The pendrin anion exchanger gene is transcriptionally regulated by uroguanylin: a novel enterorenal link

Abstract

The pendrin/SLC26A4 Cl⁻/HCO₃⁻ exchanger, encoded by the PDS gene, is expressed in cortical collecting duct (CCD) non-A intercalated cells. Pendrin is essential for CCD bicarbonate secretion and is also involved in NaCl balance and blood pressure regulation. The intestinal peptide uroguanylin (UGN) is produced in response to oral salt load and can function as an “intestinal natriuretic hormone.” We aimed to investigate whether UGN modulates pendrin activity and to explore the molecular mechanisms responsible for this modulation. Injection of UGN into mice resulted in decreased pendrin mRNA and protein expression in the kidney. UGN decreased endogenous pendrin mRNA levels in HEK293 cells. A 4.2-kb human PDS (hPDS) promoter sequence and consecutive 5′ deletion products were cloned into luciferase reporter vectors and transiently transfected into HEK293 cells. Exposure of transfected cells to UGN decreased hPDS promoter activity. This UGN-induced effect on the hPDS promoter occurred within a 52-bp region encompassing a single heat shock element (HSE). The effect of UGN on the promoter was abolished when the HSE located between nt −1119 and −1115 was absent or was mutated. Furthermore, treatment of HEK293 cells with heat shock factor 1 (HSF1) small interfering RNA (siRNA) reversed the UGN-induced decrease in endogenous PDS mRNA level. In conclusion, pendrin-mediated Cl⁻/HCO₃⁻ exchange in the renal tubule may be regulated transcriptionally by the peptide hormone UGN. UGN exerts its inhibitory activity on the hPDS promoter likely via HSF1 action at a defined HSE site. These data define a novel signaling pathway involved in the enterorenal axis controlling electrolyte and water homeostasis.

everett et al. (14) identified the gene PDS (SLC26A4) as the locus of mutations causing Pendred syndrome, an autosomal recessive disorder characterized by early-onset sensorineural hearing loss with enlargement of the vestibular aqueduct and incompletely penetrant goiter (4, 7). The ∼5-kb PDS transcript is most highly expressed in the thyroid, kidney, and inner ear (14, 44). The PDS transcript encodes the 780-amino acid pendrin protein, a member of the SLC26 anion transporter protein superfamily (12, 35). Pendrin functions as an electroneutral plasmalemmal anion exchanger. It functions in the renal cortical collecting duct (CCD) in Cl⁻/HCO₃⁻ exchange (44, 55) and Cl⁻/I⁻ exchange (23, 46). Pendrin is believed also to mediate Cl⁻/HCO₃⁻ exchange in the inner ear (36, 59) and, possibly also I⁻ transport in the thyroid (6, 60).

In the kidney, pendrin protein is located at or near the apical membrane of type B- and non-A non-B intercalated cells (IC) of the CCD (44, 45). Pendrin contributes to acid-base balance by secreting HCO₃⁻ into the tubular lumen in exchange for luminal Cl⁻ (44), and to regulation of blood pressure and systemic fluid balance via that same Cl⁻ reabsorption (43, 53, 55). Our recent deletion analysis of the 5′-flanking region of the human PDS (hPDS) gene defined both positive and negative regulatory elements in the hPDS promoter and proposed a major role for these control elements in the regulated expression of this gene in renal epithelial cells (1).

Several factors have been shown to modulate pendrin activity. Systemic HCO₃⁻ loading increases and acid loading decreases pendrin protein expression in the apical membrane of non-A IC (18, 39, 58). We have shown that pendrin is transcriptionally regulated by systemic pH and aldosterone in renal epithelial cells (1). Systemic and tubule lumen Cl⁻ concentrations regulate pendrin protein levels and activity (43, 53, 56). Recent studies indicate that pendrin-mediated Cl⁻ transport may be important in the pathogenesis of mineralocorticoid- and/or angiotensin II-induced hypertension (38, 55). Together with our demonstration that changes in extracellular Cl⁻ concentration lead to transcriptional regulation of pendrin activity (13), these data reveal pendrin's importance to transcellular Cl⁻ reabsorption in the CCD and suggest an important role for pendrin in electrolyte homeostasis and blood pressure regulation. However, the molecular mechanisms controlling pendrin activity in renal epithelial cells remain unknown.

Guanylin (GN) and uroguanylin (UGN) are low-molecular-weight peptide hormones produced mainly in the intestinal mucosa in response to oral salt load. GN and UGN resemble in structure and activity the secretory diarrhea-causing heat-stable enterotoxin (STa) of Escherichia coli (16) and induce secretion of electrolytes and water in both the intestine and kidney (24, 47). GN and UGN are coexpressed along the intestinal tract with guanylyl cyclase C (GC-C), the principal GN receptor (25). Salt ingestion induces secretion of GN and UGN into the intestinal lumen, and both hormones activate GC-C-mediated intestinal secretion of electrolytes and water (47).

GN and UGN play a major role in the regulatory link between the intestine and the kidney by increasing urinary NaCl and water excretion in response to dietary NaCl intake (but not intravenous administration), thereby serving as “intestinal natriuretic factors” (15, 24, 47). The mode of action and signaling pathways for UGN differ according to nephron segment. UGN promotes natriuresis in the proximal tubule via guanylate cyclase C (GC-C)-mediated cGMP-dependent and G-protein-dependent modulation of Na⁺/H⁺ exchange, K⁺ channels, and Na⁺-K⁺-ATPase. In contrast, the natriuretic effect of UGN on CCD principal cells involves PLA₂-mediated inhibition of K⁺ channels (ROMK) (8, 17, 41, 47, 50). However, despite accumulating data on renal expression and function of GN peptides, the cellular and molecular pathways mediating UGN action in the CCD remain poorly understood. We postulated that UGN might act on CCD non-A IC by downregulation of pendrin-mediated Cl⁻ reabsorption.

