Mechanisms and Patterns of Axonal Loss in Multiple Sclerosis

After a variable or silent initial course, MS patients often enter a progressive neurological decline leading to permanent clinical disability. Axonal loss is considered the major factor behind this process. Elucidating the etiology and pathogenesis of this phenomenon is critical to developing treatments for progressive MS. Here we discuss pathologic presentation of axonal loss in MS, patterns of axonal loss at certain disease stages, and evidence that axonal loss in MS progresses independently of clinical relapses but coincides strongly with inflammation. Mechanisms behind axonal loss in MS include T cell attack, reactive nitrogen species-induced damage, loss of myelin (demyelination), and loss of trophic support from oligodendrocytes. Also, we will review animal models of neurodegeneration that share similar disease features with MS axonopathy. Finally, we will consider the progress of therapeutic strategies aimed at axonal preservation in MS including remyelination enhancement, blockage of Na+ channels and prevention of free radical formation.


Introduction
After a variable or silent initial course, MS patients often enter a progressive neurological decline leading to permanent clinical disability.Axonal loss is considered the major factor behind this process.Elucidating the etiology and pathogenesis of this phenomenon is critical to developing treatments for progressive MS.Here we discuss pathologic presentation of axonal loss in MS, patterns of axonal loss at certain disease stages, and evidence that axonal loss in MS progresses independently of clinical relapses but coincides strongly with inflammation.Mechanisms behind axonal loss in MS include T cell attack, reactive nitrogen species-induced damage, loss of myelin (demyelination), and loss of trophic support from oligodendrocytes.Also, we will review animal models of neurodegeneration that share similar disease features with MS axonopathy.Finally, we will consider the progress of therapeutic strategies aimed at axonal preservation in MS including remyelination enhancement, blockage of Na+ channels and prevention of free radical formation.

Clinical overview of MS
In about 85% of patients, MS follows a biphasic course.In the first phase or relapsingremitting multiple sclerosis (RRMS), patients suffer transient bouts of neurological deficit caused by neuroinflammation followed by rapid recovery (Noseworthy et al., 2000).The pathologic hallmark of RRMS attacks is inflammatory demyelinated foci or lesions in the white and gray matter of the central nervous system visible on MRI.The clinical manifestations of RRMS attacks are variable from patient to patient, and may include paraesthesia (numbness), diplopia (double vision), scotoma (visual anomalies), sensory and motor disorders of limbs and cerebellar incoordination (lack of balance).In many cases, after a variable amount of time (approximately 10-20 years) patients will enter a progressive phase of the disease This more severe second stage is characterized by an irreversible neurological decline culminating in severe disability.Once this second phase is identified, patients exhibiting this biphasic couse are diagnosed with secondary progressive multiple sclerosis (SPMS).In some cases, patients will exhibit only a progressive phase of disease, which represents primary progressive multiple sclerosis (PPMS).These patients have symptomatic onset of the progressive form of the disease with no prior history of neuroinflammatory attacks.

Histology of axonal pathology 4.1 Axonal injury
Axonal injury is defined as pathological changes in the cytoarchitecture of an axon happening within a short time window before axonal death (Gentleman et al., 1993).Because debris of degenerating axons is rapidly cleared by CNS scavenger cells, quantification of axonal injury can often underestimate axonal loss accrued over long stretches of time (Trapp and Nave, 2008).Disturbance of fast axonal transport is the most accurate and sensitive method of determining axonal injury (Cochran et al., 1991;Gentleman et al., 1993;Ohgami et al., 1992, Sherriff et al., 1994a).Due to dystrophic changes, fast axonal transport mechanisms are unable to efficiently move cellular substances along the length of the axon.These substances accumulate as focal concentrations (spheroids), continuously and discontinuously, at certain points along the length of the axon (Figure 1).Axonal injury ranges from minor changes in the axoplasmic membrane to severe distentions as a result of disrupted fast axonal transport (Ferguson et al., 1997).Axonal dystrophy has caused these accumulations by interfering with fast axonal transport (from Ferguson et al., 1997).

Axonal transection
Axonal transection in MS has been a main research focus since its discovery in the late 1990s (Trapp et al., 1998;Lovas et al., 2000;Peterson et al., 2001).If axonal dystrophy exceeds a certain threshold, the axon cytoskeleton may dissolve resulting in a characteristic bulblike formation known as a transection (Figure 2).Axonal transections are also referred to as axonal end-bulbs, terminal ovoids or spheroids, or terminal swellings.Axonal end-bulbs are morphologically ovoid in shape.In acute MS white matter lesions, swellings average 40-80 µm in diameter (Trapp et al., 1998).They are typically smaller in chronic lesions.Transections are positively identified by confirming the presence of only a single axonal connection to the swelling.Swellings that have two axonal connections represent intracellular protein accumulations that are pathological but not indicative of breakage (Ferguson etl al. 1997).Immunohistochemistry using anti-amyloid precursor protein (APP) antibodies is a useful method for identifying axonal transections (Sherriff et al., 1994).APP and other fast axonal transport proteins accumulate in the end-bulb.Staining for nonphosphorylated neurofilament also detects ovoid formation (Sternberger et al., 1983).Although demyelination is a key etiological factor in transection, end-bulbs may or may not be encircled with myelin (Trapp et al., 1998;Bjartmar et al., 2001;Aboul-Enein et al., 2006).

Molecular markers of axonal pathology
Immunocytochemistry staining of amyloid precursor protein (APP) or non-phosphorylated neurofilament is commonly used to identify axonal injury and axonal transection.Staining of these factors is often used in conjunction with antibodies to myelin antigens (i.e.myelin basic protein) in order to qualitatively determine the degree of demyelination coincident with axonal damage.Although less sensitive and accurate than immunocytological methods, Bielchowsky silver stain is a classic method still used to identify the general character of nerve fibers.

