METHODS

H⁺-K⁺-ATPase Determinations

Apical membrane preparation.

Apical membranes from surface cells and from crypt cells were prepared from the distal colon of normal male Sprague-Dawley rats (200 g), as described previously (17). In brief, surface cells and crypt glands were isolated by divalent cation chelation technique using EDTA and utilized for apical membrane preparation (16). Surface cells and crypt glands were homogenized with Omni Mix (Waterbury, CT) at 2,500 and 4,000 rpm, respectively. Brush-border caps were isolated by Percoll gradient (10%) centrifugation. Apical membranes were purified by disrupting the brush-border caps by homogenization (Polytron, Westbury, NY) and centrifugation (Sorvall RC5B, SS34 rotor, 10,999 rpm, 15 min). Apical membranes used for enzyme assays were resuspended in 50 mM tris(hydroxymethyl)aminomethane (Tris) ⋅ HCl buffer (pH 7.4) containing 250 mM sucrose. Resuspended membranes were stored at −70°C until further use. Purity of the surface cell apical membrane was assessed by 10- to 12-fold enrichment of H⁺-K⁺-ATPase activity (9), whereas crypt cell apical membrane purity was assessed by the presence of a Cl⁻-dependent Na⁺/H⁺exchange activity (18).

Enzyme assay.

H⁺-K⁺-ATPase activity was measured by the method of Forbush (10) as described previously (9). In brief, 500 μl reaction mixture (50 mM Tris ⋅ HCl, 5 mM MgCl₂, 20 mM KCl, 5 mM ATP, and 10 μg protein) was incubated for 30 min at 37°C. The reaction was arrested by adding 1 ml of ice-cold stop solution (3% ascorbic acid, 1% sodium dodecyl sulfate, and 0.5% ammonium molybdate in 0.5 N HCl), and the color was developed by adding 1.5 ml of 1% sodium arsenate in 5% acetic acid. Mg²⁺-ATPase activity was measured in the absence of KCl. H⁺-K⁺-ATPase activity was measured in the presence of both Mg²⁺ and K⁺. H⁺-K⁺-ATPase activity was calculated by subtracting the Mg²⁺-ATPase activity from the activity obtained in the presence of both Mg²⁺ and K⁺. H⁺-K⁺-ATPase activity was also measured in the presence of 1 mM ouabain. H⁺-K⁺-ATPase activity measured in the presence of ouabain represents ouabain-insensitive H⁺-K⁺-ATPase activity. Ouabain-sensitive H⁺-K⁺-ATPase activity was calculated by subtracting H⁺-K⁺-ATPase activity in the presence of ouabain from H⁺-K⁺-ATPase activity in the absence of ouabain. ATPase activity is presented as micromoles P_i liberated per milligram protein per minute. Results are presented as means ± SE of three different membrane preparations; eight animals were used for each preparation. Protein was measured by the method of Lowry et al. (14).

K⁺-Dependent pH_i Determinations

Measurement of pH_i in isolated crypts.

pH_i of individual colonic cells of isolated perfused crypts was determined as previously described (16). Briefly, individual crypts were isolated using a hand dissection technique and perfused in vitro as previously described (19). After isolation, the individual crypts were transferred to a thermostatically controlled chamber mounted on the stage of an inverted microscope. After transfer the crypts were double cannulated using a series of concentric glass pipettes. The crypts were then perfused from both the basolateral and apical membrane with a Ringer buffer, pH 7.0. The luminal perfusate was then exchanged for a solution of 10 μM 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein-acetoxymethyl ester (BCECF-AM) (Molecular Probes, Eugene, OR), the precursor of the fluorescent isoform of the pH-sensitive dye BCECF. During perfusion from the lumen there was preferential dye uptake in the non-mucus-secreting cells along the length of the crypt. After a period of 10 min of dye loading the luminal solution was exchanged for a control ${HCO}_3^{-}$-free Ringer solution that contained 5 mM K⁺, for a period of 5 min. The crypt cells were imaged with an intensified charge-coupled device camera connected to an imaging system. Individual records of pH_i vs. time within single cells were obtained using a ratiometric analysis of the individual fluorescent images in a manner similar to that previously described for rabbit proximal colon cells (22).

Measurement of pH_i in polarized surface cells.

