Tumor defense by murine cytotoxic T cells specific for peptide bound to nonclassical MHC class I

Cytotoxic T Cells (CTLs) can exhibit considerable antitumor activity. Thus far, the characterized tumor peptide antigens recognized by CTLs are all presented by classical MHC class Ia molecules [human lymphocyte antigen A (HLA-A), HLA-B, and HLA-C in humans and H-2K, H-2D, and H-2L in mice]. Here we show that CTLs recognized peptides presented by nonclassical MHC class Ib molecule Qa-1b expressed by tumor cells. These CTLs conferred in vivo protection by delaying the growth of Qa-1b-expressing B78H1 melanoma cells pulsed with Qa-1b-binding peptides Cw4L or B35L and injected s.c. in C57BL/6 mice. A hierarchy of the peptides was found with regard to their ability to trigger CTLs; Cw4L stimulated a strong CTL response. The closely related and cross-reactive peptide B35L induced a weaker CTL response but was still efficient in sensitizing the target cells. Finally, Qa-1b-expressing melanoma cells without exogenous peptides were not immunogenic but possibly expressed endogenous cross-reactive antigenic peptides. The data are compatible with earlier findings that CTL activation requires relatively strong peptide antigens, whereas subsequent effector functions are also mediated by weak peptide analogues. In conclusion, CTLs mediated tumor immunity through the recognition of peptides presented by nonclassical MHC class Ib molecules. The identification of similar CTLs in humans may facilitate the vaccination of cancer patients because MHC class Ib/peptide complexes are much less polymorphic than MHC class Ia/peptide complexes.     

Mouse CD94/NKG2A is a natural killer cell receptor for the nonclassical major histocompatibility complex (MHC) class I molecule Qa-1(b)

Natural killer (NK) cells preferentially lyse targets that express reduced levels of major histocompatibility complex (MHC) class I proteins. To date, the only known mouse NK receptors for MHC class I belong to the Ly49 family of C-type lectin homodimers. Here, we report the cloning of mouse NKG2A, and demonstrate it forms an additional and distinct class I receptor, a CD94/NKG2A heterodimer. Using soluble tetramers of the nonclassical class I molecule Qa-1(b), we provide direct evidence that CD94/NKG2A recognizes Qa-1(b). We further demonstrate that NK recognition of Qa-1(b) results in the inhibition of target cell lysis. Inhibition appears to depend on the presence of Qdm, a Qa-1(b)-binding peptide derived from the signal sequences of some classical class I molecules. Mouse NKG2A maps adjacent to CD94 in the heart of the NK complex on mouse chromosome six, one of a small cluster of NKG2-like genes. Our findings suggest that mouse NK cells, like their human counterparts, use multiple mechanisms to survey class I expression on target cells.     

The effect of the proteasome inhibitor lactacystin on the presentation of transporter associated with antigen processing (TAP)-dependent and TAP-independent peptide epitopes by class I molecules

Cells were treated with two proteolytic inhibitors, N-acetyl-leucyl-leucyl-norleucinal and lactacystin, the latter reported to be a specific inhibitor for the proteasome. Both inhibitors retarded the maturation of endo-H-resistant forms of murine and human class I molecules from their endo-H-sensitive precursors in cell lines with functional TAP proteins. HLA-A2 maturation readily occurs in TAP-deficient T2 cells, and it has been shown that the peptides associated with A2 are derived from the leader segment of proteins in the secretory pathway. This maturation is inhibited by N-acetyl-leucyl-leucyl-norleucinal but not lactacystin, indicating that the proteasome is not required for the generation of HLA-A2 binding peptides in these cells. The murine class Ib molecule Qa-1b presents a leader peptide derived from D-end class I molecules to alloreactive CTL. Since this presentation is dependent on the expression of TAP proteins, we determined if this requirement reflects a need for the proteasome to process this peptide. We found that lactacystin did not inhibit the maturation of endo-H-resistant forms of Qa-1b that are dependent on this leader peptide for its maturation, nor did it inhibit the expression of this peptide-Qa-1b complex in a functional assay. Thus, unlike conventional cytosolic peptides, leader peptides (regardless of whether they are dependent on TAP for their presentation) do not require the proteasome for processing.     

