The cytokine network of Wallerian degeneration: IL-10 and GM-CSF

Wallerian degeneration (WD) is the inflammatory response of peripheral nerves to injury. Evidence is provided that granulocyte macrophage colony stimulating factor (GM-CSF) contributes to the initiation and progression of WD by activating macrophages and Schwann, whereas IL-10 down-regulates WD by inhibiting GM-CSF production. A significant role of activated macrophages and Schwann for future regeneration is myelin removal by phagocytosis and degradation. We studied the timing and magnitude of GM-CSF and IL-10 production, macrophage and Schwann activation, and myelin degradation in C57BL/6NHSD and C57BL/6-WLD/OLA/NHSD mice that display normal rapid-WD and abnormal slow-WD, respectively. We observed the following events in rapid-WD. The onset of GM-CSF production is within 5 h after injury. Production is steadily augmented during the first 3 days, but is attenuated thereafter. The onset of production of the macrophage and Schwann activation marker Galectin-3/MAC-2 succeeds that of GM-CSF. Galectin-3/MAC-2 production is up-regulated during the first 6 days, but is down-regulated thereafter. The onset of myelin degradation succeeds that of Galectin-3/MAC-2, and is almost complete within 1 week. IL-10 production displays two phases. An immediate low followed by a high that begins on the fourth day, reaching highest levels on the seventh. The timing and magnitude of GM-CSF production thus enable the rapid activation of macrophages and Schwann that consequently phagocytose and degrade myelin. The timing and magnitude of IL-10 production suggest a role in down-regulating WD after myelin is removed. In contrast, slow-WD nerves produce low inefficient levels of GM-CSF and IL-10 throughout. Therefore, deficient IL-10 levels cannot account for inefficient GM-CSF production, whereas deficient GM-CSF levels may account, in part, for slow-WD.     

Mechanisms of macrophage recruitment in Wallerian degeneration

Monocytes/macrophages are important effector cells in myelin removal during Wallerian degeneration. Experiments with the mouse mutant C57BL/Ola revealed prolonged axonal survival and reduced phagocytic cell recruitment after nerve transsection. In the present study, we compared the course of Wallerian degeneration in peripheral nerves of C57BL and C57BL/Ola mice in vivo and in vitro. In vivo experiments confirmed earlier investigations describing a delayed degeneration in the C57BL/Ola mutant compared with C57BL mice which were used as control animals without abnormal degeneration. Quite different results were seen in experiments in vitro: degenerating nerve segments of C57BL/Ola mice revealed pronounced axonal breakdown even in the absence of non-resident phagocytic cells. There was no difference in vitro compared with degenerating nerves from C57BL mice. The differences observed between the in vivo and in vitro situations suggest that axonal breakdown plays an important role in the initiation of macrophage recruitment to degenerating peripheral nerves.     

Phospholipase A2 plays an important role in myelin breakdown and phagocytosis during Wallerian degeneration

Phospholipase A(2) (PLA(2)) hydrolyzes phosphatidylcholine to lysophosphatidylcholine and arachidonic acid. The former can induce myelin breakdown and the latter, via eicosanoids, can stimulate inflammatory responses. Immunohistochemical analysis of secreted (sPLA(2)) and cytosolic (cPLA(2)) forms of the enzyme was assessed in the injured adult rat sciatic and optic nerves. sPLA(2) and cPLA(2) are expressed in the first 2 weeks in the injured sciatic nerve, which correlates with rapid Wallerian degeneration in peripheral nerves. In contrast, both forms of PLA(2) were not expressed in the optic nerve for the first 3 weeks after crush injury, which correlates with slow Wallerian degeneration in the central nervous system (CNS). In addition, PLA(2) is not expressed in the lesioned sciatic nerve of C57BL/Wld(s) mutant mice in which Wallerian degeneration is severely retarded. Blocking cPLA(2) in the transected sciatic nerve of C57BL/6 mice, which have a naturally occurring null mutation for the major from of sPLA(2), resulted in a marked slowing of myelin and axonal degradation and phagocytosis in the distal nerve segment. These results provide direct evidence of an important role for cPLA(2) in Wallerian degeneration.     

