New functions and roles of the Na⁺-H⁺-exchanger NHE3

Abstract

The sodium/proton exchanger isoform 3 (NHE3) is expressed in the intestine and the kidney, where it contributes to hydrogen secretion and sodium (re)absorption. The roles of this transporter have been studied by the use of the respective knockout mice and by using pharmacological inhibitors. Whole-body NHE3 knockout mice suffer from a high mortality rate (with only ~30% of mice surviving into adulthood), and based on the expression of NHE3 in both intestine and kidney, some conclusions that were originally derived were based on this rather complex phenotype. In the last decade, more refined models have been developed that added temporal and spatial control of NHE3 expression. For example, novel mouse models have been developed with a knockout of NHE3 in intestinal epithelial cells, tubule/collecting duct of the kidney, proximal tubule of the kidney, and thick ascending limb of the kidney. These refined models have significantly contributed to our understanding of the role of NHE3 in a tissue/cell type-specific manner. In addition, tenapanor was developed, which is a non-absorbable, intestine-specific NHE3 inhibitor. In rat and human studies, tenapanor lowered intestinal P_i uptake and was effective in lowering plasma P_i levels in patients on hemodialysis. Of note, diarrhea is seen as a side effect of tenapanor (with its indication for the treatment of constipation) and in intestine-specific NHE3 knockout mice; however, effects on plasma P_i were not supported by this mouse model which showed enhanced and not reduced intestinal P_i uptake. Further studies indicated that the gut microbiome in mice lacking intestinal NHE3 resembles an intestinal environment favoring the competitive advantage of inflammophilic over anti-inflammatory species, something similar seen in patients with inflammatory bowel disease. This review will highlight recent developments and summarize newly gained insight from these refined models.

Introduction

Sodium-hydrogen exchangers (NHE) mediate the electroneutral exchange of intracellular H⁺ for extracellular Na⁺ across cellular membranes and are important for (i) Na⁺ and water (re)absorption, (ii) cell volume regulation, and (iii) intracellular pH [50]. The mammalian NHE family is comprised of 9 distinct isoforms (NHE1-9). The major focus of this review is NHE3, which is predominantly expressed in the digestive tract (small intestine > colon) and kidney (proximal tubule and thick ascending limb of the loop of Henle) but is also found to a lesser extent in skin, nervous system, and reproductive tract [50]. Due to its role in Na⁺ (re)absorption, NHE3 plays a critical role in volume homeostasis and regulation of blood pressure. Whole-body NHE3 knockout mice (NHE3^−/−) develop hypovolemia, hypotension, mild metabolic acidosis, and diarrhea [63]. However, it is challenging to unravel the contributions of intestine vs. renal NHE3 in this model. Over the past decade, more refined models utilizing tissue-specific NHE3 knockout mice have allowed for temporal and spatial controls of NHE3 expression, thus providing new insights into the role of intestinal and renal NHE3. In addition, several novel pharmacological tools have been developed to target NHE3 (see summary of drugs discussed in Table 1), which can impact conditions such as hypertension, chronic kidney disease, hyperphosphatemia, and constipation. In this review, we will highlight recent key findings about the role of NHE3 in the kidney and intestine.

Table 1.

Summary of pharmacological tools

Drug	Mechanism of action
Acetazolamide	Carbonic anhydrase inhibitor
AVE0757	NHE3 inhibitor
Empagliflozin	SGLT2 inhibitor
Exendin-4	GLP-1 receptor agonist
Furosemide	NKCC2 inhibitor
HOE642	NHE1 (?) inhibitor
HOE694	NHE1/2 (?) inhibitor
LY3304000	NHE3 inhibitor
LY3358966	Npt2b inhibitor
N3SP	NHE3 stimulating peptide
Phlorizin	Dual SGLT1/2 inhibitor
S3226	NHE3 inhibitor
Tenapanor	Non-absorbable NHE3 inhibitor

