In spite of high frequencies of metal allergies, the structural basis for major histocompatibility complex (MHC)-restricted metal recognition is among the unanswered questions in the field of T cell activation. For the human T cell clone SE9, we have identified potential Ni contact sites in the T cell receptor (TCR) and the restricting human histocompatibility leukocyte antigen (HLA)-DR structure. The specificity of this HLA-DR–promiscuous VA22/VB17+ TCR is primarily harbored in its α chain. Ni reactivity is neither dependent on protein processing in antigen-presenting cells nor affected by the nature of HLA-DR–associated peptides. However, SE9 activation by Ni crucially depends on Tyr29 in CDR1α, an N-nucleotide–encoded Tyr94 in CDR3α, and a conserved His81 in the HLA-DR β chain. These data indicate that labile, nonactivating complexes between the SE9 TCR and most HLA-DR/peptide conjugates might supply sterically optimized coordination sites for Ni ions, three of which were identified in this study. In such complexes Ni may effectively bridge the TCR α chain to His81 of most DR molecules. Thus, in analogy to superantigens, Ni may directly link TCR and MHC in a peptide-independent manner. However, unlike superantigens, Ni requires idiotypic, i.e., CDR3α-determined TCR amino acids. This new type of TCR–MHC linkage might explain the high frequency of Ni-reactive T cells in the human population.

The induction of contact sensitivity in mice by hapten reagents such as trinitrochlorobenzene (TNCB) involves the activation of class II major histocompatibility complex (MHC)-restricted, hapten-specific, CD4+ T cells. Reports from different laboratories have indicated that the relevant antigenic epitopes in such reactions might include hapten-conjugated, MHC class II-associated peptides. This study for the first time directly demonstrates that hapten-peptides account for the majority of determinants recognized by trinitrophenyl (TNP)-specific CD4+ T lymphocytes. The sequences of those TNP carrier peptides do not have to be related to mouse proteins. Thus, we show that TNP-modified peptides derived from mouse IgG, pigeon cytochrome c or staphylococcal nuclease known to bind to I-Ab or from lambda repressor with specificity to I-Ad as well as TNP-proteins such as bovine serum albumin, ovalbumin or keyhole limpet hemocyanin all create class II-restricted hapten determinants for a number of TNP-specific T cell clones and hybridomas. All of these cells were induced with cells modified by trinitrobenzene sulfonic acid (TNBS). In addition, we present arguments indicating that individual TNP-specific helper T cells may cross-react with different TNP-peptides bound to identical class II molecules. Chemical treatment of antigen-presenting cells with TNCB or TNBS may thus result in a limited number of particularly repetitive immunodominant hapten epitopes. Immunodominant epitopes were also indicated by an overrepresentation of the TCR elements V beta 2 and V alpha 10 in I-Ab/TNP-specific T cells. Most importantly, however, we demonstrate that TNP attached to lysine 97 in the staphylococcal nuclease peptide 93-105 (i.e. a clearly "non-self" sequence) is able to prime mice for subsequent elicitation of contact sensitivity by TNCB in the absence of foreign protein. We take this to indicate that those TNP-peptide determinants defined by us as immuno-dominant are responsible for the induction of contact sensitivity to haptens.

We have recently described trinitrophenyl (TNP)-specific cytotoxic T lymphocyte (CTL) clones from C57BL/6 mice specific for hapten-modified peptides bearing a TNP-lysine in a peripheral position, i.e. in position 7 of H-2Kb-bound octapeptides. CTL recognition of such determinants is always sequence-dependent due to co-recognition of TNP as well as amino acid side chains of the carrier peptide. By the use of glycine-based designer peptides for primary induction of CTL in vitro, we have identified two sub-epitopes on individual position 7-haptenated peptides that form two TcR contact points and which can be independently recognized by cloned CTL. One of these sub-epitopes is represented by the hapten itself, the other by the amino acids tyrosine and lysine in positions 3 and 4 of the carrier peptide, respectively. Immunization with such TNP-modified peptides frequently results in the specific induction of CTL also reacting with the unmodified carrier peptides. DNA sequence analyses of the TcR revealed an extraordinary similarity of several independent TcR of CTL from individual mice and induced with different TNP-peptides. These receptor similarities clearly correlate with structural elements common to the immunizing peptides and suggest their origin from positive thymic selection of TcR on Kb-associated associated self-peptides bearing Tyr in position 3. Our data provide additional information concerning the topology of TcR binding to peptide/MHC complexes with, but also without, TNP. They also indicate a mechanism which might explain the potential of chemicals or drugs to induce autoimmune phenomena.

Hapten-specific T cell responses are responsible for chemically induced immune disorders. However, the molecular details of hapten interactions with T cell receptors (TCR) are poorly understood. Recent studies of trinitrophenyl (TNP)-specific responses revealed major histocompatibility complex-associated TNP-peptides as dominant epitopes for CD8+ and CD4+ T cells. The present study is based on the observation that two H-2Kb/TNP-specific CTL clones (II/7 and III/1), differing exclusively in two amino acids of their TCR alpha chains, also differed in their carrier specificities for various TNP-peptides. The genes of the two alpha chains and the common beta chain were cloned into expression vectors. Transfection of the TCR alpha chain of clone III/1 into a hybridoma of clone II/7 also transferred the fine specificity of clone III/1, indicating that the small alpha chain variations were indeed responsible for the different carrier specificities. Point mutations bridging the difference between the alpha chains of clones II/7 and III/1 and functional studies of the respective TCR alpha beta transfectants into a TCR-negative hybridoma revealed an unexpected result: the two receptors did not represent examples of structural complementarity for different sets of hapten-peptide conjugates; rather, they resembled two structures of principally similar specificity but of significantly different overall affinity. This was demonstrated more directly by comparing the fine specificities of III/1 transfectants expressing or not expressing the co-receptor CD8: the CD8-negative III/1 transfectant assumed a specificity pattern indistinguishable from that of a CD8-expressing, II/7-derived transfectant. Hence, comparable alterations of antigen recognition may be induced either by subtle TCR alterations or by removal of CD8, i.e. by the presence or absence of a non-polymorphic adhesion molecule.

No abstract