|Home | About | Journals | Submit | Contact Us | Français|
The level of the heterodimeric Na,K-ATPase is tightly controlled in epithelia to maintain appropriate transport function. The catalytic Na,K-ATPase α subunit is not able to exit the ER or catalyze ion transport unless assembled with the β subunit. However, requirements for the ER exit of the Na,K-ATPase β subunit that plays an additional, ion-transport-independent, role in intercellular adhesion are not clear. Exogenous β1 or β2 subunits expressed in renal MDCK cells replace endogenous β1 subunits in the α–β complexes in the ER, resulting in a decrease in the amount of the α1-bound endogenous β1 subunits by 47–61% with no change in the amount of α1 subunits. Disruption of the α1–β association by mutations in defined α1-interacting regions of either β1 or β2 subunits results in the ER retention and rapid degradation of unassembled mutants. Hence, the ER quality control system allows export only of assembled α–β complexes to the Golgi, thereby maintaining an equimolar ratio of α and β subunits in the plasma membrane, whereas the number of α1 subunits in the ER determines the amount of the α–β complexes.
The sodium pump is expressed in all animal tissues where it establishes concentration gradients for Na+ and K+ by pumping three Na+ from the cytoplasm in exchange for two extracellular K+. The ion gradients generated by the Na,K-ATPase are used in many essential cellular processes, such as osmoregulation, generation of plasma membrane potential and maintenance of intracellular pH and Ca2+ concentration, vectorial transport of many solutes, and excitability in muscle fibers and neurons. Further, the Na,K-ATPase serves as a functional signal transducer (1, 2).
The pump is a heterodimer, consisting of an α subunit that is responsible for ion transport and a β subunit that is implicated in maturation and membrane targeting of the enzyme. There are four isoforms of the Na,K-ATPase α subunit (α1, α2, α3, and α4) and three isoforms of the Na,K-ATPase β subunit (β1, β2, and β3) (3, 4). Additional tissue-specific regulatory subunits of the Na,K-ATPase that belong to the FXYD family of small membrane proteins have also been identified (5). Most data have shown that α and β subunits are always present in equimolar amounts in the isolated Na,K-ATPase (6–9). Recent data claiming that the α–β ratio can be less (10) or more (11) than one are difficult to reconcile with the 3D high-resolution structure of the enzyme that shows two specific regions of interaction between the α and β subunits at a 1:1 stoichiometry (12, 13). Given these sites of interaction, stable recruitment of additional β subunits to the α subunit or vice versa is unlikely.
ER assembly of the α with the β subunit is necessary for export of the α subunit of the Na,K-ATPase from the ER to the Golgi and formation of functional enzyme (14), suggesting that the β subunit acts as a molecular chaperone that facilitates maturation of the enzyme (15, 16).
In addition, both β1 and β2 subunits have important non-enzymatic roles such as formation and maintenance of intercellular junctions and regulation of cell migration (16–23). It is not known whether the β subunits perform these functions only as components of the α–β complexes or perhaps as individual plasma membrane proteins.
Results of studies addressing the question as to whether unassembled β subunits are present in the plasma membrane are ambiguous. A consecutive series of immunoprecipitation using the antibody against the α1 subunit resulted in depletion of MDCK cell lysates of the mature but not immature β1 subunits, indicating that the mature β1 subunits are all bound to the α1 subunit (10). Along with these data, overexpressed Xenopus oocyte Na,K-ATPase β1 and β3 subunits are retained in the ER in the absence of the α subunit in Xenopus oocytes (24–26). By contrast, the Torpedo californica β1 subunit expressed in Xenopus oocytes without the α subunit shows the mature N-glycosylation pattern, implying that it is able to traffic from the ER to the Golgi (27). Whether the unassembled Na, K-ATPase β2 subunit isoform is able to reach the plasma membrane is unknown. However, the H,K-ATPase β subunit that has more homology with the Na,K-ATPase β2 subunit than the other Na,K-ATPase β subunit isoforms is detected in the plasma membrane when expressed without the H,K-ATPase α subunit in various cell types (28–32).
In this work, analysis of the nature of the N-linked glycans of endogenous and expressed β subunits of the Na,K-ATPase enables determination of their localization either in the ER or in the post-ER compartments. The presence of high-mannose-type N-glycans shows ER location of the β subunits, while the presence of hybrid- or complex-type N-glycans in the β subunits that can only be generated by Golgi-located glycosyltransferases shows successful export of the β subunits from the ER. Renal MDCK cells were used as an expression system for YFP-linked β1 and β2 subunit isoforms and β subunit mutants in order to determine whether assembly with the α subunit is essential for ER exit of either β1 or β2 subunits. Confocal microscopy, immunohistochemistry, immunoprecipitation, surface-selective biotinylation, and Western blot analysis were employed to evaluate trafficking and molecular interactions of endogenous and exogenous Na,K-ATPase subunits. The results show that (i) association with the α subunit at 1:1 stoichiometry is a prerequisite for the export of both the β1 and β2 subunits from the ER and (ii) the amount of the α subunit in the ER is a limiting factor for α–β assembly.
