INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

CHLORIDE SECRETION in the intestine requires the activation of specific ion transporters and channels located in apical and basolateral membranes of polarized enterocytes. Basolateral transporters, which include the Na⁺-K⁺-ATPase, the Na⁺-K⁺-2Cl cotransporter, and K⁺ channels, act coordinately to elevate intracellular Cl concentration to levels above its electrochemical equilibrium potential, so that enterocytes are primed to secrete Cl through apical anion channels. Agonists that elicit elevations in intracellular concentrations of Ca²⁺ and cAMP regulate the activities of these transporters and channels and thus regulate secretion (7, 17, 28, 38, 39).

Established human intestinal cell lines, such as T84, maintain a secretory phenotype and have been used widely to examine the physiology and regulation of intestinal Cl secretion (14). Activation of apical Cl conductances by cAMP agonists in intact T84 cell monolayers has been well documented. The Cl effluxes induced by cAMP are presumed to represent Cl transport through the cystic fibrosis transmembrane conductance regulator (CFTR) (1). CFTR is highly expressed in T84 cells (18). Activation of Cl secretion by Ca²⁺-dependent agonists in T84 cells has also been well studied (3-5, 24), but the molecular identity of the channel or channels mediating this response remains undefined (1, 2, 9, 22).

Two biophysically distinct anion conductances have been described in nonpolarized human T84 and parental HT-29 cell lines. Whole cell voltage-clamp experiments have shown that isolated T84 cells grown on glass coverslips contain Ca²⁺-stimulated Cl currents that exhibit outward rectification. cAMP-induced currents from the same cells displayed linear current-voltage relationships (9, 32). Ion selectivities of the Ca²⁺-stimulated Cl currents in T84 cells differed from the ion selectivities exhibited by channels activated by cAMP (1). Experments using ¹²⁵I to trace Clsecretion similarly indicated that cAMP- and Ca²⁺-induced Cl effluxes were mediated by two separate pathways. Experimental measurements of Cl transport in well-differentiated, polarized T84 and HT-29cl.19A cells grown on permeable supports estimated by the Cl-sensitive fluorophore, 6-methoxy-N-(3-sulfopropyl) quinolinium (SPQ) (22), microelectrodes (2), or ¹²⁵I efflux (16) also support the conclusion that T84 cells express Ca²⁺-regulated Cl channels that differ from Cl channels regulated by cAMP.

Not all studies, however, have found evidence for the expression of Ca²⁺-sensitive apical membrane Cl channels on well-differentiated polarized HT-29cl.19A or T84 cells. For example, when basolateral membranes of HT-29cl.19A or T84 cells were permeabilized selectively by nystatin to examine ion conductances across the apical membrane, no evidence for Ca²⁺-stimulated Cl conductance was found (1, 12, 34). These authors concluded that Ca²⁺-stimulated Cl conductances, although present on nonpolarized cells, were not expressed on apical membranes of well-differentiated and polarized monolayers (1). However, another study using this same approach found an increase of apical Cl conductance by carbachol, suggesting that Ca²⁺-sensitive Cl channels may be present on apical membranes of T84 cells (22). Thus the presence of Ca²⁺-sensitive Cl channels in polarized colonic epithelial monolayer cultures has remained controversial.

Our aim in the present study was to reexamine the mechanisms of conductive Cl transport across apical membranes of polarized T84 cells. For these studies, we selectively permeabilized basolateral membranes of confluent monolayers grown on permeable supports with the ionophore amphotericin B (25). This allowed estimation of apical membrane conductances from measurements of transepithelial conductance. Our data show that well-differentiated human intestinal T84 cells express at least two distinct apical membrane Cl conductances. Although one can be activated only by the Ca²⁺ agonist thapsigargin, both appear to be activated by the cAMP agonist forskolin.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Cell culture. T84 cells (American Type Culture Collection, passages 45-98), a human colonic carcinoma cell line that functionally and morphologically resembles crypt intestinal epithelia, were grown as confluent monolayers in a 1:1 mixture of Dulbecco's Vogt modified Eagle's medium and Ham's F-12 medium supplemented with 15 mM HEPES buffer (pH 7.5), 14 mM NaHCO₃, 40 µg/ml penicillin, 90 µg/ml streptomycin, and 5% newborn calf serum. Monolayers were subcultured every 7 days by trypsinization with 0.1% trypsin and 0.9 mM EDTA in Ca²⁺- and Mg²⁺-free PBS and grown on collagen-coated permeable supports (area 0.3 cm², pore size 0.4 µm). All experiments described in this study were performed on cells between passages 65 and 92.

