Functional Expression of the Colonic H+,K+-ATPase α-Subunit PHARMACOLOGIC PROPERTIES AND ASSEMBLY WITH X+,K+-ATPase β-SUBUNITS

The functional and pharmacological properties of the α-subunit of the colonic H+,K+-ATPase (αC) were studied in Xenopus laevis oocytes. αC was injected with different rat β-subunits, the β-subunit of the gastric H+,K+-ATPase (βG, the only H+,K+-ATPase β-subunit identified in rat), or the β1-subunit of the Na+,K+-ATPase (β1) (associated with the basolateral Na+,K+-ATPase, but also expressed in the epithelial apical membranes of rat distal colon) (Marxer, A., Stieger, B., Quarini, A., Kashgarian, M., and Hauri, H. P. (1989) J. Cell Biol. 109, 1057-1069). The effect of the different β-subunits was studied by measuring 86Rb+ uptake (a K+ congener) in the presence or absence of Sch-28080 and ouabain. Significant Na+-independent 86Rb+ uptake was observed only when αC was coexpressed with one of the β-subunits. The expressed αCβ1 and αCβG complexes were not inhibited by Sch-28080, were only partially sensitive to ouabain (IC50 = 400-600 μM, in the presence of external 1 mM KCl), and exhibited comparable K+ activation kinetics. Coexpression of αC with epitope-tagged βG or β1, followed by immunopurification of the αβ complexes, confirmed stable assembly of αCβG and αCβ1 complexes. Since the β1-subunit, but not the α1-subunit, of Na+,K+-ATPase is expressed in the apical membrane of rat colonocytes, our data support the view that, in rat distal colon, the β1-subunit may play a surrogate role as the β-subunit for the colonic H+,K+-ATPase.

Recent molecular biological and biochemical studies indicated that the colonic H ϩ ,K ϩ -ATPase participates in the chronic adaptation to changes in K ϩ homeostasis (7). 2 While K ϩ balance is governed principally by the kidney, the colon plays a smaller but highly significant role. During chronic K ϩ restriction, active K ϩ reabsorption by epithelial cells of the renal collecting duct and distal colon serves to restore K ϩ balance (13). K ϩ -ATPase activities are expressed in these cell types, and these activities are up-regulated during chronic dietary K ϩ depletion. The K ϩ -ATPase activity in the renal collecting duct was reported to be ouabain-resistant and Sch-28080-sensitive, findings compatible with the established properties of ␣ G (14). However, the fact that expression of ␣ G mRNA is not significantly altered in the medullary collecting duct during chronic dietary K ϩ depletion (15), and that it is not expressed in the distal colon, makes it unlikely that this gene plays a role in K ϩ adaptation in either the kidney or the distal colon (16). 2 In contrast, ␣ C mRNA is principally expressed in the renal collecting duct and surface epithelial cells of the distal colon, and its abundance increases 3-5-fold in the renal medulla during chronic dietary K ϩ depletion (17). 2 Thus, ␣ C has emerged as the candidate gene most likely to mediate K ϩ conservation. Consequently, considerable effort has been devoted to define the functional properties of ␣ C .
