Using the Sleeping Beauty (SB) Transposon to Generate Stable Cells Producing Enveloped Virus‐Like Particles (eVLPs) Pseudotyped with SARS‐CoV‐2 Proteins for Vaccination

Abstract

The Sleeping Beauty (SB) transposon system is an efficient non‐viral tool for gene transfer into a variety of cells, including human cells. Through a cut‐and‐paste mechanism, your favorite gene (YFG) is integrated into AT‐rich regions within the genome, providing stable long‐term expression of the transfected gene. The SB system is evolving and has become a powerful tool for gene therapy. There are no safety concerns using this system, the handling is easy, and the time required to obtain a stable cell line is significantly reduced compared to other systems currently available. Here, we present a novel application of this system to generate, within 8 days, a stable producer HEK293T cell line capable of constitutively delivering enveloped virus‐like particles (eVLPs) for vaccination. We provide step‐by‐step protocols for generation of the SB transposon constructs, transfection procedures, and validation of the produced eVLPs. We next describe a method to pseudotype the constitutively produced eVLPs using the Spike protein derived from the SARS‐CoV‐2 virus (by coating the eVLP capsid with the heterologous antigen). We also describe optimization methods to scale up the production of pseudotyped eVLPs in a laboratory setting (from 100 µg to 5 mg). © Published 2022. This article is a U.S. Government work and is in the public domain in the USA.

Basic Protocol 1: Generation of the SB plasmids

Basic Protocol 2: Generation of a stable HEK293T cell line constitutively secreting MLV‐based eVLPs

Basic Protocol 3: Evaluation of the SB constructs by immunofluorescence assay

Basic Protocol 4: Validation of eVLPs by denaturing PAGE and western blot

Alternate Protocol 1: Analysis of SARS‐CoV‐2 Spike protein oligomerization using blue native gel electrophoresis and western blot

Alternate Protocol 2: Evaluation of eVLP quality by electron microscopy (negative staining)

Basic Protocol 5: Small‐scale production of eVLPs

Alternate Protocol 3: Large‐scale production of eVLPs (up to about 1 to 3 mg VLPs)

Alternate Protocol 4: Large‐scale production of eVLPs (up to about 3 to 5 mg VLPs)

Support Protocol: Quantification of total protein concentration by Bradford assay

INTRODUCTION

Virus‐like particles (VLPs) are nanostructures that resemble actual viruses. Enveloped VLPs (eVLPs) expressing heterologous antigens are excellent platforms to produce protein‐subunit vaccines (Roldão, Mellado, Castilho, Carrondo, & Alves, 2010), as they allow different sets of immunogens to be incorporated on the surface of VLPs to generate effective vaccine formulations.

Several expression systems, including bacteria, yeast, insect cells/the baculovirus expression vector system (BEVS), plants, and mammalian cells, exist to produce VLPs (see the Background Information section for more extensive information on expression systems to produce VLPs). Two strategies are widely used to produce VLPs: transient and stable transfection. Each of these options has its own advantages and disadvantages (see the Background Information section for more details about transient versus stable transfections).

Here, we describe a highly efficient method for generation of stable mammalian cell lines constitutively expressing eVLPs. We use the Sleeping Beauty (SB) transposable system, which has applications in both functional genomics and gene therapy. This article describes the production of eVLPs for vaccination as a novel application of this system.

As a proof of concept, we describe generation of a stable HEK293T cell line that constitutively produces murine leukemia virus (MLV)‐based eVLPs that bud from the surface of host cells and are pseudotyped with the structural Spike protein (S) from SARS‐CoV‐2.

The speed, ease, safety, cost effectiveness, and high yield of this methodology make it an excellent option for the preclinical development of a VLP‐based vaccine and for production at commercial scale.

STRATEGIC PLANNING

An overview of the process to generate a stable producer cell line constitutively secreting eVLPs is provided in Figure 1. We describe construction of the SB transposon plasmids (Basic Protocol 1) and generation of stable cell lines producing “universal eVLPs,” which can then be pseudotyped with antigens belonging to a pathogen of interest (Basic Protocol 2). “Universal eVLPs” are empty capsids formed by self‐assembly of the MLV Gag‐polyprotein that are wrapped with a naked envelope. These VLPs can be used as a versatile platform for vaccination, as they are suitable for pseudotyping with different immunogens. The flow chart in Figure 1 also outlines the different approaches for validation of the plasmid transposons and the eVLPs (Basic Protocols 3 and 4 and Alternate Protocols 1 and 2). Lastly, Basic Protocol 5 outlines small‐scale production of high‐quality eVLPs, Alternate Protocols 3 and 4 detail two alternative methods for large‐scale production of eVLPs (on the order of milligrams; see Fig. 1), and the Support Protocol describes eVLP quantification.

Figure 1.

Flow chart outlining generation of the SB plasmid constructs encoding Gag polyprotein and YFGs, represented by the letters X and XX (Basic Protocol 1) and generation of stable HEK293T cell lines constitutively delivering eVLPs that are either non‐pseudotyped or pseudotyped with a model antigen (Basic Protocol 2). In this article, the SARS‐CoV‐2 Spike protein was chosen as the antigen of interest for pseudotyping. Downstream validation of the SB constructs and the eVLPs (Basic Protocols 3 and 4 and Alternate Protocols 1 and 2); production of eVLPs, at both small scale (Basic Protocol 5) and large scale (Alternate Protocols 3 and 4); and protein concentration measurement by Bradford assay (Support Protocol) are also included.

Basic Protocol 1. GENERATION OF THE SB PLASMIDS

SB is a synthetic transposon generated by reactivation of an inactive transposon belonging to the Tc1/mariner superfamily found in salmonid fish. Tc1 DNA transposons are mobile elements encoding a single transposase gene flanked by inverted terminal repeats (ITRs) (Ivics, Izsvak, Minter, & Hackett, 1996). The transposase gene of this fossil transposon was re‐engineered by site‐directed mutagenesis to yield an optimized hyperactive transposase called SB100X transposase (Ivics, Hackett, Plasterk, & Izsvák, 1997; Mátés et al., 2009). The SB system is composed of two vectors: a transposon donor vector containing your favorite gene (YFG) and the antibiotic resistance gene cassette flanked by ITRs and an SB transposase expression vector encoding the SB100X transposase (Fig. 2). When the two vectors are co‐transfected into the mammalian host cell, the transposable element is integrated into the host cell genome through a cut‐and‐paste mechanism. The transposase enzyme, expressed in trans, binds to the ITR of the transposon, excises the transposon from the donor vector, and inserts the transposon preferentially in dispersed TA dinucleotides. There are ∼30,000 TA target sites in the human genome, the majority of which are not found in promoters or first introns of actively transcribed genes (Kowarz, Löscher, & Marschalek, 2015), reducing the likelihood of off‐target effects.

Figure 2.

Schematic representation of the SB transposition mechanism. (1) Translation of the transposase‐coding sequence. (2) Binding of the transposase to the inverted terminal repeats (ITRs). (3) The transposase excises the transposon harboring the YFG/antibiotic resistance cassette from the SB donor construct by a double‐strand break. (4) The transposase creates two single‐strand cuts in the host DNA and inserts the cassette into AT‐rich regions of the host genome. The gaps are subsequently repaired by the host DNA repair machinery.

In this protocol, we describe generation of plasmids that were used to produce eVLPs pseudotyped with SARS‐CoV‐2 proteins for vaccination (Fig. 3). Please see the Critical Parameters section for critical points that should be considered before starting the protocol.

Figure 3.

SB transposon plasmids used for generation of the stable eVLP‐producer HEK293T cell lines. (A) Schematic representation of the cloning vector pSBbi‐Bla (left) obtained from the Addgene repository (#60526) and used for construction of the pSBbi‐Gag‐Bla (right). (B) Schematic representation of the cloning vector pSBbi‐Hyg (left) obtained from the Addgene repository (#60524) and used for construction of pSBbi‐Spike‐Hygro (right). (C) Sequences of the two SfiI restriction sites with different overhangs at the 5′ and 3′ ends of YFG enable directional cloning. Expression of YFG is driven by the strong EF‐1α promoter, whereas expression of the antibiotic resistance genes [blasticidin (Bla) or hygromycin (Hygro)] is driven by the novel synthetic RPBSA promoter.

The cloning vectors pSBbi‐Bla (Addgene, plasmid #60526) and pSBbi‐Hyg (Addgene, plasmid #60524) provide the backbones for generation of pSBbi‐Gag‐Bla (Fig. 3A) and pSBbi‐Spike‐Hygro (Fig. 3B), respectively. YFG is cloned into the SfiI restriction site. If more than one antigen is to be expressed in the envelope of the VLPs, there are two additional options in terms of selectable markers (puromycin and neomycin resistance genes) that can be found at https://www.addgene.org/ (plasmids #60495 to #60526). If HEK293T cells are the preferred mammalian cell for your experiments, we do not recommend use of neomycin because these cells are resistant to this antibiotic. The cloning vectors contain a bidirectional promoter: the eukaryotic translation elongation factor 1, driving the expression of YFG on the 5′ side of the transposon, and the RPBSA promoter, driving the expression of an antibiotic resistance gene on the 3′ side (EF‐1α/RPBSA). The RPBSA promoter is composed of a fragment of the RPL13a promoter fused to a region of the RPL41 gene (Kowarz et al., 2015). The YFG/antibiotic resistance cassette is flanked by two ITRs, at both the 5′ and 3′ ends, conferring the capability of transposition when a second plasmid encoding the transposase (see Basic Protocol 2) is co‐transfected and expressed in trans.

The plasmids can be generated using custom gene synthesis and cloning services that are commercially available. We used GENEWIZ synthesis services (https://www.genewiz.com/en/Public/Services/Gene‐Synthesis). Alternatively, YFG can be PCR amplified from a previous vector containing a codon‐optimized gene. In our example, we cloned the SARS‐CoV‐2 Spike sequence in silico into the pcDNA 3.1 vector. To do this, we used sequences from the codon‐optimized pCAGGS plasmid containing the SARS‐CoV‐2 full‐length Spike glycoprotein gene‐NR‐52310 (https://www.beiresources.org/Catalog/BEIPlasmidVectors/NR‐52310.aspx) and fused the ectodomain of the Spike protein to the transmembrane (TM) and cytoplasmic tail (CT) of the vesicular stomatitis virus glycoprotein [VSV‐G; UniProtKB ‐ P03522 (GLYCO_ VSIVA)] in silico for correct assembly of the VLPs (Fig. 4B; see Critical Parameters and Basic Protocol 1 for details). We next designed two mutations within the Spike gene: (1) abrogation of the furin cleavage site and (2) two proline substitutions for stabilization of the pre‐fusion conformation (Fig. 4A; see Background Information for details). The map using this in silico approach was uploaded to the custom cloning service at GENEWIZ to synthesize the gene. The synthesized gene obtained from GENEWIZ was used as a template to PCR amplify the chimeric Spike gene and was further subcloned into the pSBbi‐Hyg plasmid.

Figure 4.

Schematic representation of the Spike protein ectodomain. (A) Diagram showing the linear structure of the SARS‐CoV‐2 Spike protein ectodomain fused to the transmembrane domain (TM; dark green) and cytoplasmic tail (CT; light green) of VSV‐G. The signal peptide (SP) is indicated at the N‐terminus (N‐) in yellow. The S1 domain is represented in pink; the receptor‐binding domain (RBD), in red; and the S2 domain, in blue. The native furin cleavage site (S1/S2), indicated by the black arrow, was abrogated (RRAR to GSAS) to avoid proteolytic degradation by the host furin serine protease. A second cleavage site, the S’2, for the transmembrane protease serine protease 2 (TMPRSS‐2) is indicated by an orange arrow. The two proline (2P) substitutions made to stabilize the S protein in the pre‐fusion conformation are indicated by the red arrow. This chimeric fusion protein was cloned into the SfiI restriction sites of the pSBbi‐Hyg plasmid, which possess different overhangs that allow for correct orientation (see Fig. 3C and Basic Protocol 1). (B) Nucleotide sequences of the TM (dark green) and CT (light green). (C) Amino acid sequences of these domains.