Isolation and characterization of murine clonogenic osteoclast progenitors by cell surface phenotype analysis

MHC class I molecules bind short peptides for presentation to CD8+ T cells. The determination of the three-dimensional structure of various MHC class I complexes has revealed that both ends of the peptide binding site are composed of polar residues conserved among all human and murine MHC class I sequences, which act to lock the ends of the peptide into the groove. In the rat, however, differences in these important residues occur, suggesting the possibility that certain rat MHC class I molecules may be able to bind and present longer peptides. Here we have studied the peptide length preferences of two rat MHC class Ia molecules expressed in the TAP2-deficient mouse cell line RMA-S: RT1-A1c, which carries unusual key residues at both ends of the groove, and RT1.Aa which carries the canonical residues. Temperature-dependent peptide stabilization assays were performed using synthetic random peptide libraries of different lengths (7-15 amino acids) and successful stabilization was determined by FACS analysis. Results for two naturally expressed mouse MHC class I molecules revealed different length preferences (H2-Kb, 8-13-mer and H2-Db, 9-15-mer peptides). The rat MHC class Ia molecule, RT1-Aa, revealed a preference for 9-15-mer peptides, whereas RT1-A1c showed a more stringent preference for 9-12-mer peptides, thereby ruling out the hypothesis that unusual residues in rat MHC molecules allow binding of longer peptides.     

Production, crystallization, and preliminary X-ray analysis of the human MHC class Ib molecule HLA-E

HLA-E is the first human class Ib major histocompatibility complex molecule to be crystallized. HLA-E is highly conserved and almost nonpolymorphic, and has recently been shown to be the first specialized ligand for natural killer cell receptors. In functional studies, HLA-E is unlike the class Ia MHC molecules in having tightly restricted peptide binding specificity. HLA-E binds a limited set of almost identical leader sequence peptides derived from class Ia molecules and presents these at the cell surface for recognition by natural killer cell receptors. We now show that the extracellular region of HLA-E forms a stable complex with beta2 microglobulin and can be refolded around synthetic peptide. Crystals of this complex formed slowly over four to six months in the presence of ammonium sulphate. The crystals diffract to 2.85 A with space group P3(1)21 and unit cell dimensions a = 182.2 A, b = 182.2 A, c = 88.4 A.     

The critical role of a solvent-exposed residue of an MHC class I-restricted peptide in MHC-peptide binding

The immunodominant ovalbumin257-264 (OVA-8, SIINFEKL) and herpes simplex virus gB496-503 (HSV-8, SSIEFARL) peptides share 50% amino acid identity (residues P1, P3, P5 and P8) and bind with comparable efficacy to the murine MHC-encoded class I molecule H-2Kb. However, these two peptides bind differently to H-2Kbm8, a natural H-2Kb variant with a substitution in four amino acids on the floor of the peptide-binding site; HSV-8 binds with high and OVA-8 with a relatively low efficacy. To investigate which of the non-homologous peptide residues were responsible for this differential binding, we used substituted peptide variants and the class I thermodynamic stabilization assay. Variation at the solvent-exposed peptide residues P6 and P7 did not appreciably influence binding. By contrast, variation at the buried P2 and, surprisingly, at the solvent-exposed P4 residue was found to be important. Transplantation of the HSV-8 P2 or P4 residues onto the OVA-8 backbone created variant peptides O2S (P2I-->S) and O4E (P4N-->E) that bound considerably better to H-2Kbm8 than OVA-8. Furthermore, the double-substituted peptide, O2S4E, bound even better, revealing a cooperative effect of the two residues. The reciprocally substituted peptides H2I and H4N, generated by grafting the OVA-8 P2 and P4 residues onto the HSV-8 backbone respectively, bound to H-2Kbm8 slightly worse than HSV-8 but the double-substituted peptide H2I4N bound as poorly as OVA-8. Effects exerted by the P4 residue, which is solvent accessible and therefore available for the TCR contact, demonstrated that exposed peptide residues can, in certain situations, influence not only the TCR contact but also MHC-peptide binding.     