The fusion proteins with YFP linked to the N-terminus of the Na,K-ATPase rat β1 or human β2 subunit (YFP-β11 and YFP-β2) were constructed as described previously (23). Y39A/F42A/Y43A (YFY/AAA) and P245G (P/G) mutants of YFP-β1 and homologous Y44A/F47A/Y48A (YFY/AAA) and P229G (P/G) mutants of YFP-β2 were constructed by site-directed mutagenesis using the QuikChange mutagenesis kit (Stratagene). The homologous residues in the rat β1 or human β2 subunits were determined based on the computational alignment (Supporting Information Figure 1). The mutagenic primers (Supporting Information Table 1) were constructed using PrimerSelect software, 5.03. Stable MDCK cell lines expressing wild-type and mutated YFP-β1 and YFP-β2 were obtained as described previously (33).
The fusion protein with YFP linked to the N-terminus of the dog Na,K-ATPase β1 subunit was constructed by cloning a cDNA encoding dog Na,K-ATPase β1 subunit into a multi-cloning site of pEYFP-C1 (Clontech) using KpnI and BamHI restriction sites. Then, the stop codon was removed in the linker between YFP and Na,K-ATPase β1 subunit by site-directed mutagenesis. A cDNA encoding dog Na,K-ATPase β1 subunit was obtained from pCIN4 vector, a kind gift from Liora Shoshani.
Cells were grown in DMEM medium (Cellgro Mediatech) containing 4.5 g/L glucose, 2 mM l-glutamine, 8 mg/L phenol red, 100 units/mL penicillin, 0.1 mg/mL streptomycin, and 10% FBS. Where indicated, 2 µg/µL swainsonine (Sigma), 100 µg/mL deoxymannojirimycin (Sigma), or 20 µg/mL cycloheximide (Sigma) was added to the cell culture medium.
Confocal microscopy images were acquired using the Zeiss LSM 510 laser scanning confocal microscope and LSM 510 software, version 3.2.
The following monoclonal antibodies were used for immunoprecipitation, immunostaining, and Western blot analysis: against the Na,K-ATPase α1 subunit, clone C464.6 (Millipore); against GFP, clones 7.1 and 13.1, that also recognize YFP (Roche Diagnostics). Also, polyclonal antibodies against the Na,K-ATPase α1 subunit (Cell Signaling) and against GFP that also recognize YFP (Clontech) were used.
Cells were fixed by incubation with 3.75% formaldehyde in PBS for 15 min at 37 °C and permeabilized by incubation with 0.1 % Triton X-100 for 5 min. Then cells were incubated with Dako protein block serum-free solution (Dako Corp.) for 30 min. Immunostaining of the Na,K-ATPase α1 subunit was performed by 1 h incubation with the monoclonal antibody against the Na,K-ATPase α1 subunit followed by 1 h incubation with Alexa Fluor 633 conjugated anti-mouse antibody (Invitrogen).
Cells grown on glass-bottom micro-well dishes (MatTek Corp.) were transfected with the plasmid encoding a fusion protein between Discosoma sp. red fluorescent protein and the marker of the ER, DsRed2-ER (Clontech), using the Lipofectamine 2000 transfection reagent (Invitrogen) according the manufacturer’s instructions. Confocal microscopy images of transfected cells were acquired 24–48 h after transfection.
Monolayers of MDCK cells grown in a 35 mm2 well of a six-well plate were rinsed twice with ice-cold PBS and lysed by incubation with 200 µL of 150 mM NaCl in 50 mM Tris, pH 7.5, containing 1% Nonidet P40, 0.5% sodium deoxycholate, and Complete protease inhibitor cocktail, 1 tablet/50 mL (Roche Diagnostics). Cell extracts were clarified by centrifugation (15000g, 10 min) at 4 °C. Then, the cell extracts were incubated with 30 µL of the protein A–agarose suspension (Roche Diagnostics) in a total volume 1 mL of the lysis buffer at 4 °C with continuous rotation for at least 3 h (or overnight) to remove the components that nonspecifically bind to protein A. Precleared supernatant was mixed with 3 µL of the polyclonal anti-YFP antibody or 20 µL of the polyclonal antibody against the Na,K-ATPase α1 subunit and incubated with continuous rotation at 4 °C for 60 min. After addition of 30 µL of the protein A–agarose suspension, the mixture was incubated at 4 °C with continuous rotation overnight. The bead-adherent complexes were washed three times on the beads, and then proteins were eluted from the beads by incubation in 35 µL of SDS–PAGE sample buffer (4% SDS, 0.05% bromophenol blue, 20% glycerol, 1% β-mercaptoethanol in 0.1 M Tris, pH 6.8) for 5 min at 80 °C.
