Thin Myelinated Axons Research Articles

Immunohistochemistry and ultrastructural characteristics of nerve endings in the oral mucosa of rat

The sensory nerve endings of the rat tongue, cheek and palate were studied using immunohistochemical staining and transmission electron microscopy analysis. The specimens were fixed in modified Karnovsky solution and embedded in Spurr resin. Substance P, calcitonin gene-related peptide (CGRP)- and protein gene product 9.5 (PGP b9.5)-containing nerve fibers in the rat tongue, cheek and palate were examined by electronic microscopical analysis and immunohistochemical localization. These fibers run very close to the basal lamina of the epithelium and extend into the filliform and fungiform papillae. Numerous plexiform fibers immunoreactive for substance P, CGRP and PGP 9.5 were found in the connective tissue of mucosa. Electron microscopic observations showed clearly immunostained nerve fibers, which are located very close to the basal lamina of epithelial cells. Some electron-dense granules may be observed in the axoplasms of both substance P and CGRP immunoreactive fibers. Several lamellar corpuscles into the subepithelial connective tissue papillae, Merkel corpuscles and numerous thin unmyelinated and myelinated axons were observed. The terminal axons revealed numerous mitochondria, neurofilaments, microtubules and clear vesicles in the base of axoplasmic protrusions. The lamellar cells showed caveolae and interlamelar spaces filled by amorphous substance. Between the lamellar cells and axoplasmic membrane, and in the adjacent lamellae region, desmosome-type junctions were observed. The quantitative and morphometric analysis showed nerve endings with an average area of 4.83 ± 3.4 μm(2) and 19.4 internal mitochondria in this site and the organized corpuscles with an average area of 79.24 ± 27.24 μm(2) and 24.23 internal mitochondria in this place. All the structures observed are involved in the transmission of pain and mechanoreceptors stimulus of these oral mucosae.

Sex-related differences in recovery of cutaneous nociception after end-to-side nerve repair in the rat

Sex-related differences in the recovery of cutaneous nociception after end-to-side nerve repair were examined in rats. Recovery of nociception in the dorsal foot was determined by skin pinch test 19 weeks after the proximal end of the distal stump of the transected peroneal nerve was sutured to the side of the adjacent intact sural nerve (end-to-side nerve coaptation). Axon sprouts in the recipient peroneal nerve were counted by light and electron microscopy. Recovery of nociception due to axon sprouting through the end-to-side coaptation was found in 87% of females and in 60% of males. The area of nociception was not significantly different (P=0.59) between females and males (13+/-8% and 11+/-9%, respectively). The number of myelinated axons in the recipient peroneal nerve (but not of unmyelinated axons) was significantly larger (P=0.028) in females (median=512, 25th and 75th percentiles: 467 and 594) than in males (median=322, 25th and 75th percentiles: 239 and 468). The majority of these axons in females and males were thin fibres, and recipient nerves in both groups were responsive to nerve pinch test. In conclusion, collateral sprouting of thin myelinated nociceptive axons into the end-to-side coapted nerve is more abundant in female than in male rats. However, recovery of cutaneous mechano-nociception due to sprouting of these axons was not different between the two sexes. Possible reasons are discussed.

The expansion of 300 CTG repeats in myotonic dystrophy transgenic mice does not induce sensory or motor neuropathy

Although many studies have been carried out to verify the involvement of the peripheral nervous system (PNS) in dystrophia myotonica (DM1) patients, the results remain controversial. The generation of DM1 transgenic mice displaying the human DM1 phenotype provides a useful tool to investigate the type and incidence of structural abnormalities in the PNS. In the present study, the morphological and morphometric analysis of semi-thin sections of sciatic and sural nerves, lumbar dorsal root ganglia (DRG) and lumbar spinal cords revealed that in DM1 transgenic mice carrying 300 CTG repeats, there is no change in the number and diameter of myelinated axons compared to wild type. Only a non-significant reduction in the percentage of thin myelinated axons was detected in electron micrographs of ultra-thin sciatic nerve sections. Analysis of the number of neurons did not reveal a loss in number of either sensory neurons in the lumbar DRG or motor neurons in the lumbar spinal cord in these DM1 mice. Furthermore, in hind limb muscle sections, stained with a neurofilament antibody and alpha-bungarotoxin, the intramuscular axon arborization appeared normal in DM1 mice and undistinguishable from that in wild-type mice. Moreover, in DM1 mice, there was no irregularity in the structure or an increase in the endplate area. Also statistical analysis did not show an increase in endplate density or in the concentration of acetylcholine receptors. Altogether, these results suggest that 300 CTG repeats are not sufficient to induce axonopathy, demyelination or neuronopathies in this transgenic mouse model.