Biochemical studies found that the apical membranes of surface epithelial cells of distal colon express distinct ouabainsensitive (18,19) and ouabain-insensitive (19,20) K ϩ -ATPase activities. Since ␣ C is the only X ϩ ,K ϩ -ATPase ␣-subunit known to be expressed in the apical membranes of these cells, these results suggested either the existence of a novel ␣-subunit, or the possibility that different ␣ C ␤-subunit complexes exhibit different ouabain sensitivities. The latter possibility gained credence with the recent conflicting reports regarding the pharmacological properties of ␣ C expressed in heterologous systems. Lee et al. (11) demonstrated that expression of ␣ C without an exogenous ␤-subunit in baculovirus-infected Spodoptera frugiperda (Sf-9) cells yielded K ϩ -ATPase activity that was resistant to high concentrations (1 mM) of ouabain, but inhibited by high concentrations (100 M) of Sch-28080. Subsequently, Cougnon et al. (12) reported that coinjection of Xenopus laevis oocytes with cRNAs encoding ␣ C and a ␤-subunit from toad bladder (␤ bl ) resulted in functional H ϩ ,K ϩ -ATPase activity that was described as ouabain-sensitive (IC 50 ϳ 970 M at 5 mM external K ϩ concentration) but Sch-28080-resistant. Neither study, however, examined the expression of an H ϩ ,K ϩ -ATPase holoenzyme comprised of ␣ C and a mammalian ␤-subunit, nor did they directly establish whether ␣ C can complex with any of the known ␤-subunits. These limitations assume considerable importance when one considers that different ␤-subunits, when coexpressed with the ␣-subunits of the gastric H ϩ ,K ϩ -ATPase or the Na ϩ ,K ϩ -ATPase, may confer unique functional properties on the holoenzyme (21,22). Accordingly, we used the oocyte expression system to examine the functional properties of ␣ C when it is coexpressed with either of two rat X ϩ ,K ϩ -ATPase ␤-subunits: the ␤-subunit of the gastric H ϩ ,K ϩ -ATPase (␤ G ), which is expressed in renal collecting duct (4) but not distal colon (23), and the ␤ 1 -subunit of the Na ϩ ,K ϩ -ATPase (␤ 1 ), which is expressed in both collecting duct (24) and in apical and basolateral membranes of colonocytes (25) in distal colon. We used this expression system instead of the baculovirus system, because Sf-9 cells exhibit endogenous K ϩ /H ϩ exchange (26) that would likely confound interpretation of K ϩ activation kinetics for a heterologously expressed H ϩ ,K ϩ -ATPase. In addition, we constructed ␤ G -and ␤ 1 -subunits bearing a common c-myc epitope to test, in a coimmunoprecipitation assay (27), whether ␣ C stably assembles with ␤ G and/or ␤ 1 . The results indicate that ␣ C can interact with both ␤-subunits, that oligomerization is required for functional activity of the expressed enzymes in this system, and that the inhibitor sensitivities and K ϩ activation kinetics of ␣ C ␤ G and ␣ C ␤ 1 holoenzymes are very similar. The ability of ␤ 1 to support functional activity of ␣ C combined with the colocalization of these subunits in the apical membrane of the colonocytes (25) suggests that ␣ C ␤ 1 likely represents at least one of the K ϩ -ATPase activities expressed in this locale, and that the existence of a "␤ C "-subunit need not be necessarily invoked.

EXPERIMENTAL PROCEDURES
Materials-The ␣ C (7) and ␣ G (5) cDNAs were gifts from Dr. G. Shull (University of Cincinnati). The ␤ 1 cDNA (28) was a gift from Dr. T. A. Pressley (Texas Tech University). The ␤ G cDNA was cloned as described below. The expression vector pAGA#2 was a gift from Dr. L. Birnbaumer (University of California at Los Angeles) (29). mAb 9E10 was purified from culture supernatants of hybridoma myc 1-9E10.2 (American Type Culture Collection, Rockville, MD). Restriction enzymes were from Promega Biotech Inc. and New England Biolabs (Beverly, MA). T7 Cap Scribe was from Boehringer Mannheim. The X. laevis oocytes were prepared and injected in Dr. L. Parent's laboratory (Department of Molecular Physiology and Biophysics, Baylor College of Medicine, Houston, TX). Oligonucleotides were synthesized by Genosys (The Woodlands, TX). The remaining reagents were of the best available quality.
Cloning of cDNA Encoding the Gastric H ϩ ,K ϩ -ATPase ␤-Subunit-The complete open reading frame of ␤ G was generated by a polymerase chain reaction using cDNA prepared from total rat stomach RNA and primers based on the published sequence (23), according to the conditions previously described by our laboratory (16). The sense oligonucleotide (5Ј-ATTCCATGGCAGCCCTGCAGGAGAAG-3Ј) contained an NcoI site (underlined), and the antisense oligonucleotide (5Ј-TACG-GTCGACTTACTTCTGTATTGTGAGCTT-3) contained a SalI site (underlined) to facilitate subcloning into pAGA#2. The correct sequence was verified by direct sequencing (30) of one of the clones.
