| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Journal of Physiology (2001), 536.3, pp. 825-839
© Copyright 2001 The Physiological Society
 |
ABSTRACT |
|---|
 TopAbstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
From the earliest functional studies, investigation of the information carried by cerebellar climbing fibres has focused on their powerful activation by somatosensory inputs (Adrian, 1943; Snider & Stowell, 1944). This line of investigation has led to the current picture of a modular organisation of the cerebellar circuitry. The origins of this view came partly from functional studies that showed a specific and detailed organisation of spino-olivocerebellar pathways projecting to precise parasagittal zones of the cerebellar cortex (Oscarsson, 1980), which have turned out to be composed of very precisely organised microzones (Garwicz et al. 1998). In parallel, anatomical studies have revealed that the Purkinje cells of each zone project to precise targets in the cerebellar nuclei, and these are accompanied by specific collateral olivonuclear and nucleo-olivary projections (Voogd & Ruigrok, 1997; Voogd & Glickstein, 1998). For some of these zones a much finer-grained microzonal arrangement is becoming clear (see Ekerot et al. 1997; Garwicz et al. 1998). This organisation is best established in the cat, but anatomical evidence has shown a similar zonal arrangement in other species. Recent electrophysiological evidence has shown that homologous zones to those described in the cat also exist in ferrets (Garwicz, 1997) and rats (Atkins & Apps, 1997; Jorntell et al. 2000), based on the activation of climbing fibres from sensory afferents. The conservation of a similar functional organisation between species, which is defined by specifically organised somatosensory inputs, implies that there are similarly conserved functions for these zones. As a consequence of this focus on activation from peripheral afferents, sensory activation is central to most hypotheses of climbing fibre function (for review see Simpson et al. 1996; De Zeeuw et al. 1998).
Many anatomical studies have revealed extensive descending projections from supraspinal structures to the inferior olive, including those areas that receive ascending afferent inputs. The presence of these pathways complicates the model that climbing fibres may signal errors signalled by sensory afferents. These descending projections arise from many locations including structures in the midbrain and close to the mesodiencephalic junction and also the cerebral cortex (for review see Brodal & Kawamura, 1980; De Zeeuw, 1990). Studies in the rat suggest that descending projections from the cerebral cortex are widespread (Brown et al. 1977; Swenson & Castro, 1983; Swenson et al. 1989). These projections reach all three subdivisions of the inferior olive; the dorsal and medial accessory olives (DAO and MAO) and the principal olive (PO). These subnuclei send climbing fibres to different cortical zones (Buisseret-Delmas & Angaut, 1993). In the cat there is some controversy over the extent of direct cortico-olivary projections. These were initially thought to be widespread, but later claimed to be much more focal (see Brodal & Kawamura, 1980; Saint-Cyr, 1983). Anatomical assessment of the projection to the inferior olive from the cerebral cortex is complicated by the need to discriminate corticofugal fibres of passage destined for other brainstem areas from fibres that make synaptic terminations in the olive. Functional studies of the influence of the sensorimotor cortex on climbing fibre pathways have in the main used electrical stimulation of the cortex (Andersson & Nyquist, 1983), and have addressed the distribution of responses. The pathways mediating these responses are largely unknown, but there are a number of possible candidates. Andersson (1984) showed that at least two pathways from the sensorimotor cortex activated climbing fibres, one from sensory cortex involved a relay through the dorsal column nuclei, another from motor cortex did not. Analysis of the pathways involved is complicated by the long latencies of the responses.
The question of whether there are functional projections from the cerebral cortex to the inferior olive, and whether such projections are inhibitory or excitatory is unresolved. Given these uncertainties, we have examined the influence of the corticofugal system on climbing fibres projecting to the rat posterior lobe cerebellar cortex using electrophysiological methods. The regions we have investigated (paravermnis and hemisphere of crus II and paramedian lobule) contain zones innervated by each of the olivary subnuclei (Buisseret-Delmas & Angaut, 1993). In the rat, c1 and c2 zones can be identified in both anterior and posterior lobes (Atkins & Apps, 1997; Jorntell et al. 2000). These have characteristic properties and are functionally similar to the zones in the cat (see Apps, 1999). Some preliminary results of this work have been published (Aggelopoulos et al. 1997).
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The data presented were obtained from experiments performed on 25 Sprague-Dawley rats (200-400 g) under general anaesthesia. All procedures were carried out under UK Home Office regulations. Anaesthesia was induced via an intraperitoneal injection of sodium pentobarbital (Sagatal, 40 mg kg-1), and subsequently maintained with single doses of 0.6 mg (I.V.) as needed to abolish limb withdrawal and corneal reflexes. Body temperature was maintained between 36 and 38 °C by a thermostatically controlled thermal blanket. The right femoral vein was cannulated for the delivery of fluids and a tracheal tube inserted. The animal was placed in a stereotaxic frame and the cerebellum was exposed by removal of part of the occipital bone. Further holes were made in the parietal bone to allow access for stimulating electrodes along the corticofugal pathway and to expose the sensorimotor cortex.
Stimulation
Corticofugal fibres were activated by electrical stimuli delivered via epoxy resin-coated stainless steel microelectrodes. The tips of these were exposed electrolytically to yield a final tip impedance of 90-120 k
. These were inserted at up to five locations in the corticofugal pathway in each experiment (see schematic diagram in Fig. 1A). The sites used were as follows.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 1. Responses to stimulation of corticofugal fibres in the sensorimotor cortex and cerebellum A, schematic diagram showing the experimental set-up. Corticofugal fibres are activated at different sites from rostral to caudal: internal capsule (IC), cerebral peduncle (CP), ponto-medullary junction (PTr), medullary pyramid (PTc) and pyramidal decussation (PTx). Antidromic volleys evoked by stimulation at these sites are recorded from the surface of the sensorimotor cerebral cortex, evoked climbing fibre responses are recorded from the cerebellum. B, averaged antidromic responses recorded from the sensorimotor cortex. These were typical surface-positive responses (Mediratta & Nicoll, 1983). Response threshold was less than 10 µA. C, climbing fibre responses evoked by stimulation of the PTr electrode at a site in the paramedian lobule (PML). The upper trace is an average (20 sweeps); the lower records are four superimposed individual responses, showing that these responses had sharp onsets and consistent latencies. D, an averaged surface-recorded climbing fibre response from crus II, evoked by stimulation through an electrode in the pyramid. The upper traces illustrate climbing fibre potentials recorded with a fine Tri-mel-coated wire electrode, the lower traces are antidromic volleys recorded simultaneously from the surface of the sensorimotor cerebral cortex. Only the first stimulus train evokes a climbing fibre response, showing a long refractory period of the climbing fibre responses. | ||
(a) The internal capsule (IC), where bundles of corticofugal fibres pass through the striatum (approximately 7.0 mm rostral to the interaural plane).
(b) The left cerebral peduncle (CP), 4.5 mm rostral to the interaural plane.
(c) In the left pyramid close to the ponto-medullary junction (PTr) at 1.0 mm rostral to the interaural plane.
(d) In the left pyramid at mid-medullary level, but still rostral to the inferior olivary nuclei (PTc) approximately 2 mm caudal to the interaural plane.