Evidence that the Rate of Wallerian Degeneration is Controlled by a Single Autosomal Dominant Gene

In a substrain of C57BL mice, C57BL/Ola, Wallerian degeneration in the distal segment of the severed sciatic nerve is extremely slow when compared to other mice. Despite this very slow degeneration in the distal segment regeneration of the motor nerves is not impaired. From suitable genetic outcrosses and backcrosses, the authors provide evidence that the rate of Wallerian degeneration in this strain is controlled by a single autosomal gene product. The authors have also shown that the rate of degeneration, in C57BL/Ola mice, is influenced by the environment in which the animals were bred and housed. Wallerian degeneration in the sciatic nerves of mice raised in isolators is slower than in those raised in a conventional animal house. This strain of mouse may prove to be of value in the understanding of nerve degeneration and regeneration.     

Evidence that Very Slow Wallerian Degeneration in C57BL/Ola Mice is an Intrinsic Property of the Peripheral Nerve

We have described a mutant mouse, C57BL/Ola, in which Wallerian degeneration following peripheral nerve transection is very slow. Our previous results suggested that recruited monocytes play a role in rapid Wallerian degeneration. The nature of the mutation in C57BL/Ola mice is not known and we have investigated whether the defect is intrinsic to the nerve or due to a defect in the circulating monocytes. We have made chimaeric mice in which bone marrow from histocompatible mice, with rapidly degenerating nerves and normal monocyte recruitment, was used to reconstitute irradiated C57BL/Ola mice and vice-versa. A substantial degree of donor repopulation of the hosts was confirmed by measures of the levels of glucose-phosphate isomerase alloenzymes in blood and tissue samples from the two different strains. The rate of degeneration of the transected sciatic nerve was found to be host-dependent, providing evidence that the mutation affects cell populations intrinsic to the nerve and not the circulating monocytes. We provide additional evidence that the peripheral nerves of C57BL/Ola mice are different from those of other mice as they degenerate at a slower rate in vitro.     

Respective roles of inflammation and axonal breakdown in the regulation of peripheral nerve hemopexin: an analysis in rats and in C57BL/Wlds mice

We have previously demonstrated that one of the peripheral nerve responses to injury is the overexpression of hemopexin (HPX). Here, we demonstrate that Wallerian degeneration is required for this response, since HPX does not increase in C57BL/Wlds mice, which display a severely impaired Wallerian degeneration. We also show that HPX synthesis is dramatically increased in macrophages during their activation or after IL-6 stimulation. However, IL-6-driven HPX overexpression occurs in vivo and in vitro in the absence of substantial macrophage invasion. We conclude that, after nerve injury, HPX overexpression occurs first in Schwann cells as a result of axotomy and is subsequently regulated by inflammation. Furthermore, our results and those already described suggest that IL-6, synthesized by the various cell types producing HPX, control nerve HPX expression via paracrine and autocrine mechanisms.     

Neurofilament redistribution in transected nerves: evidence for bidirectional transport of neurofilaments

Nerve fibers of the C57BL/6/Ola mouse exhibit very slow Wallerian degeneration following axotomy, thus allowing prolonged observation of mammalian axons separated from their cell bodies. The present study utilized teased-fiber preparations, silver histochemistry, immunocytochemistry, and electron microscopy to examine the distribution of axonal components in the distal stumps of axotomized sciatic nerves in C57BL/6/Ola mice. In examining nerve segments at varying intervals after nerve transection, we found no evidence of proximal-to-distal "emptying out" of the cytoskeleton, as would be predicted if the cytoskeleton in these transected nerves were undergoing anterograde transport as an assembled structure. Instead, we observed a gradual redistribution of cytoskeletal constituents over time, dominated by the progressive accumulation of neurofilaments at the severed ends of axons. In particular, there were massive accumulations at the proximal ends of the distal stumps. These results strongly suggest that, at least in transected nerve fibers, neurofilaments can be transported bidirectionally.     

