Calcium- and voltage-dependent ion channels in Saccharomyces cerevisiae.

Ion channels in both the tonoplast and the plasma membrane of Saccharomyces cerevisiae have been characterized at the single channel level by patch-clamp techniques. The predominant tonoplast channel is cation selective, has an open-channel conductance of 120 pS in 100 mM KCl, and conducts Na+ or K+ equally well, and Ca2+ to a lesser extent. Its open probability (Po) is voltage-dependent, peaking at about -80 mV (cytoplasm negative), and falling to near zero at +80 mV. Elevated cytoplasmic Ca2+, alkaline cytoplasmic pH, and reducing agents activate the channel. The predominant plasma membrane channel is highly selective for K+ over anions and other cations, and shows strong outward rectification of the time-averaged current-voltage curves in cell-attached experiments. In isolated inside-out patches with micromolar cytoplasmic Ca2+, this channel is activated by positive going membrane voltages: mean Po is zero at negative membrane voltages and near unity at 100 mV. At moderate positive membrane voltages (20-40 mV), elevating cytoplasmic Ca2+ activates the channel to open in bursts of several hundred milliseconds duration. At higher positive membrane voltages, however, elevating cytoplasmic Ca2+ blocks the channel in a voltage-dependent fashion for periods of 2-3 ms. The frequency of these blocking events depends on cytoplasmic Ca2+ and membrane voltage according to second-order kinetics. Alternative cations, such as Mg2+ or Na+, block the yeast plasma-membrane K+ channel in a similar but less pronounced manner.     

A Ca2-activated cation-selective channel in the basolateral membrane of the cortical thick ascending limb of Henle's loop of the mouse

The patch-clamp technique was used to investigate the properties of a cation-selective channel in the basolateral membrane of microdissected collagenase-treated fragments of cortical thick ascending limbs of Henle's loop from mouse kidney. The channel activity was seldom observed in cell-attached patches (2 out 15 studied cases). In inside-out excised patches immersed in symmetrical NaCl Ringer's solutions, the unit channel conductance was ohmic and ranged from 22 to 33 pS (mean, 26.8 +/- 0.6 pS, n = 24). When NaCl was replaced by KCl (n = 8) or sodium gluconate (n = 2) on the cytoplasmic side of the membrane, single-channel currents still reversed at 0 mV and the conductance was unchanged. The reversal potential was +28.8 +/- 0.4 mV (n = 8) when a NaCl concentration (140 vs. 42 mmol/l) gradient was applied, close to the expected value (approx. 30 mV) for a cation selective channel. The channel was found to discriminate poorly between Na+, K+, Cs+, and Li+ ions. The activity of the channel was not clearly voltage-dependent but was dependent upon the free Ca2+ concentration on the cytoplasmic side of the membrane. We conclude that the channel resembles the non-selective cation channel which has been previously described in several tissues.     

Single channel currents in adrenocortical cells

Single channel currents have been recorded from cell-attached patches of tumoral adrenocortical cells. Our experiments suggest the existence of three sets of potassium channels in the surface membrane of these cells. All channel types can be recorded in a given membrane patch but some patches have only one type of single channel currents. One channel type has a unitary conductance of about 103 pS. The other two channels have smaller conductances and opposite voltage dependence. In one case channels open on depolarization and have a single channel conductance of 31.6 pS. In the other case the probability of being in the open state increases on hyperpolarization and the single channel conductance is of 21 pS. These channels seem to be similar to the delayed and anomalous rectifying potassium channels seen in other preparations. The role of membrane ionic permeability in steroid release induced by ACTH is discussed.     

Nonselective cation channels in intact human T lymphocytes.

Nonselective cation channels were found in single channel recordings from cell-attached patches on human T lymphocytes. These channels were active under conditions that should lead to cell swelling (hypotonic bath solutions with NaCl or KCl); however, a definite dependence of activity on cell swelling has not been proven. Under these conditions similar channels were found in 20 of 23 patches from 11 different blood donors. The current-voltage relation was approximately linear for outward current (11-14 pS) and inwardly rectifying (to 23 pS) when the intact cells were depolarized with high KCl in the bath. The voltage dependence of channel activity is consistent with closing at hyperpolarized membrane potentials (Vm less than or equal to -50 mV) and block of open channels at strongly depolarized membrane potentials (Vm greater than 0 mV). Reversal potentials under all ionic gradients tested are consistent with a channel that is poorly selective between Na+ and K+ ions. Active channels in cell-attached patches were rapidly blocked by bath addition of the membrane-permeant inhibitor quinine. Channels that were active in cell-attached became quiescent after patch excision; however, two patches remained active long enough to obtain current-voltage relations. These were linear with a slope conductance for outward current of 8-11 pS. Because of the clustering of single-channel openings, detailed voltage dependence of kinetics and probability of opening were not studied.     