Electrophysiology. Studies were carried out at either 37 or 4°C with confluent monolayers plated on collagen-coated permeable supports (14) and examined 7-16 days postplating as previously described (38). Before all studies, inserts were washed with HCO₃-free medium (see Table 1) warmed to 37°C or cooled to 4°C and transferred to new 24-well tissue culture plates containing the experimental medium. To determine currents, transepithelial potentials, and conductances, a commercial voltage clamp (Bioengineering Dept., University of Iowa, Iowa City, IA) was interfaced with equilibrated pairs of calomel electrodes submerged in saturated KCl and with paired Ag-AgCl electrodes submerged in the experimental medium. Before each experiment, a blank filter was used to compensate for both the fluid resistance and resistance of the filter. In some experiments, transepithelial voltage and short-circuit current (I_sc) were continuously recorded with the aid of an analog-to-digital converter (MacLab, World Precision Instruments) and a microcomputer.

Measurement of Cl conductance of apical plasma membrane. To evaluate the ion conductance of the isolated apical plasma membrane, the polyene ionophore amphotericin B (25, 30, 31, 36) was added to the basolateral solution at a concentration of 100 µM, the lowest concentration that gave a maximal change in steady-state current as determined in preliminary experiments. Only the plasma membrane facing the amphotericin-containing solution (basolateral membrane in this case) incorporates this ionophore, thus electrically isolating the opposing (apical) plasma membrane. To study the Cl conductance of the apical plasma membrane so isolated, the voltage across the monolayer was then sequentially stepped from a holding voltage of 0 mV to values between 80 and +80 mV over a period of ~20 s. The protocol was performed before and 5 min after addition of forskolin and thapsigargin. The compositions of all solutions used in these experiments are shown in Table 1. Positive currents correspond to anion secretion and/or cation absorption.

cAMP measurement. Forskolin (100 µM) or thapsigargin (10 µM) was added to basolaterally permeabilized monolayers at 4 or 37°C. Seven minutes after stimulation, monolayers (surface area 5 cm²) were removed from their reservoirs and immersed in Hanks' balanced salt solution (HBSS⁺) at 4°C to terminate the reaction. The monolayers were rapidly cut from their plastic supports and placed in Microfuge tubes containing extraction buffer [66% ethanol, 33% HBSS⁺, and 1 mM phosphodiesterase inhibitor IBMX (Sigma)] at 4°C. Monolayers were compacted and extracts cleared of cellular debris by centrifugation. An aliquot (100 µl) of cleared cell extract was withdrawn for cAMP RIA (Du Pont-NEN) as previously described (41).

Cytosolic Ca²⁺ measurement. T84 cells cultured at subconfluent density on prepared collagen-coated 5-cm² glass coverslips (200 µl of collagen/glass coverslip) were loaded with the fluorescent Ca²⁺ indicator dye fura 2 by incubation with 5 µM of its AM derivative, fura 2-AM, in HBSS⁺ at 37°C for 60 min. Extracellular fura 2-AM was removed by washing twice with normal medium (see Table 1). The monolayer was then mounted on a modified Leiden chamber in which the coverslip constituted the bottom. Before the start of the Ca²⁺ measurements, the coverslip mounted in the Leiden chamber was maintained at either 37 or 0°C. To the chamber was added 1 ml of buffer solution at either 37 or 0°C. The unclamped temperature decreased ~4°C from 37°C and increased 4°C from 0°C during the course of each experiment.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

