A next-generation strategy for HIV-1 vaccination: computational design of a multi-epitope subunit vaccine targeting global circulating variants

Despite decades of research, no effective cure or preventive vaccine exists for human immunodeficiency virus type 1 (HIV-1). To address this gap, this study aimed to design a universal multi-epitope subunit vaccine (MESV) targeting conserved and immunogenic epitopes across diverse HIV-1 variants. Publicly available HIV-1 proteome data were analyzed using an immunoinformatics pipeline that integrates curated immunological databases and epitope-prediction servers, resulting in the identification of 44 conserved epitopes from the core proteome, including twelve newly identified relative to existing databases. The MESV was constructed using 14 linear B-cell epitopes and 30 T-cell epitopes restricted by globally conserved human leukocyte antigen (HLA) alleles, achieving 98.63% MHC-I and 99.67% MHC-II global population coverage. Molecular docking analysis revealed high binding scores and favorable energy values for MESV interactions with toll-like receptor 1 (TLR1) and toll-like receptor 2 (TLR2), suggesting stable antigen recognition. After codon optimization, the MESV (889 aa) was expressed and purified using affinity chromatography, yielding an ~95.1 kDa protein confirmed by immunoblot analysis. In silico immunological simulations demonstrated robust humoral and cellular immune responses following three doses, including elevated IgG antibody levels and increased numbers of memory B cells, cytotoxic CD8⁺ T cells, and natural killer cells. Overall, this study highlights the potential of a computationally designed MESV to overcome HIV-1 diversity and induce broad-spectrum immunity. These findings warrant further in vitro and in vivo validation to confirm immunogenicity and protective efficacy, paving the way toward the development of a universal HIV-1 vaccine.

1 Introduction

Human immunodeficiency virus (HIV) belongs to the Lentivirus genus within the Retroviridae family and was first isolated in 1983 (1). HIV has infected >75 million people worldwide, causing a chronic, incurable disease and resulting in approximately 40.4 million deaths due to acquired immunodeficiency syndrome (AIDS) (2). Developing an effective vaccine to prevent HIV infection remains an unmet medical need (3).

HIV is classified into two types, HIV-1 and HIV-2 (4). HIV-1 accounts for the most severe and advanced cases of AIDS, while HIV-2 has a more limited geographic distribution (5). HIV-1 is further divided into two main groups: Major (M) and Outlier (O). The HIV-1 group M is subdivided into nine subtypes: A, B, C, D, F, G, H, J, and K (6). HIV-1 particles contain a nucleoprotein core surrounded by a lipid membrane. The core consists of two identical copies of viral genomic RNA (7), which is approximately 9,050 nucleotides long and encodes three major polyproteins essential for viral structure and function: group-specific antigen (Gag), polymerase (Pol), and envelope (Env) (8). The genome also encodes two regulatory proteins, transactivator of transcription (Tat) and regulator of virion expression (Rev), and four accessory proteins: virion infectivity factor (Vif), viral protein U (Vpu), viral protein R (Vpr), and negative regulatory factor (Nef) (9). The Gag protein, the most abundant structural protein, plays a critical role in forming the viral core. It is cleaved into smaller proteins, including matrix (MA), capsid (CA), and nucleocapsid (NC), which are essential for viral structural integrity (9). The Gag-Pol polyprotein encodes key enzymes, reverse transcriptase, integrase, and protease, that are crucial for viral replication and maturation (10, 11). The Env protein is synthesized as a precursor, gp160, which is cleaved into the gp120 and gp41 subunits. These subunits form viral spikes that bind to CD4 receptors on host cells, facilitating viral entry (12). Among the regulatory proteins, Tat is a key regulator of viral transcription and is relatively conserved across primate lentiviruses (13). Rev is involved in transcriptional regulation and the nuclear export of incompletely or partially spliced messenger RNAs. The accessory protein Nef enhances viral infectivity and pathogenesis (14). Vif is essential for viral replication in non-permissive cells, while Vpr enhances viral infection and replication (15). The transmembrane protein Vpu regulates host membrane protein compartmentalization and promotes CD4 receptor degradation, facilitating viral replication and dissemination (16).

During a viral infection, specific toll-like receptors (TLRs) recognize distinct pathogen-associated molecular patterns (PAMPs). For example, TLR2 (often with TLR1) detects lipoproteins, while TLR3 and TLR7 recognize double-stranded and single-stranded RNA, respectively (17, 18). This recognition initiates host immune responses, including inflammation and the priming of antigen-specific adaptive immunity (19). MHC class I molecules present intracellular pathogen-derived antigens to CD8-positive (CD8⁺) T cells, inducing their differentiation into cytotoxic T lymphocytes that eliminate infected cells (20). This process also generates memory T cells, enabling a rapid and effective immune response upon future exposure (21). Meanwhile, MHC class II molecules present extracellular pathogen-derived antigens to CD4-positive (CD4⁺) T cells. Activated CD4⁺ T cells stimulate cytotoxic T cells to clear viral antigens and support B cells in producing antibodies that neutralize the pathogen (22, 23). B cell activation also leads to memory B cell formation, ensuring a rapid antibody response upon re-exposure to the virus (24).

Vaccines leverage similar immune mechanisms to establish immunity and prevent infection (25). However, despite decades of research, no licensed HIV-1 vaccine has achieved durable protection, primarily due to the virus’s rapid mutation rate and antigenic diversity. Vaccines designed using traditional approaches, such as the HIV-1 vaccine evaluated in the RV144 trial that mainly targeted the V2 loop of gp120, showed modest efficacy against HIV-1 infection, and their protection remained limited in breadth and durability. One strategy to improve vaccine efficacy is to maximize the inclusion of antigenic immunogens that are conserved across different HIV-1 variants, thereby stimulating a broadly neutralizing humoral response alongside an effective cellular immune response (26). Thus, the multi-epitope-based subunit vaccine (MESV) may offer advantages over classical approaches in cost, time efficiency, and immune response optimization (26). Indeed, two studies reported designed vaccine candidates targeting the entire HIV genome using in silico analysis; however, these vaccine candidates lacked sufficient conservation across global strains (27). Using a proteome-wide screen of the HIV-1 proteome, we generated consensus sequences and applied advanced bioinformatics and immunoinformatics tools to identify 12 novel epitopes with strong immunogenic potential. By integrating HLA screening for global population coverage, we ensured that these targets align with the genetic diversity of human populations. Additionally, we cross-validated 32 experimentally reported epitopes from the LANL database. By combining these 32 known epitopes with the 12 novel epitopes identified here, we designed an MESV candidate and validated its interaction with TLRs, along with binding stability, using in silico immune simulations. Simulation analyses further confirmed strong immune responses induced following MESV immunization.

2 Material and methods

2.1 HIV-1 proteins

The HIV-1 group M subtype B isolate strain HXB2 (total length: 11,706 bp), comprising nine protein sequences, was retrieved from UniProtKB (https://www.uniprot.org/) in FASTA format (UniProt proteome ID: UP000002241). Supplementary Figure S1 outlines the key steps involved in the generation of the MESV.

2.2 Antigenicity and physiochemical properties

The VaxiJen 2.0 server (http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html) was used to evaluate the antigenicity of the selected protein sequences, applying a cutoff score of 0.4. The physicochemical properties of individual HIV-1 proteins were analyzed using ProtParam (https://web.expasy.org/protparam/).

