WO1993007291A1 - Immunoglobulin d-associated glycans and uses thereof - Google Patents

Abstract

The receptor for IgD (IgD-R) on T lymphocytes has lectin-like properties and binds to carbohydrate moieties of IgD. Thus IgD-associated glycans bind to the IgD-R. Therefore, IgD-associated glycans, neoglycans comprising the IgD-associated glycans, and pharmaceutical compositions comprising these glycans are provided. The glycans and pharmaceutical compositions are useful in a method for inhibiting the binding of IgD to an IgD receptor, a method for inhibiting an immune response involving the interaction of T and B lymphocytes in a subject, and a method for treating a subject with an immune-mediated disease. Because the glycans can upregulate expression of IgD-R, they are also useful in a method for increasing the number of IgD receptors on the surface of a T lymphocyte, and in a method for enhancing an immune response in a subject being immunized with an antigen or vaccine.

Description

IMMUNOGLOBULIN D-ASSOCIATED GLYCANS AND USES THEREOF

BACKGROUND OF THE INVENTION

This work was supported by several U.S. government grants, including PHS Grants AI-22645, RR-03060, AG-04980, AG-04860 and AI-26113. The U.S. government therefore has certain rights in this invention.

Field of the Invention

The invention in the fields of immunology, biochemistry and medicine relates to glycans associated with IgD and their use in the stimulation of IgD receptor upregulation, in the augmentation or inhibition of immune responses and in the treatment of immune-mediated diseases.

Description of the Background Art

Immunoglobulin D (IgD) exists in both membrane-associated (surface IgD or sIgD) form and in secreted form. While the biological function of serum IgD, present at minute concentrations, is unknown, the membrane-bound form appears to play a role in a number of immunoregulatory processes (discussed below). The δ heavy chain of human IgD has four domains: Vδ, Cδ1, Cδ2 and Cδ3. An exceptionally long hinge region is located between Cδl and Cδ2. Like all other human immunoglobulins, human IgD is a glycoprotein and contains carbohydrate in its heavy chain (Spiegelberg, H. et al., Biochemistry 9:2115-2122 (1970)). The human δ chain contains asparagine-linked oligosaccharides at residue 354 in Cδ1 and at 445 and 496 in Cδ3. There are also multiple 0-linked oligosaccharides attached to serine and threonine residues in the hinge region (Takahashi, N. et al., Proc. Natl. Acad. Sci. USA 79:2850-2854 (1982)). The structures of the oligosaccharides at the three asparagine-linked glycosyl- ation sites (Mellis, S.J. et al., J. Biol. Chem. 258:11546- 11556 (1983)) and the O-glycosylation sites (Mellis, S.J. et al., J. Biol. Chem. 258:11557-11563 (1983)) of human IgD have been characterized in detail. Murine IgD also has a high content of N-linked glycans (Vasilov, R.G. et al., Eur. J. Immunol. 12:804-818 (1982); Argon, Y. et al., J. Immunol. 133:1627-1633 (1984)). However, virtually nothing was known about the role of these glycans in the biological function of IgD and of lymphocytes prior to the present invention.

IgD is expressed on the majority of mature B lympho-cytes (Rowe, D.S. et al., J. Exp. Med. 138:965-972 (1973)), although its role in the humoral immune response is still unclear. The present inventors' laboratory has recently shown the presence of receptors for IgD (IgD-R) on T cells from mice harboring IgD-secreting plasmacytomas (TEPC-1017 or TEPC-1033) or in mice injected with purified IgD produced by these plasmacytomas. Such mice exhibit significantly enhanced antibody responses of all antibody isotypes except IgD (Xue, B. et al., J. EXP. Med. 159:103-113 (1984), Swenson, CD. et al., Eur. J. Immunol. 18:13-20 (1988)). Such an augmented antibody response can be transferred from IgD- treated mice to normal mice by transfer of CD4⁺, Lyt-1⁺, CD8-T cells (Coico, R.F. et al., J. Exp. Med. 162:1852-1861 (1985)). This is the same subset of T cells that also express IgD-R on their surface, measured by their capacity to form rosettes with IgD-coated sheep erythrocytes (IgD-SRBC). These T cells have been termed Tδ cells, in keeping with the nomenclature of T cells with Fc receptors which recognize other immunoglobulin (Ig) isotypes. (Coico, R.F. et al., Nature 316:744-746 (1985)). These findings indicate that the surface IgD molecule functions as a cell membrane receptor involved in T cell-B cell interactions (Coico, R.F. et al.. Immunol. Rev. 105:45-67 (1988), Amin, A.R. et al., Research Immunol. 141:94-99 (1990)).

The mechanism by which Tδ cells augment B cell responses is unclear. Soluble IgD-binding factors released by Tδ cells may contribute to their immunoaugmenting properties (Adachi, M. et al., Proc. Natl. Acad. Sci. (USA) 85:554-558 (1988)). Tδ cells appear to interact more efficiently with slgD-bearing B cells subsequent to antigen-induced cross-linking of slgD molecules. This notion is based on observations from the present inventors' laboratory that (1) both primary and secondary antibody responses are augmented by injection of IgD prior to the first injection of antigen (Xue, B. et al., J. EXP. Med. 159:103-113 (1984)), and (2) B cells with cross-linked slgD induce upregulation of IgD-R on T cells in vivo and in vitro (see below). Cross-linking of slgD with the Cδ3 specific monoclonal antibody Hδ^a/1 causes such IgD-R upregulation. Since this antibody could sterically hinder interaction between IgD-R and the Cδ3 domain of IgD, a role for Cδ1 in this IgD-R upregulation is likely.

Fc receptors (FcR) for other immunoglobulin isocypes (IgG, IgM, IgA and IgE) are present on cells of the lymphohemopoietic system such as macrophages, granulocytes and lymphocytes. Some FcR's trigger cellular functions such as phagocytosis, antibody-dependent cytotoxicity, and the secretion of potent mediators (Metzger, H. et al., FASEB J. 2:3-8 (1988); Kinet, P.E. Cell 57:351-353 (1989)). The Fc -R principally recognizes the C 2-hinge region of IgG (Revetch, J.V. et al.. in Metzger. H. (ed.) Fc Receptors and Action of Antibodies. Amer. Soc. Microbiol., Washington, DC, 1990, pp. 211-235), whereas FceRI recognizes residues 301-376 roughly centered in the interface between C∈2 and C∈3 domains of IgE (Kinet, J.P. et al., in Metzger. H. (ed.) Fc Receptors and Action of Antibodies. Amer. Soc. Microbiol., Washington, DC, 1990, pp. 235-259). The FcμR recognizes the Cμ3 region of IgM (Mathur, A. et al., J. Immunol. 140:143-147 (1988), Mathur, A. et al., J. Immunol. 141:1855-1862 (1988)).

The specificity of IgD-R of Tδ cells for immunoglobulin of the IgD class was established by showing that IgD, at concentrations >120 μg/ml (1.0 μM), competitively inhibited the rosetting of Tδ cells with IgD- SRBC, whereas IgM, IgG1, IgG2, IgG3, IgA, and IgE failed to inhibit this interaction (Coico, R.F. et al., Nature 316:744-746 (1985), Coico, R.F. et al., Immunol. Rev. 105:45-67 (1988)).

Exposure of T cells either in vitro or in vivo to either (a) oligomeric secreted IgD molecules (TEPC-1017 or TEPC-1033), (b) antigen-crosslinked monomeric secreted IgD (such as B1-8.δ1, a monoclonal antibody of IgD isotype) or (c) B-cell surface IgD which is crosslinked (by either anti-IgD or anti-Ig), results in upregulation of their IgD-R (Coico, R.F. et al., Proc. Natl. Acad. Sci. (USA) 11:559-563 (1988)). B cells with cross-linked sIgM do not induce IgD-R upregulation. The cytokines IL-2, IL-4 and IFN- , known to participate in the immune response and in host defenses, also induce IgD-R upregulation on CD4⁺ T lymphocytes (including cloned T cells) (Coico, R.F. et al., Immunol. Rev. 105:45-67 (1988), Coico, R.F. et al., J. Immunol. 138:4-6 (1987), Amin, A.R. et al., Proc. Natl . Acad. Sci. (USA) 85:9179-9183 (1988)). Within 1-2 hours of exposure to oligomeric, aggregated or antigen-cross-linked monomeric secreted IgD, approximately 25-35% of splenic murine helper T cells (Coico, R.F. et al., Nature 316:744-746 (1985)) and 10-15% of human peripheral blood T cells (Coico, R.F. et al., J. Immunol. 145: 3556-3561 (1990)) express IgD-R, manifested by increased rosette formation with IgD-RBC. SUMMARY OF THE INVENTION

The present inventors have found that IgD-R on T-helper cells are not exclusively Fc receptors, but also bind equally well to the Fd piece of IgD, which consists of the Cδ1 and the Cδ hinge domains. This domain specificity could not be explained by the amino acid sequence, as these domains have only 26% homology (Tucker, P.W. et al., Ann. NY Acad. Sci. 399:26-40 (1982)).

Because of the absence of a significant degree of homology in the protein sequences of the two IgD domains that bound to the IgD-R, the present inventors turned their attention to the role of the IgD glycans in the interaction between IgD and the IgD-R. This approach was based in part on the present inventors' discovery, described in detail below, that mutant IgD molecules, containing either only the Cδ1 or only the Cδ3 domain, each bind to the lectin

Griffonia simplicifolia-1 (GS-1). This lectin is known to bind specifically to N-linked glycans from murine IgD

(Oppenheim, J.D. et al., J. Immunol. Methods. 130:243-250 (1990)).

The present inventors discovered that the binding of IgD to the IgD-R is blocked by several preparations, including:

(a) boiled IgD; (b) low molecular weight fragments from protease digested, boiled IgD molecules; (c) N-linked glycans isolated from IgD; and (d) Gal, GalNac, GlcNAc and neoglycoproteins containing these sugars (an explanation of all abbreviations appears below).

In contrast, the binding of IgD to the IgD-R was not blocked by the following preparations:

(a) deglycosylated IgD;

(b) low Mr fragments from IgM, IgA, or IgG2a,

(c) the monosaccharides or disaccharides mannose, glucose, fucose, mellibiose, lactose or neoglycoproteins containing these saccharides or their derivatives; or

(d) 23 of 25 Gal/GalNAc-rich bacterial polysaccharides tested.

Thus, the present invention provides a method for inhibiting the binding of IgD with an IgD receptor comprising providing to a cell having an IgD receptor an effective amount of an IgD-associated glycan.

In another embodiment, the invention provides a method for inhibiting an immune response involving the interaction of T and B lymphocytes in a subject, comprising administering to the subject an effective amount of an IgD-associated glycan, thereby inhibiting the response.

Also provided is a method for treating a subject with an immune-mediated disease, comprising administering to the subject an effective amount of an IgD-associated glycan, thereby treating the disease. The diseases for which this method is useful include autoimmune diseases in which antibody formation is enhanced such as systemic lupus erythematosus, rheumatoid arthritis, thyroiditis and the like.

The present invention is further directed to a method for increasing the number of IgD receptors on the surface of a T lymphocyte having such receptors, comprising contacting the cell with an effective amount of an IgD-associated glycan, thereby increasing the number of the receptors.

