Background of the Invention The etiology of multiple sclerosis (MS) is thought to involve one or more environmental components that predispose genetically susceptible individuals to develop the disease by disrupting immunological self-tolerance to central nervous system (CNS) myelin.

This autoimmune response then participates in the immunopathogenesis of MS, initiating or enhancing the inflammatory demyelinating response in the CNS that ultimately leads to axonal loss and irreversible neurological deficits.

The identity and mode of action of the environmental components in MS are not well understood, but one point of investigation focuses on molecular mimicry between self and pathogen-derived epitopes. Identification and characterization of the mode of action of such molecular mimics is important in the development of diagnostic and preventative protocols for MS.

Myelin oligodendrocyte glycoprotein (MOG) is a glycoprotein associated with the myelin sheath. MOG is a primary target antigen in experimental allergic encephalomyelitis (EAE) (an animal model for MS) in which an autoimmune mediated inflammatory response

causes demyelination of the CNS. MOG localization on the outer surface of the myelin sheath provides an ideal target autoantigen for antibody attack during autoimmune demyelination. Sensitizaticn with the variable-type immunoglobulin (IgV) domain of MOG initiates demyelinating autoantibody and encephalitogenic T-cell responses that act together to duplicate both the clinical course and immunopathology of MS in rodents and primates. See Storch et al. Brain Path. (accepted for publication, 1998); Kerlero et al., Eur. J. Immunol. 27: 3059-3069 (1997); Kerlero et al., J. Clin. Invest. 92: 2602-2608 (1993); Lu et al., J.

Neuro. Sci. 120: 99-106 (1993); Sun et al., J. Immunol. 146: 1490-1495 (1991). Enhanced T-and B-cell autoreactivity to MOG is also seen in a subset of MS patients, suggesting that MOG is also an important target autoantigen in the human disease. Xiao et al., J. Neuroimmunol.

31: 91-96 (1991).

Epidemiological studies have consistently identified an association between dietary factors, in particular dairy products, and the prevalence/incidence of MS. See Lauer et al., Neurol. 49: S55-61 (1997); Butcher et al., Med. Hypoth. 19: 169-178 (1986); Agranoff et al., The Lancet, November 2,1974, pp. 1061-1066. The relationship between MS prevalence and dairy product consumption in 27 countries and 29 populations all over the world has been studied with the SPEAR- MAN correlation test. Malosse et al., Neuroepidem. 11: 304-312 (1992).

A significant correlation was found between the consumption of liquid cow milk and MS prevalence. The results suggest that liquid cow milk could be a factor influencing the clinical appearance of MS. However, this possibility has, to our knowledge, never been explore in de :) th.

Butyrophilin (BTN) is a major component of the milk fat globule membrane (MFGM). The butyrophilin (Btn) gene and its protein

expression product for mouse is disclosed in Ogg et al., Mammal.

Genome 7: 900-905 (1996); and the bovine gene is disclosed in Davey et al., Gene 199: 57-62 (1997). The Btn cDNA and expression product for bovine is disclosed by Jack et al., J. Biol. Chem. 265: 14481-14486 (1990). The Btn cDNA and its expression product for humans is disclosed by Taylor en. al. Biochim. Biophys. Acta: 1306: 1-4 (1996).

Taylor also shows side by side comparison of bovine and human BTN.

See also, Banghart et al., J. Biol. Chem. 273: 4171-4179 (1998).

BTN is a glycoprotein of Mr approx. 66,000, with a single membrane anchor in the middle of the sequence. It is a member of the immunoglobulin superfamily (IgSF), a large group of proteins that function as either receptors, adhesive molecules, or components of the immune system. These proteins are characterized by the presence of Ig-like domains in the exoplasmic domain. BTN has two such domains, one of the intermediate type (Igi domain) towards the N-terminus, and another of the constant type (IgC1 domain), towards the membrane anchor.

Summary of the Invention It has previously been found that the lgl domain of BTN has a high degree of identity with the IgV domain in MOG. Gardinier et al., J.

Neurosci. Res. 33: 177-187 (1992). We have now discovered that immunization with exogenous BTN induces the production of antibodies which cross-react with endogenous MOG, and thus predispose animals to pathological autoimmune responses. This autoimmune response can be avoided by modifying the exogenous BTN to eliminate the cross- reactive regions.

