Spinocerebellar Projections Research Articles

Single cell atlas of spinal cord injury in mice reveals a pro-regenerative signature in spinocerebellar neurons.

After spinal cord injury, tissue distal to the lesion contains undamaged cells that could support or augment recovery. Targeting these cells requires a clearer understanding of their injury responses and capacity for repair. Here, we use single nucleus RNA sequencing to profile how each cell type in the lumbar spinal cord changes after a thoracic injury in mice. We present an atlas of these dynamic responses across dozens of cell types in the acute, subacute, and chronically injured spinal cord. Using this resource, we find rare spinal neurons that express a signature of regeneration in response to injury, including a major population that represent spinocerebellar projection neurons. We characterize these cells anatomically and observed axonal sparing, outgrowth, and remodeling in the spinal cord and cerebellum. Together, this work provides a key resource for studying cellular responses to injury and uncovers the spontaneous plasticity of spinocerebellar neurons, uncovering a potential candidate for targeted therapy.

Single axonal morphology reveals high heterogeneity in spinocerebellar axons originating from the lumbar spinal cord in the mouse.

Among the spinocerebellar projections vital for sensorimotor coordination of limbs and the trunk, the morphology of spinocerebellar axons originating from the lumbar cord has not been well characterized compared to those from thoracic and sacral cords. We reconstructed 26 single spinocerebellar axons labeled by biotinylated dextran injections into the gray matter of the lumbar spinal cord in mice. Axon terminals were mapped with the zebrin pattern of the cerebellar cortex. Reconstructed axons were primarily classified into ipsilaterally and contralaterally ascending axons, arising mainly from the dorsal and ventral horns, respectively. The majority of ipsilateral and contralateral axons took the dorsal-medullary and ventral-pontine pathways, respectively. The axons of both groups terminated mainly in the vermal and medial paravermal areas of lobules II-V and VIII-IXa, often bilaterally but predominantly ipsilateral to the axonal origin, with a weak preference to particular portions of zebrin stripes. The ipsilateral axons originating from the medial dorsal horn in the upper lumbar cord (n=3) had abundant (43-147) mossy fiber terminals and no medullary collaterals. The ipsilateral axons originating from the lateral dorsal horn in the lower lumbar cord (n=9) and the contralateral axons (n=14) showed remarkable morphology variations. The number of their mossy fiber terminals varied from 2 to 172. Their collaterals, observed in 17 axons out of 23, terminated mainly in the medial cerebellar nucleus, nucleus X, and lateral reticular nucleus in various degrees. The results indicated that the lumbar spinocerebellar projection contains highly heterogeneous axonal populations regarding their pathway, branching, and termination patterns.

Single axonal morphology and termination to cerebellar aldolase C stripes characterize distinct spinocerebellar projection systems originating from the thoracic spinal cord in the mouse.

The spinocerebellar projection has an essential role in sensorimotor coordination of limbs and the trunk. Multiple groups of spinocerebellar projections have been identified in retrograde labeling studies. In this study, we aimed at characterizing projection patterns of these groups using a combination of anterograde labeling of the thoracic spinal cord and aldolase C immunostaining of longitudinal stripes of the cerebellar cortex in the mouse. We reconstructed 22 single spinocerebellar axons, wholly in the cerebellum and brain stem and partly, in the spinal cord. They were classified into three groups, (a) non-crossed axons of Clarke's column neurons (NCC, 8 axons), (b) non-crossed axons of marginal Clarke's column neurons (NMCC, 7 axons), and (c) crossed axons of neurons in the medial ventral horn (CMVH, 7 axons), based on previous retrograde labeling studies. While NCC axons projected mainly to multiple bilateral stripes in vermal lobules II-IV and VIII-IX, and the ipsilateral medial cerebellar nucleus, NMCC axons projected mainly to ipsilateral stripes in paravermal lobules II-V and copula pyramidis, and the anterior interposed nucleus. CMVH axons projected bilaterally to multiple stripes in lobules II-V with a small number of terminals but had abundant collaterals in the spinal cord and medullary reticular nuclei as well as in the vestibular and cerebellar nuclei. The results indicate that, while CMVH axons overlap with propriospinal and spinoreticular projections, NCC and NMCC axons are primarily spinocerebellar axons, which seem to be involved in relatively more proximal and distal sensorimotor controls, respectively.