The present invention includes a method for inhibiting an immune response in a subject, comprising:

(a) obtaining lymphocytes including CD8+ T lymphocytes from the subject;

(b) increasing the number of IgD receptors on the surface of the CD8+ T lymphocytes by treating the lymphocytes with the above IgD-R upregulating method; and

(c) administering the treated CD8+ T lymphocytes to the subject,

thereby inhibiting the immune response.

In another embodiment, the above method is used for enhancing an immune response in a subject, the method comprising:

(a) obtaining lymphocytes including CD4+ T lymphocytes from the subject;

(b) increasing the number of IgD receptors on the surface of the CD4+ T lymphocytes by treating the lymphocytes by the above IgD-R upregulating method; and

(c) administering the treated CD4+ T lymphocytes to the subject,

thereby enhancing the immune response.

In yet another embodiment, the invention provides a method for enhancing an immune response in a subject being immunized with an antigen or vaccine, comprising administering to the subject, in combination with the. antigen or vaccine, an effective amount of an IgD-associated glycan.

In the methods for inhibiting an immune response or treating an immune-mediated disease, the glycan is preferably in non-polymerized or non-aggregated form. In the methods for increasing the number of IgD-R on T cells or in stimulating an immune response, the glycan is preferably in polymerized or aggregated form, or is a neoglycan.

Also provided is a neoglycan comprising two or more of the same or different IgD-associated glycan monomers linked to a polymer. Such a neoglycan may be a neoglycoprotein, wherein the polymer is a protein. In a preferred neoglycoprotein, the protein is serum albumin.

In all of the above glycans or neoglycans, and in the methods described above, the glycan is preferably one selected from the group consisting of:

and

wherein R¹ is NeuAc in an α2→3 linkage or H, R² is NeuAc in an α2→6 linkage or H, R³ is GlcNAc in a β1→4 linkage or H,

R⁴ is Fuc in an α1→6 linkage or H and R⁵ is Glc in an a1→3 linkage or H.

The invention is further directed to pharmaceutical compositions comprising the glycans and neoglycans described above and to the use of the pharmaceutical compositions in the methods described above. BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic representation of the constant region of the δ chain of human IgD (WAH) (Mellis et al., supra). Essential structural features of the heavy chain constant region are depicted with emphasis on oligosaccharide structure and localization (after F.W. Putnam et al., Ann. N.Y. Acad. Sci. 399:41-68 (1982)). The GalNAc-rich segment of the hinge is shown in an expanded view. Symbols (+) and (-) represent the highly charged segment of the hinge which is extremely sensitive to proteolytic degradation. Microheterogeneity of N- and O-glycosidically linked oligosaccharides is reflected by the presence or absence (±) of terminal sugar residues. Asn 354, 445 and 496 refer to locations of the N-glycosidically linked oligosaccharides. These residue numbers were calculated beginning at the N-terminus of the δ chain (SEQ ID NO:4). Their positions with respect to the first residue of the constant region is also indicated schematically below. TP refers to the tail piece of the secreted form of IgD. Relative sizes of oligosaccharide units and protein domains are not drawn to scale.

Figure 2 shows the structure of the triantennary high mannose oligosaccharides found linked to Asn354 of human IgD (WAH) (Mellis et al., supra), and indicates the abundance of various species at this site.

Figure 3 shows the structure of the dibranched complex and bisected dibranched complex oligosaccharides found linked to Asn445 of human IgD (WAH) (Mellis et al., supra), and indicates the abundance of various species at this site.

Figure 4 shows the structure of the dibranched complex and bisected dibranched complex oligosaccharides found linked to Asn496 of human IgD (WAH) (Mellis et al., supra), and indicates the abundance of various species at this site.

Figure 5 is a graph showing cross-inhibition of rosette formation by Gen.24 and KWD6 IgD. Indicator SRBC were coupled to mutant myeloma protein Gen.24 and mutant hybridoma IgD, KWD6. IgD-R⁺T cells were incubated with the given concentration of either Gen.24 or KWD6 and then rosetted with Gen.24-SRBC or KWD6-SRBC in the presence of the mutant proteins as blocking agents. Results are expressed as Mean (±SD) % Blocking of RFC (n=3). In the absence of inhibitors, the %IgD-RFC was 27-32% after subtraction of BSA-RFC values.

Figure 6 shows gel patterns of SDS-PAGE analysis of affinity-purified, naturally-degraded TEPC-1017 IgD fragments. Partially degraded IgD fragments were subjected to SDS-PAGE (7%) under non-reducing conditions (lanes A-C). Lane A: naturally degraded TEPC-1017 IgD molecules showing fragment sizes of 130, 100, 70 and 66 kDa.

Lane B: 4M MgCl₂ eluate of anti-Fcδ (Hδ^a/1)-adherent TEPC-1017 IgD fragments showing fragment sizes of 130 and 100 kDa.

Lane C: Hδ^a/1non-adherent Fabδ fragments showing molecular weights of 70 and 66 kDa.

Lane D: intact TEPC-1017 IgD molecules, reducing conditions, showing heavy (54 kDa) and light (25 kDa) chain molecules.

Lane E: same as C, but under reducing conditions.

Lanes F and G: 9% SDS-PAGE of Fabδ fragments, Western blotted and probed with anti-Fab (Lane F) and anti-IgD (Lane G). Positions of molecular weight markers given in kDa.

Figure 7 is a graph showing the induction of murine T cell IgD-R by aggregated mutant IgD molecules or IgD fragments. Murine splenic T cells were incubated with various aggregated mutant IgD molecules or IgD fragments (25 μg/ml) for 3-4 hrs, washed and rosetted with IgD-SRBC. Note upregulation of IgD-R with each of the proteins, n = 4

Figure 8 is a graph (A) and dot blot (B) showing the inability of deglycosylated IgD to inhibit IgD-rosetting by induced splenic T cells. The competitive blocking of IgD-RFC was performed using 50 μg of partially deglycosylated (DG-IgD = IgD lacking N-linked glycans) and control IgD per assay, respectively (Panel A). One μg of IgD, DG-IgD or IgG was dot blotted onto nitrocellulose paper in sets of three (Panel B). Set 1 was stained with Coomassie blue R-250. Set 2 was reacted with phosphatase-conjugated sheep anti-mouse IgD (Pel-Freeze). Set 3 was reacted with peroxidase-conjugated GS-1.

Figure 9 is a graph showing the carbohydrate specificity of IgD-R. BALB/c splenic T cells induced to express IgD-R with IL-4 (10 U/ml) were treated with various carbohydrates as blocking agents at the doses indicated. The % Inhibition was calculated as described in Example I. IL-4- induced Tδ cells rosetted with IgD-coated RBC in the absence of any blocking agent had a mean IgD-RFC value of 25 ± 2% Results are the mean ± standard error of three determinations.

Figure 10 is a graph showing competitive inhibition of human IgD-rosetting by low molecular weight IgD fragments. Human PBL were isolated from blood, and incubated with plastic dishes coated with cross-linked human IgD to upregulate human IgD-R. These cells were then rosetted with IgD-coated Ox-RBC in the presence or absence of various concentrations of potential inhibitors consisting of proteinase K (PK) digested IgD or IgG. n = 3

Figure 11 is a gel pattern from a 4-20 % SDS PAGE indicating the total digestion of human IgG and IgD by Proteinase K. Five mg samples of human IgD (160 kD) or IgG (150 kD) purified from myeloma serum were each digested with 2 mg of proteinase K for 12 hrs prior to analysis. Note that the immunoglobulins have been completely digested into low molecular weight (<14 kDa) fractions.

Figure 12 is a graph showing that heat-denatured IgD inhibits binding to IgD-R. Human PBL were first incubated with cross-linked human IgD to upregulate IgD-R, and then incubated with 50 or 100 μg control or heat-denatured IgD

(boiled 2-3 minutes, cooled at room temperature) Rosetting was with IgD-Ox-RBC (n = 3). DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention is based on the present inven¬tors' unexpected discovery that, among cellular receptors for immunoglobulin, most of which are classical Fc receptors (because they bind the Fc portion of the appropriate immunoglobulin chain), the receptor for IgD (IgD-R) is unique. This uniqueness is characterized by the absolute requirement for the presence of carbohydrate on the IgD molecule for binding to the IgD-R.

The present inventors have therefore discovered that a ligand, likely the true physiological ligand, for the IgD-R is a glycan structure. On the basis of this unexpected specificity, the present inventors have conceived of methods for using glycans to modulate immune responses.

The term "glycan," as used herein, means an unsubsti¬tuted, although possibly branched, carbohydrate, usually a polysaccharide, which represents the carbohydrate portion of a glycoside after it is separated (e.g., by hydrolysis) from the non-carbohydrate portion of the molecule.

By the term "IgD-associated glycan" is intended any glycan structure found linked to an IgD molecule, as well as any structural variant or chemical derivative thereof capable of binding to an IgD-R and disrupting the binding of the IgD-R to IgD (or an appropriate IgD-like ligand). One of ordinary skill in the art will know how to remove or add saccharide residues to the glycan structures natively associated with IgD, using standard chemical or enzymatic methods well known in the art (Whistler, R. et al. (eds.), Methods in Carbohydrate Chemistry. Volumes I-V, Academic Press, New York, 1962).

A "neoglycan" is a non-natural glycan-containing structure, such as, for example, a protein, polyamino acid, or other polymer to which one or more glycan monomers are linked. A preferred neoglycan is a "neoglycoprotein," which comprises a protein to which is linked one or more glycan monomers not naturally linked to the protein. In other embodiments, the neoglycan may comprise a non-protein polymer backbone. Also intended as neoglycans are liposomes (described below) into which the glycans or neoglycoproteins of interest are incorporated. A preferred neoglycoprotein is a serum albumin to which are attached two or more IgD-associated glycan monomers.

"Glycoproteins," of which IgD is an example, are polymers which have one or more carbohydrate chains attached to a polypeptide. Glycoproteins may contain from 4% (by weight) to more than 60% carbohydrate. The carbohydrate chains may be "oligosaccharides" having 2-10 carbohydrate residues or short "polysaccharides," usually having 10-25 residues, although some are as large as 150 residues. However, it is common in the art to term chains of 10-25 carbohydrate residues on glycoproteins as "oligosaccharides" rather than "polysaccharides" (Schachter, H. Clin. Biochem. 17, 3-14 (1984)). If the polypeptide chain of a glycoprotein is degraded to an oligopeptide, for example, by proteolytic digestion or other means, the residual molecule with carbohydrate still attached is called a "glycopeptide."

The linkage between the oligosaccharide and the protein is a glycosidic linkage, the result of a condensation (dehydration) reaction between an amino acid side chain on the protein and the anomeric carbon on the first residue of the oligosaccharide. In mammalian glycoproteins, carbohydrate linkages to protein are either N-glycosides (carbohydrate linked to the amido nitrogen of asparagine) or O-glycosides (carbohydrate linked to the hydroxyl oxygen of serine, threonine or, rarely, hydroxy- lysine). N-glycosides are more commonly found in mammalian glycoproteins than O-glycosides, but a single glycoprotein may have multiple chains, some of which are O-glycosides and some of which are N-glycosides.

The carbohydrates in O-glycosides are usually short

(2-10 residues), but longer chains may also occur. The sugar directly linked to the protein is frequently

N-acetylgalactosamine or galactose. The carbohydrates in

N-glycosides are generally somewhat longer (7-25 residues) than in O-glycosides.