2.3 Potential T-cells and linear B-cell epitopes

Potential T-cell epitopes were identified using the Immune Epitope Database and Analysis Resource (IEDB) (https://www.iedb.org/). The NetMHCpan-4.1 (EL and BA) method (2023.09 version) was applied to predict T-cell epitope binding to MHC class I alleles, including relevant human leukocyte antigen (HLA) alleles. Predictions for MHC class II epitopes were performed using the NetMHCpan-4.1 (EL) method and the IEDB-recommended 2.22 method, considering fixed peptide lengths of 9 for MHC-I; 15 for MHC-II (28). To assess the global distribution of HIV-1 recognition across MHC class I and class II alleles, the IEDB population coverage tool was employed (http://tools.iedb.org/population/). For B-cell epitope prediction, the BepiPred Linear Epitope Prediction 2.0 tool was utilized via the IEDB platform, selecting epitopes between 10 and 40 amino acids in length.

2.4 Cross-validation of MESV in HIV-1 variants

The MESV was validated against epitope sequences recorded in the Los Alamos HIV-1 Immunology Database (LANL) (https://www.lanl.gov/) to assess potential binding to CD8⁺ cytotoxic T cells and CD4⁺ T-helper cells. The MESV sequence was aligned with HIV-1 GenBank entries using QuickAlign with a 95% frequency cutoff and further validated by the ALIGN0 program. For multiple sequence alignment (MSA), ClustalW and Jalview (https://www.jalview.org/) were used within Mega11 (https://www.megasoftware.net/). Phylogenetic relationships were analyzed using the neighbor-joining method and visualized with the iTOL platform (https://itol.embl.de/).

2.5 Epitopes selection and MESV construction

Antigenicity analysis of selected epitopes was primarily conducted using VaxiJen 2.0, while allergenicity and toxicity assessments were performed using the AllerTop v.2.0 server (https://ddg-pharmfac.net/AllerTOP/) and ToxinPred, respectively (29). To minimize potential cross-reactivity, epitope sequences were screened using BLASTp (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE=Proteins) for non-human homology.

For epitope selection, multiple sequence alignment was first performed across group M using a conservation score threshold of ≥60 (sequence identity). Conserved regions were then analyzed for epitope prediction. MHC-I–binding epitopes were identified using NetMHCpan-4.1 (EL/BA) with rank values ranging from ≤0.5% to 2%, along with an IEDB class I immunogenicity score >0. MHC-II–binding epitopes were predicted using NetMHCIIpan-4.1 (EL) with rank values ranging from 2% to 10% and with multi-allelic coverage of at least three HLA-II alleles. Linear B-cell epitopes were selected based on default tool thresholds (e.g., ≥0.5). Additional filtering criteria included antigenicity (VaxiJen v2.0, viral model) ≥0.40, negative allergenicity by AllerTOP, non-toxic predictions by ToxinPred, and lack of homology to the human proteome. Epitopes meeting all the above criteria were selected for vaccine development. To enhance immunogenicity, the selected epitopes were incorporated into the MESV construct using protein linkers and adjuvants, including EAAAK, AAY, GPGPG, and AAA, along with the beta-defensin 3 adjuvant (UniProt ID: Q5U7J2) (30).

2.6 Structure prediction and stability analysis

Structural modeling of the MESV was performed using PSIPRED 4.0 (http://bioinf.cs.ucl.ac.uk/psipred/), RaptorX (http://raptorx.uchicago.edu/), and GalaxyWeb (http://galaxy.seoklab.org/). The best model was refined using GalaxyWeb (http://galaxy.seoklab.org/cgi-bin/submit.cgi?type=REFINE). The structure was validated using Ramachandran analysis, ERRAT, and PROVE scores via the UCLA-DOE LAB SAVES v6.1 (https://saves.mbi.ucla.edu/). Additionally, Z-score and knowledge-based energy values were analyzed to confirm the structural integrity of the MESV.

2.7 MESV engineering

To assess potential conformational B-cell epitopes, the MESV was analyzed using the IEDB Ellipro tool (http://tools.iedb.org/ellipro/), which combines antigen geometric features with single-amino-acid epitope propensity to predict discontinuous epitopes (31). To further evaluate structural stability, Disulfide by Design 2.0 (DbD2; http://cptweb.cpt.wayne.edu/DbD2/) was used to predict disulfide bond formation in the MESV 3D structure (32).

2.8 Receptor retrieval and molecular docking

The structures of human Toll-like receptors (TLRs), TLR1, TLR2, and the TLR1–TLR2 heterodimer (PDB ID: 2z7x), were retrieved from the Protein Data Bank (PDB) (https://www.rcsb.org/). Receptor complexes were preprocessed and cleaned using PyMOL, followed by molecular docking using HADDOCK. The CPORT file for docking was generated via HADDOCK 2.4 (https://alcazar.science.uu.nl/services/HADDOCK2.4/). Docking models were ranked based on HADDOCK scores, cluster size, root mean square deviation (RMSD), van der Waals energy, electrostatic energy, desolvation energy, and Z-score (33). Interaction analyses were visualized using LigPlot+ and PyMOL.

2.9 Molecular dynamic simulation

Molecular dynamics (MD) simulations of the MESV–TLR1, MESV–TLR2, and MESV–TLR1/2 heterodimer were conducted using the iMODS server (http://imods.chaconlab.org/), which represents a normal mode analysis. The analysis included B-factor and mobility/deformability assessments, eigenvalue variance and covariance analysis, and elastic network analysis to explore conformational dynamics.

2.10 Vaccine simulation

After MESV preparation and optimization in AutoDockTools, the MESV–TLR complexes (TLR1, TLR2, and the TLR1–TLR2 heterodimer) were solvated in a TIP3P water box containing sodium (Na⁺) and chloride (Cl^-) ions, replicating physiological conditions (34). The simulation box was configured as a dodecahedron, followed by energy minimization (using the steepest descent algorithm for 5,000 steps (nsteps = 500) with a convergence criterion of Fmax = 1,000 KJ mol^-1 nm^-1, which is step-based rather than time-based in GROMACS and the CHARMM36 force field, and a 200 ns molecular dynamics simulation (35). Longrange interactions were managed using van der Waals cut-off settings and a temperature of 300 K, with gen_velseed (–1) for randomization (36). The final MD production step utilized Parrinello–Rahman pressure coupling, with tau_p = 2.0 and ref_p = 1.0 (37). Hydrogen bond analyses were performed using Python, PyMOL, VMD, and GROMACS tools (gmx rms, gmx rmsf, gmx area, and gmx gyrate) (38).

2.11 Immune cells simulation

The strength of the immune response induced by the MESV at different time points was simulated using the C-ImmSim server (https://150.146.2.1/C-IMMSIM/index.php). The simulation parameters were set as follows: a random seed of 12345, a simulation volume of 10, and an immune model duration of 360 days. The total number of simulation steps was set to 1,050, with each step corresponding to 8 hours of simulation time. In this model, booster doses were administered three times: once after the first step, again after 84 steps, and a final booster at step 170 (39).

2.12 In silico cloning and codon optimization

Codon optimization was performed using JCAT (https://www.prodoric.de/JCat/) for expression in E. coli K12. The optimized vaccine sequence was reverse-translated, and its GC content and codon adaptation index (CAI) were calculated. Finally, SnapGene (https://www.snapgene.com/) was used to insert the optimized sequence into the pET28a(+) expression vector, generating the recombinant plasmid.

2.13 Transformation expression and purification of MESV

The gene encoding the MESV with a C-terminal 6×His tag was cloned into a pET-28a vector (IGE Biotechnology, China), which was transformed into Escherichia coli BL21 (DE3) cells (Kangti Life Technology, China). The bacteria were grown in LB medium containing 50 μg/mL kanamycin at 37 °C, and MESV protein expression was induced by the addition of 0.5 mM IPTG (Sangon Biotech, China). The bacteria were harvested and resuspended in lysis buffer containing 20 mM Tris-HCl (pH 8.0), 300 mM NaCl, 10 mM imidazole, and 1 mM PMSF (all from Sangon Biotech) using an ultrasonic processor (SCIENTZ-II, 300 W, 3 s on/5 s off, 60 min). The MESV protein was purified using a Ni–NTA resin column (Smart Lifesciences Biotechnology, China) (40). Expression of the MESV protein was evaluated by SDS-PAGE followed by Coomassie Brilliant Blue staining or immunoblot analysis using an anti-His antibody (Wuhan Sanying Biotechnology, China).

