Bicarbonate is the primary inducer of KCC3a expression in renal cortical B-type intercalated cells

Abstract

A primary function of intercalated cells in the distal tubule of the kidney is to maintain pH homeostasis. For example, type B intercalated cells secrete bicarbonate largely through the action of the apical Cl⁻/HCO₃⁻ exchanger, pendrin, which helps correct metabolic alkalosis. Since both the K-Cl cotransporter, KCC3a and pendrin colocalize to the apical region of type B and non-A, non-B intercalated cells and since both are upregulated in models of metabolic alkalosis, such as with dietary NaHCO₃ loading, we raised the possibility that apical KCC3a facilitates pendrin-mediated bicarbonate secretion, such as through apical Cl⁻ recycling. The purpose of this study was to determine if KCC3a abundance changes through intake of bicarbonate alone or through bicarbonate plus its accompanying cation, and if it requires a direct interaction with pendrin or the renin-angiotensin-aldosterone system. We observed that KCC3a protein abundance, but not mRNA, increases in a mouse model of metabolic alkalosis, achieved with dietary NaHCO₃ or KHCO₃ intake. Bicarbonate ion increases KCC3a abundance, both in vivo and in vitro, independently of the accompanying cation. Moreover, bicarbonate intake upregulates KCC3a independently of aldosterone or angiotensin II. Since NaHCO₃ intake increased KCC3a abundance in wild-type as well as in pendrin knockout mice, this KCC3a upregulation by bicarbonate does not depend on a direct interaction with pendrin. We conclude that increased extracellular bicarbonate, as observed in models of metabolic alkalosis, directly raises KCC3a abundance independently of angiotensin II, aldosterone, or changes in KCC3a transcription and does not involve a direct interaction with pendrin.

NEW & NOTEWORTHY KCC3a expression is stimulated in alkalemia. This paper shows that bicarbonate itself is mediating this effect through a posttranscriptional mechanism. The paper also shows that this phenomenon is not mediated by aldosterone or angiotensin II.

INTRODUCTION

Intercalated cells are kidney distal tubule epithelial cells that play a critical role in the control of acid-base balance. Changes in extracellular or intracellular pH or HCO₃⁻ (bicarbonate) concentration rapidly change the activity and subcellular localization of acid-base transport proteins within these cells (1). The acid-base sensing mechanism of intercalated cells has been the subject of intense research. Pendrin, a chloride-bicarbonate exchanger, is expressed on the apical plasma membrane and subapical vesicles of type B and non-A, non-B intercalated cells and is greatly upregulated in models of metabolic alkalosis. These cells secrete bicarbonate into the lumen through pendrin during an alkalotic load, thereby helping to correct the alkalosis (1). We recently demonstrated the expression of a K-Cl cotransporter, KCC3a, in pendrin-positive intercalated cells in the mouse kidney (2). Because the cotransporter transports K⁺ and Cl⁻ across the membrane into the lumen due to ionic gradients (3) and because the transporter primarily localizes with pendrin on the apical membrane of bicarbonate secreting cells in cortical connecting tubules, we proposed that the cotransporter and pendrin would act in tandem to mediate net KHCO₃ secretion while recycling Cl⁻ across the apical plasma membrane. It is well-known that metabolic alkalosis is associated with lower plasma K⁺ or hypokalemia (4). As such, KCC3a might be the long-sought electroneutral transport mechanism that accounts for K⁺ loss in alkalemia (5, 6). In support of this hypothesis, we demonstrated that KCC3a is upregulated in rodent models of metabolic alkalosis (2).

