Rat Phrenic Nerve-hemidiaphragm Preparation Research Articles

Interaction of Decamethonium with Hexamethonium or Vecuronium in the Rat

We used isobolographic analysis to investigate the interaction of decamethonium with either hexamethonium or vecuronium in the rat phrenic nerve hemidiaphragm preparation. EC50 values of decamethonium, hexamethonium, and vecuronium were (mean +/- SEM) 47.36 +/- 9.58 microM, 4.27 +/- 0.53 mM, and 5.19 +/- 1.17 microM, respectively. Combinations of drugs in concentrations corresponding to the 1:2, 1:1, and 2:1 ratios of their EC50 values were used to determine three points of each isobole. Decamethonium and hexamethonium showed antagonism: significant deviations from the line of additivity were found at EC50 ratios of 2:1 and 1:1 (P < 0.01 and P < 0.05, respectively) indicating that hexamethonium is a potent antagonist of decamethonium. For decamethonium and vecuronium none of the three points on the isobole was significantly different from the corresponding point on the line of additivity. Hexamethonium is known to be a weak antagonist at the postsynaptic nicotinic acetylcholine receptor but a potent antagonist at the presynaptic nicotinic receptor. Vecuronium is a more potent antagonist at the postsynaptic nicotinic receptor but a much weaker antagonist at the presynaptic site. It was postulated that in the rat the primary site of action of decamethonium is at the presynaptic nerve terminal. Our findings suggest that presynaptic rather than postsynaptic potency of a nondepolarizing drug determines ability to antagonize the effect of a depolarizing drug in the rat.

Antagonism of botulinum toxin-induced muscle weakness by 3,4-diaminopyridine in rat phrenic nerve-hemidiaphragm preparations

The effects of the potassium channel inhibitor and putative botulinum toxin antagonist 3,4-diaminopyridine (3,4-DAP) were investigated in vitro on the contractile properties of rat diaphragm muscle. In the presence of 100 pM botulinum neurotoxin A (BoNT/A), twitches elicited by supramaximal nerve stimulation (0.1 Hz) were reduced to approximately 10% of control in 3 hr at 37 °C. Addition of 3,4-DAP led to a rapid reversal of the BoNT/A-induced depression of twitch tension. In the presence of 100μM 3,4-DAP, antagonism of the BoNT/A-induced blockade began within 30–40 sec and reached 82% of control with a half-time of 6.7 min. The beneficial effect of 3,4-DAP was well maintained and underwent little or no decrement relative to control for at least 8 hr after addition. Application of 1 μM neostigmine 1 hr after 3,4-DAP led to a further potentiation of twitch tension, but this action lasted for < 20 min. Moreover, neostigmine caused tetanic fade during repetitive stimulation. In contrast to the efficacy of the parent compound, the quaternary derivative of 3,4-DAP, 3,4-diamino-1-methyl pyridinium produced little or no twitch potentiation up to a concentration of 1 mM. The potassium channel blocker, tetraethylammonium, generated a transient potentiation followed by a sustained depression of twitch tensions. It is concluded that 3,4-DAP is of benefit in antagonizing the muscle paralysis following exposure to BoNT/A. Co-application of neostigmine or tetraethylammonium with 3,4-DAP, however, appears to confer no additional benefit.

Evidence that the presynaptic A2a-adenosine receptor of the rat motor nerve endings is positively coupled to adenylate cyclase

