Local Influences of Myelin on Axon Cytoskeleton

Traditionally, changes in synthesis of cytoskeletal proteins, connection with suitable targets, patterns of activity, and availability of target-derived neurotrophic factors were thought to regulate axon diameter. From this viewpoint, the number of MTs and NFs (Friede and Samorajski, 1970) supplied by slow axonal transport (Hoffman et al., 1988, 1985, 1983; Lasek et al., 1983; Wujek et al., 1986) were believed to be the primary determinants of axon diameter. Synthesis of cytoskeletal and membrane proteins in neurons was presumably affected by contact with an appropriate target, largely by increasing the return of neurotrophins to the neuronal perikaryon by retrograde axonal transport ensuring neuronal survival and differentiation (Burek and Oppenheim, 1996; Thoenen, 2000). Neurotrophin uptake and retrograde transport by neurons could be enhanced by increased neuronal activity, enhancing neuronal plasticity (Thoenen, 2000). Glial cell influences on neuronal growth were thought to act primarily through mechanisms like glial production of neurotrophins (Althaus and RichterLandsberg, 2000). However, release of neurotrophins in the adult nervous system by MG was thought to be modest compared to release by targets, although it could be upregulated in some cases by nerve injury (Heumann et al., 1987) or disease (Burbach et al., 2004; Heese et al., 1998).

In fact, myelination by either SC or OL has profound effects on the local axon. The absence of myelin, whether due to demyelination or failure to form myelin, leads to characteristic alterations in the organization, biochemistry, and composition of the axonal cytoskeleton. Such changes in the neuronal cytoskeleton have a dramatic impact on neuronal function. Studies with myelin mutant mouse models, Trembler and Shiverer, provided the initial evidence for this influence.

Trembler mice have a point mutation in a gene (PMP22) expressed in SC, but not OLs (Suter and Snipes, 1995; Suter et al., 1992b). As a result, Trembler mice have a phenotype that includes demyelination of large caliber axons in the PNS with no apparent effect on myelination of CNS axons (Low and McLeod, 1975). Affected PNS SC exhibit continuing cycles of myelination and demyelination that get progressively worse, eventually resulting in a high percentage of large axons with little or no compact myelin (Low, 1976a, 1976b). In Trembler, PMP22 is a dominant mutation with several alleles (Suter et al., 1992a, 1992b; Suter and Snipes, 1995). Although specific functions of PMP22 in myelina-tion are poorly understood, a variety of human peripheral neuropathies have defects in PMP22 or in levels of its expression. For example, missense mutations lead to hypomyelinating phenotypes (Trembler, Dejerine-Sottas syndrome, and some cases of CMT1a), whereas deletions in PMP22 are implicated in hereditary neuropathy with liability to pressure palsies and PMP22 gene duplication gives rise to most cases of CMT1a in humans.

PMP22 is an SC integral membrane tetraspan protein, but most PMP22 is degraded without insertion into the plasma membrane. Interaction with an axon during myelination increases PMP22 synthesis and promotes insertion of glycosylated protein into the plasma membrane (Pareek et al., 1997). One effect of point mutations is to reduce PMP22 insertion into plasma membrane (Colby et al., 2000). The sensitivity of PNS myelin to both the presence of PMP22 and levels expressed on the myelin surface suggests that PMP22 functions in signaling to neurons. That disruption of PMP22 function or distribution leads to a neuropathy suggests a neuronal target.

Electron microscopic images of Trembler peripheral nerves in cross-section typically show many axons undergoing active demyelination or lacking myelin. Axons were smaller and thinner than normal (Ayers and Anderson, 1976). Remarkably, this reduction in axonal caliber was a local phenomenon (Aguayo et al., 1977; de Waegh and Brady, 1991; de Waegh et al., 1992). When sciatic nerve segments from Trembler mice are grafted into wild-type sciatic nerve and axons are allowed to regenerate, axon regions in the Trembler graft have a Trembler phenotype, demyelinated axons, and reduced axonal caliber in graft regions (Aguayo et al., 1977; de Waegh and Brady, 1991). In these nerves, axons proximal and distal to the graft had normal compact myelin and axon caliber. The reverse graft experiment (normal sciatic nerve segment grafted in Trembler sciatic nerve) had normal myelin and axonal caliber in the graft, but Trembler phenotype proximal and distal to the graft (Aguayo et al., 1977; de Waegh and Brady, 1991).

Axonal parameters that are altered locally included rates of slow axonal transport (de Waegh and Brady, 1990; de Waegh et al., 1992), phosphorylation of neurofilament and microtubule proteins (de Waegh et al., 1992; Kirkpatrick and Brady, 1994), and neurofilament density (de Waegh and Brady, 1991; de Waegh et al., 1992). Neurofilament phos-phorylation appeared to be the primary determinant of axonal caliber in these models (de Waegh et al., 1992), and the loss of compact myelin in Trembler PNS axons led to a significant reduction in neurofilament phosphorylation and consequential increase in neurofilament densities (Fig. 1). Two hypomyelinating transgenic mice with different severity of demyelination showed that the effects on neurofilament phosphorylation and density scaled with the severity of the demyelination (Cole et al., 1994).

