Introduction

There are at least three main reasons for the increased interest in catecholamine action on synaptic transmission in the neuromuscular junction over the last decade. First, there have been many studies confirming the effectiveness of adrenergic compounds in treating a wide range of neurodegenerative diseases. Well-known drugs used for cardiovascular and pulmonary pathologies started being offered for the treatment of congenital myasthenic syndromes (Legay 2018; Engel et al. 2015), anti-MuSK myasthenia gravis (Burke et al. 2013; Ghazanfari et al. 2014), and amyotrophic lateral sclerosis (Bartus et al. 2016). Second, direct evidence of tight contact of sympathetic axons with neuromuscular junctions has been obtained. Rudolf et al. (2013), Khan et al. (2016), Straka et al. (2018) and Rodrigues et al. (2019a) have shown, using modern immunofluorescence methods, that sympathetic innervation controls homeostasis of neuromuscular junctions. The third reason is the contradictory data and unclear mechanisms of catecholamine action on synaptic transmission (for review see Tsentsevitsky et al. 2019a).

Differences have been revealed in the effects of adrenaline (AD) and noradrenaline (NA) on both spontaneous quantal secretion and the number of quanta released in response to a nerve stimulus (quantal content). The variability of the observed effects of catecholamines may be due to different types of muscles (e.g., fast and slow, fatigued and intact), different species of animals, various experimental conditions (inhibition of acetylcholinesterase, various methods of blocking muscle contractions, preliminary nerve ending polarization). Joassard et al. (2013) established that skeletal muscle hypertrophy induced by adrenomimetics is associated with changes in the composition of skeletal muscle fibres. A decrease in the proportion of slowly contracting fibres and a corresponding increase in the proportion of fast fibres in rat skeletal muscles in response to the action of adrenomimetics were described. Such a switch in muscle fibres is particularly pronounced in a typical slow muscle of the hind limbs, the soleus muscle (Soić-Vranić et al. 2005). Recently, we described the features of the action of NA and AD in diaphragm synapses, consisting of fast and slow fibres (Tsentsevitsky et al. 2019b). Therefore, it was interesting to evaluate how exogenous adrenomimetics affect the synapses of a muscle containing fibres of the slow type, which M. soleus is.

Here, we investigated AD and NA action on the parameters of acetylcholine (ACh) quantal release—frequency of spontaneous miniature endplate potentials (F MEPPs), quantal content of the evoked endplate potentials (QC EPPs) and degree of synchrony of the release evoked by a single action potential at the neuromuscular junction in the rat slow soleus muscle.

Materials and Methods

Experimental Animals

Experiments were performed on isolated nerve-muscle preparation M. soleus of laboratory rats (10–12 months) of the Wistar strain of both sexes weighing 200–250 g. A Ringer solution of the following composition was used (in mM): NaCl, 120.0; KCl, 5.0; CaCl2, 0.4; NaHCO3, 11.0; NaHPO4, 1.0; MgCl2, 5.0; glucose, 11; pH 7.2–7.4; and temperature of 20 ± 0.3 °C (Bukharaeva et al. 2007). The solution was bubbled with 95% O2 and 5% CO2. The solution flowed through the muscle chamber continuously at the rate of 5 ml/min.

Electrophysiology

The motor nerve was stimulated with suprathreshold stimuli at a frequency of 0.5 Hz. The nerve ending action currents and endplate potentials (EPPs) were simultaneously acquired in the endplate region. Extracellular pipettes (tip diameter of 2–3 μm, input resistance of 1–3 MΩ) were positioned in the endplate region under microscope control. Spontaneous miniature EPPs (MEPPs) were recorded during the interstimuli intervals, and the averaged MEPP frequency was estimated. The ratio n/N was used as the estimate of the EPP quantal content (QC), where n is the number of successive EPPs, and N is the number of applied stimuli (Katz and Miledi 1965). The experiment lasted 1.5–2 h. A time control was carried out, which showed that the parameters of the signals during this time did not significantly change. Synaptic signals were acquired in control conditions and 20 min after the application of the drug-containing solution. The acquired signals were filtered between 0.03 Hz and 10 kHz, digitized at 3-μs intervals by a 9-bit analogue–digital converter, fed into the computer and processed by original homemade software. The methods for extracellular recordings and data acquisition and processing have been previously described in detail (Bukcharaeva et al. 1999; Bukharaeva et al. 2007; Tsentsevitsky et al. 2018). Synaptic delay was measured as the time interval between the peak of the inward presynaptic Na+ current and the time at which the rising phase of the quantal event reached 20% of its maximum (Katz and Miledi 1965; Bukcharaeva et al. 1999). Measured time parameters are shown in Fig. 2b. The limit of the synaptic delay measurement was set to 50 ms. The stability of the recording electrode position was crucial during long-term extracellular recordings. We therefore monitored the amplitudes of the nerve terminal action potentials throughout each dataset. Only experiments in which the terminal action potential changed by less than 10% during the drug application and washout were analysed (Bukcharaeva et al. 1999; Bukharaeva et al. 2007). Because the distribution of delays was not normal, the average value of synaptic delays could not be used. Histograms of the delays’ distribution were constructed. To ensure that the number of recorded uni-quantal EPPs in the 10-ms period did not depend on the variability in different preparators, the number of collected signals was normalized to an equal number of stimulations (1000 stimuli).