The action of the A2a-adenosine analogue, CGS 21680C, on electrically evoked [3H]acetylcholine ([3H]-ACh) release, and its interaction with forskolin (an activator of adenylate cyclase), MDL 12,330A (an irreversible inhibitor of adenylate cyclase), rolipram (an inhibitor of cyclic AMP specific phosphodiesterase), dibutyryl- (db-cAMP) and 8-bromo- (8-Br-cAMP) cyclic AMP analogues (substances that mimic intracellular actions of cyclic AMP), were investigated using rat phrenic nerve-hemidiaphragm preparations. CGS 21680C facilitated [3H]-ACh release. Forskolin (but not 1,9-dideoxy forskolin), rolipram, db-cAMP and 8-Br-cAMP also increased evoked neurotransmitter release in a concentration-dependent manner. When the evoked [3H]-ACh release that is dependent on stimulation of the adenylate cyclase/cyclic AMP transduction system was supramaximally stimulated by these compounds, CGS 21680C (3 nmol/l) could not further increase [3H]-ACh release. Phosphodiesterase inhibition with low concentrations (< or = 30 mumol/l) of rolipram significantly potentiated the augmenting effect of CGS 21680C (1 nmol/l) on evoked [3H]-ACh release. MDL 12,330A (an irreversible inhibitor of adenylate cyclase) decreased evoked [3H]-ACh release. The irreversible blocking action of MDL 12,330A on [3H]-ACh release was overcome by by-passing cyclase activation with db-cAMP and 8-Br-cAMP, but could not be overcome with FSK or CGS 21680C. The inhibitory effect of MDL 12,330A on evoked [3H]-ACh release was not mimicked by nifedipine. It is concluded that the increase in [3H]-ACh release caused by CGS 21680C results from activation of an A2a-adenosine receptor positively linked to the adenylate cyclase/cyclic AMP system.

In-vitro interaction between H2 antagonists and vecuronium.

The interaction between histamine H2 antagonists and the neuromuscular blocking drug vecuronium was investigated in the rat phrenic nerve-hemidiaphragm preparation. Cimetidine alone, in the concentration range 800-4000 microM produced between 14 and 74% neuromuscular paralysis with an EC50 (mean +/- s.e.) of 2900 +/- 100 microM. Ranitidine augmented the indirectly-evoked muscle response at concentrations between 30 and 160 microM but at higher concentrations, between 300 and 1800 microM, produced neuromuscular paralysis. Famotidine produced negligible and statistically insignificant (0-5%) neuromuscular paralysis at concentrations between 0.3 and 300 microM. Cimetidine (800 microM) shifted the neuromuscular concentration-effect curve of vecuronium to the left in a parallel manner, while ranitidine (160 microM) shifted it to the right. The potentiation ratio was 1.90 +/- 0.14 for cimetidine and 0.62 +/- 0.05 for ranitidine. Famotidine (30 microM) did not alter the response to vecuronium. These data indicate that higher than clinically relevant concentrations of cimetidine and ranitidine produce neuromuscular paralysis and may potentiate the action of vecuronium. Low concentrations of ranitidine may antagonize the action of vecuronium. Famotidine, in contrast, lacks significant neuromuscular effects.

Potentiation by tonic A2a‐adenosine receptor activation of CGRP‐facilitated [3H]‐ACh release from rat motor nerve endings

1. The effect of calcitonin gene-related peptide (CGRP) on [3H]-acetylcholine ([3H]-ACh) release from motor nerve endings and its interaction with presynaptic facilitatory A2a-adenosine and nicotinic acetylcholine receptors was studied on rat phrenic nerve-hemidiaphragm preparations loaded with [3H]-choline. 2. CGRP (100-400 nM) increased electrically evoked [3H]-ACh release from phrenic nerve endings in a concentration-dependent manner. 3. The magnitude of CGRP excitation increased with the increase of the stimulation pulse duration from 40 microseconds to 1 ms, keeping the frequency, the amplitude and the train length constants. With 1 ms pulses, the evoked [3H]-ACh release was more intense than with 40 microseconds pulse duration. 4. Both the nicotinic acetylcholine receptor agonist, 1,1-dimethyl-4-phenylpiperazinium, and the A2a adenosine receptor agonist, CGS 21680C, increased evoked [3H]-ACh release, but only CGS 21680C potentiated the facilitatory effect of CGRP. This potentiation was prevented by the A2a adenosine receptor antagonist, PD 115,199. 5. Adenosine deaminase prevented the excitatory effect of CGRP (400 nM) on [3H]-ACh release. This effect was reversed by the non-hydrolysable A2a-adenosine receptor agonist, CGS 21680C. 6. The nicotinic antagonist, tubocurarine, did not significantly change, whereas the A2-adenosine receptor antagonist, PD 115,199, blocked the CGRP facilitation. The A1-adenosine receptor antagonist, 1,3-dipropyl-8-cyclopentylxanthine, potentiated the CGRP excitatory effect. 7. The results suggest that the facilitatory effect of CGRP on evoked [3H]-ACh release from rat phrenic motor nerve endings depends on the presence of endogenous adenosine which tonically activates A2a-adenosine receptors. Since both CGRP and A2a-adenosine receptors are positively coupled to the adenylate cyclase/cyclic AMP system, cooperation between these receptors might occur at the second messenger transduction system level.

