C1- Transport in Gastric Microsomes AN ATP-DEPENDENT INFLUX SENSITIVE TO MEMBRANE POTENTIAL AND TO PROTEIN KINASE INHIBITOR*

Uptakes of radioactive C1- or I- by gastric micro- somal vesicles were stimulated 2- to &fold by ATP. The sensitivity of those uptakes to a C1- $ OH- ionophore and to osmotic swelling suggested they were due to transport rather than to binding. The ATP effect was labile, but dithiothreitol and methanol improved its stability. The stimulation of anion transport required magnesium; GTP and UTP were less potent than ATP whereas ADP and AMP had no effect. The apparent K, for ATP was estimated to be 2 X 10 -4 M at 22°C. The rate of the ATP-dependent transport showed satura-tion-type kinetics, with half-maximal uptake at 10 m~ for I- and 15 m~ for C1-. Nonradioactive C1-, I-, and SCN- competed with ‘’‘I- uptake while Sod’- did not. K’ valinomycin increased the ATP-dependent C1- uptake. The thermostable inhibitor of CAMP-dependent protein kinases inhibited the effect of ATP. These results suggest the existence of an anion conductance, permeant to C1-, I-, and SCN- and nonpermeant to SO4’-, which could be linked to a CAMP-dependent protein kinase. Gastric mucosa secretes HC1 down to pH 1.

Gastric mucosa secretes HC1 down to pH 1. The mechanisms responsible for this secretion are still debated, but it is becoming clear that H' and C1-are secreted by two distinct processes (l), although C1-appears obligatory for H' secretion to occur (2, 3). To account for the secretion of H', the involvement of (H+K+)-ATPase was proposed (4-6). This ATPase was suggested to be different from both the mitochondrial F1 ATPase and the plasma membrane (Na+K')-ATPase because of its lack of sensitivity to specific inhibitors such as oligomycin and ouabain, respectively. It was shown to catalyze a neutral exchange of H' against K' in purified gastric vesicles (1, 5), thus suggesting that it could possibly secrete H' in the gastric lumen in uiuo.
Mechanisms of gastric Clsecretion were mainly approached using isolated gastric mucosa mounted in Ussing chambers (2, 7, 8). Studies of ion fluxes and of transmucosal electrical parameters showed that C1-movement from serosa to lumen could be analyzed into two components: a neutral transport and an electrogenic one. On the basis of indirect evidence, the neutral transport component was suggested to be linked to H' secretion (1). To account for this link, a K-C1 channel has been suggested (1, 9). Thus, K+ trapped by the cell via the (H+K+)-ATPase would be driven back to the gastric lumen by neutral co-transport of K' and C1-.
Electrogenic C1-secretion appeared to be a widely spread * 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. phenomenon. It was described in the intestine (lo), the rectum (ll), the cornea (12), the nerves (13), and the gdls (14). In most of those tissues, C1-transport was hormone-sensitive. Furthermore, cyclic AMP and Ca2+ ionophore could mimic the effect of hormones (15)(16)(17). In none of those tissues, however, are the membranous mechanisms responsible for this C1transport elucidated. In the gastric mucosa, histamine stimulates electrogenic C1-secretion (8) and the activity of adenylate cycIase (18,19) and CAMP-dependent protein kinases (20). However, the implication of CAMP and protein kinases in the mediation of the secretory response has not been established. C1-secretion has been shown to depend upon cell metabolism and, more specifically, upon ATP synthesis (21,22).
In previous papers (23, 24) we reported evidence for an ATP-stimulated C1-transport in gastric microsomal vesicles. This ATP-stimulated C1-transport was not inhibited by SITS,' the C1-s HC03-channel inhibitor. Furthermore, it was sensitive neither to rutamycin nor to ouabain and vanadate which suggested it was not dependent on the activity of F1 or (Na+K')-ATPase. The purpose of this work was to go deeper into the characterization of this microsomal ATPdependent anion transport. We report its sensitivity to membrane potential, and furthermore, show that its ATP dependency could be due to the involvement of CAMP-dependent protein kinase($. was purified from the rat brain as previously described (25) up to the stage of the DEAE-cellulose column. It was provided at a concentra-The abbreviations used are: SITS, 4'-acetamido-4'-isothiocyano-2,2'-disulfonic acid stilbene; PEP, phosphoenolpyruvate; Hepes, 442-hydroxyethy1)-I-piperazineethanesulfonic acid; DIDS, 4,4'-diisothiocyano-2,2'-disulfonic acid stilbene.
tion of 410 pg of protein/ml in 5 mM sodium glycerophosphate, 0.2 mM EDTA, pH 7.0. The activity of the inhibitor was determined by Dr. P. Mangeat, wing a purified preparation of pancreatic protein kinase (26). A unit of inhibitor was defined as the quantity which inhibits the transfer by the purified protein kinase of 1 pmol of phosphate to protein (1 mg/ml of histone) in 1 min at 30°C.

