Immunoinformatic-driven design of a multi-epitope vaccine against Helicobacterpylori

Abstract

Helicobacter pylori (H. pylori) is present in the gastric mucosa of approximately half of the global population and is classified as a definitive type I carcinogen. Developing a logical and effective vaccine against H. pylori remains a critical need. To enhance immune protection, a multi-epitope vaccine was designed using an immunoinformatic-driven approach. The genome of the pathogen was analyzed to identify potential epitopes using several algorithms for predicting T-cell and B-cell epitopes. The resulting multi-epitope vaccine was evaluated for antigenicity, allergenicity, and physicochemical properties. The tertiary structure was predicted and validated using a Ramachandran plot. Molecular docking analysis revealed favorable interactions between the vaccine and immune receptors, while molecular dynamics simulations performed with GROMACS confirmed the stability of these interactions through key dynamic indexes. Additionally, immune simulations, codon optimization, and in silico cloning were conducted to further evaluate the vaccine candidate. In silico predictions indicate that the proposed vaccine construct may be safe and stable, holding promise as an effective measure against H. pylori-induced gastric inflammation.

Graphical abstract

Schematic overview of the vaccine design process for Helicobacter pylori. The diagram illustrates the key stages and evaluations performed to develop the final vaccine candidate.

Highlights

• •Multi-epitope vaccine design against H. pylori using immunoinformatic for enhanced immune protection.
• •Comprehensive epitope screening using advanced algorithms for T-cell and B-cell epitope prediction.
• •Stable and non-allergenic vaccine candidate with favorable antigenicity and physicochemical properties.
• •Molecular docking and dynamics simulations confirm strong binding and stability with immune receptors.
• •In silico immune simulations, codon optimization, and cloning support vaccine efficacy and feasibility.

1. Introduction

Helicobacter pylori (H. pylori) is a Gram-negative bacterium that colonizes the gastric mucosa and, on the basis of extensive epidemiological evidence, is classified as a definitive type I carcinogen [1]. It infects roughly half of the world's population [2]. Transmission occurs primarily via faecal–oral and oral–oral routes through contact with contaminated faeces, vomitus, saliva, or ingestion of tainted water and food [3]. Colonization usually begins in childhood (≤10 years) and can persist for life. Although infection is often asymptomatic, chronic colonization provokes acute and chronic inflammation of the gastric epithelium that can progress to chronic gastritis, peptic ulcer disease, gastric adenocarcinoma, and mucosa-associated lymphoid-tiessue lymphoma [4,5]. Gastric cancer is the fifth most commonly diagnosed malignancy and the third leading cause of cancer-related mortality worldwide [6,7]. Moreover, prolonged antibiotic regimens for H. pylori eradication disrupt gut-microbiota homeostasis and may precipitate additional gastrointestinal complications. Consequently, development of an effective prophylactic vaccine remains a critical global-health priority [8]. Recent advances in immunoinformatic enable rapid, cost-effective identification of B- and T-cell epitopes and rational construction of stable, immunogenic multi-epitope vaccines [9]. Despite intensive global research efforts, however, no licensed vaccine against H. pylori has yet reached the clinic [10,11].

A fundamental principle of high-performance vaccine design is the disruption of a pathogen's pathogenic mechanisms. Incorporating multiple antigens broadens and strengthens protective immunity, while epitope-based strategies further enhance safety by focusing the response on well-defined, immunodominant regions and eliminating potentially deleterious epitopes present in whole-antigen formulations. Immunological evidence demonstrates that H. pylori infection elicits robust humoral (B-cell) and cellular (T-cell) adaptive immune responses [10].

Accordingly, we used immunoinformatic workflows to construct a multi-epitope vaccine. Molecular-dynamics simulations were then performed to assess the structural stability of the standalone vaccine and its complexes with immune receptors. This fully in silico analysis provided detailed insights into the physicochemical characteristics and dynamic behavior of the construct, thereby guiding the rational development of more potent vaccines against H. pylori infection.

In the present study, we mined the complete H. pylori genome and applied a comprehensive in-silico screening pipeline to identify several open reading frames (ORFs) that have not been targeted in previous vaccine investigations. Because extracellular and outer-membrane proteins constitute the bacterium's first points of contact with host immune cells, ORFs encoding surface-exposed antigens were prioritized for inclusion. The final multi-epitope construct also incorporates a single adjuvant domain with documented anti-inflammatory and anticancer properties to enhance immunogenicity. To probe host–pathogen interactions, we performed duplicate molecular-docking analyses between the vaccine and two distinct immune receptors, thereby providing an additional layer of validation for the construct's potential efficacy. TLR2 and TLR4 were selected for molecular docking and dynamic simulation due to their pivotal role in the innate immune response to H. pylori. Numerous studies indicate that H. pylori are recognized by TLR2 and TLR4, respectively. This recognition triggers the downstream signaling pathways (e.g., NF-κB) that lead to the production of pro-inflammatory cytokines and chemokines, which are central to the immune response against this pathogen [[12], [13], [14]]. Therefore, analyzing the interaction of our designed vaccine with these specific TLRs is essential for predicting its potential immunogenicity and efficacy.

2. Methods

2.1. Retrieval of H. pylori sequence

In this study the complete genomic sequence of Helicobacter pylori (NZ_AP026446.1) with size 1,696,601 bp was retrieved from the NCBI database (https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_025998455.1/).

2.2. Identification and selection of the desired ORF

A preliminary screening of the entire H. pylori genome was conducted to identify potential vaccine targets. The complete genomic sequence was retrieved from NCBI. Open Reading Frames (ORFs) were predicted using the ORF Finder tool (https://www.ncbi.nlm.nih.gov/orffinder/).ORF finder server was configurated with a default values. Minimum ORF length was selected 75 nucleotide, standard genetic code was applied for protein translation and ‘AGT” codon was set as only start codon’ in addition nested ORFs(overlapping ORFs) were not ignored from analysis.

After data submitted,ORFs were identified through server. The next process is to find ORFs position in the cell. The subcellular localization of the resulting proteins was subsequently predicted with PSORTb v3.0.3 (https://www.psort.org/) [15].

Required data including organism type and gram strain which are bactria and negative gram respectively with the ORF sequence were submitted. The server analyzes and provides the results. Proteins localized to the extracellular space or outer membrane were prioritized for further analysis due to their high accessibility to the host immune system. In order to vaccine candidate does not include any signal peptide, we identified peptides containing signal peptide and omitted signal peptides through SignalP −0.5 server(https://services.healthtech.dtu.dk/services/SignalP-5.0/) [16]. The date required such as organism group and protein sequences in FASTA format were submitted. This server identified the peptides which have signal peptide and deleted them. The transmembrane helices position which located in membrane and into cell were detected and removed using TMHMM-2.0; those possessing such domains were excluded to focus on soluble antigens (https://services.healthtech.dtu.dk/services/TMHMM-2.0/).

