HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 278, Issue 37, 34794-34803, September 12, 2003

This Article Abstract Full Text (PDF) All Versions of this Article: 278/37/34794    most recent M302899200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Laughery, M. D. Articles by Kaplan, J. H. Search for Related Content PubMed PubMed Citation Articles by Laughery, M. D. Articles by Kaplan, J. H. Social Bookmarking                      What's this?

Mutational Analysis of - Subunit Interactions in the Delivery of Na,K-ATPase Heterodimers to the Plasma Membrane^*

Melissa D. Laughery  , Matthew L. Todd  and Jack H. Kaplan   ¶

From the Department of Biochemistry and Molecular Biology, Oregon Health and Sciences University, Portland, Oregon 97239 and the Department of Biochemistry and Molecular Genetics, University of Illinois, Chicago, Illinois 60607

Received for publication, March 21, 2003 , and in revised form, May 28, 2003.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The -subunit of the Na,K-ATPase is required to deliver functional –heterodimers to the plasma membrane (PM) of baculovirus-infected insect cells. We have investigated the molecular determinants in the -subunit for the assembly and delivery processes. Trafficking of both subunits was analyzed by Western blots of fractionated membranes enriched in endoplasmic reticulum (ER), Golgi, and PM. Heterodimer assembly was evaluated by co-immunoprecipitation, and enzymatic activity was measured by ATPase assay. Elimination of enzymatic activity by D369A point mutation of the -subunit had no effect on the compartmental distribution of the Na,K-ATPase, demonstrating that enzymatic functioning is not a prerequisite for PM delivery. Replacement of all three N-glycosylation site asparagines with glutamines produced no effect on the delivery to the PM or the activity of the enzyme, but increased susceptibility to degradation was observed. Analysis of -subunits in which the disulfide bonds were removed through substitution reveals that the bridges are important for PM targeting but not for assembly of the heterodimer. Assembly is supported by -subunits with greatly truncated extracellular domains. The presence of the amino-terminal domain and transmembrane segment is sufficient for assembly and PM delivery. Intermediate length truncated -subunits and some disulfide bridge substitution mutants assemble with the -subunit but are not able to exit the ER. We conclude that there are different and separable requirements for the assembly of Na,K-ATPase heterodimer complexes, exit of the dimer from the ER, delivery to the PM, and catalytic activity of the dimer.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The Na,K-ATPase is an ion-transporting plasma membrane (PM)¹ protein that exchanges three intracellular Na⁺ ions for two extracellular K⁺ ions in an ATP-dependent fashion. The Na,K-ATPase belongs to the P-type ATPase family of membrane enzymes, which are characterized by their ability to harness the energy of ATP hydrolysis to pump ions across the membrane against electrochemical gradients through the formation of a phosphoenzyme intermediate. The Na,K-ATPase and closely related H,K-ATPase are the only members of this family composed of two subunits, and (1). The -subunit is the P-type defining subunit of 1016 amino acid residues containing an ATP hydrolysis domain and 10 transmembrane segments capable of ion occlusion (for a review, see Ref. 2). Although the -subunit is essential for pump activity (3–6), a primary role currently attributed to the -subunit is that of a molecular chaperone to aid in the correct membrane insertion, stability, and trafficking of the -subunit to the PM (for a review, see Ref. 7).

The ₁-subunit is a glycoprotein of 303 amino acid residues consisting of a cytoplasmic amino-terminal domain of about 40 residues, a single membrane-spanning segment, and a larger extracellular carboxyl-terminal domain comprising about 240 amino acid residues. The extracellular domain has three N-linked glycosylation sites and three disulfide bonds that are conserved among all of the -subunit isoforms (Fig. 1). Several studies have investigated the role of the structural components of the -subunit in an attempt to understand how the -subunit specifically associates with the -subunit and contributes to the functioning of the Na,K-ATPase. To date, all of the major structural components of the -subunit have been implicated as being important for the expression of functional Na,KATPase heterodimer, but it is often unclear if disruption of functional pump expression is due to the inability of the -subunit to assemble with , target the heterodimer to the PM, or support Na,K-ATPase activity. The requirement for the - and -subunits to heterodimerize as a prerequisite for the targeting of the pump to the PM has been clearly established. However, the elements of the -subunit responsible for assembly and those responsible for trafficking have not been clearly identified or separated, although these processes have been attributed to the extracellular domain of the -subunit. The inability to observe functional pump at the cell surface in the oocyte and mammalian cell line expression systems has subsequently been interpreted as being due to either 1) the lack or disruption of a PM targeting signal or 2) an inability for the heterodimer to assemble. However, it is not possible to distinguish between the assembly and trafficking processes in these systems. Separation of the assembly and delivery processes is essential for understanding the roles of specific structural elements of the -subunit throughout the maturation process of functional Na,K-ATPase.

View larger version (51K): [in this window] [in a new window]   FIG. 1. Schematic diagram of wild-type and mutant -subunits. The -subunit is composed of a small amino-terminal domain, a single transmembrane segment, and a larger extracellular domain carboxyl-terminal domain. The extracellular domain contains three sites of N-linked glycosylation, represented as chains of solid hexagons. The extracellular domain also contains three conserved disulfide bonds, indicated by solid bars connecting the contributing Cys residues. -Subunit mutations investigated here either 1) removed all N-linked glycosylation by site-directed mutagenesis, 2) eliminated a single disulfide bridge by point substitution, or 3) eliminated carboxyl-terminal regions of the -subunit. The arrows represent locations of truncation mutants.

