Different subpopulations of kidney interstitial cells produce erythropoietin and factors supporting tissue oxygenation in response to hypoxia in vivo

Genetic induction of hypoxia signaling by deletion of the von Hippel-Lindau (Vhl) protein in mesenchymal PDGFR-β+ cells leads to abundant HIF-2 dependent erythropoietin (EPO) expression in the cortex and outer medulla of the kidney. This rather unique feature of kidney PDGFR-β+ cells promote questions about their special characteristics and general functional response to hypoxia. To address these issues, we characterized kidney PDGFR-β+ EPO expressing cells based on additional cell markers and their gene expression profile in response to hypoxia signaling induced by targeted deletion of Vhl or exposure to low oxygen and carbon monoxide respectively, and after unilateral ureteral obstruction. CD73+, Gli1+, tenascin C+ and interstitial SMMHC+ cells were identified as zonally distributed subpopulations of PDGFR-β+ cells. EPO expression could be induced by Vhl deletion in all PDGFR-β+ subpopulations. Under hypoxemic conditions, recruited EPO+ cells were mostly part of the CD73+ subpopulation. Besides EPO production, expression of adrenomedullin and regulator of G-protein signaling 4 was upregulated in PDGFR-β+ subpopulations in response to the different hypoxic stimuli. Thus, different kidney interstitial PDGFR-β+ subpopulations exist, capable of producing EPO in response to different stimuli. Activation of hypoxia signaling in these cells also induces factors likely contributing to improved kidney interstitial tissue oxygenation. Genetic induction of hypoxia signaling by deletion of the von Hippel-Lindau (Vhl) protein in mesenchymal PDGFR-β+ cells leads to abundant HIF-2 dependent erythropoietin (EPO) expression in the cortex and outer medulla of the kidney. This rather unique feature of kidney PDGFR-β+ cells promote questions about their special characteristics and general functional response to hypoxia. To address these issues, we characterized kidney PDGFR-β+ EPO expressing cells based on additional cell markers and their gene expression profile in response to hypoxia signaling induced by targeted deletion of Vhl or exposure to low oxygen and carbon monoxide respectively, and after unilateral ureteral obstruction. CD73+, Gli1+, tenascin C+ and interstitial SMMHC+ cells were identified as zonally distributed subpopulations of PDGFR-β+ cells. EPO expression could be induced by Vhl deletion in all PDGFR-β+ subpopulations. Under hypoxemic conditions, recruited EPO+ cells were mostly part of the CD73+ subpopulation. Besides EPO production, expression of adrenomedullin and regulator of G-protein signaling 4 was upregulated in PDGFR-β+ subpopulations in response to the different hypoxic stimuli. Thus, different kidney interstitial PDGFR-β+ subpopulations exist, capable of producing EPO in response to different stimuli. Activation of hypoxia signaling in these cells also induces factors likely contributing to improved kidney interstitial tissue oxygenation. Translational StatementTissue oxygenation is dependent on proper erythropoietin (EPO) production and local regulatory factors. We investigated different subpopulations of possible EPO-producing cells in the mouse kidney under hypoxemic stimuli and experimental kidney fibrosis. These cells comprise a heterogeneous group of interstitial cells, which share the expression of PDGFR-β. Induction of hypoxia signaling in these cells induces the expression of factors supporting tissue oxygenation. These data could provide a starting point for closer investigations of the fine regulation of tissue oxygenation. In addition, the here-defined subpopulations