Growth pattern of axons in the optic nerve of chick during myelogenesis

CNS myelin sheath is stochastically built by homotypic fusion of myelin membranes within the bounds of an oligodendrocyte process

Myelin - the multilayer membrane that envelops axons - is a facilitator of rapid nerve conduction. Oligodendrocytes form CNS myelin; the prevailing hypothesis being that they do it by extending a process that circumnavigates the axon. It is pertinent to ask how myelin is built because oligodendrocyte plasma membrane and myelin are compositionally different. To this end, we examined oligodendrocyte cultures and embryonic avian optic nerves by electron microscopy, immuno-electron microscopy and three-dimensional electron tomography. The results support three novel concepts. Myelin membranes are synthesized as tubules and packaged into "myelinophore organelles" in the oligodendrocyte perikaryon. Myelin membranes are matured in and transported by myelinophore organelles within an oligodendrocyte process. The myelin sheath is generated by myelin membrane fusion inside an oligodendrocyte process. These findings abrogate the dogma of myelin resulting from a wrapping motion of an oligodendrocyte process and open up new avenues in the quest for understanding myelination in health and disease.

Development of the visual system of the chick. II. Mechanisms of axonal guidance

The quest to understand axonal guidance mechanisms requires exact and multidisciplinary analyses of axon navigation. This review is the second part of an attempt to synthesise experimental data with theoretical models of the development of the topographic connection of the chick retina with the tectum. The first part included classic ideas from developmental biology and recent achievements on the molecular level in understanding cytodifferentiation and histogenesis [J. Mey, S. Thanos, Development of the visual system of the chick. (I) Cell differentiation and histogenesis, Brain Res. Rev. 32 (2000) 343-379]. The present part deals with the question of how millions of fibres exit from the eye, traverse over several millimetres and spread over the optic tectum to assemble a topographic map, whose precision accounts for the sensory performance of the visual system. The following topics gained special attention in this review. (i) A remarkable conceptual continuity between classic embryology and recent molecular biology has revealed that positional cellular specification precedes and determines the formation of the retinotectal map. (ii) Graded expression of asymmetric genes, transcriptional factors and receptors for signal transduction during early development seem to play a crucial role in determining the spatial identity of neurons within surface areas of retina and optic tectum. (iii) The chemoaffinity hypothesis constitutes the conceptual framework for development of the retinotopic organisation of the primary visual pathway. Studies of repulsive factors in vitro developed the original hypothesis from a theoretical postulate of chemoattraction to an empirically supported concept based on chemorepulsion. (iv) The independent but synchronous development of retina and optic tectum in topo-chronologically corresponding patterns ensures that ingrowing retinal axons encounter receptive target tissue at appropriate locations, and at the time when connections are due to be formed. (v) The growth cones of the retino-fugal axons seem to be guided both by local cues on glial endfeet and within the extracellular matrix. On the molecular level, the ephrins and their receptors have emerged as the most likely candidates for the material substrate of a topographic projection along the anterior-posterior axis of the optic tectum. Yet, since a number of alternative molecules have been proposed for the same function, it remains the challenge for the near future to define the proportional contribution of each one of the individual mechanisms proposed by matching theoretical predictions with the experimental evidence.

Identification of a proliferating marginal zone of retinal progenitors in postnatal chickens

In warm-blooded vertebrates it is generally accepted that after early stages of development new neurons are not added to the retina. Contrary to this belief, we show here that hatched chickens have a zone of proliferating cells at the peripheral margin of the retina, similar to that of fish and amphibians. We found that cells at the peripheral edge of the retina incorporated the thymidine analog BrdU and expressed the cell cycle regulator proliferating cell nuclear antigen (PCNA). Furthermore, cells in the ciliary epithelium and retinal margin coexpressed the homeodomain transcription factors Pax6 and Chx-10, similar to multipotent progenitors of embryonic retina. Expression of PCNA, Pax6, and Chx-10 in cells at the retinal margin was maintained in adult birds. Double-labeling studies showed that BrdU-labeled cells that were integrated into the retina expressed proteins found only in differentiated neurons. Increased rates of ocular growth, induced by visual deprivation, resulted in increased numbers of BrdU-labeled cells at the retinal margin. Unlike the progenitors in the retinal marginal zone of fish and amphibians, the progenitors of the chick retina do not increase their rate of proliferation in response to acute damage. Furthermore, insulin, insulin-like growth factor-I, and epidermal growth factor increased proliferation of progenitors at the retinal margin, while basic fibroblast growth factor had no effect. These results indicate that the avian retina has a marginal growth zone containing proliferating cells that share similarities with multipotent embryonic retinal progenitors and the retinal stem cells of cold-blooded vertebrates.