Bielchowsky silver stain
Before the era of immunocytochemistry, neurological studies relied on silver stains to visualize axonal pathology.Developed by Max Bielchowsky, the technique is a modified version of Santiago Ramón y Cajal's method of impregnating nerve fibers with silver.The method became known as the Bielchowsky silver stain.The classic technique is still used today to describe nerve fibers.Bielchowsky staining is particularly useful in identifying general axon morphology, such as axonal caliber and Wallerian degeneration (Galeano and Vav Ferriera, 1957;Wakita et al., 2002).However, more sensitive laboratory techniques are needed in order to detect axonal injury and transection.

Nuclear magnetic resonance (NMR) spectroscopy
NMR spectroscopy is used to non-invasively measure chemical-pathologic changes in distinct brain slices of MS patients.Most notably, NMR studies have demonstrated whole brain reductions in the neuro-axonal marker N-acetylaspartate (NAA) in MS patients compared with normal controls (Gonen et al., 2000).Because differences in NAA levels compared to controls are greater amongst older than younger MS patients, this neuronal marker is useful in charting the course of neurodegeneration in the disease (Gonen et al., 2000).Reduced levels of NAA are detectable in inflammatory foci early on in the relapsing/remitting stage of the disease (Matthews et al., 1996).Moreover, decreased NAA in the normal-appearing white matter (NAWM) of relapsing/remitting and progressive patients suggests that the neurodegenerative process is not confined to inflammatory lesions (Tourbah et al., 1996).The ability to precisely localize changes in compounds such as NAA has made serial NMR of distinct brain slices a useful tool in human clinical trials looking at drug efficacy (Destefano et al., 2007).Serial NMR has also been effective in tracking pathologic changes in animal models of MS, such as experimental autoimmune encephalomyelitis (EAE) (Stewart et al., 1991).

Amyloid precursor protein (APP)
Detection of amyloid precursor protein (APP) positive foci within axons indicates axonal injury in MS lesions (Gehrmann et al., 1995;Ferguson et al. 1997).This method derives from studies on axonal damage resulting from traumatic head injury (Gentleman et al.).APP is an integral membrane protein expressed primarily in neuronal synapses.Although its function is unknown, APP is thought to be associated with the endosomal/lysosomal system, synapse formation, neural plasticity, and iron transport (Haass et al., 1992;Ferreira et al., 1993).Normally, APP is transported by fast axonal transport along the axon (Koo et al., 1990) www.intechopen.comand is undetectable by immunocytochemical methods.However, disruptions in the integrity of axonal structure inhibit the smooth transport of APP.Subsequently, APP builds up at certain points along the axon.Axonal damage appears microscopically as distinct, elongated APP+ blots following the course of the axon, or a succession of discontinuous APP+ punctate dots that look as if they are lodged within the axon (Figure 1).Immunocytochemistry for APP is a useful method of detecting axonal injury in human and animal nerve tissue preparations (Sherriff et al., 1994b;Herrero-Herranz et al, 2008).Staining of other fast axonal transport substances is often used to determine or verify axonal injury, including synaptic vesicle terminal or synaptophisin protein (SPY) (Frischer et al. 2009).

Non-phosphorylated neurofilament
Detection of non-phosphorylated neurofilaments along axons or collected at the site of axonal transections is another efficient method of identifying axonal damage in MS (Sternberger et al., 1983;Trapp et al;1998).Neurofilament is the principal structural constituent of the axon, accounting for its shape and integrity.Phosphorylation of neurofilament ensures proper extension of neurofilament side arms (Dutta et al., 2006).This process enhances axon diameter, neurofilament spacing, and conduction velocity of the axon.Normally, axons are insulated by myelin, a multilayered lipid sheath which enhances the conduction velocity of nerve impulses and protects the axonal cell membrane from the extracellular environment.Demyelination exposes the axolemma to phosphatases which dephosphorylate neurofilament fibers and subsequently induce structural damage.Dephosphorylated neurofilament fibers show fragmentation and decreased spacing between fibers due to reduced side-arm extensions (Dutta et al., 2006).Demyelination/dysmyelination has also been shown to reduce axonal caliber and number of phosphorylated neurofilaments in animal models, as well as increase the number of nonphosphorylated neurofilament epitopes (deWaegh et al., 1992, Hsieh S-T et al. 1993).Doublelabel immunohistochemistry for non-phosphorylated neurofilament and myelin antigens is an effective method for identifying axonal pathology concomitant with demyelination (Trapp et al., 1998).In addition, axonal spheroids have been shown to contain many other substances (Table 1).Moreover, axonal spheroids are detectable in a number of other neurological conditions using immunocytochemistry for specific substances (Table 1).

Patterns of axonal pathology in MS
Although clinically axonal loss is being revealed as a pathological process independent of acute relapses (Kremenchutzky et al., 1996), inflammation and neurodegenerative processes interface to accelerate neuronal death in MS (Frischer et al., 2009).A comprehensive picture of the pathological changes taking place in axons in MS considers CNS lesions and normal brain and spinal cord tissue.Pathology in MS axons include dystrophic changes (i.e.thinning), non-terminal swellings (intracellular accumulations), intracellular vacuolization (Kornek et al., 2001;Evangelou et al., 2005;Miller and Leary, 2007) and end-bulb formation (transection).Distinct patterns of axonal pathology are present in different histopathogenetic stages of MS lesions (Table 2, Han et al., 2008;Frischer et al., 2009).Active and chronic MS white matter lesions exhibit axonal injury profiles corresponding with the degree of inflammation (Bitsch et al., 2000;Herrero-Herranz, 2008;Frischer et al., 2009).

Same as active lesion but cell associations found mostly in lesion margins
Comparable to agedmatched normal appearing white matter controls from non-MS brains Table 2. Axonal pathology within different MS white matter lesion subtypes.

