Regulation of Inositol 1,4,5-Trisphosphate Receptor-mediated Calcium Release by the Na/K-ATPase in Cultured Renal Epithelial Cells*

It is known that the Na/K-ATPase α1 subunit interacts directly with inositol 1,4,5-triphosphate (IP3) receptors. In this study we tested whether this interaction is required for extracellular stimuli to efficiently regulate endoplasmic reticulum (ER) Ca2+ release. Using cultured pig kidney LLC-PK1 cells as a model, we demonstrated that graded knockdown of the cellular Na/K-ATPase α1 subunit resulted in a parallel attenuation of ATP-induced ER Ca2+ release. When the knockdown cells were rescued by knocking in a rat α1, the expression of rat α1 restored not only the cellular Na/K-ATPase but also ATP-induced ER Ca2+ release. Mechanistically, this defect in ATP-induced ER Ca2+ release was neither due to the changes in the amount or the function of cellular IP3 and P2Y receptors nor the ER Ca2+ content. However, the α1 knockdown did redistribute cellular IP3 receptors. The pool of IP3 receptors that resided close to the plasma membrane was abolished. Because changes in the plasma membrane proximity could reduce the efficiency of signal transmission from P2Y receptors to the ER, we further determined the dose-dependent effects of ATP on protein kinase Cϵ activation and ER Ca2+ release. The data showed that the α1 knockdown de-sensitized the ATP-induced ER Ca2+ release but not PKCϵ activation. Moreover, expression of the N terminus of Na/K-ATPase α1 subunit not only disrupted the formation of the Na/K-ATPase-IP3 receptor complex but also abolished the ATP-induced Ca2+ release. Finally, we observed that the α1 knockdown was also effective in attenuating ER Ca2+ release provoked by angiotensin II and epidermal growth factor.

The Na/K-ATPase is a highly expressed integral membrane protein that hydrolyzes ATP to pump Na ϩ and K ϩ across the membrane (1). It also functions as an important signal transducer (2). Recent studies have demonstrated that cells appear to contain two functionally separable pools of Na/K-ATPase and that a majority of the cellular Na/K-ATPase is engaged in cellular activities other than pumping ions (3). Moreover, the non-pumping Na/K-ATPase apparently resides in caveolae and interacts directly with protein kinases, ion channels, and transporters (4). The interaction between the Na/K-ATPase and Src, for example, forms a functional receptor complex for cardiotonic steroids such as ouabain to activate protein tyrosine phosphorylation (5,6). Interestingly, recent studies from several laboratories have demonstrated a direct interaction between the ␣ subunit of Na/K-ATPase and inositol 1,4,5trisphosphate receptors (IP 3 Rs) 2 (7)(8)(9)(10). In addition, we have found that ouabain was capable of stimulating the formation of a functional Ca 2ϩ -signaling complex consisting of the Na/K-ATPase/Src/PLC-␥/IP 3 R in LLC-PK1 cells (9). Furthermore, we have shown that the formation of this signaling complex plays an important role in ouabain-induced Ca 2ϩ signal transduction.
The cytosolic free calcium is one of the most important cellular second messengers. Calcium enters the cytosol via the opening of Ca 2ϩ channels that either resides in the plasma membrane or in the membranes of intracellular organelles such as the ER. On the ER, IP 3 Rs are widely expressed and play an important role in converting the activation of many plasma membrane receptors into intracellular Ca 2ϩ signaling. To date, three isoforms of IP 3 Rs have been identified, sharing similar but not identical functional properties (11). The signaling events leading to the ER Ca 2ϩ release from these IP 3 Rs are initiated by the activation of phospholipase C-coupled plasma membrane receptors such as the Na/K-ATPase and P2Y receptors and subsequent generation of IP 3 , a ligand for IP 3 Rs. It has been proposed that the formation of the junctional microdomains that force the proximity of IP 3 Rs to the plasma membrane receptors provides a mechanism for defining spatially and temporally specific Ca 2ϩ signaling (12)(13)(14)(15). For instance, the forced coupling between B2 bradykinin receptors and IP 3 Rs ensures a robust Ca 2ϩ signaling when the receptor is activated by the ligand in neuronal cells (16). These findings raise the question as to whether there is a plasma membrane protein that can function as an anchor to interact and pull the ER IP 3 Rs to the proximity of the plasma membrane receptors. * This work was supported by National Institutes of Health Grants HL-36573, HL-67963, and GM-78565 and by American Heart Association Grant 0130231N. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
It is well documented that many G protein-coupled receptors and receptor tyrosine kinases are concentrated in caveolae (17,18). Because the Na/K-ATPase represents a highly abundant caveolar membrane protein, it is conceivable that the basal interaction between the caveolar Na/K-ATPase and ER IP 3 Rs could force the proximity of ER IP 3 Rs to other plasma membrane receptors. Thus, the interaction between the Na/K-ATPase and IP 3 Rs may not only be important for ouabain-induced ER Ca 2ϩ release but also play a role in other stimuli-induced Ca 2ϩ signaling. To test this hypothesis, we examined the effect of graded knockdown of Na/K-ATPase on ATP-induced ER Ca 2ϩ release. The data presented here clearly demonstrated that the cellular Na/K-ATPase regulates ER Ca 2ϩ release by interacting and targeting the IP 3 Rs in LLC-PK1 cells.

