Membrane structure at synaptic junctions in area CA1 of the rat hippocampus

In tissue from area CA1 of the rat hippocampus prepared for electron microscopic study by thin-sectioning, asymmetric synaptic junctions were found on dendritic spines, spiny dendritic shafts, and non-spiny dendritic shafts. In freeze-fractured preparations, aggregates of large particles were found on the extracellular half of the postsynaptic membrane at each of these synaptic junctions. Particle aggregate areas were measured and particle packing densities were computed at dendritic spine synapses and dendritic shaft synapses in area CA1, and compared to similar measures of particle aggregates on dendritic spines of cerebellar Purkinje cells. All of these CA1 and cerebellar synapses are excitatory and are thought to use glutamate as a neurotransmitter. There was a tendency for the dispersion of particles within individual aggregates to be less uniform in larger aggregates in both area CA1 and cerebellar cortex. Distinct particle-free zones could be distinguished in the center of particle aggregates on large "mushroom-shaped" spines in area CA1. There was no statistically significant difference between the particle densities at CA1 dendritic spines (2848 +/- 863 particles/micron2) and CA1 dendritic shafts (2707 +/- 718 particles/micron2). However, the average density of particles at cerebellar dendritic spine synapses (3614 +/- 1081 particles/micron2) was significantly greater than at dendritic spine or shaft synapses found in area CA1. Symmetric synaptic junctions were observed on the CA1 pyramidal cell somas and dendritic shafts in thin-sectioned preparations. These synapses typically exert an inhibitory action mediated by gamma-aminobutyric acid. In freeze-fracture preparations, large varicosities were found apposed to the pyramidal somal and dendritic membranes, but there were no specializations of particle distribution on either the extracellular or the cytoplasmic half of the fractured postsynaptic membranes. This finding parallels observations from freeze-fracture preparations of other GABAergic synapses in the central nervous system.     

Large myelinated club endings on the Mauthner cell in the goldfish

Summary Thin sectioning and freeze-fracturing have revealed the distribution of gap junctions and chemical synapses in the synaptic interface of the large myelinated club endings on the lateral dendrite of the goldfish Mauthner cell. In 12 samples of club endings fractured completely or nearly completely, the apposed synaptic membrane area averaged 39.090 m², of which 16.6% was occupied by gap junctions and about 4 to 5% by the active zones of chemical synapses. The numerical profile density (number per unit area of the synaptic membrane) of gap junctions varied greatly, from 1.78 to 6.30, and was mostly in inverse proportion to their size. The chemical synapses were located mainly in two places: in the circumferential rim of the synaptic membrane next to the widened extracellular space, and in the margins of intraterminal invaginations of the synaptic cleft. The axoplasm of the preterminal axon, just after losing its myelin sheath, was filled with microtubules, among which neurofilaments gathered into many small bundles. The correlation between the areas of gap junctions and the chemical synapses and the amplitude of the excitatory postsynaptic potentials (EPSP) is discussed.     

The relationship between some measures of synaptic ultrastructure as a function of distance from the soma on lamprey reticulospinal neurons

Summary Synaptic junctions located on the dendrites of lamprey (Petromyzon marinus) reticulospinal neurons labelled with intracellularly-injected horseradish peroxidase were studied. The normal ultrastructure of the synaptic junctions was defined and several quantitative measures made from each junction in order to test the hypothesis that distally-located synapses are ultrastructurally different from those located at proximal dendritic sites. A total of 820 contacts from one neuron and 279 from a second neuron ranging from 20 to 340 m from the soma were quantified. The vast majority of the presynaptic endings contained round, clear-cored vesicles and formed an asymmetrical membrane differentiation with the postsynaptic dendrite. A small fraction of the population contained flattened or pleomorphic vesicles and these synapses were equally distributed with respect to distance from the soma. Many of the terminals contained a few large dark- and clear-cored vesicles. Four quantitative measures of each synaptic contact were made. These included vesicle number, length of differentiated membrane, vesicle area and terminal area. Four ratios relating the different quantitative measures were also calculated. Each ratio or measurement from the synaptic junctions was plotted as a function of distance from the soma to determine if differences existed at any distance. It was found that synaptic junctions are uniformly similar and that distal junctions did not differ significantly (P > 0.05) from those at proximal dendritic sites. It is concluded that if distal synapses do compensate for their remote location they do this in some other way, possibly by increasing the number of synaptic contacts made by each presynaptic axon.     

