Genes That Confer the Identity of the Renin Cell

BASIC RESEARCH www.jasn.org

Genes that Confer the Identity of the Renin Cell

Eric W. Brunskill,*† Maria Luisa S. Sequeira-Lopez,‡ Ellen S. Pentz,‡ Eugene Lin,‡ Jing Yu,§ Bruce J. Aronow,* S. Steven Potter,*† and R. Ariel Gomez‡

Departments of *Pediatrics and †Cell and Developmental Biology, Cincinnati Children’s Hospital, Cincinnati, Ohio; Departments of ‡Pediatrics and §Cell Biology, University of Virginia School of Medicine, Charlottesville, Virginia

ABSTRACT Renin-expressing cells modulate BP, fluid-electrolyte homeostasis, and kidney development, but remark- ably little is known regarding the genetic regulatory network that governs the identity of these cells. Here we compared the gene expression profiles of renin cells with most cells in the kidney at various stages of development as well as after a physiologic challenge known to induce the transformation of arteriolar smooth muscle cells into renin-expressing cells. At all stages, renin cells expressed a distinct set of genes characteristic of the renin phenotype, which was vastly different from other cell types in the kidney. For example, cells programmed to exhibit the renin phenotype expressed Akr1b7, and maturing cells expressed angiogenic factors necessary for the development of the kidney vasculature and RGS (regulator of G-protein signaling) genes, suggesting a potential relationship between renin cells and pericytes. Contrary to the plasticity of arteriolar smooth muscle cells upstream from the glomerulus, which can transiently acquire the embryonic phenotype in the adult under physiologic stress, the adult juxtaglomerular cell always possessed characteristics of both smooth muscle and renin cells. Taken together, these results identify the gene expression profile of renin-expressing cells at various stages of maturity, and suggest that juxtaglomerular cells maintain properties of both smooth muscle and renin-expressing cells, likely to allow the rapid control of body fluids and BP through both contractile and endocrine functions.

J Am Soc Nephrol 22: 2213–2225, 2011. doi: 10.1681/ASN.2011040401

Renin-expressing cells are crucial in the control of renin phenotype.3 We hypothesized that maturing, BP and fluid-electrolyte homeostasis.1 In the adult adult, and recruited cells express a unique set of mammal, renin cells are located in the afferent ar- genes characteristic of the renin cell phenotype, teriole at the entrance to the glomerulus, thus their which is, in turn, vastly different from other cell name, juxtaglomerular (JG) cells (Figure 1). How- types in the kidney. Although understanding the ever, during embryonic life, renin cells are distrib- genetic regulatory network that governs the identity uted throughout the intrarenal arterial tree and in- of renin cells is of fundamental biologic and medi- side the glomeruli. With maturation, renin cells differentiate into smooth muscle cells and become restricted to a few cells in the classical JG localiza- Received April 20, 2011. Accepted August 19, 2011. tion in the adult (Figure 1).2 If an adult animal is Published online ahead of print. Publication date available at subjected to manipulations that threaten BP and www.jasn.org. S.S. Potter and R.A. Gomez contributed equally, as co-senior fluid-electrolyte homeostasis, there is an increase in authors, to this work. the number of renin cells along the preglomerular Correspondence: Dr. R. Ariel Gomez, Harrison Distinguished Pro- arteries and inside the glomerulus, resembling the fessor of Pediatrics and Biology, University of Virginia, 409 Lane embryonic pattern (Figure 1). This “recruitment” Road, MR4 Building, Room 2001, Charlottesville, VA 22908. Phone: of renin cells is achieved by the retransformation of 434-924-2525; Fax: 434-982-4328; E-mail: [email protected] arteriolar smooth muscle and mesangial cells to the Copyright © 2011 by the American Society of Nephrology

J Am Soc Nephrol 22: 2213–2225, 2011 ISSN : 1046-6673/2212-2213 2213 BASIC RESEARCH www.jasn.org

cal relevance, several problems prevented investigators from achieving this goal: (1) renin cells are very few in number (0.01 to 0.001% of the total kidney cell mass), (2) it had been almost impossible to isolate renin cells to purity, (3) in culture, renin cells stop making renin after 48 to 72 h, and (4) there had been no markers that could identify these cells independently of renin. To circumvent those problems and address the aforementioned hypothesis, we used a well-characterized Ren1c-YFP mouse that faithfully expresses YFP in renin cells throughout development and in response to physiologic challenges.4 We isolated YFP ϩcells from the kidney of newborn and adult mice, and from adult mice subjected to a physiologic challenge that elicits the retransformation of arteriolar smooth muscle cells (aSMCs) to the renin phenotype, and compared their gene profiles to those from multiple cell types of the nephron at various stages of development (Figure 1B).5 Finally, to define whether the set of genes expressed by the renin cell located at the pole of the glomerulus—the bonafide adult JG cell—is different from the set of genes expressed by other renin cells, we developed a single cell isolation and amplification procedure that allowed us to uncover the expression profile of the classical JG cell.

RESULTS

Data from 48 Affymetrix Mouse Gene 1.0 ST arrays, representing 16 different kidney samples in biologic triplicate, using Nugen RiboSpia target amplification technology, were analyzed with GeneSpring software. The samples included FACS purified renin expressing cells from newborns, adults, and adults treated with captopril. Genes with elevated expression in renin cells were sequentially screened for fold change versus total kidney cortex, Welch ANOVA (P Ͻ 0.05), yielding 1051 probesets. Further screening for fold enrichment, compared Figure 1. As the kidney matures, renin cells are restricted to the with a virtual kidney cortex made by combining the individual classical juxtaglomerular localization. (A) Top: Schematic representation compartment expression data, resulted in a list of 92 probesets of the distribution of renin cells (RC depicted in yellow) during early showing elevated expression in adult renin cells (see Concise development (left) and the progressive restriction in the location of RCs Methods for details and Supplementary Table 1 for complete to the JGA (right) during kidney ontogeny. If an adult animal is subjected to a condition that threatens BP and/or fluid-electrolyte homeo- gene lists of the 1051 and 92 gene sets). stasis, there is a recruitment of RCs along the afferent and interlobular The heat map of Figure 2A provides an overview of the gene arterioles, and within the glomeruli—by dedifferentiation of smooth expression pattern of P0, adult, and captopril-treated (re- muscle cells (SMC, red) and glomerular mesangial cells to renin-ex- cruited) adult cells. The bulk of differentially expressed genes pressing cells—in a pattern resembling the embryonic and fetal stages was associated with the newborn renin cells. Further analysis of renin distribution. Bottom: Classical depiction of the circulating renin showed that most of these genes were related to the highly angiotensin system and its acknowledged functions. EC, endothelial proliferative state of P0 cells, and included genes involved in cell depicted in blue; C, capillaries; Ang I, angiotensin I; Ang II, angio- cell division and DNA synthesis. A full list of genes differen- tensin II; ACE, angiotensin converting enzyme; Na, sodium. (B) Devel- tially expressed in P0 and adult renin cells is shown in Supple- opment of the nephron and its vasculature. Upon induction by the mentary Table 2. Interestingly, newborn cells express a signif- dividing ureteric tip, the mesenchymal cells around it condensate to icant number of factors (Reelin, Angiopoietin 2, tetraspanins, form the cap mesenchyme (CM), which, upon differentiation, evolves into the renal vesicle (RV) the s-shaped body (S), glomerulus (G), and the Lpar4, integrins, Notch receptors [Figure 4, I and L] and li- proximal tubule (PT). The JG cells and arterioles originate from a sep- gands) known to be involved in angiogenesis. The heatmap of arate group of mesenchymal precursors that differentiate in situ to form Figure 2B compares renin cells with other cells in the kidney. the afferent and efferent arterioles, (aa) and (ea), respectively. The dis- As expected, P0, adult, and captopril-treated adult cells are the tribution of renin cells in the immature kidney shown on the left is most closely related. The captopril treatment of adults resulted extensive along the aa, interlobular arterioles, and arcuate arteries. H, in a more P0-like gene expression signature, reflecting the in- loop of Henle; MD, macula densa; DT, distal tubule. creased number of cells expressing renin along the kidney vas-

2214 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 2213–2225, 2011 www.jasn.org BASIC RESEARCH