Development of macroglial cells in the embryonic chick optic nerve

Macroglia development in the embryonic chick optic nerve was immunohistochemically examined. The astrocytes with glial fibrillary acidic protein immunoreactivity were initially restricted to the retinal end of the optic nerve at stage 40, but had widely dispersed within the optic nerve in an out-side-in manner by stage 44. Oligodendrocytes with myelin basic protein immunoreactivity appeared at stage 38, and were widely distributed at stage 40. Electron microscopic observation confirmed the presence of intermediate filaments in the glial fibers and developing myelin at stages 40-41. The present results suggest that oligodendrocyte precursors undergo terminal differentiation slightly earlier than astrocyte precursors with respect to the expression of marker proteins.

Quantitative studies of the optic nerve fiber layer in the chicken retina

The optic nerve fiber layer (NFL) of the chicken retina was studied quantitatively and morphologically at 17 positions along seven radially arranged bands from the dorsal tip of the pecten oculi using electron microscopy. The number of nerve fibers was counted in areas 6 microm in width x the full thickness of the NFL. Myelinated nerve fibers in the NFL were also identified immunohistochemically using anti-myelin basic protein serum. The dorsal area (dorsal, dorso-temporal and dorso-nasal bands) in the retina had thin NFL and contained the largest number of nerve fibers, which were mainly thin and unmyelinated. The ventral area (ventral and ventro-temporal bands) had a thick NFL and contained a relatively small number of nerve fibers, many of which were myelinated. The nasal band had the thickest NFL and contained as many nerve fibers as the dorsal area, with the temporal band showing a high ratio of myelinated fibers. The band had a thick NFL and contained many nerve fibers with a relatively low ratio of myelinated fibers. The relationship between the number and composition of nerve fibers in the NFL to the chicken visual characteristics was discussed. Although the myelin in the chicken retina was loose type, the myelin-forming cells were similar in appearance to dense oligodendrocytes. retina, morphometry, myelinated fiber, nerve fiber layer.

Changes in fast axonal transport in sensory neurons during tadpole metamorphosis

Fast axonal transport of radiolabeled protein was examined in lumbar and tail dorsal root ganglion (DRG) neurons at progressive stages of bullfrog tadpole metamorphosis. Accumulation of [35S]methionine-labeled protein proximal to a lumbar peripheral nerve ligature (at a fixed distance from the DRG) increased as tadpoles advanced from premetamorphosis through prometamorphosis to metamorphic climax. The rate of increase was steeper when expressed as a percentage of protein synthesized in the neurons of origin than when expressed as a percentage of total DRG protein synthesis. Further, the increase was not secondary to a rise in protein synthesis. In contrast, fast axonal transport decreased in DRG neurons of the tail at the onset of metamorphic climax, when tail resorption is initiated. The stage-related increase in protein transport in lumbar nerves is due, at least in part, to an increased rate of transport. As determined from optically detected anterograde organelles in individual lumbar nerve axons, an approximate doubling of the fast transport rate occurred between the premetamorphic stage and metamorphic climax. In addition, the rates of organelle transport in lumbar axons of adult bullfrogs were significantly greater than in corresponding axons of tadpoles at metamorphic climax, further suggesting that organelle velocity is a developmentally regulated parameter of fast axonal transport.

Formation of the myelinated nerve fiber layer in the chicken retina

Oligodendrocytes in the ganglion cell layer, the myelinating cells in the chicken retina, were investigated morphologically and quantitatively. Oligodendroblasts divided in the inner retinal layer around the 14th day of incubation and differentiated into oligodendrocytes. The oligodendrocytes started sheathing an axon in the nerve fiber layer at the 14th day of incubation. The number of myelin lamellae increased rapidly during the first week after chicks had hatched. An immunological reaction of anti-myelin basic protein was observed on the myelin sheaths in the nerve fiber layer and on the oligodendrocytes in the ganglion cell layer. These results suggest that the oligodendrocytes form the myelinated nerve fiber layer of the chicken and that they act independently of the Müller cells during myelination.