The baculovirus expression system provides a means by which the assembly, trafficking, and activity of heterologous Na,K-ATPase can be examined separately and therefore the molecular components of the -subunit required for each process can be determined. The baculovirus system has several advantages for expression of recombinant Na,K-ATPase, because 1) no endogenous Na,K-ATPase - or -subunits are expressed in the High Five or Sf9 insect cell membranes, 2) protocols for physical separation of membrane compartments have been previously established, and 3) sufficient quantities of recombinant protein can be produced for subsequent biochemical analysis (8). In the work presented here, we specifically address the roles of 1) the phosphorylation site and catalytic activity of the Na,K-ATPase, 2) the N-linked glycosylation, 3) the disulfide bridges and the extracellular regions defined by these bridges, and 4) the amino-terminal and transmembranespanning domains in the assembly, trafficking, and activity of baculovirus-expressed sheep Na,K-ATPase. Our results demonstrate that the enzymatically inactive Na,K-ATPase (D369A) is targeted to the PM, substantiating that enzymatic function is not a prerequisite for trafficking to the surface. We also find that N-linked glycosylation of the -subunit is not necessary for assembly, delivery to the cell surface, or activity of the Na,K-ATPase. Analysis of disulfide bond mutants reveals that correct formation of the second and third disulfide bridges is essential for the exit of the heterodimer from the ER but not necessary for the assembly of the - and -subunits. Further analysis reveals that no discrete trafficking signal is present in the extracellular domain of the -subunit and that the extracellular carboxyl-terminal domain of the -subunit is necessary for enzymatic activity.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Plasmids and Mutant Construction—Mutations were created through overlapping PCR methods using sheep ₁ or ₁ cDNA as a template. Some -subunit truncation mutants required the addition of an epitope tag to allow for subsequent Western blot analysis, since the truncations eliminated the epitope for the anti- antibody. When necessary, a carboxyl-terminal FLAG epitope tag (MDYKDDDD) was incorporated through PCR methods. The tag was also added to the carboxyl terminus of the full-length wild-type -subunit to act as a control for the tag addition. PCR fragments containing mutated or cDNA were amplified by subcloning in the Pocus or Topo vectors, respectively, according to the manufacturer's instructions (Invitrogen). The Bac-to-Bac system (Invitrogen) was used to create recombinant baculovirus. Amplified mutant cDNA was ligated into multicloning site I of pFASTBACDUAL containing wild-type sheep ₁ cDNA in multicloning site II. Mutant cDNA ligated into multicloning site II of the same vector containing wild-type cDNA in multicloning site I (8–10). DH10BAC Escherichia coli, which contain the Bacmid plasmid, were transformed with appropriate pFASTBACDUAL, and T7 transposition of recombinant into Bacmid was selected for by -galactosidase activity. Bacmid was isolated and used for subsequent transfection according to the manufacturer's instructions. Disulfide bridge substitution mutations in the -subunit and the D369A -subunit mutation were confirmed by DNA sequencing of PCR amplification from isolated recombinant virus. Other mutants were confirmed by mobility shift and the presence of the FLAG tag epitope on the expressed protein.

Preparation of Recombinant Baculovirus Viral Stocks—Sf9 cells were transfected with Bacmid vector containing recombinant baculovirus DNA according to the manufacturer's protocol and allowed to produce viral particles for 5–7 days. Supernatant was removed and used to infect 200 ml of Sf9 cells at 0.5 x 10⁶ cell/ml in suspension culture for 5–7 days. Cell debris was removed by centrifugation at 1000 x g for 5 min. Supernatant was collected, stored at 4 °C, and used as viral stock for subsequent infections.

Cell Culture Maintenance—High Five and Sf9 cell lines were maintained between 0.5 and 4.0 x 10⁶ cells/ml at 27 °C in spinner flasks containing Ex-Cell^TM 405 or 420 medium (JRH Biosciences), respectively. Fresh cell lines were regenerated from frozen stocks periodically (every 2.5–3 months).

Protein Expression—Log phase high viability (determined as >98% trypan blue exclusion) High Five cells were infected in the presence of 1% ethanol (v/v) with high titer recombinant baculovirus viral stock such that estimated multiplicity of infection ranged from 1 to 10. Infections were monitored by tracking cell density and viability and then harvested between 3 and 5 days and 50–75% viability by centrifugation (1000 x g, 5 min). Supernatant was decanted, and pellets were stored at –20 °C.

Protein expression levels varied depending on cell health and age of viral stocks. Therefore, protein expression level was compared with a laboratory standard stock protein via Western analysis to gauge relative expression levels. The - and -subunits were also differentially expressed. The -subunit is generally expressed at a 4-fold higher level than the -subunit, as determined by comparison with dog kidney microsome standard and radioactivity incorporation during metabolic labeling experiments.²

Membrane Isolation—Membranes were isolated as previously described (8). Briefly, cell pellets were resuspended in HB (250 mM sucrose, 2 mM EDTA, 10 mM Tris, pH 7.4) and disrupted by Dounce homogenization. Intact cells were removed by centrifugation (1000 x g, 10 min). ER, Golgi, and PM compartments were separated by loading supernatant on a five-step sucrose gradient (2.0, 1.6, 1.4, 1.2, and 0.8 M sucrose) and ultracentrifuged in a Sw28 rotor at 25,000 rpm for 2.5 h. ER, Golgi, and PM were collected from density interfaces (1.6/1.4, 1.4/1.2, 1.2/0.8 M interfaces, respectively), diluted to 25 ml in HB, and ultracentrifuged (Ti60 rotor at 45,000 rpm for 30 min). Supernatant was decanted, and pellets were resuspended in HB containing protease inhibitors (1 µg/ml leupeptin, 2 µg/ml antipain, 1 µg/ml pepstatin, 100 µg/ml TPCK, and 100 µg/ml phenylmethylsulfonyl fluoride) and stored on ice or at –20 °C for long term storage (>1 month). Protein concentration was determined by the method of Lowry using bovine serum albumin as a standard as previously described (9).

ATPase Assay—ATPase activity assay of isolated membranes was performed for 30 min at 37 °C with quantities of membrane protein in the microgram range as previously described (10). Na,K-ATPase activity was determined as the difference in the amount of phosphate liberated in the presence and absence of 0.3 µM ouabain, expressed in µmol of P_i/mg of protein/h. ATPase values for wild-type Na,K-ATPase vary between preparations from approximately 4 to 25 µmol of P_i/mg of protein/h, depending on expression level.

