ON CLIMBING FIBER SIGNALS AND THEIR CONSEQUENCE(S)

Keywords

Abstract

1. Introduction

Despite nearly three decades of investigation the issue of climbing fiber contributions to cerebellar transactions is still unresolved. While opposing and controversial points of view exist about the roles of climbing fibers, some aspects of climbing fiber anatomy are, thankfully, agreed upon. It is accepted that all climbing fibers originate from the contralateral inferior olive (Desclin, 1974; Szentgothai & Rajkovits, 1959; see also Dow, 1942) and that the inferior olive is composed of subdivisions whose climbing fibers terminate in the cerebellar cortex in sagittally oriented zones (Groenewegen & Voogd, 1977; Jansen & Brodal, 1954; Voogd, 1964; Groenewegen et al., 1979) (Figure 1). It is further agreed that in the adult each Purkinje cell receives only one climbing fiber (Eccles et al., 1966) and that, in general, Purkinje cells within a given zone project to only one cerebellar nucleus, which receives collaterals from the inferior olivary axons that terminate as climbing fibers in that zone (Eccles, et al., 1967; De Zeeuw et al., 1994b, Wylie et al., 1994; Van der Want et al., 1989). Furthermore, the cerebellar nuclei contain GABAergic neurons that project contralaterally to those parts of the inferior olive (De Zeeuw et al., 1989; Nelson & Mugnaini, 1989) that provide the collaterals to that particular cerebellar nucleus (Dom et al., 1973; Graybiel et al., 1973). The anatomical unit composed of a particular sagittal zone with its climbing fibers and their collaterals, along with the corresponding Purkinje cells and the associated cerebellar nucleus has been called a module (Voogd & Bigar, 1980), which is considered to be the functional unit of the cerebellum (Andersson & Oscarsson, 1978; Oscarsson, 1969, 1979). Thus, it is agreed that the climbing fibers are central to parcelling the cerebellum into anatomical modules. In some parts of the cerebellar cortex, each anatomical zone is mirrored by the uniformity of its climbing fiber responses to natural stimulation (e.g., De Zeeuw et al., 1994b; Wylie et al., 1994), but in other parts, the zones break up into patches of climbing fiber responses with natural stimulation (e,g., Robertson, 1984; Robertson et al., 1982).

The olivo-cerebellar system has several remarkable physiological characteristics that make it something of a curiosity in the sensorimotor operations of the brain. One conspicuous feature is the extremely powerful excitatory action of the climbing fiber on the Purkinje cell with the resulting burst discharge of the Purkinje cell (its complex spike, Eccles et al., 1966; Thach, 1967; Ito & Simpson, 1971; Mano et al., 1989). Another unusual feature is the extremely low firing frequency of olivary neurons. In both anesthetized and awake animals, olivary neurons discharge either a single spike or a burst of spikes (2-5 spikes with an interspike interval of 2-3 milliseconds) about once or twice per second (Armstrong, 1974; Armstrong & Rawson, 1979; Crill, 1970). Even with application of chemical "excitants", such as harmaline, only 8-10 discharges per second are observed (de Montigny & Lamarre, 1973; Lamarre et al., 1971; Llins & Volkind, 1973). The limited frequency range of olivary discharge stands in contrast to the range of several hundred spikes per second for the Purkinje cell simple spikes (Thach, 1967) whose excitatory drive is conveyed by the mossy fiber-parallel fiber input (Eccles et al., 1967).

Despite the well-documented anatomy of a seemingly simple, repetitive circuit, there is little agreement about the meaning or consequences of climbing fiber activity. In trying to synthesize the various hypotheses on the function of the climbing fibers, one has the sense of looking at a drawing by Escher. Each point of view seems to account for a certain collection of findings, but when one attempts to put the different views together, a coherent picture of what the climbing fibers are doing does not appear. For the majority of researchers, the climbing fibers signal errors in motor performance, either in the usual manner of discharge frequency modulation or as a single announcement of an "unexpected event." For other investigators, the message lies in the degree of ensemble synchrony and rhythmicity among a population of climbing fibers. Beyond the matter of what messages are being carried by the climbing fiber signals (Section 2) is the more controversial issue of their consequence(s) (Sections 3 and 4). The central questions are whether climbing fiber activity a) leads to long-term depression of the strength of parallel fiber synapses; b) results in short-lasting enhancement of Purkinje cell responsiveness to mossy fiber inputs; or c) constitutes a largely independent channel with little influence on the simple spikes. In the following pages these views of the possible roles of the climbing fibers are presented in the spirit of trying to reach some consensus about their function. Each point of view is introduced by summarizing the basic observations made by the respective proponents. Subsequently, the issues of short-lasting enhancement, and synchrony and rhythmicity are addressed in Section 5 in the context of the vestibulocerebellar visual climbing fiber system, whose modulation by retinal image motion provides a palpable example of climbing fibers signaling an error in performance.

2. Climbing Fiber Messages

2.1 Climbing fiber modulation

The initial attempts to find a message in the climbing fiber activity involved the spino-olivary pathways to the anterior lobe of the vermis (Oscarsson, 1969, 1973; Thach, 1967). With natural stimulation the responses of the spino-olivary systems to this part of the cerebellum were weak or absent and the receptive fields were large and vaguely delimited. These findings were perplexing because they appeared at variance with the high degree of specificity typifying the connectivity from the inferior olive to the cerebellum. Presumably Oscarsson's initial findings with natural stimulation reflected the extensive partial spinal cord sectioning done to study separately the several spino-olivary pathways. In contrast to these early observations, subsequent investigators have been able to activate climbing fiber inputs to the vermis and pars intermedia of the anterior lobe with the use of natural stimulation, particularly cutaneous stimulation (Armstrong & Rawson, 1979; Armstrong & Edgley, 1984; Eccles et al. 1972; Gellman et al. 1983; 1985; Ishikawa et al. 1972; Leicht et al. 1977; Robertson et al. 1982; Rushmer et al., 1976). Also in contrast to the earlier findings on the inability of climbing fibers to signal peripheral events transmitted through the spinal cord, Rubia & Kolb (1978; Kolb & Rubia, 1980) found that the kinematics of passive movement of the forepaw were, in fact, reflected in the modulation of the climbing fiber responses of Purkinje cells recorded in the anterior lobe. To reveal the relation between complex spike activity and the amplitude, velocity, and acceleration of the movement a number of trials had to be summed, but such an averaging may occur during a single movement when ensembles of Purkinje cells are considered. The potential capability of ensembles of Purkinje cells to signal kinematic variables through their conjoint climbing fiber responses has been described theoretically (McCollum, 1992; Robertson & McCollum, 1989), by applying set theory to data obtained by Robertson and colleagues (Robertson, 1984; Robertson & Laxer, 1981; Robertson et al., 1982) on the somatotopy of cutaneously evoked complex spike responses in the anterior lobe of the decerebrate cat.

Because of their low discharge frequency, climbing fibers were for some time held to be unable to encode messages in the customary way, that is, with frequency coding. While it was realized that complex spikes did respond to cutaneous and muscle stimulation, the constraint on firing frequency suggested to some that climbing fibers are "phasic" (Llins, 1970, 1974) and that their responses are encoding an event rather than an on-going process (Armstrong, 1974; Rushmer et al., 1976). However, the capability of climbing fiber modulation to encode signals in the conventional manner of frequency modulation is clearly apparent in the responses of visually activated climbing fibers in the flocculonodular lobe (Figure 2) (Fushiki et al., 1994; Wylie & Frost, 1993; De Zeeuw et al., 1994b; Graf et al., 1988; Stone and Lisberger, 1990; Kano et al., 1990a,b; Kusunoki et al., 1990; Maekawa & Simpson, 1973; Simpson & Alley, 1974). These "visual" climbing fibers originate in the dorsal cap of Kooy and the ventolateral outgrowth, and signal the direction and speed of movement of large parts of the visual world across the retina. In the rabbit, they are optimally responsive to low speeds (about 1/s) (Alley et al., 1975; Barmack & Hess, 1980; Simpson & Alley, 1974). The signals of retinal image motion provide a measure of the "short comings" of the compensatory eye movement system in stabilizing the retinal image. For the low speeds of retinal image motion to which the climbing fibers are best responsive, their low firing rate does not prohibit signaling of speed and direction in the conventional manner of frequency encoding, and the modulation has both phasic and tonic components. In contrast to the complex spike modulation produced by passive paw movement (Rubia & Kolb, 1978; Kolb & Rubia, 1980), the floccular complex spike modulation produced by visual stimulation can be seen without resorting to averaging (Simpson & Alley, 1974). Similarly, Barmack et al., (1989, 1993a) found that neurons of the beta nucleus of the inferior olive, which projects to parts of the nodulus and uvula, modulate robustly in response to natural stimulation of the otoliths and vertical semicircular canals (Figure 3). For some beta nucleus neurons the response is maintained for maintained tilt, thus providing another example of "tonic" signaling by the climbing fiber system.

2.2 Climbing fibers as "comparators"

Oscarsson's (1969, 1973) original findings that the climbing fibres transmitting input from the spinal cord could not be well-activated by conventional sensory stimuli led him to hypothesize that the inferior olive is less related to reporting directly about the periphery than to reporting about interneuronal activity in the spinal cord. He proposed that the inferior olive was a "comparator" of command signals from higher centers with the activities these signals evoked at lower levels in the spinal cord, which were also influenced from the periphery (Oscarsson, 1980). In this hypothesis, climbing fibers signal modifications at lower motor levels in the spinal cord of command signals issued from higher levels. As we shall see below, this hypothesis evolved into a more explicitly stated hypothesis of error detection obtained by comparing intended with achieved movement. If the comparison occurs within the inferior olive, it is unlikely that the signals of intention and achievement are conveyed only through, respectively, the descending and ascending projections to the inferior olive because these inputs generally do not converge on the same olivary neurons (De Zeeuw, 1990). From the anatomical point of view, a comparison is much more likely to happen between the ascending and descending inputs, which are all excitatory, on the one hand and the inhibitory projections derived from the hindbrain on the other hand. Each dendritic spine of an olivary neuron receives both an inhibitory input from one of the hindbrain regions, which include the cerebellar nuclei, vestibular nuclei, nucleus prepositus hypoglossi, solitary nucleus, and dorsal column nuclei, and an excitatory input from the spinal cord, brainstem, mesodiencephalic junction or cerebral cortex (De Zeeuw et al., 1990a,b,c).

2.3 Climbing fibers as "event detectors"

As indicated above, subsequent studies showed that the climbing fibers are quite sensitive to very small mechanical perturbations (taps on the skin or the footpads), but the message that was signaled was not readily apparent. For example, touch of a footpad would yield a single climbing fiber discharge, but maintained pressure did not result in a maintained response, presumably because of the performance of the cutaneous receptors themselves. The exquisite sensitivity of the climbing fibers to pressure on the footpad led Rushmer et al. (1976; see also Armstrong, 1974) to conclude that they would provide no information during locomotion other than that the footpad had contacted an object, because the forces and displacements encountered during locomotion would insure that the probability of firing during each touchdown or lift-off would be virtually one. Therefore, they would not be able to discriminate variations in the normal range of forces and displacements. As a consequence of the response characteristics of this climbing fiber system to passive touch of the paw, Rushmer et al. proposed that it serves as an "event detector" to signal simply that the foot has made or lost contact with a weight-bearing surface. The questions then arose as to 1) whether such a stereotyped response signaled only the time of occurrence of footfall or lift-off and no other aspects of performance, and 2) whether such sensory responses occurred during voluntary movements or only during passive stimulation.

2.4 Climbing fibers as "unexpected event detectors"

The notion of the climbing fiber as an "event detector" was unsatisfying to many investigators. For example, as stated above, Kolb & Rubia (see also Ojakangas & Ebner, 1992, 1994) found that the climbing fiber system can represent movement kinematics, which is more than just the occurrence of a peripheral event. Moreover, a number of investigators showed that when animals are allowed to make voluntary movements at their own pace the complex spike activity does not occur in close correlation with specific aspects of the movement (Andersson & Armstrong, 1985, 1987; Armstrong et al., 1982; Armstrong et al., 1988; Boylls, 1980; Gellman et al. 1985; Mano et al., 1986, 1989; Thach 1968, 1970). For example, Boylls (1980) noted that in the decerebrate, walking cat some olivary cells tended to discharge at specific times during the step cycle, yet there was an absence of a strong time-locked relation between any aspect of the locomotor cycle and the firing of the olivary neurons. Such findings required modification and qualification of the "event detector" proposal. The changes in view harkened back to the proposal of Oscarsson that the inferior olive acts as a comparator of intended with achieved movement, and what was an "event detector" became an "unexpected event detector". To illustrate, Gellman et al. (1985) found in the awake cat that complex spike responses to a passively applied stimulus generally failed to occur when a similar stimulus was produced by a voluntary movement, unless the receptive field of the complex spike was "unexpectedly" brought into contact with an object during active movement. Similarly, Andersson & Armstrong (1985, 1987) showed that when an awake cat walking along a circular horizontal ladder encountered a rung that unexpectedly gave way, an increase in complex spike discharge occurred prior to the unlatched rung hitting its end stop. These two sets of observations are in line with a proposal by Oscarsson (1980) that built upon his earlier comparator hypothesis. The more explicit hypothesis is that the olive would detect perturbations of two types: perturbations in the commands introduced into lower spinal cord motor centers as a consequence of reflex activity, and perturbations in the evolving movement due to unexpected changes in load or resistance. Andersson and Armstrong recast this proposal specifically to apply to the cat walking on the horizontal ladder by pointing out that when the cat stepped on the unlatched rung a mismatch occurred between the intended and achieved trajectory of the stepping limb. The many other reports of a relation between complex spike activity and movement perturbations produced by changes in load or contact are also under the umbrella of the comparator hypothesis of error detection (e.g., Gilbert and Thach, 1977; Wang et al., 1987; Kim et al., 1987; 1988; Lou and Bloedel, 1992a, b; Ojakangas and Ebner, 1992; 1994).