EXPERIMENTAL PROCEDURES
Materials-Fura-2 AM, Image-iT FX signal enhancer, antifade kit, Alexa Fluor 488-conjugated anti-mouse and anti-goat IgG, and Alexa Fluor 555-conjugated anti-mouse IgG antibodies were obtained from Molecular Probes (Eugene, OR). U73122 and U73343 were obtained from Cayman (Ann Arbor, MI). Sulfo-NHS-biotin, immobilized streptavidin-agarose beads were obtained from Pierce. Anti-P2Y 2 receptor polyclonal antibody was obtained from Zymed Laboratories Inc.; anti-YFP polyclonal antibody was obtained from Abcam (Cambridge, MA); and anti-calnexin polyclonal antibody was obtained from StressGen Bioreagents Corp. (Victoria, British Columbia, Canada). The sources of other primary antibodies were described previously (9). All the horseradish peroxidaseconjugated secondary antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA).
Cell Preparation and Culture-LLC-PK1 cells and the Na/K-ATPase ␣1 knockdown (e.g.A4-11, PY-17, and TCN23-19) and knock-in (AAC-19) cells were maintained in the growth medium as described previously (19). When cell cultures reached about 80% confluence, cells were serum-starved for 24 h and used for the experiments.
Plasmids and Transfection-N-terminal (Ala 1 -Ser 160 ) of the rat ␣1 was cloned from rat Na/K-ATPase ␣1 cDNA provided by Dr. Pressley (Texas Tech University) and inserted in-frame into pEYFP-N1 vector (Clontech). The construct was verified by DNA sequencing. For calcium imaging, LLC-PK1 cells were transfected with various plasmids using calcium phosphate precipitation (20). Experiments were performed 24 h after transfection unless indicated otherwise.
Calcium Imaging in Live Cells-Intracellular calcium concentration was measured as described previously (9). In brief, cells were incubated with 2 M Fura-2 AM at 25°C for 40 min in a physiological salt solution (ECB) containing 100 mM NaCl, 5 mM KCl, 20 mM HEPES, 25 mM NaHCO 3 , 1 mM CaCl 2 , 1.2 mM MgCl 2 , 1 mM NaH 2 PO 4 , and 10 mM D-glucose. Coverslips were placed in a recording/perfusion chamber (model QE-1, Warner Instruments) mounted on the stage of an inverted microscope (model IX71, Olympus) equipped with a 40ϫ oil-immersion Fluor objective. Excitation light alternated between 340 and 380 nm, and emitted light was recorded at Ͼ510 nm. Ratio images were acquired and analyzed using SlideBook software (Slide-Book 4.1.0, Intelligent Imaging Innovations, Inc.). Cells were perfused with ECB at 37°C at the speed of 0.5 ml/min, and then the solution was quickly changed to a nominally calcium-free ECB supplemented with 250 M EGTA. Cells were excited for 2 ms every 5 s and monitored for less than 30 min.
Assay of IP 3 Production and PKC⑀ Translocation-GFPfused PLC-␦1 PH domain (PHD-GFP) was used to monitor PIP 2 hydrolysis and IP 3 production in live cells as described previously (9). Cells were transfected with PHD-GFP in Opti-MEM medium using Lipofectamine 2000. After 24 h, the cells were exposed to stimulus and monitored for changes in GFP signal in the plasma membrane and the corresponding cytosol.
