Transcriptomes of Major Renal Collecting Duct Cell Types in Mouse
Total Page:16
File Type:pdf, Size:1020Kb
Transcriptomes of major renal collecting duct cell types PNAS PLUS in mouse identified by single-cell RNA-seq Lihe Chena, Jae Wook Leeb, Chung-Lin Choua, Anil V. Nairc, Maria A. Battistonec, Teodor G. Paunescu˘ c, Maria Merkulovac, Sylvie Bretonc, Jill W. Verlanderd, Susan M. Walle,f, Dennis Brownc, Maurice B. Burga,1, and Mark A. Kneppera,1 aSystems Biology Center, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD 20892; bNephrology Clinic, National Cancer Center, Goyang, 10408, South Korea; cCenter for Systems Biology, Program in Membrane Biology and Division of Nephrology, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114; dDivision of Nephrology, Hypertension and Renal Transplantation, University of Florida College of Medicine, Gainesville, FL 32610; eDepartment of Medicine, Emory University School of Medicine, Atlanta, GA 30322; and fDepartment of Physiology, Emory University School of Medicine, Atlanta, GA 30322 Contributed by Maurice B. Burg, October 3, 2017 (sent for review June 22, 2017; reviewed by Andrew P. McMahon and George J. Schwartz) Prior RNA sequencing (RNA-seq) studies have identified complete count for a small fraction of the kidney parenchyma. Therefore, transcriptomes for most renal epithelial cell types. The exceptions methods were required for selective enrichment of the three cell are the cell types that make up the renal collecting duct, namely types from mouse kidney-cell suspensions. Here, we have identi- intercalated cells (ICs) and principal cells (PCs), which account for fied cell-surface markers for A-ICs, B-ICs, and PCs, allowing these only a small fraction of the kidney mass, but play critical physio- cell types to be enriched from kidney-cell suspensions by using logical roles in the regulation of blood pressure, extracellular fluid FACS. We used the resulting enrichment protocols upstream from volume, and extracellular fluid composition. To enrich these cell microfluidic-based scRNA-seq to successfully identify transcriptomes types, we used FACS that employed well-established lectin cell of all three cell types. These three transcriptomes have been surface markers for PCs and type B ICs, as well as a newly identified permanently posted online to provide a community resource. Our cell surface marker for type A ICs, c-Kit. Single-cell RNA-seq using the bioinformatic analysis of the data addresses the possible roles of IC- and PC-enriched populations as input enabled identification of A-IC–, B-IC–, and PC-selective genes in regulation of renal complete transcriptomes of A-ICs, B-ICs, and PCs. The data were transport, total body homeostasis, and renal pathophysiology. PHYSIOLOGY used to create a freely accessible online gene-expression database for collecting duct cells. This database allowed identification of Results genes that are selectively expressed in each cell type, including cell- Single-Tubule RNA-Seq in Microdissected Mouse Cortical Collecting surface receptors, transcription factors, transporters, and secreted Ducts. To provide reference data for interpretation of scRNA- proteins. The analysis also identified a small fraction of hybrid cells seq experiments in mouse, we have carried out single-tubule expressing aquaporin-2 and anion exchanger 1 or pendrin tran- RNA-seq in cortical collecting ducts (CCDs) rapidly micro- scripts. In many cases, mRNAs for receptors and their ligands were dissected from mouse kidneys without protease treatment. Data identified in different cells (e.g., Notch2 chiefly in PCs vs. Jag1 chiefly were highly concordant among 11 replicates from seven different in ICs), suggesting signaling cross-talk among the three cell types. The identified patterns of gene expression among the three types of Significance collecting duct cells provide a foundation for understanding physi- ological regulation and pathophysiology in the renal collecting duct. A long-term goal in mammalian biology is to identify the genes expressed in every cell type of the body. In the kidney, the systems biology | intercalated cell | principal cell | kidney expressed genes (i.e., transcriptome) of all epithelial cell types have already been identified with the exception of the cells that hole-body homeostasis is maintained in large part by make up the renal collecting duct, which is responsible for reg- Wtransport processes in the kidney. The transport occurs ulation of blood pressure and body fluid composition. Here, along the renal tubule, which is made up of multiple segments single-cell RNA-sequencing was used in mouse to identify tran- consisting of epithelial cells, each with unique sets of transporter scriptomes for the major collecting duct cell types: type A in- proteins. There are at least 14 renal tubule segments containing tercalated cells, type B intercalated