The influence of myelin sheaths on organization of the axonal cytoskeleton begins to explain how SC myelin can alter the functional architecture of the axon. For example, myelin sheaths formed by SCs specifically influence phos-phorylation of axonal NFs (Cole et al., 1994; de Waegh et al., 1992; Hsieh et al., 1994). Neurofilament phosphorylation was sensitive to myelination levels both in vivo and in culture (Cole et al., 1994; Starr et al., 1996). Remarkably, short gaps in compact myelin such as those found at nodes of Ranvier were sufficient to change neurofilament density (Figure 1B) and phosphorylation (de Waegh, 1990; Hsieh et al., 1994; Mata et al., 1992). Changes in phosphorylation of axonal NFs thus appear to be responsible for the characteristic structure of larger caliber axons at PNS and CNS nodes of Ranvier (Brady, 1993; de Waegh et al., 1992), with reduced caliber and increased NF density (Berthold, 1982; Scherer et al., 2004).

CNS myelin also affects local properties of the axon, with some differences in the effects. Not only fine structure (Peters et al., 1991; Raine, 1984) but also protein and lipid composition (Morell and Quarles, 1999) differ between CNS and PNS myelin. Moreover, some functions of SC in the PNS may be split between OL and astrocytes in the CNS. Shiverer is a recessive mutation with a more severe pheno-

type than Trembler. Shiverer mutant mice lack compact myelin in the CNS and develop a severe tremor in early postnatal development (Readhead and Hood, 1990). They have a mean life span of 100 to 120 d (Chernoff, 1981). The absence of compact myelin in the CNS is due to deletion of coding regions after the first exon in the myelin basic protein (MBP) gene (Roach et al., 1985). MBP is a major structural protein of CNS myelin, but a minor component of PNS myelin (Campagnoni and Macklin, 1988). Two PNS-spe-cific proteins, P0 and P2, appear to have functional overlap with MBP, so PNS myelination is near normal (Rosenbluth, 1980a, 1980b), albeit with some structural differences (Gould et al., 1995).

Whereas PNS axons in Trembler mice are subject to a constant cycle of myelination and demyelination, the absence of MBP in Shiverer means that CNS axons never see normal myelin. One advantage of studying Shiverer mouse as an animal model is that transgenic mice are available that express different levels of MBP (Popko et al., 1987). Studies with mouse strains having different levels of MBP expression indicate that MBP levels limit myelin sheath thickness (Shine et al., 1992) with an minimum of 50% of wild-type MBP required for normal thickness in CNS myelin sheaths. Mice homozygous for an MBP transgene expressed only 25% of wild-type MBP levels, resulting in a thin compact myelin with relatively few lamellae. However, this thin myelin sheath was sufficient to suppress tremors and increase life spans to near normal (Readhead et al., 1987). Availability of Shiverer and transgenic Shiverer mice homozygous for the MBP transgene (MBP/MBP) permitted evaluation of how myelin sheath thicknesses affect CNS axons (Brady et al., 1999).

Both CNS axons lacking myelin (Shiverer) and CNS axons with abnormally thin myelin sheaths (transgenic Shiverer: TG Shiverer) exhibit changes in local properties relative to CNS

Myelin Electron Microscopic Image

Figure I Local influences of myelination on the axon cytoskeleton. The use of mutant mouse strains such as Trembler show that the presence of compact myelin has a dramatic effect on the organization of axonal neurofilaments and axonal caliber. Contrast the number and density of neurofilaments in wild-type (A) and Trembler (C) axons in cross-section. The increased density of neurofilaments is accompanied by a reduction in axonal caliber and neurofilament phosphorylation (de Waegh et al., 1992). At a node of Ranvier (B), there is a gap in the myelin sheath that leads to a similar dephosphorylation of axonal neurofilaments and reduction in axonal caliber (ms: myelin sheath; sc: Schwann cell cytoplasm; mv: Schwann cell microvilli at node of Ranvier). (Adapted from Kirkpatrick and Brady, 1999.) (Micrographs by S. de Waegh and S. Brady.)

Figure I Local influences of myelination on the axon cytoskeleton. The use of mutant mouse strains such as Trembler show that the presence of compact myelin has a dramatic effect on the organization of axonal neurofilaments and axonal caliber. Contrast the number and density of neurofilaments in wild-type (A) and Trembler (C) axons in cross-section. The increased density of neurofilaments is accompanied by a reduction in axonal caliber and neurofilament phosphorylation (de Waegh et al., 1992). At a node of Ranvier (B), there is a gap in the myelin sheath that leads to a similar dephosphorylation of axonal neurofilaments and reduction in axonal caliber (ms: myelin sheath; sc: Schwann cell cytoplasm; mv: Schwann cell microvilli at node of Ranvier). (Adapted from Kirkpatrick and Brady, 1999.) (Micrographs by S. de Waegh and S. Brady.)