Materials

• Template DNA:

• Codon‐optimized DNA template of MLV Gag coding sequence (pGag‐YFP kindly provided by Lucia Rupil, Centro de Investigación y Desarrollo en Inmunología y Enfermedades Infecciosas, Cordoba, Argentina)

• Codon‐optimized DNA template of SARS‐CoV‐2‐Spike coding sequence (pCAGGS plasmid containing SARS‐CoV‐2 full‐length Spike glycoprotein gene‐NR‐52310, BEI Resources)

• Forward and reverse primers containing SfiI restriction sites (IDT)

• Herculase II Fusion DNA Polymerase kit (Agilent, cat. no. 600677), containing Herculase II reaction buffer and Herculase II Fusion DNA Polymerase, and dNTPs

• Ultrapure water, molecular grade (K‐D Medical, cat. no. RGC‐3410, or equivalent)

• DNA ladder (GeneRuler 1 kb Plus DNA Ladder, Thermo Scientific, cat. no. SM1332)

• 1% (w/v) agarose gel (UltraPure™ Agarose, Invitrogen, cat. no. 16500500, or equivalent)

• 1× TAE buffer (from OmniPur® 10× TAE Buffer, Liquid Concentrate, Sigma, cat. no. 8720‐OP, or equivalent) containing 1× GelRed (from GelRed™ 10,000× in water, VWR, cat. no. 89139‐138)

• QIAquick Gel Extraction Kit (for 50 reactions; Qiagen, cat. no. 28704)

• SB vectors:

• pSBbi‐Bla (plasmid #60526; for generation of pSBbi‐Gag‐Bla)

• pSBbi‐Hyg (plasmid #60524; for generation of pSBbi‐Spike‐Hygro)

• FastDigest SfiI (Thermo Scientific, cat. no. FD1824), including FastDigest Green Buffer

• Quick CIP (New England Biolabs, cat. no. M0525), including CutSmart™ Buffer

• 3 M sodium acetate, pH 5.2 (Quality Biological, cat. no. 351‐035‐721, or equivalent)

• Absolute ethanol, 200 proof (Warner Graham, cat. no. 201096, or equivalent), 4°C

• 70% (v/v) ethanol (from absolute ethanol, Warner Graham, cat. no. 201096, or equivalent)

• TE Buffer, pH 8.0 (Quality Biological, cat. no. 351‐011‐721, or equivalent)

• T4 DNA Ligase (New England Biolabs, cat. no. M0202L), including T4 DNA Ligase Buffer

• One Shot™ TOP10 Chemically Competent Escherichia coli (Invitrogen, cat. no. C404006)

• S.O.C. medium (Invitrogen, part of cat. no. C404006)

• LB agar plate containing 100 µg/ml carbenicillin (IPM Scientific, cat. no. 11006‐443, or equivalent)

• LB broth (Luria Bertani Broth, Miller, Quality Biological, cat. no. 340‐004‐131, or equivalent) containing 100 µg/ml ampicillin (100 mg/ml, 0.2 μm filtered; Sigma, cat. no. A5354, or equivalent) or carbenicillin

• 25% (v/v) glycerol (Thermo Scientific, cat. no. 17904)

• QIAprep Spin Miniprep Kit (Qiagen, cat. no. 27106)

• Forward and reverse sequencing primers (for Sanger sequencing; IDT)

• PCR tubes (TempAssure PCR 8‐Tube Strips, Attached Optical Caps, 0.2 ml, USA Scientific, cat. no. 1402‐4700, or equivalent)

• Thermal cycler (Eppendorf or equivalent)

• Gel imaging system (Bio‐Rad or equivalent)

• 50°C, 65°C, and 80°C heating blocks

• 37°C and 42°C water baths

• Microcentrifuge, room temperature and 4°C

• 37°C incubator with and without shaking

• 14‐ml round‐bottom tubes (Falcon™ 14 mL Polystyrene Round‐Bottom Tubes, Fisher, cat. no. 14‐959‐1B)

• Cryovials (CryoPure tubes, 2 ml, QuickSeal screw cap, Sarstedt, cat. no. 72.380.002)

• 1.5‐ml microcentrifuge tubes

• Additional reagents and equipment for agarose gel electrophoresis (see Current Protocols article: Voytas, 2000) and Sanger sequencing

NOTE: Experiments involving PCR require extremely careful technique to prevent contamination.

Generating the Sleeping Beauty plasmids

• 1Prepare PCR reactions for MLV Gag–encoding and SARS‐CoV‐2 Spike–encoding sequences by mixing ∼30 ng template DNA, 0.25 µM of each forward and reverse primer containing SfiI restriction sites, 1× Herculase II reaction buffer, 1 µl Herculase II Fusion DNA Polymerase, and 250 µM of each dNTP (the last three items all from the Herculase II Fusion DNA Polymerase kit) and topping up with ultrapure water to 50 µl in a PCR tube.

  The primers confer unique SfiI restriction sites on either end of YFG.

• 2
  Perform PCR using a thermal cycler with the following running conditions:

  Initial step:	2 min	95°C	(denaturation)
  30 cycles:	20 sec	95°C	(denaturation)
  	20 sec	65°C	(annealing)
  	30 s/kb	72°C	(extension)
  Final step:	3 min	72°C	(extension).

• 3Confirm PCR results by separating the products and a DNA ladder on a 1% agarose gel in 1× TAE buffer containing 1× GelRed. Record image with a gel imaging system.
• 4Purify PCR products from the gel using the QIAquick Gel Extraction Kit according to the manufacturer's instructions.
• 5
  Set up a 40‐µl restriction enzyme digest of each purified PCR product and each SB vector (pSBbi‐Bla and pSBbi‐Hyg):

  1. Mix 1 µg pure PCR product, 1× FastDigest Green Buffer, 1 µl FastDigest SfiI, and ultrapure water to 40 µl.
  2. Mix 4 µg SB vector, 1× FastDigest Green Buffer, 4 µl FastDigest SfiI, and ultrapure water to 40 µl.

  To facilitate the protocol, use a spectrophotometer (NanoDrop™ 2000 Spectrophotometer, Thermo Scientific, cat. no. ND‐2000, or equivalent) to measure the DNA concentration to ensure addition of sufficient DNA at the correct ratio.

• 6Incubate reactions at 50°C in a heating block for 2 hr.
• 7Perform electrophoresis of digested products and a DNA ladder on a 1% agarose gel in 1× TAE buffer containing 1× GelRed. Record image with a gel imaging system.
• 8Purify digested products from the gel using the QIAquick Gel Extraction Kit according to the manufacturer's instructions.
• 9Dephosphorylate each linearized SB vector by mixing the equivalent of 1 pmol DNA with 1× CutSmart™ Buffer, 5 U Quick CIP, and ultrapure water to 30 µl.
• 10Incubate reactions at 37°C in a water bath for 45 min. Then, inactivate at 80°C in a heating block for 2 min.
• 11Precipitate each linearized and dephosphorylated SB vector by mixing 1/10 volume of 3 M sodium acetate (pH 5.2) and 2 volumes of ice‐cold absolute ethanol into dephosphorylation reaction.
• 12Incubate at −80°C for 1 hr.
• 13Centrifuge 30 min at 20,000 × g, 4°C, in a microcentrifuge. Discard supernatant.
• 14Wash DNA pellet with 500 μl of 70% ethanol.
• 15Centrifuge pellet again for 5 min at 20,000 × g, 4°C. Discard supernatant.
• 16Air‐dry DNA pellet for 2 min and then resuspend pellet in 20 µl TE buffer (pH 8.0).
• 17
  Set up two ligation mixes to generate SB plasmids required for VLP expression:

  1. Mix SARS‐CoV‐2‐Spike insert and 50 ng pSBbi‐Hyg vector at a 3:1 ratio, add 1× T4 DNA Ligase Buffer, and top up with ultrapure water to 20 µl. Add 1 µl T4 DNA Ligase (400 U).
  2. Mix MLV Gag insert and 50 ng pSBbi‐Bla vector at a 3:1 ratio, add 1× T4 DNA Ligase Buffer, and top up with ultrapure water to 20 µl. Add 1 µl T4 DNA Ligase (400 U).
• 18Incubate ligation mixes at room temperature for 10 min. Deactivate at 65°C in a heating block for 10 min. Place on ice.

Transforming Escherichia coli

• 19Thaw one vial of One Shot™ TOP10 Chemically Competent E. coli on ice for each ligation mix.
• 20Mix 3 µl of one ligation mix from step 18 into a vial of chemically competent cells and mix gently by flicking bottom of the tube a few times.
• 21Incubate on ice for 30 min.
• 22Heat‐shock chemically competent cells in a water bath set to 42°C for 30 s.
• 23Recover cells on ice for 5 min.
• 24Add 600 µl S.O.C. medium (room temperature and without antibiotics) to the E. coli cells.
• 25Allow cells to recover for 1 hr at 37°C in an incubator with shaking at 225 rpm.
• 26Spread 50 to 100 µl of each transformation on an LB agar plate containing 100 µg/ml carbenicillin.

  It is often useful to plate several dilutions (e.g., 1/10 and 1/50 dilutions in S.O.C. medium) of each transformation mix.

• 27Incubate plates overnight at 37°C in an incubator.

Confirming the transformed Sleeping Beauty constructs

• 28Pick 3 to 5 well‐isolated colonies per transformation and inoculate each into 5 ml LB broth containing 100 µg/ml ampicillin or carbenicillin in 14‐ml round‐bottom tubes.
• 29Shake at 225 rpm overnight at 37°C in an incubator.
• 30Preserve 500 µl of each E. coli culture in 25% glycerol in cryovials and store at −80°C.
• 31Pellet each remaining E. coli culture in a 1.5‐ml microcentrifuge tube by centrifugation for 3 min at 6800 × g at room temperature in a microcentrifuge and purify plasmid DNA using the QIAprep Spin Miniprep Kit according to the manufacturer's instructions.
• 32Perform electrophoresis of purified plasmids and a DNA ladder on a 1% agarose gel in 1× TAE buffer containing 1× GelRed. Record image with a gel imaging system.

  The size of the plasmid should increase if the transformation was successful. If none of the plasmids contains the insert, repeat steps 28 to 32.

• 33Confirm quality of the transformation by Sanger sequencing with appropriate forward and reverse sequencing primers.

Basic Protocol 2. GENERATION OF A STABLE HEK293T CELL LINE CONSTITUTIVELY SECRETING MLV‐BASED eVLPs

Generating clonal stable cell lines can be a time‐consuming process that can take a few weeks’ time. Using the SB system, the time required to generate a stable eVLP‐producer cell line is dramatically reduced to ∼1 week. A HEK293T cell line co‐transfected with the plasmids of the SB transposon system [the donor plasmid encoding MLV Gag (Fig. 5A) and the SB100X transposase expression vector; see Basic Protocol 1] will constitutively express non‐pseudotyped eVLPs (Fig. 5B). These eVLPs can be used as a base platform for generation of eVLP‐based vaccines or for cancer treatment by incorporating a variety of pathogens or tumor antigens into their envelope. In this regard, they can be considered a “universal platform” that can be pseudotyped with YFG. Figure 5C shows an example of an eVLP pseudotyped with the SARS‐CoV‐2 Spike protein, and Figure 6 summarizes the steps to generate non‐pseudotyped and pseudotyped eVLPs.

Figure 5.

Cartoon showing the structure of the MLV Gag polyprotein, universal eVLPs, and pseudotyped eVLPs. (A) Cartoon representation of the MLV Gag polyprotein, made up of three different domains: Matrix (red), Capsid (green), and Nucleocapsid (yellow). (B) Schematic representation of a “universal eVLP” formed by an uncleaved Gag capsid surrounded by a host‐cell plasma membrane. These eVLPs can be used as a base platform for pseudotyping with envelope antigens of interest. (C) Schematic representation of an eVLP pseudotyped with the chimeric SARS‐CoV‐2 Spike protein (see Fig. 4), used here as a model antigen.

Figure 6.

Schematic overview of the transfection protocol. (A) Production of “universal” eVLPs (non‐pseudotyped). These particles express only an immature capsid formed by self‐assembly of the unprocessed MLV Gag polyprotein, surrounded by the envelope that is not decorated with antigens. These HEK293T cells stably produce eVLPs that can be further pseudotyped with one or more antigens. (B) Production of eVLPs pseudotyped with the SARS‐CoV‐2 Spike protein, a model antigen.

