Immunoinformatics-Based Design of Broad-Spectrum Multi-Epitope Vaccines Targeting Mutations in Emerging SARS-CoV-2 Variants

doi:10.21203/rs.3.rs-5200837/v1

Download PDF

Research Article

Immunoinformatics-Based Design of Broad-Spectrum Multi-Epitope Vaccines Targeting Mutations in Emerging SARS-CoV-2 Variants

https://doi.org/10.21203/rs.3.rs-5200837/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The ongoing COVID-19 pandemic, intensified by emerging SARS-CoV-2 mutations, highlights the urgent need for enhanced vaccines. Despite considerable efforts in vaccine design, improvements are still required in formulating vaccines targeting the novel coronavirus. This study, utilized immunoinformatics and reverse vaccinology to design multi-epitope vaccines targeting emerging variations. B and T cell epitopes were generated by analyzing the mutation sites of the prevalent variant strains, and two vaccines were designed by linking with two different adjuvants. Interaction of the model vaccines with four Toll-like receptors (TLR) revealed a relatively high affinity between vaccines and immune receptors. Codon optimization and computational cloning were conducted to validate the robustness of the multi-epitope vaccines and immunogenic simulations were performed to assess the antigenicity and antibody generation capability of the vaccine. The L455S mutation in the JN.1 variant and its adjacent F456L mutation on antibody effectiveness against the XBB variant revealed that 15 antibody structures maintained a certain level of binding affinity. This study offers an immunological evaluation from a mutation-centric perspective and integrates co-evolutionary analysis with immunoinformatics to design effective multi-epitope vaccines targeting various SARS-CoV-2 strains. The methodologies applied in this research can also be extended to the vaccine development for other pathogens.

SARS-CoV-2

Vaccine

Immunoinformatics

Epitopes

Toll-like receptors

Since the World Health Organization declared severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) a public health emergency of international concern, the global tally of COVID-19 cases has approached approximately 776 million, with around 7.06 million reported deaths as of August 4, 2024 (https://covid19.who.int/). Owing to treatments like antiviral drugs and vaccines, most individuals has gained immunity against SARS-CoV-2 through infection or vaccination [1]. However, threatens of the rapid antigenic drift of Omicron subvariants [[2],[3],[4],[5]], especially those immune-evasive variants, and the waning vaccine protection over time [[6],[7],[8],[9]] highlight the need for long term surveillance. Despite that BA.2.86 is never dominant worldwide, its descendan JN.1 has been a major variant of interest (VOI), harbouring S:L455S mutation, which contributes to its immue-evasion ability, and accounts for 70% of all circulating vatiants by February 2024 [10]. JN.1 and its two sublineages KP.2 and KP.3 account for about 20% of circulating variants by July 2024, there are chances for these variants to spread due to their immune-evading ability [11]. The emergence of these new lineages formed through mutation and recombination processes has led to successive waves of infections, highlighting the urgent need for the next generation of vaccines to ensure continued key defense against SARS-CoV-2 [12].

The novel coronavirus, with a genome length of 29.9 kb, is a non-segmented positive-sense RNA virus. Its entry into host cells occurs through binding with the angiotensin converting enzyme 2 (ACE2) receptor at specific target sites [13]. Vaccination was the primary strategy for preventing SAR-CoV-2 infections. Currently, there are about four platform technologies for approved vaccines s: inactivated virus particles, mRNA technology, adenovirus vector delivery and protein subunit formulations with adjuvants [14]. While inactivated vaccines are relatively easy to produce, they are challenging to scale up and their immunogenic efficacy may fall short of optimal standards. mRNA vaccines boast high efficacy, but accompanied by significant costs and stringent cold chain requirements [15]. Despite the widespread adoption of adenovirus vector delivery, its suitability remains under scrutiny for individuals with compromised immune function, given potential prior adenovirus exposure and resultant immunity [16].

In this context, subunit vaccines have been strategically chosen as the focus of this research. These vaccines typically exclude the entire virus, potentially enhancing safety by minimizing potential risks and adverse effects. Their production process is characterized by relative simplicity, involving specific protein antigens exclusively, thereby enhancing stability during storage and transportation [17]. Adjuvants play a pivotal role in modulating the humoral and cellular immune responses induced by subunit vaccines [[18],[19]]. They not only stimulate stronger immune responses against antigens but also possess the potential to reduce vaccine dosages and production costs [[20],[21]]. Research has demonstrated the effectiveness of Toll-like receptor (TLR) agonists in activating innate immunity across various vaccine adjuvants and immunomodulators targeting infectious diseases and cancer [22]. An in vitro study has shown that TLR4 can detect the S protein of SARS-CoV-2 [23]. Additionally, research by Chakraborty et al. has indicated that TLR5 can stimulate early signals, providing protective innate immunity against respiratory infections [24]. TLR3 has been shown to induce protective responses through TLRF pathway against previous infections with SARS-CoV and MERS-CoV [25]. A preprint study suggests that inhibiting endothelial TLR3 may be a potential mechanism for SARS-CoV-2 infection-related pulmonary vascular remodeling, and enhancing TLR3 signaling transmission could be a potential therapeutic strategy [26]. Furthermore, TLR7 can detect the binding of surface S glycoproteins with ACE2 on the virus envelope [27]. Hence, TLRs play a crucial role in initiating vital immune responses against viral infections.

In the current landscape, advancements in genomics and computational biology have introduced novel methodologies and tools for studying immune response. Particularly, the emergence of methods such as immunoinformatics, which uses a variety of computer software to deeply study various aspects of the immune response and predict effective vaccine structures, has accelerated the pace of vaccine development, consequently saving substantial time and resources [28]. This study aims to develop a broader and more effective multi-epitope vaccine against SARS-CoV-2, with particular focus on the variants HV.1, HK.3, BA.2.86, and JN.1, which are currently prevalent. HV.1 and HK.3 are descendants of the XBB.1 variant, and BA.2.86 and JN.1 are descendants of the BA.2 variant. Given that certain mutations may modify infectivity of virus, disease severity, and host interactions, potentially affecting the efficacy of certain vaccines [29], this study retraces the ancestral strains of these four variants and analyzes newly emerged mutations. Reverse vaccinology techniques and various computational tools were used to identify and analyze antigenic epitopes targeting SARS-CoV-2 based on mutation information that can bind to different HLA molecules, eliciting potent immune responses. Importantly, this study delves into the repercussions of the newly emerged L455S mutation in the JN.1 variant, specifically evaluating the effectiveness of neutralizing antibodies previously proven efficacious against XBB when encountering this mutation.

In summary, this research underscores the importance and potential of immunoinformatics and reverse vaccinology in designing multi-epitope vaccines targeting SARS-CoV-2. These methods not only aid in expediting the vaccine development process but also provide us with a deeper understanding to address the challenges posed by viral mutations.

Acquisition and analysis of new mutations in target variants

The tracing of variants HV.1, HK.3, BA.2.86, and JN.1 led to the identification of 15 noteworthy variants, accompanied by the compilation of newly emerged mutations extracted from Cov-Lineages (https://cov-lineages.org/lineage_list.html) and outbreak.info (https://outbreak.info/).

Retrieval of target proteins sequence

The animo acid sequence of S protein (Accession No.: YP_009724390.1), M protein (Accession No.: YP_009724393.1), N protein (Accession No.: YP_009724397.2), ORF1a protein (Accession No.:YP_009725295.1), ORF1b (Accession No.: BCN86436.1), ORF7b protein (Accession No.: YP_009725318.1) and ORF10(Accession No.: YP_009725255.1) were retrieved from NCBI protein database (https://www.ncbi.nlm.nih.gov/) in FASTA format.

Epitopes prediction

Cell-mediated immunity plays a pivotal role in providing resistance to diseases and promoting vaccine-induced protection against COVID-19 [30]. In humans, T-cell activation relies heavily on antigen presentation by the human leukocyte antigen (HLA) [31]. Alleles of HLA-A, HLA-B, HLA-C and HLA-DRB1 with frequency definitions exceeding 0.2 worldwide were selected from the Allele frequency net database (http://www.allelefrequencies.net/). To identify potential CD8 + epitopes interacting with HLA Ⅰ class alleles, the NetMHCpan4.1 (https://services.healthtech.dtu.dk/services/NetMHCpan-4.1/) server was employed. Epitopes of 9 and 10 amino acids in length were predicted, with binding thresholds set at 0.5% and 2% for strong and weak binding respectively. Additionally, for the prediction of CD4 + T epitopes, the NetMHCIIpan 4.3 (https://services.healthtech.dtu.dk/services/NetMHCIIpan-4.3/) server was utilized to forecast the binding affinity and percentile grade of 15 amino acids in length interacting with HLA Ⅱ class alleles. Strong and weak binder identification was based on binding strength thresholds of 1% and 5%, respectively. Moreover, B-cell epitopes play a crucial role in inducing B-lymphocytes to produce antibodies [32]. The ABCPred (https://webs.iiitd.edu.in/raghava/abcpred/) server was used to predict linear B-cell epitopes spanning 16 amino acids, with scores above 0.6 included in the candidate list.