Figure 2. The transcriptome of renin cells is vastly different from any other renal cell type. (A) Heatmap of 1051 probesets, showing differential expression in adult total kidney cortex (Ctx) and renin expressing cells from newborn (P0), adult, and captopril-treated adults (Cap). Each horizontal line represents one probeset expression, with red showing high-level expression. (B) Heatmap of renin cell-enriched genes, comparing across multiple cell types. Developmental times are given (E10.5, E13.5, Adult, P0, P1, P2, P3, P4, E15.5), as are tissue compartment (Ub, ureteric bud; Tr, trunk; Pod, podocyte; CM, cap mesenchyme; RV, renal vesicle; Endo, endothelial cells; M Endo, medullary endothelial cells; G Endo, glomerular endothelial cells; C Endo, cortical endothelial cells; Glom, glomerulus; Mes, mesangial cells; Ren, renin-expressing cells; Cap Ren, renin cells from captopril-treated adult; Ctx, total renal cortex). (C) Heatmap of 369 most specific adult renin cell genes, derived from both single cell analysis (SCAMP) and RiboSpia analysis. Ctx, total kidney cortex; Ad, adult renin cells; P0, newborn renin cells; Cap, adult captopril-treated renin cells. See also Supplementary Figure 1 and Supplementary Tables 1, 2, and 3.

culature as it occurs during development. Renin cells also show plemental Methods. To provide a uniform amplification significant gene expression similarities to mesangial cells, en- chemistry baseline for comparison, we also used SCAMP to dothelial cells, and, to a lesser extent, the renal capsule. For amplify aliquots of RNA from total adult kidney cortex. The complete gene lists with associated heat maps see Supplemen- resulting array data were screened, requiring raw signal of at tary Tables 1 through 3. least 125 (giving 10,260 probesets) and FC greater than 2.5, The transcriptome of the bonafide JG cell is likely to differ Welch t test, P Ͻ 0.05, and concordance between the RiboSpia from other renin cells. Although FACS isolation provides ex- and SCAMP microarray datasets, arriving at an adult JG cell cellent purity, the renin cell is quite rare, making purification gene expression signature of 369 genes. The heatmap (Figure challenging, and there are reports that cells outside of the JGA 2C) shows that the two target amplification strategies, SCAMP can produce renin, even in the normal adult.6,7 To insure pu- and RiboSpia, gave reproducible, overlapping, yet distinctive, rity of the JG cell, we developed a single cell amplification pro- gene lists, with some probe sets clearly amplified better by one cedure (SCAMP) that allowed us to obtain the gene profile of procedure than the other. The complete gene list for the 369- five individual YFP positive cells isolated from the JG poles of gene set is provided in Supplementary Table 1. Complete lists sieve-purified glomeruli from Ren1c-YFP mice (see Concise of biologic processes, molecular functions, regulatory tran- Methods for a brief description of the JG cell isolation, RNA scription factors, candidate target genes, and regulatory mi- amplification, and microarray analysis, and Supplemental croRNAs are included in Supplementary Table 3. Methods for details.). The excellent reproducibility (Pearson correlation coefficients: 0.86 to 0.94) and high sensitivity of Genes with Renin-Cell-Enriched Transcripts SCAMP data in comparison with two commercial systems, As expected, renin mRNA showed the highest enrichment in Nugen RiboSpia OneDirect and Miltenyi uMACS SuperAmp, adult renin cells, 41-fold (FACS-RiboSpia data, raw signal ap- are shown in Figure 2C, Supplementary Figure 1, and the Sup- proximately 11,000) when compared with total kidney cortex,

J Am Soc Nephrol 22: 2213–2225, 2011 The Genetic Network of Renin Cells 2215 BASIC RESEARCH www.jasn.org

providing an important positive control. The second highest was aldose-keto-reductase (Akr1b7), 23-fold enrichment. AKR1B7 localizes to JG cells, renal arterioles, and the glomerular mesangium in the newborn kidney, it becomes restricted to the JG cells in the adult, and it is re-expressed in afferent aSMCs that underwent retransformation to renin- expressing cells in the captopril-treated adult (Figure 3, A, C, and E). Thus, AKR1B7 fol- lows the pattern of renin expression (Fig- ure 3, B, D, and F) throughout development and in response to a homeostatic challenge. Interestingly, in animals with homozygous deletion of the renin gene,8 AKR1B7 was maintained in those cells that would have normally expressed renin (Figure 3G), indicating that AKR1B7 does not depend on renin expression for its expression and confirm- ing that AKR1B7 serves as an invaluable new independent marker of renin cells. Among the adult renin cell-specific expressed genes, it is interesting to note the high expression levels of a select sub- group, including, in addition to renin and Akr1b7, the genes Rgs5, Crip1, ATP1b2, Syne2, Plac9, Myh11, Jph2, Myo18, and Mgp. RGS5 (regulator of G protein signaling 5) is a potent GTPase-activating protein for Gi␣ and Gq␣, which is expressed in vascular smooth muscle and has been considered as a marker for pericytes, which express RGS5 at high levels.9 Rgs5 is expressed sevenfold higher in renin cells (FACS data) compared with total kidney cortex, and it is known to be expressed by the smooth muscle cells of the developing kidney arterioles and by mesangial cells in the adult animal.9 The RGS5 expression pattern overlaps, in fact, with the expression patterns of Figure 3. Akr1b7 expression is an independent marker of renin cells. Immunostaining PDGF-RB and renin,10,11 suggesting a for Akr1b7 (A, C, E, G) and renin (B, D, F, H), in consecutive sections, shows potential relationship between pericytes colocalization of Akr1b7 in renin cells in (A, B) newborn kidney. Short arrow, JG cells; and the renin cell. long arrows, arterioles and larger vessel. (C, D) Adult kidney. Arrows, JG cells. Akr1b7 Cysteine-rich intestinal protein marks renin cells at all stages of development. (E, F) Adult kidney from an animal ϩ (CRIP1) is also highly expressed in JG administered low sodium diet captopril stimulates re-expression of renin and Akr1b7. Arrows, JG cells and afferent arteriole. Akr1b7 is expressed in the same cells. The localization of CRIP1 to renin pattern as renin. (G, H) Ren1c knockout kidney:Akr1b7 marks renin cells even although cells is shown in Figure 4M. Crip1 be- the cells are unable to make renin. Arrows in (H) indicate JG cells and afferent longs to the LIM domain/double zinc arterioles that are expressing Akr1b7 in (G). finger protein family, which, in the gut, controls zinc transport. CRIP 1 may play a different role in JG correlated with the differentiation of cells to the myofibroblast cells, as a novel stress response factor whereby CRIP1 may cell lineage.12 CRIP1 protein has also been implicated in mus- provide a survival advantage. High levels of CRIP1 have been cle differentiation,13 suggesting that CRIP1 may be involved in

2216 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 2213–2225, 2011 www.jasn.org BASIC RESEARCH

Figure 4. Genes identified in the JG cell signature are expressed in JG cells and vessels. (A–D) In situ hybridization in newborn kidneys. (A) Mef2c expression in developing vessel (arrow). (B) Hey1 is expressed in the vessels (arrows) and in glomeruli of the nephrogenic zone. (C) Nr4a1 expression in an afferent arteriole (arrow). (D) Nkx3–1 is expressed in JG cells and afferent arteriole (arrow, G, glomerulus). (E) Immunostaining shows that Nkx3–1 localizes to JG cells and the afferent arteriole in kidney of a captopril-treated adult (arrow). (F–O) Immunostaining in adult kidneys. (F) Nfat is expressed in JG cells (arrow), vessels (short arrows), and some tubules. (G) Creb is expressed in renin-expressing JG cells (outline). (H) Consecutive section of (G) showing renin expression (outline). (I) Notch3 (brown) is expressed in JG cells. (L) Double staining for Notch3 (brown) and renin (purple) in a consecutive section of (I) shows coincidence of expression. Arrows in (I) and (L), JG cells. (J) Costaining for ␣SMA (purple) and renin (brown) shows expression of ␣SMA in JG cells as well as afferent arterioles. (K) Renin expression in JG cells for comparison with (J) Arrows in (J) and (K), JG cells. (M, N) Crip1 localizes to renin expressing JG cells as well as some tubules. Arrows in (M) and (N), JG cells. (O) S1P receptor is expressed in JG cells ϩ (arrow). (P) RT-PCR of RNA extracted from FACS isolated YFP cells confirms the expression of Akr1b7 and RBP-J mRNAs. (Q) Chromatin immunoprecipitation shows enrichment of phopspho-Creb at the cAMP responsive element in the renin enhancer in renin-expressing kidney cortex cells but not in skeletal muscle cells, which do not express renin. (R) RBP-J is enriched at the RBP-J element in cells from primary cultures of kidney arterial smooth muscle cells.

