MICROSTIMULATION OF THE MEDULLARY RETICULAR-FORMATION DURING FICTIVE LOCOMOTION

1. The present study was designed to determine the effects of microstimulation of the medullary reticular formation (MRF) on the locomotor activity of the cat in the absence of phasic afferent feedback from the limbs. To this end, both short(33 ms) and long (200 ms) trains of stimuli (trains of 0.2-ms pulses at 330 Hz, 35 mu A) were applied at 43 loci in the MRF (P:6-12 mm; L:0.5-1.5 mm), and in 3 loci in the medial longitudinal fasciculus(P7.5, L < 0.5 mm) during fictive locomotion in the decerebrate and paralyzed cat. The locomotor pattern was monitored by recording the activity of representative flexor and extensor muscle nerves from each of the four limbs. 2. Short trains of stimuli evoked transient excitatory and/or inhibitory responses in extensor and flexor nerves of each limb that were incorporated into the locomotor pattern. In the majority of sites, excitatory responses were obtained in the motor nerves to both flexor and extensor muscles of the fore- and hindlimbs. The exception to this rule was the ipsilateral triceps, in which the predominant response was inhibitory. The amplitude of these responses was dependent on the time of the locomotor cycle at which the stimulus was delivered, and it was always maximum during the period of activity of the respective nerve. 3. The shortest latency response in the nerves to different muscles of the forelimb averaged between 5.6 and 7.3 ms; for the hindlimbs the values were between 6.9 and 9.3 ms. 4. Changing the depth at which the stimulation was applied in any one trajectory usually produced changes only in the amplitude of the evoked responses but occasionally also caused a change in the sign of these responses, especially in the most ventral regions of the MRF. 5. At 72% of the loci (31/43), short trains of stimulation also changed the duration of the activity in the recorded nerves. These changes were often (20/31 loci) sufficiently strong to alter the duration of the overall locomotor cycle. If one considers only the largest changes produced at each locus, stimulation during the period of ipsilateral extensor activity produced an average reduction in the ipsilateral locomotor cycle duration of 12.8 +/- 8.8% (mean +/- SD), whereas stimulation when the ipsilateral flexor nerve was active produced an average increase in locomotor cycle duration of 27.1 +/- 20.8%. 6. Long trains of stimuli produced similar but larger effects than the shorter trains and always reset the locomotor rhythm. Changes in the locomotor pattern were observed at all 41 sites stimulated. At 11 of these loci, stimulation arrested the locomotor pattern; at the other 30 it produced either a reduction or an increase in the locomotor cycle duration depending on the time in the locomotor cycle at which the stimulation was given. At 19/30 loci, stimulation during the activity of the ipsilateral extensors reduced the cycle duration (mean 25.1 +/- 16.0%); at 6/30 loci it increased the cycle duration, whereas at the other 5/30 it was without effect. Stimulation during the period of activity of the ipsilateral flexors produced an increase in cycle duration at all 30 loci (mean 38.6 +/- 24.2%). 7. The results show that stimulation of the MRF during locomotion, in the absence of phasic peripheral afferent input, changes both the amplitude and the duration of nerve activity in a phase-dependent manner. It is suggested that these effects are mediated through spinal interneuronal networks, including those that are influenced by, or form part of, the central pattern generator for locomotion. It is further suggested that the relative ease with which large changes in cycle duration could be produced is, at least in part, due both to the complete absence of rhythmical feedback from the periphery, as well as to abnormal or absent phasic information from brain stem and cortical inputs. The results from these studies are discussed with respect to similar studies in the intact and decerebrate walking preparation.