Fig 2: Stage‐specific expression of LRF in the oligodendrocyte lineage during the progression of demyelination and remyelination. Spinal cord transverse sections immunostained for LRF (red) in combination with NG2 (green) to identify oligodendrocyte progenitor cells (A‐D) or GSTpi (green) to identify oligodendrocytes (E‐H'). A‐D: In the control white matter (A‐A"; PBS‐injected match to 4 wpi MHV), LRF is typically absent from NG2 cells (examples shown with green arrows) but was found in the nucleus of some NG2 cells, particularly NG2 cells with rounded nuclei (example at white arrow). In response to demyelination, there is a significant increase in the density of NG2 cells without LRF in white matter of MHV infected mice (4 wpi; B, with quantification in D). As remyelination progresses, NG2 cell density normalizes (8 wpi; C, D). NG2 cells with thickened processes indicative of a reactive phenotype remain in lesion areas at 8 wpi (C"). At both 4 and 8 wpi, lesion areas contain many LRF cells that do not express NG2 (B', C'). E‐H: In control white matter (F; PBS‐injected match to 4 wpi MHV), LRF is expressed in the vast majority of cells immunolabeled for GSTpi (white arrows in F'). During demyelination, there is a significant reduction in the density of GSTpi cells expressing LRF (4 wpi; E, G, G', G"). As remyelination progresses, GSTpi cell density normalizes (8 wpi; E, H, H'). The proportion of GSTpi cells expressing LRF is similar between normal and remyelinating white matter (8 wpi; E). DAPI nuclear stain shown in blue. One way ANOVA using n = 3 mice for each condition; p = 0.0071, F (3,8) = 8.536 for NG2 + LRF‐, p = 0.1816, F (3,8) = 2.078 for NG2 + LRF+, p = 0.0133, F (3,8) = 6.87 for GSTpi + LRF‐, p = 0.0024, F (3,8) = 12.15 for GSTpi + LRF+. Scale bars = 200 µm (A‐C, F‐H), 25 µm (all subset images). VH = spinal cord ventral horn.

Fig 3: LRF transcriptional regulation of myelin genes. Transcriptional activity was measured from transient transfections of OP cultures with expression plasmids either with (Tom‐LRF) or without (Tom) the Zbtb7a insert encoding LRF, along with firefly luciferase plasmids containing myelin promoter elements, and Renilla luciferase as an internal control. Transcription from the full‐length promoter for myelin basic protein (1323MBP) is significantly higher in defined medium (**p = 0.0017), indicating appropriate upregulation as expected during OP differentiation in vitro (A). Co‐transfection with the Tom‐LRF plasmid significantly represses transcription from 1323MBP in defined medium (****p < 0.0001; A). Further, in defined medium, LRF repression of a truncated 105 bp MBP promoter requires the intact wild type Sp1 site (*p = 0.0326; B) and is not observed with the mutation to prevent Sp1 binding (C). In the presence of mitogens, LRF does not alter transcription from the full‐length 1323MBP promoter (A) yet gains repressive activity using the truncated construct with the wild type sequence (**p = 0.0032; B) and with the Sp1 site mutated (****p < 0.0001; C). LRF expression significantly represses transcription from the 2',3'‐cyclic‐nucleotide 3'‐phoshodiesterase (CNP) promoter in both growth conditions (****p < 0.0001; D). In the presence of LRF, transcription of fatty acid synthase from FASN is significantly increased by the addition of mitogens (**p = 0.0037; E). Two‐way ANOVA using values combined from four independent experiments.

Fig 4: LRF knockdown impairs remyelination, but not oligodendrocyte repopulation, after demyelination from 0.2% cuprizone ingestion. Coronal sections from Plp/CreERT:Zbtb7a fl/fl mice were analyzed to determine the effect of LRF knockdown on the oligodendrocyte lineage stage and the extent of remyelination in the corpus callosum. Tamoxifen, or oil vehicle, was administered at the end of 6 weeks of 0.2% cuprizone and then mice were fed normal chow for 6 weeks to allow for spontaneous remyelination. A‐C: Immunoreactivity for MOG shows that remyelination has progressed after 6 weeks on normal chow in mice administered oil (A) or tamoxifen (B), as compared to the demyelination observed throughout the corpus callosum at 6 weeks of 0.2% cuprizone feeding (Figure S3). In mice administered oil, the remyelination is equivalent to myelination in age‐matched Plp/CreERT:Zbtb7a fl/fl mice that were not fed cuprizone (no cup; C). In mice administered tamoxifen, the remyelinated area of the corpus callosum is significantly reduced compared to the oil control mice (Cohen's d effect size = 2.08; C). D‐F: PLP in situ hybridization in mice administered oil (D) or tamoxifen (E) shows similar oligodendrocyte repopulation of the corpus callosum (F), which is also similar to mice that were not fed cuprizone (no cup; F). G‐I: Oligodendrocyte progenitors, identified by NG2 (green; G‐H), in mice administered oil (G) or tamoxifen (H) were found in a similar density (I). J‐L: Mature oligodendrocytes, identified by CC1 (red; J‐K) in mice administered oil (J) or tamoxifen (K) were also present at similar levels (L). cc = corpus callosum, cg = cingulum. Scale bars = 250 um. Analysis of n = 5 mice per condition. ANOVA with Tukey multiple comparisons used in C, F. Student t‐test used in I, L.

Fig 5: Role of DNA methylation on the binding of transcription factors to FCGRT. (a) Bioinformatics analysis of transcription factors potentially targeting the FCGRT locus (−1793 bp to A+1TG region). The −1058 bp to A+1TG region is shown in orange. Regions amplified by primer sets 1, 2, and 3 are indicated in magenta, cyan and green, respectively. CpG sites are represented as circles. (b) Chromatin immunoprecipitation (ChIP) assays evaluating the binding of Sp1, GABPα, and Zbtb7a, to the FCGRT locus in AC16 cells with or without Aza treatment (5 µM, 72 h). Panels show representative agarose gel electrophoresis analysis of PCR amplification products obtained with primers sets 1–3 using input samples and immunoprecipitated DNA samples. IgG controls are also shown (right panels, top, and bottom). Images were cropped from pictures taken from 3 individual agarose gels (full length pictures are included in supplementary material file 1, Figure S3). (c) Quantitative PCR analysis of ChIP samples assessing transcription factor enrichment in FCGRT regions amplified by primer sets 1 and 2. Values are expressed relative to values from IgG control antibody. Data show the mean ± SD from two independent qPCR experiments (3–4 replicates). *** P < 0.0001, 1-way ANOVA (Turkey’s test).