|Home | About | Journals | Submit | Contact Us | Français|
Enzymatic cleavage of angiotensinogen by renin represents the critical rate-limiting step in the production of angiotensin-II, but the mechanisms regulating the initial expression of the renin gene remain incomplete. The purpose of this study is to unravel the molecular mechanism controlling renin expression. We identified a subset of nuclear receptors that exhibited an expression pattern similar to renin by reanalyzing a publicly available microarray dataset. Expression of some of these nuclear receptors was similarly regulated as renin in response to physiological cues, which are known to regulate renin. Among these, only estrogen receptor α (ERα) and hepatic nuclear factor α (HNFα) have no known function in regulating renin expression. We determined that ERα is essential for the maintenance of renin expression by transfection of siRNAs targeting Esr1, the gene encoding ERα, in renin-expressing As4.1 cells. We also observed that previously characterized negative regulators of renin expression, Nr2f2 and VDR, exhibited elevated expression in response to ERα inhibition. Therefore, we tested if ERα regulates renin expression through an interaction with Nr2f2 and VDR. Renin expression did not return to baseline when we concurrently suppressed both Esr1 and Nr2f2 or Esr1 and VDR mRNAs, strongly suggesting that Esr1 regulates renin expression independent of Nr2f2 and VDR. ERα directly binds to the hormone response element (HRE) within the renin enhancer region. We conclude that ERα is a previously unknown regulator of renin that directly binds to the renin enhancer HRE sequence and is critical in maintaining renin expression in renin-expressing As4.1 cells.
As the rate-limiting step of the renin-angiotensin system (RAS), expression of the enzyme renin by renal juxtaglomerular (JG) cells is tightly regulated by multiple physiological and molecular mechanisms. The renin-expressing As4.1 cell line has been extensively used to identify genomic regions essential for renin expression.1 Two cis-elements have been found to be critical in regulating renin transcription: a proximal promoter (PPE, +6 to −117 bp) and a classic transcriptional enhancer (−2866 to −2625 bp).2–6 The PPE contains binding sites for the HOXB9/D10-PBX1b-PREP1 protein complex, C-promoter binding factor 1 and Ets-1 transcription factor.4,5 The core of the renin enhancer is evolutionarily conserved and is required to fully activate renin promoter activity in renin-expressing cells.7–9 Two independent studies reported that mice carrying a deletion of the renin enhancer exhibited diminished baseline renin expression.10,11 Other transcription factors that have been shown to regulate renin in vivo have also been reported.7,12,13
Previous studies have identified many of the transcription factors that bind within the enhancer region. The promoter-proximal end of the enhancer (−2680 to 2625) is essential in maintaining the enhancer activity. This region is comprised of four elements: the cAMP response element (CRE), the E-box, the hormone response element (HRE), and the NF-Y binding site.2,14,15 CREB and cAMP response element modulator (CREM) bind to the CRE to positively regulate renin expression.2 Both NF-κB and vitamin D receptor can compete with CREB and CREM for the CRE, resulting in decreased renin expression.16,17 Upstream stimulatory factors 1 and 2 (USF-1 and USF-2) bind to the E-box to positively regulate renin expression.2 The HRE, located at the promoter-proximal end of the E-box, consists of two TGACCT repeats separated by an atypical 10 bp spacer (DR10). Like the CRE and E-box, the HRE is critical for maintaining enhancer function.14,18 The large spacer between the two half-sites suggests that multiple hormone nuclear receptors can potentially associate with that region to regulate renin expression. These characteristics of the HRE drew us to search for transcription factors that bind this element.
Previous studies revealed that retinoic acid receptor (RAR) and retinoid X receptor (RXR) bind to the HRE.14 We also reported that two orphan nuclear receptors, Nr2f6 and Nr2f2 also bind to the HRE, but unlike RAR/RXR, both of these nuclear receptors negatively regulated renin transcription.19 These studies led to the conclusion that both positive and negative regulators bind to the HRE. The most promoter proximal binding site, Ea, matches the canonical binding sequence for nuclear factor-Y (NF-Y) and partially overlaps with the HRE. Binding of NF-Y sterically prevents proteins from binding to the HRE, resulting in down-regulation of renin expression.15 Artificially increasing the spacing between the HRE and the NF-Y binding site restores renin promoter activity.
