cAMP-dependent activation of small-conductance Cl- channels in HT29 colon carcinoma cells

SECRETION AND ABSORPTION BY COLONIC CRYPTS

The intestines play an important role in the absorption and secretion of nutrients. The colon is the final area for recapturing electrolytes and water prior to excretion, and in order to maintain this electrolyte homeostasis, a complex interaction between secretory and absorptive processes is necessary. Until recently it was thought that secretion and absorption were two distinct processes associated with either crypts or surface cells, respectively. Recently it was demonstrated that both the surface and crypt cells can perform secretory and absorptive functions and that, in fact, these functions can be going on simultaneously. This issue is important in the complexities associated with secretory diarrhea and also in attempting to develop treatment strategies for intestinal disorders. Here, we update the model of colonic secretion and absorption, discuss new issues of transporter activation, and identify some important new receptor pathways that are important modulators of the secretory and absorptive functions of the colon.

The cystic fibrosis transmembrane conductance regulator and its function in epithelial transport

CF is a well characterized disease affecting a variety of epithelial tissues. Impaired function of the cAMP activated CFTR Cl- channel appears to be the basic defect detectable in epithelial and non-epithelial cells derived from CF patients. Apart from cAMP-dependent Cl- channels also Ca2+ and volume activated Cl- currents may be changed in the presence of CFTR mutations. This is supported by recent additional findings showing that different intracellular messengers converge on the CFTR Cl- channel. Analysis of the ion transport in CF airways and intestinal epithelium identified additional defects in Na+ transport. It became clear recently that mutations of CFTR may also affect the activity of other membrane conductances including epithelial Na+ channels, KvLQT-1 K+ channels and aquaporins (Fig. 7). Several additional, initially unexpected effects of CFTR on cellular functions, such as exocytosis, mucin secretion and regulation of the intracellular pH were reported during the past. Taken together, these results clearly indicate that CFTR not only acts as a cAMP regulated Cl- channel, but may fulfill several other cellular functions, particularly by regulating other membrane conductances. Failure in CFTR dependent regulation of these membrane conductances is likely to contribute to the defects observed in CF. Currently, no general concept is available that can explain how CFTR controls this variety of cellular functions. Further studies will have to verify whether direct protein interaction, specific effects on membrane turnover, changes of the intracellular ion concentration or additional proteins are involved in these regulatory loops. At the end of this review one cannot share the provocative and reassuring title "CFTR!" of a review written a few years ago [114]. Today one might rather finish with the statement "CFTR?".

K+ and Cl- conductances in the distal colon of the rat

1. K+ and Cl- conductances and their putative regulation have been characterized in the rat colonic epithelium by Ussing-chamber experiments, whole-cell and single-channel patch-clamp recordings. 2. The apical Cl- conductance is under the control of intracellular cAMP. An increase in the concentration of this second messenger induces transepithelial Cl- secretion due to the activation of an apical 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB)- and glibenclamide-sensitive Cl- conductance. 3. In addition to the apical Cl- conductance, the basolateral membrane is equipped with Cl- channels. They are stimulated by cell swelling and play a role in cell volume regulation and transepithelial Cl- absorption. 4. The basolateral K+ conductance is under the dominant control of intracellular Ca2+. An increase in the cytosolic Ca2+ concentration leads to the opening of basolateral K+ channels, which causes a hyperpolarization of the cell membrane, indirectly supporting Cl- secretion owing to an increase in the driving force for Cl- exit. The predominant effect of cAMP on the basolateral K+ conductance is an inhibitory one, probably due to a decrease in the intracellular Ca2+ concentration. 5. The apical K+ conductance, which is involved in transepithelial K+ secretion, is stimulated by an increase in the intracellular Ca2+ concentration. 6. The differential regulation of apical and basolateral ion conductances in the epithelium of the rat distal colon provides an interesting example for the mechanisms underlying vectorial transport of ions across polarized cells.

Patch clamp on the luminal membrane of exocrine gland acini from frog skin (Rana esculenta) reveals the presence of cystic fibrosis transmembrane conductance regulator-like Cl- channels activated by cyclic AMP

Chloride channels in the luminal membrane of exocrine gland acini from frog skin (Rana esculenta) constituted a single homogeneous population. In cell-attached patches, channels activated upon exposure to isoproterenol, forskolin, or dibutyryl-cAMP and isobutyl-1-methyl-xanthine rectified in the outward direction with a conductance of 10.0 +/- 0.4 pS for outgoing currents. Channels in stimulated cells reversed at 0 mV applied potential, whereas channels in unstimulated cells reversed at depolarized potentials (28.1 +/- 6.7 mV), indicating that Cl- was above electrochemical equilibrium in unstimulated, but not in stimulated, cells. In excised inside-out patches with 25 mM Cl- on the inside, activity of small (8-pS) linear Cl--selective channels was dependent upon bath ATP (1.5 mM) and increased upon exposure to cAMP-dependent protein kinase. The channels displayed a single substate, located just below 2/3 of the full channel amplitude. Halide selectivity was identified as PBr > PI > PCl from the Goldman equation; however, the conductance sequence when either halide was permeating the channel was GCl > GBr >> GI. In inside-out patches, the channels were blocked reversibly by 5-nitro-2-(3-phenylpropylamino)benzoic acid, glibenclamide, and diphenylamine-2-carboxylic acid, whereas 4, 4-diisothiocyanatostilbene-2,2-disulfonic acid blocked channel activity completely and irreversibly. Single-channel kinetics revealed one open state (mean lifetime = 158 +/- 72 ms) and two closed states (lifetimes: 12 +/- 4 and 224 +/- 31 ms, respectively). Power density spectra had a double-Lorentzian form with corner frequencies 0.85 +/- 0.11 and 27.9 +/- 2.9 Hz, respectively. These channels are considered homologous to the cystic fibrosis transmembrane conductance regulator Cl- channel, which has been localized to the submucosal skin glands in Xenopus by immunohistochemistry (Engelhardt, J.F., S.S. Smith, E. Allen, J.R. Yankaskas, D.C. Dawson, and J.M. Wilson. 1994. Am. J. Physiol. 267: C491-C500) and, when stimulated by cAMP-dependent phosphorylation, are suggested to function in chloride secretion.

