Tracing premotor brain stem networks of orienting movements Alexej Grantyn, Etienne Olivier and Toshihiro Kitama CNRS-College de France, Paris, France zyxwvutsrqponmlkjihgfedcbaZYXWVUTSR Methods a llo wing a d ire c t m a tc hing of m o ve m e nt- re la te d firing p a tte rns and c o nne c tivity of ind ivid ua l ne urons have been use d in the a na lysis of p re m o to r ne two rks c o ntro lling o rie nting movements. Advances have been made in the d e sc rip tio n of c o d ing p ro p e rtie s of o rie nting - re la te d te c ta l o utp ut ne urons, a s we ll a s in sp e c ifying the ir d istrib ute d c o nne c tio ns in the b ra in ste m and possible modes of c o up ling to sa c c a d ic p a tte rn g e ne ra to rs in the re tic ula r fo rm a tio n. Ne w data on the properties o f sig na ls a nd c o nne c tivity p a tte rns have also b e e n o b ta ine d fo r the te c to - re c ip ie nt re tic ul+ sp ina l ne urons. At least a small p o rtio n of the ne twork p e rfo rm ing the sp a tio - te m p o ra l tra nsfo rm a tio ns o f o rie nting - re la te d te c ta l e ffe re nt sig na ls can now be described b o th in func tio na l a nd in m o rp ho lo g ic a l te rm s. C urre nt O p inio n in Ne uro b io lo g y 1993, 3:973-981 Introduction zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA During the past decade, the general views on the neuronal organization of sensorimotor systems have changed considerably. A new working hypothesis, that of parallel and distributed processing, has be- come popular and increasingly dominant. The ‘tradi- tional’ and ‘modern’ views have been confronted in a number of publications (e.g. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA [l-3,4*,5’]). Besides re- flecting the evolution of conceptions towards connec- tionism, they show that the connectionist hypothesis, if rigorously applied to biological neural networks, leads to discouraging conclusions for an experimen- talist. Indeed, it asserts that anatomical connectivity is not correlated to functional connectivity and, con- sequently, relationships between neuronal firing pat- terns and movement parameters have little or no value in predicting the structure of the network. It is easy to see that in such a case the existing physiological and morphological methods become fairly useless for un- derstanding how real neurons control real movements. On the other hand, it is admitted that brain models, connectionist or not, have a rather limited value, unless they incorporate constraints derived from experiments on live networks. Moreover, sequential signal transfor- mations and a certain degree of localization of function do coexist with parallel and distributed processing in motor systems. Studies combining functional identifl- cation of neurons by correlating their signals to move- ment parameters with identification of their individual anatomical connections are, therefore, indispensable. The possibilities and limitations of this method, first applied to the mammalian vestibule-ocular reflex in the early eighties [6,71, will be presented in this review, taking as an example the tecto-reticula-spinal system and its contribution to orienting movements. The tectmotoneuronal interface In the brain stem, the superior colliculus (SC) repre- sents, no doubt, the highest hierachical level of the mo- tor system of orienting. Its intrinsic circuits transform the multimodal, topographically organized, represen- tation of space to a topographically ordered activity profiles of efferent neurons projecting to the brain- stem tegmentum and the spinal cord 181. Efferent SC neurons encode the direction and amplitude of orient- ing movements by their location on the motor map. Their signals must be transformed before reaching the motor pools. Transformations in the temporal domain are needed because signals generated by tectal effer- ent neurons do not specify sufficiently the dynamic parameters of movements, especially those involving complex muscle synergies. The vectorial representa- tion of movement in the SC must be resolved in appro- priately weighted activities of horizontal and vertical saccadic generators. This transformation to the Carte- sian system of coordinates is well established for eye movements (reviewed in 1911, head movements in the barn owl 110,11*1, and body movements in the frog 1121. As recently reviewed by Masino 1131,a subsequent tran- Abbreviations Abd-N-abducens nuc le us; EMG - e le c tro m yo g ra m ; EPSP- e xc ita to ry p ustsyna p tic p o te ntia l; P- p ha sic ; PS- p ha sic - susta ine d ; Q Tq ua si- to nic ; RF- re tic ula r fo rm a tio n; RG c - g ig a nto c e llula r re tic ula r nuc le us; RSN- re tic ula - sp ina l ne uron; SC - sup e rio r c o llic ulus; SRBN- sa c c a d e - re la te d b urst ne uro n; TRSN- te c to re tic ulo - sp ina l ne uro n; TST- te c to sp ina l tra c t. 0 Current Biology Ltd ISSN 0959-4388 973