3 Results

3.1 The antigenicity and physiochemical properties of HIV-1 proteins

To select potential targets for vaccine development, we chose the most common strain, HIV-1 group M subtype B (isolate HXB2), to further characterize nine core proteins: Gag, Gag-Pol, Env (gp160), Tat, Nef, Vpu, Vpr, Vif, and Rev (41). The antigenicity of these proteins was predicted using the k-nearest neighbor algorithm (k = 1), based on a training set of 2,427 known allergens from various species and 2,427 non-allergens (30). The ability of each protein to induce an immune response was quantified based on its antigenic value, with values >0.4 considered indicative of strong immunogenic potential. As shown in Supplementary Table S1, eight of the nine proteins exhibited antigenic potential above this threshold, with the exception of Vpr. We additionally analyzed several physicochemical characteristics, including molecular weight, isoelectric point, stability, and amino acid composition. The molecular weights of most proteins ranged from 9 to 97 kDa, and all proteins, except for Nef and Vpu, had isoelectric points >8. The instability index was <42 for five of the nine proteins (Gag-Pol, Nef, Vpu, Vif, and gp160), suggesting that these proteins are relatively stable.

3.2 Mapping of T-cells and B-cells epitopes

To predict potential immune targets, we mapped common human leukocyte antigen (HLA) alleles associated with HIV-1 immune responses. The analysis included MHC class I alleles (e.g., HLA-A, HLA-B) corresponding to the standard HLA reference allele set and MHC class II alleles (e.g., HLA-DP*, HLA-DQ*, and HLA-DR*) (Supplementary Tables S2, S3). Epitope selection followed the recommended methodology, considering peptide lengths of 9 amino acids for MHC-I and 15 amino acids for MHC-II (42, 43). Linear B-cell epitopes (~10–40 amino acids) were predicted using the BepiPred Linear Epitope Prediction 2.0 tool via the IEDB platform. This analysis identified 186 potential epitopes across the nine proteins, which were further assessed for antigenicity, allergenicity, toxicity, and non-human homology. We applied prespecified filters, including conservation across HIV-1 proteomes, predicted antigenicity, non-allergenicity and non-toxicity, strong HLA binding with broad allele coverage, and removal of redundant or overlapping sequences. Using these criteria, we identified 44 epitopes across the following proteins for MHC-I, MHC-II, and B-cell analyses: Gag-Pol (3, 3, 4), Tat (1, 2, 1), Nef (1, 1, 1), Vpu (0, 2, 1), Vpr (1, 3, 1), Vif (1, 2, 1), Rev (1, 1, 1), Gag (1, 1, 2), and gp160 (3, 3, 2) (Table 1).

Table 1

Table 1. MESV epitopes (n = 44), antigenicity and immunogenicity prediction scores, toxicity assessment, and epitope concordance with the LANL HIV epitope database.

3.3 Identification of novel epitopes of HIV-1

The 44 selected epitopes were verified against the Los Alamos National Laboratory (LANL) HIV Immunology Database, which catalogues epitopes for cytotoxic T lymphocytes (CTL/CD8⁺) and T-helper (CD4⁺) cells. These epitopes were examined across diverse populations and subpopulations from South Africa (ZA), the United States (US), the United Kingdom (UK), Switzerland (CH), China (CN), Japan (JP), and Pakistan (PK). Of the 44 epitopes, 32 were found in the LANL immune database, while the remaining 12 epitopes—distributed among Gag-Pol (0, 0, 2), gp160 (1, 2, 2), Nef (0, 1, 0), and Vpu (0, 2, 2) for T-cell MHC class I, MHC class II, and B cells—were identified as novel, having no previous record in the LANL database (Table 1). Using a LANL exact-match rule, we classified epitopes as non-novel if they showed 40%–100% sequence identity (position-independent) or ≥30% overlap with a known epitope; all others were considered newly reported (Supplementary Table S4).

3.4 Conservation and population coverage analysis of the selected epitopes

To determine conservation, amino acid sequences for Gag, Gag-Pol, Env (gp160), Tat, Nef, Vpu, Vpr, Vif, and Rev were retrieved from the UniProtKB proteomes portal for multiple isolates, including HXB2, ARV2/SF2, 89.6 (subtype B), VI850 (subtype F1), 90CF056 (subtype H), 92BR025 and ETH2220 (subtype C), and 93BR020 (subtype F1) (44). sequence alignment (MSA; Supplementary Figure S2) revealed that epitopes on structural and functional proteins (Gag, Gag-Pol, and Nef) were highly conserved, with the highest alignment scores of 0.86, 0.85, and 0.83, respectively (Figure 1A, Supplementary Table S1). By contrast, epitopes on regulatory and accessory proteins (Vpu, Vpr, and Tat) exhibited lower conservation, with MSA scores of 0.64, 0.63, and 0.67, respectively, suggesting increased variability. The Vif and Rev proteins showed moderate conservation, with MSA scores of 0.72 (Supplementary Table S1). Phylogenetic tree analysis indicated a close evolutionary relationship between Gag and Gag-Pol (Figure 1A), suggesting shared selective pressures. By contrast, gp160 epitopes had a maximum pairwise identity of 0.93 (Supplementary Table S1) and clustered separately in the upper region of the phylogenetic tree, suggesting a distinct evolutionary trajectory. Collectively, these analyses indicate high conservation of the selected epitopes among different strains of HIV-1.

Figure 1

Figure 1. Phylogenetic analysis of HIV-1 strain clustering. (A) This phylogram illustrates the formation of HIV-1 strain clusters, with color-coding representing distinct viral proteins and countries of origin. Vertical lines indicate the clustering of HIV-1 subtypes based on reference sequences from the Los Alamos HIV sequence database. Epitopes from viral proteins (Gag, Gag-Pol, Nef, Tat, gp160, Vpr, Vif, Vpu, and Rev) were analyzed and clustered individually. The clustering patterns reveal both homogeneous and heterogeneous distributions of HIV-1 subtype B sequences among the studied countries. (B) Global population coverage of selected immune-recognizing epitopes. Top: Epitopes targeting MHC class I HLA alleles cover 98.63% of the world’s population. Bottom: Epitopes targeting MHC class II HLA alleles cover 99.67% of the global population. The line with circles (–o–) shows how coverage increases as epitopes are combined, while the bars highlight each epitope’s individual contribution. Percentages reflect the proportion of individuals predicted to recognize these epitopes based on common genetic markers (HLA types).

The IEDB population coverage tool revealed striking geographic patterns in the global distribution of HIV-1 epitope recognition. Among the 12 MHC-I–binding cytotoxic T-lymphocyte (CTL) and 18 helper T-lymphocyte (HTL) epitopes paired with their HLA alleles, Europe, North Africa, and North America demonstrated the highest population coverage (>99%) for both MHC class I and II alleles. In contrast, South Africa exhibited the lowest coverage, with 93.41% for MHC-I and 50.3% for MHC-II alleles across all epitopes. Globally, 98.63% of the population was covered by MHC-I alleles and 99.67% by MHC-II alleles (Figure 1B, Table 2). These findings suggest that vaccines incorporating these epitopes could potentially address the global burden of HIV infection.

Table 2

Table 2. Global population coverage of selected HIV-1 immunogenic epitopes based on MHC-I and MHC-II HLA alleles (IEDB data).

