Design, synthesis, and evaluation of A therapeutic vaccine candidate against lung cancer based on multi-epitopes of MAGE-A3, TGF-β2, and VEGF-A.

To confirm the precise selection of targets, a comprehensive expression analysis was characterized using UALCAN. Figure 2 A and I show the expression of VEGF-A , TGF-β2 , an d MAGE-A3 , as well as their impact on LUAD patient survival. Additionally, we utilized the Pan-cancer view of the UALCAN database to further examine the expression of these targets in tumor and normal tissues. According to UALCAN, VEGF-A mRNA expression was notably higher in LUAD cancer compared to normal tissue. This over-expression of VEGF-A is closely linked to tumor advancement and differentiation, suggesting that higher VEGF-A levels are associated with more advanced cancer stages and poorer survival rates following treatment. Furthermore, VEGF-A ’s impact extends beyond NSCLC, significantly influencing various other malignancies due to its involvement in angiogenesis. In addition, TGF-β2 is commonly expressed in both NSCLC and normal samples and its expression varies across different stages of cancer. Interestingly, its mRNA and protein were found to be more expressed in patients with advanced stages of LUAD, while they were rarely found in tissues of patients at stages 3, 2, and 1, respectively. According to online analysis from UALCAN, patients with high TGF-β2 transcriptional levels have notably lower 5-year survival rates compared to those with low expression levels. Furthermore, the presence of this factor is highly correlated with aggressive tumor characteristics and a grim prognosis, establishing it as a prime target for therapeutic interventions. MAGE-A3 serves as the foremost biomarker for diagnosing NSCLC, with elevated levels strongly correlated with lower histological grades and overall survival rates. Its highly tumor-specific nature is evident through minimal expression in normal tissues. The UALCAN algorithm has confirmed its upregulated expression, particularly in the advanced stages of LUAD. High serum levels of MAGE-A3 can indicate a poor prognosis and are often used in conjugation with imaging techniques to monitor NSCLC progression. While primarily associated with NSCLC, high MAGE-A3 expression levels have also been strongly linked to poor overall survival in LUAD and adverse outcomes in NSCLC patients. To examine the functions of our target genes, we analyzed their expression across various tissues, as well. Our search in the HPA database revealed high expression of VEGF-A in specific cancer types, while TGF-β2 displayed varying levels of expression across different cancers. Moving forward, our investigation also unveiled high expression of MAGE-A3 in certain cancer types, while indicating low or no expression in other cancers, as well as in normal tissues. The construction of the PPI network and module analysis via the ZS Revelen analysis reveals a highly interconnected network of 97 proteins with 543 interactions (Fig. 2 J). The result of the protein-protein interaction network is a highly connected network of molecules, forming a captivating graph where proteins act as the nodes and interactions as the edges. Within this network, some proteins take center stage, while others are arranged more loosely at the periphery. The complexity of the network is underscored by multiple interactions between select proteins, providing a comprehensive insight into the interplay of molecular interactions. We employed the TIMER method to evaluate tumor purity about the expression of VEGFA , TGF-β2 , and MAGE-A3 