Afferent projections from the brainstem to the three floccular zones in cats. I. Climbing fiber projections

The cat's flocculus can be divided into 3 zones on the basis of differences in their efferent projection sites. In the present study, climbing fiber projections from the inferior olive to each zone of the flocculus were studied by means of retrograde axonal transport of horseradish peroxidase (HRP). Following large injections of HRP into the flocculus, labeled cells appear in the dorsal cap and the ventrolateral outgrowth of the principal olive. No HRP-labeled somata are present in other parts of the inferior olive. Following microinjections of HRP into the rostral of caudal zones of the flocculus, labeled cells appear in the ventrolateral outgrowth and the rostral part of the dorsal cap, while, after injections into the middle zone, labeled cells are found in the caudal part of the dorsal cap. These findings show that there exists zonal organization in the climbing fiber projections to the flocculus; the rostral and caudal zones receive climbing fiber afferents from the ventrolateral outgrowth and the rostral part of the dorsal cap, while the middle zone receives those from the caudal part of the dorsal cap.     

Mossy fiber projections to the cerebellar flocculus from the extraocular muscle afferents

Inputs from extraocular muscles to the cerebellar flocculus were studied in anesthetized rabbits by recording neural responses either to electric stimulation to the nerve of the extraocular muscle or to the extraocular muscle stretch. The mossy fiber (mf) origin of the responses was identified by the similarity in the laminal profiles of these responses to those of the mossy fiber responses due to eighth nerve stimulation, the capacity to follow repetitive stimulation and the responses of simple spike discharges of Purkinje cells.The latency of mf responses evoked by electric stimulation of the afferent nerve from the superior oblique muscle was about 4.5–8.0 msec. Ramp displacements applied either to an individual extraocular muscle or to the whole eyeball also produced mf responses.Simple spike discharges of flocculus Purkinje cells were modulated by ramp displacements of extraocular muscles. Forty cells out of 47 cells examined responded with excitation and in 7 cells only inhibition was detected. Phasic, tonic and phasictonic excitations were obatained during ramp displacements. Responses of simple spike discharges were classified as direction-specific plane-specific and non-specific types depending on the convergence of afferent sources of muscles producing excitation.Loci in the flocculus responding to extraocular muscle stimulation were studied histologically and it was revealed that mf responses from extraocular muscles were obtained in extensive parts of the flocculus.     

Functional localization in the three floccular zones related to eye movement control in the cat

Anatomically, the cat's cerebellar flocculus can be divided into 3 zones on the basis of differences in their efferent projection sites^13,14. The functional differences of these 3 zones in relation to eye movement control were investigated by observing the eye movements evoked by electric stimulation of each zone of the flocculus in ketamine-anesthetized cats. Stimulation of the flocculus elicited a slow eye movement. The direction of the slow eye movement was mapped. A downward eye movement was evoked by stimulation of the caudal zone. An ipsilateral horizontal eye movement was induced from the middle zone. An upward eye movement was elicited from the rostral zone. When prolonged stimulation was applied to the flocculus, the slow eye movement was followed by nystagmus in the opposite direction. This nystagmus persisted for many seconds after cessation of stimulation (afternystagmus). Nystagmus and afternystagmus could not be elicited in deeply anesthetized cats. Possibilities as to how the stimulation leads to various eye movements are discussed.     

Zonal organization of the floccular Purkinje cells projecting to the group y of the vestibular nuclear complex and the lateral cerebellar nucleus in cats

Injections of horseradish peroxidase (HRP) were made iontophoretically in the group y of the vestibular nucleus and the lateral cerebellar nucleus (LC) in cats. Topographical distribution of labeled Purkinje cells due to the retrograde axonal transport of HRP was studied in the flocculus. It was found that Purkinje cells in the caudal part of the flocculus project mainly to the ipsilateral group y and sparsely to the ventromedial part of the ipsilateral LC.     

Afferent projection from the dorsal nucleus of the raphe to the flocculus in cats

Horseradish peroxidase (HRP) was injected iontophoretically into the cerebellar flocculus in cats. The neurons labeled with HRP were recognized in the caudal part of the dorsal nucleus of the raphe. Electrical stimulation of this region elicited orthodromic evoked potentials in the flocculus and stimulated sites of low threshold current necessary to evoke the responses were limited in a small area corresponding well with the dorsal nucleus of the raphe.     