3.5 Multi-epitope-based subunit vaccine construction

To construct the MESV, we integrated the 12 novel epitopes with 32 known epitopes (Supplementary Table S4), incorporating protein linkers and adjuvants to enhance vaccine efficacy and promote long-term protection. The EAAAK linker was placed at the beginning and end of the vaccine sequence to facilitate antigen processing and uptake. The β-defensin 3 adjuvant (UniProt ID: Q5U7J2) was added, as it has been shown to stimulate autophagosome formation and enhance antigen presentation (45). To optimize epitope connectivity, the following linkers were used: AAY linker for MHC-I T-cell epitopes; GPGPG linker for MHC-II–binding epitopes; and KK linker for B-cell epitopes. Each epitope was systematically linked for the nine target HIV-1 proteins: Gag-Pol, Tat, Nef, Vpu, Vpr, Vif, Rev, Gag, and gp160. Finally, the AAA linker was used to connect the adjuvant and epitope sequence (Supplementary Figures S3, S4).

3.6 Characterizing and profiling of the MESV

3.6.1 The antigenicity and physiochemical properties of the MESV

We evaluated the designed MESV for antigenicity, allergenicity, toxicity, and physicochemical properties using VaxiJen v2.0 and AllerTOP (Table 3). Physicochemical analysis showed that the MESV has an instability index (II) score of 37.99, indicating that the construct is stable under normal physiological conditions. The aliphatic index score of 80.43 suggests that the vaccine maintains stability across a wide range of temperatures. Furthermore, the grand average of hydropathicity (GRAVY) score of −0.391 indicates that the construct is hydrophilic, suggesting high solubility in an aqueous environment. The antigenic value of 0.72 suggests that the designed vaccine is highly antigenic. These favorable antigenicity and physicochemical properties collectively indicate the potential of the proposed vaccine for further development.

Table 3

Table 3. Physicochemical properties of MESV and evaluation of MESV-TLR1/2 and TLR1–2 heterodimer docking interactions.

3.6.2 Structure prediction, refinement, and validation of the MESV

To better understand the structural properties of the vaccine, we performed hydrophobicity and electrostatic potential analyses. The results revealed a balanced distribution of acidic and basic regions, suggesting potential stability under physiological conditions (Figures 2A–C). We then refined and validated the three-dimensional (3D) structure of the MESV. Ramachandran plot analysis demonstrated that 89.8% of residues were located in the most favored regions, with an additional 9.9% in allowed regions and only 0.3% in disallowed region (Figure 2D). We validated the vaccine structure using PROVE and ERRAT. PROVE analysis confirmed that 0% of protein atoms were outliers, while ERRAT analysis provided a quality factor of 87.14%, further supporting the structural integrity of the vaccine.

Figure 2

Figure 2. Structural characterization, evaluation, and refinement of the MESV. (A) Representation of vaccine candidates based on conserved and antigenic epitopes. Top panel, left: surface model; right: 3D structure. (B) Electrostatic potential map indicating negative (red), neutral or zero (white), and positive (blue) electrostatic potential regions. (C) Vaccine hydrophobicity indicating hydrophobic/lipophilic (yellow), neutral/intermediate hydrophobic (white), and hydrophilic (cyan) regions. (D) Evaluation of allowed and disallowed regions of the structure, including the most favored regions (within the blue dotted line), additional allowed regions (within the orange dotted line), and disallowed regions (indicated by +).

3.6.3 Stability and conformational B-cell epitopes in constructed structure of the MESV

To assess the stability of the MESV, we analyzed discontinuous B-cell epitopes in the 3D conformation. Several discontinuous epitope regions capable of binding and stimulating B-cell responses were identified (Supplementary Table S5, Supplementary Figure S6). We also predicted disulfide bond formation to evaluate the structural stability of the vaccine. Features such as the χ3 angle, energy, and Σ B-factor were assessed. Of the 49 predicted disulfide bonds, only 17 were feasible, while 32 were filtered out based on stability parameters (Supplementary Table S6). Selection of these stable disulfide bonds ensures higher structural integrity and enhanced biological stability of the vaccine.

3.7 Molecular docking and simulation analysis

We next performed molecular docking analyses to evaluate the interaction of the MESV with TLR1, TLR2, and the TLR1–TLR2 heterodimer. The docking simulations revealed a range of interactions, including hydrophobic contacts, hydrogen bonds, salt bridges, and metal ion coordination, as visualized using PyMOL and LigPlot+ (Figure 3A). The final docked structures were analyzed by evaluating interaction energies and buried surface area, averaged across the structure. Clustering was based on pairwise backbone RMSD (33). The docking scores and RMSD fluctuations were −91.1 ± 7.1 and 16.8 ± 0.5 for the MESV–TLR1 complex, −175.8 ± 10.2 and 1.5 ± 0.4 for the MESV–TLR2 complex, and −99.2 ± 6.8 and 18.1 ± 0.8 for the MESV–TLR1–TLR2 heterodimer complex, respectively. Although the refined binding energy values were −15.5 kcal/mol for MESV–TLR1 and −12.9 kcal/mol for both MESV–TLR2 and MESV–TLR1–TLR2 (Table 3), the docking score is the primary metric for ranking initial binding affinity and clearly identifies TLR2 as the preferred receptor. In addition, the MESV–TLR2 complex formed substantially more hydrogen bonds during MD simulations than the MESV–TLR1 complex, supporting enhanced interface stabilization upon dynamic relaxation. Collectively, these results demonstrate that the MESV preferentially interacts with TLR2 and achieves a more stable binding configuration during simulation.

Figure 3

Figure 3. Molecular interactions and Conformational analysis of MESV with TLR1, TLR2 and TLR1–2 heterodimer. (A) Left panel: Surface representation of the MESV (cyan) in complex with TLR1 (blue) and TLR2 (purple/red). Right panel: TLR1, TLR2, and TLR1–TLR2 heterodimer interactions with the MESV visualized using PyMOL. (B) Main-chain flexibility across the complex, with high deformability regions indicated as potential chain hinges. (C) Eigenvalue plots illustrating the calculated eigenvalues associated with each normal mode, representing motion stiffness and the energy required to deform the structure for both receptors. (D) The matrix indicating coupling between pairs of residues through correlated (red), uncorrelated (white), or anti-correlated (blue) motions.

To further evaluate structural stability, we analyzed the properties of the MESV–TLR complexes through computational simulations, including elastic network modeling, eigenvalue fluctuations, and B-factor-derived mobility assessments. A single hinge region was identified near a specific but non-intersecting structural point, suggesting structural rigidity in a biological environment. Additionally, the MESV–TLR complexes exhibited relatively high eigenvalues, confirming overall structural stability (Figures 3B–D).

MD simulations were performed to explore conformational changes and structural stability over 200 ns (Figures 4A–I). The MESV constructs were initially compact (for all three complexes) at 0 ns but underwent noticeable rearrangement during the simulation, ultimately stabilizing at the end of 200 ns (Figure 4A). Despite these changes, the overall stability of the MESV–TLR complexes remained high, particularly in the TLR2-bound state (Figures 4B–D).

Figure 4

Figure 4. Molecular dynamics simulation analysis of the MESV alone and in complex with TLR1, TLR2 and TLR1–2 heterodimer. (A). Structural evolution of the MESV (cyan) alone, shown at 0 ns and 200 ns of the simulation. (B-D). MESV (cyan) in complex with TLR1 (blue) and TLR2 (magenta). (E). RMSD plot showing that the TLR2–vaccine complex (magenta) exhibits the least structural deviation, followed by the TLR1–vaccine complex (blue), the TLR1–TLR2–vaccine complex (red), and the unbound vaccine (cyan). (F). RMSF plot illustrating reduced fluctuations across most residues for the TLR-bound states compared with the TLR1–TLR2–vaccine complex, suggesting a stabilized conformation upon binding. (G). Radius of gyration plot showing that the TLR2–vaccine complex has the most compact structure, followed by the TLR1–vaccine and TLR1–TLR2–vaccine complexes. (H). Solvent-accessible surface area (SASA) plot revealing that the TLR2–vaccine complex has the lowest exposed surface area, reflecting tighter packing. (I). Number of hydrogen bonds over time, indicating that the TLR2–vaccine complex forms significantly more hydrogen bonds.