For general references on glycoprotein terminology, see, for example, Bennington, J.L., (ed.), Saunders Dictionary & Encyclopedia of Laboratory Medicine and Technology. W.B. Saunders Co., Philadelphia, 1984; Schachter, H. Clin. Biochem. 17, 3-14 (1984); "Abbreviated terminology of oligosaccharide chains," IUB-IUPAC Recommendations - 1980, J. Biol. Chem. 257 3347-3351 (1982)).

Part of the complexity of glycoproteins arises from the many different sugar-to-sugar linkages found in the carbohydrate side chains. These linkages will be represented herein in a highly stylized shorthand. The conventions observed in these representations are:

1. The neutral sugar monomers (monosaccharides) in a polymer will usually be represented by a three letter abbreviation, e.g., Glc, glucose; Gal, galactose, etc. (see list below). For clarity, the common configuration and ring size of the sugar is implied in the abbreviation. Thus, Glu is D-glucopyranose (6-membered ring), Fuc is L-fucopyranose, etc.

2. Acidic or basic groups on a monosaccharide will be represented by additional letters appended to the three-letter abbreviation above, e.g., GlcNAc, N-acetylglucosamine, or 5-N-acetylneuraminic acid, abbreviated NeuAc.

3. The bond between two monosaccharides will be represented by a 4-character group which includes the anomeric configuration of the sugar on the left (either a or β), the carbon number of the anomeric carbon in the bond (sugar on the left), an arrow pointed toward the hydroxyl carbon in the bond, and the carbon number of the hydroxyl carbon in the bond (sugar on the right). Thus, Manβ1→4GlcNAc represents a bond between C-1 of β-D-mannose and C-4 of N-acetylglucosamine. References to specific linkages (as in the description of the specificity of an enzyme) may be written in the same manner, e.g., a (β1→3) -linkage, or simplified (β1,3- specific).

4. Unbranched structures will be presented as a string of residues and bonds without parentheses:

Manβ1→4GlcNAc/β1→4GlcNAc

5. For branched structures, when represented in linear text, where two residues are attached to different hydroxyls of the same sugar, one of the two will be written in parentheses:

Manα1→3 (Manα1→6)Man

(two mannoses linked to a single mannose, one at the C-3 hydroxyl and the other at the C-6 hydroxyl).

The following is a list of abbreviations used herein: Abbreviation Monosaccharide

Fru D-Fructose

Fuc or F L-Fucose

Gal or GL D-Galactose

GalNAc N-Acetyl-D-galactosamine

Glc or G D-Glucose

GlcA D-Glucuronic acid

GlcN D-Glucosamine

GlcNAc or GN N-Acetyl-D-glucosamine

Hex Hexose (non-specific)

Man or M D-Mannose

ManNAc N-Acetyl-D-Mannosamine

NeuAc N-Acetylneuraminic acid (sialic acid)

All asparagine-linked oligosaccharides (N-glycosides) on glycoproteins share a common precursor and so have a common "core" sequence:

Manα1→3 (Manα1→6)Manβ1→4GlcNAcβ1→4GlcNAc-Asn which can also be represented as:

The sequence differences in the oligosaccharides occur in the non-reducing end of the molecule (left end as written above), i.e., in the sugars attached to the α1,3-and α1,6-linked mannose residues of the core. Three major types of asparagine-linked oligosaccharides are found: high-mannose, complex and hybrid oligosaccharides. I n high-mannose oligosaccharides, all the residues attached to the core are mannose residues, with the total number being 5-9 (including the three in the core). These additional mannoses can have branched structures relative to the core mannoses, resulting in an "antenna-like" structure with 2 to 4 branches. If both the α1,3- and α1,6-linked mannose residues in the core have an N-acetylglucosamine (GN) attached, the oligosaccharide is a complex oligosaccharide. In complex oligosaccharides, each of these GN-initiated branches from the core is termed an antenna. (The term "antenna" is usually applied only to the branches of a hybrid or complex structure, as opposed to high-mannose structures, even if they are branched.) Since each core mannose can have several N-acetylglucosamine residues attached to it, there can be more than 2 antennae on complex oligosaccharides and the resulting structures are termed biantennary (2 GN-antennae), triantennary (3 GN-antennae), tetraantennary (4 GN-antennae), etc. The N-linked saccharides described for human IgD have biantennary (Asn496 and Asn 445) or triantennary (Asn354) structures (see Figures 1-4) (Table III of Mellis et al., J. Biol. Chem. 258:11546-11556 (1983)).

The third major type of asparagine-linked oligosac-charide is a hybrid type composed of elements of the other two types. In a hybrid oligosaccharide, the α1, 6-linked core mannose has only mannose residues attached to it (i.e., like a high-mannose structure), while the α1,3-linked core mannose has one or more GN-initiated antennae attached to it (like a complex structure). In a slight irregularity of nomenclature, the mannose-containing arm of the hybrid structure is termed a single antenna, even if it has branch points within it. The most common hybrid structures, then, would be biantennary hybrid and triantennary hybrid.

One final level of complexity occurs in most hybrid and some complex structures, wherein an N-acetylglucosamine residue is linked to the β1 ,4-linked core mannose (the innermost mannose). Structures which contain this residue are said to be "bisected", since the residue comes between the branches formed by the remainder of the structure. The bisecting N-acetylglucosamine residue contains no substituents and, in fact, limits the addition of further residues to the structure during oligosaccharide synthesis. Examples of a bisected complex structure, which is found on Asn496 of human IgD, is:

Several asparagine residues within a single IgD chain may contain an oligosaccharide chain and all chains on a single IgD are not necessarily of the same type or sequence, as is shown in Mellis et al. (supra).

Specific embodiments of the IgD-associated N-linked glycans useful according to the present invention are shown in Figures 1-4 as those structures associated with Asn354, Asn445 and Asn496 of the human IgD as defined in the WAH myeloma protein (Mellis et al., supra).

Specific embodiments of the IgD-associated O-linked glycans useful according to the present invention are shown in Figure 1, associated with Ser109 of the Cδ1 region, and Thr126, Thr127, Thr131 and Thr132 of the hinge region. The O-linked glycan may include sialic acid (NeuAc) linked in an α2,3 linkage to the Gal or in an α.2,6 linkage to the GalNAc.

The glycans of the present invention may be in either monomeric form, wherein they are non-polymerized and non-aggregated. Alternatively, the glycans of the present invention may be polymerized or aggregated. A polymerized glycan preferably comprises one or more monomeric units attached to a carrier molecule or material. More than one glycan monomer linked to a carrier molecule is referred to as a "neoglycan" (described above). Examples of carriers include proteins, such as serum albumin, or synthetic molecules such as an appropriate polyamino acid, or another polymer which will permit the glycans to maintain their structure in a manner that will enable them to bind to the IgD-R in accordance to the present invention.

Preferably, when used for in vivo administration, the carrier is one that is itself non-immunogenic in the species to which the glycan is being administered. For humans, a preferred carrier protein is human serum albumin. The glycan may be attached to the protein using any of a number of means, using well-known chemical reactions (Kieda, C. et al., FEBS Lett. 99:29-332 (1979); Monsigny, M. et al., Biol. Cell. 51:187-196 (1984)).

A preferred glycan according to the present invention is an IgD-associated glycan, described for human or murine IgD in Mellis et al., supra). and depicted in Figures 1-4. The most preferred glycans according to the present invention are the biantennary complex structures shown m Figure 1, 3 and 4.

Thus, to test a particular glycan or glycan polymer for its utility according to the present invention, the glycan or glycan polymer is incubated with T cells having IgD-R and with appropriate indicator particles, preferably erythrocytes, coated with the IgD of the appropriate species. Thus, for human T cells, ox erythrocytes coated with human IgD are preferred. For murine T cells, sheep erythrocytes coated with murine IgD are preferred. Inhibition of rosette formation by the glycan being tested, as exemplified in the Examples below, is an indication that the glycan has the requisite properties for use according to the present invention. Thus, one of ordinary skill in the art can determine without undue experimentation whether a particular glycan or glycan polymer has the capacity to interact with the IgD-R.

Since the murine IgD-R, as well as the lectin GS-1, recognize only IgD among all murine immunoglobulins, the N-linked glycans associated with the IgD molecule appear to be totally specific for IgD. This specificity is apparently reflected in the functional properties of IgD since it, and no other immunoglobulin, augments antibody production (Coico et al., 1988, supra). There appears to be a species specificity in the binding of IgD to the T cell IgD-R. For example, human IgD does not block rosette formation of mouse T cells with mouse IgD. Similarly, mouse IgD does not block rosette formation of human T cells with human IgD. In one example, a chimeric mouse-human IgD with known antigen specificity (to the dansyl hapten), produced by mouse myeloma cells, does not bind to human IgD-R. Because of this apparent species specificity of the binding of IgD to the IgD-R, the IgD-associated glycan should either be derived from the species being treated, or alternatively should have a structure that is efficacious in that species. One of ordinary skill in the art will be able to test a given glycan structure for its utility according to the present invention by using the methods described herein.

The IgD-R molecule appears to be a member of a family of cell-surface lectin-like molecules which are becoming recognized in the art. These include the IgE binding receptor known as FcεRII or CD23, the ELAM-1 molecule (also termed LECAM-2) and the MEL-14 molecule (also termed LECAM-1) (Ikuta, K. et al., Proc. Natl. Acad. Sci. USA 84:819-823 (1987); Lasky, L.A. et al., Cell 56:1045-1055 (1989); Siegelman, M. et al., Science 241:1165 (1989); Bevilacqua, M.P. et al., Science 243:1160-1165 (1989)). The FceRII has a lectin-like domain important for the interaction with IgE, but it differs from the IgD-R as described herein in that it does not predominantly recognize carbohydrate moieties of IgE (Bettler et al., Proc. Natl. Acad. Sci. USA 86:7118-7122 (1989); Vercelli et al., Nature 338:649-651 (1989)).

Some of the carbohydrate ligands of the leukocyte-endothelial cell (LEO adhesion molecules have been identified (Brandlev et al., Cell 63:861-863 (1990); Picker et al.. Cell 66:921-933 (1991)). Transfection of a specific α(1,3) fucosyl-transferase cDNA into nonmyeloid cell lines results in the de novo expression of functional ligands for LECAM-2 and cell adhesion to this molecule (Lowe, J.B. et al., Cell 63:475-484 (1990)). The present inventors and their collaborators have also found that the pentapeptide GRGDS (SEQ ID NO:1), which is the sequence of fibronectin recognized by integrins (Hynes, R., Cell 48:549-554 (1987)), does not inhibit IgD rosetting. Thus, the IgD-R is distinguishable from the GRGDS binding molecule.

The specificity of the glycan recognition by murine IgD-R involves the Gal, GalNAc and/or GlcNAc saccharides, since each of them significantly inhibit IgD-rosetting, while Man and Glc do not. This agrees with the fact that murine IgD is very rich in Gal residues (Argon, Y. et al., J. Immunol. 133:1627-1633 (1984)).

