Activation of Na (cid:49) /H (cid:49) Exchange Is Required for Regulatory Volume Decrease after Modest “Physiological” Volume Increases in Jejunal Villus Epithelial Cells*

Epithelial cell volume increases that occur because of the uptake of Na (cid:49) -cotransported solutes or hypotonic dilution are followed by a regulatory volume decrease (RVD) due to the activation of K (cid:49) and Cl (cid:50) channels. We studied the relationship of Na (cid:49) /H (cid:49) exchange (NHE) to this RVD in suspended guinea pig jejunal villus cells, using electronic sizing to measure cell volume changes and fluorescent spectroscopy of cells loaded with 2 (cid:42) ,7 (cid:42) -bis(carboxyethyl)-5(6)-carboxyfluorescein to monitor intracellular pH (pH i ). When the volume increase achieved by these cells during Na (cid:49) solute absorption was duplicated by a modest 5–7% hypotonic dilution, their pH i first acidified and then alkalinized. This alka- linization was blocked by 5-( N -methyl- N -isobutyl) amiloride (MIA; 1 (cid:109) M ), an inhibitor of NHE. The RVD subsequent to 5–7% hypotonic dilution was prevented by Na (cid:49) free medium and by amiloride and non-amiloride derivatives. The order of potency of these inhibitors was as follows: MIA > 5-( N , N -dimethyl) amiloride > cimeti- dine > clonidine, in keeping with the pattern attributa-ble to NHE-1 as the isoform of NHE responsible for in- crease in pH i after modest volume increases. A substantial 30% hypotonic dilution

substantial 30% hypotonic dilution caused acidification, and RVD following this larger volume increase was not affected by MIA. To assess the effect of hypotonicity on the activity of NHE, we measured the rate of MIA-sensitive pH i recovery from an acid load (dpH i /dt) in 5 and 30% hypotonic media. pH i recovery was faster in 5% hypotonic medium compared with isotonic medium and slowest in 30% hypotonic medium, which suggested that the activity of NHE was stimulated in the slightly hypotonic medium, but inhibited in the 30% hypotonic medium. To determine the role of activated NHE in RVD after a modest volume increase, cells were hypotonically diluted 7% in MIA to prevent RVD and then alkalinized by NH 4 Cl or acidified by propionic acid addition. Only after alkalinization was there complete volume regulation. We conclude that in Na ؉ -absorbing enterocytes, the NHE-1 isoform of Na ؉ /H ؉ exchange is stimulated by volume increases that duplicate the "physiological" volume increase occurring when these cells absorb Na ؉cotransported solutes. The subsequent alkalinization of pH i is a required determinant of the osmolyte loss that underlies this distinct volume regulatory mechanism.
A modest 5-7% volume increase in a jejunal villus epithelial cell is of interest because it duplicates the size to which these cells swell during absorption of Na ϩ solute (1)(2)(3). With any such volume increase, these cells activate K ϩ and Cl Ϫ channels, causing a regulatory volume decrease (RVD) 1 ; the resultant KCl efflux returns the volume to normal (4,5). Because reports using symmetrical mammalian cells have suggested that intracellular pH (pH i ) is a determinant of volume regulation (6 -8), our efforts to characterize RVD after modest 5-7% swelling focused on pH i and its relationship to Na ϩ /H ϩ exchange (NHE) activity as cell volume was increased experimentally. In this report, using suspended jejunal villus epithelial cells exposed to a slight hypotonic challenge (0.95 ϫ isotonic dilution) to duplicate the volume increase occurring because of Na ϩ solute absorption (5-7% swelling), we show a precise relationship between NHE activity, pH i , and the activation of ion channels for RVD. This sequence of events differs greatly from that observed after a "standard" hypotonic challenge (0.70 to 0.50 ϫ isotonic dilution), where after cell swelling of 15-20%, NHE activity is inhibited (9 -11).
