Transplantation Immunology: Peptides in the groove cause histoincompatibility
 
H.J. Stauss, C. Thomas, P. Nathan, E. Sadovnikova, 0. Neth and L. Yin   
In: Zander AR et al. (eds) Gene Technolgy, Stem Cell and Leukemia Research,
Nato ASI Series H: Cell Biology, Vol 94, Springer-Verlag,
Berlin Heidelberg New York London, pp 381-396

Imperial Cancer Research Fund, Tumour Immunology Unit, 91 Riding House Street, London WC1 8BT, United Kingdom.


Summary:

Organ and bone marrow transplantation is frequently complicated by immune responses against major and minor histocompatibility antigens. This article will discuss current concepts of the molecular nature of histocompatibility antigens. The functional units that trigger immune responses by T lymphocytes are peptides bound to the peptide binding groove of major histocompatibility complex (MHC) class I and class II molecules. In MHC mismatched transplant situations, T lymphocytes recognise donor MHC molecules displaying donor derived peptides. In MHC matched transplant situations, T cell immunity is directed against donor-specific peptides bound to the groove of MHC molecules. These donor-specific peptides are derived from polymorphic cellular proteins, named minor histocompatibility (mH) antigens. We propose that immune responses against mismatched MHC antigens and mH antigens are probably directed against a small number of immunodominant peptides. In future, it may become possible to exploit such peptides for tolerance induction, which may lead to specific downregulation of unwanted T cell responses in organ and bone marrow transplantation.


Antigen Recognition by CD4+ T helper (Th) lymphocytes and CD8+ cytotoxic T lymphocytes (CTL):

T lymphocytes do not recognise protein antigens directly, but rather peptide fragments which are presented by MHC class I and class II molecules. Peptides presented by class I molecules are recognised by CD8 expressing CTL. The antigen receptor (TCR) of CTL recognises pep tides bound to the peptide binding groove of class I molecules, which is formed by the membrane distal alfa1 and alfa2 domains (Bjorkman, et al., 1987a; Bjorkman, et al., 1987b). Simultaneously, the CD8 molecule interacts with the membrane proximal class I alfa3 domain (Salter, et al., 1989). Similarly, the TCR of Th cells recognises peptides bound to a groove formed by the membrane distal class II alfa1 and ß1 domains (Brown, et al., 1993), while the CD4 molecule interacts with the membrane proximal ß2 domain. The mechanism of MHC restricted peptide recognition by CTL and Th lymphocytes applies for the majority of known immune responses except for superantigens, which bind to the side of MHC class II molecules and cross-link TCRs of CD4 Th cells expressing certain ß-chain variable gene segments (Choi, et al., 1990; Karp and Long, 1992). MHC unrestricted recognition by CTL has also been observed with antigens that contain repeated epitopes (Barnd, et al., 1989; jerome, et al., 1991). It is possible that multivalent antigens sometimes can interact with several MHC molecules and cross-link TCRs in an unconventional fashion, leading to MHC unrestricted T cell recognition. Since there is currently no evidence of unconventional T cell recognition of allogeneic MHC and mH antigens, it is most likely that allo-recognition follows a conventional scheme involving at least four players: the TCR, the accessory molecules CD4 or CD8, the MHC molecules and the pep tides bound to the MHC groove. The peptides presented by MHC class I and class II molecules derive from distinct protein sources. Although this distinction is not absolute, class I molecules prefer peptides derived from cytosolic proteins, whereas class II molecules preferentially present peptides derived from secreted or transmembrane cell surface proteins. The MHC class I presentation pathway (Monaco, 1992) starts with cytosolic protein breakdown by the high molecular weight proteasomes consisting of multiple subunits, two of which are encoded by genes in the MHC region. Although there is currently no experimental evidence that other proteases generate peptides for MHC class I presentation, it is most likely that some proteasome independent pathway for peptide production does exists. Peptides produced in the cytosol are then transported into the lumen of the endoplasmic reticulum (ER) by two proteins called TAP1 and TAP2 (Deverson, et al., 1990; Monaco, et al., 1990; Trowsdale, et al., 1990), which associate to form functional peptide transporter complexes. Inside the ER, peptides bind to newly synthesised MHC class I molecules which leads to the formation of stable MHC/peptide complexes which then travel to the cell surface. In contrast, MHC class II molecules bind peptides in a recently identified class II compartment (Amigorena, et al., 1994) , but not in the ER. In the ER, newly synthesised MHC class II molecules bind to invariant chain which prevents binding of ER resident peptides, and which directs migration of class II molecules to the class 11 compartment. Here, invariant chain is cleaved and the MHC binding groove becomes accessible for peptide binding. The class II compartment also contains the recently identified OMA and 0MB molecules (Sanderson, et al., 1994), which are required for antigen presentation via the class II pathway. Several functions for the OM molecules have been proposed including removal of the invariant chain derived clip peptides, delivery of pep tides which can bind to class II, and induction of conformational changes in class II molecules that allow peptide binding. The peptides which are available for class II binding are primarily prod uced by proteases in the endo-lysosomal compartment. Since this compartment preferentially contains endocytosed proteins derived from the cell surface or from the extracellular environment, most of the MHC class II presented peptides originate from these proteins. In general, the pathways which produce MHC class 1 and class II presented peptides cannot discriminate between self and non-self proteins, although interferon-gamma induced up-regulation of processing molecules during viral infections may lead to preferential presentation of viral pep tides in infected cells. Under physiologic conditions, MHC class I and class II molecules are constitutively loaded with pep tides derived from self proteins. Therefore, MIIC molecules expressed by normal cells are not homogeneous, but contain large numbers of distinct peptides. For the immunobiology of tissue transplantation it is important that the T cell compartment has been rendered tolerant to complexes consisting of self MHC molecules containing self peptides. In a transplant situation immune responses can be triggered by donor MHC molecules containing donor peptides, and by syngeneic MHC molecules displaying donor peptides. Complexes of donor MHC plus donor peptides are encountered in MHC mismatched situations and trigger immune responses by allo-MHC-specific T lymphocytes. Syngeneic MHC plus donor peptides are encountered in MHC matched situations and trigger immune responses against mH antigens.