SDS-PAGE and Western Blot—Equal protein concentrations of ER, Golgi, and PM were electrophoresed in 7.5, 10, 12, or 15% acrylamide/bis gels in the presence of 2-mercaptoethanol and then transferred to nitrocellulose. Blots were blocked with 5% dry nonfat milk in PBS (140 mM NaCl, 2.7 mM KCl, 9.8 mM Na₂HPO₄, 1.8 mM KH₂PO₄) for 1 h or overnight and probed with anti- (Affinity BioReagents catalog no. MA3-930; used at 1:1 x 10⁶), anti-₁ (Affinity BioReagents catalog no. MA3-929; used at 1:4 x 10⁶), or anti-FLAG M2 (Sigma catalog no. F-3165; used at 1:1 x 10⁶) for 1 h in PBS-Tween (0.1%) with 1% milk. Blots were washed three times for 10 min in PBS-Tween and then incubated for 1 h with goat anti-mouse horseradish peroxidase-conjugated secondary antibody (Pierce catalog no. 31430; used at 1:2.5 x 10⁴) in PBS-Tween with 1% milk. Blots were washed three times for 10 min in PBS-Tween. Chemiluminescent reagents (Pierce) were used for signal detection.

Co-immunoprecipitations—ER preparations (500 µg) were diluted to 1 ml with HB and ultracentrifuged (TLA 45 rotor at 45,000 rpm, 30 min). Pellets were resuspended in 100 µl of IPB (150 mM NaCl, 10 mM KCl, 2.5 mM MgCl, 25 mM Hepes, pH 7.4, 1 µg/ml leupeptin, 2 µg/ml antipain, 1 µg/ml pepstatin, 100 µg/ml TPCK, and 100 µg/ml phenylmethylsulfonyl fluoride) supplemented with 2% n-dodecyl--maltoside. Solubilization was aided by passing the membrane suspension through a 28-gauge x 5-inch needle and incubating at room temperature (25 °C) for 25–30 min. Insoluble material was removed by centrifugation (17,500 x g, 5 min) at 4 °C.A5-µl aliquot of supernatant was taken as a Western blot standard to estimate co-immunoprecipitation efficiency. The remaining supernatant was transferred to a fresh tube, and 100 µl of polyclonal rabbit anti-loop antibody was added, brought to a final volume of 1 ml with IPB, and incubated at 4 °C with rotation for 4 h or overnight. The quantity of anti-loop antibody previously used (10 µl) (10) was found to limit the quantity of material immunoprecipitated, so antibody levels were increased 10-fold to reach saturating levels for immunoprecipitation. Protein-G-Sepharose (50 µl at 1:1 in IPB) was added and incubated at 4 °C with rotation for 4 h. Samples were centrifuged at 500 x g for 5 min, and supernatant was removed by aspiration. Beads were washed three times for 5 min each with rotation in 1 ml of IPB. After the final wash, 30 µl of 2x SDS sample buffer (125 mM Tris-Cl, pH 6.8, 20% glycerol, 4% SDS, 0.2% 2-mercaptoethanol, 0.0001% bromphenol blue) was added to the beads. After at least 30 min of incubation at room temperature, samples were centrifuged (21,000 x g, 5 min), and supernatant was loaded to SDS-PAGE for subsequent Western blot analysis.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
We have used recombinant baculovirus-infected High Five insect cells to investigate heterodimer assembly of the - and -subunits, trafficking of the heterodimer to the PM, and ATPase activity of Na,K-ATPase constructs in which the enzymatic activity of the pump has been eliminated by the mutation of the essential phosphorylation site in the -subunit (D369A) or in which the -subunit has been altered through removal of N-glycosylation, substitution of disulfide bridges, or truncation of carboxyl-terminal portions of the glycoprotein (Fig. 1). Baculovirus-encoded expression of the Na,K-ATPase in insect cells provides an advantage over other expression systems for such analysis, because the High Five cells do not express endogenous Na,K-ATPase subunits. Therefore, attributes of any Na,K-ATPase detected are those of the recombinant pump. High Five cells are capable of normal synthesis, posttranslational modification, sorting, and trafficking of the Na,KATPase (8, 9). Previously, we have observed that expression of both the - and -subunits is necessary for targeting the Na,KATPase to the PM (8). When the -subunit is expressed in the absence of the -subunit, is retained in the ER (Fig. 2A). In contrast, if the -subunit is expressed in the absence of , is delivered to the PM (Fig. 2B). When the - and -subunits are expressed together, both subunits are targeted to the PM (Fig. 2C). The addition of a carboxyl-terminal FLAG epitope to the -subunit does not alter the assembly, targeting, or activity of the heterologously expressed Na,K-ATPase, although it may slightly decrease the -subunit mobility due to increased mass and additional positive charge (Fig. 2D). It should be noted that the -subunit expressed in insect cells becomes glycosylated in the ER, and the -subunit Western blot of the ER typically shows evidence of a lower molecular weight species corresponding to unglycosylated -subunit, which runs at less than 35 kDa, as well as the diffuse glycosylated -signal ranging from 35 to 48 kDa (Fig. 2, B and C, lower blot). The complex extension of oligosaccharides in the Golgi of mammalian cells does not occur in insect cells. Therefore, the wild-type -subunits from the G and PM of insect cells are indistinguishable in mobility from the glycosylated form in the ER (Fig. 2, B, C, and D (lower blots)). The cDNAs for both the - and -subunits are contained in a single baculovirus particle, and therefore all infected cells express both subunits, but the expression levels of the - and -subunits are independent. In wild-type - expressions, 4-fold more -subunits than -subunits are generally present (data not shown). Therefore, a significant amount of -subunits observed in the PM is free -subunits not associated in a heterodimer. Likewise, solubilized protein subject to coimmunoprecipitation, which is included as a control lane in immunoprecipitation experiments probing for -subunits (Fig. 2E), contain both assembled -subunits that may be immunoprecipitated by the antibody against the -subunit and free -subunits that are unavailable for precipitation using this method.