cRNA Synthesis and Protein Expression in Xenopus Oocytes-Encoding DNAs for rat ␣ G -, ␣ C -, ␤ G -, and ␤ 1 -subunits were subcloned into pAGA#2. The recombinant molecules were linearized with HindIII or XhoI as appropriate, and capped cRNAs were synthesized using T7 RNA polymerase and T7 Cap Scribe (Boehringer Mannheim) according to the manufacturer's methods. Parametric studies demonstrated that full-length cRNAs were generated for each gene. Stage V-VI oocytes obtained from X. laevis were injected with 10 ng of cRNA or an equivalent volume of water and incubated at 19°C in modified Barth's medium (31). Three days later, 86 Rb ϩ uptake was measured according to the procedure described by Modyanov et al. (1). Briefly, the oocytes were equilibrated in solution A (90 mM NaCl (in some instances 90 mM NMDG was used), 1 mM MgCl 2 , 0.33 mM Ca(NO 3 ) 2 , 0.41 mM CaCl 2 , 5 mM BaCl 2 , and 10 mM PIPES, pH 7.4) for 15 min. Subsequently oocytes were preincubated in the presence or absence of inhibitors (ouabain or Sch-28080) in buffer A for 15 min. The oocytes were then incubated in buffer A containing 5-10 ϫ 10 6 cpm 86 Rb ϩ and 1-5 mM KCl in the presence or absence of inhibitors. All incubations were performed at room temperature, and 1 M ouabain was added to all the buffers to inhibit the activity of endogenous X. laevis Na ϩ ,K ϩ -ATPase (1). After 15 min, the reaction was stopped by aspirating the bulk of the radioactivity and washing three times with 4 ml of buffer A at 4°C. Finally, the oocytes were disrupted by pipetting up and down, transferred to scintillation vials, and counted. The quantities of 86 Rb ϩ uptake were then calculated. The blanks (no oocyte) of the experiment were routinely less than 200 cpm.
Subunit Assembly Assay-A human c-myc epitope tag, recognized by human-specific mAb 9E10 (32), was fused to the amino terminus of ␤ G and ␤ 1 by inserting a double-stranded oligonucleotide adapter (sense 5Ј-CATGGAGCAAAAGCTGATCTCCGAGGAGGACCT-3Ј; antisense, 5Ј-CATGAGGTCCTCCTCGGAGATCAGCTTTTGCTC-3Ј) that contained an initiation ATG followed by nucleotides encoding the c-myc peptide (EQKLISEEDL), into each ␤-subunit. The resultant recombinant molecules were termed c-myc-␤ G and c-myc-␤ 1 . The addition of the correct nucleotide sequence was confirmed by DNA sequencing. Oocyte proteins were metabolically labeled by coinjection of 0.345 Ci [ 35 S]methionine with the ␣ and c-myc-␤-subunit cRNAs. Three days after the injection, the oocytes were lysed, and the proteins were extracted by incubating for 30 min at 4°C in the presence of 10 mM Tris-HCl, pH 8.0, 150 mM NaCl, and 1% Triton X-100 (buffer B) (27). The insoluble material was removed by centrifugation at 10,000 ϫ g for 10 min at 4°C. The extracted proteins from 5-10 oocytes were pooled and incubated with 2 g of mAb 9E10 for 4 h at 4°C, followed by addition of 10 l (packed volume) of protein A/G plus agarose (Santa Cruz Biotechnology). After 2 h of vigorous shaking, the resin was washed six times with buffer B. The bound protein was extracted from the resin with Laemmli sample buffer containing 10% ␤-mercaptoethanol and separated on SDS-10% polyacrylamide gels. The gels were then fixed in 7% glacial acetic acid, 25% methanol and impregnated with 2,5-diphenyloxazole. The gels were dried and exposed to Kodak XAR-5 film with intensifying screens at Ϫ70°C for 1-3 days. The coprecipitation of an ␣-subunit with the c-myc-␤ G -or c-myc-␤ 1 -subunit using this protocol was interpreted as subunit assembly in accordance with previous studies of Na ϩ ,K ϩ -ATPase ␣␤ assembly (27).