The purpose of this study was to determine whether KCC3a abundance is increased following a bicarbonate challenge and, if so, whether the anion, the cation, or both, or hormones that are commonly changed in these models, such as aldosterone and angiotensin II, are significantly engaged in this process. We found that KCC3a expression was upregulated by a bicarbonate load, irrespective of the cation, as both NaHCO₃ and KHCO₃ led to similar increases in protein abundance, and that high K⁺ (or aldosterone) was not the major modulator as a high KCl diet failed to reproduce the increased expression. We found the effects long-lasting, as the 24-h stimulation of both KCC3a and pendrin expression could still be observed in bicarbonate-fed mice after 7 days. We observed that citrate (in the form of K-citrate), mimicked the bicarbonate effect. We showed that the increased expression was not due to volume depletion and angiotensin, as bicarbonate load with high-salt diet resulted in similar KCC3a increases and as the increase was not prevented by injection of the angiotensin-receptor inhibitor: losartan. Finally, we found that KCC3a abundance is regulated by bicarbonate rather than pH, in cultured mouse cortical M-1 cells. We conclude that bicarbonate itself is the signal that led to KCC3a upregulation.

MATERIAL AND METHODS

Animal Experiments

In this investigation, male C57BL/6J mice between the ages of 2–3 mo were employed. All animals were kept in ventilated cages with a 12-h light/12-h dark cycle, in a temperature-controlled and pathogen-free barrier facility. In our previous study, we noticed that different treatments had a similar impact on the abundance of KCC3a in both male and female mice. To adhere to the objective of minimizing the number of animals used whenever possible, we used only one gender in this investigation. In the first experiment, mice were maintained on regular chow and received 280 mM NaCl or NaHCO₃ in their drinking water. Acidosis was induced by providing 280 mM NH₄Cl in drinking water. In a separate experiment, mice were maintained on a normal salt or high-sodium diet (Envigo Teklad custom diet, TD.190009; 1.05% K⁺, 1.57% Na⁺, 3.38% Cl) with or without 280 mM KHCO₃ in drinking water to induce alkalosis. Losartan (100 mg/L), an antagonist of angiotensin-II receptor type 1 (AT1), was given to mice on a regular salt or Na⁺ deficient diet for 4 days (Envigo Teklad custom diet, TD.190152; 1.05% K⁺, 0% Na⁺, 0.95% Cl⁻). Another group of mice received either water, 280 mM KCl, or K-citrate with free access to their regular chow. Pendrin knockout mice were originally developed by Everett et al. (7) and characterized by us (8–11). The mice were maintained on a normal or high KHCO₃ diet (TD 140044) for 4 days. For tissue collection, mice were euthanized by isoflurane overdose, followed by cervical dislocation. All mouse studies were approved by the Vanderbilt Animal Care and Use Committee.

Antibodies

The KCC3a antibody was made in-house and characterized as described previously (2). The pendrin antibody has also been fully validated (8, 12). The β-actin antibody was purchased from Abcam (No. ab82812). Dilutions for KCC3a Ab (in-house) = 1:500 (for WB), 1:100 (for IF); Pendrin Ab (from Susan Wall laboratory) = 1:5,000, β-actin (Abcam, ab8227) = 1:3,000, and secondary HRP-conjugated antibody (Promega, W401B) = 1: 2,000.

