Immunoinformatics-Guided Design of a Multi-Epitope Vaccine Targeting WISP1 for Gastric Cancer

doi:10.21203/rs.3.rs-8247346/v1

Download PDF

Research Article

Immunoinformatics-Guided Design of a Multi-Epitope Vaccine Targeting WISP1 for Gastric Cancer

https://doi.org/10.21203/rs.3.rs-8247346/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Gastric cancer remains a leading cause of cancer-related mortality worldwide, and effective preventive or therapeutic vaccines are still lacking. WNT1‑inducible signalling pathway protein 1 (WISP1/CCN4) is a secreted matricellular protein that is overexpressed in gastric tumours and associated with poor prognosis, making it a promising immunotherapy target. We aimed to design a multi‑epitope peptide vaccine candidate targeting WISP1 using an immunoinformatics workflow.

Results

We predicted linear B‑cell epitopes from WISP1 and filtered them for antigenicity, non‑allergenicity and non‑toxicity. High‑scoring MHC class I and class II T‑cell epitopes with broad HLA coverage were then selected. The final construct combined validated B‑ and T‑cell epitopes with appropriate linkers and a TLR‑agonist adjuvant to enhance immunogenicity. Physicochemical profiling indicated that the construct is stable, soluble, hydrophilic and antigenic, with no predicted allergenicity or toxicity. Secondary and tertiary structures were modelled, refined and validated, revealing proper folding and favourable stereochemical quality. Molecular docking showed strong binding to innate immune receptors, particularly TLR4, and molecular dynamics simulations confirmed stable receptor–vaccine interactions with low structural deviation. Binding‑free energy analysis further supported these results. Immune simulations predicted robust primary and secondary immune responses characterised by sustained IgG/IgM production, increased IFN‑γ and IL‑2, and activation of memory B and T cells. Codon optimisation and in‑silico cloning suggested feasibility for experimental expression.

Conclusions

These results indicate that the proposed WISP1 multi‑epitope vaccine is antigenic, safe, structurally stable and capable of eliciting broad humoral and cellular immune responses in silico. This work provides a testable candidate for experimental validation and highlights secreted CCN family proteins as novel targets for gastric cancer vaccines.

Gastric Cancer

Immunoinformatics

Molecular Docking

Multi-Epitope Vaccine

Peptide Vaccine

WISP1

Gastric (stomach) cancer (GC) is among the deadliest tumors in the world, ranking as the fourth most frequent major driver of cancer-linked fatality, and the fifth most prevalent solid tumor (1). In spite of the remarkable progress obtained in chemotherapy, target therapies, and surgery over the past decades, late-stage GC patient survival remains poor, and overall survival at five years is less than 35% (2). As such, these problems underscore the need for new therapies that will support traditional modalities and yield sustainable clinical outcomes. Immunotherapy has been promising among the new alternatives and has the ability to achieve targeted and long-lasting anti-tumor activities through the mobilizing the host immune machinery (3).

Within the field of cancer immunotherapy, cancer vaccines against cancer have been the source of enthusiasm as a technique of stimulating cellular and humoral immunity against tumor-associated antigens (TAAs). Different TAAs, including MUC1, HER2/neu, and human telomerase reverse transcriptase (hTERT), were examined as vaccine targets in spite of being of an endogenous origin, due to their aberrantly increased expression and tumor-selective activity (4–6). These examples represent that a protein can still be useful to be used as an effective immunotherapeutic target, regardless of its origin, by revealing tumor-selective expression patterns.

WNT1-inducible signaling pathway protein 1, WISP1/CCN4, has recently been identified as an attractive candidate for GC immunotherapy. Although WISP1 is an endogenous matricellular protein involved in development and tissue repair, accumulating evidence shows that it is markedly overexpressed in GC samples relative to non-cancerous mucosa and shows a strong link to unfavorable outcomes, epithelial–mesenchymal transition (EMT), tumor microenvironment remodeling, and drug resistance. Significantly, since WISP1 is a secreted protein that is under the scrutiny of the immune system and distinguishable from basal levels of normal physiologic expression, it has the profile of a TAA and a compelling rationale for incorporation into a vaccine strategy (7–10).

Therapeutic strategies against WISP1 so far have focused primarily on monoclonal antibodies that temporarily neutralize its signaling function. While antibody therapy is promising in itself, it only provides passive immunity and must be repeated frequently. In contrast, designing a multi-epitope peptide vaccine offers potential for inducing a more long-lasting anti-tumor response by inducing long-term immunological memory against tumor cells that abnormally express WISP1. Moreover, with the application of immunoinformatics-driven epitope prediction and cross-reactivity assessment using BLAST, the incorporation of epitopes with low homology with normal host proteins is selectively permitted and so, the risk of autoimmunity is maintained low (11, 12).

While highly correlated with the onset of GC, no vaccine design approaches have been used against WISP1 yet. It is a basic gap of knowledge and an opportunity to attempt new, bioinformatics-aided vaccine construction against this TAA.

We have carefully applied systems biology, gene network models, and immunoinformatics methods in our laboratory to uncover the molecular pathways and find therapeutic targets in many cancers and diseases. We identified oncogenes and oncogenic pathways in hepatocellular carcinoma (13, 14), found the controlling function of microRNAs in Wnt/β-catenin pathway and cancer growth (15, 16), and reviewed drug resistance mechanisms in stomach and esophagus tumors through expression and network analyses (17–24). Furthermore, we have explored how artificial intelligence applications can be leveraged in pharmaceutical sciences (25) and advanced genome-level studies in cancer biology (26). Collectively, these prior works provide the methodological foundation and scientific justification for the present study, which extends our scope of work to the computational development of a therapeutic vaccine aimed at WISP1 in GC.

In this study, we employed an immunoinformatics pipeline for the development of a multi-epitope vaccine candidate targeting WISP1, aiming to stimulate robust lymphocyte-mediated responses. The objectives were: (i) to predict and confirm immunogenic B-cell, MHC classes I and II epitopes from WISP1; (ii) to design a construct incorporating linkers and an adjuvant; (iii) to establish its physicochemical and structural and immunological properties; and (iv) to study its potential binding with the receptor of the immune system through molecular docking combined with dynamic simulation and the immune system. By leveraging computational tools, this work provides a rational foundation for vaccine development against WISP1 in GC. Accordingly, we adopt a chemistry-forward, hypothesis-driven framework that couples epitope and linker selection with physicochemical developability profiling and receptor-level biophysics (docking and molecular dynamics) to guide design decisions.

Protein Sequence Retrieval and Localization

As the initial step of computational development of a GC vaccine directed at WISP1, the UniProt database (ID: O95388, CCN4_HUMAN) was used for fetching the protein sequence of WISP1. UniProt and the TMHMM database, for predicting the localization of proteins by applying Hidden Markov Models for membranes, were used for studying the localization of proteins.

Prediction of B-cell Epitopes

B-cell epitopes were identified through the ABCpred server considering the protein sequence of WISP1 (UniProt ID: O95388. CCN4_HUMAN). The ABCpred service predicts from a learned recurrent neural network. We set the threshold value at 0.51, window size at 16, and overlapping filter on. The server was provided with the sequence, and it was able to report the B-cell epitopes with the probability value above the threshold value chosen.

