Ultrastructural analysis of climbing fiber-Purkinje cell synaptogenesis in the rat cerebellum.

Previous reports have described the transient expression of the neuropeptides calcitonin gene-related peptide and neuropeptide Y in selected subsets of rat olivocerebellar compartments during embryonic and postnatal development. Using these neuropeptides as endogenous markers for olivocerebellar fibers, the aim of this electron microscopic analysis was to reveal the synaptogenetic processes occurring between climbing fibers and their target Purkinje cells, from embryonic day 19 to postnatal day 16, the period during which Purkinje cells undergo intense emission and retraction of dendrites, and climbing fibers translocate their synapses along Purkinje cell membrane surfaces. The present findings provide the first direct evidence that climbing fiber synaptogenesis starts on embryonic day 19 and that these first synapses mainly involve the Purkinje cell embryonic dendrite rather than the Purkinje cell soma. At the same age, the presence of unlabeled synapses resembling calcitonin gene-related peptide-labeled synapses in the Purkinje cell plate makes it possible to conclude that climbing fibers form a major synaptic investment on embryonic Purkinje cells, a finding that strongly supports the hypothesis of an early differentiating role of climbing fibers on cerebellar development. Furthermore, during the period of intense dendritic remodeling of Purkinje cells, 'myelin figures' were often detected in Purkinje cell dendrites suggesting that they may at least in part represent real ultrastructural markers of membrane turnover that identifies the sites where Purkinje cell dendritic rearrangement is taking place. Finally the finding that the climbing fiber terminals apposed to degenerating dendrites did not generally show signs of degeneration leads us to suggests that climbing fiber translocation from a perisomatic to a dendritic location may be driven by the Purkinje cell dendritic remodeling.

Calcitonin gene-related peptide receptor expression in the neurons and glia of developing rat cerebellum: an autoradiographic and immunohistochemical analysis.

Quantitative autoradiography (using [125I]human alpha-calcitonin gene-related peptide as a ligand) and immunofluorescence (using monoclonal antibodies directed against a purified receptor) followed by confocal analysis were applied to analyse the distribution and cellular localization of the calcitonin gene-related peptide receptor in the rat cerebellum during development. From late embryonic days to the end of the second postnatal week, during the time window of calcitonin gene-related peptide expression in climbing fibers, high levels of calcitonin gene-related peptide binding sites were found in the white matter, where immunolabeling was present in oligodendrocytes. Lower levels were found in the cerebellar cortex, where receptor immunolabeling was found in Bergmann glia in a presumptive cell surface location and, during the second postnatal week, also in the cytoplasm of Purkinje cells. From the end of the second postnatal week to adulthood, when calcitonin gene-related peptide is no longer present in climbing fibers, the number of calcitonin gene-related peptide binding sites increased in the molecular layer, where not only Bergmann glia but also Purkinje cell distal dendritic branchlets were immunolabeled in a presumptive cell surface location. Concomitantly, the number of calcitonin gene-related peptide binding sites sharply decreased in the white matter. The developmental expression of the calcitonin gene-related peptide receptor and the previously described proliferating/differentiating effects of the peptide on glial cells suggest that calcitonin gene-related peptide and its receptor may promote a coordinated development of cerebellar glial cells, an effect driven mainly by the calcitonin gene-related peptide released by climbing fibers. As a result of glia-neuron interactions, an indirect effect on the differentiation of the cerebellar neuronal circuitry is also likely to occur.

GAT-1 developmental expression in the rat cerebellar cortex: basket and pinceau formation.

In view of the key role exerted by neurotransmitter transporters in the synaptic transmission, the expression of GABA transporter GAT-1 was analysed during cerebellar development, when relevant processes of synapse maturation take place. GAT-1-immunoreactive (IR) structures started to be detected on PD 8-9, at the low molecular and Purkinje cell layer, coincident with the onset of functional inhibitory synapses on Purkinje neurons. By PD 18, GAT-1-IR structures completely ensheathed the Purkinje cell somata thus outlining the characteristic perisomatic formation, whereas GAT-1 wrapping on the axon initial segment started to be detected only at PD 15, and the mature form of the pinceau was fully developed from PD 23 on. These results, when compared with the functional maturation of the GABAergic input to Purkinje cells, indicate that GAT-1 may play a significant role in the differentiation of basket interneuron-Purkinje cell circuit.