Very Slow Retrograde and Wallerian Degeneration in the CNS of C57BL/Ola Mice

Wallerian degeneration following peripheral nerve transection in C57BL/Ola mice is very slow in comparison to other strains of mice. We show that following optic nerve transection, the axons of retinal ganglion cells in C57BL/Ola mice undergo very slow Wallerian degeneration and that retrograde degeneration of the ganglion cell bodies is much slower than in other strains of mice. The results suggest that the gene product affecting Wallerian degeneration in the peripheral nervous system (PNS) also confers a greater resistance to degeneration on central nervous system (CNS) neurons.     

Demyelination can proceed independently of axonal degradation during Wallerian degeneration in wlds mice

Peripheral nerve injury induces axonal degeneration and demyelination, which are collectively referred to as Wallerian degeneration. It is generally assumed that axonal degeneration is a trigger for the subsequent demyelination processes such as myelin destruction and de-differentiation of Schwann cells, but the detailed sequence of events that occurs during this initial phase of demyelination following axonal degeneration remains unclear. Here we performed a morphological analysis of injured sciatic nerves of wlds mice, a naturally occurring mutant mouse in which Wallerian degeneration shows a significant delay. The slow Wallerian degerenation phenotype of the wlds mutant mice would enable us to dissect the events that take place during the initial phase of demyelination. Ultrastrucural analysis using electron microscopy showed that the initial process of myelin destruction was activated in injured nerves of wlds mice even though they exhibit morphologically complete protection of axons against nerve injury. We also found that some intact axons were completely demyelinated in degenerating nerves of wlds mice. Furthermore, we observed that de-differentiation of myelinating Schwann cells gradually proceeded even though the axons remained morphologically intact. These data suggest that initiation and progression of demyelination in injured peripheral nerves is, at least in part, independent of axonal degeneration.     

Long-term consequences of impaired regeneration on facial motoneurons in the C57BL/Ola mouse

Peripheral nerves of the C57BL/Ola mouse mutant undergo markedly slowed Wallerian degeneration following injury. This is associated with impaired regeneration of both sensory and motor axons. Following a crush lesion of the facial nerve, there was no cell loss in facial nuclei of normal (C57BL/6J) adult mice, but 40% cell loss occurred in Ola mice and the survivors increased in size during the period when functional reinnervation was established. These results are interpreted as a result, first, of prolonged deprivation of target-derived trophic factor in the slowly regenerating Ola motoneurons and second, increased peripheral field size of the survivors. Within the regenerated facial nerve, there was marked heterogeneity of myelinated fibre size in Ola mice. Some Ola axons, both proximal and distal to the lesion site, had areas over twice as great as the largest 6J axons when measured 1 year following injury. A population of small diameter fibres, not observed in 6J nerves, persisted distal to the crush site in Ola nerves, and this was associated with an increase in the total number of myelinated axons in the distal nerve: on average, each parent Ola axon retained three persistent daughter axons. The delayed Wallerian degeneration in Ola mice not only impairs immediate axon regrowth, but also results in a breakdown of the normal mechanisms which regulate axon number and size in regenerating nerve.     

Loss of the Compound Action Potential: an Electrophysiological, Biochemical and Morphological Study of Early Events in Axonal Degeneration in the C57BL/Ola Mouse