Immunobiology of allo-MHC antigens:

MHC molecules are encoded by genes of the MHC cluster which maps to chromosome 6 in man and chromosome 17 in mouse. MHC class I and class II genes are extremely polymorphic, and the proteins encoded by these genes stimulate strong immune responses by T lymphocytes in MHC mismatched tissue transplantation. Allogeneic MHC class I and class II antigens are extremely immunogenic as determined by in vivo and in vitro assays. In vivo models in mice show that allogeneic MHC molecules can cause rapid skin graft rejection within 1014 days after transplantation. In vitro, T cell responses against alloMHC antigens are readily detected in naive T cell populations, and limiting dilution experiments have shown T cell precursor frequencies as high as 1/100 to 1/1000. Two models have been proposed to explain the strong immunogenicity of allogeneic MHC antigens. The high ligand density model (Bevan, 1984) postulates that allogeneic class I and class II antigens stimulate strong T cell responses because they are expressed at much higher levels than 'conventional' antigens. The model assumes that high ligand density not only stimulates high affinity T lymphocytes but also T cells expressing low affinity TCRs, while in 'conventional' immune responses against antigens which are presented at low density, high affinity T lymphocytes are stimulated exclusively. Thus, in the high ligand density model the high T cell precursor frequencies against allogeneic MHC molecules are explained by the recruitment of T cells expressing high as well as low affinity TCRs. Alternatively, it is possible that T cell responses against allogeneic MHC molecules are directed against a large number of distinct peptides bound to the groove of MHC molecules. Since the T cell compartment is tolerant only to complexes composed of self MHC plus self peptide, it is conceivable that all complexes composed of non-self MHC plus non-self peptide might stimulate T cell responses. Therefore, in the multiple ligand model (Matzinger and Bevan, 1977) it is postulated that allospecific T cell responses are directed against a large number of distinct MHC/peptide complexes expressed on the surface of allogeneic cells. Recent studies have estimated that up to 10 000 distinct peptides derived from cellular proteins may be displayed on the cell surface by MHC class I molecules (Hunt, et al., 1992). This observation is consistent with the suggestion that a large number of distinct MHC/peptide complexes are involved in T cell responses against allogeneic MHC molecules, which might account for the high precursor frequency.


