Structural and ultrastructural study of the intracranial portion of the oculomotor, trochlear and abducent nerves in dog.

The right intracranial portion of the oculomotor, trochlear and abducent nerves were removed from six adult German shepherd dogs and analysed by light and electron microscopy. In all cases the nerve sectional area was calculated. Unmyelinated and myelinated fibres were analysed and number, diameter and cross-sectional area were calculated. In myelinated fibres, also calculated were the corresponding axon area and diameter, and myelin sheath thickness. The mean number of myelinated fibres was 8543.50 +/- 1231.85 being the unmyelinated 1402 +/- 241.58 in the oculomotor nerve; 1509 +/- 223.17 and 287.67 +/- 72.28 in the trochlear nerve and 2473.00 +/- 211.41 and 231.25 +/- 92.67 respectively in the abducent. The mean diameter was 10.23 +/- 0.68 microm in myelinated and 0.43 +/- 0.21 for unmyelinated in oculomotor nerve, 10.53 +/- 0.55 microm and 0.33 +/- 0.04 for the trochlear, and 10.45 +/- 1.27 microm and 0.47 +/- 0.09 in the abducent nerve respectively. This study reveals that oculomotor, trochlear and abducent nerves of the dog show structural and ultra-structural features similar to the same nerves in other species.

Mapping and quantitative analysis of gephyrin cytoplasmic trafficking pathways in motoneurons, using an optimized Transmission Electron Microscopy Color Imaging (TEMCI) procedure.

In the present study, an optimized Transmission Electron Microscopy Color Imaging (TEMCI) procedure was used to map and quantify the pathways involved in the trafficking and subcellular targeting of gephyrin in identified abducens motoneurons. Gephyrin is a scaffolding protein, which plays a crucial role in the clustering of the GABA(A) and glycine receptors to the cytoskeleton. TEMCI associated several accurate tools: (i) nanogold immunodetection of gephyrin in motoneurons identified on the basis of their immunoreactivity to Choline Acetyl Transferase, (ii) low magnification color scale coding of gephyrin densities on series of ultrathin sections of motoneurons, which gave a map of the cytoplasmic distribution of the protein, (iii) statistical analysis of the subcellular distribution of the immunolabeling. The color map of gephyrin densities in the cell bodies reflected the distribution of inhibitory synapses over the membrane. The TEMCI analysis of motoneurons with various patterns of synaptic covering made it possible to visualize for the first time the cytoplasmic transport pathway of gephyrin towards its target at synaptic contact. A high magnification quantitative analysis, including the study of 109 inhibitory synapses, showed that most gephyrin-associated immunogold particles (67%) were located in the subsynaptic regions facing the active zones, and the second most densely occupied regions were the perisynaptic regions (19.5% of immunogold particles). A consistent proportion of the gephyrin (11.5%), significantly higher than densities present in the rest of the cytoplasm (2%), was detected in the extrasynaptic submembrane region.

Synaptic structural modification following changes in activity induced by tetanus neurotoxin in cat abducens neurons.

