Single-Cell RNA Sequencing Reveals Renal Endothelium Heterogeneity and Metabolic Adaptation to Water Deprivation

BASIC RESEARCH www.jasn.org

Sébastien J. Dumas,1,2 Elda Meta,1,2 Mila Borri,1,2 Jermaine Goveia,1,2 Katerina Rohlenova ,1,2 Nadine V. Conchinha,1,2 Kim Falkenberg,1,2 Laure-Anne Teuwen,1,2 Laura de Rooij,1,2 Joanna Kalucka,1,2 Rongyuan Chen,3 Shawez Khan,1,2 Federico Taverna,1,2 Weisi Lu,3 Magdalena Parys,1,2 Carla De Legher,1,2 Stefan Vinckier,1,2 Tobias K. Karakach ,1,2 Luc Schoonjans,1,2,3 Lin Lin,4,5 Lars Bolund,4,5 Mieke Dewerchin,1,2 Guy Eelen,1,2 Ton J. Rabelink,6 Xuri Li,3 Yonglun Luo,4,5,7,8 and Peter Carmeliet1,2,3

Due to the number of contributing authors, the afﬁliations are listed at the end of this article.

ABSTRACT Background Renal endothelial cells from glomerular, cortical, and medullary kidney compartments are exposed to different microenvironmental conditions and support specific kidney processes. However, the heterogeneous phenotypes of these cells remain incompletely inventoried. Osmotic homeostasis is vitally important for regulating cell volume and function, and in mammals, osmotic equilibrium is regulated through the countercurrent system in the renal medulla, where water exchange through endothelium occurs against an osmotic pressure gradient. Dehydration exposes medullary renal endothelial cells to extreme hyperosmolarity, and how these cells adapt to and survive in this hypertonic milieu is unknown. Methods We inventoried renal endothelial cell heterogeneity by single-cell RNA sequencing .40,000 mouse renal endothelial cells, and studied transcriptome changes during osmotic adaptation upon water deprivation. We validated our findings by immunostaining and functionally by targeting oxidative phosphorylation in a hyperosmolarity model in vitro and in dehydrated mice in vivo. Results We identified 24 renal endothelial cell phenotypes (of which eight were novel), highlighting extensive heterogeneity of these cells between and within the cortex, glomeruli, and medulla. In response to dehydration and hypertonicity, medullary renal endothelial cells upregulated the expression of genes involved in the hypoxia response, glycolysis, and—surprisingly—oxidative phosphorylation. Endothelial cells increased oxygen consumption when exposed to hyperosmolarity, whereas blocking oxidative phosphorylation compromised endothelial cell viability during hyperosmotic stress and impaired urine concentration during dehydration. Conclusions This study provides a high-resolution atlas of the renal endothelium and highlights extensive renal endothelial cell phenotypic heterogeneity, as well as a previously unrecognized role of oxidative phosphorylation in the metabolic adaptation of medullary renal endothelial cells to water deprivation.

JASN 31: 118–138, 2020. doi: https://doi.org/10.1681/ASN.2019080832

The kidney is vital for organismal homeostasis, and diaphragm) to facilitate water passage but also to has a complex vasculature to achieve this task. restrict ﬁltration of high-molecular-weight sub- Within the kidney, renal endothelial cells (RECs) stances.3 By contrast, cortical renal endothelial cells differ according to their origin and vascular bed.1,2 (cRECs) in larger vessels (arteries, arterioles) and For instance, renal endothelial cells within glo- medullary renal endothelial cells (mRECs) in the meruli (gRECs) are fenestrated (but without descending vasa recta (DVR) have a continuous

118 ISSN : 1046-6673/3101-118 JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH lining, and are surrounded by pericyte/smooth muscle cells, Significance Statement which can regulate GFR in the cortex and osmolarity gradient preservation in the medulla. In addition, the parallel arrange- The specialized vessels comprising the renal vasculature are char- ment of the DVR and the ascending vasa recta (AVR) in the acterized by highly differentiated renal endothelial cell types, but medulla causes an oxygen shunt, resulting in hypoxia in the this heterogeneity has been poorly inventoried. Using single-cell RNA sequencing, the authors developed a high-resolution atlas of 4,5 renal papilla. The molecular heterogeneity of RECs from mouse renal endothelial cells. They also investigated how medullary the cortex, glomeruli, and medulla has been investigated renal endothelial cells adapt to a switch from diuresis to antidiuresis. only using microarrays on bulk gREC, cREC, and mREC pop- This study describes the molecular and metabolic adaptation of ulations.6 Single-cell RNA sequencing (scRNA-seq) studies medullary renal endothelial cells to dehydration, and uncovers a that sequenced enriched, freshly isolated RECs to construct role for mitochondrial oxidative phosphorylation in hyperosmolarity conditions to allow for urine concentration. The authors’ atlas of a comprehensive taxonomy of RECs at high resolution are not mouse renal endothelial cells provides a resource for future studies, available. and their findings may provide insights into cardiometabolic or kidney The heterogeneity of RECs may be of particular relevance diseases involving hyperosmolarity and dehydration, in which urine for the medullary circulation. For example, dehydration ex- concentration capacity is perturbed. poses medullary cells to extreme hyperosmolarity.7–9 How the renal medullary endothelium adapts to such conditions is un- For the scRNA-seq experiment, 12-week-old male mice known, but could be critical for the function of the counter- (C57Bl6/J strain; Charles River Laboratories) were exposed current system and hence capacity to concentrate urine. to food and water ad libitum, or were water-deprived, for Because the osmolarity gradient is maximal in the medullary 12, 24, 36, or 48 hours (n=6 per group). Rodents are physio- papilla, we utilized scRNA-seq to dissect the distinct responses logically equipped to tolerate acute dehydration for periods of of mRECs, gRECs, and cRECs to dehydration, and used the as long as 24 hours without overt signs of physiologic distress metabolic gene expression signature as a guidance to function- or behavioral abnormalities.10 On the basis of previous rec- fi ally con rm novel metabolic changes. To augment the resource ommendations, longer periods of acute water deprivation are value of our REC database, we provide access to http:// acceptable procedures, but only with suitable monitoring pro- www.vibcancer.be/software-tools/rec-dehydration to explore tocols.10 Following these recommendations,10 mice were all genes that are shown in t-distributed stochastic neighbor tightly monitored assessing body weight loss (twice a day), embedding (t-SNE) plots or are listed in gene expression behavior (physical vigor, activity), and any sign of major dis- heatmaps. tress or suffering (hunched posture, fuzzy facial fur, sunken/ recessed eyes) (three times a day), up to 48 hours of water deprivation. Humane end points were defined according to these METHODS parameters. None of the dehydrated mice included in the experiments showed signs of major distress or suffering as well Animals as behavioral abnormalities. Mice were anesthetized with an Animal experiments have been approved by the Institutional intraperitoneal injection of ketamine/xylazine (Dechra, Animal Ethics Committee of Katholieke Universiteit Leuven Innovet) and blood was collected from the right ventricle. “ (KU Leuven) (approved project no. P144/2018, Renal endo- Mice were then submitted to left heart perfusion with 3 ml thelial cells in health and disease: from full characterization to PBS (1 ml/min) and perfused kidneys were collected in ice-cold ” therapeutic intervention, animal experiments 10, 11, and 15). HBSS. All 60 kidneys were collected the same day to avoid batch All animals were maintained in a room with controlled tem- effects. perature and humidity under a 12-hour light/dark cycle. No statistical method was used to predetermine sample size. REC Isolation Kidney capsules were removed and kidneys were further dis- sected under a microscope to separate cortex from medulla. Received August 23, 2019. Accepted October 1, 2019. Kidney cortices and medullae from each condition were then S.J.D., E.M., and M.B. contributed equally to this work. cut into small pieces in a collagenase digestion buffer (2 mg/ml collagenase I [C9891; Sigma-Aldrich], 27 IU/ml DNAse I Published online ahead of print. Publication date available at www.jasn.org. [D4527; Sigma-Aldrich], and HBSS [14025100; Thermo Present address: Dr. Tobias K. Karakach, University of Manitoba, Department Fisher Scientific]); 15 ml were used for cortex samples and of Pediatrics and Child Health, Winnipeg, MB, Canada. 5 ml for medulla samples. The samples were then incubated Correspondence: Prof. Peter Carmeliet, Department of Oncology, Laboratory for 40 minutes at 37°C, shaking every 10 minutes, and diges- of Angiogenesis and Vascular Metabolism, Center for Cancer Biology (CCB), tion was stopped by adding 20 ml ice-cold HBSS. Digested Vlaams Instituut voor Biotechnologie (VIB), Herestraat 49 B912, 3000 Leuven, fi Vlaanderen, Belgium, or Prof. Yonglun Luo, Department of biomedicine, cortex and medulla samples were sequentially ltered through Aarhus University, Høegh-Guldbergs Gade 10, building 1116-152, 8000 Aarhus C, 70 and 40 mm filters and glomeruli were collected from the Denmark. Email: [email protected] or [email protected] 40 mm filter used to filter the cortex samples, and were resus- Copyright © 2020 by the American Society of Nephrology pended in HBSS. RECs from cortex and medulla were enriched

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 119 BASIC RESEARCH www.jasn.org using mouse CD31 magnetic microbeads (130–097–418; high-quality cells, the total number of genes detected was Miltenyi Biotec) and QuadroMACS separators with LS col- 15,977, ranging between 1029 and 1309 (average of 1141) genes umns (Miltenyi Biotec). Collected glomeruli were further dis- per cell. Data were first normalized for sequencing depth by sociated in a trypsin digestion buffer (0.2% trypsin [25200056; dividing by the total number of UMIs in each cell, and then Thermo Fisher Scientific],100IU/mlDNAseI,100mg/ml transformed to a log scale for each cell using the Seurat (http:// heparin [Heparine LEO; LeoPharma]) for 23 minutes at 37°C. satijalab.org/seurat/) NormalizeData function.11 Trypsin was inactivated by adding 10% FBS in ice-cold HBSS. For all of the samples, the following quality control steps were Glomerular cells were centrifuged for 6 minutes at 4003g, and applied: (1) genes expressed by ,50 cells or with a row mean resuspended in wash buffer (0.1% BSA in PBS). Cortical and of ,0.001 were not considered; and (2) cells that had either medullary endothelial-enriched cells as well as glomerular cells ,300 (low-quality cells) or .3000 expressed genes (possible were incubated with CD102-Alexa647 antibody (1/300, A15452; doublets), or .10% of UMIs derived from the mitochondrial Thermo Fisher Scientific), CD45-PeCy7 antibody (1/500, 130– genome were removed. Dataset-specificcut-offvaluesand 097–418; Thermo Fisher Scientific), CD73-PerCP-Cy5.5 antibody parameter settings are listed in Supplemental Table 1. (1/100, 127213; BioLegend, for glomerular cells only), and viability dye eFluor450 (1/1000, 65–0863–14; Thermo Fisher scRNA-seq Bioinformatic Analysis Scientific) in FACS buffer (0.1% BSA, 2 mM EDTA in PBS) In Silico EC Selection for 30 minutes at 4°C. OneComp ebeads (01–1111–42; Thermo After autoscaling, the normalized data were first summarized Fisher Scientific) and cooked cells (20 minutes at 60°C) were used by principal component analysis (PCA) using the flash PCA pack- for compensation. A total of 100,000 viable endothelial cells age, and the first 15 PCAs were visualized using t-SNE (Rtsne (ECs) were FACS-sorted using a FACS Aria III (BD Biosci- package) with a perplexity value of 200 and a learning rate of ences), and collected in wash buffer (0.1% BSA in PBS). Cells 100. Graph-based clustering was performed to group cells accord- were centrifuged for 6 minutes at 4003g, and resuspended in ing to their gene expression profiles as implemented in Seurat (see 40 ml wash buffer. Cell viability and number were assessed by Supplemental Table 1 for parameter settings). Cell clusters were staining cells with acridine orange and propidium iodide annotated on the basis of canonical cell type markers, including (F23001; Logos Biosystems) and using the automatic dual- Pecam-1, Icam2, and Cdh5 (ECs); and Hbb-a1, Hbb-a2, and fluorescence cell counter system (Luna-Fl; Logos Biosystems). Hbb-bs (red blood cells) to discriminate ECs from contaminating cells. No lymphatic ECs (Lyve1, Prox1), immune cells (Ptprc), or Single-Cell Droplet-Based RNA Sequencing and other renal cell types (tubular epithelial cells [Slc27a2, Lrp2, Data Preprocessing Slc12a1, Slc12a3, Pvalb, Aqp2, Hsd11b2, Atp6v1g3, Atp6v0d2, Insrr, The single-cell suspensions of FACS-sorted gRECs, cRECs, Rhbg], podocytes [Nphs1, Nphs2], and mesangial and granular and mRECs were converted to barcoded scRNA-seq libraries cells [Myl9, Ren1]) were detected. We did not identify a separate using the Chromium Single Cell 39 Library, Gel Bead & cluster, highly expressing a stress-response signature (artifactually Multiplex Kit and Chip Kit (103 Genomics; PN-120237, resulting from the dissociation procedure).12 All downstream PN-120236), aiming for approximately 4000 cells per li- analyses were performed on ECs only. brary of control gRECs, cRECs, and mRECs, and 3000 cells of RECs from each compartment and for each dehydration EC Cluster Identification time point. Samples were processed with 103 Genomics’ V2 We used the Seurat FindVariableGenes function to identify barcoding chemistry. Cells from each sample were treated with genes with high variability (see Supplemental Table 1 for pa- the same master mix and in the same reaction vessel, by pro- rameter settings for each analysis). The normalized data were cessing every single sample in a single well of a PCR plate. All autoscaled and PCA was performed on variable genes or all samples were processed in parallel in the same thermal cycler. genes (see Supplemental Table 1 for parameter settings for Libraries were sequenced on an Illumina HiSeq4000 and each analysis), followed by t-SNE to construct a two-dimen- mapped to the mouse genome (build mm10) using CellRanger sional representation of the data. To unbiasedly group control (103 Genomics, version 2.2.0), resulting in a sequencing depth gRECs, cRECs, and mRECs, we performed PCA on highly of 63,692 reads per cell and a saturation rate of 92% on average. variable genes, and used graph-based clustering as implemen- Data from the raw unfiltered matrix was further processed ted in the FindClusters function of the Seurat package.11 In using R (version 3.4.4). After quality filtering on the basis of addition, to identify clusters of cells with discriminating gene the number of detected genes, doublet exclusion, and mito- expression patterns in all datasets, we color-coded t-SNE plots chondrial read counts (see details below), a total of 40,662 for each of the 15,977 detected genes using an in-house devel- high-quality RECs were obtained (15,419 gRECs; 11,762 cRECs; oped R/Shiny-based web tool. Cluster results were visualized 13,481 mRECs for all conditions). One of the cREC samples using t-SNE to verify that all visually identified clusters were (cREC4: cRECs isolated after 36 hours of water deprivation) captured and not underpartitioned. Underpartitioned clusters did not pass the quality control because of a low number of unique that represented two distinct biologic phenotypes were subclus- molecular identifiers (UMIs) per cell (Supplemental Table 1) and tered. Overpartitioned clusters that represent the same biologic was therefore excluded from downstream analyses. Among all phenotype were merged into a single cluster. Clusters were

120 JASN JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH considered only when containing at least 1% of the total num- subsampled by randomly selecting 575 gRECs, cRECs, and ber of cells per sample. We did not identify a separate cluster, mRECs for each control and dehydration time point and highly expressing a stress-response signature (artifactually re- analyses were performed as described,13 using the SCENIC sulting from the dissociation procedure).12 Details of clustering R package (version 1.1.0, which corresponds to RcisTarget parameters are listed in Supplemental Table 1. 1.2.0, GENIE3 1.5.4, and AUCell 1.4.1) with RcisTarget motif To obtain ranked marker gene lists for each REC cluster, we databases (RcisTargets.mm9.motifDabases.20 k). performed pairwise differential gene expression analysis for each cluster against all other clusters independently, using the Venn Diagram Dehydration Analysis limma package (version 3.34.9). The results of each differen- To evaluate the dehydration effect on the gene signature be- tial analysis were ordered on the basis of log2 fold change tween gREC, cREC, and mREC, we calculated the intersec- (genes with the highest fold change receiving the lowest rank tion of genes between the three compartments from the top number). We obtained a final ranked marker gene list for each 50 up- and downregulated genes for each dehydration time cluster by calculating the rank product for all genes in all pair- point, using Venn diagram calculation and visualization. wise comparisons. This analysis was performed on gREC, cREC, and mREC samples taken separately, by using gene expression Gene Set Enrichment Analysis in control samples only, to avoid dehydration-induced effects. Weused gene set enrichment analysis (GSEA) as implemented in Toannotate clusters, we used canonical marker genes of artery, the clusterProfiler package (version 3.6.0) to compare gene ex- capillary, and vein ECs. In addition, we searched for a coherent set pression signatures between mRECs at 48-hour dehydration and of genes involved in similar biologic processes within the top 50 control mRECs.14 Gene set analysis was performed using a set of ranking list of markers to further identify the associated REC 415 vascular-related gene sets and 857 metabolic gene sets se- phenotype. We also used gene set variation analysis (GSVA) to lected from the Molecular Signatures Database (MSigDB version confirm upregulation of the identified biologic processes in the 5.2 downloaded from http://bioinf.wehi.edu.au/software/ respective REC phenotypes (see below). Cells that could not be MSigDB/). GSEA scores were calculated for sets with a mini- unambiguously assigned to a biologically meaningful phenotype mum of ten detected genes, all other parameters were default. might represent low-quality cells and were excluded from the analysis. Because of the manual microdissection of medulla Trajectory Inference from the cortex and of the REC isolation procedure, mREC and We used the SCORPIUS package (version 1.0.2) to place cells cREC samples contained a small cluster annotated as gRECs. This onto pseudotime trajectories.15 Analysis was performed on “contaminating” cluster was removed from the analyses for these highly variable genes. Dimensionality reduction was per- two compartments. formed using the reduce dimensionality function (we used 15 dimensions, all other parameters were default). Individual GSVA RECs from each cluster were subsequently placed onto linear We used GSVAas implemented in the GSVA R-package (version pseudotime using the infer trajectory function of the SCORPIUS 1.26.0) to convert the gene-by-cell matrix into a gene-set-by-cell package with default settings. Activated RECs (angiogenic and matrix. Gene set analysis was performed using a set of 415 vas- response-to-IFN clusters) expressed a highly distinct signature cular related gene sets selected from the Molecular Signa- and could therefore not be placed on the differentiation tures Database (MSigDB version 5.2 downloaded from trajectory. http://bioinf.wehi.edu.au/software/MSigDB/). GSVA scores were only calculated for gene sets with a minimum of five de- Jaccard Similarity Analysis tected genes. All other parameters were default. To measure the similarity of mRECs subclusters from dehydrated mice compared with those from control mice, we calculated Heatmap Analysis the similarity of the top 50 marker gene sets using pairwise All heatmaps are on the basis of cluster-averaged gene expres- Jaccard similarity coefficients for all subclusters against each sion to account for cell-to-cell transcriptomic stochastics. other. The Jaccard similarity coefficient is defined as the size of Data were autoscaled for visualization. Heatmaps were pro- the intersection divided by the size of the union of sets: duced using the Heatmaply Rpackage(version0.15.2).The data matrix for each heatmap can be downloaded from the jA∩Bj jA∩Bj JðA; BÞ¼ ¼ accompanying web tool (see Data Availability below). jA∪Bj jAj þ jBj 2 jA∩Bj