In the decerebrate cat performing treadmill locomotion, Kim et al. (1987) found that about one-half of the Purkine cells showed complex spike modulation in relation to the unperturbed step cycle. The presence and absence of complex spike modulation during unperturbed locomotion in the decerebrate and intact cat, respectively, can both be accommodated by the comparator hypothesis because the decerebration altered the usual complement of signals to be compared. The difference between the findings of Boylls (see above) and Kim et al. may have its basis in the difference in the population of recorded neurons; Boylls recorded in the caudal medial accessory olive, which projects to the medial A-zone (Voogd and Bigar, 1980), whereas Kim et al. recorded complex spikes in the intermediate zones (C1 and C2).

2.5 Climbing fibers as "error detectors"

The word "unexpected" is unfelicitous because we can only presume what the animal expects. Moreover, in Andersson & Armstrong's studies the complex spike discharges were in some instances related to perturbations that the animal quite likely expected, but chose not to adjust for in the locomotor cycle. After the forelimb of the cat encountered an unlatched rung, the hindlimb also encountered the unlatched rung and this perturbation was also signaled to the cerebellum. After this sequence has occurred several times, one can reasonably presume that the cat is expecting that the hindlimb will be perturbed. Consequently, expectation or not expectation is not the issue, but rather whether a mismatch occurs between the intended and achieved movement. Gellman et al. (1985) acknowledged that "unexpected" contact detection could in some circumstances be considered error detection, but they had difficulty in reconciling the responses of climbing fibers to passive displacement in the absence of movement with error detection. Perhaps in such a situation a climbing fiber response would be indicative of an error in posture as opposed to an error in movement. If a cat in a posture of repose is disturbed by a footpad touch, then a climbing fiber response can be a signal that the achieved posture is not satisfactory. Since the cerebellum is involved in controlling both posture and movement, errors in postural performance must also be reported.

2.6 Summary

We now know that the climbing fibers can be modulated both tonically and phasically and that they can be exquisitely sensitive to sensory stimulation. While different investigators have used different phrases to describe what the climbing fibers are signaling, for the majority the climbing fibers are signaling an error in motor performance, which includes posture as well as overt movement. The visual climbing fiber signals of error are reliable and robust in comparison to those of most other climbing fiber systems. This difference is likely a consequence of the fact that the error in performance, the retinal image motion, is not computed by a neural comparator dealing with intended and achieved movement, but is computed mechanically by the movement of the eye relative to the external world. The relative movement represents the difference between the intended and achieved movement, which is detected directly at the sensory surface by specialized speed and direction selective retinal ganglion cells.

3. Interactions between Complex and Simple Spikes

The low firing frequency of the climbing fibers not only prompted thinking about them in terms of encoding phasic events, but also encouraged the idea that their low frequency should be compensated by an interactive influence on the Purkinje cell simple spikes. Proposals as to how complex spikes may magnify their influence through interaction with simple spikes run the gamut from transient to tonic and from short-lasting to long-term.

3.1 The climbing fiber pause and beyond

A transient climbing fiber effect on simple spike activity can be seen by recording spontaneous activity from Purkinje cells. After the onset of a complex spike the simple spikes pause for a variable period of time, from 10 to several hundred milliseconds, depending on the state of the animal (Bell & Grimm, 1969; Bloedel & Roberts, 1971; Granit & Philips, 1956; McDevitt et al., 1982; Murphy & Sabah, 1970; Mano et al., 1986, Sato et al., 1992, 1993). The pause in simple spike firing, which is called the climbing fiber pause (e.g. Armstrong, 1974), may be due to a combination of inactivation of the Purkinje cell membrane and the influence of climbing fiber collaterals on cerebellar cortical neurons that could decrease (via Golgi cells) transmission in the mossy fiber-granule cell pathway and could also directly inhibit (via basket cells) the Purkinje cells (Bloedel & Roberts, 1971).

After the pause, the simple spike activity has one of three general patterns (Figure 4)a return to baseline within a few milliseconds (pure pause), a gradual return to baseline over several tens of milliseconds (pause-reduction), or a rapid increase in activity to a level that is greater than baseline for a period of several tens of milliseconds (pause-facilitation). These observations of McDevitt et al. (1982) were recently confirmed and given a statistical appraisal by Sato et al. (1992). Like McDevitt et al., Sato et al. used the decerebrate, unanesthetized cat and recorded spontaneous activity of individual Purkinje cells. They also found the same types of temporal patterns in the simple spike activity subsequent to the complex spike. Of these patterns, the most common by far was pause-facilitation (71%), followed by pure pause (25%), and then pause-reduction (4%). Sato et al. also quantified the parameters of the pause and concluded that it is unlikely that the pause found in the pure pause and pause- facilitation types of Purkinje cells contributes any substantial decrease to their transient simple spike output. With regard to action potentials transmitted down the Purkinje cell axon, the pause, in fact, need not even exist because in approximately half of the Purkinje cells a burst of 2-5 action potentials was found to occur with climbing fiber activation (Ito & Simpson, 1971; see also Mano et al., 1989).

The complex spike influence on the ensuing simple spike activity is highly variable and includes the possibility that the complex spike does not effectively influence the simple spike activity, so that under physiological conditions the two systems may be considered independent, as advocated by Llins (1970, 1974). Sato et al. (1993) ascribed the variability to competition between mechanisms that facilitated and depressed the simple spike activity. A proposed mechanism for depression coupled the projection of climbing fiber collaterals to inhibitory interneurons with the observation by Kano et al. (1992) that GABAA receptor-mediated inhibitory effects on Purkinje cells in vitro can be potentiated by electrical stimulation of climbing fibers. Kano et al. also suggested that a high complex spike frequency in vivo would maintain the GABAA receptor sensitivity of Purkinje cells at a high level, whereas a low frequency would maintain it at a low level, thereby controlling the tonic simple spike firing rate (see below). For post-climbing fiber facilitation of simple spike activity Sato et al (1993) offered three tentative mechanisms: first, inhibition of inhibitory interneurons by Purkinje axon collaterals; second, inhibition of Golgi cell activity after climbing fiber activation (Schulmann & Bloom, 1981; Yamamoto et al., 1978) leading to disinhibition of the mossy fiber-granule cell relay; and third, prolonged depolarization of the Purkinje cells due to a voltage-dependent increase in calcium conductance following a complex spike (Ekerot & Oscarsson, 1981; Llins & Sugimori, 1980).

A pause in the extreme can be produced by increasing the level of inferior olive activity to 8-10 Hz, either with the use of tremogenic drugs such as harmaline or with electrical stimulation, which results in elimination of simple spike activity (Demer et al., 1985; Llinas & Volkind, 1973; Rawson & Tilokskulchai, 1981, 1982). Conversely, silencing the inferior olive, whether achieved by lesion, lidocaine, or cooling results in a gradual increase in the tonic background level of simple spike activity (Benedetti et al., 1984; Colin et al., 1980; Demer et al., 1985; Leonard & Simpson, 1986; Montarolo et al., 1982). This relation between inferior olivary activity and tonic simple spike activity has been found in all parts of the cerebellum so far studied, but its physiological meaning remains unclear (Ojakangas & Ebner, 1992, 1994). The tonic relationship between complex and simple spike activity should not be confused with the reciprocal relationship that is typically seen in the modulation of complex and simple spikes in the flocculus (Demer et al., 1985; Ghelarducci et al., 1975; Graf et al., 1988; Stone & Lisberger, 1990). Leonard & Simpson (1986) recorded the modulation of the simple spikes in the flocculus before and after silencing the climbing fibers with submicroliter injections into the dorsal cap of the inferior olive. They found that while the simple spike background activity increased on average 35%, the absolute depth of modulation surprisingly remained unchanged (Figure 5).

3.2 Gain change hypothesis

One interactive view of the consequence of climbing fiber signals is that for a short period (about 200 msec) after a complex spike the responsiveness of the Purkinje cell to its mossy fiber input is enhanced. This view of climbing fiber function, called the gain change hypothesis, evolved from a series of studies conducted by Ebner & Bloedel (1981a, b; 1984; Ebner et al., 1983). In these studies, both spontaneous and naturally evoked Purkinje cell activity recorded from decerebrate, unanesthetized cats was used to compile cross-correlograms and post-stimulus time histograms to examine the relation between a complex spike and the pattern of the ensuing simple spike activity. The findings made with each of a variety of paradigms led to the same conclusionthe climbing fiber input to a Purkinje cell frequently results in a short-lasting enhancement of its responsiveness to mossy fiber inputs. The change in responsiveness occurred during a period of up to several hundred milliseconds following a complex spike, but it did not persist in the absence of the complex spike. Increased responsiveness was manifest as an increase in the amplitude of both the excitatory and the inhibitory components of the simple spike response.

More recently, Bloedel and colleagues (Bloedel & Kelly, 1992; Lou & Bloedel, 1986, 1992a,b) have tied the gain change hypothesis to the modular organization of synchronously active climbing fibers to propose the "dynamic selection hypothesis," which bears some resemblance to the "synchronizing pulse hypothesis" of Mano et al. (1986, 1989). Under the "dynamic selection hypothesis" the climbing fibers are seen as a means to spatially focus modulation of Purkinje cells and, in turn, nuclear cell activity. According to the hypothesis, Purkinje cells that have a synchronous activation of their climbing fiber input would, through enhancement of their responsiveness to the mossy fiber input, be much more dramatically modulated than other Purkinje cells that are activated by a comparable mossy fiber input, but that are not activated by their climbing fiber input. Therefore, the climbing fiber input acts to select or emphasize those mossy fiber inputs that will produce the greatest modulation of the simple spikes. Evidence supporting the dynamic selection hypothesis has been obtained by Lou & Bloedel (1992a,b) using a perturbed locomotion paradigm in decerebrate, locomoting ferrets. The responses of 3-5 sagittally aligned Purkinje cells were recorded simultaneously in response to an intermittent perturbation of the forelimb during the swing phase, and the amplitude of the combined simple spike responses across the population of Purkinje cells correlated with the extent to which their climbing fiber inputs were synchronously activated. (The word "synchronous" as used by Bloedel and colleagues refers to a much less constrained timing relation than when used by Llins and colleagues. This difference will be elaborated below.)

3.3 Climbing fibers as "teachers"

Marr (1969) and later Albus (1971) hypothesized that the cerebellum learns to perform motor skills through climbing fiber induction of long-term changes in the strength of parallel fiber synapses on Purkinje cells. Ito (1970) put this idea into a specific context by proposing that the flocculus adjusts the performance of compensatory eye movements in response to climbing fiber signals of "visual blur" functioning as a teaching input. The hypothesis that climbing fibers function as teachers is an interactive one, but unlike the interactive gain change hypothesis, the focus is on the permanency rather than the transientness of the climbing fiber influence on the simple spike activity. A further distinction between these two interactive views lies in the polarity of the influence of the complex spike on the simple spike modulation. Ito and colleagues (Ito et al., 1982a; Ito, 1989; Kano & Kato, 1988) emphasize the reciprocal relation between complex and simple spike activity in the flocculus because that relation is the one that would be established by a long-term depressive action of climbing fibers on simple spike activity. On the other hand, recordings advanced in support of the gain change hypothesis show that complex and simple spikes often increase together (Ebner & Bloedel, 1981b; Ebner et al., 1983; Lou & Bloedel, 1992b).

Because the issue of cerebellar learning is explored in detail in several other articles in this volume (Crepel; Linden), and because our studies of the visual climbing fiber input to the flocculus have been directed toward considerations other than cerebellar plasticity, we will not re-cycle the conflicting views of the proponents and opponents of a teaching function of the climbing fibers (Bloedel, 1992; Gilbert & Thach, 1977; Ito, 1982, 1984; Kawato & Gomi, 1992; Lisberger, 1988; Lisberger & Sejnowski, 1992; Miles & Lisberger, 1981; Llins & Welsh, 1993; Ojakangas & Ebner, 1992, 1994; Schreurs & Alkon, 1993; Tempia et al., 1991; Thach et al., 1992; Thompson, 1986; Welsh & Harvey, 1989). Even so, we would like to add several observations. First, it is interesting to note that the version of the teaching hypothesis presented by Marr (1969) stated that if complex spike activity reflected a response of sensory receptors, then the movement produced by discharge of the respective Purkinje cells should act to move the receptor such that the complex spike activity decreased. If otherwise, then a positive feedback situation would exist because Marr envisioned that the climbing fibers acted to potentiate the simple spike activity. In the version of the teaching hypothesis advocated initially by Albus and later by Ito, the opposite relationship should hold because the parallel fiber input that is paired with climbing fiber input is postulated to become less effective. That is, reduction of the simple spike activity of a Purkinje cell that receives from a given climbing fiber should lead to a sensory input that opposes the occurrence of that climbing fiber's discharge. Conversely, stimulation of that Purkinje cell would be expected to evoke a motor response that, in turn, leads to a climbing fiber response. Such an association between complex and simple spikes is present for the floccular visual climbing fiber system of the rabbit (Van der Steen et al., 1994). While this reciprocal relation is consistent with the Albus-Ito version of the teaching hypothesis, its presence does not demonstrate the validity of that hypothesis. For instance, visual modulation of the floccular complex spikes cannot be held to carve out a reciprocal relation from a monolith of tonic simple spike activity because the reciprocal relation between visually induced complex and simple spike modulation is present in the flocculus of dark-reared rabbits (Soodak et al., 1988). Furthermore, concomitant activation of complex and simple spikes can occur in relation to movements that have already been well-learned (Mano et al., 1986, 1989; Ojakangas & Ebner, 1994). Also, as described below, some of the rabbit floccular Purkinje cells that showed reciprocal simple and complex spike modulation during rotation in the presence of vision, showed concomitant simple and complex spike activation during rotation in the absence of vision (De Zeeuw et al., submitted).