A low or a high dose of tetanus neurotoxin (TeNT) injected in the lateral rectus muscle of the cat causes respectively, functional block of inhibitory synapses only or of both inhibitory and excitatory synapses simultaneously in abducens neurons (González-Forero et al. [2003] J. Neurophysiol. 89:1878-1890). As a consequence, neuronal firing activity increases (at low dose) or decreases (at high dose). We investigated possible structural modifications of inhibitory synapses in response to these activity alterations induced by TeNT. We used immunofluorescence against postsynaptic (gephyrin) and presynaptic (vesicular gamma-aminobutyric acid [GABA] transporter [VGAT]) markers of inhibitory synapses in combination with cell type markers for abducens motoneurons (calcitonin gene-related peptide or choline acetyltransferase) or internuclear neurons (calretinin). Seven days after high-dose treatment, the number of gephyrin-immunoreactive (IR) clusters per 100 microm of membrane perimeter was reduced on the soma of abducens motoneurons by 55.3% and by 60.1% on internuclear neurons. Proximal dendritic gephyrin-IR clusters were also significantly altered but to a lesser degree. Partial synaptic re-establishment was observed 15 days post injection, and complete recovery occurred after 42 days. Coverage by VGAT-IR terminals was reduced in parallel. In contrast, a low dose of TeNT caused no structural alterations. With electron microscopy we estimated that overall synaptic coverage was reduced by 40% in both types of neurons after a high dose of TeNT. However, F-type terminals with postsynaptic gephyrin were preferentially lost. Thus, the ratio between F and S terminals diminished from 1.28 to 0.39 on motoneurons and from 1.26 to 0.47 on internuclear neurons. These results suggest that the maintenance of proximal inhibitory synaptic organization on central neurons is best related to neuronal activity and not to the level of inhibitory synaptic function, which was equally blocked by the high or low dose of TeNT.

Heterogeneous synaptic covering and differential charge transfer sensitivity among the dendrites of a reconstructed abducens motor neurone: correlations between electron microscopic and computer simulation data.

Ultrastructural studies on the synaptology of dendritic arborizations of motoneurones have been problematic because dendrites are very thin in relation to their great length, and most of the studies on this topic have therefore dealt with only small parts of the dendritic tree. Here we compared the ultrastructural characteristics of the axon terminals distributed along the various dendrites of a single motoneurone. For this purpose, the light microscopic 3D reconstruction of the dendritic arborization of an intracellularly labelled abducens motoneurone was combined with an electron microscopic analysis of its synaptic contacts. Dendritic profiles were randomly sampled along the various dendrites and the axon terminals they received were classified on the basis of their ultrastructural features and their GABA-immunoreactivity. It emerged that the various dendrites differed according to the type and local arrangement of their synaptic inputs. Our second aim was to incorporate the morphological data obtained into a model giving the charge transfer effectiveness T(x) of the dendritic sites. The sensitivity S(x) of T(x) to changes in the membrane resistivity (Rm) simulating various levels of tonic synaptic activity was calculated. It turned out that both the proximal and distal regions of the dendritic arborization have a dense synaptic covering and a weak sensitivity to changes in the Rm, whereas the intermediate dendrites have a sparse synaptic covering and a high sensitivity to changes in tonic synaptic activity. This pattern of organisation might mediate the "gating" of a population of synapses covering some dendritic regions in a state-dependent fashion.

Grafting of a new target prevents synapse loss in abducens internuclear neurons induced by axotomy.

The loss of afferent synaptic boutons is a prominent alteration induced by axotomy on adult central neurons. In this work we attempted to prove whether synapse loss could be reverted by reconnection with a new target. We severed the medial longitudinal fascicle of adult cats and then transplanted embryonic cerebellar primordia at the lesion site immediately after lesion. As previously shown, the transected axons from abducens internuclear neurons penetrate and reinnervate the graft [J Comp Neurol 444 (2002) 324]. By immunocytochemistry and electron microscopy we studied the synaptology of abducens internuclear neurons under three conditions: control, axotomy and transplant (2 months of survival time). Semithin sections of the abducens nucleus were immunostained against calretinin, to identify abducens internuclear neurons, and either synaptophysin (SF), to label synaptic terminals, or glial fibrillary acidic protein (GFAP) to detect the astrocytic reaction. Optical and linear density of SF and GFAP immunostaining were measured. Data revealed a significant decrease in the density of SF-labeled terminals with a parallel increase in GFAP-immunoreactive elements after axotomy. On the contrary, in the transplant group, the density of SF-labeled terminals was found similar to control, and the astrocytic reaction induced by lesion was significantly reduced. At the ultrastructural level, synaptic coverage and linear density of boutons were measured around the somata of abducens internuclear neurons. Whereas a significant reduction in both parameters was found after axotomy, cells of the transplant group received a normal density of synaptic endings. The ratio between F- and S-type boutons was found similar in the three groups. Therefore, these findings indicate that the grafting of a new target can prevent the loss of afferent synaptic boutons produced by the axotomy.