To analyze PKC activation, cells were treated and then washed with ice-cold phosphate-buffered saline. Afterward, cells were collected in a buffer containing 10 mM EGTA, 1 mM EDTA, 0.5 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 50 g/ml leupeptin, 25 g/ml aprotinin, and 20 mM Tris-HCl (pH 7.5) and homogenized in a Potter-Elvehjem homogenizer as described previously (21). The suspension was centrifuged at 100,000 ϫ g for 1 h at 4°C. The supernatant was removed (the cytosolic fraction). The pellet was suspended in the above described buffer to which Triton X-100 (1%) was added and sonicated for 10 s. After 30 min of incubation on ice, the mixture was centrifuged at 25,000 ϫ g for 10 min, and the supernatant was collected (the particulate fraction). To assay the translocation of PKC⑀, the cytosolic and the particulate fractions were subjected to immunoblot analysis as described before.
Immunocytochemistry-Cells on coverslips were fixed in 4% paraformaldehyde for 30 min at 25°C. Then the cells were permeabilized in fresh phosphate-buffered saline containing 0.5% saponin for 10 min and blocked with Image-iT FX signal enhancer for 30 min at room temperature. The cells were then incubated with primary antibodies (mouse anti-IP 3 R3 Ab, goat anti-IP 3 R2 Ab, and anti-Na/K-ATPase ␣1 monoclonal Ab) at 4°C overnight. They were washed and then incubated with suitable Alexa Fluor 488-and Alexa Fluor 555-conjugated secondary antibodies at room temperature for 1 h. Coverslips were mounted using an antifade kit. Image analyses were performed as described previously (6,19).
Biotinylation of Cell Surface Proteins-The cells were washed twice with ice-cold phosphate-buffered saline and then biotinylated as described previously (3,22). The biotinylated cells were then lysed with 1 ml of radioimmunoprecipitation buffer for 30 min at 4°C. Cell lysates were cleared by centrifugation at 14,000 ϫ g for 15 min, and the supernatants (500 g) were mixed with 200 l of streptavidin-conjugated agarose beads overnight at 4°C to recover the biotinylated proteins. The recovered proteins and 50 g cell lysates were subjected to Western blot analysis.
Immunoprecipitation of IP 3 R3 after Transient Transfection-LLC-PK1 cells were transiently transfected with eYFP or ␣NT-eYFP by calcium phosphate precipitation as described previously (20). After 24 h, cells were lysed in radioimmunoprecipitation buffer, and lysates were prepared as above. Immunoprecipitation of IP 3 R3 was conducted as described previously (9).
Statistical Analysis-All traces and immunoblots presented are representative of at least three separate experiments. All quantitative data are given as mean Ϯ S.E. Comparisons between two groups were analyzed via the Student's t test, and significance was accepted at p Ͻ 0.05.

Knockdown of Na/K-ATPase Reduces ATP-induced Ca 2ϩ
Release in LLC-PK1 Cells-Extracellular ATP is known to modulate Ca 2ϩ mobilization either through the activation of P2Y receptor coupled to a G protein, or through the direct activation of the inotropic P2X receptor. LLC-PK1 cells are known to express both P2X and P2Y receptors (23,24). Thus, we first determined the properties of ATP-induced Ca 2ϩ mobilization in LLC-PK1 cells. As depicted in Fig. 1, A and B, when cells were superfused with a nominally Ca 2ϩ -free solution, application of 20 M ATP induced a robust Ca 2ϩ transient, which reached the peak value within 1 min and then declined toward the basal level. In the presence of extracellular Ca 2ϩ , ATP stimulated a similar but prolonged Ca 2ϩ transient that apparently involved both Ca 2ϩ release and Ca 2ϩ influx. To further test whether the ATP-induced Ca 2ϩ transient in the absence of extracellular Ca 2ϩ was because of the activation of PLC, the cells were first treated with PLC inhibitor U73122 and then exposed to ATP. As depicted in Fig. 1C, addition of U73122 completely abolished the effect of ATP on intracellular Ca 2ϩ . On the other hand, the negative control U73343 did not affect ATP-induced Ca 2ϩ release (data not shown). Thus, we performed the following studies in cells incubated in a nominally Ca 2ϩ -free solution to assess the role of Na/K-ATPase in IP 3 R-mediated ER Ca 2ϩ release.
