Heterogeneity of signal transduction by Na-K-ATPase α-isoforms: role of Src interaction

Abstract

Of the four Na-K-ATPase α-isoforms, the ubiquitous α1 Na-K-ATPase possesses both ion transport and Src-dependent signaling functions. Mechanistically, we have identified two putative pairs of domain interactions between α1 Na-K-ATPase and Src that are critical for α1 signaling function. Our subsequent report that α2 Na-K-ATPase lacks these putative Src-binding sites and fails to carry on Src-dependent signaling further supported our proposed model of direct interaction between α1 Na-K-ATPase and Src but fell short of providing evidence for a causative role. This hypothesis was specifically tested here by introducing key residues of the two putative Src-interacting domains present on α1 but not α2 sequence into the α2 polypeptide, generating stable cell lines expressing this mutant, and comparing its signaling properties to those of α2-expressing cells. The mutant α2 was fully functional as a Na-K-ATPase. In contrast to wild-type α2, the mutant gained α1-like signaling function, capable of Src interaction and regulation. Consistently, the expression of mutant α2 redistributed Src into caveolin-1-enriched fractions and allowed ouabain to activate Src-mediated signaling cascades, unlike wild-type α2 cells. Finally, mutant α2 cells exhibited a growth phenotype similar to that of the α1 cells and proliferated much faster than wild-type α2 cells. These findings reveal the structural requirements for the Na-K-ATPase to function as a Src-dependent receptor and provide strong evidence of isoform-specific Src interaction involving the identified key amino acids. The sequences surrounding the putative Src-binding sites in α2 are highly conserved across species, suggesting that the lack of Src binding may play a physiologically important and isoform-specific role.

INTRODUCTION

The Na-K-ATPase was first discovered as an ion pump (23). Four different isoforms have since been identified. While the α1-isoform is expressed ubiquitously, α2 and α3 are predominantly found in myocytes and neurons, respectively; and α4 is detected in sperm (3, 4, 22, 36). In addition to its cation pumping function, the α1 Na-K-ATPase has a receptor-like property, allowing its ligands such as cardiotonic steroids (CTS) to regulate cellular signal transduction via protein and lipid kinase cascades. Mechanistically, we have proposed that the α1 Na-K-ATPase is capable of interacting with Src kinase to form a receptor complex. We further speculated that the ligand-induced changes in the conformation of Na-K-ATPase alter the activity of the associated Src kinase and consequently the transactivation of EGF receptor followed by the assembly and activation of different kinase cascades in a cell-specific manner (2, 14, 21, 32). Elucidating the molecular basis of Na-K-ATPase-mediated signal transduction is key in assessing the role of this newly appreciated cellular signaling mechanism in animal physiology and disease progression. Accordingly, we and others have made great efforts to determine whether Na-K-ATPase interacts with Src kinase. Using a combination of in vitro binding assays, colocalization, FRET analyses and functional characterization of expressed α1 mutants in cells (18, 38), we have identified two important putative binding sites in the α1 subunit of Na-K-ATPase. One is between the second cytoplasmic domain (CD2) of Na-K-ATPase α1 subunit and Src homology 2 (SH2) domain and the other, which involves the Naktide sequence, is between the nucleotide binding (N) domain of α1 subunit and Src kinase domain (32). On the basis of the NaKtide sequence from the N domain, we have developed a peptide inhibitor, called pNaKtide, that is highly effective both in vitro and in vivo in blocking Na-K-ATPase-mediated signal transduction (20). Moreover, our recent in vivo studies have shown the effectiveness of this peptide in attenuating inflammatory signaling in animal models of obesity and uremic cardiomyopathy (26, 31), which not only provides strong evidence of Na-K-ATPase/Src interaction, but also demonstrates the importance of this signaling mechanism in the progression of several chronic diseases. Using a similar approach, we have identified Y260 in the CD2 domain as a SH2 binding site upon Src-mediated phosphorylation (1, 40). Although the involvement of Src and EGF receptor in the α1 Na-K-ATPase-mediated signal transduction has been reaffirmed and is widely accepted by many laboratories, whether α1 Na-K-ATPase interacts with Src to form a receptor complex remains under intense investigation. For example, a recent study by Yosef et al. (39) concluded that there is no direct interaction between human α1 Na-K-ATPase and Src, based on their work with purified recombinant human α1 Na-K-ATPase expressed in yeast and Src kinase expressed in bacteria. This is in sharp contrast to the direct binding of Src purified from sf-9 cells to purified pig and canine kidney α1 Na-K-ATPase that we have reported, and highlights the need for additional approaches to probe the proposed model of direct interaction. Indeed, the apparent discrepancies between those studies most likely originate from the inherent limitations of the respective approaches. Hence, our study used a nonphosphorylated Src, whereas the Yosef et al. study used a bacteria-expressed Src (phosphorylated at both Y418 and Y529 sites), which can affect the activity and binding of Src to other proteins (6). On the other hand, the kidney Na-K-ATPase preparation that we used is known to contain impurities, including caveolin-1 and Src, and several additional approaches have certainly been necessary to corroborate this finding and refine the model. As mentioned earlier, such approaches have included GST pull-down assays that indicated two putative Src binding sites in the α1 polypeptide (32), in vitro and in vivo blocking effects of pNaKtide (20, 26, 31), and loss-of-signaling function in NaKtide mutant α1 expressed in mammalian cells (18).

