Blockade of neuronal facilitatory nicotinic receptors containing alpha 3 beta 2 subunits contribute to tetanic fade in the rat isolated diaphragm.

Blockade of Neuronal Facilitatory Nicotinic Receptors Containing ␣3␤2 Subunits Contribute to Tetanic Fade in the Rat Isolated Diaphragm MIGUEL FARIA, LAURA OLIVEIRA, M. ALEXANDRINA TIMO ´TEO, M. GRAC¸ A LOBO, AND PAULO CORREIA-DE-SA ´ Laborato´rio de Farmacologia, Unidade Multidisciplinar de Investigac¸a˜o Biome´dica (UMIB), Instituto de Cieˆncias Biome´dicas de Abel Salazar (ICBAS), Universidade do Porto, 4099-003 Porto, Portugal KEY WORDS neuromuscular junction; acetylcholine release; safety factor; muscle relaxants ABSTRACT Nicotinic receptor (nAChR) subtypes involved in preand postjunctional actions underlying tetanic fade were studied in rat phrenic-nerve hemidiaphragms. We investigated the ability of subtype-specific nAChR antagonists to depress nerve-evoked contractions and [ 3 H]-acetylcholine ([ 3 H]-ACh) release. Muscle tension was transiently increased during brief high frequency trains (50 Hz for 5 sec). The rank potency order of nAChR antagonists to reduce tetanic peak tension was ␣-bungarotoxin ⬎d-tubocurarine Ⰷmecamylamine ⬎hexamethonium. Reduction of maximal tetanic tension produced by dihydro-␤-erythroidine (0.03–10 ␮M), methyllycaconitine (0.003–3 ␮M), and ␣-conotoxin MII (0.001–0.3 ␮M) did not exceed 30%. Besides reduction of peak tension d-tubocurarine (0.1–0.7 ␮M), mecamylamine (0.1–300 ␮M), and hexamethonium (30–3,000 ␮M) also caused tetanic fading. With ␣-conotoxin MII (0.001–0.3 ␮M) and dihydro-␤-erythroidine (0.03–10 ␮M), tetanic fade was evident only after decreasing the safety factor of neuromuscular transmission (with high magnesium ions, 6–7 mM). The antagonist rank potency order to reduce evoked (50 Hz for 5 sec) [ 3 H]-ACh release from motor nerve terminals was ␣-conotoxin MII (0.1 ␮M) ⬎dihydro-␤-erythroidine (1 ␮M) ⬃d-tubocurarine (1 ␮M) ⬎mecamylamine (100 ␮M) ⬎hexamethonium (1,000 ␮M). When applied in a concentration (0.3 ␮M) above that producing tetanic paralysis, ␣-bungarotoxin failed to affect [ 3 H]-ACh release. Data obtained suggest that postjunctional neuromuscular relaxants interact with ␣-bungarotoxin-sensitive nicotinic receptors containing ␣1-subunits, whereas blockade of neuronal ␣3␤2-containing receptors produce tetanic fade by breaking nicotinic autofacilitation of acetylcholine release. Synapse 49:77–88, 2003. ©2003 Wiley-Liss, Inc. INTRODUCTION In myographic records, neuromuscular fade, tetanic fade, or Wedensky inhibition is the inability of a muscle to sustain tension during high frequency (30–80 Hz) motor nerve stimulation in the presence of muscle relaxants such as tubocurarine (d-TC) (Bowman, 1980; van der Kloot and Molgo´, 1994). In electrophysiological experiments, rundown of endplate potentials is also observed during high-frequency stimulation trains, an effect resulting from decreases in the quantal output (Matzner et al., 1988) without significant changes in endplate resting potentials (Magleby et al., 1981). Iontophoretic pulses of acetylcholine (ACh) delivered at a frequency of 50 Hz did not produce tetanic fade either in the absence or in the presence of d-TC (Gibb and Marshall, 1986). Moreover, these authors demonstrated that depolarization of muscle fibers by iontophoretically applied ACh could still be observed following fading of nerve-evoked endplate potentials (Gibb and Marshall, 1984). Thus, it was suggested that twitch blockade and tetanic fade are separate and inContract grant sponsor: FCT; Contract grant numbers: POCTI/36545/FCB/ 2000 and UMIB-215/94. *Correspondence to: P. Correia-de-Sa´, MD, PhD, Laborato´rio de Farmacologia, Instituto de Cieˆncias Biome´dicas de Abel Salazar (ICBAS)—Universidade do Porto, L. Prof. Abel Salazar, 2, 4099-003 Porto, Portugal. E-mail: [email protected]. Received 8 October 2002; Accepted 4 March 2003 DOI 10.1002/syn.10211 SYNAPSE 49:77–88 (2003) ©2003 WILEY-LISS, INC. dependent actions; twitch tension depression is linked to the competitive block of postsynaptic nicotinic receptors (nAChRs), whereas endplate rundown or tetanic fade result predominantly from a prejunctional action of nAChRs antagonists (Gibb and Marshall, 1986). ACh may enhance its own release during a period of repetitive motor nerve activity by activating nicotinic autoreceptors. The positive nicotinic feedback mechanism was suggested from studies measuring mechanical tension and electrophysiological signals in the presence of nicotinic antagonists (see for a review, Bowman et al., 1988). In addition, data from radiolabeled experiments showed that the nicotinic agonist 1,1-dimethyl4-phenylpiperazinium markedly increased, while nicotinic antagonists decreased, the release of [ 3 H]-ACh release in response to nerve stimulation (Vizi and Somogyi, 1989; Wessler, 1989; Correia-de-Sa´ and Ribeiro, 1994). Nicotinic autofacilitation is clearly frequencydependent and may operate as a temporary presynaptic amplifier to increase the safety factor for transmission, particularly when increased muscular strength is needed (Waud and Waud, 1971; Singh and Prior, 1998; Wood and Slater, 2001). Nicotinic receptors are a heterogeneous