Takahashi JNP Page-1 Convergent Synaptic Inputs from the Caudal Fastigial Nucleus and the Superior 1 Colliculus onto Pontine and Pontomedullary Reticulospinal Neurons

240 words 26 The caudal fastigial nucleus (FN) is known to be related to the control of eye 27 movements, and projects mainly to the contralateral reticular nuclei where excitatory (EBNs) 28 and inhibitory burst neurons (IBNs) for saccades exist (the caudal portion of the nucleus 29 reticularis pontis caudalis, NRPc, and the rostral portion of the nucleus reticularis 30 gigantocellularis, NRG, respectively). However, the exact reticular neurons targeted by caudal 31 fastigioreticular cells remain unknown. We tried to determine the target reticular neurons of 32 the caudal FN and superior colliculus (SC) by recording intracellular potentials from neurons 33 in the NRPc and NRG of anesthetized cats. Neurons in the rostral NRG received bilateral, 34 monosynaptic excitation from the caudal FNs, with contralateral predominance. They also 35 received stronger monosynaptic excitation from the caudal contralateral SC, and disynaptic 36 excitation from the rostral ipsilateral SC. These reticular neurons with caudal fastigial 37 monosynaptic excitation were not activated antidromically from the contralateral abducens 38 nucleus, but most of them were reticulospinal neurons (RSNs) that were activated 39 antidromically from the cervical cord. RSNs in the caudal NRPc received very weak 40 monosynaptic excitation from only the contralateral caudal FN, and received either 41 monosynaptic excitation only from the contralateral caudal SC, or monosynaptic and 42 disynaptic excitation from the contralateral caudal and ipsilateral rostral SC, respectively. 43 These results suggest that the caudal FN helps to control also head movements via RSNs 44 targeted by the SC, and these RSNs with SC topographic input play different functional roles 45 in head movements. 46 Takahashi JNP Page-3 INTRODUCTION 47 The cerebellum is known to play a significant role in the coordinated control of 48 complex movements. Based on anatomical studies, it is divided into three sagittal cortical 49 zones associated with their corresponding main nuclei: the fastigial, interposed and dentate 50 nuclei (Jansen and Brodal 1940; see a review by Voogd et al. 2013). To understand the 51 functional roles of these three zones, Chambers and Sprague (1955) and later, Thach (as 52 reviewed in Thach et al. 1992) performed lesion experiments, and found that among the three 53 zones, the vermis and its underlying fastigial nucleus (FN) were concerned with maintaining 54 posture and walking with its associated body and neck movements. 55 To further understand the neural circuitry of the fastigial control of complex movements, 56 afferent and efferent projections of the FN have been extensively studied using the silver 57 impregnation method (Carpenter and Nova 1960; Walberg et al. 1962a, b), the horseradish 58 peroxidase labeling method (cat, Fukushima et al. 1977; Matsushita and Hosoya 1978), the 59 autoradiographic method (cat, Moolenaar and Rucker 1976; monkey, Asanuma et al. 1983; 60 Batton et al. 1977), and the phaseolus vulgaris leucoagglutinin anterograde labeling method 61 (Homma et al. 1995). The cerebellar FN is the main source of cerebellar efferents to the lower 62 brainstem (Thomas et al. 1956). It has been generally accepted in degeneration studies that 63 fibers of the hooked bundle of Russell arise predominantly from the caudal FN and cross 64 within the cerebellum, whereas uncrossed fastigial efferent fibers arise predominantly from 65 the rostral FN and emerge from the cerebellum via the juxtarestiform body (Angaut and 66 Bowsher 1970; Carpenter et al. 1958; Cohen et al. 1958; Jansen and Jansen 1955; Rasmussen 67 1933; Voogd 1964; Walberg et al. 1962a, b). On the other hand, the selective labeling of either 68 rostral or caudal cells in the FN with autoradiographic methods revealed that some cells at all 69 rostrocaudal levels of the FN must give rise to crossed and uncrossed fibers into the brainstem, 70 but fastigioreticular fibers are almost entirely crossed and arise from all parts of the FN 71 Takahashi JNP Page-4 (Batton et al. 1977). The densest projection is to the nucleus reticularis gigantocellularis 72 (NRG) from the level of the abducens nucleus to the caudal olivary level of the medulla 73 oblongata. The major terminal regions within the NRG are located rostrally and medially 74 (Asanuma et al. 1983; Batton et al. 1977), although the rostral FN and caudal FN project 75 predominantly to the ventral and dorsal NRG, respectively (Homma et al. 1995). 76 In addition to limb and body movements, the cerebellar control of eye movements was 77 investigated in the posterior vermis (Keller et al. 1983; McElligott and Keller 1984; Ritchie 78 1976; Ron and Robinson 1973) and fastigial nucleus (Cogdell et al. 1977). Later, 79 microstimulation (Fujikado and Noda 1987; Noda and Fujikado 1987a) and recordings of 80 saccade-related Purkinje cell activity (Kase et al. 1980; Ohtsuka and Noda 1995) revealed that 81 vermal lobule VII and its adjacent lobule VIc are concerned with eye movements. The 82 destruction of Purkinje axons by focal injection of kainic acid (Noda and Fujikado 1987b) and 83 injection of bicuculline (Sato and Noda 1992) into the FN suppressed saccades evoked by 84 vermal microstimulation. These findings suggested that the caudal FN relayed impulses from 85 vermal lobules VI and VII to the brainstem oculomotor circuitry. Consistent with this 86 suggestion, neurons in the caudal FN discharge during saccades (Fuchs et al. 1993; Gruart and 87 Delgado-Garcia 1994; Helmchen et al. 1994; Hepp et al. 1982; Ohtsuka and Noda 1991). In 88 addition, some neurons in the caudal FN respond during smooth-pursuit eye movements 89 (Büttner et al. 1991; Fuchs et al. 1994). Therefore, the caudal FN is considered to be 90 concerned with eye movements. However, other studies showed that inactivation of the caudal 91 FN produced dysmetria of not only eye movements, but also head movements, showing that 92 the caudal FN is also involved in head movements in the cat (Goffart and Pélisson 1998a; 93 Goffart et al. 1998b) and monkey (Quinet and Goffart 2007). 94 As mentioned above, the general pattern of termination of fastigial fibers in the 95 brainstem appeared to be fairly well understood, but the exact projections from the caudal FN, 96 Takahashi JNP Page-5 especially the fastigial oculomotor area, were unclear. Noda and his colleagues analyzed 97 afferent and efferent projections of the vermal oculomotor region in the brainstem of the 98 monkey (Yamada and Noda 1987). Subsequent anatomical studies on the caudal FN have 99 shown that axons of neurons in the caudal one-third of the FN terminate in areas where 100 saccade-related premotor burst neurons are located [the excitatory burst neuron (EBN) area in 101 the rostral interstitial nucleus of MLF and the paramedian pontine reticular formation (PPRF) 102 (Cohen and Henn 1972; Igusa et al. 1980; Keller 1974; Luschei and Fuchs 1972), and the 103 inhibitory burst neuron (IBN) area in the paramedian pontomedullary reticular formation 104 (PPMRF) (Hikosaka and Kawakami 1977; Yoshida et al. 1982)], and among them, the most 105 extensive projection is to the contralateral NRG where IBNs are located in the monkey (Noda 106 et al. 1990; Sugita and Noda 1991). They suggest that at least a part of the labeling found in 107 the medullary reticular formation immediately caudal to the abducens nucleus may 108 correspond to the so-called IBN area. Based primarily on these anatomical findings, it is 109 generally considered that fastigioreticular neurons in the caudal FN terminate on contralateral 110 IBNs via the hook bundle (e.g., Fuchs et al. 1993; Goffart et al. 2004; Kojima et al. 2008; 111 Ohtsuka and Noda 1995; Scudder and McGee 2003). However, no experimental data are 112 available to confirm that caudal fastigioreticular neurons terminate on IBNs. With these 113 anatomical methods, the exact target neurons of these efferent fibers arising from the caudal 114 FN could not be identified as EBNs and IBNs, because the nucleus reticularis pontis caudalis 115 (NRPc) and NRG also contain reticulospinal neurons (RSNs) that innervate neck 116 motoneurons (Grantyn and Berthoz 1987; Isa and Sasaki 2002; Iwamoto and Sasaki 1990; 117 Kakei et al. 1994; Peterson et al. 1978; Shinoda et al. 2006). 118 Although anatomical data show that axon terminals of caudal fastigial neurons are 119 distributed in the NRPc and NRG, they could not determine whether these terminals actually 120 terminate on EBNs and IBNs or RSNs. However, using intracellular recordings of 121 Takahashi JNP Page-6 postsynaptic potentials (PSPs), we could identify target neurons of caudal fastigial axons 122 electrophysiologically by their antidromic responses to stimulation of their destinations or 123 morphologically by intracellular labeling of target neurons with horseradish peroxidase. In 124 spite of many anatomical studies on fastigioreticular projections, there have been far fewer 125 intracellular recording studies on synaptic inputs to their target neurons in the brainstem. 126 Reticular neurons (Ito et al.1970) and vestibular nucleus neurons (Furuya et al. 1976) have 127 been identified as targets of fastigioreticular neurons. Ito et al. (1970) analyzed the effects of 128 systematic electrical stimulation of the cerebellar nuclei on reticular neurons in the medulla 129 oblongata of the cat. Stimulation of the FN on either side produced monosynaptic excitation 130 of medullary reticular neurons, most of which were RSNs that projected to the spinal cord. 131 However, their stimulation sites in the cerebellar nucleus were located in the rostral FN, 132 where both output neurons in the rostral FN and axons from the caudal FN can be activated. 133 In addition, reticular neurons were recorded rather caudally at the level of the inferior olive 134 where IBNs are not located (Sugiuchi et al. 2005). Therefore, the target neurons of the caudal 135 FN remain unknown with respect to brainstem oculomotor and head movement systems. 136 The purpose of the present experiments was to investigate major targets in the lower 137 brainstem of fastigioreticular neurons in the caudal FN using intracellular recording 138 techniques. More specifically, to understand the neural mechanisms of the caudal FN in the 139 cerebellar control of head movements, we tried to determine the major targets of the caudal 140 FN efferent neurons in the portions of the caudal NRPc where EBNs lie, and the rostral NRG 141 where IBNs are found. Since the superior colliculus (SC) is well known to be involved in the 142 control of eye and head movements, we further examined how reticular neurons in the NRG 143 and NRPc that receive tectal inputs are influenced by the caudal FN. 144 145 Takahashi JNP Page-7 METHODS 146 Experiments were performed in 12 cats weighing 2.5-4.0 kg. Anatomical data were 147 collected from three other cats used for a previous report by Takahashi et al. (2010). Animal 148 experimentation was conducted in accordance with the “Policies on the Use of Animals and 149 Humans in Neuroscience Research” approved by the Society for Neuroscience in 1995, and 150 the "Guiding Principles for the Care and Use of Animals in the Field of Physiological 151 Sciences” (The Physiological Society of Japan, revised in 2001). The experimental protocol 152 was approved by the Animal Care Committee of Tokyo Medical and Dental University. The 153 animals were initially anesthetized with ketamine hydrochloride (Ketalar, Parke-Davis; 25 154 mg/kg, im) followed by -chloralose (40-45 mg/kg, iv, initial dose, supplemented with 155 additional doses of 10-25 mg/kg, iv, throughout the remainder of the experiment). The heart 156 rate was continuously monitored by an ECG. The body temperature was maintained between 157 38 and 39C by means of a heating pad. After setting stimulating electrodes in the brainstem 158 and FN, the animals were paralyzed by the intravenous administration of pancuronium 159 bromide (Mioblock, Organon, The Netherlands), and artificially ventilated with end-tidal CO2 160 held at 35-40 mmHg, and then intracellular recording was performed. 