New physiological aspects of renal NHE3

The generation of tubule-specific NHE3 knockout mice (NHE3^Pax8Cre), proximal tubule-specific NHE3 knockout (NHE3^Sglt2Cre), as well as thick ascending limb (TAL)-specific NHE3 knockout mice (NHE3^UmodCre) provided novel insight into their physiological functions. One NHE3 function that can be found in all renal physiology textbooks is its role in acid–base regulation; original data in NHE3^−/− mice supported this hypothesis [63, 75], where mice showed metabolic acidosis consistent with impaired HCO₃⁻ (re)absorption. Novel models with NHE3 knockout in the kidney (described above) did not identify any significant changes in blood pH or HCO₃⁻ levels [19, 39, 46, 80], questioning the overall role of renal NHE3 in acid–base homeostasis. Possible compensatory mechanisms in NHE3^Pax8Cre mice include a greater H⁺-ATPase b1-subunit expression and lower pendrin expression in the collecting duct, the latter possibly limiting HCO₃⁻ loss to maintain blood pH [19, 75]. Of note, the use of bafilomycin (inhibitor of H⁺-ATPase) did identify a contribution of this pump to HCO₃⁻ reabsorption in wild-type mice, but no compensatory role was discovered in NHE3^−/− mice [75]. In contrast to control mice, who did not respond to changes in HCO₃⁻ excretion in response to the NHE2 inhibitor HOE694, NHE3^−/− mice showed an increase in Na⁺ and HCO₃⁻ excretion, an effect described to be caused by NHE2 inhibition [1]. However, based on the lack of specificity of this compound (NHE1 vs. NHE2), it remains unclear which, and if, another isoform can explain this finding. Consistent with this, NHE2/3 double knockout mice did not show acid–base disturbances that were greater compared to those found in NHE3^−/− mice [37]. A micropuncture study did not find that fluid or Na⁺ reabsorption in the proximal convoluted tubule was affected by the NHE1 inhibitor HOE642 [71]. In NHE3^UmodCre mice, no differences in renal NHE1 expression were found compared to control mice [80].

A more specific inhibitor of NHE3, S3226, was used to determine the contribution of NHE3 in the kidney on fluid, electrolyte, and HCO₃⁻ reabsorption. In micropuncture studies, proximal tubule fluid and Na⁺ reabsorption was reduced by ~30% [71] and HCO₃⁻ by ~45% [74]. It is important to note that these two studies have conflicting results when it comes to transport in the TAL. The study by Vallon et al. [71] did not find changes in fluid, Na⁺ or K⁺ reabsorption along the TAL; however, the study by Wang et al. [74] showed that fluid and HCO₃⁻ reabsorption was reduced by ~40% during micro perfusion of the loop of Henle. The reason(s) for these differences remain elusive, but it is highly unlikely that a concentration difference of 10 μM (30 vs. 40 μM) can account for these opposing results.

Additional information possibly explaining some of these findings regarding the role of NHE3 in acid–base homeostasis comes from intestinal epithelial cell-specific NHE3 knockout mice (NHE3^{VillinCreERT2}, see “Role of intestinal NHE3 in diarrheal disease” section). New insight was also provided regarding the sex-dependent expression of NHE3. In the proximal tubule of female Sprague Dawley rats, phosphorylation of NHE3 was greater compared to male rats, and NHE3 showed a greater distribution at the base of the microvilli. Of note, NHE3 protein expression and phosphorylation in the cortex were both significantly reduced (~40%) in females compared to male C57BL/6 mice [73]. By using mice with kidney-specific knockout of the androgen receptor [24], a possible determinant of these sex differences has been discovered. In this model, NHE3 expression was lower in male mice but was unaffected in female mice. In conjunction with the changes observed in electrogenic Na⁺-bicarbonate cotransporter 1A, this could be a mechanism to limit the magnitude of the effect on ammonia excretion [24].

Another important role of renal NHE3 relates to Na⁺ homeostasis. Studies employing dietary Na⁺ challenges found that NHE3 is required to maintain steady-state plasma Na⁺ levels in response to changes in NaCl intake [19]. Despite a concerted interplay of lower glomerular filtration rate (GFR) and greater activation of distal tubule Na⁺-conserving mechanisms (Na⁺/Cl⁻ cotransporter and epithelial Na⁺ channel), plasma Na⁺ in NHE3^Pax8Cre mice was directly impacted by dietary NaCl. An additional role of NHE3 relates to urinary concentration. NHE3^Pax8Cre mice show greater fluid intake and greater urinary flow rate under baseline conditions in conjunction with significantly lower urine osmolality [19]. When fed a low NaCl diet, 18-h water deprivation unraveled that NHE3^Pax8Cre mice have a problem conserving fluid, consequently showing lower urine osmolality and a greater increase in plasma osmolality. The generation of NHE3^UmodCre mice expanded this knowledge [80]. Mice lacking NHE3 in the thick ascending limb show a mild urinary concentrating defect under baseline conditions (trend for greater fluid intake and lower urine osmolality); however, when challenged by water deprivation, no notable differences were found between genotypes (Fig. 1 and Table 2). Taken together, these data indicate that NHE3 in the medullary thick ascending limb may contribute to some extent toward establishing the medullary interstitial gradient. Of note, the lack of NHE3 in the thick ascending limb does not affect urine pH. A novel role of NHE3 for furosemide-induced urinary acidification has recently been described [7], in addition to epithelial Na⁺ channel-induced activation of vacuolar H⁺-ATPase (B1 subunit) in the connecting tubule [20, 34], accounts for urinary acidification in response to furosemide. Further studies in NHE3^UmodCre mice [80] showed that the rapid effect of furosemide on urinary acidification was not affected in NHE3^UmodCre mice; however, in contrast to control mice, the observed acidification could not be sustained. When furosemide was given chronically for 8 days, control mice showed metabolic alkalosis and hypokalemia. In contrast, these effects did not develop in NHE3^UmodCre mice. Therefore, NHE3 in the thick ascending limb is important in order to achieve sustained urinary acidification.