DIDS sensitivity of thapsigargin- and forskolin-stimulated I_sc. In intact monolayers, 10 µM thapsigargin added to the basolateral reservoir increases the transepithelial I_sc and the conductance by 12.6 ± 1.3 µA/cm² and 2.8 ± 0.6 mS/cm² (n = 8), respectively (Fig. 1). As reported in previous studies, the stimulation of adenylate cyclase by forskolin increases I_sc and conductance to a greater degree 48.4 ± 5.7 µA/cm² and 6.1 ± 0.35 mS/cm² (n = 8), respectively (Fig. 1). These data document stimulation of I_sc consistent with electrogenic Cl secretion but do not discriminate between the apical Cl efflux pathways activated by thapsigargin on one hand and forskolin on the other.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Transepithelial Cl secretion was activated by addition of thapsigargin or forskolin. Cells were grown to confluence on collagen-coated permeable supports, with conductances of <1 mS/cm2. Short-circuit current (Isc) was measured under voltage clamp at 0 mV (see METHODS). Arrows indicate addition of 10 µM thapsigargin () or 100 µM forskolin () to basolateral compartment. Note that subsequent addition of 100 µM DIDS to apical compartment reduced by ~40% forskolin-dependent Isc () and by 100% thapsigargin-dependent Isc (). Data are means of 8 experiments, each performed in triplicate.

Basolateral permeabilization does not affect integrity of apical plasma membrane and leaves tight junctions intact. To directly measure conductance changes that occur in the apical plasma membrane, we used amphotericin B to permeabilize selectively basolateral membranes. Only monovalent ions traverse the amphotericin pore. Because of its requirement for cholesterol, amphotericin permeabilizes only the plasma membrane domain in direct contact with the amphotericin B solution as previously demonstrated (25). Thus this procedure leaves the apical membrane and tight junctions intact as judged by measurement of transepithelial conductances (before amphotericin B addition, G_intact = 0.583 ± 0.032 vs. after amphotericin addition, G_apical= 2.1 ± 0.6 mS/cm², n = 20; see Fig. 2). The observed increase in conductance reflects removal of electrical resistance due to the basolateral membrane and provides indirect evidence that apical membranes and intercellular tight junctions remain intact. As previously reported (30, 31, 36), changes in Cl conductances of luminal plasma membranes under these conditions are reflected as changes in I_sc when an electrochemical gradient is established to drive Cl across the permeabilized monolayer.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Effect of amphotericin B treatment on intact T84 conductances and effect of thapsigargin and forskolin (+, present; , absent) on basolaterally permeabilized T84 conductances; n = 20-30 monolayers for each condition.

Forskolin but not thapsigargin activates an apical conductance at 0-mV membrane potential. We next examined the effects of forskolin and thapsigargin on apical membrane Cl currents in permeabilized monolayers. After permeabilization of basolateral membranes with amphotericin B, a Cl gradient was established across the electrically isolated apical membrane by placing "normal Cl medium" (buffer A) in the basolateral reservoir and "low Cl medium" (buffer B) in the apical reservoir (Na⁺ present in buffers A and B at 137 mM). These conditions are essentially identical to those established by Anderson and Welsh (1). When the monolayer was voltage clamped at 0 mV, thapsigargin did not elicit a detectable increase in apical Cl conductance. In contrast, forskolin stimulated a sizable Cl current under these same conditions (Fig. 3A). These data reproduce the results of previous studies and show that, at 0-mV membrane potential, only forskolin stimulates a Cl conductance. However, when the same experiment was performed at +15 mV potential (inside negative), both thapsigargin and forskolin elicited a clear increase in apical anion conductance (Fig. 3, B and C). These data suggest that, under appropriate conditions, apical membranes of T84 cells may exhibit both cAMP- and Ca²⁺-sensitive Cl conductances.

View larger version (10K): [in this window] [in a new window]   Fig. 3.   A: at 0-mV transmembrane potential (V = 0 mV), forskolin but not thapsigargin stimulates Isc across apical plasma membrane. A defined basolateral-to-lumen Cl gradient was established across monolayer by substitution of Na-gluconate for NaCl in solution bathing luminal side and addition of 100 µM amphotericin B to basolateral solution (normal Cl medium). First thapsigargin (10 µM) and later forskolin (100 µM) were added to basolateral solution. Under same conditions but at +15 mV transmembrane potential (V = 15 mV), both 10 µM thapsigargin (B) and 100 µM forskolin (C) stimulate Isc. Each tracing is representative of results for 3 experiments with 4 different filters in each condition.