J Am Soc Nephrol 22: 2213–2225, 2011 The Genetic Network of Renin Cells 2217 BASIC RESEARCH www.jasn.org

Table 1. Expression of smooth muscle genes in renin cells how the integrated activity of these channels control the pre- during development cise amounts of renin released to the circulation. Newborn Renin Adult Renin Cells Gene Cells Transcriptional Control of the Smooth Muscle CM RV CM RV Phenotype Renin 81.9 76.0 72.5 67.3 Analysis of the promoter regions of the genes coordinately ex- ␣-SMA 4.9 4.7 13.0 12.2 pressed in the adult renin cell for the presence of evolutionarily SM-MHC 8.2 8.7 20.3 21.5 conserved transcription factor binding sites (TFBS) allowed SM22␣ 3.1 2.2 13.7 9.7 the identification of candidate regulators. Three factors—SRF, Calponin 1 2.1 2.6 15.9 19.6 E2A (Tcf3), and IK3—gave the strongest statistics. A brief de- RGS5 34 25.8 33 25.1 scription of these TFs is provided in Table 3. Figure 5C shows RGS2 4.2 4.1 11.3 11.1 the predicted SRF targets in the renin cell. SRF binds to the CM, cap mesenchyme; RGS2, regulator of G-protein signaling 2; RGS5, serum response element of target genes and, together with regulator of G protein signaling 5; RV, renal vesicle; ␣SMA, alpha smooth muscle actin; SM-MHC, smooth muscle myosin heavy chain; SM22␣, GATA 4 and NKX2.5, directs early cardiac development. Sev- transgelin. The expression of the indicated genes is presented as fold eral members of the NKX family (NKX2.3 to 2.5 and NKX3.1, enrichment over CM or RV. Figure 5C and Figure 4, D and E) are expressed in JG cells, and, together with SRF, may contribute to the development and maintenance of their smooth muscle character. SRF is also in- the maintenance of the smooth muscle phenotype of the JG volved in the development of vascular SMCs. Together with cell. Myocardin (Myocd, 1051 gene list in Table S1), SRF directs transcription of SMC genes. Mice lacking Myocd have serious The Smooth Muscle Signature of Newborn and Adult vascular defects and die before E10.5.15 Renin Cells E2A transcription factors are critical for lineage commit- Newborn and adult renin cells express numerous genes char- ment, differentiation, and survival of lymphocytes, muscle, acteristic of the smooth muscle phenotype (Table 1). In addi- ϩϩ and neural cells.16,17 Ikaros (IK3) encodes a Zn finger DNA- tion to alpha smooth muscle actin (␣-SMA, Figure 4J), Rgs5, binding protein that regulates lymphocyte commitment and Rgs2, smooth muscle myosin heavy chain (SM-MHC, ␣ differentiation. Myh11), calponin (Cnn1), and transgelin (SM22 ) are also Another transcription factor, NFATc4 (1051 gene list in highly represented in renin cells. These findings corroborate Supplementary Table 1 and Figure 4F) is of particular interest previous work demonstrating a lineage relationship between because of its connection to calcium (see Supplementary Table 3,14 renin-expressing cells and SMCs of the renal arterioles. An 4 for the list of all transcription factors). Increase in intracellu- interesting feature is the expression of other muscle genes nor- lar Caϩϩ leads to activation of calcineurin, which dephospho- mally thought to be exclusively expressed in skeletal and/or rylates NFAT, which, in turn, is translocated to the nucleus, cardiac muscle. These include dystrophin (DMD), phospho- where it binds the promoter of target genes. Null mutations of lamban, and NKX2.3 (transcription factor involved in myo- NFATc4 and NFATc3 in mice result in defects in vascular pat- cardial development), MYO18B, and PDLIM 3 (an actinin- terning and angiogenesis, and NFAT is required for pericytes associated LIM protein whose deficiency causes dilated to coat the vessel wall.18 Given its predicted target genes, NFAT cardiomyopathy). Although the function of these genes in the may play a central role in integrating calcium signals with tran- JG cell is unexplored, it is possible that their roles in morpho- scriptional control of the smooth muscle phenotype in JG cells. genesis and contractility of skeletal and cardiac muscle are conserved in the JG cell. Transcriptional Control of the Renin Phenotype To better understand the regulation of the renin phenotype, we Expression of Genes Related to Calcium Control of analyzed the combination of expressed TFs in the adult renin- Renin Release expressing cells and the evolutionarily conserved TFBS in the Renin-expressing cells depend on calcium signaling for renin renin promoter and enhancer. We first used GeneSpring to synthesis and release. Calcium regulation of JG cells is different combine the single-cell and FACS-based datasets, requiring from other endocrine cells, with the exception of parathyroid minimum raw expression of 400, and elevated expression in cells that also respond to increase in intracellular calcium with renin cells compared with cortex, yielding 118 renin-cell-en- a decrease in hormone secretion. This inverse relationship be- riched transcription factors (Supplementary Table 4). tween calcium and renin release has been termed the calcium We also specifically examined regulation of the renin pro- “paradox.” Some of the genes expressed by renin cells involved moter. Genomatix was used to identify the TFBS with con- in calcium homeostasis are shown in Figure 5B and Table 2. served sequence and spacing between human and mouse prox- Our renin cells have a profusion of calcium, potassium, and imal promoters, with repeat sequences removed (Figure 6A). sodium channels, and other transporters that connect extracel- There was a striking concordance between the two datasets, lular signals with the cell’s milieu. It remains to be determined with most of the conserved TFBS having corresponding ex-

2218 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 2213–2225, 2011 www.jasn.org BASIC RESEARCH

Figure 5. Gene network characteristic of the adult renin cell. (A) Renin cell genes known to be involved in muscle contraction. Some of the 369 adult renin cell signature set of genes are shown in the center hexagons, with surrounding rectangles representing biologic processes (dark green), molecular functions (light green), and transcription factors (violet), with associated genes connected by lines. Genes involved in muscle contraction are highlighted in yellow. (B) Genes involved in calcium homeostasis are highlighted in yellow. (C) Transcription factors (SRF, E2A, IK3) and the smooth muscle phenotype. SRF candidate target genes are highlighted in yellow. See also Supplementary Table 3.

Table 2. Expression of genes related to Ca2 ϩ control of renin synthesis and release Gene Name Function KCNA3 (Kcna3) Potassium voltage-gated channel, shaker- Mediates the voltage-dependent potassium ion permeability of excitable related subfamily, member 3 membranes and also comprises the Ca2ϩ-activated Slo (actually 7-TM) and the Ca2ϩ-activated SK subfamilies JPH2 (Jph2) Junctophilin 2 Approximates the L calcium channels and the ryanodine receptor PLN (Pln) Phospholamban Inhibits sarcoplasmic reticulum Ca2ϩ-ATPase ϩϩ TRP6 (Trpc6) Transient receptor potential cation channel, Activated by G-protein coupled receptors; important for the Ca subfamily C, member 6 response to adrenergic stimulation and may integrate systemic responses conveyed by the renal nerves, which, in turn, are modulated ϩϩ by the regulation of intracellular Ca TRDN (Trdn) Triadin Releases calcium from the sarcoplasmic reticulum triggering muscle ϩϩ ϩϩ contraction through Ca -induced Ca release FXYD1 (Plm) Phospholemman FXYD domain-containing Plasma membrane protein that possesses channel activity; may also ion transport regulator 1 regulate Na,K-ATPase activity and modulate ion transport in extraglomerular mesangial cells and JG cells in response to changes in macula densa NaCl concentrations (tubuloglomerular feedback) HRC (Hrc) Histidine rich calcium binding protein Luminal sarcoplasmic reticulum protein that may be a target of the MEF2c transcription factor linking the smooth muscle specification ϩϩ with the regulation of intracellular Ca MGP (Mga) Matrix gla protein Extracellular matrix protein that may act as a buffer preventing wide ϩϩ variations and/or excessive free Ca accumulation in the JG cell microenvironment ϩϩ PTP4A3 (Ptp4a3) Protein tyrosine phosphatase type IV, A3 Inhibits AngII induced Ca mobilization and may play a role in the regulation of the negative feedback for renin release caused by AngII SFRP2 (Sfrp2) Secreted frizzled-related protein 2 Soluble modulator of Wnt signaling, binds to Wnt-4, and may regulate Wnt-4 signaling during kidney development; also a stimulator of angiogenesis via calcineurin-NFAT pathway and may be important for the maintenance of the JG cell’s bivalent -contractile and endocrine–phenotype pressed TFs in the renin cell. The corresponding TFs are shown tors, likely candidates for the regulation of the renin pheno- in Table 4. type. Validating the relevance of RBP-J expression, ChIP PPAR and VDR are well-known regulators of renin expres- experiments showed enrichment of binding of RBP-J to the sion.19,20 RBPF (RBP-J) is the final effector for all Notch recep- RBP-J site in the renin promoter (Figure 4R), and conditional

J Am Soc Nephrol 22: 2213–2225, 2011 The Genetic Network of Renin Cells 2219 BASIC RESEARCH www.jasn.org