Drugs

Solutions of 10 μM NA ((±)-norepinephrine (+)-bitartrate salt, Sigma–Aldrich, USA) and 10 μM L-adrenaline TCI Chemicals (Tokyo, Japan) were prepared immediately before the experiment. We conducted experiments under darkened conditions and did not observe a colour change in the adrenaline and noradrenaline solutions due to oxidation.

Statistical Analysis

The data are presented as the means ± SEM. The significance of the mean value differences was assessed by Student’s t test and the Wilcoxon signed-rank test for matched samples. The results were considered significantly different at p < 0.05, where n corresponds to the number of animals.

Results

Catecholamines can undergo autoxidation with the formation of coloured products that have a maximum absorption at 347 and 480 nm (Misra and Fridovich 1972; Palop et al. 2002). The oxidation products of catecholamines (in particular, adrenochrome and adrenolutin) exhibit fluorescence in the yellow-green region of the spectrum. We conducted a special study and compared the absorption and fluorescence spectra of oxidized products in NA solution (10 μM) and AD solution (10 μM) in control and in a perfusion solution that flowed for 20 min through an experimental bath with a neuromuscular preparation using a Lambda-25 spectrophotometer (Perkin Elmer, USA) and Fluorat-02-Panorama spectrofluorometer (Russia). The absorption and fluorescence spectra (λab 510 nm/λem 550 nm) of the perfusion solution of NA (0.0071 ± 0.0005) and AD (0.0096 ± 0.0012) did not change compared to the control (0.0061 ± 0.0008 and 0.0179 ± 0.0022, respectively), which indicates the absence of significant oxidation of NA and AD molecules under the conditions of our experiments.

The amplitude and time characteristics of the responses did not significantly change with the application of the catecholamines (Table 1). This indicated that there were no effects on the sensitivity of the postsynaptic membrane to ACh.

Table 1 Parameters of spontaneous and evoked endplate potentials after NA and AD application
Full size table

Parameters of the quantal secretion (average frequency of MEPP and EPP QC) in the control conditions and after application of 10 µM NA or AD are presented in Table 2. The changes in the average MEPP frequency with NA and AD administration are presented in Fig. 1a, b. The frequency of spontaneous ACh quantal release did not change with NA, but it was 40% higher in the presence of AD (Fig. 1d). This effect was eliminated when AD was removed.

Table 2 MEPP frequency and EPP quantal content after NA and AD application
Full size table
Fig. 1
figure 1

Changes in spontaneous quantal ACh release after application of NA or AD at a concentration of 10 µM. a Trace of spontaneous ACh release (MEPPs) in Ringer solution after NA and AD application. b Average MEPP frequencies in control conditions (black columns), with application of NA or AD (grey columns), and after removal of the catecholamines (white columns). *p < 0.05 from control value

Full size image

NA did not affect the EPP QC, but AD increased the EPP QC by 62% (Fig. 2c, Table 2). Obtained data indicated the presynaptic site of AD action, which changed the process of ACh quantal release.

Fig. 2
figure 2

Action of catecholamines on the evoked quantal ACh release. a Superposition of selected uni-quantal EPPs in control conditions and after AD application. b Evoked EPP, measured in the experiment: AP—nerve action potential, syn. delay—time interval between peak of AP and the time at which the rise phase of the quanta event reached 20% of its maximum, latency—time interval between the stimulus and peak of AP. c Average quantal content in control conditions (black columns), after catecholamine application (grey columns), and after catecholamine removal (white columns). d Average number of EPPs released with synaptic latencies not exceeding 3 ms—the period of early synchronous release in control conditions (black columns), after catecholamine application (grey columns), and after catecholamine removal (white columns). e Histograms of the synaptic latency distribution in control conditions (upper) and after AD treatment (lower), bin 10 ms

Full size image

After motor nerve action potentials, transmitter quanta are released with variable delays. The delay duration reflects the rates of depolarization—release coupling (Katz and Miledi 1965; Lin and Faber 2002), but their dispersion indicates that evoked phasic quantal release is not synchronous (Bukharaeva et al. 2007; Kaeser and Regehr 2014).

In low Ca2+/high Mg2+ external solution conditions, in the synapse of M. soleus as well as in the diaphragm (Bukharaeva et al. 2007; Tsentsevitsky et al. 2018), ACh quanta are released with varying synaptic delays after nerve action potential arrival, i.e., asynchronously (Fig. 2a). After the addition of NA, the number of quanta released during 10 ms with the 1000 stimuli did not significantly change (Fig. 2d), whereas AD application led to a decrease in the dispersion of synaptic delays and an increase in the number of responses with short delays, i.e., there were more responses with short delays in the AD than in the control conditions in the histogram of the delay distribution. That is, the secretion became more synchronous (Fig. 2e).