Recently, we have generated several Na/K-ATPase ␣1 knockdown cell lines from LLC-PK1 cells that were transfected with an ␣1-specific small interfering RNA (19). Although A4-11 cells express about 44% of ␣1 subunit in comparison with the control LLC-PK1 cells; PY-17 cells contain less than 10% of ␣1 subunit ( Fig. 2A). When these cells were exposed to 20 M ATP and monitored for changes in intracellular Ca 2ϩ , we observed a significant decrease in ATP-induced Ca 2ϩ release in both A4-11 and PY-17 cells (Fig. 2B). When the peak value was quantified ( Fig. 2C), the effect of ATP on intracellular Ca 2ϩ was proportional to the amount of cellular Na/K-ATPase. Moreover, when PY-17 cells were rescued by a rat ␣1, expression of rat ␣1 not only restored cellular Na/K-ATPase (19) but also ATP-induced ER Ca 2ϩ release (Fig. 2, B and C). Taken together, these findings indicate that knockdown of cellular Na/K-ATPase attenuated the ATP-induced ER Ca 2ϩ release. Because this reduction was observed in several knockdown cell lines, was proportional to cellular Na/K-ATPase amounts, and could be rescued by expression of exogenous Na/K-ATPase, it was unlikely a cloning artifact.
Inhibition of Na/K-ATPase by Ouabain Does Not Attenuate ATP-induced Ca 2ϩ Release-We showed previously that Na/K-ATPase knockdown resulted in a significant inhibition of cellular sodium pump activity and a subsequent increase in intracellular Na ϩ concentration (3). To rule out that the attenuation of ATP-induced Ca 2ϩ release is because of a simple inhibition of cellular Na/K-ATPase activity and subsequent increase in intracellular Na ϩ concentration, we first compared the basal Ca 2ϩ in different cells. As shown in Fig. 2B, there was no detectable difference in basal Ca 2ϩ among these cell lines. Furthermore, when LLC-PK1 cells were treated with different concentrations of ouabain and then measured for ATPinduced Ca 2ϩ release, we found that neither 50% inhibition (1 M) nor 100% inhibition (10 M) of Na/K-ATPase by ouabain (25) changed ATP-induced Ca 2ϩ release in LLC-PK1 cells (Fig. 2D).
Na/K-ATPase Knockdown Does Not Affect Purinergic Receptor Function and ER Calcium Contents-To date, pharmacological characterization and molecular cloning have identified eight P2Y receptor subtypes, of which P2Y 2 receptor is highly expressed in renal epithelial cells (26). We showed that LLC-PK1 cells expressed both IP 3 R2 and IP 3 R3 (9). To test whether changes in the cellular amount of Na/K-ATPase affect the expression of P2Y receptor and IP 3 R, we performed Western blot analysis of the total cell lysates from control LLC-PK1 cells, the knockdown and knock-in cells. As shown in Fig. 3A, knockdown of Na/K-ATPase did not change the expression of these receptors.
Binding of ATP to P2Y receptor activates PLC-␤, which subsequently catalyzes the production of second messengers diacylglycerol (DAG) and IP 3 . To determine whether Na/K-ATPase knockdown affected the P2Y receptor function, we measured the effect of ATP on DAG production using PKC⑀ activation as an indicator. As shown in Fig. 3B, cells were stimulated with 20 M ATP for 5 min. Analysis of PKC⑀ translocation demonstrated a similar activation of PKC⑀ by ATP in control LLC-PK1, Na/K-ATPase knockdown and knock-in cells (Fig. 3C).