of PDGFR-β+ interstitial cells could present new targets for a pharmacologic approach to treat EPO deficiency. Tissue oxygenation is dependent on proper erythropoietin (EPO) production and local regulatory factors. We investigated different subpopulations of possible EPO-producing cells in the mouse kidney under hypoxemic stimuli and experimental kidney fibrosis. These cells comprise a heterogeneous group of interstitial cells, which share the expression of PDGFR-β. Induction of hypoxia signaling in these cells induces the expression of factors supporting tissue oxygenation. These data could provide a starting point for closer investigations of the fine regulation of tissue oxygenation. In addition, the here-defined subpopulations of PDGFR-β+ interstitial cells could present new targets for a pharmacologic approach to treat EPO deficiency. The kidneys are the main site of oxygen-regulated erythropoietin (EPO) production. Within the kidneys, tubulointerstitial cells produce EPO1Bachmann S. Le Hir M. Eckardt K.U. Co-localization of erythropoietin mRNA and ecto-5’-nucleotidase immunoreactivity in peritubular cells of rat renal cortex indicates that fibroblasts produce erythropoietin.J Histochem Cytochem. 1993; 41: 335-341Crossref PubMed Scopus (305) Google Scholar,2Maxwell P.H. Osmond M.K. Pugh C.W. et al.Identification of the renal erythropoietin-producing cells using transgenic mice.Kidney Int. 1993; 44: 1149-1162Abstract Full Text PDF PubMed Scopus (341) Google Scholar triggered by hypoxia inducible factor 2 (HIF-2).3Paliege A. Rosenberger C. Bondke A. et al.Hypoxia-inducible factor-2α-expressing interstitial fibroblasts are the only renal cells that express erythropoietin under hypoxia-inducible factor stabilization.Kidney Int. 2010; 77: 312-318Abstract Full Text Full Text PDF PubMed Scopus (130) Google Scholar The regulation of renal EPO production occurs mainly by recruitment of EPO-producing cells that express EPO in an on/off fashion.4Eckardt K.-U. Koury S.T. Tan C.C. et al.Distribution of erythropoietin producing cells in rat kidneys during hypoxic hypoxia.Kidney Int. 1993; 43: 815-823Abstract Full Text PDF PubMed Scopus (90) Google Scholar In the normoxic kidney, few EPO-expressing cells are found at the boundary between cortex and outer medulla (OM), whereas under severe hypoxic stress, EPO-expressing cells are recruited in the whole cortex and OM.5Koury S.T. Bondurant M.C. Semenza G.L. Koury M.J. The use of in situ hybridization to study erythropoietin gene expression in murine kidney and liver.Microsc Res Tech. 1993; 25: 29-39Crossref PubMed Scopus (17) Google Scholar, 6Yamazaki S. Souma T. Hirano I. et al.A mouse model of adult-onset anaemia due to erythropoietin deficiency.Nat Commun. 2013; 4: 1950Crossref PubMed Scopus (53) Google Scholar, 7Maxwell P.H. Ferguson D.J.P. Nicholls L.G. et al.Sites of erythropoietin production.Kidney Int. 1997; 51: 393-401Abstract Full Text PDF PubMed Scopus (91) Google Scholar The identity and functional characteristics of those active and potentially EPO-producing cells are less clear. Originally, EPO-producing cells had been identified as interstitial fibroblast-like cells expressing CD73 (ecto-5′-nucleotidase).1Bachmann S. Le Hir M. Eckardt K.U. Co-localization of erythropoietin mRNA and ecto-5’-nucleotidase immunoreactivity in peritubular cells of rat renal cortex indicates that fibroblasts produce erythropoietin.J Histochem Cytochem. 