Fig. 1.

Model of renal NHE3 functions. The expression of NHE3 in the proximal tubule of the nephron is shown in yellow; the expression of NHE3 in the thick ascending limb is shown in red. Right—a model of a proximal tubule cell depicting NHE3 actions and interactions. Glucagon-like peptide 1 (GLP-1) and exendin-4 (EX-4) activate G_s protein-coupled GLP-1 receptors (GLP-1R). While there is evidence that phosphorylation of NHE3 is affected by these agonists, data in NHE3^Pax8Cre mice indicate that NHE3 is not the protein mediating the diuretic and natriuretic effect [59]. Adenosine A₁ receptors (A₁R) are G_i protein-coupled receptors that mediate diuresis and natriuresis in response to caffeine (a nonselective adenosine receptor antagonist) [61]. The inhibitory action of A₁R results in increased cAMP levels consequently inhibiting NHE3. Of note, NHE3^Pax8Cre mice provide evidence that NHE3 is dispensable for these responses and preliminary evidence suggests that these effects could be mediated by NBCe1-A [18]. The groups of gliflozins inhibit sodium-glucose cotransporters and, via a largely unknown mechanism, NHE3 activity, consequently resulting in higher urinary pH and HCO₃⁻ excretion [47]. Lithium is used as a marker of proximal tubule reabsorption; consequently, Li⁺ clearance should be greater in NHE3^Pax8Cre mice. Of note, Li⁺ disposition in NHE3^Pax8Cre mice was unaffected, implying that other transport processes are responsible for Li⁺ transport [69]. Blood pressure effects in the proximal tubule in response to Ang II are mediated by NHE3 and were shown to involve cAMP, Ca²⁺, and IRBIT (inositol 1,4,5-trisphosphate receptor-binding protein released with inositol 1,4,5-trisphosphate) [28, 40]. The NHE3 inhibitor AVE0657 significantly attenuated Ang II-induced hypertension [40]. Left—a model of a thick ascending limb cell depicting NHE3 actions and interactions. Despite the quantitatively high expression of NHE3 in the thick ascending limb, we are just beginning to understand its function in the segment. Knockout of NHE3 in the thick ascending limb does not affect acid–base or electrolyte homeostasis [80]. Of note, furosemide induces urinary acidification, which, at least in part, is mediated by NHE3 in the thick ascending limb. NHE3^UmodCre mice provided evidence that the acute urinary acidification was intact, but the sustained acidification was absent. The hypothesis was brought forward that blocking NKCC2 will provide a driving force for Na⁺ uptake via NHE3, which will subsequently lead to acidification of tubular fluid and intracellular alkalinization [7]. Acute and chronic furosemide administration drastically increases urinary Ca²⁺ excretion. Regarding the latter, the lack of NHE3 in the thick ascending limb significantly attenuated furosemide-induced urinary Ca²⁺ excretion. For simplicity, not all signaling pathways are shown and some receptors may have luminal and basolateral localizations

Table 2.

Comparison between genetically modified NHE3 models

	NHE3−/−	NHE3Pax8Cre	NHE3Sglt2Cre	NHE3UmodCre	NHE3VillinCreERT2
Blood pH	↓	↔	↔	↔	↓
Urine pH	ND	↑	ND	↔	↔
Blood pressure	↓	↓	↓	ND	ND
Urine osmolality	↓	↓	↔	↓	↔
Plasma Na+	↔	↔	↔	↔	↓
Na+ excretion	↑	↑	↑	↔	↓

This table shows the impact of organ/tissue-specific knockout on different physiological parameters. NHE3^−/− = NHE3 whole-body knockout, NHE3^Pax8Cre = NHE3 knockout along the nephron, NHE3^Sglt2Cre = NHE3 knockout in the S1/2 segment of the proximal tubule, NHE3^UmodCre = NHE3 knockout in the thick ascending limb, NHE3^{VillinCreERT2} = inducible NHE3 in intestinal epithelial cells. The downward arrow (↓) indicates lower compared to control animals, the upward arrow (↑) indicates higher compared to control animals, and the left–right arrow (↔)indicates unchanged compared to control animals. ND = not determined