Thapsigargin and forskolin activate apical Cl conductances. To test this hypothesis, the basolaterally permeabilized T84 monolayers were studied in symmetrical buffers containing Cl as the major permeant ion; monovalent cations were in the millimolar range (buffer E, Table 1). Figure 4, A and B, shows that, under symmetrical Cl concentrations, the relationships between Cl current induced by either thapsigargin or forskolin were linear. Currents induced by forskolin were much larger than currents induced by thapsigargin. Monolayer conductances induced by thapsigargin were 1 ± 0.2 mS/cm² and by forskolin 3 ± 0.4 mS/cm² at +40 mV (mean of 20 independent experiments, see Fig. 2). Same results have been found in buffers containing Cl as the sole permeant ion (choline chloride, 143 mM; CaSO₄, 1.25 mM; thapsigargin and forskolin stimulated a Cl current under these conditions, 30 and 100 µA/cm² at +40 mV, respectively). To verify that the thapsigargin- and forskolin-induced currents were specific to Cl transport, Cl was replaced by gluconate in the continued presence of low Na⁺ (buffer F, Table 1). Under these conditions, the currents normally stimulated by thapsigargin and forskolin were inhibited by nearly 100% (Fig. 4, A and B). These experiments suggest that thapsigargin and forskolin stimulate a Cl current.

View larger version (13K): [in this window] [in a new window]   View larger version (12K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Fig. 4.   Current-voltage relationships in basolaterally permeabilized monolayers of T84 treated with thapsigargin (A) or forskolin (B). Basolateral-to-apical reservoir Cl concentration ratios (in mM) were 143.50/143.50 mM Cl, Na+ present in both sides at 0.5 mM (thapsigargin, A, ; forskolin, B, ); with presence of a Cl gradient between basolateral (143.50 mM Cl) and apical compartment (6.61 mM Cl), Na+ present in both sides at 137 mM (thapsigargin, A, ; forskolin, B, ); with a low Cl concentration in both apical and cytosol basolateral compartment (6.61 mM Cl), Na+ present in both sides at 0.5 mM (thapsigargin, A, ×; forskolin, B, ×). With presence of a Cl gradient between basolateral (143 mM Cl) and apical compartment (43 mM), in absence of Na+ in both sides (thapsigargin, C, ; forskolin, C, ). Ordinates give difference current between stimulated and unstimulated cells. Tracings are representative of results from 10 different filters for each condition.

Thapsigargin- and forskolin-activated anion conductances display different ion selectivities. The anion selectivities of apical membrane conductances induced by thapsigargin or forskolin were measured directly in basolaterally permeabilized monolayers. Current-voltage relationships were examined in asymmetrical buffers containing either I or Cl as the major permeant anion (buffers A and B or C and D, Na⁺ present in buffers A-D at 137 mM, Table 1). Figure 5 shows that thapsigargin stimulated a greater increase in I_sc (anion conductance) in buffers containing I [G_apical = 2.4 ± 0.3 mS/cm² (n = 6); Fig. 5A] than in those containing Cl [G_apical = 0.8 ± 0.2 mS/cm² (n = 6); Fig. 5A]. In contrast, forskolin induced a greater increase in I_sc when buffers contained Cl rather than I as the major permeant ion [Cl: G_apical = 6 ± 0.9 mS/cm²; I: 3.3 ± 1.0 mS/cm² (n = 6)]. Both Cl and I currents, whether induced by thapsigargin or forskolin, were linear (Fig. 5, A and B).

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Effects of thapsigargin (A) and forskolin (B) on currents in basolaterally permeabilized T84 monolayers in presence of iodide solutions. Cytosol basolateral-to-apical reservoir Cl concentration ratio was 143.50/6.6.0 mM Cl, Na+ present in both sides at 137 mM (thapsigargin, ; forskolin, ). Cytosol basolateral-to-apical reservoir I concentration ratio for thapsigargin and forskolin was 136.89/6.60 mM I, Na+ present in both sides at 137 mM (thapsigargin, ; forskolin, ). Ordinates show difference current between stimulated and unstimulated cells. Tracings are representative of results from 6 different filters for each condition.