None of the Gal- or GalNAc-containing disaccharides examined inhibited IgD-rosetting (see below), indicating that none of the linkages represented in these disaccharides mimics the IgD-R-binding glycan structure as well as does free Gal. Galα1→6Glc (melibiose) has highaffinity for GS-1, but fails to inhibit IgD-rosetting, whereas, Galβ1→4Glc (lactose) does not appear to have high affinity for either GS-1 or IgD-R.

Since one monosaccharide, such as Gal, can be linked to another monosaccharide in at least 16 different ways, it is not surprising lactose and melibiose do not present the right configuration for recognition by IgD-R.

Inspection of the structures of the bacterial polysaccharides described above (Heidelberger, M. et al., Immunochem. 13:67-80 (1989)) fails to reveal a common repetitive disaccharide configuration that is peculiar to the two polysaccharides that caused partial inhibition of IgD-rosetting, K11 and K25 (see below).

As mentioned above, the lectin specificity of murine IgD-R resembles that of GS-1. Therefore, according to the present invention, the lectin GS-1 is useful as a reagent in the isolation of murine IgD-associated glycans that inhibit IgD-rosetting and are useful in various embodiments of the present invention. Thus, for example, an affinity column comprising GS-1 bound to a solid support can be used to isolate and purify from a complex mixture a glycan, glycoprotein or glycopeptide which binds to the IgD-R and can be used in accordance with the present invention.

According to the present invention, surface IgD on B cells acts not only as a receptor for antigen, but also serves as a ligand for IgD-R on T helper cells. Because of its lectin-like properties, described herein, the IgD-R on Tδ cells is concluded to be an adhesion molecule. This is further supported by the ability of the IgD-R to recognize and bind the Cδ1 region of B-cell surface IgD (Richards, M.L. et al., J. Immunol. 144:2638-2646 (1990); Amin, A.R. et al., Research Immunol. 141:94-99 (1990)).

The binding of IgD by IgD-R strengthens the cognate

T-B cell interaction central to the generation of an immune response. This interaction provides a mechanism for intracellular signalling in addition to the various secondary signals that may be generated in the T-helper cells via engagement of the IgD-R or in the B cells via engagement of their slgD.

IgD-R upregulation involves triggering of the cell in that inhibitors of protein kinase C inhibit the upregulation. Stimulation of T cells by IgD or the glycans of the present invention stimulates the translocation of protein kinase C without a concomitant Ca⁺⁺ flux.

The ultimate effect of the interaction between the

B cell slgD and the T helper cell IgD-R is the enhanced production of all antibodies of all isotypes except IgD itself (Amin, A. et al., supra). This results in an augmented primary and secondary immune responses.

Therefore, according to the present invention, the glycans described herein can be used to disrupt these T cell-B cell interactions, and are thus useful in the prevention or treatment of immune-mediated diseases, such as autoimmune diseases, that involve T cell-B cell interactions in the generation of pathological or otherwise undesired antibodies.

An effective amount of an IgD-associated glycan for inhibiting the binding of IgD with an IgD receptor according to the present invention is between about 0.01 and 5 mg/ml. More preferably, the amount is between about 0.1 and 1 mg/ml.

For inhibition of an immune response the glycan is preferably administered in an monomeric or soluble form, and is preferably given between about one week and about one day prior to immunization with an antigen.

An effective amount of an IgD-associated glycan for inhibiting an immune response involving the interaction of T and B lymphocytes in a subject according to the present invention is between about 0.01 and 1000 mg/kg body weight. More preferably, the dose is between about 0.1 and 100 mg/kg.

To treat an immune mediated disease by disrupting T cell-B cell interactions, a subject suffering from such a disease is administered an effective amount of an IgD-associated glycan according to the present invention. Preferably, a dose of between about 0.01 and 100 mg/kg body weight is administered. More preferably, the dose is between about 0.1 and 100 mg/kg. Administration may be over a prolonged period, ranging from several days to several months or years, depending on the nature and severity of the disease. The glycan is preferably administered to such subjects in non-polymerized or non- aggregated form.

The methods and compositions described herein are useful for any of a number of autoimmune diseases which involve undesired antibody production as an underlying cause or as a consequence of the pathophysiology. Such diseases include, but are not limited to, systemic lupus erythematosus, rheumatoid arthritis, myasthenia gravis, multiple sclerosis, autoimmune thyroiditis (Hashimoto's thyroiditis), Graves' disease, inflammatory bowel disease, autoimmune uveoretinitis, polymyositis and certain types of diabetes (see, Theofilopoulos, A., In: D.P. Stites, et al., eds., Basic and Clinical Immunology. Lange Medical Publica- tions, Los Altos, CA, 1988)).

The present inventors have established that human T cells also express receptors for IgD (Coico, R.F. et al., J. Immunol. 145:3556-3561 (1990)). According to the present invention, human IgD-R also exhibit lectin-like properties and are specific for glycans associated with human IgD. Human IgD-R on CD4- and CD8-bearing human peripheral blood lymphocytes are upregulated by incubation with IgD, as well as with antibodies specific for CD3, CD4 or CD8. Thus, according to the present invention IgD itself, an IgD-R-binding fragment, or most preferably, an IgD-associated glycan, or any other ligand for the IgD-R, may act as a substitute for a CD3, CD4 or CD8 ligand in engaging the T cell and triggering or enhancing a response. This is an additional case of one cell adhesion molecule replacing another in an intercellular interaction involving multiple cell adhesion molecules, which is known in the art.

An effective amount of an IgD-associated glycan for increasing the number of IgD receptors on the surface of a T lymphocyte having such receptors, according to the present invention is between about 0.001 and 1 mg/ml. More preferably, the amount is between about 0.01 and 0.1 mg/ml.

In humans, IgD-R are present on T cells of both CD4+ and CD8+ subsets. Therefore, this ligand-like structure is not limited to "classical" helper cells. Thus, the glycans of the present invention are useful not only in modulating T-B interactions involving T helper cells but are also useful in modulating the action of other types of T cells on B cell responses. The glycans may be used to upregulate the IgD-R on CD8+ cytotoxic T cells and stimulate cytotoxic activity for IgD-bearing target cells. Furthermore, the glycans may be used to disrupt interactions between T helper cell and T effector cell subsets and B cells.

Since T suppressor cells are generally of the CD8+ subclass, the glycans of the present invention are also useful in modulating the activity of these cells. For example, stimulation of T suppressor cells by binding to and upregulating the IgD-R may enhance suppressor cell activity and thereby diminish unwanted immune responses, such as those associated with immune-mediated disease.

Thus, in another embodiment, lymphocytes, preferably peripheral blood lymphocytes, may be removed from a subject, incubated in vitro with an effective amount of an IgD-associated glycan, as described herein, to upregulate the IgD-R. Such cells may then be reintroduced into the subject in order to stimulate the immune system generally, or to enhance an immune response to a particular antigen, such as a vaccine. Unfractionated populations of cells may be used, or, alternatively, isolated populations of T cells or T cell subsets (e.g., CD4+ cells cells) may be prepared, using conventional methods, before or after treatment with the glycan. Conditions for incubating the cells and concentrations of the glycans or neoglycans are those described above and in the Examples for inducing the IgD-R. Upregulation of the immune response, either by administering the glycan or treating lymphocytes in vitro with the glycan, either as a sole treatment or in combina¬tion with a vaccine as described above, is useful in any condition resulting in a weakened immune system, in particular in immunodeficiency disease such as AIDS or congenital immunodeficiency diseases.

Another embodiment is based on the fact that CD8+ cells commonly have suppressor cell activity and act to inhibit an immune response. Thus, by treating CD8+ T cells in vitro with an IgD-associated glycan to upregulate the IgD-R, as described above, and administering these cells to a subject, it is possible to inhibit an immune response and to treat an immune-mediated disease such as an autoimmune disease. Alternatively, unfractionated cells or whole T cells may be treated to upregulate the IgD-R, followed by fractionation and administration of only the CD8+ T cell subset.

The methods of the present invention may be used to enhance the immune response which becomes deficient in a proportion of aged individuals. For example, about half of aged humans do not upregulate IgD-R as well as younger adults (see Examples below). This measure correlates with total anti-influenza antibody titers in subjects immunized with influenza vaccine. Thus, individuals that did not upregulate their IgD-R had lower antibody titers, and individuals that were low responders to influenza were predictably less able or unable to upregulate their IgD-R in response to an appropriate stimulus. Therefore, according to the present invention, IgD-associated glycans may be used to upregulate the IgD-R and to increase antibody responses. In this manner, they may serve as immunological "adjuvants."

Patients with fevers of unknown origin, resembling familial Mediterranean fever, have been shown to have increased serum IgD, although no antibody specificity nor pathophysiological role for this IgD was discerned (Van der Meer, J.W.M. et al., Lancet. pp. 1087-1090 (1984 May 19)). It is possible that T cell activation through the IgD-R leading ultimately to release of pyrogenic cytokines is involved. Thus, according to the present invention, administration of an agent capable of disrupting interactions between IgD and the IgD-R is useful in treating such individuals. A preferred agent is a IgD-associated glycan, as described herein.

By understanding the role of the intercellular interaction based on the IgD-R binding to an IgD-associated glycan, as disclosed herein, the present invention provides a completely novel approach to modulating the immune response toward either greater or lesser responsiveness. Greater immune responsiveness is desirable in situations in which a subject suffers from an immunodeficiency either congenital or acquired, for example, in AIDS. In contrast, in a number of disease states, the immune system is hyperactive or generates undesired immune responses. This occurs primarily in the case of autoimmune disease, but is also observed in allergy, graft rejection and in graft-versus-host disease.

Commonly the boosting of an immune response, also known as an "adjuvant" effect, is highly desirable when immunizing or vaccinating an individual with an antigen or vaccine, which will protect that individual from a disease, such as an infectious disease. The glycans of the present invention may therefore be used in combination with an immunizing antigen preparation, such as a vaccine preparation, to provide such an adjuvant effect. For this use, the glycan and vaccine may be combined into a single formulation. Alternatively, the vaccine and glycan may be given separately, either on a concurrent or sequential basis. An effective amount of an IgD-associated glycan for enhancing an immune response in a subject being immunized with an antigen or vaccine according to the present invention is between about 0.01 and 1000 mg/kg body weight. More preferably, the amount is between about 0.1 and 100 mg/kg. For boosting the immune response, the glycan is preferably in an aggregated or polymerized form. Alternatively, as described above, CD4+ T cells treated with an IgD-associated glycan may be administered to a subject to enhance the immune response.

The preferred animal subject of the present invention is a mammal. By the term "mammal" is meant an individual belonging to the class Mammalia. The invention is particularly useful in the treatment of human subjects.

By the term "treating" is intended the administering to subjects of the IgD-associated glycans of the present invention for purposes which may include prevention, amelioration, or cure of the disease.

The present invention is also directed to a pharmaceutical composition comprising a glycan or neoglycan of the present invention. In addition to the glycan or neoglycan, the pharmaceutical composition may contain suitable pharmaceutically acceptable carriers comprising excipients and auxiliaries which facilitate processing of the active compounds into preparations which can be used pharmaceutically (see below). Preferably, the preparations, particularly those preparations which can be administered orally and which can be used for the preferred type of administration, such as tablets, dragees, and capsules, and also preparations which can be administered rectally, such as suppositories, as well as suitable solutions for administration by injection or orally, contain from about 0.01 to 99 percent, preferably from about 20 to 75 percent of active compound(s), together with the excipient.