1993; 41: 335-341Crossref PubMed Scopus (305) Google Scholar More recent studies have suggested that in addition to CD73+ cells, also other cell types are able to express EPO.8Pan X. Suzuki N. Hirano I. et al.Isolation and characterization of renal erythropoietin-producing cells from genetically produced anemia mice.PLoS One. 2011; 6e25839Crossref PubMed Scopus (85) Google Scholar Such a heterogeneity of renal EPO-producing cells has also been suggested by data from mice lacking HIF prolyl-4-hydroxylases in renal stroma cell precursors9Kobayashi H. Liu Q. Binns T.C. et al.Distinct subpopulations of FOXD1 stroma-derived cells regulate renal erythropoietin..J Clin Invest. 2016; 126: 1926-1938Crossref PubMed Scopus (78) Google Scholar and has received support by the concept that EPO-producing cells could be telocytes10Imeri F. Nolan K.A. Bapst A.M. et al.Generation of renal Epo-producing cell lines by conditional gene tagging reveals rapid HIF-2 driven Epo kinetics, cell autonomous feedback regulation, and a telocyte phenotype.Kidney Int. 2019; 95: 375-387Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar or pericytes.11Kramann R. Humphreys B.D. Kidney pericytes: roles in regeneration and fibrosis.Semin Nephrol. 2014; 34: 374-383Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar EPO-producing cells have a stellate form with multiple extensions such as pericytes,10Imeri F. Nolan K.A. Bapst A.M. et al.Generation of renal Epo-producing cell lines by conditional gene tagging reveals rapid HIF-2 driven Epo kinetics, cell autonomous feedback regulation, and a telocyte phenotype.Kidney Int. 2019; 95: 375-387Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar,12Souma T. Nezu M. Nakano D. et al.Erythropoietin synthesis in renal myofibroblasts is restored by activation of hypoxia signaling.J Am Soc Nephrol. 2016; 27: 428-438Crossref PubMed Scopus (112) Google Scholar,13Gerl K. Nolan K.A. Karger C. et al.Erythropoietin production by PDGFR-β+ cells.Pflugers Arch. 2016; 468: 1479-1487Crossref PubMed Scopus (31) Google Scholar and they are located directly adjacent to capillaries.1Bachmann S. Le Hir M. Eckardt K.U. Co-localization of erythropoietin mRNA and ecto-5’-nucleotidase immunoreactivity in peritubular cells of rat renal cortex indicates that fibroblasts produce erythropoietin.J Histochem Cytochem. 1993; 41: 335-341Crossref PubMed Scopus (305) Google Scholar,5Koury S.T. Bondurant M.C. Semenza G.L. Koury M.J. The use of in situ hybridization to study erythropoietin gene expression in murine kidney and liver.Microsc Res Tech. 1993; 25: 29-39Crossref PubMed Scopus (17) Google Scholar,12Souma T. Nezu M. Nakano D. et al.Erythropoietin synthesis in renal myofibroblasts is restored by activation of hypoxia signaling.J Am Soc Nephrol. 2016; 27: 428-438Crossref PubMed Scopus (112) Google Scholar Moreover, pericytes are key players in renal fibrosis.11Kramann R. Humphreys B.D. Kidney pericytes: roles in regeneration and fibrosis.Semin Nephrol. 2014; 34: 374-383Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar Potential EPO-producing cells are the precursors of myofibroblasts in the kidney14Souma T. Yamazaki S. Moriguchi T. et al.Plasticity of renal erythropoietin-producing cells governs fibrosis.J Am Soc Nephrol. 2013; 24: 1599-1616Crossref PubMed Scopus (135) Google Scholar and the ability of the kidneys to produce EPO ceases with interstitial myofibroblast formation in states of kidney fibrosis.12Souma T. Nezu M. Nakano D. et al.Erythropoietin synthesis in renal myofibroblasts is restored by activation of hypoxia signaling.J Am Soc Nephrol. 2016; 27: 428-438Crossref PubMed Scopus (112) Google Scholar,15Eschbach J.W. Adamson J.W. Anemia of end-stage renal disease (ESRD).Kidney Int. 1985; 28: 1-5Abstract Full Text PDF PubMed Scopus (287) Google Scholar, 16Babitt J.L. Lin H.Y. Mechanisms of anemia in CKD.J Am Soc Nephrol. 