WO1997026538A1 - Allele-specific peptide epitope strategy for vaccine development - Google Patents

Abstract

This invention provides a method for the identification of peptides that are capable of specific binding to HLA class II heterodimers expressed from exogenous DR genes in immortalized B lymphocyte cells that do not express endogenous HLA class II genes and that do not process proteins for presentation by HLA class II proteins due to defective DM gene expression. Also provided are compositions of peptides identified according to the described methods. The invention further provides a peptide derived from the Rubella E1 protein that is capable of interacting with DRB1 type D3 heterodimers.

Description

ALLELE-SPEC- FIC PEPTIDE EPITOPE STRATEGY FOR VACCINE DEVELOPMENT

Field ofthe Invention

The invention pertains to methods for identifying peptides suitable for vaccine development that are capable of binding specifically to various proteins encoded by the human histocompatibiiity region. The invention also provides compositions containing peptides capable of binding to HLA class II proteins encoded by several different DRB alleles.

Background ofthe Invention The major histocompatibiiity complex.

Two main types of T lymphocytes participate in the body's immune responses. The first of these are the T helper cells, which express the CD4 antigen and collaborate with B cells to trigger production of blood-borne immunity by providing signals necessary for B lymphocytes to secrete humoral antibodies. Helper T cells also activate other immune system components, in part by secreting hormone-like molecules called cytokines. The second main type of T cell expresses the CD8 antigen and functions to lyse virus-infected cells and to activate macrophages. These latter cells are called cytotoxic T cells, or CTLs. CTLs also are capable of producing cytokines. Among the most important of the cytokines is interleukin-2 (IL-2), which is produced by both CD4⁺ and CD8⁺ T cells (Abbas et al., 1994; Roitt, 1994, both of which are hereby incorporated by reference). One function of IL-2 is to stimulate the proliferation of T cells that have been "activated" by exposure to antigens. T helper cells begin producing IL-2 as an initial response to activation. This IL-2 stimulates the activated helper cells to proliferate and to produce more IL-2 and other cytokines. The other cytokines augment the proliferation of the activated T cells and also stimulate other cells ofthe immune system to mount an appropriate immune response. The first stage in this cascade, T cell activation, begins when a naive T cell encounters a recognizable antigen in a biologically appropriate context.

Both types of T lymphocytes recognize only antigens that are linear peptides, and can do so only when the antigens are "presented" on the surfaces of "accessory presenting cells" (APCs) in the form of discrete peptide epitopes. The T cell receptor, a specialized polymorphic protein on the surface of T cells, specifically recognizes individual peptides presented by APCs. APCs present antigenic epitopes to T cells by first proteolytically cleaving the antigen into smaller peptides. These small peptides are then transported to the cell surface in close noncovalent association with proteins produced by the APCs major histocompatibiiity complex region genes (MHC region genes). In humans, the MHC region is called "the HLA region." The actual target that T cell receptors recognize and interact with is a physical combination of an antigenic peptide and an MHC protein that is presented on the APC surface. B cells are capable in vitro and in vivo of performing the APC function for T helper cells. (Roitt, 1994; Abbas et al., 1994).

CTLs and helper T cells bear a functional correspondence to the two major classes of MHC proteins. Class I MHC proteins present processed peptides mainly to CTLs, while class II MHC proteins present antigen primarily to T helper cells. However, in humans, about 30% of CTLs interact with peptides presented by class II MHC proteins.

The binding to MHC proteins of antigen-derived peptides is a crucial step in the immune response because this binding is an essential prerequisite for T cell recognition of antigens. For class II MHC proteins, peptide binding induces a conformational change in the protein, and the MHC-peptide complex is thereafter transported to the cell surface where it is displayed to T cells, thus activating the T cell through interaction with the T cell receptor protein. Structural Features of MHC Proteins.

MHC class I and class II molecules are membrane glycoproteins that have the capacity to bind and display an enormous variety of structurally diverse peptides on the surfaces of APCs. In keeping with this capacity, the genes encoding the MHC proteins are among the most polymorphic classes of genes known. For both class I and class II, the glycoprotein is a heterodimer composed of two separate gene

SUBSTITUTE SHEET (RULE 25) products, an α polypeptide chain and a β polypeptide chain. One α chain and one β chain associate to form a functional MHC heterodimer.

The amino acid changes underlying MHC polymorphism are confined to the limited regions ofthe MHC proteins that form a groove or cleft to accommodate the antigenic peptide. The polymorphic residues occur primarily in the β-sheet floor and the inner surfaces of the α-helices that line the cleft or groove. Class I HLA molecules are composed of a polymorphic 44 kD α chain, and a non-MHC-encoded β chain of about 11 kD, known as β2 microglobulin. The β2 microglobulin is not polymorphic, therefore the peptide-binding specificity of class I molecules is provided by only the α chain, of which two domains provide the peptide-binding groove. For class II heterodimers, both the α and β chains are MHC-coded and polymorphic, and one domain from the α chain and one domain f om the β chain associate noncovalently to form the groove. Because two polymorphic chains participate in antigen binding, class II heterodimers possess greater potential for polymorphic variability than class I molecules. For MHC π molecules, the α chain is about 32-34 kilodaltons, and the β chain about 29-32 kilodaltons in size.

The gene organization within the HLA class II region is complex. The class II region has three major subregions, DQ, DP, and DR. Each of these subregions contains α genes and β genes that encode the two polypeptides that form the mature αβ class II heterodimer protein. For example, in the DR region, a DRA gene encodes a DRα chain, and a DRB gene encodes a DRβ chain. The degree of genetic polymoφhism differs among the three subregions. In the HLA DQ region, both α genes and β genes are highly polymorphic, whereas in the DP and DR regions, the β gene exhibits a higher degree of polymorphism than the α gene. In individuals heterozygous at a particular class II allele, an additional dimension of polymorphism is provided by the formation of mature heterodimers from the association of α polypeptides from one parent with β polypeptides from the other parent.

The DR subregion is of particular importance for primary immune responses, as its protein products are expressed constituitively on B lymphocytes. Moreover, it is expressed in amounts about fifty-fold higher than the amounts of DP and about ten¬ fold higher than the amounts of DQ that are expressed. Hence, the DR subregion is highly germane to vaccine development.

