Biochemical studies on the H-2K antigens of the MHC mutant bml

Biochemical analysis of the H-2K-gene product from the MHC mutant strain bml and from the C57BL/6 parent strain has been carried out in order to characterize the structural differences between parent and mutant K-gene products. Based on comparative tryptic peptide mapping of the cyanogen bromide fragments from these glycoproteins, two peptide differences were localized to the CN-Ia fragment. Partial amino-acid sequence analysis revealed two alterations in the primary structure of Kbml involving substitutions of tyrosine for arginine at position 155, and tyrosine for leucine at position 156. Both of these amino-acid replacements require a minimum of two nucleotide base changes at the nucleic acid level. These changes were the only alterations noted differentiating the Kbml and Kb glycoproteins. However, because our techniques allow us to analyze only 75 to 80 percent of the extra cellular portion of H-2Kb, it is possible there are other undetected changes. Nonetheless, the biochemical data are consistent with the hypothesis that the structural alterations noted in the Kbml mutant glycoprotein are directly related to the observed immunological specificity relative to the parent Kb molecule. Peptide comparisons of the Kb molecules of two C57BL/6 sublines and of the H-2b lymphoblastoid cell line, EL-4, disclosed no difference.     

Molecular heterogeneity of D-end products detected by anti-H-2.28 sera

Immunoprecipitation of NP-40 lysates of 125I-labeled lymph-node cells with different anti-H-2 sera and with anti-Qa-2 serum has shown that the BALB/cByA strain (H-2d, Qa-2-negative) expresses, besides H-2Ld, another molecule that is not detectable in the BALB/c-H-2dm2 strain. Electrophoresis in SDS polyacrylamide gels indicated that this molecule, provisionally designated Lq, has an apparent molecular weight of 41000 daltons, in contrast to approximately 49000 daltons for H-2Kd and H-2Ld, and 47000 daltons for H-2Dd molecules. The anti-Qa-2 serum precipitated from the Qa-2-positive strains BALB/cHeA but not from the Qa-2-negative strains BALB/cByA and BALB/c-H-2dm2 a protein that gave a very strong band corresponding to the molecular weight 41000 daltons in the gel electrophoresis. The biochemical characteristics of the Lq molecule are thus more similar to those of Qa-2 than of H-2 antigens.     

Broad recognition of cytotoxic T cell epitopes from the HIV-1 envelope protein with multiple class I histocompatibility molecules.

A few cases have been described of antigenic determinants that are broadly presented by multiple class II MHC molecules, especially murine I-E or human DR, in which polymorphism is limited to the beta chain, and the alpha chain is conserved. However, no similar cases have been studied for presentation by class I MHC molecules. Because both domains of the MHC peptide binding site are polymorphic in class I molecules, exploring permissiveness in class I presentation would be of interest, and also such broadly presented antigenic determinants would clearly be useful for vaccine development. We had defined an immunodominant determinant, P18, of the HIV-1 gp160 envelope protein recognized by human and murine CTL. To determine the range of class I MHC molecules that could present this peptide and to determine whether two HIV-1 gp160 Th cell determinants, T1 and HP53, could also be presented by class I MHC molecules, we attempted to generate CTL specific for these three peptides in 10 strains of B10 congenic mice, representing 10 MHC types, and BALB/c mice. P18 was presented by at least four different class I MHC molecules from independent haplotypes (H-2d, p, u, and q to CD8+ CTL. In H-2d and H-2q the presentation was mapped to the D-end class I molecule, and for Dd, a requirement for both the alpha 1 and alpha 2 domains of Dd, not Ld, was found. HP53 was also presented by the same four different class I MHC molecules to CD8+ CTL although at higher concentrations. T1 was presented by class I molecules in three different strains of distinct MHC types (B10.M, H-2f; B10.A, H-2a; and B10, H-2b) to CTL. The CTL specific for P18 and HP53 were shown to be CD8+ and CD4- and to kill targets expressing endogenously synthesized whole gp160 as well as targets pulsed with the corresponding peptide. To compare the site within each peptide presented by the different class I molecules, we used overlapping and substituted peptides and found that the critical regions of each peptide are the similar for all four MHC molecules. Thus, antigenic sites are broadly or permissively presented by class I MHC molecules even without a nonpolymorphic domain as found in DR and I-E, and these sequences may be of broad usefulness in a synthetic vaccine.     

