Exploring the nuclear proteins, viral capsid protein, and early antigen protein using immunoinformatic and molecular modeling approaches to design a vaccine candidate against Epstein Barr virus

Abstract

The available Epstein Barr virus vaccine has tirelessly harnessed the gp350 glycoprotein as its target epitope, but the result has not been preventive. Right here, we designed a global multi-epitope vaccine for EBV; with special attention to making sure all strains and preventive antigens are covered. Using a robust computational vaccine design approach, our proposed vaccine is armed with 6–16 mers linear B-cell epitopes, 4–9 mer CTL epitopes, and 8–15 mer HTL epitopes which are verified to induce interleukin 4, 10 & IFN-gamma. We employed deep computational mining coupled with expert intelligence in designing the vaccine, using human Beta defensin-3—which has been reported to induce the same TLRs as EBV—as the adjuvant. The tendency of the vaccine to cause autoimmune disorder is quenched by the assurance that the construct contains no EBNA-1 homolog. The protein vaccine construct exhibited excellent physicochemical attributes such as Aliphatic index 59.55 and GRAVY − 0.710; and a ProsaWeb Z score of − 3.04. Further computational analysis revealed the vaccine docked favorably with EBV indicted TLR 1, 2, 4 & 9 with satisfactory interaction patterns. With global coverage of 85.75% and the stable molecular dynamics result obtained for the best two interactions, we are optimistic that our nontoxic, non-allergenic multi-epitope vaccine will help to ameliorate the EBV-associated diseases—which include various malignancies, tumors, and cancers—preventively.

Introduction

With more than 90% of adults worldwide seroconverting to the positive extreme of infection¹ and the easy dynamics at which the virus is harmlessly acquired within the cradle to teenage years and then maintains an asymptomatic chronic infection of the B-lymphoid system until the strike of an immunocompromised which perfects the stage for a variety of pathologies including further immunocompromised, lymphomas and cancers etc.². Epstein Barr virus (EBV) is worthy of global surveillance and monitoring as it presents itself as a harmless Trojan horse housing nefarious health danger, and spreads through the most romantically innocent means: kiss. Epstein Barr virus belongs to the γ1 among the eight-herpesvirus observed in humans, genetically; the virus is armed with a double-stranded DNA genome of about 172 kb pairs containing over 70 open reading frames which permits the translation of an assortment of viral proteins with which the virus affords different forms and phase of infection²

Studies of the chronological order of the mechanisms of activity establishes five cardinal events of EBV infection which are: (1) The replication of orally acquired lytic virus; this proliferative event of the infection occurs most probably in the squamous epithelial cells of the oropharynx and in some extravagated B-lymphocytes, (2) Expansion of the infection through proliferative-growth-transforming infection of the B-lymphocytes in the oropharyngeal lymphoid tissues, (3) Repression of the proliferative-growth-transforming signals in some of this infected B-cells as they enter the B-cell pool in a stride towards immortalization of the infected B-cell, (4) Lifelong persistence of a silent asymptomatic, infected and immortal B-cell (in which the EBV antigens are tightly repressed) in the memory B-cell pool which fluxes between the blood and pharyngeal lymphoid tissue—a truly latent phase of infection and (5) Shunting of few truly latent infected B-cells into the lytic proliferative cycle within the oropharynx—this facilitates a continuous viral shedding and discharge at low dose into the throat^3,4.

Unlike the lytic infection, the latent phase of EBV infection is marked by the expression of very few numbers of antigenic proteins which persist within the host cell although expressed at a low level, especially in immunocompetent subjects, also; the latent genes or genetic translates expressed in the latent phase of EBV infection characterize the subtype of latent infection observed in a subject². A viral-immortalized memory B-lymphocyte is said to be in the Latency 0 phase when it completely represses and silences out the viral genome, but considered to be in the Latency I phase when it expresses latent membrane protein (LMP)-2A alone or co-expressed with EBV nuclear antigen EBNA-1⁵. When the immortalized B-lymphocyte assumes homing to the germinal centers of lymphoid follicles, it approaches the Latency II phase which is characterized by the expression of EBNA-1, LMP-1, LMP-2A and LMP-2B in an immunological mechanism that spares some infected lymphoblast, supporting their thriving and differentiation into memory B-cells⁶. In the stance of immunocompromise, as whispered above, EBV-infected lymphocytes denounce its asymptomatic pledge and adorn its infective armor as it presents the Latency III feature characterized by the expression of the EBV six nuclear antigens (EBNA1-6), the three LMPs and two small non-coding RNA namely EBER-1 and EBER-2^2,7.

EBV has been implicated in various autoimmune disorders with the most convincing relationship observed between EBV infection and multiple sclerosis (MS) with observed patients showing high EBV antibody titer, especially to EBNA-1 among other proposed mechanisms reviewed by Talor et al.⁸, hence for this cause EBNA-1 was withdrawn from the construct of our vaccine because of the potential autoimmune threat it possesses within the construct.

Vaccines and vaccination have been a routine methodology for immunological fortification against pathological invasions, relying on the ability of memory B-cells to recognize and foster a faster, long-lasting adaptive immune response to infection genre previously encountered before a new invasion by the same pathogen. However, till today, vaccines designed against EBV have focused on antigenic glycoproteins (especially gp350) of the herpesvirus as its epitope of interest⁹, hence the success recorded so far has failed to provide lifelong immunity against EBV primary infection, also subjects vaccinated for one strain has been found re-infected with another strain of the EBV¹⁰. Despite these discrepancies, available vaccines have successfully prevented nefarious EBV-associated malignancies⁹.

Despite the already made strides towards providing preventive vaccines against EBV infection and the malignancies they foster, the epidemiological data of EBV infection and its associated health derangement still begs that the scientific community dig for clearer water; which infers a continuous attempt to be made until “hopefully” a preventive vaccine that can supplant EBV’s infection, replication, burden on immune system an ultimately provide a lifelong immunity to EBV and its associated malignancies is achieved.

This study aims to apply an integrated knowledge of bioinformatics, computational informatics, and modeling approaches towards the design of subunit vaccine candidates against Epstein Barr Virus. The wet-lab experimental method for pinpointing these epitopes is both time-consuming and expensive due to the necessity of screening a large pool of potential candidates. Thankfully, machine learning (ML) prediction methods have alleviated this burden by narrowing down the list of potential epitope candidates for experimental trials^11,12. Machine learning approach are also not without limitations such as most machine learning tools cannot predict deterministically whether a given peptide is an epitope or not, epitope classifiers rely majorly on few classifier, and not all physicochemical properties of protein can be revealed by these machine learning tools. Implying that experimental validation is critical for validation before approval¹¹. Specifically, this study is aimed at designing a multi-epitope vaccine based on early antigen, nuclear antigen, and viral capsid protein of EBV type 1 and 2 excluding EBNA-1 which has been implicated in Multiple Sclerosis. The construct would save time for researchers working in vaccine development to screen antigenic epitope candidates from non-antigenic ones and also reduce the risk of auto-antibody reaction^11,13. The human leucocyte antigen (HLA) allele’s super-types were also analyzed to ensure a wide population coverage for the designed vaccine. This is the scientific novelty of this research paper.