Epitopes processing

The predicted epitopes were further selected for antigenicity using the Vaxijen v.2.0 server (http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html), with a threshold for viruses antigens set at 0.4 [33]. Additionally, allergenicity assessment was conducted using AllerTOP v.2.0 server (https://www.ddg-pharmfac.net/AllerTOP/), considering amino acid properties such as hydrophobicity, size, abundance, and tendencies to form helix and β-strand [34]. Furthermore, ToxinPred (http://crdd.osdd.net/raghava/toxinpred/) was employed to predict toxicity of the protein sequences. Moreover, a range of different physicochemical characteristics of these epitopes were elucidated through ProtParam tool (https://web.expasy.org/protparam/) available on the ExPASy platform. For T cell epitopes, the immunogenicity was further assessed using Class I Immunogenicity module of the Immune Epitope Database (IEDB, http://tools.iedb.org/immunogenicity/) to assess class Ⅰ epitopes. Additionally, the convolutional neural network model DeepNeo (https://deepneo.net/) was utilized to predict the immunogenicity of class Ⅱ epitopes. To determine the similarity of potential epitopes to host proteins, a BLASTP analysis (https://blast.ncbi.nlm.nih.gov/Blast.cgi) was conducted against the nonredundant protein sequence database, with Homo sapiens (NCBI taxid: 9606) selected as the target organism [35]. An e-value cutoff of 0.01 was applied to filter the results [36].

Population coverage

To ensure the effective performance of a multi-epitope vaccine, it is imperative that the epitopes possess broad HLA binding specificity, enabling to produce robust immune response across diverse populations worldwide. To assess the population coverage of the selected HLA Ⅰ and HLA Ⅱ epitopes for vaccine construction, the IEDB (http://tools.iedb.org/population/) population coverage tool was employed.

Designing of multi-epitope vaccine

The immunogenicity drawback of epitopes vaccines can be surmounted by combining antigenic epitopes with appropriate adjuvants to construct a multi-epitopes vaccine [37]. In the construction process, CTL epitopes were linked with AAY linker, while HTL epitopes were conjoined both to each other and to CTL epitopes using GGPPG linker. Subsequently, B cell epitopes followed HTL epitopes with KK linker. Additionally, a 6xHis tag (HHHHHH) was attached to the carboxyl terminus of the vaccine construct. In this study, two different adjuvants, 50s Ribosomal L7/12 (Accession No.: WP_000028878.1) [38] and Human Beta-defensin 3 (Accession No.: AAV41025.1) [39] were used, each attached at the N-terminus of the vaccine. The fusion between the adjuvant and the vaccine construct was achieved using an EAAAK linker [40]. These two vaccines with different adjuvants underwent comprehensive validation once again by Vaxijen v.2.0, AllerTOP v.2.0, ToxinPred, ProtParam and Protein-sol (https://protein-sol.manchester.ac.uk/patches) for antigenicity, allergy, toxicity, physicochemical properties and solubility, respectively.

Predicting and verifying of multi-epitope vaccine

The secondary structure elements of the complete vaccine component, such as alpha-helix, extended strand, and random coil, were predicted using the PSIPRED server (http://bioinf.cs.ucl.ac.uk/psipred/) [41]. These elements form the structural scaffold of the protein, which is crucial for predicting its function, identifying active sites, determining interaction regions and ensuring stability [42]. Subsequently, the three-dimensional structure was modelled using the I-TASSER (https://zhanggroup.org/I-TASSER/). This server employs multiple threading alignments to search for template PDB structures for modeling, and through iterative fragment assembly simulations, ultimately establishes the three-dimensional structure of the target sequence [43]. Follow this, the three-dimensional structure underwent further refinement using the Refine tool in the GalaxyWEB server (https://galaxy.seoklab.org/) to enhance model accuracy. To assess the refined tertiary structure, the ProSA-web server provided a z-score indicating discrepancies between the tertiary vaccine structure and models derived from nuclear magnetic resonance (NMR) and X-ray diffraction (XRD). Additionally, the Ramachandran plot of the vaccine structure was generated by RamachanDraw library in Python.

Prediction of B-cell conformational epitopes

The folding of a protein spatially aligns distant residues, resulting in the creation of discontinuous B-cell epitopes. Unlike linear epitopes, conformational epitopes account for a larger proportion of B-cell responses, making their prediction pivotal in vaccine design [44]. Discontinuous B-cell epitopes within the developed construct were predicted using the ElliPro server (http://tools.iedb.org/ellipro/) [45].

Disulfide engineering of multi-epitope vaccine

Disulfide bonds are integral to the function and stability of many naturally occurring proteins [46]. For multi-epitope vaccines, structural stability is essential to maintain immunogenicity and efficacy. The Disulfide by Design 2.13 server (http://cptweb.cpt.wayne.edu/DbD2/) was employed to evaluate and design potential disulfide bonds. Given that 90% of natural disulfide bonds exhibit energy values below 2.2 kcal/mol and χ³ dihedral angles peaking between − 87°and + 97°, residue pairs satisfying these criteria were selected to form disulfide bonds [47].

Molecular docking analysis of multi-epitope vaccine

Molecular docking is employed to estimate the binding affinity between vaccine construct and immune cell receptors, aiming to determine the optimal vaccine conformation to generate a protective immune response. Four immune receptors, namely TLR3 (PDB ID: 1ZIW), TLR4 (PDB ID: 3FXI), TLR5 (PDB ID: 3J0A) and TLR7 (PDB ID: 7CYN) were used in this study. The docking process was performed using the protein-protein docking panel of Schrödinger software (LLC, NY, USA, 2023-1), which is based on PIPER algorithm for sampling. The ranking of docked conformations was established by identifying the centers of highly populated clusters of the low-energy conformations [48]. The most populated clusters were selected for subsequent structural analysis.

Molecular dynamics simulation

Molecular dynamic simulation analysis is a computational approach commonly employed to analyze of the dynamic behavior of molecules within a specific environment and time frame. In the context of evaluating the binding affinity and conformational stability of vaccine construct and immune cell receptors compound, molecular dynamics simulations were conducted using Schrödinger software (LLC, NY, USA, 2023-1) over a duration of 100 nanoseconds under NPT ensemble conditions, which maintained constant particle number, pressure and temperature. During the simulation, the root mean square deviation (RMSD), root mean square fluctuation (RMSF) and radius of gyration (Rg) values of vaccine construct and receptor were monitored and analyzed. The resulting trajectory from the simulation provides insights into the conformational changes occurring complex structure over the course of the simulation period. Furthermore, the binding free affinity and dissociation constant (K_D) were calculated using the PRODIGY web server (https://wenmr.science.uu.nl/prodigy/), which relies on intermolecular contacts and properties extracted from the non-interface surface [49].

Immune simulation

The analysis of immune response elicited by the vaccine construct was conducted using the C-ImmSim immune stimulator (https://kraken.iac.rm.cnr.it/C-IMMSIM/index.php). In this study, the simulation parameters were set as follows: a randomized seed of 12345, a simulated volume of 10, and a simulated step number of 1000. The vaccine was administered every four weeks for a total of 3 doses to assess the capacity of immune cells to generate specific antibodies and various cytokines. Given that the injection time step per unit of this server corresponds to 8 hours in the real-world time, the time steps for the three injections were designated as 1, 84 and 168, respectively. This enabled the simulation to accurately simulate the temporal dynamics of the immune response following vaccination.

Codon Adaptation and In Silico Cloning

To ensure efficient expression and production in the specified host organism, protein sequences were back-transcribed using the Sequence Manipulation Suite (https://sites.ualberta.ca/~stothard/javascript/rev_trans.html) to produce DNA sequences representing the most probable non-degenerate coding sequences [50]. The optimal sequences were evaluated based on the codon adaptive index (CAI) and GC content score, which were calculated using GenScript (https://www.genscript.com/tools/rare-codon-analysis). CAI offers valuable codon usage information, with a score above 0.8 being considered favorable. Meanwhile, an optimal GC content should full within the range of 30%-70% [51]. E. coli strain was chosen as the host organism due to its favorable characteristics, such as ease of handling and suitability for mass production [52]. Finally, SnapGene software version6.0.2 (https://www.snapgene.com/free-trial) was used to adapt the sequence, which was then inserted between XhoI and SalI sites of the pET28a (+) vector.