Contribution of direct actions of the oxime HI-6 in reversing soman-induced muscle weakness in the rat diaphragm

The actions of the bispyridinium oxime HI-6 ([[[(4-aminocarbonyl)pyridino]-methoxy]methyl]-2-[(hydroxyimino)methyl]-pyridinium dichloride) were investigated in vitro on rat phrenic nerve-hemidiaphragm preparations. Isometric twitch and tetanic tensions were elicited at 37°C with supramaximal nerve stimulation at frequencies of 20 and 50 z. To approximate normal respiration patterns, trials consisting of 30 successive 0.55 s trains were alternated with 1.25 s rest periods. Under control conditions, the above stimulation pattern generated tensions that were well maintained at both frequencies. In contrast, a marked depression of muscle tension was observed in diphragms removed from rats administered 339 μg/kg soman (3 LD 50) and tested in vitro. Addition of HI-6, 4 h after soman exposure, led to a nearly complete recovery of muscle tension at 20 Hz. At 50 Hz, muscle tensions still declined especially when trains were elicited at 1.25 and 3 s intervals. The recovery by HI-6 observed in this study appears to be mediated by mechanisms unrelated to acetylcholinesterase reactivation since no increase of enzymatic activity was detected and the effect was reversed by a brief washout in oxime-free physiological solution. The results suggest that the direct action of HI-6 may play a role in restoring soman-induced diaphragmatic failure but this effect woul be significant primarily under low use conditions.

Impaired neuromuscular transmission during partial inhibition of acetylcholinesterase: the role of stimulus-induced antidromic backfiring in the generation of the decrement-increment phenomenon.

Neuromuscular transmission was studied in the rat phrenic nerve-hemidiaphragm preparation with acetylcholinesterase (AChE) partially inactivated. Enzyme inhibition resulted in (1) increased single-twitch tension of the diaphragm; (2) compound muscle action potential (CMAP) containing repetitive discharges; (3) stimulus-induced antidromic backfiring (SIAB) seen in the phrenic nerve; and (4) repetitive nerve stimulation (RNS) eliciting a decrement-increment (D-I) phenomenon (i.e., amplitude reduction maximal with the second CMAP). Using a high-calcium and low-magnesium solution, SIAB and the decrement of the second CMAP during RNS were intensified, whereas closely spaced trains and (+)-tubocurarine (TC) abolished SIAB and simultaneously prevented the decrement of the second CMAP. Importantly, low concentrations of (+)-TC prevented SIAB in the phrenic nerve, while the repetitive discharges of the CMAP and the increase in twitch tension remained unaffected. This observation suggests that preterminal nicotinic receptors stimulated by released acetylcholine induce SIAB, whereas postsynaptic events are less important in the generation of SIAB. SIAB, a presynaptic event, appears to be responsible for the transient impairment of the neuromuscular transmission, i.e., the D-I phenomenon.

INTERACTION BETWEEN BETAMETHASONE AND VECURONIUM

A possible interaction between betamethasone and vecuronium was examined in 20 rat phrenic nerve-hemidiaphragm preparations. Ten preparations were bathed in a physiological solution with betamethasone 1 mumol litre-1 added and, after a 30-min period were exposed to vecuronium at concentrations of 4, 6, 8 and 10 mumol litre-1 with vecuronium free washings between each exposure. Ten control experiments were performed also using a betamethasone-free bathing solution. In comparison with control, the betamethasone group had significantly (P = 0.0008) less depression of muscle contraction (twitch) force at all concentrations of vecuronium. The calculated ED50 (50% depression of muscle contraction force) was 5.65 mumol litre-1 for controls and 7.39 mumol litre-1 for betamethasone-pretreated preparations. This study confirms our previous clinical observations that an interaction occurs between vecuronium and betamethasone which is characterized by resistance to neuromuscular block.

