Use of immunoinformatics and the simulation approach to identify Helicobacter pylori epitopes to design a multi-epitope subunit vaccine for B- and T-cells

Background Helicobacter pylori cause a variety of gastric malignancies, gastric ulcers, and cause erosive diseases. The extreme nature of the bacterium and the implantation of this bacterium protects it against designing a potent drug against it. Therefore, employing a precise and effective design for a more safe and stable antigenic vaccine against this pathogen can effectively control its associated infections. This study, aimed at improving the design of multiple subunit vaccines against H. pylori, adopts multiple immunoinformatics approaches in combination with other computational approaches. Results In this regard, 10 HTL, and 11 CTL epitopes were employed based on appropriate adopted MHC binding scores and c-terminal cut-off scores of 4 main selected proteins (APO, LeoA, IceA1, and IceA2). An adjuvant was added to the N end of the vaccine to achieve higher stability. For validation, immunogenicity and sensitization of physicochemical analyses were performed. The vaccine could be antigenic with significantly strong interactions with TOLK-2, 4, 5, and 9 receptors. The designed vaccine was subjected to Gromacs simulation and immune response prediction modelling that confirmed expression and immune-stimulating response efficiency. Besides, the designed vaccine showed better interactions with TLK-9. Conclusions Based on our analyses, although the suggested vaccine could induce a clear response against H. pylori, precise laboratory validation is required to confirm its immunogenicity and safety status.