As a second observation, we note that when the rabbit's compensatory eye movement gain is adapted to a new, stable level, the modulation of the visual climbing fibers would continue to be substantial because of the retinal image motion still remaining (e.g., Nagao, 1983, 1988). Therefore, even though the signals of retinal image motion on the visual climbing fiber input to the flocculus are those proposed to be necessary for the climbing fibers to function as "teachers," those signals are apparently not sufficient. Complex spike modulation in conjunction with parallel fiber modulation is not by itself guaranteed to induce changes in the parallel fiber synapses. This conclusion was reached by another route by Ekerot & Kano (1985) who proposed that the original Marr and Albus hypotheses required revision to include an additional condition in order for the conjunction of complex and simple spikes to change the weight of the parallel fiber synapses. Conjunction alone is not enough. The level of inhibitory input to the Purkinje cell and its influence on the degree to which the complex spike results in calcium entry are also critical (Campbell et al., 1983; Eckerot & Oscarsson, 1981; Llins & Sugimori, 1980; Sakurai, 1987, 1990; Callaway et al., 1995). The extent to which these conditions are satisfied in the normal cerebellum is a matter of fervent debate.

4. The Olivo-cerebellar System as a "Timing Device"

The above notions about the consequences of climbing fiber signals have in common the view that the climbing fibers evoke their effects through a change in the simple spike activity. Standing apart from this view is the proposal by Llins and colleagues that the climbing fiber input acts on its own, forcing the Purkinje cells to fire in a burst-like manner to evoke an effect in the cerebellar nuclei neurons that is distinguishable from the one induced by the simple spikes (Llins, 1974; Llins, 1985; Jahnsen 1986; Llins & Muhlethaler, 1988a,b). In this proposal the climbing fibres serve as a "timing device" for movement execution (Llins, 1985, 1991; Llins & Welsh, 1993; Llins & Yarom, 1986). The timing hypothesis has at least three components: 1) the climbing fibers convey motor commands; 2) these commands are for phasic motor acts; and 3) the inferior olive controls the timing among the different components of a motor act.

4.1 Synchrony and rhythm of olivary neurons

The "timing device" proposal is based on two observations (see Figure 6). First, because of particular membrane conductances, olivary neurons (and consequently Purkinje cell complex spikes) tend to fire rhythmically, typically at a characteristic frequency of about 8-10 Hz (Llins & Yarom, 1981a,b,1986; Bloedel & Ebner, 1984; but see Keating & Thach, 1993). Second, olivary neurons show a tendency to fire synchronously, that is within a few milliseconds of each other (Bell & Grimm, 1969; Bell & Kawasaki, 1972; Llins & Yarom, 1981a), due to the fact that they are electrotonically coupled by dendrodendritic gap junctions (De Zeeuw et al., 1989; Sotelo et al., 1974; Llins, 1974; Llins et al, 1974). Multiple electrode recording has shown that complex spike synchrony is most prevalent among Purkinje cells in the same rostrocaudal band (Bell & Kawasaki, 1972; Llins & Sasaki, 1989; Sasaki et al., 1989; but see Welsh et al., 1995).

It is important to recognize that the use of the term "synchronous" by Llins and colleagues differs from that of Bloedel and colleagues (Bloedel & Kelly, 1992; Lou & Bloedel, 1992a, b) who use synchronous to describe the clustering of complex spikes of a set of Purkinje cells within a much broader period of time, typically tens of milliseconds. With such a temporal dispersion, synchrony does not necessarily mean that the timing relations among the complex spikes are due to coupling via gap junctions. At the same time, it should be noted that synchrony in the millisecond range can be achieved with chemical synapses if a strong afferent input is shared by a set of neurons. However, the importance of electrotonic coupling for the olivocerebellar system is supported by the distribution of dendritic lamellar bodies that are associated with dendrodendritic gap junctions (De Zeeuw et al., 1995); these dendritic lamellar bodies are distributed in all olivary subdivisions and their density in the inferior olive is higher than in any other area of the brain.

Lang et al. (1989, 1990; Lang, 1995) have shown that complex spike synchrony is enhanced by administration of GABA-antagonists or by lesioning the GABAergic neurons in the cerebellar nuclei that innervate the inferior olive. Many of these cerebellar nuclei GABAergic neurons terminate apposed to the dendrodendritic gap junctions inside the olivary glomeruli (De Zeeuw et al., 1989; Nelson & Mugnaini, 1989; Llinas, 1974; Llinas et al., 1974; Sotelo et al., 1974; Sotelo et al., 1986). These GABAergic neurons can vary the coupling among olivary neuronal clusters representing different muscles. In that way, particular sets of muscles may be called into play at appropriate times during a movement (Lang, 1995; Welsh et al., 1995). Indeed, dynamic repatterning of complex spike synchrony was found by Welsh et al. (1995) for a population of rat Purkinje cells recorded during rhythmic tongue movements. Lang (1995) noted that the degree of synchrony alters computational ability. Low levels of synchrony within the inferior olive reflect many small ensembles of neurons whose activity is relatively independent of each other, permitting many independent computational tasks to be performed. In contrast, a highly synchronized inferior olive has much less computational ability, but that would have advantages for task requiring simultaneous contraction of many different muscles. Llins (1991) has noted that human reaction-time movements have been reported to be paced by a normal 10 Hz physiological tremor (Goodman & Kelso, 1983), which may be mediated by the inferior olive. Moreover, harmaline increases the occurrence of synchrony and rhythmicity of olivary neurons and can, thereby, drive a 10 Hz body tremor (Llinas & Volkind, 1973).

4.2 Climbing fibers and motor commands

Many investigators of the climbing fiber system have found that complex spike activity can be influenced by sensory stimuli (see above). But what is the relation of complex spike activity to voluntary motor behavior and to motor commands in particular? This question has been addressed by focusing on complex spike activity that occurs in relation to the onset of movement. (Mano et al, 1986,1989; Fukuda et al, 1987; Lang, 1995; Lang et al, 1992; Welsh et al, 1992, 1993, 1995).

Mano et al. (1986, 1989) recorded the responses of single Purkinje cells in the intermediate and lateral parts of the cerebellar hemispheres while trained monkeys performed visually guided wrist tracking movements. For those Purkinje cells whose complex spike activity was response-locked to the movement, about 60% showed a phasic increase of complex spike firing rate at the onset of the movement. The complex spike activity increased with both rapid and slow tracking movements, but the increase was larger with faster step-tracking movements than with slower ramp-tracking movements. In most of the Purkinje cells that showed complex spikes locked to movement, the increase in complex spike firing rate occurred during "motor time," which is the period from the onset of the EMG change in the prime mover muscles to the beginning of the movement. Interestingly, the complex spikes were not response-locked when the monkey returned the manipulandum to center position after completing the tracking task, even in those Purkinje cells that showed a significant increase of complex spike activity during visually guided tracking. This difference, which was present even when differences in direction, starting point and speed were controlled for (Mano et al., 1989), may be due to the fact that the initial movement was triggered by an external stimulus and the timing constraints on performance were greater than for the return movement.

Consideration of the effect of the complex spike on the subsequent simple spike activity led Mano et al (1986, 1989) to propose the "synchronizing pulse hypothesis." Under this hypothesis the role of the climbing fibers may be to "assist the effective onset or cessation of simple spike frequency modulation when animals are required to perform voluntary movement with precise timing as in externally triggered movement." The improvement in effectiveness is held to be due to short-term actions of the complex spike on the simple spike firing pattern in combination with electrical coupling among inferior olivary neurons.

Several studies have used multiple electrode recording to examine complex spike activity in relation to spontaneous vibrissae movement (Fukuda et al., 1987; Llins, 1991; Lang, 1995) and in relation to vibrissae movement evoked by electrical stimulation of the motor cortex (Lang et al, 1992; Lang, 1995). With spontaneous movements in animals anesthetized with ketamine and in awake animals, complex spikes often occurred within 20 milliseconds prior to onset of vibrissae movement. Lang (1995) pointed out that the ability to observe complex spike timing in relation to vibrissae movement may have been aided by the fact that each vibrissa is moved by a single muscle. This simple arrangement may, in part, explain why some previous investigations (e.g., Thach, 1968; Andersson and Armstrong, 1987) found only a weak or non-existent relationship between complex spikes and the more complicated multi-joint limb movements. While complex spikes mostly preceded onset of the vibrissae movement, they also occurred soon after movement onset and thus were coincident with the movement. Lang et al. (1992) found that the probability of vibrissae movement increased when the complex spikes were more synchronous. Lang (1995) noted that the inferior olive may not only be determining the times at which the movements occur, but also may be playing a permissive role in allowing movements to occur because the movements do not necessarily have to occur at every opportunity. The permissive aspect could be a reflection of the subthreshold oscillation of olivary membrane potential, while the occurrence of a movement may be a reflection of the synchronous discharge of the coupled ensemble of olivary neurons.

In another study revealing time-locking of complex spikes to movement, Welsh et al., (1995) used multiple electrodes to record from Purkinje cells in crus 2a of alert rats trained to tongue lick in response to a tone. The complex spike population response revealed a gradual increase in activity preceding a sharp peak at the time of maximal tongue protrusion. Complex spike activity occurring after that time was suggested to be associated with tongue movements that would be required to hold the water drop on the tip of the tongue. A less pronounced second peak in climbing fiber activity occurred coincident with the closing of the mouth.

The three studies described above have in common the finding that the complex spike activity can increase in relation to rapid, precisely initiated motor behavior, but the exact relationship was different in each case. In one case the increase occured mostly between EMG onset and movement onset, in another the increase often occurred within 20 msec prior to movement onset, while in the third the increase was centered on a particular facet of the movement. Thus, in some instances the complex spikes of some Purkinje cells precede movement, while in other instances the complex spikes occur after movement onset, but still in close temporal relationship to movement onset. Those complex spikes occurring prior to the movement can be associated with the beginning of the movement, while those occurring later may be related to subtle features of the movement that occurred immediately after onset, or they may be associated with stopping the movement.

4.3 Floccular climbing fibers and motor signals

The signals of retinal image motion on dorsal cap and ventrolateral outgrowth neurons arise from midbrain projections that are non-GABAergic and presumably excitatory (Horn & Hoffmann, 1987; Nunes-Cardozo & Van der Want, 1990). In addition, the dorsal cap and the ventrolateral outgrowth are innervated by neurons in the nucleus prepositus hypoglossi (Gerrits et al., 1985; McCrea & Baker, 1985) that are partly cholinergic (Barmack et al.,1993b) and partly GABAergic (De Zeeuw et al., 1993). The GABAergic input from the prepositus hypoglossi together with the GABAergic input from the dorsal group y and the ventral dentate nucleus (De Zeeuw et al., 1994a) account for most, if not all, of the inhibitory input to the dorsal cap and ventrolateral outgrowth. As in other regions of the inferior olive (De Zeeuw et al., 1989), the inhibitory synapses account for approximately half of the synaptic input. At least one-fifth of the inhibitory synapses in the dorsal cap and ventrolateral outgrowth are located in the glomeruli and act to control the strength of coupling among olivary neurons, as discussed above. The other inhibitory synapses are located mainly elsewhere on the dendrites. Their contribution to floccular climbing fiber signaling remains to be determined, but some indication of what these inhibitory inputs may be doing is available from recordings in the awake rabbit of floccular complex spike activity during vestibular stimulation in the absence of vision (De Zeeuw et al., submitted). The complex spikes of a minority of Purkinje cells that were modulated by retinal image motion were also modulated during sinusoidal rotation about the vertical axis in the dark. This modulation was not as strong as for visual stimulation. The great majority of these Pukinje cells showed increased complex spike activity with contralateral head rotation in the dark, which is opposite to the complex spike behavior with rotation in the light. Since the polarity of the complex spike modulation during rotation in the dark was generally opposite to that in the light, the complex spike modulation in the dark was not due to an incomplete darkness. The most likely candidate for the input underlying the modulation during rotation in the dark is the nucleus prepositus hypoglossi, which contains neurons that project to the caudal dorsal cap (De Zeeuw et al., 1993) as well as neurons that have eye velocity and eye position signals associated with the horizontal vestibulo-ocular reflex (Lopez-Barneo et al., 1982; Escudero et al., 1992).

In further consideration of the possibility that motor signals are present on some floccular climbing fibers, we note that one of the five floccular zones delineated with acetylcholinesterase histochemistry (Tan et al., 1995a), -- the C2 zone, -- does not receive signals of retinal image motion (De Zeeuw et al., 1994b). The source of its climbing fibers is the rostral pole of the medial accessory olive (Tan et al., 1995b), which receives a major input from the nucleus of Darkschewitsch (De Zeeuw & Ruigrok, 1994). The part of the nucleus of Darkshewitsch that innervates the rostral pole of the medial accessory olive (Porter et al., 1993) receives input from the frontal eye fields of the cerebral cortex (Miyashita & Tamai, 1989). This connectivity suggests that the climbing fiber input to the C2 zone of the flocculus may carry motor signals that are related to voluntary gaze shifts involving both the head and eyes.

