23. Frequency of Random and Targeted Chromosomal Integration of Helper-Dependent Adenoviral Vector

For gene therapy of inherited diseases, integration of a therapeutic gene into random sites on chromosomes could potentially lead to serious problems such as cellular transformation and gene silencing. Homologous recombination (HR) between exogenous and chromosomal DNA would be an ideal strategy to obtain safe and stable gene expression stably. We constructed helper-dependent adenoviral vectors (HD AdVs) and examined the ability of AdVs to correct an insertional mutation in exon 3 of the hypoxanthine phosphoribosyl transferase (Hprt) locus in male mouse embryonic stem (ES) cells. These HD AdV contained the wild-type exon 3 sequence of the mouse Hprt gene with 6.7 kb (HD AdHprt6.7) or 18.6 kb (HD AdHprt18.6) of flanking intronic sequences, as well as the |[beta]|-geo marker gene. HR and random integration (RI) frequencies were calculated by dividing the total number of HAT- or neomycin-resistant colonies by the total number of cells plated. HR was only obtained by using the HD AdHprt18.6 vector, and the frequency was nearly 0.0002 per infected cell at an MOI of 1000. In the case of the HD AdHprt6.7 vector, no HR events were detected. The HR frequency by electroporation of a plasmid construct identical to the HD AdHprt18.6 vector was 23-fold less than that by vector infection. The structure of the exon 3 region of Hprt locus was analyzed in HAT-resistant ES clones by Southern analysis, and all the clones showed a pattern of faithful HR. Since the number of target cells might be limited in ex vivo gene therapy, such as stem cell therapy, we next investigated whether HD AdV-mediated HR can be efficiently achieved even in smaller cell populations. We successfully obtained HAT-resistant colonies at the frequency of 2.1 |[times]| 10|[minus]|4 per infected cell in a 24-well dish (25 HAT-resistant colonies out of 1.2 |[times]| 10|[minus]|5 cells infected, on average). These results clearly indicate gene correction can be achieved relatively efficiently even in smaller cell populations using AdVs. Both HD AdVs achieved similar RI frequencies of 10|[minus]|4 |[minus]| 10|[minus]|3 per cell, which was measured as a frequency to produce G418-resistant colonies. In contrast to the HR frequency, which showed an MOI-dependent increase, the RI frequency reached a plateau at an MOI of 100. In order to directly detect integrated vectors at DNA levels, we next performed PCR on ES clones infected with HD AdHprt18.6 and formed in the absence of G418 selection. Surprisingly, vector DNA was detected in 4 of the 74 samples (5 % of colonies). These results indicate that the actual RI frequency of HD AdVs is much higher than previously believed. Next, we determined the integration sites of the HD AdHprt18.6 vector by adaptor-ligated PCR. In contrast to retroviral, lentiviral and adeno-associated viral vectors, which tend to integrate near or within active genes, HD AdV had a tendency to integrate in intergenic regions (12 of 17 integrants, 71%). Small deletions of integrated vector DNA were found at the end of ITRs (16 of 17 integrations, 94%), ranging from 1 to 30 bp in size. These findings suggest that HR mediated by HD AdV is efficient and safe, and might be a viable option for ex vivo gene therapy of stem cells.