Spontaneous Acetylcholine Release Potentiation Induced by 2-Arachidonoylglycerol and Anandamide in Mouse Motor Synapses

Changes in the parameters of miniature endplate potential (MEPP) of mouse diaphragm caused by exogenous application of two classical endocannabinoids, 2-arachidonoylglycerol (2-AG) (1 μM) and anandamide (AEA) (30 μM), were compared. A slowly developing stable increase in the MEPP amplitude by 50%, without affecting the frequency of MEPPs, was caused by 2-AG. This effect was prevented by AM-251, an inverse agonist of CB1-receptors, as well as by vesamicol, a blocker of the vesicular acetylcholine (ACh) transporter. On the contrary, AEA did not cause significant changes in the MEPP amplitude but induced a slowly developing (within 2 h) increase in MEPP frequency by 75% on average. The effect of AEA was prevented by AM-251 (1 μM) as well as by blocking of L-type Ca2+-channels with nitrendipine (1 μM) and inhibition of PKA activity by H89 (1 μM). It was concluded that both 2-AG and AEA are able to exert a noncanonical facilitating presynaptic effect on spontaneous ACh release. Even though these endocannabinoids activate the same type of CB-receptors, their facilitating effects do not overlap and are strictly aimed either at potentiating the size of ACh quanta (in case of 2-AG) or increasing the frequency of MEPP (in case of AEA). We assume that different intracellular targets and signaling pathways may be involved in the differentiated facilitating effects of 2-AG and AEA in mouse neuromuscular junctions.

Retrograde signaling at synapses was discovered due to detection in neurons of a special class of chemical signals, the so-called endocannabinoids. These are lipophilic molecules derived from membrane phospholipids in response to the incoming signals received by the neuron [1]. Currently, the two best known structurally different endocannabinoids are 2-arachidonoylglycerol (2-AG) and anandamide (AEA). Both of them can be synthesized during synaptic activity, which leads to their release from postsynaptic neurons and their retrograde action on receptors. This, in turn, is accompanied by the inhibition of neurotransmitter secretion in the central synapses [2,3]. Recently, the activity of the endocannabinoid system has also been discovered at peripheral synapses [4,6]. For skeletal muscles, enzymatic systems of the synthesis and degradation of cannabinoids have been described [7][8][9] as well as endocannabinoid receptors (СВ receptors) localized both on the muscle and in the area of neuromuscular synapses [4,5,10,11]. The functional significance of the endocannabinoid system is often analyzed using synthetic agonists and antagonists of CB receptors, with their effects in synapses being both similar to and different from the effects of classical cannabinoids [1,3]. We previously showed that the synthetic CB receptor antagonist WIN 55,212-2 causes an increase in spontaneous secretion of acetylcholine (ACh) at the motor synapses of mouse diaphragm, which is prevented by the blockers of CB1 receptors [5]. It is still unknown whether 2-AG and AEA are able to cause similar effects in motor synapses or have their own effects. Hence, the present work was aimed at comparative analysis of the effects of 2-AG and AEA in the motor synapses of the mouse diaphragm and at comparison of their effects with the earlier revealed effects of WIN 55,212-2.

MATERIALS AND METHODS
Object of research. The experiments were conducted using neuromuscular isolated diaphragm muscle preparations (m. diaphragma -n. phrenicus) of adult BALB/c mice (45-60 days old) of both sexes from the Laboratory of Experimental Animals, Faculty of Biology, Moscow State University (Moscow, Russia). The maintenance of animals and experimental procedures were in accordance with the Directive 86/609/EEC on the protection of animals used for experimental and other scientific purposes. The application for animal experiments (95 g) was approved by the Committee on Bioethics of the Moscow State University. The experiment involved 27 experimental animals, one neuromuscular preparation being isolated from each animal.
