BACKGROUND Ligation injury of the L5/L6 nerve roots in rats produces behavioral signs representative of clinical conditions of neuropathic pain, including tactile allodynia and thermal and mechanical hyperalgesia. In this model, intrathecal morphine shows no antiallodynic activity, as well as decreased antinociceptive potency and efficacy. This study was designed to explore the antinociceptive activity of intrathecal clonidine alone or in combination with intrathecal morphine (1:3 fixed ratio) in nerve-injured rats. The aims, with this study, were to use nerve-injured animals to determine: (1) whether the antinociceptive potency and efficacy of intrathecal clonidine was altered, and (2) whether the combination of intrathecal morphine and clonidine would act synergistically to produce antinociception. METHODS Unilateral nerve injury was produced by ligation of the L5 and L6 spinal roots of male Sprague-Dawley rats. Sham-operated rats underwent a similar surgical procedure but without nerve ligation. Morphine and clonidine were given intrathecally through implanted catheters alone or in a 1:3 fixed ratio. Nociceptive responses were measured by recording tail withdrawal latency from a 55 degrees C water bath, and data were calculated as % maximal possible effect (%MPE). RESULTS Morphine produced a dose-dependent antinociceptive effect in both sham-operated and nerve-injured rats. The doses calculated to produce a 50 %MPE (i.e., A50) (+/-95% confidence intervals [CI]) were 15 +/- 4.9 micrograms and 30 +/- 18 micrograms, respectively. Though morphine was able to produce a maximal response (100%) in sham-operated rats, the maximal response achieved in nerve-injured animals was only 69 +/- 21.9 %MPE. Clonidine produced a dose-dependent effect, with an A50 (+/-95% CI) of 120 +/- 24 micrograms in sham-operated rats. In nerve-ligated rats, clonidine produced a maximal effect that reached a plateau of 55 +/- 10.9 %MPE and 49 +/- 10.2 %MPE at 100 and 200 micrograms, respectively, preventing the calculation of an A50. In sham-operated rats, a morphine-clonidine mixture produced maximal efficacy, with an A50 (+/-95% CI) of 15 +/- 9.2 micrograms (total dose), significantly less than the theoretical additive A50 of 44 +/- 10 micrograms. In L5/L6 nerve-ligated rats, the morphine-clonidine combination produced maximal efficacy, with an A50 (+/-95% CI) of 11 +/- 5.4 micrograms (total dose), which was significantly less than the theoretical additive A50 of 118 +/- 73 micrograms, indicating a synergistic antinociceptive interaction. The intrathecal injection of [D-Ala2, NMePhe4, Gly-ol]enkephalin (DAMGO) produced A50 values of 0.23 microgram (range, 0.09-0.6) and 0.97 microgram (range, 0.34-2.7) in sham-operated and ligated rats, respectively. Phentolamine (4 mg/kg, intraperitoneally) produced no antinociceptive effect alone and attenuated, rather than enhanced, the effect of morphine in both groups of rats. CONCLUSIONS These data show that: (1) clonidine, like morphine, loses antinociceptive potency and efficacy after nerve ligation injury, and (2) strongly suggest that a spinal combination of morphine and clonidine synergize under conditions of nerve injury to elicit a significant antinociceptive action when either drug alone may be lacking in efficacy. It is unlikely that the synergy of morphine with clonidine is due to an attenuation of spinal sympathetic outflow by clonidine, because the sympatholytic agent phentolamine produced an opposing effect on morphine antinociception. The data suggest that combinations of morphine and clonidine may prove useful in controlling pain in patients with neuropathic conditions.