Background
As a helical-shaped bacterium, Helicobacter pylori (H.pylori) is a gram-negative bacterium mostly found on the inner surface of the stomach.H. pylori is the most common cause of multiple infections, such as gastric carcinoma, peptic ulcers, mucosa-associated lymphoid tissue (MALT), and peptic ulcers [1].Based on genetic composition analysis, a notable heterogeneity was observed in the gene sequence and content of the H. pylori bacterium.
Besides, as a versatile gene reservoir, it could be easily used as an eligible tool for bacterial adaptation when new conditions occur in different human hosts.The presence of H. pylori could be quickly approved through the continuous induction of inflammatory responses as well as tissue damage that can cause extra critical clinical disorders, such as peptic ulcers (PUs), chronic gastritis (CG), osteoporosis (OP), and gastric cancer [2,3].Given the dangers of H. pylori, controlling the infection with antibiotics may reduce the incidence of the disease.Nowadays, Helicobacter pylori infections can be treated using some antibiotics; however, there are some limiting factors to these treatments, such as high-cost het, patient reluctance to take the drug, and development of severe antibiotic resistance [4].Therefore, there is a need for alternative treatment or prevention methods, such as vaccination.Attempts to develop a vaccine against H. pylori began in the early 1990s, with the knowledge that the infection caused by this bacterium was the leading cause of gastric ulcers.Despite research on the H. pylori vaccine with acceptable results in animal models and cell cultures, no vaccine has been licensed for human use [5,6].Based on recent reports, some of these factors may increase the risk of infections.Virulence factors of H. pylori could result in the production of pro-inflammatory cytokines dominant in gastric disorders and malignant proteins [1].
The lnT gene (Apolipoprotein N-acyltransferase) was first discovered in the gram-negative bacterium Salmonella enterica.However, the structure and lnT function were later studied in more detail in Escherichia coli.1-3% of the bacterial genome contains genes that encode lipoproteins in the cell membrane.These lipoproteins can play vital roles, such as nutrient uptake, protection of cell wall integrity, protein secretion, cytoplasmic folding of proteins, and pathogenicity [7].Lipoproteins are the significant components of the cell lining, being responsible for many essential functions of the cell.Powerful covalent bonds produce these components after translations in the context of "lipids," which are created by a sequential process consisting of three steps and controlled by three integral membrane enzymes.The final step of this process, creating the final mature lipoprotein, is the N-acylation of the terminal cysteine by Apolipoprotein N-acyltransferase (Lnt).This step is mainly unique to gram-negative bacteria.Synthesized proteins are prolipoproteins transported to the cytoplasmic membrane by the Sec (General secretion) or Tat (twin-arginine translocation) general secretory machinery [8].
The lnT gene encodes one of the significant coat lipoproteins of Helicobacter pylori, which stimulates immunogenicity in the host cell, producing antibodies.The cloning and antigenic properties of the lnT gene have been studied in other gram-negative bacteria, such as E. coli and S. enterica [9].The leoA gene (labile enterotoxin output A) is one of the pathogens of H. pylori GTPase, which is encoded on pathogenicity islands and potentially increases toxin release through secretory vesicles.LeoA is known as a bacterial diamine-like protein.Diamine family proteins alter membranes and generally evolve by releasing heat-sensitive enterotoxin through the membrane vesicle, released from the bacterial cell's surface.There are few reports on the functional role of bacterial diamine proteins [10].Membrane vesicles play a vital role in the protected function of the diamine family members.Reports show that this gene has the potential for immunization in the host body.Therefore, it has been introduced as a confirmed candidate for the assembly of a gene vaccine against H. pylori [11,12].
After the isolation of H. pylori bacteria from peptic ulceration disease (PUD) and gastritis patients, the iceA gene was identified.Helicobacter pylori iceA1 and iceA2 genes are the two main alleles of iceA [13].After bonding H. pylori to human epithelial cells, the expression of iceA1 was upregulated.The iceA1 genotype caused both mucosal IL-8 expression and acute antral inflammation.The bonding of iceA1 to gastric epithelial cells in vitro stimulates its transcription [14].Many studies have proposed an association between the iceA1 variant and PUD.However, no homology has been found between iceA2 and other known genes.Although iceA2 can cause asymptomatic gastritis and non-ulcer dyspepsia, its function has remained unknown [15].More than 8% of patients had this gene in connection with gastritis, more than 8% had this gene in combination with mixed lesions, and more than 6% had the iceA1 gene in combination with peptic ulceration.However, the iceA2 gene was less prevalent than the iceA1 gene in patients with mixed lesions (10%), while gastritis and peptic ulceration were prevalent in 12.5%.None of the iceA1 and iceA2 genes was prevalent in patients with gastric carcinoma.Unexpectedly, in patients with gastric carcinoma, iceA1, and iceA2 genes were more prevalent.The higher rate of the prevalence of these genes in these patients could be a sign warning against the higher risk of developing gastric carcinoma.There was a lower prevalence of positive cases of the two iceA alleles among patients with gastritis, ulcers, and mixed lesions [16].Immunoinformatics is one of the fastest, most precise, and most reliable approaches to developing vaccines against virulent pathogens.Given the advantages and feasibility of designing vaccines using immunoinformatics approaches [1], our study aimed to design a multi-epitope subunit vaccine against H. pylori.The present study used a wide range of antigenic proteins from the H. pylori proteome to design B-and T-cell epitopes.For this purpose, MHC-II binding epitopes (HTL) were used as a predictive method, and a vaccine composed of many epitopes was produced in the end.Additionally, molecular docking for the drug design, molecular dynamics for stability profiling, ''in silico expression analysis'' for validating short DNA sequences, and agent-based computational modelling were employed to check the immune response reaction and stability stimulated by the final vaccine produced.

Retrieval of Helicobacter pylori protein sequences
The amino acid sequences of leoA, IceA1-2, and APO of H. pylori were retrieved.This bacterium was retrieved from the gene bank to develop a multi-epitope subunit vaccine to evoke immune reactions against the H. pylori infection.These proteins were selected based on their reliability and target efficacy.

Antigenicity of selected H. pylori proteins
The antigenicity of the proteins was determined using the Vaxijen2.0server.Accordingly, the results showed that all selected proteins in this study had antigenic attributes.This server calculated scores 0.4356, 0.3076, 0.5319, and 0.4370 for APO, leoA, IceA1, and IceA2, respectively [17].

Prediction of cytotoxic and HTL epitopes
Based on NetCTL 1.2, 37 CTL epitopes were anticipated for the selected proteins, of which only 13 epitopes with a non-allergenic nature and a binding score were considered for this study.Similarly, using the IEDB MHC-II server resulted in CTL epitopes for the selected proteins, as Table 1 (bold columns) shows.