These previous studies indicate that renin is tightly regulated by a number of transcription factors. However, these factors do not account for the entirety of renin transcriptional regulation. We hypothesize that there are additional nuclear receptors yet to be identified that are regulators of renin expression. We took an unorthodox approach of searching for nuclear receptors, which exhibit 1) a cell-specific pattern of expression similar to renin in the kidney, and 2) exhibited the pattern of regulation similar to renin. This analysis leads us to identify estrogen receptor α (ERα) encoded by the Esr1 gene as a regulator of renin expression.
Details of the experiments using mice, transfection of siRNAs in As4.1 cells, RNA isolation, DNA Affinity Precipitation Assay, Chromatin Immunoprecipitation, Western blotting and analysis of previously published gene expression microarrays are shown in the expanded Methods section of the online only supplement.20 Care and use of mice followed the standards set forth by the National Institutes of Health (NIH) guidelines. All procedures were approved by The University of Iowa Animal Care and Use Committee.
Data was presented as mean ± SEM. The Livak method was used to calculate gene expression fold change.21 Two-way repeated-measures analysis of variance (ANOVA) with Tukey post-hoc analyses were used to calculate the statistics of concurrent transfection of two siRNAs. P≤0.05 was considered significant. Data were analyzed by use of SigmaStat (Systat Software).
To identify a candidate list of nuclear receptors that potentially regulate renin expression, we analyzed a set of microarray data from the GenitoUrinary Development Molecular Anatomy Project (GUDMAP) database.20 We compared the expression pattern of renin and the 48 nuclear receptors across various kidney cell types, including JG cells, glomerular mesangium, cortex vasculature, renal cortex, ureteric tip, capsule mesenchyme, and adult podocytes. Among all the screened kidney cell types, JG cells exhibited the most enriched renin expression followed by the glomerular mesangium. Similarly, expression of aldo-keto reductase family 1, member 7 (Akr1b7), which was reported to be co-expressed with renin in JG cells throughout murine development exhibited the most similar pattern of expression to renin (row 2 in Figure 1).20
Unexpectedly, like renin, some of the 48 nuclear receptors exhibited enriched expression in JG cells and had expression patterns similar to that of renin (Figure 1). Included in this set was Thrb, AR, Nr6a1, Esr1, Rorc, Hnf4g, Nr1i2, Ppara, Ppard, Nr1h4, Hnf4α, Esrra, VDR, Nr3c2, and Nr4a1. Many of the set of 48 nuclear receptors are also expressed in As4.1 cells (see yellow column in Figure 1).
We next assessed which of these receptors were co-regulated like renin. This was performed by pharmacologically altering renin expression in male, C57BL/6 mice. The first cohort of mice underwent captopril treatment for 10 days. Captopril treated mice exhibited increased fluid intake that peaked 7–11 days after treatment (Figure S1). The second cohort of mice was subcutaneously implanted with 50mg DOCA pellet and was given 0.15mol/L NaCl ad libitum in addition to regular chow and tap water for 21 days (DOCA-salt). Kidney weight was elevated in DOCA-salt-treated mice as previous reported (Figure S2). As expected, renin expression was markedly elevated (16.0 ± 1.06 fold) in the kidney cortex of captopril-treated mice and was reduced by 80% (0.21 ± 0.02 fold) in DOCA-salt mice (Figure 2A).
Consistent with a previous report that Akr1b7 mRNA serves as a marker for renin expression in JG cells,20 Akr1b7 expression was positively correlated with renin mRNA level in both experimental cohorts (Figure 2B). Of the nuclear receptors examined, we observed suppressed expression of thyroid hormone receptor β (Thrb), VDR, Nr2f2, hepatocyte nuclear receptor alpha (Hnfα), and Esr1 encoding estrogen receptor α (ERα) in the kidney cortex of DOCA-salt mice (Figure 2C–G). The magnitude of the repression was remarkably similar to that of renin and Akr1b7. Unlike the change induced by DOCA-salt, expression of these nuclear receptors only modestly increased in response to captopril. Nr2f6 expression exhibited an opposite pattern of expression (Figure 2H). There was no change in the expression of 8 other nuclear receptors in response to captopril or DOCA-salt (Figure S3).
Because of their concordance with renin expression, we next examined the possibility of whether these nuclear receptors are regulators of renin expression. Thyroid hormone receptor β is a known positive regulator of renin mRNA.22 Multiple studies have shown that VDR, Nr2f2, and Nr2f6 are negative regulators of renin expression.16,19,23,24 To date, however, there is no report on the role of Hnf4α and ERα in regulating renin expression. We first determined whether these two nuclear receptors are expressed in the renin-expressing As4.1 cells. By RT-qPCR, we were able to detect abundant Esr1 mRNA, but not Hnf4α (Table S1). This was consistent with our prior microarray-based analysis of the gene expression profile of As4.1 cells in which Esr1 displayed a higher expression level compared to Hnf4α (see arrowheads on right margin of Figure 1). Thus Hnf4α was eliminated from further consideration.
