Functional genomic analysis of remyelination reveals importance of inflammation in oligodendrocyte regeneration

Abstract

Tumor necrosis factor alpha (TNFalpha), a proinflammatory cytokine, was shown previously to promote remyelination and oligodendrocyte precursor proliferation in a murine model for demyelination and remyelination. We used Affymetrix microarrays in this study to identify (1) changes in gene expression that accompany demyelination versus remyelination and (2) changes in gene expression during the successful remyelination of wild-type mice versus the unsuccessful attempts in mice lacking TNFalpha. Alterations in inflammatory genes represented the most prominent changes, with major histocompatibility complex (MHC) genes dramatically enhanced in microglia and astrocytes during demyelination, remyelination, and as a consequence of TNFalpha stimulation. Studies to examine the roles of these genes in remyelination were then performed using mice lacking specific genes identified by the microarray. Analysis of MHC-II-null mice showed delayed remyelination and regeneration of oligodendrocytes, whereas removal of MHC-I had little effect. These data point to the induction of MHC-II by TNFalpha as an important regulatory event in remyelination and emphasize the active inflammatory response in regeneration after pathology in the brain.

Figure 1.

TNFα in remyelination. Electron microscopy was performed on cross-sections of the corpus callosum in wild-type and TNFα^–/– mice. Analysis is shown in untreated mice, after complete demyelination with the cuprizone, and after 2 weeks of remyelination. Scale bar, 2 μm.

Figure 2.

Expression of MHC-I and -II. A, B, Quantitative analysis of MHC-I and -II transcripts using quantitative real-time RT-PCR on total RNA isolated from the corpus callosum during demyelination and remyelination in wild-type mice. Each bar represents the averaged fold induction over untreated wild-type mice of at least three mice per genotype per time point (±SD). The scale is interrupted to accommodate the large value obtained for MHC-II with wild-type mice after 4 weeks of cuprizone treatment. C, D, Quantitative analysis of MHC-I and -II transcripts in wild-type mice compared with that in mice lacking TNFα.*p < 0.05. E, Antibodies were used to localize expression of MHC-I and -II (green) during demyelination of wild-type mice and colocalized to microglia (RCA-1, red) and astrocytes (GFAP, blue). Yellow indicates colocalization of the red and green fluorochromes. Scale bar, 12.5 μm.

Figure 3.

MHC-I and -II in remyelination. A, B, Wild-type mice, MHC-I-null mice, and MHC-II-null mice were analyzed at various time points during demyelination and remyelination. Sections containing the corpus callosum at the level of the fornix were stained for LFB/PAS, and the corpus callosum was scored by three investigators on a scale of 0 (completely myelinated) to 3 (completely demyelinated), with the presence of blue fibers indicative of intact myelin. *p < 0.05. C, Remyelination was examined further with electron microscopy. Pictures represent a cross-section of the corpus callosum at the level of the fornix in untreated mice, mice completely demyelinated with cuprizone, and mice allowed to remyelinate for 1 week. Scale bar, 2μm. D, g-ratios and percentage unmyelinated fibers were calculated from a minimum of 200 fibers per mouse per time point. *p<0.001.

Figure 4.

MHC-II and the regeneration of oligodendrocytes. A, Oligodendrocytes were identified during demyelination and remyelination in wild-type mice and mice lacking MHC-I or -II using an antibody to GST-Π (red). B, Quantification of oligodendrocytes in the corpus callosum. Each bar represents an average of the results of at least six mice per genotype per time point. *p < 0.05.