Diminished class II-associated Ii peptide binding to the juvenile dermatomyositis HLA-DQ alpha 1*0501/DQ beta 1*0301 molecule

HLA class II molecules bind and present peptide Ags to T cells, binding specific sets of peptides due to polymorphism in the peptide binding groove. Class II proteins associate with the invariant chain (Ii chain) and its derived class II-associated Ii peptide (CLIP). Ii chain association is important for normal trafficking of class II proteins to the peptide loading vesicles and for blocking premature access of peptides to HLA class II molecules during maturation. We have previously shown that juvenile dermatomyositis is associated with the HLA-DQA1*0501 allele. There is limited information available about the interaction of any DQ molecule with the Ii chain and little information about binding of individual peptides to HLA-DQalpha1*0501/DQbeta1*0301. We sequenced peptides eluted from the juvenile dermatomyositis-associated class II allele HLA-DQalpha1*0501/DQbeta1*0301. Surprisingly, we found no Ii chain or CLIP. Further examination of peptide binding to the HLA-DQalpha1*0501/DQbeta1*0301 molecule demonstrated poor CLIP binding. However, newly synthesized HLA-DQalpha1*0501/DQbeta1*0301 molecules do associate with intact Ii chain. Molecular modeling suggests that CLIP binds differently to HLA-DQalpha1*0501/DQbeta1*0301 than to DR molecules. The lack of CLIP association suggests that HLA-DQalpha1*0501/DQbeta1*0301 has access to peptides earlier in the processing pathway and so might encounter novel peptides that induce autoimmunity.     

A nonamer peptide derived from Listeria monocytogenes metalloprotease is presented to cytolytic T lymphocytes

Listeria monocytogenes is an intracellular bacterium that secretes proteins into the cytosol of infected macrophages. Major histocompatibility complex (MHC) class I molecules bind peptides that are generated by the degradation of bacterial proteins and present them to cytolytic T lymphocytes (CTL). In this study we have investigated CTL responses in L. monocytogenes-immunized mice to peptides that (i) derive from the L. monocytogenes proteins phosphatidylinositol-specific phospholipase C, lecithinase (most active on phosphatidylcholine), metalloprotease (Mpl), PrfA, and the ORF-A product and (ii) conform to the binding motif of the H2-Kd MHC class I molecule. We identified a nonamer peptide, Mpl 84-92, that is presented to L. monocytogenes-specific CTL by H2-Kd MHC class I molecules. Unlike other motif-conforming peptides derived from the secreted Mpl of L. monocytogenes, Mpl 84-92 is bound with high affinity by H2-Kd. Mpl 84-92 is the fourth L. monocytogenes-derived peptide found to be presented to CTL by the H2-Kd molecule during infection and demonstrates the importance of high-affinity interactions between antigenic peptides and MHC class I molecules for CTL priming.     

The P9 pocket of HLA-DQ2 (non-Aspbeta57) has no particular preference for negatively charged anchor residues found in other type 1 diabetes-predisposing non-Aspbeta57 MHC class II molecules

Susceptibility and resistance to type 1 diabetes are associated with MHC class II alleles that carry non-Asp and Asp at residue 57 of their beta chain respectively. The effect of Asp or non-Aspbeta57 may relate to a differential ability of distinct class II molecules to bind specific immuno-pathogenic peptides. Recent studies in man and mouse have revealed that some type 1 diabetes-predisposing non-Aspbeta57 class II molecules (i.e. DQ8, DR4Dw15 and I-Ag7) preferentially bind peptides with a negatively charged anchor residue at P9. It has been suggested that this is a common feature of type 1 diabetes-predisposing class II molecules. The molecular explanation for such a phenomenon could be that class II beta chains with Aspbeta57 form a salt bridge between Aspbeta57 and a conserved Arg of the a chain, whereas in non-Aspbeta57 molecules the Arg is unopposed and free to interact with negatively charged P9 peptide anchor residues. We have investigated the specificity of the P9 pocket of the type 1 diabetes-associated DQ2 molecule and in particular examined for charge effects at this anchor position. Different approaches were undertaken. We analyzed binding of a high-affinity binding ligand and P9-substituted variants of this peptide, and we analyzed the binding of a set of synthetic random peptide libraries. The binding analyses were performed with wild-type DQ2 and a mutated DQ2 with Ala at beta57 substituted with Asp. Our results indicate that the wild-type DQ2 (non-Aspbeta57) prefers large hydrophobic residues at P9 and that there is no particular preference for binding peptides with negatively charged residues at this position. The specificity of the P9 pocket in the mutated DQ molecule is altered, indicating that the beta57 residue contributes to determining the specificity of the P9 pocket. Our data do not lend support to the hypothesis that all non-Asp beta57 class II molecules predispose to development of disease by binding peptides with negatively charged P9 anchor residues.     

