mRNA vaccine with a targeted protein delivers strong protection against pseudorabies virus

To identify the optimal antigenic target for PRV vaccine development, we first evaluated the immunogenicity of glycoproteins B (gB) and D (gD), which have been reported to contribute significantly to the induction of neutralizing antibodies (21, 22). We first expressed the gB and gD proteins in CHO cells. The expression of these proteins was confirmed via SDS-PAGE analysis and by using specific antibodies against gB and gD, respectively (Fig. 1A and B). We believe that an effective vaccine should stimulate the production of antibodies capable of inhibiting the virus's interaction with its receptors, thereby neutralizing it. Our experiment aims to investigate the interactions between viral proteins gB and gD with cell receptors. Subsequently, we test which protein binds to and competes for these receptors. To this end, various concentrations of gB and gD were co-incubated with the PRV-EGFP reporter virus (23). It was observed that a high concentration of the gD protein effectively competes for receptor-binding sites and inhibits PRV replication in Vero E6 cells, whereas the impact of gB is minimal (Fig. 1C). The relative infection rate was quantified by measuring the EGFP-positive area, which demonstrated that the purified gD protein significantly suppressed PRV replication (Fig. 1D). Following this, individual mice were immunized with either gB or gD proteins. After the booster immunization, antibodies from the gD-immunized group demonstrated a significantly higher titer of neutralizing antibodies and completely inhibited PRV infection in Vero E6 cells. In contrast, the antibodies derived from the gB-immunized group exhibited a relatively lower titer of neutralizing antibodies (Fig. 1E and F). Based on these findings, we selected the viral gD gene as our mRNA target for subsequent experiments.

The gD mRNA vaccine was designed according to the schematic shown in Fig. 2A. The gD ectodomain of PRV HeN1 variant strain was cloned into an mRNA vector containing a 5′-cap, as well as 5′ and 3′ UTRs. Furthermore, the tissue plasminogen activator signal sequence was incorporated to enhance the efficient secretion of gD proteins from eukaryotic cells. Subsequently, the gD mRNA transcripts were encapsulated within lipid nanoparticles (LNPs) comprising cationic lipid (SM102), phospholipid with two stearoyl chains and a phosphorylcholine headgroup (DSPC), cholesterol, and DMG-2000. The morphology of the mRNA vaccines is depicted in Fig. 2B. The resulting gD mRNA-LNPs formulations were evaluated for particle size, a critical parameter influencing delivery efficiency and biodistribution. Smaller particles can more readily penetrate tissues and cells, while larger particles may exhibit distinct pharmacokinetic properties. The optimal particle size range for effective delivery is generally between 50 and 200 nanometers. Furthermore, the polydispersity index (PDI), which indicates the uniformity of the particle size distribution, was assessed. A lower PDI value signifies a narrower size distribution, which is essential for ensuring consistent performance and safety. The results demonstrated that the gD mRNA-LNPs exhibited an average particle size of 100.62 nm with a narrow distribution characterized by a PDI of 0.146 (Fig. 2C and D). To further validate gD expression in vitro, we conducted an indirect immunofluorescence assay (IFA) (Fig. 2E) and Western blot analysis (Fig. 2F) after incubating HEK293T cells with gD mRNA-LNPs or Empty LNPs. The results demonstrated efficient expression of the gD protein, with GAPDH serving as a loading control.

We next evaluated the immunogenicity of gD mRNA vaccines in mice. A total of 12 mice were randomly divided into two groups (n = 6), with one group receiving immunization using a 10 µg dose of gD mRNA vaccine and subsequent boosting with an equivalent quantity of mRNA after 2 weeks from the initial immunization. Another group was designated as the control group and received empty LNPs for vaccination (Negative control, NC). All mice were vaccinated via intramuscular injection. Serum samples were collected at different time points for PRV antibody detection (Fig. 3A). The presence of PRV gD-specific antibodies was detected using indirect ELISA, and the results demonstrated that the gD mRNA immunized group exhibited significantly elevated levels of PRV gD-specific antibodies, which persisted for at least 10 weeks post-vaccination (Fig. 3B). In contrast, no detectable PRV gD-specific antibodies were observed in the control group (Fig. 3B). Additionally, no detectable PRV gB-specific antibodies were observed in the gD mRNA immunized group or control group using indirect ELISA (Fig. 3C). Collectively, these data suggest that the gD mRNA vaccine elicited a robust-specific immune response against gD. Furthermore, a high level of PRV neutralizing antibody was elicited by this vaccine as evidenced by the detected neutralizing antibody titer in the gD mRNA immunized group (Fig. 3D).