Vaccine production
To design a unique vaccine meeting all existing standards regarding nontoxicity, antigenicity, non-allergenicity, and binding affinity, 4 HTL epitopes, four lymph cell epitopes, and 10 CTL epitopes were nominated.Also, human β-defensins were introduced to an adjuvant with GI: 159162606.Human beta-defensins are a family of epithelial cell derived antimicrobial peptides (AMPs) that protect mucosal membranes from microbial challenges.Therefore, it can be used as the best adjuvant to enhance the vaccine's performance.The defensive mechanisms of β-defensins include binding to negatively charged microbial membranes that cause cell death and chemoattraction of immune cells.
The ultimate multi-epitope vaccine was produced from 10 CTL and 13 HTL epitopes (shown in bold letters in Table 1) by joining a couple of AAY and GPGPG linkers.Besides, the adjuvant was attached to N-terminals to avoid vaccine degradation.Additionally, GPGPG and AAY linkers joined the HTL and CTL epitopes, while the EAAAK linker tended to bond with the adjuvant of the CTL epitopes.Figure 1A shows the ultimate vaccine structure.

Prediction of B cell epitopes
To predict B-cell epitopes by the BCPred server, default parameters were applied.Accordingly, the server predicted 14 linear B-cell epitopes with a length of 20 amino acids and a score more excellent than 0.90. Figure 1B shows the predicted linear and conformational epitopes.

Calculation of physiochemical properties
Some essential features were quantified, including allergenicity, instability, half-life, and relative molecular mass.The server calculated a score of -0.42 for the allergenicity of the last vaccine construct, thereby identifying the vaccine as non-allergenic.The default threshold was set as -0.4.The isoelectric point was determined to confirm the vaccine's fundamental nature.In addition, the server calculated a PI score of 9.23 and a mass of 38.718 kDa (MW).On the other hand, the in vitro half-life of the yeast E. coli (in mammals' reticulocytes) was calculated at > 30, > 20, and > 10 h.In addition, an instability index of 23.60 was obtained, confirming that the vaccine could remain stable under experimental conditions.Moreover, the aliphatic coefficient and GRAVY were calculated at 1.593 and -0.205, respectively.

Prediction of secondary structure elements
The PSIPRED web server predicted the secondary structure of the vaccine protein.Additionally, the secondary structure of the elements was predicted at 46.86% coils, 38.14% α-helix, and 10.35% β-sheet (Fig. 1C).

Modeling and refining the 3D structure
The I-TASSER web tool was utilized to model the final 3D structure of the multi-epitope vaccine.Figure 1D shows that the best fusion protein structure was obtained with a TM score of 0.56 ± 0.15, a C-score of -1.25, and an RMSD of 9.5 ± 4.6 Å.

3D structure validation
The refined 3D structure of the modeled multi-epitope vaccine was validated using RAMPAGE.The results showed   that 82.5% of the vaccine residues were within the desired area, 15.0% within the permissible area, and only 2.4% within outlier areas.The quality of the model was further assessed using the ProSA web server, resulting in a Z-score of -0.67.Moreover, the ERRAT server calculated the confidence of 92.758% for the model.Figure 2 shows the plots obtained from RAMPAGE and PROSA web servers.

Vaccine interaction analysis using TLR receptors
To improve prediction accuracy of the PatchDock server, several online tools were employed for the protein-protein docking of the vaccine.As Table 3 shows, the receptors used in this study included several TLR receptors (TLR-2, 4, 5, and 9).Based on the electrostatic complementarity and geometry of the protein's surface, the Patchdock server ranked in the top 10 interaction models.At the same time, the global energy of the docked complexes was predicted by the FireDock server.In addition, the FireDock web tool was employed to refine and rescore top complexes.The results showed that binding to the TLR-9 complex produced a good global binding energy score (-23.59kcal/mol), while the lowest binding affinity was obtained through docking with TLR-5 (-7.80 kcal/mol).Also, Chi-Chi 2 plots analysis was done (https:// saves.mbi.ucla.edu/ resul ts? job= 13483 29&p= proch eck).Numbers of residues were shown in brackets.Those in unfavorable conformation (score < -3.00) were labelled.Shading showed favorable conformation as obtain from an analysis of 163 structure at resulation 2.0A or better (Fig. 3).