2012; 23: 1631-1634Crossref PubMed Scopus (535) Google Scholar, 17Maxwell P.H. Ferguson D.J.P. Nicholls L.G. et al.The interstitial response to renal injury: fibroblast-like cells show phenotypic changes and have reduced potential for erythropoietin gene expression.Kidney Int. 1997; 52: 715-724Abstract Full Text PDF PubMed Scopus (62) Google Scholar This concept of EPO-producing cells being pericytes, however, is somewhat contradicted by the fact that classical pericytes are located mainly in the OM and to a lesser extent in the cortex.18Lemley K.V. Kriz W. Anatomy of the renal interstitium.Kidney Int. 1991; 39: 370-381Abstract Full Text PDF PubMed Scopus (203) Google Scholar The pericyte concept is also questioned by the finding that HIF stabilization in cells expressing the fibroblast cell marker PDGFR-β19Armulik A. Genové G. Betsholtz C. Pericytes: developmental, physiological, and pathological perspectives, problems, and promises.Dev Cell. 2011; 21: 193-215Abstract Full Text Full Text PDF PubMed Scopus (1714) Google Scholar strongly induces EPO expression in the interstitium of the cortex and the OM,13Gerl K. Nolan K.A. Karger C. et al.Erythropoietin production by PDGFR-β+ cells.Pflugers Arch. 2016; 468: 1479-1487Crossref PubMed Scopus (31) Google Scholar whereas HIF stabilization in cells expressing the pericyte marker NG219Armulik A. Genové G. Betsholtz C. Pericytes: developmental, physiological, and pathological perspectives, problems, and promises.Dev Cell. 2011; 21: 193-215Abstract Full Text Full Text PDF PubMed Scopus (1714) Google Scholar only negligibly induces EPO expression in the kidney.20Urrutia A.A. Afzal A. Nelson J. et al.Prolyl-4-hydroxylase 2 and 3 coregulate murine erythropoietin in brain pericytes.Blood. 2016; 128: 2550-2560Crossref PubMed Scopus (26) Google Scholar All renal cells with inducible EPO expression appear to express PDGFR-β, thus defining PDGFR-β as an essential marker of renal EPO-producing cells.13Gerl K. Nolan K.A. Karger C. et al.Erythropoietin production by PDGFR-β+ cells.Pflugers Arch. 2016; 468: 1479-1487Crossref PubMed Scopus (31) Google Scholar PDGFR-β+ cells exist in all organs but physiologically regulated EPO production in these cells appears to be restricted to the kidneys and the brain.13Gerl K. Nolan K.A. Karger C. et al.Erythropoietin production by PDGFR-β+ cells.Pflugers Arch. 2016; 468: 1479-1487Crossref PubMed Scopus (31) Google Scholar,20Urrutia A.A. Afzal A. Nelson J. et al.Prolyl-4-hydroxylase 2 and 3 coregulate murine erythropoietin in brain pericytes.Blood. 2016; 128: 2550-2560Crossref PubMed Scopus (26) Google Scholar The restriction of EPO expression to the kidneys is likely not the result of organ-specific differences of oxygen sensing, but is rather due to the selective ability of renal PDGFR-β+ cells to express EPO. This assumption is supported by the finding that cell-type–specific stabilization of HIF-proteins in PDGFR-β+ cells by deletion of the ubiquitin ligase von Hippel-Lindau induces EPO expression substantially only in the kidneys (and to a low degree also in adrenal glands).13Gerl K. Nolan K.A. Karger C. et al.Erythropoietin production by PDGFR-β+ cells.Pflugers Arch. 2016; 468: 1479-1487Crossref PubMed Scopus (31) Google Scholar This organ specificity of EPO expression raised the question about other structural and functional characteristics of renal EPO-producing PDGFR-β+ cells. So far, different tubulointerstitial cell types have been identified: cells specifically expressing CD73,1Bachmann S. Le Hir M. Eckardt K.U. Co-localization of erythropoietin mRNA and ecto-5’-nucleotidase immunoreactivity in peritubular cells of rat renal cortex indicates that fibroblasts produce erythropoietin.J Histochem Cytochem. 