Antigenicity, Allergenicity and Toxicity Assessment

The potential B-cell epitopes were evaluated using VaxiJen v2.0 (cutoff 0.5) for antigenicity, AllerTOP v2.0 for allergenic potential (27), and ToxinPred2 for toxicity (28). The non-antigenic and allergenic epitopes were excluded, and the potential vaccine construction candidates were those that were not toxic or allergen.

Identification and selection of top-scoring MHC-I and MHC-II epitopes

The epitopes of MHC-I and MHC-II of WISP1 (UniProt ID: O95388) were predicted using the IEDB server with the NetMHCpan 4.1 EL algorithm. Only those epitopes of rank < 1% were retained, and antigenicity was approximated by VaxiJen v2.0 (cut-off 0.5). The remaining epitopes were normalized and prioritized based on mean IEDB–VaxiJen scores. Potential allergenicity was estimated using AllerTOP v2.0, while toxicity was assessed with ToxinPred2. Epitopes predicted as safe (non-allergenic and non-toxic) were prioritized, and a final set of ten MHC-I and five MHC-II epitopes was retained by avoiding repetition by maintaining larger overlapped sequences.

Design of the Vaccine Framework

The chosen B-cell epitopes and MHC-II and MHC-I were introduced at the same time into a multi-epitope construct using assigned linkers (KK, GGGGS for B-cell; AAY for MHC-I; GPGPG for MHC-II; EAAAK for adjuvant). The thymosin α1 acted as an adjuvant for immune response boosting and a His-tag (HHHHHH) for simplification of purification.

Assessment of Antigenicity, Allergenicity, and Toxicity

The designed construct underwent evaluation of antigenicity through VaxiJen v2.0 (threshold 0.5), allergenic potential with AllerTOP v2.0, and toxicity prediction via ToxinPred2 based on sequence and physicochemical characteristics.

Physicochemical Properties of Candidate Vaccine

The 274-amino-acid vaccine construct was analyzed by using ProtParam (ExPASy) to determine molecular weight, half-life in various hosts, isoelectric point, aliphatic and instability indices, and GRAVY score to obtain details on hydrophobicity and stability.

Secondary and Tertiary Structure Prediction

Prediction of the secondary structure of the construct was done for GOR IV server. The probability parameters were computed from experimental evidence of known tertiary structures for building GOR. The GOR IV program took sequence into consideration in order to predict random coils, beta turns, extended strands (beta sheets) as well as alpha helices appearance.

The tertiary structure prediction was done by Robetta and AlphaFold, subsequently smoothed out with GalaxyRefine and analyzed against SAVES v6.0 (PROCHECK, ERRAT). Additional assessment was carried out using ProSA (Z-score) and MolProbity (score, clash score); subsequently, the highest quality model was chosen for subsequent analysis.

Docking with Immune Receptors

The TLR2 (PDB: 5d3i) and TLR4 (PDB: 7mlm) crystal structures were extracted, processed in UCSF Chimera, and key active-site residues were identified (Pro352, Phe322, Ile319, Leu324, Phe325, Phe349, Tyr326, Val348, and Leu317 for TLR2 (29), and Gln339, Lys263, Ser413, Arg380, Ser386, Lys341, and Arg434 for TLR4 (30)). Interactions with the vaccine construct were analyzed using ClusPro2 (31), visualized with LigPlot + v4.5.3, and binding affinities predicted by PRODIGY (32).

Simulation of Molecular Dynamics (MD)

The vaccine’s MD simulations alone and in complex with TLR2/TLR4 were performed in GROMACS 2022 using the OPLS force field (33). Systems were solvated (SPC water, 10 Å), neutralized with Na⁺/Cl⁻, equilibrated under NVT (300 K, 100 ps) and NPT (300 K, 1 bar, 100 ps), and run for 100 ns with 2 fs timesteps. Electrostatics were handled with PME (0.16 nm grid, 10 Å cutoff) and van der Waals at 1 nm. Stability was assessed via RMSD, RMSF, and Rg (34).

Binding-Free Energy (BFE) Calculation

The HawDock web server calculated the vaccine/TLR2 and vaccine/TLR4 complex's BFE. The MM-GBSA approach was utilized by this web server to calculate the interaction energy. The web server computed the binding energy by applying the following Eq. (35):

ΔG_bind = G_complex - (G_receptor + G_vaccine)

Immune Simulation

The vaccine's immunogenicity was studied using the C-ImmSim server. Three shots were administered at the time steps of 1, 84, and 168 for 1000 steps with default configurations, random seed 12345, and a volume of 10 µL.

Optimization of the Codon and Computational Cloning

The construct sequence was optimized for codons by JCat for E. coli (K12) expression with regard to GC content and CAI (36). Then, it was cloned into pET28a in SnapGene and double digestion was validated in silico by agarose simulation.

Protein Localization Analysis

UniProt classified WISP1 in the secreted protein class (Fig. 1a), which makes intuitive sense given its role in intercellular communication. Prediction by TMHMM reiterated an absence of transmembrane helices and predicted virtually all of the sequence to be localized extracellularly, in keeping with a probability of exterior location of 0.69226 (Fig. 1b). Overall, these findings indicate that WISP1 is secreted out into the tumor environment and is thus immunologically accessible and a likely candidate target for a vaccine.

Prediction of the B lymphocyte Epitopes

Several epitopes for B lymphocyte were identified in WISP1 protein (Table 1) by the ABCpred server within a score range of 0.55 to 0.95. Top-scoring epitopes were YQPKYCGVCMDNRCCI (0.95), TALSPAPTTMDFTPAP (0.93), and VGCVLDGVRYNNGQSF (0.93). Filtering antigenicity in VaxiJen v2.0 validated SPAPTTMDFTPAPL (0.5324) and FADLESYPDFSEIA (0.9919) as exceeding the threshold level of 0.5; however, PCECPPSPPRCPLG was omitted due to a negative score. There was no notable similarity found by cross-reactivity studies against the human proteome that would compromise tumor-selective specificity. These results demonstrate that WISP1 is a rich source of epitopes for B-cell that can be included in a multi-epitope vaccine to prevent GC.