Data Analysis-Quantitative data are presented as means Ϯ S.E. and were tested for significance by analysis of variance. p Ͻ 0.05 was taken as significant.

Functional
Properties of ␣ C ␤ G or ␣ C ␤ 1 Expressed in Xenopus Oocytes-Sets of oocytes were injected with cRNAs for ␣ C , ␣ G , ␤ G , ␤ 1 , ␣ C plus ␤ G , or ␣ C plus ␤ 1 . The controls were injected with water alone. 86 Rb ϩ uptake was measured 3 days later in the presence of 5 mM KCl. When the oocytes were injected with one subunit only, there was no effect on 86 Rb ϩ uptake compared to water-injected controls (Fig. 1). However, when oocytes were coinjected with ␣ C plus ␤ G or ␣ C plus ␤ 1 there was a significant increase in 86 Rb ϩ uptake compared to any group injected with one subunit alone or the control group. This 86 Rb ϩ uptake was independent of the presence of Na ϩ in the incubation medium; equimolar replacement of external Na ϩ by NMDG did not alter the 86 Rb ϩ uptake in any group. This result established that the 86 Rb ϩ uptake was not mediated by the Na ϩ ,K ϩ -ATPase.
To assess the K ϩ -activation kinetics of the ␣ C ␤ G and ␣ C ␤ 1 complexes, 86 Rb ϩ uptake was assayed in ␣ C plus ␤ G -and ␣ C plus ␤ 1 -expressing oocytes in the presence of increasing concentrations of KCl in the uptake buffer (Fig. 2). External K ϩ activated 86 Rb ϩ uptake in these oocytes in a concentration-dependent, saturable manner. Lineweaver-Burk transformation of the kinetic data revealed K1 ⁄2 values for K ϩ activation of 1.4 mM when the oocytes expressed the ␣ C ␤ G complex and 1.8 mM when the ␣ C ␤ G complex was expressed. Both values are slightly higher than the value (1.2 mM) reported for Sf-9 cells express-ing the ␣ C alone (11). The Hill coefficient for K ϩ was 1.0 in both cases (data not shown), indicating no cooperativity in the K ϩ effect.
Pharmacological Properties of ␣ C Coexpressed with Different ␤-Subunits-In the presence of 1 mM external K ϩ , ouabain inhibited 86 Rb ϩ uptake of oocytes expressing ␣ C ␤ G or ␣ C ␤ 1 in a dose-dependent manner, with IC 50 values of ϳ390 and ϳ640 M, respectively (Fig. 3). In contrast, Sch-28080, at concentrations up to 500 M, failed to inhibit 86 Rb ϩ uptake of oocytes coexpressing ␣ C ␤ G or ␣ C ␤ 1 (Fig. 4). In positive control experiments, 10 M Sch-28080 abolished 86 Rb ϩ uptake of oocytes coinjected with ␣ G plus ␤ G cRNAs, in agreement with previous reports (33).
Assembly of ␣ C ␤ G and ␣ C ␤ 1 Complexes-Since the 86 Rb ϩ uptake experiments indicated that ␣ C must be coexpressed with ␤ G and ␤ 1 for holoenzyme function, we sought to establish directly that stable ␣ C ␤ G and ␣ C ␤ 1 complexes were formed. The human c-myc epitope was added to the amino terminus of ␤ G (c-myc-␤ G ) and of ␤ 1 (c-myc-␤1) (27). Oocytes were injected with ␣ C , c-myc-␤ G , c-myc-␤ 1 , ␣ C plus c-myc-␤ G , or ␣ C plus c-myc-␤ 1 .
