Abstract Background ChRCC is a rare form of kidney cancer and presents with a poor prognosis in the metastatic setting, with limited response to immune checkpoint inhibitors (ICI) and targeted therapy. While previous studies suggested that ChRCC originates from intercalated cells of the kidney, its exact cellular origin has not been clearly defined. We therefore investigated the cellular origin of ChRCC and renal oncocytic neoplasms at single-cell resolution. Methods ChRCC, renal oncocytoma (RO) and low-grade oncocytic tumor (LOT) samples with matched normal kidney specimens were analyzed using single-cell RNA sequencing (scRNA-seq). Epithelial cells from normal samples were clustered and annotated into the distinct known cellular types found in the adult healthy human kidney. A logistic regression model (Young M.D. et al., Science, 2018) was trained on the normal epithelial clusters, using a comprehensive set of 74 marker genes (features). Subsequently, the model was tested on tumor clusters from ChRCC, RO and LOT to investigate their cellular origin through the identification of the highest predicted probabilities of similarity between normal epithelial cellular types and tumor cells. Differential gene expression and pathway analyses between ChRCC and its cell of origin were then conducted to investigate mechanisms of oncogenesis. Results Following quality-control, 46,817 cells from 5 tumors (ChRCC: n=3, RO: n=1 and LOT: n=1) and 4 normal samples were isolated for scRNA-seq analysis. Among normal samples, 784 epithelial cells were clustered and annotated into the following cellular entities: proximal tubule (PT), loop of Henle – distal tubule (LOH-DT), principal cells (PC), α-intercalated cells (ICA) and β-intercalated cells (ICB). Across all ChRCC and oncocytic tumors, 7,573 tumor cells were identified, among which 7,425 corresponded to ChRCC. For ChRCC, RO, and LOT neoplasms, the highest predicted probability of similarity was consistently identified with ICA cells (ranging from 0.60 to 0.80). This finding was validated using an external and previously published (Zhang Y. et al., Proc Natl Acad Sci USA, 2021) scRNA-seq dataset from a patient with ChRCC (n=2,853 cells). Additionally, evaluation of previously defined ChRCC-specific markers (i.e. FOXI1, RHCG, KIT, CLCNKA, and CLCNKB) in scRNA-seq data of normal kidney epithelial cells consistently showed the highest expression in ICA cells. Differential gene expression between ChRCC and ICA cells revealed a significant downregulation of MHC class I genes (i.e. HLA-A , HLA-B, and HLA-C) and HSP70 (family A) genes (i.e. HSPA1A, HSPA1B, HSPA6, and HSPA8) in ChRCC. Differential pathway analysis between ChRCC and ICA cells showed a significant enrichment of the ferroptosis, glutathione metabolism, mTOR signaling and IL-15 pathways in ChRCC. Conclusions Renal oncocytic tumors, including ChRCC, appear to originate from the α-intercalated cells of the normal human kidney. As compared to α-intercalated cells, ChRCC downregulates MHC class I genes, which may be associated with its poor response to immunotherapy. Further, the selective downregulation of HSP70 genes may in part explain ChRCC’s increased sensitivity to ferroptosis. CDMRP DOD Funding: yes