Renal NHE3 and blood pressure regulation

Given the importance of NHE3 in Na⁺ reabsorption, a role in blood pressure regulation is conceivable. Consistent with this, NHE3^Pax8Cre mice have lower systolic blood pressure [42, 63]; however, blood pressure regulation remained salt-sensitive [19]. Another novel insight comes from NHE3^Sglt2Cre mice (Fig. 1 and Table 2). Blood pressure measurements using radiotelemetry showed lower systolic, diastolic, and mean arterial blood pressure (~15 mmHg) in both sexes in mice lacking NHE3 in the proximal tubule [38–40]. Along those lines, the pressure-natriuresis response was significantly enhanced in NHE3^Sglt2Cre mice in response to similar increases in renal perfusion pressure. Angiotensin II (Ang II) is a well-known regulator of blood pressure, which is mediated, at least in part, by angiotensin 1a receptors in the proximal tubule [23]. Administration of Ang II via osmotic minipumps to control and NHE3^Sglt2Cre mice showed that both genotypes developed a pressure response; however, in NHE3^Sglt2Cre mice, this was attenuated by ~40 mmHg. By using a pharmacological NHE3 inhibitor (AVE0757), a similar pattern could be shown and the Ang II-induced increase in blood pressure could be attenuated by ~20 mmHg. Taken together, it remains to be debated whether pharmacological inhibition of NHE3 can be a viable treatment option for hypertension.

Role of renal NHE3 for caffeine-induced diuresis/natriuresis

Another interesting pharmacological aspect relates to the diuretic/natriuretic effects of caffeine (Fig. 1). In contrast to the common assumption, caffeine does not inhibit phosphodiesterases as the main mechanism of action, which only occurs under toxicological concentrations [21]. In concentrations normally observed with coffee or tea consumption, the predominant mechanism of caffeine is antagonism of adenosine receptors. Functionally, adenosine A₁ receptors have been described in the proximal tubule [77], and studies have shown that activation of adenosine A₁ receptors inhibits NHE3 activity [9]. Of note, the diuretic effect of caffeine and theophylline requires intact adenosine A₁ receptors to occur [60, 61]. Based on these findings, it would have been a reasonable assumption that the diuretic/natriuretic effects are mediated by renal NHE3. In contrast to this assumption, NHE3^Pax8Cre mice had similar diuretic/natriuretic caffeine responses compared to control mice, indicating that renal NHE3 is dispensable for caffeine-induced diuresis/natriuresis [18]. Possibly, other transporters, including but not limited to, NBCe1, might mediate renal caffeine effects. Despite these findings, RNA sequencing data and proteomics data show low to absent expression of adenosine A₁ receptors in the proximal tubule [41, 54], indicating that further studies are needed to better understand caffeine actions in the kidney.

Lack of renal NHE3 does not affect lithium pharmacokinetics

Over decades Li⁺ has been used as a measure of proximal tubule reabsorption based on the assumption that Li⁺ and Na⁺ transport go in parallel in the proximal tubule and that Li⁺ handling in further downstream nephron segments is minimal (< 25%). Administration of acetazolamide, a carbonic anhydrase inhibitor, caused fractional Li⁺ clearance to increase ~2-fold in rats, consistent with inhibition of proximal tubular reabsorption [70]. Mechanistically, it has been proposed that carbonic anhydrase II binds to the N-terminus of NHE3 and activates it; indeed, inhibition of carbonic anhydrase II with acetazolamide reduced NHE3 activity [35]. Of note, when Li⁺ pharmacokinetics was studied in NHE3^Pax8Cre mice, no differences were observed in Li⁺ bioavailability, half-life, maximum plasma concentrations, area under the curve, Li⁺ clearance, or urinary Li⁺/creatinine ratios compared to control mice [69]. Clearly, further studies are needed to determine how Li⁺ is handled along the nephron and factors that impact that, e.g. dietary P_i [69].