Active lesions
Active white matter MS lesions are lesions undergoing demyelination with heavy infiltration of HLA-II reactive macrophages, microglia and leukocytes.In short, these lesions are white matter areas undergoing a hostile immune reaction.In active lesions, reactive CD8+ cytotoxic T cells and macrophages attack the myelin sheath of axons.Microglia and macrophages are commonly found surrounding degenerating axons (Trapp et al., 1998;Barnett and Prineas, 2004).Macrophages phagocytose and clear myelin debris resultant from the destruction of myelin (Prineas and Wright, 1978;Ferguson et al., 1997;Barnett and Prineas, 2004).
Active lesions display numerous small-caliber axons and relatively large transections throughout their entirety (Frischer et al., 2009).The neurofilament fibers of dystrophic axons and swellings are commonly non-phosphorylated.Axonal spheroids or ovoids are usually positive for both non-phosphorylated neurofilament and amyloid precursor protein (APP), although ovoids positive for non-phosphorylated neurofilament positive and negative for APP have been reported (Ferguson et al., 1997).Thin axons and transections may or may not be ensheathed with myelin (Trapp et al., 1998;Bjartmar et al., 2001), although contiguous demyelination is a common feature.Lengths of non-phosphorylated axons with intact myelin occasionally extend outside of the lesion zone (also observed in chronic lesions) (Trapp et al., 1998).The degree of axonal pathology (i.e.number of ovoids, concentration of APP) correlates with the amount of inflammatory infiltrates within the lesion (Frischer et al., 2009).

Chronic active lesions
Chronic active lesions are punched-out, demyelinated lesions with inflammation and active demyelination occurring mainly at their margins.Inflammation concentrated in the margins of chronic lesions is comparable to that in the cores of active lesions.For this reason they have been called "expanding lesions" (Frischer et al., 2009).In their cores, chronic lesions contain fewer macrophages and T cells than active lesions.Inflammation in the cores of chronic active lesions is reduced compared to active lesions, and often non-existent.HLA-II reactive microglia and astrocytes are numerous in the borders of chronic lesions where inflammation and demyelination are prevalent.
Chronic lesions display fewer and smaller-diameter axonal swellings in their cores and margins than active lesions (Ferguson et al., 1997).In chronic lesions, axonal ovoids are most numerous in the margins where inflammation is concentrated, further emphasizing the relationship between axonal injury and the inflammatory reaction.Likewise, the actively demyelinating margins of chronic lesions display far more axonal ovoids than their cores (Trapp et al., 1998).However, since it is estimated that chronically demyelinated lesions show 60-70% loss of axons (Mews et al., 1998), the lack of core axonal injury may reflect axonal destruction and loss.Axonal associations with leukocytes and monocytes are similar to those seen in active lesions, although these cell-to-cell interactions are concentrated at the margins where inflammation is most active.

Chronic inactive lesions
Chronic inactive lesions are punched-out, demyelinated lesions with no detectable inflammatory or actively demyelinating processes.Chronic inactive lesions display severe tissue vacuolization and diffuse astrocytosis.They contain few HLA-II reactive cells and are generally hypocellular (Frischer et al., 2009).
Chronic inactive lesions contain few axonal spheroids in their cores and margins.Axons within chronic inactive lesions appear abnormally thicker than those within chronic active lesions, possibly due to axonal swelling associated with chronic disease processes (Shintaku et al., 1988).The degree of axonal injury and inflammation in chronic inactive lesions is equal to that seen in normal appearing white matter from aged-matched, non-MS controls (Frischer et al., 2009).This suggests that in these embattled lesions the disease process has died out.

Normal appearing white matter (NAWM): Relapsing/remitting and progressive patients
Axonal injury in the NAWM of progressive patients is more substantive compared to relapsing/remitting patients (Frischer et al., 2009).NAWM of progressive patients is characterized by diffuse inflammation, axonal injury and microglia activation.Diffuse white matter injury can be profound in these patients even with a relatively minor focal lesion load (Kutzeinigg et al., 2005).Similarly, focal demyelinating lesions show only a marginal correlation with inflammation in NAWM and cortical regions (Kutzeinigg et al., 2005).This supports the idea that the formation of demyelinated foci occurs independently of inflammatory damage taking place in normal brain regions.However, pathologic changes occurring in the NAWM do so against a background of diffuse inflammation (Kutzeinigg et al., 2005).This suggests that with disease chronicity inflammatory cells escape the control of the peripheral immune system and slowly accumulate in the whole brain.Over time, these infiltrates may inflict massive, cumulative white matter damage on a global scale, which manifests clinically as the progressive phase of the disease (Kutzeinigg et al., 2005).

Axonal pathology in relapsing/remitting and progressive MS
Up until now, we have considered axonal pathology in lesions at different pathogenic stages of MS.Here, we will discuss axonal pathology as it occurs in patients with relapsing/remitting, primary progressive, and secondary progressive forms of MS.
In all MS subtypes (relapsing/remitting, primary and secondary progressive), the amount of axonal injury in any demyelinated lesion is tightly correlated with the degree of inflammation present in the lesion (Frischer et al., 2009).This supports the idea that an inflammatory reaction is chiefly responsible for axonal injury in demyelinated zones.In line with this idea, active lesions show the highest density of axonal injury (Trapp et al., 1998).Relapsing/remitting MS patient possess a higher number of active lesions than patients of any other disease phase.Because of this, relapsing/remitting patients have a higher degree of axonal injury than patients with progressive forms of the disease (Frischer et al., 2009).
In spite of this, lesions at the same level of activity exhibit no significant differences in axonal injury between relapsing/remitting, primary progressive, and secondary progressive MS patients.That is, an active lesion from a relapsing/remitting patient contains the same extent of axonal injury as an active lesion from a progressive patient.The one important exception is with normal appearing white matter (NAWM) which shows a greater extent of axonal injury in progressive patients than in relapsing/remitting patients (Frischer et al., 2009).As discussed above, one explanation for this is that in progressive patients inflammatory cells accumulate in the whole brain and inflict massive white matter damage, whereas in relapsing/remitting patients axonal injury is limited to inflammatory demyelinating foci (Kutzeinigg et al., 2005).In summary, relapsing/remitting patients tend to exhibit high axonal injury associated with inflammatory demyelinating foci, whereas progressive patients tend to exhibit diffuse white matter damage that may be secondary to global inflammation.
Lesions from patients with progressive forms of MS (primary or secondary progressive) exhibit highly variable levels of axonal injury.However, the presence or absence of ongoing inflammation in progressive MS brains correlates with the extent of axonal damage.In order to demonstrate this, Frischer et al. classified progressive patients in a large post-mortem histological study as either pathologically active or pathologically inactive.Brains of pathologically active patients revealed the persistence of active or chronic active lesion activity (lesions with inflammatory activity).Brains of pathologically inactive patients contained inactive lesions only.Frischer et al. found that pathologically inactive patients lived significantly longer than pathologically active patients (mean 76 yrs., range 64-84 vs. mean 53 yrs., range 28-82).Chronic inactive lesions in the pathologically inactive group contained axonal injury equal to that of aged-match controls.Conversely, chronic inactive lesions in pathologically active patients exhibited significantly more axonal injury than those of the pathologically inactive group.This data suggests that in progressive MS axonal injury is intrinsically tied to an ongoing inflammatory reaction.In a subset of MS patients with significantly reduced axonal injury, the inflammatory reaction appears to die down to the level of aged-matched controls (Frischer et al., 2009).