A study of the "outward potassium channel" (OPC) in the frog oocyte. II. A study of the "inside-out" configuration

The single K-channel current reported in a previous note was also studied in "outside-out" conditions. The electrode filling solutions used for the "cell-attached" experiments faced in this case the intracellular side of the membrane patches, the extracellular side facing the bath saline, i.e. Ringer standard. The most significant observations were obtained with filling solutions with varying proportions in K/Na concentrations solutions. In the absence of Na+ ([K+] = 110 mM), the elementary conductance was still around 90 pS and the I/V diagram was again somewhat bell shaped, though the distinctive reduction of the elementary conductance began at more positive potentials (+110 mV). No inward current could be detected upon membrane repolarization also in this case. The rectification became less evident and conductance increased with increasing Na+ concentration in the filling solution, until the I/V curve became a linear one and conductance was 270 pS with standard Ringer. Distinct inward elementary currents were evident upon repolarization in these conditions. Thus a complex interaction between Na+ and K+ takes place for conduction through the outward K channel in the frog oocyte, both cations probably competing for at least one active site inside. Another interesting observation concerns the process of gating of the OPC: the open times of the elementary currents were in fact much greater in outside out experiments as compared to cell-attached experiments, probably due to the presence of Ca++ in contact with the inner membrane side. Even increasing Na+ concentration prolonged the open time duration. The gating of the OPC in the membrane was not only voltage dependent, but also Ca++ and Na+ dependent.     

A calcium and ATP sensitive nonselective cation channel in the antiluminal membrane of rat cerebral capillary endothelial cells.

The patch-clamp technique was applied to the antiluminal membrane of freshly isolated capillaries of rat brain (blood-brain barrier). With 1.3 mM Ca2+ in the bath, excision of membrane patches evoked ion channels, which could not be observed in cell-attached mode. The channel was about equally permeable to Na+ and K+ ions, but not measurable permeable to Cl- and the divalent ions Ca2+ and Ba2+. The current-voltage curve was linear in the investigated voltage range (-80 mV to +80 mV), and the single-channel conductance was 31 +/- 2 pS (n = 22). The channel open probability was not dependent on the applied potential. Lowering of Ca2+ to 1 microM or below on the cytosolic side inactivated the channels, whereas addition of cytosolic ATP (1 mM) inhibited channel activity completely and reversibly. The channel was blocked by the inhibitor of nonselective cation channels in rat exocrine pancreas 3',5-dichlorodiphenylamine-2-carboxylic acid (DCDPC, 10 microM) and by the antiinflammatory drugs flufenamic acid (greater than 10 microM) and tenidap (100 microM), as well as by gadolinium (10 microM). Thus, these nonselective cation channels have many properties in common with similar channels observed in fluid secreting epithelia. The channel could be involved in the transport of K+ ions from brain to blood side.     

Two types of K+ channels in the apical membrane of rabbit proximal tubule in primary culture

The patch-clamp technique was used to investigate ionic channels in the apical membrane of rabbit proximal tubule cells in primary culture. Cell-attached recordings revealed the presence of a highly selective K+ channel with a conductance of 130 pS. The channel activity was increased with membrane depolarization. Experiments performed on excised patches showed that the channel activity depended on the free Ca2+ concentration on the cytoplasmic face of the membrane and that decreasing the cytoplasmic pH from 7.2 to 6.0 also decreased the channel activity. In symmetrical 140 mM KCl solutions the channel conductance was 200 pS. The channel was blocked by barium, tetraethylammonium and Leiurus quinquestriatus scorpion venom (from which charybdotoxin is extracted) when applied to the extracellular face of the channel. Barium and quinidine also blocked the channel when applied to the cytoplasmic face of the membrane. Another K+ channel with a conductance of 42 pS in symmetrical KCl solutions was also observed in excised patches. The channel was blocked by barium and apamin, but not by tetraethylammonium applied to the extracellular face of the membrane. Using the whole-cell recording configuration we determined a K+ conductance of 4.96 nS per cell that was blocked by 65% when 10 mM tetraethylammonium was applied to the bathing medium.     