Immunobiology of mH antigens:

Although loci encoding mH antigens have been identified and mapped to many different chromosomes, respective genes have not yet been isolated in man. Based on classical genetic studies of recombinant inbred mice, estimates of the number of mH antigens range from fifty to several hundred (Bailey and Mobraaten, 1969). The maternally transmitted antigen (Mta) and the myxovirus resistance protein (Mx) are the only mH antigens that have been molecularly identified in mouse. Mta is derived from the mitochondrially encoded ND1 protein (Loveland, et al., 1990). Like in prokaryotes, proteins synthesised in mitochondria contain a N-formylmethionine at their amino-terminus. Peptides corresponding to the first 12 amino acids of ND1 are presented by non-classical H-2M3 class I molecules to CTL. These non-classical class I molecules have apparently evolved to present bacterial and mitochondrial antigens, since they preferentially bind peptides with formylated N-termini. The only non-formylated N-terminal amino acid that showed some binding to H-2M3 was glycine. The second molecularly identified murine mH antigen is encoded by a gene (Mx) that determines susceptibility or resistance to infection by myxovirus (Speiser, et al., 1990). The Mx gene is not present in all mouse strains and it was shown that Mx negative mice rejected skin grafts from Mx positive donors, which correlated with the detection of Mx-specific CTL responses in vitro. Skin transplantation experiments in mice have shown that graft rejection caused by mH antigens is slower than rejection caused by allo MHC antigens (Graff and Bailey, 1973). However, some mH antigens (e.g. the murine H-1 antigen) are nearly as immunogenic as MHC antigens as judged by the kinetics of graft rejection (Graff, et al., 1966). Nevertheless, in general immune responses to mH antigens are weaker than responses to allo-MHC antigens. In vitro, T cell responses against mH antigens are usually not detectable in naive T cell populations, and detection of anti-mH responses requires prior in vivo immunisation. In mice, mH antigens are defined as genetic loci which can cause skin transplant rejection when donor and recipient express the same MHC molecules. Although the composition of mH loci has not yet been fully dissected, it is likely that several genes are present in one locus. Since mH antigens usually induce responses by Th lymphocytes and by CTL, it has been suggested that distinct genes encode for antigens that are recognised by these T lymphocyte populations (Roopenian, 1992). One gene may encode a protein that is processed via the class II pathway to generate peptides that are presented by MHC class II molecules, while another gene may encode a protein that is channelled into the class I processing pathway to produce pep tides that are presented by MHC class I molecules. Stimulation of both MHC class II restricted Th cells and MHC class I restricted CTL is probably required to cause graft rejection, while stimulation of only one T cell population may be insufficient. However, it does not necessarily follow that two distinct genes are required for stimulation of Th lymphocytes and CTL. For example, studies of T cell responses against human melanomas by Rosenberg's group have shown that both MHC class II and class I presen ted peptides were derived from tyrosinase (Topalian, et al., 1994). It is therefore possible that a single gene can encode a protein that contains Th as well as CTL epitopes. Consequently, not all mH loci may contain multiple genes, although this has been shown to be the case for the murine mH loci H-3, IH-4 and HY (Roopenian, 1992).


Immunodominance:

Immunodominance has been documented most extensively by E. Sercas and colleagues (Sercarz, et al., 1993), who studied Th responses to hen egg lysozyme (HEL). Immunisation of mice with HEL stimulates strong Th responses against an immunodominant peptide, and responses to subdominant peptides are not detected. However, responses to these subdominant epitopes are readily detectable when immunodominant epitopes are absent during immunisation. Although immunodominance has not been demonstrated as rigorously for CTL. responses, there is considerable evidence that it also applies. Firstly, CTl. responses in mice infected with influenza virus, vesicular stomatitis virus, Sendai virus and lymphocytic choriomeningitis virus are directed against one or few viral peptides although these viruses express several proteins. Similarly, CTL from H-2b mice immunised with ovalbumin always recognise an immunodominant 8mer peptide presented by H-2Kb class I molecules. A recent study has shown that a subdominant peptide can be detected when mice were immunised with high doses of ovalbumin. Although the mechanisms underlying immunodominance are not fully understood, the phenomena is most likely of importance for T cell responses to allogeneic MHC and mH antigens.


