Engineering a broad-spectrum vaccine to combat emerging monkeypox virus via immunoinformatic approaches

doi:10.21203/rs.3.rs-5048713/v1

Monkeypox virus (MPXV) has caused 41,664 confirmed cases and five deaths in nonendemic regions, as reported by the World Health Organization (WHO). There is an urgent demand for effective vaccines to combat and prevent the spread of MPXV. Traditional vaccine development is low-throughput, expensive, time-consuming, and susceptible to reversion to virulence. As an alternative, a reverse vaccination approach could be used as a promising tool for designing effective and safe vaccines against MPXV. Here, the associations of MPXV proteins with viral infection were analyzed for potential immunogenic epitopes to design multi-epitope vaccine constructs based on B cell, CD4+, and CD8+ epitopes. Epitopes were selected based on allergenicity, antigenicity, and toxicity parameters. The prioritized epitopes were then combined via peptide linkers and N-terminally fused to various protein adjuvants, including PADRE, beta-defensin 3, 50S ribosomal protein L7/12, RS-09, and the cholera toxin B subunit (CTB). All the vaccine constructs were further computationally validated for their physicochemical properties, antigenicity potential, allergenicity, safety, solubility, and structural stability. The three-dimensional structure of the selected construct was also predicted. Moreover, molecular docking and molecular dynamics (MD) simulations between the vaccine and the TLR-4 immune receptor demonstrated a strong and stable interaction. The vaccine construct was codon-optimized for high expression in the E. coli platform and was ultimately cloned in silico into the pET21a(+) vector. Collectively, these results could lead to the use of innovative tools for vaccine formulation against MPXV and for the treatment of other infectious diseases.

Computational Biology

Multi-epitope vaccine

Monkeypox virus (MPXV)

Immunoinformatic approaches

Monkeypox is a zoonotic disease caused by monkeypox virus (MPXV), which is an enveloped double-stranded DNA (dsDNA) virus that is categorized in the Orthopoxvirus genus of the Poxviridae family. The symptoms of monkeypox include fever, rash, lymphadenopathy and intense asthenia; however, certain groups (children, pregnant women, and immunocompromised people) might suffer from serious complications, including secondary bacterial infections, pneumonitis, respiratory distress, sepsis, encephalitis and loss of vision due to corneal infection. The fatality rate of MPX ranges between 0 and 11%, with a self-limiting disease lasting 2–4 weeks ¹. MPXV was first discovered in monkeys in 1970; however, it was first confirmed in humans in 1970 in a 9-month-old child with suspected smallpox in the Democratic Republic of the Congo ². Thereafter, infection cases have been occasionally reported, but most MPX cases were restricted to West African and Central African nations, with a mortality rate between 1 and 10% ^3,4. Monkeypox was documented to be transmitted by travelers from human-to-human transmission or by exposure to imported animals⁵. Since early May 2022, the global incidence of monkeypox (MPX) disease caused by the virus has dramatically increased worldwide, with 92,783 laboratory-confirmed infection cases and 171 deaths reported by the World Health Organization (WHO) from 116 reporting countries (as of 19^th December 2023) ⁶. Given the widespread prevalence of MPX disease, the WHO declared this disease a public health emergency of international concern in July 2022. Due to its genetic similarity to smallpox infection, patients with severe MPXV infection are treated with antiviral medicines for smallpox virus infection, including tecovirimat, cidofovir, and brincidofovir, which are currently used to treat MPX under certain conditions; however, high-dose antiviral drug medication tends to cause hepatic malfunctions, while low-dose treatment may lead to the risk of relapse ⁷. Thus, there is an urgent demand for effective vaccines against MPXV infection to curb the number of cases.

However, generating such vaccines remains a significant challenge, and to date, no specialized vaccines have been developed specifically to protect against MPX infection. Several smallpox vaccines, including ACM2000, MVA-BN, and LC16, are available for MPX vaccination in different jurisdictions. The US FDA-approved ACAM2000 is available for smallpox and MPX vaccination. However, ACAM2000 is a live, replication-competent vaccinia virus, and there is a risk of side effects such as injection site pain, lymph node pain, pruritus, and other flu-like symptoms. Additionally, serious and long-term side effects, including myopericarditis or pericarditis, have been reported in every 20 cases out of 100,000 ACAM2000 vaccine recipients ⁸. MVA-BN, also known as JYNNEOS, a nonreplicating modified Vaccinia Ankara virus vaccine, was approved by the US FDA in 2019 for the prevention of both smallpox and monkeypox infections in the U.S. Since Jynneos is a replication-deficient live virus vaccine, it is safer than ACAM2000, particularly for vaccination in immunocompromised individuals, such as transplant recipients, patients with HIV, and atopic dermatitis ^9,10. However, limited data on the efficacy of Jynneos against MPXV infection in humans have been documented ¹¹. LC16, a third-generation smallpox vaccine, is a live, replicating attenuated vaccine derived from the Lister (Elstree) strain of vaccinia that was approved for the prevention of MPX in August 2000 in Japan ^12,13. Although preclinical evidence has demonstrated a safe profile in animal models, the LC16 vaccine is not recommended for administration to immunocompromised patients or atopic dermatitis patients ^8,14,15. Nevertheless, the efficacy of these vaccines against monkeypox is contingent on data derived from animal studies demonstrating immune protection against the monkeypox virus in nonhuman primates rather than human subjects ¹⁶.

The mutation rate of MPXV was 6-12-fold greater than previously expected, and some of these mutations led to increased transmissibility and immune evasion. Thus, there is an urgent need for novel vaccine design methods ^12,17. The drawbacks of conventional vaccine development include low throughput, high cost, and high time consumption, as well as the need for pathogen culture and identification of immunogenic components^18,19. To address these shortcomings in conventional vaccine development, reverse vaccination approaches have revolutionized vaccine design, thus enabling scientists to identify immunogenic components and highly conserved epitopes from genome sequence and proteome information without the need to isolate and culture the targeted pathogen and thus offering time and resource savings²⁰. Multi-epitope vaccines developed by immunoinformatics approaches have demonstrated robust immune protection against microbial infection because of fewer side effects, greater specificity, greater safety, lower cost production than conventional vaccines, and simultaneous induction of diverse immune responses, including innate, cellular and humoral immune responses^21,22. The critical step in developing an effective multi-epitope vaccine is target selection. Since there are no reports on specific MPXV proteins that are predominantly involved in MPXV infection, several proteins, including structural and nonstructural proteins, could play critical roles in MPXV infection and could be promising targets for vaccine development. Several MPXV proteins, such as M1R (myristylprotein), B6R (EEV type-I membrane glycoprotein), and H3L (IMV heparin binding surface protein), are associated with immune responses within poxviruses and are considered attractive targets for vaccine development²³. In addition, B9R has been reported to be involved in the pathogenesis of MPXV²⁴. F8L (DNA polymerase) has been previously reported to stimulate the CD8+ T-cell population and the release of interferon-g in rhesus macaques²⁵. Accordingly, both structural proteins and nonstructural proteins could be promising targets for vaccine development.

Here, a panel of multi-epitope vaccines was designed against MPXV using a reverse vaccination approach. The MPXV proteins were retrieved from the GenBank database and filtered through several criteria to select final candidates for vaccine construction. Both structural proteins and nonstructural proteins were selected as protein targets for multi-epitope vaccine development. According to previous studies, B6R, H3L, and F8L can stimulate interferon-γ production and activate cytotoxic T lymphocytes (CTLs) and helper T lymphocytes (HTLs) ^26–30. E4R has been reported to induce antibodies in vivo in prairie dogs ³¹. A40L, B9R, and C2L are homologous to the A38L, B8R, and K2L vaccinia viral proteins, which are reportedly crucial ligands for MHC class II and HTL ^32,33. J3R/J1l plays a pivotal role in immunomodulating and preventing apoptosis in host cells and thus facilitating viral entry and infection ³⁴. Moreover, to provide broad-spectrum protection, six different MPXV strains were used as our research objects. All candidate epitopes were then further selected based on their antigenicity, low allergenicity, and low toxicity. To enhance immunostimulation, an adjuvant protein sequence was fused to the N-terminus of each vaccine construct. Overall, our molecular docking and molecular dynamics simulation data suggested that our vaccine candidate could bind to and trigger an immune response via TLR-4. Hence, our MPXV multi-epitope vaccine construct could be a viable alternative for preventing the global outbreak of the MPX pandemic.

Monkeypox viral protein retrieval and screening Monkeypox protein sequences, including structural and nonstructural proteins retrieved from the MPXV strain Congo_2003_358 in the NCBI database (DQ011154.1), were predicted for antigenicity and allergenicity via VaxiJen 2.0 and AllergenFP1.0. From the 200 proteins retrieved from the completed genome, only 24 proteins were selected based on non-allergenicity and antigenicity, with a score higher than 0.4, as demonstrated in Table 1A and 1B. These 24 selected proteins were further utilized for epitope prediction.

Prediction of epitope candidates and assessment of allergenicity, antigenicity, and toxicity parameters.The NetMHCpan 4.1 BA and NetMHCpan 4.1 EL servers predicted 35 CTL epitopes that strongly bind to MHC class I for HLA class I alleles (HLA-A0101, HLA-A0201, HLA-A0301, HLA-A2402, HLA-A2601, HLA-B0702, HLA-B0801, HLA-B2705, HLA-B3901, HLA-B4001, and HLA-B5801). All 35 selected CTL epitopes demonstrated a positive immunogenicity score via the IEDB Class I immunogenicity tool, antigenicity with a score higher than 0.4, non-allergenicity, and non-toxicity.

HTL epitopes were predicted through NetMHCpanII 4.1 BA and NetMHCpanII 4.1 EL for MHC class II. These servers predicted 24 potential HTL epitopes with strong binding affinities against MHC class II based on the HLA alleles DRB1_0101, DRB1_0301, DRB1_0401, DRB1_0701, DRB1_0801, DRB1_0901, DRB1_1001, DRB1_1101, DRB1_1201, DRB1_1301, DRB1_1501, and DRB1_1602. Additionally, these HTL epitope candidates were tested with the IFNepitope server to selectively filter those sequences that can stimulate IFN-γ. All 24 HTL epitopes could trigger IFN-γ secretion. Out of the 24 epitopes, only one epitope had a low antigenicity score (less than 0.4); thus, 23 of the remaining HTL epitopes had high antigenicity, no allergenicity, and no toxicity.