Table 1
List of Predicted B-Cell Epitopes on WISP1
Ranking	Sequence	Starting amino acid	Score
1	YQPKYCGVCMDNRCCI	339	0.95
2	TALSPAPTTMDFTPAP	66	0.93
2	VGCVLDGVRYNNGQSF	164	0.93
2	VSLITDGCECCKMCAQ	109	0.93
3	HTLIKAGKKCLAVYQP	307	0.92
4	APLEDTSSRPQFCKWP	80	0.91
4	CKYNCTCIDGAVGCTP	183	0.91
5	CECPPSPPRCPLGVSL	96	0.90
5	AGCISTRSYQPKYCGV	331	0.90
5	HMSAPIENSXGNCCNP	25	0.90
5	RGLYCDYSGDRPRYAI	140	0.90
6	RPPRLWCPHPRRVSIP	204	0.88
7	VNAQCWPEQESRLCNL	284	0.87
7	HRNCIAYTSPWSPCST	256	0.87
7	HCCEQWVCEDDAKRPR	221	0.87
8	GKKCLAVYQPEASMNF	313	0.85
8	TSCGLGVSTRISNVNA	271	0.85
9	KTAPRDTGAFDAVGEV	237	0.83
9	EAAICDPHRGLYCDYS	132	0.83
10	RVSIPGHCCEQWVCED	215	0.82
11	CIPYKSKTIDVSFQCP	353	0.81
11	GEVEAWHRNCIAYTSP	250	0.81
12	VLWINACFCNLSCRNP	377	0.80
12	YTSPWSPCSTSCGLGV	262	0.80
13	EQESRLCNLRPCDVDI	291	0.79
13	DGAVGCTPLCLRVRPP	191	0.79
13	AMILYMEMERSHMSAP	14	0.79
14	AQQLGDNCTEAAICDP	123	0.78
15	GQSFQPNCKYNCTCID	176	0.77
16	SVMRWFLPWTLAAVTA	42	0.74
16	PNDIFADLESYPDFSE	392	0.74
16	ECCKMCAQQLGDNCTE	117	0.74
17	PTTMDFTPAPLEDTSS	72	0.73
17	NLSCRNPNDIFADLES	386	0.73
17	GVRYNNGQSFQPNCKY	170	0.73
18	AVTAAAASTVLATALS	54	0.70
18	PEASMNFTLAGCISTR	322	0.70
19	DGLGFSRQVLWINACF	369	0.69
20	ASTVLATALSPAPTTM	60	0.68
20	DVSFQCPDGLGFSRQV	362	0.68
20	GNCCNPESVMRWFLPW	35	0.68
21	PCCNHMANCCNFAMIL	2	0.55

Assessment of Antigenicity, Allergenicity, and Toxicity

We used VaxiJen, AllerTOP, and ToxinPred2 to check the predicted B-cell epitopes for both immunogenicity and safety (Table 2). Some epitopes have very significant antigenicity but were left out because they were thought to be poisonous (for example, YQPKYCGVCMDNRCCI, APLEDTSSRPQFCKWP) or allergic (for example, TSCGLGVSTRISNVNA, KTAPRDTGAFDAVGEV). On the other hand, the epitopes colored in green in Table 2, TALSPAPTTMDFTPAP, VGCVLDGVRYNNGQSF, AGCISTRSYQPKYCGV, VNAQCWPEQESRLCNL, CIPYKSKTIDVSFQCP, and PTTMDFTPAPLEDTSS, met all of the requirements, exhibiting antigenicity while lacking allergenic or toxic properties. Because of this, these sequences were chosen as strong candidates for the designed multi-epitope construct.

Table 2
Summary of B-cell Epitope Evaluation.
#	Rank	Sequence (ABCPred)	Start position	Score	Antigenicity (Vaxijen 2.0)	Allergenicity (Allertop v2.0)	Toxicity (ToxinPred)
1.	1	YQPKYCGVCMDNRCCI	339	0.95	0.7144 (PA)	NAlgn	Toxin
2.	2	TALSPAPTTMDFTPAP	66	0.93	0.5412 (PA)	NAlgn	NT
3.	2	VGCVLDGVRYNNGQSF	164	0.93	0.9813 (PA)	NAlgn	NT
4.	2	VSLITDGCECCKMCAQ	109	0.93	0.8508 (PA)	NAlgn	Toxin
5.	3	HTLIKAGKKCLAVYQP	307	0.92	0.4392 (PNA)	Na	Na
6.	4	APLEDTSSRPQFCKWP	80	0.91	0.9400 (PA)	NAlgn	Toxin
7.	4	CKYNCTCIDGAVGCTP	183	0.91	0.3555 (PNA)	Na	Na
8.	5	CECPPSPPRCPLGVSL	96	0.90	-0.1744 (PNA)	Na	Na
9.	5	AGCISTRSYQPKYCGV	331	0.90	0.8578 (PA)	NAlgn	NT
10.	5	HMSAPIENSXGNCCNP	25	0.90	0.6672 (PA)	X–?	Na
11.	5	RGLYCDYSGDRPRYAI	140	0.90	0.3679 (PNA)	Na	Na
12.	6	RPPRLWCPHPRRVSIP	204	0.88	0.4615 (PNA)	Na	Na
13.	7	VNAQCWPEQESRLCNL	284	0.87	0.6110 (PA)	NAlgn	NT
14.	7	HRNCIAYTSPWSPCST	256	0.87	0.1829 (PNA)	Na	Na
15.	7	HCCEQWVCEDDAKRPR	221	0.87	-0.1346 (PNA)	Na	Na
16.	8	GKKCLAVYQPEASMNF	313	0.85	-0.0161 (PNA)	Na	Na
17.	8	TSCGLGVSTRISNVNA	271	0.85	0.7004 (PA)	Algn	Na
18.	9	KTAPRDTGAFDAVGEV	237	0.83	0.7635 (PA)	Algn	Na
19.	9	EAAICDPHRGLYCDYS	132	0.83	0.4592 (PNA)	Na	Na
20.	10	RVSIPGHCCEQWVCED	215	0.82	0.1095 (PNA)	Na	Na
21.	11	CIPYKSKTIDVSFQCP	353	0.81	0.5438 (PA)	NAlgn	NT
22.	11	GEVEAWHRNCIAYTSP	250	0.81	0.4031 (PNA)	Na	Na
23.	12	VLWINACFCNLSCRNP	377	0.80	0.1757 (PNA)	Na	Na
24.	12	YTSPWSPCSTSCGLGV	262	0.80	0.1613 (PNA)	Na	Na
25.	13	EQESRLCNLRPCDVDI	291	0.79	0.8612 (PA)	Algn	Na
26.	13	DGAVGCTPLCLRVRPP	191	0.79	0.6115 (PA)	Algn	Na
27.	13	AMILYMEMERSHMSAP	14	0.79	0.1334 (PNA)	Na	Na
28.	14	AQQLGDNCTEAAICDP	123	0.78	0.7048 (PA)	Algn	Na
29.	15	GQSFQPNCKYNCTCID	176	0.77	0.1831 (PNA)	Na	Na
30.	16	SVMRWFLPWTLAAVTA	42	0.74	0.2474 (PNA)	Na	Na
31.	16	PNDIFADLESYPDFSE	392	0.74	0.7395 (PA)	Algn	Na
32.	16	ECCKMCAQQLGDNCTE	117	0.74	0.3011 (PNA)	Na	Na
33.	17	PTTMDFTPAPLEDTSS	72	0.73	0.6053 (PA)	NAlgn	NT
34.	17	NLSCRNPNDIFADLES	386	0.73	0.4630 (PNA)	Na	Na
35.	17	GVRYNNGQSFQPNCKY	170	0.73	0.7613 (PA)	Algn	Na
36.	18	AVTAAAASTVLATALS	54	0.70	-0.0077 (PNA)	Na	Na
37.	18	PEASMNFTLAGCISTR	322	0.70	0.3821 (PNA)	Na	Na
38.	19	DGLGFSRQVLWINACF	369	0.69	0.0860 (PNA)	Na	Na
39.	20	ASTVLATALSPAPTTM	60	0.68	0.0674 (PNA)	Na	Na
40.	20	DVSFQCPDGLGFSRQV	362	0.68	0.2236 (PNA)	Na	Na
41.	20	GNCCNPESVMRWFLPW	35	0.68	0.6844 (PA)	Algn	Na
42.	21	PCCNHMANCCNFAMIL	2	0.55	0.6965 (PA)	Algn	Na
Probable Antigen: PA, Probable Non-Antigen: PNA, Non-allergen: NAlgn, Allergen: Algn, Non-toxin: NT, Not applicable: Na

Prediction of Epitopes Binding into MHC Class I and II

After a preliminary ranking by IEDB scores, epitopes binding to MHC class I and II with rankings > 1% were excluded. VaxiJen v2.0 (threshold 0.5) was utilized for assessing the antigenicity of the remaining epitopes. The scores were then normalized and averaged based on IEDB values to rank the candidates. The top epitopes were next investigated for allergenicity (AllerTOP v2.0) and toxicity (ToxinPred2). Only sequences that were not allergenic/toxic were selected. Table 3 illustrates that the chosen epitopes, DIFADLESY (HLA-A*26:01), RISNVNAQCW (HLA-B*57:01), SPAPTTMDF (HLA-B*35:01), STRSYQPKY (HLA-A*30:02), and HTLIKAGKK (HLA-A*03:01), had high antigenicity, were not allergens/toxic, which made them ideal candidates for vaccines.

