Viral Mimetic Bacterial Outer Membrane Vesicles for Targeting Angiotensin-Converting Enzyme 2

Introduction

Extracellular vesicles (EVs) are spherical nanoparticles secreted by all biological systems, and in Gram-negative bacteria such as Escherichia coli (E. coli), they are specifically referred to as outer membrane vesicles (OMVs). They are formed by the bulging and pinching off of a portion of the outer membrane (OM) and consequently contain biologically active components.¹ Recently, OMVs have been exploited as diagnostic materials, immune boosters, vaccines, drug delivery, surrogate neutralisation tests, and intra- or inter-species communicators.^2–7

E. coli-based protein engineering technologies, widely used in many biotechnological industries, have evolved to enable the production of customized proteins. The OMVs mimic heterologous expression through this bacterial engineering, owing to the natural release mechanism of OMVs. The membranous environment of OMVs can be used to maintain the stability and exterior orientation of heterogeneous foreign antigens,⁸ so that they can be used as nanovesicle-theranostic materials. Therefore, robust target protein display on the surface of OMVs may have numerous applications in combating emerging infectious diseases.^9,10

For effective target protein display onto OMVs, we have incorporated surface display technology using antigen 43 (Ag43) of the autotransporter system of E. coli.^11,12 Ag43 is the simplest protein secretion-related polyprotein and consists of a signal peptide, a passenger domain, and a β-barrel domain to express the passenger domain onto the surface.¹³ Original passenger and β-barrel domain includes aa 138–1039 of Ag43, but it is known that the α-helical linker domain (α-HLD; aa 700–740) within the β-barrel induces surface exposure to the outside.^11,14,15 We previously published a structural analysis of a platform capable of effective surface display using the Ag43 system with eGFP.¹⁶ Through structural analysis, we identified that the 700–1039 amino acid sequence of Ag43 (hereafter referred to as Ag43β700) including α-HLD is the key amino acid sequence for surface exposure.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have recently emerged or re-emerged in many countries, and it still causes numerous infections.^17,18 The SARS-CoV-2 spike protein is the main key factor in the host receptor recognition for entry.¹⁹ They decomposed into S1 and S2 proteins by protease in the body. Generally, the S1-subunit contains the receptor-binding domain (RBD) responsible for the angiotensin-converting enzyme 2 (ACE2) binding, and the S2-subunit is responsible for membrane fusion.^20,21 Based on a previous study, we constructed viral mimetic Ag43β700_RBD OMVs that the exterior orientation of OMVs was completed by replacing the passenger domain with an RBD protein, connecting it with Ag43β700. In this study, the viral mimetic Ag43β700_RBD OMVs were demonstrated to function as versatile nanovesicles for use in a SARS-CoV-2 neutralization activity test platform and as therapeutic materials, such as vaccines, by displaying the virus spike protein on the outer membrane of the OMVs, thereby enabling effective binding to the ACE2 receptor.

Our study attempted to prove that viral mimetic Ag43β700_RBD OMVs could have provided nanovesicle-theranostic platforms: ACE2 blockade, OMVs-ELISA, intracellular uptakes, and neutralizing antibodies (nAbs) production, which can be employed for emerging infectious disease diagnosis and therapeutic applications.

Materials and Methods

Ethics Statement

All animal experiments in this study were approved by the institutional animal care and use committee of Chungbuk National University, Republic of Korea (approval no. CBNUA-1977-22-02). All animal experiments were conducted following the guidelines of the Institutional Animal Care and Use Committee (IACUC) of Chungbuk National University (approval number: CBNUA-1714-22-01) and adhered to the Guide for the Care and Use of Laboratory Animals (National Research Council, 8th edition). All experimental work, including the care and maintenance of the animals and animal housing, was carried out in a Biosafety Level Three (BSL3+) laboratory, under the recommendations and guidelines provided by the Ministry of Food and Drug Safety, Republic of Korea, and the Korea Disease Control and Prevention Agency (KCDA), Republic of Korea.

Construction of Plasmid Variants

Overall plasmid construction was based on by a previous study.¹⁶ The Ag43 gene was amplified from the genomic DNA of E. coli strain K-12. SARS-CoV-2 RNA (NCCP43326) was provided by the National Culture Collection for Pathogens (NCCP, Cheongju, Republic of Korea). SARS-CoV-2 RNA was reverse transcribed using the TOPscript™ cDNA Synthesis Kit (Enzynomics, Daejeon, Republic of Korea) to amplify the SARS-CoV-2 RBD (residues Asn331–Val524) gene. The pET28b vector was modified to express the RBD–Ag43β surface display constructs with a FLAG or 6×His-tag (GGGGS)₂ linker, and recognition sites for restriction enzymes using the primers listed in Supplementary Table S1. The Ag43β surface display constructs and strains are summarised in Table 1.