genes in NSCLC cancer (Fig. 2 K, L and M). Our analysis unveiled a compelling partial correlation between the expression of these genes and tumor purity in LUAD datasets. These findings underscore the potential influence of these gene targets on the microenvironment of lung cancer tumors. Fig. 2 The Bioinformatic analysis to confirm target selection. The expression of VEGF-A A , TGF-β2 (B), and MAGE-A3 C in normal and tumors. The expression level of VEGF-A D , TGF-β2 E , and MAGE-A3 F on LUAD patient survival. The mRNA expression in multiple human cancers is strongly higher in comparison with corresponding normal tissues VEGF-A G , TGF-β2 H , and MAGE-A3 I . The PPI network of the VEGF-A, TGF-β2, and MAGE-A3 was reconstructed using the ZS Revelen online database J . The correlation between the immune marker sets of immune cells and VEGF-A K , TGF-β2 L , and MAGE-A3 M expression in LUAD The Bioinformatic analysis to confirm target selection. The expression of VEGF-A A , TGF-β2 (B), and MAGE-A3 C in normal and tumors. The expression level of VEGF-A D , TGF-β2 E , and MAGE-A3 F on LUAD patient survival. The mRNA expression in multiple human cancers is strongly higher in comparison with corresponding normal tissues VEGF-A G , TGF-β2 H , and MAGE-A3 I . The PPI network of the VEGF-A, TGF-β2, and MAGE-A3 was reconstructed using the ZS Revelen online database J . The correlation between the immune marker sets of immune cells and VEGF-A K , TGF-β2 L , and MAGE-A3 M expression in LUAD The properties of selected epitopes for VEGF-A, TGF-β2, and MAGE-A3 proteins by different bioinformatics web tools are shown in Table 1 . Based on this table, Finally, a merged peptide SVRGKGKGQKRKRKKSKKKHHMVKISGGPHISYPPKKKRLESQQTNRRKKRALD) was selected as a multi-epitope peptide vaccine. Besides, the PSIPRED server was utilized to predict the secondary structure (i.e., α-helix, β-strand, and random coil) of the final selected peptide structure. Figure 3 A, B, C, and D present the 3D structure of VEGF-A, TGF-β, and MAGE-A3, selected merged peptides, respectively. Figure 3 E illustrates the detailed 2D structure of the final multi-epitope peptide. Fig. 3 The 3D structure of VEGF-A A , TGF-β B , MAGE-A3 C , and the final merged peptide D . The β-strands, the α-helix, and the random coils of the selected merged peptide are shown in yellow, pink, and gray colors E . The solubility of the multi-epitope peptide structure according to QuerySol was greater than 0.45, which showed that it has good solubility F The 3D structure of VEGF-A A , TGF-β B , MAGE-A3 C , and the final merged peptide D . The β-strands, the α-helix, and the random coils of the selected merged peptide are shown in yellow, pink, and gray colors E . The solubility of the multi-epitope peptide structure according to QuerySol was greater than 0.45, which showed that it has good solubility F Table 1 The properties of selected epitopes for each target protein by different bioinformatic web-tools Protein target Linear epitope Start Position Antigenicity Allergenicity Toxicity MHCI MHCII Prohibited/preferred areas Surface/non surface VEGF-A AGPGRASETMNFLLSW 172 Probable antigen 0.675 NON-ALLERGEN Non-Toxin 30 14 - Outside of the mature protein range MRCGGCCNDEGLECVP 261 Probable antigen 0.960 NON-ALLERGEN Toxin 8 1 Platelet-derived and VEGF family surface FKPSCVPLMRCGGCCN 253 Probable antigen 0.674 NON-ALLERGEN Toxin 10 6 Platelet-derived and VEGF family surface RGSASRAGPGRASETM 166 Probable antigen 0.695 NON-ALLERGEN Non-Toxin 7 3 - Outside of the mature protein range KWSQAAPMAEGGGQNH 202 Probable antigen 