Within each class II subregion of HLA class II are several genes that encode MHC polypeptides. For example, the DR subregion includes several DRB genes that encode polypeptides, namely DRB1, DRB3, DRB4, and DRB5 (Bodmer et al., 1994). However, the DR subregion probably contains only a single DRA gene. The DRA

SUBSTITUTE SHEET (RULE %>) gene encodes the DRα polypeptide. The several DRB loci encode DRβ chains, and many alleles are known at each DRB locus. Individuals may express one or two DRB loci per haplotype. (For review of HLA genetics and nomenclature, see Robinson and Nepom, 1993, which is hereby incorporated by reference.) The products ofthe various DR genes have been further classified into "types" according to their abilities to interact with a set of standard antibodies used for classification purposes. For example, using these antibodies, the DRBl -encoded polypeptides have been categorized into at least twelve "types", known as "DRI, DR2, . . ., DR12." Classification into these types depends primarily on the antigenic properties of polypeptide chains encoded by the polymorphic DRBl genes. The two most prevalent DRBl types in Caucasian populations are DR4 and DR3, which are expressed in 23% and 17%, respectively, of individuals examined (Robinson and Nepom, 1993). At least 95% of all Caucasians exhibit one of the more prevalent DRBl types, which are DRI, DR2, DR3, DR4, DR5, DR6, and DR7. An additional 5% express the less prevalent DR8, DR9, and DR10 serotypes. Among blacks and orientals, at least 98% exhibit one or more of DR types DRI -DRI 0 (Immunology of HLA, Vol. 2, 1988).

The DR types are reflected in a system of nomenclature that is used to identify each individual protein encoded within the DR region. To illustrate, at least eleven known alleles of the DRBl locus have been isolated that encode molecules all of which type as "DR4." These alleles are commonly called *0401, *0402, *0403, etc., though other systems of nomenclature exist. In this application, genotypes will be referred to using the World Health Organization nomenclature, in which a DR gene is designated by its locus followed by its allele, e.g., "DRB1*0401". Large numbers of alleles exist within the DP and DQ subregions of HLA as well. (Bodmer et al., 1994). Both DP and DQ molecules are structurally homologous to DR molecules and are likely to also be involved in presentation of peptide antigens to helper T lymphocytes. Characteristics of Peptides that Bind to MHC Proteins.

From a large number of binding studies, several general principles have emerged to describe the peptides capable of interacting with MHC molecules. The peptide-binding cleft in class I MHC proteins is constrained to accommodate peptides of only about 9 to 11 amino acid residues, while the cleft of class II proteins is open at the ends and can thus accommodate peptides that are longer and more variable in length. Peptides bound by MHC class II proteins range in length from about 13 to 25 amino acids. Because of the open cleft, class II proteins can accommodate peptides whose overall length exceeds the bound portion ofthe peptide. Thus, in contrast with class I interactions, class II interactions do not require that the peptides fit entirely within the binding cleft.

Another feature of peptide binding to MHC proteins is that the specificity of binding is not absolute. Peptides have been identified that can bind to more than one type of MHC protein, though the affinity of the binding with different types can vary over a wide range (Vignali et al., 1993). Similarly, individual MHC proteins may bind to more than one related peptide (Wucherpfennig et al., 1994). The study by Wucherpfennig et al. even demonstrated that DR-encoded proteins and DQ-encoded proteins can bind in some instances to the same peptide. Other studies have shown that peptides that bind equally well to different MHC proteins may differ in their stimulatory effects on T cells (Boehncke et al., 1993). Still, most studies have indicated that peptide binding to MHC proteins is generally type-specific.

Investigators have attempted to discern type-specific motifs, but these have proven difficult to identify (e.g., Marshall et al., 1994; Marshall et al., 1995). In one study where monosubstitutions were made in each position of a peptide to define amino acids critical for binding, most ofthe peptides continued to bind just as well as the unsubstituted peptide (Cotner et al., 1993). Other studies have revealed broad motifs, but none that were inclusive for all the peptides that bound best to a given MHC type. For example, Boehncke et al. (1993) found that only a few residues were critical for ensuring the binding of peptide epitopes to class II MHC molecules, but some amino acid changes resulted in altered biological responses. Still, even though no invariant motifs have emerged, progress has been made towards the rational design of peptides with a less than random likelihood of binding to an MHC II protein of a designated type. Numerous studies have illuminated general patterns by studying peptides that bind to particular MHC II types. For example, Sette et al. (1993a; 1993b) have identified anchor residues for DR4-binding peptides. Other studies of HLA DR, DQ, and DP, have shown that the peptides usually have three to four hydrophobic or aromatic anchors, the first and last being separated by five to eight residues (Falk et al., 1994). The Falk et al. study indicated that the peptide motifs for HLA- DR1, DR5, DQ7, and DPw4 were allele-specific and differed by spacing and occupancy of anchors. Krieger et al. (1991) identified four residues particularly important in the binding of a peptide that interacted with DR7. Hill et al. (1994) have determined a single, key hydrophobic side chain that interacts in a conserved pocket in both DRI and DR4 proteins. In another study, relatively invariant anchor residues at positions 1 and 4 were identified for a series of peptides tested for binding to DRI,

SUBSTITUTE SHEET (RULE 28) DR4 and DR11, while the anchor ^"residue at position 6 differed among the three DR types (Hammer et al., 1993). Hammer et al. concluded also that almost all DR- binding peptides contained an aromatic amino acid followed, after two variable residues, by an amino acid with hydrophobic side chains. Hammer et al. (1994) also identified a peptide-binding motif in which six of seven positions showed enrichment of certain residues, and used this information to successfully identify DR-binding sequences from natural proteins. Sidney et al. (1992) have elucidated some structural requirements for peptide binding to DR3. Thus, for those designing peptide vaccines, rough guidelines exist for the rational selection of peptides that may upon further testing prove efficacious.

Vaccine Development and the Non-responder Problem.

In recent years, there has been a great deal of interest in the development of synthetic polypeptide vaccines. Such vaccines would alleviate the dangers of infection or immune suppression associated with vaccines based on whole virus or cells, and would bypass the difficulties of obtaining large amounts of difficult-to-culture pathogens. But before a peptide vaccine can elicit a protective immune response, it must interact successfully with MHC proteins. Because MHC II proteins, especially those encoded by DR genes, participate in the humoral antibody response, it is particularly important that peptide vaccines interact with DR gene products. Ideally, one developing a peptide vaccine would first examine antigenic pathogen-encoded proteins for specific peptides, i.e., epitopes, known to provide interaction with the critical MHC class II proteins. The term "epitope" is used here to describe a peptide that interacts with the binding cleft of an MHC heterodimer and that thereafter is presented on the surface of an APC in the form of a peptide- heterodimer complex. Unfortunately, the current state of knowledge provides only rough guidelines for predicting which peptides will have the capacity to bind with class II heterodimers. Hence, vaccine designers must test numerous candidate peptide epitopes for their ability to bind to MHC proteins (Cox et al., 1988; Golvano et al., 1990; Vignali et al., 1993). Present methods for vaccine development rely heavily on testing in mice.

Bach peptide candidate is first tested individually in separate mice, then the peptides exhibiting antigenic properties are combined and tested together in additional mice. In vitro methods of vaccine testing would greatly reduce the cost of vaccine development and also would increase the rate at which test results could be determined. Kilgus et al. (1989) have proposed a strategy that provides for the in vitro selection of efficacious vaccine peptides. This method uses a multi-step competition assay to identify those candidate peptides capable of binding to the products of particular HLADR alleles. This and other "trial and error" techniques are designed to sort through large numbers of potential vaccine peptides to select ones that bind HLA, followed by testing on different arrays of APCs.