Cells were maintained for 6 days after becoming confluent in transwell inserts. Biotinylation of surface proteins was performed according to previously described procedures (34,35). Cell monolayers were biotinylated with EZ-Link sulfo-NHS-SS-biotin (Pierce) that was added into the well only (basolateral surface of the tight cell monolayers). After quenching the biotinylation reaction, cells were washed and then lysed by incubation with 200 µL of 0.15 M NaCl in 15 mM Tris, pH 8.0, with 1% Triton X-100 and 4 mM EGTA. Cell extracts were clarified by centrifugation (15000g, 10 min) at 4 °C. To isolate surface biotinylated proteins, the cell extract was incubated with 100 µL of streptavidin–agarose beads (Sigma-Aldrich) in a total volume of 1 mL of 0.15 M NaCl in 15 mM Tris, pH 8.0, with 0.5% Triton X-100 and 4 mM EGTA at 4 °C with continuous rotation for 60 min. The bead-adherent complexes were washed three times on the beads, and then proteins were eluted from the beads by incubation in 40 µL of SDS–PAGE sample buffer for 5 min at 80 °C.
Samples containing 7–8 µg of protein in 10 µL of the MDCK cell extract mixed with 10 µL of SDS–PAGE sample buffer, or immunoprecipitated proteins eluted from the protein A/G-conjugated agarose beads, or basolateral biotinylated proteins eluted from streptavidin–agarose beads were loaded onto 4–12% gradient SDS– PAGE gels (Invitrogen). Proteins were separated by SDS– PAGE using MES/SDS running buffer (0.05 M MES, 0.05 M Tris base, 0.1% SDS, and 1 mM EDTA, pH 7.3), transferred onto a nitrocellulose membrane (Bio-Rad), and detected by Western blot analysis using the appropriate primary antibody and the anti-mouse or anti-rabbit IgG conjugated to alkaline phosphatase (Promega) as a secondary antibody. Alkaline phosphatase was detected using nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate in alkaline phosphatase buffer (150 mM NaCl and 1 mM MgCl2 in 10 mM Tris-HCl, pH 9.0). Immunoblots were quantified by densitometry using Zeiss LSM 510 software, version 3.2.
Total RNA from MDCK cells was isolated using RNA-queous (Ambion). Typical RNA concentrations were 100–200 ng/µL. The RNA was measured for purity and stability with a Bioanalyser 2100 (Agilent Technologies). Total RNA (2000 ng) in total 40 µL volume was converted to cDNA by use of the Omniscript RT kit (Qiagen) and an oligo(dT)12–18 primer (Invitrogen), according to the manufacturer’s protocol. One microliter of RT product was amplified by real-time PCR using the SYBR premix ExTaq perfect real-time PCR kit (TaKaRa), according to the manufacturer’s protocol. The primers used are shown in Supporting Information Table 2. Real-time PCR was performed in eight-well strips using a DNA Engine Opticon 2 (MJ Research). The cycle of threshold (Ct) was determined for each primer set. The efficiency of amplification was determined by generating a standard curve for each primer pair using a corresponding PCR product as a template. The PCR products were purified by using the MinElute gel extraction kit (Qiagen). The range of PCR product concentrations was chosen for each primer set so that the standard curve included the values of Ct observed for the RNA samples isolated from both nontransfected and transfected cells. The resulting values of Ct were plotted against the logarithm of the PCR product copy number. Quantification of the mRNA levels was performed as described previously (22) using the β-actin mRNA as a reference. All reactions were carried out in duplicate, and three separate mRNA isolations were performed.
Susceptibility of the mutated and wild-type Na,K-ATPase β subunits to limited tryptic digestion was used to assess the effect of mutations on folding of the subunit. In order to prevent heterogeneity in N-glycan sizes, cells were treated with 100 µg/mL deoxymannojirimycin, α-mannosidase I inhibitor, for 48 h prior to the trypsin digest experiment. Deoxymannojirimycin prevents synthesis of both complex- and hybrid-type N-linked oligosaccharide chains; as a result, all newly synthesized N-glycans become high-mannose type. Incubation with the inhibitor for 48 h was found to be sufficient to replace the vast majority of the existing YFP-linked β subunits by newly synthesized high-mannose-type glycosylated forms of the subunits.
The monolayers of deoxymannojirimycin-treated cells were rinsed twice with ice-cold PBS and lysed by incubation with 200 µL of 150 mM NaCl in 50 mM Tris, pH 7.5, containing 1 % Nonidet P40 and 0.5% sodium deoxycholate. Protein concentration in cell lysates was determined using the BCA assay kit (Pierce) and adjusted to 2.3 mg/mL. Cell lysates were incubated with trypsin, which was added in the range of 0–3 µg/mL as indicated at 37 °C for 30 min. The reaction was stopped by the addition of 20 µg/mL soy trypsin inhibitor (Sigma-Aldrich) and incubation of samples on ice for 10 min. Then, samples containing 10 µL of reaction mixture combined with 10 µL of SDS–PAGE sample buffer were incubated at 80 °C for 5 min, subjected to 4–12% SDS–PAGE, and analyzed by Western blotting by using a monoclonal anti-YFP antibody.