Materials

• 0.1% (w/v) poly‐L‐lysine stock solution (Sigma‐Aldrich, cat. no. P8920)

• Sterile distilled water

• HEK293T cells (ATCC® CRL­3216™)

• PEI MAX [Polyethylenimine (PEI) HCl MAX, MW 40000, Transfection Grade, Polysciences, cat. no. 24765)

• Opti‐MEM (Opti‐MEM™ I Reduced Serum Medium, Gibco, cat. no. 31985070)

• pSBbi‐Gag‐Bla transposon construct (SB plasmid construct encoding MLV Gag driven by EF‐1α promoter and blasticidin resistance gene driven by synthetic RPBSA promoter; He, Rad, Poudel, & McLellan, 2020; Fig. 3A; see Basic Protocol 1)

• pSB100X transposase vector (pCMV(CAT)T7‐SB100, Addgene, plasmid #34879)

• 10 mg/ml blasticidin solution (InvivoGen, cat. no. ant‐bl‐1)

• pSBbi‐Spike‐Hygro transposon construct (Fig. 3B; see Basic Protocol 1)

• 100 mg/ml hygromycin solution (Hygromycin B Gold, InvivoGen, cat. no. ant‐hg‐1)

• 6‐well tissue culture plates (CytoOne 6‐well tissue culture plates, clear, individually wrapped with lids, USA Scientific, cat. no. CC7682‐7506, or equivalent; or 35‐mm tissue culture–treated Petri dishes)

• 1.5‐ml microcentrifuge tubes

• Vortex

• T75 tissue culture flasks (CytoOne T75 tissue culture flasks, 5/sleeve, USA Scientific, cat. no. CC7682‐4175, or equivalent)

• Additional reagents and equipment for preparing complete DMEM (see recipe)

NOTE: All steps described here should be performed in a BSL‐2 biosafety cabinet.

NOTE: All solutions and equipment coming into contact with cells must be sterile, and proper sterile technique should be used accordingly.

NOTE: All culture incubations are performed in a 37°C, 5% CO₂ incubator unless otherwise specified.

Generation of non‐pseudotyped “universal eVLPs”

Pretreatment of the tissue culture surface

• 1Prior to cell seeding, to treat the tissue culture surface with poly‐L‐lysine, dilute 0.1% poly‐L‐lysine stock solution 10‐fold with sterile distilled water to have a 0.01% working solution.
• 2Coat wells of 6‐well tissue culture plates with sufficient 0.01% poly‐L‐lysine to cover the cell culture surface.
• 3Incubate 5 min at room temperature.
• 4Remove poly‐L‐lysine solution, rinse surface thoroughly with sterile distilled water, and let plates dry for ≥2 hr. Use plates on the same day or store at 20°C for later use.

Cell seeding (Day 1)

• 5Prepare complete DMEM without antibiotics containing 10% fetal bovine serum (FBS; see recipe).
• 6Plate 5–8 × 10⁵ HEK293T cells/well in 2 ml complete DMEM without antibiotics.
• 7Incubate cells at 37°C, 5% CO₂ during 24 hr or until ready for transfection.

  The HEK293T cells should be 80% to 90% confluent before use. Transfect cells at a high density for high efficiency, to obtain high expression levels, and to minimize cytotoxicity.

Transfection (Day 2)

• 8One to two hours before transfection, prepare warm complete DMEM without antibiotics containing 5% FBS (see recipe) and replace growth medium with this fresh medium.

  High serum levels inhibit the efficacy of PEI Max. In most cases, low serum levels (∼5%) will produce the highest transfection efficiency.

• 9Prepare PEI MAX for PEI MAX/DNA (4:1) transfection mixture by adding 8 μg (8 µl of 1 µg/µl) PEI MAX to 250 μl Opti‐MEM in a 1.5‐ml microcentrifuge tube. Vortex solution for 5 s.

  IMPORTANT NOTE: The order of mixing in steps 9 to 11 is critical.

• 10In a separate 1.5‐ml microcentrifuge tube, prepare 2 µg of the DNA required for transfection by mixing 1.9 μg pSBbi‐Gag‐Bla transposon construct with 100 ng pSB100X transposase vector and top up to a total volume of 250 µl with Opti‐MEM.

  All amounts and volumes are given on a per‐well basis.

• 11Add diluted PEI MAX (250 µl) from step 9 to diluted DNA mix (250 µl) from step 10. Mix by pipetting up and down two or three times.
• 12Let solutions sit for 30 min in a hooded environment to allow PEI MAX/DNA complexes to form.
• 13Carefully add the 500 µl of PEI MAX/DNA mixture to a well of adherent HEK293T cells from step 7.

  Take care to gently pipet the solution down along the side of the well and not directly on the cells to avoid disrupting the monolayer of adherent cells.

• 14Return cells to 37°C, 5% CO₂ and incubate for 24 hr.

Drug selection (Day 3)

• 15Twenty‐four hours post‐transfection, prepare complete DMEM containing 5 to 15 µg/ml blasticidin (from 10 mg/ml blasticidin solution). Transfer cells to a T75 tissue culture flask and grow in complete DMEM with antibiotics for ∼5 days or until culture reaches 80% to 90% confluence.

  Test the sensitivity of HEK293T cells to blasticidin prior to transfection to determine a suitable concentration. The recommended working concentration range is 5 to 15 µg/ml.

  At this time, the surviving cell population should be constitutively expressing Gag polyprotein, so a stable high‐producer cell line has been established. The expression of the Gag polyprotein should be confirmed by immunofluorescence (see Basic Protocol 3). This assay will determine the efficiency of the transfection. Blasticidin can be removed from the medium for cell lines confirmed to stably express the Gag polyprotein.

  We recommend cryopreserving a stock of “universal eVLPs” in liquid nitrogen for use as needed for further applications.

Pseudotyping “universal eVLPs” with one or more antigens

Pretreatment of the tissue culture surface and cell seeding (Day 1)

• 16Follow steps 1 to 7 using MLV Gag non‐pseudotyped stable eVLP‐producer HEK293T cell line generated in steps 1 to 15.

Transfection (Day 2)

• 17Follow steps 8 to 14. In step 10, use pSBbi‐Spike‐Hygro transposon construct.

Drug selection (Day 3)

• 18Twenty‐four hours post‐transfection, prepare complete DMEM containing 100 to 500 µg/ml hygromycin (from 100 mg/ml hygromycin solution). Transfer cells to a T75 tissue culture flask and grow in complete DMEM with antibiotics for ∼5 days or until culture reaches approximately 80% to 90% confluence.

  Test the sensitivity of HEK293T cells to hygromycin prior to transfection to determine a suitable concentration. The recommended working concentration range is 100 to 500 µg/ml.

  Check the transfection efficiency and the co‐expression of the antigens by immunofluorescence (see Basic Protocol 3).

  Freeze several cryovials at −80°C for downstream applications and several in liquid nitrogen for long‐term storage.

Basic Protocol 3. EVALUATION OF THE SB CONSTRUCTS BY IMMUNOFLUORESCENCE ASSAY

Both error‐free expression of the coding sequences cloned into the SB transposon construct and correct assembly of the VLPs are essential for the success of eVLP production (Table 1).

Table 1.

Assays to be Performed for Validation

Type of validation	Validation output	Assay
Validation of the SB constructs	Protein expression of DNA coding sequences cloned in the different constructs	Immunofluorescence assay (IFA; Fig. 9)
	Evaluation of the stable long‐term expression of the transgenes	Monitoring expression of the transfected SB transposon construct over the course of successive passages without drug selection (Fig. 8)
Validation of the eVLPs	Presence of all proteins in the eVLPs	Western blot (Fig. 10)
	eVLP quality	Electron microscopy (negative staining; Fig. 11)

When using the SB system to produce eVLPs, there is no need to perform repeated transient transfections prior to each application. It is therefore sufficient to confirm the expression of the antigens in the HEK293T cells by immunofluorescence assay (IFA) once the stable HEK293T cell line is obtained (Basic Protocol 2). This eliminates inter‐experimental variations derived from technical factors involved in each transient transfection.

Long‐term stability of the expression of YFG or Gag polyprotein in the transfected cells (Basic Protocol 2) can be tested by using SB expression vectors containing fluorescent protein reporters (GFP, BFP, RFP) (Fig. 7). Several variants of these vectors can be found in the Addgene repository (addgene.org; Table 2). Given that YFGs are cloned in the same cassette as the genes expressing the reporter fluorescent proteins, monitoring the level of fluorescence through successive passages gives an indirect estimation of the stability of the integrated transposon and the expression of YFG. For example, we used pSBbi‐BH to express our model antigen, the Spike protein of SARS‐CoV‐2, and pSBbi‐RP for the expression of an irrelevant gene (X). Figure 8 shows that 100% of the transfected cells expressed both fluorescent proteins (RFP and BFP) and that this expression persisted for ≥5 months after transfection.

Figure 7.

Schematic representation of the optimized SB expression vectors containing fluorescent protein reporters for the SB transposase system available at addgene.org (Kowarz et al., 2015).

Table 2.

List of SB Expression Vectors Available in Addgene Repository

Name of vector	Fluorescent protein reporter	Selection marker	Addgene ID
pSBbi‐GP	Green Fluorescent Protein (GFP)	Puromycin	60511
pSBbi‐BP	Blue Fluorescent Protein (BFP)	Puromycin	60512
pSBbi‐RP	dTomato (RFP)	Puromycin	60513
pSBbi‐GH	Green Fluorescent Protein (GFP)	Hygromycin	60514
pSBbi‐BH	Blue Fluorescent Protein (BFP)	Hygromycin	60515
pSBbi‐GB	Green Fluorescent Protein (GFP)	Blasticidin	60520
pSBbi‐RB	dTomato (RFP)	Blasticidin	60522

Figure 8.

Long‐term stability of fluorescence proteins expressed in a stable HEK293T cell line. HEK293T cells, previously transfected with the expression vector pSBbi‐Gag‐Bla, were co‐transfected with the expression vectors pSBbi‐X‐RP and pSBbi‐S‐BH. The panels on the left show cells maintained in culture, without drug selection, for 15 days post‐transfection. The panels on the right show cells maintained without drug selection for 5 months post‐transfection. (A and A') Cells expressing dTomato (RFP), (B and B’) cells expressing BFP, (C and C’) merge, and (D and D’) brightfield. (A‐D) show tiling, and (A’‐D’) show a single field. Confocal images were collected using a Leica DMI8000 confocal microscope (Leica Microsystems) enabled with a 63× oil‐ immersion objective, NA 1.4. Images were acquired using constant laser intensity with an argon laser and 488‐nm wavelength for Alexa 488 excitation, and photons were collected using constant photomultiplier electronic gain between the samples to quantify the differences in absolute intensity levels from different time courses of samples. Images were collected using an automated tiling method to observe an unbiased data pool from multiple tiles (in this case, 25 tiles). Acquired images were further analyzed using Imaris image‐processing software (Bitplane USA). Scale bars: tile, 20 μm (A‐D); single field, 100 μm (A’‐D’), zoomed‐in snapshot.

Materials

• 0.1% (w/v) poly‐L‐lysine stock solution (Sigma‐Aldrich, cat. no. P8920)

• Sterile distilled water

• Stable HEK293T cell line (see Basic Protocol 2)

• 1× phosphate‐buffered saline (PBS), pH 7.4 (Gibco, Fisher Scientific, cat. no. 10‐010‐031)

• Fixative solution (see recipe)

• 0.1% (v/v) Triton X‐100 (from 10%; see recipe)

• 10% (v/v) FBS (Biowest, cat. no. S1620, or equivalent) in PBS (Gibco, Fisher Scientific, cat. no. 10‐010‐031)

• Antibody dilution buffer (see recipe)

• Primary antibody:

• Rabbit Anti‐Moloney MLV Gag Polyprotein Polyclonal Antibody, Unconjugated (MyBioSource, cat. no. MBS1490637)

• Anti‐SARS‐CoV‐2 (2019‐nCoV) Spike RBD Antibody, Rabbit PAb, Antigen Affinity Purified (Sino Biological, cat. no. 40592‐T62)

• Secondary antibody: Alexa Fluor 488 goat anti‐rabbit IgG (Invitrogen, cat. no. A11034)

• 4′,6‐diamidino‐2‐phenylindole, dihydrochloride (DAPI; Thermo Fisher Scientific cat. no. D1306)

• Antifade mounting medium (ProLong™ Diamond Antifade Mountant, Thermo Fisher, cat. no. P36961)

• Sterile 12‐ or 14‐mm round coverslips (VWR, cat. no. 89015‐725, or equivalent)

• 12‐well tissue culture plates (CytoOne 12 well tissue culture plates, USA Scientist, cat. no. CC7682‐7512)

• Microscope slides

• Wide‐field epifluorescence microscope (or confocal microscope, such as Leica DMI8000 confocal microscope, Leica Microsystems)

NOTE: Perform steps 1 to 4 in a BSL‐2 biosafety cabinet.

