Basis for intracellular retention of a human mutant of the retinal rod channel alpha subunit.

A mutant of the a subunit of the retinal rod cyclic GMP-gated channel, [Arg654(1-bp del)], corresponding to a truncated alphaR654Dstop subunit, was previously described in patients with retinitis pigmentosa: when expressed in HEK-293 cells, this mutated a subunit was retained inside the cell, but had normal channel activity in one case where it reached the plasma membrane, indicating that the mechanism of targeting is altered by the mutation, but not the function of the channel. The corresponding mutants of the bovine rod channel (alphaR656D stop), and of the closely related olfactory neuron channel (alphaR632Dstop) alpha subunits were expressed in Xenopus oocytes and their activity was analyzed by patch-clamp. Like their human homologue, these two channels have no activity, and we show that their GFP fusion proteins are accumulated into intracellular compartments. The truncation alone or the R/D mutation alone do not prevent or modify channel activity, indicating that neither the R656 residue nor the C-terminal domain downstream of R656 is necessary for homomeric channel targeting and function. Several mutations of R656 and of the preceding residues in the R656Dstop mutant disclose that the motif responsible for the absence of channel activity is an endoplasmic reticulum retention signal (KXKXXstop) in which the nature of the residues in positions -1 and -4 is determinant.

Coexpression of alpha and beta subunits of the rod cyclic GMP-gated channel restores native sensitivity to cyclic AMP: role of D604/N1201.

Coexpression of the betawt and alphawt subunits of the bovine rod channel restores two characteristics of the native channels: higher sensitivity to cAMP and potentiation of cGMP-induced currents by low cAMP concentrations. To test whether the increased sensitivity to cAMP is due to the uncharged nature of the asparagine residue (N1201) situated in place of aspartate D604 in the beta subunit as previously suggested (, Neuron. 15:619-625), we compared currents from wild-type (alphawt and alphawt/betawt) and from mutated channels (alphaD604N, alphaD604N/betawt, and alphawt/betaN1201D). The results show that the sensitivity to cAMP and cAMP potentiation is partly but not entirely determined by the charge of residue 1201 in the beta subunit. The D604N mutation in the alpha subunit and, to a lesser extent, coexpression of the betawt subunit with the alphawt subunit reduce the open probability for cGMP compared to that of the alphawt channel. Interpretation of the data with the MWC allosteric model (model of Monod, Wyman, Changeux;, J. Mol. Biol. 12:88-118) suggests that the D604N mutation in the alpha subunits and coassembly of alpha and beta subunits alter the free energy of gating by cAMP more than that of cAMP binding.

Magnesium ions promote assembly of channel-like structures from beticolin 0, a non-peptide fungal toxin purified from Cercospora beticola.

Beticolins are toxins produced by the fungus Cercospora beticola. Using beticolin 0 (B0), we have produced a strong and Mg(2+)-dependent increase in the membrane conductance of Arabidopsis protoplasts and Xenopus oocytes. In protein-free artificial bilayers, discrete deflexions of current were observed (12 pS unitary conductance in symmetrical 100 mM KCl) in the presence of B0 (approximately 10 microM) and in the presence of nominal Mg2+. Addition of 50 microM Mg2+ induced a macroscopic current which could be reversed to single channel current by chelating Mg2+ with EDTA. Both unitary and macroscopic currents were ohmic. The increase in conductance of biological membranes triggered by B0 is therefore likely to originate from the ability of this toxin to organize itself into transmembrane pores in the presence of Mg2+. The pore is poorly selective, displaying permeability ratios PCl/PK, PNa/PK and PCa/PK close to 0.3, 0.65 and 0.4, respectively. Such channel-like activity could be involved in the deleterious biological activity of the toxin, by causing the collapse of ionic and electrical gradients through biological membranes together with Ca2+ influx and scrambling of cellular signals.

Effects of cysteine modification on the activity of the cGMP-gated channel from retinal rods.