A problem that confronts any vaccine designer is that genetic variation among HLA genes results in variable responses to individual vaccine antigens. Known as the "non-responder" problem, this variability reflects the genetically-based differences among individuals with respect to the ability of their MHC-encoded proteins to bind with particular peptides (Nepom, 1989). Some individuals simply cannot respond to a particular peptide epitope, while the degree of response may differ widely among those capable of responding. For example, an appreciable percentage of recipients have failed to respond to vaccines directed against both Rubella virus and hepatitis B (Alper et al., 1989; Tingle et al., 1985). This genetic polymorphism presents a particular barrier to the development of effective synthetic peptide vaccines because such vaccines include relatively few antigenic epitopes.

Various strategies can be applied to develop human vaccines that overcome the non-responder problem. One promising avenue has involved the search within antigenic proteins for "promiscuous" or "universal" peptides capable of binding to several different class I and class II types (Chicz et al., 1993; O'Sullivan et al., 1991; Tindle et al., 1991). Other vaccine designers have linked together several peptide epitopes pre-selected for their abilities to bind to several different HLA heterodimer molecules (Berzofsky et al., 1991a; Berzofsky, 1991b; Berzofsky et al., 1987). Another strategy has been to combine different peptide epitopes capable of stimulating T helper cells and CTLs, and also capable of eliciting a neutralizing antibody response (Berzofsky, 1994).

A highly desirable synthetic peptide vaccine would be one that includes combinations of peptide epitopes all derived from the same pathogen, each peptide binding to a different HLA class II type, and which taken as a group would bind to all or nearly all HLA class II types. For humans, a pool of peptide epitopes capable of binding individually to HLA DRBl types DRI through DR10 would elicit an immune response in >98% ofthe United States ethnic Caucasian, black, and oriental populace (Immunology of HLA, Vol. 2:, 1988). Moreover, if a pool of synthetic peptides known to be specific for DRB 1 types

DR1-DR10 were available, it would be useful for in vitro assays to determine whether

SUBSTITUTE SHEET (RULE 2β) individuals in a population were immune or susceptible to the pathogen from which the group of peptides had been derived. When incubated in culture with the peptide pool, white blood cells from an immune individual would respond by proliferating and releasing cytokines such as IL-2 into the culture medium (Abbas et al., 1994; Roitt, 1994). The proliferation would be easily measurable by counting the numbers of cells in the culture, and assays for IL-2 are well-known in the art (Berzofsky et al., 1991; Abbas et al., 1994; Roitt, 1994). If single peptides instead of pooled ones were used for this in vitro assay, the results would also indicate whether the tested individual expressed the specific DR protein with which the peptide was known to interact. To prepare a pool of peptides such as described above, the general strategy would be to first identify an organism for which current vaccines are inadequate, next to identify a surface or envelope protein from that pathogen for which the amino acid sequence is known, and to analyze this sequence for motifs that may bind to HLA DRBl types DR1-DR10. Once synthesized and tested for their actual binding specificities, peptides with the desired binding properties could be tested further as a combination of individual peptides or melded into a single polypeptide containing multiple epitopes. To form a single polypeptide, the individual peptides could be linked in a linear fashion, or could be linked to form a branched structure (e.g., Okuda et al., 1993). The development of such peptide pools and multi-epitope polypeptides would be furthered by the availability of better methods for identifying peptides capable of binding with the individual HLA class II types.

Summary ofthe Invention The invention provides a method for identifying a peptide capable of binding specifically to an HLA class II heterodimer consisting of a DRα chain in association with a DRβ chain. The first step of this method involves exposing a peptide to an immortalized B lymphocyte cell line that expresses DRα and DRβ chains encoded by exogenous genes that have been transfected into the immortalized B cells. The immortalized B lymphocytes used for this method are genetically deficient in DM expression, and thus are incapable of processing native antigen proteins into peptides that can interact with HLA proteins. Thus, these cells do not cleave the peptides to which they are exposed during the assay. After exposing a peptide to these B lymphocytes, the final step in identifying a specifically-binding peptide involves performing an assay for determining whether the peptide became bound to the heterodimer expressed by the B lymphocytes to which the peptide was exposed. The invention furthermore provides compositions of peptides whose specific binding properties have been identified using the methods described herein. These compositions include groups of peptides capable of binding with at least two DR heterodimers having different DRβ chains. In its most preferred form, this composition would include a group of peptides capable of reacting with different individuals expressing DRBl types DRI, DR2, DR3, DR4, DR5, DR6, DR7, DR8, DR9, or DRlO.

Additionally, the invention provides the peptide having the amino acid sequence LRLVDADDPLLR (SEQ ID NO: 1).

Brief Description ofthe Drawings FIGURE 1 is a graph representing the relative affinity of the DR4-specific peptide El(272-285) for DRB1*0401, *0403, *0301, and *1101 molecules, as is described in Example 5.

Detailed Description Definitions:

"HLA region" - A region on chromosome six ofthe human genome where the human major histocompatibiiity complex (MHC) resides.

"Express" - For a cell to express a gene means that the gene is transcribed and translated within the cell to produce a polypeptide chain.

"HLA DRA gene" - This term refers to a gene that maps to the MHC II subregion of the HLA region, and that encodes a protein chain of about 32 to 34 kilodaltons.

"HLA DRB gene" - This term refers to a gene that maps to the MHC II subregion of the HLA region, and that encodes a protein chain of about 29 to 32 kilodaltons. Several DRB genes are known, including DRBl, DRB3, DRB4, and DRB5. "HLA class II heterodimer" - This refers to an HLA class II protein consisting of a polypeptide chain encoded by a DRA gene non-covalently associated with a polypeptide chain encoded by a DRB gene, such that together the two chains form a heterodimer that provides a groove or cleft that can bind specifically with a peptide.

"HLA DRα chain" - This is a polypeptide expressed from a DRA gene from the MHC II subregion of the HLA region. The DRα chain can associate with a polypeptide encoded by a DRB gene to form a heterodimer capable of binding and presenting peptides.

"HLA DRβ chain" - This is a polypeptide expressed from a DRB gene from the MHC II subregion of the HLA region. A DRβ chain can associate with a polypeptide encoded by a DRA gene to form a heterodimer capable of binding and presenting peptides.

SUBSTrrUTE SHEET (RULE 26) "DM genes" - These genes map to the MHC II region ofthe HLA, and contain promoter regions similar to the promoters of class II DR, DQ, and DP structural genes. Properties of DM genes are addressed in Kovats et al. 1994; Kovats et al., 1995. Two loci are known, namely DMA and DMB, which encode an α and a β chain, respectively (Bodmer et al., 1994). DM expression is required for the processing of native exogenous proteins into peptides capable of combining with MHC II heterodimers and subsequently being presented on the cell surface.

