Escherichia coli heat stable enterotoxin (ST.) caused Cl-secretion across T84 cell monolayers in a dose-dependent manner only when applied to the apical membrane surface and not when applied to the basolateral surface. Measurement of cAMP, cGMP, and free cytosolic Ca2l in response to STa suggested that cGMP alone mediated the Cl-secretory response. Studies utilizing blockers of the Na+,K+-ATPase pump, a Na+,K+,ClI cotransport system, a K+ channel, and a Cl-channel suggest that all of them participate in the Clsecretory process induced by ST.. The results suggest that the Cl-secretory response induced by STa is mediated by cGMP after the enterotoxin binds to its receptor on the apical membrane. The enterotoxin, by increasing cGMP, opens a K+ channel on the basolateral membrane as well as a Cl-channel on the apical membrane. The activation of these ion exit mechanisms, together with activations of the Na+,K+,Cl-cotransporter and the Na+,K+-ATPase pump drives Cl-exit through the Cl-channel on the apical membrane.

When applied to the basolateral (serosal) side of the T84 coIonic epithelial monolayer, taurodeoxycholate caused net Clsecretion in a dose-dependent manner with a threshold effect observed at 0.2 mM. In contrast, when applied to the apical (luminal) surface, concentrations of taurodeoxycholate below 1 mM had little or no effect. Only when the concentration of taurodeoxycholate present on the apical side was 2 1 mM did apical addition results in an electrolyte transport effect. This apical effect on electrolyte transport was associated with an abrupt increase in the permeability of the monolayer. Cyclic AMP and cyclic GMP in the T84 monolayers were not increased by the bile salt, but in the presence of extracellular Ca2", free cytosolic Ca2" increased with a graded dose effect and time course that corresponded approximately to the changes in short circuit current (ISC). The results suggest that luminal bile salts at a relatively high concentration (2 1 mM) increase tight junction permeability. Once tight junction permeability increases, luminal bile salts could reach the basolateral membrane of the epithelial cells where they act to increase free cytosolic Ca2" from extracellular sources. The resulting increases in free cytosolic Ca2", rather than in cyclic nucleotides, appear to be involved in transcellular Cl-secretion.

Confluent T8 monolayers grown on permeable supports and mounted in a modified Ussing chamber secrete chloride (Cl-) in response to prostaglandin El. The threshold stimulation was observed at 10-' M and a maximal effect at 10' M. Unidirectional flux studies showed an increase in both serosal to mucosal and mucosal to serosal CI-fluxes with 10' M prostaglandin El; the increase in serosal to mucosal Cl-flux exceeded the increase in mucosal to serosal flux, resulting in net Cl-secretion.Na' transport was not affected in either direction and the changes in net Cla flux correlated well with the changes in short circuit current. To identify the electrolyte transport pathways involved in the Cla secretory process, the effect of prostaglandin El on ion fluxes was tested in the presence of putative inhibitors. Bumetanide was used as an inhibitor for the basolaterally localized Na',K+,Ca-cotransport system whose existence and bumetanide sensitivity have been verified in earlier studies (Dharmsathaphorn et al. 1984. J. Clin. Invest. 75:462-471). Barium was used as an inhibitor for the K+ efflux pathway on the basolateral membrane whose existence and barium sensitivity were demonstrated in this study by preloading the monolayers with "Rb+ (as a tracer for K+) and simultaneously measuring MRb+ efflux into both serosal and mucosal reservoirs. Both bumetanide and barium inhibited the net chloride secretion induced by prostaglandin El suggesting the involvement of the Na+,K+,Ca-cotransport and a K+ efflux pathways on the basolateral membrane in the Clsecretory process. The activation of another Cl-transport pathway on the apical membrane by prostaglandin El was suggested by Cl-uptake studies. Our findings indicate that the prostaglandin El-stimulated Cl-secretion, which is associated with an increase in cyclic AMP level, intimately involves (a) a bumetanide-sensitive Na+,K+,Cl-cotransport pathway that serves as a C1-uptake step across the basolateral membrane, (b) the stimulation of a barium-sensitive K+ efflux mechanism on the basolateral membrane that most likely acts to recycle K+, and (c) the activation of a Cl-transport pathway on the apical membrane that serves as a Cl-exit pathway.

