immunol., 29(9): 1141-1146 (1992) "Possible Use of Similar Framework Region Amino Acid Sequences Between Human and Mouse Immunoglobulins for Humanizing Mouse Antibodies"..~
Primary Examiner: Scheiner; Toni R.
Assistant Examiner: Lucas; John
Attorney, Agent or Firm:McAndrews, Held & Malloy, Ltd.
Parent Case Text


CROSS REFERENCE TO RELATED APPLICATIONS

This application is a continuation of U.S. application Ser. No. 08/082,842, filed Jun. 23, 1993, which is a continuation-in-part of PCT/US92/10906, filed Dec. 14, 1992, which is a continuation-in-part of U.S. application Ser. No. 07/808,464, filed Dec. 13, 1991 (abandoned).
Claims

We claim:
1. A method of depleting CD5.sup.+ cells in an animal comprising the step of administering, to said animal, an effective amount of a cytotoxic protein that comprises a modified immunoglobulin variable domain, wherein said protein is selected from the group consisting of an anti-CD5 immunoglobulin molecule, an immunoconjugate comprising said molecule, and a fusion protein comprising said molecule, and wherein said modified immunoglobulin variable domain comprises at least one member selected from the group consisting of:
(a) a modified light chain variable region that has the amino acid sequence of SEQ ID No. 27;
(b) a modified light chain variable region that has the amino acid sequence of SEQ ID No. 73;
(c) a modified heavy chain variable region that has the amino acid sequence of SEQ ID No. 29; and
(d) a modified heavy chain variable region that has the amino acid sequence of SEQ ID No. 74.

2. The method of claim 1, wherein said protein further comprises a J-segment fused in frame to said variable region.

3. The method of claim 1, wherein said protein further comprises one or more constant regions fused in frame to said variable region.

4. The method of claim 3, wherein at least one constant region is derived from a different source than the source from which said variable region was derived.

5. The method of claim 4, wherein said source from which said constant region is derived is human.

6. The method of claim 1, wherein said protein is said modified anti-CD5 immunoglobulin molecule.

7. The method of claim 1, wherein said protein is said immunoconjugate.

8. The method of claim 1, wherein said protein is said fusion protein.

9. The method of claim 1, wherein said anti-CD5 immunoglobulin molecule is a single chain antibody.

10. The method of claim 1, wherein said anti-CD5 immunoglobulin molecule is an Fab.

11. The method of claim 1, wherein said anti-CD5 immunoglobulin molecule is an Fab'.

12. The method of claim 1, wherein said anti-CD5 immunoglobulin molecule is an F(ab').sub.2.

13. The method of claim 1, wherein said anti-CD5 immunoglobulin molecule is that produced by the hybridoma having ATCC Accession No. HB 11206.

14. The method of claim 1, wherein said animal has an autoimmune disease.

15. The method of claim 14, wherein said disease is systemic lupus erythematosus.

16. The method of claim 14, wherein said disease is rheumatoid arthritis.

17. The method of claim 14, wherein said disease is psoriasis.

18. The method of claim 14, wherein said disease is type I diabetes.
Description
FIELD OF THE INVENTION

The present invention generally relates to modified antibody variable domains and fragments thereof. More particularly, the invention relates to mouse antibody variable domains which are modified for administration to humans. For purposes of the present application, such modified antibody variable domains are termed "humanized antibodies" or "human-engineered antibodies." As taught herein, the humanized antibodies, or fragments thereof, according to the invention are useful, either alone or in conjugated form, in the treatment of various human diseases. The present application also teaches methods, termed "human-engineering," for preparing humanized antibodies, conjugation of humanized antibodies to various toxins, and therapeutic uses of the humanized antibodies of the invention.

BACKGROUND OF THE INVENTION

Application of unmodified mouse monoclonal antibodies in the treatment of human diseases may be problematic for several reasons. First, an immune response against the mouse antibodies may be mounted in the human body. Second, the mouse antibodies may have a reduced half-life in the human circulatory system. Third, the mouse antibody effector domains may not efficiently trigger the human immune system.

Several reports relate to eliminating the foregoing problems. For example, Junghans et al., Cancer Res., 50:1495-1502 (1990), describe the utilization of genetic engineering techniques to link DNA encoding murine variable domains to DNA encoding human constant domains, creating constructs which, when expressed, generate a hybrid mouse/human chimeric antibody.

Also by genetic engineering techniques, the genetic information from murine hypervariable complementarity determining regions (hereinafter referred to as "CDRs") may be inserted in place of the DNA encoding the CDRs of a human monoclonal antibody to generate a construct encoding a human antibody with murine CDRs. This technique is known as "CDR grafting". See, e.g., Jones et al., Nature, 321, 522-525 (1986); Junghans et al., supra.

Protein structure analysis may be used to "add back" murine residues, again by genetic engineering, to first generation variable domains generated by CDR grafting in order to restore lost antigen binding capability. Queen et al., Proc. Natl. Acad. Sci. USA, 86, 10029-10033 (1989); Co, et al., Proc. Natl. Acad. Sci. USA, 88, 2869-2873 (1991) describe versions of this method. The foregoing methods represent techniques to "humanize" mouse monoclonal antibodies.

As a result of the humanization of mouse monoclonal antibodies, specific binding activity of the resulting humanized antibodies may be diminished or even completely abolished. For example, the binding affinity of the modified antibody described in Queen et al., supra, is reported to be reduced three-fold; in Co et al., supra, is reported to be reduced two-fold; and in Jones et al., supra, is reported to be reduced two- to three-fold. Other reports describe order-of-magnitude reductions in binding affinity. See, e.g., Tempest et al., Bio/Technology, 9:266-271 (1991); Verhoeyen et al., Science, 239:1534-1536 (1988).

Examples of therapeutic targets for antibody therapy in humans are T lymphocytes, or T cells. Various T cell-reactive antibodies have been described, primarily from murine hybridomas. The specific subsets of T cells recognized by these antibodies, and their cell surface targets, are differentiated by the Clusters of Differentiation System (hereinafter referred to as the "CD System"). The CD System represents standard nomenclature for molecular markers of leukocyte cell differentiation molecules. See Leukocyte Typing III White Cell Differentiation Antigens (Michael, ed. Oxford Press 1987), which is incorporated by reference herein.

So-called "pan T cell" markers (or antigens) are those markers which occur on T cells generally and are not specific to any particular T cell subset(s). Pan T cell markers include CD2, CD3, CD5, CD6, and CD7.

The CD5 cluster antigen, for example, is one of the pan T cell markers present on about 85-100% of the human mature T lymphocytes and a majority of human thymocytes. The CD5 marker is also present on a subset, about 20%, of B cells. Extensive studies using flow cytometry, immunoperoxidase staining, and red cell lysis have demonstrated that CD5 is not normally present on hematopoietic progenitor cells or on any other normal adult or fetal human tissue with the exception of the aforementioned subpopulation of B cells.

Further information regarding the CD5 marker is found in McMichael and Gotch, in Leukocyte Typing III White Cell Differentiation Antigens (Michael, ed. Oxford Press 1987). The CD5 molecule has also been described in the literature as reactive with immunoglobulins. See, e.g., Kernan et al., J. Immunol., 33:137-146 (1984), which is incorporated by reference herein.

There are reports of attempted treatment of rheumatoid arthritis patients with monoclonal antibodies against CD4. See Horneff, et al. Arthritis and Rheumatism 34:2, 129-140 (February 1991); Goldberg, et al., Arthritis and Rheumatism, Abstract D115, 33:S153 (September 1990); Goldberg, Journal of Autoimmunity, 4:617-630 (1991); Choy, et al. Scand. J. Immunol. 36:291-298 (1992).

There are reports of attempted treatment of autoimmune disease, particularly rheumatoid arthritis, with an anti-CD7 monoclonal antibody. See Kirkham, et al., British Journal of Rheumatology 30:459-463 (1991); Kirkham, et al., British Journal of Rheumatology 30:88 (1991); Kirkham, et al., Journal of Rheumatology 19:1348-1352 (1992). Lazarovits, et al., J. Immunology, 150:5163-5174 (1993), describe attempted treatment of kidney transplant rejection with a chimeric anti-CD7 antibody. There is also a report of an attempt to treat multiple sclerosis with an anti-T12 antibody and a pan T-cell antibody (anti CD-6). Hafler, et al., Neurology 36:777-784 (1986).

None of the above attempts for therapy of human autoimmune diseases involve the use of unconjugated anti-CD5 antibodies.

Thus, there exists a need for the successful antibody therapy of T cell-mediated diseases such as autoimmune disease, graft-versus-host disease, and transplant rejection. As demonstrated by the foregoing, there also exists a need in the art for methods for the preparation of humanized antibodies useful in the treatment of various human diseases and not subject to the foregoing drawbacks.

SUMMARY OF THE INVENTION

The present invention provides methods, termed human-engineering, for preparing a modified antibody variable domain useful for administration to humans by determining the amino acids of a subject antibody variable domain which may be modified without diminishing the native affinity of the domain for antigen, while reducing its immunogenicity with respect to a heterologous species. As used herein, the term "subject antibody variable domain" refers to the antibody upon which determinations are made. The method includes the following steps: determining the amino acid sequence of a subject light chain and a subject heavy chain of a subject antibody variable domain to be modified; aligning by homology the subject light and heavy chains with a plurality of human light and heavy chain amino acid sequences; identifying the amino acids in the subject light and heavy chain sequences which are least likely to diminish the native affinity of the subject variable domain for antigen while, at the same time, reducing its immunogenicity by selecting each amino acid which is not in an interface region of the subject antibody variable domain and which is not in a complementarity-determining region or in an antigen-binding region of the subject antibody variable domain, but which amino acid is in a position exposed to a solvent containing the antibody; changing each residue identified above which aligns with a highly or a moderately conserved residue in the plurality of human light and heavy chain amino acid sequences if said identified amino acid is different from the amino acid in the plurality.

Another group of sequences, such as those in FIGS. 1A and 1B may be used to determine an alignment from which the skilled artisan may determine appropriate changes to make.

The present invention provides a further method wherein the plurality of human light and heavy chain amino acid sequences is selected from the human consensus sequences in FIGS. 5A and 5B.

