Chromatolysis: Do injured axons regenerate poorly when ribonucleases attack rough endoplasmic reticulum, ribosomes and RNA?

After axonal injury, chromatolysis (fragmentation of Nissl substance) can occur in the soma. Electron microscopy shows that chromatolysis involves fission of the rough endoplasmic reticulum. In CNS neurons (which do not regenerate axons back to their original targets) or in motor neurons or dorsal root ganglion neurons denied axon regeneration (e.g., by transection and ligation), chromatolysis is often accompanied by degranulation (loss of ribosomes from rough endoplasmic reticulum), disaggregation of polyribosomes and degradation of monoribosomes into dust‐like particles. Ribosomes and rough endoplasmic reticulum may also be degraded in autophagic vacuoles by ribophagy and reticulophagy, respectively. In other words, chromatolysis is disruption of parts of the protein synthesis infrastructure. Whereas some neurons may show transient or no chromatolysis, severely injured neurons can remain chromatolytic and never again synthesize normal levels of protein; some may atrophy or die. Ribonuclease(s) might cause the following features of chromatolysis: fragmentation and degranulation of rough endoplasmic reticulum, disaggregation of polyribosomes and degradation of monoribosomes. For example, ribonucleases in the EndoU/PP11 family can modify rough endoplasmic reticulum; many ribonucleases can degrade mRNA causing polyribosomes to unchain and disperse, and they can disassemble monoribosomes; Ribonuclease 5 can control rRNA synthesis and degrade tRNA; Ribonuclease T2 can degrade ribosomes, endoplasmic reticulum and RNA within autophagic vacuoles; and Ribonuclease IRE1α acts as a stress sensor within the endoplasmic reticulum. Regeneration might be improved after axonal injury by protecting the protein synthesis machinery from catabolism; targeting ribonucleases using inhibitors can enhance neurite outgrowth and could be a profitable strategy in vivo. © 2018 Wiley Periodicals, Inc. Develop Neurobiol, 2018

Selective vulnerability of dorsal root ganglia neurons in experimental rabies after peripheral inoculation of CVS-11 in adult mice

The involvement of dorsal root ganglia was studied in an in vivo model of experimental rabies virus infection using the challenge virus standard (CVS-11) strain. Dorsal root ganglia neurons infected with CVS in vitro show prolonged survival and few morphological changes, and are commonly used to study the infection. It has been established that after peripheral inoculation of mice with CVS the brain and spinal cord show relatively few neurodegenerative changes, but detailed studies of pathological changes in dorsal root ganglia have not previously been performed in this in vivo experimental model. In this study, adult ICR mice were inoculated in the right hindlimb footpad with CVS. Spinal cords and dorsal root ganglia were evaluated at serial time points for histopathological and ultrastructural changes and for biochemical markers of cell death. Light microscopy showed multifocal mononuclear inflammatory cell infiltrates in the sensory ganglia and a spectrum of degenerative neuronal changes. Ultrastructural changes in gangliocytes included features characteristic of the axotomy response, the appearance of numerous autophagic compartments, and aggregation of intermediate filaments, while the neurons retained relatively intact mitochondria and plasma membranes. Later in the process, there were more extensive degenerative neuronal changes without typical features of either apoptosis or necrosis. The degree of degenerative neuronal changes in gangliocytes contrasts with observations in CNS neurons in experimental rabies. Hence, gangliocytes exhibit selective vulnerability in this animal model. This contrasts markedly with the fact that they are, unlike CNS neurons, highly permissive to CVS infection in vitro. Further study is needed to determine mechanisms for this selective vulnerability, which will give us a better understanding of the pathogenesis of rabies.

Axotomized rubrospinal neurons rescued by fetal spinal cord transplants maintain axon collaterals to rostral CNS targets