View larger version (67K): [in this window] [in a new window]   FIG. 2. Distribution of Na,K-ATPase subunits in cell membranes. High Five cells were infected with recombinant baculovirus containing only wild-type -subunit cDNA (A), only wild-type -subunit cDNA (B), cDNA for both wild-type - and -subunits (C), or cDNA of wild-type -subunit and wild-type -subunit with a carboxyl-terminal FLAG tag epitope (D). Cells were fractionated, and 10 µg of membrane protein per lane was resolved on SDS-PAGE. Western blots were probed with anti-1 antibody (A and top panels of C and D) or anti- antibody (B and bottom panels of C and D). E, ER membranes (500 µg) were solubilized in 2% n-dodecyl--maltoside, and 5% of the total solubilized protein was resolved in the lane marked ER. Antibody directed at the -subunit (anti-loop) was used to precipitate, and Western blot analysis probing for the -subunit was performed on all immunoprecipitated (IP) material.

Enzymatically Inactive Na,K-ATPase—In order to determine whether the functional activity of the Na,K-ATPase was important for PM targeting of the heterodimer, we replaced the aspartate residue in the -subunit essential for enzymatic phosphorylation with an alanine (D369A). We then expressed the D369A -subunit with the wild-type -subunit in High Five cells. Cells were disrupted by Dounce homogenization and the ER, G, and PM fractions separated on a five-step sucrose gradient (8). Analogous mutation in yeast H-ATPase causes trafficking defects (11, 12). However, Fig. 3 demonstrates that both the - and -subunits of the enzymatically inactive Na,KATPase are targeted to the PM of baculovirus-infected High Five cells with a compartmental distribution like the wild-type enzyme. No Na,K-ATPase activity could be detected with the a D369A mutation.

View larger version (33K): [in this window] [in a new window]   FIG. 3. Inactive Na,K-ATPase heterodimers are transported to the PM. Enzymatic activity of the Na,K-ATPase was blocked by D369A substitution in the -subunit. D369A was expressed with wild-type -subunit in baculovirus-infected High Five cells, and 10 µg of fractionated cell membranes were subjected to Western blot analysis probing for the -subunit (A) or the -subunit (B).

Nonglycoslyated -Subunit—Wild-type ₁ has three N-linked glycosylation consensus sites (NX(S/T)) (Fig. 1). Previous work demonstrated that all three of the sites are glycosylated (13–15). In order to assess the role that these sugars might play in the assembly, targeting, and activity of the Na,K-ATPase, we replaced the Asn in each of the three glycosylation consensus sites with Gln (N158Q/N193Q/N265Q) to create the N3Q mutant. The N3Q -subunit was expressed with the wild-type sheep ₁-subunit in baculovirus-infected High Five cells, and membrane compartments were separated. On SDS-PAGE, the N3Q -subunit predominantly migrates as a tight band of 32 kDa, as expected for a -subunit lacking posttranslational modification (Fig. 4B, upper band). This 32-kDa band corresponds to the lowest molecular weight form of wild-type -subunit in the ER fraction of insect cells expressing the wild-type proteins (Fig. 2, B–D). Lower molecular weight bands evident in the N3Q mutant expression (Fig. 4B) are not observed in wild-type expression (Fig. 2C, lower blot). These smaller fragments are probably due to degradation by proteases at sites protected by the sugar moieties in wild-type -subunit. Increased proteolytic sensitivity has been previously reported for -subunits in toad bladder cells when glycosylation was inhibited (16).

View larger version (32K): [in this window] [in a new window]   FIG. 4. Effect of eliminating -subunit glycosylation on delivery of the heterodimer to the PM. High Five cells were infected with baculovirus containing cDNA of the sheep wild-type 1-subunit and the N3Q -subunit mutant. Cells were fractionated, and 10 µg of membrane protein per lane was resolved on SDS-PAGE. Western blots were probed with anti-1 (A) or anti- antibody (B).

Analysis of Western blots from each membrane compartment probed for either the -or -subunit reveals efficient delivery of both subunits from the ER to the PM (Fig. 4). This distribution pattern is also observed in wild-type co-infections (Fig. 2, C and D) (8). Likewise, distribution of the ATPase activity of the N3Q mutant was similar to wild-type Na,K-ATPase. Typically, the ATPase activities for wild-type protein are distributed between the ER, G, and PM at 25, 30, and 45%, respectively. For the heterodimer containing the N3Q -subunit, the average distribution of ATPase activity from six expressions were 20, 21, and 59% in the ER, G, and PM, respectively. The N3Q activity distribution correlates with the observed distribution of -subunit in Western blot analysis (Fig. 4A). Absolute ATPase values of the N3Q heterodimer in the PM fraction ranged from 1.5 to 6.4 µmol of P_i/mg/h in different preparations, which is comparable with wild type when protein expression level is taken into account. These results demonstrate that the lack of N-linked sugar moieties does not impede the assembly, PM targeting, or functional activity of the Na,K-ATPase.

Cys-Cys Bridge Substitution Mutants—Three disulfide bridges exist in the extracellular domain of the -subunit (Fig. 1). These disulfides are conserved throughout isoforms of both the Na,K-ATPase and the H,K-ATPase. These bridges have been implicated in enzymatic activity, because reduction of the disulfide bonds by dithiothreitol inactivates the Na,KATPase through the loss of K⁺ occlusion (6). In addition, the disulfides have been implicated in PM expression of the pump because of the inability in previous work to detect cell surface expression of heterologous protein containing disrupted disulfide bridges (14, 17). However, it is not clear from previous studies if the disulfide bridges are important in the heterodimer assembly process, the trafficking process, or both processes. To clarify the role of the disulfide bridges, we created three double point mutants in which both cysteine residues contributing to one of the disulfide bridges in the -subunit were replaced with alanine (C126A/C149A, C157A/C175A, or C213A/C276A). We also created the single point mutant C126A, which eliminates the first disulfide bridge via a single substitution. The Cys bridge mutants were expressed with the wild-type -subunit in High Five cells and subjected to fractionation and analyses as described above.

Removal of the first disulfide bridge through mutation of both Cys residues (C126A/C149A) did not disrupt targeting of the Na,K-ATPase to the PM, as evident from wild-type-like distributions for both the -subunit (Fig. 5A) and -subunit (Fig. 5B) between the membrane compartments. The -subunit Western blot of C126A/C149A (and all other bridge substitution mutants) produced distinctive banding not normally discernible in the wild-type protein in which a diffuse signal through the range of bands is normally observed (Fig. 2, B and lower blots of C and D). This banding probably corresponds to the presence of one, two, or all three glycosylation chains (Fig. 5B). The lowest molecular weight band of the C126A/C149A signal, which corresponds to the -subunit lacking N-glycosylation, is of higher relative intensity than observed with the wild-type -subunit (compare Fig. 5B with Fig. 2B and lower blots of Fig. 2, C and D) These data suggest that substitution of the Cys bridges affects the addition and/or maturation of glycosylation on the -subunit. A link between disulfide bond formation and glycosylation processing has been suggested previously (14).

