Angiotensin II increases activity of the ClC-K2 Cl− channel in collecting duct intercalated cells by stimulating production of reactive oxygen species

The renal collecting duct plays a critical role in setting urinary volume and composition, with principal cells transporting Na+ and K+ and intercalated cells mediating Cl− reabsorption. Published evidence implies Angiotensin II (Ang II) is a potent regulator of the collecting duct apical transport systems in response to systemic volume depletion. However, virtually nothing is known about Ang II actions on the basolateral conductance of principal and intercalated cells. Here, we combined macroscopic and single channel patch clamp recordings from freshly isolated mouse collecting ducts with biochemical and fluorescence methods to demonstrate an acute stimulation of the basolateral Cl− conductance and specifically the ClC-K2 Cl− channel by nanomolar Ang II concentrations in intercalated cells. In contrast, Ang II did not exhibit measurable effects on the basolateral conductance and on Kir4.1/5.1 potassium channel activity in principal cells. Although both Ang II receptors AT1 and AT2 are expressed in collecting duct cells, we show that AT1 receptors were essential for stimulatory actions of Ang II on ClC-K2. Moreover, AT1R−/− mice had decreased renal ClC-K2 expression. We further demonstrated that activation of NADPH oxidases is the major signaling pathway downstream of Ang II-AT1R that leads to stimulation of ClC-K2. Treatment of freshly isolated collecting ducts with Ang II led to production of reactive oxygen species on the same timescale as single channel ClC-K2 activation. Overall, we propose that Ang II-dependent regulation of ClC-K2 in intercalated cells is instrumental for stimulation of Cl− reabsorption by the collecting duct, particularly during hypovolemic states.

Naghmeh Hassanzadeh Khayyat ‡ , Oleg Zaika ‡ , Viktor N. Tomilin, Kyrylo Pyrshev, and Oleh Pochynyuk* From the Department of Integrative Biology and Pharmacology, The University of Texas Health Science Center at Houston, Houston, Texas, USA

