Brainstem Control of Head Movements During Orienting; Organization of The

Progress in Neurobiology 66 (2002) 205–241

Brainstem control of head movements during orienting; organization of the premotor circuits Tadashi Isa a,∗, Shigeto Sasaki b a Department of Integrative Physiology, National Institute for Physiological Sciences, Myodaiji, Okazaki 444-8585, Japan b Department of Neurophysiology, Tokyo Metropolitan Institute for Neuroscience, Musashidai, Fuchu 183-8526, Japan Received 26 July 2001; accepted 10 December 2001

Abstract When an object appears in the visual field, animals orient their head, eyes, and body toward it in a well-coordinated manner (orienting movement). The head movement is a major portion of the orienting movement. Interest in the neural control of head movements in the monkey and human have increased in the 1990’s, however, fundamental knowledge about the neural circuits controlling the orienting head movement continues to be based on a large number of experimental studies performed in the cat. Thus, it is crucial now to summarize information that has been clarified in the cat for further advancement in understanding the neural control of head movements in different animal species. The superior colliculus (SC) has been identified as the primary brainstem center controlling the orienting. Its output signal is transmitted to neck motoneurons via two major separate pathways: one through the reticulospinal neurons (RSNs) in the pons and medulla and the other through neurons in Forel’s field H (FFH) in the mesodiencephalic junction. The tecto-reticulo-spinal pathway controls orienting chiefly in the horizontal direction, while the tecto-FFH-spinal pathway controls orienting in the vertical direction. In each pathway, a subgroup of neurons functions as premotor neurons for both extraocular and neck motoneurons, while others are specified for each, which allows both coordinated and separate control of eye and head movements. Head movements almost always produce shifts in the center of gravity that might cause postural disturbances. The postural equilibrium may be maintained by transmitting the orienting command to the limb segments via descending axons of the reticulospinal and long propriospinal neurons. The SC and brainstem relay neurons receive descending inputs from higher order structures such as the cerebral cortex, cerebellum, and basal ganglia. These inputs may serve context-dependent control of orienting by modulating the activities of the primary brainstem pathways. © 2002 Elsevier Science Ltd. All rights reserved.

Contents 1. Introduction ...... 206 2. Behavioral analysis of head movements during orienting ...... 207 2.1. A common organization in the head motor system and biomechanical constraint ...... 207 2.2. General characteristics of orienting in unrestrained cats ...... 208 2.3. Visually guided orienting to moving versus stationary stimuli ...... 208 3. Primary brainstem pathways controlling orienting ...... 209 3.1. Superior colliculus (SC); brainstem center of orienting ...... 210 3.1.1. Microstimulation of the SC...... 210

Abbreviations: BC, m. biventer cervicis; BCC, m. biventer cervicis and complexus; C-RSNs, “cervical” reticulospinal neurons; CTT, central tegmental tract; CP, cerebral peduncle; CUN, cuneiform nucleus; EPSP, excitatory postsynaptic potential; FEF, frontal eye field; FFH, Forel’s field H; FR, fasciculus retroflexus; FN, fastigial nucleus; G, genu of facial nerve; HRP, horseradish peroxidase; INC, interstitial nucleus of Cajal; IO, inferior olive; LED, light emitting diode; LGN, lateral geniculate nucleus; L-RSNs, “lumbar” reticulospinal neurons; MLF, medial longitudinal fasciculus; MNs, motoneurons; NR, nucleus ruber (red nucleus); NRGc, nucleus reticularis gigantocellularis; NRPc, nucleus reticularis pontis caudalis; PAG, periaqueductal grey matter; PH, prepositus hypoglossi; PNs, phasic neurons; PSNs, phasic sustained neurons; Pyr, pyramidal tract; Rp, raphe region; RSN, reticulospinal neuron; SC, superior colliculus; SCUN, subcuneiform nucleus; SGI, stratum griseum intermediale (intermediate layer of SC); SGP, stratum griseum profundum (deep layer of SC); SGS, stratum griseum superficiale (superficial layer of SC); SN, substantia nigra; SNr, substantia nigra pars reticulata; SPL, m. splenius; TB, trapezoid body; Tec, tectum; TNs, tonic neurons; TR(S)N, tectoreticular and tectoreticulospinal neuron; VN, vestibular nucleus; III, oculomotor nucleus; VI, abducens nucleus; VII, facial nucleus ∗Corresponding author. Tel.: +81-564-55-7859; fax: +81-564-55-7790. E-mail address: [email protected] (T. Isa).

3.1.2. Lesion studies of the SC ...... 210 3.1.3. Single unit recordings in the SC ...... 212 3.1.4. Descending and ascending projections from the SC...... 212 3.2. Brainstem relay of the descending commands from the superior colliculus to neck motoneurons 213 3.2.1. Tectoreticulospinal pathways via the pontomedullary reticular formation ...... 213 3.2.2. Tectoreticulospinal pathways via the reticular formation in the mesodiencephalic junction 217 4. Effects of lesion of the brainstem relay structures on orienting head movements ...... 220 4.1. The pontomedullary reticular formation ...... 220 4.2. Forel’s ﬁeld H...... 221 5. Single unit activities of neurons in the brainstem reticular formation during orienting...... 221 5.1. Single unit activity of neurons in the pontomedullary reticular formation ...... 221 5.1.1. Directional tuning of the NRPc and NRGc neurons during orienting ...... 221 5.1.2. Firing pattern of reticular neurons and their morphological and physiological correlates . 223 5.2. Single unit activity of neurons in Forel’s ﬁeld H ...... 225 6. Differential control of horizontal and vertical components of head movements ...... 226 7. Higher order structures that regulate the primary brainstem pathways for orienting ...... 228 7.1. Pericruciate cortical areas ...... 228 7.2. Cerebellum ...... 232 7.2.1. Effects of functional inactivation ...... 232 7.2.2. Efferent projection from the fastigial nucleus ...... 233 7.3. Basal ganglia ...... 233 7.3.1. Effects of electrical stimulation ...... 233 7.3.2. Descending projection from the basal ganglia ...... 233 7.3.3. Role of basal ganglia in control of orienting triggered by volitional intention ...... 233 8. Posture adjustment during orienting head movements ...... 233 9. Other animal species ...... 234 9.1. Primates ...... 234 9.1.1. Role of the SC in orienting head movements in primates ...... 235 9.1.2. Descending pathway from the SC in primates...... 235 9.1.3. Involvement of the FEF in orienting head movements in primates ...... 235 9.2. Rodents ...... 235 9.3. Barn owl ...... 236 10. Conclusion ...... 236 References ...... 237

1. Introduction eye movements in the head-restrained condition. Stud- ies on the neural control of orienting movements in Orienting is defined as the gaze shift from one point an unrestrained condition have been done mainly in to another in space that animals make to fixate an object. cats except for early studies by Bizzi and colleagues in Orienting is observed in a wide variety of animals. Thus, primates (Bizzi et al., 1971, 1972a,b; Dichgans et al., the basic component of the neuronal pathways controlling 1973; Bizzi, 1979). Neural mechanisms for control of head orienting is presumed to be common over animal species. movements in primates have received much attention since The orienting movements are comprised of coordination of 1990. However, because of the lack of detailed knowledge different body parts such as the eyes, head, and trunk. The about the neuronal pathways controlling the orienting head body architecture differs among the animal species, which movements, interpretation of the behavioral data obtained may lead to addition of adaptive circuits to the primary com- in monkeys should largely rely on the knowledge accumu- ponent of the circuit. For instance, the oculomotor range, lated from cat studies. Thus, it is essential to summarize which is the maximum range within which the eyeballs are the knowledge obtained from cat studies to further advance allowed to move in the orbit, considerably varies among the the understanding of the neural mechanisms of orienting animal species; it is smaller in lower vertebrates and larger movements in different species of animals, especially in in primates. In contrast, the head movements play a major primates. This is a major aim of this review. role in lower vertebrates compared with higher vertebrates, Behavioral analyses of eye movements and their neu- which may be correlated with differences in the organiza- ronal mechanisms in various species of animals have been tion of the neural circuit. Unfortunately, the neural substrate reported in a number of articles. In this review, we limit of such species differences is unclear at this moment. In pri- our description to neural control of head movements dur- mates, a majority of studies have been devoted to saccadic ing orienting and will not expand our discussion to other T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 207 important issues such as the eye–head coordination and 2. Behavioral analysis of head movements during control of gaze movements, which have been reviewed in orienting detail elsewhere (Berthoz, 1989; Guitton, 1992; Sparks, 1999). The anatomy and physiology of premotor circuits for 2.1. A common organization in the head motor the saccadic eye movements were systematically described system and biomechanical constraint in a review by Moschovakis et al. (1996). Our first objective is to review the studies that analyzed Cats are quadrupedal animals, and early researchers ex- the neural pathways from the superior colliculus (SC) to the pressed doubts about applying experimental results from cats neck motoneuron in the cat. The SC or the optic tectum, has to bipedal animals, such as humans. Thus, it is important to been considered a primary center of orienting in the brain- determine whether principles obtained from the studies on stem (Hess et al., 1954; Sparks, 1986; Sprague and Meikle, the head movements of cats are applicable to the head motor 1965; Wurtz and Albano, 1980) in all vertebrates that have systems in human or other animal species. Richmond et al. been studied (see the book edited by Vanegas (1984)). It is (1999) have carefully reviewed this issue by considering supposed that the SC is involved in transformation of spa- the arrangement of the head–neck motor apparatus and its tial information of the target location into a spatio-temporal biomechanical constraint. Regarding the biomechanical con- pattern of muscle activities to move the eyes, head, and straint, fluoroscopic studies in cats showed that their verte- body. We will review the functional organization of the bral columns are oriented vertically and the atlanto–occipital neuronal pathways from the SC to the neck motoneurons, joints are in the flexed position when resting, as they are which is mainly based on the data obtained in our labora- in humans (Vidal et al., 1986). Horizontal head movements tories. We will show that descending signals from the SC of small amplitude are executed mainly around the C2 are transmitted to neck motoneurons by two major separate vertebra, while larger horizontal head movements involve pathways via relay cells in the brainstem; one chiefly con- increasing rotations around progressively lower cervical trols the horizontal component, and the other controls the vertebrates (Richmond et al., 1992) and translation by trunk vertical component of the head movements. The firing pat- and limb movements. In vertical head movements, dorsal tern of these relay cells during orienting will be reviewed extension is largely performed around the atlanto–occipital in detail. joint, while ventral flexion is executed at the joints between We have found that moving and stationary visual stim- the C5 and T2, since the atlanto–occipital joint is already uli elicit different types of orienting movements, velocity- in a flexed position at rest. Further studies by Vidal et al. and position-guided orienting, respectively. For instance, to (1988) and Graf et al. (1995) showed marked homology in catch a moving mouse, a cat makes swift orienting to the an- the head–neck system among cats, rats, and rabbits, despite ticipated location of the target. Otherwise, it misses the prey. some species differences. It is observed commonly among In contrast, orienting to a stationary object (position-guided different species of animals, though relative contribution of orienting) is not necessarily rapid and the animal is allowed translation to the whole orienting movement varies. to initiate the movement after appearance of the object in the Approximately, 20 muscles participate in the execu- fixed position. We also describe such subtypes of orienting tion of head movements. Their arrangements vary among that would give clues to in-depth understanding of orienting. quadrupeds, like cats, habitually terrestrial quadrupeds, like Sensory stimuli of various modalities induce orienting monkeys, and bipedal animals, like humans (Napier and movements. Recording studies in the SC revealed that in- Napier, 1985; Richmond et al., 2001). Since the head has puts from different sensory modalities are transformed onto to be maintained in space, dorsal neck muscles are more a common, gaze-referenced representation at or prior to developed in quadrupedal animals than in bipedal animals. arrival at the SC. Information processing downstream from For example, the m. biventer cervices and the complexus, the SC is likely similar regardless of the modality of the which constitute a major component of dorsal neck mus- target stimuli. Thus, in this review, we mainly focus on the cles in cats, are fused into a single muscle semispinalis brainstem processing for visually guided orienting that is cervicis in bipedal animals such as humans. Even among assumed to be similar to the processing for orienting to other the quadrupeds, cats use the head as the primary prehensile modalities. Several higher order structures, such as the cere- organ; however, macaque monkeys do not. Such functional bral cortex, cerebellum, and basal ganglia, have been shown repertories might have led to different architecture of neck to connect with the primary brainstem neural circuits for muscles. On the other hand, the combination pattern of orienting (the tectoreticulospinal pathway) and modulate muscle activation during head movements appears mostly their activities. Our second objective is to give an overview similar in cats and monkeys (Sasaki and Yoshimura, un- on their projections and roles in regulating the orienting published observation), which may suggest similarity in movements. the basic neural mechanisms. However, as described in Finally, the observations obtained from studying cats will the previous sections, modes of locomotion and functional be compared with those from primate and other animal repertories of head movements should be taken into account specie studies to facilitate understanding the neural mecha- when we consider neural control in different animal species nism of head movement from a phylogenical viewpoint. (see also Corneil et al., 2001). 208 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 1. (A) Trajectories of eye, head, and gaze (eye plus head) movements of a cat toward a visual stimulus, which moved in ramp with the time course as indicated in the text (visual stimulus). The initial slow head movement and the contraversive eye movement are indicated by open and closed arrows, respectively. (B) Relationship between the latencies of head movements (horizontal axis) and saccadic eye movements (vertical axis). (C) Relationship between the amplitude of head movements (horizontal axis) and saccadic eye movements (vertical axis). The oblique line indicates y = (1/2)x.

2.2. General characteristics of orienting only about half of the head in amplitude (Sasaki et al., in unrestrained cats unpublished observation). This was also in agreement with Guitton et al. (1984). In cats, virtually no solitary head Fig. 1 exemplifies, in the cat, the horizontal component of or eye movement was observed in unrestrained condition ◦ an eye movement in the orbit, a head movement, and a gaze even in the case of movements smaller than 4 (Blakemore shift in space during orienting toward a moving visual target. and Donaghy, 1980). Furthermore, linear relationship be- The head moves slowly toward a target (open arrow), while tween the amplitude of head movements and its maximum eyes move in a contraversive direction (closed arrow), pre- angular velocities has been observed, which corresponds sumably by the vestibuloocular reflex, keeping a fixed gaze to the “main sequence relationship” as described for sac- position during this phase. Then, a saccadic eye movement cadic eye movements (Bahill et al., 1975). The slope of the − is initiated to the target while the rapid head movements linear relationship was 6.12 ± 1.93 s 1 for horizontal and − continue to accelerate smoothly. After the saccade is com- 4.85±0.15 s 1 for vertical movements (Sasaki et al., 1999). pleted, the head still continues to move toward the target, and during this phase, the eyes return to the initial position 2.3. Visually guided orienting to moving by the vestibuloocular reflex as was first shown in monkeys versus stationary stimuli (Bizzi et al., 1971; Dichgans et al., 1973) and then in cats (Guitton et al., 1984; Berthoz, 1989). As shown in Fig. 1B, In ordinary life, animals are more sensitive to moving head movements preceded saccadic eye movements in a ma- objects than stationary ones. Fig. 2A and B shows compari- jority of trials. In these cases, the latencies of the initial head son of orienting movements elicited by a stationary (A) and movements were mostly shorter than 130 ms. In the case a moving visual light spot (B) in the same cat (Sasaki et al., of orienting movements with latencies longer than 150 ms, unpublished observation). In both cases, the head move- on the other hand, saccadic eye movements sometimes pre- ment led the eye movement. It is notable that latencies of ceded the head movements (Sasaki et al., unpublished obser- orienting were dramatically shortened in the case of orient- vation), which was in agreement with Guitton et al. (1984). ing toward the moving stimulus, compared with orienting to Another notable characteristic of orienting in cats was the stationary stimulus, as shown in plots of saccade laten- that amplitudes of saccadic eye movements linearly in- cies against those of head movements (Fig. 2C). Orienting creased with those of head movements as shown in Fig. 1C. movements toward the moving stimuli can be denoted as The slope of the regression line is approximately 0.5 (espe- “velocity-guided orienting”, while those toward the station- ◦ cially below 20 in amplitude), indicating that eyes move ary stimuli as “position-guided orienting”. The former might T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 209

to a moving stimulus. The cat moved the head shortly after the onset of movement of the stimulus (latencies; 50–80 ms) and the gaze shift was completed virtually in coincidence with termination of the target movement, indicating that the cat anticipated or estimated the destination of the moving visual stimulus. Further studies showed that the cat utilized the velocity information to predict the target location. Animals do not always make such “velocity-guided orienting” toward the moving object. If not, cats appeared to initiate the orienting after the light spot reached its destination and utilized the information about the terminal position of the target. This type of orienting is also classiﬁed as the “position-guided orienting” (Fig. 2B(b)). This is the same as ordinary orienting movements that have been extensively studied so far. Animals switch strategy between the above two types of orienting. Position- and velocity-guided orienting differed in kine- matics. Latencies of head movements and saccades were signiﬁcantly shorter in velocity-guided orienting than the position-guided orienting (see the previous sections). The velocity-guided orienting often appeared to be initiated with latencies <60 ms, while the movements with latencies >120 ms were apparently position-guided orienting movements. Furthermore, the maximum head angular velocity was about two times faster in the case of velocity-guided orienting than the position-guided orienting based on the estimation from the slope of the regression line of the main sequence relationship. The dynamics of orienting vary between position- and velocity-guided orienting. As a similar example, Guitton et al. (1984) showed that the slope of the main sequence relationship varies between “the visually triggered mode” and “the search mode”. The slope was larger in the latter mode (9.1◦ s−1) than in the former (4.3◦ s−1).

