Development of an mRNA-based vaccine using virus-like particles for PEDV

HEK-293T cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 15% (vol/vol) heat-inactivated fetal bovine serum (FBS) at 37°C in a 5% CO₂ atmosphere. The Vero cell line (American Type Culture Collection CCL-81) was used for PEDV propagation and cultured in DMEM supplemented with 10% (vol/vol) heat-inactivated FBS at 37°C in 5% CO₂. PEDV strains CH-HK-2021 (GenBank no. PP785988↗, G2c subtype), AH2012/12 (GenBank no. KU646831↗, G2b subtype), JS2008 (GenBank no. KC210146↗, G1 subtype), and JS-NJ2023 (GenBank no. PP580804↗, S-INDEL subtype) (Fig. S1) were isolated, maintained in our laboratory, and cultured in Vero cell culture medium supplemented with 5 µg/mL trypsin and 37.5 µg/mL pancreatin.

Based on the structural protein sequence of the PEDV variant CH-HK-2021, codon-optimized open reading frames were chemically synthesized. To enable simultaneous expression of all four structural proteins, one recombinant plasmid was constructed to carry the full-length S gene, whereas a second plasmid carried a single polycistronic cassette encoding M, N, and E proteins linked in frame by flexible glycine-serine-glycine (GSG) spacers together with a self-cleaving P2A peptide (21, 22). Both recombinant plasmids included coding sequences for the type I cap structure (N7MGPPPAM) and 5′ and 3′ untranslated regions (UTRs). S- and M+N + E-encoding mRNA were synthesized via in vitro transcription using T7 RNA polymerase and the linearized plasmid DNA templates (containing the optimal codon for the structural proteins). An equimolar mixture of S-mRNA and M+N + E-mRNA (1:1) was prepared and encapsulated into lipid nanoparticles using a microfluidic device (MaiAna I Nano L) by mixing the lipid-ethanol phase with the aqueous mRNA phase at a volumetric ratio of 1:3, yielding the PEDV VLP mRNA vaccine (Fig. 1A). The lipid formulation of the mRNA vaccine comprised SM102, 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), cholesterol (CHOL), and DMG-PEG2000, with the detailed molar ratio of the lipid components as follows: SM102:DSPC:CHOL:DMG-PEG2000 = 50:10:38.5:1.5. The particle size, polydispersity index (PDI), and zeta potential of the prepared formulation were measured as previously described (16).

The CH-HK-2021 strain was inactivated with β-propiolactone and mixed with Gel-02 adjuvant at a 4:1 (vol/vol) ratio to prepare the inactivated vaccine. One-month-old PEDV antibody-negative piglets were selected, and each piglet received 2 mL of the corresponding inactivated vaccine intramuscularly. Booster immunizations were given at 14 and 28 days post-primary vaccination (dpv). Blood samples were collected at 42 dpv, and serum was isolated for western blotting (WB) analysis.

PEDV VLP mRNA were directly added to HEK-293T cells cultured in T75 flasks without any transfection reagent. Cell supernatants were collected after 48 h of incubation; PEDV VLPs were collected after ultrasonic lysis, followed by low-speed centrifugation to remove cell debris and filtration through a 0.45 μm filter (Port Washington, New York, USA). Subsequently, the supernatants were centrifuged at 30,000 rpm and 4°C for 2 h to collect the pellets. After purification via a 30%–40%–50% discontinuous sucrose gradient, the PEDV VLP bands at the 30%–40% sucrose interface were collected, resuspended in PBS, and centrifuged again at 30,000 rpm and 4°C for 2 h to remove sucrose. Finally, the resulting pellets were resuspended in PBS and stored at −80°C.

Following the established experimental protocol (23), further processing was performed on the purified VLPs. For negative-stain transmission electron microscopy (TEM), 10 µL of purified PEDV VLPs was adsorbed onto carbon-coated 400-mesh copper grids for 5 min at ambient temperature. Excess sample was blotted away with Whatman No. 1 filter paper, and the grids were immediately stained with 1% (wt/vol) sodium phosphotungstate (pH 7.0) for 5 min. After a second blotting step to remove residual stain, the grids were air-dried and visualized on a JEM-1400 transmission electron microscope operating at 120 kV.