GAT-1 developmental expression in the rat cerebellar cortex: basket and pinceau formation.

In view of the key role exerted by neurotransmitter transporters in the synaptic transmission, the expression of GABA transporter GAT-1 was analysed during cerebellar development, when relevant processes of synapse maturation take place. GAT-1-immunoreactive (IR) structures started to be detected on PD 8-9, at the low molecular and Purkinje cell layer, coincident with the onset of functional inhibitory synapses on Purkinje neurons. By PD 18, GAT-1-IR structures completely ensheathed the Purkinje cell somata thus outlining the characteristic perisomatic formation, whereas GAT-1 wrapping on the axon initial segment started to be detected only at PD 15, and the mature form of the pinceau was fully developed from PD 23 on. These results, when compared with the functional maturation of the GABAergic input to Purkinje cells, indicate that GAT-1 may play a significant role in the differentiation of basket interneuron-Purkinje cell circuit.

Somatotopic nucleocortical projections to the multiple somatosensory cerebellar maps.

The cerebellum is organized in a series of parasagittal compartments: in C1-C3 and C2 compartments Purkinje cells receive climbing fibre afferents from the rostral part of the accessory olives, and project their axon to the nucleus interpositus anterior and posterior, respectively. Within these compartments electrophysiological studies have shown that the cutaneous input carried by climbing fibre afferents is topographically organized so as to design a map of peripheral body districts. The body map is replicated over the anterior lobe-pars intermedia and the paramedian lobule, and anatomical studies have indicated that the replication is partly due to the axonal branching of olivocerebellar neurons. The aim of this study was to analyse the presence of a somatotopic organization and of a branching pattern in the nucleocortical projections, in relation to the replicated body maps within C1-C3 and C2 compartments. By using double retrograde neuronal tracing we explored, in the cat, the topographic distribution of single- and double-labelled cells in the interposed nuclear subdivisions, after tracer injections into forelimb or hindlimb regions of the anterior lobe-pars intermedia, paramedian lobule and hemisphere (medial crus II). Most of the nucleocortical neurons were found in ipsilateral nucleus interpositus posterior, with smaller numbers in the ipsilateral nucleus interpositus anterior. Nucleocortical neurons projecting to forelimb- or hindlimb-related areas are completely segregated, the forelimb neurons being located laterally and the hindlimb neurons medially in the nucleus interpositus posterior. Within their respective domains both the forelimb and hindlimb populations projecting to the anterior lobe-pars intermedia are partly segregated from those projecting to the paramedian lobule, in that the two populations are slightly shifted along the dorsoventral axis of the nucleus. Although mostly different, some of the cells are common to the two forelimb populations, since they send axonal branches to the homologous areas of the anterior lobe and paramedian lobule. Contralateral fastigial or interposed nucleocortical projections are restricted to the anterior lobe-pars intermedia, and their neurons of origin are different from those that project to the ipsilateral cerebellar cortex: i.e. they are not a bilateral, but a separate contralateral component.

Neuropeptide Y (NPY) expression is up-regulated in the rat inferior olive during development.

The aim of this study was to analyse the developmental expression of the neuropeptide Y (NPY) in the rat inferior olivary (IO) complex by immunoperoxidase and immunofluorescence techniques. The spatial distribution of NPY-immunoreactivity (IR) did not vary during development, whereas NPY-IR intensity levels varied significantly. The peak of NPY-IR expression occurred during the second postnatal week, but differed in intensity in individual IO subnuclei, the highest levels being present in the dorsal fold of the dorsal accessory olive and in the ventro-lateral outgrowth. In the adult, NPY-IR could only be rescued in colchicine pretreated animals, but its distribution overlapped the one found during development. These findings show that NPY-IR is transiently up- regulated, during development, in specific compartments of the IO complex, and that the peptide is rescued in the same specific olivocerebellar compartments in the adult. These observations are here taken to support the hypothesis that NPY may exert different trophic-differentiating and/or neuromodulatory roles during development, when its expression is transiently up-regulated, or at adult stages, when it can be rescued, according to the different biological contexts.

Neuronal and glial localization of the GABA transporter GAT-1 in the cerebellar cortex.

