Long non-coding RNAs in the epigenetic regulation of oligodendrocyte differentiation

Abstract: Long non-coding RNAs (lncRNAs) constitute a heterogeneous class of RNAs with limited coding potential, united by an arbitrarily placed cut off of >200 ntd. The past decade has seen the emergence of lncRNAs as versatile regulators of gene expression, amidst skepticism regarding the biological usefulness of pervasive genomic transcription and its non-coding RNA products prevalent in most eukaryotes. A significant portion of lncRNAs operate in the development and functioning of the mammalian CNS. Oligodendrocytes (OLs) are the myelinating cells of the CNS that are essential for efficient saltatory conduction and axonal survival. They are derived from OL precursors (OPCs) and progress into transcriptomically heterogeneous OL sub-populations along the differentiation pathway to produce mature OLs, capable of myelination. These epigenetic transitions between different OL subpopulations are carefully regulated, spatially and temporally, by a network of transcription factors, chromatin modulators and lncRNAs. In demyelinating diseases like multiple sclerosis (MS), patients suffer immune mediated attacks against myelin. Eventually, remyelination strategies fail due to deficits in OPC migration and OL differentiation at the site of lesions. Thus, understanding molecular mechanisms governing OL differentiation and myelination is crucial not only for understanding OL function in health but also in disease, in order to develop suitable therapeutic interventions. The investigations presented in this thesis explore the role of lncRNAs and RNA-binding proteins in neurodevelopment, particularly in embryonic stem cells (ESCs) and cells of the OL lineage. Article 1 provides a resource for the protein interactome of a key pioneering transcription factor, Sox2, in different nuclear fractions of mouse ESCs. We found Sox2 to be a multifaceted regulator forming interactions with HP1 family of proteins, whose members perform as both activators and repressors in a context dependent manner. In addition to interacting with RBPs involved in post-transcriptional processes, Sox2 also interacted with Rn7sk, a well-known ncRNA involved in the regulation of transcriptional elongation at promoters and enhancers. Although they did not influence each other‘s recruitment to the chromatin, this interaction opens up the possibility for ncRNA mediated modulation of ES transcriptional programs dependent on Sox2. Article 2 draws important insights regarding lncRNAs from a broad transcriptomic resource established from single cell- as well as bulk RNA- sequencing of OL lineage cells from different developmental stages. From a subset of lncRNAs which were found to be specific for certain OL subpopulations, we investigated the role of 2610035D17Rik in modulating the expression of its neighboring gene, Sox9, a transcription factor essential for OPC specification. We decoupled the role of lncRNA transcript from its genomic locus using various loss-of-function strategies and found that the regulation of Sox9 was dependent on the regulatory elements and/or ongoing transcription at the 2610035D17Rik locus, rather than the transcript itself. In Article 4, we investigated a hitherto unexplored RNA-binding function of myelin gene expression factor 2 (Myef2), a known transcriptional repressor of myelin basic protein (MBP). To this end, we uncovered the RNA interactome of Myef2 in a mouse oligodendroglial cell line with individual nucleotide resolution CLIP (iCLIP) followed by sequencing. We show that Myef2 interacts with CUG motifs located within introns and 3‘UTRs of protein-coding genes, a finding which implicates Myef2 in post-transcriptional processes like splicing and RNA stability. Finally, in Article 3 we have identified disease specific transcriptomic profiles of OL lineage cells through single-cell RNA sequencing of OPCs and OLs derived from experimental autoimmune encephalomyelitis (EAE) mice, a model that recapitulates several aspects of MS. EAE specific OPC and OL clusters were enriched for genes involved in antigen processing and presentation (MHC class I/II). We could demonstrate that OPCs can phagocytose myelin debris and MHC-II-expressing OPCs can activate memory and effector CD4-positive T cells. These findings show OL lineage cells as active participants in MS pathology than passive targets. Further, the findings of Article 2 implicate 2610035D17Rik as a regulator of immunomodulatory properties of oligodendroglia, as 2610035D17Rik KO cells showed reduced expression of IFNγ responsive genes and elevated expression of those involved in antigen presentation, compared to the controls, following IFNγ stimulation.