Chapter 18 : Apolipoprotein Β mRNA Editing

Schematic diagram of apoB mRNA editing. The genomic structure of human apoB is at the top. The 29 exons are depicted by vertical bars, and the 28 introns are depicted by lines between the exons. The translation initiation codon and editing site are as indicated. In humans, apoB mRNA editing occurs only in the intestine. The nuclear compartment is separated from the cytoplasmic compartment by nuclear membranes. Mature apoB mRNAs are shown exported through nuclear pores into the cytoplasm, where they are translated into proteins.

Primer extension assay for the degree of apoB mRNA editing. A radiolabeled antisense oligonucleotide, shown by an arrow, downstream of the apoB mRNA (human apoB mRNA is used as an example) editing site serves as a primer for reverse transcription of the apoB mRNA mixture to be assayed. In the reverse transcriptase (RT) reaction, dATP, dCTP, dTTP, and ddGTP are added into the mix so that the primer extension reaction will stop at the first cytidine of the RNA template because of the incorporation of the chain-terminating ddGTP. When the canonical C-6666 is changed to a U by RNA editing, the reverse transcription product will be extended to the next upstream C, resulting in a longer primer extension product. The relative size of the two primer extension products can be resolved by polyacrylamide gel electrophoresis and autoradiography.

Sequence alignment of apoB mRNA from various vertebrate species, neurofibromatosis type 1 (NF1) mRNA, and NAT1 mRNA around the edited site. C-6666 in human apoB mRNA is indicated by an arrow. The chicken apoB mRNA sequence is not edited and is included for comparison. In human apoB mRNA there is a second site that is edited at about 10% efficiency compared with the canonical C-6666. It displays substantial variation in the mooring sequence region. Putative regulatory elements (regulator, spacer, and mooring sequence) are indicated by overlines. Residues that are identical to the human sequence are indicated by dashes, substitutions are indicated by uppercase letters, and insertions are indicated by lowercase letters. *, NAT1 mRNA has five mooring-like consensus sequences, but only the major one associated with an edited C is shown here, †, NF1 was originally identified by the presence of mooring-like consensus sequence, but the enzyme responsible for editing is probably not apobec-1, so the significance of the presence of the mooring-like sequence in this mRNA is unclear. Even more perplexing is the fact that in contrast to the human NF1 mRNA, the rat NF1 mRNA is not edited ( Skuse et al., 1996 ).

Tissue-specific mRNA expression of apobec-1 in mice. (A) Northern blot analysis of apobec-1 mRNA in mouse tissues. The mRNA exists in two different sizes, an ∼2.2-kb form in small intestine and an ∼2.4-kb form in liver, spleen, kidney, lung, muscle, and heart. The 4.4-kb band found in liver, lung, and spleen may represent yet another alternatively spliced apobec-1 transcript which has not been characterized. (B) Structure of the mouse apobec-1 gene and its major mRNA transcripts. The gene comprises eight exons (indicated by boxes). In the liver (and presumably other tissues except small intestine), there are two alternatively spliced transcripts, resulting from the presence of two different splice acceptor sites in exon 4. Transcription of the major small intestinal form, in contrast, is initiated within intron 3. The intestinal mRNA thus contains exons 4 (with a unique 102-nucleotide fragment, shown by a shaded box, at its 5′ end derived from the intron 3 region of the liver transcript) through 8. Panel A is reproduced with permission from Nakamuta et al. (1995).

(A) Human apobec-1 amino acid sequence. The bipartite nuclear localization signal is highlighted by two solid boxes separated by 11 residues. The sequences homologous to highly conserved sequence blocks in E. coli cytidine deaminase are underlined by three shaded boxes. The first two shaded boxes are the cytidine deaminase homology motifs. Two phenylalanine residues (F66 and F87, in bold) encompassed within these two regions were found to be important for binding to apoB mRNA in vitro. The third box shows similarity to a region of E. coli cytidine deaminase only. The leucine-rich domain is indicated by a double-arrowheaded underline. (B) Schematic drawing of the sequence motifs identified in panel A. The zinc-coordinating residues H61, C93 and C96 are shown above the first two cytidine deaminase motifs. The NLS (nuclear localization signal) and deaminase homology domains correspond to those in panel A. The individual leucine residues in the leucine-rich domain are marked by vertical lines. (C) Alignment of the last cytidine deaminase homology domain in apobec-1 with a region in the C-terminal half of E. coli cytidine deaminase. The cytidine deaminase isolated from other organisms is generally shorter than the E. coli enzyme and misses this region shown. As is evident from the alignment, the apparently high homology is entirely the result of the matching Pro and Leu residues in this regions. The significance of this short homologous stretch of sequences, which is missing in cytidine deaminase from most other organisms, is unknown.

Phylogenetic tree of the apobec-1-related sequences reconstructed by the neighbor-joining method ( Saitou and Nei, 1987 ). Only the putative catalytic domains of the various proteins consisting of 34 amino acid residues (some with fewer residues because of deletions) were aligned and analyzed. The GenBank accession numbers of these sequences are as follows: 1, L27943; 2, P19079; 3, JS0609; 4, P13652; 5, L13289; 6, Z49149; 7, U74586; 8, X84693; 9, U00037; 10, U43534; 11, U82120; 12, U88006; 13, U10439; 14, U18942; 15, P06773; 16, P21335; 17, P30314; 18, P25539; 19, P32393; 20, P00814; 21, P16006; 22, 12136; 23, P30648; 24, U22262; 25, L07114; 26, L26234; 27, U10695. The horizontal scale at the bottom of the figure indicates the numbers of amino acid substitutions per position along the tree branches.

Hypothetical model for the structure of the apoB mRNA editing enzyme complex. ABBP-1, an hnRNP A/B-type protein, binds to apoB mRNA, causing it to become extended and exposing the canonical C-6666 editing site. Other hnRNP proteins may also serve a similar function. It is unclear if these hnRNPs are involved in apoB pre-mRNA splicing, apobec-1 undergoes spontaneous dimerization. It has affinity for apoB mRNA at the editing site, but its interaction with the latter is probably enhanced by ABBP-1 and/or other auxiliary proteins. Additional auxiliary proteins that function as assemblydisassembly proteins and editing-modulating proteins are also depicted. Some of these may also bind to apobec-1, whereas others do not directly interact with it. See the text for further details.