Acquisition of neutralizing antibody structures and analysis of hotspot mutations

To assess the impact of the newly added L455S mutation on the S protein of the JN.1 variant, a comprehensive evaluation of neutralizing antibody efficacy was conducted by combining it with the adjacent F456L mutation. The antibody structure was obtained from the CoV-AbDab (https://opig.stats.ox.ac.uk/webapps/covabdab/) database. The effective neutralizing antibodies against SARS-CoV-2-Omicron XBB variant were selected, with a focus on the receptor binding domain (RBD) region of S protein. The L455S and F456L mutations analysis of neutralizing antibody was calculated using Residue Scanning Calculation module of Schrödinger software (LLC, NY, USA, 2023-1). The results post-mutation was visualized using PyMOL software.

Identification of new mutations in target variants

A total of 55 novel mutations were identified, including 35 mutations located on the S protein, 12 mutations on the ORF1a protein, 3 mutations on the M protein, 2 mutations on the N protein. Additionally, 1 new mutation each was found on the ORF1b, ORF7b and ORF10 proteins. (Fig. 1)

Identification of epitopes

Using various immunoinformatics tools, candidate epitopes were identified based on criterias including non-toxicity, non-allergenicity, and antigenic stability. Positive immunogenicity scores determined by the class Ⅰ Immunogenicity module of the IEDB database guided the selection of HLA class Ⅰ epitopes. Additionally, according to the definition of DeepNeo, HLA binding prediction score and T cell reactivity scores above 0.5 were interpreted as immunogenic neoantigen, leading to the selection of high-scoring HLA class Ⅱ epitopes. From the initial pool of 63 epitopes, those with unfavorable physical or redundant HLA restrictions were eliminated, ensuring the best overall properties. Efforts were also made to ensure non-repetitive sequences were retained in the final set. Consequently, a refined set comprising 6 HLA class Ⅰ epitopes (Table 1), 9 HLA class Ⅱ epitopes (Table 2) and 9 B cell epitopes (Table 3) were determined. All 24 selected epitopes passed the homology test, with e-values greater than 0.01. It should be noted that matches to proteins described as neutralizing monoclonal antibodies isolated from vaccinated individuals or survivors, despite being listed under Homo sapiens, were not considered relevant for autoimmunity risk, as these do not constitute naturally occurring human proteins. Thus, the 24 epitopes are pathogen specific, minimizing the risk of inducing autoimmunity. A detailed list of these epitopes was provided in Supplementary Table S1.

Table 1

Final mutated HLA class Ⅰ epitopes for vaccine construct
Mutation	Epitopes	HLA	Theoretical pI	Instability index	Half-life (h)	Aliphatic index	GRAVY	Antigenicity score	Class I Immunogenicity
S: A264D	GWTAGAADYY	A*01:01	3.80	1.67	30	30.00	-0.310	0.6426	0.17646
ORF7b: F19L	LLLLVLIML	A02:01/A02:07	5.52	8.89	5.5	335.56	3.711	0.5654	0.02472
M: T30A	LVIGFLFLA	A*02:06	5.52	8.89	5.5	216.67	3.011	1.1280	0.27318
ORF1a: A3143V	IVYIICISTK	A03:01/A11:01	8.20	-7.03	20	185.00	1.800	1.1480	0.21816
ORF1a: A3143V	LVPFWITIVY	B15:02/B15:06/B35:01/B35:05/B46:01/B56:01/B*56:02	5.52	31.49	5.5	175	1.950	1.0210	0.63248
M: A104V	ASFRLFVRTR	A11:01/A68:01	12.30	9.00	4.4	78.00	0.040	0.4450	0.2985

Table 2

Final mutated HLA Ⅱ epitopes for vaccine construct
Mutation	Epitopes	Theoretical pI	Instability index	Half-life (h)	Aliphatic index	GRAVY	Antigenicity score	HLA	MHC Binding	TCR reactivity
M: A104V	FRLFVRTRSMWSFNP	12.30	36.71	1.1	45.33	-0.233	0.7865	DRB1*04:07	0.9678	0.8775
M: A104V	FRLFVRTRSMWSFNP	12.30	36.71	1.1	45.33	-0.233	0.7865	DRB1*07:01	0.9323	0.7877
S: N211I	GYFKIYSKHTPIILV	9.53	-4.19	30	123.33	0.48	0.8472	DRB1*07:01	0.9673	0.7964
S: E554K	GTGVLTKSNKKFLPF	10.30	39.48	30	71.33	-0.160	1.1170	DRB1*11:03	0.9327	0.7955
								DRB1*14:01	0.8933	0.8350
								DRB1*14:08	0.7313	0.7888
S: A264D	GAADYYVGYLQPRTF	5.83	32.69	30	58.67	-0.273	0.8665	DRB1*12:01	0.9633	0.6774
S: A264D	GAADYYVGYLQPRTF	5.83	32.69	30	58.67	-0.273	0.8665	DRB1*15:02	0.9845	0.8344
S: H245N	GINITRFQTLLALNR	12.00	32.97	30	136.67	0.247	0.6084	DRB1*08:02	0.9816	0.7463
								DRB1*15:01	0.9951	0.8441
								DRB1*15:02	0.9963	0.8451
								DRB1*15:03	0.9952	0.8499
								DRB1*16:02	0.9974	0.8384
ORF1a: K1973R	DYRHYTPSFKKGAKL	9.82	20.37	1.1	32.67	-1.373	0.9007	DRB1*08:02	0.8570	0.8120
ORF1a: K1973R	DYRHYTPSFKKGAKL	9.82	20.37	1.1	32.67	-1.373	0.9007	DRB1*11:01	0.9397	0.7715
S: L452R	YNYRYRLFRKSNLKP	10.55	-28.93	2.8	52.00	-1.613	0.4998	DRB1*08:02	0.9934	0.9113
								DRB1*08:03	0.7089	0.8804
								DRB1*11:01	0.9955	0.9005
								DRB1*11:03	0.9947	0.9014
								DRB1*12:02	0.9880	0.8941
								DRB1*14:01	0.9918	0.9056
								DRB1*14:02	0.9662	0.8697
								DRB1*14:08	0.9797	0.8902
ORF1a: V1056L	KPTLVVNAANVYLKH	9.70	-6.38	1.3	123.33	0.147	0.6732	DRB1*03:01	0.9711	0.8329
								DRB1*09:01	0.9656	0.8384
								DRB1*12:01	0.9805	0.8685
								DRB1*12:02	0.9729	0.8666
								DRB1*14:01	0.9723	0.9052
								DRB1*14:02	0.9777	0.8110
								DRB1*14:08	0.8846	0.8799
								DRB1*15:01	0.9860	0.9037
								DRB1*15:02	0.9894	0.9043
								DRB1*15:03	0.9755	0.9069
								DRB1*16:02	0.9936	0.9004
ORF1a: V1056L	PTLVVNAANVYLKHG	9.01	-8.27	> 20	123.33	0.380	0.7320	DRB1*03:01	0.9658	0.7823
								DRB1*12:01	0.9877	0.8329
								DRB1*12:02	0.9859	0.8318
								DRB1*14:01	0.9826	0.8861
								DRB1*14:02	0.9824	0.7588
								DRB1*14:08	0.9529	0.8542
								DRB1*15:01	0.9838	0.8849
								DRB1*15:02	0.9812	0.8855
								DRB1*15:03	0.9809	0.8883
								DRB1*16:02	0.9907	0.8807

Table 3

Final mutated B cell epitopes for vaccine construct
Mutation	Epitopes	ABCpred Score	Theoretical pI	Instability index	Half-life(h)	Aliphatic index	GRAVY	Antigenicity score
S: Q14K	VFLVLLPLVSSKCVNL	0.74	8.19	100	28.74	194.38	1.906	0.9929
S: Q52H	HSTHDLFLPFFSNVTW	0.71	5.97	3.5	18.75	66.88	0.081	0.5485
S: H245N	TRFQTLLALNRSYLTP	0.84	10.83	7.2	36.84	103.75	-0.125	0.5768
S: S256L	PGDSSLGWTAGAAAYY	0.69	3.8	> 20	34.17	49.38	-0.069	0.9183
S: A264D	GWTAGAADYYVGYLQP	0.96	3.8	30	26.59	61.25	-0.119	0.7499
S: G446S	CVIAWNSNNLDSKVSG	0.72	5.83	1.2	19.11	91.25	-0.037	0.5441
S: L452R	NYNYRYRLFRKSNLKP	0.81	10.55	1.4	-26.49	48.75	-1.731	0.6178
S: P1143L	VIGIVNNTVYDPLQLE	0.77	3.67	100	23.92	151.87	0.481	0.8852
ORF1a: N2526S	HFVNLDSLRANNTKGS	0.71	8.75	3.5	16.36	73.12	-0.744	0.6026

Population coverage analysis

World population coverage calculation utilized the IEDB population tool, incorporating HLA Ⅰ and HLA Ⅱ epitopes containing crucial mutations. As shown in Fig. 2, based on HLA-restricted alleles, the predicted population coverage impressively reached 95.95%, highlighting the efficacy of the selected epitopes and indicating their suitability for a vast majority of the global population.

