NHE3 in the thick ascending limb is required for sustained but not acute furosemide-induced urinary acidification

Abstract

Na⁺/H⁺ exchanger isoform 3 (NHE3) facilitates Na⁺ reabsorption and H⁺ secretion by the kidneys. Despite stronger NHE3 abundance in the thick ascending limb (TAL) compared with the S1 and S2 segments of the proximal tubule, the role of NHE3 in the TAL is poorly understood. To investigate the role of NHE3 in the TAL, we generated and phenotyped TAL-specific NHE3 knockout (NHE3^TAL-KO) mice. Compared with control mice, NHE3^TAL-KO mice did not show significant differences in body weight, blood pH, or plasma Na⁺, K⁺, or Cl⁻ levels. Fluid intake trended to be higher and urine osmolality was significantly lower in NHE3^TAL-KO mice. Despite a similar glomerular filtration rate, NHE3^TAL-KO mice had a greater urinary K⁺-to-creatinine ratio. One proposed role of NHE3 relates to furosemide-induced urinary acidification. Acute bolus treatment with furosemide under anesthesia did not result in differences in the dose dependence of urinary flow rate, Cl⁻ excretion, or maximal urinary acidification between genotypes; however, in contrast with control mice, urinary pH returned immediately toward baseline levels in NHE3^TAL-KO mice. Chronic furosemide treatment reduced urine osmolality similarly in both genotypes but metabolic alkalosis, hypokalemia, and calciuresis were absent in NHE3^TAL-KO mice. Compared with vehicle, chronic furosemide treatment resulted in greater Na⁺-K⁺-2Cl⁻ abundance regardless of genotype. Na⁺-phosphate cotransporter 2a abundance was also greater in furosemide-treated control mice compared with vehicle treatment, an effect that was absent in NHE3^TAL-KO mice. In summary, NHE3 in the TAL plays a role in the sustained acidification effect of furosemide. Consistent with this, long-term treatment with furosemide did not result in metabolic alkalosis in NHE3^TAL-KO mice.

NEW & NOTEWORTHY Na⁺/H⁺ exchanger isoform 3 (NHE3) is very abundant in the thick ascending limb (TAL) compared with the proximal tubule. Much has been learned about the role of NHE3 in the proximal tubule; however, the function of NHE3 in the TAL remains elusive. A novel mouse model that lacks NHE3 selectively in the TAL not only shows a phenotype under baseline conditions but also identifies that NHE3 is required for sustained but not acute furosemide-induced urinary acidification.

INTRODUCTION

Na⁺/H⁺ exchanger isoform 3 (NHE3) is important for Na⁺ absorption in the intestine and kidney. In the intestine, apical NHE3 is critical for Na⁺ and fluid absorption (1), and knockout of NHE3 selectively in the small intestine and colon results in persistent diarrhea, increased mortality rate, metabolic acidosis, lower blood ${\text{HCO}}_3^{-}$ levels, hyponatremia, and hyperkalemia associated with drastically elevated plasma aldosterone levels and changes to intestinal structural integrity (2).

In the kidney, NHE3 is expressed in the apical membrane of the S1 and S2 segments of the proximal tubule (3, 4) and thick ascending limb (TAL). In the proximal tubule, NHE3 facilitates the majority of renal Na⁺ reabsorption (60%–75%) and H⁺ secretion (1, 3, 5). Ex vivo perfusion studies have shown that proximal tubule Na⁺ and ${\text{HCO}}_3^{-}$ reabsorption were reduced by ∼60% in conventional NHE3-knockout mice compared with wild-type mice. In rat micropuncture experiments, in vivo perfusion with an NHE3 inhibitor (S3226) demonstrated that the maximum reduction of Na⁺ reabsorption was only ∼30% in superficial nephrons accessible by micropuncture (6). Renal NHE3 contributes to the NaCl sensitivity of blood pressure, and studies have identified that dietary NaCl alterations directly impact total body NaCl levels by disturbing glomerulotubular balance (7). Of note, knockout of NHE3 along the nephron did not affect blood pH or ${\text{HCO}}_3^{-}$ (7).

Immunostaining of NHE3 in the TAL was qualitatively stronger compared with the proximal tubule (3, 4), suggesting that the abundance of NHE3 in this segment is greater. Supporting this, RNA-sequencing analysis from microdissected kidney tubule segments demonstrated that NHE3 mRNA is quantitatively greater in the TAL relative to the proximal tubule of rats and mice (8–10); however, protein copy numbers per cell showed the opposite (10, 11). Functionally, the role of NHE3 in the TAL is less clearly defined. Hormonal regulation of NHE3 in the TAL by arginine-vasopressin (AVP) has been described, with quantitative phosphoproteomics demonstrating AVP-dependent phosphorylation of NHE3 at serine residue 552 (12), which is an inhibitory phosphorylation site (13, 14). Furthermore, administration of angiotensin II to rats caused a significant increase in NHE3 abundance in the inner stripe of the outer medulla, a region where the medullary TAL, but not the proximal tubule, is found (15). NHE3 in the TAL contributes to acid-base regulation (16), in particular ${\text{HCO}}_3^{-}$ reabsorption (17). More recently, a new mechanism for the role of NHE3 in the TAL during furosemide-induced urinary acidification was proposed (18). Here, furosemide inhibition of Na⁺-K⁺-2Cl⁻ cotransporter (NKCC2) in the TAL increases the driving force for Na⁺ reabsorption by NHE3, which subsequently results in increased H⁺ secretion and a more acidic tubular fluid and urine.

To increase our understanding of the role of NHE3 in the TAL, we generated a TAL-specific NHE3-knockout (NHE3^TAL-KO) mouse model by crossing mice with tamoxifen-inducible Cre recombinase expression under the regulation of the uromodulin gene (encoding Tamm–Horsfall protein) with floxed NHE3 mice (19). We subsequently used this model to study the isolated role of NHE3 in the TAL of adult mice for electrolyte and acid-base homeostasis. Second, we used the model to investigate in vivo whether NHE3 in the TAL is required for urinary acidification in response to furosemide. Our results demonstrate that NHE3 in the TAL is important for a sustained but not acute acidification effect of furosemide.