PEDV VLP mRNA were directly added to HEK-293T cells cultured in T25 flasks without any transfection reagent, and the cell supernatants were collected after 24 h of incubation. Proteins were resolved by SDS-PAGE and electrotransferred onto 0.22 µm PVDF membranes (Millipore, Bedford, MA, USA). Membranes were blocked with 5% (wt/vol) skim milk in PBST (PBS supplemented with 0.05% Tween 20, pH 7.4) for 1 h at 37°C and then incubated for 2 h at 37°C with the following primary antibodies: porcine hyperimmune serum against PEDV (1:250, used for detection of M and E proteins); mouse anti-PEDV S monoclonal antibody (1:5,000); and mouse anti-PEDV N monoclonal antibody (1:5,000) all prepared in our laboratory. After three 15-min washes with PBST, membranes were probed with HRP-conjugated secondary antibody (1:10000) for 1 h at 37°C. Following three additional PBST washes, immunoreactive bands were visualized using an enhanced chemiluminescence system (Tanon 5200, Shanghai, China) and digitally captured.

The immunization schedule is shown in Fig. 2A. Twenty 6-week-old female BALB/c mice were randomly divided into four groups (n = 5). According to previous studies, 100 µL of commercial inactivated vaccine and 10 µg of mRNA-LNP vaccine were sufficient to elicit effective immune responses in mice (16, 23, 24). So, in this study, the experimental groups received an intramuscular injection of 10 µg of mRNA-LNP vaccine or 100 µL of commercial inactivated vaccine on days 0 and 14, while the control group received an equal volume of sterile PBS (100 µL). All mice were euthanized 14 days after the last immunization (i.e., on day 28). Orbital blood was collected on days 14 and 28, and serum anti-S, E, M, and N-specific IgG levels were determined by ELISA (enzyme-linked immunosorbent assay), neutralizing antibody titers were measured by neutralization assay (25). Spleens were aseptically removed to prepare single-cell suspensions, and the percentages of CD3⁺CD4⁺, CD3⁺CD8⁺ T cells, and B220⁺CD19⁺ B cells were analyzed by flow cytometry. Splenocytes were restimulated with PEDV virions for 72 h, after which lymphocyte proliferation was assessed, the culture supernatants were collected and IFN-γ and IL-4 concentrations were quantified using a mouse IFN-γ and IL-4 ELISA kit (Meimian, China) according to the manufacturer’s instructions.

Lymphocyte proliferation assays were performed by using splenocytes as described before (24). On day 28 dpv, splenic lymphocytes were aseptically isolated, transferred to 1.5 mL centrifuge tubes (1 × 10⁶ cells/tube), washed once with PBS, and centrifuged at 1,500 rpm for 5 min. The cell pellets were resuspended in 300 µL PBS containing the following fluorochrome-conjugated antibodies: PE anti-mouse CD4 (BioLegend, USA), APC anti-mouse CD8a (BioLegend, USA), FITC anti-mouse CD3e (BD Biosciences, USA), PerCP-Cy5.5 anti-mouse B220 (BD Biosciences, USA), and FITC anti-mouse CD19 (BioLegend, USA). After incubation at 4°C for 30 min in the dark, cells were centrifuged at 1,500 rpm for 5 min; the supernatants were discarded, and the pellets were rinsed twice with PBS. Cells were finally resuspended in 500 µL PBS and analyzed on an Accuri C6 Plus flow cytometer (BD Biosciences); 1 × 10⁵ events were acquired to enumerate CD4⁺ and CD8⁺ T cells. Data were processed with FlowJo v10.7.1 software.