Vaccine construction and physiochemical property prediction

The constructed vaccines were formulated with 50s Ribosomal L7/12 adjuvant (Vaccine 1) and Human Beta-defensin 3 adjuvant (Vaccine 2). Further characterization of the developed vaccines included assessments of antigenicity, allergenicity, toxicity, physicochemical properties, and solubility. The results presented in Table 4 revealed that the molecular weights of these two vaccines ranged between 52–60 kDa, which falls within the optimal range for antigenicity. Additionally, they exhibited robust antigenicity, non-allergenicity, and non-toxicity. In mammalian models, both vaccines determined a half-life of 30 hours and stability indices below 40, along with favorable hydrophilicity. Furthermore, both vaccines exhibited good solubility, with solubility scores exceeding the threshold of 0.45 and reaching 0.501.

Table 4

Characteristics of vaccines
Vaccine	Number of amino acids	Molecular weight	Theoretical pI	Half-life(h)	Instability index	Aliphatic index	GRAVY	Solubility	Antigenicity score
VC1	547	59248.92	9.86	30	23.63	89.21	-0.055	0.501	0.5762
VC2	471	52114.92	10.20	30	23.92	82.63	-0.207	0.501	0.6007

Prediction of secondary and tertiary structure of multi-epitope vaccine

The secondary structure of the multi-epitope vaccine was predicted using the PSIPRED server. The analysis revealed that vaccine 1 contained 38.21% alpha-helix (209/547), 16.82% extended strand (92/547), and 44.97% random coil (246/547). For vaccine 2, 154 amino acids (32.70%) were involved in alpha-helix formation, 95 (20.17%) in extended strand, and 222 (47.13%) in random coil (Fig S1).

The tertiary structure prediction of the vaccine was initially constructed using the I-TASSER server, which provides a C-score ranges from − 5 to 2, with higher values indicating greater confidence in the predicted models. Subsequently, refinement processes were carried out to enhance the structure fidelity, aiming to closely resemble the natural protein structure. This refinement was conducted using GalaxyRefine tool, which generated five refined structures for each vaccine. Among these, the most optimal structures were meticulously selected for further validation.

For vaccine 1 (Fig. 3a), the highest c-score was − 1.45. Refinement selected model three, which had a Global Distance Test-High Accuracy (GDT-HA) of 0.9278. The Root Mean Square Deviation (RMSD) value, a measure of the deviation, was determined to be 0.476. Ramachandran plot revealed that 81.1% of vaccine 1 residues were in the most favored region, with 14.1% in the additional allowed region, 1.5% in the generously allowed region, and 3.3% in the disallowed region(Fig. 3b). The computed Z-score for vaccine 1 was − 3.86 using the ProSA tool (Fig. 3c), which assesses the overall model quality relative to structures from various sources, such as X-ray and Nuclear Magnetic Resonance (NMR) .

For vaccine 2 (Fig. 3d), which displayed the highest c-score of -2.34 post I-TASSER prediction, refinement yielded model two with a GDT-HA of 0.9034. The RMSD value for vaccine 2 was calculated as 0.528. Analysis of the Ramachandran plot indicated that 77.3% of vaccine 2 residues were in most favored regions, with 19.3% in the additional allowed regions, 1.0% in the generously allowed regions, and 2.3% in the disallowed regions (Fig. 3e). The Z-score for vaccine 2 was determined as -4.47 (Fig. 3f).

Prediction of B-cell conformational epitopes

ElliPro tool predicted four conformational B-cell epitopes in vaccine 1, with sizes ranging from 28 to 112 residues and score values between 0.643 to 0.755. For vaccine 2, eight conformational B-cell epitopes were identified, spanning from 3 to 89 residues, with scores varying from 0.533 to 0.869. Detailed amino acid sequences and 3D structural representations of these epitopes were provided in Supplementary Table S2.

Disulfide engineering of multi-epitope vaccine

The Disulfide by Design 2.13 server indentified three potential disulfide bonds in vaccine 1 and two in vaccine 2 (Table S3). Among these, only the LYS96-ASP103 pair in vaccine 1, located on the adjuvant, exhibited favorable χ³ angle (+ 90.45°) and energy value (1.67kcal/mol), indicating potential for stability enhancement. The remaining four predicted bonds were situated within the antigenic epitopes, which may potentially compromise vaccine efficacy. Despite mutating the LYS96-ASP103 pair to cysteine, no significant improvement in overall vaccine stability was observed (Table S4). Therefore, considering both stability and antigenicity, the introduction of disulfide bonds was deemed unnecessary for the multi-epitopes vaccine.

Molecular docking analysis

The protein-protein docking panel of Schrödinger software (LLC, NY, USA, 2023-1) utilized the Piper algorithm to generate 30 docking models. These models were defined based on the centers of high-density clusters of low-energy docking structures. Optimal conformations for each docking complex were selected on cluster size. The details of the top-ranked vaccine and immune receptors are presented in Table 5. Analysis of the results reveals that vaccine1 and TLR3 exhibited relatively high cluster size, while vaccine2 and TLR4 displayed the highest cluster size.

Table 5

Docking results of vaccine 1 and vaccine 2 and with immune receptors
Vaccine	TLR	PIPER pose energy	PIPER pose score	PIPER cluster size
Vaccine 1	TLR 3	-1271.453	-78.994	60
	TLR 4	-1287.680	-3.374	46
	TLR 5	-1222.344	-369.810	57
	TLR 7	-1175.423	-509.188	45
Vaccine 2	TLR 3	-1244.063	-257.425	61
	TLR 4	-1433.962	54.049	89
	TLR 5	-1560.956	183.294	35
	TLR 7	-1141.933	-827.019	67

Molecular dynamics simulation

During the 100ns simulation, vaccine 1 consistently interacted with TLR3, TLR4, TLR5 and TLR7. RMSD was employed as a critical metric to evaluate the equilibrium status of these complexes. The trajectory analysis revealed that the TLR3 (Fig. 4a) and TLR5 (Fig. 4c) complexes stabilized at approximately 80ns and 40ns, respectively. In contrast, the TLR7 (Fig. 4d) complex exhibited notable fluctuations in the latter half of the simulation. To assess the stability of indibidual amino acid aross the entire simulation, RMSF analysis was performoed for vaccine 1 in complex with four TLRs. Among these, TLR7 exhibited the highest average RMSF (2.03 Å), followed by TLR5 (1.99 Å), TLR4 (1.69 Å), and TLR3 (1.21 Å) (Fig. 4e). Additionally, RMSF analysis revealed that the TLR-peptide complexes exhibited varying degrees of fluctuation, with the TLR7 complex showing the highest average RMSF (3.94 Å) and the TLR3 complex the lowest (2.89 Å) (Fig. 4f).

For vaccine 2, the analysis demonstrated that TLR3 (Fig. 5a), TLR5 (Fig. 5c) and TLR7 (Fig. 5d) complexes maintained a stable state at the end of the simulation, while TLR4 (Fig. 5b) complex continued to exhibit significant fluctuations. TLR5 had the highest average RMSF (2.18 Å), followed by TLR7 (1.72 Å), TLR4 (1.61 Å), and TLR3 (1.07 Å) (Fig. 5e). Among the vaccine 2 interactions, the TLR7-peptide complex showed the greatest fluctuations (5.65Å), compared to the TLR4-peptide (4.56 Å), TLR5-peptide (3.04 Å), and TLR3-peptide (2.93 Å) complexes (Fig. 5f).

Additionally, the radius of gyration (Rg) was calculated to assess the compactness of the complexes. Vaccine 1 exhibited a range of Rg values, with the TLR3 complex being the most compact (Mean Rg: 26.94 Å, Min: 26.18 Å, Max: 27.48 Å) and the TLR7 complex being the least compact (Mean Rg: 27.92 Å, Min: 26.05 Å, Max: 29.35 Å) (Fig. 6a). In contrast, vaccine 2 demonstrated generally more compact structures across all TLR complexes.The TLR3 complex again showed the highest compactness (Mean Rg: 25.14 Å, Min: 24.37 Å, Max: 25.69 Å), while the TLR7 complex had the highest Rg within Vaccine 2's interactions (Mean Rg: 25.59 Å, Min: 24.58 Å, Max: 26.32 Å) (Fig. 6b). However, the overall Rg values for vaccine 2 were lower than those for vaccine 1, indicating that vaccine 2 forms more tightly packed complexes with the TLRs.