We (37) have recently demonstrated that the α2 isoform lacks α1-like Src interaction. As such, it fails to transmit the binding of CTS to the activation of protein kinase cascades. Sequence comparison reveals that the α2 isoform lacks the identified putative Src binding sites, namely the NaKtide sequence in the N domain, and Y260 in the CD2. This gave us a unique opportunity to 1) further assess differences in Src regulation by these two isoforms and 2) demonstrate the importance of such interactions in Na-K-ATPase-mediated signal transduction using a gain-of-function approach. Specifically, we created an α2 mutant that contains the α1 NaKtide sequence and Y260 and tested whether these modifications conferred Src-dependent signaling properties. To this end, we generated a cell line expressing almost exclusively the mutant form of α2, using our established knock-down and rescue approach (21), and then conducted functional comparison of α2 mutant cells with cells expressing either α2 or α1.

EXPERIMENTAL PROCEDURES

Materials.

Cell culture medium, fetal bovine serum, trypsin, Lipofectamine 2000, and polyclonal anti-pY418-Src were purchased from Invitrogen (Carlsbad, CA). Monoclonal anti-Src antibody (B12), polyclonal anti-ERK1/2 antibody, anti-caveolin-1 antibody, and goat anti-rabbit and goat anti-mouse secondary antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Polyclonal anti-pERK1/2 was obtained from Cell Signaling (Danvers, MA). Monoclonal anti-Na-K-ATPase-β1, polyclonal anti-Na-K-ATPase-α2, and monoclonal anti-Src GD11 antibody were from Millipore. Monoclonal anti-α1 antibody (α6F) was obtained from the Developmental Studies Hybridoma Bank at the University of Iowa. The QuikChange mutagenesis kit was obtained from Stratagene (La Jolla, CA).

Cell culture.

Pig kidney epithelial cells (LLC-PK1) were purchased from the American Type Culture Collection (Manassas, VA). Na-K-ATPase-α1 knock-down cells (called PY-17), rat α1-rescued PY-17 (called AAC-19), and rat α2-rescued PY-17 (called LX-α2-4) were derived from LLC-PK1 cells (21, 37). As detailed in these earlier publications, AAC-19 and LX-α2–4 cells are similarly ouabain resistant. This resistance is conferred through expression of the naturally occurring low-affinity rat α1 in AAC-19, and the low-affinity Q108R- N119D mutant rat α2 in LX-α2-4. Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) in the presence of 10% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin in a 5% CO₂ humidified incubator. After cells had reached 95–100% confluence, cells were serum-starved for 12 h before being used for experiments. Serum starvation eliminates confounding effects of serum ouabain and other growth factors on signal transduction.

Site-directed mutagenesis.

The QuikChange mutagenesis kit was used to generate mutant α2, using an ouabain-resistant rat-α2 expression vector as the template (15). Mutations were verified by DNA sequencing.

Generation of mutant α2 stable cell lines.