161 The bone over the cerebellar vermis including lobule VII was removed so that the 162 stimulating electrodes could be introduced bilaterally into the caudal FNs. For stimulation of 163 or injection of a tracer into the caudal FN, we mapped effective stimulation sites for evoking 164 eye movements at 200-m intervals, while recording characteristic discharge patterns in 165 cerebellar molecular, Purkinje cell and granular cortical layers, white matter and the caudal 166 FN (Eccles et al. 1967). The intensity of electrical stimuli (0.2 ms duration, 2.5 ms interval, 167 30 pulses-train) was kept between 10 and 100 A. If no eye movements were evoked with 168 this current, the electrode was advanced to the next test site at 200-m intervals. When eye 169 movements were evoked, we lowered the stimulus intensity at 10-A steps and determined 170 Takahashi JNP Page-8 the lowest intensity to evoke eye movements at each stimulation site. For stimulation of the 171 caudal FN, separate stimulating electrode arrays were inclined 30-35o posteriorly from the 172 stereotaxic vertical in the parasagittal plane, and placed in the right and left FNs. The 173 stimulating electrode arrays consisted of three monopolar electrodes (diameter, 100 m), 174 insulated except at the tip, that were glued together around a central pillar, so that the tips of 175 the three electrodes were arranged dorsoventrally at 1.0-mm intervals (Sugiuchi et al. 2005, 176 Takahashi et al. 2005a, 2007) to cover the caudal part of the FN, which protrudes 177 dorsocaudally in the cat. Stimulus currents were delivered between two adjacent tips, and the 178 pillar was grounded to reduce stimulus artifacts. Negative pulses of 0.2-ms duration were 179 delivered at 500 A or less for stimulation of the SC and the FN. The parietal and occipital 180 cortex over the SC was removed bilaterally by aspiration to allow the bilateral introduction of 181 stimulating electrodes into the SCs. In nine cats, three concentric bipolar stimulating 182 electrodes (ID and OD, 0.1 and 0.3 mm, respectively; interelectrode distance along the 183 longitudinal axis, 1.0 mm) with a 1.2-1.4 mm rostrocaudal separation were placed along the 184 presumed horizontal meridian of the motor map in the SC on both sides under direct visual 185 observation by referring to the motor map of the cat SC (McIlwain 1986), and positioned in 186 the intermediate or deep layer (1.5-2.0 mm from the surface) of the SC (for details, see 187 Takahashi et al. 2005a, 2007). Ranck (1975) estimated an effective current spread of 1.0-1.5 188 mm around an electrode tip by monopolar stimulation at 500 A (200s-duration pulse) in 189 the mammalian CNS. However, since we used bipolar stimulation, the effective current 190 spread should be much less than the values estimated by Ranck (1975) (Shinoda et al. 1977). 191 Sasaki et al. (1970, 1972) estimated that 500 A could not activate fibers or cells beyond 1.0 192 mm from an electrode tip with the use of a concentric bipolar electrode of the same type as in 193 the present study. At the end of each experiment, some electrolytic lesions (30 A of cathodal 194 current for 30 s) were made to mark electrode tips, and stimulation sites were reconstructed 195 Takahashi JNP Page-9 with reference to these lesion sites histologically, using celloidin-embedded serial sections of 196 the cerebellum and brainstem stained with thionine. 197 The ventral part of the posterior vermis overlying the fourth ventricle was partially 198 removed by aspiration to facilitate the placement of recording glass microelectrodes in the 199 NRG and NRPc on the left side (Fig. 2B). The recording microelectrodes were inclined about 200 60o posteriorly from the stereotaxic vertical in the parasagittal plane. For intracellular 201 recordings from reticular neurons, glass microelectrodes filled with 0.4 M KCl or 2 M 202 K-citrate and a resistance of 15-25 M were introduced into the brainstem on the left side. 203 We first identified the location of the abducens nucleus by recording the type II responses of 204 abducens motoneurons to rotation of the turntable in the horizontal plane (Maeda et al. 1971; 205 Shinoda and Yoshida 1974), and by recording characteristic negative antidromic field 206 potentials that were evoked by stimulation of the left abducens nerve (Baker and Highstein 207 1975). Relative to this location of the abducens nucleus, reticular neurons were searched for 208 the region rostral (NRPc) and caudal (NRG) to the abducens nucleus on the left side. 209 To investigate the projection areas of the caudal FN and the SC in the brainstem, a 210 12.5% solution of dextran-biotin (DB, Invitrogen) was injected into the caudal FN in 3 cats, 211 and the SC in 3 cats. A glass microelectrode attached to a microsyringe was inserted into a 212 target, and <1.0 l of solution was injected slowly for 20-30 min. After a survival period of 213 7-10 days, animals were deeply anesthetized with ketamine hydrochloride (25 mg/kg, im), 214 followed by pentobarbital sodium (50 mg/kg, iv), and perfused with 2 L of 0.9% NaCl in 215 phosphate buffer (0.1 M, pH 7.4) followed by 2 L of a fixative solution containing 4% 216 paraformaldehyde and 0.05% glutaraldehyde with 0.2% picric acid. Frozen serial sagittal or 217 frontal sections, 70 m thick, were cut on a freezing microtome, and processed to show the 218 injected tracer by use of avidin conjugated horseradish peroxidase procedures and visualized 219 with diaminobenzidine as a chromogen (Adams 1981; Takahashi et al. 2010). 220 Takahashi JNP Page-10 RESULTS 221 Distributions of axon terminals of caudal FN neurons and tectoreticular neurons 222 To determine the exact location of axon terminals of caudal FN neurons in the 223 brainstem before intracellular recordings, we injected an anterograde tracer, DB, into the 224 caudal FN and observed the distribution of labeled axon terminals and fibers in the NRPc and 225 NRG. An injection site in the caudal FN was determined by mapping sites at which 226 stimulation effectively evoked eye movements in three cats. A stimulating electrode was 227 inclined 30 degrees caudally from the stereotaxic vertical and advanced at 200 m-intervals. 228 As it passed through the overlying cerebellar cortex (lobule VII), the electrode entered the 229 white matter between the overlying lobule VII and the dorsal border of the caudal FN. 230 Stimulation of this area evoked ipsilateral small eye movements; in the dorsal area within the 231 FN, induced eye movements were still ipsilateral (Noda et al. 1988; Quinet and Goffart 2009). 232 As the electrode was advanced further, the direction of the induced eye movements was 233 reversed to the contralateral side with a decrease in the threshold. Within the FN, background 234 activity was high, with high-frequency spontaneous negative spikes. DB was injected into this 235 area of the caudal FN (Fig. 1A). Figure 1B shows the distribution of labeled axon terminals in 236 the two levels of the brainstem: the NRPc (a), just rostral to the abducens nucleus and the 237 NRG (b), just caudal to the abducens nucleus. Labeled terminals were distributed mainly in 238 the contralateral paramedian reticular formation in the caudal part of the NRPc (Fig. 1Ba) and 239 in the rostral part of the NRG (Fig. 1Bb). 240 To compare the brainstem distributions of axon terminals of neurons in the caudal FN 241 and the SC, DB was injected into the right SC (Fig. 1C). Figure 1D shows the distribution of 242 axon terminals of labeled neurons after the injection of DB into the right SC. Most axon 243 terminals were distributed in the contralateral PPRF, just rostral to the abducens nucleus (Fig. 244 1Da), and in the contralateral PPMRF, caudal to the abducens nucleus (Fig. 1Db). Terminal 245 Takahashi JNP Page-11 distribution areas of the tectoreticular and fastigioreticular fibers extensively overlapped each 246 other in the NRG. The fastigial projection to the NRPc was much less abundant than that to 247 the NRG. We previously determined the distributions of premotor neurons terminating on 248 abducens motoneurons by using a transneuronal labeling method; the ipsilateral premotor 249 neurons were located 0.5-1.5 mm rostral to the abducens nucleus and 1.0-2.9 mm from the 250 midline, and the contralateral premotor neurons were located from the caudal abducens 251 nucleus to 1.3 mm more caudally and 0.8-1.5 mm lateral to the midline (Sugiuchi et al. 2005). 252 These anatomically-delineated premotor regions were compatible with the locations of EBNs 253 (Igusa et al. 1982) and IBNs determined electrophysiologically in the cat (Hikosaka and 254 Kawakami 1977; Yoshida et al. 1982). The present locations of caudal fastigial and tectal 255 axon terminals in the NRPc and NRG well-covered these EBN and IBN areas, respectively. In 256 the following intracellular recordings, we recorded reticular neurons in these anatomically 257 delineated EBN and IBN areas. According to Brodal (1956), the NRG in the cat occupies 258 approximately the medial two-thirds of the reticular formation lying dorsal to the rostral half 259 of the inferior olive and extending cranially to the level of the facial nucleus. Rostral to and 260 not sharply delimited from this nucleus is the NRPc. Therefore, the IBN area corresponds to 261 the most rostral portion of the NRG, and the EBN area to the most caudal portion of the 262 NRPc. 263 264 Synaptic inputs from the caudal FN and the SC to neurons in the NRG and NRPc 265 To determine the exact brainstem target neurons of fastigioreticular axons arising from 266 the caudal FN, we recorded intracellular potentials from reticular neurons in the rostral NRG 267 and caudal NRPc just caudal and rostral to the abducens nucleus, respectively, and searched 268 for reticular neurons that received synaptic inputs from the caudal FNs. At the same time, we 269 examined whether the same reticular neurons with fastigial input received inputs from both 270 Takahashi JNP Page-12 SCs. All lateralities in the present study are described with reference to the recording site, if 271 not stated otherwise. 272 Figure 2 shows a typical example of synaptic inputs from the caudal FNs (Fig. 2F) and 273 the SCs (Fig. 2G) to a neuron in the NRG. At the beginning of each experiment, we mapped 274 antidromic field potentials evoked by stimulation of the left abducens nerve and determined 275 the approximate center location of the left abducens nucleus (Fig. 2D). Reticular neurons 276 were penetrated with reference to this location and the locations of recording sites of reticular 277 neurons were reconstructed histologically, using a guide glass electrode that was left in the 278 brainstem as a landmark. This reticular neuron was located about 1 mm caudal to the 279 abducens nucleus and was not activated antidromically from the left abducens nerve (Fig. 2E). 280 Once a neuron was penetrated, we systematically examined postsynaptic potentials (PSPs) 281 evoked by stimulation of the three ventrodorsal sites in the caudal FN and three rostrocaudal 282 sites in the SC in every neuron. In this neuron, stimulation of the three ventrodorsal sites in 283 the contralateral caudal FN (sites 2-4 in Fig. 2, A and C) evoked spikes in all traces at a low 284 stimulus intensity (50 A) (Fig. 2F, 2-4). These spikes had fluctuating latencies and were 285 preceded by depolarizing potentials, indicating that these spikes were evoked orthodromically 286 from the caudal FN. Stimulation of the ipsilateral caudal FN (sites 5-7 in Fig. 2, B and C) also 287 evoked depolarizations with occasional spikes at low stimulus intensities (100 A) in the 288 same reticular neuron (Fig. 2F, 5-7). Subsequently, we examined the effects of stimulation of 289 the SC on PSPs evoked in the same reticular neuron. Stimulation of the rostrocaudal sites in 290 the contralateral SC evoked depolarizations with spikes (Fig. 2G, 8-10) and stimulation of the 291 ipsilateral SC also evoked depolarizations with spikes at longer latencies (Fig. 2G, 11-13). In 292 this way, we examined the properties of PSPs evoked by stimulation at different sites in the 293 bilateral caudal FNs and bilateral SCs in each reticular neuron. First, we will describe the 294 characteristic features of synaptic input from the bilateral caudal FNs, and then from the 295 Takahashi JNP Page-13 bilateral SCs to neurons in the NRG. Next, we will describe the features of synaptic inputs 296 from the caudal FNs and SCs to neurons in the NRPc. 297 298 Properties of synaptic input from the caudal FN to reticular neurons in the NRG 299 In the NRG neuron shown in the example in Fig. 2, stimulation of the three 300 ventrodorsal sites in the contralateral caudal FN evoked depolarizing potentials with 301 fluctuating spikes in all traces at a low stimulus intensity (50 A) (Fig. 2F, 2-4). The latencies 302 of the preceding depolarizations and evoked spikes became slightly shorter as the stimulation 303 sites were moved ventrally and rostrally (Fig. 2F2). Since the depolarizations had the shortest 304 latencies with the most ventral stimulating electrode (site 2) that was closer to the rostral FN, 305 the responses might be caused by stimulation of the rostral FN due to current spread. However, 306 depolarizations and spikes were also evoked from the more caudal and dorsal stimulation sites 307 (sites 3 and 4) at low stimulus intensities. The histological identification of the stimulation 308 sites indicated that the three electrodes were well located in the center of the caudal FN 309 mediolaterally and rostrocaudally. Based on this histological finding and the extremely low 310 stimulus intensity that was required to evoke depolarizations, it was safe to conclude that 311 large depolarizations were caused by the activation of neurons in the caudal FN, and not due 312 to current spread to the rostral FN (see the estimation of current spread with the present 313 electrodes and stimulus intensities in the Methods). Stimulation of the ipsilateral caudal FN 314 also evoked depolarizations with occasional spikes at low stimulus intensities (100 A) in the 315 same neuron (Fig. 2F, 5-7). Histological examination showed that the three stimulation sites 316 (sites 5-7) were well within the caudal FN (Fig. 2B). The anatomical data in Fig. 1 indicated 317 that the projection from the caudal FN to the ipsilateral rostral NRG was extremely sparse. 318 Therefore, it was necessary to exclude the possibility of current spread to the contralateral 319 caudal FN, and to confirm that ipsilateral FN stimulation actually evoked depolarizations in 320 Takahashi JNP Page-14 the reticular neuron rather than due to the current spread. Mediolaterally, the fastigial 321 stimulating electrodes in this example were 1.4 to the left and 1.5 mm to the right of the 322 midline, respectively. Since the medial border of the caudal FN was 0.5 to the right and 0.6 323 mm to the left of the midline, respectively, the distances from the left and right stimulating 324 electrodes to the medial border of the FN on the opposite side were 1.9 and 2.1 mm, 325 respectively. Based on the long distance (1.9 mm) and low stimulus intensities (100 A), it 326 was highly unlikely that the current spread from the stimulating electrodes in the left caudal 327 FN to the right caudal FN caused the depolarizations in the neuron shown in Fig. 2F. However, 328 it is possible that axons crossing from the contralateral fastigial nucleus were activated by the 329 ipsilateral stimulating electrodes, as will be considered further in the Discussion. 