Selective decrease of small sensory neurons in lumbar dorsal root ganglia labeled with horseradish peroxidase after ND:YAG laser irradiation of the tibial nerve in the rat

Recent electrophysiological evidence indicates that Q-switched Nd:YAG laser irradiation might have selective effects on neural impulse transmission in small slow conducting sensory nerve fibers as compared to large diameter afferents. In an attempt to clarify the ultimate fate of sensory neurons after laser application to their peripheral axons, we have used horseradish peroxidase (HRP) as a cell marker to retrogradely label sensory neurons innervating the distal hindlimb in the rat. Pulsed Nd:YAG laser light was applied to the tibial nerve at pulse energies of 70 or 80 mJ/pulse for 5 min in experimental rats. Seven days later HRP was applied to the left (laser-treated) and to the contralateral (untreated) tibial nerve proximal to the site of laser irradiation. In control animals the numbers of HRP-labeled dorsal root ganglion cells were not significantly different between the right and the left side. In contrast, after previous laser irradiation labeling was always less on the laser-treated side (2183 +/- 513 cells, mean +/- SEM) as compared to the untreated side (3937 +/- 225). Analysis of the dimensions of labeled cells suggested that the reduction of labeled cells on the laser-treated side was mainly due to a deficit in small sensory neurons. Since the conduction velocity of nerve fibers is related to the size of their somata, our histological data imply that laser light selectively affects retrograde transport mechanisms for HRP in slow conducting sensory nerve fibers.     

Retrograde Horseradish Peroxidase Transport in Motor Axons after Nd:YAG Laser Irradiation of the Tibial Nerve in Rats

We have recently demonstrated that the number of small sensory neurons of the A-δ- and C-fiber group in lumbar dorsal root ganglia labeled with horseradish peroxidase (HRP) is selectively decreased 7 days after Nd:YAG laser irradiation of the tibial nerve in the rat. In contrast, the number of large diameter sensory neurons was not affected by laser application. In an attempt to clarify the fate of motoneurons after laser irradiation of their peripheral axons, the numbers of lumbar motoneurons retrogradely labeled with HRP 7 days after Nd:YAG laser irradiation of the tibial nerve have been determined in rats. Our results show that the number of HRP-labeled motoneurons in lumbar segments L6 to L3 is not altered to a significant extent after laser irradiation of their peripheral axons (laser-treated side, 767 ± 10 cells vs control side, 808 ± 19; n = 5, mean ± SEM). In addition, no difference was detected in the mean value or the distribution of soma cross-sectional areas of labeled motoneurons on the laser-treated side and the control side. Specifically, the numbers of HRP-labeled small diameter motoneurons, which are presumably γ in type and have a conduction velocity similar to sensory neurons of the A-δ group, were not affected by laser application. Possible mechanisms of the differential vulnerability of sensory neurons as compared to motoneurons of similar size are discussed.     

Golgi-like immunoperoxidase staining of dopamine neurons in the reticular formation of the rat brainstem using antibody to tyrosine-hydroxylase

Cell bodies of sympathetic and sensory axons projecting via the superficial peroneal (SP) nerve supplying hairy skin of the distal hindlimb have been labeled retrogradely with horseradish peroxidase (HRP) on both sides of three cats in which the left SP nerve had been cut and ligated about 5 months previously. Three SP nerves from unoperated cats have also been studied. The location, size, and numbers of labeled somata have been determined from serial sections of lumbosacral dorsal root and sympathetic chain ganglia after standard histochemical processing. The numbers of myelinated fibers in each nerve have also been counted. The segmental distributions of both sympathetic and sensory cell bodies were identical bilaterally in each operated animal, but the number of labeled neurons was reduced on the lesioned side. There were only about 31% of sympathetic and about 51% of sensory somata relative to the numbers on the contralateral side. The average total number of neurons labeled from SP nerves in unoperated animals was about 8% higher than on the control side of operated animals. The average number of myelinated fibers in the neuromatized nerves was not reduced with respect to that in the contralateral nerve and both of these were not significantly different from the number in unoperated animals. The dimensions of samples of labeled sympathetic and sensory somata were reduced on the side with the neuroma, both in comparison with the contralateral side and with unlabeled neurons at the same levels. The mean cross-sectional area of profiles of sympathetic ganglion cells was 76% of the control; of sensory ganglion cells, 65% of the control. Assuming that HRP labeling was not impaired, we conclude that large numbers of neurons with unmyelinated axons had degenerated in the neuromatized cutaneous nerves.     