Kidney Western Blot

Upon harvest, kidneys were promptly flash-frozen in liquid nitrogen and either processed right away or kept at −80°C. Single kidneys were homogenized using a Potter homogenizer in 1 mL of ice-cold homogenization solution, which contained 300 mM sucrose, 50 mM Tris-HCl (pH 7.4), 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 1 mM phenylmethyl-sulphonyl fluoride, Halt protease/phosphatase cocktail (Thermo Fisher Scientific, Cat. No. 78442), and PhosSTOP phosphatase inhibitor cocktail tablet (Roche, Cat. No. 04906837001). The homogenate was centrifuged for 15 min at 6,000 rpm at 4°C, and the supernatant was transferred to a fresh tube and kept at −80°C. Protein assay was done using Bio-Rad protein assay dye reagent (Bio-Rad, Cat. No. 5000006). Protein samples (40 µg/lane) were separated on 4%–20% Mini-PROTEAN TGX Precast protein gels (Bio-Rad, Cat. No. 4561095) and transferred to a polyvinylidene fluoride membrane using a Trans-Blot Turbo Transfer System from Bio-Rad (Cat. No. 1704150). After blocking the membrane with 5% nonfat milk in TBS-Tween, the membrane was exposed to primary antibody overnight at 4°C. After several washes with TBS-Tween, the membrane was incubated with horseradish peroxidase-conjugated anti-rabbit IgG (Promega, Cat. No. W401B) at room temperature. Luminol/enhancer solution (Bio-Rad Clarity Western ECL Substrate, Cat. No. 1705061) was used for protein detection. ChemiDoc MP Imaging System (Bio-Rad, Cat. No. 12003154) was used to detect the HRP signal. Protein signal quantification was carried out using ImageJ (http://rsbweb.nih.gov/ij/) and all data were adjusted with actin. Individual data points in any given group represent a kidney or a mouse.

Immunofluorescence

Freshly isolated kidneys were fixed in 10% neutral-buffered formalin overnight and provided to the Vanderbilt Translational Pathology Shared Resource for paraffin embedding and sectioning. Sections from four to five kidneys were deposited on glass slides. The tissue sections were then deparaffinized, rehydrated in ethanol of varying strengths, and then washed with PBS at room temperature. After antigen retrieval using Citra Plus antigen retrieval reagent (BioGenex, Cat. No. HK080-9K), sections were treated with primary antibody overnight at 4°C, followed by poly-HRP-conjugated secondary antibody for 30 min, rinsed in PBS, and then incubated with a tyramide-containing signal amplification solution for 2 min at room temperature. After a final wash, the tissue sections were mounted using ProLong DAPI-containing Gold antifade reagent (Invitrogen, Cat. No. P36931) and visualized on a Zeiss LSM 880 laser scanning confocal microscope. One kidney from each of the five mice treated with NaCl and one kidney from each of the six mice treated with NaHCO₃ for 7 days were utilized in the study. For each group (NaCl or NaHCO₃−treated group), one paraffin block was made that had one kidney from every mouse within the group. As a result, one section on the slide for the NaCl-treated group contained five kidney specimens from five different mice, and one section for the NaHCO₃-treated group contained six kidney specimens from the NaHCO₃-treated group. Slides from the NaCl and NaHCO₃-treated groups were stained simultaneously, and several photos from different samples within the group were taken. The excitation and emission wavelengths of the dye used were 647 and 665, respectively. All experiments were done in a blinded manner.

M-1 Cells

Mouse cortical collecting duct M-1 cells [American Type Culture Collection (ATCC), Cat. No. CRL-2038] were grown in 60-mm polystyrene cell culture dishes (Corning, Cat. No.430166) in DMEM-F12 medium (Gibco, Cat. No. 11320-033) supplemented with 10% fetal bovine serum (R&D Systems, Cat. No. S11110) and 90 U/mL penicillin, 90 mg/mL streptomycin (Gibco, Cat. No. 15140-122) in an incubator at 37°C, 5% CO₂. For Western blot analysis, cells were washed twice with 2 mL Hanks’ balanced salt solution (Gibco, Cat. No. 14025) and lysed for 20 min on ice with 100 mL RIPA buffer (Sigma, Cat. No. R0278) supplemented with protease inhibitor cocktail (Roche, Cat. No. 05892791001, 1 tablet in 5 mL), scraped from the dish, and placed in a 1.5 mL Eppendorf tube for another 30 min on ice. The cell lysate was then pelleted for 10 min at max speed at 4°C, and supernatant was recovered for protein assay and Western blot analysis. For testing the effects of bicarbonate and pH on the cells, culture media of the following compositions were prepared: 120 mM NaCl, 35 mM NaHCO₃ or HEPES, 4.2 mM K-gluconate, 0.8 mM CaNO₃, 0.71 mM MgSO₄., 0.5 mM Na₂HPO₄, 0.52 mM NaH₂PO₄, 0.5 mM Na-pyruvate, 1.75 mM glucose, and 80 μg/mL phenol red. The media were then supplemented with MEM amino acids (50×, Gibco, Cat. No. M5550), MEM vitamins (100×, Gibco, Cat. No. 11120052), Glutamax (100×, Gibco, Cat. No.35050-061), Penicillin/Streptomycin (100×, Gibco, Cat. No. 15140-122), and 10% fetal bovine serum (Atlanta Biologicals/R&D Systems).

