An editor controlled by transcription

Abstract

The generation of proteomic diversity from a limited number of genes is a problem faced by many higher organisms with complex tissues. Alternative splicing is one well‐established solution to this problem. It has been reported that more than 70% of all human genes are alternatively spliced (Johnson et al , 2003). Another method to increase proteomic diversity is RNA editing. In animals, adenosine deaminases that act on RNA (ADARs) convert adenosines to inosines in structured or double‐stranded RNAs (dsRNAs). As inosines are interpreted as guanosines by the ribosome, A to I editing is the functional equivalent of an A to G change and can alter the coding potential of an RNA. The introduction of an inosine can also generate or delete splice sites, and therefore has a twofold impact on RNA diversity. Editing levels are highest in the brain, possibly reflecting the need for increased proteomic variation in this tissue. Well‐studied editing substrates in brain tissue include subunits of the glutamate ion channel family or the serotonin 5HT‐2c receptor. In both substrates, editing alters the coding potential of the RNA, which leads to the formation of proteins with altered properties. Interestingly, in these and many other substrates, the dsRNA structure required for editing is formed by base pairing of intronic and exonic sequences. Editing is therefore generally thought to be a co‐transcriptional process that occurs before the removal of introns. Accordingly, the speed of splicing would regulate the availability of binding sites for ADARs and thus the extent of editing. However, in ADAR2‐knockout animals, editing of the RNA that encodes glutamate receptor B (GluR‐B) …