NHE has been identified in many cell types (12,13), and four distinct isoforms have been cloned (14 -18). We have characterized NHE activation and its relationship to RVD by measuring changes in cell volume, pH i , 22 Na influx, and rates of pH i recovery from an acid load. By showing distinct differences in the response to cell swelling of different magnitudes, we have provided new insight into the mechanism of signal transduction for volume regulation in absorptive epithelial cells.

MATERIALS AND METHODS
Solutions and Reagents-Volume measurements were made on cells suspended in Na ϩ medium at a density of 30,000 cells/ml. This medium contained 140 mM NaCl, 3 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , 10 mM D-glucose, and 10 mM Hepes (pH 7.3; 295 mosm). Na ϩ -free medium and K ϩ -rich medium were made by iso-osmotic replacement of NaCl with the chloride salts of N-methyl-D-glucamine and K ϩ , respectively, and titrated to pH 7.3 with the corresponding bases. Isotonic low Na ϩ medium contained 25 mM NaCl with 115 mM N-methyl-D-glucamine and was used in all pH i recovery from ammonium prepulse experiments. Hypotonic solutions (5 and 30%) were made by an appropriate addition of distilled water. Na ϩ uptake buffer was Na ϩ medium supplemented with bovine serum albumin (type V) at 1 mg/ml.
Villus Cell Isolation and Volume Determination-Villus cells were isolated from segments of adult male (200 -300 g) guinea pig jejunum by mechanical vibration as described previously (19). We resuspended isolated cells at 0.8 -1.5 ϫ 10 6 cells/l in RPMI 1640 medium (without HCO 3 ) containing 1 mg/ml bovine serum albumin (type V) and 20 M NaHepes (pH 7.3) at 37°C. Three hours after suspension in medium, viability was 85% as assessed by trypan blue exclusion. Cell volume was measured using a Coulter Counter (model ZM) with an attached Channelyzer (C-256) as described previously (1,2,4). Villus cell volume measured electronically over a range of tonicities correlated (r ϭ 0.967) with direct measurements of cell water (4). The effect of amiloride and non-amiloride inhibitors illustrated in Fig. 8 was measured using an attenuation setting of 32. We determined relative cell volume as the ratio of cell volume under study conditions to the volume under basal conditions in isotonic medium.
pH i Measurement and Manipulation-For the fluorometric determination of pH i , villus cell suspensions (1 ϫ 10 6 cells/ml in Hepes/RPMI 1640 medium) were loaded with 2Ј,7Ј-bis(carboxyethyl)-5(6)-carboxyfluorescein by incubation with the parent acetoxymethyl ester (3.7 M) for 15 min at 37°C. After washing, 0.5-0.8 ϫ 10 6 cells were used for fluorescence determination in 2 ml of the indicated medium using a Hitachi F-4000 fluorometer with excitation at 495 nm and emission at 525 nm using 5-and 10-nm slits, respectively. We acid-loaded cells by preincubating 10 6 cells/ml for 5 min in RPMI 1640 medium containing 2.5 mM NH 4 Cl at 37°C, followed by sedimentation and resuspension in 2 ml of the indicated NH 4 ϩ -free medium. For experiments using acidloaded cells, loading with NH 4 ϩ and 2Ј,7Ј-bis(carboxyethyl)-5(6)-carboxyfluorescein was performed simultaneously as described (20). The MIA-sensitive rate of pH i recovery was the difference between pH i recovery in the presence of bafilomycin (100 nM) and Zn 2ϩ (100 M) and pH i recovery in the presence of these inhibitors and 1 M MIA. Rates of pH i recovery were determined in low Na ϩ medium (25 mM) as described above. Calibration was performed in K ϩ medium with nigericin (21) using a quench correction factor as described (22).