The role of K+ channels in colonic Cl- secretion

Cl- secretion in the rat colonic crypt base cell (bc) requires the coordinated (a) opening of Cl- channels in the luminal membrane; (b) activation of the Na+2Cl-K+ cotransporter; (c) enhanced conductive K+ exit from the cell; and (d) increased pumping by the (Na+ + K+)-ATPase. In this study we focus on the importance of conductive K+ exit. After stimulation with the cholinergic agonist carbachol (CCH, 0.1-10 mumol/l) bc respond with a marked increase in whole cell (wc) conductance and a hyperpolarization of the membrane voltage (Vm). This is paralleled by a marked increase in the (Cl- secretory) short-circuit current (Isc) in Ussing chamber studies of the intact distal colon. Current evidence favors the view that CCH, via IP3, enhances cytosolic Ca2+ activity, and that Ca2+ increases the open probability of Cl- channels indirectly and that of K+ channels directly. After stimulation with PGE2 bc also enhance the wc conductance, but this is paralleled by a marked depolarization of Vm. Again these effects correspond to a marked increase in (Cl- secretory) Isc. The depolarization and enhanced wc conductance is partly due to the activation of Cl- channels. However, current evidence suggests that these effects on Cl- channels are paralleled by an activation of K+ channels. The chromanol 293B, by inhibiting these K+ channels specifically, abolishes PGE2-induced Cl- secretion completely, but has no effect on basal K+ conductance or on CCH-induced Cl- secretion. CCH apparently activates a Ca(2+)-dependent K+ channel with a conductance of 10-20 pS, whilst PGE2 (or cAMP) activate a much smaller K+ channel. Only the latter K+ channel can be inhibited by 293B in excised patches. Noise analysis suggests that this K+ channel has a conductance of < 3 pS and fast kinetics. The complete 293B induced inhibition of Cl- secretion caused by PGE2 can be explained by the fact that PGE2 induces a marked depolarization and that this depolarization reduces the basal K+ conductance. Current evidence suggests that this inhibition of the basal K+ conductance is caused by a depolarization induced inhibition of Ca2+ entry.

Carbachol induces oscillations in membrane potential and intracellular calcium in a colonic tumor cell line, HT-29

The patch-clamp technique was used to study the effects of carbachol (CCh) on HT-29 cells. During CCh exposure, the cells (n = 23) depolarized close to the equilibrium potential for Cl- (E(Cl-); -48 mV) and the membrane potential then started to oscillate (16/23 cells). In voltage-clamp experiments, similar oscillations in whole cell currents could be demonstrated. The whole cell conductance increased from 225 +/- 25 pS in control solution to 6,728 +/- 1,165 pS (means +/- SE, n = 17). In substitution experiments (22 mM Cl- in bath solution, E(Cl-) = 0 mV), the reversal potential changed from -41.6 +/- 2.2 mV (means +/- SE, n = 9) to -3.2 +/- 2.0 mV (means +/- SE, n = 7). When the cells were loaded with the calcium-sensitive fluorescent dye, fluo 3, and simultaneously patch clamped, CCh caused a synchronous oscillating pattern of fluorescence and membrane potential. In cell-attached patches, the CCh-activated currents reversed at a relative membrane potential of 1.9 +/- 3.7 mV (means +/- SE, n = 11) with control solution in the pipette and at 46.2 +/- 5.3 mV (means +/- SE, n = 10) with a 15 mM Cl- solution in the pipette. High K+ (144 mM) did not change the reversal potential significantly (P < or = 0.05, n = 8). In inside-out patches, calcium-dependent Cl- channels could be demonstrated with a conductance of 19 pS (n = 7). It is concluded that CCh causes oscillations in membrane potential that involve calcium-dependent Cl- channels and a K+ permeability.

Heterogeneity of chloride channels in the apical membrane of isolated mitochondria-rich cells from toad skin

The isolated epithelium of toad skin was disintegrated into single cells by treatment with collagenase and trypsine. Chloride channels of cell-attached and excised inside-out apical membrane-patches of mitochondria-rich cells were studied by the patch-clamp technique. The major population of Cl- channels constituted small 7-pS linear channels in symmetrical solutions (125 mM Cl-). In cell-attached and inside-out patches the single channel i/V-relationship could be described by electrodiffusion of Cl- with a Goldmann-Hodgkin-Katz permeability of, PCl = 1.2 x 10(-14) - 2.6 x 10(-14) cm3. s-1. The channel exhibited voltage-independent activity and could be activated by cAMP. This channel is a likely candidate for mediating the well known cAMP-induced transepithelial Cl- conductance of the amphibian skin epithelium. Another population of Cl- channels exhibited large, highly variable conductances (upper limit conductances, 150-550 pS) and could be activated by membrane depolarization. A group of intermediate-sized Cl(-)-channels included: (a) channels (mean conductance, 30 pS) with linear or slightly outwardly rectifying i/V-relationships and activity occurring in distinct "bursts," (b) channels (conductance-range, 10-27 pS) with marked depolarization-induced activity, and (c) channels with unresolvable kinetics. The variance of current fluctuations of such "noisy" patches exhibited a minimum close to the equilibrium-potential for Cl-. With channels occurring in only 38% of sealed patches and an even lower frequency of voltage-activated channels, the chloride conductance of the apical membrane of mitochondria-rich cells did not match quantitatively that previously estimated from macroscopic Ussing-chamber experiments. From a qualitative point of view, however, we have succeeded in demonstrating the existence of Cl-channels in the apical membrane with features comparable to macroscopic predictions, i.e., activation of channel gating by cAMP and, in a few patches, also by membrane depolarization.

The membrane transporters regulating epithelial NaCl secretion

Ten years ago, the basic principles operating in one specific, albeit non-mammalian, exocrine gland, the rectal gland of Squalus acanthias, were described in detail. The concept emerging from these studies appeared applicable to almost any other exocrine gland, because it involved membrane transporters which are also present in mammalian epithelial cells. Meanwhile, it has become clear that the mechanisms of NaCl secretion are diverse: the mechanisms of NaCl uptake; the ion channels involved; and also the mechanisms of hormonal control. Nevertheless, several steps in NaCl secretion still appear to be uniform: (1) several signalling pathways converge and act cooperatively, (2) one primary regulatory step is the upregulation of the luminal Cl- conductance, (3) secondarily active NaCl uptake mechanisms are upregulated, (4) increasing evidence links NaCl secretion to membrane trafficking and (5) the entire machinery seems to be primed to secure cellular homeostasis in terms of cytosolic ion concentrations. This brief review summarizes the mechanisms of control of NaCl secretion. The major issues addressed are the NaCl uptake mechanisms, the ion channels involved and the cellular mechanisms coordinating secretion. The major NaCl secreting cells discussed here will be the respiratory epithelial cells, the exocrine cells of pancreatic acini and the cells of colonic crypts.

