Intracellular labeling of neurons in the medial accessory olive of the cat: III. Ultrastructure of axon hillock and initial segment and their GABAergic innervation

Morphological correlates of bilateral synchrony in the rat cerebellar cortex

Simultaneous recordings of the left and right crus IIA of the cerebellar cortex in the rat have demonstrated that Purkinje cells of both sides can be activated synchronously by their climbing fibers. Because climbing fibers arise exclusively from the contralateral inferior olive (IO), this physiological finding seems to contradict the anatomy. To define the structural basis responsible for the bilateral synchrony, we examined the possibilities that bilateral common afferent inputs to the IO and interolivary connections form the underlying mechanisms. The bilaterality of the major afferents of the olivary regions that project to crus IIA was studied using Phaseolus vulgaris leucoagglutinin as an anterograde tracer. We found that the excitatory and inhibitory projections from the spinal trigeminal nucleus and dorsolateral hump of the interposed cerebellar nucleus to the transition area between the principal olive and dorsal accessory olive were bilateral. A second possible mechanism for bilateral synchrony, which is the possibility that axons of olivary neurons provide collaterals to the contralateral side, was investigated using biotinylated dextran amine as an anterograde tracer. Labeled axons were traced and reconstructed from the principal olive and dorsal and medial accessory olive up to the entrance of the contralateral restiform body. None of these axons gave rise to collaterals. The possibility that neurons in the left and right IO are electronically coupled via dendrodendritic connections was investigated by examining the midline region of the IO. The neuropil of the left and right IO is continuous in the dorsomedial cell column. Examination of Golgi impregnations of this subdivision demonstrated that (1) many dendrites cross from one side to the other, (2) neurons close to the midline give rise to dendrites that extend into both olives, and (3) dendrites of neurons in the dorsomedial cell column frequently traverse into adjacent olivary subdivisions such as the medial accessory olive and the transition area between the principal olive and dorsal accessory olive. Sections immunostained for dendritic lamellar bodies or GABAergic terminals showed the same pattern: the neuropils of the dorsomedial cell columns on both sides form a continuum with each other as well as with the neuropil of other adjacent olivary subdivisions. Ultrastructural examination of the dorsomedial cell column demonstrated that the midline area includes many complex glomeruli that contain dendritic spines linked by gap junctions. To verify whether the complex spike synchrony observed between left and right crus IIA could indeed be mediated in part through coupled neurons in the dorsomedial cell column, we recorded simultaneously from crus IIA areas and from left and right vermal lobule IX, which receives climbing fibers from the dorsomedial cell column. In these experiments we demonstrated that the climbing fibers of all four areas, i.e., the left and right crus IIA as well as the left and right lobule IX, can fire synchronously. The present results indicate that synchronous climbing fiber activation of the left and right crus IIA in the rat can be explained by (1) bilateral inputs to the transition areas between the principal olive and dorsal accessory olive and (2) dendrodendritic electrotonic coupling between neurons of the left and right dorsomedial cell column and between neurons of the dorsomedial cell column and adjacent olivary subdivisions.

Synapses on axons of sympathetic preganglionic neurons in rat and rabbit thoracic spinal cord

Axosomatic and axodendritic synapses occur on sympathetic preganglionic neurons, but it is not yet known whether their axons receive synaptic input, which could be particularly effective at regulating sympathetic outflow. Here, we examined retrogradely labelled sympathetic preganglionic axons to see if they received synapses. Cholera toxin B subunit (CTB) or CTB conjugated to horseradish peroxidase (CTB-HRP) was used to label neurons projecting to the rat or rabbit superior cervical ganglion, the rat adrenal medulla, or the rabbit stellate ganglion. At the light microscopic level, small groups of CTB-immunoreactive axons travelled through the ventral horn near its lateral boundary, with occasional axons taking a more medial course. The axons passed through the ventrolateral funiculus to exit at the ventral roots. In parasagittal section, a few axons branched within the ventral horn, sending processes rostrally and caudally for short distances before they turned ventrally to exit the spinal cord. At the ultrastructural level, CTB-immunoreactive rat and rabbit sympathetic preganglionic axons were almost exclusively unmyelinated. In contrast, labelling with CTB-HRP revealed both myelinated and unmyelinated axons in the ventral horn, the ventrolateral white matter, and the ventral roots. CTB-HRP also allowed the detection of the initial segment of a sympathetic preganglionic axon. Synapses, with vesicles clustered presynaptically and membrane specializations postsynaptically, were found on some unmyelinated CTB-immunoreactive axons. Occasional axons received several synapses. Synapses were most common on CTB-containing axons just ventral to the intermediolateral cell column. One synapse was found on an axon within 2 microns of its origin from a proximal dendrite. Rare synapses were found several hundred micrometers ventral to the intermediolateral cell column. One branching axon had synapses just below the branch point on both the main axon and the axonal branch. These findings indicate an extensive synaptic input to the axons of at least some sympathetic preganglionic neurons. These axoaxonic synapses could have a profound effect on sympathetic activity.

