PubMed1999
The patterns of correlation found between motoneuron pools during fictive locomotion are the same whether the coherence functions used to detect the correlations are determined using pairs of rectified ENGs or motoneuron LDPs and rectified ENGs. This finding suggests that the higher frequencies in rectified ENGs (and, perhaps, EMGs) contain information about the synaptic input to motoneurons. Nevertheless, differences between the coherence functions of rectified ENG pairs and those of LDPs and rectified ENGs suggests that this information is distorted by harmonics introduced by rectification. The activities of many motoneuron pools are correlated during the flexor or extensor phase of fictive locomotion, indicating that they receive common synaptic input from branched presynaptic axons or from pools of interneurons whose activities are synchronized. Similar findings were reported by Bayev (1978), based on temporal correlations. Our results indicate that the investigated motor nuclei, which innervate muscles with actions at the hip, knee and ankle, are subject to a set of common locomotor commands. These commands are also received by inhibitory interneurons that project to the motor nuclei of antagonists, as indicated by the correlations between the hyperpolarizing phase of LDPs and activity in the rectified ENGs of antagonists. This last result is consistent with a modular organization for the spinal locomotor generator, in which one set of interneurons drives a motor pool and the inhibitory interneurons that project to the motor pool's antagonist (Jordan, 1991). However, these results also suggest that the spinal modules for locomotion may not be separable into independent unit-burst generators that produce commands for control of each joint as Grillner (1981) has suggested. Our results are more consistent with a model in which a generator distributes flexor and extensor commands to many motor pools (like the half-center model) with as yet unidentified spinal mechanisms that determine differences in the initiation and termination of activity of individual motor nuclei. Alternatively, the correlations between motor pools that we have observed could be explained by spinal mechanisms that synchronize the activity of unit-burst type generators. Despite the distribution of common locomotor commands to many functionally diverse motor nuclei, the spinal locomotor pattern generator is differentiated to the extent that some motor nuclei, like EDL and FDL, receive separate locomotor commands. This conclusion is consistent with other observations. EDL and FDL display distinctive, individualized patterns of locomotor activity that may vary in a facultative manner or in different forms of locomotion (O'Donovan et al., 1980; Trank et al., 1996). A recent study has shown that during fictive locomotion EDL and FDL motoneurons receive input from different sets of last-order interneurons than those which project to other motor pools (Degtyarenko et al., 1998). These results suggest that spinal locomotor generators are differentiated for the individualized control of some digit muscles, like FDL and EDL.