Table 1 The Plasmids and Bacterial Strains Used in the Study

Expression of Ag43-Fused Recombinant SARS-CoV-2 RBD

Bacterial colonies of E. coli strain BL21 (DE3) were transformed with the Ag43β surface display expression vector and inoculated into 5 mL Luria Bertani (LB) broth supplemented with kanamycin (50 µg/mL) and grown overnight at 37 °C with shaking at 200 rpm. The overnight cultures were inoculated into fresh LB broth supplemented with kanamycin (50 µg/mL) at a 1:100 dilution and grown to an OD₆₀₀ of approximately 0.6. The cultures were then induced to express the Ag43β surface display proteins using isopropyl β-D-1-thiogalactopyranoside (IPTG, final concentration of 0.5 mm) at 25 °C for 20 h with shaking at 200 rpm. ClearColi BL21 (DE3) (Lucigen, Wisconsin, USA) was incubated under the same conditions as BL21 (DE3), except at a dilution ratio of 1:40 and for 24 h.

Whole-Cell Trypsin Treatment and SDS-PAGE

Whole-cell trypsin treatment was performed on the E. coli strain cultures until an OD₆₀₀ of approximately 0.6 was reached, followed by centrifugation at 13,000 rpm at 4 °C for 5 min. The cell pellets were washed twice with 10 mm Tris-HCl, pH 8.0, and then resuspended in 950 µL 10 mm Tris-HCl and 50 µL trypsin (final concentration of 1 µg/mL; Sigma-Aldrich, Missouri, USA). After the incubation at 37 °C for 5 min, the cell pellets were collected by centrifugation and washed twice with 1 mL 10 mm Tris-HCl. The cell suspensions were mixed with 5× sample buffer (Biosesang, Yongin, Republic of Korea) and heated at 100 °C for 5 min. The samples were subsequently separated using SDS-PAGE at 100 V for 90 min. The relative band intensities were analysed using ImageJ software (National Institutes of Health, Maryland, USA).

Whole-Cell Trypsin Treatment and SDS-PAGE

OMVs from E. coli cells expressing Ag43-fused SARS-CoV-2 RBD were isolated as previously described,^22,23 with slight modifications. E. coli cells (OD₆₀₀ approximately 20) were collected by centrifugation at 5000 ×g at 4 °C for 10 min and resuspended in a final volume of 36 mL Tris-HCl (200 mm, pH 8.0). Next, 2.4 mL sucrose (1 M), 2.4 mL EDTA (10 mm), 2.4 mL lysozyme (10 mg/mL), and 76.8 mL distilled water were sequentially added to the cell resuspension and incubated for 10 min at room temperature. After incubation, 1.2 mL phenylmethylsulphonyl fluoride (PMSF, 100 mm) and 240 µL aprotinin (10 mg/mL) were added to the cell suspension. The suspension was then mixed with a 120 mL extraction buffer (50 mm Tris-HCl pH 8.0 with 2% v/v Triton™ X-100 and 10 mm MgCl₂) with DNase (Sigma-Aldrich) for 25 min on ice and centrifuged at 7000 ×g for 5 min. The pelleted debris was discarded, and the supernatants were transferred to new centrifuge tubes and centrifuged at 35,000 ×g for 15 min to isolate the OM fraction. The pelleted OM fraction was washed twice and filtered through 0.45-µm filters (GE Healthcare, Illinois, USA). The isolated OMVs were resuspended in 10 mm Tris-HCl (pH 8.0) or phosphate buffered saline (PBS).

Characterisation and Trypsin Treatment of OMVs

The isolated OMVs were observed using cryo-transmission electron microscopy (cryo-TEM) on a carbon-coated copper grid. The grid-adhered OMVs were cooled using liquid nitrogen and transferred to a Tecnai G2 F20 S-TWIN TMP microscope (FEI Company, Oregon, USA) for imaging. The size and concentration of OMVs were measured using qNano Gold (Izon, Christchurch, New Zealand) with a NP 200 nanopore. Trypsin treatment of the OMVs was performed in the same manner as the whole-cell treatments. The OMVs (protein concentration 0.5 mg/mL) were treated with various trypsin concentrations (0, 0.5, 2.5, and 5 µg/mL) at 37 °C for 10 min. The reaction was stopped by mixing with 5× SDS-PAGE sample buffer and heating at 100 °C for 5 min. The samples (20 µg) were separated using SDS-PAGE and then transferred to PVDF membranes for Western blotting. Anti-SARS-CoV-2 spike rabbit antibodies (Origene, Maryland, USA) were used to detect surface expression. The washed membranes were stained with WesternBright^® ECL (Advansta, California, USA), and the relative band intensities were analysed using ImageJ software.

Surface Plasmon Resonance (SPR)

The binding affinity of the ACE2 protein for the OMVs was analysed using SPR with a Biacore X-100 instrument (Cytiva, Uppsala, Sweden) at 25 °C. HBS-EP buffer (10 mm HEPES pH 7.4, 150 mm NaCl, 3 mm EDTA, and 0.005% polysorbate 20 v/v) was used as the running buffer at a flow rate of 10 μL/min. Flow cell 1 was used as a reference. To immobilise the OMVs to the CM5 sensor chip, the gold (Au) surface of the sensor chip was pre-treated with an HBS-EP buffer and activated with a 1:1 mixture of 0.05 M N-hydroxysuccinimide and 0.2 M N-ethyl-N’-(dimethylaminopropyl) carbodiimide by modifying the carboxymethyl groups of dextran. The ACE2 protein was diluted in 10 mm sodium acetate (pH 4.0) and then injected over the sensor surface to coat the surface of flow cell 2, followed by injection of 1 M ethanolamine hydrochloride (pH 8.5) to block the remaining active sites. After baseline stabilisation, different concentrations of ACE2 protein (0.1, 1, 5, 10, and 20 µM) were injected over flow cells 1 and 2 for 180 s at a flow rate of 10 μL/min and a dissociation time of 360 s. Following each experiment, the sensor chip was regenerated with 10 mm glycine (pH 2.5). Non-induced bacterial OMVs were used as controls. The SPR results were analysed using BIAevaluation software (Biacore^TM, Cytiva, Sweden) to determine the binding constants (K_D), as well as the association (k_on or ka) and dissociation (k_off or k_d) constants.