Axonal loss as a distant pathological process
Axonal loss is considered the primary cause of permanent clinical disability in patients with MS.The fact that axonal loss may approach 30-40% (Ganter et al., 1999) in the spinal cord and possibly as high as 60-70% in lesioned white matter tissue (Mews et al., 1998) demonstrates the severity of neurodegeneration in chronic disease pathogenesis.Evidence that the neurodegenerative component ensues independently of inflammatory disease derives from clinical data.These data include irresponsiveness of progressive MS to immunomodulatory therapies and inconclusive reports that pharmacological treatment of inflammatory bouts lessens morbidity and delays time of onset of progressive forms of the disease (Fisher et al., 2008;Fisniku et al., 2009).
Striking evidence that uncouples acute relapses and neurodegeneration comes from casecontrol studies examining the time of onset of progressive disability across different types of MS.In a poignant retrospective study using large cohorts of over 1000 patients per arm, Kremechutzky et al. showed that relapsing/remitting, primary progressive, and secondary progressive patients began irreversible neurological decline at about the same time (~2.6 years mean difference between time of onset).The clinical disability measured by the Expanded Disability Status Scale was similar across these patient groups and irrespective of preceding or subsequent relapse events.
Studies of the histopathogenesis of MS also indicate axonal loss as a distinct disease mechanism.Identification of degenerating nerve fibers outside inflammatory foci in the normal appearing white matter (NAWM) of post-mortem MS brains shows that the neurodegenerative process can occur independently -at least topgraphically-of inflammation and demyelination (Bjartmar et al., 2001).However, there is a strong correlation between the degree of axonal injury and inflammation in all MS lesion stages and normal appearing white matter (Bitsch et al., 2000;Herrero-Herranz, 2008;Frischer et al., 2009).Moreover, progressive patients exhibit extensive neurodegeneration against a background of massive inflammation and glial activation (Kutzeinigg et al., 2005).Although it is unclear precisely how the extent of inflammation definitely relates to axonal loss in MS.In the same way, the extent of axonal injury and the autoimmune inflammatory process concomitantly diminish to the level of aged-matched controls as the disease progresses (Frischer et al., 2009).

Mechanisms of axonal injury
Extensive research has focused on determining the mechanisms responsible for axonal loss and findings have identified several pathogenic phenomena that might be responsible.Some mechanisms are thought to initiate other pathological events downstream, ultimately leading to the demise of the axon.Initiating factors include inflammatory and immunological attack of axons, exposure of demyelinated axons to the excitotoxic MS lesion environment, and loss of trophic support by oligodendrocytes and other resident CNS cells affected by the inflammatory milieu.These initiating factors are capable of setting up an energy imbalance in the axon, leading to a state of virtual hypoxia.The result is the eventual dissolution of the axonal cytoskeleton at the site of injury.Axonal damage can also cause a type of degradation distal to the site of injury referred to as Wallerian degeneration.

CD8+ cytotoxic T lymphocytes
Major histocompatibility complex (MHC) class I-restricted CD8+ cytotoxic T lymphocytes (CTLs) are thought to directly attack axons and dendrites (neurites) and neuronal soma in MS lesions.These cells outnumber class-I restricted CD4+ T cells approximately 10 fold in actively demyelinating MS white matter (Booss J et al., 1983;Gay FW et al., 1997).Interleukin-2 (IL-2), a cytokine released by CD4+ T helper cells, along with MHC class-Iantigen recognition, initiates differentiation of CD8+ T cells into effector CTLs.Because IL-2 concentrations are increased in the serum and CSF of MS patients (Gallo et al., 1992), it is likely that in lesions CD8+ T cells differentiate into effector CTLs by recognizing MHC class I-peptide on CNS cells and receiving the IL-2 second signal.Because the CD8+ CTLs constitute a clonally expanded population, these effector cells are highly specified and reactive against a single antigen.Activated CTLs are capable of causing detrimental damage to axons.CTLs have been shown to transect MHC class I-induced axons in culture (Medana I et al., 2001).Axons and dendrites damaged by allogenic attack in vitro display degeneration characteristic of axonal injury (Manning et al., 1987).In MS lesions, effector CTLs directly contact dystrophic axons and terminal ovoids (Neumann et al., 2002).However, lesions at exceptionally early stages of demyelination contain few T or B cells (Henderson et al., 2009).Once in contact with the axon, cytotoxic granules of CTLs polarize toward the surface (Figure 3).Perforin, an effector substance contained within these granules, polymerizes to form a pore that is inserted into the axonal membrane, allowing water and salt to flow into the axon.Na+ influx initiates reversal of the Na+/Ca++ exchanger, leading to a vicious cycle of increased Ca++ intracellular accumulation (Smith et al., 2001).Granzyme, a second CTL effector substance, degrades axonal proteins, disrupting fast transport and potentially triggering apoptosis (Figure 4).