Single-Cell Regulatory Network Inference and where J is the Jaccard index and A and B are two sets of marker Clustering Analysis genes. PCA was used for visualization. Single-cell regulatory network inference and clustering (SCENIC) scans differentially expressed genes for overrepresented Crossdataset Validation of the mRECs in Dehydration transcription factor binding sites and analyzes coexpression of Subclusters from the control mREC samples were projected on transcription factors and their putative target genes. Data were all samples (respective control and all dehydration time points),

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 121 BASIC RESEARCH www.jasn.org using the scmapCluster algorithm as implemented in the scmap replicate per group) were seeded at a density of 500,000 cells package (version 1.1.5).16 We used the top 50 marker genes for per well on Seahorse XF24 tissue culture plates (Agilent) in each mREC subcluster in the control mREC taxonomy and a complete mouse EC medium (M1168; Cell Biologics). Sim- similarity threshold of 0.5; all other parameters were default. ilarly,culturedHUVECsinnormalorhyperosmolarmedium Below this threshold, cells were annotated as unassigned. The (1200 mOsm/kg) were seeded at a density of 100,000 cells per projection was visualized using a Sankey plot, similarity scores well on Seahorse XF24 tissue culture plates in normal or hyper- for each subcluster were visualized using a boxplot. osmolar (1200 mOsm/kg) EGM2 medium. For the metformin experiments, cultured HUVECs in normal medium were pre- Data Visualization treated with metformin (1, 10, or 20 mM) or control vehicle for The R implementation of the Plotly software was used for 1 hour. Oxygen consumption rate (OCR) measurements were t-SNE, violin plot, and bar graph visualization. performed at 6-minute intervals (2 minutes mixing, 2 minutes recovery, 2 minutes measuring) in a Seahorse XF24 device. The Human Umbilical Vein Endothelial Cells addition of oligomycin (1.2 mM final; Sigma-Aldrich) allowed Human umbilical vein endothelial cells (HUVECs) were for the determination of ATP-coupled oxygen consump- m fi freshly isolated from different donors after obtaining their in- tion rate (OCRATP). The addition of antimycine A (1 M nal; formed consent and under the ethical approval protocol Sigma-Aldrich) allowed for the determination of basal OCR. S57123 (Ethics Committee Research UZ/KU Leuven), as pre- HUVEC cultures were supplemented with ouabain at 0.5 mM viously described.17 The cells were routinely cultured in the immediately before OCR analyses to determine OCRATP re- + + appropriate gelatin-precoated (0.1% in PBS) cell culture plates lated to Na /K ATPase activity. The data were normalized to in M199 medium (Thermo Fisher Scientific) containing 20% protein content. FBS, 2 mM L-glutamine, 100 IU/ml penicillin, and 100 mg/ml streptomycin and endothelial cell growth factor supplements Detection of Organic Osmolytes (ECGS/heparin; Promocell) or in endothelial cell growth me- Details are provided in Supplemental Appendix 1. dium (EGM2; Promocell) supplemented with endothelial cell growth medium supplement mix (Promocell), and were regu- Biochemical Assays larly tested for mycoplasma. For experiments with galactose Plasma and Urine Sample Analyses medium, glucose-free EGM2 medium was reconstituted with Blood was collected from the right ventricle in anesthetized – 10 mM galactose in combination with dialyzed FBS (specifically mice (see above) using potassium EDTA coated syringes and 3 requested customized medium; Promocell). The cells were used centrifuged at 3000 g for 10 minutes. Plasma was collected 2 between passages 1 and 5 and all experiments were performed and stored at 20°C until biochemical measurement. Urine with HUVECs from at least three different donors. was collected directly from the bladder. Plasma creatinine, urea, sodium, total protein levels, osmolality, and aspartate aminotransferase and lactate dehydrogenase (LDH) activity, HUVEC Model of Dehydration Confluent HUVECs were adapted to hyperosmolarity by as well as urine osmolality, were measured in the Central progressively increasing the medium osmolality from 300 to Medical Laboratory of the University Hospital of Leuven. 900 mOsm/kg as follows: medium osmolality was changed to 380, 600, and 650 mOsm/kg every 12 hours, and up to Bicinchoninic Acid Assay 900 mOsm/kg by augmenting of 50 mOsm/kg every 2.5 hours. Bicinchoninic acid assay (Pierce) was used to determine pro- For seahorse experiments, the medium osmolality was in- tein content following metabolic assays and for Western blot creased up to 1200 mOsm/kg (50 mOsm/kg increase every experiments. Cells were lysed using only RIPA buffer (Thermo fi fl 2.5 hours). To achieve this higher osmolality, EGM2 was Fisher Scienti c) (for seahorse extracellular ux measure- supplemented with urea and NaCl (Sigma-Aldrich). For ments) or in the presence of protease and phosphatase inhib- the 380 mOsm/kg osmolality, urea was supplemented at itors (Roche, Vilvoorde, Belgium) (for Western blot analyses). 40 mM and NaCl at 20 mM, and for the 600 mOsm/kg, urea For organic osmolyte detection, protein pellets were dissolved was supplemented at 150 mM and NaCl at 75 mM. For adapt- in 200 mM sodium hydroxide. Bicinchoninic acid assay was ’ ing the medium from 650 to 1200 mOsm/kg, the urea molality performed following the manufacturer sguidelines. was kept constant at 150 mM, whereas NaCl was increased progressively to achieve the desired osmolality. Cells were Western Blot kept in hyperosmolar medium for at least 12 hours before Details are provided in Supplemental Appendix 1. performing any experiments. RNA Isolation and Quantitative RT-PCR Metabolic Assays RNA was collected from HUVECs cultured in normal or Seahorse Extracellular Flux Measurements hyperosmolar EGM2 (900 mOsm/kg), purified using the Freshly isolated CD31+MACS-enriched cRECs and mRECs PureLink RNA Mini Kit (Invitrogen, Life Technologies) and from control and 48-hour dehydrated mice (n=3 mice per converted to complementary DNA using the iScript cDNA

122 JASN JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH synthesis kit (Bio-Rad). RNA expression analysis was per- (CA-VIII; 1/400, 12391–1-AP; Proteintech), S100A4 (1/50, formed by Taqman quantitative RT-PCR as described,18 using A5114; Dako), PLVAP (1/40, Meca32-c; Developmental Studies premade primer sets (Integrated DNA Technologies, identifi- Hybridoma Bank), FXYD6 (1/50, 15805–1-AP; Proteintech), cation numbers: Hs.PT.58.25051081, Hs.PT.58.4615682, and CRYAB (1/200, AB13497; Abcam). Sections were then Hs.PT.58.2145446, Hs.PT.58.24360893, Hs.PT.58.39555113, incubated with the appropriate fluorescence-conjugated sec- Hs.PT.58.45350074). For comparison of gene expression be- ondary antibodies (Thermo Fisher Scientific) or followed by tween conditions, expression (normalized to HPRT as endog- amplification with the proper tyramide signal amplification enous control) is expressed relative to control condition. systems (NEL704A001KT, NEL705A001KT, NEL701A001KT; Perkin Elmer). Tyramide signal amplification was used for LDH Release Viability Assay CA-VIII, S100A4, PLVAP, and FXYD6 antibodies. Nuclei HUVECs in normal or hyperosmolar EGM2 (900 mOsm/kg) were counterstained with Hoechst (1 mg/ml, H3570; Thermo were trypsinated and seeded at a density of 100,000 cells per Fisher Scientific). Negative controls were performed omitting well in 96-well plates. The following day, cells were treated incubation with primary antibodies. Sections were mounted with rotenone (2 mM) (Sigma-Aldrich) or oligomycin on glass slides in ProLong mounting medium. Images were (1.2 mM; Sigma-Aldrich) in normal or hyperosmolar EGM2 acquired with a Zeiss LSM780 confocal microscope, coupled (900 mOsm/kg) and in normal or hyperosmolar galactose me- to ZEN software (Zeiss). Images were processed with ImageJ dium (900 mOsm/kg) for 24 hours. LDH assay kit (Cytotox- software. icity Detection Kit; Roche) was used to quantify cellular LDH release following the manufacturer’s guidelines. Renal Hypoxia Details are provided in Supplemental Appendix 1. Flow Cytometry Mitochondrial Content Total Mitochondrial Volume HUVECs cultured in normal or hyperosmolar medium were HUVECs cultured in normal or hyperosmolar EGM2 trypsinated and incubated with MitoTracker Green FM (1/5000, (900 mOsm/kg) were fixed in 4% paraformaldehyde for M7514; Thermo Fisher Scientific) for 30 minutes at 37°C in 30 minutes and then immunostained with TOMM20 (1/200; phenol red–free EBM2 medium (Promocell) supplemented Abcam) for mitochondria and Phalloidin-Alexa 488 (1/50; Life with endothelial cell growth medium supplement mix. Intensity Technologies) for actin. Images were acquired with a Zeiss of incorporated MitoTracker green was determined by fluoro- LSM780 confocal microscope, coupled to ZEN software. phore excitation with a 490 nm yellow laser and emission was Mitochondrial volumes were defined using Leica MetaMorph recorded at 516 nm. Data were recorded by flow cytometry, and AF 2.1 morphometry software package. A three-dimensional analyzed with the FlowJo 10.4.2 software (https://www.flowjo.com). volume mask was created for each cell in which the mitochon- MitoTracker staining was quantified by measuring median drial volume was calculated on the basis of the thresholded fluorescence intensity. TOMM20 signal.

Superoxide Content Measurement In Vivo Metformin Treatment HUVECs cultured in normal or hyperosmolar medium were For metformin treatment, 8- to 12-week-old male mice trypsinated and incubated with dihydroethidium (DHE; 1/250, (C57Bl6/J strain) were exposed to food and water ad libitum, D11347; Thermo Fisher Scientific) for 30 minutes at 37°C in or were water-deprived and treated with 50 mlofeithervehicle phenol red–free EBM2 medium supplemented with endothe- (water) or metformin (600 mg/kg; Sigma-Aldrich; a dose lial cell growth medium supplement mix. Intensity of incorpo- previously reported to influence urine concentration20) rated DHE was determined by fluorophore excitation with a twiceadaybygavagefor48hours(n=5–17 per group). 518 nm laser and emission was recorded at 606 nm. Data were Dehydrated mice were closely followed up, as described in recorded by flow cytometry, and analyzed with the FlowJo the Animals section. Mice were anesthetized with an intraper- 10.4.2 software (https://www.flowjo.com). DHE staining was itoneal injection of ketamine/xylazine, blood was collected quantified by measuring median fluorescence intensity. from the right ventricle and urine was collected from the bladder, and then analyzed as described above. Histology and Immunohistochemistry All methods for histology and immunostainings have been Statistical Analyses previously described.17,19 Mouse kidney samples were imme- Data are represented as mean6SEM. The presence of outliers diately fixed after collection in 4% paraformaldehyde over- was assessed using the ROUT test and eventual identified outliers nightat4°C,thendehydratedandembeddedinparaffin. were excluded from downstream analyses. Two-tailed statistical Briefly, kidneys were sectioned at 7 mm thickness and were test was used for all analyses. Statistical significance was assessed stained using the following primary antibodies: CD105 (1/50, with one sample t test or unpaired t test when comparing AF1320; R&D Systems), endomucin (1/50, 14–5851–82; two conditions, and one- or two-way ANOVAwith Bonferroni Thermo Fisher Scientific), carbonic anhydrase–related protein correction after the evaluation of variance homogeneity in

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 123 BASIC RESEARCH www.jasn.org

A kidney vasculature B experimental design C kidney compartments control

large afferent glomerular efferent arterioles heatmap all genes arteries arterioles capillaries peritubular gRECs v olute capillaries d

n high

l ute u 48 hrs

ol d a

v t l u e b s 36 hrs cRECs large n u i d o le 24 hrs veins c 12 hrs

l a mRECs im x o control dehydration r

p low c o I cortex 300mOsm kidney collection o l o medulla l p e c 400mOsm dissection kidney compartments t D capillary plexus i n 600mOsm control o g asc. f desc. cortex digestion medulla d 900mOsm vasa vasa u sieving c recta H recta t 1200mOsm e gRECs n cortical I glomeruli medullary KIDNEY e 1400mOsm cells cells VASCULATURE papilla digestion cRECs glomerular afferent arteriole mRECs GLOMERULUS cells MACS & FACS enrichment macula densa FACS proximal distal enrichment convoluted convoluted tubule tubule gRECs cRECs mRECs 15,419 cells 11,762 cells 13,481 cells capillaries juxtaglomerular apparatus single-cell RNA sequencing

efferent arteriole t-SNE 2 t-SNE 1 E F top 50 marker genes gRECs control kidney compartments control Pi16* Plat* Ehd3 Cyp4b1 Tspan7 Lpl* Pi16 Plat Ehd3 Cyb4b1 Tspan7 Lpl Igfbp3 Plvap

Npr3 t-SNE 2 t-SNE 1

Igf1 high cRECs & mRECs Cryab Igfbp7 control cRECs control Cd36 Aqp1 Plvap Igfbp3* Npr3* high Ifi27l2a low cRECs cRECs gRECs

mRECs low

G SCENIC mRECs t-SNE 2 transcription factor network t-SNE 1 kidney compartments control mRECs control Gata5 (28g) Gata5_extended (40g) Igf1 Cryab* Igfbp7* Cd36 Aqp1 Ifi27l2a* Gata2_extended (734g) Nuak2_extended (15g) Irf1 (36g) Foxp1 (10g) Sp1 (78g)

high Pparg_extended (23g) Lef1_extended (11g) t-SNE 2 t-SNE 1 Nfe2l2_extended (11g)

low

cRECs

gRECs mRECs

Figure 1. Control RECs show kidney compartment-specific gene signatures. All data in (C–F) are generated using RECs from kidneys in control condition. (A) Simplified schematic overview of the nephron and glomerular vasculature. The renal artery supplies blood to the entire kidney, and branches into afferent arterioles, which enter glomeruli to form fenestrated glomerular capillaries where blood ultrafiltration takes place. Unlike most capillary beds, glomerular capillaries exit into efferent arterioles, which branch into cortical peritubular capillaries vascularizing the proximal and distal convoluted tubules. DVR branch off from the efferent arterioles of juxta- medullary nephrons, enter the medulla and branch into a capillary plexus that drains blood into AVR. Vasa recta surround the loop of Henle and are essential for concentrating the urine. Peritubular capillaries and AVR connect to renal veins in the cortex, which drain the

124 JASN JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH

Brown–Forsythe tests, or Kruskal–Wallis tests followed by Supplemental Table 2). In addition, Plvap, a marker of fenes- Dunn tests, as appropriate, for multiple comparisons, using trated ECs with diaphragm, was highly expressed in cRECs and Prism 7 (GraphPad Prism). P,0.05 was considered statistically mRECs, but not in gRECs, consistent with previous reports1 significant. (Figure 1F). We identified markers for each REC type previously not highlighted in the literature (Figure 1, E and F,Supplemental Data Availability Table 2). Some markers were not uniformly expressed through- All raw sequencing data are available in ArrayExpress under out the clusters, but were identified only in a subset of cells within accession number E-MTAB-8145. Normalized data can be these clusters (Npr3 in cRECs; Aqp1, Cryab, and Ifi27l2a in downloaded from the accompanying web resource at http:// mRECs) (Figure 1F, Supplemental Table 2), raising the question www.vibcancer.be/software-tools/rec-dehydration. Custom of whether these clusters contained additional REC subclusters. code is available upon request. To study the underlying molecular mechanisms driving the differentiation of the different REC phenotypes, we used SCENIC.13 We identified the top ten enriched transcription fac- RESULTS tor networks in gRECs, cRECs, or mRECs, including Gata5, Gata2,andNuak2 regulons in gRECs; Irf1, Foxp1,andSp1 net- Isolation and scRNA-seq of RECs from Control and works in cRECs; and Pparg, Lef1, Nfe2l2,andIrf7 regulons in Dehydrated Mice mRECs (Figure 1G). In agreement with our findings on gRECs, The kidney vasculature displays a hierarchically organized ar- EC loss of gREC marker Gata5 causes glomerular lesions.23 chitecture1,2 (Figure 1A). To characterize the molecular heterogeneity of RECs and their adaptation to dehydration, we Intracompartment REC Heterogeneity collected kidneys from mice under euhydration (controls) and To further characterize the heterogeneity of gRECs, cRECs, mice deprived from drinking water for 12, 24, 36, and 48 hours and mRECs, we used a shared nearest neighbor modularity (Figure 1B). RECs from glomeruli, cortex, and medulla were optimization–based clustering algorithm to subcluster RECs processed separately to obtain a similar resolution for further on the basis of their gene expression signatures, resulting in characterization (Figure 1B, Supplemental Figure 1A). All 24 subclusters (Supplemental Figure 1, C–E). We used canon- gREC, cREC, and mREC single-cell samples were subjected ical marker genes of arterial (Sox17, Fbln5), fenestrated capil- to scRNA-seq, resulting in 40,662 high-quality RECs (Figure lary (Plvap, Kdr), and venous (Plvap, Nr2f2)ECs,aswellasthe 1B, Supplemental Figure 1B, Supplemental Table 1). top 50 marker genes of each subcluster to identify the different REC phenotypes. The number of RECs per subcluster as Intercompartment REC Heterogeneity well as the numbers of UMIs and genes per kidney compart- We first constructed a REC taxonomy in control conditions. A ment are displayed in Supplemental Figure 1, C–E. A detailed heatmap representing all genes expressed by gRECs, cRECs, and description of the marker genes (including novel markers, not mRECs showed that different REC types have unique transcrip- previously highlighted in the literature), with their inferred tome profiles (Figure 1, C and D). Using a rank product–based putative biologic role used to identify each subcluster, is method, we identified the top 50 marker genes for each REC reported in Supplemental Table 3. cluster (Supplemental Table 2). As reported, gRECs highly ex- gRECs exhibited five subclusters (Figure 2, A and B, Sup- pressed Ehd3, Cyp4b1,andTspan7,6,21,22 whereas mRECs plemental Figure 2A, Supplemental Tables 3 and 4): afferent showed higher expression of Igf1 and Cd366 (Figure 1, E and arteriolar ECs (G1), ECs from the terminal portion of the af- F, Supplemental Table 2). Specific markers for cRECs have not ferent arterioles associated with the juxtaglomerular apparatus yet been described. Here, we found that cRECs highly expressed (JGA) with enriched expression of cell-cell interaction genes Igfbp3, Npr3,andothernovelmarkers(Figure1,EandF, (G2) (Supplemental Figure 2B), fenestrated capillary gRECs (G3),

blood from the kidney. (B) Schematic overview of the study design. Kidney cortex, medulla, and glomeruli were enzymatically digested until obtaining single-cell suspensions. After a pre-enrichment of CD31+ cRECs and mRECs via MACS, all RECs were puriﬁed by FACS. Single-cell RNA sequencing of these RECs resulted in 15,419 high-quality gRECs; 11,762 high-quality cRECs; and 13,481 high-quality mRECs for all conditions. (C) Expression-level scaled heatmap of all genes from RECs isolated from the three kidney compartments. Scale: light blue is low expression, red is high expression. (D) t-SNE plot showing REC clusters from the three kidney compartments. (E) Ex- pression-level scaled heatmap of the top 50 marker genes of REC clusters isolated from the three kidney compartments. (F) t-SNE plots of RECs isolated from the three kidney compartments, color-coded for the expression of the indicated markers. Red arrowheads indicate cells with high expression of the indicated marker on the t-SNE plots. Scale: light blue is low expression, red is high expression. Asterisks indicate novel marker genes. (G) Expression-level scaled heatmap of the top 10 inferred transcription factor gene regulatory networks (as analyzed using SCENIC) in RECs from the three kidney compartments. Numbers between parentheses indicate the number of genes (g) that are part of the regulons for the respective transcription factors. Red corresponds to high transcription factor activity, blue corresponds to low transcription factor activity. See also Supplemental Figure 1 and Supplemental Tables 1 and 2.

