Extrinsic cytological determinants of basket and stellate cell dendritic pattern in the cerebellar molecular layer

Electron microscopic autoradiography of the uptake of [3H]GABA in dispersed cell cultures of rat cerebellums. I. The morphology of the GABAergic synapse

Dispersed cell cultures of rat cerebellums 21 days in vitro were incubated in saline containing 0.5 micrometer [3H]GABA for 5 min, fixed and prepared for electron microscopic autoradiography. Quantitative methods for analyzing the resulting autoradiographs were developed to distinguish lightly labeled neurons from heavily labeled (i.e. GABAergic) neurons. After a 3 week exposure, 2% of the presynaptic elements were found to be GABAergic. A longer incubation time of 15 min resulted in a greatly increased grain density over lightly labeled neurons, making it more difficult to distinguish them from the GABAergic neurons. When the concentration of [3H]GABA was increased to 1.0 micrometer, the grain densities over both GABAergic and other neurons increased to a similar extent. The morphology of the GABAergic synapses differed from that of the total population synapses. The GABAergic synapses had larger cross-sectional areas, lower densities of vesicles, greater numbers of vesicles, and thinner postsynaptic densities. Both the GABAergic and total population had similar lengths of the postsynaptic densities, cleft widths, vesicle sizes and types of postsynaptic elements. The morphology of the GABAergic synapse reported here is similar to that described for basket and stellate presynaptic elements in vivo.

Laminar differentiation of the hippocampus, fascia dentata and subiculum in developing rats, observed with the Timm sulphide silver method

The laminar staining of the rat hippocampal region with the Timm sulphide silver method is from studies on adult rats known to depend on the various fibersystems terminating in these laminae. In order to illustrate the development of these fibersystems the laminar differentiation of the Timm staining of fascia dentata, hippocampus and subiculum is presented for rats between 1 and 31 days old. Corresponding changes in cytoarchitectonics as revealed by thionin staining are briefly demonstrated. Even on the first postnatal day there are indications of the mature, laminar staining pattern, and between three and nine days all the laminae corresponding to the terminal fields of the major afferent and intrinsic systems appear. After 12 days only minor additions to the laminar pattern develop, but there are adjustments of absolute and relative dimensions of layers and fields, and also the staining densities of individual laminae change. These observations are in good correlation with the available information on both hippocampal neurogenesis and cytodifferentiation, and the few fiber tracing studies performed on the developing hippocampal region. Compared to the latter, which ideally marks only one system or one lamina per animal, the Timm method provides an excellent means for getting an overview of the general developmental situation at the different ages. Thus developmental gradients along septotemporal, medio-lateral and basal-apical axes are found; which should be heeded in future studies on hippocampal synaptogenesis. The observations are discussed in relation to current models for neuronal growth and formation of nervous connections.

Postnatal development of the granule cell in the kitten cerebellum

Newborn to 9-week-old kittens were anesthetized and then sacrificed by inserting a scalpel into the thoracic cavity and severing the descending aorta. The cerebella were removed and processed according to Golgi-Cox and rapid Golgi modifications. The results indicate that the dendritic processes of the granule cells undergo a marked postnatal maturation even though there is little postmigratory change in the size and shape of the cell body. A club-shaped swelling of the distal portions of the dendrites is observed by the end of the first week; this "claw" becomes multilobated by three weeks and displays elaborate digitiform projections by nine weeks. The sequence of postnatal development in the kitten appears to follow a time course similar to that observed in other species. A possible correlation between the postnatal development of the granule cell and the postnatal development of the spincocerebellar system is discussed.

The cerebellum of the bullfrog tadpole (Rana catesbeiana)

The cerebellum of the premetamorphic bullfrog tadpole differs from the cerebellum of other frog species in its morphology and maturational state. The cell mass beneath the floor of the lateral recess and bordering its lateral wall that has been reported to form the auricular lobe in other species is absent, the auricular lobe abutting the medial wall of the lateral recess instead and is continuous with the corpus cerebelli. The corpus cerebelli, although immature and yet to acquire an external granular layer, is already massive and displays an incipient molecular, Purkinje cell and granular layers. Cytodifferentiation in the auricular lobe and corpus cerebelli is similar, their constituent cells being in various stages of development. Fully mature cells are absent, but a small population of Purkinje cells and glia in the auricular lobe and along the marginal zone of the corpus cerebelli show advanced development. The orientation of these Purkinje cells is parallel to the pia and appears to approximate the course of the vestibulo-lateral commissural fibers. In the ventral part of the corpus cerebelli, developing climbing fibers are present but Purkinje cells are poorly developed.