NOTE: All solutions and equipment coming into contact with cells must be sterile, and proper sterile technique should be used accordingly.

NOTE: All culture incubations are performed in a 37°C, 5% CO₂ incubator unless otherwise specified.

• 1Place sterile 12‐ or 14‐mm round coverslips in the wells of a 12‐well tissue culture plate. Dilute 0.1% poly‐L‐lysine stock solution with sterile distilled water 10‐fold and completely cover coverslips with the 0.01% poly‐L‐lysine.

  HEK293T cells adhere weakly to glass and plastic. Poly‐L‐lysine treatment of the coverslips will enhance cell attachment and adhesion.

• 2Incubate at room temperature for 5 min. Aspirate solution and wash thoroughly twice with sterile distilled water.
• 3Let plates air‐dry for 2 hr under a hood.
• 4Seed each well with a suspension of ∼2 × 10⁵ cells from the stable HEK293T cell line from Basic Protocol 2 in 2 ml complete DMEM.

  Make sure to distribute the cells evenly.

• 5Incubate plate at 37°C, 5% CO₂ for 18 to 24 hr or until the monolayer reaches 60% to 70% confluence.

  Higher cell density will negatively affect the quality of the staining.

• 6The next day, wash cells very gently three times with 2 ml of 1× PBS (pH 7.4).
• 7Fix cells with 2 ml fixative solution for 15 min at room temperature.
• 8Wash cells once with 2 ml of 1× PBS.
• 9Permeabilize cells by incubating in 2 ml of 0.1% Triton X‐100 for 15 min at room temperature.
• 10Wash cells twice with 2 ml of 1× PBS.
• 11Incubate cells for 30 min in 2 ml of 10% FBS in PBS at room temperature to block nonspecific binding of antibodies.
• 12Wash cells once with 2 ml antibody dilution buffer.
• 13Add primary antibody diluted 1:400 in 2 ml antibody dilution buffer for 1 hr at room temperature.
• 14Wash cells three times with 2 ml antibody dilution buffer, with 5 min per wash.
• 15Add secondary antibody diluted 1:2000 (manufacturer's recommendation: 1 to 10 μg/ml) in 2 ml antibody dilution buffer for 1 hr at room temperature.
• 16Wash with 2 ml antibody dilution buffer containing 1 μg/ml DAPI and incubate for 5 min. Wash two more times with 2 ml of 1× PBS for 5 min each.
• 17Remove excess liquid from sample by gently tapping the edge of the coverslip. Then, apply 1 to 2 drops (or 20 to 60 µl) of room‐temperature antifade mounting medium directly onto a clean microscope slide and carefully lower coverslip onto the mountant to avoid trapping any air bubbles. Store in the dark and let slides air‐dry.

  Allow the mountant to warm to room temperature for 1 hr before mounting specimens.

• 18Analyze subcellular localization of the labeled antigens using a conventional wide‐field epifluorescence microscope.

  For increased optical resolution, confocal microscopy can be used.

  An illustration of immunofluorescence labeling of a HEK293T cell line expressing Gag and SARS‐CoV‐2 Spike protein is shown in Figure 9.

  A FACS sorter can be used to select for a higher‐VLP‐producer stable HEK293T cell line. Copy‐number estimation of the integrated transposon can be assessed by qPCR (Kowarz et al., 2015), and depending of the amount of the SB transposon used, it can be tuned to have between 1 and 10 transgenes copies per cell (Wächter, Kowarz, & Marschalek, 2014).

Figure 9.

Indirect immunofluorescence assay. (A) HEK293T cells expressing

MLV Gag polyprotein stained with the primary antibody Rabbit anti‐Moloney MLV Gag polyprotein polyclonal antibody diluted 1:400 and the secondary antibody Alexa Fluor 488 goat anti‐rabbit diluted 1:2000. (B) Bright‐field image. (C) HEK293T cells expressing SARS‐CoV‐2 Spike protein stained with the primary antibody rabbit anti‐SARS‐CoV‐2 Spike polyclonal antibody diluted 1:400 and the secondary antibody Alexa Fluor 488 goat anti‐rabbit diluted 1:2000. (D) Bright‐field image. The figure shows that both proteins are expressed in 100% of the cells transfected with the SB construct. The variation in expression intensity can be attributed to different copy‐number insertions. Scale bars: 100 μm.

Basic Protocol 4. VALIDATION OF eVLPs BY DENATURING PAGE AND WESTERN BLOT

This protocol is performed to confirm the identities of the proteins expressed in the eVLPs. The total protein content of the VLPs is subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS‐PAGE). First, VLPs are denatured with SDS and placed under reducing conditions. The proteins are then separated by SDS‐PAGE within a polyacrylamide gel according to their molecular weight, along with protein markers of known molecular weights, and transferred onto a nitrocellulose membrane.

Materials

• eVLP samples (see Basic Protocol 5)

• 4× LDS sample buffer (NuPAGE™ LDS Sample Buffer, 4×, Invitrogen, cat. no. NP0007)

• 10× sample‐reducing agent (NuPAGE™ Sample Reducing Agent, 10×, Invitrogen, cat. no. NP0004)

• Deionized water

• 20× MOPS SDS running buffer (NuPAGE™ MOPS SDS Running Buffer, 20×, Invitrogen, cat. no. NP0001)

• 10× NC membrane equilibration buffer (eBlot™ L1 NC Membrane Equilibration Buffer, 10×, GenScript, cat. no. L00731)

• 5× NC membrane transfer buffer (eBlot™ L1 NC Membrane Transfer Buffer, 5×, GenScript, cat. no. L00730)

• Isopropanol (Sigma, cat. no. 190764, or equivalent)

• Pre‐cast mini protein gel (NuPAGE™ 4%‐12%, Bis‐Tris, 1.0 mm, Mini Protein Gel, 10‐well, Invitrogen, cat. no. NP0321)

• Antioxidant (NuPAGE™ Antioxidant, Invitrogen, cat. no. NP0005)

• Pre‐stained protein standards (Precision Plus Protein™ Kaleidoscope™ Prestained Protein Standards, Bio‐Rad, cat. no. 1610375)

• Ponceau S solution, 0.1% (w/v) in 5% (v/v) acetic acid (Millipore Sigma, cat. no. P7170, or equivalent)

• Western blot blocking solution (see recipe)

• 1.5‐ml microcentrifuge tubes

• 95°C heating block

• Microcentrifuge

• XCell SureLock™ Mini‐Cell (Invitrogen, cat. no. EI0001)

• Electrophoresis power supply

• Blotting gel knife (Invitrogen or equivalent)

• Shallow plastic containers/trays

• Nitrocellulose membrane (Amersham™ Protran® nitrocellulose membrane, pore size 0.2 μm, Sigma, cat. no. GE10600001, or equivalent)

• Transfer sponges (eBlot™ L1 Transfer Sponge, GenScript, cat. no. L00736)

• Blotting tweezers (Invitrogen or equivalent)

• Transfer system (eBlot™ L1, GenScript, cat. no. L00686, or equivalent)

• Blotting roller (Invitrogen or equivalent)

• 1Mix each eVLP sample, 4× LDS sample buffer, and 10× sample‐reducing agent with deionized water to required working concentration (e.g., 1×) in a 1.5‐ml microcentrifuge tube.

  A typical starting amount to test is 1 µg VLPs per well. Typically, 10 µl loading sample is added to a single well of the gel.

  Example preparation for a VLP sample with an initial concentration of 1 µg/µl is as follows:

  1 µg/µl VLP sample	1.5 µl
  4× LDS sample buffer	3.75 µl
  10× sample‐reducing agent	1.5 µl
  Deionized water	8.25 µl
  Total	15 µl

• 2Boil loading samples at 95°C in a heating block for ∼5 min.
• 3Centrifuge all loading samples for 2 min at 20,000 × g in a microcentrifuge.
• 4Dilute 20× MOPS SDS running buffer in distilled water and 10× NC membrane equilibration buffer and 5× NC membrane transfer buffer in isopropanol to their working concentration of 1×.
• 5Set up a pre‐cast mini protein gel in 1× MOPS SDS running buffer within the XCell SureLock™ Mini‐Cell for SDS‐PAGE according to the manufacturer's instructions. Mix 500 µl antioxidant into upper buffer chamber.
• 6Load samples (see step 3) and 5 µl pre‐stained protein standards into their respective wells of the gel.

  For best results, load all wells with the same volume of sample buffer, even where there is no sample.

• 7Run gel at 100 V for ∼5 min using an electrophoresis power supply.

  This ensures that the entire sample enters the gel.

• 8Continue to run gel at 200 V for 45 to 50 min.
• 9Crack open gel case. Remove well separators and protruding foot of the gel with a blotting gel knife.
• 10Release gel in a shallow plastic container/tray of deionized water and allow it to equilibrate for 5 min.
• 11Meanwhile, cut out a piece of nitrocellulose membrane using a transfer sponge as a template, handling membrane with clean blotting tweezers. Equilibrate nitrocellulose membrane piece in 10 ml of 1× NC membrane equilibration buffer for 1 to 2 min.

  Always handle the nitrocellulose membrane with clean blotting tweezers to prevent contamination.

• 12Assemble transfer stack, including transfer sponges, on the transfer cassette of the transfer system (eBlot™ L1 or equivalent) according to the manufacturer's instructions. Remove any visible bubbles between nitrocellulose membrane and gel using a blotting roller.

  Handle the gel via its thicker bottom side to avoid breakage or tearing.

• 13Close transfer cassette and slot it into either channel of transfer system. Select pre‐programmed “standard” 17‐min transfer and start process.

  Use of the eBlot™ L1 Fast Wet Transfer System allows efficient blot transfer of small‐, medium‐, and large‐molecular‐weight proteins within 9 to 17 min. If this device is not available, any other conventional equipment for western blot can be used. The regular time required for the transfer process can range between 1 hr at room temperature and overnight in a cold room, depending on the particular device.

• 14When the transfer is complete, remove transfer cassette from the transfer system. Open cassette and discard transfer sponges and gel.
• 15Optional: Stain nitrocellulose membrane with Ponceau S solution for 1 to 2 min to check the quality of the transfer. Rinse membrane with deionized water to remove the Ponceau stain before moving to the next step.

  Check that no part of the membrane was obscured by an air bubble during the protein transfer.

• 16Block nitrocellulose membrane in western blot blocking solution overnight at 4°C and then proceed to Alternate Protocol 1, step 21.

  The membrane can also be blocked by rocking at room temperature for ≥1 hr.

  See Figure 10 for example results.

Figure 10.

Denaturing and native western blot analyses of the SARS‐CoV‐2 Spike and MLV Gag proteins expressed in the eVLPs. SARS‐CoV‐2 Spike protein was detected with anti‐SARS‐CoV‐2‐Spike, and MLV Gag was detected using anti‐Moloney‐MLV‐Gag polyprotein. Anti‐rabbit goat IgG H+L‐peroxidase was used as a secondary antibody. The expected sizes of the proteins are ∼180 kDa for SARS‐CoV‐2 Spike, a main band of ∼480 kDa and a weak band of ∼720 kDa for SARS‐CoV‐2 Spike trimer (similar molecular weights for SARS‐CoV‐2 trimer have been reported previously (Petruk et al., 2020), and ∼66 kDa for MLV Gag.