1.325 NON-ALLERGEN Non-Toxin 5 4 - surface DIFQEYPDEIEYIFKP 240 Probable antigen 1.155 NON-ALLERGEN Non-Toxin 22 19 Platelet-derived and VEGF family Non surface CECRPKKDRARQEKKS 308 Non antigen 0.481 NON-ALLERGEN Non-Toxin 4 7 Cysteine knot motif surface SFLQHNKCECRPKKDR 301 Non antigen −0.15 NON-ALLERGEN Toxin 11 4 Platelet-derived and VEGF family surface KGQKRKRKKSRYKSWS 329 Probable antigen 1.722 NON-ALLERGEN Non-Toxin 8 10 - surface * ITMQIMRIKPHQGQHI 282 Probable antigen 0.583 NON-ALLERGEN Non-Toxin 21 10 Platelet-derived and VEGF family surface SVRGKGKGQKRKRKKS 323 Probable antigen 2.262 NON-ALLERGEN Non-Toxin 5 4 - surface PHQGQHIGEMSFLQHN 291 Non antigen 0.445 ALLERGEN Non-Toxin 18 7 Platelet-derived and VEGF family - DVYQRSYCHPIETLVD 225 Probable antigen 0.840 NON-ALLERGEN Non-Toxin 43 9 - surface EVVKFMDVYQRSYCHP 219 Probable antigen 0.634 NON-ALLERGEN Toxin 27 18 - surface YCHPIETLVDIFQEYP 231 Probable antigen 0.975 ALLERGEN Non-Toxin 33 9 receptor binding interface/Cysteine knot motif surface ARQEKKSVRGKGKGQK 317 Probable antigen 1.988 ALLERGEN Non-Toxin 4 4 - surface MAEGGGQNHHEVVKFM 209 Probable antigen 1.666 NON-ALLERGEN Non-Toxin 13 5 - surface * EGLECVPTEESNITMQ 270 Probable antigen 0.523 NON-ALLERGEN Non-Toxin 7 2 receptor binding interface/Platelet-derived and VEGF family surface RKKSRYKSWSVYVGAR 335 Probable antigen 0.989 NON-ALLERGEN Non-Toxin 31 23 - surface ALLLYLHHAKWSQAAP 193 Non antigen 0.469 ALLERGEN Non-Toxin 26 23 - surface KARQLELNERTCRCDK 394 Probable antigen 1.034 NON-ALLERGEN Non-Toxin 8 4 VEGF heparin-binding domain surface CSCKNTDSRCKARQLE 384 Probable antigen 1.131 NON-ALLERGEN Non-Toxin 8 3 VEGF heparin-binding domain surface TTGF-β QKTIKSTRKKNSGKTP 292 Probable antigen 1.29 PROBABLE ALLERGEN Non-Toxin 9 5 - surface FAGIDGTSTYTSGDQK 278 Probable antigen 0.86 PROBABLE ALLERGEN Non-Toxin 21 1 - Non surface HCPCCTFVPSNNYIIP 253 Probable antigen 0.695 NON-ALLERGEN Toxin 38 10 tgfb-propeptide surface TSTYTSGDQKTIKSTR 284 Probable antigen 1.09 NON-ALLERGEN Non-Toxin 17 3 - surface RLESQQTNRRKKRALD 318 Probable antigen 0.842 NON-ALLERGEN Non-Toxin 9 2 Disulfide bond/ Domain surface TRKKNSGKTPHLLLML 298 Probable antigen 1.165 NON-ALLERGEN Non-Toxin 12 7 - surface HLLLMLLPSYRLESQQ 308 Non Antigen 0.297 NON-ALLERGEN Non-Toxin 50 9 Product داخلی TNRRKKRALDAAYCFR 324 Probable antigen 0.646 NON-ALLERGEN Non-Toxin 36 5 Disulfide bond/product Surface HEPKGYNANFCAGACP 364 22 Probable antigen 0.663 NON-ALLERGEN Non-Toxin 15 18 Polypeptide substrade binding site Surface LGWKWIHEPKGYNANF 358 16 Non Antigen −0.020 NON-ALLERGEN Non-Toxin 20 14 Putative homodimer interface Non Surface CPYLWSSDTQHSRVLS 378 36 Non Antigen 0.30 NON-ALLERGEN Non-Toxin 24 7 Domain Surface PEASASPCCVSQDLEP 400 58 Probable antigen 0.50 PROBABLE ALLERGEN Toxin 11 3 Putative homodimer interface Surface TILYYIGKTPKIEQLS 417 75 Probable antigen 1.114 PROBABLE ALLERGEN Non-Toxin 24 11 - Non Surface NFCAGACPYLWSSDTQ 372 30 Non Antigen 0.381 NON-ALLERGEN Non-Toxin 32 11 Putative homodimer interface Non Surface DNCCLRPLYIDFKRDL 343 1 Probable antigen 0.682 NON-ALLERGEN Toxin 21 12 Domain Surface TPKIEQLSNMIVKSCK 425 83 Probable antigen 1.294 PROBABLE ALLERGEN Non-Toxin 17 15 - Surface SRVLSLYNTINPEASA 389 47 Probable antigen 0.914 PROBABLE ALLERGEN Non-Toxin 27 10 Putative homodimer interface Non surface PLYIDFKRDLGWKWIH 349 7 Non Antigen 0.246 NON-ALLERGEN Non-Toxin 18 10 Domain Surface CPVVTTPSGSVGSLCS 118 Probable antigen 0.807 NON-ALLERGEN Non-Toxin 11 1 tgfb-propeptide ؟ VYKIDMPPFFPSETVC 103 Probable antigen 0.584 NON-ALLERGEN Non-Toxin 27 10 tgfb-propeptide Non Surface TSPTQRYIDSKVVKTR 206 Non Antigen 0.395 NON-ALLERGEN Non-Toxin 16 6 tgfb-propeptide Surface DVSAMEKNASNLVKAE 161 Probable antigen 0.648 NON-ALLERGEN Non-Toxin 13 9 tgfb-propeptide Surface PSGSVGSLCSRQSQVL 124 Probable antigen 0.590 PROBABLE ALLERGEN Non-Toxin 11 2 tgfb-propeptide - EGEWLSFDVTDAVHEW 223 Probable antigen 0.714 NON-ALLERGEN Non-Toxin 22 13 tgfb-propeptide Surface KASRRAAACERERSDE 81 Probable antigen 0.524 PROBABLE ALLERGEN Non-Toxin 5 3 tgfb-propeptide Non Surface YFRIVRFDVSAMEKNA 154 Probable antigen 0.885 NON-ALLERGEN Non-Toxin 23 13 tgfb-propeptide Surface IPPTFYRPYFRIVRFD 146 Probable antigen 1.275 NON-ALLERGEN Non-Toxin 37 26 tgfb-propeptide/ glyosylation Surface PFFPSETVCPVVTTPS 110 Non Antigen 0.376 PROBABLE ALLERGEN Non-Toxin 25 3 tgfb-propeptide Surface EFRVFRLQNPKARVPE 176 Probable antigen 1.414 PROBABLE ALLERGEN Non-Toxin 18 13 tgfb-propeptide Non Surface LCSRQSQVLCGYLDAI 131 Probable antigen 0.578 PROBABLE ALLERGEN Non-Toxin 27 9 tgfb-propeptide - RDLLQEKASRRAAACE 75 Probable antigen 0.565 PROBABLE ALLERGEN Non-Toxin 6 7 tgfb-propeptide Non Surface RVPEQRIELYQILKSK 188 Probable antigen 0.537 NON-ALLERGEN Non-Toxin 26 19 tgfb-propeptide Surface SNLVKAEFRVFRLQNP (141) 170 Probable antigen 1.025 NON-ALLERGEN Non-Toxin 35 18 tgfb-propeptide Surface MAGE-A3 QGASSLPTTMNYPLWS 66 Probable antigen 0.596 NON-ALLERGEN Non-Toxin 29 4 Disordere/ Compositional Bias Surface LGLSYDGLLGDNQIMP 182 Non Antigen 0.290 ALLERGEN Non-Toxin 8 1 "MAGE"Family/Domain Surface VLEVFEGREDSILGDP 227 Probable antigen 0.842 ALLERGEN Non-Toxin 6 1 "MAGE"Family/Domain Surface PTTMNYPLWSQSYEDS 72 Probable antigen0.509 ALLERGEN Non-Toxin 21 6 Disordered/ Compositional Bias Surface NQEEEGPSTFPDLESE 89 Non Antigen 0.423 NON-ALLERGEN Non-Toxin 4 0 Disordered/ Compositional Bias Surface HFVQENYLEYRQVPGS 249 Non Antigen −0.073 NON-ALLERGEN Non-Toxin 24 11 "MAGE"Family/Domain - GREDSILGDPKKLLTQ 233 Probable antigen 0.861 NON-ALLERGEN Non-Toxin 11 1 "MAGE"Family/Domain Surface DCAPEEKIWEELSVLE 214 0.224 NON-ALLERGEN Non-Toxin 13 5 "MAGE"Family/Domain Surface GLLGDNQIMPKAGLLI 188 Non Antigen −0.218 ALLERGEN Non-Toxin 24 5 "MAGE"Family/Domain Surface LIIVLAIIAREGDCAP 202 Non Antigen −0.221 NON-ALLERGEN Non-Toxin 6 21 "MAGE"Family/Domain Non Surface TKAEMLGSVVGNWQYF 131 Probable antigen 0.837 ALLERGEN - Non-Toxin 28 9 "MAGE"Family/Domain Surface FGIELMEVDPIGHLYI 162 Non Antigen 0.092 NON-ALLERGEN Non-Toxin 26 8 "MAGE"Family/Domain Non Surface SVVGNWQYFFPVIFSK 138 Probable antigen 1.216 ALLERGEN Non-Toxin 47 31 "MAGE"Family/Domain Surface IIAREGDCAPEEKIWE 208 Non Antigen 0.397 NON-ALLERGEN Toxin 2 6 "MAGE"Family/Domain Non Surface QIMPKAGLLIIVLAII 194 Non Antigen −1.16 NON-ALLERGEN Non-Toxin 23 28 "MAGE"Family/Domain Surface LESEFQAALSRKVAEL 101 Non Antigen 0.430 ALLERGEN Non-Toxin 23 5 "MAGE"Family Surface KASSSLQLVFGIELME 153 Non Antigen −0.119 ALLERGEN Non-Toxin 26 10 "MAGE"Family/Domain Surface ALSRKVAELVHFLLLK 108 Non Antigen 0.264 ALLERGEN Non-Toxin 45 21 "MAGE"Family Surface FPVIFSKASSSLQLVF 147 Non 0.386 NON-ALLERGEN Non-Toxin 23 16 "MAGE"Family/Domain Surface HHMVKISGGPHISYPP 288 Probable antigen 0.927 NON-ALLERGEN Non-Toxin 19 2 Domain Surface Helix TEEQEAASSSSTLVEV 33 Probable antigen0.721 NON-ALLERGEN Non-Toxin 12 1 Disordered Surface The properties of selected epitopes for each target protein by different bioinformatic web-tools Probable antigen 0.675 Probable antigen 0.960 Probable antigen 0.674 Probable antigen 0.695 Probable antigen 1.325 Probable antigen 1.155 Non antigen 0.481 Non antigen −0.15 Probable antigen 1.722 * ITMQIMRIKPHQGQHI Probable antigen 0.583 Probable antigen 2.262 Non antigen 0.445 Probable antigen 0.840 Probable antigen 0.634 Probable antigen 0.975 Probable antigen 1.988 Probable antigen 1.666 * EGLECVPTEESNITMQ Probable antigen 0.523 Probable antigen 0.989 Non antigen 0.469 Probable antigen 1.034 Probable antigen 1.131 Probable antigen 1.29 Probable antigen 0.86 Probable antigen 0.695 Probable antigen 1.09 Probable antigen 0.842 Disulfide bond/ Domain Probable antigen 1.165 Non Antigen 0.297 Probable antigen 0.646 364 22 Probable antigen 0.663 358 16 Non Antigen −0.020 378 36 Non Antigen 0.30 400 58 Probable antigen 0.50 417 75 Probable antigen 1.114 372 30 Non Antigen 0.381 343 1 Probable antigen 0.682 425 83 Probable antigen 1.294 389 47 Probable antigen 0.914 349 7 Non Antigen 0.246 Probable antigen 0.807 Probable antigen 0.584 Non Antigen 0.395 Probable antigen 0.648 Probable antigen 0.590 Probable antigen 0.714 Probable antigen 0.524 Probable antigen 0.885 Probable antigen 1.275 tgfb-propeptide/ glyosylation Non Antigen 0.376 Probable antigen 1.414 Probable antigen 0.578 Probable antigen 0.565 Probable antigen 0.537 SNLVKAEFRVFRLQNP (141) Probable antigen 1.025 Probable antigen 0.596 Disordere/ Compositional Bias Non Antigen 0.290 Probable antigen 0.842 Disordered/ Compositional Bias Non Antigen 0.423 Disordered/ Compositional Bias Non Antigen −0.073 Probable antigen 0.861 Non Antigen −0.218 Non Antigen −0.221 Probable antigen 0.837 Non Antigen 0.092 Probable antigen 1.216 Non Antigen 0.397 Non Antigen −1.16 Non Antigen 0.430 Non Antigen −0.119 Non Antigen 0.264 Non 0.386 Probable antigen 0.927 Surface Helix The peptide target’s molecular weight, as determined by the ProtParam server is 6317.52 g/mol, with an isoelectric pH of 11.92, indicating an alkaline composition. The peptide target comprises 22 residues, the estimated half-life of 1.9 h in mammalian reticulocytes, in vitro. The peptide target structure is composed of 926 atoms and its chemical formula is C 273 H 482 N 100 O 70 S 1 . The degree of hydration, or average hydrophilicity, closely correlates with protein solubility. With a predicted average GRAVY index of −1.93, the multi-epitope peptide exhibits a polar nature, effectively interacts with water, demonstrates high solubility, and shows cases of the lowest level of hydrophobicity. The solubility of the designed peptide target construct exceeded 0.45 as per QuerySol. The results of this analysis are visually presented in Fig. 3 F. Here, we used the C-immsim server to analyze and simulate the immune response to our vaccine. The perfect results were obtained from the server. In the first dose, an increase in IgM antibodies was observed, in the second dose an increase in IgG1 and IgG2, and in the third dose an increase in the concentration of all immunoglobulins. Also, an increase in IFN-γ, IL-2, Th (helper) cells, and the activated Tc cell population was observed. (Supplementary 4). The characterization of the nanoliposomes containing multi-epitope peptide was done by DLS (Fig. 4 A) and TEM (Fig. 4 B). We found that the size distribution of synthesized nanoliposomes was 65–190 nm, with a mean size of 110 nm and mean zeta potential + 30 mV. Also, the loading of multi-epitope peptides into nanoliposomes was near 98%, and only 2% were free. The stability test showed that the nanoliposomes containing multi-epitope peptide (10 mg/mL, 100 mg/mL, and with no peptide) had the same pattern. Based on their mean size and Zeta potential, all formulations were stable during 4 weeks (Supplementary 5). Fig. 4 The characterization of