qPCR

Mice were treated with 2% sucrose or 2% sucrose with 280 mM NaHCO₃ for 24 h. Kidneys were isolated and homogenized in TRIzol RNA isolation reagent (Invitrogen). After TRIzol extractions, the RNA was precipitated using isopropanol and washed twice with 70% ethanol. RNA was resuspended in diethyl pyrocarbonate (DEPC)-treated water and quantitated by measuring the optical density at 260 nm, 280 nm, and 320 nm. Because the concentration was high and variable from sample to sample, the RNA was diluted to a concentration range of 250–300 ng/μL and quantified a second time to improve precision. Reverse transcription was performed using Superscript III (Invitrogen, Cat. No. 18080051) following the manufacturer’s instruction. Reactions (20 μL) contained 1 µg RNA, 4 µL 5× buffer, 10 mM DTT, 2.5 µM Random hexamers (Invitrogen, Cat. No. 100026484), 1.25 mM MgCl₂, 250 nM dNTP, and 100 U Superscript III. After a final denaturation step, the reactions were diluted three times by adding 40 µL of water. The Applied Biosystems QuantStudio 3 Real-Time PCR system was used to carry out the quantitative PCR. Primers were utilized in two sets. One primer set covered exon 1a-2 while the other covered exon 1a-3 (Table 1). The components of the quantitative PCR reactions were 12.5 µL of SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA), 1 µL of each primer (1 µM), 7.5 µL of water, and 6 µL of cDNA. Results were expressed in relation to the housekeeping gene GAPDH. The PCR cycling conditions used were an initial denaturation step at 95°C for 2 min, followed by denaturation at 95°C for 30 s, annealing for 30 s at 65°C, and extension at 72°C for 30 s. These short steps were repeated or cycled 40 times. The melting curve analysis confirmed the presence of only one fragment for each primer set.

Table 1.

qPCR primers used to amplify KCC3a-specific transcripts

		Primer 5′…3′	Tmo	%GC
Exon 1a	Forward	ATGCATCCACCAGAAGCCAC	67.4	55
Exon 2	Reverse	CATTTCGGGCTTTTTTATGGCCGTC	73.4	48
Exon 3	Reverse	ATTCGTCTCCTTCTTCATAGTTGG	63.4	42

Statistics

As stated in the figure legends, ANOVA or two-tailed unpaired t tests were used to test the null hypothesis. This was done using GraphPad Prism 9. All data are presented as means ± SE with P < 0.05 considered significant.