A morphological study of the collateral reinnervation of the cat''s canine tooth

We examined the morphologic characteristics of the collateral reinnervation of the cat's canine tooth pulp. Collateral reinnervation was stimulated in six adult cats by transecting the ipsilateral inferior alveolar nerve and preventing it from regenerating by sealing the ligated nerve inside a nylon tube. The apices of the canine teeth were examined after 15 weeks using electron microscopy and each axon was counted and measured. Compared with contralateral control teeth, the collateral reinnervation consisted of fewer myelinated fibers and in five animals the number of nonmyelinated fibers was also smaller. In one animal the nonmyelinated count on the operated side exceeded that on the control side. The total number of axons was 33.5% of the number in control teeth, considerably fewer than seen after reinnervation by regenerating fibers in previous experiments. The structural characteristics of the collateral fibers were similar to those of regenerating nerve fibers. Compared with controls the myelinated collateral fibers were smaller and had thinner myelin sheaths. There were more nonmyelinated fibers containing only one axon. More nonmyelinated fibers had exposure of part of their axolemma to the extracellular space and there was also an increase in axoaxonal apposition.     

Ultrastructural three-dimensional reconstruction of group III and group IV sensory nerve endings ("free nerve endings") in the knee joint capsule of the cat: evidence for multiple receptive sites

The noncorpuscular endings ("free nerve endings") of thinly myelinated group III and nonmyelinated group IV afferent nerve fibers have been examined in the knee joint capsule of sympathectomized cats by transmission electron microscopy and three-dimensional reconstruction of series of semi- and ultrathin sections. The sensory ending is the most distal part of a group III or IV nerve fiber that consists only of the sensory axon and associated Schwann cells but lacks a myelin sheath and is not surrounded by perineurium. The sensory axon divides into several branches and forms a terminal tree. The branches run either as single fibers or within small Remak bundles in parallel to sensory axons of other endings; they spread along vessel walls and also extend into dense connective tissue. Each sensory axon consists of a series of spindle-shaped thick segments ("beads") connected by waist-like thin segments. Thus all axons of sensory endings have a string-of-beads appearance, which resembles that of efferent sympathetic nerve fibers. The beads of the sensory axon and the end bulb at its tip show the same ultrastructural features which are characteristic of receptive sites: an accumulation of mitochondria and glycogen particles and various vesicles in the axoplasm and "bare" areas of axolemma that are not covered by Schwann cell processes. Group III and group IV sensory endings differ in the length of their branches (up to 200 microM in group III vs. more than 300 microM in group IV), number of beads per 100 microM axon length (about seven vs. nine or ten), mean diameter of axons (0.9-1.5 microM vs. 0.3-0.6 microM), and the presence of a neurofilament core consisting of bundles of parallel microfilaments only in group III. In conclusion, we propose that the sensory part of noncorpuscular "free nerve endings" is formed by the entire terminal tree of group III or group IV nerve fibers and that the beads in the course of the sensory axon represent multiple receptive sites.     

The cell bodies of origin of sympathetic and sensory axons in some skin and muscle nerves of the cat hindlimb

Cell bodies of sensory and sympathetic axons projecting to skin and skeletal muscle of the cat hindlimb have been labeled retrogradely with horseradish peroxidase (HRP) in order to study location, size, and numbers of the somata of these neurons. HRP was applied to the freshly transected axons of nerves supplying hairy skin (superficial peroneal, SP; sural, Su), hairy and hairless skin of the paw (medial plantar, MP), or skeletal muscle (gastrocnemius-soleus, GS). Serial sections of lumbosacral dorsal root and sympathetic ganglia were studied after standard histochemical processing. Additionally, the numbers of myelinated fibers in the same nerves were determined. All sensory somata and 99.4% of sympathetic cell bodies were located ipsilaterally. Sensory somata were commonly restricted to two adjacent dorsal root ganglia (usually L6–7 for SP, MP; L7-S1 for Su, GS). Although sympathetic somata were more widely distributed rostrocaudally, their maximum frequency always occurred in the segmental ganglia immediately rostral to the sensory outflows, i.e., corresponding to rami communicantes grisei. Dimensions of sympathetic somata varied little between populations projecting to different tissues and were unimodally distributed. The size distributions of sensory somata were characterized by a peak between 10 and 20 μm radius, similar to sympathetic somata, and a varying smaller number of cells ranging up to 60 μm radius. Each nerve had a characteristic distribution profile of afferent somata. A population of very small cells was only present in GS, while the largest sensory somata in GS and MP were bigger than those in SP and Su. Numerical analysis of the data disclosed the characteristic composition of both myelinated and unmyelinated fibers in each nerve studied.     

