A common NHE3 single-nucleotide polymorphism has normal function and sensitivity to regulatory ligands

This study reports results on the functional significance of human NHE3-799C under basal conditions and in response to regulatory ligands, including a novel NHE3 inhibitor called tenapanor. We demonstrate that NHE3-799C is a common variant of NHE3 that is enriched in Asian populations; however, in contrast to our previous studies using rabbit NHE3, its presence seems to have limited clinical significance in humans and is not associated with compromised function or abnormal transport regulation.

Abstract

Na⁺/H⁺ exchanger NHE3 mediates the majority of intestinal and renal electroneutral sodium absorption. Dysfunction of NHE3 is associated with a variety of diarrheal diseases. We previously reported that the NHE3 gene (SLC9A3) has more than 400 single-nucleotide polymorphisms (SNPs) but few nonsynonymous polymorphisms. Among the latter, one polymorphism (rs2247114-G>A), which causes a substitution from arginine to cysteine at amino acid position 799 (p.R799C), is common in Asian populations. To improve our understanding of the population distribution and potential clinical significance of the NHE3-799C variant, we investigated the frequency of this polymorphism in different ethnic groups using bioinformatics analyses and in a cohort of Japanese patients with cardiovascular or renal disease. We also characterized the function of human NHE3-799C and its sensitivity to regulatory ligands in an in vitro model. NHE3-799C had an allele frequency of 29.5–57.6% in Asian populations, 11.1–23.6% in European populations, and 10.2–22.7% in African populations. PS120/FLAG-NHERF2 fibroblasts stably expressing NHE3-799C had lower total protein expression but a higher percentage of surface expression than those expressing NHE3-799R. NHE3-799C had similar basal activity to NHE3-799R and was similarly stimulated or inhibited, by serum or forskolin, respectively. Tenapanor, a small-molecule NHE3 inhibitor, dose-dependently inhibited NHE3-799R and NHE3-799C activities. The IC₅₀ values of tenapanor for NHE3-799C and NHE3-799R were significantly different, but both were in the nanomolar range. These results suggest that NHE3-799C is a common variant enriched in Asian populations, is not associated with compromised function or abnormal regulation, and is unlikely to contribute to clinical disease.

NEW & NOTEWORTHY This study reports results on the functional significance of human NHE3-799C under basal conditions and in response to regulatory ligands, including a novel NHE3 inhibitor called tenapanor. We demonstrate that NHE3-799C is a common variant of NHE3 that is enriched in Asian populations; however, in contrast to our previous studies using rabbit NHE3, its presence seems to have limited clinical significance in humans and is not associated with compromised function or abnormal transport regulation.

a transport protein, Na⁺/H⁺ exchanger isoform 3 (NHE3, SLC9A3) is primarily localized on the epithelial brush border in the small intestine, colon, gall bladder, and renal proximal tubule but is also present in the thick ascending limb of the loop of Henle in the kidney and a small population of cells in the respiratory center (5). NHE3 carries out the majority of electroneutral sodium absorption in the small intestine and kidney and plays a critical role in water and sodium homeostasis (5). Inhibition of intestinal NHE3 occurs in almost all diarrheal diseases (9) and has been suggested to contribute to sudden infant death syndrome (16).

We have reported that the NHE3 gene, SLC9A3, has more than 400 single-nucleotide polymorphisms (SNPs) in the general population, although the number of SNPs that lead to changes in amino acids is surprisingly small (22). Among the nonsynonymous polymorphisms, the rs2247114-G>A polymorphism is of particular interest because it is the only common nonsynonymous variant identified in our preliminary analysis that has a global minor allele frequency greater than 0.05. This SNP is located in exon 16 of 17 in the NHE3 gene and results in a substitution from arginine (R) to cysteine (C) at amino acid position 799 (p.R799C). Hereafter, this SNP (p.R799C) will be referred to as the NHE3-799C variant, which is distinguished from the more common NHE3-799R variant. Of potential clinical relevance, the frequency of the NHE3-799C variant varies among ethnic groups, with a relatively high prevalence in Asian populations and a low prevalence in European populations (22); however, the importance of this finding has not been investigated in detail.

We previously characterized the function of the NHE3-799C variant using rabbit NHE3 in an in vitro model, showing that it was associated with defects in intrinsic activity and trafficking. These findings suggest the possibility of a clinical phenotype such as increased susceptibility to diarrheal disorders associated with the NHE3-799C variant (22). However, human and rabbit NHE3 are only 86% identical in protein sequence, so it remains to be determined whether human NHE3-799C is also associated with abnormal function or regulation that may be of clinical significance.

Given its important role in water and sodium homeostasis, NHE3 is a drug target of interest for a variety of diseases (4, 18). Tenapanor (Ardelyx, Fremont, CA) is a small-molecule NHE3 inhibitor. It has a potent inhibitory effect by binding directly to NHE3 proteins expressed on the apical plasma membrane of intestinal epithelial cells. Tenapanor is minimally absorbed and therefore has the potential to have a low risk of systemic side effects (10). Recent clinical studies have demonstrated a potential therapeutic role of tenapanor in constipation-predominant irritable bowel syndrome (IBS-C) (3, 19). The sensitivity of NHE3-799C as a common variant to tenapanor, however, has not been characterized. Thus it is not known whether tenapanor would have the same clinical effectiveness in patients presenting with various genetic backgrounds of NHE3.