Evidence of Innervation in Talc-Induced Pleural Adhesions

To conduct a detailed morphologic and ultrastructural study of pleural adhesions following talc pleurodesis. Talc with a main particle size of 8.36 +/- 0.2 mum (mean +/- SEM) and at a dose of 200 mg/kg in a 2-mL slurry was instilled via a small catheter into the pleural cavity of 10 male rabbits. Five rabbits were killed at 1 week, and five rabbits were killed at 1 month after instillation. At autopsy, after macroscopically observing the pleural cavity, adhesions were excised from opposing pleural surfaces and processed for histopathologic, immunocytochemical, and ultrastructural study. At 1 week, all adhesions examined were mesothelium-covered fibrovascular bands containing well-developed blood and lymphatic vessels establishing a structural continuity between both pleural layers. Nerves were present in adhesions from 20% of the rabbits. They consisted of a single fascicle containing 5 to 20 thin myelinated axons of various diameters (1 to 6 microm) uniformly distributed throughout the nerve section. The anatomic location of the adhesion did not appear to influence its overall morphology. As early as at 1 week, adhesions are well-formed structures more resembling newly formed pleural tissue than a simple scar. Nerve fibers in pleural adhesions are reported for the first time, which suggests that these adhesions are potentially capable of conducting pain stimuli. Further studies are required in order to confirm our results in human pleural adhesions.

Dystrophin splice variants are distinctly localized in the hippocampus

It has previously been demonstrated that Dp71, the most abundant dystrophin protein in the brain, is mainly localized in the postsynaptic densities. Here we show the localization of Dp71f, one of the splice variants of this protein, within the CA3 region of the hippocampus. Immunopositivity occurs in the postsynaptic density of small asymmetrical axospinous and axodendritic synapses, while it is absent in the postsynaptic densities of the axospinous synapses of the large mossy fiber terminals. Dp71f immunoreactivity was found to be attached to the membranes of the mossy fibers in the stratum lucidum of the CA3 area. In a certain population of thin myelinated axons the protein seems to be present within the axon proper. These data support the notion of a physiological role of Dp71f distinct from other dystrophin isoforms present in the central nervous system.

Effects of trapidil after crush injury in peripheral nerve.

In this study, we evaluated the effects of trapidil on crush injury by monitoring nitric oxide, malondialdehyde and transforming growth factor-Beta2 levels and by transmission electron microscopy in the rat sciatic nerve. The sciatic nerve was compressed for 20 sec by using a jewelers forceps. Trapidil treatment groups were administrated a single dose of trapidil (8 mg/kg) intraperitoneally just after the injury. The crush and crush + trapidil treatment groups were evaluated on the 2nd, 7th, 15th, 30th and 45th days of the post-crush period. On the 7th and 15th days, damage in thin and thick myelinated axons, endoneural edema and mitochondrial swelling were less severe in the trapidil group histopathologically. These findings supported the idea that trapidil prevented cell damage and edema at the injury site. Day/group interaction with regard to serum nitric oxide, malondialdehyde and transforming growth factor-Beta2 levels did not show significant changes.

