Consensus Paper. Cerebellar Reserve: From Cerebellar Physiology to Cerebellar Disorders.

Cerebellar reserve refers to the capacity of the cerebellum to compensate for tissue damage or loss of function resulting from many different etiologies. When the inciting event produces acute focal damage (e.g., stroke, trauma), impaired cerebellar function may be compensated for by other cerebellar areas or by extracerebellar structures (i.e., structural cerebellar reserve). In contrast, when pathological changes compromise cerebellar neuronal integrity gradually leading to cell death (e.g., metabolic and immune-mediated cerebellar ataxias, neurodegenerative ataxias), it is possible that the affected area itself can compensate for the slowly evolving cerebellar lesion (i.e., functional cerebellar reserve). Here, we examine cerebellar reserve from the perspective of the three cornerstones of clinical ataxiology: control of ocular movements, coordination of voluntary axial and appendicular movements, and cognitive functions. Current evidence indicates that cerebellar reserve is potentiated by environmental enrichment through the mechanisms of autophagy and synaptogenesis, suggesting that cerebellar reserve is not rigid or fixed, but exhibits plasticity potentiated by experience. These conclusions have therapeutic implications. During the period when cerebellar reserve is preserved, treatments should be directed at stopping disease progression and/or limiting the pathological process. Simultaneously, cerebellar reserve may be potentiated using multiple approaches. Potentiation of cerebellar reserve may lead to compensation and restoration of function in the setting of cerebellar diseases, and also in disorders primarily of the cerebral hemispheres by enhancing cerebellar mechanisms of action. It therefore appears that cerebellar reserve, and the underlying plasticity of cerebellar microcircuitry that enables it, may be of critical neurobiological importance to a wide range of neurological/neuropsychiatric conditions.

Quantitative evaluation of cerebellar ataxia based on pathological patterns of the muscle activities.

Quantitative evaluation of cerebellar ataxia is crucial for precise evaluation of cerebellar diseases. In particular, it is essential to capture anomaly of the causal motor commands as well as the resultant movement for the ataxia. In this paper, we propose a new method to make a quantitative evaluation of the cerebellar ataxia based on EMG signals. As an experimental task, we asked subjects to perform step-tracking wrist movements with a manipulandum, and recorded wrist joint movements and muscle activities of four wrist prime movers with surface electrodes. The subjects included fourteen patients with cerebellar diseases and thirteen normal controls. We succeeded to extract two parameters from the EMG signals of the four wrist prime movers, which characterize the pathological patterns of muscle activities for the cerebellar ataxia, Total Co-contraction Level (TCL) and Directionality of Muscle Activity (DMA). We found that the two parameters were useful to characterize pathological patterns of muscle activities in cerebellar ataxia. Consequently, it is expected that our proposed method is useful not only in tracking condition of cerebellar patients but also in evaluating the effects of a treatment or neuro-rehabilitation aiming at the normalization of motor commands.

Direction of action is represented in the ventral premotor cortex.

The ventral premotor area (PMv) is a major source of input to the primary motor cortex (M1). To examine the potential hierarchical processing between these motor areas, we recorded the activity of PMv neurons in a monkey trained to perform wrist movements in different directions with the wrist in three different postures. The task dissociated three major variables of wrist movement: muscle activity, direction of joint movement and direction of movement in space. Many PMv neurons were directionally tuned. Nearly all of these neurons (61/65, 94%) were 'extrinsic-like'; they seemed to encode the direction of movement in space independent of forearm posture. These results are strikingly different from results from M1 of the same animal, and suggest that intracortical processing between PMv and M1 may contribute to a sensorimotor transformation between extrinsic and intrinsic coordinate frames.

Thalamic terminal morphology and distribution of single corticothalamic axons originating from layers 5 and 6 of the cat motor cortex.