Table 3. Transcriptional control of the smooth muscle phenotype Gene Name Function E2A (Tcf3) Transcription factor 3 Basic helix-loop-helix transcription factor critical for lineage commitment, differentiation, and survival of lymphocytes, pancreatic, muscle, and neural cells IK3 (Ikzf3) IKAROS family zinc finger 3 Zinc finger DNA-binding protein that regulates lymphocyte commitment and differentiation MYOCD (Myocd) Myocardin Expressed in renin cells and, together with SRF, directs transcription of SMC genes NFAT (Nfatc4) Nuclear factor of activated Involved in vascular patterning and angiogenesis and required for T-cells, cytoplasmic 4 pericytes to coat the vessel wall NKX (Nkx2.3, Nkx2.4, NKX-homeodomain factor Critical regulators of organ development; in JG cells; together with SRF, Nkx2.5, Nkx3.1) they may contribute to the development and maintenance of their smooth muscle character SRF (Srf) Serum response factor Member of the MADS box (MCM1, agamous, Defficiens, and SRF) superfamily of transcription factors; together with GATA 4 and NKX2.5, it directs early cardiac development and is involved in the development of vascular SMCs

of the renal vasculature, it is possible that this function may be in part accom- plished by the binding of estrogen receptors to the EREF of the renin gene. As described in Results, JG cells maintain both an endocrine and a smooth muscle phenotype. Additional studies were carried out to determine the expression of a number of the JG- cell-enriched transcripts that could be important for maintenance of their smooth muscle cell characteristics and for renin expression. Among several Figure 6. Localization of binding sites conserved in mice and humans for transcription transcription factors enriched in JG factors enriched in JG cells. (A) The renin promoter and (B) the renin enhancer. See also cells, Mef2c, Nkx3–1, Nr4a1 (Nur77), Supplementary Table 4. and Nfat are potentially involved in the maintenance of their smooth muscle deletion of RBP-J in JG cells results in a significant decrease in character and are expressed appropriately in JG cells and renin gene expression and the number of renin cells.21 Simi- vessels (Figure 4, A and D–F). The RGS proteins, which are larly, V$CREB was clearly expressed in JG cells, and ChIP ex- highly expressed in JG cells and may be involved in vessel periments showed its enrichment at the CRE of renin-express- maturation, are known to inhibit sphingosine receptor sig- ing cells (Figure 4, G and Q). We have previously shown that naling. S1P receptors are enriched in the JG cells arrays and the cAMP/CBP/p300/CREB pathway is fundamental in the localize to JG cells (Figure 4O). We found that RBP-J, the control of renin synthesis and release.4,22 final effector of Notch signaling, is important in renin ex- A similar analysis was also performed for the renin en- pression. Several other Notch pathway genes are also enhancer (Figure 6B) that begins at Ϫ2500 for the mouse and riched in JG cells, and, among them, Notch 3 and Hey1 Ϫ11,015 for the human, and, in each case, extends about 275 (Figure 4, B, I and L) were found to localize to vessels and JG bases further 5Ј. The presence of an estrogen response element cells. Thus, the aforementioned regulators are appropriately (V$EREF) is intriguing. Estrogens are known to increase ang- localized and may participate in development and/or main- iotensinogen and decrease plasma renin,23 explaining, in part, tenance of the renin cell phenotype. gender differences in the activity of the renin-angiotensin system. The molecular events that regulate the activity of the renin gene upon exposure to estrogens remain to be studied in detail. DISCUSSION Because estrogen receptors regulate transcription of genes involved in branching morphogenesis of the mammary gland, Novel features of the present work include the extensive prostate, and lung,24 and renin cells are involved in branching comparison of renin cells with numerous cells types from

2220 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 2213–2225, 2011 www.jasn.org BASIC RESEARCH

Table 4. Transcription factors in the renin proximal promoter and enhancer Gene Name Function O$VTBP (Tbp) TATA binding protein Binds to the TATA box ϳ35bp upstream of the transcription start in many genes and forms part of the RNA polymerase II preinitiation complex V$ABDB (Hoxd10) Homeobox D10 Member of the Abd-B homeobox family; expressed in the developing limb buds and involved in differentiation and limb development; mutations in this gene have been associated with Wilm’s tumor V$AP2F (Tcfap2) Activator Protein-2 Mediates transcriptional activation of the CYP11A1 gene by interacting with Sp1 V$ARID (Arid5b) AT rich interactive domain 5B Transcriptional control of lymphocyte differentiation V$CDXF (Cdx1) Vertebral caudal-related Regulates intestine-specific gene expression and enterocyte homeodomain 1 differentiation. Cdx1 is expressed in the proliferating immature epithelium during intestinal development and becomes restricted to the proliferative crypt in the adult intestine V$CREB (Creb) cAMP responsive element binding Binds to the cAMP responsive element in gene promoters and activates protein transcription V$EREF (Esrrg) Estrogen-related receptor gamma Involved in branching morphogenesis of the mammary gland, prostate, and lung V$GATA3 (Gata3) GATA binding protein 3 Involved in T cell development; mutation of this gene causes hypoparathyroidism, sensineural deafness, and kidney dysplasia V$GFI1 (Gfi1b) Growth factor independent 1b Zinc finger protein required for erythroid and megakaryocytic lineages; can be a repressor or activator depending on promoter and cell type context; alters histone methylation by recruiting histone methyltransferase to target gene promoters V$GLIF (Glis2) GLIS family zinc finger 2 Gli-related, Krüppel-like transcription factor is an activator or repressor of gene transcription, depending on the gene and promoter context and plays a role in kidney development and neurogenesis V$HAML (Cbfa2t3) Core-binding factor, runt domain, Putative breast tumor suppressor; transcriptional repressor protein alpha subunit 2, translocated to containing a zinc-finger motif common to developmental proteins, 3 (human) suggesting a potential function in regulating differentiation and morphogenesis V$HAND (Hand1) Heart and neural crest derivatives- Belongs to the bHLH family and plays an essential role in cardiac expressed protein 1 morphogenesis V$KLFS (Klf2) Kruppel-like transcription Factor 2 Modulates blood vessel maturation via smooth muscle cell migration and coating of endothelial tubes V$HOXC; V$HOXF Homeobox B8 and Homeobox D8 Hoxb8: involved in the development of the sensory neuron network; (Hoxb8; Hoxd8) knockout affects survival of spinal ganglion, causes aberrant limb reflexes and the dorsal spinal neurons are abnormally distributed Hoxd8: involved with limb and genital development; may also play a role in adult urogenital tract function and in the maturation and maintenance of lymphatic vessels V$MYOD (Myod1) Myogenic differentiation 1 A member of the bHLH group myogenic factors subfamily of transcription factors, which regulates muscle differentiation V$NR2F Nuclear receptor subfamily 2 Involved in Leydig cell development V$RBPF (RBP-J, CBF1) Recombination signal binding The final effector of the Notch-signaling pathway protein for immunoglobulin kappa J region V$RXRF (Vdr) Retinoid X receptor Ligand-activated transcription factors of the steroid hormone family of nuclear receptors, which, together with vitamin D, regulate renin gene expression V$XBBF (Rfx2) Regulatory factor X, 2 (influences Member of the regulatory factor X gene family proteins, which contain a HLA class II expression) winged helix DNA binding domain; transcriptional activator involved in regulation of gene expression during meiosis and the early development of spermatids the renal cortex at different developmental points. Further- individual adult JG cells. Specifically, we show that renin more, we developed a single cell isolation and amplification cells express a unique set of genes vastly different from other procedure that allowed us to identify the transcriptome of cell types in the kidney: They possess markers that topolog-

J Am Soc Nephrol 22: 2213–2225, 2011 The Genetic Network of Renin Cells 2221 BASIC RESEARCH www.jasn.org