Generation of mutant α2 stable cell lines was done as previously described (21). Briefly, the α1 knock-down PY-17 cells were cultured in six-well plates and transfected with the mutant α2 vector using Lipofectamine 2000. The transfected cells were selected with 3 μM ouabain for 1 wk, and the surviving ouabain-resistant colonies were collected and diluted into 96-well plates to isolate a single colony. Once the colony was expanded into a stable cell line, the expression of mutant rat α2 was verified by α2-specific antibody. Cells were then cultured in the absence of ouabain for three generations before being used for experiments.

Western blot analysis.

Western blot analysis was performed as previously described (14). Cells were washed with PBS and solubilized in ice-cold RIPA buffer (1% Nonidet P-40, 1% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 1 mM NaF, 10 μg/ml aprotinin, 10 μg/ml leupeptin, and 50 mM Tris·HCl, pH 7.4). The cell lysates were then centrifuged at 14,000 rpm, and supernatants were subjected to Western blot analysis.

Cell growth assay.

Cell growth was assessed as previously described (33). Briefly, 50,000 cells per well were seeded in triplicates in 12-well plates and cultured in 10% FBS-DMEM. At each indicated time point, cells were trypsinized and counted.

Ouabain-sensitive Na-K-ATPase activity.

The Na-K-ATPase activity was measured as previously described (37). Briefly, cells were harvested in ice-cold Skou C buffer (30 mM histidine, 250 mM sucrose, 1 mM EDTA, pH 7.4) and homogenized. After initial centrifugation (800 g for 10 min), the postnuclear fraction was further centrifuged (100,000 g for 45 min) to get crude membrane. The crude membrane pellet was resuspended in Skou C buffer and treated with alamethicin (0.1 mg/mg protein) for 10 min at room temperature and then subjected to ouabain-sensitive ATPase activity assay.

Immunoprecipitation.

Immunoprecipitation assay was performed as previously described (14). Briefly, cell lysates were incubated with monoclonal anti-Src antibody and then with protein G-agarose. After extensive washes, immunoprecipitates were collected and subjected to Western blot analysis.

[³H]ouabain binding.

To measure the surface expression of the endogenous pig Na-K-ATPase, [³H]ouabain binding was measured as described (33). Cells were cultured in 12-well plates until confluent and then serum-starved overnight. Afterward, cells were incubated in K⁺-free Krebs solution [142.4 mM NaCl, 2.8 mM CaCl₂, 0.6 mM NaH2PO4, 1.2 mM MgSO4, 10 mM glucose, 15 mM Tris (pH 7.4)] for 15 min and then exposed to 200 nM [³H]ouabain for 30 min at 37°C. At the end of incubation, the cells were washed three times with ice-cold K⁺-free Krebs solution, solubilized in 0.1 M NaOH-0.2% SDS, and counted in a scintillation counter for [³H]ouabain. Nonspecific binding was measured in the presence of 1 mM unlabeled ouabain and subtracted from total binding. All counts were normalized to protein amount.

Caveolin-rich membrane fractionation analysis.

Caveolin-rich membrane fractions were obtained as previously described (34). Briefly, cells were washed and collected in 500 mM sodium carbonate (pH 11.0) solution and homogenized using a Polytron tissue grinder and then sonicated. Cell homogenates were adjusted to 45% sucrose by addition of 90% sucrose prepared in MBS (25 mM MES, 0.15 M NaCl, pH 6.5) and placed at the bottom of an ultracentrifuge tube. The ultracentrifuge tubes were then loaded with 4 ml of 35% sucrose and 4 ml of 5% sucrose (both in MBS containing 250 mM sodium carbonate) and centrifuged at 39,000 rpm for 16–20 h in an SW41 rotor (Beckman Instruments). Twelve gradient fractions of 1 ml were collected from the top to the bottom of the centrifuge tube. Among the 12 fractions, fractions 4 and 5 were combined and diluted with 4 ml of MBS and then centrifuged at 40,000 rpm with a Beckman type 65 rotor for 1 h. The pellets were resuspended in 250 μl of MBS and termed as caveolin-enriched fraction.

Statistical analysis.

Data are presented as means ± SE. ANOVA followed by post hoc analyses were used to compare differences across groups. Statistical significance was accepted at P < 0.05.