3. Primary brainstem pathways controlling orienting

Fig. 3 illustrates a schematic drawing of the neural pathways that have been considered as playing a fundamental Fig. 2. Eye, head, and gaze movements toward a step (or “stationary visual role in control of visually guided orienting in cats. First, stimulus” (A)) and ramp (or “moving visual stimulus” (B)) movements of the visual target. In the case of orienting to the moving visual stimu- the visual information is transmitted from the retina di- lus, either “velocity-guided” (B(a)) or “position-guided” (B(b)) orienting rectly to the superficial layer of the SC. The projections are was induced. (C) Relationship between the latencies of head (horizon- retinotopically organized; the lower half of the retina (up- tal axis) and saccadic eye movements (vertical axis). Open and closed per visual field) projects to its medial part of the SC, while circles indicate the movements toward the moving (only velocity-guided the upper half projects to the lateral portion. Horizontal movements are plotted) and stationary targets, respectively. meridian in visual space runs from the rostrolateral to cau- domedial direction on the SC map. Thus, the central visual field is represented in the rostral part and the lateral periph- be related to prey catching. This process should include a eral field in the caudal portion. The SC receives projection predictive process about the future location of the object, not only directly from the retina but also indirectly through based on its position and velocity at the particular moment the cerebral cortex via the lateral geniculate nucleus (LGN). when the anticipation is made. If this process does not occur, Classical occipital visual areas (areas 17–19) project to the the animals cannot capture their target. This type of orient- superficial layer of the SC, while the frontal oculomotor ing has not been studied in detail so far, since the move- region and the other cortical areas related to motor function ments have not been investigated in an unrestrained animal. project to the deeper layer of the SC. The deeper layer Fig. 2B(a) shows examples of “velocity-guided” orienting of the SC encodes amplitude and direction (gaze vector) 210 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

who observed contraversive rapid head movements by repetitive electrical stimulation of the SC. Synergic eye and head movements induced by SC stimulation were further studied by Faulkner and Hyde (1958) and with more precise technique by Roucoux et al. (1980) in cats. These authors have established that coordinated eye and head movements are evoked by repetitive microstimulation of the SC, and their amplitudes and directions (gaze vector) changed systematically depending on the site of stimulation with close relation to the visual field of the superficial layers just above the stimulation sites. Thus, it has been established that the deeper layers of the SC are topographically organized and form a motor map encoding the gaze vector of orienting. A recent study showed that the goal-directed saccadic eye movements observed in head restrained condition change to a gaze shift of a particular vector component in the head-unrestrained condition (Paré et al., 1994). Further functional subdivision of the SC has been disclosed recently. Neurons with activ- Fig. 3. A schematic diagram of the primary pathway controlling visually ities related to fixation were found in the rostral pole of the guided orienting head movements. SC, and, thus, this region was termed “fixation zone”. Stim- ulation of this region was found to interrupt not only sac- of orienting, as has been described in the literature using cadic eye movements (Munoz and Guitton, 1989) but also electrical stimulation, lesion, and unit recording techniques head movements during the ongoing gaze shift (Paré et al., during orienting. Their output is sent to neck motoneurons 1994; Paré and Guitton, 1994). The suppression appears via intercalated neurons mainly located in the brainstem to be mediated by inhibitory effects from fixation zone to reticular formation (see the subsequent sections). The direct the orienting–generating zone in the caudal SC and also by connection from the superficial to deeper layers of the SC facilitatory effects on omnipause neurons in the brainstem had been doubted after the description by Edwards (1980). reticular formation (Paré and Guitton, 1990, 1994, 1998). However, Maeda et al. (1979) showed that electrical stim- It appears that coding of head movements might be ulation of the contralateral optic disk induced disynaptic slightly different from that of eye saccades, since the stimu- excitatory postsynaptic potential (EPSP) in tectofugal cells lation parameters to induce head and eye movements differ. in cats. Recordings from neck motoneurons revealed disy- Saccadic eye movements exhibit typical threshold effect; naptic EPSPs by stimulation in the deep and intermediate the stimulation above a certain threshold induces saccadic tectal layers but trisynaptic EPSPs from the superficial layer eye movements with particular vector, but further increase (Alstermark et al., 1992c). It was suggested that the tec- in stimulus intensities does not result in increase in move- toreticulospinal neurons (together called TR(S)Ns) were ment amplitude (Schiller and Stryker, 1972). As to the large activated monosynaptically from neurons in the superficial and rapid head movements, it was shown by Roucoux et al. layer of the SC. Furthermore, more recent anatomical studies (1980) that the amplitude of head movements increases using fine anatomical techniques (Behan and Appel, 1992; with the stimulus intensities until it reaches a plateau level Hall and Lee, 1993; Lee and Hall, 1995; Mooney et al., 1988; determined by the stimulation site in the SC motor map, Rhoades et al., 1989) and electrophysiological studies using which was further confirmed by Paré et al. (1994). Recent anesthetized hamsters (Mooney et al., 1992) and in vitro studies are accumulating evidence suggesting that the feline slice preparations from the tree shrew and rats (Lee et al., SC is involved primarily in the control of gaze shift and not 1997; Isa et al., 1998; for review see Isa and Saito, 2001) specifically of eye or head movements alone (cf. Guitton showed the existence of vertical connections from the super- et al., 1984). ficial to intermediate layer of the SC, although the functional role of the interlaminar connection in a behaving animal 3.1.2. Lesion studies of the SC has not been determined yet. We will first review how the Lesion of the SC resulted in severe impairment of ori- SC has been established as the center of orienting and then enting head movements toward the contralateral side in cats the role of neural structures in the downstream of the SC. (Sprague and Meikle, 1965). Isa et al. (1992a) analyzed the trajectories of head movements after unilateral lesion of the 3.1. Superior colliculus (SC); brainstem center of orienting SC with kainic acid using the infrared position-sensor system (the stick diagrams of movements are shown in Fig. 4). 3.1.1. Microstimulation of the SC For a few days after the unilateral lesion of the interme- Contribution of the SC to the control of orienting head diate and deep layer of the SC and the reticular formation movements was first firmly established by Hess et al. (1946), ventral to the SC (Fig. 4A), the tonic position of the head T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 211

Fig. 4. Effects of unilateral lesion of the intermediate plus deep layers of the SC and its underlying reticular formation on orienting head movements in the cat. (A) Extent of lesion made by kainic acid injection in the transverse (left panel) and in the parasagittal plane (right panel). (B–D) Orienting to the target ipsilateral to the SC lesion on the 2nd (B), 3rd (C) and 5th (D) postoperative days. Note that the cat overshot the target on the 2nd and 3rd days. Upper panels indicate schematic drawings of head position of the cat. Lower panels indicate stick diagrams of trajectories of the movement of the head axis (sampled at 100 Hz). (E–G) Orienting to the contralateral side to the SC lesion. (E) Orienting on the 2nd postoperative day. The cat oriented to the wrong direction. (F) Orienting on the 5th and postoperative days. (G) Orienting on the 7th postoperative day. The cat gradually became able to orient toward the target. Note that rotation was markedly reduced and compensated by translation. (H) Superimposed records of head positions during orienting on the 3rd and 16th postoperative days. Vertical dotted line indicates the time when the target appeared. 212 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 axis shifted to the ipsilateral side to the lesion (Fig. 4B and to the medial longitudinal fasciculus (MLF) (Altman and C). The head movements to the ipsiversive side overshot Carpenter, 1961; Coulter et al., 1979; Graham, 1977; Huerta the target for the initial few days (Fig. 4B and C), and the and Harting, 1982; Kawamura et al., 1974; Nyberg-Hansen, overshooting disappeared a few days later (Fig. 4D). The 1964; Petras, 1967). On the way, it gives off a large number contraversive head movements were initially undershooting of branches and terminals in the reticular formation in the (Fig. 4E and F). The head movements gradually recovered midbrain, pons, and medulla. In the cervical spinal cord, and 7 days after the lesion the cat became able to move the axons pass in the ventral funiculus and terminate in the the head toward the target (Fig. 4G). However, even at this gray matter. In addition to this crossed descending pathway, stage, the head movement consisted mainly of translation. there also exists the ipsilateral ascending pathway to the Rotation was still small, and the head movements remained mesodiencephalic reticular formation and the thalamus (see slower and less precise. This deficit in performing rotational the subsequent sections). head movements remained for more than 2 weeks (Fig. 4H). Intraxonal staining studies revealed more precise trajec- Thus, the residual system after the SC lesion could partially tories of individual tectofugal neurons, showing that single compensate for the head movements impaired by the SC le- tectospinal neurons project divergent collaterals in the con- sion; however, the extent of compensation was limited. In tralateral pontomedullary reticular formation. Their target other words, the SC and the underlying reticular formation regions included the medial pontomedullary reticular forma- play an essential role for precise control of orienting head tion, the abducens nucleus, the nucleus reticularis tegmenti movements with rotation. pontis, and the nucleus prepositus hypoglossi (Grantyn and Berthoz, 1985). Grantyn and Grantyn (1982) also showed 3.1.3. Single unit recordings in the SC that many of the tectospinal neurons project ascending col- Several lines of study have shown that deep layer neurons laterals to the ipsilateral mesodiencephalic region. The tar- in the caudal SC show phasic increase in activity preced- get regions of the ascending branch include the medial and ing the orienting head movements in unrestrained animals lateral aspects of the reticular formation, central gray, inter- (Harris, 1980; Munoz and Guitton, 1986; Peck, 1990; stitial nucleus of Cajal (INC), nucleus Darkschewitz, me- Straschill and Schick, 1977). Among these studies, Munoz dial aspects of the prerubral area, and the fields of Forel. and Guitton (1985) showed that antidromically identified The ascending projection was also proven for tectoreticular TR(S)Ns showed tonic discharges encoding the gaze po- neurones with projection to RSNs, which mediate disynap- sition error in head-free cats. Furthermore, these TR(S)Ns tic excitation in dorsal neck motoneurons (Alstermark et al., were divided into two subclasses; “fixation TR(S)Ns”, 1992c). Fig. 5 exemplifies the axonal trajectories of tec- which are located in the rostral pole of the SC and dis- tospinal neurons stained with horseradish peroxidase (HRP), charge maximally when the animal attentively fixates the drawn on frontal planes of the pontomedullary reticular target and show pause during the gaze shift, and “orienta- formation (Sasaki and Naito, unpublished observation). As tion TR(S)Ns”, which are located caudally in the SC and shown in the figure, the main axon of the tectospinal neurons have a visual receptive field outside the area centralize and pass in various depths of the medial portion of the reticular discharge for non-zero gaze position error and show phasic formation in or just lateral to the MLF. Collaterals projected increase in activity preceding head movements. Thus, neu- laterally, and the individual axons appeared to terminate ronal activities in the cat SC have been shown to encode within a restricted region of particular depth in the reticular gaze position error and have been involved in both fixation formation, suggesting the existence of functional compart- and execution of orientating head movements (Guitton and ments in the reticular formation along its dorsoventral axis. Munoz, 1991; Munoz and Guitton, 1991; Munoz et al., Among the above areas that receive the tectofugal projec- 1991). Grantyn and Berthoz (1985) showed that tectospinal tion, many areas have been shown to contain neurons that neurons in the deeper layer of the SC showed phasic increase project to the spinal cord. These areas include the medial in activity preceding orienting eye and head movements in pontomedullary reticular formation, the reticular formation alert head-fixed cats. They showed that tectospinal neurons in the mesodiencephalic junction, such as the INC, and the issued extensive branching and termination in the medial fields of Forel (Coulter et al., 1979; Hayes and Rustioni, pontomedullary reticular formation and proposed that the 1981; Holstege and Cowie, 1989; Huerta and Harting, 1982; tectofugal neurons consist of the “tectoreticulospinal” sys- Isa et al., 1988a; Isa and Sasaki, 1992a; Mitani et al., 1988; tem and control the orienting eye and head gaze shift. Nyberg-Hansen, 1965; Petras, 1967; Tohyama et al., 1979). The neurons in these areas mediate a significant portion of 3.1.4. Descending and ascending projections from the SC the disynaptic excitation of neck motoneurons from the SC Many neuroanatomical studies using retrograde and an- as described in the subsequent sections. terograde tracer techniques on the descending pathway In addition to the relay by the neurons in the brainstem from the intermediate and deep layers of the SC revealed reticular formation, axonal trajectories of tectospinal neu- that the major tectospinal pathway crosses within the dorsal rons at cervical segments pass through the ventral funiculus tegmental decussation and descends through the brainstem and terminate mainly in the lateral portion of laminae VII reticular formation in a position just off the midline, ventral and VIII (Nyberg-Hansen, 1965; Petras, 1967; Muto et al., T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 213

Fig. 5. Axonal trajectories of single tectospinal neurons in the pontomedullary reticular formation stained by intraaxonal injection of HRP in the cat. Trajectories of three tectospinal axons are superimposed. The rostrocaudal levels of planes D (B) and E (C) are indicated in A. Arrows in (D) and (E) indicate the stem axons.