Dowex Columns-Commercially available chloride Dowex 1-X8
was regenerated to ita formate form. This was realized by successive washings of the resin 1) with 1 N NaOH (10 liters/500 g of resin), 2) with distilled water until no trace of free C1-was detected in the rinsing water (precipitation with AgN03), 3) with 1 N formic acid (1 liter/500 g of resin), and 4) with distilled water until the pH of the rinsing water reached pH 4. The regenerated resin was stable up to 4 months at 4'C. Columns were set up by filling Pasteur pipets (10 X 0.55 cm) with regenerated Dowex 1-X8 (27). Before the test, each column was washed with 0.5 ml of 10% (w/v) bovine serum albumin to prevent protein adsorption and rinsed with 2 ml of the buffered sucrose solution containing 0.1% bovine serum albumin. So treated, the columns could be used four times and they retained as much as 99.9% of the radioactive anion while they allowed a good recovery of the proteins (Table I).

Preparation of Microsomal
Fractions-Microsomes were prepared from hog, rabbit, and human fundic stomachs. Fresh hog stomachs were kindly provided by Olida's slaughtering house (Levallois, France). Rabbit stomachs were taken from albino rabbits killed with nembutal. Human stomachs were fresh pieces of surgery. They were opened and washed with fresh tap water. Mucous was removed by rubbing the mucosa with a glass slide. Fundic mucosa was separated from the subtissue and cut into small pieces. These were diluted in sucrose solution at 4°C and homogenized (Ultraturax homogenizer) for 3 min at 4°C. Unhomogenized tissues were discarded by filtration on gauze tissue. The homogenate was then sequentially centrifuged at 4'C to separate microsomal fractions: 1) the nuclear pellets were discarded after centrifugation (6 min X 1,000 rpm) and after washing the pellets once in sucrose solution (International Equipment Co. PR-6000 centrifuge, rotor 269); 2) the mitochondrial pellets were discarded after centrifugation (7 min X 25,000 rpm) (Beckman L5-65 centrifuge, rotor 70 Ti); and 3) microsomes were pelleted by centrifugation (30 min X 39,000 rpm) and finally resuspended in sucrose solution at a concentration of 15 to 35 mg of protein/ml. Methanol was then added at a fmal concentration of 2.5 to 5% to improve the conservation.
2. Protein Determination-Proteins were estimated according to Lowry et al. (28) using bovine serum albumin as standard.
3. Anion Uptake Measurements-Anion uptake was measured as the radioactivity eluted from the Dowex columns prepared as reported under "Materials." Incubations were started by adding microsomes TABLE I Study ofprotein filtration recovery and of anion absorption capacity of the Dowex columns Samples were prepared as were those for anion uptake study (see "Experimental Procedures"). For the estimations of protein recovery, radioactive iodine was omitted and for those of microsomal proteins were omitted to prevent anion trapping and were replaced by 10% bovine serum albumin. Samples of 100 pl were eluted with 2 ml of sucrose solution. Each column was tested four successive times as in the standard assays. The values are the ratio (in per cent) of the values of the protein radioactivity eluted from the columns to the values of those of the samples introduced on the columns. Mean of NaCl and 2 to 10 x IO6 cpm of lz5I-or 36Cl-. The effect of ATP was tested by adding 1 to 3 m~ ATP, 1 to 3 mM MgS04,4 to 16 mM PEP, and 2 to 4 units of pyruvate kinase or, in the controls, 10 to 30 mM glucose, and 1 unit of hexokinase. Incubations were run for 3 min at 20-30°C, and 3 samples of 100 pl each were applied to columns and eluted with 2 ml of sucrose solution containing 0.1% bovine serum albumin. The whole volume of the eluates was counted for the determination of trapped radioactivity. 4. ATPase Activity-ATPase activity was measured in the presence of 1 m ATP, 1 m~ MgS04, 10 p g / d of oligomycin, 1 mM ouabain, 4 mM PEP, and 2 units/ml of pyruvate kinase. The incubations were run for 20 min at 30°C at pH 8.0 (50 m~ Tris/acetate). Inorganic phosphate was measured as described (29). 2. Uptake of Anions in the Presence of ATP ATP increased both the rate and the apparent extent of C1-trapping by gastric microsomes of rabbit, hog, and man. It had no effect on C1-uptake by liver microsomes similarly prepared (one experiment). The stimulation of the rate of C1influx was 2to 4-fold over the basal rate. ATP also stimulated 3-to 8-fold the rate of lZ51-uptake (Fig. 1). The ATP-stimulated uptakes (i.e. uptakes in the presence of ATP minus control) were proportional to the concentration of microsomal protein (Fig. 2). By contrast, ATP did not stimulate the rate of lz5Iefflux from hog microsomes. (Effluxes were measured using vesicles loaded with radioactive anion for 24 to 72 h at Effect of Ageing-Ageing the microsomes in 0.25 M sucrose, 50 m Hepes (pH 7.3) at 0-4°C for 1 day resulted in a full loss of the ATP effect. Conservation at 0-4°C was greatly improved by 1 to 2 m dithiothreitol and, in a less obligatory way, by methanol (2.5 to 5% final concentration). In those conditions, ATP stimulation was maintained up to a week.