ically and functionally link them to arterial and interstitial thelial cells. This cannot be attributed to endothelial cell con- pericytes, and express Akr1b7, a new and valuable marker tamination of the renin cells since the level of the specific en- for renin cells, independent from renin expression. Con- dothelial cell marker Tie-2 is low. Thus, endothelial cells and trary to arteriolar cells distant from the glomerulus, which JG cells may be related and share the expression of certain transiently express renin during development and/or a ho- genes. meostatic threat, adult JG cells maintain a dual smooth We have previously shown that the three major cells of the muscle and renin phenotype, driven by a unique transcrip- renal arteriole are present in the kidney undifferentiated mes- tional network that maintains, at all cost, the cell’s dual enchyme before arterioles and nephrons are formed.14,30 Those endocrine and contractile functions necessary for the main- cells have the capability to assemble the renal arterioles.14 We tenance of homeostasis. have shown that the differentiation, branching, and elongation The expression of Rgs5 and Rgs2 in renin cells suggests a of the renal arterial tree—which is maximal during the first potential lineage and/or functional relationship with pericytes. week of postnatal development—is intimately linked to the Pericytes, like renin cells, are mural cells that cover endothelial differentiation of the renin cells: Development of each new tubes and provide support to the vasculature. In the renal vas- arteriole encompasses the coating of the vessel with renin-ex- culature, renin cells are often encountered forming rings pressing cells.31 As the vessels mature, renin cells differentiate around the renal arterioles,25,26 not too dissimilar in appear- into SMCs.3,14 This pattern is repeated with each new branch ance to mural cells/pericytes in other vascular territories. In until the arteriolar tree is completed and the only remaining addition to their location around the renal arterioles, during renin containing cells are those at the tip of the arterioles. Al- embryonic life or homeostatic stress, renin-expressing cells are though a direct effect of locally produced angiotensin and/or occasionally found in the location of the renal interstitial peri- renin is likely responsible for arteriolar differentiation, the cytes. Although Rgs5 is involved in normal embryonic and tu- findings of our arrays suggest that renin cells produce addi- moral angiogenesis,27 and it participates in the recruitment of tional angiogeneic factors that may provide guidance cues mural cells during maturation of renal blood vessels, the role of (reelin), positional information, and cell-to-cell communi- RGS5 in kidney vascular morphogenesis seems less clear: cation (components of the Notch pathway), matrix diges- Rgs5 –/– mice are viable and fertile, and their vasculature seems tion (Timp3), endothelial elongation (angiopoietins), and to develop normally. However, Rgs5 may play a role at the end recruitment of smooth muscle cells (Enpp 1 to 3, Tspan), of arteriole maturation, when branching of the renal arterioles which, in turn, govern the proper assembly of the renal has been completed and renin cells are confined to the classical arterioles. The fact that these genes are expressed at higher JG localization. The factors that control the restricted localiza- levels in the newborn, when the renal tree is still developing, tion of renin cells near the glomerulus are unknown, but RGS suggests that they may play a role in renal arterial develop- inhibits sphingosine-1-phosphate/S1P1 receptor(s) signaling. ment during early life. Inhibition of S1P1, which is expressed in JG cells, may prevent The results of our microarray data indicate that, contrary to further migration of mature JG cells once they have reached aSMCs upstream from the glomerulus, JG cells maintain a dual their destination. endocrine and smooth muscle phenotype. This agrees with our As mentioned above, AKR1B7 is coexpressed in renin- immunocytochemical and qRT-PCR data showing the expres- expressing cells throughout development and in response to sion of renin and numerous markers and regulators of smooth physiologic challenges. Akr1b7 is re-expressed, together muscle in JG cells. with renin, along the renal arterioles, suggesting that a com- The reason for the bivalent nature of the classical JG cell mon mechanism may regulate the expression of both en- is unclear, but it is likely that the location of the cells within zymes. Interestingly, in animals with deletion of the renin the JG apparatus continually exposes them to signals from gene, Akr1b7 expression is maintained, indicating that other cells that maintain them, ready to secrete renin rapidly Akr1b7 can be used as an independent marker for cells pro- in acute situations, as is known to occur in acute hypoten- grammed to be renin-expressing cells. Akr1b7 plays an im- sion, and simultaneously contract to regulate glomerular portant role in the clearance of xenobiotics as well as in the hemodynamics. Several in vitro and in vivo studies suggest transformation of harmful aldehydes generated during hor- that JG cells have the capability to contract.32,33,34 Whereas mone synthesis. Further, Akr1b7 participates in steroid syn- many of the signals have not yet been identified, it seems thesis in the adrenal gland, where renin is also expressed in reasonable that cell-to-cell contact with other JG cells, as fetal life.28 Most recently, Akr1b7 has been shown to be cru- well as endothelial, smooth muscle, and macula densa cells, cial in the synthesis of prostaglandins,29 which, in turn, are is crucial for the maintenance of the myoepithelioid JG cell known to regulate renin synthesis and release,1 suggesting phenotype. Recently, it was found that connexin 40 (Gja5), that AKR1B7 may regulate renin release by an intracrine which is highly expressed in renin cells on our arrays, is mechanism. Further studies will be necessary to determine fundamental for the maintenance of renin expression near the role of Akr1b7 in renin cells. the glomerulus.35 Similarly, JG cells express members of the As mentioned in the Results, the gene-expression profile of Notch pathway (Notch 3, Jagged 1, Hey 1), known to transmit renin cells shares some significant similarity to that of endo- cell-to-cell signals and regulate gene expression via the transcrip-

2222 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 2213–2225, 2011 www.jasn.org BASIC RESEARCH tional regulator RBP-J. As mentioned above, the renin gene pos- CONCISE METHODS sesses a highly conserved RBP-J (V$RBPF) binding site between its proximal promoter and the renin enhancer. Our studies show Isolation of Renin Cells, RNA Purification, and Target that RBP-J binds the renin promoter in vivo. In addition, and Amplification providing functional validation to the TFs identified, we recently Renin-expressing cells were isolated by FACS from kidneys of our found that conditional deletion of RBP-J in JG cells results in a Ren1c-YFP mice,4 in which renin expression is marked by YFP using a marked decrease in the number of renin-expressing JG cells, re- high-speed digital BD FACS Aria II Cell Sorter. As a control, cells were flected by decreased circulating renin and low arterial BP.36 isolated in parallel from wild-type kidneys to provide the baseline for Whereas some of the signals conveyed by other cell types excluding unlabeled cells. Briefly, kidneys were excised from the are beginning to be unraveled, the intracellular events that mouse, decapsulated, and demedullated. The remaining cortical tis- maintain the JG cell’s unique myoepithelioid phenotype sue was minced in 0.3% collagenase A (Roche Applied Science, Indi- were unknown until now. The results of our analyses indi- anapolis, IN) in phosphate buffered saline (PBS, 14190-144; Invitro- cate that several potential transcriptional regulators (NFAT, gen, Carlsbad, CA) and incubated for 30 min at 37 °C, with frequent SFRP2, SRF, and MEF2c) may operate in conjunction with pipeting. Cells were centrifuged and the pellet resuspended in 0.05% calcium signals to maintain the smooth muscle phenotype trypsin-EDTA (25300-054; Invitrogen) and incubated at 37 °C for 5 to and thus keep the cell’s ability to contract while still possess- 10 min, with frequent pipeting, to obtain a single cell-enriched susp- ing the ability to synthesize and secrete renin. sension. Trypinsization was stopped with 10% FBS in PBS. The cells A remaining question is why JG cells retain this dual were filtered through a 100-␮m cell strainer (08 to 771-19; Thermo phenotype and the possible advantages accruing from it. Fisher Scientific Inc., Pittsburgh, PA), the red blood cells were lysed Figure 7 depicts a potential hypothetical scenario whereby (RBC lysis buffer, R7757; Sigma, St Louis, MO), and the cells filtered the cells could be either completely endocrine (left side), as through a 70-␮m cell strainer (08 to 771-2; Fisher). This cell suspen- in prolonged renal ischemia, or be completely devoid of sion was subjected to FACS to isolate YFPϩ cells. The yield of YFPϩ renin granules, as in severe hypertension and salt loading. cells was routinely 0.01 to 0.02% of the cells applied to the sorter. RNA Whereas these two opposite phenotypes may temporarily was purified and amplified as described.5 Complete protocols are accomplish the function of either producing enough renin available at GUDMAP.org (http://www.gudmap.org/Research/ in an attempt to reestablish BP/fluid electrolyte balance, or Protocols/Potter.html). Gene ontology molecular function and bio- maintain renal blood flow constant, neither of them can be logic process analysis of the JG cell gene list was conducted with Top- sustained for too long without leading to deterioration of pGene (http://toppgene.cchmc.org/). renal function and the development of systemic vascular and glomerular pathology. Therefore, to be in the homeo- Microarray Data Analysis static range, the JG cell may need to maintain a dual endo- Microarray data were screened for mappable non-Y chromosome crine-contractile phenotype, which is a myoepithelioid state probesets, with raw expression levels of 125 or greater, yielding 12,322 (Figure 7). We anticipate that the set of 369 core genes probesets. These were then screened for fold change (FC) Ͼ2 com- found in the JG cells, coregulated by the transcription fac- pared with total adult renal cortex, giving 248, 1082, and 415 probe- tors described herein, act in concert to confer the identity of sets, respectively, for renin cells from adult, P0, and captopril-treated this most intriguing cell type. adults. The pooled list, because of overlap, gave 1228 probesets, which were filtered with Welch ANOVA (P Ͻ 0.05), giving 1051 probesets, of which 226 were expressed in adult renin cells. This set of probesets was further screened, requiring at least twofold enrichment compared with a virtual renal cortex made from the individual compartment expression data, yielding 92 adult renin cell-enriched probesets.

