Renal allograft failure : a study of the drivers of epithelial cell de-differentiation

Abstract

Kidney transplantation is the gold standard renal replacement therapy for patients with end-stage renal disease. Despite advances in immunosuppressive therapy, chronic allograft dysfunction remains the commonest cause of renal allograft failure in living recipients. The typical pathology of this disease includes chronic inflammation with tubular atrophy and interstitial fibrosis. Although the origin of the excess fibroblasts and myofibroblasts remains controversial, the process of epithelial to mesenchymal transition might play a role. This study was designed to test the linked hypotheses that the immunosuppressive drug, Cyclosporine A and graft infiltrating T cells can induce allograft pathology by alteration of the bioavailability of fibrogenic TGF-β. An initial series of experiments examined the induction of mesenchymal transition by treatment of cultured human renal tubular epithelial cells with immunosuppressive concentrations of Cyclosporine A. Drug treated cells showed characteristic morphological changes associated with increased expression of the mesenchymal marker S100A4 and reduced expression of the epithelial marker E-cadherin; similar changes were induced by the addition of TGF-β1. The phenotypic change induced by Cyclosporine A was not the consequence of an increased response to autocrine TGF-β and could not be inhibited by specific blockade of the ALK5 component of the TGF-β receptor. Further studies showed in vitro that contact between activated T cells and renal tubular epithelial cells could induce mesenchymal transition by a mechanism that was dependent on activation of the TGF-β receptor complex. A final series of experiments defined a mechanism by which T cells activate latent TGF-β allowing subsequent receptor stimulation leading to either T cell or epithelial cells differentiation. The latency associated peptide binds to and inhibits native TGF-β but can be displaced by both thrombospondin-1 and neuropilin-1, producing active TGF-β. In this study it was shown that cytoplasmic thrombospondin-1 is exported and expressed on the surface of activated T cells following brief interaction with renal tubular epithelial cells; neuropilin-1 was also expressed by a mean 18% of activated human T cells. Inhibition of these two molecules with a blocking LSKL peptide sequence inhibited the normal response of activated human T cells to latent TGF-β1. This study demonstrated that both Cyclosporine A and T cells can induce renal epithelial to mesenchymal cell transition. However, the former process seems independent of TGF-β whilst the latter requires TGF-β receptor stimulation and might be regulated in vivo by T cell-mediated activation of latent TGF-β within the allograft.