2.3. Predicting antigenicity, allergenicity, and toxicity of protein sequences

Antigenicity features of ORFs were studied through the VaxiJen 2.0 server (https://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html), the threshold 0.4 was selected. It means ORFs with antigenicity score less than 0.4 were removed. VaxiJen 2.0 serve is the first server for alignment-independent protective antigen prediction. The VaxiJen server analyzes bacterial, viral, and tumor protein datasets to predict protein antigenicity [17]. Additionally, we utilized the ToxinPred server (http://crdd.osdd.net/raghava/toxinpred/) for screening in terms of toxicity [18]. The setting was set according to ToxinPred server default. Hence, a peptide fragment length of 10, the SVM(Swiss-Prot) method and SVM threshold of 0 were considered. In order to verify whether the created vaccination could contribute to an allergic chain reaction, the AllerTop v.2.0 (https://www.ddg-pharmfac.net/AllerTOP/) server was applied [19]. The antigenic proteins were also evaluated using the BLAST online tool to check for homologous results in humans and animals. For this aim, the NCBI (National Center for Biotechnology Information) server (https://blast.ncbi.nlm.nih.gov/Blast.cgi) was used, and algorithm blastp (protein-protein BLAST was selected for program selection [20].

2.4. Predicting B-Cell epitope

B-cell epitopes play a crucial role in peptide vaccine design, disease detection, and allergy research. We used the Immune Epitope Database and Analysis Resource (IEDB; https://www.iedb.org/) to predict linear B-cell epitopes from our protein sequence, running the server with default parameters. This database server assist researcher to predict T-cells, MHC1 and MHC2 according to desired sequences through access to a variety of epitope analysis and prediction tools [21]. Following the selection of the B-cell epitope prediction box the “linear epitope prediction section was chosen and finally the desired sequence with plain format was submitted. More ever selected method was Bepipred Linear Epitope Prediction 2.0 according the default. Subsequently, the final output sequences were obtained with center position of 4 and a threshold of 0.500, the result and epitope scoring were presented in table and graphical format and retrieving results was selected in plain‐text format. Predicted peptides between 5 and 25 amino acids long were retained, and any shorter than five residues were discarded [22]. Finally, we selected those epitopes with the highest prediction scores for downstream analysis. Top candidates were analyzed for immunogenicity (IEDB score), antigenicity (VaxiJen v2.0), allergenicity (AllerTOP v2.0), and toxicity (ToxinPred).

2.5. Predicting MHC-I binding epitopes

Prediction of peptide binding to major histocompatibility complex (MHC) class I molecules is a critical step for CTL epitope identification. We used the IEDB tool to MHC I binding prediction tool (http://tools.iedb.org/mhci/) the procedure of predicting of MHC-I binding epitopes is generally similar with what was followed for predicting B-Cell epitope with minor difference. I in IEBD server, predicting MHC-I section was chosen, then the desired sequence with FASTA format was entered in input box. Peptides of 12–18 amino acids were considered, in line with typical allele-specific length preferences, and all human HLA alleles, ensuring broad global population coverage were selected all other settings were left at their default values and finally the output analysis was presented tubular format. Top candidates were further evaluated for immunogenicity (IEDB score), antigenicity (VaxiJen v2.0), allergenicity (AllerTOP v2.0), and toxicity (ToxinPred). To rank binding strength, we computed the average of each peptide's IEDB binding score and VaxiJen antigenicity score, and selected the ten peptides with the highest combined averages. No duplicate or overlapping sequences were observed among the final set.

2.6. MHC-II binding prediction

Infectious agents, allergies, cancer, and autoantigens can all be identified using MHC class II binding predictions as potential epitope candidates. The IEDB (http://tools.iedb.org/mhcii/) was utilized to predict desired peptides. It was run through default parameters. Of course, all HLA alleles were selected as well as MHC-I prediction, and the epitope length was considered to be 12-18 amino acids. All procedures for MHC-I binding prediction were performed to select for MHC-II, and finally, according to the average score, the top 10 predictions were considered MHC-II candidates.

2.7. Design and assembly of a potent vaccine candidate

In the present study, MHC-I, MHC-II and B cell epitopes were joined with proper linkers to create a potent multi-epitope vaccine. It is important to note that the selected B-cell epitopes retrieved from the IEBD database with the highest score were assembled manually through linkers and no computational server was employed. Three linkers, including AAY, GPGPG, and KK, were used to bind MHCI, MHCII, and B-cell epitopes, respectively. An adjuvant was used to strengthen the immune response. Adjuvants are agents that enhance the immune response to an antigen. The adjuvant used for vaccine structure is CTxB from cholera toxin. This adjuvant facilitates the molecular targeting of antigens to immune cells. It improves both cellular and humoral immune responses through connecting to GM1 (monosialotetrahexosylganglioside) receptors on immune cells [23]. The linker EAAAK was employed to link the adjuvant to the vaccine. Then, a tag sequence including histidine residues was attached to the vaccine to increase protein expression and purification.

2.8. Investigation of allergenicity, solubility and antigenicity of the designed protein

The antigenicity of the designed vaccine candidate was evaluated using the VaxiJen v2.0 server with a threshold of 0.4.(http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html). This tool employs automatic cross-covariance (ACC) transformation to convert protein sequences into uniform vectors of key amino acid properties, enabling the classification of antigens according to their physicochemical characteristics [17]. The AllerTOP v.2 server (http://www.ddg-pharmfac.net/AllerTOP/) was used to predict the allergenicity of the designed protein. AllerTOP v.2 leverages machine learning algorithms to assess allergenicity based on key physicochemical properties, distinguishing allergens, small antigens capable of eliciting an IgE-mediated immune response—from non-allergens [24]. Protein solubility was evaluated using the Protein-Sol server (https://protein-sol.manchester.ac.uk/). Protein-Sol predicts the solubility of proteins upon overexpression in E. coli based on their amino acid composition and physicochemical properties. A solubility score above the default threshold (≥0.45) was considered indicative of good solubility [25]. To assess the potential toxicity of the designed vaccine candidate, the ToxinPred2 server (https://webs.iiitd.edu.in/raghava/toxinpred/) was employed. ToxinPred2 predicts toxicity by analyzing sequence-based features and identifying motifs associated with toxic peptides, ensuring the safety profile of the designed protein [26].