To improve our understanding of the prevalence and function of human NHE3-799C, we performed the present study to 1) determine the frequency of this variant across ethnic groups and 2) evaluate whether, and to what extent, this variant has compromised basal function or an abnormal response to stimulatory/inhibitory ligands, including tenapanor.

MATERIALS AND METHODS

Materials.

Anti-NHE3 and anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) antibodies were purchased from Novus Biologicals (Littleton, CO) and Sigma-Aldrich (St. Louis, MO), respectively. FluorSave Reagent was purchased from EMD Millipore (Billerica, MA). Nigericin was purchased from Cayman Chemical (Ann Arbor, MI). Tenapanor was provided by Ardelyx Inc. (Fremont, CA). All other reagents used in this study were purchased from either Sigma-Aldrich or Thermo Fisher Scientific (Waltham, MA).

Bioinformatics analyses.

The databases of the International HapMap Project (https://hapmap.ncbi.nlm.nih.gov/; accessed May 2016) and the National Heart, Lung, and Blood Institute (NHLBI) Exome Sequencing Project (http://evs.gs.washington.edu/EVS; accessed May 2016) were searched to determine the allele and genotype frequencies of NHE3-799C and NHE3-799R among various ethnic groups. Phylogenetic analysis was performed by searching the Ensembl genome database (http://www.ensembl.org/; accessed May 2016) (21). The function of human NHE3-799C was predicted using the Polymorphism Phenotyping v2 (PolyPhen-2) (http://genetics.bwh.harvard.edu/pph2/) (1) and Sorting Intolerant From Tolerant (SIFT) (http://sift.jcvi.org/) (11) tools. The Genotype-Tissue Expression (GTEx) project database (https://www.gtexportal.org/; accessed March 2017, dbGaP accession no. phs000424.v6.p1) was interrogated for expression quantitative trait loci (eQTL) data corresponding to the SNP causing NHE3-799C (rs2247114-G>A).

Cardiovascular or renal disease patient cohort genotyping.

As previously described (14), genomic DNA was isolated from peripheral blood leukocytes obtained from Japanese patients (of both sexes) diagnosed with cardiovascular or renal disease, who provided written informed consent. Patients’ blood samples were taken at outpatient clinics or participating hospitals (Gifu Prefectural General Medical Center, Gifu; Gifu Prefectural Tajimi Hospital, Tajimi; Japanese Red Cross Nagoya First Hospital, Nagoya; Inabe General Hospital, Inabe; Hirosaki University Hospital and Hirosaki Stroke Center, Hirosaki, Japan). The study protocol complied with the Declaration of Helsinki and was approved by the ethics committees of each participating institution. The genotypes of NHE3-799C and NHE3-799R were analyzed by Takara Bio (Kusatsu, Japan).

Plasmid construction, cell culture, and transfection.

The cDNA of human NHE3-799R was incorporated into the pcDNA3.1 vector. This construct was then used as the template for site-directed mutagenesis to produce the construct of human NHE3-799C. The sequences of the two constructs were confirmed by Sanger sequencing (Macrogen, Rockville, MD). PS120 fibroblasts with stable expression of triple FLAG-tagged NHERF2 (PS120/FLAG-NHERF2), which have no endogenous expression of plasma membrane NHEs, were used to establish stable cell lines expressing human NHE3-799R and NHE3-799C (20, 22). Transfection was performed using Lipofectamine 2000 according to the manufacturer’s protocol. Cells were cultured in Dulbecco's Modified Eagle Medium:Nutrient Mixture F-12 (DMEM/F-12) supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, 100 μg/ml streptomycin, and 1,000 μg/ml G418 in a 5% CO₂-95% air atmosphere at 37°C. Cells with stable transfection of human NHE3-799R and NHE3-799C were selected by exposure to antibiotics (G418) and regular acid challenge, as described previously (15).

Immunofluorescence.

Cells were grown to ~50% confluence on collagen type I-coated glass coverslips. Cells were washed with phosphate-buffered saline (PBS) and fixed for 15 min with 4% paraformaldehyde in PBS. After incubation with blocking/permeabilization solution [5% bovine serum albumin (BSA), 0.1% saponin in PBS] for 1 h, cells were washed with PBS and incubated overnight at 4°C with primary antibody against NHE3 (Novus NBP1-82574, rabbit polyclonal, 1:100) diluted in PBS containing 1% BSA. Cells were then washed with PBS and incubated with Hoechst (1:200) and anti-rabbit Alexa Fluor 488-conjugated secondary antibody (goat anti-rabbit immunoglobulin G, 1:200) for 1 h at room temperature. After washing with PBS, cells were mounted with FluorSave Reagent and studied using an LSM 510 confocal microscope system (Carl Zeiss Microscopy, Thornwood, NY).

Quantitative real-time PCR.

