The effect of amiloride analogs on taste responses in gerbil

Amiloride analogs that were designed to inhibit three types of Na+ transport systems (the epithelial Na+ channel, the Na+/H+ antiporter, and the Na+/Ca++ exchanger) were applied to the tongue of the gerbil to determine their effects of electrophysiological taste responses to NaCl, CaCl2, sucrose, and glutamic acid. The pattern of responses from the chorda tympani nerve indicates that the taste of NaCl is almost totally accounted for by the epithelial Na+ channel. Phenamil, an amiloride analog which specifically blocks the epithelial Na+ channel at low concentrations, suppressed the taste responses to 0.03 M NaCl by 97%. The pattern of responses also indicates that the Na+/H+ antiporter and the Na+/Ca2+ exchanger do not mediate salt taste in the gerbil. None of the amiloride analogs blocked taste responses to CaCl2, sucrose, or glutamic acid. It is concluded that the salty taste of NaCl in the gerbil is almost totally mediated by the epithelial Na+ channel, and the kinetics of this channel are identical to amiloride-sensitive sodium channels in other systems.     

Neural representation of the taste of NaCl and KCl in gustatory neurons of the hamster solitary nucleus.

NaCl and KCl are monovalent salts that can be discriminated behaviorally by hamsters on the basis of their tastes. We examined the effects of the passive Na+ channel blocker amiloride on responses to both of these salts in 34 taste-responsive neurons of the nucleus of the solitary tract (NST) in the hamster. The effects of amiloride were assessed with two different, commonly employed stimulus protocols. Additionally, concentration-response functions for each salt were measured in 37 neurons. Cells were characterized by their best response to (in M) 0. 03 NaCl, 0.1 sucrose, 0.003 HCl, 0.001 quinine hydrochloride, and 0. 1 KCl. In neurons classified as NaCl-best, amiloride reversibly blocked responses to both NaCl and KCl. In neurons classified as HCl-best, amiloride had no effect on either stimulus. In sucrose-best neurons, amiloride blocked the response to NaCl but not KCl. These results support the hypothesis that both salts are transduced by at least two different receptor mechanisms. In the NST, information arising from these different inputs is maintained in discrete populations of neurons. In addition to differences in amiloride sensitivity, the cell types also differed in their responses to the salts across concentration. At midrange salt concentrations, NaCl-best neurons were far more responsive to NaCl than KCl, whereas HCl- and sucrose-best neurons responded equivalently to the two salts at all concentrations. Because NaCl- and HCl-best cells cannot by themselves distinguish NaCl from KCl, it is the relative activity across these cell types that comprises the code for taste discrimination.     

Functional evidence for involvement of bumetanide-sensitive Na+K+2CI- cotransport in the hepatoportal Na+ receptor of the Sprague-Dawley rat.

To investigate mechanisms involved in hepatoportal Na+ sensing, responses of hepatic afferent nerve activity (HANA) to intraportal hypertonic NaCl injection were measured before, and after, intraportal infusion of inhibitors of Na+ transport systems. HANA increased in response to the intraportal injection of 0.75 M NaCl in a dose-dependent manner. The HANA response was not affected by amiloride or 4-acetamido-4'-isothiocyanato-stilbene-2,2'-disulfonic acid (SITS), but was suppressed in a dose-dependent manner by intraportal infusion of ouabain, furosemide, or bumetanide. These results indicate that the hepatoportal Na+ receptor senses the Na+ concentration via the bumetanide-sensitive Na+K+2Cl- cotransporter.     

Sodium channel but neither Na(+)-H+ nor Na-glucose symport inhibitors slow neonatal lung water clearance.

Normal clearance of alveolar liquid following birth requires active Na transport; however, the contribution of Na channels, Na-H antiports, and Na-glucose symports is unknown. We demonstrated that intraalveolar instillation of amiloride (n = 6) or the more specific Na channel blockers benzamil (n = 13) or phenamil (n = 12) before the first breath impaired lung water clearance relative to control newborns (n = 34). Benzamil and phenamil were more potent than amiloride (P less than 0.05). Neither the Na-H antiport inhibitor dimethyl amiloride (n = 7) nor the Na-glucose symport inhibitor phloridzin (n = 7) impaired lung water clearance. Ion substitution experiments with fetal rat type II alveolar epithelia demonstrated that more than 95% of their resting or terbutaline-stimulated short circuit current (Isc) depended upon Na bathing their apical membrane. Isc was decreased by amiloride (IC50 of amiloride-sensitive Isc = 0.3 x 10(-6) M) and benzamil (IC50 of benzamil-sensitive Isc = 0.3 x 10(-7) M) but was unaffected by dimethyl amiloride (10(-4) M). We conclude that in vivo postnatal clearance of fetal lung liquid can be impaired by Na channel blockers and is unaffected by blockers of Na-H antiports and Na-glucose symports. Na transport in fetal type II cells has high affinity for amiloride, and these cells likely contribute to normal neonatal lung liquid clearance.     