Development of the supraoptic decussation in the chick (Gallus gallus)

The developing supraoptic decussation (SOD), a major interhemispheric tract in birds, has been implicated in both transfer of visual information and in the modulation of brain asymmetry. Moreover little is known of its morphology during development. We have examined the development of the chick SOD, which consists of three subregions; dorsal, ventral and subventral SOD. In the dorsal SOD the total number of fibres reach 968,000 on the 19th day of incubation (E-19), falling to 570,000 by the 8th day after hatching (P-8). In the ventral SOD, the number of fibres at E-19 reach 660,000, followed by a gradual reduction in their number to about 490,000 at P-22. In the subventral SOD the number of fibres estimated was 87,000 at E-15 falling to about 36,000 P-1. Compared with adult levels, there is, respectively, a drop in the number of fibres of 44%, 25% and 69% in the dorsal, ventral and subventral SOD during development. At E-19 in both the dorsal and ventral SOD there is qualitative evidence of axonal loss; disrupted axonal profiles, increased extracellular space and cells containing lysosomal cytoplasmic inclusions indicative of macrophages. Differences were also observed in the pattern of myelination, the dorsal, ventral and subventral SOD were shown to myelinate at different rates. Thus, in a single named tract, the SOD, there are major differences in the onset, rate and extent of fibre loss and myelogenesis within its three subregions. The functional implications of these differences are considered.

Evidence for self-absorption of terminals by developing axons of retinal ganglion cells in the chick

The appearance of membrane-bound degenerative organelles in chick optic nerve axons was studied at the electron microscopic level. A semiquantitative analysis revealed a sharp increase in the number of axons containing accumulations of such organelles during the second day after hatching. In dark-reared chicks this increase was retarded, suggesting the presence of a light-influenced event in the early post-hatch period.

Substructures of paired helical filaments from Alzheimer's disease neurofibrillary tangles

Presented studies reveal that each of the approximately 10 nm filaments forming the paired helical filaments (PHF) is made up of four protofilaments. Each of the protofilaments is a beaded structure, consisting of globules connected by longitudinal bars. A cross-view of PHF shows eight globules linked by transverse bars. The transverse bars are shorter than the longitudinal bars. Comparison between PHF and neurofilament protofilaments indicates structural differences between these profiles, i.e., the globules making the PHF protofilaments are larger and the longitudinal bars are longer than those in the normal neurofilaments. A three-dimensional diagram of PHF structure is presented.

Abnormal cell relationships in Jimpy mice: electron microscopic and immunocytochemical findings

The mutant mouse strain Jimpy is characterized by a deficiency of myelin formation throughout the C.N.S. The cause of this hypomyelination is unknown. Based on previous reports, astrocytes, axons and oligodendrocytes are all altered, but no single cell type can be unequivocally defined as the primary target. Jimpy and age-matched normal mice were investigated using thin sectioning, freeze-fracturing and immunocytochemistry. We examined optic nerves and cervical spinal cords of Jimpy to determine which cells were morphologically altered during the period which precedes the onset of myelination and which cellular alterations persisted during myelinogenesis. Abnormalities of astrocytes and axons were frequently observed in Jimpy not only during myelination but also in early postnatal development before mature oligodendrocytes were present. The early astrocytic changes included hyperplasia and alterations of both cytoplasm and plasma membrane. An unusually complex network of astrocytic processes divided the axons into very small groups. During myelination, astrocytic processes were found insinuated between the axons and myelin sheath and/or within the myelin lamellae. Immunocytochemical investigations also revealed a complex network of glial fibrillary acidic protein-positive processes in contact with the majority of the axons. At stages prior to myelination axonal alterations were detected. Most of the axons were not in close contact with one another and individual axons had an undulating and irregular course. In areas where axon separation by astrocytic processes occurred, axonal diameters were more variable than the homogeneously sized axons of the normal mice. Our immunocytochemical results at stages during myelination showed not only many myelin basic protein-positive processes around axons in Jimpy but also clearly immunostained myelin sheaths. This indicates that the myelinating glia present not only produce myelin basic protein but can also incorporate it into the myelin spiral. The presented results suggest that the mouse mutant Jimpy could be a model for disturbed cell interactions in the C.N.S. Therefore, the hypomyelination may not be attributed to a defect of a single cell but rather to a deficiency in both macroglial types and, perhaps, the axon as well.