Histogenesis of mouse cerebellum in microwell cultures. Cell reaggregation and migration, fiber and synapse formation

A microwell culture system was developed for analysis of cell movements and interactions during nervous system histogenesis. Cells from trypsinized 7-day-old mouse cerebellum reaggregated within hours into clusters which later developed interconnections consisting of either sheets of migrating cells and cell processes or cables of fiber bundles with cells migrating along their surfaces. Granule cells in several stages of differentiation, basket and/or stellate neurons, some larger neurons, and two types of neuroglial cells were identified in reproducible, nonrandom patterns by scanning and transmission electron microscopy. Axonal and dendritic processes, both with growth cones, and numerous synapses were generated in vitro.

The effects of early hypo- and hyperthyroidism on the development of rat cerebellar cortex. III. Kinetics of cell proliferation in the external granular layer

The effects of early hypo- and hyperthyroidism on the rates of cell acquisition and proliferation have been studied in the external granular layer (EGL) of the developing rat cerebellar cortex at 10 days of age using quantitative autoradiographic methods. Both altered thyroid states reduce the rate of cell acquisition in the EGL, but appear to do so for different reasons. Hyperthyroidism shortens the average length of the cell cycle by decreasing the duration of the pre-DNA synthetic phase (G1), indicating that excess thyroxine may exert a direct effect on the EGL. This action involves the early onset of neuronal differentiation (cessation of proliferation)46 which presumably leads to the observed decrease in the rate of cell acquisition (increased doubling time). Such differentiating cells do not, however, leave the proliferative zone or the EGL prematurely, resulting in a reduced labeling index, mitotic index, and growth fraction as non-dividing cells dilute the proliferating cell population. Hypothyroidism, on the other hand, leads to no significant change in the length of the cell cycle or in the mitotic index, but causes a decreased labeling index and growth fraction, as well as a reduced rate of cell acquisition (increased doubling time). No significant change in the amount of cell death in the EGL could be found to explain this apparent discrepancy between the rate of cell proliferation (cell cycle length) and cell acqusiition. The answer to this puzzle appears to lie in the mitotic index, which is not affected to the same extent as the labeling index, although it is also slightly reduced. If cells were to remain longer in mitosis, this could result in a decreased labeling index and growth fraction but nearly normal mitotic index and cell cycle length (as measured using the % labeled mitoses method), since those cells dropping out of the cycling population would be counted as mitoses...

Effects of low-level x-irradiation on cat cerebella at different postnatal intervals. I. Quantitative evaluation of morphological changes

Whole head fractionated doses of 200 r and 150 r were initiated postnatally in five experimental age groups (birth, 1-week, 2-week, 3-week and 4-week) and continued over a period of 14 or 20 days to prevent reconstitution of the external granular layer. Animals irradiated at birth displayed minor deficits in behavior, which included ataxia, tremor, hypertonus and dysmetria, while animals irradiated at 1-week showed only mild symptoms of hypermetria. All other animals displayed no motor deficits. Animals irradiated at birth had smaller eyes and ears, a reduction in the size of the entire head and were susceptible to seizures. All animals were sacrified at 70 days of age. The cerebellum was found to be reduced in size and weight, the greatest deficit being seen in animals x-irradiated at the very early ages. Newborn condition animals were found to have large compliments of interneurons in the molecular layer, an established internal granular layer, and Purkinje cells were found to have a normal orientation, position, and to be unreduced in number or size. Total granule cell deficits were found to range from 83% at birth to 29% at four weeks. Quantitative changes for the molecular layer, internal granular layer, medullary layer, Purkinje cell to granule cell ratio, and granule cell density, all depicted the greatest changes in the newborn, 1-week and 2-week conditions. This experiment confirmed that the critical period in the development at which damage would result in behavioral abnormalities was from birth to five days, while for neuroanatomical abnormalities, this critical period was from birth to 18 days.