The pharmaceutical composition may be administered by any means that achieve their intended purpose. Amounts and regimens for the administration of the glycan can be determined readily by those with ordinary skill in the clinical art of treating the particular disease.

For example, administration may be by parenteral, subcutaneous, intravenous, intramuscular, intraperitoneal, transdermal, or buccal routes. Alternatively, or concurrently, administration may be by the oral route. The dosage administered will be dependent upon the age, health, and weight of the recipient, kind of concurrent treatment, if any, frequency of treatment, and the nature of the effect desired.

The pharmaceutical formulation for systemic administration according to the invention may be formulated for enteral, parenteral or topical administration. Indeed, all three types of formulation may be used simultaneously to achieve systemic administration of the glycan.

Suitable formulations for oral administration include hard or soft gelatin capsules, dragees, pills tablets, including coated tablets, elixirs, suspensions, syrups or inhalations and controlled release forms thereof. Solid dosage forms in addition to those formulated for oral administration include rectal suppositories.

Suitable injectable solutions include intravenous subcutaneous and intramuscular injectable solutions.

Suitable excipients are, in particular, fillers such as cellulose preparations and/or calcium phosphates, for example, tricalcium phosphate or calcium hydrogen phosphate, as well as binders such as starch paste, using, for example, maize starch, wheat starch, rice starch, potato starch, gelatin, tragacanth, methyl cellulose, hydroxypropylmethylcellulose, sodium carboxymethylcellulose, and/or polyvinyl pyrrolidone. If desired, disintegrating agents may be added. Auxiliaries are, above all, flow-regulating agents and lubricants. Dragee cores are provided with suitable coatings which, if desired, are resistant to gastric juices. Dye stuffs or pigments may be added to the tablets or dragee coatings.

Possible pharmaceutical preparations which can be used rectally include, for example, suppositories which consist of a combination of the glycan with a suppository base, for example, natural or synthetic triglycerides, or paraffin hydrocarbons. In addition, it is also possible to use gelatin rectal capsules.

Suitable formulations for parenteral administration include aqueous solutions of the glycan in water-soluble form. In addition, suspensions of the glycan compounds as appropriate oily injection suspensions may be administered. Suitable lipophilic solvents or vehicles include fatty oils, for example, sesame oil, or synthetic fatty acid esters, for example, ethyl oleate or triglycerides. Aqueous injection suspensions may contain substances which increase the viscosity of the suspension Optionally, the suspension may also contain stabilizers.

Other pharmaceutically acceptable carriers for the compounds according to the present invention are liposomes

(see, for example, Gregoriades, G. et al., Immunological

Adjuvants and Vaccines. Plenum Press, New York, 1989;

Michalek, S.M. et al., Curr. Top. Microbiol. Immunol.

146:51-58 (1989)). In a liposome based compositions, the glycan is contained either dispersed or variously present in corpuscles consisting of aqueous concentric layers adherent to lipidic layers. The glycan may be present both in the aqueous layer and in the lipidic layer, inside or outside, or, in any event, in the non-homogeneous system generally known as a liposomic suspension. Alternatively, the glycan may be chemically linked to one of the liposome constituents. The hydrophobic layer, or lipidic layer, generally, but not exclusively, comprises phospholipids such as lecithin and sphingomyelin, steroids such as cholesterol, more or less ionic surface active substances such as dicetylphosphate, stearylamine or phosphatidic acid, and/or other materials of a hydrophobic nature.

Having now generally described the invention, the same will be more readily understood through reference to the following examples which are provided by way of illustration, and are not intended to be limiting of the present invention, unless specified.

EXAMPLE I MATERIALS AND METHODS

Mice and cell lines

BALB/c and CB6F1 mice, 6-8 week old, were obtained from Charles River Breeding Labs. Two IgD secreting plasmacytomas TEPC-1017 and 1033 (Finkelman, F.D. et al., J. Immunol. 126:680-687 (1981)) were maintained i.p. in pristane primed BALB/c mice. The hybridoma B1-8.δ1 secreting IgD specific for the hapten 4-hydroxy-3-nitrophenyl-acetyl was a kind gift from Dr. K. Rajewsky

(Institute for Genetics, University of Cologne, F.R.G.). It was maintained i.p. in CB6F1 mice. The T cell hybridoma

(2H10) of helper phenotype and specific for cytochrome c was maintained in Click's/RPMI 1640. It was kindly provided by Dr. R.H. Schwartz (NIAID, NIH, Bethesda, MD).

Reagents

Hδ^a/1 and AMS-15 are monoclonal antibodies specific for Fcδ and Fdδ respectively (Zitron, I.M. et al., J. Exp.

Med. 152:1135-1146 (1980), Goroff, D.K. et al., J. Immunol.

136:2382-2392 (1986)). Rabbit anti-mouse IgD and Fab were prepared as previously described (Coico, R.F. et al.. Nature 316:744-746 (1985)).

Purified lectin from Griffonia simplicifolia-1 lectin

(GS-I) was donated by EY Lab (San Mateo, CA). Purified

F(ab')₂ fragments of IgG were received from Dr. W.O. Weigle

(Scripps Institute for Medical Research, La Jolla, CA). Recombinant IL-4 (rIL-4) produced by myeloma transfectants

(Karasuyama, H. et al., Eur. J. Immunol 18:94-104 (1988)) was used as a source of rIL-4, standardized by comparison with rIL-4 obtained from Dr. M. Howard (DNAX, Palo Alto,

CA). Peptides KINLGCLVIGSQPLKI (SEQ ID NO:2) (derived from Cδ1) and SSWLLCEVSGFFPENI (SEQ ID NO:3) (from Cδ3) were prepared by solid phase syjthesis (Kent, S.B.H. et al., in K. Alitalio et al. (eds.). Synthetic Peptides in Biology and Medicine. Elsevier Amsterdam, 1985, pp. 29-41), with a Biosystem peptide synthesizer 430A and their purity was determined by HPLC.

Gal, GalNAc, GlcNAc, and mannose (Man) were from Sigma (St. Louis, MO.), melibiose (Mel) from Pfanstiehl Lab. (Waukegan, IL.), and lactose (Lac) from Fluka (Ronkonkoma, NY). The purified polysaccharides from Pneumococcus and Klebsiella (Heidelberger, M, et al.. Immunochem. 13:67-80 (1989))

were made available by Dr. M. Heidelberger's laboratory. Purified Griffonia simplicifolia-1 lectin (GS-I) was obtained from EY Lab. (San Mateo, CA) and also prepared by the present inventors (Oppenheim, J.D. et al., supra). Purification of Splenic T cells

Spleen T cells were prepared by sequential depletion of adherent cells at 37°C in Petri dishes (1400-1, Nunclon, Rockilde, Denmark) and of B cells by negative selection (Wysocki, L.I. et al., Proc. Natl. Acad. Sci. (USA) 75:2844-2848 (1978)) at 4°C on Petri dishes coated with affinity-purified anti-mouse Ig (remaining Ig⁺ cells <1%). Rosette-forming cell (RFC) assay

Mouse splenic T cells or human peripheral blood lymphocytes (PBL) (2.5 × 10⁶) were incubated at 37°C for 18 hrs. in 1 ml of RPMI-1640 containing 2% fetal calf serum (FCS) with rIL-4 (10 U/ml) or crosslinked IgD (100 μg/ml) overnight. These resulting Tδ cells were tested for IgD-R in RFC assays. (Coico, R.F. et al., Nature 316:744-746 (1985)). SRBC were coated with either purified IgD, Gen.24, KWD1, KWD6 or bovine serum albumin (BSA) by the CrCl₃-coupling method (Poston, P.W. et al., J. Immunol. Methods 5:91-96 (1974)). For human cells, IgD-coated ox RBC were used. T cells surrounded by more than 3 coated SRBC (minimal definition) were scored as rosettes and the results expressed as % RFC (the % of cells forming rosettes). The majority of rosettes were appeared as daisy- or morula-like structures, with occasional RFCs showing this minimal definition. The percentage of cells rosetting with BSA-SRBC was subtracted (normal T cells: < 5 ± 1; T cell hybridomas: 24 ± 6).

Rosette inhibition assays were carried out using IgD-R⁺ T cells which had been induced with rIL-4 or IgD. IgD-R⁺ T cells were incubated with various purified test proteins, (e.g., IgD, Gen.24, KWD1, Fabδ, Fcδ, or KWD6), on ice for 30 min. and then allowed to interact with IgD-SRBC in a total volume of 300 μl. Percentages of IgD rosette inhibition were calculated with the formula:

100 - 100 × (% IgD-RFC over BSA-RFC bckgrd. blocked sample)

(% IgD-RFC over BSA-RFC bckgrd, control sample)

Purification of IgD

Purification of IgD was done by slightly modifying the method of Finkelman et al. (Finkelman, F.D. et al., J.

Immunol. 126:680-687 (1981)). IgD was also purified by affinity chromatography on an IgD specific GS-1-Sepharose column and eluted with galactose (Oppenheim, J.D. et al.,

J. Immunol . Methods 130:243-250 (1990)). Mutant IgD molecules Gen.24, KWD1 and KWD6 were affinity purified over a goat anti-mouse IgD-Sepharose column, as monitored by double diffusion in agarose gel, SDS-PAGE and ELISA.

Enzymatic treatment of IgD

Papain was purchased from Pierce Chemicals, Inc., (Rockford, IL), pronase was from Calbiochem (San Diego,

CA), and proteinase K was purchased from Sigma (St. Louis,

MO).

Purified TEPC-1017 IgD was digested at 37°C using immobilized papain as prescribed by the manufacturer. Fcδ fragments were subsequently isolated by affinity chromatography with the Fcδ-specific, Hδ^a/1-Sepharose column. Fabδ fragments of naturally degraded purified IgD were also isolated by passage through this column (See Figure 6). SDS-PAGE was performed under reducing and non- reducing conditions. Molecular weight markers are given in kDa.

Low Mr IgD fragments were prepared by pronase (F1) or proteinase K (F2) digestion of purified TEPC-1033 IgD. Ten mg IgD dissolved in 1 ml 0.1 M Tris-HCl, pH 7.4, was digested with 2 mg of pronase or proteinase K for 12 hr at 37°C, 2 mg of the same enzyme was added, the digestion continued for 12 hr, followed by a third addition of 2 mg of enzyme and another 12 hr digestion. The lowest Mr components (<5000 Da) of these digests were obtained by gel filtration on Sephadex G-25; the position of the retained small molecules was determined with tryptophan as standard. The retained fractions from these columns were pooled and lyophilized. The low Mr fraction from IgG2a was similarly prepared by pronase digestion of UPC 10 myeloma protein (F3). Unfractionated fragments, obtained by complete digestion of TEPC-1017 IgD, 3A6 hybridoma IgA (F4) and TEPC-187 IgM (F5) with proteinase K, were found to be have Mr of <5 kDa on analysis by 4-20% SDS-PAGE.

Preparation of Neoglycoproteins

Neoglycoproteins comprising monosaccharide- substituted bovine serum albumin (BSA) were prepared as previously described (Kieda et al., supra: Monsigny et al., supra); p-nitrophenyl glycosides (Sigma, St. Louis) were reduced to p-aminophenyl glycosides, converted into glycosidophenyl isothiocyanates and coupled to BSA up to approximately 20 sugar residues/mole.