Composition of the tectal and posterior commissures of the chick (Gallus domesticus)

In chick, the tectal and posterior commissures form a continuous band of axons lying in the dorsal aspect of the meso-diencephalon. In midline sagittal sections, two zones can be clearly defined. The rostral zone (RZ) contains about 290,000 axons of which 32% are myelinated. The caudal zone (CZ) contains about 911,000 fibers of which 12% are myelinated. The majority of unmyelinated fibers in RZ and CZ are between 0.15 and 0.40 micrometers in diameter. The majority of myelinated fibers in both RZ and CZ are less than 1 micrometer in diameter. It is likely that RZ corresponds to the posterior commissure while CZ corresponds to the tectal commissure.

Myelination of S1 dorsal root axons in the cat

Samples of S1 dorsal root nerve fibers from cats of different pre- and postnatal ages were examined electron microscopically with regard to axon caliber and number of myelin lamellae. Each root was examined at four different cross-sectional levels. Two levels were situation close to the spinal cord entrance on each side of the peripheral (PNS) and central nervous system (CNS) border. The third and fourth levels were located more distally. The first compact myelin lamella was observed in the CNS part of the root in a 47-day-old fetus. In the 53-day-old fetus the degree of myelination was the same in the CNS as distal in the PNS part of the root. Surprisingly, all axons appeared unmyelinated close to the PNS-CNS border and remained so for a further 10-day period. After this time lag, this part of the root became myelinated and showed a rapid increase in myelin sheath thickness. Calculations of axonal growth, mesaxonal length, and myelin volume indicated a maturation process that progressed discontinuously. Myelination did not proceed in a strict somatofugal direction, but was a regionally differentiated process. The maximal myelin production, expressed as the increase in myelin volume per Schwann cell, was found during the second to fourth postnatal months, i.e., very late in development.

Innervation of avian latissimus dorsi muscles and axonal outgrowth pattern in the posterior latissimus dorsi motor nerve during embryonic development

The distribution of the innervation to the anterior latissimus dorsi (ALD) and posterior latissimus dorsi (PLD) muscles of the chicken are described on the day of hatching and 6 weeks later using electron microscopy. In the ALD muscle, there are 5,000 muscle fibres and 374,000 endplates supplied by about 169 skeletomotor axons; in the PLD muscle, there are 12,000 focally innervated muscle fibers supplied by about 20 skeletomotor axons. On the cell surface of the muscle fibers the mean total subsynaptic area contacted by each motor axon is comparable in the ALD and PLD muscles. The growth pattern of the axons in the PLD motor nerve was described from the ninth day in ovo up to 6 weeks after hatching. The axons arrive in the PLD muscle in two successive waves: first, the large somatic axons which are already present before the ninth day in ovo and second, the small autonomic axons which continue to accumulate until hatching. The total number of somatic axons decreases from the ninth day until the hatching day when it reaches its definitive value. This decrease takes place during a period when the numbers of myofibers and of endplates dramatically increase, and it coincides with the axonal segregation by the Schwann cells. The myelination of the axons starts on the 15th day in ovo and is essentially complete upon hatching. Despite the decreasing number of somatic axons in the PLD nerve, the decrease in number of nerve endings per PLD endplate and the increasing number of PLD endplates per PLD muscle, it was found that between the 16th day in ovo and 6 weeks after hatching the mean number of axonal branches per PLD motor axon does not decrease.

Intraocular colchicine selectively destroys immature ganglion cells in chicken retina

Low doses of colchicine (greater than 0.5 microgram) injected into the eyes of young chickens irreversibly inhibit axonal transport of material labelled with [3H]fucose. This is due to almost total destruction of the retinal ganglion cells, while other retinal cell types, including the displaced amacrine cells found in the same retinal layer, survive the treatment. The neurotoxic effect of colchicine on chick retinal ganglion cells decreases after hatching, and it is suggested that the decrease in sensitivity may be related to myelination of the retinal ganglion cell axons.

Myelination of the mouse corpus callosum

Myelination has been studied in th corpus callosum of the mouse brain between birth and 240 days-of-age. Myelin sheaths were first seen at 11 days. The most rapid phase of myelination occurred between 14 and 45 days when 13.5% of axons were myelinated, but myelination continued at a reduced rate up to 240 days when 28% of axons were myelinated. The mean diameter of unmyelinated axons was more or less constant throughout the study with an overall mean diameter of 0.25 +/- 0.01 micron. Similarly myelinated axon diameter showed little variation with age with a mean diameter of 0.46 +/- 0.01 micron. This suggests that in the corpus callosum axons do not increase in size until they begin to myelinate.

