1. The laminar distribution of synaptic activity in the primary motor cortex, elicited by stimulation of intracortical, horizontal afferents, was studied in young (12-17 days old) and adult rats using the in vitro brain slice preparation. Connectivity patterns were deduced from current-source density (CSD) analyses of field potential depth profiles and were confirmed by anatomic data of retrograde cell labeling after focal injections of a fluorescent tracer. 2. According to the CSD distributions, horizontal axons in layer II/III provide strong monosynaptic input to dendrites of layer II and III pyramidal cells in a distant column, and weaker monosynaptic input to layer V and VI cells by synapsing on dendritic fields at the border of layer III and V and in deep layer V. When these pathways are activated, layer II/III cells may relay excitatory activity to upper and deep layer V, as well as to other cells in layer II/III of the same column. Axons arising from layer V provide monosynaptic input to pyramidal cells in all layers of neighboring columns, by synapsing in two dendritic fields: one in the superficial layers and the other in middle layer V. Activation of these pathways may generate a disynaptic intracolumnar input from layer II/III cells to middle layer V, as well as to other cells in layer II/III. Similar patterns of synaptic activity were elicited by stimulation from 0.45 to 2 mm distal to the recorded column. There were no apparent differences between young and adult rats in the connectivity patterns revealed by the CSD analyses. 3. Tracer injections in layer III resulted in retrograde labeling of cells in layers II/III and V, at distances >2 mm from the injection site, whereas injections in layer V resulted in retrograde labeling of cells at long distances in layer V and to a lesser extent in layer II/III. These findings indicate that neurons in layer V project, via horizontal axon collaterals, for long distances within layers III and V, whereas the horizontal axon collaterals of layer III cells are restricted, for the most part, to the superficial layers. 4. Suppression of inhibitory activity by bath application of the gamma-aminobutyric acid-A (GABA(A)) receptor antagonist bicuculline methiodide (BMI) did not alter the pattern of the CSD distributions. All synaptic currents present in the control medium were enhanced by application of BMI, although the effect was more pronounced on the polysynaptic components. In young rats, the effects of BMI were completely reversed by bath application of the N-methyl-D-aspartate (NMDA) receptor antagonist D,L-2-amino-phosphonovalerate (APV), while in adult rats APV only reduced the currents revealed by application of BMI. 5. In both young and adult rats, blockade of non-NMDA receptors by bath application of 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) revealed CNQX-resistant currents that were blocked by bath application of APV. These NMDA receptor-mediated currents were more pronounced in the superficial layers when horizontal afferents in layer II/III were stimulated, whereas they were more pronounced in layer V in response to the horizontal input from layer V. 6. These findings suggest that neurons within different representation zones in the motor cortex are connected via long, horizontal and oblique excitatory connections. The specific patterns of synaptic inputs mediated by these intrinsic connections reflect, for the most part, the anatomic arrangement of excitatory connections. The role of inhibition is primarily to limit postsynaptic responses to the more powerful inputs within organized excitatory circuits, rather than to specify the spatiotemporal pattern of activity in a network of nonspecific excitatory connections. The intracortical synaptic interactions revealed in this study may be involved in the spatiotemporal coordination of muscles during voluntary movement and in synaptic plasticity in the motor cortex.