2.9. Physicochemical properties

The physical and chemical properties of the final vaccine construct were studied using the ProtParam (http://web.expasy.org/protparam/) server, which reveals the functional and structural features of a protein [27]. This server calculates physicochemical properties, including the number of amino acids, theoretical pI, molecular weight, extinction coefficient, aliphatic index, instability index, and total hydropathic mean (GRAVY). These parameters help elucidate the nature, stability, and activity of the protein. The instability index of a protein shows its protein stability. A protein was regarded as stable if its computed protein instability index was less than 40.

2.10. Secondary and tertiary structure prediction

The secondary structure of the candidate protein was predicted using the Garnier-Osguthorpe-Robson IV (GOR IV) algorithm available on the NPS@ server (https://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_gor4.html). The GOR IV method utilizes probability parameters derived from experimentally determined protein tertiary structures, as determined by X-ray crystallography, to provide reliable predictions of α-helices, β-sheets, and coils [28]. The I-TASSER server (https://zhanggroup.org/I-TASSER/) and the Robetta web server (https://robetta.bakerlab.org/) were utilized for tertiary structure modelling. I-TASSER and Robetta use a hierarchical approach to protein structure prediction, combining threading, ab initio modelling, and iterative refinement to generate high-quality 3D models [29]. The GalaxyRefine2 web server (http://galaxy.seoklab.org/refine2) was used to refine and improve the predicted structure. The quality of the 3D model was assessed using PROCHECK (https://saves.mbi.ucla.edu) [30] for stereochemical analysis, ERRAT (https://saves.mbi.ucla.edu) for non-bonded interaction evaluation, ProSA Web (https://prosa.services.came.sbg.ac.at/prosa.php) [31] for energy profile validation, and MolProbity (http://molprobity.biochem.duke.edu) for comprehensive geometric analysis [32]. The best model was selected for the next steps.

2.11. Molecular docking

ClusPro and HADDOCK are popular web-based protein-protein docking servers. two web server used different algorithms for molecular docking. ClusPro 2.0 is an updated version of the original ClusPro server with improved algorithms and features for docking simulations and also The HADDOCK2.4 platform, accessible via its web interface at https://wenmr.science.uu.nl/haddock2.4, represents an advanced integrative modelling tool [33]. It combines heterogeneous experimental data with computational approaches to predict high-quality three-dimensional models of biomolecular assemblies [34,35].

ClusPro 2.0 web server is applied to understand the interaction between the designed vaccine structure and immune recepttor TLR2 (PDB ID: 5d3i) and TLR-4 (PDB ID: 7mlm). The key amino acids Leu317, Ile319, Phe322, Leu324, Phe325, Tyr326, Val348, Phe349, and Pro352 in the active site of TLR2 and Arg434, Ser413, Ser386, Arg380, Lys341, Lys263, and Gln339 in the active site of TLR4 were selected for molecular docking [[35], [36], [37]].

Residues involved in vaccine-receptor interactions were visualized using PyMOL (v1.1, Schrödinger, LLC) and plotted using the PDBsum-generate web server (https://www.ebi.ac.uk/thornton-srv/databases/pdbsum/Generate.html). The binding affinity for the protein-TLR2 and protein-TLR4 complexes were evaluated by PRODIGY web server(https://rascar.science.uu.nl/prodigy/) [38].

2.12. Molecular dynamics simulation

Molecular dynamics (MD) simulations are used to study the dynamic behavior of biological macromolecules such as proteins and nucleic acids. The molecular dynamics method is used to investigate the stability of the constructed vaccine model and vaccine-receptor complexes using the GROMACS 2022.1 software package. The protein and protein-receptor complexes parameter files are generated using the OPLS force field [39]. The GROMACS software itself is used to obtain the protein topology. The SPC/E model is used as the water model throughout the simulation [40]. A sufficient amount of Na+ and Cl-ions was added to the simulation box to neutralize the simulation system. The simulation system preparation involved sequential energy minimization steps: First, NVT ensemble equilibration was conducted at 300 K for 100 ps. Subsequently, pressure coupling was introduced using a Parrinello-Rahman barostat for NPT equilibration (300 K, 1.0 bar, 100 ps). This protocol was uniformly applied to all three MD simulations.

All bond constraints are performed using the Lincs algorithm. The PME method is used to calculate electrostatics with a 0.16 nm grid spacing and a 10 A cutoff and van der Waals (VDW) interactions were calculated with a cutoff of 1 nm, and the final simulation step is performed with a 2 fs time step for a duration of 150 ns to check the stability of the structures. Finally, RMSD, RMSF, H-bond, and Rg analyses are performed to check the stability or fluctuations of the structures [41,42].

2.13. Binding-free energy calculation

Protein-protein interaction (PPI) complexes were computationally predicted and analyzed using HawkDock, an integrated platform combining molecular docking and binding affinity assessment. The binding free energies of these complexes were further refined using the Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) method. This widely validated computational approach estimates interaction energetics by incorporating molecular mechanics force fields with implicit solvent models. This method accounts for key thermodynamic contributions, including van der Waals forces, electrostatic interactions, and solvation effects, providing a robust quantitative assessment of binding stability [43].

2.14. Simulation of the immune system

An immune simulation study was conducted to evaluate the immunogenicity and immune response profile of the designed vaccine candidate. The C-ImmSim server (http://150.146.2.1/C-IMMSIM/index.php), a computational tool that simulates the mammalian immune system, was used for this purpose. C-ImmSim uses position-specific scoring matrices (PSSM) and machine learning algorithms to predict immune interactions and responses [44].

2.15. Cloning and optimization

The nucleotide sequence encoding the target protein was optimized for expression in Escherichia coli K12 strain using the Java Codon Adaptation Tool (JCat). Codon optimization was performed to increase translation efficiency by replacing rare codons with those preferentially used in E. coli. The optimized nucleotide sequence was analyzed for GC content and codon adaptation index (CAI) using JCat. A CAI value close to 1.0 indicated high compatibility with the E. coli translation machinery [45].

The enhanced codon sequences of the adjuvant-containing proteins were inserted into the pET28a(+) plasmid and subsequently analyzed through the SnapGene Version 3.2.1 software to assess dual enzymatic cleavage utilizing NcoI and XhoI restriction enzymes [46].