Uptake of 22 Na-We measured the initial rate of 22 Na influx with a modified version of a procedure previously described (19). Each villus cell preparation was divided in half and resuspended at a final concentration of 5-6 mg of protein/ml in prewarmed uptake medium in a continuously stirred cuvette. This medium contained 10 M bumetanide to inhibit NaKCl 2 cotransport. Uptake was initiated by the addition of 22 Na at a concentration of 8 -10 Ci/ml. Immediately afterwards, a 500-l cell suspension was removed and added to 500 l of ice-cold 0.1 M MgCl 2 . This aliquot, which took Ͻ5 s to obtain, was taken to represent extracellular 22 Na associated with the cell pellet. Uptake was terminated after 90 s by diluting 500 l of cell suspension in an equal volume of ice-cold 0.1 M MgCl 2 , which was then gently layered on a 100-l layer of di-n-butyl phthalate/di-n-nonyl phthalate (3:2, v/v) and centrifuged in an Eppendorf microcentrifuge for 20 s. Aliquots of the supernatant were saved for counting, and the cell pellet was processed as described previously (2,4,19). Prior to the addition of 22 Na, duplicate samples were taken and processed as described above, but following aspiration of the supernatant and oil, 100 l of Triton X-100 was added to the pellet. After vigorous shaking and cell lysis, we measured protein concentration using the Bio-Rad protein assay reagent with bovine ␥-globulin as a standard. All uptake experiments were done in the presence and absence of 1 M MIA. For the uptake experiments performed under 5% hypotonic conditions, the medium was diluted with distilled H 2 O; 60 s later, the isotope was added, and uptake was allowed to proceed for 90 s. In preliminary experiments, we determined that 22 Na uptake was first-order for 110 s. The extracellular 22 Na associated with the cell pellet was subtracted from the 90-s values. Rates, expressed as nmol/ min/mg of protein, were based on five to seven experiments performed in duplicate.
Statistics-Data are reported as means Ϯ S.E. of 5 to 16 experiments performed in duplicate. Differences in means were determined using Student's t test.

5-(N-Methyl-N-isobutyl) Amiloride-sensitive Regulatory Volume Decrease-
The effect of Na ϩ -free medium on RVD following greater volume increases occurring in a very hypotonic medium is illustrated in Fig. 4. After 0.7 ϫ isotonic dilution or 0.8 ϫ isotonic dilution, the subsequent RVD was complete. Consistent with these findings, MIA (1 M) had no effect on RVD of villus cells after 0.7 ϫ isotonic dilution in Na ϩ -containing medium (extent of volume decrease ϭ 19 Ϯ 1 versus 19 Ϯ 1%, n ϭ 6). Together, the data illustrated in Figs. 1-4 suggest that when the villus cell swells after 5% hypotonic dilution, the pH i undergoes a MIA-sensitive alkalinization, and complete volume recovery is both Na ϩ -dependent and MIA-sensitive. 22 Na Influx Increased by 0.95 ϫ Isotonic Dilution-We measured the initial rate of 22 Na influx of villus cells in suspension that was MIA-sensitive under isotonic conditions or after 0.95 ϫ isotonic dilution. Under isotonic conditions (140 mM Na ϩ ) measured in the presence of bumetanide (10 M) to block NaKCl 2 cotransport, the initial rate of 22 Na influx was attenuated by 1 M MIA (46.8 Ϯ 7.1 versus 29.7 Ϯ 2.5 nmol of 22 Na/min/mg of protein; p Ͻ 0.05). When the cell medium was diluted 0.95 ϫ isotonic, this 22 Na influx rate was accelerated (73.5 Ϯ 10.6 nmol of 22 Na/min/mg; p Ͻ 0.05 versus isotonic), a response that was blocked by 1 M MIA (37.7 Ϯ 6.1 nmol of 22 Na/min/mg; p Ͻ 0.02). Clearly, the 1 M MIA-sensitive fraction of 22 Na influx was increased 2-fold after 5% hypotonic dilution.
