Electrophysiological studies in cultured mouse CNS neurones of the actions of an agonist and an inverse agonist at the benzodiazepine receptor.

The action of agents which bind with the benzodiazepine (BZ) receptor has been investigated by use of intracellular recordings from dissociated mouse neurones grown in tissue culture. The agents tested were midazolam (an agonist at the BZ receptor) and methyl 6,7-dimethoxy-4-ethyl-beta-carboline-3-carboxylate (DMCM-an inverse agonist at the BZ receptor). These were applied to the neurone under study by one of the following methods: iontophoresis; pressure application of known concentrations from blunt pipettes; directly in the perfusing medium. On only very few occasions did midazolam or DMCM have a direct effect on the membrane potential (EM) or conductance (GM) of the impaled neurone. For the neurones where direct effects were present, there was no consistent pattern of response. Neither substance affected the threshold for action potential generation. The effect of midazolam and DMCM on responses evoked by iontophoretic application of gamma-aminobutyric acid (GABA) was also investigated. Three parameters were used to quantify GABA responses: the depolarization (VGABA); the increase in GM (gGABA) measured with constant current pulses; using voltage clamp, the GABA current (IGABA). The GABA response should be quantified by a parameter which is linearly related to the number of GABA-operated channels which are conducting at any instant. VGABA does not fulfil this criterion. gGABA is an appropriate parameter, but is difficult to determine for large responses where the membrane is nearly short circuited. IGABA measured during voltage clamp fulfils this criterion. Midazolam (greater than 10(-6) M) reliably potentiated GABA responses with a parallel shift to the left of the dose-response curve. This is in agreement with biochemical studies where BZs increase the affinity of the GABA receptor for its ligand. The effect of DMCM on GABA responses was more variable. In the majority of cases GABA responses were reduced by DMCM. The threshold dose for this depression was usually around 10(-6) M, but was sometimes as low as 10(-8) M. Dose-response curves of IGABA or gGABA showed the inhibition to be of a non-competitive nature. The maximum inhibition achieved was around 70%. For a given neurone, and at doses which did not necessarily cause a reduction of the response to GABA, DMCM could antagonize the potentiating action of midazolam on GABA responses. A possible interpretation is that more than one BZ site per receptor complex must be occupied by a BZ agonist (or inverse agonist) before the functional changes for the complex as a whole can occur. Desensitization to GABA was increased by midazolam.