Edited by Mike Shipston
The renal collecting duct plays a critical role in setting urinary volume and composition, with principal cells transporting Na + and K + and intercalated cells mediating Cl − reabsorption. Published evidence implies Angiotensin II (Ang II) is a potent regulator of the collecting duct apical transport systems in response to systemic volume depletion. However, virtually nothing is known about Ang II actions on the basolateral conductance of principal and intercalated cells. Here, we combined macroscopic and single channel patch clamp recordings from freshly isolated mouse collecting ducts with biochemical and fluorescence methods to demonstrate an acute stimulation of the basolateral Cl − conductance and specifically the ClC-K2 Cl − channel by nanomolar Ang II concentrations in intercalated cells. In contrast, Ang II did not exhibit measurable effects on the basolateral conductance and on K ir 4.1/5.1 potassium channel activity in principal cells. Although both Ang II receptors AT 1 and AT 2 are expressed in collecting duct cells, we show that AT 1 receptors were essential for stimulatory actions of Ang II on ClC-K2. Moreover, AT 1 R −/− mice had decreased renal ClC-K2 expression. We further demonstrated that activation of NADPH oxidases is the major signaling pathway downstream of Ang II-AT 1 R that leads to stimulation of ClC-K2. Treatment of freshly isolated collecting ducts with Ang II led to production of reactive oxygen species on the same timescale as single channel ClC-K2 activation. Overall, we propose that Ang II-dependent regulation of ClC-K2 in intercalated cells is instrumental for stimulation of Cl − reabsorption by the collecting duct, particularly during hypovolemic states.
Hypertension is one of the major causes of morbidity and mortality affecting approximately 46% of US adults, with blood pressure in 50% of hypertensive individuals exhibiting a saltsensitive pattern (1,2). Elevated blood pressure is commonly caused by expansion of the circulating volume due to salt retention by the kidney (3). Variations in dietary salt intake regulate transport in the renal collecting duct via the renin-angiotensin-aldosterone system to shape urinary NaCl excretion and to maintain circulating volume (4,5). The collecting duct is composed of electrically uncoupled principal and intercalated cells (6,7). Principal cells perform electrogenic Na + reabsorption via the epithelial Na + channel (ENaC) localized to the apical membrane and the Na + /K + ATPase on the basolateral membrane (5,8,9). Intercalated cells are essential for maintaining acid-base balance by secreting H + via the apical V-ATPase (A-type) and HCO 3 − via pendrin (SLC26A4) in the B-type (7). In addition, both A-and B-types have the capacity to reabsorb Cl − even when ENaC activity is blocked with amiloride (10). Since both types of intercalated cells are involved, it is viewed that Cl − reabsorption could occur with little or no changes in net acid or base secretion (11). The long-standing paradigm suggests that Ang II-driven secretion of the mineralocorticoid aldosterone from adrenal gland leads to upregulation of the ENaC-dependent Na + reabsorption in the collecting duct during the volume-depleted states (6). However, cumulative evidence demonstrates aldosterone-independent direct actions of Ang II on Na + and Cl − transport in the collecting duct during variations in salt intake and in the pathophysiology of Ang-dependent hypertension (10,(12)(13)(14). In fact, kidneys have substantial capacity to locally produce Ang II. In the experimental animal models of Ang II-induced hypertension (15,16), intrarenal Ang II levels become much higher (over 100-fold) than those in plasma (17)(18)(19). Ang II binds to AT 1 and AT 2 receptors to exert its numerous physiological actions. Activation of AT 1 R promotes proliferation, vasoconstriction, antinatriuresis, salt appetite, etc. (20)(21)(22)(23)(24). AT 2 R antagonizes the actions of AT 1 R resulting in vasodilation, natriuresis, and prostaglandin release (20,22,25). Both AT 1 R (most abundantly AT 1a R isoform in mice) and AT 2 R are expressed at the apical and basolateral sides of the collecting duct cells, although AT 2 R expression is considerably lower (25)(26)(27). Chronic Ang II infusion stimulates ENaC activity well above the physiological range of regulation (13), which cannot be effectively inhibited by mineralocorticoid receptor blockade (28). Ang II also increases Cl − reabsorption, in part by stimulating apically localized HCO 3 − /Cl − exchanger pendrin in B-type intercalated cells (10). At the same time, the actions of Ang II on the basolateral conductance of the collecting duct cells are not known.
Basolateral electrical conductance of the collecting duct principal cells is almost exclusively K + selective (29,30). The most prevalent heteromeric inward rectifying K ir 4.1/5.1 40 pS potassium channel is essential for K + recycling to set up a strong hyperpolarizing resting potential on the basolateral membrane around −70 mV to establish a favorable driving force for ENaC-mediated Na + reabsorption (29,31). K ir 4.1/5.1 is also expressed in the upstream segments, most notably the distal convoluted tubule, to control NaCl reabsorption via thiazide-sensitive NCC cotransporter (32,33). Loss-of-function mutations in the KCNJ10 gene encoding K ir 4.1 subunit result in EAST/SeSAME syndrome, a complex electrolyte imbalance disorder manifested as hypotension, natriuresis, hypocalciuria, hypomagnesemia, and hypokalemic metabolic alkalosis (34,35). Consistently, Kcnj16 deletion encoding the K ir 5.1 subunit ameliorated the development of salt-sensitive hypertension in Dahl SS rats (36).
Intercalated cells of the collecting duct do not express Na + /K + ATPase and have no electrogenic basolateral potassium conductance (7,30). Instead, activity of the ClC-K2 chloride channel determines basolateral Cl − transport and sets the resting potential around −20 mV (37,38). Similarly to K ir 4.1/5.1, ClC-K2 is also expressed in the distal nephron segments, namely, the thick ascending limb and distal convoluted tubule (37,39). Inactivating mutations in the CLCNKB gene (encoding ClC-Kb, human version of ClC-K2) underlie Bartter's syndrome type III associated with hypotension, hypochloremia, and metabolic alkalosis (40)(41)(42). Of note, K ir 4.1/5.1 and ClC-K2 are expressed in all cells of the thick ascending limb and distal convoluted tubule, whereas they are separated to principal and intercalated cells of the collecting duct, respectively (43). It is possible that such mosaic architecture of the collecting duct allows independent cell typespecific regulation of Na + and Cl − transport by endocrine signals, such as Ang II.
The major focus of the current study was to explore the functional consequences and uncover the molecular mechanisms of Ang II actions on K ir 4.1./5.1 potassium and ClC-K2 chloride conductance in native collecting duct cells. Figure 1. Ang II increases basolateral chloride currents in intercalated cells of the collecting duct. A, representative macroscopic currents in individual principal cells in response to voltage steps from −90 to +60 from the holding potential of −60 mV in the control (black) and following treatment with Ang II (500 nM) for 3 min (gray). A micrograph of a typical isolated collecting duct shown on top. The expression of AQP2, a marker of principal cells (highlighted with yellow arrows), is shown with pseudocolor green. Nuclear marker DAPI is shown with pseudocolor blue. The scale bar represents 70 μM. B, currentvoltage (I-V) relations of the basolateral K + -selective conductance obtained from voltage step protocols as shown in A in the control (black) and upon treatment with Ang II (gray). The number of individual recordings is shown. Measurements were done from at least three different mice. Both SEM (smaller bars) and SD (larger bars) are shown for each measured value. C, representative macroscopic currents in individual intercalated cells in response to voltage steps from −90 to +60 from the holding potential of −60 mV in the control (black) and following treatment with Ang II (500 nM) for 3 min (gray). A micrograph of a typical isolated collecting duct shown on top. The expression of ClC-K2, a marker of intercalated cells (highlighted with white arrows), is shown with pseudocolor red. Nuclear marker DAPI is shown with pseudocolor blue. The scale bar represents 70 μM. D, current-voltage relations of the basolateral Cl − -selective conductance obtained from voltage step protocols as shown in A in the control (black) and upon treatment with Ang II (gray). The number of individual recordings is shown. Measurements were done from at least three different mice. Both SEM (smaller bars) and SD (larger bars) are shown for each measured value. * -significant change (p < 0.05) versus respective control (one-way ANOVA). DAPI, 4',6-diamidino-2-phenylindole.