RESULTS AND DISCUSSION

We observed previously that both KCC3a and pendrin abundance increase with 24 h of 280 mM NaHCO₃ in their drinking water (2). However, other studies have shown that while pendrin abundance initially increases with bicarbonate intake, it then returns to baseline after 7 days of the bicarbonate load (13). We, therefore, explored the longer-term effect of bicarbonate intake on KCC3a abundance by comparing mice treated with 280 mM NaCl (control) or NaHCO₃ in the drinking water for 7 days. Following 1 wk of NaHCO₃ treatment, we observed an increase in the abundance of both KCC3a and pendrin (Fig. 1, A and B). To verify that the increased expression observed by Western blot is related to cortical intercalated cells, we also immunostained kidney sections from the 7-day NaCl versus NaHCO₃ treatments and observed a significant increase in KCC3a signal at the apical membrane of cells showing a typical pattern of intercalated cells in the cortex (Fig. 1, C and D). Thus, our earlier report and the current observation suggest that KCC3a upregulation is essential for pendrin function to handle both short- and long-term bicarbonate challenges. Next, we sought to determine if triggering acute acidosis by NH₄Cl treatment, in contrast, decreases KCC3a abundance. To answer this question, KCC3a abundance was quantified in mice ingesting NH₄Cl or vehicle. Although chronic NH₄Cl treatment has been reported to lower pendrin abundance (14), we observed that 24 h of NH₄Cl treatment had no impact on either KCC3a or pendrin abundance (Fig. 2, A and B). We conclude that KCC3a and pendrin expression were not affected by an acute acid load. The difference in protein preparation may account for the observed difference with respect to pendrin. In contrast to Wagner et al. (13), who used membrane protein preparation for their protein quantification, we used crude protein preparation in our investigations, which comprises both membrane and intracellular proteins. Crude protein preparation is the most widely used technique for protein preparation in transporter investigations, including those involving the Na⁺/Cl⁻ cotransporter (NCC), Na⁺/K⁺/2Cl⁻ cotransporter (NKCC2), and KCC3a.

Figure 1.

Changes in KCC3a and pendrin abundance in response to long-term NaHCO₃. A and B: in contrast to NaCl treatment, a 7-day NaHCO₃ treatment increased KCC3a and pendrin abundances. For blot quantification, densitometric values were normalized to β-actin. Values are represented as means ± SE (two-tailed unpaired t test, *P < 0.05; ***P < 0.001). Individual data points in any given group represent one kidney or one mouse. C and D: Localization of KCC3a following the 7-day NaCl or NaHCO₃ treatment, respectively. NaHCO₃ causes an increase in KCC3a apical localization and intensity compared with NaCl. Scale bar = 20 µm.

Figure 2.

Short-term NH₄Cl and K-citrate treatments. A and B: a 24-h NH₄Cl treatment had no effect on expression of KCC3a and pendrin (two-tailed unpaired t test, ns, nonsignificant or P > 0.05). C: KCC3a abundance increased on K-citrate treatment compared with control (Ctrl) or KCl treatment (one-way ANOVA followed by Tukey’s multiple-comparison test, *P < 0.05). For blot quantification, densitometric values were normalized to β-actin. Values are represented as means ± SE. Individual data points in the given group represent one kidney or one mouse.

Next, we asked if the nature of the cation (K⁺ instead of Na⁺) or using a regular edible form of bicarbonate as opposed to direct bicarbonate treatment affected KCC3a abundance. Because citrate is the predominant organic anion associated with K⁺ in food such as fruits and vegetables (15), and humans consume K⁺ mostly as K-citrate, we tested high K-citrate in the diet. Note that citrate enters the Krebs cycle generating energy (ATP) and CO_2, which results in the production of bicarbonate (16). Thus, mice were fed either high KCl (280 mM) or high K-citrate (280 mM), and a significant increase in KCC3a expression was again observed with the K-citrate but not the KCl diet (Fig. 2C). These data indicated that it is the organic anion and not the cation that is responsible for the stimulation of KCC3a. In addition, because high K⁺ stimulates the release of aldosterone, our data with KCl supplementation indicate that the mineralocorticoid is not the major signaling component for the increased abundance of the cotransporter.