Discussion

In the neuromuscular junction of the rat slow M. soleus, we observed the absence of effects of NA at a concentration of 10 µM on parameters of synaptic transmission and the increased spontaneous and evoked quantal secretion in the presence of AD at the same concentration.

Catecholamines can facilitate muscle contraction—this fact was discovered more than a hundred years ago (Oliver and Schäfer 1895). However, the data regarding targeting and the mechanisms of their actions on neuromuscular transmission have been very controversial. Kuba (1970) indicated the facilitation of the action of noradrenaline on the frequency of MEPPs. However, this effect was observed under 20 µM noradrenaline (2 times higher than the concentration we used), and M. diaphragm was used. However, it was shown previously that the effects of adrenergic compounds on the synapses of fast and slow types of muscles differ. The activation of β receptors increased the evoked contractions of fast muscles and reduced the contractions of slow muscles (Bowman and Raper 1962, 1967). The magnitude and nature of the adrenergic effects vary depending on the fibre-type composition of the muscle (Juel 1988; Cairns and Dulhunty 1993; Decorte et al. 2015; Cairns and Borrani 2015). Krnjevic and Miledi (1958) wrote about the effects produced by adrenaline upon neuromuscular propagation in rats: “We have found the actions of adrenaline and noradrenaline to be to some extent unpredictable”. In addition, the effects of catecholamines on the probability of acetylcholine quanta release from nerve endings depend on the membrane potential of terminals, the concentration of extracellular calcium and the content of sodium or magnesium ions in extracellular solution (Kuba and Tomita 1971; Anderson and Harvey 1988; Ginsborg and Hirst 1971). Therefore, the differences in the experimental conditions and types of neuromuscular junctions may be the reason for the controversial data.

Previously, we found that 10 µM NA had no effect on QC of ACh release evoked by nerve stimulation at both physiological and reduced concentrations of external Ca2+ in frog and mouse diaphragm neuromuscular junctions (Bukcharaeva et al. 1999; Tsentsevitsky et al. 2018), but NA changed the degree of synchronous quantal release. In this study, we observed that NA was ineffective in the M. soleus synapse, whilst AD increased the mean EPP QC in response to nerve impulses and evaluated spontaneous secretion. Another important characteristic of the neurosecretion process is the time course of quantal secretion (Katz and Miledi 1965; Barrett and Stevens 1972; Bukharaeva and Nikolskii 2012). The release of several dozen quanta of ACh in response to a single nerve impulse does not occur simultaneously. There is asynchrony during the secretion of individual quanta, manifested by the dispersion of real synaptic latencies of uni-quantal responses recorded under low Ca2+/high Mg2+ conditions (Katz and Miledi 1965; Barrett and Stevens 1972). Analysis of the distribution of delays of individual quantal events provides insights into the intraterminal processes determining the time course of release probability (Schneggenburger and Neher 2000; Huang and Moser 2018). Our data indicated that AD shortened the release period for evoked quantal release and that the response became increasingly synchronized. As shown in our investigation of frog synapses, better synchronization of release significantly increased the size of reconstructed multi-quantal EPCs (Bukcharaeva et al. 1999). This suggested that AD facilitated synaptic transmission by making the release of quanta more synchronous.

Questions about the mechanism of AD action remain open. It must be taken into account that there are two types of α and β adrenoreceptors, including 6 subtypes, which have different sensitivities to NA and AD and are associated with different membrane and intracellular systems (potential-dependent calcium channels, potassium channels, adenylate cyclase and cAMP, phospholipase, inositol triphosphate complex) (Bylund 2007; Abdullahi et al. 2019). Previously, we showed the synchronizing effect of NA in frog synapses occurred through β1 adrenoreceptors (Bukcharaeva et al. 1999). In the neuromuscular synapse of the mouse diaphragm, the opposite effect of desynchronization of NA secretion was observed, and both α and β receptors were involved in this action (Tsentsevitsky et al. 2018, 2019b). Thus, to identify the subtype of adrenergic receptors involved in the implementation of the AD effect, a full pharmacological analysis will be carried out using specific agonists and blockers for each receptor subtype. Nevertheless, we have previously shown that the degree of synchronization of secretion in the mouse neuromuscular synapse depended on the entry of calcium ions into the nerve ending. When Ca2+ was increased, the synchrony of quantal secretion was increased (Bukharaeva et al. 2007). This relationship arises because intracellular Ca2+ determines the forward rate of Ca2+-activated vesicular fusion, such that higher Ca2+ accelerates the forward release reaction rate and makes short synaptic latencies (Schneggenburger and Neher 2000, 2005). Therefore, it can be suggested that the synchronizing effect of AD on the secretion of quanta is due to the increased calcium entering the nerve ending. This was confirmed by the observed evaluation of the spontaneous response frequency and an increase in QC. Recent evidence that the β-agonist salbutamol uses the P/Q-type Cav channel to enhance neuromuscular transmission (Rodrigues et al. 2019b) confirms this conclusion.

Thus, we may conclude that AD facilitates the neuromuscular junction by increasing calcium entry into the nerve ending and increasing spontaneous ACh release, increasing evoked transmitter release and synchronizing quantal secretion.