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 125 lse.()Epeso-ee cldhampo h o 0mre ee ngE ucutrpeoye.()Pedtm analysis subclusters. Pseudotime gREC (C) indicated the phenotypes. in subcluster 126 pseudotime sub- gREC in unrecognized markers in gene previously genes indicated indicates marker the Asterisk of 20 plot. expression top smoothed t-SNE the loess by of and visualized subclusters heatmap subclustering gREC scaled of gREC Expression-level (A) (B) condition. cluster. control in kidneys from 2. Figure AI RESEARCH BASIC GH D A

cortex t-SNE 2 t-SNE 2 t-SNE 1 G3 t-SNE 1 glomeruli C3 arteriole/efferent C2 arteriole C1 artery/large G5 arteriole/efferent G4 arteriole/efferent/juxtaglomerularapparatus* G3 capillary G2 arteriole/afferent/juxtaglomerularapparatus G1 arteriole/afferent C6 oto RC n RC xii nrcmatethtrgniy l aasonwr eeae sn RC n cRECs and gRECs using generated were shown data All heterogeneity. intracompartment exhibit cRECs and gRECs Control C7 JASN cRECs control C5 gRECs control C8 medulla www.jasn.org G2 C9 C6 postcapillaryvenule C5 capillary#2 C4 capillary#1 G4 C4 C1 papilla G1 G5 C3 C2 * * C7 vein C6 postcapillaryvenule C5 capillary#2 C4 capillary#1 C3 arteriole/efferent C2 arteriole C1 artery/large BC E activated arteries efferent afferent capillaries veins PCV capillaries arterioles arterioles capillaries arterioles Cd300lg S100a6 S100a4 Col4a1 Cxcl12 Smad6 Hspg2 Alox12 Slc6a6 Sox17 Pde2a Cldn5 Plpp3 Calca Fbln2 Plvap Cryab Isg15 Efnb2 Bmp4 Cldn5 Calca Apoe Aplnr Podxl Fbln5 Thbd Tnxb Gas6 Npr3 Ehd3 Edn1 Thbd Sgk1 Cst3 Apln CD9 Mgp Gja4 CD9 Bgn Flt1 Ifit1 Jup Rdx Kdr Klf4 Eln Irf7 Id3 Kitl Kdr Lpl top 20markergenesgRECscontrol C9 capillary/interferon* C8 capillary/angiogenic* C7 vein C1 G1 C2 top 20markergenes C3 G2 cortex cRECs control

t-SNE 2 C4 t-SNE 1 C7 endomucin G3 C5 vein CA-VIII CD105 nuclei art. C6 Car8 G4 C7 0 m20µ 0 m200µm 200µm 200µm 200 µm 50 µm C8 G5 high low C9 vein high high low low CD105 art. F C5 pseudotime analysescRECscontrol 50 µm comp. 2 comp. 2 G3 comp. 1 G5 arteriole/efferent G3 capillary G1 arteriole/afferent juxtaglomerular apparatus G4 arteriole/efferent/ juxtaglomerular apparatus G2 arteriole/afferent/ comp. 1 C7 vein C6 postcapillaryvenule C5 capillary#2 C4 capillary#1 C3 arteriole/efferent C2 arteriole C1 artery/large pseudotime analysesgRECscontrol C4 C6 G2 C7 endomucin vein art. G4 G1 C1 C3 G5 C2 50 µm JASN BgnNpr3 Cldn5 Calca Lpl Alox12 vein pseudotime CA-VIII 31: pseudotime art. 118 – 50 µm 3,2020 138, www.jasn.org BASIC RESEARCH

ECs expressing a mix of capillary and arterial EC markers (G4), the papilla (Figure 3, C and E, Supplemental Figure 3B). We and efferent arteriolar gRECs (G5). When taking advantage of a also uncovered vein-like mRECs, which likely represented pseudotime algorithm to reconstruct the trajectory of the glomer- mRECs from the papillary portion of the ascending vasa recta ular vascular bed, we observed that G4 positioned as an interme- (papillary AVR) (M7), as they showed enriched expression of diate subcluster between G3 and G5, suggesting that it represents a hyperosmolarity responsive genes, similar to papillary DVR gREC population between the glomerular capillaries and efferent RECs. This was confirmed by immunostaining of the novel arteriole, as observed also for G2 with the afferent arteriole (Figure papillary AVR marker, CRYAB, expressed in the fenestrated 2C). Thus, in analogy with G2, G4 likely represents efferent PLVAP+ mRECs from AVR in the papilla (but not in the inner arteriolar gRECs associated with the JGA (Figure 2, A–C). and outer medulla) (Figure 3, C and F, Supplemental Figure 3B). A total of nine cREC subclusters were identified (Figure 2, Although mRECs from both papillary AVR and papillary DVR D and E, Supplemental Figure 2C, Supplemental Tables 3 and exhibited distinct transcriptome profiles, they shared upregula- 4): C1, C2, and C3 represent ECs from arteries, arterioles, and ted expression of hyperosmolarity-inducible genes and other efferent arterioles, respectively; C4 and C5 presented highly novel marker genes (Supplemental Figure 3A). Pseudotime- expressed capillary markers and were characterized by high based predictions of the renal medullary vascular trajectory con- (C4) or low (C5) ApoE expression; C6 showed an intermediate firmed this analysis, positioning the arterial, capillary, and vein capillary-vein gene signature associated with postcapillary ve- subclusters in this order (Figure 3G). We also identified three new nules; and C7 included cRECs from veins. Pseudotime-based mREC subpopulations, including capillary mRECs representing predictions of the renal cortical vascular trajectory confirmed an angiogenic phenotype (M8), as well as capillary (M9) and this analysis, positioning the identified arterial, capillary, and vein venous (M10) mRECs displaying a response-to-IFN pheno- subclustersinthisorder(Figure2F).Validationattheproteinlevel type (Figure 3, A and B, Supplemental Figure 3A). by immunostaining for CA-VIII confirmed expression of this GSVAshowed high expression of genes encoding MHC class II novel venous cREC marker in endomucin-positive venous cRECs, protein complexes in response-to-IFN ECs (Supplemental Figure but not in endomucin-negative arterial cRECs24 (Figure 2, G and 4A, Supplemental Table 5). However, these RECs did not express H, Supplemental Figure 2D). We also discovered two novel cap- detectable levels of the coactivators Cd80 and Cd86. Interestingly, illary cREC subclusters: C8 showing an angiogenic phenotype, cRECs from veins also highly expressed MHC II genes (Supple- and C9 exhibiting an IFN-response phenotype (Figure 2, D and mental Figure 4A). GSVA revealed enrichment of transcripts in- E, Supplemental Figure 2C, Supplemental Tables 3 and 4). volved in angiogenesis in angiogenic clusters from the cortex and mRECs exhibited ten subclusters (Figure 3, A and B, Sup- medulla (Supplemental Figure 4B, Supplemental Table 5). Un- plemental Figure 3A, Supplemental Tables 3 and 4): M1 rep- biased correlation heatmap analysis and hierarchical clustering resented mRECs from arterioles; M2 exhibited markers of the revealed that gRECs (afferent and efferent arterioles, capillaries) DVR; and M3 likely represented mRECs from the papillary represented a separate entity, resembling each other more than portion of the descending vasa recta (papillary DVR). Although RECs from other kidney compartments, illustrating their high no known markers of the papillary DVR exist, these cells highly specification (Supplemental Figure 4C). Overall, this taxonomy expressed hyperosmolarity responsive genes, consistent with captures the complex heterogeneity of RECs between and within the high osmolarity in the papilla.25 Immunostaining for kidney compartments. We used this atlas to study the transcrip- S100A4 confirmed expression of this novel papillary DVR tome response of each REC compartment to dehydration. marker in mRECs in the papilla, but not in the inner and outer medulla (Figure 3, C and D, Supplemental Figure 3B). Fenestrated Renal Endothelial Cells Differentially Respond capillary subcluster M4 included mRECs from the medullary to Dehydration capillary plexuses. M5 represented postcapillary venous ECs, In response to acute dehydration, the kidney concentrates the and M6 included vein-like mRECs from the AVR. Immunos- urine, increases water reabsorption, and reduces glomerular fil- taining for FXYD6 confirmed expression of this novel AVR tration.7–9 To characterize the transcriptional adaptation of RECs marker in mRECs in the inner and outer medulla, but not to dehydration, mice were deprived of water for up to 48 hours,

(D) cREC subclusters visualized by t-SNE plot. Asterisks indicate newly recognized subclusters. (E) Expression-level scaled heatmap of the top 20 marker genes in cREC subclusters. (F) Pseudotime analysis of cREC subclusters and loess smoothed gene expression of the indicated markers in pseudotime in the indicated cREC subclusters. Angiogenic and response-to-IFN cREC subclusters were not included in the pseudotime analysis. (G) Simplified schematic overview of the renal cortical vasculature. Inset shows larger magnification of the boxed area. Subclusters C4 and C5 are novel; the other vascular beds have been previously documented at the anatomic level. (H) Representative micrographs of mouse kidney sections immunostained for the endothelial cell marker CD105 (red), the vein/capillary marker endomucin (gray), and the novel marker CA-VIII (green) for the vein cluster (C7). Nuclei are counterstained with Hoechst (blue). Bottom images are magnifications of the respective boxed areas. Arrrowheads indicate endothelial cells from an artery and a vein. A t-SNE plot of cRECs, color-coded for the expression of Car8 is shown below. Scale: light blue is low expression, red is high. Note: CA-VIII denotes the protein, encoded by the Car8 gene. Scale bar, 200 mm (top panels) and 50 mm (bottom panels). See also Supplemental Figures 2 and 4, and Supplemental Tables 1, 3–5. Art, artery.

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 127 BASIC RESEARCH www.jasn.org

A mRECs control B top 20 marker genes mRECs control M8 Klf4 M2 Kitl M3 arterioles Ltbp4 Aqp1 M1 arteriole descending Hpgd M2 descending vasa recta vasa recta S100a6 M4 S100a4 M3 descending vasa recta/papilla* M1 CD36 capillaries Npr3 M4 capillary M10 Jup PCV ll6st M5 postcapillary venule M9 Fxyd6 ascending Gas6 M6 ascending vasa recta M6 Cryab vasa recta Fxyd2 M7 ascending vasa recta/papilla* activated Aplnr M8 capillary/angiogenic* M5 Trp53i11 high capillaries/ Isg15 M9 capillary/interferon* M7 Ifi203 t-SNE 2 ascending Irf7 M10 ascending vasa recta/interferon* vasa recta Ifit3 t-SNE 1 Ifi35 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 low

C D E cortex nuclei nuclei CD105 CD105 PLVAP glomeruli S100A4 CD105 S100A4 FXYD6 CD105 PLVAP FXYD6 medulla papilla 200 µm 200 µm 200 µm 200 µm 200 µm 200 µm 200 µm

papilla papilla outer &

inner medulla 50 µm 50 µm 50 µm 200 µm 200 µm 200 µm 200 µm medulla outer/inner medulla

outer/inner 200 µm 200 µm 200 µm 50 µm 50 µm 50 µm 50 µm descending vasa recta/papilla ascending vasa recta

M3 S100a4 M6 Fxyd6

M1 arteriole M2 descending vasa recta M3 descending vasa recta/papilla high high M4 capillary M5 postcapillary venule

M6 ascending vasa recta t-SNE 2 low t-SNE 2 low M7 ascending vasa recta/papilla t-SNE 1 t-SNE 1

F nuclei G CD105 PLVAP pseudotime analyses CRYAB CD105 PLVAP CRYAB mRECs control M7 Cryab M7 M2 M1 M3 Aqp1 200 µm 200 µm 200 µm 200 µm M6

M5 Kdr

comp. 2 M4 high 50 µm 10 µm 50 µm 10 µm 50 µm 10 µm 50 µm 10 µm comp. 1 M1 arteriole Cryab

t-SNE 2 M2 descending vasa recta low M3 descending vasa recta/papilla t-SNE 1 pseudotime M4 capillary M5 postcapillary venule medulla papilla

outer/inner 200 µm 200 µm 200 µm 200 µm M6 ascending vasa recta M7 ascending vasa recta/papilla ascending vasa recta/papilla

Figure 3. Control mRECs exhibit intracompartment heterogeneity. All data shown were generated using mRECs from kidneys in control condition. (A) mREC subclustering visualized by t-SNE plot. (B) Expression-level scaled heatmap of the top 20 marker genes in mREC subclusters. Asterisks indicate newly recognized subclusters. (C) Schematic overview of the renal papillary vasculature. Insets show larger

128 JASN JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH

A kidney compartments B kidney compartments control vs dehydration 48 hours control vs dehydration top 50 upregulated genes top 50 downregulated genes 600 gRECs mRECs cRECs mRECs cRECs 500 cRECs mRECs 400 33 7 30 41 5 30 6 2 300 4 7 2 13 200 33 33 100 number of dysregulated genes gRECs gRECs 0 12 24 36 48 time (h)

Figure 4. mRECs are the most affected by dehydration. (A) Venn diagram of the top 50 up- and downregulated genes in RECs from the three kidney compartments at 48 hours of dehydration. (B) Number of dysregulated genes [(log2(fold change) .0.2 or ,0.2] in gRECs, cRECs, and mRECs at the different dehydration time points versus control. Note for the 36-hour time point: the cREC sample did not meet sequencing quality standards and was therefore not included in downstream analyses. See also Supplemental Figure 5 and Supplemental Table 6. using a previously reported model, documented to induce weight time point revealed kidney compartment-specific REC gene ex- loss, increased plasma urea, sodium, total protein and osmolality pression changes, with only a few genes in common (Figure 4A, levels, urine osmolality, and decreased urine volume26–30 (Figure Supplemental Figure 5I, Supplemental Table 6). Thus, RECs 1B). Water-deprived mice lost body weight, whereas their urine from different compartments respond distinctly to dehydration, and plasma osmolality, plasma sodium and total protein levels, illustrating the heterogeneity of RECs not only in baseline but and plasma LDH and aspartate transferase activities were in- also upon water deprivation. Quantification of the upregulated creased (Supplemental Figure 5, A–G). Plasma urea levels were [log2(fold change) .0.2] or downregulated [log2(fold change) slightly elevated at certain time points (Supplemental Figure 5H), ,0.2] genes in RECs from each compartment suggested that whereas plasma creatinine levels remained unchanged (,0.15 mRECs were the most affected by dehydration (Figure 4B). We mg/dl), indicating that the kidneys were still functional. therefore focused primarily on mRECs. Weperformed a differential analysis of transcriptome profiles of gRECs, cRECs and mRECs at different time points after de- Adaptation of mRECs to Hyperosmolarity during hydration compared with their respective controls (Supplemen- Dehydration tal Table 6). Analysis of the top 50 up- and downregulated genes The medullary countercurrent system enables production of in RECs from the different compartments for each dehydration urine of widely varying osmolalities, and thereby exposes

magnifications of the respective boxed areas. (D) Representative micrographs of mouse kidney sections immunostained for the endothelial cell marker CD105 (red) and the novel marker S100A4 (green) for the DVR/papilla subcluster (M3). Nuclei are counterstained with Hoechst (blue). Top and middle panels are images of the papilla; middle images are magnifications of the respective boxed areas, with arrowheads indicating CD105+S100A4+ ECs. Bottom panels are images of outer/inner medulla. A t-SNE plot of mRECs color-coded for the expression of S100a4 is shown beneath the micrographs. Scale: light blue is low expression, red is high expression. Scale bar, 200 mm (top and bottom panels) and 50 mm (middle panels). (E) Representative micrographs of mouse kidney sections immunostained for the endothelial cell marker CD105 (red), the fenestrated endothelial marker PLVAP (gray), and the novel marker FXYD6 (green) for the AVR cluster (M6). Nuclei are counterstained with Hoechst (blue). Top panels are images of the papilla; middle and bottom panels are images of outer/inner medulla; bottom images are magnifications of the respective boxed areas, with arrowheads indicating CD105+PLVAP+ FXYD6+ ECs. A t-SNE plot of mRECs color-coded for the expression of Fxyd6 is shown beneath the micrographs. Scale: light blue is low expression, red is high expression. Scale bar, 200 mm (top and middle panels) and 50 mm (bottom panels). (F) Representative micrographs of mouse kidney sections immunostained for the endothelial cell marker CD105 (red), the fenestrated endothelial marker PLVAP (gray), and the novel marker CRYAB (green) for the AVR/papilla cluster (M7). Nuclei are counterstained with Hoechst (blue). Top and middle panels are images of the papilla of the mouse kidney; middle images are magnifications of the respective boxed areas. Higher magnification of the respective boxed areas is shown in insets, with arrowheads indicating CD105+PLVAP+CRYAB+ ECs. Bottom panels are images of outer/inner medulla of the mouse kidney. A t-SNE plot of mRECs color-coded for the expression of Cryab is shown on the right of the micrographs. Scale: light blue is low expression, red is high expression. Scale bar, 200 mm (top and bottom panels), 50 mm(middle panels), and 10 mm (insets in middle panels). (G) Pseudotime analysis of mREC subclusters and loess smoothed gene expression of the indicated markers in pseudotime in the indicated mREC subclusters. Angiogenic and response-to-IFN mREC subclusters were not included in the analysis. The t-SNE plots shown in (D–F) are also represented in Supplemental Figure 4A at a larger size for reasons of clarity. See also Supplemental Figures 3 and 4 and Supplemental Tables 1, 3–5.

