Topographic Histochemstry of the Cerebellum

Optogenetics in the cerebellum: Purkinje cell-specific approaches for understanding local cerebellar functions

The cerebellum consists of the cerebellar cortex and the cerebellar nuclei. Although the basic neuronal circuitry of the cerebellar cortex is uniform everywhere, anatomical data demonstrate that the input and output relationships of the cortex are spatially segregated between different cortical areas, which suggests that there are functional distinctions between these different areas. Perturbation of cerebellar cortical functions in a spatially restricted fashion is thus essential for investigating the distinctions among different cortical areas. In the cerebellar cortex, Purkinje cells are the sole output neurons that send information to downstream cerebellar and vestibular nuclei. Therefore, selective manipulation of Purkinje cell activities, without disturbing other neuronal types and passing fibers within the cortex, is a direct approach to spatially restrict the effects of perturbations. Although this type of approach has for many years been technically difficult, recent advances in optogenetics now enable selective activation or inhibition of Purkinje cell activities, with high temporal resolution. Here we discuss the effectiveness of using Purkinje cell-specific optogenetic approaches to elucidate the functions of local cerebellar cortex regions. We also discuss what improvements to current methods are necessary for future investigations of cerebellar functions to provide further advances.

EphA4 is not required for Purkinje cell compartmentation

The Purkinje cells of both the adult and the developing cerebellar cortex are organized into parasagittal stripes or 'segments' expressing a variety of biochemical markers. We show that in the developing mouse cerebellar cortex, members of the Eph receptor gene family are expressed in mediolaterally alternating Purkinje cell segments. Since members of the Eph receptors family have been shown to play a role in hindbrain segmentation and boundary formation (Philos. Trans. R. Soc. Lond. B: Biol. Sci. 355 (2000) 993), we analyzed the effect of a null mutation of the EphA4 gene on Purkinje cell compartmentation. Using well characterized markers of Purkinje cell compartmentation in both the developing and the adult cerebellum, we observed no significant alteration in the banding pattern of these markers between the EphA4 knockout mice and their wild type controls. The ribboned pattern of migrating granule cells in the developing cerebellum also appears unaltered. The expression of other members of this gene family, including ephrin-B2, EphA2, and ephrin-A1, in a compartmentalized pattern within the Purkinje cell layer suggests a possible redundancy and/or a compensation of EphA4 function in the segmental patterning of cerebellar Purkinje cells.

Granule cells migrate within raphes in the developing cerebellum: an evolutionarily conserved morphogenic event

The early phase of granule cell migration in the developing chick cerebellum occurs within ribbons of cells moving through parasagittally arrayed gaps between Purkinje cell clusters. These parasagittal arrays of migrating granule cells, termed "granule cell raphes," also have been reported in rabbit and cat, but recent publications variously report that granule cell raphes are absent or present in rodents. By using Nissl counterstaining and Pax6 immunohistochemistry, we confirm that granule cells do migrate in raphes in the developing mouse cerebellum, and also in the primate cerebellum during a period of development that coincides with Purkinje cell compartmentation. In mouse and primate cerebellum, as in chick cerebellum, granule cell migratory streams occur at the borders of Purkinje cell clusters. GFAP immunostaining of Bergmann glial fibers shows no parasagittally localized pattern of distribution, indicating that the formation of granule cell ribbons is not prepatterned by heterogeneous distribution of radial glia. The conservation of the ribboned pattern of granule cell migration from bird to primate and the timing of this event suggest a possible role for granule cell raphes in parasagittal compartmentation of Purkinje cells. A potential mechanism for such an interaction is discussed.

Phenotype changes of inferior olive neurons following collateral reinnervation

Inferior olive neurons are able to enlarge or retract their axonic terminal fields in response to changes in the extension of their target domain. Following Purkinje cell loss, the retraction of target-deprived climbing fibres is accompanied by a size reduction in the inferior olive neuron cell bodies. Here, we asked whether perikaryal modifications also occur when inferior olivary neurons enlarge their terminal fields to innervate supernumerary targets. To achieve this aim, we carried out a morphometric analysis on the somatic compartment of inferior olive neurons in two experimental conditions known to induce an expansion of their terminal field, i.e. a subtotal 3-acetylpyridine inferior olive lesion in the adult and a unilateral transection of the inferior cerebellar peduncle in newborn rats. In both experimental conditions, the inferior olive neurons that survived the lesion showed a remarkable increase in cell body and nuclear size, although the latter change was less pronounced in the 3-acetylpyridine-treated animals. These results show that both developing and mature inferior olive neurons are capable of adjusting their perikaryal phenotype to match the modifications of their target size.