In the case of LBL, the BepiPred linear epitope prediction 2.0 server was used to predict the LBL epitopes with the default threshold of 0.5. The server yielded a total of 30 epitopes predicted to have high antigenic scores, no allergenicity, and no toxicity. The CTL, HTL, and LBL epitope repertoires were utilized for further analysis to select final candidates for vaccine construction.

Transmembrane and signal peptide screening The CTL, HTL, and LBL predicted epitopes obtained from previous experiments were further screened for transmembrane regions and signal peptides by utilizing DeepTMHMM and SignalP 6.0. DeepTMHMM was used to predict the transmembrane regions of epitopes that should be removed to avoid hydrophobic effects that are prone to aggregation, thus leading to low yields and unstable protein synthesis ^35–37 (Supplementary Dataset 1). The CTL, HTL, and LBL epitopes demonstrating potential transmembrane regions were retracted from the constructed vaccines. Our CTL, HTL, and LBL epitopes contained no transmembrane region. Finally, SignalP 6.0 was used to predict the signal peptide regions of the CTL, HTL, and LBL epitopes, and no signal peptides were found in our CTL, HTL, or LBL epitopes.

Epitope conservancy for the board spectrum coverage of vaccines across different clades. To generate broad-spectrum vaccines against various clades of Monkeypox virus, all the predicted CTL, HTL, and LBL epitopes were analyzed for epitope conservation across 7 clades of Monkeypox virus. Nearly all the epitopes covered Clade I (Zaire-96-l-16), Clade IIa (USA 2003), Clade llb-A (Nigeria 2018), Clade IIb-A.2 (USA 2002), Clade IIb-A.1 (UK P2), Clade IIb-A.1.1 (USA2021), and Clade IIb-B.1 (France 2022), with 100% conservancy. However, an HTL epitope derived from the F8L protein (SMVFEYRASTVIKGP) showed only a single amino acid variation from valine to isoleucine. One valine in Clade I, Congo_2003_358 and Zaire-96-l-16 was replaced by an isoleucine in Clade IIa (USA 2003), Clade llb-A (Nigeria 2018), Clade IIb-A.2 (USA 2002), Clade IIb-A.1 (UK P2), Clade IIb-A.1.1 (USA2021), and Clade IIb-B.1 (France 2022). In other words, the SMVFEYRASTVIKGP in Clade I became the SMVFEYRASTIIKGP in Clade II (Supplementary Table S1). Considering the amino acid properties, both valine and isoleucine are categorized as hydrophobic amino acid groups; thus, the valine to isoleucine substitution should have a negligible effect on the epitope.

Epitope homology analysis with the human proteome. To prevent autoimmune induction by epitopes, the Peptide Search server from the UniProt database was used to investigate epitopes that showed homologous sequences to the human proteome. According to the peptide match tool, the predicted epitopes were checked against the Homo sapiens proteome (Taxonomy ID: 9606). The results showed that all the epitopes were not homologous to the human proteome.

Final epitope selection and worldwide population coverage Epitopes showing nonhomologous sequences to the human proteome were further selected for the vaccine construct. To ensure broad HLA coverage, epitopes were chosen to target different alleles for both CTL and HTL. These candidates were selectively recruited based on HLA allele coverage for CTL and HTL epitopes and high antigenicity for all three types (CTL, HTL, LBL). Following this method, 9 CTL, 4 HTL, and 4 LBL epitopes were selected, as illustrated in Table 2A, 2B and 2C.

Afterwards, the selected CTL and HTL epitopes were analyzed via the IEDB population coverage tool. The distribution of HLA alleles targeted by the combination of the predicted MHC class I and II epitopes. The results showed that 94.87% of the world population across 115 nations and 21 ethnicities were covered within 16 geographic regions, as shown in Fig. 1.

Multi-epitope subunit vaccine construct. The design of the vaccine construct was formulated by combining 5 elements, including an adjuvant, 9 CTL epitopes, 4 HTL epitopes, 4 LBL epitopes, and a 6xHis tag, with the help of linkers, as illustrated in Fig. 2. Linkers have been reported to provide stability to vaccine constructs and prevent cross-interruption of each epitope, thus allowing them to independently function upon vaccination ^38,39. Five different adjuvants (cholera toxin B subunit (CTB), PADRE, beta-defensin 3, 50S ribosomal protein L7/12, and RS-09) were used in this study, resulting in five vaccine constructs. The adjuvants were N-terminally fused to the vaccine by using the EAAAK linker. The CTL epitopes were linked with AAY linkers, and GPGPG linkers were used to connect the HTL epitopes. The B-cell epitopes were linked with the help of KK linkers. HEYGAEALERAG linkers were used to connect the first HTL epitope and the last CTL epitope as well as to link the first B-cell epitope to the last HTL epitope. Additionally, the 6xHis tag was fused to the C-terminus of the vaccine construct by the RVRR linker for vaccine purification by affinity chromatography and immunodetection of vaccine expression. A schematic of each vaccine construct is depicted in Fig. 2. The antigenicity, allergenicity, toxicity, solubility, presence of signal peptide, and physicochemical properties of all five vaccine constructs were predicted and compared to those of a control vaccine without an adjuvant.

Antigenicity, allergenicity, toxicity and physicochemical properties The five vaccine constructs were evaluated and compared for antigenicity, allergenicity, toxicity, and physicochemical properties, as shown in Table 3. All five vaccine constructs demonstrated antigenicity scores higher than 0.50 according to the Vaxijen 2.0 server, which indicated the robust antigenic properties of all the multi-epitope vaccine constructs. Additionally, four vaccines were immunogenic and nontoxic; however, the beta-defensin adjuvant was toxic. The physicochemical properties predicted from the ProtParam server showed that the molecular weights of all the multi-epitope vaccine constructs ranged from ~29 to ~40 kDa. The theoretical isoelectric points (PIs) of all the constructs ranged from 5.46 to 8.71. The estimated half-lives of in vitro mammalian reticulocytes were 30 hours, >20 hours in in vivo yeasts, and >10 hours in in vivo E. coli. The instability indices of the cholera toxin B-vaccine, PADRE-vaccine, and 50S ribosomal protein L7/12 ranged from ~28 to ~35, which indicated that these constructs were stable at various temperatures ⁴⁰. However, the instability indices of the RS09-vaccine and beta-defensin-vaccine strains were greater than 65, indicating their instability ⁴⁰. The gravy scores of all the vaccines ranged from -0.427 to -0.699, indicating a hydrophilic nature ^40,41. In line with the Gravy scores, the solubility scores calculated by the SOLpro server ranged from ~0.72 to ~0.98, indicating high solubility ⁴². The aliphatic indices of cholera-vaccine, PADRE-vaccine, the RS09-vaccine, and the 50S ribosomal protein L7/12 ranged from 65.88 to 77.81, indicating high thermostability, whereas the aliphatic index of the beta-defensin-vaccine was 35.21, indicating low thermostability ⁴³. Finally, the signal peptide sequence of the vaccine was also tested; however, no signal peptide sequence was found within any of the tested vaccines. Finally, the signal peptide sequence of the vaccine was also tested; however, no signal peptide sequence was found within any of the tested vaccines. Collectively, based on the above predictions, the PADRE-vaccine demonstrated overall improvement in terms of antigenicity, solubility, and stability without compromising allergenicity or toxicity. Thus, the PADRE-vaccine was recognized as the lead vaccine construct for immunological translation and was chosen for further analysis.

Secondary structure prediction, three-dimensional structure predictions, refinement, and validation The PADRE-vaccine showed overall improvements in both immunological and physicochemical properties; therefore, it was used to predict secondary and three-dimensional structures. The secondary structure of the vaccine was predicted by using the PSIPRED 4.0 server to identify features. The predicted secondary structure included 44.6% coils, 33.7% α-helices, and 21.8% β-strands, as shown in Fig. 3. Additionally, the three-dimensional structure of the PADRE-vaccine was predicted using Alphafold 2. Five three-dimensional structure models were generated by Alphafold 2, and the model with the highest pLDDT was selected for further analysis. To improve the accuracy of the predicted three-dimensional structure model, the lead model obtained from Alphafold 2 was further refined by the GalaxyWEB server, resulting in five refined models with improved quality demonstrated by quality assessment parameters, as shown in Table 4. Model 2 was selected based on improved Molprobity and Rama favored scores of 1.055 and 98.6, respectively, which indicated the high quality of the model. Pymol was used to visualize the 3D model of the vaccine construct (Fig. 4).

Model 2 was further analyzed for validation with the SAVES 6.0 server by ERRAT and PROCHECK. The selected model exhibited a high quality score of 86.9281% according to the ERRAT. According to the plot of PROCHECK for evaluation via Ramachandran, 97.5% of the residues were found in the preferred region, 1.6% were found in the allowed region, 0.4% were found in generously allowed regions, and only 0.4% were found in forbidden regions, as shown in Fig. 5a. Additionally, the ProSa-Web server was used to compute the z-score of the model. The z-score calculated by the ProSa-web server was -4.62, as shown in Fig. 5b. Collectively, the results from ERRAT, PROCHECK, and ProSa-Web indicated the excellent quality of the predicted three-dimensional model.

Molecular docking of TLRs and selected vaccine constructs. The ClusPro 2.0 server was used to predict the binding affinity between the PADRE-vaccine and TLR4. The results included cluster size, center energy score, and lowest energy score. Among the thirty best-fit docking models, the model with the largest cluster size was prioritized for further analysis. This model yielded a score of -1375.4 for both the CES and lowest energy consumption, as shown in Table 4. The lead model obtained from the ClusPro 2.0 server was further refined via the REFINEMENT tool on the HADDOCK 2.4 server. The refined model was analyzed for Gibb’s free energy (ΔG) and dissociation constant (K_D), as well as other aspects, including the HADDOCK score, cluster size, RMSD from the overall lowest-energy structure, van der Waals energy, electrostatic energy, desolvation energy, restraint violation energy, buried surface area, and z score, as shown in Table 4. The predicted model of the vaccine-TLR4 complex was selected. The binding energy (ΔG) of this model was -18.7 kcal/mol, and the dissociation constant (K_D) was 2.00 × 10^-14 M. With respect to the van der Waals energy, electrostatic energy, desolvation energy, restraint violation energy, and buried surface area, these parameters contributed to an HADDOCK score of -808.4 +/- 7.9 for the vaccine-TLR4 complex. Additionally, the cluster size of the complex was 20. The root mean square deviation (RMSD) from the overall lowest-energy structure was 0.7 +/- 0.4. Moreover, a z score of 0 showed that this model was exactly the average score within the cluster. Overall, these parameters indicated that the selected model was stable and energetically favorable.