To evaluate the T-cell response stimulated by gD mRNA, mouse splenic lymphocytes were isolated and subjected to flow cytometry analysis of antigen-specific lymphocyte proliferation. Therefore, CD3⁺ CD4⁺ T-cells and CD3⁺ CD8⁺ T-cells were gated for analysis of antigen-specific lymphocyte proliferative responses (Fig. 3E). Subsequently, these lymphocytes were stimulated in vitro with gD mRNA-LNPs (10 µg) or empty LNPs. Additionally, we employed gE mRNA-LNPs (10 µg), another viral gene of PRV, as a more rigorous control. The results demonstrated that the groups immunized with gD mRNA exhibited a significantly higher rate of CD3⁺ CD4⁺ T-cell and CD3⁺ CD8⁺ T-cell proliferation compared to the control group (Fig. 3F and G). We further confirmed cellular immune responses by measuring IFN-γ secretion from lymphocytes stimulated with gD mRNA-LNPs. It was observed that lymphocytes from mice vaccinated with gD mRNA exhibited significantly higher levels of IFN-γ production compared to controls, which were stimulated with empty LNPs (Fig. 3H). These findings suggest that gD mRNA elicits robust PRV-specific humoral and T-cell responses in mice.

We next investigated the protective efficacy of gD mRNA vaccine against PRV challenge in mice (Fig. 4A). The gD mRNA group received two doses of the gD mRNA vaccine, each at a dosage of 10 µg via intramuscular injection. Our results demonstrated that the gD mRNA immunized group exhibited significantly elevated levels of PRV gD-specific antibodies (Fig. 4B and C) and neutralizing antibodies for the PRV HeN1 strain (Fig. 4D). Next, mice were challenged with the wild-type PRV HeN1 strain via intramuscular injection. In this study, we challenged mice with a high dose of wild-type PRV (2 × 10⁴ PFU). We observed that all mice immunized with the gD mRNA vaccine survived after challenge; however, the empty LNPs immunized group experienced 100% mortality following challenge (Fig. 4E). The clinical signs were assessed using clinical scores (Fig. S1). Additionally, viral loads in indicated tissues were detected, revealing significantly lower viral copies in the gD mRNA vaccine group compared to the empty LNPs-immunized group (Fig. S2). Furthermore, histopathological changes in the brain and lungs were evaluated. Our findings indicated that no histopathological alterations were observed in either the gD mRNA vaccine group or the negative control group. However, the group immunized with empty LNPs exhibited significant microscopic lesions in both the brain and lungs. Specifically, the alterations observed in the brains of mice treated with empty LNPs are characterized by glial cell proliferation and marked degeneration of neuronal cells (Fig. 4F). In the lungs of mice administered empty LNPs, congestion and infiltration by inflammatory cells lead to widening of alveolar septa, accompanied by partial degeneration and necrosis of bronchial epithelial cells. Additionally, fibrinous exudate is present within the alveolar spaces (Fig. 4F). Overall, the gD mRNA vaccine provided full protection against PRV challenge.

PRV typically spreads through the respiratory route. Therefore, we subsequently evaluated whether this vaccine induces mucosal protection. Following two rounds of immunization, we challenged the vaccinated mice with wild-type PRV (1 × 10³ PFU) via the intranasal route (Fig. 5A). The results demonstrated that all mice immunized with the gD mRNA vaccine survived following the intranasal challenge. In contrast, the group that received empty LNPs exhibited a 100% mortality rate after undergoing the same intranasal challenge (Fig. 5B). The clinical signs were further evaluated using clinical scores (Fig. 5C). Additionally, we assessed viral loads in the specified tissues. The results indicated a significantly lower number of viral copies in the gD mRNA vaccine group compared to the empty LNPs immunized group (Fig. 5D). Due to the potential for PRV infection to induce latent infections within the nervous system, we also evaluated viral loads in the trigeminal nerve, dorsal root ganglion, and sciatic nerve of mice that were immunized with gD mRNA-LNPs or empty LNPs following intranasal challenge. The results indicated that the gD mRNA vaccine provided complete protection against PRV intranasal challenge (Fig. 5E through G).