The mormyrid brainstem--III. Ultrastructure and synaptic organization of the medullary "pacemaker" nucleus

The ultrastructure and synaptic organization of the presumed medullary pacemaker nucleus, nucleus c of the weakly electric mormyrid fish, Gnathonemus petersii has been investigated. Nucleus c consists of about 12-15 small (20-25 micron) neurones (P-cells), which form a group situated ventrally to the medullary relay nucleus and embedded in a neuropil of myelinated fibres and dendritic processes. The P-cells often exhibit an enhanced electron density of their cytoplasm and dendroplasm. They possess several dendrites of different diameter, a short, thin axon initial segment and a thickly myelinated axon running in dorsal direction. The pacemaker neurons are interconnected by complex electronic coupling, established by somatosomatic, dendrosomatic and dendrodendritic gap junctions. Perikarya and dendrites are frequently interconnected serially by gap junctions; dendrites showed sometimes triadic gap-junction arrangement. It is suggested that this high degree of electrotonic coupling amongst the pacemaker cells represents the first level of the highly ordered synchronization processes which characterize the electric discharge command system of Gnathonemus. Pacemaker cells receive synaptic input from club endings with mixed synapses and from bouton-like terminals with chemical synapses, both of them originating from medium-sized myelinated fibres and contacting mainly neuronal perikarya and dendritic processes. The axon initial segment receives only few synaptic inputs. Bouton-like terminals were found to be of two types according to their vesicle content, namely, boutons with ovoid, clear synaptic vesicles forming Gray type-1 synapses and boutons with pleomorphic clear synaptic vesicles forming Gray type-2 synapses. Different functional roles for the two types of boutons in modulating pacemaker cell activity are suggested.     

Different kinds of axon terminals forming symmetric synapses with the cell bodies and initial axon segments of layer II/III pyramidal cells. II. Synaptic junctions

Summary Four different types of axon terminals form symmetric synapses with the cell bodies and initial axon segments of pyramidal cells in layer II/III of rat visual cortex. One type belongs to chandelier cells, and the other three kinds of terminals have origins that have not been established yet. These latter are referred to as large, medium-sized and dense terminals. The purpose of the present study was to examine the synaptic junctions formed by all four types of terminal. The synapses formed by the chandelier cell terminals are readily recognized in thin sections because of the characteristic features of both the terminals and the initial axon segments, which are the neuronal elements postsynaptic to them. In en face views of these axo-axonal synapses the junctions can be seen to have presynaptic dense projections that form a grid in which they are triagonally spaced, and have an average centre-to-centre spacing of 84 nm. As an ensemble the projections form the presynaptic grid, which usually has an oval or round outline, but may be notched on one side where projections are absent. The synaptic junctions of the large, medium-sized and dense terminals were examined by making reconstructions of the terminals from serial thin sections. It was found that at the interfaces between the axon terminals and the cell bodies of pyramidal cells, several separate synaptic junctions may be present, in addition to a number of puncta adhaerentia. Thus, there may be as many as five separate synaptic junctions and as few as one. It was also found that while the proportion of the area of the synaptic interface occupied by synaptic junctions was between 12% and 26% for dense terminals, for medium sized terminals it was 10–15%, and for the one large terminal reconstructed it was only 8%. Thus, there can be multiple synaptic junctions between each of these types of axon terminals and a pyramidal cell, and because many of the terminals forming symmetric junctions are boutons en passant, a number of vesicle release sites exist between the presynaptic axon and its postsynaptic partner. The axon terminals forming symmetric synapses in the cerebral cortex are assumed to be inhibitory, and consequently it is suggested that this arrangement of multiple release sites is designed to ensure that stimulation of the presynaptic axon results in an effective level of hyperpolarization of the postsynaptic neuron.     