We previously reported that contraction alkalosis might be responsible for the increase in KCC3a abundance observed upon 24-h water deprivation (2). Under these conditions, angiotensin II (ANG II) (17), may have also been acting on KCC3a expression. To address this possibility, we evaluated KCC3a and pendrin abundance in mice given a high-salt (HS) diet with or without 280 mM KHCO₃ and a normal-salt (NS) diet with 280 mM KHCO₃ for 24 h. The HS diet suppresses ANG II release (18), yet we found that the HS diet with KHCO₃ increased both the abundance of KCC3a and pendrin (Fig. 3, A and B). We also reported that a diet low in Na⁺ similarly increased the abundance of KCC3a (2). We postulated that this was likely caused by volume contraction and alkalosis. Diets low in Na⁺ also stimulate the production of ANG II (17). To eliminate ANG II participation more directly, we treated mice with Losartan, an angiotensin II receptor blocker (19), and showed no effect on the abundance of KCC3a in mice fed either on a normal or on a Na⁺-deficient diet (Fig. 3, C and D). The change in KCC3a abundance is therefore unaffected by systemic inhibition of ANG II. Taken together with the previous observation, this further demonstrated that bicarbonate, not ANG II, is the primary signaling molecule.

Figure 3.

The main signaling molecule responsible for the increase in KCC3a abundance is bicarbonate. A and B: KCC3a and pendrin abundance were elevated by both a normal salt (NS) and high sodium (HS) KHCO₃ diet in comparison with a high-salt diet (HS) alone (one-way ANOVA followed by Tukey’s multiple-comparison test). C and D: KCC3a abundance on a normal or Na⁺-deficient diet was unaffected by losartan [angiotensin-II receptor type 1 (AT1) antagonist]. For blot quantification, densitometric values were normalized to β-actin. Values are represented as means ± SE. *P < 0.05; **P < 0.01. Individual data points in the given group represent one kidney or one mouse. ns, nonsignificant or P > 0.05.

Pendrin and KCC3a are both colocalized and elevated by bicarbonate treatment; therefore, if there is a direct relationship between pendrin and KCC3a, pendrin gene ablation should suppress the KCC3a stimulation. Another possibility is that if KCC3a is regulated separately from pendrin and thus KCC3a will still respond to bicarbonate in pendrin knockout mice. To further assess the relationship between KCC3a and pendrin, we examined the effect of bicarbonate on KCC3a expression in pendrin knockout mice. Bicarbonate treatment for 4 days increased KCC3a abundance in pendrin knockout mice, indicating that the exchanger is not necessary to regulate KCC3a abundance upon a bicarbonate challenge (Fig. 4A).

Figure 4.

KCC3a abundance is independent of pendrin, not upregulated at the mRNA level, and not sensitive to pH. A: compared with the normal diet group, KCC3a abundance increased in pendrin knockout (KO) mice maintained on a high KHCO₃ diet for 4 days (two-tailed unpaired t test). For blot quantification, densitometric values were normalized to β-actin. Individual data points in the given group represent one kidney or one mouse. B: KCC3a transcript levels were not affected by bicarbonate treatment. Curves are for GAPDH (green), KCC3-1a-2 (red), and KCC31a-3 (blue) indicating relative abundance of the transcripts. The real-time PCR Systems software computed a relative quantification (RQ) factor, based on the delta-delta Ct (2^−ΔΔCt). The control group (Ctrl) was then set to 100% and the fold change in gene expression in the NaHCO3-treated group was calculated. Data were normalized to the expression of GAPDH. Values are represented as means ± SE. *P < 0.05. Individual data points in the given group represent one kidney or one mouse. C: expression of KCC3a in M-1 cells, wild-type (WT), and KCC3 knockout (KO) kidney samples. D: 24-h exposure of M-1 cells to acidic to alkaline pHs does not affect KCC3a expression. E: 24-h exposure of M-1 cells to increasing amounts of bicarbonate stimulates KCC3 expression. ns, nonsignificant or P > 0.05.

Next, we tested if the increase in KCC3a abundance seen with bicarbonate intake involves increased transcription. To answer this question, we used two sets of primers specific to KCC3a and examined the effect of a bicarbonate challenge on KCC3a mRNA expression. As shown in Fig. 4B, KCC3a mRNA was unaffected by bicarbonate intake, which demonstrates that the bicarbonate challenge affects KCC3a protein abundance through a posttranscriptional mechanism.