We next evaluated the immunogenicity of gD mRNA vaccines in piglets. The experimental timeline and design are schematically illustrated in Fig. 6A. There was no significant difference observed in the average daily weight gain between the group immunized with gD mRNA vaccines and the control group which was vaccinated with empty LNPs (Fig. 6B). Additionally, gD-specific antibodies were detectable 14 days post-vaccination (Fig. 6C). As a control, we also assessed the presence of gB antibodies; however, no gB antibodies were detected (Fig. 6D). Moreover, a robust PRV neutralizing antibody response was elicited at day 21 after gD mRNA immunization (Fig. 6E). To characterize cellular immunity, we performed lymphocyte proliferation assays (Fig. S3). Flow cytometric analysis showed significantly enhanced CD3⁺CD4⁺ T-cell and CD3⁺CD8⁺ T-cell proliferation indices in vaccinated piglets compared to controls (Fig. 6F and G). Furthermore, we confirmed cellular immune responses by measuring IFN-γ secretion from lymphocytes stimulated with gD mRNA-LNPs. Notably, piglets vaccinated with gD mRNA exhibited significantly elevated levels of IFN-γ production compared to control piglets (Fig. 6H). These findings indicate that in piglets, gD mRNA vaccine elicits potent PRV-specific humoral and T-cell responses.

We next investigated the protective efficacy of gD mRNA vaccine against PRV challenge in piglets. gD mRNA-LNPs-immunized group and the empty LNPs-vaccinated group were challenged with wild-type PRV HeN1 strain. The mock-infected group, serving as a negative control, was not exposed to the virus during the experimental period. It was observed that all piglets immunized with the gD mRNA vaccine survived after challenge; however, the empty LNPs-vaccinated group experienced 100% mortality following challenge (Fig. 7A). The rectal temperature and clinical signs were recorded (Fig. 7B and C).

Furthermore, gB-specific antibodies and gE-specific antibodies were detectable 10 days post-challenge in the group vaccinated with gD mRNA vaccine (Fig. 7D). Interestingly, there was a significant increase in neutralizing antibodies in the group vaccinated with gD mRNA-LNPs post-challenge (Fig. 7E). Furthermore, viral loads in serum were assessed and demonstrated lower viral copies in the gD mRNA vaccine group compared to the empty LNPs-vaccinated group (Fig. 7F). Additionally, virus shedding was evaluated by detecting viral copies in nasal and anal swabs, revealing a lower level of virus shedding in the gD mRNA vaccine group (Fig. 7G). Moreover, viral loads in indicated tissues were also evaluated, demonstrating lower viral loads in the gD mRNA vaccine group compared to the empty LNPs-vaccinated group (Fig. S4). Furthermore, histopathological changes in the brain and lungs were examined, with no significant histopathological alterations observed in the gD mRNA vaccine group, which were similar to those in the mock-infected group (Fig. S5). However, severe microscopic lesions were observed in the brain and lungs of piglets challenged with wild-type PRV in the empty LNPs-vaccinated group (Fig. S5). In pigs treated with empty LNPs, notable changes in brain tissue include proliferation of glial cells, extensive degeneration and necrosis of neuronal cells, presence of vascular cuffs, as well as occasional necrosis among glial cells. The alterations identified in the lungs of pigs receiving empty LNPs encompass congestion, significant perivascular edema, infiltration by inflammatory cells, along with fibrinous exudate within the alveolar lumen that is associated with necrotic shedding from alveolar epithelial cells (Fig. S5). The results of immunohistochemistry indicate that, in the gD mRNA vaccine group, viral antigens were undetectable (Fig. 7H). Overall, the gD mRNA vaccine provided complete protection against PRV challenge in piglets.