RMSD analysis demonstrated that the MESV–TLR2 complex achieved the most stable trajectory, reaching equilibrium earlier and maintaining a consistent RMSD plateau around ~0.8–1.0 nm, whereas the MESV–TLR1 complex exhibited slightly higher deviations (Figure 4E). The TLR1–TLR2 heterodimer complex showed moderate fluctuations but remained within an acceptable stability range throughout the simulation. RMSF analysis supported these observations, as the MESV–TLR2 complex exhibited lower residue-level flexibility, especially at the binding interface. In contrast, the MESV–TLR1 complex displayed higher fluctuations across several loop regions, indicating reduced stabilization upon MESV binding (Figure 4F). Radius of gyration (Rg) analysis indicated that the MESV–TLR2 complex maintained a relatively compact conformation, with minor fluctuations and stabilization around ~2.4–2.5 nm, whereas the TLR1-bound complexes showed slightly higher Rg values (Figure 4G). This reflects tighter molecular packing and improved structural organization in the TLR2-bound state.

Consistently, solvent-accessible surface area (SASA) analysis showed that the MESV–TLR2 complex maintained the lowest solvent-accessible surface area, indicating reduced solvent exposure and enhanced structural packing compared with the MESV–TLR1 or heterodimer complexes (Figure 4H). Lower SASA values generally correspond to more energetically favorable buried interfaces. Hydrogen bond analysis further revealed that the MESV–TLR2 complex formed the highest number of intermolecular hydrogen bonds, maintaining approximately 25–45 stable hydrogen bonds throughout the trajectory. In contrast, the MESV–TLR1 complex exhibited fewer (8–20) and more variable hydrogen bonds, indicating a less stable interaction interface (Figure 4I). The persistent hydrogen-bonding network in the MESV–TLR2 complex underscores its strong binding affinity and structural stability. These findings suggest that the MESV forms a more stable and tightly bound interaction with TLR2. Collectively, the high docking scores and favorable energy values for the MESV with TLR1, TLR2, and the TLR1–TLR2 heterodimer indicate that the produced vaccine may have the desired orientation and stability within the respective binding pockets.

3.8 Expression and purification of MESV protein

To express the MESV protein, we simulated codon optimization and expression analysis using JCAT, resulting in a CAI of 0.996 and an average GC content of 54.728%. These values suggest efficient transcription and translation potential in a bacterial expression system. We thus cloned the optimized sequence into a pET28a(+) expression vector (Supplementary Figure S6). We then expressed the His-tagged MESV protein, with a molecular weight of ~95 kDa, in E. coli, purified the MESV protein using a Ni–NTA affinity chromatography column (Figure 5A), and confirmed expression by immunoblot analysis using an anti-His antibody (Figure 5B).

Figure 5

Figure 5. Expression and purification of MESV. (A) Coomassie Brilliant Blue–stained SDS-PAGE of total E. coli BL21 (DE3) lysates and Ni–NTA eluate. Lanes: 1, protein marker (kDa); 2, uninduced lysate; 3, IPTG-induced lysate; 4, Ni–NTA-purified MESV. A prominent band at ~95 kDa appears after induction and is strongly enriched in the purified fraction. (B) Anti-His immunoblot analysis of MESV samples. The ~95 kDa band is detected in the induced and purified lanes but not in the uninduced control, confirming the identity of the His-tagged MESV.

3.9 Immune-cell profiling of the MESV

Next, we assessed the immune response induced by the MESV using the C-IMMSIM Immune System Simulator, following a three-dose vaccination regimen administered on days 1, 84, and 170. We detected elevated levels of immunoglobulin (Ig) G, including IgG1, IgG2, and IgM, as well as an increased number of CD8⁺ cytotoxic T cells. Moreover, the number of plasma cells producing these antibodies reached peak levels at different time points (Figures 6A–D). We also observed a two- to four-fold increase in the number of active, resting, internalized, or antigen-presenting dendritic cells and macrophages (Figures 6E–F), indicating effective innate immune activation. Furthermore, we noted an increased number of B lymphocytes and memory B cells (Figures 6A–B, G–H). By contrast, the change in the number of CD4+ regulatory T lymphocytes was subtle (Figure 6I). Alongside this, we also observed a peak in the total number of natural killer cells after vaccination (Figure 6L). Consistently, vaccination with a previously reported multi-epitope HIV vaccine produced an immune response similar to that of our MESV (overall kinetics/pattern) under the same C-IMMSIM Immune System Simulator settings (27), except that it induced less expansion of CD8⁺ cytotoxic T cells and slightly lower numbers of class-switched IgG and active B-cell fractions (Supplementary Figure S7). Collectively, these data suggest that the proposed MESV contributes to both short- and long-term immune protection with favorable, sustained humoral and cellular immune responses.

Figure 6

Figure 6. Immune simulation shows immune profile of the MESV vaccine. (A) Antigen count per mL over time. The x-axis represents days elapsed, and the y-axis displays antigen count. (B) B-cell population. (C) Distribution of B-cell activation states. (D) Plasma cell (PLB) population. (E) Number of dendritic cells (DC). (F) Number of macrophages (MA). (G) Number of T-helper (TH) cells. (H) CD4 T-helper lymphocytes sub-divided by entity state. (I) Number of CD4 T-regulatory lymphocytes. (J) CD8 T- cytotoxic lymphocytes. (K) CD8 T-cytotoxic lymphocytes (TC cells) per entity-state. (L) Number of Natural Killer (NK) cells.

4 Discussion

The ever-expanding genetic diversity of HIV-1 and its ability to evade immune detection pose major challenges in the development of an effective vaccine (26). Here, we screened the entire HIV-1 proteome and identified 44 epitopes, including 12 novel ones, comprising 12 MHC-I–binding CTL epitopes, 18 MHC-II–binding HTL epitopes, and 14 linear B-cell epitopes. These epitopes exhibited high antigenicity, low toxicity and allergenicity, and no homology with human proteins. Among these epitopes, 30 T-cell epitopes (combined MHC-I and MHC-II) that remained unchanged across HIV-1 variants achieved near-universal population coverage. These epitopes might provide protection for 98.63% of the global population through MHC-I alleles and 99.67% through MHC-II alleles. The conservation of these epitopes across HIV-1 variants, paired with their strong association with common HLA alleles, indicates that they are evolutionary “anchors” in the virus’s structure, efficiently recognized by immune systems worldwide (46). This dual conservation of both viral epitopes and human HLA diversity addresses a longstanding barrier in HIV vaccine development by balancing broad efficacy with genetic variability (47). By prioritizing epitopes presented by globally prevalent HLA molecules, our design avoids overreliance on region-specific immune markers, offering a solution that is as inclusive as it is innovative (48). These findings align with the growing emphasis on universal vaccine frameworks that prioritize accessibility for all populations, not just those with dominant HLA types (49).