In silico immune stimulation The C-IMMSIM v10.1 server was used to predict the immunization efficacy of the candidate vaccine. The results from stimulation using the candidate vaccine included B cells, helper T cells, cytotoxic T cells, natural killer (NK) cells, macrophages (MAs), dendritic cells (DCs), and certain cytokines. The PADRE-vaccine triggered IgM and IgG production after primary immunization and dramatically elevated the production of IgM, IgG1, and IgG2 after the secondary and tertiary responses, as shown in Fig. 6a and 6c. In line with the elevated antibody titer, total B-cell populations, particularly B isotype IgM and memory B-cell populations, were induced after the first injection (Fig. 6b). The total B cell, IgM, memory B cell, and IgG1 populations were dramatically elevated, while the B isotype IgG2 was also induced after the second and third injections, as shown in Fig. 6b. Noticeably, a decrease in the antigen concentration was detected, as shown in Fig. 5a, and these phenomena resulted from the generation of a memory B-cell population (Fig. 6b). Additionally, Fig. 6c indicates that the plasma B lymphocyte count (isotype: IgM+IgG) was noticeably elevated. The active B-cell population per state increased significantly and became consistently high after immunization, as demonstrated in Fig. 6d. According to Fig. 6e, the total help T-cell population increased significantly after the first, second and third injections. In particular, both memory and nonmemory helper T cells were dramatically elevated. Likewise, the active and resting helper T-cell populations per state could be elevated significantly, as indicated by Fig. 6f. Although the resting cytotoxic T-cell population per state decreased after the first immune response, it dramatically increased over time after the tertiary immune response, as shown in Fig. 6g. Total NK cells, the total MA population per state, especially the resting MA population, and the DC population per state (for both active and resting states) were also generated by immunization, as demonstrated in Fig. 6H, 8I, and 8J. As shown in Fig. 6K, the titer of interferon-γ (IFN-γ) was significantly increased by all three injections. Increased transforming growth factor-β (TGF-β), interleukin 10 (IL10) and interleukin 12 (IL12) levels were observed. An elevated danger level D in the same figure indicates that the risk is extremely low. This represented a favorable reaction to immunization.

Codon optimization and in silico cloning Codon optimization was performed using the Java Codon Adaptation Tool (JCAT) for protein expression in E. coli. The protein sequence of the PADRE-vaccine was reverse-translated to the DNA sequence. The DNA sequence was submitted to the JCAT server for codon optimization. The codon adaptation index (CAI) and GC content of the codon-optimized sequence were 1.0 and 51.11%, respectively (Supplementary Dataset 2). These scores indicated that the optimized nucleotide sequence was within the range of the optimum CAI (0.8 - 1.0) and GC content (30 - 70). Finally, the optimized nucleotide sequence above was in silico cloned by insertion into the pET21a vector digested with the NdeI and XhoI restriction enzymes using SnapGene 7.2.1, resulting in the recombinant plasmid pET-21a(+)-PADRE Mpox vaccine-6xHis tag, as shown in Fig. 7.

NMA via the iMODS server. The iMODS server and GROMACS software were utilized to perform molecular dynamic (MD) simulations. iMODS was used to analyze the vaccine-TLR complexes via normal mode analysis (NMA) to investigate the binding affinity of the vaccine-TLR4 complex and to investigate its stability and dynamic movements over time. The main-chain deformity was determined by measuring the extent of molecular deformation at each residue. High peaks represented the deformable regions of the vaccine-TLR complex, indicating flexible locations, while rigid regions were indicated by low values, as shown in Fig. 8. According to Fig. 8a, the vaccine-TLR4 complex demonstrated low distortion.

B-factor charts, shown in Fig. 8b, indicate the average RMSD values of the vaccine-TLR complex obtained by relating the mobility of the computed complex in NMA to the corresponding PDB field. The eigenvalues of the docked materials, illustrated in Fig. 8c, explained the motion stiffness. The required energy for structural deformation was indicated by the eigenvalue, with a lower eigenvalue representing easier deformation. The eigenvalue of the interaction between the vaccine and TLR4 was 6.239649 X 10^-7. According to Fig. 8d, the graphs of variance, which showed an inverse association with the eigenvalue, are shown. The purple color represents individual variance, while the green color represents cumulative variance for each normal mode of the vaccine-TLR complex. The covariance map of the vaccine-TLR complex illustrated coupling between two molecules in the complex (Fig. 8e). Red represents correlated atomic motions, white represents uncorrelated atomic motion, and blue represents anticorrelated atomic motion in the regions of MD depicted in Fig. 8e. The elastic network map of the vaccine-TLR4 complex shown in Fig. 8f depicts each gray dot as the formation of springs connecting the corresponding pair of atoms. Flexible regions are represented by lighter dots, whereas stiffer regions are indicated by darker dots. Collectively, these results highlighted the stable interaction between the vaccine and TLR4.

MD analysis via GROMACS. To measure the stability of the vaccine-TLR4 complex, the backbone atom RMSD was computed from the initial point in time. According to Fig. 9a, the RMSD of the complex rapidly increased from the beginning to approximately 5 ns. After that, the structure stabilized at approximately 20 ns, within the range of approximately 0.6 to 0.5 nm. However, there was an increase after 20 ns until it became substantially stable after approximately 24 ns, with a range of approximately 0.6-0.8 nm throughout the remaining simulation time. Additionally, the RMSF was analyzed to assess the flexibility of the vaccine-TLR complex. The high fluctuation in the RMSF indicated highly flexible regions of the complex. The peak corresponding to residue 636 was at 2.1589 nm, as shown in Fig. 9b. Several highly fluctuating regions were identified, including the region between residues 409 and 519 (with a range of 0.8203 to 1.287 nm), the region between residues 996 and 1222 (with a range of 0.7031 to 1.2779 nm), the region between residues 628 and 653 (with a range of 1.0358 to 2.1589 nm), and the region between residues 1267 and 1558, with a very wide range of fluctuations between 0.2887 and 1.5996 nm. The last region contained several highly fluctuating residues, such as residue 1267 (with a length of 1.5996 nm), residue 1414 (with a length of 1.4612 nm), and residue 1558 (with a length of 1.5258 nm).

Rg was analyzed to determine the structural compactness of the vaccine-TLR complex. As illustrated in Fig. 9c, the overall Rg was substantially stable throughout the simulation. Furthermore, the stability of the complex could be measured by the number of hydrogen bonds. According to Fig. 9d, the results of the MD analysis illustrated the stability of the number of hydrogen bonds within a very narrow range between 10 and 11 throughout the simulation time.

Finally, the SASA could indicate the surface area of a vaccine-TLR complex exposed to solvent molecules. These results could imply the flexibility and amenability of protein‒protein interactions. As illustrated in Fig. 9e, there was a fluctuation from the beginning of the simulation to approximately 20 ns, with a range of ~760 to ~720 nM². After 20 ns until 34 ns, the SASA became more stable. Then, there was a sudden drop at approximately 36 ns; however, the SASA maintained its stability from ~36 to 42 ns. Eventually, it increased again at a simulation time of ~43 ns, after which the SASA became stable until the end of the simulation time.

Monkeypox is a zoonotic double-stranded DNA virus in the Orthopoxviralgenus. Recently, there has been a global outbreak of monkeypox disease, with 97,962 infected patients and 184 deaths according to the data from the CDC as of June 30^th, 2024. To address the global outbreak, several smallpox vaccines, including ACAM2000, JYNNEOS (MVA-BN), and LC16, have been introduced to offer cross-protection against monkeypox ⁴⁴. This also means that nonspecific vaccines are being used to provide protection against monkeypox infection. A small study using previous data from Africa showed that the smallpox vaccine might offer approximately 85% protection against monkeypox infection ⁴⁵.

The Advisory Committee on Immunization Practices (ACIP) recommends the ACAM2000 live-attenuated vaccinia virus vaccine for specific personnel groups. These groups include laboratory personnel routinely handling human infectious orthopox viruses (monkeypox, cowpox, and variola) and medical personnel caring for vaccinia virus-infected patients ⁴⁶. However, ACAM2000 carries potential safety risks. Studies have reported occurrences of postvaccinal encephalitis, eczema vaccinatum, and progressive vaccinia following vaccination ⁴⁶. Additionally, the presence of live vaccinia virus raises concerns about secondary transmission to both vaccinated individuals and those in close contact and the potential for the development of myocarditis or pericarditis in some recipients ⁴⁷. Moreover, administration of the JYNNEOS vaccine has been associated with various side effects, including induration, chills, headache, sore throat, nausea, redness, myalgia, firmness/tightening, and pain ⁴⁴. A retrospective study conducted in a northwestern US population cohort (Oregon) identified 10 cases of cardiac events following JYNNEOS vaccination between July and October 2022 ⁴⁸. Like JYNNEOS, LC16 is a live-attenuated, nonreplicating vaccine approved for monkeypox prevention in Japan in August 2022 ⁴⁷. Most adverse events were reported as mild to moderate localized or systemic in nature. Notably, clinical trials and cohort studies have not identified serious adverse events such as encephalitis or symptomatic myocarditis ⁴⁴. While preclinical data suggest potential safety for immunocompromised patients and atopic dermatitis patients, widespread vaccination in these populations is not currently recommended due to insufficient clinical data ⁴⁴. The observed limitations of live-attenuated vaccines suggest the need to explore alternative vaccination strategies that offer comparable efficacy with reduced adverse effects.