Single Cell Amplification Procedure (SCAMP) We optimized a PCR target amplification protocol37 previously used by Cepko and colleagues.38 We added a random primer to the initial reverse transcription (RT) step to give improved full-length representation of transcripts and modified the reverse transcription Figure 7. The bivalent endocrine-contractile phenotype of JG cells and homeostatic reaction by adding T4 gene 32 protein to in- control. See text for details. crease the yield of RT-PCR products.39,40 An

J Am Soc Nephrol 22: 2213–2225, 2011 The Genetic Network of Renin Cells 2223 BASIC RESEARCH www.jasn.org exonuclease treatment step after the RT reaction was added to remove (School of Life Science, Peking University, Beijing, China) for the in excess primers, which can contribute to nonspecific products. Finally, situ hybridizations. We are grateful to Ken Gross (Roswell Park Can- we optimized the fragmentation and labeling procedure to produce cer Institute) for helpful discussions. target properly sized for Affymetrix oligonucleotide array hybridization. See Supplemental Methods for the detailed protocol and validation of SCAMP. DISCLOSURES None. RT-PCR RNA extraction, reverse transcription, and PCR were performed as described previously.41 REFERENCES Immunostaining Mice were anesthetized with tribromoethanol. The kidneys were re- 1. Keeton TK, Campbell WB: The pharmacologic alteration of renin release. Pharmacol Rev 32: 81–227, 1980 moved, weighed, and either frozen for in situ hybridization, placed in 2. Gomez RA, Lynch KR, Sturgill BC, Elwood JP, Chevalier RL, Carey RM, RNAlater for RNA extraction, or fixed in Bouin’s fixative. Immuno- Peach MJ: Distribution of renin mRNA and its protein in the develop- staining was performed on 5-␮m thick paraffin sections using the ing kidney. Am J Physiol 257: F850–F858, 1989 following primary antibodies and the appropriate Vectastain ABC kit 3. Sequeira Lopez ML, Pentz ES, Nomasa T, Smithies O, Gomez RA: (Vector Laboratories, Burlingame, CA). The primary antibodies used Renin cells are precursors for multiple cell types that switch to the renin phenotype when homeostasis is threatened. Dev Cell 6: 719– were rabbit anti-mouse renin polyclonal, 1:500, ␣-SMA, 1:10,000 728, 2004 (A2547; Sigma, St. Louis, MO), Akr1b7, 1:200 (sc-27763; Santa Cruz, 4. Pentz ES, Sequeira Lopez ML, Cordaillat M, Gomez RA: Identity of the Santa Cruz, CA), NFATc4, 1:500 (62613; Abcam, Inc., Cambridge, renin cell is mediated by cAMP and chromatin remodeling: An in vitro MA), MEF2C, 1:500 (79436; Abcam, Inc.), Crip1, 1:200 (sc-131473; model for studying cell recruitment and plasticity. Am J Physiol Heart Santa Cruz), Notch3, 1:100 (1308-NT; R&D Systems, Inc., Minneap- Circ Physiol 294: H699–H707, 2008 5. Brunskill EW, Aronow BJ, Georgas K, Rumballe B, Valerius MT, olis, MN), S1pR, 1:100 (LS-c11168 Lifespan Biosciences, Seattle, WA) Aronow J, Kaimal V, Jegga AG, Grimmond S, McMahon AP, Patterson and Nkx3.1, 1:1000 (gift of Dr. Chuck Bieberich, University of Mary- LT, Little MH, Potter SS: Atlas of gene expression in the developing land, Baltimore, MD). kidney at microanatomic resolution. Dev Cell 15: 781–791, 2008 6. Rohrwasser A, Ishigami T, Gociman B, Lantelme P, Morgan T, Cheng Chromatin Immunoprecipitation (ChIP) T, Hillas E, Zhang S, Ward K, Bloch-Faure M, Meneton P, Lalouel JM: Renin and kallikrein in connecting tubule of mouse. Kidney Int 64: ChIP was performed as described previously4 using chromatin from 2155–2162, 2003 adult kidney cortex and skeletal muscle for the phopsho-Creb ChIP, 7. Prieto-Carrasquero MC, Botros FT, Pagan J, Kobori H, Seth DM, and from arterial smooth muscle cells of the renin lineage4 in the Casarini DE, Navar LG: Collecting duct renin is upregulated in both RBP-J studies. The antibodies used were Phospho-Creb (Ser 133, kidneys of 2-kidney, 1-clip goldblatt hypertensive rats. Hypertension 9198S; Cell Signaling Technology, Inc., Danvers, MA) and RBP-J 51: 1590–1596, 2008 8. Takahashi N, Lopez ML, Cowhig JE Jr, Taylor MA, Hatada T, Riggs E, (AB5790; Millipore, Billerica, MA). Lee G, Gomez RA, Kim HS, Smithies O: Ren1c homozygous null mice are hypotensive and polyuric, but heterozygotes are indistinguishable In situ Hybridization from wild-type. J Am Soc Nephrol 16: 125–132, 2005 In situ hybridization in kidneys from newborn mice was performed as 9. Cho H, Kozasa T, Bondjers C, Betsholtz C, Kehrl JH: Pericyte-specific described previously.42 expression of Rgs5: Implications for PDGF and EDG receptor signaling during vascular maturation. FASEB J 17: 440–442, 2003 10. Gaengel K, Genove G, Armulik A, Betsholtz C: Endothelial-mural cell Public Data Availability signaling in vascular development and angiogenesis. Arterioscler All data are available on the public GUDMAP website (GUDMA- Thromb Vasc Biol 29: 630–638, 2009 P.org), which includes links to GEO, where the array data are also 11. Bondjers C, Kalen M, Hellstrom M, Scheidl SJ, Abramsson A, Renner available. O, Lindahl P, Cho H, Kehrl J, Betsholtz C: Transcription profiling of platelet-derived growth factor-B-deficient mouse embryos identifies RGS5 as a novel marker for pericytes and vascular smooth muscle cells. Am J Pathol 162: 721–729, 2003 ACKNOWLEDGMENTS 12. Latonen L, Jarvinen PM, Laiho M: Cytoskeleton-interacting LIM-domain protein CRP1 suppresses cell proliferation and protects from stress-induced cell death. Exp Cell Res 314: 738–747, 2008 The support of the GUDMAP program is greatly appreciated. SP and 13. Pomies P, Louis HA, Beckerle MC: CRP1, a LIM domain protein impli- BA are supported by 3UO1DK70251. RAG is supported by cated in muscle differentiation, interacts with alpha-actinin. J Cell Biol R37HL066242 and RO1HL096735. MLSSL is supported by 139: 157–168, 1997 KO8DK75481. JY is supported by 1R01DK085080. We thank Niloo- 14. Sequeira Lopez ML, Pentz ES, Robert B, Abrahamson DR, Gomez RA: far Latifi, Yan Hu, and Cristina Monteagudo for immunostainings; Embryonic origin and lineage of juxtaglomerular cells. Am J Physiol Renal Physiol 281: F345–F356, 2001 Ruth Castellanos-Rivera for Notch3 immunostaining and the RBP-J 15. Li S, Wang DZ, Wang Z, Richardson JA, Olson EN: The serum re- ϩ PCR in YFP cells; William H. Wilson IV for initial bioinformatics sponse factor coactivator myocardin is required for vascular smooth discussions; Magali Cordaillat for the ChIP data; and Yanru Dou muscle development. Proc Natl Acad SciUSA100: 9366–9370, 2003

2224 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 2213–2225, 2011 www.jasn.org BASIC RESEARCH