5. Visual Climbing Fibers in the Rabbit Vestibulocerebellum

5.1 Modules of the rabbit flocculus

In attempting to reach some consensus about several of the prominent views of climbing fiber function, we have performed experiments in the rabbit flocculus where the visual climbing fiber input is known to carry a reliable and robust error signal of retinal image motion. A functional overview of the modular input-output relations of the visual climbing fibers in the rabbit's flocculus is shown in Figure 7.

The flocculus of the rabbit consists of five zones (1, 2, 3, 4 and C2) whose borders can be delineated in the floccular white matter by using acetylcholinesterase (AChE) staining (De Zeeuw et al., 1994b; Tan et al.,1995 a,b; Van der Steen et al., 1994). The climbing fiber projections to zones 1 and 3 are derived from the rostral dorsal cap and ventrolateral outgrowth; the climbing fiber projections to zones 2 and 4 are derived from the caudal dorsal cap; and the climbing fiber projection to zone C2 is derived from the rostral pole of the medial accessory olive (Tan et al. 1995b). In zones 1-4 the complex spikes are optimally modulated by rotational optokinetic stimulation about either the vertical axis or about a horizontal axis approximately perpendicular to the plane of the ipsilateral anterior semicircular canal (Simpson et al., 1981; Graf et al., 1988; Kano et al., 1990b; Kusunoki et al., 1990; Wylie & Frost, 1993). These optimal axes have a geometry similar to that of the best-response axes of the semicircular canals and to the axes about which the three pairs of extraocular muscles rotate the eye (see Figure 2). The horizontal axis complex spikes in the rabbit flocculus can be divided into two classes based on ocular dominance. Those dominated by the contralateral eye respond best to rotation about an axis at about 45 contralateral azimuth; those dominated by the ipsilateral eye respond best to rotation about an axis at about 135 ipsilateral azimuth. We refer to the corresponding Purkinje cells as contra-45 and ipsi-135 neurons, respectively. The vertical axis (VA) neurons are located in zones 2 and 4, while the contra-45 and ipsi-135 neurons are located in zones 1 and 3. The climbing fibers of zone C2 do not respond to optokinetic stimulation (De Zeeuw et al., 1994b).

5.2 Simple spike transients in the rabbit flocculus

The gain change hypothesis (Ebner & Bloedel, 1981a, b; 1984; Ebner et al. 1983) asserts that the climbing fiber input to a Purkinje cell can produce a short-lasting enhancement of the responsiveness to mossy fiber inputs. We investigated the patterns of simple spike transients of Purkinje cells in the flocculus of awake rabbits presented with vestibular stimulation in the light and dark, and with optokinetic stimulation. Vestibular stimulation was provided by rotating the restrained rabbit about the vertical axis with the use of a servo-controlled turntable; optokinetic stimulation was provided by a planetarium projector rotating about the vertical axis. Eye movements were recorded using the scleral search coil technique (Robinson, 1963), and Purkinje cells, identified by the presence of a brief pause in simple spike activity following a complex spike, were recorded extracellularly. Complex spike and simple spike peristimulus time histograms (PSTH) were computed along with complex spike-simple spike cross-correlograms.

A statistical method similar to that of Sato et al. (1992) was used to assign the Purkinje cells to one of three categories pure pause, pause-facilitation, or pause-reduction on the basis of the cross-correlogram patterns obtained with at least 100 sequential complex spikes . To be considered a pause-facilitation cell the average simple spike activity during at least one 20 msec period within the first 50 msec following the end of the brief pause had to be significantly greater (p<0.01) than the average simple spike activity during the 100 msec prior to the onset of the complex spike. This criterion is less restrictive than that of Sato et al. since the 20 msec period used to test for significance was not constrained to begin immediately following the end of the pause in the cross-correlogram. A similar test was used to assign cells to the pause-reduction category. From our sample of 43 Purkinje cells, 26 were pure pause (Figure 8), 12 were pause-facilitation (Figure 9) and 5 were pause-reduction. Even with our more liberal statistical criterion, the percentage of pause-facilitation cells was less than half of the 60-70% we had anticipated from the findings of Ebner & Bloedel (1981a), McDevitt et al. (1982) and Sato et al. (1992). Also, the duration of the facilitation was substantially shorter than anticipated. Part of these differences may be attributed to the decerebration employed by these investigators because in the awake monkey Mano et al. (1986) found, as we did in the awake rabbit, that the pure pause category was the predominant one (66%), with the remaining Purkinje cells in the pause-facilitation category. In this context, it is noteworthy that in another study in the awake monkey, Ojakangas and Ebner (1994) found that about 75% of the Purkinje cells were in the pure pause category.

The significance of the pause and any ensuing simple spike transients for the control of slowly changing compensatory eye movements eludes us. The duration of the pause (10-20 msec) of the pure pause and pause-facilitation Purkinje cells is very brief in comparison to the time course of the compensatory eye movements produced in response to the stimuli employed. The reciprocity often seen between simple and complex spikes in the flocculus is not due to the brief pause, because, as also found by several other investigators (Demer et al., 1985; Leonard, 1986; Sato et al., 1992), it is too brief and too infrequent to make a substantial contribution to the reciprocal relationship. The duration, size, and infrequent occurrence of the transient facilitations, and reductions, indicate that they too make little contribution to the overall simple spike modulation. On the other hand, transients in simple spike activity following complex spikes do occur in the flocculus. Perhaps their contributions will become apparent in relation to other aspects of eye movements such as voluntary gaze shifts, which in the rabbit typically occur as a combined eye and head movement.

5.3 Synchrony and rhythmicity in the rabbit vestibulocerebellum

Complex spike synchrony was investigated by recording from pairs of Purkinje cell in the ventral nodulus of ketamine-anesthetized rabbits (Wylie et al., 1995). Like the flocculus, the ventral nodulus is comprised of zones. Three of the four zones receive visual climbing fibers from the dorsal cap and ventrolateral outgrowth. The climbing fibers of the most medial zone are not responsive to retinal image motion, but receive a vestibular climbing fiber input from the beta subnucleus. Two zones (VA1 and VA2) receive their climbing fiber input from the caudal dorsal cap, which contains VA neurons. The zone between the two VA zones contains contra-45 and ipsi-135 neurons, which receive their climbing fibers from the rostral dorsal cap and the ventrolateral outgrowth (zone HA) (Balaban & Henry, 1988; Barmack et al., 1989, 1993a; Barmack & Shojaku, 1992; Kano et al., 1990a,b; Katayama & Nisimaru, 1988; Shojaku et al., 1991; Wylie et al., 1994).

In confirmation of the studies of Llins and colleagues described above, we found that pairs of Purkinje cells from the same zone often showed complex spike synchrony, which we defined as the tendency for the complex spikes of the two Purkinje cells to discharge within at most 2 msec of each other. To assess the temporal relationship of a cell pair, cross-correlograms were constructed using 1, 2, 5, 10 and 20 msec binwidths for 100 bins on either side of time zero. To quantify the temporal relationship of a cell pair, the cross-correlation coefficient was used as a synchrony index (Sasaki et al., 1989; Sugihara et al., 1993; Wylie et al, 1995). The tendency of a neuron pair to fire within a given time period was determined as significant if one of the two time-zero bins was denoted as a peak. A time-zero bin was denoted as a peak if, a) it had a value of at least 5, b) it was the highest bin, and c) it was at least 3 standard deviations above the mean. The activity was considered to be synchronous when one of the two time-zero peaks in the cross-correlograms with 1 or 2 msec bins was significant. The cross-correlogram for two Purkinje cells from the VA1 zone is shown in Figure 6E. During a 500 sec period one cell had 330 spontaneous complex spikes while the other had 365. On 53 occasions, the complex spikes discharged within a millisecond of each other. Of 82 pairs consisting of two Purkinje cells in the same zone, 33 (40%) showed a tendency to fire within 2 msec. Included in these 82 are 7 pairs comprised of a contra-45 Purkinje cell and an ipsi-135 Purkinje cell located in the HA zone. Five of these pairs showed a synchronous relationship. The occurrence of complex spike synchrony has recently been confirmed in the flocculus of the awake rabbit (Figure 10).

We also recorded from 16 pairs consisting of one Purkinje cell in each of the VA1 and VA2 zones, which are spatially separated by the 1mm wide HA zone. Nonetheless, 6 pairs showed a synchronous relationship. Moreover, 3 of 14 pairs consisting of one floccular VA Purkinje cell and one nodulus VA Purkinje cell showed a synchronous relationship, which is notable when one considers that the climbing fibers to these zones are probably of different lengths. Sugihara et al. (1993) have shown that despite the fact that some olivocerebellar branches in crus 2a are considerably longer than others, their conduction times are quite uniform (but see Aggelopoulous et al., 1994).

The value of the synchrony index varied depending upon the nature of the response of the particular complex spike pair to optokinetic stimulation. For cross-correlograms computed at the 2 msec binwidth, the largest average synchrony index was 0.039 for 39 pairs of complex spikes that were recorded in the same zone and responded best to optokinetic stimulation about the vertical axis (Wylie et al., 1995). While this value may appear small, it is about an order of magnitude larger than the value expected for two random independent spike trains having a mean firing rate in the usual range of complex spike firing rates (Sugihara et al., 1993). Even so, some may still question the importance of this degree of synchrony. As two tentative answers, it is suggested that the effect of synchrony on the cerebellar nuclei may not be directly proportional to the strength of the synchrony (Welsh, personal communication), and that changes in the pattern of the synchrony across the cortex could be more important than changes in the value of the synchrony index (Welsh et al., 1995).

We do not know the functional significance of synchrony between different floccular and nodular modules that receive similar climbing fiber inputs. In broad terms, the nodulus and ventral uvula are involved in control of the velocity storage mechanism, whereas the flocculus is important for controlling the gain of the vestibulo- ocular and optokinetic reflexes (Cohen et al., 1992; Waespe et al., 1983, 1985). In addition, both the nodulus (Ito et al., 1982b; Nagao, 1983) and the flocculus (Stahl & Simpson, 1995: De Zeeuw et al., submitted; Ito et al., 1982b; Nagao, 1983) influence the phase of compensatory eye movements, but in opposite directions. Thus, the modules in the nodulus and flocculus may be synchronized to coherently combine different aspects of compensatory eye movements.

Because retinal image motion effectively modulates the complex spike activity of nodular and floccular Purkinje cells, we investigated whether that stimulus would affect the strength of the temporal relation of these complex spikes. Purkinje cell pairs with complex spikes of the same class were recorded in anesthetized rabbits during spontaneous activity and during retinal image rotation about the preferred axis (Wylie et al., 1995). As a quantitative measure of synchrony, the cross-correlation coefficient was calculated during rotation in the On direction, during rotation in the Off direction, and during spontaneous activity for 53 pairs of Purkinje cells. To compare these three conditions, ratios were calculated for each comparison (e.g., (On-spont)/(On+spont), and a t-test was performed versus the null hypothesis that there was no difference (i.e., (On- spont)/(On+spont) = 0). For cross-correlation coefficients obtained from 2 msec binwidth correlograms, the mean of the ratio for On versus spontaneous was significantly greater than zero (0.18, p<0.02); the means of the two other ratios (Off vs. spont; On vs. Off) were not significantly different from zero.

The level of synchrony may be different in awake animals for two reasons. First, compensatory eye movements will occur in response to visual and vestibular stimulation and Welsh et al. (1992; 1993, 1995) have noted that complex spike synchrony in crus 2a increases during tongue movements. Second, as outlined above, Lang et al. (1989, 1990; Lang, 1995) have shown that removal of the GABAergic input to the inferior olive increases the degree of complex spike synchrony. The GABAergic input to the dorsal cap and the ventrolateral outgrowth is from the nucleus prepositus hypoglossi (De Zeeuw et al., 1993), the ventral dentate nucleus and the dorsal group y (De Zeeuw et al., 1994a). Neurons in prepositus hypoglossi and dorsal group y are modulated during visual-vestibular stimulation in alert preparations (Chubb et al., 1984; Lopez-Berneo et al., 1982; McFarland & Fuchs, 1992; Partsalis et al., 1993) and this modulation may result in stronger synchrony.

Many nodular Purkinje cells showed rhythmic complex spike activity as revealed in auto-correlograms (Figure 6F; Wylie et al., 1995)). Thirty-two percent (36 of 114) displayed one or more peaks in their complex spike auto-correlogram. The average rhythm was 8 Hz (range 2.5-12.5 Hz), in agreement with the findings of Llinas & Yarom (1981a, b; 1986). For the 36 rhythmic cells, one to six peaks on each side of the correlogram were apparent (mean = 2.4 peaks). While most cells (25) had two or more peaks, 11 had only one peak. (Cells showing a single peak in the auto-correlogram would accurately be described as producing pairs of action potentials separated by a characteristic time interval). Thirty-five of the 114 Purkinje cells were recorded for sufficient time to obtain more than 600 complex spikes. A 10msec binwidth auto-correlogram with 1000 bins on either side of time-zero was constructed for these cells. On the basis of visual inspection the initial part of these auto-correlograms was judged to be rhythmic for 9 of the 35 cells, as exemplified in Figure 6F. The rhythm strength for these 9 cells was quantified by first determining the best-fit sine wave for the initial 2 or 3 periodicities of the auto-correlogram and then calculating the ratio of the amplitude to the mean of the sine wave. For the cell shown in Figure 6F, the rhythm strength was 0.38 for the first 2 periodicities of the auto-correlogram. The average rhythm strength for the 9 cells was 0.35 (range 0.26-0.53); the average frequency of the best-fit sine wave was 8.0Hz (range 4.8-11.5Hz).

We found that the occurrence of synchronous firing and rhythmic oscillation were independent. Synchrony occurred with or without the rhythmic firing of one or both neurons in the pair. These observations, made in the anesthetized rabbit, are in contrast to those made in the slice preparation where spontaneous oscillations of olivary neurons occurred synchronously in all cells examined (Llins & Yarom, 1986).