DIDS affects Cl conductances stimulated by both thapsigargin and forskolin. As shown in Fig. 1, the addition of DIDS to the apical compartment of intact T84 monolayers inhibited differentially transepithelial I_sc stimulated by thapsigargin or forskolin (100 vs. 40%, respectively). We found nearly the same differential inhibition of DIDS on thapsigargin- or forskolin-induced I_sc in basolaterally permeabilized monolayers studied in symmetrical buffers containing Cl as the major permeant ion (buffer E). DIDS inhibited completely I_sc elicited by thapsigargin [G_apical = 1.2 ± 0.2 mS/cm² vs. G_apical = 0.1 ± 0.1 mS/cm² with DIDS (n = 5); Fig. 6A]. In contrast, DIDS inhibited forskolin-induced Cl currents by 57% [G_apical = 4.1 ± 0.7 mS/ cm² vs. G_apical = 1.6 ± 0.8 mS/cm² with DIDS (n = 5); Fig. 6B]. DIDS had a similar effect on I_sc induced by 1 mM 8-bromo-cAMP or 10 mM forskolin applied to permeabilized monolayers (data not shown). This last result verifies the initial analysis using 100 µM forskolin. Thus the anion conductances observed are not attributable to an artifact seen only at 100 µM forskolin concentration. In the presence of transepithelial Cl gradients in basolaterally permeabilized monolayers, basal Cl currents were not affected by apical DIDS treatment [24 ± 4 vs. 20.3 ± 2.7 µA/cm² at +40 mV (n = 6)]; these results are in accordance with previous studies (8, 40). Thus the effect of thapsigargin on Cl current cannot easily be attributed to the swelling-activated Cl channel (29). Taken together, these data confirm results obtained on intact monolayers and provide further evidence that T84 apical membranes exhibit distinct Cl conductances.

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Effect of apical addition of 100 µM DIDS on 10 µM thapsigargin (A)- and 100 µM forskolin (B)-induced conductances in basolaterally permeabilized monolayers of T84 (ICl). In these experiments, apical and cytosol basolateral buffers contained Cl as major permeant ion, Na+ present in both sides at 0.5 mM (see Table 1). Ordinates show difference currents between stimulated and unstimulated cells in absence (thapsigargin, ; forskolin, ) and presence of DIDS (thapsigargin, ; forskolin, ). Tracings are representative of results from 5 different filters for each condition.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Effect of apical addition of 100 µM DIDS on 100 µM forskolin-induced conductances in basolaterally permeabilized T84 monolayers. Basolateral-to-apical reservoir Cl concentration ratio (in mM) was 136.89/6.60, Na+ present in both sides at 0.5 mM. Ordinate shows difference current between stimulated and unstimulated cells in absence () and presence of DIDS (). Tracings are representative of results from 10 different filters for each condition.

Low temperature affects Cl efflux stimulated by forskolin but not by thapsigargin. Previous studies have shown that activation of cAMP-dependent apical membrane Cl channels in permeabilized T84 monolayers was markedly attenuated by reducing incubation temperatures to 5°C (11, 21), perhaps via prevention of channel recruitment to the plasma membrane (11). The effects of temperature on Ca²⁺-sensitive channels have not been similarly studied. Thus, to further compare the apical membrane Cl conductances in T84 cell monolayers, we examined the effect of temperature on thapsigargin-induced and forskolin-induced I_sc in permeabilized monolayers. Permeabilized monolayers were incubated in symmetrical buffers containing Cl as the major permeant ion; monovalent cations were in the millimolar range (Cl: 143.25; K⁺: 0.43; Na⁺: 0.34 mM; see buffer E, Table 1) at 4 or 37°C for 90 min before treatment with thapsigargin or forskolin. At 4°C, T84 cells maintain intact tight junctions as determined by low transepithelial conductances (0.8 ± 0.4 mS/cm²). The ultimate magnitude of the thapsigargin-induced Cl secretory response was not affected at 4°C (Fig. 8A). In contrast, the forskolin-induced Cl secretory response was diminished by incubation at 4°C as previously reported (Fig. 8B and Ref. 11). Thus the activation of Cl conductances by thapsigargin and forskolin also differ in temperature sensitivity. Interestingly, the magnitude of the residual forskolin-activated current at 4°C was similar to that of the temperature-insensitive thapsigargin-stimulated current, and both currents were inhibited by DIDS to equivalent degree (Fig. 8, A and B). Permeabilized monolayers were incubated in symmetrical buffers containing Cl or I (see Table 1, buffer A or C); the ion selectivity of the DIDS-sensitive component of forskolin-induced current at 4°C is I > Cl (Fig. 9). These data demonstrate that the fraction of the forskolin-activated conductance which is DIDS sensitive and cold resistant may represent the same Cl conductance as that stimulated by thapsigargin.