ACE2-Blockade LF-Strip Assay

To verify the affinity of Ag43β700_RBD OMVs for ACE2, an ACE2-blockade lateral flow (LF) assay was conducted. The ACE2-blockade LF strip was designed as previously described studies.²⁴ Briefly, the Au nanoparticles (AuNPs) were conjugated to rabbit immunoglobulin G (IgG) (control) or RBDs under pH 8.4. Nitrocellulose membranes were blocked with PBS and dried at 37 °C for 1 h. Human ACE2 protein (1 mg/mL) as the test (T) line, 1 mg/mL anti-IgG antibody as the reaction (R) line, and 1 mg/mL anti-rabbit IgG antibody as the control (C) line were manually spotted onto the nitrocellulose membrane. Conjugation pads were blocked with PBS containing 2% bovine serum albumin (BSA) and dried at 37 °C for 4 h. AuNP–RBD or IgG were applied to the conjugation pad and dried at 37 °C for 1 h. After applying Ag43β700_RBD OMVs of varying concentrations onto the strip, the band intensity was quantified using ImageJ.

Enzyme-Linked Immunosorbent Assay (ELISA)

Indirect ELISAs were conducted using anti-SARS-CoV-2 RBD IgG. Ninety-six-well maxibinding immunoplates (SPL Life Sciences, Pocheon, Republic of Korea) were coated with 50 µL recombinant SARS-CoV-2 spike protein RBD (Sigma-Aldrich) or Ag43β700_RBD OMVs dilutions ranging from 0.07 × 10⁶ to 4.00 × 10⁷ particles/mL in PBS overnight at 4 °C. The plates were then washed three times with 200 µL 0.05% PBS-Tween (PBS-T) and blocked for 2 h at 25 °C using blocking buffer [3% (w/v) BSA in PBS-T]. The plates were washed three times and incubated with anti-SARS-CoV-2 S rabbit IgG (diluted 1:1000 in blocking buffer) for 2 h at 25 °C. After washing, 50 µL secondary horseradish peroxidase-conjugated anti-rabbit IgG (diluted 1:10,000 in blocking buffer) was added to the wells and incubated for 2 h at 25 °C. The plates were then washed and 100 µL 3.3′,5,5′-tetramethylbenzidine (TMB) substrate solution (Thermo Scientific, USA) was added to each well and incubated for 30 min. The reaction was stopped with 100 µL 0.16 M sulphuric acid, and the absorbance of each well was measured at 450 nm using a microplate reader (SpectraMax M2e, Molecular Devices, California, USA).

Confocal Fluorescence Microscopy of OMV Uptake by Vero E6 Cells

Vero E6 cells (African green monkey kidney cells, ATCC^®, CRL-1586™) were cultured in Dulbecco’s Modified Eagle Medium (DMEM; Welgene, Gyeongsan, Republic of Korea) supplemented with 10% (v/v) foetal bovine serum (FBS; Gibco (USA) and 1% Antibiotic-Antimycotic (Gibco, New York, USA) at 37 °C in a 5% CO₂ incubator. Ag43β700_RBD OMVs (2.0 × 10¹² particles/mL) were mixed with 1% v/v DiI cell-labelling solution (Invitrogen, Massachusetts, USA) and incubated at 37 °C for 30 min. To remove the residual DiI, the labelled Ag43β700_RBD OMVs were centrifuged at 35,000 ×g for 30 min. The pelleted DiI-labelled Ag43β700_RBD OMVs were then resuspended in PBS. For the cell-uptake assays, Vero E6 cells were seeded into 12-well plates (4.8 × 10⁶ cells/well), covered with glass coverslips, and cultured overnight. Next, 50 µL Ag43β700_RBD OMVs (3.0 × 10¹² particles/mL) in fresh culture medium were added to each well and incubated for 1 h at 37 °C. The attached cells were washed three times with PBS and fixed with 4% w/v paraformaldehyde. The cells were then washed three times with PBS and mounted in mounting medium with DAPI (VectorLab, California, USA). Images were acquired using a confocal laser scanning microscope LSM 880 (Zeiss, Oberkochen, Germany).

Time-Dependent ACE2 Recognition by OMVs in Vero E6 Cells

Cells were seeded into a 6-well plate (3 × 10⁵ cells/well) and incubated for 24 h. OMVs (10 µg) were then added to the wells and incubated for 10, 30, 60, and 120 min. The cell culture media containing OMVs was then discarded, and the cells were washed three times with Dulbecco’s PBS. The collected protein (10 µg) was separated using 10% SDS-PAGE and then subjected to Western blotting with anti-ACE2 rabbit (Abcam) and anti-actin rabbit (Invitrogen) antibodies to detect ACE2 recognition by the OMVs.