Alternate Protocol 1. ANALYSIS OF SARS‐CoV‐2 SPIKE PROTEIN OLIGOMERIZATION USING BLUE NATIVE GEL ELECTROPHORESIS AND WESTERN BLOT

Viral envelope glycoproteins are typically oligomers, and their quaternary structure is essential for virus entry. They bind to host cell receptors and mediate attachment to and fusion with the host membrane, leading to delivery of the viral genome into the cytoplasm of the host cell. These proteins are the main target of neutralizing antibodies, and in many cases, correct oligomerization on the surface of the VLPs is critical for antibody recognition. The SARS‐CoV‐2 Spike protein used here as a model antigen is a trimer. Native gel electrophoresis can be performed to confirm that the Spike protein's trimeric state is maintained on the surface of the eVLPs. First, the eVLP sample is mixed with Triton X‐100, a non‐ionic detergent that solubilizes the proteins in the sample but maintains them in their native conformation. Addition of Coomassie G‐250 to the solubilized proteins displaces the detergent and confers a net negative charge on the proteins in the sample. The proteins are then separated within a polyacrylamide gel according to their molecular masses, along with protein markers of known masses, and transferred onto a polyvinylidene fluoride (PVDF) membrane. The band corresponding to the Spike protein on the membrane is finally visualized by western blotting.

Materials

• eVLP samples (see Basic Protocol 5)

• 4× sample buffer (NativePAGE™ Sample Buffer, 4×, Invitrogen, cat. no. BN2003)

• 10% (v/v) Triton X‐100 (see recipe)

• Deionized water

• 5% G‐250 sample additive (NativePAGE™ 5% G‐250 Sample Additive, Invitrogen, cat. no. BN2004)

• 20× running buffer (NativePAGE™ Running Buffer, 20×, Invitrogen, cat. no. BN2001)

• 10× membrane equilibration buffer (eBlot™ L1 PVDF Membrane Equilibration Buffer, 10×, GenScript, cat. no. L00734)

• 5× membrane transfer buffer (eBlot™ L1 PVDF Membrane Transfer Buffer, 5×, GenScript, cat. no. L00733)

• Isopropanol (Sigma, cat. no. 190764, or equivalent)

• 20× Cathode Buffer Additive (NativePAGE™ Cathode Buffer Additive, 20×, Invitrogen, cat. no. BN2002)

• Pre‐cast mini protein gel (NativePAGE™ 4%‐16%, Bis‐Tris, 1.0 mm, Mini Protein Gel, 10‐well, Invitrogen, cat. no. BN1002)

• Unstained protein standard (NativeMark™ Unstained Protein Standard, Invitrogen, cat. no. LC0725)

• Absolute methanol, anhydrous, ≥99.8% (VWR, cat. no. MK301668, or equivalent)

• 10% (v/v) acetic acid (see recipe)

• Western blot blocking solution (see recipe)

• Anti‐Moloney‐MLV‐Gag‐polyprotein rabbit polyclonal antibody (Rabbit Anti‐Moloney MLV Gag Polyprotein Polyclonal Antibody, Unconjugated, MyBioSource, cat. no. MBS1490637)

• Anti‐SARS‐CoV‐2‐Spike rabbit polyclonal antibody (Anti‐SARS‐CoV‐2 (2019‐nCoV) Spike RBD Antibody, Rabbit PAb, Antigen Affinity Purified, Sino Biological, cat. no. 40592‐T62)

• Anti‐rabbit goat antibody‐peroxidase (Peroxidase AffiniPure Anti‐Rabbit Goat IgG, H+L, Jackson ImmunoResearch, cat. no. 111‐035‐003)

• Western blot wash solution (see recipe)

• Peroxidase and luminol/enhancer reagents (from Clarity™ Western ECL Substrate, Bio‐Rad, cat. no. 1705060)

• 1.5‐ml microcentrifuge tubes

• XCell SureLock™ Mini‐Cell (Invitrogen, cat. no. EI0001)

• Electrophoresis power supply

• 10‐ml serological pipet

• Blotting gel knife (Invitrogen or equivalent)

• Shallow plastic containers/trays

• PVDF membrane (eBlot™ L1 PVDF Membrane, GenScript, cat. no. L00735, or equivalent)

• Blotting tweezers (Invitrogen or equivalent)

• Transfer sponges (eBlot™ L1 Transfer Sponge, GenScript, cat. no. L00736)

• Transfer system (eBlot™ L1, GenScript, cat. no. L00686, or equivalent)

• Blotting roller (Invitrogen or equivalent)

• Laboratory rocker

• Autoradiography cassette (Fisher Scientific, cat. no. FBVC810, or equivalent) or gel imaging system (Bio‐Rad or equivalent)

• Autoradiography film (Western Blotting autoradiography film, Blu‐Lite UHC, Chemglass, cat. no. CLS‐1915‐115, or equivalent)

• Film‐processing machine (Nikon or equivalent)

Blue native gel electrophoresis

• 1Prepare 1× sample buffer (from 4×) and 0.5% Triton X‐100 (from 10%) with deionized water in a 1.5‐ml microcentrifuge tube and solubilize each eVLP sample in this solution.

  A typical starting amount to test for SARS‐CoV‐2‐Spike is 500 ng VLPs per well. Typically, 10 µl loading sample is added to a single well of the gel.

  Example preparation for a VLP sample with an initial concentration of 1 µg/µl is as follows:

  1 µg/µl VLP sample	1 µl
  4× sample buffer	5 µl
  10% (v/v) Triton X‐100	1 µl
  Deionized water	13 µl
  Total	20 µl

• 2Mix well by flicking bottom of the tube a few times and incubate eVLP samples on ice for 15 min.
• 3Add 5% G‐250 sample additive to solubilized sample to a concentration that is 1/4 of the final concentration of detergent in the sample.

  In the example above, the final concentration of Triton X‐100 is 0.5%. To obtain 1/4 of this concentration (i.e., 0.125%), add 0.5 µl of 5% G‐250 sample additive to the 20 µl loading sample.

• 4Dilute 20× running buffer and 10× PVDF membrane equilibration buffer in deionized water to their working concentration of 1×. Dilute 5× PVDF membrane transfer buffer to a working concentration of 1× using deionized water and isopropanol to a final concentration of 10%.

  For a 1‐L solution, use 200 ml of 5× PVDF membrane transfer buffer, 100 ml isopropanol, and 700 ml deionized water.

• 5Prepare Light Blue Cathode, Dark Blue Cathode, and Anode Buffers from 20× running buffer, deionized water, and/or 20× Cathode Buffer Additive according to the manufacturer's instructions. Cool all buffers to 4°C before use.
• 6Set up a pre‐cast mini protein gel in an XCell SureLock™ Mini‐Cell for blue native gel electrophoresis in a cold room according to the manufacturer's instructions.

  Follow the manufacturer's steps for detergent‐containing samples and for western blotting as the downstream experiment.

• 7Load samples and 5 µl unstained protein standard into their respective lanes of the gel.

  The samples and protein standard are loaded prior to filling up the upper buffer because the Dark Blue Cathode Buffer impedes visibility, making loading difficult. For best results, load all lanes with the same volume of sample buffer, even where there is no sample.

• 8Fill upper buffer chamber to the brim with Dark Blue Cathode Buffer (∼200 ml; from step 5). Add 500 to 600 ml Anode Buffer to lower buffer chamber.
• 9Run gel at 150 V for 30 to 35 min using an electrophoresis power supply.

  At this point, the dye front should be at 1/3 of the way through the gel. If not, continue the electrophoresis until the dye front is 1/3 of the way.

• 10Pause run. Aspirate Dark Blue Cathode Buffer from the upper buffer chamber using a 10‐ml serological pipet. Fill empty upper buffer chamber with ∼200 ml Light Blue Cathode Buffer.
• 11Continue gel electrophoresis at 250 V for 50 to 60 min or until the dye front reaches the foot of the gel.
• 12Crack open gel case. Remove well separators and protruding foot of the gel with a blotting gel knife.
• 13Release gel in a shallow plastic container/tray of deionized water and allow it to equilibrate for 5 min.
• 14Meanwhile, soak a PVDF membrane in absolute methanol for a few seconds to activate, handling membrane with clean blotting tweezers. Equilibrate PVDF membrane in 10 ml of 1× PVDF membrane equilibration buffer for 1 to 2 min.

  Always handle the PVDF membrane with clean blotting tweezers to prevent contamination.

• 15Assemble transfer stack, including transfer sponges, on the transfer cassette of the transfer system (eBlot™ L1 or equivalent) according to the manufacturer's instructions. Remove any visible bubbles between nitrocellulose membrane and gel using a blotting roller.

  Handle the gel via its thicker bottom side to avoid breakage or tearing.

• 16Close transfer cassette and slot it into either channel of the transfer system. Select pre‐programmed “standard” 17‐min transfer and start process.
• 17When the transfer is complete, remove transfer cassette from the transfer system. Open cassette and discard transfer sponges and gel.

  Check that no part of the membrane was obscured by an air bubble during the protein transfer.

• 18Fix proteins on PVDF membrane in 10% acetic acid for 15 min. Rinse membrane in deionized water.
• 19De‐stain membrane by removing excess Coomassie dye by soaking membrane briefly (≤5 s) in absolute methanol. Remove membrane from methanol and rinse in deionized water when the protein standards are sufficiently visible.
• 20Block membrane by incubating it in western blot blocking solution overnight at 4°C.

  The membrane can also be blocked by rocking at room temperature for ≥1 hr.

Western blotting of the transferred proteins

• 21Dilute anti‐Moloney‐MLV‐Gag‐polyprotein rabbit polyclonal antibody 1:10,000, anti‐SARS‐CoV‐2‐Spike rabbit polyclonal antibody 1:2000, and anti‐rabbit goat antibody‐peroxidase 1:5000 with western blot blocking solution.
• 22Incubate membrane with the appropriate primary antibody solution for 1 hr.

  Always handle each membrane with clean blotting tweezers to avoid contamination.

  Ensure that the entire surface of the membrane is covered by the antibody solution.

• 23Wash membrane in 20 to 30 ml western blot wash solution on a laboratory rocker for 5 min. Repeat this washing step two more times with fresh western blot wash solution.
• 24Incubate membrane with the peroxidase‐conjugated secondary antibody solution for 1 hr.

  Ensure that the entire surface of the membrane is covered by the antibody solution.

• 25Wash membrane in 20 to 30 ml western blot wash solution on a laboratory rocker for 5 min. Repeat this washing step two more times with fresh western blot wash solution.
• 26Mix equal volumes of peroxide reagent and luminol/enhancer reagent to generate ECL substrate.

  Prepare a sufficient volume of the ECL substrate to cover the entire surface of each membrane.

• 27Incubate membrane with ECL substrate for 2 to 5 min. Blot off excess ECL substrate from membrane.
• 28Place membrane in between plastic sheets within an autoradiography cassette. Alternatively, capture chemiluminescence on the membrane using a gel imaging system.
• 29Expose autoradiography film to membrane in a darkroom and develop film in a film‐processing machine.

  Typically, start with an exposure time of 1 min and adjust accordingly to obtain a sharp image.

• 30Mark positions of the protein standards on the developed film to estimate the sizes of the protein bands.

  Scan the developed and marked film for recording purposes.

  See Figure 10 for example results.

Alternate Protocol 2. EVALUATION OF eVLP QUALITY BY ELECTRON MICROSCOPY (NEGATIVE STAINING)

A quality assessment of the eVLPs should be performed for each preparation. Electron microscopy (negative staining) is a powerful tool to assess both the purity and the integrity of eVLPs and is highly recommended to be done at this stage. This technique also identifies the relative amount of contaminating exosomes or debris that may exist in each preparation.

Materials

• eVLP samples (see Basic Protocol 5)

• Distilled water

• Uranyl acetate (Ted Pella, cat. no. 19481)

• Grids (200‐mesh carbon‐coated copper grids, Electron Microscopy Sciences, cat. no. CFC200‐Cu)

• Glow discharger (SPI Plasma Prep III or equivalent)

• Filter papers

• Petri dishes or grid storage box

• 120‐kV transmission electron microscope with digital camera

NOTE: If your proteins are unstable and will break down overnight, they can be stabilized by fixing them with 2% paraformaldehyde and 2.5% glutaraldehyde (from Karnovsky's EM grade Fixative Kit, Electron Microscopy Sciences, cat. no. 15720) in 0.1 M Sorensen's buffer (from Sorensen's Phosphate Buffer 0.2 M, pH 7.2, Electron Microscopy Sciences; cat. no. 11600‐05) overnight at 4°C.