To address these shortcomings, a multi-epitope vaccine was proposed as a candidate for this study because, compared with traditional vaccines, this novel type of vaccine has a reduced capacity for eliciting pathological adverse effects due to the absence of unwanted components ⁴⁹. These unwanted components could be screened by using immunoinformatic tools. In tests with animal models and early clinical trials, multi-epitope vaccines induced cellular and humoral immune responses and elicited robust immune responses from individual epitopes against tumors and microbial and viral infections ^49,50. In this study, epitopes were recruited from monkeypox viral proteins, as sequences of both structural and nonstructural viral proteins were screened for only strong binding to the CTL, HTL, and LBL epitopes. Only strongly bound CTL epitopes that could induce class I immunogenicity and strongly bound HTLs that could induce interferon-γ were recruited for the vaccine construct. Moreover, the highest antigenic LBL epitopes were also identified. These epitopes were also screened for allergenicity, toxicity, transmembrane regions, and human analogous proteomes. Among the 200 monkeypox viral proteins, epitopes were derived from 8 proteins, namely, A40L, B6R, B9R, C2L, E4R, F8L, H3L, and J3R/J1L. These proteins were found to potentially induce the adaptive immune system and cytokines. In particular, previous studies have shown B6R and H3L to be immunodominant, as they stimulate CTLs, HTL, LBL, and interferon-γ ^26–30. The vaccinia viral proteins B5R and H3L, homologous proteins of the monkeypox viral protein B6R and H3L, respectively⁵¹, were found to be immunodominant^32,33,52,53. F8L is another highly dominant immunogenic protein that was found to trigger both HTL and CTL ^54,55. This protein was also tested in vivo and found to induce CTL in nonhuman primates ⁵⁵. E4R is another monkeypox viral protein that was reported to induce antibodies in vivo in prairie dogs ³¹. Moreover, the monkeypox viral proteins A40L and C2L were not directly mentioned in previous studies. However, the vaccinia viral proteins A38L and K2L are homologous to A40L and C2L, respectively ⁵¹. They were found to be crucial ligands for MHC class II and HTL ^32,33. B9R is homologous to the vaccinia viral protein B8R ⁵¹. This protein was found to be immunodominant in CTL stimulation ^56,57. Finally, J3R/J1L, which are chemokine binding proteins, play critical roles in modulating and preventing apoptosis in host cells to facilitate viral infection ³⁴. The homologous vaccinia viral protein B29R of J3R/J1L was reported to be involved in acute and memory CTL stimulation with recombinant antigen ⁵⁸.

The epitopes selected from the proteins above were investigated for epitope conservation. According to previous studies on vaccine design, there are still no vaccines that can provide broad-spectrum protection against all subclades. Even though the most virulent clade II, with fewer deadly effects, caused a global outbreak in 2022 with subclade IIb ^59–61, highly fatal clade I was observed during the outbreak again from 2023-2024 ^59,61. Therefore, this study aimed to provide broad-spectrum protection from both highly lethal and highly virulent clades because clade I strains are generally considered the deadliest clade with less virulence, whereas clade II strains are highly virulent but less lethal ^59–61. Accordingly, the complete genome of Clade I Zaire-96-l-16 and 7 variants of Clade II, including Clade IIa USA 2003, Clade llb-A Nigeria 2018, and Clade IIb-A.2 USA 2022, Clade IIb-A.1 UK 2018 (UK P2), Clade IIb-A.1.1 USA 2021, Clade IIb-B.1 France 2022 was employed in the IEDB Epitope Conservancy Tool. Almost all the epitopes selected for the vaccine construct exhibited 100% conservancy. Only one HTL epitope, SMVFEYRASTVIKGP, presented a single amino acid substituted from valine to isoleucine. The valine in the Clade I strains was substituted by isoleucine in Clade IIa, Clade llb-A, Clade IIb-A.2, Clade IIb-A.1, Clade IIb-A.1.1, and Clade IIb-B.1. However, valine to isoleucine substitution had a negligible effect on the protein because both valine and isoleucine are categorized as hydrophobic amino acids.

The vaccine construct model constructed from selected epitopes was improved further with adjuvants and linkers. Studies have shown that linkers incorporated into vaccine constructs can enhance vaccine stability ^62,63. These linkers are thought to function by facilitating spatial separation between individual epitopes, thereby minimizing potential steric hindrance and ensuring the independent immunogenicity of each epitope upon vaccination ⁶². Moreover, adjuvants were also fused to the vaccine model to improve the immune response. Moreover, adjuvants can exert immunostimulatory effects through various mechanisms, including antigen depot-mediated promotion of prolonged exposure to the immune system, activation of innate immunity via pathogen recognition receptor (PRR) engagement, immune cell costimulation, immunomodulation, and particularly antigen-presenting cell (APC) maturation ^64,65. The results of the physicochemical property analyses in this study could indicate the improvement of the vaccine model, especially the best selected model with PADRE adjuvants. Compared to other designed vaccine models, the PADRE adjuvant vaccine had the highest antigenicity, highest solubility, and improved stability without being potentially allergenic or toxic.

After that, the PADRE adjuvant vaccine model was further subjected to molecular docking and MD with TLR4 because this TLR4 plays a crucial role in antiviral infection ^66,67. The TLR4 signaling complex might physically and hydrophobically interact with the hydrophobic pocket of a coreceptor called MD-2 and hydrophobic fusion peptides on viral proteins ⁶⁶. Another potential factor is glycosylation, as all TLR4-activating viral proteins undergo glycosylation ⁶⁶. Moreover, the way viral proteins assemble into units (oligomeric state) might also play a role in the interaction with the TLR4 signaling complex. ⁶⁶. Therefore, the TLR4 pdb file was retrieved from the Protein Data Bank (PDB ID: 4G8A); however, some missing residues were indicated by the pdb file information. Consequently, Modeler10.4 ^68,69 and EMBOSS NEEDLE ⁷⁰ were used to perform pdb file improvements before docking and MD.

The results of molecular docking with the ClusPro 2.0 server ⁷¹ demonstrated a favorable score of -1375.4 for both the center energy score and lowest energy. Similarly, the binding energy and dissociation constant determined by the HADDOCK 2.4 server ^72–74 were favorable, with a binding energy (ΔG) of -18.7 kcal/mol and K_D of 2.00 × 10^-14 M. Nevertheless, these docking servers have limitations. With rigid-body estimation, only moderate molecular conformations were allowed upon binding ⁷⁵. In fact, an important aspect of correctly approximating binding geometries is flexibility ⁷⁶ Additionally, a reliable approximation of key thermodynamic observations (such as the binding free energy) is not possible ⁷⁶. Furthermore, only a rigid picture of the binding process is produced, leading to a lack of approximation of kinetic quantities ⁷⁶. To address these limitations, the iMOD server ⁷⁷ was used to analyze the rigidity and deformability of the vaccine-TLR4 complex, which indicated its stability. In addition, GROMACS software ^78,79 was used to analyze the stability and dynamics of the complex structure within a timeframe of 50 ns and under more realistic conditions of solvation, neutralization (with Na and Cl ions), 310 K temperature and 1 atm pressure. Moreover, immunization simulations performed with the C-ImmSim server ^80–82 supported the idea that the PADRE-vaccine has the potential to bind effectively to TLR4 because the results of immunization indicated both innate and adaptive immune responses.

In vitro and in vivo studies are indispensable for validating this broad-spectrum multi-epitope vaccine design. The critical steps include protein expression in E. coli to confirm its expression as predicted by its physicochemical properties, as well as in vivo studies in model organisms. Furthermore, the potential of this vaccine design can be significantly enhanced by exploring advanced delivery systems such as mRNAs and outer-membrane vesicles (OMVs). These additional studies will not only affirm the efficacy and safety of the vaccine but also open new avenues for innovative approaches in vaccine development, offering a robust defense against diverse and evolving viral threats.

Retrieval of monkeypox viral proteins. The protein sequences of the monkeypox virus Clade I (Congo/Central Africa) were retrieved from the GenBank database with the accession number DQ011154.1. This clade has the highest lethality and causes severe symptoms compared to other virulent clades. ^59,60. Although clade IIb caused the global pandemic in 2022, it had a lower mortality rate than did clade I ^59,60. However, infected patients in clade I were recently observed in the Democratic Republic of the Congo between 2023 and 2024 ^59,61,83. According to the lethality rate and reemergence of Clade I, the MPXV strain Congo_2003_358 was utilized as a benchmark for the collection of MPX proteins.

Out of the 200 protein candidates, 24 proteins, including structural and nonstructural proteins retrieved from the complete genome, were selected based on antigenicity, allergenicity, and toxicity. The VaxiJen v2.0 ^84–86 server (http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html) was used to screen for antigenic viral proteins. The accuracy of the VaxiJen test for determining the protective antigen distribution of a viral protein ranges from 70 to 89% ⁸⁵. Additionally, AllergenFP v1.0 (https://ddg-pharmfac.net/AllergenFP/) was used to predict the allergenicity of proteins ⁸⁷. Only proteins with an antigenic probability score greater than 0.4 and no allergies were further utilized for epitope prediction.

Prediction of cytotoxic T lymphocytes (CTLs), helper T lymphocytes (HTLs), and linear B lymphocytes (LBLs).The cytotoxic T-lymphocyte (CTL) epitopes of the selected MPXV proteins were predicted by using both NetMHCpan 4.1 BA and NetMHCpan 4.1 EL (https://services.healthtech.dtu.dk/services/NetMHCpan-4.1/) as prediction methods ⁸⁸. The HLA alleles, which included HLA-A0101, HLA-A0201, HLA-A0301, HLA-A2402, HLA-A2601, HLA-B0702, HLA-B0801, HLA-B2705, HLA-B3901, HLA-B4001, and HLA-B5801, were selected for screening of CTL epitopes. Since the vast majority of peptides presented by MHC class I include 9 amino acids, the peptide length was set at 9 amino acids. Additionally, only CTL epitopes demonstrating strong binding results were selected. The ability of the predicted epitopes with strong binding potential to trigger immunogenicity was further investigated via the Class-I Immunogenicity tool in the IEDB (http://tools.iedb.org/immunogenicity/) ⁸⁹ before further assessment.

Similarly, the helper T-lymphocyte (HTL) epitopes of each viral protein were predicted by NetMHCpanII 4.1 BA and NetMHCpanII 4.1 EL (https://services.healthtech.dtu.dk/services/NetMHCIIpan-4.1/) ⁹⁰. DRB1_0101, DRB1_0301, DRB1_0401, DRB1_0701, DRB1_0801, DRB1_0901, DRB1_1001, DRB1_1101, DRB1_1201, DRB1_1301, DRB1_1501, and DRB1_1602 were selected as targeted HLA alleles to screen for HTL epitopes. The length of the epitopes was set to 15 amino acids by default. Only the epitopes with strong binding results were selected. The interferon-γ activation potential of the HTL epitopes was subsequently predicted by IFNepitope (http://crdd.osdd.net/raghava/ifnepitope/predict.php) ⁹¹

Furthermore, the linear B-lymphocyte (LBL) epitopes of monkeypox proteins were predicted using the Bepipred Linear Epitope Prediction 2.0 server (https://services.healthtech.dtu.dk/services/BepiPred-2.0/) with the default threshold of 0.5 ⁹². All the predicted epitopes (CTL, HTL, and LBL) were validated for allergenicity, antigenicity, and toxicity screening.