Materials and methods

Target organism antigen sequence retrieval

The target antigens’ proteome sequences were obtained for this work from the National Center for Biotechnology Information (NCBI). Preferentially, the antigen sequences for the two genotypes of Epstein Barr virus were retrieved (EBV-1 and EBV-2)¹⁴. Ten distinct antigens, including EBNA-2, EBNA-3A, EBNA-3B, EBNA-3C, EBNA-LP, EBV-EA, EBV-VCA, LMP-1, LMP-2A, and LMP-2B, were understudied to analyze and create a multi-epitope vaccine. Figure 1 below displays the sequential chart for the approaches used.

Figure 1.

Sequential methodology flow chat.

B-cell epitope prediction

B-cell epitope-binding receptors are present on the surface of B lymphocytes¹⁵. ABCpred server was utilized to predict the potential B-cell epitopes present in the retrieved antigenic proteins. The threshold for the epitope predictions in 16-mers was 0.51. Based on the score acquired from the trained recurrent neural network, the server has been trained to rank the predicted epitopes based on their score. The peptides’ higher score values indicate that a B-cell epitope is more likely to be effective¹⁶. The epitopes predicted by this server were then verified and cross-checked using the BCepred server¹⁷.

MHC-1 epitope prediction

Early management of Epstein-Barr virus infection is associated with the expansion of CD8+ T lymphocytes, which are specific for various EBV proteins generated during the lytic and latent stages of viral infection¹⁸. The MHC-1 binding feature of the IEDB server was used to predict the Cytotoxic T-cell epitopes¹⁹ exploiting IEDB recommended 2020.09 (NetMHCpan EL 4.1) prediction method using humans as the MHC source with HLA-A*01:01 and HLA-A*02:01 alleles. For both alleles, the epitope predictions were predicted in 9-mers. The server uses a group of neural networks trained on all peptides with lengths between 8 and 16, and the networks trained on all peptide lengths have much better performance.

MHC-II epitope prediction

Prediction of helper T-cell epitopes is a vital step in the design of an epitope-based vaccination because these cells promote the spread of potent humoral and cellular responses by promoting CTLs to grow optimally and maintaining a robust cytotoxic T-cell response²⁰. Additionally, EBV-transformed B-cells retain the characteristics of antigen-activated B-cell blasts and may interact with CD4+ T cells to perform inhibitory or supportive activities²¹. The NetMHCIIpan server with DRB1 loci of DRB1_0101, DRB1_0301, DRB1_0401, DRB1_0701, DRB1_0803, DRB1_1001, DRB1_1101, DRB1_1201, DRB1_1301 was employed to accurately predict potential Helper-T cell epitopes. For strong binders (SBs) and weak binders (WBs), the prediction’s accuracy was set at 2% and 10%, respectively²². For further investigation, epitopes having the potential to be only strong binders were chosen. The percentage Rank is a transformation that allows for interspecific MHC binding prediction comparisons and normalizes prediction scores across various MHC molecules²².

IFN-gamma, IL-4, and IL-10 inducibility assessment of MHC-II epitopes

An immediate response by the infected cell is one hallmark of a viral infection. This entails the production of certain cytokines, commitment to apoptosis, and activation of a built-in antiviral defense system. These incidents aid in limiting viral spread and reducing viral replication²³. The ability of the predicted helper T cell epitopes to be IFN-gamma, IL-4, and IL-10 inducible was accessed through IFNepitope (IL-4pred and IL-10pred servers) respectively. The IFNepitope server was utilized in a hybrid method, incorporating predictions from the motif and machine learning approaches²⁴. As a prediction model, IFN-gamma versus non-IFN-gamma was employed. To choose exclusive motifs in IL-4 induce and non-inducing MHC class II binding peptides, IL-4pred employs the publicly accessible program MERCI and the learning approach support vector machine (SVM)²⁵. The prediction method was based on a hybrid motif that incorporates the two approaches.

Antigenicity, allergenicity, and toxicity evaluation of retrieved epitopes

An antigen’s antigenicity determines its capacity to elicit an immune response and the generation of memory cells. Highly antigenic epitopes should be employed in the construction of vaccines²⁶. To evaluate the antigenicity nature of B-cell epitopes, MHC-1, and MHC-II epitopes, Vaxijen server 2.0 was utilized using the viral model and threshold of 0.4²⁷. To cross-validate the antigenic epitopes predicted by Vaxijen, epitopes with antigenic scores ≥ 0.4 were submitted to Virvacpred. The user-friendly and durable interface of Virvapred allows for the accurate and quick prediction of viral antigens²⁸. Later, epitopes with antigenic scores ≥ 0.5 were chosen for analysis. Consequently, no allergenic properties must be present in the epitopes used to make the vaccine²⁹ and no toxicity property³⁰. AllerTop server and Toxinpred server were used to access the epitope allergenicity and toxicity^25,31.

Human homology and conservancy test

Unlike epitopes stemming from highly mutable genomic regions, embracing conserved epitopes will likely enhance security across various species. Using the BLASTP tool on NCBI, the chosen epitopes from the previous step were accessed to confirm their organism uniqueness and, more importantly, to determine whether or not they are human protein homologs. The epitopes selected as non-homolog have an e-value ≤ 0.05³². To investigate the conservancy between epitopes and proteins, the conservational analysis of a few selected epitopes was performed using the IEDB server³³.

Epitope selection and vaccine construct

Epitopes that were successively predicted using different immunoinformatic methods and passed all evaluations were then utilized for the creation of chimeric polyvalent vaccines. The adjuvant beta-defensin-3 was utilized to increase the immunogenicity of the vaccination design³⁴. Also, the choice of selection of the adjuvant was based on the TLRs activated by Epstein Barr virus (TLR-2 and TLR-9). It has been verified that beta defensin-3 activates TLR 2 and TLR 9^35,36. The adjuvant was linked to the first B cell epitope using an EAAAK linker, while the HTL epitopes and B-cells were connected using a GPGPG linker, and the CTL epitopes were attached from the N to C terminal using an AAY linker³⁷.

