Search

Polymyositis (PM) and dermatomyositis (DM) are the prototypical inflammatory diseases of skeletal muscle. In PM, CD8+ T cells invade and destroy muscle fibers, whereas humoral effector mechanisms prevail in DM. We studied the expression of the cytotoxic mediator perforin in inflammatory cells in PM and DM muscle by semiquantitative PCR, immunohistochemistry and confocal laser microscopy. Similar levels of perforin mRNA were expressed in PM and DM, and abundant perforin-expressing CD3+CD8+ and CD3+ CD4+ T cells were observed in both diseases. However, there was a striking difference in the intracellular localization of perforin. In DM, perforin was distributed randomly in the cytoplasm of the inflammatory T cells. In contrast, 43% of the CD8+ T cells that contacted a muscle fiber in PM showed perforin located vectorially towards the target muscle fiber. The results suggest (a) that the random distribution of perforin in the cytoplasm of muscle-infiltrating T cells observed in DM reflects nonspecific activation, and (b) that the vectorial orientation observed only in PM reflects the specific recognition via the T cell receptor of an antigen on the muscle fiber surface, pointing to a perforin- and secretion-dependent mechanism of muscle fiber injury.

The human T cell response to the myelin basic protein (MBP) has been studied with respect to T cell receptor (TCR) usage, HLA class II restriction elements, and epitope specificity using a total of 215 long-term MBP-specific T cell lines (TCL) isolated from the peripheral blood of 13 patients with multiple sclerosis (MS) and 10 healthy donors. In most donors, the anti-MBP response was exceedingly heterogeneous. Using a panel of overlapping synthetic peptides spanning the entire length of human MBP, at least 26 epitopes recognized by human TCL could be distinguished. The MBP domain most commonly recognized was sequence 80-105 (31% of MS TCL, and 24% of control TCL). Sequence 29-48 was recognized more frequently by control-derived TCL (24%) than by TCL from MS patients (5%). The MBP epitopes were recognized in the context of DRB1 *0101, DRB5*0101, DRB1*1501, DRB1*0301, DRB1*0401, DRB1*1402, and DRB3*0102, as demonstrated using a panel of DR gene-transfected L cells. The TCR gene usage was also heterogeneous. V beta 5.2, a peptide of which is currently being used in a clinical trial for treatment of MS patients, was expressed by only one of our TCL. However, within this complex pattern of MBP-specific T cell responses, a minority of MS patients were found to exhibit a more restricted response with respect to their TCL epitope specificity. In these patients 75-87% of the TCL responded to a single, patient-specific cluster of immunodominant T cell epitopes located within a small (20-amino acid) domain of MBP. These nested clusters of immunodominant epitopes were noted within the amino acids 80-105, 108-131, and 131-153. The T cell response to the immunodominant epitopes was not monoclonal, but heterogeneous, with respect to fine specificity, TCR usage, and even HLA restriction. In one patient (H.K.), this restricted epitope profile remained stable for > 2 yr. The TCR beta chain sequences of TCL specific for the immunodominant region of HK are consistent with an oligoclonal response against the epitopes of this region (80-105). Further, two pairs of identical sequences were established from TCL generated from this patient at different times (June 1990 and June 1991), suggesting that some TCL specific for the immunodominant region persisted in the peripheral repertoire. The possible role of persistent immunodominant epitope clusters in the pathogenesis of MS remains to be established.