H-2M3a violates the paradigm for major histocompatibility complex class I peptide binding

The major histocompatibility (MHC) class I-b molecule H-2M3a binds and presents N-formylated peptides to cytotoxic T lymphocytes. This requirement potentially places severe constraints on the number of peptides that M3a can present to the immune system. Consistent with this idea, the M3a-Ld MHC class I chimera is expressed at very low levels on the cell surface, but can be induced significantly by the addition of specific peptides at 27 degrees C. Using this assay, we show that M3a binds many very short N-formyl peptides, including N-formyl chemotactic peptides and canonical octapeptides. This observation is in sharp contrast to the paradigmatic size range of peptides of 8-10 amino acids binding to most class I-a molecules and the class I-b molecule Qa-2. Stabilization by fMLF-benzyl amide could be detected at peptide concentrations as low as 100 nM. While N-formyl peptides as short as two amino acids in length stabilized expression of M3a-Ld, increasing the length of these peptides added to the stability of peptide-MHC complexes as determined by 27-37 degrees C temperature shift experiments. We propose that relaxation of the length rule may represent a compensatory adaptation to maximize the number of peptides that can be presented by H-2M3a.     

Maturation of Qa-1b class I molecules requires beta 2-microglobulin but is TAP independent

Two conformationally distinct and stable forms of Qa-1b, one strongly associated with beta 2-microglobulin (beta 2m) and the other associated with a novel molecule, gp44, were observed during immunochemical studies on the expression of Qa-1b molecules in mouse spleen cells. Both forms are efficiently processed and expressed at the cell surface. However, a large proportion of Qa-1b was found to be disulfide linked to gp44 without any detectable beta 2m. In TAP1-deficient mice, both forms undergo carbohydrate processing and are expressed on the cell surface, suggesting that they may traffic using a pathway not requiring a TAP association step. Consistent with this, size exclusion chromatography of newly synthesized class I molecules shows that high molecular mass complexes containing H-2Kk do not contain Qa-1b. Although Qa-1b can be stably expressed without beta 2m, there was no maturation of either form in cells from beta 2m-deficient mice where heavy chains were rapidly degraded. These results suggest that Qa-1b, like most other class I molecules, requires beta 2m for an initial folding step. However, beta 2m is not essential for subsequent processing of Qa-1b molecules.     

AMP analogs: their function in the activation of glycogen phosphorylase b

A series of AMP analogs has been selected in order to better understand the structural requirements (a) for the efficient binding of the activator molecule at the correct site on phosphorylase b from rabbit skeletal muscle and (b) for the activation which is observed. Two types of activation are known, according to Black and Wang [J. Biol. Chem. 243, 5892-5898 (1968)]: either a cooperative response with respect to the activator concentration (like the one which is obtained for AMP itself) or a non-cooperative response observed in the case of IMP. It is shown that the 5'-phosphate moiety is absolutely required for the analog to bind at the correct site (adenine or adenosine bind at another enzymic site), and that the free enthalpy, delta G, corresponding to the association process varies in a complex manner with respect to the substitution of the different positions of the AMP molecule. Moreover, the differences delta G (analog) - delta G (AMP) = delta G obtained for two types of substitution separately do not add up to the same energy difference as the one obtained when the two substitutions are made simultaneously on the AMP molecule. It appears that all the mononucleotides which have been tested up to now may be divided into two classes. Class I (AMP class) is characterized, apart from a strong activation, by the following features: (a) one molecule of analog expels two molecules of bound glucose 6-phosphate as it binds on the enzyme; (b) bound analog protects slowly one crucial cysteinyl residue against attack by 5,5'-dithio-bis(2-nitrobenzoic acid) at 4 degrees C; (c) association of two molecules of dimer is strengthened at 4 degrees C in the presence of the analog. Class II (IMP class) is associated with a weak activation and with the following set of properties: (a) a single molecule of bound glucose 6-phosphate is released as the first molecule of analog binds on the dimer; (b) two slowly reacting cysteinyl residues per subunit are immediately protected against 5,5'-dithio-bis(2-nitrobenzoic acid) by the binding of the analog at 4 degrees C; (c) the analog dissociates the low amount of tetramer which is present at 4 degrees C in the absence of AMP into two molecules of dimer. These results are discussed according to a plausible scheme of transconformations taking place in glycogen phosphorylase b, a model which has been derived earlier by relaxation studies.     