16. Jones ME, Kondo M, Zhuang Y: A tamoxifen inducible knock-in allele and prostaglandin F2alpha are regulators of adrenal endocrine func- for investigation of E2A function. BMC Dev Biol 9: 51, 2009 tions. PLoS ONE 4: e7309, 2009 17. Aronheim A, Shiran R, Rosen A, Walker MD: The E2A gene product 30. Gomez RA, Norwood VF: Recent advances in renal development. Curr contains two separable and functionally distinct transcription activa- Opin Pediatr 11: 135–140, 1999 tion domains. Proc Natl Acad SciUSA90: 8063–8067, 1993 31. Reddi V, Zaglul A, Pentz ES, Gomez RA: Renin-expressing cells are 18. Graef IA, Chen F, Chen L, Kuo A, Crabtree GR: Signals transduced by associated with branching of the developing kidney vasculature. JAm Ca(2ϩ)/calcineurin and NFATc3/c4 pattern the developing vascula- Soc Nephrol 9: 63–71, 1998 ture. Cell 105: 863–875, 2001 32. Peti-Peterdi J: Multiphoton imaging of renal tissues in vitro. Am J 19. Kong J, Qiao G, Zhang Z, Liu SQ, Li YC: Targeted vitamin D receptor Physiol Renal Physiol 288: F1079–F1083, 2005 expression in juxtaglomerular cells suppresses renin expression inde- 33. Peti-Peterdi J, Toma I, Sipos A, Vargas SL: Multiphoton imaging of pendent of parathyroid hormone and calcium. Kidney Int 74: 1577– renal regulatory mechanisms. Physiology (Bethesda) 24: 88–96, 2009 1581, 2008 34. Takenaka T, Suzuki H, Okada H, Hayashi K, Kanno Y, Saruta T: Mecha- 20. Desch M, Schreiber A, Schweda F, Madsen K, Friis UG, Weatherford nosensitive cation channels mediate afferent arteriolar myogenic con- ET, Sigmund CD, Sequeira Lopez ML, Gomez RA, Todorov VT: In- striction in the isolated rat kidney. J Physiol 511(Pt 1):245–253, 1998 creased renin production in mice with deletion of peroxisome prolif- 35. Kurtz L, Schweda F, de WC, Kriz W, Witzgall R, Warth R, Sauter A, erator-activated receptor-gamma in juxtaglomerular cells. Hyperten- Kurtz A, Wagner C: Lack of connexin 40 causes displacement of sion 55: 660–666, 2010 renin-producing cells from afferent arterioles to the extraglomerular 21. Castellanos Rivera RM, Monteagudo MC, Pentz ES, Glenn ST, Gross mesangium. J Am Soc Nephrol 18: 1103–1111, 2007 KW, Carretero O, Sequeira-Lopez MLS, Gomez RA: The transcriptional 36. Castellanos Rivera RM, Monteagudo MC, Pentz ES, Glenn ST, Gross regulator RBP-J regulates the number and plasticity of renin cells. KW, Carretero O, Sequeira-Lopez MLS, Gomez RA: Transcriptional Physiol Genomics 2011 Jul 12. [Epub ahead of print] regulator RBP-J regulates the number and plasticity of renin cells. 22. Lopez ML, Gomez RA: The renin phenotype: roles and regulation in Physiol Genomics 43: 1021–1028, 2011 the kidney. Curr Opin Nephrol Hypertens 19: 366–371, 2010 37. Brady G, Iscove NN: Construction of cDNA libraries from single cells. 23. Komukai K, Mochizuki S, Yoshimura M: Gender and the renin-angio- Methods Enzymol 225: 611–623, 1993 tensin-aldosterone system. Fundam Clin Pharmacol 24: 687–698, 38. Cherry TJ, Trimarchi JM, Stadler MB, Cepko CL: Development and 2010 diversification of retinal amacrine interneurons at single cell resolution. 24. Heldring N, Pike A, Andersson S, Matthews J, Cheng G, Hartman J, Proc Natl Acad SciUSA106: 9495–9500, 2009 Tujague M, Strom A, Treuter E, Warner M, Gustafsson JA: Estrogen 39. Villalva C, Touriol C, Seurat P, Trempat P, Delsol G, Brousset P: receptors: How do they signal and what are their targets. Physiol Rev Increased yield of PCR products by addition of T4 gene 32 protein to 87: 905–931, 2007 the SMART PCR cDNA synthesis system. Biotechniques 31:81–83, 25. Pentz ES, Lopez ML, Kim HS, Carretero O, Smithies O, Gomez RA: 2001 Ren1d and Ren2 cooperate to preserve homeostasis: Evidence from 40. Jefferies D, Farquharson C: Effects of choice of reverse-transcriptase mice expressing GFP in place of Ren1d. Physiol Genomics 6: 45–55, enzyme and use of T4 gene 32 protein on banding patterns in agarose 2001 gel differential display. Anal Biochem 308: 192–194, 2002 26. Casellas D, Dupont M, Kaskel FJ, Inagami T, Moore LC: Direct visu- alization of renin-cell distribution in preglomerular vascular trees dis- 41. Gomez RA, Pentz ES, Jin X, Cordaillat M, Sequeira Lopez ML: CBP and sected from rat kidney. Am J Physiol 265: F151–F156, 1993 p300 are essential for renin cell identity and morphological integrity of 27. Berger M, Bergers G, Arnold B, Hammerling GJ, Ganss R: Regulator of the kidney. Am J Physiol Heart Circ Physiol 296: H1255–H1262, 2009 G-protein signaling-5 induction in pericytes coincides with active ves- 42. Yu J, Carroll TJ, Rajagopal J, Kobayashi A, Ren Q, McMahon AP: A sel remodeling during neovascularization. Blood 105: 1094–1101, Wnt7b-dependent pathway regulates the orientation of epithelial cell 2005 division and establishes the cortico-medullary axis of the mammalian 28. Barski OA, Tipparaju SM, Bhatnagar A: The aldo-keto reductase su- kidney. Development 136: 161–171, 2009 perfamily and its role in drug metabolism and detoxification. Drug Metab Rev 40: 553–624, 2008 29. Lambert-Langlais S, Pointud JC, Lefrancois-Martinez AM, Volat F, Manin M, Coudore F, Val P, Sahut-Barnola I, Ragazzon B, Louiset E, Supplemental information for this article is available online at http://www.jasn. Delarue C, Lefebvre H, Urade Y, Martinez A: Aldo keto reductase 1B7 org/

J Am Soc Nephrol 22: 2213–2225, 2011 The Genetic Network of Renin Cells 2225 Erratum

CORRECTION related with YRPW motif 1, should be Heyl (HeyL), which represents the gene hairy/enhancer-of-split related with YRPW Brunskill EW,Sequeira-Lopez MLS, Pentz ES, Lin E, YuJ, Aronow motif-like, in Figure 4B and legend and in the ﬁfth line above BJ, Potter SS, Gomez RA: Genes That Confer the Identity of the the Discussion section of page 8 of the PDF document. Renin Cell. J Am Soc Nephrol 22: 2213–2225, 2011. The Gene We apologize for the error and the inconvenience this may symbol Hey1, which represents the gene hairy/enhancer-of-split have caused.

J Am Soc Nephrol 23: 567, 2012 ISSN : 1046-6673/2303-567 567 Supplemental Experimental Procedures

Single Cell AMplification Procedure (SCAMP)

The tools available for the analysis of gene expression patterns of very small samples, including single cells, remain imperfect. They can be divided into three basic biochemistries. One method uses multiple rounds of in vitro transcription1, mediated by a

T7 promoter incorporated into the cDNA. This approach has been used with success for

small laser capture samples and even single cells.2-5 Nevertheless, the procedure is

arduous, requires days to complete, the results are variable6, and available commercial

kits specify input RNA amounts far exceeding single cell quantities. A second method,

designated RiboSpia, uses a displacement DNA polymerase reaction7 to drive a single

round of amplification. The RiboSpia OneDirect kit from Nugen is stated to work with

RNA from even a single cell. A chief disadvantage is the very high cost. A third method

uses PCR based target amplification. Perhaps surprisingly, this has been shown to be

capable of providing faithful representation from even extremely small starting samples.8

Detailed protocol for PCR target amplification in single cells [optimized from the

protocol9 previously used by Cepko and colleagues.10 1 μl of RNA, or even a single cell,

was added to 5 μl of RT-Lysis Buffer. RT-Lysis Buffer has 47 μl Lysis Buffer, 1 μl

RNAseout (Invitrogen), 1 μl dNTP (2.5 mM each, Takara), 0.5 μl oligodT (20 ng/μl), 0.5

μl oligodT+N (5 ng/μl). Oligonucleotides used in these studies were supplied by Oligos

Etc. (Wilsonville, OR) and contained the sequences 5’-

TATAGAATTCGCGGCCGCTCGCGATTTTTTTTTTTTTTTTTTTTTTTT (Oligo dT) and 5’- TATAGAATTCGCGGCCGCTCGCGATTTTTTTTTTTTTTTTTTTTTTTTNNNNNN

(Oligo dT+N). Lysis buffer consists of 100 μl 10X Roche PCR Buffer, 5 μl NP-40, 50 μl

O 0.1 M DTT, 60 μl 25 mM MgCl2, 785 μl H2O. The samples were heated to 65 C 2 min,

cooled to 4OC, and centrifuged. The random primer added in the initial reverse

transcription (RT) step gives improved full-length representation of transcripts which

improves results when using, for example, the Affymetrix Gene ST arrays, which carry

probes across multiple exons, and are not strongly biased for the 3’ ends of genes.

RT reactions were set up by adding 0.8 μl Superscript Mix, incubating 25OC for 5 min,

37OC for 30 min, heat-inactivated at 70OC for 10 min, cooled to 4OC and centrifuged.