To seek further support, we also measured ATP-induced IP 3 production in live cells using a GFP-fused PLC-␦1 PH domain protein as a probe (27). The imaging analysis was conducted in the control LLC-PK1 cells and another ␣1 knockdown TCN23-19 cell line. Like PY-17 cells, TCN23-19 cells expressed about 10% of Na/K-ATPase in comparison with the control LLC-PK1 cells (19). Moreover, the ATP-induced ER Ca 2ϩ release was similarly attenuated as in PY-17 cells (Fig. 3D). Importantly, expression of rat ␣1 in TCN23-19 cells also restored the effect of ATP on ER Ca 2ϩ release (data not shown). However, unlike PY-17 cells, these cells did not express YFP, making it easier to run GFP imaging analysis and immunostaining. As depicted in Fig. 3E, cells were transiently transfected with the vector expressing GFP-fused PLC-␦1 PH domain. In unstimulated cells, the PH domain bound to PIP 2 and mainly associated with the plasma membrane. When PIP 2 was hydrolyzed by PLC to produce IP 3 in response to the addition of 20 M ATP, the PH domain translocated into the cytoplasm together with IP 3 . This resulted in an increase in intensity  of cytosolic GFP and a concomitant decrease in the plasma membrane GFP signal and thus an overall increase in the ratio of cytosol versus membrane GFP signal (Fig. 3E). Consistent with the findings of Fig. 3B, we found that the Na/K-ATPase knockdown did not significantly change the ATP-induced IP 3 production. Taken together, the above findings indicate that the attenuation of ATP-induced Ca 2ϩ release in Na/K-ATPase knockdown cells is unlikely because of the changes in the number and function of P2Y receptors.
To rule out that the Na/K-ATPase knockdown reduced ER Ca 2ϩ contents, we determined cellular Ca 2ϩ after the cells were treated with TG. TG-induced ER Ca 2ϩ depletion has been used by many investigators as an indirect measurement of ER calcium content (28). We found no significant difference in the amount of Ca 2ϩ released by TG treatment in control LLC-PK1 cells and Na/K-ATPase knockdown cells (Fig. 3F).
Knockdown of Na/K-ATPase Changes the Cellular Distribution of IP 3 Rs-Because the Na/K-ATPase interacts with IP 3 R, it is quite possible that this interaction plays an important role in regulating the subcellular distribution of IP 3 Rs. To test this hypothesis, the distribution of IP 3 receptor subtypes was investigated by immunocytochemistry. As depicted in Fig. 4, A and B, IP 3 R2 or IP 3 R3 antibody-labeled LLC-PK1 cells showed a punctated ER staining pattern. Interestingly, there was also a clear cell periphery distribution of both IP 3 R2 and IP 3 R3. Such staining pattern has been reported previously in confluent Madin-Darby canine kidney cells (29,30). To analyze whether the IP 3 R resided at the cell periphery was co-localized with the Na/K-ATPase, LLC-PK1 cells were double labeled with a polyclonal anti-IP 3 R2 and a monoclonal anti-␣1 antibody. As expected, Na/K-ATPase showed a continuous plasma membrane staining pattern and was co-localized with IP 3 R2 (Fig. 4C).
Because anti-IP 3 R3 antibody produced a much clearer cell periphery signal, we repeated the same staining of IP 3 R3 in the TCN23-19 cells. Although the overall density of IP 3 R3 in TCN23-19 cells was similar to that in the control LLC-PK1 cells, knockdown of the Na/K-ATPase caused a significant change in subcellular distribution of IP 3 R3. Specifically, the cell periphery distribution of the IP 3 R3 was totally abolished (Fig. 4D).
Because IP 3 R could reside in the plasma membrane (22), the cell periphery distributed IP 3 R may come from two pools as follows: one from the IP 3 R in the plasma membrane and the other from the IP 3 R in the ER membrane that is in close proximity to the plasma membrane. To distinguish these two pools, we performed the cell surface biotinylation study. Quantitative analysis indicated that about 1.8% of total cellular IP 3 R3 was biotinylated in LLC-PK1 cells. Because the same procedure failed to detect any calnexin, a known 90-kDa integral ER membrane protein, in the biotinylated samples (Fig. 4E), the detected IP 3 R3 most likely resided in the plasma membrane, which is consistent with the prior observations (22). When the same experiments were performed in both TCN23-19 and PY-17 cells, we detected the same amount of biotin-labeled IP 3 R3 as in the control LLC-PK1 cells (Fig. 4, E and F). Thus, the disappearance of cell periphery IP 3 R3 in TCN23-19 cells was likely because of the redistribution of IP 3 R3s that resided in proximity to the plasma membrane. 