The process of decompaction and myelin removal after C7 ventral root avulsion in the cat

An electron microscopical study was carried out on the ventral horn in order to investigate the microvascular changes after C7 ventral root avulsion in cats. Endothelial cells: At 2 days after avulsion the endothelial cells contained vacuoles filled with fibrous-like substances. After 14 days the endothelial phagosomes also contained myelin sheath-like and "soap-bubble" structures. Tight junctions between the endothelial cells remained present without exception. From 14 up to 90 days, intraluminal debris was observed. Edema and glia: From 2 up to 30 days after avulsion perivascular edema was noted around blood vessels and polymorphonuclear granulocytes were found mainly in the peri-endothelial space. Eight days after avulsion, the number of astrocytic processes around the blood vessels and the phagocytic activity of perivascular cells increased. Myelin sheath-like structures were encountered in phagosomes of the pericytes. After 14 days the distribution of astrocytic processes around the blood vessels had stabilized and remained so until day 90 after avulsion. In the same period the phagocytic activity decreased, and the myelin sheath-like material in the perivascular cell phagosomes gradually disappeared. The amount of microglial cells around the blood vessels showed an increase after 30 days survival and then stabilized. These results indicate transport of debris from the neuropil across the endothelial cells into the blood vessel lumen after ventral root avulsion.

Innervation of "painful" lumbar discs

STUDY DESIGN

The authors investigated the innervation of discographically confirmed degenerated and "painful" human intervertebral discs.

OBJECTIVE

To determine the type and distribution patterns of nerve fibers present in degenerated human intervertebral discs.

SUMMARY OF BACKGROUND DATA

The innervation of intervertebral discs has previously been extensively described in fetal and adult animals as well as humans. However, little is yet known about the innervation of severely degenerated human lumbar discs. The question may be posed whether a disc that has been removed for low back pain possesses an increased innervation compared with normal discs.

METHODS

The presence of nerve fibers was investigated using acetylcholinesterase enzyme histochemistry, as well as neurofilament and substance P immunocytochemistry. From 10 degenerated and 2 control discs, the anterior segments were excised and their nerve distribution studied by examining sequential sections.

RESULTS

In all specimens, nerve fibers of different diameters were found in the anterior longitudinal ligament and in the outer region of the disc. In 8 of 10 degenerated discs, fibers were also found in the inner parts of the disc. Substance P-immunoreactive nerve fibers were sporadically observed in the anterior longitudinal ligament and the outer zone of the anulus fibrosus.

CONCLUSIONS

Findings indicate a more extensive disc innervation in the severely degenerated human lumbar disc compared with normal discs. The nociceptive properties of at least some of these nerves are highly suggested by their substance P immunoreactivity, which provides further evidence for the existence of a morphologic substrate of discogenic pain.

A re-evaluation of the ultrastructural localization of 5'-nucleotidase activity in the developing rat cerebellum, with a cerium-based method

The membrane ectoenzyme 5'-nucleotidase converts 5'-AMP into adenosine which, in the nervous tissue, plays an important role as intercellular messenger. Moreover, during histogenesis, 5'-nucleotidase seems to be related to cell proliferation and migration. Conflicting data are reported in the literature about the localization (neuronal or glial) of 5'-nucleotides in the rat cerebellum. In the present report we have analyzed the distribution of 5'-nucleotidase activity with electron microscopy, using a cerium-based method, at different postnatal histogenetic stages (postnatal days (PND) 11, 17, 28). On PND 11 and 17, rims of reaction product outlined the plasma membranes of some neuroblasts in the external granular layer and of parallel fibers and some migrating cells in the developing molecular layer. Positivity was frequently observed on membranes of adjacent neuronal cells and glial processes. Moderate activity was also present on the membranes of granule cells and of mossy fiber rosettes and granule cell dendrites constituting the cerebellar glomeruli within the internal granule cell layer. At PND 28, the reaction product was slightly reduced in some localizations. Cytochemical patterns prove that the cerium-based method is suitable for demonstration of 5'-nucleotidase-specific activity. In fact, a continuous and fine reaction product appears strictly linked to the cell membranes, and no unevenly scattered precipitates can be observed. Data suggest that, during cerebellar histogenesis, 5'-nucleotidase may be involved in the mechanisms of cell migration and proliferation. However, in adulthood, prominent localization of the reaction product on neuronal elements suggests a major role in neuromodulation processes for the enzyme.

