Coordination of locomotor activity in the lamprey: role of descending drive to oscillators along the spinal cord.

In the lamprey and most fish, locomotion is characterized by caudally propagating body undulations that result from a rostrocaudal phase lag for ipsilateral burst activity. One of the mechanisms that might contribute to rostrocaudal phase lags is a gradient of oscillator burst frequencies along the spinal cord that presumably are produced, in part, by descending drive from the brain. The purpose of the present study was to test whether a gradient of oscillator frequencies does exist along the lamprey spinal cord. First, during brain-initiated locomotor activity in in vitro brain/spinal cord preparations, the cycle times (=1/frequency) of locomotor activity generated by the functionally isolated rostral spinal cord (activity blocked in middle and caudal cord) were significantly shorter than control cycle times when the entire spinal cord was generating locomotor activity. Second, the cycle times of locomotor activity generated by the functionally isolated caudal cord (activity blocked in rostral and middle cord) were significantly longer than control cycle times for activity generated by the entire spinal cord. Thus, no one region of the spinal cord appears to dictate the overall cycle times of locomotor activity generated by the entire spinal cord, although overall cycle times tended to be closest to those of the isolated rostral spinal cord. Finally, although short- and long-distance coupling as well as oscillator frequency gradients probably contribute to rostrocaudal phase lags of spinal locomotor activity, the asymmetrical nature of short-distance coupling, in which the descending component dominates, appears to be the main factor.

Coordination of spinal locomotor activity in the lamprey: long-distance coupling of spinal oscillators.

The extent and strength of long-distance coupling between locomotor networks in the rostral and caudal spinal cord of larval lamprey were examined with in vitro brain/spinal cord preparations, in which spinal locomotor activity was initiated by chemical microstimulation in the brain, as well as with computer modeling. When locomotor activity and short-distance coupling were blocked in the middle spinal cord for at least 40 segments, burst activity in the rostral and caudal spinal cord was still coupled 1:1, indicating that long-distance coupling is extensive. However, in the absence of short-distance coupling, intersegmental phase lags were not constant but decreased significantly with increasing cycle times, suggesting that long-distance coupling maintains a relatively constant delay rather than a constant phase lag between rostral and caudal bursts. In addition, under these conditions, intersegmental phase lags, measured between rostral and caudal burst activity, were significantly less than normal, and the decrease was greater for longer distances between rostral and caudal locomotor networks. The above result could be mimicked by a computer model consisting of pairs of oscillators in the rostral, middle, and caudal spinal cord that were connected by short- and long-distance coupling. With short-distance coupling blocked in the middle spinal cord, strychnine was applied to either the rostral or caudal spinal cord to convert the pattern locally from right-left alternation to synchronous burst activity. Synchronous burst activity in the rostral spinal cord resulted in a reduction in right-left phase values for burst activity in the caudal cord. These results also could be mimicked by the computer model. Strychnine-induced synchronous burst activity in the caudal spinal cord did not appear to alter the right-left phase values of rostral burst activity. Taken together, the experimental and modeling results suggest that the descending and ascending components of long-distance coupling, although producing qualitatively different effects, are comparatively weak. In particular, the descending component of long-distance coupling appears to become progressively weaker with increasing distance between two given regions of spinal cord. Therefore, short-distance coupling probably contributes substantially to normal rostrocaudal phase lags for locomotor activity along the spinal cord. However, short-distance coupling may be more extensive than "nearest neighbor coupling."

Descending control of turning locomotor activity in larval lamprey: neurophysiology and computer modeling.