6. Conclusion

We have attempted to put some of the prominent hypotheses of climbing fiber function into the context of the visual climbing fiber input to the rabbit's vestibulocerebellum. With regard to the gain change hypothesis, the majority of floccular Purkinje cells in the awake, behaving rabbit showed only brief pauses in the simple spike activity that followed a complex spike. While transient increases and decreases in simple spike activity did occur following the climbing fiber pause in some Purkinje cells, their functional meaning for the control of slowly changing compensatory eye movements escapes us. Perhaps their contribution is to be found in relation to rapid, voluntary eye movements associated with changes in gaze. With regard to climbing fibers and cerebellar learning, we can say from our studies in the flocculus that the error message is right, but apparently that alone is not sufficent to ensure plasticity. With regard to synchrony and rhythmicity of complex spikes, both were observed in the vestibulocerebellum. Even though the synchrony changed in relation to visual stimulation, its contribution to control of compensatory eye movements awaits elucidation in the awake animal.

In sum, the functional meaning of the climbing fiber input to the cerebellum remains an intriguing question that must be answered prior to understanding cerebellar performance. Although many of the electrophysiological characteristics of this afferent system are known, and a number of hypotheses have been developed as to its function, a consensus has not been reached and the function of the climbing fibers remains an enigma. This situation reflects, in part, our limited understanding of the various ways in which cerebellar cortical inhibitory interneurons influence the consequences of the climbing fiber signals.

Acknowledgments

This research was supported by grants from NWO (R 95-260), KNAW, and NIH (NS-13742). D.R. Wylie was supported by a postdoctoral fellowship from NSERC (Canada). We thank Evgeney Buharin and Ilan Kerman for technical assistance.

References

Aggelopoulos, N.C., Duke, C., & Edgley, S.A. (1994) Non-uniform conduction time in the olivocerebellar pathway in the anaesthetized cat. Journal of Physiology (London) 476.P: 26-27.

Albus, J. S. (1971) A theory of cerebellar function. Mathematical Bioscience 10:25-61.

Alley, K. A., Baker, R. & Simpson, J. I. (1975) Afferents to the vestibulo-cerebellum and the origin of the visual climbing fibers in the rabbit. Brain Research 98:582-589.

Andersson, G. & Armstrong, D. M. (1985) Climbing fibre input to b zone Purkinje cells during locomotor perturbation in the cat. Neuroscience Letters Supplement 22:S27

Andersson, G. & Armstrong, D. M. (1987) Complex spikes in Purkinje cells in the lateral vermis of the cat cerebellum during locomotion. Journal of Physiology (London) 385:107-134.

Andersson, G. & Oscarsson, O. (1978) Climbing fiber microzones in cerebellar vermis and their projection to different groups of cells in the lateral vestibular nucleus. Experimental Brain Research 32:565-579.

Armstrong, D. M. (1974) Functional significance of the inferior olive. Physiological Reviews 54:358-417.

Armstrong, D. M., Campbell, N. C., Edgley, S. A., Schild, R. F. & Trott, J. R. (1982) Investigations of the olivocerebellar and spino-olivary pathways. In: CerebellumNew Vistas eds. S. L. Palay, V. Chan-Palay. Springer-Verlag.

Armstrong, D. M., & Edgley, S. A. (1984) Discharges of Purkinje cells in the paravermal part of the cerebellar anterior lobe during locomotion in the cat. Journal of Physiology (London) 352:403-424.

Armstrong, D. M., Edgley, S. A. & Lidierth, M. (1988) Complex spikes in Purkinje cells of the paravermal part of the anterior lobe of the cat cerebellum during locomotion. Journal of Physiology (London) 400:405-414.

Armstrong, D. M., & Rawson, J. A. (1979) Activity patterns of cerebellar cortical neurons and climbing fibre afferents in the awake cat. Journal of Physiology (London) 289:425-448.

Balaban, C. D. & Henry, R. T. (1988) Zonal organization of olivo-nodulus projections in albino rabbits. Neuroscience Research 5:409-423.

Barmack, N. H., Fagerson M., Fredette, B. J., Mugnaini, E. & Shojaku, H. (1993a) Activity of neurons in the beta nucleus of the inferior olive of the rabbit evoked by natural vestibular stimulation. Experimental Brain Research 94:203-215.

Barmack, N. H. & Hess, D. T. (1980) Multiple-unit activity evoked in the dorsal cap of inferior olive of the rabbit by visual stimulation. Journal of Neurophysiology 43:151-164.

Barmack, N. H., Mugnaini, E. & Nelson, B. J. (1989) Vestibularly-evoked activity of single neurons in the beta nucleus of the inferior olive. In: The Olivocerebellar System in Motor Control, Experimental Brain Research Series 17, ed. P. Strata. Springer-Verlag.

Barmack, N. H. & Shojaku, H. (1992) Representation of a postural coordinate system in the nodulus of the rabbit cerebellum by vestibular climbing fiber signals. Vestibular and Brain Stem Control of Eye, Head and Body Movements, eds. H. Shimazu and Y. Shinoda, Japan Scientific Societies Press.

Barmack, N. H., Fagerson, M. & Errico, P. (1993b) Cholinergic projection to the dorsal cap of the inferior olive of the rat, rabbit, and monkey. Journal of Comparative Neurology 328:263-281.

Bell, C. C. & Grimm, R. J. (1969) Discharge properties of Purkinje cells recorded on single and double microelectrodes. Journal of Neurophysiology 32:1044-1055.

Bell, C. C. & Kawasaki, T. (1972) Relations among climbing fiber responses of nearby Purkinje cells. Journal of Neurophysiology 35:155-169.

Benedetti, F., Montarolo, P. G. & Rabacchi, S. (1984) Inferior olive lesion induces long-lasting functional modifications in the Purkinje cells. Experimental Brain Research 55:368-371.

Bloedel, J. R. (1992) Functional heterogeneity with structural homogeneity: How does the cerebellum operate? Behavioral and Brain Sciences 15:666-678.

Blodel, J. R. & Ebner, T. J. (1984) Rhythmic discharge of climbing fibre afferents in response to natural peripheral stimuli in the cat. Journal of Physiology (London) 352:129-146.

Bloedel, J. R. & Kelly, T. M. (1992) The dynamic selection hypothesis: a proposed function for the cerebellar sagittal zones. In: The Cerebellum Revisited, eds. R. Llins and C. Sotelo. Springer-Verlag.

Bloedel, J.R. & Roberts, W.J. (1971) Action of climbing fibers in cerebellar cortex of the cat. Journal of Neurophysiology. 34:17-31.

Boylls, C.C. (1980) Contributions to locomotor coordination of an olivo-cerebellar projection to vermis in the cat: experimental results and theoretical proposals. In:The Inferior Olivary Nucleus: Anatomy and Physiology eds. J. Courville, C. de Montigny & Y. Lamarre. Raven Press.

Callaway, J. C., Lasser-Ross, N. & Ross, W. N. (1995) IPSPs strongly inhibit climbing fiber-activated [Ca2+]i increases in the dendrites of cerebellar Purkinje neurons. Journal of Neuroscience 15:2777-2787.

Campbell, N. C., Ekerot, C.-F. & Hesslow, G. (1983) Interaction between responses in Purkinje cells evoked by climbing fibre impulses and parallel fibre volleys in the cat. Journal of Physiology (London) 340:225-238.

Chubb, M. C., Fuchs, A. F. & Scudder, C. A. (1984) Neuron activity in monkey vestibular nuclei during vertical stimulation and eye movements. Journal of Neurophysiology 52:724-742.

Cohen, H., Cohen, B., Raphan, T. & Waespe, W. (1992) Habituation and adaptation of the vestibuloocular reflex: a model of differential control by the vestibulocerebellum. Experimental Brain Research. 90:526-538.

Colin, F., Manil, J. & Desclin, J. C. (1980) The olivocerebellar systsm. Delayed and slow inhibitory effects: an overlooked salient feature of the cerebellar climbing fibers. Brain Research 187:3-27.

Crill, W. E. (1970) Unitary multiple-spiked responses in the cat inferior olive nucleus. Journal of Neurophysiology 33:199-209.

De Zeeuw, C.I. (1990a) Ultrastructure of the cat inferior olive. PH. D. thesis, Erasmus University Rotterdam.

De Zeeuw, C. I. & Ruigrok, T. J. H. (1994) Olivary neurons in the nucleus of Darkschewitsch in the cat receive excitatory monosynaptic input from the cerebellar nuclei. Brain Research 653:345-350.

De Zeeuw, C. I., Holstege, J. C, Ruigrok, T. J. H. & Voogd, J. (1990a) Mesodiencephalic and cerebellar terminals end upon the same dendritic spines within the glomeruli of the cat and rat inferior olive: An ultrastructural study using a combination of (3H)leucine and WGA-HRP anterograde tracing. Neuroscience. 34:645-655.

De Zeeuw, C. I., Ruigrok, T. J. H., Holstege, J. C., Jansen, H. J., & Voogd, J. (1990b) Intracellular labeling of neurons in the medial accessory olive of the cat. II. Ultrastructure of dendritic spines and their GABAergic innervation. Journal of Comparative Neurology 300:478-494.

De Zeeuw, C. I., Ruigrok, T. J. H., Holstege, J. C., Schalekamp, M. P. A., & Voogd, J. (1990c) Intracellular labeling of neurons in the medial accessory olive of the cat. III. Ultrastructure of the axon hillock and initial segment and their GABAergic innervation. Journal of Comparative Neurology 300:495-510.

De Zeeuw, C. I., Holstege, J. C., Ruigrok, T. J. H. & Voogd, J. (1989) Ultrastructural study of the GABAergic, the cerebellar, and the mesodiencephalic innervation of the cat medial accessory olive: Anterograde tracing combined with immunocytochemistry. Journal of Comparative Neurology 284:12-35.

De Zeeuw, C. I., Wentzel, P. & Mugnaini, E. (1993) Fine structure of the dorsal cap of the inferior olive and its GABAergic and non-GABAergic input from the nucleus prepositus hypoglossi in rat and rabbit. Journal of Comparative Neurology 327:63-82.

De Zeeuw, C. I., Gerrits, N. M., Voogd, J., Leonard, C. S. & Simpson, J. I. (1994a) The rostral dorsal cap and ventrolateral outgrowth of the rabbit inferior olive receive a GABAergic input from dorsal group y and the ventral dentate nucleus. Journal of Comparative Neurology 341:420-432.

De Zeeuw, C. I., Wylie, D. R., DiGiorgi, P.L., & Simpson, J. I. (1994b) Projections of individual Purkinje cells of identified zones in the flocculus to the vestiublar and cerebellar nuclei in the rabbit. Journal of Comparative Neurology 349:428-447.

De Zeeuw, C. I., Hertzberg, E. & Mugnaini, E. (1995) The dendritic lamellar body: a new neuronal organelle putatively associated with dendrodendritic gap junctions. Journal of Neuroscience 15(2):1587-1604.

De Zeeuw, C. I., Wylie, D. R., Stahl, J. S., & Simpson, J. I. Phase relations of Purkinje cells in the rabbit flocculus during compensatory eye movements. Journal of Neurophysiology (submitted).

Demer, J.L., Echelman, D.A., & Robinson, D.A. (1985) Effects of electrical stimulation and reversible lesions of the olivocerebellar pathway on Purkinje cell activity in the flocculus of the cat. Brain Research 346:22-31.

Desclin, J. C. (1974) Histological evidence supporting the inferior olive as the major source of cerebellar climbing fibers in the rat. Brain Research 77:365-384.

Dom, R., King, J. S. & Martin, G. F. (1973) Evidence for two direct cerebello-olivary connections. Brain Research 57:498-501.

Dow, R. S. (1942) The evolution and anatomy of the cerebellum. Biological Reviews 17:179-220.

Ebner, T.J. & Bloedel, J. R. (1981a) Temporal patterning in simple spike discharge of Purkinje cells and its relationship to climbing fiber activity. Journal of Neurophysiology 45:933-947.

Ebner, T.J. & Bloedel, J. R. (1981b) Role of climbing fiber afferent input in determining responsiveness of Purkinje cells to mossy fiber inputs. Journal of Neurophysiology 45:962-971.

Ebner, T.J. & Bloedel, J. R. (1984) Climbing fiber action on the responsiveness of Purkinje cells to parallel fiber inputs. Brain Research 309:1822-186.

Ebner, T. J., Yu, Q. & Bloedel, J. R. (1983) Increase in Purkinje cell gain associated with naturally activated climbing fiber input. Journal of Neurophysiology 50:205-219.

Eccles, J. C., Ito, M. & Szentagothai, J. (1967) The Cerebellum as a Neuronal Machine. Springer.

Eccles, J. C., Llins, R. & Sasaki, K. (1966) The excitatory synaptic action of climbing fibres on the Purkinje cells of the cerebellum. Journal of Physiology (London) 182:268-296.

Eccles, J. C., Sabah, N. H., Schmidt, R. F. & Tborkov, H. (1972) Cutaneous mechanoreceptors influencing impulse discharges in cerebellar cortex. III. In Purkyne cells by climbing fiber input. Experimental Brain Research 15:484-497.

Ekerot, C. F. & Kano, M. (1985) Long-term depression of parallel fibre synapses following stimulation of climbing fibres. Brain Research 342:357-360.

Ekerot, C.-F. & Oscarsson, O. (1981) Prolonged depolarization elicited in Purkinje cell dendrites by climbing fibre impulses in the cat. Journal of Physiology (London) 318:207-221.