3. Result

3.1. Retrieval of H. pylori sequence

To design an H. pylori vaccine candidate, the sequences were downloaded. To identify ORFs, the ORF FINDER server was used. 8209 ORFs were found. Then, PSORT prediction tools were applied to determine subcellular localization, and PSORT prediction tools were used to screen extracellular and outer membrane ORFs. We found 23 extracellular ORFs and 38 ORFs located in the outer membrane. Subsequently, through SignalP −0.5 server the ORFs including signal peptide were recognized and signal peptide was removed from ORFs that had it. Seven out of 23 extracellular ORFs had peptide signals, and 27 out of 38 outer membranes had signal peptides. Therefore, ORFs were classified into four groups: ex (Extracellular) without s.p (Signal peptide), ex with s.p, out (Outer membrane) without s.p and out with s.p. ORF15 (Flagellin B) was selected based on immunogenicity assessment and non-toxic nature related to extracellular ORFs without signal peptides with an antigenicity score of 0.7911 to find B-cell, MHC-I and MHC-II epitopes (Table 1). The results demonstrated a great antigenic nature, as well as non-toxic and non-allergenic properties.

Table 1.

ORF15 (Flagellin B) selected according to high score, immunogenicity assessment, non-toxic nature.

Name	Score	Toxinpred	Allergenicity
ORF15(Flagellin B)	0.7911	Non-toxin	Probable non-allergen

3.2. Prediction of B-Cell epitope

To identify the B-cell epitope, the IEDB server was applied. This server finds seven B-cell epitopes. Subsequently, three B-cells were selected based on appropriate structure.

In the next step, the predicted epitope was screened for antigenicity, allergenicity and toxicity. The VaxiJen server revealed antigenicity scores of 1.393, 1.258 and 0.6929 for three B-cell epitope candidates which were acceptable scores. Toxicity and allergenicity evaluation showed that the desired epitopes were not toxic and allergen (Table 2).

Table 2.

B-Cell epitopes selected for vaccine design based on high score, immunogenicity assessment, non-toxic nature.

Epitope	Toxicity	Allergenicity	Antigenicity
NNISVTQVNVKAAESQIRDVDF	Non-toxin	Probable non-allergen	1.393
KAMDEQIK	Non-Toxin	Probable non-allergen	1.2589
QNNRDLSSSLEKLSSGLRINKAAD	Non-Toxin	Probable non-allergen	0.6929

3.3. Prediction of MHC class I-restricted cytotoxic T lymphocyte (CTL) epitopes

The MHC-I binding epitopes were predicted through the IEBD server for all available HLA alleles. Among a number of MHC-I predicted epitopes, 10 epitopes were selected as highly potent vaccine candidates. Subsequently, two epitopes were selected based on the average of IEDB score and antigenicity score, and the vaccine's desired structure. The IEDB class I immunogenicity tool determined the vaccine's immunogenicity score.

We calculated the average score of the antigenic and IEDB scores. In conclusion, the effective epitopes were selected based on characteristics such as high score (good binder), and the average of IEDB and antigenicity scores, antigenic (Table 3).

Table 3.

MHC-I epitopes selected based on high average score, immunogenicity assessment, non-toxic nature.

Epitope	HLA allele	IEDB-score	Antigenicity-score	Average	Toxicity	Allergecity
EASMDIQGR	HLAA∗68:01	0.897427	1.828100	0.806743269	Non-toxin	Probable non-allergen
KLSSGLRINK	HLA-A∗03:01	0.938462	1.086200	0.681961121	Non-toxin	Probable non-allergen

3.4. MHC-II binding epitopes prediction

MHC-II binding epitopes were predicted using the IEDB server for all HLA alleles. As shown in Table, two epitopes were selected with a length of 14-18 amino acids. All of the eight epitopes had good scores with the average of IEDB score and antigenicity score. Furthermore, toxic and allergen assessment were performed to confirm that the potential vaccine does not cause toxicity and allergy in humans (Table 4).

Table 4.

MCH-II epitopes selected for vaccine structure based on the average of the IEDB score and antigenicity score.

Epitope	HLA allele	IEDB-score	Antigenicity-score	Average	Toxicity	Allergecity
NANGAQAETNSQGI	HLA-DQA1∗05:01/DQB1∗03:01	0.6417	2.0606	0.754386714	Non-toxin	Probable non-allergen
FRINTNIAALTSHAVGVQ	HLA-DRB1∗01:01	0.9376	1.1757	0.716159562	Non-toxin	Probable non-allergen

3.5. Design and assembly of a potent vaccine candidate

A total of two MHC-I and two MHC-II epitopes and three B-Cell epitopes were assembled to design the multi-epitope vaccine. Selection criteria for final epitopes are based on their score average, taking into account antigenicity, allergenicity, and toxicity. The sequences AAY, GPGPG and KK as linkers merged MCH-I, MCH-II and B-Cell epitopes. CTxB as an adjuvant joined through EAAAK linker to vaccine construct. The vaccine construct was designed to include specific epitopes and visualized using SnapGene software (version 3.2.1) (Fig. 1). Finally, a histidine tag was placed at the end of the vaccine sequence. The ultimate successful design of the multi-epitope vaccine is presented in Fig. 1.

Fig. 1.

Schematic diagram of potent vaccine candidate sequence including two MHC-I and two MHC-II epitopes and three B-Cell epitopes linked together by EAAAK linker.

3.6. Prediction of antigenic features, allergenic tendencies, solubility behavior, and toxic effects in protein sequences

The antigenic potential of the designed protein was assessed using the VaxiJen v2.0 server. The analysis revealed that the vaccine candidate exhibited a strong likelihood of eliciting an immune response, with an antigenicity score of 0.8561. To evaluate allergenicity, the AllerTOP v.2 web server was used, confirming that the designed protein was non-allergenic. Additionally, Toxicity assessment was performed using the ToxinPred2 server, and the results indicated that the vaccine candidate was non-toxic. In addition, the solubility of the vaccine candidate was predicted using Protein-Sol, and it was calculated that the designed protein is soluble, with a score of 0.632.

3.7. Physicochemical properties

Evaluating the physicochemical properties of a vaccine is critical to assessing its efficacy and safety. The ProtParam server was used to predict various chemical and physical properties, as summarised in Table 5. The vaccine candidate consists of 259 amino acids with a molecular weight (MW) of 28382.35 Da. The instability index was calculated to be 27.02, predicting a stable nature of the vaccine candidate. Proteins with an instability index below a certain threshold (typically 40) are considered to be stable in vivo. The predicted isoelectric point (pI) of the vaccine is 9.63. The estimated half-life of the vaccine is greater than 30 h in mammalian cells, greater than 20 h in yeast and 10 h in E. coli. The grand average of hydropathicity (GRAVY) was calculated to be −0.435, with negative values indicating hydrophilicity and positive values indicating hydrophobicity. Additionally, the aliphatic index of 80.73 reflects the protein's thermal stability, with higher values indicating greater stability. This analysis and prediction highlights the favorable physicochemical properties of the vaccine candidate, supporting its potential stability and efficacy.