cells, and principal cells. The at least 16 epithelial cell types (1, 2). A systems-level under- information was used to create a publicly accessible online re- standing of renal function depends on knowledge of which source. The data allowed identification of genes that are selec- protein-coding genes are expressed in each of these cell types. tively expressed in each cell type, which is informative for cell- Most renal tubule segments contain only one cell type, and the level understanding of physiology and pathophysiology. genes expressed in these cells have been elucidated through the application of RNA sequencing (RNA-seq) or serial analysis of Author contributions: L.C., J.W.L., S.M.W., D.B., M.B.B., and M.A.K. designed research; L.C., gene expression applied to microdissected tubules from rodent J.W.L., C.-L.C., A.V.N., M.A.B., T.G.P., M.M., and J.W.V. performed research; S.M.W., D.B., and kidneys (2, 3), which identify and quantify all mRNA species M.A.K. contributed new reagents/analytic tools; L.C., J.W.L., C.-L.C., A.V.N., M.A.B., T.G.P., (i.e., transcriptomes) expressed in them. The exception is the M.M., S.B., J.W.V., S.M.W., D.B., M.B.B., and M.A.K. analyzed data; and L.C., J.W.L., C.-L.C., A.V.N., M.A.B., T.G.P., M.M., S.B., J.W.V., S.M.W., D.B., M.B.B., and M.A.K. wrote the paper. renal collecting ducts, which are made up of at least three cell Reviewers: A.P.M., University of Southern California; and G.J.S., University of Rochester types, known as type A intercalated cells (A-ICs), type B in- Medical Center. tercalated cells (B-ICs), and principal cells (PCs). Single-tubule The authors declare no conflict of interest. RNA-seq applied to collecting duct segments provides an ag- gregate transcriptome for these three cell types. Hence, to Published under the PNAS license. identify separate transcriptomes for A-ICs, B-ICs, and PCs, it is Data deposition: The sequences and metadata reported in this paper have been depos- ited in the Gene Expression Omnibus (GEO) database, https://www.ncbi.nlm.nih.gov/geo necessary to carry out RNA-seq at a single-cell level. Recent (accession no. GSE99701). advances in single-cell RNA-seq (scRNA-seq) have facilitated 1To whom correspondence may be addressed. Email: [email protected] or knepperm@ our understanding of heterogeneous tissues like brain (4), lung nhlbi.nih.gov. (5), pancreas (6), and retina (7). However, a barrier to success This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. with such an approach exists because collecting duct cells ac- 1073/pnas.1710964114/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1710964114 PNAS | Published online October 31, 2017 | E9989–E9998 Downloaded by guest on October 5, 2021 untreated mice (Dataset S1). The single-tubule RNA-seq data for A + - CD+ + mouse CCDs are provided as a publicly accessible web page GFP :GFP c-Kit :DBA DBA+:c-Kit+ (https://hpcwebapps.cit.nih.gov/ESBL/Database/mTubule_RNA-Seq/). Asb15 Slc4a1 Ptgds Among the most abundant transcripts in mouse CCDs are those Hepacam2 Oxgr1 Fxyd4 typical of PCs (e.g., Aqp2, Aqp3, Scnn1b, Scnn1g, Kcnj1,and Rhbg Cpsf4l Aqp2 Atp6v0d2 Kit Npnt Avpr2) and ICs (e.g., Car2, Slc4a1, Slc26a4, Rhcg,andAtp6v1b1). Aqp6 Aqp6 Apela Wscd2 Dmrt2 Avpr2 Identification of Cell-Surface Markers for ICs. To identify potential Slc4a9 Tmem61 Mcoln3 cell-surface marker proteins that can be used for FACS enrich- Slc26a4 Avpr1a Tmem45b ment of ICs, we used transgenic mice that express GFP driven by + Avpr1a Atp6v1g3 C1qa the promoter for the B1 subunit of the H -ATPase (Atp6v1b1) (8). Atp6v1g3 Hepacam2 Hsd11b2 Atp6v1b1 is known to be expressed in A- and B-ICs and is abun- Foxi1 Ociad2 C1qc dant in rat connecting tubule (CNT), CCD, and outer medullary Atp6v1c2 Adgrf5 Ptgs1 collecting duct (2), the segments that contain ICs. We used en- Spink8 Atp6v0d2 C1qb zymatic tissue dissociation and FACS to enrich GFP-expressing Nefl Guca2a Kcne1 + Atp6v1c2 Csf1r (GFP ) cells and carried out RNA-seq to quantify mRNA abun- Kit + − Dagla Mme Apoe dance levels for all expressed genes in GFP -cells vs. GFP -cells. + − Insrr Slc4a9 Rnf186 Fig. 1A shows the 24 transcripts with GFP :GFP mRNA ex- Slc4a1 Bsnd Rasl11a pression ratios greater than 50 based on two pairs of samples Serpinb9 Aldh1l1 Scin isolated on different days (full listing of ratios is provided in Slc8a1 Rcan2 Tmem229a Dataset S2). Consistent with the idea that these are IC-selective Pkib Foxi1 Trf genes, 12 of 24 of the transcripts in Fig. 1A are already widely Bglap3 Serpinb9 Tacstd2 Slc26a7 Slc2a4 Kcnj1 known to be expressed in ICs (shown in boldface). Notably, there Gcgr Scnn1g are two transcripts that code for potential cell surface marker Ccbe1 proteins, specifically Hepacam2 and Kit (also known as c-Kit). 0 800 060015 Both are integral membrane proteins with long extracellular N- terminal regions (i.e., type I membrane proteins).