Description of the neuronal and synaptic bases of memory in the vertebrate central nervous system (CNS) requires a CNS stimulus-response pathway that is defined and accessible, has the capacity for adaptive change, and clearly contains the responsible substrates. This study was an attempt to determine whether the spinal stretch reflex (SSR), the initial, purely spinal, portion of the muscle stretch response, which satisfies the first requirement, also satisfies the second, capacity for adaptive change. Monkeys prepared with chronic fine-wire biceps electromyographic (EMG) electrodes were trained to maintain elbow position and a given level of biceps background EMG activity against constant extension torque. At random times, a brief additional extension torque pulse extended the elbow and elicited the biceps SSR. Under the control mode, reward always followed. Under the SSR increases or SSR decreases mode, reward followed only if the absolute value of biceps EMG from 14 to 24 ms after stretch onset (the SSR interval) was above or below a set value. Animals performed 3,000-6,000 trials/day over data-collection periods of up to 15 mo. Background EMG and the initial 30 ms of pulse-induced extension remained stable throughout data collection. Under the SSR increases or SSR decreases mode, SSR amplitude (EMG amplitude in the SSR interval minus background EMG amplitude) changed appropriately. Change was evident in 5-10 days and progressed over at least 4 wk. The SSR increased (SSR increases) to 140-190% control amplitude or decreased (SSR decreases) to 22-79%. SSR change did not regress over 12-day gaps in task performance. A second pair of biceps electrodes, monitored simultaneously, supplied comparable data, indicating that SSR amplitude change occurred throughout the muscle. Beyond 40 ms after pulse onset, elbow extension was inversely correlated with SSR amplitude. The delay between the SSR and its apparent effect on movement is consistent with expected motor-unit contraction time. The data demonstrate that the SSR is capable of adaptive change. At present the most likely site(s) of the mechanism of SSR amplitude change are the Ia synapse and/or the muscle spindle. Available related evidence suggests persistent segmental change may in fact come to mediate SSR amplitude change. If so, such segmental change would constitute a technically accessible fragment of a memory.
Adaptive plasticity in the primate spinal stretch reflex: evidence for a two-phase process