View larger version (63K): [in this window] [in a new window]   FIG. 5. Membrane distribution and assembly of heterodimers lacking the -subunit first disulfide bridge. The first disulfide bridge of the -subunit was removed by mutation of both (C126A/C149A; A–C) or one (C126A; D–F) of the Cys residues forming the bond. Each bridge mutation was expressed with the wild-type -subunit in High Five cells. Membranes were isolated, and 10 µg of ER, G, and PM protein was separated by SDS-PAGE. Western blots were probed with anti- (A and D) or anti- antibody (B and E). The -subunit was immunoprecipitated from solubilized ER membranes (C and F), and the blots were probed for the -subunit. 5% of the solubilized ER membrane was loaded beside the co-immunoprecipitation (IP) lane as a control.

Co-immunoprecipitation of the C126A/C149A -subunit by an antibody directed at the -subunit (anti-loop) confirmed that the assembly of the first bridge C126A/C149A mutant -subunit into the heterodimer occurs in the ER (Fig. 5C). The difference between the banding pattern of the immunoprecipitations of disulfide bridge mutants (Figs. 5, C and F, and 6, C and F) and wild-type (Fig. 2E) is a reflection of the effects on processing of the mutants referred to above. Distribution of ouabain-sensitive ATPase activity of the heterodimer containing the C126A/C149A -subunit was 22, 27, and 50% in the ER, G, and PM, respectively, which is essentially the same distribution as wild type. The total value of Na,K-ATPase activities in the PM fractions ranged from 2.0 to 8.9 µmol of P_i/mg/h and were a reflection of variable protein expression levels from different infections.

View larger version (52K): [in this window] [in a new window]   FIG. 6. Substitution of second or third Cys-Cys bridges of the -subunit disrupts PM delivery of the heterodimer. Both Cys residues contributing to either the second (C157A/C175A; A–C) or third (C213A/C276A; D–F) disulfide bridge were replaced with Ala, and subunits were expressed with wild-type -subunit in High Five cells. Western blots of 10 µg of protein from each compartment were probed for (A and D) or (B and E). Co-immunoprecipitation (IP) yields are shown next to 5% of solubilized ER membrane protein (C and F).

An altered trafficking pattern was observed when the first disulfide bridge was eliminated through the removal of only one (C126A) rather than both (C126A/C149A) Cys residues. The C126A mutation causes a slight increase in ER retention for the -subunit and a more drastic retention of the -subunit (Fig. 5, D and E). Sufficient amounts of C126A heterodimers are trafficked to the PM to produce an ouabain-sensitive ATPase activity of 4 µmol of P_i/mg/h in the PM fraction. Despite the slight retention of the -subunit and extensive retention of the -subunit, the distribution of Na,K-ATPase activity between ER, G, and PM compartments with the C126A -subunit was 21, 30, and 49%, respectively, which is similar to wild type. This implies that a greater proportion of the Na,KATPase that is folded incorrectly (inactive) is retained in the ER. Since the -subunit is expressed at higher levels than the -subunit and the -subunit is the catalytic subunit, only the distribution of the -subunit need be considered in relation to the observed distribution of ATPase activity. The observation of normal ATPase distribution, along with ER retention suggests that two pools of the C126A -subunit may exist: a correctly folded and functional pool, which is targeted to the PM in the form of the heterodimer, and a misfolded pool, which is retained in the ER. Co-immunoprecipitation of the C126A -subunit by anti-loop antibody demonstrates that the heterodimer assembles in the ER (Fig. 5F). However, it cannot be determined whether this co-immunoprecipitated -subunit represents only correctly folded C126A or both correctly folded and misfolded C126A.

The - and -subunits are both predominantly confined to the ER when either the second (C157A/C175A) or third (C213A/C276A) disulfide bridges are eliminated through substitution of both contributing Cys residues (Fig. 6, A, B, D, and E). These data clearly demonstrate that the formation of the second and third disulfide bonds plays an important role in enabling exit from the ER. Despite ER retention, co-immunoprecipitation clearly shows that the second and third bridge substitution mutants are capable of assembly with the -subunit (Fig. 6, C and F). Although the second and third disulfide bond mutants are largely retained in the ER, the small amount of heterodimer detectable in the PM is sufficient to produce small but measurable levels of ouabain-sensitive Na,K-ATPase activity (1.5–2.1 µmol of P_i/mg/h). No significant ATPase activity was detectable in the ER or G compartments, indicating that the few heterodimers trafficked to the PM represent a small population of functional enzyme.

Extracellular Domain Truncations—We have shown above with the full-length -subunit that the disulfide bridges play an integral role in the production of heterodimer capable of efficiently exiting the ER. This ER retention may occur because the structure(s) produced by the extracellular domain and stabilized by the disulfide bonds are necessary for the presentation of a PM signal in the heterodimer. Alternatively, the Cys bridges may prevent unfolding or misfolding of the domain and retention of the misfolded protein in the ER though interactions with chaperones. In order to distinguish between these possibilities, we constructed a series of truncation mutants in which the carboxyl-terminal region beyond the first cysteine of each of the disulfide bond was truncated. If ER retention of the disulfide substitution mutants is due to recognition of misfolded protein, we hypothesized that removal of misfolded segments may alleviate ER retention, whereas if the disulfides are important for producing a PM targeting signal, their removal would result in ER retention. These truncations terminated the normal coding sequence of the -subunit prior to the first Cys of each of the disulfide bonds at residues 212, 157, and 125, as depicted by the arrows in Fig. 1. Coding sequence for residue 213 was replaced with a stop codon to create the 212 construct. A carboxyl-terminal FLAG tag epitope (MDYKDDDD) was added beyond residue 157 or 125 to create the 157FLAG and 125FLAG constructs. The FLAG tag was necessary because these constructs lacked the epitope for the anti- antibody. The FLAG tag was also added to the full-length wild-type -subunit to determine whether the presence of the epitope altered trafficking, assembly, or activity of the Na,K-ATPase. The behavior of the wild-type -FLAG was the same as wild type (Fig. 2, compare C and D).

