High-throughput UHPLC-MS/MS analysis of Cystatin C proteoforms and creatinine in human sweat for kidney function...

Chronic kidney disease (CKD) is a major global health concern, requiring timely and accurate assessment of kidney function for effective patient management. Creatinine and Cystatin-C (CysC) are well-established biomarkers for estimating glomerular filtration rate (eGFR), a fundamental parameter for CKD monitoring. However, routine clinical assessment relies on invasive blood sampling and analytical methods with limited molecular specificity, particularly for CysC, which is predominantly measured using immunoassays. Sweat has recently emerged as a promising non-invasive biofluid that enables repeated sampling without patient discomfort; however, analytical strategies capable of selectively quantifying CysC and its clinically relevant proteoforms are still lacking. Here, we present the first validated, non-invasive, and patient-friendly protocol for the simultaneous determination of creatinine, intact CysC, and its most abundant circulating proteoform, 3-proline hydroxylated cystatin-C (3prOH-CysC), in human sweat. The workflow integrates iontophoresis-based sweat induction using pilocarpine or carbachol as stimulating agent, sweat collection on bovine serum albumin-pretreated filter paper to minimize analyte adsorption, micro-extraction by packed sorbent sample clean-up and analyte preconcentration, and proteoform-resolved analysis by liquid chromatography-tandem mass spectrometry. The protocol showed near-quantitative recovery (about 85% for native CysC and 3prOH-CysC and about 95% for creatinine), excellent linearity (R² > 0.995), low LOD (0.01 mg/dL for creatinine, 0.4 μg/L for native CysC, and 0.3 μg/L for 3prOH-CysC), and good overall intra- and inter-day precision (RSD <15%). Application of the method to a clinical pilot cohort of stage 3-5 CKD patients demonstrated for the first time the presence of native CysC and 3prOH-CysC as well as two oxidized CysC proteoforms, namely Ox(M14)-CysC and Ox(M41)-CysC. Moreover, significant sweat-blood correlations for creatinine (ρ = 0.62, p = 0.003), native CysC (ρ = 0.56, p = 0.017), and 3prOH-CysC (ρ = 0.75, p < 0.001) were observed. These relationships were independent of age and body mass index. Overall, this study provides proof-of-concept evidence supporting the feasibility of proteoform-resolved LC-MS/MS analysis of sweat for kidney function assessment. The molecular specificity afforded by this approach, beyond that of conventional immunoassays, highlights its potential for biomarkers characterization and warrants further investigation toward patient-centered and longitudinal CKD assessment strategies.