US20040234531A1

US20040234531A1 - Chimeric human leukocyte antigen and epitope-bearing molecules having immunosuppressant activity

Info

Publication number: US20040234531A1
Application number: US10/476,104
Authority: US
Inventors: Sofia Casares; Teodor Brumeanu; Constantin Bona
Original assignee: Mount Sinai School of Medicine
Current assignee: Icahn School of Medicine at Mount Sinai
Priority date: 2001-05-07
Filing date: 2002-05-07
Publication date: 2004-11-25
Also published as: US20080311133A1; EP1392356A1; EP1392356A4; US20040258697A1; US6811785B2; WO2002089837A1; US20020164340A1

Abstract

The present invention relates to chimeric molecules that comprise human major histocompatibility complex elements, linked to epitopes of interest, that have been joined together to form dimers or higher multimers. Such chimeric molecules may be used to treat or prevent diseases associated with autoimmunity, particularly diabetes. It is based, at least in part, on the discovery that a chimeric molecule comprising an IILA-DR element linked to an epitope of GAD65, an antigen associated with autoimmune diabetes, stimulated the secretion of the inhibitory cytokine IL-10 from CD4 T cells of Type I diabetic patients.

Description

1. INTRODUCTION

The present invention relates to chimeric molecules that comprise human major histocompatibility complex elements, linked to epitopes of interest, that have been joined together to form dimers or higher multimers. Such chimeric molecules may be used to treat or prevent diseases associated with autoimmunity, particularly diabetes. It is based, at least in part, on the discovery that a chimeric molecule comprising an HLA-DR element linked to an epitope of GAD65, an antigen associated with autoimmune diabetes, stimulated the secretion of the inhibitory cytokine IL-10 from CD4 T cells of Type 1 diabetic patients.

2. BACKGROUND OF THE INVENTION 2.1 Diabetes as an Autoimmune Disease

Type 1 diabetes, also known as “insulin dependent diabetes mellitus” (“IDDM”) is usually diagnosed before the age of 30 (hence its former name, “juvenile diabetes”) and occurs in 1 out of 800 people in the United States (see www. niaid.nih.gov/publications/autoimmune.htm#diabetes). It has been shown to be caused by an autoimmune attack on cells of the pancreas, called “β cells” which produce insulin. The resulting depletion of insulin results in an inability to absorb, store and utilize carbohydrates such as glucose, and therefore produces symptoms of weakness, weight loss, excessive hunger and thirst—“starvation amidst plenty” (see my.webmd.com/content/dmk/dmk_article_—40027). The inability to process ingested carbohydrates produces high levels of sugar in the blood, which is associated with a number of pathologies, including cardiovascular disease, renal failure, digit and limb amputation and blindness. Because most of the insulin-producing cells of the body are destroyed, patients with type 1 diabetes are typically dependent on insulin treatment for virtually their entire lives.

A number of observations and experimental results demonstrate the pathogenicity of T cells in type 1 diabetes. Pancreatic β cell islets from pre-diabetic and diabetic animal models are infiltrated with T cells (Miyakazi et al., 1985, Clin. Exp. Immunol. 60: 622-625). Neonatal thymectomy prevents the disease (Like et al., 1982, Science 216: 644-646). Transplantation of islets into the thymus induced T-cell tolerance and prevention of diabetes (Herold et al., 1992, J. Exp. Med. 176: 1107-1114). The disease is transferred by autoreactive bone marrow-derived cells, splenic T cells and T cell clones isolated from infiltrated islets (Serreze et al., 2988, Diabetes 37:252-255; Wicker et al., 1986, Diabetes 35:855-860; Peterson et al., 1994, J. Immunol. 153:2800-2806). Adoptive transfer of protective CD4⁺ T cells prevents diabetes (Zekzer et al., 1997, Diabetes 46:1124-1132). Non-obese diabetic severe-combined immunodeficient (NOD-SCID) mice do not develop the disorder (Bowman et al., 1994, Immunol Today 15:115-120, whereas the same mice expressing a monoclonal diabetogenic T cell receptor (“TCR”) develop insulitis and diabetes (Verdaguer et al., 1997, J. Exp. Med. 186:1663-1676).

Both genetic susceptibility and environmental factors contribute to the pathogenesis of type 1 diabetes; genetically, it is controlled by major histocompatibility complex (“MHC”) class II molecules (Todd et al., 1987, Nature 329: 599-604, and see section 3.1, below) and non-MHC alleles (Buzzetti et al., 1998, Diabetes Metab. Rev. 14: 111-128). The association between MHC class II and autoimmune diabetes in both humans and animal models strongly suggests a pathogenic role for CD4⁺ T cells. This is because MHC class II molecules are required for thymic education of precursors and for the presentation of peptides to CD4⁺ T cells (Nepon et al., 1991, Annu. Rev. Immunol. 9: 493-525).

Patel et al. (1997, Proc. Natl. Acad. Sci. U.S.A. 94:8082-8087) studied the association between type 1 diabetes and MHC class II genotype. They report that the strongest associations of type 1 diabetes susceptibility in Caucasians are with the HLA-DQ loci (human major histocompatibility antigens are referred to as Human Leukocyte Antigens, hence “HLA”; for explanation and nomenclature, see section 3.1, below), specifically with alleles that carry a neutral residue (Val, Ala, or Ser) at position 57 of the DQβ chain, such as the HLA-DQB1*0302 allele, while other alleles encoding an Asp residue at the same position are negatively related to susceptibility (Todd et al., 1987, Nature 329: 599-604). In addition, 90% of patients with type 1 diabetes express HLA-DR3 or HLA-DR4 (Nepom and Erlich, 1991, Annu. Rev. Immunol. 9: 493-525). HLA-DRB1*0405 is associated with type 1 diabetes in populations of Mexican Americans (Erlich et al., 1993, Nat. Genet. 3: 358-363), Sardinians (Cucca et al., 1995, Hum. Immunol. 43: 301-308), and Belgians (Van der Auwera et al., 1995, Diabetes 44: 527-530), whereas HLA-DRB1*0403 confers a protective effect. HLA-DRB1*0405 increases the effect of HLA-DQB1*0302 when found on the same haplotype, while HLA-DRB1*0403 acts in a dominant manner to confer protection (Cucca et al., 1995, Hum. Immunol. 43: 301-308;Van der Auwera et al., 1995, Diabetes 44: 527-530)), even when linked to DQB*0302 with the DRB1*0301-DQB1*0201 susceptibility haplotype present on the other chromosome. HLA-DRB1*0401 correlates with type 1 diabetes in northern Europeans. The degree of contribution of HLA-DRB1*04 alleles to disease has been suggested to have the following order: HLA-DRB1*0405, *0402, *0401, *0404, *0403, and *0406 (Undlien et al., 1997, Diabetes 46: 143-149).

Several β-islet cell proteins have been identified as potential disease targets (Nepom, 1995, Curr. Opin. Immunol. 7: 815-830). Eighty percent of prediabetic individuals followed in families with a history of diabetes, and most recent onset diabetic patients, have been found to have serum autoantibodies directed against glutamic acid decarboxylase (GAD; EC 4.1.1.15; Baekkeskov et al., 1987, J. Clin. Invest. 79: 926-934). Furthermore, half of new-onset patients have T cell proliferative responses to GAD (Hagopian et al., 1995, J. Clin. Invest. 95:1505-1511; Harrison et al., 1993, Lancet 341: 1365-1369). Finally, GAD isoform 2 (GAD65) has been used in tolerogenic regimens to prevent disease in susceptible nonobese diabetic (NOD) mice (Tian et al., 1996, Nat. Med. 2: 1348-53; Tisch et al., 1993, Nature 366: 72-75; Kaufman et al., 1993, Nature 366: 69-72). Patel et al. (1997, Proc. Natl. Acad. Sci. U.S.A. 94:8082-8087) have defined the immunodominant epitopes for GAD65 (shown in Table 1), demonstrated that these epitopes are processed by human antigen-presenting cells (APCs) and presented in an HLA-DR4-restricted manner, and have found some of these epitopes to elicit T cell responses in DR0401 patients.

TABLE 1*


Position	Sequence	SEQ ID NO:

116-130	MNILLQYVVKSFDRST	4

	PRLIAFTSEHSHFSL	10

271-285	PRLIAFTSEHSHFSL

356-370	KYKIWMHVDAAWGGG	13

376-390	KHKWKLSGVERANSV	15

481-495	LYNIIKNREGYEMVF	18

511-525	PSLRTLEDNEERMSR	22

546-560	SYQPLGDKVNFFRMV	24

551-565	GDKVNFFRMVISNPA	25

556-570	FFRMVISNPAATHQD	26

566-580	ATHQDIDFLIEEIER	27

Another β-cell protein which appears to function as an autoantigen is proinsulin (Griffin et al., 1995, Am. J. Pathol. 147:845-857). Rudy et al. (1995, Mol. Med. 1(6): 625-333 report that a region of proinsulin, amino acid residues 24-33, bears homology to a region at the C-terminus of GAD65 (506-518) Preproinsulin and proinsulin T cell epitopes are described in Congia et al., 1998. Proc. Natl. Acad. Sci. U.S.A. 95:3833-3838, including an immunodominant HLA DRB1*0401-restricted peptide at residues 73-90, having a minimal core epitope with the amino acid sequence LALEGSLQK (SEQ ID NO: 45).

The immunodominance of an epitope is related to the HLA phenotype. For example, Raju et al., 1997, Hum. Immunol. 58:21-29 observed a differential pattern of recognition of epitopes on human pre-proinsulin (HPI) polypeptide presented by the HLA DQ8 allele as compared to HLA DQ6. The sequences 1-24 and 44-63 were immunodominant in DQ8 transgenic mice while DQ6 transgenic mice primarily recognized sequences 14-33 and 74-93 of HPI.

Analysis of the autoimmune nature of type 1 diabetes has led to attempts at prevention by two major approaches. First, there is the possibility of general immune suppression achieved by the administration of cyclosporin A (Bach, 1994, Endocrin. Rev. 15: 516-542), the anti-inflammatory cytokines interleukin 4(“IL-4”), IL-10, IL-11, IL-13 and transforming growth factor-β (“TGF-β) (Cameron et al., 1997, J. Immunol. 159:4686-4692; Nitta et al., 1998, Hum. Gen. Ther. 9:1701-1707; Nicoletti et al., 1999, Diabetes 48:2333-2339; Zaccone et al., 1999. Diabetes 48:1522-1528; and Piccorillo et al., 1998. J. Immunol. 161:3950-3956), antibodies specific for the proinflammatory cytokines interferon-γ (“IFN-γ”) and tumor necrosis factor-α (“TNF-α), TCR-CD3 complex Herold et al., 1992, Diabetes 41:457-464), CD4 and CD8 coreceptors (Wang et al., 1987. Diabetes 36:535-538; Shizuku et al., 1988, Science 240:659-662), CD25 activation marker (Kuttler et al., 1999, J. Mol. Med. 77:226-229) or CD40 ligand (CD40L) and B7-2 costimulatory molecules (Balasa et al., 1997, J. Immunol. 159:4620-4627; Lenschow et al., 1995, J. Exp. Med. 181:1145-1155). These reagens have a variety of side effects that call into question the ethics of applying them to asymptomatic humans.

A second avenue of treatment involves the possibility of using immunospecific therapy to induce tolerance in autoreactive T cells after administration of immunodominant peptides derived from the major β cell antigens. Prevention of diabetes in animal models has been achieved through oral, intranasal, intravenous, or subcutaneous administration of B9 peptide from the insulin B chain (Daniel and Wegmann, 1996, Proc. Natl. Acad. Aci. U.S.A. 93:956-960); p217, p247, p290 and p524 peptides from glutamic acid decarboxylase 65 (“GAD65”)(Tian et al., 1996, J. Exp. Med. 183:1561-1567; Tian et al., 1996, Nature Med. 2:1348-1353; Maron et al., 1999, J. Autoimmun. 12:251-258; and Tisch et al., 1999, J. Immunol. 163:1178-1187), or p12 and p27 peptides from the heat shock protein hsp60 (Elias et al., 1991, Proc. Natl. Acad. Sci. U.S.A. 88:3088-3091; Bockova et al., 1997, J. Autoimmun. 10:323-329). However, some studies have shown a lack of protection by peptide-based therapy (Petersen et al., 1997, Autoimmunity 25:129-138; Funda et al., 1998, APMIS 106:1009-1016), mostly because of the short lifespans of the peptides in blood, which ranges from 22 seconds to 10 minutes (Ishioka et al., 1994, J. Immunol. 152:4310-4319).

Latent Autoimmune Diabetes of Adults (“LADA”) may represent a significant proportion of patients diagnosed with adult onset type 2 diabetes, hitherto believed to generally represent a genetic rather than an autoimmune condition. In a Sardinian study, antibodies directed toward GAD65 and IA-2, markers of Type 1 diabetes, were found among Type 2 diabetes siblings from Sardinian multiplex families despite excluding those who had been treated with insulin during the first 4 years of disease, supporting the importance of investigating markers of Type 1 diabetes in studies of Type 2 diabetes (Maioli et al., 2002, Diabetes Res. Clin. Pract. 56:41-47). See also Tan and Lim, Singapore Med. J. 42:513-516. Falorni et al. (2000, J. Clin. Endocrinol. 85(1):309-316) report that the presence of GAD65Ab binding to COOH-terminal epitopes is strongly associated with a need for insulin requirement in patients carrying a diagnosis of type 2 diabetes, suggesting that these patients may in fact be suffering from LADA.

2.2 MHC/Epitope Chimeric Molecules

A major drawback of peptide-based therapy as a potential treatment for type I diabetes has been overcome by the genetic engineering of soluble peptide-MHC chimeras (pMHC”)(Casares et al., 1998, Biotech Gen. Engineer. Rev. 15:159-198). The soluble dimeric pMHC class II chimera “mDEF” is on an immunoglobulin scaffold consisting of murine major histocompatibility element I-E ^dlinked to an eleven amino acid residue peptide derived from influenza A/PR8/34 virus hemaglglutinin (“HA(110-120)”) and the Fcγ2A constant region. The DEF molecule has been found to bind specifically to cognate T cells (Casares et al., 1997, Protein Engineer. 10:1295-1301) and has a half life of 50 hours in blood circulation (Brumeanu et al., 2001, Int. Rev. Immunol. 20:301-323; Casares et al., 2001, Protein Engineer. 14:195-200). This molecule has remarkable modulatory effects on T cell function. At low TCR-CD4 occupancy, T cells were polarized toward the T helper type 2 (“T _H2”) phenotype and had subsequent bystnader inhibitory effects on CD8⁺ T cell function as a result of IL-2 deprivation (Casares et al., 1999, J. Exp. Med. 188:1633-1640). At high TCR-CD4 occupancy, DEF induced T cell anergy (Brumeanu et al., 2001, Int. Rev. Immunol. 20:301-323; Casares et al., 2001, Protein Engineer. 14:195-200). Similar soluble pMHC class I chimeras have been engineered (Hamad et al., 1998, J. Exp. Med. 188: 1633-1640; Appel et al., 2000, J. Biol. Chem. 275:312-321) and the down-regulatory effects of these reagents on antigen-specific CD4⁺ T cells confirmed in vitro.

The theory behind such molecules is that T cells recognize and respond to antigens when they are in a specific molecular context—namely, when the T cell receptor encounters an antigen together with an MHC molecule. Cytotoxic T lymphocytes (which bear CD8 surface antigen, and are referred to as “CD8” T lyphocytes”) will only kill cells bearing a foreign antigen in the context of a self MHC class I molecule (for definitions of class I and II molecules, see section 3.1, below). Analogously, helper T lymphocytes (which bear CD4 surface antigen, and are referred to as “CD4 ⁺” T lymphocytes”) will only proliferate in response to a foreign antigen in the context of a self MHC class II molecule on the surface of a specialized cell of the immune system referred to as an “Antigen Presenting Cell” (“APC”) (for review, see Davies, 1997, Introductory Immunobiology, Chapman & Hall, New York, pp. 177-223). It is believed that when a T cell encounters an antigen in the context of the appropriate MHC molecule but without costimulatory molecules present, for example, on the APC, it “short circuits” and becomes unresponsive. DEF is a molecule which presents the TCR with an epitope and an MHC element, but no costimulatory molecule.

This mechanism of T cell unresponsiveness seems to account for the peripheral self-tolerance of tissue-specific antigens (Guerder et al., 1995, Int. Rev. Immunol. 13:135-146). Based on this knowledge, soluble antigen presenting molecules have been designed as potential immunomodulatory agents with varying degrees of success (Nag et al., 1996, Cell. Immunol. 170:25-33; Sharma et al., 1991, Proc. Natl. Acad. Sci. U.S.A. 88: 11465-11469; Spack et al., 1995, J. Autoimmunity 8: 787-807; Abastado et al., 1995, J. Exp. Med. 182: 439-447; Altman et al., 1996, Science 274: 94-96; Godeau et al., 1992, J. Biol. Chem. 267: 24223-24229; Scheirle et al., 1992, J. Immunol. 149: 1994-1999; Scott et al., 1996, J. Exp. Med. 183: 2087-2095; Stem and Wiley, 1992, Cell 68: 465-477; International Patent Application No. PCT/US95/02698, Publication No. WO95/23814 by Kappler and Marrack; Kozono et al., 1994, Nature 369: 151-154; Mottez et al., 1995, J. Exp. Med. 181: 493-502; Rhode et al., 1996, J. Immunol. 15: 4885-4891).

Epitope-bearing major histocompatibility complex chimeric molecules, including DEF, are described in International Patent Application No. PCT/US97/20023, Publication No. WO 99/09064 by Casares et al., and the possible utility of such molecules in diabetes is suggested (see also Brumeanu and Casares, 2001, Int. Rev. Immunol. 20: 301-331; Casares and Brumeanu, July 2001, Cur. Mol. Med. 1:357-378). Such molecules have been demonstrated to selectively deliver toxic molecules to antigen specific T cells in mice (Casares et al., 2001, Nature Biotechnol. 17:1-9). Prior to the present invention, however, it was not known or expected that chimeric molecules bearing human major histocompatibility elements together with autoimmune disease associated antigens would be effective in modulating the human immune response toward treatment or prevention of specific autoimmune diseases such as diabetes.

3. SUMMARY OF THE INVENTION

The present invention relates to chimeric proteins that comprise human major histocompatibility complex (HLA) Class II elements, linked to epitopes which are associated with the pathogenesis of human autoimmune diseases, where the chimeric proteins have been joined together to form dimers or higher multimers. Such chimeric molecules may be used to treat or prevent diseases associated with autoimmunity.

In particular embodiments, the invention provides for methods and compositions for treating or preventing autoimmune diabetes. According to such embodiments, HLA class II elements are linked to epitopes of pancreatic cell proteins such as GAD65 and proinsulin. Particularly useful are the combination of HLA class II elements which are epidemiologically associated with diabetes, such as HLA-DR4 alleles, with peptide fragments of GAD65 or proinsulin bearing immunodominant epitopes, such as peptide 271-285 of GAD65 or peptide 32-46 (also known as the B9-23 peptide) or peptide 73-90 of proinsulin. The invention is based, at least in part, on the discovery that a chimeric molecule comprising HLA-DR4 elements linked to an epitope of GAD65, an antigen associated with autoimmune diabetes, stimulated the secretion of the inhibitory cytokine IL-10 from CD4 T cells of Type 1 diabetic patients.

In addition to providing for HLA-/epitope chimeric molecules, the present invention encompasses methods of diagnosing and treating autoimmune disease. In a particular non-limiting subset of embodiments, the invention provides for the use of chimeric HLA/epitope molecules in the treatment of type 1 diabetes presenting as either early onset disease, previously known as juvenile diabetes, or Latent Autoimmune Diabetes of Adult (“LADA”). Because of its protracted onset, LADA may be particularly responsive to treatment according to the invention.

3.1 Major Histocompatibility Complex Antigens

Major histocompatibility complex (“MHC”) antigens are used by the immune system to determine whether or not to react. For example, these antigens are the basis for acceptance or rejection, by a recipient host, of a donor graft tissue—the “histocompatibility” of donor and recipient cells. If the MHC antigens of donor and recipient are the same or very similar, the graft should be accepted by the host and a graft-versus-host reaction should not occur.

At a molecular level, MHC antigens participate in the interaction between a T cell and its target; in order to generate an immune reaction toward a non-self target antigen, a T cell must encounter the target antigen in the context of an MHC antigen recognized as “self.” The portion of the target antigen recognized by the T cell (via its T cell receptor, or “TCR”) is an “epitope”. There can be multiple epitopes within a given antigen, and they may interact with different TCRs. The phenomenon whereby a T cell will only act in the context of a particular MHC antigen is referred to as “MHC restriction”.

