cis-Oleamide (cOA) is a brain lipid that regulates physiological sleep. It is often classified as an endocannabinoid although in vitro evidence for a CB1 receptor interaction is lacking (e.g. Boring et al. 1996). cOA is a blocker of neural Na+ channels/burst firing (Verdon et al. 2000; Nicholson et al. 2001), it positively modulates GABAA receptors (Verdon et al. 2000) and it blocks gap junctions (Guan et al. 1997). These actions are common to anticonvulsant drugs. We address the hypothesis that cOA may be an endogenous anticonvulsant.
Male Wistar rats (ca 150 g) were humanely killed. ACSF (NaCl 135 mM, KCl 3, MgCl2 1, CaCl2 2, NaH2PO4 1.25, NaHCO3 16, glucose 10) with 4AP (100 µM) was superfused for 45 min over saggital hippocampal slices (400 µm) at 35 °C. Extracellular field recordings were obtained from the CA3c subfield. 0.1 % BSA (for cOA experiments only) and 0.1 % DMSO were added to salines containing hydrophobic ligands. Values reported below are % means ± S.E.M. (normalised to pretreatment). Statistical analysis was as stated below. A P value of < 0.05 was judged significant.
4AP perfusion resulted in the appearance of both primary, inter-ictal bursts (90Ð150 ms duration, all slices) and secondary (> 0.25Ð3 s, 3 of 49 slices) ictal-like epileptiform events in the CA3 region of the hippocampus. cOA (64 mM) and WIN55212-2 (5 mM), after 60 min, both decreased the frequency of the primary events compared with time-matched controls (cOA 71.5 ± 1.7, n = 6, P < 0.001; WIN 47.25 ± 2.5, n = 4, P < 0.0001; controls 97.50 ± 2.1, n = 4, unpaired t tests). Carbamazepine (CBZ 100 mM) reversibly increased the frequency of primary epileptiform events (CBZ after 30 min 147.0 ± 7.6, n = 7, P < 0.001; wash 98.0 ± 4.0, n = 4, P > 0.05; control 101.0 ± 2.2, n = 7, paired t tests). CBZ (320 mM) caused a very transient increase in the frequency (after 1Ð2 min), but then markedly reduced their incidence after 15 min of perfusion (48.75 ± 9.3, n = 4, P < 0.05; control 99.25 ± 2.529, n = 4, paired t test). Pre-treatment of the slices for 30 min with the selective and potent CB1 antagonist AM251 (5 mM) did not affect the ability of cOA to suppress the frequency of epileptiform activity (cOA and AM251 67.33 ± 5.2, n = 3, P > 0.05, unpaired t tests) but fully blocked the effects of WIN5512-2 (WIN55212-2 and AM251 95.41 ± 3.35, n = 2, P < 0.01). Qualitatively, all three ligands rapidly and completely blocked the ictal-like discharges at these concentrations.
Factors limiting the incidence of type II discharges in this model (highly sensitive to anticonvulsants and cOA) are poorly understood. cOA exhibits anticonvulsant effects which (unlike WIN55212-2) are not mediated via CB1 receptors.
This work was supported by The Wellcome Trust, College of Pharmacy Practice, RPharm Society of GB and NI.
All procedures accord with current UK legislation.