Caveolin-1 Deficiency Inhibits the Basolateral K+ Channels in the Distal Convoluted Tubule and Impairs Renal K+ and Mg2+ Transport.

Kcnj10 encodes the inwardly rectifying K(+) channel Kir4.1 in the basolateral membrane of the distal convoluted tubule (DCT) and is activated by c-Src. However, the regulation and function of this K(+) channel are incompletely characterized. Here, patch-clamp experiments in Kcnj10-transfected HEK293 cells demonstrated that c-Src-induced stimulation of Kcnj10 requires coexpression of caveolin-1 (cav-1), and immunostaining showed expression of cav-1 in the basolateral membrane of parvalbumin-positive DCT. Patch-clamp experiments detected a 40-pS inwardly rectifying K(+) channel, a heterotetramer of Kir4.1/Kir5.1, in the basolateral membrane of the early DCT (DCT1) in both wild-type (WT) and cav-1-knockout (KO) mice. However, the activity of this basolateral 40-pS K(+) channel was lower in KO mice than in WT mice. Moreover, the K(+) reversal potential (an indication of membrane potential) was less negative in the DCT1 of KO mice than in the DCT1 of WT mice. Western blot analysis demonstrated that cav-1 deficiency decreased the expression of the Na(+)/Cl(-) cotransporter and Ste20-proline-alanine-rich kinase (SPAK) but increased the expression of epithelial Na(+) channel-α. Furthermore, the urinary excretion of Mg(2+) and K(+) was significantly higher in KO mice than in WT mice, and KO mice developed hypomagnesemia, hypocalcemia, and hypokalemia. We conclude that disruption of cav-1 decreases basolateral K(+) channel activity and depolarizes the cell membrane potential in the DCT1 at least in part by suppressing the stimulatory effect of c-Src on Kcnj10. Furthermore, the decrease in Kcnj10 and Na(+)/Cl(-) cotransporter expression induced by cav-1 deficiency may underlie the compromised renal transport of Mg(2+), Ca(2+), and K(+).


RESULTS
To examine whether cav-1 is expressed in the basolateral membrane of the DCT, we carried out immunostaining in the wildtype (WT) and cav-1 KO mice. Figure 1A is a florescence microscope image at a low magnification showing the expression of cav-1 in the kidney of WT mice. The positive cav-1 staining was observed in both renal cortex and medullary region, and this staining was specific because no cav-1 immunostaining was observed in the KO mice (Figure 1B). We confirmed the previous report that cav-1 was expressed in the glomerulus, DCT, cortical collecting duct (CCD), and renal vessels. 19,20 Microscopic examination with a high magnification revealed that cav-1 was expressed in the glomerulus ( Figure 1C), DCT ( Figure 1C), and CCD ( Figure 1D), whereas the cav-1 staining was not visible in the KO mice ( Figure 1E). Moreover, immunostaining image demonstrates that cav-1 is highly expressed in the basolateral   (Figure 1, F-H), which is identified by positive immunostaining of parvalbumin, a Ca 2+ -binding protein exclusively expressed in the DCT. 21 Because previous study showed that Kir4.1 was expressed in the basolateral membrane of DCT, 13 it is reasonable to assume that Kir4.1 and cav-1 are colocalized in the basolateral membrane of DCT cells. Our previous study demonstrated that c-Src phosphorylated Kcnj10 and that the inhibition of src-family protein tyrosine kinase (SFK), a family of c-Src-like tyrosine kinase, 22 decreased the basolateral 40 pS K + channel activity in DCT1. 9 Therefore, we next examined whether cav-1 was required for the stimulatory effect of c-Src on Kir.4.1 in Kcnj10-transfected HEK293T cells with the perforated whole-cell recording. Figure 2A shows a set of typical recording of Ba 2+ -sensitive K + currents measured in the cells transfected with Kcnj10, Kcnj10+c-Src, Kcnj10+cav-1, and Kcnj10+cav-1+c-Src. Figure 2B is a Western blot demonstrating that the total expression of Kcnj10 protein was similar among four groups. From the inspection of Figure 2A, it is apparent that c-Src alone did not significantly increase K + currents without cav-1 cotransfection (224622 pA/pF at -60 mV in Kcnj10-transfected cells versus 230621 pA/pF in Kcnj10+c-src-transfected cells) (n=6). In contrast, in cav-1 cotransfected HEK cells, c-Src significantly increased the K + currents to 360626 pA/pF (n=6), whereas the expression of cav-1 had no significant effect on K + currents (210623 pA/pF) ( Figure 3A). Therefore, c-Src-induced stimulation of Kcnj10 requires cav-1 participation. Because Kcnj10 protein expression was similar with or without cav-1 ( Figure 2B), the synergistic effect of cav-1 on c-Src-induced stimulation of Kcnj10 channel activity was most likely caused by posttranslational modulation.
