Action potential initiation in unmyelinated afferents comes in from the cold

Abstract

Understanding the mechanisms of signal transduction and action potential generation within unmyelinated nerve endings allows us an insight into the first stages of nociception, the raw sensory input that leads to the perception of pain. Signal transduction in A-fibres has been well described, and the highly circumscribed distribution of Na+ channels in myelinated nerve at nodes of Ranvier has an impact on the transduction processes as well as impulse conduction. The results of classical experiments, for example those of Lowenstein & Mendelson (1965) on mechanical transduction in the Pacinian corpuscle, have shown that action potential initiation takes place at the first node, separately from mechanical transduction at the encapsulated ending. Clearly, mammalian unmyelinated endings might be expected to be different from this, because their Na+ channels are distributed more evenly, but how different? In this issue of The Journal of Physiology, Carr et al. (2009) have used three characteristics of unmyelinated nerve endings sensitive to cold and innervating the cornea in the isolated guinea pig eye to investigate the site of action potential initiation in single axons, and resolve it to sites (plural) close to the termination. The characteristics that make these tiny afferents yield their secrets in this study are the way action currents from individual endings can be recorded using an electrode pressed against the cornea, the slow conduction velocity, allowing the resolution of distance by a functional collision technique, and the presence of on-going activity with a cool cornea. Carr and colleagues have now provided functional evidence that is it likely not only that transducer channels and Na+ channels operate in the same membrane space within corneal leash fibres, but that, presumably depending on the magnitude of the generator potential, the action potential initiation site can move, and that more than one initiation site can exist within the receptive field of a single cold-sensitive axon. As the initiation site moves backwards away from the corneal surface with a larger generator potential, it may integrate the local responses provided by several individual collateral branches. The previous and present work of Richard Carr, James Brock and colleagues on corneal afferents (e.g. Brock et al. 1998; Carr et al. 2003) has provided important evidence as to how unmyelinated afferent endings work. The discovery that many of the Na+ channels conferring excitability in these small endings (although probably not all) appear insensitive to the marine toxin tetrodotoxin (TTX) implicates the transient TTX-resistant channel, NaV1.8, in impulse generation. This result fits appositely with the finding that in the NaV1.8 knock-out, nociceptor signalling in the cold is lost, including responses to noxious cold (Zimmermann et al. 2007). We surmise that nociceptor signalling in the cold must be initiated by impulse generation in normal endings where substantial generator potentials exist, substantial enough for other Na+ channel subtypes to drop out through inactivation. This reasoning implicates NaV1.8 in maintaining excitability in the face of long lasting depolarizations, and may be an important part of the explanation as to why NaV1.8 is a key player in neuroma discharge (Roza et al. 2003). One problem in understanding function in small diameter afferents is that it is not possible to measure the membrane potential. We can guess at a value, and also hope that measurements of membrane potential made in patch-clamped cell bodies are telling us something applicable, so that we might be in a position to understand what various ion channels are doing in the endings. In contrast, extracellular recordings of action current in single afferent endings have provided vital direct information about the state of Na+ channels in the ending, and the degree of channel inactivation. For example, observations on action current amplitude made in the cornea, where external hyperpolarizing currents are employed (Carr et al. 2002), can be compared with measurements of Na+ channel voltage dependence and behaviour made in voltage-clamped and current-clamped cell bodies (e.g. Blair & Bean, 2003), potentially providing a safer route to form an estimate of nerve ending membrane potential. Applying hyperpolarizing currents to an antidromically activated ending allowed the impulse to invade closer to the termination, by locally reducing Na+ channel inactivation, and increasing the amplitude of action currents. Carr et al. (2002) made such a comparison and estimated −30 mV for a working cold receptor, a potential that happens to be close to the activation threshold for NaV1.8. In conclusion, using a technically demanding electrophysiological approach, Carr et al. (2009) have made an important contribution to our knowledge of the functional properties of the membrane of unmyelinated nerve endings in the cornea and the origin of the nerve impulse in small afferents. Unfortunately, at least at the moment, it seems technically impossible to make similar recordings in mouse that would allow the investigation of the knock-out of ion channel subtypes on nerve ending function.