MHC induction
In the CNS, disparate cell lineages interact to ensure that the powerful immune defenses of the brain are not misguided.Autoimmune diseases such as MS must overcome these cell-tocell interactions that hold the immune system at bay.For example, neurons are able to suppress MHC expression in surrounding cells, in particular microglia and astrocytes, through ligand-receptor interactions (Neumann H, 2001).However, disruption of Na+ membrane-dependent electrical activity has been shown to decrease the efficacy of these interactions and increase induction of MHC class-I by inflammatory cytokines such as interferon-.In the case of MS, demyelination and axonal injury are both capable of disrupting Na+ flow in axons.This helps explain the increased CTL differentiation and activation of microglia and astrocytes.Moreover, MHC class-I induction on neurons and axons makes these cells viable targets for immunological attack.

Nitric oxide intermediates
Activated microglia and macrophages are readily seen surrounding and engulfing degenerating axons and end-bulbs (Lassmann, 2003).These monocytes are thought to introduce toxic effector molecules capable of damaging axons.Two of these that have been identified are proteases and reactive nitrogen species.Proteases degrade intracellular structural elements, causing disturbances in fast axonal transport (Anthony et al., 1998).High levels of nitric oxide (NO) radicals have been identified in MS lesions (Smith and Lassmann., 2002).NO radicals cause injury to mitochondria and disrupt axonal cytoarchitecture (Smith et al. 2001;Smith and Lassmann, 2002).NO radical-induced mitochondrial injury reduces generation of ATP and, if exceeding a certain threshold, triggers apoptotic mechanisms by cleavage and activation of caspases.Also, reactive nitrogen species disrupt the permeability of the axonal membrane, causing an influx of extracellular ions and axonal swelling.At low concentrations, NO species have been shown to block impulse conduction in axons even in the absence of structural damage (Redford et al., 1997).This demonstrates how clinical deficit may manifest in MS patients before dramatic degenerative changes are detectable in axons (Lassmann et al., 2003).Because reactive nitrogen species can also cause demyelination and oligodendrocyte damage, these substances fit well into the pathogenetic sequence of events in MS (Lassmann et al., 2003).In line with this concept, early structural changes in mitochondria and oxidative DNA damage has recently been implicated in axonal injury preceding demyelination (Nikić et al., 2011;Haider et al., 2011).

Demyelination
The loss of myelin and glia causes severe dysregulation of physiologic and homeostatic mechanisms in axons ultimately ending in cell death.The duration from onset of injury to neuronal death depends on compensatory and reparative mechanisms of the cellular milieu of the CNS in addition to the extent further inflammatory reactions accelerate the process.The most striking evidence that loss of myelin is behind neurological progression in MS are post-mortem histological studies showing the tight correlation between inflammatory demyelination and axonal loss (Frischer et al., 2009).Strong support also comes from experiments showing that myelin-protein deficient mice lose axons over time (Griffiths et al., 1998).The relative slowness of axonal loss in these mice is comparable to the course of neurodegenerative progression in MS (Nave and Trapp, 2008).Further evidence includes histological data showing a 60-70% reduction in axonal density in chronically demyelinated lesions compared with normal tissue of the same area, and a 30% reduction in fresh lesions (Mews et al., 1998).A similar percentage is seen in the spinal cord when the absolute number of nerve fibers are counted in a tract (Bjartmar et al., 2000).

Energy imbalance in demyelinated axons
A state of energy imbalance is the hallmark predecessor to cell death.Because demyelinated axons are in a state of increased demand and decreased generation of energy in the form of ATP, chronic energy imbalance might significantly contribute to axonal loss in MS.In demyelinated axons, Na+ channels become diffusely arrayed along the cell surface in contrast to normal clustering around nodes of Ranvier (Figure 5, Craner et al., 2004;Black et al., 2007b).These axons have an increased energy requirement per action potential.Disruption of ion transport caused by cytoarchitectual disturbances further increases the metabolic load of the axon.Simultaneously, various pathologic mechanisms reduce generation of ATP.These mechanisms include disruption of the mitochondrial membrane by ambient nitric oxide and free radicals as well as down-regulation of mitochondrial protein transcripts (Mahad et al., 2008).Using novel multiphoton imaging technology, a recent study identified mitochondrial pathology taking place within axonal swellings ensheathed with normal myelin in an animal model of multiple sclerosis (Nikić et al., 2011).This data shows that energy imbalance in the axon can occur even in the absence of local demyelination.
Energy imbalance in the demyelinated axon leads to a reduction in protein synthesis and fast axonal transport.The cytoplasmic density of Na+/K+ATPase exchanger drops, and consequently the electrochemical gradient critical to the functionality and survival of the cell is compromised (Young et al., 2008).The cell is overwhelmed with inrushing Na+ and is unable to compensate.Rising intracellular Na+ triggers a reversal of the Na+/Ca++ exchanger, leading to a Ca++mediated neurodegenerative cascade of events not unlike those seen in ischemia.This deleterious state has been called virtual hypoxia.

Oligodendrocyte decimation
The oligodendrocyte is an extraordinary glial cell derived from neuroepithelial stem cell progenitors.Its characteristic arbor projects numerous angular processes capable of myelinating up to 40 nerve axons (Weinshenker, 1994;Weiner, 1998;Compston and Coles, 2002).Classically oligodendrocyte processes wrap around the cell membrane of axons and produce concentric myelin layers.After myelination takes place, oligodendrocyte processes retain their physical connection with the myelin sheath of axons.In vitro studies using human oligodendrocytes and neurons cultured concomitantly show that molecular trophic factors released by oligodendrocytes significantly impact the survival and proliferation of axons (Wilkins et al., 2003).Moreover, small molecule secretions by neurons and axons are thought to influence the structural viability and intracellular mechanisms of local oligodendrocytes well after the myelination phenomenon.Therefore, a bidirectional trophic interplay between neurons and oligodendrocytes supports the healthy maintenance of both these cell populations.In MS, degenerating oligodendroyctes are numerous in white matter tissue adjacent to inflammatory foci and have been identified in NAWM regions distant from lesions (Henderson et al., 2009).In both lesions and NAWM regions, oligodendrocyte processes "die back" from the axonal myelin sheath in some cases (Lucchinetti et al., 2000).It is unclear whether the dying back phenomenon is resultant from contact with degenerating axons or dystrophy intrinsic to the oligodendrocyte.It is clear, however, that the environment of the MS lesion has a profound affect on oligodendrocyte survival.Compared to NAWM in both MS patients and controls, chronic lesions contained markedly decreased numbers of oligodendrocytes and oligodendrocyte progenitor cells (Kulmann et al., 1999).Disruption in the secretion, integrity, and uptake of trophic factors released and/or delivered by neurons and glia might contribute to the oligodendrocyte demise in MS.Other possibilities include NO-mediated cytotoxicity, T cell-induced killing, or the presence of a currently unknown toxin or virus affecting oligodendrocyte health.Our group is currently studying molecular profiles of MS oligodendrocytes in order to find out what factors influence their survival.

