Protective role of epithelium in the guinea pig airway

We developed an in vitro system to assess the role of the epithelium in regulating airway tone using the intact guinea pig trachea (J. Appl. Physiol. 64: 466-471, 1988). This method allows us to study the response of the airway when its inner epithelial surface or its outer serosal surface is stimulated independently. Using this system we evaluated how the presence of intact epithelium can affect pharmacological responsiveness. We first examined responses of tracheae with intact epithelium to histamine, acetylcholine, and hypertonic KCl when stimulated from the epithelial or serosal side. We then examined the effect of epithelial denudation on the responses to these agonists. With an intact epithelium, stimulation of the inner epithelial side always caused significantly smaller changes in diameter than stimulation of the outer serosal side. After mechanical denudation of the epithelium, these differences were almost completely abolished. In the absence of intact epithelium, the trachea was 35-fold more sensitive to histamine and 115-fold more sensitive to acetylcholine when these agents were applied to the inner epithelial side. In addition, the presence of an intact epithelium almost completely inhibited any response to epithelial side challenge with hypertonic KCl. These results indicate that the airway epithelial layer has a potent protective role in airway responses to luminal side stimuli, leading us to speculate that changes in airway reactivity measured in various conditions including asthma may result in part from changes in epithelial function.     

Pharmacological differentiation of epithelium-derived relaxing factor from nitric oxide

We examined the possibility that nitric oxide is one of the epithelium-derived relaxing factors in guinea pig airways. First we studied whether nitric oxide could relax isolated tracheal strips, and then we examined the effects of known inhibitors of endothelium-dependent relaxation (EDR) in the vascular system [hemoglobin, methylene blue, and NG-monomethyl-L-arginine (L-NMMA)] on epithelium-dependent relaxation (EpDR) induced by hyperosmotic stimuli in perfused whole tracheal preparations. Mannitol (160 mM in Krebs-Henseleit solution) applied to the epithelial surface was used as an osmotic stimulus to induce EpDR after carbachol-induced contraction (2 microM, serosal side). Nitric oxide produced concentration-dependent and complete relaxation of epithelium-denuded tracheal strips. Preincubation of the whole trachea with hemoglobin significantly inhibited osmotic-induced EpDR (P less than 0.05), but preincubation with methylene blue and L-NMMA did not. Hemoglobin introduced into the epithelial side after EpDR induced by hyperosmotic stimuli reversed relaxation, but methylene blue and L-NMMA did not. These results suggest that, although EpDR and vascular EDR have some pharmacological similarities and nitric oxide can relax airway smooth muscle, nitric oxide is not responsible for osmotic-induced EpDR.     

Role of the epithelium in airway smooth muscle responses to relaxant agonists.

We studied the role of the guinea pig tracheal epithelium in modulating tracheal smooth muscle responses to the relaxant agonists albuterol, sodium nitroprusside, and theophylline. We used an in vitro preparation that allowed separation of the fluids bathing the luminal (internal) and serosal (external) surfaces of the trachea, and bronchodilators were administered to either surface of carbachol-contracted tracheae. All three drugs produced dose-dependent relaxation. However, albuterol and nitroprusside were less potent (concentration that produced half-maximal effect increased by 100- and 32-fold, respectively) when given to the epithelial side with the epithelium intact compared with the epithelium denuded or compared with serosal administration with the epithelium intact. These differences were not observed for theophylline, where smooth muscle responses were independent of either the side of stimulation or of the presence or absence of the epithelium. Direct measurements of the diffusion of theophylline across the tracheal wall in the presence or absence of epithelium showed that after 5 h of incubation with a fixed luminal concentration of theophylline, only 1.7% had diffused across the tracheal wall with the epithelium intact. This increased to only approximately 3.3% when the epithelium was denuded. These results suggest that the epithelial is a relatively weak barrier for lipophilic agents but has a major role as a diffusion barrier to hydrophilic substances.     