THE distribution and neuronal or glial localization of GAT-1, a high affinity GABA transporter, in the cerebellar cortex was analysed by means of double-label immunofluorescence experiments with GAT-1 combined with calbindin D, synaptophysin, or GFAP antibodies. In the Purkinje cell (Pc) layer, prominent synaptic GAT-1-immunoreactivity (IR) was found in the axon terminals of basket cells surrounding the Pc axon hillock. GAT-1-IR was also found in neuronal and glial processes ensheathing Pc somata and dendrites. Numerous immunoreactive fibres and puncta originating from basket and stellate cells, or from Golgi cells, were also detected in the molecular or granular layer, respectively. These observations suggest that GAT-1 is involved in the termination of the action of GABA at the inhibitory synapses of all cerebellar interneurones, primarily of basket cell terminals at the Pc axon hillock. GAT-1 in the astroglial processes presumably plays the additional role of regulating the extracellular concentrations of GABA.

Developmentally regulated expression of alpha- and beta-calcitonin gene-related peptide mRNA and calcitonin gene-related peptide immunoreactivity in the rat inferior olive.

Immunohistochemical methods have revealed the transient neonatal expression of calcitonin gene-related peptide (CGRP) in olivocerebellar compartments, and it has been hypothesized that this peptide plays a role in the development of olivocerebellar connectivity. Furthermore, the distribution of the CGRP binding sites in the cerebellar cortex also favors this hypothesis. In this study, the pattern of postnatal expression of alpha- and beta-CGRP mRNAs in the inferior olive (IO) complex was analyzed using in situ hybridization histochemistry with RNA probes complementary to specific sequences of alpha- and beta-CGRP mRNAs, and the results were compared with the pattern of CGRP immunoreactivity. High levels of alpha-CGRP mRNA expression were found in specific subnuclei of the IO complex, i.e., the medial part of the dorsal fold of the dorsal accessory olive, the beta nucleus, the dorsal cap, the caudal third of the medial accessory olive, and the rostral part of the dorso-medial cell column; in the same subnuclei beta-CGRP mRNA was detected. The olivary distribution of the two CGRP mRNA coincided with that of CGRP immunoreactivity. The expressions of alpha-CGRP mRNA and CGRP immunoreactivity were restricted to the first 2 postnatal weeks, the peak being reached at the end of the first week; beta-CGRP mRNA was transiently expressed in the same olivary compartments, but only from postnatal day 6 to 9. In general, the alpha-CGRP signal was also more intense than the beta-CGRP signal. The present findings indicate that the alpha- and beta-CGRP mRNA expression in the olivary complex is under developmental control and restricted to specific olivocerebellar compartments. The data provide a basis for the transient expression of a CGRP olivocerebellar compartment and further support the hypothesis of a role for CGRP in the complex postnatal cerebellar phenomena of connectivity reshaping and synapse stabilization.

CGRP as a marker of the climbing fibers during the development of the cerebellum in the rat.

Olivary fibers express CGRP-LI until the nest phase of their development and never in the next postnatal steps, that is, the "capuchon" stage. Together with other results presented in this volume regarding the postnatal development of CGRP binding sites in the cerebellar cortex, these immunohistologic findings suggest a role for CGRP in reshaping connectivity and in synapse stabilization of cerebellar circuitry during postnatal development.

L-homocysteic acid mediates synaptic excitation at NMDA receptors in the hippocampus.

beta-p-Chlorophenylglutamate (Chlorpheg), a specific L-homocysteate (L-HC) uptake blocker, was tested on the L-HC- and L-glutamate-induced currents and on the excitatory postsynaptic potentials (EPSPs) evoked in CA1 rat hippocampal neurons by Schaffer collaterals stimulation. In the presence of tetrodotoxin (TTX; 1 microM), Chlorpheg (0.5-2 mM) potentiated L-HC- but not L-glutamate-induced currents. In normal magnesium containing medium and at resting membrane potential, Chlorpheg (1.5-1 mM) increased the amplitude and duration of the EPSPs evoked by Schaffer collaterals stimulation. This effect was prevented by bath application of the N-methyl-D-aspartate (NMDA) receptor antagonist CPP (20 microM). Chlorpheg enhanced also the NMDA component of the EPSP, evoked in the presence of 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 10 microM), bicuculline (20 microM) and glycine (100 microM). This effect was blocked by CPP (20 microM). It is concluded that L-HC is an endogenous NMDA agonist at the Schaffer collateral-CA1 synapse.

l-Homocysteate Preferentially Activates N-methyl-D-aspartate Receptors to CA1 Rat Hippocampal Neurons.