Immune Cytokine Test in Raw 264.7 Cells

Cytokine assays were performed using Raw 264.7 cells (ATCC^®, TIB-71™) cultured under the same conditions and procedures as those described for Vero E6 cells. The cells were seeded into a 96-well plate (3 × 10⁴ cells/well) and incubated for 24 h. Different concentrations of OMVs (50, 5, 0.5, and 0.05 µg/mL) and experimental controls (100 ng/mL lipopolysaccharide (LPS) and 50 µg/mL control OMVs) were then added to each well and incubated for 24 h. The cell culture media were then collected and centrifuged at 1500 ×g at 4 °C for 10 min to remove debris. Cytokines were then detected using cytokine multiplex ELISA (Arigo, Taiwan) according to the manufacturer’s protocol.

Mouse Immunisation and ELISA for Viral nAbs

Eight-week-old female B6.Cg-Tg (K18-ACE2) 2Prlmn/J mice (strain no. 034860) were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and housed in separate groups prior to infection. For the neutralisation test, the test group of female C57/Bl6 mice (n = 5) was immunised intranasally with Ag43β700_RBD OMVs (8.0 × 10⁸ particles/mouse) four times (days 0, 1, 2, and 2.5). The control group was immunised intranasally with non-induced control OMVs. The mice were observed for 6 days, and changes in body weight were recorded as an indication of morbidity. Blood samples were obtained using the retro-orbital plexus puncture method, and the collected serum was stored at −20 °C. To detect the presence of nAbs, ELISAs (Thermo Scientific) were performed according to the manufacturer’s instructions. The percentage of neutralisation was calculated as:

where sample abs represents the absorbance of the tested sample and negative control abs represents the absorbance of the negative control provided with the kit at 450 nm.

Virus Infection in K18-hACE2 Mice and Treatment Regimen

The β-CoV/Korea/KCDC03/2020 (NCCP43326) virus was provided by the NCCP. Molnupiravir (MPV) was obtained from MedChemExpress (Monmouth Junction, NJ, USA). The MPV stock solution was prepared with a vehicle containing 1% dimethyl sulphoxide and 20% sulfobutylether-β-cyclodextrin (MedChemExpress) in 0.9% NaCl and stored at –20 °C. The experimental mice were divided into four groups, with 10 mice in each group: group 1, infection-only; group 2, MPV (20 mg/kg); group 3, control OMVs; and group 4, Ag43β700_RBD OMVs. For infection, mice were anesthetised by isoflurane inhalation and intranasally inoculated with 50 µL 5 mouse lethal dose (MLD₅₀) β-CoV/Korea/KCDC03/2020 on day 0. On the same day, MPV was administered orally. For the survival test, OMVs (control and Ag43β700_RBD OMVs) were administered intranasally 1 h before infection and 1, 24, and 48 h post-infection. After infection and treatment, the mice were carefully monitored for body weight changes and survival from days 0 to 14. The mice were sacrificed when they lost ≥25% of their body weight before infection (as measured 1 day pre-infection). The lung and brain tissues from SARS-CoV-2-infected mice were collected at 5 days post infection (DPI) (n = 5/group) to determine the viral titres and perform histopathological and immunohistochemical analyses.

Viral Titration and Histopathological and Immunohistochemical Analyses

Lung and brain tissues from the SARS-CoV-2-infected mice were collected, homogenised by bead disruption, and centrifuged at 12,000 rpm for 5 min using TissueLyser II (Qiagen, Hilden, Germany). Ten-fold serially diluted virus-containing samples were then used to infect Vero E6 cells seeded in a 96-well plate at 2 × 10⁶ per well and cultured in a 5% CO₂ incubator at 37 °C for 4 days or until an apparent cytopathic effect was detected and visualised with 0.2% Crystal Violet staining. Viral titre was calculated and quantified according to the Reed and Muench method and expressed as log₁₀ 50% tissue culture infection dose (TCID₅₀) per mg of tissue. For the histopathological and immunohistochemical analyses, the tissues collected at 5 DPI were fixed in 10% formalin neutralisation buffer (Sigma-Aldrich) and processed for hematoxylin and eosin (H&E) and immunohistochemical staining. A rabbit anti-SARS-CoV-2 nucleocapsid polyclonal antibody was used as the primary antibody (Sino Biologicals, China) and further processed using a Ventana DISCOVERY ULTRA (Roche, USA) system. All slides were visualised using a Pannoramic^® 250 Flash III slide scanner and evaluated using CaseViewer software (3DHISTECH Ltd., Budapest, Hungary).

Statistical Analysis

All statistical analyses were performed using the GraphPad Prism 9 software (GraphPad, San Diego, CA, USA). Statistically significant differences between more than two groups were determined using one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparisons test. The Kaplan–Meier method was used for survival analyses, and differences were calculated using the log-rank Mantel–Cox test with multiple comparisons.