family. Diversity in nAChR types is derived from the variability of genes encoding for receptor subunits and their pentameric subunit combinations (namely, associations between ␣1–␣9 and ␤1–␤4 subunits) (Lukas et al., 1999). Subunits confer distinct structural and functional properties (e.g., calcium permeability, desensitization rate, phosphorylation sensitivity) to the nAChR types that they form. Several lines of evidence indicate that neuronal and muscular nicotinic receptors present at the neuromuscular junction differ in their pharmacological profiles (e.g., Gibb and Marshall, 1984; 1986; Vizi et al., 1987). However, these findings are not yet clearly integrated with molecular definitions of the receptor subtypes that participate in preand postjunctional actions underlying changes of neuromuscular transmission. The muscle endplate nAChR is best characterized and has an (␣1) 2 (␤1)ε␦subunit configuration (Schuetze and Role, 1987; Salpeter et al., 1988). In contrast, the current tally of neuronal nAChR subunits in mammals is eleven (␣2–␣7, ␣9, ␣10, ␤2–␤4) with an additional subunit, ␣8, identified in avian species. The present work was designed to study the role of facilitatory nicotinic autoreceptors block on tetanic fade and to investigate the type of nAChR that might be involved in its operation. For this purpose we tested the effects of several nicotinic antagonists, d-tubocurarine (d-TC), hexamethonium (HEX), mecamylamine (Meca), dihydro-␤-erythroidine (DH-␤-E), methyllycaconitine (MLA), ␣-bungarotoxin (BTX), and ␣-conotoxin MII (CTX MII) on tension responses and [ 3 H]- ACh release triggered by brief high-frequency trains (50 Hz for 5 sec) delivered to the rat phrenic nervehemidiaphragm preparations. Like Meca, HEX was also first recognized as a ganglionic nAChR-blocking agent; these agents exert their effects acting as nicotinic channel blockers and are considered noncompetitive antagonists. DH-␤-E is a competitive neuronal nicotinic antagonist with a degree of selectivity for receptors containing ␣4␤2 and ␣3␤2 subunits (ChavezNoriega et al., 1997). The Delphinium alkaloid, MLA, competitively antagonizes ␣7 nAChRs (K i ⬃1 nM) and, unlike BTX, discriminates between neuronal ␣7 and muscle-type ␣1 receptors. BTX was instrumental in the isolation and purification of muscular ␣1-containing nAChRs, but it is also a highly potent and selective antagonist at ␣7 nAChRs (K 1 ⬃1 nM) without interacting with ␣/␤heteromers (for a review, see Dwoskin and Crooks, 2001). CTX MII, a 16-residue polypeptide from the venom of the piscivorous cone snail Conus magus, has a high degree of selectivity for ␣3␤2-containing neuronal nAChRs (Cartier et al., 1996), although it might also block receptors containing ␣6 subunits with a high affinity (e.g., Kuryatov et al., 2000). MATERIALS AND METHODS Rats (Wistar, 150–200 g) of either sex (Charles River, Barcelona, Spain) were kept at a constant temperature (21°C) and a regular light (06.30–19.30 h) dark (19.30–06.30 h) cycle with food and water ad libitum. The animals were killed by stunning followed by exsanguination. Animal handling and experiments carried out at ICBAS followed the guidelines of the International Council for Laboratory Animal Science (ICLAS). The experiments were carried out in vitro on left phrenic nerve-hemidiaphragm preparations (4–6 mm width). Each muscle was superfused with Tyrode’s solution (pH 7.4) with the following composition (mM): NaCl 137, KCl 2.7, CaCl 2 1.8, MgCl 2 1, NaH 2 PO 4 0.4, NaHCO 3 11.9, glucose 11.2, and choline 0.001, at 37°C. This solution was continuously gassed with a mixture of O 2 (95%) and CO 2 (5%). Nerve stimulation conditions The left phrenic nerve was stimulated with an extracellular glass-platinum suction electrode placed near its first division branch, to avoid direct stimulation of muscle fibers (indirect stimulation). To evaluate drug Abbreviations ACh acetylcholine BTX ␣-bungarotoxin CTX GIIIB ␮-conotoxin GIIIB CTX MII ␣-conotoxin MII DH-␤-E dihydro-␤-erythroidine HEX hexamethonium Meca mecamylamine MLA methyllycaconitine MT-7 muscarinic toxin 7 nAChRs nicotinic acetylcholine receptors d-TC d-tubocurarine 78 M. FARIA ET AL. effects on muscle contractile properties, direct stimulation of muscle fibers was delivered through a pair of platinum electrodes placed at each side of the diaphragm near its costal insertion (field stimulation). Supramaximal intensity (current strength of 8 mA), rectangular pulses of 0.04 ms (indirect stimulation), or 1-ms (field stimulation) duration were used to achieve firing synchronization, thus reducing the number of silent units (motoneurons and/or muscle fibers) that might make interpretation of data difficult. The pulses were delivered by a Grass S48 (Quincy, MA, USA) stimulator coupled to a stimulus isolation unit (Grass SIU5) operating in a constant current mode. The stimulation parameters were continuously monitored on an oscilloscope (Meguro, MO-1251A, Japan) and were within the same range used in previous studies with this preparation (e.g., Wessler and Kilbinger, 1986; Correia-de-Sa´ et al., 2000). Muscle tension recordings When recording tension responses, the innervated diaphragm strips were mounted vertically in a