MATERIALS AND METHODS

Animals

All animal experimentation was conducted in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals (National Institute of Health, Bethesda, MD) and was approved by the University of South Florida Institutional Animal Care and Use Committee (3338 R). Floxed NHE3 (NHE3^loxlox) female mice (19) were crossed with hemizygous male C57BL/6N-Umod^{em1(cre/ERT2)Amc}/J mice (Stock No. 030601, Jackson Laboratory, Bar Harbor, ME) that express tamoxifen-inducible Cre recombinase under the control of the uromodulin gene, which is expressed exclusively in the TAL and early distal tubule (20). From the F₁ progeny, male mice heterozygous for the floxed NHE3 allele (NHE3^{lox/wild-type}) and Umod^{em1(cre/ERT2)Amc}/J were bred to female NHE3^lox/lox mice to generate final breeder pairs in the F₂ progeny. To generate male experimental NHE3^TAL-KO and NHE3^lox/lox (control) mice, female NHE3^lox/lox mice were bred to male NHE3^TAL-KO mice. Mice were genotyped by PCR from genomic DNA isolated from ear punch. Genotyping for the floxed allele was as previously described (4); TAL-Cre/ER^T2 genotyping was performed according to Protocol No. 29530 (Jackson Laboratory). NHE3 deletion was induced by tamoxifen (200 mg·kg⁻¹ body wt), which was initially dissolved in 5% (vol/vol) of ethanol followed by adding 95% (vol/vol) of corn oil and given by oral gavage (volume: 1% of body wt) for three consecutive days. An equal dose of tamoxifen was given to control and NHE3^TAL-KO mice. Littermate age-matched, 3- to 6-mo-old mice were used for experiments. Male mice were used except where indicated in Fig. 3. All acute physiological experiments were performed between 0900 and 1200.

Figure 3.

Baseline physiological analysis in control and thick ascending limb-specific Na⁺/H⁺ exchanger isoform 3 knockout (NHE3^TAL-KO) mice. NHE3^TAL-KO mice showed a trend for higher fluid intake (A) associated with a significantly lower urine osmolality (B) compared with control mice. No differences were found in plasma osmolality (C), food intake (D), hematocrit (E), and glomerular filtration rate (GFR; F) between genotypes. Data were analyzed by an unpaired Student’s t test. In addition to single data, means ± SE values are shown. n = 5 or 6 male and 5 female mice/genotype for A and D, n = 10 male and 10 female mice/genotype for B, n = 9 male mice/genotype for C, n = 13–15 male and 6 female mice/genotype for E, and n = 7 or 8 male and 8 female mice/genotype for F. *P < 0.05 vs. control. bw, body weight.

Physiological Analysis

Body weight and fluid and food intake were determined for 10 days starting after tamoxifen administration (21). Spontaneous-voided urine was collected daily by reflex urination/gentle bladder massage and holding mice over a clean Petri dish (22). Values from multiple days were averaged. Ten days after tamoxifen administration, blood was collected under brief isoflurane anesthesia from the retroorbital plexus for analysis.

Acute Responses to Furosemide

Acute responses to furosemide were assessed in anesthetized mice as previously described (23). Mice were anesthetized with thiobutabarbital (100 mg·kg⁻¹ body wt ip, 2 μL·g⁻¹ body wt, Sigma-Aldrich, St. Louis, MO) and ketamine (100 mg·kg⁻¹ body wt im, 2 μL·g⁻¹ body wt) (24). Mice were placed on an operating table with a servo-controlled heating plate (RT, Effenberger, Munich, Germany) to maintain body temperature at 37.5°C. The trachea was cannulated, and 100% oxygen was blown toward the tracheal tube throughout the experiment. The right jugular vein was cannulated for continuous maintenance infusion of 2.25% BSA in 0.85% NaCl at a rate of 500 µL·h⁻¹·30 g body wt⁻¹. A catheter was placed into the bladder for urine collection. The aforementioned preparation took ∼15 min, after which mice were allowed another 45 min to stabilize before baseline urine was collected for three consecutive 5-min periods. Mice were treated with increasing doses of furosemide (0.1, 0.3, 1, 3, and 10 mg·kg⁻¹ body wt) administered via bolus application (0.5 µL·g⁻¹ body wt, infusion rate: 25 µL·min⁻¹). After each bolus, allowing 2 min for furosemide distribution, urine was collected for two consecutive 5-min periods before the next bolus was applied. The urinary flow rate was calculated based on urine weight, which was determined gravimetrically. Urine was analyzed as described in the Analysis of Urine and Plasma Samples.

Chronic Responses to Furosemide

Chronic responses to furosemide (40 mg·kg⁻¹ ip given twice a day at 08:00 and 20:00) were studied in awake mice. Urine was collected at baseline and after 8 days of furosemide administration. On day 8, mice were anesthetized with isoflurane and a blood collection was performed. Subsequently, the left kidney was removed and immediately snap-frozen. Tissue was processed for immunoblot analysis (22).

Eighteen-Hour Water Deprivation

Water deprivation was conducted overnight (18 h) (25, 26). Blood was taken from the retroorbital plexus for the determination of hematocrit and osmolality, after brief isoflurane anesthesia, before and after water deprivation. At the same time, spontaneous-voided urine was collected for the determination of urine osmolality. After reintroduction of the water bottle, fluid intake was determined after 3, 6, and 24 h.

Analysis of Urine and Plasma Samples

Blood chemistry was determined by an OPTI CCA-TS2 blood gas analyzer using an E-Cl⁻ type cassette (OPTIMedical, Roswell, GA). Plasma and urine osmolality were measured using an Osmomat 3000 (Gonotec, Berlin, Germany). Urine clinical chemistry was performed using commercially available assays that were modified to work with small volumes (7, 25, 27). Urinary Cl⁻ was analyzed by the thiocyanide method (Stanbio Laboratory, Boerne, TX), and urinary creatinine was determined using Infinity Creatinine Liquid Stable Reagent (Thermo Fisher Scientific, Middletown, VA). Urinary pH was determined using a pH electrode immediately after collection (No. 9810BN, Thermo Fisher Scientific, Pittsburgh, PA).