Results

Development of OMVs with Exterior Display of Viral Spike Protein RBD

We used Ag43β autotransporter derivatives including α-HLD to generate OMVs displaying the SARS-CoV-2 spike protein RBD (Figure 1A). We analysed the morphology of the OMVs and confirmed successful exterior display of the SARS-CoV-2 spike protein RBD through trypsin treatment (Figure 1B and C) and validated their antigenicity using a SARS-CoV-2 Antigen Self-Test Kit. Representative lateral flow (LF) strip assays were used to evaluate the band intensity level for each concentration of Ag43β700_RBD-OMVs, and the positive interaction between Ag43β700_RBD-OMVs and anti-RBD antibodies was clearly visible in the T line at concentrations above 10⁷ particles/mL (Figure 1D). These results indicate RBD protein was properly exposed to the outside of OMVs.

Figure 1 Construction of Ag43β700_RBD OMVs and working confirmation using commercially available COVID-19 Ag strip assay kit. (A) Schematic representation of the composition of E. coli Ag43β700_RBD (SARS-CoV spike protein RBD, amino acid residues 331–524). The SP, RBD, His-tag, glycine linker, central α-HLD, and β-barrel domain are indicated in yellow, canary, violet, grey, red, and deep grey, respectively. The insert indicates the predicted three-dimensional structure of Ag43β700_RBD on the exterior of the OMVs. (B) Size distribution of the Ag43β700_RBD OMVs as determined by qNano. Cryo-transmission electron microscopy (TEM) analysis was used to visually analyse the Ag43β700_ RBD OMV morphology and verify the particle size. Scale bar = 200 nm. (C) Extracellular trypsin treatment and Western blot analysis of Ag43β700_RBD OMVs. Significantly reduced RBD signals are seen in the Ag43β700_RBD OMVs. (D) Representative lateral flow (LF)-strip assay at different Ag43β700_RBD OMV concentrations. Weakly positive red test lines are visible at the “T” location of the LF-strip assay at concentrations above 7.4 × 107 Ag43β700_RBD OMVs. Purified RBDs induced a clearly visible test line. No test line is visible on the PBS/LB-strip assay. Control lines are visible at the “C” location of all the LF-strip assays.

ACE2-Blockade LF-Strip Assay and Confirmation of Serodiagnostic Antigens

We tested the ability of ACE2-blockade in providing competitive advantages to early-stage virus infection with natural ACE2. Efficient display of SARS-CoV-2 RBD antigens (amino acid residues 331–524) at the surface of E. coli OMVs was achieved using the Ag43β700 display platform. An additional binding assay was performed using SPR to assess whether direct binding of potential therapeutics to Ag43β700_RBD OMVs can effectively block viral entry at the first infection step (Supplementary Figure S1). Various hACE2 concentrations ranging from 0.1 to 20 µM were passed through the chip surface, and Au signal changes at different OMV concentrations were recorded. hACE2 showed high binding affinity for immobilised Ag43β700_RBD OMVs with a K_D of 4.82 × 10^–7 M. This finding corresponds to previous studies, which report the binding affinity of Ag43β700_RBD OMVs for hACE2 as comparable to that of the native viral RBD protein for hACE2.^25–27 The control OMVs exhibited a low binding affinity (K_D = 1.01 × 10^–3 M). These results suggest that the displayed RBD retains its native structure, which is an important requirement for successful antigen presentation on the exterior surface of OMVs.

Subsequent analytical evaluation of hACE2 blockade using Lateral Flow (LF)-hACE2 strip assays indicated that the interaction between the RBD and hACE2 is immediately blocked by the presence of the Ag43β700_RBD OMVs (Figure 2A). The band intensity at the T line disappeared as the Ag43β700_RBD OMV concentration increased (Figure 2B). Regression analysis based on the T/C ratio confirmed that the Ag43β700_RBD OMVs inhibit the binding of the RBD to hACE2. The Ag43β700_RBD OMVs described in this study may therefore be useful for blocking viral entry into ACE2-expressing cells, thereby preventing or treating COVID-19.

Figure 2 Neutralising ability of Ag43β700_RBD OMVs and ELISA to confirm binding activity between Ag43β700_RBD OMVs and anti-SARS-CoV-2 spike RBD antibodies. (A) Illustration of the LF-hACE2 strip assay. Addition of phosphate buffered saline (PBS) does not block binding of the gold nanoparticle (AuNP)–RBD-conjugates to hACE2 and the RBD–ACE2 complex subsequently creates a visible test (T) line. Addition of Ag43β700_RBD OMVs can block binding of the AuNP–RBD-conjugates to hACE2, thereby decreasing the line density at the T line. The control (C) and reaction (R) lines on the LF-hACE2 strip indicate that the test ran properly. (B) Photos of the LF-hACE2 strip assays indicating the neutralising ability of Ag43β700_RBD OMVs using different concentrations of AuNP–RBD OMVs (upper). Signals of strips exposed to different Ag43β700_RBD OMVs concentrations were analyzed using ImageJ software (lower). (C) Illustration of Ag43β700_RBD OMV-based indirect ELISA platform to identify anti-SARS-CoV-2 spike RBD antibodies. Ag43β700_RBD OMVs are bound to the surface of 96-well plates. (D) Ag43β700_RBD OMV concentration–response curves for binding anti-RBD antibodies in the OMV-ELISA.