Escudero, M., de la Cruz, R. R. & Delgado-Garcia, J. M. (1992) A physiological study of vestibular and prepositus hypoglossi neurons projecting to the abducens nucleus in the alert cat. Journal of Physiology (London) 458:539-560.

Fukuda, M., Yamamoto, T., & Llins, R. (1987) Simultaneous recordings from Purkinje cells of different folia in the rat cerebellum and their relation to movement. Society for Neuroscience Abstracts 13:603.

Fushiki, H., Sato, Y., Miura, A. & Kawasaki, T. (1994) Climbing fiber responses of Purkinje cells to retinal image movement in cat cerebellar flocculus. Journal of Neurophysiology 71:1336-1350.

Gellman, R., Gibson, A. R. & Houk, J. C. (1985) Inferior olivary neurons in the awake cat: Detection of contact and passive body displacement. Journal of Neurophysiology 54:40-60.

Gellman, R., Houk, J. C. & Gibson, A. R. (1983) Somatosensory properties of the inferior olive of the cat. Journal of Comparative Neurology 215:228-243.

Gerrits, N. M., Voogd, J. & Magras, I. N. (1985) Vestibular afferents of the inferior olive and the vestibulo-olivo-cerebellar climbing fiber pathway to the flocculus in the cat. Brain Research 332:325-336.

Ghelarducci, B., Ito, M. & Yagi, N. (1975) Impulse discharges from flocculus Purkinje cells of alert rabbit during visual stimulation combined with horizontal head rotation. Brain Research 87:66-72.

Gilbert, P. F. C. & Thach, W. T. (1977) Purkinje cell activity during motor learning. Brain Research 128:309-328.

Goodman, D. & Kelso, J. (1983) Exploring the functional significance of physiological tremor: a biospectroscopic approach. Experimental Brain Research 49:419-431.

Graf, W., Simpson, J. I. & Leonard, C. S. (1988) Spatial organization of visual messages of the rabbit's cerebellar flocculus. II. Complex and simple spike responses of Purkinje cells. Journal of Neurophysiology 60:2091-2121.

Granit, R. & Phillips, C. G. (1956) Excitatory and inhibitory processes acting upon individual Purkinje cells of the cerebellum in cats. Journal of Physiology (London) 133:520-547.

Graybiel, A. M., Nauta, H. J. W., Lasek, R. J. & Nauta, W. J. H. (1973) A cerebello-olivary pathway in the cat: An experimental study using autoradiographic tracing techniques. Brain Research 58:205-211.

Groenewegen, H. J. & Voogd, J. (1977) The parasagittal zonation within the olivocerebellar projection. I. Climbing fiber distribution in the vermis of the cat cerebellum. Journal of Comparative Neurology 174:417-488.

Groenewegen, H. J., Voogd, J. & Freedman, S. L. (1979) The parasagittal zonation within the olivocerebellar projection. II. Climbing fiber distribution in the intermediate and hemispheric parts of cat cerebellum. Journal of Comparative Neurology 183:551-602.

Horn, A. K. E. & Hoffmann, K. P. (1987) Combined GABA immunocytochemistry TMB/HRP histochemistry of pretecto nuclei projecting to the inferior olive in rats, cats, and monkeys. Brain Research 409:135-138.

Ishikawa, J., Kawaguchi, S. & Rowe, M. J. (1972) Actions of afferent impulses from muscle receptors on cerebellar Purkinje cells. II. Responses to muscle contraction. Effects mediated via the climbing fiber pathway. Experimental Brain Research 16:104-114.

Ito, M. (1970) Neurophysiological aspects of the cerebellar motor control system. International Journal of Neurology 7:162-176.

Ito, M. (1972) Neural design of the cerebellar motor control system. Brain Research 40:81-84.

Ito, M. (1982) Cerebellar control of the vestibulo-ocular reflex-around the flocculus hypothesis. Annual Review of Neuroscience 5:275-296.

Ito, M. (1984) The Cerebellum and Neuronal Control. Raven Press.

Ito, M. (1989) Long-term depression. Annual Review of Neuroscience 12:85-102.

Ito, M. & Simpson, J. (1971) Discharges in Purkinje cell axons during climbing fibre activation. Brain Research 31:215-219.

Ito, M., Nisimaru, N. & Yamamoto, M. (1977) Specific patterns of neuronal connections involved in the control of the rabbit's vestibulo-ocular reflexes by the cerebellar flocculus. Journal of Physiology (London) 265:833-854.

Ito, M., Sakurai, M. & Tongroach, P. (1982a) Climbing fibre induced depression of both mossy fibre responsiveness and glutamate sensitivity of cerebellar Purkinje cells. Journal of Physiology (London) 324:113-134.

Ito, M., Jastreboff, P. J. and Miyashita, Y. (1982b) Specific effects of unilateral lesions in the flocculus upon eye movements in albino rabbits. Experimental Brain Research 45:233-242.

Jahnsen, H. (1986) Extracellular activation and membrane conductances of neurones in the guinea-pig deep cerebellar nuclei in vitro. Journal of Physiology (London) 372:149-168.

Jansen, J. & Brodal, A. (1954) Aspects of Cerebellar Anatomy. Oslo: Johan Grundt Tanum Forlag.

Kano, M. & Kato, M. (1988) Mode of induction of long-term depression at parallel fibre-Purkinje cell synapses in rabbit cerebellar cortex. Neuroscience Research 5:544-556.

Kano, M., Kano, M.-S., Kusunoki, M. & Maekawa, K. (1990a) Nature of optokinetic response and zonal organization of climbing fiber afferents in the vestibulocerebellum of the pigmented rabbit. II. The nodulus. Experimental Brain Research 80:238-251.

Kano, M., Rexhausen, U., Dreessen, J., & Konnerth, A. (1992) Synaptic excitation produces a long-lasting rebound potentiation of inhibitory synaptic signals in cerebellar Purkinje cells. Nature (London) 356:601-604.

Kano, M-S., Kano, M. & Maekawa, K. (1990b) Receptive field organization of climbing fiber afferents responding to optokinetic stimulation in the cerebellar nodulus and flocculus of the pigmented rabbit. Experimental Brain Research 82:499-512.

Katayama, S. & Nisimaru, N. (1988) Parasagittal zonal pattern of olivo-nodular projections in rabbit cerebellum. Neuroscience Research 5:424-438.

Kawato, M. & Gomi, H. (1992) The cerebellum and VOR/OKR learning models. Trends in Neuroscience 15:445-453.

Keating, J. G. & Thach, W. T. (1993) Complex spike activity in the awake behaving monkey: non-clock-like discharge. Society of Neuroscience Abstracts 19:980.

Kim, J. H., Wang, J.-J. & Ebner, T. J. (1987) Climbing fiber afferent modulation during treadmill locomotion in the cat. Journal of Neurophysiology 57:787-802.

Kim, J. H., Wang, J.-J. & Ebner, T. J. (1988) Alterations in simple spike activity and locomotor behavior associated with climbing fiber input to Purkinje cells in a decerebrate walking cat. Neuroscience 25:475-489.

Kolb, F. P. & Rubia, F. J. (1980) Information about peripheral events conveyed to the cerebellum via the climbing fiber system in the decerebrate cat. Experimental Brain Research 38:363-373.

Kooy, F. H. (1916) The inferior olive in vertebrates. Thesis.Haarlem: Bohn.

Kusunoki, M., Kano, M., Kano, M.-S. & Maekawa, K. (1990) Nature of optokinetic response and zonal organization of climbing fiber afferents in the vestibulo-cerebellum of the pigmented rabbit. I. The flocculus. Experimental Brain Research 80:225-237.

Lamarre, Y., Montigny, C., Dumont, M. & Weiss, M. (1971) Harmaline-induced rhythm activity of cerebellar and lower brain stem neurons. Brain Research 32:246-250.

Lang, E. J. (1995) Synchronicity, rhythmicity, and movement: the role of the olivocerebellar system in motor coordination. PhD. thesis New York University.

Lang, E. J., Sugihara, I., & Llins, R. (1989) Intraolivary injection of picrotoxin causes reorganization of complex spike activity. Society of Neuroscience Abstracts 15:77.5.

Lang, E. J., Sugihara, I., & Llins, R. (1990) Lesions of the cerebellar nuclei, but not of the mesencephalic structures, alters the spatial pattern of complex spike synchronicity as demonstrated by multiple electrode recordings. Society Neuroscience Abstracts 16:370.3.

Lang, E. J., Sugihara, I., & Llins, R. (1992) The ability of motor cortex stimulation to evoke vibrissal movements is modulated by a 10 Hz signal arising in the inferior olive. Society of Neuroscience Abstracts 18:178.6.

Leicht, R., Rowe, M. J. & Schmidt, R. F. (1977) Mossy and climbing fiber inputs from cutaneous mechanoreceptors to cerebellar Purkyne cells in unanesthetized cats. Experimental Brain Research 27:459-477.

Leonard, C. S. (1986) Signal Characteristics of Cerebellar Purkinje Cells in the Rabbit Flocculus during Compensatory Eye Movements. Ph.D. Thesis, New York University.

Leonard, C. S. & Simpson, J. I. (1986) Simple spike modulation of floccular Purjinje cells during the reversible blockade of their climbing fiber afferents. In: Adaptive processes in visual and oculomotor systems., ed. E. Keller & D. Zee. Pergamon.

Leonard, C. S., Simpson, J. I. & Graf, W. (1988) Spatial organization of visual messages of the rabbit's cerebellar flocculus. I. Typology of inferior olive neurons of the dorsal cap of Kooy. Journal of Neurophysiology 60:2073-2090.

Lisberger, S. G. (1988) The neuronal basis for learning of simple motor skills. Science Wash. DC 242:728-735.

Lisberger, S. G. & Pavelko, T. A. (1988) Brain stem neurons in modified pathways for motor learning in the primate vestibulo-ocular reflex. Science Wash. DC 242:771-773.

Lisberger, S. G. & Sejnowski, T. J. (1992) Motor learning in a recurrent network model based on the vestibulo-ocular reflex. Nature (London) 360:159-161.

Llins, R. (1970) Neuronal operations in cerebellar transactions. In: The Neurosciences: Second Study Program, ed. F.O. Schmitt. Rockefeller University Press.

Llins, R. (1974) Eighteenth Bowditch lecture: Motor aspects of cerebellar control. Physiologist 17:19-46.

Llins, R. (1985) Functional significance of the basic cerebellar circuit in motor coordination. In: Cerebellar Functions, eds. J.R. Bloedel, J. Dichgans, and W. Precht, Springer-Verlag.

Llins, R. (1991) The noncontinuous nature of movement execution. In: Motor Control: Concepts and Issues , eds. D.R. Humphrey & H.-J. Freund. John Wiley.

Llins, R., Baker, R. & Sotelo, C. (1974) Electrotonic coupling between neurons in cat inferior olive. Journal of Neurophysiology 37:560-571.

Llins, R. & Mhlethaler, M. (1988a) An electrophysiological study of the In Vitro, perfused brain stem-cerebellum of adult guinea-pig. Journal of Physiology (London) 404:215-240.

Llins, R. & Mhlethaler, M. (1988b) Electrophysiology of guinea-pig cerebellar nuclear cells in the In Vitro brain stem-cerebellar preparation. Journal of Physiology (London) 404:241-258.

Llins, R. & Sasaki, K. (1989) The functional organization of the olivo-cerebellar system as examined by multiple Purkinje cell recordings. European Journal of Neuroscience 1:587-602.

Llins, R. & Sugimori, M. (1980) Electrophysiological properties of in vitro Purkinje cell dendrites in mammalian cerebellar slices. Journal of Physiology (London) 305:197-213.

Llins, R. & Volkind, R. A. (1973) The olivo-cerebellar system: Functional properties as revealed by harmaline-induced tremor. Experimental Brain Research 18:69-87.

Llins, R., & Welsh, J. P. (1993) On the cerebellum and motor learning. Current Opinion in Neurobiology 3:958-965.

Llins, R. & Yarom, Y. (1981a) Electrophysiology of mammalian inferior olivary neurones in vitro. Different types of voltage-dependent ionic conductances. Journal of Physiology (London) 315:549-567.

Llins, R. & Yarom, Y. (1981b) Properties and distribution of ionic conductances generating electroresponsiveness of mammalian inferior olivary neurones In Vitro . Journal of Physiology (London) 315:569-584.

Llins, R. & Yarom, Y. (1986) Oscillatory properties of guinea-pig inferior olivary neurones and their pharmacological modulation: An In Vitro study. Journal of Physiology (London) 376:163-182.

Lou, J-S. & Bloedel, J. R. (1986) The responses of simultaneously recorded Purkinje cells to the perturbations of the step cycle in the walking ferret: a study using a new analytical method-the real time post synaptic response (RTPR). Brain Research 365:340-344.

Lou, J-S. & Bloedel, J. R. (1992a) Responses of sagittally aligned Purkinje cells during perturbed locomotion: synchronous activation of climbing fiber inputs. Journal of Neurophysiology 68:570-580.

Lou, J-S. & Bloedel, J. R. (1992b) Responses of sagittally aligned Purkinje cells durng perturbed locomotion: relation of climbing fiber activation to simple spike modulation. Journal of Neurophysiology 68:1820-1833.

Lopez-Berneo, J., Darlot, C., Berthoz, A. & Baker, R. (1982) Neuronal activity in prepositus nucleus correlated with eye movement in the alert cat. Journal of Neurophysiology 47:329-352.

Maekawa, K. & Simpson, J. I. (1973) Climbing fiber responses evoked in vestibulo-cerebellum of rabbit from visual system. Journal of Neurophysiology 36:649-666.