Development of innervation within syngeneic thymus tissue transplanted under the kidney capsule of the nude mouse: a light and ultrastructural microscope study

The developing thymus, prior to the onset of its functional and structural organization, is innervated by the autonomic nervous system (ANS). The present study extends these earlier findings by analyzing at the light and ultrastructural microscope level the distribution of ANS nerves within normal 18-day-old embryo thymus, adult thymus, and thymic tissue transplanted under the kidney capsule of syngeneic nude mice. The results of this study showed that prior to birth the murine thymus is innervated by AChE-positive fibers that are distributed at the corticomedullary boundaries and throughout the adjacent cortex, as well as to cells beneath the thymic capsule. This pattern of distribution remains constant during adult life. The examination of the various thymic tissues by ultrastructural analysis demonstrated that myelinated and nonmyelinated fiber bundles penetrate the thymic capsule and the interface between the kidney and the thymus transplant. The myelinated fibers measured 2 micron or less in diameter; the nonmyelinated fibers were 1 micron in diameter. Myelination did not accompany the nerves into the parenchyma of the normal gland but was observed in the intralobular trabeculae of the transplanted thymus. Subcapsular nerves form a network that terminates among the thymocytes, whereas intrathymic nerves enter the parenchyma in bundles along the vasculature and interlobular septa before penetrating into the deeper layers of the thymic cortex. Some larger nerve fibers terminate in the corticomedullary boundaries and in the interlobular septa. Smaller fibers form en passant boutons near parenchymal cells. The innervation of both the normal thymus and the transplanted thymus prior to the onset of thymic immune function supports a role for ANS innervation in the development of thymic competency.     

Sodium channel Na(v)1.6 is expressed along nonmyelinated axons and it contributes to conduction

Nodes of Ranvier in myelinated fibers exhibit a complex architecture in which specific molecules organize in distinct nodal, paranodal and juxtaparanodal domains to support saltatory conduction. The clustering of sodium channel Na(v)1.6 within the nodal membrane has led to its identification as the major nodal sodium channel in myelinated axons. In contrast, much less is known about the molecular architecture of nonmyelinated fibers. In the present study, Na(v)1.6 is shown to be a significant component of nonmyelinated PNS axons. In DRG C-fibers, Na(v)1.6 is distributed continuously from terminal receptor fields in the skin to the dorsal root entry zone in the spinal cord. Na(v)1.6 is also present in the nerve endings of corneal C-fibers. Analysis of compound action potential recordings from wildtype and med mice, which lack Na(v)1.6, indicates that Na(v)1.6 plays a functional role in nonmyelinated fibers where it contributes to action potential conduction. These observations indicate that Na(v)1.6 functions not only in saltatory conduction in myelinated axons but also in continuous conduction in nonmyelinated axons.     

Effects of 4-aminopyridine on rapidly and slowly conducting axons of rat corpus callosum

The action potentials of some normal mammalian peripheral myelinated and non-myelinated nerve fibers are known to differ with respect to the role of voltage-dependent potassium currents in membrane repolarization. Because comparable differences have not been fully established for mammalian cerebral axons, the present work examined the influence of the potassium blocker, 4-aminopyridine (4-AP), on field potentials in the rat corpus callosum. Three successive negative waves could be detected in field potentials obtained with recording electrodes positioned several millimeters from the site of callosal stimulation. Spatial and functional properties of these field components indicated that the two shorter-latency waves were due to activity in callosal fibers. Furthermore, the discontinuity of the first and second waves, not only with respect to latency but also in pattern of recruitment and in refractoriness, suggested that they largely reflected activity in respective myelinated and nonmyelinated groups of callosal fibers. When 4-AP was injected into the callosum, or superfused onto surgically exposed callosal fibers, the duration and amplitude of the second negative wave were markedly increased. The first negative wave either was not affected or was reduced in amplitude by 4-AP application. This contrasting effect was identical to that known to distinguish normal myelinated and nonmyelinated mammalian peripheral nerve fibers and provided some evidence suggesting that mammalian cerebral axons have a functional organization similar to that of peripheral fibers.     