Sex-related difference in collateral sprouting of nociceptive axons after peripheral nerve injury in the rat

Possible sex-related differences in the extent of collateral sprouting of noninjured nociceptive axons after peripheral nerve injury were examined. In the first experiment, peroneal, tibial, and saphenous nerves were transected and ligated in female and male rats. Eight weeks after nerve injury, skin pinch tests revealed that the nociceptive area of the noninjured sural nerve in the instep skin expanded faster in females; the final result was a 30% larger increase in females than in males. In the second experiment, the end-to-side nerve anastomosis was used as a model for axon sprouting. In addition to the previous procedure, the end of an excised peroneal nerve segment was sutured to the side of the intact sural nerve. Eight weeks later, collateral sprouting of nociceptive axons into the anastomosed peroneal nerve segment was assessed by the nerve pinch test and axon counting. There was no significant difference with respect to the percentages of male and female rats with a positive nerve pinch test. The number of myelinated axons in the anastomosed nerve segment was significantly larger in female (456 ± 217) than in male (202 ± 150) rats, but the numbers of unmyelinated axons were not significantly different. In normal sural nerves, the numbers of either all myelinated axons or thin myelinated axons did not significantly differ between the two sexes. Therefore, the more extensive collateral axon sprouting observed in female than in male rats is probably due to the higher sprouting capacity of thin myelinated sensory axons in females.

Nerve fibre composition of the palmar cutaneous branch of the median nerve and clinical implications

Fifteen fresh human cadaver hands were dissected, using ×2.8 loupe magnification, to study the subcutaneous innervation at the site of the incision (in the line with the radial border of the ring finger) for standard open carpal tunnel decompression. Subcutaneous nerve branches were detected and traced proximally to determine their origin. Morphometric analysis of nerve cross sections from the site of the incision and from the main nerve trunk proximal to cutaneous arborisation was performed using light and transmission electron microscopy and a computer-based image analysis system. At the site of the incision, the ulnar sub-branch (US) of the palmar cutaneous branch of the median nerve (PCBMN), which innervates the skin over the hypothenar eminence, was found in 10 of 15 cases. Branches from the ulnar side were not detected. The main trunk of PCBMN consisted on average of 1000 (SD 229) myelinated axons arranged in 1–4 fascicles. In the US of the PCBMN there were on average 620 (SD 220) myelinated axons, 80% of them smaller than 40 μm 2 i.e. thin myelinated axons, and on average 2037 (SD 1106) unmyelinated axons, arranged in 1–3 fascicles. The ratio of the number of myelinated axons in the US and the main trunk of the PCBMN was on average 63% (SD 19%). Frequency distribution of cross-sectional areas of myelinated axons shows no significant difference between the US and the main nerve trunk of the PCBMN. The importance of incision trauma to subcutaneous innervation of palmar triangle is emphasised and possible mechanisms of scar discomfort are discussed.

Sensory nerve endings in the anterior cruciate ligament (Lig. cruciatum anterius) of sheep

This study examines the structure of sensory nerve endings in the sheep anterior cruciate ligament (ACL). Three types of nerve endings are found: free nerve endings (FNE), Ruffini corpuscles, and lamellated corpuscles. The FNE (more than 100) are found subsynovially. The afferent nerve fibres are either thin myelinated axons (Adelta) or C fibres with diameters of 1-2 microm. FNE have been reported to function as thermoreceptors and polymodal nociceptors. In addition, FNE are also seen between fascicles of collagen fibres, often close to blood vessels. Part of this group may be efferent autonomic fibres controlling local blood flow. The corpuscles are seen subsynovially and between fascicles of connective tissue close to the attachment points of the ACL. A ligament contains about 20 Ruffini corpuscles, which are mainly located in the subsynovial connective tissue. They consist of cylinders formed from perineural cells surrounding the afferent myelinated axons (diameters 4-5 microm) with enlarged nerve terminals anchored between collagen fibres. These enter in bundles from the surrounding connective tissue at one open pole, pass through the length of the cylinder, and leave at the other pole. Functionally, Ruffini corpuscles have been described as slowly adapting stretch receptors. Lamellated corpuscles (usually between 5 and 15) are found in the subsynovial connective tissue. The afferent myelinated axon has a diameter of 4-6 microm, and the nerve terminal is located in the centre of numerous layers formed by lamellated terminal glial cells and by a perineural capsule. They are known to function as rapidly adapting pressure receptors. The most important function of the ACL is its mechanical function, but additional sensory functions must be considered triggering reflex mechanisms in case of extreme positioning or overload.