Inhibition of SFK has been shown to decrease the basolateral K + channel activity in mouse DCT1. 9 We next examined whether cav-1 was also required for the effect of PP1, an inhibitor of SFK, on Kcnj10. We performed the perforated whole-cell recording in HEK293T cells transfected with Kcnj10 or Kcnj10+cav-1. From the inspection of Figure 3B, it is apparent that the inhibition of SFK with PP1 had no effect on the Ba 2+ -sensitive K + current in Kcnj10-transfected cells (210615 pA/pF, n=5). In contrast, PP1 decreased the K currents to 110612 pA/pF (n=6) in cav-1 cotransfected cell (Figure 3, A and B). We also repeated the experiments in which the effect of SFK inhibition on K + currents was examined in HEK293T cells transfected with Kcnj10/16 alone or cav-1 +Kcnj10/16. Results are summarized in Figure 3C showing that PP1 inhibited Kcnj10/16 only in the cav-1 cotransfected cells but not in those without cav-1 transfection (Kcnj10/16: 280612 pA/pF; Kcnj10/16+PP1: 260610 pA/pF; Kcnj10/16+ cav-1: 275612 pA/pF; Kcnj10+cav-1+PP1: 150615 pA/pF; n=4-6). Therefore, results suggest that cav-1 is critically involved in the modulation of SFK effect on Kcnj10. In addition to c-Src, with-no-lysine kinase 4 (WNK4) is expressed in the DCT and plays a role in the regulation of NCC. 23,24 To determine whether the synergistic effect of cav-1 on c-Src-mediated stimulation of Kcnj10 was specific, we examined the effect of WNK4 on Kcnj10 in the cav-1 cotransfected cells with the perforated whole-cell recording ( Figure 3D). The expression of WNK4 had no effect on Kir4.1 in Kcnj10-transfected cells or in the cells transfected with cav-1+Kcnj10 (control: 234622 pA/pF; cav-1: 240623 pA/pF; WNK4: 230615 pA/pF; WNK4+cav-1: 230620 pA/pF; n=6). Therefore, the expression of cav-1 specifically modulates the effect of c-Src on Kir.4.1.