Altered peripheral myelination in mice lacking GABAB receptors

Emerging evidence implicates gamma-aminobutyric acid type B (GABA(B)) receptors in peripheral nervous system (PNS) functions. In order to elucidate which biochemical, morphological and functional parameters of peripheral nerve fibers depend on GABA(B) receptors we studied GABA(B1)-deficient mice, which are devoid of functional GABA(B) receptors. Here, we show that GABA(B1)-deficient mice exhibit morphological and molecular changes in peripheral myelin, including an increase in the number of irregular fibers and increases in the expression levels of the myelin proteins PMP22 and P0. Moreover, the number of small myelinated fibers and small neurons of the lumbar dorsal root ganglia is higher in GABA(B1)-deficient mice than in wild-type littermates. We further show that GABA(B1)-deficient mice exhibit gait alterations and reduced allodynia. In summary, our findings implicate GABA(B) receptors in the PNS myelination process and raise the possibility that PNS alterations contribute to the sensory phenotypes of GABA(B1)-deficient mice.     

Rett syndrome: findings suggesting axonopathy and mitochondrial abnormalities

We report the histopathologic findings of 3 sural nerve biopsies and 1 muscle biopsy from 3 patients with Rett syndrome. The 3 sural nerve biopsies demonstrated a few ultrastructural abnormalities, including the presence of many Pi-granules and mitochondrial changes in the cytoplasm of Schwann cells, occasional bands of Büngner and onion-bulb formations, and mitochondrial alterations in myelinated axons. Morphometric analysis disclosed reduction in the number of large myelinated fibers with normal densities in comparison to those of an age-matched normal control. Light microscopic examination of the biopsied muscle from a 6-year-old patient with Rett syndrome revealed the existence of many small, dark, angulated fibers with NADH-TR staining. Ultrastructural investigation of the muscle confirmed the presence of the dumbbell-shaped mitochondria. Peripheral nerve involvement and the possibility of mitochondrial abnormalities in Rett syndrome were suggested by the results.     

Slowly conducting,low-threshold components of sensory nerve potentials in peripheral neuropathy: A microneurographic study

By recording the averaged potential from a microelectrode inserted into the nerve, small late components of the sensory nerve action potential can be analyzed. We studied the thresholds of the late components in median nerves of 15 normal subjects and 37 patients with peripheral neuropathies. Under normal conditions, the late components reflected the activities of small myelinated fibers which had high thresholds. In the patients with peripheral neuropathies, a part of the late components often had abnormally low thresholds, occasionally the lowest. The presence of low-threshold, late components was related to the pathologies of the different types of neuropathy and correlated with the numbers of regenerating clusters in sural nerve biopsies. These slowly conducting, low-threshold components may originate from regenerating or remyelinating fibers, and therefore should have prognostic value. © 1997 John Wiley & Sons, Inc. Muscle Nerve 20: 961–968, 1997     

Photochemically-induced ischemic injury of the rat sciatic nerve: a light- and electron microscopic study

Lesion of presumably ischemic origin of the rat sciatic nerve was induced photochemically by laser irradiation combined with systemic administration of a photosensitizing organic dye, erythrosin B. We have studied the pathologic features of the nerve after the photochemical insult with light- and electronmicroscopy and related them to behavioral signs of neuropathic pain. At the irradiated nerve site, occlusion of blood vessels was seen and the vessels were packed with aggregated thrombocytes, fibrins and deformed erythrocytes, supporting the notion that photochemical reaction caused intraneural ischemia. The degree of the nerve injury at the center of irradiation was related to the duration of the laser exposure. Brief irradiation (30 seconds) only caused identifiable injury to myelinated fibers, whereas longer irradiation (2 minutes) caused greater injury to myelinated and unmyelinated fibers, characterized by extensive axonal degeneration and demyelination. The rats irradiated for 2 minutes, but not 30 seconds, exhibited neuropathic pain-like behaviors, expressed as mechanical and cold allodynia. The nerve injury was most severe 7 days after ischemia and regeneration of both myelinated and unmyelinated fibers was observed 3 months later. The nerve caudal to the irradiation exhibited Wallerian degeneration 7 days after the insult, whereas at 10 mm proximal to the irradiation the nerve was largely normal. It is thus concluded that photochemically induced intraneural ischemia caused injury to both myelinated and unmyelinated fibers, with myelinated fibers being more susceptible. However, the development of neuropathic pain-like behaviors may require injury to the unmyelinated fibers.     

