A place in the world for RNA editing

Introduction RNA editing describes the alteration of an RNA's informational capacity other than by splicing, 5'and 3'-end formation, and the creation of hypermodified bases, it can be divided into insertion or deletion editing, in which the RNA is cleaved and bases added or removed (see Simpson and Thiemann, 1995 [this issue of Cell]), and substitution or modification editing, in which the RNA is not cleaved (this minireview). Substitution editing is not a single process; rather, it is a series of distinct and probably separately derived traits (Table 1). With the establishment of systems in vitro for RNA editing, it has become possible to dissect the processes involved and to trace their evolutionary relationships. Surprisingly, some of these arcane processes are related, albeit distantly. C to U and U to C Editing of RNA Expressed in the Nucleus The tissue-specific editing of apolipoprotein B (apoB) mRNA in mammals is an early posttranscriptional event that converts a glutamine (CAA) to stop codon (UAA) (Navaratnam et al., 1995). This generates apoB48 (241 kDa), which is required for dietary lipid absorbtion. Full-length apoB100 (512 kDa) is made in the liver and transports endogenously synthesized cholesterol and triglyceride in the circulation. The establishment of a system in vitro for apoB mRNA editing has allowed the catalytic subunit of the editing enzyme to be identified (Navaratnam et al., 1995; Teng et al., 1993). It is a cytidine deaminase with homology to the Escherichia coil enzyme (Figure 1); designated APOBECI (forapoB mRNA editing cytidine deaminase I; Navaratnam et al., 1993). E. coil cytidine deaminase has two core domains of similar tertiary structure (Betts et al., 1994). One contains the active site with zinc at its center. The other forms a lid covering the active site cleft. In APOBECI the tertiary structure of the catalytic domain and mechanism of catalysis are conserved (Driscoll and Zhang, 1994; Yamanaka et al., 1994; Navaratnam et al., 1995). The enzyme has uniquely acquired the capacity to bind to AU-rich RNA through residues involved in zinc coordination, proton transfer, and the formation of the al~a structure that encompasses the active site (Navaratnam et al., 1995). The domain that forms the lid is absent. Interaction with an AU-rich sequence downstream of the edited C is crucial for editing. The target for the apoB mRNA editing enzyme is in a 22 nt sequence with 4 nt upstream of the edited C and a core downstream sequence (5'-UGAUCAG UAUA-3', +5 to +15) in which most alterations reduce or abolish editing (Navaratnam et al., 1995). Downstream and overlapping with this sequence is the AU-rich binding site (5'-UAUAU U-3', +12 to +17) for APOBECI. As APOBECI forms a homodimer, it is plausible that one subunit of the dimer binds the AU-rich sequence and positions the other for editing the C at a fixed distance upstream (Lau et al., 1994). APOBECI alone is not competent for editing. APOBECI requires other proteins, widely produced in cells that neither make APOBECI nor apoB mRNA (Teng et al., 1993; Driscoll and Zhang, 1994; Yamanaka et al., 1994; Navaratnam et al., 1995), to provide its RNA binding specificity. Although ultraviolet cross-linking has identified proteins of around 43 and 60 kDa that interact specifically with key nucleotides immediately downstream of the editing site (5'-UGAU-3', +5 to +8), the importance of these proteins in RNA recognition and as part of a larger editosomal complex (27S; 1400 kDa) that assembles at the editing site is uncertain (Navaratnam et al., 1995; Harris et al., 1993). APOBECI is expressed in the testes, ovary, and spleen, which do not make apoB, and editing activity is present in these tissues (Teng et al., 1993; Driscoll and Zhang, 1994; Yamanaka et al., 1994; Navaratnam et al., 1995). apoB mRNA editing is unlikely therefore to be unique. Other targets for the enzyme most probably exist.