View larger version (11K): [in this window] [in a new window]   Fig. 8.   Effect of temperature on thapsigargin (A)- and forskolin (B)-induced Cl conductances in basolaterally permeabilized T84 monolayers. In these experiments, apical and cytosol basolateral buffers contained Cl as major permeant anion, Na+ present in both sides at 0.5 mM (see Table 1). Ordinates show difference currents between stimulated and unstimulated cells treated with agonist at 37°C (thapsigargin, ; forskolin, ) and 4°C (thapsigargin, ; forskolin, ) in absence and presence of 100 µM DIDS; addition inhibits completely thapsigargin-stimulated current at 4°C (thapsigargin, A, ×; forskolin, B, ×). Tracings are representative of results from 3 different filters for each condition.

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Effects of 100 µM forskolin on currents in basolaterally permeabilized T84 monolayers measured in iodide solutions at 4°C. Basal Na+ present in both sides at 0.5 mM; basolateral-to-apical reservoir Cl concentration ratio (in mM) was 143.50/6.6.0, Na+ present in both sides at 137 mM (). Cytosol basolateral-to-apical reservoir I concentration ratio (in mM) was 136.89/6.60, Na+ present in both sides at 137 mM (). Ordinate shows difference currents between stimulated and unstimulated cells. Tracings are representative of results from 3 different filters for each condition.

Effects of thapsigargin and forskolin on intracellular cAMP and Ca²⁺ concentrations. In some epithelial cells, stimulation of cAMP also results in elevated intracellular Ca²⁺ (10). To investigate the possibility that forskolin may activate a Ca²⁺-sensitive Cl conductance by raising cytosolic Ca²⁺ (in addition to cAMP), the effect of forskolin treatment on [Ca²⁺]_i was examined. T84 cells grown on glass coverslips and loaded with the fluorescent indicator fura 2 were used for these studies (35). Cells incubated at 4 and 37°C were studied. Ca²⁺ response was observed in cells near the center of confluent islands. As shown in Fig. 10, resting levels of [Ca²⁺] were between 100 and 150 nM (133 ± 16 nM; n = 28 cells from 3 different coverslips). Application of forskolin at 37 or 4°C caused small but readily detected and consistently observed intracellular Ca²⁺ elevations (33 ± 3 nM; n = 28 cells from 3 different coverslips). As expected, stimulation with thapsigargin (10 µM) raised [Ca²⁺]_i approximately twofold above resting levels at both temperatures (at 37°C, [Ca²⁺]_i = 165 ± 12 nM vs. [Ca²⁺]_i = 150 ± 14 nM at 4°C; n = 28 cells from 3 different coverslips), although the time required to reach maximal [Ca²⁺] levels was greater at 4°C. Thapsigargin addition before or after forskolin addition produced similar increases in [Ca²⁺]_i. Thapsigargin had no effect on intracellular cAMP at either temperature (Fig. 11), whereas forskolin increased cAMP 30-fold above resting levels at 37°C and 3-fold above resting levels at 4°C. These data provide direct evidence that forskolin treatment can induce a small but detectable increase in intracellular Ca²⁺ at both 4 and 37°C. These data are consistent with previous studies in different cell types (10) but differ from previous studies performed in isolated T84 cells (43). The small Ca²⁺ response that we have found after forskolin addition might be due to the polarized state of T84. Thus the ability of forskolin to produce a small elevation in [Ca²⁺]_i may contribute to its induction of DIDS-sensitive temperature-insensitive anion conductance.