The effect of sulfhydryl reagents on the activity of the cGMP-gated channel from bovine retinal rods was studied by measurements of 8-Br-cGMP-(cGMP)-induced calcium efflux from rod membrane vesicles and records of 8-Br-cGMP-dependent sodium currents through channels incorporated into planar lipid bilayers. N-ethylmaleimide and mersalyl (thiol blockers) as well as diamide (dithiol-disulfide conversion agent) have a dual effect on the channels activity: at low concentration, they increase the apparent affinity for cyclic nucleotide ("activation") at the same time inducing a loss of cooperativity for nucleotide binding; at higher concentration, N-ethylmaleimide and diamide produce a reduction of the amplitude and initial rate of the calcium release at saturating nucleotide concentration, while mersalyl is shown to reduce the activity of the channels in bilayer experiments ("inhibition"). Nitric oxide precursors have no effect. The results suggest that blocking at least 1 of the 3 cytoplasmic cysteine residues situated close to the cGMP-binding site in each channel subunit by N-ethylmaleimide, mersalyl, or diamide (forming a dimer between 2 subunits) increases the affinity for the nucleotide. Inhibition is produced by blocking at least one of the 2 other cytoplasmic sulfhydryl groups (N-ethylmaleimide, mersalyl, oxidized glutathione) or the 2 others (diamide, intrasubunit bridge), and may concern a process of channel inactivation. The 3 cytoplasmic sulfhydryl groups are accessible when the channels are in the open state, but not (or much less) accessible when the channels are in the closed state.

Gating of retinal rod cation channel by different nucleotides: comparative study of unitary currents.

Single channels are observed after incorporation of native vesicles from bovine rod outer segment membranes into planar lipid bilayers. The activity of a single channel in the presence of cGMP is compared to that induced by the analog 8-bromo-cGMP and by cAMP. At +80 mV, K0.5 is about 3 microM for 8Br-cGMP, 18 microM for cGMP and 740 microM for cAMP. In cAMP, the amplitude of the current is smaller than in cGMP or 8Br-cGMP and depends on the filter cut-off frequency. The open/closed transition rates of the channel are slightly slower with 8Br-cGMP than with cGMP while they are 5 to 10 times faster with cAMP. Addition of Ni2+ ions to either cGMP or cAMP increases the open probability: the open/closed transition rates and amplitude of the current in cAMP are then comparable to those in cGMP. A dual effect of the addition of cAMP on the cGMP- or 8Br-cGMP-dependent activity previously reported (Furman, R.E., Tanaka, J.C. 1989. Biochemistry 28:2785-2788) is observed with a single channel: addition of subthreshold cAMP concentrations to cGMP (or to 8Br-cGMP) markedly increases Po; addition of cAMP concentrations higher than about 70 microM progressively accelerates the kinetics and reduces the amplitude to values observed in cAMP alone. The results are discussed in relation with the model previously proposed to account for the existence of four current levels (Ildefonse, M., Bennett, N. 1991. J. Membrane Biol. 123:133-147).

Single-channel study of the cGMP-dependent conductance of retinal rods from incorporation of native vesicles into planar lipid bilayers.

Unitary currents through cGMP-dependent channels of retinal rods are observed following incorporation into planar lipid bilayers of native vesicles from purified rod outer segment membranes washed free of soluble and peripheral proteins. The influence of the concentration of cGMP, inhibitors (cis-diltiazem, tetracaine and Ag+) and divalent cations (Ca2+, Mg2+, and Co2+) on the conductance and open probability of the channel is described, as well as the voltage dependence of these effects. The cGMP dependence suggests the existence of four binding sites for cGMP and reveals that sequential binding of four cGMP molecules corresponds to the opening of four discrete conductance levels. Finally, we provide conclusive evidence that activated G-protein does not directly inactivate the cGMP-dependent channels of bovine retinal rods.

Direct activation of cGMP-dependent channels of retinal rods by the cGMP phosphodiesterase.