(e) In the left pyramid just rostral to the pyramidal decussation and caudal to the inferior olivary nucleus (PTx). This was localised by measuring the position of the obex and inserting the electrode at an angle of 30 deg to the vertical, tip anterior. The electrode entered at a point 0.5 mm lateral and 0.5 mm caudal to the obex, and was advanced to a depth of approximately 4.5 mm below the surface, where it was most effective in evoking antidromic responses in the motor cortex. This location is approximately 5.5-6 mm caudal to the interaural plane, approximately 1 mm caudal to the caudalmost parts of the inferior olive.
In all cases the positions of these electrodes were optimised by adjusting them to the location where antidromic field potentials evoked in the sensorimotor cortex by stimuli delivered through the electrodes had lowest thresholds. Placement of the cerebral cortical recording electrode was determined by the electrophysiological and anatomical description of Donoghue & Wise (1982). Antidromic volleys are illustrated in Fig. 1A. The CP, PTr and PTc electrodes, once positioned, were fixed in place with dental acrylic cement. The skin edges and cotton wool soaked in agar were used to form a pool around the exposed brain surface, which was covered in warm mineral oil to prevent tissue drying and cooling. Rectangular cathodal pulse stimuli (0.1 or 0.2 ms duration) were delivered through the electrodes, with a common anode placed on the skull surface, at a frequency of 0.5-1 Hz. Single stimuli and trains of up to three stimuli at frequencies of 333-500 Hz were used to evoke climbing fibre responses. Peripheral afferents were activated by stimuli delivered to the distal limbs via pairs of percutaneous stainless steel pins inserted approximately 5 mm in the ventral surface of the footpads. Stimuli were delivered through these at twice the threshold to evoke a visible response from the sensorimotor cortex. These did not cause substantial limb twitches.
Recording
Activation of cerebral corticofugal fibres was assessed from antidromically evoked potentials recorded from the surface of the sensorimotor cortex using silver ball electrodes. Climbing fibre responses were recorded from the surface of the cerebellum as surface-positive potentials, using either fine silver ball electrodes (up to 0.4 mm tip diameter) or finer blunt-ended Tri-mel-coated silver wire electrodes (0.12 mm tip diameter) with the indifferent placed on the adjacent skull. Climbing fibre responses were recorded from the cerebellar surface as large, positive-going responses (see Oscarsson, 1968), which were refractory for up to 80 ms after a first stimulus (Armstrong & Harvey, 1968). Small mossy fibre responses were seen in some experiments, but these were not prominent under the barbiturate anaesthesia used in this study (see Gordon et al. 1973). Two to three stimuli were routinely used to evoke potentials, in each case ensuring that the final stimulus was the one that evoked the responses. Signals were sampled online at high frequency (20 kHz) and also stored on digital audio tape for subsequent analysis off-line (10 kHz). Responses were analysed both as single responses and as averages of 20 sweeps. Latencies were always measured from the onset of the effective stimulus artefact to the onset of the response (except for antidromic responses where the short latencies of the onsets made measurement difficult). Conduction velocities were estimated from the distance between stimulation sites (measured from the stereotaxic positions of the electrodes and verified from reconstruction using serial histological sections) divided by the difference in response latency, therefore excluding utilisation times in the calculation.
At the end of each experiment electrolytic lesions were made to mark the locations of the tips of each stimulating electrode, the animals were killed with an overdose of sodium pentobarbital, the left ventricle was cannulated and the animal perfused with 4 % paraformaldehyde solution. The brain was cut in serial 50 µm sections from which the electrode tracks and tip locations could be reconstructed.
Lesions
In order to disrupt the corticospinal tract, electrolytic lesions were produced in the caudal pyramid in some experiments, by applying a 1 s pulse of 100 V through a low-impedance electrode positioned in the medullary pyramid. In other experiments lesions of the dorsal column nuclei were made mechanically with a pair of fine watchmaker's forceps. Recordings of potentials evoked in the sensorimotor cortex from limb afferent stimulation allowed assessment of the extent of the disruption to the ascending sensory pathways. In all cases, post-mortem histology confirmed the location of the electrodes and the extent of the lesions.
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In all experiments typical surface-positive climbing fibre responses could be recorded from the cerebellar surface after stimulation of corticofugal fibres. These were prominent in the paramedian lobule (PML) and at several locations in crus II. Figure 1B-D illustrates the appearance of these responses. The largest potentials were positive going and typical for surface-recorded climbing fibre potentials (see Oscarsson, 1968; Andersson & Nyquist, 1983). Smaller, surface-negative responses could be recorded medially and laterally to the zone where large positive responses were seen, reflecting activation of distant Purkinje cells. In all cases these potentials were refractory 40 ms after a first stimulus (see Fig. 1D; Armstrong & Harvey, 1968). Mossy fibre-mediated responses were excluded from the recordings by their morphology, latency and refractory properties (Eccles et al. 1967; and see Jorntell et al. 2000). Moreover, deep barbiturate anaesthesia as employed in our experiments suppressed the mossy fibre component of responses, as previously described by others (e.g. Gordon et al. 1973).
Verification that responses were evoked from the cerebral corticofugal pathway
Great care was taken to position the stimulating electrodes to ensure that they were located in the cerebral corticofugal pathway and not other structures. Antidromically evoked volleys recorded from the sensorimotor cortex were used as a guide; the final position being that at which volleys could be reliably evoked by currents of less than 10 µA in all cases (see Fig. 1A). These locations were always confirmed histologically after the experiment. Cerebellar climbing fibre responses were reliably evoked with brief trains of up to three stimuli (usually less than 30 µA), delivered from electrodes positioned in the cerebral corticofugal pathway (single stimuli rarely evoked climbing fibre responses). To ensure that the optimal site for evoking these responses was from the corticofugal pathway, we verified that the sizes of both antidromic cortical responses and orthodromically evoked climbing fibre responses changed in tandem as the electrode was moved (not illustrated).
