Hypokalemia and Pendrin Induction by Aldosterone

Aldosterone plays an important role in regulating Na-Cl reabsorption and blood pressure. Epithelial Na+ channel, Na+-Cl− cotransporter, and Cl−/HCO3− exchanger pendrin are the major mediators of Na-Cl transport in the aldosterone-sensitive distal nephron. Existing evidence also suggests that plasma K+ concentration affects renal Na-Cl handling. In this study, we posited that hypokalemia modulates the effects of aldosterone on pendrin in hyperaldosteronism. Chronic aldosterone infusion in mice increased pendrin levels at the plasma membrane, and correcting hypokalemia in this model almost completely blocked pendrin upregulation. However, hypokalemia induced by a low-K+ diet resulted in pendrin downregulation along with reduced plasma aldosterone levels, indicating that both hypokalemia and aldosterone excess are necessary for pendrin induction. In contrast, decreased plasma K+ levels were sufficient to increase Na+-Cl− cotransporter levels. We found that phosphorylation of mineralocorticoid receptor that prevents aldosterone binding in intercalated cells was suppressed by hypokalemia, which resulted in enhanced pendrin response to aldosterone, explaining the coordinated action of aldosterone and hypokalemia in pendrin regulation. Finally, to address the physiological significance of our observations, we administered aldosterone to mice lacking pendrin. Notably, plasma K+ levels were significantly lower in pendrin knockout mice (2.7±0.1 mmol/L) than in wild-type mice (3.0±0.1 mmol/L) after aldosterone infusion, demonstrating that pendrin alleviates hypokalemia in a state of aldosterone excess. These data indicate that the decreased plasma K+ levels promote pendrin induction by aldosterone, which, in concert with Na+-Cl− cotransporter, counteracts the progression of hypokalemia but promotes hypertension in primary aldosterone excess.

A ccumulating data suggest that aldosterone plays a major role in the development of hypertension. Clinical studies have demonstrated that primary aldosteronism (PA) affects 5% to 20% of individuals with hypertension. 1,2 Moreover, increased aldosterone levels predispose normotensive individuals to hypertension. 3 Aldosterone and its receptor, the mineralocorticoid receptor (MR), regulate extracellular fluid volume and plasma K + levels by controlling ion transport processes in the distal nephron. Classical targets of aldosterone in this segment are the principal cells of the collecting duct, which express the amiloride-sensitive, epithelial Na + channel (ENaC). 4,5 Aldosterone regulates ENaC by increasing its apical membrane levels and channel activity. The latter process requires the proteolytic cleavage of the inhibitory domain in the extracellular loop by aldosterone-induced serine proteases. 4,6,7 The reabsorption of Na + through ENaC generates a lumen-negative transepithelial voltage, which provides the driving force for Cl − reabsorption and K + secretion. 5,8 In the collecting duct, transcellular Cl − transport occurs through pendrin in renal intercalated cells, [9][10][11] another cell type that comprises the collecting duct.
Pendrin, encoded by SLC26A4, is a Cl − /HCO 3 − exchanger expressed specifically in β-intercalated cells. Pendrin has been shown to regulate the acid/base balance by excreting HCO 3 − into the urine. 10 In addition to its role in alkali secretion, however, accumulating studies have demonstrated the key role of pendrin in maintaining extracellular fluid volume homeostasis by regulating Cl − reabsorption. 9,[11][12][13] Importantly, mice lacking the pendrin gene (Pds −/− ) show volume contraction, low blood pressure, and increased Cl − excretion compared with wildtype mice when they are on a low Na-Cl diet. 14,15 Cl − reabsorption in the cortical collecting duct disappears in Pds −/− mice, explaining the impaired ability to retain fluid. 11 In addition, double knockout of pendrin and Na + -Cl − cotransporter (NCC) present in the distal convoluted tubules (DCT) leads to severe volume depletion and hypotension, 16 demonstrating that NCC and pendrin play compensatory roles in Cl − and fluid volume conservation. Conversely, mice overexpressing pendrin show salt-sensitive hypertension and are very sensitive to Cl − intake. 17 Mineralocorticoids can upregulate pendrin, 12 and Pds −/− mice show blunted pressor responses to deoxycorticosterone 12 and aldosterone, 18 indicating that pendrin mediates fluid volume regulation by aldosterone. We previously showed that this process is modulated by MR phosphorylation at S843 in the ligand-binding domain (MR S843-P ), which prevents ligand binding and MR signaling in intercalated cells. 19 The pathophysiological roles of MR S843-P in hypertension have remained unknown.
The plasma K + concentration is increasingly recognized as an important regulator of renal ion transport mechanisms. In DCT cells, plasma K + levels alter NCC phosphorylation levels independently of aldosterone, [20][21][22][23] which is implicated in the inverse relationship between dietary K + intake and blood pressure levels. 24,25 Stimulation of NCC by hypokalemia also plays a pathogenic role in PA. 26 In addition to NCC, previous findings have implicated K + intake in regulating pendrin expression, 19,27 and we have shown that dietary K + loading decreases pendrin levels. 19 To date, the role of the plasma K + concentration in regulating pendrin remains unclear, especially in the context of hyperaldosteronism. In this study, we tested the hypothesis that low plasma K + levels potentiate the effect of aldosterone on pendrin expression, thereby counteracting the progression of hypokalemia in a state of aldosterone excess.