Naturally processed peptides from HLA-DQ7 (alpha1*0501-beta1*0301): influence of both alpha and beta chain polymorphism in the HLA-DQ peptide binding specificity

Self peptides bound to HLA-DQ7 (alpha1*0501-beta1*0301), one of the HLA molecules associated with protection against insulin-dependent diabetes mellitus, were characterized after their acid elution from immunoaffinity-purified HLA-DQ7 (alpha1*0501-beta1*0301) molecules. The majority of these self peptides derived from membrane-associated proteins including HLA class I, class II, class II-associated invariant chain peptide and the transferrin-receptor (TfR). By in vitro binding assays, the specificity of these endogenous peptides for HLA-DQ7 (alpha1*0501-beta1*0301) molecules was confirmed. Among these peptides, the binding specificity of the TfR 215-230 self peptide was further examined on a variety of HLA-DQ and DR dimers. Several findings emerged from this analysis: (1) this peptide displayed HLA-DQ allelic specificity, binding only to HLA-DQ7 (alpha1*0501-beta1*0301); (2) when either the DQalpha or DQbeta chain was exchanged, little or no binding was observed, indicating that specificity of HLA-DQ peptide binding was determined by polymorphic residues of both the alpha and beta chains. (3) Unexpectedly, the TfR 215-230 self peptide, eluted from DQ, was promiscuous with regard to HLA-DR binding. This distinct DR and DQ binding pattern could reflect the structure of these two molecules as recently evidenced by crystallography.     

The Qa-1b molecule binds to a large subpopulation of murine NK cells

Recent studies on human NK cells have demonstrated that the NK cell CD94/NKG2 receptors bind to the nonclassical MHC class I molecule HLA-E. A functional CD94/NKG2 complex has not yet been identified in rodents, but cDNA encoding rat and mouse CD94 and NKG2 have recently been cloned, suggesting that CD94/NKG2 receptors may exist in species other than man. The mouse nonclassical MHC class I molecule Qa-1 shares several features with HLA-E. This suggests that Qa-1 may be similarly recognized by murine NK cells. To study the ability of Qa-1 to bind to murine NK cells, we have produced a soluble tetrameric form of Qa-1b. In the present study, we demonstrate that Qa-1b tetramers distinctly bind to a large subset of fresh or IL-2-activated NK1.1+/CD3- splenocytes independently of the expression of Ly49 inhibitory receptors. Binding occurs whether NK cells have evolved in an MHC class I-expressing or in an MHC class I-deficient environment. Our data suggest the existence of a Qa-1-recognizing structure on a large subpopulation of murine NK cells that may be similar to the human CD94/NKG2 heterodimeric complex.     

