An approach to rapid distributed manufacturing of broad spectrum anti-viral griffithsin using cell-free systems to mitigate pandemics

Abstract

This study describes the cell-free biomanufacturing of a broad-spectrum antiviral protein, griffithsin (GRFT) such that it can be produced in microgram quantities with consistent purity and potency in less than 24 h. We demonstrate GRFT production using two independent cell-free systems, one plant and one microbial. Griffithsin purity and quality were verified using standard regulatory metrics. Efficacy was demonstrated in vitro against SARS-CoV-2 and HIV-1 and was nearly identical to that of GRFT expressed in vivo. The proposed production process is efficient and can be readily scaled up and deployed wherever a viral pathogen might emerge. The current emergence of viral variants of SARS-CoV-2 has resulted in frequent updating of existing vaccines and loss of efficacy for front-line monoclonal antibody therapies. Proteins such as GRFT with its efficacious and broad virus neutralizing capability provide a compelling pandemic mitigation strategy to promptly suppress viral emergence at the source of an outbreak.

Keywords: Antiviral griffithsin; Cell-free protein synthesis (CFPS); Distributed manufacturing; Pandemic preparedness; SARS-CoV-2.

Figure 1.

Upper: Proof of concept for GRFT bioprocess (top) includes expression using a cell-free system (CFS), heat induced purification and polishing chromatography to yield active and pure GRFT for use at the next pandemic hotspot. Lower: A distributed manufacturing platform which can be deployed in a shipping container to make biologics safely and effectively at the site of the future viral outbreak. The container includes concentrated, dried solutions for reconstitution of cell-extracts, required CFPS reagents, and purification buffers, as well as validated analytics for real-time process monitoring and control.

Figure 2.

(a) GRFT yields from an E. coli CFPS system comparing two different codon-optimized constructs, tagless-GRFT.std and tagless-GRFT.opt. Reactions were performed in triplicate at a 25 μL scale overnight at 30 °C. C1, tagless-GRFT.std; C2, tagless-GRFT.opt; C3, tagless-GRFT.opt + 2 mM Serine/Glycine; C4, tagless-GRFT.opt with 50 mM KGlu (normal is 175mM); C5, tagless-GRFT.opt using BL21 Extract; C6, sfGFP control. (b) Tagged-GRFT yields from various plasmid concentrations in ALiCE^® CFPS system. Reactions were performed in triplicate at the 250 μL scale overnight at 25 °C. Total GRFT concentration was determined in pooled samples. Error bars represent standard deviation (SD) of the mean (n=3).

Figure 3.

(a) GRFT amounts in various harvest and purification fractions as estimated by Western blotting analysis (Supplementary Figure 3) in the primary capture step of GRFT purification for tagged- and tagless-GRFT constructs expressed in ALiCE^® and E. coli cell-free systems, respectively. PS1, tagged-GRFT Harvest; PS2, tagged-GRFT IMAC FT; PS3, tagged-GRFT IMAC W1; PS4, IMAC Elution; PS5, tagged-GRFT Post Precipitation; PS6, tagless-GRFT.opt Harvest; PS7, tagless-GRFT.opt Post Precipitation. PS1–7 were then utilized in a mass balance to determine protein recovery. (b) (%) recovery obtained for IMAC (tagged-GRFT, PC1) and heat-induced purification (tagged-GRFT, PC2 and tagless-GRFT.opt, PC3). Error bars represent SD of the mean (n=2).

Figure 4.

LC/MS spectra Total Ion Chromatogram (TIC), MS spectrum, and deconvolution of E. coli and ALiCE^® post precipitation products in sterile-filtered 1X PBS (pH 7.4) (Supplementary Method 1). (A) E. coli expressed tagless-GRFT.opt (MW 12.68 kDa) (Left-top). (B) ALiCE^® expressed tagged-GRFT (MW 14.53 kDa) (Middle) (C) NIH GRFT standard (MW 14.49 kDa) (Right-bottom).

Figure 5.

(a) Side by side comparison of gp120 ELISA for Tagged-GRFT (ALiCE^®) and tagless-GRFT.opt (E. coli) cell-free expressed products while evaluating binding at varied concentrations of GRFT. (b) Activity of GRFT in SARS-CoV-2 pseudovirus assays showing inhibition of pseudoviral entry. Samples tested in duplicate, and experiment repeated twice. Visualization using luciferase endpoint. Error bars represent SD of the mean (n = 4).