Vaccine construct validation

The antigenicity attribute of a vaccine is what allows it to trigger an immunological response in the host. High antigenicity is required for a successful vaccine design to elicit an immunological response³⁸. To evaluate the construct’s potential for antigenicity, Vaxijen server was used²⁷ and the test was also confirmed using a server specifically built for viral antigen-Virvacpred²⁸. A vaccine’s ability to trigger allergic reactions can be determined by screening for allergenicity, which is a crucial step³⁹. The AllerTop server³¹ was exploited to evaluate the construct allergenicity and also cross-validated using the AllercatPro 2.0 server. By comparing the query protein’s amino acid sequence and 3D structure to the server’s list of known allergens, AllercatPro 2.0 predicts the query protein’s allergenicity potential l⁴⁰. A good vaccination construct should have no potential to cause an allergic reaction, just like it should have zero toxicity¹⁵. The toxicity level of the chimeric polyvalent vaccine construct was evaluated using the Toxinpred2 server⁴¹ and also the test was repeated using ToxDL server to validate the result⁴². Additionally, using NCBI’s BLASTp software, the vaccine construct was examined for its similarity to human proteome³⁷.

Construct validation across the genotypes of Epstein Barr virus and EBNA-1 antigen

The construct’s credibility to elicit an immune response for both versions of the target organism is of utmost importance because we employed a conventional strategy to design a novel vaccine candidate for both genotypes of Epstein-Barr virus (EBV 1 and 2). The NCBI’s BLASTP tool was used to perform the homology search⁴³. Additionally, because EBNA-1 antigen was excluded due to its autoimmune disease⁴⁴, we validated our construct to not have a homology sequence to that found in EBNA-1 that might have sued during epitope linkage through the same homology search.

Physicochemical property survey

A vaccine design must have improved physicochemical qualities, including the capacity to interact with the aqueous environment, good stability, temperature tolerance, and good solubility, among many others. These attributes were checked using the Expaxy Protparam server. The server predicted the construct Grand Average Hydropathicity (GRAVY), Molecular Weight, Aliphatic Index, Instability Index, and half-life⁴⁵.

Secondary structure and tertiary structure prediction

The secondary structure of the vaccine construct was predicted using the Self-Optimized Prediction Method (SOPMA). SOPMA predicts parameters that take into account the coils, sheets, twists, and helices^46,47. The 3D model of the vaccine is needed to act as a ligand for the target receptor before molecular docking. The I-TASSER server was used to accurately predict the 3D structure of the novel chimeric protein construct. The de-novo method employed the threading methodology, which made use of many templates^48,49. I-TASSER makes five model predictions with TM, RMSD, and C-score values ranging from − 5 to 2. The most effective model was verified and used for more analysis.

3D-structure refinement and validation

Structure validation and refining are required to further boost the confidence of the polyvalent construct’s projected tertiary structure²⁰. Refinement was done by the GalaxyRefine tool⁵⁰. The server applies dynamics stimulation following sequence input to carry out structural perturbations and relaxation⁵¹. The selected best model was validated using the Ramachandran plot server⁵². The main output of this server is a Ramachandran plot depicting amino acids in highly favorable regions, favorable regions, and questionable regions. Disulfide by Design 2 was used to check for the presence of disulfide bonds in the vaccine construct⁵³. Finally, the ProSA web server predicted the construct Z-score⁵⁴. ProSA-web focuses on specific requirements for confirming protein structures discovered using theoretical calculations, NMR spectroscopy, and X-ray research⁵⁵.

Molecular docking, binding affinity assessment, and interaction visualization

Molecular docking is a helpful computational approach that not only determines the optimal inclination for interaction between receptor and ligand but also offers helpful details about the specific interaction^56–58. To perform the molecular docking, four (4) TLRs were used as receptors retrieved from AlphaFold Protein structure database namely TLR 1 (ID-Q15399), TLR 2 (ID-O60603), TLR 4 (ID-A0A7D5XLN4), and TLR 9 (ID-B4E0A1)⁵⁹. AlphaFold is a DeepMind AI system that creates cutting-edge predictions of protein structures from their amino acid sequences⁵⁹. The structure preparation, which involves protonation, energy minimization, the inclusion of partial charges, and the determination of binding affinity, were all done using the Molecular Operating Environment (MOE) software version 2015.10⁶⁰. Each TLRs were input as the target receptor while the vaccine construct was used as the ligand. After the successful molecular docking, the PDBsum web server was used for the interaction⁶¹.

Population coverage on selected epitopes

The tool at IEDB that measures population coverage was used to determine how well the vaccination was covered by the human population⁶². Thus, the distribution of HLA alleles among the global population is essential for the creation of an effective multi-epitope vaccination⁶³. HLA alleles were introduced, and a variety of ethnic groupings and global geographic regions were chosen.

Molecular dynamics simulation

Molecular dynamics (MD) simulation is an essential and effective technique to evaluate the induced structural stability and molecular interactions in the protein complex during dynamics^63,64. The molecular dynamics of the best hit complexes were accessed through the CABS-flex web server to evaluate the fluctuation map of the docked complex⁶⁵. A single protein structure must be entered into the cabs-flex server for it to provide a residue fluctuation profile and accompanying analysis⁶⁶.

Immune simulation

Utilizing the C-ImmSim server, in-silico immune simulation was done to estimate the multi-epitope vaccine construct’s real-life immunogenic profile⁶⁷. The minimum recommended time between the first and second dosage for the majority of vaccines now in use is 4 weeks²⁶. For our immunological simulation, three injections were given, spaced 4 weeks apart, each comprising 1000 vaccine construct units. The C-ImmSim server calculates simulation runtimes using a time-step scale. Each time step in this scale corresponds to 8 h in real life. The three injection points were set at time steps 1, 84, and 172, respectively, with a total of 1050 time steps for the simulation. All other parameters were left at their default values.

Construct functionality test

Despite being highly antigenic, a vaccine’s architecture has a high chance of inducing or mediating other biological processes without being allergic or harmful. Using the protein function prediction (PFP) web service we verified that our vaccine design only served to mediate immunological response. Molecular function, biological process, and cellular component annotations are predicted by the PFP algorithm using PSI-BLAST (version 2.2.6) with statistical significance scores (P-value) and estimated accuracy within a certain range of edges on the GO-directed acyclic graph (DAG)⁶⁸.

Codon optimization and in-silico cloning

Codon optimization was done using the JCat online software for a specific expression host⁶⁹, supplying a sequence that has been optimized and has a high codon adaptive index (CAI) and GC content percentage (GC%). The ideal GC% range is between 30 and 70%; anything less could result in transcriptional and translational deficiencies. A CAI score of 0.8 to 1 is ideal and indicates better gene expression in a particular organism⁷⁰. In the next step, SnapGene software was utilized to do a virtual in-silico cloning using the enhanced nucleotide sequence as a template. At restriction sites XhoI and Xbai, the sequence was introduced into the plasmid vector pET28a ( +).

Result

Target organism antigen sequence retrieval

Each considered antigens were preferentially retrieved for the two types of EBV in FASTA format. Table 1 shows the accession number, amino acid length, and submission date for all retrieved proteins.