Neural representation of salts in the rat solitary nucleus: brain stem correlates of taste discrimination

One mechanism of salt taste transduction by gustatory receptor cells involves the influx of cations through epithelial sodium channels that can be blocked by oral application of amiloride. A second mechanism is less clearly defined but seems to depend on electroneutral diffusion of the salt through the tight junctions between receptor cells; this paracellular pathway is insensitive to amiloride. Because the first mechanism is more sensitive to sodium salts and the second to nonsodium salts, these peripheral events could underlie the ability of rats to discriminate sodium from nonsodium salts on the basis of taste. Behavioral experiments indicate that amiloride, at concentrations that are tasteless to rats, impairs a rat's ability to discriminate NaCl from KCl and may do so by making both salts taste like KCl. In the present study, we examined the neural representation of NaCl and KCl (0.05-0.2 M), and mixtures of these salts with amiloride (0, 3, and 30 microM), to explore the neural correlates of this behavioral result. NaCl and KCl were represented by distinct patterns of activity in the nucleus of the solitary tract. Amiloride, in a concentration-dependent manner, changed the pattern for NaCl to one more characteristic of KCl, primarily by reducing activity in neurons responding best to NaCl and sucrose. The effect of amiloride concentration on the response to 0.1 M NaCl in NaCl-best neurons was virtually identical to its effect on behavioral discrimination performance. Modeling the effects of blocking the amiloride-insensitive pathway also resulted in highly similar patterns of activity for NaCl and KCl. These results suggest that activity in both the amiloride-sensitive and -insensitive pathways is required for the behavioral discrimination between NaCl and KCl. In the context of published behavioral data, the present results suggest that amiloride-sensitive activity alone is not sufficient to impart a unique signal for the taste of sodium salts.     

Amiloride increases sodium chloride taste detection threshold in rats

The epithelial sodium-channel blocker amiloride has been shown to inhibit sodium responses in the 7th cranial nerve of the rat. In the signal detection task used in this study, amiloride (100 microM) treatment raised the NaCl threshold by approximately 1 log10 unit. The inhibition constant for amiloride was 1 microM at 0.013 M NaCl. Because the NaCl intake of adult rats has been shown to be related to the level of dietary NaCl exposure early in development, rats were exposed by way of maternal diet to 1 of 3 diets (0.1% NaCl, n = 8; 1.0% NaCl, n = 8; 3.0% NaCl, n = 9) from conception through weaning, to determine whether this treatment affects taste sensitivity. At Postnatal Day 30, rats were placed on 1.0% NaCl chow. This treatment did not affect NaCl detection or amiloride sensitivity in adulthood. The amiloride-induced shifts in NaCl sensitivity functions imply that the transcellular sodium transduction pathway is necessary for normal NaCl detection in the rat.     

Lack of amiloride sensitivity in SHR and WKY glossopharyngeal taste responses to NaCl.

To explore possible functional strain differences in taste receptors located on the posterior tongue, we recorded electrophysiological taste responses from the glossopharyngeal nerve of spontaneously hypertensive (SHR) and Wistar-Kyoto (WKY) rats. Multifiber responses to a concentration series (0.5 M to 2.0 M) of NaCl, KCl and NH4Cl were recorded before and after lingual application of the epithelial sodium transport blocker, amiloride. Responses to a concentration series (0.0025 M to 0.1 M) of quinine hydrochloride were also recorded. When expressed relative to the 0.5-M NH4Cl response, responses to the monochloride salts were equivalent between SHR and WKY. Surprisingly, NaCl responses were not suppressed by the sodium transport blocker, amiloride. This is in direct contrast to the dramatic suppression observed in the chorda tympani. Also, relative responses to quinine were greater in the glossopharyngeal nerve of SHR than WKY. These results indicate that taste receptors innervated by the glossopharyngeal nerve lack amiloride sensitivity and that posterior taste receptor function to monochloride salts is equivalent between SHR and WKY.     