330 The effects of an increase in stimulus intensity on the evoked depolarizations were 331 examined by stimulating an identical site in the caudal FN (Fig. 3A) with increasing stimulus 332 intensities (Fig. 3B). The threshold for evoking depolarizing potentials was less than 50 A, 333 and their onset latencies were 1.2 ms. As the stimulus intensities increased, the latencies of the 334 depolarizations decreased by 0.1-0.3 ms, and their amplitudes gradually increased and were 335 almost saturated at 400 A. The maximal amplitudes of the evoked depolarizations were 7-8 336 mV. These depolarizations were monophasic with a summit time of 1.0-1.7 ms and decayed 337 monotonically to the baseline resting potentials. This feature of the evoked depolarizations 338 from the caudal FN is consistent with the excitatory postsynaptic potentials (EPSPs) recorded 339 in medullary reticular neurons during stimulation of the rostral FN (Ito et al. 1970). 340 341 Identification of the synaptic nature of fastigioreticular inputs to NRG neurons 342 To determine whether the depolarizations evoked by stimulation of the caudal FN were 343 EPSPs or disinhibition that could be attributed to a decrease in inhibitory postsynaptic 344 potentials (IPSPs), we injected Cl into cells by passing a hyperpolarizing current through a 345 Takahashi JNP Page-15 recording glass microelectrode (Fig. 3, C-I). In most of the penetrated cells, no 346 hyperpolarizing potentials were evoked by stimulation of the SC and the FN. However, 347 stimulation of the contralateral abducens nucleus occasionally evoked large 348 hyperpolarizations or small depolarizations followed by large hyperpolarizations, as shown in 349 Fig. 3C14. These potentials were not evoked at low stimulus intensities, but rather at 350 intensities as high as 400–500 A. Although the origin of these hyperpolarizations is 351 unknown, these hyperpolarizing potentials could be a useful indicator of whether or not 352 enough Cl was injected into the cell by the passage of hyperpolarizing current. After the 353 intracellular injection of Cl, early depolarizations were not changed, but later 354 hyperpolarizations were reversed to depolarizing potentials (Fig. 3D14). The difference 355 between the evoked potentials before and after Cl injection (Fig. 3E) indicated that the 356 reversed component was produced by IPSPs (Eccles 1964). After the IPSPs were reversed to 357 depolarizing potentials with Cl injection, the configurations of depolarizing potentials evoked 358 by either the contralateral FN (Fig. 3F, 2-4) or ipsilateral FN (Fig. 3F, 5-7) did not change (Fig. 359 3G, 2-4 and Fig. 3G, 5-7, respectively). Therefore, these FN-evoked depolarizations were 360 considered to be EPSPs (Eccles 1964). Similarly, the configurations of depolarizing potentials 361 evoked by stimulation of the contralateral (Fig. 3H10) and ipsilateral SC (Fig. 3H11) before 362 Cl injection did not change after Cl injection (Fig. 3I10 and Fig. 3I11, respectively), 363 indicating that the essential component of these SC-evoked depolarizing potentials was 364 EPSPs. Similar findings were confirmed in five reticular cells, where hyperpolarizations 365 evoked by stimulation of the contralateral abducens nucleus were available. 366 367 Distribution of stimulation sites in and around the caudal FN for evoking EPSPs 368 To examine stimulation sites in the caudal FN that effectively evoked EPSPs in reticular 369 neurons, we compared EPSPs evoked in reticular neurons by stimulation of the rostral (Fig. 370 Takahashi JNP Page-16 4C), middle (Fig. 4D) and caudal parts (Fig. 4E) of the caudal FNs in three experiments. 371 Figure 4A shows a reconstruction of the three stimulating electrode tracks in the contralateral 372 caudal FN in the three different experiments. The photomicrograph shown in Fig. 4B shows a 373 parasagittal section of the caudal FN which included the most effective position for the 374 stimulating electrodes shown in Fig. 4D. Stimulation of the middle of the contralateral caudal 375 FN evoked EPSPs from three sites (Fig. 4D, 2-4) with slight predominance for the most 376 ventral site (site 2). In this case, the three stimulation sites were well located within the caudal 377 FN. In contrast, in the other two cases (Fig. 4, C and E), all of the three stimulating electrodes 378 were not located within the FN, but some electrodes were located in the white matter just 379 outside the caudal FN. In Fig. 4C, the largest EPSPs were evoked from the most ventral site 380 (Fig. 4C2) where the stimulating electrode was located within the FN, but only smaller EPSPs 381 were evoked from the other two more dorsal sites where the stimulating electrodes were 382 located in the white matter just dorsal to the FN (Fig. 4C, 3 and 4). In Fig. 4E, larger EPSPs 383 were evoked from the most dorsal site (Fig. 4E4) where the electrode was located within the 384 FN, and slightly smaller EPSPs were evoked from the more ventral sites (Fig. 4E, 2 and 3) 385 where the stimulating electrodes were located in the white matter adjacent to the caudal 386 border of the FN. A similar tendency was observed when the ipsilateral caudal FN was 387 stimulated in the same three experiments (Fig. 4C, 5-7, D, 5-7 and E, 5-7). In the ipsilateral 388 caudal FN, effective stimulation sites were also located within the caudal FN, which were 389 similar to those in the contralateral FN, since the contraand ipsilateral stimulating electrodes 390 were fixed in parallel at the same rostrocaudal level and inserted into the bilateral FNs 391 simultaneously on a single manipulator. The results obtained in these three experiments 392 confirmed that the monosynaptic EPSPs were evoked in NRG neurons by stimulation of the 393 caudal FN rather than the white matter surrounding the caudal FN. 394 395 Takahashi JNP Page-17 Identification of target neurons of caudal fastigioreticular neurons in the rostral NRG 396 Next, we specifically searched for reticular neurons in the rostral NRG that received 397 excitation from the caudal FN. However, the rostral NRG contains not only IBNs but also 398 reticulospinal neurons (RSNs) that innervate neck and limb motoneurons (Eccles et al. 1975; 399 Isa and Sasaki 2002; Ito et al. 1970; Kakei et al. 1994; Matsuyama and Jankowska 2004; 400 Peterson et al. 1978, 1979). To examine whether NRG neurons with caudal fastigial input 401 were IBNs or RSNs, a bipolar stimulating electrode was placed in the contralateral abducens 402 nucleus for the antidromic identification of IBNs (Sugiuchi et al. 2005; Takahashi et al. 403 2005a), and two stimulating electrodes were placed in the ventral funiculus at the second 404 cervical segment (C2) of the spinal cord on the same side as the recording site (n = 6). Figure 405 5 shows an example of a reticular neuron in an identification experiment. This neuron was 406 penetrated in the NRG just caudal to the left abducens nucleus and received excitation from 407 both the contralateral (Fig. 5D, 2-4) and ipsilateral caudal FN (Fig. 5D, 5-7). EPSPs were 408 larger from the contralateral FN than from the ipsilateral FN, and were larger from the ventral 409 part (sites 2 and 5) than from more dorsal sites in the caudal FNs. This pattern of input from 410 the bilateral FNs was typical of NRG neurons. This neuron was not activated antidromically 411 from either the ipsilateral abducens nerve (Fig. 5C1) or the contralateral abducens nucleus 412 (Fig. 5C14), but was activated antidromically at 1.0 ms from the ipsilateral C2 (Fig. 5B). 413 Therefore, this penetrated neuron was not an abducens motoneuron nor a premotor IBN, but 414 instead was regarded as an RSN that projected ipsilaterally to C2. Among the 27 neurons 415 examined, 25 NRG neurons were regarded as RSNs, since they were activated antidromically 416 from C2 and their latencies were 0.8-1.3 ms (mean ± SD = 1.1 ± 0.2 ms). The NRG neurons 417 that we identified as receiving excitation from the caudal FN were not IBNs, because none of 418 them were activated antidromically from the contralateral abducens nucleus. 419 The latencies of EPSPs evoked by stimulation of the contralateral FN ranged from 0.8 420 Takahashi JNP Page-18 to 1.9 ms (1.1 ± 0.2 ms, n = 38), most of which were less than 1.4 ms (Fig. 5F). As will be 421 shown in Fig. 7, the latencies of antidromic spikes in caudal FN neurons evoked by 422 stimulation of the contralateral IBN area were 0.7-1.0 ms. Therefore, virtually all of these 423 contralateral FN-evoked EPSPs in RSNs in the rostral NRG were considered to be 424 monosynaptic. The range of latencies of the ipsilateral FN-evoked EPSPs was 0.8-2.1 ms (1.1 425 ± 0.3 ms, n = 38) (Fig. 5G). These latencies were on average comparable to those of the 426 contralateral FN-evoked EPSPs, but the latencies of the ipsilateral EPSPs tended to be slightly 427 shorter than the contralaterally-evoked EPSPs in each individual NRG neuron. 428 429 Properties of synaptic input from the SCs to NRG neurons 430 In the NRG neurons described so far (Figs. 2, 3 and 5), excitation was evoked by 431 stimulation of the bilateral SCs. As shown in the example in Fig. 2G, stimulation of the 432 contralateral SC evoked depolarizations with rather fixed-latency spikes at 300 A (Fig. 2G, 433 8-10), whereas stimulation of the ipsilateral SC evoked slowly-rising depolarizations with 434 spikes at fluctuating latencies at 300 A (Fig. 2G, 11-13). This pattern of input from both SCs 435 was characteristic of NRG neurons; NRG neurons received synaptic inputs from both SCs, 436 and the contralateral input was stronger with a shorter latency. These depolarizing potentials 437 were considered to be EPSPs, since the intracellular injection of Cl into the penetrated cell 438 did not reverse the depolarizing potentials to hyperpolarizing potentials (compare Fig. 3H, 10 439 and 11 with Fig. 3I, 10 and 11, respectively). 440 Previous studies showed that RSNs in the NRG received excitatory input from the SC, 441 but no studies have investigated the topographic input patterns from different parts of the 442 bilateral SCs to single RSNs innervating neck muscles. This information is important for 443 understanding the function of the tecto-reticulospinal system in the control of head 444 movements, since it is well known that there is a motor map for eye and head movements in 445 Takahashi JNP Page-19 the SC (Freedman and Sparks 1996; Paré et al. 1994; Roucoux et al. 1980). To examine the 446 topographic input patterns from the rostrocaudal sites in the SCs, we placed three stimulating 447 electrodes rostrocaudally along the horizontal meridian of the motor map in each SC, and 448 compared the amplitudes and latencies of EPSPs evoked from individual stimulation sites in 449 each NRG neuron. As shown before, Fig. 5D showed a typical example of an RSN that 450 received inputs from the caudal FNs. The same RSN also received excitation from the 451 bilateral SCs, but the pattern of input from the contralateral SC was very different from that 452 from the ipsilateral SC (Fig. 5E). Contralateral input was not evoked from the most rostral site, 453 and strongest caudally (site 10), whereas ipsilateral inputs were almost evenly evoked from 454 the rostral and caudal sites in the SC (sites 11–13), with a slight rostral predominance. As in 455 this example, the patterns of synaptic input from the ipsilateral and contralateral SC appeared 456 to be different in each RSN. A detailed analysis of the ipsilateral synaptic potentials revealed a 457 notch in the rising phase of the EPSPs (arrow in Fig. 5E13), indicating that there are two 458 components of EPSPs with different latencies in this RSN. Figure 6D and E show latency 459 histograms of EPSPs evoked in NRG neurons by stimulation of the contralateral (Fig. 6D) 460 and ipsilateral SC (Fig. 6E). The latencies of contra-SC-evoked EPSPs ranged from 0.7 to 1.7 461 ms (1.2 ± 0.3 ms, n = 38). Since the shortest excitation from the SC to abducens motoneurons 462 (1.5-2.5 ms) was regarded as disynaptic via EBNs in the PPRF (Izawa et al. 1999), EPSPs 463 with latencies shorter than 1.5 ms were most likely monosynaptic in most NRG neurons. In 464 contrast, the latency of the excitation from the ipsilateral SC was usually longer than that from 465 the contralateral SC. The latencies of ipsi-SC-evoked EPSPs were distributed into two groups, 466 one ranging from 1.1 to 1.6 ms (n = 21) and the other ranging from 1.8 to 2.6 ms (n = 17) (1.7 467 ± 0.4 ms, n = 38) (Fig. 6F). Therefore, the early group consisted of monosynaptic EPSPs and 468 the late group was considered to be disynaptic. 