[Experimental anatomical data on the structure of retrobulbar nervous plexus in man and various laboratory animals]. / Anatomo-éksperimental'nye dannye o stroenii retrobul'barnogo nervnogo spleteniia u cheloveka i nekotorykh laboratornykh zhivotnykh.

Retrobulbar nervous plexuses were studied in man, albino rat, guinea pig and rabbit by means of macro- and micropreparation with elective dyes use, histotopogrammes and in the experiment with sympathetic superior cervical node removal. It was established that models characteristic for mammals were present in this plexus structure. Branches of abducent nerve conducting sympathetic fibres to the eyeball are the part of the plexus in rodents and the rabbit. In all the cases sympathetic fibres to the eyeball pass without contact with ciliary ganglion, apart from guinea pig, in which portion of sympathetic fibres passes through additional ciliary ganglia. Only in man ciliary ganglion has sensory root although sympathetic fibres are not discovered in trigeminal nerve branches.

Interspecies variation in axon-myelin relationships.

The primary objective of this paper was to determine the extent and nature of interspecies differences in axon calibre and myelin sheath thickness and in the various relationships between these. Morphometric analysis of the axon perimeter-myelin sheath thickness relationship was performed on an equivalent nerve fibre population in a mammal, the rat, a bird, the chicken, an amphibian, the frog, a bony fish, the trout, and a cartilaginous fish, the dogfish. The abducent nerve was studied. It is especially suitable for this purpose because its fibres are closely similar in type and in peripheral distribution across the species studied. The relationship differed substantially between species. Differences were present in its setting, as described by the positions of the scatterplots, in the g ratio and in the regression and correlation data relating the parameters. Both parameters were markedly larger in the fish species than in all of the others. In addition, in rat, chicken, frog and trout, where large and small fibre classes could be differentiated clearly, the setting of the relationship between the two parameters was different for the two classes. In the main, variation in each of the parameters was greater between than within species. The larger fibres in the fish species were closely similar in axon perimeter and sheath thickness despite their long evolutionary separation. From this study and from others in the series, it may be concluded that there is no fixed or constant relationship between axon calibre and the thickness of the surrounding myelin sheath. Each nerve tends to have its own particular relationship and this differs between species.

Abducens internuclear and ascending tract of deiters inputs to medial rectus motoneurons in the cat oculomotor nucleus: synaptic organization.

Abducens internuclear and ascending tract of Deiters (ATD) inputs to medial rectus motoneurons in the oculomotor nucleus are important for conjugate horizontal movements. In the present study, the organization of these separate populations of neurons and their synaptic connections with medial rectus motoneurons in the cat oculomotor nucleus have been examined by light and electron microscopy by using retrograde and anterograde axonal tracers. Consistent with the patterns of retrograde horseradish peroxidase labeling, the abducens internuclear projection is predominantly, if not exclusively, contralateral, whereas the ATD projection is exclusively ipsilateral, as demonstrated by anterograde autoradiographic and biocytin labeling. Both populations of synaptic endings contain spheroidal synaptic vesicles and establish synaptic contacts with modest postsynaptic densifications. In addition, ATD synaptic endings frequently are associated with subjunctional dense bodies and subsurface cisternae. The two populations of excitatory inputs differ, however, in their soma-dendritic distribution. The majority of abducens internuclear synaptic endings contact distal dendrites, whereas the majority of ATD synaptic endings contact proximal dendrites or somata. Abducens internuclear synaptic endings furthermore have a higher density of mitochondria than ATD synaptic endings. The more proximal location of ATD synaptic endings is consistent with the faster rise time and earlier reversal to polarizing currents of ATD excitatory postsynaptic potentials in comparison to those evoked by the abducens internuclear pathway as determined electrophysiologically. Given the differences in the physiologic signals conveyed by the abducens internuclear (eye velocity and eye position) and ATD (head velocity) pathways, the findings in this study suggest that the soma-dendritic stratification of the two inputs to medial rectus motoneurons may provide a means for the separate control of visuomotor and vestibular functions, respectively.

