Failure to generate atheroprotective apolipoprotein AI phenotypes using synthetic RNA/DNA oligonucleotides (chimeraplasts)

Abstract Background Elevated plasma high‐density lipoprotein (HDL), and its major constituent apolipoprotein AI (apoAI), are cardioprotective. Paradoxically, two natural variants of apoAI, termed apoAI Milano and apoAI Paris , are associated with low HDL, but nevertheless provide remarkable protection against heart disease for heterozygous carriers and may even lead to longevity. Both variants arise from point mutations and have Arg 173 and Arg 151 to Cys substitutions, respectively, which allow disulphide‐linked dimers to form. Potentially, synthetic RNA/DNA oligonucleotides (chimeraplasts) can permanently correct single point mutations in genomic DNA. Here, we use a variation of such targeted gene repair technology, ‘gain‐of‐function chimeraplasty’, and attempt to enhance the biological activity of apoAI by altering a single genomic base to generate the atheroprotective phenotypes, apoAI Milano and apoAI Paris . Methods We targeted two cultured cell lines that secrete human apoAI, hepatoblastoma HepG2 cells and recombinant CHO‐AI cells, using standard 68‐mer chimeraplasts with polyethyleneimine (PEI) as carrier and then systematically varied several experimental conditions. As a positive control we targeted the dysfunctional APOE2 gene, which we have previously converted to wild‐type APOE3 . Results Conversion of wild‐type apoAI to apoAI Milano proved refractory, with limited correction in CHO‐AI cells only. However, a successful conversion to apoAI Paris was achieved, as demonstrated by polymerase chain reaction‐restriction fragment length polymorphism (PCR‐RFLP) analysis and direct genomic sequencing. Unexpectedly, attempts with a new batch of 68‐mer chimeraplast to enhance conversion, by using different delivery vehicles, including chemically modified PEI, failed to show a base change; nor could conversion be detected with an 80‐mer or a 52–76‐mer series. In contrast, when a co‐culture of CHO‐E2 and CHO‐AI cells was co‐targeted, a clear conversion of apoE2 to apoE3 was seen, whereas no apoAI Paris could be detected. When the individual chimeraplasts were analysed by denaturing electrophoresis only the active apoE2‐to‐E3 chimeraplast gave a sharp band. Conclusions Our findings suggest that different batches of chimeraplasts have variable characteristics and that their quality may be a key factor for efficient targeting and/or base conversion. We conclude that, although an evolving technology with enormous potential, chimeraplast‐directed gene repair remains problematical. Copyright © 2003 John Wiley & Sons, Ltd.