conventional 10-ml capacity isolated organ bath chamber. Tetani (5 sec long) delivered with a frequency of 50 Hz were applied once every 15 min. Directand nerveinduced tetanic responses were recorded isometrically at a resting tension of 50 mN with a force transducer and displayed on a Hugo-Sachs (Germany) recorder. After the initial stabilization period, these experimental conditions allowed a well-preserved tetanic pattern for several hours in the absence of test drugs. Solutions were changed transferring the inlet tube of the peristaltic pump (Gilson, Minipuls3, France) from one flask to another. The flow rate was 20 ml min ⫺1 during the first min after changing the solutions and 5 ml min ⫺1 until the next changeover of solutions. Test drugs were allowed to contact with the preparations at least 12 min before tetanus; incubation time with BTX was prolonged to 45 min in some of the experiments. The tension produced at the beginning of tetanic stimulation (a) was compared with that obtained at the end of tetanic stimulation (b) (cf. Silva et al., 1999). The ratio R(R⫽b/a) obtained after drug addition was taken as a percentage of that observed before any drug administration. Zero percent represents equality between ratios. Positive and negative values represent increment and fading of the tetanic tension, respectively. In order to reduce the safety margin of neuromuscular transmission (Paton and Waud, 1967; Wood and Slater, 2001), MgCl 2 (6–7 mM) was added to the bath in some of the experiments. Osmolarity was maintained by equimolar substitution of NaCl. Elevation of magnesium ions to 6 and 7 mM decreased the amplitude of nerve-evoked tetanic responses by 36 ⫾5% (n⫽5) and 74 ⫾6% (n⫽4), respectively. Isotope experiments The procedures used for labeling the preparations and measuring evoked [ 3 H]-ACh release were previously described (Correia-de-Sa´ et al., 1991) and used with minor modifications. Experiments were performed in the absence of cholinesterase inhibitors to prevent unphysiological extracellular accumulation of ACh. Phrenic nerve-hemidiaphragm preparations were mounted in Perspex chambers of 3 ml capacity through which solutions flowed. After a 30-min equilibration period, the perfusion was stopped and the nerve endings were labeled for 40 min with 1 ␮M[ 3 H]-choline (specific activity 2.5 ␮Ci nmol ⫺1 ) under electrical stimulation at 1 Hz frequency. After the end of the labeling period the preparations were again superfused (15 ml min ⫺1 ) and the nerve stimulation stopped. From this time onwards, hemicholinium-3 (10 ␮M) was present to prevent uptake of choline. After a 60-min period of washout the perfusion was stopped. Bath samples (2 ml) were automatically collected every 3 min by emptying and refilling the organ bath with the solution in use, using a fraction collector (Gilson, FC 203B) coupled to a peristaltic pump (Gilson, Minipuls3) programmed device. Aliquots (0.5 ml) of the incubation medium were added to 3.5 ml of Packard Insta Gel II (Meriden, CT, USA) scintillation cocktail. Tritium content of the samples was measured by liquid scintillation spectrometry (% counting efficiency: 40 ⫾2%) after appropriate background subtraction, which did not exceed 5% of samples tritium content. The radioactivity was expressed as DPM g ⫺1 of wet weight of the tissue determined at the end of the experiment. After the loading and washout periods, the preparation contained 5,542 ⫾248 ⫻10 3 DPM g ⫺1 and the resting release was 132 ⫾12 ⫻10 3 DPM g ⫺1 in 3 min (n⫽8). When the fractional release was calculated, this value proved to be 2.38 ⫾0.14% of the radioactivity present in the tissue at the first collected sample. [ 3 H]-ACh release was evoked stimulating the phrenic nerve with brief high-frequency trains (50 Hz for 5 sec, 40 ␮s pulse width). Two stimulation periods were used: at 12 min (S 1 ) and at 39 min (S 2 ) after the end of washout (zero time). Electrical stimulation of the phrenic nerve increased the release of [ 3 H]-ACh in a Ca 2⫹ - and tetrodotoxin-sensitive manner (Correia-de-Sa´ et al., 2000), while the output of [ 3 H]-choline remained unchanged (Wessler and Kilbinger, 1986), thus indicating that ACh comes mainly from vesicle exocytosis from depolarized nerve terminals. It is unlikely that nonquantal ACh release (Katz and Miledi, 1977) account for the total amount of ACh released upon electrical stimulation of the phrenic nerve. This assumption is based on findings indicating that the spontaneously releasable neuronal pool of ACh is not labeled with [ 3 H]-choline nor is it released by electrical nerve stimulation (Molenaar et al., 1987), and it is completely exhausted (within minutes) in the presence of hemicholinium-3 ␣3␤2 NICOTINIC RECEPTORS ON MOTONEURONS 79 (Nikolsky et al., 1991). Therefore, evoked [ 3 H]-ACh release was calculated by subtracting the basal tritium outflow from the total tritium outflow during each stimulation period (Correia-de-Sa´ et al., 1991). Test drugs were added 15 min before S 2 . In some experiments, incubation time with BTX was prolonged to 45 min and S 2 was delivered at the 69 th min after the end of washout. Drug effects were expressed by the ratios S 2 /S 1 , i.e., the ratio between