Table 1.

Accession number, amino acid (AA) length and submission date for retrieved antigens.

Antigen	Accession number	AA length	Submission date
AG876
EBNA-2	Q1HVF7.1	641	23-Feb-2022
EBNA-3A	Q69138.1	925	29-Sep-2021
EBNA-3B	Q1HVG4.2	946	29-Sep-2021
EBNA-3C	Q69140.1	1069	23-Feb-2022
EBNA-LP	Q1HVI8.1	506	29-Sep-2021
EBV-EA	ABB89234.1	404	15-Oct-2009
EBV-VCA	ABB89273.1	1381	15-Oct-2009
LMP-1	Q1HVB3.1	371	23-Feb-2022
LMP-2A	ABB89217.1	497	15-Oct-2009
LMP-2B	ABB89219.1	378	15-Oct-2009
AKATA
EBNA-2	AFY97831.1	486	28-Dec-2012
EBNA-3A	AFY97830.1	944	28-Dec-2012
EBNA-3B	AFY97829.1	918	28-Dec-2012
EBNA-3C	AFY97856.1	1009	28-Dec-2012
EBNA-LP	AFY97832.1	506	28-Dec-2012
EBV-EA	BAU51465.1	404	27-Apr-2016
EBV-VCA	BAU51503.1	1381	27-Apr-2016
LMP-1	AFY97906.1	371	28-Dec-2012
LMP-2A	AFY97825.1	497	28-Dec-2012
LMP-2B	AFY97826.1	378	28-Dec-2012
B59-8
EBNA-2	P12978.1	487	02-Jun-2021
EBNA-3A	P12977.2	944	03-Aug-2022
EBNA-3B	P03203.3	938	23-Feb-2022
EBNA-3C	P03204.1	992	23-Feb-2022
EBNA-LP	Q8AZK7.1	506	02-Jun-2021
EBV-EA	CAD53407.1	404	26-Jul-2016
EBV-VCA	YP_401697.1	1381	13-Aug-2018
LMP-1	P03230.1	386	25-May-2022
LMP-2A	YP_401631.1	497	13-Aug-2018
LMP-2B	YP_401632.1	378	13-Aug-2018
GD1
EBNA-2	Q3KSV2.1	451	02-Jun-2021
EBNA-3A	Q3KST2.2	935	29-Sep-2021
EBNA-3B	Q3KST1.1	938	29-Sep-2021
EBNA-3C	Q3KST0.1	1009	03-Aug-2022
EBNA-LP	P0C732.1	506	29-Sep-2021
EBV-EA	P0CW72.1	221	02-Jun-2021
EBV-VCA	P0C704.1	1381	02-Dec-2020
LMP-1	P0C741.1	366	23-Feb-2022
MUTU
EBNA-2	AFY97916.1	486	28-Dec-2012
EBNA-3A	AFY97915.1	944	28-Dec-2012
EBNA-3B	AFY97914.1	938	28-Dec-2012
EBNA-3C	AFY97990.1	992	28-Dec-2012
EBNA-LP	AFY97917.1	506	28-Dec-2012
EBV-EA	AFY97929.1	404	28-Dec-2012
EBV-VCA	AFY97965.1	1381	28-Dec-2012
LMP-1	AFY97987.1	386	28-Dec-2012
LMP-2A	AFY97909.1	497	28-Dec-2012
LMP-2B	AFY97910.1	378	28-Dec-2012

B-cell epitope prediction

The linear B-cell epitopes were predicted using the ABCpred server and cross-validated using the BCepred server to successfully develop a vaccination candidate. From each antigen, only the first ten predicted epitopes were collected and examined further. Before further analysis, repetitive antigen epitopes were also eliminated. Only 6 epitopes were ultimately used in the vaccine’s development after a satisfactory review. The chosen epitopes and their corresponding assessment scores are shown in Table 2.

Table 2.

Antigenicity, toxicity, allergenicity, human homology, conservancy, IF-gamma, IL-4, IL-10 inducibility test of selected epitopes.

Epitope	Antigenicity (Vaxijen ≥ 0.4)	Antigenicity (Virvacpred ≥ 0.5)	Allergenicity	Toxicity	Human homology	Conservancy (%)	IFN-gamma	IL-4	IL-10
B-Cell
DLSIHGRPRPRTPEWP	1.3340	0.8442	Non-allergen	Non-toxic	Non-homolog	100	Not-applicable
RSGRRAPCVYTADLDI	0.8598	0.9121	Non-allergen	Non-toxic	Non-homolog	100
RSWRMGYRTHNLKVNS	0.7935	0.9812	Non-allergen	Non-toxic	Non-homolog	100
TRGGQEPRRVRRRVLV	0.6315	0.5000	Non-allergen	Non-toxic	Non-homolog	100
RGDRSEGPGPTRPGPP	0.4671	0.9215	Non-allergen	Non-toxic	Non-homolog	100
SPLDWDPLGYDVGHGP	1.5754	0.7078	Non-allergen	Non-toxic	Non-homolog	85.71
MHC-1
QLDDLGYPL	0.6830	0.5331	Non-allergen	Non-toxic	Non-homolog	100	Not-applicable
VATRRQSVY	0.8985	0.990	Non-allergen	Non-toxic	Non-homolog	100
RLGATIWQL	0.7761	0.7061	Non-allergen	Non-toxic	Non-homolog	100
PLDWDPLGY	2.3604	0.7489	Non-allergen	Non-toxic	Non-homolog	85.71
MHC-II
WKLLSPYRTWRMGYR	0.5125	0.5000	Non-allergen	Non-toxic	Non-homolog	100	Inducer	Inducer	Inducer
KTSYRSDEAEEAQST	0.6544	0.5000	Non-allergen	Non-toxic	Non-homolog	100	Inducer	Inducer	Inducer
HSDYQPLGTQDQSLY	0.8193	0.8360	Non-allergen	Non-toxic	Non-homolog	100	Inducer	Inducer	Inducer
RQAIRDRRRNPASRR	0.6657	0.9893	Non-allergen	Non-toxic	Non-homolog	100	Inducer	Inducer	Inducer
VTDFSVIKAIEEEHR	0.6030	0.7831	Non-allergen	Non-toxic	Non-homolog	100	Inducer	Inducer	Inducer
SVGNPSLSVIPSNPY	0.5870	0.8949	Non-allergen	Non-toxic	Non-homolog	85.71	Inducer	Inducer	Inducer
SLGNPSLSVIPSNPY	0.6359	0.7829	Non-allergen	Non-toxic	Non-homolog	85.71	Inducer	Inducer	Inducer
ETTQTLRFKTKALAV	1.1302	0.7450	Non-allergen	Non-toxic	Non-homolog	83.33	Inducer	Inducer	Inducer
CVNTEFKDLSRMTDG	1.1258	0.8543	Non-allergen	Non-toxic	Non-homolog	94.74	Inducer	Inducer	Inducer

MHC-1 epitope prediction

To facilitate an immune response, T-cells are crucial in the development of vaccines. The MHC-1 program of the IEDB was used to retrieve MHC-1, also known as CTL epitopes. The development of the novel vaccine candidates made use of the 4 epitopes found in 9-mers that had a high propensity for binding to MHC-1. Table 2 displays the chosen epitopes and the results of their evaluations.