Anatomical compartments in the white matter of the rabbit flocculus

The white matter of the rabbit flocculus is subdivided into five compartments by narrow sheets of densely staining acetylcholinesterase-positive fibers. The most lateral compartment is continuous with the C2 compartment of the paraflocculus and contains the posterior interposed nucleus. The other four compartments are numbered from lateral to medial as floccular compartments 1, 2, 3, and 4 (FC1-4). FC1-3 continue across the posterolateral fissure into the adjacent folium (folium p) of the ventral paraflocculus. FC4 is present only in the rostral flocculus. In the caudal flocculus FC1 and FC3 abut dorsal to FC2. Fibers of FC1-4 can be traced into the lateral cerebellar nucleus and the floccular peduncle. The presence of acetylcholinesterase in the deep stratum of the molecular layer of the flocculus and ventral paraflocculus distinguishes them from the dorsal paraflocculus. The topographical relations to the flocculus and the floccular peduncle with group y and the cerebellar nuclei are discussed.

Pyridoxal kinase immunoreactivity in rabbit brain

Murine polyclonal antibody against purified bovine brain pyridoxal kinase (EC 2.7.1.35) was generated and showed cross-reactivity with rabbit brain pyridoxal kinase. This antibody was used to immunohistochemically examine the distribution of pyridoxal kinase in the rabbit brain. The cytoplasm of neuronal cells and neuroglial cells in the cerebral cortex, hippocampal region, brain nuclei and cerebellar cortex showed positive staining with various degrees of intensity. The neuronal cells and surrounding fibers in some brain nuclei, such as the area tegmentalis ventralis or the substantia nigra, showed intense staining. The neuronal cells of the hippocampal region showed somewhat weak reactivity, but some with intense reactivity were found sparsely distributed and positive staining fiber networks of a very low density were also observed.

Interconnections of the upper ventral rami of the human sacral plexus: a reappraisal for dorsal rhizotomy in neurostimulation operations

The extension of a dorsal rhizotomy in bladder stimulation patients is partly determined by connections between the ventral rami of the second, third, and fourth sacral spinal nerves. The literature is inconclusive on interconnections of these ventral rami in the human sacral plexus. The sacral plexuses of ten human cadavers were dissected in this gross anatomy study. In nine cases a branch connecting the ventral rami of the second and third sacral spinal nerves was found. Electron microscopy demonstrated the presence of thick myelinated fibers in this branch. In the male plexuses this branch formed the only link between the second sacral spinal segment and the pelvic plexus. The ventral ramus of the second sacral nerve always contributed to the pudendal nerve, whereas involvement of the ventral rami of the first and third sacral nerves differed individually and intersexually.

Transient expression of stage-specific embryonic antigen-1 (CD15) in the developing dorsal rat spinal cord