Table 3
List of epitopes binding to MHC class I molecules.
#	Allele	Peptide	score_IEDB	Antigenicity (VaxiJen_ Score)	Normalized_ antigenicity (VaxiJen_ Score)	Mean_score (VaxiJen+ IEDB/2)		Toxicity (Toxin_pred)
1.	HLA-A*26:01	DIFADLESY	0.910506	1.1379	0.612630559	0.761568279	PNAl	NT
2.	HLA-B*57:01	RISNVNAQCW	0.855398	1.1531	0.620814041	0.738106021	PNAl	NT
3.	HLA-B*35:01	SPAPTTMDF	0.980229	0.8277	0.445622914	0.712925957	PNAl	NT
4.	HLA-A*30:02	STRSYQPKY	0.783257	0.8812	0.474426618	0.628841809	PNAl	NT
5.	HLA-A*02:03	GVSTRISNV	0.642072	1.01	0.543770862	0.592921431	PNAl	NT
6.	HLA-A*01:01	AICDPHRGLY	0.78995	0.6249	0.336438032	0.563194016	PNAl	NT
7.	HLA-A*01:01	ICDPHRGLY	0.775061	0.6186	0.333046194	0.554053597	PNAl	NT
8.	HLA-B*35:01	NPNDIFADL	0.594541	0.9373	0.504630128	0.549585564	PNAl	NT
9.	HLA-A*30:02	CVLDGVRY	0.053162	1.8574	1	0.526581	PNAl	NT
10.	HLA-A*03:01	HTLIKAGKK	0.382064	1.2461	0.670884031	0.526474016	PNAl	NT
Probable non-allergen: PNAl, Non-Toxin: NT

The last group of MHC-II epitopes (Table 4) was chosen because they had high mean scores, were not allergens, and were not poisonous. The epitopes are NDIFADLESYPDF (HLA-DRB1*04:05), CVLDGVRYNNGQSFQPN (HLA-DRB3*02:02), TGAFDAVGEVEAWHR (HLA-DRB1*09:01), DVDIHTLIKAGKK (HLA-DRB1*11:01), and ATALSPAPTTMDFTP (HLA-DRB1*04:05). All of them had great antigenicity and safety, which made them perfect for the vaccine build.

Table 4
List of MHC class II epitopes.
#	Allele	Peptide	Score IEDB	Antigenicity (VaxiJen Score)	Normalizedantigenicity (VaxiJen_ Score)	Mean score (VaxiJen + IEDB/2)	Allergenicity (AllerTOP_ v.2.0)	Toxicity (Toxin_pred)
1.	HLA-DRB1*04:05	NDIFADLESYPDF	0.844	1.0016	0.714	0.779	PNAl	NT
2.	HLA-DRB3*02:02	CVLDGVRYNNGQSFQPN	0.446	1.0534	0.751	0.598	PNAl	NT
3.	HLA-DRB1*09:01	TGAFDAVGEVEAWHR	0.329	1.1482	0.819	0.574	PNAl	NT
4.	HLA-DRB1*11:01	DVDIHTLIKAGKK	0.492	0.8836	0.630	0.561	PNAl	NT
5.	HLA-DRB1*04:05	ATALSPAPTTMDFTP	0.698	0.5911	0.421	0.559	PNAl	NT
Probable non-allergen: PNAl, Non-Toxin: NT

Design and Assessment of the Construct

To construct a vaccine, six B lymphocyte, nine MHC-I, and one MHC-II epitopes were implemented. Their folding and function were preserved with linkers (KK and GGGGS for B-cell, AAY for MHC-I, GPGPG for MHC-II, and EAAAK for adjuvant integration). Thymosin α1 (Tα1) was chosen because it has established clinical safety and immunomodulatory activity. It is a licensed therapeutic agent for viral infections and for immunotherapy of cancers, in which it boosts both innate and adaptive immunity. Of particular relevance to a therapeutic tumor vaccine, Tα1 induces Th1-type immunity, stimulates dendritic cells and T cells, and enhances antigen presentation by upregulating MHC antigens. These activities would all be highly advantageous for a therapeutic tumor vaccine because they would all promote strong cytotoxic T-cell immunity to tumor antigens like WISP1 (37). Also, a His-tag was introduced at the C-terminal aimed at facilitating purification and identification. The construct (Fig. 2) was of robust antigenicity, predicted not to be allergenic, and confirmed not to be toxic, thus indicating strong promise as a vaccine candidate.

Physicochemical Characteristics

The vaccine construct, composed of 274 amino acids, weighs about 29.98 kDa and has a predicted isoelectric point of 6.23, indicating it is somewhat acidic (Table 5). Its stability varies depending on the environment: the predicted half-life was estimated at 1.9 h in human cells, > 20 h in yeast, and > 10 h in E. coli. Its stability score indicates that it holds up well and aliphatic index shows its intermediate thermostability, indicating that it is a candidate of promise. Its GRAVY score of -0.481 indicates that it has a hydrophilic nature that would allow for favorable water-base interactions, and these properties cumulatively bring it to the point of being a candidate of quality for experimental study.

Table 5
Predicted physicochemical attributes of the multi-epitope vaccine evaluated using ProtParam that reflect construct stability and hydrophilicity.
Amount	Physicochemical features
Amino acids	274
Molecular weight	29979.53
PI (theoretical)	6.23
Predicted half-life	1.9 h in mammalian reticulocytes (in vitro) > 20 h in yeast (in vivo) > 10 h in E. coli (in vivo)
Index of instability	33.89 (stable)
Aliphatic index	57.96
GRAVY (Grand average of hydropathicity)	-0.481

Secondary Structure Prediction

In silico secondary structure analysis of the construct (Fig. 3) showed the following distributions of elements: alpha helix (50 residues, 18.25%), extended strands (55 residues, 20.07%), beta turns (26 residues, 7.34%), and random coils (169 residues, 61.68%). These results show that the vaccine has a mix of secondary structure elements, which is suitable for its functionality and stability.

The pictorial depiction of the secondary structure prediction reveals regions of different elements across the sequence, and the coils, beta strands, and alpha helices are clearly distinguished. A major region of helicity structure with stretches of extended strand differentiated by turns and coils is anticipated, typical for a stable protein construct.