MHC-II epitope prediction

MHC-II epitopes, also referred to as HTL cells, are essential for the development of vaccines because they offer comparable assistance for the growth and maintenance of an effective Cytotoxic T-lymphocyte epitope function. The 15-mers epitopes were obtained from NetMHC-IIpan taking into account 9 human alleles. Finally, 9 HTL epitopes were linked to create the vaccine candidate. Prediction scores for individual epitopes are shown in Table 2.

IFN-gamma, IL-4, and IL-10 inducibility assessment of MHC-II epitopes

The ability of the chosen MHC-II epitopes to produce cytokines that can activate immune cells to produce antibodies was examined using an inducibility test. All of the chosen antibodies were examined and found to be capable of inducing IL-4, IL-10, and IFN-gamma. Table 2 displays the analysis findings.

Antigenicity, allergenicity, and toxicity evaluation of retrieved epitopes

The utilization of highly antigenic, non-toxic, and non-allergenic epitopes is essential in the development of vaccines. The antigenicity ratings for B-cell, MHC-1, and MHC-II epitopes for Vaxijen vary from 0.4 to 1.5, 0.6 to 2.3, and 0.5 to 0.8, respectively, whereas those for Virvacpred range from 0.5 to 0.9 for all class of epitopes. All of the chosen epitopes were non-allergenic and non-toxic as shown in Table 2.

Human homology and conservancy test

A vaccine’s capacity to maximize its effects depends on how unique it is to the host’s body. When the epitopes employed in the vaccine design were examined for non-human analogs, none were discovered to have human analogs. To give a wider epitope security, a conservancy test was then conducted on the chosen epitopes. For the B-cell epitope, CTL epitope, and HTL epitope, the test yielded conservancy percentages ranging from 85.71 to 100%, 85.71 to 100%, and 83.33 to 100%, respectively. The homology and conservancy test across several epitopes is shown in Table 2.

Primary vaccine construct

Using the chosen epitopes predicted above, a multi-epitope subunit vaccination candidate was constructed. The N-terminal of the possible vaccine candidate construct, which was connected to the other components by an EAAAK linker, had a beta defensin-3 adjuvant attached to it. The following mode was used to connect the several epitopes that make up the vaccine design using GPGPG and AAY linker: Adjuvant—EAAAK—B-cell (1)—GPGPG—B-cell (2)—GPGPG—B-cell (3)—GPGPG—B-cell (4)— GPGPG—B-cell (5)—GPGPG—B-cell (6)—GPGPG—HTL (1)—GPGPG—HTL (2)—GPGPG—HTL (3)—GPGPG—HTL (4)—GPGPG—HTL (5)—GPGPG—HTL (6)—GPGPG—HTL (7)—GPGPG—HTL (8)—GPGPG—HTL (9) AAY—CTL (1)—AAY—CTL (2)—AAY—CTL (3)—AAY—CTL (4).

Vaccine construct validation

To verify the integrity of the vaccine construct, it was submitted to several servers. We evaluated its antigenicity, toxicity, allergenicity, and similarity to humans. Table 3 shows that the construct is antigenic, non-toxic, non-allergenic, and non-homologous, according to the outcome.

Table 3.

Antigenicity, Toxicity, Allergenicity and Human homology test.

Parameter	Antigenicity (VAXIJEN)	Antigenicity (VIRVACPRED)	Allergenicity (ALLERTOP)	Allergenicity (ALLERCATPRO)	Toxicity (TOXINPRED2)	Toxicity (TOXDL)	Human homology
Score	0.6833	0.8001	Non-allergen	Non-allergen	Non-toxin	Non-toxin	Only Homolog to human beta defensin-3

Construct validation across the genotypes of Epstein Barr virus and EBNA-1 antigen

We corroborated our findings by conducting a homology search to the EBNA-1 protein after ruling out the EBNA-1 antigen, which has been linked to multiple sclerosis (MS). Additionally, as we created a unique vaccine candidate using a universal strategy to protect against both types of EBV (types 1 and 2), we validated our construct by running a homology search for both types. The outcome indicated that vaccine construct had no substantial hit to EBNA-1 protein while homolog to the two genotypes of EBV. The outcome suggests that the vaccine formulation can protect against both EBV strains.

Physicochemical property survey

Based on the sequence of amino acids used in the vaccine construct, Expaxy Protparam estimates several vaccine construct parameters. The construct contains a total of 421 amino acids, and its theoretical PI is 9.97, indicating that the construct has basic behavioral characteristics. The overall number of positively charged amino acid residues is 57, whereas the total number of negatively charged residues is 32. This results in a net charge of 25, which suggests that the amino acid is well-soluble in water. As the threshold for a construct should be below 40, the instability index of the construct was predicted to be 40.48, indicating a soluble construct. Mammalian reticulocyte half-life is anticipated to be 30 h, yeast half-life is predicted to be > 20 h, and Escherichia coli half-life is predicted to be > 10 h. Thermal stability in a protein construct is conferred by the aliphatic index (AI). The vaccine construct’s AI, which was computed, was 59.55, indicating a high aliphatic index. Thermostability rises with increasing AI. Last but not least, a protein construct’s Grand Average Hydropathicity (GRAVY) should be negative because this will improve the protein’s ability to interact with the water environment. The novel vaccination candidate’s GRAVY calculated value is − 0.710. This suggests that our vaccine’s construct will interact well with the water environment.

Secondary structure and tertiary structure prediction

SOPMA calculated that 78 (18.53%), 82 (19.48%), 18 (4.28%), and 43 (57.72%) amino acid residues were involved in the assembly of the vaccine candidate’s alpha helix, extended strand, beta-turn, and random coil respectively, as illustrated in Fig. 2. The five models for the vaccine candidate tertiary were predicted by the I-TASSER computer. Each structure’s reliability is based on its confidence score. The confidence scores for the anticipated 5 models range from − 1.33 to − 3.74. It was decided to use the model with the greatest C-score value (Fig. 3). The predicted TM score for the model is 0.55 ± 0.15, and its root-mean-square deviation (RMSD) is 10.1 ± 4.6 Å.

Figure 2.

Vaccine candidate secondary structure prediction.

Figure 3.

3D structure of the chimeric vaccine candidate visualized using UCSF Chimera.