We investigated the axonal morphology of single corticothalamic (CT) neurons of the motor cortex (Mx) in the cat thalamus, using a neuronal tracer, biotinylated dextran amine (BDA). After localized injection of BDA into the Mx, labeled CT axons were found ipsilaterally in the thalamic reticular nucleus (TRN), the ventroanterior-ventrolateral complex (VA-VL), the central lateral nucleus (CL), the central medial nucleus, and the centromedian nucleus, but with the primary focus in the VA-VL. The terminals in the VA-VL formed a large laminar cluster, which extended approximately in parallel with the internal medullary lamina. The laminar organization mirrored morphologic features of single CT axons. We reconstructed the trajectories of 25 single CT axons that arose from layer V (16 axons) or layer VI (9 axons) and terminated in the VA-VL. Terminals of single CT axons that originated from both layer V and layer VI were confined within a laminar structure about 700 microm thick, suggesting the existence of laminar input organization in the VA-VL. Otherwise, the two groups of the CT axons showed contrasting features. All of the CT axons derived from layer VI gave rise to a few short collaterals to the TRN and then formed extensive arborization with numerous small, drumstick-like terminals in the VA-VL. On the other hand, the CT axons arising from layer V gave rise to collaterals whose main axons descended into the cerebral peduncle. Each collateral projected to the VA-VL or CL without projection to the TRN and formed a few small clusters of giant terminals. The two groups of CT neurons in the same cortical column had convergent rather than segregated termination in the VA-VL. However, the terminals of layer VI CT neurons were distributed diffusely and widely in the VA-VL, whereas the terminals of layer V CT neurons were much more focused and surrounded by the terminals of the former group. These contrasting features of the two types of CT projections appear to represent their different functional roles in the generation of motor commands and control of movements in the Mx.

Muscle and movement representations in the primary motor cortex.

What aspects of movement are represented in the primary motor cortex (M1): relatively low-level parameters like muscle force, or more abstract parameters like handpath? To examine this issue, the activity of neurons in M1 was recorded in a monkey trained to perform a task that dissociates three major variables of wrist movement: muscle activity, direction of movement at the wrist joint, and direction of movement in space. A substantial group of neurons in M1 (28 out of 88) displayed changes in activity that were muscle-like. Unexpectedly, an even larger group of neurons in M1 (44 out of 88) displayed changes in activity that were related to the direction of wrist movement in space independent of the pattern of muscle activity that generated the movement. Thus, both "muscles" and "movements" appear to be strongly represented in M1.

Cerebellar input to corticothalamic neurons in layers V and VI in the motor cortex.

To investigate whether corticothalamic (CT) neurons in the motor cortex (Mx) receive cerebellar input via the ventroanterior-ventrolateral nucleus of the thalamus (VA-VL), we recorded intracellular potentials from neurons in the Mx of anesthetized cats and examined effects of stimulation of the VA-VL and the brachium conjunctivum on them. After this electrophysiological identification, horseradish peroxide (HRP) was injected iontophoretically into the recorded neurons for morphological analysis. We identified 34 neurons as CT neurons by their antidromic response to stimulation of the VA-VL, of which 13 were layer VI CT neurons and 21 were layer V CT neurons. A majority of the CT neurons of both layers VI and V received monosynaptic excitatory postsynaptic potentials (EPSPs) from the VA-VL and di- or polysynaptic EPSPs from the cerebellum. The laminar distribution and morphological characteristics of single CT neurons receiving cerebellar input were analyzed on 19 HRP-labeled CT neurons. Eight layer V and six layer VI CT neurons were reconstructed from serial sections. All the reconstructed layer VI CT neurons were modified pyramidal neurons whose apical dendrites ended in layer III or V, and all the stained layer V CT neurons were typical pyramidal neurons, although the laminar and tangential distribution of recurrent collaterals of these neurons varied from neuron to neuron.

Projection pattern of single corticocortical fibers from the parietal cortex to the motor cortex.

Arborization of single corticocortical (CC) axons projecting from the parietal cortex to the motor cortex (Mx) was analysed using an intraaxonal staining technique in the cat. Stem axons arising from cell bodies in area 5 ramified repeatedly into numerous terminal branches in the Mx, forming 2-6 patches (0.2-0.8 mm in diameter) separated by a terminal-free gap. Axon terminals were distributed mainly in layers II and III and sparsely in layers V, VI and I. This feature is quite similar to that of thalamocortical axons and other corticocortical fibres. Thus the patchy organization may be a basic input structure for afferents of the Mx and play a role in generation of adequate motor output patterns in the Mx.

Morphology of single axons of tectospinal neurons in the upper cervical spinal cord.