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 129 BASIC RESEARCH www.jasn.org

ABmRECs CmRECs control & dehydration mRECs control & dehydration correlation heatmap control & dehydration pseudotime

AVR control 12-24 hours 36 hours DVR high 36-48 hours 48 hours control control 12 hours 12 hours 24 hours 24 hours low 36 hours control 48 hours 12 hours 24 hours control 36 hours 12 hours 48 hours 36 hours 24 hours comp. 2 t-SNE 2 48 hours capillary t-SNE 1 comp. 1

D E F mRECs top 50 mRECs top 50 mRECs upregulated genes downregulated genes control vs dehydration control high 12 hours 24 hours 36 hours 48 hours low heat shock immediate cytoskeleton DNA damage high high proteins early genes remodeling /growth arrest

low low control control 12 hours 24 hours 36 hours 24 hours 36 hours 48 hours 48 hours 12 hours

G mRECs H mRECs WATER cell volume regulation- control vs dehydration TRANSPORT water related genes cell volume regulation-related genes Slc6a6 Aqp1 MYO-INOSITOL POLYOL Slc2a1 BIOSYNTHESIS PATHWAY water Akr1b3 Sord glucose-6P glucose Atp1a1 TAURINE TRANSPORT Atp1b3 Isyna1 Akr1b3 + + 2Na 2Na Fxyd2 myo-I3P sorbitol - - Fxyd5 Cl Slc6a6 Cl high taurine taurine Slc38a2 Impa1 Sord Slc12a2 MYO-INOSITOL SODIUM POTASSIUM ATPase Slc5a3 TRANSPORT myo-inositol fructose Isyna1 + + myo-inositol 3Na 3Na Impa1 + Slc5a3 + Atp1a1 Fxyd2 low 2Na 2Na Aqp1 Atp1b3 Fxyd5 + + 2K 2K high SODIUM/AMINO ACID control TRANSPORT 12 hours 24 hours 36 hours 48 hours + + Na Na amino Slc38a2 amino acid acid low

Figure 5. mRECs show a typical transcriptomic response to hyperosmolarity in dehydration. (A) Correlation heatmap of mRECs from the control condition and at different dehydration time points. Scale: red indicates a high transcriptome similarity, blue indicates a low transcriptome similarity. (B) t-SNE plot color-coded for mRECs from the control condition and at different dehydration time points. (C) Pseudotime analysis of mRECs from control conditions, and at 12–24 and 36–48 hours of dehydration. (D and E) Expression-level scaled heatmap of the top 50 upregulated (D) and downregulated (E) genes in mRECs at different dehydration time points. Red corresponds to high gene expression levels, blue corresponds to low gene expression levels. (F) Expression-level scaled heatmap of genes encoding heat shock proteins, or genes involved in cytoskeleton remodeling, DNA damage, and growth arrest, and immediate early genes in mRECs from control condition and at the different dehydration time points. Red corresponds to high expression levels, blue corresponds to low expression levels. (G) Expression-level scaled heatmap of cell volume regulation–related genes in mRECs from control condition and at the different dehydration time points. Red corresponds to high expression levels, blue corresponds to low expression

130 JASN JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH medullary renal cells to harsh conditions (elevated levels of (similarity threshold .0.5; Supplemental Figure 6B). Thus, solutes such as sodium chloride and urea) upon dehydration. no novel dehydration-evoked cluster was identified in mRECs. Although the adaptation of medullary epithelial cells to survive in such hostile milieu are described,25,31,32 little is known mRECs Adapt Their Metabolic Gene Expression about the mechanisms via which mRECs cope with and adapt Signature upon Dehydration to a hyperosmolarity challenge. We observed a shift in gene GSEA of mRECs from control versus 48-hour dehydrated expression as early as 12 hours of water deprivation in mRECs mice, in combination with expression heatmap analysis, re- (Figure 5, A and B). Their response to dehydration occurred in vealed upregulation upon dehydration of gene sets involved in two phases: an early adaptation phase (12–24 hours) and a late hypoxia/hypoxic response, but surprisingly, also in inorganic response (36–48 hours) (Figure 5, A and B). Pseudotime anal- ion transmembrane transport (genes encoding complex sub- ysis of mRECs showed that the arterial-capillary-venous units of OXPHOS) (Figure 6, A and B, Supplemental Table 7), trajectory was affected the most at 36–48 hours of water dep- suggesting that mRECs might be exposed to deeper hypoxia, rivation and that dehydration affected medullary capillaries yet upregulated OXPHOS (see below). Upregulation of the most (Figure 5C). Differential analysis of mREC transcrip- OXPHOS genes has never been reported in medullary epithe- tome profiles in control and at 12, 24, 36, and 48 hours of lial cells exposed to hyperosmolarity. Consistent with a water deprivation largely confirmed this result (for the upre- report highlighting an increase in hypoxia inducible factor gulated genes, the 36-hour time point seemed to be a transi- 1a–positive cells in the renal medulla upon dehydration,35 tion phase) (Figure 5, D and E). we observed an increase in pimonidazole-stained hypoxic In the early phase, we observed a transient increase in ex- areas in the inner kidney medulla from dehydrated mice pression of genes encoding heat shock proteins and immediate (48 hours) (Supplemental Figure 5J), validating the GSEA re- early genes, and initial changes in expression of genes involved sults. Also, SCENIC revealed that the Epas1 regulon (encoding in cytoskeleton remodeling, DNA damage, and growth arrest, hypoxia-inducible factor 2a) was upregulated in mRECs after as reported for renal epithelial cells upon dehydration or hy- dehydration (Figure 6C). We also identified upregulation of perosmolarity33 (Figure 5F). In the late phase, the changes in Rad21 regulon, involved in DNA repair,36,37 at all dehydration expression of genes involved in cytoskeleton remodeling, time points, whereas the Pparg regulon was downregulated DNA damage, and growth arrest became more prominent (Figure 6C). (Figure 5F). Other transcriptome changes related to genes in- Considering that the unbiased GSEA yielded an upregula- volved in cell volume regulation, especially encoding trans- tion of mainly metabolic pathway genes, together with hyp- porters and biosynthetic enzymes necessary to increase the oxia, we performed a GSEA focusing on metabolic pathways intracellular concentration of inert osmolytes and sodium as only (Figure 6D, Supplemental Table 7). This analysis revealed well as water transport to equilibrate intra- and extracellular an upregulation of ribosome- and proteasome-related genes, osmolarity.25,31,32 In agreement, mRECs upregulated the suggesting high protein turnover (Figure 6, E and F). Also, expression of genes involved in taurine transport (Slc6a6), genes involved in OXPHOS and glycolysis were upregulated myo-inositol transport (Slc5a3, transiently), sodium-coupled in mRECs after 36–48 hours of dehydration (Figure 6, neutral amino acid transport (Slc38a2), sorbitol synthesis B and G), whereas expression of TCA cycle–related genes (Akr1b3), Na+/K+ ATPase (Atp1a1, Atp1b3, Fxyd2, Fxyd5), did not show prominent changes, as visualized by metabolic and water transport (Aqp1) (Figure 5, G and H). Not previ- pathway mapping (Figure 6H). Consistently, freshly isolated ously reported as a response mechanism to hyperosmolarity, mRECs from mice exposed to 48 hours of water deprivation we observed a transcriptome switch from myo-inositol trans- showed a higher level of oxygen consumption coupled to ATP port (Slc5a3)tosynthesis(Isyna1) (Figure 5, G and H). Con- synthesis, whereas oxygen consumption was not affected in trary to what is described in epithelial cells,34 transcript levels cRECs from the same animals (Figure 6I). Taking advantage of 2 of the Na+/K+/Cl (NKCC) cotransporter (Slc12a2)werede- the mREC taxonomy, we observed that OXPHOS genes were creased in mRECs (Figure 5G). enriched in papillary mRECs compared with other mRECs in mRECs from dehydration samples contained the same both control and 48 hours dehydration conditions (Figure 6J). number of similar mREC subclusters as control mRECs (Sup- In contrast to mRECs, which launched a response in two plemental Figure 6A). In addition, by using scMap16 and the clearly distinct phases, each time point after dehydration top 50 marker genes of each mREC subcluster (M1–M10), largely evoked a different response in gRECs (Supplemental mRECs from dehydration samples were confidently assigned Figure 7, A and B). In addition, changes in expression of the to mREC subclusters in the control mREC taxonomy aforementioned genes involved in cell volume regulation,

levels. (H) Pathway map showing changes in transcript levels of genes belonging to myo-inositol and polyol pathways, and genes encoding transporters related to cell volume regulation in mRECs after 48 hours of dehydration compared with control. Scale bar: red corresponds to gene upregulation after dehydration, gray indicates that the change in gene expression did not reach the fold change threshold to be color-coded. See also Supplemental Figure 7 and Supplemental Table 1.

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 131 BASIC RESEARCH www.jasn.org

A mRECs B mRECs C mRECs control vs dehydration 48 hours oxidative SCENIC geneset enrichment analysis phosphorylation transcription factor network

hydrogen ion transmembrane transport Max_extended (1706g) inorganic cation transmembrane transport inorganic ion transmembrane transport E2f4_extended (920g) ion transmembrane transport CI Jund_extended (120g) transmembrane transport Bbx (14g) cation transmembrane transport hypoxia Rad21_extended (900g) secretome of lung cancer and endothelium CII Epas1 (22g) Hif1a targets dn CIII carbohydrate transport high Taf7 (52g) high enriched in mRECs 48 hours dehydration Irf1 (36g) active ion transmembrane transport CIV hypoxia Nr3c1 (25g) p53 targets up positive regulation of endothelial cell proliferation CV Pparg_extended (23g) IFNbeta response up low low VEGF signaling pathway VEGF/VEGFR pathway

enriched in cell migration involved in sprouting angiogenesis control mRECs control control 12 hours 24 hours 48 hours 36 hours -2 -1 0 1 2 12 hours 24 hours 36 hours 48 hours

mRECs mRECs DEmRECs F G mRECs ribosomal proteasome control vs dehydration 48 hours glycolysis geneset enrichment analysis genes genes metabolic genesets Slc2a1 metabolism of mRNA Hk1 metabolism of RNA Pfkl respiratory chain Pfkm electron transport chain metabolism of proteins Pfkp amide biosynthesis Aldoa hydrogen ion transmembrane transport Tpi1 hydrogen transport oxidoreductase activity acting on NADPH Gapdh enriched in mRECs nucleoside triphosphate metabolism 48 hours dehydration high high Pgk1 high positive regulation of glycoprotein metabolism Pgam1 positive regulation of glycogen metabolism cAMP catabolism Eno1 positive regulation of NO biosynthesis Pkm positive regulation of glucose metabolism Ldha bile acid transmembrane transport regulation of glucose metabolism low low Slc16a3 low enriched in DNA catabolism control mRECs proteoglycan metabolism proteoglycan biosynthesis control control control

-2 -1 0 1 2 12 hours 24 hours 36 hours 48 hours 12 hours 24 hours 36 hours 48 hours 12 hours 24 hours 36 hours 48 hours H I GLUCOSE TRANSPORT ATP-linked respiration glucose mRECs 10 * Slc2a1 control vs dehydration 48 hours 8

6 GLYCOLYSIS ATP OXIDATIVE PHOSPHORYLATION 4 glucose complex I complex III complex IV complex V OCR 2 Uqcrfs1 Cyc1 Atp5a1 Atp5g1 Hk1 Ndufa Ndufs Cox4i1 Cox7 Uqcrc Uqcrb Atp5b Atp5h (pMoles/min/µg protein) Ndufb Ndufv Cox5 Cox8 0 Uqcrh Uqcrq Atp5c1 Atp5j glucose-6P Ndufc complex II Cox6 CTRL DH CTRL DH Uqcr10 Uqcr11 Atp5d Atp5k Gpi1 cRECs mRECs Atp5e Atp5o fructose-6P Sdh Atp5f1 Atp5l Pfk succinate Sucla2 J mRECs

fructose-1,6BP succinyl-CoA fumarate Aldoa Ogdh Fh1 CI G3P Tpi1 malate a-ketoglutarate Gapdh TCA CYCLE 3PG Idh1 Mdh CII Pgam1 isocitrate CIII oxaloacetate Eno1 high Aco2 Pdh CIV Pkm cis-aconitate pyruvate acetyl-CoA citrate high oxidative phosphorylation Ldhb Cs Aco2 CV Acly Ldha low lactate low papilla papilla medulla medulla

CTRL DH (48h)

Figure 6. mRECs metabolically adapt to dehydration. (A) GSEA of mRECs after 48 hours of dehydration compared with controls. (B) Expression-level scaled heatmap of oxidative phosphorylation-related genes for mRECs from the control condition and at different dehydration time points. (C) Expression-level scaled heatmap of the top up- and downregulated inferred transcription factor gene regulatory networks (as analyzed using SCENIC) in mRECs in control condition and at different dehydration time points.

132 JASN JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH metabolism, protein turnover, and stress response were differ- over anaerobic metabolism41–43 (Figure 7F). By contrast, con- entingRECs(SupplementalFigure7,C–H).Overall,the trol ECs exposed to the same medium were not significantly adaptive response of gRECs to dehydration was qualitatively affected by OXPHOS inhibitors, highlighting a key role and quantitatively different, likely because these cells are not for OXPHOS in EC survival in hyperosmolarity conditions exposed to such high hyperosmolarity and are influenced by (Figure 7F). their local microenvironment. Furthermore, treatment of mice with metformin, a multifunctional compound that inhibits complex I of OXPHOS OXPHOS Upregulation Promotes EC Survival and Urine and oxygen consumption in ECs (Supplemental Figure 8C; Concentration consistent with previous reports44,45), during dehydration el- To investigate whether the upregulation of the OXPHOS gene evated plasma urea levels, without altering plasma creatinine signature and oxygen consumption in mRECs after dehydra- levels (,0.12 mg/dl), consistent with more severe dehydration tion was triggered by hyperosmolarity, we developed an in vitro (Figure 7G).26 Notably, metformin treatment largely prevented model using HUVECs, by exposing them to a progressive in- the increase in urine osmolality during dehydration, reflecting crease in urea and NaCl concentration in the culture medium, an impaired capability in urine concentration (Figure 7H). the main osmolytes in the kidney medullary interstitium,38 Overall, mRECs adapted to hyperosmolarity upon dehydration until reaching an osmolality of 900–1200 mOsm/kg in the me- by upregulating an OXPHOS gene signature and oxygen con- dium corresponding to osmolality values close to the maximal sumption, whereas treatment with a multifunctional OXPHOS urine concentration capacity in humans.38 Consistent with blocker (see Limitations of the Study) impaired the physiologic the mREC data in dehydration, HUVECs exposed to hyper- ability of mice to concentrate urine to spare body fluid. osmolarity upregulated several osmolytes necessary to maintain cell volume, such as sorbitol and glutamine, among others (Supplemental Figure 8A). They also upregulated genes DISCUSSION encoding the different complex subunits of OXPHOS (Figure + + 7A), and augmented OCRATP (Figure 7B). The Na /K ATPase Renal Endothelial Cell Heterogeneity in Health pump plays a critical role in maintaining physiologic concen- Using scRNA-seq, we revealed a total of 24 different REC trations of intracellular sodium, especially in hyperosmolarity subpopulations, of which eight were previously not recog- conditions, a process requiring high amounts of ATP.25,39,40 As nized. In the glomerular vasculature, we identified a new in mRECs during dehydration (see above), protein levels of gREC subpopulation (G4) presumably lining the juxtaglo- this pump were upregulated in HUVECs exposed to hyperos- merular section of the efferent arteriole. Interestingly, gRECs molarity (Supplemental Figure 8B). In agreement, treatment of in afferent arterioles associated with the JGA (G2) highly ex- hyperosmolarity-exposed HUVECs with ouabain, a Na+/K+ pressed genes encoding cell-cell interaction molecules, possibly ATPase inhibitor, inhibited the increase in OCRATP (Figure 7B), to control processes (renin production, tubulo-glomerular suggesting that the increase in OXPHOS is necessary to sustain feedback) that require close interactions between gRECs, renin- ATP production for sufficient Na+/K+ ATPase activity. producing cells, and other cells in the JGA.46 For instance, en- Hyperosmolarity-induced upregulation of OXPHOS was riched Dll4 expression in this gREC cluster might signal to associated with an increase in mitochondrial content and re- Notch-expressing granular cells to maintain renin produc- active oxygen species levels (Figure 7, C–E). To study the function.6,47 Consistent with their role in fine-tuning GFR,48 gRECs tional relevance of the increased OXPHOS gene expression, from afferent and efferent arterioles highly expressed marker we treated hyperosmolarity-exposed HUVECs with OXPHOS genes encoding vasoregulatory molecules. In the cortical inhibitors (complex I inhibitor rotenone; complex V inhibitor vasculature, we deconvoluted cortical capillaries in two sub- oligomycin). OXPHOS inhibition compromised HUVEC sur- populations, characterized by high (C4) or low (C5) ApoE vival when cells were cultured in a hyperosmolar medium, in expression and other transcriptome signatures. In the medul- which glucose was replaced by galactose to increase aerobic lary vasculature, we identified novel markers of DVR mRECs,

Numbers between parentheses indicate the number of genes (g) that are part of the regulons for the respective transcription factors. Red corresponds to high transcription factor activity, blue corresponds to low transcription factor activity. (D) GSEA using only metabolic gene sets of mRECs after 48 hours of dehydration compared with controls. (E–G) Expression-level scaled heatmap of ribosomal genes (E), proteasome genes (F), and glycolysis-related genes (G) for mRECs from the control condition and at different dehydration time points. (H) Pathway map showing changes in transcript levels of metabolic genes of glycolysis, TCA cycle, and oxidative phosphorylation in mRECs after 48-hour dehydration compared with controls. Scale bar: red corresponds to gene upregulation after dehydration, gray indicates that the change in gene expression did not reach the fold change threshold to be color-coded. (I) OCRATP of cRECs and mRECs isolated from control mice and from mice subjected to 48 hours of dehydration (DH). Data are from n=3 independent experiments, for which three mice were pooled each time. (J) Expression-level scaled heatmap of oxidative phosphorylation-related genes for mRECs from papillary subclusters or other medullary subclusters in the control condition and after 48 hours dehydration. See also Supplemental Figure 5 and Supplemental Table 7. Statistical test: unpaired t test, *P,0.05.

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 133 BASIC RESEARCH www.jasn.org

A OXPHOS complex subunits B C mitochondrial content gene expression ATP-linked respiration mitotracker green 3 NDUFA7 SDHB UQCR10 COX4I1 ATP5H 2.0 800 * * * 1.5 600 * 2 # ATP 1.0 400 OCR 1 * 0.5 200 mRNA copies/ * median fluorescence intensity (fold change)

* (pMoles/min/µg protein) 0 0.0 1,000 copies of HPRT1 0 control + - + - + - + - + - hyperosmolarity -+-+-+-+-+ control control hyperosm. hyperosm. hyperosm.ouabain +

DEmitochondrial content F TOMM20 staining reactive oxygen species OXPHOS inhibition Dihydroethidium (DHE) cell death Nuclei Phalloidin TOMM20 100 2.0 2.5 # * 80 2.0 1.5 * # 60

control 1.5 1.0 40 20 µm 20 µm 1.0 cell death (%) (fold change) 0.5 0.5 20 median fluorescence intensity (fold change) 0.0 0

Total mitochondrial volume 0.0 rotenone -+- -+- -+- -+- oligomycin --+ --+--+--+

hyperosmolarity 20 µm 20 µm control control galactose galactose hyperosm. hyperosm. control hyperosmolarity

G in vivo OXPHOS inhibition H in vivo OXPHOS inhibition plasma urea level urine osmolality 200 4000 * 150 # # 3000 100

mg/L 2000

50 mmol/kg H2O 1000

0 0

DH DH CTRL CTRL

DH + MET DH + MET CTRL + MET CTRL + MET

Figure 7. ECs survive hyperosmolarity during dehydration by upregulating oxidative phosphorylation. (A) Quantitative RT-PCR analysis of OXPHOS complex subunit-encoding transcripts (NDUFA7, SDHB, UQCR10, COX4I1, ATP5H) in control and hyperosmolarity- exposed HUVECs (n=4). (B) OCRATP of control and hyperosmolarity-exposed HUVECs, in the absence or presence of ouabain (0.5 mM) (n=3). (C) MitoTracker green median fluorescence intensity of control and hyperosmolarity-exposed HUVECs (n=10). (D) Left: representative images of control and hyperosmolarity-exposed HUVECs stained for the mitochondrial marker TOMM20 (red), the cytoskeleton phalloidin (green), and the nucleus (Hoechst, blue). Scale bar: 20 mM. Right: total mitochondrial volume per cell of TOMM20-stained control and hyperosmolarity-exposed HUVECs (n=4). (E) DHE median fluorescence intensity of control and hyperosmolarity-exposed HUVECs (n=8). (F) Cell death quantification of control and hyperosmolarity-exposed HUVECs in the presence or absence of OXPHOS inhibitors (rotenone 2 mM; oligomycin 1.2 mM) in control medium or medium in which glucose was replaced by galactose (n=3–6). (G and H) Plasma urea levels (G) and urine osmolality (H) of mice treated with vehicle or metformin (600 mg/kg, twice a day) and submitted to 48 hours of dehydration (DH) (n=5–17). Data are mean6SEM. See also Supplemental Figure 8. Statistical tests: one-sample t test, unpaired t test, one- and two-way ANOVA/Bonferroni, *P,0.05 compared with control; #P,0.05 compared with hyperosmolarity or dehydration.