Renal NHE3 mediates Na⁺ retention in nephrotic syndrome

Nephrotic syndrome is characterized by protein wasting and edema; however, the changes in renal transporters and channels that contribute to the occurrence of edema show a time dependence. At the early activation phase (2 days) of nephrotic syndrome, NHE3 protein expression was ~2-fold greater compared to control mice [10], whereas at the later retention phase (3 days), the difference disappeared. In the subsequent steady-state phase (5 days), NHE3 protein expression was reduced by 50%. In a rat model of nephrotic syndrome, NHE3 protein expression was comparable after 6 days, but was ~2-fold greater after 12 days compared to the control group. However, after 18 days any significant differences disappeared [78]. The greater expression also corresponded to a ~2-fold greater NHE3 activity in nephrotic rats [2]. Taken together, these data indicate that greater renal NHE3 expression and activity play a critical role in the development of edema. Of note, the contribution of Na⁺ transport in further downstream nephron segments cannot be neglected to contribute to this process.

Role of renal NHE3 in acute kidney injury

Acute kidney injury (AKI) is a multifactorial disease that, despite our progress in understanding its entity, still has a high mortality rate. By using the NHE3 inhibitor S3226, it has been shown that ischemia-reperfusion injury-induced AKI can be improved as evidenced by a smaller decline in GFR and better kidney function/structure [30]. Of note, our own unpublished observations show that S3226 has many non-specific systemic effects when administered to mice lacking renal NHE3. In a different AKI model, ischemia-reperfusion injury of the lower limb, GFR was reduced by 98% (nearly complete loss of kidney function), and blood urea nitrogen (BUN) increased by 825% in control mice [48, 49]. In contrast, NHE3^KspCadCre mice were protected and showed only a 25% decrease in GFR and a 105% increase in BUN. Further studies in LLC-PK1 (kidney cell line of a male pig) suggested that preservation of mitochondrial biogenesis might be one of the mechanisms [49]. Of note, the Ksp-cadherin cre mouse model targets the thick ascending limb, distal nephron, and collecting duct [64], suggesting that NHE3 in the thick ascending limb is mediating the observed beneficial effects in AKI. By using different models of AKI in rats (cisplatin, volume depletion + Ang II receptor blockade and ischemia-reperfusion injury) it could be demonstrated that the expression of renal NHE3 as well as the abundance of NHE3 in urinary exosomes increased in response to injury. Similarly, in patients with sepsis-induced AKI, a ~10-fold increase of NHE3 in urinary exosomes was present [83]. These data show that NHE3 could be a promising tool for the diagnosis, course, treatment efficacy, and potential outcomes in human AKI.

Role of NHE3 in mediating pleiotropic effects of gliflozins

Drugs targeting sodium-glucose co-transporters (“gliflozins”) are a new mainstay in the treatment of diabetes mellitus and heart failure [12, 13, 55, 62]. In addition to their glucosuric effect, other pleiotropic effects include reductions in blood pressure, body weight, and others [6, 8, 13, 72]. Evidence is accumulating that some of these effects are mediated by coordinated transport processes in the proximal tubule involving NHE3, the latter being the cause for inhibition of proximal tubular Na⁺ reabsorption [42, 71]. Consistent with this idea, experiments in NHE3^Pax8Cre mice identified that tubular NHE3 is required for the acute natriuretic effect of empagliflozin and for the effect on volume status when empagliflozin was administered chronically [46] (Fig. 1). In humans, a novel formula employing multi-ion urine anion gap calculation showed that empagliflozin increased the gap compared to placebo, providing indirect evidence for NHE3 mediating this process [58]. Despite NHE3^−/− mice showing metabolic acidosis [63, 75], other models with the knockout of NHE3 along different parts of the nephron do not show metabolic acidosis [19, 46, 47, 80]. Consistent with these studies in mice, SGLT2-mediated inhibition of NHE3 did not cause differences in venous blood pH between patients treated with empagliflozin or placebo [58]. In contrast, urinary pH was significantly more alkaline in NHE3^Pax8Cre mice [18, 19, 46, 47], and a study in healthy young volunteers showed that urinary pH levels increased after empagliflozin treatment [3]. Of note, differences exist between these two human studies in terms of the effects of empagliflozin on NHE3 expression. The study by Rao et al. [58] showed that 14 days after treatment, the expression of NHE3 in urinary extracellular vesicles was reduced compared to placebo treatment; on the other hand, the study by Biancalana et al, [3] did not see differences in mRNA or protein expression/phosphorylation of NHE3 in exfoliated tubular cells after 6 h of treatment. Whether this is a time-dependent effect or due to the substantial variability between samples in the latter study remains to be determined. Of note, no changes in NHE3 expression or phosphorylation were observed in mice treated with empagliflozin for 15 weeks [47]; however, NHE3 phosphorylation was significantly higher in Akita mice (a type 1 diabetes mellitus model) chronically treated with empagliflozin, possibly indicating a functional NHE3 inhibition by empagliflozin. The fact that higher NHE3 phosphorylation was observed in diabetic models treated with SGLT2 inhibitors [44, 47] but not in non-diabetic conditions [44, 47], implies that diabetes mellitus changes the functional properties of NHE3. Recently, other factors influencing expression patterns have been discovered, e.g., NHE3 plays a permissive role in SGLT2 expression [47]. By using the originally discovered compound phlorizin (a dual SGLT1/2 inhibitor) in micropuncture experiments, insight was provided for the existence of a functional interaction between SGLTs and NHE3 in the proximal tubule at the single nephron level [53]. Additional regulation of expression has been found to be controlled by clock genes, which regulate circadian expression patterns of proteins. Noteworthy is that the circadian clock protein period (Per)1 regulates the circadian expression of SGLT1 in the kidney but not the circadian expression of SGLT2 [65].

