The entire proteome sequence data in FASTA format was acquired by downloading from UniProt, a freely accessible database providing information on protein sequences and functions (17). Their functional importance and pathogenicity based on their essential proteins. To attenuate the risk of triggering an autoimmune response, making sure the chosen proteins lacked homology with human proteins was essential (18). To achieve this, the BlastP bioinformatics tool was employed. BlastP is a widely used bioinformatics tool that conducts sequence similarity (19). This involved compare query sequence against a comprehensive database of protein sequence. Finding the similarities between query and database protein sequences is the primary goal of this procedure (20). Moreover, the VaxiJen server was employed to assess the antigenicity of the chosen proteins by applying a predetermined 0.05 threshold (21).

Cytotoxic T lymphocyte (CTL) cells plays a vital part in the immune system by admit certain antigens (22). To identify CTL epitopes, the interested protein sequences were given that in FASTA format, CTLPred was used in MHC class I complex (23). Integrated epitope prediction was utilized, focusing on selecting CTL epitopes with consensus scores below 2. A lower score indicates an enhanced capability of these epitopes to bind to MHC-I molecules, making them favorable candidates for the construction of vaccine sequences (24). The purpose of CTLPred is to identify putative T cell epitopes in a particular protein sequence by emphasizing their capacity to bind to MHC-I molecules, which makes it easier for cytotoxic T cells to recognize them (25). We utilized the VaxiJen 2.0 server to evaluate the antigenicity of these chosen epitopes. When referring to an epitope’s ability to elicit an immune response, the term “antigenicity” is used (26). To predict allergenicity and toxicity, additional tools such as AllerTOP 2.0 and ToxinPred servers were utilized (27). This comprehensive approach ensures that the selected epitopes not only possess the potential to activate the immune system but are also non-allergic and non-toxic, making them promising candidates for vaccine development (28).

Helper T lymphocytes (HTL) play a crucial role as integral components within the adaptive immune system, demonstrating the capability to coordinate both cellular and antibody-mediated immune responses directed against foreign pathogens (29). In the context of vaccine development, epitopes on Helper T lymphocyte (HTL) cells that interact with MHC class II alleles are essential. These T cells collaborate synergistically with B cells to generate antibodies that specifically target and neutralize active pathogenic cells. This collaborative effort is integral as they work in conjunction with Cytotoxic T Lymphocytes (CTL) and macrophages to collectively combat and eliminate harmful pathogens (30).

B-cell epitopes play a pivotal role in initiating an adaptive immune response and serve as fundamental components in the design of vaccines (29). The ABCPred bioinformatics tool, which is adept at predicting antigenic epitopes or antibody-binding sites within protein sequences, was utilized in the current investigation to predict B-cell epitopes (31). The goal of ABCPred is to anticipate linear B-cell epitopes, or certain areas of proteins that are recognized and bound to by antibodies (32). A threshold value of 0.5 was employed in this tool to predict linear B-cell epitopes within the protein sequences. The VaxiJen 2.0, ToxinPred, and AllerTOP 2.0 servers were utilized to assess the antigenicity, toxicity, and allergenicity of the predicted B-cell epitopes (33). This thorough evaluation ensures that the predicted B-cell epitopes not only possess the potential to stimulate an immune response (antigenicity) but are also non-toxic and non-allergenic, rendering them suitable candidates for inclusion in vaccine constructs (34).

By carefully integrating adjuvant and epitopes from Cytotoxic T-lymphocyte sources, Helper T-lymphocyte and B-cell the multi-epitope vaccine (MEV) sequence was constructed. Specific linkers were used to ensure effective functionality in the integrated design (35). Choosing an adjuvant to control the immunological response that the vaccine evoked was a crucial decision. In this particular case, the adjuvant chosen was the subunit B of cholera enterotoxin (36).

The construction of the MEV sequence followed a strategic approach:

The MEV sequence underwent a series of thorough assessments and analyses to guarantee its safety and efficacy as a candidate vaccine (41): Homology Testing: To verify that the MEV sequence did not exhibit significant similarity to human proteins, a BlastP search against the Homo sapiens proteome was conducted with neglect framework (42). This step aimed to avoid potential autoimmune responses. Antigenicity and Immunogenicity Evaluation: The antigenicity and immunogenicity profiles of the MEV sequence were assessed through analysis using the VaxiJen and IEDB servers, respectively (43). The purpose of this phase was to assess the sequence’s ability to elicit an immunological response. Using the ProtParam service, the modified vaccine’s physicochemical characteristics were ascertained (44). These properties included:

These factors provide important information about the MEV sequence’s general properties, charge, and stability, contributing to the assessment of its potential as an effective vaccine (43). These parameters help assess the vaccine’s stability and suitability for practical applications (46). Toxicity and Allergenicity Prediction: To ensure the vaccine’s safety, its toxicity was assessed using the ToxinPred server, and its potential for inducing allergies was evaluated through the AllerTOP server (47). These analyses were conducted to confirm that the vaccine candidate was non-toxic and devoid of allergic responses (48).