Recovery from Intracellular Acidification-Since 0.95 ϫ isotonic dilution caused the villus cells to acidify, we sought to determine if the activation of MIA-sensitive NHE during the 5% volume increase was secondary to this acidification. The villus cells were acidified using the ammonium prepulse technique (23, 24). As illustrated in Fig. 5, villus cells acidified to a pH i of Ӎ6.95 by ammonium prepulse exhibit recovery. The addition of 2.5 mM NH 4 Cl to these cells caused an alkaliniza-tion of 0.15 Ϯ 0.02 pH units (n ϭ 4) (Fig. 5A). The pH i then declined (0.09 Ϯ 0.02 pH units) over the next 5 min and did not change thereafter. When these cells were resuspended in isotonic medium (NH 4 Cl Ϫ -free), the time course of recovery was first-order for 5 min and was ϳ41% complete in 10 min (Fig.  5B). When cells were suspended in 0.95 ϫ isotonic medium, they recovered their pH i with a time course that was first-order for 5 min and Ӎ57% complete in 10 min (Fig. 5C). The rate of pH i recovery measured over the first 5 min was faster in 0.95 ϫ isotonic medium than in isotonic medium (dpH i /dt ϫ 10 Ϫ2 pH units/min ϭ 3.48 Ϯ 0.27 versus 1.59 Ϯ 0.25, n ϭ 6; p Ͻ 0.001). We then assessed the effect of MIA on the rate of pH i recovery. In isotonic medium (Fig. 5D), MIA slightly diminished pH i recovery (Fig. 5E). In 0.95 ϫ isotonic medium, MIA substantially diminished pH i recovery (Fig. 5, F and G). The MIAsensitive rate of pH i recovery was greater in 0.95 ϫ isotonic medium compared with isotonic medium ((1.18 Ϯ 0.15) versus (0.43 Ϯ 0.03) ϫ 10 Ϫ2 pH units/min, n ϭ 5; p Ͻ 0.001). Because the MIA-sensitive rate of pH i recovery in isotonic medium was low, we speculated that identifying and controlling amilorideinsensitive sources of pH i recovery would better resolve the MIA-sensitive component of pH i recovery from an acid load in these cells.
The effects on pH i recovery following ammonium prepulse of bafilomycin A, a potent and selective inhibitor of type V H ϩ -ATPase, and of Zn 2ϩ , an inhibitor of H ϩ conductance, are illustrated in Fig. 6. The initial rate of pH i recovery was firstorder and 18% complete (Fig. 6, A). Bafilomycin (100 nM) reduced the extent of pH i recovery (⌬pH i /2 min ϭ 64.1 Ϯ 9.1%; p Ͻ 0.005) (Fig. 6, A and C). The inclusion of Zn 2ϩ (100 M) in the presence of bafilomycin further diminished the extent of pH i recovery (⌬pH i /2 min ϭ 25.4 Ϯ 2.9%; p Ͻ 0.02) (Fig. 6, B  and C). These results suggest that Ӎ75% of pH i recovery in isotonic medium is amiloride-insensitive.
Effect of Hypotonicity on Initial Rate of MIA-sensitive pH i Recovery from Intracellular Acidification-To determine the effect of hypotonicity on the activity of NHE, we measured the initial rate of MIA (1 M)-sensitive pH i recovery following ammonium prepulse in the presence of bafilomycin and Zn 2ϩ in  ). D, q, 0.93 ϫ isotonic dilution in Na ϩ -free medium. p Ͻ 0.001 in all cases where RVD was inhibited (n ϭ 6). Volume was measured electronically and is expressed relative to the isotonic control.

FIG. 4. Effect of Na ؉ -free medium on RVD after greater volume
increases. E, 0.70 ϫ isotonic dilution; q, 0.8 ϫ isotonic dilution (n ϭ 6). Volume was measured electronically and is expressed relative to the isotonic control. media of different tonicities (Fig. 7). In isotonic media, MIA (1 M) had only a slight effect on pH i recovery (Fig. 7A), but in 0.95 ϫ isotonic medium, MIA completely blocked pH i recovery (Fig. 7B). In 0.70 ϫ isotonic medium, the villus cells continued to acidify; MIA attenuated this acidification (Fig. 7C). The initial rates of MIA-sensitive pH i recovery are summarized in Table I. In all cases, the starting pH i was the same. In isotonic medium, the initial rate of MIA-sensitive pH i recovery was (1.16 Ϯ 0.23) ϫ 10 Ϫ2 pH units/min. This rate was substantially increased in 0.95 ϫ isotonic medium ((3.25 Ϯ 0.26) ϫ 10 Ϫ2 pH units/min; p Ͻ 0.001). In 0.70 ϫ isotonic medium, the rate was inhibited ((0.31 Ϯ 0.20) ϫ 10 Ϫ2 pH units/min; p Ͻ 0.02). Clearly, the activation of MIA-sensitive Na ϩ /H ϩ exchange by 5% hypotonic swelling was not due to intracellular acidification.
