ATP1AL1, a Member of the Non-gastric H,K-ATPase Family, Functions as a Sodium Pump*

The human ATP1AL1-encoded protein (an α subunit of the human non-gastric H,K-ATPase) has previously been shown to assemble with the gastric H,K-ATPase β subunit (gH,Kβ) to form a functionally active ionic pump in HEK 293 cells. This pump has been found to be sensitive to both SCH 28080 and ouabain. However, the86Rb+-influx mediated by the ATP1AL1-gH,Kβ heterodimer in HEK 293 cells is at least 1 order of magnitude larger than the maximum ouabain-sensitive proton efflux detected in the same cells. In this study we find that the intracellular Na+ content in cells expressing ATP1AL1 and gH,Kβ is two times lower than that in control HEK 293 cells in response to incubation for 3 h in the presence of 1 μm ouabain. Moreover, analysis of net Na+efflux in HEK 293 expressing the ATP1AL1-gH,Kβ heterodimer reveals the presence of Na+ extrusion activity that is not sensitive to 1 μm ouabain but can be inhibited by 1 mm of this drug. In contrast, ouabain-inhibitable Na+ efflux in control HEK 293 cells is similarly sensitive to either 1 μm or 1 mm ouabain. Finally,86Rb+ influx through the ATP1AL1-gH,Kβ complex is comparable to the 1 mm ouabain-sensitive Na+ efflux in the same cells. The data presented here suggest that the enzyme formed by ATP1AL1 and the gastric H,K-ATPase β subunit in HEK 293 cells mediates primarily Na+,K+ rather than H+,K+ exchange.

The human ATP1AL1 gene encodes an ␣ subunit of a nongastric H,K-ATPase (1). The non-gastric H,K-ATPases form a distinctive subset within a larger family of P-type potassiumdependent ion-transporting ATPases (ion pumps) that drive the import of K ϩ ions in exchange for Na ϩ ions (Na,K-ATPase) or for protons (H,K-ATPase) (2)(3)(4). All of these pumps are heterodimers, composed of a large polytopic ␣ subunit (ϳ110 kDa) predicted to span the membrane 10 times and a heavily glycosylated ␤ subunit (core protein is ϳ35 kDa) that crosses the bilayer once in a type II orientation.
The cDNAs encoding three non-gastric H,K-ATPase ␣ subunits have been isolated and characterized. These include the colonic H,K-ATPase from rat (5), guinea pig (GenBank TM and EMBL data bank accession number D21854), and rabbit (6), the ␣ subunit of the H,K-ATPase from toad bladder epithelium (7), and the protein encoded by the human ATP1AL1 gene (8). These proteins manifest ϳ75-85% identity to one another, and each of them is equally distant from both the Na,K-and H,K-ATPase ␣-polypeptides (ϳ65% identity). A cDNA encoding a ␤ subunit polypeptide has also been isolated from a library prepared from toad bladder (9). This protein is ϳ35% identical to the Na,K-ATPase and the gastric H,K-ATPase ␤-polypeptides. Although no additional non-gastric H,K-ATPase ␤ subunits have yet been isolated from any mammalian species, it has been clearly demonstrated that the Na,K-ATPase and the gastric H,K-ATPase ␤-subunits (gH,K␤) 1 can assemble with the non-gastric H,K-ATPase ␣-polypeptides either when these proteins are co-expressed in heterologous expression systems (1, 10 -12) or in vivo (13,14).
In Xenopus oocyte expression studies, each of the non-gastric H,K-ATPase ␣ subunit proteins, including ATP1AL1, have been shown to catalyze K ϩ (Rb ϩ ) influx (1, 8, 10 -12). For the toad bladder enzyme and ATP1AL1, these fluxes are moderately sensitive to both ouabain and SCH 28080 (1,7,10), whereas the colonic pump was inhibitable only by ouabain (11,12). The pump-expressing oocytes could acidify the extracellular milieu in an extracellular K ϩ -dependent fashion. Moreover, this acidification could be inhibited by both ouabain and SCH 28080 in a manner similar to that found in 86 Rb ϩ -uptake experiments (7,10,11). However, measurements quantitating these proton effluxes were not possible in the oocyte studies.