In the C57BL/Ola (Ola) mouse strain there is a marked slowing of axonal disintegration during Wallerian degeneration. The locus of the mutation controlling this phenomenon (slow Wallerian degeneration- Wld^s ) has been mapped to chromosome 4, and its protective effect decreases with advancing age. Using biochemical, electrophysiological and histological techniques, the present study was undertaken to determine whether neurofilament phosphorylation and stability are altered or whether calcium-activated proteases are absent in the sciatic nerves of Ola mice. A compound action potential was detectable only when neurofilaments were present and normal axonal architecture was seen. In 1-month-old Ola mice, compound action potentials and neurofilaments were still detectable at 21 days post-transection, whereas both were undetectable by 2 days in BALB/c and C57BL/6J (6J) mice of the same age. Neurofilament levels declined faster with advancing Ola age, confirming previous results, whereas degeneration slowed in ageing BALB/c and 6J mice. In vitro and in vivo degeneration rates were comparable in BALB/c and 6J nerves. Ola nerves, however, showed more rapid decline in vitro than in vivo . Ola and BALB/c nerves frozen and then thawed and incubated in the presence of calcium ions and the ionophore A23187 were not resistant to degradation by intrinsic proteases. Even when a compound action potential could no longer be elicited, however, a majority of nerves still had >50% of myelinated and unmyelinated axons whose electron microscopic profiles appeared normal. Thus, it appears that the first event in Wallerian degeneration in the Ola mouse is a change at the plasma membrane-a transected nerve becomes unable to conduct a compound action potential. Degeneration of the cytoskeleton is a later, separable event.     

Neuroprotection induced by mucosal tolerance is epitope-dependent: conflicting effects in different strains

The ability to cope with ongoing neurodegeneration after injury to the central nervous system of mammals differs among strains and depends in part on the animal's ability to manifest a T-cell-mediated protective response. After CNS injury, strain-related differences were observed. Moreover, the post-injury effect of naturally occurring regulatory CD4+CD25+ T cells was found to differ in different strains. In this study, using partially injured optic nerves of Balb/c/OLA and C57BL/6J mice as models, we observed strain-related differences in the T-cell-mediated protection obtained by antigens administered via the nasal route. Active immunization with myelin-related antigens emulsified in complete Freund's adjuvant had a beneficial effect on both strains, whereas mucosal administration of the same antigens was destructive in mice of the Balb/c/OLA strain but protective in C57BL/6J mice.     

Radiation-induced Reductions in Macrophage Recruitment Have Only Slight Effects on Myelin Degeneration in Sectioned Peripheral Nerves of Mice

Macrophage recruitment into the distal nerve stump of the cut or crushed sciatic or saphenous nerves of C57BL/6J mice was reduced by prior whole body irradiation. This procedure was successful in keeping the numbers of cells stained with the mouse macrophage-specific antibody F4/80 to the levels found in unsectioned nerves. Quantitative image analysis of immunostained sections showed that the rate of loss of myelin basic protein was identical in nerves from irradiated and unirradiated mice up to 5 days but thereafter was slower in macrophage-deprived nerves. Similar analysis of semithin sections stained with toluidine blue detected more undegenerated myelin in the nerves from irradiated mice 10 days after operation. Quantitative counts made from electron micrographs of the sectioned nerves at 7 days also showed slightly less extensive myelin breakdown in the nerves from irradiated mice. Complete removal of myelin from some Schwann cells can occur without macrophages, but macrophages accelerate the removal of myelin in the later stages of Wallerian degeneration. It is concluded that there are two phases to the breakdown of myelin in peripheral nerves undergoing Wallerian degeneration: an initial stage entirely dependent on the activity of Schwann cells and a later stage dependent on both Schwann cells and the presence of macrophages.     

Absence of Wallerian Degeneration does not Hinder Regeneration in Peripheral Nerve

Wallerian degeneration of the distal stump of a severed peripheral nerve involves invasion by myelomonocytic cells, whose presence is necessary for destruction of myelin and for initiating mitosis in Schwann cells (Beuche and Friede, 1984). Degeneration of the distal ends of the axons themselves is assumed to occur by autolytic mechanisms. We describe a strain of mice (C57BL/6/Ola) in which leucocyte invasion is slow and sparse. In these mice, confirming Beuche and Friede, myelin removal is extremely slow. A new finding is that axon degeneration is also very slow. This is a consequence of lack of recruitment of myelomonocytic cells for if such recruitment is prevented in other mouse strains by a monoclonal antibody against the complement type 3 receptor (Rosen and Gordon, 1987) axon degeneration is again slowed. We have also, surprisingly, found that nerve regeneration in the C57BL/6/Ola mice is not impeded by the presence of largely intact axons in the distal stump and absence of recruited cells, myelin debris and the absence of Schwann cell mitosis.     