In general, human engineering according to the above methods may be used to generate antibodies useful in the treatment of various diseases against which monoclonal antibodies generally may be effective. However, humanized antibodies possess the additional advantage of reducing the immunogenic response in the treated patient in the same manner and potentially to a greater extent than observed for chimeric antibodies (see LoBuglio, et al, Proc Natl. Acad. Sci. USA, 86:4220-4224 (1989) and Bruggemann, et al., J. Exp. Med., 170:2153-2157 (1989).

The present invention also discloses products and pharmaceutical compositions useful in the treatment of myriad human diseases which may be targeted by an antibody. In particular, products prepared by the foregoing methods include a modified H65
mouse monoclonal variable domain. Additionally, DNA sequences encoding the modified H65 variable domain are provided.

Modified antibody variable domains which are products of the methods of the present invention may be used, inter alia, as components of various immunoglobulin molecules such as Fab, Fab', and F(ab').sub.2 domains, single chain antibodies, and Fv or single variable domains.

The present invention provides novel proteins comprising a human-engineered antibody variable domain which are specifically reactive with a human CD5 cell differentiation marker. Preferred human-engineered anti-CD5 antibodies according to the present invention may have a binding affinity for CD5 of less than 2.times.10.sup.-9 M. In a preferred embodiment, the present invention provides proteins comprising the he3 light and heavy chain variable regions as shown in SEQ ID NOS: 73 and 74, respectively. DNA encoding certain he3 proteins is shown in SEQ ID NOS: 75 and 76.

In a preferred embodiment of the present invention, the protein comprising a human-engineered antibody variable region is an intact he3 immunoglobulin deposited with the American Type Culture Collection, 12301 Parklawn Drive, Rockville, Md.
20852 as ATCC Accession No. HB 11206.

Also in a preferred embodiment of the invention, the protein comprising a human-engineered antibody variable region is a Fab, F(ab').sub.2 fragment, or a single-chain antibody.

Proteins according to the present invention may be made by methods taught herein and in co-pending, co-owned U.S. patent application Ser. No. 07/808,464 by Studnicka, incorporated by reference herein. Modified antibody variable domains made by such methods may be used in therapeutic administration to humans either alone or as part of an immunoconjugate or immunofusion as taught in co-owned, co-pending U.S. patent application Ser. No. 07/787,567 filed Nov. 4, 1991 by Bernhard, et al. and co-owned, co-pending U.S. patent application Ser. No. 08/064,691, filed May 19, 1993 by Better, et al. (Attorney Docket No. 27129/31394). Proteins according to the present invention may also be applied to determine T cell levels in order to aid in the diagnosis of human autoimmune disease states. Proteins according to the present invention are useful in the treatment of human diseases and particularly useful in the treatment of autoimmune diseases. Additionally, other T cell-mediated diseases such as graft-versus-host disease or tissue transplant rejection may be treated with proteins according to the invention.

In a therapeutic treatment or diagnostic regimen, the whole protein may be used, a fragment of the protein, such as a Fab or F(ab').sub.2 region may be used, or a single-chain antibody may be used. Alternatively, an immunoconjugate or an immunofusion comprising the protein or fragment may be used. A fragment or single chain form of the presently-claimed antibodies are especially useful in applications in which no constant region is required.

The present invention also provides methods for treatment of autoimmune diseases, wherein animal models are predictive of the efficacy of treatment in humans. Finally, the present invention includes pharmaceutical compositions containing the humanized antibodies according to the invention.

Proteins, specifically he3 antibodies, according to the present invention are all useful in diagnostic procedures, wherein it is desirable to detect, identify, or isolate CD5 antigens. Such antibodies may be labelled for diagnostic identification of CD5 antigen.

Additional aspects and applications of the present invention will become apparent to the skilled artisan upon consideration of the detailed description of the invention which follows.

BRIEF DESCRIPTION OF THE DRAWING

FIGS. 1A and 1B are alignments of the amino acid sequences of the light and heavy chains, respectively, of two murine antibody variable domains [HYH (HYHEL-10 Fab-lysozyme complex), MCPC (IgA Fab MCPC603-phosphocholine complex), and two human antibody variable domains NEWM (Ig Fab' NEW) and KOL (IgG1 KOL)] by criteria of sequence and structural homology;

FIG. 2 is a schematic depiction of the structural relationships between the amino acid residues of the light chain of the variable domain;

FIG. 3 is a schematic depiction of the structural relationships between the amino acid residues of the heavy chain of the variable domain;

FIG. 4 is a schematic representation of an antibody variable domain;

FIGS. 5A and 5B are alignments of the consensus amino acid sequences for light (FIG. 5A) the subgroups of light chains [hK1 (human kappa light chain subgroup 1), hK3 (human kappa light chain subgroup 3), hK2 (human kappa light chain subgroup 2), hL1 (human lambda light chain subgroup 1), hL2 (human lambda light chain subgroup 2), hL3 (human lambda light chain subgroup 3), hL6 (human lambda light chain subgroup 6), hK4 (human kappa light chain subgroup 4), hL4 (human lambda light chain subgroup
4) and hL5 (human lambda light chain subgroup 5] and heavy chains (FIG. 5B) [hH3 (human heavy chain subgroup 3), hH1 (human heavy chain subgroup 1) and hH2 (human heavy chain subgroup 2)], respectively, of human antibody variable domains;

FIGS. 6A and 6B are alignments of human light (FIG. 6A) chain consensus sequence hK1 with the actual (h65) and low-risk modified (prop) light chain sequences of the H65 mouse monoclonal antibody variable domain and of human heavy (FIG. 6B) chain consensus sequence hH3 with the actual (h65) and modified (prop) heavy chain sequences of the H65 mouse monoclonal antibody variable domain, respectively;

FIGS. 7A and 7B are listings of the nucleotide sequences of the oligonucleotides utilized in the construction of the genes encoding modified V/J-regions of the light (FIG. 7A) and heavy (FIG. 7B) chains of the H65 mouse monoclonal antibody variable domain;

FIGS. 8A and 8B are listings of the nucleotide sequences of the genes encoding modified V/J-regions of the heavy (FIG. 8B) and light (FIG. 8A) chains, respectively, of the H65 mouse monoclonal antibody variable domain;

FIG. 9 is a graph of the results of a competitive binding assay showing that the H65 antibody variable domain modified by a method according to the present invention retains the antigen-binding capability of the natural H65 antibody variable region;

FIGS. 10A and 10B are alignments of human light (FIG. 10A) chain consensus hK1 and heavy (FIG. 10B) chain consensus hH1 with the light and heavy chain sequences, respectively, of the variable domain of human antibody EU, unmodified murine antibody TAC, murine antibody TAC modified according to the present invention (prop) and murine antibody TAC modified according to a different method (Que);

FIG. 11 is a graph of he3 IgG and he3 Fab binding to CD5 found on Molt-4M cells, demonstrating that such binding is improved over that of cH65 IgG and cH65 Fab;

FIG. 12 is a graph showing the effects of anti-Lyt-1 (murine anti-CD5) administration on the severity of collagen-induced arthritis in DBA/1J mice;

FIGS. 13A and 13B are schematic depictions of human T cell recovery in spleen and blood, respectively from PBMC/SCID mice following treatment with H65 monoclonal antibody (hereinafter referred to as "MoAb");

FIGS. 14A and 14B are schematic depictions of human T cell recovery in spleen and blood, respectively from PBMC/SCID mice following treatment with H65-based F(ab').sub.2 fragment;

FIG. 15 is a graph of the effects of OX19 MoAb on the severity of DR BB rat collagen-induced arthritis;

FIGS. 16A and 16B are alignments of human light chain consensus sequence hK1 with the actual (h65) and low and moderate risk modified (prop) light chain sequences of the H65 mouse monoclonal antibody variable domain and of human heavy chain consensus sequence hH3 with the actual (h65) and modified (prop) heavy chain sequences of the H65 mouse monoclonal antibody variable domain, respectively;

FIG. 17 is a graph showing results of competitive binding experiments using humanized single chain antibodies and he3 Fab to compete .sup.123 i-labeled cH65 IgG; open circles represent the pING3326 single chain antibody (V.sub.L -V.sub.H); open squares represent the pING3337 single chain antibody (V.sub.H -V.sub.L); and closed circles represent he3 Fab; and

FIG. 18 is a graph showing results of a competitive binding experiment using single chain antibodies and single chain antibody fusion proteins.

DETAILED DESCRIPTION OF THE INVENTION

Animal models of T cell-mediated autoimmune diseases were studied using therapeutic protocols with anti-T cell antibodies, especially anti-CD5 (Examples 1-3). Anti-CD5 antibodies were found to be particularly useful in several therapeutic regimens, as they were able to deplete the number of T cells in various lymphoid organs and also reduce the pathological effects of T cells. These studies provide an example of one therapeutic target (CD5) for the development of methods for the humanization of murine anti-T cell antibodies.

The present invention provides novel proteins and fragments comprising a humanized antibody variable region, and particularly an he3 variable region which is specifically reactive with a human CD5 cell differentiation marker. The invention also provides anti-CD5 antibodies with an affinity of less than about 2.times.10.sup.-9 M.

The terms "humanized," "human-like," or "human-engineered" refers to an immunoglobulin wherein the constant regions have at least about 80% or greater homology to human immunoglobulin, and wherein some of the nonhuman (i.e. murine) variable region amino acid residues may be modified to contain amino acid residues of human origin.

Humanized antibodies may be referred to as "reshaped" antibodies. Manipulation of the complementarity-determining regions (CDRs) is one means of manufacturing humanized antibodies. See, e.g., Jones, et al., Nature 321:522-525 (1988); Riechmann, et al., Nature 332:323-327 (1988). For a review article concerning chimeric and humanized antibodies, See Winter et al. Nature 349:293-299 (1991).

Construction of humanized antibody variable domains according to the present invention may be based on a method which includes the steps of: (1) identification of the amino acid residues of an antibody variable domain which may be modified without diminishing the native affinity of the domain for antigen while reducing its immunogenicity with respect to a heterologous species; (2) the preparation of antibody variable domains having modifications at the identified residues which are useful for administration to heterologous species; and (3) use of the humanized antibodies of the invention in the treatment of autoimmune diseases in humans. The methods of the invention are based on a model of the antibody variable domain described herein which predicts the involvement of each amino acid in the structure of the domain.