Tertiary Conformation Prediction

Three-dimensional structure modeling was performed using Robetta and AlphaFold. The model exhibited outstanding quality, as denoted by a MolProbity score of 1.78 (86% similarity to best-ranked structures) and clash score of 4.14 (96% similarity to optimal structures). ProSA-web validation generated the Z-score of − 7, typical for native proteins. Additional validation was performed using ERRAT, which yielded an overall quality score of 90.94%, and Ramachandran plot analysis, which indicated 98.3% of residues in favored/allowed regions and 1.7% in disallowed regions (Fig. 4). Collectively, these results verify an accurate backbone conformation appropriate for the next steps, including molecular docking and dynamics investigations.

Molecular Docking with Immune Receptors

TLR2 is the most abundant TLR found in gastric tumors, whereas high TLR4 expression has been connected with elevated risk for gastric cancer (38). Docking of the construct with TLR2 and TLR4 was conducted using ClusPro 2.0, and binding affinities were also assessment by PRODIGY and PDBsum. Parameters analyzed were binding energy, weight score, dissociation constant (Kd), and active site interactions, with the residue contacts visualized by PDBsum (Fig. 5 and Table 6). The vaccine demonstrated favorable interactions with the receptors with particularly strong binding affinity for TLR4.

Table 6
Binding evaluation results among designed candidate vaccine and TLR2/TLR4.
Complex	ΔG (kcal mol-1)	Weighted score	Number of hydrogen bonds	PRODIGY (Kd)
Vaccine construct–TLR2	-9.0	-1255.0	2	2.4e-07
Vaccine construct–TLR4	-17.8	-1067.0	21	9.1e-14

Molecular Dynamics Analysis

MD simulations were performed on unbound TLR2 and TLR4 for 100 ns to stabilize their structures under physiological conditions. The vaccine construct was also simulated alone and in complex with both receptors. Mean RMSD values were 0.31 nm for TLR2, 0.24 nm for TLR4, 0.94 nm for the vaccine, 1.28 nm for the vaccine–TLR2 complex, and 0.68 nm for the vaccine–TLR4 complex. The last one showed the best stability (Fig. 6a). The structural rigidity (Rg) was also stable, and the means were 3.0, 3.1, 2.21, 3.95, and 3.33 nm for TLR2, TLR4, vaccine, vaccine–TLR2, and vaccine–TLR4, respectively (Fig. 6b). The RMSF analysis showed that there were very small changes in residue levels across all systems. TLR4 complexes were the most stable, whereas TLR2 complexes only showed changes of less than 1 nm (Fig. 6c).

Binding-Free Energy Calculation

The Hawdock web server predicted binding free energies between the construct and TLR4 as well as TLR2 by MM/GBSA method. Total binding energy comprised contributions from Gibbs free energy, SASA, van der Waals, and electrostatic interactions. As illustrated in Table 7, the SASA, electrostatic, and Van der Waals energies were suitable energies in the interaction vaccine-TLR receptors, also electrostatic energy was favorable energy in the interaction. However, unfavorable energy in the interaction vaccine with TLR2 and TLR4 was the Gibss energy. The lower vaccine–TLR4 interaction energy indicated greater affinity of the construct to TLR4 compared to TLR2.

Table 7
MM/GBSA binding free energy components (electrostatic, Van der Waals, Gibbs, SASA) for two complexes (vaccine–TLR2/TLR4). The minor total energy of the TLR4 complex indicates stronger affinity.
Energy	TLR2-vaccine complex	TLR4-vaccine complex
van der Waals energy (KJ/mol)	-98.88	-191.96
Electrostatic energy (KJ/mol)	-210.02	-594.12
Gibbs energy (KJ/mol)	272.75	680.15
SASA energy (KJ/mol)	-11.39	-26.57
Total energy (KJ/mol)	-47.55	-132.5

Immune Simulation

C-ImmSim predicted robust humoral and cellular responses, with elevated primary and secondary responses and reduced antigen levels (Fig. 7a–i). B-cell and memory B-cell populations rose after each dose and persisted for 350 days (Fig. 7a). Antibody levels (IgG, IgM) increased with each injection, while antigens decreased (Fig. 7b). Active and memory TH cells remained elevated (Fig. 7c–d), TC cells expanded steadily (Fig. 7e–f), and NK cells and macrophages were enhanced (Fig. 7g–h). Cytokines IFN-γ, IL-10, and TGF-β were also upregulated (Fig. 7i).

Optimization of the Codon and Computational Cloning

Back-translation and codon optimization of the construct were performed using JCat. The optimized vaccine nucleotide sequence had 1.0% CAI and 55.72% GC content. After codon optimization, 834 nucleotides was the sequence length of the designed structure. The simulation of the optimized construct sequence in pET-28a ⁽⁺⁾ by SnapGene indicated the construct’s sequence is suitable for cloning in pET-28a ⁽⁺⁾ (Fig. 8a). The beginning and end of the sequence of the construct were digested with XbaI as well as XhoI enzymes, respectively, and the double digestion procedure by these enzymes showed the presence of the designed construct (828 bp) and the vector (pET-28a⁽⁺⁾, 5231 bp) (Fig. 8b).

This study presents a rational, immunoinformatics-based design of a multi-epitope vaccine directed against WISP1 (CCN4) for GC. In line with the study objective, we first verified that WISP1 is a secreted, extracellular protein, thereby accessible to humoral and cellular immune recognition. Building on this, we prioritized B lymphocyte and T lymphocyte epitopes with favorable antigenicity, non-allergenicity, as well as non-toxicity, and assembled them into a construct with well-validated linkers and an immunostimulatory adjuvant (thymosin-α1). Structural modeling and stringent quality checks (ProSA, Ramachandran, ERRAT, MolProbity) supported a well-folded tertiary structure. Functionally, the construct docked favorably to human TLR2 and, more strongly, to TLR4, and remained stable during 100-ns MD simulations. MM/GBSA estimates further indicated a lower binding free energy (more favorable) for the TLR4 complex than for TLR2, consistent with the MD readouts. Finally, in silico immune simulations suggested robust primary and secondary responses with durable B-cell and T-cell memory and induction of Th1-skewing cytokines. Altogether, these results provide a computational foundation for developing a WISP1-directed therapeutic vaccine candidate for GC.

Recent literature supports several pillars of this design. First, CCN4/WISP1 has emerged as a tumor-promoting matricellular protein that shapes the tumor microenvironment, EMT, and immune contexture across cancers; these properties have renewed interest in CCN4 as an immune-visible target when aberrantly overexpressed in tumors (9). (For broader CCN4 biology and immune–stroma interactions, see contemporary reviews synthesizing evidence across solid tumors (39).) Second, the decision to engineer a construct capable of engaging TLR4 is biologically coherent: TLR agonists are among the most validated vaccine adjuvant classes for eliciting durable Th1-biased responses and potent antigen presentation, with TLR4 agonism (e.g., MPLA) translated to licensed human vaccines and actively optimized in current adjuvant platforms (40–42). Third, our choice of Tα1 as the intrinsic adjuvant is supported by recent clinical-immunological evidence showing that Tα1 augments dendritic-cell activation, enhances antigen presentation, and reinforces T-cell effector and memory responses, mechanisms directly aligned with therapeutic cancer vaccination (43, 44). Finally, multi-epitope vaccine pipelines that combine epitope filtering for antigenicity, allergenicity/toxicity screening, structural validation, and docking/MD against innate sensors have matured, with contemporary exemplars indicating that such in silico designs can yield constructs that are stable, immunogenic, and ready for experimental testing (45, 46).