Attenuation of stimulated Ca2+ influx in colonic epithelial (HT29) cells by cAMP

In HT29 colonic epithelial cells agonists such as carbachol (CCH) or ATP increase cytosolic Ca2+ activity ([Ca2+]i) in a biphasic manner. The first phase is caused by inositol 1,4,5-trisphophate-(Ins P3-) mediated Ca2+ release from their respective stores and the second plateau phase is mainly due to stimulated transmembraneous Ca2+ influx. The present study was undertaken to examine the effect of increased adenosine 3',5'-cyclic monophosphate (cAMP) (forskolin 10 micromol/l = FOR) on the Ca2+ transient in the presence of CCH (100 micromol/l). In unpaired experiments it was found that FOR induced a depolarization and reduced cytosolic Ca2+ ([Ca2+]i, measured as the fura-2 fluorescence ratio 340/380 nm) significantly. Dideoxyforskolin had no such effect. The effect of FOR was abolished when the cells were depolarized by a high-K+ solution. In further paired experiments utilizing video imaging in conjunction with whole-cell patch-clamp, [Ca2+]i was monitored separately for the patch-clamped cell and three to seven neighbouring cells. In the presence of CCH, FOR reduced [Ca2+]i uniformly from a fluorescence ratio (345/380) of 2.9 +/- 0.12 to 1.8 +/- 0.07 in the patch-clamped cell and its neighbours (n = 48) and depolarized the membrane voltage (Vm) of the patch-clamped cells significantly and reversibly from -54 +/- 7.4 to -27 +/- 5.9 mV (n = 6). In additional experiments Vm was depolarized by 15-54 mV by various increments in the bath K+ concentration. This led to corresponding reductions in [Ca2+]i. Irrespective of the cause of depolarization (high K+ or FOR) there was a significant correlation between the change in Vm and change in [Ca2+]i. These data indicate that the cAMP-mediated attenuation of Ca2+ influx is caused by the depolarization produced by this second messenger.

Clusters of Cl- channels in CFTR-expressing Sf9 cells switch spontaneously between slow and fast gating modes

The Sf9 insect Spodoptora frugiperda cell line was used for heterologous expression of the cloned human cystic fibrosis transmembrane conductance regulator (CFTR) cDNA, or the cloned beta-galactosidase gene, using the baculovirus Autographa califonica as the infection vector. Using application of the patch-clamp technique, evidence for functional expression of CFTR was obtained according to the following three criteria. Firstly, whole-cell currents recorded 2 days after infection with CFTR revealed a statistically significant increase of membrane conductance, approximately 25 times above that of mock-infected control cells, with the reversal potential of the major current component being governed by the chloride equilibrium potential (ECl). Secondly, in contrast to uninfected cells and cells infected with beta-galactosidase, the membrane conductance to chloride of CFTR-injected cells was stimulated by cytosolic adenosine 3',5'-cyclic monophosphate (cAMP), which was raised by exposing the cells to 10 microM forskolin. Thirdly, recordings of currents through single channels in excised outside-out membrane patches of CFTR-infected cells revealed channels which were clearly different from the native insect chloride channel. Excised outside-out patches of CFTR-infected and forskolin-stimulated cells exhibited wave-like gating kinetics of well-resolved current transitions. All-point Gaussian distributions revealed contributions from several (five to nine) identical channels. Such channels, in excised outside-out patches, studied with a pipette [Cl-] = 40 mM and a bath [Cl-] = 150 mM, rectified the current in agreement with simple electrodiffusion and with a single-channel Goldman-Hodgkin-Katz permeability, PCl = 1. 34 x 10(-14) +/- 0.23 x 10(-14 )cm3/s (n = 5), corresponding to a physiological single-channel conductance of 2.8 +/- 0.5 pS (VM = ECl) and a limiting conductance, gamma150/150, = 7.7 +/- 1.3 pS ([Cl-]Bath = [Cl-]Cell = 150 mM). Currents recorded from multichannel excised outside-out patches could shift from the above mode of resolvable unitary conductance transitions to one which was too fast to reveal the dwell-times of closed and open states. During periods characterized by noisy currents, the variance (sigma2) of current fluctuations about their stationary mean value depicted a U-shaped function of membrane potential, with a minimum value at a pipette potential where the chloride current was shown to be zero. Thus, it can be concluded that the current fluctuations are caused by fast gating of channels specific for chloride ions. Switching back and forth between the two gating modes of clusters of chloride channels occurred from moment to moment in excised patches when the membrane potential was held at a constant value indicating cooperative gating as a result of interaction between neighbouring chloride channels.

N-Acetyl-L-cysteine and its derivatives activate a Cl- conductance in epithelial cells

N-Acetyl-L-cysteine (NAC) is a widely used mucolytic drug in patients with a variety of respiratory disorders including cystic fibrosis (CF). The beneficial effects of NAC are empirical and the exact mechanism of action in the airways remains obscure. In the present study we examined the effects on whole-cell (wc) conductance (Gm) and voltage (Vm) of NAC and the congeners S-carboxymethyl-L-cysteine (CMC) and S-carbamyl-L-cysteine (CAC) and L-cysteine in normal and CF airway epithelial cells. L-Cysteine (1 mmol/l) had no detectable effect. The increase in Gm (delta Gm) by the other compounds was concentration dependent and was (all substances at 1 mmol/l) 3.8 +/- 1.4 nS (NAC; n = 11), 4.2 +/- 1.0 nS (CMC; n = 16) and 3.8 +/- 1.6 nS (CAC; n = 18), respectively. The changes in Gm were paralleled by an increased depolarization (delta Vm) when extracellular Cl- concentration was reduced to 34 mmol/l: under control conditions = -4.1 +/- 2.1 versus 10.2 +/- 2.1 mV in the presence of NAC, CMC, CAC (n = 36). In the presence of NAC, CMC and CAC, the reduction in Cl- concentration was paralleled by a reduction of Gm by 2.1 +/- 0.4 nS (n = 35), indicating that all substances acted by increasing the Cl- conductance. Analysis of intracellular pH did not reveal any changes by any of the compounds (1 mmol/l). A Cl- conductance was also activated in HT29 colonic carcinoma and CF tracheal epithelial (CFDE) cells but not in CFPAC-1 cells, which do not express detectable levels of delta F508-CFTR, suggesting that the presence of CFTR may be a prerequisite for the induction of Cl- currents. Next we examined the ion currents in Xenopus oocytes microinjected with CFTR-cRNA. Water-injected oocytes did not respond to activation by forskolin and 3-isobutyl-1-methylxanthine (IBMX) (delta Gm = 0.08 +/- 0.04 microS; n = 10) and no current was activated when these oocytes were exposed to NAC or CMC. In contrast, in CFTR-cRNA-injected oocytes Gm was enhanced when intracellular adenosine 3',5'-cyclic monophosphate (cAMP) was increased by forskolin and IBMX (Gm = 4.5 +/- 1.3 microS; n = 8). Gm was significantly increased by 0.74 +/- 0.2 microS (n = 11) and 0.46 +/- 0.1 microS (n = 10) when oocytes were exposed to NAC and CMC, respectively (both 1 mmol/l). In conclusion, NAC and its congeners activate Cl- conductances in normal and CF airway epithelial cells and hence induce electrolyte secretion which may be beneficial in CF patients. CFTR appears to be required for this response in an as yet unknown fashion.