Two main mechanisms may contribute to vaccine protection against HIV-1 infection: one involves inducing neutralizing antibodies (nAbs) against HIV-1 envelope immunogens to eliminate the virus, while the other targets non-neutralizing factors, such as non-nAbs and immune cells, to control viral infection through alternative immune mechanisms (50, 51). Because the HIV-1 envelope glycoprotein gp160 is anchored in both viral and host cell membranes, it serves as the primary target for nAbs. To enhance vaccine efficacy, the MESV designed in this study included multiple epitopes that may induce nAbs, such as IRGKVQKEYAFFYKLDIIPIDNDT in the hypervariable V2 region of gp160, which is a key target for nAb production against diverse HIV-1 variants (52). In addition to this B-cell epitope, the MESV incorporates MHC-I epitopes (e.g., RVKEKYQHL) and MHC-II epitopes from gp160, which can be presented to induce T-cell-mediated cytotoxicity for viral clearance (53). Moreover, we included novel epitopes within immunogenic regions of gp160, such as the V3 loops, to enhance immune activation. Epitopes incorporated into the MESV from Gag-Pol (e.g., GSEELRSLY and IPLTEEAEL) may also elicit CTL responses targeting both Gag and Pol components (54). Additionally, the incorporation of epitopes from Nef can further enhance the cellular immune response (55).

The designed MESV comprises 889 amino acids with a predicted molecular mass of ~95.1 kDa. This falls well within the typical size range for recombinant protein antigens and multi-epitope fusion constructs (56, 57). Given these parameters, it is unsurprising that the MESV demonstrated multiple immunogenic advantages in silico, including strong immune activation and the potential for long-term immune protection. This modular design not only resists antigenic drift but also permits rapid adaptation to emerging strains, offering a scalable solution for regions burdened by HIV-1 diversity. By harmonizing conserved epitopes from structural and functional proteins, the MESV represents a pivotal step toward a universally effective HIV-1 vaccine, bridging the gap between computational innovation and clinical application.

5 Conclusions

Overall, the high antigenicity, non-allergenicity, favorable solubility, low toxicity, and strong conservation of the selected epitopes suggest that the MESV is a promising multi-epitope vaccine candidate against diverse HIV-1 variants. Its high binding affinity to TLR2 and TLR1 receptors involved in immune recognition of pathogen-associated molecular patterns supports its potential to elicit both antibody-mediated and cell-mediated immune responses. While in silico analyses provide valuable insights, further in vitro and in vivo validation is essential to fully characterize immunogenic properties and mechanisms of action. Nonetheless, this study highlights the potential of computer-aided vaccine design in developing effective strategies against the global spread of HIV and potentially other infectious diseases.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Ethics statement

Ethical approval was not required for the study involving humans in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and the institutional requirements.

Author contributions

SK: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. ZA: Data curation, Formal analysis, Investigation, Methodology, Software, Writing – original draft. MH: Conceptualization, Data curation, Formal analysis, Methodology, Software, Visualization, Writing – review & editing. LL: Data curation, Formal analysis, Investigation, Validation, Visualization,Writing – review & editing. MI: Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – review & editing. LY: Formal analysis, Funding acquisition, Project administration, Supervision, Validation, Visualization, Writing – review & editing. SZ: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – review & editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This study was supported by National Natural Science Foundation of China (32170712), Shenzhen Medical Research Fund (B2302022), and the Shenzhen Key Laboratory Foundation (ZDSYS20200811143757022).

Conflict of interest

The authors declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fviro.2025.1705632/full#supplementary-material

Supplementary Figure 1 | Overview of the in silico approach for MESV vaccine design and validation.

Supplementary Figure 2 | Phylogenetic analysis of HIV samples from the LANL database, highlighting the top amino acid sequences of selected epitopes.

Supplementary Figure 3 | Schematic design of the vaccine construct.

Supplementary Figure 4 | Secondary structure of the MESV, showing alpha-helices, beta-strands, and coil regions.

Supplementary Figure 5 | Conformational B-cell epitope on the surface of MESV.

Supplementary Figure 6 | MESV cloned into pET-28a(+) using NcoI and XhoI restriction sites, the codon-optimized MESV insert is highlighted in blue.

Supplementary Figure 7 | Comparative immune-simulation results for the multi-epitope HIV construct of Hashempour et al., 2024, generated in C-ImmSim using the same parameters as MESV.

Supplementary Table 1 | Antigenicity, physicochemical properties, and epitope conservation of MESV in the HIV-1 Group M, subtype B isolate HXB2 proteome.

Supplementary Table 2 | Immunogenic selected epitopes for MHC-II (HTL) cell-recognized HLA alleles.

Supplementary Table 3 | Immunogenic selected epitopes for MHC-I (CTL) cell-recognized HLA alleles.

Supplementary Table 4 | Cross-reference of MESV epitopes with the LANL HIV molecular immunology database showing exact sequence identity matches (closest partial/overlap hits) and sequence logos.

Supplementary Table 5 | Prediction of conformational B-cell epitopes based on 3D structural modeling.

Supplementary Table 6 | Disulfide engineering details of the vaccine structure, including residue numbers, names, chain identifiers, and bond characteristics such as χ3 angles, energy values, and Σ B-factors.

References

1. Sharp PM and Hahn BH. Origins of HIV and the AIDS pandemic. Cold Spring Harb Perspect Med. (2011) 1:a006841. doi: 10.1101/cshperspect.a006841

PubMed Abstract | Crossref Full Text | Google Scholar

2. Astha, Arti A, Ankita S, Neetu C, Priyanka S, and Sanjeev K. A study on prevalence of HIV and its associated risk factors in patients visiting tertiary care center of North India. Asian J Med Sci. (2023) 14:58–64. doi: 10.3126/ajms.v14i1.45997

Crossref Full Text | Google Scholar

3. Landovitz RJ, Scott H, and Deeks SG. Prevention, treatment and cure of HIV infection. Nat Rev Microbiol. (2023) 21:657–70. doi: 10.1038/s41579-023-00914-1

PubMed Abstract | Crossref Full Text | Google Scholar

4. Bekker LG, Beyrer C, Mgodi N, Lewin SR, Delany-Moretlwe S, Taiwo B, et al. HIV infection. Nat Rev Dis Primers. (2023) 9:42. doi: 10.1038/s41572-023-00452-3

PubMed Abstract | Crossref Full Text | Google Scholar

5. Mercer A. Protection against severe infectious disease in the past. Pathog Glob. Health. (2021) 115:151–67. doi: 10.1080/20477724.2021.1878443

PubMed Abstract | Crossref Full Text | Google Scholar

6. Williams ME and Cloete R. Molecular modeling of subtype-specific tat protein signatures to predict tat-TAR interactions that may be involved in HIV-associated neurocognitive disorders. Front Microbiol. (2022) 13:866611. doi: 10.3389/fmicb.2022.866611

PubMed Abstract | Crossref Full Text | Google Scholar

7. Aiken C and Rousso I. The HIV-1 capsid and reverse transcription. Retrovirology. (2021) 18:29. doi: 10.1186/s12977-021-00566-0

PubMed Abstract | Crossref Full Text | Google Scholar

8. Jha V, Rustagi K, Gharat K, Sonawane N, Rathod M, Patel R, et al. Human immunodeficiency virus type 1: Role of proteins in the context of viral life cycle. J Adv. Biotechnol Exp Ther. (2022) 5, 307–19. doi: 10.5455/jabet.2022.d117

Crossref Full Text | Google Scholar

9. Klasse PJ, Sanders RW, Cerutti A, and Moore JP. How can HIV-type-1-Env immunogenicity be improved to facilitate antibody-based vaccine development? AIDS Res Hum Retroviruses. (2012) 28:1–15. doi: 10.1089/aid.2011.0053

PubMed Abstract | Crossref Full Text | Google Scholar

10. Girard MP, Osmanov S, Assossou OM, and Kieny MP. Human immunodeficiency virus (HIV) immunopathogenesis and vaccine development: a review. Vaccine. (2011) 29:6191–218. doi: 10.1016/j.vaccine.2011.06.085

