Our understanding of quantal transmission developed initially through the skeletal neuromuscular junction (Fatt & Katz, 1952), and it took about 30 years for a similar phenomenon to be confirmed in autonomic nerves (Blakeley & Cunnane, 1979; Hirst & Neild, 1980). The first convincing demonstration of non-quantal release of acetylcholine at the neuromuscular junction occurred about 30 years ago (Katz & Miledi, 1977), and it is only now, in this issue of Experimental Physiology, that similar effects are shown for parasympathetic junctions (Abramochkin et al. 2009). One problem hampering the study of parasympathetic nerve terminals is their diffuse innervation, forming complex and irregular three-dimensional networks amidst other cells types. The electrical coupling between smooth and cardiac muscles in situ makes the interpretation of electrophysiological experiments more complex, while the presence of G-protein-coupled receptors (such as muscarinic receptors) on parasympathetic targets rather than ligand-gated ion channels (such as nicotinic receptors) makes the detection of neurotransmitter release inherently less sensitive.
Abramochkin and colleagues (2009) show that when acetylcholinesterase is pharmacologically inhibited, several cardiac parameters (including heart rate) are sensitive to the muscarinic receptor antagonist, atropine. This suggests that there is tonic release of a cholinergic agonist, almost certainly acetylcholine, from within the atria that persists when vesicular release is prevented with botulinum toxin. Further evidence that acetylcholine itself is involved comes from experiments with hemicholinium III, which inhibits choline transport into the nerve terminals, hence preventing local acetylcholine production and reducing the atropine-sensitive effects. The effectiveness of the botulinum-toxin-induced inhibition of vesicular release was elegantly demonstrated using a fluorescent indicator of vesicle cycling, FM1-43. One of the anticholinesterases used, armin, has a contentious geopolitical history, but its use is not critical for the interpretation of the study because of the additional use of hexamethonium.
The relative contributions of non-quantal and quantal transmitter release from parasympathetic nerves in normal conditions still need to be established, although such ratios do not necessarily reflect their relative importance. For example, at the skeletal neuromuscular junction the non-quantal leak of acetylcholine may be two orders of magnitude higher than the quantal release associated with the endplate potentials (reviewed by Katz & Miledi, 1977), yet the importance of quantal transmission for skeletal muscle contraction can be seen in the effectiveness of botulinum toxin for relaxing skeletal muscle, for which it is commonly used in general anaesthesia and therapeutically, to manage muscles that spasm. Botulinum toxin is also useful for treating some smooth muscles that spasm, such as in the oesophagus (achalasia) or in the overactive bladder; in both cases, botulinum toxin is an effective yet temporary treatment that causes the muscles to relax because they are no longer driven by the quantal (or packeted and vesicular) release of acetylcholine. Could non-quantal acetylcholine release be important outside the heart? Perhaps, but the pleomorphism and functional diversity of autonomic targets makes generalization difficult.
It is worth considering whether separate targeting of non-vesicular acetylcholine release might have some therapeutic benefit. So far, functional effects of such release have only been shown in the presence of acetylcholinestase inhibition, so as yet there is no evidence for an important role in normal cardiac physiology. However, acetylcholinesterase inhibitors are not infrequently given, in the context of reversing non-depolarizing neuromuscular blocking drugs or for myasthenia gravis, so it may be that during such pharmacotherapy non-vesicular acetylcholine release has some acute cardiac effects. Abramochkin et al. (2009) suggest that a low concentration of acetylcholine released from the parasympathetic nerves may be trophic or anti-apoptotic, both reasonable and testable hypotheses.
For all of the very elegant studies presented by Abramochkin et al. (2009), the non-vesicular release of acetylcholine still needs further investigation. For example, to be sure that acetycholine was the cholinergic agonist activating muscarinic receptors it would be helpful to assay acetylcholine more directly, but the required sensitivity (while maintaining specificity) is difficult to achieve. Non-neuronal sources of acetylcholine (for a review, see Kawashima & Fujii, 2008) might also play a role. While Abramochkin et al. (2009) wisely block ganglionic transmission in their model system, it would be interesting to explore the role of non-vesicular acetylcholine release in the cardiac intrinsic plexus, an understudied network that should be capable of significant information processing and signalling. Let us hope that research into autonomic junctions does not continue to lag 30 years behind its somatic cousin.