Proline-rich synapse-associated protein-1 and 2 (ProSAP1/Shank2 and ProSAP2/Shank3)—scaffolding proteins are also present in postsynaptic specializations of the peripheral nervous system

Proline-rich synapse-associated protein-1 and 2 (ProSAP1/Shank2 and ProSAP2/Shank3) were originally found as synapse-associated protein 90/postsynaptic density protein-95-associated protein (SAPAP)/guanylate-kinase-associated protein (GKAP) interaction partners and also isolated from synaptic junctional protein preparations of rat brain. They are essential components of the postsynaptic density (PSD) and are specifically targeted to excitatory asymmetric type 1 synapses. Functionally, the members of the ProSAP/Shank family are one of the postsynaptic key elements since they link and attach the postsynaptic signaling apparatus, for example N-methyl-d-aspartic acid (NMDA)-receptors via direct and indirect protein interactions to the actin-based cytoskeleton. The functional significance of ProSAP1/2 for synaptic transmission and the paucity of data with respect to the molecular composition of PSDs of the peripheral nervous system (PNS) stimulated us to investigate neuromuscular junctions (NMJs), synapses of the superior cervical ganglion (SCG), and synapses in myenteric ganglia as representative synaptic junctions of the PNS. Confocal imaging revealed ProSAP1/2-immunoreactivity (-iry) in NMJs of rat and mouse sternomastoid and tibialis anterior muscles. In contrast, ProSAP1/2-iry was only negligibly found in motor endplates of striated esophageal muscle probably caused by antigen masking or a different postsynaptic molecular anatomy at these synapses. ProSAP1/2-iry was furthermore detected in cell bodies and dendrites of superior cervical ganglion neurons and myenteric neurons in esophagus and stomach. Ultrastructural analysis of ProSAP1/2 expression in myenteric ganglia demonstrated that ProSAP1 and ProSAP2 antibodies specifically labelled PSDs of myenteric neurons. Thus, scaffolding proteins ProSAP1/2 were found within the postsynaptic specializations of synapses within the PNS, indicating a similar molecular assembly of central and peripheral postsynapses.     

Ultrastructure of the rectifying electrotonic synapses between giant fibres and pectoral fin adductor motor neurons in the hatchetfish

Summary Synapses formed by giant fibres on pectoral fin adductor motor neurons were identified by horseradish peroxidase (HRP) injection. The synapses were distributed in clusters on the somata and proximal dendrites of the motor neurons. All of the labelled synapses contained synaptic vesicles and often had clearly defined active zones characteristic of chemical synapses. Some synapses also showed gap junctions with the motor neuron soma, often directly adjacent to an active zone. The gap junctions were asymmetrical, with a thick layer of electron dense material on the postsynaptic side. Previous electrophysiological data indicate that giant fibre inputs to motor neurons are purely electrotonic and that these electrical synapses rectify.     