The Triton X-100 extracts were immunoprecipitated with mAb 9E10 (27), and samples were analyzed by SDS-PAGE and fluorography. Fig. 5 demonstrates that mAb 9E10 did not immunoprecipitate ␣ C when it was expressed alone (without a ␤-subunit). However, when ␣ C cRNA was coinjected with c-myc-␤ G cRNA, the two subunits were coprecipitated, indicating stable assembly. As predicted from the 86 Rb ϩ uptake data, stable assembly between ␣ C and c-myc-␤ 1 was also evident when the oocytes were coinjected with cRNAs for these subunits (Fig. 5). The relative amounts of ␣ C ␤ 1 and ␣ C ␤ G coprecipitated in this assay were roughly comparable. The immunoprecipitated ␤ G and ␤ 1 were in the core glycosylated (narrow band at ϳ50 kDa) and fully glycosylated (broad band at 60 -70 kDa) forms. Both the core and fully glycosylated ␤ 1 -subunits migrated more rapidly than the corresponding glycosylated forms of ␤ G on SDS-PAGE, as observed by others (34). The explanation for such differences in mobility is unknown. A band that migrated near ␣ C (labeled ? in Fig. 5) was also immunoprecipitated from oocytes injected with cRNA for ␤ G or ␤ 1 , regardless of whether ␣ C cRNA was coinjected. Presumably this band represents the endogenous ␣ 1 -subunit of the Na ϩ ,K ϩ -ATPase. FIG. 2. K ؉ -dependent activation of 86 Rb ؉ uptake in oocytes coexpressing ␣ C and ␤ G -or ␤ 1 -subunit. Top panel, oocytes were coinjected with ␣ C and ␤ G (diamonds) or with ␣ C and ␤ 1 (circles). Three days later, 86 Rb ϩ uptake was measured in the presence of 90 mM NaCl at the indicated external KCl concentrations. Bottom panel, the Lineweaver-Burk plot of the same data. Each point on the plot represents the data obtained from 8 -10 oocytes. Straight lines were fitted by method of least squares. Correlation coefficients (R) were 0.9 for both ␣ C ␤ G and ␣ C ␤ 1 .   FIG. 3. Effect of ouabain concentration on 86 Rb ؉ uptake in oocytes expressing ␣ C ␤ G or ␣ C ␤ 1 complexes. 86 Rb ϩ uptake in oocytes coexpressing ␣ C and ␤ G or ␣ C and ␤ 1 was determined in the presence of 1 mM KCl with addition of the indicated ouabain concentrations. Symbols as in Fig. 1. Eight to 10 oocytes were used for group.

FIG. 4. Effect of Sch-28080 concentration on 86
Rb ؉ uptake in oocytes expressing ␣ C ␤ G , ␣ C ␤ 1 , or ␣ G ␤ G complexes. 86 Rb ϩ uptake, in oocytes coexpressing ␣ C and ␤ G , ␣ C and ␤ 1 , or ␣ G and ␤ G was determined in the presence of 1 mM KCl, 90 mM NaCl and the indicated concentrations of Sch-28080. **p Ͻ 0.01 versus no Sch-28080. Open bars, no inhibitor; closed bars, incubated in presence of Sch-28080. Ten oocytes were used in each group.

DISCUSSION
The colonic H ϩ ,K ϩ -ATPase plays a central role in the regulation of K ϩ absorption by the colon (11,18,35,36) and kidney (37). However, the functional and pharmacologic properties of this isoform, and its requirements for association with a ␤-subunit remain ambiguous. We used the oocyte expression system to study the properties of ␣ C interacting with differing ␤-subunits. The rat ␤ G -subunit was selected for study because it is the only known mammalian H ϩ ,K ϩ -ATPase ␤-subunit, and it is coexpressed with ␣ C in the renal collecting duct (4,38). The rat Na ϩ ,K ϩ -ATPase ␤ 1 -subunit was chosen because immunoreactivity for this protein was found in the apical membrane of rat colonocytes (25), a membrane domain in which Na ϩ -independent K ϩ -ATPase activity, but not Na ϩ ,K ϩ -ATPase activity or ␣ 1 -subunit immunoreactivity, was observed. We therefore hypothesized that both ␤ G and ␤ 1 would support ␣ C functional activity.