In addition to the ACE2-blockade assays, we developed Ag43β700_RBD OMV-based ELISAs as a rapid and reliable tool for analysing the seroprevalence of SARS-CoV-2-specific antibodies (Figure 2C). Presentation of RBD proteins on the exterior surface of OMVs offers potential benefits over soluble RBD proteins, including unidirectional and spatially defined protein orientation, sufficient quantity, and a pseudoviral membrane environment. As shown in Figure 2D, Ag43β700_RBD OMVs were used as serodiagnostic antigens, and their performance was compared with that of recombinant viral RBD (Supplementary Figure S2). High sensitivity was exhibited with a limit of detection of 5.42 × 10⁶ particles/mL (Figure 2D). Both recombinant RBD- and OMV-based ELISAs exhibited similar binding signals across a range of antigen dilutions (RBD, 0–30.0 ng/mL; Ag43β700_RBD OMVs, 0.07–40.0 × 10⁶ particles/mL). The results indicate that the Ag43β700_RBD OMV-based ELISA can determine the antibody titre in vaccinated individuals.

Intracellular Uptake of Ag43β700_RBD OMVs in vitro

We confirmed that the Ag43β700_RBD OMVs present the RBD protein on their surface, and that Ag43β700_RBD OMVs bind to ACE2, thereby providing a basis for the potential application of Ag43β700_RBD OMVs as a target-specific drug delivery vehicle for antiviral therapeutics.^28,29 To confirm in vitro ACE2 binding of Ag43β700_RBD OMVs, DiI-labelled Ag43β700_RBD OMVs were incubated with Vero E6 cells, which are ACE2 receptor-positive cells (Figure 3A).

Figure 3 Intracellular uptake of Ag43β700_RBD OMVs in vitro. (A) Schematic illustration explain Ag43β700_RBD OMV uptake into host cells is mediated by the interaction between the viral RBD protein and the hACE receptor. Confocal fluorescence microscopy images of the binding of Vero E6 cells with DiI-labelled Ag43β700_RBD (bottom) and control (top) OMVs. Scale bar: 20 μm. (B) Relative ACE2 expression levels in Vero E6 cells incubated with Ag43β700_RBD OMVs for different binding times (10, 30, 60, and 120 min) determined by Western blot analysis using an anti-ACE2 antibody (see Supplementary Figure S3). The red-dotted line represents value 1 of the non-treated (NT) group. (C) ELISA for immune cell cytokines (TNFα, IL-6, and IL-1β). Statistical differences were analysed using an independent sample one sample t-test. **p = 0.0075; ***p = 0.0004; *p = 0.0174; ****p < 0.001; ***p = 0.0001; **p = 0.0049 (TNFα)], [N.D; **p = 0.0012; ***p = 0.0003; ***p = 0.0006; *p = 0.0127; NS; N.D (IL-6)], and [N.D; **p = 0.0044; ***p = 0.0006; ***p = 0.0003; ***p = 0.0003; N.D; N.D (IL-1β)].

To investigate whether ACE2-Ag43β700_RBD OMVs are involved in the regulation of ACE2 expression, we used Western blotting to examine ACE2 expression in the presence of Ag43β700_RBD OMVs. Our results indicated that ACE2 expression is upregulated following Ag43β700_RBD OMV treatment (Figures 3B and S3), which suggests that Ag43β700_RBD OMVs may regulate ACE2 at the protein level. Similar observations of ACE2 upregulation mediated by immune signalling and oxidative stress pathways have been reported in SARS-CoV-2-infected epithelial cells.^30–32 Ag43β700_RBD OMVs may therefore lead to ACE2 glycosylation and activation of downstream signal transduction,^33,34 but the characteristics of Ag43β700_RBD OMVs that induce ACE2 glycosylation require further investigation.

With respect to safety, the OMVs generated from Gram-negative bacteria may cause pyrogenicity. We subsequently assessed the levels of crucial representative pro-inflammatory cytokines in SARS-CoV-2 infection, including tumour necrosis factor alpha (TNFα), interleukin (IL)-6, and IL-1β, and found elevated pro-inflammatory cytokine expression levels following Ag43β700_RBD OMV treatment. As shown in Figure 3C, 100 ng/mL LPS caused a significant increase in Raw 264.7 monocytes/macrophages, but this increase was comparable to the increase seen in cells treated with a relatively high OMV concentration (50 µg/mL, 10⁹ particles/mL). Moreover, the same OMV concentration did not induce cell death (data not shown), which may be attributed to Ag43β700_RBD OMVs from the LPS-mutant E. coli failing to exhibit specific responses with lethal toxic components, such as LPS.³⁵ However, it still remains unclear why IL-6 and IL-1β were rarely induced, while TNFα was highly induced.