Traumatic brain injury causes delayed motor and cognitive impairment in a mutant mouse strain known to exhibit delayed wallerian degeneration

Delayed Wallerian degeneration after neuronal injury is a feature of the C57BL/Wld^s mouse mutant. In the present study, we examined the effect of unilateral controlled cortical impact (CCI) on motor and cognitive performance in C57BL/6 and C57BL/Wld^s mice. Performance on a beam-walking task was impaired in both injured groups over the first 3 weeks; however, between 28 and 35 days post injury, C57BL/6 mice continued to improve whereas C57BL/Wld^s mice showed increased footfaults. In a spatial learning task, C57BL/Wld^s animals performed consistently better than C57BL/6 mice when tested 7–10 days and 14–17 days following CCI. C57BL/Wld^s mice also demonstrated improved working memory performance as compared with C57BL/6 mice when trained on days 21–22 after injury; this effect was lost on days 23 and 24, and was not evident in other animals tested in the same task at 28–31 days following injury. These results indicate a marked delay in motor and cognitive impairment following CCI in C57BL/Wld^s mice compared with injured C57BL/6 controls. This is consistent with previous work showing delayed temporal evolution of neuronal degeneration in C57BL/Wld^s mice and suggests CCI may be a suitable model for examining the functional consequences of traumatic brain injury (TBI) in genetically altered mice. J. Neurosci. Res. 53:718–727, 1998. © 1998 Wiley-Liss, Inc.     

The Effectiveness of the Gene Which Slows the Rate of Wallerian Degeneration in C57BL/Ola Mice Declines With Age

The rate of Wallerian degeneration is unusually slow in severed axons of mice of the C57BL/Ola strain. Within mice of that strain we have now found that the rate of degeneration increases with the age of the animal. In 4-week-old mice nerve stimulation evokes muscle contractions even 5 days after sciatic nerve section and compound action potentials can be recorded in the distal nerve stump up to 3 weeks after section. In 1-year-old animals no action potentials can be excited 5 days after nerve section. Heterozygous mice carrying only one copy of the dominant gene show the same age-related decline in viability of the distal nerve stump after axotomy, and the rate of decline is no greater than for homozygous mice. The more rapid rate of degeneration of severed axons of mice of the C57BL/6J strain was affected in the opposite way by age, degeneration occurring more slowly in older animals.     

Poor growth of Mammalian motor and sensory axons into intact proximal nerve stumps

Wallerian degeneration is very slow in the mouse strain now known as C57BL/Ola. Sensory axon regrowth following peripheral nerve lesions is very poor in these animals but motor axons succeed in reinnervating the distal nerve stump even while the majority of severed axons are still intact (Lunn et al., Eur. J. Neurosci., 1, 27 - 33, 1989). To see if motor axons could grow into a completely undegenerated portion of nerve, the proximal stumps of the peroneal and tibial nerves were sutured together in six BALB/c mice and the ability of large motor and sensory fibres from the tibial nerve to grow into the peroneal nerve was examined electrophysiologically in four of them. For the acute experiment the peroneal nerve was cut approximately 7 mm central to the point of suture to the tibial nerve. Both at 2 weeks and 7 weeks after surgery the size of the potential recorded in the ventral roots on stimulating the portion of peroneal nerve into which tibial axons were directed to grow was only approximately 8% of the potential recorded when the tibial nerve was itself stimulated. The potential recorded in the dorsal roots was only approximately 2%. Counts of axon numbers in electron micrographs showed a small but non-significant increase over normal in the number of unmyelinated axons in the peroneal nerves which had been connected to the tibial nerve in this way. It is concluded, in agreement with Langley and Anderson (J. Physiol., 31, 365 - 391, 1904), that axon growth into intact nerves is extremely limited in mammals and that the distal nerve stump of C57BL/Ola mice, although it degenerates very slowly, is not therefore equivalent to an intact peripheral nerve.     