Neurons that maintain extensive axon collaterals proximal to the site of axotomy may be better able to survive injury. Early lesions of the rubrospinal tract lead to retrograde cell death of the majority of axotomized immature neurons. Transplants of fetal spinal cord tissue rescue axotomized rubrospinal neurons and promote their axonal regeneration. Rubrospinal neurons develop many of their axon collaterals postnatally. The present study tests the hypothesis that the axotomized rubrospinal neurons that are rescued by transplants and regenerate their axons are those neurons that have established axon collaterals to targets rostral to the lesion. Neonatal rats received a transplant of fetal spinal cord tissue placed into a midthoracic spinal cord hemisection. One month after transplantation, the retrogradely transported fluorescent tracers fast blue (FB) and diamidino yellow (DY) were used to identify rubrospinal neurons with collaterals to particular targets. FB was injected either into the interpositus nucleus of the cerebellum or into the gray matter of the cervical enlargement to identify collaterals to these targets, and DY was injected into the spinal cord approximately 5 mm caudal to the transplant and lesion site to label retrogradely the neurons that regenerated their axons. Double labeling was observed in the axotomized neurons of the red nucleus after tracer injections into the cervical spinal cord but not after injections into the cerebellum. This labeling pattern indicates that axotomized rubrospinal neurons that are rescued and regenerate axons caudal to the transplant maintain axon collaterals at cervical spinal cord levels. Cerebellar collaterals do not appear to play a role in the survival and regrowth of axotomized rubrospinal neurons.

Ventral horn motoneurons 10, 20 and 52 weeks after T-9 spinal cord transection

To determine if transneuronal degeneration occurs in ventral horn motoneurons caudal to a spinal cord transection, we completely transected the spinal cord at T-9 in seven-week-old female rats. Ten, 20 or 52 weeks later, the motoneurons of the right sciatic nerve of transected and control rats were retrogradely labeled with Fluoro-Gold. There were no differences between control and transected rats in numbers or rostrocaudal distribution of labeled motoneurons at either 10, 20 or 52 weeks. At 20 weeks, there was no significant difference between control and transected rats in mean cross-sectional area of labeled neurons. We conclude that transneuronal degeneration did not occur.

Retrograde transport of fluoro-gold in corticospinal and rubrospinal neurons 10 and 20 weeks after T-9 spinal cord transection

Retrograde labeling with horseradish peroxidase is greatly diminished in corticospinal and rubrospinal neurons axotomized by complete T-9 spinal cord transection. We found, 10 or 20 weeks after a complete T-9 cord transection, that the number of corticospinal and rubrospinal neurons retrogradely labeled after Fluoro-Gold insertion into a new transection at T-1 did not differ from that of controls. While transection alters uptake, transport, and/or intracellular metabolism of some transportable substances, it does not affect the ability of the neurons to be retrogradely labeled with Fluoro-Gold.

Cytological and cytochemical (RNA) studies on rubral neurons after unilateral rubrospinal tractotomy: the impact of GM1 ganglioside administration

The rubrospinal tract (RST) was cut unilaterally at C2-3 segment in 21 rats that were killed 3, 7, 10, 14, 28, 60, and 90 days later. Additionally, 14 rats, killed 14 or 28 days after lesioning, were treated postoperatively by daily intraperitoneal injections of GM1 ganglioside. Six unoperated, untreated rats served as controls. In untreated animals, axotomized neurons of the magnocellular division of the red nucleus (RN) exhibited cytoplasmic, nuclear, and nucleolar atrophy 7-10 days postoperatively. Atrophy progressed through the 90th postoperative day. Regression analyses disclosed a bimodal pattern to cytoplasmic and nucleolar atrophy, with an initial rapid phase changing to a slower but progressive mode from 14 days postoperatively. Nuclear atrophy proceeded in a unimodal manner. GM1 treatment did not affect these atrophic processes. Neuronal loss did not occur in the axotomized RN through the 60th postoperative day. Axotomized neurons of untreated rats showed significant and progressive reductions in mean somal (cytoplasmic) and nucleolar RNA from, respectively, the 7th and 14th postoperative day. GM1 partly prevented these RNA losses. Both in treated and untreated rats, spinal cord lesions contained many axonal sprouts 2 to 4 weeks after surgery, but newly generated axons did not traverse the rostro-caudal extent of any lesion.

Loss of neurons in the red nucleus after spinal cord transection

Red nucleus neurons, particularly those of the caudal one-half of the nucleus, die or severely atrophy following complete spinal cord transection at T9. The size of residual horseradish peroxidase-labeled cells was smaller at 10 and 15 weeks, but those survivors which could be labeled at 25 weeks were normal in size. Hematoxylin and eosin-stained sections of the red nucleus at 52 weeks postoperative showed loss of cells from all size groups.

