Systematic approaches to central nervous system myelin

Abstract

Rapid signal propagation along vertebrate axons is facilitated by their insulation with myelin, a plasma membrane specialization of glial cells. The recent application of 'omics' approaches to the myelinating cells of the central nervous system, oligodendrocytes, revealed their mRNA signatures, enhanced our understanding of how myelination is regulated, and established that the protein composition of myelin is much more complex than previously thought. This review provides a meta-analysis of the > 1,200 proteins thus far identified by mass spectrometry in biochemically purified central nervous system myelin. Contaminating proteins are surprisingly infrequent according to bioinformatic prediction of subcellular localization and comparison with the transcriptional profile of oligodendrocytes. The integration of datasets also allowed the subcategorization of the myelin proteome into functional groups comprising genes that are coregulated during oligodendroglial differentiation. An unexpectedly large number of myelin-related genes cause-when mutated in humans-hereditary diseases affecting the physiology of the white matter. Systematic approaches to oligodendrocytes and myelin thus provide valuable resources for the molecular dissection of developmental myelination, glia-axonal interactions, leukodystrophies, and demyelinating diseases.

Fig. 1

Central nervous system myelin. a The optic nerve of an adult, wild-type mouse was visualized by transmission electron microscopy upon fixation by high-pressure freezing and freeze substitution. Several myelinated axons are shown in cross section. Note the periodic arrangement of myelin membranes (electron micrograph kindly provided by W. Möbius). b One-dimensional gel-separation of CNS myelin. Myelin purified from wild-type mouse brains was separated by SDS-PAGE in different buffer systems providing improved resolution either in the low- (MES) or high- (MOPS) molecular weight range. Proteins were visualized with colloidal Coomassie (Coom., 5 μg protein load) or silver staining (0.5 μg protein load). Bands are denoted, which are constituted by known myelin proteins according to mass spectrometric identification. MAG myelin associated glycoprotein; TUBA α-tubulin; CNP 2′,3′-cyclic nucleotide phosphodiesterase; SIRT2 sirtuin 2; GAPDH glyceraldehyde-3-phosphate dehydrogenase; CLDN11 claudin 11/OSP; MOG myelin oligodendrocyte glycoprotein; PLP proteolipid protein; MBP myelin basic protein; CKB brain creatine kinase; CA2 carbonic anhydrase 2; MOBP myelin-associated oligodendrocytic basic protein. In bands marked with arrowheads, only proteins not previously associated with myelin were identified

Fig. 2

Assembling a compendium of CNS myelin proteins. a The number of proteins identified in different approaches to the CNS myelin proteome is plotted. The total number of myelin-associated proteins is unknown. Transmembrane proteins (black) were systematically predicted by TMHMM2, Phobius, and TMpred software. Proteins derived from mitochondria (which are diminished but not entirely lost during myelin purification) were predicted by Cello and Wolfpsort software and according to the literature. T [55]; V [54]; R [57]; W [53]; B [58]; J [7]; D [59]. I [56] provide datasets for mouse (I-m) and human (I-h) myelin. The integration of all datasets (‘All’) yields a catalogue of 1,261 proteins for which a unique gene identifier was available. b Single and multiple identifications. For all proteins identified in CNS myelin it was plotted in how many approaches they were identified. Note that fewer than half of the proteins (48 %) were identified in only one approach. c Cross-study reproducibility. For all approaches to the CNS myelin proteome it was plotted which percentage of identified proteins were additionally identified in at least one other approach. Note the high overall reproducibility. The seemingly low reproducibility of the dataset B [58] is due to the subfractionation of myelin by ion exchange chromatography (IEX) (see text for details)

Fig. 3

Venn diagram comparing the number of proteins identified in human versus rodent CNS myelin (a), and in CNS versus PNS myelin (b) according to [52]

Fig. 4

Profiles of myelin-associated mRNAs in oligodendrocytes. a The mRNA abundance profiles in oligodendrocyte progenitor cells (OPC), pre-myelinating, post-mitotic oligodendrocytes (OL), and myelinating oligodendrocytes (mOL), as determined by [9], were filtered for the proteins identified by MS in purified myelin. Upon k-means clustering, the normalized mRNA-abundance profiles were plotted with regard to the differentiation stage. Genes with significant mRNA-abundance changes were categorized in eight clusters. Known myelin-related genes are in bold, and genes encoding proteins probably derived from mitochondrial, blood or nuclear contamination are in gray. mRNAs in the clusters ‘UP’ or ‘ASCENDING’ display significantly increased abundance during oligodendrocyte differentiation while mRNAs in the clusters ‘DOWN’ or ‘DESCENDING’ are significantly suppressed during development. b The numbers of mRNAs per cluster are given