134 JASN JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH including, for instance, Scin encoding Scinderin (binding the renal epithelial cells, which primarily rely on glycolysis for ATP water channel aquaporin-2 in collecting duct epithelial cells49). production.25,60 The high expression of Scin and Aqp1 in DVR mRECs raises the Interestingly, although ECs are glycolysis-addicted and question whether a similar interaction of Scinderin with AQP1 produce most of their ATP glycolytically,61 mRECs upregula- may be operational in mRECs. Other novel markers of DVR ted OXPHOS gene expression and oxygen consumption upon mRECs included genes involved in vasodilation (Adipor2, dehydration, while being exposed to deeper hypoxia, possibly Hpgd), consistent with the role of DVR mRECs in regulating to use the available oxygen more efficiently. Furthermore, medullary blood flow and osmolarity gradient formation.50 OXPHOS inhibition compromised EC survival in hyperos- We also identified new markers of mRECs of the DVR and molar conditions in vitro and largely prevented urine concen- AVR in the medullary papilla, which were poorly characterized trationindehydratedmicein vivo. Such a role for OXPHOS in the past. For instance, papillary AVR mRECs highly ex- in the response to hyperosmolarity has not been documented pressed Cryab involved in the hyperosmolarity response51 previously, also not in medullary epithelial cells. We speculate and Car2, the loss of which causes polyuria and a failure to that hyperosmolarity-exposed mRECs upregulate OXPHOS to augment the medullary osmolarity and thereby to concentrate provide ATP for energy-demanding processes such as protein the urine.52 synthesis, volume and osmolyte content regulation, and main- We also uncovered capillary and vein mREC and cREC tenance of intracellular sodium concentration requiring so- populations showing a response-to-IFN phenotype as well dium export driven by the Na+/K+ ATPase,25,39,40 especially as antigen processing and presentation-related genes, but because ATP production from glycolysis can be compromised not Cd80 and Cd86, suggesting participation in immune sur- if a fraction of the glycolytic intermediates is shunted for os- veillance as semiprofessional antigen presenting cells, capable molyte (sorbitol; myo-inositol) synthesis in hyperosmolarity, of activating only antigen-experienced immune cells.53 Angio- and OXPHOS is a more productive pathway to generate ATP genic capillary mRECs and cRECs were also identified, al- per mole of glucose than glycolysis. Hyperosmolarity-exposed though whether they are involved in repair/regeneration of mRECs might also upregulate OXPHOS to produce “metabolic damaged RECs requires further study. water” during mitochondrial respiration, resulting from the oxidation of organic materials.62–65 Whether and to which ex- Renal Endothelial Cell Response to Dehydration tent metabolic water production by hyperosmolarity-exposed The fundamental question of how mRECs adapt to dehydra- mRECs contributes to cell volume regulation in dehydration tion and survive in hyperosmolarity conditions has not re- remains to be determined. ceived much attention. This is highly relevant because ECs In addition to providing a detailed REC taxonomy in base- have long half-lives, and mRECs are exposed to inhospitable line condition, this study also offers insight in the transcrip- conditions during dehydration, which would be lethal to other tome adaptation of RECs to dehydration at the single-cell cell types in the body. Dehydration-exposed mRECs upregu- level, highlighting a previously unrecognized role of lated expression of genes involved in heat shock proteins, cy- OXPHOS in the mREC response to hyperosmolarity and toskeleton rearrangement, and DNA damage/growth arrest, the ability to concentrate urine in dehydration. As such, this consistent with a response to the hypertonic effect of high study may constitute grounds for examining the response of NaCl and to the denaturating effect of urea on cellular mac- medullary epithelial cells upon hyperosmolarity challenge in romolecules,32,38 as described in renal inner medulla epithelial the future. Furthermore, the publicly available extensive data- cells.32 mRECs from dehydrated mice also highly expressed set is a rich resource for future data mining. genes, involved in proteasome degradation, possibly of dena- tured proteins, and increased the expression of ribosomal genes, Limitations of the Study possibly to compensate for the enhanced protein turnover.54,55 There were several limitations of our study. First, we used Medullary epithelial cells accommodate hyperosmolarity metformin as a non-toxic compound to inhibit OXPHOS in by accumulating organic metabolites as inert osmolytes.25,32,56–58 RECs in dehydrated mice in vivo, and acknowledge that this In agreement, mRECs upregulated the expression of genes involved compound may have additional activities66,67 and can act in- in water transport, taurine/sodium and amino acid/sodium directly, and hence more direct proof is required in the future. cotransport, and myo-inositol and sorbitol synthesis, as well Second, although we did not identify REC subclusters as genes involved in Na+/K+-ATPase–mediated sodium export expressing a tissue dissociation artifact gene signature, we can- during dehydration, presumably to equilibrate osmotically not exclude the possibility that the REC isolation procedure with their hypertonic interstitium while preventing toxic intra- might have influenced REC transcriptomes. To minimize such cellular Na+ accumulation. The low medullary oxygen, resulting effect, we used similar dissociation protocols. Third, unfortu- in part from the low blood flow rate in the vasa recta, is an nately, the insufficient number of cRECs at 24 hours of de- inevitable accompaniment of urinary concentration, especially hydration precluded us from confidently characterizing their in antidiuretic conditions.59 Deeper hypoxia in the medulla possible different transcriptome response to dehydration. during dehydration may contribute to the upregulated expres- Fourth, although we demonstrate that freshly isolated mRECs sion of glycolytic genes in mRECs, as also occurs in medullary from dehydrated mice have elevated oxygen consumption,

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 135 BASIC RESEARCH www.jasn.org additional studies are required to further characterize possible (81670855), a Guandgong Province Leading Expert Program grant and the elevations of OXPHOS and oxidative stress in vivo by comple- Key Program of Guangzhou Scientific Research Plan (3030901006074). The mentary quantitative methods. Finally, we also acknowledge that work of Prof. Carmeliet is supported by the Vlaams Instituut voor Biotech- nologie TechWatch program, long-term structural Methusalem funding by water deprivation may evoke additional systemic and local ef- the Flemish Government, grants from the Research Foundation Flanders fects, such as reducing caloric intake, BP,and urine volume, and (FWO-Vlaanderen), Foundation against Cancer (2016-078), Kom op Tegen Kanker altered mechanostimulatory stimuli (lower blood and urine (Stand up to Cancer, Flemish Cancer Society), European Research Council Proof of flow),10,26,28,30 which may contribute to the observed findings. Concept (ERC-713758), and ERC advanced research grant (EU-ERC743074). This study was sponsored by Regenerative Medicine Crossing Borders.

ACKNOWLEDGMENTS SUPPLEMENTAL MATERIAL Dr. Dumas, Dr. Meta, Ms. Borri, Dr. Goveia, and Dr. Rohlenova de- This article contains the following supplemental material signed all experiments. Dr. Dumas, Dr. Meta, Ms. Borri, Dr. Chen, online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ and Dr. Lu setup renal endothelial cell isolation protocol and per- ASN.2019080832/-/DCSupplemental. formed in vivo and cell culture experiments, as well as biochemical Supplemental Figure 1. EC selection and data metrics for gRECs, and molecular assays. Dr. Dumas, Dr. Meta, Ms. Borri, Ms. Con- cRECs, and mRECs. chinha, Dr. Falkenberg, and Dr. Teuwen prepared scRNA-seq sam- Supplemental Figure 2. Identiﬁcation of gREC and cREC subclusters. ples. Dr. de Rooij and Ms. Parys performed FACS sorting of RECs. Supplemental Figure 3. Identiﬁcation of mREC subclusters. Dr. Lin and Prof. Luo performed 10X single-cell sequencing. Dr. Supplemental Figure 4. gREC, cREC, and mREC subcluster Goveia, Dr. Khan, and Dr. Karakach preprocessed scRNA-seq data. analysis. Dr.Dumas,Dr.Meta,Dr.Goveia,andDr.Rohlenovaperformedthe Supplemental Figure 5. REC molecular adaptation to dehydration. bioinformatics analysis. Dr. Dumas and Dr. Vinckier performed Supplemental Figure 6. Response of mREC subpopulations to immunostaining and confocal imaging. Dr. Dumas, Dr. Meta, Ms. dehydration. Borri, Dr. Kalucka, and Ms. De Legher performed metabolic ex- Supplemental Figure 7. Molecular and metabolic adaptation of periments. Mr. Taverna setup the online database. Dr. Schoonjans, gRECs to dehydration. Dr. Lin, Prof. Bolund, Prof. Dewerchin, Dr. Eelen, Prof. Rabelink, Supplemental Figure 8. In vitro dehydration model characterization. Prof. Li, and Prof. Luo provided advice and discussed results. Supplemental Appendix 1. Supplemental methods, metabolic Dr.Dumas,Dr.Meta,andProf.Carmelietwrotethemanuscript. assays (detection of organic osmolytes), Western blot, and histology Prof. Carmeliet conceptualized the study. All authors discussed and immunohistochemistry (renal hypoxia). results and commented on the manuscript. Supplemental Table 1. scRNA-seq data processing and visualization. We thank Ms. A. Manderveld and Ms. A. Carton for their assis- Supplemental Table 2. Top 50 marker genes for gRECs, cRECs, and tance with immunostaining, Dr. B. Ghesquière for the mass spec- mRECs in control. trometry experiments, and laboratory members and Prof. N. Supplemental Table 3. Molecular taxonomy of phenotypes of Chandel (Chicago, IL) for their feedback and discussion. freshly isolated mouse RECs in control. Supplemental Table 4. Top 50 marker genes for gREC, cREC, and mREC phenotypes in control. DISCLOSURES Supplemental Table 5. GSVA of cREC and mREC phenotypes in None. control. Supplemental Table 6. Differential analyses of cRECs, mRECs, and gRECs in control versus different time points of dehydration. FUNDING Supplemental Table 7. GSEA of mRECs in control versus 48-hour dehydration. Dr. Dumas and Dr. Falkenberg are supported by Marie Curie-individual fellowships. Ms. Conchinha, Dr. Goveia, Dr. Kalucka, and Dr. Rohlenova are supported by the “Fonds voor Wetenschappelijk Onderzoek” (FWO), Dr. Teuwen by University of Antwerp. Dr. Khan is supported by a grant REFERENCES from Kom op Tegen Kanger (Stand up to Cancer, Flemish Cancer Society). Prof. Bolund and Prof. Luo are supported by the Sanming Project of 1. Molema G, Aird WC: Vascular heterogeneity in the kidney. Semin Medicine in Shenzhen (SZSM20162074). Dr. Lin is supported by the Nephrol 32: 145–155, 2012 Lundbeck Foundation (R219-2016-1375) and a Sapere Aude starting grant 2. Jourde-Chiche N, Fakhouri F, Dou L, Bellien J, Burtey S, Frimat M, et al.: from the Danmarks Frie Forskningsfond (8048-00072A). The work of Endothelium structure and function in kidney health and disease. Nat Prof. Luo is supported by Beijing Genomics Institute-Research, the Danish Rev Nephrol 15: 87–108, 2019 Research Council for Independent Research (DFF-1337-00128), the Sapere 3. Kriz W: Fenestrated glomerular capillaries are unique. JAmSoc Aude Young Research Talent Prize (DFF-1335-00763A), and Aarhus Univer- Nephrol 19: 1439–1440, 2008 sity strategic grant (AU-iCRISPR). Prof. Li is supported by the State Key 4. Eckardt KU, Bernhardt WM, Weidemann A, Warnecke C, Rosenberger Laboratory of Ophthalmology, Zhongshan Ophthalmic Center at the Sun C, Wiesener MS, et al.: Role of hypoxia in the pathogenesis of renal Yat-Sen University and the National Natural Science Foundation of China disease. Kidney Int Suppl S46–S51, 2005

136 JASN JASN 31: 118–138, 2020 www.jasn.org BASIC RESEARCH

5. Pallone TL, Edwards A, Mattson DL: Renal medullary circulation. Compr 27. Cui FX, Liu HQ, Zou ZM, Li H: Metabolic responses to water deprivation Physiol 2: 97–140, 2012 in C57BL/6J mice using a proton nuclear magnetic resonance-based 6. Brunskill EW, Sequeira-Lopez ML, Pentz ES, Lin E, Yu J, Aronow BJ, metabonomics approach. Rsc Adv 5: 80142–80149, 2015 et al.: Genes that confer the identity of the renin cell. J Am Soc Nephrol 28.FaracoG,WijasaTS,ParkL,MooreJ,AnratherJ,IadecolaC:Water 22: 2213–2225, 2011 deprivation induces neurovascular and cognitive dysfunction through 7. Otani H, Kaya M, Tsujita J: Effect of the volume of fluid ingested on vasopressin-induced oxidative stress. J Cereb Blood Flow Metab urine concentrating ability during prolonged heavy exercise in a hot 34: 852–860, 2014 environment. J Sports Sci Med 12: 197–204, 2013 29. Morris M, Li P, Callahan MF, Oliverio MI, Coffman TM, Bosch SM, et al.: 8. Popkin BM, D’Anci KE, Rosenberg IH: Water, hydration, and health. Neuroendocrine effects of dehydration in mice lacking the angiotensin Nutr Rev 68: 439–458, 2010 AT1a receptor. Hypertension 33: 482–486, 1999 9. Bongers CCWG, Alsady M, Nijenhuis T, Tulp ADM, Eijsvogels TMH, 30. Nagakura A, Hiyama TY, Noda M: Na(x)-deficient mice show normal Deen PMT, et al.: Impact of acute versus prolonged exercise and vasopressin response to dehydration. Neurosci Lett 472: 161–165, dehydration on kidney function and injury. Physiol Rep 6: e13734, 2010 2018 31. Berl T: How do kidney cells adapt to survive in hypertonic inner me- 10. Rowland NE: Food or fluid restriction in common laboratory animals: dulla? Trans Am Clin Climatol Assoc 120: 389–401, 2009 Balancing welfare considerations with scientific inquiry. Comp Med 32. Burg MB, Ferraris JD, Dmitrieva NI: Cellular response to hyperosmotic 57: 149–160, 2007 stresses. Physiol Rev 87: 1441–1474, 2007 11. Satija R, Farrell JA, Gennert D, Schier AF, Regev A: Spatial re- 33. Hall A: Rho GTPases and the actin cytoskeleton. Science 279: 509–514, construction of single-cell gene expression data. Nat Biotechnol 33: 1998 495–502, 2015 34. Ikebe M, Nonoguchi H, Nakayama Y, Tashima Y, Tomita K: Upregula- 12. van den Brink SC, Sage F, Vértesy Á, Spanjaard B, Peterson-Maduro J, tion of the secretory-type Na(+)/K(+)/2Cl(-)-cotransporter in the kidney Baron CS, et al.: Single-cell sequencing reveals dissociation-induced by metabolic acidosis and dehydration in rats. JAmSocNephrol12: gene expression in tissue subpopulations. Nat Methods 14: 935–936, 423–430, 2001 2017 35. Manotham K, Tanaka T, Ohse T, Kojima I, Miyata T, Inagi R, et al.: A 13. Aibar S, González-Blas CB, Moerman T, Huynh-Thu VA, Imrichova H, biologic role of HIF-1 in the renal medulla. Kidney Int 67: 1428–1439, Hulselmans G, et al.: SCENIC: Single-cell regulatory network inference 2005 and clustering. Nat Methods 14: 1083–1086, 2017 36. Bauerschmidt C, Arrichiello C, Burdak-Rothkamm S, Woodcock M, Hill 14. Yu G, Wang LG, Han Y, He QY: clusterProfiler: An R package for com- MA, Stevens DL, et al.: Cohesin promotes the repair of ionizing radia- paring biological themes among gene clusters. OMICS 16: 284–287, tion-induced DNA double-strand breaks in replicated chromatin. Nu- 2012 cleic Acids Res 38: 477–487, 2010 15. Cannoodt R, Saelens W, Saeys Y: Computational methods for trajectory 37. Xu H, Balakrishnan K, Malaterre J, Beasley M, Yan Y, Essers J, et al.: inference from single-cell transcriptomics. Eur J Immunol 46: 2496– Rad21-cohesin haploinsufficiency impedes DNA repair and enhances 2506, 2016 gastrointestinal radiosensitivity in mice. PLoS One 5: e12112, 2010 16. Kiselev VY, Yiu A, Hemberg M: Scmap: Projection of single-cell RNA- 38. Sands JM, Layton HE: The physiology of urinary concentration: An seq data across data sets. Nat Methods 15: 359–362, 2018 update. Semin Nephrol 29: 178–195, 2009 17. De Bock K, Georgiadou M, Schoors S, Kuchnio A, Wong BW, Cantelmo 39. Weigand KM, Swarts HG, Fedosova NU, Russel FG, Koenderink JB: AR, et al.: Role of PFKFB3-driven glycolysis in vessel sprouting. Cell Na,K-ATPase activity modulates Src activation: A role for ATP/ADP 154: 651–663, 2013 ratio. Biochim Biophys Acta 1818: 1269–1273, 2012 18. Carmeliet P, Moons L, Luttun A, Vincenti V, Compernolle V, De Mol M, 40. Whittam R, Willis JS: Ion movements and oxygen consumption in kid- et al.: Synergism between vascular endothelial growth factor and ney cortex slices. JPhysiol168: 158–177, 1963 placental growth factor contributes to angiogenesis and plasma 41. Aguer C, Gambarotta D, Mailloux RJ, Moffat C, Dent R, McPherson R, extravasation in pathological conditions. Nat Med 7: 575–583, et al.: Galactose enhances oxidative metabolism and reveals mito- 2001 chondrial dysfunction in human primary muscle cells. PLoS One 6: 19. Cantelmo AR, Conradi LC, Brajic A, Goveia J, Kalucka J, Pircher A, et al.: e28536, 2011 Inhibition of the glycolytic activator PFKFB3 in endothelium induces 42. Marroquin LD, Hynes J, Dykens JA, Jamieson JD, Will Y: Circumventing tumor vessel normalization, impairs metastasis, and improves chemo- the Crabtree effect: Replacing media glucose with galactose increases therapy. Cancer Cell 30: 968–985, 2016 susceptibility of HepG2 cells to mitochondrial toxicants. Toxicol Sci 97: 20. Efe O, Klein JD, LaRocque LM, Ren H, Sands JM: Metformin improves 539–547, 2007 urine concentration in rodents with nephrogenic diabetes insipidus. 43. Warburg O, Geissler AW, Lorenz S: [On growth of cancer cells in media JCI Insight 1: 88409, 2016 in which glucose is replaced by galactose]. Hoppe Seylers Z Physiol 21. Karaiskos N, Rahmatollahi M, Boltengagen A, Liu H, Hoehne M, Chem 348: 1686–1687, 1967 Rinschen M, et al.: A single-cell transcriptome atlas of the mouse glo- 44. Detaille D, Guigas B, Chauvin C, Batandier C, Fontaine E, Wiernsperger merulus. J Am Soc Nephrol 29: 2060–2068, 2018 N, et al.: Metformin prevents high-glucose-induced endothelial cell 22. Takemoto M, He L, Norlin J, Patrakka J, Xiao Z, Petrova T, et al.: Large- death through a mitochondrial permeability transition-dependent scale identification of genes implicated in kidney glomerulus devel- process. Diabetes 54: 2179–2187, 2005 opment and function. EMBO J 25: 1160–1174, 2006 45. Detaille D, Vial G, Borel AL, Cottet-Rouselle C, Hallakou-Bozec S, Bolze 23. Messaoudi S, He Y, Gutsol A, Wight A, Hébert RL, Vilmundarson RO, S, et al.: Imeglimin prevents human endothelial cell death by inhibiting et al.: Endothelial Gata5 transcription factor regulates blood pressure. mitochondrial permeability transition without inhibiting mitochondrial Nat Commun 6: 8835, 2015 respiration. Cell Death Discov 2: 15072, 2016 24. dela Paz NG, D’Amore PA: Arterial versus venous endothelial cells. Cell 46. Kurtz L, Madsen K, Kurt B, Jensen BL, Walter S, Banas B, et al.: High- Tissue Res 335: 5–16, 2009 level connexin expression in the human juxtaglomerular apparatus. 25. Neuhofer W, Beck FX: Cell survival in the hostile environment of the Nephron, Physiol 116: 1–8, 2010 renal medulla. Annu Rev Physiol 67: 531–555, 2005 47. Castellanos Rivera RM, Monteagudo MC, Pentz ES, Glenn ST, Gross 26. Bekkevold CM, Robertson KL, Reinhard MK, Battles AH, Rowland NE: KW, Carretero O, et al.: Transcriptional regulator RBP-J regulates the Dehydration parameters and standards for laboratory mice. JAmAssoc number and plasticity of renin cells. Physiol Genomics 43: 1021–1028, Lab Anim Sci 52: 233–239, 2013 2011

