TEMPORALLY DISTINCT MECHANISMS OF USE-DEPENDENT DEPRESSION AT INHIBITORY SYNAPSES IN THE RAT HIPPOCAMPUS IN-VITRO

1. gamma-aminobuturic acid-B (GABA(B)) autoreceptor-dependent and -ind ependent components of paired-pulse depression (PPD) at inhibitory syn apses in area CA3 of the rat hippocampus were studied using whole-cell recording techniques. Inhibitory fibers were activated directly in th e presence of the ionotropic glutamate receptor antagonists 6,7-dinitr oquinoxaline-2,3-dione (20 mu M) and D-2-amino-5-phosphonovalerate (20 mu M). 2. When pairs of monosynaptic inhibitory postsynaptic currents (eIPSCs) were evoked with an interstimulus interval of 200 ms, the am plitude of the second response (eIPSC(2)) was depressed when compared with the first (eIPSC(1)). The GABA(B) receptor agonist baclofen (10 m u M) depressed both responses, but eIPSC(1) was depressed more than eI PSC2, resulting in PPD that was comparatively smaller. Addition of the GABA(B) receptor antagonist CGP 55845A(1 mu M) completely reversed de pression of eIPSC(1) by baclofen and increased the amplitude of eIPSC( 2) above the control value, such that PPD in the combination of baclof en and CGP 55845A was equivalent to that in baclofen alone. The ratio eIPSC(2)/eIPSC(1) was 0.64 under control conditions, 0.77 in the prese nce of baclofen, and 0.79 in the presence of baclofen and CGP 55845A. These results demonstrate the existence of two components of PPD at in hibitory synapses, one that depends on activation of GABA(B) autorecep tors (GABA(B) receptor-dependent PPD) and one that does not (GABA(B) r eceptor-independent PPD). 3. When the number of inhibitory fibers acti vated was lowered by decreasing the stimulus intensity, eIPSC(2)/eIPSC (1) was 0.76 under control conditions, 0.75 in the presence of baclofe n, and 0.76 in the presence of baclofen and CGP 55845A. These results indicate that GABA(B) receptor-dependent PPD requires activation of se veral presynaptic inhibitory neurons, whereas GABA(B) receptor-indepen dent PPD does not. 4. The timecourses of the GABA(B)-dependent and -in dependent components of PPD were compared by varying the interstimulus interval in the absence and presence of CGP 55845A. GABA(B)-dependent PPD was maximal at an interstimulus interval of 100 ms and was undect ectable at 1 s. In contrast, GABA(B)-independent PPD was maximal at 5 ms and 1 s, was slightly less pronounced at intermediate intervals (50 -200 ms), and was present at intervals as long as 5 s. 5. GABA(B)-inde pendent PPD was not blocked by antagonists at opioid receptors(10 mu M naloxone) or muscarinic acetylcholine receptors(10 mu M atropine). GA BA(B)-independent PPD could not be accounted for by a decrease in driv ing force because of Cl- redistribution. 6. When the probability of ne urotransmitter release was lowered by increasing the extracellular con centration of Mg2+ from 1 mM to 5-8 mM, GABA(B)-independent PPD at int ermediate interstimulus intervals (20-200 ms) was diminished or revers ed into paired-pulse facilitation (PPF). GABA(B)-independent PPD at in terstimulus intervals of 500 ms or greater was not changed by increasi ng Mg2+. 7. Postsynaptic currents could be elicited in an all-or-none manner at a threshold intensity when the stimulus intensity was increa sed slowly from 1 mu A to similar to 20 mu A. These minimal evoked inh ibitory postsynaptic currents (meIPSCs) presumably resulted from the a ctivation of single inhibitory fibers. GABA(B)-independent PPD of meIP SCs was similar to that of eIPSCs; meIPSC(2)/ meIPSC(1) was 0.73 at an interstimulus interval of 100 ms. PPD was observed of meIPSCs that we re unaffected by baclofen or CGP 55845A. GABA(B) independent PPD did n ot change the rise or decay kinetics of meIPSCs, but was accompanied b y an increase in the variability of meIPSC(2). These results indicate that GABA(B)-independent PPD occurs at synapses made by single inhibit ory neurons.