In this study, we demonstrate that UGN-injected mice show a decrease in both pendrin protein and mRNA in the kidney compared with control mice. In addition, we demonstrate a UGN-dependent decrease in endogenous pendrin mRNA levels in human embryonic kidney (HEK293) cells. Using human PDS (hPDS) 5′-flanking DNA sequences, we show that UGN decreases hPDS promoter activity in transfected HEK293 cells. This inhibitory effect of UGN on the hPDS promoter likely occurs via a heat shock element (HSE), located at −1119 to −1115 on the promoter. These findings provide the first direct evidence that UGN may regulate pendrin activity and that this regulation occurs at the transcriptional level.

MATERIALS AND METHODS

Animal studies.

All animal studies were conducted under a protocol approved by the Institutional Animal Care and Use Committee of the Faculty of Medicine, Technion.

Eight-week-old male ICR mice were kept in a light- and temperature-controlled room and were fed ad libitum with standard laboratory mouse chow. At time 0, mice received injections into the tail vein of UGN (37.6 μg/kg in isotonic saline) or of vehicle only (isotonic saline). This dose of UGN significantly increases urine volume and urine Na excretion in mice (8). After 2 or 24 h, mice were anesthetized by intraperitoneal administration of Sagatal (pentobarbitone sodium; Rhone Merieux, Essex, UK) and euthanized. Kidneys were harvested for determination of pendrin mRNA and protein levels.

Immunofluorescence microscopy of mouse kidney.

Cryosections (5 μm) of freshly harvested kidneys were fixed 30 min at 4°C in 100% ethanol followed by sequential treatments with 100% acetone for 1 h at room temperature (RT), 0.2% Triton X-100 in PBS (150 mM NaCl, 15 mM sodium phosphate, pH 7.4) for 10 min at RT, and three 10-min washes with PBS. Sections were blocked in 1% BSA in PBS at RT for 15 min and exposed to rabbit anti-pendrin antibody (kindly provided by Dr. Dominique Eladari, Université René Descartes, Paris, France) (43) diluted 1:100 in PBS containing 1% BSA overnight at 4°C, followed by three 10-min washes with PBS. Sections were then exposed to FITC-conjugated donkey anti-rabbit IgG (Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:50 in PBS containing 1% BSA for 1 h at RT in a dark chamber, followed by three 10-min washes with PBS. Stained sections were mounted in UltraCruz Mounting Medium (Santa Cruz Biotechnology) containing 4,6-diamidino 2-phenylindole dihydrochloride (DAPI) for nuclear staining. Immunostained slides were visualized at ×20 magnification with a Zeiss Axiovert 135 epifluorescence photomicroscope. Images were acquired with an attached CCD camera (Roper Scientific/Princeton Instruments, Trenton, NJ) controlled by Image-Pro Version 5 (MediaCybernetics, Bethesda, MD). Relative pendrin polypeptide abundance in the kidney was quantitated as previously described, with modifications (54). In each of four independent experiments, five kidney sections from UGN-treated mice and five sections from UGN-untreated mice were scored independently by two investigators blinded to the treatment status of the mice. Fifteen to twenty randomly selected visual fields from each section were scored for the number of pendrin-positive cells per visual field. Results were normalized to the value for UGN-untreated mice in each of the four independent experiments. Antibody specificity was demonstrated by the absence of staining when the primary antibody was omitted.

Immunoblot analysis.

Harvested kidneys were homogenized with a sonicator at 4°C in RIPA buffer (25 mM Tris·HCl, pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS) containing the Complete protease inhibitors (Roche, Mannheim, Germany). Samples were incubated on ice for 30 min and centrifuged at 10,000 g for 10 min at 4°C. Protein concentration of the supernatant was determined by Bradford assay (Sigma-Aldrich, St. Louis, MO). Proteins were separated on 10% SDS polyacrylamide gels (Pierce, Thermo Scientific, Rockford, IL), transferred to nitrocellulose membranes (Schleicher and Schuell, Dassel, Germany), and probed with rabbit anti-pendrin antibody (9) (kindly provided by Dr. A. J. Griffith, National Institutes of Health/National Institute of Deafness and Other Communication Disorders) with 1:200 dilution or mouse monoclonal anti-actin antibody (42 kDa, Sigma) overnight at 4°C. After washing, blots were incubated with the secondary antibodies (donkey anti-rabbit with 1:50,000 dilution and donkey anti-mouse with 1:20,000 dilution) IgG conjugated with horseradish peroxidase (Jackson ImmunoResearch laboratories, West Grove, PA) for 1 h at RT. Blots were assayed with ECL using Pierce ECL Western Blotting Substrate (Pierce, Thermo Scientific) to visualize bound antibodies. For semiquantitative estimate of pendrin protein, densitometric analysis was carried out by Totalab densitometry software (Non-linear Dynamics, Newcastle-upon-Tyne, UK), and pendrin values were normalized to actin.

Quantitative real-time PCR analysis of mouse kidney.

Harvested kidneys were homogenized with a Polytron PT3000 homogenizer in TRI Reagent (Sigma-Aldrich) at 20,000 rpm. Total kidney RNA was prepared using TRI Reagent (Sigma-Aldrich) followed by phenol/chloroform extraction and isopropanol precipitation. Reverse transcription of 2 μg of total mouse kidney RNA was carried out using Moloney murine leukemia virus reverse transcriptase (MMLV-RT) with random hexamer primers (Promega, Madison, WI). Quantitative PCR experiments were performed using ABsolute Blue QPCR Rox MIX (Thermo Scientific, Waltham, MA) with Assay on Demand primers (Applied Biosystems, Foster City, CA) for gene expression analysis of mouse Pds (Mm 00442308), and mouse transient receptor potential cation channel, subfamily M, member 6 (Trpm6; Mm 00463112_m1) serving as the control, with mouse TATA box binding protein (Tbp; Mm 00446973) serving as a housekeeping gene for normalization. All measurements were performed in duplicate in an ABI Prism 7000 Cycler (Applied Biosystems).

Cell lines.