axons with normal myelin sheaths (wild-type). These differences are analogous in some, but not all, respects to those seen in Trembler PNS axons. Relative to wild-type, axonal diameters of optic axons were reduced in Shiverer mice; NF densities were increased in Shiverer; and NFH phosphoryla-tion was reduced (Brady et al., 1999; Kirkpatrick et al., 2001). However, changes in these parameters were less pronounced than corresponding changes in Trembler PNS nerve (de Waegh et al., 1992). In TG Shiverer mice with 25% of normal MBP and fewer layers of compact myelin than wild-type, these same parameters were intermediate between Shiverer and wild-type (Brady et al., 1999). In adult myelinated axons, Na(v) 1.2 channels are restricted to unmyelinated zones like the axon hillock and Na(v) 1.6 channels are restricted to the nodes of Ranvier (Boiko et al., 2001). Consistent with this pattern, Na(v) 1.2 channels are abundant on nonmyelinated fibers and appear first in development, but Na(v) 1.6 channels do not appear in significant amounts until myelin forms. Targeting of sodium channel isoforms is altered in Shiverer axons (Boiko et al., 2001) with a distribution similar to that seen in nonmyelinated axons (i.e., Na(v)1.2) all along the axons and very little Na(v) 1.6 is detected.

Other local axonal parameters were the same in TG Shiverer and wild-type optic axons. For example, slow axonal transport rates were increased in Shiverer, but were indistinguishable in TG Shiverer and wild-type nerves (Brady et al., 1999). These observations indicate that OL influence local axonal parameters during development and that the full effect of myelination is not expressed until the thickness of the myelin sheath approaches wild-type levels.

Changes in other myelin-related proteins also affect local axonal parameters. For example, absence of either PMP22 or connexin 32 (another SC protein) leads to specific changes in axonal properties and architecture (Neuberg et al., 1999). Both of these knockouts allow formation of compact myelin that is subsequently removed and exhibit axonal atrophy. Notably, there are also changes in the distribution of a nodal Kv-channel in these mice (Neuberg et al., 1999). Connexin 32 mutant mice also have changes in the axonal cytoskeleton similar to those seen in Trembler along with indications of altered axonal transport (Sahenk and Chen, 1998).

Mice with a null mutation in proteolipid protein (PLP) exhibit axonal swellings and degeneration consistent with a local effect on fast axonal transport (Griffiths et al., 1998). A detailed study of axons in PLP null mice (Edgar et al., 2004) showed significant accumulations of axonal vesicles and alterations in the molecular motor protein cytoplasmic dynein. This action of PLP-deficient myelin was local because similar phenotypes were seen in axonal segments surrounded by PLP-null OL in a Shiverer background (Edgar et al., 2004).

Mice lacking myelin-associated glycoprotein (MAG) form compact myelin sheaths, but these sheaths are not closely apposed to the axonal membrane (Yin et al., 1998).

With increasing age, the axons in MAG knockout mice develop significantly smaller calibers and reduced neurofilament phosphorylation. Another indication that MAG affects neuronal properties is found in the observation that MAG is also an important modulator of neurite growth (Filbin, 1995). A similar phenotype is observed in mice that lack the ability to make complex gangliosides (Sheikh et al., 1999), raising the possibility that complex gangliosides are ligands for MAG (Vyas et al., 2002). Whether another MAG ligand, Nogo receptor (Barton et al., 2003; Liu et al., 2002), influences axonal properties is not yet known. Studies on mice lacking both MAG and the complex gangliosides (Marcus et al., 2002) suggest that these two components interact to stabilize glial-axonal interactions. Developmental studies (Yin et al., 1998) suggest that MAG-based signaling is not solely responsible for local alteration of axonal parameters, but interactions between myelin and axons mediated by MAG influence phosphorylation of axonal cytoskeletal proteins and axonal caliber (Dashiell et al., 2002).

Local axonal parameters of both CNS and PNS fibers are affected by dysmyelination or demyelination, but the changes are not identical. What they share is an increase in neurofilament density coupled to a reduction in both neurofilament phosphorylation and axonal caliber. The stability of axonal microtubules is also affected in both mouse models, but the composition of axonal microtubules is affected differentially (Kirkpatrick and Brady, 1994; Kirkpatrick et al., 2001). Slow axonal transport is also affected by lack of myelin in both CNS and PNS, but is increased over wildtype in Shiverer axons that have never seen myelin and slower in Trembler PNS axons that have been demyelinated (Brady et al., 1999; de Waegh and Brady, 1988, 1990; de Waegh et al., 1992). These effects all result from local effects of MG on axons, but equally striking are differences in neuronal properties that result from changes in gene expression in the neuron.

Peripheral Neuropathy Natural Treatment Options

Peripheral Neuropathy Natural Treatment Options

This guide will help millions of people understand this condition so that they can take control of their lives and make informed decisions. The ebook covers information on a vast number of different types of neuropathy. In addition, it will be a useful resource for their families, caregivers, and health care providers.

Get My Free Ebook


Post a comment