Measurement of Glomerular Filtration Rate in Conscious Mice

Glomerular filtration rate (GFR) measurements were performed by determining the plasma kinetics of FITC-sinistrin (Fresenius-Kabi, Linz, Austria) following an intravenous bolus injection (28). FITC-sinistrin (2% in isotonic saline) was injected into the retroorbital plexus of mice (2 µL·g⁻¹ body wt) during brief isoflurane anesthesia. At 3, 5, 7, 10, 15, 35, 56, and 75 min after injection, blood was collected from the tip of the tail into Na⁺-heparinized 10-µL microcaps (Hirschmann Laborgeräte, Eberstadt, Germany). After centrifugation, plasma was diluted at 1:10 in 0.5 mol·L⁻¹ HEPES (pH 7.4), and fluorescence was determined with a fluorospectrometer (Cytation 3, Biotek, Winooski, VT). GFR was calculated using a two-compartment model of two-phase exponential decay (GraphPad Prism, San Diego, CA).

Quantitative RT-PCR

In a different cohort of control and NHE3^TAL-KO mice, the left kidney was removed for immunoblot analysis and the right kidney was separated into the cortex and renal medulla and processed for mRNA analysis. Tissue was homogenized using a Tissue Tearor (Bartlesville, OK) and QIAshredder (Qiagen, Valencia, CA) according to the manufacturer’s instructions. Total RNA was purified using the RNeasy Plus Mini Kit (Qiagen). cDNA was made by reverse transcribing total RNA using a QuantiTect reverse transcription kit (Qiagen) according to the manufacturer’s instructions. Semiquantitative RT-PCR was performed using TaqMan Universal PCR Master Mix (Fisher Scientific) in a QuantStudio 3 (Applied Biosystems, Foster City, CA). The template concentration was 12.5 ng cDNA per 25-µL reaction (performed in triplicate) and was used in conjunction with primer pairs specific for NHE3 (Mm01352478_g1, efficiency: 98%) and Sglt2 (Mm00453831_m1, efficiency: 98%) with GAPDH (Mm99999915_g1) as a reference gene. Data analysis used the ΔΔC_t method (where C_t is threshold cycle) and normalized to GAPDH expression and compared with control (2).

Immunofluorescent Staining

Tissue preparation, kidney sectioning, and immunolabeling were performed as previously described (2, 29) using antibodies against Lotus tetragonolobus lectin (LTL; 1:1,000, B-1325-2, Vector Laboratories, Burlingame, CA), NHE3 (1:500, SPC-400, StressMarq Biosciences, Victoria, BC, Canada), and NKCC2 (1:100, sc-293222, Santa Cruz Biotechnology, Dallas, TX). DyLight 649 streptavidin (1:1,000, SA-5649-1, Vector Laboratories), Texas red 595 goat anti-rabbit (1:1,500, TI-1000, Vector Laboratories), and Alexa Fluor 488 donkey anti-mouse (1:1,500, A-21202, Life Technologies, Carlsbad, CA) were used for the visualization of LTL, NHE3, and NKCC2, respectively. A mouse-on-mouse kit (BMK-2202, Vector Laboratories) was used to reduce endogenous mouse IgG staining. Sections were mounted using HardSet Antifade Mounting Medium with DAPI (H-1500, Vector Laboratories). An Olympus (Center Valley, PA) FV1000 MPE Multiphoton Laser Scanning Microscope with a ×40 UPLFL objective lens (numerical aperture: 0.75) was used for imaging of labeled sections.

Immunoblot Analysis

Kidney tissue was homogenized in dissection buffer (250 mmol·L⁻¹ sucrose and 10 mmol·L⁻¹ triethanolamine buffer) containing protease inhibitor cocktail (Roche Applied Science, Penzberg, Germany) and Halt phosphatase inhibitor cocktail (Thermo Fisher Scientific). Homogenates were centrifuged at 4,000 g before protein quantity was determined using a standard BCA assay (Thermo Fisher Scientific). Gel samples of equal concentration were made by the addition of Laemmli sample buffer (final concentration of 0.1 mol·L⁻¹ SDS and 15 mg·L⁻¹ DTT). Samples were heated at 65°C for 15 min before Western blot analysis. Proteins were transferred to polyvinylidene difluoride membranes and immunoblotted with rabbit polyclonal antibodies against NHE3 (AB3085, Millipore, Billerica, MA; characterized in Ref. 30), type II Na⁺-phosphate cotransporter (Npt2a) (27), NKCC2 (31), phospho-Thr⁵⁸ Na⁺-Cl⁻ cotransporter (NCC) (32), H⁺-ATPase (33), anion exchanger 2 (AE2) (34), pendrin (35), electrogenic Na⁺-bicarbonate cotransporter 1 (NBCe1) (36), H⁺-K⁺-ATPase M79 (37), Na⁺-K⁺-ATPase [No. 06-520, Merck Millipore, Darmstadt, Germany (7)], and mouse monoclonal antibody against NCC (38). For detailed antibody information, see Table 1. Detection was performed with ECL Plus (Amersham, Piscataway, NJ). Secondary antibodies were from Dako (Jena, Germany), and sites of antibody/antigen interaction were visualized using the Enhanced Chemiluminescence System (GE Healthcare, Buckinghamshire, UK) and an Image-Quant LAS 4000 imager (GE Healthcare). Coomassie-stained gels were used to adjust for equal protein loading for immunoblot analysis. Densitometric analyses were performed using Image Studio Lite (Qiagen).

Table 1.