3 Receptors-The above data suggest that the Na/K-ATPase may function as a plasma membrane scaffold, forcing the proximity and facilitating the coupling between the ER IP 3 R and the plasma membrane P2Y receptor. Should this be the case, we would expect that exposure of PY-17 cells to a higher concentration of ATP might compensate for the deficiency in coupling of ER IP 3 R to the P2Y receptor if higher concentration of ATP can activate additional receptors. To test this, we first compared the effects of 20, 50, and 100 M ATP on PKC⑀ activation and found that the activation of PKC⑀ reached the maximal level by 50 M ATP in both LLC-PK1 and PY-17 cells (Fig. 5, A and B). The effects of ATP on ER Ca 2ϩ release were also dose-dependent (Fig. 5C). Significant effect was observed when LLC-PK1 cells were exposed to 1 M ATP, and the maximal effect was reached at 10 M ATP. Thus, in LLC-PK1 cells, partial activation of the plasma membrane P2Y receptors by 10 M ATP was sufficient to induce the maximal ER Ca 2ϩ release. However, when ATP concentration curve on ER Ca 2ϩ release was determined in PY-17 cells, we found that 50 M ATP was required to produce the maximal Ca 2ϩ release (Fig. 5C). Moreover, the effect of 50 M ATP on Ca 2ϩ release in PY-17 cells was only comparable with that produced by 1 M ATP in LLC-PK1 cells. Finally, 1 M ATP failed to elicit ER Ca 2ϩ release in PY-17 cells. To be sure that this decrease in ATP sensitivity is not a cloning artifact of PY-17 cells, we repeated the ATP concentration curve on ER Ca 2ϩ release in TCN23-19 cells, observing an essentially identical dose-response curve as in PY-17 cells (data not shown). Taken together, our data clearly indicate that cellular Na/K-ATPase plays an important role for efficient signal transmission from the activated plasma membrane P2Y receptors to the ER IP 3 R.

The N Terminus of the Na/K-ATPase Functions as a Dominant Negative Regulator of ATP-induced ER Ca 2ϩ
Release-Results from recent studies indicate that the N terminus of the Na/K-ATPase ␣1 subunit binds to IP 3 R directly (9,10). To detect whether the effect of Na/K-ATPase on ER Ca 2ϩ release is because of the direct interaction between Na/K-ATPase and IP 3 R, we constructed a YFP fusion protein containing the ␣1 N terminus (amino acids 1-160) (␣NT-eYFP). As shown in Fig.  6A, the expression of ␣NT-eYFP or eYFP in LLC-PK1 cells was dependent on the amount of vectors used in the transfection. When intracellular Ca 2ϩ was measured in the transfected cells (indicated by the eYFP signal), we found that expression of ␣NT-eYFP, but not eYFP, dose-dependently inhibited ATPinduced ER Ca 2ϩ release. Complete inhibition was achieved when LLC-PK1 cells were transfected with 2.0 g of ␣NT-eYFP (Fig. 6, B and C). As an additional control, the effects of ATP on ER Ca 2ϩ release were also measured in cells that were cultured on the same coverslip but were not transfected (no detectable YFP signal). As expected, these untransfected cells showed intact ability to release ER Ca 2ϩ in response to ATP stimulation (data not shown).
To further test if ␣NT-eYFP can disrupt the interaction between Na/K-ATPase and IP 3 R, the LLC-PK1 cells were transfected with 2.0 g of eYFP and ␣NT-eYFP constructs, respectively. After 24 h, the total cell lysates (including both the transfected and untransfected cells) were immunoprecipitated with an anti-IP 3 R3 antibody and then probed for co-precipitated ␣1. As depicted in Fig. 7A, expression of the N terminus caused more than 45% reduction in the co-precipitated ␣1 subunit. Because the transfection efficiency was about 50%, these findings clearly demonstrated that the ␣1 N terminus was an effective dominant negative regulator of the interaction between the Na/K-ATPase and IP 3 R. Knockdown of the Na/K-ATPase Also Inhibits Angiotensin II and EGF-induced Ca 2ϩ Release-If the Na/K-ATPase is a key scaffold protein for bringing IP 3 R to the proximity of plasma membrane, we would expect that ␣1 knockdown should also affect other stimuli-induced ER Ca 2ϩ release. To test this, we determined the effect of Na/K-ATPase knockdown on other G protein-coupled receptors as well as EGF receptor-mediated Ca 2ϩ release. We have used EGF as a positive control to study ouabain-induced protein tyrosine phosphorylation in LLC-PK1 cells (25). Others have shown that angiotensin II stimulates signal transduction in LLC-PK1 cells (31). As illustrated in Fig.  8A, angiotensin II was effective in stimulating ER Ca 2ϩ release in LLC-PK1 cells. Like ATP, its effect on Ca 2ϩ was significantly diminished in the Na/K-ATPase knockdown PY-17 cells (Fig.  8B). Moreover, when the same experiments were repeated with EGF, we found that EGF-induced Ca 2ϩ release was also attenuated (Fig. 8, C and D).

