A Spatial Multiomics Atlas and Resource for Immune-Fibrotic Niches in Chronic Kidney Allograft Dysfunction

Abstract

Chronic allograft dysfunction (CAD) is a major cause of late kidney-allograft failure, yet its spatially organized cellular and metabolic mechanisms remain poorly resolved because human biopsies are scarce and experimental models are imperfect. Here, we established a tacrolimus-modulated accelerated mouse model of CAD and constructed a cross-species multiomics atlas by integrating mouse single-cell RNA sequencing, flow cytometry, and multiplex imaging with human single-nucleus RNA sequencing, Visium HD spatial transcriptomics, and spatial metabolomics. Cross-species analysis revealed conserved immune, stromal, epithelial, and myeloid programs and delineated 2 anatomically distinct fibrotic niches. Cortical fibrosis was associated with endothelial/pericyte remodeling, altered cholesterol-biosynthesis transcriptional programs, and reduced regional cholesterol-related signals. Medullary fibrosis was enriched for proximal-tubule S3 [PT(S3)] cells with glycolysis-associated remodeling, increased regional lactate-related signals, and adjacent tissue-resident memory T-cell domains displaying an atypical exhaustion-like program. Glutamine/glutamate metabolism-associated programs accompanied inferred monocyte-to-macrophage state remodeling. In renal tubular epithelial cells, 2-deoxy-D-glucose partially attenuated transforming growth factor-β1-induced E-cadherin loss and α-smooth muscle actin induction, providing focused functional support for glycolysis-associated tubular remodeling. We additionally provide an exploratory kidney-transplant multiomics resource for data visualization, auxiliary summarization of platform-derived results, and gene/pathway querying. Together, this study provides a spatially resolved, cross-species framework for identifying candidate immune-metabolic mechanisms and prioritizing experimentally testable targets in CAD.