In the decerebrated rabbit, application of fentanyl to the fourth ventricle inhibits reflex responses of medial gastrocnemius (MG) motoneurones evoked by electrical stimulation of the sural nerve, an effect that is partially reversed by blockade of spinal 5-HT_1A receptors and completely blocked by spinalization (Clarke & Ward 2000). The present study investigates the possible involvement in the effects of fentanyl of spinal 5-HT_1B/1D, 5-HT₂ and 5-HT₃ receptors.

Experiments were performed on 22 rabbits decerebrated under halothane (2 %) or isoflurane (2-4 %)-nitrous oxide anaesthesia. Reflexes were evoked in the MG muscle nerve by electrical stimulation of the ipsilateral sural nerve at an intensity sufficient to activate C-fibres (mean 124 times threshold, 0.2 ms pulses). Reflexes were analysed in three time bands relative to the stimulus: 5-12 ms (phase 1); 12-100 ms (phase 2) and 100-250 ms (phase 3). Fentanyl was administered to the fourth ventricle in incrementing doses of 3, 7 and 20 µg kg^-1 at intervals of 10 min to give a cumulative dose of 30 µg kg^-1. When reflexes had recovered (usually 1 h), the 5-HT_1B/1D antagonist GR 127935 (n = 7), the total 5-HT₂ receptor antagonist ICI 170809 (Clarke et al. 1996, n = 7) or the 5-HT₃ receptor antagonist tropisetron (n = 8) was given intrathecally at 300 µg. After a further 40-60 min, fentanyl administration was repeated. Experiments were terminated by I.V. administration of 2 ml saturated KCl solution.

In the control state in all groups of animals, fentanyl induced a dose-related inhibition of all three phases of MG responses (Wilcoxon test, P < 0.01). For phase 1 reflexes, responses were reduced to median values of 31 % (inter-quartile range, IQR, 26-46 %); 16 % (IQR, 0-48 %) and 45 % (IQR 21-74 %) of pre-fentanyl levels prior to GR 127935, ICI 170809 and tropisetron, respectively. ICI 170809 significantly (Wilcoxon test, P < 0.05) reduced phase 1 reflexes to a median of 65 % (IQR 42-74 %) of pre-drug levels, but neither GR 127935 nor tropisetron had a significant effect on reflexes per se (Wilcoxon test, P > 0.1). After ICI 170809, intraventricular fentanyl inhibited the phase 1 response to a median of 44 % (IQR 28-55 %) of pre-fentanyl values, a significantly smaller effect than in the control state (Wilcoxon test, P < 0.04). Similar reductions were seen in the effects of fentanyl on phase 2 and 3 reflexes. After tropisetron and GR 127935, the effects of fentanyl on phase 1 reflexes were statistically indistinguishable from those obtained before the drug (Wilcoxon test, P > 0.4). However, inhibition of phase 2 and 3 reflexes was reduced after GR 127935 (Wilcoxon tests, P < 0.04).

These data suggest that 5-HT_1B/1D and 5-HT₂, but not 5-HT₃, receptors contribute to the inhibition of reflexes after intraventricular administration of fentanyl.

This work was supported by BBSRC. GR 127935 was a gift of GlaxoSmithKline.