Further analysis of the binding characteristics of protein-protein complexes between the multi-epitope vaccines and TLR receptors was conducted using the PRODIGY (Protein binding energy prediction) tool. The analysis examined the interaction types (ICs), near-interface sequence (NIS), residue properties, ΔG, and Kd values of the protein complexes. The ICs types included charged/charged, charged/polar, charged/non-polar, polar/polar, non-polar/polar, and non-polar/non-polar interactions.

Based on the data in Table 6, it can be inferred that both vaccine1 and vaccine 2 exhibit the smallest ΔG and Kd values in the interaction with TLR5. It suggests that the protein complexes formed between TLR5 and the designed multi-epitopes vaccines are more stable, indicating a higher propensity for interaction between TLR5 and these vaccines. Additionally, TLR3 also demonstrates favorable binding capability, with ΔG values of -20.2 kcal/mol and − 17.4 kcal/mol, and Kd values of 1.6e-15 and 1.8e-13 when interacting with vaccine1 and vaccine 2, respectively. In contrast, the binding affinities of TLR4 and TLR7 to these vaccines is lower compared to TLR3 and TLR5.

Table 6

Interaction between constructed vaccines and immune receptors
Properties	Vaccine 1				Vaccine 2
Properties	TLR3	TLR 4	TLR 5	TLR 7	TLR 3	TLR 4	TLR 5	TLR 7
ΔG (kcal mol-1)	-20.2	-11.6	-23.5	-14.3	-17.4	-12.0	-26.0	-13.6
Kd (M) at 25 ℃	1.6e-15	3.3e-09	6.1e-18	3.4e-11	1.8e-13	1.7e-09	8.3e-20	1e-10
ICs-charged/charged	32	8	25	10	24	17	33	4
ICs-charged/polar	34	15	32	14	24	15	26	16
ICs-charged/apolar	57	19	53	29	36	22	61	33
ICs-polar/polar	4	3	5	2	1	3	5	6
ICs-polar/apolar	33	21	55	26	28	15	57	25
ICs-apolar/apolar	46	12	63	39	41	10	48	69
NIS-charged	23.24	22.67	21.88	23.8	23.37	21.82	23.73	23.82
NIS-apolar	42.75	43.26	46.07	41.39	39.47	40.68	44.08	38.6

Immune simulation

Using the C-Immsim software, the impact of triple injections of multi-epitope vaccines on the human immune system was evaluated. Both vaccines elicited sufficient innate and adaptive immune responses. Upon stimulation with vaccine 1, B lymphocytes primarily engaged in antigen presentation functions during the initial phase. Presenting-2 B lymphocytes reached a peak of 650 cells/mm³ within a short period, followed by a significant proliferation of active B lymphocytes, reaching a peak of 700 cells/mm³ after the third stimulation (Fig. 7a). Additionally, the numbers of active and resting CD4 T-helper cells reached peaks of 8000 cells/mm³ and 3500 cells/mm³, respectively, after the third stimulation (Fig. 7b). Following primary immunization, the number of CD8 T-cytotoxic cells in the immune system increased rapidly, reaching a peak around 50th day. However, after pathogen clearance, the body needs to maintain immune system balance and prevent excessive immune responses, resulting in a decrease in the number of duplicating CD8 T-cytotoxic cells during this period (Fig. 7c). Furthermore, immunoglobulin levels gradually increased, including IgG, IgG1, IgG2, and IgM levels. The elevation of these antibodies may enhance the ability to clear pathogens and improve protective immune responses to subsequent infections. During and after the second and third vaccine administrations, antigen concentrations decreased (Fig. 7d). Notably, vaccine1 induced high levels of cytokines such as INF-g, TGF-b, IL-10, IL-18, and IL-2 (Fig. 7e). These cytokines play crucial roles in regulating immune responses and maintaining immune system balance. Additionally, NK cells were observed to be in an active state (Fig. 7f).

For vaccine 2, the peak of Presenting-2 B lymphocytes after the third inoculation was marginally higher than that observed with vaccine1 (Fig. 8a). Additionally, as depicted in Fig. 8b, the count of active CD4 T-helper cells surpassed that of vaccine 1 after the third stimulation, albeit slightly. Analogous to vaccine1, the peak of CD8 T-cytotoxic cells occurred around day 50 post-vaccine 2 stimulation (Fig. 8c). Concurrently, there was a gradual rise in immunoglobulin levels. However, it was noteworthy that after the third stimulation, the level of IgM (350,000 cells/mm3) was slightly lower than that observed with vaccine 1 (400,000 cells/mm3) (Fig. 8d). Among the cytokines showing elevated levels, IL-2 reached its 650,000 cells/mm3 and 500,000 cells/mm3 after the second and third vaccinations, respectively (Fig. 8e). Furthermore, NK cells exhibited heightened activity (Fig. 8f).

Codon adaptation and Insilco-cloning

The reverse-transcribed codon DNA sequences for Vaccine1 and Vaccine2 measure 1641 base pair and 1413 base pair in length, respectively. Both sequences exhibit a perfect Codon Adaptation Index (CAI) of 1.0, coupled with GC content of 57%, indicating their optimal adaptation for efficient expression in the host system. Subsequently, restriction sites of XhoⅠ (158) and SalI (179) were added to the N and C terminals of each vaccine sequence. Using the SnapGene tool, there modified sequences were computationally cloned into the pET28a (+) vector (Fig. 9). The final genome size after insertion of vaccine 1 and vaccine 2 were 6998bp and 6770 bp, respectively.

Structural analysis of hotspot mutation L455S/F456L

Out of the initial 506 antibodies obtained, 482 models were excluded, resulting in 24 antibodies for further analysis. These 24 antibodies corresponded to 62 PDB structures. Specifically, mutations in residues L455S and F456L were analyzed in the RBD region of the spike protein, focusing on the fab portion of 46 selected structures. Detailed residue scanning results can be found in Supplementary Table S5. As indicated in the table, using a threshold of 3 kcal/mol of Δ Affinity value, 15 PDB structure exhibited sustained neutralizing capabilities. Inspection of these 15 PDB structure in PyMOL revealed that the L455S and F456L mutations did not directly engage with the antibodies. Notably, 7 structures exhibited increased polar interactions within the RBD region following mutation. The residue scanning results for these 7 structures are detailed in Table 7. Specifically, in the structures 7WRJ, 7Y0W, 7YAD and 7YR0, the double mutations L455S and F456L resulted in 2 interactions between S455 and P491, marking an increase of one interaction compared to the pre-mutation state. Similarly, in structures 7TLY, 7R6W, and 7XCK, the number of polar interactions increased by one post-mutation, attributed to the mutated S455. Prior to mutation, 7TLY had 2 polar interactions, while 7R6W and 7XCK each had 3. The schematic diagrams of these three cases are illustrated in Fig. 10.

Table 7

Residue scanning results for 7 structures exhibiting increased polar interactions within the RBD region post-mutation
PDB ID	Antibody Fab	Mutation	Δ Affinity (kcal/mol)	Δ Stability (kcal/mol)
7WRJ	BD55-4637	L455S	-0.18	14.07
		F456L	-0.06	3.62
		L455S + F456L	-0.20	15.42
7Y0W	BD55-5514/BD55-5840	L455S	-0.07	13.61
		F456L	-0.05	6.25
		L455S + F456L	-0.10	12.39
7R6W	S2X35/S309	L455S	-0.16	17.02
		F456L	-0.03	8.40
		L455S + F456L	-0.19	14.41
7TLY	S309	L455S	0	13.64
		F456L	0	7.15
		L455S + F456L	-0.01	15.32
7XCK	S309	L455S	0	18.16
		F456L	0	6.89
		L455S + F456L	0	17.08
7YAD	S309	L455S	0	12.14
		F456L	0	3.55
		L455S + F456L	0	11.92
7YR0	S309	L455S	-0.01	11.04
		F456L	0	2.87
		L455S + F456L	-0.01	6.43

The SARS-CoV-2 virus, responsible for the COVID-19 pandemic, exhibits a high mutation rate, resulting in numerous variants [53]. Some of these variants confer resistance to existing vaccines and antiviral drugs, complicating efforts to control the virus. The ongoing evolution of SARS-CoV-2 variants poses significant challenges to global health, societal stability, and economic recovery [54], underscoring the urgent need for continuous surveillance and characterization of these variants to maintain effective pandemic control measures, including vaccination strategies [55]. Immunoinformatics plays a pivotal role in this process, particularly in epitope mapping, aiding the identification of immune processes and developing peptide-based vaccines [56]. Although mRNA-based vaccines have proven highly effective, their production and distribution costs limit accessibility in low- and middle-income countries. In contrast, peptide-based vaccines, which trigger immune responses by incorporating highly antigenic peptide fragments, offer advantages such as high stability and lower development costs [57]. This approach has been widely applied in studies of other pathogens, including Ebola virus [58], Influenza virus [59], Human papillomavirus (HPV) [60], Monkeypox virus (MPXV) [[61],[62]], and Hepatitis C virus (HCV) [63]. Most peptide vaccines targeting SARS-CoV-2 focused primarily on the spike protein[[64],[65]]. However, our study adopts a different approach. we have designed multi-epitope vaccines by targeting hotspot mutations in prevalent SARS-CoV-2 variants, leveraging structural biology and immunoinformatics. We analyzed 15 noteworthy variants and identified 55 related mutations, ultimately selecting 24 epitopes for vaccine construction. These selected epitopes not only exhibit superior immunogenic and physicochemical properties but are also predicted to be non-self peptides and to avoid triggering an autoimmune response. A comprehensive search of the IEDB database revealed limited experimental validation directly related to the specific mutated epitopes of interest. Most of the available data is not mutation-specific, complicating direct comparisions with our designed epitopes. It is precisely for this reason that we have addressed a critical gap by providing predictive insights into the epitopes of these recent mutations, potentially preempting immune escape mechanisms in the evolving virus.

