Funny currents are becoming serious players in nociceptor's sensitization

The first step in sensing environmental signals like touch, temperature or injury by cutaneous sensory receptors is the detection of the stimulus by transducer molecules that respond to specific (i.e. adequate) stimuli generating a local, graded receptor potential. This depolarization is subsequently transformed at the nerve ending into a barrage of action potentials that travels along the axon to the CNS.

Besides standard Hodgkin–Huxley-type Na⁺ and K⁺ channels, many other ion channels shape the firing of cutaneous sensory receptors. Among them, HCN channel subunits give rise to hyperpolarization-activated inward currents (I_h). I_h currents were originally described in the heart, where they play a critical role in cardiac pacemaking (Brown et al. 1979). In the heart they were also known as funny currents (I_f) for their unusual voltage dependence. I_h currents are present in different neuronal types at various levels of the nervous system, playing important roles in membrane potential oscillations and excitability (Pape, 1996). In visceral and somatic peripheral sensory neurons it has been known for some time that expression of I_h is heterogeneous. I_h is particularly prominent among large-diameter, fast-conducting, myelinated sensory neurons (Scroggs et al. 1994; Cabanes et al. 2003), while in small-diameter sensory neurons, large I_h currents have only been observed in cold-sensitive thermoreceptor neurons (Viana et al. 2002). I_h is thought to play some role in driving neuropathic pain (Chaplan et al. 2003).

Mammalian HCN channel subunits come in four different flavours (HCN1–4), each coded by a different gene (Robinson & Siegelbaum, 2003). All can bind cyclic adenosine monophosphate (cAMP) on their intracellular C-terminus, resulting in a shift in the voltage-operating range of I_h towards less negative potentials. The functional properties of native I_h currents vary substantially in neurons, including their kinetics and cAMP sensitivity, and these differences depend largely on the molecular composition of the channels. Thus, HCN2 and HCN4 are very sensitive to intracellular cAMP while HCN1 and HCN3 are not.

When local inflammation occurs at peripheral tissues a fraction of nociceptors innervating the injured area experience a reduction in firing threshold and develop a low-frequency ongoing activity (Perl et al. 1976). This ‘sensitization’ makes them excitable to innocuous forces and enhances the magnitude of their response to injurious stimuli, leading ultimately to spontaneous pain and ‘tenderness’ of the inflamed tissues to innocuous and noxious stimuli (allodynia and hyperalgesia). The cellular mechanisms of peripheral nociceptor sensitization are still incompletely known.

In this issue of The Journal of Physiology, Momin and colleagues (Momin et al. 2008) provide novel insights into the role played by I_h in shaping the electrical activity of different subpopulations of rodent sensory neurons during inflammatory and neuropathic pain. They used cell body size as a rough but simple criterion to classify cultured sensory neurons assuming that large-diameter neurons correspond with fast-conducting thick-myelinated fibres, which serve non-nociceptive functions; medium-sized cells include a mixed population of thin-myelinated neurons that cover non-nociceptive and nociceptive roles and the smallest and slowest neurons that would give rise to unmyelinated axons, the majority of which are nociceptors (Fang et al. 2005). Chemical stimulation was additionally used to validate this functional classification.

The authors noticed that about half of small sensory neurons expressed a slowly activating I_h with strong sensitivity to cAMP. Their small size and frequent response to capsaicin and ATP suggested a nociceptive phenotype. Remarkably, inhibition of I_h with ZD7288 fully abrogated the enhanced excitability produced by PGE₂, suggesting that modulation of HCN channels is the main contributor to the enhanced excitability produced by this pro-inflammatory mediator. In addition, H-89 had no effect, ruling out effects mediated by activation of PKA. In contrast, all large sensory neurons expressed a fast activating I_h with no sensitivity to cAMP changes. Medium size neurons showed either fast or slow I_h, and only the latter was sensitive to cAMP.

In HCN1(−/−) mice, I_h was absent in 75% of large neurons and the remaining showed only a slowly activating I_h but lacking cAMP sensitivity. In medium neurons the fast I_h was also eliminated while the slow I_h remained intact. In medium size neurons, the slowly activating I_h had variable sensitivity to cAMP. The I_h in small neurons was unaffected by deletion of HCN1 and the properties suggest a major contribution of HCN2. Finally, a very small population of small-diameter neurons also expressed a fast-activating I_h, and were sensitive to menthol, consistent with the characteristics of I_h reported previously for cold thermoreceptors (Viana et al. 2002).

In an inflammatory pain model, using intraplantar injection of PGE₂, mechanical and heat hyperalgesia were unaffected or only mildly impaired in HCN1(−/−) mice. It will be interesting to compare this result with other models of inflammation. In a neuropathic pain model, that involves partial ligation of the sciatic nerve, the most prominent change in HCN1(−/−) mice was a marked reduction in cold allodynia. Whether this deficit reflects alterations in cold thermoreceptor excitability needs to be addressed in the future. It is notable and surprising that basal mechanical sensitivity and mechanical hyperalgesia were unaffected in HCN1(−/−) mice, despite the prominent expression of this subunit in large sensory neurons, putative mechanoreceptors.

In summary, we learned in this elegant study that the molecular and functional phenotype of I_h varies in the different subpopulations of sensory neurons. Up to now, enhanced firing of nociceptors by PGE₂ was thought to take place through modulation of different Na⁺ and K⁺ currents, and TRPV1 channels (Gold et al. 1996; Lopshire & Nicol, 1998). Without directly addressing the merits of these hypotheses, the new results shift the spotlight to HCN channels, a novel perspective with potential therapeutic implications for pain. Nevertheless, this will have to take into account the fact that I_h currents are prominent in many tissues and systemic therapeutic agents targeting I_h for the treatment of pain may have unwelcome side-effects.