3D-structure refinement and validation

The structure predicted by the I-TASSER modeler was refined and any potential mistakes were fixed using the galaxy refine tool. Model 1 (Fig. 4A) was selected from the server’s list of 5 models based on the following criteria: GDT-HA of 0.9691, RMSD of 0.353, MolProbity of 2.509, Clash score of 20.3, weak rotamers of 1.2, and Rama favored 85.9. By creating a Ramachandran map, the chosen model was validated. According to the analysis of the results, 92.491% of the vaccine construct residue are in the very favored region, 5.802% are in the preferred region, and 1.706% are in the questionable region (Fig. 4B). Disulfide engineering was used to identify potential disulfide bonds in the effort to stabilize the improved vaccination model. 33 pairs of residues were estimated to be useful for improving stability for the improved construct. Only 3 pairs of residues were chosen out of all potential pairings after being evaluated for energy level less than or equal to 2.0 and Chi 3 value between 87 and + 97 degrees, as shown in Fig. 4C. A Z-score of 3.04 was reported as shown in Fig. 4D after ProSa web server was used to assess the quality and probable errors in the vaccination 3D model.

Figure 4.

Vaccine 3D refinement and validation. (A) Galaxy refined model, (B) Ramachandran plot map, (C) Disulfide engineered structure showing possible disulfide bonds in yellow color, (D) ProSa web z-score plot.

Molecular docking, binding affinity assessment, and interaction visualization

A vaccine needs to engage well with its target receptor to successfully elicit an immune response. It was possible to accurately predict the binding posture and affinities between the vaccine and TLR 1, 2, 4, and 9 by using the Molecular Operating Environment (MOE) software. The outcome produced 100 alternative poses for each docking, and the best pose (respectively shown in Fig. 5A–5D) was selected based on the binding affinity score. According to the results, TLR 1, TLR 2, TLR 4, and TLR 9 had binding scores of − 29.81, − 40.13, − 28.93, and − 35.89 kcal/mol, respectively. Figure 5E–H depict the visualization of the docked complex, which reveals that 21 amino acids from the vaccine interact with TLR 1 (Fig. 5A) with no hydrogen bond, 39 residues do so with TLR 2 while forming seven hydrogen bonds (Fig. 5B), 19 with TLR 4 while forming three hydrogen bonds (Fig. 5C), and 29 residues do so with TLR 9 while forming two hydrogen bonds (Fig. 5D).

Figure 5.

Docked complex between (A) TLR 1-Vaccine, (B) TLR 2-Vaccine, (C) TLR 4-Vaccine, and (D) TLR 9-Vaccine (Vaccine structure in white while TLRs in yellow in surface mode) (E) Interaction visualization between TLR 1-Vaccine (Chain A-TLR 1, Chain B-Vaccine), (F) Interaction visualization between TLR 2-Vaccine (Chain A-TLR 2, Chain B-Vaccine), (G) Interaction visualization between TLR 4-Vaccine (Chain A-TLR 4, Chain B-Vaccine), (H) Interaction visualization between TLR 9-Vaccine (Chain A-TLR 9, Chain B-Vaccine as visualized by PDBSUM web server. Amino acids participating in hydrogen bond are also represented with blue lines).

Population coverage on selected epitopes

MHC class I and class II epitopes, together with related HLA alleles, were used to analyze population coverage throughout 16 geographical regions of the world available in the IEDB database. The outcome showed that all epitopes had a world coverage of 85.75% when class I and class II were combined (Table 4). Europe has the most coverage (92.52%), while Central America has the lowest (23.76%).

Table 4.

MHC 1 and MCH II combined population coverage with Europe having the highest hit (92.52%) and Central America having the lowest hit (23.76%).

Population/area	Class combined
	Coverage (%)	Average_hit
Central Africa	65.83	6.97
Cantral America	23.76	2.24
East Africa	79.21	9.08
East Asia	62.01	6.12
Europe	92.52	12.14
North Africa	79.33	8.87
North America	88.56	11.03
Northeast Asia	51.03	4.78
Oceania	62.26	6.07
South Africa	55.63	5.21
South Asia	52.24	9.59
Southeast Asia	80.91	4.7
Southwest Asia	50.53	5.78
West Africa	64.22	7.59
West Indies	70.99	8.56
World	78.73	10.13
Average	67.27	7.26
Standard deviation	17.0	2.59

Molecular dynamics simulation

Molecular dynamics was used to assess the degree of ligand fluctuation to determine the stability of the vaccination candidate within the hotspot of the target receptor. The most successful hit complex, TLR 2 and TLR 4, was subjected to a molecular dynamics simulation. The findings of the molecular dynamics simulation for the TLR 2-Vaccine complex are shown in Fig. 6A along with the RMSF fluctuation map. The complex trajectory’s movement showed that there was minimal fluctuation in the coordinates within the ranges of 1.3 and 5.5 Å. This shows that there is little distortion and a consistent binding for the vaccine. The hydrogen-participating amino acid residues 401 (1.4130 Å), TYR 400 (1.4020 Å), and PRO 215 (3.0510 Å) maintained a stable trajectory. This might provide adequate stability in the receptor’s pocket. The RMSF map for the TLR 9-Vaccine, with a trajectory range from 0 to 5.7 Å, is likewise shown in Fig. 6B. The amino acid involved in the hydrogen bond, PRO: 155 (2.7980 Å), ARG: 157 (2.5280 Å), likewise preserved stability. Spring models in the colors yellow-orange-red show the degree of contact and stability of the binding complex system (Fig. 6C,D).

Figure 6.

Molecular dynamics simulation plot showing (A) RMSF plot for TLR 2-Vaccine, (B) RMSF plot for TLR 9-Vaccine, (C) Contact map for TLR 2-Vaccine, (D) Contact map for TLR 9-Vaccine.

Immune simulation

By using an in silico immune simulation technique over 1050 simulation steps, we evaluated the vaccine construct’s ability to induce an immunological response. This technique is employed, among other things, to analyze the ability of the vaccine design to trigger an immune response and its antigens. By investigating the vaccine candidate, B cells, T cells, and memory cells that produce immune responses to combat viral infections were evaluated. Our in silico experiments’ findings demonstrated the vaccine candidate we developed’s potency. Results showed that the T-cell populations (helper T cells and cytotoxic T cells), B-lymphocytes, Natural killer cells, Dendritic cells, Macrophages, Epithelial cells, and Cytokines with durable memory cells were produced to a maximum peak and found to be crucial participants in the elicitation of initial and subsequent immunological responses (Fig. 7).

Figure 7.