Building on these insights, we assessed the global population coverage of the selected T-cell epitopes, which was found to be 95.95%. These epitopes were linked using AAY, GGPPG, KK, and EAAAK linkers. To enhance the immunogenicity of the vaccine, we employed two different adjuvants: 50s Ribosomal L7/12 and Human Beta-defensin 3, while maintaining the sama 24 epitopes in both vaccines. The 50s Ribosomal L7/12, an adjuvant derived from bacterial ribosomal proteins, significantly enhances T-cell and B-cell immune responses by activating dendritic cells (DCs) and macrophages, thereby promoting antigen presentation and strengthening adaptive immune responses. Beta-defensin is an antimicrobial peptide and a chemoattractant that facilitates interactions between the innate and adaptive immune systems [66]. It is known to induce innate immune responses, promote cytokine release, and activate specific immune responses, exhibiting immunomodulatory properties [67]

The pI values of vaccine 1 and vaccine 2 were calculated as 9.86 and 10.2, respectively. While elevated pI values are generally associated with protein instability under extreme pH conditions [68], our theoretical predictions suggest that these constructs do not inherently indicate instability under physiological conditions. Both constructs demonstrate favorable stability characteristics, with instability index of 23.63 for vaccine 1 and 23.92 for vaccine 2. Additionally, the protein-sol score for both constructs were 0.501, indicating that these vaccine structures are likely to maintain their structural integrity and functionally in a physiological environment. The molecular weights of the constructs are less than 110kDa, further enhance the feasibility of efficient purification, which is advantageous for large-scale production. Furthermore, the negative GRAVY indices that the constructs are hydrophilic, suggesting good interaction with water molecules, which is beneficial for purification and stability. Secondary structure predictions reveal substantial alpha-helices and beta-sheets content, indicative of regions capable of eliciting strong antibody responses. The 3D structures of the optimized multi-epitope vaccines were modeled and refined to ensure high-quality predictions. Ramachandran plots revealed that 96.7% and 97.7% of residues in vaccine 1 and vaccine 2, respectively, were located in allowed regions, indicating proper folding and structural integrity. The Z-scores of -3.86 and − 4.47 further support the structural stability of the models. Additionally, B-cell conformational epitope analysis identified four and eight discontinuous epitopes in vaccine 1 and vaccine 2, respectively, highlighting their potential to elicit a potent humoral immune response. These findings underscore the suitability of these constructs for further vaccine development.In vaccine development, TLR-mediated immune responses play a vital role in inducing both innate and adaptive immunity [69]. Molecular docking and dynamics simulations of the vaccine with TCR receptors indicated that our predicted vaccine showed stronger binding affinity to TLR5 and TLR3 receptors compared to TLR4 and TLR7. Similarly, during a 100-nanosecond simulation, TLR5 and TLR3 demonstrated relative stability. Furthermore, immunological simulation analyses revealed that both vaccines could induce B-cell and T-cell immune responses.

To ensure successful expression of the vaccine in the host, we inserted these vaccines into the pET28a (+) vector, selecting restriction enzymes XhoⅠ and SalI, which have similar buffer environments. After insertion, the final lengths of vaccine1 and vaccine2 were 6998bp and 6770bp, respectively. Overall, this study provides a foundation for further research into the potential efficacy of computer-designed vaccines. This approach can reduce the time, cost, labor, and risk of failure in vaccine development. The multi-epitope vaccine design can potentially overcome immune escape mechanisms, although the accuracy of the predictions is limited and may result in false positives or negatives. Therefore, the predicted outcomes of this study require experimental and clinical validation to ensure the designed vaccine's protective efficacy against SARS-CoV-2. Future studies will focus on validating these findings through in vitro and in vivo assays, which will further refine and validate the vaccine design.

This study not only evaluates SARS-CoV-2 Variants from the perspective of cellular immunity by designing multi-epitope vaccines but also explores the impact of key mutations, L455S and F456L, on neutralizing antibodies from the humoral immunity standpoint. Among the 62 PDB structures of 24 antibodies studied, 46 underwent mutation analysis for the L455S and F456L regions in the RBD, revealing that only 15 PDB structures retained neutralizing capability. However, for the L455S and F456L mutations, we did not obtain epitopes that were antigenic, non-allergenic, non-toxic, immunogenic, and had favorable physicochemical properties. Interestingly, although our analysis found that the Δ Affinity value between the RBD region and antibodies in these 7 structures was less than 3 kcal/mol and the number of polar interactions increased, the L455S and F456L mutations did not occur in regions of direct interaction. This suggests that while the mutations may not directly interfere with the antibody binding sites, they could still affect the overall interaction dynamics. Therefore, the effectiveness of neutralizing antibodies against these mutations requires further investigation to comprehensively understand their potential impact on antibody efficacy.

This study utilized a novel immunoinformatics approach to design multi-epitope vaccines targeting critical mutations in SARS-CoV-2. The designed vaccines demonstrated promising immune response and strong affinity to immune receptors. Our research contributes to the ongoing effort to combat COVID-19 by identifying potential epitopes effective against various SARS-CoV-2 variants, supporting the strategy of developing multivalent vaccines that offer broader immunity. Our study also evaluated the effects of two key mutations, L455S and F456L, on 46 existing antibodies that are effective against XBB variants, and found that 15 antibody structures were able to maintain effectiveness. Additionally, our study emphasizes the crucial role of computational tools in epitope prediction and vaccine design, which can expedite the development process. By combining advanced immunoinformatics techniques, we can enhance our preparedness and response to future challenges posed by COVID-19.

Author information

Authors and Affiliations

Key Laboratory of DGHD, MOE, School of Life Science and Technology, Southeast University, Nanjing, 210018, China

Xueyin Mei, Wanrong Xie, Jian Li

Department of Bioinformatics, School of Biomedical Engineering and Informatics, Nanjing Medical University, Nanjing, 211166, China

Xue Lin

Jiangsu Provincial Center for Disease Control and Prevention, Nanjing, 210009, China

Liguo Zhu

Contributions

Xueyin Mei: Formal analysis, Writing-original draft, Writing-review & editing. Wanrong Xie: Validation, Investigation. Xue Lin: Methodology, Data curation. Liguo Zhu: Resources, Data curation. Jian Li: Supervision, Conceptualization.

