Cells and virus

HEK293T cells (ATCC, CRL-3216) or BHK-21 cells (kindly provided by Dr. Zhang Shuye) were cultured with complete Dulbecco’s modified Eagle’s medium supplemented with 10% (v/v) fetal bovine serum and 1% penicillin–streptomycin at 37 °C in 5% CO₂ incubator. CVS-11 rabies challenge virus (GenBank: GQ918139.1) was generously provided by Wuxi Xin Lian Xin Biotech Co., Ltd (Wuxi, China).

Analysis of functional T-cell epitopes of rabies large protein and identification of epitope-enriched fragments

Given the complexity of human leukocyte antigen (HLA) polymorphisms, 12 HLA-I and 54 HLA-II supertypes^22,23,24 were used in this study to predict restricted epitopes with maximum population coverage (covering more than 90% of selected HLA-I supertypes and 95% of selected HLA-II supertypes in the global population)²⁵, to trigger both CD4⁺ and CD8⁺ T cells for the development of potent and long-lasting immunity²².

The NetMHC pan 4.1 and NetMHCII pan 4.0 servers were used to predict the location of potential HLA-I or HLA-II epitopes of rabies large proteins from five strains: PM1503 (accession number: DQ099525), CTN-1 (accession number: FJ959397), aG (accession number: GQ412744), ERA (accession number: EF206707), and CVS-11 (accession number: GQ918139). All potential epitopes with high-affinity binding (the thresholds were 2% ranked for HLA-I and 5% ranked for HLA-II) were selected for further analysis. Then, a custom computer-based algorithm was developed to analyze the enriched regions among the high-affinity binding epitopes. Meanwhile, the amino acid sequences from the five large rabies proteins were aligned through the Clustal Omega program²⁵. The conservation score of amino acids in each position was also calculated. Next, the values of the corresponding region’s strong-binding affinity epitope score multiplied by its amino acid conservation score were used as a criterion to intercept the T-cell epitopes of large proteins. Additionally, the original three to five amino acids were contained within the N and C-terminal peptide flanks to promote the efficient presentation of the intercepted T-cell epitopes²⁶. Then, allergenicity and safety tests were conducted using AllerTOP V.2²⁷、AlgPred 2.0²⁸ 、AlgPred 2.0-IgE²⁹、AllergenFP V1.0, and AllerCatPro 2.0³⁰. Finally, the recombinant RABV-LT sequence was successfully constructed.

Vaccines

The mRNA vaccines used in this study were nucleoside-modified mRNAs obtained by completely replacing the uridine triphosphate (UTP) with 1-methylpseudourine-5’-triphosphate (Nanjing Synthgene Medical Technology Co., Ltd, China).

RABV-G mRNA vaccine expressing optimized rabies glycoprotein (PM, GenBank: AJ871962) was synthesized through in vitro transcription using a T7 high-yield RNA transcription kit following the manufacturer’s protocols (Novoprotein, Shanghai, China) on linearized plasmid templates, as previously reported²¹. RABV-LT sequences were also cloned into the linearized plasmid template, similar to the method used for the RABV-G mRNA vaccine. Then, an internal ribosome entry site (IRES) from the encephalomyocarditis virus (EMCV) was introduced between the RABV-G and RABV-LT genes, creating a dual-immunogen mRNA, referred to as RABV-G-LT. The carboxy-terminal end of the RABV-G-LT gene includes an Escherichia coli dihydrofolate reductase (DHFR) domain, along with a FLAG tag for detecting RABV-LT expression. After transcription, the mRNAs were capped with the vaccinia capping system and an mRNA cap 2′-O-methyltransferase (Novoprotein, Shanghai, China), followed by purification using LiCl and 70% ethanol as we previously described²¹. The mRNAs were evaluated by agarose gel electrophoresis, and protein expression was detected in 293 T cells (see below). Finally, the mRNA-LNP vaccines were formulated by mixing the mRNAs in the aqueous solution and an ionizable cationic lipid mixture in ethanol using the previously described self-assembly method³¹. The concentration of the mRNA-LNP vaccines was calculated through a Quant-iT RiboGreen assay (Invitrogen, R11490).

The Inactivated Rabies vaccine in the study was obtained from Sigma–Aldrich (Product Code: EPR0100000, assigned the activity of 10 IU/vial, France).

Protein expression and identification

RABV-G mRNA, RABV-LT mRNA, or RABV-G-LT mRNA was transfected into 293 T cells using Lipofectamine 3000 Reagent (Thermo Fisher Scientific, L3000001) to confirm in vitro expression. In brief, the cells were plated into six-well plates 12 h before transfection. Then, mRNAs and Lipofectamine 3000-mRNA were mixed in Opti-MEM (Gibco) in a ratio of 1:2 (4 μg mRNA vs. 8 μL of transfection reagent), and the mixture was added to each well. The cells transfected with RABV-G or RABV-LT mRNA were harvested 24 h later. For RABV-G-LT mRNA, the TMP132 (Sigma–Aldrich) was added to the cell culture medium 6 h before collection. All the cells were lysed with 4× sodium dodecyl sulfate loading buffer (TaKaRa) for Western blot analysis. The primary antibody Rab-50 (Santa Cruz) was used to identify RABV-G (glycoprotein) protein, and the monoclonal ANTI-FLAG M2 antibody (Sigma–Aldrich) was used to detect RABV-LT protein, followed by peroxidase-conjugated goat anti-mouse immunoglobulin G (IgG) H&L (Yeasen, China).