In the passive immunization and challenge trial, six PEDV-uninfected pregnant sows were selected and intramuscularly injected with either PEDV VLP mRNA vaccine (100 µg per sow) or an equivalent volume of PBS at 1 month before farrowing; the vaccinated sows received a booster dose of the same amount 14 days later. On the day of farrowing, serum and colostrum samples were collected from each sow for antibody testing. After 5 days of lactation, five neonatal piglets from each sow were randomly chosen and orally challenged with the PEDV CH-HK-2021 strain at 2 × 10⁴ TCID₅₀ per piglet. Rectal swabs and Fecal consistency scores were collected daily post-challenge. At 13 days post-challenge (dpc), the piglets were euthanized and necropsied; small-intestinal tissues and contents were collected for histopathological examination and PEDV RNA load quantification, respectively.

ELISA was performed to quantify IgG and IgA antibodies specific for the PEDV-S, E, M, and N. As described before (26), recombinant eukaryotic PEDV-S, E, and M proteins and prokaryotically expressed N proteins (100 ng/well) were coated in wells of a 96-well plate, suspended in bicarbonate buffer (100 mM, pH 9.6), and incubated at 4°C overnight. The coating solution was discarded, washed three times with PBST, and incubated with 3% BSA for 2 h at 37°C. The addition of serum or colostrum collected from mice or sows to be tested (1:50), after incubation for 30 min at 37°C, the wells were washed three times with PBST. Then, 100 µL of goat anti-pig/mouse IgG or goat anti-pig/mouse IgA [both HRP conjugates] was added to each well and incubated at 37°C for 30 min. Then, 100 µL of TMB substrate solution was added to each well and incubated at 37°C for 15 min. Finally, the reaction was stopped by adding 50 µL of H₂SO₄ (2 M) to each well, and the absorbance at 450 nm (OD₄₅₀) was measured.

As described before (16), porcine and mouse serum were heat-inactivated at 56°C for 30 min, whereas intestinal tissues were directly homogenized and centrifuged to collect the supernatants. They were then serially twofold diluted, mixed with an equal volume of PEDV (200 TCID₅₀/100 µL), and incubated at 37°C for 1 h. Monolayers of Vero cells in a 96-well tissue culture plate were then inoculated with the mixture. After incubation at 37°C for 2.5 h, the mixture was discarded, and the wells were washed three times with DMEM. DMEM trypsin (10 µg/mL), which is necessary for the virus to enter the cells, was then added to each well, and the cells were incubated at 37°C for 3–5 days.

Following the viral challenge, the piglets were monitored daily for clinical signs of diarrhea. Diarrhea was assessed by scoring fecal consistency. Fecal consistency was scored on a 0–3 scale: 0, well-formed, normal feces; 1, soft and pasty but still shaped; 2, semi-liquid diarrhea with scant solid residue; 3, watery diarrhea without any formed matter. Piglets with a score ≥2 were defined as diarrheic (27). At necropsy, tissues from the jejunum and ileum were collected and fixed in formalin for histological examination. The dehydrated tissues were treated with xylene, embedded in paraffin wax, sliced, and mounted on slides. The slides were subjected to hematoxylin and eosin (H&E) staining and an immunohistochemistry (IHC) assay. PEDV antigen was detected in fixed tissue sections by IHC using our laboratory-produced PEDV-specific monoclonal antibody (1:200 dilution) according to the protocol described before (28). Additionally, intestinal tissue samples were collected from piglets for IgA quantification.

Viral RNA copies in anal swabs and intestinal tissues of piglets were quantified by TaqMan real-time RT-qPCR (29). Total RNA was extracted from 10% (wt/vol) suspensions of swabs or homogenized intestinal mucosa, reverse-transcribed using a commercial kit (Vazyme, China), and amplified with Taq Pro HS Probe Master Mix (Vazyme) according to the manufacturer’s protocol. Primer–probe sequences targeting the PEDV N gene were: forward, 5′-GTCTGAAAAGCCAATCATTC-3′; reverse, 5′-TTGCCTCTGTTGTTACTC-3′; probe, 5′-CTGTTGTTGCCATTGCCACGA-3′.