Ang II increases ClC-K2-dependent basolateral conductance in intercalated collecting duct cells
The renal collecting duct is a heterogeneous nephron segment containing electrically uncoupled principal and intercalated cells exhibiting different morphology and physiological functions (6,7). We first used patch clamp electrophysiology in freshly isolated collecting duct to assess Ang II actions on the basolateral conductance in principal and intercalated cells. Since electrical conductance of the apical membrane is much lower than the conductance of the basolateral membrane for both cell types (29,44), the changes in macroscopic whole cell current chiefly reflect alterations in the electrical conductance of the basolateral membrane. Figure 1A shows representative macroscopic currents from aquaporin type 2 (AQP2)-positive principal cells of freshly isolated collecting ducts before and after application of Ang II (500 nM for 3 min). The respective current-voltage relations demonstrate notable inward rectification and reversal around −70 mV (Fig. 1B), which is characteristic of the K + -selective conductance via K ir 4.1/5.1 channel, as we and others have reported previously (29,31). However, we did not observe any significant changes in the amplitude of the K ir 4.1/5.1-mediated K + current in principal cells after treatment with Ang II (Fig. 1, A and B).
The AQP2-negative intercalated cells exhibited anionselective conductance with a reversal around −20 mV (Fig. 1, C and D), which is mediated by the ClC-K2 Cl − channel on the basolateral membrane (45). Of importance, application of Ang II (500 nM for 3 min) significantly increased the amplitude of the Cl − -dependent current by almost 2-fold. These results show that Ang II increases the basolateral conductance specifically in the intercalated cells of the collecting duct.
We next assessed the effects of Ang II on the basolateral conductance of collecting duct cells at the single channel level. As shown in the representative experiment ( Fig. 2A) and the summary graph (Fig. 2B), application of Ang II (500 nM) did not affect the open probability of the 40 pS K ir 4.1/5.1 channel, the dominant K + channel in the basolateral membrane of the principal cells (29,31). In contrast, Ang II significantly increased the open probability of the 10 pS ClC-K2 channel (46) in a reversible manner in intercalated cells (Fig. 3A). As summarized in Figure 3B, the mean open probability was 0.32 ± 0.06, 0.52 ± 0.05, and 0.31 ± 0.06 in the control, after Ang II application, and following washout with control medium, respectively. Ang II increased the ClC-K2 open probability in a dose-dependent manner. As shown in Figure 3C, Ang II concentrations higher than 5 nM exhibited a significant stimulatory effect on single channel ClC-K2 activity. It is worth mentioning that similar levels of interstitial Ang II were reported in the kidney (18) arguing for the physiological Ang II stimulates ClC-K2 in intercalated cells relevance of Ang II actions on ClC-K2-dependent Cl − conductance in the intercalated cells.
Ang II acts on AT 1 receptor to regulate ClC-K2 activity and expression in the collecting duct Expression of both AT 1 and AT 2 receptors was reported in the collecting duct cells (25)(26)(27). Thus, we next tested which receptor types are instrumental in transducing stimulatory Ang II actions on ClC-K2. Pretreatment with AT 1 R blocker, losartan (1 μM for 3 min), had no effect on basal ClC-K2 open probability but precluded activation of the channel by Ang II (Fig. 4A). As summarized in Figure 4B, the mean open probability was 0.32 ± 0.04, 0.31 ± 0.04, and 0.29 ± 0.04 in the control, after pretreatment with losartan, and following Ang II in the presence of losartan, respectively. Stimulation of AT 2 receptors with a selective agonist, CGP42112 (100 nM for 3 min), did not change the ClC-K2 open probability being 0.31 ± 0.06 and 0.32 ± 0.06 in the control and after the agonist, respectively (Fig. 4C). Moreover, stimulation of Mas receptors with Ang 1-7 also had no effect on ClC-K2 activity (Fig. S2). The mean P o was 0.34 ± 0.05, 0.34 ± 0.05, and 0.33 ± 0.05 in the control, following application of Ang 1-7 (500 nM for 3 min), and washout with control medium, respectively. Altogether, the results in Figure 4 strongly suggest the dominant role for AT 1 R receptor in ClC-K2 activation by Ang II in intercalated cells.
We further tested the significance of AT 1 R on renal ClC-K2 expression in mice with genetic deletion of this receptor type. As shown on the representative Western blot in Figure 5A and summarized in Figure 5B, the intensity of the ClC-K-reporting signal was nearly halved in kidney homogenates from AT 1 R −/− compared with that of WT mice. However, it has to be mentioned that the ClC-K antibodies recognize both ClC-K1 (expressed in the thin ascending limb (47)) and ClC-K2 (expressed in the distal nephron segments (43)). Regardless, the results in Figure 5 indicate that AT 1 R at least partially controls ClC-K2 expression in the kidney.