469 Another example of SC input to an RSN is shown in Fig. 6. This RSN received stronger 470 Takahashi JNP Page-20 excitation from three ventrodorsal sites in the contralateral FN (Fig. 6B, 2-4) and weaker 471 excitation from the two ventral sites in the ipsilateral FN at 300 A (Fig. 6B, 5 and 6). The 472 same RSN received excitation from the three sites in the contralateral SC, and the excitation 473 was larger from the more caudal SC at a latency of 0.8 ms (Fig. 6C10). On the other hand, the 474 ipsilateral excitation was largest from the most rostral SC at a latency of 1.7 ms (Fig. 6C11), 475 and during double-pulse stimulation of the same site, the EPSPs evoked from the second 476 stimuli were much larger than those evoked from the first stimuli (Fig. 6C11, dotted line in 477 the middle traces). The presence of this temporal facilitation indicates that these EPSPs were 478 evoked disynaptically from the ipsilateral SC. Since the projection from the contralateral SC 479 to the NRG was extensive (Fig. 1Db), monosynaptic excitation was considered to be directly 480 conveyed by tectoreticular neurons via the predorsal bundle. However, the projection from the 481 SC to the NRG on the same side was very weak (Fig. 1Db). Therefore, we sought to 482 determine which pathway conveyed the ipsilateral excitation to NRG neurons. The caudal FN 483 has been shown to project to the NRG and the SC (Hirai et al. 1982; May and Hall 1986; May 484 et al. 1990). If single neurons in the caudal FN project to both the SC and NRG on the 485 opposite side with their axon collaterals, stimulation of the SC will cause monosynaptic 486 excitation in ipsilateral NRG neurons by the axon reflex with single FN fibers. 487 To examine the above possibility, we recorded the activity of a caudal FN neuron and 488 searched for spikes that were activated antidromically by stimulation of the contralateral 489 rostral NRG (sites 9 and 10 in Fig. 7A). Once the neuron was activated antidromically, we 490 further examined whether the same neuron was activated from the contralateral caudal NRPc 491 (sites 7 and 8) and the bilateral SCs (sites 1–6). Figure 7, B-G shows an example of a caudal 492 FN neuron that projected to both the SC and the NRG. This neuron was activated 493 antidromically from the rostral part of the contralateral SC at a latency of 1.2 ms (Lt) (Fig. 494 7B4) and from the contralateral NRG at a latency of 0.9 ms (Lc) (Fig. 7B, 9 and 10), but was 495 Takahashi JNP Page-21 not activated from the contralateral NRPc, even at 500 A (Fig. 7B, 7 and 8). As in this 496 example, caudal FN neurons that were activated antidromically from the NRG were not 497 activated from the NRPc, except for one neuron, and the latencies of antidromic spikes 498 evoked from the NRG ranged from 0.7 to 1.0 ms (0.8 ± 0.2 ms, n = 15). 499 When single FN neurons were activated from both the rostral NRG and the SC on the 500 opposite side as in this example, we performed a spike collision test (Shinoda et al. 1976, 501 1977) to calculate the conduction times of spikes along axon collaterals of an FN neuron, as 502 shown in Fig. 7G. Spikes followed double-pulse stimuli of the SC at an interval of 0.7 ms (Rt) 503 (Fig. 7Cc) and those of the NRG at an interval of 0.5 ms (Rc) (Fig. 7Dc). When stimuli for the 504 right NRG were applied 1.4 ms and 1.5 ms (Itc) after stimuli for the right SC, spike collision 505 did and did not occur, respectively (Fig. 7E). When stimuli for the NRG were applied 1.5 ms 506 and 1.6 ms (Ict) before stimuli for the SC, spike collision did and did not occur, respectively 507 (Fig. 8F). Using these measured values, the conduction times of spikes along axon branches 508 from the branching point of the stem axon to the SC (Xt) and the NRG (Xc) were calculated 509 based on the following equations (see Shinoda et al. (1976) for derivation of the equations) : 510 Xt = 1/2 (Ict + Lt – Lc – Rt) = 0.7 511 512 Xc = 1/2 (Itc + Lc – Lt – Rc) = 0.4 513 514 Therefore, the conduction time of spikes from the SC to the NRG on the same side was 1.1 515 ms (= Xt + Xc) in this neuron. Similarly, this conduction time in eight caudal FN neurons 516 ranged from 0.8 to 1.2 ms (1.0 ± 0.1 ms). This finding suggests that the monosynaptic 517 excitation evoked from the ipsilateral SC in reticular neurons in the NRG might be caused by 518 the axon reflex of single FN neurons with axon collaterals to both the SC and the NRG. 519 As shown in the examples in Figs. 2G, 5E and 6C, the general pattern of input from the 520 SCs to NRG neurons was that (1) the inputs from the bilateral SCs were both excitatory, but 521 not reciprocal, (2) the input from the contralateral SC was larger than that from the ipsilateral 522 Takahashi JNP Page-22 SC, (3) the contralateral input was larger from the more caudal SC, (4) the ipsilateral input 523 was larger from the more rostral SC or was almost equal from the rostrocaudal sites in the SC, 524 and (5) contra-SC-evoked EPSPs had monosynaptic latencies but the ipsi-SC-evoked EPSPs 525 had both monosynaptic and disynaptic latencies. However, another type of input pattern from 526 the bilateral SCs was found in an RSN in Fig. 8. This RSN received excitation from the 527 caudal FNs. The synaptic input from the ipsilateral SC was as usual; i.e., EPSPs were evoked 528 from all rostrocaudal sites, but larger EPSPs were evoked from the more rostral sites (Fig. 8D, 529 11-13). Double-pulse stimulation of the same site (site 11) showed temporal facilitation in 530 which the second stimuli evoked larger EPSPs than the first stimuli (Fig. 8D11, the bottom 531 traces). The synaptic input from the contralateral SC was different from the usual topographic 532 pattern; EPSPs were evoked from all rostrocaudal sites and larger EPSPs were evoked from 533 more rostral site (Fig. 8D8). Therefore, this RSN received excitation from throughout the 534 rostrocaudal extent of the SCs with rostral dominance. This pattern of synaptic input from the 535 SCs was found in three RSNs. 536 537 Comparison of fastigial and tectal inputs to NRPc and NRG neurons 538 To compare synaptic inputs from the caudal FN to neurons in the caudal NRPc and the 539 rostral NRG, we recorded intracellular potentials from reticular neurons in the NRPc just 540 rostral to the abducens nucleus. Figure 9 shows such an NRPc neuron. Stimulation of the 541 contralateral FN evoked very small EPSPs (Fig. 9B2), and stimulation of the ipsilateral FN 542 did not evoke any EPSPs in this NRPc neuron (Fig. 9B, 5-7). Since stimulating electrodes 543 were placed in the middle of each caudal FN, NRG neurons had much larger EPSPs from the 544 caudal FN on either side in the same animal. Therefore, we excluded the possibility that this 545 weak fastigial input was due to the ill-placed stimulating electrodes in the caudal FN. 546 Intracellular penetration into this cell was excellent, since stimulation of the contralateral SC 547 Takahashi JNP Page-23 evoked EPSPs with spikes at 500 A, and clear EPSPs from all three rostrocaudal electrodes 548 even at 100 A, with the largest EPSPs from the caudal electrode (Fig. 9C). However, 549 short-latency synaptic inputs from the ipsilateral SC were not evoked in this NRPc neuron and 550 only late small depolarizations were observed (Fig. 9C, 11-13). Figure 10 shows another 551 pattern of synaptic input from the bilateral SCs to an NRPc neuron. In this neuron, stimulation 552 of the caudal FNs did not evoke large EPSPs, but stimulation of the most ventral site in the 553 contralateral FN evoked very small EPSPs (Fig. 10C2). EPSPs with similar amplitudes were 554 evoked from all the rostrocaudal sites in the ipsilateral SC (Fig. 10D, 11-13), but larger EPSPs 555 with spikes were evoked from the more caudal site in the contralateral SC (Fig. 10D10). 556 Out of 20 NRPc neurons examined, 18 neurons were antidromically activated from the 557 ipsilateral cervical spinal cord (C2), and none of them were antidromically activated from the 558 ipsilateral abducens nerve or the contralateral abducens nucleus. The effect of stimulation of 559 the ipsilateral abducens nucleus was not examined in this study. Among the 20 NRPc neurons, 560 5 had very small EPSPs from the contralateral caudal FN, as shown in Fig. 9B2. The other 561 neurons did not show any short-latency EPSPs. Clear EPSPs could not be evoked in any of 562 these NRPc neurons from the ipsilateral caudal FN. As to the tectal inputs in the NRPc 563 neurons, large EPSPs were always evoked from the contralateral SC, and the amplitudes of 564 the evoked EPSPs were largest and their latencies were shortest from the most caudal site (site 565 10), as exemplified in the two NRPc neurons. In contrast, the pattern of synaptic input from 566 the ipsilateral SC was varied; one group of NRPc neurons had no traceable EPSPs and the 567 other group had EPSPs with nearly equivalent amplitudes from all rostrocaudal sites in the 568 ipsilateral SC. The latencies of contra-SC-evoked EPSPs were 1.2-1.8 ms (1.6 ± 0.2 ms, n = 569 20), and those of ipsi-SC-evoked EPSPs were 1.4-2.5 ms (1.7 ± 0.3 ms, n = 8). 570 Takahashi JNP Page-24 DISCUSSION 571 The present study has characterized the patterns of synaptic inputs from the bilateral 572 caudal FNs and SCs to cells in the rostral NRG and the caudal NRPc (Fig. 11A). Neurons in 573 the rostral NRG received monosynaptic excitation from both caudal FNs, with contralateral 574 predominance. None of these NRG neurons with fastigial excitation were activated 575 antidromically from the contralateral abducens nucleus, but virtually all of them were RSNs, 576 since they were activated antidromically from the second cervical spinal cord. In contrast, 577 neurons in the caudal NRPc received only very much weaker excitation from the contralateral 578 caudal FN and no excitation from the ipsilateral caudal FN. Virtually all of them were RSNs. 579 The same NRG and NRPc neurons received excitation from the bilateral SCs. NRG neurons 580 received stronger monosynaptic excitation from the more caudal contralateral SC, and 581 stronger disynaptic excitation from the more rostral ipsilateral SC. NRPc neurons received 582 either monosynaptic excitation only from the contralateral caudal SC, or monosynaptic 583 excitation from the contralateral SC and stronger disynaptic excitation from the more rostral 584 ipsilateral SC. This difference in fastigial and tectal synaptic input patterns between NRG and 585 NRPc neurons may reflect different functional roles of NRG and NRPc neurons in the control 586 of head movements. We will first summarize the present results in relation to previous studies, 587 and then discuss the presumed functional roles of RSNs in the NRG and NRPc that had 588 different patterns of synaptic input from the bilateral caudal FNs and topographic input from 589 the bilateral SCs in the control of head movements. 590 591 Properties of synaptic input from the caudal FN to NRG and NRPc neurons 592 Ito et al. (1970) recorded intracellular potentials while stimulating the rostral FN and 593 showed large monosynaptic EPSPs in RSNs in the medullary reticular formation. The 594 present intracellular recordings confirm their finding. Mori et al. (1998) activated 595 Takahashi JNP Page-25 fastigiobulbar fibers in the hook bundle, and observed extracellular spike activity at the 596 monosynaptic range in RSNs in the medullary NRG. Eccles et al. (1975) stimulated the FN 597 and found increased spike activity at the monosynaptic range in pontine and medullary RSNs. 598 In both experiments, the hook bundle or the rostral FN was stimulated. Since stimulation of 599 the hook bundle and the rostral FN activates fibers arising from rostral and caudal fastigial 600 neurons, none of these studies could determine whether caudal FN axons directly terminate on 601 reticular neurons in the NRPc and NRG. 602 The present study identified different patterns of input in NRG and NRPc neurons 603 following caudal FN stimulation. NRG neurons had monosynaptic excitation from the 604 contralateral caudal FN, and smaller monosynaptic excitation with higher thresholds from the 605 ipsilateral caudal FN. In contrast, NRPc neurons received little or no excitation from the 606 contralateral FN, and no excitation from the ipsilateral FN. The present findings are consistent 607 with the anatomical finding, because the caudal fastigial projection to the rostral NRG appears 608 to be considerably stronger than that to the caudal NRPc (Homma et al. 1995; Noda et al. 609 1990). Ipsilateral stimulation of the caudal FN was much less effective for evoking excitation 610 in NRG and NRPc neurons than contralateral stimulation. This finding corresponds to the 611 anatomical evidence that axon terminals of the FN in the NRG and NRPc were dense 612 contralaterally and extremely sparse ipsilaterally (Homma et al. 1995). Although we cannot 613 exclude the possibility of an ipsilateral caudal FN projection to the NRG, fibers passing from 614 the contralateral FN through the ipsilateral FN may be the most likely source of the activity 615 observed in ipsilateral FN stimulation (Batton et al. 1977; Ito et al. 1970). The shorter latency 616 of ipsi-FN-evoked EPSPs vs. contra-FN-evoked EPSPs supports this interpretation (Fig. 5G). 617 Nonetheless, we cannot rule out the possibility that some of the effects we observed were 618 caused by the axon reflex of mossy fiber neurons other than primary vestibular neurons. 619 It has been well established that the NRPc and NRG are the origins of different types of 620 Takahashi JNP Page-26 reticulospinal tracts (Holstege and Kuypers 1982; Isa and Sasaki 2002; Ito et al. 1970; 621 Peterson et al. 1974). RSNs are functionally classified into two groups in the NRPc and NRG. 622 RSNs in the NRPc receive stronger input from the SC than from the motor cortex, whereas 623 RSNs in the NRG receive stronger input from the motor cortex than from the SC (Alstermark 624 et al. 1985, 1992a, b; Iwamoto and Sasaki 1990; Iwamoto et al. 1990). The present finding 625 indicates that NRG and NRPc neurons with differential fastigial inputs have different 626 functional roles in the control of head movements. Grantyn and others have reported that 627 single RSNs in the NRPc with axon collaterals to both the abducens nucleus and the cervical 628 cord (Grantyn et al. 1987) are involved in the control of synergistic eye and head movements 629 during orienting movements (Grantyn and Berthoz 1987). The caudal FN may control eye and 630 head movements separately, since RSNs in the NRPc receive only little or no input from the 631 caudal FN. In contrast, RSNs in the NRG receive stronger cerebral input than those in the 632 NRPc, and lobule VII and the caudal FN are influenced by the cerebral input, suggesting that 633 RSNs in the NRG are more involved in volitional head movements. 