Spike timing and synaptic dynamics at the awake thalamocortical synapse

Thalamocortical (TC) neurons form only a small percentage of the synapses onto neurons of cortical layer 4, but the response properties of these cortical neurons are arguably dominated by thalamic input. This discrepancy is explained, in part, by studies showing that TC synapses are of high efficacy. However, TC synapses display activity-dependent depression. Because of this, in vitro measures of synaptic efficacy will not reflect the situation in vivo, where different neuronal populations have widely varying levels of "spontaneous" activity. Indeed, TC neurons of awake subjects generate high rates of spontaneous activity that would be expected, in a depressing synapse, to result in a chronic state of synaptic depression. Here, we review recent work in the somatosensory thalamocortical system of awake rabbits in which the relationship between TC spike timing and TC synaptic efficacy was examined during both thalamic "relay mode" (alert state) and "burst mode" (drowsy state). Two largely independent methodological approaches were used. First, we employed cross-correlation methods to examine the synaptic impact of single TC "barreloid" neurons on a single neuronal subtype in the topographically aligned layer 4 "barrel" - putative fast-spike inhibitory interneurons. We found that the initial spike of a TC burst, as well as isolated TC spikes with long preceding interspike intervals (ISIs) elicited postsynaptic action potentials far more effectively than did TC impulses with short ISIs. Our second approach took a broader view of the postsynaptic impact of TC impulses. In these experiments we examined spike-triggered extracellular field potentials and synaptic currents (using current source-density analysis) generated through the depths of a cortical barrel column by the impulses of single topographically aligned TC neurons. We found that (a) closely neighboring TC neurons may elicit very different patterns of monosynaptic activation within layers 4 and 6 of the aligned column, (b) synaptic currents elicited by TC impulses with long preceding ISIs were greatly enhanced in both of these layers, and (c) the degree of this enhancement differed reliably among neighboring TC neurons but, for a given neuron, was very similar in layers 4 and 6. Thus, results generated by both methodological approaches are consistent with the presence of a chronic depression at the awake TC synapse that is relieved by long ISIs. Since long ISIs necessarily precede TC "bursts", our results are consistent with the notion that these events powerfully activate cortical circuits.