Previous study showed that SFK activated Kcnj10 by stimulating tyrosine phosphorylation at Tyr. 9 To examine whether the mutation of Tyr 9 could affect cav-1induced synergistic effect on SFK-induced stimulation of Kcnj10, we performed the experiments in the cells transfected with Kcnj10 Y9F in which the tyrosine phosphorylation site was mutated. Figure 4 is of typical recordings showing that the K + currents in KCNJ10 Y9F -transfected cells were significantly smaller (75610 pA/pF, n=5) than those in the control cells (transfected with KCNJ10). Moreover, the expression of c-Src modestly increased the K + currents from 76610 pA/pF to 105610 pA/pF (n=6) in cav-1 cotransfected cells. Therefore, cav-1 modulates the SFK effect on Kcnj10, at least in part, by facilitating the phosphorylation of Kcnj10 at Tyr. 9 After demonstrating that cav-1 was required for the effect of SFK on Kcnj10 in a cell model, we further explored the role of cav-1 in the regulation of Kcnj10 in cav-1 KO mice. If cav-1 is required for facilitating the stimulatory effect of SFK on the basolateral 40 pS K + channel, the disruption of cav-1 should mimic the effect of SFK inhibition and decrease the 40 pS K + channel activity in the basolateral membrane of the DCT. Figure  5A is a typical single-channel recording made in a cell-attached patch of DCT1 of cav-1 KO mice. As in WT mice, we identified a 40 pS inwardly rectifying K + channel in the DCT, and this 40 pS K + channel was the only type of K + channel detected in the basolateral membrane of DCT1 from total 66 patches in cav-1 KO mice. The absence of Kcnj10 homotetramer (20 pS K + channel) in the DCT suggests that Kcnj10 prefers to interact with Kcnj16 (Kir.5.1) in vivo. This notion was also supported by the report that the 20 pS K + channel was abundant in the DCT of Kcnj16 2/2 mice. 25 Moreover, the mean channel open probability in cav-1 KO mice (0.4260.02, n=10) was not significantly different from those of WT animals (0.4660.02, n=10) ( Figure 5B). However, the probability of finding this 40 pS K + channel was significantly lower in cav-1 KO mice (25 patches with the K channel from total 66 patches) than in WT mice (18 patches with K channel from total 35 patches) ( Figure 5C). Therefore, the mean K + channel number per patch was significant less in DCT1 of cav-1 KO mice (1.1760.23) than in WT mice (2.560.45) ( Figure 5D). To examine whether the disruption of cav-1 reduced the expression of Kcnj10 in the kidney, we performed the immunostaining in WT and KO mice. Figure 5E shows immunostaining images with a low magnification in the kidney of WTand KO mice. From the inspection of Figure 5E, it is apparent that the intensity of Kcnj10 staining is lower in KO mice than WTmice. This is further confirmed by the microscopic examination, with a large magnification showing that the immunostaining intensity of the basolateral Kcnj10 in KO mice is weaker than in WTmice ( Figure 5F). This suggests that the lack of cav-1 decreased the surface expression of the functional 40 pS K + channel in the DCT1.
The notion that the lack of cav-1 decreased the basolateral K channel activity in the DCT was further supported by experiments in which whole-cell K + currents were measured with the perforated patch. As discussed previously, the DCT1 cells have no apical K + channel, and they have a lack of coupling with neighboring cells. 9,13 Therefore, the whole-cell K currents were equal to the basolateral K + conductance. Figure 6A is a set of whole-cell recording showing that K + currents of DCT1 in cav-1 KO mice (513672 pA) were significantly lower than those in WT mice (1,3006205 pA) (n=8). The inhibition of SFK decreased the K + current of DCT1 in WTmice to 470680 pA (n=7) ( Figure 6B), whereas it had no significant effect on the basolateral K + conductance of DCT in cav-1 KO mice (440680 pA) (n=8). This suggests that the disruption of cav-1 significantly inhibits the basolateral K + conductance in DCT1 and that the decrease in K + conductance is most likely caused by the defective regulation of the 40 pS K + channel in the basolateral membrane by SFK. Because Kcnj10 plays a dominant role in determining the cell membrane potential in DCT1, 13 a decrease in the basolateral K + conductance is expected to depolarize the cell membrane potential. This notion was supported by measuring K + reversal potential using the perforated whole-cell recording with 5 mM K + in the bath and 140 mM K + in the pipette. The results summarized in Figure 6C from five measurements show that the K + reversal potential was 6563 mV in DCT1 of WT mice, and it depolarized to 4962 mV in the presence of SFK inhibitor. In contrast, the K + reversal potential was significantly lower in DCT1 of cav-1 KO mice (5063 mV) than in WT mice, and it was also not significantly different from the value in the presence of SFK inhibitor (4763 mV).