The 212 -subunit and the -subunit are predominantly ER-retained (Fig. 7, A and B). Co-immunoprecipitation of the ER fraction shows that 212 is pulled down by the anti-loop antibody and therefore associates with the -subunit in the ER (Fig. 7C). However, the efficiency of co-immunoprecipitation is lower than normally observed (compare relative intensities of ER standard and IP in Fig. 7C with relative intensities in Fig. 5, C and F, and/or Fig. 6, C and F). No significant ouabain-sensitive ATPase activity was detected in any membrane compartment for the 212 -subunit and wild-type expression.

View larger version (55K): [in this window] [in a new window]   FIG. 7. Truncation beyond third disulfide bridge causes ER retention. The extracellular domain of the -subunit beyond and including the third disulfide bridge was removed (212) and expressed in High Five insect cells with wild type . Western blots containing 10 µg of fractionated membrane protein/lane were probed with antibody against the -subunit (A) or -subunit (B). Co-immunoprecipitation of ER membrane protein was performed by precipitation with anti-loop antibody against the -subunit and probing the Western blot for the -subunit (C). As a control, 5% of the protein available for immunoprecipitation was resolved in the lane marked ER. IP, immunoprecipitation.

The 157FLAG -subunit does not contain any N-glycosylation sites and therefore migrates as a tight band through SDS-PAGE at the expected size of 18 kDa (Fig. 8B). When expressed with the -subunit, both subunits were primarily retained in the ER, although some protein could be clearly detected in the G and PM fractions (Fig. 8, A and B). Direct testing of heterodimer assembly by co-immunoprecipitation experiments could not be performed, because the 157FLAG -subunit aggregates and is not soluble in any detergents normally used for Na,K-ATPase or H,K-ATPase solubilization prior to immunoprecipitation (data not shown). No significant ouabain-sensitive ATPase activity could be detected in any membrane compartment for the 157FLAG and wild-type expression.

View larger version (50K): [in this window] [in a new window]   FIG. 8. Membrane distributions of the - and -subunits when lacks the extracellular domain beyond the second Cys-Cys bridge. Removal of the carboxyl-terminal domain beyond residue 157 and the addition of a carboxyl-terminal FLAG epitope produced the 157FLAG -subunit, which was expressed in High Five cells with the wild-type -subunit. Fractionated membranes (10 µg) were subjected to Western blot analysis probing with anti-1 (A) or anti-FLAG antibody (B).

The 125FLAG truncation eliminates all three extracellular Cys bridges. Expression of 125FLAG with wild-type primarily results in retention of 125FLAG in ER (Fig. 9B). However, sufficient levels of 125FLAG are targeted to the PM to allow the -subunit to be delivered to the PM (Fig. 9A). This is probably due to the higher expression level of the -subunit than the -subunit (see "Experimental Procedures"). The 125FLAG -subunit is capable of efficient assembly with the -subunit, as demonstrated by co-immunoprecipitation (Fig. 9C). As with other truncation mutants, no ouabain-sensitive ATPase activity is detectable in any membrane fraction with 125FLAG.

View larger version (45K): [in this window] [in a new window]   FIG. 9. The effect of truncating beyond residue 125 on subunit distribution and assembly. 125FLAG was created by removal of the carboxyl-terminal domain of the -subunit beyond residue 125 and the addition of a FLAG tag epitope and then expressed in High Five cells with wild-type . Western blots of ER, Golgi, and PM were probed for (A) or FLAG (B). C, co-immunoprecipitation (IP) from ER membrane was performed with anti-loop antibody and Western blot probed with anti-FLAG. The lane marked ER represents 5% of the solubilized material used for precipitation.

Extracellular Domain Deletion—It had been previously concluded that the transmembrane and amino-terminal domains of the -subunit provide no specific interactions responsible for heterodimeric assembly and stabilization (18). We constructed the extensively truncated 75FLAG mutant to test this claim. To ensure proper membrane insertion, some predicted extracellular residues were retained, and the carboxyl-terminal FLAG epitope was added to allow for Western blot detection (Fig. 1). Fractionation of expressed 75FLAG -subunit and wild-type -subunit demonstrates that both subunits are delivered to the PM (Fig. 10, A and B). Co-immunoprecipitation of 75FLAG -subunit by anti-loop antibody confirms the ability of the 75FLAG -subunit to assemble into a heterodimer with the -subunit (Fig. 10C). However, reduced yields of immunoprecipitated 75FLAG -subunit indicate a weaker interaction in the heterodimer. No ouabain-sensitive ATPase activity was observed in any membrane compartment with the 75FLAG mutation.

View larger version (42K): [in this window] [in a new window]   FIG. 10. Assembly and trafficking of the -subunit lacking an extracellular domain. The extracellular domain was eliminated beyond residue 75, and a carboxyl-terminal FLAG tag epitope was incorporated. In High Five cells, the resulting 75FLAG -subunit was expressed with the wild-type -subunit, and 10 µg of fractionated membrane protein was separated by SDS-PAGE. Western blots were probed with anti- antibody (A) or anti-FLAG antibody (B). The ER fraction was solubilized with n-dodecyl--maltoside (ER lane of C represents 5% of solubilized material), subject to precipitation with antibody directed at the -subunit, and probed with anti-FLAG for 75FLAG (C).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
In the present work, we have utilized baculovirus-infected insect cells to investigate the assembly and delivery to the PM of Na,K-ATPase heterodimers. Although the -subunit is delivered to the PM when expressed in the absence of the -subunit, the reverse is not true. In the absence of the -subunit, the -subunit is retained in the ER (8). This demonstrates a chaperone-like role for the -subunit in the trafficking process. Here, we separate the overall process into the steps of 1) assembly of heterodimer units in the ER and 2) the delivery of the heterodimer or individual subunits from the ER to the PM. We systematically examine the roles of extracellular components of the -subunit in the assembly, trafficking, and functioning of the Na,K-ATPase. Our data show that although assembly of units must occur prior to exit from the ER, assembly is not sufficient for PM delivery, because several mutant -subunits, such as the second and third disulfide bridge substitution mutants, are capable of assembly with the -subunit but remain in the ER.