1993; 41: 335-341Crossref PubMed Scopus (305) Google Scholar,2Maxwell P.H. Osmond M.K. Pugh C.W. et al.Identification of the renal erythropoietin-producing cells using transgenic mice.Kidney Int. 1993; 44: 1149-1162Abstract Full Text PDF PubMed Scopus (341) Google Scholar the mesenchymal progenitor cell marker Gli1,21Kramann R. Wongboonsin J. Chang-Panesso M. et al.Gli1+ pericyte loss induces capillary rarefaction and proximal tubular injury.J Am Soc Nephrol. 2017; 28: 776-784Crossref PubMed Scopus (90) Google Scholar the extracellular matrix protein tenascin C (TNC),22He W. Xie Q. Wang Y. et al.Generation of a tenascin-C-CreER2 knockin mouse line for conditional DNA recombination in renal medullary interstitial cells.PLoS One. 2013; 8: 79839Crossref PubMed Scopus (15) Google Scholar or markers for contractile pericytes.19Armulik A. Genové G. Betsholtz C. Pericytes: developmental, physiological, and pathological perspectives, problems, and promises.Dev Cell. 2011; 21: 193-215Abstract Full Text Full Text PDF PubMed Scopus (1714) Google Scholar,23Bergers G. Song S. The role of pericytes in blood-vessel formation and maintenance.Neuro Oncol. 2005; 7: 452-464Crossref PubMed Scopus (1074) Google Scholar We were interested to see which of these renal interstitial cell (RIC) populations carry PDGFR-β and if they are capable of producing EPO in response to HIF stabilization. To address these questions we used in situ hybridization to analyze colocalization of various markers with PDGFR-β. Moreover, we studied if cells carrying these markers are able to produce EPO under hypoxemic stimuli, targeted inhibition of HIF degradation, and pathologic conditions. With the production of EPO, renal interstitial PDGFR-β+ cells contribute to the maintenance of normal tissue oxygenation. Considering their important function as an oxygen sensor and for hormone production, the question was raised whether EPO-producing cells coexpress other functional regulatory proteins in states of hypoxia. Localization and quantification of PDGFR-β-expressing cells was performed by RNAscope. On 5-μm transverse mouse kidney sections through the renal papilla, we counted an average of 6500 PDGFR-β+ RICs, which were distributed over all kidney zones (Figure 1). Approximately 50% of PDGFR-β+ RICs were located in the cortex, 40% in the OM (even distribution between outer and inner stripe) and 10% in the inner medulla (IM). In addition to interstitial cells, glomerular mesangial cells expressed PDGFR-β. The localization and quantification of CD73 expression was analyzed by in situ hybridization for CD73 and PDGFR-β mRNA. CD73 expression was located in interstitial cells, glomerular mesangial cells, and proximal tubular cells, most prominently in the S3 segment. Counting the colocalization with PDGFR-β revealed that approximately 60% of all interstitial PDGFR-β+ cells coexpressed CD73. Coexpression occurred along a clear corticomedullary gradient (Figure 1, top row) and was highest in the cortex (>90% of PDGFR-β+ RICs). In the OM, coexpression only occurred in the outer zone (45% of PDGFR-β+ RICs). In the inner zone of the OM and in the IM, CD73/PDGFR-β coexpression was undetectable. Quantification of Gli1+ (glioma-associated antigen 1) cells was investigated by immunohistochemical analysis of cells expressing green fluorescent protein (GFP) after Gli1 promoter–driven Cre recombinase activity (Gli1CreERT2/+ mT/mG mice). This approach was necessary, as endogenous Gli1 expression levels are too low to allow for a robust quantification. Approximately 15% of all PDGFR-β+ RICs costained for Gli1-driven GFP in reporter mice, with the highest coexpression in the outer zone of the OM. In