Morphology of single axons of tectospinal (TS) neurons was investigated by intraaxonal injection of horseradish peroxidase (HRP) at the upper cervical spinal cord of the cat. TS axons were electrophysiologically identified by their direct responses to stimulation of the contralateral superior colliculus (SC). None of these axons responded to thoracic stimulation at Th2. Three-dimensional reconstructions of the axonal trajectories were made from 20 well-stained TS axons at C1-C3. Cell bodies of these axons were located in the intermediate or deep layers of the caudal two-thirds of the SC. Usually, TS axons had multiple axon collaterals, and up to seven collaterals were given off per stem axon [2.7 +/- 1.6 (mean +/- S.D.); n = 20]. Collaterals had simple structures and ramified a few times mainly in the transverse plane. The number of terminals for each collateral was small. These collaterals terminated in the lateral parts of laminae V-IX, mainly in laminae VI, VII, and VIII. There were usually gaps free from terminal arborizations between adjacent collaterals, because the rostrocaudal spread of each collateral (mean = 700 microns) was narrower than the intercollateral interval (mean = 2,500 microns). Seven of the 19 TS axons had terminals in the lateral parts of laminae V-VIII, with little projection to lamina IX, and the other 12 axons had terminals in lamina IX besides the projection to the lateral parts of laminae V-VIII. Axon terminals in lamina IX did not appear to make contacts with the somata or proximal dendrites of retrogradely labeled motoneurons, but contacts were found with the somata of counterstained interneurons in the lateral parts of laminae V-VIII. Three spinal interneurons (two in lamina VIII and one in lamina V at C1) that received monosynaptic excitation from the SC were stained, and their axonal trajectories were reconstructed. They had multiple axon collaterals at C1-C2 and mainly projected to laminae VIII and IX, with smaller projections to lamina VII. Many axon terminals of the interneurons were found in multiple neck motor nuclei, where some of them made contacts with retrogradely labeled motoneurons. The present finding provides evidence that the direct TS projection to the spinal cord may influence activities of multiple neck muscles, mainly via spinal interneurons, and may play an important role in control of head movement in parallel with the tectoreticulospinal system.

Four convergent patterns of input from the six semicircular canals to motoneurons of different neck muscles in the upper cervical cord.

This study was performed to investigate the pattern of input and the pathways from the six semicircular canals to motoneurons of various neck muscles in anesthetized cats. Intracellular postsynaptic potentials from neck motoneurons were recorded in response to electrical stimulation of the six ampullary nerves. The results showed that motoneurons of a particular neck muscle have a homogeneous convergent pattern of input from the six semicircular canals; there are four patterns of input from the six semicircular canals to motoneurons of various neck muscles; and the trisynaptic connection between the semicircular canal nerves and neck motoneurons was identified in addition to the disynaptic connection.

Morphology of single axons of tectospinal and reticulospinal neurons in the upper cervical spinal cord.

Single axons of tectospinal (TS) and reticulospinal (RS) neurons were stained with intraaxonal injection of HRP after electrophysiological identification, and their axonal trajectory was reconstructed at C1-C3 of the cat. TS neurons were located in the intermediate or deep layers of the caudal two-thirds of the superior colliculus (SC) and had multiple axon collaterals (up to seven collaterals) per stem axon). Collaterals had a simple structure, ramified several times mainly in the transverse plane, and terminated in the lateral parts of laminae V-VIII. More than half also had terminals in lamina IX. Terminals of TS neurons did not appear to make contacts with either the somas or proximal dendrites of retrogradely-labeled motoneurons in lamina IX, but clear contacts were found on counterstained interneurons in the lateral part of laminae V-VIII. Here, we examined three stained spinal interneurons receiving monosynaptic excitation from the SC. These interneurons had multiple axon collaterals mainly in laminae VII-IX, and made extensive contacts with retrogradely-labeled motoneurons of multiple neck muscles. Stem axons of single RS neurons receiving input from the contralateral SC ran in the ventromedial funiculus and gave off multiple axon collaterals to laminae VII-IX over at least several cervical segments. Their terminal boutons appeared to make contact with both the somas and proximal dendrites of retrogradely-labeled neck motoneurons. Single RS neurons made contacts with motoneurons of different neck muscles. These results provide evidence for functional synergies at the level of single RS neurons and spinal interneurons for neck movements. The present finding indicates that the direct TS projection to the spinal cord may influence the activity of multiple neck muscles mainly via spinal interneurons, and plays an important role in control of head movement in parallel with the tecto-reticulospinal system.

Cerebellar and cerebral inputs to corticocortical and corticofugal neurons in areas 5 and 7 in the cat.