Immunodominance among mH antigens

The MHC matched mouse strains C57Bl./6 and BAl.B/B differ by more than 29 mH loci (Bailey and Mobraaten, 1969). Early work by Wettenstein and colleagues has shown that C57Bl./6 mice immunised with BAl.B/B spleen cells generated CTL. responses against only two immunodominant mH loci, while the majority of mH loci were immunologically silent (Wettstein, 1986 ). Silence of these loci was not due to lack of immunogenicity but rather to the overriding effects of immunodominant loci. Omission of dominant mH antigens on the immunising cells revealed CTL. responses against subdominant mH antigens that were previously undetectable (Wettstein, 1986 ). The mechanisms of immunodominance among mH loci are currently not well understood. One possibility is that immunodominant loci contain multiple genes providing several peptide epitopes that are presented by MHC class II and class I molecules, which might lead to strong stimulation of Th cells and CTL To investigate this possibility , we have analysed pep tides that are involved in CTL. responses of BAl.B/B mice against mH mismatched C57Bl./6 stimulators (Yin, et al., 1993). Peptide purification by HPLC showed that only two HPLC fractions contained CTL recognised pep tides. One of the fractions contained peptides that were presented by H-2Kb class 1 molecules, while the other HPLC fraction contained H-2Db presented peptides. Further HPLC purification did not provide any evidence that more than one peptide was present in each of these HPLC fractions. The Kb presented peptides were found to be derived from the minor H-1 locus, while the Ob presented peptides were derived from another, unidentified mH locus. Together, these experiments provide evidence that immunodominance of these two mH loci was not caused by multiple CTL stimulating peptides. Only one, or a small number of closely related peptides, which are strongly immunogenic and account for the immunodominance of the analysed mH loci.


Peptide dominance in anti-allo MHC T cell responses?

There is now considerable evidence that a large proportion of CTL specific for allogeneic MHC class 1 molecules is peptide dependent. Immune responses to allogeneic MHC class II molecules have not been studied as extensively as anti-class I responses, but there is no reason to believe that there will be any fundamental difference. The first formal demonstration of peptide dependent allo-recognition came from experiments with human cells transfected with murine class I molecules. The human transfectants were not recognised by murine allo-specific CTL, but recognition was restored when the human cells were coated with peptides prepared by cyanobromide cleavage of proteins isolated from the cytosol of mouse cells (Heath, et al., 1989). We have used limiting dilution analysis to show that the majority of CTL clones raised against the allogeneic H-2Kb class I molecule were peptide dependent (Aosai, et al., 1991). Rotzschke et al. have documented that different CTL clones specific for a given allogeneic class I molecule recognised distinct HPLC purified peptide fractions (Rotzschke, et al., 1991). Although these studies clearly indicate peptide dependent allorecognition, it is not clear whether all peptides presented by allogeneic class I molecules are recognised by CTL, or whether a limited number of strongly immunogenic peptides dominate CTL responses. As mentioned above, immunodominance has been observed in many different situations, so that it would be surprising if peptide recognition in the context of allogeneic MHC molecules would be an exception. Recent findings support the idea that immunodominant pep tides are involved in CTL responses against non-self MHC class I molecules. For example, Henderson et al. have shown that five independently derived anti-HLA-A2 murine CTL clones recognised the same A2 presented peptide (Henderson, et al., 1993). Also, Udaka et al. have sequenced a peptide (p2Ca) that was recognised by an allo-CTL clone specific for H-2Ld (Udaka, et al., 1992). In an elegant study J. Connolly has analysed CTL responses of Ld negative dm2 mice stimulated with Ld expressing BALB/c cells (Connolly, 1994). In limiting dilution assays, approximately half of anti-Ld CTL precursors showed specificity for the p2Ca peptide. Interestingly, immunodominance of the p2Ca peptide was only observed in responder mice that expressed the Vß8 gene segments of the TCR-ß chain, but not in Vß8 negative strains. The experiments by Connolly indicate that a single peptide can stimulate up to 50% of alloreactive CTL, and that the immunodominant peptides may sometimes preferentially stimulates CTL precursors expressing certain TCR variable region gene segments.


Downregulation of allo-responses with immunodominant peptides?