There are two classes of MHC molecules, referred to as “class I” and “class II” antigens. Cytotoxic T cells bearing the marker CD8, which act to destroy virally infected cells or cancer cells, perceive target antigen in the context of class I MHC molecules. Helper T cells bearing the marker CD4, which promote the antibody-mediated type of immune response, interact with target antigen in the context of class II MHC molecules.

Class II molecules are heterodimers comprised of and β chains, each having two domains and being of approximately the same length. The genes encoding class II molecules in humans are arranged in the following order: DP {β2-α2-β1-α1}—DQ {β2-α2-β3-β1-α1}—DR {β1-β2-β3-β9-α1}.

With respect to nomenclature, human MHC antigens are referred to as Human Leukocyte Antigens (“HLAs”) and murine MHC antigens are termed “H-2 antigens.” Class I molecules are encoded at the A, B and C loci in man and the K, D and L loci in the mouse. Class It molecules are encoded at the DP, DQ and DR loci in man and the I-A and I-E loci in the mouse.

In naming human HLA genes, one type of nomenclature used herein, for example, HLA-DR4, is a somewhat shorthand approach to a second, more recent and specific nomenclature, for example HLA-DRB1*0401. In the first type of nomenclature, e.g., HLA-DR4, the class II molecule is expressed from the DR locus and falls in serological group 4. As technology has advanced, differences have been found amongst antigens that fall within serologic group 4—these are alleles, and are designated, in the second type of nomeclature, preceded by an asterisk. “B1” in HLA-DRB1 refers to the β1 subregion of the DR locus, encoding the β1 domain.

4. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1A-G. mDEF protects doubly transgenic (“dTg”) mice against diabetes. (A) Cumulative incidence of diabetes in groups of 15-day-old prediabetic dTg mice injected intravenously every five days with mDEF (N=12 mice), MOPC-173 (n=8), HA(110-120) peptide (30 ng, n=8 or 3 μg, n=8) or saline (n=14). (B) Frequency of infiltrated and noninfiltrated islets in the pancreata of 3-month-old dTg mice that were either untreated or treated every 5 days with mDEF. Forty islets were counted, with the use of microscopes, for each pancreas. Data are means ±s.e.m. of four mice per group. (C,D) Histological examination of pancreatic islets from 3-month-old normoglycemic dTg mice that were treated with mDEF and age-matched diabetic dTg untreated mice. Sections of pancreata were stained with (C) hematoxylin and eosin (“H&E”) or with (D) rabbit anti-insulin antibody followed by HRP-goat anti-rabbit Ig antibody. Insulin was revealed in the islets as a red-brownish color. (E) Reversal of diabetes in mice treated with a single i.v. injection of mDEF (2 μg, n=16 mice). Nine of sixteen mice showed normoglycemia after treatment. Diabetes was not reversed by MOPC-173 (2 μg, n=8) or HA(110-120) synthetic peptide (30 ng, n=8). (F,G) Histological examination of pancreatic islets from 4 month-old HA-Tg mice treated priodically with mDEF. (F) H&E staining and (G) insulin detection. [0025]
FIG. 2A-C. mDEF induced anergy in splenic TCR-Tg T cells. (A) Proliferative response of splenic T cells from mDEF-treated and untreated dTg mice after in vitro stimulation with HA(110-120) synthetic peptide (5 μg/ml), HA peptide (5 μg/ml)+rIL-2(20U/ml) or ConA (5 μg/ml). (B) Cytokine production of splenic cells stimulated with HA(110-120) synthetic peptide (5 μg/ml). Data are means ±s.d. from three mice. (C) Flow cytometric analysis of splenic T cells from pools of the same mice used in (A), stained with monclonal FITC-6.5, PE-CD4 and PerCP-CD8. The percentages of CD4+ T cells in the spleens of mice from groups I to IV were 29.9, 27.7, 34.8 and 38.2, respectively, the percentages of CD8+ T cells were 4.1, 3.7, 4.3 and 5.0, respectively, and the percentages of 6.5+ T cells in the spleens and of CD4+ and CD8+ cell subsets are shown No differences were found within the 6.5+ CD4-CD8 T cells in any group (2.6±0.6%). [0026]
FIG. 3. Inhibition of ZAP-70 phosphorylation in mDEF-treated dTg mice. Splenic cells were restimulated for 5 hours with 6.5 mAb or anti-CD3, immunopreciptated with [0027] anti-ZAP 70, then phosphorylation of ZAP-70 was assessed by immunoblotting (upper panels). ZAP-70 was revealed on the same membrane, which was stripped off and reprobed with rabbit anti-ZAP 70 followed by HRP protein A (lower panels). Arrow denotes the position of phosphorylated ZAP-70.
FIG. 4A-F. DEF does not affect the development of TCR-Tg thymic precursors. (A) Proliferative responses of and (B) cytokine production by thymic T cells from untreated or mDEF-treated dTg mice after in vitro stimulation with HA(110-120) synthetic peptide (5 μg/ml). Data were means ±s.d. from three mice. (C-F) Three-color FACS analysis of thymic T cells from mice used in (A), which were stained with FITC-6.5, PE-anti-CD4 and PerCP anti-CD8. (C) Frequency of TCR-Tg (6.5+) T cells in the thymus. (D) Expression of CD4 and CD8 on gated 6.5+ thymic cells. The percentages of cells are indicated in each quadrant. (E,F) Frequency of 6.5+ cells in (E) CD4+CD8+ double positive and (F) CD4+CD8− single positive thymocytes. No significant differences were observed in the expression of 6.5 within the CD4-CD8− double negative (39.0±6.5%) and CD4-CD8+ single positive (26.4±4.5%) thymocyte populations in the four groups of mice. [0028]
FIG. 5A-E. Assessment of diabetogenicity for mDEF-anergized T cells. (A) Splenic CD4+ T cells from untreated dTg mice were either left unpulsed or pulsed in vitro with mDEF (50 μg/ml) and transferred intravenously into HA-Tg-RAG-2[0029] ^−/− mice (4×10⁴cells/mouse, eight mice per group). Changes in (A) blood glucose concentrations and (B) mortaliy rates in the mice were assessed. (C) Histological examination of pancreata from HA-Tg-RAG-2^−/−mice transferred with mDEF-pulsed (lower panels) or unpulsed (upper panels) splenic CD4+ T cells seven days after cell transfer. (D) Negatively sorted splenic CD4+ T cells were left untreated or pulsed for 24 hours with mDEF (50 μg/ml), labeled with CFSE (2 μM) and transferred into HA-Tg-RAG-2^−/−(5×10⁶cells per mouse). Groups of two mice were killed on days 1 and 4 after cell transfer and pancreas-infiltrating cells were stained with clonotypic PE-6.5. (D) Fluorescence intensity of splenic CD4+ T cells labeled with 2 μM of CFSE (filled histogram) and sequential halving CFSE concentrations to determine the CSFE intensity in dividing cells (open histogram). Discontinuous plots showing the fluorescence background of nonlabeled cells. (E) CSFE fluorescence intensity of gated 6.5+ pancreatic T cells from mice killed on day 1 or day 4 after cell transfer. The percentages of 6.5+ T cells in the pancreata are shown.
FIG. 6A-H. mDEF stimulates regulatory IL-10 secreting CD4+ T cells in the pancreas. (A) Proliferative response of and (B) cytokine production by pancreatic T cells from untreated and mDEF-treated dTg mice after in vitro stimulation with HA(110-120) synthetic peptide (5 μg/ml). Data are means ±s.d from three mice. (C,D) Flow cytometric analysis of pancreatic T cells from the mice used in (A), which were stained with FITC-6.5. (C) Expression of 6.5+ T cells within the pancreatic population. (D) Expression of TCR-Tg (6.5+) within the CD4+ poopulation. The percentages of CD4+ T cells in the pancreata of mice from groups I to IV were 44.9, 42.7, 49.6 and 47.3, respectively, and the percentages of CD8+ T cells were 2.9, 3.3, 3.0 and 3.4, respectively. The percentages of 6.5+ T cells are shown. (E) Negatively sorted CD4+ T cells from the pancreata of untreated dTg mice were pulsed with mDEF (4 μg/ml) for 32 hours (monensin was added for the last 6 hours). This was followed by staining with FITC-6.5 and PE-anti-mouse IL-10. Control cells were stained with FITC-rat IgG1 and PE-rat IgG1 isotype controls. The percentages of positve cells in each quadrant are shown. The number of IL-10-secreting CD4+ cells that were not pulsed with DEF was <0.92%. (F) Cumulative incidence of dibetes in HA-Tg-RAG-2[0030] ^−/−mice that received negatively sorted pancreatic CD4+(pCD4+) T cells (1×10⁵cells per mouse) from untreated dTg mice that either had or had not been incubated with mDEF (4 μg/ml) (n=8 mice per group). (G) Proliferative response of splenic TCR-Tg T cells (4×10⁵) from untreated dTg mice cultured with negatively sorted pancreatic CD4+ T cells (1×10⁵) that either had or had not been pulsed with mDEF (4 μg/ml). Anti-mouse IL-10 or rat IgG1 (100 μg/ml) were added to the cultures; data are means ±s.d of triplicate samples. (H) HA-Tg-RAG-2^−/−mice that had received mDEF-pulsed pancreatic CD4+ T cells (1×10⁵cells/mouse, n=12 mice) were challenged one week later with 5×10⁴splenic CD4+ (spCD4+) T cells from untreated dTg mice. Additional groups of mice (n=8) were injected twice a week with neutralizing anti-IL-10 (JES-2A5) or rat IgG1 (300 μg). Controls (n=12) received only 5×10⁴CD4+ splenic T cells.
FIG. 7A-B. mDEF up-regulates the expression of CD62L on pancreatic TCR-Tg T cells and favors migration to the peripheral lymphoid organs. CD8-depleted splenic T cells (10[0031] ⁷) from untreated dTg mice were transferred intravenously into HA-Tg-RAG-2^−/−mice. Mice either were, or were not, injected intravenously with mDEF (2 μg) on days 1, 2 and 4 after cell transfer. (A) Total numbers of TCR-Tg T cells were measured by FACs analysis with FITC-6.5 in the pancreata, spleens and popliteal lymph nodes of mice that were killed 7 days after cell transfer. Data are means ±s.d. from two mice. The total number of 6.5+ T cells in spleens and lymph nodes of control (untreated) mice was <10⁵. (B) Gated pancreatic 6.5+ T cells from the mice used in (A) were examined for the expression of CD62L by FACS analysis with FITC-6.5 and allophycocyanin-anti-mouse CD62L. CD62L expression on donor 6.5+ splenic T cells before transfer as well as CD62L expression on the 6.5+ pancreatic T cells from the adoptively transferred mice that were either untreated or treated with mDEF are shown. The percentage of positive cells are shown; the dashed lines indicate fluorescence background.
FIG. 8A-H. Genetic construction and characterization of human DEF molecules. (A) Genetic construction of human DEF molecules, here as incorporated into p2BAC vector. (3) Schematic representation of the human DEF molecule. (C) nucleotide sequence, as shown by sequencing gel, of DEF-GAD65. (D) nucleotide sequence, as shown by sequencing gel, of DEF-Pi. (E) nucleotide sequence, as shown by sequencing gel, of control DEF without epitope-bearing peptide (for comparison with (C) and (D)). (F) The soluble DEF-GAD65 chimeric molecules were analyzed by SDS-PAGE using 12.5% homogeneous PhastGels. [0032] Lanes 1 to 3 indicate samples separated under non-reducing conditions. Lane 1 contains 14-4-4 mouse IgG2a monoclonal antibody control; lane 2 contains human IgG1 (ZM-2) monoclonal antibody from which the hinge region and Fc domain were used to construct the DEF-DR4_GAD271chimeria; and lane 3 contains the DEF-GAD65 molecule. (G,H) The DEF-DR4_GAD271chimera and the control molecules were separated by SDS-PAGE under non-reducing (lanes 1-6) and reducing (lanes 7-12) conditions, probed with anti-human IgG1 antibody and anti-HLA-DR antibodies, as follows: 14-4-4 mAb ( lanes 1,4,7 and 10), human IgG1 ( lanes 2, 5, 8 and 11), and DEF-GAD65 ( lanes 3, 6, 9 and 12). The DEF-GAD65 chimera was revealed with noth anti-human IgG1, and anti-HLA DR antibodies. Arrows indicate the molecular size (kDa) of the prestained molecular markers (BioRad): myosin (203 kDa), beta-galactosidase (118 kDa), bovine serum albumin (86 kDa), ovalbulin (51.6 kDa), carbonic anhydrase (34.1 kDa) and lysozyme (29 kDa).
FIG. 9A-C. Potency of human DEF molecules in activating murine specific T cells. (A) Lymph node cells (3×10[0033] ⁵) from HLA-DR4 transgenic mice immunized with p271-285 GAD65 peptide were cultured for five dats in the presence of spleen cells (2×10⁵) from naive HLA-DR4 mice and 5, 10 or 15 μg/ml of GAD65 peptide or 0.05, 0.5 or 1 μg/ml of DEF-GAD65. (B) Using the same protocol as for (A), cells from mice immunized withp73-90 proinsulin peptide wee stimulated with the same doses of proinsulin peptide or DEF-Pi. For both (A) and (B), ³H-thymidine (1 μCi) was added for the last 18 hours of culture, then the cells were harvested on Skatron filter papers and the radioactivity was measured in a β-scintillation chamber (Pharmacia LKB, Uppsala, Sweden). The proliferative responses are expressed as Δcpm=specific cpm−background cpm. (C) Cytokines in the cell culture supernatants collected at day 5 were assayed by ELISA following the manufacturer's instructions (Biosource Int., Camarillo, Calif.).
FIG. 10A-D. Human DEF molecules stimulate IL-10 secretion by peptide-specific CD4 T cells from HLA-DR*0401 diabetic patients. PBMC from HLA-DR*0401 patients with [0034] type 1 diabetes were cultured in the presence of either p271-285 GAD65 peptide, p73-90 proinsulin peptide, DEF-GAD65, DEF-Pi, PHA, Human IgG1, or no antigen, for 5 days. (A) ³H-thymidine (1 μCi) was added for the last 18 hours of culture, then cells were harvested on Skatron filter papers and the radioactivity was measured in β-scintillation chamber (Pharmacia LKB, Uppsala, Sweden). Proiliferative response was expressed as the index of proliferation=specific cpm/background cpm. (B) Quantification of IL-10 secretion in cell culture supernatants collected at day 5 of cell culture by ELISA following the manufacturer's instructions (Biosource Int., Camarillo, Calif.). (C) Quantification of IFNγ secretion in cell culture supernatants collected at day 5 of cell culture by ELISA following the manufacturer's instructions (Biosource Int., Camarillo, Calif.). No secretion of IL-2 or IL-4 was detected in the supernatants. (D) Inhibition of the IL-10 response to human DEF molecules by anti-CD4 antibodies. PBMC were stimulated in vitro with DEF-GAD65 or DEF-Pi (5 micrograms per ml) in the presence of 50 micrograms per ml of anti-human CD4 (RPA-T4, Pharmingen), rat IgG1 (Pharmingen), or no antibody/IL-10 was determined in supernatants at day 5 of stimulation.
FIG. 11A-B. Secretion of IL-10 by CD4 T cells on stimulation with DEF-GAD65. PBMC from patient no. 19 were stimulated in vitro with GAD65, DEF-Pi or IgG1 followed followed by three-color staining with anti-human CD3-FITC, CD4-PE and IL-10-APC monoclonal antibodies (Pharmingen). (A) depicts production of IL-10 in cells stimulated by each agent, and shows that 8.1 percent of cells stimulated with DEF-GAD65 produced intracellular IL-10. (B) shows that the majority of DEF-GAD65-induced IL-10-secreting cells were CD3+CD4+ T cells FIG. 12 depicts a reaction scheme for producing the multimeric complexes of the present invention. [0035]
FIG. 13 depicts the separation of multimeric complexes by size exclusion chromatography and the characterization of the multimeric complexes. [0036]
FIG. 14A-D provide results of studies demonstrating the life span of the multimeric complexes in blood circulation and distribution in lymphoid organs. [0037]
FIG. 15A-C are graphs demonstrating thymidine incorporation and cytokine production in TCR-HA T cells exposed to various concentrations of multimers. [0038]
FIG. 16A-D show the results of FACS analysis used to determine the percent of apoptosis of T cells after exposure to multimers.[0039]

5. DETAILED DESCRIPTION OF THE INVENTION

For clarity of description, and not by way of limitation, the detailed description of the invention is divided into the following subsections: [0040]
(i) general structure of HLA/epitope chimeric molecules; [0041]
(ii) specific HLA/epitope chimeric molecules; [0042]
(iii) complexes of HLA/epitope chimeric molecules and [0043]
(iv) clinical uses of HLA/epitope chimeric molecules. [0044]

5.1 General Structure of HLA/Epitope Chimeric Molecules

The basic structure of an HLA/epitope chimera of the invention comprises the following elements: (i) a class II HLA element; (ii) an epitope of interest; and (iii) a means of linking the HLA element and epitope together in an orientation whereby the epitope is presented to a TCR in the context of HLA. Each of these components is more thoroughly discussed in the sections that follow. [0045]
The chimeric molecule of the invention is referred to herein as a protein, however, such “chimeric proteins” as defined herein may comprise non-protein components, including, but not limited to, carbohydrate residues, chemical crosslinking agents, lipids, etc. (see below). [0046]
The class II HLA element comprises at least the minimal structure necessary to interact with both the epitope and a TCR. Generally, this will require the presence of some or all of the extracellular domains of the α and/or β chains of a DP, DQ or DR antigen. If a portion of an α and/or β chain is used, it is particularly desirable to determine that the final construct is capable of selectively binding to a TCR of interest, for example, using fluorescence activated cell sorting as set forth in [0047] section 6, below.
An epitope of interest is an epitope of an antigen toward which an immune response is desireably suppressed. The epitope may be a peptide, may be a molecule comprising amino acids as well as non-amino acid components, or may be a molecule which lacks amino acid residues altogether. The epitope may be comprised in a larger molecule which may or may not further comprise amino acid residues. The epitope may be recognized as implicated in the causation of an autoimmune disease, or it may be a newly discovered epitope. [0048]
The means of linking the HLA element and epitope together permits the association of the two and further permits interaction of both with a TCR. For example, in particular embodiments of the invention, the epitope is linked to the extracellular portion of an α or β class II HLA chain where the HLA chain associates with its partner in the final product—according to such embodiments, it is desirable that the epitope be positioned so that it can associate with both α and β chain components. In one approach, the epitope is attached to a free end of the α or β chain to increase conformational mobility. A linker molecule which improves this mobility may optionally be inserted at the fixed end of the epitope. In another approach, the epitope may be comprised in a linker molecule that joins α and β chain components, designed to permit freedom of movement of the chain components relative to each other. [0049]
In non-limiting embodiments of the invention, the primary structure is a heterodimer comprised of two units. In one subset of embodiments, the first unit comprises an HLA class II α chain component, and the second unit comprises an HLA class II β chain component linked to an epitope of interest. In another, alternative subset of embodiments, the first unit comprises an HLA class II β chain component, and the second unit comprises an HLA class II α chain component linked to an epitope of interest. In a heterodimer formed of the two units, the α and β chains form a noncovalent association in which they both interact with the epitope. [0050]
Preferably such heterodimers are joined together into multimers of at least two heterodimers. This may be accomplished by fusing one or both HLA class II components to a multimerizing structure, which tends to form multimers with similar or dissimilar structures. In particular nonlimiting embodiments, one unit comprises at least a portion (comprising at least one constant region domain) of an immunoglobulin constant region to act as a multimerizing structure. Preferably, an immunoglobulin hinge region or portion thereof may be incorporated between the constant region portion and the HLA class II chain portion. If this is the unit comprising the epitope, the epitope and the immunoglobulin constant region portion are located at opposite ends of the unit relative to each other. Association between HLA partner chain components bring together one pair of units, and the constant region of one of the joined units can be linked to a second constant region from a second pair of joined units, for example by disulfide bonds, as in a native immunoglobulin molecule. The result is a dimer of heterodimers. It is itself referred to herein as a dimer because it consists of two like units bound together and functionally has two HLA/epitope structures for associating with TCRs. A schematic representation of such a dimer (of heterodimers) is depicted in FIG. 8A. [0051]
Such multimers may be joined together to form larger complexes as described in Sections 5.3 and 8, below. [0052]
For example, a chimeric molecule of the invention may be constructed as follows. An HLA class II allele positively associated (causation rather than protection) with autoimmune diabetes, for which nucleic acid sequence of at least extracellular components of the α and β chains is available, may be selected. As each of the α and β chains has two extracellular domains, one proximal to and one distal to the cell membrane in the native molecule, the HLA class II components of the invention preferably comprise at least the membrane distal α and β domains and more preferably comprise all or at least 80 percent, preferably 90 percent of both extracellular domains of each chain. The boundaries of the extracellular domains relative to the transmembrane domain and cytoplasmic domains are known in the art. [0053]
For each of the α and β domain(s) to be incorporated into the chimeric molecule, primers may be designed, using the known nucleotide sequences, to produce DNA encoding the desired regions by reverse transcriptase-polymerase chain reaction (“RT-PCR”) using, as template, mRNA harvested from cells having the same HLA class II genotype. In this way, a cDNA encoding an HLA a chain component and a cDNA encoding a HLA β chain component may be produced. [0054]
DNA encoding the epitope, where the epitope is comprised in a peptide, may be produced using standard techniques, including chemical or recombinant synthesis. Where sequence encoding the epitope is to be fused to an HLA component, such epitope-encoding sequence may be incorporated into a PCR primer used to produce the cDNA encoding the HLA component. [0055]
DNA encoding a multimerizing structure may also be produced using known techniques. For example, where the multimerizing structure is an immunoglobulin constant (Fc) region, cDNA encoding the Fc region may be obtained using either available cloned DNA or by RT-PCR using as template RNA from an appropriate antibody producing cell line. [0056]
DNA encoding the HLA α component, the HLA β component, the epitope, and the multimerizing structure may then be combined in a number of permutations. For example, in a first set of embodiments, DNA encoding the epitope and DNA encoding the multimerizing structure may be linked to opposite ends of the cDNA encoding the HLA β component. According to these embodiments, the α component is comprised in a separate molecule, and binds non-covalently to the β component. The resulting heterodimer may be itself dimerized by association between multimerization structures, for example, between the Fc regions under reducing conditions. [0057]
Alternatively, in a second set of embodiments, DNA encoding the epitope and DNA encoding the multimerizing structure may be linked to opposite ends of the cDNA encoding the HLA β component. Here, the HLA β component is comprised in a separate molecule. [0058]
In a third possible set of embodiments, DNA encoding the epitope is linked to one end of each of the cDNAs encoding the α and β components, and DNA encoding the multimerization structure is linked to the opposite end of one of those cDNAs. [0059]
It may be desirable to incorporate linker regions which enhance the secondary and tertiary structures of the molecule. For example, DNA encoding a flexible linker molecule having small side chains may be interposed between the DNA encoding the epitope and the cDNA encoding the HLA component. Further, DNA encoding a flexible linker, such as an immunoglobulin hinge region, may be interposed between cDNA encoding the multimerization structure and cDNA encoding the HLA component. [0060]
For expression, the foregoing constructs may be operatively linked to a suitable promoter element. It may further be desirable to incorporate other elements that favor expression and/or secretion, such as transcriptional start and stop signals, polyadenylation signals, translational start and stop sites, ribosome binding sites, and leader sequences. Such elements are known in the art. [0061]
The chimeric molecules of the invention may be expressed in bacterial, yeast, insect, or animal cells. [0062]
In preferred, non-limiting embodiments, the chimeric molecules of the invention may be expressed in a baculovirus/insect cell system. For example, chimeric constructs encoding both units of the heterodimer may be inserted, under the control of separate promoters, into the p2BAC system (Invitrogen) (see [0063] Section 8, below).
The chimeric molecules of the invention may further be expressed in eukaryotic cells, such as, but not limited to, mammalian cells, where the foregoing constructs are operably linked to a promoter active in such cells. Furthermore, the foregoing constructs may be used to produce transgenic animals, such as, but not limited to, mice and goats. [0064]
In specific non-limiting embodiments of the invention, a chimeric molecule may be linked to a toxic agent, such as, but not limited to, doxorubicin, ricin, or diptheria toxin; an immunosuppressive agent such as cyclosporin or calicheamicin; a radioactive agent, an MRI contrast agent, or a metal chelate. [0065]