Total RNA was extracted from NHE3-799C- and NHE3-799R-transfected cells using PureLink RNA Mini Kit and subsequently reverse-transcribed to complementary DNA (cDNA) using SuperScript VILO Master Mix. Quantitative real-time PCR was performed using Power SYBR Green Master Mix on a QuantStudio 12K Flex real-time PCR system (Applied Biosystems, Foster City, CA). cDNA samples were run in triplicate and the resultant cycle threshold (C_t) values for NHE3 mRNA expression were normalized to the values for 18S rRNA. The following primers were used: NHE3 forward, 5′-GTCTTCCTCAGTGGGCTCAT-3′; NHE3 reverse, 5′-ATGAGGCTGCCAAACAGG-3′; 18S rRNA forward, 5′-GCAATTATTCCCCATGAACG-3′; 18S rRNA reverse, 5′-GGGACTTAATCAACGCAAGC-3′.

Immunoblotting.

Cells were rinsed three times with PBS and harvested in PBS by scraping. After centrifugation (5 min at 2,000 g), the cell pellet was solubilized in lysis buffer (60 mM HEPES, 150 mM NaCl, 3 mM KCl, 5 mM EDTA trisodium, 3 mM EGTA, 1 mM Na₃PO₄, 1% Triton X-100, pH 7.4) containing a protease inhibitor cocktail, and then homogenized by sonication. After spinning down the insoluble cellular debris by centrifugation (10 min at 5,000 g), the supernatant was collected as the protein lysate. Protein concentration was measured using the bicinchoninic acid method. Proteins were denatured at 70°C for 10 min, separated by SDS-PAGE on a 10% acrylamide gel, and transferred onto a nitrocellulose membrane. After blocking with 5% nonfat milk, the blot was probed with primary antibodies against NHE3 (Novus NBP1-82574, rabbit polyclonal, 1:500) and GAPDH (Sigma G8795, mouse monoclonal, 1:5,000) overnight at 4°C, followed by fluorescence-labeled secondary antibodies against rabbit and mouse immunoglobulin G (1:10,000) for 1 h at room temperature. Detection and quantification of protein bands were performed using an Odyssey system and Image Studio software (LI-COR Biosciences, Lincoln, NE).

Cell surface biotinylation.

Cells were grown to ~80% confluence in 100-mm Petri dishes and rinsed with PBS and borate buffer (154 mM NaCl, 1 mM boric acid, 7.2 mM KCl, 1.8 mM CaCl₂, pH 8.0), before being incubated with 0.5 mg/ml NHS-SS-biotin to label plasma membrane proteins. Unbound NHS-SS-biotin was removed using quenching buffer (20 mM Tris, 120 mM NaCl, pH 7.4). Cells were then washed with PBS, and the cell pellet was harvested and solubilized in lysis buffer, as described above. After homogenization by sonication, the protein lysate was centrifuged (10 min at 5,000 g) to remove insoluble cell debris. A small proportion of the protein lysate was collected as the total lysate; the rest was incubated with avidin-agarose overnight at 4°C. After centrifugation (10 min at 2,000 g) to sediment the avidin-agarose beads, the supernatant was collected as the intracellular fraction; the avidin-agarose beads were washed with lysis buffer containing 0.1% Triton X-100. Biotinylated proteins were then eluted from the beads into SDS buffer (5 mM Tris·HCl, 1% SDS, 10% glycerol, 1% 2-mercaptoethanol, pH 6.8) and collected as the surface fraction. Protein lysates from the total lysate, surface fraction, and intracellular fraction were immunoblotted as described above. The percentage of surface expression of NHE3 was calculated as previously reported (2).

Measurement of NHE3 activity.

NHE3 activity was measured by fluorometry using the intracellular pH (pH_i)-sensitive dye BCECF-AM, as described previously (13). In brief, cells were grown on glass coverslips to ~70–90% confluence and incubated with 10 μM BCECF-AM in Na⁺/NH₄Cl solution (98 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgSO₄, 1 mM NaH₂PO₄, 25 mM glucose, 20 mM HEPES, 40 mM NH₄Cl, pH 7.4) for 20 min at 37°C to allow dye loading and acidification. Cells were then mounted in a fluorometer (Photon Technology International, Birmingham, NJ) and superfused with TMA⁺ solution [138 mM tetramethylammonium chloride (TMACl), 5 mM KCl, 2 mM CaCl₂, 1 mM MgSO₄, 1 mM NaH₂PO₄, 25 mM glucose, 20 mM HEPES, pH 7.4], followed by Na⁺ solution (138 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgSO₄, 1 mM NaH₂PO₄, 25 mM glucose, 20 mM HEPES, pH 7.4) for Na⁺-dependent pH_i recovery. For each coverslip, pH_i was calibrated using K⁺-clamp solutions set at pH 6.0, 6.6, and 7.3 with 10 μM nigericin. Data were analyzed using Origin 6.0 software (OriginLab, Northampton, MA). The maximum velocity (V_max) and Michaelis-Menten constant (K_m [H⁺]_i) were determined using the Hill equation. At least three coverslips were analyzed in each experiment (set as n = 1), and experiments were repeated four times. In addition, intrinsic NHE3 activity was calculated by normalizing the basal activity according to the amount of NHE3 expressed on the plasma membrane, based on the results of immunoblotting and cell surface biotinylation, as reported previously (22).

