Structural organization of the fastigial nucleus. I. Dendrites and axonal pathways

Modular output circuits of the fastigial nucleus for diverse motor and nonmotor functions of the cerebellar vermis

The cerebellar vermis, long associated with axial motor control, has been implicated in a surprising range of neuropsychiatric disorders and cognitive and affective functions. Remarkably little is known, however, about the specific cell types and neural circuits responsible for these diverse functions. Here, using single-cell gene expression profiling and anatomical circuit analyses of vermis output neurons in the mouse fastigial (medial cerebellar) nucleus, we identify five major classes of glutamatergic projection neurons distinguished by gene expression, morphology, distribution, and input-output connectivity. Each fastigial cell type is connected with a specific set of Purkinje cells and inferior olive neurons and in turn innervates a distinct collection of downstream targets. Transsynaptic tracing indicates extensive disynaptic links with cognitive, affective, and motor forebrain circuits. These results indicate that diverse cerebellar vermis functions could be mediated by modular synaptic connections of distinct fastigial cell types with posturomotor, oromotor, positional-autonomic, orienting, and vigilance circuits.

Integration of multiple motor segments for the elaboration of locomotion: role of the fastigial nucleus of the cerebellum

This chapter provides a conceptual overview of the role and operation of higher structures of the central nervous system (CNS) in the control of posture and locomotion in the mammal, including the nonhuman primate and the human. Both quadrupedal and bipedal locomotion require the integrated neural control of multiple body segments against gravity. During development, and in selected instances in the adult, motor learning is required, particularly for merging anticipatory and reactive CNS processes, the latter being necessary after tripping and stumbling. We have recently found that the fastigial nucleus (FN) of the cerebellum in the cat plays a particularly important role in the control of locomotion, by virtue of its critical position in uniting the cerebro-cerebellar and the spino-cerebellar loops of neural activity that participate in the integrated control of multiple body segments. Further understanding of the CNS structures that achieve this integration has come from our recent study of an intact nonhuman primate, the Japanese monkey, Macaca fuscata, as it learns to elaborate bipedal locomotion rather than its normal quadrupedal fashion. Based on findings from these two animal species, we now present a model of the overall integrated control of posture and locomotion that features the combined operation of parallel and distributed neural circuitry throughout the CNS.

Instigation and control of treadmill locomotion in high decerebrate cats by stimulation of the hook bundle of Russell in the cerebellum

In high decerebrate cats, pulse train microstimulation of a restricted region of the midline cerebellar white matter produced a generalized increase in postural muscle tone in the neck, trunk, and limb extensor muscles, and air-stepping of all four legs on a stationary surface. On the moving belt of a treadmill, such stimulation produced well coordinated, fore- and hindlimb locomotion as evoked by stimulating the mesencephalic locomotor region (MLR). Microinjection of a neural tracer into the cerebellar locomotion-inducing site resulted in a bilateral retrograde labeling of cells limited to the fastigial nuclei simultaneously with anterograde labeling of fibers projecting bilaterally to the medial pontomedullary reticular formation (mPMRF) the vestibular complex and upper cervical segments. These results have led to our proposition that the effective cerebellar locomotor region (CLR) corresponds to the midline region of the hook bundle of Russell. Passing through this structure are crossed fastigioreticular and fastigiovestibular fibers, together with fastigiospinal fibers. Subsequently, we showed that CLR stimulation resulted in simultaneous short-latency synaptic activation of long-descending reticulospinal and vestibulospinal cells with high synaptic security. Clearly, the fastigial nucleus possesses potential capability to recruit and regulate posture- and locomotor-related subprograms which are distributed within the brainstem and spinal cord by the in-parallel activation of fastigiospinal, fastigioreticular, and fastigiovestibular pathways.Key words: cerebellar locomotor region (CLR), fastigial nucleus, hook bundle of Russell, reticulospinal cell, vestibulospinal cell.