We have previously demonstrated that HEK 293 cells expressing both the ATP1AL1-encoded protein and the gastric H,K-ATPase ␤ subunit exhibit 86 Rb ϩ uptake activity which is different from that mediated by the endogenous Na,K-ATPase and which is sensitive to both SCH 28080 and to ouabain (K i values ϳ131 and 42 M, respectively) (1). In addition, these cells exhibit acid extrusion activity (during recovery of intracellular pH after acid loading through an NH 4 ϩ pulse) that is independent of extracellular sodium and that can be significantly inhibited by 1 mM ouabain. Comparison of the Rb ϩ and proton fluxes we measure demonstrates that Rb ϩ influx mediated by this pump is 10 times larger than the proton efflux catalyzed by the same enzyme. Moreover, while the acid extrusion activity mediated by the pump shows a marked pH dependence, the 86 Rb ϩ uptake activity present in the cells expressing the ATP1AL1-gH,K␤ complex is not stimulated by the acute lowering of intracellular pH. These data suggest, therefore, that the enzyme formed by the ATP1AL1-encoded protein and the gastric H,K-ATPase ␤ subunit might exchange potassium for another cation.
Our recent results provide further support for the concept that ATP1AL1 is not an H,K-ATPase but rather acts as an Na,K-ATPase. In our previous study we found that coexpression of ATP1AL1 and gH,K␤ in HEK 293 cells conferred upon these cells an ability to grow in the presence of 1 M ouabain. None of the untransfected cells, nor cells singly transfected with only one of the pump subunit cDNAs, could grow in the presence of ouabain. More detailed analysis reveals that the intracellular Na ϩ content in the cells expressing ATP1AL1 and gH,K␤ is two times lower than that in control HEK 293 cells after incubation for 3 h in medium containing 1 M ouabain. In contrast, there is no difference between transfected and control cells in media with no or with 1 mM ouabain. Further study indicates that a Na ϩ extrusion activity sensitive to 1 mM but not to 1 M ouabain is present only in HEK 293 cells expressing the ATP1AL1-gH,K␤ heterodimer. Moreover, 86 Rb ϩ influx mediated by the ATP1AL1-gH,K␤ complex measured under these conditions is comparable to the 1 mM ouabain-sensitive Na ϩ efflux in the same cells. These observations suggest that the ATP1AL1-gH,K␤ enzyme expressed in HEK 293 cells mediates primarily Na ϩ ,K ϩ rather than H ϩ ,K ϩ exchange.

EXPERIMENTAL PROCEDURES
Transfection and Cell Culture-All plasmid constructions and stably transfected lines of HEK 293 cells have been previously obtained and characterized (1). Cell culture procedures were performed as before (1). Ouabain dependence of the cell growth was tested as described (15) with modifications. The cells were seeded on a 24-well plate (5⅐10 4 cells per well, to allow the cells to reach almost 100% confluency in 4 days under normal conditions). On the following day, cell media were replaced with the same media containing different concentrations of ouabain (two wells were analyzed for each concentration of ouabain). After an additional 3 days of growth in the presence or absence of ouabain, cells were washed twice with phosphate-buffered saline (150 mM NaCl, 10 mM Na 2 HPO 4 /NaH 2 PO 4 , pH 7.4) at room temperature and lysed with 1% CHAPS at 4°C for at least 1 h. The cell protein content was determined using BCA Protein Assay (Pierce) (16).