1996). Furthermore, by intraaxonal staining technique, was supposed to be mediated by RSNs. The authors also Muto et al. (1996) showed that termination on the soma showed that these RSNs mediate convergent inputs from and/or proximal dendrites of identified neck motoneurons the cerebral cortex by observing spatial facilitation between was not observed. On the other hand, they labeled three the tectofugal and corticofugal fibers in neck motoneurons. spinal interneurons that receive monosynaptic tectal excitation. These neurons were shown to terminate on identified 3.2. Brainstem relay of the descending commands neck motoneurons, suggesting the existence of a disynaptic from the superior colliculus to neck motoneurons pathway from the SC to neck motoneurons mediated via spinal interneurons. Olivier et al. (1995) have claimed that Extensive electrophysiological and anatomical studies there is a monosynaptic projection to neck motoneurons by in cats led to agreement that the reticular formation in the observing post-spike facilitation of EMG activities in neck pons, medulla, and mesodiencephalic junction contains a muscles, but the above anatomical studies and electrophysi- large number of premotor neurons that mediate the descend- ological studies (Alstermark et al., 1992a) so far have failed ing signal from the SC to neck motoneurons. Most of these to supply evidence supporting the existence of monosynap- premotor neurons also receive convergent inputs from the tic connection of tectospinal fibers with neck motoneurons. cerebral cortex. Actually, the rostrocaudal levels of location The first electrophysiological investigation of effects of these premotor neurons were first studied systematically from the SC in dorsal neck motoneurons was performed by by investigating the spatial facilitation of disynaptic tectal Anderson et al. (1971). They observed disynaptic excitation EPSPs with EPSPs evoked by stimulating the pyramidal from the contralateral SC. The authors suggested that the (Pyr) tract either rostral or caudal to lesions made at vari- disynaptic excitation is mediated by RSNs in the brain- ous levels by Alstermark and colleagues as described in the stem. They also reported weak disynaptic effects from the subsequent sections. ipsilateral SC. The tectal effect in neck motoneurons was further systematically studied by Alstermark et al. (1992c). 3.2.1. Tectoreticulospinal pathways via the Systematic tracking with the stimulating electrodes in the pontomedullary reticular formation mesencephalon revealed that large disynaptic excitation Alstermark et al. (1985) first studied the location of relay was induced mainly from the intermediate and deep layers cells mediating the disynaptic Pyr excitation in neck mo- of the caudal half of the contralateral SC. The excitation toneurons. Disynaptic Pyr EPSPs were induced in dorsal 214 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 6. Location of intercalated neurons mediating the disynaptic Pyr excitation in neck motoneurons. The effect of Pyr transection on disynaptic Pyr EPSPs recorded in a SPL motoneuron in acutely anesthetized and paralyzed cat: (A) shows records following the electrical stimulation of the Pyr on the contralateral side above the lower Pyr lesion (C) made at 3 mm rostral to the Pyr decussation; the records in (B) were obtained from a SPL motoneuron by the Pyr stimulation between the lower and upper Pyr lesions (C); (C) a schematic drawing of excitatory (open circle) and inhibitory (closed circle) pathways from the contralateral Pyr to neck motoneurons (modified from Alstermark et al., 1985). neck motoneurons (m. splenius) by stimulation of the Pyr Morphological studies showed the direct termination of between the transection of the Pyr at the caudal medullary RSNs in the motor nuclei of neck muscles in the upper cer- level and that made at the rostral pontine level (Fig. 6). On vical segments (Fig. 8). Axonal trajectory of single RSNs the other hand, when stimulation of the Pyr was performed that receive monosynaptic tectal excitation was studied caudal to the lesion made at the caudal medulla, disynaptic by intraaxonal injection of HRP in the cervical segments excitation almost completely disappeared, and only long (Iwamoto et al., 1988; Sasaki and Iwamoto, 1999; Kakei latency EPSPs with more than trisynaptic linkage remained et al., 1994; Shinoda et al., 1996). RSNs in the NRPc pass (not illustrated). All these results suggested that the disynap- through the ventral funiculus and those in the NRGc pass tic Pyr EPSPs are mediated by RSNs in the pons and medulla through the ventral portion of the lateral funiculus. Both (see Fig. 6C). Based on the above findings, Alstermark and groups of RSNs project collaterals to and terminate in the colleagues systematically investigated the location of the motor nucleus of neck muscles. Shinoda et al. (1996) showed intercalated neurons from the SC by checking the spatial that single RSNs are connected with different groups of facilitation of the tectal EPSPs with Pyr EPSPs in animals neck motoneurons, suggesting the existence of functional with Pyr lesion at different levels of the tract (Alstermark synergy of neck muscles controlled by single RSNs. In et al., 1992a,b). These studies revealed that intercalated neu- addition to motor nuclei, projection to and termination in rons are located mainly in the caudal brainstem, i.e. in the laminae VII and VIII was confirmed. Monosynaptic exci- pons and medulla. Sasaki and coworkers further studied the tatory connection of the RSNs in the NRPc (Iwamoto and candidates of the revealed interneurons mediating the disy- Sasaki, 1990) and NRGc (Sasaki, 1999) with dorsal neck naptic tectal and Pyr EPSPs (Iwamoto et al., 1990; Iwamoto motoneurons has been confirmed by spike-triggered aver- and Sasaki, 1990). The authors systematically recorded from aging of synaptic potentials in neck motoneurons induced a large number of RSNs in the caudal brainstem and found by single RSNs (Fig. 9). These results presented the direct that RSNs in the nucleus reticularis pontis caudalis (NRPc) evidence of the monosynaptic connections from the RSNs and nucleus reticularis gigantocellularis (NRGc) receive to neck motoneurons, which have been suggested by the monosynaptic excitation from the contralateral SC and the earlier studies by Peterson and coworkers (Peterson et al., cerebral peduncle. They observed that disynaptic EPSPs 1974, 1975, 1978). Iwamoto and Sasaki (1990) and Sasaki induced in neck motoneurons by stimulation of the con- (1999) further clarified with the spike triggered averaging tralateral SC were completely collided by the stimulation at technique that connection of individual RSNs was highly the lower cervical segments (Fig. 7B), while stimulation of specific and targeted either to the lateral head flexor sple- the lumbar segments did not induce monosynaptic EPSPs nius (SPL) motoneurons or head elevator biventer cervicis in neck motoneurons (Fig. 7A). Based on the findings, they and complexus (BCC) motoneurons (Fig. 10). concluded that among the RSNs, mainly those terminating In addition to connection with motoneurons at the level at the lower cervical level (C-RSNs) are connected with of the cervical spinal cord, morphological studies using neck motoneurons, while those projecting to the lumbar intraaxonal staining of single RSNs with HRP clarified that segments (L-RSNs) are only partially, if at all, involved in those RSNs in the NRPc project divergent axon collater- mediating the tectal and Pyr excitation in neck motoneurons. als within the brainstem reticular formation. The detailed T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 215

Fig. 7. Occlusion of the tectal and Pyr EPSP induced in neck motoneurons by the volley evoked from the C7. The monosynaptic EPSP evoked by stimulation of the C7 ventral funiculus but not from the L1. The tectal EPSP (B, top) completely occluded (B, bottom) the C7 EPSP (B, middle). The traces in the right column show expanded traces of the left (a pert indicated by arrows in the lowest record) (from Iwamoto et al., 1990).

Fig. 8. Axonal trajectories of single C-RSNs in the NRPc (A) and in the NRGc (B) at the C2 spinal; segment, stained by intraaxonal injection of HRP (Iwamoto et al., 1988). analysis of the branching pattern of the RSNs in the NRPc and vestibular nucleus (VN) and the NRGc. In contrast, the have been made by Grantyn and Berthoz (1987), who RSNs in the NRGc project few collaterals in the medullary stained the RSNs with bursting activities preceding ori- reticular formation (Fig. 11B), suggesting differentiation be- enting movements in head-ﬁxed awake cats and by Sasaki tween RSNs in the NRPc and NRGc. Furthermore, Grantyn (1992) in anesthetized cats (Fig. 11A, C and D). The tar- et al. (1992) noted that some RSNs in the NRPc (pha- gets included the abducens nucleus, prepositus hypoglossi, sic type RSNs) lack collateral projections to the abducens 216 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 9. A direct evidence of monosynaptic excitatory connection of a C-RSN with dorsal neck motoneuron, investigated by the “spike-triggered averaging technique”: (A) identification of a C-RSN (antidromic activation from the C6 but not from the L1), and orthodromic firing from the contralateral SC (tectum; Tec); (B) location of the C-RSN (asterisk) shown in this figure and other C-RSNs, which exhibited monosynaptic connection with neck motoneurons; (C) experimental arrangements of the spike triggered averaging of a unitary EPSP. The uppermost trace indicates the extracellular spikes of the C-RSNs used as trigger for averaging. The middle trace indicated the intracellular potential of the neck motoneuron averaged (over 500 times). The lowest trace is the extracellular field potential. The positive–negative waves in the middle and lowest records, which appear in close relation with the triggered spike, are cross talk of the action potential to the recording system (from Iwamoto et al., 1988).

Fig. 10. Pattern of synaptic connection of RSNs in the NRGc with SPL and BCC motoneurons. (A) Locations of motoneurons in a horizontal plane. Open symbols indicate motoneurons in which unitary EPSPs from the NRGc-RSN were observed, and filled circles were those in which a unitary RSN-EPSP could not be detected. Size of symbols represents relative amplitudes of the EPSPs. Circles indicate SPL and squares indicate BCC motoneurons. (B) Examples of single RSN-EPSPs recorded in four SPL (left panel) and no PSPs in three BCC motoneurons (right panel). A total of 560–1800 sweeps were averaged in these BCC records. Arrows indicate onset time of triggering spike. (C) Summary diagram showing the number of BCC and SPL motoneurons with (filled column) and without (open column) connections for ten NRG-RSNs. Each row represents the number of motoneurons examined for their connection with one RSN (Sasaki, 1999). Note the muscle specificity of connections. T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 217

Fig. 11. Axonal trajectories of single RSNs in the pontomedullary reticular formation, revealed by intraaxonal staining with HRP: (A and B) illustrate low magnification view of axonal trajectories of two RSNs in the NRPc (A) and three RSNs in the NRGc (B). Note that the NRPc-RSNS project a number of axon collaterals in the pontomedullary reticular formation on their way to the spinal cord, while the collateral projection was rare in the three (two uncrossed and one crossed) RSNs in the NRGc. High magnification view of reconstructed axonal trajectories of the two RSNs in the NRPc (shown in (A)) is shown in (C) and (D). The rostrocaudal level of (C) and (D) are indicated in the insets. nucleus. These observations suggest that a population of revealed that the disynaptic excitation is mediated by RSNs RSNs in the NRPc are involved in the control of combined in the caudal brainstem, in common with the tecto-reticular eye and head movements and send signals related to execu- and cortico-reticular fibers. tion of gaze shift simultaneously to neurons in various nuclei besides extraocular and neck motoneurons. Whereas, the 3.2.2. Tectoreticulospinal pathways via the reticular pontomedullary reticular formation includes other popula- formation in the mesodiencephalic junction tions of RSNs that receive excitation from the SC but do not The first systematic study on ascending projection from project to the extraocular motor nuclei and presumably are the SC was performed by Harting et al. (1980), using antero- involved in control of head movements and/or movements grade tracers and revealed that the SC projects to the retic- of other body parts (Grantyn et al., 1992). The targets of ular formation in the mesodiencephalic projection. Grantyn RSNs in the NRPc and NRGc are summarized in Fig. 15A. and Grantyn (1982) showed that a population of tectofugal In addition to the SC and cerebral peduncle, stimula- neurons bifurcate into descending and ascending branches tion of the midbrain tegmentum, the nucleus cuneiformis, and the ascending branches project to the reticular formation induced disynaptic EPSPs in ipsilateral neck motoneurons in the mesodiencephalic junction. Alstermark et al. (1992c) (Alstermark et al., 1992c). Testing of spatial facilitation systematically mapped the effective region for evoking 218 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 tectal disynaptic effects on neck motoneurons and showed We found that mono- and disynaptic EPSPs were evoked that disynaptic tectal EPSPs could be induced by stimu- from FFH in neck motoneurons and further investigated the lating the ascending tectal branches which could be traced descending pathways mediating these EPSPs in cats (Isa up to the mesencephalic tegmentum and fields of Forel et al., 1988a,b; Isa and Sasaki, 1992a,b; Isa and Itouji, 1992). (Forel’s field H (FFH)). Electrical stimulation of this region Injection of HRP into the upper cervical spinal cord resulted in alert cats has been shown to produce dorsiflexion of the in retrograde labeling of a number of neurons in FFH (and head (Hess, 1956; Hassler, 1972). The mesodiencephalic also in INC) on the ipsilateral side. Injection of HRP into the junctional regions including FFH and the INC have been NRGc resulted in retrograde labeling of a greater number of shown to control vertical eye movements (Pasik et al., 1969; neurons in these regions. Thus, neurons in FFH were shown Büttner et al., 1977; Nakao and Shiraishi, 1985; Nakao et al., to project descending projections to the NRGc and some of 1990). It has been strongly suggested that FFH and INC are them further down to the upper cervical spinal cord. Injec- involved in control of orienting movement especially in the tion of WGA-HRP into FFH revealed that descending fibers vertical direction. Stimulation of the INC has been shown to from FFH terminated mainly in the NRGc, while sparse ter- elicit monosynaptic EPSPs in neck motoneurons; however, mination was observed in the NRPc (Isa and Sasaki, 1992a). the INC is more likely related to tonic posture of the head Termination in the ventral horn of the upper cervical spinal (Fukushima et al., 1978; for review see Fukushima, 1987). cord was observed (Isa and Sasaki, 1992a). Anterograde

Fig. 12. Projection of FFH neurons to the cervical spinal cord and monosynaptic connection with BCC motoneurons: (A–C) systematic antidromic threshold mapping that reveals the axonal trajectories of a single FFH neuron. Experimental arrangement is indicated in A and antidromic spike of a FFH neuron from the C3 segment is indicated in (B). (C) Results of systematic antidromic threshold mapping at four different rostrocaudal levels in the caudal portion of the C3 segment. Threshold of antidromic activation is indicated by the diameter of the circles (see the inset) and antidromic latencies from individual points are indicated by numerals in the figure (in ms); (D–G) monosynaptic EPSP recorded in a BCC motoneurons following stimulation of the ipsilateral FFH. (D) EPSPs in a BCC motoneuron evoked by stimulation of FFH. Short segmental latency EPSPs with fixed onset suggests the monosynaptic origin. (E) Double shock stimulation revealed marked temporal facilitation, suggesting the existence of potent disynaptic pathway. Note an arrow which points to a notch on the rising phase of the EPSP, which indicates the onset of disynaptic component ((F) and (G)) systematic tracking with stimulating electrode at two different rostrocaudal levels. Numerals in (G) correspond to the stimulation sites indicated in (F). The figure shows that effective site to induce the monosynaptic EPSP is restricted to inside FFH (modified from Isa and Sasaki, 1992b). T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 219 labeling study was also performed by Holstege and Cowie EPSPs and IPSPs were recorded in RSNs mainly in the (1989), who reported the termination of FFH neurons in cer- NRGc but also observed in the NRPc-RSNs following stim- vical segments. Projection to neck motor nuclei of single ulation of FFH (Fig. 13A and B). Spike triggered averaging FFH neurons was clarified by systematic antidromic thresh- of the membrane potential of RSNs by action potentials old mapping (Fig. 12A–C). Stimulation of FFH elicited of FFH neurons revealed monosynaptic excitatory connec- monosynaptic EPSPs as judged from the segmental latencies tion from FFH neurons to RSNs (Fig. 13C and D). Since measured from the conduction volley (Fig. 12D and upper these RSNs are known to project to neck motoneurons, a record in Fig. 12E; Isa and Sasaki, 1992b). When double major portion of the disynaptic EPSPs was assumed to be shock stimulation is applied, large disynaptic EPSPs (start- mediated by the FFH-reticulospinal pathway. ing after the notch in the rising phase indicated by an ar- The axonal trajectories of individual FFH neurons were row in Fig. 12E, lower record) are additionally induced in revealed by systematic antidromic threshold mapping tech- the head elevator BCC motoneurons, but there is virtually nique in the pons, medulla, and cervical spinal cord. FFH no effect in lateral head flexor SPL motoneurons (Isa et al., neurons were classified into two subgroups from their pro- 1988a; Isa and Sasaki, 1992b) (Fig. 12D–G). Thus, it was jection pattern (Isa et al., 1988b; Isa and Itouji, 1992). Type shown that FFH neurons exert excitatory effects directly on I neurons project both to the oculomotor nucleus and to the neck motoneurons that control the vertical component of NRGc (some of them project further down to the cervical head movements. spinal cord) (Fig. 14A–D), while type II neurons lack pro- We further analyzed the projection of FFH neurons to jection to the oculomotor nucleus but project to the NRGc, the pontomedullary reticular formation. Stimulation of FFH and some of them descend further down to the cervical induced negative field potentials in the NRPc and NRGc. spinal cord (Fig. 14E–H). The former neurons may con- The field potential was larger in the medioventral part of the trol both eye and head movements, while the latter control NRGc. Intracellular recordings revealed that monosynaptic mainly head movements as immediate premotor neurons.

Fig. 13. Monosynaptic excitatory connection from FFH neurons to RSNs in the pontomedullary reticular formation (modified from Isa and Sasaki, 1992b). (A and B) Intracellular recording from a C-RSN in the NRPc. Identification of the C-RSN is indicated in the insets of (A). (B) This shows the effect of stimulation in FFH at different stimulus strength. Short and fixed onset of the EPSPs suggests the monosynaptic origin of the EPSPs. Amplitude of the EPSPs induced from different stimulation sites are indicated by diameters of the circles (see the inset) plotted at the individual stimulation sited in (A). (C and D) Direct evidence of the monosynaptic excitatory connection from FFH to RSNs revealed by the spike triggered averaging technique. (C) This shows the experimental arrangement. In (D), the action potential of the FFH neuron is indicated on the top and averaged membrane potential (averaged over 512 times) of the RSN triggered by the spike is indicated in the middle and the bottom trace indicates the average of the extracellular field potential. 220 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 14. Axonal trajectories of two different types of FFH neurons in the mesencephalon and pontomedullary reticular formation as revealed by systematic antidromic threshold mapping technique (Isa et al., 1988b). Parts (A–D) show the axonal trajectory of a type I FFH neuron that project to the oculomotor nucleus (III): (A) indicates the location of the cell and (B) shows the antidromic activation from the III, NRGc and C1 level of the spinal cord. (D) This indicates the results of the antidromic threshold mapping. Diameters of the circle indicate the antidromic threshold and latencies of the antidromic spikes are indicated by numerals in the ﬁgure (in ms). (C) This indicates the longitudinal level of mapping in (D) with corresponding numbers; (E–H) axonal trajectory of type II FFH neuron that lacks collateral projection to the III. The same arrangement as (A)–(D) (modiﬁed from Isa et al., 1998).