MD simulation of TLR-vaccine complexes
MD simulations of the final docking vaccine complexes were performed to calculate fluctuations and residual stability using the selected receptors (TLR-2, 4, 5, and 9).Table 4 shows the results of these simulations.In addition, RMSF was assessed to quantify residual fluctuations; on the other side, RMSD was assessed to quantify the stability of each system (Fig. 4A).In total, 30 ns simulations were determined for all systems, reporting RMSD variables of 0.44 ± 0.05 nm for TLR-9, 1.35 ± 0.09 nm for TLR-4, 3.31 ± 090 nm for TLR-2, and 3.38 ± 0.03 nm for TLR-5.RMSD results showed that the complex TLR-9 with the vaccine was more stable over the simulation time and could establish better connections than other TLRs.The residual fluctuation (RMSF) results in Table 4 show that RMSF changes in TLR-2, 4, 5, and 9 were more significant than the vaccine structure, except for the TLK-9 vaccine complex.In addition, RMSF changes in the vaccine (0.184 nm) were more significant in the presence of Tlk-9 (0.14 nm), indicating that the structure of the vaccine in this complex was more flexible.Figure 4B shows the radius of the gyration (Rg) between the total vaccine complex and TLR-2, 4, 5, and 9. Rg shows the degree of compaction and the level of protein availability.
As the images show, TLK-4 and TLK-9 were more compact than TLK-2 and TLK-5 overtime during the simulation.In addition, the Rg value of TLK-9 (3.30 nm) was the smallest among TLK-2 (4.59 nm), TLK4 (4.64 nm), and TLK-5 (5.73 nm); accordingly, this indicates that the designed vaccine interacted better with TLK-9, being consistent with the docking results.Figure 5A shows the distance between TLR-2, 4, 5, and 9 and the vaccine at the nanometer scale.Fluctuations of the TLK-5 distance were more significant, while within a specific range, TLK-4 and TLK-9 decreased at the end of the simulation time.The distance between TLK-5 and the vaccine increased with the vaccine having little exposure to TLK-5.
As Fig. 5B shows, the number of hydrogen bonds between TLK-2 and TLK-5 was within the range of 1,160, while in TLK-4 and TLK-9 simulations for the first two nanoseconds, the number of hydrogen bonds started from 1,950 and then fluctuated within the range 1,650; this indicates that TLK-4 and TLK-9 had more solubility and could interact better.