Where indicated, the total cell lysates were treated with EndoH from Streptomyces plicatus (Glyco-Prozyme Inc.) according to the manufacturer’s instructions.
Statistical analysis was performed using Student’s t test (GraphPad Prism 4 software and Microsoft Excel). Statistical significance is specified in the figure legends.
In MDCK cells, the major endogenous Na,K-ATPase subunit isoforms are α1 and β1. The endogenous β1 subunit contains two N-glycosylated forms that are detected as a broad band at 45–55 kDa and a sharper band at ~40 kDa (Figure 1B, bands M and I, respectively). Only the top band is found in the basolateral plasma membrane (Figure 3A, central panel), showing that the lower band represents the intracellular immature fraction of the subunit and the higher band represents the mature plasma membrane form.
The mature form of the β1 subunit is resistant to EndoH digestion, while the immature form is digested by EndoH producing the deglycosylated form of the β1 subunit (Figure 1B, right panel). EndoH cleaves hybrid- or high-mannose-type N-glycans but not the complex-type N-glycans from the glycoproteins (Figure 1A). Therefore, the mature form of the β1 subunit contains only complex-type N-glycans, whereas N-glycans of the immature form are either hybrid or high-mannose type, or a mixture of the two types.
To further define the type of the N-glycans of the intracellular fraction of β1 subunit, MDCK cells were incubated with inhibitors of N-glycan processing such as swainsonine and deoxymannojirimycin that preserve the hybrid type or high-mannose type of N-glycans, respectively (Figure 1A). The electrophoretic mobility of the immature form of the β1 subunit in the control cells is similar to the mobility of the β1 subunit in deoxymannojirimycin-treated cells but greater than that in swainsoninetreated cells (Figure 1B, left and central panels), showing that N-glycans of the immature form are actually high-mannose but not hybrid type. Similarly, expressed YFP-β1 and YFP-β2 contain mature and immature forms. Experiments with deoxymannojirimycin showed that N-glycans of the immature forms of YFP-β1 and YFP-β2 are also of the high-mannose type (Supporting Information Figure 2).
High-mannose-type N-glycans are added to glycoproteins cotranslationally in the ER. Although processing of N-glycans starts immediately after their covalent linkage to the asparagines of N-glycosylation sites, all N-glycans preserve the high-mannose structure while they reside in the ER. Only after export of the β subunit to the Golgi are the high-mannose N-glycans transformed to hybrid- and complex-type glycans due to the action of Golgi-located mannosidases and glycosyltransferases (Figure 1A). Since the immature forms of the β subunits are high-mannose but not hybrid type, they must reside in the ER.
Therefore, the relative content of the high-mannose form of the β subunit in whole cell lysates quantitatively reflects the fraction of the β subunits present in the ER. Similarly, the relative content of the high-mannose form of the β subunit in the α1-coprecipitated fractions reflects the ER-located portion of the α1-assembled β subunits prior to their export to the Golgi.
Expression of YFP-linked rat β1 subunit decreases the amount of the complex-type endogenous β1 subunit in the whole cell lysate by 52%, without a significant change in the amount of the α1 subunit (Figure 2A,B). The amount of the complex-type endogenous β1 subunit is also decreased with expression of YFP-β2 and YFP-linked dog β1 subunit by 47% and 61%, respectively (Figure 2A,B), but not with expression of the P/G mutant of YFP-β2 that does not interact with the α1 subunit (see below, Figure 5).
Quantitative PCR shows no significant difference in the amount of mRNA of the endogenous α1 or β1 subunit in nontransfected and YFP-β1-transfected cells (Supporting Information Figure 3), indicating that the decrease of the amount of the endogenous β1 subunit in the YFP-β1-expressing cell line is not due to decreased transcription of the gene encoding the β1 subunit. Also, there is no change in the amount of the ER-located high-mannose endogenous β1 subunits in the whole cell lysate (Figure 2A,C), showing that translation of the β1 subunit is not inhibited by expression of exogenous β subunits. The total amount of the ER-located β subunits (endogenous and expressed) is increased in all transfected cells (Figure 2C), consistent with the increase in the total message for the β1 subunit (endogenous and expressed) due to transfection.
The amount of the α1 subunit is similar in nontransfected and transfected cells (Figure 2B). The decrease in the complex-type endogenous β1 subunit is compensated for by expression of the exogenous β subunits, so that the total amount of the complex-type β subunits (endogenous and expressed) is not significantly different from the amount of the complex-type endogenous β1 subunit in nontransfected cells (Figure 2B). Therefore, the ratio between the total amount of complex-type β subunits and the α1 subunit is not altered by expression of exogenous β subunits.