Electrophysiology. The mice were swiftly decapitated; their abdominal and thoracic cavities were opened, and the left half of the diaphragm with the entering phrenic nerve was extracted. The isolated neuromuscular preparation was placed into a 3-mL experimental chamber perfused with oxygenated (95% O 2 , 5% CO 2 ) Liley solution [12]. Spontaneous miniature endplate potentials (MEPPs) were recorded intracellularly using glass microelectrodes filled with 2.5 M KCl (microelectrode tip resistance, [15][16][17][18][19][20][21][22][23][24][25]. Signals were recorded with a Neuroprobe Amplifier Model 1600 (A-M Systems, United States) or Axoclamp-2B (Molecular Devices, United States) and, using an Е-154 ADC/DAC module (L-Card, Russia) with the PowerGraph 6.0 interface, saved into a PC hard drive for further analysis in MiniAnalysis (Synaptosoft, United States). The criteria of successful insertion of the microelectrode into the endplate region were stable membrane potential (MP) and MEPP 10-90% rise time of <1 ms. It ceased when MP decreased by more than 5 mV of the initial value during the recording, and the signals from this synapse were not included in the sample for further analysis. In each of the studied synapses, MEPPs were recorded for 120 s. In the control, MEPPs were recorded in 5-7 synapses, after which the neuromuscular preparation had been perfused with Liley solution containing the substances under study. When investigating the effect of 2-AG or AEA, we recorded the spontaneous activity of different synapses during 2-h perfusion of neuromuscular preparation with a solution containing the respective endocannabinoid. When studying the mechanism of endocannabinoid effects, the respective reagent was added to the perfusion solution 15 min before the addition of 2-AG or AEA. The spontaneous activity of different synapses was then recorded under these conditions for 2 h. No less than three neuromuscular preparations were used in each series of experiments.
Data analysis and statistics. We evaluated the MP of muscle fibers, the time course of MEPPs (rise time and half-decay), their amplitude, and frequency. The parameters of the MEPP time course did not change significantly compared to the control in all series of experiments. In order to neutralize the driving force of potential shift during MP changes, we standardized the values of MEPP amplitudes to MP -70 mV by the formula А st = A × (-70/MP), where A is the recorded MEPP amplitude and А st is the standardized MEPP amplitude. Statistical data analysis was performed with GraphPad Prism 6.0. The data in the text and in the figures are presented as "the mean ± standard error of the mean." The normal distribution of MEPP parameters was assessed by the D'Agostino-Pearson normality test. The reliability of difference scores between the series was assessed by one-way analysis of variance in case of normal distributions of values in each of the compared samples. The Kruskal-Wallis test was used in case of distribution differing from the normal value in at least one sample. The differences between the samples were considered significant at p < 0.05 (n is the number of tested synapses).
Materials. The following agents were used in the experiments: endocannabinoids AEA (Merck, United States) and 2-AG, СB1 receptor inverse agonist AM 251; protein kinase A inhibitor H89; phospholipase C inhibitor U73122; L-type Ca 2+ channel blocker, nitrendipine (Torcris Bioscience, United States); the inhibitor of vesicular ACh transporter, vesamicol (Enzo Life Sciences, United States). The stock solutions of all substances except for H89 were prepared with DMSO (dimethyl sulfoxide) (Helicon, Russia) as a solvent. The solvent for H89 was deionized H 2 O. The final concentration of DMSO in the working solution was no more than 0.01% (v/v) and had no independent effect on the parameters of spontaneous activity in mouse motor synapses.

RESULTS AND DISCUSSION
In the first part of the work, the changes in the MEPP parameters of neuromuscular synapses of the mouse diaphragm were studied under the conditions of exposure to 2-AG (1 μM). It was shown that the average value of muscle fiber MP did not change during its 2-h exposure to 2-AG. No changes in the average frequency of MEPPs were recorded in this period of time: it was 0.46 ± 0.04 Hz (n = 16) in the control, 0.46 ± 0.04 Hz (n = 16) and 0.38 ± 0.04 Hz (n = 17, p > 0.05) in the first and second hour of application, respectively. In addition, there was a gradual increase in the MEPP amplitude from 1.35 ± 0.11 mV in the control to 2.03 ± 0.24 mV by the second hour of the exposure to 2-AG (p < 0.05), which persisted after 1-h washing (Fig. 1A). It is known that the increase in the MEPP amplitude can be of both pre-and postsynaptic nature [13]. The nature of the effect of 2-AG was studied by the inhibition of ACh transport into synaptic vesicles by vesamicol (1 μM). Vesamicol per se has no effect on the MEPP amplitude during long-term    [4,12], but 2-AG in its presence proved to be unable to increase the MEPP amplitude (Fig. 1b). Since 2-AG can influence both types of CB receptors but CB1 is more widespread in the nervous system [1], the receptor effect of 2-AG was tested with the involvement of the CB1 receptor inverse agonist AM 251 (1 μM) extensively used to prevent the development of CB1-mediated effects. AM 251 per se did not cause any changes in the parameters of MEPP [4,5]; however, 2-AG in its presence did not induce an increase in the MEPP