Basal Uptake
Characterization of the Uptake-I-and C1-uptakes were assumed to represent transport events because 1) they were reversed by 5 X to M of the C1-e OH-ionophore, tributyltin (30) and 2) the magnitude of the ATP-induced uptakes correlated the volume of the microsomal vesicles as monitored by the osmolarity of the suspension medium (Fig.  3). Furthermore, uptakes extrapolated almost to zero for infinite osmolarity suggesting a low contribution of adsorption phenomena. ATP Specificity-ATP stimulation required Mg". GTP and UTP could partially mimic the ATP effect, while AMP and ADP, i.e. ATP in the presence of glucose and hexokinase, had no effect ( Table 11). The apparent K , for ATP was estimated to be 1.  and Iwere not affected by the temperature up to 30°C. At higher temperatures, transport capacity progressively decreased. This was likely due to alteration of vesicle permeability.
p H Influence-Maximal ATP effect was obtained at pH 7.8 which also corresponded to minimal basal transport (Fig. 4).
Anion Concentration-The rates of the ATP-dependent C1-and Iinfluxes were saturable functions of the anion concentrations. Half-maximal rate of ATP-dependent Iinflux was obtained with 10 m~ I- (Fig. 5) and maximal rate was  Competition studies were carried out in the presence of ATP, using lZ5Ias tracer (Fig. 6). I-, C1-, and SCNcompeted with lz5I-, while SO?' up to 80 m~ did not. These observations suggested that C1-, I-, and SCN-might be transported by the same ATP-dependent mechanism while Sod2is not.
Influence of Membrane Potential-ATP-dependent C1transport was markedly increased by the addition of K' and valinomycin (Fig. 7). The addition of valinomycin or potassium alone had no effect. The K+ concentration dependence of the K' valinomycin effect was consistent with the hypothesis of a K+-generated potential according to the Nernst equation. The magnitude of the effect was furthermore dependent upon C1-concentration (being higher for high [Cl-]) which suggested it could be due to attenuation of a selfcreated potential that limits C1-uptake.
Mechanism of ATP InvoZuement-The presence of an anion-sensitive ATPase was investigated. In the presence of 10 pg/ml of oligomycin and 1 mM ouabain that inhibited 20 to 35% of the ATPase activity of the microsomes, the activity was estimated to be 90 m o l Pi liberated/min/mg of protein at 30°C. C1-and I-(4 to 20 m) stimulated basal activity (0 to 23%). HC03-, which is known to stimulate an ATPase in gastric microsomes (31,32), had no effect on our preparations.