Calcium is not involved in the cAMP-mediated stimulation of Cl- conductance in the apical membrane of Necturus gallbladder epithelium

The permeability properties of the forskolin-stimulated Cl- conductance in the apical membrane of Necturus gallbladder epithelium and the possible participation of intracellular Ca2+ in its stimulation have been investigated. The anion selectivity sequence as derived from biionic potential measurements (SCN- > I- approximately NO3- > Br- > Cl- >> ISE-) differed from the sequence derived from measurements of apical membrane resistance (NO3- approximately Br- approximately Cl- > SCN- > I- approximately ISE-). Accordingly, the conductance was inhibited by SCN- and I- which, from the potential measurements, appeared to be more permeable than Cl-. This finding agrees with observations of the cystic fibrosis transmembrane conductance regulator (CFTR) Cl- channel reported recently. However, none of the commonly used Cl- channel blockers, such as 5-nitro-2-(3-phenylpropylamino)-benzoic acid (NPPB), 4,4'-diisothiocyanatostilbene-2,2'-disulphonic acid (DIDS), anthracene-9-carboxylic acid (9-AC) and glibenclamide reduced this conductance in Necturus gallbladder. In contrast to the situation in most other epithelia, elevation of intracellular Ca2+ concentration ([Ca2+]i) by ionomycin stimulated only K+ conductance and not that of Cl- in the apical cell membrane. Chelation of intracellular Ca2+ did not prevent the stimulation of Cl- conductance by forskolin. This indicates that [Ca2+]i does not have even a permissive role in the cyclic adenosine monophosphate-(cAMP)-mediated stimulation process, as would have been expected if exocytosis was involved. Further evidence against the involvement of exocytosis in the stimulation process came from the observation that the stimulation was not associated with an increase in apical membrane capacitance and was not suppressed by disruption of the cytoskeleton by preincubation of the tissue with cytochalasin D. The data indicate that Necturus gallbladder epithelium contains homologues of the CFTR Cl- channel which reside permanently in the apical cell membrane and which can be stimulated by a cAMP-dependent phosphorylation process without involvement of cell Ca2+ or exocytosis.

Cl- channel inhibition by glibenclamide is not specific for the CFTR-type Cl- channel

As long as the question of which channels are responsible for cAMP-mediated epithelial Cl- secretion remains unsolved, it is still important to search for specific inhibitors that might help to relate macroscopic to microscopic events. Following the report by Sheppard and Welsh (J Gen Physiol 100: 573, 1992) that glibenclamide inhibits whole-cell Cl- currents in genetically manipulated fibroblasts expressing the cystic fibrosis transmembrane conductance regulator (CFTR), we have studied the effect of glibenclamide on different types of Cl- channels of HT29 and T84 cells at the single-channel level. Our results confirm that micromolar concentrations of glibenclamide inhibit the linear, low-conductance Cl-channel, which appears to represent CFTR and show that the inhibition results from a typical flicker block. However, the same concentrations of glibenclamide inhibit also the outwardly rectifying intermediate conductance Cl- channel which, potentially, may contribute to transepithelial Cl- secretion.

A new class of inhibitors of cAMP-mediated Cl- secretion in rabbit colon, acting by the reduction of cAMP-activated K+ conductance

Previously we have shown that arylaminobenzoates like 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB), which are very potent inhibitors of NaCl absorption in the thick ascending limb of the loop of Henle, are only poor inhibitors of the cAMP-mediated secretion of NaCl in rat colon. This has prompted our search for more potent inhibitors of NaCl secretion in the latter system. The chromanole compound 293 B inhibited the equivalent short-circuit current (Isc) induced by prostaglandin E2 (n = 7), vasoactive intestinal polypeptide (VIP, n = 5), adenosine (n = 3), cholera toxin (n = 4) and cAMP (n = 6), but not by ionomycin (n = 5) in distal rabbit colon half maximally (IC50) at 2 mumol/l from the mucosal and at 0.7 mumol/l from the serosal side. The inhibition was reversible and paralleled by a significant increase in transepithelial membrane resistance [e.g. in the VIP series from 116 +/- 16 omega.cm2 to 136 +/- 21 omega.cm2 (n = 5)]. A total of 25 derivatives of 293 B were examined and structure activity relations were obtained. It was shown that the racemate 293 B was the most potent compound within this group and that its effect was due to the enantiomer 434 B which acted half maximally at 0.25 mumol/l. Further studies in isolated in vitro perfused colonic crypts revealed that 10 mumol/l 293 B had no effect on the membrane voltage across the basolateral membrane (Vbl) in non-stimulated crypt cells: -69 +/- 3 mV versus -67 +/- 3 mV (n = 10), whilst in the same cells 1 mmol/l Ba2+ depolarised Vbl significantly. However, 293 B depolarised Vbl significantly in the presence of 1 mumol/l forskolin: -45 +/- 4 mV versus -39 +/- 5 mV (n = 7). Similar results were obtained with 0.1 mmol/l adenosine. 293 B depolarised Vbl from -40 +/- 5 mV to -30 +/- 4 mV (n = 19). This was paralleled by an increase in the fractional resistance of the basolateral membrane. VIP had a comparable effect. The hyperpolarisation induced by 0.1 mmol ATP was not influenced by 10 mumol/l 293 B: -75 +/- 6 mV versus -75 +/- 6 mV (n = 6). Also 293 B had no effect on basal K+ conductance (n = 4). Hence, we conclude that 293 B inhibits the K+ conductance induced by cAMP. This conductance is apparently relevant for Cl- secretion and the basal K+ conductance is insufficient to support secretion.