Ang II increases ClC-K2 activity in the intercalated cells by activating NOX signaling cascade and generating reactive oxygen species
Activation of AT 1 R by Ang II can stimulate a variety of intracellular cascades in renal epithelial cells (48). Thus, we next aimed to determine the downstream effector of AT 1 R in mediating ClC-K2 activation in intercalated cells of the collecting duct. As demonstrated by the representative patch clamp experiment in Figure 6A, inhibition of the G q/11phospholipase C pathway with U73122 (10 μM) did not Ang II stimulates ClC-K2 in intercalated cells significantly alter single channel ClC-K2 activity and did not prevent stimulatory actions of Ang II (500 nM). As summarized in Figure 6B, the mean P o was 0.33 ± 0.03, 0.29 ± 0.03, 0.51 ± 0.04, and 0.31 ± 0.03 in the control, upon pretreatment with U73122 for 3 min, after application of Ang II in the continued presence of the blocker, and following washout with  Ang II stimulates ClC-K2 in intercalated cells control medium, respectively. Activation of AT 1 R can stimulate phosphoinositide 3 kinase (PI3-K), which is capable of acutely increasing ClC-K2 activity (46). However, pretreatment with PI3-K inhibitor LY294002 (20 μM) for 3 min did not prevent the upregulation of ClC-K2 by Ang II (Fig. 6C). Furthermore, inhibition of phospholipase A2, another potential downstream effector of AT 1 R, with AACOCF 3 (30 μM) significantly inhibited basal ClC-K2 P o from 0.36 ± 0.03 to 0.24 ± 0.03 but did not abolish Ang II-induced increases in ClC-K2 P o to 0.33 ± 0.03 (Fig. 6D). Overall, we concluded that G q/11 -phospholipase C, PI3-K, and phospholipase A2 do not play a significant role in the stimulation of ClC-K2 activity by Ang II in intercalated cells of the collecting duct.
Abundant published evidence demonstrates a marked increase in reactive oxygen species (ROS) levels in renal tubule cells treated with Ang II (49). Thus, we next quantified the action of Ang II on ClC-K2 upon pretreatment with NADPH oxidase (NOX) inhibitor apocynin (100 μM). Apocynin did not affect ClC-K2 basal activity but precluded stimulatory actions of Ang II on the channel (Fig. 7A). As summarized in Figure 7B, the mean P o was 0.32 ± 0.04, 0.30 ± 0.04, 0.29 ± 0.04, and 0.31 ± 0.04 in the control, upon pretreatment with apocynin for 3 min, after application of Ang II in the continued presence of the NOX blocker, and following washout with control medium, respectively. These results support the view that Ang II increases ClC-K2 activity in a NOX-dependent manner.
We next monitored generation of ROS in response to Ang II in freshly isolated split-opened collecting ducts using fluorescence microscopy. As shown in Figure 8A, the overall intensity of the ROS-reporting signal was markedly increased after pretreatment with Ang II (500 nM) for 15 min. We further stained the tested collecting ducts with AQP2 to quantify Ang II-induced ROS generation in principal (AQP2-positive) and intercalated (AQP2-negative, shown with white arrows) cells. As summarized in Figure 8B, principal cells exhibited larger ROS levels than intercalated cells at the baseline. Pretreatment with Ang II significantly increased intensities of the ROSreporting fluorescent signal in both cell types, with the stimulatory effect being moderately more pronounced in principal cells. Finally, we explored the time course of ROS generation in principal and intercalated cells in response to acute administration of Ang II (500 nM). As shown in Figure 8C, we detected a rapid and reversible increase in the magnitude of ROS-reporting fluorescent signal in both cell types within 5 min of Ang II application. Consistently with the results in Figure 8B, the overall effect was moderately larger in principal cells. Overall, the results in Figure 8 demonstrate that Ang II is capable of increasing ROS production in freshly isolated collecting ducts. Moreover, the time course of Ang II-induced ROS generation (Fig. 8C) closely follows the time course of ClC-K2 activation by Ang II (Fig. 3A) strongly implying that Ang II increases ClC-K2 P o in a ROS-dependent manner.