"Cells deficient in DM expression" - In such cells, messenger RNA transcribed from DMA and DMB genes is reduced or absent. These cells are defective in the ability to proteolytically process antigenic proteins into peptides that can be presented by HLA class T proteins.

"Endogenous gene" - An endogenous gene is one that is naturally present in a cell's DNA.

"Exogenous gene" - An exogenous gene is one that is not naturally present in a cell's DNA, but that is introduced into the cell by artificial means. Such means may include fusion with another cell or with liposomes, introduction of a vector carrying the gene, such as an animal virus or a plasmid or other vector, or by other means such as may be devised.

"Epitope" - An epitope is a peptide that interacts with the binding cleft of an MHC heterodimer and thereafter is presented on the surface of an APC in the form of a peptide-heterodimer complex.

"Bind specifically" - For a peptide to "bind specifically" to an HLA class II heterodimer means that the peptide will combine with the heterodimer to form a complex measurable in accord with the binding assay of Example 2. This invention provides a method for identifying a peptide that binds specifically to an HLA class II protein heterodimer consisting of a DRα chain in association with a DRβ chain that are expressed from exogenous genes in an immortalized B lymphocyte cell line that is deficient in DM expression and that does not express endogenous HLA class II genes. Whether the peptide is bound to the HLA DR heterodimer protein can be determined by any standard method. Using the methods of this invention, one can create a series of B cell lines, each of which expresses one ofthe DRBl types known as DRI, DR2, DR3, DR4, DR5, DR6, DR7,

DR8, DR9, and DR10. By exposing a given peptide to each of the cell lines in this group, and determining whether the peptide binds to each DRBl type, one can determine the binding specificity of the peptide for the various DRBl types represented by this group of cell lines. The invention furthermore contemplates

SUBSTITUTE SHEET (RULE 25) application of the methods herein^" to types expressed by DRB3, DRB4, and DRB5 genes. Prior to this invention, it was not known that exogenous HLA proteins expressed in B lymphocytes deficient in antigen processing retained their ability to bind specifically with peptides. The invention further provides compositions of peptides identified by applying the method described above. Such compositions include two (or more) peptide epitopes derived from the same polypeptide, the first of these epitopes being capable of binding specifically with an HLA class II protein expressed by a first DRB gene, and the second epitope being capable of binding specifically with an HLA class II protein expressed by a second DRB gene. Such compositions include peptides that bind to MHC proteins encoded by DRBl genes, including protein types DRI, DR2, DR3, DR4, DR5, DR6, DR7, DR8, DR9, and DR10. It is contemplated within the scope of this invention that in some instances a single peptide may bind specifically with more than one DR type. In a presently preferred embodiment, the first DRB gene is DRB 1 type DR3, and the second DRB gene is DRB 1 type DRB4.

The invention provides further the peptide LRLVDADDPLLR (SEQ ID NO:l), which has been demonstrated here to bind specifically with a protein encoded by a DRBl type DIG gene.

A method is thus described herein for circumventing the genetic barrier that has constrained efforts to design universally effective polypeptide vaccines. The present invention employs genetically engineered B lymphocyte cell lines to offer a novel and improved strategy for identifying peptide epitopes capable of binding to all of the major types of class π HLA proteins. Such epitopes, once identified, can be administered simultaneously to provide an efficacious vaccine that will avoid the non- responder problem. Moreover, these epitopes can be used in an in vitro assay to identify individuals who are immune to the pathogen from which the antigenic protein was derived.

The subject binding assay exploits a B cell line established from a patient with the "bare lymphocyte syndrome" (BLS), a condition characterized by the absence of HLA DP, DQ, or DR products on the surface of B lymphocytes, macrophages, and dendritic cells. In these patients, class I MHC expression usually is normal, but they suffer from severe immune deficiencies and are highly susceptible to viral infections (Klein et al., 1993). Numerous investigators have established B lymphocyte cell lines from BLS patients and upon further study of the cells have learned that while intact HLA II genes are present in these patients, their cells do not transcribe these genes (Benichou and Strominger, 1991; Kovats et al., 1994; Steimle et al., 1995;

SUBSTITUTE SHEET (RULE 28) Hauber et al., 1995). One line of BLS cells, BLS-1 cells, has been intensely studied and shown to be deficient in proteolytic processing of antigens to which the cells are exposed (Kovats et al., 1995).

The subject binding assay utilizes BLS cells that were first transfected with a cloned DRA gene encoding the HLA DRα chain. The cell line thus established served as "recipient cells" for transfection with other cloned genes encoding DRβ chains corresponding to different alleles of the DRBl locus. The transfected DRβ chain genes included DRBl types DR4, DR3, and DR5. BLS cells expressing both a transfected DRα and a transfected DRβ chain are hereafter called "transferrent cell lines." Thus, continuous B cell lines can be created that express any MHC II type that is required for vaccine development, limited only by the availability of cloned HLA class II genes. A large number of class π HLA genes already have been identified and sequenced, and cell lines carrying these genes deposited at the European Collection of Animal Cell Cultures, from which the cells are available to all interested parties (Bodmer et al., 1994). Additional cloned HLA genes can be easily obtained using routine methods. (See, e.g., Kwok et al., 1988; Kwok et al., 1989; Krieger et al., 1991; Hiraiwa et al., 1990.).

For the subject binding assay, transferrent BLS cell lines are exposed to peptides derived from antigenic proteins to determine whether each peptide can bind to the particular DR heterodimer being expressed by that transferrent line. A peptide capable of binding to a cell line expressing a given DR type thus is likely to participate in eliciting an immune response if administered to a patient expressing that same DR type. Thus, the identification of such peptides is a useful first step in selecting peptides for further testing as candidate vaccine peptides. By transfecting B cells from BLS patients with specific cloned HLA DR genes, cell lines are created that present peptides in the context of specific DR types, hence providing a reagent for rapidly identifying peptide epitopes capable of binding specifically to MHC II heterodimers expressed from the transfected genes. Using such cell lines, individual peptides each capable of binding to a different MHC II type can be identified and the peptides so identified can be combined into a single pool for use in development of vaccines to which a major proportion ofthe population will be immunologically responsive.

In one application of this method, peptides with the desired binding capacities are tested in an animal model for their ability to elicit an in vivo immune response. Animals useful for this purpose include transgenic mice that express immunologically functional human HLA class II heterodimers. Methods for creating such mice are described by Fugger et al. (1994), and Altmann et al. (1995), both of which are hereby incoφorated by reference.