Cation selective channel in fetal alveolar type II epithelium.

A cation selective channel was identified in the apical membrane of fetal rat (Wistar) alveolar type II epithelium using the patch clamp technique. The single channel conductance was 23 +/- 1.2 pS (n = 16) with symmetrical NaCl (140 mM) solution in the bath and pipette. The channel was highly permeable to Na+ and K+ (PNa/PK = 0.9) but essentially impermeant to chloride and gluconate. Membrane potential did not influence open state probability when measured in a high Ca2+ (1.5 mM) bath. The channel reversibly inactivated when the bath was exchanged with a Ca(2+)-free (less than 10(-9) M) solution. The Na+ channel blocker amiloride (10(-6) M) applied to the extracellular side of the membrane reduced P(open) relative to control patches; P(control) = 0.57 +/- 0.11 (n = 5), P(amiloride) = 0.09 +/- 0.07 (n = 4, p less than 0.01), however, amiloride did not significantly influence channel conductance (g); g(control) 19 +/- 0.9 pS (n = 5), 18 +/- 3.0 pS (n = 4). More than one current level was observed in 42% (16/38) of active patches; multiple current levels (ranging from 2 to 6) were of equal amplitude suggesting the presence of multiple channels or subconductance states. Channel activity was also evident in cell attached patches. Since monolayers of these cells absorb Na+ via an amiloride sensitive transport mechanism we speculate that this amiloride sensitive cation selective channel is a potential apical pathway for electrogenic Na+ transport in the alveolar region of the lung.     

K+ channel cAMP activated in guinea pig gallbladder epithelial cells

In guinea pig gallbladder epithelial cells, an increase in intracellular cAMP levels elicits the rise of anion channel activity. We investigated by patch-clamp techniques whether K(+) channels were also activated. In a cell-attached configuration and in the presence of theophylline and forskolin or 8-Br-cAMP in the cellular incubation bath, an increase of the open probability (P(o)) values for Ca(2+)-activated K(+) channels with a single-channel conductance of about 160 pS, for inward current, was observed. The increase in P(o) of these channels was also seen in an inside-out configuration and in the presence of PKA, ATP, and cAMP, but not with cAMP alone; phosphorylation did not influence single-channel conductance. In the inside-out configuration, the opioid loperamide (10(-5) M) was able to reduce P(o) when it was present either in the microelectrode filling solution or on the cytoplasmic side. Detection in the epithelial cells by RT-PCR of the mRNA corresponding to the alpha subunit of large-conductance Ca(2+)-activated K(+) channels (BK(Ca)) indicates that this gallbladder channel could belong to the BK family. Immunohistochemistry experiments confirm that these cells express the BK alpha subunit, which is located on the apical membrane. Other K(+) channels with lower conductance (40 pS) were not activated either by 8-Br-cAMP (cell-attached) or by PKA + ATP + cAMP (inside-out). These channels were insensitive to TEA(+) and loperamide. The data demonstrate that under conditions that induce secretion, phosphorylation activates anion channels as well as Ca(2+)-dependent, loperamide-sensitive K(+) channels present on the apical membrane.     

Non-selective cation channel on pancreatic duct cells.

We have identified a non-selective cation channel on pancreatic duct cells. These epithelial cells secrete the bicarbonate ions found in pancreatic juice; a process controlled by the hormone secretin, which uses cyclic AMP as an intracellular messenger. The non-selective channel is located on both apical and basolateral plasma membranes of the duct cell, is equally permeable to sodium and potassium, and has a linear I/V relationship with a single-channel conductance of about 25 pS. Channel opening requires the presence of 1 microM Ca2+ on the cytoplasmic face of the membrane, and is also increased by depolarization. Intracellular ATP, ADP, magnesium, and a rise in pH all decreased channel activity. The channel was not affected by 10 mM TEA, 1 mM Ba2+ or 0.5 mM decamethonium applied to the cytoplasmic face of the membrane, but 0.5 mM quinine caused a flickering block which was more pronounced at depolarizing potentials. We observed the channel only rarely in cell-attached patches on unstimulated duct cells, and acute exposure to stimulants did not cause channel activation. However, after prolonged stimulation, the proportion of cell-attached patches containing active channels was increased 9-fold. The role of this channel in pancreatic duct cell function remains to be elucidated.     