To determine whether the exogenous β subunits replace the endogenous β1 subunit in α–β heterodimers, we examined the amount of the α1-bound endogenous and exogenous β1 subunits in nontransfected and YFP-β1-expressing cells. Immunoprecipitation of the α1 subunit in nontransfected MDCK cells results in coprecipitation of the endogenous β1 subunit. In YFP-β1-expressing cells, immunoprecipitation of the α1 subunit coprecipitates both endogenous β1 subunits and the added YFP-β1 subunits (Figure 3A, right panel), indicating that the α1 subunit associates not only with the endogenous β1 subunit but also with the expressed YFP-β1 subunit.
The α1-coprecipitated fractions of the β1 subunit and YFP-β1 contain both complex-type and high-mannose forms (Figure 3A, right panel). The amount of the complex-type endogenous β1 subunits bound to α1 subunits is decreased by 48% in YFP-β1-expressing cells as compared to nontransfected cells (Figure 3A, right panel, and Figure 3B). A similar decrease in the ratio between the complex-type endogenous β1 subunits and the α1 subunits is observed in the whole cell lysate and in the basolateral membrane (Figure 3A,B), suggesting that all complex-type β1 subunits are bound to the α1 subunits in nontransfected and transfected cells.
The presence of high-mannose forms in α1-bound endogenous and exogenous β subunits (Figure 3A, right panel) reflects assembly of α–β complexes in the ER, where the β subunits have only high-mannose-type N-glycans. These high-mannose forms comprise 2–3% of the total amount of α1-bound β1 subunits (Figure 3A, right panel, and Figure 3C), that likely reflects the relative residence time of the α–β complexes in the ER.
The percentage of the high-mannose form of the endogenous β1 subunit is greater in whole cell lysate than in the α1-coprecipitated fractions both in nontransfected and in YFP-β1-expressing cells. Similarly, the percentage of high-mannose YFP-β1 is greater in the total cell lysate than in the α1-coprecipitated fractions (Figure 3A, right panel, and Figure 3C). These results indicate that some of the high-mannose ER-resident endogenous and exogenous β1 subunits are not assembled with α1 subunits.
A comparison of the percentages of the high-mannose forms in the α1-coprecipitated fractions and cell lysates shows that 24% of high-mannose β1 subunits are assembled with the α1 subunits in nontransfected cells and only 11% of the ER-resident endogenous β1 subunits are bound to the α1 subunits in YFP-β1-expressing cells (Figure 3C). Since the amount of the high-mannose β1 subunits is similar in transfected and nontransfected cells (Figure 2C), these data show that expression of YFP-β1 decreases the amount of α1-bound high-mannose endogenous β1 subunits in the ER about 2-fold. Consistent with these data, a similar decrease is observed in the amount of the α1-bound complex-type endogenous β1 subunits in YFP-β1-expressing cells as compared to nontransfected cells (Figure 3A, right panel, and Figure 3B).
Therefore, all complex-type, but not all high-mannose-type, endogenous β1 subunits are assembled with the α1 subunits in both nontransfected and YFP-β1-expressing cell lines. Expression of YFP-β1 decreases the amount of α1-bound endogenous β1 subunits approximately by 50% both in the ER and in the basolateral membrane.
To determine whether expression of YFP-β1 affects the degradation rate of the endogenous β1 subunits, we examined changes of the basolateral and cellular contents of the β1 subunit during incubation of cells in the presence of the inhibitor of protein synthesis, cycloheximide. The basolateral content of the endogenous β1 subunit decreases by approximately 15% after 8 h of cell incubation with cycloheximide (Supporting Information Figure 4). A similar decrease is observed for the basolateral α1 subunit and YFP-β1 (Supporting Information Figure 4). No significant differences in the rates of decrease of the β1 subunit are found in transfected and nontransfected cells (Supporting Information Figure 4), indicating that expression of YFP-β1 does not affect degradation of the plasma membrane β1 subunits.
Since all complex-type β1 subunits and only a minor fraction of high-mannose β1 subunits are α1-associated, the amount of complex-type and high-mannose forms of the β1 subunits was quantified in cell lysates during cell incubation with cycloheximide to determine whether the α1-bound and unassembled β1 subunits are degraded differently.
In both nontransfected and YFP-β1-expressing cell lines, the amount of the complex-type β1 subunit decreases only slightly after 5 h of incubation with the inhibitor (Figure 4A,B). In contrast, the amount of high-mannose β1 subunits rapidly decreases during incubation of nontransfected or transfected cells with cycloheximide. Its content is reduced by 50–60% after 1 h (Figure 4A,B). Similarly, the high-mannose YFP-β1 also is decreased much faster than complex-type YFP-β1 as seen in Figure 4A.