Stimulation of a restricted region in the midline cerebellar white matter evokes coordinated quadrupedal locomotion in the decerebrate cat

In the reflexively standing acute decerebrate cat, we have previously shown that pulse train microstimulation of the hook bundle of Russel in the midline of the cerebellar white matter, through which crossed fastigiofugal fibers decussate, augments the postural tone of neck, trunk, fore-, and hindlimb extensor muscles. In the present study we examined the possible role of such stimulation in evoking locomotion as the animal is supported by a rubber hammock with its feet contacting the moving surface of a treadmill. We were able to provoke well-coordinated, bilaterally symmetrical, fore- and hindlimb movements, whose cycle time and pattern were controlled by appropriate changes in stimulus intensity and treadmill speed. We carefully and systematically mapped this cerebellar locomotor region (CLR) through repeated dorsoventral penetrations with a glass-coated tungsten microelectrode in a single animal and between animals. We found that the optimal locus for evoking locomotion was centered on the midline, at Horsley-Clarke coordinates H0 and P7.0, and extended over a rostrocaudal and dorsolateral range of approximately 0.5 mm. The lowest effective stimulus intensity at the optimal site was in the range of 5-8 microA. Along penetration tracks to left or right of the midline, effective stimulus intensity increased and evoked locomotor patterns were no longer symmetrical, but rather shifted toward the contralateral limbs. In the same animals, controlled locomotion was evoked by stimulating the mesencephalic locomotor region (MLR). With concomitant stimulation of the optimal sites in the CLR and the MLR, each at subthreshold strength, locomotor movements identical to those seen with suprathreshold stimulation of each site alone were evoked. With concomitant stimulation at suprathreshold strength for each site, locomotion became vigorous, with a shortened cycle time. After making ablative lesions at either the CLR or MLR (unilateral or bilateral), controlled locomotion was still evoked at the prior stimulus strength by stimulating the remaining site. Together, these results demonstrate that selective stimulation of the hook bundle of Russel in the midsagittal plane of the cerebellar white matter evokes "controlled" locomotion identical to that evoked by stimulating the MLR. We have shown that the fastigial nucleus is one of the supraspinal locomotion inducing sites and that it can independently and simultaneously trigger brain stem and spinal locomotor subprograms formerly believed to be the domain of various brain stem regions including the MLR and the subthalamic locomotor region.

Cerebellar-induced locomotion: reticulospinal control of spinal rhythm generating mechanism in cats

In a decerebrate cat (locomotor preparation), stimulation of a restricted region along the midline cerebellar white matter has been found to evoke generalized augmentation of postural muscle tone on a stationary surface (Asanome et al. 1998. Neurosci. Res. 30: 257-269) and "controlled" locomotion on the surface of a moving treadmill. Characteristics of cerebellar-evoked locomotion were similar to those of mesencephalic locomotor region-evoked "controlled" locomotion on the same animal. Microinjection of a neural tracer (CTb-HRP) into the lesioned stimulus site of the cerebellar white matter resulted in both retrograde labelling of cells in the fastigial nuclei, bilaterally, and anterograde labeling of fibers descending to the brain stem. These results indicated that the effective cerebellar stimulus site (cerebellar locomotor region) corresponded to the midline region of the hook bundle of Russell (Rasmussen, A. T., 1933. J. Comp. Neurol. 57: 165-197), through which crossed fastigioreticular, fastigiovestibular, and fastigiospinal fibers pass. In this study, contribution of reticulospinal systems to the control of cerebellar-evoked locomotion was extensively studied. By stimulating the cerebellar locomotor region and the MLR in the same animal, a majority of antidromically identified pontomedullary reticulospinal cells were synaptically activated. The results of the present study demonstrated that fastigial cells with crossed fastigioreticular fibers and reticulospinal fibers play a crucial role in the control of posture and locomotion in the locomotor preparation.