Effects of P-glycoprotein expression on cyclic AMP and volume-activated ion fluxes and conductances in HT-29 colon adenocarcinoma cells

The tissue distribution of P-glycoprotein (Pgp) and the structurally related cystic fibrosis transmembrane conductance regulator (CFTR) is apparently mutually exclusive, particularly in epithelial; where one protein is expressed the other is not. To study the possible function(s) of Pgp and its potential effects on CFTR expression in epithelia, HT-29 colon adenocarcinoma cells, which constitutively express CFTR, were pharmacologically adapted to express the classical multidrug resistance (MDR) phenotype (Pgp+). Concomitant with the appearance of Pgp and MDR phenotype (drug resistance, reduced drug accumulation and increased drug efflux), CFTR levels and cAMP-stimulated Cl conductances were markedly decreased compared to wild-type HT-29 (Pgp-) cells (as shown using the whole cell patch clamp technique). Removal of drug pressure led to the gradual decrease in Pgp levels and MDR phenotype, as evidenced by increased rhodamine 123 accumulation (Pgp-Rev). Concomitantly, CFTR levels and cAMP-stimulated Cl- conductances increased. The cell responses of Pgp/Rev cells were heterogeneous with respect to both Pgp and CFTR functions. We also studied the possible contribution to Pgp to hypotonically activated (HCS) ion conductances. K+ and Cl- effluxes from Pgp- cells were markedly increased by HCS. This increase was twice as high as that induced by the cation ionophore gramicidin; it was blocked by the Cl- channel blocker DIDS (4,4'-disothiocyano-2,2'-disulfonic stilbene) and required extracellular Ca2+. In Pgp+ cells, the HCS-induced fluxes were not significantly different from those of Pgp- cells. Verapamil (10 microM), which caused 80% reversal of Pgp-associated drug extrusion, failed to inhibit the HCS-evoked Cl- efflux of Pgp+ cells. Similarly, HCS increased Cl- conductance to the same extent in Pgp-, Pgp+ and Pgp-Rev cells. Verapamil (100 microM), but not 1,9-dideoxyforskolin (50 and 100 microM), partially inhibited the HCS-evoked whole cell current (WCC) in all three lines. Since the inhibition by verapamil was not detected in the presence of the K+ channel blocker Ba2+ (3 mM), it is suggested that verapamil affects K+ and not Cl- conductance. We conclude that hypotonically activated Cl- and K+ conductances are similar in HT-29 cells irrespective of Pgp expression. Expression of high levels of Pgp in HT-29 cells confers no physiologically significant capacity for cell volume regulation.

cAMP-dependent activation of ion conductances in bronchial epithelial cells

The cAMP-dependent activation of Cl- channels was studied in a bronchial epithelial cell line (16HBE14o-) in fast and slow whole-cell, and cell-attached patch-clamp experiments. The cells are known to express high levels of cystic fibrosis transmembrane conductance regulator mRNA and protein. Isoproterenol, forskolin and histamine (all 10 mumol/l) reversibly and significantly depolarized the membrane voltage (Vm) and increased the whole-cell Cl- conductance significantly by 34.0 +/- 0.9 (n = 3), 18.1 +/- 2.7 (n = 50), and 25 +/- 4.5 (n = 37) nS respectively. The effect of histamine was blocked by cimetidine (10 mumol, n = 5) but not by diphenhydramine (10 mumol/l, n = 4), which suggests binding of histamine to H2 receptors. The forskolin-induced current was not inhibited significantly by 4,4'-diisothiocyanatostilbene-2,2'-disulphonic acid (0.5 mmol/l, n = 9) nor glibenclamide (10 mumol/l, n = 3) and had an anion-permeability sequence of Cl = Br- > I- (n = 9). In cell-attached recordings forskolin (10 mumol/l) increased the conductance of the patched membrane from 65.5 +/- 13.6 pS to 150.8 +/- 33.2 pS (n = 30). Although the conductance was increased significantly, clear ion channel events occurring in parallel with the current activation were not detected in the cell-attached membrane. In 4 out of 30 cell-attached recordings single-channel currents were observed. These channels, with a single-channel conductance of about 6 pS, were already active before forskolin was added. No effect of forskolin on the channel amplitude, open probability or kinetics of these channels was observed.(ABSTRACT TRUNCATED AT 250 WORDS)

Forskolin and PMA pretreatment of HT29 cells alters their chloride conductance induced by cAMP, Ca2+ and hypotonic cell swelling

HT29 cells were preincubated with forskolin (10(-5) mol/l, FORHT) or phorbol 12-myristate 13-acetate (PMA) (10(-7) mol/l, PMAHT) for 20 h, which has been shown previously and is also shown here, to upregulate and downregulate, respectively, the expression of the cystic fibrosis transmembrane conductance regulator (CFTR). CFPAC-1 cells underwent the same protocols. HT29 cells were examined by slow (SWC) and fast (FWC) whole-cell patch-clamp techniques. The results of SWC and FWC were indistinguishable and were pooled. CFPAC-1 cells were examined with FWC. The membrane voltage (V) of FORHT was -41.8 +/- 1.4 mV (n = 77) and that of PMAHT was -43.6 +/- 2.4 mV (n = 76). The conductance (G) of FORHT (9.4 +/- 0.9 nS, n = 77) was significantly larger than that of PMAHT (3.7 +/- 0.4 nS, n = 76). Acute application of forskolin (10(-5) mol/l, FOR) plus 0.5 mmol/l 8-(4-chlorophenylthio)-cAMP (cAMP) depolarized V by 12 (FORHT) and 8 (PMAHT) mV, respectively. The acute increase of G by FOR plus cAMP was by 7.6 +/- 1.9 nS for FORHT (n = 22) and only 2.2 +/- 1 nS for PMAHT (n = 13). ATP (10(-4) mol/l) depolarized V in both types of cells. It enhanced G by 16.7 +/- 4.1 nS in FORHT (n = 14) and significantly less (by 5.5 +/- 1.2 nS, n = 14) in PMAHT. Also the G increase lasted longer in FORHT. Neurotensin (NT, 10(-8) mol/l) also had a stronger and longer lasting effect in FORHT.(ABSTRACT TRUNCATED AT 250 WORDS)