The localization of CD15 (synonyms: stage-specific embryonic antigen-1 (SSEA-1), 3(alpha)-fucosyl-N-acetyl-lactosamine or FAL), which is implicated in neuronal differentiation, in the developing dorsal rat spinal cord was studied by immunocytochemistry. A embryonal day 9 (E9), SSEA-1 was detected in the neural ectoderm and, at E11, in cells near the ventricle of the matrix layer. This localization indicated that SSEA-1 is present in proliferating premigratory cells of the rat spinal cord. Between E12 and E16, cells of the alar plate expressed SSEA-1. Expression of the antigen was restricted to neuroblasts that will form the dorsal horn. SSEA-1, therefore, can be used at this stage as a marker for a subdivision of the matrix layer. At E14, the dorsal root entrance zone showed SSEA-1. This indicated that SSEA-1 is associated with ingrowing primary afferents. From E16 on, SSEA-1 was present in the dorsal raphe, which suggested a function for SSEA-1 in the guidance of developing fibres. After E17, the antigen was also found within the dorsal mantle layer. SSEA-1 was first present in Rexed's laminae II, IV and V. Later on in development the antigen was detected only in Rexed's laminae II (substantia gelatinosa). These distribution patterns indicated that SSEA-1 is present on migratory and/or postmigratory cells. In addition, SSEA-1 is associated with small-diameter dorsal root fibres, the C fibres and A(sigma) fibres, that terminate within the substantia gelatinosa. After birth, SSEA-1 was present throughout the dorsal horn, probably as a result of the myelination of the fibres.

The expression of CD15 in dissociated cultured rat dorsal root ganglion cells

This study describes the presence of CD15 in dorsal root ganglia neurons in five experimental conditions: chemically defined medium and the same medium with added nerve growth factor, retinoic acid or antibodies against insulin or tyrosine phosphate. Positive astrocyte controls were used to differentiate the monoclonal antibodies that did not react with CD15. Those monoclonal antibodies which detected CD15 in this positive control were also used to study CD15 positivity in dorsal root ganglion cells. This study shows: (i) masking of the CD15 antibody, which influences the detection capacity of the monoclonal antibodies used; (ii) that CD15 discerns two subpopulations of DRG neurons: a CD15-positive and a CD15-negative population; (iii) that CD15 expression is not involved in the outgrowth of protrusions or the wrapping by non-neuronal cells of DRG neurons.

Expression of the carbohydrate epitope 3-fucosyl-N-acetyl-lactosamine (CD15) in the vertebrate cerebellar cortex

The distribution of the carbohydrate epitope CD15 was investigated on paraffin sections of the brains of man and mammals (monkey, dog, rabbit, rat, mouse, dolphin), reptile, bird and fish by means of immunohistochemistry. This paper demonstrates a differential expression of the CD15 epitope in the cerebella of these various vertebrates. CD15 positivity was found on glial cells and neuronal structures. In adult brains two major distribution patterns were distinguished: one with very intense labelling of the molecular layer, for which the rat is representative, the other with very low immunoreactivity in this layer (mouse). Amongst the rodents (mouse, rat and rabbit), as well as the monkey and human, the positivity in the molecular layer could be attributed to Bergmann fibres of the Golgi epithelial cells. A typical parasagittal band pattern, present in the mouse molecular layer for CD15, which is absent in rat and rabbit molecular layer, is present during human cerebellar development. CD15 positivity on neuronal structures is found on parallel fibres in the developing human, on the lower stellate cells in the dog, and in climbing fibres of the dolphin and, presumably, the catfish too. Moreover, within the parrot cerebellum, large CD15-positive mossy fibre-like endings are found just at the infraplexiform layer.

Spontaneous and stimulated firing in cultured rat suprachiasmatic neurons

Neurons from the suprachiasmatic nucleus (SCN) of the hypothalamus, the site of a circadian pacemaker in mammals, were isolated from embryonic rat. After mechanical dissociation neurons were brought into culture for 1-2 weeks, using a chemically defined medium. Recordings were made from 74 bipolar neurons using two different configurations of the patch-clamp technique. During cell attached patch recordings, 45% of neurons fired spontaneously. The mean firing rate was 0.7 +/- 0.6 Hz and the firing pattern was irregular. In whole cell recordings 73% of the investigated neurons showed spontaneous activity with an irregular firing pattern. The mean spontaneous firing rate with an intracellular Cl- concentration of 145 mM was 1.0 +/- 0.6 Hz. The resting membrane potential of the bipolar neurons was estimated to be -62 +/- 24 mV. An intracellular Cl- concentration of 145 mM depolarised the membrane potential. It also increased the probability of spontaneous firing. A depolarising current stimulus produced an action potential with a threshold voltage of -46 +/- 9 mV. Suprathreshold stimuli resulted in repetitive firing with a mean frequency of 12 +/- 4 Hz. The minimum interspike interval was 52 +/- 14 ms. All action potentials either occurring spontaneously or elicited by current stimuli were abolished by the Na(+)-channel blocker TTX. These results indicate that our cultured neurons have some electrophysiological properties in common with SCN neurons in brain slices and in vivo.