634 635 Head movements controlled by the caudal FN 636 Although the caudal FN is known to play an important role in the control of eye 637 movements, the present results indicate that the caudal FN is also associated with head 638 movements via a fastigio-reticulo-spinal pathway. There were few investigations of the caudal 639 FN and head movements. Recently, Goffart and his colleagues showed that injection of 640 muscimol into the caudal FN caused ipsilateral deviation of the head in the cat (Goffart and 641 Pelisson 1998a) and the monkey (Quinet and Goffart 2005). In both species, gaze shifts are 642 dysmetric after caudal FN inactivation; i.e., ipsilateral hypermetria and contralateral 643 hypometria (Goffart et al. 1998a, b). The gaze dysmetria is associated with changes in 644 eye-head coupling that are vivid in the monkey but not in the cat. 645 Takahashi JNP Page-27 The synaptic connection between the caudal FN and RSNs found in the present study 646 helps explain the head movement deficits observed after muscimol inactivation of the caudal 647 FN. Since cerebellar nucleus neurons have high spontaneous activity (Thach 1967), RSNs that 648 receive strong excitatory influence from the caudal FN are active in an upright position at rest 649 before movements (Eccles et al. 1975; Isa and Naito 1995). When this excitatory influence is 650 eliminated after inactivation, spontaneous activity of the RSNs and, in turn, the activity of 651 neck motoneurons will decrease on the non-injected side, so that the neck muscle tonus on the 652 same side decreases and the head deviates to the injected side. Although head 653 movement-related neurons have not been discovered yet in the fastigial nucleus, we speculate 654 that, if such neurons fire in a manner similar to saccade-related FN neurons (Fuchs et al. 655 1993; Ohtsuka and Noda 1991), these neurons will increase their activity at the end of 656 ipsilateral head movements and at the beginning of contralateral head movements. Under 657 these conditions, an increase in spike activity in RSNs on the side opposite to the injected side 658 will not occur at the end of head movements ipsilateral to the injected side. Consequently, 659 appropriate breaking force will not be generated by neck muscles on the non-injected 660 contralateral side, and ipsilateral hypermetria of the head movement will result. This 661 prediction is confirmed by observations in the cat (Goffart et al. 1998), but not in the monkey 662 (Quinet and Goffart 2007). On the other hand, at the beginning of contralateral head 663 movements, an increase in spike activity in RSNs on the side opposite to the injected side will 664 not occur. Thus, the turning force of the head toward the non-injected side will decrease, and 665 contralateral hypometria of the head movement will result as shown in the cat and monkey 666 (Goffart et al. 1998; Quinet and Goffart 2007). Further information on the activity of caudal 667 FN neurons during head movements will be needed to understand whether this interpretation 668 is correct. 669 670 Takahashi JNP Page-28 Neural pathways from the SCs to NRG and NRPc neurons 671 We characterized the patterns of input from the different rostrocaudal locations of the 672 SC to NRPc and NRG neurons. Both NRPc and NRG neurons received strong monosynaptic 673 excitation from the contralateral SC. Huerta and Harting (1982) showed anatomical evidence 674 that tectofugal neurons terminate in the dorsal portion of the NRPc, and as the predorsal 675 bundle passes caudally, distribute throughout the NRG. Grantyn and Grantyn (1982) showed 676 that single tectofugal neurons have multiple axon collaterals in the brainstem on their way to 677 the spinal cord. Most of the contralateral SC-evoked EPSPs in the NRG neurons were 678 monosynaptic (Fig. 6D), whereas the ipsilateral SC-evoked EPSPs were monosynaptic and 679 disynaptic (Fig. 6E). However, since the present anatomical data showed that ipsilateral 680 projection from the SC to the rostral NRG which contains IBNs (Fig. 1Db) and the caudal 681 NRPc which contains EBNs (Fig. 1Da) was extremely sparse, the ipsilateral monosynaptic 682 excitation of RSNs was not likely to be conveyed by direct tectoreticular axons. One of the 683 possible sources for this monosynaptic excitation was the axon reflex activity within FN 684 fibers projecting to both the SC and the NRG, since caudal fastigial neurons extensively 685 project to the rostral SC (Hirai et al. 1982; May and Hall 1986; May et al. 1990). This 686 possibility was confirmed by spike collision experiments (Fig.7) in which the calculated 687 values were in agreement with the latencies of the ipsi-SC-evoked excitation in NRG neurons. 688 Other possible monosynaptic pathways from the ipsilateral SC to NRG and NRPc neurons 689 remain undetermined. 690 Inactivation of the mesencephalic reticular formation produced severe deficits in head 691 posture (Fukushima 1987; Hess 1956; Klier et al. 2001; Waitzman et al. 2000). Consistent 692 with these lesion experiments, tectoreticular neurons (TRNs) from the rostral SC extensively 693 terminate on FFH and INC neurons on the same side which mediate saccade signals to 694 vertical ocular motoneurons (Sugiuchi et al. 2013). Similarly, they terminate on FFH and INC 695 Takahashi JNP Page-29 neurons that mediate head movement signals to the cervical cord (Isa and Naito 1994). Some 696 FFH neurons have axon collaterals in the caudal NRPc and rostral NRG (Isa and Sasaki 1992). 697 Therefore, it is most likely that the present disynaptic excitatory input from the ipsilateral SC 698 is conveyed via FFH and INC neurons to RSNs in the NRG that terminate on neck 699 motoneurons. This trisynaptic pathway from the SC to neck motoneurons parallels the 700 disynaptic pathway from the SC via spinally-projecting FFH neurons to neck motoneurons 701 (Alstermark et al. 1992a). It may be functionally important that, in the pathways via RSNs in 702 the NRG, the cerebral and tectal inputs to neck motoneurons can be modulated at RSNs in the 703 NRG by control of the caudal FN. 704 705 Topographic and convergent patterns of input from the SCs to NRG and NRPc neurons 706 Topographic input patterns from different sites in the motor map of the SC and 707 convergent input patterns from the SCs have not yet been systematically examined in single 708 RSNs in the rostral NRG and the caudal NRPc, using intracellular recordings. Virtually all 709 NRG neurons and some NRPc neurons receive bilateral excitatory inputs from the SCs, and 710 some NRPc neurons receive only excitation from the contralateral SC. These patterns of input 711 from the SCs to RSNs in the NRG and NRPc were very different from those observed in 712 brainstem neurons in the saccade-generating system, since EBNs, IBNs and abducens 713 motoneurons always receive excitation from the contralateral SC and inhibition from the 714 ipsilateral SC (Izawa et al. 1999; Sugiuchi et al. 2005). Another difference in the tectal input 715 pattern lies in the strength of synaptic inputs from the rostral and caudal sites in the SCs. In 716 EBNs, IBNs and abducens motoneurons, synaptic input is stronger from more caudal SCs, 717 whereas in NRG and NRPc neurons, synaptic input is stronger from the rostral part in the 718 ipsilateral SC and usually stronger from the caudal part in the contralateral SC. 719 The SC plays an important role in the control of orienting eye and head movements. 720 Takahashi JNP Page-30 According to a motor map for representing eye movements in the SC (Guitton et al. 1980; 721 McIlwain 1986; Roucoux et al. 1980), large horizontal saccades are represented in the caudal 722 SC, and large upward and downward saccades in the medial and lateral parts of the rostral SC, 723 respectively. TRNs projecting to the FFH and INC are distributed in the rostral SC where 724 vertical eye movements are represented as a vertical meridian (Sugiuchi et al. 2013; Takahashi 725 et al. 2007, 2010, 2011), whereas TRNs projecting to horizontal EBNs and IBNs are 726 distributed throughout the entire rostrocaudal extent of the horizontal meridian (Takahashi et 727 al. 2010). When the head is free to move, microstimulation of the SC in cats (Guillaume and 728 Pelisson 2001; Paré et al. 1994) and in primates (Freedman et al. 1996; Klier et al. 2001) 729 evokes coordinated eye-head gaze shifts. Therefore, stimulation of a particular part of the SC 730 may induce eye and head movements in the same plane, as stimulation of a particular 731 semicircular canal nerve induces eye and head movements in the same plane as the stimulated 732 semicircular canal (Suzuki and Cohen, 1964). The discharge of single neurons in the SC is 733 better correlated with gaze than with either eye or head movements (Freedman and Sparks 734 1997; Munoz et al. 1991), and the SC issues command signals related to coordinated eye-head 735 gaze shifts in the cat (Guillaume and Pelisson 2006; Matsuo et al. 2004; Pelisson et al. 2001) 736 and the monkey (Corneil et al. 2002; Freedman et al. 1996; Klier et al. 2001). However, 737 inactivation of the SC only occasionally resulted in slight increases in the reaction times of 738 head movements associated with gaze shifts (Walton et al. 2008). Based on these previous 739 findings on the SC topography for gaze, it is likely that RSNs in the NRPc with predominant 740 input from the more caudal SC and no ipsilateral input may be related to horizontal head 741 movements, whereas RSNs in the NRG and some RSNs in the NRPc with predominant inputs 742 from the bilateral more-rostral SCs may be related to vertical head movements. The 743 innervation pattern of single RSNs by the bilateral SCs is very similar to that of single neck 744 motoneurons in the vestibulocollic system. In the vestibulocollic system, motoneurons of all 745 Takahashi JNP Page-31 neck muscles examined receive excitation from the contralateral horizontal semicircular canal, 746 whereas extensor and flexor neck muscle motoneurons receive excitation from the bilateral 747 anterior and posterior semicircular canals, respectively (Shinoda et al. 1994, 1997; Sugiuchi et 748 al. 2004). All NRG and NRPc RSNs with strong excitation from the contralateral caudal sSC 749 are considered to be related to horizontal head movements, because all dorsal neck muscles 750 on one side contract together during voluntary horizontal head movements (Corneil 2001), as 751 well as in the vestibulocollic reflex, and RSNs in the NRPc with axon collaterals to both 752 abducens nucleus are associated with horizontal head movements (Grantyn and Berthoz 1987). 753 Similarly, NRG and NRPc RSNs with predominant input from the rostral SCs are considered 754 to be involved in vertical head movements. This interpretation is in agreement with the 755 following findings. Since the NRG receives a substantial descending projection from the 756 caudal vestibular nuclei, Fagerson and Barmack (1995) examined the peripheral origins of the 757 vestibularly-modulated activity in NRG neurons, and found that more than 85% of them 758 responded to vertical vestibular stimulation, indicating that NRG neurons are related to 759 vertical head movements. Furthermore, there were a variety of RSNs in the NRPc and NRG 760 that had horizontal, upward oblique, or downward oblique preferred directions of ipsilateral 761 head movements (Isa and Naito 1995). In our previous studies, we demonstrated that the 762 reciprocal inhibitory patterns of the SC saccade system and the vestibuloocular system were 763 similar, implying that the SC saccade output system may use the semicircular canal 764 coordinates, like the vestibuloocular system (Sugiuchi et al. 2013; Takahashi et al. 2007, 2010, 765 2011). The present results strongly suggest that the SC head movement system may also use 766 the semicircular canal coordinates. 767 768 Neural implementation of Listing’s law in head movements 769 Voluntary eye movements do not have torsional components, and have only horizontal 770 Takahashi JNP Page-32 and vertical components. This is referred to as Listing’s law (von Helmholtz 1867). For 771 vertical upward eye movements, for example, it is most likely that TRNs in the medial parts 772 of the two rostral SCs fire simultaneously to cancel torsional components of eye movements 773 in opposite directions induced by the two SCs (Takahashi et al. 2007). This simultaneous 774 activation of the bilateral medial SCs is caused by excitatory commissural connections 775 between the symmetric parts of the rostral SCs, which are considered to be responsible for 776 implementing Listing’s law in saccadic eye movements (Takahashi et al. 2005a, 2010, 2011). 