The inflammatory mediator adenosine caused sustained Cl- secretion across monolayers of T84 cells. The effect was promptly reversed by the adenosine receptor antagonist 8-phenyltheophylline and appeared to be mediated through an adenosine A2-receptor [rank order of potency: 5'-(N-ethyl)-carboxamido-adenosine (NECA) greater than adenosine greater than (-)-N6-(phenylisopropyl)adenosine (PIA) greater than or equal to (+)-PIA]. High doses of adenosine and its analogues increased cellular adenosine 3',5'-cyclic monophosphate (cAMP) but not guanosine 3',5'-cyclic monophosphate (cGMP) or free cytosolic Ca2+. However, lower concentrations of adenosine had maximal effects on Cl- secretion with little or no effect on cAMP. In other respects, Cl- secretion resembled that induced by cAMP-mediated secretagogues such as vasoactive intestinal peptide (VIP). Addition of both low and high doses of NECA activated basolateral K+ and apical Cl- channels, exhibited synergism with Ca2(+)-mediated secretagogues, did not produce additive effects with VIP or Escherichia coli heat-stable enterotoxin, and was associated with cAMP-dependent protein kinase-mediated protein phosphorylation. The results suggest that either adenosine mobilizes an intracellular pool of cAMP that is extremely efficiently coupled to the cAMP-dependent protein kinase and is thereafter rapidly destroyed or that second messenger(s) other than cAMP, cGMP, or Ca2+ are able to activate Cl- secretion in the T84 cell line. In the latter case, such messenger(s), as yet unidentified, might represent a final common pathway for cyclic nucleotide-activated Cl- secretion.

The mast cell mediator, histamine, induces a rapid and transient increase in chloride secretion across monolayers of the human colonic epithelial cell line, T84. Threshold stimulation occurred at 3 X 10(-6) M histamine and a maximal effect at 10(-4) M. The effect was reproduced by the H1 agonists 2-methylhistamine and 2-pyridylethylamine, but not by the H2 agonists 4-methylhistamine and dimaprit, suggesting the involvement of an H1 receptor. Additionally, histamine's action was inhibited by an H1 antagonist, diphenhydramine, but not by an H2 antagonist, cimetidine. Histamine treatment increased free cytosolic calcium levels, but not those of adenosine 3',5'-cyclic monophosphate (cAMP) or guanosine 3',5'-cyclic monophosphate (cGMP). The mechanism of chloride secretion induced by histamine resembled that of carbachol, in that both 1) were associated with an increase in free cytosolic calcium, 2) had a site of activation at a basolaterally localized K+ channel, and 3) were potentiated by both cAMP- and cGMP-mediated secretagogues. These results suggest that histamine may act as an intestinal secretagogue via direct interactions with epithelial cells.

Adenosine and its agonists, 5'-(N-ethylcarboxamido)adenosine and N6-(D-2-phenylisopropyl)-adenosine, induced a sustained increase in chloride secretion when added to either the apical or basolateral aspect of monolayers of the human colonic epithelial cell line T84. Secretion was induced with identical kinetics by addition to both sides, but apical addition was less potent. The rank order of potency of the agonists on either side was consistent with the presence of an adenosine A2-receptor, but the apical and basolateral receptors differed in both their ability to stimulate increases in adenosine 3',5'-cyclic monophosphate, and their susceptibility to down-modulation by chronic exposure to 5'-(N-ethylcarboxamido) adenosine in culture.

The influence of quinidine, a putative K+ channel blocker, on Cl- secretion induced by vasoactive intestinal polypeptide (VIP) was investigated. Quinidine inhibited Cl- secretion induced by VIP in T84 cell monolayers. A similar inhibitory effect of quinidine on Cl- secretion was observed in an isolated human colon. However, in the isolated human colon, which absorbs Na+ avidly, inhibition of Na+ absorption predominated. In the T84 cell, the half-maximal inhibition by quinidine occurred at 60 microM, while 300 microM almost completely inhibited the VIP-induced Cl- secretion. Mucosal addition of quinidine was at least equally effective compared with serosal addition, suggesting that quinidine acts on the apical membrane or intracellularly. Quinidine had little or no effect on VIP-stimulated Cl- efflux in the first 15 min after its addition, suggesting that blockage of the Cl- exit pathway on the apical membrane is an unlikely mechanism. Similarly, quinidine did not inhibit the K+-recycling mechanism on the basolateral membrane in the first 15 min after its addition. The initial inhibitory action of quinidine corresponded better with its ability to decrease cellular ATP levels. Our study suggests that the depletion of cellular ATP levels may explain the initial inhibitory action of quinidine on electrolyte transport in the intestine, while the late effect is multifactorial.