Animals
This study was approved by the institutional review board in the Teikyo University Review Board No. 14-018. Male C57BL/6 mice and Sprague-Dawley rats were obtained from Tokyo Laboratory Animals Science (Japan). Pds −/− mice with 129/Sv background were developed by Everett et al 12,28 and were obtained from Jackson Laboratory (United States). Wild-type mice with 129/Sv background were also obtained from Jackson Laboratory. The details of the study protocol are described in Figure S1 in the online-only Data Supplement and are also available in the extended Materials and Methods in the online-only Data Supplement.

Western Blot
Western blotting was performed as described previously. 29 Plasma membrane fraction was purified from total kidneys using plasma membrane isolation kit (Minute, Invent Biotechnologies, United States). Enrichment of plasma membrane proteins but not cytoplasmic proteins was validated in the laboratory ( Figure S2). Tubulin (for total cell lysates) and Coomassie brilliant blue staining of the gel (for plasma membrane proteins; Figure S3) were used to monitor the identical loading of different samples. 30,31 The rest of the Methods is available in the online-only Data Supplement.

Statistical Analysis
The data are summarized as means±SEM. Unpaired t test was used for comparisons between 2 groups. For multiple comparisons, statistical analysis was performed by ANOVA followed by Tukey post hoc tests. P values <0.05 was considered statistically significant.