To determine if ERα is a regulator of renin expression, we utilized small interfering RNA (siRNA) mediated gene silencing and investigated whether interference with expression of the Esr1 gene would alter expression of endogenous renin mRNA. Esr1 mRNA expression was reproducibly knocked down by 70% (0.31 ± 0.05 vs. NC, Figure 3A). Suppressing expression of Esr1 resulted in a 55% reduction of renin mRNA expression (0.44 ± 0.21 vs. NC, Figure 3B). This data suggests that ERα is required for maximal expression of renin under baseline condition and suggests that ERα is a positive regulator of renin transcription. Surprisingly, treatment of As4.1 with β-estradiol had no effect on the level of Esr1 or renin mRNA after 6, 12 or 24 hours of incubation (Figure S4).
We next asked if ERα protein can bind to the HRE in the renin enhancer because the HRE is a partial inverted sequence of the canonical estrogen receptor response element (ERE). First, nuclear extracts from As4.1 cells were mixed with biotin-labeled HRE probes and subjected to a DNA affinity precipitation assay (DAPA). Western blot was performed to detect eluted proteins. A probe where the critical nucleotides in the HRE were mutated was used as a control.19 ERα protein was detected in proteins eluted from the WT probe but not from the MUT probe, signifying that ERα is able to bind to the HRE sequence with high specificity (Figure 4A). Nr2f2 served as the positive control since we showed it can bind to the HRE in previous studies.19 USF2 which binds to the neighboring E-box, but not the HRE was used as a negative control.2 Second, we tested whether ERα binds to the renin enhancer in chromatin in As4.1 cells. Chromatin immunoprecipitation (ChIP) showed that ERα antibody was able to pull down ~30-fold (29.2 ± 7.2) enrichment of the renin enhancer compared to IgG (Figure 4B). USF2, which binds to the neighboring E-Box sequence within the enhancer (included in this assay), was also enriched (9.0 ± 5.6). Importantly, there was no binding of USF2 nor ERα to a region 10 kb upstream of the renin enhancer, providing confidence in the selectivity of the ChIP assay. To further validate the specificity of ERα enrichment within the enhancer and validate the siRNA-mediated knock down of ERα at the protein level, we performed ChIP assay after siRNA-mediated Esr1 knock down in As4.1 cells. In the NC siRNA chromatin sample, USF2 antibody was able to pull-down approximately a 20-fold (21.5 ± 6.2) enrichment of the renin enhancer sequence; and ERα antibody was able to pull down approximately 60-fold (59.7 ± 19.7 of IgG) of enrichment of the renin enhancer. However, the ERα ChIP signal was significantly reduced after siEsr1 (Figure 4C).
Since renin expression is tightly regulated by both positive and negative modulators, we considered the possibility that the significant reduction of renin followed by ERα suppression might be attributable to the altered expression of other transcription factors rather than only ERα per se. Thus, we assessed whether expression of other nuclear receptors were altered upon siEsr1 transfection. Indeed siEsr1 augmented expression of Nr2f6, VDR, and farnesoid X receptor (Nr1h4) in siEsr1 transfected cells (Figure 3C–E). There was no consistent change in the expression of 10 other nuclear receptors in response to both siRNAs targeting Esr1 (Figure S5).
Nr2f6 and VDR are known negative regulators of renin, whereas there is no published information on the role of Nr1h4 in regulating renin expression. This finding led us to test whether the suppression of renin expression, caused by the loss of ERα, is due to the elevated expression of the negative regulators. To accomplish this we simultaneously transfected siRNAs targeting both Esr1 and Nr2f6 into As4.1 cells, and asked if suppressing Nr2f6 restored renin expression. In this experiment, the induction of Nr2f6 mRNA by siESR1 was only modest and not statistically significant (Figure 5A). siRNA targeting Nr2f6 effectively knocked down Nr2f6 expression but did not significantly decrease Esr1 expression. As before, siRNA targeting Esr1 alone decreased renin mRNA. As we reported previously, there was a modest induction of renin by Nr2f6 knockdown alone.19 Contrary to our hypothesis, siRNA mediated knockdown of both ERα and Nr2f6 preserved a decrease in renin expression. In fact, combined suppression of both ERα and Nr2f6 appeared to synergistically suppress renin expression (Figure 5A). Similar results were observed when we knocked down ERα and VDR simultaneously (Figure 5B). Thus, the reduction of renin expression mediated by the loss of ERα does not depend on the induction of Nr2f6 and VDR.