In addition, we found another group of neurons (type Ia), 4. Effects of lesion of the brainstem relay which projected to the oculomotor nuclei but lacked projec- structures on orienting head movements tion to the NRGc and were interpreted to be speciﬁed for the control of eye movements. The characteristic pattern of 4.1. The pontomedullary reticular formation descending projection of types I and II FFH neurons is summarized in Fig. 15B. The diagram shows the homology of To elucidate the functional role of the pontomedullary FFH neurons to the RSNs in the NRPc and NRGc (Fig. 15A) reticular formation in control of orienting head movements, with respect to projection to motoneurons innervating ex- we performed local lesion using kainic acid injection in traocular and neck muscles. awake animals (Isa and Sasaki, 1988; Sasaki et al., 1999). As to the input organization, the effect of stimulation of When the NRPc and the rostral one third of the NRGc the ipsilateral SC has been investigated in alert cats (Nakao were impaired (Fig. 16G), cats could not make rapid orient- et al., 1988, 1990; Isa and Naito, 1994), and excitatory ing head movements and saccadic eye movements toward action of mono- or disynaptic latencies was observed in the ipsiversive direction to the lesion (Fig. 16B–E). Slow FFH neurons. Other input origin from higher structures translation of the head remained; however, swift rotation has not been examined using electrophysiological tech- could not be observed (Fig. 16D). Lesions restricted either niques yet. to the NRPc or NRGc produced only limited deﬁcits in T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 221

Fig. 15. A schematic drawing of descending projection of two types of RSNs in the NRPC/NRG (A) and two types (types I and II) of FFH neurons (B) (modiﬁed from Isa and Itouji, 1992).

orientating head movements. Thus, at such gross level, the 5. Single unit activities of neurons in the brainstem NRPc and NRGc appeared to be able to partially compen- reticular formation during orienting sate the function of each other. Functional differentiation between the two structures required unit recordings during 5.1. Single unit activity of neurons in the orienting (see Section 4.2). In these animals, a vertical com- pontomedullary reticular formation ponent of movements was fairly normal when the head was oriented toward the intact side, while the vertical movements Single unit activities of neurons in the brainstem retic- and oblique movements were severely damaged when the ular formation have been investigated in awake cats dur- head was oriented to the lesioned side (Sasaki et al., 1999). ing a variety of free movements (Siegel and Tomaszewski, Besides our study, Suzuki et al. (1989) showed that injection 1983), during orienting in head-free cats (Isa and Naito, of ibotenic acid into the medullary reticular formation caused 1995; Sasaki et al., 1996), and in head-restrained awake initial excitatory effects that caused ipsiversive head turn and cats (Delgado-Garcia et al., 1988; Vidal et al., 1982, 1983; later impairment of ipsiversive rapid head movements. This Grantyn and Berthoz, 1987; Grantyn et al., 1987, 1992, result supported that swift orienting head movements are 1993). controlled mainly by the ipsilateral pontomedullary reticular formation. 5.1.1. Directional tuning of the NRPc and NRGc neurons during orienting 4.2. Forel’s field H We analyzed the single unit activities of the pontomedullary reticular neurons during orienting movements We also investigated the effect of lesion of FFH by kainic in alert head-free cats (Isa and Naito, 1995). Neurons were acid injection. Unilateral injection induced marked asym- identified as relay cells that receive mono- or oligosynaptic metry in tonus of neck muscles, but after bilateral injection, excitatory inputs from the contralateral SC and/or cere- the asymmetry disappeared. After bilateral FFH lesion, ver- bral peduncle by stimulation through chronically implanted tical rotation of the head was severely impaired. Vertical electrodes. A portion of these neurons was successfully translation partially remained, which was executed mainly identified as RSNs by antidromic activation from the cer- by limbs (Fig. 17C–E; Isa et al., 1992b). In these animals, vical spinal cord. A majority of neurons fired maximally horizontal head movements were not impaired. These re- during orienting in a particular direction. The directional sults suggested that FFH is involved mainly in control of preference of the reticular neurons was investigated by the vertical component of head movements, while the NRPc comparing the activity during movements toward eight and a part of the NRGc mainly regulate the horizontal different directions, each of which was separated by 45◦ component of head movements. Single unit recording study (Fig. 18C). Activity of neurons with directional preference described in the subsequent sections further supports this was well fitted by cosine function (Fig. 18D). Most neurons hypothesis. in the NRPc (24/25) showed preference for movements in 222 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 16. Effects of unilateral kainic acid lesion of the NRPc and the rostral portion of the NRGc on horizontal orienting head movements: (A) schematic drawing of the orienting task; (B and C) EMG, EOG and head displacement during the task in an intact cat (B) and after lesion (C). The onset of the had movement is indicated by the vertical dotted line. When the target was moved to the lesioned side, the animal made slow (C(1a)) or no movement of the head (C(1b)), while swift movement to the intact side C2. Arrows indicate the time when the target started to move; R, right; L, left. (D) Examples of the head displacements after the lesion. Upper and lower traces are head turning to the lesions side and intact side, respectively. Ten traces are superimposed, aligned on the presentation of the visual stimulus (E) amplitudes and the maximum angular velocity during head movements are plotted. Control trials, movements toward the lesioned side; and movements toward the intact side are indicated by closed triangles, open circles and closed circles, respectively. The extent of lesion is shown in (F) and (G). To identify the loss of RSNs, HRP was injected into the C2 gray matter and RSNs were retrogradely labeled (closed circles) using tetramethylbenzidine (TMB) method. (F) This indicates the lateral projection of the extent of the lesion (modiﬁed from Isa and Sasaki, 1988). T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 223

Fig. 17. Effects of bilateral lesion of FFH: (A) EMG, EOG and head displacements during upward orienting in intact animals; (B) extent of lesion at different rostrocaudal levels, indicated by “hatch”; (C) trajectories of upward head movements in the control, 3 and 8 days after lesion; (D) stick diagrams of the upward head movements in (C), sampled at 100 Hz. (E) EMG and EOG records during orienting movements. Note that saccadic eye movements are not induced, while the vestibuloocular reflex could be clearly observed (from Isa et al., 1992a). the ipsilateral horizontal direction (between −38 and 16◦, many reticular neurons in the NRPc and NRGc increased Figs. 19 and 23A). Similarly, a majority of neurons in the firing in close temporal relation to saccade and neck EMG lateral portion of the NRGc (1–2 mm from the midline) activity. They classified them into three subtypes according had preference in ipsilateral horizontal direction (0–10◦), to their firing pattern (Grantyn et al., 1992). We observed while medially and ventrally located neurons tended to neurons with similar categories in alert head-free cats. PNs have preference for movements in the oblique or vertical fired just before saccadic eye movement and ceased firing directions (Figs. 19 and 23B). This can be explained by the before its termination (Fig. 20A). Total number of their observation that FFH neurons project preferentially to the spikes during bursts was correlated with angular velocities medial portion of the NRGc and in the deeper part of the (horizontal component) of saccades and head movements. NRPc. These results gave critical evidence that the NRPc Phasic type neurons were further subdivided into “long and the major portion of the NRGc are involved mainly in lead” and “medium lead” type neurons. Phasic sustained the control of horizontal orienting. neurons (PSNs) started phasic firing just before the head The results that the NRGc contains neurons with preferred prelude (phasic component) and continued firing during direction to oblique or vertical directions fit the observation the head movements until the termination (sustained com- by Delgado-Garcia et al. (1988), who found phasic neurons ponent) (Fig. 20B). The phasic component was correlated (PNs) with preference to vertical movements in (the dorsal with the head velocity, while the sustained component was portion of) the medullary reticular formation. In addition to correlated with the head position. Phasic sustained type the directional preference, difference in the degree of modu- neurons were further subdivided into augmenting type and lation depth was observed between the neurons in the NRPc plateau type. In addition, tonic neurons (TNs) did not show and NRGc. The NRPc neurons exhibited more prominent phasic component but exhibited tonic firing related to the phasic increase in activity and large modulation depth com- position of gaze (Fig. 20C). Behavioral correlates of their pared with the NRGc neurons. This reflects relatively high activities have suggested that the PNs are mainly concerned baseline activity of neurons in the NRGc compared with the with the velocity-sensitive system, while the PSNs are in- NRPc (Isa and Naito, 1995). volved in both the position-sensitive and velocity-sensitive systems. Grantyn et al. (1992) showed that PNs and TNs 5.1.2. Firing pattern of reticular neurons and their in the NRPc and the NRGc lack collateral projection to the morphological and physiological correlates abducens nucleus and other reticular structures, while the Berthoz (1989) have made systematic analysis of neu- PSNs projected extensive collaterals to the abducens nucleus ronal activities during orienting in head-fixed awake cats by and other reticular structures. The PSNs may play a signif- simultaneously monitoring eye movements and neck EMG icant role in controlling synergic eye and head movements, activity (Vidal et al., 1982, 1983; Grantyn and Berthoz, 1987; while the PNs can control head movements independent of Grantyn et al., 1987, 1992, 1993). The authors showed that eye movements. 224 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 18. An example of the activity of an identified “orienting-related” RSN in the NRPc during orienting movements in eight different directions and its directional preference (modified from Isa and Naito, 1995): (A) location of the cell is indicated with a closed circle on a parasagittal view of the brain stem (1.00 mm from the midline) movements; (B) rastergram and perimovement histogram over eight ipsiversive horizontal movements are illustrated with horizontal and vertical head position (hH and vH) and velocities (hH and vH), horizontal and vertical EOG signals (hEOG and vEOG) and rectified EMG records of mSPL and Mbc; (C) rastergrams, average of rectified EMG of mSPL and mBC, and average of perimovement histograms of unit activity are shown for movements in each direction. Timing of visual stimulus is indicated with a small square below each rastergram; (D) fitting of directional tuning of activity with cosine function. Mean and SD of data in (C) are plotted (from Isa and Naito, 1995).

Our recent study (Sasaki et al., unpublished observation) showed increase in activity whose onset was locked to timing showed the neuronal activities of the pontomedullary reticu- of target presentation (Isa and Naito, 1995). Fig. 21 shows lar neurons during velocity-guided and position-guided ori- the activity of an identiﬁed RSN in the NRPc. This neu- enting. In the case of velocity-guided orienting (Fig. 2), both ron showed phasic increase in activity locked to the visual PNs and PSNs were activated. Especially, PSNs start to ﬁre stimulus at the latency of approximately 40 ms (Fig. 21B). preceding the initial small drift of the head (Fig. 1A, open When the gaze movement was initiated with longer latency arrow), while PNs are activated apparently in coincidence (lower traces in A), the increase in activity continued until with the synchronous eye and head movements. Whereas, the onset of the movement-related increase in activity. Simi- in the case of position-guided orienting, virtually only PNs lar type of activities in RSNs have been reported by Grantyn were activated. These results suggested that phasic and (1989). This type of activity suggested that the activity of PSNs were activated differentially in a context dependent RSNs is not exclusively motor, but includes activity that is manner. purely visual in nature. Thus, the very early visual signal In addition to the change in activity related to execution reaches the spinal cord already 40 ms after presentation of of movements, some proportion of pontine reticular neurons stimulus via RSNs. T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 225

Fig. 19. Locations of 25 pontine and 21 medullary reticular neurons investigated in Isa and Naito (1995) are plotted in parasagittal plane. Preferred direction of each neuron is indicated by arrow direction. Inset indicates the arrangement of preferred direction (from Isa and Naito, 1995).

1994). Among them, 20 neurons showed phasic increase in activity that preceded the onset of head movements by 20–100 ms (Fig. 22). The neuron shown in Fig. 22 exhibited preference for upward direction (the preferred direction was 86◦). All of the 18 tested neurons were excited from the ipsilateral SC at the latency of 0.8–1.8 ms, which suggested mono- or disynaptic linkage. Seven of the 18 tested neurons were identified as descending neurons by antidromic activation from the caudal medullary reticular formation. The magnitude of phasic increase in activity was linearly correlated with vertical angular velocity of head movements. Directional preference was systematically investigated in 12 neurons and the directional tuning could be closely fitted by cosine function. Most of them showed preference for movements in the upward direction (Figs. 22B and 23C). Furthermore, most FFH neurons with phasic increase in activity preceding upward head movements exhibited increase in activity lagging the onset of downward head movements (Fig. 22A). This delayed increase in activity well coincided with the increase in the EMG activity of the head elevator m. biventer cervicis. This suggested that the late increase in activity of FFH neurons is involved in terminating the downward head movements by activating the head elevator muscles during the late phase of movements. Interestingly, activity of most FFH neurons during ipsiversive upward and contraversive upward movements was Fig. 20. Activity of “phasic” (A), “phasic sustained” (B) and “tonic” (C) virtually the same, suggesting that neurons on both sides type neurons in the pontomedullary reticular formation during orienting. of FFH showed similar activation during oblique move- The unit activity is presented with gaze and head movement records. ments (see Fig. 22A and B). Repetitive electrical stimulation on one side of FFH induced ipsiversive oblique up- 5.2. Single unit activity of neurons in Forel’s field H ward movements accompanying torsion (Fig. 24A and B), and combined stimulation of both sides induced purely up- We recorded from a total of 63 neurons in FFH during ward movements (Fig. 24C). These results suggested that orienting movements in alert head-free cats (Isa and Naito, the CNS involves a neuronal constraint to activate FFH 226 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 21. Example of a RSN with “stimulus-locked activity”: (A) spike activity of single trials during movements in ipsilateral horizontal direction are aligned with timing of visual stimulation (vis.). Timing of onset of head movements are indicated with triangles (mov.). (B) Rastergram and average peristimulus (left) and perimovement (right) histograms (0◦ direction) (modiﬁed from Isa and Naito, 1995).

on both sides to the same degree, and equal bilateral ac- during movements in either direction. For instance, SPL tivation of FFH encodes the vertical component of head and m. complexus are recruited unilaterally during horizon- movements. tal movements, while activated bilaterally during vertical movements (Roucoux et al., 1989). Then, the question is how the premotor commands are formulated to control 6. Differential control of horizontal and vertical these neck muscles for execution of movements in various components of head movements directions. All of the above results strongly suggested that horizontal and vertical components of head movements From the viewpoint of muscle types, horizontal move- are controlled by separate channels in the downstream of ments are executed by muscles with a high proportion of the SC. fast fibers, and among these, the splenius muscle shows Masino and Knudsen (1990) suggested that the same strat- the most vigorous activation (Richmond and Vidal, 1988). egy is used in control of head movements in barn owls. The In contrast, vertical movements appear to be produced by authors took advantage of the refractory period in inducing recruiting mainly extensor muscles containing a high pro- head movements that follows repetitive stimulation in one portion of slow fibers. Especially, m. biventer cervicis and side of the SC. They stimulated both sides of the SC sequen- occipitoscapularis, which are tonically active when the head tially within a time interval shorter than the refractory period is raised or head is stationary in most postures, are impor- and carefully observed the movements induced by the sec- tant (Richmond et al., 1992). However, recruitment of a ond stimulus. If the movements induced by the second stim- given muscle is not strictly divided into horizontal or ver- ulus had a common horizontal component with movements tical dichotomy. Or rather, a given muscle can be recruited induced by the first, even though the vertical components T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 227

Fig. 22. Directional preference of phasic activity of an FFH neuron: (A) activity of an FFH neuron during orienting in eight directions. Rastergram, averaged rectiﬁed EMG records of mSPL and mBC, and the averaged perimovement histogram of the unit activity. (B) Polar plot of the activity of 12 well-analyzed upward preferring neurons in FFH. The records were standardized to give a maximum value of 100%. (C) Location of the 20 FFH neurons that showed phasic increase in activity preceding movement. Closed circles (n = 19) preferred upward, while the remaining one neuron preferred downward direction (modiﬁed from Isa and Naito, 1994).