the cortex and in the inner zone of the OM, only few PDGFR-β+ cells coexpressed Gli1-driven GFP. In the IM, Gli1-driven GFP expression was almost undetectable (Figure 1, second row). For the quantification of SMMHC expression, as a marker for contractile pericytes, immunohistochemistry was performed on reporter mice expressing GFP in all cells with Cre activity under control of the SMMHC promoter (SMMHCCreERT2/+ mT/mG mice). SMMHC was detected in smooth muscle cells of renal vessels and in interstitial cells. Approximately 10% of all interstitial PDGFR-β+ cells coexpressed SMMHC in the OM. In the cortex and in the IM, colocalization was less than 5% (Figure 1, third row). Across the kidney, approximately 25% of all PDGFR-β+ RICs coexpressed TNC. Coexpression of TNC by PDGFR-β+ cells occurred along a reverse corticomedullary gradient. Colocalization rate was low in the cortex and in the outer zone of the OM (<5%), but increased from 65% in the inner zone of the OM to 100% in the IM (Figure 1, bottom row). Detection of either CD73+, Gli1-GFP+, SMMHC-GFP+, or TNC+ interstitial cells without coexpression of PDGFR-β was very low (<3%), suggesting that these are all pure subpopulations of interstitial PDGFR-β+ cells. In total, we counted on average 3900 CD73+, 950 Gli1-driven GFP+, 600 SMMHC+, and 1600 TNC+ interstitial cells among 6500 PDGFR-β+ cells on a kidney section. Table 1 gives an overview of the degree of colocalization between PDGFR-β and the respective second markers as well as the colocalization between the additional markers among each other.Table 1An overview of the degree of colocalization between PDGFR-β and the respective second markers as well as the colocalization between the additional markers among each otherCell populationCD73Gli1SMMHCTNCPDGFR-β60151025CD73—1500Gli160—55SMMHC08—5TNC032—EPO in Vhl-KO mice75<5510EPO, erythropoietin; TNC, tenascin C.All numbers are percentages.First row: percentage of different subpopulations regarding interstitial PDGFR-β+ cells. Rows 2–5: percentage of colocalization of different subpopulations among each other. To calculate the percentage for the colocalization between different subpopulations, the respective subpopulation named in the left column was set to 100%. For example, 15% of all CD73+/PDGFR-β+ interstitial cells also expressed Gli1, but there was no colocalization between CD73 and SMMHC or TNC, respectively. Last row: percentage of different subpopulations on PDGFR-β+ EPO-producing cells in Vhl-KO mice. Open table in a new tab EPO, erythropoietin; TNC, tenascin C. All numbers are percentages. First row: percentage of different subpopulations regarding interstitial PDGFR-β+ cells. Rows 2–5: percentage of colocalization of different subpopulations among each other. To calculate the percentage for the colocalization between different subpopulations, the respective subpopulation named in the left column was set to 100%. For example, 15% of all CD73+/PDGFR-β+ interstitial cells also expressed Gli1, but there was no colocalization between CD73 and SMMHC or TNC, respectively. Last row: percentage of different subpopulations on PDGFR-β+ EPO-producing cells in Vhl-KO mice. On kidney sections of wild-type mice, we counted on average 10 EPO mRNA–expressing cells (Figure 2a) that were all positive for PDGFR-β and CD73. Induction of EPO expression in PDGFR-β+ cells was achieved by the inducible deletion of the von Hippel-Lindau (Vhl) gene (PDGFR-βCreERT2/+ Vhlfl/fl, referred to as PDGFRβ-Vhl-KO), which led to the stabilization of hypoxia inducible transcription factors.13Gerl K. Nolan K.A. Karger C. et al.Erythropoietin production by PDGFR-β+ cells.Pflugers Arch. 2016; 468: 1479-1487Crossref