The cisternal organelle as a Ca(2+)-storing compartment associated with GABAergic synapses in the axon initial segment of hippocampal pyramidal neurones

The axon initial segment of cortical principal neurones contains an organelle consisting of two to four stacks of flat, membrane-delineated cisternae alternating with electron-dense, fibrillar material. These cisternal organelles are situated predominantly close to the synaptic junctions of GABAergic axo-axonic cell terminals. To examine the possibility that the cisternal organelle is involved in Ca2+ sequestration, we tested for the presence of Ca(2+)-ATPase in the cisternal organelles of pyramidal cell axons in the CA1 and CA3 regions of the hippocampus. Electron microscopic immunocytochemistry using antibodies to muscle sarcoplasmic reticulum ATPase revealed immunoreactivity associated with cisternal organelle membranes. The localisation of Ca(2+)-ATPase in cisternal organelles was also confirmed by enzyme cytochemistry, which produced reaction product in the lumen of the cisternae. These experiments provide evidence for the presence of a Ca2+ pump in the cisternal organelle membrane, which may play a role in the sequestration and release of Ca2+. Cisternal organelles are very closely aligned to the axolemma and the outermost cisternal membrane is connected to the plasma membrane by periodic electron-dense bridges as detected in electron micrographs. It is suggested that the interface acts as a voltage sensor, releasing Ca2+ from cisternal organelles upon depolarisation of the axon initial segment, in a manner similar to the sarcoplasmic reticulum of skeletal muscle. The increase in intra-axonal Ca2+ may regulate the GABAA receptors associated with the axo-axonic cell synapses, and could affect the excitability of pyramidal cells.

Axon hillocks and initial segments in spinal trigeminal nucleus with emphasis on synapses including axo-axo-axonic contacts

As a part of a continuing study of the feline spinal trigeminal nucleus, the fine structure and synaptic arrangements on the axon hillock and axon initial segment of neurons in this region are described here. Transmission electron microscopy has been used to characterize qualitatively the axon hillock and initial segment and associated synapses in pars interpolaris. Axon hillocks and initial segments are easily identified in continuity with somata or as isolated profiles in the neuropil, and they receive synaptic contacts: these we regard as axo-axonic. The presynaptic terminals contain either mainly round or mainly flattened synaptic vesicles and have Type I (asymmetric) or Type II (symmetric) thickenings respectively at their contacts with the axon hillock or initial segment. I report here also the unusual arrangement of three separate axons in a serial synaptic complex. Some of the round vesicle Type I contacts onto the axon hillock-initial segment region also receive Type II contacts from one or more flattened vesicle terminals, thus forming an axo-axo-axonic complex. These flattened vesicle terminals lack the usual features of a presynaptic dendrite. It has been shown that in this nucleus some round vesicle terminals, especially those postsynaptic to flattened vesicle terminals, are primary afferents from the periphery. Therefore the round vesicle terminal presynaptic to the axon hillock-initial segment region, some of which are included in the axo-axo-axonic complex may also be a primary afferent directly contacting the spike generator area of the relay neuron and under presynaptic control of a flattened vesicle synapse. The latter may possibly be an intrinsic contact. This strategic situation of round vesicle terminals and the axo-axo-axonic complex at the axon hillock or initial segment has major implications relevant to the overall output of these neurons.

Electron microscopy of in vivo recorded and intracellularly injected inferior olivary neurons and their GABAergic innervation in the cat