Apical K+ channels in Necturus taste cells. Modulation by intracellular factors and taste stimuli

The apically restricted, voltage-dependent K+ conductance of Necturus taste receptor cells was studied using cell-attached, inside-out and outside-out configurations of the patch-clamp recording technique. Patches from the apical membrane typically contained many channels with unitary conductances ranging from 30 to 175 pS in symmetrical K+ solutions. Channel density was so high that unitary currents could be resolved only at negative voltages; at positive voltages patch recordings resembled whole-cell recordings. These multi-channel patches had a small but significant resting conductance that was strongly activated by depolarization. Patch current was highly K+ selective, with a PK/PNa ratio of 28. Patches containing single K+ channels were obtained by allowing the apical membrane to redistribute into the basolateral membrane with time. Two types of K+ channels were observed in isolation. Ca(2+)-dependent channels of large conductance (135-175 pS) were activated in cell-attached patches by strong depolarization, with a half-activation voltage of approximately -10 mV. An ATP-blocked K+ channel of 100 pS was activated in cell-attached patches by weak depolarization, with a half-activation voltage of approximately -47 mV. All apical K+ channels were blocked by the sour taste stimulus citric acid directly applied to outside-out and perfused cell-attached patches. The bitter stimulus quinine also blocked all channels when applied directly by altering channel gating to reduce the open probability. When quinine was applied extracellularly only to the membrane outside the patch pipette and also to inside-out patches, it produced a flickery block. Thus, sour and bitter taste stimuli appear to block the same apical K+ channels via different mechanisms to produce depolarizing receptor potentials.     

Modification of cardiac sodium channels by carboxyl reagents. Trimethyloxonium and water-soluble carbodiimide

In TTX-sensitive nerve and skeletal muscle Na+ channels, selective modification of external carboxyl groups with trimethyloxonium (TMO) or water-soluble carbodiimide (WSC) prevents voltage-dependent Ca2+ block, reduces unitary conductance, and decreases guanidinium toxin affinity. In the case of TMO, it has been suggested that all three effects result from modification of a single carboxyl group, which causes a positive shift in the channel's surface potential. We studied the effect of these reagents on Ca2+ block of adult rabbit ventricular Na+ channels in cell-attached patches. In unmodified channels, unitary conductance (gamma Na) was 18.6 +/- 0.9 pS with 280 mM Na+ and 2 mM Ca2+ in the pipette and was reduced to 5.2 +/- 0.8 pS by 10 mM Ca2+. In contrast to TTX-sensitive Na+ channels, Ca2+ block of cardiac Na+ channels was not prevented by TMO; after TMO pretreatment, gamma Na was 6.1 +/- 1.0 pS in 10 mM Ca2+. Nevertheless, TMO altered cardiac Na+ channel properties. In 2 mM Ca2+, TMO-treated patches exhibited up to three discrete gamma Na levels: 15.3 +/- 1.7, 11.3 +/- 1.5, and 9.8 +/- 1.8 pS. Patch-to-patch variation in which levels were present and the absence of transitions between levels suggests that at least two sites were modified by TMO. An abbreviation of mean open time (MOT) accompanied each decrease in gamma Na. The effects on channel gating of elevating external Ca2+ differed from those of TMO pretreatment. Increasing pipette Ca2+ from 2 to 10 mM prolonged the MOT at potentials positive to approximately -35 mV by decreasing the open to inactivated (O-->I) transition rate constant. On the other hand, even in 10 mM Ca2+ TMO accelerated the O-->I transition rate constant without a change in its voltage dependence. Ensemble averages after TMO showed a shortening of the time to peak current and an acceleration of the rate of current decay. Channel modification with WSC resulted in analogous effects to those of TMO in failing to show relief from block by 10 mM Ca2+. Further, WSC caused a decrease in gamma Na and an abbreviation of MOT at all potentials tested. We conclude that a change in surface potential caused by a single carboxyl modification is inadequate to explain the effects of TMO and WSC in heart. Failure of TMO and WSC to prevent Ca2+ block of the cardiac Na+ channel is a new distinction among isoforms in the Na+ channel multigene family.     

Single potassium channels of human glioma cells.

Single potassium channels in the membrane of human malignant glioma cells U-118MG were studied using the technique of patch clamp in cell-attached and inside-out configurations. Three types of potassium channels were found which differed from each other under conditions close to physiological in their conductance and gating characteristics. The lowest-conductance channel (20 pS near the reversal potential) showed a mild outward rectification up to 45 pS at positive voltages and spontaneous modes of high and low activity. At extreme values of potentials its activity was generally low. The intermediate conductance channel had an S-shaped I-V curve, giving a conductance of 63 pS at reversal, and a low and voltage independent opening probability. The high-conductance (215 pS) channel was found to be activated by both membrane potential and Ca2+ ions and blocked by internal sodium at high voltages. The current-voltage curves of all three channel types displayed saturation.     

