SYNAPTIC INTERACTIONS IN PRIMATE MOTOR CORTEX: RELATIONS BETWEEN CONNECTIVITY AND INTRACORTICAL LOCATION
To investigate intracortical microcircuits within primate motor cortical areas, we documented pairs of neurons whose synaptic interactions were identified by spike-triggered averaging (STA) of intracellularly recorded membrane potentials and whose relative location was histologically identified. Average synchronous excitation potentials (ASEPs) were the most commonly observed feature in STAs between neuron pairs in all cortical layers (about 70%). This synchrony spread broadly for distances of more than 4 mm, gradually decreasing in amplitude and probability with cell separation. Excitatory postsynaptic potentials (EPSPs) were observed among 9% of the neuron pairs, whose separation extended for distances of more than 4 mm. The peak amplitudes of EPSPs were inversely correlated with cell separation. Most source neurons were located in layer II-III or layer V, while the postsynaptic target neurons had wide laminar distribution. The probability of finding excitatory connections was not uniform within the cortical space. The connectivity between neuron pairs was relatively dense within 1.0 mm, and became sparse for distances between 1.0 and 2.0 mm, and showed a second peak for distance between 2.0 and 4.0 mm. Inhibitory postsynaptic potentials (IPSPs) and/or average synchronous inhibitory potentials (ASIPs) were observed for 10% of neuron pairs. Most of these were separated by less than 1.0 mm, suggesting that their influence was restricted within their own and neighboring columns. The major source neurons that provide these inhibitory effects were located within layer II-III. We conclude that excitatory synaptic effects (ASEPs and EPSPs) are widely distributed and omni-directional within primate motor cortex, while inhibitory connections (ASIPs and IPSPs) are restricted within columnar dimensions, and are predominantly directed from superficial to deeper layers. The appearance of independent zones that had no functional connectivity with neighboring columns indicates that primate motor cortical areas may not be organized in a uniform way. This would support the sparse coding theory and multiple representations of cortical output in the motor cortex.