The obtained data were plotted using GraphPad Prism version 8.0 for Windows (version 11; GraphPad Software Inc), and statistical analysis was performed by one-way ANOVA or two-way ANOVA, followed by Tukey’s multiple comparisons test for comparisons among groups. The data are presented as Mean ± SD, and the significance of differences between groups was indicated by letters and asterisks. The number of letters (a, b, c, d, e) was only used to distinguish the hierarchy of differences among multiple groups and had no corresponding relationship with specific P-value ranges: groups labeled with the same letter indicated no significant difference between them, while groups labeled with different letters indicated a significant difference between them (at P<0.05, the order of difference levels was a>b > c>d > e). The asterisk annotation criteria were as follows: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, no significant difference.

The preparation strategy of PEDV VLP mRNA is illustrated in Fig. 1A. All formulated mRNA-LNP batches exhibited average particle diameters of approximately 75 nm and polydispersity indices below 0.1; the zeta potential was approximately −2.29 mV, and the encapsulation efficiency exceeded 90% (Fig. 1B), all of which collectively demonstrate high formulation quality and stability of the VLP mRNA vaccine.

To determine the expressions of PEDV S, E, M, and N proteins, the VLP mRNA were transfected into HEK293T cells, and WB assay showed the expressions of S, N, M, and E proteins at sizes of approximately 180, 55, 29, and 15 kDa, respectively (Fig. 1C), which revealed that the four structural proteins of PEDV were expressed successfully. To explore whether the co-expressed S, N, E, and M proteins could assemble into VLPs, the purified VLP samples were observed by TEM. TEM images showed that the VLPs exhibit the spike structures, with a diameter of approximately 100 nm, and are similar to coronavirus particles in terms of characteristics (Fig. 1D).

Mice were immunized with the PEDV VLP mRNA, S mRNA, and the dual inactivated vaccines according to the schedule shown in Fig. 2A. Throughout the study, no adverse reactions were observed. ELISA results (Fig. 2B and C) revealed that after two immunizations, serum IgG and IgA antibodies specific for PEDV-S increased progressively; the IgG and IgA titers in the VLP mRNA and S-mRNA groups were significantly higher than those in the PEDV/TGEV dual-inactivated vaccine and PBS control groups. In addition, serum IgG titers directed against E, M, and N proteins in the VLP mRNA and inactivated vaccine immunized groups were also significantly increased after the second immunization (Fig. 2D). Analyses of intestinal contents showed that IgG and IgA levels in the VLP mRNA group were significantly higher than those in all other experimental groups (Fig. 2E). The neutralizing antibody titers to the PEDV CH-HK-2021 strain of serum and intestinal contents in the VLP mRNA group reached approximately 1:2⁷, which was significantly higher than those in the S mRNA-immunized and inactivated-vaccine-immunized groups (Fig. 2F).

To detect the immune response induced by the VLP mRNA vaccine, mouse splenic lymphocytes were isolated at 28 dpv, and B- and T-lymphocyte subpopulations were characterized and counted by flow cytometry. The results showed that the percentages of CD3⁺CD8⁺ (Fig. 3A), CD3⁺CD4⁺ (Fig. 3B) T cells, and CD19⁺B220⁺ B lymphocytes (Fig. 3C) in the VLP mRNA group were higher than those in the S mRNA, inactivated vaccine, and the PBS control groups. The lymphocytes were re-stimulated in vitro with the PEDV viral solution (1:10) to analyze the antigen-specific proliferative response of the lymphocytes. VLP mRNA immunized group showed the highest stimulation index (Fig. 3D). To further characterize the cellular immune response induced by the VLP mRNA vaccine, we measured IFN-γ and IL-4 secreted by virus restimulated lymphocytes by ELISA (Fig. 3E and F). The results showed that the levels of IFN-γ in three vaccines immunized mice were all significantly increased. However, the level of IL-4 was only increased in the VLP mRNA immunized group. All data revealed that the VLP mRNA vaccine induces higher levels of humoral and cellular immunity.