N-Glycanase Treatment and Oligosaccharide Purification

IgD was treated with N-glycanase (PNGase F) as previously described (Oppenheim et al., supra). Ten mg purified TEPC-1017 IgD was treated with N-glycanase according to the manufacturer's specifications (Genzyme, Cambridge, MA). A positive control, transferrin, was treated similarly. All samples were then tested for the presence of glycans by dot-blot analysis using staining with digoxigenin succinyl-ε-amidocaproic acid hydrazide (Boehringer Mannheim Biochemica Glycan Detection Kit) to monitor the reaction.

The partially deglycosylated IgD (DG-IgD) and the released asparagine-linked (N-linked) oligosaccharides were purified as follows: The reaction mixture was passed through a PD-10 desalting column according to the manufacturer's specifications (Pharmacia). The protein and

(salt + oligosaccharide) fractions were pooled individually and lyophilized. The DG-IgD was concentrated and fractionated on a Superose 6 FPLC column to isolate the intact IgD molecules. This DG-IgD was then chromatographed on a GS-1 Sepharose column. The unbound fraction was passed over Extracti-Gel-D to remove traces of detergent according to the manufacturer's specifications (Pierce Chem. Inc.) and was extensively dialyzed against DPBS before use. The control IgD was treated similarly, but without enzyme.

The released glycans (in the salt fraction) were purified by chromatography on GS-1 Sepharose column. The bound glycans were eluted with glycine-HCL, pH 3.0. The sample was neutralized and lyophilized. After dissolving in 2 ml of distilled water, the carbohydrate content was estimated by the Anthrone method.

The DG-IgD was analyzed for its reactivity with peroxidase conjugated GS-1 and anti-IgD antibody in agarose gel immunodiffusion and in dot blot assays. The released glycans were shown to inhibit immunoprecipitin reactions between IgD and GS-1 in immunodiffusion gels. EXAMPLE II

IgD-R Independently Recognizes Cδ1 and Cδ3 Regions

Mouse IgD, unlike human IgD, lacks A Cδ2 heavy chain domain (Tucker, P.W. et al., Science 208:1353-1360 (1980)) and consists of the Cδ1, Cδ-hinge and Cδ3 domains. To determine the part of the IgD molecule which IgD-R recognized, binding the ability of mutant IgD molecules lacking one or more heavy chain domains to CD4⁺Tδ cells was tested in rosetting assays. In addition, the ability of the mutant IgDs to inhibit rosetting of CD4⁺ T hybridoma cells or splenic Tδ cells with IgD-coated SRBC was examined.

All the mutant IgD molecules tested have intact V_H and V_L domains. KWD1 lacks the Cδ1 domain (Mountz, J.D. et al., J. Immunol. 145:1583-1591 (1990)), and KWD6 lacks both Cδl and Cδ-hinge (Table 1). The deletions of these domains were confirmed by Northern blots and by ELISA with monoclonal anti-δ antibodies (Table 1). In contrast, Gen.24, produced by a spontaneous variant of the IgD-producing plasmacytoma, TEPC-1017, includes Cδl and part of Cδ-hinge but lacks Cδ3 (Thiele, C.J. et al., J. Immunol. 134.:1251-1256 (1985)). All the mutant IgD molecules bind to the lectin GS-1, which is known to be specific for N-linked glycans of murine IgD (Oppenheim, J.D. et al., J. Immunol. Methods 130:243-250 (1990)).

The results were as follows. Intact dimeric (TEPC- 1017) and monomeric (B1-8.δ1) IgD blocked IgD-rosetting to approximately the same extent, The mutant IgD molecules

(Table 1), had similar binding activity, in particular when compared by IC₅₀ (the molar concentration of IgD (or mutant) needed for approximately 50% inhibition).

Considering that TEPC-1017 is a 260 kDa dimer, its effectiveness (on a molar basis) is comparable to that of Gen.24 (100 kDa) and KWDl (90 kDa). In addition, all of the mutant proteins, when coating SRBC, resulted in rosette formation by either CD4⁺ splenic Tδ cells or 2H10 (CD4⁺, IgD-R⁺) T hybridoma cells, though KWD1-coated RBC were somewhat less effective (Table 1).

To confirm the results with KWD6, attempts were made to isolate Fcδ molecules. However, Fcδ fragments of sufficient size homogeneity could not be obtained for proper inhibition studies. The heterogeneous preparation obtained (which reacted with Hδ^a/1 but not with rabbit anti-Fab and had an average size of 40 kDa) inhibited IgD-RFC by only 23±1% at 120 μg/ml.

TABLE 1

PROPERTIES OF TEPC-1017, MUTANT IgD MOLECULES AND FABδ

Property DNA/mRNA or Protein from:

Gen.24 KWDl KWD6 IgD Fabδ

2H+2L Protein M.W. (kDa) 1 100 90 85 135 66-70 mRNA size (kb)² 1.15 1.1 1.05 1.75

DNA/mRNA Reactivity with probe:

V_H + + + +

C_δ1 + - - +

C_δH ± + - +

C_δ3 - + + +

Protein Reactivity with:³

Rabbit anti-IgD +++ +++ ++++ +++ +++ Rabbit anti-Fab ++ ++ ++ ++ +++ Hδ^a/1 (mab to Fcδ) - +++ +++ ++ - AMS-15 ++++ +/- +/- ++++ ND GS-1-peroxidase ++ ++ + +++ +

Percentage RFC with:⁴

Splenic Tδcells 28+3 15±2 23±2 21±2 ND 2H10 cells 57±8 37±4 46±3 69±5 ND

% Inhibition of IgD-RFC by:⁵

100 μg ND ND ND 91±5 ND 40-50 μg 82±6 94±4 ND 88±8 ND 20-25 μg 88±2 90±1 ND 69±4 50±2 10-12.5 μg 60±8 73±3 ND 47±2 12±2 5-6 μg 43±6 63±5 ND 0 0

¹ By SDS-PAGE; 2H + 2L chain given.

² Northern blots on RNA from mutant protein-producing cells, using domain- specific oligomer cDNA probes.

3 Agglutination of coated SRBC and/or by Western blotting and ELISA. GS-1 = lectin from G. simplicifolia. ND=not determined.

4 SRBC coated with each protein were tested for binding to IgD RFC. BSA- RFC background (already) subtracted: 5±1% for splenic T cells; 24±6% for 2H10 cells.

5 Test T cells rosetted with TEPC-1017 IgD-SRBC in the absence of blocking agents gave 30-38% RFC. Blocking proteins were added at indicated amounts. Other Ig isotypes showed no inhibitory effect in this assay.

EXAMPLE III

Fd and Fc Regions of IgD Compete for Binding to IgD-R

The above results indicate that KWD6 and Gen.24, despite their lack of Cδ1 + Cδ-hinge and Cδ3 domains, respectively, each contain some structure or determinant recognized by the IgD-R. The identity of this determinant was further examined (See Figure 5) by RFC cross-blocking experiments. KWD6 and Gen.24 were equally effective in blocking rosetting with Gen.24-coated SRBC, while KWD6 was quantitatively more effective than Gen.24 in blocking rosetting with KWD6-coated SRBC. These results demonstrate that the (Cδ1 + Cδ-hinge) and Cδ3 domains of IgD can independently bind to the IgD-R and competitively inhibit each other's binding to the same receptor.

EXAMPLE IV

The Common Binding Site for IgD-R is Not Due to

Amino Acid Sequence Homology

The tailpiece of the secreted murine IgD is considerably longer (21 residues) than that of human IgD

(Burton, D.R., In: Metzger, H. (ed.) Fc Receptors and

Action of Antibodies. Amer. Soc. Microbiol., Washington,

D.C., 1990, pp. 31-54). All of the mutant molecules presumably share this C-terminal sequence. Experiments were performed to determine whether these residues played a role in binding to the IgD-R. Fabδ fragments were isolated by passing spontaneously degraded purified IgD over an Fcδ-specific Hδ^a/1-Sepharose affinity column. As shown in Figure 6, two fragments of approximately 66-70 kDa as well as a 90 kDa fragment were present in a stored preparation of TEPC-1017 IgD. Hδa/1-Sepharose bound the 90 kDa and intact IgD (130 kDa), but did not bind the 66 and 70 kDa fragments. Reduction of the 66 and 70 kDa fragments generated Ig light chains (25 kDa) and 32 and 36 kDa heavy chain fragments. Western blotting of the unreduced IgD fragments showed reactivity with both rabbit anti-Fab and rabbit anti-IgD (Figure 6). In ELISA, they reacted with AMS-15 (a Fdδ-specific antibody) but not with Hδa/1. The Fabδ fragments were estimated to be >95% pure by ELISA. As was true for the mutant proteins, Fabδ fragments, bound to IgD-R, as shown by their ability to inhibit Tδ rosette formation with SRBC coated with intact IgD (Table 1). Control IgG Fab molecules did not have this activity. On a molar basis, Fabδ was less effective an inhibitor of IgD-rosetting than was the Gen.24 molecule. This discrepancy could be explained by the lower avidity of a single δ chain compared to a double δ chain. The effectiveness of the Fabδ in rosette inhibition supports the conclusion that the C-terminal amino acid residues of secreted IgD are not necessary for binding to IgD-R.

Cross-inhibition by molecules containing Cδ3 and (Cδ1+hinge) suggests the presence of common determinants in these domains, for example due to amino acid sequence homology or common carbohydrate moieties. Cδ1 (residues 28-40) and Cδ3 (residues 24-36) share 6 of 13 amino acids (Tucker, P.W. et al., Science 208:1353-1360 (1980)). Therefore, these two peptides together with their neighboring residues (16- mers) were synthesized and tested as inhibitors of IgD-rosetting. No inhibition by the peptides was seen using concentrations as high as 300 μg/ml. Thus, the only common determinant that might explain the result is a carbohydrate determinant.

In addition, aggregated IgD mutant molecules and fragments were tested for their ability to induce IgD-R in murine T cells. Murine splenic T cells were incubated with various aggregated mutant IgD molecules or IgD fragments

(25 μg/ml) for 3-4 hrs, washed and rosetted with IgD-SRBC.

The results are shown in Figure 7. Each of the proteins tested caused upregulation of IgD-R.

The results described above show that the T cell receptor for IgD reacts with two entirely non-overlapping portions of the IgD molecule, the Cδ1 and the Cδ3 domain.

The cross-inhibition of rosetting between various mutant IgD molecules indicates that regions of both Cδ1 and Cδ3 are involved in the interaction with the same receptor. It is unlikely that the tailpiece of IgD was responsible for the cross-inhibition of rosetting by the mutant molecules. Moreover, the results could not readily be explained by homology of the Cδ1 and Cδ3 polypeptide backbone structure, since corresponding oligopeptides did not inhibit rosetting. Furthermore, this peptide region is highly homologous with the murine Cμ4 (Tucker, P.W. et al., ScieSce 208: 1353-1360 (1980)), whereas IgM fails to interact with IgD-R (Coico, R.F. et al., Nature 316:744-746 (1985)).

It can be concluded that the receptor on T cells for IgD is not limited to the Fc region, and it therefore should be referred to as IgD-R rather than a Fcδ receptor.