Backfiring of the isolated rat phrenic nerve does not collide with impulse propagation following repetitive nerve stimulation at 1?50 Hz

Acetylcholinesterase inhibition with neostigmine in the isolated rat phrenic nerve-hemidiaphragm preparation induced axonal backfiring and repetitive compound muscle action potentials following single nerve stimulation. The duration of backfiring and the repetitive compound muscle action potentials did not exceed 55 ms. With repetitive nerve stimulation at frequencies ranging from 1 to 50 Hz, backfiring was present only with the first stimulus and the amplitude of the second compound muscle action potential was maximally reduced, while the subsequent responses recovered gradually. However, the amplitudes of the concomitant antidromic nerve action potentials remained unchanged during the entire train of stimulation. Lack of nerve action potential amplitude changes and the short duration of backfiring of the first nerve action potential exclude a collision phenomenon of backfiring with the nerve action potential induced by the second stimulus. Moreover, the duration of the repetitive compound muscle action potentials did not exceed the duration of backfiring. Therefore, the prolongation of the muscle membrane refractory period by reexcitation following backfiring cannot explain the decrement of the second compound muscle action potential.

Interaction of sevoflurane, isoflurane, enflurane and halothane with non-depolarizing muscle relaxants and their prejunctional effects at the neuromuscular junction.

The interaction of four inhalational anesthetics (sevoflurane, isoflurane, enflurane and halothane) with pancuronium and vecuronium and also their prejunctional actions at the neuromuscular junction were quantitatively studied using rat phrenic nerve-hemidiaphragm preparations. To investigate the prejunctional effects of inhalational anesthetics, a train-of-four ratio (T4/T1) and the tetanus ratio (the ratio of the final response to the initial response during tetanus) were evaluated. All four inhalational anesthetics markedly potentiated the neuromuscular blockade of twitch response caused by either pancuronium or vecuronium with halothane and enflurane being the most potent both on a % concentration basis and on a MAC (minimum alveolar concentration) basis. Although none of the four inhalational anesthetics had any effects on the T4/T1 ratio, they produced variable effects on the tetanus ratio. Sevoflurane had little effect on the tetanus ratio, whereas 1 and 2% isoflurane and 1, 2 and 3% enflurane increased the tetanus ratio and 5% halothane and 5% enflurane significantly reduced the tetanus ratio. Halothane and enflurane had the most potent depressant action of the four inhalational anesthetics both on the % concentration basis and on the MAC basis. These results indicate that the main site of action of inhalational anesthetics is a postjunctional site at the neuromuscular junction and that they do not seem to act on prejunctional sites at the concentrations used in clinical situations.

Neuromuscular Effects of Respiratory and Metabolic Acid-Base Changes In Vitro With and Without Nondepolarizing Muscle Relaxants