Survival and subsequent regeneration of olfactory neurons after a distal axonal lesion

If an axonal lesion is made close enough to the cell body, injured olfactory neurons degenerate and are replaced by new nerve cells arising from undifferentiated mucosal basal cells. Therefore, under these conditions neural regeneration occurs through a process similar to neuronal development. The use of the long garfish olfactory nerve has revealed that neuronal death is not an inevitable consequence of an axonal injury and that the extent of cell death depends on the distance between the site of injury and the perikaryon. A lesion located up to 40 mm from the cell body induces the death of all mature neurons. Between 60 and 100 mm an increasing proportion of neurons survive the injury (from 60 to more than 90%) and are able to regenerate their distal segment. During regeneration, two populations of growing axons have been characterized. The fastest growing fibres (5.8 mm per day) correspond to the small population of neurons which were already growing at the time of the lesion and are able to survive an injury at any distance along the nerve. The majority of the regenerating fibres grow at 0.8 mm per day and corresponds to the damaged mature neurons. Elongation velocities were not affected by the distance from injury to cell body or by the mode of neural repair (development or regeneration). The initial delay between the injury and the beginning of elongation increases linearly with distance at a rate of 1 day cm-1 and is independent of the elongation velocity of the growing neurons. This indicates that the mechanism responsible for the beginning of axonal growth is initiated at or near the cell body and involves the entire axonal stump and not only the area surrounding the crush site. The ability of nerve cells to survive an injury may depend on the length of the axonal stump remaining attached to the cell body and on the level of protein synthesis at the time of the crush. From preliminary results it can be hypothesized that the length of the initial delay is determined by a reorganization of the cytoskeletal elements in the proximal axonal stump.

The axon reaction of lamprey spinal interneurons

Axotomy and partial denervation of giant interneurons (GIs) and lateral cells (LCs) were produced by complete spinal transection in the larval lamprey spinal cord. Both cell types demonstrated a reduction in cytoplasmic basophilia, increase in cell size, nuclear eccentricity, and formation of a chromophilic nuclear cap. This was quantified in the case of cell diameter. During the first 8 weeks of recovery, the GIs with the largest diameters were found progressively further from the scar and this peak change moved at approximately 0.5 mm/day. The increase in size of GIs remained up to 20 weeks post-transection, long after the time required for their axons to regenerate across the scar and form functioning synapses. GIs injected intracellularly with horseradish peroxidase (HRP) also showed this increase in diameter as well as a simplification of their dendritic trees. Intracellular recordings from GIs revealed changes in the frequency and amplitude of spontaneous synaptic input. In the first two weeks after transection, spontaneous excitatory postsynaptic potentials (EPSPs) and inhibitory postsynaptic potentials (IPSPs) were less frequent than in control cells. After 6 weeks of recovery they became more frequent than in control cells. EPSPs predominated in axotomized GIs, while in control cells they constituted only 36% of the total of spontaneous potentials. A reversible increase in the amplitude of these EPSPs occurred at 3-4 weeks of recovery time. The resting membrane potential was significantly reduced by the 6th week after transection and returned to normal after the 22nd week.

A horseradish peroxidase study of the rubral and cortical afferents to the lateral reticular nucleus in the rat

The origin and organization of supraspinal afferents to the lateral reticular nucleus (LRN) in the rat were studied by means of the retrograde axonal transport of horseradish peroxidase (HRP). HRP was deposited into the LRN via both dorsal (stereotaxic) and ventral (microsurgical) routes. The entire cerebrum, brainstem, and cerebellum were surveyed for retrogradely labelled neurons. Significant projections arose from the contralateral red nucleus and the contralateral frontoparietal cortex. The rubral projection arose from neurons in the caudal two-thirds of the red nucleus. Ventrally and ventrolaterally located neurons projected to rostrolateral LRN, while dorsal and dorsomedial neurons projected to rostromedial LRN. The projection from the cerebral cortex arose from neurons located in layer V of the frontoparietal region. Rubral and cerebrocortical projections overlap in the rostral LRN, making this region of the nucleus a site of integration of descending inputs with ascending spinal signals.