Effects of low-level x-irradiation on cat cerebella at different postnatal intervals. III. Changes in the morphology of interneurons in the molecular layer

The whole head of infant kittens was irradiated with fractionated doses of 150 R and 200 R at different postnatal intervals. Experimental age conditions consisted of a newborn, 1-week, 2-week, 3-week, and a 4-week age condition while the age of sacrifice remained constant at 70 days. Going analysis revealed that the interneurons found in the molecular layer of the newborn, 1-week and 2-week condition were basket cells. Stellate cells were found in increasing numbers in the 3-week and 4-week conditions. Basket calls were found to occupy the entire molecular layer in the newborn and 1-week conditions resulting in an abnormal axonal plexus. The position of the basket cell soma was found to be in the same plane as its axonal projection. The branching point of the primary dendrite of Purkinje cells was found to correlate with the amount of molecular layer occupied by the abnormal basket cell plexus. Interneurons in the molecular layer were found to show different dendritic growth patterns dependent upon where their soma was located. Interneurons in the deep molecular layer showed only a decreased dendritic field, whereas interneurons in the middle and superficial moledular layer had an abnormal growth of dendrites into the deep molecular layer. A hypothesis is presented to account for the decreased dendritic arborization and the increased length of the primary dendrite before branching, the ectopic basket cell and their abnormally directed dendritic growth, and the difference in behavioral deficits between the newborn and the 1-week conditions.

The growth of the dendritic trees of Purkinje cells in irradiated agranular cerebellar cortex

The heads of noenatal Wistar rats were irradiated with 200 rads daily from birth to the 10th day post-partum. Ten litters each containing 5 animals were killed at 30 days post-partum and their brains treated by the Golgi-Cox technique. The dendritic trees of 24 Purkinje cells were analysed using the quantitative technique of network analysis, and comparisons made between parameters obtained from 20 normal Purkinje cells. All dendritic trees in agranular irradiated cortex were markedly reduced in size (as indicated by total dendritic length and total number of segments) although mean path lengths were normal. Segment lengths were normal over proximal branches, but uniformly increased over distal branches. Abnormal appendages, called 'giant spines' were observed on many dendrites. They were often some 10 mum in length and their presence effectively reduced segment lengths, increased the frequency of trichotomy and deviated growth from the normal random terminal pattern so that long collateral branching topologies were formed. Nevertheless, trichotomy was uniformly reduced in those trees without 'giant spines' and the distribution of branching patterns suggested that growth had proceeded by random terminal dichotomy. These results demonstrate that the development of dendritic trees is retarded in the agranular irradiated cerebellum, where synaptogenesis is very greatly reduced below normal. The quantitative changes in segment lengths, size of trees, and trichotomy accord with those predicted by the filopodial synaptogenic hypothesis of dendritic growth formulated by Vaughn et al. 99, whilst the results of the topological analysis suggest that branching is established by a degree of non-random interaction between growing dendrites and their substrate. 'Claw-like' dendritic complexes within some Purkinje cell trees may have been induced by aberrent fibre bundles of few surviving granule cells.

Retrosplenial cortex in the rhesus monkey: a cytoarchitectonic and Golgi study

The laminar and cellular structure of retrosplenial cortex in the rhesus monkey was studied with Nissl stained and rapid Golgi impregnated tissue and the results were used to evaluate morphological features of a cortical transition zone. The granular layer of retrosplenial granular cortex is composed primarily of small, density packed, star pyramidal cells. These cells branch within the granular layer itself, while the apical dendrite enters layer I where it branches infrequently or not at all. This cell type is similar to the star pyramid first described by Lorente de No except in its areal and laminar distribution. Cytoarchitectonic observations of retrosplenial agranular cortex show, that, although this area is relatively "agranular" in comparison to other cortical areas, it does possess an incipient layer II and layer IV. These layers are composed mainly of small and medium sized pyramidal cells, but many non-pyramidal cell types were found in these and other layers in this area in rapid Golgi preparations. Stellate cells with beaded or smooth, lightly spinous dendrites were found throughout layer I-IV, while fusiform cells with smooth or very lightly spinous dendrites appear in layers III-VI. Areas surrounding retrosplenial cortex in the posterior cingulate region were also evaluated in Nissl and Golgi preparations including the indusium griseum, subiculum (dorsal to the corpus callosum) and area 23. The laminar and cellular constitutents of retrosplenial cortex were then evaluated in the context of cortical architectonic transition. The transition from one cellular layer in the indusium griseum to five cellular layers in area 23 is made by the addition of layers II, III, IV and VI in retrosplenial cortex to the one ganglionic layer of the indusium griseum and subiculum. Besides the addition and subdivision of layers in retrosplenial cortex, two aspects of cell morphology were found to change in this region. First, the structure of pyramidal cells progressively changes from those in the indusium griseum which have predominently round or oval somata and a preponderance of apical and few basal dendrites to those in layer V of retrosplenial cortex and area 23 which have pyramidal shaped somata and a great number of basal dendrites which branch frequently and spread horizontally for hundreds of microns. Second, there is a change in the number and distribution of non-pyramidal cell types. Evidence was not found that the indusium griseum, dorsal subiculum or layer V of retrosplenial granular cortex contain a significant number of stellate or fusiform cells. At the retrosplenial granular/agranular border, though, these cells gradually begin to constitute a greater proportion of the cell population and in area 23 form a major component of layer IV...