Organization of pontocerebellar projections to identified climbing fiber zones in the rat

The organization of pontocerebellar projections to the paravermis and hemisphere of the posterior cerebellum of the rat was studied in relation to the organization of climbing fibers. Small injections of cholera toxin subunit B were placed in the cerebellar cortex at locations predetermined by evoked climbing fiber potentials from selected body parts or based on coordinates. The injection site was characterized with respect to the zebrin pattern and by the distribution of retrogradely labeled neurons in the inferior olive. The following zones were studied: hindlimb-related zones C1 and C2 of lobule VIII; forelimb-related zones C1, C2, and D0/D1 of the paramedian lobule; and face-related zones A2 of the paramedian lobule and C2 and D0 of crus 2B. The results show that the distribution of pontine neurons is closely related to the climbing fiber somatotopy. Injections centered on face-related zones result in distribution of pontine neurons within the pontine core region. Forelimb regions surround this core, whereas hindlimb regions are mostly supplied by caudal pontine regions and by a single patch of more rostrally located neurons. This distribution fits well with published data on the somatotopy of the corticopontine projection from the rat primary somatosensory cortex. However, apart from differences in the participation of ipsilaterally projecting cells, the distribution of pontine neurons does not change significantly when the injection covers different zones of the same lobule such as C1 and C2 of lobule VIII; C1, C2, and D0/D1 of the paramedian lobule; A2 of the paramedian lobule; and C2 and D0 of crus 2B.     

The fine topography of climbing fiber projections to the paramedian lobule of the cerebellum in the cat

Motor cortex and somatosensory afferents of climbing fibers (CFs) which terminate in the two rostral folia of the paramedian lobule of the cerebellum were studied in pentobarbital-anesthetized cats. CF responses were elicited in all but one Purkinje cell recorded within the paramedian lobule by stimulation of one or more pericruciate sites; 67% of these also had a peripheral receptive field or were excited by stimulation of the superior radial, sciatic, or infraorbital nerve. CFs responded to stimulation of peripheral nerves at either short latency (14 to 34 ms) or long latency (<120 ms), or gave a mixture of both responses. Those responding at long latency generally did not have peripheral receptive fields. The forelimb was predominantly represented in folia a and b, although it was noted that the relationship between bodily representation and surface landmarks was variable between cats. The proximal portion of the limb was represented medially and the distal part laterally. A clear relation was established between the location of the lowest-threshold cortical site and the peripheral afferents which evoked a CF response in any given cell. The cortical site was generally in a region controlling movements of the body parts on which the peripheral receptive field was found. Two populations of CFs exist in this region, one receiving convergent and complementary cortical and peripheral information and another which, under the present experimental conditions, is excited only from the cortex. It is suggested that these two populations may project as interdigitating bands which run across the long axis of the folia.     

Afferent projections to the caudal part of the dorsal nucleus of the raphe in cats

Horseradish peroxidase (HRP) was iontophoresed into the caudal part of the dorsal nucleus of the raphe (DNR) in cats. Labeled neurons with HRP were recognized in the medial and the superior vestibular nucleus. Electrical stimulation of the medial or the superior vestibular nucleus elicited orthodromic evoked potentials and unitary responses in the caudal part of the DNR. These projections may be involved in eye movement control. In addition, labeled neurons were located in the magnocellular division of the alaminar spinal trigeminal nucleus. This projection may be a part of the pathway conveying somatosensory inputs from the face to the cerebellar flocculus.     

Electrophysiological study of the nucleus of the optic tract that transfers optic signals to the nucleus reticularis tegmenti pontis—the visual mossy fiber pathway to the cerebellar flocculus

Neuronal activities (n = 43) in the pretectal region in rabbits were recorded. They were orthodromically activated from the optic chiasm (latency, 1.86 ± 0.35msec) and antidromically from the ipsilateral nucleus reticularis tegmenti pontis (Nrt) (latency, 0.97 ± 0.22sec). Thirty-one (72%) neurons were in the nucleus of the optic tract (NOT), four (9%) in the anterior pretectal nucleus (PA) and seven (16%) in the border between NOT and PA. These findings demonstrate that the NOT is involved in the visual mossy fiber pathways to the flocculus and may contribute to optokinetic eye movements.     

