POTENTIAL MECHANISMS OF DAMAGE

Axonal Damage Consequent to Demyelination

One of the potential mechanisms accountable for axonal loss following demyelination is Wallerian degeneration, whereby axons degenerate distal to the site of damage. Wallerian degeneration has been recognized as a contributor to axonal damage in MS.113

Reduced mitochondrial activity, due to increased energetic demand of demyelinated axons, can partially explain the decreased NAA levels observed in MS brain. Following demyelination, sodium channels, which are normally located at the nodes of Ranvier, become dispersed along the length of the demyelinated axon in order to restore nerve conduction.81,114 The increased number of sodium channels along the demyelinated axon requires a large quantity of sodium ions to be pumped back into the extracellular space by the energy-dependent Na+/K+-ATPase. Because adenosine triphosphate (ATP) concentrations are limited, accumulating sodium ions in the axoplasm results in impairment of Na+/K+-ATPase pump function. As a consequence, increased intra-cellular Na+ concentrations reverse the Na+/Ca2+ exchanger and permits Ca2+ entry into the cell. Further exacerbating this disrupted Ca2+ homeostasis in axons are reduced levels of the plasma membrane Ca2+ ATPase (PMCA2) efflux pump, which is seen in EAE,115,116 resulting in increased intracellular Ca2+ causing death of spinal cord neurons in vitro.117 In turn, increased axoplasmic concentration of Ca2+ activates proteases, impairs mitochondrial operation, depolymerizes microtubules, and compromises axonal transport.118,119

However, because axonal damage is detected also in areas devoid of demyelinating lesions and is barely controlled by current immunomodulatory therapies that limit the number and extent of demyelinating lesions, alternative pathogenetic mechanisms might come into play.

Axonal Damage Consequent to Direct Cytotoxic Attack

The exact mechanisms that underlie the progression of axonal damage are largely unknown. In vitro experimental evidence has shown CD4+ T cells can contribute to neurodegeneration in a mouse model of Parkinson’s Disease (PD).120 In addition, polyclonally activated CD4+/CD8+ effector cells align along axons and the soma of human neurons, causing neuronal cell death in vitro, but not death of oligodendrocytes or astrocytes.121 It has been proposed that loss of spinal motor neurons is the consequence of CD4+ and CD8+ T-cell infiltration in animal models of demyelination.122 However, the number of CD8+ T cells correlates with the extent of axonal damage in MS and the detection of CD8+ T cells in the brain,123 blood,124 and CSF125 of patients, in excess of CD4+ T cells,123 have suggested CD8+ T cells as important effectors of axonal damage. Indeed, accumulation of APP in damaged axons correlates with the number of macrophages and CD8+ T cells within MS lesions,15 CD8+ T cells closely interact with demyelinated axons in MS lesions126 and are capable of inducing direct damage to cultured murine neurons.127 Dying spinal motor neurons are surrounded by CD4+ and CD8+ T cells in EAE and postmortem MS tissue.122 Consistent with the proposed cytotoxic role of CD8+ T cells, depletion of CD4+ T cells in MS patients shows no improvement on relapse rate or inflammatory activity when visualized by MRI.128,129 Depletion of both CD8+ and CD4+ T cells, however, decreases relapse rate and new lesion formation but provides only little improvement of neurological symptoms.56,130 This suggests that CD8+ T cells are potential mediators of axonal damage. One possibility is that CD8+ T cells may cause damage indirectly, by targeting oligodendrocytes and myelin. For example, CD8+ T cells have been found to mediate the lysis of cultured murine oligodendrocytes131 and myelin in cerebellar slice cultures with axonal damage occurring as a bystander effect.132 The alternative possibility is that CD8+ T cells and possibly natural killer cells directly damage the axon due to the release of the cytolytic molecule perforin, a membrane pore-forming protein found in intracellular granules within these cells.133 The evidence to support the role for perforin in MS disease progression is substantiated from both the analysis of MS blood samples and the phenotype of mice with genetic ablation of the gene encoding for perforin. For example, high numbers of myelin basic protein (MBP)-reactive perforin mRNA expressing blood mononuclear cells (MNCs) are detected in MS patients134 and increased perforin expression is found in CD4+ T cells during MS disease exacerbation135 and correlates with the extent of brain lesions detected by MRI.136 Importantly, perforin-deficient mice are protected from axonal damage in the spinal cord and suggest that this might be a prominent mechanism of axonal degeneration.137

Axonal injury is therefore, at least in part, independent of demyelinating activity, and this concept may bear important implications for future therapeutic strategies aimed at preventing axonal loss.

