Using CRISPR/Cas9 to screen an Autographa californica multiple nucleopolyhedrovirus vector genome for gene essentiality in the production of a HIV-Gag virus like particle

Abstract

The baculovirus expression vector system (BEVS) has been widely used for producing recombinant proteins and virus-like particles (VLPs) due to its high protein expression, ability for some complex post translational modifications (PTMs), and high scalability. Of particular interest is the baculovirus Autographa californica multiple nucleopolyhedrovirus (AcMNPV), which is the most used virus for the BEVS. This baculovirus is amenable to genetic modifications, yet other than replacing or removing a few genes, this baculovirus has not seen major modifications. A hypothesis in the Aucoin lab centers around the idea that more product can be made if less resources are funnelled away from this goal [1]. Recently, a transfection-infection (TI) assay has been described [2], where the baculovirus genome can be scrutinized. Briefly, a stable Sf9 cell line expressing Cas9 (Sf9Cas9) [3], which has been shown to produce similar levels of green fluorescent protein and baculovirus as the native/parental Sf9 cells, is transfected with a plasmid hosting a single guide ribonucleic acid (sgRNA) targeting a specific baculovirus genomic sequence and subsequently infected with a baculovirus vector that is able to express a gene of interest/reporter product. However, given the large genome of the baculovirus, and the number of genes that can be scrutinized, higher throughput methodologies are required. This work addresses some of the engineering challenges related to the scale of the T-I assay and the associated analyses. These changes are then implemented and used to evaluate the essentiality of 21 late or very late genes in AcMNPV to produce BV and VLPs. More specifically, the effect of targeting these specific genes is being examined when expressing an enveloped virus-like particle under the control of the p6.9 promoter. Moving the T-I assay from a 6-well plate (2.5x106 cells/well in 2mL of media), to a 12-well plate required a change in cell seeding density (0.9x106cell/well in 0.8mL of media) to obtain similar infection and production dynamics. Further, using a human immunodeficiency virus (HIV) group specific antigen (Gag) fused with a green fluorescent protein (GFP) virus-like particle (VLP), or Gag-GFP for short, the effect of targeting genes could be assessed on the release of VLPs in the culture supernatant using flow cytometry. Dilutions of cell culture supernatant in a 96-well plate allowed dilutions to be done in parallel, which increased the throughput of VLP quantification using a 96-well plate compatible flow cytometer. Although VLP concentrations were consistently 10%-20% higher when diluted in the plate condition compared to microcentrifuge tubes, this amount was deemed to be acceptable due to the consistent effect. Furthermore, VLP stability over a 7-hour period was tested. It was found that VLPs started degrading 1.5 hours after initial dilution. This was found to be long enough for 15 samples to be run through the flow cytometer at a time and maintain VLP integrity. Controls were assessed for the T-I assay. One negative control involved only infecting cells with the baculovirus coding for Gag-GFP (infected control). A second negative control involved the use of transfection reagent without any plasmid followed by infection with the baculovirus coding for Gag-GFP (transfection reagent control). A third negative control involved a complete transfection step with a plasmid coding for a sgRNA that was not directed towards any baculovirus sequence. The latter is referred to a scrambled sequence control (or scrambled control for short). A fourth control involved a complete transfection step with a plasmid coding for mKate2, which is a red fluorescent protein (mKate2 control). It was observed that the transfection reagent control, the scrambled control, and the mKate2 control produced higher amounts of budded virus compared to the infected only control. Not only did we determine that that the transfection reagent increased infectious baculovirus in our assay, we determined that the scrambled control was the most appropriate control for the T-I assay. Of the 21 genes scrutinized as part of this work, deletion of lef-2, bion, lef-1, pkip and lef-11 all lowered budded virus production, which was consistent with previous studies. Deletion of pif2 lowered budded virus but has been previously reported to be non-essential. Deletion of ptp, Ac4, pk1, and fgf showed some indication that budded virus production was impaired, but failed to consistently meet the threshold for significance at a 95% confidence interval across all constructs tested. Deletion of v-ubi, Ac38, lef-12, and Ac51, which were previously reported to have lower levels of budded virus upon gene deletion, were found to have no effect in our study. Finally, Ac19 and env displayed no effects on budded virus and VLP production. These final two genes are possible candidates for genes that may be non-essential and should be screened further.