SYNAPTIC PROLIFERATION IN THE MOTOR CORTEX OF ADULT CATS AFTER LONG-TERM THALAMIC-STIMULATION

1. One of the hypotheses for information storage in the CNS postulates the induction of structural changes in synaptic circuits. This postulate predicts that behavioral experiences produce changes in neural activity that subsequently induce synaptogenesis in the mature CNS. Available data indicate that the establishment of engrams for novel motor acts may involve alterations of synaptic interactions within the primary motor cortex. The present study examines the hypothesis that patterns of synaptic circuitry and of synaptic activation are rearranged after enhanced neural activity in pathways projecting to the motor cortex. 2. Electrodes implanted in the ventroposterolateral (VPL) nucleus of the thalamus were used for long-term stimulation (20-mu-A, 4 days) of afferents to the motor cortex in freely behaving, adult cats. This stimulation primarily affected corticocortical inputs from the somatosensory cortex (area 2) to area 4-gamma of the motor cortex. Electron microscopy and stereological procedures were used to compare the numerical density (N(upsilon)) of various types of synapses in layers II/III of the stimulated (experimental) motor cortex with the N(upsilon) of the corresponding synapses in the contralateral (control) hemisphere. 3. Long-term stimulation produced a significant increase (25.6%) in synaptic N(upsilon) in experimental motor cortex. This increase was due primarily to an increase in the N(upsilon) of asymmetrical synapses with dendritic spines. The numbers of symmetrical synapses, and of asymmetrical synapses with dendritic shafts, were not affected by long-term stimulation. 4. Synaptic active zones [calculated by measuring the lengths of postsynaptic densities (PSDs)] were significantly longer in experimental motor cortex. Lengthening of PSDs occurred selectively in asymmetrical synapses with dendritic shafts (28% increase). 5. The N(upsilon) of synapses having perforations in their PSDs perforated synapses) was significantly higher in experimental hemispheres. Also increased was the incidence of synapse-associated polyribosomes, which are most commonly found at the base of dendritic spines. An increase in the number of perforated synapses and of polyribosomes are both morphological hallmarks of synaptogenesis. 6. The percentages of synapses having different curvatures (i.e., presynaptically concave, convex, or flat) were similar in experimental and in control motor cortex. This indicates that long-term stimulation did not produce an interconversion of synapses from one morphological class to another. 7. Electrophysiological recordings and current source-density (CSD) analyses of field potentials in the motor cortex induced by microstimulation in the somatosensory cortex revealed that long-term stimulation resulted in the appearance of a new component in these field potentials. In experimental motor cortex, corticocortical stimulation resulted in the appearance of a short-latency current sink in layers II-III. In control motor cortex, these field potentials were purely monophasic, and the CSD revealed only a relatively slow, late current sink in layer V. These changes in the CSD can be attributed to a proliferation of synapses belonging to corticocortical afferents to the motor cortex. 8. The findings indicate that enhanced neural activity induces synaptic proliferation in the motor cortex. This synaptogenesis was evidenced by an increase in the N(upsilon) of specific classes of synapses, by an increase in synaptic structural features that are indicative of synaptogenesis, and by alterations in the patterns of synaptic activation. We propose that this use-dependent synaptic proliferation may participate in processes of motor learning and memory.