Synergistic activation of non-rectifying small-conductance chloride channels by forskolin and phorbol esters in cell-attached patches of the human colon carcinoma cell line HT-29cl.19A

Cell-attached patch-clamp studies with the human colon carcinoma HT-29cl.19A cells revealed a small chloride channel with a unitary conductance of 6.5 pS at 70 mV and 4.6 pS at -70 mV clamp potential after cAMP was increased by activation of adenylyl cyclase by forskolin. Usually channels inactivated upon patch excision, but in a few excised patches the channels stayed active and displayed a linear I/V relation in symmetrical (150 mmol/l) chloride solutions with a conductance of 7.5 pS. A 16-fold increase in channel incidence was observed when forskolin and phorbol 12,13-dibutyrate (PDB) were present together. The open probability was voltage-independent and was not different in the presence of forskolin plus PDB or with forskolin alone. The conductance sequence of the channel as deduced from outward currents carried by five different anions including chloride was: Cl- > BR- > NO3- > gluconate > I-. The permeability sequence deduced from the reversal potentials was NO3- > or = Br- > Cl- > I- > gluconate. With iodide in the pipette the conductance decreased strongly. Moreover, the inward current was reduced by 61%, indicating a strong inhibition of the chloride efflux by iodide. Similarly, the forskolin-induced increase of the short-circuit current (Isc) in confluent filter-grown monolayers was strongly reduced by iodide in the apical perfusate. Iodide also increased the fractional resistance of the apical membrane and repolarized the membrane potential, indicating an inhibitory action on the forskolin-induced increase of the apical chloride conductance.(ABSTRACT TRUNCATED AT 250 WORDS)

Phosphorylation-regulated low-conductance Cl- channels in a human pancreatic duct cell line

A low-conductance Cl- channel has been identified in the apical membrane of the human pancreatic duct cell Capan-1 using patch-clamp techniques. Cell-attached channels were activated by the vasoactive intestinal polypeptide (VIP, 0.1 mumol/l), dibutyryl-adenosine 3',5'-cyclic monophosphate (db-cAMP, 1 mmol/l), 8-bromo adenosine 3',5'-cyclic monophosphate (8-Br-cAMP, 1 mmol/l), 3-isobutyl-1-methyl-xanthine (IBMX, 100 mumol/l) and forskolin (10 mumol/l). No channel activity was observed in non-stimulated control cells. In both cell-attached and excised inside-out patches, the channel had a linear current/voltage relationship and a unitary conductance of 9 pS at 23 degrees C and 12 pS at 37 degrees C. Its opening probability was not voltage dependent although pronounced flickering was induced at negative potentials. Anionic substitution led to the selectivity sequence Cl- > I- >> > HCO3- > gluconate. In inside-out excised patches, the channel activity declined spontaneously within a few minutes. Reactivation of silent excised channels was achieved by adding protein kinase A (PKA, in the presence of ATP, cAMP and Mg2+). Conversely, active channels were silenced in the presence of alkaline phosphatase. The PKA-activated Cl- channel was 4,4'-diisothiocyanatostilbene-2,2'-disulphonic acid (DIDS, 100 mumol/l) and 4-acetamido-4'-isothiocyanatostilbene-2,2'-disulphonic acid (SITS, 100 mumol/l) insensitive, but was blocked by diphenylamine-2-carboxylic acid (DPC, 100 mumol/l). These results demonstrate that the apical low-conductance Cl- channel in Capan-1 is regulated on-cell by VIP receptors via cAMP and off-cell by PKA and phosphatases.(ABSTRACT TRUNCATED AT 250 WORDS)

Evidence for agonist-induced export of intracellular Ca2+ in epithelial cells

There is increasing evidence that some agonists not only induce intracellular Ca2+ increases, due to store release and transmembranous influx, but also that they stimulate Ca2+ efflux. We have investigated the agonist-stimulated response on the intracellular Ca2+ activity ([Ca2+]i) in the presence of thapsigargin (10(-8) mol/l, TG) in HT29 and CFPAC-1 cells. For CFPAC-1 the agonists ATP (10(-7)-10(-3) mol/l, n = 9), carbachol (10(-6)-10(-3) mol/l, n = 5) and neurotensin (10(-10)-10(-7) mol/l, n = 6) all induced a concentration-dependent decrease in [Ca2+]i in the presence of TG. Similar results were obtained with HT29 cells. This decrease of [Ca2+]i could be caused by a reduced Ca2+ influx, either due to a reduced driving force for Ca2+ in the presence of depolarizing agonists or due to agonist-regulated decrease in Ca2+ permeability. Using the fura-2 Mn2+ quenching technique we demonstrated that ATP did not slow the TG-induced Mn2+ quench. This indicates that the agonist-induced [Ca2+]i decrease in the presence of TG was not due to a reduced influx of Ca2+ into the cell, but rather due to stimulation of Ca2+ export. We used the cell attached nystatin patch clamp technique in CFPAC-1 cells to examine whether, in the presence of TG, the above agonists still led to the previously described electrical changes. The cells had a mean membrane voltage of -49 +/- 3.6 mV (n = 9). Within the first 3 min ATP was still able to induce a depolarization which could be attributed to an increase in Cl- conductance.(ABSTRACT TRUNCATED AT 250 WORDS)

Small and intermediate conductance chloride channels in HT29 cells

Recently, it has been shown that intermediate conductance outwardly rectifying chloride channels (ICOR) are blocked by cytosolic inhibitor (C. I.) found in the cytosol of human placenta and epithelial cells. C. I. also reduced the baseline current in excised membrane patches of HT29 cells. In the present study, this effect of C. I. was characterized further. Heat treated human placental cytosol was extracted in organic solvents and dissolved in different electrolyte solutions. It is shown that the reduction of baseline conductance (g(o)) is caused by inhibition of small non-resolvable channels, which are impermeable to Na+ and SO4(2-), but permeable to Cl-. The regulation of these small Cl(-)-conducting channels (g(o)) and of ICOR was examined further. First, no activating effects of protein kinase A (PKA) on the open probability (Po) of the ICOR or on the g(o)) were observed. The Po of the ICOR was reduced by 22% in a Ca(2+)-free solution. g(o) was insensitive to changes in the Ca2+ activity. The effects of C. I. from a cystic fibrosis (CF) placenta and the CF pancreatic duct cell line CFPAC-1 were compared with the effects of corresponding control cytosols, and no significant differences between CF and control cytosols were found. We conclude that the excised patches of HT29 cells contain ICOR and small non-resolvable Cl(-)-conducting channels which are similarly inhibited by C. I.(ABSTRACT TRUNCATED AT 250 WORDS)