Trafficking of Inactive Na,K-ATPase—Point mutation of the enzymatic site of phosphorylation in PMA1 ATPase in the yeast leads to misfolding and intracellular retention of the pump (11, 12). Likewise, in studies where the catalytically essential Asp of the Na,K-ATPase was replaced with Asn, the appearance of the pump at the cell surface in Xenopus oocytes could not be detected (19). These studies implied that the presence of catalytic activity might be important for the processing and delivery of P-type ATPases. Subsequently, the Asp to Asn catalytic mutant of Na,K-ATPase has been detected in crude membrane preparations from NIH3T3 cells (20), and cell surface trafficking of point-mutated Na,K-ATPase expressed in yeast has been observed (21), indicating that the catalytically inactive Na,K-ATPase was properly targeted to the PM in yeast. As Pederson et al. (21) have recognized, the retention of catalytic Asp mutants differs among P-type ATPases and in various cell expression systems; therefore, the ability to overexpress relatively high levels of ATPase in the absence of endogenous pump subunits provides a distinct advantage in the ability to access catalytically inactive mutants. Pederson et al. (21) report in yeast that the Asp mutant is delivered to the surface (as confirmed in our studies), whereas in oocytes delivery is not achieved. Here, we have utilized the baculovirus-infected insect cell system and directly observed the distribution between the ER, G, and PM of sheep Na,K-ATPase containing the D369A mutation. We find that the inactive D369A Na,K-ATPase, like wild type, is delivered to the PM. These results clearly demonstrate that 1) functional activity is not a requirement for PM delivery and 2) the D369A -subunit of the Na,K-ATPase, unlike the yeast H-ATPase, achieves a correctly folded and delivery-competent conformation.

Glycosylation of the -Subunit—The ₁-subunit in vivo has three N-linked glycosylation sites in its extracellular domain. In the present work, we show that removal of all N-linked glycosylation, by replacing the three asparagine residues at the points of sugar attachment with glutamines, produces no significant effect on Na,K-ATPase assembly or PM delivery when expressed in High Five insect cells. The distribution of protein (and its activity) among the membrane compartments is unaltered in comparison with the fully glycosylated wild-type protein. This confirms previous conclusions from work in oocytes, sensory neurons, and epithelial cells in which drug inhibition of cellular glycosylation did not alter expression or activity of Na,K-ATPase composed of the ₁ isoform from various species (16, 22–24). Although it had been previously concluded from work with Asn to Gln mutants in oocytes that one sugar residue was necessary for the -subunit to acquire an assembly-competent conformation (14), we have demonstrated that the completely unglycosylated -subunit is capable of assembly, delivery to the PM, and enzymatic activity.

Glycosylation is not necessary for assembly, expression, or function of the Na,K-ATPase for most subunit isoform combinations (25), but it does seem to contribute to the protease resistance of the Na,K-ATPase. It has been shown that resistance to tryptic digestion is compromised by drug inhibition of glycosylation of (16), but the degradation rate in cultured neuron cells is unaltered (24), and stable intracellular pools of -subunit mutants lacking glycosylation have been observed in oocytes (14). Our data provide evidence of protease susceptibility of the N3Q -subunit in membranes isolated from High Five cells, which is not observed with the glycosylated wild-type -subunit (see Fig. 4, lower bands in B). In light of our data and work by others, we suggest that the N-linked glycosylation of the ₁-subunit is of minimal importance for the formation, trafficking, or functioning of the ₁₁ Na,K-ATPase but that the glycosylation may serve to protect the ₁-subunit from degradation by cellular proteases and thereby play a role in the stability of the Na,K-ATPase.

Role of Disulfide Bonds in Expression and Functioning of the Na,K-ATPase—To date, all Na,K-ATPase and H,K-ATPase -subunits identified contain six conserved extracellular Cys residues that give rise to three disulfide bonds. Previous work has implicated some or all of these disulfides as being essential for function, assembly, and/or expression (6, 14, 17, 26). Disruption of these bridges through site-directed mutagenesis has consistently shown that blocking the formation of the second and third bridges causes a more severe disruption than the mutation of the first bridge (14, 17, 26). However, the conclusions drawn regarding the ability of these Cys bridge-substituted -subunits to assemble with and stabilize the -subunit vary. Noguchi and co-workers (26) found that the -subunit of the T. californica with the first bridge removed could assemble with the -subunit but produced little ouabain-sensitive ATPase activity, whereas neither the second nor third bridge substitution mutants were able to assemble with the -subunit. Beggah et al. (14) found that Xenopus -subunits with the first bridge substituted could assemble with the -subunit and cause a small increase over background in the number of active pumps at the cell surface and that substituted second or third disulfide bridges were able to transiently assemble with the -subunit.

The results we have obtained in the insect cell expression system clarify the contributions of each of the disulfide bonds in the assembly and trafficking of the Na,K-ATPase heterodimer. Co-immunoprecipitation of mutant -subunits with an antibody directed at the -subunit from ER fractions of each of the disulfide bridge substitution mutants clearly demonstrates that these -subunits are capable of heterodimer assembly (Fig. 5, C and D, and Fig. 6, C and D). However, substitution of the second and third disulfide bridge in particular and to some degree the first disrupts the ability of the mutant -subunits and of the heterodimer to be efficiently targeted to the PM. These results indicate that the disulfide bonds, especially the two carboxyl-terminal disulfides, are important for making the -subunit, and thereby the heterodimer, trafficking-competent.