Unlike other methods for humanization of antibodies, which advocate replacement of the entire classical antibody framework regions with those from a human antibody, the methods described herein introduce human residues into the variable domain of an antibody only in positions which are not critical for antigen-binding activity and which are likely to be exposed to immunogenicity-stimulating factors. The present methods are designed to retain sufficient natural internal structure of the variable domain so that the antigen-binding capacity of the modified domain is not diminished in comparison to the natural domain.

Data obtained from the analysis of amino acid sequences of antibody variable domains using the MacImdad (Molecular Applications Group, Stanford, Calif.) three-dimensional molecular modeling program, in conjunction with data obtained from previous theoretical studies of hypervariable region structure and data obtained from the crystal structures of the HYH (HYHEL-10 Fab-lysosyme complex, Brookhaven structure "3HFM"), MCPC (IgA Fab MCPC603-phosphocholine complex, Brookhaven structure "2MCP"), NEWM (Ig Fab' NEW, Brookhaven structure "3FAB") and KOL (IgG1 KOL, Brookhaven structure "2IG2") antibody variable domains from the Brookhaven database (Brookhaven National Laboratory, Upton, N.Y.), are utilized to develop the antibody variable domain model.

FIGS. 1A and 1B provide the sequences of the four antibody variable domains which have been crystallized. The amino acid sequences of the light and heavy chains of HYH (SEQ ID NOS: 1 and 5, respectively), MCPC (SEQ ID NOS. 2 and 6, respectively), NEWM (SEQ ID NOS. 3 and 7, respectively) and KOL (SEQ ID NOS. 4 and 8, respectively) are shown, wherein the exclamation points "!" in the MCPC light chain sequence at position 30x, the MCPC heavy chain sequence at positions 52x and 98x, the NEWM light chain at position 30x, the KOL light chain at position 93x, and the KOL heavy chain sequence at position 98x, stand for the amino acid sequences NSGNQK (SEQ ID NO: 9), NKG (SEQ ID NO: 10), GST (SEQ ID NO: 11), AG, SL and HGFCSSASC (SEQ ID NO: 12), respectively which are variations in the length of hypervariable loop sequences among the various antibodies. The amino acid positions in FIGS. 1A and 1B, 2, and 3 are numbered according to Kabat et al., Sequences of Proteins of Immunological Interest, Fourth Edition, U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health (1987) (hereinafter referred to as "Kabat"), with the exception of those designated with a lower-case "x", which are variations in length of hypervariable loops which Kabat has numbered as "a,b,c,d . . . ".

FIGS. 2 and 3 comprise depictions of the structure of the light and heavy chains, respectively, wherein each chain is displayed "unfolded" into a flattened beta sheet structure so that interactions among the residues are easier to visualize. The strands of folded polypeptide chains are represented as thick vertical lines, connected by eight beta-turn loops. Three of the loops are identified as antigen-binding loops or CDRs, one is accessory to the loops, and the remaining four at the "bottom" of the variable domain are not involved in antigen binding. The amino and carboxy termini of the variable domain are symbolized by small black dots at the ends of the polypeptide chains. Each amino acid position is represented as either a circle, a triangle, or a square. The covalent disulfide bond between the two cysteines at positions 23 and 88 in the light chain and the covalent disulfide bond between positions 22 and 92 in the heavy chain are each shown as a thick horizontal line. All of the residues in each chain are shown on the map, including antigen-binding residues and framework residues. Solid slanted lines (either single or double) connecting pairs of residues which are adjacent in three-dimensional space but not in linear sequence, represent one or two hydrogen bonds between the mutually aligned amino nitrogens and carbonyl oxygens in the backbones of the residues.

The analysis of each amino acid position to determine whether the position influences antigen binding and/or is immunogenic was based upon the information in FIGS. 1A, 1B, 2 and 3, as well as the additional variable region structural information in the following paragraphs.

The basic structure of the antibody variable domain is strongly conserved. The variable domain is composed of a light chain (or subunit) and a heavy chain (or subunit), which are structurally homologous to each other and which are related by a pseudo-two-fold axis of rotational symmetry. At the "top" of the variable domain, the region farthest away from the constant domain, there are six antigen-binding loops which are built upon a larger structural framework region. The variable domain is functionally distinct from the constant domain, being connected only by two highly flexible chains and pivoting on both "ball-and-socket" joints formed by five amino acids in the heavy and light chains.

Each subunit, light or heavy, resembles a "sandwich" structure, composed of two layers of antiparallel beta sheets with a propeller twist in three-dimensional space. Each amino acid chain folds back on itself repeatedly to create nine distinct strands. Three-and-one-half of these strands form the "outside" beta-sheet layer of each subunit and the other five-and-one-half form the "inside" layer. The various strands in each layer are extensively hydrogen-bonded to each other. The two beta-sheet layers within the subunit are held together by a single covalent disulfide bond and by numerous internal hydrophobic interactions. The sequences involved in bonding the strands of the subunits together are called "framework" sequences.

Certain amino acids, either in antigen-binding sequences or in framework sequences, do not actually bind antigen but are critical for determining the spatial conformation of those residues which do bind. Each antigen-binding loop requires a properly formed "platform" of buried residues, which provides a surface upon which the loop folds. One or more of the loop residues often will be buried in the platform as an "anchor" which restricts the conformational entropy of the loop and which determines the precise orientation of antigen-contacting sidechains. Thus, the shapes of the residues which make up the platform contribute to the ultimate shape of the antigen-binding loop and its affinity for specific antigens.

Amino acid sidechains exist in various different chemical environments within the subunits. Some residues are exposed to the solvent on the outer accessible surface while other residues are buried in hydrophobic interactions within a subunit. Much of the immunoglobulin variable domain is constructed from antiparallel beta sheets which create amphipathic surfaces, such that the "inside" surface is hydrophobic and the "outside" surface is hydrophilic. The outside is exposed to solvent, and therefore is also exposed to the humoral environment when the domain is in the circulatory system of an animal. Amino acid sidechains which are completely exposed to the solvent and which do not physically interact with other residues in the variable domain are likely to be immunogenic and are unlikely to have any structural importance within the immunoglobulin molecule. A highly schematic representation of the variable domain is shown in FIG. 4, wherein thick lines represent peptide bonds and shaded circles denote amino acid sidechains.

The two subunits of antibody variable domains adhere to each other via a hydrophobic interface region which extends along the inside beta-sheet layer from the border of the variable domain with the constant domain to the antigen-binding loops. Amino acid sidechains from both subunits interact to form a three-layered "herringbone" structure. Some of these interfacial residues are components of the antigen-binding loops, and thus have a direct effect upon binding affinity. Every residue in the interface is structurally important because the conformation of the binding regions is strongly influenced by changes in the conformation of the interface.

The foregoing data and information on the structure of antibody variable domains aids in a determination of whether a particular amino acid of any variable domain is likely to influence antigen binding or immunogenicity. The determination for each amino acid position is represented by a pair of symbols (e.g., + and +, in the lines labelled "bind" and "bury", respectively) in FIGS. 1A, 1B, (and also in FIGS. 5A, 5B, 6A, 6B, 10A and 10B). In each of these pairs, the first symbol relates to antigen binding, while the second symbol relates to immunogenicity and framework structure. Tables 1, 2, and 3, below, set out the significance of the symbols and possible pairings.

TABLE 1 ______________________________________ First Symbol In Pair (Ligand Binding) ______________________________________ + Little or no direct influence on antigen- binding loops, low risk if substituted o Indirectly involved in antigen-binding loop structure, moderate risk if changed - Directly involved in antigen-binding loop conformation or antigen contact, great risk if modified ______________________________________

TABLE 2 ______________________________________ Second Symbol In Pair (Immunogenicity And Structure) ______________________________________ + Highly accessible to the solvent, high immunogenicity, low risk if substituted o Partially buried, moderate immunogenicity, moderate risk if altered - Completely buried in subunit's hydrophobic core, low immunogenicity, high risk if changed = Completely buried in the interface between subunits, low immunogenicity, high risk if modified. ______________________________________

TABLE 3 ______________________________________ Significance Of Pairs ______________________________________ ++ Low risk Highly accessible to the solvent and high immunogenicity, but little or no effect on specific antigen binding o+, +o, oo Moderate Risk Slight immunogenicity or indirect involvement with antigen binding any - or = High risk Buried within the subunit core/interface or strongly involved in antigen binding, but little immunogenic potential ______________________________________

The pairings set out in Tables 1-3 indicate that making mouse-to-human modifications at positions which have a pair of low risk symbols (++) (i.e., a symbol in the "bind" line and a symbol in the "bury" line corresponding to one position) results in a major reduction in therapeutic immunogenicity with little chance of affecting binding affinity. At the opposite end of the spectrum, modifying positions which have a pair of high risk symbols (--) may degrade or abolish binding activity with little or no actual reduction in therapeutic immunogenicity. There are 73 low risk positions in the variable domain (38 in the light chain and 35 in the heavy chain) which are indicated by circles in the lines labelled "risk" in FIGS. 1A, 1B, 5A, 5B, 6A, 6B,
10A and 10B. There are 29 moderate risk positions in the variable domain (12 in the light chain and 17 in the heavy chain) as indicated by the triangles in the lines labelled "risk" in FIGS. 1A, 1B, 5A, 5B, 6A, 6B, 10A, and 10B.

The results of the above analysis may be applied to consensus sequences for the different subgroups of antibody variable domains because the structural characteristics they represent are highly conserved, even among various species. FIGS. 5A and
5B thus set out and align the consensus sequences (derived from Kabat) of the subgroups of light (hK1, SEQ ID NO: 13; hK3, SEQ ID NO: 14; hK2, SEQ ID NO: 15; hL1 SEQ ID NO: 16; hL2, SEQ ID NO: 17; hL3, SEQ ID NO: 18; hL6, SEQ ID NO: 19; hK4, SEQ ID NO:
20; hL4, SEQ ID NO: 21; and hL5, SEQ ID NO: 22) and heavy chains (hH3, SEQ ID NO: 23; hH1, SEQ ID NO: 24; and hH2, SEQ ID NO: 25) of antibody variable domains with the pairings representing the structural characteristics of each amino acid position, wherein the consensus sequences for the hL6, hK4, hL4, hL5 and hH2 subgroups were derived from less than twenty actual light or heavy chain sequences.