Effects of CTL, HTL, and B-cell epitopes on allergenicity, antigenicity, and toxicity parametersThe allergenicity of all the selected epitopes of HTL, CTL and LBL was determined using AllergenFP v1.0 ⁸⁷. The Vaxijen v2.0 server was used to evaluate antigenicity with a threshold of 0.4 ^84–86. The ToxinPred2 server (https://webs.iiitd.edu.in/raghava/toxinpred2/) was used for toxicity prediction, with an SVM (Swiss-Prot)-based prediction method ⁹³. Only peptides demonstrating nonallergenic, antigenic, or nontoxic properties were selected for analysis of epitope conservation.

Screening of transmembrane regionsBefore vaccine construction, candidate epitopes were predicted for transmembrane regions. The transmembrane regions of all the selected epitopes were predicted by DeepTMHMM (https://dtu.biolib.com/DeepTMHMM) ⁹⁴. The predicted transmembrane segments were omitted from the vaccine construction to avoid low yields or unstable protein expression due to hydrophobic effects ^35–37.

Epitope conservation for the broad-spectrum coverage of vaccines across different clades To generate a broad-spectrum vaccine against various clades of the monkeypox virus, all the selected CTL, HTL, and LBL epitopes were investigated for epitope conservation across seven monkeypox virus clades, namely, clade I (Zaire-96-l-16), clade IIa (USA 2003), clade llb-A (Nigeria 2018), clade IIb-A.2 (USA 2022), clade IIb-A.1 (UK 2018 P2), clade IIb-A.1 (USA 2021), and clade IIb-B.1 (France 2022), using the IEDB Epitope Conservancy Tool (http://tools.iedb.org/conservancy/) ⁹⁵. Only highly conserved epitopes demonstrating percent identity higher than 90% were retrieved for vaccine construction.

Autoimmune screening by homology analysis of epitopes with human proteomes. To avoid the possibility of inducing autoimmune diseases or cross-reactivity in the host, all the selected CTL, HTL, and LBL epitopes obtained in section 2.5 were blasted against the Homo sapiens proteome (Taxonomy ID: 9606) via the Peptide Search server (https://www.uniprot.org/peptide-search) ⁹⁶. Only epitopes exhibiting nonhomologous identity to the human proteome were subsequently selected for vaccine construction.

Population coverage. To quantify the global impact of the vaccine, the IEDB population coverage tool (http://tools.iedb.org/tools/population/iedb_input) was used to analyze the worldwide distribution of HLA alleles targeted by our selected MHC class I and II epitopes ⁹⁷. The HLA allele genotypic frequencies from the Allele Frequency database (http://www.allelefrequencies.net/) were used to provide information on genetic variations in human populations ⁹⁸. This database offers data for more than 115 countries and categorizes 21 ethnicities within 16 geographic regions ⁹⁸. It also allows users to define custom populations with specific allele frequencies. The program calculates population coverage for various groups and generates an average value.

Construction of the multi-epitope subunit vaccine. The multi-epitope vaccines were constructed by combining the final epitopes of CTL, HTL, and LBL with separate linkers (AAY, HEYGAEALERAG, GPGPG, and KK). To enhance immunogenicity and prolong innate and adaptive immunity ^38,39, a peptide adjuvant was fused to the multi-epitope vaccine. The five peptide adjuvants used in this study included PADRE, beta-defensin 3, 50S ribosomal protein L7/12, RS-09, and the cholera toxin B subunit (CTB). Each adjuvant was N-terminally fused to a multi-epitope vaccine, yielding six multi-epitope vaccine constructs.

Physicochemical, antigenicity, allergenicity, toxicity, and solubility of the vaccine. The physiochemical properties of each multi-epitope vaccine construct, including molecular weight, grand average hydropathicity (GRAVY), theoretical isoelectric point (pI), instability index, and half-life, were evaluated using the ProtParam tool of the ExPASy database server (https://web.ExPASy.org/protparam/) ⁴⁰. The antigenicity and toxicity of the designed vaccines were determined using AllergenFP v. 1.0 (http://ddg-pharmfac.net/AllergenFP/) ⁸⁷ and the ToxinPred2 server (https://webs.iiitd.edu.in/raghava/toxinpred2/) ⁹³, respectively. Allergenicity was identified by VaxiJen v2.0 (http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html) ^84–86. The SOLpro server (http://scratch.proteomics.ics.uci.edu/) was used to predict the solubility of the designed vaccines via an SVM-based method ⁴². Additionally, the SignalP6.0 server (https://services.healthtech.dtu.dk/services/SignalP-6.0/) was used to identify signal peptide regions of the recombinant proteins ⁹⁹.

Secondary prediction, tertiary structure predictions, refinement, and validationThe secondary structure features (α-helixes, β-sheets, and random coils) of the selected vaccine constructs were predicted using the PSIPRED server (http://bioinf.cs.ucl.ac.uk/psipred/) ¹⁰⁰. The tertiary structure of the selected vaccine was predicted using AlphaFold2 by implementing a fast homology search of MMseqs2 with AlphaFold2 ¹⁰¹. Following the initial three-dimensional (3D) modeling, the primary 3D models with the highest scores were further refined by submission to the GalaxyRefine module on the GalaxyWEB server (https://galaxy.seoklab.org/) ^102,103. This tool optimizes unreliable loops and terminal regions through an optimization-based refinement method and generates five refined models from each initial structure ¹⁰⁴. The model with the lowest MolProbity score was selected as the most well defined model for each vaccine construct. These chosen models were further evaluated for quality validation using ERRAT and PROCHECK from the SAVES v6.0 (https://saves.mbi.ucla.edu/) and ProSA-Web (https://prosa.services.came.sbg.ac.at/prosa.php) datasets.

The ERRAT program evaluated overall structural integrity by quantifying nonbonded interactions between neighboring atoms ¹⁰⁵. Models with an overall quality factor exceeding 85 were considered reliable. The Ramachandran plot generated by the PROCHECK tool was used to evaluate the stereochemical accuracy of the protein backbone, with more residues located in favored regions indicating increased model accuracy ^106–109. Finally, through z-score prediction, the ProSA-Web server provided an overall quality assessment in which positive values suggested potential structural issues ^110,111.

Molecular docking of Toll-like receptor 4 (TLR4) and the lead vaccine construct.

To determine whether the lead vaccine can trigger an immune response via TLR4, molecular docking evaluations between the lead vaccine and TLR4 (PDB ID: 4G8A) were performed to predict the occurrence of persistent interactions between the complexes ¹¹².

To bridge the gaps caused by missing residues in the retrieved PDB file of TLR4, MODELLER10.4 was used to locate missing amino acid residues prior to molecular docking analysis to obtain the most accurate outcomes ^68,69. After the PDB file was improved, the ClusPro 2.0 server (https://cluspro.org/login.php) was used to measure the binding energy ⁷¹ between the lead multi-epitope vaccine and the TLR4 receptor. Thirty model structures were generated for the docking scenario, and the ten models demonstrating the largest clusters were prioritized for further analysis.

To enhance structural precision, the top-ranked model obtained from the ClusPro 2.0 server was further refined using the HADDOCK 2.4 server (https://wenmr.science.uu.nl/haddock2.4/) ^72,113. This server optimizes complex structures by strategically rearranging side chains within restricted interfaces and employing soft rigid-body refinement techniques^72,113. Subsequently, the binding energies of the refined docked models were assessed using the PRODIGY server (https://bianca.science.uu.nl/prodigy/), which provided quantitative insights into the binding affinities of the peptides ^72–74. PyMOL was used to visualize the structures of the figures.

In silico immune stimulationThe selected vaccine was analyzed via the C-IMMSIM tool (https://kraken.iac.rm.cnr.it/C-IMMSIM/) to predict the immune efficacy of our designed vaccine ^80–82. The C-IMMSIM server can identify epitopes through position-specific scoring matrix (PSSM) approaches to simulate immune interactions ^80–82. For the immune simulation, the simulation time steps were set to 1050, where the 1-time step equals 8 h. The time step of each injection point was set at 1, 84, or 168. The dose of each antigen injection was set at 1,000, and the time interval between each injection was four weeks, which is the recommended interval for vaccination.

Codon optimization and in silico cloning The vaccine construct was subjected to codon optimization by using the Java Codon Adaptation Tool (JCAT) (http://jcat.de/) ¹¹⁴ with E. coli strain K12 as the organism. After codon optimization, the vaccine sequence was inserted into the expression vector pET21a (+), which was digested with NdeI and XhoI using Snap Gene 7.2.1.

Molecular dynamics (MD) simulation The best docking model obtained from molecular docking analysis was analyzed by the iMODS server to predict the stability of the vaccine-TLR complex based on normal mode analysis (NMA) to predict rigidity and deformability⁷⁷. The iMODs server provides information on deformability graph, B-factor graph (RMSD values), eigenvalues, variance plot, covariance map, and elastic network map. Additionally, the structural dynamics and stability of the docked vaccine-TLR complex were analyzed by using GROMACS v2024 software ^78,79. The CHARMM36 forcefield ^115,116 was adopted to create the topology file for MD analysis. The TIP3P water model ^116–118 was also utilized for solvation of the vaccine-TLR complex. Afterwards, Na and Cl ions were added to neutralize the system, followed by minimizing the system energy. The NVT of each system was performed at 310 K, and the modified Berendsen thermostat ¹¹⁹ was used to perform temperature equilibration. A Parrinello–Rahman barostat ¹²⁰ was also employed. The pressure was adjusted to 1 atm for the NPT simulation with a time constant of 2 ps. Then, the Leapfrog algorithm was used to integrate the equations of motion with a time step of 2.0 fs.

The vaccine-TLR complex system was run for a 50 ns simulation under constant temperature and pressure. The LINCS algorithm ¹²¹ was used for the constraint of hydrogen bonds. The particle mesh Ewald (PME) method ¹²² was used to analyze the long-range electrostatic energy. Additionally, short-range Coulomb electrostatic and van der Waals cutoffs of 1.0 nm were used to perform MD. The results of the MD analysis performed with GROMACS software included the root mean square deviation (RMSD), root mean square fluctuations (RMSF), radius of gyration (Rg), hydrogen bonds, and total solvent accessible surface area (SASA) of the system.