• 1Glow‐discharge grids with a glow discharger for 10 s to make them hydrophilic.
• 2Adsorb 5 µl eVLP sample on grid for 1 min.
• 3Wick off liquid by touching the grid edge with a filter paper.
• 4Rinse grid by adding a 10‐µl drop of distilled water onto the grid.
• 5Wick off water (see step 3).
• 6Prepare 2% (w/v) uranyl acetate stain in distilled water.

  CAUTION: Uranyl acetate is radioactive. Use PPE and collect and dispose of spent stain and items that come in contact with the stain as per your laboratory safety protocols.

  This stain is light sensitive. Cover the tube with aluminum foil and store the stain in a refrigerator.

• 7Add 5 µl of 2% aqueous uranyl acetate stain to grid and stain for 1 min.
• 8Wick off stain (see step 3).
• 9Place stained grids in a Petri dish with a filter paper or store in a grid storage box for long‐term storage.
• 10Image under a 120‐kV transmission electron microscope with a digital camera.

  Please see Figure 11 for example results.

Figure 11.

Negatively stained transmission electron microscopy (TEM) images of eVLPs expressing (A) SARS‐CoV‐2 Spike protein or (B and C) the receptor‐binding domain (RBD) of the Spike protein. Scale bars: 100 nm. Imaged on a Hitachi 7800 transmission electron microscope using a bottom‐mounted AMT camera at 80 kV.

Basic Protocol 5. SMALL‐SCALE PRODUCTION OF eVLPs

According to need, the eVLP production system can be scaled. In Basic Protocol 5 and Alternate Protocols 3 and 4, we describe three procedures to reach different levels of yield that can be achieved in a laboratory setting (Fig. 12). Table 3 summarizes the eVLP yields at different scales of production.

Figure 12.

Schematic diagram showing the steps to produce three different output levels of eVLPs. (A) Cells are seeded in six T175 tissue culture flasks until they are confluent. (B) Cells are seeded in three five‐layer tissue culture flasks and concentrated using a Centricon Plus‐70 centrifugal filter device. (C) Cell are seeded in a Cell Factory System (10 layers) and concentrated using a Vivaflow 50R filter operated through a Vivaflow peristaltic pump. (D) The supernatants containing eVLPs from the three different methods (A, B, or C) are purified and concentrated by ultracentrifugation through a 20% (w/v) sucrose cushion. After 2 hr of ultracentrifugation at 100,000 × g, the pellets of VLPs will be concentrated at the bottom of the tube.

Table 3.

eVLP Yields at Different Scales of Production

	Surface area	eVLP yield
6 × T175 flasks	1050 cm2	100‐300 µg
3 × five‐layer flasks (equivalent to 15 × T175 flasks)	2625 cm2	1‐3 mg
Cell Factory System (10 layers each; equivalent to 36 × T175 flasks)	6320 cm2	3‐5 mg

Materials

• Stable producer HEK293T cell lines expressing eVLPs of interest (see Basic Protocol 2)

• Complete DMEM (see recipe; make fresh), 37°C

• 20% (w/v) sterile sucrose (Sigma, cat. no. S0389; in 1× PBS)

• Sterile 1× PBS (Gibco, Fisher Scientific, cat. no. 10‐010‐031)

• T175 flasks (Falcon® Tissue Culture Flasks, Corning, 175 cm, plug seal, VWR, cat. no. 29185‐258)

• 50‐ml conical centrifuge tubes (CellPro™ 50 ml Centrifuge Tube, Polypropylene, Conical bottom, Alkali Scientific, cat. no. CW5602, or equivalent)

• Benchtop refrigerated centrifuge (Eppendorf Centrifuge 5810R, Marshall Scientific, cat. no. EP‐5810R, or equivalent), 4°C

• 0.45‐µm membrane filters (Medical Millex‐HP Syringe Filter Unit, 0.45 µm, polyethersulfone, 33 mm, Millipore Sigma, cat. no. SLHPM33RS)

• Ultracentrifuge tubes (38.5 ml, Open‐Top Thinwall Polypropylene Tubes, 25 × 89 mm, Beckman Coulter, cat. no. 326823)

• Scale

• Ultracentrifuge (Beckman Coulter Optima XPN‐80, cat. no. A95765, or equivalent) with swinging‐bucket rotor, 4°C

• Additional reagents and equipment for Bradford assay (see Support Protocol)

NOTE: All the procedures should be performed in a BSL‐2 biosafety cabinet.

NOTE: All solutions and equipment coming into contact with cells must be sterile, and proper sterile technique should be used accordingly.

NOTE: All culture incubations are performed in a 37°C, 5% CO₂ incubator unless otherwise specified.

• 1To obtain 200 ml supernatant containing eVLPs, seed six T175 tissue culture flasks with ∼5 × 10⁶ stable producer HEK293T cells expressing eVLPs of interest in ∼32 ml complete DMEM and incubate at 37°C, 5% CO₂ until cells reach 95% to 100% confluence and the supernatant becomes yellow (about 3 to 4 days), indicating that a large amount of VLPs have been secreted into the medium.
• 2Collect supernatant into 50‐ml conical centrifuge tubes and centrifuge 15 min at 2000 × g, 4°C, to remove cellular debris.
• 3Pool clarified supernatant and filter through a 0.45‐µm membrane filter.
• 4Aliquot eVLP suspension into clean ultracentrifuge tubes. Carefully underlay suspension with 5 ml sterile 20% sucrose cushion (in 1× PBS).

  Do not exceed 32 ml to avoid spillage during centrifugation, and loading <28 ml could result in collapsing of the tubes. Top up with 1× PBS to the appropriate volume when necessary.

  Using a pipettor on a very slow setting, slowly layer the 5 ml of 20% sucrose solution under the sample supernatant, placing the tip of the pipet at the bottom of the tube and trying to avoid any mixing of the suspension and the sucrose.

• 5Balance weights of the tubes accurately using a scale.
• 6Centrifuge suspensions for 2 hr at 100,000 × g, 4°C, in an ultracentrifuge with a swinging‐bucket rotor.

  The eVLP pellet will be formed underneath the sucrose cushion (Fig. 12D).

• 7Decant supernatant and leave tubes upside down for 1 to 2 min on a rack to drain excess supernatant. Then, plug open end of each tube with tissue paper.

  Generally, the pellet is well adhered to the bottom of the tube and will not be dislodged during decanting. However, as a precaution, it is recommended to turn the tube upside down only once and to handle with care.

• 8Remove tissue paper when the supernatant is completely drained.
• 9Resuspend eVLP pellet in ∼100 µl sterile 1× PBS by gently pipetting up and down.

  The volume can vary, depending on the size of the pellet and the downstream applications. Some researchers add 15% (w/v) trehalose (D‐(+)‐trehalose, anhydrous, Thermo Fisher, cat. no. J66006.18) to preserve the stability and integrity of their eVLPs during storage (Wang et al., 2017). Caution: trehalose may affect the antigenicity of capsid expressed proteins, which should be assessed for each eVLP preparation.

• 10Measure protein concentration by Bradford assay (see Support Protocol).

Alternate Protocol 3. LARGE‐SCALE PRODUCTION OF eVLPs (UP TO ABOUT 1 to 3 mg VLPs)

To generate larger amounts of eVLPs than in Basic Protocol 5, the stable producer HEK293T cell lines (Basic Protocol 2) should be grown in five‐layer tissue culture flasks (Fig. 12B).

Additional Materials (also see Basic Protocol 5)

• Trypsin‐EDTA (0.25%), phenol red (Gibco, cat. no. 25200114)

• Five‐layer tissue culture flasks (Falcon® 875 cm² Cell Culture Multi‐Flask, 5‐layer with Vented Cap, Corning, cat. no. 353144, or 5‐Layer Cell Culture Flask, Vent Cap, Thomas Scientific, cat. no. 1194Z19)

• Centrifugal filter device (Centricon™ Plus‐70 Centrifugal Filter Units, Millipore Sigma, cat. no. UFC710008)

• 1To obtain 450 ml supernatant containing eVLPs (approximate yield: 1 to 3 mg total protein), seed two T175 tissue culture flasks with ∼5 × 10⁶ stable producer HEK293T cells in ∼32 ml complete DMEM to produce eVLPs of choice and incubate at 37°C, 5% CO₂.
• 2When the cells reach ∼95% confluence (∼2 days), aspirate medium and add 2 to 3 ml trypsin‐EDTA. Gently swirl trypsin over the cells and remove excess quickly.
• 3Leave remaining trypsin for 1 to 2 min at room temperature (or 1 min at 37°C) and resuspend cells in each flask by gentle pipetting with 10 ml complete DMEM to inactivate the trypsin.

  The cells should not show any visible clumping.

• 4Pool cell suspensions and top up to 450 ml with complete DMEM.
• 5Use this cell suspension to seed five‐layer tissue culture flasks (150 ml per flask).

  A comprehensive explanation of how to successfully operate these flasks can be found in the following links:

  https://www.youtube.com/watch?v=wO2QxWsEFzo

  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3369669/

• 6Incubate cells at 37°C, 5% CO₂ until the cells reach about 95% to 100% confluence and the supernatant becomes yellow (about 3 to 4 days), indicating that a large amount of VLPs have been secreted into the medium.
• 7Decant supernatant in large centrifuge tubes and centrifuge 15 min at 2000 × g, 4°C, to remove cellular debris.
• 8Pool clarified supernatant and pass through a 0.45‐µm membrane filter.
• 9Concentrate eVLP suspension using a centrifugal filter device and an ultracentrifuge until the desired volume is obtained.

  CAUTION: Centrifugation of eVLPs can increase the risk of shedding of incorporated viral glycoproteins within the viral envelope. It is a good idea to assess the stability of the pseudotyped eVLPs after concentration to monitor for loss of antigen.

  The time until the desired volume is obtained depends on the capacity of the ultracentrifuge rotor. If a Beckman SW28 Centrifuge Rotor 6‐Position is available, it is possible to concentrate an initial volume of ∼192 ml supernatant (loading 32 ml per ultracentrifuge tube, with six tubes total). Two rounds of centrifugation for ∼45 min each at 4°C in a refrigerated benchtop centrifuge should be sufficient. The maximum centrifugal force during concentration should not exceed 3500 × g, and during recovery, it should not exceed 1000 × g, according to the manufacturer's instructions.

• 10Follow steps 4 to 9 of Basic Protocol 5.
• 11Measure protein concentration by Bradford assay (see Support Protocol).

Alternate Protocol 4. LARGE‐SCALE PRODUCTION OF eVLPs (UP TO ABOUT 3 to 5 mg VLPs)

Using a Cell Factory System, with 10 layers, production of VLPs can be substantially increased (Fig. 12C).

Additional Materials (also see Basic Protocol 5)

• Trypsin‐EDTA (0.25%), phenol red (Gibco, cat. no. 25200114)

• 10‐layer Cell Factory System (Nunc™ EasyFill™ Cell Factory System, 10 layers, Thermo Scientific, cat. no. 140410, or equivalent)

• 250‐ml conical centrifuge tubes (Thermo Scientific, cat. no. 376814, or equivalent)

• 2000‐ml bottles (Nalgene™ Square PETG Media Bottles with Closure, 2000 ml, closure size: 53B, Thermo Fisher Scientific, cat. no. 2019‐2000)

• Vacuum filter units (Nalgene™ Rapid‐Flow™ Sterile Single Use Vacuum Filter Units, 0.45 µm, PES, 1 L, Thermo Fisher Scientific, cat. no. 167‐0045)

• Vivaflow device (Vivaflow 50R Crossflow Cassette Hydrosart 100,000 MWCO, Sartorius, cat. no. VF05H4) with pump (Vivaflow Pump (115V), Easy load pump head (size 16), tubing, 500 ml sample diafiltration reservoir, Sartorius, cat. no. VFS204)

• 1To obtain a final volume of 2 L supernatant containing a large amount of eVLPs (approximate yield: 3 to 5 mg total protein), seed two T175 tissue culture flasks with ∼5 × 10⁶ stable producer HEK293T cells in ∼32 ml complete DMEM and incubate at 37°C, 5% CO₂ until cells reach ∼95% confluence (∼2 days).
• 2When the cells are ready, aspirate medium and add 4 to 5 ml trypsin‐EDTA. Gently swirl trypsin over the cells and remove excess quickly.
• 3Leave for about 1 to 2 min at room temperature (or 1 min at 37°C) and resuspend cells in each flask by gentle pipetting with 20 ml complete DMEM to inactivate trypsin.