Using the Rubella virus El protein as a model system, the data described here focus on the DRBl gene to illustrate the efficacy of the proposed approach for identifying useful epitopes. Rubella vaccination is nearly universal in this country, but is associated with a substantial frequency of reported non-response (Tingle et al., 1985). The experiments described in Examples 5 and 6 show that two different epitopes from the Rubella El protein were recognized by their respective DR types both separately and when the epitopes were linked in tandem. These epitopes correspond to amino acids 254-266, which binds to type DR3 molecules, and to amino acids 272-285, which binds to type DR4 molecules. Experiments were performed to verify that the El (253-285) peptide, which contains both epitopes, remained intact throughout the selection binding assay.

EXAMPLES Example 1. Cell Lines and Transfectants

Human B cells used for this study were derived from a patient with congenital bare lymphocyte syndrome immunodeficiency, and were rendered immortal by transformation with Epstein-Barr virus (Kovats et al., 1994). Using the influenza hemagglutinin peptide HA (307-319), which binds indiscriminately to most DR types (O'SulIivan et al., 1991), it has been demonstrated that BLS-1 cells transfected with cloned DRA and DRB genes could present peptide but could not mediate the presentation of native influenza hemagglutinin (Kovats et al., 1994). Later studies (Kovats et al., 1995) showed also that BLS-1 cells lacked essential factors required for transcription of endogenous MHC class II genes. Kovats et al. (1994; 1995) noted further that when exogenous class II genes were expressed in these cells, the heterodimers were structurally abnormal. Moreover, the inability of these cells to present native antigens was shown to result from an additional defect that resided in the class LT-like DM gene, which maps to the HLA class II region and whose product is required for normal processing and presentation of whole protein antigens (Kovats et al., 1995). Several other BLS patients with APCs deficient in antigen processing have been identified by Mellins et al. (1990).

To create cell lines expressing specific DR types, BLS-1 cells were first transfected with cDNA corresponding to a HLA-DRA gene by cocultivation with retroviral vectors as previously described (Kovats et al., 1994; Kovats et al., 1995). Transfected cells expressing DRA were then used as recipient cells for transfection with various cloned DRBl genes (Kovats et al., 1994; Kovats et al., 1995). Thus, the α and β chains encoded by the exogenous DR genes were able to associate within the doubly-transfected recipient cells to form an MHC class π heterodimer. The transfected DR genes were transcriptionally regulated by a promoter derived from cytomegalovirus that functions at high efficiency in the BLS-1 cells. Expression of transfected HLA-DR molecules in the BLS-1 cells was verified in each case by immunocytofluorimetry (Kovats et al., 1994). Transfected BLS-1 cells used in the assays below included BLS-l.DR4Dw4, BLS-1.DR3, BLS-l.DR4wl3, BLS-l.DR4wl4 and BLS-1.DR5, which carry the DRA gene and single DRBl genes which, respectively, were DRB1*0401, *0301, *0403, *0404, and *1101. These correspond, respectively, to DR types 4, 3, 4, 4, and 5. Of these cell lines, all except the line expressing DRB 1*0403 were previously described (Kovats et al., 1994; Kovats et al., 1995).

Example 2. Binding Assay To determine whether individual peptides could be bound specifically by the products of particular transfected DR genes, IO⁷ BLS-1 cell lines, prepared in accordance with Example 1 and expressing exogenous DRA and DRBl genes, were incubated with biotinylated Rubella El peptide in RPMI 1640 (Grand Island Biologicals) with 5% dialyzed fetal calf serum for three hours at 37°C. Biotinylation was performed using standard techniques (Harlow et al., 1988, which is hereby incoφorated by reference). Unbound biotinylated peptide was then removed from the incubated cells by washing three times with phosphate buffered saline (0.01 M phosphate, 0.15 M NaCl, pH 7.2). Following the washes, cells were lysed in Tris- buffered saline (137 mM NaCl, 5 mM Kcl, pH 7.2, and containing 15 mg/1 of phenol red) with 1% Nonidet P40 (Particle Data Laboratories), 5O mM disodium ethylenediaminetetraacetate (EDTA), and protease inhibitors (phenylmethylsulfonyl fluoride (PMSF), feupeptin, and pepstatin). Lysates were centrifuged at 14,000 g for 10 minutes to remove cell debris.

Supematants from each lysate were incubated with streptavidin immobilized on beads (Ultralink, Pierce) for 60 min each at 4°C to capture complexes consisting of biotinylated peptide bound to MHC II heterodimers expressed by the transfected exogenous genes. To remove fortuitously-captured proteins, i.e., proteins not bound to biotinylated peptide, the streptavidin beads were washed twice with 50 mM Tris (pH 8.5), 500 mM NaCl, 0.5% desoxycholate (Sigma), 50 μM PMSF and then three times with 50 mM sodium phosphate (pH 7), 0.4 M urea, 50 μM PMSF. Specifically- captured HLA-peptide complexes were then eluted by boiling in 2% SDS, 0.02 M sodium phosphate (pH 7), 0.4 M urea, 0.1 mM EDTA, 100 μM PMSF. Eluted

SUBSTITUTE SHEET (RULE2B) fractions were electrophoresed ^"on 10% SDS poiyacrylamide gels, transferred to nitrocellulose, and developed for Western blotting using anti-DRBl monoclonal antibody XD5.A111 (Steimle et al., 1995). Binding of the anti-DR monoclonal antibody to protein on the Western blot was visualized by addition of goat anti-mouse Ig-horseradish peroxidase (Jackson Labs) and chemiluminescence assay with exposure to Amersham Hyperfilm-MP for 1-15 min. DRβ internal standards were co- electrophoresed in alternate lanes on the poiyacrylamide gel. Specific binding of the tested peptide was indicated when a sample derived from a cell lysate yielded a luminescent band that co-migrated with the internal DRβ standard. This assay can differentiate peptides that bind specifically from those that do not in that the absence of a luminescent co-migrating band reveals that a given tested peptide cannot bind specifically with the DR protein expressed by the tested cell line. The ability to differentiate binding and non-binding peptides is an important aspect of an assay for specific binding. Example 3. T Cell Assays

Cytotoxic CD4⁺ T lymphocytes specific for the Rubella El (272-285) CTL peptide have been previously described (Ou et al., 1994). This is the AtRCV2 cell line. Transfected BLS-1 cells were used as targets for recognition by this CTL line in the presence or absence of specific El peptides. The chromium release assay was performed essentially as described in Mishell et al. (1980), which is hereby incoφorated by reference. After incubation at 37°C for 1 h with 1 μM biotinylated El peptides in RPMI 1640, 1 x IO⁶ AtRCV2 cells were washed once with complete medium and internally labeled for 1 h at 37°C with 100 μCi of Na₂ ⁵¹CrO₄ (Amersham). Radiolabelled target cells were washed four times with medium, then incubated for 4 h at 37°C with T cells at different effectoπtarget ratios in round- bottomed 96-well plates (Nunc). All assays were performed in triplicate. Percent specific lysis was calculated according to the formula: ((Experimental counts per minute (CPM) - Spontaneous CPM Released)/(Maximal CPM Released - Spontaneous CPM Released)). Example 4. Peptide Synthesis

Peptides were synthesized with an Applied Biosystems 432 Peptide Synthesizer. For biotinylation, excess F-moc ε-amino caproic acid from Novachem in N,N-dimethyformamide (DMF) was added to the reaction vessel after deprotection of the N-terminal amino acid residue. Two caproic acid residues were added and the final coupling reaction was then carried out with excess biotin in 50% dimethylsulfoxide and 50% DMF. The peptides were then cleaved from the resin -lo¬

using 95% trifluoracetic acid, 5%^'thioanisole, and 5% 1,2-ethanedithiol. The peptides were precipitated in methyl-t-butyl ether and extracted with 30% acetonitrile in 0.1% trifluoroacetic acid. Residual resins were removed by filtration through a polypropylene column with a filter disc (Isolab, QS-Q). The filtrate was then lyophilized.