Neutralisation Assays of Ag43β700_RBD OMVs in vivo

To further examine whether Ag43β700_RBD OMVs induce nAbs, we examined antibody titres and neutralising activity. Mice immunised with Ag43β700_RBD OMVs via intranasal inoculation at 8.0 × 10⁸ particles/mouse on days 0, 1, 2, and 2.5 exhibited no significant decrease in body weight during 6 days of monitoring (Figure 4A). Serum samples were collected from the mice to determine the nAb levels, and the percentage neutralisation was assessed using ELISA. A cut-off of 20% was used to distinguish positive (+nAb) and negative (–nAb) samples. We found that intranasal Ag43β700_RBD OMV administration induced high serum nAb levels with 39–58% neutralisation (Figure 4B), which suggests that the use of Ag43β700_RBD OMVs as an adjuvant induces a more comprehensive immune response. Several studies have suggested that intranasal vaccination may contribute to the production of nAbs against a variety of respiratory viruses, including SARS-CoV-2.^36,37 However, there are few reports to demonstrate the necessity of adjuvants for inducing intranasal immune efficacy.^38,39 We demonstrated intranasal immunisation with Ag43β700_RBD OMVs induce nAbs, indicating that Ag43β700_RBD OMVs alone may induce both antibody- and robust cell-mediated immunity. Ag43β700_RBD OMVs may therefore have in vivo applicability in vaccine strategies to boost SARS-CoV-2 nAbs.

Figure 4 Neutralisation assays in vivo. (A) Immunisation and blood-draw schedule for C57/Bl6 mice and body weight changes (%). (B) Serum ELISA against SARS-CoV-2 RBD. Ag43β700_RBD OMVs improved the neutralising antibody response. The presence of neutralising sera disrupted the ACE2–RBD interaction, thus decreasing the measured signal. For commercial kits, sera presenting neutralization >20% were considered positive for the presence of neutralizing antibodies, according to the manufacturer’s protocol.

Intranasal Immunisation with Ag43β700_RBD OMVs

The complexity of SARS-CoV-2 pathogenesis necessitates multidisciplinary approaches for combating SARS-CoV-2 infection. Currently, there are numerous different therapeutic approaches available, such as cell-based drug delivery therapies, monoclonal antibody/plasma-mediated immunotherapies, and vaccine therapeutics. Recent studies have indicated that intranasal vaccination effectively induces mucosal immunity and prevents mucosal pathogen invasion.⁴⁰ We evaluated the efficacy of Ag43β700_RBD OMVs for intranasal immunisation against SARS-CoV-2 (Figure 5A) and found that infected mice lost weight from 4 to 9 DPI (Figure 5B). Mice immunised with Ag43β700_RBD OMVs showed less than 5% weight loss 1 day after administration and then exhibited significant weight loss protection at 9 DPI. It shows a nearly similar recovery rate compared to MPV (positive control), which has been approved by the FDA as a COVID-19 treatment.⁴¹ Whereas there was no significant protection from weight loss in the infection-only and control OMV-treated mice (Figure 5B). Furthermore, mice immunised with Ag43β700_RBD OMVs began to recover their body weight from 8 DPI, whereas mice treated with MPV only began to recover their body weight from 9 DPI. Ag43β700_RBD OMV-immunised mice therefore exhibited greater recovery than MPV-treated mice. Mice immunised with Ag43β700_RBD OMVs exhibited 43% survival (Figure 5C). All tested mice were sacrificed at 5 DPI and, as SARS-CoV-2 mainly replicates in the lung and brain of K18-hACE2 mice,^42,43 these tissues were harvested for viral titre analyses. Compared with the viral titres in the lung tissues of the infection-only group, the viral titres in the lung tissues of both the MPV-treated and the Ag43β700_RBD OMV-immunised groups were reduced to a similar level (Figure 5D). In contrast to the lung tissues, the brain tissue of the Ag43β700_RBD OMV-immunised group showed a 10-fold greater decrease in viral titre than that of the MPV-treated group (Figure 5D). As brain-related abnormalities, such as pathological changes and brain fog, due to SARS-CoV-2 infection are continuously being reported,^44,45 Ag43β700_RBD OMVs may reduce the pathological effects of SARS-CoV-2 in the brain.

Figure 5 Protection of hACE2 mice against SARS-CoV-2 challenge following intranasal Ag43β700_RBD OMV administration. (A) Schematic representation of the immunisation schedule. K18-hACE2 mice (n = 5/group) were intranasally immunised 1 day prior to infection with SARS-CoV-2. Three booster doses were then administered at 1, 24, and 48 h post infection. (B) Weight change (%). (C) Survival percentage (%). (D) Analysis of the viral titres in the lung and brain tissues. Immunisation with Ag43β700_RBD OMVs decreased the viral titre in the SARS-CoV-2-infected mice. (E) Histological evaluation of lung and brain sections from K18-hACE2 mice following intranasal infection with 5 MLD50 SARS-CoV-2 5 days post-infection.

To evaluate the antiviral efficacy of Ag43β700_RBD OMV-immunisation, we performed H&E staining and immunohistochemistry on the lung and brain tissues harvested at 5 DPI. Similar to naïve tissue, the lung and brain tissues obtained from Ag43β700_RBD OMV-immunised mice exhibited less severe lesions and reduced inflammation (Figure 5E). Furthermore, immunohistochemical staining for SARS-CoV-2 revealed reduced staining in the lung and brain tissues of Ag43β700_RBD OMV-immunised mice (Figure 5E). These results suggest that intranasal Ag43β700_RBD OMV immunisation has the potential to prevent viral infection.