Antibody/Complement-Mediated Lesion

We have previously discussed experimental evidence linking GM pathology, characterized by axonal loss and empty myelin sheaths to the injection of NFL into mice, thereby suggesting an antibody-mediated mechanism of damage.43 Furthermore, adoptive transfer of transiently expressed axonal glycoprotein 1 (TAG-1)-specific T cells into rats elicited a disease course with a preferential GM pathology, with WM demyelination only occurring after MOG antibody injection,138 thereby reflecting GM disease detected by early MRI measurements in MS patients. In further support of the hypothesis that neurodegeneration may underlie the pathogenesis of MS are also the findings of autoantibodies to neuronal and axonal antigens in the serum and CSF in a subset of MS patients, as well as the presence of meningeal B-cell infiltrates.44 The autoantibodies and antigens include neurofilament,139 axolemma enriched fractions,140 neurofascin,141 and contactin 2/TAG-1.138 However, some of these axonal or neuronal targets, namely intracytoplasmic proteins, may represent markers of injury rather than direct mediators of attack. Thus, neurodegeneration may underlie disease progression and WM demyelination, at least in a subset of patients.

Damage Caused by Dysfunctional Glial-Neuronal Crosstalk

The axonal degeneration detected in mice with deletion of genes encoding for myelin proteins suggested the existence of a bidirectional trophic support between myelin and neurons, affecting the long-term survival of neurons.51 For example, neurons and oligodendrocytes can exchange small metabolites necessary for myelin membrane lipid synthesis.142 The axonal degeneration detected in Plp-null mice, for instance, has been proposed to result from impaired transport of the nicotinamide adenine dinucleotide+-dependent deacetylase sirtuin 2, a cytoplasmic deacetylase implicated in axo-glial support. Similarly, ablation of a crucial peroxisomal enzyme in oligodendrocytes has been associated with dramatic axonal degeneration143 and the subsequent presence of inflammatory infiltrates.144 Thus, the loss of trophic support and deregulation of energy metabolism may destabilize metabolically isolated axons, resulting in their damage.51

Damage Caused by Exposure to Glutamate and Cytokines

We have discussed evidence linking inflammation to axonal damage even in the absence of demyelination. Meningeal infiltrates and diffuse microglial activation within the parenchyma have been associated with the release of cytokines and increased levels of the excitatory neurotransmitter glutamate. It is conceivable that these important mediators might directly modulate axonal function, within GM cortical areas (where the neurons are exposed to the meningeal infiltrates) or within normal appearing WM regions (characterized by diffuse microglial infiltration in the absence of demyelinating lesions). Recent evidence suggests that exposure of cultured neurons to glutamate and cytokines results in impaired axonal transport followed by axonal transections.103 This process was found to be mediated by a Ca2+-dependent export of the enzyme histone deacetylase 1 (HDAC1) from the nucleus to the axoplasm, with consequent disruption of mitochondrial and cargo transport along axons.103 Pharmacological inhibitors of HDAC1 (ie, MS-275), but not inhibitors of other cytosolic HDACs (ie, tubacin), were able to prevent the damage and partially restore axonal transport in neurons exposed to cytokines and glutamate, thereby suggesting protein acetylation as an important therapeutic target for axonal damage in demyelinating disorders.145

Exposure of cultured neurons to glutamate and cytokines results in impaired axonal transport followed by axonal transections. Pharmacological inhibitors of HDAC1 (ie, MS-275), but not inhibitors of other cytosolic HDACs (ie, tubacin), were able to prevent the damage and partially restore axonal transport in neurons exposed to cytokines and glutamate, thereby suggesting protein acetylation as an important therapeutic target for axonal damage in demyelinating disorders.