#6014 SINGLE-CELL SPATIAL TRANSCRIPTOMICS OF THE KIDNEY IN HEALTH AND DISEASE DEFINES INJURY-SPECIFIC DOMAINS AND IDENTIFIES NOVEL CELL-CELL INTERACTIONS

Abstract

Abstract Background and Aims Single-cell sequencing has revealed an unexpected diversity of cell types throughout the body. However, the loss of spatial context in many single-cell sequencing techniques hampers our understanding of cell-cell interactions, which are central to almost all (patho-)physiological processes, particularly in highly complex organs such as the kidney. In this study, we aim to profile the spatial organization of, and cell-cell interactions within, the healthy and diseased mouse kidney with single-cell resolution. We focus on changes induced by acute kidney injury (AKI), since AKI is highly prevalent and can progress to chronic kidney disease (CKD), with no targeted treatment strategy to prevent this AKI-to-CKD transition existing to date. Method To model ischemic AKI in the mouse, we performed bilateral ischemia-reperfusion injury on C57BL/6J mice. Kidneys were collected in the transition phase from AKI to CKD at 4 weeks after AKI (n = 3). Kidneys from non-surgery mice were used as controls (n = 3). To characterize the spatial complexity of the kidney with single-cell resolution we used seqFISH+, a sequential fluorescence in situ hybridization approach that allows the multiplexing of thousands of genes by sequential hybridization and confocal imaging, thus enabling RNA-quantification at the single-transcript level. Results Quantifying and clustering 1300 genes in 230,422 cells identified all major cell types of the kidney and revealed changes in the kidneys’ cellular composition at 4 weeks post AKI: While the cortical vasculature and cells of the proximal tubule segment 3 were reduced, the abundance of injured proximal tubule cells (identified by Havcr1 and Vcam1 expression), immune cells and fibroblasts increased. Clustering the identified cell types on their neighbors within a 30um radius revealed distinct spatial domains comprising different combinations of cell types. Certain domains were consistently present across samples and corresponded to the known regional organization of the kidney, thus providing an internal validation. Other domains developed de novo after AKI, illustrating the structural changes induced by AKI. One of these injury-specific domains was comprised of injured proximal tubule cells, macrophages and fibroblasts, while another mainly comprised macrophages, dendritic cells and T cells. Ligand-receptor analysis within domains revealed novel cell-cell interactions, for example highlighting a Crlf1-expressing fibroblast population in close proximity to injured proximal tubule cells, which express the interleukin-6 cytokine Clcf1, a usually co-secreted binding partner of Crlf1. Zooming in on one injury-specific domain, we found that the fraction of fibroblasts and macrophages in the vicinity of injured proximal tubule cells increases with the degree of injury-related gene expression in these cells, suggesting a causal role of injured proximal tubule cells in defining a “pathogenic niche” associated with kidney disease progression. Conclusion This study provides a spatial characterization of the kidney with unprecedented resolution, highlights AKI-induced structural changes in the kidney, defines injury-specific spatial domains and reveals cell-cell-interactions relevant to disease progression from AKI to CKD.