Synaptic plasticity in the oedematous human cerebral cortex

Synaptic plastic changes are fundamental events which occur spontaneously during development, maturity and aging processes or can be induced by injury or trauma. To examine lesion-induced synaptic plasticity, cortical biopsies were taken from the frontal, parietal, temporal and occipital cortex of living patients during neurosurgical treatment of brain trauma, brain tumours and vascular malformations, and processed for transmission electron microscopy. Enlargement of both pre- and postsynaptic endings, irregularly shaped, lobulated, stellate and bifurcated presynaptic endings and conformational changes of dendritic spines were observed. Numerous flat, curved and invaginated axodendritic and axospinous asymmetric synapses were distinguished and a smaller proportion of axodendritic and axosomatic symmetric synapses. Activated or sensitized synapses showed numerous frontline spheroid synaptic vesicles, prominent dense presynaptic dense projections and increased length of synaptic membrane complex. Perforated synapses, multiple synapses and serial synapses were also found evincing synaptic splitting and formation of new synaptic connections. The overall images suggest increased number of excitatory circuits, which were correlated with the tonico-clonic convulsion or post-traumatic seizures observed in some patients. Numerous coated vesicles were observed in pre- and postsynaptic structures. Increased number of polyribosomes were found in the dendritic shafts. The dilated spine apparatus, the coated vesicles and the increased number of polyribosomes seem to represent a system for synthesis, transport and storage of synaptic proteins for the formation of new synapses. Coexisting synaptic plasticity and synaptic degeneration were observed in the patients under study. Dendritic and astrocyte synapse-like junctions were also characterized.     

Elements of molecular machinery of GABAergic signaling in the vertebrate cholinergic neuromuscular junction

It is generally accepted that gamma-aminobutyric acid (GABA) is a signaling molecule abundant in central synapses. In a number of studies though, it has been shown that GABA signaling functions in the peripheral nervous system as well, in particular, in the synapses of sympathetic ganglia. However, there exists no firm evidence on the presence of GABAergic signaling cascade in the intercellular junctions of the somatic nerve system.By the use of immunohistochemistry methods, in the synaptic area of cholinergic neuromuscular contact in rat diaphragm, we have detected glutamate decarboxylase, the enzyme involved in synthesis of GABA, molecules of GABA, and also GAT-2, a protein responsible for transmembrane transport of GABA. Earlier we have also shown that metabotropic GABA_B receptors have overlapping localization in the same compartment. Moreover, activation of GABA_B receptors affects the intensity of acetylcholine release. These data taken together, allows us to suggest that in the mammalian cholinergic neuromuscular junction, GABA is synthesized and performs certain synaptic signaling function.     

The formation of photoreceptor synapses in the retina of larvalXenopus

Summary The first appearance and early development of synapses between photoreceptors, bipolar cells and horizontal cells were studied in the retina of larvalXenopus, between stages 37/8 and 46. Synapse morphology was reconstructed from examination of serial sections. Additional data were obtained from E-PTA stained tissue. Photoreceptors initially contacted horizontal cell dendrites via ribbon synapses at the photoreceptor basal surface (stages 37/8-39). Later, horizontal cell dendrites penetrated the receptor as digitiform processes (stages 39–40) which subsequently expanded extensively within the invagination (stages 40–44). At stages 41–46, a photoreceptor ribbon generally was associated with two horizontal processes in a synaptic dyad complex; each ribbon participated in 1 or 2 such complexes. Bipolar cell dendrites made contact with receptors via superficial or invaginating basal-type junctions. Such junctions were observed first at stage 39, but increased greatly in number beginning at stage 42. Basal junctions were characterized by marked staining of the paramembranous receptor cell cytoplasm and a 9–13 nm wide cleft containing an electron-dense material. No association of bipolar cell dendrites with photoreceptor ribbons was noted.Junctional complexes presumed to be gap junctions first were observed at stage 39 between photoreceptor bases and basal processes emitted by neighbouring photoreceptors.     

Effect of chronic vitamin E deficiency on the synapses of cerebellar glomeruli in young rats

A morphometric investigation has been carried out on the synaptic junctions in the cerebellar glomeruli of young-adult rats chronically deprived of vitamin E for 10 months and control animals of the same age. The following parameters were evaluated: the average length of the synapses (L), the numerical (NV) as well as the surface (SV) density of the synaptic contact zones. The results from these experimental groups were compared with data from young, adult and old rats. The results obtained show a significant decrease of the surface density of the synaptic contact zones in old and alpha-tocopherol deprived young-adult (11-month-old) rats as compared to younger and normally fed animals. This reduction of the synaptic contact area seems to be due to the marked decline in the number of synapses found in both cases. The average size (L) of the synaptic junctions, on the other hand, was increased in alpha-tocopherol deficient rats as compared to normally fed littermates. The significant reduction of the synaptic contact area in old and vitamin E deprived young rats supports the hypothesis that a common denominator may be responsible to explain this alteration. Because of the recognized protective role of alpha-tocopherol against free radical attacks on plasma membranes, the present findings support an involvement of membrane structural alterations in aging as well as in vitamin E deficiency.     