Example 5. Binding of Rubella Peptide El (272-285^ to

DRBl Alleles Expressed bv BLS-1 Cells

Biotinylated El (272-285) prepared according to Example 4 was tested for binding to various DRBl types expressed on BLS transferrent cell lines. Cells were lysed and the binding assay conducted as described in Example 2. After peptides were eluted from the HLA-peptide affinity-purified complexes, the eluate was subjected to poiyacrylamide gel electrophoresis and Western blotting analysis. Western blots were developed with monoclonal antibody directed to invariant regions of the HLA-DRβ chain as described in Example 2. This method was employed first to demonstrate the capacity of exogenous

HLA genes expressed in BLS-1 cells to display specific binding. These first experiments used the Rubella El peptide (272-285), which was known previously to bind specifically to DRBl type DR4 HLA proteins (Ou et al., 1994). This Rubella peptide has the sequence GEVWVTPVIGSQAR (SEQ ID NO:2). Biotinylated El (272-285) was tested for binding to transferrent BLS-1 cells expressing HLA

DR4Dw4, which is a DRBl type DR4 HLA protein. Cells were exposed to 0, 10, 20,

40, and 80 μM ofthe peptide. Detection ofthe peptide-MHC complex was achieved by lysis ofthe cells followed by affinity purification using streptavidin, and subsequent

SDS poiyacrylamide gel and Western blotting analysis with a monoclonal antibody to HLA-DRB1.

Gel electrophoresis of the eluted material revealed a distinct DRβ chain band in all samples that had received biotinylated peptide, but not in the control lane, which represented BLSl.DR4Dw4 cells in the absence of peptide. Moreover, the results indicated that the peptide had bound to the protein expressed by the transferrent cell line, and that the amount of peptide bound increased proportionally with the amount of peptide added to the cells. Dose-dependent binding of peptide was seen in the range from 10-80 μM for the DRB 1*401 molecule. This dose dependency reinforced the conclusion that the binding had been specific.

Other experiments further illustrated the ability of this method to define the type-specific binding pattern for a given peptide. The DR4-specific peptide El

(272-285) was incubated with BLS cells expressing five different DRBl types. The BLS-1 transferrents expressed either DRBl type DR3, type DR5, or one of three different type DR4 alleles. After being incubated with 0, 20, or 40 μM of peptide, the cells were lysed, the peptide-MHC complex affinity purified as before, and the peptide-bound MHC protein recovered and examined by gel electrophoresis as described. Densitometric quantification of the results indicated that the three DR4 proteins and the DRS protein bound to the target peptide, while the DIG protein failed to bind the peptide. The relative affinity of El (272-285) for these five MHC proteins is illustrated in FIGURE 1. Thus, this method provides a means for determining the relative affinity of a given peptide for an array of different MHC proteins.

Example 6. Design and Testing of Peptides that Will Bind to Predetermined HLA Proteins BLS transferrent cells can be used to facilitate the rational design of efficacious peptide vaccines. For the present studies, individual peptide motifs for binding to specific HLA types are extrapolated from studies of peptides eluted from HLA molecules (Rammensee et al., 1995). For example, all DR molecules apparently prefer to bind peptides with a primary hydrophobic anchor position, W in the El (272-285) peptide and V in the El (254-266) peptide, whereas secondary anchor positions are more allele-specific. DIG prefers a negatively charged amino acid in the fourth position relative to the primary anchor, whereas DR4 does not. By utilizing such general principles, a set of peptides can be constructed having a reasonable possibility of having the desired specificity, and tested using the methods of the subject invention. Other investigators have published on the subject of predicting the motifs characterizing peptides capable of binding particular HLA types (Verreck et al., 1994; Sette et al., 1993a; Sette et al., 1993b; Hammer et al., 1993; Hill et al., 1994; Chicz et al., 1993; Wucheφjennig et al., 1994).

Using this information as the basis for prediction of type-specific motifs, the Rubella El protein sequence was scanned for peptides that would be potentially suited for high affinity interaction with HLA type DR3. One such motif was identified in the sequence 254-266, LRLVDADDPLLR (SEQ ID NO:l), adjacent to but not overlapping with the El (272-285) epitope. Biotinylated El (254-266) was then tested for binding to multiple DRBl alleles, as was the longer peptide El (253-285), having the amino acid sequence RLRLVDADDPLLRtapgpGEVWVTPVIGSQAR (SEQ ID NO:3), and encompassing both epitopes. (Lower case letters represent spacer amino acids, and underlined amino acids are anchor residues that interact directly with the MHC binding cleft.) El (254-266) was incubated under binding conditions (as in Example 2) with four BLS transferrents, one of which expressed the DR3 allele *0301, the others of which expressed three different DR4 alleles, namely *0401, *0403, and *0404. Peptide binding was assayed as described in Example 2. The results indicated that of the transferrent lines tested, only those expressing type DR3 bound the El (254-266) peptide. Hence, the El (245-266) peptide could be useful in developing a peptide- based Rubella vaccine. These results further demonstrate that the BLS-1 transferrent cells provide a convenient model system for the rapid evaluation of the binding specificity of peptides derived from larger antigenic proteins. The next experiment demonstrated that extended peptides encompassing epitopes for more than one HLA type can still bind with their MHC counteφarts. The peptide El (253-285), which combines epitopes specific for DIG (254-266) and for DR4 (272-285) was biotinylated in accord with the procedure of Example 2. Then 0, 40, and 80 μM of the biotinylated peptide were incubated with four different antigen- presenting transferrent cell lines, expressing the DR alleles *0301, *0401, *0403, and *0404. The extent of binding was assayed as before. The four cell lines bound to the extended peptide with approximately equal intensity. These results thus illustrate that despite these cells' deficiency in proteolytically processing antigenic proteins, the exogenous MHC proteins they express can accommodate and specifically bind epitopes in the context of a peptide that is longer than one epitope.

