This dissertation investigates why remyelination becomes inefficient with age in multiple sclerosis (MS) and how this failure drives neuronal loss. Key findings include a novel characterization of age-dependent remyelination failure in the optic nerve/retina system, demonstrating significant retinal ganglion cell (RGC) loss in middle-aged animals. A transcriptomic atlas reveals that remyelinating oligodendrocytes adopt a distinct state that impairs myelin lipid biosynthesis and is markedly worse in middle-aged mice. Finally, genetic ablation of Map3k12 (encoding DLK) provides robust, long-term RGC neuroprotection and hints at partial functional recovery, positioning DLK inhibition as a promising neuroprotective strategy for MS/optic neuritis.