Wallerian degeneration in the optic nerve of the wabbler-lethal (wl/wl) mouse.

Optic nerve pathology was studied in C57BL/6J wabbler-lethal (wl/wl) and control (+/+) mice at postnatal age of 4 weeks (P28). Qualitative light and ultrastructural pathology in wl/wl animals conformed to the criteria of primary axonal (Wallerian) degeneration. Most optic nerve axons in mutant animals appeared normal, as did oligodendroglia, the degree of myelination, the integrity and maturity of vascular elements, astroglia, and most myelin. Still, degenerating axons surrounded by somewhat normal myelin and axons with thickened myelin sheaths were prevalent in wl/wl mice. Dysmyelination or hypomyelination was not evident. At P28, pathology appeared more prominent in large diameter fibers. In the optic nerve of wl/wl mice, axonal degeneration preceded myelin disruption, adding this nerve to other previously reported systems undergoing Wallerian degeneration in this mutant.     

Motor Neuron Death Induced by Axotomy in Neonatal Mice Occurs More Slowly in a Mutant Strain in Which Wallerian Degeneration is Very Slow

In contrast to motor neurons of adults, motor neurons of neonatal mice die if their axons are cut. We have examined the extent, the time course and the loss of susceptibility with age to such induced death in C57BL Wld ^s mice. This is a strain which has a dominant autosomal mutation which dramatically slows the rate of degeneration of axons separated from their cell bodies. Following axotomy in neonatal animals the total number of motor neurons killed is no less in C57BL/ Wdl ^s mice than in two other strains (C57BL/6J/Ola and BALB/c/Ola). Indeed, the susceptibility to axotomy persists to a later age in C57BL/ Wld ^s mice. However, the rate of cell death is significantly slower than in the two other strains; in C57 BL/Wld ^s mice under a week old at the time of sciatic nerve section only-16% of motor neurons have been lost 3 days after axotomy, whereas in the other mice -45–84% of the final loss has by then occurred. This result extends previous work which showed that retrograde degeneration of retinal ganglion cells of adult mice was slower in C57 BL/Wld ^s mice (V. H. Perry et al., Eur. J. Neurosci. , 2 , 408–413, 1991). It is therefore possible that Wallerian degeneration of axons shares features in common with retrograde nerve cell death and that the Wld ^s mutation may throw light on aspects of this complex process.     

Differences in Sympathetic Innervation of Mouse DRG Following Proximal or Distal Nerve Lesions

Nerve injury leads to novel sympathetic innervation of the dorsal root ganglion (DRG). We have hypothesized previously that the degenerating nerve increases the sympathetic sprouting in the DRG and pain after chronic sciatic constriction injury (CCI) by virtue of its influence on sensory and sympathetic axons spared by the injury. However, L5 spinal nerve ligation and transection (SNL) results in the complete isolation of the L5 DRG from the degenerating stump, yet sympathetic axons invade the ganglion, and sympathetically dependent pain develops. We investigated the role of Wallerian degeneration in both sympathetic sprouting and neuropathic pain in these two models of painful peripheral neuropathy by comparing responses of normal C57B1/6J and C57B1/Wld^smice in which degeneration is impaired. After CCI, Wld^smice, unlike 6J mice, did not develop thermal or mechanoallodynia or sympathetic innervation of the L5 DRG. After SNL, both strains developed mechanoallodynia and sympathetic sprouts in L5, but only 6J mice developed thermal allodynia. Observation of the origins of the invading sympathetic axons revealed that after CCI, sympathetics innervating blood vessels and dura (probably intact) sprouted into the ganglion, but after SNL sympathetics (probably axotomized) invaded from the injured spinal nerve. Based on these findings, we hypothesize that there are two mechanisms for sympathetic sprouting into DRG, differentially dependent on Wallerian degeneration. Analysis of pain behavior in these animals reveals that (i) mechanoallodynia and sympathetic innervation of the DRG tend to coincide and (ii) thermal allodynia and Wallerian degeneration, but not sympathetic innervation of the DRG tend to coincide.