Synapse formation in the mouse olfactory bulb. I. Quantitative studies

A quantitative study of synapse formation in the mouse olfactory bulb has been carried out using serial sections. Volumetric synaptic density as well as absolute number of synapses per olfactory bulb for eight distinct synaptic types have been determined at 15 different ages, from the beginning of synapse formation at embryonic day 14 (E14) to postnatal day 44 (P44). Synapses are first found in appreciable numbers at E15 when both axo-dendritic and a few dendro-dendritic synapses occur in the presumptive glomerular layer. Initial synapse formation correlates closely with the reorientation of mitral cells from a primitive tangenital to a definitive radial direction. Synapse formation by mitral cell dendrites occurs after mitral cell axons have grown into the future olfactory cortical areas, either simultaneous with or before synapse formation by these axons. Virtually all synaptic types detected in adults have been found on the day of birth, consistent with the idea that olfaction is an important sensory modality for newborn mice. Volumeric density of a given synaptic type generally increases 50--100 times during development while the absolute number increases about 1,000 times. Synapses in glomeruli develop more precociously than those in the time of origin and differentiation of the principal postsynaptic elements of these two divisions (mitral cells and internal granule cells). Correlation of the time of synapse formation of various synaptic types with their putative excitary or inhibitory role determined in adult studies suggests that excitatory synapses generally form somewhat earlier, although throughout nearly all of synaptic development, both excitatory and inhibitory synapses are present. Reciprocal dendro-dendritic synapses in the external plexiform layer appear to have a special mode of formation. It is suggested that a granule-to-mitral dendro-dendritic synapse only forms next to an already existing mitral-to-granule synapse on the same gemmule.

The organization and postnatal development of the commissural projection of the rat somatic sensory cortex

Anterograde and retrograde tracing experiments have been used to demonstrate the origin and terminal distribution of commissural fibers in the first somatosensory cortex (SI) of the rat. The commissural fibers originate from pyramidal cells of all layers, but predominantly from layers III and V. The fibers terminate in a series of approximately vertical bands. In each of these there are concentrations of terminals extending from the inner portion of the molecular layer to the deep portion of layer III as well as in the superficial part of layer V, and in layer VI. Discrete vertical bands of cortex are reciprocally connected across the midline to give both the origin and terminal regions of the projection a patchy or "columnar" appearance. The commissural fibers arise from and terminate in areas of the cortex that lie between and alongside the aggregations of granule cells that distinguish SI of the rat. No commissural fibers terminate within the aggregations of layer IV cells themselves but the more superficial terminal ramifications may come to overlie these aggregations. A heterotopic projection to the contralateral second somatosensory cortex has been observed and is similar in form to the homotopic projection to SI. Many commissural fibers have crossed the midline in the corpus callosum by the day of birth but lie in the underlying white matter and do not enter the cortical plate until at least the third postnatal day. During the first postnatal week these fibers grow somewhat diffusely into the maturing cortex and their topographic and laminar pattern of distribution attains its adult characteristics by the end of the first week. Commissural axons, thus, arise from immature cells but the maturation of cell form seems to precede the ingrowth of these axons and the acquisition of commissural synapses.