The effects of metabolic (bicarbonate, [HCO3]) and respiratory (carbon dioxide, PCO2) acid-base changes on indirectly elicited twitch tension with and without nondepolarizing neuromuscular blocking agents were compared in a rat phrenic nerve-hemidiaphragm preparation. Ionized calcium [Ca2+] and magnesium [Mg2+] concentrations in the modified Krebs' solution were measured and kept constant. Likewise, twitch was altered when pH changes were produced by altering either PCO2 or [HCO3]. Decreasing pH either by increasing PCO2 or by decreasing [HCO3] significantly decreased (P less than 0.01) twitch, by 9.5 +/- 0.6 (SEM, n = 8) and 10.6 +/- 1.5%, respectively. Increasing pH by decreasing PCO2 or by increasing [HCO3] significantly increased (P less than 0.01) twitch, by 5.6 +/- 0.9 and 7.9 +/- 0.6%, respectively. After a partial depression of twitch by nondepolarizing neuromuscular blocking agents, the effects of PCO2 and [HCO3] changes were again assessed. Decreasing pH by increasing PCO2 or by decreasing [HCO3] intensified d-tubocurarine (dTc) (28.2 +/- 1.6 and 32.0 +/- 2.9%, respectively) and vecuronium (23.0 +/- 1.4 and 36.8 +/- 3.2%, respectively) block, whereas it reversed metocurine (1.2 +/- 2.2% NS and 2.9 +/- 1.3%, respectively) and pancuronium (8.3 +/- 1.5 and 11.5 +/- 3.0%, respectively) block. Conversely, increasing pH by decreasing PCO2 or by increasing [HCO3] antagonised dTc (12.8 +/- 2.2 and 13.6 +/- 1.8%, respectively) and vecuronium (25.3 +/- 1.7 and 25.0 +/- 3.0%, respectively) block, whereas it potentiated metocurine (4.2 +/- 0.6 and 8.0 +/- 1.1%, respectively) and pancuronium (11.0 +/- 1.2 and 17.5 +/- 2.0%, respectively) block. Except where indicated, all changes in block described above were statistically significant.(ABSTRACT TRUNCATED AT 250 WORDS)

Release of Calcitonin Gene‐Related Peptide‐Like Immunoreactive Substance from Neuromuscular Junction by Nerve Excitation and Its Action on Striated Muscle

In a rat phrenic nerve-hemidiaphragm preparation, calcitonin gene-related peptide (CGRP) increased the twitch contraction induced by nerve or transmural stimulation dose dependently. Either electrical or high K+ stimulation of the phrenic nerve caused release of a CGRP-like immunoreactive substance (CGRP-LIS) in a Ca2(+)-dependent manner. Electrical stimulation of the phrenic nerve also increased the cyclic AMP content in diaphragm. This increase was not observed in Ca2(+)-free medium and was blocked by antiserum against CGRP. These results indicate that excitation of the motor nerve causes release of CGRP-LIS at nerve terminals and that the released CGRP-LIS increases the cyclic AMP content of skeletal muscles and potentiates twitch contraction.

Comparison of the Effects of Calcium and the Calcium Channel Stimulant Bay k 8644 on Neomycin‐Induced Neuromuscular Blockade

The effects of calcium and the calcium channel stimulant Bay k 8644 on neomycin-induced neuromuscular blockade were evaluated using rat phrenic nerve-hemidiaphragm preparations in vitro. neomycin showed maximum potency in inducing neuromuscular blockade when the calcium concentration in the bath was 1 mM. Higher calcium concentrations (1.5, 2 and 4 mM) produced a gradual decrease in neomycin potency, manifested as a progressive shift to the right of neomycin concentration-response curves. Bay k 8644 (0.1 and 1 microM) did not significantly modify indirectly elicited diaphragm contractions per se, nor did it antagonize neomycin-induced neuromuscular blockade. A higher concentration of Bay k 8644 (10 microM) antagonized indirectly elicited contractions. In conclusion, our results show that Bay k 8644, at concentrations causing calcium channel stimulation at the cardiac and vascular level, did not exert stimulant effects in a nerve-skeletal muscle preparation, suggesting that its action may be tissue-selective to a certain degree.

Pertussis toxin prevents the inhibitory effect of adenosine and unmasks adenosine-induced excitation of mammalian motor nerve endings.

Pertussis toxin (PTX), which blocks certain classes of guanine nucleotide binding proteins (G proteins), consistently blocked the inhibitory effects of adenosine (100 microM-250 microM) on quantal acetylcholine (ACh) secretion in rat phrenic nerve hemidiaphragm preparations. PTX pretreatment also highlighted long-lasting increases in evoked ACh release elicited by adenosine. The results suggest that specific G proteins are involved in mediating the inhibitory effects of adenosine at motor nerve endings.