The observation that immune responses to allogeneic MHC molecules and mH antigens may be dominated by strongly immunogenic peptides, raises the question whether such peptides can be exploited to downregulate T cell responses. Tolerance induction with synthetic peptides has been demonstrated with transgenic mice expressing the LCMV glycoprotein (GP) in pancreatic islet cells (Aichele, et al., 1994). When these mice were repeatetly injected intraperitoneally with high doses of synthetic peptides corresponding to a CTL epitope in GP, they were subsequently resistant to T cell mediated induction of autoimmune diabetes. Oral antigen administration is another way of inducing antigen-specific unresponsiveness (Weiner, et al., 1994). It seems that the mechanism of unresponsiveness might depend upon the dose of orally given antigen (Friedman and Weiner, 1994). Low dose antigen appears to preferentially induce TGF-ß producing regulatory Th2 lymphocytes, which suppress responses by Th1 cells and CTL. In contrast, high dose antigen is less likely to induce regulatory T cells, and is more likely to cause deletion of antigen responsive T cells. A recent report by Sun et al. indicates that it may be possible to enhance induction of unresponsiveness by oral antigens (Sun, et al., 1994). Cholera toxin (CT) consists of a central subunit A surrounded by five B subunits. Intact CT is a potent immune-stimulating oral adjuvant leading to stimulation of immune responses against oral antigens rather then tolerance induction. Sun et al. showed that recombinant subunit B has the opposite effect, since it can enhances induction of oral tolerance against proteins coupled to the B subunit. Ongoing clinical trials in multiple sclerosis patients and rheumatoid arthritis patients may benefit from this observation, since it seems possible to improve the design of tolerance inducing oral vaccines. Peptides corresponding to sequences of MHC class I and class II molecules have been shown to inhibit T cell responses. One of the first studies showed that peptides derived from residues 98-113 of the alfa2 domain of HLA-A2 inhibited recognition by most A2-specific CTL (Perham, et al., 1987). These HLA-A2 peptides probably inhibited CTL recognition by binding to the TCR. Subsequently, it was shown that peptides corresponding to the HLA-B7 residues 75-84 could induce tolerance in a rat heart allograft model (Nisco, et al., 1994). A combination of HLA-B7 peptides together with low doses of cyclosporine A resulted in long-term heart allograft survival in most rats, while treatment with peptides alone or cyclosporine A alone had little effect. In this study, oral peptide feeding was as efficient as i. v. injection in inducing heart allograft survival. The mechanism by which HLA-B7 pep tides can suppress immune responses against rat MHC molecules is currently unclear. Recently, it has been described that these peptides bind to two T cell molecules of 72kDa and 74kDa, and that binding is associated with an increase in intracellular Ca2 + (Krensky and Clayberger, 1994). These observations suggest that the immunesuppressing effect of these HLA-derived pep tides may not be specific for immune responses against allo-MHC molecules. A similar lack of specificity is encountered with peptides corresponding to the conserved binding sites for CD8 and CD4 on MHC class I and class II molecules, respectively. In vitro, these peptides can inhibit differentiation of human CTL precursors and the mixed lymphocyte reaction of freshly isolated peripheral blood lymphocytes (Clayberger, et al., 1994). Sayegh et al. have described a HLA-DR2 derived peptide which inhibits proliferation in the mixed lymphocyte reaction, but not proliferation induced by mitogens or mumps (Krensky and Clayberger, 1994) .This peptide corresponds to residues 182-94 of the transmembrane region of the DR alfa2 chain and shows binding to several HLA-DR molecules. Therefore, this peptide is probably involved in indirect allo-recognition, where peptides from allogeneic MHC molecules are presented to T cells in the context of self MHC products. Evidence that indirect allo-recognition can result in tissue transplant rejection comes from experiments with MHC class II knockout mice (Auchincloss, et al., 1993). Normal mice receiving MHC class II negative skin transplants mounted a CD4 T cell responses which led to graft rejection. These CD4 T cells recognised donor derived peptides presented by host MHC class II molecules (Lee, et al., 1994). In a rat model it was shown that indirect allo-recognition may be susceptible to peptide-specific tolerance induction. Peptides derived from rat MHC class II ß chains RT1.Bu and RT1.Du, which are probably involved in indirect allorecognition, have been given orally to LEW rats which resulted in the down-regulation of allo-responses in vivo and in vitro (Sayegh, et al., 1992).
The described results obtained with MHC derived peptides are encouraging and provide evidence that tolerance induction with pep tides is feasible. The identification of strongly immunogenic pep tides which dominate T cell responses against allogeneic MHC molecules and mH antigens may lead to their exploitation for tolerance induction. Synthetic peptides corresponding to mH antigens can be used directly for tolerance induction, since they can be presented by host MHC molecules. In contrast, peptides derived from the groove of allogeneic MHC molecules cannot be used directly, since they will not be presented properly by host MHC molecules. Therefore, complexes of donor MHC molecules plus donor peptide will be required for induction of donor-specific tolerance. Such complexes can be produced using recombinant MHC molecules generated in bacteria or yeast, which can be refolded with synthetic peptides to obtain homogeneous MHC/peptide complexes. An easier way to enrich for MHC molecules containing defined synthetic peptides is to acid treat intact cells in the presence of peptides, which leads to partial MHC denaturation, dissociation of endogenous peptides and binding of synthetic peptides. It has been shown that professional antigen presenting cells treated in this way were much more efficient in stimulating peptide-specific T cell responses than untreated A PC (Langlade-Demoyen, et al., 1994). Similarly, it is conceivable that peptide loading of non professional antigen presenting cells with immunodominant peptides may greatly enhance their ability to induce peptide-specific tolerance.