Mano, N., Kanazawa I. & Yamamoto, K. (1986) Complex-spike activity of cerebellar Purkinje cells related to wrist tracking movement in monkey. Journal of Neurophysiology 56:137-158.

Mano, N., Kanazawa I. & Yamamoto, K. (1989) Voluntary movements and complex-spike discharges of cerebellar Purkinje cells. In: The Olivocerebellar System in Motor Control, Experimental Brain Research Series 17, ed. P. Strata. Springer-Verlag.

Marr, D. (1969) A theory of cerebellar cortex. Journal of Physiology (London), 202:437-470.

McCollum, G. (1992) Rules of combination that generate climbing fiber tactile receptive fields. Neuroscience 50,3:707-725.

McCrea, R. A. & Baker, R. (1985). Anatomical connections of the nucleus prepositus of the cat. Journal of Comparative Neurology 237:377-407.

McDevitt, C. J., Ebner, T. J. & Bloedel, J. R. (1982) The changes in Purkinje cell simple spike activity following spontaneous climbing fiber inputs. Brain Research 237:484-491.

McFarland, J. L. & Fuchs, A. F. (1992) Discharge patterns in nucleus prepositus hyposglossi and adjacent medial vestibular nucleus during horizontal eye movement in behaving macaques. Journal of Neurophysiology 68:319-332.

Miles, F. A., Braitman, D. J. & Dow, B. M. (1980) Long-term adaptive changes in primate vestibuloocular reflex. IV. Electrophysiological observations in flocculus of adapted monkeys. Journal of Neurophysiology 43:1477-1493.

Miles, F. A. & Lisberger, S. G. (1981) Plasticity in the vestibulo-ocular reflex: a new hypothesis. Annual Review of Neuroscience 4:273-299.

Miyashita, E. & Tamai, Y. (1989) Subcortical connections to frontal "oculomotor" areas in the cat. Brain Research 502:75-87.

Montarolo, P. G., Palestini, M. & Strata, P. (1982) The inhibitory effect of the olivocerebellar input on the coerebellar Purkinje cells in the rat. Journal of Physiology (London) 332:187-202.

Montigny, C. de, & Lamarre, Y. (1973) Rhythmic activity induced by harmaline in the olivo-cerebello-bulbar system of the cat. Brain Research 53:81-95.

Murphy, J. T. & Sabah, N. H. (1970) The inhibitory effect of climbing fiber activation and cerebellar Purkinje cells. Brain Research 19:486-490.

Nagao, S. (1983) Effects of vestibulo-cerebellar lesions upon dynamic characteristics and adaptation of vestibulo-ocular and optokinetic responses in pigmented rabbits. Experimental Brain Research 53:36-46.

Nagao, S. (1988) Behavior of floccular Purkinje cells correlated with adaptation of horizontal optokinetic eye movement response in pigmented rabbits. Experimental Brain Research 73:489-497.

Nelson, B. J. & Mugnaini, E. (1989) Origins of GABAergic inputs to the inferior olive. In: The Olivocerebellar System in Motor Control, Experimental Brain Research Series 17, ed. P. Strata. Springer-Verlag.

Nues-Cardozo, B. & Van der Want, J. J. L. (1990) Ultrastructural organization of the retinal-pretecto-olivary pathway. A combined WGA-HRP retrograde/GABA immunohistochemical study. Journal of Comparative Neurology 291:313-327.

Ojakangas, C. L. & Ebner, T. (1992) Purkinje cell complex spike changes during a voluntary arm movement learning task in the monkey. Journal of Neurophysiology 68:2222-2236.

Ojakangas, C. L. & Ebner, T. J. (1994) Purkinje cell complex spike activity during voluntary motor learning: relationship to kinematics. Journal of Neurophysiology 72:2617-2630.

Oscarsson, O. (1969) The sagittal organization of the cerebellar anterior lobe as revealed by the projection patterns of the climbing fiber system. In: Neurobiology of Cerebellar Evolution and Development, ed. R. Llins, AMA.

Oscarsson, O. (1973) Functional organization of spinocerebellar paths. In: Handbook of Sensory Physiology, Vol. II, ed. A. Iggo. Springer-Verlag.

Oscarsson, O. (1979) Functional units of the cerebellum-sagittal zones and microzones. Trends in Neuroscience 2:143-145.

Oscarsson, O. (1980) Functional organization of olivary projection to the cerebellar anterior lobe. In: The Inferior Olivary Nucleus: Anatomy and Physiology, eds. Courville, J., DeMontigny, C., and Lamarre, Y. New York: Raven Press.

Partsalis, A. M., Zhang, Y. & Highstein, S. M. (1993) The y group in vertical visual-vestibular interactions and VOR adaptation in the squirrel monkey. Society of Neuroscience Abstracts 19:138.

Porter, C. M., Van Kan, P. L. E., Horn, K. M., Bloedel, J. R. & Gibson, A. R. (1993) Functional divisions of cat rMAO. Society for Neuroscience Abstracts 19:1216.

Rawson, J.A. & Tilokskulchai (1981) Suppression of simple spike discharges of cerebellar Purkinje cells by impulses in climbing fibre afferents. Neuroscience Letters 25:125-130.

Rawson, J.A. & Tilokskulchai (1982) Climbing fibre modification of cerebellar Purkinje cell responses to parallel fibre inputs. Brain Research 237:492-497.

Robertson, L.T. (1984) Topographic features of climbing fiber input in the rostral vermal cortex of the cat cerebellum. Experimental Brain Research 55:445-454.

Robertson, L. T. & Laxer, H. D. (1981) Localization of cutaneously elicited climbing fibre responses in lobule V of the monkey cerebellum. Brain Behavior and Evolution 18:157-168.

Robertson, L. T., Laxer, K. D. & Rushmer, D. S. (1982) Organization of climbing fiber input from mechanoreceptors to lobule V vermal cortex of the cat. Experimental Brain Research 46:281-291.

Robertson, L.T. & McCollum, G. (1989) Ensembles of climbing fiber tactile receptive fields encode distinct information for various cerebellar regions. In: The Olivocerebellar System in Motor Control, Experimental Brain Research Series 17, ed. P. Strata. Springer-Verlag.

Robinson, D. A. (1963) A method of measuring eye movement using a scleral search coil in a magnetic field. Transactions on Bio-Medical Electronics BME-10:137-145.

Rubia, F. J. & Kolb, F. P. (1978) Responses of cerebellar units to a passive movement in the decerebrate cat. Experimental Brain Research 31:387-401.

Rushmer, D. S., Roberts, W. J. & Augter, G. K. (1976) Climbing fiber responses of cerebellar Purkinje cells to passive movement of the cat forepaw. Brain Research 106:1-20.

Sakurai, M. (1987) Synaptic modification of parallel fibre-Purkinje cell transmission in in vitro guinea pig cerebellar slices. Journal of Physiology (London ) 394:463-480.

Sakurai, M. (1990) Calcium is an intracellular mediator of the climbing fiber induction of cerebellar long-term depression. Proceedings of the National Academy of Science 87:3383-3385.

Sasaki, K., Bower, J. M. & Llins, R. (1989) Multiple Purkinje cell recording in rodent cerebellar cortex. European Journal of Neuroscience 1:572-586.

Sato, Y., Miura, A., Fushiki, H. & Kawasaki, T. (1992) Short-term modulation of cerebellar Purkinje cell activity after spontaneous climbing fiber input. Journal of Neurophysiology 68,6:2051-2062.

Sato, Y., Miura, A., Fushiki, H. & Kawasaki, T. (1993) Barbiturate depresses simple spike activity of cerebellar Purkinje cells after climbing fiber input. Journal of Neurophysiology 69:1082-1090.

Schreurs, B. G. & Alkon, D. L. (1993) Rabbit cerebellar slice analysis of long-term depression and its role in classical conditioning. Brain Research 631:235-240.

Schulmann, J.A. & Bloom, F.E. (1981) Golgi cells of the cerebellum are inhibited by inferior olive activity. Brain Research 210:350-355.

Shojaku, H., Barmack, N. H. & Mizukoshi, K. (1991) Influence of vestibular and visual climbing fiber signals on Purkinje cell discharge in the cerebellar nodulus of the rabbit. Acta Otolaryngologica 481:242-246.

Simpson, J. I. & Alley, K. E. (1974) Visual climbing fiber input to rabbit vestibulocerebellum: A source of direction-specific information. Brain Research 82:302-308.

Simpson, J. I., Graf, W. & Leonard, C. (1981) The coordinate system of visual climbing fibers to the flocculus. In: Progress in Oculomotor Research, Developments in Neuroscience, vol. 12, eds. A. Fuchs & W. Becker, Elsevier/North-Holland.

Simpson, J. I., Leonard, C. S. & Soodak, R. E. (1988) The accessory optic system of rabbit. II. Spatial organization of direction selectivity. Journal of Neurophysiology 60:2055-2072.

Soodak, R. E. & Simpson, J. I. (1988) The accessory system of rabbit. I. Basic visual response properties. Journal of Neurophysiology 60:2037-2054.

Soodak, R. E., Croner, L. J., & Graf, W. (1988) Development of the optokinetic reference frame of floccular Purkinje cells in rabbit. Society for Neuroscience Abstracts 14:758.

Sotelo, C., Gotow, T. & Wassef, M. (1986) Localization of glutamic-acid-decarboxylase-immunoreactive axon terminals in the inferior olive of the rat, with special emphasis on anatomical relations between GABAergic synapses and dendrodendritic gap junctions. Journal of Comparative Neurology 252:32-50.

Sotelo, C., Llins, R. & Baker, R. (1974) Structural study of the inferior olivary nucleus of the cat. Morphological correlates of electrotonic coupling. Journal of Neurophysiology 37:541-559.

Stahl, J. S. & Simpson, J. I. (1995) Dynamics of rabbit vestibular nucleus neurons and the influence of the flocculus. Journal of Neurophysiology 73:1396-1413.

Stone, L. S. & Lisberger, S. G. (1990) Visual responses of Purkinje cells in the cerebellar flocculus during smooth-pursuit eye movements in monkeys. II. Complex spikes Journal of Neurophysiology 63:1262-1275.

Sugihara, I., Lang, E. J. & Llins, R. (1993) Uniform olivocerebellar conduction time underlies Purkinje cell complex spike synchronicity in the rat cerebellum. Journal of Physiology (London) 470:243-271.

Szentgothai, J. & Rajkovits, K. (1959) Uber den Urspring der Kletterfasern des Kleinjirns. Zeitschrift fr Anatomie und Entwicklungsgeschichte 121:130-141.

Tan, J., Gerrits, N. M., Nanhoe, R.S., Simpson, J. I. & Voogd, J. (1995a) Zonal organization of the climbing fiber projection to the flocculus and nodulus of the rabbit. A combined axonal tracing and acetylcholinesterase histochemical study. Journal of Comparative Neurology 356:1-22.

Tan, J., Simpson, J. I., & Voogd, J. (1995b) Anatomical compartments in the white matter of the rabbit flocculus. Journal of Comparative Neurology 356:23-50.

Tempia, F., Dieringer, N., & Strata, P. (1991) Adaptation and habituation of the vestibulo-ocular reflex in intact and inferior olive-lesioned rats. Experimental Brain Research 86:568-578.

Thach, W. T., Jr. (1967) Somatosensory receptive fields of single units in the cat cerebellar cortex. Journal of Neurophysiology 30:675-696.

Thach, W. T., Jr. (1968) Discharge of Purkinje and cerebellar nuclear neurons during rapidly alternating arm movement in the monkey. Journal of Neurophysiology 31:785-797.

Thach, W. T., Jr. (1970) Discharge of cerebellar neurons related to two maintained postures and two prompt movements. II. Purkinje cell output and input. Journal of Neurophsyiology 33:537-547.

Thach, W. T., Goodkin, H. P. & Keating, J. G. (1992) The cerebellum and the adaptive coordination of movement. Annual Review of Neuroscience 15:403-442.

Thompson, R. F. (1986) The neurobiology of learning and memory. Science 223:941-947.

Van der Steen, J., Simpson, J. I., & Tan, J. (1994) Functional and anatomical organization of three-dimensional eye movements in rabbit cerebellar flocculus. Journal of Neurophysiology 72:31-46.

Van der Want, J. J. L., Wiklund, L., Guegan, M., Ruigrok, T. & Voogd, J. (1989) Anterograde tracing of the rat olivocerebellar system with phaseolus vulgaris leucoagglutinin (PHA-L). Demonstration of climbing fiber collateral innervation of the cerebellar nuclei. Journal of Comparative Neurology 288:1-18.

Voogd, J. (1964) The Cerebellum of the Cat. Structure and Fibre Connections. Thesis. Van Gorcum, Assen, The Netherlands.

Voogd, J. & Bigar, F. (1980) Topographical distribution of olivary and cortico-nuclear fibers in the cerebellum: a review. In: The Inferior Olivary Nucleus: Anatomy and Physiology , eds. J. Courville, C. de Montigny, Y. Lamarre, Raven.

Waespe, W., Cohen, B. & Raphan, T. (1983) Role of the flocculus and paraflocculus in optokinetic nystagmus and visual-vestibular interactions: effects of lesions. Experimental Brain Research. 50:9-33.

Waespe, W., Cohen, B. & Raphan, T. (1985) Dynamic moduification of the vestibulo-ocular reflex by the nodulus and the uvula. Science Wash. D.C. 228:199-202.

Wang, J.-J., Kim, J. H. & Ebner, T. J. (1987) Climbing fiber afferent modulation during a visually guided, multi-joint arm movement in the monkey. Brain Research 410:323-329.

Welsh, J. P. & Harvey, J. A. (1989) Cerebellar lesions and the nicitating membrane reflex: performance deficits of the conditioned and unconditioned response. Journal of Neuroscience 9:299-311.

