1. Standard intracellular recordings from CA1 pyramidal neurones in in vitro hippocampal slices have been used to investigate the effects of excitatory amino acid antagonists and adrenergic agents on evoked synaptic potentials. 2. Ortho- and antidromic stimulation were conducted with remotely placed electrodes in order to minimize the possibility of stimulating the interneurones directly. In addition to the excitatory postsynaptic potential (EPSP), orthodromic stimulation evoked an inhibitory sequence consisting of a fast and slow inhibitory postsynaptic potential (IPSP). The slow-IPSP was blocked by intracellular injection of QX 314. Antidromic stimulation evoked a relatively pure fast-IPSP. 3. In seven neurones the differential effects of glutamatergic receptor blockers on the fast-IPSP were investigated. The N-methyl-D-aspartate (NMDA) receptor blocker, DL-2-amino-5-phosphonovaleric acid (APV) was added after the full effect of the non-NMDA receptor blocker, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) had been achieved. In three neurones, APV had no additional blocking effect, while in the remaining four neurones, both the ortho- and antidromically evoked IPSPs were reduced by 20-50%. This suggest that NMDA receptors participate in the activation of some GABAergic interneurones, which was further confirmed by showing that the IPSP was enhanced by Mg2+-free medium. 4. In the presence of CNQX (10-mu-M) and APV (50-mu-M) together, the ortho- and antidromically evoked fast-IPSPs were greatly reduced. A small 'residual' IPSP remained which was best studied by depolarizing the neurone to around -50 mV. With maximum stimulation, this amounted to 26.3 +/- 15.4% (mean +/- S.E.M., n = 15) of the control IPSP evoked by orthodromic stimulation and 41 +/- 14.6% of the control IPSP evoked by antidromic stimulation. The following statements apply equally to the ortho- and antidromically activated residual IPSPs. 5. The residual IPSP was completely blocked by low concentrations of bicuculline, indicating that it is mediated by GABA(A) receptors. When compared with a control IPSP of similar amplitude, the residual IPSP was found to have a faster rise time and time-to-peak, but a similar decay time. 6. Neither the muscarinic cholinergic antagonist, atropine nor the presynaptic glutamate agonist, L-2-amino-4-phosphonobutyric acid (L-APB) had any effect on the residual IPSP. 7. The residual IPSP was completely blocked by the adrenergic beta-receptor antagonist, L-propranolol (50-100-mu-M). A similar blocking effect was obtained with the specific beta-1-receptor antagonist, atenolol (20-mu-M) whereas the beta-2-receptor antagonist erythro-dl-1-(7-methylindan-4-yloxy)-3-isopropylaminobutan-2-ol (ICI 118.551; 20-mu-M) had no effect. 8. Atenolol reduced the control fast-IPSP in a dose-dependent manner with a maximal effect obtained at 10-mu-M. When paired-pulse stimulation was used, atenolol reduced the conditioning and test IPSPs to a similar extent. Atenolol was without effect on the IPSP evoked by direct stimulation of the interneurones. 9. At a concentration which had no direct effect on the pyramidal neurone, the noradrenergic alpha-2-receptor agonist, clonidine (4-mu-M) reduced the control fast-IPSP by about 20%. This action is most likely mediated by operation of presynaptic alpha-2-receptors, which decrease the release of noradrenaline. 10. Possible locations of the beta-1-receptor have been considered. The most likely explanation is that beta-1-receptors mediate a fast excitation of feedback and feedforward GABAergic interneurones. The time course of this action is, however, much more rapid than would be expected for an event which is mediated by an increase in cyclic AMP, which is the established mechanism of beta-1-receptor-mediated effects. 11. In conclusion, the results provide evidence that the presynaptic mechanisms underlying feed-forward and feedback inhibition in area CA1 consist of two parts: a glutamatergic component and a noradrenergic component mediated by beta-1-receptors.