Morphometric nerve fiber analysis and aging process of the human abducent nerve.

Myelinated nerve fibers of the human abducent nerve were analyzed with a new staining method that permits simultaneous observation of the axon and surrounding myelin sheath. The following equipment was employed for the measurements: an image-analyzing digitizer, a microscope equipped with a drawing tube (or camera lucida), and a computer for data storage and statistical analysis. The numbers, transverse areas, and circularity ratios of axons were measured in 10 human abducent nerves. The average number was 1,997 with a definite decrease with age, and the average area was 3.90 micron 2.

Unmyelinated nerve fiber analysis of the human abducent nerve.

Unmyelinated nerve fibers of the abducent nerve have occasionally been observed with electron microscopes, but, to our knowledge, they have never been evaluated from the morphometric point of view. We analyzed the aging process of the unmyelinated nerve fibers in the human abducent nerve with the help of a new staining method suitable for morphometric research on the nervous system. We studied numbers and transverse areas of unmyelinated fibers of the abducent nerve in 10 cadavers. Our findings were that (1) these fibers were distributed diffusely, (2) their number decreased with age, and (3) the mean transverse area did not change with age. Most of the unmyelinated axons were thinner than the myelinated axons. These results may be important for analysis of clinical signs in relation to aging and ophthalmologic functions.

Effects of botulinum neurotoxin type A on abducens motoneurons in the cat: ultrastructural and synaptic alterations.

The synaptic alterations induced in abducens motoneurons by the injection of 3 ng/kg of botulinum neurotoxin type A into the lateral rectus muscle were studied using ultrastructural and electrophysiological techniques. Motoneurons identified by the retrograde transport of horseradish peroxidase showed a progressive synaptic stripping already noticeable by four days post-injection which increased over the study period. By 35 days post-injection, the normal coverage of motoneurons by synaptic boutons (66.4 +/- 4.0%) significantly decreased to 27.2 +/- 4.0%. Synaptic boutons detached by a widening of the subsynaptic space but remained apposed by synaptic contacts and desmosomes to the motoneuron. Detachment did not affect equally flat and round vesicle-containing boutons. The control motoneuron had almost equal numbers of both types of boutons, but after 35 days post-injection the ratio of round to flat vesicle-containing boutons was 1.20 +/- 0.01. Synaptic boutons impinging on motoneurons showed signs of alterations in membrane turnover, as indicated by an increase in the number of synaptic vesicles and a decrease in the number of coated vesicles and synaptic vesicles near the active zone. Abducens motoneurons had a transient increase in soma size by 15 days that returned to normal at 35 days, but no signs of chromatolysis or organelle degeneration were seen. Accompanying the swelling of motoneurons, a 15-fold increase in the number of spines, very infrequent in controls, was observed. Spines located in the soma and proximal dendritic trunk received synaptic contacts from both flat and round vesicle-containing boutons that could be either partly detached or completely attached to the motoneuron. An increased turnover of the plasmatic membrane of the motoneuron was observed, as indicated by a four-fold increase in the number of somatic coated vesicles. Animals were implanted with bipolar electrodes in the ampulla of both horizontal semicircular canals for evoking contralateral excitatory and ipsilateral inhibitory postsynaptic potentials. Motoneurons were antidromically identified from the lateral rectus muscle. Synaptic potentials of vestibular origin were recorded in abducens motoneurons. In the period between two and six days post-injection, a complete abolition of inhibitory synaptic potentials was observed. By contrast, excitatory synaptic potentials remained, but were reduced by 82%. The imbalance between excitatory and inhibitory inputs to motoneurons induced a progressive increase of firing frequency within a few stimuli applied to the contralateral canal. Between 7 and 15 days post-injection, both excitatory and inhibitory postsynaptic potentials were virtually abolished and remained so up to the longest time checked (105 days). Some motoneurons recorded beyond 60 days post-injection showed signs of recovery of excitatory postsynaptic potentials. During the whole time-span studied, presynaptic wavelets were present, indicating no affecting of the conduction of afferent volleys to the abducens nucleus. Taken together, these data indicate that botulinum neurotoxin at high doses causes profound synaptic alterations in motoneurons responsible for the effects seen in the behavior of motoneurons recorded in alert animals.