Stimulation/inhibition of NHE3 activity.

Cells that were ~70–90% confluent on glass coverslips were subjected to serum starvation for 3 h, followed by incubation with 10 μM BCECF-AM in Na⁺/NH₄Cl solution ± 10% dialyzed serum or 10 μM forskolin for 20 min at 37°C. Cells were then mounted in the fluorometer and NHE3 activity was determined as described above. In each experiment, at least three coverslips were studied for each condition (basal, serum, and forskolin), and the activities following serum/forskolin treatment were compared with the basal activity, which was set as 100%. Experiments were repeated four times.

Determination of IC₅₀ of tenapanor.

Cells were grown on glass coverslips to ~70–90% confluence. After incubation with 10 μM BCECF-AM in Na⁺/NH₄Cl solution for 20 min at 37°C, cells were mounted in the fluorometer and superfused with TMA⁺ solution with various concentrations of tenapanor for 10 min. The solution was then switched to Na⁺ solution with the same concentration of tenapanor to allow Na⁺-dependent pH_i recovery, followed by K⁺-clamp solutions for calibration of pH_i. The rate of pH_i recovery (ΔpH_i/Δt) over the initial 30 s following the switch to Na⁺ solution was analyzed. The effects of serial concentrations of tenapanor (0, 0.03, 0.1, 0.3, 1, 3, 10, 30, and 100 nM) on NHE3-mediated initial pH_i recovery were investigated, and at least two coverslips were studied for each concentration of tenapanor in each experiment. Five experiments were repeated. The IC₅₀ (the concentration of inhibitor that inhibits the response by 50%) of tenapanor for human NHE3-799R and NHE3-799C was calculated using a logistic regression model and Origin 6.0 software.

Statistical analysis.

Data are presented as means ± SE. Statistical analyses were conducted using the Student’s t-test or Pearson's χ-squared test where appropriate. P < 0.05 was considered statistically significant.

RESULTS

NHE3-799C is a nonconservative variant enriched in Asian populations.

The prevalence of NHE3-799R and NHE3-799C in various ethnic populations is shown in Table 1. Bioinformatics analyses showed that NHE3-799R was the major genotype over NHE3-799C in most ethnic populations (Table 1) and was highly conserved among mammals (Fig. 1); NHE3-799C was a nonconservative variant and was probably exclusive to humans. However, the allele frequency of NHE3-799C varied greatly across the ethnic groups: 29.5–57.6% in Asian populations, 11.1–23.6% in European populations, and 10.2–22.7% in African populations. In particular, NHE3-799C seemed to be mostly enriched in East Asian populations, with an allele frequency of 39.0–57.6% among Chinese and Japanese groups. To confirm these database observations, we measured the allele frequency of NHE3-799C in Japanese patients who had cardiovascular or renal disease (Table 2). The allele frequency of NHE3-799C in this population was 54.0%, which was similar to the database frequency. Taken together, these results suggest that NHE3-799C is a nonconservative variant that is enriched in Asian populations.

Table 1.

Prevalence of NHE3-799C (corresponding to allele A) and NHE3-799R (corresponding to allele G) across ethnic groups

Ethnic Group	Population Code	Population Description	Allele: Frequency (Count)	Genotype: Frequency (Count)	Data Source
East Asian	HCB	Han Chinese in Beijing, China	A: 0.500 (43)	A/A: 0.256 (11)	HapMap
			G: 0.500 (43)	A/G: 0.488 (21)
				G/G: 0.256 (11)
	CHB	Han Chinese in Beijing, China	A: 0.390 (32)	A/A: 0.146 (6)	HapMap
			G: 0.610 (50)	A/G: 0.488 (20)
				G/G: 0.366 (15)
	CHD	Chinese in Denver, CO	A: 0.576 (98)	A/A: 0.306 (26)	HapMap
			G: 0.424 (72)	A/G: 0.541 (46)
				G/G: 0.153 (13)
	JPT	Japanese in Tokyo, Japan	A: 0.541 (93)	A/A: 0.233 (20)	HapMap
			G: 0.459 (79)	A/G: 0.616 (53)
				G/G: 0.151 (13)
South Asian	GIH	Gujarati Indians in Houston, TX	A: 0.295 (52)	A/A: 0.057 (5)	HapMap
			G: 0.705 (124)	A/G: 0.477 (42)
				G/G: 0.466 (41)
American	MEX	Mexican ancestry in Los Angeles, CA	A: 0.296 (29)	A/A: 0.041 (2)	HapMap
			G: 0.704 (69)	A/G: 0.510 (25)
				G/G: 0.449 (22)
European	TSI	Tuscans in Italy	A: 0.236 (41)	A/A: 0.046 (4)	HapMap
			G: 0.764 (133)	A/G: 0.379 (33)
				G/G: 0.575 (50)
	CEU	Utah residents (CEPH) with Northern and Western European Ancestry	A: 0.111 (25)	A/A: 0.009 (1)	HapMap
			G: 0.889 (201)	A/G: 0.204 (23)
				G/G: 0.788 (89)
		European American	A: 0.130 (1,118)	A/A: 0.022 (93)	NHLBI Exome Sequencing Project
			G: 0.870 (7,474)	A/G: 0.217 (932)
				G/G: 0.761 (3,271)
African	YRI	Yoruba in Ibadan, Nigeria	A: 0.102 (23)	A/G: 0.204 (23)	HapMap
			G: 0.898 (203)	G/G: 0.796 (90)
	ASW	African ancestry in Southwest USA	A: 0.112 (11)	A/G: 0.224 (11)	HapMap
			G: 0.888 (87)	G/G: 0.776 (38)
	LWK	Luhya in Webuye, Kenya	A: 0.194 (35)	A/A: 0.033 (3)	HapMap
			G: 0.806 (145)	A/G: 0.322 (29)
				G/G: 0.644 (58)
	MKK	Maasai in Kinyawa, Kenya	A: 0.227 (65)	A/A: 0.056 (8)	HapMap
			G: 0.773 (221)	A/G: 0.343 (49)
				G/G: 0.601 (86)
		African-American	A: 0.126 (554)	A/A: 0.018 (40)	NHLBI Exome Sequencing Project
			G: 0.874 (3,848)	A/G: 0.215 (474)
				G/G: 0.766 (1,687)