In agreement with the majority of studies of X ϩ ,K ϩ -ATPases (9, 12, 23), our 86 Rb ϩ uptake data provide clear evidence that ␣ C requires a ␤-subunit for functional activity in the oocyte expression system. Cougnon et al. (12), using ␣ C coexpressed in oocytes with an amphibian ␤-subunit, reached a similar conclusion. As predicted from our 86 Rb ϩ uptake data, ␣ C was coprecipitated with either the ␤ G -or ␤ 1 -subunit in the assembly assay, confirming the formation of stable heterodimers. The fact that comparable amounts of ␣ C ␤ 1 and ␣ C ␤ G complexes were coprecipitated in the assembly assay (Fig. 5) suggests that the assembly efficiency and/or stability of these heterodimers are quite similar. Using an identical assembly assay, Lemas et al. (27) found that the Na ϩ ,K ϩ -ATPase ␣ 1 -subunit can assemble with ␤ 1 or ␤ G , and they identified a 26 amino acid region of the Na ϩ ,K ϩ -ATPase ␣ 1 sufficient to allow stable interaction with these ␤-subunits. In addition, Jaisser et al. (22) found that ␤ 1 could support functional activity of ␣ G . Our data, therefore, extend the range of potential ␣␤ pairs to include ␣ C ␤ G and ␣ C ␤ 1 , and lend further support to the concepts that X ϩ ,K ϩ -ATPase ␣-subunits contain a conserved assembly domain for ␤-subunit association, and that there is no remarkable ␣/␤ isoform selectivity in the assembly process.
Previous studies suggested that the different ␤-subunit isoforms could confer different K ϩ -activation kinetics on the Na ϩ ,K ϩ -ATPase. Coexpression of the Bufo Na ϩ ,K ϩ -ATPase ␣ 1 -subunit with the Bufo Na ϩ ,K ϩ -ATPase ␤ 1 -subunit, Na ϩ ,K ϩ -ATPase ␤ 3 -subunit, or rabbit H ϩ ,K ϩ -ATPase ␤ G -subunit in Xenopus oocytes resulted in different K ϩ -activation kinetics for the various holoenzymes. The Na ϩ ,K ϩ -ATPase ␣ 1 /H ϩ ,K ϩ -ATPase ␤ G enzyme performed as a Na ϩ ,K ϩ pump with a much lower apparent affinity for K ϩ , both in the presence and absence of external Na ϩ , compared to the Na ϩ ,K ϩ -ATPase ␣ 1 / Na ϩ ,K ϩ -ATPase ␤ 1 and Na ϩ ,K ϩ -ATPase ␣ 1 /Na ϩ ,K ϩ -ATPase ␤ 3 pumps (39). In the present report, however, ␣ C ␤ 1 and ␣ C ␤ G enzymes exhibited comparable K ϩ -activation kinetics. Thus the ability of different ␤-subunits to alter this functional property may be restricted to specific X ϩ ,K ϩ -ATPase ␣-subunits or to hybrid ion pumps of specific species.
Structure-function studies of the Na ϩ ,K ϩ -ATPase ␣-subunit indicated that amino acids in several transmembrane regions (H1, H2, H5, and H6), the first extracellular loop, and the third extracellular loop (specifically Cys 104 , Tyr 108 , Gln 111 , Asp 121 , Asn 122 , and Tyr 308 , Phe 786 , Leu 793 , Thr 797 , Phe 863 , Arg 880 ), contributed to ouabain sensitivity (40,41). Comparison of the amino acid composition of ␣ C with the Na ϩ ,K ϩ -ATPase ␣-subunits indicates that most of these amino acids are conserved, with the notable exception that Tyr 108 and Phe 786 of the Na ϩ ,K ϩ -ATPase ␣-subunit are substituted with Phe and Tyr in ␣ C . The specific role of these two amino acids in conferring relative ouabain-insensitivity to ␣ C has not yet been studied. It has also been suggested that Phe 124 and Asp 137 (rat ␣ G sequence) (6) are required to confer Sch-28080-sensitivity to ␣ G , but the ␣ C and ␣ G sequences are identical at these two positions. Given evidence that ␣ C is Sch-28080-insensitive (12) (present report), sequences other than these two amino acids must be implicated in Sch-28080 binding affinity to the ␣-subunit. One candidate motif for conferring Sch-28080 sensitivity is the sequence GDLT (amino acids 131-134 of ␣ G ), which is conserved in the known ␣ G -subunits (all of which are Sch-28080-sensitive) of amphibians, birds and mammals, but is absent from ␣ C (6). Pharmacologic analysis of chimeric ␣ G /␣ C molecules and site-directed mutants of ␣ C should clarify this question.