Ultrastructure of substance P immunoreactive elements in the periaqueductal gray matter of the rat

Substance P (SP) is a non-opioid peptide that generates a potent, analgesia when injected into the periaqueductal gray matter (PAG)The aim of this study was to investigate the fine neuronal structures and synaptic circuits involved in SP action in rats by means of electron microscopy, using immunocytochemical (ICC) pre-embedding methods. A conventional ultrastructural study, carried out to interpret the ICC data correctly, shows small sized nerve cell bodies with a high nucleus–cytoplasmic ratio; absence of an extensive granular endoplasmic reticulum; and few axo-somatic contacts having symmetrical and asymmetrical junctions in equal proportions. The large neuropil is characterized by numerous thin unmyelinated axons and axodendritic synapses mainly showing pleomorphic vesicles and asymmetrical junctions. The ICC analysis showed moderately labeled nerve cell bodies with the same structural, synaptic, and dimensional features as the negative cells. In the neuropil SP immunoreactivity is shown by dendrites, synapses, and thin elements which are unidentifiable structurally. No SP terminals synapsing on SP nerve cell bodies were found and only occasional SP light labeled terminals synapsing on negative perikarya were seen. The SP boutons generally have pleomorphic vesicles and asymmetrical junctions. On the basis of these data a possible excitatory activity of PAG SP synapses could be hypothesized. This activity would take place on postsynaptic neurons generally at a dendritic level. Our ultrastructural findings give support to an excitatory role carried out by SP neurons of the PAG, as suggested by the role of PAG circuitry on spinal nociception.     

Intrinsic differences in sensitivity to 5-HT between high- and low-output terminals innervating the same target

The differential action of neuromodulators on synapses of various efficacy provides additional fine tuning of synaptic regulation beyond frequency induced plasticity. We used the well-characterized high- and low-output motor nerve terminals, of the tonic and phasic neuromuscular junctions (NMJs) in the walking leg extensor muscle of the crayfish, to investigate differential actions of serotonin (5-HT) since both terminals innervate the same target. The excitatory postsynaptic potentials of the tonic NMJ are enhanced to a greater extent than for the phasic NMJs during exposure to 5-HT (100 nM). Macropatch current recordings at identified sites along the motor nerve terminals and quantal analysis indicate that mean quantal content is substantially increased by 5-HT. The overall probability of vesicular release increases to a greater extent at tonic terminals than at phasic terminals when exposed to 100 nM 5-HT. Measures in the area (i.e. charge) of spontaneous quantal currents indicate no difference in postsynaptic receptivity to the glutamatergic synaptic transmission upon exposure to 5-HT. The results provide new details concerning differential modulation of low- and high-output synapses present on the same target tissue.     

Enlargement of synaptic size as a compensative reaction in aging and dementia.

A quantitative investigation has been carried out on synaptic contact zones of dentate gyrus supragranular layer and cerebellar glomeruli in autoptic samples from adult, old and demented patients. During physiological aging and senile dementia, the synaptic average area was significantly increased as compared to adult values in both the CNS areas investigated. Conversely, the number of contacts and their total surface contact area per unit volume of tissue were decreased. Current literature reports that, in animal models, enlarged synapses undergo perforations and splitting to modify synaptic connectivity. As against these assumptions, the increased synaptic size observed in our study appears to represent a compensative reaction of old and demented CNS to counteract the reduction in number and in total contact area of the synaptic junctions.     

Morphological correlates of synaptic transmission in lamprey spinal cord.

