Transcript Deconvolution Reveals Heterogeneous Cell Populations in Microdissected Nephron Segments of the Rat Kidney

Abstract

Introduction: The kidney tubule comprises distinct segments with specialized epithelium, including proximal tubules (PT) subsegments S1, S2, and S3, descending (DTL) and ascending (ATL) thin limbs, thick ascending limb (TAL) medullary (MTAL) and cortical (CTAL) subsegments, distal convoluted tubule (DCT), connecting tubule (CNT), and collecting ducts (CD) containing principal (PC) and intercalated (IC) cells. Single-cell RNA sequencing (scRNAseq) is a crucial tool in kidney research. However, it assigns cells to nephron segments based on transcriptome similarity, disregarding anatomical context and spatial information, critical for kidney function. Hypothesis: We hypothesized that cellular deconvolution of microdissected tubule segment transcriptomes would identify mixed cell populations coexisting on different parts of the nephron. Methods: We employed CibersortX to estimate cell proportions in microdissected rat nephron segment transcriptomes (GSE56743), using rat kidney epithelial cell transcriptomes from a scRNAseq dataset (GSE137869) as reference. Raw sequencing files from the Sequence Read Archive (SRA) were downloaded and reprocessed. The number of detected genes was 34,322 for bulk RNAseq and 15,282 for scRNAseq, with a shared gene space of 13,373 used for deconvolution. Results: We discovered that PT.S1 segments primarily consist of S1 cells (81%) and some S2 cells (12%). Microdissected PT.S2 segments contain a mix of S1 (18%), S2 (58%), and S3 (19%) cells. PT.S3 segments were primarily composed of S3 cells (75%). DTL and ATL segments displayed a high percentage of the corresponding cell type. Microdissected MTAL segments were a mix of MTAL (25%) and CTAL (69%) cells, while microdissected CTAL segments were almost exclusively composed of MTAL cells (95%). DCT segments exhibited 80% DCT cells with some principal (2%) and intercalated (1%) cells. In CNT, CCD, and OMCD, we observed a continuous increase in PC abundance (58%, 72%, and 78%, respectively) along with a decrease in IC (32%, 23%, and 11%, respectively). Microdissected IMCD contained PC (58%), IC (4%), and 18% of a distinct cell type with IC characteristics not found in other segments. Conclusions: Our analysis supports the well-established cell heterogeneity in connecting tubules and the collecting duct system, including the existence of a specialized IMCD cell. Additionally, we identified mixed cell populations in the S2 segment of the proximal tubule and the medullary thick ascending limb. In the case of S2, the high proportion of cells from adjacent S1 and S3 segments suggests a differentiation continuum from S1 to S3 with diffused anatomical boundaries. We identified two different TAL cell types, the MTAL phenotype restricted to the MTAL and the CTAL phenotype present in higher abundance in both subsegments. These cell types may correspond to the medullary "rough surface" and cortical "smooth surface" cells described in rat TAL. In summary, the terminology used in scRNAseq cluster assignment referring to specific nephron segments is misleading, as it encompasses anatomical regions with multiple coexisting cell types. This highlights the need for caution when interpreting such data. This work was supported in part by CWRU BGT670107 and NIH-NIDDK DK128304 grants to AGV. This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.