Dissolution of the axonal cytoskeleton
As described above, various mechanisms interface to ultimately bring intracellular Ca++ concentrations to dangerous heights.These mechanisms include direct T cell-mediated axonal injury, diffusion of Na+ channels along demyelinated internodal regions of axons, and free radical-mediated increased membrane permeability and mitochondrial damage.Reduction in mitochondrial output establishes an energy imbalance within the cell, leading to stalled fast axonal transport and skewed ion channel densities in the axonplasmic membrane.Rising intracellular Ca++ activates Ca++ dependent proteases which degrade the axonal structural framework.The activity of ambient proteases and the loss of trophic factors from oligodendrocytes are thought to amplify this degenerative process in the MS lesion environment (Anthony et al., 1998, Wilkins et al., 2003).The active disintegration of the intra-axonal microtubule network and decreased ATP availability stall the machinery of fast axonal transport.Proteins like APP accumulate, causing swelling and increased tension in a manner not unlike that seen in acute traumatic head injury (Povlishock, 1992).Furthermore, disabled fast transport arrests movement of Ca++ channel proteins along axonal microtubules.Subsequent insertion of these channels into the membrane results in Ca++ channels discontinuously arrayed along the surface of MS lesion axons (Kornek et al. 2001).The result is a vicious cycle of Ca++ influx which eventually leads to the demise of the axon.The positive feedback loop of protease activation-cytoskeletal disruption-ion channel insertion continues as more Ca++ rushes into the cell (Smith et al., 2001).Ultimately, swelling and structural damage reach a point where the axonal cytoskeleton dissolves (Lassmann et al., 2003).The axon is transected, and conduction along the nerve fiber inhibited.Organelles and fast transport proteins such as APP remain and continue to accumulate in the terminus, causing axonal end-bulb formations with characteristic morphology (Povlishock, 1992).

Wallerian degeneration
Anterograde "Wallerian degeneration" or secondary degeneration has been proposed as a mechanism of axonal loss.In this process, acute ultrastructural changes at some point along the axon can cause degeneration in regions of the nerve fiber not immediately associated with the site of injury.Nerves passing through a focus of inflammatory demyelination might therefore manifest pathological changes distal to the injured or transected site (Shindler et al., 2008).In this scope, degenerating nerves in the NAWM of MS patients might be spatially distinct from inflammatory foci but retain an intimate link to the cellular changes taking place at those sites of disease activity.That is, if axons and lesion are contained within the same tract, secondary degeneration helps explain the loss of axons in seemingly unaffected white matter areas contralateral to lesioned tissue (Evangelou et al., 2000a;Lovas et al., 2000;Evangelou et al., 2000b;Evangelou et al., 2004).Demyelination however is not the only pathological phenomenon in MS capable of inducing Wallerian degeneration.Any subtle change in cytoarchitecture, due to its effects on fast axonal transport, will be felt downstream.Immunological attack of axons by CD8+ CTLs is capable of initiating secondary degeneration in the absence of demyelination (Manning et al., 1987).In the same vein, NO-mediated axonal injury by monocytes could potentiate degenerative changes elsewhere in fiber tracts.Loss of important trophic factors can cause transynaptic secondary neuronal loss, which in turn leads to degeneration of the projecting axons (Wilkins et al., 2003).EAE is an inducible autoimmune condition in animals characterized by T cell mediated inflammatory demyelination and neurodegeneration (Gold et al., 2006;Steinman and Zamvil, 2006).Animals are inoculated with CNS proteins often with complete Freund's adjuvant (CFA) to stimulate an inflammatory reaction to myelin antigens.Pertussis toxin is used in these models to break down the blood brain barrier and allow immune cell infiltration.Depending on the protocol and animal genotype, EAE can follow a monophasic, relapsing/remitting, or chronic course.In mice, symptoms usually become apparent within 2 weeks of inoculation.The classic first symptom is loss of tonus in the tail leading to paralysis, followed by loss of motor function in the hind and then forelimbs.Symptoms associated with MS also present in EAE, including ataxia, optic neuritis, emotional liability, and cognitive dysfunction.
Axonal pathology in EAE mice morphologically resembles changes seen in MS patients (Soulika et al., 2009).EAE mice induced by immunization with myelin oligodendrocyte protein (MOG) peptide 35-55 show axonal vacuolization and fragmentation, as well as brain atrophy (Bannerman et al., 2005;Wang et al., 2005;Bannerman and Hahn, 2007;Herrero-Herranz et al., 2008;Jones et al., 2008).C57BL/6 mice immunized with MOG in CFA exhibit intra-axonal accumulations of APP and hyperphosphorylated heavy chains (NF-H).Neurological deficits manifest clinically in EAE mice at approximately the same time as characteristic SMI32+ spheroids are detectable in spinal cord axons (Figure 6, Soulika et al., 2009).These spheroids contain high concentrations of Toll-like receptor 8 (TLR8) protein (Soulika et al., 2009) shown to cause neuronal death via apoptosis and inhibit neurite formation (Ma et al., 2006(Ma et al., , 2007)).In this same EAE model, innate immunity infiltrates are detectable early on in the disease course.Neutrophils in the early innate immunological wave permeabalize the blood brain barrier and perform neurotoxic effector functions via a contact-dependent, protease-mediated mechanism (Dinkel et al., 2004).Moreover, a progressive, symmetric, severe loss of small diameter dorsal corticospinal axons is evident approximately 3 months post-innoculation (Soulika et al., 2009).At this time, inflammation is diminished to the level of controls, while genes associated with the activation of innate immunity continue to be expressed (Soulika et al. 2009).These data indicate that the innate arm of the immune system may play a key role in late-onset axonal loss in the MOGinduced EAE model (Weiner, 2009).