5.2 Specific HLA/Epitope Chimeric Molecules

This section provides non-limiting examples of specific components which may be assembled into HLA/epitope chimeric molecules, as described above, as well as specific HLA/epitope constructs. [0066]
5.2.1 Epitopes [0067]
The following are particular non-limiting examples of peptides and proteins that comprise epitopes relevant to particular autoimmune diseases. Where specific peptides are recited, since the epitopes are contained within the recited peptides, the boundaries of the peptides may be extended by amino acid sequences occuring in the parent protein, for example by at least five amino acids. Further, the following peptides may be comprised in larger peptides, provided that they retain at least some (preferably, at least 25 percent) of the immunogenic activity which they exhibit in the recited peptides. [0068]
The present invention provides for the use of peptide fragments of proteins toward which an autoimmune response is generated in an autoimmune disease. Such peptide fragments are between 5 and 200, preferably between 5 and 100, and more preferably between 5 and 30 amino acids in length. [0069]
Where a peptide is found to comprise an epitope, it is understood that the determination of the minimum amino acid sequence comprising the epitope could be achieved using standard techniques, so that peptide fragments of those recited herein may be used according to the invention. Similarly, it is understood that epitope-containing peptides may be comprised in larger peptides, either naturally occurring or engineered. [0070]
By way of definition, in a subset of embodiments according to the invention, a dimer of heterodimers, formed by epitope linked to HLA class II components, where the dimer is formed by immunoglobulin constant regions linked by disulfide bonds, is referred to as a “DEF” molecule. Where such a molecule is assembled of murine components, it is referred to as “mDEF.”[0071]

Epitopes associated with insulin dependent diabetes mellitus, including type 1 diabetes and LADA, are epitopes of interest according to the invention. One protein known to contain such epitopes is GAD65. The sequence of human GAD65 is set forth in GenBank Accession No. M81882, referencing Bu et al., 1992, Proc. Natl. Acad. Sci. U.S.A. 89:2115-2119. There is a discrepancy of one amino acid between the amino acid sequence in GenBank M81882 and the numbering system used by Patel et al., 1997, Proc. Natl. Acad. Sci. U.S.A. 94:8082-8087. Where peptides were described in Patel, the numbering used in the paper will be maintained; otherwise, residue numbers are according to the GenBank database. Suitable GAD65 epitope-containing peptides include, but are not limited to:


peptide 1-20	(MASPGSGFWSFGSEDGSGDS;;
	SEQ ID NO: 1)

peptide 21-40	(ENPGTARAWCQVAQKFTGGI;,
	SEQ ID NO: 2)

peptide 61-80	(SQPPRAAARKAACACDQKPC;;
	SEQ ID NO: 3)

peptide 116-	(MNILLQYVVKSFDRST;;
130	SEQ ID NO: 4)

peptide 126-	(FDRSTKVIDFHYPNE;;
140	SEQ ID NO: 5)

peptide 207-	(YEIAPVFVLLEYVT;;
220	SEQ ID NO: 6)

peptide 231-	(PGGSGDGIFSPGGAISNMYA;;
250	SEQ ID NO: 7)

peptide 251-	(MMIARFKMFPEVKEKGMAAL;;
270	SEQ ID NO: 8)

peptide 261-	(EVKEKGMAALPRLIAFTSEH;;
280	SEQ ID NO: 9)

peptide 271-	(PRLIAFTSEHSHFSL;;
285	SEQ ID NO: 10)

peptide 281-	(SHFSLKKGAAALGIGTDSVI;;
300	SEQ ID NO: 11)

peptide 311-	(IPSDLERRILEAKQKGFVPF;;
330	SEQ ID NO: 12)

peptide 356-	(KYKIWMHVDAAWGGG;;
370	SEQ ID NO: 13)

peptide 361-	(MHVDAAWGGGLLMSRKHKWK;;
380	SEQ ID NO: 14)

peptide 376-	(KHKWKLSGVERANSV;;
390	SEQ ID NO: 15)

peptide 381-	(LSGVERANSVTWNPHKMMGV;;
400	SEQ ID NO: 16)

peptide 471-	(VDKCLELAEYLYNIIKNREG;;
490	SEQ ID NO: 17)

peptide 481-	(LYNIIKNREGYEMVF;;
495	SEQ ID NO: 18)

peptide 497-	(GKPQHTNVCFWYIPPSLRTLE;;
517	SEQ ID NO: 19)

peptide 506-	(FWYIPPSLRTLED;;
518	SEQ ID NO: 20)

peptide 509-	(IPPSLRTLEDNEERMSRLSK;;
528	SEQ ID NO: 21)

peptide 511-	(PSLRTLEDNEERMSR;;
525	SEQ ID NO: 22)

peptide 521-	(ERMSRLSKVAPVIKARMMEYGTTMVSYQPLGDKVNF
585	FRMVISNPAATHQDIDFLIEEIERLGQDL;;
	SEQ ID NO: 23)

peptide 546-	(SYQPLGDKVNFFRMV;;
560	SEQ ID NO: 24)

peptide 551-	(GDKVNFFRMVISNPA;;
565	SEQ ID NO: 25)

peptide 556-	(FFRMVISNPAATHQD;;
570	SEQ ID NO: 26)

and

peptide 566-	(ATHQDIDFLIEEIER;.
580	SEQ ID NO: 27).

Another protein known to contain epitopes associated with autoimmune diabetes is proinsulin. The sequence of human proinsulin is:


	MALWMRLLPLLALLALWGPDPAAAFVNQHLCGSHLVEALYLVCGERGFF

	YTPKTRREAEDLQVGQVELGGGPGAGSLQPLALEGSLQKRGIVEQCCTSICSL

	YQLENYCN.
	(SEQ ID NO: 28)

	Proinsulin peptides which may be used include, but

	are not limited to:

	peptide 1-18	(MALWMRLLPLLALLALWG;;
		SEQ ID NO: 29)

	peptide 1-24	(MALWMRLLPLLALLALWGPDPAAA;;
		SEQ ID NO: 30)

	peptide 11-25	(LALLALWGPDPAAAFV;;
		SEQ ID NO: 31)

	peptide 20-36	(PAAAFVNQHLCGSHLV;;
		SEQ ID NO: 32)

	peptide 24-36	(AFVNQHLCGSHLV;;
		SEQ ID NO: 33)

	peptide 30-49	(CGSHLVEALYLVCGERGFF;;
		SEQ ID NO: 34)

	peptide 32-46	(also referred to as the “B9-23 peptide

		in the literature)

		(SHLVEALYLVCGERG;
		(SEQ ID NO: 35)

	peptide 35-50	(LVEALYLVCGERGFFY;;
		SEQ ID NO: 36)

	peptide 35-80	(LVEALYLVCGERGFFYTPKTRREAEDLQVGQVELG

		GGPGAGSLQPL;;
		SEQ ID NO: 37)

	peptide 40-59	(YLVCGERGFFYTPKTRREAE;;
		SEQ ID NO: 38)

	peptide 44-60	(GERGFFYTPKTRREAED;;
		SEQ ID NO: 39)

	peptide 44-63	(GERGFFYTPKTRREAEDLQV;;
		SEQ ID NO: 40)

	peptide 52-67	(PKTRREAEDLQVGQVE;;
		SEQ ID NO: 41)

	peptide 63-78	(VGQVELGGGPGAGSLQ;;
		SEQ ID NO: 42)

	peptide 73-90	(GAGSLQPLALEGSLQKRG;;
		SEQ ID NO: 43)

	peptide 74-90	(GERGFFYTPKTRREAED;;
		SEQ ID NO: 44)

	peptide 80-88	(LALEGSLQK;;
		SEQ ID NO: 45)

	peptide 85-101	(SLQKRGIVEQCCTSICS;;
		SEQ ID NO: 46)

	peptide 84-110	(GSLQKRGIVEQCCTSICSLYQLENYCN;;
		SEQ ID NO: 47)

	and

	peptide 97-110	(CTSICSLYQLENYCN;.
		SEQ ID NO: 48)

The present invention also provides for the use of epitopes associated with [0074] type 1 diabetes occuring in p277 from heat shock protein 60 (Wicker et al. (1996) J. Clin. Invest. 98:8082; Horvath et al., 2002, Immunol. Lett. 80:155-162), including but not limited to peptide 437-460.
Yet another protein containing epitopes associated with autoimmune diabetes is islet antigen 2 (“IA-2”), particularly peptide 805-820. Articles of interest regarding IA-2 include Seissler et al., 2002, Horm Metab Res 34:186-191; Farilla et al., 2002, Eur. J. Immunol., PubMed ID# 11981830; Bearzatto et al., 2002, J. Immunol. 168:4202-4208; Williams et al., 2002, Diabetologia 45:217-213; and Horhonen et al., 2002, Diabetes Metab Res. Rev. 18:43-48. [0075]
Proteins containing epitopes relevant to autoimmune diseases other than diabetes include myelin basic protein (“MBP”), proteolipid protein (“PLP”) and myelin oligodendrocyte glycoprotein (“MOG”) (associated with multiple sclerosis); human cartilage glucoprotein 39 and collagen II (associated with rheumatoid arthritis); wheat gliadin (associated with celiac disease); acetyl choline receptor (associated with myasthenia gravis) and thyroid stimulating hormone receptor (autoimmune thyroiditis). [0076]
Specific peptides comprising epitopes of interest relevant to autoimmune diseases other than diabetes include: peptides associated with multiple sclerosis such as PLP 38-49 (ALTGTEKLIETY; SEQ ID NO:49); PLE 91-104 (YTTGAVRQIFGDYK; SEQ ID NO:50); PLP 115-128 (TVTGGQKGRGSRGQ; SEQ ID NO:51; PLP 195-208 (SIGSLCADARMYGV; SEQ ID NO:52) (Kinkel et al., 1992, Neurology 42 (Suppl. 3): 159 (abstr. 87P); and PLP 89-106 (GFYTTGAVRQIFGDYKTT; SEQ ID NO:53); PLP 40-60 (TGTEKLIETYFSKNYQDYEYL:SEQ ID NO:54) (Pelfrey et al., 1993, J. Neuroimmunol. 46:3342; MBP 84-103 (Steimnan et al., 1995, Mol. Med. Today 1:79-83); and p95-116 from PLP (Kawamura et al (2000) J. Clin. Invest. 105: 977-984. [0077]
5.2.2 HLA Class II Molecules [0078]
The following HLA Class II molecules have been associated with autoimmune diabetes, and are therefore of interest. Where a serological group, e.g., DR3 or DR4, is recited, all alleles within that group are encompassed. Where particular alleles are recited, any variations of that allele (referred to herein as “subtypes”) are encompassed. Where specific alleles or subtypes are recited, the disclosure is not to be construed as limited to such alleles or subtypes. [0079]
HLA Class II antigens associated with diabetes which may be used according to the invention include: HLA-DQ, particularly alleles that carry a neutral residue (Val, Ala, or Ser) at position 57 of the DQβ chain, such as the HLA-DQB1*0302 allele; HLA-DR3; HLA-DQw2; HLA-DQw3; HLA-DQw8 and HLA-DR4, for example the alleles HLA-DRB1*0405, *0402, *0401, *0404, *0403, and *0406. Some of these alleles have been further subdivided into subtypes. An updated list of alleles and nucleic acid and sequence information may be obtained via the EMBL-European Bioinformatics Institute linked ImMunoGeneTics database, having a web address at www.ebi.ac.uk/imgt/index.html. Information relating to the nucleic acid and amino acid sequences of particular alleles are as follows: HLA-DQB1*0302 has IMGT/HLA No. HLA00627, with the following GenBank accession numbers: K01499, L34097, M25326 and M65038; HLA-DRB1*0405 is referenced by IMGT by three subtypes, HLA-DRB1*04051, HLA-DRB1*04052, and HLA-DRB1*4053, where HLA-DRB1*04051 has IMGT/HLA No. 00690, with GenBank accession numbers AF029271, L13875 and M15070, HLA-DRB1*04052 has IMGT/HLA No. 00691, with GenBank accession numbers D49952 and D50889, and HLA-DRB1*04053 has IMGT/HLA No. 01551; and HLA-DRB1*0401 is referenced by IMGT as two subtypes, HLA-DRB1*04011 and HLA-DRB1*04012; where HLA-DRB1*04011 has IMGT/HLA No. 00685, with GenBank accession numbers AF029267, M20541, K02776, M20550, M17381 and M20548 and HLA-DRB1*04012 has IMGT/HLA No. 00686, with GenBank accession number X96851. HLA Class II molecules identified in the future as associated with autoimmune diabetes may also be used according to the invention. [0080]
HLA Class II molecules associated with authoimmune diseases other than diabetes which are of interest and may be used according to the invention include: HLA-DR2 (associated with multiple sclerosis); HLA-DR4, HLA-DQw3 and HLA-DR1, HLA-DQw1 and HLA-D QKRAA+ (associated with rheumatoid arthritis); HLA-DR2, HLA-DQw1 (associated with multiple sclerosis); HAL-DQ2, HLA-DQ8, and HLA-DR3, HLA-DQw2 and HLA-DR7, HLA-DQw2, especially HLA-DQ2 (α1*0501, β1*0201) (associated with celiac disease); HLA-DR4, HLA-DQw3, and HLA-DR6, DQw (associated with pemphigus vulgaris); HLA-DR8 and HLA-DR5 (associated with pauciarticular juvenile rheumatoid arthritis); HLA-DR3, HLA-Dqw2, and HLA-DR2, HLA-DQw1 (associated with systemic lupus erythematosis); HLA-DR3 (associated with Sjogren's syndrome); HLA-DR2, HLA-DQw1 (associated with narcolepsy), HLA-DR3, HLA-DQw2 (associated with Graves' disease); and HLA-DR3, HLA-DQw2 (associated with dermatitis herpetiformis). [0081]
Because certain associations have been drawn between specific HLA Class II and immune reactivity toward particular epitopes of interest, it is desireable to combine such HLA class II elements with said epitopes. Specific non-limiting embodiments of combinations of HLA element and epitope are as follows. In autoimmune diabetes, epitopes associated with HLA-DR4 are GAD65 peptides 271-285, 506-518, 241-255, 556-575 and 176-195, preproinsulin peptide 24-36, and islet antigen-2 peptide 805-820; and epitopes associated with HLA-DQ8 are GAD65 peptides 201-220, 231-250, 247-266, 509-528, and 253-265, proinsulin peptides 32-46, 1-24 and 44-63, and hsp60 peptide 437-460. In rheumatoid arthritis, epitopes associated with HLA-DR4 include collagen type II peptide 260-273 and epitopes associated with HLA-DQ8 include collagen type II peptide 554-573. [0082]
5.2.3 Specific Constructs [0083]
The following are specific non-limiting examples of HLA/epitope constructs according to the invention. [0084]
A chimeric molecule of the invention may be constructed from HLA components wherein an α component derives from HLA-DRA1*0101 and a β component derives from HLA-DRB1*0401 as follows. The complete sequences of these particular α and β chains are known in the art, and are available via, respectively, in Steven et al., 1995, Tisue Antigens 45:258-280 and IMGT/1HLA No. HLA00685. See also NCBI Sequence Viewer (www.ncbi.nim.nig.org). At least 80 percent and preferably 90 percent of the native extracellular α and β chains are desirably utilized in the chimeric molecules. cDNAs encoding these molecules may be produced, using, as template, mRNA from an HLA-DRB1*0401 cell line, and using as primers nucleic acid sequence from [0085] regions 5′ of α and 3′ of α to produce cDNA encoding the HLA α component (e.g., regions 5′ of α, +1 to +6, 5′-ATC AAA GAA GAA CAT GTG -3′ (SEQ ID NO:57) and, 3′ of α, +173 to +178, 5′-CCA GTG CTT GAG AAG AGG-3′ (SEQ ID NO:58)) and from regions 5′ of β and 3′ of β to produce cDNA encoding the HLA β component (e.g., regions 5′ of β, +1 to +6, 5′-GGG GAC ACC CGA CCA CGT-3′ (SEQ ID NO:59) and, 3′ of β, +183 to +188, 5′-CCA TTC CAC TGT GAG AGG-3′ (SEQ ID NO:60) (Steven et al., 1995, Tissue Antigens 45:258-280).
DNA encoding an epitope associated with autoimmune diabetes, such as, but not limited to, one of the above listed GAD65 or proinsulin peptides, may then be fused to the cDNA encoding the α or β component prepared as described in the preceding paragraph. Alternatively, said DNA may be incorporated into the PCR primers used to generate said DNAs. Preferably, DNA encoding a low steric hindrance linker molecule (e.g., a peptide comprising a majority of amino acids having small side chains such as glycine and alanine) is interposed between the cDNA encoding the HLA component and the DNA encoding the epitope; flexibility, and consequent interaction with HLA components, may be optimized by changing the composition or length of the linker. Preferably, DNA encoding a peptide comprising one of the following peptides may be used: peptide 271-285, 116-130 or 551-565 of GAD65 or peptide 73-90 of proinsulin. [0086]
As generally discussed in Section 5.1, a cDNA encoding an HLA component such as an HLA-DRA1*0101 or HLA-DRB1*0401 component may be linked to cDNA encoding a human immunoglobulin constant region. In one embodiment, the cDNA encoding the immunoglobulin constant region and DNA encoding a peptide comprising the epitope may be linked on either side of a cDNA enclosing an HLA component. As a specific, non-limiting example, a construct may comprise DNA encoding immunoglobulin constant region/immunoglobulin hinge region/HLA-DRB1*0401 β chain extracellular domain/flexible linker/peptide comprising GAD65 or proinsulin epitope. This construct may be operably linked to a suitable promoter element, and used to express one unit of a heterodimer; the other unit may be expressed from a construct comprising cDNA encoding HLA-DRA1*0101 operably linked to a suitable promoter element. The translation products of each of these constructs may be linked to a leader sequence to provide for secretion, depending upon the expression system used. The units may be expressed separately and then combined to form heterodimer or, preferably, may be coexpressed. A schematic diagram of one non-limiting example of a heterodimer produced by such constructs, comprising either GAD65 271-285 peptide or proinsulin 73-90 peptide, is depicted in FIG. 8B. [0087]
In a second series of specific embodiments, a chimeric molecule of the invention may be constructed using an ax component derived from HLA-DQA1*0301 and a β component derived from HLA-DQB1*0302 as follows. The sequences of the α and β chains are available (Steven et al., 1995, Tissue Antigens 45:258-280). At least 80 percent and preferably 90 percent of the native extracellular α and β chains are desirably utilized in the chimeric molecules. cDNAs encoding these molecules may be produced, using, as template, mRNA from an HLA-DQB1*0302 cell line, and using as primers nucleic acid sequence from [0088] regions 5′ of α and 3′ of α to produce cDNA encoding the HLA β component (e.g., regions 5′ of α, +1 to +6, 5′-GAA GAC ATT GTG GCT GAC-3′ (SEQ ID NO:61) and 3′ of α, +171 to +176, 5′ AGG CTC ATC CAG GCC CCA-3′ (SEQ ID NO:62)) and from regions 5′ of β and 3′ of β to produce cDNA encoding the HLA β component (e.g., regions 5′ of β, +1 to +6, 5′ AGA GAC TCT CCC GAG GAT -3′ (SEQ ID NO:63), and 3′ of β, +171 to +176, 5′-CTC CAC GTG GCA GGT GTA-3′ (SEQ ID NO:64)) (Steven et al., 1995, Tissue Antigens 45:258-280).
DNA encoding an epitope associated with autoimmune diabetes may then be linked to one of the α or β component molecules produced according to the preceding paragraph, optionally with a linker peptide between. DNA encoding peptides comprising the following peptides are preferred: the insulin B9-23 peptide (proinsulin peptide 32-46) and GAD65 peptides 126-140, 207-220, 497-517, and 537-557. Other sequences needed for expression and dimerization, including promoter sequences, Fc-encoding sequences, hinge region-encoding sequences, etc., may be incorporated as set forth in examples supra. [0089]