Biologically and clinically, a WISP1-targeted immunotherapy is attractive for several reasons. As a secreted TAA enriched in GC and implicated in tumor–stroma remodeling (7, 47), WISP1 offers (i) immune accessibility without the constraints of intracellular antigen processing for B-cell recognition; (ii) the potential to disrupt protumorigenic signaling axes and EMT indirectly via immune pressure; and (iii) a route to broaden patient coverage by combining multiple B lymphocyte and T lymphocyte epitopes across MHC classes I and II alleles. The observed preference for TLR4 engagement is also practical: TLR4-primed antigen presentation favors the cross-priming of CD8⁺ T lymphocyte and the production of durable memory, both central to sustained control of solid tumors (48). Moreover, the predicted Th1-skewed cytokine profile in our immune simulations (e.g., IFN-γ elevation) is congruent with antitumor efficacy and with the known immunobiology of Tα1 adjuvanticity (37, 49).

This work has several strengths. Methodologically, we used conservative, consensus-style epitope triaging (antigenicity, non-allergenicity, non-toxicity), applied state-of-the-art tertiary modeling and multi-metric quality control, and tested receptor engagement and complex stability with docking, MD, and MM/GBSA. Conceptually, we focused on a secreted, immuno-visible, and functionally relevant TAA; operationally, we embedded an adjuvant (Tα1) with contemporary evidence of clinical safety and mechanistic suitability for cancer immunotherapy.

Limitations should also be acknowledged. All findings are in silico and require empirical validation. Epitope immunogenicity can differ in primary human APCs and across HLA haplotypes; proteasomal processing, epitope competition, and epitope dominance were not experimentally assessed. While docking and MD suggest preferential TLR4 engagement, receptor binding in cells is influenced by membrane context and co-receptors (e.g., CD14/MD-2), which were not modeled here. Furthermore, although C-ImmSim offers useful first-pass predictions, it cannot substitute for in vivo immunogenicity, safety, and antitumor efficacy testing. Finally, WISP1 expression heterogeneity across GC subtypes and stages will need careful patient-selection strategies in eventual trials.

Future studies should therefore (i) validate epitope presentation by mass spectrometry and measure T- and B-cell responses in human PBMCs and HLA-diverse murine models; (ii) compare Tα1 with, or combine it alongside, TLR4/TLR3 agonists (e.g., MPLA, poly I:C) to optimize magnitude and quality of responses; (iii) confirm receptor engagement using cell-based TLR reporter assays and assess downstream NF-κB/IRF signaling; (iv) perform tumor-bearing mouse studies to quantify antitumor activity, memory, and safety; and (v) explore rational combinations with checkpoint inhibitors or chemotherapy, in keeping with contemporary cancer-vaccine strategies. Parallel translational work should map WISP1 expression and secretion in GC cohorts and investigate whether serum WISP1 could serve as a pharmacodynamic or stratification biomarker.

In summary, our data nominate WISP1 as a tractable, immune-visible TAA in GC and provide a well-vetted, multi-epitope vaccine candidate with predicted TLR4-favored engagement, structural stability, and durable immunogenicity. By integrating rigorous epitope selection with structural and receptor-level validation, and by selecting an evidence-based adjuvant, the present work lays a credible foundation for preclinical testing and, ultimately, for evaluating WISP1-targeted vaccination as a novel therapeutic avenue in gastric cancer.

This work describes the first bioinformatics-based design of a WISP1-targeting vaccine with multiple epitopes against GC. By combining careful epitope selection with immunogenicity and safety filtering, structure-based validation, and receptor-level docking and dynamics, we derived a construct that should be stable, immunogenic, and have the ability to engage TLR4 to facilitate sustained antitumor responses. These observations align with the study aim of nominating WISP1 as a rational, immune-accessible target for vaccination and extend therapeutic approaches to extracellular matricellular proteins, an otherwise relatively unstudied class of target molecules in cancer therapy. Although the interpretations reflect computational predictions and can’t reflect the full richness of the tumor microenvironment complexity and inter-patient variability, they constitute a plausible basis warranting experimental assessment. Here, follow-up studies should emphasize in vitro demonstration of dendritic cell activation and induction of cytokine production, complemented by in vivo evaluation of antitumor activity and safety of the lead construct of interest in appropriate GC model systems. Overall, the present work places WISP1-directed vaccination on the table as a promising option for expanding immunotherapeutic intervention strategies in GC.

CAI: Codon Adaptation Index; EMT: Epithelial–Mesenchymal Transition; GRAVY: Grand Average of Hydropathicity; MD: Molecular Dynamics; MM/GBSA: Molecular Mechanics/Generalized Born Surface Area; Rg: Radius of Gyration; RMSD: Root Mean Square Deviation; RMSF: Root Mean Square Fluctuation; SASA: Solvent Accessible Surface Area; TAA: Tumor-Associated Antigen; Tα1: Thymosin alpha 1; WISP1 (CCN4): WNT1-Inducible Signaling Pathway Protein 1.

Funding

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

MSR and NMD conceived the study. NMD, BS and KA curated the data, performed the research and formal analysis. NMD, BS and KA wrote the manuscript and prepared the figures. All authors read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Consent to participate

Animals, humans, or clinical data were not a part of this investigation.

Consent for publication

Not applicable.

Availability of data and materials

All data generated or analyzed during this study are included in this article. The WISP1 sequence was retrieved from UniProt (ID: O95388), and structural data were retrieved from the Protein Data Bank (PDB IDs: 5D3I and 7MLM).

Acknowledgment

The authors thank Iran University of Medical Sciences and Semnan University of Medical Sciences for assistance. The authors used Grammarly, QuillBot and ChatGPT (OpenAI, GPT‑4, accessed 2025) solely for language editing and clarity. No AI tools were used to generate scientific content, analyze data, interpret results or create figures. All authors are responsible for the final manuscript.

Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229–63. 10.3322/caac.21834.
Hsu J-T, Lin Y-N, Chen Y-F, Kou H-W, Wang S-Y, Chou W-C, et al. A comprehensive overview of gastric cancer management from a surgical point of view. Biomedical J. 2025;48(4):100817. https://doi.org/10.1016/j.bj.2024.100817.
Ghemrawi R, Abuamer L, Kremesh S, Hussien G, Ahmed R, Mousa W, et al. Revolutionizing Cancer Treatment: Recent Advances in Immunotherapy. Biomedicines. 2024;12(9). 10.3390/biomedicines12092158.
Gheybi E, Salmanian AH, Fooladi AAI, Salimian J, Hosseini HM, Halabian R, et al. Immunogenicity of chimeric MUC1-HER2 vaccine against breast cancer in mice. Iran J Basic Med Sci. 2018;21(1):26–32. 10.22038/ijbms.2017.25686.6335.
Sueangoen N, Thuwajit P, Yenchitsomanus P-T, Thuwajit C. Public neoantigens in breast cancer immunotherapy (Review). Int J Mol Med. 2024;54(1):65. 10.3892/ijmm.2024.5388.
Ruzzi F, Riccardo F, Conti L, Tarone L, Semprini MS, Bolli E, et al. Cancer vaccines: Target antigens, vaccine platforms and preclinical models. Mol Aspects Med. 2025;101:101324. https://doi.org/10.1016/j.mam.2024.101324.
Chen H, Zhang X, Zhang Z, Li G, Li X, Yang S, et al. Identification and validation of CCN family genes to predict the prognosis in gastric cancer. Discov Oncol. 2024;15(1):610. 10.1007/s12672-024-01459-2.
Zhang LH, Wang Y, Fan QQ, Liu YK, Li LH, Qi XW, et al. Up-regulated Wnt1-inducible signaling pathway protein 1 correlates with poor prognosis and drug resistance by reducing DNA repair in gastric cancer. World J Gastroenterol. 2019;25(38):5814–25. 10.3748/wjg.v25.i38.5814.
Gaudreau P-O, Clairefond S, Class CA, Boulay P-L, Chrobak P, Allard B, et al. WISP1 is associated to advanced disease, EMT and an inflamed tumor microenvironment in multiple solid tumors. OncoImmunology. 2019;8(5):e1581545. 10.1080/2162402X.2019.1581545.
Khoshdel F, Mottaghi-Dastjerdi N, Yazdani F, Salehi S, Ghorbani A, Montazeri H, et al. CTGF, FN1, IL-6, THBS1, and WISP1 genes and PI3K-Akt signaling pathway as prognostic and therapeutic targets in gastric cancer identified by gene network modeling. Discover Oncol. 2024;15(1):344. 10.1007/s12672-024-01225-4.
Xi Y, LaCanna R, Ma HY, N'Diaye EN, Gierke S, Caplazi P, et al. A WISP1 antibody inhibits MRTF signaling to prevent the progression of established liver fibrosis. Cell Metab. 2022;34(9):1377–e938. 10.1016/j.cmet.2022.07.009.
Hamley IW. Peptides for Vaccine Development. ACS Appl Bio Mater. 2022;5(3):905–44. 10.1021/acsabm.1c01238.
Rahimi-Farsi N, Shahbazi T, Ghorbani A, Mottaghi-Dastjerdi N, Yazdani F, Mohseni P, et al. Network-based analysis of candidate oncogenes and pathways in hepatocellular carcinoma. Biochem Biophys Rep. 2025;43:102086. https://doi.org/10.1016/j.bbrep.2025.102086.
Rahimi-Farsi N, Ghorbani A, Mottaghi-Dastjerdi N, Shahbazi T, Bostanian F, Mohseni P, et al. Comprehensive system biology analysis of microRNA-101-3p regulatory network identifies crucial genes and pathways in hepatocellular carcinoma. J Genetic Eng Biotechnol. 2025;23(1):100471. https://doi.org/10.1016/j.jgeb.2025.100471.
Rahimi H, Farzadifar E, Mottaghi-Dastjerdi N, Ghorbani A, Amerizadeh F, Pasdar A. Integrative Systems Biology Analysis of MicroRNA Regulation in Wnt Signaling Pathway During Breast Cancer Progression. Int J Cancer Manage. 2025;18(18):e164192. https://doi.org/10.5812/ijcm-160657.
Saeidi M, Pasdar A, Rahmani F, Ghorbani A, Mottaghi N, Amerizadeh F. A Prospective to Regulatory Role of miRNAs on Wnt/β-catenin Signaling and Its Crosstalk to the Other Cellular Pathways in Tumorigenesis of Glioblastoma by a Systems Biology Approach. Int J Cancer Manage. 2025;18(1). https://doi.org/10.5812/ijcm-156834.
Mottaghi-Dastjerdi N, Soltany-Rezaee-Rad M. AI-Powered Insights into Drug Resistance in Gastric Cancer: A Path Toward Precision Therapy. Iran J Pharm Research: IJPR. 2025;24(1):e159954. 10.5812/ijpr-159954.
Yazdani F, Mottaghi-Dastjerdi N, Shahbazi B, Ahmadi K, Ghorbani A, Soltany-Rezaee-Rad M, et al. Identification of key genes and pathways involved in T-DM1-resistance in OE-19 esophageal cancer cells through bioinformatics analysis. Heliyon. 2024;10(18). 10.1016/j.heliyon.2024.e37451.
Khoshdel F, Mottaghi-Dastjerdi N, Yazdani F, Salehi S, Ghorbani A, Montazeri H et al. CTGF, FN1, IL-6, THBS1, and WISP1 genes and PI3K-Akt signaling pathway as prognostic and therapeutic targets in gastric cancer identified by gene network modeling. 2023. 10.1007/s12672-024-01225-4
Mottaghi-Dastjerdi N, Ghorbani A, Montazeri H, Guzzi PH. A systems biology approach to pathogenesis of gastric cancer: gene network modeling and pathway analysis. BMC Gastroenterol. 2023;23(1):248. 10.1186/s12876-023-02891-4.
Mottaghi-Dastjerdi N, Soltany-Rezaee-Rad M, Sepehrizadeh Z, Roshandel G, Ebrahimifard F, Setayesh N. Identification of novel genes involved in gastric carcinogenesis by suppression subtractive hybridization. Hum Exp Toxicol. 2015;34(1):3–11. 10.1177/0960327114532386.
Mottaghi-Dastjerdi N, Soltany-Rezaee-Rad M, Sepehrizadeh Z, Roshandel G, Ebrahimifard F, Setayesh N. Gene expression profiling revealed overexpression of vesicle amine transport protein-1 (VAT-1) as a potential oncogene in gastric cancer. Indian J Biotechnol. 2016;15:161–5.
Mottaghi-Dastjerdi N, Soltany-Rezaee-Rad M, Sepehrizadeh Z, Roshandel G, Ebrahimifard F, Setayesh N. Genome expression analysis by suppression subtractive hybridization identified overexpression of Humanin, a target gene in gastric cancer chemoresistance. Daru. 2014;22(1):14. 10.1186/2008-2231-22-14.
Soltany-Rezaee-Rad M, Mottaghi-Dastjerdi N, Setayesh N, Roshandel G, Ebrahimifard F, Sepehrizadeh Z. Overexpression of FOXO3, MYD88, and GAPDH identified by suppression subtractive hybridization in esophageal cancer is associated with autophagy. Gastroenterol Res Pract. 2014;2014(1):185035. 10.1155/2014/185035.
Mottaghi-Dastjerdi N, Soltany-Rezaee-Rad M. Advancements and Applications of Artificial Intelligence in Pharmaceutical Sciences: A Comprehensive Review. Iran J Pharm Res. 2024;23(1):e150510. 10.5812/ijpr-150510.
Farsiu N, Ghorbani A, Karbanowicz TP, Mottaghi-Dastjerdi N, Veltri P, Guzzi PH. Discovery of long non-coding RNAs in naïve CD4 + T cells in response to initiating antiretroviral therapy at acute or chronic phase of HIV-1 infection. Acta Virol. 2024;68–2024. 10.3389/av.2024.11572.
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.
Sharma N, Naorem LD, Jain S, Raghava GPS. ToxinPred2: an improved method for predicting toxicity of proteins. Brief Bioinform. 2022;23(5). 10.1093/bib/bbac174.
Koymans KJ, Feitsma LJ, Brondijk TH, Aerts PC, Lukkien E, Lossl P, et al. Structural basis for inhibition of TLR2 by staphylococcal superantigen-like protein 3 (SSL3). Proc Natl Acad Sci U S A. 2015;112(35):11018–23. 10.1073/pnas.1502026112.
Su L, Athamna M, Wang Y, Wang J, Freudenberg M, Yue T, et al. Sulfatides are endogenous ligands for the TLR4-MD-2 complex. Proc Natl Acad Sci U S A. 2021;118(30):e2105316118. 10.1073/pnas.2105316118.
Kozakov D, Hall DR, Xia B, Porter KA, Padhorny D, Yueh C, 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.
Koohnavard F, Ahmadi K, Eftekhar E, Edalatmanesh MA. Computational screening of FDA-approved drugs to identify potential aromatase receptor inhibitors for polycystic ovary syndrome. J Biomol Struct Dyn. 2023;41(24):15507–19. 10.1080/07391102.2023.2190411.
Shahbazi B, Arab SS, Mafakher L, Azadmansh K, Teimoori-Toolabi L. Computational assessment of pigment epithelium-derived factor as an anti-cancer protein during its interaction with the receptors. J Biomol Struct Dynamics. 2023;41(10):4575–91. https://doi.org/10.1080/07391102.2022.2069863.
Weng G, Wang E, Wang Z, Liu H, Zhu F, Li D, et al. HawkDock: a web server to predict and analyze the protein-protein complex based on computational docking and MM/GBSA. Nucleic Acids Res. 2019;47(W1):W322–30. 10.1093/nar/gkz397.
Nuryana I, Laksmi FA, Dewi KS, Akbar FR, Nurhayati, Harmoko R. Codon optimization of a gene encoding DNA polymerase from Pyrococcus furiosus and its expression in Escherichia coli. J Genet Eng Biotechnol. 2023;21(1):129. 10.1186/s43141-023-00605-7.
Dominari A, Hathaway Iii D, Pandav K, Matos W, Biswas S, Reddy G, et al. Thymosin alpha 1: A comprehensive review of the literature. World J Virol. 2020;9(5):67–78. 10.5501/wjv.v9.i5.67.
Meliț LE, Mărginean CO, Mărginean CD, Mărginean MO. The Relationship between toll-like receptors and helicobacter pylori‐related gastropathies: still a controversial topic. J Immunol Res. 2019;2019(1):8197048. 10.1155/2019/8197048.
Singh K, Oladipupo SS. An overview of CCN4 (WISP1) role in human diseases. J Translational Med. 2024;22(1):601. 10.1186/s12967-024-05364-8.
Jeon D, Hill E, McNeel DG. Toll-like receptor agonists as cancer vaccine adjuvants. Hum Vaccin Immunother. 2024;20(1):2297453. 10.1080/21645515.2023.2297453.
Fensterheim BA, Young JD, Luan L, Kleinbard RR, Stothers CL, Patil NK, et al. The TLR4 Agonist Monophosphoryl Lipid A Drives Broad Resistance to Infection via Dynamic Reprogramming of Macrophage Metabolism. J Immunol. 2018;200(11):3777–89. 10.4049/jimmunol.1800085.
Roman F, Burny W, Ceregido MA, Laupèze B, Temmerman ST, Warter L, et al. Adjuvant system AS01: from mode of action to effective vaccines. Expert Rev Vaccines. 2024;23(1):715–29. 10.1080/14760584.2024.2382725.
Liu K, Kong L, Cui H, Zhang L, Xin Q, Zhuang Y, et al. Thymosin α1 reverses oncolytic adenovirus-induced M2 polarization of macrophages to improve antitumor immunity and therapeutic efficacy. Cell Rep Med. 2024;5(10):101751. 10.1016/j.xcrm.2024.101751.
Yao S, Huang Q, Zou Y, Liu T, Yang Y, Huang T, et al. The efficacy and safety of thymosin alpha-1 combined with lenvatinib plus sintilimab in unresectable hepatocellular carcinoma: a retrospective study. Sci Rep. 2025;15(1):13960. 10.1038/s41598-025-97160-7.
Li X, Wen H, Xiao X, Ren Z, Tan C, Fu C. Design of a novel multi-epitope vaccine candidate against endometrial cancer using immunoinformatics and bioinformatics approaches. J Biomol Struct Dyn. 2024;42(21):11521–37. 10.1080/07391102.2023.2263213.
Motamedi H, Ari MM, Shahlaei M, Moradi S, Farhadikia P, Alvandi A, et al. Designing multi-epitope vaccine against important colorectal cancer (CRC) associated pathogens based on immunoinformatics approach. BMC Bioinformatics. 2023;24(1):65. 10.1186/s12859-023-05197-0.
Chang K-S, Chen S-T, Sung H-C, Hsu S-Y, Lin W-Y, Hou C-P, et al. WNT1 Inducible Signaling Pathway Protein 1 Is a Stroma-Specific Secreting Protein Inducing a Fibroblast Contraction and Carcinoma Cell Growth in the Human Prostate. Int J Mol Sci. 2022;23(19):11437. 10.3390/ijms231911437.
Lavelle EC, McEntee CP. Vaccine adjuvants: Tailoring innate recognition to send the right message. Immunity. 2024;57(4):772–89. 10.1016/j.immuni.2024.03.015.
Wei Y, Zhang Y, Li P, Yan C, Wang L. Thymosin α-1 in cancer therapy: Immunoregulation and potential applications. Int Immunopharmacol. 2023;117:109744. https://doi.org/10.1016/j.intimp.2023.109744.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
17 Dec, 2025
Reviews received at journal
11 Dec, 2025
Reviewers agreed at journal
08 Dec, 2025
Reviews received at journal
08 Dec, 2025
Reviewers agreed at journal
07 Dec, 2025
Reviewers agreed at journal
07 Dec, 2025
Reviews received at journal
07 Dec, 2025
Reviewers agreed at journal
06 Dec, 2025
Reviewers invited by journal
06 Dec, 2025
Editor assigned by journal
06 Dec, 2025
Editor invited by journal
05 Dec, 2025
Submission checks completed at journal
05 Dec, 2025
First submitted to journal
05 Dec, 2025