Vestibular afferents of the inferior olive and the vestibulo-olivo-cerebellar climbing fiber pathway to the flocculus in the cat

The projection of the vestibular nuclei to the inferior olive was investigated by means of anterograde transport of tritiated leucine. Following injections in the medial and descending vestibular nuclei, terminal labeling was found ipsilaterally in the dorsomedial cell column, subnucleus beta and the caudal medial accessory olive, while the latter also received afferents from the nucleus prepositus hypoglossi. At the contralateral side termination in the dorsomedial cell column and the medial accessory olive was found after injections in the nucleus vestibularis superior and group Y. The ventrolateral outgrowth and different parts of the principal olive also received afferents from these two nuclei and also from ventral parts of the lateral cerebellar nucleus. The dorsal cap was labeled exclusively from the contralateral nucleus prepositus hypoglossi. The termination in the inferior olive of the vestibular afferents is compared with the projection from a number of pretectal nuclei. Furthermore the consequences of the divergence and convergence of both types of projections at the level of the inferior olive is discussed in relation to the subsequent climbing fiber projection to the flocculus.     

Neuronal pathway from floccular caudal zone contributing to vertical eye movements in cats—role of group y nucleus of vestibular nuclei

In ketamine-anesthetized cats, electric microstimulation of the group y nucleus of the vestibular nuclei evoked a slow and smooth upward eye movement. Destruction of the group y nucleus eliminated a slow and smooth downward eye movement evoked by stimulation of the caudal zone of the flocculus. These data support the interpretation that Purkinje cell activity in the caudal zone of the flocculus can evoke vertical eye movements by inhibiting the activity of neurons of the group y nucleus.     

Mossy fiber projections from the cuneate nucleus to the cochlear nucleus in the rat

A reciprocal connection is known to exist between the cuneate nucleus, which is a first-order somatosensory nucleus, and the cochlear nucleus, which is a first-order auditory nucleus. We continued this line of study by investigating the fiber endings of this projection in the cochlear nucleus of rats using the neuronal tracer Phaseolus vulgaris leucoagglutinin in combination with ultrastructural and immunocytochemical analyses. In the cochlear nucleus, mossy fiber terminals had been described and named for their morphologic similarity to those in the cerebellum, but their origins had not been discovered. In the present study, we determined that the axonal projections from the cuneate region gave rise to mossy fiber terminals in the granule cell regions of the ipsilateral cochlear nucleus. The cuneate mossy fibers appear to be excitatory in nature, because they are filled with round synaptic vesicles, they make asymmetric synapses with postsynaptic targets, and they are labeled with an antibody to glutamate. The postsynaptic targets of the mossy fibers include dendrites of granule cells. This projection onto the granule cell interneuron circuit of the cochlear nucleus indicates that somatosensory cues are intimately involved with information processing at this early stage of the auditory system. © 1996 Wiley-Liss, Inc.     

Collateral projections from vestibular nuclear and inferior olivary neurons to lobules I/II and IX/X of the rat cerebellar vermis: a double retrograde labeling study

Axon collateral projections to various lobules of the cerebellar cortex are thought to contribute to the coordination of neuronal activities among different parts of the cerebellum. Even though lobules I/II and IX/X of the cerebellar vermis are located at the opposite poles in the anterior–posterior axis, they have been shown to receive dense vestibular mossy fiber projections. For climbing fibers, there is also a mirror‐image‐like organisation in their axonal collaterals between the anterior and posterior cerebellar cortex. However, the detailed organisation of mossy and climbing fiber collateral afferents to lobules I/II and IX/X is still unclear. Here, we carried out a double‐labeling study with two retrograde tracers (FluoroGold and MicroRuby) in lobules I/II and IX/X. We examined labeled cells in the vestibular nuclei and inferior olive. We found a low percentage of double‐labeled neurons in the vestibular nuclei (2.1 ± 0.9% of tracer‐labeled neurons in this brain region), and a higher percentage of double‐labeled neurons in the inferior olive (6.5 ± 1.9%), especially in its four small nuclei (18.5 ± 8.0%; including the β nucleus, dorsal cap of Kooy, ventrolateral outgrowth, and dorsomedial cell column), which are relevant for vestibular function. These results provide strong anatomical evidence for coordinated information processing in lobules I/II and IX/X for vestibular control.     