A preparation has also been developed that permits imaging studies of surface cell pH_i. These imaging studies were performed in situ in full-thickness rat distal colon with simultaneous independent superfusion of the serosal and mucosal aspects of the tissue. After rats were killed, ∼3 cm of full-thickness distal colon was removed and opened, and fecal debris was removed with${HCO}_3^{-}$-free Ringer. With a series of O-rings, the full-thickness segment was mounted in a two-sided chamber that was used for fluorescence imaging; the basic design of this chamber was adapted from those described in recent studies of colonic mucosa (4, 8, 12). This chamber is bound by two glass coverslips with a volume of ∼100 μM per side and was mounted on the stage of a Zeiss IM-35 inverted microscope, equipped for epi-illumination (20). Luminal and basolateral (i.e., “bath”) solutions were switched via computer-controlled five-way valves with zero dead space. Chamber temperature was monitored and maintained at 37°C by prewarming the bath solution with a flow rate of ∼3 ml/min.

Surface cells were loaded with the pH-sensitive dye BCECF (18) by switching the mucosal solution to anN-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES)-Ringer solution containing 10 μM BCECF-AM and then temporarily halting flow for ∼15 min. Reinitiating mucosal flow after dye loading minimized extracellular accumulation of dye. The dye uptake pattern was restricted to the surface epithelial cell, similar to that described in other studies (8, 12). At the end of each experiment, the absence of both mucosal-to-serosal leak and surface cell viability was confirmed with trypan blue (Sigma, St. Louis, MO). Experiments in which cells failed to exclude trypan blue were not used (<1% of experiments in which BCECF was initially loaded).

Solutions.

The HEPES-Ringer solution contained (in mM) 125 NaCl, 5 KCl, 1.0 CaCl₂, 1.2 MgSO₄, 2 NaH₂PO₄, 10.2 glucose, and 32 HEPES. KCl replaced NaCl (and vice versa) in selected experiments. The pH of all solutions was 7.40 at 37°C unless otherwise indicated. High-K⁺/nigericin calibration solution contained (in mM) 10 μM nigericin, 105 KCl, 1.0 CaCl₂, 1.2 MgSO₄, 2 H₃PO₄, 10.5 glucose, 32.2 HEPES, and 46.4N-methyl-d-glucamine (NMDG); pH was adjusted with HCl or NMDG. All solutions were equilibrated with air and adjusted to an osmolality of 300–310 mosmol/kg with NaCl or mannitol.

Optical system.

A solid-state intensified TV system, with digital image acquisition and analysis, was used in these studies, as described previously (20). Briefly, fluorescence emission or intensity of emitted light (I) was monitored at 530 nm while dye was alternately excited at 440 and 490 nm. Between excitation periods, the cells were in the dark, minimizing photobleaching of dye and photodynamic damage. Data pairs of 440/490 were acquired as often as one per 2.5 s, but the sampling rate (typically once per 5–30 s) was matched to the speed of pH_i changes. A program was used to outline the boundaries of individual surface cells and to group pixels by cells. Typically, two to four dye-loaded surface cells were selected randomly for analysis from each preparation. For each cell, in each sample, I₄₉₀ and I₄₄₀ values were obtained, thus providing the I₄₉₀/I₄₄₀ratio. Each experiment was concluded with a single-point calibration (3), using the nigericin/high-K⁺technique to clamp pH_i at pH 7.00 (21).

Statistics

One-tailed Student’s t-test was used, with P < 0.05 considered statistically significant.

RESULTS

Distribution of H⁺-K⁺-ATPase Isoform Activity in Surface and Crypt Cells

Surface cells.

Both ouabain-sensitive and ouabain-insensitive H⁺-K⁺-ATPase activities have previously been identified in apical membranes isolated predominantly from surface cells (9). HKcα message and protein level were expressed only in surface (and the upper 20% of crypt) cells (11,13). It is not known whether both ouabain-sensitive and ouabain-insensitive isoforms of H⁺-K⁺-ATPase are expressed solely in surface cells or in both surface and crypt cells; therefore, H⁺-K⁺-ATPase activity was assayed in apical membranes isolated from surface and crypt cells. As presented in Fig. 1, H⁺-K⁺-ATPase activity (94.0 ± 11.3 nmol P_iliberated ⋅ mg protein⁻¹ ⋅ min⁻¹) in surface cell apical membrane was partially inhibited by 1 mM ouabain (50.9 ± 7.0 nmol ⋅ mg protein⁻¹ ⋅ min⁻¹). The ouabain-insensitive fraction of H⁺-K⁺-ATPase activity was 54% of the total activity. These results are similar to those previously presented (9).

Crypt cells.