RESULTS

Generation and characterization of mutant α2-expressing cell lines LY-a2 and LY-b2.

To understand the structural basis for the observed differences between α1 and α2 in signal transduction, we used an ouabain-resistant rat-α2 expression vector as the template (15) and generated mutant α2-expressing stable cell lines by constructing the following mutations: A258Y (corresponding to Y260 in α1) in the CD2 sequence, and P414A, T417L, and K432Q in NaKtide sequence (corresponding to A416, L419, and Q434 in α1) (Fig. 1A). These mutations reconstituted two putative α1 Src-binding sites in α2. We then transfected PY-17 cells with the mutant α2 cDNA. PY-17 cells were derived from pig kidney LLC-PK1 cells, and the expression of endogenous α1 Na-K-ATPase was reduced ~90% by the stable expression of α1-specific siRNA (21). We have found that knock-in of rat α1 or other ouabain-resistant mutants into PY-17 cells could reduce the expression of endogenous pig α1 further, producing stable cell lines that express over 95% of exogenous Na-K-ATPase, which makes it possible to study the expressed mutant without significant interference from endogenous α1 Na-K-ATPase (18, 38). Ouabain selection of mutant α2 cDNA-transfected PY-17 cells resulted in numerous clones. After initial random screening of multiple clones, six clones with either barely detectable α1 expression or less than that in PY-17 cells were selected and expanded in the absence of ouabain for three generations. Western blot analyses revealed varying degrees of α1 and α2 expression in these clones (Fig. 1B). Two clones, named LY-a2 and LY-b2, were further expanded and analyzed because they barely expressed α1, as revealed by Western blot. The rat α1-rescued PY-17 cells (called AAC-19) and wild-type α2-rescued PY-17 cells (called LX-α2-4) were used as control. As depicted in Fig. 1C, LY-a2 and LY-b2 cells expressed similar amounts of α2 compared with LX-α2-4 cells. No α2 signal was detected in PY-17 or AAC-19 cells. When α1 expression was measured, the expression level of endogenous α1 was barely detectable and was much lower than that in the parent PY-17 cells (Fig. 1D).

Fig. 1.

Generation and characterization of mutant α2-expressing cell lines. A: comparison of human CD2 and NaKtide sequences in α1 and α2 Na-K-ATPase. Residues mutated in the present study are shown in bold. B: representative Western blots of α1 and α2 expression in different cell lines is shown; PY-17 and LX-α2-4 cells were used as controls. Mutant clones LY-a1–3 and LY-b1–3 are shown. LY-a2 and LY-b2 were selected for further study. C and D: cell lysates were prepared from different cell lines and then subjected to Western blot analysis. Tubulin was measured in the same membrane and used as a loading control. In D, half amount of AAC-19 cell lysates was loaded to see clear detection of α1 in PY-17 cells without overexposure of the film. A representative Western blot is shown and quantitative data shown as means ± SE (n = 5) are presented. **P < 0.01 vs. AAC19.

Pumping function of the mutant α2 Na-K-ATPase.

To assess the pumping function of mutant α2, we first checked the expression of β1 subunit. As shown in Fig. 2A, expression of mutant-α2 in LY-a2 produced the same degree of β1 expression and glycosylation as AAC-19 and LX-α2-4 cells, indicating that the expressed mutant α2 was able to assemble with the β1 subunit into a fully functional Na-K-ATPase in the established stable cell line similar to wild-type α1 and α2. However, although β1 expression and glycosylation in LY-b2 were much higher than in PY-17 cells, they were lower than those detected in AAC-19, LX-α2-4, and LY-a2. Therefore, we focused most of our functional analyses in LY-a2 cells.

Fig. 2.

Characterization of the α/β assembly, Na-K-ATPase activity and the expression of endogenous α1 in mutant α2 cells. A: cell lysates were prepared from different cell lines and then analyzed by Western blotting for expression of core and glycosylated Na-K-ATPase-β1, using tubulin as a loading control. Representative Western blot and quantification are shown (n = 5). *P < 0.05 vs. PY-17, **P < 0.01 vs. AAC-19. B: whole cell lysates were prepared and assayed for ouabain-sensitive ATPase activity as described under experimental procedures. Values are means ± SE (n = 4), and statistical analyses indicate no significant difference between AAC-19 cells and other cell lines. C: cells were cultured up to 100% confluence and then subjected to ouabain binding assays according to the protocol described in experimental procedures. Values are means ± SE (n = 3). **P < 0.01 vs. PY-17.

