The glutamate-mediated synaptic responses of neocortical pyramidal cell to fast-spiking interneuron (pyramidal-FS) connections were studied by performing paired recordings at 30-33 degrees C in acute slices of 14- to 35-day-old rats (n = 39). Postsynaptic fast-spiking (FS) cells were recorded in whole cell configuration with a patch pipette, and presynaptic pyramidal cells were impaled with sharp intracellular electrodes. At a holding potential of -72 mV (near the resting membrane potential), unitary excitatory postsynaptic potentials (EPSPs) had a mean amplitude of 2.1 +/- 1.3 mV and a mean width at half-amplitude of 10.5 +/- 3.7 ms (n = 18). Bath application of the N-methyl-D-aspartate (NMDA) receptor antagonist D(-)2-amino-5-phosphonovaleric acid (D-AP5) had minor effects on both the amplitude and the duration of unitary EPSPs, whereas the alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionate (AMPA)/kainate receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) almost completely blocked the synaptic responses. In voltage-clamp mode, the selective antagonist of AMPA receptors 1-(4-aminophenyl)-3-methylcarbamyl-4-methyl-7,8-methylenedioxy-3, 4-dihydro-5H-2,3-benzodiazepine (GYKI 53655; 40-66 microM) blocked 96 +/- 1.9% of D-AP5-insensitive unitary excitatory postsynaptic currents (EPSCs), confirming the predominance of AMPA receptors, as opposed to kainate receptors, at pyramidal-FS connections (n = 3). Unitary EPSCs mediated by AMPA receptors had fast rise times (0.29 +/- 0.04 ms) and amplitude-weighted decay time constants (2 +/- 0.8 ms; n = 16). In the presence of intracellular spermine, these currents showed the characteristic rectifying current-voltage (I-V) curve of calcium-permeable AMPA receptors. A slower component mediated by NMDA receptors was observed when unitary synaptic currents were recorded at a membrane potential more positive than -50 mV. In response to short trains of moderately high-frequency (67 Hz) presynaptic action potentials, we observed only a limited temporal summation of unitary EPSPs, probably because of the rapid kinetics of AMPA receptors and the absence of NMDA component in these subthreshold synaptic responses. By combining paired recordings with extracellular stimulations (n = 11), we demonstrated that EPSPs elicited by two different inputs were summed linearly by FS interneurons at membrane potentials below the action potential threshold. We estimated that, in our in vitro recording conditions, 8 +/- 5 pyramidal cells (n = 18) should be activated simultaneously to make FS interneurons fire an action potential from -72 mV. The low level of temporal summation and the linear summation of excitatory inputs in FS cells favor the role of coincidence detectors of these interneurons in neocortical circuits.