Lesions of the corticospinal tract rostral to the inferior olive
To further confirm that responses were being evoked from the corticospinal pathway and to establish whether relays in rostral parts of the neuraxis mediate the responses evoked from the corticofugal pathway, focal electrolytic lesions were produced in the pyramid in four animals via electrodes placed in the pyramidal tract at mid-medullary level, about 2 mm rostral to the olive. Responses evoked from the stimulation of corticofugal fibres at more rostral sites (PTr and CP) were tested before and after the lesions (see Fig. 2A). Figure 2B illustrates the electrolytic lesion, which has interrupted most of the corticofugal fibres. These lesions dramatically reduced the climbing fibre responses that could be evoked from more rostral sites (e.g. Fig. 2C). Note that responses evoked by stimulation of cutaneous afferents from the ipsilateral forelimb (iFL) provided a positive control that these lesions did not dramatically alter the excitability of the inferior olive. Among the group of animals, climbing fibre responses evoked from the corticofugal fibres were significantly reduced to 36 % (S.E.M., ±4 %) of pre-lesion levels (Student's t test, 1 tail, P < 0.02), whereas those evoked from the ipsilateral forelimb were unchanged. Although these lesions were localised to the pyramidal tract, they rarely included the entire structure. Not surprisingly, therefore, climbing fibre responses evoked from the corticospinal tract were never completely abolished (Fig. 2C), although they were greatly reduced throughout both folia of the PML and crus II. These results thus indicate that at least a large proportion of the responses evoked from the stimulation of corticofugal fibres in these experiments were not mediated by relays in the rostral brainstem.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 2. Effects of lesions of the pyramidal tract in the low medulla A, climbing fibre responses evoked by stimulation of ipsilateral forelimb cutaneous afferents (iFL - upper records) and corticofugal fibres (PTr - lower records). The black traces were recorded before a lesion of the corticofugal fibres in the low medulla, the grey traces were recorded after the lesion. The climbing fibre responses evoked from the PTr stimulus are diminished after the lesion, the iFL evoked responses are little changed. B, a dark field photomicrograph of the post-mortem histology, showing the centre of the electrolytic lesion. Although the lesion did not destroy all corticofugal fibres, it was confined to the pyramid. C compares the effects of lesions of this type from four animals. The sizes of the post-lesion responses are plotted as a percentage of the pre-lesion (control) climbing fibre response (means ± S.E.M.). | ||
Conduction velocities
In order to estimate the conduction velocities of the corticofugal axons that evoked the climbing fibre responses, the latency of climbing fibre responses evoked by stimuli delivered at several different rostro-caudally spaced sites in the corticofugal pathway were compared (Fig. 3). Antidromic responses evoked by these stimuli in the sensorimotor cortex provided an assessment of the conduction velocity of the fastest corticofugal fibres. The antidromic potentials had short onset latencies (often < 1 ms), making accurate assessment of response onset difficult. In the light of this, differences in the latency of the antidromic potentials were measured from the major inflexion of the response (illustrated in Fig. 3C). The peak latencies of antidromic responses in the sensorimotor cortex were less than 1 ms from the internal capsule (IC), increasing to 1.5-2.0 ms for responses to stimuli delivered in the mid-medulla (PTc). From these latencies and the known distances between stimulation sites, the conduction velocity of the fastest corticofugal fibres was estimated. From 12 experiments where multiple sites were stimulated, the mean estimated conduction velocity was 18.7 ± 2.3 m s-1 (mean ± S.E.M.), in good agreement with published values (Porter & Sanderson, 1964; Mediratta & Nicoll, 1983; Landry et al. 1984; Stewart et al. 1990; Babalian et al. 1993).
 |
View larger version [in this window] [in a new window] |
|
|
Figure 3. Conduction velocity of the corticofugal fibres The records illustrated in A-C were taken from an experiment in which corticofugal fibres were stimulated through electrodes in CP, PTr and PTc. A, climbing fibre responses evoked at a site in crus II from all three stimulation sites, four sweeps superimposed in each case. The onset region (marked by the box) is shown on an expanded timebase in B. C shows averaged antidromic volleys recorded from the sensorimotor cortex by stimuli delivered to the same electrodes. In this case the dashed lines mark the inflexion of the responses, which was used because the onset was not always easily discriminable from the stimulus artefact. Note the different timescales. D, bar chart summarising conduction velocity estimates based on latency differences as shown in B and C. Estimates of antidromic (AD) conduction velocity (from records such as those in C), which would be dominated by the fastest conducting fibres, are substantially greater than the estimated conduction velocities of fibres that evoke climbing fibre responses (CFRs). Consistent conduction velocity estimates were obtained between IC and CP, CP and PTr and PTr and PTc. | ||
Climbing fibre responses evoked in response to the stimulation of corticofugal fibres had onset latencies of 8-12 ms, substantially in excess of the olivocerebellar conduction time (3-5 ms; Sugihara et al. 1993). In comparison to the antidromic responses, there were larger differences between the onset latencies of climbing fibre responses evoked from different sites in the corticofugal pathway in all experiments. Figure 3A shows the climbing fibre responses evoked from the same sites as the antidromic volleys shown in Fig. 3C. The onsets of these responses are shown on the same timebase as the antidromic volleys in Fig. 3B. The difference in onset latency of the climbing fibre responses evoked from the cerebral peduncle in the midbrain (CP) and the pyramidal tract mid-medulla (PTc) is about 2 ms (Fig. 3B), whereas the difference between the latencies of the antidromic volleys evoked by the same stimuli is less than 0.4 ms (Fig. 3C). The large difference between the latencies of the climbing fibre responses (Fig. 3A) cannot easily be explained by synaptic relays in the rostral brainstem. Firstly, the responses were significantly reduced by lesions confined to the caudal pyramid (which is composed solely of corticofugal axons). Secondly, there was a significant linear correlation between the stimulating electrode to recording electrode distance and latency (y = 10.7 + 0.68x; r = 0.725; t statistic = 22.6, P < 0.001; F statistic = 26.7, P < 0.001), which would be difficult to explain if there were an intervening synapse between any of the stimulating electrode locations. These imply that the greater difference in latency originates from a lower conduction velocity in the corticofugal fibres responsible for the activation of climbing fibres. By comparing the onset latencies of responses evoked from different sites we were able to estimate the conduction velocities of the fibres responsible for climbing fibre activation (Fig. 3D). From the difference between the latencies of responses evoked from the CP and PTr, the conduction velocity of the fibres responsible for climbing fibre activation was estimated to be 2.3 ± 0.4 m s-1 (mean ± S.E.M.; 8 experiments). Similarly the conduction velocity estimated from responses evoked from the IC and the CP was 1.8 ± 0.4 m s-1 (mean ± S.E.M.; 3 experiments) and that estimated from responses evoked from the PTr and PTc was 1.3 ± 0.4 m s-1 (mean ± S.E.M.; 4 experiments). There was no significant difference between these independent velocity measurements (Kruskal-Wallis one-way ANOVA of ranks). There was, however, a highly significant difference between these estimates and the estimated velocity of the fastest conducting corticofugal fibres that were responsible for the antidromic volleys (Kruskal-Wallis one-way ANOVA of ranks, using Dunn's method for multiple comparisons with a control group, P < 0.001). These data imply that the responses were mediated through a population of slowly conducting corticofugal axons.
Responses at different sites
Climbing fibre responses were evoked at different sites in the PML and crus II, consistent with corticofugal input to specific zones. Two major areas were consistently seen; one site was medially close to the paravermal groove, and larger responses were seen over a broader area further laterally. This pattern was seen in both PML and crus II, although the more medial responses were only seen in a narrow part of crus II, and the lateral responses were only seen in a narrow zone of PML. Examples of responses from these different locations are illustrated in Fig. 4B, where they are aligned to the effective stimulus. The responses in the medial zone had consistently shorter latencies than the responses seen further laterally. The approach described previously, stimulating the corticofugal pathway at several different locations, was used to examine responses evoked at both sites. In all cases, consistently longer-latency responses were evoked from more rostral stimulation sites. Figure 4C shows the distributions of responses evoked at these sites from the CP and rostral pyramid (PTr). These differences yielded estimates of conduction velocity that were similar for each zone at about 2 m s-1 (Fig. 4D), implying that a similar population of slow conducting cerebral corticofugal axons generated the climbing fibre responses in the two areas.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 4. Climbing fibre responses evoked at different cortical sites A, superimposed averaged responses evoked at medial and lateral sites in the PML and at a site in lateral crus II in response to stimulation in the medullary pyramid (PT). A short-latency response is recorded medially in the PML (latency approximately 8 ms), with longer-latency responses laterally in PML and crus II. Lower trace illustrates a response recorded laterally in crus II (approximately 12 ms onset latency). In all cases the third stimulus of the train was the effective stimulus. The voltage calibration bar represents 0.1 mV for the PML responses and 0.25 mV for the crus II response. B, distribution of short- and long-latency responses across the PML. The early responses ( | ||
Are the climbing fibre responses generated by corticospinal fibres?