In the consensus sequences set out in FIGS. 5A and 5B, upper case amino acid designations indicate that the amino acid is present at that location in about 90% to about 100% of the known human sequences (excluding small incomplete fragments) of that subgroup (i.e., is "highly conserved"); whereas lower case amino acid designations indicate that the amino acid is present at that location in about 50% to about 90% of the known human sequences in that subgroup (i.e., is "moderately conserved"). A lower case "x" denotes conservation in less than about 50% of the known sequences in that subgroup (i.e., a "poorly conserved" position).

The information presented in FIGS. 5A and 5B on the relationship of a particular amino acid in a sequence of an antibody variable domain to the structure and antigen-binding capacity of the domain is sufficient to determine whether an amino acid is modifiable. Additional structural studies, such as those on which FIGS. 5A and 5B are based, are not required.

Thus, according to the present invention, FIGS. 5A and 5B may be used to prepare, for example, a modified mouse antibody variable domain that retains the affinity of the natural domain for antigen while exhibiting reduced immunogenicity in humans by the following steps. The amino acid sequences of both the light chain and the heavy chain from the mouse variable domain are first determined by techniques known in the art (e.g., by Edman degradation or by sequencing of a cDNA encoding the variable domain). Next, the consensus sequences set out in FIGS. 5A and 5B for human antibody variable domains are examined to identify both a light chain consensus and a heavy chain consensus sequence that are the most homologous to the particular mouse subunit sequences that are to be modified. The mouse sequences are aligned to the consensus human sequences based on homology either by sight or by using a commercially available computer program such as the PCGENE package (Intelligenetics, Mountain View, Calif.).

FIGS. 5A and 5B are then used again to identify all of the "low risk" or "moderate risk" positions at which the mouse sequence differs significantly from the chosen human consensus. The mouse amino acid residues at these low risk and moderate risk positions are candidates for modification. If the human consensus is strongly conserved at a given low risk or moderate risk position, the human residue may be substituted for the corresponding mouse residue. If the human consensus is poorly conserved at a given low risk or moderate risk position, the mouse residue is retained at that position. If the human consensus is moderately conserved at a specific position, the mouse residue is normally replaced with a human residue, unless the mouse residue occurs at that position in at least one of the sequences (e.g., in Kabat) on which the human consensus sequence is based. If the mouse residue does occur at that position in a human sequence then the mouse residue may be retained.

Other criteria may be important to the determination of which identified residues of a variable region are to be modified. For example, since the side chain of proline is connected to both its .alpha.-carbon and its peptide nitrogen, free rotation is restricted around the carbon-nitrogen bond (the Ramachandran .phi. angle). Therefore, wherever there is a proline in a sequence, the shape of the backbone is distorted and that distortion can influence other residues involved in antigen binding. The presence or absence of a proline residue at any point in the amino acid sequence is a structurally important feature. If the mouse sequence contains a proline at a certain location, it is likely that its presence is necessary for a proper backbone and framework conformation and proline is preferably retained. If the mouse sequence does not contain a proline at a location where the human consensus sequence has one, it is likely that substituting a proline in the mouse sequence would affect proper conformation of the sequence, therefore the mouse residue is preferably retained. Where a proline at a particular position involving proline is changed from mouse to human, such a change is considered to be at least moderate risk even if that position would otherwise be low risk.

Similarly, insertions and deletions in a mouse sequence, relative to a human consensus framework, are normally preserved intact. If the mouse sequence has an alteration in the length and spacing of the variable region backbone, it is likely that the alteration is necessary to provide a surface for proper folding of the antigen-binding loops. The alteration is preferably retained in a modified version of the sequence.

Residues participating in the interface between the light and heavy chains of a variable domain are also preferably left intact in a modified version. They are all designated high risk, with = symbols on the "bury" lines in FIGS. 1, 5, 6, 10. The sidechains in the interface region are buried deep within the structure, so they are unlikely to elicit a therapeutic immunogenic response in a heterologous species.

Once a modified sequence has been designed, DNAs encoding the complete variable domain may be synthesized [via oligonucleotide synthesis as described, for example, in Sinha et al., Nucleic Acids Res., 21:4539-4557 (1984)], assembled [via PCR as described, for example in Innis, Ed., PCR Protocols, Academic Press (1990) and also in Better et al. J. Biol. Chem. 267:16712-16118 (1992)], cloned and expressed [via standard procedures as described, for example, in Ausubel et al., Eds., Current Protocols in Molecular Biology, John Wiley & Sons, New York (1989) and also in Robinson et al., Hum. Antibod. Hybridomas, 2:84-93 (1991)], and finally tested for specific antigen binding activity [via competition assay as described, for example, in Harlow et al., Eds., Antibodies: A Laboratory Manual, Chapter 14, Cold Spring Harbor Laboratory, Cold Spring Harbor (1988) and Munson et al., Anal. Biochem., 107:220-239 (1980)].

Humanized antibodies according to the present invention may be incorporated into an immunoconjugate for use in the treatment of various human diseases. For example, treatment of certain autoimmune diseases with immunotoxin conjugates is described in co-pending, co-owned U.S. patent application Ser. No. 07/759,297 filed Sep. 13, 1991, and U.S. patent application Ser. No. 07/988,430, filed Dec. 9, 1992, both of which are incorporated by reference herein. An immunoglobulin such as an anti-T-cell immunoglobulin may be conjugated to a cytotoxic molecule. The cytotoxic molecule to which the immunoglobulin is conjugated may be any of a number of toxins such as lectin A or a ricin A chain. The above-referenced '297 application also describes use of an anti-CD5 antibody conjugated to a ricin A chain providing an anti-T-cell immunotoxin. Humanized antibodies of the invention may also be used in immunofusions with, for example, gelonin toxin as taught in co-owned, co-pending U.S. patent application Ser. No. 08/064,691, filed May 12, 1993 (Attorney Docket No. 27129/31394).

Humanized antibodies according to the present invention include he3 and fragments thereof which display increased content of human amino acids and a high affinity for human CD5 cell differentiation marker. The he3 antibody is a humanized form of the mouse H65 antibody in which the moderate risk changes described below were made in both variable regions. Such humanized antibodies have less immunogenicity and have therapeutic utility in the treatment of autoimmune diseases in humans. For example, because of their increased affinity over existing therapeutic monoclonal antibodies such as H65, he3 antibodies of the invention may be administered in lower doses than H65 anti-CD5 antibodies in order to obtain the same therapeutic effect. The he3 variable regions are also useful in increasing potency over H65 anti-CD.sub.5 antibodies when used as a portion of an immunoconjugate or immunofusion protein.

The he3 proteins according to the present invention may also be used in the treatment of graft-versus-host disease. Laurent et al. Bone Marrow Transplantation, 4:367-371 (1989), incorporated by reference herein, reports that administration of a murine anti-CD5 Fab-RTA conjugate may greatly reduce the likelihood of graft-versus-host disease by causing an ex vivo purge of T cells from donor bone marrow prior to transplantation. See also, Antin et al., Blood, 78:2139-2149 (1991); Kernan et al., J. Am. Med. Assoc., 259:3154-3157 (1988), both incorporated by reference herein.

Alternatively, anti-CD5 antibodies and particularly human-engineered anti-CD5 antibodies of the present invention may be utilized in an unconjugated form for the therapy of autoimmune diseases. Such antibodies and uses are detailed below.

A general description of various autoimmune diseases is found in The Autoimmune Diseases (Rose & Mackey, eds 1985). Autoimmune diseases may be characterized, inter alia, by abnormal immunological regulation which results in excessive B cell activity and diminished, enhanced, or inappropriate T cell activity. Such altered T cell activity may result in excessive production of autoantibodies. Although the autoimmune diseases are complex and diverse in their manifestations, they possess the common feature of a malfunctioning immune system. Therapeutic depletion of circulating T cells through the administration of an anti-pan T cell immunoglobulin improves the clinical course of patients with autoimmune disease. For anti-CD5 antibody therapy, the additional depletion of CD5 B cells may have a further beneficial effect since CD5 B cells have been implicated in some autoimmune diseases.

An example of an anti-pan T cell immunoglobulin is a CD5 antibody which is primarily reactive with a surface antigen of mature T cells, but is also reactive with 10-20% of mature B cells. Clinical data obtained using an anti-pan T cell immunoglobulin in models of autoimmune diseases in non-human animals are predictive of the effects of using such immunoglobulins as therapy against human autoimmune diseases. Once prepared, humanized antibodies are then useful in the treatment of autoimmune disease. In this regard, an anti-CD5 monoclonal antibody is presented as an example of a preferred embodiment of the invention.

For the purpose of the present invention, an immunoglobulin, such as an antibody, is "reactive" with or "binds to" an antigen if it interacts with the antigen, forming an antigen-immunoglobulin complex. The antigen is generally a unique surface protein or marker. A most preferred marker is the CD5 antigen cluster.

An anti-pan T cell immunoglobulin may be obtained from a number of sources. It is reactive with most mature T cells or with both T cells and subsets of other lymphoid cells, such as B cells or natural killer (NK) cells. The immunoglobulin may be synthetic or recombinant, including genetically-engineered immunoglobulins such as chimeric immunoglobulins, humanized antibodies, hybrid antibodies, or derivatives of any of these.

Chimeric immunoglobulins, antibodies or peptides comprise fused portions from different species produced by chimeric DNA. Chimeric DNA is recombinant DNA containing genetic material from more than one mammalian species. Chimeric immunoglobulins include one portion having an amino acid sequence derived from, or homologous to, a corresponding sequence in an immunoglobulin, antibody or peptide derived from a first gene source while the remaining segment of the chain(s) is homologous to corresponding sequences from another gene source. For example, a chimeric antibody peptide may comprise an antibody heavy chain with a murine variable region and a human constant region. The two gene sources will typically involve two species, but will occasionally involve different sources from one species.

Chimeric immunoglobulins, antibodies, or peptides are typically produced using recombinant molecular and/or cellular techniques. Specifically, chimeric antibodies have variable domains of both light and heavy chains which mimic the variable domains of antibodies derived from one mammalian species, while the constant portions are homologous to the sequences in antibodies derived from a second, different mammalian species.

