Enhanced protein expression in the baculovirus/insect cell system using engineered SUMO fusions

Abstract

Recombinant protein expression in insect cells varies greatly from protein to protein. A fusion tag that is not only a tool for detection and purification, but also enhances expression and/or solubility would greatly facilitate both structure/function studies and therapeutic protein production. We have shown that fusion of SUMO (small ubiquitin-related modifier) to several test proteins leads to enhanced expression levels in Escherichia coli. In eukaryotic expression systems, however, the SUMO tag could be cleaved by endogenous desumoylase. In order to adapt SUMO-fusion technology to these systems, we have developed an alternative SUMO-derived tag, designated SUMOstar, which is not processed by native SUMO proteases. In the present study, we tested the SUMOstar tag in a baculovirus/insect cell system with several proteins, i.e. mouse UBP43, human tryptase beta II, USP4, USP15, and GFP. Our results demonstrate that fusion to SUMOstar enhanced protein expression levels at least 4-fold compared to either the native or His(6)-tagged proteins. We isolated active SUMOstar tagged UBP43, USP4, USP15, and GFP. Tryptase was active following cleavage with a SUMOstar specific protease. The SUMOstar system will make significant impact in difficult-to-express proteins and especially to those proteins that require the native N-terminal residue for function.

Figure 1

Analysis of the expression of GFP in Sf9 cells. Sf 9 cells (2 × 10⁶) were infected at an MOI of 0.05 with baculovirus harboring various GFP constructs. The cells were harvested at 120 hours post infection and lysed. (A) Equal amounts total, soluble protein were analyzed by SDS-PAGE and the separated proteins stained with Coomassie Blue. (B) Cell lysates expressing gp67His₆SUMOstar- and gp67His6-GFP constructs were also analyzed by western blot using an anti-His₆ antibody as the primary antibody. Images of the stained gels or western blots were made using a Fuji imager. Fuji image analysis software was used to estimate the relative amounts of each protein. Molecular weights, in kDa, of the marker proteins are shown next to each image. N.b. although the calculated MW of SUMO is 11.5 kDa, it migrates with an apparent MW of ~20 kDa on SDS-PAGE. The positions of native GFP and His₆SUMOstar-GFP are indicated. Gp67 = the gp67 signal sequence, His6 = the hexahis tag, SUMO = the SUMO fusion partner; and SUMO* = the SUMOstar fusion partner.

Figure 2

Analysis of the expression of UBP43 in Sf9 cells. Cells (10⁴) in 24-well plates were infected with baculovirus harboring either His₆-UBP43 (H), SUMOstar-UBP43 (S*), or His₆SUMOstar-UBP43 (HS*) at different MOI as indicated. The cells were harvested at 140 hours post infection and SDS-PAGE analysis as described in the Materials and Methods. Equivalent amounts of cells were loaded in each lane. Following electrophoresis, the separated proteins were stained with Coomassie Blue. Image collection and analysis was carried out as described in the legend to Fig. 1.

Figure 3

Analysis of the secretion of Tryptase βII and GFP into the medium by Sf9 cells. (A) Cells (10⁴) were infected with baculovirus harboring either gp67-Tryptase (G), gp67SUMO-Tryptase (GS) or gp67SUMOstar-Tryptase (GS*) at MOIs of 0.1, 1 or 5 in 6 well plates. The corresponding fusions to GFP were only analysed at an MOI of 1. Conditioned medium from each well was collected at 72 hr post infection and clarified by centrifugation at 300 × g for 10 min. The clarified samples were concentrated using a 0.5 ml Microcon concentrator (MWCO 10 kDa, Millipore, Billerica, MA) at 14,000 × g for 40 min. Equal amounts of the concentrated media were loaded in each lane. (B) Aliquots of the cells expressing either G-Tryptase or GS*-tryptase were lysed and analyzed by SDS-PAGE in order to determine if the fusion proteins were being retained within the cells. Following electrophoresis, the separated proteins were stained with Coomassie Blue. Image collection and analysis was carried out as described in the legend to Fig. 1.

Figure 4

Isolation secreted SUMOstar-Tryptase from Sf9 conditioned medium. Secreted gp67His₆SUMOstar-Tryptase from 100 mL of conditioned medium was captured in batch mode by mixing with Ni-NTA as described in the Materials and Methods section. (A) SDS-PAGE analysis of aliquots of the starting condition medium (L), the unbound material (FT), the Wash (W) and pooled and concentrated eluate fractions (E). (B) The pooled and concentrated eluate fraction was assayed for tryptase activity as with and without removal of the SUMOstar fusion partner by incubation with SUMOstar protease using Chromozym TH as a substrate for tryptase as described in the Materials and Methods section. (●) Complete reaction mixture with SUMOstar protease, heparin, and SUMOstar-tryptase. ( ) Reaction mixture with heparin and SUMOstar-tryptase but without SUMOstar protease. ( ) Reaction mixture minus SUMOstar-tryptase and SUMOstar protease. In a separate experiment SUMOstar protease was found to have no activity against the tryptase substrate (data not shown.)

Figure 5

Isolation of intracellular USP4 and USP15 and activity determination. Isolation of His₆SUMOstar-enzymes from Sf9 cell lysates was carried out as described in the Materials and Methods. (A) SDS-PAGE analysis of aliquots of samples from the isolation of USP4, L = the clarified cell lysate, FT = the unbound fraction from the Ni-NTA column, W = the low imidazole wash from the column and E = the pooled and concentrated elution fractions. The expected position of His₆SUMOstar-USP4 is indicated. (B) SDS-PAGE analysis of aliquots of samples from the isolation of USP15. Samples labed as per (A). The expected position of His₆SUMOstar-USP15 is indicated. The ability of each purified enzyme to process Ub-PLA2 to release active PLA2 was carried out as described in the Materials and Methods. (C) Ub-PLA2 processing activity of either 5 ( ) or 10 ( ) μL of the concentrated pool of USP4. (D) Ub-PLA2 processing activity of either 5 ( ) or 10 ( ) μL of the concentrated pool of USP15. For both, the line represents the fit of the data to equation 1. Boxed points were excluded from the calculation due to depletion of the PLA2 substrate.