PubMed Abstract | Crossref Full Text | Google Scholar

11. Wei M, Zhao X, Liu M, Huang Z, Xiao Y, Niu M, et al. Inhibition of HIV-1 assembly by coiled-coil domain containing protein 8 in human cells. Sci Rep. (2015) 5:14724. doi: 10.1038/srep14724

PubMed Abstract | Crossref Full Text | Google Scholar

12. Modi PK, Jaiswal S, and Sharma P. Regulation of neuronal cell cycle and apoptosis by microRNA 34a. Mol Cell Biol. (2016) 36:84–94. doi: 10.1128/MCB.00589-15

PubMed Abstract | Crossref Full Text | Google Scholar

13. Cribbs SK, Crothers K, and Morris A. Pathogenesis of HIV-related lung disease: immunity, infection, and inflammation. Physiol Rev. (2020) 100:603–32. doi: 10.1152/physrev.00039.2018

PubMed Abstract | Crossref Full Text | Google Scholar

14. Wallet C, Rohr O, and Schwartz C. Evolution of a concept: From accessory protein to key virulence factor, the case of HIV-1 Vpr. Biochem Pharmacol. (2020) 180, 1–21. doi: 10.1016/j.bcp.2020.114128

PubMed Abstract | Crossref Full Text | Google Scholar

15. Nodder SB and Gummuluru S. Illuminating the role of vpr in HIV infection of myeloid cells. Front Immunol. (2019) 10:1606. doi: 10.3389/fimmu.2019.01606

PubMed Abstract | Crossref Full Text | Google Scholar

16. Ruiz A, Guatelli JC, and Stephens EB. The Vpu protein: new concepts in virus release and CD4 down-modulation. Curr HIV Res. (2010) 8:240–52. doi: 10.2174/157016210791111124

PubMed Abstract | Crossref Full Text | Google Scholar

17. Bieback K, Lien E, Klagge IM, Avota E, Schneider-Schaulies J, Duprex WP, et al. Hemagglutinin protein of wild-type measles virus activates toll-like receptor 2 signaling. J Virol. (2002) 76:8729–36. doi: 10.1128/JVI.76.17.8729-8736.2002

PubMed Abstract | Crossref Full Text | Google Scholar

18. Jin MS, Kim SE, Heo JY, Lee ME, Kim HM, Paik SG, et al. Crystal structure of the TLR1-TLR2 heterodimer induced by binding of a tri-acylated lipopeptide. Cell. (2007) 130:1071–82. doi: 10.1016/j.cell.2007.09.008

PubMed Abstract | Crossref Full Text | Google Scholar

19. Kim CG, Kye YC, and Yun CH. The role of nanovaccine in cross-presentation of antigen-presenting cells for the activation of CD8(+) T cell responses. Pharmaceutics. (2019) 11, 1–28. doi: 10.3390/pharmaceutics11110612

PubMed Abstract | Crossref Full Text | Google Scholar

20. Abdellrazeq GS, Fry LM, Elnaggar MM, Bannantine JP, Schneider DA, Chamberlin WM, et al. Simultaneous cognate epitope recognition by bovine CD4 and CD8 T cells is essential for primary expansion of antigen-specific cytotoxic T-cells following ex vivo stimulation with a candidate Mycobacterium avium subsp. paratuberculosis peptide vaccine. Vaccine. (2020) 38:2016–25. doi: 10.1016/j.vaccine.2019.12.052

PubMed Abstract | Crossref Full Text | Google Scholar

21. Ron R, Moreno E, Martínez-Sanz J, Brañas F, Sainz T, Moreno S, et al. CD4/CD8 ratio during human immunodeficiency virus treatment: time for routine monitoring? Clin Infect Dis. (2023) 76:1688–96. doi: 10.1093/cid/ciad136

PubMed Abstract | Crossref Full Text | Google Scholar

22. Cui B, Lin H, Yu J, Yu J, and Hu Z. Autophagy and the immune response. Adv Exp Med Biol. (2019) 1206:595–634. doi: 10.1007/978-981-15-0602-4_27

PubMed Abstract | Crossref Full Text | Google Scholar

23. Chen Z, Yu J, Guo Z, Chen S, Li Y, Zhou Q, et al. Harnessing the power of tumor-draining lymph nodes: unveiling predictive biomarkers for immune checkpoint inhibitor. Explor Target. Antitumor. Ther. (2025) 6:1002315. doi: 10.37349/etat.2025.1002315

PubMed Abstract | Crossref Full Text | Google Scholar

24. Roth GA, Picece V, Ou BS, Luo W, Pulendran B, and Appel EA. Designing spatial and temporal control of vaccine responses. Nat Rev Mater. (2022) 7:174–95. doi: 10.1038/s41578-021-00372-2

PubMed Abstract | Crossref Full Text | Google Scholar

25. Kumar N, Sood D, and Chandra R. Design and optimization of a subunit vaccine targeting COVID-19 molecular shreds using an immunoinformatics framework. RSC. Adv. (2020) 10:35856–72. doi.org/10.1039/D0RA06849G

PubMed Abstract | Google Scholar

26. Haynes BF, Wiehe K, Borrow P, Saunders KO, Korber B, Wagh K, et al. Strategies for HIV-1 vaccines that induce broadly neutralizing antibodies. Nat Rev Immunol. (2023) 23:142–58. doi: 10.1038/s41577-022-00753-w

PubMed Abstract | Crossref Full Text | Google Scholar

27. Hashempour A, Khodadad N, Akbarinia S, Ghasabi F, Ghasemi Y, Nazar M, et al. Reverse vaccinology approaches to design a potent multiepitope vaccine against the HIV whole genome: immunoinformatic, bioinformatics, and molecular dynamics approaches. BMC Infect Dis. (2024) 24:873. doi: 10.1186/s12879-024-09775-2

PubMed Abstract | Crossref Full Text | Google Scholar

28. Mishra SK, Senathilake KS, Kumar N, Patel CN, Uddin MB, Alqahtani T, et al. Exploratory algorithms to devise multi-epitope subunit vaccine by examining HIV-1 envelope glycoprotein: An immunoinformatics and viroinformatics approach. PloS One. (2025) 20:e0318523. doi: 10.1371/journal.pone.0318523

PubMed Abstract | Crossref Full Text | Google Scholar

29. Garg VK, Avashthi H, Tiwari A, Jain PA, Ramkete PW, Kayastha AM, et al. MFPPI–multi FASTA protParam interface. Bioinformation. (2016) 12:74. doi: 10.6026/97320630012074

PubMed Abstract | Crossref Full Text | Google Scholar

30. Dimitrov I, Naneva L, Doytchinova I, and Bangov I. AllergenFP: allergenicity prediction by descriptor fingerprints. Bioinformatics. (2014) 30:846–51. doi: 10.1093/bioinformatics/btt619

PubMed Abstract | Crossref Full Text | Google Scholar

31. Sun P, Ju H, Liu Z, Ning Q, Zhang J, Zhao X, et al. Bioinformatics resources and tools for conformational B-cell epitope prediction. Comput Math. Methods Med. (2013) 2013:943636. doi: 10.1155/2013/943636

PubMed Abstract | Crossref Full Text | Google Scholar

32. Craig DB and Dombkowski AA. Disulfide by Design 2.0: a web-based tool for disulfide engineering in proteins. . BMC Bioinf. (2013) 14:346. doi: 10.1186/1471-2105-14-346

PubMed Abstract | Crossref Full Text | Google Scholar

33. Dominguez C, Boelens R, and Bonvin AM. HADDOCK: a protein– protein docking approach based on biochemical or biophysical information. J Am Chem Soc. (2003) 125:1731–7. doi: 10.1021/ja026939x