In nontransfected and YFP-β1-expressing cell lines, only a minor fraction of the ER-located high-mannose β1 subunits is assembled with the α1 subunits as found by precipitation by the anti-α1 antibody (Figures 3C and and4C).4C). These α1-bound high-mannose β1 subunits disappear after 1 h incubation with cycloheximide (Figure 4C), indicating that they are rapidly exported to the Golgi and processed to hybrid and complex type. In contrast, unassembled high-mannose β1 subunits are not able to traffic from the ER to Golgi. Hence, in the absence of protein synthesis, a decrease in the amount of these high-mannose β1 subunits reflects mainly the ER-associated degradation (ERAD) of unassembled β1 subunits. Therefore, the ER-retained unassembled β1 subunits are degraded much more rapidly than α1-associated complex-type β1 subunits.
To determine whether binding to the α subunit is necessary for the ER export of the β2 subunit of the Na,K-ATPase, several point mutations were introduced into known sites of α–β interaction in the β2 subunit. According to the crystal structure of the Na,K-ATPase at 2.4 Å resolution (2ZXE) (13), association between the α1 subunit and the β1 subunit is supported by interaction between both the extracellular and transmembrane domains of the two subunits (Figure 5A). P245 in the β1 subunit is adjacent to a stretch of amino acid residues 246YYGK that directly interact with the 7–8 extra-cytoplasmic loop of the α1 subunit (13). Y39, F42, and Y43 interact with the residues in the seventh transmembrane domain of the α1 subunit (13). Mutations of homologous residues were introduced into YFP-β2, producing P229G (P/G) and Y44A/F47A/Y48A (YFY/AAA) mutants of YFP-β2 (Supporting Information Figure 1).
The wild-type YFP-β2 is found mostly in the basolateral membranes of MDCK cells where it is colocalized with the endogenous α1 subunit (Figure 5C, panel YFP-β2 (WT)). Antibody against YFP coimmunoprecipitates the endogenous Na,K-ATPase α1 subunit with YFP-β2 (Figure 5B, lane WT). These results show that YFP-β2 is associated with the α1 subunit.
YFY/AAA and P/G mutations in YFP-β2 do not alter susceptibility of the subunits to trypsin digestion (Supporting Information Figure 5), indicating that the mutations do not affect overall conformation and folding of the subunit. The YFY/AAA and P/G mutations result in major and full retention of YFP-β2 inside the cells, respectively (Figure 5C). The mutant YFY/AAA shows incomplete colocalization with the endogenous α1 subunit (Figure 5C). The P/G mutant of YFP-β2 shows no colocalization with the α1 subunit (Figure 5C). No α1 subunits are detected in the ER (Figure 5C, middle panels). Immunostaining for specific markers of the intracellular compartments indicates that the mutants are partially or completely retained in the ER. Colocalization of the P/G mutant of YFP-β2 with the ER marker is shown as an example (Figure 5D). The extent of the ER retention detected by confocal microscopy (Figure 5C) correlates with the increase in the relative content of the ER-resident high-mannose form of YFP-β2 in the mutants (Figure 5B, top panel) and inversely correlates with the efficiency of coimmunoprecipitation of the α1 subunit with YFP-β2 (Figure 5B, bottom panel).
Homologous mutations in YFP-β1, Y39A/F42A/Y43A and P245G, also impair α–β association and increase the ER retention of unassembled mutants without affecting protein folding (not shown).
It is well established that the Na,K-ATPase α1 subunit is unable to exit the ER unless it is assembled with the β subunit in various expression systems (14). Equally, the results presented in this paper demonstrate that assembly with the α1 subunit is also required for the export of either β1 or β2 subunit from the ER in MDCK cells.
The dependence of the export of the β2 subunit from the ER to Golgi on the α1 subunit association is shown by the results of site-directed mutagenesis of the β2 subunit in the two regions of its specific interaction with the α1 subunit. Both the YFY/AAA mutation in the transmembrane domain and the P/G mutation in the extracellular domain of the β2 subunit partially or completely disrupt α–β association as determined by loss in both coimmunoprecipitation and microscopic colocalization of YFP-β2 and α1 subunit (Figure 5B,C) without affecting protein folding as assessed by susceptibility of the mutants to trypsin digestion (Supporting Information Figure 5). Proportionately to the degree of impairment of α–β association, unassembled mutants are partially or completely retained in the ER as detected by colocalization with an ER marker (Figure 5D) and by the quantitative increase in the relative content of the ER-resident high-mannose form of the β subunits (Figure 5B, top panel). Therefore, association with the α1 subunit is required for export of the β2 subunit from the ER.