Augmentation of postural muscle tone induced by the stimulation of the descending fibers in the midline area of the cerebellar white matter in the acute decerebrate cat

In a reflexively standing acute decerebrate cats, the cerebellar white matter was systematically stimulated and the effects on the level of postural muscle tone were studied. A stimulating microelectrode was placed systematically at 0.1-0.5 mm increments from H + 2 to H - 2 at levels ranging from P7.0 to P8.0 rostrocaudally and mediolaterally from LR0 to L1.5 or R1.5. Stimuli delivered to the restricted region of the cerebellar white matter along its midline resulted in simultaneous and bilateral augmentation of tonic activities in the neck, lumbar back, fore- and hindlimb extensor muscles along with increased levels in the forces exerted by each of the left and the right fore- and hindlimbs. Effective stimulus regions were located in the cerebellar white matter rostral and ventral to the most rostral part of the fastigial nucleus. Microinjection of a retrograde neural tracer, cholera-toxin b subunit conjugated horseradish peroxidase (CTb-HRP), into the lesioned effective stimulus sites resulted in a retrograde labeling of cells in the fastigial nuclei, bilaterally. All these results suggest that the augmentation of postural muscle tone was evoked by a selective activation of fastigiofugal fibers which course through the 'hook bundle'.

GABA-ergic transmission in deep cerebellar nuclei

gamma-Aminobutyric acid (GABA) is the inhibitory transmitter released at Purkinje cell axon terminals in deep cerebellar nuclei (DCN). Neurons in DCN also receive excitatory glutamatergic inputs from the inferior olive. The output of DCN neurons, which depends on the balance between excitation and inhibition on these cells, is involved in cerebellar control of motor coordination. Plasticity of synaptic transmission observed in other areas of the mammalian central nervous system (CNS) has received wide attention. If GABA-ergic and/or glutamatergic synapses in DCN also undergo plasticity, it would have major implications for cerebellar function. In this review, literature evidence for GABA-ergic synaptic transmission in DCN as well as its plasticity are discussed. Studies indicate that fast inhibitory postsynaptic potentials (IPSPs) and currents (IPSCs) in neurons of DCN are mediated by GABAA receptors. While GABAB receptors are present in DCN, they do not appear to be activated by Purkinje cell axons. The IPSPs undergo paired-pulse, as well as frequency-dependent, depressions. In addition, tetanic stimulation of inputs can induce a long-term depression (LTD) of the IPSPs and IPSCs. Excitatory synapses do not appear to undergo long-term potentiation or LTD. The LTD of the IPSP is not input-specific, as it can be induced heterosynaptically and is associated with a reduced response of DCN neurons to a GABAA receptor agonist. Postsynaptic Ca2+ and protein phosphatases appear to contribute to the LTD. The N-methyl-D-aspartate receptor-gated, as well as the voltage-gated Ca2+ channels are proposed to be sources of the Ca2+. It is suggested that LTD of GABA-ergic transmission, by regulating DCN output, can modulate cerebellar function.

Projections from the central cervical nucleus to the cerebellar nuclei in the rat, studied by anterograde axonal tracing

Projections from the central cervical nucleus (CCN) to the cerebellar nuclei were examined following injections of Phaseolus vulgaris-leucoagglutinin or cholera toxin subunit B into the C1-C3 segments in the rat. Labeled axons and terminals were immunohistochemically demonstrated. Labeled spinocerebellar fibers arising from the CCN entered the cerebellum through the inferior and the superior cerebellar peduncles. Labeled mossy fiber terminals were seen in lobules I-VI, sublobule VIIb, lobules VIII and IX, and the copula pyramidis of the cerebellar cortex. Labeled axons ran toward the cerebellar cortex, through and between the medial and the interpositus nuclei, and gave off collateral axons and terminal axons to the cerebellar nuclei. The projections to the cerebellar nuclei were predominantly contralateral to the cells of origin. Labeled terminals were distributed from the medial to the ventrolateral part of the middle subdivision of the medial nucleus throughout its rostrocaudal extent. Labeled terminals were also seen in the lateral part of the medial nucleus and in the border region between the medial nucleus and the interpositus nuclei, which corresponds to the rostromedial extension of the posterior interpositus nucleus. In the anterior interpositus nucleus, labeled terminals were distributed dorsoventrally in the middle third of the mediolateral extent. They were more numerous in the rostrodorsal part of this area. Labeled terminals were distributed dorsally and caudally in the medial third of the posterior interpositus nucleus. No labeled terminals were seen in the caudomedial subdivision and the dorsolateral protuberance of the medial nucleus, the dorsolateral hump region and the lateral nucleus. The present study demonstrates that the CCN projects to specific areas of the cerebellar cortex and the medial and the interpositus nuclei.

