K⁺ channels play an essential role in setting the resting membrane potential and controlling the excitability of neurons, and so represent potentially attractive targets for the treatment of pain.

Recently, the novel anticovulsant retigabine was shown to have anti-neuropathic activity in two models of chronic pain, whilst having no effect in a model of acute pain (Rostock et al. 2000). The action of retigabine is mediated through enhancement of currents generated by neuronal KCNQ channels (KCNQ2-5) (Main et al. 2000; Wickenden et al. 2000; Tatulian et al. 2001). These channels are the molecular correlates of the M-channel (Wang et al. 1998), which carries a slowly activating, non-inactivating, voltage-dependent K⁺ current (I_M) that dampens excitability (Brown, 1988).

In this study, using whole-cell perforated-patch recording, we sought the presence of M-current in cultured dorsal root ganglion (DRG) neurons from rats (17 days old, humanely killed according to approved Schedule 1 methods). M-current was the dominant subthreshold sustained current in all small cells tested (capacitance 20.4 ± 1.1 pF, n = 30; numbers are means and S.E.M. throughout), of which 16/22 cells were sensitive to capsaicin. M-current was also present in the majority (9) of large cells (capacitance > 100 pF, n = 10) tested, but in contrast to small cells, large dendrotoxin-sensitive (DTX, 100 nM) and Cs⁺-sensitive (1 mM) currents were also observed.

The kinetics and pharmacology of M-current in small DRG neurons were further characterized. M-current activated at ~-60 mV and deactivated slowly (t_fast = 76.4 ± 9.9 ms, t_slow = 583 ± 134 ms, n = 9). I_M was inhibited by the M-channel blocker linopirdine (IC₅₀ 2.1 ± 0.2 µM; n = 8), its analogue XE991 (IC₅₀ 0.26 ± 0.01 µM; n = 6), Ba²⁺ (IC₅₀ 0.3 ± 0.04 mM; n = 4) and TEA (IC₅₀ 1.1 ± 0.08 mM; n = 7). As expected, retigabine (10 µM) enhanced I_M in a voltage-dependent manner (EC₅₀ values: 0.18 ± 0.02 and 1.19 ± 0.07 µM at -20 and -50 mV, respectively, n = 7). Furthermore, linopirdine (10 µM) and retigabine (10 µM) reduced and increased the threshold of firing, respectively.

RT-PCR confirmed the presence of all four neuronal KCNQ subunits in whole DRG, though KCNQ4 was absent at the single cell level. Immunocytochemical data provided further evidence for the presence of KCNQ subunits in both small and large DRG neurons.

The role of I_M in the processing of neuropathic pain was investigated by recording the responses of dorsal horn spinal neurones in both naive and neuropathic (spinal nerve ligation) rats (halothane/nitrous oxide-anaesthetized). Spinal retigabine (10-90 µg) exerted dose-related inhibition of both the electrically and low- and high-intensity mechanical and thermal evoked neuronal responses in naive animals. Nociceptive primary afferent C-fibre responses and measures of spinal cord hyperexcitability were most susceptible to retigabine, whereas Aβ-fibre evoked responses were spared.

Finally, the effects of activating M-current were examined in a behavioural model of inflammatory hyperalgesia. Hyperalgesia was assessed using a dual weight averager (Clayton et al. 1997). Following intraplantar administration of carrageenan (2 %, 100 µl), animals treated with vehicle only distributed 21 ± 3 % of their hindleg load onto the inflamed paw. Retigabine (5 mg kg^-1 P.O.) produced a reversal of the decrease in weight bearing on the inflamed pore (41 ± 2 %), while animals treated with both retigabine and XE991 (5 mg kg^-1 P.O.) distributed a similar weight to those treated with vehicle alone (28 ± 3 %).

Together, these findings suggest that I_M helps to control excitability in nociceptors and that it may represent a novel therapeutic target for the treatment of pain.

This work was supported by the UK Medical Research Council and the European Union (grant QLG3-CT-1999-00827).