Discharge Properties of Brain Stem NeuronsProjecting to the Flocculusin the Alert Cat. I. Medial VestibularNucleus GUY CHERON, MIGUEL ESCUDERO,AND EMILE GODAUX Laboratory of Neurosciences, (Jniversity of Mons-Hainaut, Mons 7000, Belgium; and I'Qboratory of Neurosciences, University of Sevilla, Seville 41012, Spain SUMMARY A N D CONCLUSIONS 1. The aim of this study was to characterize the signalstransmit- ted by neurons of the medial vestibular nucleus (MVN) to the middle zone of the flocculus in alert cats. 2. Bipolar stimulating electrodes were implanted into the middle zone of each flocculus, because this zone is known to be involved in the control of horizontal eye movements. Correct implantation of the stimulating electrodeswas ensured by 1) recording of Pur- kinje cells whose activity was related to horizontal eye movements and 2) elicitation of slow abduction of the ipsilateral eye upon electrical stimulation. 3. The rostral two-thirds of the MVN were investigatedby mi- croelectrodesduring stimulation of both flocculi. Antidromically activated neurons were found only in the central part of the ex- plored area.Forty-four units were activated from the contralateral, eight from the ipsilateral flocculus. Neurons could never be acti- vated from both flocculi. 4. Neurons included in this study were MVN neurons that had / ) to be antidromically activated from one flocculus and 2) to modulate their firing rate during the horizontal vestibuloocular re- flex (VOR) elicited by sinusoidal stimulation(0.1 Hz; 10, 20, 30, or 40'). The 39 neurons matching both criteria were classified in 2 groups: 22 neurons changed their firing rate during spontaneous horizontal eye movements (EM-neurons), 17 modulated their ac- tivity only during head rotation and were labeled vestibular-only neurons (VO-neurons). 5. Sufficient data were obtained from 13 EM-neurons to allow a quantitativeanalysis.Among those, 12 were activatedfrom the contra- lateral and I from the ipsilateralflocculus. Their sensitivityto horizontal eye position during intersaccadic fixation was 3.54 t 2.75 (SD) spikes's r/deg. Eight EM-neurons behaved as type I neurons,five as type II neurons. During the slow phases of the VOR, all of these neurons combined some head-velocitysensitivity (1.50 + 0.43 spikes's-'/ deg.s-') with some horizontaleye-position sensitivity (3.61 + 2.45 spikes's-t/deg). Additionally,seven ofthese neurons presented a sen- sitivityto eye velocity(1.34 + 0.55 spikes's-'/deg's-'). The phase differencebetweenthe modulation of firing rate and eye position varied substantiallybetween neurons. The observed phase lead with respect to eye posiûonrangedfrom 2 to 110" (41.9 + 31.8"). 6. Sufficient data were obtained from 10 VO-neurons to allow a quantitative analysis.Among those, nine were activated from the contralateral and one from the ipsilateral flocculus. All of these neuronsbehavedas type I neurons.The sensitivity to head velocity was L64 + 1.07spikes ' s -'ldeg ' s -r . The phase lead of the modu- lation of spike activity with respect to head velocity ranged from 4.5 to 30.5" (16.4 + 8.9'). Z. We conclude that the MVN provides the horizontal zone of the flocculus (with a strong contralateral preference)with informa- tion about head velocity (through VO-neurons and EM-neurons) and about eye velocity and position (through EM-neurons). JouRNAL oF NEURoPHYSIoLocY Vol.76, No.3, September 1996. Printed in U.S.A. I NTRO D U C T I ON A major challenge in current neuroscience is to elucidate the function of natural neural networks.Because of its struc- tural homogeneity, the cerebellarcortex is a suitable target for this challenge. Although the basiccircuitry of the cerebel- lum has been extensivelystudied (Eccles et al. i967; Ito 1984; Llinas and Hillman 1969), its mode of information processing remains unknown.To understand this processing, it is crucial to study a part of the cerebellar cortex that afferent and etTerent connections are well known anatomi- cally and that controls easily measurable movements. The cerebellar flocculus is known to control eye move- ments by acting on a limited number of premotor brain stem nuclei (Ito et al. 1917; Lisberger and Fuchs 1978;Noda and Suzuki 1919a-c; Noda and Warabi 1986; Waespe et al. 198 l;Waespe and Henn 1981), and its anatomic organiza- tion is well known. In the cat, three different floccular zones have been described (rostral, middle, and caudal), whose Purkinje cells project onto different parts of the vestibular nuclear complex (Sato et al. 1982). The rostral and caudal zones control the vertical recti and the oblique muscles, whereasthe middle zone controls the horizontal recti (Sato et al. 1988; Sato and Kawasaki 1984,1981). The output of the zonesprovide three operational units (Oscarsson 1979) through which the flocculus controls eye movements (Sato and Kawasaki1984,1990,1991). The distribution of climbing f,bersin the flocculus stresses its modular character, because fibers from different parts of the contralateralinferior olive terminate in discrete cortical strips. In the cat, Gerrits and Voogd (1982) distinguished six different strips that convey direction-specific information (Graf et al. 1988; Leonard et al. 1988; Stahl and Simpson 1992) and seem congruent with the three Purkinje cells out- put zones. In contrast to the selective projections of climbing fibers in discrete zones, mossy fiber input apparently spreads across all the floccular zones ( Genits et al. 1984; Satoet al. 1983b ) . However, the fact that a given nucleus projects to the three floccular zones simultaneouslydoes not imply that it sends identical signals to thesethree zones.The flocculus receives mossy fibersprojections from five main sources: /) ttre ves- tibular nuclear complex, 2) the perihypoglossal nuclei (Epema et al. 1990;Langer et al. 1985a,b; Satoet ali.1983b; Tan and Gerrits 1992;Tan et al. 1995),3) the paramedian and lateral reticular nuclei (Satoet al. 1983b) ,4) thenucleus reticularis tegmentipontis (Gerrits et al. 1984), and 5) the 0022-3071196 $5.00 Copyright O 1996 The American Physiological Society t7 59