H⁺-K⁺-ATPase activity was also measured in apical membranes isolated from crypt cells and is presented in Fig. 1. H⁺-K⁺-ATPase activity of crypt cell apical membranes was substantially lower than that of surface cell apical membranes (24.0 ± 4.2 vs. 94.0 ± 11.3 nmol ⋅ mg protein⁻¹ ⋅ min⁻¹). In contrast to surface cell apical membranes, H⁺-K⁺-ATPase of crypt cell apical membranes was almost completely (96%) inhibited by 1 mM ouabain. These results indicate that the ouabain-insensitive H⁺-K⁺-ATPase isoform is expressed only in apical membranes of surface cells and that the ouabain-sensitive isoform is the only H⁺-K⁺-ATPase isoform expressed in apical membranes from crypt cells. Because ouabain-insensitive H⁺-K⁺-ATPase is present in surface but not in crypt cell apical membranes, a spatial distribution identical to that of HKcα message and protein (11, 13), it is likely that HKcα encodes the ouabain-insensitive H⁺-K⁺-ATPase α-subunit.

Although these results demonstrate the presence of both ouabain-sensitive and ouabain-insensitive H⁺-K⁺-ATPase isoforms in surface cell apical membranes, it is possible that only the ouabain-insensitive isoform is present in surface cells and that the ouabain-sensitive H⁺-K⁺-ATPase activity identified in this surface cell apical membrane preparation represents contamination of this preparation by crypt cell apical membranes. Because the method to separate surface and crypt cells may result in some cross-contamination, additional studies were performed to determine K⁺-dependent regulation of pH_i, using video fluorescent microscopy. This method visually analyzes individual cells and should resolve whether the presence of both ouabain-sensitive and ouabain-insensitive H⁺-K⁺-ATPase activity in surface cell apical membranes represents partial contamination of the surface cell preparation with crypt cells.

K⁺-Dependent Regulation of pH_i

Crypt cells.

To study K⁺-dependent regulation of pH_i in crypt cells, a recently developed preparation of isolated colonic crypts (16, 19) was used that permits separate perfusion of apical (lumen) and basolateral (bath) membranes. An acid load was induced by removing Na⁺ from both lumen and bath solutions while maintaining lumen K⁺ concentration at 5 mM.¹Figure 2 demonstrates that the addition of 20 mM K⁺ to the lumen solution increased pH_i to its baseline value. The presence of 0.5 mM ouabain before the readdition of 20 mM K⁺ completely prevented K⁺-dependent pH_i recovery (Fig.3A). These studies demonstrate that K⁺-dependent pH_i recovery from an acid load in crypt cells is exclusively ouabain sensitive (Table1) and is consistent with the presence of the ouabain-sensitive H⁺-K⁺-ATPase isoform in crypt cell apical membranes.

Surface cells.

Apical K⁺-dependent acid extrusion was also studied in polarized surface colonocytes, using a newly developed preparation. After an NH₄Cl-induced acid load¹ (2), pH_i recovered rapidly to resting pH_i values in the presence of both Na⁺ and K⁺. In the absence of both Na⁺ and K⁺, pH_i did not alkalinize toward resting pH_i values until either K⁺ or Na⁺ was returned to the mucosal solution. pH_i recovery was initiated after the addition of 5 mM mucosal K⁺, and further recovery was observed on increasing mucosal K⁺concentration to 20 mM (Fig. 3B). Recovery in the presence of either 5 or 20 mM mucosal K⁺ was not inhibited by 1.0 mM mucosal ouabain (Fig. 3B and Table 1). Because HKcα message and protein are localized to surface cells but not crypt cells, these studies of K⁺-dependent regulation of pH_i provide further support for the high probability that the functional activities encoded by HKcα isoform are not ouabain sensitive.

Role of voltage.

To exclude the possibility that K⁺-dependent recovery of pH_i in response to an acid load is a result of membrane depolarization and not apical membrane electroneutral H⁺/K⁺exchange, experiments were performed with barium in both surface and crypt cells. Two separate sets of experiments were performed. In one series of experiments the effect of luminal 5 mM barium on basal pH_i was determined. Barium will depolarize the apical membrane, so if K⁺-dependent recovery of pH_i were a result of membrane depolarization, the addition of barium should have caused an intracellular alkalization. Table 2demonstrates that the luminal addition of barium to either surface or crypt cells did not alter basal pH_i.

In a second series of studies, barium was added to the luminal solution before the induction of an acid load. The acid load was induced in surface cells by a prepulse of NH₃/NH₄Cl and in crypt cells by the removal of lumen and bath Na⁺. Table 2 demonstrates that barium failed to alter the recovery of pH_i in response to an acid load in either surface or crypt cells. These two sets of experiments exclude the possibility that voltage as a result of K⁺-induced apical membrane depolarization is responsible for the observed pH_i recovery from an acid load that is observed after the addition of K⁺ (Figs. 2 and 3).