Endogenous peptides bound to HLA-A3 possess a specific combination of anchor residues that permit identification of potential antigenic peptides

A motif specific to peptides that bind to the human class I major histocompatibility complex molecule HLA-A3 was identified by sequence analysis of HPLC fractions containing endogenous peptides. Twenty-six different sequences were obtained, 19 of which were nonamers. The majority of these endogenous peptide sequences contained Leu at position (P)2, while most sequences contained Tyr or Lys at P9. In addition, Phe was shared by 16 sequences at P3. Synthetic peptides corresponding to endogenous peptide sequences were shown to bind to HLA-A3. The importance of Leu at P2 and Tyr or Lys at P9 ("anchor" residues) for peptide binding to HLA-A3 was demonstrated by the following results: (i) peptides GLFGGGGGY, GLFGGGGGK, and GLGGGGFGY, but not GLFGGGGGV, specifically bound to HLA-A3 and (ii) six nonapeptides from within the influenza A nucleoprotein, matrix, and polymerase proteins, selected for synthesis based upon their possession of P2 and P9 anchor residues, were shown to bind HLA-A3. In contrast, none of a set of eight peptides that bound to HLA-A2, or six that bound to HLA-B27, bound detectably to HLA-A3. These findings provide a rationale for the design and selection of peptides that can be recognized by HLA-A3-restricted T cells.     

HLA-E surface expression depends on binding of TAP-dependent peptides derived from certain HLA class I signal sequences

Previous studies showed that HLA-E was expressed in lymphoblastoid cell line (LCL) 721.221 cells, but surface expression was lacking. To determine the signals controlling surface expression, we constructed a series of hybrid genes using complementary portions derived from the HLA-E and HLA-A2 genes. In this manner, a hybrid of HLA-E was identified, designated AEH, which differed from HLA-E by having the HLA-A2 signal sequence substituting for the HLA-E leader peptide. Transfection of LCL 721.221 cells with AEH induced HLA-E surface expression. Analysis of peptides bound to HLA-E revealed that a nonamer peptide derived from the A2 signal sequence was the predominant peptide bound. LCL 721.221 cells transfected with certain class I genes, including HLA-G, were also sufficient to promote peptide binding and HLA-E surface expression without increasing the level of HLA-E heavy chain synthesis. Peptides bound to HLA-E consisted of nine amino acids, with methionine at position 2 and leucine in the carboxyl-terminal position, and were nearly identical to the leader sequence-derived peptide previously shown to be a predominant peptide bound to the murine Qa-1 Ag. Signal peptides derived from certain HLA-B proteins with threonine in position 2 only marginally up-regulated HLA-E surface expression in .221 cells. An examination of HLA-E peptide binding in the TAP negative cell line .134 indicated that peptide binding to HLA-E was dependent on a functional TAP heterodimer regardless of whether peptide was available in cis, as in the AEH construct, or in trans, as in the class I transfectants of .221 cells.     

Indices of carbohydrate and lipid metabolism in vasopressin-replete and -deficient New Zealand genetically hypertensive rats

Intracellular antigens are continually presented to cytotoxic T lymphocytes by major histocompatibility complex (MHC) class I molecules, which consist of a polymorphic 43 kDa heavy chain and a 12 kDa soluble subunit beta 2-microglobulin (beta 2m), and which bind an 8-10 amino-acid antigenic peptide. The assembly of this trimolecular complex takes place in the lumen of the endoplasmic reticulum (ER) and almost certainly requires cofactors. Most MHC class I molecules in the ER that have not yet acquired peptide are simultaneously bound to the transporter associated with antigen processing (TAP), to the 48 kDa glycoprotein tapasin and to the lectin-like chaperone calreticulin, in a multicomponent 'loading complex'. Previous studies have shown that a mutant MHC class I molecule T134K (in which Thr134 was changed to Lys) fails to bind to TAP. Here, we show that this point mutation also disrupted, directly or indirectly, the interaction between MHC class I molecules and calreticulin. T134K molecules did not present viral antigens to T cells even though they bound peptide and beta 2m normally in vitro. They exited the ER rapidly as 'empty' MHC class I complexes, unlike empty wild-type molecules which are retained in the ER and degraded. We show here that, paradoxically, the rapid exit of empty T134K molecules from the ER was dependent on a TAP-derived supply of peptides. This implies that MHC class I assembly is a two-stage process: initial binding of suboptimal peptides is followed by peptide optimisation that depends on temporary ER retention.     