Determination of NHE Isoform Activated by 5% Hypotonicity-We measured the relative pharmacological sensitivities of RVD after 0.95 ϫ isotonic dilution to several NHE inhibitors, both amiloride and non-amiloride derivatives. Cimetidine (25 M) attenuated the rate of RVD in comparison with clonidine (50 M) (Fig. 8A). The relative volume of cells treated with cimetidine at 5 min was greater than that of cells treated with clonidine (relative volume ϭ 1.02 Ϯ 0.01 versus 1.00, n ϭ 7; p Ͻ 0.05). Concentration-response profiles for inhibition of RVD after 0.95 ϫ isotonic dilution are illustrated in Fig. 8B. Cimetidine was six times more potent than clonidine. The EC 50 values of cimetidine (20 M) and clonidine (130 M) were greater than those of 5-(N,N-dimethyl) amiloride (1 M) and 5-(N-methyl-N-isobutyl) amiloride (220 nM). The order of po-tency of these inhibitors of the isoform of NHE activated by the 5% volume increase was as follows: MIA Ͼ 5-(N,N-dimethyl) amiloride Ͼ cimetidine Ͼ clonidine.
To determine the effect of extracellular K ϩ on the isoform of NHE activated by 5% hypotonicity, we measured the change in pH i of the villus cells suspended in isotonic K ϩ -rich medium. After 0.95 ϫ isotonic dilution, the increase in pH i was substantially diminished in 140 mM K ϩ compared with 3 mM K ϩ (⌬pH i /3 min ϭ 0.020 Ϯ 0.010 versus 0.070 Ϯ 0.010 pH units, n ϭ 6; p Ͻ 0.001).
NH 4 Cl-induced Alkalinization Allows RVD When NHE Is Inhibited-Since increasing osmolyte influx (Na ϩ ) when the villus cells are losing K ϩ and Cl Ϫ for RVD seems counterintuitive, we measured the volume and ⌬pH i in cells hypotonically diluted 7% in the presence of MIA (1 M) and following the addition of 1 mM NH 4 Cl (Fig. 9). As previously observed, MIA (1 M) prevented RVD after cell swelling following 7% hypotonic dilution (Fig. 9A). The addition of 1 mM NH 4 Cl to the swollen cells caused RVD in the presence of MIA. Within 2 min of the addition of NH 4 Cl, these cells started to shrink, and RVD was complete in the next 3 min (relative volume ϭ 1.03 Ϯ 0.01 versus 1.00 Ϯ 0.01; p Ͻ 0.001). Immediately after the addition of 1 mM NH 4 Cl, the villus cells alkalinized (Fig. 9B). This alkalinization (0.086 Ϯ 0.010 pH units, n ϭ 15) was no different than that measured in these cells following 5% hypotonic dilution (Fig. 2). We then acidified the pH i of comparably treated cells to show the converse of the alkalinization experiment (Fig.  9, C and D). Sodium propionate (2 mM) added to cells hypotonically diluted 7% in the presence of MIA had no effect on the inhibited RVD (Fig. 9C). The addition of sodium propionate to these cells caused an acidification (0.086 Ϯ 0.016 pH units, n ϭ 10) that remained stable for the next 5 min of the experiment (Fig. 9D). This experiment suggests that it is alkalinization of pH i caused by the NHE activated by modest swelling that signals the ion conductances for the subsequent volume regulation. DISCUSSION Our results indicate that a modest cell volume increase of 5-7% activates NHE, while an increase of 15% caused by standard hypotonic dilution inhibits NHE. Furthermore, RVD following the modest volume increase of 5-7% absolutely requires activated NHE. We base this interpretation of our results on experiments that isolated the function of NHE during modest volume increases to show amiloride-sensitive alkalinization of pH i and increases in both 22 Na influx and pH i recovery from an acid load in slightly hypotonic (0.95 ϫ isotonic) medium. Evidence that activated NHE was required for this RVD came from experiments showing that extracellular Na ϩ was required for the subsequent cell shrinkage and inhibitor sensitivity of RVD. Our results also indicate that it is the alkalinization of pH i from activated NHE that is required for RVD after modest volume increases as transient alkalinization of pH i caused cell shrinkage when an amiloride derivative had prevented volume regulation.