Human embryonic kidney (HEK293) cells were grown and maintained in DMEM/F-12 supplemented with 10% fetal calf serum, 2 mM glutamine, 50 IU/ml penicillin, and 50 μg/ml streptomycin (Biological Industries, Beit Ha'Emek, Israel) at 37°C in a humidified atmosphere of 95% air-5% CO₂.

Quantitative real-time PCR analysis of cell lines.

Confluent HEK293 cells were plated in six-well dishes (0.5 × 10⁶ cells/well). After 24 h in serum-containing regular media, cells were washed with PBS and placed in media supplemented with 1 μM UGN or media without UGN as the control containing 0.2% BSA. Following 24-h exposure to these media, cells were washed with PBS and total RNA was isolated, reverse-transcribed, and analyzed as with mouse kidney (see above). Assay on Demand primers (Applied Biosystems) for quantitative real-time PCR analysis of human PDS (Hs 00166504_m1), human TRPM6 (Hs 00214306_m1), and human TBP (Hs 99999910_m1) were utilized for gene expression analysis.

RT-PCR analysis of cell lines.

Total RNA isolated with Tri-reagent (Sigma-Aldrich) from HEK293 cells was reverse-transcribed using MMLV-RT and random hexamers (Promega). The resultant cDNA was amplified by PCR using the Stratagene Herculase II Fusion DNA Polymerase (Agilent Technologies, Cedar Creek, TX) with the following oligonucleotide primers: heat shock factor (HSF) 1 forward (gataacctgcagaccatgctgag), HSF1 reverse (ctgtgtagtgcaccagctgcttc), and HSF2 forward (gttactgatgataatgcagatg), HSF2 reverse (ctcatcttcattgtcatcttc). cDNA products were separated on 1% agarose gels.

Generation of promoter/reporter constructs.

Plasmids containing 5′- flanking fragments of the hPDS promoter cloned upstream of a luciferase reporter gene in the vector pGL3-basic vector were previously characterized in our laboratory (1). As illustrated in Fig. 1, the 5′-ends of the fragments were positioned at nt −4185 (pL4), −2060 (pL2), −1433 (pL1.4), −1342 (pL1.3), −1253 (pL1.2), −1140 (pL1.1), and −1044 (pL1), respectively. All fragments extended up to but did not include the initiator ATG.

Fig. 1.

Human pendrin gene (hPDS) promoter constructs in pGL3-basic vectors. 5′-Truncations of progressively increasing length resulting in progressively smaller promoter fragments were PCR amplified from human genomic DNA, then cloned upstream of the luciferase reporter gene in pGL3-basic (1; see text for details).

These luciferase reporter constructs were purified using a Genopure plasmid maxi kit according to the manufacturer's protocol (Roche). Empty pGL3-basic vector served as the negative control. Promoter activity was estimated from the ratio of luciferase activity in cells expressing pGL3-basic promoter constructs to that of cell expressing promoterless pGL3-basic vector, studied in parallel for each experiment.

For studies of reporter gene activity driven by distal hPDS promoter fragments, a DNA region from position −1153 to −1044 was PCR generated with human genomic DNA and the primers F1/R1 (Table 1). In addition, a region 5′-truncated by 50 bp and positioned between −1101 and −1044 was produced by annealing F2/R2 primers (Table 1). The two DNA fragments were cloned into the pGL3-promoter vector and were termed pJ110pr and pJpr58, respectively. In these experiments, promoter activity was determined from the ratio of hPDS promoter-driven luciferase activity in the pGL3-promoter vector to that of the empty pGL3-promoter vector.

Table 1.

PDS primers used in PCR for generation of distal hPDS promoter constructs and of mutated constructs

F1	GACCACGGACCTCTTCCTC
R1	CTCACTCATCCCGTTACTAATC
F2	CAGGGCCTGGCCCGCCCCTGACCTCGCAACCCTTGAGATTAGTAACGGGATGAGTGAG
R2	CTCACTCATCCCGTTACTAATCTCAAGGGTTGCGAGGTCAGGGGCGGGCCAGGCCCTG
F3	CTGCTCTACTTTAAGGAGTCC
R3	GGACTCCTTAAAGTAGAGCAG

PDS, Pendrin gene; F, forward; R, reverse.

In all transfection experiments, plasmid pCH110 (Pharmacia, Uppsala, Sweden), containing a LacZ gene driven by the cytomegalovirus (CMV) promoter, was used to normalize for transfection efficiency.

Generation of mutated constructs.

A site-specific mutation in a HSF binding site [predicted by computer analysis (21) of the 4.2-kb 5′-DNA fragment to reside between nt −1119 and −1115 upstream of the hPDS initiator ATG codon] was introduced into the pL1.4 plasmid (Fig. 1) and the 110-bp insert-containing plasmid pJpr110. This was performed using PCR with Stratagene Herculase II Fusion DNA Polymerase (Agilent Technologies) with mutagenic primers F3/R3 (Table 1). The parental DNA template was digested with DpnI overnight, and the mutated plasmid was transformed into competent E. coli strain DH5α. Subsequently, small- and large-scale plasmid preparations were performed using a Genopure Plasmid Maxi kit (Roche). The presence of the mutation was verified by DNA sequencing.

Transient transfections and reporter gene assays.

HEK293 cells were plated into quadruplicate 24-well dishes (5 × 10⁴ cells/well) in serum-containing media. Cotransfections were performed using Fugene 6 (1.2 μl/well, Roche) with 0.3 μg of reporter plasmid and 0.3 μg of pCH110, and cells were incubated at 37°C for 48 h. For enzymatic assays, cells were washed with PBS and lysed by incubating in 225 μl/well of mammalian protein extraction reagent (M-Per, Pierce, Cheshire, UK) for 20 min at RT. Cell lysates were centrifuged, and the supernatant was aliquoted (50 μl/well) into 96-well plates. Fifty microliters of Luciferase Assay Reagent (Promega) were automatically added, and the light intensity of the reaction was immediately read in a Clarity Luminescence Microplate Reader (Biotek, Winoosky, VT) for a period of 10 s. Luciferase activity was normalized to β-galactosidase activity, which was measured in identical cell lysates as follows: 160 μl of ortho-nitrophenyl-β-d-galactopyranoside (ONPG) substrate (Sigma-Aldrich) were added to 50 μl of cell lysate in each well of a 96-well plate. The reaction mixture was incubated for 30 min at 37°C, or until yellow color developed. β-Galactosidase measurements were performed by a spectrophotometer with a 405-nm filter. Measurements of luciferase and β-galactosidase were carried out in duplicate.