Group I metabotropic glutamate receptors on second-order baroreceptor neurons are tonically activated and induce a Na+-Ca2+ exchange current

The nucleus tractus solitarius (NTS) is essential for coordinating baroreflex control of blood pressure. The baroreceptor sensory fibers make glutamatergic synapses onto second-order NTS neurons. Glutamate spillover activates Group II and III presynaptic metabotropic glutamate receptors (mGluRs) on the baroreceptor central terminals to inhibit synaptic transmission, but the role of postsynaptic mGluRs is less understood. We used whole cell patch-clamping in anatomically identified second-order baroreceptor neurons in a brain stem slice to test whether Group I, II, and III mGluRs had postsynaptic effects at this first central synapse in the baroreceptor afferent pathway. The Group I agonist DHPG induced a depolarization and spiking that was mimicked by endogenous glutamate. Group I mGluR blockade prevented the depolarization and slightly hyperpolarized the neurons, suggesting a small tonic Group I mGluR activation. The DHPG-induced inward current consisted of voltage-dependent and -independent components; the former was blocked by TEA and the latter was blocked by replacing extracellular NaCl with LiCl or Tris-HCl. The DHPG current was potentiated in a Ca2+-free external solution and was diminished by intracellular dialysis with BAPTA and by perfusion with Na+-Ca2+ exchanger blockers, KB-R7943 or 3',4'-dichlorobenzamil. Intracellular dialysis with GDPbetaS or heparin and perfusion with the PLC inhibitor U-73122 or the Ca2+-calmodulin inhibitor W-7 significantly decreased the DHPG current. The data suggest that Group I mGluRs on baroreceptor neurons are functional; are activated by endogenous glutamate; and activate a Na+-Ca2+ exchanger through G-protein, PLC, IP3, and Ca2+-calmodulin mechanisms to excite the cell, thus providing postsynaptic mechanisms to enhance or prolong baroreceptor signal transmission.     

Volume regulation by Necturus gallbladder: apical Na+-H+ and Cl(-)-HCO-3 exchange

Necturus gallbladder epithelial cells exhibited volume regulatory swelling when exposed to a hypertonic mucosal bathing solution. The initial, osmotically induced shrinkage was followed by a rapid increase in cell volume back to the control value despite continuing hypertonicity of the mucosal perfusate. This volume regulatory increase occurred by osmotic water flow accompanying the transient cellular uptake of NaCl from the mucosal bathing solution. Volume regulatory increase required Na+ and Cl- in the mucosal bath; it was inhibited by amiloride or 4-acetamido-4'-isothiocyanostilbene-2,2'-disulfonic acid but not by bumetanide or ouabain. The K1/2 for Na+ was 2.8 mM, the K1/2 for Cl- was 1.9 mM, and maximum velocity of fluid flow into the cell for both ions was greater than 10 x 10(-6) cm/s. Both volume regulatory increase and transepithelial fluid absorption involve NaCl flux across the apical membrane into the cells, but the nature of the NaCl fluxes differ in the two processes. During volume regulatory increase NaCl enters the cells by parallel Na+-H+ and Cl(-)-HCO-3 exchanges, whereas during transepithelial fluid absorption NaCl enters the cell by the coupled flux of NaCl.     

Evidence for Na+ independent active secretion of K+ and HCO 3 − by rat salivary duct epithelium

In order to elucidate whether or not active secretion of potassium and bicarbonate by the rat submaxillary duct epithelium operates independently of sodium reabsorption, Na+ transport was blocked by amiloride, which is known to inhibit Na+ entry from lumen into cell. With 10(-4) M amiloride in HCO - 3 -Ringer at the luminal side, the transepithelial electrical potential difference approached zero, the Na+ conductance of the luminal cell membrane was drastically reduced, and the K+ conductance was significantly reduced. Net K+ secretion was reduced by 80%, whereas net HCO - 3 secretion was significantly increased. The remaining 20% of net K+ secretion proceeded at zero net Na+ transport and in the absence of significant chemical and electrical potential differences between lumen and interstitium of the duct. This active component of net K+ secretion was accompanied by an equal rate of active HCO - 3 secretion. These findings confirm the independence of this active secretion of K+ and HCO - 3 from Na+ transport. They indicate an electrically neutral secretion of K+ and HCO - 3, probably by the postulated luminal K+ -H+ -exchange mechanism. The 80% of net K+ secretion, which were abolished by amiloride together with Na+ reabsorption, seem to be functionally coupled with Na+ transport. The linkage of K+ -to- Na+ is probably mediated by a luminal carrier exchanging Na+ for K+ and H+.     

