Efficient recombinant adeno-associated virus production by a stable rep-cap HeLa cell line correlates with adenovirus-induced amplification of the integrated rep-cap genome

Background A possible procedure for the production of clinical grade recomb inant adeno-associated virus type 2 (rAAV) would include the use of packagi ng cell lines, harboring the rep-cap gene and the vector, combined with a r eplication defective adenoviral plasmid to provide the helper activities. S everal studies have already shown that rAAV can be efficiently assembled by infecting the stable packaging cell line with adenovirus. However, the dir ect comparison with an adenoviral plasmid has never been reported. Methods To investigate this point, a clone of HeLa and 293 cells harboring one to two rep-cap copies per cell genome (HeRC32 and 293RC21, respectively ) were generated. Recombinant AAV was produced by transiently transfecting the AAVCMVLacZ vector and supplying the adenoviral helper activities by eit her wild-type adenovirus or an adenoviral plasmid (pAdc). As a control, rAA V was similarly produced from naive HeIa and 293 cells additionally transfe cted with a rep-cap plasmid. Results Despite satisfactory rAAV yields from Hela and 293 cells, we show t hat those from HeRC32 and 293RC21 cells dramatically decrease when adenovir us is replaced by the adenoviral plasmid (pAdc). The analysis performed to identify the factors hampering efficient rAAV assembly by HeRC32 cells in t he presence of pAdc shows that: (1) while upon adenovirus infection the int egrated rep-cap genome undergoes a dramatic amplification leading to a 100- fold increase in the rep-cap copy number, no amplification is detected upon transfection of pAdc; (2) in pAdc-transfected HeRC32 cells, the intracellu lar localization of the adenovirus E4orf6 and E1B-55kDa proteins is abnorma l as compared to adenovirus-infected cells. Conclusions This study documents that stable rep-cap cells lines are severe ly hampered for rAAV assembly when a replicative adenovirus is substituted with an adenoviral plasmid. Furthermore, our results also suggest that the lack of amplification of the rep-cap genes, eventually combined with the al tered distribution of the adenoviral proteins, E4orf6 and E1B-55kDa, is rel ated to the low rAAV yields observed under these conditions. Copyright (C) 2000 John Wiley & Sons, Ltd.