Evidence against the smooth endoplasmic reticulum as a continuous channel for the retrograde axonal transport of horseradish peroxidase

The involvement of the axonal smooth endoplasmic reticulum as a channel for the retrograde axonal transport of horseradish peroxidase (HRP) has been tested by analysing serial sections of 52 HRP-positive organelles in chick optic nerves. The enzyme marker was injected in the posterior, contralateral optic tectum 10 h before fixation of young chicks. The two optic nerves, retinas and optic tecta were incubated for electron microscopic demonstration of HRP. Thin sections of the retinas and tecta and serial thin sections of the optic nerves were studied in some cases with the aid of a goniometer. Of the 52 organelles, 42% had a tubular shape, 46% were oval and 12% were multivesicular bodies. None of the organelles was found to have continuities with other membranous structures, including tubules or cisternae of the smooth endoplasmic reticulum. In 10 cases, the smooth endoplasmic reticulum was followed in serial sections over a length of up to 4 micrometer. In every case, the reticulum appeared to form a continuous system although some tubular extensions apparently ended blindly near other organelles. In neither the 10 series of serial sections nor in any other individual micrographs did any recognizable profile of the smooth endoplasmic reticulum contain HRP. Measurements of the thickness of the membranes of HRP-containing organelles, of the smooth endoplasmic reticulum and of plasmalemma were made, since these membranes have been distinguished on the basis of their thickness in other cells. The plasmalemma in the axons was about 20% thicker than that of the smooth endoplasmic reticulum, and about 9% thicker than that of HRP-labeled organelles. The membrane of the smooth endoplasmic reticulum and HRP-organelles could also be distinguished by this means. It is concluded that in chick retinal ganglion cell axons, HRP is not transported in a retrograde direction via a continuous channel of smooth endoplasmic reticulum.

Neuropathology of phenylacetate poisoning in rats: an experimental model of phenylketonuria

Results of this investigation indicate that the suckling rat treated with phenylacetate should be a useful new model for studying the pathogenesis of phenylketonuria and neuronal development. Both cerebellar and retinal neurons of postnatally treated rats are vulnerable to the adverse effects of phenylacetate. Morphological changes observed in the cerebellum, retina, and optic nerve of treated animals during the fourth to twenty-first days of life consist of regional reduction in the size of cerebellar vermis lobules IV, V, VIa, and IX, 35 to 40% reduction in thickness of the molecular layer, accumulation of cerebellar external granular cells and retinal neuroblastic cells, fewer parallel fibers in the cerebellar cortex, and fewer myelinated axons in the optic nerve.

A quantitative comparison between the ganglion cell populations and axonal outflows of the visual streak and periphery of the rabbit retina

A vertical density profile of the ganglion cells 2 mm temporal of the optic nerve head in the rabbit retina has been produced by counting somata in the cresyl-violet-stained, ganglion cell layer of a flat-mounted retina. Somata classified as ganglion cells were characterized by obvious Nissl staining in an extensive cytoplasm and typically had diameters greater than 9 micrometer. The accuracy of the profile, and thus of the classification criteria, has been substantiated by electron micrographic determination of the numbers of ganglion cell axons arising within local regions of known area on the same retina. This study indicates that Vaney and Hughes' estimate ('76) of 547,100 presumed ganglion cells in the rabbit retina should be changed to 373,500 ganglion cells. The latter value is within the statistical error of their optic nerve count of 394,000 fibers. The mean diameter of ganglion cells 6 mm from the visual streak in the inferior periphery (density: 550 cells/mm2) was 28% greater than that of cells on the peak of the streak (density: 5,400 cells/mm2), although the form of the ganglion cell diameter distribution did not change markedly with eccentricity. The increase in the mean size of ganglion cells in the periphery appeared to be approximately matched by an increase in the size of their axons. Larger axons became myelinated farther from the edge of the myelinated band than did smaller axons. Within the ganglion cell layer there was another population of cells which were quite distinct from the obvious neuroglia: Their nuclei were similar to those of the larger ganglion cells and many appeared to have Nissl granules within their limited cytoplasm. About half of this heterogeneous population was classified as "coronate cells," which were characterized by the partial nuclear encapsulation of their eccentric cytoplasm.