PubMed Scopus (31) Google Scholar On kidney sections of PDGFRβ-Vhl-KO mice, we counted on average 4600 interstitial cells expressing EPO mRNA. In line with this, PDGFRβ-Vhl-KO mice were polycythemic and had increased plasma EPO levels (Table 2) after tamoxifen feeding. EPO expression was strongly induced in the cortex and the OM but only to a minor extent in the IM. Approximately 60% of EPO+ RICs were located in the cortex, 35% in the OM, and less than 5% were located in the IM (Figure 2b). In addition, on average 70 mesangial cells per kidney section expressed EPO mRNA (Figure 3).Table 2Hematocrit values, plasma EPO concentrations, and relative EPO mRNA expression levels in control mice, PDGFR-βCreERT2/+ Vhlfl/fl mice, Gli1CreERT2/+ Vhlfl/fl mice, and SMMHCCreERT2/+ Vhlfl/fl mice after feeding a tamoxifen containing chow for 4 weeksGenotypeHematocrit values (%)Plasma EPO concentrations (pg/ml)EPO mRNA expression level (a.u.)Control49.8 ± 0.6137.9 ± 18.81.0 ± 0.2PDGFR-βCreERT2/+ Vhlfl/fl71.9 ± 1.5121,026.0 ± 13,019.0183.1 ± 10.9Gli1CreERT2/+ Vhlfl/fl70.8 ± 0.81335.9 ± 95.620.8 ± 4.5SMMHCCreERT2/+ Vhlfl/fl69.2 ± 1.61278.9 ± 93.09.6 ± 1.5a.u., arbitrary units; EPO, erythropoietin.Values are the mean ± SEM of 8 mice in each group. Open table in a new tab Figure 3Erythropoietin (EPO) mRNA expression in glomeruli, presumably mesangial cells. (a–c) EPO expression was found in 1–3 intraglomerular cells in approximately 35% of glomeruli (encircled) per kidney section of PDGFRβ-Vhl-KO mice. Bars = 50 μm. To optimize viewing of this image, please see the online version of this article at www.kidney-international.org/.View Large Image Figure ViewerDownload Hi-res image Download (PPT) a.u., arbitrary units; EPO, erythropoietin. Values are the mean ± SEM of 8 mice in each group. EPO and CD73 mRNA expression was localized by in situ hybridization on kidney sections of PDGFRβ-Vhl-KO mice. Coexpression mirrored the zonal gradient observed for the colocalization of CD73 with PDGFR-β (Figures 2c and 4a ). It was highest in the cortex, here more than 95% of EPO+ RICs expressed CD73. In the OM, coexpression occurred mostly in the outer zone (50% CD73+ EPO cells). EPO/CD73 coexpression in the inner zone of the OM and in the IM was absent. In summary, over 75% of all EPO+ RICs coexpressed CD73, suggesting an important role of interstitial CD73+ cells for hypoxia-induced EPO expression. To test for the ability of Gli1+ cells to express EPO, we used mice with a conditional deletion of Vhl in Gli1+ cells (Gli1CreERT2/+ Vhlfl/fl). After tamoxifen treatment, mice became polycythemic and had increased plasma EPO concentrations and renal EPO mRNA expression levels (Table 2). No extrarenal EPO mRNA induction was detectable. A total of 90% of EPO-expressing cells were located in the outer stripe of the OM. The cortex contained approximately 10% EPO-expressing cells (Figure 2d). All EPO+ cells coexpressed PDGFR-β (Figure 4b). In situ hybridization for EPO and SMMHC on kidney sections of PDGFRβ-Vhl-KO mice showed that approximately 5% of all EPO+ RICs coexpressed SMMHC. SMMHC coexpression with EPO was mostly found in the OM (Supplementary Figure S1), where approximately 13% of EPO cells coexpressed SMMHC (Figure 4c). However, SMMHC is also strongly expressed by smooth muscle cells in the walls of arteries and arterioles. Because these cells do commonly not express PDGFR-β, hypoxia signaling in these cells was not triggered in PDGFRβ-Vhl-KO mice. To investigate the ability of SMMHC+ cells to produce EPO, we generated SMMHCCreERT2/+ Vhlfl/fl mice. After induction, mice were polycythemic and had increased plasma EPO concentrations and renal EPO mRNA expression