Electron microscopic serial analysis of GABA presynaptic terminals on the axon hillock and initial segment of labeled abducens motoneurons in the rat.

The aim of the present study was to provide a quantitative analysis of the synapses made onto the axon hillock and initial segment of rat abducens motoneurons retrogradely or intracellularly stained with HRP. GABA-immunoreactive terminals contacting these axons were visualized using a postembedding procedure. The presynaptic terminals contained either spherical or pleomorphic vesicles. gamma-Aminobutyric acid (GABA)-immunoreactive axon terminals, which belonged to this last category, were distributed both onto axon hillocks and the proximal part of initial segments. The percentage of axonal membrane covered by synapses ranged from 44.1 to 68.2%. A quantitative analysis performed on a series of ultrathin sectioned terminals contacting the axon of an intracellularly labeled motoneuron revealed a significant correlation between the length of membrane apposition of the terminals and their perimeter or surface area, and also between the area of membrane apposition and terminal volume. GABA-immunoreactive terminals had a mean perimeter and volume that were larger than those of unlabeled axon terminals. The number of active zones was correlated with the area of apposition. Some hypotheses concerning the functional role of the GABAergic innervation of this particular part of the neuron are discussed.

Utricular input to cat extraocular motoneurons.

Intracellular recordings from 200 identified extraocular motoneurons in the bilateral III, IV and VI cranial nuclei were studied to determine the connectivities between the utricular nerve and the extraocular motoneurons in cats. Stimulating electrodes were placed within the left utricular nerve, while other branches of the vestibular nerve were removed. Monosynaptic and disynaptic connections between the utricular nerve and the ipsilateral abducens motoneurons and interneurons were recorded as described previously. Stimulation of the utricular nerve evoked longer latency depolarizing and hyperpolarizing potentials in contra- and ipsilateral medial rectus motoneurons, respectively. Depolarizing and hyperpolarizing potentials with longer latencies were also recorded in the ipsilateral inferior oblique and contralateral trochlear motoneurons. The short and longer latency circuits between the utricular nerve and extraocular motoneurons may play a role in stabilizing the retinal image during head tilt and horizontal linear acceleration.

A light and electron microscope study of rat abducens nucleus neurons projecting to the cerebellar flocculus.

Injection of horseradish peroxidase (HRP) into the cerebellar flocculus of the rat was employed to identify neurons in the abducens nucleus that project to the flocculus. The number, ultrastructural features and precise localisation of these neurons in the nucleus were examined. They were present bilaterally and represented about 7% of the total neuronal population of each nucleus. They were localised principally in the dorsomedial area of the cranial half of each nucleus and did not display the typical ultrastructural features of motoneurons. It is concluded that the localisation and ultrastructural characteristics of these HRP-positive neurons are useful for distinguishing them from other neuronal populations within the nucleus.

GABAergic innervation of rat abducens motoneurons retrogradely labelled with HRP: quantitative ultrastuctural analysis of cell bodies and proximal dendrites.

