High-level transient expression, extraction, and purification of recombinant beta-glucuronidase from agroinfiltrated lettuce

Transient (temporary) recombinant protein expression in plant tissues was investigated as a potential alternative to stable expression in transgenic plants. Agroinfiltration followed by transient expression is a system whereby post-harvest, non-transgenic plant material is infused with a suspension of bacteria (Agrobacterium tumefaciens) modified to include a gene for the desired protein. The bacteria are capable of transferring the gene to plant cell nuclei and within days the desired protein is produced by the plant.; Scouting studies evaluated vacuum infiltration of five types of germinated seeds and eleven types of leaves with three bacterial strains for expression of the beta-glucuronidase (GUS) gene. The activity of the resulting GUS protein was measured using histochemical and spectrophotometric assays and a novel digital imaging method. Romaine lettuce and A. tumefaciens strain C58C1 were selected as a platform to develop engineering data on agroinfiltration as a production system. Factors including light, bacterial concentration, temperature, vacuum, and pH were tested to establish operating conditions. The best conditions produced 0.16% GUS protein based on dry lettuce tissue. Lettuce leaves expressing GUS were extracted with several buffer combinations, and the yield and stability were determined. Dithiothreitol was found to be the most important component in the extraction buffer. Of three models examined, a first-order kinetic model best described inactivation of GUS, explaining the increase in GUS inactivation rate in freeze-dried leaves with increasing temperature. Anion exchange chromatography was compared to aqueous two-phase extraction (ATPE) for purification of GUS. Separation of spiked E. coli-derived GUS from lettuce extract using chromatography and ATPE was virtually identical for yield and purification efficiency. However, for lettuce-derived GUS, separation by chromatography produced non-Gaussian behavior and ATPE revealed different partition coefficients compared to the spiked E. coli-derived GUS. A hypothesis is that a portion of the active GUS produced by transient expression is subtly different than E. coli-derived GUS due to post-translational modification to the protein that occurs in plants but not bacteria.