The question of whether and to what extent the phenotype/functions differ in these two types of primary afferent dorsal root ganglion (DRG) neurone will be examined. A number of broad questions about functions of C- and A-fibre afferent neurones will be raised and possible answers for some of these will be discussed. Differences in electrophysiological membrane properties between both nociceptive and non-nociceptive and between C-fibre and A-fibre nociceptive neurones are becoming increasingly clear. Of the phenotypic properties specific to nociceptive neurones, many are expressed in the following order of magnitude: C > Aδ > Aβ. These include the following: slow conduction velocity (CV) (by definition), broader action potentials (APs), greater AP overshoot, lower fibre firing frequency, and higher incidence of spontaneous activity and expression of the NGF receptor trkA. The possible contribution of several Na⁺ channel subunit types to these nociceptive properties will be considered. Interestingly mean AHP (afterhyperpolarisation) durations, which are also indicative of the nociceptive nature of the unit, are equally long in nociceptive neurones of all CVs. All these properties appear likely to limit the rate of information transfer along nociceptive fibres, especially the slower conducting units.

There is little evidence of discontinous functional differences between Aδ and Aβ nociceptive neurones, but rather a gradual shift of properties with increasing conduction velocity (CV) away from the extreme (C-fibre) form of the nociceptive phenotype, perhaps linked with the decreasing trkA expression (and therefore possibly lower influence of NGF) in the faster conducting units. Functionally, both CV ranges include mechano-heat units and moderate pressure receptive units.

The normal limitation of firing rate in nociceptive neurones can be reduced or removed by tissue inflammation. Inflammation, via upregulation of tissue NGF, alters many membrane properties over a few days, thus increasing the potential rate of information transfer especially along the slowly conducting nociceptive C-fibres. In addition, the proportion of C-fibre units that show spontaneous activity increased > 3-fold to 50 % of units 4 days after tissue inflammation. Are these spontaneously active units responding to inflammatory mediators in the tissues? If so, such mediators are the adequate stimuli for these units, making them detectors of actual rather than potential tissue damage. In this context the possible role of C-fibre units that are normally unresponsive to noxious mechanical or thermal stimuli will also be considered.

Since so much important work on the properties of putative nociceptors is carried out on isolated neurones in culture, it would be helpful to establish clear methods of distinguishing electrophysiologically or immuno-cytochemically (a) between non-nociceptive and nociceptive neurones and (b) between C-fibre and A-fibre nociceptive neurones. Since trkA-like immunoreactivity (LI) labels such a high proportion of A- and C-fibre nociceptive units, and since A-fibre somata are neurofilament rich, double-labelling to show these could provide a clear distinction between A- and C-fibre nociceptive neurones and other neurones. However, since some D hair units also show weak trkA-LI, some caution is needed in relation to these units. Electrophysiologically D hair units could be distinguished from nociceptive units by their short AP and AHP durations and smaller AP overshoots. In in vivo experiments on somatic afferent units, it is hard to test units that do not have superficial receptive fields with heat or noxious chemicals. Thus the only units which clearly do not have polymodal receptive characteristics are high threshold mechanoreceptive (HTM) units with superficial receptive fields. Interestingly, there is an emerging pattern that DRG neurones with such receptive fields tend to be devoid of peptides, while those with deeper receptive fields appear more likely to express both peptides and the NGF receptor trkA. In the context of linking afferents and tissue homeostasis, it may worth considering the possibility that units with trkA and peptides may be those that detect deep tissue damage and inflammation, while those with superficial receptive fields may be those that provide the first response to the external environment, and thus may contribute to reflex withdrawal responses. It is therefore interesting that a higher proportion of rapidly conducting A-fibre than C-fibre mechanical nociceptors do not respond to heat. Is it therefore possible that the faster conducting, non-peptidergic units are more concerned with withdrawal reflexes than the slower conducting peptidergic NGF-dependent units, and that the latter may detect tissue damage and inflammation? Following this line of thought, it may be the depth or location of a receptive field that defines it as nociceptive, rather than a particularly high threshold at the receptor terminal. Thus the consideration of depth or location of receptive field may be important in understanding the relative functions of C- and A-fibre nociceptive units.

This work was supported by a Wellcome Trust, UK grant to S.N.L.