In this quantitative electron microscopic study we investigated the distribution of GABA axon terminals on rat abducens motoneurons by combining retrograde labelling of montoneurons with post-embedding immunodetection of GABA. We analysed the synapses on 13 cell bodies and 60 proximal dendritic profiles distributed along the entire rostro-caudal extent of the nucleus. For each of these two compartments, we analysed 1754 and 1176 axon terminals in contact with 6042 and 3299 microns of postsynaptic membrane. The axon terminals were classified as Sv-type (containing spherical vesicles) or Pv-type (containing pleomorphic vesicles). The GABAergic terminals contained pleomorphic vesicles and established mainly symmetrical synaptic contacts. Their apposition lengths were greater than those of unlabelled terminals. On cell bodies, the percentage of GABAergic synaptic covering varied from 2.5% to 14.1% and the synaptic frequency of GABAergic axon terminals varied from 0.6% to 8.9%. These two parameters were significantly correlated with the diameter of the motoneurons. The percentage of synaptic covering and synaptic frequency were smaller on dendrites of small motoneurons than on those of large ones. The proximal dendrites of small motoneurons had a lesser GABAergic innervation than large ones. The total synaptic covering and frequency were smaller on somata than on dendrites. However, the percentage of synaptic covering by GABA terminals was higher on cell bodies than on proximal dendrites.

Electrotonic clusters in the dendritic arborization of abducens motoneurons of the rat.

Following reconstruction with high spatial resolution of the 3-D geometry of the dendritic arborizations of two abducens motoneurons, we simulated the distribution of electronic voltage over the whole dendritic tree. Here, we demonstrate that the complex stochastic electronic structure of both motoneurons can be reduced to a statistically significant small set of well discriminated clusters. These clusters are formed by dendritic branches belonging to different dendrites of the neuron but with similar electronic properties. A cluster analysis was performed to estimate quantitatively the partition of the branches between the dendritic clusters. The contents of the clusters were analysed in relation to their stability under different values of specific membrane resistivity (Rm), to their remoteness from the soma and their location in 3-D space. The cluster analysis was executed in a 2-D parameter space in which each dendritic branch was described by the mean electrotonic voltage and gradient. The number of clusters was found to be four for each motoneuron when computations were made with Rm = 3 k omega.cm2. An analysis of the cluster composition under different Rm revealed that each cluster contained invariant and variant branches. Mapping the clusters upon the dendritic geometry of the arborizations allowed us to describe the cluster distribution in terms of the 3-D space domain, the 2-D path distance domain and the total surface area of the tree. As the cluster behaviour reflects both the geometry and the changes in the neuronal electrotonic structure, we conclude that cluster analysis provides a tool to handle the functional complexity of the arborizations without losing relevant information. In terms of synaptic activities, the stable dendritic branches in each cluster may process the synaptic inputs in a similar manner. The high percentage of stable branches indicates that geometry is a major factor of stability for the electrotonic clusters. Conversely, the variant branches introduce the conditions for mechanisms of functional postsynaptic plasticity.

Stochastic geometry and electronic architecture of dendritic arborization of brain stem motoneuron.

We describe how the stochastic geometry of dendritic arborization of a single identified motoneuron of the rat affects the local details of its electrotonic structure. After describing the 3D dendritic geometry at high spatial resolution, we simulate the distribution of voltage gradients along dendritic branches under steady-state and transient conditions. We show that local variations in diameters along branches and asymmetric branchings determine the non-monotonous features of the heterogeneous electrotonic structure. This is defined by the voltage decay expressed as a function of the somatofugal paths in physical distances (voltage gradient). The fan-shaped electrotonic structure demonstrates differences between branches which are preserved when simulations are computed from different values of specific membrane resistivity although the absolute value of their voltages is changed. At given distances from soma and over long paths, some branches display similar voltages resulting in their grouping which is also preserved when specific membrane resistivity is changed. However, the mutual relation between branches inside the group is respecified when different values of specific membrane resistivity are used in the simulations. We find that there are some invariant features of the electrotonic structure which are related to the geometry and not to the electrical parameters, while other features are changed by altering the electrical parameters. Under transient conditions, the somatofugal invasion of the dendritic tree by a somatic action potential shifts membrane potentials (above 10 mV) of dendritic paths for unequal distances from the soma during several milliseconds. Electrotonic reconfigurations and membrane shifts might be a mechanism for postsynaptic plasticity.