Superscript mix consists of 3 μl Superscript III (Invitrogen), 0.5 μl RNAseout

(Invitrogen), 0.3 μl T4 gene 32 protein (New England Biolabs). For exonuclease reactions, 0.5 μl exonuclease (New England Biolabs) was added, incubated 37OC for 30

min, heat-inactivated at 80OC for 25 min, cooled to 4OC and centrifuged. Tailing

reactions were carried out by adding 3 μl TdT mix, incubated 37OC for 20 min, heat- inactivated at 70OC 10 min, cooled to 4OC and centrifuged. TdT mix contains 0.15 μl of

100 mM dATP (Roche), 0.3 μl 10XPCR buffer (Roche), 1.37 μl H2O, 0.5 μl TdT

(Roche), 0.5 μl RNaseH (Ambion) and 0.18 μl of 25mM MgCl2. PCR reactions were

carried out by adding 90 μl PCR mix, and cycling at 95OC for 1 min, 95OC for 1 min,

37OC for 5 min, 72OC for 16 min, 93OC for 40 sec, 67OC for 1 min, 72OC for 6 min + 6

sec per cycle, repeating the cycling conditions 34 times, 72OC for 10 min, 4OC forever.

The PCR mix consists of 10μl 10X PCR buffer, 10μl dNTPs (2.5mM), 2μl Oligo dT

(1μg/ul), 1μl LA-TAQ polymerase and 67μl of H2O. PCR reagents were purchased from Takara Bio. The resulting amplified cDNA products were purified using Qiagen PCR

purification kit (Qiagen).

The following optimized fragmentation and labeling procedure produces target

properly sized for Affymetrix oligonucleotide array hybridization. 5 μg of cDNA were fragmented in DNase I fragmentation mix containing two units DNase I (Roche) and 1X

One-PHor buffer (100mM K-glutamate, 0.25mM Tris-acetate, 0.1mM Mg-Acetate). The cDNA was incubated at 37OC for 13 min, heat-inactivated at 95OC for 15 min, cooled to

4OC and centrifuged. The fragmented cDNAs were biotinylated by adding 1.5 μl of

1mM Bio-N6-ddATP (ENZO Life Scences) and 1 μl of TdT (Roche). The samples were incubated at 37OC for 90 min, heat-inactivated at 65OC for 15 min., cooled to 4OC and centrifuged.

To evaluate the quality of SCAMP data we performed comparisons with two commercial systems, Nugen RiboSpia OneDirect and Miltenyi μMACS SuperAmp, which is also PCR based. Amplifications were performed starting with 25 and 50 picograms (pg) of RNA standard that was made from whole newborn mice. Total RNA content per cell will depend on cell type, but is generally in the range of 5-30 pg per cell

11;12. Miltenyi provided a technician that performed the μMACS SuperAmp PCR based procedure, according to their protocols and with their equipment, which includes special columns, while we carried out the Nugen OneDirect and the SCAMP protocols.

Technical replicates were performed, in each case starting with the same newborn mouse homogenate standard RNA, but performing independent amplifications and microarray hybridizations, with Affymetrix Mouse Gene 1.0 ST arrays. A total of sixteen test small sample amplifications and array hybridizations were

carried out, including two using Miltenyi μMACS SuperAmp (one 25 pg and one 50 pg),

four with Nugen OneDirect (two 25 pg and two 50 pg), and ten with SCAMP (three 10

pg, four 25 pg and three 50 pg). Pearson Correlation Coefficients (PCC) were calculated

to provide a measure of reproducibility using GeneSpring GX 11.0.2. The PCC values

generated were 0.917 for μMACS SuperAmp, and 0.924-0.949 for Nugen OneDirect and

0.860-0.941 for SCAMP. The 10 pg samples, only attempted with SCAMP, gave more

noise, as might be expected, but all three systems gave excellent reproducibility.

M25 M25 M50 0.92 M50 D25 0.81 0.82 D25 D25 0.79 0.82 0.92 D25 D50 0.80 0.83 0.94 0.94 D50 D50 0.81 0.84 0.94 0.93 0.95 D50 S10 0.76 0.80 0.79 0.77 0.78 0.79 S10 S10 0.77 0.82 0.79 0.78 0.79 0.79 0.86 S10 S10 0.79 0.84 0.80 0.79 0.80 0.81 0.87 0.88 S10 S25 0.75 0.81 0.79 0.78 0.80 0.80 0.87 0.88 0.89 S25 S25 0.74 0.79 0.79 0.78 0.79 0.79 0.86 0.87 0.87 0.89 S25 S25 0.71 0.77 0.78 0.77 0.79 0.79 0.85 0.86 0.86 0.88 0.87 S25 S25 0.71 0.77 0.78 0.77 0.78 0.78 0.85 0.86 0.87 0.89 0.87 0.88 S25 S50 0.71 0.78 0.78 0.76 0.78 0.78 0.87 0.87 0.88 0.90 0.89 0.89 0.90 S50 S50 0.72 0.78 0.77 0.77 0.78 0.78 0.87 0.87 0.88 0.91 0.90 0.89 0.90 0.92 S50 S50 0.73 0.80 0.79 0.78 0.80 0.80 0.88 0.89 0.91 0.92 0.90 0.91 0.93 0.94 0.93 Pearson correlation coefficients for Miltenyi (M), D (OneDirect) and S (SCAMP)

methods. Numbers next to M, D, S designate picograms of total RNA used.

We further tested the quality of the SCAMP data. One important measure of

target amplification quality is sensitivity, or the ability to detect low abundance

transcripts. Analysis of the array data indicated that SCAMP gave greater sensitivity than either OneDirect or SuperAmp, with ANOVA results showing SCAMP detection of

expression of significantly more genes, and with higher mean probe level intensities

(Figure S1). In addition, for Affymetrix Gene ST arrays there is a quality metric

designated “area under the curve”, or AUC. This is a measure of relative signal intensities

for the exons of about 100 housekeeping genes, compared to introns for the same genes.

A higher AUC therefore indicates more specific signal. The average AUC values for the

50 pg samples were 0.870 for SCAMP, 0.825 for OneDirect, and only 0.700 for

SuperAmp. SCAMP gave clearly superior data as measured by this key quality metric.

These overall results show that SCAMP is a cost effective, easily executed system for the

target amplification of extremely small samples, yielding high quality data.

We therefore used the SCAMP procedure for target amplification of the five

individual YFP positive juxtaglomerular renin producing cells.

REFERENCES

1. Van Gelder RN, von Zastrow ME, Yool A, Dement WC, Barchas JD, Eberwine JH: Amplified RNA synthesized from limited quantities of heterogeneous cDNA. Proc Natl Acad Sci U S A 87:1663-1667, 1990

2. Kamme F, Salunga R, Yu J, Tran DT, Zhu J, Luo L, Bittner A, Guo HQ, Miller N, Wan J, Erlander M: Single-cell microarray analysis in hippocampus CA1: demonstration and validation of cellular heterogeneity. J Neurosci 23:3607-3615, 2003

3. Potter SS, Hartman HA, Kwan KM, Behringer RR, Patterson LT: Laser capture- microarray analysis of Lim1 mutant kidney development. Genesis 45:432-439, 2007

4. Bennett MR, Czech KA, Arend LJ, Witte DP, Devarajan P, Potter SS: Laser capture microdissection-microarray analysis of focal segmental glomerulosclerosis glomeruli. Nephron Exp Nephrol 107:e30-e40, 2007

5. Brunskill EW, Aronow BJ, Georgas K, Rumballe B, Valerius MT, Aronow J, Kaimal V, Jegga AG, Grimmond S, McMahon AP, Patterson LT, Little MH, Potter SS: Atlas of gene expression in the developing kidney at microanatomic resolution. Dev Cell 15:781-791, 2008

6. Clement-Ziza M, Gentien D, Lyonnet S, Thiery JP, Besmond C, Decraene C: Evaluation of methods for amplification of picogram amounts of total RNA for whole genome expression profiling. BMC Genomics 10:246, 2009

7. Kurn N, Chen P, Heath JD, Kopf-Sill A, Stephens KM, Wang S: Novel isothermal, linear nucleic acid amplification systems for highly multiplexed applications. Clin Chem 51:1973-1981, 2005

8. Iscove NN, Barbara M, Gu M, Gibson M, Modi C, Winegarden N: Representation is faithfully preserved in global cDNA amplified exponentially from sub- picogram quantities of mRNA. Nat Biotechnol 20:940-943, 2002

9. Brady G, Iscove NN: Construction of cDNA libraries from single cells. Methods Enzymol 225:611-623, 1993

10. Cherry TJ, Trimarchi JM, Stadler MB, Cepko CL: Development and diversification of retinal amacrine interneurons at single cell resolution. Proc Natl Acad Sci U S A 106:9495-9500, 2009

11. Copois V, Bret C, Bibeau F, Brouillet JP, Del RM, Berthe ML, Maudelonde T, Boulle N: Assessment of RNA quality extracted from laser-captured tissues using miniaturized capillary electrophoresis. Lab Invest 83:599-602, 2003

12. Kane MD, Jatkoe TA, Stumpf CR, Lu J, Thomas JD, Madore SJ: Assessment of the sensitivity and specificity of oligonucleotide (50mer) microarrays. Nucleic Acids Res 28:4552-4557, 2000

Supplemental Figure 1