Quantitation of conductance pathways in antral gastric mucosa

The magnitude of cellular and shunt conductance of Necturus gastric antral mucosa was studied by (a) comparing the cellular PD response to transepithelial PD response during changes of ionic activity in the serosal bathing solution and (b) by measurement of current spread within the epithelial sheet. Using constant product KCl changes cellular resistance was 6,788 omegacm2 and shunt resistance was 1,803 omegacm2. Deletion of HCO3- from the serosal solution produced similar but quantitatively smaller changes in PD. Using HCO3- deletion cellular resistance was 7,338 omegacm2 and shunt resistance was 1,973 omegacm2. Measurement of current spead within the mucosa avoids changing ionic gradients yet gave very similar results; cellular resistance was 8,967 omegacm2 and shunt resistance was 2,947 omegacm2. The shunt contribution to transepithelial conductance ranged from 75.2 to 79.0%. Shunt selectivity was assessed using KCl dilution potentials, where mucosal dilution gave a small change in tissue PD compatible with an anion/cation selectivity ratio of 1.16 across the shunt, whereas serosal dilution effect was dominated by a PD change across the serosal membrane of the cell.     

Cold and hyperosmolar fluids in canine trachea: vascular and smooth muscle tone and albumin flux

We have studied the effects of liquids of various osmolalities and temperatures on the tracheal vasculature, smooth muscle tone, and transepithelial albumin flux. In 10 anesthetized dogs a 10- to 13-cm length of cervical trachea was cannulated to allow instillation of fluids into its lumen. The cranial tracheal arteries were perfused at constant flow, with monitoring of the perfusion pressures (Ptr) and the external tracheal diameter (Dtr). Control fluid was Krebs-Henseleit solution (KH) with NaCl added to result in a 325-mosM solution (isotonic). Hypertonic solutions were KH with NaCl (warm hypertonic) or glucose (hypertonic glucose) added to result in a 800-mosM solution. All solutions were at 38 degrees C, with isotonic and the hypertonic NaCl solutions also given at 18 degrees C (cold isotonic and cold hypertonic). Fluorescent labeled albumin was given intravenously, and the change in fluorescence in the fluid was measured during each 15-min period. Changing from warm isotonic to cold isotonic decreased Dtr and Ptr. Changing from warm isotonic to warm hypertonic or hypertonic glucose decreased Ptr with no change in Dtr. The cold hypertonic responses were not different from cold isotonic responses. Warm hypertonic solution increased albumin flux into the tracheal lumen over a 15-min period to three times that of the control period, persisting for 15 min after replacement with warm isotonic solution. Cooling induces a vasodilation and smooth muscle contraction of the trachea, whereas hypertonic solutions result in vasodilation and, if osmolality is increased with NaCl, an increase in albumin flux into the tracheal lumen.     

Evidence for A1 and A2 adenosine receptors in guinea pig trachea

The adenosine analogs [5'-N-ethylcarboxamideadenosine (NECA), 2-Chloro-adenosine (2-ClA), R-phenylisopropyladenosine (R-PIA), N6-cyclohexyl adenosine (CHA), and N6-cyclopentyladenosine (CPA)] produced both relaxation and contraction responses in isolated guinea-pig trachea. A concentration-related relaxation response was observed in trachea which were precontracted with either histamine or KC1. This response followed an order of analog potency that was indicative of the A2 receptor subtype (NECA greater than 2-ClA greater than R-PIA greater than CPA greater than CHA). Theophylline, an adenosine-receptor antagonist, blocked this relaxation response. In addition, a concentration-related contractile response was produced with adenosine analogs in those trachea that were not previously contracted. In contrast, the contractile response followed an analog potency indicative of the A1 receptor subtype (R-PIA greater than 2-ClA = CPA = CHA). This contractile response was not mediated by cholinergic, adrenergic or histaminergic receptors. 2-ClA induced a biphasic response, while NECA only relaxed these tissue under basal tone. Unlike the relaxation response, these contractile responses were not attenuated by theophylline, but were blocked by 1,3 dipropyl-8-(2 amino-4-chlorophenyl)xanthine (PACPX). These findings confirm the existence of two subpopulations of adenosine receptors in guinea pig trachealis muscle.     

Stimulation frequency-dependent nonadrenergic noncholinergic airway responses of the guinea pig

We examined the inhibitory and excitatory components of the nonadrenergic noncholinergic (NANC) innervation of the guinea pig airways by in vivo and in vitro methods. Electrical stimulation of the vagus in chloralose-urethan-anesthetized guinea pigs after cholinergic and adrenergic blockade produced peripheral airway constriction (insufflation pressure) and tracheal relaxation (pouch pressure). Vagal stimulation was applied for 90 s at 5-V pulses of 2-ms duration at frequencies of 5, 15, 25, and 35 Hz in each group (n = 6). The pouch relaxation peaked at 15 Hz. The insufflation pressure was highest at 5 Hz. Field stimulations of the same frequencies were applied on tracheal spirals and lung parenchymal strips. The maximal relaxation of the trachea occurred at 15-35 Hz. The lung parenchymal strip tensions increased almost linearly as the frequency increased from 5 to 35 Hz. The results of the study indicated a frequency-dependent response for both excitatory and inhibitory components of the NANC, which operate at different frequencies for optimal responses.     

