386. Chromosomal Integration Sites of Double-Stranded Linear DNA in Mouse ES Cells

Abstract

In gene therapy approaches for inherited diseases, efficient transduction of target cells as well as stable maintenance of the therapeutic gene are required. Major viral vectors currently in use for efficient chromosomal integration in gene therapy include retroviral, lentiviral and adeno-associated viral (AAV) vectors. Recent large-scale analysis of chromosomal integration sites revealed the potential risk of insertional mutagenesis caused by these vectors. While retroviral and lentiviral vectors tend to integrate into transcription units, DNA viral vectors, such as AAV and adenoviral vectors, reportedly have no such strong preference, at least in tissue culture cells. To investigate if less frequent intragenic integration is unique to DNA viral vectors or common to all types of exogenous DNA (viral and non-viral), we cloned chromosomal junctions of stably integrated plasmid DNA in mouse ES cells using a plasmid rescue strategy. For this, the PGKneo plasmid was linearized and electroporated into mouse ES cells. Genomic DNA was extracted from G418-resistant cell populations, digested with enzymes that do not cut within the plasmid sequence and used to transform E. coli DH10B by electroporation. Sequence analysis of 125 colonies demonstrated that 39.2 % (49/125) of the integrations were in transcription units, which is not statistically different from the frequencies by AAV integration (38.8 %) and by computer simulation (34.8 %) as reported by Miller et al . Previous studies have reported a significant integration preference of AAV vectors for CpG islands (4.0%) and we have observed a similar bias with adenoviral vectors. However, our present analysis showed no significant difference between the frequencies of plasmid integration into CpG islands (0.80 %; 1/125) and that of a computational simulation (0.84%). These results suggest that it is a natural fate of exogenous DNA (both non-replicating viral DNA genome and plasmid DNA) to integrate without a significant preference into genes. This data supports the hypothesis that exogenous DNA integrates into preexisting chromosomal double-strand breaks by the non- homologous end-joining pathway. However, the higher integration frequency into CpG islands by DNA viral vectors than by plasmid DNA suggests that the integration pathway of DNA viral vectors is a little more complicated and that perhaps extra factors, presumably proteins binding to the end of viral genomes, play some role in the selection of integration sites.