Manufacturing of adenovirus vectors: production and purification of helper dependent adenovirus.

Adenoviral vector (AdV) of the third generation also known as helper-dependent adenoviral vector (HDV) is an attractive delivery system for gene therapy applications. However, obtaining high quality-grade HDV in sufficient amount remains a challenge that hampers the extensive use of this vector in preclinical and clinical studies. Here we review recent progress in the large-scale manufacturing of HDV. The production of HDV is now amenable to large-scale volume with reduced process duration under optimized rescue and co-infection conditions. Also, efficient downstream processing of HDV with acceptable recovery of HDV and minimal contamination by the helper virus is described.

An efficient process for the purification of helper-dependent adenoviral vector and removal of helper virus by iodixanol ultracentrifugation.

The preparation of large amount of purified helper-dependent adenoviral vector material is hampered by the lack of development of downstream processes with proven records on separation and recovery efficiencies. In order to facilitate the use of clinical-grade helper-dependent virus material for large-scale in vivo studies, a three-step purification scheme consisting of (1) an anion-exchange chromatography for initial capturing of virus, (2) a shallow iodixanol density gradient ultracentrifugation for the removal of helper virus from helper-dependent virus, and (3) a size-exclusion chromatography for the removal of iodixanol and residual protein contaminants as a polishing step was developed. The use of a fast iodixanol density ultracentrifugation step was highly effective in separating infectious helper-dependent virus from contaminating helper virus. The overall downstream processing scheme gave 80% infectious particle yield. The contamination ratio of helper virus in the helper-dependent virus preparation are reduced from 2.57 to 0.03% corresponding to a reduction of helper virus by factors of 85 by two iodixanol purification steps. It was also demonstrated that size-exclusion chromatography is an excellent step for the removal of iodixanol and polishing of the final helper-dependent virus preparation.

Development of a suspension serum-free helper-dependent adenovirus production system and assessment of co-infection conditions.

Helper-dependent adenovirus (HDAd), deleted in all viral protein-coding sequences has been designed to reduce immune response and favor long-term expression of therapeutic genes in clinical programs. Its production requires co-infection of E1-complementing cells with helper adenovirus (HAd). Significant progresses have been made in the molecular design of HDAd, but large scale production remains a challenge. In this work, a scalable system for HDAd production is designed and evaluated focusing on the co-infection step. A human embryo kidney 293 (293) derived cell line, the 293SF/FLPe was generated to produce efficiently HDAd while restricting the packaging of HAd. This cell line was adapted to grow in suspension and in serum-free medium. Multiplicity of infection (MOI) of HDAd ranging from 0.1 to 50 was evaluated in presence of HAd at a MOI of 5. Optimal MOIs for HDAd amplification were found in the range of 5-10. HAd contamination was only 1%. These results were validated in a 3 L bioreactor under controlled operating conditions where a higher HDAd yield of 2.6 x 10(9) viral particles (VP)/mL or 3.5 x 10(8) infectious units (IU)/mL of HDAd was obtained.

Insights into adenoviral vector production kinetics in acoustic filter-based perfusion cultures.

One of the major limitations in the production of adenoviral vectors is the reduction in cell-specific productivity observed for increasing cell density at infection in batch cultures. This observation strongly suggests some nutrient depletion and/or metabolite inhibition in the media. These limitations have been partially overcome through other feeding strategies, such as fed-batch and sequential batch operations. To improve these results, we evaluated perfusion as a strategy to increase the volumetric productivity of HEK-293 cell cultures, by allowing productive infection at higher cell densities. An acoustic cell separator was employed in consideration of the increased shear sensitivity of the cells during the infection phase. The effects of perfusion rate and cell density at infection on the production of a recombinant adenovirus expressing the GFP were investigated. The perfusion mode allowed successful infection at cell densities in the range of 2.4-3 x 10(6) cell/mL, while maintaining a similar cell specific productivity (17,900 +/- 2400 VP/cell) to that of a batch infected at a low cell density (5 x 10(5) cell/mL). The highest virus concentrations (4.1 +/- 0.6 x 10(10) VP/mL) were attained for a feed rate of 2 vol/d and constituted a fivefold increase compared to a batch with medium replacement. Rapid assessment of the infection status was achieved through the use of on-line monitoring of respiration, fluorescence, and biovolume. Analysis of the kinetics of nutrient consumption and metabolite production revealed that a reduction in specific productivity is correlated with reduced metabolic activity.