Two different receptor sites for Ca2+ and Na+ in frog taste responses

Distilled water, 1 mM CaCl2 and 500 mM NaCl (pH 4.5) are effective stimuli which excite chemoreceptors of the frog tongue. To learn whether or not these taste stimuli react with different taste receptor sites, a proteolytic enzyme was topically applied to the tongue dorsum. Responses were recorded from the frog glossopharyngeal nerve during taste stimulation. After application of 0.1% pronase E to the dorsal tongue surface, the magnitude of the NaCl response remained unchanged, but the magnitude of the water and CaCl2 responses was markedly decreased. The selective suppression by the pronase E treatment indicates that there are two different receptor sites for Ca2+ and Na+ in the frog taste receptor cell and the receptor sites responsible for the generation of the water and the Ca2+ response may be composed of a protein.     

Na+-H+ exchange activity in taste receptor cells

mRNA for two Na(+)-H(+)-exchanger isoforms 1 and 3 (NHE-1 and NHE-3) was detected by RT-PCR in fungiform and circumvallate taste receptor cells (TRCs). Anti-NHE-1 antibody binding was localized to the basolateral membranes, and the anti-NHE-3 antibody was localized in the apical membranes of fungiform and circumvallate TRCs. In a subset of TRCs, NHE-3 immunoreactivity was also detected in the intracellular compartment. For functional studies, an isolated lingual epithelium containing a single fungiform papilla was mounted with apical and basolateral sides isolated and perfused with nominally CO(2)/HCO(3)(-)-free physiological media (pH 7.4). The TRCs were monitored for changes in intracellular pH (pH(i)) and Na(+) ([Na(+)](i)) using fluorescence ratio imaging. At constant external pH, 1) removal of basolateral Na(+) reversibly decreased pH(i) and [Na(+)](i); 2) HOE642, a specific blocker, and amiloride, a nonspecific blocker of basolateral NHE-1, attenuated the decrease in pH(i) and [Na(+)](i); 3) exposure of TRCs to basolateral NH(4)Cl or sodium acetate pulses induced transient decreases in pH(i) that recovered spontaneously to baseline; 4) pH(i) recovery was inhibited by basolateral amiloride, 5-(N-methyl-N-isobutyl)-amiloride (MIA), 5-(N-ethyl-N-isopropyl)-amiloride (EIPA), HOE642, and by Na(+) removal; 5) HOE642, MIA, EIPA, and amiloride inhibited pH(i) recovery with K(i) values of 0.23, 0.46, 0.84, and 29 microM, respectively; and 6) a decrease in apical or basolateral pH acidified TRC pH(i) and inhibited spontaneous pH(i) recovery. The results indicate the presence of a functional NHE-1 in the basolateral membranes of TRCs. We hypothesize that NHE-1 is involved in sour taste transduction since its activity is modulated during acid stimulation.     

Influence of voltage and extracellular Na+ on amiloride block and transport kinetics of rat epithelial Na+ channel expressed in Xenopus oocytes

We expressed the three subunits of the epithelial amiloride-sensitive Na(+) channel (ENaC) from rat distal colon heterologously in oocytes of Xenopus laevis and analysed blocker-induced fluctuations in current using conventional dual-microelectrode voltage-clamp. To minimize Na(+) accumulation we performed all experiments in low-Na(+) solutions (15 mM). Noise analysis revealed that control or ENaC-injected oocytes did not exhibit spontaneous relaxation noise. However, in ENaC-expressing oocytes, amiloride induced a distinct Lorentzian component in the power density spectra. With three amiloride concentrations and a linear analysis of the respective changes in the corner frequency f(c) (2 pi f(c) plot) we determined the rate constants k(on) and k(off) for the amiloride-ENaC interaction. At a clamp potential (V(m)) of -60 mV k(on) was 80.8 +/- 5.1 microM(-1) s(-1) and k(off) 15.4 +/- 4.2 s(-1). The half-maximal blocker concentration (K(mic,ami)) was 0.19 microM (V(m)=-60 mV). While k(on) was voltage-independent in the range -50 to -100 mV, k(off) and K(mic,ami) decreased significantly with increasing membrane hyperpolarization, resulting in an increased affinity of amiloride for its binding site on ENaC. Increasing extracellular [Na(+)] ([Na(+)](o)) led to saturation of ENaC. Subsequent noise analysis revealed that single-channel current increased non-linearly with [Na(+)](o) and that saturation was not due to a reduction in the number of open channels. The apparent affinity of Na(+) for its binding site on the channel was voltage dependent and increased with hyperpolarization. Noise analysis revealed that k(on) and k(off) for amiloride decreased with increasing [Na(+)](o), while the affinity of the amiloride-binding site did not change. These findings show that the affinity of rat intestinal ENaC for amiloride is voltage dependent and is influenced non-competitively by [Na(+)](o), indicating that Na(+) and amiloride do not compete for the same binding site at the channel.     