Spinal cord projections to the cerebellum in the mouse.

The projections from the spinal cord to the cerebellar cortex were studied using retrograde neuronal tracers. Thus far, no study has shown the detailed topographic mapping of the projections from the spinal neuron clusters to the cerebellar cortex regions for experimental animals, and there are no studies for the mouse. Tracers Fluoro-Gold and cholera toxin B were injected into circumscribed regions of the cerebellar cortex, and retrogradely labeled spinal cord neurons were mapped throughout the spinal cord. Spinal projections to the cerebellar cortex were mainly from five neuronal columns--central cervical nucleus, dorsal nucleus, lumbar and sacral precerebellar nuclei, and lumbar border precerebellar cells--and from scattered neurons located in the deep dorsal horn and laminae 6-8. The spinocerebellar projections to the cortex were mainly to the vermis. All five precerebellar cell columns projected to both anterior and posterior parts of the cerebellar cortex. Results of this study provide an amendment to the known rostral and caudal boundaries of the precerebellar cell columns in the mouse. Scattered precerebellar neurons in the most caudal deep dorsal horn and laminae 6-8 projected exclusively to the anterior part of the cerebellar cortex. In this study, no labeled spinal neurons were found to project to the lobules 6 and 7 of the cerebellar vermis, the flocculus, and the paraflocculus. Spinocerebellar neurons were located bilaterally, but the majority of the projections were contralateral for the central cervical nucleus, and ipsilateral for the remaining spinal precerebellar neuronal clusters.

Axon Sorting within the Spinal Cord Marginal Zone via Robo-Mediated Inhibition of N-Cadherin Controls Spinocerebellar Tract Formation

The axons of spinal projection neurons transmit sensory information to the brain by ascending within highly organized longitudinal tracts. However, the molecular mechanisms that control the sorting of these axons within the spinal cord and their directed growth to poorly defined targets are not understood. Here, we show that an interplay between Robo and the cell adhesion molecule, N-cadherin, sorts spinal commissural axons into appropriate longitudinal tracts within the spinal cord, and thereby facilitates their brain targeting. Specifically, we show that d1 and d2 spinal commissural axons join the lateral funiculus within the spinal cord and target the cerebellum in chick embryos, and that these axons contribute to the spinocerebellar projection in transgenic reporter mice. Disabling Robo signaling or overexpressing N-cadherin on these axons prevents the formation of the lateral funiculus and the spinocerebellar tract, and simultaneously perturbing Robo and N-cadherin function rescues both phenotypes in chick embryos. Consistent with these observations, disabling Robo function in conditional N-cadherin knock-out mice results in a wild-type-like lateral funiculus. Together, these findings suggest that spinal projection axons must be sorted into distinct longitudinal tracts within the spinal cord proper to project to their brain targets.

Fluorescence mapping of afferent topography in three dimensions

Neural circuits are organized into complex topographic maps. Although several neuroanatomical and genetic tools are available for studying circuit architecture, a limited number of methods exist for reliably revealing the global patterning of multiple topographic projections. Here we used wheat germ agglutinin (WGA) conjugated to Alexa 555 and 488 for dual color fluorescent mapping of parasagittal spinocerebellar topography in three dimensions. Using tissue section and wholemount imaging we show that WGA-Alexa tracers have three main characteristics that make them ideal tools for analyses of neural projection topography. First, the intense brightness of Alexa fluorophores allows multi-color imaging of patterned afferent projections in wholemount preparations. Second, WGA-Alexa tracers robustly label the entire trajectory of developing and adult projections. Third, long tracts such as the adult spinocerebellar tract can be traced in less than 6h. Moreover, using WGA-Alexa tracers we resolved a level of complexity in the compartmentalized topography of the spinocerebellar projection map that has never before been appreciated. In summary, we introduce versatile tracers for rapidly labeling multiple topographic projections in three dimensions and uncover wiring complexities in the spinocerebellar map.