Comparative study of glutamate decarboxylase immunoreactive boutons in the mammalian inferior olive

An antiserum raised against rat glutamate decarboxylase was used to map GABAergic boutons in the inferior olive of rabbit, cat, rhesus monkey, and human. A description of the human periolivary region is also included. The inferior olive of each species contained a dense GABAergic innervation, but immunostaining intensities varied among regions. These intensities were evaluated visually and photometrically, and the sizes and frequencies of occurrence of boutons in various olivary subnuclei were measured. The beta nucleus in all species was intensely immunostained and contained the largest boutons. The caudal subdivision of the dorsal accessory olive stained with a lower intensity than the beta nucleus, but contained similarly large GABAergic boutons. By visual analysis, the rostral subdivision and the subnucleus a of the medial accessory olive and the principal olive were stained with an intermediate intensity, and these regions contained small GABAergic boutons. Photometric analysis of focal regions of the neuropil, however, revealed species differences in teh staining intensity of the principal olive, which was lowest in rabbits and highest in primates. In all species, the lowest immunostaining intensity was observed in the subnucleus b of the medial accessory olive. Species variations in bouton sizes and regional staining intensities were observed in the dorsal cap and the dorsomedial cell column. The heterogeneous staining pattern and regional variation of bouton size argue for the existence of separate GABAergic projections to discrete regions of the inferior olive. Since glutamate decarboxylase immunostaining patterns in the olive are largely similar across species, the afferent projections producing these patterns may also be similar.

The rat inferior olive as seen with immunostaining for glutamate decarboxylase

Boutons presumed to use gamma-aminobutyric acid as neurotransmitter (GABAergic boutons) were detected by glutamate decarboxylase (GAD) immunocytochemistry in all regions of the rat inferior olive. The remarkably high concentration of these boutons allowed a clear visualization of olivary subnuclei boundaries. Regional variations in GAD immunostaining intensity were observed within the nuclear complex and were graded both visually and photometrically. The regional staining variations, for the most part, followed subnuclei boundaries and olivary zonal compartments that have been delineated by the topography of climbing fibre projections. Some subnuclei were grouped by similar staining intensities. The beta nucleus and a medial region in the ventral fold of the dorsal accessory olive were most intensely immunostained, followed by the subnucleus c of the medial accessory olive. Lower staining intensities were observed in the dorsomedial cell column, the dorsal fold of the dorsal accessory olive and the dorsal cap. The lowest intensities were observed in the subnuclei a and b of the medial accessory olive, the ventrolateral outgrowth, the rostral lamella of the medial accessory olive, the principal olive, and the lateral part of the ventral fold of the dorsal accessory olive. The factors contributing to the variations in immunostaining intensity (bouton size and frequency of occurrence) were investigated. The largest boutons were observed in the beta nucleus. Intermediate sized boutons were observed in the dorsomedial cell column, dorsal cap and the dorsal fold of the dorsal accessory olive. The smallest boutons were present in the remaining regions of the inferior olive, including the principal olive, the rostral lamella of the medial accessory olive, and the ventral fold of the dorsal accessory olive. The medial region of the dorsal accessory olive ventral fold contained a higher density of GABAergic boutons than other regions. GABAergic bouton size and innervation density therefore largely accounted for the variations in GAD immunostaining intensity. This study provides a map of the rat inferior olive based on the distribution of GABAergic nerve terminals, and may serve as a basis for characterizing different GABAergic afferent systems in the inferior olive.