Fig. 23. Preferred directions of 25 pontine (A) and 21 medullary (B) and 12 FFH (C) neurons investigated in Isa and Naito (1994, 1995). were in the opposite directions, the horizontal component horizontal components were in the opposite directions, the of the movement induced by the second stimulus was elim- second stimulus could induce only a horizontal component inated by the refractoriness and only the vertical component of its original vector. Our single unit recording and lesion of its original vector was induced. When the second stimu- studies suggested that the NRPc neurons are controlling lus had a common vertical component with the ﬁrst, but the the horizontal component and FFH neurons control the 228 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 24. Effects of repetitive microstimulation of FFH (400 Hz, 40 trains; frontal view of the head movements, sampled at 200 Hz, during the stimulation period (100 ms)): (A) stimulation of the left-side; (B) stimulation of the right-side; (C) simultaneous bilateral stimulation (from Isa and Naito, 1994).

vertical component (Isa and Naito, 1994, 1995). Descending 7. Higher order structures that regulate the primary commands from both structures converge at the level of the brainstem pathways for orienting NRGc, where neurons with various preferred directions are intermingled (Fig. 23B). A variety of regions in the cerebral cortex, basal ganglia, The above-described hypothesis about the horizon- and cerebellum (chiefly the fastigial nucleus (FN)) have tal/vertical dichotomy of command mediating systems been shown to be involved in control of orienting move- should face the problem associated with the posture. Since ments by experiments using microstimulation, lesion, single the activation pattern of neck muscles may vary depend- unit recording, and neuroanatomical techniques. Among ing on the posture of the neck, questions may arise how these, we review the structures that directly project to the the command generating system should deal with the neurons involved in the tectoreticulospinal pathways. posture-dependent muscle activation pattern. In this regard, Thomson and coworkers showed, based on the activation 7.1. Pericruciate cortical areas pattern during horizontal head movements with different postures, that neck muscles of cats can be divided into Intracortical microstimulation studies in alert cats clari- two subgroups (Thomson et al., 1994, 1996). Five muscles fied that various areas are involved in control of orienting: (obliquus capitis inferior, SPL, levator scaplae, complexus, the medial and lateral bank of the presylvian sulcus, the and BC) displayed activation patterns that did not change fundus of the coronal sulcus and ventral bank of the anterior when the cats adopted a different neck posture. Most of ectosylvian sulcus. In addition, several occipital regions are these muscles are dorsally located and span many cer- involved in control of orienting movements. These areas vical joints. On the other hand, five other neck muscles include areas 17–19, 21a, 21b, 20a, 20b, PS, 7p, AMLS, (semispinalis cervicis, longissimus capitis, levator scaplae PMLS, VLS, ALLS, PLLS, DLS, ALG, EVA and insula ventralis, scalenus anterior, and obliquus capitis superior) (Sprague et al., 1977). are modulated dependent on the posture. These muscles Hassler (1966) first performed systematic stimulation are deeper and laterally located and generally span fewer studies of the frontal cortex in cats and showed differ- cervical joints. These observations suggest that the basic ent subregions of the area 6 (6a␤,6a␣,6a␦,6a␥, 6if-fu) motor program generated in the SC may regulate mainly the evoked various types of head movement (turning, rotation, invariantly activated muscle groups in the dorsal side of the horizontal, vertical, ipsiversive, or contraversive move- neck, while the neck posture selectively modulates the ac- ments). Schlag and Schlag-Rey (1970) further showed tivation pattern of ancillary muscles. Organizing the motor that eye movements with various directions were induced output in this manner might simplify the task of computing by stimulation of the precruciate region, 6a␤, and named the appropriate patterns of neck muscle activation. If we can this area “the frontal oculomotor region” in cats. Guitton rest on this assumption, the strategy to divide the command and Mandl (1978a,b) studied the effects of microstimula- generation system of the head movement into two broadly tion and unit activity in the frontal oculomotor region and tuned horizontal and vertical vector components may help showed that electrical stimulation of the frontal oculomotor to reduce the number of degrees of freedom and simplify region induced phasic activation of dorsal neck muscles, the computation associated with control of the complex and neurons in this region showed phasic activity preceding motor system such as the neck. combined eye movements and neck EMG activities. Thus, T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 229 they suggested that the frontal oculomotor region of cats is 1984; Isa and Sasaki, unpublished observation). Following involved in control of orienting head movements. injection of HRP into the NRPc, retrogradely labeled cor- Using the anterograde degeneration technique, Kuypers ticoreticular neurons were distributed mainly in the medial (1958) showed that neurons in the pericruciate region project portion of the ventral bank in the pericruciate sulcus on the to and make termination in the medial pontomedullary retic- bilateral sides, concentrated in the areas 6a␣,6a␤, the pre- ular formation and showed some regional differences in the frontal cortex, and medial half of area 4if, areas 6if-fu and projection pattern in the cat. Miyashita and Tamai (1989) 4s-fu. Smaller numbers of neurons were also found in areas analyzed the descending projection from the frontal ocu- 3, 1, 2, 43, and both dorsal and ventral banks of sulcus ecto- lomotor region by using an anterograde tracing technique. sylvius (s.esyl) (Fig. 25D). These regions include the frontal They showed projections to the deep layers of the SC, and oculomotor region of the cat (Schlag and Schlag-Rey, 1970; the reticular formation in the diencephalon, mesencephalon, Tamai et al., 1987). In contrast, injection of HRP into the pons and medulla, including regions such as the fields of NRGc resulted in retrograde labeling of neurons mainly in Forel, the INC, posterior commissure nucleus, cuneiform the central to lateral portion of the ventral bank of the peri- nucleus, paramedian pontine reticular formation (PPRF) and cruciate gyrus, the neck and forelimb region of the area 4␥ dorsal medullary reticular formation. Projection to the SC (Nieoullon and Rispal-Padel, 1976). Furthermore, smaller and more rostral region was predominantly ipsilateral, how- numbers of labeled neurons were found in areas 6a␣,6a␤, ever, the contralateral projection dominated at the medullary 6if-fu and 4s-fu, areas 3, 1, 2, 43 and both ventral and dor- level. The variation between the cortico-reticular projec- sal banks of s.esyl (Fig. 25C). These results were consistent tions from different areas in the pericruciate region was with the results of anterograde labeling by Matsuyama and intensively studied by Matsuyama and Drew (1997) using Drew (1997). an anterograde tracer Phaseolus vulgaris-leucoaggulutinin We further analyzed the axonal trajectories of single cor- (PHA-L). They showed that the hindlimb region of the mo- ticofugal neurons in the pericruciate region by systematic tor cortex (4␥) projects weakly but mainly to the NRGc, antidromic threshold mapping techniques (Isa and Sasaki, while the forelimb region projects more densely to both unpublished observation). As shown in Fig. 26A–C, a sin- the NRPc and NRGc. The major cortical input to the pon- gle neuron in the area 6a␤ projected to the bilateral sides tomedullary reticular formation originates from the areas of the pontine reticular formation. The neurons appeared to 6a␤ and 4␥. The former projected mainly to the NRPc terminate in the middle part of the NRPc (and in the nucleus while the latter to the NRGc. reticularis tegmenti pontis). Fig. 26D–F shows the axonal Differential control of the NRPc and NRGc were also trajectories of a single corticospinal neuron in the forelimb studied by retrograde labeling of cortical neurons by in- region of area 4␥. This neuron projected to the ventral por- jecting HRP in the respective areas (Keizer and Kuypers, tion of the NRPc, the caudal part of the NRGc, leaving some

Fig. 25. Distribution of cortico-reticular neurons projecting to the NRGc (C) and NRPc (D) in the ventral and dorsal bank of the cruciate sulcus on the ipsilateral side of injection in the cat. Injection sites are indicated in the text. The arrangement of (C) and (D) (unfolding of the sulcus) is explained in (A) and (B). Each section was 100 ␮m thick. Each dot indicates single labeled cell with HRP. 230 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 26. Axonal trajectories of single cortical neurons in area 6a␤ and area 4␥ in the pontomedullary reticular formation as revealed by systematic antidromic threshold mapping technique: (A–C) a neuron in area 6a␤. Location of the cell is indicated as a closed circle in (A) and results of antidromic threshold mapping are indicated on the parasagittal plane in (B) (1.0 mm lateral from the midline) and on the coronal planes (the rostrocaudal level is indicated by the corresponding numbers in (B). Diameters of the circles indicate the threshold for antidromic activation (see insets) and numerals indicate latencies in ms; (D–F) axonal trajectory of single corticospinal neurons in area 4␥. The same arrangement as (A–C); (D) location of the cell; (E) antidromic threshold mapping in the parasagittal plane; (F) antidromic threshold mapping in the frontal plane (the rostrocaudal level is indicated in (E)). T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 231

Fig. 27. Comparison of EPSPs induced in RSNs in the NRPc ((C), left) and NRGc ((C), right) from pericruciate cortical regions. Stimulation strength was 1 mA: (A) stimulation sites; (B) location of recorded RSNs; (C) EPSPs induced by cortical stimulation at the location with the corresponding number in A; (D) a schematic drawing of the cortico-reticulo-spinal connection. empty zone in the rostral part of the NRGc. It also pro- no difference in the cortical input pattern was observed jected to the nucleus reticularis magnocellularis in the caudal between the C-RSNs and L-RSNs. These results suggested medulla on its way to the cervical spinal cord. The neuron that the frontal oculomotor region is involved in control of projected to the bilateral sides of the medullary reticular for- orienting movements via the pathway through the SC and mation. It is noted that the neuron in area 4␥ (Fig. 26D–F) also via the pathway directly to the reticular formation in had faster axonal conduction velocity than that in area 6a␤ the pons and medulla, bypassing the SC. (Fig. 26A–C) according to the antidromic latencies. Similar In addition to the cortical input origin, the RSNs in the results were obtained in another 15 corticoreticular neurons, NRPc and in the NRGc differ in the relative strength of thus, it has been suggested that there may be further func- the inputs from the SC and cerebral cortex. Field poten- tional differentiation between the NRPc and NRGc (see the tial recordings in the reticular formation and stimulation of subsequent paragraphs). cortico- and tectoreticular fibers showed that the rostral por- Intracellular recording studies from the RSNs combined tion of the pontomedullary reticular formation receives a with electrical stimulation of the pericruciate region in stronger input from the SC and the caudal region receives anesthetized cats supported the above anatomical results. stronger input from the pyramid (Alstermark et al., 1992a). Many authors showed that cortical stimulation induced Iwamoto et al. (1990) compared amplitudes of monosynap- monosynaptic EPSPs in RSNs in the pontomedullary retic- tic EPSPs evoked by stimulation of the SC and cerebral ular formation (He and Wu, 1985; Magni and Willis, peduncle in anesthetized cats. They clarified that RSNs in 1964; Pilyavsky and Goskin, 1978). We studied the the NRPc receive stronger excitation from the contralateral cortico-reticular projection by intracellular recordings from tectum (SC) than from the contralateral cerebral peduncle, the RSNs in the anesthetized cats and showed that stim- while those in the NRGc receive stronger excitation from the ulation of the pericruciate cortex induced monosynaptic cerebral peduncle than from the SC (Fig. 28). These results EPSPs in the RSNs on bilateral sides. RSNs in the NRPc were also confirmed by analysis of negative field potential received stronger excitation from the medial portion of induced by stimulation of the SC and the cerebral peduncle the pericruciate gyrus bilaterally, while those in the NRGc (Iwamoto et al., 1990). received stronger excitation from the more lateral portion Stimulation of the frontal oculomotor region has been (Fig. 27A–C) including area 4␥. In addition to the differ- shown to induce purely horizontal and vertical as well as ences in amplitude, duration of the EPSPs was longer in oblique eye or head movements (Guitton and Mandl, 1978a; RSNs in the NRPc suggesting slower conduction of corti- Schlag and Schlag-Rey, 1970; Tamai et al., 1983). Then the coreticular axons of neurons in area 6 than those in area 4 question may arise as to how these vertical and horizontal as mentioned above. Furthermore, as shown in Fig. 27D, movements are produced. We have described that there 232 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 28. Comparison of tectal and Pyr EPSPs in RSNs in the NRPc and NRGc: (A) amplitudes of tectal monosynaptic EPSPs are plotted against the location of the recorded cells; (B) same as (A) for Pyr EPSPs; (C) parasagittal plane of the brainstem. Arrow indicates the standard plane RC = 0mm in the rostrocaudal coordinates used in horizontal axes in (A) and (B); (D) Examples of tectal and Pyr EPSPs from four RSNs located in R1, R1, C4 and C5 planes; (E) plots of ratio of tectal EPSPs amplitude vs. Pyr EPSPs against the location of the RSNs recorded. Three RSNs had no tectal EPSPs (three lowest points in plane C4) (from Iwamoto et al., 1990). exist two descending pathways from the frontal oculomotor oculomotor and visual functions (Carpenter and Batton, region to the extraocular and neck motoneurons; one via the 1982). The Purkinje cells in the vermis project to the FN. SC and another via the direct projection to the brainstem Pelisson and coworkers have investigated the effects of reticular formation. In the latter case, the vertical compo- functional block of the FN by injection of muscimol on the nent is controlled by the pathway through FFH and a part of head-free gaze shift in cats (Goffart and Pelisson, 1998; Gof- NRGc, while the horizontal component is regulated chiefly fart et al., 1998). Injection into the caudal portion of the FN by the pathway via the NRPc and partly through the NRGc. (cFN) resulted in hypermetria to the ipsilateral and hypome- This suggests that neurons in the frontal oculomotor region tria to the contralateral side of injection and slight slowing which are engaged in the control of orienting movements down of both eye and head movements to bilateral sides. with various horizontal and vertical components project to Relative contribution of eye and head to the amplitude of the these horizontal and/or vertical brainstem centers with vary- gaze shift remained constant in the ipsiversive movements, ing weight of projection. Studies of the projection pattern while a small increase in the head contribution was observed of single axons of neurons in the frontal oculomotor region in the contraversive gaze shift. On the other hand, when encoding particular vector of orienting movements may muscimol was injected to the rostral part of the FN (rFN), suggest such differential projections, but this issue remains typical gaze dysmetria was observed; ipsiversive movements as an open question at this moment. became hypermetric and contraversive movements hypo- metric. Relative contribution of head movements to the gaze 7.2. Cerebellum shift decreased in ipsiversive movements, while it increased for contraversive displacements. Furthermore, rFN block 7.2.1. Effects of functional inactivation also modified latency of the gaze shift; latency of ipsiversive The medial cerebellum (vermis) has been shown to be movements was shortened, while the latency of contraversive involved in control of movements of the body axis includ- movements was prolonged (Pelisson et al., 1998). These re- ing the head (Sprague and Chambers, 1954). The cerebellar sults indicated that the FN is critically involved in control of vermis, chiefly its medio-posterior portion receives inputs orienting gaze shift. Moreover, the results suggest the func- from a large number of brainstem structures related to the tional distinction between rFN and cFN as to the control of T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 233 orienting gaze shift. The cFN appears to be mainly related to showed that head turning induced by stimulation of the gain control of orienting movements, while the rFN seems to caudate nucleus remained after ablation of the pericruciate be involved in control of initiation and coordination of neck cortex, suggesting that there should be a more direct link muscle activity during orienting as well as the gain control. from the caudate nucleus to RSNs, bypassing the pericruciate cortex. One of the possible pathways is the caudate— 7.2.2. Efferent projection from the fastigial nucleus substantia nigra pars reticulata (SNr)—SC-reticulospinal The FN neurons project through two distinct efferent pathway. Joseph and Boussaoud (1985) showed the activity tracts, an uncrossed and a crossed (hook) bundle. They of the SNr related to orienting eye and head movements. project to the pontomedullary reticular formation as well However, it is also known that SNr neurons directly project as to the vestibular nuclei, the prepositus hypogolossi, the to the RSNs, bypassing the SC (Manetto and Lidsky, 1987; spinal cord, deep layers of the SC, and the thalamic nuclei Perciavalle, 1987; Schneider et al., 1985). The relative con- (Carpenter and Batton, 1982; Gruart and Delgado-Garc´ıa, tribution of the two pathways, one is mediated via SC and 1994; Homma et al., 1995). Projection to the brain stem the other bypassing the SC, is still unclear. of reticular formation is chiefly to the medullary reticular formation, mainly from the contralateral rFN via the hook 7.3.3. Role of basal ganglia in control of orienting bundle. Different effects caused by inactivation of the rFN triggered by volitional intention and cFN have been attributed to the different projection Orienting behaviors can be triggered not only by external from the rFN and cFN. stimulus but also by the internal drive. The orienting move- Mori and colleagues selectively stimulated the hook bun- ments caused by such “volitional intention” can be observed, dle to activate the fastigiofugal fibers and observed monosy- for instance, when the animal is searching for an unseen naptic excitation in a majority of RSNs in the NRGc (Mori but expected target and/or looks at the remembered location et al., 1998; see also Eccles et al., 1975; Ito et al., 1970). In of the target. As a model of these behaviors, the “mem- addition to the fastigio-reticular pathway, some of the fasti- ory guided saccades” are most often studied, using only giofugal neurons project further down to the spinal cord. non-human primates and humans (see reviews by Gaymard The fastigiospinal neurons mainly terminate in the inter- et al., 1998; Goldman-Rakic et al., 1990; Hikosaka et al., mediate zone in the upper cervical segments. Besides the 2000). Higher order structures such as the basal ganglia and fastigio-reticulo-spinal pathway, Wilson et al. (1978) found cerebral cortex are involved in control of memory guided that the rFN contains fastigiospinal neurons terminating at saccades. Several neurological diseases such as Parkinson’s the upper cervical cord and stimulation of this area induced disease (Hodgson et al., 1999; Blekher et al., 2000) and pa- monosynaptic EPSPs in some neck motoneurons, suggesting tients with specific brain injuries (for instance, in the frontal the monosynaptic connection of fastigiospinal neurons with eye field, parietal cortex and dorsolateral prefrontal cortex, neck motoneurons. How these projections are related to the the caudate nucleus) (Ploner et al., 1999) have been reported effects of functional block of the FN remains to be studied. regarding deficiency in memory-guided saccades.