This and previous studies show that the regulation of pendrin and KCC3a has some similarities, but also some differences. Although the abundance of both proteins increases in response to bicarbonate intake, the transcript of pendrin (Slc26a4), but not KCC3a, increases in this model (20). Moreover, while pendrin is regulated by aldosterone and angiotensin II (21, 22), the present study indicates that KCC3a is not. Finally, while pendrin abundance is severely decreased in KCC3 knockout mice (2), the reverse is not true as KCC3a expression was unaffected by the absence of the exchanger in the pendrin knockout mice. These latest observations suggest that pendrin function relies on KCC3a, whereas KCC3a functions independently of pendrin.

Because distinguishing bicarbonate from pH effects is challenging in vivo, we employed an in vitro cell culture model. In mouse cortical M-1 cells, which express intercalated cells markers like pendrin, V-ATPase, and Gpr116 (23, 24), we detected KCC3a expression (Fig. 4C). Thus, we used this cell line to assess the effect of alkaline pH and bicarbonate on KCC3a abundance. To do so, we varied pH and bicarbonate concentration independently. M-1 cells were cultured at different pHs at 37°C for 24 h in HCO₃⁻/CO₂-free, HEPES-buffered media. Conversely, we cultured these cells at different bicarbonate concentrations. As shown, varying pH over 24 h in the absence of bicarbonate did not affect KCC3a abundance (Fig. 4D), whereas increasing bicarbonate concentration markedly stimulated KCC3a abundance (Fig. 4E). This finding suggests that bicarbonate is the primary signaling molecule causing the increase in KCC3a abundance.

Like KCC3a and pendrin, the Na⁺-dependent Cl⁻/HCO₃⁻ exchanger AE4 (SLC4A9) is found predominantly in type-B cells (25, 26). Recent studies proposed that AE4 is a bicarbonate sensor at the basolateral membrane and an entry point for bicarbonate into β-intercalated cells (27). These studies also show that the exchanger is involved in the intercalated cell’s response to an alkaline load as under these conditions, pendrin expression is upregulated in wild-type mice but not in AE4 knockout mice (28). Thus, AE4 function is critical for the stimulating effect of alkalosis on pendrin. Whether the stimulation of KCC3a is also AE4-dependent remains to be determined, but it is likely as both KCC3a and pendrin expression are coregulated by a bicarbonate load. Carbonic anhydrases (CA), which catalyze the reversible hydration of CO₂ resulting in the production of bicarbonate, are likely to also contribute to the increased intracellular bicarbonate concentration observed in conditions of alkalemia. Isoforms such as CA2, CA15, and CA12 are expressed in B-type intercalated cells. How the intracellular bicarbonate concentration act on KCC3a expression is currently unknown. A protein possibly involved is soluble adenylyl cyclase (sAC) (29, 30). This enzyme is uniquely sensitive to bicarbonate (31).

DATA AVAILABILITY

All data generated or analyzed during the described study are included in this published article.

GRANTS

This work was supported by the National Institute of Diabetes and Digestive and Kidney grants R01DK093501 (to E.D.), and R01DK119793 (to S.M.W.), and NIH Office of the Director grant DP5OD033412 (to A.S.T.) and by Fondation Leducq Grant 17CVD05 (to E.D.).

DISCLOSURES

E. Delpire is an editor of the American Journal of Physiology-Cell Physiology and was 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

M.Z.F. and E.D. conceived and designed research; M.Z.F., A.S.T., S.M.W., and E.D. performed experiments; M.Z.F., A.S.T., S.M.W., and E.D. analyzed data; M.Z.F., A.S.T., R.K., S.M.W., and E.D. interpreted results of experiments; M.Z.F. and E.D. prepared figures; M.Z.F. and E.D. drafted manuscript; M.Z.F., A.S.T., R.K., S.M.W., and E.D. edited and revised manuscript; M.Z.F., A.S.T., R.K., S.M.W., and E.D. approved final version of manuscript.