Abnormal cerebellar output in rats with an inherited movement disorder

Biochemical and metabolic mapping techniques have consistently identified the deep cerebellar nuclei (DCN) of the genetically dystonic rat as a site of abnormality. Extracellular single-unit recording techniques were used to assess the functional significance of these findings in affected rats and normal littermates between 16 and 25 days of age. Cells in the medial nucleus of the mutant rats had significantly increased spontaneous firing rates in comparison with cells from normal rats. In both the medial and the interpositus nuclei, cells from the mutants fired more rhythmically than those from the normal rats. When harmaline was administered systemically to activate the olivo-cerebellar system, in normal rats, increased firing rate and bursting patterns of activity were seen. There was no reliable change in the average firing rate or rhythmicity of cells in the medial nucleus of the dystonic rats, although previous studies have shown that harmaline activates neurons in the inferior olive in the mutants. It is likely that naturally stimulated olivary activity also fails to modulate cerebellar output in this model of inherited movement disorder. Anatomical studies did not reveal any consistent changes in the number of Purkinje cells, the volume of the DCN, or the soma size of DCN neurons. Since the electrophysiological findings cannot be ascribed to a loss of the Purkinje cells that normally provide an inhibitory input to the cerebellar nuclei, the results of this study indicate the presence of a functional defect in the control of cerebellar output in the dystonic rat that accounts for the failure of these animals to display harmaline tremor and which may be critical to the motor syndrome.

Olivocerebellar and cerebelloolivary connections of the oculomotor region of the fastigial nucleus in the macaque monkey

Anatomical connections of the caudal portion of the fastigial nucleus (FN) with the inferior olive (IO) were studied in macaque monkeys with wheat-germ-agglutinin-conjugated horseradish peroxidase (WGA/HRP) and HRP. When injected HRP was confined to a caudal portion of the FN, retrogradely labeled Purkinje cells (P cells) appeared in the oculomotor vermis. We defined the area that receives the projection from vermal lobule VII as the fastigial oculomotor region. The same HRP injection resulted in retrograde labeling of IO neurons in an area of group b (of Bowman and Sladek: J. Comp. Neurol. 152:299-316, '73) of the contralateral medial accessory olive (MAO). This area was designated as the Z-portion because in the coronal section it appears like the letter "Z." Retrogradely labeled IO neurons were also found in the Z-portion when HRP was injected into the oculomotor vermis, indicating that neurons in this portion project to both the fastigial and vermal oculomotor regions. Anterogradely labeled axons from the contralateral fastigial oculomotor region also terminated in the Z-portion. When the effective site included a region anterior to the fastigial oculomotor region, labeled P cells appeared in lobule V and labeled IO neurons appeared in group a. Labeled terminals of fastigial fibers were also found in group a. When the effective site included a region ventral to the oculomotor region, labeled P cells appeared in vermal lobules VIII and IX and labeled IO neurons appeared in caudal parts of a and b, in addition to group c. HRP injection into the posterior interposed nucleus (PIN) resulted in labeling of P cells in the paravermal zone and of IO neurons in the rostral two-thirds of the MAO and the dorsal accessory olive (DAO). The location of the labeled terminals coincided with the region where the densest labeling of IO neurons was found. Thus, the olivary projections to both the cerebellar cortex and deep cerebellar nuclei and the nucleoolivary projection exhibited a closely related topographical organization.

Autoradiography of mossy fiber terminals in the fastigial nucleus of the cat

Terminal boutons of mossy fiber collaterals in the fastigial nucleus originating from the nucleus reticularis tegmenti pontis, the bulbar reticular formation, and the medial vestibular nucleus were studied with high-resolution autoradiography in order to examine their ultrastructural features and synaptic relations. Labeled mossy fiber boutons ranged in size from 0.5 to 5 micron in diameter, and they all contained clear and spherical vesicles in an electron-lucent matrix, mitochondria, and some fine tubular elements. These boutons form asymmetric synapses with dendritic profiles of different sizes. No evidence was found for mossy fiber termination on the soma of fastigial neurons. Two types of mossy fiber terminals were distinguished on the basis of the aggregation of synaptic vesicles: one type with clustered vesicles and one type with densely packed vesicles, occurring in equal number from all sources. Furthermore, the applicability of the congruity hypothesis is confirmed for the general identification of terminals.