777 Listing’s law also applies to head movements (Straumann et al. 1991; Tweed and Vilis 778 1988). Rostral TRNs projecting to the FFH have commissural excitation from the symmetric 779 part of the contralateral SC (Takahashi et al. 2007). The tectal excitatory commissural 780 connections in the head movement system are responsible for an increase of activity in single 781 FFH neurons in not only contralateral, but also ipsilateral oblique upward head movements in 782 the cat (Isa and Naito 1994), since the contralateral projection of the SC to the FFH is 783 extremely weak (Sugiuchi et al. 2013). The present result indicates that one group of RSNs in 784 the NRG and NRPc receive convergent excitatory inputs from the rostral parts of the bilateral 785 SCs (Fig. 11A). To view it another way, TRNs in a particular part of the rostral SC send their 786 outputs to RSNs in the symmetric parts of the NRG and NRPc on both sides, and the RSNs in 787 turn project to motoneurons innervating identical neck muscles, either extensors on both sides 788 or flexors on both sides (Fig. 11B). These neural connections suggest that when a cat executes 789 oblique vertical head movements to the contralateral side, the rostral SC sends out command 790 signals via bilateral RSNs to identical neck muscles on both sides. In addition, the same part 791 of the SC may activate TRNs in the symmetric part of the opposite SC. Consequently, TRNs 792 in the symmetric rostral parts of the bilateral SCs give rise to command signals, and contract 793 identical functional groups of neck muscles on both sides, so that oppositely-directed 794 torsional components of head movements induced by individual SCs cancel out each other, 795 Takahashi JNP Page-33 and the SC may result in generating head movements that conform to Listing’s law. Therefore, 796 the present neural connections including the tectal excitatory commissural connection are 797 considered to provide a neural basis for implementing Listing’s law in head movements. 798 799 Targets of NRPc and NRG neurons with large caudal fastigial input 800 It is assumed that fastigioreticular axons arising from the caudal FN directly innervate 801 burst neurons in the brainstem saccade generator, especially IBNs, on the contralateral side 802 via the hook bundle (Fuchs et al. 1993; Ohtsuka and Noda 1991, 1995), although the 803 experimental evidence is still lacking. In the present study, we searched in the cat for neurons 804 in the rostral NRG that received monosynaptic excitation from the contralateral FN. These 805 neurons were not activated antidromically from the contralateral abducens nucleus, and 806 instead most of them were activated antidromically from the ipsilateral cervical spinal cord. If 807 the penetrated neurons had been IBNs, they should have been activated antidromically from 808 the contralateral abducens nucleus (Hikosaka and Kawakami 1977; Sugiuchi et al. 2005; 809 Yoshida et al. 1982). If the connection between the caudal FN and contralateral IBNs exists, 810 ipsilateral FN stimulation should evoke disynaptic IPSPs via contralateral IBNs in penetrated 811 neurons in addition to the monosynaptic EPSPs observed in the present study. However, 812 IPSPs were not observed in the NRG neurons. Furthermore, the NRG neurons with large FN 813 input had only excitatory inputs from the bilateral SCs, which also supports the finding that 814 the examined neurons were not IBNs, since IBNs received excitation from the contralateral 815 SC and inhibition from the ipsilateral SC (Sugiuchi et al. 2005; Takahashi et al. 2005b; 816 Yoshida et al. 1982). Therefore, the present study could not confirm that contralateral caudal 817 FN neurons directly terminate on IBNs. Unlike a widely accepted assumption in the field, it 818 remains to be demonstrated whether IBNs receive direct excitation from the contralateral FN. 819 Since IBNs and EBNs are smaller than large RSNs, it might be more difficult to make stable 820 Takahashi JNP Page-34 intracellular recordings from these neurons. Further experiments might be needed to reliably 821 conclude that this direct connection exists by recording intracellular potentials from 822 electrophysiologically-identified IBNs and examining the nature of synaptic inputs evoked by 823 stimulation of the caudal FNs, since the existence or absence of this pathway is essential for 824 understanding the neural mechanism of the caudal FN in the cerebellar control of saccadic eye 825 movements. 826 827 Takahashi JNP Page-35 References828Adams JC. Heavy metal intensification of DAB-based HRP reaction Product. J Histochem829Cytochem 29: 775, 1981.830Alstermark B, Pinter MJ, Sasaki S. Pyramidal effects in dorsal neck motoneurones of the831cat. J Physiol (Lond) 363: 287-302, 1985.832Alstermark B, Pinter MJ, Sasaki S. Tectal and tegmental excitation in dorsal neck833motoneurones of the cat. J Physiol (Lond) 454: 517-532, 1992a.834Alstermark B, Pinter MJ, Sasaki S. Descending pathways mediating disynaptic excitation835of dorsal neck motoneurones in the cat: brain stem relay. Neurosci Res 15: 42-57, 1992b.836Angaut P, Bowsher D. Ascending projections of the medial cerebellar (fastigial) nucleus: an837experimental study in the cat. Brain Res 24: 49-68, 1970.838Asanuma C, Thach WT, Jones EG. Brainstem and spinal projections of the deep cerebellar839nuclei in the monkey, with observations on the brainstem projections of the dorsal column840nuclei. Brain Res 286: 299-322, 1983.841Baker R, Highstein SM. Physiological identification of interneurons and motoneurons in the842abducens nucleus. Brain Res 91: 292-298, 1975.843Batton RR IIIrd, Jayaraman A, Ruggiero D, Carpenter MB. Fastigial efferent projections844in the monkey: an autoradiographic study. J Comp Neurol 174: 281-305, 1977.845Brodal A. The Reticular Formation of the Brain Stem. Anatomical Aspects and Functional846Correlations. Illinois, USA: Charles C Thomas Pub, 1956.847Büttner U, Fuchs AF, Markert-Schwab G, Buckmaster P. Fastigial nucleus activity in the848alert monkey during slow eye and head movements. J Neurophysiol 65: 1360-1371, 1991.849Carpenter MB, Brittin GM, Pines J. Isolated lesions of the fastigial nuclei in the cat. J850Comp Neurol 109: 65-89, 1958.851Carpenter MB, Nova HR. Descending division of the brachium conjunctivum in the cat:852 Takahashi JNP Page-36 a cerebello-reticular system. J Comp Neurol 114: 295-305, 1960.853Chambers WW, Sprague JM. Functional localization in the cerebellum. I. Organization in854longitudinal cortico-nuclear zones and their contribution to the control of posture, both855extrapyramidal and pyramidal. J Comp Neurol 103: 105-129, 1955.856Cogdell B, Hassul M, Kimm J. Fastigial evoked eye movement and brain stem neuronal857behavior in the alert monkey. Arch Otolaryngol 103: 658-666, 1977.858Cohen B, Henn V. Unit activity in the pontine reticular formation associated with eye859movements. Brain Res 46: 403-410, 1972.860Cohen D, Chambers WW, Sprague JM. Experimental study of the efferent projections from861the cerebellar nuclei to the brainstem of the cat. J Comp Neurol 109: 233-259, 1958.862Corneil BD, Olivier E, Munoz DP. Neck muscle responses to stimulation of monkey863superior colliculus. I. Topography and manipulation of stimulation parameters. J Neurophysiol86488: 1980-1999, 2002.865Corneil BD, Olivier E, Richmond FJ, Loeb GE, Munoz DP. Neck muscles in the rhesus866monkey. II. Electromyographic patterns of activation underlying postures and movements. J867Neurophysiol 86: 1729-1749, 2001.868Eccles JC. The Physiology of Synapses. Berlin: Springer-Verlag, 1964.869Eccles JC, Ito M, Szentágothai J. The Cerebellum as a Neuronal Machine. Berlin:870Springer-Verlag, 1967.871Eccles JC, Nicoll RA, Schwarz WF, Táboriková H, Willey TJ. Reticulospinal neurons with872and without monosynaptic inputs from cerebellar nuclei. J Neurophysiol 38: 513-530, 1975.873Fagerson MH, Barmack NH. Responses to vertical vestibular stimulation of neurons in the874nucleus reticularis gigantocellularis in rabbits. J Neurophysiol 73: 2378-2391, 1995.875Freedman EG, Sparks DL. Activity of cells in the deeper layers of the superior colliculus876of the rhesus monkey: evidence for a gaze displacement command. J Neurophysiol 78:877 Takahashi JNP Page-37 1669-1690, 1997.878Freedman EG, Stanford TR, Sparks DL. Combined eye-head shifts produced by electrical879stimulation of the superior colliculus in rhesus monkey. J Neurophysiol 76: 927-952, 1996.880Fuchs AF, Robinson FR, Straube A. Role of the caudal fastigial nucleus in saccade881generation. I. Neuronal discharge pattern. J Neurophysiol 70: 1723-1740, 1993.882Fuchs AF, Robinson FR, Straube A. Participation of the caudal fastigial nucleus in883smooth-pursuit eye movements. I. Neuronal activity. J Neurophysiol 72: 2714-2728, 1994.884Fujikado T, Noda H. Saccadic eye movements evoked by microstimulation of lobule VII of885the cerebellar vermis of macaque monkeys. J Physiol (Lond) 394: 573-594, 1987.886Fukushima K. The interstitial nucleus of Cajal and its role in the control of movements of887head and eyes. Prog Neurobiol 29: 107-192, 1987.888Fukushima K, Peterson BW, Uchino Y, Coulter JD, Wilson VJ. Direct fastigiospinal fibers889in the cat. Brain Res 126: 538-542, 1977.890Furuya N, Kawano K, Shimazu H. Transcerebellar inhibitory interaction between the891bilateral vestibular nuclei and its modulation by cerebellocortical activity. Exp Brain Res 25:892447-463, 1976.893Goffart L, Chen LL, Sparks DL. Deficits in saccades and fixation during muscimol894inactivation of the caudal fastigial nucleus in the rhesus monkey. J Neurophysiol 92:8953351-3367, 2004.896Goffart L, Pelisson D. Orienting gaze shifts during muscimol inactivation of caudal fastigial897nucleus in the cat. I. Gaze dysmetria. J Neurophysiol 79: 1942-1958, 1998a.898Goffart L, Pelisson D, Guillaume A. Orienting gaze shifts during muscimol inactivation of899caudal fastigial nucleus in the cat. II. Dynamics and eye-head coupling. J Neurophysiol 79:9001959-1976, 1998b.901 Takahashi JNP Page-38 Grantyn A, Berthoz A. Reticulo-spinal neurons participating in the control of synergic eye902and head movements during orienting in the cat. I. Behavioral properties. Exp Brain Res 66:903339-354, 1987.904Grantyn A, Grantyn R. Axonal patterns and sites of termination of cat superior colliculus905neurons projecting in the tecto-bulbo-spinal tract. Exp Brain Res 46: 243-256, 1982.906Grantyn A, Ong-Meang Jacques V, Berthoz A. Reticulo-spinal neurons participating in the907control of synergic eye and head movements during orienting in the cat. II. Morphological908properties as revealed by intraaxonal injections of horseradish peroxidase. Exp Brain Res 66:909355-377, 1987.910Gruart A, Delgado-Garcia JM. Signalling properties of identified deep cerebellar nuclear911neurons related to eye and head movements in the alert cat. J Physiol 471: 37-54, 1994.912Guillaume A, Pelisson D. Gaze shifts evoked by electrical stimulation of the superior913colliculus in the head-unrestrained cat. I. Effect of the locus and of the parameters of914stimulation. Eur J Neurosci 14: 1331-1344, 2001.915Guillaume A, Pelisson D. Kinematics and eye-head coordination of gaze shifts evoked from916different sites in the superior colliculus of the cat. J Physiol (Lond) 577: 779-794, 2006.917Guitton D, Crommelinck M, Roucoux A. Stimulation of the superior colliculus in the alert918cat. I. Eye movements and neck EMG activity evoked when the head is restrained. Exp Brain919Res 39: 63-73, 1980.920Helmchen C, Straube A, Buttner U. Saccade-related activity in the fastigial oculomotor921region of the macaque monkey during spontaneous eye movements in light and darkness. Exp922Brain Res 98: 474-482, 1994.923Hepp K, Henn V, Jaeger J. Eye movement related neurons in the cerebellar nuclei of the924alert monkey. Exp Brain Res 45: 253-264, 1982.925 Takahashi JNP Page-39 Hess WR. Hypothalamus und Thalamus. Experimental-Dokumente. [Hypothalamus and926thalamus, Documentary Pictures]. Stuttgart: George Thieme Verlag, 1956.927Hikosaka O, Kawakami T. Inhibitory reticular neurons related to the quick phase of928vestibular nystagmus-their location and projection. Exp Brain Res 27: 377-396, 1977.929Hirai T, Onodera S, Kawamura K. Cerebellotectal projections studied in cats with930horseradish peroxidase or tritiated amino acids axonal transport. Exp Brain Res 48: 1-12,9311982.932Holstege G, Kuypers HGJM. The anatomy of brain stem pathways to the spinal cord in cat.933A labeled amino acid tracing study. Prog Brain Res 57: 145-175, 1982.934Homma Y, Nonaka S, Matsuyama K, Mori S. Fastigiofugal projection to the brainstem935nuclei in the cat: an anterograde PHA-L tracing study. Neurosci Res 23: 89-102, 1995.936Huerta MF, Harting JK. Tectal control of spinal cord activity: neuroanatomical937demonstration of pathways connecting the superior colliculus with the cervical spinal cord938grey. Prog Brain Res 57: 293-328, 1982.939Igusa Y, Sasaki S, Shimazu H. Excitatory premotor burst neurons in the cat pontine reticular940formation related to the quick phase of vestibular nystagmus. Brain Res 182: 451-456, 1980.941Isa T, Naito K. Activity of neurons in Forel’s field H during orienting head movements in942alert head-free cats. Exp Brain Res 100: 187-199, 1994.943Isa T, Naito K. Activity of neurons in the medial pontomedullary reticular formation during944orienting head movements in alert head-free cats. J Neurophysiol 74: 73-95, 1995.945Isa T, Sasaki S. Descending projections of Forel's field H neurones to the brain stem and the946upper cervical spinal cord in the cat. Exp Brain Res 88: 563-579, 1992.947Isa T, Sasaki S. Brainstem control of head movements during orienting; organization of the948premotor circuits. Prog Neurobiol 66: 205-241, 2002.949 Takahashi JNP Page-40 Ito M, Udo M, Mano N, Kawai N. Synaptic action of the fastigiobulbar impulses upon950neurones in the medullary reticular formation and vestibular nuclei. Exp Brain Res 11: 29-47,9511970.952Iwamoto Y, Sasaki S. Monosynaptic excitatory connections of reticulospinal neurones in the953nucleus reticularis