July 1, 2018 (Vol. 38, No. 13)
Henry Chiou Ph.D. Associate Director Thermo Fisher Scientific
Jonathan F. Zmuda Ph.D. Director of R&D Thermo Fisher Scientific
Maya Yovcheva R&D Scientist and R&D Lead Thermo Fisher Scientific
Kenneth Thompson Ph.D. R&D Scientist Thermo Fisher Scientific
Sara Barnes R&D Scientist Thermo Fisher Scientific
Katalyn Irvin R&D Scientist Thermo Fisher Scientific
Melissa Cross R&D Scientist Thermo Fisher Scientific
Natasha Lucki Ph.D. Product Manager Thermo Fisher Scientific
Employing the ExpiSf™ Chemically Defined Sf9 Insect Cell Expression System
The baculovirus expression vector system (BEVS) provides a versatile platform for the expression of individual recombinant proteins as well as multimeric protein complexes, virus-like particles, membrane proteins, and proteins that are toxic to mammalian cells. Recently, BEVS has demonstrated particular utility in the commercial-scale manufacture of vaccines for human and veterinary applications.
Among the advantages of BEVS are the ability to express proteins with enhanced post-translational modifications compared to bacteria, lower costs of reagents and fewer equipment needs compared to mammalian expression, and a demonstrated record of scalability for the commercial production of vaccines and cell therapy reagents.1
BEVS platforms, however, continue to possess a number of shortcomings compared to plasmid-based mammalian transient systems including: lot-to-lot variability of the expression media due to the presence of undefined components such as yeastolate, lower protein titers, and longer overall time to protein. These factors can significantly hamper product development at a time when manufacturers are under continual pressure to develop vaccines and other biologics more quickly and efficiently.
Unlike many mammalian expression systems that transitioned to chemically-defined media formulations more than a decade ago, insect expression systems have continued to rely on poorly defined, yeastolate-containing media formulations that can impart significant lot-to-lot variability on cell growth, baculovirus production, and protein expression.
Yeastolate-containing media are further limited in their capacity to sustain high-density cell growth and consequently support high-cell-density baculovirus infections, two critical aspects required to increase biomass and improve protein titers on a per volume basis.
Compared to mammalian transient protein expression, the time from gene to protein for insect systems is significantly longer, typically on the order of 3–4 weeks, due primarily to the time required to generate a baculovirus stock using adherent Sf9 cells followed by subsequent amplification of virus through P1, P2, and/or P3 stocks to obtain sufficient virus for protein expression runs.
To address the shortcomings of traditional insect expression systems, the ExpiSf™ Expression System was developed to eliminate the variability associated with yeastolate-containing culture medium, to increase protein titers, and to reduce total time to protein—all to streamline research such as vaccine development from bench to clinic.
Similar to the Expi293 and ExpiCHO mammalian expression systems, a systems-based approach was employed during development of the ExpiSf Expression System.
As an initial step in this process, the first chemically defined (CD), yeastolate-free insect cell culture medium, ExpiSf CD Medium, was developed to support high-density cell growth, high-titer virus production, and enhanced protein expression. Multiple lots of ExpiSf CD Medium were formulated and compared for cell growth and protein expression levels; consistent growth kinetics (Figure 1A) and protein titers (Figure 1B) were obtained across four different lots of ExpiSf CD Medium.
Next, Gibco Sf9 cells were adapted into ExpiSf CD Medium through extensive long-term passaging to generate the high-density ExpiSf9 cell line. ExpiSf9 cell growth characteristics were compared to traditional Sf9 cells by culturing Sf9 cells in various yeastolate-containing media for at least 10 passages to ensure adaptation.
Compared to Sf9 cells grown in yeastolate-containing media, ExpiSf9 cells grown in ExpiSf CD Medium achieved higher peak viable cell densities in standard shake flask cultures (>20 × 106 cells/mL vs. 2–10 × 106 cells/mL; Figure 1C), approximately double the peak density of the next highest density media formulation. ExpiSf9 cells also possessed a broad log-phase growth range spanning from approximately 4–12 × 106 viable cells/mL, with a stable doubling time of ~24 hours.
