We analyzed how the GABA, receptor agonist baclofen (10-50 mu M) influences
the activity induced by 4-aminopyridine (4-AP, 50 mu M) in the CA3 area of
hippocampal slices obtained from 12- to 25-day-old rats. Interictal and ic
tal discharges along with synchronous GABA-mediated potentials occurred spo
ntaneously in the presence of 4-AP. Baclofen abolished interictal activity
(n = 29 slices) and either disclosed (n = 21/29) or prolonged ictal dischar
ges (n = 8/29), whereas GABA-mediated potentials occurred at a decreased ra
te. The N-methyl-D-aspartate (NMDA) receptor antagonist 3,3-(2-carboxypiper
azine-4-yl)-propyl-1-phosphate (CPP, 10 mu M, n = 8) did not modify the GAB
A-mediated potentials or the ictal events recorded in 4-AP + baclofen. In c
ontrast ictal, activity, but not GABA-mediated potentials, was blocked by t
he non-NMDA receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX,
10 mu M, n = 5). Most baclofen effects were reversed by the GABA, receptor
antagonist CGP 35348 (1 mM; n = 4). Baseline and transient increases in [K
+](0) associated with the 4-AP-induced synchronous activity were unaffected
by baclofen. Baclofen hyperpolarized CA3 pyramids (n = 8) recorded with K-
acetate-filled electrodes by 4.8 +/- 1.3 mV and made spontaneous, asynchron
ous hyperpolarizing and depolarizing potentials disappear along with interi
ctal depolarizations. GABA-mediated synchronous long-lasting depolarization
s (LLDs) and asynchronous depolarizations were also studied with KCl-filled
electrodes in 4-AP + CPP + CNQX (n = 6); under these conditions baclofen d
id not reduce LLD amplitude but abolished the asynchronous events. Dentate
hilus stimulation at 0.2-0.8 Hz suppressed the ictal activity recorded in 4
-AP + baclofen (n = 8). Our data indicate that GABA, receptor activation by
baclofen decreases transmitter release leading to disappearance of interic
tal activity along with asynchronous excitatory and inhibitory potentials.
By contrast, GABA-mediated LLDs and ictal events, which reflect intense act
ion potential firing invading presynaptic inhibitory and excitatory termina
ls respectively, are nor abolished. We propose that the proconvulsant actio
n of baclofen results from I) block of asynchronous GABA-mediated potential
s causing disinhibition and 2) activity-dependent changes in hippocampal ne
twork excitability.