5.3 Complexes of HLA/Epitope Chimeric Molecules

The chimeric molecules of the invention are preferably supplied as multimers (at least two) of heterodimers as described in Sections 5.1 and 5.2. A dimer may be produced by incorporating immunoglobulin constant region into at least one unit of the heterodimer. Higher multimers may be achieved by crosslinking such heterodimers, where crosslinking may be accomplished using agents and methods known in the art. [0090]
In preferred embodiments of the invention, heterodimers comprising an immunoglobulin constact region may be covalently linked through a carbohydrate residue of the immunoglobulin region element by a polyalkylene glycol linker. [0091]
Polyalkylene glycols and the use thereof to modify proteins are known in the art. In accordance with the present invention, a difunctional polyalkylene glycol, preferably a difunctional polyethylene glycol (PEG), is used to cross link the chimeric molecules via a carbohydrate residue on the immunoglobulin constant region element. [0092]
PEG (HO—(CH[0093] ₂CH₂O)_n—CH₂CH₂—OH) is derivatized at both termini to provide activated difunctional PEG. Preferably PEG is modified to contain primary amino groups at both termini to provide diamino-PEG (NH₂—(CH₂CH₂O)_n—CH₂CH₂—NH₂). The molecular weight of the PEG is not limited but is preferably from 1000-5000, and more preferably about 3500. Diamino-PEG having a molecular weight of 3400 is commercially available from Shearwater Corporation (Huntsville, Ala.) and is particularly preferred in accordance with the present invention.
The chimeric molecules may be covalently linked through a carbohydrate residue of the immunoglobulin constant region by an enzymatic method as follows. The carbohydrate residue of the immunoglobulin constant regions of the chimeric molecules is desialylated by incubation with one or more neuraminidases. Free sialic acid (N-acetyl neuraminic N-acetyl acid, NANA) may be removed for example by dialysis. The desialylated molecules are reacted with galactose oxidase to oxidize the terminal galactose (Gal) residues, and with diamino-PEG. This reaction results in the formation of Schiff bases between the C6-aldehyde groups of the Gal residues and the amino terminal groups of the diamino-PEG. The Schiff bases may be stabilized by reductive alkylation, for example with pyridine borane. The resulting multimeric complexes may be purified by methods known in the art, including for example dialysis and size exclusion chromatography. Particularly preferred multimers, such as tetramers and octamers, may be identified and isolated by size exclusion chromatography. [0094]
In another embodiment, the present invention provides a method of making multimeric complexes of at least two chimeric molecules wherein the chimeric molecules comprise an immunoglobulin constant region element and two MHC elements associated with a peptide of interest, and wherein at least two of the chimeric molecules are covalently linked through a carbohydrate residue of the immunoglobulin constant region by a polyalkylene glycol linker. [0095]
The method comprises contacting the chimeric molecules with one or more neuraminidases under conditions to desialylate carbohydrate residues of the immunoglobulin constant region element to provide desialylated chimeric molecules, and contacting the desialylated chimeric molecules with galactose oxidase and diamino-polyalkylene glycol under conditions to covalently link two or more chimeric molecules. [0096]
In a preferred embodiment, the one or more neuraminidases are a mixture of neuraminidases from [0097] Arthrobacter ureafaciens and Clostridium perfringens. Following desialylation, free sialic acid may be removed, for example by dialysis.
In another preferred embodiment the diamino-polyalkylene glycol is diamino-polyethylene glycol. In a particularly preferred embodiment, the diamino-polyethylene glycol has a molecular weight of about 3400 ((NH[0098] ₂)₂-PEG_3,400, Shearwater Corporation, Huntsville, Ala.).
In another preferred embodiment, mild reducing conditions, for example treatment with pyridine borane, are used to stabilize the Schiff bases formed between the amino groups of PEG and the terminal galactose residues of the carbohydrate residues of the immunoglobulin constant region. [0099]
In another preferred embodiment, the resulting multimeric complexes are purified by methods known in the art, including for example size exclusion chromatography. [0100]
Conditions for desialylation and for derivatizing immunoglobulins with a monofunctional PEG derivative are disclosed in International Patent Application Publication Nos. WO96/40731 and WO96/36357, which disclosures provide reaction conditions and parameters that may be used in the practice of the present method. WO96/40731 and WO96/36357 are incorporated herein by reference. [0101]
In a particularly preferred embodiment, the present method comprises incubating the chimeric molecules overnight at 37° C. with neuraminidase from [0102] Arthrobacter ureafaciens and Clostridium perfingens (Calbiochem-Novobiochem Intern. Inc., La Jolla, Calif.) in 0.1 M phosphate buffer pH 5.5 containing 5 mM CaCl₂. Free sialic acid released by neuraminidase is removed by dialysis against PBS pH 7.4. Desialylated chimeric molecules are incubated for 48 hours at 37° C. with galactose oxidase (GAO, Sigma Chemical Co., St. Louis, Mo.) and diamino-polyethylene glycol bifunctional cross linker with a molecular mass of 3,400 Da ((NH₂)₂-PEG_3,400, Shearwater Corporation, Ala.). The Schiff bases formed between the aldehyde groups generated by GAO at the 6^thcarbon of terminal galactose residues and the amino groups of PEG are stabilized on mild reduction with 80 mM of pyridine borane (PB)(Aldrich). The reaction mixture is dialyzed against phosphate buffered saline (PBS) in SPECTRA/POR bags (100,000 MWCO, Sigma), and multimers are separated by size exclusion chromatography.
In alternative embodiments of the invention, if lysines and sulfhydryl groups are available for cross-linking, one may consider using a heterobifunctional amine/sulfhydryl reactive agent such as AMAS, BMPS, EMCS, sulfo-EMCS, GMBS, sulfo-GMBS, sulfo-KMUS, MBS, sulfo-MBS, SBAP, SIA, SIAB, sulfo-SIAB, SMCC, LC-SMCC, SMPB, SMPH, sulfo-SMPB, SVSB, BMPA, EMCA, KMUA, SMPT, sulfo-LC-SMPT, SPDP, LC-SPDP, and sulfo-LC-SPDP. Additional information may be found in the Pierce pamphlet and/or in Hermanson, 1995, “Bioconjugate Technologies”, Academic Press, Inc., Pierce Product #20002GJ, and Wong, 1991, “Chemistry of Protein Conjugation and Cross-Linking,” CRC Press,Inc., Pierce Product No. 15010GJ. [0103]

5.4 Clinical Uses of HLA/Epitope Chimeric Molecules

The HLA/epitope chimeric molecules of the invention may be used for diagnosing and treating autoimmune diseases including, but not limited to, diabetes (juvenile and LADA), rheumatoid arthritis (adult and juvenile), celiac disease, systemic lupus erythematosis, multiple sclerosis, Sjogren's syndrome, pemphigus vulgaris, dermatitis herpetiformis, Graves disease, and narcolepsy. [0104]
The present invention provides for a method of diagnosing an autoimmune disorder or identifying a predisposition to an autoimmune disorder in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric molecule which comprises HLA class II components and an epitope which has a positive correlation with the autoimmune disease, where the HLA class II components in the chimeric molecule are compatible with HLA class II determinants of the subject. Preferably, the chimeric molecule is provided as a multimer. The presence of such T cells may be considered to positive correlate with the existence of or a predisposition to the autoimmune disease asssociated with said epitope in the subject, and to indicate that the subject may benefit from treatment with such chimeric molecules to ameliorate or prevent the disease. [0105]
The T cells may be detected in vivo or in vitro from a sample taken from the subject. Preferably, the T cells are collected as peripheral blood mononuclear cells from a blood sample collected from the subject. [0106]
The chimeric molecules are heterodimers and multimers thereof, prepared as set forth in the foregoing sections. [0107]
The ability of a chimeric molecule to selectively bind to a T cell of a patient may be determined by standard laboratory methods which detect T cell/molecule binding. As one example, binding of the chimeric molecule to a T cell may be detected by detectably labeling a component of the chimeric molecule, and then detecting the presence of the label, e.g., labeling the chimeric molecule with a radioactive label such as [0108] ³⁵S-methionine. As a second example, binding of a chimeric molecule to a T cell may be detected using a detectably labeled antibody that binds to the chimeric molecule; e.g., where the chimeric molecule comprises an immunoglobulin constant region, binding can be detected using detectably labeled anti-Ig constant region antibody. The binding of said labeled antibody could be detected, where the label is a fluorescent label, by fluorescent microscopy or fluorescence activated cell sorting (“FACS”; see Section 7, below).
It may be desirable to determine that the binding of a chimeric molecule to a T cell is selective. For example, T cells bound to chimeric molecule collected by FACS could then be tested for their ability to proliferate or secrete cytokine in the presence of chimeric molecules or the a nominal peptide containing the same epitope as is comprised in the chimeric molecules. Methods for determining proliferation and cytokine secretion are set forth in the following sections. [0109]
The HLA class II components in the chimeric molecule are considered to be compatibile with HLA class II determinants of the subject if they are of the same or a similar serotype. For example, a chimeric molecule comprising components of a particular allele of serotype HLA-DR4 can be used in diagnostic methods of subjects having the same alele or another HLA-DR4 allele; in other words, any subject having the HLA-DR4 serotype. Preferably, the HLA components of the chimeric molecule are as similar as possible to HLA class II determinants carried by the subject to be tested. [0110]
According to one set of particular non-limiting embodiments, the present invention provides for a method of diagnosing autoimmune diabetes or identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric molecule which comprises HLA class II components and an epitope of a protein selected from the group consisting of GAD65 and proinsulin, where the HLA class II components in the chimeric molecule are compatible with HLA class II determinants of the subject. Preferably the HLA class II determinants are selected from the group consisting of HLA-DQB1*0302, HLA-DR3, HLA-DQw2, HLA-DQw3, HLA-DQw8, and HLA-DR4. The presence of such T cells may be considered to positively correlate with the existence of or a predisposition to autoimmune diabetes, and to indicate that the subject may benefit from treatment with such chimeric molecules to ameliorate or prevent that disease. [0111]
In one specific non-limiting embodiment, the subject has an HLA-DR4 determinant and the chimeric molecule comprises HLA-DR4 components and a GAD65 epitope associated with autoimmune diabetes, such as, but not limited to, peptide 271-285, peptide 116-130, or peptide 551-565. [0112]
In another specific non-limiting embodiment, the subject has an HLA-DR4 determinant and the chimeric molecule comprises HLA-DR4 components and a proinsulin epitope associated with autoimmune diabetes, such as, but not limited to, peptide 73-90. [0113]
In yet another specific embodiment, the subject has an HLA-DQ8 determinant and the chimeric molecule comprises HLA-DQ8 components and a proinsulin epitope associated with autoimmune diabetes, such as, but not limited to, peptide 32-46. [0114]
Where a predisposition is present, the present invention provides for methods of preventing an autoimmune disease in an individual comprising administering, to a subject having a predisposition, a preventative amount of a chimeric molecule comprising HLA class II components and an epitope which has a positive correlation with the autoimmune disease. An amount of chimeric molecule is considered to be “preventative” if it stabilizes or slows progression of clinical markers and/or suppresses immunoresponsiveness, e.g., via IL-10 secretion (see below). In a preferred embodiment, the autoimmune disease which is prevented is diabetes. A delay in the onset of disease is considered prevention, even where “prevention” is not absolute (permanent). [0115]
The present invention further provides for methods of treating a human subject suffering from an autoimmune disorder comprising administering, to the subject, an effective amount of a chimeric molecule comprising HLA class II components and an epitope which has a positive correlation with the autoimmune disease, where the HLA class II components in the chimeric molecule are compatible with HLA class II determinants of the subject. Preferably, the chimeric molecule is provided as a multimer. Preferably, the HLA components are sufficiently similar so as not to provoke an undesirable immune reaction in the subject. [0116]
An effective amount of chimeric molecule is an amount which, when administered to the subject, decreases baseline glucose levels by at least 25 percent within one week and/or increases IL-10 levels by at least two-fold within one week and/or induces anergy in the subject's T cells, defined herein as a decrease, relative to pretreatment levels, in the in vitro proliferative response to the administered epitope by at least 30% within 3 days on administration of chimeric molecule. [0117]
A mixture of chimeric molecules comprising different HLA class II component and/or different epitope-containing peptides may be administered, for example where the individual is heterozygous for HLA class II alleles and each allele is associated with a predisposition for autoimmune disease. [0118]
Such treatment may also be effected ex vivo, by subjecting T cells collected from the subject to an effective concentration of said chimeric molecule. An effective concentration is an amount which increases the amount of IL-10 by treated T cells by at least two-fold within one week, and/or which, when said T cells are reintroduced into the subject, decreases baseline glucose levels by at least 25 percent within one week, and/or which induces anergy, as defined above, in the subject's T cell response toward the administered epitope. [0119]
The chimeric molecules of the invention may be administered in a suitable pharmaceutical carrier. The chimeric molecules of the invention may be stored frozen and/or in lyophilized form prior to use. The chimeric molecules of the invention may be comprised in formulations which further include stabilizing agents. [0120]
In specific, non-limiting embodiments of the invention, the dosage of chimeric molecules may be such as to produce a local concentration of between about 0.01 and 10 μg/ml, and preferably between about 0.05 and 5 μg/ml. In specific nonlimiting embodiments of the invention, the dosage per kilogram may be between about 0.75 and 750 μg/kg, and preferably between about 3.75 and 375 μg/kg, and/or, for a human subject, between about 50 μg and 50 mg, and preferably between about 0.25 μg and 25 mg or between about 0.25 and 5 mg. The compositions of the invention may be administered to a subject orally, subcutaneously, intramuscularly, intravenously, intraarterially, intravaginally, intrarectally, intraperitaoneally, intrathecally, topically, or by inhalation. Sustained release formulations, including tissue or organ implants, may also be used. [0121]
The effectiveness of said treatment methods may be improved by using chimeric molecules comprising HLA/epitope pairs which have been correlated with the autoimmune disease. Further, it may be determined whether the chimeric molecules bind to T cells of the subject to be treated, using methods as set forth above. [0122]
Preferably, the treatment is instituted prior to or early in the course of the disease. Furthermore, preferably the treatment regimen comprises repeated individual treatments, at a frequency of preferably at least once a week, more preferably at least every 5 days. The interval between doses may be adjusted based upon the responsiveness of the subject to treatment. [0123]
According to one set of particular non-limiting embodiments, the present invention provides for a method of treating autoimmune diabetes in a human subject, in need of such treatment, having an HLA class II determinant selected from the group consisting of HLA-DQB1*0302, HLA-DR3, HLA-DQw2, HLA-DQw3, HLA-DQw8, and HLA-DR4, comprising administering, to the subject, an effective amount of a chimeric molecule which comprises HLA class II components and an epitope of a protein selected from the group consisting of GAD65 and proinsulin, where the HLA class II components in the chimeric molecule are compatible with HLA class II determinants of the subject. Preferably, the chimeric molecules of the invention are provided in multimeric (i.e., at least dimeric) form. [0124]
In one specific non-limiting embodiment, the subject to be treated has an HLA-DR4 determinant and the chimeric molecule comprises HLA-DR4 components and a GAD65 epitope associated with autoimmune diabetes, such as an epitope of peptide 271-285, peptide 116-130, and or peptide 551-565. [0125]
In another specific non-limiting embodiment, the subject to be treated has an HLA-DR4 determinant and the chimeric molecule comprises HLA-DR4 components and a proinsulin epitope associated with autoimmune diabetes, such as peptide 73-90. [0126]
In yet another specific embodiment, the subject to be treated has an HLA-DQ8 determinant and the chimeric molecule comprises HLA-DQ8 components and a proinsulin epitope associated with autoimmune diabetes, such as, but not limited to, as contained in peptide 32-46 (insulin B9-23 peptide). [0127]
The present invention further provides for methods of treating a human subject suffering from an autoimmune disorder comprising administering, to the subject, an effective amount of a chimeric molecule comprising HLA class II components, an epitope which has a positive correlation with the autoimmune disease, and a toxic agent, where the HLA class II components in the chimeric molecule are compatible with HLA class II determinants of the subject. Preferably, the chimeric molecule is provided as a multimer. [0128]
According to these methods, the chimeric molecule specifically targets autoimmune reactive T cells for delivery of the toxic or immunosuppressive agent. The agent may be, for example, doxorubicin, ricin, cyclosporin, calicheamicin, a radioactive compound or diptheria toxin. [0129]
According to one set of particular non-limiting embodiments, the present invention provides for a method of treating autoimmune diabetes in a human subject, in need of such treatment, having an HLA class II determinant selected from the group consisting of HLA-DQB1*0302, HLA-DR3, HLA-DQw2, HLA-DQw3, HLA-DQw8, and HLA-DR4, comprising administering, to the subject, an effective amount of a chimeric molecule which comprises HLA class II components, an epitope of a protein selected from the group consisting of GAD65 and proinsulin, and a toxic agent such as, but not limited to, doxorubicin, where the HLA class II components in the chimeric molecule are compatible with HLA class II determinants of the subject. Preferably, the chimeric molecules of the invention are provided in multimeric (i.e., at least dimeric) form. In one specific non-limiting embodiment, the subject to be treated has an HLA-DR4 determinant and the chimeric molecule comprises HLA-DR4 components and a GAD65 epitope associated with autoimmune diabetes, such as peptide 271-285. In another specific non-limiting embodiment, the subject to be treated has an HLA-DR4 determinant and the chimeric molecule comprises HLA-DR4 components and a proinsulin epitope associated with autoimmune diabetes, such as peptide 73-90. In still further non-limiting embodiments, the subject to be treated has a HLA-DQ8 allele and the chimeric molecule comprises HLA-DQ8 components and proinsulin peptide 32-46 (insulin B 9-23); or the subject has both HLA-DR4 and HLA-DQ8 alleles and is treated with a mixture of chimeric molecules including chimeric molecules comprising HLA-DR4 components and the GAD65 peptide 271-285 and chimeric molecules comprising HLA-DQ8 components and the proinsulin peptide 32-46. [0130]