DISCUSSION
In this study we reported two major findings. First, knockdown of the Na/K-ATPase changed the subcellular distribution of IP 3 R and attenuated the ER Ca 2ϩ release provoked by the activation of a number of plasma membrane receptors. Second, disruption of the interaction between the Na/K-ATPase and IP 3 R by overexpression of the ␣1 N terminus functioned as the Na/K-ATPase knockdown and reduced ATP-induced ER Ca 2ϩ release. These new findings suggest a novel structural and functional role of Na/K-ATPase, which may serve as a focal point for recruiting the ER IP 3 Rs into proximity of the plasma membrane receptors and facilitating IP 3 R-mediated Ca 2ϩ signaling (Fig. 9).
The Na/K-ATPase Enhances the Efficiency of Signal Transmission from the Plasma Membrane Receptor Activation to the ER Ca 2ϩ Release-We provided evidence that the Na/K-ATPase is required for several extracellular stimuli to efficiently stimulate ER Ca 2ϩ release in LLC-PK1 cells. Specifically, both knockdown of the Na/K-ATPase and overexpression of the N terminus of the ␣1 subunit were sufficient to attenuate ATPinduced Ca 2ϩ signaling. Although the knockdown of Na/K-ATPase did result in an increase in intracellular Na ϩ concentration, it is unlikely that the de-sensitization of the receptor activation-provoked ER Ca 2ϩ release is because of the simple inhibition of the pumping function of the Na/K-ATPase. First, although PY-17 cells and A4-11 cells exhibited similar pumping activity (3), ATP-induced ER Ca 2ϩ release was further reduced in PY-17 cells that expressed less Na/K-ATPase (Fig.  2B). Second, complete inhibition of Na/K-ATPase by ouabain did not change ATP-induced ER Ca 2ϩ release in LLC-PK1 cells (Fig. 2D).
Finally, there was no detectable change in basal Ca 2ϩ in the Na/K-ATPase knockdown cells (Fig. 2B), and the ER Ca 2ϩ content was similar among different cell lines (Fig. 3F). As shown in Figs. 3 and 5, the Na/K-ATPase knockdown specifically attenuated the IP 3 -induced Ca 2ϩ signaling but not DAG-provoked activation of PKC⑀. A similar pattern of Ca 2ϩ regulation has been reported in neuronal cells exposed to agonists of muscarinic M1 and B2 bradykinin receptors (16). Specifi-  cally, it was found that although both receptors were equally effective at activating PKC, only the B2 receptor agonist could elicit robust ER Ca 2ϩ release from the IP 3 Rs. Mechanistically, this difference in receptor-mediated Ca 2ϩ signaling is because of the different distribution of these receptors in the plasma membrane. Although B2 receptors and IP 3 Rs are physically coupled, the M1 receptors are remote from IP 3 Rs, leading to diffusion of IP 3 over a significant distances and attenuation of IP 3 -induced Ca 2ϩ signaling. Consistent with this mode of regulation, we observed that knockdown of Na/K-ATPase shifted the ATP concentration curve to the right (Fig. 5). Needless to say, other possibilities may also be in play. For example, knockdown of Na/K-ATPase reduces the interaction between the Na/K-ATPase and IP 3 R, which may simply reduce the sensitivity of IP 3 R to IP 3 or change the channel opening properties as suggested previously (10). These issues have to be addressed in future studies.
In addition to the reduction in ATP sensitivity, the maximal effect of ATP on Ca 2ϩ release was also attenuated by Na/K-ATPase knockdown (Fig. 5). Although the exact molecular basis of this defect is unknown, it is of great interest to note the existence of a well characterized Ca 2ϩ /calmodulin-mediated negative feedback mechanism in the cell (32,33). This regulatory mechanism serves as a filter to efficiently inhibit slow rather than rapid rises in IP 3 . If the local IP 3 does not reach a sufficient concentration with appropriate speed, the second messenger will fail to or cannot sufficiently initiate the Ca 2ϩ signaling. Taken together, our data support the notion that the Na/K-ATPase plays an important role in coupling the activation of the plasma membrane receptors to the opening of ER IP 3 Rs.