Glycerol gangliotomy of the second dorsal cervical root in rats: an experimental study to evaluate a minimal invasive approach for the treatment of the chronic cervicogenic headache.

Glycerol is a known agent in the therapy of chronic tic douloureux. It has been used for about 20 years in percutaneous, retrogasserian minimal-invasive rhizotomy, although the pharmacological mechanism of the pain relief involved remains unclear. To investigate glycerol treatment as a possible replacement for invasive approaches in the therapy of chronic cervicogenic headaches, we performed an experimental study on the pathomorphologic action of anhydrous glycerol injection into the second upper cervical dorsal root ganglion (DRG) of rats. Glycerol injections into the second cervical ganglion were investigated light- and electron-microscopically in a series of 40 rats for survival times of up to 30 days. We detected an unspecific overall effect on sensory neurons and satellite cells, as well as on myelinated and unmyelinated axons and Schwann cells. This could be detected after 5 days and sometimes led to degeneration of most of the neurons. Contralateral saline injections as a control showed no morphological effects. The loss of afferent fiber connections to the posterior horn of the myelon could be detected by immunohistochemical labeling of reactive astrocytes. Our results show a glycerol-induced deterioration of the cytoarchitecture of the neurons and their glial satellite cells. The effects on the ganglion cells appear to have been mediated by membrane disturbances and loss of glial integrity. These observations are contrary to previously reported results indicating the specific effect of glycerol on thin myelinated sensory axons.

Immunocytochemical distribution of Ca 2+-independent protein kinase C subtypes ( δ, ϵ, and ζ) in regenerating axonal growth cones of rat peripheral nerve

In the peripheral nerve, regenerating axonal sprouts usually emanate at nodes of Ranvier, and extend as growth cones along the inner surface of Schwann cells and/or through Schwann cell columns in the distal nerve segment. In order to elucidate the significance of Ca 2+-independent protein kinase C in nerve regeneration, localizations of δ, ϵ and ζ subtypes were examined immunocytochemically in sprouts and growth cones of regenerating axons, as well as in normal intact nerves in the rat sciatic nerve. In normal nerves, intense immunoreactivities of δ, ϵ and ζ subtypes were present in axons of both myelinated and unmyelinated fibres. Subcellularly, the distribution of these subtypes in the axoplasm was patchy, and discontinuous in the axolemma and subaxolemmal peripheral zones of myelinated nerves. Some thin myelinated axons showed no immunoreactivity for ϵ subtype. Schwann cells of both myelinated and unmyelinated fibres had moderate immunoreactivities for each subtype. In areas of nerve regeneration, axonal sprouts at nodes of Ranvier, and growth cones extending along Schwann cell basal laminae, had intense immunoreactivities for δ, ϵ and ζ subtypes which are distributed diffusely throughout the axoplasm, and on the entire axolemma. In the sprouts, immunoreactivity for ϵ subtype was strong on the axolemma, but weak or almost absent in the axoplasm. These data, together with those of our previous study, indicate that Ca 2+-independent protein kinase C subtypes ( δ, ϵ and ζ) have basically the same distribution patterns as those of Ca 2+-dependent subtypes in sprouts and growth cones of regenerating axons, as well as in normal intact axons; albeit ϵ subtype is somewhat different in distribution and intensity from δ and ζ subtypes. It is suggested that Ca 2+-independent subtypes are involved in maintaining growth cone activities along with the Ca 2+-dependent subtypes.