Differentiation of Bergmann glia cells in the cerebellum: a golgi study

The migration of Bergmann glia cells from the external granular layer to the Purkinje cell layer and the differentiation of their processes were analyzed in Golgi-Cox preparations of the cerebellum of the rat. During their migration they showed numerous cytoplasmic excrescences and long thick processes, which suggested that they may follow ameboid form of movement for their migration. The differentiation of their processes was observed to involve a series of complex events related to retraction of the thick apical processes from the tips of which sheaves of filiform processes emerged, growth in the number and the thickness of the filiform processes, resorption of some of these filiform processes, cyclic changes in the appearance, resorption and reappearance of excrescences on the filiform processes, and changes in size and shape of the growth cones. These findings are discussed in the context of the role played by the Bergmann glia cells in the postnatal developmental events in the cerebellum, and the dynamic nature of differentiation of their processes.

Differentiation of Purkinje cells and their relationship to other components of developing cerebellar cortex in man

The differentiation of Purkinje cells and their relationship to other components of the developing cerebellar cortex were analyzed by the Golgi impregnation method and by electron microscopy in human specimens of various pre- and postnatal ages. The three stages of Purkinje cell maturation that have been previously recognized in other species are also evident in man: the first stage occupies primarily the fourth fetal month (12-16 weeks); the second stage lasts through the fifth, sixth and seventh feta months (16-28 weeks); the third stage extends throughout the remaining period of intrauterine life and the first postnatal year and continues at a slow rate thereafter. During the first stage, Purkinje cells are distributed in a layer, several rows deep. Their bipolar somas are relatively smooth and have only a few processes at the apical and basal cell poles. In the 3-month period of the second stage, Purkinje cells become gradually organized into a single row. Their somas become invested with additional randomly oriented dendritic processes and numerous somatic spines (pseudopodia). The first morphologically well-defined synapses appear on the Purkinje cell somatic spines and on their immature dendritic shafts at the beginning of the second stage and become more prominent during the period from 18 to 24 weeks. In the third stage, the dendritic arbor becomes flattened in the plane transverse to the folium and somatic spines disappear. Spines appear on the secondary and tertiary dendrites between the twenty-fourth and twenty-eighth fetal weeks and continue to increase in number during the entire third stage as new dendritic branches develop. These observations indicate that cellular maturation and synaptogenesis in the primate cerebellum differ from these events in non-primate species, with respect to time of birth, in the relative duration of each phase and in the total time necessary for neuronal differentiation. The protracted time of differentiation and the slow growth of Purkinje cell dendrites in man may be due to the numerically complex relationships existing between granule and Purkinje cells. It is probably not simply a reflection of the larger size of human Purkinje cells and their dendrites.

Early formation of synapses in the molecular layer of the fetal rat cerebellum

The initial phase of synaptogenesis in the molecular layer of the albino rat cerebellum has been studied quantitatively and qualitatively with the electron microscope. Using both aldehyde-osmium (AL-OS) and ethanolic phosphotungstic acid (E-PTA) preparations, synapses were observed as early as embryonic day 19 throughout the entire rostro-caudal extent of the vermal molecular layer. The undeveloped nature of the elements participating in the formation of these synapses precluded their identification, for at this stage of cerebellar development, the Purkinje cells have not developed characteristic dendrites and both the massive production of cells within the external granule layer and the migration of neuroblasts through the molecular layer are in their initial phases. Evidence of synapse formation at this early stage of molecular layer development indicates that the rudiments of neuronal systems are present before the bulk of the molecular layer has come into existence and before most of the constituent elements of the cerebellum have differentiated. From embryonic day 19 to embryonic day 22, the number of E-PTA synaptic contacts was always less than the number of AL-OS synapses. This decrepancy and the variations observed in the morphology of the presynaptic components of E-PTA stained synaptic contacts, indicate that E-PTA morphology, and particularly the development of presynaptic dense projections, represents a late stage of synapse formation.

