A stable platform for the production of virus-like particles pseudotyped with the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) spike protein

Abstract

In this study, we showed that a codon optimized version of the spike (S) protein of SARS-CoV-2 can migrate to the cell membrane. However, efficient production of Moloney murine leukemia (MLV) infectious viral particles was only achieved with stable expression of a shorter S version in C-terminal (ΔS) in MLV Gag-pol expressing cells. As compared to transient transfections, this platform generated viruses with a 1000-fold higher titer. ΔS was 15-times more efficiently incorporated into VLPs as compared to S, and that was not due to steric interference between the cytoplasmic tail and the MLV capsid, as similar differences were also observed with extracellular vesicles. The amount of ΔS incorporated into VLPs released from producer cells was high and estimated at 1.25 μg/mL S2 equivalent (S is comprised of S1 and S2). The resulting VLPs could potentially be used alone or as a boost of other immunization strategies for COVID-19.

Keywords: COVID-19; Coronavirus; Moloney; Retrovirus; SARS-CoV-2; VLP; Vaccine; Virus-like particle.

Fig. 1

Expression of S protein at the surface of 293 cells. FACS analysis of untransfected and transiently transfected cells with plasmids encoding the Galv envelope, the full-length S protein, and the ΔS version. S was detected with an anti-S1 antibody. The results shown are representative of three independent experiments.

Fig. 2

Transduction efficiency of different GFP pseudotyped vectors produced in transient transfections. (A) Expression of ACE2 at the surface of 293-ACE2 cells measured by FACS analysis with an anti-ACE2 antibody. The dark grey histogram represents parental 293 cells. The results shown are representative of two independent experiments. Two days after infection of 293-ACE2 cells, titers of VSV-G-, Galv-, S- and ΔS-pseudotyped vectors were (B) measured by FACS analysis or (C) evaluated by fluorescence microscopy. Values presented are the mean ± SD of three independent experiments. Titers of S- and ΔS-pseudotyped viruses were too low to be detected by FACS analysis. Fluorescent and bright-field pictures are displayed. The envelope pseudotype and the volume used for infection are indicated. ND: not determined.

Fig. 3

Characterization of stable VLPs producer cells. (A) S expression was measured by FACS analysis of 293GP, 293GP-S and 293GP-ΔS cells with an anti-S1 antibody. (B) GFP fluorescence of 293GP-Galv/GFP, 293GP-S/GFP and 293GP-ΔS/GFP measured by FACS analysis. The results shown are representative of three independent experiments.

Fig. 4

Transduction efficiency of GFP pseudotyped vectors released from stable producers. (A) Fluorescent and bright-field pictures are displayed. The envelope pseudotype and the volume used for infection are indicated. (B) Titers of Galv-, S- and ΔS-pseudotyped vectors produced from stable producers were measured by FACS analysis two days after infection. Values presented are the mean ± SD of three independent experiments. ND: not determined.

Fig. 5

Quantification of S and ΔS incorporated into VLPs. (A) Western blot analysis from concentrated supernatants of 293GP and 293 cells using anti-S2 and anti-p30 antibodies. Different amounts of Fc-tagged S2 were also loaded on the gel to quantify S2 in VLPs. (B) Differences between S and ΔS incorporation into VLPs. All the bands detected by the anti-S2 antibody in S- and ΔS-containing samples were quantified and normalized with the signal obtained for MLV p30. Values presented are the mean ± SD of three independent experiments analyzed twice in Western blot. (C) Western blot analysis of SARS-CoV-2 S protein in cellular extracts. Signals for S2, S, and multimeric forms of S were detected with the anti-S2 antibody. The Gag precursor pr65 was detected with the anti-p30 antibody, and β-tubulin was assessed for loading comparison.

Fig. 6

Incorporation of SARS-CoV-2 ΔS into MLV VLPs. Western blot analysis with antibodies against S2 and p30 on collected fractions separated with an iodixanol velocity gradient of (A) 293-ΔS and (B) 293GP-ΔS supernatants. The arrow below the blot indicates the density gradient. The results shown are representative of two independent experiments.