Antibodies used in the present study

Target	Source	Use	Concentration/Dilution	Validation
NHE1	sc-136239, Santa Cruz Biotechnology, Dallas, TX	WB	1 μg/mL	Specific IHC labeling of the basolateral membrane of distal tubule/collecting duct segments, single protein band of ∼90 (∼115 glycosylated) kDa on WB, signal absent in kidney-specific NHE1 knockout mice (unpublished observations)
NHE3	AB3085, Millipore, Billerica, MA	WB	2 μg/mL	Specific IHC labeling of the apical brush border of proximal tubules, single protein band of ∼75 kDa on WB, signal absent in NHE3 knockout mice (2,3)
Npt2a	Researcher generated (39)	WB	1 μg/mL	Specific IHC labeling of the apical brush border of proximal tubules, single protein band of ∼80 kDa on WB, regulation of band intensities by parathyroid hormone (40)
NKCC2	Researcher generated (31)	WB	0.2 μg/mL	Specific IHC labeling of the apical membrane of cells in the TAL, single protein band of ∼160 kDa on WB of renal outer medulla homogenates that is absent in knockout mice (41)
NCC	Researcher generated (38)	WB	0.1 μg/mL	Single protein band of ∼140 kDa on WB of renal cortex homogenates that is absent in NCC knockout mice or mock NCC-transfected cells (38)
pT58 NCC	Researcher generated (32)	WB	0.01 μg/mL	Specific IHC labeling of the distal convoluted tubule, single protein band of ∼140 kDa on WB of renal cortex homogenates (32) or NCC-transfected cells that is absent in cells expressing mutant (T58A) form (42)
H+-ATPase B1 subunit	Researcher generated (43)	WB	0.1 μg/mL	Single protein band of ∼62 kDa on WB of rat kidney membrane preparations that disappears on peptide preabsorption, specific labeling of intercalated cells in the kidney collecting duct by IHC that is absent in knockout mice (44)
AE2	Researcher generated (45)	WB	1 μg/mL	Specific IHC labeling of cells throughout the renal tubule (46), single protein band on WB of lysates from AE2-transfected cells absent in mock-transfected cells (34)
Pendrin	Researcher generated (35)	WB	0.1 μg/mL	Specific IHC labeling of type B and non-A, non-B intercalated cells (35), single protein band of ∼85 kDa on WB of kidney lysates
NBCe1	Researcher generated (36)	WB	1 μg/mL	Single protein band of ∼120 kDa on WB of lysates from the kidney cortex that is absent from lysates from the outer and inner medulla (36), specific IHC labeling of the basolateral membrane of proximal tubule cells that is absent in knockout mice (47)
H+-K+-ATPase	Researcher generated (37)	WB	1 μg/mL	Nonglycosylated and core-glycosylated forms of protein only observed in H+-K+-ATPase-expressing cells and not mock-transfected cells (37)
Na+-K+-ATPase α1-subunit	No. 06-520, Merck Millipore, Darmstadt, Germany	WB	0.01 μg/mL	Single protein band of ∼100 kDa on WB of lysates from the kidney that is increased by high aldosterone (48)

AE2, anion exchanger 2; IHC, immunohistochemistry; NBCe1, electrogenic Na⁺-bicarbonate cotransporter 1; NCC, Na⁺-Cl⁻ cotransporter; NHE, Na⁺-H⁺ exchanger; NKCC2, Na⁺-K⁺-2Cl⁻ cotransporter; Npt2a, type II Na⁺-phosphate cotransporter; pT58 NCC, phospho-Thr⁵⁸ Na⁺-Cl⁻ cotransporter; WB, Western blot.

Statistical Analyses

Data are expressed as means ± SE. An unpaired Student’s t test was performed to analyze statistical differences between the two groups. One- and two-way ANOVA or two-way mixed-effects ANOVA followed by the two-stage linear step-up procedure of Benjamini, Krieger, and Yekutieli or Tukey multiple comparison tests, as indicated in the figures, were used to test for significant differences between genotype or experimental conditions. All data were analyzed via GraphPad Prism, v. 8.3 or SigmaPlot, v. 12.5. Significance was considered at P < 0.05.

RESULTS

Confirmation of NHE3 Knockout in the TAL

NHE3^TAL-KO mice were born at the predicted Mendelian frequencies, appeared grossly indistinguishable from control mice, and developed normally. Mice were treated with tamoxifen, and qualitative immunofluorescence was used to assess NHE3 distribution in the kidney of control and NHE3^TAL-KO mice (Fig. 1). LTL was used to guide the identification of proximal tubules, and NKCC2 was used as a marker for the TAL. Staining intensity and NHE3 localization in the proximal tubule were not different between control and NHE3^TAL-KO mice. In contrast, NHE3 labeling in the TAL was completely absent in NHE3^TAL-KO mice (Fig. 1). Kidneys were dissected into the cortex and medulla and processed for mRNA analysis. Expression of SGLT2, expressed in the S1 and S2 segments of the proximal tubule, was ∼93% lower in the medulla versus the cortex, independent of genotype (Fig. 2A), confirming the purity of the kidney dissection. In contrast, NHE3 mRNA was significantly lower in the cortex (by ∼32%) and medulla (by ∼90%) of NHE3^TAL-KO mice compared with control mice (Fig. 2B). Similar results were found at the protein level, with NHE3 abundance in the cortex and medulla of NHE3^TAL-KO mice ∼55% and ∼90% lower, respectively, compared with control mice (Fig. 2C). The specificity of the antibody was confirmed using homogenates from an NHE3^{loxloxPax8Cre} mouse (7), which lacks renal NHE3 expression.

Figure 1.

Thick ascending limb (TAL)-specific Na⁺/H⁺ exchanger isoform 3 (NHE3) knockout (NHE3^TAL-KO) mice lack immunofluorescent labeling of NHE3 in the TAL of the kidney. Representative confocal images of NHE3 localization in control and NHE3^TAL-KO mice are shown. In control mice, NHE3 labeling (red) overlapped in the S1 and S2 segments of the proximal tubule with Lotus tetragonolobus lectin (LTL; proximal tubule marker). Na⁺-K⁺-2Cl⁻ cotransporter (NKCC2; green signal) is a marker of the TAL and identifies overlapping expression of NHE3 in the TAL (arrowheads, yellow signal in the merged image). In contrast, no NHE3 labeling (arrows) was found in the TAL of NHE3^TAL-KO mice (lack of yellow labeling); however, NHE3 staining in the proximal tubule was preserved (magenta signal in the merged image), indicating selective deletion of NHE3 in the TAL.

Figure 2.

Na⁺/H⁺ exchanger isoform 3 (NHE3) mRNA and protein expression are significantly reduced in thick ascending limb-specific NHE3 knockout (NHE3^TAL-KO) mice. A: as a marker of the purity of the preparation, mRNA expression of Na⁺-glucose transporter 2 (SGLT2; expressed in the S1 and S2 segments of the proximal tubule) was used. Independent of genotype, SGLT2 mRNA expression was almost completely absent in the medulla vs. cortex preparation. B: NHE3 mRNA expression in NHE3^TAL-KO mice was ∼30% lower in the cortex and ∼94% lower in the medulla compared with control mice. C: compared with control mice, NHE3 protein expression (molecular weight: ∼93 kDa) in NHE3^TAL-KO mice was ∼55% and ∼90% lower in the cortex and medulla, respectively. Data were analyzed by an unpaired Student’s t test. In addition to single data, means ± SE values of n = 4–8 male mice/genotype are shown. *P < 0.05 vs. control. WT, wild type.