Localization of serotonin immunoreactivity in the opossum cerebellum

We have used the indirect antibody peroxidase-antiperoxidase technique to analyze the course of serotonin (5-hydroxytryptamine; 5HT) fibers to the deep cerebellar nuclei; the distribution of serotonin within the nuclei; the continued course of 5HT fibers to the cerebellar cortex; and the lobular and laminar distribution of this indoleamine in the cerebellar cortex. Only rarely are fibers found in either the restiform body or the brachium pontis. However, a distinct bundle of serotoninergic axons is present in the medial aspect of the brachium conjunctivum. Axons arise from this bundle and course dorsally into the neuropil of the deep cerebellar nuclei. The densest immunostaining is present in posterior and ventral regions of all four cerebellar nuclei. Within the nuclei large (24% of total) and small (76% of total) varicosities are present. The average distance between varicosities on individual axons is 3.85 micron (S.D. = 1.2). The innervation of the cerebellar cortex is derived primarily from fibers that course through the deep nuclei. At levels caudal to the deep nuclei a single midsagittal band courses into lobules VIII and IX. In the cerebellar cortex, serotoninergic axons and varicosities are present in all lobules; however, the fiber density is not uniform. The densest distribution is present in vermal lobule VIII and the dorsal folia of lobule IX. Within the granule cell layer of lobules VIII and IX, immunoreactive elements form a midsagittal band, and to a lesser degree, two parasagittal bands. Beaded serotoninergic fibers course through the deep and middle portion of the granule cell layer and give rise to a plexus at the border between the Purkinje cell and granule cell layers. Within this plexus axons extend long distances in the transverse and sagittal planes. Long beaded axons oriented in the transverse plane of the folia are also present in the deep molecular layer. A few radial serotoninergic fibers ascend to the pial surface and give rise to very short tangential branches. In all three cortical layers, both large (19% of total) and small (81% of total) varicosities are present. The average distance between varicosities on individual fibers is 5.3 micron (S.D. = 2.2).(ABSTRACT TRUNCATED AT 400 WORDS)

The perineuronal net in the fastigial nucleus of the rat cerebellum. A Golgi and quantitative study

The morphological study of the rat fastigial nucleus with the Golgi-Rio Hortega method showed the presence of glial perineuronal nets surrounding the large neurons, but not the small ones. This perineuronal net appeared as a mesh of tenuous glial processes which covers the neuronal perikarya and proximal dendrites. The small alveolate compartments in this mesh seem to correspond to the holes for the synaptic boutons. Our results also indicate that the perineuronal net is derived from interneuronal protoplasmic and velate astrocytes. Using camera lucida drawings of this perineuronal net we have made a quantitative estimation of the size and density of synaptic boutons on these large neurons. The average numerical density of synaptic boutons was about 19 per 100 micron 2 of the neuronal surface, the mean area of the synaptic holes being 2.5 micron 2. Furthermore, the quantitative data evidence that about 52.5% of the neuronal surface is presumably occupied by synaptic boutons whereas the remaining 47.5% is covered by the glial processes of the perineuronal net. Semithin sections prepared from thick Golgi sections were used for the cytological study of the neurons surrounded by this glial pericellular network. The possible functional significance of the perineuronal net in the regulation of synaptic transmission in the fastigial cerebellar nucleus is briefly discussed.

Fastigial unit activity during voluntary movement in primates

To examine the proposition that fastigial n. of cerebellum provides a fast feedback pathway to suprasegmental structures for spinal information regarding movement. Macaca irus were trained to make flexion and/or extension voluntary wrist movements. Fastigial neuron activity was then correlated with force, velocity, handle position and with shoulder or forearm muscle activity. From 200 units (75% participating), our data establish that fastigial neurons are uniformly recruited after movement onset, processing force-velocity information. Units were found specifically correlated with force or velocity alone. Fastigial activity is strategically placed to provide a 'correction' signal path for motor performance.