JASN 31: 118–138, 2020 Single-Cell Atlas of Renal Endothelium 137 BASIC RESEARCH www.jasn.org

48. Ito S, Abe K: Contractile properties of afferent and efferent arterioles. 58. Kinne RK: The role of organic osmolytes in osmoregulation: From Clin Exp Pharmacol Physiol 24: 532–535, 1997 bacteria to mammals. JExpZool265: 346–355, 1993 49. Noda Y, Horikawa S, Katayama Y, Sasaki S: Identification of a mul- 59. Brezis M, Rosen S: Hypoxia of the renal medulla--its implications for tiprotein “motor” complex binding to water channel aquaporin-2. disease. N Engl J Med 332: 647–655, 1995 Biochem Biophys Res Commun 330: 1041–1047, 2005 60. Bagnasco S, Good D, Balaban R, Burg M: Lactate production in isolated 50. Pallone TL, Turner MR, Edwards A, Jamison RL: Countercurrent ex- segments of the rat nephron. Am J Physiol 248: F522–F526, 1985 change in the renal medulla. Am J Physiol Regul Integr Comp Physiol 61. Eelen G, de Zeeuw P, Simons M, Carmeliet P: Endothelial cell metab- 284: R1153–R1175, 2003 olism in normal and diseased vasculature. Circ Res 116: 1231–1244, 51. Lee SD, Choi SY, Lim SW, Lamitina ST, Ho SN, Go WY, et al.: TonEBP 2015 stimulates multiple cellular pathways for adaptation to hypertonic 62. Mellanby K: Metabolic water and desiccation. Nature 150: 21, 1942 stress: Organic osmolyte-dependent and -independent pathways. 63. Ziegler EE, Filer LJ, International Life Sciences Institute-Nutrition Am J Physiol Renal Physiol 300: F707–F715, 2011 Foundation: Present Knowledge in Nutrition,Washington,D.C.,ILSI 52. Krishnan D, Pan W, Beggs MR, Trepiccione F, Chambrey R, Eladari D, Press, International Life Sciences Institute, 1996 et al.: Deficiency of carbonic anhydrase II results in a urinary concen- 64. Takei Y, Bartolo RC, Fujihara H, Ueta Y, Donald JA: Water deprivation trating defect. Front Physiol 8: 1108, 2018 induces appetite and alters metabolic strategy in Notomys alexis: 53. Blum JS, Wearsch PA, Cresswell P: Pathways of antigen processing. Unique mechanisms for water production in the desert. Proc Biol Sci Annu Rev Immunol 31: 443–473, 2013 279: 2599–2608, 2012 54. Dall’Asta V, Bussolati O, Sala R, Parolari A, Alamanni F, Biglioli P, et al.: 65. Rutkowska J, Sadowska ET, Cichon M, Bauchinger U: Increased fat Amino acids are compatible osmolytes for volume recovery after catabolism sustains water balance during fasting in zebra finches. JExp hypertonic shrinkage in vascular endothelial cells. Am J Physiol Biol 219: 2623–2628, 2016 276: C865–C872, 1999 66.ForetzM,GuigasB,BertrandL,PollakM,ViolletB:Metformin:From 55. Law RO: Amino acids as volume-regulatory osmolytes in mammalian mechanisms of action to therapies. Cell Metab 20: 953–966, 2014 cells. Comp Biochem Physiol A Comp Physiol 99: 263–277, 1991 67. Rena G, Hardie DG, Pearson ER: The mechanisms of action of 56. Lang F, Busch GL, Ritter M, Völkl H, Waldegger S, Gulbins E, et al.: metformin. Diabetologia 60: 1577–1585, 2017 Functional significance of cell volume regulatory mechanisms. Physiol Rev 78: 247–306, 1998 57. Handler JS, Kwon HM: Regulation of renal cell organic osmolyte See related editorial, “Endothelial Cell Identity, Heterogeneity and Plasticity in transport by tonicity. Am J Physiol 265: C1449–C1455, 1993 the Kidney,” on pages 1–2.

AFFILIATIONS

1Department of Oncology, Laboratory of Angiogenesis and Vascular Metabolism, Katholieke Universiteit Leuven (KU Leuven), Leuven, Belgium; 2Laboratory of Angiogenesis and Vascular Metabolism, Center for Cancer Biology, Vlaams Instituut voor Biotechnologie (VIB), Leuven, Belgium; 3State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-Sen University, Guangzhou, Guangdong, China; 4Lars Bolund Institute of Regenerative Medicine, Beijing Genomics Institute (BGI)-Qingdao, BGI-Shenzhen, Qingdao, China; 5Department of Biomedicine, Aarhus University, Aarhus, Denmark; 6Division of Nephrology, Department of Internal Medicine, The Einthoven Laboratory for Vascular and Regenerative Medicine, Leiden University Medical Center, Leiden, The Netherlands; 7China National GeneBank, Beijing Genomics Institute (BGI)-Shenzhen, Shenzhen, China; and 8Qingdao-Europe Advanced Institute for Life Sciences, Beijing Genomics Institute (BGI)-Qingdao, Qingdao, China

138 JASN JASN 31: 118–138, 2020 2019-09-27 1

SUPPLEMENTAL MATERIAL

SINGLE CELL RNA SEQUENCING REVEALS RENAL ENDOTHELIUM HETEROGENEITY AND (METABOLIC) ADAPTATION TO WATER DEPRIVATION

Sébastien J. Dumas1-3*, Elda Meta1,3*, Mila Borri1,3*, Jermaine Goveia1,3, Katerina Rohlenova1,3, Nadine V. Conchinha1,3, Kim Falkenberg1,3, Laure-Anne Teuwen1,3, Laura de Rooij1,3, Joanna Kalucka1,3, Rongyuan Chen2, Shawez Khan1,3, Federico Taverna1,3, Weisi Lu2, Magdalena Parys1,3, Carla De Le- gher1,3, Stefan Vinckier1,3, Tobias K. Karakach1,3&, Luc Schoonjans1-3, Lin Lin4,5, Lars Bolund4,5, Mieke Dewerchin1,3, Guy Eelen1,3, Ton J. Rabelink6, Xuri Li2#, Yonglun Luo4,5,7,8#, Peter Carmeliet1-3#

SUPPLEMENTAL MATERIAL: TABLE OF CONTENTS

Supplemental Figures

Figure S1. EC selection and data metrics for gRECs, cRECs and mRECs

Figure S2. Identification of gREC and cREC subclusters

Figure S3. Identification of mREC subclusters

Figure S4. gREC, cREC and mREC subcluster analysis

Figure S5. REC molecular adaptation to dehydration

Figure S6. Response of mREC subpopulations to dehydration.

Figure S7. Molecular and metabolic adaptation of gRECs to dehydration

Figure S8. In vitro dehydration model characterization

Supplemental Methods

Metabolic Assays: detection of organic osmolytes

Western Blot

Histology and Immunohistochemistry: renal hypoxia

Supplemental Tables

Table S1. scRNA-seq data processing and visualization

Table S2. Top 50 marker genes for gRECs, cRECs and mRECs in control

Table S3. Molecular taxonomy of phenotypes of freshly isolated mouse RECs in control

Table S4. Top 50 marker genes for gREC, cREC and mREC phenotypes in control

Table S5. Geneset variation analysis of cREC and mREC phenotypes in control

Table S6. Differential analyses of cRECs, mRECs and gRECs in control vs different timepoints of dehydration

Table S7. Geneset enrichment analyses of mRECs in control vs 48 hour-dehydration

2019-09-27 2

SUPPLEMENTARY FIGURE LEGENDS

Figure S1. EC SELECTION AND DATA METRICS FOR GRECS, CRECS AND MRECS (RELATED TO FIGURE 1, 2 AND 3)

(A) FACS strategy to sort RECs from kidney cortex, medulla and glomeruli. cRECs and mRECs were purified by FACS, sorting CD45-CD102+ cells, excluding CD45+ leukocytes. gREC were isolated by FACS sorting CD45-CD73-CD102+ cells, thereby excluding CD45+ leukocytes and CD73+ mesangial cells. (B) t-SNE plots of RECs from the three kidney compartments in control and dehydration conditions, color-coded for the expression of the indicated markers (Pecam1, Cdh5 and Icam2 for ECs; Hbb-a1, Hbb-a2 and Hbb- bs for red blood cells that were excluded from downstream analyses). Red arrowheads indicate cells with high expression of the indicated marker on the t-SNE plot. Scale: light blue is low expression, red is high expression. (C-E) Bar graphs showing the number and percentage of analyzed RECs per cluster (left panel), and violin plots showing the number of genes and unique molecular identifiers (UMIs) (right panel) for gRECs (C), cRECs (D) and mRECs (E) in control conditions.

Figure S2. IDENTIFICATION OF GREC AND CREC SUBCLUSTERS (RELATED TO FIGURE 2)

(A) t-SNE plots of gRECs from the control condition, color-coded for the expression of the indicated markers. Red arrowheads indicate cells with high expression of the indicated marker on the t-SNE plot. Scale: light blue is low expression, red is high expression. (B) Expression level-scaled heatmap of cell- cell interaction-related genes in gRECs from the control condition. (C) t-SNE plots of cRECs from the control condition, color-coded for the expression of the indicated markers. Red arrowheads indicate cells with high expression of the indicated marker on the t-SNE plot. Scale: light blue is low expression, red is high expression. (D) Representative images of mouse kidney sections used as negative controls for the micrographs displayed in Figure 2H. Incubation with CD105 (red), endomucin (grey) and CA-VIII (green) was omitted. Nuclei are counterstained with Hoechst (blue). Scale bar, 200 µm.

2019-09-27 3

Figure S3. IDENTIFICATION OF MREC SUBCLUSTERS (RELATED TO FIGURE 3)

(A) t-SNE plots of mRECs from the control condition, color-coded for the expression of the indicated markers. Red arrowheads indicate cells with high expression of the indicated marker on the t-SNE plot. Scale: light blue is low expression, red is high expression. (B) Representative images of mouse kidney sections used as negative controls for the micrographs displayed in Figures 3D-F. Incubation with CD105 (red), PLVAP (grey) and CRYAB (green) was omitted. Nuclei are counterstained with Hoechst (blue). Scale bar, 200 µm.

Figure S4. GREC, CREC AND MREC SUBCLUSTER ANALYSIS (RELATED TO FIGURES 2 AND 3)

(A,B) Expression level-scaled heatmap of interferon activated- and antigen presentation-related genes (A) and angiogenesis-related genes (B) in cRECs and mRECs from the control condition. Scale: light blue is low expression, red is high expression. (C) Correlation heatmap of all gREC, cREC and mREC subclusters. Scale: red indicates a high transcriptome similarity, blue indicates a low transcriptome similarity.

Figure S5. REC MOLECULAR ADAPTATION TO DEHYDRATION (RELATED TO FIGURE 4 AND 6)

(A) Body weight of mice subjected to control condition or water deprivation over time, expressed as percentage of initial body weight. (B-H) Urine osmolality (mOsm/kg) (B), plasma osmolality (mOsm/kg) (C), plasma sodium level (mM) (D), plasma total protein level (g/L) (E), plasma AST activity (U/L) (F), plasma LDH activity (U/L) (G), and plasma urea level (mg/dL) (H) of mice subjected to control condition or water deprivation over time. (I) Venn diagram of the top 50 up- (left) and downregulated (right) genes in RECs from the three kidney compartments at 12 hours of dehydration (top), 24 hours of dehydration (middle) and 36 hours of dehydration (bottom). Note for the 36 hours timepoint: the cREC sample did not meet sequencing quality standards and was therefore not included in downstream analyses. (J) (Top) Representative micrographs of kidney sections from mice subjected to control condition or 48

2019-09-27 4

hours of water deprivation, stained for the hypoxia probe pimonidazole (brown). (Bottom) Pimonida- zole staining quantification as expressed as percent of measured area for cortex and medulla from mice subjected to control condition or 48 hours of dehydration (DH). Data are mean ± SEM; n=3-6 mice/condition. Statistical tests: unpaired t-test, One- and Two-way ANOVA/Bonferroni, Kruskal-Wallis/Dunn’s. *P<0.05.

Figure S6. RESPONSE OF MRECS SUBPOPULATIONS TO DEHYDRATION. (RELATED TO FIGURE 5)

(A) Principal component analysis plot of Jaccard similarity coefficient of mREC subpopulations in control and dehydrated mice. (B) scMap cluster projection of the control mREC phenotypes to all mRECs (i.e. including mRECs from the 12, 24, 36 and 48 hours of dehydration). Similarity scores of mRECs clusters are provided as boxplots. Unassigned cells are indicated in dark grey.

Figure S7. MOLECULAR AND METABOLIC ADAPTATION OF GRECS TO DEHYDRATION (RELATED TO FIGURE 5 AND 6)

(A) Correlation heatmap of gRECs from the control condition and at different dehydration time points. Scale: red indicates a high transcriptome similarity, blue indicates a low transcriptome similarity. (B) t- SNE plot color-coded for gRECs from the control condition and at different dehydration time points. (C- H) Expression level-scaled heatmaps of genes encoding heat shock proteins, or genes involved in cytoskeleton remodeling, DNA damage and growth arrest, and immediate early genes (C), cell volume regulation related genes (D), ribosome-related genes (E), proteasome-related genes (F), glycolysis-related genes (G) and oxidative phosphorylation-related genes (H), in gRECs from the control condition and at different dehydration time points. (I) Pathway map showing changes in transcript levels of metabolic genes from glycolysis, TCA cycle and OXPHOS in gRECs after 48 hours of dehydration compared with controls. Blue corresponds to low expression levels after dehydration; grey indicates that the change in gene expression did not reach the fold change threshold to be color-coded.

2019-09-27 5

Figure S8. IN VITRO DEHYDRATION MODEL CHARACTERIZATION (RELATED TO FIGURE 7)

(A) Measurement of osmolyte content in control and hyperosmolarity (hyperosm)-exposed ECs: sorbitol, taurine, alanine, glutamine and glycerophosphocholine (n=3). (B) Left: Representative blot of Na+/K+ ATPase α1 subunit and β-actin in control and hyperosmolarity (hyperosm)-exposed ECs. Right: Densi- tometric quantification of the immunoblot signal of the Na+/K+ ATPase α1 subunit in control and hyperosmolarity (hyperosm)-exposed ECs (n=4). (C) Oxygen consumption rate (OCR) of control HUVECs pre- treated with metformin (1 mM, 10 mM, 20 mM) or control vehicle for 1 hour. (n=3). Data are mean ± SEM. Statistical test: unpaired t-test, One-way ANOVA/Bonferroni. *P<0.05.

2019-09-27 6

SUPPLEMENTAL METHODS

METABOLIC ASSAYS

Detection of organic osmolytes: Medium was removed and control HUVECs and hyperosmolarity-exposed HUVECs (900mOsm/kg) were washed with ice cold 0.9% NaCl. Osmolytes were extracted by adding 300 µL of a 80% methanol (in water) extraction buffer containing 2 µM of deuterated (d27) myristic acid (as internal standard) to the cells. Following extraction, precipitated proteins and insolubilities were removed by centrifugation at 20.000 x g for 15 min at 4°C. The supernatant was transferred to the appropriate mass spectrometer vials. Measurements were performed using a Dionex UltiMate 3000 LC System (Thermo Scientific) in-line connected to a Q-Exactive Orbitrap mass spectrometer (Thermo Sci- entific). 15 µl of sample was injected and loaded onto a Hilicon iHILIC-Fusion(P) column (Achrom). A linear gradient was carried out starting with 90% solvent A (LC-MS grade acetonitrile) and 10% solvent B (10 mM ammoniumacetate pH 9.3). From 2 to 20 minutes the gradient changed to 80% B and was kept at 80% until 23 min. Next a decrease to 40% B was carried out to 25 min, further decreasing to 10% B at 27 min. Finally, 10% B was maintained until 35 min. The solvent was used at a flow rate of 200

µl/min, the column temperature was kept constant at 25°C. The mass spectrometer operated in negative ion mode, settings of the HESI probe were as follows: sheath gas flow rate at 35, auxiliary gas flow rate at 10 (at a temperature of 260°C). Spray voltage was set at 4.8 kV, temperature of the capillary at 300°C and S-lens RF level at 50. A full scan (resolution of 140.000 and scan range of m/z 70-1050) was applied. For the data analysis we used an in-house library and sorbitol, glycerophosphocholine, taurine, alanine and glutamine were quantified (area under the curve) using the XCalibur 4.0 (Thermo Scientific) software platform. Measured values were normalized to protein content.

2019-09-27 7

WESTERN BLOT

Protein lysates were separated by SDS-PAGE under reducing conditions, transferred to a nitrocellulose membrane, and analyzed by immunoblotting. Primary antibody used was rabbit anti-Na+/K+ ATPase α1 subunit (1/1000, 3010, Cell Signaling) and mouse anti-β actin (1/1000, A5441, Sigma) in 5% bovine serum albumin (BSA), appropriate secondary antibody was from Cell Signaling Technology (1:2000, Anti- Rabbit IgG HRP-linked #7074; 1:2000, Anti-Mouse IgG HRP-linked #7076) in 5% bovine serum albumin (BSA). Signal was detected using the ECL system (Pierce) according to the manufacturer’s instructions.

Densitometric quantifications of bands were done with Fiji software (https://fiji.sc).

HISTOLOGY AND IMMUNOHISTOCHEMISTRY

Renal Hypoxia: Renal hypoxia was detected after injection of 60 mg/kg pimonidazole hydrochloride (Hypoxyprobe kit, Chemicon-Millipore, Merck) into 48h-dehydrated and normally hydrated mice (kidneys were collected 2 hours after injection). To visualize the formation of pimonidazole adducts, kidney paraffin sections were immunostained with Hypoxyprobe-1-Mab1 following the manufacturer’s instructions and counterstained with hematoxylin. Pimonidazole staining was quantified using Leica Met- aMorph AF 2.1 morphometry software package (Leica).

2019-09-27 8

SUPPLEMENTARY TABLES

Table S1. SCRNA-SEQ DATA PROCESSING AND VISUALIZATION (RELATED TO FIGURE 1-5, FIGURE S1-S3; S6)

Table S2. TOP 50 MARKER GENES FOR GRECS, CRECS AND MRECS IN CONTROL (RELATED TO FIGURE 1)

Table S3. MOLECULAR TAXONOMY OF PHENOTYPES OF FRESHLY ISOLATED MOUSE RECS IN CONTROL (RELATED TO FIGURE

2-3, FIGURE S2-S4)

Table S4. TOP 50 MARKER GENES FOR GREC, CREC AND MRECS PHENOTYPES IN CONTROL (RELATED TO FIGURE 2-3,

FIGURE S2-S3)

Table S5. GENESET VARIATION ANALYSIS OF CREC AND MREC PHENOTYPES IN CONTROL (RELATED TO FIGURE 2-3, FIGURE S2-S4)

Table S6. DIFFERENTIAL ANALYSES OF CRECS, MRECS AND GRECS IN CONTROL VS DIFFERENT TIMEPOINTS OF DEHYDRATION

(RELATED TO FIGURE 4, FIGURE S5)

Table S7. GENESET ENRICHMENT ANALYSES OF MRECS IN CONTROL VS 48 HOUR DEHYDRATION (RELATED TO FIGURE 6)

Table S3: Molecular taxonomy of phenotypes of freshly isolated mouse RECs in control conditions. Related to Figures 2 and 3.