Table 5.

Prediction of physical and chemical properties for the designed vaccine.

Physical and chemical properties	Value
Molecular weight	28382.35
Theoretical pI	9.63
Total number of vaccine amino acid	259
The estimated half-life (mammalian reticulocytes, in vitro)	30 h
The estimated half-life (yeast, in vivo)	>20 h
The estimated half-life (Escherichia coli, in vivo)	>10 h
The instability index	27.02
This classifies the protein	stable
Aliphatic index	80.73
Grand average of hydropathicity (GRAVY)	−0.435

3.8. Prediction of secondary and tertiary structure of vaccine construct

The study and predict of the secondary structure of the designed vaccine reveals that it contains 43.63% alpha helix (α-helix), 15.06% extended strand (Extended Strand), and 41.31% random coil (Random Coil). To study the 3D structure of the vaccine construct, I-TASSER and the Robetta server were used. Predicted tertiary structures were evaluated using MolProbity, ProSA-web and SAVES v6.1 servers. The best model was selected. The MolProbity analysis yielded a Clash score of 5.49 for the protein, positioning it in the 92nd percentile compared to other structures, reflecting its high structural integrity. The MolProbity score of 1.56 further reinforced its reliability, placing it in the 94th percentile. The three-dimensional model of the vaccine candidate was evaluated using ProSA-web, which assessed the energy plot and Z-score. The protein's Z-score of −5.80 was within the typical range of 10 to −20 observed for native protein structures, as shown in Fig. 2. Analysis of the Ramachandran plot revealed that 99.60% of the amino acids fell within preferred and allowed regions, with only 0.40% in disallowed regions, underscoring the plausibility of the predicted structure. In addition, ERRAT analysis, a tool that detects irregularities in protein structures by analyzing deviations in atomic distributions, yielded a score of 99.19% for the modelled structure, indicating its high quality (Fig. 3). Taken together, these results confirm the reliability and accuracy of the predicted tertiary structure (Table 6).

Fig. 2.

Structural prediction of the designed vaccine construct: (A) predicted secondary structure elements (α-helices, β-sheets, and coils) and (B) three-dimensional model showing the spatial arrangement of vaccine components.

Fig. 3.

Tertiary structure prediction of the vaccine candidate through computational analysis: (a) structural integrity assessment using ProSA-web, (b) stereochemical quality evaluation via Ramachandran plot, and (c) atomic-level prediction employing the ERRAT algorithm.

Table 6.

Prediction and evaluation of the secondary structure of the designed vaccine.

MLRC
Alpha helix	(Hh)	113 is 43.63%
310 helixes	(Gg)	0 is 0.00%
Pi helix	(Ii)	0 is 0.00%
Beta bridge	(Bb)	0 is 0.00%
Extended strand	(Ee)	39 is 15.06%
Beta turn	(Tt)	0 is 0.00%
Bend region	(Ss)	0 is 0.00%
Random coil	(Cc)	107 is 41.31%
Other states		0 is 0.00%

3.9. Molecular docking

Studying the interaction between the vaccine and immune receptors is an essential step in vaccine design. Molecular docking was studied using ClusPro 2.0 and HADDOCK web serever. The interaction between the designed vaccine candidate and TLR4 and TLR2 receptors was analyzed using ClusPro 2.0 and the HADDOCK web server. Criteria such as lowest energy, most negative weighted score, and highest number of binding residues, especially at the active sites, were prioritized in ClusPro 2.0 web server. The HADDOCK web server sorts structures according to the HADDOCK score. The results showed strong binding interactions between the vaccine candidate and the active sites of TLR2 (Fig. 4) and TLR4(Fig. 5). ClusPro 2.0 calculated weighted scores were −1561.2 for the vaccine-TLR2 complex and −922.5 for the vaccine-TLR4 complex. The HADDOCK web server calculated HADDOCK scores were −135.6±5.7 and −76.9±12.9 for vaccine-TLR2 and vaccine-TLR4 complex.

Fig. 4.

Molecular docking analysis of the vaccine-TLR2 interaction: (a) 3D structural representation of the vaccine construct (red) bound to human TLR2 (green), and (b) atomic-level interaction profile between TLR2 chain-A and the vaccine's chain-B, highlighting key binding residues.

Fig. 5.

Structural analysis of the vaccine-TLR4 docking complex: (a) three-dimensional representation showing the vaccine construct (red) bound to human TLR4 (blue), and (b) atomic-level interaction mapping between TLR4's chain-Z and the vaccine's chain-A, revealing critical binding interfaces.

The PRODIGY web server and PDBsum were used to analyze the interaction between vaccine candidates with TLR2 and TLR4. The output of two web servers calculated binding energy (ΔG) and optimal dissociation constant (Kd) (Table 7). The best model was selected for the next steps.

Table 7.

Evaluating molecular binding results between the designed vaccine candidate and TLR2 and TLR4.

Complex	Method	ΔG (kcal mol-1)	PRODIGY (Kd)
Vaccine candidate–TLR2	Cluspro	−12.3	9.4e−10
Vaccine candidate–TLR2	HADDOCK	−11.5	3.6E−9
Vaccine candidate–TLR4	Cluspro	−21.7	1.3e−16
Vaccine candidate–TLR4	HADDOCK	−10.3	2.7e-8

Key residues involved in the interaction included Ile319, Phe322, Phe325, Tyr326, Val348, and Phe349. Residues Pro352 for TLR2, Arg380, Lys341, Lys263, and Gln339 are also important for interacting with TLR4.The interacting residues were visualized using PyMOL (PyMOL Molecular Graphics System, version 1.1, Schrodinger, LLC) and further analyzed using the PDBsum web server.