Conclusion:

Although immuno-suppressive drugs such as cyclosporine A can mediate acceptance of histoincompatible tissue transplants, their longterm use creates considerable problems. Induction of antigen-specific tolerance would be preferable over generalised immune-suppression. Animal models have provided evidence that synthetic peptides derived from sequences of MHC class I and class II molecules can prolong graft survival. The mode of action and the specificity of some of these MHC derived pep tides is not yet clear. It is likely, that immunodominant peptides present in the peptide binding groove of MHC molecules can be identified and used for peptide-specific downregulation of T cell responses against allogeneic MHC molecules and against mH antigens. In future, an approach using sub-therapeutic doses of immuno-suppressive drugs in combination with tolerance inducing peptides derived from MHC sequences and from MHC bound peptides, may lead to long-term graft survival with minimal side effects.


References:

Aichele, P.. Kyburz, D., Ohashi, P.S., Odermatt, B., Zinkernagel, R.M., Hengartner, H. and Pircher, H. ( 1994) Peptide-induced T -cell tolerance to prevent autoimmune diabetes in a transgenic mouse model. Proc Natl Acad Sci USA. 91: 444-8.

Amigorena, S., Drake, l.R., Webster, P. and Mellman, I. (1994) Transient accumulation of new class II MHC molecules in a novel endocytic compartment in B lymphocytes [see comments]. Nature. 369: 113-20.

Aosai, F., Ohlen, C., Ljunggren, H.C., Hoglund, P., Franksson, L., Ploegh, H., Townsend, A., Karre, K. and Stauss, H.l. (1991) Different types of allospecific CTL clones identified by their ability to recognize peptide loading-defective target cells. Eurjlmmunol. 21: 2767-74.

Auchincloss, H.l., Lee, R., Shea, S., Markowitz, l.S., Crusby, M.l. and Climcher, L.H. ( 1993) The role of "indirect" recognition in initiating rejection of skin grafts from major histocompatibility complex class 11deficient mice. Proc NatlAcad Sci USA. 90: 3373-7.

Bailey, D. and Mobraaten, L. ( 1969) Estimates of the number of loci contributing to the histoincompatiibility between C57BL/6 and BALB/c strains of mice. 7: 394-400.

Barnd, D.L., Lan, M.S., Metzgar, R.S. and Finn, 0.1. ( 1989) Specific, major histocompatibility complex-unrestricted recognition of tumor-associated mucins by human cytotoxic T cells. Proc Natl Acad Sci U S A. 86: 715963.

Bevan, M.l. (1984) Immunol. Today. 5: 128.

Bjorkman, P.l., Saper, M.A., Samraoui, B., Bennett, W.S., Strominger, l.L. and Wiley, D.C. (1987a) The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature. 329: 512-8.

Bjorkman, P.l., Saper, M.A., Samraoui, B., Bennett, W.S., Strominger, l.L. and Wiley, D.C. ( 1987b) Structure of the human class I histocompatibility antigen, HLA-A2. Nature. 329: 506-12.

Brown, l.H., lardetzky, T.S., Corga, l.C., Stern, L.]., Urban, R.C., Strominger, l.L. and Wiley, D.C. (1993) Three-dimensional structure of the human class II histocompatibility antigen HLA-DR1 [see comments]. Nature. 364: 33-9.