NOTE 1: Expression patterns of all genes in Tables S4, S5 and S6 (also of those not listed in the text, figures or Tables S4, S5 and S6) can be explored via the accompanying online web tool available from https://endotheliomics.shinyapps.io/rec_dehydration/ (username: [email protected]; password: scRECpaper).

KIDNEY CLUS COMPARMENT ASSIGNED PHENOTYPES -TER EXPRESSED GENES TYPICAL OF FIG. TABLE NR • arterial ECs (Sox17, Sema3g, Gja4) (Corada et al., 2013; Fang et al., 2017; Kutschera et al., 2011) S2A • vascular integrity & elastic fiber assembly (Ltbp4, Fbln5, Bmp4) (Noda et al., 2013; Tojais et al., 2017) S2A • tight junctions (Cldn5) (Morita et al., 1999) afferent arteriole G1 • Connexin 37 (Gja4), known to be present in gRECs from afferent arterioles but not efferent arterioles (Zhang and Hill, 2005) • vasotone regulation (Edn1, Alox12, S1pr1) (Cantalupo et al., 2017; Ma et al., 1991; Takeya et al., 2015; Yiu et al., 2B, S2A 2003) • intermediate capillary-like and arterial-like EC phenotype (Kdr, CD300lg, Efnb2, Dll4, Gja4, Gja5) (Buschmann et al., 2B 2010; Fang et al., 2017; Kamba et al., 2006; Rosivall and Peti-Peterdi, 2006; Shutter et al., 2000; Wang et al., 1998; S2A, S2B Zhao et al., 2018) • gap junctions (Gja5), known to be present in gRECs from afferent arterioles but not efferent arterioles (Zhang and S2A portion of the Hill, 2005), may contribute to the tubuloglomerular feedback in the juxtaglomerular apparatus (Just et al., 2009; afferent arteriole GLOMERULI G2 Kurtz et al., 2010; Sorensen et al., 2012) S4 associated with • chemokine receptor, described to be expressed by RECs in contact with renin+ cells at this location (Cxcr4) 2B, S2A, the juxtaglomerular (Takabatake et al., 2009) S2B apparatus • cell-cell interaction (Dkk2, Gja5, Dll4, Efnb2, Cx3cl1, Cxcr4, Ntn4, Lifr, Nrp1, Fas, F2r, Cdh5, Tbxa2r, Unc5b) 2B, S2B (Cunningham et al., 2000; Davis et al., 2019; Imaizumi et al., 2004; Just et al., 2009; Komhoff et al., 1998; Kurtz et al., 2010; Lejmi et al., 2008; Min et al., 2011; Sata et al., 2000; Shutter et al., 2000; Sorensen et al., 2012; Takabatake et al., 2009; Wang et al., 1998; Welti et al., 2013; Yamaguchi et al., 2012) • capillary ECs (Kdr) (Kamba et al., 2006) 2B, S2A • known glomerular capillary EC marker (Ehd3) (Patrakka et al., 2007) 2B capillary G3 • Tgfβ/bmp signaling pathways (Eng, Smad6, Smad7, Xiap, Hipk2), involved in glomerular capillary formation (Cai et al., 2012; Liu et al., 1999; Shang et al., 2013; Ueda et al., 2008; Van Themsche et al., 2010) portion of the G4 • intermediate capillary-like and arterial-like EC phenotype (Kdr, Sox17) (Corada et al., 2013; Kamba et al., 2006) S2A efferent arteriole • absence/low expression of connexins 37 and 40 (Gja4, Gja5) (Zhang and Hill, 2005) S2A 2

associated with the • immune cell adhesion & extravasation, endothelial permeability (CD9, Rdx, Gas6, Podxl, Sgk1, Pde2a, Clic1, Icam2, juxtaglomerular Endrb) (Halai et al., 2014; Horrillo et al., 2016; Koehl et al., 2017; Koss et al., 2006; Ni et al., 2019; Reyes et al., 2018; apparatus Su et al., 2014; Surapisitchat et al., 2007; Xu et al., 2016) • arterial ECs (Sox17) (Corada et al., 2013) S2A • absence/low expression of connexins 37 and 40 (Gja4, Gja5) (Zhang and Hill, 2005) S2A • vasotone regulation (Calca) (Reslerova and Loutzenhiser, 1998) 2B, S2A efferent arteriole G5 • prevention of coagulation (Thbd) (Isermann et al., 2001) 2B • hyperosmolarity-responsive genes (Klf4, S100a4, Slc6a6, Cryab, S100a6, Ptprr, CD200, Ebf1, Slc38a2) (Alfieri et al., 2001; Izumi et al., 2015; Maallem et al., 2008; Schulze Blasum et al., 2016) • arterial ECs (Sox17, Sema3g, Gja4, Gja5, Jag1) (Buschmann et al., 2010; Corada et al., 2013; Fang et al., 2017; High S2C et al., 2008; Kutschera et al., 2011) • suppression of calcification (Mgp) (Bjorklund et al., 2018) • tight junction (Cldn5) (Morita et al., 1999) large artery C1 • vascular integrity & elastic fiber assembly (Eln, Ltbp4, Fbln5, Fbln2, Bmp4) (Chapman et al., 2010; Noda et al., 2013; 2E Tojais et al., 2017; Wagenseil and Mecham, 2012; Walker et al., 2015) • vasotone regulation (Ace, Edn1, S1pr1) (Arendshorst et al., 1990; Ma et al., 1991; Yiu et al., 2003) • response to shear stress (Pi16) (Hazell et al., 2016) • arteriolar ECs (Sox17, Cxcl12) (Corada et al., 2013; Poulos et al., 2018) S2C arteriole C2 • vasotone regulation (Alox12) (Cantalupo et al., 2017; Ma et al., 1991; Takeya et al., 2015; Yiu et al., 2003) • kidney function biomarker & angiostatic mediator (Cst3) (Benndorf, 2018; Li et al., 2018; Shlipak et al., 2013) • arteriolar ECs (Sox17, Kitl) (Corada et al., 2013; Poulos et al., 2018) S2C • vasotone regulation (Calca) (Reslerova and Loutzenhiser, 1998) S2C efferent arteriole C3 • prevention of coagulation (Thbd) (Isermann et al., 2001) CORTEX S4 • shared marker genes with cluster G5 (Calca, Thbd, Rpl8, S100a6, CD200, Rplp0, Ifi27l2a…) 2E • fenestrated capillaries ECs (Kdr, Plvap) (Dimke et al., 2015; Stan et al., 1999) S2C capillary #1 C4 • lipid metabolism (Plpp3, Apoe, Thrsp) (Busnelli et al., 2018; Huang and Mahley, 2014; Yao et al., 2016) 2E, S2C • microvascular remodeling (Id3) (Lee et al., 2014) • fenestrated capillaries ECs (Kdr, Plvap) (Dimke et al., 2015; Stan et al., 1999) 2E, S2C • VEGF receptors (Kdr, Flt1, Nrp1) (Welti et al., 2013) 2E, S2C capillary #2 C5 • Insulin-growth factor binding (Igfbp5, Igfbp3, Insr) (Pollak, 2012) • blood volume & sodium excretion regulation (Npr3) (Matsukawa et al., 1999; Potter, 2011) • intermediate fenestrated capillary-like and vein-like EC phenotype (Kdr, Plvap, Nr2f2) (Dimke et al., 2015; Stan et S2C al., 1999; You et al., 2005) • regulation of endothelial permeability (Jup) (Nottebaum et al., 2008) postcapillary venule C6 • extracellular matrix (Tnxb, Hspg2, Ltbp1) (Gubbiotti et al., 2017; Ikuta et al., 2001; Robertson et al., 2015; Unsold et 2E al., 2001; Valcourt et al., 2015)

• vascular development/angiogenesis (Pbx1, Arghap31) (Caron et al., 2016; Charboneau et al., 2005) • vein ECs (Nr2f2, Plvap) (Stan et al., 1999; You et al., 2005) S2C vein C7 • immune cell adhesion & extravasation, endothelial permeability (Cd9, Gas6) (Ni et al., 2019; Reyes et al., 2018) • fenestrated capillary ECs (Kdr, Plvap) (Dimke et al., 2015; Stan et al., 1999) S2C capillary/angiogenic C8 • angiogenic EC markers (Gpihbp1, Esm1, Col4a1, Col4a2, Trp53i11, Apln, Aplnr, Plk2, Fscn1) (del Toro et al., 2010; 2E, Yang et al., 2015; Zhao et al., 2018) S2C, S4B • fenestrated capillary ECs (Kdr, Plvap) (Dimke et al., 2015; Stan et al., 1999) S2C capillary/interferon C9 • interferon-stimulated genes (Isg15, Ifit1, Ifit3, Ifi203, Ifit3b, Ifit2, Irf7, Ifi204) (Schneider et al., 2014) 2E, S2C • arteriolar ECs (Sox17, Fbln5, Kitl) (Corada et al., 2013; Poulos et al., 2018) S3A • vascular integrity & elastic fiber assembly (Ltbp4) (Noda et al., 2013) arteriole M1 • similar marker genes found in arterioles from cortex and glomeruli: G1, G5, C2 and C3 (Thbd, Calca, Klf4, Tgfb2, 3B, S3A Tm4sf1, Tsc22d1, Id1, Slc6a6, Cd24a, Kitl) • shear stress (Pi16, Klf2, Klf4) (Clark et al., 2011; Hazell et al., 2016; Wang et al., 2010) • arteriolar ECs (Sox17, Gja4, Fbln5, Cxcl12) (Corada et al., 2013; Poulos et al., 2018) S3A • water and urea transport (Aqp1, Slc14a1) (Kim et al., 2002; Pallone et al., 2000) 3B, S3A descending vasa recta M2 • vasotone regulation (Hpgd, Edn1, Adipor2) (Fesus et al., 2007; Silldorff et al., 1995) • tight junctions (Cldn5) (Morita et al., 1999) papillary portion of • arterial ECs (Sox17, Fbln5) (Corada et al., 2013) S3A the descending vasa M3 • vasotone regulation (Alox12) (Ma et al., 1991) recta • hyperosmolarity-responsive genes (s100a4, s100a6) (Nielsen et al., 1995; Schulze Blasum et al., 2016) 3B, S3A • fenestrated capillary ECs (Kdr, Plvap) (Dimke et al., 2015; Stan et al., 1999) S3A MEDULLA • fatty acid transport and metabolism (Cd36, Plpp3) (Busnelli et al., 2018; Son et al., 2018) S4 capillary M4 • VEGF receptors (Kdr, Flt1, Nrp1) (Welti et al., 2013) S3A • blood volume & sodium excretion regulation (Npr3) (Matsukawa et al., 1999; Potter, 2011) • intermediate fenestrated capillary-like and vein-like EC phenotype (Kdr, Plvap, Nr2f2, Ephb4) (Dimke et al., 2015; S3A Stan et al., 1999; Wang et al., 1998; You et al., 2005) postcapillary venule M5 • regulation of endothelial permeability (Jup, Bmpr2, Il6st) (Alsaffar et al., 2018; Benn et al., 2016; Nottebaum et al., 2008) • vein ECs (Nr2f2, Plvap) (Pannabecker and Dantzler, 2006; Stan et al., 1999; You et al., 2005) S3A ascending vasa recta M6 • angiopoietin receptor (Tek), necessary for ascending vasa recta development (Kenig-Kozlovsky et al., 2018) • immune cell adhesion & extravasation, endothelial permeability (Gas6) (Ni et al., 2019) papillary portion of • vein ECs (Nr2f2, Plvap) (Stan et al., 1999; You et al., 2005) S3A the ascending vasa M7 • hyperosmolarity-responsive genes (Cryab, Fxyd2, CD9) (Izumi et al., 2015; Sheikh-Hamad et al., 1996) 3B, S3A recta • anaerobic glycolysis (Ldha, Aldoa, Gapdh) (Chen et al., 2017; Eelen et al., 2015) capillary/angiogenic M8 • fenestrated capillary ECs (Kdr, Plvap) (Dimke et al., 2015; Stan et al., 1999) S3A 4

• angiogenic EC markers (Gpihbp1, Col4a1, Col4a2, Trp53i11, Esm1, Aplnr, Plk2) (del Toro et al., 2010; Yang et al., 3B, S3A, 2015; Zhao et al., 2018) S4B • fenestrated capillary ECs (Kdr, Plvap) (Dimke et al., 2015; Stan et al., 1999) S3A capillary/interferon M9 • interferon-stimulated genes (Isg15, Ifi203, Ifit3b, Ifit2, Irf7, Ifitm3, Ifi204) (Schneider et al., 2014) 3B, S3A ascending vasa • vein ECs (Nr2f2, Plvap) (Stan et al., 1999; You et al., 2005) S3A recta/interferon M10 • interferon-stimulated genes (Ifit3, Ifit1, Ifi44, Iigp1, Irgm1, Ifi35) (Schneider et al., 2014) S3A

References:

Alfieri, R.R., Petronini, P.G., Bonelli, M.A., Caccamo, A.E., Cavazzoni, A., Borghetti, A.F., and Wheeler, K.P. (2001). Osmotic regulation of ATA2 mRNA expression and amino acid transport System A activity. Biochem Biophys Res Commun 283, 174-178.

Alsaffar, H., Martino, N., Garrett, J.P., and Adam, A.P. (2018). Interleukin-6 promotes a sustained loss of endothelial barrier function via Janus kinase-mediated STAT3 phosphorylation and de novo protein synthesis. Am J Physiol Cell Physiol 314, C589-C602.

Arendshorst, W.J., Chatziantoniou, C., and Daniels, F.H. (1990). Role of angiotensin in the renal vasoconstriction observed during the development of genetic hypertension. Kidney Int Suppl 30, S92-96.

Benn, A., Bredow, C., Casanova, I., Vukicevic, S., and Knaus, P. (2016). VE-cadherin facilitates BMP-induced endothelial cell permeability and signaling. J Cell Sci 129, 206-218.

Benndorf, R.A. (2018). Renal Biomarker and Angiostatic Mediator? Cystatin C as a Negative Regulator of Vascular Endothelial Cell Homeostasis and Angiogenesis. J Am Heart Assoc 7, e010997.

Bjorklund, G., Svanberg, E., Dadar, M., Card, D.J., Chirumbolo, S., Harrington, D.J., and Aaseth, J. (2018). The role of matrix Gla protein (MGP) in vascular calcification. Curr Med Chem.

Buschmann, I., Pries, A., Styp-Rekowska, B., Hillmeister, P., Loufrani, L., Henrion, D., Shi, Y., Duelsner, A., Hoefer, I., Gatzke, N., et al. (2010). Pulsatile shear and Gja5 modulate arterial identity and remodeling events during flow-driven arteriogenesis. Development 137, 2187-2196.

Busnelli, M., Manzini, S., Parolini, C., Escalante-Alcalde, D., and Chiesa, G. (2018). Lipid phosphate phosphatase 3 in vascular pathophysiology. Atherosclerosis 271, 156-165.

Cai, J., Pardali, E., Sanchez-Duffhues, G., and ten Dijke, P. (2012). BMP signaling in vascular diseases. FEBS Lett 586, 1993-2002.

Cantalupo, A., Gargiulo, A., Dautaj, E., Liu, C., Zhang, Y., Hla, T., and Di Lorenzo, A. (2017). S1PR1 (Sphingosine-1-Phosphate Receptor 1) Signaling Regulates Blood Flow and Pressure. Hypertension 70, 426-434. 5

Caron, C., DeGeer, J., Fournier, P., Duquette, P.M., Luangrath, V., Ishii, H., Karimzadeh, F., Lamarche-Vane, N., and Royal, I. (2016). CdGAP/ARHGAP31, a Cdc42/Rac1 GTPase regulator, is critical for vascular development and VEGF-mediated angiogenesis. Sci Rep 6, 27485.

Chapman, S.L., Sicot, F.X., Davis, E.C., Huang, J., Sasaki, T., Chu, M.L., and Yanagisawa, H. (2010). Fibulin-2 and fibulin-5 cooperatively function to form the internal elastic lamina and protect from vascular injury. Arterioscler Thromb Vasc Biol 30, 68-74.

Charboneau, A., East, L., Mulholland, N., Rohde, M., and Boudreau, N. (2005). Pbx1 is required for Hox D3-mediated angiogenesis. Angiogenesis 8, 289-296.

Chen, Y., Fry, B.C., and Layton, A.T. (2017). Modeling glucose metabolism and lactate production in the kidney. Math Biosci 289, 116-129.

Clark, P.R., Jensen, T.J., Kluger, M.S., Morelock, M., Hanidu, A., Qi, Z., Tatake, R.J., and Pober, J.S. (2011). MEK5 is activated by shear stress, activates ERK5 and induces KLF4 to modulate TNF responses in human dermal microvascular endothelial cells. Microcirculation 18, 102-117.

Corada, M., Orsenigo, F., Morini, M.F., Pitulescu, M.E., Bhat, G., Nyqvist, D., Breviario, F., Conti, V., Briot, A., Iruela-Arispe, M.L., et al. (2013). Sox17 is indispensable for acquisition and maintenance of arterial identity. Nat Commun 4, 2609.

Cunningham, M.A., Rondeau, E., Chen, X., Coughlin, S.R., Holdsworth, S.R., and Tipping, P.G. (2000). Protease-activated receptor 1 mediates thrombin-dependent, cell- mediated renal inflammation in crescentic glomerulonephritis. J Exp Med 191, 455-462.

Davis, S.M., Collier, L.A., Goodwin, S., Lukins, D.E., Powell, D.K., and Pennypacker, K.R. (2019). Efficacy of leukemia inhibitory factor as a therapeutic for permanent large vessel stroke differs among aged male and female rats. Brain Res 1707, 62-73. del Toro, R., Prahst, C., Mathivet, T., Siegfried, G., Kaminker, J.S., Larrivee, B., Breant, C., Duarte, A., Takakura, N., Fukamizu, A., et al. (2010). Identification and functional analysis of endothelial tip cell-enriched genes. Blood 116, 4025-4033.

Dimke, H., Sparks, M.A., Thomson, B.R., Frische, S., Coffman, T.M., and Quaggin, S.E. (2015). Tubulovascular cross-talk by vascular endothelial growth factor a maintains peritubular microvasculature in kidney. J Am Soc Nephrol 26, 1027-1038.

Eelen, G., de Zeeuw, P., Simons, M., and Carmeliet, P. (2015). Endothelial cell metabolism in normal and diseased vasculature. Circ Res 116, 1231-1244.

Fang, J.S., Coon, B.G., Gillis, N., Chen, Z., Qiu, J., Chittenden, T.W., Burt, J.M., Schwartz, M.A., and Hirschi, K.K. (2017). Shear-induced Notch-Cx37-p27 axis arrests endothelial cell cycle to enable arterial specification. Nat Commun 8, 2149.

Fesus, G., Dubrovska, G., Gorzelniak, K., Kluge, R., Huang, Y., Luft, F.C., and Gollasch, M. (2007). Adiponectin is a novel humoral vasodilator. Cardiovasc Res 75, 719-727.

Gubbiotti, M.A., Neill, T., and Iozzo, R.V. (2017). A current view of perlecan in physiology and pathology: A mosaic of functions. Matrix Biol 57-58, 285-298.

Halai, K., Whiteford, J., Ma, B., Nourshargh, S., and Woodfin, A. (2014). ICAM-2 facilitates luminal interactions between neutrophils and endothelial cells in vivo. J Cell Sci 127, 620-629. 6

Hazell, G.G., Peachey, A.M., Teasdale, J.E., Sala-Newby, G.B., Angelini, G.D., Newby, A.C., and White, S.J. (2016). PI16 is a shear stress and inflammation-regulated inhibitor of MMP2. Sci Rep 6, 39553.