In order to test whether the climbing fibre responses evoked by stimulation of corticofugal fibres were evoked by fibres projecting caudal to the inferior olive, corticospinal fibres were activated at the decussation of the corticospinal tract, caudal to the level of the inferior olive, in three experiments. Climbing fibre responses were reliably evoked from stimulation within the decussation of the pyramidal tract (PTx) in all of the experiments, at a location where weak stimuli (< 10 µA) could evoke antidromic cerebral cortical responses. The presence of substantial ascending inputs to the inferior olive raises the possibility that stimuli delivered in the region of the decussation may have activated ascending pathways. To examine this possibility, the distributions of responses evoked from the PTx and PTr across the cerebellar surface were compared. The distributions of potentials evoked from the two sites were very similar, both in the form of the individual traces and in the general distribution of responses of different amplitudes (Fig. 5). Note that the lack of somatotopic organisation in the corticospinal fibres (Dawnay & Glees, 1986; Coleman et al. 1997) means that a similar distribution of responses would be expected if the stimuli delivered were submaximal. Particularly in crus II, substantial climbing fibre field potentials could be evoked by stimulation of corticofugal fibres in the absence of climbing fibre potentials evoked from the peripheral afferents tested: here the PTx evoked responses did not co-localise with the peripheral afferent responses, but did covary with the PTr evoked responses. In addition, we verified that the PTx stimuli did not evoke synaptic field potentials typical of activation of the ascending sensory pathways in the sensorimotor cortex, which would follow activation of the ascending dorsal column-medial lemniscal pathway. These data suggest that at least some of the corticofugal fibres that activate climbing fibre responses project into the spinal cord as corticospinal axons.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 5. Responses evoked from stimulation of corticofugal fibres caudal to the inferior olive A, responses (averages of 20 sweeps) evoked from pyramid rostral to the inferior olive (PTr) and caudal to the inferior olive at the level of the pyramidal decussation (PTx), from four sites distributed across crus II. The recording sites are illustrated on the diagram of the rotated hemicerebellum. B, the distribution of responses across the hemicerebellum. PTr ( | ||
Lesions of the dorsal column nuclei
In the cat some of the climbing fibre responses evoked by stimulation of the cerebral cortex (those evoked by stimulation over the posterior sigmoid gyrus (somatosensory) cortex) are relayed via the dorsal column nuclei, while responses evoked from stimulation over the anterior sigmoid gyrus (motor cortex) are not (Andersson, 1984). The long latencies of the climbing fibre responses evoked from corticofugal pathways might accommodate a relay in the dorsal column nuclei. To examine whether the responses seen in this study were relayed through the dorsal column nuclei, controlled mechanical lesions were made that included the bodies of both the gracile and cuneate nuclei. These lesions were made while monitoring sensory evoked potentials generated in the sensorimotor cortex by limb afferent stimulation. An example of part of such a lesion is illustrated in Fig. 6B. In order to examine whether the lesions influenced climbing fibre responses at different cerebellar locations, we examined the effects on responses recorded in crus II and medially in the PML. Typical records of climbing fibre responses recorded from medial PML and lateral crus II, before (thick grey traces) and after (thin black traces) a lesion of the dorsal column nuclei are illustrated in Fig. 6A. Climbing fibre responses and somatosensory cortex (Cx) responses evoked by stimulation of ipsilateral fore- and hindlimb afferents (iFL and iHL, respectively) are also shown. Notice that lesions that essentially abolish the peripheral evoked responses in the sensorimotor cortex, as well as the climbing fibre responses evoked from the iFL in PML, had relatively small effects on the climbing fibre responses evoked from the pyramidal tract (PT). The iHL climbing fibre response was largely unaffected by these lesions, implying a relay elsewhere. This is consistent with a direct projection from spino-olivary neurones in the lumbosacral cord to the rostral DAO (see Horn et al. 1998). The presence of large climbing fibre responses after the lesion implies that climbing fibre excitability was not greatly reduced by the lesion. The differential effect of these lesions of PT and iFL responses implies that the PT responses were not substantially relayed through the dorsal column nuclei. Lesions of this type were made in five experiments, in the first two of which very extensive lesions were made. To analyse these data, we have plotted the relative reduction of the PT evoked climbing fibre responses against the relative reduction of the iFL evoked responses (Fig. 6C). For the crus II responses there was no correlation between the effect on the iFL and PT evoked climbing fibre responses. For the PML evoked responses there was a significant negative linear regression (y = 46 + 0.52x; P < 0.05), with an intercept at 46 %. Thus a lesion that abolished the iFL response would be expected to reduce the PT response to about half its original size. One interpretation of this result would be that the PT stimulation activated the climbing fibres projecting to the PML through two pathways, one that relayed via the dorsal column nuclei and one that did not (see Discussion).
 |
View larger version [in this window] [in a new window] |
|
|
Figure 6. Effects of lesions of the dorsal column nuclei A, climbing fibre responses and cerebral cortical responses (Cx) evoked by stimulation of peripheral afferents and the PT. Control responses (thick grey traces) can be compared to those evoked after ablation of the dorsal column nuclei (thin black traces). Responses to stimulation of corticofugal fibres (PT) were mostly unaffected by the lesions, whereas responses to stimulation of forelimb afferents (iFL) were greatly reduced. Sensory evoked responses to forelimb or hindlimb afferents (iHL) in the sensorimotor cortex were effectively abolished. Climbing fibre responses to hindlimb afferents were reduced, but not greatly so. The traces are averages of 20 responses. B, a camera lucida drawing of a histological reconstruction of the lesion in this experiment: the shaded area shows the extent of the lesion. C, the relative effect of these lesions on the climbing fibre responses evoked by PT stimuli against the effect on responses evoked by iFL afferents. The plots show a greater effect of the lesions on iFL than PT responses. The crus II responses show no correlation, but there is a significant linear regression for the PML responses (P < 0.05). The intercept of this regression line is about 50 %, implying that a lesion which abolishes the iFL response would reduce the PT response by half. | ||
Distribution of responses
A complete map of the distribution of climbing fibre responses evoked from the corticospinal tract and from limb afferent nerves throughout crus IIb and rostral and caudal folia of the PML was obtained in eight experiments from sequential recordings across each folium. Partial maps (1-2 folia) were obtained in other experiments. Our aim here was to obtain an overview of the general organisation of the projections, rather than an exhaustive analysis of the microzones, which would require a higher level of resolution in the recordings given the narrowness of some of these zones. For these experiments we used a fixed stimulus paradigm of three supramaximal stimuli to the corticofugal pathways.
