Pergamon Neuroscience Vol. 74, No. 4, pp. 1161 1173, 1996 Copyright (tD 1996 IBRO. Published by ElsevierScience Ltd Printed in Great Britain PII: S0306-4522(96)00184-4 03064522/96 $15.00+0.00 INTERACTION BETWEEN THE CAUDAL BRAINSTEM AND THE LAMPREY CENTRAL PATTERN GENERATOR FOR LOCOMOTION A. H. COHEN,* L. GUAN, J. HARRIS, R. JUNGt and T. KIEMEL Department of Zoology, University of Maryland, College Park, MD 20742, U.S.A. Abstract--Because of its remarkable simplicity and the robustness of the isolated preparation, the lamprey has been used as a model system to study locomotion and its central pattern generator. The function of the spinal cord is relatively well understood in this context, but the role of the brain or even the caudal brainstem remains less so. We here present a study of the interaction between the caudal brainstem and the spinal pattern generator for locomotion. We show that the interaction is highly complex, with both feedforward input from the brainstem to spinal cord and feedback input from the spinal cord to brainstem playing a significant role in the motor output during locomotion. The brainstem, when diffusely stimulated pharmacologically, can initiate fictive locomotion, or it can disrupt or alter the ongoing D-glutamate initiated motor output. The nature of the disruptions vary greatly, and can induce generalized irregularity, while the alterations can include accelerating or decelerating of the bursting. All behaviors are displayed with spectrograms of the motor nerve discharge. We also show that the unstimulated brainstem can disrupt as well as slow the bursting, but in a complex fashion. Finally, a slow episodic behavior initiated from the caudal brainstem is also described. This can be elicited either by D-glutamate to the brainstem or by ascending activity from the spinal cord pattern generator. Thus, we demonstrate that the interaction between the brainstem and the spinal cord during the production of locomotion is highly complex. The locomotion that is exhibited by the combined brainstem-spinal cord preparation is extremely variable. This is in striking contrast to the variability of the locomotor output pharmacologically induced in the spinal cord alone. The latter preparation exhibits remarkable regularity, or upon occasion, irregularity, but not the routine irregularity or the systematic up and down changes in frequency seen with the brainstem present. However, the pattern of frequency changes induced by the brainstem is not predictable, and remains to be understood. Copyright ,~) 1996 IBRO. Published by Elsevier Science Ltd. Key words: motor systems, neural oscillator, spectrogram, spinal cord. A common view of motor function is that "com- mand neurons" initiate motor output from central pattern generators (CPGs). Interestingly, it has often been found, that once motor output begins, there is phasic feedback from the CPG to the command neurons, s'11'12 One of the systems in which this organization has been described is the lamprey. Nuclei in the lamprey brainstem can initiate loco- motion, 13'15 and they in turn receive phasic input from the CPG in the spinal cord once fictive loco- motion begins. 11 The phasic ascending activity leads to periodic membrane potential oscillations in the brainstem neurons, oscillations that are sufficient to evoke spiking in many of the brainstem neurons with some of the spiking described as synchronous with that of the rostral segments of the spinal cord. While such a feedback loop could set up sustained motor activity, the opposite is most descriptive of fictive *To whom correspondence should be addressed. ]Present address: Center for Biomedical Engineering and Department of Physiology and Biophysics, Wenner-Gren Laboratory, University of Kentucky, Lexington, KY 40506-0070, U.S.A. Abbreviation: CPG, central pattern generator. swimming elicited by the brainstem, as the motor activity is typically short-lived, 3 perhaps more reflec- tive of the mixed feedback that has been described by Dubuc et aL 7 What role the feedback does play is not known. A proposal from Grillner and his col- leagues ll was that the phasic descending brainstem activity maintains the rostral segments at a frequency or excitability that is higher than from the more caudal segments and thereby helps to generate the rearward travelling wave seen in the spinal cord activity. Vinay and Grillner 2° have now shown that the feedback acts instead to slow the rhythm. In this report, we further investigate the dynamic interaction of the brainstem-spinal cord loop and the role it might play in regulation of locomotion. Through the use of the isolated lamprey spinal cord- brainstem preparation with pharmacological stimula- tion of the brainstem and/or the spinal cord to evoke motor activity, we provide evidence that the inter- action between the ascending activity from the loco- motor CPG and the brainstem is more complex than previously thought. Vinay and Grillner 2° found that one effect of the continuous descending activity from the brainstem is to reduce the frequency of fictive 1161