Non-invasive quantitative assessment of kidney injury using near-infrared autofluorescence imaging

Introduction Tubulointerstitial fibrosis represents the final common pathway to kidney failure, with tubular epithelial cell (TEC) injury playing a critical role. Current methods for assessing TEC injury rely on invasive biopsies or lack temporal resolution, limiting real-time monitoring of disease progression and therapeutic responses. Methods Near-infrared (NIR) autofluorescence was evaluated for non-invasive TEC injury assessment in three mouse models: unilateral ureteral obstruction (UUO; consistent tubulointerstitial injury), folic acid nephropathy (variable injury severity), and unilateral ischemia-reperfusion injury (IRI; acute tubular injury), without exogenous probes. Mechanistic studies used BALB/c- Cpox nct mice, which harbor a spontaneous missense mutation reducing coproporphyrinogen oxidase activity, an enzyme that metabolizes coproporphyrinogen III in heme biosynthesis. Kidney slice cultures were treated with the mitochondrial complex III inhibitor antimycin A or the mitochondrial reactive oxygen species scavenger MitoTEMPO. Human kidney samples (non-cancerous region of nephrectomies) were examined ex vivo . Results NIR autofluorescence correlated with TEC injury, oxidative stress markers, and fibrosis across all three models. Spectroscopic analysis identified coproporphyrin III as the likely primary fluorophore. Cpox mRNA and CPOX protein were consistently reduced in injured kidneys. CPOX was expressed in proximal tubules and lost in damaged cells. BALB/c- Cpox nct mice showed elevated baseline autofluorescence, which increased further after UUO, with HPLC confirming accumulation of the porphyrin metabolite coproporphyrin III. In kidney slice cultures, antimycin A reduced CPOX protein and enhanced autofluorescence; MitoTEMPO reversed both effects. To test whether autofluorescence can track therapeutic responses, finerenone (non-steroidal, mineralocorticoid receptor antagonist) was administered in UUO and bilateral IRI models; treated mice showed reduced autofluorescence that correlated with improved histology. In 34 human samples, autofluorescence correlated inversely with eGFR and positively with periodic acid-Schiff stain injury scores. Conclusions NIR autofluorescence enables non-invasive, real-time assessment of kidney injury without external agents. This approach has immediate applications for longitudinal monitoring in preclinical research demonstrating proof-of-concept for future development of clinically applicable imaging approaches.