Corresponding authors

Correspondence to Xue Lin, Liguo Zhu or Jian Li

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing Interest

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

Sun K, Bhiman JN, Tempia S, et al. SARS-CoV-2 correlates of protection from infection against variants of concern. Nat Med Published online July. 2024;26. 10.1038/s41591-024-03131-2.
Wang Q, Guo Y, Iketani S, et al. Antibody evasion by SARS-CoV-2 Omicron subvariants BA.2.12.1, BA.4 and BA.5. Nature. 2022;608(7923):603–8. 10.1038/s41586-022-05053-w.
Ito J, Suzuki R, Uriu K et al. Convergent evolution of SARS-CoV-2 Omicron subvariants leading to the emergence of BQ.1.1 variant. Nat Commun. 2023;14(1):2671. Published 2023 May 11. 10.1038/s41467-023-38188-z
Cao Y, Yisimayi A, Jian F, et al. BA.2.12.1, BA.4 and BA.5 escape antibodies elicited by Omicron infection. Nature. 2022;608(7923):593–602. 10.1038/s41586-022-04980-y.
Cao Y, Jian F, Wang J, et al. Imprinted SARS-CoV-2 humoral immunity induces convergent Omicron RBD evolution. Nature. 2023;614(7948):521–9. 10.1038/s41586-022-05644-7.
Chemaitelly H, Tang P, Hasan MR, et al. Waning of BNT162b2 Vaccine Protection against SARS-CoV-2 Infection in Qatar. N Engl J Med. 2021;385(24):e83. 10.1056/NEJMoa2114114.
Goldberg Y, Mandel M, Bar-On YM, et al. Waning Immunity after the BNT162b2 Vaccine in Israel. N Engl J Med. 2021;385(24):e85. 10.1056/NEJMoa2114228.
Andrews N, Stowe J, Kirsebom F, et al. Covid-19 Vaccine Effectiveness against the Omicron (B.1.1.529) Variant. N Engl J Med. 2022;386(16):1532–46. 10.1056/NEJMoa2119451.
Collie S, Nayager J, Bamford L, Bekker LG, Zylstra M, Gray G. Effectiveness and Durability of the BNT162b2 Vaccine against Omicron Sublineages in South Africa. N Engl J Med. 2022;387(14):1332–3. 10.1056/NEJMc2210093.
Rubin R, As. COVID-19 Cases Surge, Here's What to Know About JN.1, the Latest SARS-CoV-2 Variant of Interest. JAMA. 2024;331(5):382–3. 10.1001/jama.2023.27841.
Kaku Y, Yo MS, Tolentino JE, et al. Virological characteristics of the SARS-CoV-2 KP.3, LB.1, and KP.2.3 variants. Lancet Infect Dis. 2024;24(8):e482–3. 10.1016/S1473-3099(24)00415-8.
Arevalo-Romero JA, Chingaté-López SM, Camacho BA, Alméciga-Díaz CJ, Ramirez-Segura CA. Next-generation treatments: Immunotherapy and advanced therapies for COVID-19. Heliyon. 2024;10(5):e26423. 10.1016/j.heliyon.2024.e26423. Published 2024 Feb 19.
Ysrafil Y, Sapiun Z, Astuti I, et al. Designing multi-epitope based peptide vaccine candidates against SARS-CoV-2 using immunoinformatics approach. Bioimpacts. 2022;12(4):359–70. 10.34172/bi.2022.23769.
Barouch DH. Covid-19 Vaccines - Immunity, Variants, Boosters. N Engl J Med. 2022;387(11):1011–20. 10.1056/NEJMra2206573.
Jiesisibieke ZL, Liu WY, Yang YP, Chien CW, Tung TH. Effectiveness and Safety of COVID-19 Vaccinations: An Umbrella Meta-Analysis. Int J Public Health. 2023;68:1605526. 10.3389/ijph.2023.1605526. Published 2023 Jul 7.
Lotfi H, Mazar MG, Ei NMH, Fahim M, Yazdi NS. Vaccination is the most effective and best way to avoid the disease of COVID-19. Immun Inflamm Dis. 2023;11(8):e946. 10.1002/iid3.946.
Samaranayake LP, Seneviratne CJ, Fakhruddin KS. Coronavirus disease 2019 (COVID-19) vaccines: A concise review. Oral Dis. 2022;28(Suppl 2):2326–36. 10.1111/odi.13916.
Yin Q, Luo W, Mallajosyula V, et al. A TLR7-nanoparticle adjuvant promotes a broad immune response against heterologous strains of influenza and SARS-CoV-2. Nat Mater. 2023;22(3):380–90. 10.1038/s41563-022-01464-2.
Hartmeier PR, Ostrowski SM, Busch EE, Empey KM, Meng WS. Lymphatic distribution considerations for subunit vaccine design and development. Vaccine. 2024;42(10):2519–29. 10.1016/j.vaccine.2024.03.033.
Sun B, Yu S, Zhao D, Guo S, Wang X, Zhao K. Polysaccharides as vaccine adjuvants. Vaccine. 2018;36(35):5226–34. 10.1016/j.vaccine.2018.07.040.
Baljon JJ, Kwiatkowski AJ, Pagendarm HM, et al. A Cancer Nanovaccine for Co-Delivery of Peptide Neoantigens and Optimized Combinations of STING and TLR4 Agonists. ACS Nano. 2024;18(9):6845–62. 10.1021/acsnano.3c04471.
Luchner M, Reinke S, Milicic A. TLR Agonists as Vaccine Adjuvants Targeting Cancer and Infectious Diseases. Pharmaceutics. 2021;13(2):142. Published 2021 Jan 22. 10.3390/pharmaceutics13020142
Zhao Y, Kuang M, Li J, et al. SARS-CoV-2 spike protein interacts with and activates TLR41. Cell Res. 2021;31(7):818–20. 10.1038/s41422-021-00495-9.
Chakraborty C, Sharma AR, Bhattacharya M, Sharma G, Lee SS, Agoramoorthy G. Consider TLR5 for new therapeutic development against COVID-19. J Med Virol. 2020;92(11):2314–5. 10.1002/jmv.25997.
Birra D, Benucci M, Landolfi L, et al. COVID 19: a clue from innate immunity. Immunol Res. 2020;68(3):161–8. 10.1007/s12026-020-09137-5.
Farkas D, Bogamuwa S, Piper B, et al. A role for Toll-like receptor 3 in lung vascular remodeling associated with SARS-CoV-2 infection. Preprint bioRxiv. 2023. 10.1101/2023.01.25.524586. 2023.01.25.524586. Published 2023 Jan 25.
Yazdanpanah F, Hamblin MR, Rezaei N. The immune system and COVID-19: Friend or foe? Life Sci. 2020;256:117900. 10.1016/j.lfs.2020.117900.
Humayun F, Cai Y, Khan A, et al. Structure-guided design of multi-epitopes vaccine against variants of concern (VOCs) of SARS-CoV-2 and validation through In silico cloning and immune simulations. Comput Biol Med. 2022;140:105122. 10.1016/j.compbiomed.2021.105122.
Hessel SS, Dwivany FM, Zainuddin IM, et al. A computational simulation appraisal of banana lectin as a potential anti-SARS-CoV-2 candidate by targeting the receptor-binding domain. J Genet Eng Biotechnol. 2023;21(1):148. 10.1186/s43141-023-00569-8. Published 2023 Nov 28.
Hosseini SA, Zahedipour F, Mirzaei H, Kazemi Oskuee R, Potential. SARS-CoV-2 vaccines: Concept, progress, and challenges. Int Immunopharmacol. 2021;97:107622. 10.1016/j.intimp.2021.107622.
Mirzaei HR, Pourghadamyari H, Rahmati M, et al. Gene-knocked out chimeric antigen receptor (CAR) T cells: Tuning up for the next generation cancer immunotherapy. Cancer Lett. 2018;423:95–104. 10.1016/j.canlet.2018.03.010.
Kalkanlı Taş S, Kırkık D, Öztürk K, Tanoğlu A. Determination of B- and T- cell epitopes for Helicobacter pylori cagPAI: An in silico approach. Turk J Gastroenterol. 2020;31(10):713–20. 10.5152/tjg.2020.19154.
Doytchinova IA, Flower DR. VaxiJen: a server for prediction of protective antigens, tumour antigens and subunit vaccines. BMC Bioinformatics. 2007;8:4. 10.1186/1471-2105-8-4. Published 2007 Jan 5.
Dimitrov I, Bangov I, Flower DR, Doytchinova I. AllerTOP v.2–a server for in silico prediction of allergens. J Mol Model. 2014;20(6):2278. 10.1007/s00894-014-2278-5.
Camacho C, Coulouris G, Avagyan V, et al. BLAST+: architecture and applications. BMC Bioinformatics. 2009;10:421. 10.1186/1471-2105-10-421. Published 2009 Dec 15.
Bhattacharya S, Banerjee A, Ray S. Development of new vaccine target against SARS-CoV2 using envelope (E) protein: An evolutionary, molecular modeling and docking based study. Int J Biol Macromol. 2021;172:74–81. 10.1016/j.ijbiomac.2020.12.192.
Ud-Din M, Albutti A, Ullah A, et al. Vaccinomics to Design a Multi-Epitopes Vaccine for Acinetobacter baumannii. Int J Environ Res Public Health. 2022;19(9):5568. 10.3390/ijerph19095568. Published 2022 May 4.
Abraham Peele K, Srihansa T, Krupanidhi S, Ayyagari VS, Venkateswarulu TC. Design of multi-epitope vaccine candidate against SARS-CoV-2: a in-silico study. J Biomol Struct Dyn. 2021;39(10):3793–801. 10.1080/07391102.2020.1770127.
Sirohi PR, Gupta J, Somvanshi P, Prajapati VK, Grover A. Multiple epitope-based vaccine prediction against SARS-CoV-2 spike glycoprotein. J Biomol Struct Dyn. 2022;40(8):3347–58. 10.1080/07391102.2020.1846626.
Alshiekheid MA, Dou AM, Algahtani M, et al. Bioinformatics and immunoinformatics assisted multiepitope vaccine construct against Burkholderia anthina. Saudi Pharm J. 2024;32(1):101917. 10.1016/j.jsps.2023.101917.
McGuffin LJ, Bryson K, Jones DT. The PSIPRED protein structure prediction server. Bioinformatics. 2000;16(4):404–5. 10.1093/bioinformatics/16.4.404.
Yang Y, Gao J, Wang J, et al. Sixty-five years of the long march in protein secondary structure prediction: the final stretch? Brief Bioinform. 2018;19(3):482–94. 10.1093/bib/bbw129.
Yang J, Zhang Y. I-TASSER server: new development for protein structure and function predictions. Nucleic Acids Res. 2015;43(W1):W174–81. 10.1093/nar/gkv342.
Haste Andersen P, Nielsen M, Lund O. Prediction of residues in discontinuous B-cell epitopes using protein 3D structures. Protein Sci. 2006;15(11):2558–67. 10.1110/ps.062405906.
Ponomarenko J, Bui HH, Li W, et al. ElliPro: a new structure-based tool for the prediction of antibody epitopes. BMC Bioinformatics. 2008;9:514. 10.1186/1471-2105-9-514. Published 2008 Dec 2.
Dombkowski AA, Sultana KZ, Craig DB. Protein disulfide engineering. FEBS Lett. 2014;588(2):206–12. 10.1016/j.febslet.2013.11.024.
Craig DB, Dombkowski AA. Disulfide by Design 2.0: a web-based tool for disulfide engineering in proteins. BMC Bioinformatics. 2013;14:346. 10.1186/1471-2105-14-346. Published 2013 Dec 1.
Kozakov D, Hall DR, Xia B, et al. The ClusPro web server for protein-protein docking. Nat Protoc. 2017;12(2):255–78. 10.1038/nprot.2016.169.
Xue LC, Rodrigues JP, Kastritis PL, Bonvin AM, Vangone A. PRODIGY: a web server for predicting the binding affinity of protein-protein complexes. Bioinformatics. 2016;32(23):3676–8. 10.1093/bioinformatics/btw514.
Stothard P. The sequence manipulation suite: JavaScript programs for analyzing and formatting protein and DNA sequences. Biotechniques. 2000;28(6):1102–4. 10.2144/00286ir01.
Zaib S, Rana N, Areeba, et al. Designing multi-epitope monkeypox virus-specific vaccine using immunoinformatics approach. J Infect Public Health. 2023;16(1):107–16. 10.1016/j.jiph.2022.11.033.
Ahmad S, Demneh FM, Rehman B, et al. In silico design of a novel multi-epitope vaccine against HCV infection through immunoinformatics approaches. Int J Biol Macromol. 2024;267(Pt 2):131517. 10.1016/j.ijbiomac.2024.131517.
Frasson I, Diamante L, Zangrossi M, et al. Identification of druggable host dependency factors shared by multiple SARS-CoV-2 variants of concern. J Mol Cell Biol. 2024;16(3):mjae004. 10.1093/jmcb/mjae004.
Hattab D, Amer MFA, Al-Alami ZM, Bakhtiar A. SARS-CoV-2 journey: from alpha variant to omicron and its sub-variants [published correction appears in Infection. 2024 Apr 29. doi: 10.1007/s15010-024-02283-0]. Infection. 2024;52(3):767–786. 10.1007/s15010-024-02223-y
Li P, Liu Y, Faraone JN, et al. Distinct patterns of SARS-CoV-2 BA.2.87.1 and JN.1 variants in immune evasion, antigenicity, and cell-cell fusion. mBio. 2024;15(5):e0075124. 10.1128/mbio.00751-24.
Tomar N, De RK. Immunoinformatics: an integrated scenario. Immunology. 2010;131(2):153–68. 10.1111/j.1365-2567.2010.03330.x.
Banerjee A, Santra D, Maiti S, Energetics. and IC50 based epitope screening in SARS CoV-2 (COVID 19) spike protein by immunoinformatic analysis implicating for a suitable vaccine development. J Transl Med. 2020;18(1):281. Published 2020 Jul 10. 10.1186/s12967-020-02435-4
Alizadeh M, Amini-Khoei H, Tahmasebian S et al. Designing a novel multi–epitope vaccine against Ebola virus using reverse vaccinology approach. Sci Rep. 2022;12(1):7757. Published 2022 May 11. 10.1038/s41598-022-11851-z
Yuan L, Li X, Li M, et al. In silico design of a broad-spectrum multiepitope vaccine against influenza virus. Int J Biol Macromol. 2024;254(Pt 3):128071. 10.1016/j.ijbiomac.2023.128071.
Sanami S, Rafieian-Kopaei M, Dehkordi KA, et al. In silico design of a multi-epitope vaccine against HPV16/18. BMC Bioinformatics. 2022;23(1):311. 10.1186/s12859-022-04784-x. Published 2022 Aug 2.
Pritam M. Exploring the whole proteome of monkeypox virus to design B cell epitope-based oral vaccines using immunoinformatics approaches. Int J Biol Macromol. 2023;252:126498. 10.1016/j.ijbiomac.2023.126498.
Sanami S, Nazarian S, Ahmad S, et al. In silico design and immunoinformatics analysis of a universal multi-epitope vaccine against monkeypox virus. PLoS ONE. 2023;18(5):e0286224. 10.1371/journal.pone.0286224. Published 2023 May 23.
Ahmad S, Demneh FM, Rehman B, et al. In silico design of a novel multi-epitope vaccine against HCV infection through immunoinformatics approaches. Int J Biol Macromol. 2024;267(Pt 2):131517. 10.1016/j.ijbiomac.2024.131517.
Arshad SF, Rehana R, Saleem MA, et al. Multi-epitopes vaccine design for surface glycoprotein against SARS-CoV-2 using immunoinformatic approach. Heliyon. 2024;10(2):e24186. 10.1016/j.heliyon.2024.e24186. Published 2024.
Sanami S, Zandi M, Pourhossein B, et al. Design of a multi-epitope vaccine against SARS-CoV-2 using immunoinformatics approach. Int J Biol Macromol. 2020;164:871–83. 10.1016/j.ijbiomac.2020.07.117.
Dhanushkumar T, Selvam PK. Rational design of a multivalent vaccine targeting arthropod-borne viruses using reverse vaccinology strategies. Int J Biol Macromol. 2024;258(Pt 1):128753. 10.1016/j.ijbiomac.2023.128753.
Dhople V, Krukemeyer A, Ramamoorthy A. The human beta-defensin-3, an antibacterial peptide with multiple biological functions. Biochim Biophys Acta. 2006;1758(9):1499–512. 10.1016/j.bbamem.2006.07.007.
Pelegrine D, Gasparetto C. Whey proteins solubility as function of temperature and pH. LWT—Food Sci Technol. 2005;38(1):77–80. 10.1016/j.lwt.2004.03.013.
Fitzgerald KA, Kagan JC. Toll-like Receptors and the Control of Immunity. Cell. 2020;180(6):1044–66. 10.1016/j.cell.2020.02.041.