the evoked [ 3 H]-ACh release during the second stimulation period (in the presence of the test drug) and the evoked [ 3 H]-ACh release during the first stimulation period (without the test drug). Percentage values shown in figures correspond to percentage changes in S 2 /S 1 ratios as compared with the S 2 /S 1 ratio in control experiments (0.83 ⫾0.06, n⫽ 6); zero percent represents identity between ratios. Positive and negative values represent facilitation and inhibition of evoked [ 3 H]-ACh release, respectively. None of the drugs used significantly (P⬎0.05) changed basal tritium outflow. Materials and solutions Chemicals used were: ␣-Bungarotoxin (BTX), choline chloride, dihydro-␤-erythroidine hydrobromide (DH-␤- E), hemicholinium-3, hexamethonium bromide (HEX), mecamylamine (Meca), methyllycaconitine citrate (MLA), pirenzepine dihydrochloride, d-tubocurarine chloride (d-TC) (Sigma, St. Louis, MO, USA); ␣-conotoxin MII (CTX MII) (Tocris Cookson, UK); ␮-conotoxin GIIIB (CTX GIIIB), muscarinic toxin 7 (Peptide Institute, Japan); [methyl3 H]-choline chloride (ethanol solution, 80 Ci mmol ⫺1 ) (Amersham, UK). Aqueous stock solutions were stored as frozen aliquots at ⫺20°C. Dilutions of these stock solutions were made daily and appropriate controls were done. The pH of the superfusion solution did not change by the addition of drugs in the maximum concentrations applied to the preparations. Statistics The data are expressed as mean ⫾SE, from nexperiments. Statistical significance of experimental results was analyzed by one-way analysis of variance (ANOVA) followed by Dunnett’s modified t-test. P⬍ 0.05 was considered to represent significant difference. RESULTS Blockade of muscle-type nicotinic receptors containing ␣1-subunits reduce tetanic tension Changes in the amplitude of tetanic peak tension produced by nicotinic receptor antagonists was taken as a measure of the postjunctional activity of these compounds. Figure 1 shows that BTX (0.003–0.1 ␮M), d-TC (0.1–0.7 ␮M), Meca (0.1–300 ␮M), and HEX (10– 3,000 ␮M) decreased tetanic peak tension in a concentration-dependent manner. Failure to detect nerveevoked diaphragm contractions in response to tetanic trains (50 Hz for 5 sec) was observed with BTX (0.1 ␮M), d-TC (0.7 ␮M), Meca (300 ␮M), and HEX (3,000 ␮M). As BTX exhibits slow binding kinetics and its action is essentially irreversible, we performed experiments where the preincubation time was prolonged from 12 to 45 min. Reduction of tetanic peak tension following a 45-min contact with BTX (0.003–0.1 ␮M) was not statistically different (P⬎0.05) from that obtained using a 12-min incubation period (data not shown). Depression of tetanic peak tension with DH- ␤-E (0.03–10 ␮M), which blocks preferentially ␣4␤2and ␣3␤2-containing receptors, MLA (0.003–3 ␮M), a preferential ␣7-receptor antagonist, and CTX MII (0.001–0.3 ␮M), a selective ␣3␤2-receptor antagonist that also blocks receptors containing the ␣6 subunit, did not exceed 30%. Thus, reduction in tetanic peak tension has an antagonist profile with a rank order of potency of BTX ⬎d-TC ⰇMeca ⬎HEX. This is in agreement with previous studies suggesting that nicotinic receptors localized on skeletal muscle contain ␣1subunits (Schuetze and Role, 1987; Salpeter et al., 1988). nAChR underlying tetanic fade possess a distinct antagonist profile from the muscular receptor type Figure 2A shows pen-recorder traces of nerve-evoked muscle contractions obtained during short high-frequency trains (50 Hz for 5 sec). In control conditions, a Fig. 1. Concentration–response curves for the effects of nicotinic receptor antagonists on the amplitude of nerve-induced tetanic peak tension. Tetanic responses were elicited once every 15 min by stimulating the phrenic nerve with brief high-frequency trains (50 Hz for 5 sec, 40 ␮s pulse width). d-Tubocurarine (d-TC), hexamethonium (HEX), mecamylamine (Meca), dihydro-␤-erythroidine (DH-␤-E), methyllycaconitine (MLA), ␣-bungarotoxin (BTX), and ␣-conotoxin MII (CTX MII) were applied in a cumulative manner and contacted the preparation at least 12 min before recordings. The ordinates are percentage of maximal tetanic peak tension (100%) observed in control conditions (in the absence of nicotinic antagonists). The vertical bars represent ⫾SE of 3–6 experiments (for each curve) and are shown when they exceed the symbols in size. 80 M. FARIA ET AL. brief facilitation (b ⬎a) was evident during the course of tetani (50 Hz for 5 sec), i.e., muscle tension was transiently increased when high-frequency repetitive pulses were delivered to the nerve. Increasing concentrations of d-TC (0.1–0.7 ␮M, Fig. 2A), Meca (0.1–300 ␮M), and HEX (10–3,000 ␮M) caused a very intense fade (b ⬍a) of tetanic contractions (Fig. 3) in parallel with a significant reduction of the maximal tetanic tension (cf. Fig. 1). Neither BTX (0.003–0.1 ␮M) nor MLA (0.003–3 ␮M) produced tetanic fade (Fig. 3), albeit BTX (0.003–0.1 ␮M) strongly depressed the maximal tetanic tension (Figs. 1, 2A). To investigate the possibility of a direct action of the nicotinic antagonists on muscle contractile properties, we studied the effect of d-TC on tetanic tension induced by direct muscle stimulation. When applied in