Peripheral taste responses in genetically hypertensive rats

In two-bottle preference-aversion tests, the spontaneously hypertensive rat (SHR) tolerates higher concentrations of NaCl than the normotensive Wistar-Kyoto (WKY). In contrast, the inbred Dahl salt-sensitive (S/JR) and inbred Dahl salt-resistant (R/JR) rat show similar preferences for NaCl. In order to determine if taste receptor function was also altered between the hypertensive rat and its normotensive control, we recorded electrophysiological taste responses from the chorda tympani (CT) nerve in SHR, WKY, S/JR and R/JR rats. Responses to a concentration series (0.05 M to 0.5 M) of NaCl, NaAcetate, KCl, NH4Cl and CaCl2 were recorded before and after lingual application of amiloride hydrochloride, an epithelial sodium transport blocker. When expressed relative to the 0.5 M NH4Cl response, responses to the majority of stimuli were equivalent between the SHR and WKY. By comparison, relative responses to NaCl were greater in the R/JR than S/JR; however, the magnitude of amiloride suppression was equivalent between these two strains. Relative responses to the majority of the remaining salts did not differ between the S/JR and R/JR. These results suggest that taste receptor function may be equivalent between the hypertensive rat and its normotensive control.     

Sodium detection during the water absorption response in Rana pipiens

Although taste in vertebrates is typically associated with specialized receptors in the lingual epithelium, Hoff and Hillyard reported that the toad, Bufo punctatus, is able to "taste" sodium with the abdominal skin. This was reflected in a differential behavioral response to hypertonic NaCl. The present study tests for the presence of such abdominal chemoreceptors in the frog Rana pipiens. The experiment was a five-condition design in which frogs were placed on filter paper saturated with: deionized water, 250 mM NaCl, 350 mM NaCl, 12.9 microM amiloride, or 350 mM NaCl + 12.9 microM amiloride. The time that the frogs remained on the test substrate before moving to a surface of deionized water was recorded. It was necessary to dehydrate the frogs to 80% of their body weight to elicit a behavioral response to the NaCl whereas dehydration to 90% of their body weight has been reported effective in Bufo punctatus. The frogs displayed significantly shorter mean times to move on both concentrations of NaCl compared to deionized water, with the shortest times occurring when 350 mM NaCl was used. Amiloride alone did not have an effect upon times to move to deionized water, but did significantly reduce the response to 350 mM NaCl. Movement to amiloride + 350 mM NaCl did not differ significantly from that to deionized water. The results indicate that Rana pipiens, like Bufo punctatus, have epithelial chemoreceptors for the detection of NaCl on hydrated surfaces and that these receptors, like those of mammals, are amiloride sensitive.     

Thyroid hormone control of Na+-K+-ATPase and K+-dependent phosphatase in rat heart.

To assess the possible role of the Na+ pump in mediating physiological responses to thyroid hormone in the rat myocardium, we examined the effects of L-3,5,3'-triiodothyronine (T3) on the activities of the closely associated enzymes, Na+-K+-dependent adenosine triphosphatase (Na-K-ATPase) and K+-dependent p-nitrophenyl phosphatase (K-dep-pNPPase). In hypothyroid rats, administration of T3 (50 microng/100 g body wt) resulted in significant increases (greater than 50%) in Na-K-ATPase and K-dep-pNPPase activities in both crude homogenates and microsomal fractions of the rat ventricle. Significant effects on Na-K-ATPase activity were also attained with low doses (1 microng/100 g body wt) of T3. A method was developed for assaying K-dep-pNPPase activity in cardiac slices. With this technique, enhancement in K-dep-pNPPase activity of 89.2% was found in ventricle slices after treatment of hypothyroid rats with T3 (50 microng/100 g body wt), implying that augmentation of the capacity of the Na+ pump is achieved in vivo. The potent analogue, L-3,5-diiodo-3' isopropyl thyronine (isopropyl T2) had the same effects on cardiac growth and Na-K-ATPase as T3, in hypothyroid rats. In contrast, the relatively inactive isomer, L-3,3',5'-triiodothyronine (reverse T3) had no significant effect on the heart weight-to-body weight ratio or on ventricular Na-K-ATPase activity.     