The results also demonstrate that the mutant molecules and Fabδ share with IgD the ability to bind to GS-1, a lectin which binds specifically to N-glycans isolated from IgD but does not bind to deglycosylated IgD. This suggested to the present inventors that the carbohydrate moieties may be playing an important role in the binding of IgD to IgD-R and in the modulation of the IgD-R.

EXAMPLE V

Low Mr IgD Fragments Block IgD Interaction with IgD-R Initially, experiments were performed to analyze the consequences of heat denaturation and complete proteolytic enzymic digestion of IgD on its ability to bind to the IgD-R. TEPC-1033 IgD, after boiling for 10 min, no longer reacted with polyclonal goat anti-IgD antiserum but still competitively inhibited IgD-rosette forming cells (RFC) when T cells were incubated with TEPC-1017 IgD-coated sheep red blood cells (SRBC) (Table 2).

Complete digestion of IgD with the proteinase enzymes pronase or proteinase K resulted in fragments of Mr < 5000 Da which could still competitively inhibit rosette formation. Low Mr fragments of immunoglobulins of other isotypes, including IgM, IgA, or IgG2a, produced by the same treatment, failed to inhibit significantly IgD rosette formation (Table 2), even at concentrations of 150 μg per assay. These findings demonstrate not only that the inhibition was specific for IgD fragments, but also that contaminating enzymes and/or reagents used to prepare the IgD fragments did not cause the IgD-RFC inhibition.

TABLE 2

COMPETITIVE INHIBITION OF IgD-ROSETTING BY LOW Mr

IgD FRAGMENTS WITH AFFINITY FOR GS-1 LECTIN

Blocking Agent^* % Inhibition of IgD-RFC^** - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Expt, . 1 Expt, . 2 Exot, .3

Whole TEPC-1033/1017 IgD (50 μg) 69 ± 1 70 ± 4 74 ± 2

Boiled IgD (10 min at 100°C) 54 ± 1 - -

Low Mr Fragments of Ig Digested With:

F1: IgD, Pronase 75 ± 3 58 ± 1 _

F2. IgD, Proteinase K 55 ± 3 51 ± 1 55 ± 5

F3. UPC 10 IgG2a, Pronase 7 ± 6 - - F4: 3A6 IgA, Proteinase K - 2 ± 5 F5: TEPC 187 IgM, Proteinase K - - 2 ± 5

F1 + F2 _ 62 ± 4 -

F1 + F2 (boiled, 10 min at 100°C) - 45 ± 8 -

F1 + F2 (absorbed with GS-1-Sepharose)) - 2 ± 3 - F1 + F2 (absorbed with BSA-Sepharose) - 49 ± 3 -

Where appropriate, results are expressed as Mean ± SD

* Expts. 1 and 2 : blocking agents were prepared from

TEPC-1033 IgD, 50 μg/assay. Expt. 3: boiled fragments were from 75 μg TEPC-1017 IgD, IgM or IgA. Since all fragments had Mr < 5 kDa, fragments were not further fractionated before assay. In Expts. 1 and 2, low Mr fractions were adsorbed by passage over lectin GS-1- coupled Sepharose 4B.

** Splenic T cells were induced to express IgD-R by

overnight incubation with IL-4 (10 U/ml). Control IgD- RFC values (i.e., no blocking agent added) were 35±6% (Expt. 1), 34±3% (Expt. 2), and 26±2% (Expt. 3), n = 3 or 4. EXAMPLE VI

Lectin GS-1 Binds to and Absorbs IgD-RFC-Inhibiting Moietv

A lectin derived from Griffonia simplicifolia (GS-1) with specificity for Gal and GalNAc binds to murine IgD but not to any other murine Ig or any other proteins in murine ascites fluid (Oppenheim et al., supra). Like the IgD-R, GS-1 appears to be uniquely specific for IgD among all mouse Ig isotypes.

IgD which was completely digested by protease retained its ability to form a precipitate with GS-1 upon double diffusion in agar. Neither intact nor digested IgM or IgA showed any such precipitation reaction.

To investigate the possible contribution of IgD- associated carbohydrates in the IgD digests to the rosette- inhibiting activity, the ability of GS-1 to absorb out the inhibitory activity was assessed. Absorption with Sepharose-bound GS-1, which binds secreted murine IgD via the N-linked carbohydrates (Oppenheim et al., supra), fully removed the IgD-inhibitory activity from the pronase and proteinase K digests, whereas BSA-Sepharose did not (Table 2)

EXAMPLE VII

Role of N-linked Carbohydrates in the IgD/IgD-R Interaction

N-linked sugars were removed from IgD by treatment with the enzyme N-glycanase. This "deglycosylated" IgD (DG-IgD) no longer bound to GS-1 and failed to cause significant inhibition of IgD-rosetting (Figure 8A). DG- IgD caused 12±13% inhibition as compared to 74±2% inhibition by monomeric B1-8.δ IgD and 80±4% inhibition by dimeric TEPC-1017 IgD. The DG-IgD retained its ability to bind to anti-IgD antibodies (Figure 8B).

The carbohydrates released from IgD during hydrolysis by N-glycanase were purified using affinity chromatography with GS-1-Sepharose and tested for their capacity to block rosetting (Table 3). A dose-dependent inhibition of IgD-rosetting was obtained when the indicator erythrocytes were coated with intact IgD and also when they were coated with the mutant IgD molecules Gen.24 or KWD6 (Table 3). The concentration which resulted in 50% inhibition (IC₅₀) was approximately 10-15 μg of N-glycans per assay. As shown by their reactivity with GS-1 (see below/above), both the mutant IgD proteins contain N-linked glycans.

TABLE 3

COMPETITIVE INHIBITION OF MDRINE IgD-RFC BY PURIFIED N-LINKED GLYCANS

OBTAINED FROM IgD

Blocking % IgD-RFC (% Blocking)

SRBC Coated with: Agent (μg)* EXDt.1 ExDt. 2

TEPC-1017 IgD None 28 ± 2 25 ± 2

TEPC-1017 IgD IgD (50) 4 ± 1 (86) 3 ± 0.4 (88) TEPC-1017 IgD Glycan (5) 9 ± 0.3 (68) 8 ± 0.5 (69) TEPC-1017 IgD Glycan (2.5) ND 19 ± 2 (27)

Gen.24 Buffer 24 ± 0.3 ND

Gen.24 Glycan (5) 9 ± 1 (63) ND

KWD6 Buffer 29 ± 1 ND

KWD6 Glycan (5) 12 ± 1 (59) ND

Splenic T cells were induced to express IgD-R by overnight incubation with IL-4 (10 U/ml).

*

Blocking agents were all derived from TEPC-1017 IgD. Low Mr fractions containing N-glycans from IgD were passed over GS-1- Sepharose. Adherent glycans were eluted with glyciune-HCl, pH

3.0, neutralized to pH 7.0 with 1M Trie and lyophilized. Equal amounts of similarly neutralized glycine-HCl were used as

_** control "buffer" .

n=3 . Control RFC values (BSA-RFC) were < 2% and have been subtracted.

EXAMPLE VIII

Monosaccharides Competitively Inhibit Binding of

IgD to IgD-R

Since the GS-1 lectin binds N-acetylgalactosamine (GalNAc) as well as galactose (Gal) (Murphy, L.A. et al..

J. Biol. Chem. 252:4739-4743 (1977)), studies were performed to examine whether these sugars could also inhibit IgD-RFC.

GalNAc caused a highly significant, dose-related inhibition of rosette formation (Figure 9). However, assuming an average molecular size of 10 sugar residues for the isolated mixture of N-glycans from IgD, this monosaccharide was much less effective (IC₅₀ = 0.1 mM) than the IgD-associated N-glycans (IC₅₀ = 5 μM). Gal and GlcNAc were less effective than GalNac. Significant competitive inhibition of IgD-RFC was not seen with mannose (Man), glucose (Glc), or the disaccharides, lactose (Lac), melibiose (Mel), β-D-Gal-[1-3]-D-GalNAcorα-D-Gal-(1-4)-D- Gal.

Neoglycoproteins based on bovine serum albumin (BSA), such as α-D-GalNAc-BSA, α-D-GlcNAc-BSA and α-D-Gal-BSA, when added at a concentration of 3 μM protein to Tδ cells, blocked IgD-rosetting by 76.1%, 43% and 39.8% respectively.

Dimeric TEPC 1017 IgD causes 50% inhibition at 0.15 μM. On the other hand, α-D-Lac-BSA, α-D-Man-BSA, α-D-Man-6- Phosphate-BSA, α-L-Fuc-BSA and β-D-Glc-BSA did not cause significant inhibition at the same concentrations. These results are consistent with those obtained with the freed monosaccharides. Moreover, the three inhibitory neoglycoproteins coprecipitate with GS-1 upon double diffusion in agar at 4°C, whereas the non-inhibitory neoglycoproteins do not.

Various Gal-/GalNAc-rich purified polysaccharides of bacterial origin were also tested for their ability to inhibit IgD-rosetting. These included pneumococcal polysaccharides S1, S4, S8, S11a, S13, S14, S15 and S29, and Klebsiella polysaccharides K11, K12, K16, K18, K21, K22, K23, K24, K25, K27, K31, K38, K41, K51, K53, K56, K74 and K83 (Heidelberger, M. et al.. supra)). Two of these polysaccharides gave a substantial degree of inhibition at each of two concentrations (25 and 100 μg per RFC mixture): K11 caused 29±9% and 64±13% inhibition and K25 resulted in 30±1% and 50±8% inhibition of IgD-RFC. None of the other polysaccharides inhibited by more than 10% at 25 μg per RFC mixture.

DISCUSSION

In view of the high content of N-linked glycans in murine IgD (Argon et al., supra: Vasilov et al., supra)) and the previous demonstration that N-linked glycans of IgD are solely responsible for the binding of IgD to the Gal/GalNAc IgD-specific GS-1 lectin (Oppenheim et al.. supra), N-linked glycans from IgD were tested for their ability to inhibit IgD-rosetting and found active at very low concentrations. Assuming an approximate content of carbohydrate for IgD of 10%, the effectiveness of the glycans on a w/v basis as compared to intact IgD indicates that it alone is responsible for the rosette inhibition. A change in tertiary structure could of course have contributed to the absence of inhibitory activity in DG-IgD. In addition, it is possible that the protein backbone structure contributes to the stabilization of IgD-IgD-R complexes after initial binding of the glycans by the receptors.

Thus, the present observations show that the IgD-R functions as a lectin in its interaction with IgD. This has an interesting resemblance to other cell-surface lectins known in the art, including FceRII (CD23), ELAM-1

(LECAM-2) and MEL-14 (LECAM-1) (Lasky et al., supra:

Siegelman et al., supra: Bevilacqua, M.P. et al., supra;

Brandley et al., supra: Picker et al., supra: Low et al., supra)

Among the Ig-specific receptors the IgD-R are therefore unique, since there is an absolute requirement for the presence of carbohydrate on its ligand for binding. Fc -R does not have a strict requirement for glycans on IgG for its binding (Peppard, J.V. et al., Mol. Immunol. 26.495-500 (1989)). Although FceRII exhibits a lectin-like domain in its structure important for the interaction with IgE, it does not predominantly recognize carbohydrate moieties of IgE (Bettler et al., Proc. Natl. Acad. Sci. USA 86:7118-7122 (1989)).