Fig. 1.

NHE3-799R is highly conserved among mammals, whereas NHE3-799C is a nonconservative variant. Phylogenetic analysis was performed by searching the Ensembl genome database (accessed May 2016). NHE3-799R (corresponding to allele G) was highly conserved among mammals, whereas NHE3-799C (corresponding to allele A) was a nonconservative variant and was probably exclusive to humans.

Table 2.

Genotypes of NHE3-799C (corresponding to allele A) and NHE3-799R (corresponding to allele G) in Japanese patients with cardiovascular disease (CVD) or chronic kidney disease (CKD)

Genotype	Patients with CVD, % (n = 2,357)	Patients with CKD*, % (n = 1,501)	Total Patients†, % (n = 4,200)
A/A	30.0	29.0	28.9
A/G	49.0	49.0	50.2
G/G	21.0	22.0	20.9

^*Patients with CKD included those with an estimated glomerular filtration rate (eGFR) < 60 ml·min⁻¹·1.73 m⁻².

^†The total number of patients included an additional 342 patients who had no diagnosis of CVD and an eGFR > 60 ml·min⁻¹·1.73 m⁻².

Enrichment of NHE3-799C in Asian populations is not associated with cardiovascular and renal disease risk.

Of the Japanese patients who were genotyped, 28.9% were homozygous for NHE3-799C and 50.2% were heterozygous for NHE3-799C/NHE3-799R (Table 2). Examination of the relationship between diagnosis [cardiovascular disease (CVD) and chronic kidney disease (CKD)] and genotype frequency showed that the heterozygous and homozygous frequencies did not significantly differ (CVD: P = 0.58; CKD: P = 0.60), indicating that the homozygous NHE3-799C genotype is unlikely to be associated with an increased risk of CVD or CKD.

Homogeneous expression of human NHE3 in mixed clones of transfected cell lines.

To characterize human NHE3-799C in vitro, we transfected NHE3-null fibroblasts (PS120/FLAG-NHERF2) to establish cell lines stably expressing human NHE3-799C and NHE3-799R. For each transfected cell line, mixed cell clones were used for experiments following selection by antibiotics and regular acid challenge. First of all, we verified the homogeneity of NHE3 expression among mixed clones within each transfected cell line by immunofluorescence. As shown in Fig. 2, almost all cells transfected with human NHE3-799R or NHE3-799C had positive NHE3 immunostaining with slight variation between cell clones within each cell line. Additionally, NHE3 was mostly localized in the intracellular compartment in this fibroblast model, which is consistent with our previous finding (6).

Fig. 2.

Homogeneous expression of NHE3 among cell clones of transfected cell lines. PS120/FLAG-NHERF2 cells were transfected with human NHE3-799R or NHE3-799C, followed by selection by antibiotics and regular acid challenge. For each stably transfected cell line, mixed cell clones were used for experiments. Homogeneity of NHE3 expression among mixed clones was verified by immunofluorescence. NHE3 was expressed in almost all cells with slight variation between cell clones within each cell line. Scale bar, 50 µm.

Protein/mRNA expression of human NHE3-799R and NHE3-799C.

We assessed protein/mRNA expression of human NHE3-799R and NHE3-799C in transfected cells, as shown in Fig. 3. The total protein expression and mRNA expression of NHE3 were significantly lower in the NHE3-799C-transfected cells (protein: 45.0 ± 4.3%, P < 0.05, n = 4; mRNA: 71.6 ± 2.3%, P < 0.05, n = 3) compared with the NHE3-799R-transfected cells (set as 100%). In contrast, the percentage of surface expression of NHE3 was significantly higher in the cells expressing NHE3-799C compared with those expressing NHE3-799R (12.4 ± 0.4% vs. 8.1 ± 1.4%, P < 0.05, n = 3). Thus there was a moderately lower level of NHE3 protein expression on the plasma membrane in the NHE3-799C-transfected cells (68.9%) compared with the NHE3-799R-transfected cells (set as 100%).