In all experiments, 48 h after transfection, the medium was replaced with fresh medium containing 0.2% BSA instead of 10% fetal calf serum to prevent any artifacts (hormones, growth factors, etc.) arising from the serum. Experimental media contained 1 μM UGN (Peptides International, Louisville, KY). Control experiments were carried out with regular medium. Cells were exposed to experimental media for 24 h and analyzed as outlined above.

Transfection of cell lines with small interfering RNA.

Reverse transfection of HEK293 cells with small interfering (siRNA) was carried out with siPort NeoFX (Ambion, Austin, TX) as per the manufacturer's protocol. Briefly, siPORT NeoFX (5 μl/well) was diluted into DMEM (100 μl/well) for each six-well plate used and incubated for 10 min at RT. Human HSF 1 (hHSF1) Silencer Select siRNA (4392420;ID:s6950, Ambion) and Silencer Negative Control siRNA (AM4611, Ambion) were diluted into 100 μl of DMEM to a final concentration of 5 nM. The diluted siPORT NeoFX and the diluted siRNA were combined and incubated for 10 min at RT. The transfection complexes were then dispensed into the empty six-well plates (200 μl/well). One hour before transfection, healthy growing, adherent cells were trypsinized and resuspended in normal growth medium at 1 × 10⁵ cells/ml. Then, 2.3 ml containing 1 × 10⁵ cells/ml HEK293 cells were overlaid into each well of the six-well plates. Forty-eight hours after transfection, the medium was replaced with fresh medium containing 0.2% BSA and 1 μM UGN. Control experiments were carried out with medium containing 0.2% BSA. In addition, the fresh media contained the transfection mixes of siPORT NeoFX and appropriate siRNA. Following 24-h exposure to these media, cells were washed with PBS and total RNA was isolated using a GeneJET RNA Purification kit (Fermentas, Vilnius, Lithuania) and reversed transcribed as described above for quantitative real-time PCR analysis of mouse kidneys. Quantitative PCR experiments were performed using qPCR GreenMaster with ROX (Larova, Teltow, Germany) with the following oligonucleotides: HSF1 forward (gataacctgcagaccatgctgag) and HSF1 reverse (ctgtgtagtgcaccagctgcttc) for gene expression analysis of HSF1, and actin forward (tgacggggtcacccacactgtgcccatcta) and actin reverse (gcattgcggtggacgatggaggg) for analysis of β-actin gene expression (control). Analysis of hPDS and hTBP were performed as described above for quantitative real-time PCR analysis of cell lines.

Data analysis.

Data comparisons were made with Student's unpaired t-test for grouped independent data. P < 0.05 was considered significant.

RESULTS

Effect of UGN on pendrin protein abundance in mouse kidneys.

To explore whether UGN modulates pendrin activity, we first examined the effect of this peptide hormone on pendrin protein levels in mouse kidneys. For this purpose, cryosections of kidneys harvested from mice 24 h after UGN injection were probed with pendrin antibody (see materials and methods). As shown in Fig. 2, A and B, a marked decrease (30%) in total cortical pendrin polypeptide abundance was observed in kidneys of UGN-injected mice compared with control mice. These results were further corroborated by Western blot analysis (Fig. 2, C and D), which showed a similar magnitude of decrease in pendrin protein abundance in kidneys of UGN-treated mice.

Fig. 2.

Effect of uroguanylin (UGN) on pendrin protein abundance in mouse kidney. A: exemplar immunofluorescence images of pendrin staining in kidneys of control and UGN-injected mice obtained 24 h after injection. Cryosections of mouse kidneys were probed with pendrin antibody and FITC-coupled secondary antibody (see materials and methods). B: relative pendrin protein abundance in the kidney was determined by analysis of multiple immunofluorescence images (see materials and methods). UGN treatment reduced pendrin protein abundance by 30% relative to control mice. Values are mean ± SE of 4 independent experiments. No secondary antibody staining was observed in the absence of primary antibody (data not shown), indicating antibody specificity. **P < 0.01. C: immunoblot analysis of total kidney protein (10 μg/lane) from 3 control mice and 3 UGN-treated mice. Probing the blot with anti-pendrin antibody displayed a band comigrating with the 110-kDa molecular mass marker. Antibody against the 42 kDa β-actin polypeptide was used to control for protein loading. Reduced pendrin protein expression is evident in kidneys of UGN-treated mice relative to control mice. D: relative pendrin abundance in blots was determined by densitometric analysis; n = 3 animals/group. *P < 0.05.

Effect of UGN on Pds mRNA expression in mouse kidneys.

To explore whether UGN influences transcription of the pendrin gene, we examined the effect of this modulator on pendrin mRNA levels in mouse kidneys. The effect of UGN on mRNA levels of the distal tubular Mg²⁺ channel Trpm6 served as a control. As shown in Fig. 3, A and B, kidneys of UGN-injected animals displayed a 40% reduction in pendrin mRNA levels at 24 h after injection compared with control, noninjected mice. The apparent UGN-induced reduction at 2 h was not statistically significant. No change in Trpm6 mRNA level was evident at either time point.

Fig. 3.