3.10. Molecular dynamics

To evaluate the stability of the designed protein structures and their complexes with receptors, MD simulations were performed for a duration of 150 ns. The unbound TLR2 and TLR4 receptors remained stable during the 150 ns simulation, with average RMSD values of 0.316 nm and 0.244 nm, respectively. The designed vaccine protein itself also demonstrated stability, converging with an average RMSD of 0.605 nm. Following complex formation, the vaccine-TLR2 complex reached equilibrium after approximately 20 ns and maintained a stable trajectory with a mean RMSD of 1.006 nm and low fluctuations. Similarly, the vaccine-TLR4 complex was stable throughout the simulation, exhibiting an average RMSD of 0.428 nm (Fig. 6A). In conclusion, all systems demonstrated stability with low fluctuations during the simulation time. Rg, a parameter used to assess compressibility changes during MD simulations, was also analyzed. Rg measures the distribution of protein atoms around their axis and provides insight into the overall dimensions and compactness of the protein. Stable Rg values indicate consistent protein folding and complex stability. As shown in the graph, the designed protein, as well as its complexes with TLR4 and TLR2, maintained stable compression levels throughout the simulation (Fig. 6B).

Fig. 6.

Molecular dynamics analysis of vaccine-receptor complexes: (A) root-mean-square deviation (RMSD) trajectories, (B) radius of gyration (Rg) fluctuations, and (C) root-mean-square fluctuation (RMSF) profiles for the engineered vaccine alone, TLR2/TLR4 alone and in complex with TLR2/TLR4 receptors. (D) Number of H-bonds in the vaccine-TLR2 and vaccine-TLR4 complexes.

Finally, RMSF analysis provides critical insights into the dynamic behavior and structural stability of the vaccine candidate both in its free state and when bound to the immune receptors TLR2 and TLR4. The results clearly show that this candidate vaccine interacts stably and significantly with both TLR receptors, but the nature of these interactions has key differences.

The candidate vaccine in the free state has favorable structural stability, with controlled flexibility in peripheral regions essential for molecular interactions. Complexation with TLR4 results in a remarkably stable structure, with a significant reduction in fluctuations in the binding interface. This high stability is due to the optimized binding mechanism and the presence of a network of strong molecular interactions in this complex. In contrast, the complex with TLR2, while still showing acceptable stability, has greater flexibility in some functional regions (Fig. 6C). Taken together, these results demonstrate the stability and robustness of the designed proteins and their complexes during simulation. Analysis of hydrogen bond formation during the 150 ns molecular dynamics simulation provided important insights into the stability and nature of the binding interactions between the designed vaccine candidate and the TLRs. The results indicated a stronger hydrogen bond network with TLR4 compared to TLR2. The vaccine-TLR4 complex maintained a consistently higher average of 13.88 hydrogen bonds throughout the simulation. In contrast, the vaccine-TLR2 complex sustained an average of approximately 11.15 bonds (Fig. 6D). Furthermore, both complexes exhibited low fluctuations, confirming a stable and specific binding interface.

3.11. Binding-free energy calculation

The binding free energies between the vaccine and TLR4 and TLR2 were evaluated by the Hawdock web server, predicted with the MM/GBSA method. The total binding energy is divided into electrostatic, Van der Waals, SASA and Gibbs energies. The binding free energy calculation result showed that the Van der Waals, SASA, and electrostatic energies were favorable energies in the interaction between vaccine-TLR receptors, but the Gibbs energy had a negative role in the interaction between vaccine-TLR receptors. The vaccine protein showed greater binding affinity to TLR4 than to TLR2, as indicated by the lower binding energy of the vaccine-TLR4 complex (Table 8).

Table 8.

Binding-Free Energy Calculations for Vaccine-TLR4 and Vaccine-TLR2 Complexes with the MM/GBSA Method. This result showed that Van der Waals, Electrostatic, Gibbs, and SASA energies have a critical role in the interaction between the vaccine and TLR2 and TLR4. A lower binding energy in the vaccine-TLR4 complex implies higher binding affinity for TLR4 than for TLR2.

Energy	TLR2-vaccine complex	TLR4-vaccine complex
van der Waals energy (KJ/mol)	−120.85	−270.04
Electrostatic energy (KJ/mol)	−683.95	−1398.71
Gibbs energy (KJ/mol)	753.19	1527.15
SASA energy (KJ/mol)	−13.57	−37.86
Total energy (KJ/mol)	−65.18	−179.46

3.12. Immune simulation

The immune response elicited by the vaccine candidate was simulated using the C-ImmSim server, a computational tool designed to model the dynamics of the immune system. The simulation was performed to mimic a realistic immune response over several weeks. Key immune parameters, including antibody production, cytokine levels, and immune cell populations, were analyzed to evaluate the immunogenicity and efficacy of the vaccine candidate. An increase in IgM levels indicates the primary host response to the pathogen. In addition, the secondary response to the designed protein as an antigen is measured by an increase in IgG1 and IgG2 levels (Fig. 7a), the B cell population (Fig. 7b), and the TH cell population (Fig. 7c). The results of our in silico immune simulation demonstrated the vaccine's potential to induce a strong immune response. Specifically, the model predicted a significant rise in the levels of various cytokines and interleukins, with a pronounced increase in IFN-γ post-immunization (Fig. 7d).As shown, IFN-γ levels increased after three injections. Memory responses are also essential for long-term immunity in vaccine design, and our results have shown that the levels of memory B and T cells are maintained at high levels.

Fig. 7.

Computational immune simulation of the vaccine candidate using C-ImmSim server: Immunological profiling following three-dose administration (timepoints: 1, 336, 672) showing (a) antigen-antibody dynamics, (b) B-lymphocyte population kinetics, (c) helper T-cell (TH) proliferation, and (d) cytokine secretion patterns.

3.13. Codon optimization and in silico cloning of the designed candidate vaccine

Codon optimization was carried out using the JCat tool. Following this process, the optimized sequence of the designed structure measured 891 nucleotides in length. The optimized sequence exhibited a codon adaptation index (CAI) of 1 and a GC content of 47.74%. To verify clonability, the optimized vaccine sequence was simulated in the pET-28a (+) vector using SnapGene software, confirming its compatibility (Fig. 8. Additionally, double digestion with NcoI and XhoI restriction enzymes revealed the presence of the vaccine candidate fragment (785 bp) alongside the pET-28a (+) vector (5231 bp) (Fig. 8b.).

Fig. 8.

a. Cloning of the designed protein gene into the pET-28a vector (shown in green), 8b. Informatics evaluation of the cloning of the designed protein by double digest.