Fig. 3.

NHE3-799C-transfected cells have lower total protein/mRNA expression but a higher percentage of surface protein expression than NHE3-799R-transfected cells. Total protein expression and mRNA expression were studied by immunoblotting and quantitative real-time PCR, respectively. Surface protein expression was determined by cell surface biotinylation and immunoblotting. A and B: the total protein expression of NHE3 was significantly lower in the NHE3-799C-transfected cells than in the NHE3-799R-transfected cells (55.0% reduction, P < 0.05, n = 4). C: quantitative real-time PCR revealed lower mRNA expression in the NHE3-799C-transfected cells compared with the NHE3-799R-transfected cells (28.4% reduction, P < 0.05, n = 3). D and E: the percentage of surface expression of NHE3 was significantly higher in the NHE3-799C-transfected cells than in the NHE3-799R-transfected cells (means ± SE: 12.4 ± 0.4% vs. 8.1 ± 1.4%, P < 0.05, n = 3). IF, intracellular fraction; SF, surface fraction; TL, total lysate. *P < 0.05.

The protein/mRNA expression in stably transfected cells is dependent on a number of factors such as efficiency of transfection, efficiency of selection, site of DNA integration, and variation between mixed clones. Using the transfected cell model, we were not able to compare the endogeneous expression of human NHE3-799C and NHE3-799R. To have an idea of the potential effect of the NHE3-799C genotype on endogenous protein/mRNA expression, we searched the GTEx eQTL database and found that NHE3-799C was associated with a dose-dependent reduction in mRNA expression in different human tissues such as thyroid, testis, and esophagus (Fig. 4). However, these eQTL data are limited by the small sample size, especially in the homozygous NHE3-799C group.

Fig. 4.

Association of NHE3 genotype and mRNA expression across human tissues. Expression quantitative trait loci (eQTL) analysis was performed by searching the Genotype-Tissue Expression (GTEx) project database (accessed March 2017). NHE3-799C (corresponding to allele A) was associated with lower mRNA expression compared with NHE3-799R (corresponding to allele G) across human tissues, including but not limited to thyroid, testis, and esophagus mucosa, which were the top three tissues ranked by P value.

Human NHE3-799R and NHE3-799C have comparable basal activity and are similarly stimulated or inhibited by regulatory ligands.

As shown in Fig. 5, PS120/FLAG-NHERF2 cells stably expressing human NHE3-799C appeared to have a lower V_max activity compared with those expressing NHE3-799R under the basal condition, but the difference was not statistically significant after four repeats (694.9 ± 61.7 vs. 891.4 ± 75.2 µM/sec, P = 0.11). Furthermore, there was no significant difference between human NHE3-799C and NHE3-799R in K_m (H⁺)_i (0.41 ± 0.05 vs. 0.42 ± 0.05, P = 0.89). After normalizing the basal activity by the amount of NHE3 protein expressed on the plasma membrane (based on the results of immunoblotting and surface biotinylation), we found that human NHE3-799C had similar intrinsic activity (113.2%) to NHE3-799R (set as 100%), indicating that human NHE3-799C is not functionally compromised. This is in agreement with the results of in silico analyses that consistently showed that human NHE3-799C was unlikely to have impaired function (PolyPhen-2: benign; SIFT: tolerated). Moreover, human NHE3-799R and NHE3-799C both responded normally by upregulating activity following serum stimulation (NHE3-799R: +26.4 ± 7.8%, P < 0.05; NHE3-799C: +32.4 ± 10.0%, P < 0.05), and by downregulating activity following forskolin inhibition (NHE3-799R: −31.2 ± 8.9%, P < 0.05; NHE3-799C: −35.7 ± 8.7%, P < 0.05).

Fig. 5.

NHE3-799C-transfected cells have similar basal activity to NHE3-799R-transfected cells and are similarly stimulated or inhibited, by serum or forskolin, respectively. A: basal NHE3 activity was not significantly different between cells stably expressing human NHE3-799C and NHE3-799R (means ± SE: V_max: 694.9 ± 61.7 vs. 891.4 ± 75.2 µM/s, P = 0.11, n = 4). B: the NHE3-799C-transfected cells and NHE3-799R-transfected cells both responded normally by upregulating activity following exposure to 10% dialyzed serum (means ± SE: +32.4 ± 10.0% and +26.4 ± 7.8%, n = 4), and by downregulating activity following exposure to 10 µM forskolin (means ± SE: −35.7 ± 8.7% and −31.2 ± 8.9%, n = 4). ns, not significant; V_max, maximum velocity. *P < 0.05.

Tenapanor inhibits the activity of human NHE3-799R and NHE3-799C dose-dependently in the nanomolar range.