pontis caudalis with dorsal neck motoneurones in the cat. Exp Brain Res95480: 277-289, 1990.955Iwamoto Y, Sasaki S, Suzuki I. Input-output organization of reticulospinal neurones, with956special reference to connexions with dorsal neck motoneurones in the cat. Exp Brain Res 80:957260-276, 1990.958Izawa Y, Sugiuchi Y, Shinoda Y. Neural organization from the superior colliculus to959motoneurons in the horizontal oculomotor system of the cat. J Neurophysiol 81: 2597-2611,9601999.961Jansen J, Brodal A. Experimental studies on the intrinsic fibers of the cerebellum. II. The962cortico-nuclear projection. J Comp Neurol 73: 267-321, 1940.963Jansen J, Jansen J, Jr. On the efferent fibers of the cerebellar nuclei in the cat. J Comp964Neurol 102: 607-632, 1955.965Kakei S, Muto N, Shinoda Y. Innervation of multiple neck motor nuclei by single966reticulospinal tract axons receiving tectal input in the upper cervical spinal cord. Neurosci Lett967172: 85-88, 1994.968Kase M, Miller DC, Noda H. Discharges of Purkinje cells and mossy fibres in the cerebellar969vermis of the monkey during saccadic eye movements and fixation. J Physiol (Lond) 300:970539-555, 1980.971Keller EL. Participation of medial pontine reticular formation in eye movement generation in972monkey. J Neurophysiol 37: 316-332, 1974.973 Takahashi JNP Page-41 Keller EL, Slakey DP, Crandall WF. Microstimulation of the primate cerebellar vermis974during saccadic eye movements. Brain Res 288: 131-143, 1983.975Klier EM, Wang H, Crawford JD. The superior colliculus encodes gaze commands in976retinal coordinates. Nat Neurosci 4: 627-632, 2001.977Kojima Y, Iwamoto Y, Robinson FR, Noto CT, Yoshida K. Premotor inhibitory neurons978carry signals related to saccade adaptation in the monkey. J Neurophysiol 99: 220-230, 2008.979Luschei E, Fuchs AF. Activity of brain stem neurons during eye movements of alert980monkeys. J Neurophysiol 35: 445-461, 1972.981Maeda M, Shimazu H, Shinoda Y. Rhythmic activities of secondary vestibular efferent982fibers recorded within the abducens nucleus during vestibular nystagmus. Brain Res 34:983361-365, 1971.984Matsuo S, Bergeron A, Guitton D. Evidence for gaze feedback to the cat superior colliculus:985discharges reflect gaze trajectory perturbations. J Neurosci 24: 2760-2773, 2004.986Matsushita M, Hosoya Y. The location of spinal projection neurons in the cerebellar nuclei987(cerebellospinal tract neurons) of the cat. A study with the horseradish peroxidase technique.988Brain Res 142: 237-248, 1978.989Matsuyama K, Jankowska E. Coupling between feline cerebellum (fastigial neurons) and990motoneurons innervating hindlimb muscles. J Neurophysiol 91: 1183-1192, 2004.991May PJ, Hall WC. The cerebellotectal pathway in the grey squirrel. Exp Brain Res 65:992200-212, 1986.993May PJ, Hartwich-Young R, Nelson J, Sparks DL, Porter JD. Cerebellotectal pathways in994the macaque: implications for collicular generation of saccades. Neuroscience 36: 305-324,9951990.996McElligott JG, Keller EL. Cerebellar vermis involvement in monkey saccadic eye997movements: microstimulation. Exp Neurol 86: 543-558, 1984.998 Takahashi JNP Page-42 McIlwain JT. Effects of eye position on saccades evoked electrically from superior colliculus999of alert cats. J Neurophysiol 55: 97-112, 1986.1000Moolenaar GM, Rucker HK. Autoradiographic study of brain stem projections from1001fastigial pressor areas. Brain Res 114: 492-496, 1976.1002Mori S, Matsui T, Kuze B, Asanome M, Nakajima K, Matsuyama K. Cerebellar-induced1003locomotion: reticulospinal control of spinal rhythm generating mechanism in cats. Ann N Y1004Acad Sci 860: 94-105, 1998.1005Munoz DP, Guitton D, Pelisson D. Control of orienting gaze shifts by the tectoreticulospinal1006system in the head-free cat. III. Spatiotemporal characteristics of phasic motor discharges. J1007Neurophysiol 66: 1642-1666, 1991.1008Noda H, Fujikado T. Involvement of Purkinje cells in evoking saccadic eye movements by1009microstimulation of the posterior cerebellar vermis of monkeys. J Neurophysiol 57:10101247-1261, 1987a.1011Noda H, Fujikado T. Topography of the oculomotor area of the cerebellar vermis in1012macaques as determined by microstimulation. J Neurophysiol 58: 359-378, 1987b.1013Noda H, Murakami S, Yamada J, Tamaki Y, Aso T. Saccadic eye movements evoked by1014microstimulation of the fastigial nucleus of macaque monkeys. J Neurophysiol 60: 1036-1052,10151988.1016Noda H, Sugita S, Ikeda Y. Afferent and efferent connections of the oculomotor region of1017the fastigial nucleus in the macaque monkey. J Comp Neurol 302: 330-348, 1990.1018Ohtsuka K, Noda H. Saccadic burst neurons in the oculomotor region of the fastigial nucleus1019of macaque monkeys. J Neurophysiol 65: 1422-1434, 1991.1020Ohtsuka K, Noda H. Discharge properties of Purkinje cells in the oculomotor vermis during1021visually guided saccades in the macaque monkey. J Neurophysiol 74: 1828-1840, 1995.1022 Takahashi JNP Page-43 Paré M, Crommelinck M, Guitton D. Gaze shifts evoked by stimulation of the superior1023colliculus in the head-free cat conform to the motor map but also depend on stimulus strength1024and fixation activity. Exp Brain Res 101: 123-139, 1994.1025Pelisson D, Goffart L, Guillaume A, Catz N, Raboyeau G. Early head movements elicited1026by visual stimuli or collicular electrical stimulation in the cat. Vision Res 41: 3283-3294,10272001.1028Peterson BW, Anderson ME, Filion M. Responses of pontomedullary reticular neurons to1029cortical, tectal and cutaneous stimuli. Exp Brain Res 21: 19-44, 1974.1030Peterson BW, Pitts NG, Fukushima K. Reticulospinal connections with limb and axial1031motoneurons. Exp Brain Res 36: 1-20, 1979.1032Peterson BW, Pitts NG, Fukushima K, Mackel R. Reticulospinal excitation and inhibition1033of neck motoneurons. Exp Brain Res 32: 471-489, 1978.1034Quinet J, Goffart L. Saccade dysmetria in head unrestrained gaze shifts after muscimol1035inactivation of the caudal fastigial nucleus in the monkey. J Neurophysiol 93: 2343-2349,10362005.1037Quinet J, Goffart L. Head-unrestrained gaze shifts after muscimol injection in the caudal1038fastigial nucleus of the monkey. J Neurophysiol 98: 3269-3283, 2007.1039Quinet J, Goffart L. Electrical microstimulation of the fastigial oculomotor region in the1040head-unrestrained monkey. J Neurophysiol 102: 320-336, 2009.1041Ranck JB. Which elements are excited in electrical stimulation of mammalian central1042nervous system: a review. Brain Res 98: 417-440, 1975.1043Rasmussen AT. Origin and course of the fasciculus uncinatus (Russell) in the cat, with1044observations on other fiber tracts arising from the cerebellar nuclei. J Comp Neurol 57:1045165-197, 1933.1046 Takahashi JNP Page-44 Ritchie L. Effects of cerebellar lesions on saccadic eye movements. J Neurophysiol 39:10471246-1256, 1976.1048Ron S, Robinson DA. Eye movements evoked by cerebellar stimulation in the alert monkey.1049J Neurophysiol 36: 1004-1022, 1973.1050Roucoux A, Guitton D, Crommelinck M. Stimulation of the superior colliculus in the alert1051cat. II. Eye and head movements evoked when the head is unrestrained. Exp Brain Res 39:105275-85, 1980.1053Sasaki K, Kawaguchi S, Matsuda Y, Mizuno N. Electrophysiological studies on1054cerebello-cerebral projections in the cat. Exp Brain Res 16: 75-88, 1972.1055Sasaki K, Staunton HP, Dieckmann G. Characteristic features of augmenting and recruiting1056responses in the cerebral cortex. Exp Neurol 26: 369-392, 1970.1057Sato H, Noda H. Saccadic dysmetria induced by transient functional decortication of the1058cerebellar vermis. Exp Brain Res 88: 455-458, 1992.1059Scudder CA, McGee DM. Adaptive modification of saccade size produces correlated1060changes in the discharges of fastigial nucleus neurons. J Neurophysiol 90: 1011-1026, 2003.1061Shinoda Y, Arnold AP, Asanuma H. Spinal branching of corticospinal axons in the cat. Exp1062Brain Res 26: 215-234, 1976.1063Shinoda Y, Ghez C, Arnold A. Spinal branching of rubrospinal axons in the cat. Exp Brain1064Res 30: 203-218, 1977.1065Shinoda Y, Sugiuchi Y, Futami T, Ando N, Kawasaki T. Input patterns and pathways from1066the six semicircular canals to motoneurons of neck muscles. I. The multifidus muscle group. J1067Neurophysiol 72: 2691-2702, 1994.1068Shinoda Y, Sugiuchi Y, Futami T, Ando N, Yagi J. Input patterns and pathways from the six1069semicircular canals to motoneurons of neck muscles. II. The longissimus and semispinalis1070muscle groups. J Neurophysiol 77: 1234-1258, 1997.1071 Takahashi JNP Page-45 Shinoda Y, Sugiuchi Y, Izawa Y, Hata Y. Long descending motor tract axons and their1072control of neck and axial muscles. Prog Brain Res 151: 527-561, 2006.1073Shinoda Y, Yoshida K. Dynamic characteristics of response to horizontal head angular1074acceleration in vestibular pathway in the cat. J Neurophysiol 37: 653-673, 1974.1075Straumann D, Haslwanter T, Hepp-Reymond MC, Hepp K. Listing’s law for eye, head1076and arm movements and their synergic control. Exp Brain Res 86: 209-215, 1991.1077Sugita S, Noda H. Pathways and termination of axons arising in the fastigial oculomotor1078region of macaque monkeys. Neurosci Res 10:118-136, 1991.1079Sugiuchi Y, Izawa Y, Takahashi M, Na J, Shinoda Y. Physiological characterization of1080synaptic inputs to inhibitory burst neurons from the rostral and caudal superior colliculus. J1081Neurophysiol 93: 697-712, 2005.1082Sugiuchi Y, Kakei S, Izawa Y, Shinoda Y. Functional synergies among neck muscles1083revealed by branching patterns of single long descending motor-tract axons. Prog Brain Res1084143: 411-421, 2004.1085Sugiuchi Y, Takahashi M, Shinoda Y. Input-output organization of inhibitory neurons in the1086interstitial nucleus of Cajal projecting to the contralateral trochlear and oculomotor nucleus. J1087Neurophysiol 110: 640-657, 2013.1088Suzuki JI, Cohen B. Head, eye, body and limb movements from semicircular canal nerves.1089Exp Neurol 10: 393-405,1964.1090Takahashi M, Sugiuchi Y, Izawa Y, Shinoda Y. Commissural excitation and inhibition by1091the superior colliculus in tectoreticular neurons projecting to omnipause neuron and inhibitory1092burst neuron regions. J Neurophysiol 94: 1707-1726, 2005a.1093Takahashi M, Sugiuchi Y, Izawa Y, Shinoda Y. Synaptic inputs and their pathways from1094fixation and saccade zones of the superior colliculus to inhibitory burst neurons. Ann N Y1095Acad Sci 1039: 209-219, 2005b.1096 Takahashi JNP Page-46 Takahashi M, Sugiuchi Y, Shinoda Y. Commissural mirror-symmetric excitation and1097reciprocal inhibition between the two superior colliculi and their roles in vertical and1098horizontal eye movements. J Neurophysiol 98: 2664-2682, 2007.1099Takahashi M, Sugiuchi Y, Shinoda Y. Topographic organization of excitatory and Inhibitory1100commissural connections in the superior colliculi and their functional roles in saccade1101generation. J Neurophysiol 104: 3146-3167, 2010.1102Takahashi M, Sugiuchi Y, Shinoda Y. Commissural inhibition between bilateral superior1103colliculi for saccades and bilateral vestibular nuclei for vestibulo-ocular reflex(VOR). Ann N1104Y Acad Sci Suppl 1233: 152-174, 2011.1105Thach WT, Goodkin HP, Keating JG. The cerebellum and the adaptive coordination of1106movement. Annu Rev Neurosci 15: 403-442, 1992.1107Thach WT J. Somatosensory receptive fields of single units in cat cerebellar cortex. J1108Neurophysiol 30: 675-696, 1967.1109Thomas DM, Kaufman RP, Sprague JM, Chambers WW. Experimental studies of the1110vermal cerebellar projections in the brain stem of the cat (fastigiobulbar tract). J Anat 90:1111371-385, 1956.1112Tweed D, Vilis T. Listing’s law for the head. Soc Neurosci Abstr 14: 958, 1988.1113Von Helmholtz H. Handbuch der Physiologischen Optik. Hamburg, Voss, 1867.1114Voogd J. The Cerebellum of the Cat. Structure and Fiber connexions. Assen: Van Gorcum,11151964.1116Voogd J, Shinoda Y, Ruigrok TJH, Sugihara I. Cerebellar nuclei and the inferior olivary1117nuclei: Organization and Connections. In: Handbook of the Cerebellum and Cerebellar1118Disorders. Schmahmann J, Koibuchi N, Rossi F. (Eds). Heidelberg, Germany:1119Springer-Verlag, vol.1, pp.337-436, 2013.1120 Takahashi JNP Page-47 Walberg F, Pompeiano O, Brodal A, Jansen J. The fastigiovestibular projection in the cat.1121An experimental study with silver impregnation methods. J Comp Neurol 118: 49-75, 1962a.1122Walberg F, Pompeiano O, Westrum LE, Hauglie-Hanssen E. Fastigioreticular fibers in the1123cat. An experimental study with silver methods. J Comp Neurol 119: 187-199, 1962b.1124Walton MM, Bechara B, Gandhi NJ. Effect of reversible inactivation of superior colliculus1125on head movements. J Neurophysiol 99: 2479-2495, 2008.1126Waitzman DM, Silakov VL, DePalma-Bowles S, Ayers AS. Effects of reversible1127 inactivation of the primate mesencephalic reticular formation. II. Hypometric vertical1128saccades. J Neurophysiol 83: 2285-2299, 2000.1129Yamada J, Noda H. Afferent and efferent connections of the oculomotor cerebellar vermis in1130the macaque monkey. J Comp Neurol 265: 224-241, 1987.1131Yoshida K, McCrea R, Berthoz A, Vidal PP. Morphological and physiological1132characteristics of inhibitory burst neurons controlling horizontal rapid eye movements in the1133alert cat. J Neurophysiol 48: 761-784, 1982.113411351136 Takahashi JNP Page-48 Legends1137Fig. 1. Brainstem distribution of axon terminals of neurons in the caudal fastigial nucleus1138(FN) (A and B) and the superior colliculus (SC) (C and D). A: frontal section of the1139 cerebellum showing a site at which dextran-biotin (DB) (0.35 l) was injected into the right1140 caudal FN (black area). B: distribution of anterogradely-labeled axon terminals in the1141brainstem after the injection of DB into the FN shown in A. All labeled