Choi, Y.W., Herman, A., DiCiusto, D., Wade, T., Marrack, P. and Kappler, l. ( 1990) Residues of the variable region of the T -cell-receptor beta-chain that interact with S. aureus toxin superantigens. Nature. 346: 471-3.

Clayberger, C., Lyu, S.C., DeKruyff, R., Parham, P. and Krensky, A.M. ( 1994) Peptides corresponding to the CD8 and CD4 binding domains of HLA molecules block T lymphocyte immune responses in vitro. j Immunol. 153: 946-51.

Connolly, l.M. (1994) The peptide p2Ca is immunodominant in allorecognition of Ld by ß-chain Vß8+ but not Vß8- strains. Proc. Natl. Acad. Sci. 91: 11482-486.

Deverson, E.V., Cow, I.R., Coadwell, W.J., Monaco, l.J., Butcher, C.W. and Howard, l.C. ( 1990) MHC class II region encoding proteins related to the multidrug resistance family of transmembrane transporters [see comments]. Nature. 348: 738-41.

Friedman, A. and Weiner, H.l. (1994) Induction of anergy or active suppression following oral tolerance is determined by antigen dosage. Proc Natl Acad Sci U S A. 91: 6688-92.

Graff, R. and Bailey, D. (1973) The non-H-2 histocompatibility loci and their antigens. Transplant Rev. 15: 26-49.

Graff, R., Hildemann, W. and Snell, G. (1966) Histocompatibility genes of mice. VI. allografts in mice congenic at various non-H-2 histocompatibility loci. 4: 425-437.

Heath, W.R., Hurd, M.E., Carbone, F.R. and Sherman, l.A. ( 1989) Peptidedependent recognition of H-2Kb by alloreactive cytotoxic T lymphocytes. Nature. 341: 749-752.

Henderson, R.A., Cox, A.l., Sakaguchi, K., Appella, E., Shabanowitz, J., Hunt, D.F. and Engelhard, V.H. (1993) Direct identification of an endogenous peptide recognized by multiple HlA-A2.1-specific cytotoxic T cells. PNAS. 90: 10275-10279.

Hunt, D.F., Henderson, R.A., Shabanowitz, J., Sakaguchi, K., Michel, H., Sevilir, N., Cox, A.l., Appella, E. and Engelhard, V .H. ( 1992 ) Characterization of peptides bound to class I MHC Molecule HlA-A2.1 by mass spectrophotometry. Science. 255: 1261-1263.

Jerome, K.R., Barnd, D.l., Bendt, K.M., Boyer, C.M., Taylor, P.J., McKenzie, I.F., Bast, R.J. and Finn, O.J. (1991) Cytotoxic T-lymphocytes derived from patients with breast adenocarcinoma recognize an epitope present on the protein core of a mucin molecule preferentially expressed by malignan t cells. Cancer Res. 51: 2908-16.

Karp, D.R. and long, E.O. (1992) Identification of HlA-DR1ß chain residues critical for binding staphylococcal enterotoxins A and E. ] Exp Med. 175: 415-424.

Krensky, A.M. and Clayberger, C. ( 1994) The induction of tolerance to alloantigens using HlA-based synthetic peptides. Curr. Opin. Immunol. 6: 791-796.

langlade-Demoyen, P., levraud, J.P., Kourilisky, P. and Abastado, cJ.P. ( 1994) Primary cytotoxic T lymphocyte induction using peptidestripped autologous cells. Intern. Immunol. 6: 1759-66.

lee, R.S., Grusby, M.J., Glimcher, l.H., Winn, H.J. and Auchincloss, H.J. ( 1994) Indirect recognition by helper cells can induce donor-specific cytotoxic T lymphocytes in vivo. ]Exp Med. 179: 865-72.

Loveland, B., Wang, C.R., Yonekawa, H., Hermel, E. and lindahl, K.F. ( 1990) Mlatemally transmitted histocompatibility antigen of mice: a hydrophobic peptide of a mitochondrially encoded protein. Cell. 60: 97180.

Matzinger, P. and Bevan, M.J. (1977) Cell. Immunol. 29: 1. Monaco, J.J. ( 1992) A molecular model of MHC class-l-restricted antigen processing. Immunol Today. 13: 173-9.

Monaco, J.J., Cho, S. and Attaya, MI. ( 1990) Transport protein genes in the murine MHC: possible implications for antigen processing. Science. 250: 1723-6.