Tenapanor inhibited the basal activity of human NHE3-799R in a dose-dependent manner in the nanomolar range; concentrations lower than 0.1 nM had minimal effect and concentrations higher than 30 nM inhibited almost 100% of basal activity (Fig. 6). A similar dose-dependent nanomolar inhibitory effect of tenapanor on human NHE3-799C was also observed (data not shown). The IC₅₀ values of tenapanor for NHE3-799C and NHE3-799R were similar, both being in the nanomolar range (2.41 ± 0.69 vs. 0.87 ± 0.34 nM, P < 0.05); however, the IC₅₀ value of tenapanor for NHE3-799C was consistently higher than that for NHE3-799R.

Fig. 6.

Tenapanor dose-dependently inhibits the activity of human NHE3-799C and NHE3-799R in the nanomolar range. A: representative data showing the dose-dependent inhibitory effect of tenapanor on the basal activity of human NHE3-799R. B: representative data showing the estimation of IC₅₀ of tenapanor for the two NHE3 variants using a logistic regression model. The IC₅₀ value of tenapanor was significantly higher for human NHE3-799C than for NHE3-799R (means ± SE: 2.41 ± 0.69 vs. 0.87 ± 0.34 nM, P < 0.05, n = 5).

DISCUSSION

The present study shows that human NHE3-799C, a common variant enriched in Asian populations, is not associated with compromised function, or abnormal sensitivity to stimulatory and inhibitory ligands, and is therefore unlikely to be associated with altered susceptibility to diarrheal diseases secondary to dysfunction/dysregulation of sodium absorption. In addition, this study does not find evidence of association between homozygous NHE3-799C genotype and cardiovascular or renal disease in an Asian patient cohort. Taken together, these findings suggest that homozygous expression of NHE3-799C does not result in a clinical phenotype.

In a previous study, we found that rabbit NHE3-799C had significantly lower basal activity and surface expression than rabbit NHE3-799R (22); these results are not consistent with those of the present study. This may be explained by the differences between human and rabbit NHE3, which are 86% identical in protein sequence, with most of the mismatches in the COOH-terminal domain that harbors the amino acid 799 position. It is possible that the substitution at amino acid position 799, which is not conserved among mammals, is tolerated in human NHE3 but not in the rabbit counterpart.

The NHE3 protein comprises a well-structured NH₂-terminal transmembrane domain (amino acids 1–456) and a COOH-terminal cytoplasmic domain (amino acids 457–834), and has both structured and nonstructured domains (8). The NH₂- and COOH-terminal parts of NHE3 have different functions: Na⁺/H⁺ exchange is carried out by the NH₂-terminal domain, whereas short-term regulation and interaction of NHE3 with its modulatory binding partners occurs in the COOH-terminal domain (5). Given that the R799C substitution is located in the COOH-terminal rather than in the NH₂-terminal domain, it can be predicted that NHE3-799C is less likely to have abnormal kinetics (e.g., elevated K_m) or compromised intrinsic function compared with variants located in the NH₂-terminal domain, examples of which cause a congenital diarrheal phenotype (9). This is confirmed by our findings in this study that human NHE3-799C and NHE3-799R had similar intrinsic activity and K_m. Initially, we hypothesized that NHE3-799C might have abnormal regulation, because the substitution occurs in the COOH-terminal domain that is potentially necessary for the effects of several ligands, including tyrosine kinases and the tyrosine kinase inhibitor genistein. However, this hypothesis is not supported by the present study, because the stimulatory effect of serum and the inhibitory effect of forskolin were both preserved in human NHE3-799C. In fact, the R799C substitution is located in the tail of the COOH-terminal domain, which is distant from the major regulatory domains that we have proposed previously, although this area does have multiple phosphorylation sites (6). It is therefore reasonable to assume that NHE3-799C is not associated with a major regulatory abnormality, although it may have differential sensitivity to a particular regulatory partner.

Based on our preliminary analysis, we performed a further investigation into the prevalence of NHE3-799C in various ethnic groups. We confirmed the previous finding that NHE3-799C was enriched in Asian populations, in particular East Asian populations. This indicates that NHE3-799C, as a nonconservative variant, may have a potential evolutionary role or clinical significance in a population-specific manner. Potential NHE3-implicated clinical diseases include diarrheal diseases, in which dysfunction/dysregulation of NHE3 is a common finding (9, 17), and cardiovascular and renal diseases, in which NHE3 may be implicated through its role in blood pressure regulation (12). The present study does not find any potential contribution of the homozygous NHE3-799C genotype to these diseases based on the following observations: First, NHE3-799C had similar intrinsic function and responded to stimulatory/inhibitory compounds in a manner similar to NHE3-799R. NHE3-799C is therefore unlikely to be associated with dysfunction/dysregulation of sodium absorption that may lead to diarrhea, unlike a number of genetic mutations located in the NH₂-terminal transmembrane domain of NHE3 that cause congenital sodium diarrhea (9). In addition, we searched several databases of acute diarrhea and did not find an NHE3 polymorphism that could be identified as a predisposing factor to acute diarrhea in developing countries. Similarly, our search for a role of NHE3 polymorphisms in diarrhea-predominant irritable bowel syndrome (IBS-D) in collaboration with Mayo Clinic failed to reveal any significant association. However, these two unpublished preliminary studies are insufficient for publication. Second, in a large cohort of Japanese patients with CVD and CKD, no evident relationship was established between diagnosis of these diseases and the homozygous NHE3-799C genotype. As such, the homozygous NHE3-799C genotype is unlikely to be associated with an increased risk of diarrhea, CVD, and CKD.