The three-dimensional (3D) structure of a protein is critical to its stability and function. Your vaccine’s tertiary structure was created using a multi-step process that guaranteed the protective structure’s accuracy and predicted its function: The initial vaccine 3D structure was generated using the Alpha- fold 2 server (49). GalaxyRefine Server: The projected 3D structure was further refined and enhanced through the GalaxyRefine server (50). This action was taken to improve the precision and excellence of the forecasted protein structure. RAMPAGE Server: To validate the refined structure, you used the RAMPAGE server (51). RAMPAGE assesses the quality of a protein structure by examining its stereochemical properties, such as bond angles and dihedral angles. This validation step helps ensure that the refined structure conforms to expected structural standards (52). ProSA-web Server: For structural verification ProSA server was used. The structural quality were find by using this server, the lower score have more valid structure (53).

The discontinuous B-cell epitopes were located using the ElliPro program (54). The tertiary structure was also determined using MEV. ElliPro uses the protein’s 3D structure as input to detect areas where amino acids conjoin to form epitope (24). The PI score is used for identifying discontinuous B-cell epitopes. Higher PI scores specify that the part of protein to be involved in antibody binding (55). This information is important for vaccine construct that might interact with antibodies and immune response (56).

By creating disulfide connections between cysteine residues, disulfide engineering seeks to improve the stability of a protein’s three-dimensional structure (57). These bonds are essential for maintaining the stability of the molecular interactions inside the protein (58). The improved model was presented in Design v2.0. This specialized server is designed to predict and recommend the placement of disulfide bonds within a protein structure. It achieves this by analyzing both the protein’s amino acid sequence and its structural information (59).

Establishing an effective immunological response relies significantly on the interaction between a vaccine protein and host immune cells (60). To assess how well your MEV could bind to human immunological receptors, molecular docking research was conducted (61). One specific immunological receptor investigated in this research is Toll-like receptor 4 (TLR4) (62). The Toll-like Receptor 4 (TLR4) is pivotal in the human immune system as it recognizes and responds to foreign microbes with high sensitivity and specificity (63). It responds to molecular patterns associated with pathogens (PAMPs), such as lipooligosaccharides and lipopolysaccharides, which can activate TLR4-mediated immune responses (64).

Normal mode analysis is frequently included in in silico studies to assess the stability of protein-protein complexes (65). This assessment involves analyzing the dynamics of proteins and contrasting their behavior with their typical modes (66). One method to study protein mobility is through normal mode analysis (NMA), which assesses the common motions of proteins within their inherent coordinates (67). In my research, the iMODS server was employed for this purpose (68). The iMODS (NMA) server provides valuable insights into the intrinsic motions of a multiplex of proteins (69). It calculates various parameters, including eigenvalues, covariance, B-factors, and deformability, to characterize the protein’s dynamics and stability (70). Eigenvalues: These values indicate the primary chain’s deformability. Smaller eigenvalues suggest that the protein structure is more prone to bending or undergoing deformations (71). This property is directly associated with the energy required to induce such deformations. Covariance: Covariance analysis helps identify correlated motions within the protein structure, shedding light on how different parts of the protein move together (72). Higher B-factors indicate greater flexibility and mobility. The dynamics of the protein-protein complex under study can be explored by using the iMODS server for normal mode analysis (73). This information is crucial for assessing how the complex behaves and whether it is suitable for its intended function in the context of a vaccine or other biomedical application (74).

Through utilization of C-ImmSim, investigators have the capability to computationally simulate and forecast the immune reactions elicited by the MEV construct upon its introduction into the organism (75). This in silico methodology provides significant utility in evaluating the prospective efficacy of the vaccine and acquiring understanding of its immunological ramifications, circumventing the necessity for exhaustive in vivo or in vitro experimentation (76).