6. EXAMPLE

Down-Regulation of Diabetogenic CD4⁺ T Cells by a Soluble Dimeric Peptide-MHC Class II Chimera

6.1 Materials and Methods

6.1.1 Mice [0131]
TCR-Tg (BALB/c, H-2[0132] ^d) mice expressed the 14.3 d HA-specific TCR, which recognized the HA(110-120) epitope of A/PR/8/34 influenza virus in association with I-E^d. HA-Tg (B10.D2, H-2^d) mice expressed the HA protein from the same virus in pancreatic β cells under the control of the rat insulin promoter. The TCR-Tg and HA-Tg mice were crossed to generate dTg mice (Radu et al., 1999, Autoimmunity 30:199-207). HA-Tg RAG-2^−/−mice on a B10.D2 background were as described in Sarukan et al., 1998, Immunity 6:563-570. Mice were housed in pathogen-free conditions and were used according to the guidelines of the Institutional Animal Care Committee at Mount Sinai School of Medicine.
6.1.2 Antigens [0133]
The mDEF molecule comprised I-E[0134] ^dα and I-E^dβ extracellular domains that were dimerized through an Fcγ2a fragment of the COOH terminus of I-E^dβ (Casares et al., 1997, Protein Engineer. 10:1295-1301). In addition, the HA(110-120) (SFERFEIFPKE; SEQ ID NO: 55) CD4+ T cell epitope from the A/PR/8/34 influenza virus and (GGGS)₃linker (SEQ ID NO: 56) were covalently linked to the NH₂terminus of I-E^dβ. Recombinant mDEF protein was produced in a baculovirus SF9 insect cell system and purified by affinity chromatography on a goat anti-γ2a sepharose column.
The HA(110-120) synthetic peptide was prepared with solid-phase Fmoc technology and purified by reverse phase HPLC on a C2/C18 column (Amersham-Pharmacia Biotech, Piscataway, N.J.). The purity of the synthetic peptide was assessed by amino acid sequencing at the Protein Core Facility at Mount Sinai School of Medicine. MOPC-173 myeloma cells secrete an IgG2a that was used as isotype control for the FCγ2a portion of mDEF. [0135]
6.1.3 Treatment of dTG Mice with Antigens [0136]
Fifteen day-old prediabetic and diabetic mice were injected in the tail vein with mDEF or MOPC-173 (2 μg) or with HA(110-120) peptide (30 ng or 3 μg) in saline (0.2 ml) every 5 days up to the age of 5 months. At the age of 15 days, the HA-Tg littermates were injected intravenously with mDEF (2 μg) every 5 days up to the age of 5 months. [0137]
6.1.4 Diabetes Monitoring [0138]
The blood glucose concentrations in mice were monitored with the Accu-Chek Advantage (Boehringer Mannheim, Indianapolis, Ind.). Mice were considered diabetic when glycemia was >200 mg/dl after two consecutive measurements. [0139]
6.1.5 Histopathological Analysis [0140]
Pancreata were fixed in a 10% solution of buffered formalin, embedded in paraffin, and then sections were cut in stair-wise (7 μm per section). Staining was done using the Mayer hematoxylin-eosin (H&E) technique. For each organ, ten sections were analyzed. Staining for intracellular insulin was done with polyclonal rabbit anti-insulin (Santa Cruz Biotechnologies, Santa Cruz, Calif.) and revealed with a horseradish peroxidase (HRP)-goat anti-rabbit conjugate (Southern Biotechnologies, Birmingham, Ala.). [0141]
6.1.6 Tissue Preparation [0142]
Single cell suspensions of splenic, pancreatic and thymic cells were obtained as described in Radu et al., 1999, Autoimmun. 30:199-207. [0143]
6.1.7 Cell Purification [0144]
Negatively sorted CD4+ T cells were obtained from the spleens and the pancreata of dTg mice with the use of a mouse CD4 subset column kit (R&D Systems, Minneapolis, Minn.). Depletion of CD8+ T cells from spleens was done with magnetic Dynabeads coupled to rat anti-mouse CD8 (Dynal, Oslo, Norway), following the manufacturer's instructions. Labeling of CD4+ T cells with 2 μm of CFSE was done with the Vybrant cell tracer kit (Molecular Probes, Eugene, Oreg.), following the manufacturer's instructions. [0145]
6.1.8 T Cell Proliferation Assays and Cytokine Production [0146]
The T cell proliferative responses to HA peptide were measured by analysis of [0147] ³H-thymidine incorporation as described in Casares et al., 1999, J. Exp. Med. 190:543-553. Briefly, 5×10⁵cells were stimulated in vitro with HA peptide (5 μg/ml) or ConA (5 μg/ml) for 4 or 2 days, respectively. ³H-thymidine (1 μCi) was added for the last 18 hours of culture. The proliferative response (Δcpm) was calculated as specific cpm−background cpm. After 3 days of culture, cytoline (IL-2, IL-4, IFN-γ and IL-10) concentrations in the cell culture supernatants were quantified with the use of ELISA kits (Biosource International, Camarillo, Calif.) following the manufacturer's instructions.
6.1.9 Flow Cytometry [0148]
The following monoclonal antibodies were used for FACS analysis: fluorescein isothiocyanate (FITC-)-rat anti-mouse 6.5 (anti-TCR-Tg clonotypic), phycoerythrin (PE)-6.5, PE-rat anti-mouse CD4 (Pharmingen, San Diego, Calif.), peridinin chlorophyll protein (PerCP)-rat anti-mouse CD3 (Pharmingen), PE-anti-mouse CD3 (Pharmingen), FITC anti-mouse IgG2a (Boehringer Mannheim, Indianapolis, Ind.), PE-anti-mouse IL-10 (JES-2A5, Pharmingen), and isotype controls conjugated to FITC, PE, PerCP or allophycocyanin (Pharmingen). Intracellular staining of IL-10 was done with the Cytofix/cytoperm plus kit with Golgistop (Pharmingen), following the manufacturer's instructions. [0149]
6.1.10 Signal Transduction [0150]
ZAP-70 phosphorylation was determined in splenic T cells (20×10[0151] ⁶) from mDEF-treated or untreated dTg mice that were collected 1 day after treatment, and from splenic T cells from untreated animals that were pulsed, or not, in vitro for 18 hours with mDEF (μg/ml). Cells were restimulated with plastic-immobilized 6.5 clonotypic mAb or soluble 2C11 anti-CD3ε (10 μg/ml) for 5 hr. Total cell lysates were precipitated with protein A/G agarose beads coated with rabbit anti-ZAP-70 IgG (Santa Cruz Biotechnology), and the protein A/G agarose immunoprecipitate was separated by SDS-PAGE under denatuing and reducing conditions with 12.5% homogeneous PHAST gels (Amersham-Pharmacia). Gels were electrotransferred on PVDF membranes, blocked overnight with 3% bovine serum albumin (BSA), probed for 24 hours with HRP-RC-20P anti-phosphotyrosine (Transduction Laboratories, San Diego, Calif.) and the HRP was developed by chemoluminescence with the Lumi-Light plus kit (Boehringer Mannheim). Membranes were stripped off by acidic treatment with glycine-HCl, at pH 2.8, for 30 minutes at room temperature, blocked with 3% BSA, reprobed with rabbit anti-ZAP-70, incubated with HRP-protein A (Sigma Chemicals, St. Louis, Mo.) and the HRP enzyme developed by chemoluminescence.
6.1.11 Adoptive Transfer Experiments [0152]
Negatively sorted CD4+ splenic or pancreatic T cells from untreated dTg mice were cultured for 24 hours in the presence of mDEF or no antigen, then washed and infused in HA-Tg-RAG-2[0153] ^−/−mice. Some of the mice that had received pancreatic CD4+ T cells were challenged one week later with CD4+ splenic T cells. In a second set of experiments, HA-Tg-RAG2^−/−mice were infused with CD8-depleted splenic T cells from untreated dTg mice. The adoptively transferred mice were then injected, or not, with mDEF (2 μg) on days 1, 2 and 4 after transfer. Mice were assessed for the development of glycemia on a daily basis.

6.2 Results

6.2.1 mDEF Prevents Diabetes and Restores Normoglycemia [0154]
To investigate the effect of mDEF, we used double transgenic mice that expressed both the influenza A/PR/8/34 virus hemagglutinin (HA) under the control of the rat insulin promoter on pancreatic β cells and the HA(110-120) specific I-E[0155] ^d-restricted TCR on T cells (Radu et al., 1999, Autoimmunity 30:199-207). These mice are referred to as double transgenic (“dTG”) mice. Mice expressing both transgenes on a hybrid BALB/c x B.10.D2 (H-2^d) background have a higher incidence and faster onset of diabetes than the same mice on a BALB/C background (Id.; Sarukhan et al., 1998, EMBO J. 17 71-80). This is associated with a non-MHC encoded predisposition to the differentiation of autoreactive T cells toward the pathogenic Th1 phenotype facored by the B10.D2 background (Scott et al., 1994, Immunity 1:73-83). These dTg mice were used to investigate the effect of mDEF before the onset of disease.
Fifteen-day-old prediabetic mice ([0156] glycemia 104±28 mg/dl) were injected intravenously every 5 days with mDEF (2 μg) for 4.5 months. Control groups were age-matched dTg mice that were injected with equimolar amounts of soluble HA(110-120) synthetic peptide (30 ng), which corresponded to the amount of HA peptide carried by 2 μg of mDEF, 100-fold higher concentrations of HA peptide (3 μg), NPOC-173 (a mouse IgG2a control for the Fcγ2a component of mDEF) or sterile saline solution (which was the equivalent of no antigen).
Ten of twelve mice (83.3%) injected with mDEF resisted diabetes onset during the entire treatment period of 4.5 months. In contrast, the control mice developed diabetes within the first ten weeks of life (FIG. 1A). The degree and severity of insulitis in the mDEF-treated mice was reduced, as shown by a higher frequency of peri-islet insulitis and non-infiltrated islets than in the control (untreated) mice, which developed massive intra-islet infiltration and lacked insulin production in most of the islets (FIG. 1B-D). The mDEF-treated mice that developed diabetes (2 of 12) also showed heavy pancreatic infiltration that was comparable to control diabetic mice. Interruption of mDEF treatment resulted in the development of diabetes within 15±6 days. Higher doses of mDEF (20 μg/mouse) also prevented the onset of diabetes, but the period of protection was not extended if the treatment was interrupted. [0157]
We next investigated whether DEF was able to restore normoglycemia in diabetic mice when given at the onset of hyperglycemia (275±44 mg/dl). A single dose of mDEF (2 μg) administered intravenously restored normoglycemia within 2 days in 9 of 16 mice (56%, glycemia 127±21); this lasted up to 8 days (FIG. 1E). Periodic treatment with mDEF induced long-term normoglycemia in these mice. However, mDEF did not restore normoglycemia when administered to [0158] diabetic mice 2 weeks after the onset of disease (n=8 mice, glycemia 484±35). dTg mice survived several weeks after the onset of hyperglycemia (Sarukhan et al., 1998, EMBO J. 17:71-80). Our results show that ablation of the autoimmune process by mDEF in long-term diabetic mice was not sufficient to induce the recovery of sufficient numbers of β cells or the production of insulin required to restore normoglycemia.
In addition to dTg mice we used single Tg mice that expressed HA under the control of the rat insulin promoter (referred to hereafter as HA-Tg mice). Unlike dTg mice, HA-Tg mice are tolerant of HA pancreatic antigen, and they do not develop diabetes (Lo et al., 1992, Eur. J. Immunol. 22:1013-1022). When injected every 5 days over a three month preiod, mDEF did not induce pancreatic insulitis or hyperglycemia (FIGS. [0159] 1F-G), which indicated that it does not exert toxic effects or break T cell tolerance to pancreatic self (HA) antigen.
Our results indicated that recurrent administration of small doses of mDEF prevented the onset of diabetes in greater than 80 percent of prediabetic mice, restored normoglycemia and insulin secretion in about 60 percent of recent-onset diabetic mice and did not break self-tolerance to HA in mice that were resistant to diabetes. [0160]
6.2.2 Induction of Splenic Anergy by mDEF [0161]
We also examined the kinetics of HA-specific T cell responses in the spleens and thymi of normoglycemic mice treated with single or multiple doses of mDEF. Several groups of dTg mice were compared: untreated mice (group I), mice treated once with mDEF and analyzed one day later (group II) and mice treated periodically with mDEF and analyzed one day (group III) or one week (group IV) after the final injection. [0162]
Cells were stimulated in vitro with HA peptide. Compared to the control mice (group I), the proliferative responses of and cytokine secretion by splenic T cells from groups II and III were reduced (FIGS. [0163] 2A-B). This indicated that mDEF induced T cell unresponsiveness shortly after treatment. However, splenic T cells from group IV mice (analyzed 7 days after mDEF injection) proliferated and secreted mainly IL-2, IFN-γ and IL-10. IL-10 secretion was not detected in cultures from control mice.
mDEF-induced unresponsiveness of the splenic TCR-Tg T cells from dTg mice did not result from clonal deletion. First, flow cytometric analysis showed no decrease in the frequency of CD4+ or CD8+ TCR-Tg T cells, and staining with the monoclonal antibody 6.5 (clonotypic for the HA-specific TCR) did not alter expression of this TCR (FIG. 2C). The latter also showed that T cell hyporesponsiveness did not result from TCR down-modulation. This was in agreement with published data, which showed expression of the transgenic TCR on mDEF-anergized CD4+ T cells from TCR-Tg mice was unaltered (Brumeanu et al., 2001, Int. Rev. Immunol. 20:301-323). Instead, mice injected periodically with mDEF and analyzed 7 days after the last injection (group IV) showed an increase in the TCR-Tg CD4+ T cell population. Second, mDEF-induced hyporesponsive T cells regained their proliferative capacity on in vitro stimulation with HA peptide and recombinant IL-2 (rIL-2) (FIG. 2A). Thus, we inferred that mDEF-induced unresponsiveness does not result from clonal deletion. [0164]
Because mDEF binds to the HA-specific TCR and CD4 on specific T cells (Casares et al., 1997, Protein Engineer. 10:1295-1301; Casares et al., 1999, J. Exp. Med. 190:543-553), anergy could have resulted from the TCR being occupied in such a way that the T cells were unable to interact with cognate ligand on antigen-presenting cells (APCs). Flow cytometric analysis—which used three-color staining with anti-Fcγ2a (for the Fe portion of mDEF), 6.5 mAb (for the transgenic TCR) and antibodies to CD3 (for other T cell subsets)—was used to rule out this possibility. We found that in dTg mice (n=3) injected with mDEF (20 μg) and killed five hours later, mDEF was not stably bound to TCR-Tg T cells or to other T cells. DEF-treated and untreated mice showed <0.2% Fcγ2a+[0165] 6.5+ and Fcγ2a+CD3+ T cells in the spleen.
The T cell response to concanavalin A (“ConA”) was detected in all groups of mice, although the magnitude of the proliferative response was reduced in mDEF-treated mice (FIG. 2A). This showed that the anergy induced by mDEF did not affect the entire population of CD4+ T cells, but rather it affected those specific for the HA peptide. The specificity of mDEF-induced anergy was also assessed in naive BALB/c mice, as these mice have a low frequency of HA-specific cells in the periphery (Casares et al., 1999, J. Exp. Med. 190:543-553; Bot et al., 1998, J. Immunol. 160:4500-4507). BALB/C mice were injected with mDEF; 24 hours later the proliferation responses of splenic T cells to ConA were measured. The similar T cell proliferation indices (Δcpm) obtained for untreated (181,082±3340) and mDEF-treated (208,031±9383) mice showed that mDEF does not induce the anergy of polyclonal CD4+ T cell populations and, hence, anergy was strictly related to TCR-Tg T cells. [0166]
mDEF-induced anergy of CD4+ TCR-Tg T cells occurred in the context of altered early TCR signalling. ZAP-70 kinase was barely, if at all, phosphorylated in splenic CD4+ TCR-Tg T cells from mice killed one day after the last injection of mDEF: group II and group III (FIG. 3). ZAP-70 was not phosphorylated in these cells, and its structural integrity was unaltered, which suggested a mechanism of negative signaling rather than deprivation through protein degradation. In contrast, in vitro stimulation with mAb 2C11 (anti-CD3), but not 6.5 mAb, led to ZAP-70 phosphorylation, which indicated that only CD4+ TCR-Tg T cells, not other cell populations, were rendered anergic by mDEF. [0167]
Lack of ZAP-70 phosphorylation was also found in negatively sorted splenic TCR-Tg CD4+ T cells that were pulsed in vitro with DEF. The in vitro mDEF-anergized T cells did not proliferate upon stimulation with HA in the presence of APCs (control cpm 51,455±5987; mDEF-pulsed Δcpm 725±324), but they did not proliferate vigorously when stimulated with HA-pulsed APCs and rIL-2 (control Δcpm 63,967±8567; mDEF-pulsed Δcpm 59,367±5745). Taken together, these results showed that DEF-induced T cell unresponsiveness is an IL-2 reversible anergy. Anergy was antigen-specific and resulted from the early alteration of TCR signalling. [0168]
mDEF treatment was not followed by thymic clonal deletion because the HA-specific T cell response of thymic precursors from mDEF-treated mice was similar to that in control mice (FIGS. [0169] 4A-B). In addition, flow cytometric analysis showed that mDEF did not affect the development of various subsets of TCR-Tg T cell precursors (FIGS. 4C-F). mDEF did not alter the continuous thymic output of autoreactive T cells. This explains restoration of the TH1 response in the spleens of mice 1 week after mDEF treatment (FIGS. 2A-B) and, thus, the requirement for recurrent DEF administration in order to achieve long-term protection against diabetes.
6.2.3 Anergized T Cells Show Impaired Pancreatic Migration [0170]
To determine whether the anergic CD4+ T cells wre still diabetogenic, we used HA-Tg recombination-activating [0171] gene 2 deficient (“RAG-2^−/−”) mice. These animals develop severe insulitis and diabetes after the transfer of splenic CD4+ T cells from dTg mice (Sarukhan et al., 1998, Immunity 6:563-570). Negatively sorted CD4+ T cells from the spleens of dTg mice that were anergized in vitro with DEF were transferred into HA-Tg-RAG-2^−/−mice. Control mice were HA-Tg-RAG-2^−/−recipients that received CD4+ T cells that were incubated in medium alone. We used in vitro—rather than in vivo—anergized T cells to avoid any contamination with diabetogenic T cells that had newly arrived from the thymus, as mDEF does not affect thymic output. We observed a delay in the onset of hyperglycemia and mortality in mice that had received mDEF-anergized T cells (FIGS. 5A-B). We found that the delay in diabetes onset correlated with the degree of pancreatic infiltration. Pancreata from mice that had received anergic T cells 7 days previously showed a lower degree of insulitis and higher insulin content in islets compared to control mice (FIGS. 5B-C).
We next investigated the fate of mDEF-anergized CD4+ T cells. Negatively sorted CD4+ T cells from untreated dTg mice were first anergized with DEF in vitro, then labeled with carboxyfluorescein diacetate succinimidyl diester (“CFSE”) before transfer into HA-Tg-RAG-2[0172] ^−/−mice. Control cells were treated with medium alone and labeled with sequential halving concentrations of CFSE to determine the decrease in dye fluorescence after successive cell divisions (FIG. 5D).
Using flow cytometry, we determined the frequency and rate of cell division of the labeled cells in the pancreata of [0173] mice 1 and 4 days after adoptive cell transfer. After 1 day, TCR-Tg T cells were found in the pancreata of control mice (FIG. 5E, upper left panel), but few cells were found in the pancreata of mice that had received anergic T cells (FIG. 5E, upper right). Four days after cell transfer, the frequency of TCR-Tg T cells increased more than tenfold in control mice compared to mice that received anergic T cells (FIG. 5E, lower panels). However, once these cells reached the pancreata, both control and anergic T cells showed similar cell division rates (for several generations), as shown be CFSE intensity. Meanwhile, few TCR-Tg cells were found in the spleens of either group of mice.
These results show that mDEF-anergized TCR-Tg T cells had an impaired ability to migrate to and infiltrate the HA-expressing pancreata, which explains the delay in diabetes onset in HA-Tg-RAG-2[0174] ^−/−recipient mice. It is likely that the small fraction of mDEF-pulsed T cells that were still able to migrate to the pancreas were not efficiently anergized by mDEF or had sufficiently recovered from anergy in vivo. This data is in agreement with results which showed tht recurrent administration of mDEF was mandatory for long-term prevention of diabetes in dTg mice via the continuous silencing of peripheral T cells and T cells that were newly derived from the thymus.
6.2.4 mDEF Stimulated TR1-Like Cells in Pancreas [0175]
In contrast to the hyporesponsiveness of splenic T cells from mDEF-treated mice (groups I and III), pancreas-infiltrating T cells from these mice proliferated and secreted predominantly IL-10 on in vitro stimulation with HA (FIGS. [0176] 6A-B). Control (untreated) mice secreted IL-2 and IFN-γ, but not IL-10. Staining with 6.5 anti-clonotypic TCR antibody showed a similar frequency of TCR-Tg CD4+ T cells in the pancreta of both control and mDEF-treated mice (FIGS. 6C-D). These results showed that mDEF induced the pancreatic T cells to secrete IL-10.
Seven days after treatment, pancreatic T cells secreted IL-10 and IL-2. Restoration of the TH1 response was likely mediated by repopulation with T cells that had just arrived from the thymus and/or from the peripheral lymphoid organs. [0177]
To determine whether the IL-10 secreting T cells were diabetogenic, pancreatic CD4+ T cells pulsed in vitro with mDEF or medium alone (control) were transferred into HA-Tg-RAG-2[0178] ^−/−mice. The pancreatic TCR-Tg CD4+ T cells pulsed in vitro with mDEF also secreted IL-10, as determined by intracellular staining with 6.5 and IL-10 mAbs (FIG. 6E). This was consistent with the in vivo results. These mice did not develop diabetes (FIG. 6F) and some degree of peri-islet infiltrations was observed 70 days after cell transfer, which indicated tht the T_Rcells persisted in the recipient for a long period of time.
In addition, we found that the IL-10 secreting regulatory T (T[0179] _R) cells generated in vitro suprressed HA-specific splenic T cells from untreated dTg mice via IL-10 secretion. Suppression was reversed by the addition of neutralizing anti-IL-10 (FIG. 6G). This also occurred with the pancreatic T_Rcells generated in vivo upon mDEF treatment. Pancreatic T cells from mice from groups III and IV (which were periodically treated with mDEF, and were analyzed one day or one week later) inhibited the proliferative responses of splenic T cells (cultured at a 1:1 ratio) by 73 percent and 53 percent, respectively. Inhibition correlated with the amount of IL-10 secreted (FIGS. 6A-B). Immunomodulation via IL-10 secretion suggested that these antigen-specific T_Rcells were T_R1-like (Levings and Roncarolo, 2000, J. Allergy Clin. Immunol. 106:109-112).
To determine whether [0180] T _R1 cells are able to suppress the diabetogenicity of TCR-Tg T cells in vivo, CD4+ pancreatic T cells from dTg mice that had been stimulated in vitro with mDEF were transferred into HA-Tg-RAG-2^−/−mice. One week later the mice were challenged with splenic CD4+ T cells from untreated dTg mice+anti-IL-10 or with rat IgG1 isotype controls. The mice that received T _R1 cells showed delayed diabetes onset in 11 of 12 mice. One of these mice resisted developing the disease for more than 4 months. In contrast, all control mice developed diabetes within 18 days of T cell transfer (FIG. 6H). Administraton of neutralizing anti-IL-10 reduced protection, which showed that T _R1 cells suppressed the diabetogenic TCR-Tg CD4+ T cells in vivo.
6.2.5 mDEF Fosters Trafficking of Pancreatic T Cells [0181]
Because mDEF stimulated IL-10 secreting T cells in pancreas, we next addresed the question of whether IL-10 detected in the spleens of [0182] dTg mice 1 week after mDEF administration (FIG. 2A, group IV) could have been secreted by pancreatic T _R1 cells that had migrated in the spleen. We found that IL-10 secreted in spleens 1 week after mDEF administration was associated with reduced amounts of IL-10 in the pancreas (FIGS. 2B and 6B, group IV). This was also associated with an increase in the frequency of TCR-Tg CD4+ T cells in the spleen and a reduction of this population in pancreas (FIGS. 2C and 6C, D, group IV).
When transferred into HA-Tg-RAG-2[0183] ^−/−mice, TCR-Tg CD4+ T cells from the untreated dTg mice preferentially migrated to the pancreas (FIGS. 5C-e). Therefore, we investigated whether treatment of adoptively-transferred HA-Tg-RAG-2^−/−mice with mDEF induced the migration of T cells from pancreas to the peripheral lymphoid organs. CD8-depleted splenic T cells from untreated dTg mice were transferred into HA-Tg-RAG-2^−/−mice. These mice were then injected or not with DEF (2 μg) 24 hours, 48 hours, and 96 hours after cell transfer. Three days later, we looked for TCR-Tg T cells in the pancreata, spleens and popliteal lymph nodes. In control mice, most of the TCR-Tg (6.5+) cells were found in pancreas; in mDEF-treated mice, however, most of the cells were found in the pancreata and peripheral lymphoid organs (FIG. 7A). This showed that mDEF treatment induced a change in T cell behavior in the pancreas.