The basolateral K + conductance plays a key role in providing the driving force for Clexit through the basolateral Clchannel. We previously demonstrated that the disruption of Kcnj10 also inhibited the basolateral Clconductance in the DCT of Kcnj10 2/2 mice. 13 Therefore, we used the perforated wholecell recording to examine the Clconductance in the DCT of both WT and cav-1 KO mice. Figure 7A is a recording showing NPPB-sensitive Clcurrents in DCT1 of WT and cav-1 KO mice, and results from six experiments are summarized in Figure 7B. It is apparent that Clcurrents in the DCT1 of KO mice (-4906120 pA at -60 mV) were significantly lower than those of WT mice (-920690 pA at -60 mV). Therefore, the disruption of cav-1 also inhibited the basolateral Clconductance in the DCT1, which should increase the intracellular Clconcentration, thereby inhibiting WNK autophosphorylation and WNK-SPAK interaction. 26,27 We previously demonstrated that the disruption of Kcnj10 decreased the expression of NCC and SPAK in Kcnj10 KO mice. 13 Therefore, we examined the expression of SPAK in WT and cav-1 KO mice. Figure 7C is a Western blot showing that the lack of cav-1 inhibited the expression of full-length SPAK by 52%610% (n=4). Because SPAK has been shown to stimulate NCC activity through phosphorylation, [28][29][30] it is conceivable that the disruption of cav-1-induced inhibition of SPAK should reduce the NCC phosphorylation. Figure 7D is a Western blot showing the expression of Thr 53 -phosphorylated NCC, which is an index of activated NCC, 31 and the total NCC in the kidney from the WT and cav-1 KO mice. In comparison with the control value, the disruption of cav-1 decreased the expression of phosphorylated NCC and total NCC by 80%610% and 45%610% (n=5), respectively.
We also examined the expression of NCC with a florescence microscope. Figure 8A is an image with a low magnification showing that the expression of NCC is lower in the KO mice than in WTmice. This is further confirmed by the microscopic examination with a large magnification showing a sharp staining of NCC in the WT mice in comparison with the KO mice ( Figure 8B). To examine whether the decrease in NCC expression was the result of suppressing transcription, we carried out real-time quantitative PCR to examine the mRNA level of NCC and SPAK, and Figure 8C shows an agarose gel of the PCR product of NCC and SPAK. The results from four experiments are summarized in Figure 8D, demonstrating that the disruption of cav-1 did not alter the mRNA of both fulllength SPAK and kidney-specific SPAK, whereas it increased mRNA of NCC by 40%610% (n=4). Therefore, the decrease in NCC and SPAK protein expression in cav-1 KO mice was caused by a post-translational regulation. The increase in NCC mRNA in cav-1 KO mice may be caused by a compensation mechanism. The effect of the cav-1 disruption on NCC was specific because the expression of TRPV5, which is expressed in the apical membrane of the DCT, was the same between WT and KO mice (n=4) ( Figure 8E). 32 Because cav-1 is also expressed in the basolateral membrane of the CNTand CCD, we next examined whether the disruption of cav-1 affected basolateral K channel and apical epithelial sodium channel (ENaC) in the CNT. We conducted the perforated whole-cell recording to measure the K + reversal potential of the CNT, which is an indication of the basolateral K + conductance and the membrane potential. The results summarized in Figure 9A show that the K + reversal potential in the CNTof cav-1 KO mice (6162 mV, n=4) was slightly but significantly (P,0.03) lower than those of WT mice (6963 mV, n=4). However, the depolarization amplitude in the CNT was modest in comparison with those in DCT1, suggesting that Kcnj10 in the CNT might not play a dominant role in determining the basolateral membrane potential. We also examined the expression of ENaC in WT and cav-1 KO mice with Western blot and immunostaining. Figure 9B is a representative Western blot demonstrating that the expression of ENaC-a in the KO mice increased by 110%620% (P,0.01, n=4) in comparison with WT mice, whereas the expression of ENaC-b increased by 25%610% (P.0.05, n=5). This is further confirmed by fluorescence microscopic examination showing ENaC-a staining ( Figure 9C) and ENaC-b ( Figure 9D). Therefore, the disruption of cav-1 in the CCD decreased the basolateral K conductance and increased ENaC expression.
