Expression of hepatitis B surface antigen in the methylotrophic yeast Pichia pastoris using the GAP promoter

High-level expression and efficient assembly of Hepatitis B surface Antigen (HBsAg) particles have been reported in Pichia pastoris by integrating a single copy of the HBsAg gene under the control of the alcohol oxidase (AOX1) promoter. However, the time taken to reach peak product concentration is usually very long ( approximately 240 h). In this paper, we describe the expression of HBsAg in P. pastoris using the recently described glyceraldehyde-3-phosphate dehydrogenase (GAP) promoter. Unlike the previously described AOX1 promoter based system (in which biomass is generated first followed by methanol-induced antigen production), biomass generation and antigen production occur simultaneously in medium containing glycerol or glucose. Maximal levels of HBsAg expression in case of the single copy AOX1 integrant (attained after 6 days of induction) exceeded the levels of antigen produced by the single copy GAP integrant. However, this was offset by continuous antigen production by the GAP clone. In an attempt to further enhance antigen production levels of the GAP clones, we isolated multicopy Pichia integrants containing up to four copies of the GAP promoter-driven constitutive expression cassette using the Zeocin screening procedure. The data demonstrated a direct correlation between the gene dosage and the levels of HBsAg expressed by the GAP clones. The effect of copy number was additive and the four copy clone resulted in about four-fold higher yield of HBsAg. The majority of HBsAg produced in the constitutive expression system was found to be of particulate form, based on sedimentation behaviour and particle-specific ELISA, suggesting that it has the potential to serve as an effective immunogen. These particles were sensitive to thiol reagents. We also explored the possibility of secreting the GAP expressed HBsAg in P. pastoris. In-frame fusion of the Saccharomyces cerevisiae alpha-factor secretion signal under the constitutive GAP promoter resulted in secretion of approximately 20 nm HBsAg particles as evidenced by electron microscopy. However, the levels of secreted HBsAg particles were very low, presumably due to the inherent hydrophobicity of the HBsAg molecule and the consequent propensity for membrane association. Our studies show that secretion is not a good strategy for expression of HBsAg in P. pastoris. The data also suggests that intracellular production of HBsAg under the GAP promoter using multicopy expression cassettes can indeed serve as an effective alternative to the AOX1 promoter. Further, the GAP promoter based system obviates the need to use and extensively monitor methanol during recombinant antigen production. Finally, this constitutive system has the potential for continuous culture wherein several batches of recombinant protein-containing biomass can be harvested from a single initial fermentation.