The Nonpumping Na/K-ATPase as a Structural Anchor for Targeting ER IP 3 Rs into Junctional Microdomains-There is solid evidence that a privileged communication between the plasma membrane receptors and the apposing ER is established in many cells to ensure specificity and efficiency in Ca 2ϩ signaling (16,34,35). Structurally, these communications occur in junctional microdomains where the plasma membrane and the opposing ER are separated by a small gap of about 10 -25 nm (36). Interestingly, in neuronal cells the coupling between metabotropic glutamate receptors and ER IP 3 Rs appears to be mediated by the Homer family of proteins (37).
Caveolae are uncoated invaginations of the cell surface and are highly enriched in signaling proteins, including receptor tyrosine kinases, G protein-coupled receptors, Ca 2ϩ -ATPase, and several PKC isoforms (38). Direct measurements of calcium changes in endothelial cells suggest that caveolae may be sites that regulate intracellular Ca 2ϩ concentration and Ca 2ϩdependent signal transduction (39). Moreover, there is evidence that caveolae are in close contact with the ER and these contacts are mechanically stable (40). Because the nonpumping Na/K-ATPase represents an abundant caveolar protein, our new findings lead us to speculate that the nonpumping Na/K-ATPase may serve as a structural anchor for recruiting and targeting IP 3 R to the caveolae/ER junctional microdomain (Fig.  9). This notion is supported by the following observations. First, overexpressing the ␣1 N terminus was sufficient to block the interaction between Na/K-ATPase and IP 3 Rs (Fig. 7). Consequently, it also diminished ATP-induced ER Ca 2ϩ release (Fig.  6). Second, there was a pool of IP 3 R3 resided at the cell periphery in LLC-PK1 cells. This pool of IP 3 R3 disappeared in the Na/K-ATPase knockdown cells (Fig. 4). Finally, the Na/K-ATPase is known to interact with ankyrin and spectrins (41)(42)(43). These interactions will certainly provide structural stability for the formation of junctional Ca 2ϩ signaling microdomains.
It is important to note that in addition to serving as a plasma membrane anchor for targeting IP 3 Rs, the nonpumping Na/K-ATPase is also associated with PLC-␥ and Src family kinases in the caveolae (6,9). It is known that Src family kinases can phosphorylate IP 3 Rs and make them more sensitive to IP 3 stimulation (44). Thus, it is logical for us to suggest that the formation of junctional microdomain of the Na/K-ATPase-PLC-Src-IP 3 R complex ensures the efficient signal transmission from the activation of plasma membrane receptors into ER Ca 2ϩ release (Fig. 9). This notion was further supported by the fact that the Na/K-ATPase knockdown not only de-sensitized ATP-, but also angiotensin II-induced as well as EGF-induced ER Ca 2ϩ release.
In conclusion, our study highlights a novel role of the Na/K-ATPase in regulation of ER IP 3 R-mediated Ca 2ϩ mobilization. It is worth noting that the expression of Na/K-ATPase is highly regulated by physiological stimuli, and significant down-regulation occurs under many pathological conditions (45)(46)(47). Thus, changes in cellular Na/K-ATPase could have profound effects on plasma membrane receptor-initiated Ca 2ϩ signaling. Moreover, de-sensitization of the receptor-mediated Ca 2ϩ signaling by the down-regulation of Na/K-ATPase may lead to an increased release of the ligands via a feedback mechanism, which could result in an overstimulation of the plasma mem- FIGURE 9. The role of Na/K-ATPase in the formation of the junctional Ca 2؉ signaling microdomain. Caveolar Na/K-ATPase interacts with the ER IP 3 R through its N terminus. This interaction could be further stabilized by the cytoskeleton protein such as ankyrin (8,41) and enhanced by ouabain-activated signaling pathways (7,9,10). Functionally, this interaction between Na/K-ATPase and IP 3 R helps targeting IP 3 R to the proximity of PLC-coupled plasma membrane receptors (e.g. GPCR, EGFR, and Na/K-ATPase), thus facilitating the IP 3 R-mediated Ca 2ϩ release from the ER. SOCs, store-operated calcium channels; PM, plasma membrane; N, N terminus; C, C terminus; CaM, calmodulin.