DIFFERING INTERACTIONS BETWEEN HEXAMETHONIUM AND TUBOCURARINE, PANCURONIUM OR ALCURONIUM AT THE NEUROMUSCULAR JUNCTION

The action of hexamethonium has been investigated in the rat phrenic nerve-hemidiaphragm preparation, alone and in combination with the neuromuscular blocking agents tubocurarine, pancuronium and alcuronium. Hexamethonium alone in concentrations between 3.55 x 10(-3) and 7.1 x 10(-3) mol litre-1 produced neuromuscular blockade in a dose-dependent manner. Low concentrations of hexamethonium antagonized the neuromuscular blocking effect of all three neuromuscular blocking drugs, less with tubocurarine than with the other two. Increasing the concentration of hexamethonium produced potentiation of the neuromuscular blocking effect, this being greater with tubocurarine than with either pancuronium or alcuronium. The cholinesterase activity in rat diaphragm homogenates was inhibited by hexamethonium. This inhibition was only significant at concentrations greater than those which resulted in antagonism and cannot, therefore, explain the observed antagonism. The mechanism of these observations is discussed with respect to the known behaviour of a combination of antagonists acting within a receptor system.

The influence of respiratory-induced acid-base changes on the action of non-depolarizing muscle relaxants in rats.

The influence of respiratory-induced acid-base changes on the action of non-depolarizing muscle relaxants was investigated using the rat phrenic nerve-hemidiaphragm preparation. Changes in pH were induced by changes in the CO2 concentration aerating Krebs' solution. In the absence of muscle relaxants, an increase in CO2 from 5% to 7.5% decreased (P less than 0.01) indirectly elicited twitch tension by 5.4 +/- 0.7% (mean +/- SEM), while a decrease in CO2 from 5% to 2.5% increased (P less than 0.01) twitch tension by 2.3 +/- 0.7%. With a change in CO2 from 2.5% to 7.5%, partial neuromuscular blockade produced by d-Tc or vecuronium was augmented (P less than 0.01), while that produced by metocurine, pancuronium, or alcuronium was reduced (P less than 0.01). With the change in CO2 from 7.5% to 2.5%, the neuromuscular blockade produced by d-Tc or vecuronium was reduced (P less than 0.01), while that produced by metocurine, pancuronium, or alcuronium was augmented (P less than 0.01). Dose-response study showed that 2.5% CO2 shifted the dose-response curves for d-Tc and vecuronium to the right (P less than 0.01) from those with 5% CO2, whereas 7.5% CO2 shifted them to the left (P less than 0.05). In contrast, neither 2.5% CO2 or 7.5% CO2 significantly shifted the dose-response curves for metocurine or pancuronium from those with 5% CO2. Their dose-response curves with 2.5% CO2 were to the left, instead of to the right, of those with 7.5% CO2 (P less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)

Interactions of Vecuronium and Atracurium in an In Vitro Nerve-Muscle Preparation

Atracurium and vecuronium were compared when given alone and in combination in the in vitro rat phrenic nerve-hemidiaphragm preparation stimulated via the phrenic nerve. The slopes of the log dose response curves of atracurium and vecuronium were parallel; their ED50s were 1.12 +/- 0.0035.10(-5)M and 5.89 +/- 0.16.10(-6)M, respectively. The combination's log dose response curves were significantly shifted to the left when compared with those of either relaxant alone; an increased potency is displayed by the combination. These observations indicate nondepolarizing muscle relaxant synergy for the combination of equal proportions of vecuronium and atracurium. The synergistic interaction of vecuronium and atracurium in this in vitro-model is not dependent on pharmacokinetic factors such as uptake, distribution, and biodegradation as are present in the in vivo animal models and in humans. Synergy of vecuronium and atracurium in vitro is a new finding and is consistent with hypotheses of multiple receptor sites and different modes of action of the "competitive" neuromuscular blocking agents. This degree of synergy, seen in the in vitro animal data, if extrapolatable to humans, is probably of little clinical significance.