It is of interest to note that Ca⁺⁺ is required in the interaction between IgE and FceRII (Richards et al., supra), as it is for the interaction between IgD-R and IgD

(Amin, A.R. et al., FASEB J. 4:A2203 (1990)) and for a variety of "C-type" lectins (Drickamer, K., J. Biol. Chem.

263:9557-9560 (1988)).

The results reported herein indicate that the portion of the IgD molecule available on the surface of B cells is capable of binding to IgD-R of T cells, pointing to the possibility that Fabδ-antigen complexes, released from the surface of B cells by cleavage of the IgD molecule, could function in the regulation of the immune response by upregulating IgD-R on T cells. Since the IgD idiotype is present in such complexes, a T cell-mediated idiotype- specific influence on the immune response (Bourgois, A. et al.. Eur. J. Immunol. 7:210-213 (1977)), could be an integral part of the immunoregulatory effect.

EXAMPLE IX

Interactions of Human IgD and Human IgD-R. and

Upregulation of Human IgD-R

Human PBL were isolated from blood, and incubated with plastic dishes coated with cross-linked human IgD to upregulate human IgD-R. These cells were then rosetted with IgD-coated Ox-RBC in the presence or absence of various concentrations of potential inhibitors consisting of proteinase K digested IgD or IgG. The results, shown in Figure 10, indicate that proteinase K digested human IgD was capable of inhibiting rosette formation. The fact that this digestion was complete is shown in Figure 11. Five mg samples of human IgD (160 kD) or IgG (150 kD) purified from myeloma serum were each digested with 2 mg of proteinase K for 12 hrs prior to analysis by 4-20 % SDS PAGE. The results show that both immunoglobulins have been completely digested into low molecular weight (<14 kDa) fractions.

The next experiment tested whether heat-denatured human IgD could bind to IgD-R. Human PBL were first incubated with cross-linked human IgD to upregulate IgD-R, and then incubated with 50 or 100 μg control or heat-denatured IgD (boiled 2-3 minutes, cooled at room temperature). Rosetting was with IgD-Ox-RBC. As shown in Figure 12, heat-denatured IgD bound to the IgD-R as evidenced by its inhibition of rosette formation. Table 4 shows the upregulation of IgD-R by antibodies to T cell surface molecules. Anti-CD4 and anti-CD8 antibodies, each reactive with a subset of T cells, indeed induces upregula-tion of IgD-R on fewer cells than does anti-CD3. The sum of percentages of IgD-RFC detected after stimulation with anti- CD4 and anti CD8 is approximately equal to that seen after stimulation with IgD. The percent of CD8 cells having IgD-R is usually higher than for CD4 cells, both in young and aged individuals.

TABLE 4

Upregulation of IgD-R on Human T Cells by

Anti-CD3, -CD4 and -CD8

Mean Increment¹ in %IgD-RFC + SD (N) in PBL from:

Aged IgD- Aged IgD Agent²(hrs) Young Adults Responders Nonrespondera

IgD (2) 6.3 ±0.3 (4) 6.2 ±2.4 (4)

IgD (18) 11.3 ±4.3 (6) 9.1 ±3.3 (5) 0.3 ±0.5 (7)

Anti-CD3 (2) 5.6 ±4.3 (3) 5.9 ±4.1 (4)

Anti-CD3 (18) 12.2 ±4.4 (6) 10.7 ±5.4 (5) 0.0 ±0.4 (7)

Anti-CD4 (18) 5.5 ±4.2 (6) 3.9 ±2.4 (5) 0.0 ±0.4 (7)

Anti-CD8 (18) 6.6 ±4.5 (6) 7.1 ±4.8 (5) 0.2 ±0.4 (7)

¹ Background IgD-RFC subtracted was <2%

₂ Cells were exposed to dishes coated with IgD or the

monoclonal antibodies specific for CD3, CD4 or CD8 at concentrations of 25 μg/dish.

The references cited above are all incorporated by reference herein, whether specifically incorporated or not.

Having now fully described this invention, it will be appreciated by those skilled in the art that the same can be performed within a wide range of equivalent parameters, concentrations, and conditions without departing from the spirit and scope of the invention and without undue experimentation.

While this invention has been described in connection with specific embodiments thereof, it will be understood that it is capable of further modifications. This application is intended to cover any variations, uses, or adaptations of the inventions following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains and as may be applied to the essential features hereinbefore set forth as follows in the scope of the appended claims.

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT: Thorbecke, Gertrude Jeanette

Amin, Ashok R.

Oppenheim, Joel D.

(ii) TITLE OF INVENTION: IMMUNOGLOBULIN D-ASSOCIATED

GLYCANS AND USES THEREOF

(iii) NUMBER OF SEQUENCES: 3

(iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: Browdy and Neimark

(B) STREET: 419 Seventh Street, N.W.

(C) CITY: Washington

(D) STATE: DC

(E) COUNTRY: USA

(F) ZIP: 20004

(v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFTWARE: PatentIn Release #1.0, Version #1.25

(vi) CURRENT APPLICATION DATA:

(A) APPLICATION NUMBER:

(B) FILING DATA:

(C) CLASSIFICATION:

(viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: Livnat, Shmuel

(B) REGISTRATION NUMBER: 33,949

(C) REFERENCE/DOCKET NUMBER: THORBECKE=1

(ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: (202)628-5197

(B) TELEFAX: (202)737-3528

(2) INFORMATION FOR SEQ ID NO:1:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 5 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS : single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID NO:1:

Gly Arg Gly Asp Ser

5

(2) INFORMATION FOR SEQ ID NO: 2:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 16 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS : single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 2

Lys Ile Asn Leu Gly Cys Leu Val Ile Gly Ser Gln Pro Leu Lys Ile

5 10 15

(2) INFORMATION FOR SEQ ID NO: 3

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 16 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: peptide

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 3:

Ser Ser Trp Leu Leu Cys Glu Val Ser Gly Phe Phe Pro Glu Asn Ile

5 10 15

Claims

WHAT IS CLAIMED IS:

1. A method for inhibiting the binding of IgD with an IgD receptor comprising providing to a cell having an IgD receptor an effective amount of an IgD-associated glycan.

2. A method according to claim 1 wherein said glycan is in non-aggregated and non-polymerized form.

3. A method according to claim 1, wherein said glycan is selected from the group consisting of :

and

wherein R¹ is NeuAc in an α2→3 linkage or H, R² is NeuAc in an α2→6 linkage or H, R³ is GlcNAc in a β1→4 linkage or H, R⁴ is Fuc in an α1→6 linkage or H and R⁵ is Glc in an α1→3 linkage or H.

4. A method for inhibiting an immune response involving the interaction of T and B lymphocytes in a subject, comprising administering to said subject an effective amount of an IgD-associated glycan, thereby inhibiting said immune response.

5. A method according to claim 4 wherein said glycan is in non-polymerized or non-aggregated form.

6. A method according to claim 4, wherein said glycan is selected from the group consisting of:

and

wherein R¹ is NeuAc in an α2→3 linkage or H, R² is NeuAc in an α2→6 linkage or H, R³ is GlcNAc in a β1→4 linkage or H, R⁴ is Fuc in an α1→6 linkage or H and R⁵ is Glc in an α1→3 linkage or H.

7. A method for treating a subject with an immunemediated disease, comprising inhibiting the immune response in the subject with a method according to claim 4, thereby treating said disease.

8. A method for treating a subject with an immunemediated disease, comprising inhibiting the immune response in the subject with a method according to claim 5, thereby treating said disease.

9. A method for treating a subject with an immunemediated disease, comprising inhibiting the immune response in the subject with a method according to claim 6, thereby treating said disease.

10. A method according to claim 7 wherein said disease is an autoimmune disease.

11. A method for increasing the number of IgD receptors on the surface of a T lymphocyte having such receptors, comprising contacting T lymphocytes with an. effective amount of an IgD-associated glycan, thereby increasing the number of said IgD receptors.

12. A method according to claim 11 wherein said glycan is in aggregated or polymerized form.

13. A method according to claim 11 wherein said glycan comprises a neoglycoprotein.

14. A method according to claim 11, wherein said glycan is selected from the group consisting of:

and

wherein R¹ is NeuAc in an α2→3 linkage or H, R² is NeuAc in an α2→6 linkage or H, R³ is GlcNAc in a β1→4 linkage or H, R⁴ is Fuc in an α1→6 linkage or H and R5 is Glc in an α1→3 linkage or H.

15. A method for inhibiting an immune response in a subject comprising:

(a) obtaining lymphocytes including CD8+ T lymphocytes from the subject;

(b) increasing the number of IgD receptors on the surface of said CD8+T lymphocytes by treating said lymphocytes by a method according to the claim 11; and

(c) administering the treated CD8+T lymphocytes to said subject,

thereby inhibiting said immune response.

16. A method for enhancing an immune response in a subject comprising:

(a) obtaining lymphocytes including CD4+ T lymphocytes from the subject;

(b) increasing the number of IgD receptors on the surface of said CD4+ T lymphocytes by treating said lymphocytes by a method according to claim

11,

(c) administering the treated CD4+T lymphocytes to said subject,

thereby enhancing said immune response.

17. A method for enhancing an immune response in a subject being immunized with an antigen or vaccine, comprising administering to said subject, in combination with the antigen or vaccine, an effective amount of an IgD- associated glycan.

18. A method according to claim 17, wherein said glycan is in aggregated or polymerized form.

19. A method according to claim 17 wherein said glycan comprises a neoglycoprotein.

20. A method according to claim 17, wherein said glycan is selected from the group consisting of:

and

wherein R¹ is NeuAc in an α2→3 linkage or H, R² is NeuAc in an α2→6 linkage or H, R³ is GlcNAc in a β1→4 linkage or H, R⁴ is Fuc in an α1→6 linkage or H and R⁵ is Glc in an α1→3 linkage or H.

21. A neoglycan comprising two or more of the same or different IgD-associated glycan monomers linked to a polymer.

22. A neoglycan according to claim 21 which is a neoglycoprotein, wherein said polymer is a protein.

23. A neoglycoprotein according to claim 22, wherein said protein is serum albumin.

24. A neoglycan according to claim 21 incorporated into a liposome.

25. A neoglycan according to claim 21, wherein said glycan is selected from the group consisting of:

and

26. A pharmaceutical composition useful for enhancing an immune response in a subject comprising a neoglycan according to claim 21 and a pharmaceutically acceptable excipient.

27. A pharmaceutical composition useful for enhancing an immune response in a subject comprising a neoglycoprotein according to claim 22 and a pharmaceutically acceptable excipient.

28. A pharmaceutical composition useful for enhancing an immune response in a subject comprising a neoglycoprotein according to claim 23 and a pharmaceutically acceptable excipient.

29. A pharmaceutical composition useful for enhancing an immune response in a subject comprising a neoglycan according to claim 24 and a pharmaceutically acceptable excipient.

30. A pharmaceutical composition useful for enhancing an immune response in a subject comprising a neoglycan according to claim 25 and a pharmaceutically acceptable excipient.

31. Apharmaceutical composition useful for blocking an immune response in a subject comprising an IgD-associated glycan and a pharmaceutically acceptable excipient.

32. A pharmaceutical composition according to claim 31 wherein said glycan is in non-polymerized or non-aggregated form.