The projection of the nucleus reticularis tegmenti pontis and adjacent regions of the pontine nuclei to the central cerebellar nuclei in the cat

The projection of the nucleus reticularis tegmenti pontis (NRTP) and the pontine nuclei (NP) to the central cerebellar nuclei (CCN) was investigated by means of anterograde transport of tritiated leucine. Although termination was found in all the CCN, it was most pronounced in the lateral nucleus and the lateral aspect of the posterior interposed nucleus. The extreme lateral aspect of the anterior interposed nucleus and the caudal part of the fastigial nucleus received a projection of modest intensity. Termination in the infracerebellar nucleus and group Y is likely to be present but could not be confirmed with certainty from the light microscopical material. The contribution from the NP was small and originated from the dorsolateral and dorsal paramedian subdivisions of the NP. Within the NRTP the total area giving rise to projections to the CCN was extensive, and the origin of the projections to the individual CCN overlapped considerably. The projection of the NRTP to the ventrocaudal part of the lateral nucleus was found in conjunction with a projection to the ventrolateral part of the posterior interposed nucleus. Both projections seemed to branch off the fiber bundle terminating in the ventral paraflocculus. Similar correlations could be established in the projection of the NRTP to the dorsal paraflocculus and crus II of the ansiform lobule with other parts of the lateral and posterior interposed nuclei. It was concluded that the transverse, lobular organization of mossy fibers, which differs fundamentally from the longitudinal, modular organization of climbing fibers, is maintained in the collateral projection to the CCN. The results are further discussed in relation to the corticonuclear projection and the engagement of the NRTP and different parts of the CCN in pontocerebellar circuits.     

Afferent projections to the ventral lateral geniculate nucleus in the cat

Horseradish peroxidase (HRP) injections were made into the dorsal lateral geniculate nucleus (LGN_d) and ventral lateral geniculate nucleus (LGN_v) of the cat in order to define afferent projections to LGN_v. These were found from the superior colliculus, contralateral LGN_v, dorsal median raphe nucleus, locus coeruleus, ipsilateral pretectum, and various portions of visual cortex. While many cortical areas project to LGN_v (17, 18, 19, 21 and lateral suprasylvian), the heaviest input arises from areas 17 and 20. The cell bodies of origin are in layer 5 in contrast to layer 6 which projects to LGN_d.     

Modification of the projection from the sensory cortex to the motor cortex following the elimination of thalamic projections to the motor cortex in cats

Examination of the projection from area 2 of the sensory cortex to the motor cortex revealed substantial changes following lesion of the ventrolateral nucleus of the thalamus. These observed changes were as follows. (1) The polarity of the evoked potentials elicited by area 2 stimulation reversed in the depth of the motor cortex whereas in normal animals, there was no reversal. (2) The amplitude of area 2-elicited EPSPs in the motor cortical neurons became greater following the lesion of VL. (3) The shape of the observed EPSPs was characterized by multiple peaks whereas in normal animals, the EPSPs were generally smooth and monophasic. (4) Neurons receiving a short-latency input from area 2 were distributed throughout the depths of the motor cortex whereas in normal animals, they were located only in the upper layers (layers II and III). (5) Intracellular injection of HRP revealed that the neurons receiving short-latency input were not restricted to typical stellate type cells, but also included bipolar or bitufted neurons with elongated cell bodies and polarized arborizations. These neurons were located in the superficial (II and III) as well as in the deep (V) layer. It is concluded that the elimination of thalamic input resulted in the reinforcement of the corticocortical input to the motor cortex. The subsequently observed corticocortical projection extended to neurons did not originally innervated by the association fibers. The results suggested that functional recovery following thalamic lesion is partly due to reorganization of projections from the sensory cortex to the motor cortex.     

Rostro-caudal distribution of reticulospinal projections from different brainstem nuclei in the lamprey

The reticulospinal (RS) system in the lamprey is responsible for the control of locomotion, postural corrections and steering. To perform these functions, the RS system has to affect different muscular compartments along the body axis selectively. In this study, the possibility that RS neurones in different nuclei may project to different parts of the spinal cord, was investigated. The rostro-caudal extent of single RS axons was defined by stimulating them antidromically while recording from their cell body. All recorded mesencephalic RS neurones projected to the caudal tip of the spinal cord. Of the rhombencephalic RS neurones, 26% of the recorded neurones did not reach the caudalmost fourth of the spinal cord and this proportion varied between the anterior (18%), middle (17%) and posterior (36%) rhombencephalic reticular nuclei. For these RS axons, the level of termination covered the whole rostro-caudal extent of the spinal cord. No correlation was found between the length of an axon and its conduction velocity or between the length of an axon and the rostro-caudal position of its cell body in the nuclei.     