Role of intestinal NHE3 in diarrheal disease

The role of NHE3 in the intestine has been studied for decades and has been well characterized as the major mechanism by which Na⁺ is absorbed from the gut lumen. This, in combination with the accompanying trans/paracellular water absorption, has implicated that intestinal NHE3 plays a critical role in body fluid and blood pressure homeostasis [50]. Similarly, inhibition or lack of NHE3 in the intestine has been linked to diarrheal disease due to the effect of increasing luminal Na⁺ levels. In humans, mutations in NHE3 [4, 29] or in signaling molecules regulating NHE3 activity cause congenital sodium diarrhea (CSD), characterized by intrauterine onset of diarrhea, hyponatremia, severe dehydration, and metabolic acidosis [31, 45]. The generation of NHE3^−/− mice confirmed these findings; mice exhibited distended abdomens, fluid-filled intestinal loops, and alkaline diarrhea [63]. To exclude the contribution of renal NHE3 to the intestinal phenotype observed in whole-body NHE3 knockout mice, we generated intestinal epithelial cell-specific NHE3 knockout mice by using a non-inducible villin-Cre driver. Unfortunately, this model had high mortality, with only one out of > 50 offspring surviving ~2 weeks [11]. Therefore, we generated inducible intestinal epithelial cell-specific NHE3 knockout mice (NHE3^{VillinCreERT2}) by using a tamoxifen-inducible villin-Cre driver, which resulted in ~25% mortality rate 3 weeks after induction [14]. Of note, no Cre expression is observed in the kidney (despite villin being normally expressed in the kidney) of these villin-Cre mice possibly due to certain chromosomal environments [43]. We were able to demonstrate that knockout of intestinal NHE3 in adulthood mimics human CSD [82]; mice had persistent alkaline diarrhea, hyponatremia, and metabolic acidosis, in conjunction with renin-angiotensin-aldosterone system activation and volume contraction. NHE3 plays an important role in intestinal structural integrity, as NHE3^{VillinCreERT2} mice had increased intestinal permeability and epithelial apoptosis, and ~45% of the mice developed signs of cryptitis [82]. To determine if there were differences in intestinal transit time, we fasted control and NHE3^{VillinCreERT2} mice overnight and then gave them an oral gavage with 5% lissamine green solution in water (1% of body weight). After 5 min, intestines were harvested, and the distance from the pylorus to the ileocecal valve was measured, along with the distance traveled by the blue/green dye front (data expressed as percent of intestinal segment length). We found that compared with control mice, NHE3^{VillinCreERT2} mice had a ~20% significantly faster intestinal transit time (Fig. 2).

Fig. 2.

Intestine-specific NHE3 knockout increases intestinal motility. Five days after tamoxifen administration, mice were administered lissamine green via oral gavage and euthanized 5 min after administration. Motility was significantly greater in NHE3^{VillinCreERT2} mice compared with control mice. In addition to the single data summary, data are shown and are expressed as mean ± SEM and were analyzed by Student’s t-test. *P < 0.05 versus control

With studies providing evidence that lack of NHE3 increases intestinal luminal fluidity, it is not surprising that NHE3 became a pharmacological target for constipation in 2019 (Fig. 3). Tenapanor (Ibsrela^®) is a non-absorbable, intestinal-specific, NHE3 inhibitor that has been used to successfully treat irritable bowel syndrome with constipation in adults [5]. Similarly, patients with cystic fibrosis are at risk for constipation and intestinal obstructions and could benefit from treatments that increase intestinal content fluidity. Recently, it was shown that administration of tenapanor reduced intestinal fluid absorption and increased alkaline output in mice that either lack the cystic fibrosis transmembrane regulator (CFTR) protein or express the most frequent mutant CFTR protein (F508del) [67] and could prevent intestinal obstructions in CFTR knockout mice [68].