Increase in cytosolic Ca2+ regulates exocytosis and Cl- conductance in HT29 cells

Increases of cytosolic Ca2+, as occur with agonists such as ATP, neurotensin (NT), hypotonic cell swelling and ionomycin, enhance the membrane conductance (GM) and hence the input conductance (GI) of HT29 cells. In the present study we have examined whether these increases in GM are paralleled by exocytosis. To this end the membrane capacitance (CM) of HT29 cells was measured by patch clamp techniques. Two methods to monitor CM were used: a direct method (DM) and a phase tracking method (PTM). With the DM the following results were obtained. NT (10(-8) mol/l, n = 9) increased GM and CM significantly from 2.4 +/- 0.3 nS and 23.5 +/- 3 pF to 32 +/- 8 nS and 27.3 +/- 3.1 pF respectively. ATP (10(-4) mol/l, n = 29) had a very similar effect. GM and CM were increased from 5.7 +/- 1 nS and 36 +/- 4.4 pF to 111 +/- 21 nS and 44 +/- 5.4 pF respectively. Hypotonic cell swelling (160 mosmol/l, n = 18) had a comparable effect: GM and CM were increased from 4.9 +/- 1 nS and 30 +/- 4.1 pF to 46 +/- 10 nS and 37 +/- 4.9 pF respectively. Ionomycin (10(-7) mol/l, n = 4) gave similar results. With the PTM it was possible to monitor the rapid changes in GM and CM, as they were induced by ATP (n = 42) and NT (n = 29), with high time resolution.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of forskolin on crypt cells of rat distal colon. Activation of nonselective cation channels in the crypt base and of a chloride conductance pathway in other parts of the crypt

We recently showed that prostaglandin E2 (PGE2) causes depolarization in cells at the base of isolated crypts from rat distal colon by activating nonselective cation channels. In order to investigate whether PGE2 acts via intracellular cyclic adenosine monophosphate (cAMP), the effect of forskolin on cell potential and on whole-cell current was investigated using the slow whole-cell patch-clamp method with nystatin. In addition, effects of forskolin in cells at other sites along the crypt were investigated. At the crypt base, the unstimulated cells had a resting potential of -70.6 +/- 1.3 mV (n = 25). When forskolin was added to the bath, the cells depolarized to -21.1 +/- 1.5 mV (n = 25). This depolarization was inhibited by substitution of all Na+ in the bath solution by N-methyl-D-glucamine (NMDG+) or by addition of flufenamic acid (50 mumol/l), a blocker of nonselective cation channels, to the bath. In contrast, the Cl- channel blocker 5-nitro-2-(3-phenylpropylamino)benzoic acid (NPPB, 50 mumol/l) did not affect the depolarization. Moving along the crypt, the resting potential was -66.8 +/- 1.8 mV (n = 11) in the mid-crypt and -48.1 +/- 2.9 mV (n = 9) in cells of the upper part of the crypt. Forskolin caused a strong depolarization to about -20 mV in all parts of the crypt. In contrast to cells at the base, this depolarization was only partly diminished by substitution of Na+ by NMDG+, whereas substitution of bath Cl- by gluconate caused an initial further depolarization, followed by a repolarization to the cell's resting potential.(ABSTRACT TRUNCATED AT 250 WORDS)

The outwardly rectifying Cl- channel is not involved in cAMP-mediated Cl- secretion in HT-29 cells: evidence for a very-low-conductance Cl- channel

The patch-clamp technique and transepithelial current measurements in conjunction with analysis of transepithelial current noise were employed in order to clarify the role of the outwardly rectifying, depolarization-induced Cl- channel (ORDIC) during cAMP-mediated Cl- secretion in HT-29/B6 cells. Confluent monolayers growing on permeable supports were used in order to ensure the apical location of measured Cl- channels. The ORDIC needed to be activated by excision and/or depolarization, and was found in both cAMP-stimulated and non-stimulated cells. Both 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB) and 4,4'-dinitro-2,2'-stilbenedisulphonate (DNDS) induced fast flickery-type blocks of the ORDIC at low, micromolar blocker concentrations and were used as a probe for ODIC. However, these substances were ineffective in blocking transepithelial forskolin-induced Cl- secretion of monolayers in Ussing chambers. No inhibitory effect at all was detected for DNDS up to 1 mmol/l. NPPB blocked the ORDIC at low concentrations (IC50 = 0.5 +/- 0.3 mumol/l) by reducing its open probability, but NPPB did not block forskolin-induced Cl- secretion unless high concentrations were used (IC50 = 240 +/- 10 mumol/l). In order to exclude effects of NPPB other than on the apical Cl- channel, transepithelial measurements were performed in basolaterally amphotericin-permeabilized, forskolin-stimulated preparations, and a serosal-to-mucosal Cl- gradient was applied as a driving force. Under these conditions, NPPB's inhibitory effects were also very small. Noise analysis of this gradient-driven Cl- current showed a very-low-frequency Lorentzian noise component (fc = 1.4 +/- 0.2 Hz), which was not compatible with Lorentzians predicted from single-channel gating of ORDIC.(ABSTRACT TRUNCATED AT 250 WORDS)

Small-conductance chloride channels induced by cAMP, Ca2+, and hypotonicity in HT29 cells: ion selectivity, additivity and stilbene sensitivity

