1. Extracellular action potentials were recorded from antidromically activated efferent neurons in visual, somatosensory, and motor cortex of the awake rabbit using low-impedance metal microelectrodes. Efferent neurons were also activated by current pulses delivered near the soma [juxtasomal current pulses (JSCPs)] through the recording microelectrode. Action potentials generated by JSCPs were not directly observed (because of the stimulus artifact), but were inferred with the use of a collision paradigm. Efferent populations studied include callosal neurons [CC (n = 80)], ipsilateral corticocortical neurons [C-IC (n = 21)], corticothalamic neurons of layer 6 [CF-6 (n = 57)], and descending corticofugal neurons of layer 5 [CF-5, corticotectal neurons of the visual cortex (n = 48)]. 2. Most CC neurons (45/46) and all C-IC (8/8) and CF-6 neurons (39/39) were directly activated by JSCPs at near-threshold intensities. Some CF-5 neurons (9/38), however, showed evidence of indirect activation. All efferent classes had similar current thresholds (means 1.85-2.10 microA) to direct activation by JSCPs, and thresholds were inversely related to extracellular spike amplitude. For each neuron, the range of JSCP intensities that generated response probabilities of between 0.2 and 0.8 was measured, and this "range of uncertainty" was significantly greater in CF-5 neurons (mean 32.7% of threshold) than in CC (mean 19.0%) or CF-6 (mean 20.4%) neurons. 3. Several factors indicate that the threshold of efferent neurons to JSCPs is very sensitive to excitatory and inhibitory synaptic inputs. Iontophoretic applications of gamma-aminobutyric acid (GABA) increased the threshold to JSCPs, and glutamate reduced the threshold. Electrical stimulation of afferent pathways at intensities just below threshold for eliciting action potentials resulted in a dramatic decrease in JSCP threshold. This initial short-latency threshold decrease was specific to stimulation of particular afferent pathways and is thought to reflect excitability changes associated with EPSPs. Examination of such subliminal responses revealed subthreshold synaptic inputs that were not revealed by examination of all-or-none action potentials. In contrast to the specificity of the short-latency threshold decrease, a long-lasting increase in JSCP threshold was seen in virtually all neurons after stimulation of each of the afferent pathways tested. This increase in threshold usually began 20-40 ms after stimulation, lasted for 100-200 ms, and is thought to reflect excitability changes associated with a long-lasting inhibitory postsynaptic potential (IPSP) seen in many cortical neurons. 4. Many neurons in primary somatosensory cortex of rat, cat, and rabbit have no demonstrable receptive fields.(ABSTRACT TRUNCATED AT 400 WORDS)