RVD following a modest volume increase was prevented by the non-amiloride derivatives cimetidine and clonidine, with cimetidine being six times more potent than clonidine. This observation provides strong evidence that NHE-1 is the isoform activated by the 5-7% volume increase. Studies using transfectants of NHE isoforms have shown that clonidine is more potent than cimetidine in inhibiting NHE-2 and NHE-3, while only with NHE-1 is this order of potency reversed (25,26). K ϩ , a weak competitive inhibitor of NHE-1, but not of NHE-2 or NHE-3 (26), prevented the alkalinization of pH i stimulated by the slight (0.95 ϫ isotonic) hypotonic dilution, supporting the interpretation that the isoform responsible for these pH i effects is NHE-1. Jejunal villus epithelial cells possess three isoforms of NHE (17,27,28). NHE-3 and NHE-2 are found on the apical membrane and are both more sensitive to clonidine than to cimetidine, while NHE-1, which is more sensitive to cimetidine than to clonidine, has been localized to the basolateral membrane of villus cells (27)(28)(29). We found that MIA was more potent than dimethyl amiloride in preventing RVD after the modest 5-7% volume increase, but this hierarchy is the same for NHE-1, -2, and -3 (26). After a 5% volume increase, but not a 15% volume increase, RVD, the alkalinization of pH i , the increased 22 Na influx, and recovery of pH i from an acid load were all prevented by a low concentration of the N-5-alkyl amiloride derivative. As cimetidine was more potent than clonidine in preventing this RVD, when taken together, our data strongly suggest that NHE-1 is the isoform of NHE activated during the modest 5-7% volume increase in the villus cells.
The fact that the MIA-sensitive rate of pH i recovery from an acid load was accelerated in cells suspended in 0.95 ϫ isotonic medium but inhibited in cells suspended in the standard hypotonic (0.70 ϫ isotonic) medium suggests that intracellular acidification is not directly related to activation of NHE-1. The well documented kinetic asymmetry of NHE-1 in symmetrical cells (12), evidenced by a sigmoidal relationship between the Na ϩ /H ϩ exchange rate and internal H ϩ concentration, suggests that such cooperativity is because of an intracellular H ϩ modifier site, distinct from the H ϩ transport site, and that regulation of NHE occurs via changes in the affinity of this internal H ϩ modifier site for intracellular H ϩ (12,30). We measured differences in the rate of pH i recovery from an acid load after resuspending the villus cells in media of different tonicities, but with [Na ϩ ] o reduced to 25 mM. We used these conditions because others, using A6 cells, an epithelial cell line that exhibits both apical and basolateral NHE (31), have observed that pH i changes due to basolateral NHE are greatest at low Na ϩ concentrations (32). Our findings of a substantial increase in pH i recovery in 0.95 ϫ isotonic medium compared with isotonic low Na ϩ medium are in accord with these data from A6 cells. The inhibition of MIA-sensitive pH i recovery in 0.70 ϫ isotonic medium but with the same Na ϩ concentration is consistent with reports of the effect of substantial hypotonicity on NHE in symmetrical cells. In nominally HCO 3 -free medium, after acid loading, osteosarcoma cell suspensions undergo amiloride-sensitive pH i changes that are diminished in 0.70 ϫ isotonic medium (9). Kinetic analysis of these data demonstrated that inhibition of the exchanger was due to decreased V max without a change in apparent affinity for H ϩ or Na ϩ . Single cell analysis of ⌬pH i after 0.70 ϫ hypotonic dilution confirmed that NHE was inhibited following cell swelling (10). We have clearly shown that the MIA-sensitive pH i recovery from an acid load increased in 0.95 ϫ isotonic medium, but decreased in 0.70 ϫ isotonic medium compared with isotonic controls. We conclude that the activation of basolateral NHE-1  during the modest 5-7% volume increase in