Author contributions. J.P. designed all the research, performed all the research, analyzed the data, and wrote the paper. P.L. and P.J. performed three-dimensional structure prediction, structural refinement, in silico cloning, and data analysis. T.J. predicted the physicochemical properties. B.M. conceptualized the project, designed the research, analyzed the data, and wrote the paper.

Acknowledgments. We thank Dr. Charnon Pattiyanon, Jarukit Suchat, Yan Trostin, and Raweeroj Thongdee from CMKL University for kindly providing the Apex server used in the molecular dynamics simulation in this research.

Additional Information - competing financial interests. The authors declare no competing financial interests.

Data availability statement

All data generated or analyzed in this study are provided within the manuscript and supplementary information files.

Das, T. et al. Efficacy of smallpox approved tecovirimat (Tpoxx) drug against Monkeypox: current update and futuristic prospects. Int J Surg 109, 1528–1530 (2023).
Yang, Z.-S. et al. The first case of monkeypox virus infection detected in Taiwan: awareness and preparation. International Journal of Infectious Diseases 122, 991–995 (2022).
Damon, I. K. Status of human monkeypox: clinical disease, epidemiology and research. Vaccine 29, D54–D59 (2011).
Huang, Y., Mu, L. & Wang, W. Monkeypox: epidemiology, pathogenesis, treatment and prevention. Sig Transduct Target Ther 7, 1–22 (2022).
Petersen, E. et al. Human Monkeypox: Epidemiologic and Clinical Characteristics, Diagnosis, and Prevention. Infect Dis Clin North Am 33, 1027–1043 (2019).
Charniga, K. et al. Estimating the Incubation Period of Monkeypox Virus during the 2022 Multi-National Outbreak. http://medrxiv.org/lookup/doi/10.1101/2022.06.22.22276713 (2022) doi:10.1101/2022.06.22.22276713.
Shamim, M. A. et al. The use of antivirals in the treatment of human monkeypox outbreaks: a systematic review. International Journal of Infectious Diseases 127, 150–161 (2023).
Sah, R. et al. Monkeypox (Mpox) vaccines and their side effects: the other side of the coin. Int J Surg 109, 215–217 (2023).
Sah, R. et al. FDA’s authorized “JYNNEOS” vaccine for counteracting monkeypox global public health emergency; an update – Correspondence. Int J Surg 107, 106971 (2022).
Petersen, B. W. et al. Vaccinating against monkeypox in the Democratic Republic of the Congo. Antiviral Res 162, 171–177 (2019).
Titanji, B. K., Tegomoh, B., Nematollahi, S., Konomos, M. & Kulkarni, P. A. Monkeypox: A Contemporary Review for Healthcare Professionals. Open Forum Infectious Diseases 9, ofac310 (2022).
Tan, C., Zhu, F., Pan, P., Wu, A. & Li, C. Development of multiepitope vaccines against the monkeypox virus based on envelope proteins using immunoinformatics approaches. Frontiers in Immunology 14, (2023).
Kenner, J., Cameron, F., Empig, C., Jobes, D. V. & Gurwith, M. LC16m8: an attenuated smallpox vaccine. Vaccine 24, 7009–7022 (2006).
Bamouh, Z. et al. Investigation of Post Vaccination Reactions of Two Live Attenuated Vaccines against Lumpy Skin Disease of Cattle. Vaccines (Basel) 9, 621 (2021).
Yokote, H. et al. Safety of Attenuated Smallpox Vaccine LC16m8 in Immunodeficient Mice. Clin Vaccine Immunol 21, 1261–1266 (2014).
Gruber, M. F. Current status of monkeypox vaccines. npj Vaccines 7, 1–3 (2022).
Isidro, J. et al. Phylogenomic characterization and signs of microevolution in the 2022 multicountry outbreak of monkeypox virus. Nat Med 28, 1569–1572 (2022).
Sanami, S. et al. In silico design and immunoinformatics analysis of a universal multiepitope vaccine against monkeypox virus. PLOS ONE 18, e0286224 (2023).
Bidmos, F. A., Siris, S., Gladstone, C. A. & Langford, P. R. Bacterial Vaccine Antigen Discovery in the Reverse Vaccinology 2.0 Era: Progress and Challenges. Frontiers in Immunology 9, (2018).
Delany, I., Rappuoli, R. & Seib, K. L. Vaccines, Reverse Vaccinology, and Bacterial Pathogenesis. Cold Spring Harb Perspect Med 3, a012476 (2013).
Aziz, S. et al. Contriving multiepitope vaccine ensemble for monkeypox disease using an immunoinformatics approach. Front Immunol 13, 1004804 (2022).
Amanna, I. J. & Slifka, M. K. Contributions of humoral and cellular immunity to vaccine-induced protection in humans. Virology 411, 206–215 (2011).
Fantin, R. F. & Coelho, C. H. Human antibody responses to circulating monkeypox virus emphasize the need for the first mpox-specific vaccine. The Lancet Microbe 0, (2024).
Abrahim, M., Guterres, A., Neves, P. C. da C. & Bom, A. P. D. A. The emergence of new lineages of the Mpox virus could affect the 2022 outbreak. 2022.07.07.498743 Preprint at https://doi.org/10.1101/2022.07.07.498743 (2023).
Song, H. et al. Characterizing monkeypox virus specific CD8+ T-cell epitopes in rhesus macaques. Virology 447, 181–186 (2013).
Keasey, S. et al. Proteomic Basis of the Antibody Response to Monkeypox Virus Infection Examined in Cynomolgus Macaques and a Comparison to Human Smallpox Vaccination. PLOS ONE 5, e15547 (2010).
Song, S. et al. Analysis of binding and authentic virus-neutralizing activities of immune sera induced by various monkeypox virus antigens. Immunol Res (2024) doi:10.1007/s12026-024-09499-0.
Tang, D. et al. Recombinant proteins A29L, M1R, A35R, and B6R vaccination protects mice from mpox virus challenge. Front. Immunol. 14, (2023).
Xia, A. et al. Cross-reactive antibody response to Monkeypox virus surface proteins in a small proportion of individuals with and without Chinese smallpox vaccination history. BMC Biol 21, 205 (2023).
Zhang, N. et al. Multivalent mRNA vaccines against monkeypox enveloped or mature viron surface antigens demonstrate robust immune response and neutralizing activity. Sci. China Life Sci. 66, 2329–2341 (2023).
Townsend, M. B. et al. Humoral Immunity to Smallpox Vaccines and Monkeypox Virus Challenge: Proteomic Assessment and Clinical Correlations. Journal of Virology 87, 900–911 (2013).
Duke-Cohan, J. S. et al. The heterogeneity of human antibody responses to vaccinia virus revealed through use of focused protein arrays. Vaccine 27, 1154–1165 (2009).
Lorente, E. et al. Proteomics Analysis Reveals That Structural Proteins of the Virion Core and Involved in Gene Expression Are the Main Source for HLA Class II Ligands in Vaccinia Virus-Infected Cells. J. Proteome Res. 18, 900–911 (2019).
Couñago, R. M. et al. Structures of Orf Virus Chemokine Binding Protein in Complex with Host Chemokines Reveal Clues to Broad Binding Specificity. Structure 23, 1199–1213 (2015).
Kaur, J., Kumar, A. & Kaur, J. Strategies for optimization of heterologous protein expression in E. coli: Roadblocks and reinforcements. International Journal of Biological Macromolecules 106, 803–822 (2018).
Lajmi, A. R., Wallace, T. R. & Shin, J. A. Short, Hydrophobic, Alanine-Based Proteins Based on the Basic Region/Leucine Zipper Protein Motif: Overcoming Inclusion Body Formation and Protein Aggregation during Overexpression, Purification, and Renaturation. Protein Expression and Purification 18, 394–403 (2000).
Shortle, D., Stites, W. E. & Meeker, A. K. Contributions of the large hydrophobic amino acids to the stability of staphylococcal nuclease. ACS Publications https://pubs.acs.org/doi/pdf/10.1021/bi00487a007 (2002) doi:10.1021/bi00487a007.
Crofts, K. F. et al. An Analysis of Linker-Dependent Effects on the APC Activation and In Vivo Immunogenicity of an R848-Conjugated Influenza Vaccine. Vaccines 11, 1261 (2023).
Ryu, K. A., Slowinska, K., Moore, T. & Esser-Kahn, A. Immune Response Modulation of Conjugated Agonists with Changing Linker Length. ACS Chem. Biol. 11, 3347–3352 (2016).
Gasteiger, E. et al. Protein Identification and Analysis Tools on the ExPASy Server. in The Proteomics Protocols Handbook (ed. Walker, J. M.) 571–607 (Humana Press, Totowa, NJ, 2005). doi:10.1385/1-59259-890-0:571.
Kyte, J. & Doolittle, R. F. A simple method for displaying the hydropathic character of a protein. J Mol Biol 157, 105–132 (1982).
Magnan, C. N., Randall, A. & Baldi, P. SOLpro: accurate sequence-based prediction of protein solubility. Bioinformatics 25, 2200–2207 (2009).
Ikai, A. Thermostability and aliphatic index of globular proteins. J Biochem 88, 1895–1898 (1980).
Sah, R. et al. Monkeypox (Mpox) vaccines and their side effects: the other side of the coin. Int J Surg 109, 215–217 (2023).
McCarthy, M. W. Therapeutic strategies to address monkeypox. Expert Rev Anti Infect Ther 20, 1249–1252 (2022).
Petersen, B. W., Harms, T. J., Reynolds, M. G. & Harrison, L. H. Use of Vaccinia Virus Smallpox Vaccine in Laboratory and Health Care Personnel at Risk for Occupational Exposure to Orthopoxviruses - Recommendations of the Advisory Committee on Immunization Practices (ACIP), 2015. MMWR Morb Mortal Wkly Rep 65, 257–262 (2016).
Gruber, M. F. Current status of monkeypox vaccines. npj Vaccines 7, 1–3 (2022).
Sharff, K. A., Tandy, T. K., Lewis, P. F. & Johnson, E. S. Cardiac events following JYNNEOS vaccination for prevention of Mpox. Vaccine 41, 3410–3412 (2023).
Zhang, L. Multiepitope vaccines: a promising strategy against tumors and viral infections. Cell Mol Immunol 15, 182–184 (2018).
Lennerz, V. et al. Immunologic response to the survivin-derived multiepitope vaccine EMD640744 in patients with advanced solid tumors. Cancer Immunol Immunother 63, 381–394 (2014).
Likos, A. M. et al. A tale of two clades: monkeypox viruses. J Gen Virol 86, 2661–2672 (2005).
Davies, D. H. et al. Vaccinia virus H3L envelope protein is a major target of neutralizing antibodies in humans and elicits protection against lethal challenge in mice. J Virol 79, 11724–11733 (2005).
Drexler, I. et al. Identification of vaccinia virus epitope-specific HLA-A*0201-restricted T cells and comparative analysis of smallpox vaccines. Proc Natl Acad Sci U S A 100, 217–222 (2003).
Grifoni, A. et al. Defining antigen targets to dissect vaccinia virus and monkeypox virus-specific T-cell responses in humans. Cell Host & Microbe 30, 1662-1670.e4 (2022).
Song, H. et al. Characterizing monkeypox virus specific CD8+ T-cell epitopes in rhesus macaques. Virology 447, 181–186 (2013).
Cao, D. et al. Redirecting anti-Vaccinia virus T-cell immunity for cancer treatment by AAV-mediated delivery of the VV B8R gene. Molecular Therapy - Oncolytics 25, 264–275 (2022).
Wang, Y., Flesch, I. E. A. & Tscharke, D. C. Vaccinia Virus CD8+ T-Cell Dominance Hierarchies Cannot Be Altered by Prior Immunization with Individual Peptides. Journal of Virology 83, 9008–9012 (2009).