  The cell should not present any visible clumps.

• 4Pool two cell suspensions and top up to 2 L with complete DMEM.
• 5To seed the cells, add total volume of the suspension to a 10‐layer Cell Factory System.

  A succinct explanation of how to successfully operate these flasks can be found at the following link:

  https://www.youtube.com/watch?v=y5wI5tWVaHU

• 6Incubate at 37°C, 5% CO₂ until cells reach confluence and supernatant becomes yellow (about 4 to 5 days), indicating that a large amount of VLPs have been secreted into the medium.
• 7Collect supernatant in 250‐ml conical centrifuge tubes and centrifuge 15 min at 2000 to 4000 × g, 4°C, to remove cellular debris.
• 8Pool clarified supernatant into a 2000‐ml bottle and filter through a 0.45‐µm membrane filter using a vacuum filter unit.
• 9Concentrate suspension using a Vivaflow device with a pump until the desired volume is reached following the manufacturer's instructions.

  If the starting volume is 2 L, the final volume after concentration should be ∼10‐fold less. Small final volumes can lead to protein precipitation (the formation of aggregates) and should be avoided.

  CAUTION: Centrifugation of eVLPs can increase the risk of shedding incorporated viral glycoproteins within the viral envelope. It is a good idea to assess the stability of the pseudotyped eVLPs after concentration to monitor for loss of antigen.

• 10Follow steps 4 to 9 of Basic Protocol 5.
• 11Measure protein concentration by Bradford assay (see Support Protocol).

QUANTIFICATION OF TOTAL PROTEIN CONCENTRATION BY BRADFORD ASSAY

The total protein concentration for the protein from Basic Protocol 5 and Alternate Protocols 1 and 2 can be determined using standard protein assays such as BCA or Bradford. Here, we describe a protocol for Bradford measurement.

Materials

• Bio‐Rad Protein Assay Kit II (Bio‐Rad, cat. no. 5000002), containing dye reagent concentrate and bovine serum albumin (BSA) standard

• Deionized water

• 1× PBS

• Protein samples (see Basic Protocol 5 or Alternate Protocol 3 or 4)

• Whatman #1 filter or equivalent

• Funnel

• Eppendorf tubes

• 96‐well plate

• Plate reader (SpectraMax i3 plate reader, Molecular Devices, or equivalent)

Preparation of 1× Bradford reagent

• 1Prepare Bradford dye reagent by diluting 1 part dye reagent concentrate from the Bio‐Rad Protein Assay Kit II with 4 parts 1× PBS.
• 2Filter through Whatman #1 filter or equivalent, inserted into a funnel, to remove particles.

  The diluted dye reagent can be kept for ≤2 weeks at 4°C.

Preparation of the standards

• 3Reconstitute lyophilized BSA standard from the Bio‐Rad Protein Assay Kit II with deionized water.

  Use the reconstituted standard within 60 days or store aliquots at −20°C.

• 4Prepare at least 3 to 5 concentrations of BSA standard (at least in duplicate) in Eppendorf tubes 1× PBS as protein standards.

  The concentrations should be inclusive of the concentration of the protein solution to be tested.

  For example, concentrations could include 50 µg/ml, 100 µg/ml, 200 µg/ml, 300 µg/ml, 400 µg/ml, 600 µg/ml, and 800 µg/ml.

Bradford assay

• 5Combine 5 µl of each standard (step 4) with 200 µl diluted dye reagent (step 2) in wells of a 96‐well plate. Do the same with the protein samples.
• 6Incubate at room temperature for ≥10 min.

  Absorbance will increase over time. Samples should be incubated for no more than 1 hr.

• 7Read sample absorbance at 595 nm with a plate reader.

  The apparatus will give you the concentration already calculated. Alternative, copy the absorbances and do your own calculations using Prism software. Determine sample absorbance by interpolation from the protein standard curve (linear regression or second‐order polynomial best fit).

REAGENTS AND SOLUTIONS

Acetic acid, 10%

• 50 ml acetic acid, glacial (Fisher Scientific, cat. no. BP2401‐500, or equivalent)

• 450 ml ddH₂O

• Swirl to mix

• Store ≤1 year at room temperature

CAUTION: Preparation should be conducted in a fume hood, as glacial acetic acid is caustic and can burn the skin or lungs if exposed.

Antibody dilution buffer

• 1× PBS (Gibco, Fisher Scientific, cat. no. 10‐010‐031)

• 1% (v/v) FBS (Biowest, cat. no. S1620, or equivalent)

• 1% (v/v) normal goat serum (NGS; Sigma, cat. no. G9023)

• Store ≤3 weeks at 4°C

Complete DMEM

• 450 ml Dulbecco's modified Eagle's medium (DMEM), high glucose (Millipore Sigma, cat. no. D5796)

• 25 or 50 ml FBS (Biowest or equivalent), heat inactivated for 30 min at 56°C (5% or 10% final)

• 5 ml 200 mM GlutaMAX™ (100×; Gibco, cat. no. 35050‐061) (2 mM final)

• 2.5 ml pen/strep (10,000 U/ml penicillin, 10,000 µg/ml streptomycin; Gibco, cat. no. 15140‐122) (100 U/ml and 100 µg/ml final)

• 0.25 ml 50 mg/ml gentamicin (Gibco, cat. no. 15750078) (25 µg/ml final)

• Prepare fresh immediately before use

Fixative solution

• 25 ml ddH₂O

• 5 ml 10× PBS

• 20 ml 10% formaldehyde, methanol free, ultrapure (Polysciences, cat. no. 04018‐1)

• 40 µl 27% glutaraldehyde solution, grade 1, 25% (Sigma‐Aldrich, cat. no. G5882)

• Store ≤1 month at room temperature

Triton X‐100, 10%

• 100 µl Triton™ X‐100 (Sigma‐Aldrich, cat. no. X100, or equivalent)

• 900 µl ddH₂O

• Vortex to mix

• Store ≤1 year at room temperature

Western blot blocking solution

• 25 g instant non‐fat dry milk

• 1× PBS, pH 7.4, to 500 ml

• Mix until completely dissolved

• Store ≤1 week at 4°C

This solution is 5% (w/v) milk in PBS.

Western blot wash solution

• 1 L 1× PBS, pH 7.4

• 500 µl TWEEN® 20 (Sigma‐Aldrich, cat. no. P1379, or equivalent)

• Mix until completely dissolved

• Store ≤1 week at room temperature

This solution is 0.05% TWEEN 20 in PBS.

COMMENTARY

Background Information

VLPs are nanostructures that resemble actual viruses but are non‐infectious and safe. They are empty capsids that are unable to replicate because they are devoid of genetic material. Hence, they are attractive molecular tools for gene therapy and vaccine development. Chimeric eVLPs expressing heterologous antigens are excellent platforms to produce protein subunit vaccines (Roldão et al., 2010). Due to their particulate nature and the repetitive distribution of immunogens on their envelope, they elicit strong humoral and cellular immunity. In addition, they are strong activators of dendritic cell maturation, resulting in the development of strong adaptive immune responses (Zepeda‐Cervantes, Ramírez‐Jarquín, & Vaca, 2020). The possibility of expressing a variety of immunogens simultaneously on the surface of these eVLPs makes them a particularly attractive and versatile platform for generation of different vaccine formulations. Several expression systems are currently used to produce VLPs, such as bacteria, yeast, insect cells/the BEVS, plants, and mammalian cells. Bacteria and yeast are excellent systems for VLP production when low cost and scaling up are priorities (Tanner & Lehle, 1987). The BEVS and plants offer another good platform to produce high‐quality VLPs expressed at high levels. However, these systems have fundamental limitations due to differences in their glycosylation pathways compared to mammalian cells (Kost, Condreay, & Jarvis, 2005; Margolin et al., 2021). Extensive glycosylation is found on the envelope proteins of human viruses (Goffard et al., 2005; Grant, Montgomery, Ito, & Woods, 2020; Hebert, Zhang, Chen, Foellmer, & Helenius, 1997; Ohuchi, Ohuchi, Feldmann, & Klenk, 1997; Ohuchi, Ohuchi, Garten, & Klenk, 1997; von Messling & Cattaneo, 2003), and this can affect both the immunogenicity and the antigenicity of the VLPs produced. This is the case for envelope proteins from Ebola, influenza, HIV, and SARS‐CoV‐1 and SARS‐CoV‐2. Hence, the type of expression system used to produce eVLPs can greatly impact the success of the derived vaccine. Complex naturally occurring glycosylation pathways are only present in mammalian cell expression systems, and this must be taken into consideration when eukaryotic cells are chosen to produce pseudotyped eVLPs. For example, the HEK293, CHO, and BHK‐21 cell lines can produce recombinant proteins with complex patterns of glycosylation suitable for human use.

Another important parameter to consider is whether the VLPs are produced transiently or constitutively (by stable transfection). Each of these two options has its own advantages and disadvantages. Transiently transfected genetic material is internalized into cells with high efficiency but is not integrated into the host genome and is lost due to environmental factors and cell division. Therefore, foreign genes are only expressed for a limited time (Kim & Eberwine, 2010). The main advantage of transient expression is that the process, from obtaining the purified plasmid construct to protein expression, takes only a few days. This is very useful for diffraction techniques in structural biology (Nettleship, Assenberg, Diprose, Rahman‐Huq, & Owens, 2010). However, production of VLPs for vaccination during preclinical studies often requires scaling up to liter volume quantities generated from multiple transient transfections. This requires larger amounts of purified plasmid and more transfection reagents and can result in batch‐to‐batch variability. In this case, stable transfection is the method of choice. Once a stable cell line is established, the expression of proteins is fast. Higher yields are usually obtained, and no variability typical of transient transfection occurs. However, a significant drawback to producing stable cell lines is the length of time it takes to make stable integrants of the recombinant DNA in the host cell genome. This process can take 4 weeks to 6 months (Nettleship et al., 2010), usually because the frequency of integration is extremely low, with only approximately 1 in 10⁴ cells in a transfection stably integrating DNA (see Current Protocols article: Mortensen, Chestnut, Hoeffler, & Kingston, 2001). The DNA introduced into the cell harbors a selectable marker that confers resistance to drugs, which applies a positive selection pressure that facilitates the recovery of rare integrations into the host nuclear genome, resulting in the constitutive, stable expression of the transgene in the host cell. However, random insertion of foreign DNA has a potentially deleterious effect if integrated at sites that affect host cell fitness or viability. One approach to overcome this problem is to target the integration of YFG at a specific genomic location, such as via the Flp‐in system (O'Gorman, Fox, & Wahl, 1991). The advantage of using such a targeted approach is the production of a stable cell line; however, it typically takes ≥3 months to generate such a line. This process is lengthy largely because it requires multiple transfections [to first insert the Flp Recombination Target (FRT) site and then the heterologous protein of interest]. Further, it is often not possible to express more than one heterologous protein reliably and stably. Moreover, if eVLPs pseudotyped with more than one heterologous protein are required, transient transfection of the second target protein is typically performed, which introduces batch effects and variations in valency that cannot be easily controlled.

To overcome many of the drawbacks listed above, we re‐purposed a stable transposon system, known as the Sleeping Beauty (SB) transposon system, which uses the non‐viral Tc1 Mariner transposon to promote gene integration at TA sites within the host genome. This significantly decreases the probability of insertions at transcriptional units. It is fast and highly efficient, and more importantly, it can be adapted to stably express three or more genes in a matter of 1 to 2 weeks. Depending on your needs, if only one envelope antigen needs to be incorporated, a stable cell line can be generated by co‐transfection of YFG with a GAG polyprotein to constitutively generate eVLPs pseudotyped with a specific antigen. The GAG polyprotein typically used for this purpose is derived from either the HIV‐1 lentivirus or MLV. Alternatively, if the aim is to produce a more versatile platform involving the incorporation of a suite of different envelope antigens, it may prove advantageous to generate a “universal” cell line that stably expresses the GAG polyprotein to constitutively produce non‐pseudotyped eVLPs that can be pseudotyped with one or more antigens by additional downstream transfections with plasmids encoding YFG(s). This is possible because the GAG polyprotein, which plays a fundamental role in the correct assembly of immature virus particles and budding, can self‐assemble and generate empty capsids. Its sole expression in the host cell is sufficient for the organization of the empty capsid and for budding of the VLP from the host cell.