Measurements of Intracellular Sodium and Potassium Contents-Cells were subcultured onto a 6-well plate and grown to ϳ80 -90% confluency (2-3 days). They were treated with 10 mM sodium butyrate in the media for 16 -24 h prior to the experiment. On the day of the experiment, the media were replaced with the same containing 0, 1 M, or 1 mM ouabain (two wells for each concentration of ouabain) plus or minus 1 mM amiloride or 200 M bumetanide, and cells were incubated for an additional 3 h in a humidified incubator at 37°C under 5% CO 2 atmosphere (10 mM of sodium butyrate was present throughout the experiment as well). After this period, cells were washed three times with ice-cold Na ϩ -free isotonic solution containing 160 mM N-methyl-D-glucammonium (NMDG) chloride, pH 7.4, and lysed in 2 ml of 1% CHAPS solution as above. Concentrations of sodium and potassium in the cell lysate were measured by means of flame photometry on a 400 Flame Photometer (Corning, New York). The cell lysate protein concentration was determined as in the previous paragraph. Finally, the intracellular sodium and potassium content was expressed as nanomoles per g of total cell protein.
Na ϩ Efflux Measurements-Experiments were performed on the cells prepared in the same manner as for the steady state experiments above. On the day of the experiment, cells were preloaded with sodium by washing three times (5 ml per well) with potassium-free solution ("STD/ OK" solution, 145 mM NaCl, 1 mM CaCl 2 , 1.2 mM MgSO 4 , 2 mM NaH 2 PO 4 , 32 mM HEPES, pH 7.4, at 37°C) plus or minus ouabain, and incubated in the last wash for 1 h at the same temperature (all of the following steps were performed at 37°C as well). One-hour incubation of these cells in STD/OK solution is enough for the intracellular sodium concentration to reach a plateau at ϳ100 mM. 2 After the sodium load period, Na ϩ efflux was initiated by replacing STD/OK solution with 2 ml per well of STD/ONa buffer (which is STD/OK buffer containing 5 mM KCl and with NaCl substituted by an equimolar concentration of NMDGCl). After incubation for different times, sodium efflux was terminated by rinsing cells twice with ice-cold isotonic NMDGCl solution. Cells were lysed in 1% CHAPS, and intracellular sodium content was determined as described above.
86 Rb ϩ Uptake Measurements-86 Rb ϩ uptake was performed as described previously (1) except that the cells were preloaded with Na ϩ by incubating in STD/OK solution at 37°C with or without ouabain for 1 h. 86 Rb ϩ uptake was initiated by switching the incubation solution from STD/OK to STD/ONa (with 5 mM RbCl instead of KCl). The uptake was terminated after incubation at 10 min at 37°C by washing 6 times with the ice-cold isotonic NMDGCl solution. Cells were lysed in 2% CHAPS as above and analyzed (1).
Statistics-Unless noted, data are reported as the mean Ϯ S.E. We used the unpaired Student t test to assess levels of significance. A p value less than 0.05 is considered significant. The Na ϩ efflux data were fitted by the Igor graphics and data analysis program (WaveMetrics, Lake Oswego, OR), which exploits the Levenberg-Marquardt algorithm (17). The intracellular sodium content at different time points ([Na ϩ i ] t ) is presented as a fraction of the 0 time point ([Na ϩ i ] 0 ) and plotted versus time. Each time course was fitted to a single exponential as ([ , where values of A, B, and k Na are obtained from the fitting and correspond to ([ , and a rate constant (min Ϫ1 ), respectively ([Na ϩ i ] ϱ Ϫ is the intracellular sodium content at time infinity).