Immune simulation response showing (A) Antigen and immunoglobulins production, (B) B-lymphocytes total count, memory cells and sub-divided in isotypes IgM, IgG1 and IgG2, (C) Plasma B lymphocytes count sub-divided per isotype (IgM, IgG1, and IgG2), (D) CD4 T-helper lymphocytes count showing total and memory counts, (E) CD8 T-cytotoxic lymphocytes count showing total and memory count, (F) Natural killer cells total count, (G) Dendritic cells which present antigen to MCH 1 and II, (H) Macrophage total count, (I) Epithelial cells total count, (J) Concentration of cytokines and interleukins.

Construct functionality test

We examined potential molecular function, biological process, and cellular component of the vaccine construct in an effort to further confirm the validity of the unique vaccine candidate. Table 5 below displays the best hit for the results.

Table 5.

Various function exacted by the vaccine candidate.

Function	Description	PFP score
Molecular function	CCR6 chemokine receptor binding	857.86
	Chemoattractant activity	670.68
Biological process	Chemotaxis	897.00
	Defense response	557.31
	Immune response	224.30
	Immune system process	180.75
Cellular component	Intracellular	1512.81

Codon optimization and in-silico cloning

Using the sequence length, CAI value, and GC content of 1263 nucleotides, 1.0, and 60.25%, respectively, the vaccine construct was codon optimized by the JCat service. Our findings indicate a higher likelihood of multi-epitope vaccine expression in the bacterial system because the ideal range of GC content is 30–70%. Additionally, multiple cloning sites (MCS) of the pET28a(+) expression vector were used to frame the XhoI and Xbai restriction sites at the N- and C-terminal of the optimized DNA sequence, respectively (Fig. 8). The pET28a(+) plasmid is a universal vector for the cloning and production of recombinant proteins due to the presence of a poly-histidine affinity tag.

Figure 8.

Cloning map generated by SnapGene using XhoI and Xbai restriction site of pET28a( +). The optimized nucleotide sequence for the vaccine construct is depicted in red color on the map.

Discussion

The best and most affordable method for halting and controlling deadly virus outbreaks is vaccination. A robust humoral and cellular immune response against the target virus and infected cells is a requirement for a successful vaccine. The antigen chosen has a significant impact on the vaccine’s effectiveness⁷¹. Identification of epitopes on the surface of antigen is considered important for designing an epitope-based peptide vaccine (EBPV)^11,52. Different strategies have been used to create a therapeutic EBV vaccine that targets the envelop glycoproteins^72–75, with less consideration on EBV nuclear antigens (EBNA 1, EBNA 2, EBNA 3A, EBNA 3B, EBNA 3C, EBNA-LP), capsid protein (EBV-VCA), latent proteins (LMP 1, LMP 2A, LMP 2B), and early antigen (EBV-EA). EBNA 2 and EBNA-LP, two nuclear antigens that work as transcription activators in tandem to control the expression of viral and cellular genes involved in the start and upkeep of cell proliferation, are the first viral genes produced in EBV-infected B cells^76,77. The viral membrane latent genes LMP 1 and LMP 2 are upregulated by EBNA 2 and EBNA-LP, and the BamHI-C promoter, which controls the production of all nuclear antigens, is also controlled⁷⁸.

In a research conducted by Lanz et al.⁷⁹, Multiple sclerosis (MS) and the EBNA 1 antibody were found to be related. The central nervous system protein GlialCAM and the EBNA 1 antibody were found to interact in a way that encourages autoreactive lymphocytes to assault the myelin sheath (CNS). Additionally, in a mouse model, immunization against EBNA 1 worsens MS. In the current study, we developed a novel vaccine candidate with nuclear antigens, capsid proteins, and early antigen to provide therapeutic effectiveness against EBV such that the antibodies produced against the vaccine will be effective in inhibiting the first viral proteins expressed (EBNA 2 and EBNA-LP), which make it easier to regulate the expression of the other EBNA and LMP genes and reprogram cell gene, as well as providing therapeutic effectiveness against EBV⁸⁰. EBNA-3A and 3C have also been reported to be essential for B-lymphocyte growth while EBNA-2 and LMP-1 are essential for cell immortalization. When EBNA-2’s ability to connect to the cellular DNA-binding protein CBF1 and imitate active Notch signaling is successfully inhibited, viral multiplication and the expression of additional antigens are always stopped.

The most antigenic linear B-cell and T-cell epitopes from individual EBV antigens were included into the vaccine design. Though, high antigenicity is not the only parameter to be considered while choosing epitopes, we also based our judgment on their toxicity profile, allergenicity, and human homology. According to reports, linear B cell epitopes contribute to the neutralization of viruses³⁰ while MHC class I antigen presentation, cytotoxic T lymphocyte (CTL) rearrangement, and activation of T helper cells are essential for creating chimeric vaccines, they are also an essential component of the adaptive immune response^47,58. All of the antigenic epitopes that had been chosen were examined for their allergenicity, toxicity, human homology, conservancy, IL-4, IL-10, and IFN-gamma (the last three assessments were only made for MHC-II) and were found to be promising epitopes for the development of a sub-unit protein vaccine due to the results of their tests. Due to their repeat in both types of EBV, 6 B-cell epitopes, 4 MHC-1 epitopes, and 9 MHC-II epitopes were given preference after this study (see Table 2). Together with human Beta-defensin 3 adjuvant, these epitopes were fused. This adjuvant has been discovered to activate the TLR target of EBV (TLR 2 and 9)^35,36. Most viruses including EBV evoke primary innate immune response during infection by activating specific TLR(s). EBV has the capability to activate monocytes and plasmacytoid dendritic cells (pDCs), release of IL-8 and MCP-1, and recognition of CpG motifs in the genome of EBV by working together with TLR9 and TLR2^81,82. Using EAAK, GPGPG, and AAY linker, the adjuvant epitope, B-cell epitopes, and T-cell epitopes were connected (see primary vaccine construct above). An optimal vaccine’s ability to prompt a quicker immune response depends on how well the adjuvant chosen to boost the host immune response works^83,84. By directing antigen to antigen presentation cells (APCs), releasing cytokines that trigger Th1 or Th2 immune responses, generating cell-mediated immunity, and lowering the amount of antigen required for protective vaccination, adjuvants frequently improve vaccine effectiveness⁵⁸. To validate the efficacy of the preventive vaccine construct for the two genotypes of EBV, the vaccine construct was tested across both genotypes. According to the results, both genotypes’ antigen proteomes had a high percentage of significant hits. This demonstrated that both EBV type 1 and 2 immune responses can be induced by the construct. Blasting was done throughout the EBNA 1 antigen to ensure there were no hit sequence matches, and the results showed no substantial hits to the proteome make-up of the antigen.