Example 7. T Cells Can Recognize and Interact with Peptides Bound bv Transferrent BLS Cells The following experiment illustrates that peptide presented by BLS transferrent cells can be recognized by appropriate effector T cells. As discussed earlier in Example 5, the Rubella peptide El (272-285) represents an epitope previously known to bind with type DR4 HLA proteins. Hence, this peptide (combined with aDR4 protein) is capable of interacting with "DR4-restricted" T cells. The term "restricted" refers to the fact that individual T cells can recognize a peptide antigen only when the antigen is complexed with an MHC molecule that the T cell recognizes as "self." Hence, T cells derived from any given individual are "restricted" by the MHC class π molecules expressed by that individual's cells. The El (272-285) epitope thus is recognized by the T cell receptor expressed by AT177C5 cells, which are a clonal derivative ofthe human ATRVC2 T cell line, and which are restricted by the HLA-DRB1 molecule (Ou et al., 1994). The individual from whom AT177C5 cells were originated expressed a DR4 allele of DRBl, hence this line of T cells can recognize the El (272-285) peptide only in the context of a DRBl type DR4 or other related MHC protein.

The recognition by DR4-restricted AT177C5 cells of the El (272-285), El (254-266), and El (253-285) peptides was tested when the peptides were presented by BLS transferrent cells carrying either the DRB 1 *0403 gene, the DRB 1 *0301 gene, or transfected with both DRBl *0403 and DRB 1*0301 to create a "heterozygous" DR3/4 genotype (results shown in TABLE 1). Biotinylated El peptides were used to sensitize target cells at 1 μM. After the transferrent BLS cells were pulsed with the peptides, specific cytotoxicity was determined in ⁵¹Cr release assays (described in Example 3) at the effectoπtarget (E:T) ratios shown in TABLE 1.

The results indicated that the DR4-restricted T cells recognized the El (272-285) epitope both as a discrete peptide and when presented in the context ofthe longer El (253-285) sequence. The T cells did not lyse BLS cells pulsed only with the El (254-266) peptide, which was consistent with these T cells' being DR4- restricted. As expected, lysis of AT177C5 cells occurred only when the peptides containing amino acids 272-285 were presented by BLS-1 cells expressing DR4, but not when presented by BLS cells expressing only the DIG allele.

TABLE 1 % specific cytotoxicity

Peptide E E--TT rraattiioo D DRIG3 DR3/4 DR4

254-266 10 6.6 0 6.5

3 0 0 0

1 0 0 1.8

272-285 10 0 79.7 100

3 0 78.6 90.9

1 0 64.9 73.1 253-285 10 8.7 85.7 81.5

3 4.2 70.0 79.5

1 0 63.1 68.6

The data illustrated in TABLE 1 further indicate that the "heterozygous" cells expressing both type DR3 and DR4 were capable of binding multi-epitope peptides. Gel electrophoresis of the affinity-purified peptide-MHC complexes from heterozygous cells indicated that the El (253-285) peptide, which contains both epitopes, bound simultaneously to DR3 and DR4 proteins (data not shown). Hence, this model system provides a means for determining which multi-epitope peptides are efficacious in the absence of DM-mediated processing. Multi-epitope vaccines that do not require processing are highly^" desirable because they offer the prospect of vaccine preparations with greatly increased intrinsic immunogenicity (Ada, 1993). Example 8. Synthetic Polypeptide Vaccine for Tuberculosis The 30 kilodalton major secretory protein (30 kd protein) of Mycobacterium tuberculosis is an excellent candidate for vaccine development. It has been demonstrated that protective immunity against tuberculosis can be induced by vaccination with major extracellular proteins of Mycobacterium tuberculosis, including the 30 kd protein. (Horwitz, et al., 1995). Although Horwitz et al. used only conventional animal models to test whole proteins for their efficacy in protecting guinea pigs against tuberculosis, a peptide vaccine containing the pertinent epitopes is likely to also provide protection.

The 30 kd protein is examined for regions of amino acid sequence that conform with motifs associated with specific binding to various DRBl types.

Candidate peptides are identified and synthesized that incoφorate these motifs, and that have potential specificities for DR 1, DR2, DR3, DR4, DR5, DR7, DR8, DR9, and DRI 0.

To identify peptides with the desired binding capacities, BLS transferrent cells expressing each of the above-listed DR types are created by transfection with cloned DR genes as described in Example 1. Each cell Une is incubated with each candidate peptide, and binding assays are performed as described in Example 2. Peptides that bind specifically to each of the cell lines are selected for further testing as vaccine candidates.

Peptides chosen as vaccine candidates are first tested in transgenic mice expressing human MHC II genes (Fugger et al., 1994; Altmann et al., 1995). Peptides that elicit an immune response in these mice are then administered to human volunteers either as a solution of all of the peptides, or as a solution of a single polypeptide molecule into which all ofthe peptide epitopes have been linked together.

While the preferred embodiment of the invention has been illustrated and described, it will be appreciated that various changes can be made therein without departing from the spirit and scope ofthe invention.

SUBSTITUTE SHEET (RULE 2£5) REFERENCES

1. Abbas, A.K., et al., Cellular and Molecular Immunology, 2d ed., W.B. Saunders Co. (1994).

2. Roitt, Ivan, Essential Immunology, 8th ed., Blackwell Scientific Publications (1994).

3. Bodmer, J.G., et al., Tissue Antigens, 44:1-18 (1994).

4. Robinson, D M., and G. Nepom, Autoimmune Dis., 13:255-272 (1993).

5. Immunology of HLA, Vol. 2: Immunogenetics and Histocompatibiiity. Springer- Verlag, New York, 1988.

6. Vignali, D.AA et al., Eur. J. Immunol, 23:1602-1607 (1993).

7. Wucheφfennig, K.W., et al., J. Exp. Med, 179:279-290 (1994).

8. Boehncke, W.-H, et al., J. Immunol, 150:331-341 (1993).

9. Marshall, K W, et al., J. Immunol, 152:4946-4957 (1994).

10. Marshall, K W, et al., J. Immunol, 154:5927-5933 (1995).

11. Cotner, Tom, Autoimmunity, 16:57-67 (1993).

12. Sette, A, et al., J. Immunol, 146:2331-2340 (1993a).

13. Sette, A. et al., J. Immunol, 151:3163-3170 (1993b)

14. Falk, K., et al., Immunogenetics, 39:230-242 (1994).

15. Krieger, J.I., et al, J. Immunol, 146(7):2331-2340 (1991).

16. Hill, CM, et al, J. Immunol, 152:2890-2898 (1994).

17. Hammer, J, et al. Cell, 74:197-203 (1993).

18. Hammer, J, et al, Proc. Natl. Acad Sci. USA, 91 :4456-4460 (1994).

19. Sidney, J, et al, J. Immunol, 149(8):2634-2640 (1992).

20. Cox, J. H, et al, Eur. J. Immunol, 18:2015-2019 (1988).

21. Golvano, J, et al, Eur. J. Immunol, 20:2363-2366 (1990).

22. Kilgus, J, et al, Proc. Natl. Acad Sci. USA, 86: 1629-1633 (1989). 23. Nepom, G, N. E. J. Med, -321:7 '51-752 (1989).