RESULTS
In our previous studies (1), we found that the ouabain-sensitive 86 Rb ϩ influx mediated by the ATP1AL1-gH,K␤ heterodimer in HEK 293 cells was 10 times larger than the ouabain-sensitive proton efflux in the same cells (ϳ0.1 mmol H ϩ ⅐l Ϫ1 ⅐min Ϫ1 versus ϳ(0.8 -1.6) mmol Rb ϩ ⅐l Ϫ1 ⅐min Ϫ1 ). In addition, whereas the acid extrusion activity was pH i -dependent with the maximum at pH i ϳ6.3, the 86 Rb ϩ uptake activity appeared not be affected by acute lowering of intracellular pH (1). It is likely, therefore, that the ATP1AL1-gH,K␤ complex mediates the exchange of potassium primarily for a cation other than protons. We speculated that Na ϩ serves as the counter ion for the transport catalyzed by ATP1AL1-gH,K␤. This possibility is consistent with our finding that this pump can substitute for the Na ϩ ,K ϩ -dependent ATPase in maintaining intracellular ionic homeostasis. Stably transfected HEK 293 cell lines expressing both ATP1AL1 and the gastric H,K-ATPase ␤ subunit can grow in media containing 1 M ouabain, whereas wild type HEK 293 cells, HEK 293 cells transfected with empty expression vector, or cells expressing only one of the subunit proteins are highly sensitive to this concentration (1). A more detailed characterization of this difference in ouabain resistance has been obtained from measurements of ouabain-inhibitable cell growth, which are presented in Fig. 1. Wild type HEK 293 cells, HEK 293 transfected with the expression vector alone (HEK NEO) and clone 50 (expressing ATP1AL1 alone) demonstrate growth inhibition by ouabain which exhibits a single low K i of about 200 nM, reflecting the high affinity for ouabain of the endogenous Na,K-ATPase (1). In contrast, three cell lines expressing the ATP1AL1-gH,K␤ heterodimer (cell lines 119, 152, 114) clearly exhibit a second inhibition constant of about 50 M, which is in good agreement with the ouabain sensitivity of the 86 Rb ϩ uptake mediated by the ATP1AL1-gH,K␤ complex (K i ϳ 42 M (1)). It should be noted that the isolation of ouabain-resistant cells expressing both ATP1AL1 and gH,K␤ is a common event. Four lines of transfected HEK 293 cells expressing both subunits were initially isolated and selected by means of immunofluorescence analysis. Three of these cell lines demonstrate a substantial cell growth at concentrations of ouabain equal to or higher than 1 M (Fig. 1). In addition, a small fraction of the cells (ϳ7%) from the fourth cell line (data not shown) expressing the active pump complex can also survive at concentrations of ouabain greater than 1 M.
To elucidate further the mechanism underlying the ouabain resistance in 119 cells, we studied the effect of the selective inhibition of the endogenous Na,K-ATPase and the ATP1AL1-gH,K␤ enzyme by 1 M and 1 mM ouabain, respectively, on the steady state levels of intracellular sodium (Na i ) and potassium (K i ) contents (determined by means of flame photometry, see "Experimental Procedures"). As can be seen in Fig. 2A, the presence of either 1 M or 1 mM ouabain is sufficient to maxi-2 R. Zahler, personal communication.
mally inhibit the endogenous Na,K-ATPase in HEK 293 cells and both produce a ϳ6.7-fold increase in Na i in these cells. In 119 cells, a similar ϳ5.5-fold elevation in Na i is obtained with 1 mM ouabain, whereas incubation with 1 M ouabain leads to only a ϳ3.0-fold increase. Intracellular potassium content changes in the opposite manner under the same conditions (Fig. 2B). Neither bumetanide (200 M) nor amiloride (1 mM) appears to alter the intracellular sodium or potassium contents in response to the different ouabain concentrations in the media. This observation rules out any involvement of either the Na ϩ , K ϩ , Cl Ϫ cotransporter or the Na ϩ /H ϩ exchanger in maintenance of low Na i in 119 cells. Since the ATP1AL1-gH,K␤ complex is fully inhibited by 1 mM ouabain and only partially blocked by a 1 M concentration of the drug, these data suggest that Na ϩ extrusion directly catalyzed by this pump could account for the observed differences in the intracellular sodium content.