Basolateral outward rectifier chloride channel in isolated crypts of mouse colon

Single channel patch-clamp techniques were used to demonstrate the presence of outwardly rectifying chloride channels in the basolateral membrane of crypt cells from mouse distal colon. These channels were rarely observed in the cell-attached mode and, in the inside-out configuration, only became active after a delay and depolarizing voltage steps. Single channel conductance was 23.4 pS between -100 and -40 mV and increased to 90.2 pS between 40 and 100 mV. The channel permeability sequence for anions was: I(-) > SCN(-) > Br(-) > Cl(-) > NO(3)(-) > F(-)> SO(4)(2-) approximately gluconate. In inside-out patches, the channel open probability was voltage dependent but insensitive to intracellular Ca(2+) concentration. In cell-attached mode, forskolin, histamine, carbachol, A-23187, and activators of protein kinase C all failed to activate the channel, and activity could not be evoked in inside-out patches by exposure to the purified catalytic subunit of cAMP-dependent protein kinase A. The channel was inhibited by 5-nitro-2-(3-phenylpropylamino)benzoate, 9-anthracenecarboxylic acid, and DIDS. Stimulation of G proteins with guanosine 5'-O-(3-thiotriphosphate) decreased the channel open probability and conductance, whereas subsequent addition of guanosine 5'-O-(2-thiodiphosphate) reactivated the channel.     

Basic properties and potential regulators of the apical K+ channel in macula densa cells

These studies examine the properties of an apical potassium (K+) channel in macula densa cells, a specialized group of cells involved in tubuloglomerular feedback signal transmission. To this end, individual glomeruli with thick ascending limbs (TAL) and macula densa cells were dissected from rabbit kidney and the TAL covering macula densa cells was removed. Using patch clamp techniques, we found a high density (up to 54 channels per patch) of K+ channels in the apical membrane of macula densa cells. An inward conductance of 41.1 +/- 4.8 pS was obtained in cell-attached patches (patch pipette, 140 mM K+). In inside- out patches (patch pipette, 140 mM; bath, 5 mM K+), inward currents of 1.1 +/- 0.1 pA (n = 11) were observed at 0 mV and single channel current reversed at a pipette potential of -84 mV giving a permeability ratio (PK/PNa) of over 100. In cell-attached patches, mean channel open probability (N,Po, where N is number of channels in the patch and Po is single channel open probability) was unaffected by bumetanide, but was reduced from 11.3 +/- 2.7 to 1.6 +/- 1.3 (n = 5, p < 0.02) by removal of bath sodium (Na+). Simultaneous removal of bath Na+ and calcium (Ca2+) prevented the Na(+)-induced decrease in N.Po indicating that the effect of Na+ removal on N.Po was probably mediated by stimulation of Ca2+ entry. This interpretation was supported by studies where ionomycin, which directly increases intracellular Ca2+, produced a fall in N.Po from 17.8 +/- 4.0 to 5.9 +/- 4.1 (n = 7, p < 0.02). In inside- out patches, the apical K+ channel was not sensitive to ATP but was directly blocked by 2 mM Ca2+ and by lowering bath pH from 7.4 to 6.8. These studies constitute the first single channel observations on macula densa cells and establish some of the characteristics and regulators of this apical K+ channel. This channel is likely to be involved in macula densa transepithelial Cl- transport and perhaps in the tubuloglomerular feedback signaling process.     

Ion channels in mammalian proximal renal tubules.