Spinocerebellar Projections from the Cervical and Lumbosacral Enlargements in the Chicken Spinal Cord

In birds, spinocerebellar (SC) projections to the cerebellar cortex have not been understood well. We examined SC fiber terminal fields originating from the cervical and lumbosacral enlargements (CE and LSE, respectively) in the chicken. SC fiber terminals show parasagittal bands in the granular layer. Labeled terminals from the CE were distributed primarily in folia II–V and IX. Parasagittal bands of labeled terminals from the CE were not clearly separated in folia II and III but were clearly separated in folia IV and V. In folium IX, labeled terminals were diffusely distributed in all subfolia with no evidence of banding. The numbers of bands were 5 in folium II, 12 in folium III and 7 in folia IV and V at maximum. Labeled terminals from the LSE were distributed primarily in folia II–VI and IX. Labeled terminals from the LSE were arranged in 4 bands in folium II and in 8 bands in folium III at maximum. Parasagittal bands from the LSE in folia IV and V were not clearly separated. In folium VI, the numbers of parasagittal bands was 6 at maximum. In folium IX, labeled terminals were mainly found in subfolium IXc forming 6–8 parasagittal bands. There were more parasagittal bands of labeled terminals from the CE than from the LSE. The topography of SC fiber terminals from the CE was different from that of SC fiber terminals from the LSE.

Paths, elongation, and projections of ascending chick embryonic spinal commissural neurons after crossing the floor plate

The paths of embryonic chick spinal commissural neurons originating from the lumbosacral (LS) 2 spinal segment and born around Hamburger-Hamilton stage (HH) 18 were observed by labeling the axons with an in ovo electroporation method designed to limit the electroporated area to ~ one somite length. After crossing the floor plate, these axons followed two major paths, one ventral and one dorsal, and a minor path running between the major ones. These axons reached the brachial region by HH28, passed through the cervical region at HH29, and entered the medullary area by HH30. The dorsal axons entered the developing cerebellum by HH33, crossed the midline again, and spread into the rostral–ipsilateral area of the developing cerebellum so that most of them were confined to lobules II–III by HH39. A small population of ventrally running axons turned to enter the cerebellum, and the rest entered the superior medullary velum between the cerebellum and the midbrain. The LS2-originating axons that ascended ipsilaterally into the cerebellum followed a single path, and their extension was delayed compared with that of the commissural axons. Some of the ipsilateral axons innervated the cerebellum; the rest entered the superior medullary velum. These direct observations of the formation of part of the spinocerebellar projections in chick will be a useful reference for future analyses of the underlying mechanisms.

Cerebellar inputs to motor cortex.

The macaque cerebellar nuclei all project topically onto a common thalamic field that is somatotopically organized in its projection to motor cortex. The complete overlap (except at the cellular level) of dentate and interpositus (and possibly fastigius and vestibular nuclei) projection onto the somatotopic thalamic field implies a complete body representation within each cerebellar nucleus, rather than a preferential representation of trunk in fastigius, proximal limb in interpositus and digits in dentate, as is sometimes supposed. Dentate receives from association cortex and generates the earliest signals, which assist motor cortex in initiating goal-directed movements. Interpositus receives the spinocerebellar projection and provides a fast input to motor cortex from the periphery, perhaps used in transcortical 'reflex' responses and in the control of oscillation. Fastigius and vestibular nuclei provide an opportunity for labyrinthine control of motor cortex activities-even for the digits. What is unique about cerebellar input to motor cortex? Recent work has emphasized two aspects: switching of a cerebellar signal on or off through Purkinje cell inhibition, and adjusting the magnitude of the signal to optimize motor performance.