Although this study does not find any clinical significance of NHE3-799C, it is still possible that NHE3-799C is implicated in particular diseases. Further research is encouraged to investigate whether 1) the NHE3-799C variant responds differently to specific clinically relevant compounds, which may subsequently cause a clinical phenotype; 2) the frequency of NHE3-799C is significantly higher in a particular patient population with a clinical disorder other than cardiovascular/renal diseases; and 3) NHE3-799C is associated with an alteration in the expression/function of the neighboring genes of NHE3, which may lead to a downstream event that does not involve NHE3. Indeed, the NHE3 gene overlaps with the EXOC3 gene located on the opposing DNA strand. Whereas analysis of GTEx eQTL data showed that NHE-799C was associated with reduced mRNA expression of NHE3, analysis of EXOC3 mRNA levels showed the opposite effect: increased expression was associated with the NHE3-799C variant (data not shown). This issue can be addressed using the human enteroid model (7) derived from donors who are homozygous for NHE3-799C and NHE3-799R, because this model often maintains the genotype and phenotype of the patient from whom the enteroids are produced.

Tenapanor is a novel NHE3 inhibitor characterized by high potency, minimal absorption, and a low risk of systemic side effects. It is administrated orally and binds to NHE3 expressed on the apical surface of intestinal epithelial cells, blocking Na⁺/H⁺ exchange. As a result, NHE3-mediated electroneutral sodium absorption is reduced, which is the molecular basis for the use of tenapanor in the treatment of diseases such as IBS-C (3). This study compared the inhibitory effect of tenapanor on the major form (NHE3-799R) with a common variant (NHE3-799C) of NHE3, and showed that tenapanor dose-dependently inhibited NHE3-799R and NHE3-799C activities in the nanomolar range. Although the IC₅₀ value of tenapanor was about three times higher for NHE3-799C (2.41 ± 0.69 nM) than for NHE3-799R (0.87 ± 0.34 nM), the difference is indeed minor (in the nanomolar range), and therefore is unlikely to be clinically significant. These findings suggest that the clinical effectiveness of tenapanor is likely to be similar in patients presenting with homozygous expression of NHE3-799C and in those with NHE3-799R. Tenapanor binds to the NH₂-terminal transmembrane domain of NHE3; hence, it is possible that NHE3 variants located near the binding site of tenapanor in the NH₂-terminal domain, but not NHE3-799C, which is located in the tail of the COOH-terminal domain, may exhibit differential sensitivity to tenapanor in a rare group of patients.

The present study has a number of limitations. First, the data regarding the frequency of NHE3-799C across different ethnic groups are limited by the small sample size in the populations available in the databases and in our Japanese patient cohort. Further genotyping data from larger and more diverse populations are merited. Second, we used a transfected cell model that lacks endogenous NHE3 expression, whose expression of NHE3 is influenced by several factors. This is likely to have caused at least partly the significant difference in total protein expression between the cell lines stably expressing human NHE3-799C and NHE3-799R. Owing to the use of this transfected cell model, we could not make any conclusion concerning the endogenous protein/mRNA expression of human NHE3-799C and NHE3-799R in enterocytes. This issue can be addressed using human enteroids with endogenous expression of the two NHE3 variants, but it is beyond what we could achieve as part of this study.

In summary, this study demonstrates that NHE3-799C is a nonconservative variant that is enriched in Asian populations but is not functionally compromised. Further studies are needed to show whether the variant has any evolutionary role or clinical significance in a population-specific fashion.

GRANTS

This study was supported in part by National Institute of Diabetes and Digestive and Kidney Diseases Grants RO1-DK-26523, RO1-DK-61765, P01-DK-072084, P30-DK-089502, and R24-DK-99803 and AstraZeneca (Mölndal, Sweden). Ardelyx Inc. (Fremont, CA) partially funded this study (grants to M. Donowitz and C.-M. Tse).

DISCLOSURES

A. Walentinsson is an employee of AstraZeneca. P. J. Greasley is an employee of and has ownership interest in AstraZeneca. M. Donowitz serves on the Scientific Advisory Board of Ardelyx Inc. M. Donowitz and C.-M. Tse have a grant from Adelyx Inc. for collaborative research. J. Yin, B. Cha, R. Sarker, and X. C. Zhu have no conflicts of interest.

AUTHOR CONTRIBUTIONS

J.Y., C.-M.T., X.C.Z., and M.D. conceived and designed research; J.Y., C.-M.T., B.C., R.S., and A.W. performed experiments; J.Y., C.-M.T., B.C., R.S., and A.W. analyzed data; J.Y., C.-M.T., and M.D. interpreted results of experiments; J.Y. prepared figures; J.Y. drafted manuscript; J.Y., C.-M.T., X.C.Z., P.J.G., and M.D. edited and revised manuscript; C.-M.T., P.J.G., and M.D. approved final version of manuscript.