Functional heritage: the evolution of chimeric RNA into a gene

Abstract

Once believed to be unique features of neoplasia, chimeric RNAs are now being discovered in normal physiology. We speculated that some chimeric RNAs may be functional precursors of genes, and that forming chimeric RNA at the transcriptional level may be a 'trial' mechanism before the functional element is fixed into the genome. Supporting this idea, we identified a chimeric RNA, <i>HNRNPA1L2-SUGT1 (H-S)</i>, whose sequence is highly similar to that of a 'pseudogene' <i>MRPS31P5</i>. Sequence analysis revealed that <i>MRPS31P5</i> transcript is more similar to <i>H-S</i> chimeric RNA than its 'parent' gene, <i>MRPS31</i>. Evolutionarily, <i>H-S</i> precedes <i>MRPS31P5</i>, as it can be detected bioinformatically and experimentally in marmosets, which do not yet possess <i>MRPS31P5</i> in their genome. Conversely, <i>H-S</i> is minimally expressed in humans, while instead, <i>MRPS31P5</i> is abundantly expressed. Silencing <i>H-S</i> in marmoset cells resulted in similar phenotype as silencing <i>MRPS31P5</i> in human cells. In addition, whole transcriptome analysis and candidate downstream target validation revealed common signalling pathways shared by the two transcripts. Interestingly, <i>H-S</i> failed to rescue the phenotype caused by silencing <i>MPRS31P5</i> in human and rhesus cells, whereas <i>MRPS31P5</i> can at least partially rescue the phenotype caused by silencing H-S in marmoset cells, suggesting that <i>MRPS31P5</i> may have further evolved into a distinct entity. Thus, multiple lines of evidence support that <i>MRPS31P5</i> is not truly a pseudogene of <i>MRPS31</i>, but a likely functional descendent of <i>H-S</i> chimera. Instead being a gene fusion product, <i>H-S</i> is a product of cis-splicing between adjacent genes, while <i>MRPS31P5</i> is likely produced by genome rearrangement.