A decrease in the cell membrane potential is expected to diminish the driving force for transcellular Ca 2+ and Mg 2+ transport, whereas a low NCC expression in the DCT and a high ENaC expression in the CNT/CCD would affect renal Na and K transport in aldosterone-sensitive distal nephron. This possibility was examined by carrying out the metabolic cage study using age-and sex-matched cav-1 KO and WT mice. Table 1 summarizes the results, showing that the mouse body weight and dietary intake of Na + , K + , Mg 2+ , and Ca 2+ were similar in both groups. However, the urinary K + and Mg 2+ excretion were significantly higher in KO mice than in WT mice. Although urinary Na + and Ca 2+ excretion also have a trade to be higher in KO mice than in WT mice, the difference was not significant. Therefore, the disruption of cav-1 impairs the Mg 2+ absorption and increases K + excretion.
We also measured the plasma concentrations of Na + , K + , Mg 2+ , and Ca 2+ in WT and the KO mice. From the inspection of Figure 10

DISCUSSION
In this study, we demonstrated that the disruption of cav-1 inhibited the basolateral 40 pS K channel in DCT1 and caused membrane depolarization. Therefore, cav-1 KO mice could be used as an alternative animal model to study the effects of loss-of-function mutations of Kir.4.1 on the electrolyte transport in the DCT. The mechanism by which the lack of cav-1 decreases the number of the 40 pS K + channel in the basolateral membrane is not clear. cav-1 plays an important role in mediating a variety of cellular functions, such as mediating the signal transduction and ion channel trafficking. 18,33 For instance, cav-1 has been shown to regulate the delivery of glutamate transporter in C6 glioma cells 34 and to inhibit Kir.1.1 (renal outer medullary K + ). 35 However, the observation that c-Src-induced stimulation of Kir4.1 requires the involvement of cav-1 strongly suggests that a decrease in the basolateral 40 pS K + channel activity in the DCT of cav-1 KO mice was, at least in part, the result of the defective regulation of Kir.4.1 by SFK.
Our previous study demonstrated that Kir4.1 was the substrate of SFK and that tyrosine residue 9 was phosphorylated by c-Src. 9 The role of SFK in the regulation of the basolateral 40 pS K + channel was demonstrated by the observation that the inhibition of SFK decreased the basolateral 40 pS K + channel activity and depolarized the basolateral membrane potential in DCT1. However, because inhibiting SFK decreased Kir4.1 activity only in the cav-1 cotransfected cells, but not in Kcnj10 alone transfected HEK293T cells, this strongly suggests that cav-1 is required for the effect of SFK on Kcnj10. In this regard, cav-1 is highly expressed in the basolateral membrane of DCT and has been shown to be associated with Kir4.1 and Ca 2+ -sensing receptor (CaSR). 12,19,36 Also, it has been shown that cav-1 interacts directly with c-Src and the interaction domain is within residues 82-101. 37 Therefore, it is possible that cav-1 provides a microenvironment for the interaction between SFK and Kir.4.1. Relevant to this speculation is the report that cav-1 serves as a scaffolding protein to sequester Ras protein and heme oxygenase-1 in caveolae. 38,39 Because SFK plays an important role in stimulating the 40 pS K + channels in the basolateral membrane of the DCT, the disruption of cav-1 impairs the stimulatory effect of SFK on the basolateral K + channels, thereby decreasing the basolateral K + conductance. 9 Therefore, the defective regulation of Kir4.1 by SFK is one of the possible mechanisms by which the disruption of cav-1 inhibits the basolateral K + conductance in the DCT.