24. Alper, CA, et al, N E. J. Med, 321(11):708-712 (1989).

25. Tingle, A.J, et al, J. Infec. Dis., 151(2):330-336 (1985).

26. Chicz, R.M, et al, J. Exp. Med, 178:27-47 (1993).

27. O'SulIivan, D, et al, J. Immunol, 147:2663-2669 (1991).

28. Tindle, et al, Proc. Natl Acad Sci. USA, 88:5887-5891 (1991).

29. Berzofsky, J.A, et al, J. of Clin. Invest. , 88:876-884 (1991a).

30. Berzofsky, J.A, FASEB J., 5:2412-2417 (1991b).

31. Berzofsky, J.A. et al, Immunol. Rev., 98:9-52 (1987).

32. Berzofsky, J.A, et al, PCT Patent Application No. PCT/US94/05142 (1994).

33. Okuda, K, et al, J. Molec. Recognition, 6:101-109 (1993).

34. Kovats, S, et al, J. Exp. Med, 179:2017-2022, (1994).

35. Kovats, S, et al, J. Clin. Invest, 96:217-223 (1995).

36. Klein, et al, J. Pediatrics, 123 :921-928 (1993).

37. Benichou, B, and Strominger, Proc. Natl. Acad Sci. USA, 88:4285-4288 (1991).

38. Steimle, V, et al. Genes & Dev., 9:1021-1032 (1995).

39. Hauber, I, et al, J. Exptl. Med, 181: 1411-1423 (1995).

40. Kwok, W.W, et al. In Immunology of HLA, Vol. 2: Immunogenetics and Histocompatibiiity. Springer- Verlag, New York, 1988.

41. Kwok, W.W, et al, J. Immunol 143:3598 (1989).

42. Krieger, J.I, et al, J. Immunol 146:2331 (1991).

43. Hiraiwa, AK, et al, Proc. Natl Acad Sci. USA 87:8051 (1990).

44. Fugger, L, et al, Proc. Natl Acad Sci. USA, 91:6151-6155 (1994).

45. Altmann, D.M, et al, J. Exp. Med, 181:867-875 (1995).

46. Mellins, E, et al. Nature, 343 :71-74 (1990). 47. Harlow, E, and D. Lane, Eds, Antibodies, A Laboratory Manual. Cold Spring Harbor Press, p. 341 (1988).

48. Ou, D, et al. Human Immunology, 39: 177-187 (1994).

49. Mishell, B.B, and S.M. Shiigi. Selected Methods in Cellular Immunology, W.H. Freeman & Co. (1980).

50. Rammensee, H. G, et al, Immunogenet. 41 : 178-228 (1995).

51. Verreck, et al, Eur. J. Immunol. 24:375-79 (1994).

52. Ada, G, "Vaccines," Ch.37, in Fundamental Immunology, 3d ed, W.E. Paul, Ed, pp. 1309-1352 (1993).

53. Horwitz, M.A, et al, Proc. Natl. Acad. Sci. USA 912:1530-34 (1995).

SUBSTITUTE SHEET (RULE 28) SEQUENCE LI STING

( 1 ) GENERAL INFORMATION :

(i) APPLICANT: Nepom, Gerald T. (ii) TITLE OF INVENTION: Allele-Specific Peptide Epitope

Strategy for Vaccine Development (iii) NUMBER OF SEQUENCES: 3 (iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: Dennis Shelton

(B) STREET: 1420 5th Avenue, Suite 2800

(C) CITY: Seattle

(D) STATE: Washington

(E) COUNTRY: U.S.A.

(F) ZIP: 98101-2347 (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 #1.30 (vi) CURRENT APPLICATION DATA:

(A) APPLICATION NUMBER: not yet assigned

(B) FILING DATE:

(C) CLASSIFICATION: (viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: Shelton, Dennis K.

(B) REGISTRATION NUMBER: 26,997

(C) REFERENCE/DOCKET NUMBER: VMRC18231 (ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: (206) 682-8100

(B) TELEFAX: (206) 224-0779

(2) INFORMATION FOR SEQ ID NO:l:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 12 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide

(A) DESCRIPTION; El (254-266) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(v) FRAGMENT TYPE: internal (vi) ORIGINAL SOURCE:

(A) ORGANISM: Rubella virus (xi) SEQUENCE DESCRIPTION: SEQ ID NO:l:

Leu Arg Leu Val Asp Ala Asp Asp Pro Leu Leu Arg

1 5 10 (2) INFORMATION FOR SEQ ID NO:2: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 14 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide

(A) DESCRIPTION: El (272-285) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(v) FRAGMENT TYPE: N-terminal (vi) ORIGINAL SOURCE:

(A) ORGANISM: Rubella virus (xi) SEQUENCE DESCRIPTION: SEQ ID NO:2:

Gly Glu Val Trp Val Thr Pro Val Ile Gly Ser Gin Ala Arg 1 5 10

(2) INFORMATION FOR SEQ ID NO:3:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 32 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: peptide

(A) DESCRIPTION: El (253-285) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(v) FRAGMENT TYPE: internal (vi) ORIGINAL SOURCE:

(A) ORGANISM: Rubella virus (xi) SEQUENCE DESCRIPTION: SEQ ID NO:3:

Arg Leu Arg Leu Val Asp Ala Asp Asp Pro Leu Leu Arg Thr Ala Pro 1 5 10 15

Gly Pro Gly Glu Val Trp Val Thr Pro Val Ile Gly Ser Gin Ala Arg

Claims

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. A method for identifying a peptide that binds specifically to an HLA class II heterodimer consisting of a DRα chain in association with a DRβ chain, comprising the steps of exposing the peptide to an immortalized B lymphocyte cell in which the DRα and DRβ chains of the heterodimer are expressed from exogenous genes, said cell being deficient in DM expression and endogenous HLA class II expression, and determining whether the peptide has bound to the heterodimer.

2. A peptide comprising the amino acid sequence LRLVDADDPLLR (SEQ ID NO: 1).

3. A composition of peptides identified according to Claim 1, the composition comprising first and second peptide epitopes both derived from the same polypeptide, the first epitope being capable of binding specifically with an HLA class π protein comprising the gene product of a first DRB gene, and the second epitope being capable of binding specifically with an HLA class II protein comprising the gene product of a second DRB gene.

4. The composition of Claim 3, wherein the first and second DRB genes are DRBl genes selected from the group consisting of DRBl types DRI, DR2, DR3, DR4, DRS, DR6, DR7, DR8, DR9, and DR10.

5. The composition of Claim 4, wherein the first DRB gene is DRBl type DR3, and the second DRB gene is DRBl type DR4.