The cationic conductances of purified bovine retinal rod membranes were studied by incorporation of vesicles into planar lipid bilayers. When the membranes were stripped of all peripheral proteins [guanine nucleotide-binding protein (G protein) and cGMP phosphodiesterase (3',5'-cyclic-GMP 5'-nucleotidohydrolase), EC 3.1.4.35], sodium and calcium fluxes were almost only observed in the presence of cGMP. Reconstitution experiments in which purified cGMP phosphodiesterase alone or with G protein were reassociated to the vesicles in proportions similar to those found in the native rod provide evidence for a direct interaction between the cGMP-dependent channel protein and the phosphodiesterase. (i) In its inhibited state, phosphodiesterase markedly stimulates the activity of the channels in the presence of cGMP (situation in the dark-adapted rod) but is not capable of activating the channels in the absence of cGMP. (ii) In the absence of cGMP, activation of the phosphodiesterase by G protein with GTP bound (equivalent to photoexcitation) induces the opening of cation channels that have the same conductance for sodium ions as cGMP-activated channels (20-22 pS, with two sublevels of about 7 pS and 13 pS).

Excitation contraction coupling in skeletal muscle: evidence for a role of slow Ca2+ channels using Ca2+ channel activators and inhibitors in the dihydropyridine series.

Ca2+ current and tension have been simultaneously recorded from single twitch fibres of the semi-tendinosus of Rana esculenta in a medium containing a physiological Ca2+ concentration (1.8 mM). Under appropriate conditions it can be shown that tension develops in two phases. The first is rapid and reaches its maximum before activation of the inward Ca2+ current. The second phase is slower and with a time course which appears to be correlated with that of the inward current. Nifedipine, a specific Ca2+ channel inhibitor greatly reduced ICa2+ and the slower component of tension. Bay K8644 a Ca2+ channel activator, which has receptors on T-tubule, increased ICa2+ and the slow component of tension. These results indicate that a slow component of skeletal muscle contraction is related to the inward Ca2+ current flowing through dihydropyridine sensitive voltage-dependent Ca2+ channels.

Possible role of Na+ ions in intracellular Ca2+ release in frog auricular trabeculae.

Membrane potential-current and mechanical tension of frog atrial muscle were studied in a Ca and Mg-free solution containing 1 mmol/l EGTA (Ca-free solution). Exposure to Ca-free solution resulted in a shortening of action potential duration within 1.5 min and a subsequent lengthening which were paralleled by changes in magnitude and duration of the contraction. Similarly, the slow inward current quickly disappeared and progressively reappeared with a quite slower inactivation time-course. Its reversal potential varied with [Na]0 as for a pure Na current. By 12 min in Ca-free solution, the tension-voltage relation could be interpreted as the sum of two components correlated with the slow inward current and the membrane potential respectively. Contractures in response to sustained large depolarizations had similar time courses in Ca-free solution and Ringer's containing Na-Ca exchange blockers (Mn2+ 15 mmol/l or La3+ 3 mmol/l). Intracellular Na loading by voltage-clamp depolarizations (40 mV from the resting potential for 100 ms, at 0.2 Hz) in the presence of Veratrine (7.5 X 10(-6) g/ml) caused a large progressive increase in tonic tension. An intracellular Ca2+ release is invoked, partly related to Na+ entry and partly to membrane potential changes. The potential dependent part could be influenced by intracellular Na+.

Tityus gamma toxin, a high affinity effector of the Na+ channel in muscle, with a selectivity for channels in the surface membrane.

Toxin gamma from the venom of Tityus serrulatus scorpion produces a partial block of the surface Na+ channel in frog muscle. This block occurs with no change in the voltage-dependence or in the kinetics of the remaining surface Na+ current. The partial blockade of Na+ channel activity occurs with no change in tubular Na+ currents nor in twitch tension. The maximum effect of the toxin is attained at concentrations as low as 3 X 10(-10) M. Hyperpolarization to potentials more negative than the resting potential (E = -90 mV) reduces or abolishes the effect of the toxin. Radioiodinated toxin gamma binds to frog muscle membranes with a very high affinity corresponding to a dissociation constant of about 1 X 10(-11) M. Data obtained with both rabbit and frog muscle indicate that toxin gamma is specific for Na+ channels in surface membranes. Toxin gamma does not seem to bind to Na+ channels in T-tubule membranes. The biochemical data are in good agreement with electrophysiological studies and data on contraction. There is one Tityus gamma toxin binding site per tetrodotoxin binding site in surface membranes. Competition experiments have confirmed that Tityus gamma toxin binds to a new toxin receptor site on the Na+ channel structure. This site is the same that the toxin II from Centruroides suffusus binding site, but this toxin has 100 times less affinity for the Na+ channel than Tityus gamma toxin.