Substance P immunoreactive nerve terminals in the dorsolateral nucleus of the tractus solitarius: roles in the baroreceptor reflex

Physiological and light microscopic evidence suggest that substance P (SP) may be a neurotransmitter contained in first-order sensory baroreceptor afferents; however, ultrastructural support for this hypothesis is lacking. We have traced the central projections of the carotid sinus nerve (CSN) in the cat by utilizing the transganglionic transport of horseradish peroxidase (HRP). The dorsolateral subnucleus of the nucleus tractus solitarius (dlNTS) was processed for the histochemical visualization of transganglionically labeled CSN afferents and for the immunocytochemical visualization of SP by dual labeling light and electron microscopic methods. Either HRP or SP was readily identified in single-labeled unmyelinated axons, myelinated axons, and nerve terminals in the dlNTS. SP immunoreactivity was also identified in unmyelinated axons, myelinated axons, and nerve terminals in the dlNTS, which were simultaneously identified as CSN primary afferents. However, only 15% of CSN terminals in the dlNTS were immunoreactive for SP. Therefore, while the ultrastructural data support the hypothesis that SP immunoreactive first-order neurons are involved in the origination of the baroreceptor reflex, they suggest that only a modest part of the total sensory input conveyed from the carotid sinus baroreceptors to the dlNTS is mediated by SP immunoreactive CSN terminals. Five types of axo–axonic synapses were observed in the dlNTS. SP immunoreactive CSN afferents were very rarely involved in these synapses. Furthermore, SP terminals were never observed to form the presynaptic element in an axo–axonic synapse with a CSN afferent. Therefore, SP does not appear to be involved in the modulation of the baroreceptor reflex in the dlNTS.     

Peripheral nerve endoneurial microangiopathy and necrosis in rats with insulinoma

Peripheral nerve pathology related to chronic hyperinsulinemia and hypoglycemia has yet to be fully explored. Here we conducted a systematic quantitative analysis of morphological alterations in peripheral sensory and motor nerve fibers and endoneurial microvasculature in longstanding insulinoma-carrying rats (I-rats; n=12). Age-matched normal rats (n=6) served as controls. Over the 15-month observation period, two of I-rats developed paresis of the hind limbs when their blood glucose level fell below 1.7 mmol/l. These animals showed a massive myelinated fiber loss associated with active degeneration of residual myelinated fibers and multiple endoneurial microvascular occlusions at the sciatic nerve level. The rest of the non-paretic I-rats showed a decreased density of large myelinated fibers with axonal degeneration in the peroneal nerve and an increased density of small myelinated fibers with preserved morphology in the sural nerve. This was associated with endoneurial microangiopathic changes indicative of endoneurial ischemia/hypoxia in the sciatic and peroneal nerves, and an increase in endoneurial microvascular density in the sciatic and sural nerves. In conjunction with previous data, these findings suggest that the observed increase in endoneurial microvascular density may be a compensatory response to endoneurial ischemia/hypoxia induced by chronic hyperinsulinemia in I-rats without paresis. In conclusion, the present study showed characteristic morphological alterations in peripheral sensory and motor nerve fibers associated with microangiopathy indicative of endoneurial ischemia/hypoxia in the sciatic and peroneal nerves, and provides the first evidence for the occurrence of endoneurial necrosis in the sciatic nerve, to which the hind limb paresis can be ascribed in I-rats.     

Abnormal Schwann Cell-Axon Interaction In EBF-2 "Knock Out" Mice

Sometimes standard sensory nerve conduction studies show normal results in patients with definite symptoms of sensory polyneuropathy. This is usually explained because standard neurophysiological tests evaluate only large myelinated fibers and do not assess the slowest conducting fibers, more distal segments of the nerves and tactile receptors. Tactile stimulation is a test, not routinely available, that assesses the function of tactile receptors and conduction of fibers that are depolarized by these receptors. During conventional sensory nerve conduction studies (in patients and healthy subjects) through surface electrodes, where we slowly increased the intensity of the stimulus, we occasionally observed a sensory response characterized by a particular morphology with two peaks. After several experiments (performed in the neurophysiological laboratories of Catholic University of Rome and of University of Uppsala) we argued that the double component of the response is the expression of the stimulation of tactile receptors (and depolarization of their related fibers). Therefore an electrical stimulation through conventional EMG equipment allows us to assess function of tactile receptors (and related nerve fibers). This observation may have important diagnostic application in clinical practice to evaluate suspected polyneuropathies negative to neurophysiological conduction studies.     