Functional studies have suggested the presence of both ouabain-sensitive and -insensitive components of K ϩ -ATPase activity in rat distal colon (11,18,19,20,42) and kidney (43). In addition, a polyclonal antibody directed against the amino terminus of ␣ C inhibited both ouabain-sensitive and -insensitive components of K ϩ -ATPase activity in apical membranes prepared from distal colon, and it specifically labeled only the apical membrane of the distal colon epithelium (11). Since this antibody did not label the basolateral membrane, where the Na ϩ ,K ϩ -ATPase resides in these cells, and did not inhibit Na ϩ ,K ϩ -ATPase activity in membranes from rabbit renal medulla, this antibody does not appear to cross-react with the Na ϩ ,K ϩ -ATPase ␣ 1 -subunit. Moreover, whereas 1 mM ouabain failed to inhibit K ϩ -ATPase activity in ␣ C -expressing Sf-9 cells, it dramatically inhibited (by 75%) K ϩ -ATPase activity of colonic apical membranes. Integration of these data with the present results supports the idea that, in the apical membrane of the distal colon epithelium, ␣ C ␤ 1 holoenzymes contribute the Sch-28080-insensitive and relatively ouabain-insensitive com-FIG. 5. Assembly of ␣ C plus c-myc-␤ G and ␣ C plus c-myc-␤ 1 complexes. Oocytes were injected with ␣ C , c-myc-␤ G , ␣ C plus c-myc-␤ G , ␤ 1 , or ␣ C plus c-myc-␤ 1 in the presence of 0.345 Ci of [ 35 S]methionine. The Triton X-100 extracted proteins were immunoprecipitated with mAb 9E10 and protein A/G PLUS-agarose and analyzed by SDS-PAGE and fluorography as described under "Experimental Procedures." Representative fluorographs are shown. Eight to 10 oocytes were used in each group. c-myc-␤ G , ␤-subunit of the gastric H ϩ ,K ϩ -ATPase with the human c-myc epitope at the amino terminus; c-myc-␤ 1 , ␤ 1 -subunit of the Na ϩ ,K ϩ -ATPase with the human c-myc epitope at the amino terminus; cg␤ G , core glycosylated c-myc-␤ G ; cg␤ 1 core-glycosylated c-myc-␤ 1 ; fg␤ G , fully glycosylated c-myc-␤ G ; fg␤ 1 , fully glycosylated c-myc-␤ 1 ; ␤ G , expected position of c-myc-␤ G core protein; ␤ 1 , expected position of cmyc-␤ 1 core protein.
ponent of K ϩ -ATPase activity, and that a novel X ϩ ,K ϩ -ATPase ␣-subunit, bearing an amino terminus antigenically similar to ␣ C , mediates the more ouabain-sensitive fraction of Na ϩ -independent K ϩ -ATPase activity. Given the fact that the ouabain sensitivities of ␣ C ␤ G and ␣ C ␤ 1 were virtually indistinguishable, it seems less likely that a novel ␤-subunit could confer ouabain sensitivity on ␣ C . If ␣ C indeed interacts with ␤ 1 in vivo, it is logical to predict that these two subunits will be coordinately up-regulated in the colonocyte apical membrane during chronic K ϩ deprivation. Studies are presently under way in our laboratory to test this hypothesis.
In conclusion, the present studies show that both ␤ G and ␤ 1 can interact with ␣ C , each creating a functional H ϩ ,K ϩ -ATPase that is Sch-28080-insensitive and only partially sensitive to ouabain. The functional and pharmacological properties of ␣ C ␤ G and ␣ C ␤ 1 holoenzymes are quite similar. Our data support the view that, in the rat distal colon, ␤ 1 may play a surrogate role as the ␤-subunit for the colonic H ϩ ,K ϩ -ATPase.