Discussion
In the current study, we explored the direct effects of Ang II on the basolateral conductance in principal and intercalated cells of the collecting duct (Fig. 9). We show that Ang II acts on AT 1 R to activate NOX and trigger subsequent ROS generation in both cell types. This pathway stimulates the basolateral Cl − conductance and ClC-K2 activity in intercalated cells but does not alter K + -selective conductance and K ir 4.1/5.1 activity in principal cells despite a moderately greater Ang IIinduced ROS generation in this cell type.
Activation of the basolateral Cl − conductance (Fig. 1) and the single channel ClC-K2 activity (Fig. 3) by Ang II provides a strong support to the idea that Ang II stimulates transcellular Cl − reabsorption by the collecting duct intercalated cells. This is in line with the previously reported upregulation of the apical Cl − /HCO 3 − exchanger, pendrin, by Ang II in intercalated cells (10). Although it is commonly believed that appreciable Cl − movement occurs in a paracellular manner in the collecting duct (30% for rabbits (50)) secondary to the electrogenic Na + reabsorption via ENaC by the principal cells, the tight junctions are only marginally more selective for Cl − versus Na + (ratio is 1.2-1.3:1) (51) indicating rather passive concomitant NaCl flux, as it similarly occurs in upstream nephron segments (52). Furthermore, blockade of ENaC with amiloride did not abolish lumen-to-bath Cl − movement in perfused collecting ducts (10). We recently showed that ClC-K2 activity in the intercalated cells is inversely related to dietary Cl − intake but not to aldosterone (45). Thus, it is plausible to propose that elevations of Ang II during volume depletion (which is also chloride depletion) increases ENaC-mediated sodium and pendrin/ClC-K2-dependent Cl − reabsorption by acting on principal and intercalated collecting duct cells, respectively (10,53). Furthermore, Ang II seems to be also critical to determine ClC-K2 expression. Indeed, we found markedly lower ClC-K2 levels in the kidney in mice Collecting ducts from at least three different mice were used.
Ang II stimulates ClC-K2 in intercalated cells lacking AT 1 R (Fig. 5). Future studies are necessary to carefully determine the role of Ang II in the regulation of ClC-K2 activity and expression during variations in dietary salt intake. It is generally believed that the basolateral K ir 4.1/5.1 channel plays a critical role in setting the resting membrane potential of the basolateral membrane in the collecting duct and upstream segments, such as distal convoluted tubule (32,54). This, in turn, determines the transepithelial voltage to control NaCl reabsorption. Ang II augments chloride reabsorption in the collecting duct by stimulating apical Cl − /HCO 3 − exchange (10) and basolateral Cl − exit via ClC-K2 (Figs. 1 and 3) in intercalated cells and increases the apical ENaC-mediated sodium entry in principal cells (14,53). In this regard, the lack of stimulatory effects of Ang II on K ir 4.1/5.1 (Figs. 1 and 2) requires a comment. It has been determined that conductance of the basolateral membrane is approximately 10 times larger than conductance of the apical membrane in principal cells (29,44) with ENaC activity being a rate-limiting step in determining the rate of Na + reabsorption (6). This means that the tandem of Na + /K + ATPase and K ir 4.1/5.1 has more than enough capacity to perform basolateral Na + exit even in the presence of increased ENaC-dependent Na + entry in response to Ang II. Thus, potential stimulation of K ir 4.1/5.1 activity by Ang II would not further augment Na + reabsorption by the principal cells. On the other side, hyperpolarization of the basolateral membrane (due to augmented K ir 4.1/5.1 activity) sets up a favorable driving force for the apical K + secretion via ROMK (K ir 1.1) channel (55). Indeed, we recently found that elevated dietary K + intake increases K ir 4.1/5.1 activity in the collecting duct to facilitate urinary K + excretion (45). In turn, apical K + secretion by the principal cells decreases the electrical driving force for Cl − reabsorption by intercalated cells.
Of interest, Ang II has been shown to inhibit ROMK activity (56), which would aid coordination of Na + and Cl − reabsorption by principal and intercalated cells, respectively, in the absence of augmented K + secretion. Overall, it is reasonable to propose that such architecture allows adaptation of the collecting duct cells to different physiological stimuli by switching from Na + /K + exchange during hyperkalemia to predominantly NaCl reabsorption during hypovolemia. Our results suggest a critical role of AT 1 receptors in mediating stimulatory signal of Ang II to ClC-K2 (Fig. 4). Pretreatment with the AT 1 R blocker, losartan, abolished increases in the ClC-K2 open probability in response to Ang II. Furthermore, we detected a 50% reduction of ClC-K expression in the kidney of AT 1 R −/− mice (Fig. 5) suggesting both regulatory and permissive roles of Ang II-AT 1 R cascade for regulation of ClC-K2-dependent chloride reabsorption in the collecting duct. At this stage, we are not able to determine the relative contribution of the apical and basolateral receptors in this regulation. The basolateral membrane was more readily available in patch clamp studies in freshly isolated