Plasmid fermentation for gene therapy and vaccination

Plasmid-based gene therapy and vaccination require the production of purified plasmid DNA. There is a current understanding that supercoiled plasmid DNA is the best form for administration and FDA guidelines state that a specification for the minimum level of supercoiled plasmid DNA, in the final product, should be made. Currently, plasmids of 10 kb or smaller are being used in gene therapy trials, but if plasmid-based vaccination is to become a reality then much larger plasmids, with the ability to carry multi-variant genes, will be required, this raises questions of feasibility of production. This study examined two main issues. Firstly a fed-batch strategy to maximise the level of supercoiling of a 6.9 kb plasmid resulting from fermentation of a recombinant Escherichia coli strain was developed. Secondly issues relating to the production of larger plasmids, for gene therapy and vaccination, were examined. A fed-batch fermentation strategy for the production of a 6.9 kb plasmid, pSVp, in £ coli DH5a, on a semi-defined medium, was developed. Starvation of amino acids was shown to induce plasmid amplification and in batch fermentation a maximum biomass of 3.5 g/L dry cell weight (DCW) and maximum plasmid yields of 40 mg/L and 11.3 mg/g DCW were achieved. However, amplified plasmid levels were not maintained. After 31 h 90 % of plasmid was in the supercoiled form but only remained so for 1 h. Maximum plasmid yield and maximum plasmid supercoiling did not occur simultaneously. Hence a strategy to delay amplification was investigated. A dual feeding strategy consisting of a linear amino acids feed and an exponential D-glucose feed was employed. In fed-batch culture a mean maximum biomass of 4.9 g/L dry cell weight and mean maximum plasmid yields of 44 mg/L and 9.1 mg/g DCW were achieved. 90 % of plasmid was in the supercoiled form after 25 h and remained at this level until harvest at 35 h. An important consideration in the production of large plasmids is the ability to supply sufficient oxygen to the fermentation. E. coli DH5a harbouring either the plasmid pBGS18, a 4.4 kb pUCbased plasmid, or pQR150, a 20 kb derivative of pBGS18, was grown in D-glucose-limited chemostat culture to investigate the effects of plasmid size and recombinant protein production on oxygen demand. Under conditions where no recombinant protein was expressed the cellular oxygen demand of the two strains was not significantly different. When recombinant protein was expressed cells harbouring pBGS18 demonstrated a statistically insignificant increase in mean specific oxygen uptake rate while those harbouring pQR150 demonstrated a statistically significant increase. It was concluded that plasmid size does not significantly affect oxygen demand. The increase in oxygen demand reported with an increase in plasmid size by other researchers is due to production of recombinant protein. The production of a 116 kb plasmid, p5176, in E. coli DH10p on complex and semi-defined media was investigated. Fermentations on complex medium were poorly reproducible while those on semi-defined medium were highly reproducible. Maximum plasmid yields were comparable between the two media, while higher biomass yields and lower oxygen requirements were observed with semi-defined medium. Lysis and primary recovery were investigated and indicated that, using current methods, the manufacture of a large plasmid product may be difficult. A fed-batch fermentation strategy which allows an increased yield of supercoiled plasmid DNA has been developed. It has been established that oxygen supply is not an issue in the production of large plasmids and a base for the production of large plasmids has been established for future work.