Discussion
Due to the dangers of H. pylori, controlling its infections with antibiotics may reduce the incidence of the disease.Some antibiotics are now available for treating H. pylori infections; however, high costs, patient unwillingness to regularly use the drug and the development of severe antibiotic resistance are among the factors limiting the adequacy of these treatments [4].Hence, there is an urgent need for an alternative treatment or a prevention method, such as vaccination.Attempts to develop a vaccine against H. pylori began in the early 1990s, with the knowledge that the infection caused by this bacterium was the leading cause of gastric ulcers.Despite research on the H. pylori vaccine and acceptable results in animal models and cell cultures, no vaccine has been licensed for human use [5,18].In this regard, gene vaccines have attracted increasing attention due to their cost-effectiveness, effectiveness, and high immunogenicity in terms of humoral or cellular immunity, or even both [19,20].
Contrary to classical approaches to vaccination, modern research has made it possible to produce "subunit vaccines" that accommodate microbial-specific protein sequences capable of effectively stimulating the reaction.Thanks to ample information on microbial genomes and proteomes, developing effective vaccines against various infections have been possible.Conventional approaches to vaccine production are less effective, time-consuming, and expensive, so they are now almost obsolete.Designing vaccines through a relatively stable, safe, inexpensive, specific, and more practical immunoinformatics approach saves time and money [1,21].To develop a subunit vaccine against the H. pylori infection, this bacterium was used to identify target sequences using the protein advanced immunoinformatics approach.As mentioned earlier, the most detectable toxic agents in H. pylori infections are APO, leoA, iceA1, iceA2, cag, PaI, vacA, cagþ, dupA, oipA, dup A, and babA.The lnt (Apolipoprotein N-acyltransferase) gene encodes lipoproteins that line the cell [7][8][9][10].Lipoproteins, a significant component of cell lining, contribute to several essential cell functions.In addition, they are generated by covalent linkage after lipid translations through a 3-stage sequential process directed by three integral membrane enzymes.These lipoproteins play vital roles, including nutrient uptake, protection of cell wall integrity, protein secretion, phytoplasma folding of proteins, and pathogenicity [7,8].The leoA (labile enterotoxin output A) gene is one of the pathogens of H. pylori GTPase, which is encoded in pathogenicity islands and potentially increases toxin release through secretory vesicles.LeoA is known as a bacterial diamine-like protein.Diamine family proteins alter membranes and generally evolve by releasing heat-sensitive enterotoxin through the membrane vesicle from the bacterial cells' surface.There are few reports on the functional role of bacterial diamine proteins [22,23].Membrane vesicles play an essential role in the protective function of the members of the diamine family.
Reports show this gene has the potential for immunogenicity in the host body.Therefore, it has been suggested as a suitable candidate for producing a gene vaccine against H. pylori [24].The H. pylori iceA gene, induced in contact with the epithelium during bacterial attachment to the gastric mucosa, has two species (iceA1 and iceA2).Evidence shows an association between H. pylori iceA1 and peptic ulcers.There was a significant association between an iceA1 virulence factor and the iceA1 gene with antibiotic mutations associated with clarithromycin.Most iceA1-positive isolates were mutant and resistant to clarithromycin.This issue confirms the relationship between resistance and the presence of virulence factor iceA1 [21,25,26].
The immunogenicity of proteins was evaluated via the thorough analysis of online servers [27,28].Moreover, online tools were utilized to produce a rugged design for the subunit vaccine.Epitopes T-lymphocyte and B-lymphocyte were specified from the sequences of the selected proteins [29,30].T-cell receptors (TCRs) generally create an immune response to MHC-I and MHC-II molecules by activating T-lymphocytes via MHC-binding antigens or antigen-presenting cells (APCs).Based on the results of the H. pylori proteome immunoassay, the vaccine protein developed was able to cover a wide range of MHCs of classes I and II as well as linear B epitopes with similar physicochemical and structural attributes.Production of plasma and memory B cells would make it possible to protect the body against pathogen-related and other specific antigens [31,32].The final design of the vaccine included multiple binding epitopes of MHC I and MHC II that improved host immunity in the presence of adjuvants.Key parameters involved in the study were evaluated accurately to neutralize possible allergic reactions to the developed vaccine.Accordingly, the PSIPRED v3.3 server was employed for SS and 3D formations.The quality of the refined 3D model was verified using various servers.In addition, interactions between the designed vaccine and TLR-2, TLR-4, TLR-5, and TLR-9 were evaluated using the docking process, and docking complexes were stabilized using MD simulations.In this scientific study, immunoinformatic approaches confirmed the vaccine's effectiveness and stability.
Moreover, an agent-based model predicted an accurate response after vaccine sequencing.The vaccine designed in the present study was antigenic, non-allergenic, and effective in controlling H. pylori infections.Future clinical studies are recommended to assess the efficacy of this vaccine further.H. pylori is a cause of gastric infections and sometimes gastric cancer [33,34].This study aimed to design a novel multi-epitope subunit vaccine for the H. pylori antigen using a thorough immunoinformatics approach.To predict the H. pylori epitope, various immunoinformatics tools were employed.Several B-and T-cell epitopes were analyzed and combined with proper adjuvants and linkers to improve immunization of the new vaccine.According to the results, the vaccine had good antigenicity, solubility, allergenicity, tertiary structure, and physicochemical properties.
This study simulated molecular dynamics and binding of the vaccine and TLR-2, 4, 5, and 9.In addition, the binding stability and affinity of the TLK complexes and the vaccine were estimated.Besides, the cellular immune response to antigens was confirmed via immunoassay in silicon.The results obtained in this study were experimentally valuable, so they can be used in developing an experimental vaccine against H. pylori.

Collection of proteins
At this stage, we chose four proteins to produce an appropriate definition for the vaccine against H. pylori.The proteins included Apolipoprotein N-acyltransferase (APO), leoA, IceA1, and IceA2, all of which have been proven effective in managing the H. pylori infection.The most suitable proteins for designing a vaccine are those affecting the virulence of pathogens.Protein sequences were downloaded from the "protein" section of the https:// www.ncbi.nlm.nih.gov/ prote in/ server with the GenBank IDs of ACX97396.1,ACX97803.1,AAL59391.1,and AAG49536.1.