collecting ducts (see Fig. 1). However, Ang II elicited a rapid increase in ROS levels in intercalated cells within split-opened area of the collecting duct having exposed the apical membrane (Fig. 8). Thus, our results favor the scenario where apically and basolaterally localized AT 1 R produce complementary stimulatory effects on ClC-K2 activity via ROS production. Moreover, both intratubular and interstitial Ang II levels in the kidney were shown to be in the nanomolar range (18), which corresponds to the established concentration range of ClC-K2 regulation by Ang II (Fig. 3C). This provides a direct support to the notion that this regulation is physiologically relevant. Although expression of AT2 and Mas receptors have been reported in the collecting duct cells with proposed antihypertensive roles by promoting natriuresis and diuresis (25)(26)(27)57), their stimulation with CGR42112 and Ang 1-7, respectively, did not affect ClC-K2 activity in the intercalated cells ( Fig. 4C and  Fig. S2). It is possible that the vasoprotective branch of the renin-angiotensin system might play a more pronounced role during volume-expanded hypertensive states and diabetic nephropathy (57). Furthermore, animal sex should be also taken into account, since AT2 receptor expression is higher in females (58). Future studies should carefully determine the contribution of AT 1 receptor-dependent and -independent mechanisms in the regulation of ClC-K2mediated Cl − reabsorption in the collecting ducts during normotensive and hypertensive states.
We found Ang II-AT 1 R increases ClC-K2 activity in the intercalated cells by stimulating NOX and following ROS production (Figs. 7 and 8). Of importance, our results demonstrate a striking similarity between the time courses of upregulation of the ClC-K2 open probability (Fig. 3) and ROS generation (Fig. 8) in response to Ang II. Although it is common that ROS can induce covalent modification of specific cysteine residues to alter the functional status of the target proteins, this effect is poorly reversible and does not fit with the observed gradual restoration of the basal ClC-K2 activity upon Ang II washout (Figs. 3 and 6). Of interest, it was shown that an increase in superoxide levels significantly increased intracellular pH, whereas increases in peroxide levels led to intracellular acidification (59,60). Of note, ClC-K2 exhibits a remarkably steep pH dependence, with acidic (<7.0) and alkaline (>7.5) media leading to acute and reversible decreases and increases in channel activity, respectively (46,61). A potential mechanism likely involves protonation/deprotonation of a histidine residue (H497) of the channel (62). Thus, it is plausible to propose that the Ang II-AT 1 R-NOX pathway promotes generation of superoxide to increase cytosolic pH to activate ClC-K2 in intercalated cells. At the same time, K ir 4.1/5.1 is also known to be sensitive to pH (31,63). It is possible that Ang II leads to generation of both superoxide and peroxide in principal cells but Ang II stimulates ClC-K2 in intercalated cells mostly superoxide in intercalated cells. This would explain the higher ROS (superoxide and peroxide together) levels shown in Figure 8 but potentially little pH changes and thus lack of K ir 4.1/5.1 regulation by Ang II in principal cells. We will consider to investigate this intriguing aspect in the future.
Although ROS-dependent redox-signaling processes contribute significantly to the normal cellular responses to Ang II, excessive ROS accumulation drives proinflammatory and profibrotic actions of Ang II contributing to endothelial dysfunction, fibrosis, and the development of hypertension (64). Indeed, we previously showed that Ang II increases ENaC activity by increasing ROS in the collecting duct principal cells (14) and this is an important mechanism for stimulation of ENaC-dependent Na + reabsorption in response to dietary sodium deficiency (53). However, chronic Ang II infusion stimulates ENaC activity far beyond the physiological range independently of aldosterone thus contributing to excessive volume retention and hypertension (13). Cl − is the principal extracellular anion accounting for over 70% of the total negative ion content. Jointly with Na + , Cl − is the major contributor to extracellular volume and osmolarity. Although abnormal regulation of Na + balance is considered to be central for the development of elevated blood pressure, accumulated evidence argues that the Cl − component might play an even more important role in the pathology of salt-sensitive hypertension (65). For instance, Dahl salt-sensitive or stroke-prone spontaneously hypertensive rats develop elevated blood pressure when fed with high-NaCl diet but not high Na + bicarbonate or other Cl − substitutes (66)(67)(68). This article demonstrates that physiologically relevant Ang II levels stimulate ClC-K2 activity and by extension transcellular Cl − reabsorption in the collecting duct, the site directly involved in the regulation of urinary electrolyte excretion to match dietary intake and setting salt sensitivity of blood pressure. Overall, we propose that upregulation of ClC-K2 by Ang II is critical for protection of the circulatory volume during hypovolemic states, whereas overactivation of ClC-K2 might contribute to the pathophysiology of Ang II-dependent hypertension.