Here, we chose the SARS‐CoV‐2 Spike protein as a model antigen to pseudotype naked universal eVLPs. As described in the next section (Critical Parameters and Troubleshooting), several important points should be considered to guarantee the correct expression of YFG on the surface of the eVLPs. Fully understanding the protein structure is critical for this purpose. As an example, here, we describe the structural characteristics of our model antigen, the SARS‐CoV‐2 Spike protein, which is responsible for virus entry into the host cell and is the main target of neutralizing antibodies. It is a trimeric type 1 fusion glycoprotein composed of two subunits, S1 and S2. The S1 subunit contains the receptor‐binding domain (RBD), which binds to the human angiotensin‐converting enzyme 2 (hACE2) receptor, and the S2 subunit promotes the fusion of the viral envelope with the host cell membrane during the entry process. The Spike protein is cleaved by the cellular protease furin at the S1/S2 cleavage site, exposing a second cleavage site, the S’2, that is cut by the transmembrane protease serine protease 2 (TMPRSS‐2). This proteolytic activation induces a transition from the metastable pre‐fusion trimer to the stable post‐fusion conformation, triggering virus‐host membrane fusion. It has been demonstrated previously that abrogation of the furin cleavage site (RRAR to GSAS) as well as mutations in two prolines within the hinge region of S2 (K986P, V987P) is sufficient to stabilize the Spike protein in a pre‐fusion conformation, which increases its ability to induce levels of neutralizing antibodies necessary to confer protection (Bos et al., 2020, Wrapp et al., 2020). The N‐terminus of the Spike protein harbors an endoplasmic reticulum (ER) signal sequence. The C‐terminus possesses a TM domain and a CT that contains an ER retrieval signal that is used to retain the protein in the ER‐Golgi intermediate compartment (ERGIC) prior to assembly and budding of the viral particles from the endocytic compartment.

Critical Parameters and Troubleshooting

Basic Protocol 1

Prior to cloning, it is important to create all constructs in silico, and several aspects should be taken into consideration:

a. Codon optimization is an important step when expressing a heterologous gene construct in HEK293T cells. This procedure will improve gene expression and increase the translation efficiency of YFG. Multiple algorithms are available online to assess codon usage that do not alter the amino acid sequence. In our study, we used the free tool provided by GENEWIZ (https://www.genewiz.com/en/Public/Services/Gene‐Synthesis/Codon‐Optimization).

b. Careful examination of the sequences after codon optimization should be performed given that new restriction sites that might interfere with potential downstream subcloning can be created.

c. To ensure the correct assembly of the envelope antigens and budding of the eVLPs at the plasma membrane of the host cell, each SB transposon construct should encode the ectodomain of the envelope protein fused to the TM domain and CT of VSV‐G. It has been shown that swapping the CT and TM domains of heterologous proteins for those of the VSV‐G protein greatly increases the efficiency of pseudotyping (Garrone et al., 2011; Kirchmeier et al., 2014). The MLV Gag polyprotein is localized to the cytosolic face of the plasma membrane of the host cell and colocalizes with the CT of VSV‐G. This interaction initiates self‐assembly, forming the empty capsid and resulting in subsequent budding of the eVLPs from the host cell.

d. Any ER retrieval signal of YFG needs to be removed to avoid the protein's accumulation near the ERGIC and to allow for correct targeting to the plasma membrane of the host cell (see Background Information). For example, we removed the ER retrieval signal at the C‐terminus of the mature Spike protein, which would otherwise have retained the protein in the ERGIC and fused the ectodomain of the protein to the CT and TM domains of the VSV‐G protein to promote appropriate assembly of YFG at the plasma membrane of the host cell, the site of VLP budding and release.

e. Virus envelope proteins are typically glycoproteins expressed as oligomers (Xiao, Feng, Chakraborti, & Dimitrov, 2004). Trimers are common three‐dimensional structures that confer important immunogenic properties (Habte, Banerjee, Shi, Qin, & Cho, 2015; Kovacs et al., 2012). They possess trimerization domains that stabilize conformational epitopes important for generation of neutralizing antibodies. In situations in which the trimeric domain is removed during the construction of chimeric fusion envelope proteins, incorporation of a heterologous foldon domain into the ectodomain is recommended (Meier, Güthe, Kiefhaber, & Grzesiek, 2004).

f. In the case of production of eVLPs for vaccination, the size of the antigen may negatively impact the quality of the eVLPs, such as leading to poor expression on the envelope or less yield of VLPs. In such cases, identification of the regions involved in pathogenicity and their immunogenic capability might be required. Sometimes, the sole expression of these regions is enough to confer anti‐pathogen protection.

Basic Protocol 2

Generation of a stable HEK293T cell line constitutively secreting MLV‐based “naked” eVLPs (non‐pseudotyped)

a. It is important to maintain the right ratio between the amount of the transposon construct and the SB100X transposase to avoid overexpression inhibitory effects (OIEs) that occur when the transposase is present in excess (Grabundzija et al., 2010). Optimal results are obtained when using low amounts of the transposase vector (5% to 10% of total transfected DNA) (Kowarz et al., 2015).

b. Genomic integration frequency is another aspect to consider depending on the application. Single insertions are advantageous to limit loss‐of‐function mutations and promote safe gene transfer practices in human gene therapeutic applications, whereas multiple insertions are required for somatic mutagenesis for cancer discovery (DeNicola, Karreth, Adams, & Wong, 2015; Grabundzija et al., 2010). The number of insertions can be tuned by varying the transposon construct dosage. For instance, at low transposon dosage (∼10 ng), most of the transfected cells have a single insertion. At high transposon dosage (∼1.9 µg), an average of 2 to 6 insertions genome‐wide are typically obtained (Kowarz et al., 2015). In the protocol described here, a high transposon dosage was used to promote high levels of expression for the transfected gene to increase the output of eVLPs generated (Turchiano et al., 2014).

Pseudotyping “naked” eVLPs with one or more antigens

a. If co‐transfection with two SB transposon plasmid constructs is required, the DNA/Opti‐MEM solution to generate the PEI MAX/DNA complex should be prepared by mixing 0.95 µg plasmid 1 (pSBbi‐X‐Hygromycin) with 0.95 µg plasmid 2 (pSBbi‐XX‐HA‐Puromycin) and 100 ng pSB100X vector. Alternatively, sequential transfections with each construct are likewise possible.

b. Although three or more antigens can be displayed on the lipid envelope of eVLPs (Serradell et al., 2019; Yan, Wei, Guo, & Sun, 2015; Nooraei et al., 2021), there is a limit imposed by the number of antibiotic resistance genes cloned into the series of SB plasmids available at Addgene (plasmids #60495 to #60526). Currently, there are four different selectable markers: blasticidin, hygromycin, puromycin, and neomycin. If HEK293T cells are used for transfection, neomycin is not advisable because HEK293T cells are immortalized with the SV40 large T antigen, which increases protein expression and confers neomycin resistance.

c. We recommend monitoring the long‐term expression of YFG. Previous reports using this system indicate that HEK293T cells expressing a reporter GFP are stable for ≥1 month (Kowarz et al., 2015). We monitored the expression of fluorescent proteins in one of our stable HEK293T cell lines transfected with the reporter genes encoding RFP and BFP (Fig. 8), and the heterologous gene expression persisted for ≥5 months in the absence of drug selection.

Understanding Results

Several mammalian cell lines are available to produce eVLPs by either transient or stable transfection. The methodology and platform used here allow generation of a highly efficient multivalent vaccine in a short time, overcoming the laborious and time‐consuming processes of the systems currently employed in the field. If the expression of multiple antigens is required, the SB transposon system allows for generation of stable cell lines that constitutively express eVLPs, with a dramatic reduction of the time and reagents needed to produce large amounts of eVLPs compared to the currently available systems. The limitation of selectable markers available for the SB plasmids on Addgene (up to three when using HEK293T cell lines) can be easily overcome, generating different VLPs expressing different sets of envelope proteins that can be administrated as a pool.

The methodology presented here also facilitates large‐scale production of eVLPs that allow completion of immunization schedules using the same batch of VLPs during preclinical studies, eliminating variation of batch‐to‐batch production when transient transfection is performed.

Time Considerations

Basic Protocol 1

Generation of the SB plasmids takes approximately 7 to 8 days.

Basic Protocol 2

Use of the SB system significantly reduces the time required to generate a stable cell line expressing YFG. However, use of this system to produce eVLPs for vaccination requires a thorough understanding of how the empty capsids are formed and the selection of appropriate antigens as envelope proteins for their capability to generate neutralizing antibodies against the pathogen of interest.

Generation of a stable HEK293T cell line constitutively releasing non‐pseudotyped MLV‐based eVLPs takes ∼8 days. Once this stable cell line is established, the secreted eVLPs can be considered a versatile “universal platform” suitable for generation of any eVLP‐based subunit vaccine. Further transfections with any plasmid containing coding sequences for the expression of envelope glycoproteins of interest will be required.

When one or two antigens need to be expressed on the envelope of the eVLPs, only one transfection with the two additional plasmids needs to be carried out. The entire process of pseudotyping “universal eVLPs” with 1 or more antigens will take again ∼8 days.

Basic Protocols 3 and 4 and Alternate Protocols 1 and 2

Validation of the SB constructs and VLPs requires 2 days for indirect IFA (2 days), 2 days for SDS‐PAGE under reducing conditions and western blot, 2 days for native blue gel electrophoresis and western blot, and 30 to 45 min for electron microscopy (negative staining) from preparation of one grid to imaging. Sample preparation only takes 10 to 15 min. The most time‐consuming part is imaging, as that would depend on the sample quality.

Basic Protocol 5

The time required to produce about 100 to 300 μg eVLPs is approximately 3 to 4 days, depending on the initial inoculum.

Alternate Protocol 3

Increasing the yield up to about 1 to 3 mg VLPs takes approximately 6 to 7 days in total. This includes expansion until appropriate cell confluence is reached in the three multilayer flasks (4 to 5 days), depending on the initial inoculum seeded in the multilayer flask. One day is needed for removal of cell debris by centrifugation, filtration, and concentration, and 1 day is needed for purification by ultracentrifugation.

Alternate Protocol 4

The time required for increasing the yield up to about 3 to 5 mg VLPs will be increased by 2 to 3 days with respect to Alternate Protocol 3, as it will require an additional step to expand the cells before seeding the Cell Factory System. The cells are first grown in two T175 tissue culture flasks, and when they reach confluence, they are passaged into the Cell Factory System. After 5 days of incubation at 3°C, 5% CO₂, the supernatant should be usually ready for concentration and ultracentrifugation.

Support Protocol

Total protein concentration measured by Bradford assay takes 2 hr.

Author Contributions

Viviana Pszenny: Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Writing ‐ original draft, Writing ‐ review and editing; Erick Tjhin: Formal analysis, Investigation, Methodology, Validation, Visualization, Writing ‐ review and editing; Eliza V.C. Alves‐Ferreira: Investigation, Methodology, Writing ‐ review and editing; Stephanie Spada: Methodology; Fadila Bouamr: Methodology; Vinod Nair: Investigation, Methodology; Sundar Ganesan: Investigation, Methodology; Michael E. Grigg: Project administration, Resources, Supervision, Writing ‐ review and editing.

Conflict of Interest

The authors declare no conflict of interest.

Pszenny, V. , Tjhin, E. , Alves‐Ferreira, E. V. C. , Spada, S. , Bouamr, F. , Nair, V. , Ganesan, S. , & Grigg, M. E. (2022). Using the Sleeping Beauty (SB) transposon to generate stable cells producing enveloped virus‐like particles (eVLPs) pseudotyped with SARS‐CoV‐2 proteins for vaccination. Current Protocols, 2, e575. doi: 10.1002/cpz1.575

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.