Spinocerebellar projections in the pigeon with special reference to the neck region of the body

Spinocerebellar projections in the pigeon with special reference to the neck region of the body

Spinocerebellar projections in the pigeon with special reference to the neck region of the body

Neck sensory information is important for control of head and body movements in all vertebrates. Neuroanatomic tracing methods were used to study the pathways of neck afferent systems. Both the projection of primary afferent fibers and of secondary afferent pathways to brainstem and cerebellum were investigated with the anterograde transport of dextran amines as tracers (biotinylated dextran amine and tetramethyl rhodamine dextran amine). For comparison, the projections of spinocerebellar systems of wing and leg were studied also. Complementary experiments using retrograde tracers (Fast Blue, tetramethyl rhodamine dextran amine, rhodamine isothiocyanate) injected into the cerebellum served to corroborate the results of the anterograde tracing experiments. Primary neck afferent fibers terminated in the spinal gray substance with dense terminal fields in laminae I to V of the dorsal horn and lamina IX of the ventral horn as well as in the marginal nuclei located at the lateral border of the spinal cord. In the brainstem, dense terminal fields were seen in deep layers of the medullary dorsal horn, in the external cuneate nucleus, and in group x. Secondary neck afferents arising from ventral horn cells showed a significant projection to the descending and medial vestibular nuclei and to the medial cerebellar nucleus. Terminals were found both in the anterior and the posterior cerebellum. A quantitative evaluation disclosed that most terminals of neck afferents distributed in lobules II-IV of the anterior cerebellum and lobule IX of the posterior cerebellum. With injections aimed at spinocerebellar neurons located into the cervical and lumbosacral enlargements, no projections were found in the vestibular or deep cerebellar nuclei. Projections from the cervical enlargement were concentrated in lobules III-V and those from the lumbosacral enlargement in lobules III-VI. This points to a rostrocaudal somatotopic representation of neck, wing, and leg in the anterior cerebellum. The results of the retrograde tracing experiments support such a somatotopic organization.

Distribution of the sacral neurones of origin of the ascending spinal tracts with axons passing through the lateral funiculi of the lowermost thoracic segments: an experimental HRP study in the cat.

The locations of 249 cell bodies of the ascending tract neurones in the grey matter of S1-S3 segments of the spinal cord were reconstructed by histochemical staining, after their axons (or axonal collaterals) at the level of the Thl3 segment were injected with horseradish peroxidase (HRP). In three cats in which the injections of HRP were restricted to the lateral part of the lateral funiculi (llf), about 84% of 159 retrogradely labelled cells were found on the contralateral side, while about 16% were located ipsilaterally. They were the most numerous in S2, S3 and S1 segments, respectively, and the neurones were distributed mainly in the lateral laminae I-VII, medial laminae V, VI and lamina VIII. In three other animals in which the injections of the marker were limited to the dorsal part of the lateral funiculi (dlf), 84 of the 90 ascending tract neurones were found to be distributed in the S2 and S3 segments both ipsi- (lateral laminae III-V) and contralaterally, (lateral laminae IV and V as well as the medial laminae VII and VIII) in similar numbers. The remaining six of the 90 cells with only contralateral projections at the dorsolateral funiculus at Thl3 were scattered within the S1 segment. These data are consistent with the results of studies on sacral spinocerebellar, spinothalamic and spinoreticular projections, as well as the localization of sacral spinocervical and priopriospinal neurones. They may also imply the importance of the bilateral fiber course of the neurones of origin of ascending tracts in the S2 and S3 segments within the dorsolateral funiculus.

The electrophysiological study of axonal branching of the C6-C7 neurones of the cat's spinal cord projecting to the lateral reticular nucleus and the cerebellum.