We next measured and compared the ouabain-sensitive ATPase activity in AAC-19, LX-α2-4 and LY-a2 cells to assess the ion-pumping function of the mutant-α2 Na-K-ATPase. As shown in Fig. 2B, Na-K-ATPase activity was similar among those three cell lines. To further assess the level of endogenous pig α1 expression in the mutant-rescued cells, we conducted [³H]ouabain binding analyses (33). Because ouabain binds to the endogenous pig α1, but not to the expressed mutant, this binding assay allows us to assess the surface expression of endogenous α1 in the presence of highly expressed ouabain-resistant α2 in the rescued cells. The parental PY-17 cells were used as a control. As depicted in Fig. 2C, expression of endogenous α1 in either α2 or mutant α2 rescued cells was further reduced by more than 60% when compared with PY-17 cells. Because the expression level of α1 Na-K-ATPase in PY-17 cells is already too low to allow ouabain-activated signal transduction (21), we consider it reasonable to assume that the level of α1 Na-K-ATPase remaining in LY-a2 cells would not be sufficient to mediate ouabain-induced signal transduction. Moreover, LY-a2 cells have ouabain-sensitive ATPase activity comparable to that of control AAC-19 cells. Therefore, we decided to use this cell line in the following experiments to characterize the signaling functions of mutant α2 Na-K-ATPase. Rat α1-rescued AAC-19 cells were used as a positive control and the wild-type α2-rescued LX-α2-4 as a negative control for comparison. We also occasionally used LY-b2 cells as a control to verify a number of findings obtained in LY-a2 cells.

Expression of mutant α2 restores caveolin-1 expression and redistributes Src and Na-K-ATPase into caveolae fraction.

Caveolin-1 is a scaffolding protein and the main component of caveolae, integral in ouabain-induced signal transduction (9, 25, 34). We have shown that knockdown of α1 expression increased the endocytosis and degradation of caveolin-1, resulting in a significant decrease in total cellular caveolin-1 in PY-17 cells (7). As reported, this defect could be rescued by the expression of rat α1 in AAC-19 cells (8), but not by rat α2 in LX-α2-4 cells (37), which is consistent with the fact that α2 lacks α1-like signaling function. When caveolin-1 content was measured in LY-a2 cells, we found that expression of mutant α2 in LY-a2 was sufficient to restore caveolin-1 expression (Fig. 3A).

Fig. 3.

Expression of mutant α2 restores caveolin-1 expression and redistributes Src and Na-K-ATPase into caveolae fraction. A: cell lysates from different cell lines were separated by SDS-PAGE and analyzed by Western blotting for caveolin-1. Representative Western blot is shown, and quantitative data are means ± SE (n = 5). B: cell lysates from AAC-19, LX-α2-4, and LY-a2 were collected by sucrose density gradient fractionation as described in experimental procedures. Fractions 4/5-11 of equal volume were subjected to SDS-PAGE. Representative Western blot of caveolin-1 in different fractions is shown. C and D: α1 and α2 Na-K-ATPase and Src in different fractions are shown as described in B. Percentage of Na-K-ATPase and Src in caveolin-enriched fractions (4/5) was measured. Values are means ± SE (n = 3 for B–D). *P < 0.05; **P < 0.01 vs. AAC-19; #P < 0.05 vs. LX-α2-4.

To further assess the α1-like effect of the α2 mutant, we used a well-established detergent-free and carbonate-based density gradient fractionation procedure (34) and prepared caveolin-enriched fractions from LY-a2, AAC-19, and LX-α2-4 cells. Western blot analyses of the fractions revealed that the distribution of caveolin-1 was maximal in light-density fractions 4/5, as expected when this fractionation procedure is used, and similar in the three cell lines (Fig. 3B). However, both α2 and Src were distributed out of caveolin-enriched fractions into high-density fractions in LX-α2-4 cells. In contrast, the expression of mutant α2 returned Src to the caveolin-1-enriched fractions in LY-a2 cells. This pattern of distribution due to the expression of mutant α2 was similar to that in AAC-19 cells (Fig. 3, C and D). These data again indicated an intact α1-like signaling property for the α2 mutant.

