Characterization of virus-like particles from the BEVS platform with a special focus on gradient ultracentrifugation

Abstract

Coronaviruses continue to be an evolving threat, and Severe Acute Respiratory Syndrome (SARS) Virus-Like-Particles (VLPs) may be suitable as vaccine candidates. Early indications were that a SARS-VLP could be generated based on a universal betacoronavirus sequence using only Spike, Envelope and Membrane structural proteins. This process involved the generation of VLPs through the insect cell-baculovirus expression vector system (IC-BEVS), followed by semi-purification via gradient ultracentrifugation and subsequent fractionation. Collected fractions were analyzed by bicinchonic acid assay, flow virometry, immunoblotting, and multi-angle dynamic light scattering. Particle concentration, size, DNA content, and surface protein composition was determined for semi-purified samples.

This study provides a clearer understanding of biological dynamics in VLP generation workflows, by comparing HIV-VLPs, SARS-VLPs, baculovirus, and other co-produced particles. It was expected that significant concentrations of particles would not be observed in uninfected Sf9 cell culture supernatant samples, baculovirus alone would be observed in non-VLP producing samples, and VLPs would be observed in addition to baculovirus in VLP major structural protein producing samples. MADLS was used to assess the size and concentration of all particles produced and semi-purified, since all samples would be treated identically prior to measurement. It can be concluded that though both types of VLPs possess similarly-derived envelopes, their behavior is highly dependent on structural protein composition.

In terms of particle diameter, VLPs were expected to range between 50 nm and 250 nm, and baculovirus at 250 nm and greater. Post semi-purification, it was noted that some particles in the expected size range of baculovirus and VLP were detected in all samples, between fractions of densities 1.05 to 1.21 g/mL. Further analysis elucidated that uninfected samples had significantly lower measurement accuracy and a larger range of mean concentrations and sizes for those particles, which indicates a lack of reliability. Particles in the expected size range for VLPs found in non-VLP producing baculovirus samples were far more consistent, and determined to be real particles between densities 1.07 to 1.17 g/mL. These may be extracellular vesicles produced by insect cells under stress of infection, which have been documented in literature. Particles in the expected size range for baculovirus for this sample were also found in this density range. In VLP producing samples, particles were found that were consistent with expected size ranges for VLPs and baculovirus in the density ranges of 1.07 to 1.17 g/mL. Data from multiple orthogonal methods confirmed that the fraction containing the highest concentration of non-fluorescent HIV-VLPs, SARS-VLPs, and baculovirus was that with a density of 1.13 g/mL. Fluorescent HIV-VLPs by contrast, appeared less dense and had a maximum concentration at a density of 1.10 g/mL. In examining MADLS as a method of better characterizing the SARS-VLP, significant knowledge about the IC-BEVS was obtained which will assist in further developing this system and associated downstream purification processes.