7.3. Basal ganglia

7.3.1. Effects of electrical stimulation 8. Posture adjustment during orienting head Since the classical study of Ferrier (1873), a number of movements studies have shown that electrical stimulation of the striatum on one side induces contraversive movements of the head Orienting head movements always accompany the shift in and body (Cools, 1973; Forman and Ward, 1957; Laursen, the center of gravity that may cause disturbance in equilib- 1962; Pycock, 1980). More recently, Ohno and colleagues rium. Thus, postural adjustment prior and during orienting studied this issue more systematically in cats (Akaike et al., is essential for proper execution of orienting. Measurement 1989; Kitama et al., 1991; Ohno and Tsubokawa, 1987). of loads on forelimbs during orienting revealed synchronous The authors showed that stimulation of the caudate nucleus change in load on limbs with head movements (Yoshimura induced contraversive head turn combined with contraver- and Sasaki, unpublished observation) (Fig. 29). The load on sive saccadic eye movements. The EMG activity pattern of the contralateral forelimb to head turning increased, reach- the neck muscles during the stimulation-induced head turn ing peak after approximately 100 ms and then decreased was similar as voluntary orienting movements of the head. rapidly to its original level. This change in load may be a The effective site in the caudate nucleus was mainly in the feed-forward mechanism that can eventually compensate for caudal portion of the head of the caudate. the shift in the center of mass caused by the head movement. The load on the ipsilateral forelimb changes nearly like a 7.3.2. Descending projection from the basal ganglia mirror image of the contralateral forelimb. The load changes The effect of activities of the basal ganglia to the reticu- tended to be larger in cases where the cats protruded their lospinal pathway may be largely mediated via the thalamus heads in more forward direction and put more weight on and motor cortex. However, Ohno and Tsubokawa (1987) their forelimbs. 234 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Fig. 29. Posture adjustment during orienting toward a moving (ramp) Fig. 30. Eye, head and gaze movements during an orienting movement visual stimulus (upper lane) in the cat. The trajectory of eye and head in a monkey. movements are indicated (middle lane) and load changes on the contralateral and ipsilateral forelimbs. adjustment. The C-RSNs have been shown to terminate in the area where the long propriospinal neurons are located Repetitive electrical stimulation of the medial pon- (Sasaki, 1997; Matsushita et al., 1979). The excitatory con- tomedullary reticular formation (NRPc and NRGc) induced nection of the RSNs with long propriospinal neurons was ipsiversive orienting movements of the head and, in paral- shown by Alstermark et al. (1991). Thus, the long pro- lel, increase in the load on the contralateral forelimb and priospinal neurons are supposed to receive the same de- decrease in the load on the ipsilateral forelimb with laten- scending command as motoneurons from RSNs during ori- cies of less than 10 ms. This pattern of load change closely enting movements. resembled those observed during visually guided orienting movements as described in the previous sections. The intraaxonal staining of presumed C-RSNs at the level 9. Other animal species of brachial segments showed that they descended chiefly in the ventral funiculus and issued collaterals and terminals to 9.1. Primates laminae VII, VIII, and medial trunk motor nuclei, but only sparse branches were found in the motor nuclei of fore- The oculomotor range of both human and non-human pri- limb motoneurons in addition to neck motoneurons (Sasaki, mates is larger than 50◦. Accordingly, contribution of head 1997). Thus, the C-RSNs are suggested to be involved in movements is less crucial in primates for orienting smaller postural adjustment of forelimbs during orienting by send- than 50◦ (Guitton and Volle,1987; Fuller, 1992; Stahl, 1999). ing the same descending command to forelimb motoneurons It has been reported that orienting movements smaller than as those sent to neck motoneurons mainly via the interneu- 20◦ are performed chiefly with saccadic eye movements rons in the brachial segments to coordinate the movements and contribution of head movements needs to be considered of the forelimbs and head. mainly in the case of orienting with larger amplitude. Bizzi Moreover, it is well known that load change in hindlimbs et al. reported that eye movements preceded the head in gaze is coordinated with forelimbs during posture adjustment shifts smaller than 20◦, but the head preceded the eye move- (Coulmance et al., 1979). We also found that load in ments in larger gaze shifts (Bizzi et al., 1972b). Solitary eye hindlimbs changed in a similar manner as forelimbs during movements without associated head movements were often orienting, i.e. increase on the contralateral side and decrease observed, especially when the monkey was searching its en- on the ipsilateral side (Yoshimura and Sasaki, unpublished vironment. However, we could observe the coordinated eye observation). In preliminary experiments, we observed that and head movements even during orienting less than 20◦ the activity of L-RSNs, which project to the lumbar seg- with their heads unrestrained when they were trained to get ments, is modulated during orienting (Sasaki, unpublished a reward by retrieving a small piece of food placed in a observation). Thus, they likely control posture adjustment small hole underneath the target position (Fig. 30; Sasaki of hindlimbs during orienting. and Yoshimura, unpublished observation). Under such a con- In addition to these reticulospinal pathways, long pro- dition, head movements always preceded the saccadic eye priospinal neurons, which are located in the cervical seg- movements. Thus, coupling of eye and head movements is ments and project to lumbar segments, may be involved more flexible in primates than in cats. It may depend on how in coordination of forelimbs and hindlimbs during posture the animals are raised and trained during the experimental T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 235 sessions. Bizzi et al. (1972b) studied the activity of SPL mus- Cowie and Robinson (1994) systematically mapped the cles during horizontal orienting in monkeys and described effects of microstimulation in the medial pontomedullary differences between two modes of gaze movements. In a reticular formation in monkeys. They showed that stimula- “triggered” mode, in response to an unexpected visual tar- tion of the medullary reticular formation induced ipsiver- get, the agonist muscles show a phasic discharge, followed sive head movements without gaze shifts; the eyes moved by a tonic activity, and antagonist muscles are suppressed in toward the contralateral side by vestibuloocular reflex. Sys- the meantime. In contrast, in a “predictive” mode, in which tematic mapping revealed that both ipsiversive horizontal the animals are trained for a long period and orient the head movements and movements including vertical vector could to an expected target, there is a gradual increase in activity be induced from the medullary reticular formation, which of the agonist muscle and a progressive decrease of activity fits our observation of single unit recordings in the pon- in antagonist muscles. Thus, neck muscles are controlled in tomedullary reticular formation in cats (Section 5.1). a context dependent manner. 9.1.3. Involvement of the FEF in orienting head 9.1.1. Role of the SC in orienting head movements movements in primates in primates van der Steen et al. (1986) showed that lesion of the Contribution of the SC to orienting head movements has FEF in monkeys showed partial impairment of orienting eye been shown not only in cats but also in monkeys. Earlier and head movements. After recovery from the impairment studies by Stryker and Schiller (1975) and Robinson and in movement dynamics, relative neglect in the contralateral Jarvis (1974) showed that in primates the role of the SC hemifield remained when a pair of visual stimuli was simul- in orienting is essentially oculomotor and its contribution taneously presented in both hemifields. Thus, effects of le- to head movements is negligible. However, recent studies sion of the FEF in monkeys showed similar effects as lesion showed that stimulation of the SC induces head movements of the frontal oculomotor region in cats. A recent study by (Cowie and Robinson, 1994; Freedman et al., 1996; Klier Tyson and Keating (2000) showed that electrical stimulation et al., 2001). However, a relationship between the stimulation of the FEF induced combined eye and head movements in parameter and movements is different from that in the case monkeys. All the above results suggest that the FEF in mon- of eye movements. Induction of saccadic eye movements keys is involved in control of orienting head movements. by electrical stimulation of the SC has an all or none-like nature (Robinson, 1972; Schiller and Stryker, 1972); stim- 9.2. Rodents ulation above particular stimulus strength evokes saccades with particular vector and increase in stimulus strength does The effects of stimulation of the SC in rodents have been not much affect the vector of the evoked saccades. In con- extensively studied by Dean et al. (see reviews by Dean trast, the amplitude and velocity of head movements con- et al., 1988, 1989), who revealed that the SC stimulation tinues to increase in parallel to the increase in stimulus in- induces two different repertories of movements: (1) orient- tensities until they reach a particular value depending on ing responses toward the contralateral side of stimulation the stimulation site within the SC. Freedman and Sparks and (2) avoidance behaviors. Lesion studies revealed that (1997) further clarified that the motor-related activity in descending axons, which cross the midline at the predorsal the SC was associated with the amplitude and direction decussation, control the former types of movements, while of the gaze shift while it was only weakly correlated with the uncrossed ipsilateral descending pathway controls the each of eye or head movements. These results support the latter movements. The crossed descending pathway termi- “gaze displacement hypothesis”, which states that the SC nates in the medial pontomedullary reticular formation and encodes a single gaze displacement command, instead of controls orienting movements while some fibers descend “separate channel hypothesis”, which states that separate further down to the spinal cord. These pathways have been commands for eye and head movements are generated by described in cats and primates (cats: Nyberg-Hansen, 1964; the SC. Petras, 1967; Huerta and Harting, 1982; and primates: Cowie et al., 1994; Robinson et al., 1994), as well as rodents. 9.1.2. Descending pathway from the SC in primates However, the oculomotor range of rodents is smaller than In primates, by using injection of anterograde tracer tech- that of felines and primates. The maximal amplitude of nique, Cowie et al. (1994) showed that the deep layer of the saccade-like rapid eye movements induced by the SC stim- SC projects to the contralateral medial pontomedullary retic- ulation is around 10◦ in rats (McHaffie and Stein, 1982). ular formation and the medial reticular formation projects to The major effect of the SC stimulation is induction of head the ventral horn of the ipsilateral cervical spinal cord. Robin- and body movements. As to the relationship between the son et al. (1994) showed that the medial pontomedullary stimulation strength and response amplitude, King et al. reticular formation includes RSNs projecting to the cervi- (1991) claimed that the response amplitude increases lin- cal spinal cord. These results suggested that monkeys have early with the stimulus intensity and suggested that the ori- a similar organization of the tectoreticulospinal system as enting system in rats is different from the pulse-generating cats in relation to the control of the head movements. system like the saccade generator in primates. However, the 236 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 conclusion is based on observation of head movements horizontal, vertical and roll components of movements are showed that head-motor system is not a pulse generating encoded by anatomically distinct neural circuits” (Masino system in cats (Roucoux et al., 1980; Paré et al., 1994) or and Knudsen, 1992, 1993). Moreover, a forebrain region primates (Freedman et al., 1996). So far no experiment has that they refer to as the archistriatal gaze fields (AGFs) con- been performed to measure eye movements during unre- trol orienting head movements via the pathway through the strained orienting movements induced either by electrical optic tectum and also directly through the midline tegmen- or natural stimulus in rodents. tum bypassing the tectum (Knudsen et al., 1995). These results suggest that both the cat and barn owl have similar 9.3. Barn owl neuronal organization in the tecto-reticulo-spinal pathway in control of orienting head movements. The head movements induced by electrical stimulation of the SC have been well studied in the barn owl whose oculomotor range is small, and the head movements consist of a 10. Conclusion major portion of orienting gaze shift. The head movements induced by the electrical stimulation of the SC exhibit kinet- During the past 10 years, a large amount of information ics similar to saccadic eye movements in primates, and spa- has been accumulated as to the orienting gaze movements tial map and metrics of the SC-induced saccadic head move- in various species of animals, including rats, cats, owls, and ments have been studied by du Lac and Knudsen (1990). primates. Apparently, the contribution of eye and head move- Masino and Knudsen (1990) showed that in the downstream ments to the whole gaze shift varies among different ani- of the SC, the horizontal and vertical components of head mal species depending on the size of the oculomotor range. movements are controlled by different neural pathways (see The degree of coupling of the eye and head movements also Section 6). Further studies showed that “at the level of the varies among species, and such differentiation is likely to midbrain tegmentum there exists a three-dimensional Carte- occur in the brainstem neural circuits. The anatomical stud- sian representation of head-orienting movements such that ies suggested that similar regions in the brain stem reticu-