The activity of cerebellar nuclear neurones in relation to stimuli which evoke a pectoral fin reflex in dogfish

Extracellular single-unit recordings from the cerebellar nucleus were classified into type I and type II units on the basis of their spontaneous discharges. Type I units discharged at a very regular frequency, giving interspike interval histograms with narrow distributions. Type II units had irregular discharges. Type I units were identified as cerebellar nuclear units by their antidromic responses to stimulation of the contralateral brachium conjunctivum (b.c.) in the mid-brain and by their inhibitory responses to stimulation of the cerebellar cortex. Type II units were not driven antidromically by b.c. stimulation but were inhibited by stimulating the cerebellar cortex. Activity of the nuclear neurones was monitored following subcutaneous electrical stimulation of a fin that elicits a reflex elevation. 67% of units responded, the majority with an increased discharge frequency (excitation, 59%) but some with a decreased discharge frequency ('inhibition', 8%). Latencies of both excitatory and inhibitory responses were greater than 50-400 msec. Most excitatory responses lasted for at least 500 msec; several lasted for more than 10 sec. Inhibitory responses lasted for about 500 msec. With units tested by bilateral fin stimulation, the same qualitative response was obtained whichever fin was stimulated. These results are discussed in relation to the known responses of cerebellar Purkinje cells recorded under similar experimental conditions.

A light microscopic investigation of the afferent connections of the lateral reticular nucleus in the cat

The topographical organization of the projections to the lateral reticular nucleus (LRN) of the cat was investigated using the horseradish peroxidase (HRP), silver-impregnation and autoradiographic tracing methods. Following injection of HRP into the LRN, labelled cells were found mainly within Rexed's laminae VII and VIII of the spinal cord, the contralateral red nucleus, the ventro-rostral aspect of the contralateral fastigial nucleus and the contralateral anterior sigmoid and coronal gyri of the cerebral cortex. Animals with injections of tritiated amino acids placed within the pericruciate cortex, red nucleus or fastigial nucleus were processed for autoradiography. In a corresponding series of animals, electrolytic lesions were placed selectively into the above sources of reticular afferents, and terminal degeneration within the LRN was studied by light microscopy. An extensive input from the spinal cord was found to terminate predominantly on the ipsilateral side throughout the rostrocaudal extent of the LRN, except for a small ventromedial area of the rostral parvocellular division and a small rostromedial area of the magnocellular division. The cortical projection terminated diffusely within the middle one-half of the contralateral magnocellular division, while the rubral projection terminated extensively within the contralateral subtrigeminal division and the dorsolateral region of the rostral magnocellular and neighbouring parvocellular divisions. The rubral projection did not overlap the cortical projection. The fastigial nucleus projected sparsely to the contralateral LRN, mainly to the medial aspect of the rostral two-thirds of the magnocellular division, with less to the parvocellular and subtrigeminal divisions. The LRN therefore receives spinal and supraspinal projections within at least its rostral one-half, and these terminate within specific areas in a partially overlapping fashion, whereas the caudal one-half is primarily a spinal receiving region. No convergence of the rubral and sensorimotor cortical projections was evident.

Vestibular inputs to the fastigial nucleus; evidence of convergence of macular and ampullar inputs

1. Experiments have been undertaken on 11 decerebrate cats to investigate the effects of natural vestibular stimulation on the activity of cerebellar fastigial neurons. 2. From recordings in the rostral portion of the nucleus during sinusoidal lateral (roll) and horizontal (yaw) rotation, distinctive patterns of response were observed. 3. The majority of neurons sensitive to vestibular stimulation showed responses to a single modality of vestibular activation. During lateral tilt some neurones showed positional sensititivy, others gave responses related tothe velocity of movement. Other neurones responded in phase with the velocity of movement in the horizontal plane. 4. Aside from these neuronal responses, others provided indications of a convergence of inputs from different sets of vestibular receptors. In particular, several neurons showed a pattern of response that indicated tht they received inputs from otolith receptors and ampullar receptors of the vertical canal. At low velocities of movement their response was positional but with inreasing velocity the magnitude of the response increased and there was a marked phase shift of the discharge towards head velocity. 5. Neurons responding to horizontal rotation often showed positional responses during lateral tilt. There were also indications of a convergence of ampullar inputs from both vertical and horizontal canals. 6. The neural pathways mediating these resonses are discussed in consideration of previous neuroanatomical and neurophysiological data. We consider it likely that several pathways may act to evoke the patterns of response observed, and a role of the cerebellar cortex is indicated.

