In addition to binding chromatin modifiers, there is increasing evidence that lncRNAs can bind and positively regulate transcription by both transcription factors and hormone receptors. A clear example are the neural lineage-specific lncRNAs RMST, Paupar and Evf-2 that can bind and positively regulate the function of transcription factors Sox2, Pax6, and Dlx2, respectively 61-63. lncRNAs PRNCR1 and PCGEM1 can cooperatively bind the androgen receptor and positively regulate gene activation in prostate cancer cells 64.
In contrast, there is also evidence that lncRNAs can serve to bind and antagonize transcriptional regulators and chromatin-modifiers: lncRNA PANDA binds transcription factor NF-YA and causes a failure to activate transcriptional targets 65 whereas, SchLAP1 binds chromatin-remodelling enzyme SNF5 and antagonizes its binding to targets in prostate cancer cell lines 66. Moreover, lncRNA Gas5 serves contains a ‘decoy’ binding site for the glucocorticoid receptor and inhibits its binding to downstream targets 67. In addition, Jpx, a lncRNA that is transcribed divergently ~10 kb upstream of the Xist locus. Before the onset of X-inactivation, the XIST promoter is bound by CTCF. Jpx is upregulated upon X-inactivation and binds CTCF. This binding allows Jpx to ‘evict’ CTCF from the Xist promoter, and X-inactivation proceeds68. Therefore, in addition to allowing for the active regulation of chromatin state, lncRNAs can also cooperate or antagonize a numerous of transcriptional regulators in the nucleus to control gene expression programs 68.
Alternative splicing represents a mechanism for generating extraordinary diversity from a limited set of gene loci. Differential use of splice isoforms can, for example, change the DNA-binding affinity of a transcription factor to allow for the maintenance of pluripotency in ESCs 69. The lncRNA Gomafu has a conserved UACUAAC tandem repeat motif, which mimics the intron branch point sequence 70. Gomafu can therefore bind and sequester splicing factors away from pre-mRNA. Specifically, Gomafu can bind the splicing factors SRSF1 and QKI, causing alternative splicing of at least two mRNA targets (71). Knockdown or overexpression of Gomafu causes opposite changes in exon selection of these targets, further suggesting that Gomafu expression levels alone can directly influences alternative splicing of specific transcripts. Another lncRNAs involved in the regulation of splicing is MALAT1. Such transcript localizes to nuclear paraspeckles, subnuclear domains that are thought to be centered for mRNA splicing 72. Mechanistically, MALAT1 binds SRSF1, an RNA splicing factor 73. Antisense oligo-mediated depletion of MALAT1 causes widespread alternative splicing of hundreds of exons. It can achieve this in part through the regulation of splicing factor recruitment to nuclear paraspeckles and pre-mRNA targets. Interestingly, three independently generated transgenic mice that lack MALAT1 do not have any changes in alternative splicing or SR protein localization 74-76. Contradictorily, MALAT1 involvements in splicing rather irregular as it was reported in a study that there is no influence of oncogenic transcription factor B-myb in fibroblast as previously described 73,131. These mouse studies suggest that lncRNA phenotypes may be highly dependent on cell cultures conditions and cellular context, and highlight the need to carefully design knockout strategies for lncRNAs. It is clear that a straight genetic knockout may not phenocopy the results of an acute depletion in culture using siRNA or similar methods.