High, F.A., Lu, M.M., Pear, W.S., Loomes, K.M., Kaestner, K.H., and Epstein, J.A. (2008). Endothelial expression of the Notch ligand Jagged1 is required for vascular smooth muscle development. Proc Natl Acad Sci U S A 105, 1955-1959.

Horrillo, A., Porras, G., Ayuso, M.S., and Gonzalez-Manchon, C. (2016). Loss of endothelial barrier integrity in mice with conditional ablation of podocalyxin (Podxl) in endothelial cells. Eur J Cell Biol 95, 265-276.

Huang, Y., and Mahley, R.W. (2014). Apolipoprotein E: structure and function in lipid metabolism, neurobiology, and Alzheimer's diseases. Neurobiol Dis 72 Pt A, 3-12.

Ikuta, T., Ariga, H., and Matsumoto, K.I. (2001). Effect of tenascin-X together with vascular endothelial growth factor A on cell proliferation in cultured embryonic hearts. Biol Pharm Bull 24, 1320-1323.

Imaizumi, T., Yoshida, H., and Satoh, K. (2004). Regulation of CX3CL1/fractalkine expression in endothelial cells. J Atheroscler Thromb 11, 15-21.

Isermann, B., Hendrickson, S.B., Zogg, M., Wing, M., Cummiskey, M., Kisanuki, Y.Y., Yanagisawa, M., and Weiler, H. (2001). Endothelium-specific loss of murine thrombomodulin disrupts the protein C anticoagulant pathway and causes juvenile-onset thrombosis. J Clin Invest 108, 537-546.

Izumi, Y., Yang, W., Zhu, J., Burg, M.B., and Ferraris, J.D. (2015). RNA-Seq analysis of high NaCl-induced gene expression. Physiol Genomics 47, 500-513.

Just, A., Kurtz, L., de Wit, C., Wagner, C., Kurtz, A., and Arendshorst, W.J. (2009). Connexin 40 mediates the tubuloglomerular feedback contribution to renal blood flow autoregulation. J Am Soc Nephrol 20, 1577-1585.

Kamba, T., Tam, B.Y., Hashizume, H., Haskell, A., Sennino, B., Mancuso, M.R., Norberg, S.M., O'Brien, S.M., Davis, R.B., Gowen, L.C., et al. (2006). VEGF-dependent plasticity of fenestrated capillaries in the normal adult microvasculature. Am J Physiol Heart Circ Physiol 290, H560-576.

Kenig-Kozlovsky, Y., Scott, R.P., Onay, T., Carota, I.A., Thomson, B.R., Gil, H.J., Ramirez, V., Yamaguchi, S., Tanna, C.E., Heinen, S., et al. (2018). Ascending Vasa Recta Are Angiopoietin/Tie2-Dependent Lymphatic-Like Vessels. J Am Soc Nephrol 29, 1097-1107.

Kim, Y.H., Kim, D.U., Han, K.H., Jung, J.Y., Sands, J.M., Knepper, M.A., Madsen, K.M., and Kim, J. (2002). Expression of urea transporters in the developing rat kidney. Am J Physiol Renal Physiol 282, F530-540.

Koehl, B., Nivoit, P., El Nemer, W., Lenoir, O., Hermand, P., Pereira, C., Brousse, V., Guyonnet, L., Ghinatti, G., Benkerrou, M., et al. (2017). The endothelin B receptor plays a crucial role in the adhesion of neutrophils to the endothelium in sickle cell disease. Haematologica 102, 1161-1172.

Komhoff, M., Lesener, B., Nakao, K., Seyberth, H.W., and Nusing, R.M. (1998). Localization of the prostacyclin receptor in human kidney. Kidney Int 54, 1899-1908.

Koss, M., Pfeiffer, G.R., 2nd, Wang, Y., Thomas, S.T., Yerukhimovich, M., Gaarde, W.A., Doerschuk, C.M., and Wang, Q. (2006). Ezrin/radixin/moesin proteins are phosphorylated by TNF-alpha and modulate permeability increases in human pulmonary microvascular endothelial cells. J Immunol 176, 1218-1227. 7

Kurtz, L., Madsen, K., Kurt, B., Jensen, B.L., Walter, S., Banas, B., Wagner, C., and Kurtz, A. (2010). High-level connexin expression in the human juxtaglomerular apparatus. Nephron Physiol 116, p1-8.

Kutschera, S., Weber, H., Weick, A., De Smet, F., Genove, G., Takemoto, M., Prahst, C., Riedel, M., Mikelis, C., Baulande, S., et al. (2011). Differential endothelial transcriptomics identifies semaphorin 3G as a vascular class 3 semaphorin. Arterioscler Thromb Vasc Biol 31, 151-159.

Lee, D., Shenoy, S., Nigatu, Y., and Plotkin, M. (2014). Id proteins regulate capillary repair and perivascular cell proliferation following ischemia-reperfusion injury. PLoS One 9, e88417.

Lejmi, E., Leconte, L., Pedron-Mazoyer, S., Ropert, S., Raoul, W., Lavalette, S., Bouras, I., Feron, J.G., Maitre-Boube, M., Assayag, F., et al. (2008). Netrin-4 inhibits angiogenesis via binding to neogenin and recruitment of Unc5B. Proc Natl Acad Sci U S A 105, 12491-12496.

Li, Z., Wang, S., Huo, X., Yu, H., Lu, J., Zhang, S., Li, X., Cao, Q., Li, C., Guo, M., et al. (2018). Cystatin C Expression is Promoted by VEGFA Blocking, With Inhibitory Effects on Endothelial Cell Angiogenic Functions Including Proliferation, Migration, and Chorioallantoic Membrane Angiogenesis. J Am Heart Assoc 7, e009167.

Liu, A., Dardik, A., and Ballermann, B.J. (1999). Neutralizing TGF-beta1 antibody infusion in neonatal rat delays in vivo glomerular capillary formation 1. Kidney Int 56, 1334- 1348.

Ma, Y.H., Harder, D.R., Clark, J.E., and Roman, R.J. (1991). Effects of 12-HETE on isolated dog renal arcuate arteries. Am J Physiol 261, H451-456.

Maallem, S., Wierinckx, A., Lachuer, J., Kwon, M.H., and Tappaz, M.L. (2008). Gene expression profiling in brain following acute systemic hypertonicity: novel genes possibly involved in osmoadaptation. J Neurochem 105, 1198-1211.

Matsukawa, N., Grzesik, W.J., Takahashi, N., Pandey, K.N., Pang, S., Yamauchi, M., and Smithies, O. (1999). The natriuretic peptide clearance receptor locally modulates the physiological effects of the natriuretic peptide system. Proc Natl Acad Sci U S A 96, 7403-7408.

Min, J.K., Park, H., Choi, H.J., Kim, Y., Pyun, B.J., Agrawal, V., Song, B.W., Jeon, J., Maeng, Y.S., Rho, S.S., et al. (2011). The WNT antagonist Dickkopf2 promotes angiogenesis in rodent and human endothelial cells. J Clin Invest 121, 1882-1893.

Morita, K., Sasaki, H., Furuse, M., and Tsukita, S. (1999). Endothelial claudin: claudin-5/TMVCF constitutes tight junction strands in endothelial cells. J Cell Biol 147, 185-194.

Ni, J., Lin, M., Jin, Y., Li, J., Guo, Y., Zhou, J., Hong, G., Zhao, G., and Lu, Z. (2019). Gas6 Attenuates Sepsis-Induced Tight Junction Injury and Vascular Endothelial Hyperpermeability via the Axl/NF-κB Signaling Pathway. Frontiers in Pharmacology 10.

Nielsen, S., Pallone, T., Smith, B.L., Christensen, E.I., Agre, P., and Maunsbach, A.B. (1995). Aquaporin-1 water channels in short and long loop descending thin limbs and in descending vasa recta in rat kidney. Am J Physiol 268, F1023-1037.

Noda, K., Dabovic, B., Takagi, K., Inoue, T., Horiguchi, M., Hirai, M., Fujikawa, Y., Akama, T.O., Kusumoto, K., Zilberberg, L., et al. (2013). Latent TGF-beta binding protein 4 promotes elastic fiber assembly by interacting with fibulin-5. Proc Natl Acad Sci U S A 110, 2852-2857. 8

Nottebaum, A.F., Cagna, G., Winderlich, M., Gamp, A.C., Linnepe, R., Polaschegg, C., Filippova, K., Lyck, R., Engelhardt, B., Kamenyeva, O., et al. (2008). VE-PTP maintains the endothelial barrier via plakoglobin and becomes dissociated from VE-cadherin by leukocytes and by VEGF. J Exp Med 205, 2929-2945.

Pallone, T.L., Edwards, A., Ma, T., Silldorff, E.P., and Verkman, A.S. (2000). Requirement of aquaporin-1 for NaCl-driven water transport across descending vasa recta. J Clin Invest 105, 215-222.

Pannabecker, T.L., and Dantzler, W.H. (2006). Three-dimensional architecture of inner medullary vasa recta. Am J Physiol Renal Physiol 290, F1355-1366.

Patrakka, J., Xiao, Z., Nukui, M., Takemoto, M., He, L., Oddsson, A., Perisic, L., Kaukinen, A., Szigyarto, C.A., Uhlen, M., et al. (2007). Expression and subcellular distribution of novel glomerulus-associated proteins dendrin, ehd3, sh2d4a, plekhh2, and 2310066E14Rik. J Am Soc Nephrol 18, 689-697.

Pollak, M. (2012). The insulin and insulin-like growth factor receptor family in neoplasia: an update. Nat Rev Cancer 12, 159-169.

Potter, L.R. (2011). Natriuretic peptide metabolism, clearance and degradation. FEBS J 278, 1808-1817.

Poulos, M., Redmond, D., Gutkin, M., Ramalingam, P., and Butler, J.M. (2018). Single-Cell Characterization of the HSC-Supportive Bone Marrow Vascular Microenvironment. Blood 132, 2577-2577.

Reslerova, M., and Loutzenhiser, R. (1998). Renal microvascular actions of calcitonin gene-related peptide. Am J Physiol 274, F1078-1085.

Reyes, R., Cardenes, B., Machado-Pineda, Y., and Cabanas, C. (2018). Tetraspanin CD9: A Key Regulator of Cell Adhesion in the Immune System. Front Immunol 9, 863.

Robertson, I.B., Horiguchi, M., Zilberberg, L., Dabovic, B., Hadjiolova, K., and Rifkin, D.B. (2015). Latent TGF-beta-binding proteins. Matrix Biol 47, 44-53.

Rosivall, L., and Peti-Peterdi, J. (2006). Heterogeneity of the afferent arteriole--correlations between morphology and function. Nephrol Dial Transplant 21, 2703-2707.

Sata, M., Suhara, T., and Walsh, K. (2000). Vascular endothelial cells and smooth muscle cells differ in expression of Fas and Fas ligand and in sensitivity to Fas ligand-induced cell death: implications for vascular disease and therapy. Arterioscler Thromb Vasc Biol 20, 309-316.

Schneider, W.M., Chevillotte, M.D., and Rice, C.M. (2014). Interferon-stimulated genes: a complex web of host defenses. Annu Rev Immunol 32, 513-545.

Schulze Blasum, B., Schroter, R., Neugebauer, U., Hofschroer, V., Pavenstadt, H., Ciarimboli, G., Schlatter, E., and Edemir, B. (2016). The kidney-specific expression of genes can be modulated by the extracellular osmolality. FASEB J 30, 3588-3597.

Shang, Y., Doan, C.N., Arnold, T.D., Lee, S., Tang, A.A., Reichardt, L.F., and Huang, E.J. (2013). Transcriptional corepressors HIPK1 and HIPK2 control angiogenesis via TGF-beta- TAK1-dependent mechanism. PLoS Biol 11, e1001527.

Sheikh-Hamad, D., Ferraris, J.D., Dragolovich, J., Preuss, H.G., Burg, M.B., and Garcia-Perez, A. (1996). CD9 antigen mRNA is induced by hypertonicity in two renal epithelial cell lines. Am J Physiol 270, C253-258. 9

Shlipak, M.G., Mattes, M.D., and Peralta, C.A. (2013). Update on cystatin C: incorporation into clinical practice. Am J Kidney Dis 62, 595-603.

Shutter, J.R., Scully, S., Fan, W., Richards, W.G., Kitajewski, J., Deblandre, G.A., Kintner, C.R., and Stark, K.L. (2000). Dll4, a novel Notch ligand expressed in arterial endothelium. Genes Dev 14, 1313-1318.

Silldorff, E.P., Yang, S., and Pallone, T.L. (1995). Prostaglandin E2 abrogates endothelin-induced vasoconstriction in renal outer medullary descending vasa recta of the rat. J Clin Invest 95, 2734-2740.

Son, N.H., Basu, D., Samovski, D., Pietka, T.A., Peche, V.S., Willecke, F., Fang, X., Yu, S.Q., Scerbo, D., Chang, H.R., et al. (2018). Endothelial cell CD36 optimizes tissue fatty acid uptake. J Clin Invest 128, 4329-4342.

Sorensen, C.M., Giese, I., Braunstein, T.H., Brasen, J.C., Salomonsson, M., and Holstein-Rathlou, N.H. (2012). Role of connexin40 in the autoregulatory response of the afferent arteriole. Am J Physiol Renal Physiol 303, F855-863.

Stan, R.V., Kubitza, M., and Palade, G.E. (1999). PV-1 is a component of the fenestral and stomatal diaphragms in fenestrated endothelia. Proc Natl Acad Sci U S A 96, 13203- 13207.

Su, Y., Qadri, S.M., Cayabyab, F.S., Wu, L., and Liu, L. (2014). Regulation of methylglyoxal-elicited leukocyte recruitment by endothelial SGK1/GSK3 signaling. Biochim Biophys Acta 1843, 2481-2491.

Surapisitchat, J., Jeon, K.I., Yan, C., and Beavo, J.A. (2007). Differential regulation of endothelial cell permeability by cGMP via phosphodiesterases 2 and 3. Circ Res 101, 811- 818.

Takabatake, Y., Sugiyama, T., Kohara, H., Matsusaka, T., Kurihara, H., Koni, P.A., Nagasawa, Y., Hamano, T., Matsui, I., Kawada, N., et al. (2009). The CXCL12 (SDF-1)/CXCR4 axis is essential for the development of renal vasculature. J Am Soc Nephrol 20, 1714-1723.

Takeya, K., Wang, X., Kathol, I., Loutzenhiser, K., Loutzenhiser, R., and Walsh, M.P. (2015). Endothelin-1, but not angiotensin II, induces afferent arteriolar myosin diphosphorylation as a potential contributor to prolonged vasoconstriction. Kidney Int 87, 370-381.

Tojais, N.F., Cao, A., Lai, Y.J., Wang, L., Chen, P.I., Alcazar, M.A.A., de Jesus Perez, V.A., Hopper, R.K., Rhodes, C.J., Bill, M.A., et al. (2017). Codependence of Bone Morphogenetic Protein Receptor 2 and Transforming Growth Factor-beta in Elastic Fiber Assembly and Its Perturbation in Pulmonary Arterial Hypertension. Arterioscler Thromb Vasc Biol 37, 1559-1569.

Ueda, H., Miyazaki, Y., Matsusaka, T., Utsunomiya, Y., Kawamura, T., Hosoya, T., and Ichikawa, I. (2008). Bmp in podocytes is essential for normal glomerular capillary formation. J Am Soc Nephrol 19, 685-694.

Unsold, C., Hyytiainen, M., Bruckner-Tuderman, L., and Keski-Oja, J. (2001). Latent TGF-beta binding protein LTBP-1 contains three potential extracellular matrix interacting domains. J Cell Sci 114, 187-197.

Valcourt, U., Alcaraz, L.B., Exposito, J.Y., Lethias, C., and Bartholin, L. (2015). Tenascin-X: beyond the architectural function. Cell Adh Migr 9, 154-165. 10

Van Themsche, C., Chaudhry, P., Leblanc, V., Parent, S., and Asselin, E. (2010). XIAP gene expression and function is regulated by autocrine and paracrine TGF-beta signaling. Mol Cancer 9, 216.

Wagenseil, J.E., and Mecham, R.P. (2012). Elastin in large artery stiffness and hypertension. J Cardiovasc Transl Res 5, 264-273.

Walker, A.E., Henson, G.D., Reihl, K.D., Morgan, R.G., Dobson, P.S., Nielson, E.I., Ling, J., Mecham, R.P., Li, D.Y., Lesniewski, L.A., et al. (2015). Greater impairments in cerebral artery compared with skeletal muscle feed artery endothelial function in a mouse model of increased large artery stiffness. J Physiol 593, 1931-1943.

Wang, H.U., Chen, Z.F., and Anderson, D.J. (1998). Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4. Cell 93, 741-753.

Wang, W., Ha, C.H., Jhun, B.S., Wong, C., Jain, M.K., and Jin, Z.G. (2010). Fluid shear stress stimulates phosphorylation-dependent nuclear export of HDAC5 and mediates expression of KLF2 and eNOS. Blood 115, 2971-2979.

Welti, J., Loges, S., Dimmeler, S., and Carmeliet, P. (2013). Recent molecular discoveries in angiogenesis and antiangiogenic therapies in cancer. J Clin Invest 123, 3190-3200.

Xu, Y., Zhu, J., Hu, X., Wang, C., Lu, D., Gong, C., Yang, J., and Zong, L. (2016). CLIC1 Inhibition Attenuates Vascular Inflammation, Oxidative Stress, and Endothelial Injury. PLoS One 11, e0166790.

Yamaguchi, I., Tchao, B.N., Burger, M.L., Yamada, M., Hyodo, T., Giampietro, C., and Eddy, A.A. (2012). Vascular endothelial cadherin modulates renal interstitial fibrosis. Nephron Exp Nephrol 120, e20-31.

Yang, H., Fang, L., Zhan, R., Hegarty, J.M., Ren, J., Hsiai, T.K., Gleeson, J.G., Miller, Y.I., Trejo, J., and Chi, N.C. (2015). Polo-like kinase 2 regulates angiogenic sprouting and blood vessel development. Dev Biol 404, 49-60.

Yao, D.W., Luo, J., He, Q.Y., Wu, M., Shi, H.B., Wang, H., Wang, M., Xu, H.F., and Loor, J.J. (2016). Thyroid hormone responsive (THRSP) promotes the synthesis of medium- chain fatty acids in goat mammary epithelial cells. J Dairy Sci 99, 3124-3133.

Yiu, S.S., Zhao, X., Inscho, E.W., and Imig, J.D. (2003). 12-Hydroxyeicosatetraenoic acid participates in angiotensin II afferent arteriolar vasoconstriction by activating L-type calcium channels. J Lipid Res 44, 2391-2399.

You, L.R., Lin, F.J., Lee, C.T., DeMayo, F.J., Tsai, M.J., and Tsai, S.Y. (2005). Suppression of Notch signalling by the COUP-TFII transcription factor regulates vein identity. Nature 435, 98-104.

Zhang, J., and Hill, C.E. (2005). Differential connexin expression in preglomerular and postglomerular vasculature: accentuation during diabetes. Kidney Int 68, 1171-1185.

Zhao, Q., Eichten, A., Parveen, A., Adler, C., Huang, Y., Wang, W., Ding, Y., Adler, A., Nevins, T., Ni, M., et al. (2018). Single-Cell Transcriptome Analyses Reveal Endothelial Cell Heterogeneity in Tumors and Changes following Antiangiogenic Treatment. Cancer Res 78, 2370-2382.