Fastigial modulation of brainstem behavioral mechanisms

Stimulation of the rostral, but not caudal, half of the cerebellar fastigial nucleus was found to induce coordinated eating and grooming behaviors in the cat. These behaviors persisted, unchanged, after lesions of the ascending fastigial projections in the superior cerebellar peduncle, suggesting that such behaviors are produced by direct fastigio-bulbar pathways. These results support the suggestion that there is a close anatomical association between fastigial systems mediating elicited behaviors and those mediating fastigial pressor responses. The present results are also consistent with the growing view that cerebellar systems can exert important modulatory influences on neurobehavioral mechanisms.

The organization of afferents to the cerebellar cortex in the cat: projections from the deep cerebellar nuclei

The topography of the cerebellar nucleo-cortical projection was investigated in the cat by experiments employing the horseradish peroxidase (HRP) technique or by combined HRP-autoradiographic methods. The results of the HRP studies extend previous findings showing that neurons in the deep nuclei project to the cerebellar cortex in an orderly way. Thus, it appears that the cortex of the vermis-proper receives projections from neurons located predominately in the fastigial nucleus. Intermediate and lateral zones of mid-vermal cerebellar cortex are projected on by neurons located in the interposed and dentate nuclei. Crus II receives input from neurons located predominately in the dentate nucleus, while the paramedian lobule is projected on by neurons located in a large postero-dorsal sector of the interposed nucleus and in a smaller medial strip of the dentate nucleus. Neurons in the ventral part of the dentate nucleus and the lateral part of the interposed nucleus send fibers to the paraflocculus. The nucleo-cortical pathway to the flocculus and nodulus arises largely from a population of neurons located in a ventral region stretching from the medial border of the dentate nucleus to the lateral border of the fastigial nucleus. The results of experiments using the combined HRP-autoradiographic method show that clusters of neurons in the deep cerebellar nuclei project back to the cerebellar cortical areas from which they receive input, establishing a fairly precise feedback loop between the cerebellar cortex and deep nuclei.

The location of spinal projection neurons in the cerebellar nuclei (cerebellospinal tract neurons) of the cat. A study with the horseradish peroxidase technique

The distribution of spinal projection neurons was studied in the cerebellar nuclei of the cat following injections of horseradish peroxidase (HRP) into the cervical, thoracic and lumbar cord. HRP-positive (labeled) neurons were found in the medial (fastigial) and the posterior interpositus nuclei on the side contralateral to the cervical injection, being most numberous in cases with injections between the C2 and the C3 segments. In the medial nucleus (M) labeled neurons were distributed in the central to the caudal portions, and there was a conspicuous group of labeled small neurons extending from the ventrolateral part to the intermediate zone between the M and the anterior interpositus nucleus. With an increasing number of medium-sized neurons, this neuronal group persisted caudally in a similar position, ventromedial to the posterior interpositus nucleus (IP). Labeled large neurons were seen in the medial third of the IP. In the two cases labeled neurons of medium and small sizes were equal in number, and the neurons of the IP constituted about 10% of the total number of the spinal projection neurons. The present study suggests that the neurons of the M and the IP, including those of the intermediate group located between the two, project the bulk of the crossed descending fibers as far caudally as the C2 and the C3 segments.

Brain stem afferents to the fastigial nucleus in the cat demonstrated by transport of horseradish peroxidase