Supplementarydata.docx
Supplementary Materials Supplementary table is available in the online version this article. Supplementary Material 1 Supplementary Material 2
SupplementarydataTableS5.xlsx

Download PDF

Reviewers agreed at journal
14 Oct, 2024
Reviewers invited by journal
11 Oct, 2024
Editor assigned by journal
09 Oct, 2024
First submitted to journal
03 Oct, 2024

You are reading this latest preprint version

Immunoinformatics-Based Design of Broad-Spectrum Multi-Epitope Vaccines Targeting Mutations in Emerging SARS-CoV-2 Variants

Status:

Version 1

Abstract

Figures

Introduction

Materials and methodology

Acquisition and analysis of new mutations in target variants

Retrieval of target proteins sequence

Epitopes prediction

Epitopes processing

Population coverage

Designing of multi-epitope vaccine

Predicting and verifying of multi-epitope vaccine

Prediction of B-cell conformational epitopes

Disulfide engineering of multi-epitope vaccine

Molecular docking analysis of multi-epitope vaccine

Molecular dynamics simulation

Immune simulation

Codon Adaptation and In Silico Cloning

Acquisition of neutralizing antibody structures and analysis of hotspot mutations

Results

Identification of new mutations in target variants

Identification of epitopes

Population coverage analysis

Vaccine construction and physiochemical property prediction

Prediction of secondary and tertiary structure of multi-epitope vaccine

Prediction of B-cell conformational epitopes

Disulfide engineering of multi-epitope vaccine

Molecular docking analysis

Molecular dynamics simulation

Immune simulation

Codon adaptation and Insilco-cloning

Structural analysis of hotspot mutation L455S/F456L

Discussion

Conclusions

Declarations

References

Supplementary Files

Status:

Version 1