Intracellular recordings and current and single-electrode voltage-clamp techniques were used to study the membrane responses of CA1 pyramidal neurons to bath application of l-homocysteic acid (l-HC) in the rat hippocampal slice preparation. In control artificial cerebrospinal fluid (ACSF), l-HC (25 - 250 microM) depolarized the membrane and induced a burst-like firing pattern. Both the membrane depolarization and the burst firing were blocked by the N-methyl-d-aspartic acid (NMDA) receptor antagonists d-(-)-2-amino-5-phosphonovaleric acid (AP-5, 50 microM), d-(-)-2-amino-7-phosphonoheptanoic acid (AP-7, 50 microM) and (+/-)-3-(2-carboxy-piperazin-4-yl)-propyl-1-phosphonic acid (CPP, 20 microM). In ACSF containing tetrodotoxin (1 microM), l-HC (100 - 300 microM) induced at resting membrane potential a depolarization which was associated with a small increase in input conductance. These effects were unaffected by 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 10 - 20 microM) but were fully blocked by AP-5, AP-7 (50 microM) and CPP (10 - 20 microM). In voltage-clamp experiments, l-HC induced slow inward currents which were voltage-dependent between - 70 and - 30 mV and reversed polarity near 0 mV. The l-HC-induced inward current was unaffected by CNQX (10 - 20 microM) but was strongly reduced by AP-5 or AP-7 (50 microM). The l-HC-induced inward current was temperature-dependent. Between - 60 and - 70 mV, its amplitude increased by 320% when the temperature was lowered from 33 to 22 degrees C. The l-HC-induced current was also potentiated by the specific l-HC uptake blocker beta-p-chlorophenylglutamate (Chlorpheg, 0.5 - 2 mM). These data suggest that l-HC preferentially activates NMDA receptors in CA1 hippocampal neurons.

Modulation of cerebellar CGRP binding sites induced by climbing fibre activation.

The neuropeptide CGRP is transiently expressed at neonatal stages in the rat olivocerebellar system, while high affinity binding sites for the peptide are permanently expressed in the molecular layer of the adult cerebellum. In this study, an increase in the density of cerebellar high affinity binding sites for calcitonin gene-related peptide (CGRP) is induced by harmaline, given at a dose appropriate to induce tremor. The present results, taken together with the demonstration that harmaline is a potent stimulator of the olivocerebellar system, suggest that the expression of cerebellar CGRP receptors can be modulated by the level of activity in the cerebellar afferents.

CGRP expression in the rat olivocerebellar system during postnatal development.

CGRP-like immunoreactivity was studied in the cerebellar cortex and precerebellar nuclei of neonatal rats. From postnatal day (PD) 0 to PD 13-15, CGRP immunoreactivity was transiently found in fiber-like structures around Purkinje cells, defining parasagittal bands. Following the same time course, transient labeling was also found in the inferior olive neurons which, in the adult, distribute their axons to the olivocerebellar compartments corresponding to the immunoreactive bands. It is concluded that CGRP is transiently expressed in specific climbing fiber compartments.

Spatial distribution of axon collaterals of single inferior olive neurons.

The aim of this study was to define the overall distribution pattern of the axon collaterals of single inferior olive (IO) neurons in relation to the multiple somatotopic maps defined by the climbing fiber (CF) input through the cerebellar cortex. In a previous study (Rosina and Provini: Brain Res. 289:45-63, '83), it was shown that the IO neurons supply interlobar collaterals to pairs of somatotopically related areas in the intermediate part of the anterior lobe (PIAL), in the paramedian lobule (PML), in crus II, and in the simple lobule, within strips C1 to D2. The residual branches then could either distribute within single folia or to adjacent folia within each somatotopically defined cerebellar area or both. We studied whether or not the IO axons branch over neighboring folia of the face-forelimb (FL) areas of PIAL and PML and how this interfolial branching relates to the interlobar collateralization by using the multiple fluorescent retrograde tracing technique. The main results of the study were as follows: the axons from neurons in IO subdivisions that are related to strips C1-C3 give off two interfolial branches in the FL area of PIAL and practically no interfolial collaterals are given in the FL area of PML; and the neurons that give off interfolial collaterals also give interlobar branches. From these data we have inferred the general branching pattern of the IO neurons that convey FL information to PIAL and PML. Each neuron gives off two interlobar collaterals: the branch directed to PIAL splits again into two interfolial collaterals, while each of these three collaterals should give off about three branches within each target folium to account for the ten collaterals estimated to be present in the cat. The distribution pattern of IO axon collaterals proposed here suggests that the same CF-relayed information may interact, at the Purkinje cell level, with different sets of mossy fiber inputs. The effect of this interaction would be to modulate the motor commands forwarded to specific muscle groups in relation to the different conditions under which a given movement is executed.