Homologous mutations in YFP-β1 also impair α–β association and increase ER retention of unassembled mutants, confirming previously published data on inability of the unassembled β1 subunit to traffic from the ER to Golgi (10, 24–26). Additional evidence is presented here by comparing the effects of expression of the YFP-linked rat β1 subunit on the amount of the endogenous Na,K-ATPase subunits in different cellular fractions. Transfection decreases the ratio between the amount of the endogenous complex-type β1 subunit and amount of the α1 subunit to the same extent in the whole cell lysate, basolateral membrane, and α1-immunoprecipitated fractions (Figure 3A,B). These data demonstrate that all complex-type β1 subunits are assembled with the α1 subunit in both nontransfected and transfected cells. The alternative possibility that the unassembled β1 subunits are delivered to the plasma membrane but are rapidly endocytosed and degraded in transfected cells is ruled out by the experiments showing the same rate of degradation of the plasma membrane resident Na,K-ATPase subunits in both transfected and nontransfected cells (Supporting Information Figure 4).
As for the Na,K-ATPase β1 and β2 subunits, the Na,K-ATPase β3 subunit requires assembly with the α subunit to exit the ER (24–26). Therefore, the inability to exit the ER without α-association is a property common to all three Na,K-ATPase β subunit isoforms. In contrast, the homologous H,K-ATPase β subunit is able to reach the plasma membrane when expressed without the H,K-ATPase α subunit in various cell types (28–32). This ability to exit the ER, at least in part, can be explained by association of the H,K-ATPase β subunit with the endogenous Na,K-ATPase α subunit (32). However, the presence of the expressed H,K-ATPase β subunit and the absence of the endogenous Na,K-ATPase α subunit in the apical membranes of polarized MDCK and LLC-PK1 cells show that a fraction of the H,K-ATPase β subunits is able to exit the ER without the α subunits (30–32). The efficiency of the ER export of the H, K-ATPase β subunit is greatly increased by coexpression of the H, K-ATPase α subunit (36), suggesting that plasma membrane trafficking of the unassembled β subunit is due to a “leak” of the ER quality control system. One of the possible reasons for this difference of the H,K-ATPase β subunit from homologous Na, K-ATPase β subunits might be lack of interaction of the H, K-ATPase β subunit with BiP (26), the ER chaperone responsible for the ER retention of unassembled subunits of many oligomeric proteins, including the Na,K-ATPase subunits (26). It is also possible that α-unassembled H,K-ATPase and Na,K-ATPase β subunits differ from each other in binding affinity to the ER lectins facilitating the ER to Golgi trafficking.
Expression of rat YFP-β1 significantly decreases the amount of the endogenous β1 subunit but does not alter the amount of the α1 subunit (Figures 2 and and3)3) or mRNA level of either subunit (Supporting Information Figure 3) or degradation of the plasma membrane resident subunits (Supporting Information Figure 4).
Similar to our data, expression of flag-conjugated sheep β1 subunit in MDCK cells did not alter mRNA level of endogenous subunits or the amount of the α1 subunit but decreased the level of the endogenous β1 subunit (37). Based on these results, the authors concluded that translation of the endogenous β1 subunit is repressed by expression of the exogenous β1 subunit by an unknown mechanism (37).
However, separate detection and quantification of the complex-type and high-mannose forms of the endogenous β1 subunit as performed here favor a different interpretation of these results. We find that only the complex-type fraction of the endogenous β1 subunit is reduced due to expression of exogenous β subunits, but the amount of the high-mannose, ER-resident, form does not change (Figure 2), implying that translation of the endogenous β1 subunit is not repressed by exogenous β subunits. The presence of exogenous YFP-β1 decreases the amount of the α1-bound endogenous β1 subunits both in the ER (Figure 3C) and in the basolateral membrane (Figure 3B) by approximately 50%. The amount of the α1 subunit does not change, while the complex-type YFP-β1 compensates for the decrease in the complex-type endogenous β1 subunit (Figure 2B). These data show that expressed YFP-β molecules compete with the endogenous β1 subunits for a limited number of the α1 subunits in the ER and replace the endogenous subunits in about half of the newly formed α–β complexes. Only assembled α–β complexes are exported from the ER to Golgi, while the unassembled β subunits are retained in the ER and degraded through the ER-associated degradation (ERAD) pathway. The ER-retained unassembled β1 subunits are degraded much more rapidly than the α1-bound complex-type β1 subunits in both transfected and nontransfected cell lines (Figure 4), which can explain why a 50% decrease in the α1-bound endogenous β1 subunits is not accompanied by an equivalent increase in the amount of the ER-retained unassembled endogenous β1 subunits (Figure 2). Therefore, a decrease in the abundance of the endogenous β1 subunit by expression of exogenous β subunits, as shown here (Figures 2 and and3)3) and previously (37), is likely due to competition of the β subunits for association with the limited number of the α1 subunits and rapid degradation of the ER-retained unassembled β subunits. This explanation is more straightforward than translational repression (37). In agreement with our interpretation, expression of the P/G mutant of YFP-β2 that is not able to assemble with the endogenous Na,K-ATPase α1 subunit does not change the amount of the endogenous β1 subunits (Figure 2). Similarly, the β1/β2 chimera that is not associated with the α1 subunit does not affect the amount of the endogenous β1 subunits in MDCK cell subunits (37).