Trigeminocerebellar projections to the posterior lobe in the cat, as studied by anterograde transport of wheat germ agglutinin-horseradish peroxidase.

Cerebellar projections of the nucleus interpolaris and oralis of the spinal trigeminal nucleus were studied in the cat by anterograde transport of wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP). Injections of WGA-HRP into these nuclei labeled many mossy fiber terminals mainly ipsilaterally in the rostral folium of lobule IX (IXa or IXa + b), the simple lobule, the anterior part (sublobule A) of the paramedian lobule and the posterior part of crus II. Labeled terminals were also seen in the anterior lobe, lobules VI and VII, the anterior part of crus I, and the paraflocculus dorsalis. Projection fields in the horizontal plane of lobules were reconstructed from a series of transverse sections through each folium of lobule IX, the paramedian lobule, and the posterior part of crus II on the ipsilateral side. In sublobule IXa + b, labeled terminals were distributed in five longitudinal areas extending along the apicobasal axis of the sublobule. These five areas were located in the apical two-thirds of the ipsilateral half of the sublobule. Labeled terminals were distributed in five longitudinal areas in sublobule A (the rostral part) of the paramedian lobule. In the posterior part of crus II, four aggregations of labeled terminals were present in cross sections through a lobule. They were distributed in the apicobasal extent of the lobules. The present study indicates that the projection fields of trigeminocerebellar fibers are longitudinally arranged along the apicobasal axis of the cerebellar lobules.     

Neural activity of nucleus reticularis tegmenti pontis — the origin of visual mossy fiber afferents to the cerebellar flocculus of rabbits

Activities of 55 neurons were extra- and intracellularly recorded in the nucleus reticularis tegmenti pontis (Nrt) of anesthetized rabbits. The cells were antidromically activated from the flocculus as well as orthodromically from the optic tracts. They were antidromically activated from either the ipsilateral or contralateral flocculus (latency, 0.94 msec) but not from the flocculi on both sides. There was no preference for projection to either the ipsilateral or contralateral flocculus. Twenty out of 55 neurons were orthodromically activated from the ipsilateral and 15 neurons from the contralateral optic tract. The remaining 20 neurons were excited from the optic tract on both sides. The orthodromic latencies were all in the same range with a mean of 4.7 msec. These findings demonstrate the existence of neurons in the Nrt that transfer visual signals to the flocculus through mossy fiber afferents. The indicate that Nrt neurons may contribute to cotrolling eye movements through the flocculus.     

Pontine and lateral reticular projections to the c<Subscript>1</Subscript> zone in lobulus simplex and paramedian lobule of the rat cerebellar cortex

Spatial localization and axonal branching in mossy fiber projections to two rostrocaudally-separated regions of the 'forelimb' c1 zone in lobulus simplex and paramedian lobule were studied in rats using a retrograde double-labelling tracer technique. In four animals the two cortical regions were localized electrophysiologically and each was micro-injected with tracer material, yielding a total of eight different cases. Single- and double-labelled cell bodies were plotted in the basal pontine nucleus (BPN), nucleus reticularis tegmenti pontis (NRTP), and the lateral reticular nucleus (LRN). As a control, cells labelled in the contralateral inferior olive were also counted. The parts of the c1 zone in lobulus simplex and the paramedian lobule were found to receive mossy fiber inputs from similar regions of BPN, NRTP and LRN. Double-labelled cells were not found in NRTP but were present in BPN and LRN (on average 6% and 25% of the smaller single-labelled population, respectively). The incidence of double-labelled cells in the olive and LRN was positively correlated, but no relation was found between olive and BPN, suggesting a zonal organization within the mossy fiber projections from LRN, but not from the pons. In quantitative terms, the c1 zone in lobulus simplex received a greater density of mossy fiber projections from BPN, NRTP and LRN than the c1 zone in the paramedian lobule. This suggests that the two parts of the same cerebellar cortical zone differ, at least partially, in regard to their inputs from three major sources of mossy fibers. This is consistent with the modular hypothesis and could enable a higher degree of parallel processing and integration of information within different parts of the same zone.