DISCUSSION

Physiological studies of active K⁺absorption in the distal colon of the rat have revealed the presence of two distinct K⁺ transport processes (9, 15). In these studies of active K⁺ absorption, which were performed across isolated colonic mucosa under voltage-clamp conditions, K⁺ absorption consists of mucosal Na⁺-sensitive and mucosal Na⁺-insensitive components (15). Aldosterone stimulated the mucosal Na⁺-sensitive, but not the mucosal Na⁺-insensitive, component of active K⁺ absorption (15). The mucosal Na⁺-sensitive component was also ouabain insensitive, whereas the mucosal Na⁺-insensitive component was ouabain sensitive. In addition, two distinct H⁺-K⁺-ATPase fractions, ouabain sensitive and ouabain insensitive, were also identified in studies with apical membranes isolated from rat distal colon (9). Neither of these studies was designed, however, to identify the spatial localization of either the active K⁺-absorptive process or H⁺-K⁺-ATPase (9, 15).

In contrast, these present studies were planned to establish whether these different K⁺-dependent processes were exclusively present in surface and/or in crypt epithelial cells. The results of these studies provide compelling evidence that a single K⁺-dependent transport process is present in crypt epithelial cells. K⁺-dependent pH_i recovery from an acid load in crypt cells was completely ouabain sensitive (Figs. 2 and3A), whereas H⁺-K⁺-ATPase activity in apical membranes isolated solely from crypt cells was 96% inhibited by ouabain (Fig. 1). Thus it appears that there is but a single K⁺-dependent transport process in crypt epithelial cells, one that is ouabain sensitive. These studies also suggest that it is likely that only one K⁺-dependent pH_i transport process is present in surface epithelial cells and that, in contrast to that in crypt cells, this K⁺-dependent pH_i transport process is ouabain insensitive (Fig. 3B).

The results of these physiological studies must be interpreted in relation to the recent cloning and localization of the colonic H⁺-K⁺-ATPase. HKcα cDNA was cloned from rat colon and is a member of a gene family of related P-type ATPases (7, 11). Both HKcα message and protein have been localized to surface (and to 20% of upper crypt) epithelial cells of the rat distal colon, and Western blot analysis and immunocytochemical studies have established that HKcα protein is present in the apical membrane of surface cells (11, 13). Controversy exists regarding the ouabain sensitivity of the H⁺-K⁺-ATPase that is encoded by this HKcα cDNA (5, 6, 13). The H⁺-K⁺-ATPase activity expressed in Sf9 cells by HKcα cDNA was ouabain insensitive (13). In contrast, the ouabain sensitivity of⁸⁶Rb uptake after the injection of HKcα cRNA into Xenopus oocytes depended on the specific β-subunit cRNA that was coinjected. Coinjection of an amphibian bladder β-subunit resulted in 100% ouabain-sensitive ⁸⁶Rb uptake (6), whereas the ouabain-sensitive fraction of⁸⁶Rb uptake by oocytes coinjected with either gastric H⁺-K⁺-ATPase β-subunit cRNA or Na⁺-K⁺-ATPase β₁-subunit cRNA was less than 50% of total ⁸⁶Rb uptake (5).

Although the ouabain sensitivity of H⁺-K⁺-ATPase activity and that of K⁺-dependent pH_i recovery in crypt cells manifest excellent concordance, there was modest dissociation of their inhibition by ouabain in surface cells. Because pH_i recovery in surface cells was measured by video fluorescent microscopy in individual cells, ouabain insensitivity of pH_i recovery undoubtedly represents a surface cell function. In contrast, it is not unlikely that the apical membranes isolated from surface cells are partially contaminated with ouabain-sensitive H⁺-K⁺-ATPase activity from crypt apical membranes. This suggestion is based on our recent demonstration (16) that Cl⁻-dependent Na⁺/H⁺exchange is the sole Na⁺/H⁺exchange in crypt apical membrane vesicles, whereas in surface cell apical membrane vesicles Cl⁻-dependent Na⁺/H⁺exchange is 27% of total Na⁺/H⁺exchange.

In these studies an intracellular acid load was induced either by the removal of lumen and bath Na⁺ in crypt cells or by an NH₃/NH₄Cl prepulse in surface cells.¹ We established in both surface and crypt cells that in the absence of lumen Na⁺, recovery from an acid load required lumen K⁺. Although this represents the first identification of a K⁺-dependent recovery from an acid load in small or large intestinal epithelia, we do not as yet know whether this represents an important control mechanism to regulate pH_i in colonocytes or is merely an expression of H⁺/K⁺exchange that mediates transepithelial K⁺ movement.

The results of these present studies, combined with the recent localization of HKcα message and protein to surface epithelial cells (11, 13), provide compelling evidence that HKcα cDNA encodes a protein that is ouabain insensitive and that the ouabain-sensitive H⁺-K⁺-ATPase isoform is both exclusively localized in crypt cells and encoded by a cDNA that has not as yet been cloned.