To evaluate the breadth of humoral immunity induced by the gD mRNA vaccine, sera collected at 28 days post-vaccination (dpv) from immunized mice and piglets were analyzed for cross-reactivity and neutralizing activity against heterologous PRV strains. Initially, we evaluated the conservation of gD amino acid sequences across various PRV strains. We retrieved 37 gD amino acid sequences from the NCBI GenBank database, which included 4 sequences from genotype I, 5 sequences from genotype II classic strains, and 28 sequences from genotype II variant strains. Subsequently, we employed MegaAlign software to analyze the conservation of the gD protein's amino acid sequence among different PRV strains. The results indicated that the sequence of the PRV gD protein is conserved, with only a few amino acids exhibiting mutations across all aligned sequences (Fig. 8A and B). To evaluate the cross-reactivity and broad-spectrum neutralizing activity elicited by the gD mRNA vaccine, we selected HeN1, SC, TJ, and Bartha K61 from various branches within the phylogenetic tree. IFA revealed that sera from both gD mRNA-vaccinated mice (Fig. S6A) and piglets (Fig. S6B) exhibited robust cross-reactivity with four genetically distinct PRV strains: HeN1, SC, TJ, and Bartha K61. Neutralization assays further demonstrated broad-spectrum protective potential. Serum from vaccinated mice (Fig. 8C) and piglets (Fig. 8D) effectively neutralized SC, TJ, and Bartha K61 strains, with neutralization titers comparable to those observed against the homologous HeN1 strain. This cross-neutralizing activity correlated with the conserved epitope recognition profile observed in IFA, confirming that gD mRNA immunization elicits antibodies targeting conserved antigenic regions across divergent PRV strains.

Human embryonic kidney cells (HEK293T, ATCC CRL-11268) and African green monkey kidney cells (Vero-E6, ATCC CRL-1586) were cultured in Dulbecco's modified Eagle's medium (DMEM, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Excell, Australia) at a temperature of 37°C with a CO₂ concentration of 5%. The PRV HeN1 strain (GenBank Accession No. KP098534↗), SC strain (GenBank Accession No. KT809429↗), TJ strain (GenBank Accession No. KJ789182↗), and Bartha K61 strain (GenBank Accession No. JF797217↗) were stored at −80°C in our laboratory. The PRV HeN1 EGFP strain was maintained as described in our previous report (29, 32). The Anti-GAPDH antibody was procured from Proteintech (Proteintech, China). PE/Cyanine5 anti-mouse CD3 antibody and APC/Cyanine7 anti-mouse CD8a antibody were procured from Biolegend (Biolegend, USA). The FITC rat anti-mouse CD4 antibody was sourced from BD-Pharmingen (BD-Pharmingen, USA). Mouse anti-porcine CD3E-SPRD antibody, mouse anti-porcine CD4-FITC antibody, and mouse anti-porcine CD8A-PE antibody were obtained from SouthernBiotech (SouthernBiotech, USA). Anti-mouse IgG-FITC was acquired from Sigma (Sigma, USA). Additionally, mouse monoclonal antibodies specific for the PRV gB were generously provided by Prof. Zhi-jun Tian and Jinmei Peng at our institute.

The gB and gD genes of the PRV HeN1 strain (GenBank Accession: KP098534↗) were codon-optimized and synthesized by Sangon Biotech (Sangon Biotech, China). Subsequently, the ectodomain of either gB or gD was cloned into the pb513B vector, which incorporates a Flag tag along with a 6× His tag. A stable CHO cell line expressing these constructs was then established as previously described (33). The CHO cells stably expressing the recombinant protein were established through puromycin selection. The recombinant protein present in the supernatant of cultured cells was collected and purified using Ni-NTA resin affinity chromatography (GenScript, USA) according to the manufacturer's instructions. SDS-PAGE and Western blot analyses were conducted to confirm the purity of the isolated protein. The concentration of the purified proteins was quantified utilizing a BCA Protein Assay Kit (Solarbio, China), after which these proteins were stored at –80°C.

To investigate the effect of gB or gD on viral replication, different concentrations of gB or gD protein were co-incubated with the PRV-EGFP reporter virus (multiplicity of infection = 0.01) on Vero E6 cells for 2 hours. After washing three times with cold PBS, the samples were cultured in DMEM containing 2% FBS. PBS was used as the control. After 36 hours of infection, the infected cells were examined under a microscope. The extent of infection was determined by quantifying the area occupied by EGFP-positive cells using Image J software.