levels (Table 2). No extrarenal EPO mRNA induction was found in these mice. Localization of EPO mRNA in these kidneys (Figure 2e) was similar to the colocalization of EPO and SMMHC in kidneys of PDGFRβ-Vhl-KO mice (Supplementary Figure S1). Smooth muscle cells of arteries and arterioles did not express EPO. EPO and TNC mRNA was localized on sections of PDGFRβ-Vhl-KO mice. Coexpression of EPO and TNC was less than 5% in the cortex and in the outer zone of the OM. In the inner stripe of the OM, approximately 30% of EPO+ RICs coexpressed TNC. In the IM, all EPO-expressing cells colocalized with TNC (Figures 2f and 4d). However, there were a large number of TNC+ cells in the IM, which did not express EPO. In total, we counted an average of 3500 CD73+, 200 SMMHC+, 450 TNC+, and 160 Gli1-inducible EPO-expressing cells among the average 4600 EPO+ cells on kidney sections of PDGFRβ-Vhl-KO mice (Table 1). To evaluate the contribution of different PDGFR-β subpopulations to the EPO production in wild-type mice with nongenetic induction of the hypoxia-signaling pathway, we exposed mice to either low oxygen (8% O2 for 3 hours) or carbon monoxide (0.1% CO for 4 hours). RNAscope for EPO mRNA on kidney sections of these mice showed a recruitment of EPO-producing cells predominantly along the corticomedullary border and in the cortex for both conditions (Figure 5a and b). Compared with normoxic conditions the number of EPO+ cells increased 24-fold under low oxygen conditions and approximately 100-fold after CO exposure. The recruitment of EPO-producing cells occurred in a cluster-like fashion. These cells coexpressed CD73 (Figure 6) along with PDGFR-β. Some EPO-expressing cells were positive for Gli1, but none for SMMHC or TNC (Table 3).Figure 6Double fluorescent in situ hybridization for erythropoietin (EPO) (green) and CD73 (red) mRNA on kidney sections of wild-type mice exposed to (a) low oxygen or (b) CO. Under both conditions, EPO-producing cells were mainly located in the cortex in a grouped fashion. The recruited EPO+ cells were mostly part of the CD73+ subpopulation. Circles indicate glomeruli. Bars = 50 μm. To optimize viewing of this image, please see the online version of this article at www.kidney-international.org/.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table 3Average number of EPO-producing cells per transverse kidney section, their localization, and contributing subpopulations of PDGFR-β+ interstitial cellsTreatment/genotypeAverage number of EPO+ cells per sectionLocalization of EPO+ cellsSubpopulationsUntreated wild-type10 ± 3>99% cortex<1% osOMCD738% O2 for 3 h242 ± 3580% cortex20% osOMCD73, Gli10.1% CO for 4 h1096 ± 11378% cortex21% osOM<1% isOMCD73, Gli1Wild-type UUO8 ± 350% cortex8% OM42% IMCD73, TNCPDGFRβ-Vhl-KO UUO1732 ± 11465% cortex34% OM<1% IMCD73, Gli1, SMMHC, TNCEPO, erythropoietin; IM, inner medulla; isOM, inner stripe of outer medulla; OM, outer medulla; osOM, outer stripe of outer medulla; TNC, tenascin C; UUO, unilateral ureteral obstruction. Open table in a new tab EPO, erythropoietin; IM, inner medulla; isOM, inner stripe of outer medulla; OM, outer medulla; osOM, outer stripe of outer medulla; TNC, tenascin C; UUO, unilateral ureteral obstruction. In chronic kidney disease in human patients, EPO production in the kidney declines drastically.15Eschbach J.W. Adamson J.W. Anemia of end-stage renal disease (ESRD).Kidney Int. 1985; 28: 1-5Abstract Full Text PDF PubMed Scopus (287) Google Scholar,16Babitt J.L. Lin H.Y. Mechanisms of anemia in CKD.J Am Soc Nephrol. 2012; 23: 1631-1634Crossref PubMed Scopus (535) Google Scholar Therefore, we induced kidney fibrosis in mice using 10 days’ unilateral ureteral o