Only 24% of high-mannose β1 subunits are bound to α1 subunits (Figure 3C). The unassembled β1 subunits are degraded faster than α-bound subunits (Figure 4). These data imply that less than a quarter of the newly synthesized β1 subunits assemble with the α1 subunit. Consistently, pulse–chase experiments show that β1 subunits are synthesized in 5:1 excess over the α1 subunit in MDCK cells (38).The newly synthesized immature β1 subunits are rapidly degraded, and only after 60 min, which coincides in time with maturation of the β1 subunits, are the α1 and β1 subunits detected with a 1:1 stoichiometry (38). Therefore, the number of α1 subunits in the ER appears to be the limiting factor that determines the number of α1–β1 complexes in MDCK cells.
Expression of the wild-type YFP-β1 (rat or dog) or YFP-β2 results only in an insignificant change of both the α1 subunit and total amount of the mature Na,K-ATPase β subunits (endogenous and exogenous), so that the ratio between the total amount of complex-type β subunits and α1 subunits remains constant (Figure 2). These data indicate that, similar to the endogenous β subunits, the exogenous β subunits form equimolar complexes with the endogenous α1 subunits.
It is likely that the ER quality control that allows export only of assembled α–β complexes is a common mechanism that ensures a 1:1 ratio between α and β subunits in the plasma membranes of all animal cells. In most cells, rapid degradation of unassembled subunits via ERAD maintains this ratio close to one also in intracellular membranes. Indeed, the equimolar ratio between α and β subunits was found not only in purified Na, K-ATPase preparations but also in total microsome membrane fractions in many different cell types, including renal, liver, and brain cells (39–42). However, in certain cells, such as the Xenopus oocytes and rat cardiac myocytes, the α subunit is present in an excess in intracellular membrane fractions (43, 44). In Xenopus oocytes, the functioning α–β complexes are expressed in the plasma membrane, while a significant amount of unassembled α subunits is accumulated in the ER (43). In cardiac myocytes, the majority of the active pumps containing α and β subunits at 1:1 stoichiometry is detected in caveolae, and a 5-fold excess of the α subunits over β subunits is found in caveolin-free membrane fractions (44). The majority of caveolin-free fractions contain ER markers implying that this excess of the α subunit is mostly due to the ER retention of unassembled α subunits. In addition, it is possible that two subunits of the Na,K-ATPase might disassemble after internalization and undergo distinct endocytic and degradation pathways, as the authors suggested based on unequal ratios between the α and β subunits detected in early and late endosomal fractions (44).
Both the β1 and β2 subunits play important roles in intercellular adhesion and cell migration. The Na,K-ATPase β2 subunit was first described as an adhesion protein in glial cells (AMOG) that mediates neuron–astrocyte interaction and neuron migration (17, 18). Overexpression of the Na,K-ATPase β1 subunit in nonpolarized CHO cells facilitates formation of intercellular junctions (21). The β1 subunit of the Na,K-ATPase is necessary for formation and stabilization of adherens junctions in MDCK cells (19, 23). The Na,K-ATPase β1 subunit is also involved in regulation of cell motility (20). Both the tightness of intercellular junctions and cell motility depend on the extent of N-glycan branching of the Na,K-ATPase β1 subunit (22). Furthermore, formation of tight cell monolayers from relatively motile dispersed MDCK cells is accompanied by changes in the level of expression of specific glycosyltransferases that result in a gradual decrease in N-glycan branching (22), suggesting that intercellular adhesion and cell migration are regulated via glycosyltransferase-mediated remodeling of N-glycans of the Na,K-ATPase β subunit.
The evidence presented here on the absence of unassembled β subunits in the plasma membrane indicates that both β1 and β2 subunits can regulate cell adhesion and migration only as components of the α–β complexes. Therefore, the adhesive function of the β subunits cannot be modulated independently of ion transport activity by altering β subunit abundance on the cell surface. However, cells can regulate cell adhesion and/or migration by remodeling of the structure of N-glycans of the β subunits without affecting the quantity and, hence, activity of the pump.
In conclusion, the relative amounts of the Na,K-ATPase subunits in the plasma membrane of MDCK cells is tightly regulated. The ER quality control system allows export of only equimolar α–β complexes to the Golgi, thus ensuring that no unassembled subunits are present in the plasma membrane. The number of the pumps in the membrane is determined by the amount of the α subunit in the ER, the limiting factor for the formation of α–β complexes.
The authors are thankful to Dr. Liora Shoshani for providing the cDNA encoding the dog Na,K-ATPase β1 subunit.
†Supported by NIH Grants DK077149 and DK058333.
1Abbreviations: YFP-β1 and YFP-β2, the fusion proteins between the yellow fluorescent protein and the Na,K-ATPase β1 subunit and β2 subunit, respectively.
SUPPORTING INFORMATION AVAILABLE
Tables 1 and 2 and Figures 1–5 as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org.