4. Discussion

H. pylori is a significant public health issue, infecting around 50% of the global population [47]. The International Agency for Research on Cancer (IARC) has categorized it as a definitive group 1 carcinogen due to its strong association with the development of gastric cancer [48]. There is a clear need for an effective vaccine against H. pylori, as it can help to prevent gastric inflammation and reduce the risk of gastric cancer [[49], [50], [51], [52], [53]]. While antibiotic treatments for H. pylori infection are available, they present several challenges, including relapse, increased antibiotic resistance, and disruption of the gut microbiome [54]. Vaccines represent a safer and more efficacious alternative to conventional approaches, which are often limited by low efficacy and high production costs [51,[55], [56], [57]]. To overcome these challenges, reverse vaccinology has emerged as an innovative strategy for the rational design of highly targeted subunit vaccines [58,59]. By leveraging genomic and proteomic data of microbes, screening and evaluating potential vaccine antigens have become more feasible [60]. Through reverse vaccinology, researchers can identify and assess vaccine candidates based on the genetic and protein information of pathogens, such as H. pylori. This approach holds promise for developing effective and affordable subunit vaccines against this global health threat [51,56,61,62]. In addition to discovering new protective antigens using computational methods, bioinformatics programs can accurately identify immunodominant epitopes (B and T cell epitopes) crucial for humoral and cellular immunity. Previous research focused on immunoinformatic to design a novel multiepitope vaccine against H. pylori, emphasizing the importance of these epitopes in vaccine development [51,56,61,63,64]. In this study, by analyzing the genome of the pathogenic organism, potential epitopes were identified that can elicit both T-cell and B-cell immune responses. The use of multiple epitopes in the vaccine design aims to enhance overall immune protection against H. pylori infection. From the initial pool of 8209 H. pylori ORFs encoding surface-exposed and secreted proteins, we prioritized candidates based on immunogenicity scores. Our comprehensive immunoinformatic pipeline included: (1) MHC class I/II binding affinity predictions to identify potential T-cell epitopes, and (2) systematic screening of B-cell epitopes. This multi-parametric approach enabled the identification of the most promising vaccine targets that could potentially elicit both humoral and cellular immune responses against H. pylori infection.

Ehensively considered in the evaluation of epitopes. ORF15 (Flagellin B) was selected based on immunogenicity assessment and non-toxic nature related to extracellular ORFs without signal peptides, with an antigenicity score of 0.7911 to find B-cell, MHC-I and MHC-II epitopes. The results demonstrated a great antigenic nature, non-toxic and non-allergenic properties.

Flagellin B (FlaB) in H. pylori plays several key roles in bacterial motility, colonization, and pathogenesis. The protein is essential for the assembly of the bacterial flagellum, which enables H. pylori to move through the viscous gastric mucus layer and reach the epithelial surface. Additionally, FlaB contributes to the bacterium's ability to establish persistent infection by facilitating adherence to host cells and modulating immune responses. Its structural properties make it a potential target for vaccine development and therapeutic interventions against H. pylori infections [51,[65], [66], [67]]. It is a minor structural component of the flagellar filament, located proximal to the hook, contributing to the overall structure and stability of the flagella. FlaB is essential for proper flagella formation and bacterial motility, with mutant strains lacking it exhibiting irregular flagella and reduced colonization abilities. Like the major flagellin FlaA, FlaB helps evade recognition by Toll-like receptor 5 (TLR5), aiding H. pylori in avoiding the host innate immune response. The FlaB gene is regulated by a distinct promoter from FlaA, enabling differential expression and control of flagellar assembly. Despite being less abundant than FlaA, FlaB is crucial for maintaining proper flagellar structure and function, supporting H. pylori's colonization and immune evasion strategies [66,[68], [69], [70], [71]].

We chose Flagellin B (FlaB), also identified as ORF15, as the target for their vaccine design after a multi-step screening process. Starting with 8209 potential protein sequences (ORFs) from the Helicobacter pylori genome, they first selected proteins located outside the cell or on its outer membrane. Among these candidates, Flagellin B was selected because it demonstrated a high antigenicity score of 0.7911, indicating its strong potential to provoke an immune response. Furthermore, computational analysis confirmed its safety, predicting it to be non-toxic and non-allergenic. Several studies have investigated FlaA and FlaB of Helicobacter pylori as vaccine or diagnostic targets. A DNA vaccine based on the FlaA gene showed that FlaA, the major flagellin subunit, can induce strong immune responses in mice, supporting its potential as a vaccine antigen [72]. Another immunoinformatic study designed a multi-epitope peptide vaccine incorporating epitopes from HpaA, FlaA, FlaB, and Omp18, which was predicted to be non-toxic, non-allergenic, and able to bind multiple MHC alleles, demonstrating that combining FlaA and FlaB with other virulence proteins can provide broader epitope coverage [51]. In addition, a study on an antigenic determinant of FlaA (bp 1345–1395) identified a highly immunogenic fragment with diagnostic value, showing sensitivity and specificity of ∼90% for IgG and 100% for IgM in ELISA, suggesting FlaA's utility as a serological marker [73].

Subunit vaccine design has progressed significantly from single-antigen approaches to sophisticated multi-epitope platforms, offering distinct immunological advantages. The multi-epitope strategy employed in this study addresses key limitations of conventional vaccines by simultaneously eliciting broad-spectrum neutralizing antibodies and blocking multiple pathogenic pathways through targeted epitope combinations [51,61,[74], [75], [76]]. Building on these principles, we engineered a novel chimeric antigen by systematically arranging identified epitopes into optimized configurations, then evaluated its immunogenic potential in combination with a selected adjuvant to enhance both cellular and humoral immune responses.

Recent preclinical studies have demonstrated that cholera toxin subunit B (CTxB), an established mucosal adjuvant incorporated in H. pylori multiepitope vaccine designs, significantly boosts both immunogenic potency and antigenic recognition of the vaccine constructs [54]. H. pylori infections typically initiate at the human mucosal surface, making oral or intranasal routes of administration more suitable. Although epitope-based vaccines are weakly immunogenic when administered mucosally, robust mucosal protein adjuvants can significantly enhance their immunogenicity [61,68,[77], [78], [79], [80]].

CT is a well-known bacterial toxin with strong mucosal adjuvanticity, commonly used in animal models for anti-H. pylori vaccines. The CTxB-adjuvanted multiepitope construct forms a fusion protein rather than a de novo protein entity. This adjuvant incorporation can induce conformational modifications in the antigenic structure, potentially compromising the reliability of prediction. Our findings suggest that adjuvant integration may be optimally performed post multiepitope vaccine development to preserve structural fidelity [68,80,81].