We also measured sodium efflux in 119 and HEK 293 cells directly by determining the Na i recovery time course following a sodium load (as described under "Experimental Procedures"). Data from one representative experiment (taken from three independent trials) performed on each cell type are depicted in Fig. 3 (A and B). The data are presented as a time course of changes in the ratio [Na ϩ i ] t /[Na ϩ i ] 0 (where [Na ϩ i ] t and [Na ϩ i ] 0 represent the intracellular sodium contents at different time points and at the 0 time point, respectively). As can be seen in Fig. 3A, the rate of the sodium efflux in 119 cells is differentially sensitive to the presence of 1 M or 1 mM ouabain. In contrast, 1 M ouabain inhibits Na i recovery rate to the same extent as 1 mM ouabain in untransfected HEK 293 cells (Fig.  3B). The same conclusion can be drawn from a comparison of the rate constants for Na ϩ efflux (k Na ) obtained from the fitting procedure (see "Experimental Procedures" and Table I). Thus, in 119 cells k Na in the presence of 1 mM ouabain is ϳ1.86 lower than that in the presence of 1 M of the drug (0.1738 Ϯ 0.0129 min Ϫ1 versus 0.3239 Ϯ 0.0479 min Ϫ1 , p Ͻ 0.02), which in turn is significantly lower (in ϳ1.5 times) than the rate constant in the absence of ouabain (0.3239 Ϯ 0.0479 min Ϫ1 versus 0.486 Ϯ 0.0386 min Ϫ1 , p Ͻ 0.05). On the other hand, in HEK 293 cells, k Na in the presence of either 1 M or 1 mM ouabain is ϳ4.2 times less than that in the absence of the drug (Table I, p Ͻ 0.005), and there is no statistically significant difference between effect of these two different concentrations of ouabain on the rate constant (Table I, p Ͼ 0.4). Therefore, there is a sodium extruding activity in 119 cells that is absent in the control HEK 293 cells and that demonstrates sensitivity to ouabain similar to that of the 86 Rb ϩ uptake activity mediated by the ATP1AL1-gH,K␤ heterodimer in the same cells. Finally, we measured the 86 Rb ϩ uptake mediated by the ATP1AL1-gH,K␤ heterodimer in 119 cells under the sodium pre-loading conditions employed in the Na ϩ efflux experiment. We find that these two activities are of comparable magnitude, ϳ(0.099 Ϯ 0.046) nmol of Na ϩ ⅐g of cell protein Ϫ1 ⅐10 min Ϫ1 versus ϳ(0.195 Ϯ 0.024) nmol Rb ϩ ⅐g of cell protein Ϫ1 ⅐10 min Ϫ1 , for the Na ϩ efflux and 86 Rb ϩ uptake, respectively (Table II). As we noted above, we have previously shown that the rate of proton extrusion is at most 1/10 of the Rb ϩ uptake driven by ATP1AL1-gH,K␤ expressed in transfected HEK 293 cells (9). Thus, while it is clear that ATP1AL1 can mediate the efflux of protons in exchange for K ϩ , our data demonstrate that Na ϩ ,K ϩ exchange is the predominant transport catalyzed by this pump.  3. Na ؉ efflux in 119 and HEK 293 cells. Cells were grown on 6-well plates for 2-3 days to 80 -100% confluency and were treated with 10 mM sodium butyrate for 18 -24 h overnight before the experiment. On the day of the experiment, cells were loaded with sodium by incubation cells resistance to ouabain at concentrations that completely inhibit the growth of the cells expressing only the highly ouabain-sensitive endogenous Na,K-ATPase. The conclusion that ATP1AL1 functions predominantly as an Na,K-ATPase under physiological conditions would also explain the 1 order of magnitude difference between the 86 Rb ϩ uptake mediated by the complex and the proton efflux through the same pump. Similarly, the observation that ATP1AL1-gH,K␤-catalyzed 86 Rb ϩ uptake appears to be insensitive to decreases in intracellular pH whereas the proton extruding activity demonstrates a marked stimulation by lowering pH i (1) is consistent with the model.
Recently, Cougnon et al. (18) have found that the steady state level of intracellular Na ϩ is approximately two times lower in Xenopus oocytes expressing a combination of the colonic H,K-ATPase ␣ subunit with the gastric H,K-ATPase ␤ subunit than in oocytes expressing gH,K␤ alone or the gastric H,K-ATPase ␣ and ␤ subunits together. A similar change is noted in oocytes co-expressing the toad Na,K-ATPase ␣ and ␤ subunits. The decrease in intracellular Na ϩ in Xenopus oocytes expressing colonic H,K-ATPase was also observed when diffusive Na ϩ influx was elevated by expression of the functional amiloride-sensitive epithelial Na ϩ channel. This effect is dependent on the presence of potassium in the extracellular milieu and could be inhibited by 2 mM ouabain, a concentration known to inhibit the colonic H,K-ATPase (11,12). This observation suggests that the colonic H,K-ATPase is also likely to be capable of active outward Na ϩ transport and lends additional support for our interpretation of the data presented here.