Immunoglobulins of the present invention may be monoclonal antibodies (hereinafter referred to as "MoAbs") of the IgM or IgG isotype of murine, human or other mammalian origin. Most preferably, such a MoAb is reactive with the CD5 antigen found on both T and B cells. MoAbs from other animal species may be prepared using analogous non-human mammalian markers.

In addition to the human-engineering methods of the current invention, a variety of methods for producing MoAbs are known in the art. See, e.g., Goding, Monoclonal Antibodies; Principles and practice (2d ed., Academic Press 1986), which is incorporated by reference herein. Less preferred forms of immunoglobulins may be produced by methods well-known to those skilled in the art, such as by chromatographic purification of polyclonal sera to produce substantially monospecific antibody populations.

Monoclonal antibodies specifically directed against human CD5 antigen may be obtained by using combinations of immunogens and screening antigens which have only the human CD5 antigen in common or by a screening assay designed to be specific for only anti-CD5 monoclonals. For example, production of monoclonal antibodies directed against CD5 may be accomplished by 1) immunization with human T cells expressing the CD5 antigen followed by screening of the resultant hybridomas for reactivity against a non-human cell line transfected with human CD5 (constructed in a manner similar to that described in Nishimura, et al., Eur. J. Immunol., 18:747-753 (1988)); 2)immunization with a non-human cell line transfected with human CD5 followed by screening of the resultant hybridomas for reactivity against a human T cell line expressing the CD5 antigen; 3) immunization with human or non-human cell lines expressing human CD5 followed by screening of the resultant hybridomas for ability to block reactivity of existing anti-CD5 monoclonals with a human T cell line; 4) immunization with human or non-human cell lines expressing human CD5 followed by screening of the resultant hybridomas for reactivity with purified native or recombinant CD5
antigen; or 5) immunization with a recombinant derivative of the human CD5 antigen followed by screening of the resultant hybridomas for reactivity against a human T cell line expressing CD5.

A preferred monoclonal antibody for use in preparing humanized antibodies according to the present invention is produced by hybridoma cell line XMMLY-H65 (H65) deposited with the American Type Culture Collection, 12301 Parklawn Drive, Rockville, Md. 20852 on and given the Accession No. HB 9286. A preferred antibody is prepared as disclosed herein using the human-engineered forms of the murine H65 antibody.

The generation of human MoAbs to a human antigen is also known in the art. See, e.g., Koda et al. Hum. Antibod. Hybridomas, 1(1):15-22 (1990). Generation of such MoAbs may be difficult with conventional techniques. Thus, it may be desirable to modify the antigen binding regions of the non-human antibodies, e.g., the F(ab').sub.2 or hypervariable regions (CDRs), and fuse them to human constant regions (Fc) or framework regions by recombinant DNA techniques to produce substantially human molecules using general modification methods described in, for example, EP publications 173,494 and 239,400, which are incorporated by reference herein.

Alternatively, one may isolate DNA sequences which encode a human MoAb or portions thereof which specifically bind to the human T cell by screening a DNA library from human B cells according to the general protocols outlined by Huse et al., Science 246:1275-1281 (1989); Marks, et al., J. Mol. Biol. 222:581-597 (1991) which are incorporated by reference herein, and then cloning and amplifying the sequences which encode the antibody (or binding fragment) of the desired specificity.

In addition to the immunoglobulins specifically described herein, other "substantially homologous" modified immunoglobulins may be readily designed and manufactured utilizing various recombinant DNA techniques known to those skilled in the art. Modifications of the immunoglobulin genes may be readily accomplished by a variety of well-known techniques, such as site-directed mutagenesis. See, Gillman et al., Gene 8:81-97 (1979); Roberts, et al., Nature 328:731-734 (1987), both of which are incorporated by reference herein. Also, modifications which affect the binding affinity of the antibody may be selected using the general protocol outlined by Marks, et al., J. Biol. Chem., 267:16007-16010 (1992), which is incorporated by reference herein.

In the present invention, an immunoglobulin, antibody, or peptide is specific for a T cell if it binds or is capable of binding T cells as determined by standard antibody-antigen or ligand-receptor assays. Examples of such assays include competitive assays, immunocytochemistry assays, saturation assays, or standard immunoassays such as ELISA, RIA, and flow cytometric assays. This definition of specificity also applies to single heavy and/or light chains, CDRs, fusion proteins, or fragments of heavy and/or light chains, which bind T cells alone or are capable of binding T cells if properly incorporated into immunoglobulin conformation with complementary variable regions and constant regions as appropriate.

In some competition assays, the ability of an immunoglobulin, antibody, or peptide fragment to bind an antigen is determined by detecting the ability of the immunoglobulin, antibody, or peptide to compete with a compound known to bind the antigen. Numerous types of competitive assays are known and are discussed herein. Alternatively, assays which measure binding of a test compound in the absence of an inhibitor may also be used. For instance, the ability of a molecule or other compound to bind T cells may be detected by labelling the molecule of interest directly, or it may be unlabelled and detected indirectly using various sandwich assay formats. Numerous types of binding assays such as competitive binding assays are known. See, e.g., U.S. Pat. Nos. 3,376,110, 4,016,043; Harlow et al., Antibodies: A Laboratory Manual, (Cold Spring Harbor Publications, N.Y. 1988), which are incorporated by reference herein.

Assays for measuring binding of a test compound to one component alone rather than using a competition assay are also available. For instance, immunoglobulins may be used to identify the presence of a T cell marker. Standard procedures for monoclonal antibody assays, such as ELISA, may be used. See, Harlow and Lane, supra. For a review of various signal producing systems which may be used, see U.S. Pat. No. 4,391,904, which is incorporated by reference herein.

Other assay formats may involve the detection of the presence or absence of various physiological or chemical changes which result from an antigen-antibody interaction. See Receptor-Effector Coupling--A Practical Approach (Hulme, ed., IRL Press, Oxford 1990), which is incorporated by reference herein.

Humanized antibodies of the present invention may be administered to patients with a disease having targetable cellular markers. Such diseases include, but are not limited to, autoimmune diseases such as lupus (including systemic lupus erythematosus and lupus nephritis), scleroderma diseases (including lichen sclerosis, morphea and lichen planus), rheumatoid arthritis and the spondylarthropathies, thyroiditis, pemphigus vulgaris, diabetes mellitus type 1, progressive systemic sclerosis, aplastic anemia, myasthenia gravis, myositis including polymyositis and dermatomyositis, Sjogren's disease, collagen vascular disease, polyarteritis, inflammatory bowel disease (including Crohn's disease and ulcerative colitis), multiple sclerosis, psoriasis and primary biliary cirrhosis; other diseases mediated by T cells, such as tissue transplant rejection and graft versus host disease; diseases caused by viral infections; diseases caused by fungal infections; diseases caused by parasites; and the like.

Immunoglobulins, antibodies or peptides according to the invention may be administered to a patient either singly or in a cocktail containing two or more antibodies, other therapeutic agents, compositions, or the like, including, but not limited to, immunosuppressive agents, potentiators and side-effect relieving agents. Of particular interest are immunosuppressive agents useful in suppressing allergic or other undesired reactions of a host. Immunosuppressive agents include prednisone, prednisolone, dexamethasone, cyclophosphamide, cyclosporine, 6-mercaptopurine, methotrexate, azathioprine, and gamma globulin. All of these agents are administered in generally accepted efficacious dose ranges such as those disclosed in the Physician's Desk Reference, 41st Ed. (1987). In addition to immunosuppressive agents, other compounds such as an angiogenesis inhibitor may be administered with the anti-pan T immunoglobin. See Peacock, et al., Arthritis and Rheum. 35 (Suppl.), Abstract, No. B141 (Sept. 1992).

Anti-pan T cell immunoglobulins may be formulated into various preparations such as injectable and topical forms. Parenteral formulations are preferred for use in the invention, most preferred is intramuscular (i.m.) or intravenous (i.v.) administration. The formulations containing therapeutically effective amounts of anti-pan T cell antibodies are either sterile liquid solutions, liquid suspensions or lyophilized versions and optionally contain stabilizers or excipients. Lyophilized compositions are reconstituted with suitable diluents, e.g., water for injection, saline, 0.3% glycine and the like, at a level of from about 0.01 mg/kg of host body weight to about 10 mg/kg or more of host body weight.

Typically, the pharmaceutical compositions containing anti-pan T cell immunoglobulins are administered in a therapeutically effective dose in a range of from about 0.01 mg/kg to about 5 mg/kg body weight of the treated animal. A preferred dose range of the anti-pan T cell antibody is from about 0.02 mg/kg to about 2 mg/kg body weight of the treated animal. The immunoglobulin dose is administered over either a single day or several days by daily intravenous infusion. For example, for a patient weighing 70 kg, about 0.7 mg to about 700 mg per day is a preferred dose. A more preferred dose is from about 1.4 mg to about 140 mg per day.

Anti-pan T cell immunoglobulin may be administered systemically by injection intramuscularly, subcutaneously, intrathecally, intraperitoneally, into vascular spaces, or into joints (e.g., intraarticular injection at a dosage of greater than about
1 .mu.g/cc joint fluid/day). The dose will be dependent upon the properties of the anti-pan T cell immunoglobulin employed, e.g., its activity and biological half-life, the concentration of anti-pan T cell antibody in the formulation, the site and rate of dosage, the clinical tolerance of the patient involved, the autoimmune disease afflicting the patient and the like as is well within the knowledge of the skilled artisan.

The anti-pan T cell immunoglobulin of the present invention may be administered in solution. The pH of the solution should be in the range of about pH 5.0 to about 9.5, preferably pH 6.5 to 7.5. The anti-pan T cell immunoglobulin or derivatives thereof should be in a solution having a pharmaceutically acceptable buffer, such as phosphate, tris (hydroxymethyl) aminomethane-HCl, or citrate and the like. Buffer concentrations should be in the range from about 1 to about 100 mM. A solution containing anti-pan T cell immunoglobulin may also contain a salt, such as sodium chloride or potassium chloride in a concentration from about 5 mM to about 150 mM. An effective amount of a stabilizing agent such as albumin, a globulin, a detergent, a gelatin, a protamine, or a salt of protamine may also be included and may be added to a solution containing anti-pan T cell immunoglobulin or to the composition from which the solution is prepared. Systemic administration of anti-pan T cell immunoglobulin is typically made every two to three days or once a week if a chimeric or humanized form is used. Alternatively, daily administration is useful. Usually administration is by either intramuscular injection or intravascular infusion.