Baseline Physiological Analysis in Control and NHE3^TAL-KO Mice

Body weight was not significantly different between genotypes 10 days after tamoxifen administration (control: 32 ± 1 g and NHE3^TAL-KO: 32 ± 1 g, n = 18 and 19, respectively). Measurement of fluid and food intake when mice were housed in their home cages showed that NHE3^TAL-KO mice had a tendency for higher fluid intake (Fig. 3A), but this did not reach statistical significance (P = 0.08). Urine osmolality (Fig. 3B) was significantly lower (∼400 mmol·kg⁻¹, P < 0.05) in NHE3^TAL-KO mice compared with control mice. No significant differences between genotypes were observed in plasma osmolality, food intake, hematocrit, or GFR (Fig. 3, C–F).

Blood analysis showed no significant differences between genotypes in the concentrations of Na⁺, K⁺, Cl⁻, or pH (Fig. 4, A–D). Urinary Na⁺- and Cl⁻-to-creatinine ratios as well as urinary pH were not significantly different between genotypes (Fig. 4, E–H). In contrast, urinary K⁺/creatinine (Fig. 4F) was slightly but significantly elevated in NHE3^TAL-KO mice compared with control mice.

Figure 4.

Thick ascending limb-specific Na⁺/H⁺ exchanger isoform 3 knockout (NHE3^TAL-KO) mice show no electrolyte or acid-base abnormalities. No differences were observed between genotypes in plasma Na⁺ (A), K⁺ (B), Cl⁻ (C), blood pH (D), urinary Na⁺-to-creatinine (E), Cl⁻-to-creatinine ratios (G), or urinary pH (H). F: NHE3^TAL-KO mice had significantly higher urinary K⁺-to-creatinine ratios compared with control mice. Data were analyzed by an unpaired Student’s t test. In addition to single data, means ± SE values of n = 23–25 male mice/genotype are shown. *P < 0.05 vs. control.

Acute Furosemide Responses in Control and NHE3^TAL-KO Mice

To test the role of NHE3 in the TAL for furosemide-induced urinary acidification, experiments were performed in anesthetized mice. Furosemide inhibits NKCC2, which, in addition to transporting Na⁺, K⁺, and Cl⁻, can transport ${\text{NH}}_4^{+}$, which could acidify the urine (49). Bolus intravenous administration of furosemide resulted in comparable diuretic (ED₅₀: 1.8 ± 0.3 vs. 1.8 ± 0.4 mg·kg⁻¹; Fig. 5A) and chloruretic (ED₅₀: 1.6 ± 0.3 vs. 2.1 ± 0.5 mg·kg⁻¹; Fig. 5B) dose-response curves in control and NHE3^TAL-KO mice. Similar results were obtained when urinary Cl⁻ excretion was corrected by urinary creatinine (Fig. 5C). Relative to baseline, urinary pH significantly decreased at a dose of 0.3 mg·kg⁻¹ in control mice; however, higher doses did not further decrease urinary pH in the timeframe examined (Fig. 6, Aand B). The maximum decrease was ∼0.45 ± 0.1 pH units. In contrast, urinary pH did not significantly differ from baseline in NHE3^TAL-KO mice at a dose of 0.3 mg·kg⁻¹ (Fig. 6, Aand B). At a dose of 1 mg·kg⁻¹, urinary pH significantly decreased versus baseline (−0.6 ± 0.1 pH units), reaching a maximum response that was comparable to control mice. In contrast to control mice, this response was immediately followed by a return toward baseline levels (Fig. 6, Aand B). No additional effects were observed at 3 and 10 mg·kg⁻¹ compared with baseline or previous dose.

Figure 5.

Acute diuretic and chloruretic responses to furosemide are similar in control and thick ascending limb-specific Na⁺/H⁺ exchanger isoform 3 knockout (NHE3^TAL-KO) mice. After administration of increasing doses of furosemide under anesthesia, the diuretic (A), chloruretic (B), and urinary Cl⁻-to-creatinine ratio (C) dose-response curves showed no differences between genotypes. Data were fit to a four-parameter logistic curve, and ED₅₀ values were calculated with SigmaPlot. Data are expressed as means ± SE; n = 10–12 male mice/genotype.

Figure 6.

Thick ascending limb-specific Na⁺/H⁺ exchanger isoform 3 knockout (NHE3^TAL-KO) mice lack a sustained urinary acidification in response to furosemide. A: in control mice, furosemide at doses of 0.3 and 1 mg·kg⁻¹ resulted in a dose-dependent decrease in urinary pH; however, higher doses of furosemide (3 and 10 mg·kg⁻¹) did not further acidify the urine. In NHE3^TAL-KO mice, furosemide significantly decreased urinary pH only at a dose of 1 mg·kg⁻¹, followed by an acute increase toward baseline levels. B: single data are shown to better illustrate the lack of change at 1 and 3 mg·kg⁻¹ in control mice, which contrasted with NHE3^TAL-KO mice, where urinary pH recovered and remained at a constant higher level. In A, data are shown as means ± SE; in B, single data and means ±SE values of n = 10–12 male mice/genotype are shown. #P < 0.05 vs. vehicle for control. §P < 0.05 vs. vehicle in NHE3^TAL-KO mice. †P < 0.05 vs. previous period NHE3^TAL-KO mice. Data were analyzed by two-way mixed-effects ANOVA followed by the two-stage linear step-up procedure of Benjamini, Krieger, and Yekutieli.