Fig. 31. A schematic diagram of summary of the neuronal circuitries controlling visually triggered orienting head movements described in this review. Abbreviations—IN: spinal interneurons. T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 237 lar formation are involved in the head movements in these Alstermark, B., Pinter, M.J., Sasaki, S., 1992a. Descending pathways animal species. However, the detailed structural and elec- mediating disynaptic excitation of dorsal neck motoneurones in the trophysiological analyses on the neural circuits have been cat: brain stem relay. Neurosci. Res. 15, 42–57. performed only in cats. Thus, it is necessary to obtain the Alstermark, B., Pinter, M.J., Sasaki, S., 1992b. Descending pathways mediating disynaptic excitation of dorsal neck motoneurones in the information about the basic neuronal circuits in other an- cat: facilitatory interactions. Neurosci. Res. 15, 32–41. imal species, but at the present stage, we should not un- Alstermark, B., Pinter, M.J., Sasaki, S., 1992c. Tectal and tegmental derestimate the information obtained from the experimental excitation in dorsal neck motoneurones of the cat. J. Physiol. (London) studies obtained in cats. Fig. 31 shows the schematic dia- 454, 517–532. gram of the neural circuitries controlling the visually guided Altman, J., Carpenter, M.B., 1961. Fiber projections of the superior colliculus in the cat. J. Comp. Neurol. 116, 157–177. orienting head movements described in this review. One of Anderson, M.E., Yoshida, M., Wilson, V.J., 1971. Influence of superior the major findings obtained in cats is that the horizontal colliculus on cat neck motoneurons. J. Neurophysiol. 34, 898–907. and vertical components of head movements are controlled Bahill, A.T., Clark, M.R., Stark, L., 1975. The main sequence a tool for by different brainstem nuclei. Such differential control of studying human eye movements. Math. Biosci. 24, 191–204. horizontal and vertical movements is also observed in the Behan, M., Appel, P.P., 1992. Intrinsic circuitry in the cat superior barn owl, and results of microstimulation of the brainstem colliculus: projections from the superficial layers. J. Comp. Neurol. 315, 230–243. strongly suggested that it is also the case in primates (Cowie Berthoz, A., 1989. Cooperation and substitution of the saccadic system and Robinson, 1994). Thus, differential control of horizon- and the reflexes of vestibular origin: should the “reflex” concept be tal and vertical head movements is a common feature over revised? Rev. Neurol. (Paris) 145, 513–526. species. Such differentiation fits the biomechanical architec- Bizzi, E., 1979. Strategies of eye–head coordination. Prog. Brain Res. 50, ture of the head–neck system (Vidal et al., 1986) and func- 795–803. Bizzi, E., Kalil, R.E., Tagliasco, V., 1971. Eye–head coordination in tional differentiation of the neck muscles (Richmond and monkeys: evidence for centrally patterned organization. Science 173, Vidal, 1988). It will be interesting to discuss the differential 452–454. control of horizontal and vertical movements in relation to Bizzi, E., Kalil, R.E., Morasso, P., Tagliasco, V., 1972a. Central the vestibular coordinate system. The vestibular system is programming and peripheral feedback during eye–head coordination already highly developed in lower vertebrate species such as in monkeys. Bibl. Opthalmol. 82, 220–232. fish. In contrast, the orienting system acquired its acuity and Bizzi, E., Kalil, R.E., Morasso, P., 1972b. Two modes of active eye–head coordination in monkeys. Brain Res. 40, 45–48. predominantly developed in higher vertebrates, which may Blakemore, C., Donaghy, M., 1980. Co-ordination of head and eyes in the suggest that it is “phylogenetically new”. Then it would not gaze changing behavior of cats. J. Physiol. (London) 300, 317–335. be a strange idea that the newly developed system adopted Blekher, T., Siemers, E., Abel, L.A., Yee, R.D., 2000. Eye movements in the coordinate system of the old. The strategy to parcel the Parkinson’s disease: before and after pallidotomy. Invest. Ophthalmol. movements into the horizontal and vertical functional syn- Vis. Sci. 41, 2177–2183. Büttner, U., Büttner-Ennever, J.A., Henn, V., 1977. Vertical eye movement ergies may help reduce the number of degrees of freedom related unit activity in the rostral mesencephalic reticular formation of of an otherwise too complex skeletomotor system such as the alert monkey. Brain Res. 130, 239–252. the neck and would contribute as a neural constraint to solve Carpenter, M.B., Batton, R.R., 1982. Connections of the fastigial nucleus the ill-posed problem impinged on the visuomotor transfor- in the cat and monkey. In: Palay, S.L., Chan-Palay, V. (Eds.), The mation process. Cerebellum—New Vistas. Springer, Berlin, pp. 250–295. In Section 2, we described different motor coordination Cools, A.R., 1973. Chemical and electrical stimulation of the caudate nucleus in freely moving cats: the role of dopamine. Brain Res. 58, during “velocity-guided” and “position-guided” orienting 437–451. movements. However, the neural substrate of such differ- Corneil, B.D., Olivier, E., Richmond, F.J., Loeb, G.E., Munoz, D.P., 2001. ence is still unclear. Since the fundamental neural circuits Neck muscles in the rhesus monkey. II. Electromyographic patterns involved in the control of orienting movements have been of activation underlying postures and movements. J. Neurophysiol. 86, well clarified as described in this review, the neural basis of 1729–1749. Coulmance, M., Gahery, Y., Massion, J., Swett, J.E., 1979. The placing such context-dependent modulation should be studied by us- reaction in the standing cat: a model for the study of posture and ing single unit recording from the related structures in alert movement. Exp. Brain Res. 37, 265–281. head-free animals. Coulter, J.D., Bowker, R.M., Wise, S.P., Murray, E.A., Castiglioni, A.J., Westlund, K.N., 1979. Cortical, tectal and medullary descending pathways to the cervical spinal cord. Prog. Brain Res. 50, 263–279. References Cowie, R.J., Robinson, D.L., 1994. Subcortical contributions to head movements in macaques. I. Contrasting effects of electrical stimulation Akaike, N., Ohno, T., Tsubokawa, H., 1989. EMG activities of neck of a medial pontomedullary region and the superior colliculus. J. muscles underlying lateral flexion of the neck during head-turning Neurophysiol. 72, 2648–2664. induced by electrical stimulation of the caudate nucleus in cats. Cowie, R.J., Smith, M.K., Robinson, D.L., 1994. Subcortical contributions Neurosci. Res. 6, 397–410. to head movements in macaques. II. Connections of a medial Alstermark, B., Pinter, M.J., Sasaki, S., 1985. Pyramidal effects in dorsal pontomedullary head-movement region. J. Neurophysiol. 72, 2665– neck motoneurones of the cat. J. Physiol. (London) 363, 287–302. 2682. Alstermark, B., Isa, T., Tantisira, B., 1991. Pyramidal excitation in long Dean, P., Redgrave, P., Mitchell, I.J., 1988. Organization of efferent propriospinal neurons in the cervical segments of the cat. Exp. Brain projections from superior colliculus to brainstem in rat: evidence for Res. 84, 569–582. functional output channels. Prog. Brain Res. 75, 27–36. 238 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Dean, P., Redgrave, P., Westby, G.W., 1989. Event or emergency? Grantyn, A., Berthoz, A., 1987. Reticulo-spinal neurons participating in Two response systems in the mammalian superior colliculus. Trends the control of synergic eye and head movements during orienting in Neurosci. 12, 137–147. the cat. I. Behavioral properties. Exp. Brain Res. 66, 339–354. Delgado-Garcia, J.M., Vidal, P.P., Gómez, C., Berthoz, A., 1988. Vertical Grantyn, A., Ong-Meang Jacques, V., Berthoz, A., 1987. Reticulo-spinal eye movements related signals in antidromically identified medullary neurons participating in the control of synergic eye and head reticular formation neurons in the alert cat. Exp. Brain Res. 70, 585– movements during orienting in the cat. II. Morphological properties 589. as revealed by intra-axonal injections of horseradish peroxidase. Dichgans, J., Bizzi, E., Morasso, P., Tagliasco, V., 1973. Mechanisms Exp. Brain Res. 66, 355–377. underlying recovery of eye–head coordination following bilateral Grantyn, A., Berthoz, A., Hardy, O., Gourdon, A., 1992. Contribution labyrinthectomy in monkeys. Exp. Brain Res. 18, 548–562. of reticulospinal neurons to the dynamic control of head movements: du Lac, S., Knudsen, E.I., 1990. Neural maps of head movement vector presumed neck bursters. In: Berthoz, A., Vidal, P.P., Graf, W. (Eds.), and speed in the optic tectum of the barn owl. J. Neurophysiol. 63, The Head–Neck Sensory Motor System. Oxford University Press, 131–146. Oxford, pp. 318–329. Eccles, J.C., Nicoll, R.A., Schwarz, W.F., Taborikova, H., Willey, T.J., Grantyn, A., Olivier, E., Kitama, T., 1993. Tracing premotor brain stem 1975. Reticulospinal neurons with and without monosynaptic inputs networks of orienting movements. Curr. Opin. Neurobiol. 3, 973–981. from cerebellar nuclei. J. Neurophysiol. 38, 513–530. Gruart, A., Delgado-Garc´ıa, J.M., 1994. Signalling properties of identified Edwards, S.B., 1980. The deep cell layers of the superior colliculus: deep cerebellar nuclear neurons related to eye and head movements in their reticular characteristics and structural organization. In: Hobson, the alert cat. J. Physiol. (London) 478, 37–54. J.A., Brazier, M.A.B. (Eds.), The Reticular Formation Revisited. Raven Guitton, D., 1992. Control of eye–head coordination during orienting gaze Press, New York, pp. 193–209. shifts. Trends Neurosci. 15, 174–179. Faulkner, R.F., Hyde, J.E., 1958. Coordinated eye and body movements Guitton, D., Mandl, G., 1978a. Frontal ‘oculomotor’ area in alert cat. I. evoked by brainstem stimulation in decerebrated cats. J. Neurophysiol. Eye movements and neck activity evoked by stimulation. Brain Res. 21, 171–182. 149, 295–312. Ferrier, D., 1873. experimental researches in cerebral physiology and Guitton, D., Mandl, G., 1978b. Frontal ‘oculomotor’ area in alert cat. pathology. West Riding Lunatic Asylum Med. Rep. 3, 30–96. II. Unit discharges associated with eye movements and neck muscle Forman, D., Ward, J.W., 1957. Responses to electrical stimulation of the activity. Brain Res. 149, 313–327. caudate nucleus in cats in chronic experiments. J. Neurophysiol. 20, Guitton, D., Douglas, R.M., Volle, M., 1984. Eye–head coordination in 230–243. cats. J. Neurophysiol. 52, 1030–1050. Guitton, D., Munoz, D.P., 1991. Control of orienting gaze shifts by Freedman, E.G., Sparks, D.L., 1997. Activity of cells in the deeper layers the tectoreticulospinal system in the head-free cat. I. Identification, of the superior colliculus of the rhesus monkey: evidence for a gaze localization, and effects of behavior on sensory responses. J. displacement command. J. Neurophysiol. 78, 1669–1690. Neurophysiol. 66, 1605–1623. Freedman, E.G., Stanford, T.R., Sparks, D.L., 1996. Combined eye–head Guitton, D., Volle, M., 1987. Gaze control in humans: eye–head gaze shifts produced by electrical stimulation of the superior colliculus coordination during orienting movements to targets within and beyond in rhesus monkeys. J. Neurophysiol. 76, 927–952. the oculomotor range. J. Neurophysiol. 58, 427–459. Fukushima, K., 1987. The interstitial nucleus of Cajal and its role in the Hall, W.C., Lee, P., 1993. Interlaminar connections of the superior control of movements of head and eyes. Prog. Neurobiol. 29, 107–192. colliculus in the tree shrew. I. The superficial gray layer. J. Fukushima, K., van der Hoeff-van Halen, R., Peterson, B.W., 1978. Direct Comp. Neurol. 332, 213–223. excitation of neck motoneurons by interstitiospinal fibers. Exp. Brain Harris, L.R., 1980. The suoperior colliculus and movements of the head Res. 33, 565–581. and eye in the cats. J. Physiol. (London) 300, 367–391. Fuller, J.H., 1992. Comparison of head movement strategies among Harting, J.K., Huerta, M.F., Frankfurter, A.J., Strominger, N.L., Royce, mammals. In: Berthoz, A., Vidal, P.P., Graf, W. (Eds.), The Head–Neck G.J., 1980. Ascending pathways from the monkey superior colliculus: Sensory Motor System. Oxford University Press, Oxford, pp. 101–112. an autoradiographic analysis. J. Comp. Neurol. 192, 853–882. Gaymard, B., Ploner, C.J., Rivaud, S., Vermersch, A.I., Pierrot-Deseilligny, Hassler, R., 1966. Exprapyramidal motor areas of cat’s frontal lobe: their C., 1998. Cortical control of saccades. Exp. Brain Res. 123, 159–163. function and architectonic differentiation. Int. J. Neurol. 5, 301–306. Goffart, L., Pelisson, D., 1998. Orienting gaze shifts during muscimol Hassler, R., 1972. Supranuclear structures regulating binocular eye and inactivation of caudal fastigial nucleus in the cat. I. Gaze dysmetria. head movements. Bibl. Ophthalmol. 82, 207–219. J. Neurophysiol. 79, 1942–1958. Hayes, N.L., Rustioni, A., 1981. Descending projections from brainstem Goffart, L., Pelisson, D., Guillaume, A., 1998. Orienting gaze shifts and sensorimotor cortex to spinal enlargements in the cat. Exp. Brain during muscimol inactivation of caudal fastigial nucleus in the cat. II. Res. 41, 89–107. Dynamics and eye–head coupling. J. Neurophysiol. 79, 1959–1976. He, X., Wu, C.P., 1985. Connections between pericruciate cortex and Goldman-Rakic, P.S., Funahashi, S., Bruce, C.J., 1990. Neocortical medullary reticulospinal neurons in cat: an electrophysiological study. memory circuits. Cold Spring Harbour Symp. Quant. Biol. 55, 1025– Exp. Brain Res. 61, 109–116. 1038. Hess, W.R., 1956. Hypothalamus and thalamus. Documentary Pictures. Graham, J., 1977. An autoradiographic study of the efferent connections Georg Thieme, Stuttgart. of the superior colliculus in the cat. J. Comp. Neurol. 173, 629–654. Hess, W.R., Burgi, S., Bucher, V., 1946. Motorische Funktion des Graf, W., de Waele, C., Vidal, P.P., 1995. Functional anatomy of the Tectal-und Tegmentalgebietes. Monatsschr. Psychiatr. Neurol. 112, 1– head–neck movement system of quadrupedal and bipedal mammals. J. 52. Anat. 186, 55–74. Hikosaka, O., Takikawa, Y., Kawagoe, R., 2000. Role of the basal ganglia Grantyn, A., 1989. How visual inputs to the ponto-bulbar reticular in the control of purposive saccadic eye movements. Physiol. Rev. 80, formation are used in the synthesis of premotor signals during orienting. 953–978. Prog. Brain Res. 80, 159–170. Hodgson, T.L., Dittrich, W.H., Henderson, L., Kennard, C., 1999. Eye Grantyn, A., Grantyn, R., 1982. Axonal patterns and sites of termination movements and spatial working memory in Parkinson’s disease. of cat superior colliculus neurons projecting in the tecto-bulbo spinal Neuropsychologia 37, 927–938. tract. Exp. Brain Res. 46, 243–256. Holstege, G., Cowie, R.J., 1989. Projections from the rostral Grantyn, A., Berthoz, A., 1985. Burst activity of identified mesencephalic reticular formation to the spinal cord: an HRP and tecto-reticulo-spinal neurons in the alert cat. Exp. Brain Res. 57, 417– autoradiographical tracing study in the cat. Exp. Brain Res. 75, 265– 421. 279. T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 239

Homma, Y., Nonaka, S., Matsuyama, K., Mori, S., 1995. Fastigiofugal King, S.M., Dean, P., Redgrave, P., 1991. Bypassing the saccadic pulse projection to the brainstem nuclei in the cat: an anterograde PHA-L generator: possible control of head movement trajectory by rat superior tracing study. Neurosci. Res. 23, 89–102. colliculus. Eur. J. Neurosci. 3, 790–801. Huerta, M.F., Harting, J.K., 1982. Tectal control of spinal cord activity: Kitama, T., Ohno, T., Tanaka, M., Tsubokawa, H., Yoshida, K., 1991. neuroanatomical demonstration of pathways connecting the superior Stimulation of the caudate nucleus induces contraversive saccadic eye colliculus with the cervical spinal cord grey. Prog. Brain Res. 57, movements as well as head turning in the cat. Neurosci. Res. 12, 293–328. 287–292. Isa, T., Endo, T., Saito, Y., 1998. The visuo-motor pathway in the, local Klier, E.M., Wang, H., Crawford, J.D., 2001. The superior colliculus circuit of the rat superior colliculus. J. Neurosci. 18, 8496–8504. encodes gaze commands in retinal coordinates. Nat. Neurosci. 4, 627– Isa, T., Itouji, T., 1992. Axonal trajectories of single Forel’s field H 632. neurones in the mesencephalon, pons and medulla oblongata in the Knudsen, E.I., Cohen, Y.E., Masino, T., 1995. Characterization of cat. Exp. Brain Res. 89, 484–495. a forebrain gaze field in the archistriatum of the barn owl: Isa, T., Itouji, T., Sasaki, S., 1988a. Excitatory pathways from Forel’s field microstimulation and anatomical connections. J. Neurosci. 15, 5139– H to head elevator motoneurones in the cat. Neurosci. Lett. 90, 89–94. 5151. Isa, T., Itouji, T., Nakao, S., Sasaki, S., 1988b. Subtypes of neurones in Kuypers, H.G., 1958. An anatomical analysis of cortico-bulbar connections Forel’s field H as defined by their axonal projection. Neurosci. Lett. to the pons and lower brain stem in the cat. J. Anat. 92, 198–218. 90, 95–99. Laursen, A.M., 1962. Movements evoked from the region of the caudate Isa, T., Itouji, T., Sasaki, S., 1992a. Control of head movements in nucleus in cats. Acta Physiol. Scand. 54, 175–184. the cat: two separate pathways from the superior colliculus to neck Lee, P., Hall, W.C., 1995. Interlaminar connections of the superior motoneurones and their roles in orienting movements. In: Shimazu, colliculus in the tree shrew. II. Projections from the superficial gray H., Shinoda, Y. (Eds.), Vestibular and Brain Stem Control of Eye, to the optic layer. Vis. Neurosci. 12, 573–588. Head and Body Movements. Japan Scientific Societies Press, Tokyo, Lee, P.H., Helms, M.C., Augustine, G.J., Hall, W.C., 1997. Role of pp. 275–284. intrinsic synaptic circuitry in collicular sensorimotor integration. Proc. Isa, T., Itouji, T., Sasaki, S., 1992b. Control of head movements via Forel’s Natl. Acad. Sci. U.S.A. 94, 13299–13304. field H. In: Berthoz, A., Vidal, P.P., Graf, W. (Eds.), The Head–Neck Maeda, M., Shibazaki, T., Yoshida, K., 1979. Labyrinthine and visual Sensory Motor System. Oxford University Press, Oxford, pp. 345–350. inputs to the superior colliculus neurons. Prog. Brain Res. 50, 735–743. Isa, T., Naito, K., 1994. Activity of neurons in Forel’s field H during Magni, F., Willis, W.D., 1964. Cortical control of brain stem reticular orienting head movements in alert head-free cats. Exp. Brain Res. 100, neurons. Arch. Ital. Biol. 102, 418–423. 187–199. Manetto, C., Lidsky, T.I., 1987. Influences of the basal ganglia on the Isa, T., Naito, K., 1995. Activity of neurons in the medial pontomedullary medullary reticular formation. Neurosci. Lett. 75, 278–282. reticular formation during orienting movements in alert head-free cats. Masino, T., Knudsen, E.I., 1990. Horizontal and vertical components of J. Neurophysiol. 74, 73–95. head movement are controlled by distinct neural circuits in the barn Isa, T., Saito, Y., 2001. The direct visuo-motor pathway in mammalian owl. Nature 345, 434–437. superior colliculus; novel perspective on the interlaminar connection. Masino, T., Knudsen, E.I., 1992. Anatomical pathways from the optic Neurosci. Res. 41, 107–113. tectum to the spinal cord subserving orienting movements in the barn Isa, T., Sasaki, S., 1988. Effects of lesion of paramedian pontomedullary owl. Exp. Brain Res. 92, 194–208. reticular formation by kainic acid injection on the visually triggered Masino, T., Knudsen, E.I., 1993. Orienting head movements resulting horizontal orienting movements in the cat. Neurosci. Lett. 87, 233–239. from electrical microstimulation of the brainstem tegmentum in the Isa, T., Sasaki, S., 1992a. Descending projections of Forel’s field H barn owl. J. Neurosci. 13, 351–370. neurones to the brain stem and the upper cervical spinal cord in the Matsushita, M., Ikeda, M., Hosoya, Y., 1979. The location of spinal cat. Exp. Brain Res. 88, 563–579. Isa, T., Sasaki, S., 1992b. Mono- and disynaptic pathways from Forel’s neurons with long descending axons (long descending propriospinal field H to dorsal neck motoneurones in the cat. Exp. Brain Res. 88, tract neurons) in the cat: a study with the horseradish peroxidase 580–593. technique. J. Comp. Neurol. 184, 63–80. Ito, M., Udo, M., Mano, M., Kawai, N., 1970. Synaptic action of Matsuyama, K., Drew, T., 1997. Organization of the projections from the fastigiobulbar impulses upon neurons in the medullary reticular the pericruciate cortex to the pontomedullary brainstem of the cat: a formation and vestibular nuclei. Exp. Brain Res. 11, 29–47. study using the anterograde tracer Phaseolus vulgaris-leucoagglutinin. Iwamoto, Y., Sasaki, S., 1990. Monosynaptic excitatory connections of J. Comp. Neurol. 389, 617–641. reticulospinal neurones in the nucleus reticularis pontis caudalis with McHaffie, J.G., Stein, B.E., 1982. Eye movements evoked by electrical dorsal neck motoneurones in the cat. Exp. Brain Res. 80, 277–289. stimulation in the superior colliculus of rats and hamsters. Brain Res. Iwamoto, Y., Sasaki, S., Suzuki, I., 1988. Descending cortical and tectal 247, 243–253. control of dorsal neck motoneurons via reticulospinal neurons in the Mitani, A., Ito, K., Mitani, Y., McCarley, R.W., 1988. Descending cat. Prog. Brain Res. 76, 97–108. projections from the gigantocellular tegmental field in the cat: cells of Iwamoto, Y., Sasaki, S., Suzuki, I., 1990. Input–output organization of origin and their brainstem and spinal cord trajectories. J. Comp. Neurol. reticulospinal neurones, with special reference to connections with 268, 546–566. dorsal neck motoneurones in the cat. Exp. Brain Res. 80, 260–276. Miyashita, E., Tamai, Y., 1989. Subcortical connections of frontal Joseph, J.P., Boussaoud, D., 1985. Role of the cat substantia nigra pars ‘oculomotor’ areas in the cat. Brain Res. 502, 75–87. reticulata in eye and head movements. I. Neural activity. Exp. Brain Mooney, R.D., Nikoletseas, M.M., Hess, P.R., Allen, Z., Lewin, A.C., Res. 57, 286–296. Rhoades, R.W., 1988. The projection from the superficial to the deep Kakei, S., Muto, N., Shinoda, Y., 1994. Innervation of multiple neck layers of the superior colliculus: an intracellular horseradish peroxidase motor nuclei by single reticulospinal tract axons receiving tectal input injection study in the hamster. J. Neurosci. 8, 1384–1399. in the upper cervical spinal cord. Neurosci. Lett. 172, 85–88. Mooney, R.D., Huang, X., Rhoades, R.W., 1992. Functional influence Kawamura, K., Brodal, A., Hoddevik, G., 1974. The projection of the of interlaminar connections in the hamster’s superior colliculus. J. superior colliculus onto the reticular formation of the brain stem: an Neurosci. 12, 2417–2432. experimental anatomical study in the cat. Exp. Brain Res. 19, 1–19. Mori, S., Matsui, T., Kuze, B., Asanome, M., Nakajima, K., Matsuyama, Keizer, K., Kuypers, H.G., 1984. Distribution of corticospinal neurons with K., 1998. Cerebellar-induced locomotion: reticulospinal control of collaterals to lower brain stem reticular formation in cat. Exp. Brain spinal rhythm generating mechanism in cats. Ann. NY Acad. Sci. 860, Res. 54, 107–120. 94–105. 240 T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241