GANGLIONIC AXONS IN MOTOR ROOTS AND PIA MATER

In addition to motor axons and preganglionic axons, ventral roots contain unmyelinated or thin myelinated sensory axons and postganglionic sympathetic axons. It has been said that ventral roots channel sensory axons to the CNS. However, it now seems that these axons end blindly, shift to the pia or loop and return towards the periphery and that these units reach the CNS via dorsal roots. Sensory ventral root axons project from a variety of somatic or visceral receptors; some of them are third branches of dorsal root afferents and some seem to lack a CNS projection. Many ventral root afferents contain substance P (SP) and/or calcitonin gene-related peptide (CGRP). These fibres are not affected by neonatal capsaicin treatment and they cannot induce radicular or pial extravasation. Some thin ventral root axons are sympathetic and relate to blood vessels. Afferents containing SP and/or CGRP and sympathetic axons also occur in the spinal pia mater. The sensory axons mediate pain. They might also have vasomotor, tissue-regulatory and/or mechanoreceptive functions. The motor roots of cranial nerves IV, VI and XI contain unmyelinated axons arranged like in ventral roots outside the autonomic outflow. However, the motor root of cranial nerve V channels some unmyelinated axons into the CNS. The occurrence of thin axons in ventral roots and pia mater changes during development and ageing. After peripheral nerve injury, ipsilateral ventral roots and pia are invaded by new sensory and postganglionic sympathetic axons. © 1997 Elsevier Science Ltd. All Rights Reserved.

Structural basis of neuron-to-neuron cross-excitation in dorsal root ganglia.

Lanthanum was used as a tracer substance to determine whether small molecules in the bulk extracellular space in dorsal root ganglia have access to the narrow cleft that separates sensory neurons from their surrounding satellite cell sheath. Results showed that lanthanum is able to diffuse into this cleft, especially when the tissue is incubated with the tracer before fixation. Lanthanum gained access to the cleft at the seam where adjacent satellite cell processes meet. There appears to be preferential access in the axon hillock-initial segment region. Large diameter light neurons, which generally support fast conducting myelinated axons and carry information about non-nociceptive sensory events, proved more likely to admit lanthanum than small diameter dark neurons, which tend to have thin myelinated and unmyelinated axons and typically carry nociceptive information. Peripheral axotomy triggered a reduction in the access of lanthanum to the neuron-satellite cell cleft. These data bear on the mechanism of non-synaptic cell-to-cell cross-excitation within dorsal root ganglia, and in particular, lend support to the hypothesis that this interaction is mediated chemically rather than electrically.

Amino acid immunoreactivity in corticospinal terminals.

Anterogradely labeled corticospinal axons and their terminals were identified after injections of wheat germ-agglutinin conjugated to horseradish peroxidase in the sensorimotor cortex of rats. Thin myelinated axons were labeled in the corticospinal tract. Terminal labeling was densest in the internal basilar nucleus and laminae III-IV of the dorsal horn throughout the spinal cord; electron microscopical observations were mainly from the cervical enlargement. Labeled terminals were most often small and dome-shaped with densely packed, clear round vesicles and sparse mitochondria. These terminals established asymmetric synapses with small dendrites or spines and were never involved in axoaxonic contacts. Postembedding immunocytochemistry was used to study the subcellular distribution of glutamate, aspartate, and gamma-aminobutyric acid (GABA). Corticospinal terminals appeared enriched in glutamate, but not GABA. Some corticospinal terminals appeared enriched in aspartate, though the labeling was less selective than in the case of glutamate. GABA immunolabeling was very dense in about 20% of terminals. These were most often small, rich in mitochondria, and made symmetric synapses; they were not anterogradely labeled from the cortex. Quantitative analysis on double immunolabeled material allowed a direct comparison of particle density for different antigens in the same section. Terminals with a high density of particles coding for glutamate were not enriched with GABA, and terminals immunolabeled for GABA were not enriched with glutamate. There was no significant correlation between glutamate and aspartate immunolabeling in corticospinal terminals; a subpopulation of these terminals may be enriched in aspartate. Aspartate immunolabeling was consistently higher in dendrites postsynaptic (in the plane of section) to corticospinal terminals than in other dendrites.(ABSTRACT TRUNCATED AT 250 WORDS)

Sensory nerve endings in the beak skin of Japanese quail.