Effects of mutant α2 Na-K-ATPase expression on basal Src and ERK activities.

We (20, 32) have previously reported that the α1 Na-K-ATPase interacts with Src and keeps it in an inactive state, thereby modulating Src effectors such as ERKs. On the other hand, α2 Na-K-ATPase lacks Src-regulatory function; consequently, LX-α2-4 cells exhibit an increase in basal Src/ERK activity due to the absence to the tonic downregulatory effect (37). To assess whether the mutant α2 Na-K-ATPase is able to interact with and downregulate basal Src/ERKs like α1, we compared total Src and active Src in AAC-19, LX-α2-4, and LY-a2 cells. As shown in Fig. 4A, there was no difference in total Src expression among those three cell lines. However, Src kinase activity in LY-a2 cells was significantly lower than that in LX-α2-4 cells as measured by Western blot analysis of the phosphorylation of Y418 (pY418; Fig. 4A), although it was still higher than that in AAC-19 cells. We also investigated ERK1/2, a downstream target in the signaling cascade. As depicted in Fig. 4B, ERK1/2 activity in LY-a2 cells also decreased compared with LX-α2-4 cells but was still higher than in AAC-19 cells. This is in concordance with the observed difference in Src activity in these cell lines. To further verify that expression of α2 mutant is sufficient to reduce basal ERK1/2 activity, we repeated the experiments in LY-b2 cells. As depicted in Fig. 4C, ERK1/2 activity was also significantly reduced in LY-b2 cells compared with LX-α2-4 cells.

Fig. 4.

Regulation of basal kinase activity by mutant- α2. A: cell lysates were prepared from different cell lines and then subjected to Western blot analysis for phospho- (p)Y418 Src and total cellular Src(c-Src). B and C: total cell lysates were analyzed for total ERK1/2 and pERK1/2 as described in A. Representative Western blot is shown, and quantitative data are means ± SE (n = 6 for A and B; n = 3 for C). **P < 0.01 vs. LX-α2-4 cells.

Effects of mutant α2 expression on ouabain-induced signal transduction.

To assess whether mutant α2 Na-K-ATPase is capable of functioning as a receptor for cardiotonic steroids to regulate protein kinases like α1, we treated LY-a2 cells with ouabain. As depicted in Fig. 5, ouabain stimulated Src and ERKs in LY-a2 cells in a time- and dose-dependent manner. To further verify that the expression of mutant α2 is sufficient for ouabain to activate ERK, we exposed LY-b2 cells to ouabain. As shown in Fig. 5C, ouabain also activated ERK in a dose-dependent manner in LY-b2 cells. To probe whether ouabain-induced activation of ERKs is due to the activation of receptor Na-K-ATPase in a Src-dependent manner, we pretreated cells with pNaKtide or Src inhibitor PP2. We have shown that pNaKtide is a specific antagonist of Na-K-ATPase-mediated signal transduction (20). These experiments demonstrate that both pNaKtide and PP2 were able to completely block ouabain-induced activation of ERKs in LY-a2 cells (Fig. 6).

Fig. 5.

Effects of ouabain on mutant α2 cells. A: LY-a2 cells were treated with different concentrations of ouabain for 10 min, and cell lysates were subjected to Western blot analysis of Src pY418 and pERK1/2. B: LY-a2 cells were exposed to 100 μM ouabain for different times, and cell lysates were subjected to Western blot analysis of Src pY418 and pERK1/2. C: LY-b2 cells were treated with different concentrations of ouabain for 10 min, and cell lysates were subjected to Western blot analysis of Src pY418 and pERK1/2. Representative Western blot is shown, and quantitative data are means ± SE (n = 5). *P < 0.05, **P < 0.01 vs. control.

Fig. 6.