Welsh, J. P., Lang, E. J., & Llins, R. (1993) The microstructure of coherence in the olivocerebellar system during rhythmic movement in normal and deafferented rats. Society of Neuroscience Abstracts 19: 529.9.

Welsh, J. P., Lang, E. J., Sugihara, I. & Llins, R. (1992) Rhythmic olivo-cerebellar control of skilled tongue movement in relation to patterned hypoglossal nerve activity. Society of Neuroscience Abstracts 18:178.7.

Welsh, J. P., Lang, E. J., Sugihara, I. & Llins, R. (1995) Dynamic organization of motor control within the olivocerebellar system. Nature (London) 374:453-457.

Wylie, D. R. & Frost, B. J. (1993) Responses of pigeon vestibulocerebellar neurons to optokinetic stimulation. II. The 3-dimensional reference frame of rotation neurons in the flocculus. Journal of Neurophysiology 70:2647-2659.

Wylie, D. R., De Zeeuw, C. I., DiGiorgi, P. L., & Simpson, J. I. (1994) Projections of individual Purkinje cells of identified zones in the ventral nodulus to the vestibular and cerebellar nuclei in the rabbit. Journal of Comparative Neurology 349:448-463.

Wylie, D. R., De Zeeuw, C. I. & Simpson, J. I. (1995) Temporal relations of the complex spike activity of Purkinje cell pairs in the vestibulocerebellum of rabbits. The Journal of Neuroscience 15:2875-2887.

Yamamoto, C., Yamashita, H. & Chujo, T. (1978) Inhibitory action of glutamic acid on cerebellar interneurons. Nature (London) 262:786-787.

Yamamoto, M. (1979) Vestibulo-ocular reflex pathways of rabbits and their representation in the cerebellar flocculus. Progress in Brain Research 50:451-457.

[Note: Figures avaiable only in hard copy version]

Figure Captions

Figure 1. The cerebellar modules of the cat. A shows a schematic diagram of the longitudinal zonal organization of the inferior olivary projection to the cerebellar cortex (top) and a representation of a horizontal section through the deep cerebellar nuclei (fastigial, anterior and posterior interposed, and lateral) and the lateral vestibular (Deiters') nucleus (bottom). B shows a horizontal reconstruction of the inferior olive. Stereotactic levels are indicated to the left (level 14 is caudal; level 9 is rostral). Corresponding symbols in A and B indicate reciprocal connections between the inferior olive and the cerebellar nuclei . Likewise, each cortical zone receives from the correspondingly marked subdivision of the inferior olive. The corticonuclear relations are indicated with arrows (i.e. Purkinje cells of Zone A project to the fastigial nucleus; Purkinje cells of Zone B project to Deiters' nucleus, etc.). Some of the finer subdivisions of the olivo-cerebellar projection, such as that to the flocculus (see Figure 7), are not illustrated. ANSI-ansiform lobule; -subnucleus beta; D-dentate (lateral cerebellar) nucleus; DAO-dorsal accessory olive; DC-dorsal cap of Kooy; Dei-Deiters' nucleus (lateral vestibular nucleus); dl-dorsal lamella of the principal olive; dmcc-dorsomedial cell column; IP-posterior interposed nucleus; MAO-medial accessory olive; PFLD-dorsal paraflocculus; PFLV-ventral paraflocculus; PMD-paramedial lobule; PO-principal olive; SI-simple lobule; vl-ventral lamella of the principal olive; vlo-ventrolateral outgrowth; A,B,C1,C2,C3 and D, designations of the longitudinal cortical zones; c-caudal, r-rostral (from Groenewegen et al. 1979).

Figure 2. Responses of the complex spikes of floccular Purkinje cells to optokinetic stimulation. These panels provide an overview of the spatial organization of the preferred and null axes characterizing the rotation selectivity of the visual climbing fiber responses for each of the three principal classes (A, B, and C) of Purkinje cells in the rabbit's cerebellar flocculus. A "planetarium" projected stimulus was rotated about each of three orthogonal axes. The peristimulus histogram (15 repetitions, 100 ms binwidth) next to each axis depicts the firing rate during clockwise (CW) and counterclockwise (CCW) constant speed (0.5/s) rotation. CW and CCW indicate the sense of visual world rotation when viewed from the rabbit along the axis that is directed toward the respective histogram. The ordinate scale on the vertical axis histogram applies to all histograms. The Purkinje cells were classified according to which one of the three axes of rotation produced the deepest modulation and according to which eye was dominant. The numbers on the equator of the reference sphere enclosing the rabbit's head indicate the ipsilateral and contralateral azimuth angles. The neuron type shown in A responds best to rotation about the vertical axis (VA neurons). The neuron types shown in B and C both respond best to rotation about a horizontal axis oriented at 45 contralateral azimuth/135 ipsilateral azimuth. They differ with respect to ocular dominance: the neuron in B is ipsilateral dominant (posterior (135) axis or ipsi-135 neuron) whereas the neuron in C is contralateral dominant (anterior (45) axis or contra-45 neuron). D shows the distribution of the preferred axes for dominant eye stimulation for 10 anterior (45) axis Purkinje cells and 10 posterior (135) axis Purkinje cells. For each cell the orientation of the preferred axis was determined by fitting azimuthal tuning data with a sine curve. The line indicating the preferred axis orientation is drawn from the dominant eye. The solid triangle indicates that the recordings were made from the left flocculus, as was also the case for A, B, and C . The spatial organization of climbing fiber responses to rotation is similar to that of the vestibular semicircular canals and eye muscles. That is, the VA neurons respond best to rotation about an axis that is approximately perpendicular to the horizontal vestibular canals and the plane of the horizontal recti. Likewise, the ipsi-135 and contra-45 neurons repond best to rotation about an axis that is approximately perpendicular to the ipsilateral anterior canal (and contralateral posterior canal), and the plane of the ipsilateral vertical recti (and contralateral obliques) (adapted from Graf et al. 1988).

Figure 3. Neuronal activity in the -nucleus evoked by vestibular stimulation. A shows direction-selective modulation of a -nucleus neuron in response to sinusoidal vestibular stimulation about the longitudinal axis. The lower trace indicates position of the head and body with an upward deflection indicating right side up. This neuron increased its activity when the rabbit was rotated onto its left side. B and C show peristimulus time histograms (compiled from 40 cycles) of the activity of the same olivary neuron in response to rotation about the longitudinal axis (see lower trace) at two frequencies, 0.05 Hz (B) and 0.20 Hz (C). Each bin represents 2. D and E show the distribution of the "null axes" of -nucleus neurons. These axes were determined by sinusoidally rotating the animal about a horizontal axis and shifting the head about the vertical axis, either clockwise or counterclockwise, until the neuron was no longer modulated. The "null axes" clustered around two loci that correspond to optimal orientations of pairs of vertical semicircular canals. The null axes in D are consistent with vestibular modulation originating from neurons with right posterior-left anterior semicircular canal sensitivity. These neurons were clustered in the rostral -nucleus on the right side. The null axes in E are consistent with vestibular modulation originating from neurons with right anterior-left posterior semicircular canal sensitivity. These neurons were clustered in the caudal -nucleus on the right side (adapted from Barmack et al. 1993a).

Figure 4. Changes in simple spike discharge rates following spontaneous climbing fiber discharges of Purkinje cells in unanesthetized, decerebrate cats. A-D show post-stimulus time histograms (PSTH) of simple spike (SS) activity triggered by 100 spontaneous climbing fiber inputs, time zero (t = 0) indicating the time at which the triggering complex spike occurred (bin width = 1 ms). Each histogram was obtained from a different neuron. E shows a plot of the number of Purkinje cells with the indicated ratio of the average simple spike rate in the period from 10 to 50 ms following the complex spike trigger to the average simple spike rate in the period from 400 to 440 ms (from McDevitt et al., 1982).

Figure 5. Effect of lidocaine block of the inferior olive on the simple spike activity of floccular Purkinje cells in the rabbit. A shows the effect on spontaneous simple spike activity produced by blocking the complex spike. The arrows indicate the times of the lidocaine injections into the dorsal cap. The simple spike firing rate gradually increased by 55% following blockade of the complex spikes. B and C show two examples of visually modulated VA Purkinje cells in the presence and absence of complex spike modulation. The optokinetic stimulation was produced by rear-projecting a random dot pattern (70x70) onto a tangent screen centered on the optic axis of the ipsilateral eye. The CONTROL histograms were compiled from 10 stimulus cycles before lidocaine injections and the LIDOCAINE histograms were compiled from 10 stimulus cycles after the complex spike activity was suppressed. Both cells have stereotypical simple spike modulation (simple spikes and complex spikes are reciprocal) whose absolute value is unaffected by silencing the complex spike modulation even though the simple spike spontaneous rate increased. Abbreviations: SS, simple spikes; CS, complex spikes; t-n, temporal to nasal (in reference to the ipsilateral eye); n-t, nasal to temporal (from Leonard, 1986; see also Leonard & Simpson, 1986).

Figure 6. Synchronous and rhythmic complex spike activity of Purkinje cells. A-D are adapted from Sasaki et al., 1989. A-C show the auto- and cross-correlograms of complex spike activity of three cells recorded simultaneously from the superficial vermis of the rat. The relative location of these cells is shown in D. The cell at position A is the "master cell" (M), which is located 500 m rostral to the neuron at position B and 500m lateral to the neuron at position C. The area of the dots at positions B and C represents the degree of cross-correlation between each cell and the master cell. A shows the auto-correlogram of the complex spike activity recorded at position A. As indicated by the two peaks occurring at about 120 and 240 msec, this cell had a characteristic frequency of about 8 Hz. B shows the cross-correlogram of the two sagittally aligned cells, located at positions A and B. The presence of the time-zero peak indicates that these two cells tended to fire within the same msec (see inset in B). In contrast, the cross-correlogram for the cells at positions A and C, shown in C, does not have a peak. E and F show examples of synchrony and rhythmicity in the ventral nodulus of the anesthetized rabbit (Wylie et al., 1995). E shows the cross-correlogram between two VA1 zone neurons that have a synchronous temporal relationship. F shows an auto-correlogram of a third VA1 neuron with 3 to 4 peaks spaced about 85 msec apart, indicating a characteristic frequency of about 12 Hz.

Figure 7. Input-output relations between retinal image motion and the floccular modular control of extraocular muscles in the rabbit. Signals from retinal ganglion cells are processed in various parts of the accessory optic system before being differentially distributed to the inferior olive. Each panel (A, B, and C) relates to one of the three classes of olivary neurons that are distinguished on the basis of the orientation of their preferred axis for responding to visual world rotation and their ocular dominance (see Figure 2). Each class is located within a particular part of the dorsal cap and ventrolateral outgrowth and gives rise to a specific group of climbing fibers that traverse particular acetylcholinesterase delineated white matter compartments to innervate the Purkinje cells of particular zones. Two zones receive their input from the caudal dorsal cap and two other zones receive their input from the rostral dorsal cap and ventrolateral outgrowth. Note that the 135 and 45 classes of climbing fibers both project to the same two floccular zones. The Purkinje cells in the floccular zones, in turn, have a differential projection to the vestibular complex from which particular sets of extraocular muscles are controlled. Activation of these sets of muscles causes eye rotation about axes that are spatially close to the preferred axes of the visual climbing fibers. AOS, accessory optic system; DTN, LTN, MTN, dorsal, lateral, and medial terminal nucleus; VTRZ, visual tegmental relay zone (based on Ito et al., 1977; Yamamoto, 1979; Leonard et al., 1988; Graf et al., 1988; Soodak & Simpson, 1988; Simpson et al., 1988; Van der Steen et al., 1994; Tan et al., 1995a, b; De Zeeuw et al., 1994b)

Figure 8. Neural activity of a "pure pause" floccular Purkinje cell recorded in the awake rabbit during sinusoidal vestibular stimulation in the light at 0.05 Hz. A shows the complex and simple spike PSTHs compiled over 6 cycles (binwidth 100 msec) and averages of the turntable and (inverted) eye positions. Note the reciprocity in the complex and simple spike modulation. The complex spike activity increased during rotation toward the side of recording. B shows the complex spike-simple spike cross-correlogram (1 msec bins) compiled from the same 6 cycles. After a brief pause, the simple spike activity rapidly returned to an average rate that was not significantly different from the average rate during the 100 msec pre-complex spike period.

Figure 9. Neural activity of a "pause-facilitation" floccular Purkinje cell recorded in the awake rabbit during sinusoidal vestibular stimulation in the dark at 0.05 Hz. A shows the complex and simple spike PSTHs compiled over 6 cycles (binwidth 100 msec) and averages of the turntable and (inverted) eye positions. The simple spike activity was modulated, but the complex spike activity was not. B shows the complex spike-simple spike cross-correlogram (1 msec bins) compiled from the same 6 cycles. After a brief pause, the simple spike activity rapidly rose to a level that for a short time was higher than the remainder of the analysis period. For each 20 msec period that began at the times between the two lines on the time axis of the correlogram, the average simple spike rate was significantly greater (p<0.01) than the average rate during the 100 msec pre-complex spike period.

Figure 10. Synchronous complex spike activity recorded from a pair of floccular Purkinje cells in the awake rabbit. Both of these Purkinje cells received visual climbing fibers of the contra-45 axis class. Panel A shows the average planetarium position, the average vertical position component of the compensatory eye movement, and the peristimulus time histograms (PSTHs; 50 msec bins) for the complex spike activity of the two Purkinje cells cumulated over 40 cycles. A clear synchronous relationship between the firing of the complex spikes of these two Purkinje cells is seen in B as a peak in their cross-correlogram.