As in Kcnj10 KO mice, the disruption of cav-1 also leads to a decrease in NCC expression. 13 Because cav-1 was not detected in the apical membrane of DCT, it is unlikely that cav-1 may directly be involved in regulating NCC membrane trafficking. 19 Therefore, the disruption of cav-1-induced decrease in NCC expression may be the results of the depolarization of the DCT cell membrane potential. We have previously demonstrated the expression of NCC and SPAK were decreased in homozygous Kcnj10 2/2 and heterozygous Kcnj10 +/2 mice; however, heterozygous mice have a normal growth, as do WT mice. 13 Moreover, the decrease in NCC and SPAK expression was closely correlated with the basolateral membrane potential. Because the disruption of cav-1 also depolarized the membrane potential of DCT1, this should be, at least in part, responsible for decreasing NCC and SPAK expression in cav-1 KO mice. The depolarization in the basolateral membrane potential is expected to decrease the Clexit by the basolateral Clchannels, thereby increasing the intracellular Cl concentration. In addition, the disruption of cav-1 may also directly inhibit the basolateral Clchannel activity, thereby synergizing the effect of Kcnj10 inhibition on the intracellular Cllevel. An increase in the intracellular Clconcentration leads to inhibit the WNK-SPAK interaction, thereby decreasing SPAK expression and the SPAK-induced stimulation of NCC phosphorylation, which is essential for the activity of NCC. [28][29][30]40 Therefore, it is possible that a decrease in SPAK expression is expected to inhibit NCC phosphorylation, thereby decreasing NCC expression. It has been reported that the SPAK-induced phosphorylation of NCC inhibited ubiquitylation of the thiazide-sensitive NCC in the DCT. 41 Although the disruption of cav-1 reduced the expression of NCC, the net renal Na + excretion was not significantly larger in cav-1 KO mice than in WT mice. It is possible that a decrease in NCC expression in the DCT should increase Na + delivery to the CNT/CCD, thereby stimulating Na + absorption through ENaC in these nephron segments, thereby offsetting the effect of low NCC expression. This speculation was also supported by the observation that the expression of ENaC-a was upregulated in cav-1 KO mice. As a consequence of enhancing Na + absorption in the CNT/CCD, K secretion in these nephron segments increased. This speculation is supported by the observation that the renal K + excretion was significantly enhanced in cav-1 KO mice and that they developed hypokalemia.
In addition to hypokalemia, the disruption of cav-1 also induced hypocalcemia and hypomagnesemia. It has been previously reported that the disruption of cav-1 impaired renal calcium reabsorption and led to hypercalciuria and urolithiasis. 42 The mechanism by which the disruption of cav-1 causes hypomagnesemia and hypocalcemia is not completely understood. However, a decrease in the basolateral K + conductance in the DCT should partially contribute to increasing renal excretion of Mg 2+ and Ca 2+ in the DCT because both transcellular Mg 2+ and Ca 2+ transport processes are electrogenic processes. Therefore, a decrease in the basolateral K conductance in the DCT should diminish the driving force for the entry of Mg 2+ and Ca 2+ across the apical membrane, thereby inhibiting their absorption in the DCT. Furthermore, the disruption of cav-1 may affect the function of Ca 2+ -sensing receptor (CaSR), thereby affecting electrolyte transport in the kidney. This possibility is further enhanced by the observation that Kcnj10 interacted with CaSR and cav-1 and was inhibited by CaSR activation. 12 However, further study is required to address the role of CaSR in decreasing the basolateral K + conductance of the DCT in cav-1 KO mice. Figure 11 is a cell scheme illustrating the mechanism by which the disruption of cav-1 impairs the transport in the DCT. The lack of cav-1 in the basolateral membrane impairs the stimulatory effect of SFK on Kir.4.1, thereby decreasing the basolateral K + conductance and depolarizing the cell membrane potential. A depolarization of the cell membrane potential decreases the driving force for Mg 2+ and Ca 2+ entry and increases their excretion. Also, a decrease in K conductance reduces the Clexit across the basolateral membrane and increases the intracellular Clconcentration, which inhibits SPAK-induced stimulation of NCC and increases the Na + delivery to the CNT/CCD. Therefore, the disruption of cav-1 could mimic the loss-of-function mutations of Kcnj10induced tubulopathy. We conclude that cav-1 plays a role in the stimulation of the basolateral K + channels by SFK in the DCT and that the basolateral K + channel critically participates in regulating the expression of NCC and SPAK in the DCT. The disruption of cav-1induced inhibition of the basolateral K + channels and NCC in the DCT may be partially responsible for renal Mg 2+ and K + waste.