The purpose of the present study was to examine the mechanisms that produce natural spontaneous turning maneuvers in larval lamprey. During swimming, spontaneous turning movements began with a larger-than-normal bending of the head to one side. Subsequently, undulations propagated down the body with greater amplitude on the side ipsilateral to the turn. During turning to one side, which usually occurred within one cycle, the amplitude and duration of ipsilateral muscle burst activity as well as overall cycle time increased significantly with increasing turn angle. In in vitro brain/spinal cord preparations, brief electrical stimulation applied to the left side of the oral hood at the onset of locomotor burst activity on the right side of the spinal cord produced turninglike motor activity. During the perturbed cycle, the duration and amplitude of the burst on the right as well as cycle time were significantly larger than during preceding control cycles. In several lower vertebrates, unilateral stimulation in brain stem locomotor regions elicits asymmetric, turninglike locomotor activity. In the lamprey, unilateral chemical microstimulation in brain stem locomotor regions elicited continuous asymmetric locomotor activity, but there was little change in cycle time, as occurs during the single turning cycles in whole animals. The descending mechanisms responsible for producing turning locomotor activity were examined with the use of a computer model consisting of left and right phase oscillators in the spinal cord that were coupled by net reciprocal inhibition. With relatively weak reciprocal coupling, a brief unilateral descending excitatory input to one oscillator produced effects ipsilaterally, but there was little effect on the contralateral oscillator. Turninglike patterns could be produced by each of the following modifications of the model: 1) unilateral descending input and relatively strong reciprocal coupling; 2) unilateral descending input that phase shifted as well as increased the amplitude of the waveform generated by an oscillator on one side; and 3) brief descending modulatory inputs that excited the oscillator on one side and inhibited the contralateral oscillator. In all three cases, there was an increase in "burst" duration ipsilateral to the excitatory input and an increase in cycle time, similar to turning locomotor activity in whole animals. It is likely that turning maneuvers are mediated by descending modulatory inputs primarily to the spinal oscillator networks, which control the timing of burst activity, but perhaps also to motoneurons for axial musculature.

Coupling of spinal locomotor networks in larval lamprey revealed by receptor blockers for inhibitory amino acids: neurophysiology and computer modeling.

1. Receptor blockers for inhibitory amino acids were applied to part or all of the spinal cord of larval lamprey during brain stem-initiated locomotor activity. Blocking glycinergic inhibition with strychnine applied to the entire spinal cord converted the locomotor pattern from left-right alternation to synchronous left-right bursting. The results suggest that left and right oscillators are connected by relatively strong reciprocal inhibitory (glycinergic) connections in parallel with weaker reciprocal excitatory connections. This possible organization was supported by results from a computer model consisting of left and right oscillators connected by reciprocal inhibition and excitation in parallel. In addition, the results suggest that reciprocal inhibition is not required for left-right rhythmicity but rather is involved primarily with phasing of left-right activity. 2. Locally blocking glycinergic inhibition with strychnine in the rostral spinal cord resulted in synchronous left-right burst activity in that region of the cord as well as in more caudal areas of the cord in which reciprocal inhibition should still be functional. 3. Blocking glycinergic inhibition in the caudal spinal cord converted the pattern in that region of the cord to left-right synchronous activity. The effects in the ascending direction on the burst patterns in more rostral areas of the spinal cord were less than those mentioned above in the descending direction with application of strychnine to the rostral spinal cord. 4. With glycinergic inhibition or GABAergic inhibition blocked in the entire spinal cord, stable longitudinal coupling along the spinal cord persisted. This and the neurophysiology results mentioned above suggest that the main mechanism for longitudinal coupling between locomotor networks in adjacent regions of the spinal cord is ipsilateral excitatory connections and not crossed inhibitory connections. This possible organization was supported by results from a computer model, which consisted of a pair of oscillators in the more rostral and more caudal spinal cord that could be connected by various types of coupling schemes. 5. The neurophysiological data above suggest that ipsilateral, excitatory coupling is stronger in the descending direction than in the ascending direction. In the computer model, a dominant descending coupling is a necessary requirement to produce positive longitudinal phase lags.

Role of excitatory amino acids in brainstem activation of spinal locomotor networks in larval lamprey.

An in vitro brain/spinal cord preparation from larval lamprey was used to determine the role of excitatory amino acid (EAA) receptors in the descending activation of spinal locomotor networks. The general EAA receptor blockers KYN, PDA, and DGG completely blocked locomotor activity initiated from the brainstem. The NMDA receptor blocker APV and the non-NMDA receptor blocker DNQX usually attenuated but did not block locomotor activity. Relatively long and short cycle times were attenuated about equally by APV or DNQX, and therefore the attenuation was not cycle time dependent. Receptor blockers for EAAs attenuated locomotor activity, but often with little or no change in the cycle time of burst activity. Although both NMDA and non-NMDA receptors for EAAs are important for the descending initiation of locomotor activity in the lamprey, it is unclear whether these receptors are concentrated in areas of the spinal locomotor networks that control cycle time.