This study is concerned with the distribution and ultrastructure of sensory nerve endings in the beak skin of adult Japanese quail (Coturnix coturnix japonica). The following nerve endings were found: free nerve endings, clusters of dermal Merkel nerve endings, Herbst corpuscles and Ruffini corpuscles. The latter were found only in the dermis of the tip of the upper beak. The remaining endings were present in the skin of all areas of upper and lower beak. Free nerve endings were supplied by either thin myelinated axons or unmyelinated C-fibers and were localized in the dermis close to the basal layer of the epidermis. Merkel cells formed clusters (up to 50) localized below and between the epidermal cones of the beak skin. Disc-shaped thickenings of nerve endings were squeezed between individual Merkel cells. Small Herbst corpuscles were found in the dermis close to the epidermal cones of the beak skin. Large Herbst corpuscles occurred in deep layers of the dermis. The Ruffini corpuscles were cylindrical in shape (80 microns x 400 microns) and arranged in groups of up to ten corpuscles. Each corpuscle was surrounded by an incomplete fibrous capsule.

Nodes of Ranvier and myelin sheath dimensions along exceptionally thin myelinated vertebrate PNS axons.

The trigeminal alveolar branch in the lower jaw of the cichlid Tilapia mariae was examined by light and electron microscopy on single and serial sections, and by light microscopy on teased fibre preparations. The principal purpose was to find out if the exceptionally thin myelinated axons (d < 1 micron) present in this nerve possess true nodes of Ranvier, and to determine the dimensions of their myelin sheaths. This necessitated analysis of the whole size range of myelinated fibres, with respect to nodal and internodal morphology. The results show that the exceptionally thin myelinated fibres exhibit primitive nodal regions, with patches of axolemmal undercoating, and few Schwann cell processes in the node gap. This contrasts with the more complex nodal organization seen in larger trigeminal alveolar branch fibres. For the whole population of myelinated fibres the number of myelin lamellae increases rectilinearly with axon diameter, and sheath length increases with fibre diameter according to a logarithmic expression. The myelin sheaths of the exceptionally thin trigeminal alveolar branch fibres are composed of 10-20 lamellae, and extend 35-50 microns along the axon. These results show that the structural complexity of nodal regions in the trigeminal alveolar branch decreases with decreasing fibre size, that the exceptionally thin myelinated trigeminal alveolar branch fibres possess primitive nodes and that they have very short myelin sheaths. Our crude theoretical calculations suggest that these fibres might be capable of saltatory conduction.

Viscero-sympathetic reflex responses to mechanical stimulation of pelvic viscera in the cat

Viscero-sympathetic reflex responses to mechanical stimulation of urinary bladder and colon were studied in cutaneous vasoconstrictor (CVC) neurones supplying hairy skin, in muscle vasoconstrictor (MVC) neurones supplying skeletal muscle and in sudomotor (SM) neurones supplying the sweat glands of the central paw pad of the cat hindlimb. The cats were anaesthetized, paralysed and artificially ventilated. The vasoconstrictor activity was recorded from the axons of the postganglionic fibres that were isolated in filaments from the respective peripheral hindlimb nerves. The activity in the sudomotor neurones was monitored by recording the fast skin potential changes occurring on the surface of the central paw pad. Afferents from the urinary bladder and from the colon were stimulated by isotonic distension and isovolumetric contraction of the organs. Most CVC neurones with ongoing activity were inhibited by these stimuli; only a few CVC neurones were excited. The MVC and SM neurones were generally excited by the visceral stimuli, yet the size of the evoked skin potential changes was variable. The reflex responses elicited in the sympathetic outflow to the cat hindlimb by stimulation of visceral afferents from the pelvic organs are uniform with respect to the different types of afferent input system but differentiated with respect to the efferent output systems. Graded stimulation of the visceral afferents from the urinary bladder by isotonic pressure steps elicited graded reflex responses in CVC (threshold < 30 mmHg) and MVC neurones (threshold < 20 mmHg) and a graded increase of the arterial blood pressure (threshold < 20 mmHg). These graded reflex responses are closely related to the quantitative activation of sacral afferent neurones with thin myelinated axons innervating the urinary bladder that are also responsible for eliciting the micturition reflex, but not to the quantitative activation of sacral afferent neurones with unmyelinated axons. The latter have thresholds of 40–50 mmHg intravesical pressure at which the size of the vesico-sympathetic reflexes in the vasoconstrictor neurones was about 50% of maximal size. This does not exclude the fact that activation of unmyelinated vesical afferents contributes to the vesico-sympathetic reflexes.