Structural relationships among the H-2 D-regions of murine MHC haplotypes

The number of genes encoding functional Ag-presenting molecules in the D region of the murine MHC differs among haplotypes. For example, the H-2b D region contains a single "D/L" gene, H-2Db, whereas the d-haplotype encodes two, H-2Dd and Ld. Using D/L specific oligonucleotide probes, we have found that, as with H-2d, the q- and v-haplotypes contain two D/L genes, whereas the other haplotype examined have one. Hybridization analysis using cloned probes that map between H-2Dd and Ld revealed similar structures in each of the three haplotypes (d, q, and v) which have "duplicated" D regions. Two approaches were used to examine allelic relationships among the D/L genes. First, the 5' region of the H-2Db gene was sequenced, and found to be more similar to H-2Ld than to H-2Dd. Second, oligonucleotide probes that distinguish H-2Ld from H-2Dd revealed H-2Ld-related genes in several haplotypes, including the duplicated haplotypes H-2q and H-2v. Analogous probes specific for H-2Dd, however, did not detect similar sequences in the other haplotypes. We interpret these results to mean that the three duplicated D regions arose from a common duplication event, and share the five gene structure of the D region cluster defined in H-2d. However, subsequent events have generated sequence divergence at the D-locus.     

Biochemical studies on the H-2K antigens of the MHC mutantbm1

Biochemical analysis of the H-2K-gene product from the MHC mutant strainbml and from the C57BL/6 parent strain has been carried out in order to characterize the structural differences between parent and mutant K-gene products. Based on comparative tryptic peptide mapping of the cyanogen bromide fragments from these glycoproteins, two peptide differences were localized to the CN-Ia fragment. Partial amino-acid sequence analysis revealed two alterations in the primary structure of K^bml involving substitutions of tyrosine for arginine at position 155, and tyrosine for leucine at position 156. Both of these amino-acid replacements require a minimum of two nucleotide base changes at the nucleic acid level. These changes were the only alterations noted differentiating the K^bml and K^b glycoproteins. However, because our techniques allow us to analyze only 75 to 80 percent of the extra cellular portion of H-2K^b, it is possible there are other undetected changes. Nonetheless, the biochemical data are consistent with the hypothesis that the structural alterations noted in the K^bml mutant glycoprotein are directly related to the observed immunological specificity relative to the parent K^b molecule. Peptide comparisons of the K^b molecules of two C57BL/6 sublines and of the H-2^b lymphoblastoid cell line, EL-4, disclosed no difference.     

BALB/c-H-2 db : A newH-2 mutant in BALB/cKh that identifies a locus associated with theD region

A newH-2 mutant, BALB/c-H-2 ^db , is described. This mutant originated in BALB/c, is inbred, and is coisogenic with the parental BALB/cKh strain. The mutation is of the loss type since BALB/c-H- ^db rejects BALB/c, but not vice versa. Complementation studies have localized the mutation to theD region of theH-2 complex. A cross between BALB/c-H-2 ^db and B10.D2-H-2 ^da failed to complement for either BALB/c or B10.D2 skin grafts, indicating that these are two separate mutations at the same locus (Z2). Direct serological analysis and absorption studies revealed that, with one exception, theH-2 andIa specificities of BALB/c and BALB/c-H-2 ^db are identical. In particular,H-2.4, the H-2D^d private specificity, is quantitatively and qualitatively identical in the two strains. The exception is that of the specificities detected by antiserum D28b: (k×r)F₁ anti-h, which contains anti-H-2.27, 28, and 29. These specificities appear to be absent from theH-2 ^db mutant since they are not detected directly or by absorption. Other public specificities are present in normal amounts,e.g., the reaction with antisera to H-2.3, 8, 13, 35, and 36. The reaction with antiserum D28 (f×k)F₁ anti-s, which contains antibodies to H-2.28, 36, and 42, is the same in both strains. Antiserum made between the two strains (H-2 ^db anti-H-2 ^d ) reacts like an anti-H-2 serum, in that it reacts with both T and B cells by cytotoxicity, but is not a hemagglutinating antibody. The serum reacts as does the D28b serum in both strain distribution and in cross-absorption studies. We conclude that theH-2 ^db mutation occurred at a locus in theD region, resulting in the loss of the H-2.28 public serological specificity and of a histocompatibility antigen. Whether these are one and the same antigen is not yet known. The data, in view of other evidence, imply that the public and private specificities are coded for by separate genes.Abbreviations used in this paper are as follows CML cell-mediated lysis - MLR mixed lymphocyte reaction - GVHR graft-versus-host reaction - RFC rosette-forming cells - RAM-Ig rabbit anti-mouse IgG     