the villus cells is not a consequence of the intracellular acidification normally observed after cell swelling. Approximately 75% of the pH i recovery from a moderate acid load in the villus cells was insensitive to the N-5-alkyl amiloride derivative. Our finding that bafilomycin inhibited 36% of the pH i recovery suggests that a type V H ϩ -ATPase contributes to pH i homeostasis in these cells (33). This ATPase, which has been localized to the apical membrane of urinary bladder epithelial cells (34) and the plasma membrane of peritoneal macrophages (35), has been shown to contribute to pH i recovery from acid loads in the presence of amiloride in both peritoneal (20) and alveolar (36) macrophages. We also observed that Zn 2ϩ , in the presence of this selective inhibitor of vacuolar H ϩ -ATPase, further reduced by ϳ39% the pH i recovery of the villus cells. The concentration of Zn 2ϩ used in our experiments has been shown by others to block H ϩ conductance in snail neurons (37) and human granulocytes (38). Furthermore, unequivocal results using transfectants of NHE-1 have shown that H ϩ conductance, which is Zn 2ϩ -sensitive, may be dissociated from NHE activity and that substantial alkalinizations of pH i due to NHE-1 still occur in the presence of ZnCl 2 (39). As such, the sensitivities of pH i recovery from an acid load to bafilomycin and to Zn 2ϩ suggest that both H ϩ conductance and a type V ATPase substantially contribute to pH i homeostasis in villus cells. Cellular acidification after cell swelling because of hypotonic dilution (0.70 to 0.50 ϫ isotonic) has been observed in several symmetrical cell types. The source of acidification has been speculated to be conductive OH Ϫ efflux through volume-activated Cl Ϫ channels (10), increased glycolytic metabolic activity (8), or inhibited NHE activity (9). We found that cell swelling of 15% following this standard hypotonic dilution inhibited NHE activity. Clearly, several mechanisms contribute to swellinginduced cellular acidification since modest 5-7% volume increases similar to swelling caused by Na ϩ solute absorption cause villus cells to acidify as well as to activate NHE-1. Far from being a "housekeeping" function, the activation of NHE-1 may be an essential requirement for RVD following "physiological" volume increases, when these cells swell during Na ϩ solute absorption.
Alkalinizing the pH i of cells swollen 5% in the presence of MIA by-passed inhibition and allowed complete RVD, whereas acidifying the pH i had no effect on the inhibited volume reduction. The extent of this alkalinization, induced by NH 4 Cl addition, was comparable to that observed in cells swollen after 0.95 ϫ isotonic dilution. This finding suggests that Na ϩ influx resulting from activated NHE-1 is osmotically neutral and that it is the change in pH i that is a determinant of the osmolyte loss (K ϩ and Cl Ϫ ) required for volume regulation. Previously, we reported that RVD following swelling because of the uptake of D-glucose was sensitive to the high conductance Ca 2ϩ -activated . Volume was measured electronically and is expressed relative to the isotonic control.
(maxi-K) K ϩ channel blocker charybdotoxin, while RVD following a greater swelling of 15% caused by the standard 0.70 ϫ isotonic dilution was insensitive to the toxin (3). Since the calcium gating of charybdotoxin-sensitive K ϩ conductance is exquisitely sensitive to alkaline pH i (40), the activation of NHE-1 during the modest 5-7% volume increase may serve as the source of the required alkalinization for Ca 2ϩ gating of the charybdotoxin-sensitive K ϩ loss. Because villus cells acidify as they swell 15% of their isotonic volume, we speculate that a different K ϩ conductance is activated for RVD following larger, "non-physiological" volume increases. As villus cells swell, the extent of that swelling is a key determinant of changes in pH i , which in turn serve to signal the subsequent volume regulation.