View larger version (18K): [in this window] [in a new window]   Fig. 10.   Effects of sequential addition of forskolin (Fors) and thapsigargin (Thap) on cytosolic [Ca2+] ([Ca2+]i) in T84. [Ca2+]i was measured at either 37 () or 4°C () in fura 2-loaded cells on glass coverslips. Ca2+ response was observed in cells near center of confluent "islands." A single representative experiment for each temperature (37 and 4°C) is given, and data are those of 14 cells from same coverslip. Tracings are representative of results from 3 different coverslips.

View larger version (10K): [in this window] [in a new window]   Fig. 11.   Effect of temperature on generation of cAMP by basolaterally permeabilized T84 cells treated with forskolin (100 µM) and with thapsigargin (10 µM). Apical and basolateral buffers contained Cl as major permeant ion. Experiments were carried out at 37 (A) or 4°C (B). Ctrl, resting level of cAMP; Thap and Fors, level of cAMP after addition of thapsigargin and forskolin, respectively. Data are presented as picomoles of cAMP generated per monolayer of T84 cells and represent 3 filters each assayed in duplicate.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The results of these studies show that selective permeabilization of basolateral membranes of T84 cells permits effective electrical isolation of apical membranes for analyses of anion conductances after agonist stimulation, as previously described (30, 31, 36). Such preparations avoid the complexities associated with regulation of various basolateral transporters, pumps, and channels that contribute to agonist-elicited I_sc and also permit analyses of apical membrane conductance in the polarized state. We find that apical anion conductance is stimulated by both thapsigargin and forskolin. However, the anion currents induced by forskolin or thapsigargin exhibit different magnitudes and anion selectivities. Furthermore, they are differentially sensitive to the anion channel blocker DIDS and cold temperature, indicating that T84 cells exhibit two forms of apical membrane anion channels. Our results also show that the ion selectivity (Cl < I) and temperature dependence of the DIDS-sensitive component of the forskolin-induced I_sc are similar to the ion selectivity and temperature dependence of the thapsigargin-induced I_sc. Thus the forskolin response appears to represent apical anion conductance of two types, one of which has features common to the response elicited by thapsigargin.

The T84 colon carcinoma cell has frequently been used as a model for Cl secretion, and considerable efforts have been directed at defining Cl channels by using different methods such as patch-clamp, microelectrode, or permeabilization techniques. Whole cell voltage-clamp studies with T84 cells reveal two different Cl conductances, one Ca²⁺ dependent and the other one cAMP dependent (1, 9). In addition, the anion selectivity of the Ca²⁺-stimulated Cl conductance in T84 cells differs from the that of cAMP-dependent Cl conductance (1). The use of the ¹²⁵I efflux technique in subconfluent cells and recently on T84 cell monolayers also suggests that Ca²⁺ and cAMP induce two different anion efflux pathways (5, 16). In contrast, results from studies using a technique in which the basolateral membranes of confluent T84 cells are permeabilized by nystatin are contradictory (1, 22). A speculative conclusion has been made that Ca²⁺-stimulated Cl conductances observed in cells grown on permeable supports are absent in cells grown on glass coverslips (1). On the other hand, studies using intracellular microelectrodes or using fluorescent dye (SPQ) have detected evidence favoring the presence of both Ca²⁺- and cAMP-stimulated Cl conductances in T84 cells grown to confluence on permeable supports (2, 22).

The present study shows that intracellular elevation of Ca²⁺ or cAMP activates distinct Cl conductances in basolaterally permeabilized monolayers. In symmetrical buffers, the conductances stimulated by forskolin or by Ca²⁺ each have linear relationships over the range ± 80-mV membrane potential. The conductances, however, display different biophysical characteristics. Experiments on the halide permselectivity between Cl and I indicate that G_anion(cAMP) has the permeability sequence, Cl > I. These results are in accordance with previous studies (1, 6) and suggest that the observed conductance probably represents transport through CFTR (42). In contrast, G_anion(Ca) has the opposite permeability sequence, I > Cl. We also show that G_Cl(cAMP) is only partially inhibited by DIDS in contrast to the G_Cl(Ca), which is completely inhibited by the addition of DIDS. Thus these data are consistent with the presence of at least two distinct Cl conductances in T84 monolayers, as also suggested by previous patch-clamp and microelectrode studies (2, 9).