Nerve fibers of the eighth nerve and their distribution to sensory nerves of inner ear in the bullfrog

The number of both myelinated and unmyelinated nerve fibers has been determined of the eighth nerve and each of its sensory branches in the bullfrog. No significant differences were found in the size distribution of the myelinated nerve fibers among the three ampullary nerves or between the saccular and lagenar nerves. Comparison of these two functional groups also showed no significant differences in the myelinated nerve fiber size distribution between the angular acceleration and otolithic receptors, collectively the equilibrium receptors. The wide variability, that was found not only among the animals but also between the right and left sides, most likely resulted in the absence of differences. Highly significant differences were noted when the myelinated nerve fiber diameter distributions to the acoustic organs were compared to those of the equilibrium organs. The most notable difference was the absence of any myelinated nerve fibers whose diameters were greater than 8 μm.     

Clinical, electrophysiological and histological analysis of adult onset chronic sensory neuropathy

We reported clinical, electrophysiological and pathological findings in 10 patients with acquired, slowly progressive sensory neuropathy. Group I patients which had delayed sensory nerve conduction velocities (SNCV) or no response of SCV in electrophysiological studies were observed to have marked deep sensory disturbances. On the other hand, Group II who had abnormalities of both SNCV and motor nerve conduction velocities electrophysiologically has tendency to show superficial sensory disturbances more dominantly than that of deep sensation. Histological findings of biopsied sural nerve in Group I were mainly composed of diminution of large myelinated fibers, in Group II the diminution of large myelinated fibers with decreased unmyelinated fibers were observed. However, there were no significant difference between Group I and Group II in the quantitative analysis of myelinated fibers. There were no definite tendency under the basic disease or background factors of these 10 patients. Only 3 of these 10 patients had immunological abnormalities such as RA factors or antinuclear antibodies.     

Effect of NGF on the survival of rat retinal ganglion cells following optic nerve section

The ability of NGF (2.5S subunit) to support the survival of adult rat retinal ganglion cells (RGCs) and optic nerve fibers after intracranial section of the optic nerve was investigated. NGF was injected intraocularly at a dose of 3 micrograms/injection every 2.3 d from the day of axotomy to analysis. Control animals received cytochrome c injections. The survival of RGCs was analyzed in whole-mounted retinas after either cresyl violet staining or labeling with HRP applied to the proximal stump of the optic nerve. Survival times were 5 and 7 weeks. Diameter distribution and number of myelinated optic nerve fibers were assessed in ultrathin cross sections of the optic nerve. We found that RGCs surviving axotomy were much more numerous following NGF treatment compared with controls. Large-size cells were, in particular, preserved by NGF treatment. The quantitative ultrastructural studies indicated that the number of myelinated optic nerve fibers at 5 and 7 weeks postaxotomy was significantly greater in the NGF group with respect to the cytochrome c group. In agreement with the results obtained at the level of the RGCs, large-diameter axons were, in particular, preserved. We conclude that NGF injected intraocularly is effective in promoting the survival of RGCs and optic nerve fibers at least for a period as long as 7 weeks after intracranial section of the optic nerve.     

Growth-related order of the retinal fiber layer in goldfish

The retinal fiber layer and the juxtaretinal portion of the optic nerve of goldfish have been studied with light and electron microscopy in order to determine whether the age-related order of fibers in the nerve originates in the retina. In the retina, no patent spaces (channels) were noted. The fibers ran in fascicles and consisted of two classes: nonmyelinated fibers, which ran superficially (close to the vitreal surface), and "myelinated" fibers, which ran more deeply and were loosely wrapped by processes presumed to be glial. The myelinated fibers were larger and presumably older. The nonmyelinated fibers are believed to be the young ones, from the peripheral, more recently generated, ganglion cells, for the following reasons. (1) Their size and cytoskeletal elements were typical of young axons. (2) They were the only axons in peripheral retina. (3) They were continuous with the nonmyelinated fibers in the nerve, previously shown to be the young ones. (4) When retinal axons were cut peripherally, the degenerating axons were in the superficial part of the fiber layer. (5) Growth cones, presumably from the newest ganglion cells, were always observed at the most superficial position in the fiber layer, in direct contact with the basal lamina of the inner limiting membrane superficially and nonmyelinated fibers deeply. The nonmyelinated fibers always clustered together in the retinal fiber layer and occupied the most central portion in the cross-section of the optic nerve head. Thus, the age-related organization of fibers in the nerve is established in the retina. These results are discussed in the context of growth, with the aim of evaluating the relative importance of four factors that might influence the intraretinal course of the growth cone. Its interactions with other fibers and with the basal lamina of the inner limiting membrane seem to be more important than interactions with the glial end feet or guidance into open, preformed channels.