concentrations above those that caused complete neuromuscular block, d-TC (1 and 5 ␮M) and BTX (0.3 ␮M) were virtually devoid of effect on muscle tension induced by tetanic field stimulation (50 Hz for 5 sec) (Fig. 2B). In addition, measurements of tetanic tension and fading may depend on the threshold for activation of muscle action potential, which would be misinterpreted as being due to an indirect presynaptic effect. To evaluate this possibility we compared the depression of contractile responses caused by the nicotinic receptor antagonists with the effect of ␮-conotoxin GIIIB (CTX GIIIB), a sodium channel blocker in skeletal muscle with no action on channels present in axons and motor nerve terminals (Cruz et al., 1985). CTX GIIIB (0.01–0.6 ␮M) decreased tetanic peak tension in a concentration-dependent manner with no obvious fading phenomena being associated (Fig. 2). Muscular paralysis was obtained with CTX GIIIB (0.6 ␮M) when tetanic trains (50 Hz for 5 sec) were delivered either indirectly, to the phrenic nerve trunk, or directly, to the muscle fibers. Data from direct muscle stimulation protocols and the absence of tetanic fading in the presence of CTX GIIIB clearly contrast with the results obtained with d-TC (see Fig. 2), further indicating that a presynaptic action might be involved. Neuronal ␣3␤2 nAChR block reduces [ 3 H]-ACh release triggered by high-frequency trains To address the role of neuromuscular blocking agents on prejunctional nicotinic receptors, we compared their ability to produce tetanic fade and to decrease [ 3 H]-ACh release evoked by high-frequency stimulation trains (50 Hz for 5 sec). Figure 4A illustrates the time course of tritium outflow in experiments where CTX MII (0.1 ␮M), d-TC (1 ␮M), and BTX (0.3 ␮M) were applied 15 min before S 2 . As can be seen from these typical experiments, evoked [ 3 H]-ACh release was decreased in the presence of CTX MII (0.1 Fig. 2. Comparison between the effects of two nicotinic receptor antagonists (d-tubocurarine and ␣-bungarotoxin) and ␮-conotoxin GIIIB, a sodium channel blocker in skeletal muscle, on tetanic muscle tension induced by high-frequency (50 Hz for 5 sec) stimulation trains delivered to the phrenic nerve (indirect stimulation, A) or to diaphragm muscle fibers (direct stimulation, B). A: Typical recording traces of nerve-evoked hemidiaphragm contractions obtained during brief tetanic trains (50 Hz for 5 sec), in the absence (Ctr) and in the presence of d-tubocurarine (d-TC, 0.3 and 0.7 ␮M), ␣-bungarotoxin (BTX, 0.03 and 0.1 ␮M), and ␮-conotoxin GIIIB (CTX GIIIB, 0.3 and 0.6 ␮M). The small horizontal lines indicate the duration of tetanic stimulation (5 sec); vertical calibration: 50 mN. B: Note that d-TC (1 or5␮M) and BTX (0.3 ␮M), applied in concentrations above those required to cause complete neuromuscular block were virtually devoid of effects on the contractile responses induced by direct muscle stimulation. In contrast, CTX GIIIB (0.6 ␮M) abolished contractions caused by direct muscle depolarization. Fig. 3. Effects of nicotinic receptor antagonists on tetanic fading induced by high-frequency (50 Hz for 5 sec) stimulation trains delivered to the phrenic nerve. Tetanic fade was calculated as the ratio (R) between the tensions recorded at the end (b) and at the beginning (a) of the tetanic response (R ⫽b/a) (see Fig. 2). d-Tubocurarine (d-TC), hexamethonium (HEX), mecamylamine (Meca), dihydro-␤-erythroidine (DH-␤-E), methyllycaconitine (MLA), ␣-bungarotoxin (BTX), and ␣-conotoxin MII (CTX MII) were applied in a cumulative manner and contacted the preparation at least 12 min before recordings. On the ordinate, ratio (R) is expressed as a percentage of that obtained in control (Ctr) conditions (in the absence of nicotinic antagonists), taken as 100%. The vertical bars represent ⫾SE of 3–6 experiments (for each curve) and are shown when they exceed the symbols in size. ␣3␤2 NICOTINIC RECEPTORS ON MOTONEURONS 81 ␮M) and d-TC (1 ␮M), but not when BTX (0.3 ␮M) was added. The antagonist rank potency order to inhibit (by about 50–70%) the release of [ 3 H]-ACh (50 Hz for 5 sec) was CTX MII (0.1 ␮M) ⬎DH-␤-E (1 ␮M) ⬃d-TC (1 ␮M) ⬎Meca (100 ␮M) ⬎HEX (1000 ␮M) (Fig. 4B). The lack of BTX effect at the prejunctional level was observed even when the time of incubation was prolonged to 45 min and it was applied in a concentration (0.3 ␮M) above that necessary to cause complete muscular paralysis (data not shown). It is also worth noting that application of BTX did not significantly (P⬎ 0.05) affect the basal tritium outflow (Fig. 4A; but see e.g., Apel et al., 1995). This fully agrees with data showing that failure of BTX (0.003–0.1 ␮M) to induce presynaptic rundown of tetanic contractions can be dissociated from its ability to decrease tetanic peak tension (see above). Thus, the present results suggest that autofacilitation of ACh release from motor nerve terminals is probably mediated by ␣3␤2containing nicotinic receptors sensitive to blockade by CTX MII. Reducing the safety factor of neuromuscular transmission significantly potentiates tetanic fade caused by neuronal nicotinic receptor antagonists In general, the amount of transmitter released per nerve impulse is greater