In previous studies, apical Cl conductances in confluent T84 monolayers grown on permeable supports were measured by permeabilizing basolateral membranes with nystatin and then applying an apically directed gradient of Cl across the monolayer with the voltage clamped to 0 mV. Under these conditions in the presence of amphotericin rather than nystatin, we observed similar results to those reported, finding that Cl secretion was activated by the addition of forskolin but not thapsigargin (1). When, however, the transmonolayer potential was clamped at +15 mV, currents were activated both by thapsigargin and forskolin. Our results suggest the possibility that there is less discrimination between Na⁺ and Cl currents at 0 mV than at other membrane potentials. Thus, in our studies when higher driving forces were applied, a thapsigargin-induced Cl conductance was readily apparent. The Cl conductances induced by thapsigargin, however, were much smaller than those observed for Cl transport through forskolin-induced conductances. In other words, when stronger driving forces were applied, Cl transport through the small thapsigargin-induced conductance became detectable.

The finding that thapsigargin induced an elevation of the intracellular Ca²⁺ without affecting cAMP levels raises the possibility that elevations of intracellular Ca²⁺ activate apical Cl conductances in T84 cells. Because low temperatures did not affect the Ca²⁺-dependent Cl conductance or the intracellular Ca²⁺ increase, it is reasonable to suggest that Ca²⁺ could directly activate this Cl conductance. In addition, our observation that the cAMP-stimulated Cl conductance was partially DIDS sensitive in basolaterally permeabilized as well as in intact monolayers raises the possibility that cAMP may stimulate two distinct Cl conductances. We show that the temperature-insensitive component of forskolin-induced current, like the thapsigargin-induced current, was DIDS sensitive and displayed a permselectivity of I  Cl. These results are consistent with the possibility that a fraction of the Cl conductance induced by forskolin represents transport through a Ca²⁺-activated, DIDS-sensitive Cl channel. In T84 cells, forskolin can thus activate, in addition to CFTR, the Ca²⁺-sensitive conductance activated by thapsigargin alone. These results are consistent with previous studies showing that cAMP activates a Cl conductance which is DIDS sensitive (20, 26) and shows preference for I over Cl. Our data, however, do not allow us to distinguish the possibility that T84 cells may contain two distinct DIDS-sensitive conductances, one activated by forskolin and the other by thapsigargin.

Our data do not identify a mechanism by which forskolin might activate the Ca²⁺-sensitive conductance. Although forskolin induces a small increase in intracellular Ca²⁺ at both 4 and 37°C, these data do not exclude the possibility that cAMP may activate the DIDS-sensitive conductance directly or through activation of protein kinase(s).

From the results of these and previous studies (2, 5, 16, 22, 33), we propose the following working model of Cl secretion in T84 cells via two apical conductive pathways (Fig. 12). The first is activated by cAMP and not sensitive to inhibition by DIDS. This conductance exhibits temperature sensitivity (inhibition at 4°C), a large Cl conductance, ion selectivity of Cl > I, and presumably represents CFTR. The second is activated both by thapsigargin and by forskolin. This conductance exhibits little or no temperature sensitivity, a much smaller Cl conductance, and ion selectivity of I > Cl. Forskolin may activate the DIDS-sensitive conductance via cAMP or Ca²⁺ second messengers. Further studies, however, are needed to identify the mechanism by which forskolin may activate this Ca²⁺-sensitive conductance.

View larger version (22K): [in this window] [in a new window]   Fig. 12.   Cellular model for regulation of Cl secretion in T84. Ca2+ activates a Cl channel that is completely inhibited by DIDS and insensitive to cold temperature. Forskolin activates via protein kinase A (PKA), a cold-temperature-sensitive Cl channel only partially inhibited by DIDS, as well as a second, temperature-insensitive anion channel that is completely DIDS inhibited and of different anion selectivity. Activation of this 2nd channel occurs by an unknown pathway that may involve Ca2+ and/or cAMP and/or other second messengers. PCl and PI, Cl and I permeabilities; CFTR, cystic fibrosis transmembrane conductance regulator.