Ascending projections of the cervical spinal cord neurones to the lateral reticular nucleus (LRN) and the restiform body (RB) were electrophysiologically investigated in alpha-chloralose anaesthetized cats. The main purpose of the study was to demonstrate whether the some of C6/C7 segments neurones gave off collateral branches that reached both the LRN and RB. Antidromic action potentials were recorded from 50 neurones located in laminae IVVIII of the grey matter of C6/C7 segments following stimulation of the LRN and RB with tungsten electrodes. Analysis of the antidromic responses allowed to distinguish three populations of neurones according to their projection patterns, 41% cells ascended exclusively to either one of the two structures investigated (26% to LRN and 15% to RB), while in the remaining cases (59%) collateral projections ascended to both LRN and RB. The axonal conduction velocities were comprised in the range of 14-90 m/s and no differences were found when compared mean values for particular groups of neurones. In conclusion, the results obtained indicate that in a great number of C6/C7 neurones both spinoreticular and spinocerebellar projections are present. It is likely that afferent input from forelimbs may be directly of indirectly transmitted to the cerebellum through the same neurones.

Rostral cerebellar malformation (rcm/rcm): a murine mutant to study regionalization of the cerebellum.

A recently described recessive mouse mutant, rostral cerebellar malformation (rcm/rcm), demonstrates a swaying gait at approximately 12 days of age (Lane et al. [1992] J. Hered. 83:315-318). The mutant cerebellar (Cb) phenotype consists of cerebellar tissue that extends rostrally, beyond the usual distinct anterior cerebellar boundary, into the midbrain (Lane et al. [1992] J. Hered. 83:315-318; Ackerman et al. [1997] Nature 386:838-842). Interestingly, the cerebellar ectopia occurs in the absence of any significant alterations in the distribution of nuclear groups within the brainstem. The ectopic Cb tissue is 1) adherent to the posterior and lateral aspects of the inferior colliculus and to the lateral aspect of the rostral brainstem and 2) contains acellular regions within the inner granular layer (igl) and ectopic, calbindin-immunoreactive Purkinje cells (PCs) deep to the igl. Within the Cb proper, PC organization, as revealed by zebrin II immunoreactivity, is generally normal. In the ectopic Cb tissue PCs also exhibit a banded zebrin distribution. Analysis of the spinocerebellar projection in the mutant suggests a lobular distribution similar to that seen in the normal mouse. Within the anterior region, however, the normal parasagittal banding pattern is somewhat obscured. Spinocerebellar innervation of the ectopic Cb tissue exists, but it is almost exclusively to the region adjacent to the caudal inferior colliculus. In conjunction with the recent finding that the mutation appears to affect a UNC-5-like receptor protein for netrin-1, a molecule that may be involved in axonal guidance and cell migration (Ackerman et al. [1997] Nature 386:838-842), our results suggest that this mutant is an important model for the analysis of cerebellar development and regionalization.

Quantitative analysis of cuneocerebellar projections in rats: differential topography in the anterior and posterior lobes

The distribution of wheatgerm agglutinin-horseradish peroxidase-labelled mossy fibre terminals of internal and external cuneate projections to the cerebellar anterior and posterior lobes were quantitatively analysed in adult rats. Computer-based image analysis mapped the spatial distribution of labelled cuneocerebellar terminals in two-dimensional reconstructions of the unfolded cortex. Cuneocerebellar projections are mainly ipsilateral in their distribution. Cuneate projections to the anterior lobe vermis-medial paravermis terminate in well-circumscribed, irregularly shaped patches. These terminal patches are aligned and form a longitudinally continuous, parasagittally oriented stripe in the lateral vermis-medial paravermis of lobules I–V. These terminal patches represent the topographically organized divergent projections of different parts of the internal and external cuneate nuclei. Cuneocerebellar projections to the lateral paravermis-hemisphere, particularly in the posterior part of lobule V, are organized as a transversely oriented band of terminals. Cuneocerebellar projections to the posterior lobe terminate mainly in three transversely oriented bands of terminals located at the junction between lobules. An anterior band of terminals was located in lobule VI anteriorly and was continuous with the band of terminals located in the posterolateral part of lobule V at the junction of these two lobules. Intermediate and posterior transversely oriented bands of terminals were located at the VII–VIII and VIII–IX junctions, respectively. Cuneocerebellar projections to these three bands largely appear to represent convergent projections from different parts of the cuneate nuclei. These findings are discussed in relation to similarly analysed and previously reported findings on the organization of lower thoracic-upper lumbar spinocerebellar projections and in the context of how cuneocerebellar somatosensory input may be differentially organized and processed in disparate areas of the cerebellar cortex.