Further experiments need to be performed to characterize in detail the (Na ϩ ,H ϩ ),K ϩ exchange mediated by the ATP1AL1-gH,K␤ pump. Whereas the data presented in this paper demonstrate that the pump-catalyzed Rb ϩ and Na ϩ fluxes are of similar magnitudes, it will be important to measure the cation stoichiometry directly and to determine its dependence upon intracellular pH. It will also be necessary to measure the ATPase activity of the pump and actual affinity for sodium. We have attempted to measure the ATPase activity of ATP1AL1 expressed in HEK 293 cells. Unfortunately, we have not yet succeeded in detecting K ϩ -dependent ATPase activity in 119 cells that is insensitive to 1 M ouabain but is inhibited by 1 mM concentration of this drug. Presumably the membrane isolation or permeabilization conditions applied to date have not maintained ATP1AL1 function. Future experiments will involve further efforts to identify assay conditions in which ATP1AL1driven ATP hydrolysis can be quantitated.
The physiological relevance of the Na ϩ ,K ϩ exchange mediated by the ATP1AL1-gH,K␤ pump and by the colonic H,K-ATPase is not clear at present. Several laboratories have reported that sodium is involved in regulating the activity of pump-driven K ϩ absorption in the renal distal tubule and colon. Thus, earlier studies demonstrated that an Na,K-ATPaselike activity that appeared not to be aldosterone-regulated is present in the medullary collecting ducts of rats fed a low potassium diet (19,20) or from rats with furosemide-induced potassium wasting (21). This activity was proposed to be located in the apical membrane of medullary collecting duct cells (19,20).
Other investigators have found that Na ϩ not only reduces K ϩ absorption through electroneutral, SCH 28080-sensitive H,K-ATPase in the perfused cortical collecting duct of K ϩdepleted rabbits (22,23) but can apparently be transported from the lumen to bath by the same pump when the lumen K ϩ concentration was low (23). Recently, elegant biochemical experiments have revealed that there are at least three distinguishable K-ATPase activities distinct from that of Na,K-ATPase present at several sites along the rat nephron (24). One of these, referred to as the type III K-ATPase activity, can be equally stimulated by either Na ϩ or K ϩ . Potassium depletion appears to boost levels of expression of this ATPase activity. K ϩ depletion also leads to an increase in the expression of the colonic H,K-ATPase ␣ subunit both at the mRNA and protein levels in the cortical and/or outer medullary collecting ducts (25)(26)(27)(28)(29). It is not clear, however, if colonic H,K-ATPase and/or ATP1AL1 could possibly account for type III ATPase activity, since the pharmacological profile of the colonic H,K-ATPase and ATP1AL1 determined in expression studies is dramatically different from that of the type III ATPase. Although the type III ATPase is sensitive to ouabain with K i of ϳ20 M and to SCH 28080 with K i of ϳ1 M (24), ATP1AL1 is inhibited by ouabain and SCH 28080 with K i values of ϳ40 and ϳ130 M Rb ϩ uptake and Na ϩ efflux (nmol ⅐ g of cell protein Ϫ1 ⅐ 10 min Ϫ1 ) sensitive to either 1 M or 1 mM ouabain in HEK 293 and 119 cells were computed as the difference between total fluxes at 10 min in the presence of 0, 1 M, or 1 mM of the drug. 86 Rb ϩ uptake was determined as described under "Experimental Procedures" and Na ϩ efflux was calculated as (Na ϩ to Na ϩ i ) where Na ϩ 0 and Na ϩ i are the intracellular sodium content at 0 and 10 min, respectively. Each value represents mean Ϯ S.E. of three independent experiments.  (1), respectively, and the colonic H,K-ATPase exhibits a K i for ouabain of ϳ(0.5-1) mM and no susceptibility to SCH 28080 (11,12). Measurements of ATP1AL1-catalyzed ATP hydrolysis will be required to determine whether Na ϩ can substitute for K ϩ as it does in the renal type III ATPase activity.