We were unable to find responses evoked by stimulation of corticofugal fibres on the vermis, except in its most lateral part close to the paravermal groove. The responses here extended across onto the medial hemisphere and are considered with the projections to the medial hemisphere. In lobules VI and VII these responses were not at locations where the peripheral afferents we tested evoked climbing fibre responses. Responses were found in several identifiable zones on the hemisphere with a consistent pattern moving laterally from the paravermal groove to the most lateral extent of the exposure. This progression was similar in the different folia, but with some differences. In crus II, the largest responses evoked from corticofugal fibres were consistently found in a sagittal band in the middle of the exposed surface, centred 2-3 mm lateral to the paravermal groove. These corticofugal responses were co-localised in areas of cortex where bilateral peripheral limb inputs also evoked climbing fibre responses at relatively long latencies (about 20 ms, see Fig. 7C). These bilateral limb responses were generally labile, yet they could be recorded over a broad zone in crus II (about 1 mm mediolaterally). The characteristics of this region are those of the c2 zone (Jorntell et al. 2000). Large climbing fibre responses were also evoked by corticofugal fibres in areas medial and lateral to this zone in crus II. Laterally we were unable to evoke substantial responses to stimulation of any of the peripheral afferent stimuli we tested, although PT stimuli evoked large synchronous responses (Fig. 7E). Medial to the c2 zone there were responses to iFL inputs in some experiments, but these were usually small and in a narrow band close to the paravermal groove, whereas PT evoked responses were larger and were consistently found (Fig. 7A). In a region in between these areas the responses evoked from the PT were small. In the rostral folium of the PML a similar pattern of responses was observed. The region of bilateral limb responses was less easily defined than in crus II and was narrower, being located about 2 mm lateral to the paravermal groove (Fig. 7D). As in crus II, large responses were evoked from the PT, but not from the peripheral stimuli we used, in the region lateral to this zone (Fig. 7F). Medial to the c2 zone, at locations 0.4-2 mm lateral to the paravermal groove, large responses were evoked by corticofugal stimuli and these were accompanied by short-latency responses to iFL and occasionally to iHL (Fig. 7B). These are the characteristics of the c1 zone in the anterior lobe (Jorntell et al. 2000). In the caudal folium of the PML the region of bilateral responses (c2 zone) was not found in two experiments and was at the limit of the exposure in others. Large corticofugal evoked responses were seen at these locations. Areas lateral to this region were not exposed. The most prominent features of the caudal folium of the PML were large, short-latency responses to iHL, and sometimes iFL, in a region around the paravermal groove and up to 2 mm lateral to it (Fig. 7B). These were consistent with a c1 zone location and were accompanied by large responses to PT stimulation.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 7. Distribution of responses across the cerebellar hemisphere This figure illustrates the distribution of climbing fibre responses evoked from the PTr, ipsilateral forelimb (iFL), contralateral forelimb (cFL) and ipsilateral hindlimb (iHL). Responses were mapped across the entire cortex, but for simplicity, only selected sites are shown. All traces are averages of 20 responses to stimuli of fixed intensity. These include locations from crus II and PML close to the paravermal groove (A and B), mid-hemisphere (C and D) and laterally on the hemisphere (E and F). Note that at sites C and D large responses were evoked bilaterally from forelimb afferents as well as from the PTr. The vertical calibrations with each trace are 0.1 mV for iFL responses (except D, 0.2 mV), 0.2 mV for all iHL and cFL responses, and 0.2 mV for all PTr responses (except C, 0.5 mV). | ||
The copula pyramidis (hemisphere lobule VIII) was examined in four experiments. As described previously, large responses were evoked in this region from ipsilateral hindlimb afferents, and this region is considered to include a hindlimb c1 zone (Atkins & Apps, 1997). Substantial responses to PT stimulation were also seen in this region (not illustrated), although problems of access precluded further analysis.
Our interpretation of this pattern is that climbing fibres projecting to a c1 zone located in the medial hemisphere, immediately lateral to the paravermal groove, receive input from corticofugal pathways. This zone is relatively broad in PML, but becomes narrower in crus II. Further laterally is a c2 zone where bilateral limb inputs activate responses at longer latency, and PT responses are also seen at that location. Opposite to the c1 zone, the c2 zone appears to be narrow in PML and broader in crus II. This organisation is represented schematically in Fig. 8.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 8. Summary of distribution of responses This diagram attempts to summarise the distribution of responses in PML and crus II. Responses from the PT were evoked just lateral to the paravermal groove; in crus II these were not accompanied by limb responses, but in PML they were found at sites where responses from iFL and/or iHL also evoked responses. We have tentatively labelled this region c1. Further laterally is a zone where climbing fibre responses were evoked from both forelimbs and the PT, which should correspond to c2. Further laterally climbing fibre responses were evoked from the PT but not from any of the sources of peripheral afferents we tested. | ||
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Origin of responses
In these experiments we describe a powerful excitatory input from cerebral corticofugal fibres to the climbing fibres projecting to the rat posterior lobe cerebellar hemisphere. A key issue is the pathway through which these responses are mediated and a number of possible routes exist. The simplest route would be a direct connection from corticofugal fibres to the inferior olive. Alternatives would involve relays at sites that might include rostral brainstem structures (e.g. red nucleus and surrounding region, nucleus of Darkschewitsch) and the dorsal column nuclei, among others (see Brodal & Kawamura, 1980). The large reductions of responses to stimulation at rostral sites by focal electrolytic lesions through electrodes placed in the pyramidal tract mid-medulla only 2 mm rostral to the inferior olive indicate that the responses seen under our experimental conditions were not evoked through relays in the rostral brainstem to any great extent. This is also supported by the progressive shortening of the latencies of responses evoked from electrodes at progressively more caudal locations in the corticofugal pathway (Fig. 3). The consistent conduction velocity estimates between the stimulating sites argue against a relay closely associated with the cerebral corticofugal pathway rostral to the level of the inferior olive. These argue for activation of a progressively shorter pathway at more caudal locations. If the responses are not mediated by a relay in the rostral brainstem, then the dorsal column nuclei are candidates for a relay. Substantial lesions of the dorsal column nuclei did not abolish the responses (Fig. 6), although they did reduce the responses in the PML. In the cat, blockade of transmission through the dorsal column nuclei abolished climbing fibre responses evoked in the paravermal cortex by stimulation of the sensory cerebral cortex, but not those evoked from stimulation of the motor cortex (Andersson, 1984). These responses were recorded at the same locations on the cerebellum, implying that alternative pathways converged from the sensory and motor cortex on the same climbing fibres. There are extensive projections from the cerebral cortex to the dorsal column nuclei (see Armand, 1982; Swenson & Castro, 1983; Molinari et al. 1996) and from the dorsal column nuclei to the rostral accessory olives and PO, which send climbing fibres to areas of the hemisphere (Molinari et al. 1996). Cuneo-olivary and gracilo-olivary projection neurones are located principally within the caudal portions of the dorsal column nuclei (Molinari, 1984; McCurdy et al. 1998). Our lesions of the dorsal column nuclei are thus unlikely to have spared corticofugal projections. The lesions did effectively remove the peripheral inputs to climbing fibres. It is therefore unlikely that the dorsal column nuclei are a major relay for the responses described here. The regression analysis does suggest different influences of dorsal column lesions on responses recorded at different locations in the PML and crus II. The PML responses were recorded at a location where iFL and iHL evoked short-latency climbing fibre potentials and are thus consistent with a c1 zone location. These were affected by the dorsal column lesion, as shown by the significant negative regression in Fig. 6C, but a significant part of the response survived these lesions (it remained at about 50 % of original size). This finding could be interpreted as indicating that the responses had two components, one relayed through the dorsal column nuclei, one not. This would parallel the findings of Andersson (1984), that in the cat c3 zone climbing fibre responses evoked from the sensory cortex were abolished by dorsal column lesions, whereas motor cortex evoked responses in the same zone were unaffected. The crus II responses were recorded from an area where climbing fibre responses were evoked from the iFL and contralateral forelimb (cFL) at long latency, consistent with a c2 zone location.