MHC-I binding epitope (CTL) prediction
The NetCTL 1.2 server, an online web tool (http:// www.cbs.dtu.dk/ servi ces/ NetCTL/), was used to predict CTL epitopes for four virulent proteins (APO, leoA, IceA1, and IceA2).The 9-mer epitopes were predicted using the introduced server for the query protein based on MHC-I binding, C-terminal cleavage, and transport-associated protein (TAP) scores.In this server, an artificial neural network is available for C-terminal cleavage, which bonds peptides to MHC-I, while the weight matrix could obtain tap transport efficiency.The foretoken threshold of epitopes was set at 0.75 [35].

B-cell epitope prediction
B lymphocytes (B cells) are crucial in empowering the immune system by secreting necessary antibodies for long-term immunity.Aimed to produce B-cell linear epitopes, the online server of BCPREDS (B-cell epitope prediction server) (http:// ailab.ist.psu.edu/ bcpred/) was used.This server predicted 20-mer linear B-cell epitopes using the kernel technique.The 20-mer linear B-cell epitopes of BCPred employed a backing vector machine (SVM) algorithm to predict B-cell (linear) epitopes [36].

Vaccine sequence creation
A collection of HTL and CTL epitopes were chosen with the presupposition of having a major binding score and a non-allergenic nature.The selected CTL epitopes were linked to each other using an AAY linker to create an ultimate multi-epitope vaccine construct.On the other hand, the selected HTL epitopes were linked to each other using GPGPG linkers.The employed linkers improved epitope representation, and proper separation was achieved using them.There were two main reasons for using these linkers; firstly, they could successfully barricade the formation of functional epitopes (neo-epitopes) and could effectively improve epitope presentation [37][38][39].Moreover, aiming at improving immunogenicity, the EAAAK linker was used as an adjuvant to add human β-defensins to the N-terminus of the vaccine.Linkers (AYY, KK, eaaak and GPGPG) play vital roles in producing an extended conformation (flexibility), protein folding, and separation of functional domains, and therefore, make the protein structure more stable [27].

Allergenicity profiling
To ensure that the vaccine would not cause any allergic responses, the online antigenicity prediction server of AlgPred2.0was used (https:// webs.iiitd.edu.in/ ragha va/ algpr ed2).The introduced server would use several algorithms (SVMc + MAST + IgEepitope + ARPs BLAST) to check the allergenicity of the inquiry sequence [40].The results of this server are dependable, at the precision of 85% at the -0.4 threshold.

Evaluation of physiochemical properties
Physicochemical properties of the vaccine sequence were evaluated using the online freely available web server of ProtParam (http:// web.expasy.org/ protp aram/) [42].The ProtParam server (References/Documentation) computed all required properties of GRAVY, the instability index, theoretical pI, in vivo half-life, in vitro half-life, the aliphatic score, molecular weight, and amino acid composition.

Prediction of secondary structure
The secondary structure of the vaccine sequence was evaluated using the PSIPRED protein structure prediction server (http:// bioinf.cs.ucl.ac.uk/ psipr ed/) that provided results with high accuracy [27].Besides, the PSI-Blast tool was used to identify proteins showing homology to our vaccine construct.
Next, these sequences were customized to create a position-specific scoring matrix (PSSM).Moreover, the introduced server utilized a neural network to process the PSSM and methodize secondary structure elements [43].

D structure prediction of the vaccine
The I-TASSER (Iterative Threading Assembly Refinement) server is a homology modeling tool utilized for 3D or tertiary modeling of the multi-epitope vaccine (https:// zhang.lab.ccmb.med.Umich.Edu/I-TASSER/).The I-TASSER server is an integrated platform used to predict and computerize the performance and structure of the protein.In addition, it utilized the sequence-to-structureto-function paradigm to identify analogous structures from the Protein Data Bank (PDB)53n.Starting from an aminoalkanoic acid sequence, I-TASSER generated 3D atomic models from a wide range of repeated structural assembly simulations and threading alignments.Accordingly, a template modeling (TM) value less than 0.17 showed a random similarity, while a TM greater than 0.5 indicated a model with careful topology [44].