Therefore, the present invention is directed to a method of producing

a modified non-human BTN, comprising providing a non-human BTN comprising an exoplasmic domain comprising an 191 domain and an IgC1 domain, and an cytoplasmic domain comprising a B30.2 region, and removing or modifying the 191 domain from the BTN. Preferably, the non-human BTN is bovine BTN.

The invention also includes a food containing a modified non-human BTN, wherein the modified BTN is lacking an 191 domain or contains a modified 191 domain. Preferably, the food is a dairy product.

The invention also includes a chimeric BTN, comprising an exoplasmic domain comprising a human lgl domain and a non-human IgC1 dumain, and an cytoplasmic domain comprising a non-human B30.2 region.

Preferably, the non-human IgC1 domain and the non-human B30.2 region are bovine. A DNA which encodes this chimeric BTN is also included, as is a vector containing the DNA.

The invention further includes a method of reducing a cross-reactive T-cell or antibody response to MOG in an animal upon introduction of exogenous BTN into the animal, the method comprising removing a portion of the exogenous BTN which elicits a cross-reactive T-cell or antibody response to MOG in the animal, and thereafter introducing the modified BTN into the animal. Preferably, the animal is a human and the exogenous BTN is bovine BTN.

The invention still further includes a method of reducing a cross- reactive T-cell or antibody response to MOG in a human upon introduction of non-human BTN into the human, the method comprising replacing a portion of the exogenous BTN which elicits a cross-reactive T-cell or antibody response to MOG in the human with a portion of human BTN which corresponds with the portion of the exogenous BTN to be replaced, to form a chimeric BTN, and thereafter introducing the

chimeric BTN into the cnimal. Preferably, the exogenous BTN is bovine BTN.

The invention also includes a food containing a modified non-human BTN, wherein the non-human BTN has been modified by removing a portion of the non-human BTN which elicits a cross-reactive T-cell or antibody response to MOG in a human ingesting the non-human BTN.

Preferably, the food is a dairy product.

The invention further includes a food containing a chimeric BTN, wherein the chimeric BTN has been made by replacing a portion of a exogenous BTN which elicits a cross-reactive T-cell or antibody response to MOG in a human ingesting the exogenous BTN with a portion of human BTN which corresponds in function with the portion of the exogenous BTN which is replaced. Preferably, the food is a dairy product.

The invention still further includes a DNA which encodes a modified non-human BTN, wherein the expressed modified non-human BTN is lacking an lgi domain. Preferably, the non-human BTN is bovine BTN.

The invention also includes a DNA which encodes a modified non- human BTN, wherein in the expressed modified non-human BTN a portion of the non-human BTN which elicits a cross-reactive T-cell or antibody response to MOG in a human upon introduction of the non- human BTN into the human is absent. Preferably, the non-human BTN is bovine BTN.

The invention furtherncludes a DNA which encodes a chimeric BTN, wherein in the expressed chimeric BTN a portion of a non-human BTN which elicits a cross-reactive T-cell or antibody response to MOG in a human ingesting the non-human BTN is replaced with a portion of human BTN which corresponds with the portion of the non-human BTN

which is replaced. Preferably, the non-human BTN is bovine BTN.

The invention further includes a transgenic mammal in which the endogenous Btn gene is optionally removed, the mammal containing a rDNA construct in at least the mammary epithelial cells of the mammal, the rDNA construct comprising any of the DNA sequences recited above, and a promoter sequence which is operably linked to and drives expression of the DNA sequence to produce a modified or chimeric BTN, wherein the rDNA construct is integrated in the mammal such that the DNA sequence is expressed in the mammary gland of the transgenic mammal and the modified or chimeric BTN is present in the milk of the mammal.

The invention further includes a method of producing a modified or chimeric BTN, comprising producing milk in a transgenic mammal as described above, and collecting the milk produced by the transgenic animal.

The invention also includes diagnostic applications. One such method is a method for detecting susceptibility to multiple sclerosis in a human patient from ingesting non-human BTN, comprising administering at least the 191 domain of the non-human BTN to a patient to generate a T- cell or antibody response in the patient to the 191 domain, collecting a sample from the patient containing T-cells or antibodies, incubating the sample with a synthetic peptide comprising the IgV domain of human MOG; and determining any binding between the T-cells or antibodies in the sample and the synthetic peptide, thereby detecting susceptibility to multiple sclerosis in a patient from ingesting the non-human BTN.

The invention also includes a method for detecting susceptibility to multiple sclerosis in a patient from ingesting bovine BTN, comprising collecting a sample from the patient containing at least the IgV domain

of human MOG, incubating the sample with an antibody which is specific for the 191 domain of bovine BTN, and determining any binding between the antibody and the sample, thereby detecting susceptibility to multiple sclerosis in a patient from ingesting the bovine BTN.