PP2 and pNaKtide abolish ouabain-induced activation of ERK1/2 in LY-a2 cells. LY-a2 cells were pretreated with 1 μM pNaKtide for 1 h or 1 μM PP2 for 30 min and then treated with 100 μM ouabain for the indicated durations. Cell lysates were collected and subjected to Western blot analysis of pERK1/2. Representative Western blot is shown, and quantitative data are means ± SE (n = 3). *P < 0.05, **P < 0.01 vs. respective control condition.

Expression of mutant-α2 increases the formation of Na-K-ATPase/Src complex.

We have previously reported that anti-Src antibody could coprecipitate α1 Na-K-ATPase from AAC-19 cells (32). However, only one-third of that amount of α2 was coprecipitated with Src from LX-α2-4 cells under the same experimental conditions (37). To reaffirm that the mutant α2 has α1-like capacity of Src interaction, we immunoprecipitated Src from LY-a2 and LX-α2-4 cells and then probed for α2. As depicted in Fig. 7, anti-Src antibody coprecipitated significantly more α2 from LY-a2 cells than from LX-α2-4 cells.

Fig. 7.

Mutant-α2 coimmunoprecipitates with Src and could interact with Src to form a functional signaling complex. Total cell lysates were immunoprecipitated with an anti-Src antibody as described in experimental procedures and analyzed by Western blotting for α2 Na-K-ATPase and c-Src. Representative Western blot is shown, and the quantitative data are means ± SE (n = 4). **P < 0.01.

Expression of mutant α2 Na-K-ATPase restores cell growth.

We reported that cell proliferation was significantly reduced in LX-α2-4 cells compared with that in AAC-19 (37). Given that α2 mutant exhibits α1-like signaling properties, we examined cell proliferation in LX-α2-4, LY-a2, and AAC-19 cells. As depicted in Fig. 8, LY-a2 grew at comparable rate as in AAC-19 cells, which is much faster than that of LX-α2-4 cells.

Fig. 8.

Effects of mutant-α2 expression on cell growth. LY-a2, AAC-19, and LX-α2-4 cells were seeded at a density of 50,000 cells/well in 12-well plates, cultured, and counted at the indicated times. Values are means ± SE (n = 4). **P < 0.01 vs. LX-α2-4 cells at the same time point.

DISCUSSION

We report here the generation of a mammalian cell line (LY-a2) that expresses a gain of function mutant α2 Na-K-ATPase in which sequences important for Src binding were mutated to corresponding sequences of the α1 polypeptide. Functional studies reveal that LY-a2 cells have a total Na-K-ATPase activity similar to that in α2-expressing LX-α2-4 cells. However, LY-a2 cells exhibited α1-like, but not α2-like, signaling phenotypes and growth, providing further support for the importance of Src interaction in Na-K-ATPase-mediated signal transduction and revealing the structural basis of isoform-specific functions of Na-K-ATPase.

Role of Na-K-ATPase and Src interaction in signal transduction.

We and others have demonstrated an important role of Src in Na-K-ATPase-mediated signal transduction. To further test the importance of two putative Src binding sites in Na-K-ATPase-mediated Src regulation and signal transduction, we mutated the above four amino acids in α2 to the corresponding α1 sequence. As depicted in Fig. 2, the mutant exhibited pumping function similar to those of wild-type α2, and was fully capable of rescuing the expression of the β-subunit. Most importantly, expression of mutant α2 led to an α1-like phenotype of Src regulation and signal transduction in LY-a2 cells. These changes include an increase in caveolin-1 expression and a decrease in basal Src and ERK activity. Moreover, when cell lysates were fractionated on density gradients, we found that Src and mutant α2 Na-K-ATPase were redistributed into the low-density caveolin-1-enriched fraction as in α1 cells. In contrast, wild-type α2 and Src were distributed toward higher-density fractions in LX-α2-4 cells. Accordingly, ouabain activated both Src and ERK in a dose- and time-dependent manner in α2 mutant cells. Finally, when cell proliferation was measured, unlike α2 cells, expression of α2 mutant made these cells grow as fast as the α1 cells, again pointing out the importance of Na-K-ATPase-mediated Src regulation in control of cell growth as we previously reported (33). Clearly, the mutant α2 gained α1-like function in Src regulation and signal transduction. Taken together, the loss-of-function mutations in α1 (18) and gain-of-function mutations in α2 reported here provide strong evidence of direct interaction between the α1 Na-K-ATPase and Src being important for Src regulation and ouabain-activated signal transduction.