Neurodegeneration in mice lacking expression of 2'-3'-cyclic nuclotide 3'phosphodiesterase (CNP)
CNP is an intrinsic protein of oligodendrocytes associated with cytoskeletal elements (Nishizawa et al., 1985) and detectable in purified myelin (Kurihara et al, 1967).Engineered mice deficient in CNP (CNP-/-) display profound axonal loss, leading to a severe neurodegenerative disorder and premature death (Lapp-Siefke et al., 2003).This animal model demonstrates that oligodendrocytes provide support for axons independent of the physical stability of myelin itself.Degenerating neurons of CNP-/-mice display pathologic changes characteristic of MS (Povlishock, 1992) including axonal swellings filled with microtubules, dense bodies, multivesicular bodies, and mitochondria.These www.intechopen.comNeurodegeneration 138 cells were associated with phagocytic activity, reminiscent of macrophages and microglia often surrounding dystrophic axons in MS lesions.Moreover, the progressive neurological decline of CNP-/-mice is not unlike that seen in progressive MS patients.During the first 3 months, the knockout strain is indistinguishable from wildtype.This brings to mind the proposed clinical silence of the neurodegenerative component of MS preceding the progressive phase of the disease (Bjartmar and Trapp, 2003;Kremenchutsky et al., 2006).At 4 months CNP-/-mice begin to show hindlimb impairments, convulsions, and ataxia.At 6 months, the mice are unable to grasp a round bar with their hindlimbs, and display a reduction in overall brain size.Less than 20% of CNP-/-mice reach 13 months.Those that do are unable to balance on a round bar for 1 second, and exhibit profound muscle weakness, weight loss, gait abnormalities, and kyphosis.The CNP-/strain is of promising value to the MS research community.Not only does the model uncouple axonal loss from structurally intact myelin (demyelination), but it also demonstrates a relationship between intracellular processes of the oligodendrocyte and axonal survival.

Overview of therapeutic strategies to counteract neuronal-axonal loss in MS
Developments in our understanding of axonal loss in MS have spearheaded the relatively recent realignment of our thinking about the disease as a neurodegenerative as well as immune-mediated disease process.The increased understanding of axonal loss in MS has opened exciting research avenues geared toward developing novel therapies.These therapies aim to increase neuroprotection, enhance the remyelinating capacity of the CNS, and address the incongruity of trophic support supposed in the multiple sclerosis CNS environment by replacing both important molecules and decimated cell populations.Elucidating the mechanisms of nerve degeneration and how they interface with inflammatory demyelination during the biphasic clinical course of MS remains a critical step toward therapeutic victory over the disease.

Remyelination
The brain has evolved mechanisms to counteract axonal degeneration.Most of the reparative focus is on rebuilding the myelin architecture devastated by the inflammatory milieu (Prineas et al., 1993).Oligodendrocytes are responsible for repairing demyelinated axons through a process called remyelination.By repairing ultrastructural deficits accumulated through loss of myelin, remyelination not only partially restores saltatory conduction (Kriss et al., 1988) but also ameliorates degenerative processes occurring elsewhere in the axon such as secondary Wallerian degeneration (Irvine et al., 2008).In line with this, patients with robust remyelination live longer compared to patients showing few remyelinated brain regions (Patrikios et al., 2006).
Chronic MS disease processes diminish the ability for specific brain regions to remyelinate (Hanafy and Sloane, 2011).For example, hyaluronic acid accumulates in chronic MS lesions and inhibits remyelination and oligodendrocyte progenitor maturation (Figure 7, Sloane et al., 2010b).Activation of other molecular cascades such as the Notch1 pathway, Wnt pathway, and inflammation are thought to contribute to remyelination blockade (Hanafy and Sloane, 2011).The consequence of this is failure of the axon to successfully remediate injury and energy imbalance, further instigating dystrophic changes, transection, and secondary degeneration.A large sector of MS research is geared toward developing therapeutics that promote and amplify remyelination in MS.Strategies include enhancing oligodendrocyte proliferation and progenitor recruitment in chronic lesions (Patel et al., 2010), grafting exogenous oligodendrocyte progenitor cells in proximity to affected tissue (Wang et al., 2011), and blocking or genetically deleting receptors that inhibit remyelination (Mi et al., 2007;Sloane et al., 2010b).These efforts center around the idea that boosting remyelination capacity in MS patients would ameliorate long-term clinical outcomes and provide neuroprotection against progressive disease.