Cell Culture and Transient Transfection
We used HEK293T cells (American Type Culture Collection) for the transient expression of flagtagged or GFP-tagged Kcnj10. The cells were grown in Dulbecco's modified Eagle medium (DMEM; Invitrogen), supplemented with 10% FBS (Invitrogen) in 5% CO 2 and 95% air at 37°C. Cells were grown to 50%-70% confluence for transfection, and the corresponding cDNA was simultaneously applied to the cells using TurboFect transfection reagent (Fermentas). Briefly, a cDNA cocktail (0.5 mg Kcnj10) was diluted with 200 ml serum-free DMEM and further mixed with 4 ml Turbofect transfection reagent for the transfection of cells cultured in a 35-mm Petri dish. Cells transfected with the vector alone were used as a control, and their background currents were subtracted from that of the experimental groups. After 15 minutes of incubation at room temperature, the mixture of the transfection agents was applied to the cells followed by an additional 24 hours of incubation before use.

Preparation of DCT1
C57BL/6 mice and cav-1 KO mice with C57BL/6 background (either sex, 4 weeks old) were purchased from The Jackson Laboratory (Bar Harbor, ME). The mice were fed with a control diet and had free access to water. After mice were euthanized by cervical dislocation, we perfused the left kidney with 5 ml collagenase type 2 (1 mg/1 ml) containing L-15 medium (Life Technology). The collagenase perfused kidney was removed, and the renal cortex was cut with a sharp razor. We followed the method described previously and in the Supplemental Material to dissect the DCT/CNT. Supplemental Figure 1 shows a typical DCT/CNT image. 9 The patch-clamp experiments were performed in the DCT as evidenced by the presence of NCC mRNA (Supplemental Figure 2).

Perforated Whole-Cell Recording
An Axon 200A patch-clamp amplifier was used for the measurement of the whole-cell recording, and K currents were low-pass filtered at 1 KHz, digitized by an Axon interface (Digidata 1322). Data were analyzed using the pCLAMP Software System 9 (Axon). We carried out the perforated whole-cell patch-clamp experiments on HEK293T cells or the DCT1 at room temperature as described previously. 43 For measuring the whole cell K currents, the cells/DCT1 were incubated with a bath solution containing 140 mM KCl, 1.8 mM MgCl 2 , 1.8 mM CaCl 2 , and 10 mM HEPES (pH 7.4). For measuring the K reversal potential using perforated whole-cell patch, the DCT was superfused with HEPES-buffered NaCl solution containing 140 mM NaCl, 5 mM KCl, 1.8 mM CaCl2 , 1.8 mM MgCl2 and 10 mM HEPES (pH=7.4). Borosilicate glass (1.7-mm OD) was used to make the patch-clamp pipettes with a Narishege electrode puller. The pipette had a resistance of 2-4 MV when filled with 140 mM KCl. The tip of the pipette was filled with pipette solution containing 140 mM KCl, 2 mM MgCl 2 , 1mM EGTA, and 5 mM HEPES (pH 7.4). The pipette was then backfilled with amphotericin B (20 mg/0.1 ml) containing the pipette solution. After forming a high-resistance seal (.2 GV), the membrane capacitance was monitored until the whole-cell patch configuration was formed.
The whole-cell K current was determined by adding 1 mM Ba 2+ in the bath solution.

Single Patch-Clamp Study in DCT1
An Axon200B patch-clamp amplifier was used to record the single channel current. The currents were low-pass filtered at 1 KHz and digitized by an Axon interface. Channel activity defined as NP o (a product of channel number and open probability) was calculated from data samples of 60 seconds duration in the steady state as follows: NP o ¼ ∑ðt 1 þ 2t 2 þ ::::::it i Þ where t i is the fractional open time spent at each of the observed current levels. The composition of the pipette solution for the singlechannel recording was the same as that for the whole-cell recording, whereas the bath solution was HEPES-buffered 140 mM NaCl.