In the design of a multiepitope vaccine, we applied two epitopes of MHC-I, two epitopes of MHC-II and three epitopes of B cells connected by different linkers and interacted by adjuvant. At the end of the sequence, His-tag was added for the purification of the produced vaccine by chromatography. The 6xHis tag is a helpful tool for the purification, identification, and quantification of protein-protein interactions within the candidate vaccine model. It allows for the efficient purification of proteins and their interactors, facilitating the analysis of protein structures and their interactions. To improve the immunogenicity of the epitope vaccine, flexible linkers (GPGPG) were applied to link MHC-II epitopes, preserving their structural features and conformational immunogenicity. Bioinformatics analysis revealed that the multiepitope peptide vaccine is highly antigenic, non-allergenic, stable, and water soluble, making it an effective and safe vaccine candidate. To prevent the formation of neo-epitopes from epitope linkage sites and enhance epitope presentation, a KK linker was inserted flanking the immunodominant epitopes. The lysine linker (–KK–) can target the lysosomal protease, cathepsin B, which is the primary enzyme responsible for MHC-II antigen presentation. Moreover, the AAY (Ala-Ala-Tyr) linker serves as a proteasomal cleavage site in mammalian cells. This means that epitopes linked using AAY are effectively separated within cells, reducing junctional immunogenicity. Additionally, the AAY linker enhances the immunogenicity of the multi-epitope vaccine by facilitating the processing and presentation of individual epitopes to the immune system. The EAAAK peptide linker was used to join the adjuvants to the C- and N-termini of the designed vaccine. EAAAK is a rigid α-helix forming linker with intramolecular hydrogen bonding, resulting in a compact backbone structure [[82], [83], [84], [85]]. Compared to flexible linkers, rigid linkers, such as EAAAK, have several advantages. They maintain a fixed distance between functional domains, minimizing interference between epitopes and preserving their individual properties. This allows for efficient separation of domains in a bifunctional fusion protein, ensuring that each domain functions effectively. By incorporating these linkers, the vaccine design aimed to optimize epitope processing and presentation to the immune system, potentially leading to improved immunogenicity and efficacy [84,[86], [87], [88]].

The multi-epitope vaccine designed against Helicobacter pylori was evaluated for its antigenicity, allergenicity, and physicochemical properties. The vaccine had a molecular weight of 28,382.35 Da, with a pI of 9.63, indicating its basic nature. Its solubility, enhanced by antigenicity, was also assessed. The protein showed a low instability index of 27.02, confirming stability suitable for vaccine development. The aliphatic index (80.73) suggested good thermal stability, while the negative GRAVY score (−0.435) indicated hydrophilicity. The secondary structure consisted of 43.63% helix, 15.06% extended strand, and 41.31% random coil. The tertiary structure was validated with the Ramachandran plot, showing 99.6% of residues in preferred regions. Docking studies revealed a stable interaction and strong binding affinity to TLR4 (−1076.2 kcal/mol) compared to TLR2 (−343.9 kcal/mol), suggesting effective immune response initiation. The results showed that the vaccine can effectively bind to its targets, with a negative ΔG, indicating strong potential for immune activation via Toll-like receptors (TLRs), which play a crucial role in recognizing pathogens like H. pylori. The vaccine is predicted to be immunogenic, non-allergenic, and stable, making it a promising candidate for further development. In a Man Cui et al. study, a multi-epitope vaccine against Helicobacter pylori was designed by linking epitopes from nine target proteins using specific linkers: GPGPG for B-cell, AAY for CTL, and KK for HTL epitopes, enhancing epitope presentation and preventing neo-epitope formation. The vaccine's physicochemical properties were analyzed, revealing a molecular weight of 49.47 kDa, a pI of 9.7, and an instability index of 15.98, indicating stability. The aliphatic index (71.44) suggested thermostability, and a GRAVY score of −0.423 indicated hydrophilicity. The vaccine was predicted to be soluble with a solubility score of 0.498. Secondary structure analysis showed 30.13% α-helix, 21.40% extended strand, 7.86% β-turn, and 40.61% random coil. Tertiary structure modelling confirmed 89.2% of residues in favored regions of the Ramachandran plot, with a ProSA z-score of −5.15, indicating reliability. Docking studies indicated strong binding affinities to TLR2 (−343.9 kcal/mol) and TLR4 (−1076.2 kcal/mol), suggesting effective immune response initiation. Immunogenicity scores from VaxiJen (0.9674) and ANTIGENpro (1.0720) exceeded the threshold, while allergenicity and toxicity analyses confirmed the vaccine to be non-allergenic and non-toxic. These findings support the vaccine's potential for further development [51].

The complex was then studied using MD simulation to assess the stability of the vaccine-receptor complex. The dynamic nature of MD simulation allowed the vaccine candidate to adopt optimal conformations and interactions. The simulation results confirmed the stability of the complexes, indicating long-lasting interactions between the receptors and the designed vaccine. It is worth noting that additional bacterial and animal studies are necessary to determine whether the vaccine we developed can elicit a specific immune response against H. pylori infection. Then, MD simulation was conducted to assess the stability and behavior of the vaccine-receptor complex over time. This analysis provides insights into the long-term interactions and potential efficacy of the vaccine in a dynamic environment. The MD simulation confirms that the structure exhibits strong stability, compactness, and effective interactions, indicating that the vaccine is likely to induce a robust immune response when binding to these receptors.

5. Conclusion

This study presents a rationally designed multi-epitope vaccine candidate against H. pylori with robust computational prediction. Through rigorous immunoinformatics approaches, we identified conserved B- and T-cell epitopes from the key virulence factor Flagellin B, demonstrating strong antigenicity and non-allergenic properties. Molecular docking revealed stable interactions with HLA alleles and TLRs, while 150 ns MD simulations confirmed complex stability. The constructed vaccine exhibited favorable physicochemical properties, solubility, and induction of both humoral and cellular immune responses in silico approach.

While the insilico analyses presented in this study are promising, it is essential to conduct rigorous experimental validation in animal models and human subjects to confirm the safety, efficacy, and immunogenicity of the vaccine before it is translated into clinical practice. This research represents a significant step towards addressing the global burden of H. pylori-related diseases and underscores the potential of immunoinformatic in vaccine development against infectious pathogens.

Ethical approval

Not applicable (This study is based on computational analysis and does not involve human or animal subjects).

Consent for publication

Not applicable.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement

Nasim Rahimi-Farsi: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software. Fatemeh Bostanian: Conceptualization, Data curation, Writing – original draft. Khadijeh Ahmadi: Software, Validation, Visualization, Writing – original draft. Alireza Zangooie: Conceptualization, Writing – review & editing. Mahsa Sedighi: Conceptualization, Writing – review & editing. Behzad Shahbazi: Investigation, Project administration, Supervision, Validation. Negar Mottaghi-Dastjerdi: Project administration, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

These data are available from the corresponding author upon reasonable request.