Chronic Effects of Furosemide Treatment in Control and NHE3^TAL-KO Mice

To study the chronic effects of NHE3 in furosemide-induced acidification, mice were treated for 8 days with furosemide. Vehicle treatment (Fig. 7A) did not significantly change urine osmolality in control and NHE3^TAL-KO mice (−180 ± 140 vs. −8 ± 214 mmol·kg⁻¹, not significant). In contrast, furosemide treatment significantly reduced urine osmolality in control mice (−1,230 ± 238 mmol·kg⁻¹) and NHE3^TAL-KO mice (−1,510 ± 77 mmol·kg⁻¹), with no significantly different effects between genotypes. In contrast to acute experiments under anesthesia (Fig. 6), no difference in urinary pH was observed in response to furosemide (Fig. 7B). Consistent with furosemide-induced hypokalemia and metabolic alkalosis, blood K⁺ (Fig. 7C) and blood pH (Fig. 7D) were significantly lower and higher, respectively, in furosemide-treated control mice compared with vehicle-treated mice. Neither plasma K⁺ (Fig. 7C) nor blood pH (Fig. 7D) was significantly different between vehicle- and furosemide-treated NHE3^TAL-KO mice. Furosemide also increases urinary Ca²⁺ excretion (50). Consistent with this, in control mice, the urinary Ca²⁺-to-creatinine ratio increased approximately threefold compared with baseline (Fig. 7E). In contrast, in furosemide-treated NHE3^TAL-KO mice, the urinary Ca²⁺-to-creatinine ratio increase was ∼1.5-fold compared with baseline, which was not statistically significant, and remained significantly lower compared with furosemide-treated control mice. Vehicle treatment did not affect urinary Ca²⁺-to-creatinine ratios in control or NHE3^TAL-KO mice.

Figure 7.

Thick ascending limb-specific Na⁺/H⁺ exchanger isoform 3 knockout (NHE3^TAL-KO) mice lack hypokalemia, metabolic alkalosis, and calciuresis in response to chronic furosemide treatment. A: chronic (8 day) furosemide treatment significantly reduced urine osmolality in control and NHE3^TAL-KO mice; vehicle treatment was without effect in either genotype. B: urinary pH was not affected by genotype or treatment. In control mice, furosemide treatment resulted in hypokalemia (C), metabolic alkalosis (D), and calciuresis (E). In NHE3^TAL-KO mice, these effects were absent. Data were analyzed by repeated-measures two-way ANOVA (A, B, and E) and two-way ANOVA (C and D), with both followed by a Tukey’s multiple comparison test. In addition to single data, means ± SE values of n = 6 or 7 male mice/genotype are shown. *P < 0.05 vs. control. #P < 0.05 vs. baseline same genotype.

Response to Water Deprivation

To test for the role of NHE3 in the TAL in urinary concentrating ability, mice were challenged by an 18-h water deprivation. Urine osmolality increased significantly in control mice (change: 1,255 ± 121 mmol·kg⁻¹) and NHE3^TAL-KO mice (change: 1,489 ± 132 mmol·kg⁻¹), which was not significantly different between genotypes (Supplemental Fig. S1A: https://doi.org/10.6084/m9.figshare.19705429.v1). In response to water deprivation, both genotypes showed a significant increase in plasma osmolality (Supplemental Fig. S1B) and hematocrit (Supplemental Fig. S1C) as well as a decrease in body weight (Supplemental Fig. 1D) that were not significantly different between control and NHE3^TAL-KO mice. Fluid intake, up to 24 h after water deprivation, was also not significantly different between control mice (7.6 ± 0.5 mL) and NHE3^TAL-KO mice (8.5 ± 0.4 mL; Supplemental Fig. S1E).

Profiling of Transporters, Pumps, and Channels Along the Nephron in Response to Chronic Furosemide Treatment in Control and NHE3^TAL-KO Mice

Expression levels of NKCC2 were greater in furosemide-treated mice compared with vehicle-treated mice, regardless of genotype (Fig. 8A). In the distal convoluted tubule, the expression (Fig. 8B) and phosphorylation (Fig. 8C) of NCC at Thr⁵⁸ were greater in both genotypes in response to furosemide treatment compared with vehicle-treated mice. Levels of the Na⁺-K⁺-ATPase α-subunit (Fig. 8D) were not affected by treatment or genotype. Full-length Npt2a and NBCe1 are expressed in the luminal and basolateral membrane of the proximal tubule, respectively. In response to chronic furosemide treatment in control mice, full-length Npt2a (Fig. 9A) expression was significantly lower (∼45%, P < 0.05) compared with vehicle-treated control mice. Levels were not significantly affected by treatment in NHE3^TAL-KO mice. Levels of NBCe1 (Fig. 9B) were not affected by treatment or genotype. Full-length AE2 is localized to the basolateral membrane in the TAL. No differences in expression levels were observed between genotypes or treatments (Fig. 9C). Expression levels of nongastric H⁺-K⁺-ATPase (Fig. 9D), found from the TAL to the inner medullary collecting duct, were not affected by treatment or genotype. Levels of the H⁺-ATPase β1-subunit (Fig. 9E), located in type A, type B, and non-A, non-B intercalated cells, were not affected by treatment or genotype. Levels of the Cl⁻/${\text{HCO}}_3^{-}$ exchanger pendrin (Fig. 9F), expressed in type B and non-A, non-B intercalated cells, were not significantly affected by treatment or genotype. No compensatory changes were observed in Na⁺/H⁺ exchanger isoform 1 (NHE1) expression between genotypes, and NHE1 was not affected by furosemide treatment (Supplemental Fig. S2).

Figure 8.

Furosemide (Furo)-induced greater abundance of Na⁺-K⁺-Cl⁻ cotransporter (NKCC2) in control (Con) mice is absent in thick ascending limb-specific Na⁺/H⁺ exchanger isoform 3 knockout (NHE3^TAL-KO) mice. A: chronic furosemide treatment (8 days) resulted in significantly higher NKCC2 expression (molecular weight: ∼121 kDa) in control and NHE3^TAL-KO mice compared with vehicle (Veh) treatment. B and C: Na⁺-Cl⁻ cotransporter (NCC) expression (B; molecular weight: ∼160 kDa) and phosphorylation at Thr⁵⁸ (pT58 NCC; C; molecular weight: ∼160 kDa) were significantly higher in response to furosemide treatment independent of genotype. D: expression of the Na⁺-K⁺-ATPase α-subunit (molecular weight: ∼112 kDa) was not affected by treatment or genotype. Data were analyzed by two-way ANOVA followed by a Tukey’s multiple comparison test. In addition to single data, means ± SE values of n = 6 male mice/genotype are shown. §P < 0.05 vs. vehicle same genotype.

Figure 9.