Pendrin and NCC Induction by Aldosterone Is Ameliorated by K + Supplementation
Excessive aldosterone production in PA is associated with decreased plasma K + levels. Interestingly, however, it is also known that only a subpopulation of PA patients actually develops hypokalemia. 2 We investigated whether a plasma K + decrease mediates the effects of aldosterone in the distal nephron. Mice received a continuous infusion of aldosterone (60 ng/h) for 2 weeks; 1 group of mice received K + supplementation via drinking water to prevent a decrease in plasma K + levels (study 1). Aldosterone infusion for 2 weeks caused a significant decrease in plasma K + levels (3.1±0.1 mmol/L in aldosterone-infused mice versus 4.5±0.1 mmol/L in controls; P<0.001), an effect that was blocked by K + supplementation (4.6±0.1 mmol/L; not significant versus control group; Figure 1A). Water intake and urinary volumes increased, whereas urinary pH decreased by K + supplementation in aldosterone-infused mice (Table S1).
ENaC activation by aldosterone involves proteolytic cleavage of the inhibitory domain, which reduces the molecular weight of ENaCγ from 85 kDa (uncleaved form) to 70 kDa (active, cleaved form). 4,32 Western blot analysis of an isolated plasma membrane fraction revealed that the levels of pendrin, NCC, and the active, cleaved form of ENaCγ were elevated in aldosterone-infused mice (by 1.9-, 2.3-, and 2.2-fold for pendrin, NCC, and cleaved ENaCγ, respectively; P<0.001 versus control; Figure 1B). Notably, the correction of hypokalemia by K + supplementation almost completely abolished the upregulation of pendrin (1.1-fold increase; P=0.471 versus control and P<0.001 versus aldosterone infusion group; Figure 1B). In aldosterone-infused mice, K + intake also reduced plasma membrane NCC levels. In contrast, K + loading did not decrease but instead enhanced ENaCγ cleavage (by 3.4-fold versus control; P<0.001). Immunofluorescence analysis revealed that pendrin was strongly expressed at the apical membrane of intercalated cells in the collecting duct in aldosterone-infused mice, which was profoundly attenuated by K + supplementation ( Figure 1C; Figure S4). These data indicate that high aldosterone is sufficient to activate ENaCγ, whereas the decrease in plasma K + is important for the induction of pendrin and NCC induction by aldosterone.
We also evaluated whether the well-described, blood pressure-lowering effects of oral K + intake 25 were associated with the changes in pendrin and NCC, using a rat model of hypertension. Uninephrectomized rats that received aldosterone (0.75 μg/h) showed a progressive increase in blood pressure, which was ameliorated by preventing hypokalemia (170±2 mm Hg in aldosterone-infused rats versus 155±1 mm Hg in aldosterone-infused rats that received K + at 4 weeks; P<0.01; Figure S5A), consistent with previous observations using similar models. 33 In addition, urinary albumin excretion tended to be reduced by K + supplementation (4.5±1.6 versus 1.4±0.3 mg/d; P=0.079). In the plasma membrane fraction of the kidney, we found that K + supplementation abolished the increases in pendrin and NCC expression, but not in cleaved ENaCγ in this model ( Figure S5B).

Reversal of Hypokalemia by Amiloride Abrogates Pendrin Induction by Aldosterone
The downregulation of pendrin by K + supplementation in the aldosterone infusion model suggests the involvement of plasma K + levels or oral K + intake. Previous studies have also suggested a role for Cl − in pendrin regulation. 34 To clarify the mechanism underlying our observations and to test whether changes in the plasma K + concentration mediate the effects of aldosterone on pendrin, aldosterone-infused mice were treated with amiloride, a K + -sparing diuretic (15 mg/L in drinking water) that inhibits ENaC but does not directly affect pendrin (study 2). As expected, amiloride prevented the decrease in the plasma K + levels in aldosterone-infused mice (3.4±0.3 mmol/L in aldosterone-infused mice versus 4.7±0.1 mmol/L in aldosterone plus amiloride group; P<0.001). In the plasma membrane fraction, pendrin upregulation in aldosteroneinfused mice was fully blocked by amiloride ( Figure S6). These data demonstrate that hypokalemia mediates the induction of pendrin by aldosterone.

Hypokalemia Induced by K + Restriction Increases NCC but Decreases Pendrin Expression
The results described above indicate that aldosterone and plasma K + play a role in the regulation of pendrin. Given that hypokalemia increases NCC phosphorylation independently of aldosterone, 21,23,26 we tested whether hypokalemia could directly induce pendrin expression in the absence of hyperaldosteronism by feeding mice a low-K + diet (containing 0.3% Na-Cl, which is the Na-Cl content in a normal-K + diet; study 3). When wild-type mice were changed from a normal-K + diet to low-K + diet, they showed significant decrease in plasma K + levels (3.1±0.1 mmol/L for the low-K + diet versus 4.6±0.1 mmol/L for the normal-K + diet; P<0.0001; Figure 2A). Consistent with previous reports showing an increase in NCC phosphorylation by K + depletion, 21,23,26 NCC levels in the plasma membrane were increased by a low-K + diet (by 1.5-fold; P=0.0121; Figure 2B and 2C). In contrast, a low-K + diet resulted in the significant decrease in pendrin levels (P<0.001). Consistent with the changes in pendrin levels, urinary pH tended to be decreased by the low-K + diet (7.32±0.24 in normal-K + diet group versus 6.82±0.07 in low-K + diet group; P=0.083). The levels of cleaved ENaCγ and the ratio of cleaved versus total ENaCγ were also decreased in the low-K + group. Low-K + diet resulted in the significant decrease in plasma aldosterone levels ( Figure 2D), likely explaining the reduced levels of pendrin and cleaved ENaCγ. Accordingly, immunofluorescence microscopy revealed that the apical staining of NCC was intensified whereas that of pendrin was reduced in mice fed a low-K + diet relative to controls ( Figure 2E and 2F; Figure S4). These data indicate that the decrease in plasma K + not only induces NCC phosphorylation, but also enhances plasma membrane NCC expression. In contrast, hypokalemia alone is not sufficient to induce pendrin expression, which also requires hyperaldosteronism.