This paper reports on the detailed morphology of inferior olivary neurons in the cat following electrophysiological examination, intracellular injection with horseradish peroxidase, and gamma aminobutyric acid (GABA) immunocytochemistry. The activity of olivary cells was recorded intracellularly in vivo and their response to mesodiencephalic stimulation was tested. In a number of cases their response to stimulation of the contralateral superior cerebellar peduncle was also tested. Mesodiencephalic stimulation resulted in monosynaptic, and superior peduncle stimulation in disynaptic activation of cells in the medial accessory and principal olivary subdivisions. Rebound olivary activity was usually only found after mesodiencephalic stimulation. Light microscopic investigation of osmicated and Araldite embedded Vibratome sections was facilitated considerably when performing the osmication in a glucose solution. Peroxidase labeled olivary cells, like that earlier described for Golgi-impregnated material, possess a complex globular dendritic geometry. Especially, and unlike Golgi material, the abundance of exceptionally long and complex spiny appendages could be appreciated. The axons usually stemmed from first order dendrites and did not give rise to recurrent axon collaterals. The ultrastructural analysis of this material, mainly from serial sections, was combined with postembedding GABA immunohistochemistry. In this way, GABAergic as well as non-GABAergic profiles were studied in conjunction with HRP labeled cellular elements. The GABAergic terminals usually contained pleomorphic vesicles and made symmetrical synapses whereas non-GABAergic terminals nearly always formed asymmetrical synapses and contained round or oval vesicles. Most, if not all, HRP labeled spiny appendages were incorporated in glomeruli. A particular spiny appendage may contribute more than one spine head to a glomerular core, which, on average, consisted of spiny elements of six different neurons. A glomerular core is surrounded by approximately the same amounts of GABAergic and non-GABAergic boutons. Also, all spiny appendages, and most of their individual spine heads, are contacted by GABAergic as well as non-GABAergic boutons. Spiny appendages on the axon hillock may be incorporated in dendritic glomeruli, however, most synapses with the hillock were made by GABAergic boutons. The combined physiological and morphological observations imply that 1) the cerebellar nuclei can exert an excitatory influence on inferior olivary neurons through a mesodiencephalic relay, 2) the GABAergic nucleo-olivary input seems to be capable of diminishing the oscillatory tendencies of olivary neurons, and 3) the mesodiencephalic (non-GABAergic) and cerebellar (GABAergic) input may subserve a timing function since these inputs systematically impinge upon the same olivary spines.

Fine structure of the dorsal cap of the inferior olive and its GABAergic and non-GABAergic input from the nucleus prepositus hypoglossi in rat and rabbit

The dorsal cap of the inferior olive is involved in the control of eye movements and is excited by inputs from the midbrain. In the present study we attempted to determine the inhibitory input to this nucleus in rat and rabbit. The projection from the nucleus prepositus hypoglossi to the dorsal cap was studied in the light microscope by anterograde tracing of Phaseolus vulgaris-leucoagglutinin and lesion-induced depletion of glutamic acid decarboxylase immunoreactivity, and in the electron microscope by anterograde tracing of wheat germ agglutinin-coupled horseradish peroxidase combined with GABA immunocytochemistry. We show that the nucleus prepositus hypoglossi projects bilaterally to the dorsal cap, contralaterally to the ventrolateral outgrowth, and ipsilaterally to the medial accessory olive. After lesioning of the nucleus prepositus hypoglossi, the caudal dorsal cap was depleted of most of its glutamic acid decarboxylase-immunoreactive terminals while the rostral dorsal cap and the ventrolateral outgrowth were depleted of a minor part. Ultrastructural analysis indicates that the majority, but not all, of the terminals from the nucleus prepositus hypoglossi in the dorsal cap are GABA-positive. These GABA-positive and GABA-negative terminals form predominantly symmetric and asymmetric synapses; most of them synapse on dendrites outside and inside glomeruli, frequently in association with dendrodendritic gap junctions, while a small minority are axosomatic. None of the terminals from the nucleus prepositus hypoglossi was found to form a crest synapse, although synapses of this kind were predominantly formed by GABAergic terminals. This study shows that the dorsal cap receives a major inhibitory input from the nucleus prepositus hypoglossi, the terminals of which are located at strategic positions on the olivary neurons.

The GABAergic cerebello-olivary projection in the rat

Immunocytochemical detection of glutamate decarboxylase (GAD), the predominant biosynthetic enzyme of gamma-aminobutyric acid (GABA), reveals the presence of a dense GABAergic innervation in all parts of the inferior olive. One brain center that provides a substantial projection to the inferior olive is the cerebellar nuclei, which contain many small GABAergic neurons. These neurons were tested as a source of GABAergic olivary afferents by combining retrograde tract tracing with GAD immunocytochemistry. As expected from previous studies, injections of wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP) into the inferior olive retrogradely label many small neurons in the interposed and lateral cerebellar nuclei and the dorsal part of the lateral vestibular nucleus, and fewer neurons in the ventro-lateral region of the medial cerebellar nucleus. These projections are predominantly crossed and are topographically arranged. The vast majority, if not all, of these projection neurons are also GAD-positive. The relative contribution of this projection to the GABAergic innervation of the inferior olive was tested by lesion of the cerebellar nuclei, or the superior cerebellar peduncle. Within 10 days the lesion eliminates most GAD-immunoreactive boutons in the principal olive, the rostral lamella of the medial accessory olive, the ventrolateral outgrowth, and the lateral part of the dorsal accessory olive ventral fold. Thus, the effectiveness of this depletion demonstrates that the cerebellar nuclei provide most of the GABAergic innervation to regions of the inferior olive known to receive a cerebellar projection. Moreover, when the lateral vestibular nucleus is damaged, the dorsal fold of the dorsal accessory olive is depleted of GABAergic boutons. The synaptic relations that boutons of the GABAergic cerebello-olivary projection share with olivary neurons were investigated at the electron microscopic level by GAD-immunocytochemistry, anterograde degeneration of the cerebellar axons or anterograde transport of WGA-HRP. All of these methods confirm that GABAergic, cerebello-olivary axon terminals contain pleomorphic vesicles, and synapse on various portions of olivary neurons, and especially on dendritic spines within glomeruli, often in very close proximity to the gap junctions that characteristically couple the dendritic profiles. These results demonstrate four major points: that virtually all of the GABAergic, and presumably inhibitory, neurons of the cerebellar and dorsal lateral vestibular nuclei are projection neurons; that a large portion of the inferior olive receives GABAergic afferents from the cerebellar nuclei; that a portion of the dorsal accessory olive receives GABAergic afferents from the dorsal lateral vestibular nucleus; and that cerebello-olivary fibers often synapse near gap junctions, and therefore could influence electrical coupling of olivary neurons.

