Urinary Clusterin is a Biomarker of Renal Epithelial Senescence and Predicts Human Kidney Disease Progression

Cellular senescence is characterized by generally irreversible cell cycle arrest and changes in secretory activity, with senescent renal epithelia proposed as drivers of kidney fibrosis. The lack of noninvasive biomarkers represents an obstacle to the design of human trials of senescent cell-depleting medications.

Proteomic analysis was performed on urine from patients with chronic kidney disease (CKD) alongside immunofluorescence staining of paired kidney biopsies (n = 51). Enzyme-linked immunosorbent assays (ELISAs) and immunofluorescence staining were performed in a second cohort of matched urine and kidney biopsies (n = 53). Spatial transcriptomic analysis was performed on kidney tissue from benign and fibrotic kidney disease (n = 13). Clusterin and senescence markers were analyzed in vitro by quantitative polymerase chain reaction (PCR) in irradiated human renal epithelia. Urinary biomarker concentrations were quantified by ELISA (n = 322) to assess their ability to predict patient outcomes (end-stage kidney disease or > 40% renal functional loss).

P21⁺Ki67^- epithelial senescence correlated with age and inversely with renal function. Urinary clusterin-to-creatinine ratio (uCCR) correlated tightly with P21⁺Ki67^- epithelial senescence in both matched urine and kidney biopsy cohorts (rho > 0.5, P < 0.001) and predicted levels of senescence after adjusting for other variables. Clusterin was upregulated transcriptomically in CDKN1A (p21) expressing epithelia in vitro and in vivo. An elevated uCCR predicted adverse renal end points in a cohort of patients with CKD after adjusting for baseline estimated glomerular filtration rate (eGFR), urinary albumin-to-creatinine ratio (uACR), age, systolic blood pressure, and sex.

uCCR represents a surrogate for histologic quantification of p21⁺Ki67^- senescent renal epithelia and predicts outcomes in human kidney disease independent of existing clinical risk factors.