Adeno-associated virus type 2 integrates preferentially into the AAVS1 locus on chromosome 19 of the human genome. It was reported previously that transfection with two plasmids, one for Rep and the other carrying a transgene flanked by inverted terminal repeats (ITRs), enables preferential integration of the latter into AAVS1. Aiming at increasing the frequency of AAVS1-specific integration, the Rep-to transgene-plasmid ratio necessary to achieve a higher frequency of site-specific integration was examined. 293 cells were co-transfected with the Rep78 plasmid and an ITR-flanked Neo gene at different ratios. G418-resistant clones were selected randomly. Extensive Southern blot analysis showed an optimum range of Rep78 expression. In that range, approximately 20 % of clones harboured the Neo gene at AAVS1. Excess Rep expression, however, resulted in 'abortive' integration of the Neo gene, a rearrangement of AAVS1 without transgene integration. Rep78 appeared to cause abortive integration more extensively than Rep68. Deleterious effects of the Rep protein on the AAVS1 locus should be considered to develop an improved AAVS1-targeted system.

Targeted integration of foreign DNA is ideal for gene therapy, particularly when target cells such as hematopoietic cells actively divide and proliferate. Adeno-associated virus (AAV) has been shown to integrate its genome into a defined locus, AAVS1 (19q13.3-qter). The inverted terminal repeat (ITR) and Rep proteins are responsible for this site-specific integration, and a system has been developed that delivers a gene preferentially into AAVS1 by using these components of AAV. We examined whether this system could be applied to gene transfer into K562 cells. Two rep expression plasmids were tested, 1 driven by the cytomegalovirus (CMV) promoter (pCMVR78) and the other under the translational control of an internal ribosome entry site (pMGiR78) with mouse mammary tumor virus promoter. K562 cells were cotransfected with a rep plasmid and a plasmid containing a neo gene flanked by the ITRs. G418-resistant clones were isolated and analyzed by Southern blot analysis and fluorescence in situ hybridization (FISH). Southern blot analysis suggested AAVS1-specific integration of the neo gene in 6 (35%) of 17 clones when K562 cells were transfected with pMGiR78 by lipofection. FISH located the neo gene on chromosome 19 in 5 of these 6 clones (29%). Eight (32%) of 25 clones obtained by electroporation with pCMVR78 had the neo gene at AAVS1, according to Southern blot analysis, and 4 of these 8 clones (16%) were positive according to FISH analysis. These results suggest that site-specific integration of foreign DNA can be achieved at a significantly high rate in human hematopoietic cells using the AAV components.

Adeno-associated virus (AAV) vectors are potentially useful vehicles for the delivery of therapeutic genes into human cells. To determine the optimal expression pattern of AAV proteins (Rep78, Rep68, Rep52, Rep40, and Cap proteins) for packaging the recombinant AAV genome, helper plasmids were split into two portions. In this study, two sets of split-type helper plasmids were prepared; i.e., 1) a Rep expression plasmid (pRep) and Cap expression plasmid (pCap), and 2) a large Rep expression plasmid (pR78/68) and small Rep plus Cap expression plasmid (pR52/ 40Cap). When AAV vectors were produced using these sets of split-type helper plasmids at various ratios, the optimal ratio of (large) Rep expression plasmid and Cap expression plasmid was 1 to 9 for both sets. More importantly, the titers were comparable to or even higher than that of a conventional helper plasmid (pIM45) (4.9 ± ± ± ±2.1× × × ×10 11 vector particles/10 cm dish for pRep and pCap; 2.9 ± ± ± ±1.6× × × ×10 11 vector particles/10 cm dish for pR78/68 and pR52/40Cap; and 1.8 ± ± ± ±0.16× × × ×10 11 particles/10 cm dish for pIM45). Western analysis of AAV proteins suggests that the expression of a relatively small amount of large Rep and a large amount of Cap is important for optimal vector production. The present study shows that the AAV helper plasmid can be split without losing the ability to package the recombinant AAV genome, and provides us with valuable basic information for the development of efficient AAV packaging cell lines.