Baur, K. et al. Immediate-Early Expression of a Recombinant Antigen by Modified Vaccinia Virus Ankara Breaks the Immunodominance of Strong Vector-Specific B8R Antigen in Acute and Memory CD8 T-Cell Responses. Journal of Virology 84, 8743–8752 (2010).
Desai, A. N. et al. Implications of the 2023-2024 MPXV clade I outbreak in the Democratic Republic of Congo to global public health. Clin Microbiol Infect S1198-743X(24)00213–1 (2024) doi:10.1016/j.cmi.2024.04.016.
Kugelman, J. R. et al. Genomic Variability of Monkeypox Virus among Humans, Democratic Republic of the Congo. Emerg Infect Dis 20, 232–239 (2014).
Vakaniaki, E. H. et al. Sustained Human Outbreak of a New MPXV Clade I Lineage in the Eastern Democratic Republic of the Congo. Nat Med 1–1 (2024) doi:10.1038/s41591-024-03130-3.
Chen, X., Zaro, J. L. & Shen, W.-C. Fusion protein linkers: Property, design and functionality. Advanced Drug Delivery Reviews 65, 1357–1369 (2013).
Gooyit, M., Miranda, P. O., Wenthur, C. J., Ducime, A. & Janda, K. D. Influencing Antibody-Mediated Attenuation of Methamphetamine CNS Distribution through Vaccine Linker Design. ACS Chem. Neurosci. 8, 468–472 (2017).
Azmi, F., Ahmad Fuaad, A. A. H., Skwarczynski, M. & Toth, I. Recent progress in adjuvant discovery for peptide-based subunit vaccines. Hum Vaccin Immunother 10, 778–796 (2014).
Nagpal, G., Chaudhary, K., Agrawal, P. & Raghava, G. P. S. Computer-aided prediction of antigen presenting cell modulators for designing peptide-based vaccine adjuvants. J Transl Med 16, 181 (2018).
Halajian, E. A., LeBlanc, E. V., Gee, K. & Colpitts, C. C. Activation of TLR4 by viral glycoproteins: A double-edged sword? Front. Microbiol. 13, (2022).
Lester, S. N. & Li, K. Toll-like receptors in antiviral innate immunity. J Mol Biol 426, 1246–1264 (2014).
Fiser, A., Do, R. K. & Sali, A. Modeling of loops in protein structures. Protein Sci 9, 1753–1773 (2000).
Sali, A. & Blundell, T. L. Comparative protein modeling by satisfaction of spatial restraints. J Mol Biol 234, 779–815 (1993).
Madeira, F. et al. The EMBL-EBI Job Dispatcher sequence analysis tools framework in 2024. Nucleic Acids Res gkae241 (2024) doi:10.1093/nar/gkae241.
Kozakov, D. et al. The ClusPro web server for protein‒protein docking. Nat Protoc 12, 255–278 (2017).
Honorato, R. V. et al. Structural Biology in the Clouds: The WeNMR-EOSC Ecosystem. Front Mol Biosci 8, 729513 (2021).
Vangone, A. & Bonvin, A. M. Contacts-based prediction of binding affinity in protein–protein complexes. eLife 4, e07454.
Xue, L. C., Rodrigues, J. P., Kastritis, P. L., Bonvin, A. M. & Vangone, A. PRODIGY: a web server for predicting the binding affinity of protein‒protein complexes. Bioinformatics 32, 3676–3678 (2016).
Desta, I. T., Porter, K. A., Xia, B., Kozakov, D. & Vajda, S. Performance and Its Limits in Rigid Body Protein‒Protein Docking. Structure 28, 1071-1081.e3 (2020).
Gioia, D., Bertazzo, M., Recanatini, M., Masetti, M. & Cavalli, A. Dynamic Docking: A Paradigm Shift in Computational Drug Discovery. Molecules 22, 2029 (2017).
Lopéz-Blanco, J. R., Garzón, J. I. & Chacón, P. iMod: multipurpose normal mode analysis in internal coordinates. Bioinformatics 27, 2843–2850 (2011).
Abraham, M. J. et al. GROMACS: High performance molecular simulations through multilevel parallelism from laptops to supercomputers. SoftwareX 1–2, 19–25 (2015).
Szilárd, P., Abraham, M. J., Kutzner, C., Hess, B. & Lindahl, E. Tackling Exascale Software Challenges in Molecular Dynamics Simulations with GROMACS. in vol. 8759 3–27 (2015).
Ragone, C. et al. Identification and validation of viral antigens sharing sequence and structural homology with tumor-associated antigens (TAAs). J Immunother Cancer 9, e002694 (2021).
Rapin, N., Lund, O., Bernaschi, M. & Castiglione, F. Computational immunology meets bioinformatics: the use of prediction tools for molecular binding in the simulation of the immune system. PLoS One 5, e9862 (2010).
Stolfi, P. et al. In-silico evaluation of adenoviral COVID-19 vaccination protocols: Assessment of immunological memory up to 6 months after the third dose. Front. Immunol. 13, (2022).
Masirika, L. M. et al. Ongoing mpox outbreak in Kamituga, South Kivu province, associated with monkeypox virus of a novel Clade I sublineage, Democratic Republic of the Congo, 2024. Euro Surveill 29, 2400106 (2024).
Doytchinova, I. A. & Flower, D. R. Identifying candidate subunit vaccines using an alignment-independent method based on principal amino acid properties. Vaccine 25, 856–866 (2007).
Doytchinova, I. A. & Flower, D. R. VaxiJen: a server for prediction of protective antigens, tumor antigens and subunit vaccines. BMC Bioinformatics 8, 4 (2007).
Doytchinova, I. A. & Flower, D. R. Bioinformatic Approach for Identifying Parasite and Fungal Candidate Subunit Vaccines. TOVACJ 1, 22–26 (2008).
Dimitrov, I., Naneva, L., Doytchinova, I. & Bangov, I. AllergenFP: allergenicity prediction by descriptor fingerprints. Bioinformatics 30, 846–851 (2014).
Reynisson, B., Alvarez, B., Paul, S., Peters, B. & Nielsen, M. NetMHCpan-4.1 and NetMHCIIpan-4.0: improved predictions of MHC antigen presentation by concurrent motif deconvolution and integration of MS MHC eluted ligand data. Nucleic Acids Res 48, W449–W454 (2020).
Calis, J. J. A. et al. Properties of MHC class I presented peptides that enhance immunogenicity. PLoS Comput Biol 9, e1003266 (2013).
Reynisson, B. et al. Improved Prediction of MHC II Antigen Presentation through Integration and Motif Deconvolution of Mass Spectrometry MHC Eluted Ligand Data. J Proteome Res 19, 2304–2315 (2020).
Dhanda, S. K., Vir, P. & Raghava, G. P. S. Designing of interferon-gamma inducing MHC class-II binders. Biol Direct 8, 30 (2013).
Jespersen, M. C., Peters, B., Nielsen, M. & Marcatili, P. BepiPred-2.0: improving sequence-based B-cell epitope prediction using conformational epitopes. Nucleic Acids Res 45, W24–W29 (2017).
Gupta, S. et al. In Silico Approach for Predicting Toxicity of Peptides and Proteins. PLOS ONE 8, e73957 (2013).
Hallgren, J. et al. DeepTMHMM predicts alpha and beta transmembrane proteins using deep neural networks. 2022.04.08.487609 Preprint at https://doi.org/10.1101/2022.04.08.487609 (2022).
Bui, H.-H., Sidney, J., Li, W., Fusseder, N. & Sette, A. Development of an epitope conservancy analysis tool to facilitate the design of epitope-based diagnostics and vaccines. BMC Bioinformatics 8, 361 (2007).
Zaru, R., Orchard, S., & UniProt Consortium. UniProt Tools: BLAST, Align, Peptide Search, and ID Mapping. Curr Protoc 3, e697 (2023).
Bui, H.-H. et al. Predicting population coverage of T-cell epitope-based diagnostics and vaccines. BMC Bioinformatics 7, 153 (2006).
Gonzalez-Galarza, F. F. et al. Allele frequency net database (AFND) 2020 update: gold-standard data classification, open access genotype data and new query tools. Nucleic Acids Research 48, D783–D788 (2020).
Teufel, F. et al. SignalP 6.0 predicts all five types of signal peptides using protein language models. Nat Biotechnol 40, 1023–1025 (2022).
Buchan, D. W. A. & Jones, D. T. The PSIPRED Protein Analysis Workbench: 20 years on. Nucleic Acids Res 47, W402–W407 (2019).
Mirdita, M. et al. ColabFold: making protein folding accessible to all. Nat Methods 19, 679–682 (2022).
Heo, L., Park, H. & Seok, C. GalaxyRefine: Protein structure refinement driven by side-chain repacking. Nucleic Acids Res 41, W384-388 (2013).
Lee, G. R., Heo, L. & Seok, C. Effective protein model structure refinement by loop modeling and overall relaxation. Proteins 84 Suppl 1, 293–301 (2016).
Ko, J., Park, H., Heo, L. & Seok, C. GalaxyWEB server for protein structure prediction and refinement. Nucleic Acids Res 40, W294-297 (2012).
Colovos, C. & Yeates, T. O. Verification of protein structures: patterns of nonbonded atomic interactions. Protein Sci 2, 1511–1519 (1993).
Laskowski, R. A., MacArthur, M. W., Moss, D. S. & Thornton, J. M. PROCHECK: a program to check the stereochemical quality of protein structures. Journal of Applied Crystallography 26, 283–291 (1993).
Laskowski, R. A., Rullmannn, J. A., MacArthur, M. W., Kaptein, R. & Thornton, J. M. AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR. J Biomol NMR 8, 477–486 (1996).
Laskowski, R. A., MacArthur, M. W. & Thornton, J. M. PROCHECK : validation of protein-structure coordinates. urn:isbn: F, 684–687 (2012).
Morris, A. L., MacArthur, M. W., Hutchinson, E. G. & Thornton, J. M. Stereochemical quality of protein structure coordinates. Proteins 12, 345–364 (1992).
Sippl, M. J. Recognition of errors in three-dimensional structures of proteins. Proteins 17, 355–362 (1993).
Wiederstein, M. & Sippl, M. J. ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res 35, W407-410 (2007).
Lengauer, T. & Rarey, M. Computational methods for biomolecular docking. Curr Opin Struct Biol 6, 402–406 (1996).
Honorato, R. V. et al. The HADDOCK2.4 web server for integrative modeling of biomolecular complexes. Nat Protoc (2024) doi:10.1038/s41596-024-01011-0.
Grote, A. et al. JCat: a novel tool to adapt codon usage of a target gene to its potential expression host. Nucleic Acids Res 33, W526-531 (2005).
Lopes, P. E. M., Guvench, O. & MacKerell, A. D. Current status of protein force fields for molecular dynamics simulations. Methods Mol Biol 1215, 47–71 (2015).
Mackerell, A. D. Empirical force fields for biological macromolecules: overview and issues. J Comput Chem 25, 1584–1604 (2004).
Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L. Comparison of simple potential functions for simulating liquid water. The Journal of Chemical Physics 79, 926–935 (1983).
Price, D. J. & Brooks, C. L., III. A modified TIP3P water potential for simulation with Ewald summation. The Journal of Chemical Physics 121, 10096–10103 (2004).
Berendsen, H. J. C., Postma, J. P. M., van Gunsteren, W. F., DiNola, A. & Haak, J. R. Molecular dynamics with coupling to an external bath. The Journal of Chemical Physics 81, 3684–3690 (1984).
Parrinello, M. & Rahman, A. Strain fluctuations and elastic constants. The Journal of Chemical Physics 76, 2662–2666 (1982).
Hess, B., Bekker, H., Berendsen, H. J. C. & Fraaije, J. G. E. M. LINCS: A linear constraint solver for molecular simulations. Journal of Computational Chemistry 18, 1463–1472 (1997).
Darden, T., York, D. & Pedersen, L. Particle mesh Ewald: An N⋅log(N) method for Ewald sums in large systems. The Journal of Chemical Physics 98, 10089–10092 (1993).