The vanadate-sensitive potassium absorption in the rat distal colon appears to include both Na ϩ -sensitive, ouabain-insensitive as well as Na ϩ -insensitive, ouabain-sensitive, components (30). Since both the expression of the colonic H,K-ATPase (25,26) and the activity of the sodium-sensitive component of the vanadate-sensitive K ϩ absorption activity (30) are elevated in the sodium-restricted rats, it has been proposed that the colonic H,K-ATPase is responsible for the Na ϩ -sensitive ouabain-insensitive component of colonic K ϩ restriction (26). Two types of acid extruding activity, Na ϩ -dependent and Na ϩ -independent, have also been detected at the lumen side of the rat colon (31). Both of these activities can be significantly inhibited by either ouabain or orthovanadate, but only the Na ϩ -independent component is dependent on the presence of mucosal K ϩ . Sodium depletion has been shown to stimulate the Na ϩindependent fraction of this acid extruding activity in the rat colon without affecting the Na ϩ -dependent component. The K-ATPase activities detected in the colon, however, do not exhibit a dependence on the presence of Na ϩ (32-34). Thus, whether a simple concordance can be drawn to relate the (Na ϩ ,H ϩ ),K ϩ exchange mediated by the ATP1AL1-gH,K␤ pump and colonic H,K-ATPase to this complex pattern of activities involved in K ϩ absorption in the nephron and in the colon remains to be seen.
Both the Na,K-ATPase and the gastric H,K-ATPase can exhibit mixed ion specificities in certain circumstances but not under normal physiological conditions (35)(36)(37). Comparison of the sequences of the Na,K-and gastric H,K-ATPase fourth transmembrane segments (TM4) reveals that they differ at 7 of 28 positions. It is interesting to note that mutagenesis studies performed on the Na,K-ATPase ␣ subunit demonstrate that Asn-331, 3 in the fourth putative transmembrane domain, plays a critical role in determining the sodium affinity of the pump (40). Recent chimera studies in our laboratory 4 indicate that the TM4 residues play an important role in establishing distinct cation selectivities of these pumps. Substitution of the Na,K-ATPase fourth transmembrane domain residues with their gastric H,K-ATPase counterparts (Leu-324, Asn-331, Thr-345, Phe-333, Tyr-340, and Ser-354, respectively) produces an (H,Na),K-ATPase activity that may preferentially function as an H,K-ATPase at low pH. It is interesting to note that the sequences of the ATP1AL1 and colonic H,K-ATPase TM4s differ from that of the Na,K-ATPase by only one residue. In the positions of the residues that appear to be important for proton transport in the gastric H ϩ ,K ϩ pump, ATP1AL1 and the colonic H,K-ATPase possess the same residues found in the comparable positions of the sodium pump. The conservation of Na,K-ATPase sequence in the TM4s of ATP1AL1 and the colonic H,K-ATPase may account, at least in part, for the Na,K-ATPase-like behavior of these pumps. Mutagenesis and chimera experiments employing ATP1AL1, the gastric H,K-ATPase, and the Na,K-ATPase are under way to define further the structural basis for the cation transport properties of these pumps.
In conclusion, the data presented here demonstrate that the ATP1AL1 co-expressed with the gastric H,K-ATPase ␤ subunit in HEK 293 cells functions to mediate predominantly Na ϩ ,K ϩ exchange. The kinetic properties of the (Na ϩ ,H ϩ ),K ϩ exchange driven by the ATP1AL1-gH,K␤ pump as well as it's physiologic role in the maintenance of ionic homeostasis remain to be elucidated.