than that required to trigger an action potential in the muscle fiber, although the transmission safety margin may become critical in pathological conditions (e.g., myasthenic syndromes). Because of the high safety factor of neuromuscular transmission, depression of nerve-evoked muscle contractions due to prejunctional acting drugs might not always reflect the magnitude of transmitter release inhibition (see for a review, see Wood and Slater, 2001). This might explain why nicotinic blocking agents like CTX MII (0.1 ␮M) and DH-␤-E (1 ␮M) are more potent to inhibit (⫺56 ⫾9%, n⫽5 and ⫺50 ⫾6%, n⫽4, respectively) evoked [ 3 H]-ACh release (Fig. 4) than to cause depression of tetanic contractions (Figs. 5, 6). Increasing magnesium concentration in the bathing fluid is a useful strategy to decrease the safety factor of Fig. 4. Effects of nicotinic receptor antagonists on [ 3 H]-ACh release from the rat motor nerve endings evoked by high-frequency (50 Hz for 5 sec) trains. A: Time course of tritium outflow from rat phrenic nerve terminals in the absence (Control, F) and in the presence (Œ)of ␣-bungarotoxin (BTX, 0.3 ␮M), d-tubocurarine (d-TC, 1 ␮M), and ␣-conotoxin MII (CTX MII, 0.1 ␮M). After the labeling and washout periods (zero time), [ 3 H]-ACh release was elicited by stimulating the phrenic nerve twice (S 1 at the 12th min and S 2 at the 39th min) at a frequency of 50 Hz for 5 sec. Tritium outflow was measured in samples collected every 3 min. The nicotinic antagonists were applied at least 12 min before S 2 (as indicated by the horizontal bars). Note that spontaneous tritium outflow was not significantly modified in the presence of the drugs. B: Inhibitory effects of ␣-conotoxin MII (CTX MII, 0.1 ␮M), ␣-bungarotoxin (BTX, 0.3 ␮M), dihydro-␤-erythroidine (DH-␤-E, 1 ␮M), d-tubocurarine (d-TC, 1 ␮M), mecamylamine (Meca, 100 ␮M), and hexamethonium (HEX, 1000 ␮M) on evoked [ 3 H]-ACh release from motor nerve terminals determined as described in (A). The ordinates are percentage inhibition of [ 3 H]-ACh release as compared to control, in the absence of added drugs. Each column represents pooled data from 4–11 experiments. The vertical bars represent ⫾SE. *P⬍0.05 (one-way ANOVA followed by Dunnett’s modified t-test) when compared with zero percent. 82 M. FARIA ET AL. synaptic transmission by reducing the amount of transmitter being released per stimulation pulse (del Castillo and Katz, 1954; Paton and Waud, 1967). We reported previously that increments of the magnesium content in the buffer (up to 8.5 mM) decreased the evoked [ 3 H]-ACh release by 53 ⫾9% (n⫽4), without affecting agonist-induced presynaptic nicotinic facilitation (Correia-de-Sa´ and Ribeiro, 1994). To probe for the apparent discrepancy existing between the magnitude of transmitter release inhibition and the fading phenomena, we reevaluated the effects of DH-␤-E (0.03–10 ␮M) and CTX MII (0.001–0.3 ␮M) in conditions where the safety margin of neuromuscular transmission was partially reduced with high magnesium concentrations (6–7 mM). Figure 5 illustrates pen-recorder traces taken from representative experiments with DH-␤-E (3 and 10 ␮M) and CTX MII (0.1 and 0.3 ␮M), where nerve-induced tetanic peak tension was reduced by about 50% using 6–7 mM MgCl 2 . The amplitude of traces obtained with normal (1 mM) and high magnesium concentrations were normalized to facilitate comparison (see calibration bar in the figure). Depressions of tetanic peak tension due to both DH-␤-E (0.03–10 ␮M) and CTX MII (0.001–0.3 ␮M) were significantly (P⬍0.05) potentiated upon increasing magnesium content in the buffer (Fig. 6). A complete neuromuscular block was obtained with 0.3 and 10 ␮M concentrations of DH-␤-E in the presence of 7 and 6 mM MgCl 2 , respectively. It is worth noting that magnesium (6–7 mM) slightly enhanced tetanic facilitation (see R values in Figs. 6B, D). This is a well-known phenomenon seen during high-frequency repetitive nerve stimulation when quantal output is low (see e.g., van der Kloot and Molgo´, 1994). Albeit, preservation of the tetanic ascendant in high magnesium solutions, simultaneous application of DH-␤-E (0.03–10 ␮M) significantly (P⬍0.05) enhanced fading (Fig. 6B). Tension at the end of tetani was virtually abolished when 1 and 10 ␮M DH-␤-E was added in the presence of 7 and 6 mM MgCl 2 , respectively. Under low quantal conditions, CTX MII (0.001– 0.3 ␮M) significantly (P⬍0.05) attenuated tetanic facilitation in a concentration-dependent manner (Fig. 6D). CTX MII (0.3 ␮M, n⫽3) decreased the R ratio (b/a) from 1.50 ⫾0.02 (n⫽3, normal Tyrode buffer) to 1.20 ⫾0.05 (n⫽3) when 7 mM MgCl 2 was added to the incubation media. d-TC (0.7 ␮M), Meca (300 ␮M), and HEX (3,000 ␮M) used in concentrations high enough to decrease the safety factor of neuromuscular transmission induced the rundown of tetanic contractions until complete neuromuscular block (see Fig. 3). These achievements did not require the raising of magnesium concentration, because these agents may simultaneously block preand postjunctional nicotinic sites. Blockade of muscarinic M 1 autoreceptors reduce transmitter release without producing tetanic fade In addition to the short-term nicotinic positive feedback mechanism, ACh