Although myelin serves similar functions in the PNS and CNS, its main protein constituent differs: the type I integral membrane protein P0 (protein zero) is present in PNS myelin, whereas the tetraspan membrane protein PLP (proteolipid protein) resides in the CNS. It is thought that P0 was initially the primary structural protein of both PNS and CNS myelin — which first appeared 440 million years ago in cartilaginous fishes — but was replaced by PLP in the CNS after the divergence of the bony fishes 400 million years ago. What could be the benefits of the P0 to PLP conversion during evolution?

To address this issue, Yin and colleagues reversed the evolutionary step by generating transgenic mice that expressed P0, rather than PLP, in the CNS. In these P0-CNS animals, the level of P0 expression in the CNS is similar to that of PLP in normal mice, and replacing PLP with this ancestral protein in the CNS does not affect the expression of other myelin proteins. In addition, P0, like PLP, is able to stabilize compact CNS myelin, and its distribution is indistinguishable from that of PLP in wild-type mice. Electron microscopy studies show that P0-CNS myelin is structurally similar to its PNS counterpart, and has greater periodicity (membrane spacing) than that of wild-type CNS myelin.

Next, the researchers studied the effect of replacing PLP with P0 by measuring the animals' motor function. In P0-CNS mice, motor performance was normal at 6 months of age, but declined significantly (by 90%) by 1 year — by which time 50% of the animals had died. These observations are consistent with the precocious accumulation of the amyloid precursor protein — a reliable indicator of axonal pathology in primary myelin disease affecting PLP-deficient mice — in the brains of P0-CNS mice at a young age. So, PLP, but not P0, might provide trophic support for axons in the CNS, thereby delaying the onset of neurodegeneration.

This elegant study indicates that the shift from P0 to PLP during CNS myelin evolution was associated with an important neuroprotective function of myelin-forming glia. This finding may further our understanding of human myelin diseases, in which a spectrum of neurological disabilities is associated with null mutations, deletions and point mutations in the PLP gene.