Animal experiments

Female BALB/c mice (6–8 weeks old, specific-pathogen-free) were purchased from Suzhou Hua Chang Biological Co., Ltd Female rhesus macaques (~3 years old) obtained from a domestic source (Ningbo, China) were all housed in the Shanghai Public Health Clinical Center (SPHCC) animal facility. The animal experiments were conducted according to the recommendations of the SPHCC Guide for the Care and Use of Laboratory Animals.

The immunization schedules were presented thoroughly in the relevant figures. For mouse immunization, all the mRNA vaccines were injected into the thigh of the hindlimb intramuscularly at a dose of 3 μg in a 100 μL volume. The mice in the negative control group were injected intramuscularly with empty LNP. Further, the mice were immunized using an inactivated vaccine intramuscularly (1 IU/mL). The sera or spleen samples were collected at different time points according to the immunization schemes shown for specific antibody analysis or T-cell responses. Following vaccination, the mice were challenged with 20 or 40 MLD₅₀ (mouse LD₅₀, 50 μL per mouse) of CVS-11 intramuscularly. Then, the body weight and survival were monitored daily. Three mice in each group were euthanized 7 or 12 days post-infection (dpi), and the spinal cord or brain tissues were collected to quantify viral loads and histopathological analysis (equivalent portions of the brain tissues).

Female rhesus macaques were randomly assigned to five groups (N = 5 in each vaccination group, N = 4 for empty-LNP negative control group): RABV-G mRNA (3 or 10 μg per macaque), RABV-G-LT mRNA (3 or 10 μg per macaque), and empty-LNP. All the vaccines (volume: 1 mL) were intramuscularly injected into the left deltoid muscle with a prime-boost regimen at 6-week intervals. Blood was collected at various times to isolate the serum and PBMCs (Peripheral Blood Mononuclear Cells).

Antibody analysis

RABV-G-specific antibodies in mice or rhesus macaque sera were analyzed using ELISA. The RABV-G protein was obtained from AtaGenix (China) and precoated with 1 μg/mL. The sera were diluted 200-fold in the first well, followed by a twofold serial dilution. For IgM or total IgG, horseradish peroxidase (HRP)-conjugated Rabbit anti-Mouse IgM Ab (Abclonal, China) or HRP-conjugated anti-mouse IgG (Yeasen, China) at a 1:5000 dilution was added. For the subclass of antibodies, biotin-conjugated goat anti-mouse IgG1, IgG2a, and IgG2b (1:5000; Abcam) were incubated in the plates. Then, streptavidin-conjugated HRP (SA-HRP) (1:5000) (Yeasen) was incubated. Finally, the plates were colored using the substrate OPD (Sigma–Aldrich), and the absorbance at 490 nm was measured using a Synergy microplate reader (Bio-Tek). The endpoint titers were determined at the highest dilution, and the cutoff value was defined as yielding an OD value twice that of the empty LNP control. All the sera needed to be inactivated at 56 °C for 30 min to evaluate virus neutralization. The anti-RABV-neutralizing antibodies in the sera were determined using the FAVN (Fluorescent Antibody Virus Neutralization) test according to the World Health Organization protocol as described earlier³².

Enzyme-linked immunosorbent spot assay mice

The enzyme-linked immunosorbent spot (ELISPOT) assay was performed following the manufacturer’s protocols (BD Bioscience). Briefly, the plates were precoated with anti-mouse interferon (IFN)-γ captured antibodies (5 μg/mL) overnight at 4 °C. Further, 2 × 10⁵ splenocytes were added to the plates in duplicates and stimulated with the RABV-G peptide pools (5 μg/mL), RABV-LT peptide pools (5 μg/mL), or the media for 20–24 h at 37 °C. This was followed by incubation sequentially with biotin-conjugated mouse anti-IFN-γ antibody and alkaline phosphatase-conjugated SA. Finally, the plates were colored using an AEC substrate reagent. The spots in plates were scanned and counted with a BioSpot plate reader (ChampSpot 437III; Beijing Sage Creation Science Co., Ltd).

Macaques

ELISPOT assay was performed with the nonhuman primate IFN-γ kits (Mabtech) following the manufacturer’s protocols. A total of 1 × 10⁵ PBMCs were stimulated with RABV-G peptide pools (2 μg/mL), RABV-LT peptide pools (2 μg/mL), or the media as negative control. The plates were incubated at 37 °C for 48 h. The spots were scanned and analyzed using an ELISPOT reader, as mentioned earlier.