In the passive immunization study section, we further evaluated the immunogenicity of the PEDV VLP mRNA vaccine in pregnant sows. For this purpose, we vaccinated six pregnant sows with either the VLP mRNA vaccine or PBS and received a booster vaccination at 14 dpv (Fig. 4A). Serum and colostrum were collected from sows to detect anti-PEDV antibody levels. Immunized sows showed no adverse reactions after vaccination. PEDV-S-conjugated IgG and IgA antibody titers in serum (Fig. 4B and D) and colostrum (Fig. 4C and E) were significantly elevated after immunization. IgG levels of anti-PEDV E, M, and N proteins in sow serum and colostrum (Fig. 4F and G) were also induced by VLP mRNA immunization. The neutralizing antibody titers against the different subtype PEDV strains of CH-HK-2021 (G2c subtype), AH2012/12 (G2b subtype), JS2008 (G1 subtype), and JS-NJ2023 (S-INDEL subtype) reached titers of 1:2⁶–2⁸ in serum (Fig. 4H), which exhibited a relatively good broad-spectrum neutralizing effect.

The ability of the VLP mRNA vaccine to confer passive immunoprotection to suckling piglets was next investigated. After 5 days of lactation, piglets from immunized and control sows were challenged with PEDV (Fig. 5A). Antibody levels in piglet sera were determined, PEDV-S-specific IgG and IgA antibodies were significantly induced (Fig. 5B and C), and neutralizing antibody titers against the CH-HK-2021 strain reached approximately 1:2⁶ (Fig. 5D), whereas no specific antibodies were detected in the sera of piglets born to control sows, suggesting that serum antibodies were transferred through colostrum. Besides, the antibodies against PEDV N, E, and M proteins in serum of VLP mRNA immunized piglets were also increased (Fig. 5E). After the challenge, piglets born to control sows developed diarrhea at 2 dpc, which progressed to severe watery diarrhea and lasted for up to 10 days. However, piglets born to immunized sows started to have diarrhea at 4 dpc and maintained only mild diarrhea for approximately 5 days (Fig. 5F). PEDV RNA levels in the feces of piglets born to immunized sows were significantly lower than those of control piglets (Fig. 5G). No piglets died in the immunized group after challenge, whereas one piglet died in the control group on each of days 6, 8, and 12 following challenges (Fig. 5H).

The gross and histological lesions in piglets born to immunized and control sows after disease onset were analyzed. As shown in Fig. 5I, no significant gross lesions were found in the intestinal tissues of piglets born to immunized sows, whereas the intestinal tissues of piglets born to the control sows showed thinning of the intestinal wall, pneumatosis, and hemorrhage. Pathological examination showed no obvious histopathological changes in the ileum of piglets born to immunized sows, whereas piglets born to control sows showed obvious histopathological changes in the ileum, including blunted and atrophied villi, vacuolization, and disappearance of intestinal crypts and villous structures. Immunofluorescence analysis revealed obvious PEDV aggregation in the atrophied ileal villi of control piglets. In contrast, only sparse PEDV-positive cells were observed in the ileum of piglets from immunized sows. Moreover, the jejunal villus height-to-crypt depth ratio was significantly higher in piglets born to immunized sows than in the control group (Fig. 5J). PEDV RNA loads in the jejunum and ileum were notably higher in control piglets compared with those from immunized sows (Fig. 5K). Additionally, intestinal IgA levels were determined, and the results demonstrated significantly elevated IgA concentrations in the intestinal tissues of piglets born to immunized sows (Fig. 5L). These results suggest that the PEDV VLP mRNA vaccine confers effective passive immunity against PEDV challenge in newborn piglets.

All data have been included in the article and supplemental material. All reported data and any additional information required to reanalyze the data are available from the lead contact upon reasonable request.