Latencies
In their classical work on cerebocerebellar interactions, Allen & Tsukahara (1974) considered the inferior olive to be innervated by slow corticospinal axons. This was based on a short publication (Kitai et al. 1969) in which the latencies of climbing fibre responses to stimuli delivered in the subcortical white matter below the sensorimotor cortex and in the cerebral peduncle in the cat were compared. A direct projection was assumed and it was estimated that the cerebral corticofugal axons that activate climbing fibres conduct at about 9 m s-1, compared to a velocity in excess of 60 m s-1 for the fastest conducting corticospinal fibres. The cerebellar recording zone was not described, and the possibility of a relay in the brainstem was not discussed. The data presented here reveal a similar arrangement in a different species (rat), and also provide evidence that the responses were not mediated by a relay in the upper brainstem. The ratio of the conduction velocities of the 'slow' sub-population of cerebral corticofugal axons that generate climbing fibre responses (approximately 2 m s-1), and the velocities of the most rapidly conducting cerebral corticofugal axons (estimated from the antidromic volleys to be about 19 m s-1), is of a similar order to that in the cat (i.e. 9 m s-1:60 m s-1). The estimated conduction velocity of the fastest corticofugal fibres in the rat (19 m s-1) is in agreement with values for the fastest corticospinal axons from numerous previous studies (Porter & Sanderson, 1964; Mediratta & Nicoll, 1983; Landry et al. 1984; Stewart et al. 1990; Babalian et al. 1993).
According to Hursh (1939) conduction velocities as low as those of the fibres that activate climbing fibres would be observed in myelinated axons with a cross-sectional diameter of approximately 0.3 µm. In addition to fine myelinated axons, the pyramid of the rat contains a high proportion of unmyelinated axons (Leenen et al. 1985) and, if the stimuli we used could activate such small axons, they could have contributed to the activation of climbing fibres. From the current data, we cannot determine the cortical region that gives origin to the corticofugal axons that activate climbing fibre responses. The mean onset latency of responses evoked from the PTr site (ponto-medullary junction) was about 10 ms. Assuming that they were mediated through collaterals projecting directly to the inferior olive as the pyramidal tract passes it (approximately 4 mm caudal to the stimulation site), a conduction velocity of about 2 m s-1 would equate with a conduction delay of about 2 ms. This would be coupled to a conduction time in the olivocerebellar fibres to crus II of 4.16 ± 0.36 ms (range 3-5 ms; Sugihara et al. 1993). This leaves 3-4 ms unaccounted for. If these responses were mediated via a direct pathway some slowing might be expected in the collateral branches, and there will be some synaptic delay at the inferior olive. In the mouse, collateral branches from pyramidal tract axons have been described as arising caudal to the olive then ascending into the nucleus (Terashima, 1995), a course that would also involve some delay. However, the latencies do not provide a conclusive argument as to the synaptic linkage.
The simplest explanation of our electrophysiological findings is that the responses in crus II and at least a proportion of the responses in PML (which survived dorsal column lesions) were mediated by an excitatory projection from the sensorimotor cortex which directly activates the inferior olive or relays in its close proximity. Consistent with this notion, anterograde tracing experiments have demonstrated direct projections from the sensorimotor cortex to the relevant subnuclei of the inferior olive in the rat (Swenson et al. 1989; see following subsection). This does not imply that this is the only route through which cerebral cortical areas can influence the climbing fibre system: the absence of actions via indirect pathways may be the result of depression by the anaesthesia used in these experiments.
Distribution of responses
The development of modern precise tracing methods has revealed a very high degree of organisation of the cerebellar cortex into olivo-cortico-nuclear modules. Some of these include very narrow cortical territories, much smaller than could be adequately examined with the electrophysiological methods used here. These studies do, however, reveal a broad organisational arrangement, which is in broad agreement with the organisation reported by others (Buisseret-Delmas & Angaut 1993; Atkins & Apps, 1997; Jorntell et al. 2000). Mapping the responses across the PML and crus II revealed a consistent pattern of responses to corticofugal fibre stimulation and from peripheral afferents. A consistent feature of both crus II and PML is a c2 zone, relatively broader in crus II than in the PML, where long-latency responses were evoked from both ipsilateral and contralateral limb afferents (see Fig. 8). Similar responses have been described for the PML by Atkins & Apps (1997), and also for a region in the anterior lobe by Jorntell et al. (2000). Substantial responses were evoked from corticofugal fibres at putative c2 sites. This would be consistent with the substantial direct projections from the sensorimotor cortex to the olivary subnucleus projecting to this cortical zone (the rostral MAO) in the rat (Swenson et al. 1989). This is not the case in the cat, where the rostral MAO is devoid of projections from the cerebral cortex (see Brodal & Kawamura, 1980; Saint-Cyr, 1983), but is innervated by afferents from the rostral brainstem, some of which are themselves targets for cortical projections (see Brodal & Kawamura, 1980, for discussion). Prominent shorter-latency climbing fibre responses were evoked from ipsilateral afferents in more medial areas of the PML, and in some cases in crus II. These properties are consistent with a c1 zone and have also been described before in the PML (Atkins & Apps, 1997; Teune et al. 1998) and anterior lobe (Jorntell et al. 2000). Again these responses, which are mediated by climbing fibres originating from the rostral DAO, were co-localised with substantial responses to stimulation of corticofugal afferents. Similar co-localisation of limb evoked responses, and responses evoked by stimulation of the sensorimotor cortex, has been described in the cat for the c1 and c3 zones (Miller et al. 1969; Andersson, 1984), which share many properties with the c1 zone of the rat. Some of these cerebral corticofugal responses may be mediated by direct projections to the rostral DAO from the cerebral cortex, as there is anatomical evidence that such direct connections are present (cat: Brodal & Kawamura, 1980; Saint-Cyr, 1983; rat: Swenson et al. 1989). In crus II, large climbing fibre responses were evoked by corticofugal fibres in a region lateral to the c2 zone, and these were not co-localised with short-latency limb afferent-evoked responses. In the cat a c3 zone exists lateral to the c2 zone, in both anterior and posterior lobes. The c3 zone shares many characteristics with the c1 zone, notably short-latency ipsilaterally evoked climbing fibre responses. One key difference between the cat and rat anterior lobe found by Jorntell et al. (2000) was the absence of a clear c3 zone. At least in crus II we were able to explore the cerebellar surface lateral to the c2 zone and were also unable to find a zone with the characteristics of the c3 zone. This area might instead be consistent with a d zone, which is innervated by climbing fibres from the principal olive, again a site for direct cortico-olivary fibres in the rat (Swenson et al. 1989).