Peptide selection by an MHC H-2Kb class I molecule devoid of the central anchor ("C") pocket

The peptide-binding site of the murine MHC class I molecule H-2Kb contains a deep C pocket, that is critical for peptide binding, as it accepts the anchor phenylalanine or tyrosine residue located in the middle (position 5, P5F/Y) of H-2Kb binding peptides. H-2Kb predominantly binds octameric peptides. By both criteria, H-2Kb is unique among the known murine and human class I molecules, none of which have a deep C pocket or preferentially select octamers. We investigated the relative importance of the C pocket in peptide selection and binding by the MHC. An MHC class I H-2Kb variant, Kbw9, predicted to contain no C pocket, was engineered by replacing valine at MHC9 with tryptophan. This mutation drastically altered the selection of peptides bound to Kbw9. The Kbw9 molecule predominantly, if not exclusively, bound nonamers. New peptide anchor residues substituted for the loss of the P5F/Y:C pocket interaction. P3P/Y, which plays an auxiliary role in binding to Kb, assumed the role of a primary anchor, and P5R was selected as a new primary anchor, most likely contacting the E pocket. These experiments demonstrate that the presence of a deep C pocket is responsible for the selection of octameric peptides as the preferred ligands for Kb and provide insight into the adaptation of peptides to a rearranged MHC groove.     

The structural basis of MHC control of collagen-induced arthritis; binding of the immunodominant type II collagen 256-270 glycopeptide to H-2Aq and H-2Ap molecules

The Aq major histocompatibility complex (MHC) class II molecule is associated with susceptibility to murine collagen-induced arthritis (CIA), whereas the closely related H-2Ap molecule is not. To understand the molecular basis for this difference, we have analyzed the ability of H-2Aq and H-2Ap molecules (referred to as Aq and Ap) to bind and present collagen type II (CII)-derived glycosylated and non-glycosylated peptides. T cell clones specific for the immunodominant CII 256-270 peptide and restricted to both Aq and Ap molecules were identified. When these clones were incubated with CII protein and either Aq- or Ap-expressing antigen-presenting cells (APC), only Aq-expressing APC were able to induce stimulation. With the use of A(beta) transgenic mice this could be shown to be solely dependent on the MHC class II molecule itself and to be independent of other MHC- or non-MHC genes. Peptide binding studies were performed using affinity-purified MHC class II molecules. The CII 256-270 peptide bound with lower affinity to the Ap molecule than to the Aq molecule. Using a set of alanine-substituted CII 256-270 peptides, MHC class II and T cell receptor (TCR) contacts were identified. Mainly the side chains of isoleucine 260 and phenylalanine 263 were used for binding both the Aq and Ap molecule, i.e. the peptide was orientated similarly in the binding clefts. The major TCR contact amino acids were lysine 264, which can be posttranslationally modified, and glutamic acid 266, which is the only amino acid in the heterologous peptide which differs from the mouse sequence. Glycosylation at positions 264 and 270 of the CII 256-270 peptide did not change the anchor positions used for binding to the Aq or Ap molecules. The autologous form of the peptide (with aspartic acid at position 266) bound with lower affinity to the Aq molecule as compared with the heterologous peptide. The variable affinity displayed by the immunodominant CII 256-270 peptide for different MHC class II molecules, the identification of MHC and TCR contacts and the significance of glycosylation of these have important implications for the understanding of the molecular basis for inherited MHC class II-associated susceptibility to CIA and in turn, for development of novel treatment strategies in this disease.     

Selective release of some invariant chain-derived peptides from HLA-DR1 molecules at endosomal pH

The predominant peptides bound to major histocompatibility complex class II molecules expressed on human B cells are derived from a relatively limited number of self proteins. To determine whether any of the prebound self peptides might be released in endosomes during recycling, water-soluble HLA-DR1 molecules were incubated with a high affinity synthetic peptide at pH 4.0 and 7.0 at 37 degrees C. The resulting bound peptide repertoire was then acid extracted, and separated by reversed-phase high performance liquid chromatography. Using a combination of mass spectrometry and ultraviolet spectroscopy, prebound self peptides and newly bound synthetic peptide were characterized. Most self peptides bound to HLA-DR1 were not appreciably released during extended exposure to pH 4.0. However, some invariant chain-derived peptides were uniquely released at this pH.