Sodium taste detectability in rats is independent of anion size: the psychophysical characteristics of the transcellular sodium taste transduction pathway

There are two known sodium transduction pathways in the rat gustatory system. The transcellular pathway is blocked by amiloride, and the paracellular pathway is limited by the anion gluconate. The contribution of each pathway to sodium detection was assessed. Sodium gluconate (NaGlu) and NaCl thresholds did not differ, implying that the paracellular pathway is not necessary for normal sodium detection. Adding 100 microM amiloride raised both NaCl and NaGlu thresholds but did not abolish all performance to NaGlu, indicating that some chemical cue was present at high concentrations. Rats were also exposed to one of three NaCl diets (0.12%, 1.0%, or 6.0% NaCl) through maternal and ad lib intake from Embryonic Day 1 through testing in adulthood. No differences across dietary groups were found for NaCl or NaGlu threshold with or without amiloride. Thus, this developmental dietary treatment does not appear to affect taste sensitivity to sodium subserved through either transduction pathway. Collectively, these data suggest that the transcellular transduction pathway is both necessary and sufficient for normal sodium detection.     

Na(+)-H+ exchange is not important for pancreatic HCO3- secretion in the pig.

Pancreatic inter- and intralobular duct cells extrude H(+)-ions to interstitial fluid when they secrete HCO3- to pancreatic juice. This study assesses the potential importance of Na(+)-H(+)-ion exchange for H(+)-ion extrusion and secretion of HCO3-, using the Na(+)-H+ exchange blockers amiloride and hexamethylene-amiloride. Intracellular pH (pHi) in inter- and intralobular pancreatic duct epithelium was measured using BCECF fluorescence. H(+)-ion efflux was measured using a NH4Cl prepulse, acid-loading technique. In HCO3(-)-free media, pHi recovery following acid loading was blocked by amiloride (10(-4) M) and hexamethylene-amiloride (10(-6) M), demonstrating amiloride- and hexamethylene-amiloride-sensitive Na(+)-H+ exchange. However, 5 x 10(-6) M hexamethylene-amiloride did not reduce secretin-dependent pancreatic HCO3- secretion in vivo. Maximal H(+)-efflux through Na(+)-H+ exchange was 1.5 +/- 0.2 mumol min-1 ml cell volume-1, i.e. less than 1% of estimated net H(+)-ion efflux during HCO3- secretion. Conclusion: amiloride- and hexamethylene amiloride sensitive Na(+)-H+ exchange is not important for secretin-dependent pancreatic HCO3- secretion in the pig. Other mechanisms for H+ extrusion dominate.     

Block by amiloride and its derivatives of mechano-electrical transduction in outer hair cells of mouse cochlear cultures.

1. The effects of amiloride and amiloride derivatives on mechano-electrical transducer currents in outer hair cells of the cultured neonatal mouse cochlea were examined under whole-cell voltage clamp. 2. At -84 mV transducer currents were reversibly blocked by the extracellular application of the pyrazinecarboxamides amiloride, benzamil, dimethylamiloride, hexamethyleneiminoamiloride, phenamil and methoxynitroiodobenzamil with half-blocking concentrations of 53, 5.5, 40, 4.3, 12 and 1.8 microM, respectively. Hill coefficients were determined for all but the last of these compounds and were 1.7, 1.6, 1.0, 2.2 and 1.6, respectively, suggesting that two drug molecules co-operatively block the transducer channel. 3. Both the structure-activity sequence for amiloride and its derivatives and the mechanism of the block of the transducer channel appear to be different from those reported for the high-affinity amiloride-sensitive epithelial Na+ channels but similar to those of stretch-activated channels in Xenopus oocytes. 4. The block by all pyrazinecarboxamides was voltage dependent with positive membrane potentials releasing the block. The form of the voltage dependence is consistent with a voltage-independent binding of the drug to a site that is accessible at hyperpolarized but not at depolarized potentials, suggesting that the transducer channel undergoes a voltage-dependent conformational change. The channel was not blocked by 1 mM amiloride from the intracellular side at either negative or positive membrane potentials. 5. The kinetics of the block were studied using force steps or voltage jumps. The results suggest that the drug binding site is only accessible when the transducer channel is open (open-channel block) and that the channel cannot close when the drug molecules are bound. 6. The time dependence and voltage dependence of the block together reveal that the transducer channel has at least two open conformational states, the transition between which is voltage dependent.