axon terminals in two1142 consecutive 70-m-thick sections are plotted in representative frontal sections (a) 1.0 mm1143 rostral to the rostral border of the abducens nucleus and (b) 0.5 mm caudal to the caudal1144border of the abducens nucleus. Each dot indicates one axon terminal in B and D. C: frontal1145 section of the midbrain showing a site at which DB (1.0 l) was injected into the SC (black1146 area). D: distribution of anterogradely-labeled axon terminals in the brainstem after the1147injection of DB into the right SC shown in C. IO, inferior olive; VI n, abducens nerve; VII n,1148facial nerve; SO, superior olive; G, facial genu; NVII, facial nucleus; Py, pyramis; CG, central1149gray; CP, cerebral peduncle; Lt, left; Rt; right.11501151Fig. 2. Experimental setup and properties of synaptic inputs from the caudal fastigial nuclei1152(FNs) and the superior colliculi (SCs) to a reticular neuron in the nucleus reticularis1153gigantocellularis (NRG). A-C: lateral (A, B) and dorsal views (C) of the brainstem showing1154the experimental setup. Stimulating electrodes were placed in the left abducens nerve (site 1),1155the right FN (sites 2-4 in A and C), the left FN (sites 5-7 in B and C), the right SC (sites 8-101156in A and C), the left SC (sites 11-13 in B and C) and the right abducens nucleus (site 14). IC,1157inferior colliculus; PN, pontine nucleus; VI Nucl, abducens nucleus; Lt.SC, Rt.SC, left and1158right superior colliculus; Lt. VI n, left abducens nerve; VI, abducens nucleus; NRG, nucleus1159reticularis gigantocellularis; Lt. FN, Rt.FN, left and right fastigial nucleus; VI,VIIA, VIIB,1160lobules VI,VIIA, VIIB. In this diagram and similar diagrams in the following figures, a1161 Takahashi JNP Page-49 recording site of a reticular neuron is indicated by an open cell that was penetrated with a1162microelectrode on the left side, and the locations of stimulating electrodes were checked on1163the surface of the bilateral SCs based on the photographs taken after each experiment. D:1164 antidromic field potentials in the left abducens nucleus evoked at 100 A by stimulation of1165 the left abducens nerve. Note the typical negative field potentials of antidromic spikes in the1166abducens nucleus. E: no intracellular response or negative field potentials evoked by1167antidromic activation of the left abducens nerve (site 1). This reticular neuron was recorded11681.2 mm caudal to the abducens nucleus (D), and was not activated from the ipsilateral1169 abducens nerve at 200 A. F: properties of postsynaptic potentials (PSPs) evoked by1170 stimulation of the contralateral FN at 50 A (2-4) and the ipsilateral FN at 100 A (5-7) in the1171 same reticular neuron as in E. Large depolarizing potentials with spikes were evoked from the1172stimulation sites in the bilateral FNs. The number attached to each panel (2-7) corresponds to1173the stimulation site in the bilateral FNs (sites 2-7). In each panel, top and bottom traces:1174intracellular potentials and juxtacellular field potentials recorded just outside the penetrated1175cell, respectively. The same arrangements for stimulating electrodes and their response traces1176are used in the following figures, if not stated otherwise. G: properties of PSPs evoked by1177 stimulation of the contralateral (sites 8-10) and ipsilateral SC (sites 11-13) at 300 A in the1178 same reticular neuron as in F. Large depolarizations with spikes were evoked from the1179contralateral SC (8-10) and the ipsilateral SC (11-13).11801181Fig. 3. Synaptic nature of fastigial and collicular inputs to reticular neurons in the rostral1182NRG. A: parasagittal section of the cerebellar vermis showing stimulation sites in the right1183(contralateral) caudal FN (sites 2-4). B: effects of stimulus intensity on depolarizing PSPs1184evoked by stimulation of the contralateral caudal FN in a neuron in the NRG. C-I: effects of1185Cl injection into another NRG neuron on FN(F, G) and SC-evoked PSPs (H, I) in the same1186 Takahashi JNP Page-50 experiment as in A and B. C, D: depolarizing PSPs followed by hyperpolarizing PSPs evoked1187 by stimulation of the contralateral abducens nucleus (site 14 in Fig. 2C) at 400 A before (C)1188and after the injection of Cl (F) into the penetrated cell. Note that the late hyperpolarizing1189PSPs were reversed to depolarizing PSPs, indicating that they were inhibitory postsynaptic1190potentials (IPSPs). E: Reversed IPSPs after Cl injection into the cell. The averaged PSPs in C1191were subtracted from those in D. Arrow indicates the onset of stimulus. F, G: PSPs evoked by1192stimulation of the contralateral (2-4) and ipsilateral caudal FN (5-7) before (F) and after1193injection of Cl (G). H, I: PSPs evoked by stimulation of the contralateral (10) and ipsilateral1194SC (11) before (H) and after Cl injection (I). Note that after the late hyperpolarizing IPSPs1195were reversed to depolarizing IPSPs in D, PSPs evoked by stimulation of the bilateral FNs1196and SCs did not change their configurations.11971198Fig. 4. Effects of stimulation sites in and around the caudal FN on PSPs evoked in reticular1199neurons in the NRG. A: a reconstruction of the stimulating electrode positions in the right1200(contralateral) caudal FN in three experiments (c, d, e). Three electrode tracks (c-e) were1201reconstructed on serial sections of the cerebellum in each experiment and projected to a1202representative parasagittal section (1.5 mm lateral from the midline). Filled and open circles,1203stimulation sites located within the FN and its adjacent white matter, respectively. Broken line1204on the left indicates the stereotactic vertical. B: photomicrograph of the parasagittal section1205showing stimulation site 2 in Ad in the right caudal FN (arrowhead). C-E: excitatory1206postsynaptic potentials (EPSPs) evoked by stimulation of different sites in the contralateral1207(2-4) and ipsilateral caudal FN (5-7). Stimulating electrode positions were located rostral (C)1208and caudal (E) to those in D. C: large monosynaptic EPSPs evoked by stimulation of the1209deepest site of the stimulating electrodes located in the contralateral FN (c in A). D: large1210monosynaptic EPSPs evoked by stimulation of the three dorsoventral sites (d in A). The1211 Takahashi JNP Page-51 excitation from the deepest site was slightly larger and shorter. E: larger monosynaptic1212excitation evoked by stimulation of the most superficial site of the stimulating electrodes1213located in the contralateral FN (e in A).12141215Fig. 5. Identification of a reticulospinal neuron (RSN) in the rostral NRG that received1216excitation from the caudal FNs. A: experimental setup. In addition to 14 stimulating1217electrodes (Fig. 2C), two stimulating electrodes were placed in the ventral funiculus of the1218 cervical spinal cord (C2) on the left side (site C2). FFH, Forel’s field H. B: antidromic spikes1219 evoked in an all-or-none manner at threshold (150 A) by stimulation of the ipsilateral second1220 cervical spinal cord (site C2). C: no antidromic field potentials evoked by stimulation of the1221 left abducens nerve (site 1) at 500 A. This neuron was recorded 0.8 mm caudal to the caudal1222 border of the abducens nucleus. IPSPs evoked by contra-abducens nucleus stimulation (C14)1223were hyperpolarized and constant in amplitude during recording. D, E: EPSPs evoked from1224the bilateral FNs (D) and from the bilateral SCs (E). F, G: latency histograms of EPSPs in1225neurons of the rostral NRG evoked by stimulation of the contralateral FN (F) and the1226ipsilateral FN (G) (n=38).12271228Fig. 6. Properties of fastigial and tectal synaptic inputs to an RSN in the rostral NRG. A:1229experimental setup. B, C: properties of synaptic inputs from the bilateral FNs (B) and from1230the bilateral SCs (C). B: monosynaptic EPSPs evoked by stimulation of the contralateral (2-4)1231 and ipsilateral caudal FN (5-7) at 300 A. C: topographic pattern of synaptic input from the1232 rostrocaudal sites in the contralateral (8-10) and the ipsilateral SC (11-13). The broken line in1233C11 shows the PSPs evoked by single stimuli. The excitation evoked by the second stimuli1234was larger, indicating the presence of temporal facilitation. D, E: latency histograms of EPSPs1235in NRG neurons evoked by stimulation of the contralateral SC (D) and the ipsilateral SC (E)1236 Takahashi JNP Page-52 (n=38). Note the two groups of neurons with different EPSP latencies in E.12371238Fig. 7. Intracellular recordings from a left fastigial output neuron with axon branches that1239projected to the SC and the brainstem on the contralateral side (BS). A: schematic diagram of1240the experimental setup. Stimulating electrodes were placed rostrocaudally in the bilateral SCs1241(1-6), dorsoventrally in the excitatory burst neuron (EBN) area (7, 8) in the caudal NRPc and1242in the inhibitory burst neuron (IBN) area (9, 10) in the rostral NRG. B: stimulation of three1243 rostrocaudal sites in the ipsilateral SC (1-3 in the left SC) and the contralateral SC (4-6 in the1244right SC) and the contralateral EBN area in the caudal NRPc (7-8) and the contralateral IBN1245 area in the NRG (9-10) at 500 A. Note that this neuron was activated from site 4 in the right1246 SC, and sites 9 and 10 in the right IBN area of the NRG. C, D: antidromic activation of the1247caudal FN neuron in an all-or-none manner at thresholds from the contralateral SC (C, site 4)1248and the contralateral IBN region (D, site 10) (a), and double-shock stimulation of the same1249sites to determine refractory periods of the cell (b and c). Antidromic spikes were evoked with1250double-pulse stimuli from site 4 at intervals of 0.8 ms (Cc), but not at intervals of 0.7 ms (Cb)1251for SC stimuli, and from site 10 at intervals of 0.6 ms (Dc), but not at intervals of 0.5 ms (Db).1252Refractory periods were considered to be 0.7 ms for site 4 and 0.5 ms for site 10. E, F: spike1253collision test between SC-evoked and NRG-evoked antidromic spikes. Stimuli for the1254contralateral NRG (site 10) were applied 1.4 ms (a) and 1.5 ms (b) after stimuli for the1255contralateral SC (site 4) (E), and 1.5 ms (a) and 1.6 ms (b) before stimuli for the contralateral1256SC (site 4) (F). G: schematic diagram of axonal conduction times for each part of the axon of1257this caudal FN neuron calculated using the above collision test values. The latencies of1258SC-evoked and NRG-evoked antidromic spikes were 1.2 ms and 0.5 ms, respectively.12591260Fig. 8. Properties of fastigial and tectal synaptic inputs to an RSN in the NRG. A:1261 Takahashi JNP Page-53 experimental setup. B: no intracellular response to stimulation of the left abducens nerve at1262 200 A (1) and the contralateral abducens nucleus at 500 A (14). C, D: properties of1263 synaptic inputs from the FNs (C) and from the SCs (D). This RSN in the NRG received larger1264EPSPs from the rostral part of the contralateral SC than from the caudal part of the1265contralateral SC. D: temporal facilitation of the evoked EPSPs. The bottom traces in D11; the1266EPSPs evoked by the first (broken line) and the second stimuli (solid line) relative to the onset1267of the stimuli (arrow). The EPSPs evoked at 1.9 ms were facilitated but the EPSPs at 1.2 ms1268were not.12691270Fig. 9. Properties of fastigial and tectal synaptic inputs to an RSN in the nucleus reticularis1271pontis caudalis (NRPc). A: the experimental setup. B: properties of PSPs evoked by1272 stimulation of the contralateral (sites 2-4) and ipsilateral FN (sites 5-7) at 500 A in an RSN.1273 No PSPs were evoked from the stimulation sites in the ipsilateral FN even with double-pulse1274 stimuli at 500 A (5-7). C: properties of EPSPs evoked by stimulation of the contralateral1275 (sites 8-10) and ipsilateral SC (sites 11-13) at 500 A in the same RSN as in B. Stimulation of1276 the contralateral SC evoked large EPSPs with spikes at 500 A (8-10), and at 100 A, rostral1277 stimulation evoked small EPSPs (8, 9), and caudal stimulation evoked large EPSPs with1278spikes (10). Double-pulse stimulation of the ipsilateral SC (sites 11-13) did not evoke visible1279 short-latency EPSPs at 500 A. D: no antidromic spikes or antidromic field potentials were1280 evoked from the ipsilateral abducens nerve in this neuron.12811282Fig. 10. Properties of fastigial and tectal synaptic inputs to an RSN in the caudal NRPc. A:1283the experimental setup. B: no antidromic spike was evoked from the ipsilateral abducens1284nerve in this neuron. C: properties of PSPs evoked by stimulation of the contralateral (sites1285 Takahashi JNP Page-54 2-4) and ipsilateral FN (sites 5-7) at 500 A in the same RSN as in B. No monosynaptic PSPs1286 were evoked from the stimulating sites in the FNs even with double-pulse stimuli at 500 A1287 (2-4). D: properties of EPSPs evoked by stimulation of the contralateral (8-10) and ipsilateral1288 SC (11-13) at 500 A in the same RSN as in C.1289 1290Fig. 11. A: summary of inputs from the caudal FN and the SC to RSNs in the rostral NRG1291and caudal NRPc. Open neurons indicate excitatory neurons. Dotted lines indicate a presumed1292disynaptic excitatory pathway. rSC, cSC, rostral and caudal SC; cFN, caudal fastigial nucleus;1293VI, the abducens nucleus; NRPc, nucleus reticularis pontis caudalis; NRG, nucleus reticularis1294gigantocellularis; C2, second cervical spinal cord; FFH, Forel’s field H; INC, interstitial1295nucleus of Cajal. B: schematic neural diagram showing the postulated output pathways from1296the SC to neck muscles on both sides. Note that identical functional groups of neck muscles1297on both sides are activated during oblique vertical head movements. RST, reticulospinal tract.1298CN, tectal excitatory commissural neuron.1299