The axon reaction of the goldfish mauthner cell and factors that influence its morphological variability

The axon reaction of the goldfish Mauthner cell, elicited by spinal cord transection, included somatic swelling, nuclear eccentricity, chromatolysis, nuclear infolding, and a perinuclear buildup of basophilic material. The latter three changes were found most consistently and showed gradations which were ranked quantitatively. The time of onset of chromatolysis and nucleus-associated changes depended upon the distance of the wound from the Mauthner cell soma. Specifically, for Mauthner axons cut at 5, 10.5, and 20 mm distal to their somata, the approximate postoperative times of onset were 10, 20, and 40 days, respectively. Mauthner cells axotomized 42 mm distally did not display a consistent axon reaction. Cell atrophy and death were not found in cells axotomized 10.5, 20, or 42 mm from their somata up to 285 postoperative days, but were observed at the longer postoperative intervals (421 days) in neurons cut 5 mm distally and were consistently found in neurons axotomized less than 1.6 mm from their somata. The axon reactions of Mauthner cells within a pair were frequently different. This variability cannot be explained by the influence of cut site or postoperative interval and is hypothesized to result from different metabolic conditions of the individual cells.

Neuronal and oligodendrocytic response to cortical injury: ultrastructural and cytochemical changes

A needle wound was made in the adult rat cerebral cortex. Responses of neurons and oligodendrocytes at the site of injury were followed over a period of 450 days and correlations made between morphological and enzyme cytochemical changes to clarify some phenomena previously unresolved. Evidence from acid phosphatase activity in degenerating neurons showed no increase in the number of cytochemically stained lysosomal profiles nor changes in the subcellular localization of the acid phosphatase reaction product. Our observations indicated that the majority of dying neurons were not digested by their own acid phosphatase 'autodigestion' but by the process of heterodigestion. The time-course study revealed that not all the traumatized neurons were eliminated but some persisted permanently in an attenuated 'atrophic' state. The atrophic neurons were small in size with low cytoplasmic-nuclear ratios and exhibited low levels of glucose-6-phosphatase and cytochrome oxidase activities. The acid phosphatase activity was slightly increased as evidenced by cytochemically stained hypertrophic Golgi cisternae and a slight increase in the number of lysosomes. The low level of enzyme activities concerned with carbohydrate metabolism reflected the low metabolic activity in atrophic neurons whilst an increase in Golgi-lysosomal enzyme activity suggested some anabolic process necessary for their survival. Oligodendrocytes displayed only minor changes in morphology, and their glucose-6-phosphatase and cytochrome oxidase activities were normal, suggesting that these cells have little or no involvement in the repair of a cerebral wound. The absence of significant changes in lysosomal acid phosphatase activity indicated a minimal role, if any, of oligodendrocytes in the process of phagocytosis.

A qualitative electron microscopic study of the corticopontine projections after neonatal cerebellar hemispherectomy

The present study shows that 3--5 days following lesions of the dentate and interposed nuclei in normal adult rats degenerating axons and axon terminals can be detected in the contralateral pontine gray. The degenerating axon terminals form Gray's type I axo-dendritic contacts with fine and intermediate dendrites measuring between 0.8--2.4 microns. The present study also investigates, by electron microscopy, the synaptic rearrangement of the sensorimotor corticopontine projections following neonatal left cerebellar hemispherectomy. Following neonatal left cerebellar hemispherectomy, the right sensorimotor and adjacent cortex (SMC) presents a very dense ipsilateral and a modest amount of contralateral corticopontine projections in contrast with a predominantly ipsilateral corticopontine projection seen in the normal adult rat. As with the ipsilateral corticopontine projection seen in the normal adult animal, the bilateral corticopontine projections seen in the experimental animals form contacts with dendrites suggestive of Gray's type I synapses. While the corticopontine projections in normal control animals form synapses with fine dendrites measuring 0.2--1.2 micron the corticopontine projections in the experimental animals form synaptic relations with fine dendrites and with intermediate dendrites measuring 0.2--2.4 microns. As the normal cerebellopontine fibers from the dentate and interposed nuclei also form axo-dendritic synapses on fine and intermediate dendrites and the contracts formed are also of Gray's type I synapses, it is possible that some of the newly formed corticopontine fibers in the experimental animals might have replaced the cerebellopontine fibers synapsing on intermediate dendrites. Synaptic rearrangement appears to take place as suggested by the presence of synaptic complexes in which one axon terminal contacts two or more dendrites or two or more axon terminals contact one dendrite. Such complexes are frequently seen to undergo degeneration following the right SMC lesion in the experimental animals. Other complex synaptic structures are also present in both the right and left pontine gray in the experimental animals. They are not seen to undergo degeneration following the right SMC lesions. Occasional features of neuronal reaction could still be seen in both sides of the pontine gray for as long as 3--6 months after the neonatal cerebellar lesions.