Inhibition by Ba of background current and of its modifications by carbachol in frog atrium.

In previous work on frog atrium [11], we have shown that the carbachol-induced K current (iCCh) and the inward rectifying K channel (iK1) are both inhibited by cesium ions. A more detailed study [1, 2] indicates that the dose-response and the voltage dependence of the inhibition by Cs of iK1 and iCCh are similar. These results are in favor of a modulation by acetylcholine of the iK1 channels. Such a conclusion has been put forward by [3] and by [9]. However, Cs ions can also block the pacemaker currents iK2 [8] and if [5]. An 'if like' component seems to be present in frog atrium (Bonvallet and Ojeda, unpublished results). For these reasons, we have studied the action of barium which can be considered as a more specific inhibitor of the inward rectifying K channels in skeletal muscle [13], [14] and in heart muscle [4]. We report here results obtained on two types of preparations: one in which carbachol increases the background conductance (typical effect) and another one (four fibres from the same frog heart) where carbachol decreases the background conductance (atypical effect) as observed by Carmeliet and Ramon (1980) in sheep Purkinje fibres. In both cases, the results show that, in presence of 5 mM barium, carbachol has no effect on the background current.

Block by Cs of K current iK1 and of carbachol induced K current iCch in frog atrium.

1. In frog atrium, Cs ions block both the inward rectifier iK1 and the carbachol induced K current iCch. 2. Both iK1 and iCch display a high affinity for Cs with a K0.5 of 4 X 10(-5) M for iK1 and of 8 X 10(-5) M for iCch at V = -50 mV. 3. Block of both iK1 and iCch is strongly voltage dependent. When fitted by the block model of Woodhull (1973), delta is greater than 1 for the two currents. 4. From these similarities, action of Cch on frog atrium K permeability could be interpreted as a modification of iK1.

Differences in the properties of Na+ channels in muscle surface and T-tubular membranes revealed by tetrodotoxin derivatives.

The effect of ethylenediamine derivatives of tetrodotoxin (enTTXI and enTTXII) on frog skeletal muscle was studied both electrophysiologically and biochemically. Electrophysiological experiments with one of these molecules (enTTXI) showed that the concentrations needed to block the early phase of the inward sodium current (K0.5 = 7 nM) are much lower than those needed to block the late phase of inward current or muscle contraction (K0.5 = 40 nM). Conversely, tubular Na+ channels are more sensitive to enTTXII than are surface Na+ channels. Toxin binding to isolated muscle membranes was studied using 3H-enTTXI and 3H-enTTXII. The first derivative (3H-enTTXI) has a higher affinity Kd = 8 nM) for Na+ channels in the surface membrane than for Na+ channels in the T-tubular membrane (Kd greater than 20 nM). In contrast 3H-enTTXII has a higher affinity for the tubular Na+ channel (Kd = 0.2 nM) than for the receptor in surface membranes (Kd = 4nM). We conclude that Na+ channels in muscle surface and T-tubular membranes have different toxin-binding properties, which must reflect a difference in molecular structure.

Centruroides toxin, a selective blocker of surface Na+ channels in skeletal muscle: voltage-clamp analysis and biochemical characterization of the receptor.

This paper describes the effects of a toxin from the scorpion Centruroides suffusus suffusus on frog skeletal muscle. The main findings are the following, (i) Centruroides toxin (CssII) blocks the early phase of the inward sodium current in the muscle that arises from influx via Na+ channels in the surface membrane, but it does not affect the late phase of the inward current that represents flux through Na+ channels in the T-tubule membranes, (ii) CssII, in marked contrast to tetrodotoxin, does not affect contraction of the muscle, (iii) Measurements of the binding of 125I-labeled CssII to a partially purified membrane preparation from the muscle indicate that the Kd of the CssII--receptor complex is approximately 0.4 nM. The half-life for the dissociation of this complex is 3 min at 22 degrees C and 16 min at 2 degrees C. Binding of the radiolabeled toxin varies markedly with pH and becomes insignificant at pH greater than 8.5. Proteolytic digestion of the membrane preparation decreases its ability to bind CssII, suggesting that the receptor is a protein. (iv) The number of binding sites for a radiolabeled derivative of tetrodotoxin on the membrane preparation was similar to that for CssII. However, neither tetrodotoxin nor any of seven other neurotoxins and some local anesthetics that alter the functioning of the Na+ channel have any effect on the binding of CssII to the muscle membrane. These results therefore indicate that CssII belongs to a different class of neurotoxins that has a different receptor on the Na+ channel.