Alternatively, anti-pan T cell immunoglobulin is formulated into topical preparations for local therapy by including a therapeutically effective concentration of anti-pan T cell immunoglobulin in a dermatological vehicle. Topical preparations may be useful to treat skin lesions such as psoriasis and dermatitis associated with lupus. The amount of anti-pan T cell immunoglobulin to be administered, and the anti-pan T cell immunoglobulin concentration in the topical formulations, will depend upon the vehicle selected, the clinical condition of the patient, the systemic toxicity and the stability of the anti-pan T cell immunoglobulin in the formulation. Thus, the physician will necessarily employ the appropriate preparation containing the appropriate concentration of anti-pan T cell immunoglobulin in the formulation, as well as the amount of formulation administered depending upon clinical experience with the patient in question or with similar patients.

The concentration of anti-pan T cell immunoglobulin for topical formulations is in the range from about 0.1 mg/ml to about 25 mg/ml. Typically, the concentration of anti-pan T cell immunoglobulin for topical formulations is in the range from about 1 mg/ml to about 20 mg/ml. Solid dispersions of anti-pan T cell immunoglobulin as well as solubilized preparations may be used. Thus, the precise concentration to be used in the vehicle may be subject to modest experimental manipulation in order to optimize the therapeutic response. Greater than about 10 mg of anti-pan T cell immunoglobulin/100 grams of vehicle may be useful with 1% w/w hydrogel vehicles in the treatment of skin inflammation. Suitable vehicles, in addition to gels, are oil-in-water or water-in-oil emulsions using mineral oils, petrolatum, and the like.

Anti-pan T cell immunoglobulin may be optionally administered topically by the use of a transdermal therapeutic system (Barry, Dermatological Formulations, p. 181 (1983)). While such topical delivery systems have been designed largely for transdermal administration of low molecular weight drugs, by definition they are capable of percutaneous delivery. They may be readily adapted to administration of anti-pan T cell immunoglobulin or derivatives thereof and associated therapeutic proteins by appropriate selection of the rate-controlling microporous membrane.

Preparations of anti-pan T cell immunoglobulin either for systemic or local delivery may be employed and may contain excipients as described above for parenteral administration and other excipients used in a topical preparation such as cosolvents, surfactants, oils, humectants, emollients, preservatives, stabilizers and antioxidants. Any pharmacologically acceptable buffer may be used, e.g., Tris or phosphate buffers.

Administration may also be intranasal or by other nonparenteral routes. Anti-pan T cell immunoglobulin may also be administered via microspheres, liposomes or other microparticulate delivery systems placed in certain tissues including blood.

Anti-pan T cell immunoglobulin may also be administered by aerosol to achieve localized delivery to the lungs. This is accomplished by preparing an aqueous aerosol or liposomal preparation. A nonaqueous (e.g., fluorocarbon propellent) suspension may be used. Sonic nebulizers preferably are used in preparing aerosols. Sonic nebulizers minimize exposing the anti-pan T cell antibody or derivatives thereof to shear, which can result in degradation of anti-pan T cell immunoglobulin.

Ordinarily, an aqueous aerosol is made by formulating an aqueous solution or suspension of anti-pan T cell immunoglobulin together with conventional pharmaceutically acceptable carriers and stabilizers. The carriers and stabilizers will vary depending upon the requirements for the particular anti-pan T cell immunoglobulin, but typically include nonionic surfactants (Tweens, Pluronics, or polyethylene glycol), innocuous proteins such as serum albumin, sorbitan esters, oleic acid, lecithin, amino acids such as glycine, buffers, salts, sugars, or sugar alcohols. The formulations are sterile. Aerosols generally may be prepared from isotonic solutions.

Alternatively, administration may be by an oral route, such as the method described by Steiner, et al. (U.S. Pat. No. 4,925,673), where the protein is delivered after encapsulation with a proteinoid substance.

Each of the foregoing materials and methods are illustrated by way of the following examples, which are not to be construed as limiting the invention. All references cited herein are incorporated by reference.

EXAMPLES

Example 1

A. The Use Of Anti-CD5 (Anti Lyt 1) In The Prophylactic Treatment Of Collagen Induced Arthritis In DBA/IJ Mice

Collagen-induced arthritis ("CIA") is a widely utilized model of human rheumatoid arthritis. CIA is characterized by a chronic polyarticular arthritis which can be induced in rodents and in primates by immunization with homologous or heterologous, native Type II collagen. The resulting arthritis resembles rheumatoid arthritis because there are similar histopathologic sequelae, cellular and humoral immune responses and restricted association with specific major histocompatibility complex ("MHC") haplotypes.

Native, heterologous Type II collagen emulsified with complete Freund's adjuvant induces an arthritis-like autoimmune reaction in DBA/IJ mice after a single intradermal tail injection. The mice were obtained from Jackson Laboratories, Bar Harbor, Me. Initially, the arthritis is noticeable as a slight swelling of one or more digits in the fourth week post-immunization. The chronic phase of CIA continually worsens over the ensuing 8 weeks as the arthritis progresses from the digits into the remaining peripheral articulating joints and eventually ends with ankylosis of the involved joints. The histopathology of CIA is characterized by lymphocyte infiltration of the joint space, synovial MHC class II expression and pannus formation. Not all joints are involved on every mouse, so there is a spectrum of arthritic severity. In a group of ten or more mice, the overall arthritic severity develops in a linear fashion over the course of 10-12 weeks.

The CIA model was used to test the potential efficacy of a monoclonal antibody directed against the pan-T cell surface antigen, Lyt-1, the murine equivalent of CD5. The antibody was administered to the mice before the immunization with Type II collagen. Normal DBA/I mice were also treated with a single 0.4 mg/kg i.v. injection of anti-Lyt-1 and were sacrificed after 72 hours for FACS analysis and for in vitro proliferation assays on spleen and lymph node cells. Any efficacy of this antibody would indicate a beneficial T cell-directed approach in rheumatoid arthritis via the CD5 surface antigen.

B. Effects of Anti-CD5 (Anti-Lyt-1) on DBA/IJ Spleen Cells and Peripheral Lymph Nodes

Antibody 53-7.313 is a rat IgG.sub.2a monoclonal antibody (ATCC Accession No. TIB 104) reactive with all alleles of the mouse lymphocyte differentiation antigen, Lyt-1. The IND1 antibody is a mouse IgG.sub.1, anti-human melanoma antibody used as a negative control (XOMA Corp., Berkeley, Calif.). All other antibodies were obtained from Pharmingen Inc. (San Diego, Calif.) as direct conjugates for quantitation on a Becton-Dickinson FACScan instrument.

Male DBA/IJ mice, age 6-8 weeks, were administered a single intravenous dose of either phosphate buffered saline, IND1 or anti-CD5 (anti-Lyt-1) via the tail vein at 0.4 mg/kg in 0.1 ml of phosphate buffered saline. Mice were sacrificed for analysis three days after dosing. Single cell suspensions of spleens and peripheral lymph nodes were prepared by standard procedures and 1.times.10.sup.6 cells were stained with the respective antibodies for fluorescence activated cell sorter (FACS) analysis. Proliferation assays were also performed to provide a second measure of T cell depletion. Cells (1.times.10.sup.5 /well) were stimulated with Concanavalin A, Interleukin-2 ("IL-2"), IL-2 and H57.597 (a pan .alpha.,.beta. T cell receptor antibody) or the Staphylococcal enterotoxins A and B. Cells were cultured for a total of 72 hours and proliferation was quantitated by the addition of .sup.3 H-methylthymidine for the last 24 hours. After 72 hours, the cells were harvested with an Inotech INB-384 harvesting and counting system, which collects the cells onto glass fiber filters with subsequent gas proportional beta particle detection. Results are generally expressed as the mean of triplicate wells.+-.SEM in Tables 4 and 5.

C. FACS Analysis Of Lymph Node And Spleen Cells

FACS analysis of lymph node cells ("LNC") and spleen cells ("SPC") from each treatment group (n=3/group) were analyzed for percent expression of .alpha.,.beta. T cell receptor, CD3, CD4, CD5, and CD8. The results are presented in Table 4.

TABLE 4 __________________________________________________________________________ FACS Analysis Of Anti-CD5 (Anti-Lyt-1) Treated DBA/1J Mice TREAT- CELL MENT TYPE .alpha., .beta.TCR CD3 CD4 CD8 CD5 __________________________________________________________________________ PBS LNC 80.2 .+-. 2.2% 79.8 .+-. 1.6% 58.7 .+-. 1.4% 19.4 .+-. 2.6% 80.0 .+-. 0.6% IND1 LNC 82.5 .+-. 1.3% 82.6 .+-. 1.9% 60.9 .+-. 2.0% 21.1 .+-. 1.5% 78.5 .+-. 1.2% .alpha.CD5 LNC *62.7 .+-. 5.8% *62.4 .+-. 1.0% *42.0 .+-. 1.9% 21.1 .+-. 0.2% *56.0 .+-. 2.6% PBS SPC 18.0 .+-. 2.8% 25.0 .+-. 0.1% 16.5 .+-. 2.1% 4.10 .+-. 0.5% 23.1 .+-. 0.1% INDI SPC 19.3 .+-. 1.6% 22.8 .+-. 1.4% 13.9 .+-. 0.8% 4.20 .+-.
0.3% 20.8 .+-. 1.5% .alpha.CD5 SPC 14.0 .+-. 0.3% *13.8 .+-. 0.4% *8.07 .+-. 0.3% *2.40 .+-. 0.1% *11.0 .+-. 0.1% __________________________________________________________________________

In Table 4, statistical significance was determined by Analysis of Variance followed by Duncan's New Multiple Range post-hoc test. These data indicate that administration of anti-CD5 (anti-Lyt-1) antibody results in a significant depletion of peripheral T lymphocytes at the 72 hour time point. The results could not be explained by residual circulating antibody as other T cell markers (CD3, etc.) are also depleted to a similar extent.