Nisco, S., Vriens, P., Hoyt, G., Lyu, S.C., Farfan, F., Pouletty, P., Krensky, A.M. and Clayberger, C. ( 1994) Induction of allograft tolerance in rats by an HLA class-I-derived peptide and cyclosporine A. J Immunol. 152: 3786-92.

Perham, P., Clayberger, C., lorn, S.L., Ludwig, D.S., Schoolnik, G.K. and Krensky, A.M. ( 1987) Inhibition of alloreactive cytotoxic T lymphocytes by pep tides from the alfa2 domain of HLA-A2. Nature. 325: 625-628.

Roopenian, D.C. ( 1992) What are minor histocompatibility loci? Anew look at an old question. Immunol Today. 13: 7-10.

Rotzschke, 0., Falk, K., Faath, S. and Rammensee, H.G. (1991) On the nature of peptides involved in T cell alloreactivity. J Exp Med. 174: 1059-71.

Salter, R.D., Norment, A.M., Chen, B.P., Clayberger, C., Krensky, A.M., Littman, D.R. and Parham, P. (1989) Polymorphism in the alpha 3 domain of HLA-A molecules affects binding to CD8. Nature. 338: 345- 7.

Sanderson, F., Kleijmeer, M.1., Kelly, A., Verwoerd, D., Tulp, A., Neefjes, 1.1., Geuze, H.1. and Trwsdale, 1. (1994) Accumulation of HLA-DM, a regulator of antigen presentation, in MHC class II compartments. Science. 266: 1566-1569.

Sayegh, M.H., Khoury, S.1., Hancock, W.W., Weiner, H.L. and Carpenter, C.B. (1992) Induction of immunity and oral tolerance with polymorphic class II major histocompatibility complex allopeptides in the rat. Proc Natl Acad Sci U S A. 89: 7762-6.

Sercarz, E.E., Lehmann, P.V., Ametani, A., Benichou, G., Miller, A. and Moudgil, K. ( 1993) Dominance and crypticity of T cell antigenic determinants. Annu Rev Immunol. 11: 729-66.

Speiser, D.E., lurcher, T., Ramseier, H., Hengartner, H., Staeheli, P., HaIler, 0. and linkemagel, R.M. ( 1990) Nuclear myxovirus-resistance protein Mx is a minor histocompatibility antigen. Proc Natl Acad Sci U S A. 87: 2021-5.

Sun, 1.B., Holmgren, 1. and Czerkinsky, C. ( 1994) Cholera toxin B subunit: An efficient transmucosal carrier-delivery system for induction of perigheral immunological tolerance. Proc. Natl. Acad. Sci. USA. 91: 10795-799.

Topalian, S.L., Rivoltini, L., Mancini, M., Markus, N.R., Robbins, P.F., Kawakami, Y. and Rosenberg, S.A. ( 1994) Human CD4+ T cells specifically recognise a shared melanoma-associated antigen encoded by the tyrosinase gene. Proc. Natl. Acad. Sci. USA. 91: 9461-9465.

Trowsdale, 1., Hanson, I., Mockridge, I., Beck, S., Townsend, A. and Kelly, A. ( 1990) Sequences encoded in the class II region of the MHC related to the 'ABC' superfamily of transporters [see comments]. Nature. 348: 7414.

Udaka, K., Tsomides, T.1. and Eisen, H.N. (1992) A naturally occurring peptide recognized by alloreactive CD8+ cytotoxic T lymphocytes in association with a class I MHC protein. Cell. 69: 989-98.

Weiner, H.L., Friedman, A., Miller, A., Khoury, S.1., al, S.A., Santos, L., Sayegh, M., Nussenblatt, R.B., Trentham, D.E. and Hafler, D.A. (1994) Oral tolerance: immunologic mechanisms and treatment of animal and human organ-specific autoimmune diseases by oral administration of autoantigens. Annu Revlmmunol. 12: 809-37.

Wettstein, P.1. (1986) Immunodominance in the T-cell response to multiple non-H-2 histocompatibility antigens. II. Observation of a hierarchy among dominant antigens. Immunogenetics. 24: 24-31.

Yin, L., Poirier, G., Neth, 0., Hsuan, 1.1., Totty, N.F. and Stauss, H.1. (1993) Few peptides dominate CTL responses to single and multiple minor histocompatibility antigens. Int. Immunol. 5: 1003-1009.