The main findings from the current study are that 1) contrary to expectation, a number of nuclear receptor transcription factors exhibit a pattern of gene expression in the developing and adult kidney that mirrors the expression of renin, 2) like renin, expression of 5 nuclear receptors exhibit a marked decrease in expression in response to DOCA-salt, 3) unlike renin, none of the same nuclear receptors exhibited an increase in expression in response to angiotensin converting enzyme inhibition, and 4) ERα, which binds to the HRE within the renin enhancer is required for baseline expression of renin and acts independent of VDR and Nr2f2 and does not require exogenous estrogen.
Renin expression in JG cells is precisely regulated by both extracellular and intracellular stimuli through the actions of transcription factors, but the entirety of the mechanisms underlying this transcriptional regulation has not been fully elucidated. Detailed molecular studies have clearly defined two regions of the renin regulatory region that are required for maximal expression of the renin gene, a proximal promoter and a distal enhancer.1,25 Deleting the distal enhancer in vivo diminishes the level of renin expression and blunts the response to physiological cues regulating expression of the gene.11,26 The enhancer consists of a complex series of linked regulatory elements which both stimulate and inhibit expression of renin, but how these elements act in concert to regulate renin expression in response to physiological cues remains unclear. Major elements of the enhancer (distal to proximal) include an E-box, cAMP-response element (CRE), HRE, and a binding site of NF-Y.2,14,15 We originally showed that retinoic receptor α (RARα) and retinoid X receptor (RXR) bind to the HRE and regulate renin expression in response to retinoids.14 However, electrophoretic mobility shift assays suggested that many proteins may bind to the sequence. In an effort to identify these proteins we later performed yeast one hybrid and identified Nr2f2 and Nr2f6 as HRE binding proteins with the capacity to antagonize enhancer activity.19,24 Attempts to identify other HRE-binding protein which may help explain the mechanisms of renin regulation have not been successful.
The publication of a comprehensive comparison of expression of renin to other genes across the genome in the kidney during development, in adults, and in response to physiological perturbations, provided a novel opportunity to search for other transcription factors that are co-expressed, or enriched, in renin expressing cells.20 Since several members of the nuclear hormone receptor family bind to the HRE to regulate renin, we focused on that family of transcription factors, and adopted a working yet unorthodox hypothesis that nuclear receptors co-expressed with renin, and co-regulated like renin, may regulate renin. ERα was selected from among a group of nuclear receptors which appeared to exhibit enriched expression in JG cells and was highly expressed in As4.1 cells. None of these genes were exclusively expressed in JG cells and each exhibited a varying cell-selectivity of expression. Expression of Esr1 exhibited only a modest increase in expression in the kidney in response to captopril despite exhibiting increased expression in adult JG cells treated with captopril compared with untreated JG cells. It remains unclear if the relative increase in Esr1 expression in response to captopril reflects an increase in the level of expression per cell or reflects, like renin, and increase in the number of cells expressing Esr1. Interestingly, like renin, expression of Esr1 (and several other nuclear hormone receptors) decreased approximately 10-fold in response to DOCA-salt. DOCA-salt was previously reported to induce the expression of over 2000 genes but decrease expression of only 50 genes in kidney from Wistar rats.27 Whether renin and Esr1 were among those was not reported.
siRNAs targeting Esr1 blunted baseline renin expression. This along with data showing that ERα has the capacity to bind to the HRE in the renin enhancer in vitro, and binds to the HRE under baseline conditions in chromatin in vivo would generally support a conclusion that ERα is among the many transcriptional regulators of renin. However, this regulation could be indirect, e.g. through another transcription factor. Indeed, silencing ERα also simultaneously increased expression of two known negative regulators of renin expression, VDR and Nr2f6.19,28 This led us to assess if the decrease in renin mRNA by siEsr1 was mediated by an increase in Nr2f2 or VDR. However, this mechanism was effectively ruled out because simultaneous silencing of ERα and Nr2f2, or ERα and VDR, did not restore expression of renin to baseline levels. The only other report we could find in the literature directly linking ERα to renin gene expression was a study of ERα-receptor deletion in the subfornical organ of female mice which results in a 1.4-fold increase in renin expression in the lamina terminalis.29 This suggests that regulation of renin by ERα may be cell-specific, its expression blunted by ERα in the brain, and stimulated by ERα in juxtaglomerular cells.