Although retrograde and anterograde degeneration studies have provided important information concerning brain stem afferents to the fastigal nucleus (FN), these data may be incomplete and should be confirmed by axonal transport methods. Attempts were made to inject horseradish peroxidase (HRP) unilaterally into the FN in a series of adult cats. Animals were perfused with dextran and a fixative solution of paraformaldehyde and glutaraldehyde in 0.1 M phospate buffer. Representative sections were treated by the Graham and Karnovsky ('66) method. Selective HRP injections in one FN resulted in retrograde transport of the marker to Purkinje cells of the ipsilateral vermis and distinctive appendages of the contralateral medial accessory olivary (MAO) nucleus (nucleus beta and the dorso-medial cell column). Retrograde transport of the label was found bilaterally in cells of the medial (MVN) and inferior (IVN) vestibular nuclei, in cell group x and in the nucleus prepositus (PP). Labeled vestibular neurons, most numerous in MVN, were identified in dorsal, caudal and lateral regions, with a slight ipsilateral preponderance. Only a few neurons in caudal, dorsal and lateral regions of the IVN were labeled and none of these included cells of group f. Labeled cells in the caudal third of PP were greatest ipsilaterally. Rostral and caudal injections of FN labeled smaller numbers of cells in MVN, IVN, cell group x and PP. HRP injections of FN and portions of lobules VIII and IX resulted in bilateral retrograde labeling of larger numbers of cells in MVN, IVN and cell group x, and ipsilateral labeling of cells in group y and the interstitial nucleus of the vestibular nerve. Injections of HRP into basal folia of lobules V and VI resulted in retrograde transport of the marker to cells of the medial and dorsal accessory olivary nuclei contralaterally, and to cells of the ipsilateral accessory cuneate nucleus. Transport of label injected into portions of the pyramis was detected in parts of the contralateral MAO and bilaterally in parts of the pontine and reticulotegmental nuclei. This study suggests that the principal afferents of the fastigial nucleus arise from: (1) Purkinje cells of the ipsilateral vermis, (2) restricted portions of the contralateral MAO (nucleus beta and dorsomedial cell column), (3) portions of the MVN and IVN (bilaterally) and (4) caudal parts of the PP. Secondary vestibular inputs to the fastigial nucleus probably are relayed mainly by Purkinje cells in the cerebellar cortex.

Anatomical and physiological evidence for a cerebellar nucleo-cortical projection in the cat

Combined neuroanatomical and electrophysiological experiments were performed to test the hypothesis that axon collaterals of neurons in the cerebellar nuclei project to the cerebellar cortex in cats. The anatomical studies demonstrated that (a) following the injection of tritiated leucine into the deep cerebellar nuclei, labeled fibers could be traced into the granular layer of the cerebellar cortex, and (b) following the injection of horseradish peroxidase into the cerebellar cortex, retrogradely labeled horseradish peroxidase-positive neurons were identified in the deep nuclei. The electrophysiological experiments confirmed the anatomical findings. Neurons in the dentate and interposed nuclei, identified by their antidromic activation from the brachium conjunctivum, could also be activated antidromically from the cerebellar surface. Collision experiments demonstrated that projections from the deep cerebellar nuclei to the cerebellar cortex are in part collaterals of efferent neurons projecting through the brachium conjunctivum. Care was taken to ensure that all recordings were obtained from the region of cell somata in order to minimize the likelihood of recording from neuronal elements passing through the cerebellar nuclei. These combined neuroanatomical and electrophysiological studies provide strong evidence supporting the existence of a collateral system from cerebellar output neurons to the cerebellar cortex. The existence of this collateral system emphasizes that the cerebellar cortex and cerebellar nuclei may comprise a functional unit in which these collaterals may serve as a substrate for feedback control of the cerebellar cortex by the cerebellar output.

Influences on colonic and small intestinal motility by the cerebellar fastigial nucleus

The fastigial influence on intestinal motility was investigated in acute experiments on chloralosed cats. Motility was recorded both from the small and large intestine. Electrical stimulation of the rostral fastigial pole produced, in combination with a blood pressure rise, increased motor activity in ileum and colon while jejunum could respond with either increased on decreased motility. The intestinal responses were neither secondary to changes in intestinal blood flow, nor to baroreceptor reflexes induced by the increased blood pressure. The excitatory responses were not due to increased parasympathetic activity since sectioning of such pathways failed to abolish the responses. Instead, interruption of adrenergic sympathetic discharge, accomplished either by guanethidine or by sectioning of relevant nerves, aid eliminate the responses, indicating that the fastigial effects were mediated by suppression of prevailing adrenergic tone. Noxious stimuli to the abdomen, including laparotomy, inhibit intestinal motility by a reflex increase in adrenergic discharge. It is suggested that fastigial influence on intestinal motility is mainly due to suppression of this reflex.