Taking advantage of the higher densities that ExpiSf9 cells achieve in ExpiSf CD Medium, an expression enhancer (ExpiSf Enhancer) was developed to work in conjunction with the ExpiSf CD Medium to allow for consistent, high-density infection of ExpiSf9 cells at 5 × 106 cells/mL, leading to multifold improvements in protein titers compared to traditional Sf9 workflows, in which cells are infected at 1–2 × 106 cells/mL in yeastolate-containing media (Figure 1D).
Since optimal baculovirus infection of Sf9 cells is thought to occur when cells are in log-phase growth, the extended log-phase growth range of the ExpiSf9 cells, in conjunction with the ExpiSf Enhancer, allows for ExpiSf9 cells to be infected at significantly higher densities than traditional insect workflows, thereby increasing protein titers on a per volume basis.
Lastly, to reduce overall time to protein, the ExpiFectamine™ Sf Transfection Reagent was developed to allow for gentle, nontoxic transfection of high-density ExpiSf9 suspension cultures with large bacmid DNA. Incorporating suspension-based transfection for baculovirus generation enabled easy preparation of 100 mL (or greater) of high titer (1–5 × 109 infectious virus particles/mL), high-quality P0 virus, completely eliminating the need for additional virus amplification for typical liter and subliter workflows.
Additionally, suspension-based bacmid transfection shortens overall time to protein by up to 50% (Figure 1E) by eliminating the need for virus amplification while at the same time removing the risk of deleterious virus passaging effects (whereby gene incorporation/protein expression levels decrease during generation of P1+ virus stocks), ensuring that the highest quality baculovirus is used for protein expression runs.
Figure 1. Performance characteristics of the ExpiSf Expression System. (A) Four different lots of ExpiSf CD Medium demonstrated consistent growth of ExpiSf9 cells with peak viable cell densities (VCDs) of ~20 × 106 cells/mL. (B) Green fluorescent protein (GFP) titers were 4–5-fold higher in all lots of ExpiSf CD Medium tested compared to a traditional Sf9 workflow using yeastolate-containing (YC) medium. (C) ExpiSf9 cells in ExpiSf CD Medium exhibited superior cell growth compared to five yeastolate-containing insect cell media. (D) ExpiSf Enhancer, used in conjunction with ExpiSf CD Medium and ExpiSf9 cells, generated 3-fold higher GFP titers than a traditional Sf9 workflow; ExpiSf Enhancer nearly doubled protein titers compared to the ExpiSf System without enhancer addition. (E) The ExpiSf Expression System reduced the time required to go from bacmid DNA to protein expression by half by eliminating the need for virus amplification via direct generation of high-volume and high-titer P0 virus stocks.
Comparison of ExpiSf to Traditional Sf9-Based Workflows
The expression levels of three different proteins—green fluorescence protein, human Fc fusion protein, and tumor necrosis factor-alpha (TNF-a)—were compared between the ExpiSf system and a traditional Sf9 workflow in which cells are infected at 1–2 × 106 cells/mL in yeastolate-containing media.
Compared to five different yeastolate-containing media tested using traditional Sf9 workflows, the high-density ExpiSf Expression System generated 3–5-fold improvements in expression levels across the three proteins tested (Figure 2A).
G-protein-coupled receptors are commonly expressed in insect cells, in part because of the potential for toxicity in mammalian cells, as well as the desire for reduced glycosylation for structural biology studies. Cannabinoid receptor 2 (CB2) was expressed in the ExpiSf system and in a traditional Sf9 workflow.
Optimal cell harvest time was determined to be 48 hours post infection in the ExpiSf corresponding to a cell viability of 70% at the time of harvest; longer infection times led to decreased viability without improving CB2 expression per cell (Figures 2B & 2C). CB2 expression (as measured by total number of CB2 molecules per cell by quantitative flow cytometry) was 10-fold higher in the ExpiSf Expression System compared to the traditional Sf9 workflow (Figure 2D).
