Circulating CD8+CD28^null T Cytotoxic Cells in Polymyositis—A Possible Biomarker? Comment on the Article by Pandya et al

Abstract

We read with interest the recent report by Pandya et al on the cytotoxic effects of CD28 null T cells on myocytes from patients with polymyositis (PM) (1). They cocultured autologous muscle cells with purified peripheral blood CD81 and CD41 T cells from 5 patients with PM and showed that CD28 null T cells exerted myotoxic activity by increasing the secretion of granzyme B, whereas CD281 T cells did not. The activation of this intracellular pathway was further corroborated by the results of experiments in which perforin polarization was down-regulated by treatment with selective inhibitors.Many years ago we also focused on this topic, investigating whether CD81 T cytotoxic cells circulating in the peripheral blood might differentiate PM from dermatomyositis (DM) (2). By that time in the early 1990s, reported findings from some studies suggested that CD28 antigen expression on the cell surface might identify T cells with cytotoxic functions (3,4). Therefore, in our study we investigated, by flow cytometry, for possible differences in CD81CD281 circulating T cells between PM patients (n 5 12), DM patients (n 5 10), and compared to healthy donors (n 5 16). We found meaningful differences in the distribution of CD81 T cell subsets across groups, but we misinterpreted the implications of our results since we started with the assumption that the CD81 CD281 T cells were the cytotoxic ones (2). In light of more recent studies providing evidence that the lack of CD28 surface antigen defines T cytotoxic cells (5,6) and after reading the article by Pandya et al (1), we reviewed our old data.PM patients displayed the highest percentage of peripheral blood CD81CD28 null T cells (mean 6 SD 13.1 6 2%, versus 5.8 6 1% in patients with DM [P , 0.05] and 6.4 6 2% in healthy controls [P , 0.05]). Similar differences were also detected when absolute numbers of CD81CD28 null T cells were measured. We also investigated the state of functional activation, namely, the migratory phenotype of CD81 T cells, by assessing the coexpression of CD25 and HLA-DR surface antigen, which are up-regulated upon T cell activation, and lymphocyte function-associated antigen 1 (LFA-1), which promotes migration of activated T cells through endothelial cells into extravascular tissue, and adhesion to the target cells (7). Compared to healthy donors, CD81HLA-DR1 T cells were significantly increased in myositis patients, but with no significant difference between the PM and DM groups (mean 6 SD 16.3 6 2% and 16.0 6 1%, respectively).Upon T cell activation, LFA-1 shows a bimodal expression, and a "dim" or "bright" LFA-1 phenotype can be easily recognized based on mean fluorescence intensity seen on flow cytometry. Indeed, the proportion of CD81LFA-1 bright T cells was significantly higher among patients with PM (mean 6 SD 73.4 6 7%) than among patients with DM (45.5 6 9%) or healthy controls (40.2 6 10%) (both P , 0.05), suggesting that activated peripheral blood CD81 T cells are more prone to leak from blood streaming into PM muscles. To investigate this wor...