In the plasma membranes of mammalian proximal renal tubules single ion channels were investigated mainly in isolated tubules perfused on one side, in isolated nonperfused (collapsed) tubules and in primary cell cultures. With these techniques, the following results were obtained: in the luminal membrane of isolated one-sided perfused tubules of rabbit and mouse S3 segments, K(+)-selective channels with single-channel conductance (g) of 33 pS and 63 pS, respectively, were recorded. In primary cultures of rabbit S1 segments, a small-conductance (42 pS) as well as a large-conductance (200 pS) K+ channel were observed. The latter was Ca2(+)- and voltage-sensitive. In cultured cells a Ca2(+)-activated, nonselective cation channel with g = 25 pS was also recorded. On the other hand, an amiloride-sensitive channel with g = 12 pS, which was highly selective for Na+ over K+, was observed in the isolated perfused S3 segment. In the basolateral membrane of isolated perfused S3 segments, two types of K+ channels with g = 46 pS and 36 pS, respectively, were observed. The latter channel was not dependent on cytosolic Ca2+ in cell-excised patches. A K+ channel with g = 54 pS was recorded in isolated nonperfused S1 segments. This channel showed inward rectification and was more active at depolarizing potentials. In isolated perfused S3 segments, in addition to the K+ channels also a nonselective cation channel with g = 28 pS was observed. This channel was highly dependent on cytosolic Ca2+ in cell-free patches. It can be concluded that the K+ channels both in the luminal and contraluminal cell membrane are involved in the generation of the cell potential. Na+ channels in the luminal membrane may participate in Na+ reabsorption, whereas the function of a basolateral cation channel remains unclear. Recently, single anion-selective channels were recorded in membranes of endocytotic vesicles, isolated from rat proximal tubules. Vesicles were enlarged by the dehydration/rehydration method and investigated with the patch clamp technique. The Cl- channel had a conductance of 73 pS, the current-voltage curve was linear and the channel inactivated at high negative clamp potentials. It is suggested that this channel is responsible for charge neutrality during active H+ uptake into the endosomes.     

Biophysical properties and metabolic regulation of a TASK-like potassium channel in rat carotid body type 1 cells

The single channel properties of TASK-like oxygen-sensitive potassium channels were studied in rat carotid body type 1 cells. We observed channels with rapid bursting kinetics, active at resting membrane potentials. These channels were highly potassium selective with a slope conductance of 14-16 pS, values similar to those reported for TASK-1. In the absence of extracellular divalent cations, however, single channel conductance increased to 28 pS in a manner similar to that reported for TASK-3. After patch excision, channel activity ran down rapidly. Channel activity in inside-out patches was markedly increased by 2 and 5 mM ATP and by 2 mM ADP but not by 100 microM ADP or 1 mM AMP. In cell-attached patches, both cyanide and 2,4-dinitrophenol strongly inhibited channel activity. We conclude that 1) whilst the properties of this channel are consistent with it being a TASK-like potassium channel they do not precisely conform to those of either TASK-1 or TASK-3, 2) channel activity is highly dependent on cytosolic factors including ATP, and 3) changes in energy metabolism may play a role in regulating the activity of these background K+ channels.     

Serotonin stimulates a Ca2+ permeant nonspecific cation channel in hepatic endothelial cells.

The contraction of hepatic endothelial cell fenestrae after exposure to serotonin is associated with an increase in intracellular Ca2+ which is derived from extracellular Ca2+, is inhibited by pertussis toxin and is not associated with activation of phosphoinositol turnover or cAMP. Using cell-attached and excised patches in primary cultures of rat hepatic endothelial cells, we identified a serotonin-activated channel with conductance of 26.4+/-2.3 pS. The channel was also permeant to Na+, K+ and Ca++ but not to anions. In cell-attached patch recordings, addition of serotonin to the bath significantly increased channel activity with Ca2+ or Na+ as the charge-carrying ions. This channel provides a mechanism whereby serotonin can raise the cytosolic Ca2+ concentration in hepatic endothelial cells.     

L-glutamate activates excitatory and inhibitory channels in Drosophila larval muscle

Muscle fibers from Drosophila larvae show an L-glutamate-sensitive membrane potential. Bath-applied L-glutamate depolarizes the muscle in the range from 0.5 to 20 microM. Greater concentrations of the agonist repolarize the fibers. The repolarizing effect disappears if chloride is replaced by sulfate in the external medium. Intracellular recordings show the occurrence of depolarizing and hyperpolarizing spontaneous miniature postsynaptic potentials (smpp). Patch-clamp studies indicate the presence of two types of receptor channels: (i) an anion-selective channel activated by both L-glutamate and GABA. In outside out-patches, bathed in symmetrical 140 mM Cl- and 200 microM GABA, the channel displays conductance substates of 40, 80 and 110 pS. In the presence of 200 microM L-glutamate only the 40 and 80 pS substates are observed; (ii) a cation-selective channel activated only by L-glutamate that has a conductance of 104 pS in cell-attached patches (128 mM Na+ outside). The presence of these two types of receptor channels in Drosophila muscle may explain the effect of bath-applied L-glutamate on membrane potential and the presence of inhibitory and excitatory smpp.