This improvement was due to both per cell increases in expression levels as well as the significantly higher density of ExpiSf9 cells in a given volume compared to the traditional Sf9 workflow.
Figure 2. Comparison of the ExpiSf Expression System to traditional Sf9 workflows using various yeastolate-containing media. (A) Expression levels of an Fc fusion protein, green fluorescent protein (GFP), and tumor necrosis factor-alpha (TNF-a) were on average >4, >3, or >5-fold higher in the ExpiSf expression system than those obtained using various yeastolate-containing media in a traditional Sf9 workflow. (B) Optimization of CB2 G-protein-coupled chemokine receptor harvest time and (C) post-infection viability kinetics in the ExpiSf Expression System. (D) Increased total CB2 expression levels obtained in the ExpiSf Expression System compared to traditional Sf9 workflow in Sf-900 II medium; increased expression of CB2 is due to both higher per cell expression as well as greater cell density in a given volume for the ExpiSf Expression System.
Glycosylation Patterns and Protein Functionality
Although high protein yields are desirable, resultant proteins are less valuable if they are aggregated, misfolded, degraded, or improperly glycosylated. The quality of the proteins expressed in the ExpiSf system was compared to the quality of the same proteins expressed using a traditional Sf9 workflow with yeastolate-containing medium.
Using secreted alkaline phosphatase (SEAP) as a model protein, glycosylation patterns generated in the ExpiSf were shown to be highly comparable to those of a traditional Sf9 insect workflow, with the predominate glycoforms being Man3, Man3F, and Man6 in both instances (Figure 3A). SDS-PAGE showed a single band with a molecular weight of ~57 kD for both workflows (Figure 3B).
The biological activity of TNF-α expressed in the ExpiSf system and by traditional Sf9 workflow was assessed using an NF-κB luciferase reporter gene assay. HIS-tagged TNF-a was expressed and purified by Ni-NTA, and its concentration was determined by A280. TNF-a was expressed at >4-fold higher levels in ExpiSf compared to the traditional Sf9 workflow (Figure 3C); reporter gene assay results demonstrated equivalent biological activity (relative luminescence units; RLUs) for the proteins, respectively (Figure 3D).
In summary, the ExpiSf Expression System represents a significant advance in insect cell protein expression in terms of media consistency, protein yield, and time. It enables researchers to streamline protein expression and vaccine development to shorten time lines from bench to clinic.
Figure 3. Comparison of glycosylation patterns and biological activity for proteins expressed in the ExpiSf Expression System and by traditional Sf9 workflow. (A) Glycosylation patterns of secreted alkaline phosphatase (SEAP) were highly similar in the ExpiSf Expression System and in a traditional insect workflow using Sf900-II medium. (B) SDS-PAGE of SEAP purified from the ExpiSf Expression System and a traditional Sf9 workflow using Sf900-II medium. (C) The ExpiSf Expression System generated >4-fold higher TNF-a expression levels compared to a traditional Sf9 workflow. (D) TNF-a activity, as measured by luciferase-based NF?B reporter gene assay, showed equivalent biological response for protein generated in the ExpiSf Expression System and by traditional Sf9 workflow.
1. Felberbaum RS. The baculovirus expression vector system: A commercial manufacturing platform for viral vaccines and gene therapy vectors. Biotechnol. J. 2015. 10: 702–714.
Maya Yovcheva is an R&D scientist and R&D lead for the ExpiSf Expression system at Thermo Fisher Scientific. Her colleagues are R&D scientists Kenneth Thompson, Ph.D., Sara Barnes, Katalyn Irvin, and Melissa Cross. Natasha Lucki, Ph.D., is a product manager, Henry Chiou, Ph.D., is an associate director, product management, and Jonathan F. Zmuda, Ph.D. (Jon.Zmuda@thermofisher.com), is a director of R&D.
The ExpiSf™ Expression System is for Research Use Only. Not for use in diagnostic procedures. For additional information about the ExpiSf system, please go to: thermofisher.com/expisf.