Selective stimulation of jugular ganglion afferent neurons in guinea pig airways by hypertonic saline

Pedersen, Karen E., Sonya N. Meeker, Margerita M. Riccio,and Bradley J. Undem. Selective stimulation ofjugular ganglion afferent neurons in guinea pig airways by hypertonicsaline. J. Appl. Physiol. 84(2):499-506, 1998.We evaluated the ability of hyperosmolar stimulito activate afferent nerves in the guinea pig trachea and main bronchiand investigated the neural pathways involved. By usingelectrophysiological techniques, studies in vitro examined the effectof hyperosmolar solutions of sodium chloride (hypertonic saline) onguinea pig airway afferent nerve endings arising from either vagalnodose or jugular ganglia. The data reveal a differential sensitivityof airway afferent neurons to activation with hypertonic saline.Afferent fibers (both A and C fibers) with cell bodies located injugular ganglia were much more sensitive to stimulation with hypertonicsaline, compared with afferent neurons with cell bodies located innodose ganglia. Additional studies in vivo demonstrated that inhalationof aerosols of hypertonic saline induced plasma extravasation in guineapig trachea that was mediated via tachykininNK₁ receptors. Identification of adifferential sensitivity of guinea pig airway afferent nerves tohypertonic saline leads to the speculation that airway responses tohyperosmolar stimuli may result from activation of afferent neuronsoriginating predominantly from the jugular ganglion.

     

Action of Antidiuretic Hormone on the Equivalent Pore Radius at both Surfaces of the Epithelium of the Isolated Toad Skin

A previously described method (1) allows the observation of swelling and shrinking of the epithelial cells of the isolated toad skin, when the solution bathing either the outer or inner side of the skin is modified. Thus, the concentration of probing molecules of graded size, isotonic to the epithelial cells, across each face of the isolated toad skin can be determined. These concentrations have been used for the estimation of the equivalent pore radius at the outer and inner face of the skin epithelium, following the approach of Goldstein and Solomon for red cells (3). An equivalent pore radius of 4.5 A for the outer surface, and one of 7 A for the inner surface have been obtained. Antidiuretic hormone had an effect only when added to the inner side. This effect was only at the outer surface and is interpreted as widening of the 4.5 A pores to about 6.5 A. A model membrane, formed by narrow and wide pores in series, may explain some of the apparent inconsistencies previously observed.     

Relaxant activity of atriopeptins in isolated guinea pig airway and vascular smooth muscle

Atriopeptins are circulating peptide hormones which are secreted by atrial tissue and act at the kidney. Because the atriopeptins survive passage through the pulmonary circulation, they also may be involved in the modulation of airway or pulmonary vascular smooth muscle tone. Using in vitro organ bath techniques, atriopeptins were found to induce potent concentration-dependent relaxation of isolated guinea pig trachea, and pulmonary artery with a rank order of potency: atriopeptin III greater than atriopeptin II greater than atriopeptin I. Atriopeptin-induced smooth muscle relaxation was observed to be a direct response since it was not mediated by activation of relaxant VIP receptors, beta-adrenergic receptors, or H2 receptors nor affected by cyclooxygenase inhibition or denuding of the vasculature or trachea of endothelial and epithelial cells. The time course of atriopeptin II-induced relaxation of the pulmonary artery was transient in contrast to the prolonged relaxations on the trachea. The transient relaxant responses of atriopeptin II on pulmonary artery were not due to metabolism of atriopeptin II to atriopeptin I by angiotensin-converting enzyme since pretreatment with captopril did not augment the response. These results seem to indicate that distinct atriopeptin receptors may exist in airway and pulmonary arterial smooth muscle and that activation of these relaxant receptors may play an important role in the regulation of pulmonary vascular and bronchomotor tone.     