Pontocerebellar system linking the two hemispheres by intracerebellar branching.

Origin and percentage of bilaterally projecting pontocerebellar neurons whose axons branch within the cerebellum and link the two cerebellar hemispheres were studied in cats using double retrograde fluorescent tracing and lesion techniques. These pontocerebellar neurons, which can be viewed as a separate component of the pontocerebellar system, account for 5-10% of the pontine neurons projecting to the lateral hemisphere. The system is mainly composed of neurons the axons of which recross within the cerebellum. The pontocerebellar interhemispheric system described here is likely to be the basis for the previously described intracerebellar commissural system. The bilateral distribution of these fibers may provide an important substrate for the coordination of bilaterally performed movements.

Somatotopy of climbing fiber branching to the cerebellar cortex in cat.

The study was aimed at determining the distribution and incidence of the inferior olive axonal branching as related to the somatotopic organization of the climbing fiber system in cat. Multiple fluorescent tracing was used. In a first set of experiments, spectrally different fluorescent tracers were injected into somatotopically defined areas of the anterior lobe-pars intermedia and of the paramedian lobule. Retrogradely labeled cells were found to be localized in well-segregated face-forelimb or hindlimb olivary domains, within the various subdivisions. A large and consistent number of the olive cells were found to send axonal branches to face-forelimb or hindlimb regions of the anterior lobe-pars intermedia and of the paramedian lobule. It was also found that inferior olive branching is restricted to somatotopically corresponding areas of the two lobes. In a second set of experiments the same technique was used to check whether the inferior olive neurons which project to the forelimb areas of the anterior lobe-pars intermedia and the paramedian lobule also send collaterals to the face-forelimb areas of crus II (medial crus II). The data show that different sets of neurons projecting either to the face-forelimb areas of crus II and anterior lobe-pars intermedia or of crus II and paramedian lobule coexist within the face-forelimb related olivary domains with the set of neurons projecting to the face-forelimb areas of the anterior lobe-pars intermedia and paramedian lobule. Neurons branching to the 3 cerebellar regions were observed only occasionally. Therefore, at least for the face-forelimb-related cerebellar areas, the climbing fiber branching can be said to originate in sets of inferior olive neurons that connect different pairs of somatotopically homologous cerebellar cortical areas. The branching pattern is present in all the cortical zones studied, namely C1-C3, C2, D1 and D2, and appears to be a general feature of the olivocerebellar system.

Longitudinal and topographical organization of the olivary projection to the cat ansiform lobule.

This study on the organization of the olivary projections to the ansiform lobule in the cat was aimed at defining the longitudinal zonal pattern and the internal topography within the zones. The horseradish peroxidase method was used. Two types of injections were made: large injections covering the full extent of small groups of folia, and small injections aimed to be restricted to the single zones proposed by Voogd. It was shown that olivary projections to these parts of the cerebellum originate from the dorsal and medial accessory nuclei and from the principal nucleus, in agreement with previous studies. Moreover it was found that the ventral lamella and the dorsal lamella (and bend) of the principal nucleus give rise to two distinct, non-overlapping, cerebellar projections to D1 and D2 zones, respectively. It is thus concluded that in the cat four separate olivo-cerebellar strips, corresponding to the C2, C3, D1 and D2 zones of Voogd, are present in the ansiform lobule. In addition a transverse rostro-caudal organization was found, in which discrete regions of each olivary subdivision are connected to discrete areas of the crural cortex, within each olivo-cerebellar strip. The old idea of a point-to-point topography in the olivo-cerebellar system is consequently still valid, being compatible with the now proven longitudinal zonal pattern of the olivo-ansiform projection. The fact that the two subdivisions of the principal olive, here shown to project to separate D1 and D2 zones of the ansiform lobule, receive specific sets of afferents suggests that the two cortical zones are part of two different cerebellar operational units.