D structure refinement of the vaccine
Validation of the 3D structure was refined using the three web tools of RAMPAGE47 (https:// saves.mbi.ucla.edu/), ERRAT46, and ProSA-web45 (https:// prosa.servi ces.came.sbg.ac.at/prosa.php).The RAMPAGE server used PROCHECK principles for validation through investigating the Ramachandran plot.The ERRAT server worked based on the non-bonded interactions inside a given structure.Finally, Prosa-Web plotted general priority and underlined errors calculated in the 3D query.Next, UCSF Chimera software (www.rbvi.ucsf.edu/ chimera) was used to visualize the model [45][46][47].Additionally, structure correction and energy minimization were performed using YASARA software.

Evaluation of physiochemical properties
Physicochemical characteristics of the fusion protein were evaluated using a ProtParam software program (http:// web.expasy.org/ cgi-bin/ protp aram).

The interaction analysis vaccine with TLR receptors
Human toll-like receptors of 2, 4, 5, and 9 (TLR-2, TLR-4, TLR-5, and TLR-9) and the vaccine were used.In addition, to conduct a thorough molecular docking study, the PatchDock server was used.Next, a candidate transformation test was done on complementary patched structures.Besides, their 3D structure was obtained from the RCSB PDB server (https:// www.rcsb.org/) with IDs 6nig, 3fxi, and 3joa.Since toll-like 9 had no precise structure, its 3D structure was modeled and validated according to the abovementioned steps.Next, the transformations of every candidate were compared using the scores obtained through the geometric fit and its atomic desolvation energy.The input parameters for docking analysis were considered ligands and the PDB coordinate file for each protein.The root-mean-square deviation (RMSD) clustering was set at 4Å, and the complex type was modified to the protein-ligand type [48].

Molecular dynamics simulation
According to the docking results, dynamic simulations were run to determine the effects and behavior of the vaccine for TLK-2, TLK-4, TLK-5, and TLK-9 receptors.Accordingly, MD simulations were performed by GROMACS 2020.1 software and force field G43a1.Besides, the SPC model was chosen for water molecules.Na + and Cl-ions were added to them to neutralize and eliminate undesired interactions.In addition, energy minimization was applied to the system for NVT particles in the system (N), the system's fixed volume (V), and temperature (T), in 300 0K, in 5 ns, and on the NPT system by the pressure of 1 bar, in 12 ns, and at 0.001 femtosecond time intervals.Besides, non-bond interactions were performed at 10 Å by the Particle Mesh Ewald (PME) method.Additionally, root mean square deviation (RMSD), root mean square fluctuation (RMSF), the radius of gyration, protein-protein hydrogen bonds, protein-solution hydrogen bonds, and the target protein's secondary structure were obtained.

Fig. 1 A
Fig. 1 A Arrangement of the HTL and CTL epitopes in the final vaccine construct using different linkers.B B-cell epitopes as predicted by the BCPred server; in the 3D structure of the vaccine, linear B-cell epitopes are presented in magenta, while conformational B-cell epitopes are presented in pink.C Predicting the secondary structure and solubility of the vaccine construct; the results of the PSIPRED server for predicting the secondary structure of the vaccine construct showed 46.86% coils, 38.14% alpha-helices, and 10.35% beta-strands.D Multi-template modeling of the final 3D structure of the vaccine; the relevant elements are depicted in different colors, loops in sky blue, beta-sheets in colored blue, and helices in red

Fig. 3
Fig.3Chi-Chi 2 plots analysis.Numbers of residues were shown in brackets.Those in unfavorable conformation (score < -3.00) were labelled.Shading showed favorable conformation as obtain from an analysis of 163 structure at resulation 2.0A˚ or better

Fig. 4 AFig. 5 A
Fig. 4 A Fluctuations and residual stability (i.e.RMSDs) of the whole system obtained from 20 ns of simulations.B The radius of gyration (Rg) of the whole systems obtained from 20 ns of simulations

Table 2
Helper T-Lymphocyte epitopes and their scores anticipated using the IEDB MHC class II server It shows the difference of alleles *

Table 4
The average and standard deviations of the radius of gyration (Rg), area per residue, root-mean-square deviation (RMSD), root mean square fluctuations (RMSF) (Chain A includes; TLR-2, 4, 5 and 9, and Chain B includes the vaccine), the hydrogen bond between protein-protein, H-bond between protein-solvation (number), between the vaccine with the selected receptors (TLR-2, 4, 5 and 9)