Science. 2011 Jul 1;333(6038):53-8. Epub 2011 May 19.
Widespread RNA and DNA sequence differences in the human transcriptome.
Li M, Wang IX, Li Y, Bruzel A, Richards AL, Toung JM, Cheung VG.
In this paper authors show widespread differences in the sequences derived from DNA and RNA levels in human (DNA and RNA were extracted from B-cells in 27 human individuals). The differences between exome and transcriptome are quite significat and might point out that unknown RNA editing mechanism exist in human cells, other than already known editing mechanisms which mediate A->G and C->U changes.
In total they discovered 28,766 editing events residing in exons of 4,741 known genes. They discovered all 12 possible categories of changes of the bases.
However it is important to point out than their criteria for defining an editing event is rather weak, at least 10% of the reads from RNA-Seq needs to carry the base that is different from the DNA sequence. They were considering all sites that are covered with at least 10 RNA-Seq reads, which theoretically leaves a possibility they were analyzing sites with only one RNA-Seq read that contain a mismatched base (10% of 10 reads). This indicates they might have been analyzing a significant number of sequencing errors, actually.
The examples of similar editing events present in multiple individuals or when other tissue were analyzed, in combination with identification of their editing events in the human proteome, strengthens their claim made in this paper and indicates that many of these events might be true.
Although total number of editing events is 28,766, as not every event is present in all individuals, each individual possessed smaller number of editing events - ranging from 282 to 1,863. Editing events were more common in 3ÚTR exons then in other exons. The levels of mRNA that actually contain the changed base were presented by the distribution of the percentage of RNA-Seq reads with the changed base and this gives bimodal distribution with the peaks close to 0 and 100%. As many of such edited mRNA were minor variants it is not surprising their corresponding peptides were not found massively by mass-spec analysis in this study.