The dye Procion brown was used to identify in the light and electron microscope, synaptic contacts made between monosynaptically coupled neurons in the lamprey spinal cord whose synaptic interaction had been recorded. Synaptic contacts were made on different dendrites of the postsynaptic cell at different distances from the soma. Some of the contacts were made on dentritic spines and some on the smooth shaft of the dentrites. Serial sections through synaptic contacts made on dendritic processess of the postsynaptic cells were used for three-dimensional reconstruction of the synapses using computer graphics techniques. The computer reconstructions and detailed examination of the serial EM micrographs revealed the large proliferation of membrane involved in making these en passant synapses as well as the morphological changes due to stimulation of the presynaptic axon. These changes include depletion of synaptic vesicles and formation of complex vesicles and synaptic cisternae. Besides chemical synaptic contacts, four electrotonic contacts were located, confirming the mixed electrochemical synaptic response recorded from the postsynaptic cell. The mean quantum content was estimated and compared with the estimate of the available transmitter pool, assuming the quantal release hypothesis applies at these synapses. The total transmitter pool was estimated by counting all synaptic vesicles in all synaptic contacts. It was estimated that about 6% of the total transmitter pool is available for release at these synapses. This compares with less than 1% at the neuromuscular junction and about 20% at sympathetic synapses. These results support the hypothesis that synaptic vesicles may be recycled as described by Heuser and Reese (22) at the neuromuscular junction. Ongoing studies are investigating the effect on a variety of synaptic junctions to stimulation for different periods of time of presynaptic axons. The methods described in this study can also be used to test the models of synaptic interaction on dendritic trees described by Rall (39) and Jack and Redman (24).     

UNC-13 is required for synaptic vesicle fusion in C. elegans.

We analyzed the synaptic physiology of unc-13 mutants in the nematode C. elegans. Mutants of unc-13 had normal nervous system architecture, and the densities of synapses and postsynaptic receptors were normal at the neuromuscular junction. However, the number of synaptic vesicles at neuromuscular junctions was two- to threefold greater in unc-13 mutants than in wild-type animals. Most importantly, evoked release at both GABAergic and cholinergic synapses was almost absent in unc-13 null alleles, as determined by whole-cell, voltage-clamp techniques. Although mutant synapses had morphologically docked vesicles, these vesicles were not competent for release as assayed by spontaneous release in calcium-free solution or by the application of hyperosmotic saline. These experiments support models in which UNC-13 mediates either fusion of vesicles during exocytosis or priming of vesicles for fusion.     

Ultrastructure of substance P immunoreactive elements in the periaqueductal gray matter of the rat.

Substance P (SP) is a non-opioid peptide that generates a potent analgesia when injected into the periaqueductal gray matter (PAG). The aim of this study was to investigate the fine neuronal structures and synaptic circuits involved in SP action in rats by means of electron microscopy, using immunocytochemical (ICC) pre-embedding methods. A conventional ultrastructural study, carried out to interpret the ICC data correctly, shows small sized nerve cell bodies with a high nucleus-cytoplasmic ratio; absence of an extensive granular endoplasmic reticulum; and few axo-somatic contacts having symmetrical and asymmetrical junctions in equal proportions. The large neuropil is characterized by numerous thin unmyelinated axons and axo-dendritic synapses mainly showing pleomorphic vesicles and asymmetrical junctions. The ICC analysis showed moderately labeled nerve cell bodies with the same structural, synaptic, and dimensional features as the negative cells. In the neuropil SP immunoreactivity is shown by dendrites, synapses, and thin elements which are unidentifiable structurally. No SP terminals synapsing on SP nerve cell bodies were found and only occasional SP light labeled terminals synapsing on negative perikarya were seen. The SP boutons generally have pleomorphic vesicles and asymmetrical junctions. On the basis of these data a possible excitatory activity of PAG SP synapses could be hypothesized. This activity would take place on postsynaptic neurons generally at a dendritic level. Our ultrastructural findings give support to an excitatory role carried out by SP neurons of the PAG, as suggested by the role of PAG circuitry on spinal nociception.     