D. Effects Of Anti-CD5 (Anti Lyt-1) Administration On Proliferation Analysis

In vitro proliferation assays were performed on mice from each treatment group (n=3/group) in response to Concanavalin A, IL-2, IL-2+H57, Staphylococcal enterotoxin A and B ("SEA" and "SEB"). The results are presented in Table 5.

Overall, these data indicate that there is an observable and functional depletion of DBA/IJ T peripheral lymphocytes 72 hours after a single (0.4 mg/kg) intravenous dose of anti-CD5 (anti-Lyt-1) antibody.

E. Effects Of Anti-CD-5 (Anti Lyt-1) On Collagen-Induced Arthritis in DBA/IJ Mice

Male DBA/IJ mice, age 6-8 weeks, were administered the antibodies 53-7.313 anti-CD5 (anti-Lyt-1), IND1 (anti-melanoma) or phosphate buffered saline (PBS) in two intravenous (0.4 mg/kg) doses 48 hours apart starting four days prior to immunization with 100 .mu.g of bovine type II collagen emulsified with an equal volume of Freund's complete adjuvant to a final injection volume of 100 .mu.l. Each dose group was comprised of ten mice. Mice were monitored weekly starting on Day 21 after immunization. Individual mice were scored for arthritic severity by grading each paw on a scale from 0 to 2. A score of 1 indicated swelling in up to two digits and a score of 2 indicated swelling in more than two digits up to total paw involvement and ankylosis of the large joint in the later time points. An individual mouse could have a maximum arthritic severity score of 8. Mice were monitored until day 80 after collagen immunization and then were sacrificed by cervical dislocation. Results are expressed as the mean arthritic score for each dose group.

TABLE 5 __________________________________________________________________________ Proliferation Analysis Of Anti-CD5 (Anti-Lyt-1) Treated DBA/1J Mice TREAT- Concanavalin MENT A IL-2 IL-2 + H57 SEA SEB __________________________________________________________________________ IND1 26547 .+-. 3501 1181 .+-. 234 11341 .+-. 1663 12324 .+-. 1968 8747 .+-. 2025 .alpha.CD5 *11561 .+-. 4375 *593 .+-. 274 *4090 .+-. 2383 *5568 .+-. 2576 *1138 .+-.
350 __________________________________________________________________________

Statistical significance was determined by a Repeated Measures Analysis of Variance with one between subjects variable (antibody treatment). A Repeated Measures Analysis was necessary as each mouse was continually monitored for the duration of the study. Thus, the arthritic scores for consecutive days cannot be considered as independent observations contributing to the overall degrees of freedom in the F test for significant differences among groups. A Repeated Measures Analysis uses the degrees of freedom from the number of individuals per group instead of the number of observations. A typical between subjects Analysis of Variance may be inappropriate and may indicate false significance among the treatment groups. A comparison of means in the Treatment by Day after Immunization was done to determine the significance of anti-CD5 (anti-Lyt-1) treatment relative to PBS and IND1 control groups.

The changes in arthritic score during the course of the study are shown in FIG. 12, where circles indicate PSS, open boxes represent Ind1, and closed boxes represent anti-CD5 (anti-Lyt-1). The overall conclusion in FIG. 12 is that administration of the anti-CD5 (anti-Lyt-1) antibody prior to collagen immunization caused a significant decrease in the resulting severity of arthritis. In all of the treatment groups, the appearance of visible symptoms initiated at approximately 30 days after immunization and progressed linearly until the end of the study. The anti-CD5 (anti-Lyt-1) treatment group began to show ameliorated arthritic symptoms at day 48 and never developed arthritis to the same extent as the other two groups. The onset of arthritis was not significantly delayed by the anti-CD5 (anti-Lyt-1) treatment.

In conclusion, the intravenous administration of a rat monoclonal antibody reactive to the mouse equivalent of CD5, Lyt-1, is able to significantly decrease T lymphocytes in the spleen and in peripheral lymph nodes after a single 0.4 mg/kg dose. This T cell decrease is the probable mechanism for the significant (p<0.01) decrease in arthritic severity seen with the same anti-CD5 (anti-Lyt-1) dose prior to type II collagen immunization and provides evidence for therapeutic efficacy of .alpha.-CD5 antibodies.

Example 2

The Use Of OX19 Monoclonal Antibody In The Prophylactic Treatment Of Collagen Induced Arthritis In Diabetes-Resistant BB Rats

Collagen-induced arthritis (CIA) in the diabetes-resistant Biobreeding (DR BB) rat is a particularly relevant animal model of human rheumatoid arthritis, in that the DR BB rat RTl.D.beta. gene encodes a nucleotide sequence homologous to the human HLA-DR.beta. gene reported to be associated with rheumatoid arthritis susceptibility. In this model, DR BB rats are administered a single intradermal tail injection of heterologous Type II collagen emulsified with incomplete Freund's adjuvant. Development of the arthritis is considerably faster than in the DBA/1J CIA model. Onset of clinical signs occurs 1.5 to 2 weeks after collagen immunization, with peak swelling observed a few days after onset. Incidence is generally quite high (>85% of animals immunized). The swelling is generally severe, involves the entire footpad and ankle joint, and is restricted to the hindlimbs. Histopathological examination has revealed that the arthritis begins as a proliferative synovitis with pannus formation at the joint margins that is followed by a bidirectional erosion of both the outer (unmineralized) and inner (mineralized) layers of cartilage.

This experiment uses the DR BB CIA rat model to assess the efficacy of a MoAb, OX19 directed against the equivalent of the CD5 antigen in the rat. The antibody was administered to the rats prior to immunization with Type II collagen. Normal Sprague-Dawley rats were also treated with a single 0.5 mg/kg i.v. injection and were sacrificed after 3 hours for evaluation of MoAb binding to T cells, or after 2 days for quantitation of T cells in lymphoid tissues using flow cytometry.

A. Effects Of OX19 MoAb On T Cells In Lymphoid Tissues Of Normal Sprague-Dawley Rats

OX19 MoAb is a mouse IgG1 directed against the equivalent of rat CD5 antigen present on rat T cells. OX19 hybridoma is available from the European Collection of Animal Cell Cultures (ECACC) and has ECACC No. 84112012. H65 MoAb, a mouse IgG1
reactive against human CD5, was used as an isotype matched negative control. Fluorescein-conjugated antibodies directed against surface antigens on rat pan-T cells (W3/13), CD4 cells (W3/25) and CD8 cells (OX8) were obtained from Accurate Chemical and Scientific Corporation, Westbury, N.Y. for flow cytometric quantitation of T cells in rat lymphoid tissues. Phycoerythrin-conjugated goat anti-mouse IgG1 (Caltag Laboratories, South San Francisco, Calif.) was used to detect OX19 MoAb bound to rat T cells in a two-color analysis.

Male Sprague-Dawley rats (Simonsen Laboratories, Gilroy, Calif.), 100 to 150 grams, were divided into treatment groups, to which a single i.v. bolus injection of OX19 MoAb (0.5 mg/kg) or control MoAb (0.5 mg/kg) in phosphate buffered saline containing 0.1% Tween 80 (PBS/Tween) was administered. Animals were sacrificed at 3 hours (binding experiment) or 2 days (depletion experiment) after dosing. Single cell suspensions of blood, spleens and lymph nodes were prepared by standard procedures and 1.times.10.sup.6 cells were stained with appropriate antibodies for FACS analysis.

B. Binding Of OX19 MoAb To Rat T Cells In Vivo

Blood, spleen and lymph node cells from one animal in each treatment group were analyzed for the percentages of CD4 and CD8 T cells, and percentage of CD4 and CD8 T cells that also stained positively for surface-bound mouse IgG1 (CD4, CD4/MIgG1, CD8, or CD8/MIgG, respectively). The results are presented in Table 6.

TABLE 6 ______________________________________ Binding Of (Anti-CD5) OX19 MoAb To Rat T Cells In Vivo % Positive Cell CD8/ Tissue Treatment CD4 CD4/mIgG1* CD8 mIgG1* ______________________________________ Blood H65 MoAb 47.0 6.7 11.1 5.7 OX19 8.7 96.2 4.1 70.2 Spleen H65 MoAb 23.1 14.8 4.4 20.6 OX19 MoAb 16.4 84.8 3.4 73.6 Lymph Node H65 MoAb 66.9 4.2 7.4 6.5 OX19 MoAb 54.7 96.2 7.3 96.8 ______________________________________

As shown in Table 6, T cells were depleted from the blood at 3 hours after OX19 MoAb administration. Almost all of the T cells that remained in the blood, and most of those present in the spleen and lymph nodes in the OX19 MoAb-treated rat also stained positively for surface-bound mouse IgG1, indicating that the dose of OX19 MoAb used was sufficient to saturate most of the T cells in these major lymphoid organs. These results provide doses useful in therapeutic applications.

C. Effect of OX19 MoAb Treatment On T Cell Subpopulations In Rat Lymphoid Tissues

Blood, spleen and lymph node cells from two animals in each treatment group were analyzed for percentage of pan-T, CD4 and CD8 cells. The results are presented in Table 7 as the mean of the two animals.

TABLE 7 ______________________________________ FACS Analysis Of Tissues From OX19 (Anti-CD5) MAb-Treated Rats % Positive Cells Tissue Treatment Pan-T CD4 CD8 ______________________________________ Blood H65 MoAb 61.8 50.4 12.0 OX19 MoAb
47.0 37.3 8.8 Spleen H65 MoAb 36.0 25.3 7.1 OX19 MoAb 21.5 9.9 5.0 Lymph Node H65 MoAb 74.5 62.7 13.1 OX19 MoAb 33.8 24.9 4.3 ______________________________________

As shown in Table 7, OX19 MoAb treatment resulted in depletion of T cells from all tissues examined as compared to treatment with the control MoAb. These results also provide appropriate doses to be used in therapeutic applications using antibodies according to the invention.