may increase its own release by acting at muscarinic M 1 receptors on motor nerve terminals (see e.g., Wessler, 1989; Oliveira et al., 2002). As illustrated in Figure 7A, the muscarinic M 1 receptor antagonist pirenzepine (10 nM) inhibited (26 ⫾5%, n ⫽ 4) the release of [ 3 H]-ACh induced by 50 Hz frequency trains. This inhibitory action was mimicked (37 ⫾6%, n⫽5) by the muscarinic toxin 7 (MT-7, 1 nM, data not shown) isolated from the venom of the green mamba (Dendroaspis angusticeps), which exhibits high subtype selectivity for M 1 (pK B ⬃9.8) receptors (Adem and Karlsson, 1997). Pirenzepine (1–30 nM), applied cumulatively at least 12 min before recordings, decreased tetanic peak tension in a concentration-dependent manner. Depression of tetanic contractions due to pirenzepine (1–30 nM) did not exceed 20% in normal Tyrode solution (MgCl 2 , 1 mM) (Fig. 7B), but it was significantly (P⬍0.05) Fig. 5. Representative recordings of nerve-evoked hemidiaphragm tetani (50 Hz for 5 sec) in the presence of dihydro-␤-erythroidine (DH-␤-E, 3 and 10 ␮M) and ␣-conotoxin MII (CTX MII, 0.1 and 0.3 ␮M): influence of the safety margin of neuromuscular transmission. Recordings in Aand C(normal quantal output) were obtained in normal Tyrode buffer (MgCl 2 , 1 mM), and those in Band D(low quantal output) were obtained in high magnesium conditions (MgCl 2 , 6–7 mM). DH-␤-E (3 and 10 ␮M) and CTX MII (0.1 and 0.3 ␮M) were applied in a cumulative manner for periods of 12 min before each 50 Hz frequency stimulation train. The small horizontal line indicates the duration of tetanic stimulation (5 sec). Please note that the amplitude of traces obtained in normal (Aand C, vertical calibration: 50 mN) and high (Band D, vertical calibration: 25 mN) magnesium conditions were normalized to facilitate comparisons. ␣3␤2 NICOTINIC RECEPTORS ON MOTONEURONS 83 potentiated (46 ⫾7%, n⫽3) when the magnesium content in the buffer was raised to 6 mM (Fig. 7C). In contrast to the findings obtained with several nicotinic channel blockers (e.g., DH-␤-E, CTX MII, d-TC), pirenzepine (30 nM) was virtually devoid of effect on tetanic facilitation even after decreasing the safety factor of neuromuscular transmission. Under these conditions, percent variation of R ratio (b/a) was not higher than 3–7% when compared to controls obtained in normal Tyrode’s solution (1.23 ⫾0.07, n⫽3) or after raising magnesium levels (1.35 ⫾0.06, n⫽3). DISCUSSION In this study we demonstrate that the rat neuromuscular junction is equipped with ␣3␤2-containing neuronal nAChRs mediating facilitation of ACh release, in addition to the classical muscle-type nAChR containing the ␣1 subunit. Subtype-specific nicotinic antagonists (e.g., BTX, DH-␤-E, CTX MII) had distinct profiles to inhibit evoked [ 3 H]-ACh release and to depress tetanic peak tension, clearly indicating that preand postjunctional receptors have different pharmacological properties. Unlike tetanic peak depression due to the “pure” muscular relaxing agent, BTX, fading of tetanic contractions induced by DH-␤-E and CTX MII consisted primarily of the inhibition of ACh release from motor nerve terminals. This provides further support for the hypothesis that tetanic fade is due to an underlying attenuation of nicotinic autofacilitation rather than to a use-dependent block of postjunctional nicotinic receptors (Wilson and Nicholson, 1997). Due to the high safety margin of neuromuscular transmission, repercussions of the fine-tuning nicotinic modulation of transmitter release at the postjunctional level require significant decreases in the synaptic quantal content, like those observed during high-frequency trains or after increasing the magnesium content in the buffer. In contrast to the pivotal role of nAChRs in autonomic neurotransmission and to initiate muscle contraction, neuronal nAChRs are considered to exert a modulatory influence (Wonnacott, 1997). In native neuronal systems, knowledge of the subunit composition of nAChRs is generally lacking and only a few major subtypes have been identified. These include ␣4␤2* nAChR, which is relatively abundant in the CNS (Flores et al., 1996). The other major subtype is comFig. 6. Effects of dihydro-␤-erythroidine (DH-␤-E, 0.03–10 ␮M) and ␣-conotoxin MII (CTX MII, 0.001–0.3 ␮M) on tetanic peak tension and fade (R ⫽b/a) in conditions where the safety factor of neuromuscular transmission was reduced. Tetanic responses were induced once every 15 min by stimulating the phrenic nerve with brief high-frequency trains (50 Hz for 5 sec, 40 ␮s pulse width). DH-␤-E (0.03–10 ␮M, Aand B) and CTX MII (0.001–0.3 ␮M, Cand D) were applied in a cumulative manner and contacted the preparation at least 12 min before recordings. A,C: The ordinates are percentage of maximal tetanic peak tension (100%) determined in the absence of nicotinic antagonists. B,D: The ratio (R) is expressed as a percentage of that obtained in the absence of nicotinic antagonists, taken as 100%. R values for each set of experiments are indicated for comparison. The vertical bars represent ⫾SE of nexperiments and are shown when they exceed the symbols in size. *P⬍0.05 (one-way ANOVA followed by Dunnett’s modified t-test) as compared with the effect of each nicotinic receptor antagonist in normal Tyrode buffer (MgCl 2 1 mM). 