Résumé :
|
Myology textbooks describe muscle fibroblasts on the grounds of EM studies as independent cells and envisage them only as collagen-producing cells. Lack of muscle fibroblast marker likely represents one cause of our ignorance about this cell subset. While studying resident muscle MPs, we identified CD34 as a reliable marker of muscle fibroblasts. CD34 immunostaining and electron microscopy identified muscle and fascia fibroblasts as reticular cells (FRCs) with very long, extremely thin, and interconnected cell processes. Sorted CD34+ cells were negative for hematopoietic (CD45), myeloid (CD11b), endothelial (CD31) and human satellite cell (CD56) markers. Immunostaining for Cx43 identified gap junctions at the level of FRC interconnexions, both in vivo and in vitro. They appeared permeable to small Dye molecules as assessed by the scrape-loading method. Elisa multiplex assay for 16 cytokines (LuminexTM), performed on culture supernatants of FRCs collected from intact or notexin-injured muscle, displayed selective and strongly activable production of KC, IL5, MCP1, and IL6, suggesting FRC pro-migratory and/or differentiating effects on neutrophils, eosinophils, monocytes, and B-cells, respectively. In contrast, muscle FRC produced no IL2, and, at steady state, expressed very little MHC-1 molecule in vivo and in vitro (6%, likely reflecting activation during cell sorting). FRC expression of MHC-1 molecules was upregulated by IFNg in vitro (up to 40%) and by an inflammatory milieu in vivo (myositis). FRCs selectively induced moderate CD8+ T-cell proliferation in mixed leucocyte reaction (MLR) tests with allo>auto MLR efficiency suggesting antigen presenting cell function by the minor population of MHC-1+ FRCs. The effect was independent of IL15, as demonstrated by allo-MLRs performed in presence of blocking anti-IL15 Abs and by using FRCs sorted from IL-15-/- mice. In our in vitro study reflecting situation of a nearly homeostatic tissue, the amplified CD8+ T-cells exhibited a poorly activated phenotype (with a minority of CD44hiCD62Llo , PD1+, CD69+, Ly6C+ cells) and had downregulated their TNFa (1.4 vs 57%) and IFNg (4 vs 28%) production in presence of FRCs. Both immunophenotyping and functional tests showed no implication of FRCs in Treg amplification.In conclusion, interconnected muscle FRCs may play a sentinel role by sensing danger signals and releasing a selective set of inflammatory cytokines. Like myofibers, they poorly express MHC-1 molecules at steady state but can upregulate these molecules upon stimulation and elicit selective CD8+ T-cell proliferation. Contribution of FRCs in pathophysiology of eosinophilic fasciitis and polymyositis deserves further investigations.
|