SUBCLINICAL ACUTE KIDNEY DISEASE: URINARY PROTEOMIC DIVERGENCE FOLLOWING NEPHROTOXIC AND ISCHAEMIC INJURY

Abstract

Defective renal repair following acute kidney injury (AKI) may lead to acute kidney disease (AKD), defined by persistent functional and/or structural abnormalities lasting up to three months. However, subclinical AKD often goes undetected due to the lack of specific diagnostic tools, as conventional biomarkers such as plasma creatinine (pCr) may remain within normal limits despite underlying structural damage. To identify urinary proteomic signatures associated with subclinical AKD of differing etiologies, we induced AKI in Wistar rats using either toxic (cisplatin-induced) or ischemic (unilateral ischaemia-reperfusion, I/R) models, and monitored renal function, renal histopathology and urine composition over 42 days. Renal function, as assessed by pCr and creatinine clearance, was rapidly and severely impaired following treatment with cisplatin and I/R, and normalized afterwards within 10 and 2 days, respectively. Despite functional recovery, histological analysis revealed persistent tissue injury--characterized by tubular dilation, necrosis, inflammation, and interstitial fibrosis--particularly in cisplatin-treated animals at day 42. Urinary proteomic analysis identified 53 and 23 proteins that were differentially excreted with the urine in the cisplatin and I/R groups, respectively, compared to controls, and 11 proteins differed from one AKI etiology to the other. Additionally, 29 proteins appeared exclusively post-AKI, irrespective of cause. Most of these proteins are predominantly involved in the immune response, complement activation, haemostasis, and gluconeogenesis. These findings suggest that urinary proteomic fingerprints may serve as sensitive indicators of subclinical AKD, reflecting underlying structural damage even in the absence of overt functional impairment. Such profiles could offer etiology-specific insights, while also enabling the early detection of AKD across diverse injury contexts.