Transport pathways in canine lingual epithelium involved in sweet taste

The responses of canine lingual epithelium to D-glucose weremeasured in an Ussing chamber to determine the possible contributionof the osmotic changes of taste cells to the response of saccharides.With the mucosal solution containing 50 mM NaCl, 2 mM HEPES,pH 7.4 (solution A) and the serosal solution containing Krebs—Henseleit(KH) buffer the addition of up to 0.5 M D-glucose in the mucosalsolution increased the short circuit current (I_sc) in a sigmoidalmanner. The D-glucose-stimulated I_sc was inhibited by 0.1 mMamiloride or 1 mM ouabain added to either the mucosal or theserosal solution, and partially inhibited by 5 mM BaCl₂ addedto the serosal solution. The inhibition by these three compoundswas also observed in the presence of 0.5 M NaCl. Ouabain alsoinhibited transport when added to solution A. These experimentssuggest that in canine lingual epithelium the paracellular pathwaypermits molecules as large as ouabain (mol. wt 586) to diffusefrom the mucosal to the serosal solution and vice versa underall osmotic conditions. These results may explain the phenomenonof intravascular taste. Such is not the case in rat tongue whereouabain only inhibited transport when added to the serosal solution.Increasing the osmolality of the serosal KH buffer by additionof relatively membrane-impermeable saccharides such as sucroseor L-glucose did not significantly alter the I_sc, whereas makingthe serosal KH solution hypo-osmotic resulted in a transientdecrease in I_sc. These data suggest that the increase in I_scinduced by saccharides, such as D-glucose, is not simply anosmotic response of the epithelium but more likely the consequenceof saccharides binding weakly to receptors. That the responseto both salts by themselves and in the presence of saccharidesexhibits the same cation selectivity, and that both are inhibitedby amiloride, ouabain, BaCl₂ and LaCl₃ suggest that in caninelingual epithelia, in contrast to rat epithelium, the responsesto hyperosmotic concentrations of salts and saccharides mightoccur via the same transcellular pathways.     

Effects of Pb<Superscript>2+</Superscript> ions on Na<Superscript>+</Superscript> transport in the isolated skin of the toad <Emphasis Type="Italic">Pleurodema thaul</Emphasis>

The effects induced by lead ions on the short-circuit current (SCC) and on the potential difference (V) of the toad Pleurodema thaul skin were investigated. Pb²⁺ applied to the outer (mucosal) surface increased SCC and V and when applied to the inner (serosal) surface decreased both parameters. The stimulatory effect, but not the inhibitory action, was reversible after washout of the metal ion. The amiloride test showed that the increase was due principally to stimulation of the driving potential for Na⁺ (V-E_Na+) and that inhibition was accompanied by a reduction in the V-E_Na+ and also by a significant decrease in skin resistance indicating possible disruption of membrane and/or cell integrity. The effect of noradrenaline was increased by outer and decreased by inner administration of Pb²⁺. The results suggest that mucosal Pb²⁺ activates toad skin ion transport by stimulating the V-E_Na+ and that serosal Pb²⁺, with easier access to membrane and cellular constituents, inactivates this mechanism, revealing greater toxicity when applied to the inner surface of the skin. Abbreviations: SCC – short-circuit current; V – potential difference; V-E_Na+– driving potential for Na⁺; ENaC – epithelial sodium channel; R_Na+– active sodium resistance; RS – passive or shunt resistance; G_Na– active sodium conductance; GS – passive or shunt conductance; Gmax – total conductance; EC₅₀– half-maximal excitatory concentration; IC₅₀– half maximal inhibitory concentration; NA – noradrenaline.     

A novel concentric double-barrelled calcium-selective microelectrode for small cells

A novel concentric design of double-barrelled Ca2+-selective microelectrode, with an inner pipette tip that protrudes beyond an outer one, has recently been developed and is described. This configuration of pipettes was produced from concentric capillaries in one step using a horizontal pipette puller. For the tip of the inner barrel to protrude, Corning 1724 aluminosilicate glass was selected, as it has a higher melting point than the 1723 glass which is used for the outer barrel. To reduce electrode resistance the inner capillary was best made with a triangular shape. It was preferentially silanized in a dry box by injection of methyltrichlorosilane into only the inner barrel. The Ca2+ neutral carrier-based liquid membrane (ETH 1001) was back-filled from the tip to the shank of the inner pipette and above this CaCl2 solution was added. KCl, which contained EGTA and was buffered to pCa 7, was used to fill the reference barrel. These Ca2+ electrodes showed linear response with slope approximately equal to 30 mV for changes in Ca2+ concentration between 10(-3) and 10(-7) M in the presence of constant [K+]. They offer a number of advantages including a low noise level achieved by the presence of the external concentric KCl electrode, and a simple mechanical structure that allows applications to a variety of small cells.     