Effect of UGN on Pds/PDS mRNA levels in mouse kidney (A and B) and in HEK293 renal epithelial cells (C). A and B: Pds and mouse transient receptor potential cation channel, subfamily M, member 6 (Trpm6) mRNA levels were measured by real-time PCR in kidneys of mice 2 (A) and 24 h (B) after UGN injection. Results represent relative levels of Pds and Trpm6 mRNAs in mice injected with UGN relative to control mice injected with isotonic saline. All values were normalized to Tbp mRNA levels (see materials and methods). Values are means ± SE of 3 (2 h) and 5 (24 h) experiments, respectively, each consisting of 4 mice. UGN treatment reduced Pds mRNA levels after 2 h by 22% (not significant; A) and after 24 h by 40% (P < 0.01; B) relative to unexposed control mice, whereas Trpm6 mRNA levels remained unchanged. **P < 0.01. C: cells were exposed for 24 h to medium without or with UGN (1 μM), after which total RNA was extracted for real-time PCR of PDS and TRPM6 mRNAs. Results represent relative change in TBP-normalized levels of PDS mRNA and TRPM6 mRNA in UGN-exposed cells relative to unexposed control cells (control UGN; see materials and methods). Values are means ± SE of 4–5 independent experiments, each performed in duplicate. UGN reduced PDS mRNA level by 35%, whereas TRPM6 mRNA level remained unchanged. **P < 0.01.

Effect of UGN on endogenous PDS mRNA level in HEK293 cells.

To further explore the influence of UGN on transcription of the pendrin gene, we examined the effect of this peptide on endogenous PDS mRNA levels in HEK293 cells, which express native pendrin mRNA and pendrin protein (1). For this purpose, cells were exposed for 24 h to medium without or with 1 μM UGN (see materials and methods). As shown in Fig. 3C, UGN decreased endogenous PDS mRNA levels by 35% compared with control medium. In contrast, UGN treatment did not change TRPM6 mRNA levels. These findings indicate that UGN specifically regulates PDS mRNA levels.

Effect of UGN on hPDS promoter activity.

We next examined whether UGN modulates human PDS transcription at the level of the hPDS promoter. For this purpose, hPDS reporter constructs with progressive 5′-truncations encoding 4,185 bp (pL4), 2,060 bp (pL2), 1,433 bp (pL1.4), and 1,044 bp (pL1) of the hPDS promoter (Fig. 1) were transiently transfected into HEK293 cells (see materials and methods). As demonstrated in Fig. 4, exposure of transfected HEK293 cells to UGN (1 μM) caused a 20–25% decrease in luciferase activity driven by the hPDS promoter fragments of lengths 4, 2 and 1.4 kb compared with cells exposed to control medium. In contrast, no inhibition of reporter gene activity by UGN was observed with the 1-kb promoter fragment.

Fig. 4.

Effect of UGN on the hPDS promoter in HEK293 cells. A: schematic representation of the hPDS promoter constructs in pGL3-basic used in these experiments. B: cells were transfected with 0.3 μg pGL33-basic or 0.3 μg pGL3-basic containing inserts of decreasing size corresponding to the progressively truncated 5′-flanking region of hPDS (pL4, pL2, pL1.4, pL1). Cells were exposed to 1 μM UGN or control medium for 24 h, and luciferase activity was measured. Transfection efficiency was controlled for by cotransfection with 0.3 μg pCH110, and luciferase activity was normalized to β-galactosidase activity. Data represent the % change in luciferase activity in cells exposed to experimental medium (with 1 μM UGN) relative to cells exposed to control medium (without UGN). Values are means ± SE of 3–5 independent experiments, each performed in quadruplicate. UGN decreased hPDS promoter activity in HEK293 cells, and this inhibitory effect essentially disappeared upon shortening the promoter fragment from 1.4 to 1 kb. In these and subsequent promoter-reporter experiments, transfection efficiency was 70–80%, and luciferase activity in untreated control cells was comparable from one batch to another. **P < 0.01.

This finding suggests that UGN modulates activity of the hPDS promoter through a UGN response element (URE) located within the 389 bp between nt −1433 and −1044 upstream of the hPDS translational initiator codon.

To further define the hypothesized URE located within this 389-bp promoter region, deletion analysis was performed. Plasmids pL1, pL1.1, pL1.2, pL1.3, and pL1.4 (Fig. 5A) were transfected into HEK293 cells and exposed to media without or with 1 μM UGN. As demonstrated in Fig. 5A, HEK293 cell transfection with pL1.4 (1.4 kb), pL1.3 (1.3 kb), pL1.2 (1.2 kb), and pL1.1 (1.1 kb) resulted in similar magnitudes (20–25%) of UGN-induced inhibition of luciferase activity, whereas transfection with pL1 (1.0 kb) caused a marked decrease in the UGN-induced effect. These findings suggest the presence of a putative URE within the 96 bp between nt −1140 and −1044 of the hPDS promoter.

Fig. 5.

Effect of UGN on a 400-nt region of hPDS promoter in HEK293 cells. Depicted at left are schematics of the hPDS promoter constructs in pGL3 basic (A) or in pGL3 SV40 minimal promoter vectors (B) used in these experiments. A: cells were transfected with 0.3 μg pGL3-basic or 0.3 μg pGL3-basic containing progressive 5′-truncated hPDS promoter segments (pL1.4, pL1.3, pL1.2, pL1.1, pL1). B: cells were transfected with 0.3 μg pGL3 SV40 minimal promoter vector or 0.3 μg pGL3-SV40 containing the 110-bp hPDS promoter fragment from nt −1044 to −1153 (pJpr110) or the 5′-truncated 58-bp fragment corresponding to nt −1044 to −1101 (pJpr58). In both A and B, cells were exposed to 1 μM UGN or control medium for 24 h, and luciferase activity was measured. Transfection efficiency was controlled for by cotransfection with 0.3 μg pCH110, and luciferase activity was normalized to β-galactosidase activity. Data represent the % change in luciferase activity in cells exposed to experimental medium (with 1 μM UGN) relative to cells exposed to control medium (without UGN). As shown in A, UGN decreased hPDS promoter activity in HEK293 cells. This UGN-induced effect was evident in cells transfected with the 1.4- to 1.1-kb PDS promoter fragments, and markedly diminished when the fragment length was shortened from 1.1 to 1 kb. As demonstrated in B, UGN inhibited luciferase activity in cells transfected with the 110-bp promoter fragment. This UGN-induced effect markedly decreased when the fragment size was shortened to 58 bp. Values are means ± SE of 3–6 independent experiments, each performed in quadruplicate. **P < 0.01.