Differential lack of class I H-2d antigen expression by sublines of the BALB/c S49 T cell lymphoma

Five different sublines of the BALB/c murine S49.1 T cell lymphoma were found to exhibit distinct patterns of absence of detectable H-2d class I major histocompatibility antigen expression. The results were demonstrated and verified by a) the generation of H-2Kd-, H-2Dd,Ld-, and H-2Ld-specific cytotoxic T lymphocytes that were assayed on S49.1 target cell lines, b) antibody-mediated cytotoxicity with the use of anti-H-2d monoclonal reagents, and c) flow microfluorometry. The five lines investigated were S49.1, T-25, T-25ADH, Thy-1-, and 100/0. None of these lines expressed detectable levels of Ld. S49.1 expressed both Kd and Dd, T-25 and T-25ADH expressed Dd but not Kd or Ld, Thy-1- expressed Kd but not Dd or Ld, and 100.0 did not express any detectable amounts of Kd, Dd, or Ld. These results indicate that K and D (and L) antigens can be expressed independently of each other and suggest that expression of class I antigens is controlled in a locus-specific manner.     

Isolation and antigenic characterization of the product of a third polymorphic H-2 locus, H-2L.

The serology, immunochemistry, and genetics of the product(s) of a third H-2 locus, H-2L (previously designated D') have been studied by using an antiserum raised in BALB/c H-2db mutant mice against tissues from the wild type strain, BALB/c. Genetic mapping studies and sequential immunoprecipitation experiments both indicate that this antiserum reacts specifically with L molecules. These results imply that an H-2L product is antigenically undetectable in BALB/c-H-2db mice and that the lesion in this mutant is confined to the H-2L and not the H-2D locus. Two new specificities, H-2.64 and H-2.65, are defined by the reactivity of anti-L serum on allogeneic cells, and the strain distribution of these specificities suggests the existence of at least three H-2L alleles. This third H-2 locls is therefore polymorphic and in view of this and other similarities to the H-2K and H-2D loci, it must be considered in any evolutionary models dealing with the origin of multiple subloci of the major histocompatibility complex.     

MHC class I recognition by NK receptors in the Ly49 family is strongly influenced by the beta 2-microglobulin subunit

NK cell recognition of targets is strongly affected by MHC class I specific receptors. The recently published structure of the inhibitory receptor Ly49A in complex with H-2Dd revealed two distinct sites of interaction in the crystal. One of these involves the alpha1, alpha2, alpha3, and beta2-microglobulin (beta2m) domains of the MHC class I complex. The data from the structure, together with discrepancies in earlier studies using MHC class I tetramers, prompted us to study the role of the beta2m subunit in MHC class I-Ly49 interactions. Here we provide, to our knowledge, the first direct evidence that residues in the beta2m subunit affect binding of MHC class I molecules to Ly49 receptors. A change from murine beta2m to human beta2m in three different MHC class I molecules, H-2Db, H-2Kb, and H-2Dd, resulted in a loss of binding to the receptors Ly49A and Ly49C. Analysis of the amino acids involved in the binding of Ly49A to H-2Dd in the published crystal structure, and differing between the mouse and the human beta2m, suggests the cluster formed by residues Lys3, Thr4, Thr28, and Gln29, as a potentially important domain for the Ly49A-H-2Dd interaction. Another possibility is that the change of beta2m indirectly affects the conformation of distal parts of the MHC class I molecule, including the alpha1 and alpha2 domains of the heavy chain.     