The Java Codon Adaptation Tool (JCat) was employed in the study to carry out several essential tasks related to codon optimization and adaptation (77). The first step entails back translation, converting the designed vaccine’s amino acid sequence into a nucleotide (DNA) sequence (78). This process is known as back translation. The optimization make sure that the gene is expressed efficiently in the host organism (79). The CAI value is indicates that how well optimized codon matches the codon of host organism (80). A higher CAI value indicates a better match. The optimized sequence GC content is determined. This is important because the GC content of the host organism may change gene expression (81). The protein sequence of the vaccine was fed into the JCat server, and strain K12 of E. coli was selected as the host (82). Using SnapGene 3.2.1 software, the optimized sequence was included into the pET30a (+) expression vector following the codon optimization process (83). This vector is frequently used to ensure that the vaccine design is correctly translated and transcribed in the host organism by expressing recombinant proteins in E. coli (84).

The entire proteome of H. parainfluenzae (strain T3T1) consists of 1974 proteins. By using the web server Geptop-2.0, 370 essential proteins inferred from these. By eliminating human homologous by an online analysis using BlastP, there were 180 non-homologous protein recognized. To assess these antigenicity values were used. The top two proteins with the highest antigenicity were composed of extracellular proteins, and they were subsequently selected for additional screening ( Table 1 ).

Based on the H. parainfluenzae target protein, 42 CTL epitopes 12-mer were prioritized. For the purpose of developing vaccines, the top seven epitopes exhibiting high immunogenicity and antigenicity as well as being non-toxic and non-allergenic were selected using a variety of technologies to evaluate antigenicity, immunogenicity, allergenicity, and toxicity (85) ( Table 2 ). 37 unique HTL epitopes were identified, and the top four were chosen for vaccination based on their cytokine-induced capabilities ( Table 3 ). Similarly, among the selected 19 LBL epitopes, two were deemed suitable for vaccine development, considering predictions related to toxicity, allergenicity, and antigenicity ( Table 4 ).

All of the chosen epitopes were included in the building of the vaccine; the MHC-I, LBL, and MHC-II epitopes were linked together using AAY, KK, and GPGPG linkers, respectively (86). The presentation of the epitope and immunization are improved by these linkers (39). The first CTL epitopes were added as an adjuvant, coupled with cholera enterotoxin component B. These epitopes were utilized to inhibit the generation of junctional epitopes, facilitated by the incorporation of the EAAAK linker. Linkers made of EAAAK increase structural stability (87). The final 344 amino acid vaccination demonstrates the arrangement of different epitopes and linkers.

Population coverage is considered to be a significant parameter when it comes to MEV construction (88). At present, a total of 7 CTL and 4 HTL epitopes have been chosen to determine the population coverage, considering their respective alleles. Statistical analysis reveal that the shared coverage of 90.00% worldwide for chosen epitopes. The greatest population coverage was recorded in Japan that was 98.41%. While the reported population coverage for other countries were South Korea for 93.48%, Mexico for 90.85% and Spain for 79.32%. In short, this investigation validated that the filtered epitopes would prove to be potential candidate to construct MEV ( Figure 1 ).

When the produced vaccination’s homology was compared to the human proteome, no similarities were found. Our vaccine was found to be highly antigenic, non-allergenic, and non-toxic after its allergenicity, antigenicity, and toxicity were evaluated. ProtParam was utilized to assess the physiochemical properties (89). The resultant construct exhibited a molecular weight (MW) of 38,262.33 kDa and an isoelectric point (pI) of 9.37. The design of the vaccine showed an average half-life of thirty hours in vitro, more than twenty hours in vivo (Yeast), and more than ten hours in vivo (E. coli). The estimated GRAVY (grand average of hydropathy) was -0.271. Given all of these characteristics, H. parainfluenzae appears to be a suitable candidate for immunization.

To anticipate the secondary structure of the vaccine, the PSIPRED and SOPMA servers were deployed (90). The sequence analysis revealed that the alpha-helix comprised 25.58% of the sequence, consisting of 88 residues. Additionally, 30.52% of the sequence adopted extended chains with 105 residues, while 34.30% of the vaccine’s construct, encompassing 118 amino acids, exhibited a coil structure.

The web resource AlphaFold2 was used to estimate the tertiary structure of H. parainfluenzae (91). Furthermore, the Galaxy-refine server was employed to improve the structural refinement of the vaccine (50). Analysis of the improved model using the Ramachandran plot indicated that 90.3% of amino acids were situated in the favored regions, 8.3% were in the allowed regions, and only 1.3% were in the outlier regions, based on the enhanced model study ( Figure 2 ). The improved model score found in the quality-check analysis of the ERRAT. These results show that the enhanced model has an exceptional 91.613 quality score ( Supplementary Figure 2 ).