The invention also includes the use of defined human MHC class 11 binding motifs to identify the presence and/or frequency of non-human BTN/human MOG cross-reactive responses.

The invention also includes a method of identifying a patient at risk of developing a pathogenic autoimmune response to MOG upon exposure to non-human BTN, the method comprising screening a gene of a human patient to identify amino acid polymorphisms in the gene with domains which are at least 50% identical to the Igl domains of the non- human BTN.

The invention further includes the use of synthetic or native peptides comprising sequences from a non-human BTN to induce immunological tolerance to the non-human BTN in a human patient.

The invention still further includes the use of a DNA vaccine containing a construct based on native, recombinant or modified mRNA, cDNA or genomic DNA encoding for a non-human BTN to induce immunological tolerance to the non-human BTN in a human patient.

Brief Description of the Drawings Fig. 1 shows representative spinal cord histopathology of actively BTN-Ig immunized DA rats (a) and adoptive transfer of BTN, 6_$,-specific T-cell lines in naive DA rats (b-d).

Fig. 2 shows the mutually cross-reactive T-cell response provoked from amino acid sequence 74-90 of MOG and BTN.

Fig. 3 shows the potentially pathogenic autoantibody response to

MOG induced by BTN: (a) sera from DA and BN immunized with BTN- Ig; (b) epitope mapping of DA rat immunized with BTN-Ig; (c) Anti-BTN- Ig antisera specifically stain the surface of MOG-transfected LTK cells (filled histogram) compared to sham transfected LTK cells (line histogram).

Fig. 4 shows a comparison between the amino acid region of bovine BTN and rat MOG.

Fig. 5 shows comparisons of amino acid sequences of cow, human, guinea pig and mouse BTN.

Fig. 6 shows (a) the structure of bovine BTN, and (b) the cleavable signal at the N-terminus (dotted line and arrow), the 191 and IgC1 domains and the exoplasmic domain in the B30.2 region in the C- terminal cytoplasmic domain.

Detailed Description of the Invention Bovine BTN is shown in Fig. 6 (a). BTN is the major integral protein associated with the MFGM in milk of many species that may play a role in the mechanism of milk secretion. See U. S. Provisional Application 60/025,006 filed on August 20,1996, hereby incorporated by reference.

All lactating species express BTN, which is characterized by the two domains shown most clearly in Fig. 5, and described above. The first domain (Igl) of the protein is composed of the first 120 amino acids of the N-terminal region, and it is this region that is shown to have 50% identity with the IgV domain of MOG. This 50% identity is consistent with or conserved from species to species. Within the 191 domain, BTN stimulates the production of both antibodies and T-cells which cross- react with MOG. Autoimmune responses between autologous MOG and BTN are normally minimized due to evolutionary pressure over the past

60 million years (plus). However, an autoimmune response to MOG is initiated by exposure to BTN from another mammal. Thus, the problem arises on exogenous protein ingestion. One predisposed (i. e., one having an appropriate phenotype and exposure to BTN at an inappropriate age) can produce a B-cell or T-cell response to BTN which cross-reacts with MOG, thus priming the individual for MS. Upon elimination of the portion of BTN which produces a cross-reaction with MOG, therefore, one can prevent triggering MS in an individual having the predisposition discussed above.

The term"exogenous BTN"broadly refers to BTN which is not native to the animal into which it is introduced. When speaking in terms of human introduction, the term"exogenous BTN"means non-human BTN.

Preferably, the term means bovine BTN, or any other non-human BTN which is typically ingested by humans in dairy products, such as goat BTN or sheep BTN.

In the present invention, it is preferred to digest the lgl domain of the exogenous BTN prior to consumption of this milk fat protein in a dairy product. Thereby, one can eliminate the threat of contracting MS through this route. Digestion of the 191 domain may be accomplished using a protease, such as trypsin. Protein digestion can be terminated, if need be, by protease inhibitors such as trypsin inhibitor. Controlled protein digestion is well known to those of skill in the art.

In another embodiment of the invention, one may humanize at least a portion of the Igl domain of non-human BTN (i. e., from cows or other species providing milk and/or its products), thereby forming a chimeric BTN and removing the potential pathogenic exogenous domain of the protein. One method involves replacing a portion of the non-human BTN which elicits a cross-reactive T-cell or antibody response to human

MOG with a portion of human BTN which corresponds with the portion of the non-human BTN to be replaced.