Interestingly, the nature of the mutation eliminating the first disulfide bridge alters the ability to traffic (Fig. 5). The elimination of the disulfide through the single mutation C126A causes significant ER retention of the -subunit, but removal of both Cys (C126A/C149A) produces a heterodimer that is delivered to the PM. It is tempting to postulate that the presence of Cys¹⁴⁹ may cause incorrect disulfide formation during the folding pathway between Cys¹⁴⁹ and some other Cys residue and that this formation may lead to an inappropriate fold or trap the -subunit in a folding intermediate. The necessity of correct disulfide formation of the second and the third bridges in the -subunit extracellular domain may explain why free Cys residues are not found in the extracellular domain of any -subunit.

The cause of ER retention of the disulfide bridge-substituted mutants can be interpreted in one of two ways. As discussed above, the lack of disulfide formation probably leads to a misfolding of the -subunit extracellular domain. This misfolding could 1) disrupt or mask a structure or signal that is essential for trafficking to the PM or 2) cause ER retention due to recognition of the misfolded protein. We reasoned that if ER retention was due to recognition of misfolded sections of the -subunit, the retention could be overcome through removing the misfolded region of the extracellular domain. We therefore constructed a series of mutants in which the -subunits were truncated just prior to the beginning of each of the disulfide bridges. Our results indicate that elimination of larger stretches of the carboxyl terminus enabled greater fractions of heterodimer to reach the PM compartment (compare ER and PM fractions of anti- Western blots in Figs. 7, 8, 9, 10). Our findings lead us to conclude that the ER retention of the Na,KATPase that we observe is a result of the mutant -subunits being recognized as misfolded proteins. We speculate that this retention is mediated through the interaction of the -subunit with ER-resident chaperone proteins. It is possible that different observations made in various cell systems are due to the effects of cell-specific chaperones. We are currently investigating the role of molecular chaperones in the assembly process. Our current findings highlight the danger of identifying trafficking signals or interaction motifs based on the ER retention of mutant subunits, which may be misfolded.

-Subunit PM Targeting Signal—Numerous examples of short amino acid stretches serving as targeting motifs have been recorded in the literature. Specific examples include KDEL, dileucine, and tyrosine-based signals. Generally, these series of amino acids are thought to mediate targeting of cargo through the specific interaction with other proteins, such as an adaptor or receptor proteins (for reviews, see Refs. 27 and 28). Previous work in which the -subunit carboxyl-terminal residues were sequentially removed (18, 29, 30) led to the suggestion that the final 10 amino acid residues of the -subunit serve as a PM targeting motif due to the observed ER retention of this -subunit (29). However, our results indicate that these residues do not serve an essential role in delivery of the -subunit to the PM.

Elimination of a PM target signal would be expected to completely abolish the PM delivery of the heterologously expressed Na,K-ATPase. However, our data demonstrate that -subunits with carboxyl-terminal truncations much larger than 10 residues did not eliminate PM delivery. We found that -subunits with larger truncations enabled more -subunits and therefore heterodimers to arrive at the PM (see Figs. 7, 8, 9, 10). Strikingly, the 75FLAG -subunit, which eliminates almost the entire extracellular domain, enables delivery of the heterodimer to the PM. It has previously been demonstrated that Na,K-ATPase with -subunits lacking the amino-terminal domain are transported to the PM (3, 31). Similarly, heterodimers in which the -subunit transmembrane domain was replaced with a transmembrane region from another unrelated protein were delivered to the PM, leading to the suggestion that the extracellular domain alone was involved in associations (18). Together, our results and this body of data lead us to suggest that no specific PM targeting motif exists in the -subunit.

Regions of - Interaction—Significant effort has been made to identify the regions of the - and -subunits that are involved in the heterodimeric assembly of the Na,K-ATPase. Although the amino-terminal (32) and transmembrane (33, 34) domains have been implicated in having specific interactions with the -subunit, the -subunit extracellular domain has been identified as being necessary and sufficient for assembly with the -subunit (18, 35, 36). Specifically, the extracellular domain of avian ₁-subunit residues 63–123 have been shown to interact with the M7M8 extracellular loop of the -subunit through yeast two-hybrid analysis (36). These residues correspond to the region from Glu⁶³, which is predicted to be just outside of the membrane, to Asp¹²⁵, which is just prior to the first Cys residue of the first disulfide bridge in the sheep ₁-subunit (see Fig. 1). Here, we have shown that the -subunit truncation, 75FLAG, which only retains 13 of the 60 residues included in the yeast two-hybrid experiments, is capable of assembly with the -subunit as determined by co-immunoprecipitation (see Fig. 10C). This suggests that when the cytoplasmic and transmembrane regions are intact, the extracellular domain is not absolutely necessary for heterodimer assembly. We propose that interactions between both the transmembrane and extracellular domains of the - and -subunits contribute to associations between the subunits and contribute to the stability of the heterodimer. Removal of some of these interactions through selective truncation does not sufficiently destabilize the heterodimer to disrupt association but may affect the overall stability of the interaction.

In summary, although the -subunit of the Na,K-ATPase is structurally the simpler of the two subunits, it is clear that the -subunit contributes to the functional characteristics of the Na,K-ATPase in a complex manner. This inherent complexity is emphasized by the different observations obtained with a variety of expression systems and with different -subunit isoforms. It is difficult to draw many wide ranging conclusions from any single series of studies in a single expression system. However, certain limited conclusions can be made. First, we find no evidence of a specific identifiable signaling sequence at precise locations in the -subunit that directs its PM location. Second, the sets of interactions with the -subunit that stabilize the heterodimer probably contain contributions from multiple regions of the -subunit, not just the extracellular domain as had been previously suggested. Third, although assembly of - and -subunits is required for delivery of stable, functional Na,K-ATPase to the PM, such assembly is not sufficient to ensure that the heterodimer will be efficiently targeted to the PM. Finally, in most cases where structural modifications in the -subunit result in an increase in ER retention, trafficking is not an all-or-none phenomenon, and some PM delivery still occurs.

	FOOTNOTES

 
^* This work was supported by National Institutes of Health Grant HL 30315 (to J. H. K.) and by an American Heart Association Fellowship (to M. D. L.). 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.

^¶ To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Genetics, University of Illinois, 900 S. Ashland Ave., Chicago, IL 60607. Tel.: 312-355-2732; Fax: 312-355-1765; E-mail: kaplanj{at}uic.edu