Anti-curare effect of plasma from patients with thermal injury

Following severe thermal injury, patients are resistant to non-depolarizing muscle relaxants. Although this resistance has been well documented clinically, little is known about its etiology. We have tested the hypothesis that circulating factors contribute to the decreased potency of neuromuscular blockers following burns. The potencies of d-tubocurarine (2 μM) or pancuronium (2 μM) dissolved in plasma from either burned or control human subjects were tested on the indirectly stimulated (0.2 Hz) rat phrenic nerve-hemidiaphragm preparation. The muscle relaxants produced less neuromuscular blockade when dissolved in plasma from burned patients than when they were dissolved in plasma from controls. Thus, circulating factors are involved in the decreased potency of non-depolarizing neuromuscular blocking drugs.

EFFECT OF VERAPAMIL ON Ca2³ INFLUX TO RAT PHRENIC NERVE-DIAPHRAGM PREPARATION

The effect of verapamil on contractility and on influx of calcium (Ca2+) to the rat phrenic nerve-hemidiaphragm preparation was studied in vitro, at rest and during electrical stimulation at 0.2-100 Hz. Calcium content of the bathing solution was measured using the Technicon method (SMAC, Sequential Multiple Analyser/Computer). The results showed that verapamil (a) reduced the twitch and tetanic contractions in a non-competitive manner, and (b) reduced the influx of Ca2+ to the rat hemidiaphragm preparation. Greater reductions occurred as the rate of stimulation increased.

Differential release of [ 3H]acetylcholine from the rat phrenic nerve-hemidiaphragm preparation by electrical nerve stimulation and by high potassium

Neuronal transmitter stores of the phrenic nerve were labelled under different conditions. Subsequently, transmitter release evoked by electrical nerve stimulation and by a high potassium-low sodium solution was studied. Incubation of the end-plate preparation with [ 3H]choline at rest led to the synthesis of [ 3H]acetylcholine which could not be released by electrical nerve stimulation but it was released by high potassium-low sodium solution, independent of the presence of extracellular calcium. When the end-plate preparation was labelled during stimulation at 1 Hz, prolonged periods of electrical nerve stimulation released 83% of the total releasable [ 3H]transmitter pool in a completely calcium-dependent manner. After exhaustion of the electrically releasable pool, high potassium-low sodium solution still caused a significant outflow. Without a preceding exhaustion of the [ 3H]acetylcholine pool, high potassium-low sodium solution released a similar amount in the absence of extracellular calcium or after pretreatment with the intracellular calcium chelating substance, Quin-2. When evoked transmitter release was studied at different temperatures (36, 26 and 16°C) Q 10 values of 1.6 and 1.0 were found for the release caused by electrical nerve stimulation and high potassium-low sodium solution (calcium-independent effect), respectively. After labelling during a short interval (2 min) but at a high stimulation rate (50 Hz), only 72% of the releasable [ 3H]transmitter could be released by electrical nerve stimulation, whereas the outflow due to the calcium-independent effect of high potassium-low sodium solution increased from 17 (labelling during stimulation at 1 Hz) to 28%. It is suggested that the calcium-independent effect of high potassium-low sodium solution reflects the release of acetylcholine from the cytoplasmic compartment, as this outflow occurred after labelling at rest and increased when cytoplasmic synthesis was enhanced by a high loading stimulation. In contrast to high potassium-low sodium solution, propagated nerve activity cannot release acetylcholine synthesized at rest (presumed to be cytoplasmic), but only [ 3H]acetylcholine synthesized during quantal release (presumed to be vesicular). The absolute requirement of extracellular calcium for electrically stimulated release suggests an exocytotic release mechanism. The low Q 10 value of 1.6 does not fit into the concept of a carrier- or channel-operated release mechanism. High potassium-low sodium solution triggers both calcium-dependent release (exocytosis) and calcium-independent liberation of acetylcholine, which escapes from the cytoplasmic compartment by a diffusion-like mechanism. Accordingly, it is important to consider that propagated nerve activity and high potassium-low sodium solution can cause the release of transmitter by different mechanisms from different compartments.