Given our findings, it is interesting that estrogen related receptor α (ERRα), an orphan member of the nuclear receptor superfamily was previously reported to regulate renin expression.30 However, despite previous data showing that inhibition of ERRα caused a 3-fold increase in renin expression in As4.1 cells, our studies revealed that inhibition of ERRα with two different siRNAs (which decreased expression of ERRα to 0.29±0.29 and 0.57±0.28, n=5, respectively) had no effect on expression of renin in As4.1 cells (1.06±0.41 and 1.03±0.35, respectively). Further studies would be required to understand these discrepant experimental results.
The identification of ERα as a potential regulator of renin is interesting in light of the effect of estrogen on RAS activity and the long held hypothesis that estrogen deficiency in menopause may play a role in increasing RAS activity.31 Therefore, it was notable that 17-β estradiol treatment of As4.1 cells had no effect on expression of Esr1 or renin mRNA. Similarly, there was no change in expression of renin mRNA in primary chorionic cells exposed to exogenous estrogen although the importance of ERα itself was not experimentally tested.32 This suggests that unless an endogenous ERα ligand is synthesized in As4.1 cells, its activity on renin expression does not require ligand binding. Although ERα-bound ligand is the classical model for transcriptional activation in response to endogenous or exogenous estrogen, estrogen-independent activation of ERα transcriptional activity was reported to occur in breast cancer cells with increased levels of ERα.33 It was hypothesized that the activation of ERα function in the absence of ligand (termed “concentration-inducible”) is explained by a conformational change in ERα, when the concentration of ERα is high, that looks similar to the conformation of active ERα in the presence of estrogen. Of course, it is unclear if the level of ERα in As4.1 cells, a fully transformed cell line derived from a renal tumor, reaches that level.
In addition to the classical ligand, many nuclear receptors including ERα can be activated by a variety of extracellular signals in a ligand-independent manner. For example, ERα can be activated through the cAMP/Protein kinase A (PKA) pathway. cAMP signaling is able to stimulate ERα transcriptional activity via PKA-mediated phosphorylation of coactivator-associated arginine methyltransferase 1 (CARM1), which then associates with ERα, resulting in the recruitment of ERα co-activators and transcriptional activation of ERα target genes.34 This is interesting when one considers that there is a CRE directly adjacent to the HRE in the renin enhancer, which when mutated causes a complete loss of enhancer activity, coupled with data suggesting that PKA may be constitutively activated in As4.1 cells.2 Thus it remains possible that constitutive activity of PKA in these cells signals constitutive activity of ERα to maintain renin expression.
Females usually exhibit a lower blood pressure compared to their age-matched males, at least until menopause at which time blood pressure rises. The link between estrogen levels and RAS activity has been studied extensively.35 Indeed, it is well established that plasma renin activity increases in women during the luteal phase of the menstrual cycle, a time when plasma estrogen levels are high.36 Since expression of hepatic and renal angiotensinogen is induced by 17β-estradiol, it comes as no surprise that plasma Ang-II levels also increase with increasing circulating estrogen.37 It is therefore notable and surprising that despite the known relationship between estrogen and plasma renin activity and Ang-II, and the importance of the post-menopausal increase in blood pressure in women, that there is surprisingly little information on the regulation of renin gene expression in response to estrogen. Our identification of ERα as a potential regulator of renin gene expression, at least in As4.1 cells, should spark new studies directly assessing the importance of this transcriptional pathway, the relative importance of ligand-dependent and ligand-independent ERα-mediated renin transcription, and whether mechanisms of renin expression through ERα differ in before and after menopause when the levels of estrogen markedly differ. These would provide a more comprehensive assessment to determine if our identification of ERα as a regulator of renin expression plays a role in the cardiovascular protection offered by estrogen prior to menopause and the increased cardiovascular risk after menopause.
We thank Debbie Davis for providing instruction for the Captopril and DOCA-Salt treatment, and to Xuebo Liu for helpful advice.
Sources of Funding: This work was supported through research grants from the NIH to CDS (HL084207, HL048058, HL125603, and HL062984), MLSSL (DK091330 and DK096373), and RAG (HL066242 and DK096373). The authors also gratefully acknowledge the generous research support of the Roy J. Carver Trust.
Conflict of Interest/Disclosure: None