Mechanisms of spontaneous relaxation of smooth muscle (guinea pig taenia coli) depolarized in a hyperkalemic medium]

Experiments were performed on isolated strips of guinea pig taenia coli by the double sucrose-gap method. The artificial node was depolarized with potassium solution (from 120 to 167.7 mM KCl). When the bathing solution contained 0.4 mM Ca and the temperature was equal to 25 degrees C then potassium contracture was followed by fast relaxation. The muscular tone changed slightly during rectangular pulse of hyperpolarizing current, after switching off the current muscle generated a transient contractile response. The amplitude of such off-responses increased in some range with increasing in strength and duration of conditioning current. Treatment of muscle with compound D-600 resulted in a reduction of muscular tone and elimination of off-responses. The addition of Na ions to potassium solution (substitution of 47.7 mM KCl with the same quantity of NaCl) reduced muscular tone and enhanced the relaxation after off-responses. In sodium-free potassium solution each off-response was followed by increasing muscular tone but when the bathing solution contained Na ions this increase of the tone was not observed. The data obtained strongly suggest that the spontaneous relaxation of smooth muscle which was contracted in K-solution resulted from: 1) inactivation of calcium channels of surface membrane, 2) sequastration of Ca ions by intracellular storange sites, 3) extrusion of Ca in extracellular space (in part by means of Na-Ca exchange diffusion).     

Transport numbers in biomembranes from emf measurements.

Initial membrane potentials of biological periderm and cuticular membranes have been measured with KCl solution using Ag/AgCI electrodes. For the emf measurements, the concentration in both compartment were first brought to equilibrium with 0.01 M KCI solution. After equilibration the concentration in one compartment (inner surface of the membrane) was kept constant as 0.01 M and the concentration in the other compartment was varied between 10(-4) M and 1 M. The procedure was reversed as the concentration in the outer surface of the membrane was fixed and the other side was varied. Transport number of K(+) ion corresponding to the each of the two surface layers (homogeneous and heterogeneous) of the membrane was evaluated from the initial time emf measurements. The results obtained were compared with charged polysulfone with polyester supported membrane.     

PROTOPLASMIC ASYMMETRY IN NITELLA AS SHOWN BY BIOELECTRIC MEASUREMENTS

Using multinucleate cells of Nitella 2 or 3 inches in length it is possible to kill one end with chloroform without producing at the other any immediate alteration which can be detected by our present methods. When a spot in external contact with sap is killed its potential difference falls approximately to zero and it is therefore possible to measure the potential difference across the protoplasm at any desired point merely by leading off from that point to the one where the protoplasm has been killed. The results indicate that the inner and outer protoplasmic surfaces differ, for when both surfaces are in contact with the same solution (cell sap) there is an electromotive force of about 15.9 millivolts, the inner surface being positive to the outer (i.e. the positive current tends to flow from the inner surface through the electrometer to the outer surface). The situation resembles that in Valonia where the corresponding value (with Valonia sap applied to the outside) has been reported as about 14.5 millivolt (the inner surface being positive to the outer). It would seem appropriate to designate this as radial polarity.     

DIFFERING RATES OF DEATH AT INNER AND OUTER SURFACES OF THE PROTOPLASM : I. EFFECTS OF FORMALDEHYDE ON NITELLA

When protoplasm dies it becomes completely and irreversibly permeable and this may be used as a criterion of death. On this basis we may say that when 0.2 M formaldehyde plus 0.001 M NaCl is applied to Nitella death arrives sooner at the inner protoplasmic surface than at the outer. If, however, we apply 0.17 M formaldehyde plus 0.01 M KCl death arrives sooner at the outer protoplasmic surface. The difference appears to be due largely to the conditions at the two surfaces. With 0.2 M formaldehyde plus 0.001 M NaCl the inner surface is subject to a greater electrical pressure than the outer and is in contact with a higher concentration of KCl. In the other case these conditions are more nearly equal so that the layer first reached by the reagent is the first to become permeable. The outer protoplasmic surface has the ability to distinguish electrically between K⁺ and Na⁺ (potassium effect). Under the influence of formaldehyde this ability is lost. This is chiefly due to a falling off in the partition coefficient of KCl in the outer protoplasmic surface. At about the same time the inner protoplasmic surface becomes completely permeable. But the outer protoplasmic surface retains its ability to distinguish electrically between different concentrations of the same salt, showing that it has not become completely permeable. After the potential has disappeared the turgidity (hydrostatic pressure inside the cell) persists for some time, probably because the outer protoplasmic surface has not become completely permeable.     