Overall there is a good congruence between the anatomical and functional findings. However, the fact that the corticofugal fibres that mediate climbing fibre responses are of fine calibre may make them difficult to identify with anterograde tracers, especially particulate tracers (e.g. horseradish peroxidase). To some extent this may have been responsible for some of the reported discrepancies in the prevalence of cortical projections to the inferior olive (Saint-Cyr, 1983).
One interesting finding was that the latency of corticofugal responses evoked in the lateral c2 zone was 1-2 ms longer than those recorded medially in the c1 zone, a pattern that parallels the latencies of peripheral responses recorded in these respective cerebellar zones. In the cat, stimulation of various areas of the sensorimotor cortex evoke responses in the CZ zone (Andersson & Nyquist, 1983).
The cerebro-olivary projection as a collateral of the corticospinal tract
The results of this study imply that the climbing fibre responses evoked by stimulation of corticofugal fibres are mediated through a population of fibres that conduct slowly in comparison to the fastest corticofugal fibres. Since similar responses were evoked by stimulation at the level of the decussation, many of these fibres may continue into the spinal cord as corticospinal fibres. We have very little knowledge of the information carried by these more slowly conducting fibres, in any species. The similarity of the relation between the slow corticofugal fibres that activate the inferior olive and the fastest conducting cortical output fibres in animals with such different cerebral cortical organisation (rat and cat) is remarkable. This, together with the finding that the activation of climbing fibres to different zones in the posterior lobe is also via a set of relatively slowly conducting corticofugal fibres, implies that this is an important principle in the control of climbing fibres by the cerebral cortex.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ADRIAN, E. D. (1943). Afferent areas in the cerebellum connected with the limbs. Brain 66, 289-315 | |
| AGGELOPOULOS, N. C., BAKER, M. R., JAVID, M. & EDGLEY, S. A. (1997). Activation of climbing fibre pathways to the rat cerebellar cortex by cerebral corticofugal pathways. Journal of Physiology 504.P, 113P | |
| ALLEN, G. I. & TSUKAHARA, N. (1974). Cerebrocerebellar communication systems. Physiological Reviews 54, 957-1006 | [Medline] |
| ANDERSSON, G. (1984). Demonstration of a cuneate relay in a cortico-olivo-cerebellar pathway in the cat. Neuroscience Letters 46, 47-52 | [Medline] |
| ANDERSSON, G. & NYQUIST, J. (1983). Origin and sagittal termination areas of cerebro-cerebellar climbing fibre paths in the cat. Journal of Physiology 337, 257-285 | [Abstract] |
| APPS, R. (1999). Movement-related gating of climbing fibre input to cerebellar cortical zones. Progress in Neurobiology 57, 537-562 | [Medline] |
| ARMAND, J. (1982). The origin, course and terminations of corticospinal fibers in various mammals. Progress in Brain Research 57, 329-360 | [Medline] |
| ARMSTRONG, D. M. & HARVEY, R. J. (1968). Responses of a spino-olivocerebellar pathway in the cat. Journal of Physiology 194, 147-168 | [Medline] |
| ATKINS, M. J. & APPS, R. (1997). Somatotopical organisation within the climbing fibre projection to the paramedian lobule and copula pyramidis of the rat cerebellum. Journal of Comparative Neurology 389, 249-263 | [Medline] |
| BABALIAN, A., LIANG, F. & ROUILLER, E. M. (1993). Cortical influences on cervical motoneurons in the rat: recordings of synaptic responses from motoneurons and compound action potential from corticospinal axons. Neuroscience Research 16, 301-310 | [Medline] |
| BRODAL, A. & KAWAMURA, K. (1980). Olivocerebellar projection: a review. Advances in Anatomy, Embryology and Cell Biology 64, 1-138 | [Medline] |
| BROWN, J. T., CHAN-PALAY, V. & PALAY, S. L. (1977). A study of afferent input to the inferior olivary complex in the rat by retrograde axonal transport of horseradish peroxidase. Journal of Comparative Neurology 176, 1-22 | [Medline] |
| BUISSERET-DELMAS, C. & ANGAUT, P. (1993). The cerebellar olivo-corticonuclear connections in the rat. Progress in Neurobiology 40, 63-87 | [Medline] |
| COLEMAN, K. A., BAKER, G. E. & MITROFANIS, J. (1997). Topography of fibre organisation in the corticofugal pathways of rats. Journal of Comparative Neurology 381, 143-157 | [Medline] |
| DAWNAY, N. A. & GLEES, P. (1986). Somatotopic analysis of fibre and terminal distribution in the primate corticospinal pathway. Brain Research 391, 115-123 | [Medline] |
| DE ZEEUW, C. I. (1990). Ultrastructure of the cat inferior olive. PhD thesis, Erasmus University, Rotterdam | |
| DE ZEEUW, C. I., SIMPSON, J. I., HOOGENRAAD, C. C., GALJART, N., KOEKKOEK, S. K. & RUIGROK, T. J. (1998). Microcircuitry and function of the inferior olive. Trends in Neurosciences 21, 391-400 | [Medline] |
| DONOGHUE, J. P. & WISE, S. P. (1982). The motor cortex of the rat: cytoarchitecture and microstimulation mapping. Journal of Comparative Neurology 212, 76-88 | [Medline] |
| ECCLES, J. C., SUSAKI, K. & STRATA, P. (1967). Interpretation of the potential fields generated in the cerebellar cortex by a mossy fibre volley. Experimental Brain Research 3, 58-81 | |
| EKEROT, C.-F., GARWICZ, M. & JORNTELL, H. (1997). The control of forelimb movements by intermediate cerebellum. Progress in Brain Research 114, 423-429 | [Medline] |
| GARWICZ, M. (1997). Sagittal zonal organization of climbing fibre input to the cerebellar anterior lobe of the ferret. Experimental Brain Research 117, 389-398 | [Medline] |
| GARWICZ, M., JORNTELL, H. & EKEROT, C.-F. (1998). Cutaneous receptive fields and topography of mossy fibres and climbing fibres projecting to cat cerebellar C3 zone. Journal of Physiology 512, 277-293 | [Abstract/Full Text] |
| GORDON, M., RUBIA, F. J. & STRATA, P. (1973). The effect of pentothal on the activity evoked in the cerebellar cortex. Experimental Brain Research 17, 50-62 | [Medline] |
| HORN, K. M., HAMM, T. M. & GIBSON, A. R. (1998). Red nucleus stimulation inhibits within the inferior olive. Journal of Neurophysiology 80, 3127-3136 | [Abstract/Full Text] |