Humoral immunity is established through the production of antibodies by B-lymphocytes. Consequently, an effective vaccination should incorporate optimal B-cell epitope domains to stimulate the production of antibodies. The vaccine construct’s 15 conformationally-discontinuous epitopes of 3-47 residues ( Supplementary Table 2 ). with score ranging 0.524-0.989 (as shown in Figure 3 ) and 12 linearly-continuous epitopes were predicted using ABCPred 2.0 with default parameters ( Supplementary Table 1 ). The conformational B-cell epitopes were visualized during the vaccine production process using PyMOL v.1.3, a molecular graphics system ( Figure 3 ).

Thirty-three amino acid residue pairings with the ability to create disulfide bonds were found by the DbD2 server. Subsequent evaluation based on chi3 and B-factor energy parameters narrowed down the selection to three specific pairs: PRO74-HIS78, GLY223-GLN245, and PHE246-ALA264. In each case, a cysteine residue was introduced to facilitate the formation of the disulfide bond. The evaluation process involved screening residues within the range of −87 to +97 chi3 values and ensuring an energy value of < 2.2. These criteria were employed to identify residues suitable for disulfide bond formation, enhancing the stability of the protein ( Supplementary Figure 1 ).

TLR4 (PDB ID: 2z66) receptor was considered for docking, consisting of four chains. ClusPro was used to perform protein-protein docking between the vaccine design and TLR4 receptor to identify the optimal binding pose, assess stability, and determine binding affinity. Among ten models produced by ClusPro, the best docking complex was chosen by considering two points (1) the largest number of cluster and (2) possessing minimum energy. The first model fulfilled the criteria and had 87 members in cluster and -1098.4 kcal/mol energy, exhibiting the stability of the docking complex. Interaction between receptor and MEV was determined through PDB sum which unveiled that MEV showed better interaction with chain D of the receptor, formed 24 hydrogen bonds more than any other chain ( Table 5 and Supplementary Table 3 ). The thermodynamic parameters for the binding energy of the docking complex were obtained via the PRODIGY website. Specifically, the Gibbs free energy ΔG (ΔG = RT ln Kd) or equilibrium dissociation constant Kd was found to be 5.6x10-11 at 37oC and ΔG -14.o kcal/mol ( Figure 4 ).

Normal mode analysis between MEV and receptor docked complex was performed through iMODS which provided insights into deformability of residues. Deformability refers to how easily a protein can undergo any change in its three-dimensional structure. Peaks in the graph show high deformability indicating how easily protein three-dimensional structure can change. A comparison of the docked complex’s PDB and NMA can be seen in the B-factor graph. The graph’s greater peaks for NMA than PDB indicated that NMA data was expected to predict higher B-factors than PDB data. On the other hand, it clarifies that B-factor values found experimentally using PDB data were not as flexible or mobile as those indicated by computational simulations using NMA. The amount of energy needed to distort a structure is measured by its eigenvalue. The graph’s lower eigenvalues demonstrated that only a little amount of energy was required to distort the structure, confirming the molecular motion’s notable flexibility and stability. The variance graph has an inverse relation to the Eigenvalue graph. The red color indicates individual while the green color refers to cumulative variance. Relation between residues is indicated by the covariance matrix. The correlated, uncorrelated, and anti-correlated experience of amino acids is represented by red, white, and blue colors respectively. The docking complex showed a better correlation between pairs of residues which ensure the stability of dock complex ( Figure 5 ).

Both primary and secondary immune responses are important when it comes to particular immune responses to a disease. IgG and IgM concentrations are significantly higher in the first reaction. The antigen is then reduced as a result of an increase in IgM, IgG1, IgG2, and IgG1 antibodies in both the secondary and primary stages. It is known that the cytokine and interleukin responses work quite well in this process. Reports show that the immunization produced an effective immune response, which resulted in the pathogen being eliminated on consecutive visits. These findings demonstrate the efficacy of the immune system in mounting a robust and adaptive response to combat the pathogen, ultimately contributing to an effective defense mechanism against the infection ( Figure 6 ).

To improve the vaccine protein’s expression efficiency in the E. coli host system, in silico cloning and codon optimization were carried out. The codons in the vaccination protein were modified to match the intended E. coli K12 host’s codon use. The Codon Adaptation Index (CAI) was calculated to be 0.9, indicating a high level of alignment, and the GC content of the optimized DNA was determined to be 49.32%.A CAI score of 1.0 was considered as an optimal compromise for efficient expression. The E. coli vector pET 30a (+) had the synthesized codon sequence carefully placed between the Nco1 and particular restriction sites. Furthermore, the resultant clone’s overall size of 5788 base pairs guaranteed that the expression vector’s optimized vaccination protein sequence would integrate successfully ( Figure 7 ).