Effect of ouabain and furosemide on pepsinogen secretion by gastric glands in vitro

Gastric glands were isolated from rabbit stomach and pepsinogen secretion was measured after stimulation with isoproterenol, forskolin, 8-bromo cyclic adenosine monophosphate (8-bromo cAMP), cholecystokinin octapeptide (CCK-OP), carbachol, and hyperosmolar medium. The responses to these stimuli in medium containing 143 mM Na+ and 5.4 mM K+ (normal medium) were compared with responses to the same stimuli in media containing either 0 Na+ and 5.4 mM K+, or 143 mM Na+ and O K+. In addition, the effects of ouabain and furosemide on secretion elicited by these stimuli were determined. Medium containing 0 Na+ inhibited all stimuli. Medium containing 0 K+ inhibited the action of 8-bromo cAMP and stimuli postulated to be mediated by cAMP. Ouabain inhibited the same stimuli as O K+ medium, and, in addition, inhibited the response to hyperosmolar medium. However, ouabain enhanced the response to CCK-OP. Furosemide inhibited the response to hyperosmolar medium but had no effect on the action of any secretagogue employed. Intraglandular [Na+] increased and [K+] decreased after exposure to K+-free medium or ouabain. cAMP content of the glands was assayed after stimulation with both isoproterenol and hyperosmolar medium. Isoproterenol and hyperosmolar medium significantly increased cAMP levels. The results are discussed in relation to possible involvement of ion transport or intracellular ion concentration in the secretory process.     

Effects of cadmium on Na+ transport in the isolated skin of the toad Pleurodema thaul

Cadmium ions applied to either (outer or inner) surface of the isolated toad skin dose-dependently increased the short-circuit current (SCC), the potential difference (V) and the active sodium conductance (G(Na)) in the concentration range 0.07-0.50mM. Maximal stimulatory effect was over 30% with an EC(50) of about 0.1mM. The effect of the highest concentration used (0.75mM) decreased considerably, and when it was applied to the inner surface (10 experiments), induced between 30% and 40% inhibition of the electric parameters in four experiments. Pretreatment with amiloride inverted the stimulatory effect of externally applied Cd(2+), suggesting competitive action on the apical Na(+) channel. The effect of noradrenaline (NA) was increased after outer application of Cd(2+) and decreased after inner application of the metal: the latter effect might be due to cadmium inhibition of the activity of Na(+),K(+)-ATPase. On the other hand, pretreatment with amiloride was followed by partial although transient reversal of its effects by serosal Cd(2+), which might be explained by action of cadmium on cytoplasmic lysine residues concerned with Na(+) channel gating. The amiloride test showed that the increment of the electric parameters was due principally to stimulation of the driving potential for Na(+) (V-E(Na(+))) and that inhibition was accompanied by a reduction in the V-E(Na(+)) and by a significant decrease in skin resistance indicating possible disruption of membrane or cell integrity. These data are in favor of the possibility that externally applied Cd(2+) activates toad skin ion transport, partly by increasing apical sodium conductance and also by stimulating the V-E(Na(+)), and that internally applied Cd(2+), with easier access to membrane and cellular constituents, may inhibit the sodium pump.     

On the role of calcium in the mechanism of ADH action

With an increased influx of Ca2+ in the cytoplasm, the response of cells to ADH in the urinary bladder of the frog was lowered by addition of ionophore A23187 from the side of the basolateral cell membrane, but inhibited when it was added from the apical cell membrane. The removal of calcium by EGTA from the serosal surface was accompanied by a sharp increase of osmotic permeability not only to water, but also to inulin; while when calcium was removed from the mucosal surface of the urinary bladder, osmotic permeability was not changed. After being added to the Ringer solution from the outer surface of the apical cell membrane, the inhibitors of Ca2+ channels (verapamil, Ni2+, Mn2+, Co2+) decreased the effect of ADH. These data indicate that Ca2+ applied onto the outer surface of apical plasma membrane plays an important role in the action of ADH.