Hypokalemia Decreases MR Phosphorylation at S843
Our results support a mechanism in which low plasma K + levels potentiate the effect of aldosterone in intercalated cells, resulting in pendrin induction. We previously demonstrated that phosphorylation of MR at S843 (MR S843-P ) prevents ligand binding selectively in intercalated cells, regulating MR signaling. 19 Given that this phosphorylation is increased by hyperkalemia, 19 we expected that K + depletion would lead to The amount of protein loaded on the gel was monitored using Coomassie brilliant blue (CBB) staining. C, Kidney sections stained for α-pendrin (green) and DAPI (blue) in the indicated animals. Pendrin staining is increased in the apical membrane in aldosterone-infused mice, which is blocked by K + supplementation (see also Figure S4, which shows low-power field images). Bar represents 50 μm. Data are expressed as means±SEM; **P<0.01. Aldo indicates mice received aldosterone infusion (n=9); Aldo+K, mice received aldosterone and 1% KCl in the drinking water (n=10); Ctrl, control mice received sham operation (n=7); and ENaC, epithelial Na + channel. dephosphorylation of MR, producing receptor competence. To test this hypothesis, we evaluated MR S843-P levels in mice eating a low-K + diet. Western blot analysis revealed that the low-K + diet caused a significant decrease in MR S843-P (a 46% reduction; P=0.004; Figure 3), with no change in total MR.

Low Plasma K + Levels Potentiate Pendrin Induction by Aldosterone
Given that K + depletion reduced the phosphorylation of MR, converting to the active, competent form, we expected that aldosterone-mediated pendrin induction would be increased under these conditions. However, the biological effect might not be apparent in mice that do not exhibit elevated aldosterone levels. We therefore evaluated the effects of exogenous aldosterone infusion on pendrin levels in mice eating a low-K + diet (study 4). Because changes in plasma K + levels during the experiment could alter MR S843-P levels and thereby confound the analysis, we performed experiments to determine the dose of aldosterone that did not alter the plasma K + levels. As shown in Figure 4A, the plasma K + levels were not affected 7 days after the low-dose aldosterone (12 ng/h) in mice eating a normal-K + diet or on a low-K + diet. However, the low-dose aldosterone infusion significantly increased plasma aldosterone levels both in normal-K + and low-K + groups ( Figure 4B).
In the plasma membrane fraction of the kidney, the lowdose aldosterone produced a modest but significant increase in cleaved ENaCγ levels in mice fed normal-K + and low-K + diets ( Figure 4C and 4D), confirming the effects of aldosterone in the distal nephron. The magnitude of this effect was comparable between the groups. In contrast, although low-dose aldosterone increased pendrin expression by as much as 2.8-fold in mice on a low-K + diet (P=0.008; Figure 4C and 4E), it did not alter pendrin levels in mice on a normal-K + diet (P=0.31). Moreover, MR S843-P levels were lower in the low-K + group of aldosterone-infused mice ( Figure S7). These data demonstrate that pendrin induction by aldosterone is influenced by plasma K + levels and are consistent with hypokalemia promoting the Figure 2. K + restriction increases Na + -Cl − cotransporter (NCC), whereas it decreases pendrin. A, Plasma K + concentrations in mice eating a normal-K + or a low-K + diet (n=9). B, Effect of low-K + diet on plasma membrane expression of the indicated proteins in the kidney. Blots show biological replicates. The amount of protein loaded on the gel was monitored using Coomassie brilliant blue (CBB) staining. C, Quantification of the expression levels described in (B). D, Plasma aldosterone concentrations in mice eating a normal-K + or a low-K + diet. E and F, Kidney sections stained for α-NCC (E, red) or α-pendrin (F, green) and DAPI (blue) in the indicated animals. Apical staining of NCC is increased, whereas that of pendrin is decreased in mice eating a low-K + diet (see also Figure S4). Bars represent 50 μm. Data are expressed as means±SEM; *P<0.05; **P<0.01. ENaC indicates epithelial Na + channel.