The products of transferred H-2 genes express determinants that restrict hapten-specific cytotoxic T cells

Cell lines into which cloned H-2 genes had been introduced (i.e., transformants) were used to correlate the genes and their products that are capable of functioning as H-2 restriction elements for hapten-self-(AED and TNP) specific cytotoxic T cells (CTL). These transformants provided a unique system in which major histocompatibility restricted (MHC) T cell recognition could be examined by using cells that express only H-2Ld or only H-2Dd gene products. BALB/c (H-2d) anti AED-self CTL lysed both the H-2Ld and Dd transformants, but not parental, i.e., untransformed, cells. The AED-self lysis of the Ld and Dd transformants was shown to be specifically inhibited by anti-H-2Ld and anti H-2Dd monoclonal antibody, respectively. In contrast to these results, BALB/c anti TNP-self CTL were found to lyse readily the Dd but not Ld transformed lines, supporting reports indicating that H-2Ld-restricted TNP-self CTL could not be detected. The results of this study thus demonstrate that the cell surface products encoded by these transferred MHC class I genes contain self determinants recognized by CTL.     

The MHC class I molecule H-2Dp inhibits murine NK cells via the inhibitory receptor Ly49A.

MHC class I molecules strongly influence the phenotype and function of mouse NK cells. NK cell-mediated lysis is prevented through the interaction of Ly49 receptors on the effector cell with appropriate MHC class I ligands on the target cell. In addition, host MHC class I molecules have been shown to modulate the in vivo expression of Ly49 receptors. We have previously reported that H-2Dd and H-2Dp MHC class I molecules are able to protect (at the target cell level) from NK cell-mediated lysis and alter the NK cell specificity (at the host level) in a similar manner, although the mechanism behind this was not clear. In this study, we demonstrate that the expression of both H-2Dd and H-2Dp class I molecules in target cells leads to inhibition of B6 (H-2b)-derived Ly49A+ NK cells. This inhibition could in both cases be reversed by anti-Ly49A Abs. Cellular conjugate assays showed that Ly49A-expressing cells indeed bind to cells expressing H-2Dp. The expression of Ly49A and Ly49G2 receptors on NK cells was down-regulated in H-2Dp-transgenic (B6DP) mice compared with nontransgenic B6 mice. However, B6DP mice expressed significantly higher levels of Ly49A compared with H-2Dd-transgenic (D8) mice. We propose that both H-2Dd and H-2Dp MHC class I molecules can act as ligands for Ly49A.     

Induction of the H-2 D antigen during B cell activation

Mitogenic activation causes increased expression of class I Ag of the MHC in mouse B cells. The increased expression was seen in flow cytometry analysis for both K and D in k as well as d haplotypes. A more detailed molecular analysis was carried out for H-2Dd. Increased expression (10- to 20-fold) of the H-2 Dd gene was detected at both protein and messenger RNA levels, and the time course for the accumulation of H-2 Dd protein on the cell surface parallels the increase in the steady-state messenger RNA levels. The increase in H-2 Dd expression in small B cells stimulated with LPS is detectable after 10 h of culture. The present data provide molecular and serologic evidence about alterations in the expression of the H-2 Dd Ag, previously identified as a B cell activation antigen B7.2. Our results indicate a new significance for the function and regulation of the MHC during immune responses, and suggest that the class I molecules may serve some role in the B cell activation process.     

Induction of cytotoxic T cells to a cross-reactive epitope in the hepatitis C virus nonstructural RNA polymerase-like protein.

Cytotoxic T lymphocytes (CTL) have been found to mediate protection in vivo against certain virus infections. CTL also may play an important role in control of infection by hepatitis C virus (HCV), but no CTL epitopes have yet been defined in any HCV protein. The nonstructural protein with homology to RNA polymerase should be a relatively conserved target protein for CTL. To investigate the epitope specificity of CTL specific for this protein, we used 28 peptides from this sequence to study murine CTL. Mice were immunized with a recombinant vaccinia virus expressing the HCV nonstructural region corresponding to the flavivirus NS5 gene (RNA polymerase), and the primed spleen cells were restimulated in vitro with peptides. CTL from H-2d mice responded to a single 16-residue synthetic peptide (HCV 2422 to 2437). This relatively conserved epitope was presented by H-2d class I major histocompatibility complex (MHC) molecules to conventional CD4- CD8+ CTL but was not recognized by CTL restricted by H-2b. Moreover, exon shuffle experiments using several transfectants expressing recombinant Dd/Ld and Kd demonstrated that this peptide is seen in association with alpha 1 and alpha 2 domains of the Dd class I MHC molecule. This peptide differs from the homologous segments of this nonstructural region from three other HCV isolates by one residue each. Variant peptides with single amino acid substitutions were made to test the effect of each residue on the ability to sensitize targets. Neither substitution affected recognition. Therefore, these conservative mutations affected peptide interaction neither with the Dd class I MHC molecule nor with the T-cell receptor. Because these CTL cross-react with all four sequenced isolates of HCV in the United States and Japan, if human CTL display similar cross-reactivity, this peptide may be valuable for studies of HCV diagnosis and vaccine development. Our study provides the first evidence that CD8+ CTL can recognize an epitope from the HCV sequence in association with a class I MHC molecule.     

