ncRNA mediated modes of gene regulation (TGS and TGA) and psuedogene directed regulation of protein-coding gene function

Figure 1: Model for non-coding RNA directed transcriptional gene silencing (TGS) in human cells. Small antisense non-coding RNAs can be generated (sequences determined via the use of a computational algorithm) to target silent state epigenetic changes to targeted loci. The result of this targeting is the epigenetic remodeling of the target locus, which results in transcriptional gene silencing of the target site. Recent findings suggest that long non-coding RNAs, that are antisense to gene promoters and possibly expressed from bidirectionally transcribed loci, are endogenous effectors driving this process in human cells (reviewed in (Knowling and Morris, 2011; Morris, 2009)).

Figure 2: Model for transcriptional gene activation (TGA) or de-repression of gene expression in human cells. Genes that are under long non-coding RNA mediated transcriptional regulation (A-D) can be targeted with small RNAs or antisense oligonucleotides (E-F) leading to a loss of RNA directed transcriptional gene silencing and subsequence gene activation/de-repression. Examples of gene activation can be found with tumor suppressor genes PTEN, p21, p15 and with the transcription factor Oct-4 (Hawkins and Morris, 2010; Johnsson et al., 2013; Morris et al., 2008; Yu et al., 2008).

Figure 3: Pseudogene mediated regulation of PTEN transcription and translation. (A) The PTEN pseudogene 1 (PTENpg1) expresses both sense and 2 antisense isoforms from chromosome 9 in human cells. (B) The alpha PTENpg1 antisense variant functions in Trans to target epigenetic silencing complexes consisting of DNMT3a and other possible proteins such as EZH2 to the promoter for the PTEN protein-coding gene on chromosome 10 resulting in (C) epigenetic and transcriptional silencing. (D) The beta PTENpg1 antisense variant interacts in an RNA:RNA dependent manner to (E) localize PTENp1 sense transcripts to the cytoplasm where miRNA sponging may occur. This model is based on work presented in Johnsson et al (Johnsson et al., 2013).