Immunostaining
Mice were anesthetized with ketamine (100 mg/kg) and clonidine analogue (10 mg/kg), and the abdomens were cut open for perfusion of kidneys with PBS containing heparin (40 unit/ml) followed by 20 ml of 4% paraformaldehyde. After perfusion, the kidneys were removed and subjected to post-fixation with 4% paraformaldehyde for 12 hours. The kidneys were dehydrated and cut to 8-10 mM slices with Leica1900 cryostat (Leica). The tissue slices were dried at 42°C for 1 hour. The slides were washed with 13 PBS for 15 minutes and permeabilized with 0.3% Triton dissolved in 1XPBS buffer containing 1% BSA and 0.1% lysine (pH=7.4) for 15 minutes. Kidney slices were blocked with 2% horse serum for 30 minutes at room temperature and were then incubated with primary antibodies (parvalbumin and cav-1) for 12 hours at 4°C. Slides were thoroughly washed with 13 PBS and followed by an addition of second antibodies mixture in 0.4% Triton dissolved in 13 PBS for 2 hours at room temperature.

Preparation of Protein Samples, Western Blot, and Quantitative Real-Time PCR
The proteins for Western blot were obtained from HEK293T cells transfected with the corresponding DNA or mouse renal cortex. The tissues were prepared with ice-cold homogenization solution containing 250 mM sucrose, 50 mM Tris-HCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 1% protease, and phosphatase inhibitor cocktails titrated to a pH of 7.6. Tissue samples were homogenized in ice-cold homogenization solution using a tissue homogenizer. After homogenization, protein concentration was measured using the DC Protein Assay Kit (Bio-Rad, Hercules, CA). The proteins were separated by electrophoresis on 4%-15% SDS-polyacrylamide gels and transferred to nitrocellulose membrane. The membranes were blocked with LI-COR blocking buffer (PBS). An Odyssey Infrared Imaging System (LI-COR, Lincoln, NE) was used to scan the membrane at a wavelength of 680 or 800 nM. The

Metabolic Cage Study
We followed the protocol for the metabolic cage study described previously. 44 Briefly, we used eight WT mice and eight cav-1 KO mice for the study, and both WT and KO mice were age and sex matched. Two mice were housed in a cage for the study, and they stayed in the cage for 3-4 days to allow them to be adapted with the environment. After training, mice were fed with a normal diet (1%K+0.3%Na) for an additional 3 days, and their 24-hour food intake, urine output, and body weight were recorded daily. The urine samples were collected for measuring the electrolytes profile. We took the mean value from 3-day measurements for each cage. At the end of the experiments, the blood samples were collected through cardiac puncture. Plasma and urinary Na + , K + , Ca 2+ , and Mg 2+ concentrations were measured at the Yale O'Brienne Kidney Center, and urinary electrolyte profiles were presented as milliequivalent per gram of body weight per 24 hours.

Experimental Materials and Statistics
cav-1 (sc-894) and parvalbumin (sc-7449) antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). We obtained Flag (f-3165) and ENaC-b (SAB5200106) antibodies from Sigma-Aldrich (St. Louis, MO), c-Src (#2123) antibody from Cell Signaling Technology (Danvers, MA), and actin antibody (ab3280) from Abcam, Inc., (Cambridge, MA). The antibodies for Kcnj10 (APC-035), ENaC-a (ASC-030, and TRPV5 (ACC-035) were from Alomone Labs (Jerusalem, Israel). The antibody for NCC was a gift from Robert Hoover at Emory University, whereas antibodies of SPAK and Thr 53 phosphorylated NCC were kindly provided by David Ellison/James McCormick at Oregon Health & Science University. 31 The data are presented as mean6SEM. We used the paired t test or one-way ANOVA to determine the statistical significance. Figure 11. Cell scheme illustrating the mechanism by which the downregulation of cav-1 impairs the regulation of the basolateral Kir4.1 by SFK in the DCT1, thereby inhibiting NCC and decreasing the driving force for transcellular Mg 2+ /Ca 2+ absorption in the DCT. Dotted lines represent a diminished signaling.