Intracellular labeling of neurons in the medial accessory olive of the cat: I. Physiology and light microscopy

This study is the first of three reports on the detailed morphology of horseradish peroxidase injected neurons in the medial accessory olive of the cat. Intracellular, in vivo recordings of olivary cells were made and their response to mesodiencephalic stimulation was tested. In 44 units a short latency action potential could be recorded, which was very suggestive of a monosynaptic excitatory pathway. The short latency response was frequently followed by a long latency (mean 188 msec) or rebound action potential. Recordings were followed by intracellular iontophoresis of horseradish peroxidase. A total of 21 neurons, all located within the medial accessory olive were chosen for morphological analysis. Cells could be divided into two categories on the basis of their overall morphological appearance. Type I cells (n = 5) had sparsely branching dendrites that radiated away from the soma and were usually found in the caudal part of the medial accessory olive. The axon usually originated from the soma. Type II cells (n = 16) were located more rostrally. They had larger cell bodies with dendrites that ramified extensively, forming a globular structure (mean diam. 338 microns). The axon usually originated from a first order dendrite. No recurrent axon collaterals were observed on either type I or II cells. Both cell types carried long and complex spiny appendages; however, they were most numerous on the second and higher order dendrites of type II cells. Since the soma of these cells is usually not found in the centre of its dendritic field, even if the cell is located in the center area of the neuropil, it is suggested that the dendritic trees of up to 100 neurons may be intricately interwoven, establishing clusters with intensive intercommunication by means of dendritic gap junctions. The abundance, length and complexity of the spiny appendages suggest an important role in this process, but may also be relevant instruments in enhancing the computational capabilities of these neurons, especially in time sensitive processes. When relating the physiological and the morphological results, it was noted that both type I and type II cells could respond to mesodiencephalic stimulation and were both able to trigger a rebound action potential. No significant correlations were found between cell size and the latency of the rebound.

Intracellular labeling of neurons in the medial accessory olive of the cat: II. Ultrastructure of dendritic spines and their GABAergic innervation

In order to describe the morphology of dendritic spines of identified neurons in the cat inferior olive together with their gamma-aminobutyric acid (GABA) synaptic input, a technique was used combining intracellular labeling of horseradish peroxidase with postembedding gold-immunocytochemistry. With this technique physiologically identified olivary cells were reconstructed with the light microscope, and the horseradish peroxidase reaction product and immunogold labeling were subsequently examined in serial sections at the ultrastructural level. In addition, a degenerating neuron was observed, resulting in a triple labeling in single ultrathin sections. Quantitative and three-dimensional analysis showed that the dendritic spines were composed of long, thin stalks ending in one or more spine heads. The spines of cells located in the caudal half of the medial accessory olive (type I cells, characterized by dendrites which run away from the soma) were found to be less complex than those of cells located rostrally in this olivary subnucleus (type II cells, characterized by dendrites which tend to turn back towards the soma). Most, if not all, of the spines of both cell types were located within glomeruli. On average, the spines within individual glomeruli originated from 6 different dendrites (with a maximum of 8). Different spines within the same glomerulus were never derived from different dendrites of the same olivary neuron, but single spines frequently gave rise to several spine heads, which could be located either within different glomeruli or inside a single glomerulus. The glomerular spine heads originating from the same spine were rarely located near one another. All spines and most of the spine heads were contacted by both GABAergic and non-GABAergic terminals. Most of the GABAergic terminals contained pleomorphical vesicles and displayed symmetric synapses whereas the non-GABAergic terminals showed usually round to oval vesicles and asymmetric synapses.