Role of GABA in the extraocular motor nuclei of the cat: a postembedding immunocytochemical study.

The GABAergic innervation of the extraocular motor nuclei in the cat was evaluated using postembedding immunocytochemical techniques. The characterization of GABA-immunoreactive terminals in the oculomotor nucleus was carried out at the light and electron microscopic levels. GABA-immunopositive puncta suggestive of boutons were abundant in semithin sections throughout the oculomotor nucleus, and were found in close apposition to somata and dendrites. Ultrathin sections revealed an extensive and dense distribution of GABA-immunoreactive synaptic endings that established contacts with the perikarya and proximal dendrites of motoneurons and were also abundant in the surrounding neuropil. GABAergic boutons were characterized by the presence of numerous mitochondria, pleiomorphic vesicles and multiple small symmetrical synaptic contacts. The trochlear nucleus exhibited the highest density of GABAergic terminations. In contrast, scarce GABA immunostaining was associated with the motoneurons and internuclear neurons of the abducens nucleus. In order to further elucidate the role of this neurotransmitter in the oculomotor system, retrograde tracing of horseradish peroxidase was used in combination with the GABA immunostaining. First, medial rectus motoneurons were identified following horseradish peroxidase injection into the corresponding muscle. This was carried out because of the peculiar afferent organization of medial rectus motoneurons that contrasts with the remaining extraocular motoneurons, especially their lack of direct vestibular inhibition. Semithin sections of the oculomotor nucleus containing retrogradely labeled medial rectus motoneurons and immunostained for GABA revealed numerous immunoreactive puncta in close apposition to horseradish peroxidase-labeled somata and in the surrounding neuropil. At the ultrastructural level, GABAergic terminals established synaptic contacts with the somata and proximal dendrites of medial rectus motoneurons. Their features and density were similar to those found in the remaining motoneuronal subgroups of the oculomotor nucleus. Second, oculomotor internuclear neurons were identified following the injection of horseradish peroxidase into the abducens nucleus to determine whether they could give rise to GABAergic terminations in the abducens nucleus. About 20% of the oculomotor internuclear neurons were doubly labeled by retrograde horseradish peroxidase and GABA immunostaining. A high percentage (80%) of the oculomotor internuclear neurons projecting to the abducens nucleus showed immunonegative perikarya. It was concluded that the oculomotor internuclear pathway to the abducens nucleus comprises both GABAergic and non-GABAergic neurons and, at least in part, the GABA input to the abducens nucleus originates from this source. It is suggested that this pathway might carry excitatory and inhibitory influences on abducens neurons arising bilaterally.

Quantitative ultrastructural study of cat abducens interneurons.

Abducens interneurons project to the medial rectus subdivision of the contralateral oculomotor nucleus and have a role in the mediation of horizontal conjugate gaze. Two types of interneurons have been reported based on shape (fusiform and circular) and differences in organelles. In this study, 41 abducens interneurons from three young adult cats were examined in the transmission electron microscope and were classified, on the basis of eccentricity (e), as circular (e less than 0.82; N = 18) or fusiform (e greater than or equal to 0.82; N = 23). The volume fraction of nucleus, nucleolus, mitochondria, rough endoplasmic reticulum, polyribosomes, and Golgi complex and the number of synaptic profiles per 100 microns of membrane were determined. Although there is a qualitative difference in these cell types, statistical analyses indicate no significant differences in organelle content, mean area, number of synaptic profiles per 100 microns of membrane, or diameter. While the differences in shape may possibly indicate different functional groups, this notion is not supported by the variables examined or by physiological studies.