Fig. 3.

Model for intestinal NHE3 function. Intestinal NHE3 is the major contributor to Na⁺ reabsorption in the gut. Studies using intestinal epithelial cell–specific NHE3 knockout mice have helped unravel that lack of NHE3 leads to alkaline diarrhea, hyponatremia, and metabolic acidosis [82]. In addition, the gut microbiome shifts in favor of more inflammophilic species [15, 79]. Infectious diarrhea caused by Vibrio cholera toxin (CT), enterotoxigenic E. coli heat-stable enterotoxin (STa), and C. difficile toxin B (TcdB) all lead to increases in cGMP, cAMP, and the activation of PKA that consequently inhibits NHE3 activity. Of note, adenylyl cyclase isoform 6 (AC6) is essential for CT-induced increases in cAMP and CFTR activity [17]. Enteropathogenic E. coli (EPEC), while not secreting a classical toxin, stimulates the PKA-dependent E3 ubiquitin ligase Nedd4-2, leading to the internalization of NHE3. In addition, EPEC inhibits the activity and surface expression of the chloride/bicarbonate exchanger SLC26A3 (downregulated in adenoma, DRA) [22], which together with inhibition of NHE3 leads to osmotic diarrhea. The novel NHE3-stimulating peptide (N3SP) partially reverses cAMP/cGMP-dependent inhibition of NHE3 and may lessen intestinal fluid loss during infectious diarrhea [84]. Along those lines, the pharmacological NHE3 inhibitors tenapanor and LY3304000 inhibit intestinal P_i uptake [33, 76]. For tenapanor, this was shown to occur via inhibiting paracellular P_i absorption. Of note, this effect might be species dependent considering that different effects between mice, rats, and humans are observed [51, 76, 81]. In mice, knockout or inhibition of NHE3 results in increased Npt2b expression [81] and enhanced P_i uptake [76, 81]. Combining LY3304000 with the Npt2b inhibitor LY3358966 in rats showed additive effects and resulted in a significantly smaller P_i uptake compared to respective single drug treatments [76]. For simplicity, not all signaling pathways are shown

Role of intestinal NHE3 in infectious diarrhea

Diarrhea caused by gastrointestinal infection with microorganisms such as Vibrio cholerae, Clostridium difficile, and various pathogenic strains of Escherichia coli has been shown to be associated with impaired Na⁺ and fluid absorption. For example, cholera toxin (Fig. 3) increases cyclic adenosine monophosphate (cAMP) which activates protein kinase A (PKA) leading to phosphorylation and inhibition of NHE3 activity [66]. Similarly, the C. difficile toxin B decreases NHE3 activity via a Rho-GTPase-sensitive redistribution of NHE3 away from the apical membrane consequently altering the gut microbiome such that the environment favors colonization by C. difficile [16, 26]. Recently, a study confirmed that NHE3 is decreased in a C. difficile toxin B-dependent manner; however, they did not see changes in the cytoskeletal protein group ezrin/radixin/moesin (ERM) that is linked to the NHE regulatory factor (NHERF) family proteins that help anchor NHE3 to the apical membrane, suggesting there may be another mechanism by which NHE3 is depleted during C. difficile infection [52]. The enterotoxigenic E. coli (ETEC) heat-stable enterotoxin STa has been shown to inhibit NHE3 via both cAMP-dependent PKA and protein type II cGMP-dependent kinase (PKG II) activity [27, 45]. In contrast, enteropathogenic E. coli (EPEC) inhibits NHE3 without secreting an enterotoxin, although the signaling mechanism has still not been fully elucidated. Recently, it was shown that inhibition of human NHE3 is dependent on the PKA-dependent E3 ubiquitin ligase Nedd4-2, which ubiquitinates and facilitates internalization of NHE3 (Fig. 3); consequently, Nedd4-2 potentiates intestinal fluidity and may exacerbate the severity of diarrhea from cholera and EPEC [32]. A novel pharmacological peptide, N3SP, stimulates NHE3 activity and was shown to partially reverse cAMP/cGMP-dependent inhibition of NHE3 in vitro, preventing cholera toxin- and ETEC-induced fluid secretion in intestinal loops of mice [84].