Purkinje cell dendritic alterations after transient developmental injury of theexternal granular layer

To examine the influence of granule cell axons on morphogenesis of their major synaptic target, the Purkinje cell dendrites, a transient injury of the external granular layer of developing rat cerebellum was induced by injecting 10 mg/kg methylazoxymethanol acetate on postnatal days 1, 2, 3 and 4. The antiproliferative action of the drug resulted in a diminution of the mitotic population which slowed the expansion of the internal granular and molecular layers. This was followed by a reconstitution of a thicker than normal external granular layer and a late phase of extensive growth. Perturbations of Purkinje cell dendritic morphology induced by this altered timing of granule cell accumulation were studied with Golgi-Cox impregnation techniques. Systematic observation were made of a number of defined abnormalities of Purkinje cells which included somas positioned below the normal monolayer; elongated dendrites; multiple somatic dendrites; S-shaped dendrites; dendrites disoriented away from the radial direction; and portions of dendrites below the soma of origin. The results indicated that most of these 'abnormalities' occur to a small extent in normal cerebellum but that these experimental conditions induced a bias in favor of their appearance. Some dendritic abnormalities, i.e., dendrites tilted out of the sagittal plane, dendrites directed toward the white matter, and apical dendrites growing vertically between cells of the external granular layer, were not noted in normal cerebellum and we attribute these particular changes to unique developmental possibilities occuring only after a more or less severe injury. This analysis of bias in dendritic morphology subsequent to subtle developmental injury provides informat on on processes of normal development and on the genesis of phylogenetic variants of Purkinje cell morphology.

Timing of major ontogenetic events in the visual cortex of the rhesus monkey

The place and time of origin, the migration and eventual disposition of neurons of the monkey visual cortex were studied by autoradiography in animals killed at various intervals after 3H-thymidine pulse labeling at embryonic (E) and early postnatal (P) stages. All neurons destined for the visual cortex are generated during about a 2-month period between E45 and E102. Neuron position in the cortical laminae correlates systematically with time of cell origin; neurons destined for deeper cortical positions are generated earlier, and more superficial ones progressively later. Thus, most neurons in layer VI are born between E45 and E60, in layer V between E60 and E70, in layer IV between E70 and E80, and in layers III and II between E80 and E102. No neurons, but numerous glia, are generated within the cortical plate itself. Initially young neurons are produced almost exclusively in the ventricular zone. Later they are probably generated in both ventricular and subventricular zones, and by the end of the proliferative period the subventricular zone becomes the predominant source of new cells. At the time when all neurons destined for the monkey visual cortex have already been produced, primary fissures are barely indicated on the cerebral surface and no secondary fissure appears as yet. Autoradiographic analyses indicate that at early stages young neurons move to the cortical plate relatively synchronously and at a fast rate, whereas at later stages there are considerable differences in the rates of cell migration. At early stages when the migration pathway is relatively short, the external process of the ventricular cell may stretch across almost the entire migratory distance. It is possible that nuclei move without interruption within their own cylinders of cytoplasm, a mechanism which might account for the rapid, synchronous movement of cell bodies as seen in the autoradiographic material. During later stages when young neurons, mostly of subventricular origin, move across a distance which is more than 10 times the length of their leading process, migrating cells follow radial glial guides across the widened intermediate zone and through densely packed cortical plate. This type of cell displacement seems to proceed less synchronously and require more time. It is proposed that two different mechanisms of cell displacement might exist: one in operation at early stages of cortical formation when mostly ventricular cells migrate, and another utilized by subventricular cells at later stages.

Weaver mutant mouse cerebellum: defective neuronal migration secondary to abnormality of Bergmann glia

Previous work showed that in cerebella of mice homozygous for the autosomal mutation weaver, wv, most postmitotic granule cell neurons die during the first 2 weeks after birth close to their site of genesis in the external granular layer. Analysis of the less severely affected heterozygotes by electron microscopy and autoradiography indicates that granule cell death occurs several days after cell genesis and is secondary to failure of their somas to migrate across the molecular layer to the granular layer. This migration defect in turn appears secondary to a hitherto unrecognized disorder of Bergmann glial cells, the cells that normally guide the young neurons in their migration. In +/wv cerebella, Bergmann glial processes are enlarged and irregular in caliber, electronlucent, and often vacuolated; in wv/wv, Bergmann cell processes are almost absent. The primary genetic abnormality remains undefined, but the gene dosage effect, here recognized at a cellular level for the first time in a mammalian neurological mutant, suggests that even though neuronal death serves as the most prominent and clinically relevant phenotypic expression, the Bergmann glial abnormality may actually be closer to the primary cellular target of the wv genetic locus.