The URE was localized more precisely within this 96-bp promoter sequence through finer scale deletion analysis. The 110-bp hPDS promoter sequence from nt −1153 to −1044 as well as a 5′-truncated 58-bp sequence between nt −1101 and −1044 were cloned into the pGL3-promoter upstream of the SV40 minimal promoter and transfected into HEK293 cells. Figure 5B illustrates that in cells transfected with the plasmid pJpr110 containing the 110-bp hPDS promoter sequence, UGN decreased hPDS mRNA levels by 20%. However, truncation of the promoter fragment from nt −1153 to −1101 in plasmid pJpr58 completely abolished the UGN-induced effect on reporter gene activity. Taken together, these experiments suggest that the 52-bp DNA sequence between nt −1153 and −1101 contains a regulatory element that is necessary for transcriptional regulation by UGN.

HSE dependence of hPDS promoter regulation by UGN.

Computer analysis of this segment (21) revealed at nt −1119 to −1115 (CTTCT) a HSE (AGAAN). We therefore examined whether the effect of UGN on the promoter was HSE dependent. For this purpose, a nucleotide within the HSE motif was point-mutated in both pL1.4 and pJpr110 (CTACT; see materials and methods). HEK293 cells transfected with these constructs were exposed to UGN as above. As shown in Fig. 6A, the decrease in pL1.4 activity following UGN treatment was greatly diminished in cells transfected with the mutant promoter fragment. Similarly, as demonstrated in Fig. 6B, the UGN-induced reduction in luciferase activity seen in cells transfected with wild-type HSE-containing pJpr110 plasmid was completely abolished in the corresponding plasmid with the mutant HSE site. These experiments suggest that UGN modulates hPDS activity by a mechanism that requires the promoter's HSE site at nt −1119 to −1115.

Fig. 6.

Role of a defined heat shock element (HSE) in the effect of UGN on the hPDS promoter in HEK293 cells. A: effect of UGN on the 1.4-kb 5′-flanking region. Cells were transfected with 0.3 μg pGL3-basic or 0.3 μg pGL3-basic containing the 1.4-kb 5′-flanking region of hPDS (pL1.4) or 0.3 μg pL1.4 harboring a point-mutated HSE site (pL1.4mut). Cells were exposed to 1 μM UGN or control medium for 24 h, and luciferase activity was then measured. B: effect of UGN on the 110-bp HSE-containing DNA region. Cells were transfected with 0.3 μg pGL3-SV40 minimal promoter vector or 0.3 μg pGL3-SV40 containing the 110-bp hPDS promoter fragment with a wild-type HSE site (pJpr110) or 0.3 μg pJpr110 harboring a point-mutated HSE site (pJpr110mut). Cells were exposed to UGN or control as in A, and luciferase activity was measured and normalized to β-galactosidase activity. Data represent the % change in luciferase activity in cells exposed to experimental medium (with 1 μM UGN) relative to cells exposed to control medium (without UGN). Whereas UGN inhibited luciferase activity in cells transfected with the wild-type 1.4-kb promoter fragment (A) or the wild-type 110-bp fragment (B), mutating the HSE site in either of these fragments markedly diminished the UGN-induced effect. Values are means ± SE of 3–5 independent experiments, each performed in quadruplicate. *P < 0.05.

HSF1 regulates hPDS promoter activity through the HSE.

Since the HSE motif is recognized by HSF1 and HSF2, we analyzed the presence of HSFs in HEK293 cells. RT-PCR analysis in HEK293 cells revealed expression of mRNA encoding HSF1 but not HSF2 (data not shown). Therefore, we tested by quantitative real-time PCR the effect of UGN on hPDS mRNA levels in cells in which HSF1 was knocked down by siRNA transfection (Fig. 7; see materials and methods). Transfection of HSF1 siRNA reduced endogenous HSF1 mRNA levels in HEK293 cells to 24% of the levels in control cells (Fig. 7A). As demonstrated in Fig. 7B, the 30% reduction in hPDS mRNA levels by UGN treatment of HEK293 cells transfected with control siRNA (middle) was completely abolished in HEK293 cells transfected with HSF1 siRNA (right).

Fig. 7.

Effect of heat shock factor (HSF) 1 small interfering RNA (siRNA) on endogenous PDS mRNA levels in UGN-treated HEK293 cells. HEK293 cells were transfected with HSF1 siRNA or control siRNA before treatment with medium containing or lacking UGN (1 μM). See materials and methods. Subsequently, total RNA was extracted and real-time PCR analysis of the HSF1 mRNA (A) or PDS mRNA (B) was performed. Results in A represent relative change in HSF1 mRNA in cells transfected with HSF1 siRNA compared with cells treated with control siRNA. Shown in B is PDS mRNA in 1) cells transfected with HSF1 siRNA compared with cells treated with control siRNA (left), 2) cells transfected with control siRNA and exposed to experimental medium (with UGN) relative to control siRNA-transfected cells exposed to a medium without UGN (middle), or 3) cells transfected with HSF1 siRNA and exposed to experimental medium (with UGN) relative to HSF1 siRNA-transfected cells exposed to a medium without UGN (right). Values were normalized to the housekeeping genes actin (A) or TBP (B). See materials and methods. Values are mean ± SE of 4 independent experiments, each performed in duplicate. As demonstrated in A, HSF1 siRNA transfection reduced HSF1 mRNA expression by 76%. As shown in B, HSF1 siRNA transfection had no effect on PDS mRNA level (left), whereas UGN reduced PDS mRNA level by 30% in cells that were transfected with control siRNA (middle), and PDS mRNA level remained unchanged in UGN-treated cells that were transfected with HSF1 siRNA (right). *P < 0.05.

