Expression and characterization of full length and truncated versions of major outercapsid protein VP2 of bluetongue virus in bacterial and insect cells

Abstract

The spread of bluetongue virus (BTV) to previously disease-free regions which prohibit the use of the current BTV live-attenuated vaccine has highlighted the need for a new generation of vaccines (Ferrari, De Liberato et al. 2005; Veronesi, Hamblin et al. 2005). Subunit vaccines are one of the attractive alternative strategies. Subunit vaccines against BTV would target the outercapsid protein VP2, the main neutralization-specific antigen (Huismans, van der Walt et al. 1987; Roy, Urakawa et al. 1990; Roy, French et al. 1992; Roy, Bishop et al. 1994). A subunit vaccine based on the use of BTV-VP2 may be achieved by either using VP2 by itself or by means of virus-like particles (VLPs) on which VP2 proteins are exposed. In VLPs, the VP2 is co-expressed with other capsid and core proteins to form a particle that resembles the intact BTV. The BTV-VLP vaccine strategy is advantageous since it presents the neutralizing epitopes of more than one viral protein in a more authentic manner as found on the virus itself (Huismans, van der Walt et al. 1987; Roy, Urakawa et al. 1990; Roy, French et al. 1992; Roy, Bishop et al. 1994). However there are difficulties associated with large scale production and a decrease in the stability of the particles over time (Berg, Difatta et al. 2005; Wang, Zhao et al. 2006). Studies have already demonstrated the vaccine potential of BTVVP2 by itself (Huismans, van der Walt et al. 1987; Roy, Urakawa et al. 1990; Roy, French et al. 1992; Roy, Bishop et al. 1994). However if BTV-VP2 is to be used by itself as a single subunit vaccine, it is important that the protein is expressed under conditions where it is correctly folded and soluble. Solubility refers to the capacity of the expressed antigen to fold into an ordered tertiary structure that authentically exposes the neutralizing epitopes to the immune system (Dinner, Sali et al. 2000; Dobson 2003). However non-native interactions within and between in vitro synthesized viral proteins such as BTV-VP2 often leads to protein aggregation or insolubility. The immune response against aggregated or insoluble proteins is generally very poor. This problem of aggregation and insolubility may be alleviated to an extent by generating truncated versions of the protein from which hydrophobic regions that promote aggregation have been deleted leaving only the major neutralizing epitopes of the antigen (Fukumoto, Xuan et al. 2003; Bonafe, Rininger et al. 2009; Liu, Zeng et al. 2009; Seo, Pyo et al. 2009). The focus of the research presented in this dissertation was to evaluate the solubility of full-length BTV(10)-VP2 and truncated versions thereof after expression in a prokaryotic and baculovirus-Sf9 expression system. The full-length BTV(10)-VP2 (956 amino acids) gene and genes encoding truncated versions of BTV(10)-VP2 i.e. BTV(10)-VP2(aa450) (amino acid 1 to 450) and BTV(10)- VP2(aa650) (amino acid 1 to 650) were cloned into the bacterial expression vector pET160-DEST and the baculovirus expression vector pDEST™8. The C-terminal hydrophobic regions which might contribute to aggregation or insolubility of the protein when expressed in vitro were deleted from these truncated BTV(10)-VP2 proteins. The truncated proteins however still contained BTV neutralizing epitopes that were predicted from literature. The prokaryotic expression of the full-length BTV(10)-VP2 and the other truncated recombinant BTV(10)-VP2 proteins was carried out in E. coli BL21 Star DE3 expression strain. The initial pilot expression study confirmed high level expression of the recombinant proteins. The study also revealed that these proteins were insoluble. The optimization of the prokaryotic expression in order to increase the yield of soluble proteins by means of differential inducer concentrations, fermentation temperature and harvesting times did not produce soluble BTV(10)-VP2 and truncated BTV(10)-VP2 proteins. Previous studies have demonstrated the role of L-arginine in the recovery of soluble proteins from aggregation by reversing aggregation (Tsumoto, Umetsu et al. 2003). However in the current study, arginine treatment of the inclusion body and bacterial lysate containing the BTV(10)- VP2 and truncated recombinant proteins did not release soluble proteins. No soluble recombinant BTV(10)-VP2 proteins were detected when the recombinant proteins were expressed in BL21 host cells over-expressing heat-shock proteins (hsps) and chemical chaperones. However when the different recombinant proteins were co-expressed with the molecular chaperones dnaK-dnaJ-GrpE, it resulted in a fraction of soluble recombinant BTV(10)-VP2 proteins. In particular, approximately 50% of the total expressed BTV(10)-VP2(aa450) protein was soluble while approximately 20% of the total expressed BTV(10)-VP2(aa650) and full-length BTV(10)-VP2 were found soluble when coexpressed with dnaK-dnaJ-GrpE chaperones. These recombinant proteins could be eluted from a nickel affinity column further confirming that these proteins are in fact soluble. Interestingly the coexpression of the BTV(10)-VP2(aa450) protein with the above chaperones in combination with chaperones groEL-groES or only groEL-groES did not produce any soluble proteins. Baculovirus-insect expression of the aforementioned BTV(10)-VP2 recombinant proteins was carried out in Spodoptera frugiperda 9 (Sf9) cells. High level expression of the recombinant proteins was confirmed by an initial pilot expression study conducted at 42 hours post infection (p.i.). The pilot study also revealed that the recombinant proteins were insoluble. Arginine treatment of the lysate released a small fraction of soluble BTV(10)-VP2(aa450) and BTV(10)-VP2(ORF) proteins only detectable with immunoblot analysis using the anti-BTV(10) IgY antibodies. The amount of solubilized proteins was however too small to justify the cost associated with this expression system.