Moschovakis, A.K., Scudder, C.A., Highstein, S.M., 1996. The Perciavalle, V., 1987. Substantia nigra influences on the reticulospinal microscopic anatomy and physiology of the mammalian saccadic neurons: an electrophysiological and ionophoretic study in cats and system. Prog. Neurobiol. 50, 133–254. rats. Neuroscience 23, 243–251. Munoz, D.P., Guitton, D., 1985. Tectospinal neurons in the cat have Peterson, B.W., Anderson, M.E., Filion, M., 1974. Responses of discharges coding gaze position error. Brain Res. 341, 184–188. ponto-medullary reticular neurons to cortical, tectal and cutaneous Munoz, D.P., Guitton, D., 1986. Presaccadic burst discharges of stimuli. Exp. Brain Res. 21, 19–44. tecto-reticulo-spinal neurons in the alert head-free and -fixed cat. Brain Peterson, B.W., Maunz, R.A., Pitts, N.G., Mackel, R.G., 1975. Patterns Res. 398, 185–190. of projection and branching of reticulospinal neurons. Exp. Brain Res. Munoz, D.P., Guitton, D., 1989. Fixation and orientation control by the 23, 333–351. tecto-reticulo-spinal system in the cat whose head is unrestrained. Rev. Peterson, B.W., Pitts, N.G., Fukushima, K., Mackel, R., 1978. Neurol. 145, 567–579. Reticulospinal excitation and inhibition of neck motoneurons. Munoz, D.P., Guitton, D., 1991. Control of orienting gaze shifts by the Exp. Brain Res. 32, 471–489. tectoreticulospinal system in the head-free cat. II. Sustained discharges Petras, J.M., 1967. Cortical, tectal and tegmental fiber connections in the during motor preparation and fixation. J. Neurophysiol. 66, 1624–1641. spinal cord of the cat. Brain Res. 6, 275–324. Munoz, D.P., Guitton, D., Pelisson, D., 1991. Control of orienting Pilyavsky, A.I., Goskin, A.P., 1978. Investigation of the cortico- gaze shifts by the tectoreticulospinal system in the head-free cat. reticulospinal connections in cats. Neuroscience 3, 99–103. III. Spatiotemporal characteristics of phasic motor discharges. J. Ploner, C.J., Rivaud-Pechoux, S., Gaymard, B.M., Agid, Y., Neurophysiol. 66, 1642–1666. Pierrot-Deseilligny, C., 1999. Errors of memory-guided saccades in Muto, N., Kakei, S., Shinoda, Y., 1996. Morphology of single axons of humans with lesions of the frontal eye field and the dorsolateral tectospinal neurons in the upper cervical spinal cord. J. Comp. Neurol. prefrontal cortex. J. Neurophysiol. 82, 1086–1090. 372, 9–26. Pycock, C.J., 1980. Turning behavior in animals. Neuroscience 5, 461– Nakao, S., Shiraishi, Y., 1985. Direct excitatory and inhibitory synaptic 514. inputs from the medial mesodiencephalic junction to motoneurons Rhoades, R.W., Mooney, R.D., Rohrer, W.H., Niloletseas, M.M., Fish, innervating extraocular oblique muscles in the cat. Exp. Brain Res. S.E., 1989. Organization of the projection from the superficial to the 61, 62–72. deep layers of the hamster’s superior colliculus as demonstrated by the Nakao, S., Shiraishi, Y., Oda, H., Inagaki, M., 1988. Direct inhibitory anterograde transport of Phaseolus vulgaris-leucoagglutinin. J. Comp. projection of pontine omnipause neurons to burst neurons in the Forel’s Neurol. 283, 54–70. field H controlling vertical eye movement-related motoneurons in the Richmond, F.J.R., Vidal, P.P., 1988. The motor system: joints and muscles cat. Exp. Brain Res. 70, 632–636. of the neck. In: Peterson, B.W., Richmond, F.J.R. (Eds.), Control of Nakao, S., Shiraishi, Y., Li, W.B., Oikawa, T., 1990. Mono- and disynaptic Head Movement. Oxford University Press, New York, pp. 1–21. excitatory inputs from the superior colliculus to vertical saccade-related Richmond, F.J.R., Thomson, D.B., Loeb, G.E., 1992. Electromyographic neurons in the cat Forel’s field H. Exp. Brain Res. 82, 222–226. studies of neck muscles in the intact cat. I. Patterns of recruitment Napier, J.R., Napier, P.H., 1985. The Natural History of Primates. The underlying posture and movement during natural behaviors. Exp. Brain MIT Press, Cambridge, MA. Res. 88, 41–58. Nieoullon, A., Rispal-Padel, L., 1976. Somatotopic localization in cat Richmond, F.J., Corneil, B.O., Singh, K., 1999. Animal models of motor motor cortex. Brain Res. 105, 405–422. systems: cautionary tales from studies of head movement. Prog. Brain Nyberg-Hansen, R., 1964. The location and termination of tectospinal Res. 123, 411–416. fibers in the cat. Exp. Neurol. 9, 212–227. Richmond, F.J., Singh, K., Corneil, B.D., 2001. Neck muscles in the rhesus Nyberg-Hansen, R., 1965. Sites and mode of termination of reticulo-spinal monkey. I. Muscle morphometry and histochemistry. J. Neurophysiol. fibers in the cat: an experimental study with silver impregnation 86, 1717–1728. methods. J. Comp. Neurol. 124, 71–100. Ohno, T., Tsubokawa, H., 1987. Regional differences in the cat caudate Robinson, D.A., 1972. Eye movements evoked by collicular stimulation nucleus as to the effectiveness in inducing contraversive head-turning in the alert monkey. Vision Res. 12, 1795–1808. by electrical stimulation. Neurosci. Res. 4, 497–516. Robinson, D.L., Jarvis, C.D., 1974. Superior colliculus neurons studied Olivier, E., Grantyn, A., Kitama, T., Berthoz, A., 1995. Post-spike during head and eye movements of the behaving monkey. J. facilitation of neck EMG by cat tectoreticulospinal neurones during Neurophysiol. 37, 533–540. orienting movements. J. Physiol. (London) 482, 455–466. Robinson, F.R., Phillips, J.O., Fuchs, A.F., 1994. Coordination of gaze Paré, M., Guitton, D., 1990. Gaze-related activity of brainstem omnipause shifts in primates: brainstem inputs to neck and extraocular motoneuron neurons during combined eye–head gaze shifts in the alert cat. pools. J. Comp. Neurol. 346, 43–62. Exp. Brain Res. 83, 210–214. Roucoux, A., Guitton, D., Crommelinck, M., 1980. Stimulation of the Paré, M., Crommelinck, M., Guitton, D., 1994. Gaze shifts evoked by superior colliculus in the alert cat. II. Eye and head movements evoked stimulation of the superior colliculus in the head-free cat conform when the head is unrestrained. Exp. Brain Res. 39, 75–85. to the motor map but also depend on stimulus strength and fixation Roucoux, A., Crommelinck, M., Decostre, M.-F., 1989. Neck muscle activity. Exp. Brain Res. 101, 123–139. activity in eye–head coordinated movements. Prog. Brain Res. 80, Paré, M., Guitton, D., 1994. The fixation area of the cat superior colliculus: 351–362. effects of electrical stimulation and direct connection with brainstem Sasaki, S., 1992. Reticulospinal control of head movements in the cat. omnipause neurons. Exp. Brain Res. 101, 109–122. In: Berthoz, A., Vidal, P.P., Graf, W. (Eds.), The Head–neck Sensory Paré, M., Guitton, D., 1998. Brain stem omnipause neurons and the control Motor System. Oxford University Press, Oxford, pp. 311–317. of combined eye–head gaze saccades in the alert cat. J. Neurophysiol. Sasaki, S., 1997. Axonal branching and termination of cervical 79, 3060–3076. reticulospinal neurons in the cat brachial segments. Neurosci. Lett. Pasik, P., Pasik, T., Bender, M.B., 1969. The pretectal syndrome in 228, 83–86. monkeys. I. Disturbances of gaze and body posture. Brain 92, 521–534. Sasaki, S., 1999. Direct connection of the nucleus reticularis Peck, C.K., 1990. Neuronal activity related to head and eye movements gigantocellularis neurons with neck motoneurons in cats. Exp. Brain in cat superior colliculus. J. Physiol. (London) 421, 79–104. Res. 128, 527–530. Pelisson, D., Goffart, L., Guillaume, A., 1998. Contribution of the Sasaki, S., Isa, T., Naito, K., 1999. Effects of lesion of pontomedullary rostral fastigial nucleus to the control of orienting gaze shifts in the reticular formation on visually triggered vertical and oblique head head-unrestrained cat. J. Neurophysiol. 80, 1180–1196. orienting movements in alert cats. Neurosci. Lett. 265, 13–16. T. Isa, S. Sasaki / Progress in Neurobiology 66 (2002) 205–241 241

Sasaki, S., Iwamoto, Y., 1999. Axonal trajectories of the nucleus reticularis Suzuki, S.S., Siegel, J.M., Wu, M.F., 1989. Role of pontomedullary gigantocellularis neurons in the C2–C3 segments in cats. Neurosci. reticular formation neurons in horizontal head movements: an ibotenic Lett. 264, 137–140. acid lesion study in the cat. Brain Res. 484, 78–93. Sasaki, S., Naito, K., Oka, M., 1996. Firing characteristics of neurones Tamai, Y., Fujii, T., Nakai, M., Komai, N., Tsujimoto, T., 1983. Eye in the superior colliculus and the pontomedullary reticular formation movements evoked by electrical stimulation of the frontal cortex in during orienting in unrestrained cats. Prog. Brain Res. 112, 99–116. the coronary sulcus of the cat. Jpn. J. Physiol. 33, 305–308. Schiller, P.H., Stryker, M., 1972. Single unit recording and stimulation Tamai, Y., Nakai, M., Miyashita, E., 1987. Types of monocular eye in superior collisulus of the alert rhesus monkey. J. Neurophysiol. 35, movements studied by intracortical microstimulation in the cat’s 915–924. coronal sulcus. Jpn. J. Physiol. 37, 741–747. Schlag, J., Schlag-Rey, M., 1970. Induction of oculomotor responses by Thomson, D.B., Loeb, G.E., Richmond, F.J., 1994. Effect of neck posture electrical stimulation of the prefrontal cortex in the cat. Brain Res. 22, on the activation of feline neck muscles during voluntary head turns. 1–13. J. Neurophysiol. 72, 2004–2014. Schneider, J.S., Manetto, C., Lidsky, T.I., 1985. Substantia nigra projection Thomson, D.B., Loeb, G.E., Richmond, F.J., 1996. Effect of neck posture to medullary reticular formation: relevance to oculomotor and related on patterns of activation of feline neck muscles during horizontal motor functions in the cat. Neurosci. Lett. 62, 1–6. rotation. Exp. Brain Res. 110, 392–400. Shinoda, Y., Kakei, S., Muto, N., 1996. Morphology of single axons of Tohyama, M., Sakai, K., Salvert, D., Touret, M., Jouvet, M., 1979. Spinal tectospinal and reticulospinal neurons in the upper cervical spinal cord. projections from the lower brain stem in the cat as visualized by the Prog. Brain Res. 112, 71–84. horseradish peroxidase technique. I. Origin of reticulospinal tracts and Siegel, J.M., Tomaszewski, K.S., 1983. Behavioral organization of reticular their funicular trajectories. Brain Res. 173, 383–403. formation: studies in the unrestrained cat. I. Cells related to axial, Tyson, A.T., Keating, E.G., 2000. Electrical stimulation of the frontal eye limb, eye, and other movements. J. Neurophysiol. 50, 696–716. ﬁeld in a monkey produces combined eye and head movements. J. Sparks, D.L., 1986. Translation of sensory signals into commands for Neurophysiol. 84, 1103–1106. control of saccadic eye movements: role of primate superior colliculus. van der Steen, J., Russell, I.S., James, G.O., 1986. Effects of unilateral Physiol. Rev. 66, 118–171. frontal eye-ﬁeld lesions on eye–head coordination in monkey. J. Sparks, D.L., 1999. Conceptual issues related to the role of the superior Neurophysiol. 55, 696–714. colliculus in the control of gaze. Curr. Opin. Neurobiol. 9, 698–707. Vanegas H., 1984. In: Vanegas, H. (Ed.), Comparative Neurology of Optic Sprague, J.M., Chambers, W.W., 1954. Control of posture by reticular Tectum. Plenum Press, New York. formation and cerebellum in the intact, anesthetized and unanesthetized Vidal, P.P., Roucoux, A., Berthoz, A., 1982. Horizontal eye position-related and in the decerebrated cat, anesthetized and unanesthetized and in activity in neck muscles of the alert cat. Exp. Brain Res. 46, the decerebrated cat. Am. J. Physiol. 176, 52–64. 448-453. Sprague, J.M., Meikle, T.H., 1965. The role of the superior colliculus in Vidal, P.P., Corvisier, J., Berthoz, A., 1983. Eye and neck motor signals visually guided behavior. Exp. Neurol. 11, 114–146. in periabducens reticular neurons of the alert cat. Exp. Brain Res. 53, Sprague, J.M., Levy, J., Di Berardino, A., Berlucchi, G., 1977. 16–28. Visual cortical areas mediating form discrimination in the cat. J. Vidal, P.P., Graf, W., Berthoz, A., 1986. The orientation of the cervical Comp. Neurol. 172, 441–488. vertebral column in unrestrained awake animals. I. Resting position. Stahl, J.S., 1999. Amplitude of human head movements associated with Exp. Brain Res. 61, 549–559. horizontal saccades. Exp. Brain Res. 126, 41–54. Vidal, P.P., De Waele, C., Graf, W., Berthoz, A., 1988. Skeletal geometry Straschill, M., Schick, F., 1977. Discharges of superior colliculus neurons underlying head movements. Ann. NY Acad. Sci. 545, 228–238. during head and eye movements of the alert cat. Exp. Brain Res. 27, Wilson, V.J., Uchino, Y., Maunz, R.A., Susswein, A., Fukushima, K., 1978. 131–141. Properties and connections of cat fastigiospinal neurons. Exp. Brain Stryker, M.P., Schiller, P.H., 1975. Eye and head movements evoked by Res. 32, 1–17. electrical stimulation of monkey superior colliculus. Exp. Brain Res. Wurtz R.H., Albano, 1980. Visual-motor function of the primate superior 23, 103–112. colliculus. Ann Rev. Neurosci. 3, 189–226.