Inotropic effects of potassium rich solutions of frog cardiac muscles.

1. Inotropic effects of potassium rich solutions on frog cardiac muscle have been studied bith in current clamp and voltage clamp conditions, using a double sucrose gap apparatus. 2. Potassium rich solutions cause either a positive or a negative inotropic effect, together with an increase or a decrease in the duration of the action potential, according to the preparation. 3. The phasic phase of contraction and the slow inward current are decreased in amplitude; the reversal potential of the slow inward current is shifted towards more negative values. 4. The tonic phase of contraction is first decreased, the increased; the effects are correlated with modifications of the background current, initially in the inward, then in an outward direction. 5. The tension level obtained in contracture experiments is increased or decreased, according to the direction of the changes in the background current. 6. The effects of potassium rich solutions are still observed in he presence of ouabain, suggesting that they are independent of any effect on the sodium-potassium pump. 7. The effects of potassium rich solutions are still observed when external sodium is replaced by sucrose; they disappear (except the effect on the background current) when external sodium is replaced by lithium. 8. The results, which indicate that potassium ions play a role in the regulation of the intracellular concentration of calcium ions, are discussed in relation to a possible K/Ca exchange mechanism, to the Na/Ca exchange and to the role of intracellular calcium stores.

[Demonstration of intracellular release of calcium induced by sodium ions in the auricular trabeculae of the frog]. / Physiologie animale. -Mise en évidence d'une libération intracellulaire de calcium induite par les ions sodium au niveau de trabécules auriculaires de Grenouille.

In a Ca-free, Mg-free medium containing EGTA (10-3M) auricular trabecles develop a slow inward current which is a pure sodium current. After 12 min in this medium, it is still possible to obtain a phasic mechanical activity which shows a perfect correlation with the current. This kind of behavior indicates that a mechanism of sodium-induced calcium release is present at the level of some internal sites.

Evidence for an action of sodium ions in the activation of contraction of twitch muscle fibre.

Simultaneous records in current clamp conditions of potential and of contraction of frog skeletal twitch muscle fibre have shown that a part of the contraction directly depends on the intervention of sodium ions (increase in presence of veratrine, decrease in lithium Ringer). The results confirm previous voltage clamp data and suggest a mechanism of sodium induced calcium release.

Existence of a sodium current in the tubular membrane of frog twitch muscle fibre; its possible role in the activation of contraction.

1. The membrane current of frog twitch muscle fibre has been recorded together with contraction in the test gap of a double sucrose gap apparatus. 2. In Ringer, the inward current sometimes shows 2 phases with the same threshold and the same reversal potential: a rapid one (early inward current) and a slower one (late inward current). The mechanical threshold is near the inward current threshold. The amplitude of the contraction increases progressively with depolarizations, without modification of its time to peak. Experiments with variation of the pulse duration and with conditioning depolarization show that a part of the contraction seems to be correlated with the late inward current. 3. Experiments in a sodium free solution, with low TTX concentration, and on glycerol treated fibres show that the late inward current corresponds to a tubular sodium current. 4. A method is described to separate the two phases of inward current. The smooth development of the current-voltage relation of the late inward current, its diminution without modification in time to peak under the action of TTX, and the exponential decay of its tail current all suggest that the tubular membrane potential is sufficiently well controlled. 5. In the experiments where the tubular membrane potential seems to be controlled, a part of the contraction depends on the tubular sodium current, perhaps involving a mechanism of sodium induced calcium release.