Cl-Transport in
More consistent is the finding that the thermostable inhibitor of CAMP-dependent protein kinases (25) prevented the stimulation of the rate of Iinflux by ATP (Fig. 8). Halfmaximal inhibition was obtained with a concentration as low FIG. 6. Competition studies between I-, Cl-, SCN-, SO,*-  These vesicles, when fresh, showed a low transport capacity for C1-in contrast to the high transport capacity of gastric mucosa in uiuo. Moreover, vesicular C1-transport was not sensitive to ATP. A passive conductance for C1-was, however, evident after ageing or limited proteolysis but its physiological meaning was unclear (33). In this work, basal and ATPdependent uptakes of C1-and I-were characterized in fresh vesicles. The ATP effect was shown to be rapidly lost after ageing in the absence of dithiothreitol. This lability could suggest that the H+-transporting vesicles, as prepared by The electrical nature of the ATP-dependent C1-transport in vesicles was suggested by its capacity to respond to the membrane potential created by K+ and valinomycin. It must be noted, however, that the microsomal preparation used is in all likelihood heterogeneous and therefore the effects of K+ and valinomycin could stem from an action in another enzyme system, for example the (H+K+)-ATPase. For instance, it has been suggested that the above enzyme can generate a potential in the presence of K+ and valinomycin (34). This possibility seems unlikely because KC and valinomycin stimulation of C1-transport was not inhibited by dicyclohexylcarbodiimide although this agent is known to be a potent inhibitor of the (H+K+)-ATPase (5).

[MI
To date, the role of ATP in the transport of ions by subcellular systems has been mostly studied in relation to ATPases (35). Gastric microsomes contain electrogenic ATPases such as mitochondrial or (Na+K+)-ATPases, the activity of which could be responsible for the creation of a potential able to drive C1-into the vesicles. The possible implication of those enzymes was, however, rejected because we found that rutamycin, oligomycin, ouabain, and vanadate had no effect on (21-or Itransport (23, this work). The possible implication of the (H+K+)-ATPase is more difficult to examine because no specific inhibitor is available. However, our results argue for an independent mechanism. Zn2+ (1 m~) and dicyclohexylcarbodiimide (1 mM) were described to inhibit, respectively, 90 and 82% of the (H+K+)-ATPase activity (5); in our hands, Zn2+ had no effect on C1-transport and 1 m~ dicyclohexylcarbodiimide was either ineffective or only partially inhibitory when preincubated prior to the assay. A strong argument against the implication of the H+K+ pump is the lack of specificity of the transport system for K'. Thus, replacement of Na+, routinely used as cation in our assays, by K+ did not modify the ATP-sensitive C1-transport although the (H+K+)-ATPase is selective for K+, not Na+ (5).
Anion-sensitive ATPase were mainly described as HC03-stimulated ATPases (31,32) but singularity of these as a new type of ATPase has been contested because mitochondrial ATPase is also stimulated by HC03-(29,31). The presence of a so far unidentified C1"sensitive ATPase in our material cannot, however, be ruled out.
The protein kinase inhibitor used in this study, i.e. the M, = 26,000 thermostable protein isolated by Walsh (25) is suggested to be specific for the CAMP-dependent subclass of protein kinases. This substance inhibited the ATP-stimulated anion uptake. This inhibition did not appear to be due to the glycerophosphate buffer in which the agent was provided. In Fig. 8, showing the dose-response curve to the inhibitor, buffer concentration was kept constant in all the samples, including those without inhibitor. Neither was the inhibition due to an effect on the ATP-regenerating system (phosphoenolpyruvate; pyruvate kinase) because we found that it did not alter the rate of synthesis of pyruvate by the pyruvate kinase. Synthesis of pyruvate was measured by the oxidation of PNADH + H+ catalyzed by lactic dehydrogenase in excess.
Our result does suggest that the catalytic subunit of a CAMPdependent protein kinase might be implicated in the regulation of the gastric microsomal anion transport. This finding brings on a new interest in the mechanism of C1-transport in gastric mucosa as well as in other epithelia. Further studies are, however, needed to characterize this protein kinase dependency.