Hypokalemia Induced by Aldosterone Is Aggravated in Pendrin Knockout Mice
Recent studies indicate that the stimulation of NCC by hypokalemia serves as a mechanism to conserve K + by reducing distal Na + reabsorption and K + excretion. 26,35 However, compensatory actions of pendrin and NCC in renal K + handling have been implicated by findings in subjects with Pendred syndrome. 36 This autosomal recessive disorder results from a loss-of-function mutation in SLC26A4 (encoding pendrin) and is characterized by sensorineural hearing loss and thyroid abnormality. Remarkably, treatment with thiazide induces severe hypokalemia in this disorder, 36 indicating the role of pendrin (as well as NCC) in controlling plasma K + levels. Given these data, we expected that pendrin induction in a state of aldosterone excess might act to prevent the progression of hypokalemia.
To test this possibility, wild-type and pendrin knockout (Pds −/− ) mice (which show normokalemia at baseline) 11,28 received continuous infusion of aldosterone (60 ng/h) for 2 weeks (study 5). Urinary pH was significantly lower in Pds −/− mice receiving aldosterone than that in wild-type mice receiving aldosterone ( Figure 5A), consistent with the role of pendrin in excreting HCO 3 − . Of note, after 2 weeks of aldosterone infusion, Pds −/− mice showed significantly lower levels of plasma K + than did wild-type mice (2.7±0.1 versus 3.0±0.1 mmol/L; P=0.01; Figure 5B). These data are consistent with pendrin counteracting the progression of hypokalemia in hyperaldosteronism.