The new vaccine candidate’s physicochemical characteristics validated its validity as a credible design. The molecular weight (MW) was 450.6 kDa, which is greater than the recommended cutoff value of 110 kDa for a promising vaccine candidate²³. The construct theoretical PI was predicted to be 9.97, giving the vaccine candidate a basic nature. The vaccine construct’s instability index is 40.48, which is just a little bit above the cutoff of 40⁸⁵. The construct’s calculated aliphatic index (59.55) demonstrated encouraging thermostability, implying high stability in temperate regions⁸⁶. The GRAVY value indicates how proteins behave in a water environment. A negative GRAVY value means that the protein is well soluble in water⁸⁷. The construct assumes a GRAVY score of − 0.710 indicating hydrophilic interaction with the water environment, thus, further validating the potential of the vaccine construct.

To better understand the nature of the construct and how it interacts with the target receptor, the secondary and tertiary structure of the vaccine is crucial in the development of a vaccine candidate²³. The analysis of the secondary gave mostly random coils (57.72%). After predicting the tertiary structure of the construct which gave a confidence score of − 1.33 assuming the highest score among the 5 predicted models with a TM score of 0.55 ± 0.15, it was then subjected to refinement and validation. The similarity between the two structures is gauged by the TM-score. A TM-score greater than 0.5 denotes a valid topology model, whereas one less than 0.17 denotes random similarity⁸⁸. Thus, the model provided as the predicted tertiary structure is credible and reliable. The refinement of the construct was done to improve its integrity. The refined model has an improved Rama favored from 71.6 to 85.9, its poor rotamers reduced from 13.7 to 1.2 and GDT-HA score of 0.9691. All these results suggest an improved integrity to the structure of the vaccine. To further validate the improved model, Ramachadran plot analysis gave a 92.491% amino acid residue in highly favorable region, 5.802% are in preferred region while about 1.706% are in questionable region. This has further proved better vaccine 3D construct. The vaccine’s model quality is closer to the protein structure revealed by X-ray crystallography, according to ProSa Web’s Z-score of − 3.04, which indicates greater vaccine^43,50

The best way to determine whether a vaccine candidate will interact successfully with the target protein receptor is by molecular docking⁷¹. The vaccine 3D structure was docked against TLR 1, TLR 2, TLR 4 and TLR 9. The result obtained showed a binding score of − 29.81, − 40.13, − 28.93, − 35.89 kcal/mol for TLR 1, TLR 2, TLR 4, and TLR 9 respectively. The high binding affinity have donned by TLR 2 and TLR 9 is not a surprise since report has proved that these two TLRs get activated by EBV² and also received inducing help from the adjuvant used. The interacting amino acid showed greater amount of vaccine residues interacting with TLRs. TLR 2 and TLR 9 also showed hydrogen bond interaction (7 and 2 respectively) which can increase the stability of the vaccine around the receptor. The molecular dynamics analysis depicts a steady binding affinity (see Fig. 6A,B).

The effectiveness of the vaccine design to generate an immune response was examined using immune simulation that mimicked the human host body system. The IgM and IgG counts against the antigen created by the plot reached 700,000. Natural killer cells, dendritic cells, macrophages, epithelial cells, the production of cytokines, memory B cells, and memory T cells—the memory in B cells lasting several months—were all clearly developing. This is a sign of humoral immunity and is necessary to support the immune response. Additionally, helper T cells were strongly activated, demonstrating the vaccine construct’s ability to provide EBV protection.

A protein functionality test was carried out to further confirm the specific function of the vaccination candidate. The outcome demonstrated that the design would only have one biological or molecular function, triggering an immune response and a defensive response. The nucleotide sequence’s (Guanine–Cytosine) GC content and the value of the codon adaptive index (CAI) will always describe how easy it is to clone and express a construct. Between 0.8 and 1.0 is the CAI threshold for an enhanced nucleotide sequence⁸⁹ and GC content should only range between 30 and 70%¹². The constructed multi-peptide vaccine showed a CAI value of 1.0 and GC content of 60.25%. This shows better expression and cloning ability. This will increase cloning success into vector for large scale production during clinical trials.

With an average of 85.75% of the world’s population, the population coverage of the used T-cell epitopes demonstrated a wide range of efficiency across various ethnicities and geographical regions. This further certifies that the vaccine construct will cover enough world population. Bioinformatics, computational, and modeling techniques are being used in this work to build a potent and effective vaccine candidate against EBV, and all results thus far have shown that this candidate vaccine is not only effective but also safe, non-allergenic, and non-toxic.

Conclusion

The process involved in vaccine production in old times takes several years to become effective and available for consumer use. Moreover, these vaccines are rather expensive. However, the advent of Bioinformatics and computer aided therapeutic discoveries has irrevocably helped in the area of vaccine design. Vaccine as a form of infection prevention, stands as a major approach to effectively roll out diseases. The use of immunoinformatic approach to vaccine construction is not only fast but also help to accurately use antigenic, non-toxic and non-allergic regions of target organism to construct a potential vaccine candidate. The result of the novel multi-epitope vaccine construction showed 6 B-cell epitopes, 4 CTL epitopes and 9 HTL epitopes that are highly antigenic, non-toxic, non-allergic and non-human homolog. The vaccine has been found to interact well with TLR2 and TLR9 and from this result, it can be conclusively said that the vaccine will have full potential of inducing immune response. Also, it was found that the novel vaccine candidate has a world coverage of 85.75% which implies a great geographical coverage. Though, computationally, it has been well validated that the multi-epitope vaccine construct will effectively knock-out EBV even before the expression of EBNA 1, further in-vivo and in-vitro analysis is projected to validate its potency. The in-vivo study will include animal model studies which is directly administering the vaccine to a laboratory animal. The process will involve challenge study, monitoring of the models, protection assessment, and dosing study. In-vitro assay on the other hand will involve a cell culture assay where immune cell activation, antibody assay, T-cell response, and neutralization assay will be studied.

Abbreviations

GP

Glycoprotein

HTL

Helper T lymphocyte

CTL

Cytotoxic T lymphocyte

TLR

Toll like receptor

MS

Multiple sclerosis

EBNA-1

Epstein Barr virus nuclear antigen-1

GRAVY

Grand average hydropathicity

MHC

Major histocompatibility complex

SB

Strong binders

WB

Weak binders

BLAST

Basic local alignment search tool

NMR

Nuclear magnetic resonance

RMSD

Root mean square of deviation

MOE

Molecular Operating Environment

HLA

Human leukocyte alleles

MD

Molecular dynamics

GC

Guanine–cytosine

CAI

Codon adaptive index

Author contributions

Conceptualization, original draft writing, reviewing, and editing: EKO, TOO, OEE, OQB, MPO. Formal analysis, investigations, funding acquisition, reviewing, and editing: ATO, TTO, BA, DAA, OAO. Resources, data validation, data curation, and supervision: SEO, AMS, MB, YAY, TIA.

Data availability

All materials used will be available upon reasonable request. All requests should be directed to the corresponding author.

Competing interests

The authors declare no competing interests.