To define potential alcohol binding sites in the neuronal nicotinic acetylcholine receptor (nAChR) we used cysteine mutagenesis and sulfhydryl-specific labeling. The basis of this strategy is that covalent addition of an alkylthiol group to a cysteine in an alcohol binding site will mimic the action of an irreversibly bound alcohol. Each amino acid in the extracellular region of the second transmembrane segment of the nAChR subunit alpha2 was mutated to cysteine. The resulting alpha2 subunits were coexpressed with wild-type beta4in Xenopus laevis oocytes, and the responses were studied using two-electrode voltage clamp. Of the 11 mutants tested, 2 fulfilled criteria for participation in an alcohol binding site: alpha2(L262C)beta4 and alpha2(L263C)beta4. Covalent binding of propanethiol to these cysteines did not change acetylcholine (ACh) affinity, but modified ACh maximal response in both cases: it increased for alpha2(L263C)beta4 and decreased for alpha2(L262C)beta4. The same modifications on ACh responses were obtained with ethanol on alpha2(L263C)beta4 and octanol on alpha2(L262C)beta4. This suggested that alcohol binding at L263 enhances receptor function, whereas binding at L262 inhibits function. We studied different n-alcohols (ethanol, butanol, pentanol, and octanol), as well as isoflurane and urethane, on these two mutants. Covalent binding of propanethiol to the cysteines revealed changes in the alcohol modulation consistent with an excitatory site (L263) or an inhibitory site (L262) being no longer accessible to alcohol. Thus, n-alcohols appear to act on both sites and their ability to enhance (short-chain), inhibit (long-chain), or produce no effect (intermediate-chain) depends upon their relative action at these two sites.

Recent investigations have suggested that abnormal increases in brain cyclic 3',5'-adenosine monophosphate (cAMP) may play a role in epileptogenesis. Therefore, the effect of three commonly used antiepileptic drugs on cAMP accumulation in rat cortex slices was investigated. Ouabain, a depolarizing agent which produces seizures when applied to rat cortex, produced a five- to sevenfold increase in cAMP accumulation, and both carbamazepine and and phenytoin inhibited this increase. Ouabain stimulation may be mediated by the release of endogenous adenosine, and carbamazepine antagonized adenosine stimulation of cAMP accumulation whereas phenytoin did not. Carbamazepine had no effect on adenosine efflux. The augmentation of cAMP accumulation by norepinephrine was inhibited by carbamazepine and phenobarbital but slightly increased by phenytoin. If increases in brain cAMP are involved in epileptogenesis, the antagonism of cAMP accumulation by antiepileptic drugs may play a role in their anticonvulsant action.

Ncitro1og.y S r r v i w . Vrtoruns Administrution Metliial Ci>ntc)r. and Departmentoj' Neurology, Univi,rsity of' Colorado Hcalt h Scienc.c~ Cen tor I Dc,nwr, ColoradoSummary: Adenosine and its immediate metabolites, inosine and hypoxanthine, were measured in mouse brain following the induction of electroshock seizures and after a subconvulsive series of electric shocks. Electroshock seizures resulted in a marked and prolonged rise in inosine, with maximal values at 5 min. Hypoxanthine increased more slowly but reached high levels by 10 min. Adenosine was unchanged. Phenytoin and to a lesser extent phenobarbital reduced these effects. Following the subconvulsive stimulus. 15 single shocks over an interval of 5 sec. inosine increased rapidly, adenosine rose slightly, and hypoxanthine did not change. Both phenytoin and phenobarbital blocked these increases in adenosine and inosine. Early elevations in inosine may play some role in seizure generation and propagation. The high levels of inosine and hypoxanthine found after recovery may be involved in the termination of epileptic activity, possibly by interacting with the benzodiazepine receptor for which they are ligands.Recent observations have suggested that adenosine and its immediate metabolites, inosine and hypoxanthine, may play a role in epileptogenesis. The injection of adenosine, inosine, or hypoxanthine into rat cortex at concentrations down to lo-' M produces epileptiform discharges and seizures (Lewin, 1976). These purines increase in rat brain following electrical stimulation in vivo (Schultz and Lowenstein, 1978). The release of adenosine and its metabolites from cortical slices is enhanced by electrical stimulation (Pull and Mcllwain, 1972) and by depolarizing agents (Shimizu et al., 1970); and phenytoin, but not phenobarbital, inhibits the emux elicited by ouabain and veratridine (Lewin and Bleck, 1976). Therefore, it was elected to measure adenosine, inosine, and hypoxanthine in mouse brain at intervals following the induction of electroshock seizures, as well as after a subconvulsive series of electric shocks. The effect of the prior administration of phenytoin or phenobarbital was also studied. METHODSSeizures were induced through electrodes applied to the shaved scalp just anterior to the ears. A 60-Hz square-wave current at 100 V and with a pulse duration of 5 msec was applied for 5 sec. The sub-

No abstract