84 M. FARIA ET AL. prised of ␣7 subunits, which form homomeric receptors in both central and peripheral nervous systems (Chen and Patrick, 1997). A variety of heteromeric ␣3* nAChRs may also exist in the peripheral nervous system; the predominant subtype contains ␣3 and ␤4 subunits but may also assemble with ␣5 and/or ␤2 subunits (Conroy and Berg, 1995; Flores et al., 1996). Attempts to identify the subunit composition of nAChRs present at motor nerve terminals have been made using nicotinic antagonists lacking subtype selectivity (e.g., Vizi et al., 1995). In this article we show that the antagonist rank potency order to produce neuromuscular tetanic fade and to reduce the release of [ 3 H]-ACh evoked by high-frequency trains was CTX MII ⬎DH-␤-E ⬃d-TC ⬎Meca ⬎HEX. Neither BTX nor MLA caused tetanic fade and/or modified the release of [ 3 H]-ACh, virtually excluding the involvement of neuronal BTX-sensitive receptors (like ␣7*, ␣8*, and ␣9*) in the nicotinic positive feedback mechanism. Others have found that BTX produced a disproportionately higher reduction on the amplitude of miniature endplate potentials when compared to the amplitude of endplate potentials recorded during brief tetanic trains (Domet et al., 1995). However, one cannot exclude the possibility that this transient facilitation of ACh release prior to the establishment of the neuromuscular block could be due to phospholipase A 2 activity (Fathi et al., 2001), which is a known contaminant of several commercially available snake toxins, including BTX (see e.g., Apel et al., 1995). Equipotency between DH- ␤-E and d-TC to inhibit evoked [ 3 H]-ACh release, together with the higher potency of these agents as compared to Meca, make the involvement of ␣4*- and ␣3␤4containing receptors also highly improbable (see e.g., Dwoskin and Crooks, 2001). Since, DH-␤-E is a relatively weak antagonist at ␣3␤4- (ganglionic-like) and (␣1) 2 ε␤1␦- (muscle-type) receptors as compared to the ␣3␤2 subtype (Chavez-Noriega et al., 1997) and evoked [ 3 H]-ACh release was highly sensitive to CTX MII (a preferential ␣3␤2-antagonist at the nanomolar concentration range) (Cartier et al., 1996), our data indicate that facilitatory nAChRs located at the rat motor nerve terminals exhibit a ␣3␤2 subunit configuration. Immunohistochemical studies performed in mouse diaphragms support the view that ␣3-containing nAChRs may exist at the prejunctional level (Tsuneki et al., 1995). The involvement of CTX MII-sensitive neuronal ␣6-heteroceptors (Kuryatov et al., 2000) cannot be excluded from the present data, despite the fact that in the chick retina these receptors are also blocked by nanomolar concentrations of MLA (Vailati et al., 1999). The antagonist profile found in the release experiments clearly differs from that concerning depression of tetanic peak tension (BTX ⬎d-TC ⰇMeca ⬎HEX), which is a well-known phenomenon resulting from the blockade of muscle-type ␣1-containing nAChRs (Schuetze and Role, 1987; Salpeter et al., 1988). It is not surprising that d-TC, Meca, and HEX could simultaneously depress tetanic peak tension, reduce evoked [ 3 H]-ACh release, and induce tetanic fade, as nonselective agents can antagonize cooperatively both neuronal and muscular nAChRs (see e.g., de Oliveira and Oliveira, 1999). In contrast, transmitter release inhibition caused by neuronal nAChR antagonists, like DH- ␤-Eand CTX MII, was associated with a mild fading phenomenon. This apparent discrepancy was attenuated once the safety factor of neuromuscular transmission was decreased by reducing the probability of transmitter release using high magnesium concentrations in the buffer (del Castillo and Katz, 1954; Paton and Waud, 1967; see also Correia-de-Sa´ and Ribeiro, Fig. 7. Effect of the muscarinic M 1 -receptor antagonist, pirenzepine, on transmitter release and muscular tension induced by highfrequency (50 Hz for 5 sec) stimulation trains delivered to the phrenic nerve. A: Time course of tritium outflow from rat phrenic nerve terminals in the absence (Control, F) and in the presence (Œ)of pirenzepine (10 nM), applied 15 min before S 2 (as indicated by the horizontal bar). After the labeling and washout periods (zero time), [ 3 H]-ACh release was elicited by stimulating the phrenic nerve twice (S 1 at the 12th min and S 2 at the 39th min) at a frequency of 50 Hz during 5 sec. Tritium outflow was measured in samples collected every 3 min. B,C: Typical recording traces of nerve-evoked hemidiaphragm contractions observed during brief tetanic trains (50 Hz for 5 sec), in the absence (Ctr) and in the presence of pirenzepine (10 and 30 nM), obtained in normal Tyrode buffer (B, MgCl 2 , 1 mM) and in high magnesium conditions (C, MgCl 2 , 6 mM). Pirenzepine (1–30 nM) was applied in a cumulative manner for periods of 12 min before tetani. The small horizontal line indicates the duration of tetanic stimulation (5 sec). Please note that the amplitude of traces obtained in normal (B, vertical calibration: 50 mN) and high (C, vertical calibration: 25 mN) magnesium conditions were normalized to facilitate comparisons. ␣3␤2 NICOTINIC RECEPTORS ON MOTONEURONS 85