Identification of lncRNAs involved in neuronal differentiation and intellectual disability

Recent studies have assigned important functions to lncRNAs in gene regulation and protein interactions. Since many of these lncRNAs emerged recently during vertebrate and primate evolution, a crucial role in the human brain is anticipated. Here, we aimed at identifying candidate lncRNAs associated with neuronal development and intellectual disability (ID). To do this, we combined genomic annotations (i.e. LNCipedia database) with functional genomics data (neuron-specific H3K4me3, REST binding & DNaseI hypersensitivity). For each of these three potential filters, we performed an enrichment analysis of ID genes and genome wide association studies (GWAS) hits for central nervous system (CNS) disorders, both on RefSeq protein-coding genes (ID genes & GWAS hits) and LNCipedia lncRNA transcripts (GWAS hits only). We found the neuron-specific H3K4me3 mark to confer the highest specificity for genes involved in ID and neurodevelopment. Applying this mark as a filter resulted in a set of 4188 lncRNA transcripts, of which 53 harbour a GWAS hit for CNS disorders. As the presence of such a SNP directly implicates these lncRNA loci in neuropathogenesis, we focused during subsequent analyses on this set of 53 lncRNAs. This approach was complemented by extensive expression profiling of all protein-coding genes and ca. 23,000 lncRNA transcripts in 15 human tissues, among which 8 brain samples. This allowed us to construct coexpression profiles for the lncRNAs identified by our filtering strategy (30 out of 53 were covered on the expression array). Using Gene Set Enrichment Analysis (GSEA) we assigned putative functions to the selected lncRNAs. For 13 candidates, highly correlated protein-coding genes were enriched in gene sets linked to neuronal processes and development, underscoring the relevance of the strategy used here. Subsequently, these lncRNAs will be functionally validated.