Projections from the cervical enlargement to the cerebellar nuclei in the rat, studied by anterograde axonal tracing.

Spinocerebellar projections from the cervical enlargement originate from neurons in the medial part of lamina VI and the central part of lamina VII. In the present study, the topographic projections of the cervical enlargement to the cerebellar nuclei were examined by anterograde tracing with biotinylated dextran in the rat. Following injections of the tracer into the spinal cord at levels between the C5 and T1 segments, anterogradely labeled axons and terminals were immunohistochemically demonstrated in the cerebellar nuclei. Unilateral injections revealed that projections are bilateral, but predominantly ipsilateral, to the cells of origin. Labeled axons entered the media nucleus from its rostrodorsal and rostromedial aspects. Labeled terminals were distributed to dorsal and medial parts of the middle subdivision at its rostral levels and to medial parts of the caudomedial subdivision of the medial nucleus. Most axons terminated in the middle subdivision. Single axons were seen to course rostrocaudally in the medial nucleus and give off terminal axons to both subdivisions. A few labeled terminals were seen in the dorsolateral protuberance of the medial nucleus, and in the anterior interpositus and the posterior interpositus nuclei. No labeled terminals were seen in the lateral cerebellar nucleus. The present study demonstrates that spinocerebellar neurons in laminae VI and VII of the cervical enlargement project to dorsomedial areas of the medial nucleus at rostral levels, bilaterally but predominantly ipsilaterally. It is suggested that these areas specifically receive cutaneous and muscular input related to the forelimb movement.

Partial ablation of the neonatal external granular layer disrupts mossy fiber topography in the adult rat cerebellum.

The spinocerebellar projection in the rat is compartmentalized in an array of parasagittal bands of mossy fiber terminals. These bands align reproducibly with bands of Purkinje cells that differentially express zebrin II. To investigate whether this alignment is obligatory, Purkinje cell and mossy fiber compartmentation has been compared in the rat cerebellum where the cytoarchitecture was contorted by neonatal administration of methylazoxymethanol. Methylazoxymethanol ablates many granule cell precursors, leaving a much reduced external granular layer, and adult rats that received a single methylazoxymethanol injection at birth showed varying degrees of abnormal cerebellar foliation. Zebrin II immunocytochemistry nevertheless revealed no fundamental abnormality in the Purkinje cell compartments. However, despite the normal Purkinje cell compartmentation being retained, the spinocerebellar mossy fiber-Purkinje cell topography is disrupted by methylazoxymethanol treatment. The normal precise array of parasagittal mossy fiber terminal fields becomes blurred across the lobule, and the normal clear banding is difficult to follow. These data suggest that, despite the early topography being dependent on the Purkinje cells, the granule cell-mossy fiber interactions also regulate the topography of the spinocerebellar projection.