For the vaccination of BALB/c mice, a total of 15 female BALB/c mice were randomly assigned to three groups. The mice received intramuscular injections of 100 µL (100 µg) of purified gB or gD protein in the hind leg, or phosphate-buffered saline (PBS), at weeks 0 and 2. During the initial immunization, the proteins were emulsified with Freund's complete adjuvant (Sigma, USA). In subsequent booster administrations, they were combined with Freund's incomplete adjuvant (Sigma, USA). Blood samples were collected from the retro-orbital sinus using glass capillaries. Serum samples were obtained 2 weeks after booster immunization and stored at −20 °C for neutralizing antibody detection.

HEK293T cells treated with gD mRNA-LNPs or empty LNPs were washed with PBS and fixed with 4% paraformaldehyde (PFA) for 15 minutes at room temperature and washed twice with PBS as described in our previous studies (34–39). Then, the cells were permeabilized with 0.25% Triton X-100 for 10 minutes at 4°C and blocked with 2% bovine serum albumin (BSA) (Coolaber, China) for 30 minutes at 37°C. After washing three times with PBS, the cells were subsequently incubated with PRV gD-specific primary antibodies at 37°C for 1 hour and washed three times with PBS. The cells were then incubated with appropriate secondary antibodies at 37°C for 1 hour and washed three times with PBS. The cell nuclei were stained with 4′,6-diamino-2-phenylindole (DAPI) for 15 minutes at 37°C and washed three times with PBS. Finally, the fluorescence signals were detected via microscope.

The western blot was performed as described in our previous reports (33, 40–42). Briefly, HEK293T cell lysates or PRV gB and gD proteins were mixed with loading buffer for SDS-PAGE. The separated proteins were transferred to PVDF membranes (Millipore, USA). The membranes were blocked with 5% nonfat milk in PBS for 1 hour and incubated with an antibody overnight at 4°C, followed by washing and incubation with the appropriate secondary antibody for 1 h at room temperature. Finally, the membranes were visualized using the Odyssey CLx imaging system.

The glycoprotein D of the HeN1 strain of PRV was employed as the reference amino acid sequence (GenBank Accession: KP098534↗). The ectodomain (1–355 aa) was cloned into an mRNA vector, followed by in vitro transcription using T7 RNA polymerase and a linearized plasmid DNA template containing optimized codons for the glycoprotein D. The lipids, including cationic lipid (SM102), phospholipid with two stearoyl chains and a phosphorylcholine headgroup (DSPC), cholesterol, and DMG-2000, were dissolved in anhydrous ethanol at a mass ratio of 50:10:38.5:1.5. Subsequently, the mRNA was dissolved in an aqueous phase prepared using citrate buffer solution at pH 4.0 (50 mM). The microfluidic device (LNP-S1-L, Fluidiclab, China) was utilized to package the mRNA with an alcohol phase to water phase ratio of 1:3 (43). The resulting mixture was then diluted with RNase-free PBS buffer and concentrated through ultrafiltration employing a 30 kDa cutoff membrane. To adjust the mRNA concentration to 60 µg/mL and the sucrose concentration to 10%, an equal volume of PBS solution containing 20% sucrose was added. Finally, the gD mRNA-LNPs vaccine preparation underwent filtration through a 0.22 µm membrane before being stored at −20°C after packaging. The particle size and PDI were measured by NanoBrook 90Plus device (Brookhaven Instruments Corporation, USA) according to the instructions.

HEK293T cells were seeded in 12-well plates. When the cells reached 70%–80% confluence, cells were treated with 2 µg gD mRNA-LNPs. The empty LNPs were served as the control group. After 24 hours of incubation, the expression of mRNA was analyzed by IFA and western blot (30, 44).

Serum-neutralizing antibodies against PRV were determined by performing a neutralization assay. The assay was performed as previously described with minor modifications (30, 49). Serum samples from mice and piglets were inactivated by heating at 56°C for 30 min and then twofold serial dilutions with DMEM. A similar amount of PRV HeN1 EGFP strain, HeN1 strain, SC strain, TJ strain, or Bartha K61 strain (200 TCID₅₀) was mixed thoroughly with 50 µL of the diluted inactivated serum for 2 h at 37°C. Subsequently, the mixtures of 100 µL were added to a 96-well plate with monolayer Vero E6 cells for 2 h at 37°C. Two hours later, the cells were washed three times with cold PBS and cultured in DMEM containing 2% FBS. Blank cells were set up as a negative control. The plates were incubated for 48 hours at 37°C with 5% CO₂. Neutralizing antibodies were calculated using the Reed-Muench method.