Reagents and animals
All chemicals and materials were from Sigma, VWR, and Tocris unless noted otherwise and were of reagent grade. For experiments, C57BL/6J mice (Charles River Laboratories) and B6.129P2-Agtr1 tm1Unc /J (AT 1a R −/−, JAX strain #002682, the dominant AT 1 receptor isoform in the kidney (69)) 6 to 10 weeks old were used. In order to minimize sex-related variations in the measured experimental parameters, only males were used for experiments. Animal use and welfare adhered to the NIH Guide for the Care and Use of Laboratory Animals following protocols reviewed and approved by the Animal Care and Use Committee of the University of Texas Health Science Center at Houston.

Tissue isolation
The procedure for isolation of the collecting ducts suitable for electrophysiology followed previously published protocols (45,70,71). Briefly, mice were sacrificed by CO 2 administration followed by cervical dislocation, and the kidneys were removed immediately. Kidneys were cut into thin slices (<1 mm) with slices placed into ice-cold Ringer solution containing (in mM) 150 NaCl, 5 KCl, 1 CaCl 2 , 2 MgCl 2 , 5 glucose, and 10 Hepes (pH 7.35). Straight cortical-tomedullary sectors, containing approximately 30 to 50 renal tubules, were isolated by microdissection using watchmaker forceps under a stereomicroscope. To dissolve the basal lamina and to get direct access to the basolateral membrane, isolated sectors were further incubated in the Ringer solution containing 0.8 mg/ml collagenase type I (Alfa Aesar) and 5 mg/ml of dispase II (Roche Diagnostics) for 20 min at 37 C followed by extensive washout. Individual collecting ducts were visually identified by their morphological features (pale color; coarse surface) and were mechanically isolated from the sectors by microdissection. The collecting ducts were further verified by positive expression of AQP2 water channel with immunofluorescent microscopy, as detailed below. Isolated collecting ducts were attached to a 5 x 5 mm cover glass coated with poly-L-lysine. A cover glass containing a collecting duct was placed in a chamber mounted on an inverted Nikon Eclipse Ti microscope and perfused with the Ringer solution at room temperature. The samples were used within 1 to 2 h after isolation. For each experimental condition, collecting ducts from at least three different mice were analyzed.

Whole cell currents and membrane potential in isolated collecting ducts
Whole cell currents in collecting duct cells were measured under voltage-clamp conditions in the perforated-patch mode with gigaohm seals formed on the basolateral membrane, as described (45,70). Patch clamp recordings were acquired with an Axopatch 200B (Molecular Devices) patch clamp amplifier interfaced via a Digidata 1440 (Molecular Devices) to a computer running the pClamp 10.7 (Molecular Devices). The bath solution was (in mM): 150 NaCl, 5 KCl, 1 CaCl 2 , 2 MgCl 2 , 5 glucose, and 10 Hepes (pH 7.35). Freshly made amphotericin-B, 400 μM (Enzo Life Sciences) was dissolved in the pipette solution containing (in mM):150 KAcetate, 5 KCl, 2 MgCl 2 , and 10 Hepes (pH 7.35) by ultrasonication. Recording pipettes had resistances of 3 to 5 MΩ. Electrical recordings were made once the access resistance from the pipette to the cell interior reduced to less than 15 MΩ, usually 5 to 10 min after achieving a pipette-to-membrane seal resistance of 5 to 10 GΩ. The capacity of individual cells (15 pF) was manually compensated. Principal and intercalated cells were further distinguished by their electrical properties with principal cells having K + -selective cation conductance and a highly negative resting membrane potential around −70 mV, whereas intercalated cells have Cl − -selective anion conductance and resting membrane potential around −20 mV, as we demonstrated previously (29).

Single channel recordings in isolated collecting ducts
The activity of K ir 4.1/5.1 and ClC-K2 channels in freshly isolated collecting ducts was determined in cell-attached patches on the basolateral membrane of principal and intercalated cells, respectively, under voltage-clamp conditions, as previously described (45,70). Recording pipettes had resistances of 8 to 10 MΩ. Bath and pipette solutions were (in mM): 150 NaCl, 5 KCl, 1 CaCl 2 , 2 MgCl 2 , 5 glucose, and 10 Hepes (pH 7.35); and 150 KCl, 2 MgCl 2 , and 10 Hepes (pH 7.35). In the cell attached configuration, the actual voltage applied to a membrane patch (V patch ) is a sum of the pipette voltage and the resting basolateral membrane potential of principal (V basolateral , which is close to −70 mV for principal and −20 mV for intercalated cells, see Fig. 1, B and D). Currents were low-pass filtered at 1 kHz with an eight-pole Bessel filter (Warner Instruments). Events were inspected visually prior to acceptance. Channel activity (NP o ) and open probability (P o ) were assessed using Clampfit 10.7 (Molecular Devices). Channel activity in individual patches, defined as NP o , was calculated using the following equation: NP o = (t 1 + 2t 2 + …+nt n ), where N is the number of active channels (K ir 4.1/5.1 or ClC-K2) in a patch and t n is the fractional open time spent at each of the observed current levels. P o was calculated by dividing NP o by the maximal number of simultaneously active channels within a patch (N) as defined by all-point amplitude histograms. For representation, current traces were filtered at 200 Hz and corrected for a slow baseline drift as necessary.