These findings provide strong evidence for the involvement of HSF1 in the regulation of the pendrin gene by UGN.

DISCUSSION

In this study, we describe transcriptional regulation of the hPDS gene encoding pendrin, a major anion exchanger in the CCD, by UGN, an important modulator of fluid and electrolyte homeostasis. We demonstrate a UGN-induced decrease in both pendrin protein expression (Fig. 2) and pendrin mRNA level (Fig. 3, A and B) in kidneys harvested from mice that were injected with this modulator as well as a UGN-induced reduction in endogenous pendrin mRNA level in renal cells (Fig. 3C). In addition, we demonstrate that UGN decreases hPDS promoter activity in transfected renal cells (Figs. 4 and 5). We show that UGN modulates hPDS promoter activity in large part via a HSE located between nt −1119 and −1115 of the hPDS promoter (Figs. 5 and 6), and we demonstrate that HSF1 likely mediates this UGN-induced modulation (Fig. 7). These findings suggest that pendrin gene expression is subject to HSE/HSF1-mediated transcriptional regulation by UGN.

The renal CCD plays a vital role in acid-base homeostasis and electrolyte balance and includes principal cells and IC (types A, B, and non-A, non-B) (2, 45). Pendrin functions as an apical Cl⁻/HCO₃⁻ exchanger in type B, and non-A, non-B IC cells of the CCD (44, 51, 55). In addition to its role in secreting bicarbonate and maintaining acid-base balance, pendrin is also involved in NaCl balance and blood pressure regulation (43, 44, 53, 55). Pendrin activity has been shown to be modulated by several factors including systemic pH (18, 39, 58), body Cl⁻ stores (43, 53, 56), H₂O intake (23), as well as the hormones aldosterone (55) and angiotensin II (38). We have demonstrated that systemic pH (1), aldosterone (1), and extracellular Cl⁻ level (13) regulate pendrin activity at the transcriptional level.

The GN peptides play a major role in regulating extracellular fluid volume, electrolyte homeostasis, as well as blood pressure (28). Their well-known function is the prevention of hypernatremia and hypervolemia, following an oral salt load (17, 47). The GN peptides, which are produced in the intestine as well as in the kidney and several other organs, exert their effect on water and electrolyte balance by increasing Na⁺, Cl⁻, HCO₃⁻, and water secretion in the intestine as well as by increasing urinary excretion of Na⁺, Cl⁻, K⁺, and water without changing renal blood flow and glomerular filtration rate (8, 15, 19, 47). The effect of the GN peptides, which essentially act as intestinal natriuretic factors, on renal electrolyte and water handling is achieved via both endocrine and paracrine mechanisms (8, 47). In this important role, the GN peptides join other well-known regulatory systems of body fluid and electrolyte balance, including the renin-angiotensin-aldosterone system, arginine-vasopressin, atrial natriuretic peptide (ANP) and its homologs, and the nitric oxide (NO) system (17, 47). Similar to the ANP and NO systems, the cellular function of the GN peptides, at least in the intestine and the proximal tubule, is achieved via the action of the intracellular second messenger cGMP (17, 47).

The nephron segments identified to date as target sites for GN/UGN action include the proximal tubule (47, 49), and the principal cells of the CCD (48, 50). In the proximal tubule, UGN inhibits luminal Na⁺/H⁺ exchange and K⁺ channels and basolateral Na⁺-K⁺-ATPase through mechanisms dependent on GC-C, cGMP (47, 49), and G proteins (49). GC-C is not expressed in the CCD (41, 50), and UGN acts on CCD principal cells to inhibit ROMK K⁺ channels via GC-C-independent signaling through PLA₂ (49). However, the natriuresis, kaliuresis, diuresis, and, in particular, the chloriuresis induced by UGN (8, 48) are not fully explained by these cellular pathways. Thus discovery and characterization of additional GC-C-independent mechanisms of UGN action on the renal tubule remain of great interest. We have therefore examined the role of pendrin activity in the regulation of electrolyte and water excretion by UGN.

In the present study, which aimed to investigate the effect of UGN on the expression of the pendrin gene, we have used the models of the intact mouse, the mouse kidney, and cultured renal cell lines. The mouse model allowed exploration of the effect of the peptide hormone UGN in a physiologically relevant environment. Cell lines transfected with hPDS promoter constructs enabled direct examination of the promoter's role in UGN regulation of pendrin activity, independently of other numerous modulators of pendrin activity such as pH, Cl⁻, and aldosterone, among others.

The experiments using mouse kidney and HEK293 cells, showing the UGN-induced effect on pendrin at the PDS mRNA level, followed by the experiments using transfected renal cells, demonstrating the effect of this hormone at the PDS promoter level, clearly indicated that UGN regulates pendrin expression at the transcriptional level. These studies culminated with the experiments demonstrating that the effect of UGN on the PDS promoter likely involves binding of transcription factor HSF1 to an HSE located between nt −1119 and −1115 within the promoter.

The heat shock response (HSR) is a highly conserved, fundamental defense mechanism that protects living cells against proteotoxic stress such as heat, infection, and inflammation, pharmacological agents. and other stresses (3, 30). The HSFs comprise a group of transcription factors that regulate the HSR (3, 30, 31). Mammalian genomes encode three homologs of HSF including HSF-1, HSF-2, and HSF-4 (20, 37, 40).

The HSFs exert their regulatory activity by binding to specific promoter elements (HSEs) consisting of tandem repeats of the sequence 5′-nGAAn-3′ (3, 30). HSEs were first defined upstream of cytoprotective heat shock genes including heat shock protein (HSP) 70, HSP90, HSP27, and other molecular chaperonins of the HSR network (30, 31), and are conserved from yeasts to humans. Upon exposure to heat shock and other stresses, inactive HSF1 monomers in cytoplasm undergo trimerization. HSF1 trimers translocate to the nucleus, undergo phosphorylation and sumoylation, bind to the HSE, and induce transcription of the target gene (3).