1. In the parietal cortex (Px, areas 5 and 7), the organization and characteristics of cerebellar and cerebral inputs and their effects on efferent neurons were investigated with the use of intracellular and extracellular recording techniques in the anesthetized cat. 2. Evoked field potential analysis revealed that two regions of the Px, the caudal bank of the ansate sulcus (Ans. S.) and the crown of the suprasylvian gyrus (Ssyl. G.), received converging input from the dentate and the interpositus nucleus. The cerebellar input to the caudal bank of the Ans. S. was relayed via the ventrolateral region of the ventroanterior-ventrolateral (VA-VL) complex of the thalamus, whereas the cerebellar input to the crown of the Ssyl. G. was relayed via the dorsomedial region of the VA-VL complex. 3. A total of 176 neurons was recorded intracellularly in the Px to examine inputs from the cerebellum. Of these, 72 neurons were corticocortical neurons projecting to the motor cortex (Mx), and 48 were corticofugal neurons to the pontine nucleus (PN). Intracellular staining with horseradish peroxidase revealed that the former corticocortical neurons were layer III pyramidal neurons and the latter corticofugal neurons were layer V pyramidal neurons. 4. Stimulation of the brachium conjunctivum (BC) produced di- or polysynaptic excitatory postsynaptic potentials (EPSPs) in corticocortical neurons projecting to the Mx and corticofugal neurons to the pontine nucleus in the Px. The characteristics of BC-evoked EPSPs were different between the bank of the Ans. S. and the crown of the Ssyl. G. In the bank of the Ans. S., the slope of the rising phase of the BC-evoked EPSPs was steeper, and their minimum latency was shorter by 0.8 ms than those in the crown of the Ssyl. G. These differences may reflect differences in the terminal distribution and conduction velocity of the thalamocortical fibers relaying cerebellar input to these two parietal areas. 5. Stimulation of the Mx produced mono- or disynaptic EPSPs in both corticocortical neurons projecting to the Mx and corticofugal neurons projecting to the pontine nucleus in the Px. For each neuron, effective sites for inducing EPSPs were distributed very widely and sometimes covered both areas 4 and 6. Extensive corticocortical projection from the Mx to the Px was confirmed by injection of an anterograde tracer into the Mx. 6. These data indicate that neurons in the Px receive inputs from both the cerebellum and the Mx and send outputs to the Mx and the cerebellum.(ABSTRACT TRUNCATED AT 400 WORDS)

Innervation of multiple neck motor nuclei by single reticulospinal tract axons receiving tectal input in the upper cervical spinal cord.

Axons of reticulospinal neurons (RSNs) activated monosynaptically by stimulation of the contralateral superior colliculus (SC) were stained with intraaxonal injection of horseradish peroxidase in the cat upper cervical spinal cord. Stem axons of single RSNs gave rise to multiple axon collaterals to laminae IX, VIII and VII over a few cervical segments. Single RSNs made contacts with retrogradely labeled neck motoneurons of different neck muscles. Therefore, RSNs were regarded as mediating output of the SC to functionally different groups of neck muscles simultaneously. The result gave evidence of neural implementation of a functional synergy for a neck movement at a single neuron level.

Thalamocortical organization in the cerebello-thalamo-cortical system.

In the cat, the cerebellum projects via the ventroanterior-ventrolateral (VA-VL) complex of the thalamus to the motor and premotor cortices and also to the parietal association cortex. Cerebellar inputs to each of these regions have been characterized electrophysiologically by depth profiles of cortical potentials following stimulation of the brachium conjunctivum and of the VA-VL complex, and morphologically by the laminar distribution of thalamocortical (TC) terminations, in aggregate and at the single-axon level. One population of TC neurons terminated mainly in layer I and was associated with late surface negative potentials. A second population, with terminations in layers III and IV, was associated with early deep negative potentials. Terminations in layer III of the motor cortex formed multiple patches about 1-1.5 mm wide (mediolateral), which aligned to form 2-5 mm stripes extended rostrocaudally. This pattern correlates with the configuration of individual TC axons, which have two to six separate terminal patches distributed over 6 mm (rostrocaudal). The wide morphological divergence of single TC axons in the cortex may imply functionally multiple innervation of different efferent columns. Alternately, along with other inputs, it may permit a highly dynamic output selection from multiple representations, for example, of a variety of muscle groups in different combinations.