You are reading this latest preprint version

Immunoinformatics-Guided Design of a Multi-Epitope Vaccine Targeting WISP1 for Gastric Cancer

Status:

Version 1

Abstract

Background

Results

Conclusions

Figures

Introduction

Methods

Protein Sequence Retrieval and Localization

Prediction of B-cell Epitopes

Antigenicity, Allergenicity and Toxicity Assessment

Identification and selection of top-scoring MHC-I and MHC-II epitopes

Design of the Vaccine Framework

Assessment of Antigenicity, Allergenicity, and Toxicity

Physicochemical Properties of Candidate Vaccine

Secondary and Tertiary Structure Prediction

Docking with Immune Receptors

Simulation of Molecular Dynamics (MD)

Binding-Free Energy (BFE) Calculation

Immune Simulation

Optimization of the Codon and Computational Cloning

Results

Protein Localization Analysis

Prediction of the B lymphocyte Epitopes

Assessment of Antigenicity, Allergenicity, and Toxicity

Prediction of Epitopes Binding into MHC Class I and II

Design and Assessment of the Construct

Physicochemical Characteristics

Secondary Structure Prediction

Tertiary Conformation Prediction

Molecular Docking with Immune Receptors

Molecular Dynamics Analysis

Binding-Free Energy Calculation

Immune Simulation

Optimization of the Codon and Computational Cloning

Discussion

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1