Whereas the role of the HSF/HSE system in regulating the activity of HSPs mediating the HSR has been well established, little is known about the full range of the biological target genes for the HSFs. In addition to the HSP genes, HSEs have been also described in genes encoding proteins with nonchaperonin functions (52), including proteins involved in transport processes such as the multiple drug resistance genes (57) and occludin (11). Whole genome analysis of the plant Arabidopsis thaliana (5) and the yeast Saccharomyces cerevisiae (20) identified small-molecule solute transport proteins, among other proteins, as direct transcriptional targets of plant and yeast HSF, respectively. The HSP90 complex has been shown to regulate the ANP receptor in HEK293 cells (26), and ANP induces HSP32 in human endothelial cells (22).

Our mutation analysis (Fig. 6) combined with RNA-silencing experiments (Fig. 7) provides strong evidence for the binding of HSF1 to the HSE of the hPDS promoter as playing an essential role in the modulation of the pendrin gene by UGN. Our study is the first report of a mammalian kidney solute transporter transcriptionally regulated by the HSF/HSE system in a hormone-specific manner. Very recently, the GLT1 glutamate transporter was shown to be transcriptionally regulated by HSF1 in neuroprogenitor lines of neuroblastoma/glioma cells (27).

The intestinal GN system of teleost fishes plays an important role in seawater adaptation (10, 17, 61). UGN mRNA is upregulated in the intestine and kidney of eels when they are exposed to the NaCl-rich environment of seawater (10, 61). The conserved primary structure of UGN throughout vertebrate evolution suggests that UGN-mediated regulation of systemic Na⁺ balance is the mammalian counterpart of UGN-mediated osmoregulation in those teleost fishes which live in either freshwater or seawater (17). Our findings demonstrating the involvement of the HSF/HSE axis, a stress-stimulated pathway, in the UGN-induced modulation of pendrin gene transcription support this notion and raise the possibility of an adaptive chloriuric response to osmotic or salt load stress mediated in the kidney by this novel UGN-HSR-pendrin connection. However, the signaling pathway whereby UGN triggers HSF1 biding to the HSE of the PDS promoter to activate transcription remains to be clarified.

Recently, Goy and coworkers (32, 33, 34, 42) have thoroughly investigated the biochemical profile and biological activity of the UGN system. These studies have shown that proUGN is the endocrine agent released from the intestine into the circulation or produced in the kidney (33) in response to oral salt intake and that it is converted in the kidney to active UGN (42). The same authors have provided evidence for the presence of two UGN isomers with saluretic activity, A and B, which markedly differ in their activity profile (34). Isomer A likely exerts its effect on the kidney through a GC-C-dependent pathway, whereas isomer B may achieve its saluretic effect via a distinct, still undefined receptor. The relative contribution of the propeptides, the active peptides, and their topoisomers in mediating the pendrin-dependent, chloriuretic response of UGN remains a subject for future investigation.

The interaction between the ANP and HSR systems (22, 26) and our current findings on the UGN-HSR connection suggest the importance of future investigation of possible cross talk between the GN peptides and the ANPs.

The data in this study demonstrate transcription regulation of pendrin, but do not address possible UGN-induced regulation of pendrin activity at translational or posttranslational levels. The techniques applied for mRNA and protein determination in our study do not define the proportional contributions of transcriptional and translational regulations of the PDS gene and its gene product to the observed modulatory response. Noteworthy in this regard is the well-established effect of the heat shock system on transporter protein trafficking, as exemplified by the HSP70-induced regulation of aquaporin-2 trafficking in the rat kidney (29).

In conclusion, our findings provide the first direct evidence that pendrin-mediated Cl⁻/HCO₃⁻ exchange in the renal tubule is regulated at the transcriptional level by UGN. UGN exerts its activity on the hPDS promoter, likely via the HSE site located between positions nt −1119 and −1115.

Taken together, our in vivo and in vitro findings provide evidence that the inhibitory effect of UGN on pendrin activity occurs along the entire pathway of pendrin production: from HSE-mediated changes in activity of its gene promoter, changes at the mRNA level, and, finally, changes at the pendrin protein level. The relative contribution of each of these changes to the action of UGN on pendrin activity is a subject for future research. Nevertheless, all observed changes could be explained by a single inhibitory action on transcription via the HSF/HSE system.

Our findings identify the pendrin Cl⁻/HCO₃⁻ exchanger of the non-A IC of the CCD as an important renal target of UGN, the “intestinal natriuretic peptide,” and provide a possible novel explanation for a significant part of UGN-induced chloriuresis. Our study provides insight into a novel signaling pathway involved in the enterorenal link controlling electrolyte and water homeostasis. Further studies utilizing various experimental models may shed more light on the transcriptional effects of the GN peptides on the pendrin gene, on the role of the heat shock system in these effects, and, thus, on acid-base balance, renal salt homeostasis, and blood pressure regulation in health and disease.

GRANTS

I. Zelikovic was supported by the USA-Israel Binational Science Foundation, by the Rappaport Institute for Research in the Medical Sciences, and by the Dr. Y. Rabinovitz Research Fund, Technion-Israel Institute of Technology; J. Rozenfeld received support from the Dora and Sydney Gabrel Fund, Rambam Medical Center, Haifa, Israel; S. L. Alper was supported by National Institutes of Health (NIH) Grants DK43495 and DK34854 (Harvard Digestive Diseases Center) and by the USA-Israel Binational Science Foundation; S. L. Carrithers was supported, in part, by NIH grants DK070374 and DK089892 and by KSTC-184-512-08-046.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: J.R., O.T., S.L.A., and I.Z. provided conception and design of research; J.R., O.T., O.K., L.A., E.E., and I.Z. performed experiments; J.R., O.T., E.E., and I.Z. analyzed data; J.R., O.T., and I.Z. interpreted results of experiments; J.R., O.T., and I.Z. prepared figures; J.R., S.L.A., and I.Z. drafted manuscript; J.R., S.L.C., S.L.A., and I.Z. edited and revised manuscript; J.R., S.L.A., and I.Z. approved final version of manuscript.