The potential importance of the conserved α2-specific sequences.

Recent studies have revealed major differences in the physiological functions of each Na-K-ATPase isoform. For example, the α2 isoform is known to play an important role in the regulation of intracellular Ca²⁺ levels in myocytes, as it coresides with the Na/Ca exchanger (5, 19). This is consistent with the studies of SWAP mice (ouabain-sensitive α1 Na-K-ATPase mutant and ouabain-resistant α2 Na-K-ATPase mutant) (12). We and others have shown that α1 Na-K-ATPase represents an important signaling mechanism in many different types of cells including cardiac myocytes. In the heart, activation α1 Na-K-ATPase by cardiotonic steroids is capable of protecting myocytes from ischemic reperfusion injury (11, 28–30), whereas chronic stimulation of this signaling mechanism by either endogenous or infused cardiotonic steroids induces cardiac hypertrophy and fibrosis in vivo by increasing the generation of reactive oxygen species (13, 16, 17, 24, 27). We (26) have further demonstrated the importance of this signaling mechanism in the development and progression of experimental uremic cardiomyopathy, and the effectiveness of blocking this signaling by pNaKtide in attenuating and reversing the pathological changes in the same animal models. Similarly, other laboratories have noted that replacement of endogenous ouabain-resistant α1 with an ouabain-sensitive α1 mutant increased both cardiac hypertrophy and fibrosis in pressure overload models (35). Interestingly, transgenic overexpression of α2 Na-K-ATPase reduced cardiac hypertrophy (10). Retrospectively, these in vivo findings are consistent with our data, indicating that α1 is most likely responsible for cardiotonic steroid-induced remodeling of the heart, whereas α2, with its role predominantly in ion pumping, may protect the heart from pressure overload-induced hypertrophy. To this end, it is of great interest to note the following two novel findings reported here. First, α1 and α2 appear to reside in different cellular compartments or membrane microdomains, as shown in Fig. 3. Interestingly, gain-of-Src binding mutations resulted in a redistribution of α2 Na-K-ATPase to compartments where α1 resides and allows a close contact with signaling partners of α1 Na-K-ATPase, namely Src and caveolin-1. Therefore, it is plausible that gain-of-Src interacting function could change the cellular localization of α2, making it less able to exert its physiological function (e.g., Ca²⁺ regulation). Second, the fact that sequences surrounding the putative Src binding sites in the α2 are highly conserved across different species of animals (Fig. 9) also suggests that lack of Src binding in α2 is of critical importance for α2-specific functions. These speculations need to be experimentally tested in the future; however, the reported gain-of-function mutant α2 would make it possible to directly test this hypothesis in vivo. In short, we reaffirm that α1 Na-K-ATPase differs from α2 in its ability to interact with and regulate Src. This difference is due, at least in part, to the conserved Src binding sites in the α1 and the conserved lack of Src binding sites in the α2 isoform of Na-K-ATPase.

Fig. 9.

Sequences of α2 CD2 and NaKtide in mammals. Sequence comparisons reveal that the observed sequence differences between human α2 and α1 are conserved in all mammalian α2 subunits. Residues corresponding to human α1 NaKtide sequence are underlined. Residues mutated in the present study are shown in bold.

GRANTS

This work is supported by National Heart, Lung, and Blood Institute Grant HL-109015.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

H.Y., X.C., J.Z., J.X.X., and M.B. performed experiments; H.Y., X.C., J.Z., J.X.X., M.B., and Z.X. analyzed data; H.Y., X.C., J.Z., J.X.X., M.B., S.V.P., and Z.X. interpreted results of experiments; H.Y., X.C., J.Z., and M.B. prepared figures; H.Y. and S.V.P. drafted manuscript; X.C., J.Z., J.X.X., M.B., S.V.P., and Z.X. edited and revised manuscript; X.C., J.Z., J.X.X., M.B., S.V.P., and Z.X. approved final version of manuscript; Z.X. conceived and designed research.