Synaptology of the command (pacemaker) nucleus in the brain of the weakly electric fish, Sternarchus (Apteronotus) albifrons

The high frequency electric emission of the weakly electric fish Sternarchus (Apteronotus) albifrons depends on the pacemaker activity of a specific brainstem nucleus located in the ventral part of the rhombencephalic reticular formation. The general morphology and fine structure of this nucleus has been investigated, with particular reference to its synaptic connections. Three neuronal components could be distinguished in the nucleus; namely large cells of 80–100 μm diameter, small cells of 30–50 μm diameter and bundles of thin, myelinated fibres. These elements are embedded in a network of thick myelinated fibres. The large cells have a few small and short dendrites whereas the small neurons have long branching dendrites. Large and small neurons possess thick myelinated axons, but only those of the latter show branching patterns and send collaterals which have intranuclear courses only. Two types of synaptic terminals have been found on both neurons: large club endings exclusively with gap junctions and small bouton-like terminals with polarized chemical synapses. Serial semi-thin and ultra-thin sections revealed that the large club endings belong to the pacemaker cells, whereas the small terminals are found in the thin myelinated axons of extranuclear origin. The findings indicate that the small neurons are connected 1) to each other and 2) to the large neurons, by way of their large myelinated axons. Both, small (pacemaker) as well as large (relay), neurons receive chemical synapses from myelinated fine fibers probably originating from higher encephalic centers. Thus, electric organ discharge rhythm can be modulated at the level of pacemaker as well as of the relay cells. No somatosomatic, dendrodendritic or dendrosomatic connections were found between large, small or large and small cells.

Intra-axonalcorpora amylacea in ventral and lateral horns of the spinal cord

Deposits similar to corpora amylacea were observed by electron microscopy and light microscopy within thin myelinated axons in the spinal gray matter in various diseases. By electron microscopy the deposits consisted of randomly interlacing short linear structures of various thickness which were different from tangles of neurofilaments. Sometimes they contained dense granules and electron-dense floccules. Though their incidence was relatively high in some cases of degenerative or metabolic disease of the central nervous system, they were not specific to any disease but seemed to be related to aging, indicating a peculiar aspect of chronic degeneration of the axons.

An electron microscopic study of the laminar pattern and mode of termination of afferent fibre pathways in the somatic sensory cortex of the cat.

In a series of experiments, the extrinsic afferent pathways to the somatic sensory areas have been selectively interrupted and the distribution and mode of termination of degenerating synaptic endings studied at intervals of 2 to 6 days. Degenerating commissural fibres terminate on spines attached to dendrites of medium and small size in all cortical layers, but the endings are concentrated in the deeper parts of layers I and III and in layer IV. Cortical association fibres passing from SII to SI end on spines of small and medium-sized dendrites mainly in the intermediate layers (III, IV and V) of the cortex, but a small number are invariably seen in the molecular layer. Some of the latter are probably derived from thin myelinated axons which spread radially from a lesion and run just beneath the pia mater. Degenerating thalamo-cortical axons terminate on spines and to a lesser extent on shafts of dendrites of small diameter, mainly in layer IV but with overlap into adjacent parts of layers III and V and with an additional small but consistent number ending in the molecular layer. Thus, the molecular layer and layer IV receive the terminations of all extrinsic afferents, while the relations of these to the other laminae and their mode of termination is, in each case, slightly different. One interpretation of the results is that all three sets of extrinsic afferents terminate in the middle portion of the apical dendritic tree of pyramidal neurons, and that thalamo-cortical fibres have additional terminals on stellate neurons. The functional implications of this arrangement are discussed.