Functional intestinal NHE3 is necessary for a healthy gut microbiome

It has become clear that changes to the ionic composition and luminal pH in the intestine can directly affect the composition of the gut microbiome. Several studies in humans and mice showed that NHE3 deficiency can predispose patients to microbial dysbiosis and the development of inflammatory bowel disease [25, 56]. In an adoptive T cell transfer colitis model, NHE3 status was the most significant determinant of the gut microbial community [36]. Further, in studies where germ-free mice underwent fecal microbiota transfer from NHE3-deficient mice, it was shown that there was accelerated onset and severity of inflammation, demonstrating a causative relationship between NHE3-induced dysbiosis and colitis [25]. Our studies in inducible NHE3^{VillinCreERT2} mice expanded on this and demonstrated that intestinal NHE3 is required for a healthy microbiome; lack of intestinal NHE3 created an intestinal microenvironment that favored the competitive advantage of inflammophilic over anti-inflammatory species [79]. Given the recognized role of inflammation and microbial dysbiosis in the initiation and progression of colorectal cancer, it is important to study the role of NHE3 in these processes. Laubitz et al. [36] found that NHE3 expression is reduced in patients with sporadic colorectal cancer and that lack of NHE3 accelerated the development of colorectal cancer and increased tumor burden in a novel NHE3-deficient multiple intestinal neoplasia mouse model. Importantly, when there was loss of NHE3 in the surface epithelium of colonic tumors, there was increased proliferation, decreased apoptosis, and reduced adhesion to the extracellular matrix proteins, thereby promoting cancer progression.

Role of NHE3 in intestinal phosphate (P_i) absorption

Tenapanor not only inhibits Na⁺ absorption mediated via NHE3 but also secondarily inhibits P_i absorption. In clinical studies, tenapanor reduced plasma P_i in patients on hemodialysis with hyperphosphatemia treated with or without P_i binders. This was accompanied by reduced fibroblast growth factor 23 (FGF23), a major hormone contributing to the development of left ventricular hypertrophy in chronic kidney disease (CKD). Based on these findings, the use of tenapanor (XPHOZAH^®) for the control of plasma P_i in adult patients with CKD on dialysis was recently granted US regulatory approval by the Federal Food and Drug Administration in October 2023. Despite this approval, several contradictory data exist regarding its mode of action. This discussion has spurred new research regarding the mechanism(s) of how intestinal P_i is absorbed. The current working hypothesis is that tenapanor reduces intestinal P_i absorption by inhibition of paracellular P_i transport, and not via the transcellular pathway [33]. This was somewhat surprising considering that the intestinal P_i transporter Npt2b was thought to be responsible for the majority of intestinal P_i absorption. To better understand the role of intestinal NHE3 in P_i homeostasis, we performed studies in NHE3^{VillinCreERT2} mice. Surprisingly, in contrast to studies using tenapanor in humans and rats, we found that lack of intestinal NHE3 enhanced, rather than inhibited, intestinal P_i absorption which was associated with increased intestinal Npt2b expression [81]. Along the same line, a recent study showed that a novel NHE3 inhibitor, LY3304000 (Fig. 3), robustly increased intestinal P_i absorption in mice [76]. In contrast, LY3304000 modestly inhibited intestinal P_i absorption in rats, and there was a synergistic effect on the inhibition of P_i absorption when the NHE3 inhibitor was combined with an Npt2b inhibitor LY3358966 [76]. Taken together, this suggests that species differences might exist between mice, rats, and humans and warrants further studies to explore why these differences occur.

Future directions

Significant advances have been made regarding the role of NHE3 by using tissue/cell type-specific animal models. Despite these strides, important questions remain unanswered. For example, why do kidney-specific NHE3 knockout mouse models not show systemic acid–base disturbances? In contrast, intestine epithelial-specific NHE3 knockout mice show overt metabolic acidosis as a consequence of diarrhea (intestinal loss). The data are consistent with NHE3 playing a role in renal HCO₃⁻ reabsorption, but knockout in the kidney can normally be compensated by other transporters and other nephron segments. Alternatively, as originally proposed by Preisig et al. [57], an additional mechanism that significantly contributes to H⁺ secretion (Na⁺-independent and amiloride-insensitive, e.g., H⁺-ATPase) might be contributing to maintaining acid–base homeostasis. Regarding the role of NHE3 in the intestine, several questions remain unanswered. We need a better understanding of the mechanism(s) on how tenapanor affects intestinal P_i transport as well as the quantitative contribution of paracellular vs. transcellular P_i transport. Along those lines, further studies are needed to determine if paracellular transport can be regulated in a hormonal/paracrine manner, and the role of NHE3 in this process. Of note, despite diarrhea being a major side effect of tenapanor in humans (16% vs. 4% placebo), no long-term clinical studies have focused on how NHE3 inhibition can result in changes in the composition of the gut microbiome. These effects need to be considered, particularly in the context of a potential faster progression of colorectal cancer.

Data Availability

No datasets were generated or analysed during the current study.