Discussion
We demonstrate that pendrin expression in intercalated cells is regulated by the interaction of aldosterone and plasma K + concentrations. An increase in plasma aldosterone levels stimulates ENaC in principal cells. In DCT cells, hypokalemia directly increases NCC levels independently of aldosterone. However, unlike the regulation of ENaC or NCC, we found that neither aldosterone nor hypokalemia alone was sufficient to induce pendrin; hypokalemia allowed the pendrin induction . Hypokalemia potentiates pendrin induction by aldosterone. A, Plasma K + concentrations in the indicated animals. Plasma K + levels were unchanged by low-dose aldosterone in normal-K + and low-K + groups. NS indicates not significant. B, Plasma aldosterone concentrations in the indicated animals. Plasma aldosterone levels were significantly increased by low-dose aldosterone in both normal-K + and low-K + groups. C, Effect of aldosterone on plasma membrane levels of the indicated proteins in the kidney. Blots show biological replicates. The amount of protein loaded on the gel was monitored using Coomassie brilliant blue (CBB) staining. D and E, Quantification of cleaved epithelial Na + channel (EnaCγ; D) and pendrin (E) in the plasma membrane described in (C). Data are expressed as means±SEM; *P<0.05; **P<0.01. Low K indicates mice on a low-K + diet; Low K+low aldo, mice on a low-K + diet received low-dose (12 ng/h) aldosterone; Normal K, mice on a normal-K + diet; and Normal K+low aldo, mice on a normal-K + diet received low-dose (12 ng/h) aldosterone (n=6 each group). by aldosterone in intercalated cells (Figure 6), which was associated with MR S843-P dephosphorylation. Our model indicates that variations in plasma K + levels have a major effect on the response of pendrin to aldosterone and that pendrin induction prevents the progression of hypokalemia in hyperaldosteronism (as discussed below).
Recent studies have demonstrated that pendrin is a key component of the electroneutral Na-Cl reabsorption machinery in the collecting duct. 16,19,[37][38][39] In this segment, Cl − transport through pendrin is coupled to Na + reabsorption through ENaC ( Figure S8). 13,15,18 Moreover, evidence indicates that pendrin stimulates ENaC activity and abundance. 13,15,18 In addition to this mechanism, pendrin can also work in tandem with Na + -dependent, Cl − /HCO 3 − exchanger in β-intercalated cells, 37,39 partly mediating Na-Cl reabsorption in this segment. Electroneutral Na-Cl reabsorption in the collecting duct eliminates the lumen-negative potential that drives the K + secretion through renal outer medullary K + channel, thereby reducing renal K + secretion. 38,40 These data also have important clinical implication for patients with PA. In this study, we showed that the coexistence of hyperaldosteronism and low plasma K + levels stimulates pendrin, and that the genetic knockout of pendrin exacerbates hypokalemia in aldosterone infusion model. These data, together with the previous evidence that Pds −/− mice show a blunted pressor response to aldosterone, 18 indicate that the pendrin induction prevents the progression of plasma K + decrease, but instead promotes hypertension in aldosterone excess. In subjects with PA, it has been demonstrated that only a subgroup of cases show hypokalemia. 2 Recent studies have suggested a role of NCC in this process, which decreases distal Na + reabsorption and K + excretion. 26,35 In addition to NCC, however, our data highlight the key role of pendrin in counteracting hypokalemia. This study suggests that the decrease in plasma K + levels potentiates pendrin induction in hyperaldosteronism, which, in concert with NCC, alleviates the progression of hypokalemia. In addition to the effects of pendrin and NCC in counteracting hypokalemia, decreased plasma K + levels may blunt aldosterone secretion in PA, mitigating ENaC activation and renal K + secretion.
A potential limitation of this study is that we did not demonstrate the mechanisms by which plasma K + modulates MR S843-P levels in intercalated cells. In DCT cells, hypokalemia alters the membrane potential by influencing K + exit likely through the inwardly rectifying K + channel Kir4.1 (encoded by KCNJ10). 23,41 Given that there is scant evidence for K + conductive pathways at the basolateral membrane in β-intercalated cells, 42 distinct mechanisms are likely to be involved. Another potential limitation is that it is unclear from the study what levels of plasma aldosterone induce pendrin expression in hypokalemia, although the data demonstrate that both aldosterone and hypokalemia are necessary for pendrin upregulation.

Perspectives
Our study demonstrates that pendrin expression in intercalated cells is regulated by the interaction of aldosterone and plasma K + . This mode of regulation is distinct from those of NCC in DCT cells or ENaC in principal cells, illustrating the diverse mechanisms underlying renal Na-Cl transport. Reduced plasma K + levels potentiate pendrin induction by aldosterone, which, along with NCC induction by low plasma K + , counteracts the progression of hypokalemia. Given the role of pendrin Figure 6. Regulation of epithelial Na + channel (ENaC) and pendrin by aldosterone in the collecting duct. In a state of aldosterone excess, aldosterone binds to mineralocorticoid receptor (MR) and activates ENaC in principal cells, which promotes electrogenic Na + reabsorption and lumen-negative potential in the collecting duct. This drives K + secretion through apical K + channel renal outer medullary K + channel (ROMK) in principal cells, resulting in the decrease in plasma K + levels. Hypokalemia causes MR S843-P dephosphorylation in β-intercalated cells, allowing aldosterone binding to MR in these cells and increasing pendrin at the plasma membrane. and NCC in fluid volume regulation, these mechanisms prevent plasma K + decrease but promote hypertension. These data are also relevant to the well-described, however, poorly characterized antihypertensive effects of high K + intake. Hyperaldosteronism is a complex disorder associated with hypokalemia and metabolic alkalosis. The pathogenic role of acid/base disturbance in a state of aldosterone excess may require further evaluation. The identity of the kinase regulating MR S843-P is the area of future research.