Urine peptides as potential biomarkers of graft lesions in kidney graft recipients

Anne-Laure Gagez¹, Pierre Marquet^1,7, Yann Le Meur^2,3, Nalia Taguiera⁴, François Bayle⁵, Elisabeth Cassuto-Viguier⁶, Michel Delahousse⁸, Louis N. Gastinel¹.

¹INSERM UMR-S850, France, University of Limoges; ²Department of Nephrology-Transplantation, CHU Limoges; ³Department of Nephrology-Transplantation, CHU Brest; ⁴Roche Pharma; ⁵Department of Nephrology-Transplantation, CHU Grenoble; ⁶Department of Nephrology-Transplantation, CHU Nice; ⁷Department of Pharmacology, Toxicology and Pharmacovigilance, CHU Limoges; ⁸Department of Nephrology-Transplantation, CHU Suresnes; France.

Non-invasive monitoring of the graft using urine biomarkers in kidney transplant recipients can lead to the early diagnosis of renal dysfunction, helping to avoid or reduce the use of renal allograft biopsies and improve graft survival. Our pilot study was designed as an ancillary study of a large French multicenter clinical trial in kidney transplant patients followed over one year post-transplantation. It involved systematic biopsies at M3 and M12, immediately preceded by urine sample collection, and biopsies with paired urine samples in case of clinical suspicion of acute rejection at any time. The natural urinary peptidome of 57 adult kidney transplant patients was analyzed using nanoHPLC coupled off-line with MALDI TOF-TOF mass spectrometry. 22 presented no sign of rejection or chronic nephropathy (ST for stable), 17 with histological signs of chronic nephropathy grade 1 (CAN1) and 18 with histological signs of acute rejection (AR). Patients were randomly assigned to a training (n=38) or a test set (n=19). Univariate and multivariate statistical analyses were used in the training set to identify, among 39 sequenced peptides, those with the most discriminative relative abundance. We validated further in the test set: an association of 2 peptides (collagen a1-V[1525-1541] and collagen XV[1092-1108]) discriminating RA and non-RA, with sensitivity (Sn) = 100% and specificity (Sp) = 66%; 2 peptides able to discriminate CAN1 and non-CAN1 (collagen a1-V[1525-1541] and uromodulin[567-584]), with Sn = 67% and Sp = 92%; and one peptide (collagen a1-V[1526-1539]) discriminating ST and non-ST patients, with Sn = 100% and Sp = 50%. Studies are ongoing in order to verify the discriminative power of these peptides in a larger sample cohort by using SRM methods, and to study other analytes yet not sequenced by using data mining and statistical learning methods.