Western blotting
Immediately after dissection kidneys were placed on ice, decapsulated, and homogenized in three volumes of ice-cold lysis buffer containing 50 mM TrisCl, 5 mM EDTA and 1% Triton X-100 (pH 7.5) supplemented with Complete Mini protease and PhosSTOP phosphatase inhibitor cocktails (Roche Diagnostics). The homogenates were centrifuged at 1000g for 15 min at +4 C, and the sediment was discarded. Protein concentration was determined with a Bradford assay using bovine serum albumin as a standard. The samples (40 μg/lane) were separated on 9% polyacrylamide gels at 150 V for 90 min and transferred to a nitrocellulose membrane for 70 min at 100 V. Equal protein load was verified by Ponceau red staining using standard procedures. Nitrocellulose membranes were incubated with primary anti-ClC-K antibodies (rabbit polyclonal, 1:1000 Alomone Labs, Cat. # ACL-004) overnight at +4 C. Upon washout (three times for 10 min in TBS-Tween), the membrane was incubated with peroxidase-conjugated goat anti-rabbit (1:10,000, Jackson ImmunoResearch Laboratories) secondary antibodies for 1 h at room temperature. Blots were quantified using ImageJ 1.50e software (NIH). The intensities of the studied protein bands were normalized to the total signal of the respective line in Ponceau red staining.

Total ROS detection
Freshly isolated split-opened collecting ducts were loaded with the oxidative stress detection reagent (Enzo Life Sciences, ENZ-51011) in 1X Wash Buffer for 45 min at room temperature according to the manufacturer's protocol. Fluorescent images were recorded with the same exposure time (4 ms) with a Nikon Ti-S Wide-Field Fluorescence Imaging System (Nikon Instruments) integrated with Lambda XL light source (Sutter Instrument) and QIClick 1.4 megapixel monochrome CCD camera (QImaging) via NIS Elements 4.3 Imaging Software (Nikon Instruments). The collecting ducts were imaged with Nikon fluorescence microscope with excitation at 490 nm and emission at 525 nm using a 40X Nikon Super Fluor objective, and regions of interest were drawn for individual cells. The efficiency of the total ROS detection kit was tested on subconfluent mpkCCD c14 cells, a generally accepted model of the collecting duct principal cells, as we similarly did previously (14). As shown on the representative micrographs in Fig. S1A and summarized in Fig. S1B, addition of ROS inducer pyocyanin (200 μM for the last 20 min of incubation) drastically increased the intensity of ROS-reporting fluorescent signal, which was largely precluded by concomitant incubation with the negative control reagent (N-acetyl-L-cysteine, 10 mM) for 30 min. For the experiments, Ang II (500 nM) or vehicle were added to the freshly isolated split-opened collecting ducts for the last 10 min of the incubation with the oxidative stress detection reagent. In another set of experiments, the time course of changes in ROS in response to perfusion of Ang II (500 nM) was assessed by sampling the fluorescent intensities at 525 nm every 15 s.

Immunofluorescent microscopy
Immediately after ROS measurements, split-opened collecting ducts were fixed with 10% neutral buffer formalin for 15 min at room temperature. After fixation, the samples were permeabilized by addition of 1% SDS in PBS for 10 min and washed in PBS for 5 min. Nonspecific staining was blocked with 1% BSA in PBS for 1 h at room temperature. The samples were incubated overnight at +4 C with anti-Aquaporin 2 antibody (1:4000 dilution; Alomone Labs, Cat. # AQP-002). After washing with PBS, the samples were incubated with goat anti-rabbit IgG labeled with Alexa Fluor 594 (1:1000 dilution; Invitrogen) for 1 h at room temperature in the dark. For experiments with double staining, the samples were incubated with anti-Aquaporin 2 antibody (1:4000 dilution; Alomone Labs, Cat # AQP-002) overnight at 4 C. After washing with PBS, the samples were incubated with goat antirabbit IgG labeled with Alexa Fluor 488 (1:2000 dilution; Invitrogen) for 1 h at room temperature in the dark. Subsequently, nonspecific staining was blocked with 10% rabbit serum for 30 min and samples were incubated with anti-CLC-K antibody (1:500 dilution; Alomone, Cat # ACL-004) conjugated with goat anti-rabbit IgG labeled with Alexa Fluor 594 for 2 h at 37 C in the dark. After washing with PBS (three times for 5 min) the samples were stained with 4',6-diamidino-2-phenylindole (500 nM concentration, Calbiochem) to visualize nuclei. The samples were dehydrated and mounted with Fluoromount-G (SouthernBiotech, Cat# 0100-01). Labeled tubules were examined with an inverted Nikon Eclipse Ti Ang II stimulates ClC-K2 in intercalated cells fluorescent microscope using a 40X Plan-Fluor (1.3 NA) objective. Samples were excited with 405-and 561-nm laser diodes, and emission was captured with a 16-bit Cool SNAP HQ 2 camera (Photometrics) interfaced to a PC running NIS elements software.

Data analysis
All summarized data are reported as mean ± SEM and ±SD for nonpaired experiments and mean ± SEM for paired patch clamp studies, as indicated in respective figure legends. Statistical comparisons were made using one-way ANOVA with post hoc Tukey test or one-way repeated measures ANOVA with post hoc Bonferroni test (for paired experiments within the same group). p Value less than 0.05 was considered significant.

Data availability
All data from this study are contained within the article including Supplemental Information. Conflict of interest-The authors declare that they have no conflicts of interest with the contents of this article.