The pancreatic T cells migrated to lymphoid organs, possibly as a result of up-regulation of adhesion molecules such as CD62 ligand (“CD62L”). Indeed, pancreatic TCR-Tg T cells from adoptively transferred HA-Tg-RAG-2 ^−/−mice treated with mDEF showed higher expression of CD62L (FIG. 7B, middle panel) than the donor cells (FIG. 7B, left panel) or cells from the pancreata of control (untreated) mice (FIG. 7B, right panel). We also found that pancreatic TCR-Tg T cells from prediabetic dTg mice on mDEF therapy showed increased expression of CD62L (Table 2). Up-regulation of CD62L was restricted to TCR-Tg T cells, as the population of cells that did not express the transgenic TCR (6.5−) showed no differences in CD62L expression in either treated or untreated mice. In both HA-Tg-RAG-2^−/−mice and dTg mice treated with mDEF, CD62L was also up-regulated on splenic TCR-Tg T cells, although overall expression was lower than that in the pancreas. Increased expression of CD62L on pancreatic T cells after mDEF treatment may well favor their migration from the pancrea; it may also explain the presence of IL-10 secreting T cells in the spleens of dTg mice 1 week after the last injection of mDEF.

TABLE 2


Cells	Gated on 6.5 + CD62L+	Gated on 6.5 − CD62L+

Group	46.2 ± 3.2	47.08 ± 8.1
I
Group	57.1 ± 2.2	48.6 ± 5.4
II
Group	61.7 ± 3.8	51.8 ± 9.2
III
Group	50.2 ± 2.6	50.2 ± 5.5
IV

7. EXAMPLE

A Human Peptide-MHC Class II Chimera

7.1. Construction of Human DEF Molecules

A human DEF molecule was constructed comprising the extracellular domains of HLA-DR(α1*0101.-β*0401), with either the p271-285 GAD65 peptide (PRLIAFTSEHSHFSL; SEQ ID NO:10) or the proinsulin peptide p73-90 (GAGSLQPLALEGSLQKRG; SEQ ID NO: 43) covalently linked to the N-terminus of the HLA-DRB1*0401 chain. The construct used is depicted in FIG. 8A. The resulting HLA-DR4/peptide complexes were dimerized via a human Fcγ1 fragment covalemntly linked to the C-terminus of the HLA-DRB1*0401 chain (see FIGS. 8A and 8B). The resulting molecules were named according to their incorporated GAD65 or proinsulin peptide, DEF-GAD65 and DEF-Pi, respectively. Sequencing gels of partial sequences of the constructs producing DEF-GAD65, DEF-Pi and DEF are shown in FIGS. [0185] 8A-C, respectively.
In preparing these constructs, the gene encoding the extracellular domains of α and β chains of HLA-DRB1*0401 was obtained using RT-PCR from mRNA of Boleth cells, a HLA-DR(α[0186] 1*0101, -β10401) lymphoblastoid cell line. Codons at positions 214 (ACG=Thr) and 215 (AGC=Ser) were mutagenized to ACT=Thr and AGT=Ser to create a SpeI restriction site. The primers used had the nucleic acid sequences:
For cloning of HLA-DRA*0101, the primers were: [0187]

5′-ACGGCTTAAGACTCCCAACAGAGCGCCCAAGAA-3′

(SEQ ID NO: 65)

and

5′-GCCGTCTAGATTACCAGTGCTTGAGAAGAGGCT-3′.

(SEQ ID NO: 66)

For cloning the peptide and linker, the primers were:


5′-ATCGGAATTCCTCCTCTGGCCCCTGGTCCTG-3′
(SEQ ID NO: 67)

and

5′-

CGATGGATCCACCACCTCCTAGTGAAAAGTGTGAATGTTCTGATGTAAAT

GCAATTAGTCGTGGTCGGGTGTCCCCAGC-3′.
(SEQ ID NO: 68)

For cloning HLA-DRB1*0401, the primers were: [0189]

5′-

ATCGGGATCCGGTGGAGGGGGAAGTGGAGGTGGAGGGTCTCCACGTTTCT

TGGAGC-3′

(SEQ ID NO: 69)

and

5′-GAGGACTAGTCAGGCTTGGGTGCTCCAC-3′.

(SEQ ID NO: 70)
This allowed the assembly of genes encoding β1*0401 and Fcγ components in a natural frame. The cDNA coding for the hinge, CH2 and CH3 domains of human IgG1 was obtained from the ZM-1 hybridoma obtained from the American Type Culture Collection (“ATCC”), Manassas, Va., by RT-PCR using the [0190] primers 5′-ACTGACTAGTCCTCTCACAGTGGAATGGGCAGAGCCCAAATCTTGTGAC-3′ (SEQ ID NO:71) and 5′-CAGTAAGCTTTCATTTACCCGGAGACAGGGAGAG-3′ (SEQ ID NO:72). The β chain of DEF-Pi was obtained from the gene encoding DEF-GAD65 by deletion of sequences encoding p271-285 of GAD65 and replacing them with sequence encoding the proinsulin p73-90 peptide. The β chain of a control DEF was obtained by assembling the genes encoding the extracellular domains of HLA-DRB1*0401 and human Fcγ1.

7.2 Characterization of Human DEF Molecules

The human DEF molecules constructed according to the preceding section were produced in a SF9/baculovirus system and purified using a protocol involving affinity chromatography using an anti-human Igγ1-Sepharose 4B column, similar to the procedure reported in Casares et al., 1997, Protein Engineer. 10: 1295-1301. The molecular size of DEF-GAD65 and DEF-Pi under non-reducing conditions was found to be approximately 185 kDa (FIGS. [0191] 8F-G). After boiling and disulfide reduction the chimeric molecules showed a component of 60 kDa that appears to correspond to monomeric peptide/DRβ1*0401/Fcγ1. Dissociated MHC chain under boiling condition was detected as a separate component of 32 kDa. These results demonstrate that the recombinant human DEF molecules were secreted by SF9 cells as a soluble dimer of 185 kDa.
The structural integrity of human DEF molecules was analyzed by Western Blot. The intact DEF-GAD65 and DEF-Pi chimeric molecules were recognized by both anti-human IgG1 and anti-HLA-DR antibodies (FIGS. 8F and G). After boiling and reduction, the chimeric molecules showed a component of 60 kDa corresponding in molecular size to half of the molecule without the MHCα chain. The 60 kDa component, but not the dissociated MHC a chain, were identified by both anti-human IgG1 and anti-HLA-DR antibodies (FIG. 8H). [0192]

7.3 Immunomodulatory Effects of Human DEF Molecules

7.3.1. Effects in a Murine Model [0193]
HLA-DR*0401 +/+, Abb −/− mice (C56BL/6, Taconic, Germantown, N.Y.) expressthe α1β1 domains of HLA-DR*0401 linked to the α2β2 domains of I-E[0194] ^d, and at the same time are deficient for endogenous MHC class II (1-A^b, I-E^b) molecules (Abb −/−).
Groups of five mice were immunized in the footpads with either 100 micrograms of p271-285 GAD65 peptide or p73-90 proinsulin peptide in Complete Freund's Adjuvant (“CFA”). Lymph nodes were harvested ten days later and cells from GAD65-immunized mice were stimulated in vitro with various doses of p271-285 synthetic peptide, DEF-GAD65 or empty DEF molecules. Cells from proinsulin-immunized mice were stimulated in vitro with various doses of p73-90 proinsulin peptideDEF-Pi, or empty (control) DEF. The amount of peptide carried by the human DEF molecules was normalized on a molar basis, and the immunopotency determined as described in Casares et al., 1999, 3. Exp. Med. 190: 543-553 and Stordeur and Goldman, 1998, Int. Rev. Immunol. 16:501-522. The molecular weights of p271-285 GAD65 and p73-90 proinsulin are 1730 and 1850 Da, respectively. The molecular weights for the human DEF molecules, as set forth above, is approximately 185 kDa. Half-maximal proliferation of of stimulated cells was obtained with 1 micromolar of p271-285, 2.5 micromolar of p73-90, 2.5 nanomolar of DEF-GAD65 or 0.5 nanomolar DEF-Pi (FIGS. 9A and B). [0195]
The results of these experiments showed that DEF-GAD65 and DEF-Pi molecules were, respectively, 400 and 5000 more potent in activating T cells than the synthetic peptides. These data are in agreement with studies which showed that murine DEF is 100 more potent than the nominal peptide in activating specific T cells (Casares et al., 1999, J. Exp. Med. 190: 543-553). [0196]
In addition, cells from both groups of mice stimulated with DEF-GAD65 or DEF-Pi, respectively, produced high amounts of IL-10 (FIG. 9C). Cells from GAD65-immunized mice stimulated in vitro with the GAD65 peptide also secreted IL-10. In contrast, cells from proinsulin-immunized mice stimulated with the nominal peptide secreted IL-2 but not IL-10. [0197]
In mice, the p271-285 of human GAD65 is a self-peptide expressed by the pancreatic β cells, whereas the p73-90 of human proinsulin is a foreign peptide. Immunization of HLA-DR4 transgenic mice with GAD65 antigen or p271-285 peptide does not induce insulins or diabetes (Wicker et al., 1996, J. Clin. Invest. 98:2597-2603; Patel et al., 1997, Proc. Natl. Acad. Sci. U.S.A. 94: 8082-8087; Abraham and David, 2000, Cur. Opin. Immunol. 12:122-129). Thus, the lack of diabetogenicity on immunization with GAD65 can be explained by the induction of non-diabetogenic/regulatory IL-10 secreting T cells, since IL-10 is a powerful inhibitory cytokine for Th1 cells (Stordeur and Goldman, 1998, Int. Rev. Immunol. 16:501-522). [0198]
The T cell response on stimulation with the human DEF molecules was specific, since, in control experiments, incubation of cells from both groups of mice with human IgG1 did not induce proliferation or cytokine secretion. It has been demonstrated that the Th2 immune deviation induced by murine DEF occurs through DEF binding to both TCR and CD4 co-receptor (Casares et al., 1999, J. Exp. Med. 190: 543-553). The addition of anti-mouse CD4 monoclonal antibody GK1.5, Pharmingen, San Diego, Calif.) to the cultures prior to stimulation with human DEF molecules completely abolished their prolifeative response and IL-10 secretion, indicating the ability of HLA-DR to bind to murine CD4. Binding of HLA-DR molecules to mouse CD4 has also been reported in Rosloniec et al., 1998, J. Immunol. 160:2573-2578 and Pan et al., 1998, J. Immunol. 161: 2925-2929. The results indicate that the interaction of human DEF molecules with cognate TCR and CD4 co-receptor stimulate specific T cells toward the Th2-like phenotype and secretion of IL-10. [0199]
7.3.2. Effects on Human Diabetic Patients' T-Cells [0200]
Fifteen HLA-DR*0401 patients diagnosed with [0201] type 1 diabetes, ages 8-21, were used in the following studies. Mononuclear cells (2×10⁵) were isolated from peripheral blood mononuclear cells (“PBMCs”) from patient nos. 18 and 19 (both haplotype B1*0401/B1*0402) and were stimulated with p271-285 GAD65 peptide (15 micrograms per ml) or DEF-GAD65 (5 micrograms per ml). Mononuclear cells from the other patients were stimulated with either p271-285 GAD65 or p73-90 proinsulin peptide at 15 micrograms pre ml or DEF-GAD65 or DEFiPi molecules at 5 micrograms per ml. In addition, T cells from patients 13, 26, 42, 43, 44 and 61 were stimulated with phytohemagglutinin (“PHA”). Cells from all the patients were also stimulated with human IgG1 at 5 micrograms per milliliter. The results are shown in FIG. 10A-D.
T cells from patients stimulaed in vitro with human DEF molecules, but not with synthetic peptides, secreted high amounts of IL-10 (FIG. 10B). Only T cells from patient no. 23 secreted higher levels of IFNγ than IL-10 on stimulations with DEF-Pi (FIG. 10C). No cytokine secretion was detected on stimulation with human IgG1. It is noteworthy that DEF-GAD65 stimulation induced an overall higher IL-10 response as compared to DEF-Pi. Addition to the cultures of neutralizing anti-human IL-10 antibodies (clone JES3-9D7, Pharmingen, 100 micrograms per ml) did not increase the proliferative response to DEF-Gad65, indicating that a lower proliferative response to DEF-GAD65 was not due to IL-10 mediated suppression of T cells. It is known that IL-10 exerts inhibitory effects preferentially on Th1 cells (Stordeur and Goldman, 1998, Int. Rev. Immunol. 12:122-129). [0202]
Addition of anti-human CD4 antibodies to the cultures stimulated with human DEF molecules abrogated the secretion of IL-10 (FIG. 10D). This indicates that the human DEF molecules immunomodulate T cells through the binding to both TCR and CD4 co-receptor. [0203]
Next, it was investigated whether the CD4 T cells were responsible for DEF-induced Il-10 secretion. Ficoll-isolated PBMC from patient 19 (showing T cell responses to DEF-GAD65 but not DEF-Pi) were stimulated for 48 hours with human IgG1, DEF-GAD65 or DEF-Pi molecules, and then were surface stained for CD3 and CD4 and intracellularly stained for IL-10. As shown in FIG. 11A, cells stimulated with DEF-GAD65, but not with DEF-Pi or IgG1, produced IL-10. Analysis of IL-10 producing T cells indicated that the majority of these cells (71.2%) were CD3+CD4+ T cells (FIG. 11B), indicating that human DEF molecules induce IL-10 secretion by specific CD4 cells. [0204]
Thes studies indicate that human DEF molecules are powerful immunomodulators of autoreactive T cells by polarizing them toward a protective Th2 phenotype and secretion of IL-10, a potent immunosuppressive cytokine for diabetogenic Th1 cells. [0205]