Neurobeachin, a protein implicated in membrane protein traffic and autism, is required for the formation and functioning of central synapses

The development of neuronal networks in the brain requires the differentiation of functional synapses. Neurobeachin (Nbea) was identified as a putative regulator of membrane protein trafficking associated with tubulovesicular endomembranes and postsynaptic plasma membranes. Nbea is essential for evoked transmission at neuromuscular junctions, but its role in the central nervous system has not been characterized. Here, we have studied central synapses of a newly generated gene-trap knockout (KO) mouse line at embryonic day 18, because null-mutant mice are paralysed and die perinatally. Although the overall brain architecture was normal, we identified major abnormalities of synaptic function in mutant animals. In acute slices from the brainstem, both spontaneous excitatory and inhibitory postsynaptic currents were clearly reduced and failure rates of evoked inhibitory responses were markedly increased. In addition, the frequency of miniature excitatory and both the frequency and amplitudes of miniature inhibitory postsynaptic currents were severely diminished in KO mice, indicating a perturbation of both action potential-dependent and -independent transmitter release. Moreover, Nbea appears to be important for the formation and composition of central synapses because the area density of mature asymmetric contacts in the fetal brainstem was reduced to 30% of wild-type levels, and the expression levels of a subset of synaptic marker proteins were smaller than in littermate controls. Our data demonstrate for the first time a function of Nbea at central synapses that may be based on its presumed role in targeting membrane proteins to synaptic contacts, and are consistent with the 'excitatory–inhibitory imbalance' model of autism where Nbea gene rearrangements have been detected in some patients.     

Synaptic structure of the lateral line lobe nucleus in mormyrid fish

Electron microscopical observations reveal a complex synaptic structure (Fig. 1) in the nucleus of the lateral line lobe (nLLL). Different types of axosomatic and axoaxonic synapses are demonstrated to be in contact with the large cells. The results furnish morphological evidence for electrotonic transmission (by way of club endings with gap junctions) at this level of the electrosensory pathway of mormyrid fish. A new ultrastructural finding is the existence of presynaptic gap junctions on the unmyelinated surface area of the club endings.     

Unipolar brush cells develop a set of characteristic features in primary cerebellar cultures

The unipolar brush cells (UBCs) are a class of excitatory interneurons recently discovered in the cerebellar granular layer. UBCs differ morphologically and biochemically from granule cells, although they share the same mossy fiber and Golgi cell inputs. To elucidate development of the UBCs, we sought to ascertain their presence in primary cerebellar cultures and the class-specific properties they develop in vitro outside of the context of the tissue. By light and electron microscopy, we demonstrate that primary cultures from embryonic and postnatal mouse and rat cerebella contain UBC-like neurons that are highly polarized and can be distinguished from granule cells on several grounds. Granule cells are more numerous in dissociated postnatal cultures than in embryonic cultures; express little, if any, calretinin immunoreactivity; and develop dendritic processes devoid of typical claw-like endings, but provided with small synaptic junctions. By contrast, UBC-like neurons occur more frequently in embryonic cultures than in postnatal cultures, are intensely calretinin-positive, and develop characteristic cell organelles, dendrites, and large synapses. In embryonic cultures, the UBC-like neurons have a clear nucleus and contain a special cytoplasmic array of ringlet subunits, resembling the botrysome. At 12–28 days in vitro, the UBC dendrites contain abundant mitochondria, are provided with clusters of non-synaptic appendages, and engage in glomerular arrays together with large and small axon terminals. The large terminals contain round synaptic vesicles, form extensive, asymmetric synapses with the cell bodies and the dendrites of the UBC-like neurons, and resemble mossy terminals, while the small terminals contain pleomorphic vesicles, form symmetric synaptic junctions, and resemble Golgi terminals. In postnatal cultures grown for 12 days, UBC-like neurons are rare and resemble in most aspects the cells observed in embryonic cultures, although they rarely develop elaborate dendritic brushes.