| HUMPHREY, D. R., CORRIE, W. S. & RIETZ, R. (1978). Properties of the pyramidal tract neuron system within the precentral wrist and hand area of primate motor cortex. Journal of Physiology (Paris) 74, 215-226 | |
| HURSH, J. B. (1939). Conduction velocity and diameter of nerve fibres. American Journal of Physiology 127, 131-139 | |
| JORNTELL, H., EKEROT, C., GARWICZ, M. & LUO, X. L. (2000). Functional organization of climbing fibre projection to the cerebellar anterior lobe of the rat. Journal of Physiology 522, 297-309 | [Abstract/Full Text] |
| KITAI, S. T., OSHIMA, T., PROVINI, L. & TSUKAHARA, N. (1969). Cerebro-cerebellar connections mediated by fast and slow conducting pyramidal tract fibres of the cat. Brain Research 15, 267-271 | [Medline] |
| LANDRY, P., WILSON, C. J. & KITAI, S. T. (1984). Morphological and electrophysiological characteristics of pyramidal tract neurons in the rat. Experimental Brain Research 57, 177-190 | [Medline] |
| LEENEN, L. P., MEEK, J., POSTHUMA, P. R. & NIEUWENHUYS, R. (1985). A detailed morphometrical analysis of the pyramidal tract of the rat. Brain Research 359, 65-80 | [Medline] |
| MCCURDY, M. L, HOUK, J. C. & GIBSON, A. R. (1998). Organization of ascending pathways to the forelimb area of the dorsal accessory olive in the cat. Journal of Comparative Neurology 392, 115-133 | [Medline] |
| MEDIRATTA, N. K. & NICOLL, J. A. (1983). Conduction velocities of corticospinal axons in the rat studied by recording cortical antidromic responses. Journal of Physiology 336, 545-561 | [Abstract] |
| MILLER, S., NEZLINA, N. & OSCARSSON, O. (1969). Projection and convergence patterns in climbing fibre paths to cerebellar anterior lobe activated from cerebral cortex and spinal cord. Brain Research 14, 230-233 | [Medline] |
| MOLINARI, H. H. (1984). Ascending somatosensory projections to the dorsal accessory olive: an anatomical study in cats. Journal of Comparative Neurology 223, 110-123 | [Medline] |
| MOLINARI, H. H., SCHULTZE, K. E. & STROMINGER, N. L. (1996). Gracile, cuneate, and spinal trigeminal projections to inferior olive in rat and monkey. Journal of Comparative Neurology 375, 467-480 | [Medline] |
| OSCARSSON, O. (1968). Termination and functional organization of the ventral spino-olivocerebellar path. Journal of Physiology 196, 453-478 | [Medline] |
| OSCARSSON, O. (1980). Functional organisation of the olivary projection to the cerebellar anterior lobe. In The Inferior Olivary Nucleus, ed. COURVILLE, J., DE MONTIGNY, C. & LAMARRE, Y. , pp. 279-289. Raven Press, New York | |
| PORTER, R. & SANDERSON, J. H. (1964). Antidromic cortical response to pyramidal tract stimulation in the rat. Journal of Physiology 170, 355-370 | |
| SAINT-CYR, J. A. (1983). The projection from the motor cortex to the inferior olive in the cat. An experimental study using axonal transport techniques. Neuroscience 10, 667-684 | [Medline] |
| SIMPSON, J. I., WYLIE, D. R. & DE ZEEUW, C. I. (1996). On climbing fibre signals and their consequence(s). Behavioural and Brain Sciences 19, 384-398 | |
| SNIDER, R. S. & STOWELL, A. (1944). Receiving areas of the tactile, auditory and visual systems in the cerebellum. Journal of Neurophysiology 7, 331-357 | |
| STEWART, M., QUIRK, G. J. & AMASSIAN, V. E. (1990). Corticospinal responses to electrical stimulation of motor cortex in the rat. Brain Research 508, 341-344 | [Medline] |
| SUGIHARA, I., LANG, E. J. & LLINAS, R. (1993). Uniform olivocerebellar conduction time underlies Purkinje cell complex spike synchronicity in the rat cerebellum. Journal of Physiology 470, 243-271 | [Abstract] |
| SWENSON, R. S. & CASTRO, A. J. (1983). The afferent connections of the inferior olivary complex in rats. An anterograde study using autoradiographic and axonal degeneration techniques. Neuroscience 8, 259-275 | [Medline] |
| SWENSON, R. S., SIEVERT, C. F., TERREBERRY, R. R., NEAFSEY, E. J. & CASTRO, A. J. (1989). Organization of cerebral cortico-olivary projections in the rat. Neuroscience Research 7, 43-54 | [Medline] |
| TERASHIMA, T. (1995). Course and collaterals of corticospinal fibers arising from the sensorimotor cortex of the reeler mouse. Developmental Neuroscience 17, 8-19 | [Medline] |
| TEUNE, T. M., VAN DER BURG, J., DE ZEEUW, C. I., VOOGD, J. & RUIGROK, T. J. (1998). Single Purkinje cell can innervate multiple classes of projection neurons in the cerebellar nuclei of the rat: a light microscopic and ultrastructural triple-tracer study in the rat. Journal of Comparative Neurology 392, 164-178 | [Medline] |
| VOOGD, J. & GLICKSTEIN, M. (1998). The anatomy of the cerebellum. Trends in Neurosciences 21, 370-375 | [Medline] |
| VOOGD, J. & RUIGROK, T. J. (1997). Transverse and longitudinal patterns in the mammalian cerebellum. Progress in Brain Research 114, 21-37 | [Medline] |
Acknowledgements
We wish to thank the Wellcome Trust for their support. M.R.B. is a member of the University of Cambridge MB/PhD Programme and is supported by the Medical Research Council and by a James Baird studentship.
Corresponding author
S. A. Edgley: Department of Anatomy, Downing Street, Cambridge CB2 3DY, UK.
Email: sae1000{at}cus.cam.ac.uk
This article has been cited by other articles:
|
|
 |
|
  R. Ackerley, J. Pardoe, and R. Apps A novel site of synaptic relay for climbing fibre pathways relaying signals from the motor cortex to the cerebellar cortical C1 zone J. Physiol., October 15, 2006; 576(2): 503 - 518. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Pardoe, S. A. Edgley, T. Drew, and R. Apps Changes in Excitability of Ascending and Descending Inputs to Cerebellar Climbing Fibers during Locomotion J. Neurosci., March 17, 2004; 24(11): 2656 - 2666. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) were maximal 2.8 mm lateral to the midline and the long-latency responses (
) were maximal 4.4 mm lateral to the midline. C, bar chart summarising latencies of responses at the medial and lateral sites to stimuli delivered to the CP and PTr: there are consistent differences between the latencies of responses evoked from these stimulation sites at both recording sites. D, estimated conduction velocities of responses recorded from medial PML, lateral PML and crus II from latency differences such as those shown in C.
, dashed lines) and PTx (
, continuous lines) responses follow a similar distribution.