Background: The multidrug resistance (MDR) gene product P-glycoprotein (P-gp) is a transmembrane efflux pump for hydrophobic, potentially toxic compounds, including the immunosuppressant cyclosporine A (CsA). We have previously shown that CsA increases P-gp expression in proximal tubule and endothelial cells in vitro. The aim of the present study was to investigate the in vivo relevance of these observations in renal allograft biopsies from CsA-treated patients.
Methods: P-gp expression was determined by immunohistochemistry of paraffin sections using two different monoclonal antibodies (UIC2 and MRK16). Biopsies were taken from CsA-treated renal transplant patients with different histopathological diagnoses (N = 79) and were compared with biopsies from normal human kidneys (N = 13) or with allograft biopsies from patients under a CsA-free immunosuppression (N = 15). Moreover, biopsies from 10 donor kidneys before implantation and during rejection episodes ("zero biopsies") were investigated.
Results: P-gp expression in biopsies with acute tubular necrosis (ATN; N = 10) after CsA treatment was significantly higher in arterial endothelia, proximal tubules, and epithelial cells of Bowman's capsule (BC), whereas P-gp was sparsely induced in CsA nephrotoxicity (N = 19) compared with controls. Acute cellular (N = 30) and vascular rejection (N = 10) or chronic allograft nephropathy (N = 10) after CsA was associated with strong P-gp expression in infiltrating leukocytes and increased P-gp expression in arterial endothelia, proximal tubules, and BC. In contrast, biopsies of patients treated with a CsA-free immunosuppression regimen did not show increases in P-gp expression compared with controls. Zero biopsies showed a weak, homogeneous, nonpolarized expression of P-gp in tubules and an increased expression of P-gp after CsA therapy in the brush border, arterial endothelia, and BC.
Conclusions: CsA treatment was associated with increased P-gp expression in parenchymal cells of kidney transplants with ATN, acute or chronic transplant rejection, but P-gp was not increased in patients with CsA nephrotoxicity. This indicates that CsA induces its own detoxification by P-gp and that inadequate up-regulation of P-gp in renal parenchymal cells contributes to CsA nephrotoxicity. Increased expression of P-gp in infiltrating leukocytes correlated with the severity of allograft rejection, suggesting that P-gp may decrease the immunosuppressive efficacy of CsA. Thus, individual differences in the P-gp induction response of CsA-exposed renal parenchymal cells and/or infiltrating leukocytes may predispose to either CsA nephrotoxicity or rejection, respectively.