Brain enzyme histochemistry following stabilization by microwave irradiation

The activities of various enzymes present in brain homogenates were assayed biochemically (a) with no pretreatment, (b) following a standard microwave treatment in saline and (c) after a standard microwave treatment in formalin. All enzyme activity was lost after the microwave - formalin in treatment. Following microwave - saline treatment, the activities of alkaline phosphatase, 5'-nucleotidase, isocitrate and succinate dehydrogenases were reduced. In contrast, the activities of lactate and malate dehydrogenases were unchanged, and that of acetylcholinesterase apparently increased. Analogous outcomes were seen following attempted histochemical demonstrations of these enzymes. Thus satisfactory histochemical demonstration of all enzymes was achieved (except with alkaline phosphatase, lactate and malate dehydrogenases) following the microwave-saline pretreatment. Since acid phosphatase, catalase and peroxidase were also successfully demonstrated, it seems that microwave-saline pretreatments permit both retention of sufficient enzyme activity for histochemical demonstration to occur and retention of sufficient structural integrity for critical morphological investigations. Since the failure to stain the sites of lactate and malate dehydrogenases is not due to microwave inactivation of these enzymes, their demonstration may be possible by varying the staining procedures.

Glial fiber pattern in the developing chicken cerebellum: vimentin and glial fibrillary acidic protein (GFAP) immunostaining

The possible relation between glial fibers and the formation of longitudinal granule cell migration patterns that occur in the cerebellar anlage of the chicken was investigated by immunocytochemistry of vimentin (monoclonal antibody) and glial fibrillary acidic protein (polyclonal antibody against GFAP, PGF) on fixed and unfixed brain tissues. In addition, neuronal development was studied with a monoclonal antibody for neurofilament. Vimentin was present in radial and tangential fibers in the cerebellar anlage during granule cell migration in almost all parts of the anlage. However, no specific topographic relation of vimentin and GFAP to the migration pattern of granule cells was observed. In adults, Bergmann fibers and astroglia were stained with vimentin antiserum and not with GFAP antiserum. Conclusions are that radial fibers do not determine the formation of longitudinal cytoarchitectonic patterns in the chick cerebellum and that vimentin is the main cytoskeletal component of Bergmann fibers and astroglial cells in embryonic and adult chicken cerebellum.

Cerebellar mutations affecting the postnatal survival of Purkinje cells in the mouse disclose a longitudinal pattern of differentially sensitive cells

The pattern of surviving Purkinje cells (PCs) was investigated in three cerebellar mutant mice with severe postnatal PC death. Two of these mutations, nervous (nr) and Purkinje cell degeneration (pcd) mutations are already well characterized. The third mutation is a new one, which appeared spontaneously in DW/J-Pas mice and was called tambaleante (tbl). PCs were identified by immunocytochemistry using an antibody against vitamin D-dependent calcium-binding protein which labels all the PCs in adult control mice. In each of the three mutations, surviving PCs are arranged according to a different and reproducible pattern which is symmetric relative to the midline. In NR and young PCD mutants, PCs are closely packed in broad sagittal bands. In TBL, they are more loosely arranged in a rather patchy pattern. In PCD and in TBL mutants the death of resistant PCs is only shortly delayed but in NR there is little change in the number of surviving PCs after 3 months. The differential sensitivity of subsets of PCs to the effect of nr, pcd, and tbl mutations is topographically determined. These results provide a new evidence of the PC heterogeneity which has been previously demonstrated by histochemical and immunohistochemical techniques. Moreover, in the anterior vermis of control mice, three thin sagittal bands of PCs are labeled by the Q113 monoclonal antibody. Similarly, in the anterior lobe of the NR cerebellum, the thin longitudinal strips of missing PCs coincide with the absence of Q113 immunoreactivity: in this region the nr mutation affects specifically the survival of Q113 positive cells. However, other clusters of Q113 immunoreactive PCs do survive in NR mice suggesting that susceptibility to the nr mutation and Q113 positivity are two independent markers of the underlying PC compartmentalization.

The microwave Rio-Hortega technique: a 24 hour method

Application of microwaves in histochemistry and cytochemistry generally speeds up the technique. Microwaves stimulate diffusion into the tissue and influence the proteins and membrane of the cell. Silver impregnation techniques for the brain, such as the fast Rio-Hortega or the slow Golgi-Cox technique, normally require a minimum time period of 7 days and 20 days respectively. Using microwaves, the Rio-Hortega technique can be completed within 24 h. In sections prepared from mature brains, good silver impregnation of cell bodies, of axons and their terminals, and of dendrites and their spines are obtained. An explanation is given as to why the method cannot be further reduced in time. To our knowledge, this is the first report of the application of microwave irradiation for colouring pieces of tissue.