Strategies to identify long noncoding RNAs involved in gene regulation

Long noncoding RNA (lncRNA) is operationally defined as RNA longer than 200 bases that does not encode mRNA, rRNA or tRNA [1,2]. Although several lncRNAs have been sporadically identified and characterized in the past 20 years, genome-wide identification of lncRNAs has only recently become possible with the advent of high-throughput sequencing technologies of cDNA (RNA-seq). Evidence that this field is gaining momentum can be seen in the most recent report of the ENCODE (Encyclopedia of DNA Elements) project published in September 2012, which described 9,640 lncRNA loci in comparison to 20,687 protein-coding genes in 15 human cell lines [3-5]. This ratio of lncRNAs and protein-coding genes underscores the potential magnitude and diversity of the biological effects mediated by lncRNAs. Indeed, despite the fact that only about 100 lncRNAs have been functionally characterized to date [4], it has become clear that lncRNAs are involved in almost every aspect of cellular and molecular biology. LncRNAs control cell differentiation, development, cancer progression, and cell metabolism, among other cell functions. At the gene expression level, lncRNAs regulate all processes of RNA metabolism including chromatin modification, transcription, splicing, RNA transport, and translation. LncRNAs themselves are transcribed from intergenic regions, exons, introns, and their overlapping regions (Figure 1A and 1B). At the mechanistic level, lncRNAs serve as “scaffolds” providing platforms to assemble RNA-protein complexes, “guides” to recruit RNAprotein complexes to target genes, and “decoys” by binding to and sequestering regulatory proteins away from their target DNA sequences [1,2]. Given the recent appreciation for the biological importance of lncRNAs, it is now clear that, regardless of the research project or field, one needs to ask whether lncRNAs are essential mechanistic components of the biological process under consideration. The first step to addressing this question is to