Lower type II Na⁺-phosphate cotransporter (Npt2a) expression in response to chronic furosemide (Furo) treatment in control (Con) mice is absent in thick ascending limb-specific Na⁺/H⁺ exchanger isoform 3 knockout (NHE3^TAL-KO) mice. A: chronic furosemide treatment resulted in significantly lower Npt2a (molecular weight: ∼75 kDa) expression in control mice compared with vehicle (Veh), an effect absent in NHE3^TAL-KO mice. B–F: no differences were observed in the levels of electrogenic Na⁺-bicarbonate cotransporter 1 (NBCe1; B; molecular weight: ∼116 kDa), anion exchanger 2 (AE2; C; molecular weight: ∼180 kDa), nongastric H⁺-K⁺-ATPase (D; molecular weight: ∼100 kDa), H⁺-ATPase β1-subunit (E; molecular weight: ∼57 kDa), or the Cl⁻/${\text{HCO}}_3^{-}$ exchanger pendrin (F; molecular weight: ∼95 kDa) between treatment or genotype. Data were analyzed by two-way ANOVA followed by a Tukey’s multiple comparison test. In addition to single data, means ± SE values of n = 6 male mice/genotype are shown. §P < 0.05 vehicle same genotype.

DISCUSSION

Studies to define the general role of NHE3 in the TAL were hampered by the lack of a suitable animal model. To navigate this problem, we generated NHE3^TAL-KO mice and used this model to determine the contribution of NHE3 in the TAL to renal acid-base/electrolyte/water homeostasis as well as its contribution to furosemide-induced urinary acidification. In this model, NHE3 mRNA and protein were greatly reduced in the medulla, with immunohistochemistry confirming that this is due to an almost complete absence of NHE3 in the TAL, whereas proximal tubule NHE3 expression is unaffected. Despite the Cre-driver mouse line we used being tamoxifen-inducible, it was previously reported that a small number of tdTomato-positive cells exist within a subpopulation of uromodulin-positive cells (51), suggesting tamoxifen-independent Cre-mediated recombination. Consequently, we have not studied the phenotype of NHE3 knockout before and after tamoxifen administration.

Under standard (baseline) conditions, there were no significant differences between control and NHE3^TAL-KO mice in body weight, food intake, and plasma concentrations of Na⁺, K⁺, and Cl⁻. Fluid intake tended to be higher, and when combined with the observed lower urine osmolality in NHE3^TAL-KO mice without any differences in kidney function, plasma osmolality, or hematocrit, it suggests that NHE3^TAL-KO mice have a mild defect in the renal handling of water but remain in fluid balance. Mice with knockout of NHE3 along the entire nephron also have a mild urinary concentrating defect (7); however, when NHE3^TAL-KO mice were challenged by water deprivation, no phenotype was detected, implying that other mechanisms are fully capable to compensate for the lack of NHE3 in the TAL under these conditions. Furthermore, under baseline conditions, blood and urine pH in NHE3^TAL-KO mice were not different from controls, suggesting that lack of NHE3 in the TAL does not cause an acid-base phenotype. This is in contrast to mice deficient in NHE3 along the entire nephron, which have a higher urinary pH (7). This suggests that NHE3 activity in the proximal tubule is more important for renal ${\text{HCO}}_3^{-}$ reabsorption than NHE3 activity in the TAL.

In addition to being a diuretic, furosemide causes urinary acidification (52). The mechanism for this was originally proposed to be dependent on epithelial Na⁺ channel-induced activation of vacuolar H⁺-ATPase (B1 subunit) in the connecting tubule (53, 54). In rats, 7 days of furosemide treatment increased H⁺-ATPase B1 subunit abundance in the cortex and outer medulla (∼1.8-fold and ∼2.4-fold, respectively) without affecting mRNA expression (55). However, we did not observe changes in the protein abundance of the H⁺-ATPase B1 subunit or other transporters/pumps involved in acid-base regulation after 8 days of furosemide treatment. Previous studies have identified that in adult NHE3^−/− mice Na⁺/H⁺ exchanger isoform 8 expression is increased (56). Our data expand this knowledge and show that the ubiquitously expressed NHE1 isoform is not increased in the absence of NHE3 in the TAL and not regulated by furosemide. Whether other Na⁺/H⁺ exchanger isoforms compensate in NHE3^TAL-KO mice remains to be determined.

More recently, the effects of furosemide have been suggested to be mediated by NHE3 activity in the TAL, with urinary acidification markedly reduced after pharmacological NHE3 inhibition because of a decrease in intracellular Na⁺ concentration, increasing the driving force for NHE3 (18). Our novel mouse line provided us the ideal opportunity to test this mechanism for furosemide-induced urinary acidification and has provided significant new insights. In acute experiments under anesthesia, furosemide induced a rapid and sustained decrease of urinary pH in control mice. NHE3^IEC-KO mice showed a response of similar magnitude. However, this acidification was followed by an immediate recovery to levels not significantly different from baseline, indicating that NHE3 activity is required for sustained furosemide-induced urinary acidification. In control mice, doses of >1 mg·kg⁻¹ did not result in an additional effect on urinary acidification despite having a significant effect on urinary flow rate and urinary Cl⁻ excretion. In contrast to the acute experiments, chronic furosemide administration had no effect on urinary acidification. The reason for this remains elusive, but we do not see furosemide-induced urinary acidification during short-term (3-h) studies when mice are housed in metabolic cages (unpublished observations). A similar lack of urinary acidification in response to furosemide has been observed by others. For example, acute studies in humans, 15-min collection after furosemide administration, showed an increase in urinary pH in conjunction with an increase in ${\text{HCO}}_3^{-}$ excretion (57, 58). This effect was suggested to be caused by furosemide-inhibiting carbonic anhydrase (50, 58). Other possibilities include changes in titratable acid excretion and/or ammonium trapping (59). Taken together, furosemide-induced urinary acidification might depend on species and/or other experimental conditions.