Table 1. Selected monkeypox proteins based on antigenic and nonallergenic properties.

1a

Structural Protein
Protein Name	Product	Antigenicity	Allergenicity
A14L	IMV membrane protein	0.5019	Non-allergen
A15L	Phosphorylated IMV membrane protein	0.4759	Non-allergen
A15.5 L	IV and IMV membrane protein	0.7480	Non-allergen
A30L	IMV surface protein (Envelope protein)	0.6212	Non-allergen
A40L	CD47-like putative membrane protein	0.4361	Non-allergen
B6R	EEV type-I membrane glycoprotein	0.5786	Non-allergen
B16R	IFN-alpha/beta-receptor-like secreted glycoprotein	0.5407	Non-allergen
B19R	Serine protease inhibitor-like SPI-1 protein	0.7784	Non-allergen
B21R	Putative membrane-associated glycoprotein	0.5246	Non-allergen
C2L	Serine protease inhibitor-likeprotein SPI-3	0.5702	Non-allergen
F7R	Membrane protein	0.4334	Non-allergen
H3L	IMV heparin binding surface protein	0.4538	Non-allergen
L5L	Llate 16 kDa putative membrane protein	0.7559	Non-allergen
M1R	Myristylprotein	0.6339	Non-allergen
M5R	Putative membrane protein	0.4468	Non-allergen

1b

Nonstructural Protein
Protein Name	Product	Antigenicity	Allergenicity
A28L	A-type inclusion protein	0.4602	Non-allergen
D3R	Secreted epidermal growth factor-like protein	0.4947	Non-allergen
D10L	Host-range	0.5965	Non-allergen
E4R	Uracil-DNA glycosylase	0.5759	Non-allergen
F1L	Poly-A polymerase catalytic subunit VP55	0.4419	Non-allergen
F3L	Double-stranded RNA binding protein	0.4127	Non-allergen
F7R	Membrane protein	0.4334	Non-allergen
F8L	DNA polymerase	0.4103	Non-allergen
J3R/J1L	Chemokine-binding protein	0.7563	Non-allergen

Table 2a. Finalized CTL epitopes for multi-epitope vaccine construction.

Protein	CTL epitope	Antigenic score	Allergenicity	Toxicity	Class I Immunogenicity Score	Binding affinity	Human proteome
A40L	RHIWLLCKL	0.5947	NA	NT	0.0734	Strong Binding	NH
B6R	LSCNGETKY	0.8019	NA	NT	0.0100	Strong Binding	NH
B9R	KIGPPTVRL	1.1705	NA	NT	0.0907	Strong Binding	NH
C2L	IYFKGTWQY	0.8395	NA	NT	0.0098	Strong Binding	NH
C2L	TRDPLYIYK	0.4908	NA	NT	0.0919	Strong Binding	NH
F8L	RTIDIDETI	1.0682	NA	NT	0.3232	Strong Binding	NH
	YPPPRYITV	0.4807	NA	NT	0.1672	Strong Binding	NH
	ETIELGERY	1.4370	NA	NT	0.2798	Strong Binding	NH
H3L	QEKRDVVIV	1.4396	NA	NT	0.1551	Strong Binding	NH

NA = Non-allergen

NT = Non-toxin

NH = Non-homologous

Table 2b. Finalized HTL epitopes for multi-epitope vaccine construction.

Protein	HTL epitope	Antigenicity	Allergenicity	Toxicity	IFNg epitope	Binding affinity	Human proteome
B6R	DSGYHSLDPNAVCET	0.9057	NA	NT	Positive	Strong Binding	NH
E4R	TGVIDYKGYNLNIID	1.4293	NA	NT	Positive	Strong Binding	NH
F8L	SMVFEYRASTVIKGP	0.5161	NA	NT	Positive	Strong Binding	NH
F8L	RSLETDLRSEFDSRS	0.9620	NA	NT	Positive	Strong Binding	NH

NA = Non-allergen

NT = Non-toxin

NH = Non-homologous

Table 2c. Finalized LBL epitopes for multi-epitope vaccine construction.

Protein	LBL epitope	Antigenicity	Allergenicity	Toxicity	Human proteome
B6R	LTSTETSFND	1.5197	NA	NT	NH
B9R	GDYKVEEYCTG	0.9854	NA	NT	NH
H3L	KSGGL	2.1517	NA	NT	NH
J3R/J1L	EEEKDSDIKTHPV	0.6156	NA	NT	NH

NA = Non-allergen

NT = Non-toxin

NH = Non-homologous

Table 3. Antigenicity, allergenicity, toxicity and physicochemical properties of the vaccine candidates.

Properties	Adjuvant
Properties	Cholera Toxin B subunit (CTB)	PADRE	RS09	50S Ribosomal L7/L12	Beta-defensin
Antigenicity (Vaxijen)	0.5979	0.6419	0.6340	0.5566	0.6321
Allergenicity (Allergenfp v.1.0)	Non-allergen	Non-allergen	Non-allergen	Non-allergen	Non-allergen
Toxicity (ToxinPred V.2)	Non-Toxin	Non-Toxin	Non-Toxin	Non-Toxin	Toxic
Number of amino acids	375	285	279	402	332
Molecular weight (Da)	41956.23	31673.59	30989.55	43738.24	37236.12
Theoretical isoelectric point (pI)	7.64	7.68	6.32	5.46	8.71
The estimated half-life¹	30 hours	30 hours	30 hours	30 hours	30 hours
The estimated half-life²	>20 hours	>20 hours	>20 hours	>20 hours	>20 hours
The estimated half-life³	>10 hours	>10 hours	>10 hour	>10 hour	>10 hour
Aliphatic index	69.07	68.28	65.88	77.81	35.21
Instability index	35.18	30.79	65.88	28.13	65.6
Grand average of hydropathicity (GRAVY)	-0.616	-0.599	-0.699	-0.427	-0.68
Solubility	0.83	0.93	0.87	0.98	0.72
Signal Peptide	No signal Peptide	No signal peptide	No signal peptide	No signal peptide	No signal peptide

The estimated half-life¹ = mammalian reticulocytes, in vitro

The estimated half-life² = yeast, in vivo

The estimated half-life³ = Escherichia coli, in vivo

Table 4. The docked PADRE-vaccine-TLR4 complex was analyzed by ClusPro 2.0, and refined docking analysis was performed with the HADDOCK 2.4 server.

Tool	Parameter	PADRE-Vaccine-TLR4 complex
Tool	Parameter	PADRE-Vaccine-TLR4 complex	ClusPro2.0	Center	-1375.4
Lowest Energy	-1375.4		ClusPro2.0
HADDOCK 2.4 server	K_D (M) at 25 °C	2.00×10^-14
	ΔG (kcal mol^-1)	-18.7
	HADDOCK score	-808.4 +/- 7.9
	Cluster size	20
	RMSD from the overall lowest-energy structure	0.7 +/- 0.4
	Van der Waals energy	-426.0 +/- 3.7
	Electrostatic energy	-1417.1 +/- 28.0
	Desolvation energy	-99.1 +/- 6.6
	Restraints violation energy	0.0 +/- 0.0
	Buried Surface Area	12977.8 +/- 131.7
	Z Score	0

The authors declare no competing interests.

SupplementaryInformation.docx

Engineering a broad-spectrum vaccine to combat emerging monkeypox virus via immunoinformatic approaches

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Materials and Methods

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1