RNA editing in mammals is exclusively of the substitutional type. Two types of substitutional RNA have been reported in mammals, namely adenosine to inosine (A to I) and cytosine to uracil (C to U) RNA editing (1,2). All forms of substitutional RNA editing result in an alteration of the amino acid sequence encoded by the edited mRNA, emphasizing the relevance of this model of genetic regulation in creating diversity at the level of the transcriptome. A to I RNA editing modifies pre-mRNA transcripts of the neuronal calcium-gated glutamate receptor (GluR) and serotonin (5HT) receptor-type 2c mRNA, as well as viral RNA such as hepatitis delta virus (2). A to I RNA editing is mediated through the enzymatic action of distinct members of a family of adenosine deaminases acting on double-stranded RNA (ADARs). At least three members of this gene family have been characterized, each with partially overlapping target specificity, suggesting that a range of potential targets exist for each member (1,2). A to I RNA editing proceeds via enzymatic deamination at the C6 position of the purine ring. The reading frame of the edited transcript is altered because the inosine residue is recognized as guanosine by the translational apparatus, making the net change an A to G. The selectivity of A to I RNA editing depends upon an intronic sequence which can fold and undergo base pairing with an adjacent exonic complementary sequence, resulting in the formation of a double-stranded structure encompassing the editing site (2,10). This requirement for a double-stranded RNA template is an important distinction between A to I and C to U RNA editing since the former requires a pre-mRNA template containing intronic regions which form a double-stranded conformation with a complementary region of the exon containing the edited base. A to I RNA editing is thus confined to unspliced RNA transcripts. There is a second biochemical distinction between A to I and C to U RNA editing. Specifically, A to I RNA editing is mediated by ADAR gene products that are functionally competent as single recombinant proteins. ADARs do not exhibit a requirement for additional cofactors in order to mediate enzymatic activity since each protein contains both double-stranded RNA-binding domains as well as a deaminase domain which together permits them to act as modular-editing enzymes in vitro (2,10).
Three examples of C to U substitutional RNA editing have been described in mammals and each will be reviewed to provide a framework for a detailed discussion of the underlying mechanisms. These include the RNAs encoding apolipoproteinB (apoB), neurofibromatosis type 1 (NF1), and the novel target of apobec-1 (NAT1). ApoB RNA editing is the original and most extensively described form of C to U substitutional editing and will be the major focus of this review. NF1 RNA editing is less well understood, but the available information suggests that certain features may be shared with apoB RNA editing. NAT1 is a novel gene with homology to the translational repressor eIF4G and was discovered in a genetic screen for targets of apobec-1 (described in Section II.C).
In addition to these well-described examples, there are other reports of substitutional RNA editing in mammals, each involving U to C changes. These include RNA editing of the Wilms' tumor susceptibility gene (WT1) in rodent and human tissues, and U to C RNA editing of the mouse mitochondrial 16S transcript in certain tissues (11). In regard to the possible U to C editing of the WT1 RNA, an intriguing report almost a decade ago (12) indicated that a single nucleotide change was observed in both rat and human tissues at a conserved position (nucleotide 839) which resulted in a change in the amino acid sequence from leucine (CTC) to proline (CCC). U to C RNA editing modified ~ 30% of the endogenous WT1 transcripts and the change was further predicted to result in a loss of repressor activity with functional implications in both renal growth and malignant transformation (12). However, two more recent reports failed to confirm these earlier findings. In one report, a series of 15 Wilms' tumor samples was analyzed and revealed only the wild-type (CTC) sequence (13). A second report demonstrated less than 1% RNA editing in rat and mouse kidney and no editing in a human sample (14). Our summary conclusions from these studies suggest that methodological or strain-specific differences may account for the divergent results. From a practical perspective, this uncertainty precludes any formal conclusions concerning the incidence, extent, or potential biological significance of WT1 RNA editing at this time.
Another example of U to C RNA editing was recently reported in which this modification occurred in a chimeric transcript containing 16S mitochondrial RNA joined by a linker sequence of 121 nt to the 5' end of ribosomal RNA (15). The workers found a single U to C change in the chimeric 16S RNA isolated from sperm and testis RNA, and further demonstrated both edited and unedited forms of the RNA in somatic tissues such as spleen and liver (11). The authors formally excluded such possibilities as cloning artefact, genetic polymorphisms, and amplification of a pseudogene, as well as RNA polymerase misincorporation. Although the mechanism of this apparent U-to-C-editing reaction was not addressed, potential explanations include transamination and transglycosylation of the targeted nucleoside base. Similarly, the function of the chimeric RNA remains unknown and the impact of this putative RNA-editing event awaits further study.
Was this article helpful?