Despite a lack of furosemide-induced urinary acidification from spontaneous urine collections in our long-term experiments, control mice developed metabolic alkalosis. However, an effect of furosemide on acid-base status was absent in NHE3^TAL-KO mice, highlighting the importance of this segment for long-term acid-base regulation in response to furosemide. Metabolic alkalosis is usually associated with hypokalemia, which can be compensated for by muscle and kidneys (60). Furosemide causes flow-induced K⁺ secretion, which requires functional epithelial Na⁺ channels as a prerequisite (61), which might further worsen hypokalemia. Consistent with this, in control mice, plasma K⁺ was significantly lower in response to furosemide compared with vehicle treatment. In contrast, plasma K⁺ remained unaffected in NHE3^TAL-KO mice, possibly because of a lack of furosemide-induced changes in blood pH. We hypothesize that the exchange of H⁺ versus K⁺ in muscle cells is the dominating factor for hypokalemia because similar urinary flow rates and urine osmolalities in response to furosemide were observed between genotypes. However, our study cannot completely exclude that changes in the expression levels/activity of the renal outer medullary K⁺ channel (62) and/or maxi K⁺ channel exist (63) between the genotypes that affect the kaliuretic response. The observed higher urinary K⁺-to-creatinine ratio in NHE3^TAL-KO mice under baseline conditions could also relate to flow-induced K⁺ secretion considering the lower urine osmolality and trend for greater fluid intake, both indicators of a higher urinary flow rate.

Furosemide treatment is associated with an increase in urinary Ca²⁺ excretion (50, 64, 65). Mechanistically, the furosemide-induced increase in Ca²⁺ excretion is proposed to be due to a reduction in the lumen-positive transepithelial voltage and consequently a reduction in paracellular Ca²⁺ (and Mg²⁺) transport in the TAL (66). In addition, increased Na⁺ and Cl⁻ reabsorption in the distal convoluted tubule is associated with decreased Ca²⁺ reabsorption. Regarding the latter, it was shown that the Ca²⁺-sensing receptor can increase NCC activity via with-no-lysine kinases acting upon the Ste20-related proline alanine-rich kinase pathway resulting in inhibition of Ca²⁺ reabsorption (67). The furosemide-induced calciuresis is associated with an increased risk of hyperparathyroidism in patients with normal kidney function (68) and with chronic kidney disease (69). A reduction in plasma Ca²⁺ increases parathyroid hormone, which can reduce Npt2a abundance in response to furosemide treatment; however, if this is the primary effect or if off-target effects on carbonic anhydrase in the proximal tubule (70) play a role remains to be determined. In humans, furosemide increased urinary P_i excretion in the first hour after administration, which was followed by a reduction of P_i excretion in the subsequent hour if volume contraction was not corrected (71). Clearance experiments in dogs showed that furosemide increased fractional P_i excretion, an effect that was dependent on parathyroid hormone (72). Our data expand this previous knowledge by showing that Npt2a abundance is reduced by ∼50% in control mice chronically treated with furosemide. In contrast, NHE3^TAL-KO mice do not have significantly lower Npt2a abundance or calciuresis after furosemide. The reason(s) for this is unknown, but it could imply that NHE3 in the TAL plays a role in Ca²⁺ reabsorption. The fact that NHE3^TAL-KO mice lack the sustained urinary acidification in response to furosemide and that the Ca²⁺-sensing receptor is altered by extracellular pH (73) could have secondary consequences on claudin 14 and 16, both regulators of paracellular cation permeability in the TAL (74).

Furosemide treatment induces a compensatory increase in NKCC2 expression (75), a mechanism speculated to contribute to diuretic resistance (76). Our results confirm this phenomenon, with chronic furosemide treatment increasing the abundance of NKCC2 (75) and NCC (77, 78) as well as phosphorylation of NCC (20, 67, 78, 79). In addition, the growth of the distal tubule has been described as a response to increased NaCl delivery into the distal tubule (80, 81). Despite the possible increase in distal convoluted tubule mass, we did not find greater Na⁺-K⁺-ATPase expression in whole kidneys in response to furosemide treatment, which is consistent with unaffected mRNA expression in response to furosemide treatment in Wistar rats (82). However, another study found that Na⁺-K⁺-ATPase activity was increased in the distal convoluted tubule and collecting duct of furosemide-treated rats (83).

In summary, we have generated a novel TAL-specific NHE3 knockout model. Using this tool, we were able to demonstrate that under normal conditions, the contribution of NHE3 in the TAL for acid-base regulation is minor or can be compensated for in distal nephron segments. We also provide new evidence for a role of NHE3 in the TAL for furosemide-induced acidification, with acute urinary acidification unaffected in NHE3^TAL-KO mice but an immediate recovery to baseline levels, suggesting a role for NHE3 in sustained urinary acidification. Consistent with this, NHE3^TAL-KO mice lack metabolic alkalosis in response to long-term treatment with furosemide. Of note, it is possible that under certain conditions (low ${\text{HCO}}_3^{-}$ or metabolic acidosis), NHE3 in the TAL may play a permissive role for NaCl and acid-base homeostasis that is not observed under baseline conditions or when NKCC2 is maximally inhibited by furosemide.

SUPPLEMENTAL DATA

Supplemental Figs. S1 and S2: https://doi.org/10.6084/m9.figshare.19705429.v1.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grant 1R01DK110621 (to T.R.), Veterans Affairs Merit Review Award IBX004968A (to T.R.), and American Heart Association Transformational Research Award 19TPA34850116 (to T.R.). Financial support for this work was also provided by NIDDK Diabetic Complications Consortium (RRID: SCR_001415, www.diacomp.org) Grants DK076169 and DK115255 (to T.R.). L.T. was supported by American Heart Association Postdoctoral Fellowship 828731. Further funding to R.A.F. was provided by the Novo Nordisk Foundation (NNF21OC0067647, NNF17OC0029724, and NNF19OC0058439), the Leducq Foundation (17CVD05), and the Independent Research Fund Denmark (0134-00018B).

DISCLOSURES

T.R. and R.A.F. are Associate Editors of the American Journal of Physiology-Renal Physiology but are not involved and did not have access to information regarding the peer-review process or final disposition of this article. An alternate editor oversaw the peer-review and decision-making process for this article. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

J.X. and T.R. conceived and designed research; J.X., L.T., J.A.D.R., R.A.F., and T.R. performed experiments; J.X., L.T., J.A.D.R., R.A.F., and T.R. analyzed data; J.X., L.T., J.A.D.R., R.A.F., and T.R. interpreted results of experiments; J.X., R.A.F., and T.R. prepared figures; J.A.D.R. and T.R. drafted manuscript; J.X., L.T., J.A.D.R., R.A.F., and T.R. edited and revised manuscript; J.X., L.T., J.A.D.R., R.A.F., and T.R. approved final version of manuscript.