Previous studies in HT29 cells utilizing the cell-attached nystatin (CAN) method [Greger R, Kunzelmann K (1991) Pflügers Arch 419:209-211] have revealed that the Cl- channels induced by cAMP or by increasing cytosolic Ca2+, e.g. by addition of ATP, and by hypotonic cell swelling share in common their conductance, which was so small in our studies [Kunzelmann et al. (1992) Pflügers Arch (in press)] that we could not resolve it at the single-channel level. This prompted the question whether these Cl- conductances can be distinguished in terms of their ion selectivity and sensitivity towards inhibitors. Whether these pathways are additive or not was also examined. The present study utilized the whole-cell patch-clamp and the CAN methods. A total of 160 patches were studied. In whole-cell patches 8-(4-chlorophenylthio)-cAMP (cAMP, 0.1 +/- 1 mmol/l) induced a significant depolarization by 5 mV and a twofold increase in conductance (G) from 6.2 +/- 1.5 nS to 11.7 +/- 3.2 nS (n = 15). Total replacement of Cl- by Br- and I- in cAMP-treated cells hyperpolarized the membrane voltage (V) significantly from -35 +/- 2.8 to -39 +/- 3.4 and -45 +/- 3.3 mV respectively, but had no detectable effect on G, which was 11.9 +/- 3.3 nS in the case of Br- and 11.8 +/- 3.3 nS in the case of I-. Hence, the permselectivity of the cAMP pathway was I- > Br- > Cl-, but the conductances for these anions were all indistinguishable.(ABSTRACT TRUNCATED AT 250 WORDS)

cAMP and Ca2+ act co-operatively on the Cl- conductance of HT29 cells

Previous studies in HT29 cells have revealed that the Cl- channels induced by cAMP or by increasing cytosolic Ca2+, e.g. by addition of ATP, and by hypotonic cell swelling share in common all examined properties, such as ion selectivity and blocker sensitivity. In addition, it was shown that conductances induced by either pathway were not additive. Therefore all three pathways apparently act on the same type of small conductance Cl- channel. In CFPAC-1 cells the general properties of the Cl- conductance were identical. However, the cAMP response was absent. In both cell types the Ca(2+)-mediated conductance response was transient. Here we examine the kinetics of the conductance increases induced by neurotensin (NT, 10(-8) mol/l) or ATP (10(-5) mol/l) in HT29 and CFPAC-1 cells using the slow (nystatin) or fast whole cell patch clamp technique, and we ask whether cAMP influences these kinetics. In the continuous presence of NT the conductance response in both cell types was very transient. It collapsed with a time constant (tau) of 39 (30-56 s) in HT29 and of 33 (27-41 s) in CFPAC-1 cells. The ATP response was also transient with a tau of 49 (42-57 s) in HT29 cells and 102 (77-152 s) in CFPAC-1 cells. Pre-treatment by membrane permeable cAMP (10(-3) mol/l) enhanced the baseline conductance in HT29 but not in CFPAC-1 cells. Furthermore, the ATP- and NT-induced conductance increases became significantly less transient in HT29 but not in CFPAC-1 cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Transmitter-induced changes of the membrane voltage of HT29 cells

The colonic carcinoma cell line HT29 was used to examine the influence of agonists increasing cytosolic cAMP and Ca2+ activity on the conductances and the cell membrane voltage (Vm). HT29 cells were grown on glass cover-slips. Cells were impaled by microelectrodes 4-10 days after seeding, when they had formed large plaques. In 181 impalements Vm was -51 +/- 1 mV. An increase in bath K+ concentration from 3.6 mmol/l to 18.6 mmol/l or 0.5 mmol/l Ba2+ depolarized the cells by 10 +/- 1 mV (n = 49) or by 9 +/- 2 mV (n = 3), respectively. A decrease of bath Cl- concentration from 145 to 30 mmol/l depolarized the cells by 11 +/- 1 mV (n = 24). Agents increasing intracellular cAMP such as isobutylmethylxanthine (0.1 mmol/l), forskolin (10 mumol/l) or isoprenaline (10 mumol/l) depolarized the cells by 6 +/- 1 (n = 13), 15 +/- 3 (n = 5) and 6 +/- 2 (n = 3) mV, respectively. In hypoosmolar solutions (225 mosmol/l) cells depolarized by 9 +/- 1 mV (n = 6). Purine and pyrimidine nucleotides depolarized the cells dose-dependently with the following potency sequence: UTP greater than ATP greater than ITP greater than GTP greater than TTP greater than CTP = 0. The depolarization by ATP was stronger than that by ADP and adenosine. The muscarinic agonist carbachol led to a sustained depolarization by 27 +/- 6 mV (n = 5) at 0.1 mmol/l, and to a transient depolarization by 12 +/- 4 mV (n = 5) at 10 mumol/l. Neurotensin depolarized with a half-maximal effect at around 5 nmol/l.(ABSTRACT TRUNCATED AT 250 WORDS)

Small-conductance Cl- channels in HT29 cells: activation by Ca2+, hypotonic cell swelling and 8-Br-cGMP

The present study demonstrates the activation of Cl- channels in HT29 cells by agonist (ATP, neurotensin, carbachol) increasing cytosolic Ca2+, by hypotonic cell swelling and by cGMP. Cell-attached nystatin patch-clamp (CAN) as well as slow and fast whole-cell recordings were used. The cell membrane potential was depolarized in a dose-dependent manner with half-maximal effects at 0.4 mumol/l for ATP, 60 pmol/l for neurotensin and 0.8 mumol/l for carbachol. The depolarization, which was caused by Cl- conductances increases, occurred within 1 s and was accompanied by a simultaneous and reversible increase of the input conductance of the cell-attached membrane from 295 +/- 32 pS to 1180 +/- 271 pS (ATP; 10 mumol/l, n = 21) and 192 +/- 37 pS to 443 +/- 128 pS (neurotensin; 1 nmol/l, n = 8). The effects of the agonists could be mimicked by ionomycin (0.2 mumol/l), suggesting that an increase in intracellular Ca2+ was responsible for the activation of Cl- channels. The depolarization was followed by a secondary hyperpolarization. Hypotonic cell swelling also depolarized the cells and induced an increase in the membrane conductance. With 120 mmol/l NaCl the depolarization was 10 +/- 0.8 mV and the cell-attached conductance increased from 228 +/- 29 pS to 410 +/- 65 (n = 26) pS. NaCl at 90 mmol/l and 72.5 mmol/l had even stronger effects. Comparable conductance increases were also obtained when the different agonists or hypotonic cell swelling were examined in whole cell experiments. 5-Nitro-2-(3-phenylpropylamino)-benzoate (1 mumol/l) did not prevent the effects of Ca(2+)-increasing hormones and of hypotonic solutions.(ABSTRACT TRUNCATED AT 250 WORDS)