amplitude (Fig. 1b). The findings suggest that the effects of 2-AG can be a result of its presynaptic receptor effect on CB1 receptors. Their activation triggers the intracellular signaling cascade targeting the stimulation of ACh loading into vesicles and, thereby, the increase in the size of spontaneously secreted quanta. An analogous effect of vesamicol, i.e., prevention of delayed increase in the MEPP amplitude under the influence of endocannabinoid receptor agonists, has been described recently for neuromuscular synapses of the mouse diaphragm [4]. The second part of the work was devoted to the analysis of the effects of another endocannabinoid-AEA (30 μM). As in the case of 2-AG, AEA did not cause any statistically significant changes in the MP of muscle fibers. In contrast to the MEPP amplitude increasing in the presence of 2-AG, the AEA-sensitive parameter of MEPPs proved to be their frequency: there was a marked increase in its average value from 0.37 ± 0.03 Hz (n = 15) in the control to 0.65 ± 0.05 Hz by the second hour of АЕА application (n = 17, p < 0.05) (Fig. 2a). At the same time, there were no reliable changes in the MEPP amplitude: the mean value of this parameter was 1.61 ± 0.14 mV in the control and 1.64 ± 0.14 (p > 0.05) by the second hour of exposure to AEA. The receptor specificity of AEAinduced increase in the MEPP frequency was also tested using AM 251 (1 μM). It was shown that the AEA-induced increase in the MEPP frequency is completely prevented in the presence of CB1 receptor inverse agonist (Fig. 2b). This fact suggests that, as in the case of 2-AG, the AEA-induced effect is specific and seems to be associated with the activation of CB1 receptors.
The possible causes of increased MEPP frequency in motor synapses can be the increased activity of some enzymes: protein kinase A (PKA), phospholipase C (PLC), and/or and increased Ca 2+ level in the terminal with the involvement of different types of Ca 2+ -channels [5,14]. Hence, we further investigated whether the AEA-induced increase in the MEPP frequency will manifest itself in the presence of PLC or PKA inhibitors as well as blocking of Ca 2+ -channels. It turned out that the PLC inhibitor U73122 (5 μM), which does not affect the MEPP parameters per se [12], is incapable of preventing the AEA-induced increase in the MEPP frequency (Fig. 2b). This fact contradicts the previous data showing that, in the case of activation of CB1 receptors by their synthetic agonist WIN 55,212-2, the statistically significant increase in MEPP frequency was prevented by the blocking of PLC and protein kinase C activities but not by the inhibition of PKA [5]. In the present work, with AEA used as a CB1 receptor agonist, the increase in MEPP frequency was prevented by the PKA inhibitor H89 (1 μM) but not by the PLC inhibitor (Fig. 2d). Finally, nitrendipine, the selective blocker of voltagedependent L-type Ca 2+ channels, which had no reliable effect on MEPP parameters per se [15], also prevented the AEA-induced increase in MEPP frequency (Fig. 2e). The latter fact indicates the Ca 2+ -dependence of the AEA-induced increase in MEPP frequency in motor nerve terminals, which is in agreement with our previous data on the dependence of the WIN 55,212-2-induced increase in MEPP frequency on stored Ca 2+ and the activity of ryanodine receptors [5], which are usually activated and release Ca 2+ from the stores in response to Ca 2+ entry through the L-type Ca 2+ -channels. The ability of these channels, in the case of intensification of their activity due to the effect of enzymes (PKA) or their agonist BayK8644, to induce Ca 2+ entry into the terminal and lead to an increase in MEPP frequency has been described previously [16,17]. We have shown for the first time that the AEA-induced increase in MEPP frequency is related not only to the function of presynaptic L-type Ca 2+ -channels but also to the activation of PKA in the terminals of motor synapses. This suggests that the effect of AEA is mediated in this case by activation of not the canonical G i protein but another, probably G s -type, protein. As it is known, this type of G protein is a stimulator of the adenylyl cyclase signaling cascade aimed at the activation of PKA. In turn, one on the targets of such signaling in mouse motor nerve terminals is presynaptic L-type Ca 2+ -channels. The potential involvement of the signaling pathway stimulating adenylyl cyclase and PKA in implementation of the effect of endocannabinoids is much less studied than the classical activation of G i -dependent signaling pathways. At the same time, such assumption seems to be justified due to the discovery of numerous manifestations of biased agonism in endocannabinoids when their receptors are affected [18]. In particular, the ability of CB1 receptors either to directly activate G s protein [19] or to inhibit G i protein and, thereby, to enhance the tonic activity of G s protein-coupled cascade and PKA under the exposure to different agonists is under discussion [20]. Finally, the ability of endocannabinoid receptor agonists to activate the cascade with the involvement of G s protein and PKA not directly but via coupling to β-arrestins has been described recently [21]. The ques- tion about the particular type of G proteins and reactions mediated by the latter, which stimulates the activity of L-type Ca 2+ -channels and PKA, leading to a stable AEA-induced increase in MEPP frequency in motor synapses, is to be addressed in further experiments.