Peptide-specific spots were determined after the deduction of spots of the negative control wells (wells without peptides) from spots of RABV-G peptides or RNBA-LT peptides. Results were considered as the average number of peptide-specific spots in replicate wells.

Flow cytometry assay

Antigen-specific T-cell response was analyzed using intracellular cytokine staining (ICS). Briefly, 1 × 10⁶ isolated splenocytes of mice were suspended in phosphate-buffered saline containing 0.2% bovine serum albumin and counted, followed by stimulation with RABV-G peptide pools (1 μg/mL), RABV-LT peptide pools (1 μg/mL), or the media. Golgi-Plug (BD Biosciences) was added at 1:1000 dilution after 1 h. The cells were stained with Amcyan LIVE/DEAD Aqua (BioLegend) and incubated with antibodies against CD3 (Percpcy5.5, Clone 17A2; BioLegend), CD4 (AF700, Clone RM4-5; BD Biosciences), and CD8 (fluorescein isothiocyanate, Clone 53-6.7; BioLegend). After fixation and permeabilization (BD Biosciences), the intracellular staining was employed through staining with antibodies against anti-IFN-γ (PE, Clone XMG1.2; BD Biosciences), anti-IL-2 (APC, Clone JES6-5H4; BioLegend), or anti-TNF-α (BV605, Clone MP6-XT22; BioLegend). The cells were collected using BD FACSAria III flow cytometer (BD Biosciences) and then analyzed using FlowJo 10 software.

Viral RNA extraction and RT-PCR quantification

The brain or spinal cord tissues were collected using the RNAzol RT Reagent (Molecular Research Center) following the manufacturer’s protocols. Then, the total RNA was extracted using the RNA isolation kit (Direct-zol RNA Miniprep Plus kit; Zymo Research). The concentration was measured using the NanoDrop reader (BioTek). One-step reverse transcription–polymerase chain reaction (RT-PCR) was carried out using the HiScript II One-Step qRT-PCR SYBR Green Kit (Vazyme) and a Bioer real-time PCR system to compare rabies viral RNA copies in each group.

The oligo primers used for rabies nucleoprotein were as follows:

Forward: 5′-AATGCGACGGTTATTGCTGC-3′; reverse: 5′-TGCCACGTCGGTCTTTGTTA-3′.

The RT-PCR cycling was performed as follows: 50 °C for 15 min, 95 °C for 2 min, followed by 40 cycles of 95 °C for 10 s and 60 °C for 30 s. The relative fold change was calculated using the 2^−ΔΔCt method and normalized based on the non-infection control (blank group).

Histopathological assay and RNAscope in situ hybridization (ISH) assay

The brain tissues from mice were collected, fixed immediately with 4% (v/v) paraformaldehyde, and embedded in paraffin to sagittally cut paraffin sections with a 4–5 μm thickness. Following deparaffinization, the histopathological changes in the brains of mice in the infected group were examined using hematoxylin and eosin staining and viewed under a light microscope.

RNAscope in situ hybridization (ISH) was performed to analyze rabies viral RNA expression in brain sections. The viral nucleoprotein (NP) RNA served as the target due to its high conservation across different rabies virus strains. For this assay, the NP-specific RNAscope probe (V-RABV-gp4 (220268) from ACDBio) was employed to detect viral RNA. The RNA ISH assay was conducted using the RNAscope Multiplex Fluorescent Reagent Kit V2 (Advanced Cell Diagnostics, 323100).

Statistical analysis

In the study, the data were presented as mean ± SEM. The titer Data were shown as GMT ± geometric SD. One-way ANOVA followed by Tukey’s multiple comparisons tests and Kruskal–Wallis (survival data) was used for statistical analysis using GraphPad Prism (Version 9.4). P values < 0.05 (^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001; ns, not significant) indicated statistically significant differences.

Ethics statement

All animal experiments in this study were approved by the Institutional Animal Care and Use Committee (IACUC) of Shanghai Public Health Clinical Center. Macaque studies and related experimental procedures were approved by the Laboratory Animal Welfare and Ethics Committee of Shanghai Public Health Clinical Center (SPHCC) (approval number: 2022-A038-01).

All mouse experiments were conducted after approval by the Laboratory Animal Welfare and Ethics Committee of Shanghai Public Health Clinical Center (SPHCC) (approval number: 2021-A073-02). All infection experiments were performed in the ABSL2 laboratory following guidelines of Environmental Health and Safety. All work was performed with approved standard operating procedures and safety conditions for the RABV. The procedures used for anesthesia and euthanasia of study animals followed tenets of the ARRIVE reporting guidelines. During the RABV challenge study, the mice with body weight drop more than 30%, severe paralysis, or inability to feed were euthanized using carbon dioxide inhalation.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.