Input-output organization of the ventrolateral nucleus of the thalamus.

Input-output organization of the ventrolateral nucleus (VL) of the thalamus was analyzed electrophysiologically and morphologically at the single cell level. Virtually all pyramidal tract neurons in the motor cortex and area 6 received convergent inputs from the dentate (DN) and the interpositus (IN) nuclei and about 60% of thalamocortical (TC) neurons received convergent inputs from both the DN and the IN. Anterograde labelling following focal injection of Phaseolus vulgaris leukoagglutinin and intracellular staining of TC axons showed that the terminals in layer III tended to aggregate into patches about 1-1.5 mm wide in a frontal plane, which were arranged in longitudinal strips about 2-5 mm long in a rostrocaudal direction.

Spinal commissural neurons mediating vestibular input to neck motoneurons in the cat upper cervical spinal cord.

Spinal commissural neurons (CNs) activated di- or trisynaptically by stimulation of ipsilateral vestibular afferents were stained with intraaxonal injection of horseradish peroxidase in the cat upper cervical spinal cord. Stem axons of CNs in lamina VIII or VII, after crossing the midline, had ascending and/or descending main branches that gave off multiple axon collaterals to laminae IX and VIII over a few cervical segments. Terminal boutons appeared to make contact with proximal dendrites and somata of retrogradely-labelled neck motoneurons. Therefore, these CNs were regarded as mediating vestibular afferent input to contralateral neck motoneurons trisynaptically at the shortest.

Two modes of cerebellar input to the parietal cortex in the cat.

The characteristics of cerebellar input to the parietal cortex through the ventroanterior-ventrolateral (VA-VL) complex of the thalamus were investigated in the adult cat by using combined electrophysiological and anatomical methods. Two distinct parietal regions were activated by stimulation of the cerebellar nuclei (CN). In the first region located in the depth of the bank of the ansate sulcus, stimulation of the CN induced early surface positive-deep negative potentials and late surface negative-deep positive potentials. In this cortical area, potentials of similar shape and time course were evoked at a shorter latency by stimulation of the ventrolateral part of the VA-VL complex where large negative field potentials were evoked by stimulation of the CN. After injection of the anterograde tracer Phaseolus vulgaris leucoagglutinin (PHA-L) in this part of the VA-VL complex, axon terminals of thalamocortical (TC) fibers were found in layers I, III and IV in the depth of the bank of the ansate sulcus and layers I and III in the motor cortex. In the second region located in the suprasylvian gyrus, late surface negative-deep positive potentials were evoked by stimulation of the CN and similar potentials were evoked at a shorter latency from the dorsomedial part of the VA-VL complex where large cerebellar-evoked potentials could be recorded. PHA-L injection in this thalamic region stained TC fibers and their terminals in layer I of the suprasylvian gyrus, and in layers I and III of the motor cortex. The laminar distribution of TC axon terminals in two different regions of the parietal cortex could account for the depth profiles of the cerebellar- and the thalamic-evoked potentials in each region. These results show that cerebellar information is conveyed to two separate areas in the parietal cortex by two different TC pathways.

Parietal projection of thalamocortical fibers from the ventroanterior-ventrolateral complex of the cat thalamus.

Anterograde labelling following focal injections of Phaseolus vulgaris leucoagglutinin was used to identify the parietal distribution of thalamocortical (TC) fibers from the ventroanterior-ventrolateral (VA-VL) complex of the cat thalamus. In injections in the ventrolateral or the caudal part of the VA-VL complex, labelled TC fibers were distributed in layers I, III and IV of the parietal areas 5a and/or 5b, whereas in injections located more rostrally or dorsomedially, labelled TC fibers were almost confined to layer I.

Distribution of terminals of thalamocortical fibers originating from the ventrolateral nucleus of the cat thalamus.

Anterograde labelling following focal injections of Phaseolus vulgaris leucoagglutinin was used to identify the threedimensional cortical distribution of thalamocortical (TC) fibers from the ventrolateral nucleus of the thalamus of the cat. The labelled TC fibers were distributed usually in layers I and III of the motor cortex and the terminals in layer III tended to aggregate into patches about 1-1.5 mm wide in a mediolateral direction. These patches were arranged in longitudinal strips about 2-5 mm long in a rostrocaudal direction and were separated by gaps of terminal free area.