Oxidative stress induces the expression of the major histocompatibility complex in murine tumor cells.

The effect of t-butyl hydroperoxide (t-BOOH) on the induction of the Major Histocompatibility Complex (MHC) class I genes has been studied in two cell clones (B9 and G2) of the methylcholanthrene-induced murine fibrosarcoma GR9. These two clones were selected based on their different biological and biochemical behavior specially related to their tumor induction capability when injected into a BALB/c mouse. t-BOOH (0.125 mM) induced the expression of H-2 molecules in both cell clones. In B9 cell clone, in which MHC basal expression is very low or absent, t-BOOH significantly induced H-2Kd, H-2Dd and H-2Ld molecules. In G2 cell clone the expression of MHC class I genes was also enhanced by the xenobiotic, the effect being especially significant on the H-2Ld molecule which is not expressed under basal conditions. H-2 molecules expression was accompanied by the activation of the transactivator factor NF kappa B. These results suggest that oxidative stress may modulate the antigen expression of tumor cells and thus the immune response of the host organism. Basal levels of oxidative parameters, such as anti-oxidant enzymes, malondialdehyde (MDA) and the DNA damaged base 8-hydroxy-2'-deoxyguanosine (8-OHdG), showed differences between the two fibrosarcoma cell clones.     

"Physiological interaction" does not explain the H-2 requirement for recognition of virus-infected cells by cytotoxic T cells.

B10.A (H-2Kk, H-2Dd) ectromelia-immune T cells from secondary responses in vitro were protent killers of both infected L929 (H-2Kk H-2Dk) and infected P-815 (H-2Kd, H-2Db) target cells. Specific competition with unlabelled targets showed that two separate T cell subsets were responsible for lysis of infected L929 and infected P-815 cells. One hypothesis to account for this (a form of "physiological interaction") is that T cells which kill one target e.g. infected L929) display only one out of two possible self-complementary recognition structures, in this example the H-2Kk alloantigen, not H-2Dd, whereas T cells that lyse infected P-815 targets display only H-2Dd, not H-2Kk. This hypothesis was tested and seems untenable because of the following results: A.TH (H-2Ks, H-2Dd) ectromelia-immune, secondary cytotoxic T cells which killed infected SJL/J (H-2Ks, H-2Ds) targets were themselves inactivated by pre-incubation with SJL/J cytotoxic T cells generated in one-way mixed lymphocyte reaction (MLR) against BALB/c (H-2Kd, H-2Dd). A.TL (H-2Ks, H-2Dd) ectromelia-immune secondary cytotoxic T cells which killed infected BALB/c targets were themselves inactivated by BALB/c cytotoxic T cells generated in MLR against SJL/J. Thus, virus-immune T cells which lyse infected targets by virtue of shared H-2K are also displaying H-2D alloantigen, and vice versa.     

Role of endogenous peptides in murine allogenic cytotoxic T cell responses assessed using transfectants of the antigen-processing mutant 174xCEM.T2.