Partial ablation of the neonatal external granular layer disrupts mossy fiber topography in the adult rat cerebellum

The spinocerebellar projection in the rat is compartmentalized in an array of parasagittal bands of mossy fiber terminals. These bands align reproducibly with bands of Purkinje cells that differentially express zebrin II. To investigate whether this alignment is obligatory, Purkinje cell and mossy fiber compartmentation has been compared in the rat cerebellum where the cytoarchitecture was contorted by neonatal administration of methylazoxymethanol. Methylazoxymethanol ablates many granule cell precursors, leaving a much reduced external granular layer, and adult rats that received a single methylazoxymethanol injection at birth showed varying degrees of abnormal cerebellar foliation. Zebrin II immunocytochemistry nevertheless revealed no fundamental abnormality in the Purkinje cell compartments. However, despite the normal Purkinje cell compartmentation being retained, the spinocerebellar mossy fiber-Purkinje cell topography is disrupted by methylazoxymethanol treatment. The normal precise array of parasagittal mossy fiber terminal fields becomes blurred across the lobule, and the normal clear banding is difficult to follow. These data suggest that, despite the early topography being dependent on the Purkinje cells, the granule cell-mossy fiber interactions also regulate the topography of the spinocerebellar projection. © 1996 Wiley-Liss, Inc.

Development of the spinocerebellar projection in the prenatal mouse.

An abundance of information is available concerning the spinocerebellar projection in adult mammals. However, only a few studies have attempted a developmental analysis of this important projection system in early postnatal and/or prenatal animals. The present study provides an analysis of the development of the projection from the spinal cord to the cerebellum in fetal mice using anterograde tracing techniques in an in vitro preparation. After applications of biocytin to the caudal cervical spinal cord, anterogradely labelled fibers were present in the brainstem of embryonic day 12 (E12/13) mice, however, there was no indication of label in the cerebellum. At E13/14, labelled fibers were evident in the rostrolateral portions of the cerebellum/isthmus region. By E15/16, labelled spinocerebellar fibers had progressed farther into the cerebellum and were seen crossing the midline in a very superficial position. At older ages, the number of crossing fibers increased, and they became more ventrally positioned within the cerebellum. At E17/18 and E18/19, labelled spinocerebellar fibers were observed to branch and invade deeper portions of the cerebellum including the cerebellar nuclei. However, at E18/19, there was no indication of the parasagittal organization characteristic of this projection in the adult animal. The results of this study indicate that spinocerebellar fibers are present within the cerebellum significantly earlier than the development and differentiation of their primary targets, the granule cells. Furthermore, these data suggest that spinocerebellar fibers may form associations with cerebellar nuclear cells during fetal development.

Developing mossy fiber terminal fields in the rat cerebellar cortex may segregate because of Purkinje cell compartmentation and not competition.

Many mossy fiber afferent projections to the rat cerebellar cortex terminate in parasagittal bands. In particular, the anterior lobe vermis of the cerebellum contains alternating bands of mossy fibers from the spinal cord and external cuneate nuclei. The cerebellar cortical efferents, the Purkinje cells, are also organized in parasagittal bands. These can be revealed by immunochemical staining for the antigen zebrin II, which is selectively expressed by bands of Purkinje cells. In some cases, the boundaries between mossy fiber terminal fields align with identified transitions between zebrin+/- sets of Purkinje cells, whereas others are located within apparently homogeneous Purkinje cell compartments. Two theories can explain the terminal-field topography: In one view, mossy fiber terminals segregate during development, because growth cones from different sources compete for common territory. Alternatively, mossy fiber growth cones directly recognize chemically distinct target territories, and activity-dependent mechanisms play only minor roles. To explore these issues, two sets of experiments were performed. First, the terminal-field map of the neonatal spinocerebellar projection was compared to the Purkinje cell compartmentation as revealed by anticalbindin immunocytochemistry. Second, subsets of spinocerebellar mossy fiber afferents were ablated early in postnatal development, and the consequences for the neighboring cuneocerebellar terminal fields were mapped in the adult with reference to the zebrin II+/- compartments. These experiments revealed no evidence that competitive interactions constrain the mossy fiber terminal-field distribution but, rather, suggest that the organization of the mossy fiber projections follows the compartmentation of the Purkinje cells.