By"corresponds"herein is meant that the two portions (i. e., the portion of human BTN and the portion of non-human BTN) have the highest percentage amino acid identity possible when the entire amino acid sequences are compared. Percentage identity in this application is determined using a variety of algorithms known in the art. An example of a useful algorithm in this regard is the algorithm of Needleman and Wunsch, which is used in the"Gap"program by the Genetics Computer Group. This program finds the alignment of two complete sequences that maximizes the number of matches and minimizes the number of gaps. Another useful algorithm is the algorithm of Smith and Waterman, which is used in the"BestFit"program by the Genetics Computer Group.

This program creates an optimal alignment of the best segment of similarity between two sequences. Optimal alignments are found by inserting gaps to maximize the number of matches using the local homology algorithm of Smith and Waterman. It is preferred to use the algorithm of Needlemas and Wunsch to compare amino acid identity in the present case.

Human and non-human BTN are shown in alignment in Fig. 5. From this Figure, and the algorithms referred to above, those of skill in the art can ascertain which portions of human BTN should replace corresponding portions of non-human BTN.

The present invention includes genetic manipulation of the endogenous BTN gene of an animal, such that at least a portion of the human Igl domain is introduced into the germ line of the animal, replacing the corresponding portion of the endogenous Igl domain.

Vector DNA comprising the animal endogenous gene with the desired

altered sequence and appropriate genetic and selection markers is introduced into blastocytes by micro-injection, which is then implante into pseudo-pregnant female recipients. Thus, subsequent generations will also express the desired humanized polypeptide in their milk.

The invention also contemplates diagnostic applications. For example, immunoassays may be conducted using MOG-and/or BTN- transfected cell lines, and/or recombinant extracellular domains expressed in eukaryotic systems to identify pathogenic antibody responses directed against defined epitopes of human MOG and/or non- human (e. g., bovine) BTN. Synthetic or native peptides comprising sequences from a non-human BTN may be used to determine susceptibility to MS in a human patient. The immunoassays may be in any form known to those of skill in the art. For instance, direct and indirect binding assays, competitive assays, sandwich assays and the like are generally described in U. S. Patent Nos. 4,642,285; 4,376,110; 4,016,043; 3,879,262; 3,852,157; 3,850,752; 3,839,153; 3,791,932; and Harlow and Lane, Antibodies, A Laboratory Manual. Cold Spring Harbor Publications, NY (1988), each of which is hereby incorporated by reference.

Immunoassays may also be conducted based on the use of defined MHC class 11 binding motifs to identify the presence/frequency of BTN/MOG cross-reactive T-cell responses.

In addition, genetic screening may be conducted to identify amino acid polymorphisms in human MOG and related genes with domains which are homologous to the lgl domains of exogenous (i. e., non-human, e. g., bovine) BTN (see Henry et al, Immunogenetics 46: 383-395 (1997); Ruddy et al., Genome Res. 7: 441-456 (1997); Tazi-Ahnini et al., Immunogenetics 47: 55-63 (1997)) to identify those at risk of developing

a pathogenic autoimmune response to MOG upon exposure to exogenous BTN.

The invention also is directed to the use of synthetic or native BTN peptides to induce immunological tolerance to BTN, and thus autoimmune tolerance to MOG. The peptides may be administered nasally, orally, intravenously or subcutaneously. Alternatively, a DNA vaccine containing a construct based on native, recombinant or modified mRNA, cDNA or genomic DNA encoding for BTN may be used. The vaccine is injected into the muscle tissue, for example, where the nucleic acid is taken up by host cells which then start expressing the protein.

The protein then serves as an antigen which stimulates an immune response.

EXAMPLE 1: Testing Molecular Mimicry between BTN and MOG Lewis, Dark Agouti (DA) and Brown Norway (BN) rats were used to test the molecular mimicry between BTN and MOG. Three strains of rats were used to provide a level of genetic diversity that is representative of the genetic heterogeneity seen in man. Although all three strains are susceptible to MOG-induced EAE, immunization with bovine MFGM or BTN-Ig failed to induce clinical disease. However, histopathological evaluation of the CNS revealed that both immunogens induced a sub-clinical inflammatory response in the CNS of DA rats, but not in Lewis or BN rats (Fig. 1a). This observation demonstrated that BTN elicits an autoreactive response in the CNS in a strain-dependent manner, but did not clearly identify the target antigen.