One model to explain the high frequency of alloreactive T cells proposes that allogeneic MHC molecules are recognized together with host cell-derived peptides. A model system was developed to investigate the relevance of this mechanism by expression of H-2Dd or H-2Ld in 174xCEM.T2 (T2) cells. This human cell line contains a mutation in its Ag-processing pathway that should restrict the association of endogenous peptides with cell surface class I molecules. CTL generated by stimulating C57BL/6 (H-2b) responder cells with H-2Dd or H-2Ld transfectants of the human B cell line C1R or the murine T cell lymphoma EL4 were assayed for their ability to recognize alloantigenic determinants on these transfectants. The major fraction of the H-2Dd-specific allogeneic CTL response, generated in a MLC or under clonal limiting dilution conditions, was composed of T cells that recognized H-2Dd expressed on C1R or EL4 cells, but failed to recognize this molecule on T2 cells. Clonal analysis indicated that approximately one-third of these CTL recognized determinants that were unique to H-2Dd expressed on C1R stimulator cells whereas the remainder recognized determinants that were also found on EL4 transfectants. Less than 10% of H-2Dd-reactive CTL recognized the T2 transfectant, and these clones also killed C1R-Dd and EL4-Dd. This result suggests that the great majority of H-2Dd-specific alloreactive CTL recognize determinants that are formed by a complex of H-2Dd with endogenous peptides that are absent or significantly reduced in T2 cells. Based on recognition of human or murine transfectants, these CTL exhibit some level of specificity for the structure or composition of the bound peptides. Examination of allogeneic CTL specific for H-2Ld revealed populations similar to those described for H-2Dd. In addition, a major new population was present that recognized determinants shared between C1R-Ld and T2-Ld but not present on EL4-Ld. These results are consistent with the idea that the alloreactive response to H-2Ld is also largely dependent on the presence of bound peptide. However, they also may indicate that the H-2Ld molecule expressed on T2 cells is occupied by one or more peptides that are shared with other human, but not murine, cells. The significance of these results to current models of alloreactivity is discussed.     

H-2D end confers dominant protection from IL-10-mediated acceleration of autoimmune diabetes in the nonobese diabetic mouse.

BALB/c mice that express IL-10 as a transgene in their pancreatic beta cells (Ins-IL-10 mice) do not develop diabetes, even after crossing to nonobese diabetic (NOD) mice ((Ins-IL-10 x NOD)F(1) mice). However, backcross of F(1) mice to NOD mice (NOD.Ins-IL-10 mice) results in N2 and N3 generations that develop accelerated diabetes. In this study, we found that NOD.Ins-IL-10 mice that expressed BALB/c-derived MHC molecules (NOD.Ins-IL-10(H-2(g7/d)) mice) were protected from diabetes. This protection associated with peri-islet infiltration and preserved beta cell function. Moreover, expression of I-A(d) and I-E(d) MHC class II molecules of BALB/c origin was not responsible for protection, but NOD.Ins-IL-10 mice that expressed BALB/c MHC class I D(d) molecules (NOD.Ins-IL-10(H-2(g7/d)) mice) did not develop diabetes. To directly test the possibility of a protective role of H-2D(d) in the development of accelerated diabetes, we generated transgenic mice expressing D(d) under the control of the MHC class I promoter. We found that double transgenic NOD.Ins-IL-10-D(d) mice developed accelerated diabetes in a fashion similar to NOD.Ins-IL-10 mice that were D(d) negative. Microsatellite analysis of H-2D(d)-linked loci confirmed association between BALB/c-derived alleles and protection of NOD.Ins-IL-10(H-2(g7/d)) mice. These results suggest a control of H-2D(d)-linked gene(s) on IL-10-mediated acceleration of autoimmune diabetes and dominant protection of the D(d) region in NOD.Ins-IL-10 mice.     

Synthesis of a defective H-2 histocompatibility antigen by mutant mouse strain BALB/c-H-2 dm2

Mutant mouse strain BALB/c-H-2 ^dm2 (dm2), which fails to express the H-2L^d histocompatibility antigen associated with the wild type, BALB/c, synthesizes instead a smaller molecule that is structurally related to H-2L^d but does not carry detectable alloantigenic determinants. This new protein, p40, is a membrane glycoprotein found in dm2 cells but not in BALB/c. p40 was detected by electrophoresis of dm2 glycoprotein preparations and by immunoprecipitation with heterologous H-2-specific antibodies. The p40 molecule is found associated with intracellular membranes but was not detected at the cell surface. Peptide mapping studies suggest that dm2 carries an alteration in the H-2L ^d structural gene, which prevents proper maturation of the protein product.     

Soluble class I MHC with beta2-microglobulin covalently linked peptides: specific binding to a T cell hybridoma

Soluble forms of the mouse MHC class I molecule, Dd, were produced in which the peptide binding groove was uniformly occupied by peptides attached via a covalent flexible peptide linker to the N terminus of the associated beta2-microglobulin. The MHC heavy chain and beta2-microglobulin were firmly associated, and the molecules displayed an Ab epitope requiring proper occupancy of the peptide binding groove. Soluble Dd containing a covalent version of a well-characterized Dd-binding peptide from HIV stimulated a T cell hybridoma specific for this combination. Furthermore, a tetravalent version of this molecule bound specifically with apparent high avidity to this hybridoma.