8. EXAMPLE

Multivalent MHC Class II-Peptide Chimeras

8.1 Materials and Methods

8.1.2 Mice [0206]
Transgenic (Tg) BALB/c mice expressing the 14.3d TCR that recognizes the HA110-120 peptide of the hemagglutinin (HA) protein of PR8 influenza virus in association with I-E[0207] ^dclass II molecules were provided by Dr. von Boehmer (Dana Farber Institute, Miss.). The seven to eight-week-old BALB/c mice were purchased from Jackson Laboratories, USA.
8.1.3 Cells [0208]
Primary T-cells specific for the immunodominant CD4 T cell epitope HA110-120 (TCR-HA T cells) were obtained from the spleens of Tg mice. Approximately 32% of T-cells in the spleen of these mice express the TCR-HA transgene as determined by fluorescence activated cell sorting (FACS) using 6.5.2 anti-TCR-HA clonotypic monoclonal antibody (mAb). The clonotypic 6.5.2 mAb (rat IgG1/k) was provided by Dr. J. Caton (NIH). The 14.3-1 T-cell hybridoma cells (TcH) expressing the 14.3 d TCR-HA were provided by Dr. K. Karjalainen (Basel Institute for Immunology, Switzerland). [0209]
8.1.4 MHC/Peptide Chimera [0210]
The genetically engineered soluble, bivalent MHC class II peptide chimera consists of the I-E[0211] ^dα and I-E^dβ extracellular domains that were dimerized through a murine Fcγ2a fragment at the C-termini of I-E^dβ chains. The HA110-120 (SFERFEIFPKE, SEQ ID NO: 55) CD4 T cell epitope of HA of influenza virus A/PR/8/34 (Haberman et al., 1990, J. Immunol. 145:3087) was covalently linked to the N-terminus of I-E^dβ chains (Casares et al., 1997, Protein Engineering 10:1295-1301). The soluble, bivalent chimeric protein was chromatographically purified on a goat anti-mouse γ2a-Sepharose column from the cell culture supernatants of SF9 insect cells infected with baculovirus expressing both I-E^dα and I-E^dβ/HA110-120/Fcγ2a genes, as previously described (Casares et al., 1997, supra). The soluble bivalent chimeric protein binds stably and specifically to cognate TCR on CD4+ T-cells (Casares et al., 1997, supra).
8.1.5 Enzymatically-Mediated Synthesis of Multimers [0212]
Soluble, bivalent chimeric protein (5 mg) was incubated overnight at 37° C. with 500 mU of neuraminidases from [0213] Arthrobacter ureafaciens and Clostridium perfringens (Calbiochem-Novobiochem Intern. Inc., La Jolla, Calif.) in 5 ml of 0.1 M phosphate buffer pH 5.5 containing 5 mM CaCl₂. Free sialic acid released by neuraminidases was removed by dialysis against phosphate buffered saline (PBS) pH 7.4. Desialylated bivalent chimeric protein was incubated for 48 hours at 37° C. with 100 U galactose oxidase (GAO, Sigma Chemical Co., St. Louis, Mo.) and 5 mg of diamino-polyethylene glycol bifunctional cross linker with a molecular mass of 3,400 Da ((NH₂)₂-PEG_3,400, Shearwater Corporation, Ala.). The Schiff bases formed between the aldehyde groups generated by GAO at the 6^thcarbon of terminal galactose residues and the amino groups of PEG were stabilized on mild reduction with 80 mM of pyridine borane (PB)(Aldrich) (FIG. 12). The reaction mixture was dialyzed against PBS in SPECTRA/POR bags (100,000 MWCO, Sigma), and multimers were separated by size exclusion chromatography.
8.1.6 Chromatographic Separation of Complexes [0214]
Multimers were separated by size-exclusion chromatography in a [0215] Superose 6 column (Amersham-Pharmacia Biotech) equilibrated in PBS. The reaction mixture (200 μl) was applied in the column at 1 ml/min flow rate, and fractions were collected at 1 minute intervals. The recovery yield for each protein peak was calculated on the chromatographic profile using the UN-SCAN-IT analysis software version 5.1 (Silk Scientific Corp., Calif.). To identify the PEG polymer, some 0.025 ml from each fraction were reacted with 0.025 ml of Nessler's Reagent (Sigma). PEG polymer was detected as a whitish precipitate. The peak tubes were measured for the protein content using the Biuret microassay, since the Biuret reagent does not interfere with PEG polymers (Brumeanu et al., 1995, J. Immunol. 154:3088-3095).
8.1.7 SDS-PAGE and Western Blot Analyses [0216]
Some 5 μg of chromatographically purified multimers were electrophoresed in 4-12% polyacrylamide gradient gels (PhastGels, Amersham-Pharmacia, N.J.) under denaturing and non-reducing conditions, and the gels were silver stained according to the manufacturer's instructions. In parallel experiments, 5 μl of blood serum and tissue extracts from mice injected intravenously (i.v.) with [0217] ¹²⁵I-radiolabeled multimers were analyzed at various intervals of time by SDS-PAGE under denaturing and non-reducing conditions using 4-12% gradient PhastGels. Samples of tissue extracts containing ¹²⁵I-labeled multimers were prepared from spleen, thymus, lymph nodes, and brain. The extracts were obtained by tissue homogenization, and then cleared of debris by centrifugation. The supernatants were collected and precipitated for 2 hours at room temperature with 50% of saturated ammonium sulfate (SAS), and the SAS precipitates were dialyzed extensively at 4° C. in SPECTRA/POR bags (100,000 MWCO) against PBS containing a cocktail of protein inhibitors (Complete kit, Boehringer Mannheim, Germany). The protein concentration in the dialyzed preparation was adjusted at 5 mg protein/ml with sample buffer containing 5% of 2-mercaptoethanol (2ME), and separated by SDS-PAGE in 4-12% gradient PhastGels. Gels were electrotransferred onto PVDF membranes (0.45 μm), and the radioactive bands were identified upon exposure of membranes onto Kodak X-OMAT films (Sigma).
The specificity of the glycosidic bonds generated by (NH[0218] ₂)₂-PEG_3,400linker between the N-glycan moieties of chimeric dimers was determined by Western blot. Samples of purified multimers (5 μg) were digested for 2 hours at 37° C. with PGNase F (0.01 U/μg protein, Sigma) in the presence of 5% 2ME, and electrophoresed on 10-15% gradient PhastGels. Gels were electrotransferred onto PVDF membranes (0.45 μm), and the membranes were blocked overnight at 4° C. with 5% fat free milk (Carnation, Nestlé Food Company, Glendale, Calif.) in PBS, then washed with 0.05% Tween 20 in PBS, and incubated for 2 hours at room temperature with ¹²⁵I-labeled goat anti-mouse γ2a Ab (2×10⁵cpm/100 cm²membrane) in PBS containing 1% BSA and 0.05% Tween 20. The membranes were washed with 0.05% Tween 20 in PBS, and exposed onto Kodak X-OMAT films.
8.1.8 Cytofluorometric Analyses [0219]
The 14-3-1 TcH (1×10[0220] ⁵) expressing 14.3d TCR-HA were incubated for 30 minutes on ice with 2 μg/ml of the purified multimers in PBS/BSA 1% containing or not 100 μg/ml 6.5.2 anti-TCR clonotypic mAb. Cells were washed in cold PBS/BSA 1% NaN₃0.05%, and bound multimeric molecules were stained for 30 minutes on ice with a goat anti-γ2a-FITC conjugate (Boehringer Mannheim). The fluorescence intensity was measured among 10,000 cells in a FACSCalibur instrument (Becton Dickinson, Calif.) after subtraction of the background generated by the secondary Ab-FITC conjugate. To determine the extent of apoptosis induced by the multimers in TCR-HA T-cells, two-color FACS analysis was used. Cells were stained for 30 minutes on ice with 2 μg of 6.5.2 clonotypic mAb-FITC conjugate and 2 μg of anti-Anexin V-PE conjugates (PharMingen, Calif.), and the 6.5.2⁺/Anexin V⁺ T-cells were scored among 10,000 events using the FACSCalibur instrument.
8.1.9 Thymidine Incorporation Assay [0221]
The proliferative capacity of TCR-HA T-cells on exposure to the multimers was determined by thymidine incorporation assay ([0222] ³H-TdR). Spleen cells (10⁶) from TCR-HA Tg mice were incubated for 72 hours with various concentrations of purified multimers, or medium alone. Tritiated thymidine (1 μCi/well) was added to the cultures for the last 24 h, cells were harvested on filter paper (Squadron Inc., Sterling, Va.), and the radioactivity (cpm) was measured in a β-scintillation chamber (Amersham-Pharmacia Biotech).
8.1.10 Cytokine Assays [0223]
Cytokine production was determined in the cell culture supernatants of spleen cells (10[0224] ⁶) from Tg mice incubated for 48 hours with 10 μg/ml and 50 μg/ml of multimers. The amount of IL-2, IL-4, and IFN-γ was measured by ELISA according to the manufacturer's instructions (Cytoscreen mIL-2 and Cytoscreen mIL-4 ELISA kits, Biosource International, Calif.).
8.1.11 Blood Clearance and Organ Distribution [0225]
The multimers (100 μg in 100 μl of PBS) were radiolabeled with [0226] ¹²⁵I using the conventional chloramine method. Two BALB/c mice per group were injected in the tail vein with ²⁵I-labeled multimers (50×10⁶cpm) in PBS (200 μl). The clearance rates and index of distribution in organs for the radiolabeled multimers were calculated as previously described (Brumeanu et al., 1995, J. Immunol. 154:3088-3095). Briefly, 15 minutes after injection, the time required for uniform distribution of the radiolabeled material, the mice were bled from the tail vein, and radioactivity (cpm) in 20 μl of serum was measured in a γ-counter (Pharmacia LKB). The total radioactivity injected per mouse (TRI) was estimated 15 minutes after injection, on the basis that 7.3% of body weight is blood and 55% of the blood volume is serum. The total residual radioactivity (TRR) was estimated in blood at the time of withdrawal. The index of distribution of multimers in the organs was expressed as a percent from TRR at the time of collection (% ID), according to the formula: % ID=1-(TRR/TRI)×100 (Brumeanu et al., 1995, J. Immunol. 154:3088-3095).

8.2 Characterization of Multimeric Complexes

Chimeric multimers synthesized as described in Example 1 were separated by size exclusion chromatography and characterized by SDS-PAGE and Western blot, the results of which are depicted in FIG. 13. [0227]
A [0228] superose 6 HR 10/30 column was calibrated at 1 min/ml with mouse IgM (≧800 kDa), thyroglobulin (660 kDa), ferritin (440 kDa), catalase (250 kDa), mouse IgG (150 kDa), cytochrome c (14.4 kDa) in PBS. The mixture of multimers was dialyzed, applied on the column, and the tubes were collected at 1 minute intervals (continuous line). Each tube was tested for PEG presence by Nessler's reagent (negative reaction (−), weak positive reaction (±), and strong positive reaction (+). Some free, residual PEG polymer that was not dialyzed out eluted in the salt volume of the column (+).
The insert in FIG. 13 represents chromatographically purified multimers analyzed by SDS-PAGE under denaturing and non-reducing conditions ([0229] lane 1, octamer; lane 2, tetramer, and lane 3, dimer). Digestion with PGNase F under reducing conditions and identification with ¹²⁵I-goat anti-γ2a Ab by Western blot revealed that multimers were composed of identical monomeric units of ˜80 kDa (lane 4, octamer; lane 5, tetramer, and lane 6, dimer).
The foregoing results show that the enzymatically-mediated cross linking of chimeric dimers via diamino-PEG[0230] _3,400polymer led mainly to the generation of tetramers and octamers as found by size exclusion chromatography. Their relative molecular masses estimated in the peak tubes were 375 kDa and 720 kDa, respectively. Quantification of the corresponding chromatographic peaks showed a yield recovery of 55% for tetramer, and 32.5% for octamer. Some 2% of highly multimerized complex (MW≧800 kDa) and 9.5% of the dimer were also separated by chromatography. Accordingly, a small amount of dimer (9.5%) did not react with PEG polymer either because of less accessibility of the galactose acceptors, or because of less galactose acceptors per dimeric molecule. Using a galactose oxidase (GAO)/tolidine-horseradish peroxidase coupled assay, it was found that the number of galactose acceptors per molecule of dimer was on average 10.5. A small fraction of the dimers expressing lower amount of carbohydrates, which contained on the average 3.7 galactose acceptors per molecule, had previously been separated by anion-exchange chromatography (Casares et al., 1997, Protein Engineering 10:1295-1301). This may account for the lack of cross linking by PEG polymer for 9.5% of the dimers. The amount of PEG in the tetramer and octamer as detected by Nessler's reagent was considerably lower than in the peak of free PEG (MW≦5 kDa). Lack of PEG in the peak of dimer and presence of PEG in the peaks of octamer and tetramer indicated that PEG did not co-elute with these proteins but rather was strongly attached to them.
The SDS-PAGE analysis confirmed the composition of the chromatographic peaks as consisting of multimers with molecular masses of 170 kDa for dimer, 365 kDa for tetramer, and 700 kDa for octamer (FIG. 13 insert, [0231] lanes 1, 2, and 30. Digestion of multimers with PGNase F under denaturing and reducing conditions followed by blotting with ¹²⁵I-radiolabeled goat anti-γ2a Ab revealed a major component of ˜80 kDa that corresponded to the monomeric unit of the dimer (FIGURE insert, lanes 4, 5, and 60. This clearly demonstrates that PEG polymer was able to covalently cross link the dimers through the N-glycan moieties. Together, the results demonstrate that: (1) the enzymatically-mediated cross linking of dimers by diamino-PEG_3,400polymer generated covalently linked tetramers and octamers through their carbohydrate moieties, and (2) multimers can be efficiently cleared of residual adducts by size exclusion chromatography.

8.3 Binding of Multimeric Complexes to Cognate T Cells

The multimers of the previous examples were tested by FACS for their ability to bind to TCR-HA T-cells. The fluorescence intensity of 14-3-1 TcH expressing TCR-HA upon incubation with purified multimers and secondary Ab (goat anti-γ2a-FITC) was between 33.2 and 38.7%. The fact that differences in the fluorescence intensity of multimers were not detected suggests that either FACS analysis cannot distinguish discrete alterations in their affinity binding constants to cognate TCR, or the incorporated PEG cross linker in the multimers can interfere with the binding of secondary Ab. However, none of multimers bound to 14-3-1 TcH when the cells were preincubated with 6.5.2 clonotypic mAb. This clonotypic mAb inhibits the binding of the dimer to 14-3-1 TcH (Casares et al., 1999, supra). In aggregate, the results demonstrate that the enzymatically-mediated multimerization of the chimeric molecule did not affect its ability to bind specifically to cognate TCR. [0232]

8.4 Blood Clearance and Organ Distribution of Multimers

The multimers of the foregoing examples were tested in naïve BALB/c mice for their life span and stability in blood circulation and lymphoid organs. [0233]
Life span of multimers in blood circulation and distribution in the lymphoid organs were determined as follows. Chromatographically purified multimers were radiolabeled with [0234] ¹²⁵Iodine, injected i.v. in naïve BALB/c mice, and the blood clearance, organ distribution, and the degradation patterns in blood and in lymphoid organs were determined as described. FIG. 14A shows the persistence of multimers in blood circulation; FIG. 14B shows the degradation patterns of multimers in blood circulation at various intervals of time after injection. Lane 1, ¹²⁵I-labeled octamer before injection; lane 2, ¹²⁵I-labeled octamer at 24 hours; lane 3, ¹²⁵I-labeled octamer at 48 hours; lane 4, ¹²⁵I-labeled octamer at 72 hours; lane 5, ¹²⁵I-labeled octamer at 96 hours; lane 6, ¹²⁵I-labeled tetramer before injection; lane 7, ¹²⁵I-labeled tetramer at 24 hours; lane 8, ¹²⁵I-labeled tetramer at 48 hours; lane 9, ¹²⁵I-labeled tetramer at 72 hours; lane 10, ¹²⁵I-labeled tetramer at 96 hours; lane 11, ¹²⁵I-labeled dimer before injection; lane 12, ¹²⁵I-labeled dimer at 24 hours; lane 13, ¹²⁵I-labeled dimer at 48 hours; lane 14, ¹²⁵I-labeled dimer at 72 hours; and lane 15, ¹²⁵I-labeled dimer at 96 hours after injection. FIG. 3C shows the index of distribution (ID (%) of multimers in lymphoid organs and brain. The bars represent the mean value per group of mice with a ±SD of 1.3% for spleen, 0.4% for lymph nodes, 1.4% for thymus, and 0.2% for brain. The bars marked for serum represent the % ID in serum 15 minutes after the i.v. injection, and they were assigned as the total radioactivity injection (% ID=100). The (% ID) values in organs were calculated in relation to the total radioactivity in blood as described. FIG. 14D illustrates the degradation patterns of multimers in the lymphoid organs, 48 hours after the i.v. administration. Lanes 1, 2, and 3, ¹²⁵I-labeled octamer, tetramer, and diner, respectively, in spleen homogenate. Lanes 4, 5, and 6, ¹²⁵I-labeled octamer, tetramer, and dimer, respectively, in homogenates from lymph nodes. Lanes 7, 8, and 9, ¹²⁵I-labeled octamer, tetramer, and dimer, respectively, in thymus homogenate.
All multimers showed longer half-life in blood circulation than a genetically engineered immunoglobulin (IgHA) expressing the HA110-120 peptide in the CDR3 loop of VH domain (Brumeanu et al., 1993, J. Exp. Med. 178:1795-1799). The half-life of multimers in blood was 50 hours (FIG. 14A). The electrophoretic analysis showed similar patterns of degradation of multimers in blood (FIG. 14B). During the first 24 hours there was no detectable degradation, whereas after 24 hours the degradation occurred progressively in all multimers. However, intact molecules of multimers were still detected 72 hours after injection. [0235]
Multimers were detected in spleen, lymph nodes and thymus (FIG. 14C) where they persisted as intact molecules for 48 hours after injection (FIG. 14D). However, the degradation process was slightly higher in spleen than in thymus and lymph nodes for all multimers, presumably because of APC's ability to uptake and process the proteins. None of multimers were detected in brain, indicating their inability to cross the hematoencephalic barrier. Longer life of multimers than the immunoglobulins in vivo may account for the high amount of carbohydrate moieties expressed by multimers. The persistence of multimers as intact molecules in blood and lymphoid organs also demonstrated that the PEG-galactose imidic bonds were resistant to endoglycosidases. [0236]