To determine whether a cross-reactive T-cell response to MOG was involved, we initially analyzed the ability of a panel of overlapping BTN-Ig peptides (B6387, B76100 and B7490) as well as to BTN-Ig to

stimulate encephalitogenic MOG-specific T-cell lines derived from each strain. A specific proliferative response was seen (Fig. 2a). A similar cross-reactive T-cell response was identified in BTN-specific T-cell lines selected in vitro using the BTN peptide representing the overlapping sequence B7490 (BDHIAX that also recognized the corresponding MOG peptide M7490 (MESIG) (Fig. 2b). Both MOG-and BTN/BDHIA-specific T- cell lines exhibited a similar level of proliferation in response to the homologous peptides, BDHIA and MESIG The antigen-specific response of these T-cell lines was restricted to major histocompatibility complex (MHC) class 11, as demonstrated by blocking assays using the mAbs OX6 (RT1 B), OX17 (RT1 D) and OX18 (class I MHC) (Figs. 2c and 2d).

We next demonstrated that immunization with BTN initiates an encephalitogenic MOG cross-reactive T-cell response in vivo by adoptive transfer into naive DA recipients. The i. v. injection of 5 x 106 to 10'BDHIA-specific T-cell lines induced an intense inflammatory response in the CNS of all recipients (Fig. 1 b) that initiated mild clinical disease characterized by weight loss and tail/hind limb paraparesis (Fig. 1c). Detailed immunohistochemical analysis of the lesions revealed that they consisted of large number of T-cells concentrated in the perivascular space. Migration of infiltrating T-cells into the parenchyme and macrophage recruitment into the CNS were both minimal (Fig. 1d), a situation similar to that described following the adoptive transfer of T-cells specific for MOG in the Lewis rat. These observations clearly demonstrate that"molecular mimicry"between MOG-and BTN-derived T-cell epitopes is sufficient to trigger an encephalitogenic class 11 MHC-restricted T-cell response. The MHC haplotype-dependence of this response reflects the importance of conserving both MHC anchor residues between the two antigens, while

at the same time sufficient epitope structure to ensure T-cell activation.

These limitations will restrict the induction of a MOG-reactive T-cell response by BTN to specific MHC class 11 haplotypes, in this case RT1. Bave ion the DA rat. The importance of the role of MHC rather than "background genes"in determining this response was illustrated by the fact that a second RT1. BaV'haplotype strain, ACI, mounted an identical encephalitogenic BpHIA-specific T-cell response (data not shown).

However, unlike the strict MHC haplotype-dependence of the cross-reactive T-cell response, a cross-reactive antibody response to MOG was observed in all three rat strains following immunization with BTN-Ig. The anti-MOG titer after immunization with BTN was greatest in BN rats and lowest in Lewis rats (BN » DA>Lewis) (Fig. 3a) (Lewis not shown). This was not due to the induction of a response to the histidine-tag on these bacterial products, as epitope mapping revealed that this response was directed against multiple linear MOG epitopes, in particular the sequences 1-39 and 63-125 (Fig. 3b). Moreover, in BN rats the titer was such that antibody binding could be directly demonstrated by FACS analysis of MOG-transfected fibroblasts (Fig.

3c). This is a crucial observation in that it demonstrates that the processing and preser. tation of BTN-Ig initiates an antibody response to MOG epitopes exposed at the cell surface, a prerequisite if the autoantibody response is to be pathogenic in vivo. Sensitization with exogenous BTN can therefore disrupt B-cell tolerance to induce a pathogenic autoimmune response to MOG across a wide range of genotypes.

EXAMPLE 2: Construction of Transgenic Animals The production of transgenic mammals containing an exogenous DNA

sequence encoding a desired protein or polypeptide in its germ! ine is accomplished by procedures well-known in the art. For example, see Rosen, U. S. Patent 5,304,489 (transgenic mice) and Clark et al., U. S.

Patent 5,322,775 (transgenic sheep), each of which is hereby incorporated by reference. The process generally comprises injecting the desired DNA into the pronucleus of fertilized oocytes, transferring the oocytes to pseudopregnant surrogate mothers, and screening the offspring for integration and expression of the foreign gene. To construct the transgenic animals of the invention, the injected DNA is a rDNA construct comprising a DNA sequence encoding a modified or chimeric BTN and a promoter sequence which is operably linked to and drives expression of the DNA sequence. Preferably, the endogenous Btn gene is removed to prevent the secretion of endogenous BTN.