The splenic lymphocytes from mice or peripheral blood lymphocytes from piglets at 28 dpv were isolated to assess the proliferation of PRV gD-specific T cells using the CellTrace Violet Cell Proliferation Kit (Invitrogen, USA), following the manufacturer's instructions. The splenic lymphocytes or peripheral blood lymphocytes were stained with CellTrace Violet. When they are stimulated by specific antigens, they undergo division and proliferation, leading to a decrease in the violet fluorescence signal. This reduction can be utilized to assess cell proliferation status. Subsequently, the stained cells by CellTrace Violet were seeded into a 12-well round-bottom plate at a density of 1,000,000 cells per well and stimulated with gD mRNA-LNPs (10 µg), gE mRNA-LNPs (10 µg), or empty LNPs at 37°C with 5% CO₂ for 72 hours. Empty LNPs and gE mRNA-LNPs stimulated cells were used as negative controls. The splenic lymphocytes from mice were then stained with PE/Cyanine5 anti-mouse CD3 antibody (Biolegend, USA), APC/Cyanine7 anti-mouse CD8a antibody (Biolegend, USA), and FITC rat anti-mouse CD4 antibody (BD-Pharmingen, USA). The peripheral blood lymphocytes from piglets were stained with mouse anti-porcine CD3E-SPRD antibody (SouthernBiotech, USA), mouse anti-porcine CD4-FITC antibody (SouthernBiotech, USA), and mouse anti-porcine CD8A-PE antibody (SouthernBiotech, USA). After three washes with PBS, flow cytometric analysis was performed using an Apogee flow cytometer (UK). Data analysis was conducted using FlowJo version 10.8.

The cellular immune responses in the vaccinated piglets were evaluated using IFN-γ precoated enzyme-linked immunospot (ELISPOT) kits (Mabtech, USA) following the manufacturer's instructions. Briefly, the plates were blocked with RPMI 1640 medium (Gibco, USA) containing 10% FBS and incubated for 30 minutes. Peripheral blood lymphocytes from piglets were seeded at a density of 300,000 cells per well and stimulated with gD mRNA-LNPs (5 µg) or empty LNPs followed by incubation at 37°C with 5% CO₂ for 48 hours.

Empty LNPs stimulated cells were used as negative controls. The plates were washed five times with PBS and then incubated at room temperature for 2 hours with a detection antibody. Streptavidin-HRP was added to the plates and incubated for an additional hour. After final washes, the TMB substrate solution was applied and subsequently stopped with water. The air-dried plates were analyzed using an ELISpot reader (AID, Germany) to determine the number of spot-forming cells per million cells.

The spleen lymph single-cell suspensions from mice were seeded into 12-well plates at a density of 1,000,000 cells per well and incubated with gD mRNA-LNPs (10 µg) and Empty LNPs for 72 hours at 37°C. The cells stimulated by empty LNPs were used as the negative control. Subsequently, the cells were centrifuged and the supernatants were collected to measure IFN-γ levels using commercial mouse IFN-γ ELISA kits (Cusabio, China), following the manufacturer's instructions. The concentration of IFN-γ in the samples was determined based on standard curves.

The viral DNA from blood, tissues, and swab samples was extracted using the TIANamp Genomic DNA Kit (Tiangen, China) following the manufacturer's instructions. Quantification of the copy numbers of PRV's gB gene was performed using qRT-PCR, as previously described in our reports (23, 29, 40).

A total of 37 gD amino acid sequences were obtained from the NCBI GenBank database, comprising 4 sequences of genotype I, 5 sequences of genotype II classic strains, and 28 sequences of genotype II variant strains. Subsequently, the MegaAlign software was employed to analyze the conservation of the gD protein amino acid sequence across various PRV strains, while phylogenetic analysis was conducted using MEGA software. Sera collected at 28 days post-vaccination (dpv) from immunized mice and piglets were evaluated through IFA and neutralization assays against several PRV strains, including HeN1, SC, TJ, and Bartha K61, which represent different branches in the phylogenetic tree.

Statistical analysis was conducted via GraphPad Prism 8.0 software. Statistical significance was analyzed via t tests. A P value < 0.05 was considered statistically significant. Statistical significance is indicated by * P < 0.05, ** P < 0.01, *** P < 0.001, and **** P < 0.0001.