8.5 Immunoregulatory Effects of Multimers

The potency of multimers in stimulating cognate T-cells was compared at three different degrees of TCR/CD4 occupancy, 1, 10, and 50 μg/ml of multimer per 10[0237] ⁶splenic TCR-HA Tg cells. Since the frequency of TCR-HA T-cells in spleen of these mice is 30-33%, the TCR/CD4 occupancy corresponds respectively, to 0.7, 7.0, and 35 pMole multimer per cell.
The TCR-HA splenic T-cells from transgenic mice were exposed to the various concentrations of multimers for three days, and the [0238] ³H-TdR assay of thymidine incorporation in proliferating T-cells was determined as described hereinabove (FIG. 15A). At low occupancy (0.7 pMole/cell), octamer was one time more potent than tetramer, and 7 times more potent than dimer in stimulating TCR-HA T-cells (FIG. 15A). For as much as ten times higher occupancy (7.0 pMole/cell), there was an inverse relation between the valence and immunopotency of multimers. Thus, dimer was 1.5 times more potent than tetramer, whereas octamer did not stimulate the cells. At the highest occupancy (35 pMole/cell), none of ligands stimulated the cells.
The cytokine production was assessed in the cell culture supernatants after two days of continuous exposure to 10 μg/ml (FIG. 15B) and 50 μg/ml (FIG. 15C) of multimers. The cytokine values (pg/ml) are indicated as mean of duplicate wells. The ±SD for IL-2 measured in duplicate wells was 12.5 pg/ml, for IL-4 was 21 pg/ml, and for IFN-γ was 32.7 pg/ml. [0239]
The relation of valence-to-potency correlated with the pattern of cytokine secretion. At low TCR/CD4 occupancy, the valence of the multimeric paralleled the increase in IL-4 secretion (FIG. 15B), and at the highest occupancy where none of multimers stimulated the cells, the IL-2 and IL-4, secretion was barely detected. In contrast, the increase in valency paralleled the IFN-γ secretion (FIG. 15C). [0240]
TCR-HA Tg T-cells exposed for three days to 50 μg/ml of multimers under the conditions described for [0241] ³H-TdR assay were used to determine the percent of apoptosis by two-color FACS analysis using the 6.5.2 mAb-FITC and anti-Anexin V-PE conjugates as described. Results are shown in FIG. 16 (16A, control, 16B, dimer, 16C, tetramer, and 16D, octamer). The percentage of Anexin V⁺ apoptotic cells within the gated population of 6.5.2⁺ cells (upper right corner) was calculated using the CELLQuest analysis software.
The increase in IFN-γ secretion correlated with a high percentage of apoptotic TCR-HA T-cells in the case of octamer, but not in the case of dimer and tetramer (FIG. 16). In contrast, the unresponsiveness of TCR-HA T-cells induced by dimer and tetramer at low and high TCR/CD4 occupancy was not due to apoptosis but rather to anergy, since the unresponsiveness of these cells was reverted by exogenous IL-2 added to the cultures. [0242]
The extent of T-cell unresponsiveness induced by multimers was correlated with the increase in IFN-γ secretion. Cells exposed to octamer at high TCR/CD4 occupancy underwent rapid apoptosis after secreting large amounts of IFN-γ. [0243]
The foregoing results indicated that depending on the degree of TCR/CD4 occupancy, the increase in valency correlated with distinct immunomodulatory effects on cognate T-cells. Thus, at low TCR occupancy, the increase in valency paralleled its potency to induce T-cell proliferation and Th2-like cytokine secretion, whereas at high occupancy was inversely proportional with the extent of T-cell unresponsiveness and cytokine secretion. The fact that multimers (1) exhibit long life in blood circulation, (2) penetrate the lymphoid organs, (3) persist in these organs as intact molecules for as long as 72 hours, and (4) exhibit immunomodulatory effects on cognate T-cells demonstrate that this class of antigen-specific TCR/CD4 ligands can be used to modulate the effector functions of pathogenic T-cells in various infectious and autoimmune diseases. Also, the multimers are useful to identify antigen-specific CD4[0244] ⁺ memory T-cells that can persist in vivo for a long period of time at a low frequency.
Various publications are cited herein, the contents of which are hereby incorporated by reference in their entireties. [0245]
1 72 1 20 PRT Homo sapiens 1 Met Ala Ser Pro Gly Ser Gly Phe Trp Ser Phe Gly Ser Glu Asp Gly 1 5 10 15 Ser Gly Asp Ser 20 2 20 PRT Homo sapiens 2 Glu Asn Pro Gly Thr Ala Arg Ala Trp Cys Gln Val Ala Gln Lys Phe 1 5 10 15 Thr Gly Gly Ile 20 3 20 PRT Homo sapiens 3 Ser Gln Pro Pro Arg Ala Ala Ala Arg Lys Ala Ala Cys Ala Cys Asp 1 5 10 15 Gln Lys Pro Cys 20 4 16 PRT Homo sapiens 4 Met Asn Ile Leu Leu Gln Tyr Val Val Lys Ser Phe Asp Arg Ser Thr 1 5 10 15 5 15 PRT Homo sapiens 5 Phe Asp Arg Ser Thr Lys Val Ile Asp Phe His Tyr Pro Asn Glu 1 5 10 15 6 14 PRT Homo sapiens 6 Tyr Glu Ile Ala Pro Val Phe Val Leu Leu Glu Tyr Val Thr 1 5 10 7 20 PRT Homo sapiens 7 Pro Gly Gly Ser Gly Asp Gly Ile Phe Ser Pro Gly Gly Ala Ile Ser 1 5 10 15 Asn Met Tyr Ala 20 8 20 PRT Homo sapiens 8 Met Met Ile Ala Arg Phe Lys Met Phe Pro Glu Val Lys Glu Lys Gly 1 5 10 15 Met Ala Ala Leu 20 9 20 PRT Homo sapiens 9 Glu Val Lys Glu Lys Gly Met Ala Ala Leu Pro Arg Leu Ile Ala Phe 1 5 10 15 Thr Ser Glu His 20 10 15 PRT Homo sapiens 10 Pro Arg Leu Ile Ala Phe Thr Ser Glu His Ser His Phe Ser Leu 1 5 10 15 11 20 PRT Homo sapiens 11 Ser His Phe Ser Leu Lys Lys Gly Ala Ala Ala Leu Gly Ile Gly Thr 1 5 10 15 Asp Ser Val Ile 20 12 20 PRT Homo sapiens 12 Ile Pro Ser Asp Leu Glu Arg Arg Ile Leu Glu Ala Lys Gln Lys Gly 1 5 10 15 Phe Val Pro Phe 20 13 15 PRT Homo sapiens 13 Lys Tyr Lys Ile Trp Met His Val Asp Ala Ala Trp Gly Gly Gly 1 5 10 15 14 20 PRT Homo sapiens 14 Met His Val Asp Ala Ala Trp Gly Gly Gly Leu Leu Met Ser Arg Lys 1 5 10 15 His Lys Trp Lys 20 15 15 PRT Homo sapiens 15 Lys His Lys Trp Lys Leu Ser Gly Val Glu Arg Ala Asn Ser Val 1 5 10 15 16 20 PRT Homo sapiens 16 Leu Ser Gly Val Glu Arg Ala Asn Ser Val Thr Trp Asn Pro His Lys 1 5 10 15 Met Met Gly Val 20 17 20 PRT Homo sapiens 17 Val Asp Lys Cys Leu Glu Leu Ala Glu Tyr Leu Tyr Asn Ile Ile Lys 1 5 10 15 Asn Arg Glu Gly 20 18 15 PRT Homo sapiens 18 Leu Tyr Asn Ile Ile Lys Asn Arg Glu Gly Tyr Glu Met Val Phe 1 5 10 15 19 21 PRT Homo sapiens 19 Gly Lys Pro Gln His Thr Asn Val Cys Phe Trp Tyr Ile Pro Pro Ser 1 5 10 15 Leu Arg Thr Leu Glu 20 20 13 PRT Homo sapiens 20 Phe Trp Tyr Ile Pro Pro Ser Leu Arg Thr Leu Glu Asp 1 5 10 21 20 PRT Homo sapiens 21 Ile Pro Pro Ser Leu Arg Thr Leu Glu Asp Asn Glu Glu Arg Met Ser 1 5 10 15 Arg Leu Ser Lys 20 22 15 PRT Homo sapiens 22 Pro Ser Leu Arg Thr Leu Glu Asp Asn Glu Glu Arg Met Ser Arg 1 5 10 15 23 65 PRT Homo sapiens 23 Glu Arg Met Ser Arg Leu Ser Lys Val Ala Pro Val Ile Lys Ala Arg 1 5 10 15 Met Met Glu Tyr Gly Thr Thr Met Val Ser Tyr Gln Pro Leu Gly Asp 20 25 30 Lys Val Asn Phe Phe Arg Met Val Ile Ser Asn Pro Ala Ala Thr His 35 40 45 Gln Asp Ile Asp Phe Leu Ile Glu Glu Ile Glu Arg Leu Gly Gln Asp 50 55 60 Leu 65 24 15 PRT Homo sapiens 24 Ser Tyr Gln Pro Leu Gly Asp Lys Val Asn Phe Phe Arg Met Val 1 5 10 15 25 15 PRT Homo sapiens 25 Gly Asp Lys Val Asn Phe Phe Arg Met Val Ile Ser Asn Pro Ala 1 5 10 15 26 15 PRT Homo sapiens 26 Phe Phe Arg Met Val Ile Ser Asn Pro Ala Ala Thr His Gln Asp 1 5 10 15 27 15 PRT Homo sapiens 27 Ala Thr His Gln Asp Ile Asp Phe Leu Ile Glu Glu Ile Glu Arg 1 5 10 15 28 110 PRT Homo sapiens 28 Met Ala Leu Trp Met Arg Leu Leu Pro Leu Leu Ala Leu Leu Ala Leu 1 5 10 15 Trp Gly Pro Asp Pro Ala Ala Ala Phe Val Asn Gln His Leu Cys Gly 20 25 30 Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe 35 40 45 Phe Tyr Thr Pro Lys Thr Arg Arg Glu Ala Glu Asp Leu Gln Val Gly 50 55 60 Gln Val Glu Leu Gly Gly Gly Pro Gly Ala Gly Ser Leu Gln Pro Leu 65 70 75 80 Ala Leu Glu Gly Ser Leu Gln Lys Arg Gly Ile Val Glu Gln Cys Cys 85 90 95 Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn 100 105 110 29 18 PRT Homo sapiens 29 Met Ala Leu Trp Met Arg Leu Leu Pro Leu Leu Ala Leu Leu Ala Leu 1 5 10 15 Trp Gly 30 24 PRT Homo sapiens 30 Met Ala Leu Trp Met Arg Leu Leu Pro Leu Leu Ala Leu Leu Ala Leu 1 5 10 15 Trp Gly Pro Asp Pro Ala Ala Ala 20 31 16 PRT Homo sapiens 31 Leu Ala Leu Leu Ala Leu Trp Gly Pro Asp Pro Ala Ala Ala Phe Val 1 5 10 15 32 16 PRT Homo sapiens 32 Pro Ala Ala Ala Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val 1 5 10 15 33 13 PRT Homo sapiens 33 Ala Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val 1 5 10 34 19 PRT Homo sapiens 34 Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg 1 5 10 15 Gly Phe Phe 35 15 PRT Homo sapiens 35 Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly 1 5 10 15 36 16 PRT Homo sapiens 36 Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr 1 5 10 15 37 46 PRT Homo sapiens 37 Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr 1 5 10 15 Thr Pro Lys Thr Arg Arg Glu Ala Glu Asp Leu Gln Val Gly Gln Val 20 25 30 Glu Leu Gly Gly Gly Pro Gly Ala Gly Ser Leu Gln Pro Leu 35 40 45 38 20 PRT Homo sapiens 38 Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr Arg 1 5 10 15 Arg Glu Ala Glu 20 39 17 PRT Homo sapiens 39 Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr Arg Arg Glu Ala Glu 1 5 10 15 Asp 40 20 PRT Homo sapiens 40 Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr Arg Arg Glu Ala Glu 1 5 10 15 Asp Leu Gln Val 20 41 16 PRT Homo sapiens 41 Pro Lys Thr Arg Arg Glu Ala Glu Asp Leu Gln Val Gly Gln Val Glu 1 5 10 15 42 16 PRT Homo sapiens 42 Val Gly Gln Val Glu Leu Gly Gly Gly Pro Gly Ala Gly Ser Leu Gln 1 5 10 15 43 18 PRT Homo sapiens 43 Gly Ala Gly Ser Leu Gln Pro Leu Ala Leu Glu Gly Ser Leu Gln Lys 1 5 10 15 Arg Gly 44 17 PRT Homo sapiens 44 Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr Arg Arg Glu Ala Glu 1 5 10 15 Asp 45 9 PRT Homo sapiens 45 Leu Ala Leu Glu Gly Ser Leu Gln Lys 1 5 46 17 PRT Homo sapiens 46 Ser Leu Gln Lys Arg Gly Ile Val Glu Gln Cys Cys Thr Ser Ile Cys 1 5 10 15 Ser 47 27 PRT Homo sapiens 47 Gly Ser Leu Gln Lys Arg Gly Ile Val Glu Gln Cys Cys Thr Ser Ile 1 5 10 15 Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn 20 25 48 15 PRT Homo sapiens 48 Cys Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn 1 5 10 15 49 12 PRT Homo sapiens 49 Ala Leu Thr Gly Thr Glu Lys Leu Ile Glu Thr Tyr 1 5 10 50 14 PRT Homo sapiens 50 Tyr Thr Thr Gly Ala Val Arg Gln Ile Phe Gly Asp Tyr Lys 1 5 10 51 14 PRT Homo sapiens 51 Thr Val Thr Gly Gly Gln Lys Gly Arg Gly Ser Arg Gly Gln 1 5 10 52 14 PRT Homo sapiens 52 Ser Ile Gly Ser Leu Cys Ala Asp Ala Arg Met Tyr Gly Val 1 5 10 53 18 PRT Homo sapiens 53 Gly Phe Tyr Thr Thr Gly Ala Val Arg Gln Ile Phe Gly Asp Tyr Lys 1 5 10 15 Thr Thr 54 21 PRT Homo sapiens 54 Thr Gly Thr Glu Lys Leu Ile Glu Thr Tyr Phe Ser Lys Asn Tyr Gln 1 5 10 15 Asp Tyr Glu Tyr Leu 20 55 11 PRT Homo sapiens 55 Ser Phe Glu Arg Phe Glu Ile Phe Pro Lys Glu 1 5 10 56 12 PRT Homo sapiens 56 Gly Gly Gly Ser Gly Gly Gly Ser Gly Gly Gly Ser 1 5 10 57 18 DNA Homo sapiens 57 atcaaagaag aacatgtg 18 58 18 DNA Homo sapiens 58 ccagtgcttg agaagagg 18 59 18 DNA Homo sapiens 59 ggggacaccc gaccacgt 18 60 18 DNA Homo sapiens 60 ccattccact gtgagagg 18 61 18 DNA Homo sapiens 61 gaagacattg tggctgac 18 62 18 DNA Homo sapiens 62 aggctcatcc aggcccca 18 63 18 DNA Homo sapiens 63 agagactctc ccgaggat 18 64 18 DNA Homo sapiens 64 ctccacgtgg caggtgta 18 65 33 DNA Homo sapiens 65 acggcttaag actcccaaca gagcgcccaa gaa 33 66 33 DNA Homo sapiens 66 gccgtctaga ttaccagtgc ttgagaagag gct 33 67 31 DNA Homo sapiens 67 atcggaattc ctcctctggc ccctggtcct g 31 68 79 DNA Homo sapiens 68 cgatggatcc accacctcct agtgaaaagt gtgaatgttc tgatgtaaat gcaattagtc 60 gtggtcgggt gtccccagc 79 69 56 DNA Homo sapiens 69 atcgggatcc ggtggagggg gaagtggagg tggagggtct ccacgtttct tggagc 56 70 28 DNA Homo sapiens 70 gaggactagt caggcttggg tgctccac 28 71 49 DNA Homo sapiens 71 actgactagt cctctcacag tggaatgggc agagcccaaa tcttgtgac 49 72 34 DNA Homo sapiens 72 cagtaagctt tcatttaccc ggagacaggg agag 34

Claims

We claim:

1. A chimeric protein heterodimer comprising a first and a second subunit, wherein (a) the first subunit comprises a protein comprising (i) at least one domain of a human immunoglobulin constant region; (ii) at least 90 percent of the extracellular domains of a first Class II Human Leukocyte Antigen component selected from the group consisting of an alpha chain and a beta chain; and (iii) a peptide comprising an epitope associated with a human autoimmune disease; and (b) the second subunit comprises a protein comprising at least 90 percent of the extracellular domains of a second Class II Human Leukocyte Antigen component selected from the group consisting of an alpha chain and a beta chain, wherein if the first component is a beta chain, the second component is an alpha chain and if the first component is an alpha chain, the second component is a beta chain.

2. A dimer comprising two chimeric proteins according to claim 1.

3. The dimer of claim 2, where the two chimeric proteins are joined by disulfide bonds.

4. The dimer of claim 2, where the two chimeric proteins are joined by chemical crosslinking other than a disulfide bond.

5. A multimer comprising at least three chimeric proteins according to claim 1.

6. The multimer of claim 5, wherein the chimeric proteins are joined by chemical crosslinking.

7. The multimer of claim 6, where the chemical crosslinker is a polyalkylene glycol.

8. The chimeric protein of claim 1, linked to a toxic agent.

9. The chimeric protein of claim 8, where the toxic agent is doxorubicin.

10. A chimeric protein heterodimer comprising a first and a second subunit, wherein (a) the first subunit comprises a protein comprising (i) at least one domain of a human immunoglobulin constant region; (ii) at least 90 percent of the extracellular domains of a first Class II Human Leukocyte Antigen component selected from the group consisting of an alpha chain and a beta chain; and (iii) a peptide comprising an epitope associated with autoimmune diabetes; and (b) the second subunit comprises a protein comprising at least 90 percent of the extracellular domains of a second Class II Human Leukocyte Antigen component selectee from the group consisting of an alpha chain and a beta chain, wherein if the first component is a beta chain, the second component is an alpha chain and if the first component is an alpha chain, the second component is a beta chain.

11. A dimer comprising two chimeric proteins according to claim 10.

12. The dimer of claim 11, where the two chimeric proteins are joined by disulfide bonds.

13. The dimer of claim 11, where the two chimeric proteins are joined by chemical crosslinking other than a disulfide bond.

14. A multimer comprising at least three chimeric proteins according to claim 1.

15. The multimer of claim 14, wherein the chimeric proteins are joined by chemical crosslinking.

16. The multimer of claim 15, where the chemical crosslinker is a polyalkylene glycol.

17. The chimeric protein of claim 10, linked to a toxic agent.

18. The chimeric protein of claim 17, where the toxic agent is doxorubicin.

19. A chimeric protein heterodimer comprising a first and a second subunit, wherein (a) the first subunit comprises a protein comprising (i) at least one domain of a human immunoglobulin constant region; (ii) at least 90 percent of the extracellular domains of a first Class II Human Leukocyte Antigen component selected from the group consisting of an alpha chain and a beta chain; and (iii) a peptide comprising a peptide fragment of human glutamic acid decarboxylase 65; and (b) the second subunit comprises a protein comprising at least 90 percent of the extracellular domains of a second Class II Human Leukocyte Antigen component selected from the group consisting of an alpha chain and a beta chain, wherein if the first component is a beta chain, the second component is an alpha chain and if the first component is an alpha chain, the second component is a beta chain.

20. A dimer comprising two chimeric proteins according to claim 19.

21. The dimer of claim 20, where the two chimeric proteins are joined by disulfide bonds.

22. The dimer of claim 20, where the two chimeric proteins are joined by chemical crosslinking other than a disulfide bond.

23. A multimer comprising at least three chimeric proteins according to claim 19.

24. The multimer of claim 23, wherein the chimeric proteins are joined by chemical crosslinking.

25. The multimer of claim 24, where the chemical crosslinker is a polyalkylene glycol.

26. The chimeric protein of claim 19, linked to a toxic agent.

27. The chimeric protein of claim 26, where the toxic agent is doxorubicin.

28. The chimeric protein of claim 19, wherein the Class II Human Leukocyte Antigen beta chain is of serotype HLA-DR4.

29. The chimeric protein of claim 19, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

30. The chimeric protein of claim 20, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

31. The chimeric protein of claim 21, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

32. The chimeric protein of claim 22, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

33. The chimeric protein of claim 23, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

34. The chimeric protein of claim 24, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

35. The chimeric protein of claim 25, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

36. The chimeric protein of claim 26, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

37. The chimeric protein of claim 27, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

38. The chimeric protein of claim 28, wherein the glutamic acid decarboxylase peptide fragment comprises the sequence PRLIAFTSEHSHFSL (SEQ ID NO:10).

39. A chimeric protein heterodimer comprising a first and a second subunit, wherein (a) the first subunit comprises a protein comprising (i) at least one domain of a human immunoglobulin constant region; (ii) at least 90 percent of the extracellular domains of a first Class II Human Leukocyte Antigen component selected from the group consisting of an alpha chain and a beta chain; and (iii) a peptide comprising a peptide fragment of human proinsulin; and (b) the second subunit comprises a protein comprising at least 90 percent of the extracellular domains of a second Class II Human Leukocyte Antigen component selected from the group consisting of an alpha chain and a beta chain, wherein if the first component is a beta chain, the second component is an alpha chain and if the first component is an alpha chain, the second component is a beta chain.

40. A dimer comprising two chimeric proteins according to claim 39.

41. The dimer of claim 40, where the two chimeric proteins are joined by disulfide bonds.

42. The dimer of claim 40, where the two chimeric proteins are joined by chemical crosslinking other than a disulfide bond.

43. A multimer comprising at least three chimeric proteins according to claim 40.

44. The multimer of claim 43, wherein the chimeric proteins are joined by chemical crosslinking.

45. The multimer of claim 44, where the chemical crosslinker is a polyalkylene glycol.

46. The chimeric protein of claim 39, linked to a toxic agent.

47. The chimeric protein of claim 46, where the toxic agent is doxorubicin.

48. The chimeric protein of claim 39, wherein the Class II Human Leukocyte Antigen beta chain is of serotype HLA-DR4.

49. The chimeric protein of claim 39, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

50. The chimeric protein of claim 40, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

51. The chimeric protein of claim 41, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

52. The chimeric protein of claim 42, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

53. The chimeric protein of claim 43, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

54. The chimeric protein of claim 44, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

55. The chimeric protein of claim 45, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

56. The chimeric protein of claim 46, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

57. The chimeric protein of claim 47, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

58. The chimeric protein of claim 48, wherein the proinsulin peptide fragment comprises the sequence GAGSLQPLALEGSLQKRG (SEQ ID NO: 43).

59. The chimeric protein of claim 39, wherein the proinsulin peptide fragment comprises the sequence LALEGSLQK (SEQ ID NO: 45).

60. The chimeric protein of claim 48, wherein the proinsulin peptide fragment comprises the sequence LALEGSLQK (SEQ ID NO: 45).

61. A method of identifying a predisposition to an autoimmune disorder in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 1, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

62. A method of identifying a predisposition to an autoimmune disorder in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 2, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

63. A method of identifying a predisposition to an autoimmune disorder in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 5, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

64. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 10, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

65. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 11, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

66. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 14, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

67. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 19, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

68. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 20, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

69. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 23, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

70. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 28, where the HLA class II components in the chimeric protein are compatible with HLA class I determinants of the subject.

71. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 29, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

72. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 30, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

73. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 33, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

74. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 38, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

75. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 39, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

76. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 40, where the HLA class II components in the chimeric protein are compatible with HLA class I determinants of the subject.

77. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 43, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

78. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 48, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

79. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 49, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

80. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 50, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

81. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 58, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

82. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 59, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

83. A method of identifying a predisposition to autoimmune diabetes in a human subject comprising detecting the presence of T cells in the subject which bind to a chimeric protein according to claim 60, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

84. A method of treating a human subject suffering from an autoimmune disorder comprising administering, to the subject, an effective amount of a chimeric protein according to claim 1, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

85. A method of treating a human subject suffering from an autoimmune disorder comprising administering, to the subject, an effective amount of a chimeric protein according to claim 2, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

86. A method of treating a human subject suffering from an autoimmune disorder comprising administering, to the subject, an effective amount of a chimeric protein according to claim 5, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

87. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 10, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

88. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 11, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

89. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 14, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

90. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 19, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

91. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 20, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

92. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 23, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

93. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 28, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

94. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 29, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

95. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 30, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

96. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 33, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

97. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 38, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

98. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 39, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

99. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 40, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

100. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 43, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

101. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 48, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

102. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 49, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

103. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 50, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

104. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 58, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

105. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 59, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.

106. A method of treating a human subject suffering from autoimmune diabetes comprising administering, to the subject, an effective amount of a chimeric protein according to claim 60, where the HLA class II components in the chimeric protein are compatible with HLA class II determinants of the subject.