The slow development of the effects of endocannabinoids demonstrated in the present work, which takes approximately 2 h since the beginning of their application, had been previously observed in motor synapses when studying the effects of CB receptor agonists on MEPP parameters [4,5] as well in the synapses of the central nervous system under CB-mediated modulation of long-term plasticity [2,18]. Such dynamics of manifestation of the effects seems to be a consequence of the involvement of signaling pathways of different lengths and different targets [22]. The particular causes of the slowly developing effects of endocannabinoids in motor synapses are an issue that undoubtedly requires further studies.
Thus, the comparative analysis of the effects of two structurally different endocannabinoids has shown for the first time that 2-AG and AEA act via presynaptic receptors and cause an enhancement of spontaneous ACh secretion in mouse motor synapses: 2-AG increases the MEPP amplitude without changing their frequency, while AEA increases the MEPP frequency without affecting their amplitude.
The examples of the facilitating effects of endocannabinoids on neurotransmitter secretion, though described in literature, are few [18,23]. They are currently considered as manifestation of a much broader (than has been previously supposed) range of the functional activity of endocannabinoids in synapses. There is active discussion on the ability of CB receptors to displace agonists and the possibility of their interaction not only with canonical G i protein but also with other types of G proteins (G q and/or G s ) [18,22] as well as the interaction between CB receptors and β-arrestins, which triggers separate signaling cascades. The heteroreceptor complexes formed by CB1 receptors and other membrane receptors of cells can also mediate the atypical effects of endocannabinoids [24]. Thus, the divergence of pathways, targets, and effective time range of endocannabinoids mentioned in literature is now explained by the pluripotency of these signaling molecules [22,24]. The complex determinancy and ambiguousness of physiological effects, typical for endocannabinoids, apparently have also been manifested in motor synapses and caused the differently directed facilitating effects of 2-AG and AEA, despite the fact that they activate the same presynaptic CB1 receptors. The detailed mechanisms of such effects of 2-AG and AEA mediated by CB1-type receptors are yet to be elucidated.
In our opinion, the long-lasting effects of exogenous 2-AG and AEA becoming apparent due to their long-term action can serve as a model of tonic consti-tutive effects of endocannabinoids in motor synapses. The possible contribution of tonic effects of endocannabinoids to the maintenance of the homeostasis of neurons and synapses in early ontogenesis is also discussed with regard to the synapses of the central nervous system [25]. It is obvious, however, that the effects of endocannabinoids may vary under different conditions of synaptic functioning, in particular, depending on the release of endocannabinoids from muscle fibers or perisynaptic Schwann cells at rest or during muscle contraction. Their manifestations are yet to be elucidated.
Summing up the data obtained in the present work, we can state that the analysis of the effects of two classical endocannabinoids (2-AG and AEA) in mouse motor synapses has shown their noncanonical oneway effect in the form of induction of slowly developing potentiation of spontaneous ACh secretion for the first time. Despite activation of the same type of receptors (CB1 type), 2-AG and AEA caused a targeted increase in either MEPP amplitude (reflecting the increasing size of ACh quanta) or MEPP frequency (reflecting the increase in the frequency of spontaneous release of ACh quanta). We have shown that stimulating effects targeting different parameters of MEPPs presuppose the involvement of different signaling pathways and their targets, including the activation of PKA and L-type Ca 2+ -channels, which is coupled to the AEA-induced increase in MEPP frequency and the increase in the loading of ACh into synaptic vesicles caused by 2-AG.
Overall, the findings expand our notions of possible manifestations of the functional activity of endocannabinoids, not only in the on-demand mode in response to induced synaptic activity but probably also in the tonic mode in quiescent synapses. Obviously, the complete comprehension of multiple and ambiguous effects of the function of the endocannabinoid signaling system in motor synapses require further analysis.

FUNDING
This study was supported by the Russian Foundation for Basic Research (project no. 19-04-00616a).

COMPLIANCE WITH ETHICAL STANDARDS
Conflict of interests. The authors declare that they have no conflict of interests.
Statement on the welfare of animals. All applicable international, national, and/or institutional guidelines for the care and use of animals were followed.