Discussion

Delivery platform technology for vaccine design to treat infectious diseases has become a crucial tool for enhancing the stability of biomolecules, depending on the lipid composition and ratio. Various delivery systems including lipid nanoparticles, liposomes, and EVs are ongoing to improve the delivery efficacy of biomolecules and minimize potential side effects.⁴⁶ Among these, OMVs have also been used as a notable material for vaccine delivery platforms.^47,48 Intrinsic properties of OMVs that encapsulate many biomolecules (eg RNA, protein, metabolite) have been well demonstrated. Moreover, they are natural materials, and side effects are significantly lower than synthetic particles. Naturally released OMVs (nOMVs) contain various immunological substances including pathogen-associated molecular patterns (PAMPs), and they are highly attractive as vaccine materials due to their ability to induce immediate innate immune responses.⁹ The representative OMV vaccine, BEXSERO, is currently used for Neisseria meningitidis (N. meningitidis) serogroup B in many countries, and various pharmaceutical companies have been developing new vaccines by leveraging the advantages of OMVs.¹⁰

Spontaneously released OMVs have low yields, hindering their use for industrial purposes. To overcome this, the development of generalized modules for membrane antigen (GMMA) or mimetic OMVs through chemical treatment methods is commonly employed. In this regard, detergent-based OMVs (dOMVs) production achieves over 10⁷ times greater efficiency compared to nOMVs,¹⁶ and it also reduces OMV endotoxicity.

Unlike bacteria, viruses evolve through rapid mutations, leading to the emergence of various variants. To respond to these variants, fast and accurate vaccine development technologies are needed. The most straightforward approach is to develop a modified vaccine tailored to the variant’s mutant site like an mRNA vaccine.⁴⁹ Our platform is designed with a restriction enzyme site positioned between the signal peptide and the β-barrel sequence. By simply inserting the mutated sequence, the system can be effortlessly reconfigured into a target-specific construct, streamlining the display process. This design facilitates rapid and precise responses to diverse variants and can also be applied to other viral species, such as Dengue, Zika, and Influenza.

In addition to the various advantages mentioned earlier for vaccine applications, it can also be used as a diagnostic tool particularly for neutralizing antibodies (nAbs) diagnosis. Following viral infection or vaccination, the body produces nAbs, which target a structural component of the virus to mitigate or block viral infectivity.⁵⁰ Effective defence against viral infection requires enough nAbs. Detection of nAbs allows evaluation of vaccine efficacy,⁵¹ serological diagnosis,⁵² and epidemiological investigation.⁵³ Currently, the plaque-reduction neutralisation test (PRNT) is the most widely used virus neutralisation test (VNT) to measure nAbs. However, PRNTs are complex and require large serum samples and long culture times for plaque formation.⁵⁴ Various other methods, such as pseudovirus neutralisation tests (pVNTs) and surrogate virus neutralisation tests (sVNTs), are also used.^55,56 In our data, Figure 2C and D demonstrate that Ag43β700_RBD OMVs can be applied as a diagnostic material for evaluating neutralizing antibody activity.

Similar studies using OMVs against SARS-CoV-2 are summarized in Table 2. OMVs extracted from various bacteria have been explored as vaccine candidates against SARS-CoV-2 through multiple approaches, including i) inactivated viruses with OMVs, ii) recombinant RBD proteins with OMVs, and iii) one-step synthesis recombinant RBD-OMVs. However, our research is a multi-platform capable of both diagnosis and treatment for COVID-19. The findings suggest that it can serve as an antigen, a platform for neutralizing antibody diagnostics, and an intranasal vaccine material.

Table 2 Comparison with Ag43β700_RBD and Other SARS-CoV-2 Approach Strategy

Although this study provides theranostic observations, future directions remain. This study was conducted using the small RBD protein (aa 331–524 in Figure 1A), indicating the need for further investigation to apply larger proteins. Additionally, exploring the OMV loading of foreign targets (eg drugs, antigens, mRNA, etc.) is needed to enhance the multivalent approach. Lastly, to understand whether our viral mimetic OMVs can also neutralize other coronaviruses, additional in vitro and in vivo studies will be needed to determine the binding specificity and affinity of circulating variants of concern (VOCs).¹⁸

This study represents the first case of Ag43β700_RBD OMVs in theranostic approaches against SARS-CoV-2, which may have great potential for synergistic antiviral performance as diagnostic and intranasal vaccine candidates.

Conclusion

Mitigation of virus infections requires rapid, accurate diagnostic methods, as well as therapeutics for disrupting spike protein–host interactions. This study demonstrated the potential of Ag43β700 OMVs in a theranostic approach for SARS-CoV-2 diagnosis and therapy. We demonstrated the diagnostic performance of Ag43β700_RBD OMVs. Furthermore, we found that intranasal immunisation with Ag43β700_RBD OMVs provides strong protection against SARS-CoV-2 by inducing a robust immune response, as well as a detectable level of neutralising antibody against the SARS-CoV-2 RBD. In this study, we focused on the SARS-CoV-2, further investigations are required to cross-validation with other variants or viruses. In conclusion, Ag43β700_RBD OMVs may be applied as a multifaceted theranostic platform for further emerging infectious diseases.