Na+ channel blockers
Increased Na+ influx into axons creates a state of virtual hypoxia in MS axons and contributes to axonal dissolution (Waxman, 2008).Irregular densities of Na+ channels detectable along demyelinated axons (Black et al., 2007a) along with altered expression of Na+ channels in MS neurons provide strong support for this idea (Craner et al., 2004).
Building on these observations, in vitro experiments show that selective Na+ channel blockers can prevent axonal injury from anoxia and ischemia (Stys et al., 1992a(Stys et al., , 1992b(Stys et al., , 1996;;Fern et al., 1993;).Using a MOG-induced EAE mouse model, Lo et al. showed that administration of the Na+ channel blocker phenytoin ameliorates 28-30 day clinical outcomes in EAE and reduces loss of corticospinal axons from 63% to 28% (Lo et al., 2002).Other studies have verified the neuroprotective effects of Na+ channel blockers in EAE (Bechtold et al. 2004).Moreover, phenytoin-treated mice in the Lo et al. study displayed 75% less inflammatory infiltration than normal EAE mice, possibly accounting for the dramatically improved clinical scores.This reduction has been attributed to hyperdensities of sodium channels on activated macrophages and microglia in the CNS of EAE mice (Craner et al., 2005).Blockage of Na+ channels presumably curtails neurotoxic immune effector functions of these cells, including NO radical production and phagocytosis, and diminishes immune-mediated axonal destruction.Infiltrating monocytes in MS lesions also show increased sodium channel expression (Craner et al., 2005).At present, three large scale phase-I human clinical trials have been launched to examine the efficacy of Na+ channel blockers in treating different stages of MS (Waxman, 2008).However, the progress of these trials has been disturbed by emerging evidence of acute disease exacerbation following withdrawl of Na+ channel blocker therapy.Black et al. found that EAE mice treated with phenytoin experienced marked inflammatory infiltration and increased vascular permeability once the drug was removed, with death occuring in approximately 50% of these mice (Black et al., 2007).

Preventing free radical damage
NO is a double-edged sword in MS.On the one hand, NO is thought to modulate immunological functions.For example, prevention of NO production has been shown to exacerbate some forms of EAE (Giovanni et al., 1998).On the other, NO-associated free radical damage has been implicated in the pathogenesis of axonal loss in MS and the clinical manifestations of the disease (Redford et al., 1997;Smith et al., 2001Smith et al., , 2002)).NO can be converted into reactive nitrogen species, such as peroxynitrate (ONOO(-)), and inflict cellular damage (Jack et al., 2007).In EAE, inducible nitric oxide synthase (iNOS) shows up in perivascular macrophages before demyelination, and is associated with transient functional disturbance in axons (Aboul-Enein et al., 2006).This brings to mind early, reversible pathological changes seen in axons before inflammatory demyelination (About-Enein et al., 2006;Nikić et al., 2011).Furthermore, iNOS is highly expressed in actively demyelinating, remyelinating, and chronic MS lesions, and is extensive in normal brain matter regions (Hill et al., 2004;Broholm et al, 2006).Because of these findings, blocking formation of reactive nitrogen species may be an attractive therapeutic target for the treatment of axonal loss in MS.
The neuroprotective ability of the drug glatiramer acetatate (GA, trade name Copaxone) is under active investigation.Subcutaneous injection of GA is currently used for the treatment of relapsing/remitting MS.GA is thought to work by inducing Th2/3 cells to cross the blood brain barrier, accumulate in the CNS, and express IL-10, TGF-, and neurotrophic factors such as brain-derived neurotrophic factor (BDNF) (Aharoni et al., 2003).GA has been reported to prevent free radical formation in peripheral blood adherent mononuclear cells in vitro (Iarlori et al., 2008).One of these radicals, superoxide anion (O(2)(-)), reacts with NO to produce peroxynitrate.Peroxynitrate has been associated with extensive oligodendrocyte cell death in a subset of MS lesions (Jack et al., 2007).Early animal studies on the neuroprotective effects of GA show promise.Spinal cords of glatiramer acetate-treated EAE mice show significantly reduced demyelination and enhanced remyelination compared to controls (Aharoni et al., 2008).Moreover, when treatment was initiated at the time of manifestation of clinical symptoms, or even in the chronic disease phase when extensive demyelination had already accumulated, GA dramatically reduced pathological damage.Randomized human clinical trials measuring disease activity and burden using state-of-theart imaging technology are currently investigating the neuroprotective effects of GA treatment in MS patients (García-Martín et al., 2010).As more is discovered regarding the pathogenesis of axonal loss and mechanisms of substances such as GA, selectively shutting down the neurodegenerative machinery of MS will become more feasible.

Conclusion
Since Charcot's identification of axonal pathology in MS in 1869, two main attributes of the disease are clear.First, independent of the relapsing/remitting phase of MS, a pathologic process of accumulating axonal loss is chiefly responsible for the currently untreatable neurological decline in MS.Second, although it might not be the causative factor, inflammation definitely spatiotemporally coincides with the extent of axonal injury in affected tissue.The presence of cytotoxic CTLs targeted against axons, free radical-induced mitochondrial damage, and mislocalization of membrane ion channels point to an etiology that is immunological in nature.Loss of oligodendrocyte trophic support or other protective functions also cannot be disregarded as a cause of axonal dystrophy.

Fig. 1 .
Fig. 1.Axonal injury in an active multiple sclerosis lesion.Immunocytochemistry staining with an anti-amyloid precursor protein (APP) antibody reveals stained spheroid elements 40-80 µm in diameter or in a beaded string-like appearance along the length of axons.Axonal dystrophy has caused these accumulations by interfering with fast axonal transport (fromFerguson et al., 1997).

Fig. 2 .
Fig. 2. Axonal transections in a multiple sclerosis lesion.Non-phosphorylated neurofilament is green and myelin is red.An end-bulb (arrow) retains a single axonal connection and is therefore terminal.Axons stain positive for non-phosphorylated neurofilament where myelin is discontinuous (Adapted from Trapp and Nave, 2008).

Fig. 4 .
Fig. 4. A CD8+ CTL recognizes MHC-I on an MS axon and releases effector substances.Granzyme helps degrade the intracellular fast transport network.Perforin inserts into the axoplasmic membrane and allows Na+ to freely diffuse into the cell.Rising intracellular Na+ leads to detrimental effects in the axon.

Fig. 6 .
Fig.6.Axonal spheroids (arrowhead) are evident in spinal cord axons in a mouse model of MOG-induced EAE.Although demyelination of the tract is evident, many of the spheroids are encircled with intact myelin (long arrow).This pathology is consistent with axonal changes in MS (adapted fromSoulika et al., 2009).

Fig. 7 .
Fig. 7. MS axon passing through a chronic MS lesion.Oligodendrocytes migrate into the lesion environment to replace lost myelin on the axon (remyelination).Hyaluronan accumulates in the chronic MS lesion, preventing oligodendrocyte maturation and possibly influencing migration into the lesion.