PubMed Abstract | Crossref Full Text | Google Scholar

34. Al-Zayadi FQJ, Shakir AS, Kareem AS, Ghasemian A, and Behmard E. Design of a novel multi-epitope vaccine against Marburg virus using immunoinformatics studies. BMC Biotechnol. (2024) 24:45. doi: 10.1186/s12896-024-00873-2

PubMed Abstract | Crossref Full Text | Google Scholar

35. Wacha AF and Lemkul JA. charmm2gmx: an automated method to port the CHARMM additive force field to GROMACS. J Chem Inf. Model. (2023) 63:4246–52. doi: 10.1021/acs.jcim.3c00860

PubMed Abstract | Crossref Full Text | Google Scholar

36. Alharbi M, Alshammari A, Alsabhan JF, Alzarea SI, Alshammari T, Alasmari F, et al. A novel vaccine construct against Zika virus fever: insights from epitope-based vaccine discovery through molecular modeling and immunoinformatics approaches. Front Immunol. (2024) 15:1426496. doi: 10.3389/fimmu.2024.1426496

PubMed Abstract | Crossref Full Text | Google Scholar

37. Samantray S, Schumann W, Illig A-M, Carballo-Pacheco M, Paul A, Barz B, et al. (2020).

Google Scholar

38. Hollingsworth SA and Dror RO. Molecular dynamics simulation for all. Neuron. (2018) 99:1129–43. doi: 10.1016/j.neuron.2018.08.011

PubMed Abstract | Crossref Full Text | Google Scholar

39. Pan C, Yue H, Zhu L, Ma GH, and Wang HL. Prophylactic vaccine delivery systems against epidemic infectious diseases. Adv Drug Delivery Rev. (2021) 176:113867. doi: 10.1016/j.addr.2021.113867

PubMed Abstract | Crossref Full Text | Google Scholar

40. Hochuli E, Bannwarth W, Döbeli H, Gentz R, and Stüber D. Genetic approach to facilitate purification of recombinant proteins with a novel metal chelate adsorbent. Bio/technology. (1988) 6:1321–5. doi: 10.1038/nbt1188-1321

Crossref Full Text | Google Scholar

41. Junqueira DM and Almeida SE. HIV-1 subtype B: Traces of a pandemic. Virology. (2016) 495:173–84. doi: 10.1016/j.virol.2016.05.003

PubMed Abstract | Crossref Full Text | Google Scholar

42. Malhotra U, Holte S, Dutta S, Berrey MM, Delpit E, Koelle DM, et al. Role for HLA class II molecules in HIV-1 suppression and cellular immunity following antiretroviral treatment. J Clin Invest. (2001) 107:505–17. doi: 10.1172/JCI11275

PubMed Abstract | Crossref Full Text | Google Scholar

43. Salvadori L, Santana F, and Marcos EVC. Frequency of alleles and haplotypes of the human leukocyte antigen system in Bauru São Paulo, Brazil. Rev Bras Hematol. e Hemoterapia. (2014) 36:108–14. doi: 10.5581/1516-8484.20140026

PubMed Abstract | Crossref Full Text | Google Scholar

44. Jager S, Cimermancic P, Gulbahce N, Johnson JR, McGovern KE, Clarke SC, et al. Global landscape of HIV-human protein complexes. Nature. (2011) 481:365–70. doi: 10.1038/nature10719

PubMed Abstract | Crossref Full Text | Google Scholar

45. Funderburg N, Lederman MM, Feng Z, Drage MG, Jadlowsky J, Harding CV, et al. Human β-defensin-3 activates professional antigen-presenting cells via Toll-like receptors 1 and 2. Proc Natl Acad Sci. (2007) 104:18631–5. doi: 10.1073/pnas.0702130104

PubMed Abstract | Crossref Full Text | Google Scholar

46. Boomgarden AC and Upadhyay C. Progress and challenges in HIV-1 vaccine research: A comprehensive overview. Vaccines (Basel). (2025) 13. doi: 10.3390/vaccines13020148

PubMed Abstract | Crossref Full Text | Google Scholar

47. Sundaramurthi JC, Ashokkumar M, Swaminathan S, and Hanna LE. HLA based selection of epitopes offers a potential window of opportunity for vaccine design against HIV. Vaccine. (2017) 35:5568–75. doi: 10.1016/j.vaccine.2017.08.070

PubMed Abstract | Crossref Full Text | Google Scholar

48. Ahir-Bist S, Chavan V, Padmaja Samant M, Nanavati R, Mehta P, and Mania-Pramanik J. Human leukocyte antigens class I and class II alleles associated with vertical human immunodeficiency virus transmission - an exploratory study from Mumbai, India. Indian J Dermatol Venereol Leprol. (2023) 89:159. doi: 10.25259/IJDVL_654_19

PubMed Abstract | Crossref Full Text | Google Scholar

49. Rodrigue KD, Paul YAVD, Cézaire AAA, Germaine B, Roseline AM, André IK, et al. HLA class I and II polymorphism in ivorian subjects: HIV-1 infected versus uninfected. Open J Immunol. (2025) 15:1–28. doi: 10.4236/oji.2025.151001

Crossref Full Text | Google Scholar

50. Burton DR. Advancing an HIV vaccine; advancing vaccinology. Nat Rev Immunol. (2019) 19:77–8. doi: 10.1038/s41577-018-0103-6

PubMed Abstract | Crossref Full Text | Google Scholar

51. Su H, Zhang J, Yi Z, Khan S, Peng M, Ye L, et al. A human monoclonal antibody neutralizes SARS-CoV-2 Omicron variants by targeting the upstream region of spike protein HR2 motif. hLife. (2024) 2:126–40. doi: 10.1016/j.hlife.2024.02.001

Crossref Full Text | Google Scholar

52. Karasavvas N, Billings E, Rao M, Williams C, Zolla-Pazner S, Berman PW, et al. Plasma igG to linear epitopes in the V2 and V3 regions of HIV-1 gp120 correlate with a reduced risk of infection in the RV144 vaccine efficacy trial. PloS One. (2013) 8:e75665. doi: 10.1371/journal.pone.0075665

PubMed Abstract | Crossref Full Text | Google Scholar

53. Marino F. Gaining insight into posttranslationally modified HIV antigens and their underlying characteristics. Proteomics. (2018) 18:e1800041. doi: 10.1002/pmic.201800041

PubMed Abstract | Crossref Full Text | Google Scholar

54. Moog C, Dereuddre-Bosquet N, Teillaud JL, Biedma ME, Holl V, Van Ham G, et al. Protective effect of vaginal application of neutralizing and nonneutralizing inhibitory antibodies against vaginal SHIV challenge in macaques. Mucosal Immunol. (2014) 7:46–56. doi: 10.1038/mi.2013.23

PubMed Abstract | Crossref Full Text | Google Scholar

55. Larijani MS, Sadat SM, Bolhassani A, Pouriayevali MH, Bahramali G, and Ramezani A. In silico design and immunologic evaluation of HIV-1 p24-nef fusion protein to approach a therapeutic vaccine candidate. Curr HIV Res. (2018) 16:322–37. doi: 10.2174/1570162X17666190102151717

PubMed Abstract | Crossref Full Text | Google Scholar

56. Tarrahimofrad H, Rahimnahal S, Zamani J, Jahangirian E, and Aminzadeh S. Designing a multi-epitope vaccine to provoke the robust immune response against influenza A H7N9. Sci Rep. (2021) 11:24485. doi: 10.1038/s41598-021-03932-2

PubMed Abstract | Crossref Full Text | Google Scholar

57. Maleki A, Russo G, Parasiliti Palumbo GA, and Pappalardo F. In silico design of recombinant multi-epitope vaccine against influenza A virus. BMC Bioinf. (2022) 22:617. doi: 10.1186/s12859-022-04581-6

PubMed Abstract | Crossref Full Text | Google Scholar