Example 3

Effect Of OX19 MoAb Treatment On Development Of Collagen-Induced Arthritis In DR BB Rats

The ability of OX19 MoAb to prevent the development of collagen-induced arthritis was next measured in a manner similar to that described above in the mouse model. Male DR BB/Wor rats (obtained from the University of Massachusetts breeding facility; 8 per treatment group), age 6 weeks, were administered i.v. injections of OX19 MoAb (0.5 mg/kg), control MoAb (0.5 mg/kg) or buffer (PBS/Tween) on day 7 and day 4 prior to immunization at the base of the tail on day 0 with 0.3 mg of bovine Type II collagen emulsified in 0.15 ml of incomplete Freund's adjuvant. Rats were scored daily for arthritis beginning 8 days after collagen immunization. Severity was graded on a scale from 0 to 2, with a score of 1 indicating moderate swelling and a score of 2 indicating severe swelling. An individual animal could have a maximum arthritic severity score of 4 if there was bilateral hindlimb involvement.

The changes in arthritic score during the course of the study are shown in FIG. 15 and the arthritic incidence for each treatment group is presented in Table 8 and provides additional evidence of the therapeutic efficacy of anti-CD5 antibodies.

TABLE 8 ______________________________________ Effect Of OX19 (Anti CD5) MoAb Treatment On Arthritis Incidence Total Total arthritics Arthritics Score of "2" Score of "2" (1 or both (Both hind (1 or both (Both hind TREATMENT hind limbs) limbs) hind limbs) Limbs) ______________________________________ PBS/Tween 7/8 (88%) 7/8 (88%) 7/8 (88%) 5/8 (63%) Control MoAb 7/8 (88%) 4/8 (50%) 6/8 (75%) 4/8 (50%) OX19 MoAb 0/8 (0%) 0/8 (0%) 0/8 (0%) 0/8 (0%) ______________________________________

Control (buffer and control MoAb-treated) rats developed severe, predominantly bilateral hindlimb arthritis between days 10 and 14 with high incidence (88% for both groups). Treatment with OX19 MoAb completely prevented development of arthritis (0% incidence).

In conclusion, a 0.5 mg/kg intravenous dose of a mouse MoAb directed against the rat equivalent of CD5 was found to saturate and subsequently deplete T cells from lymphoid tissues of normal rats. This T cell depletion is the probable mechanism for the complete inhibition of arthritis development observed when the MoAb was administered prior to Type II collagen immunization in DR BB rats and provides additional evidence for the therapeutic efficacy of anti-CD5 antibodies.

Example 4

Preparation Of XMMLY-H65 Anti-Pan T Cell Immunoglobulin

The murine monoclonal antibody produced by cell line XMMLY-H65 (hereinafter referred to as "MoAbH65") is reactive with the human CD5 antigen. The cell line XMMLY-H65 was deposited with the American Type Culture Collection, 12301 Parklawn Drive, Rockville, Md., 20852 and designated as Accession No. HB9286.

MoAbH65 was produced after immunization of BALB/c mice with the human T-cell line HSB-2 originally isolated from a patient with T-cell acute lymphocytic leukemia. Adams, et al. Can. Res. 28:1121 (1968). The murine myeloma cell line P3 7
NS/1-Ag-1-4 of Kohler et al. Ernr. J. Immunol. 6:292 (1976) was fused with spleen cells from an immunized mouse by the technique of Galfre et al., Nature 266:550 (1977). One of the resulting hybrid colonies was found to secrete a MoAb that recognizes a pan-T-lymphocyte antigen with a molecular weight of 67 kD, expressed on approximately 95% of peripheral T-lymphocytes [Knowles, Leukocyte Typing II, 1, (E. Reinherz, et al. eds., Springer Verlag (1986)]. This antigen is not present on the surface of any other hematopoietic cells, and the antibody itself has been tested for binding to a large range of normal human tissues and found to be negative for all cells except for T-lymphocytes and a subpopulation of B lymphocytes.

The H65 antibody-producing hybrid cell line was cloned twice by limiting dilution and was grown as ascites tumors in BALB/c mice.

MoAbH65 was purified from mouse ascites by a modification of the method of Ey et al. Immunochem. 15:429 (1978). In brief, the thawed mouse ascites was filtered to remove lipid-like materials and was diluted with 2 to 3 volumes of 0.14M NaPO.sub.4, pH 8.0, before application onto an immobilized protein A-Sepharose column of appropriate size. The unbound materials were removed from the column by washing with 0.14M NaPO.sub.4, pH 8.0, until no further change in absorbance at 280 nm was seen. A series of column washes with 0.1M sodium citrate (pH 6.0, pH 5.0, pH 4.0, and pH 3.0) were then performed to elute bound antibody.

Peak fractions were pooled, adjusted to pH 7.0 with saturated Tris base, and concentrated by using a cell stirred with Amicon YM10 membrane (Amicon, Lexington, N.Y.). An antibody solution was then dialyzed against phosphate-buffered saline (PBS), pH 7.0, and was stored frozen at -70.degree. C.

MoAb H65 is of the IgG.sub.1 subclass, as determined by double diffusion in agar with the use of subclass-specific antisera (Miles-Yeda, Ltd. Rehovot, Israel). The serologic characteristics of this antibody and the biochemical characteristics of the gp67 (i.e., CD5) antigen were examined during the First International Workshop on Human Leukocyte Differentiation Antigens (Paris, 1982). MoAb H65 (workshop number: T34), and nine other MoAbs were found to have the same serologic pattern and to immunoprecipitate the gp67 antigen. Knowles, in Reinherz, et al., Leukocyte Typing II, 2:259-288 (Springer-Verlag, 1986). In other studies, MoAb H65 has been shown to block the binding of FITC-conjugated anti-Leu-1 (Becton Dickson, Mountain View, Calif.) on CD5+ cells indicating that both antibodies recognize the same epitope on the CD5 molecule or determinants that are located in such a configuration as to result in blocking by steric hindrance.

Example 5

Depletion Of Human T Cells From SCID Mice By Treatment With H65 MoAb

Severe combined immunodeficient (CB.17 scid/scid; SCID) mice maintain human lymphoid cells for several months following transplantation of human peripheral blood mononuclear cells (PBMC). Such chimeric mice, referred to as PBMC/SCID mice, have functional human cells, as shown by the presence of human Ig in their serum. PBMC/SCID mice maintain human T cells in tissues such as spleen and blood. Human T cells present in PBMC/SCID mice are predominantly of a mature phenotype and express T cell antigens, including CD3, CD5, CD7, and CD4 or CD8. In addition, most T cells appear to be activated memory cells, as judged by the expression of HLA-DR and CD45RO. These engrafted T cells appear to be functional since (a) they may provide help to B cells to produce anti-tetanus toxoid antibodies, (b) they produce soluble interleukin-2 receptor (sIL-2R) which may be detected in plasma, and (c) they proliferate in response to mitogenic anti-human CD3 monoclonal antibodies supplemented with IL-2 in vitro.

Because of the presence of human T and B cells, PBMC/SCID mice offer an in vivo model system in which to evaluate the efficacy of anti-human T cell drugs, such as H65 MoAb, a mouse IgGI directed against human CD5. The therapeutic efficacy of such anti CD5 antibodies was demonstrated in Examples 1-3 above.

The SCID mice were obtained from Taconic, Germantown, N.Y., and at 6 to 7 weeks of age were injected with 200 mg/kg cyclophosphamide intraperitoneally (i.p.) to ensure engraftment of human PBMC. Two days later, 25 to 40.times.10.sup.6 human PBMC, isolated by Ficoll-Hypaque density gradient centrifugation from lymphapheresis samples obtained from normal donors (HemaCare Corporation, Sherman Oaks, Calif.), were injected intraperitoneally.

At 2 to 3 weeks after PBMC injection, the mice were bled from the retro-orbital sinus and levels of human immunoglobulin (Ig) and human sIL-2R in plasma were quantified using sandwich ELISAs. Mice with low or undetectable levels of these human proteins were eliminated from the study and the remainder were divided into the various treatment groups (6 per group). The mice were then administered H65 MoAb (0.2 or 0.02 mg/kg/day), H65-based F(ab').sub.2 fragment (2 mg/kg/day) or vehicle (buffer) intravenously (i.v.) for 10 consecutive daily injections. One day after the last injection, the mice were bled and spleens were collected. Single cell suspensions of blood cells and splenocytes were prepared by standard methods. Recovered cells were then assayed for human T cell surface markers using flow cytometry.

Cells (2.times.10.sup.5) were stained with the following FITC- or PE-conjugated Abs (Becton-Dickinson, Mountain View, Calif.): HLe-1-FITC (anti-CD45), Leu-2-FITC (anti-CD8), and Leu-3-PE (anti-CD4). Samples were analyzed on a FACScan using log amplifiers. Regions to quantify positive cells were set based on staining of cells obtained from naive SCID mice. The absolute numbers of human antigen-positive cells recovered from SCID tissues were determined by multiplying the percent positive cells by the total number of cells recovered from each tissue sample. The total number of leukocytes in blood was calculated using a theoretical blood volume of 1.4 ml/mouse. Statistical comparisons between treatment groups were made using the Mann-Whitney U test.

The number of human T cells (CD4 plus CD8 cells) recovered from spleens and blood of PBMC/SCID mice following treatment with H65 MoAb or vehicle (control) is shown in FIGS. 13A and 13B, wherein the dash in the figures represents the median value. Significantly (pL 0.05) lower numbers of T cells were recovered from spleens and blood of mice treated with either 0.2 or 0.02 mg/kg/day H65 MoAb as compared to vehicle-treated mice.

In contrast, treatment with 2 mg/kg/day of an H65-based F(ab').sub.2 fragment did not significantly deplete human T cells from spleens or blood, even though a 10 to 100-fold higher dose was used (FIGS. 14A and 14B). Median values in FIGS. 14A and B are indicated by dashes.

These results indicate that an anti-human CD5 MoAb depletes human T cells in an experimental animal model in a manner similar to the depletion of T-cells, demonstrated in Examples 1-3 with anti-mouse as anti-rat CD5. Because anti-CD5 antibodies, including H65, were therapeutically effective, humanized anti-CD5 antibodies with comparable affinities but without significant immunogencity would be useful. The ability of this MoAb to deplete human T cells from SCID mice is apparently dependent on the Fc portion of the MoAb, as an F(ab').sub.2 fragment was ineffective.

Example 6

Identification Of Low Risk Residues in A Mouse Variable Domain

A method of the present invention was utilized to prepare modified antibody variable domains by identi