the nanoliposomes containing selected merged peptides by TEM A and DLS B . The size distribution of synthesized nanoliposomes was 65-190 nm, with a mean size of 110 nm and mean zeta potential +30 mV The characterization of the nanoliposomes containing selected merged peptides by TEM A and DLS B . The size distribution of synthesized nanoliposomes was 65-190 nm, with a mean size of 110 nm and mean zeta potential +30 mV Figure 5 A illustrates the serum titration curve of vaccinated mice with nanoliposomes containing multi-epitope peptide. This test revealed that IgG antibodies against the selected peptide were produced in both doses, 10 mg/ml and 100 mg/ml. Even up to 10,000 times dilution, the antibodies remained statistically significant ( P  ≤ 0.05). Based on our preliminary data, we tested logarithmic concentrations of vaccine, including 100, 10, 1, 0.1, and 0.01 mg/mL. We found that antibody titers had no significant difference at concentrations below 10 mg/ml. Also, the concentrations above 100 mg/mL were unstable and precipitated. Fig. 5 The serum titration curve A . Here, mice were immunized 3 times with nanoliposomes at doses of 10 mg/ml and 100 mg/ml. This test showed that in both doses, IgG antibodies were produced against selected peptides and significantly increased even up to 1:10,000 dilution. Cytokine analysis shows high levels of interleukin 4 B , interleukin 6 C , interleukin 10 D , tumor necrosis factor E , and interferon-gamma F in mice vaccinated with 10 mg/mL and 100 mg/mL of peptide vaccine. * means a significant difference compared with control at P<0.05, n=6, the sample size was n=6 mice per each group The serum titration curve A . Here, mice were immunized 3 times with nanoliposomes at doses of 10 mg/ml and 100 mg/ml. This test showed that in both doses, IgG antibodies were produced against selected peptides and significantly increased even up to 1:10,000 dilution. Cytokine analysis shows high levels of interleukin 4 B , interleukin 6 C , interleukin 10 D , tumor necrosis factor E , and interferon-gamma F in mice vaccinated with 10 mg/mL and 100 mg/mL of peptide vaccine. * means a significant difference compared with control at P<0.05, n=6, the sample size was n=6 mice per each group Figures 5 B-Fshow the serum levels of of various cytokines including, interleukin 4 (A), interleukin 6 (B), interleukin 10 (C), tumor necrosis factor (D), and interferon-gamma (E) in mice that vaccinated by the peptide vaccine three times at doses of 10 mg/ml and 100 mg/ml. Both vaccinated groups showed a significant increase in all investigated cytokines compared to the control group. Moreover, vaccinated mice with a dose of 100 mg/ml exhibited significantly higher levels of the cytokines than those receiving a dose of 10 mg/ml. Figure 6 A and B show the percentage of cell viability and cell apoptosis, respectively, in the A549 cells after 24 and 48 h of exposure to serum of vaccinated mice at the doses of 10 mg/ml and 100 mg/ml. Our results showed that the amount of cell viability attenuated at both doses and exposure times, while the number of apoptotic cells elevated at both doses and exposure times. Considering the cell viability and apoptosis, a significant difference has been seen between doses of 100 mg/ml and 10 mg/ml ( P  ≤ 0.05). Fig. 6 The result of MTT assay A and apoptosis assay B in A549 cancer cells after 24 and 48 h of exposure to diluted serum of treated mice. This test showed that the amount of cell viability decreased and the number of apoptotic cells was increased in both doses and in both exposure times. The expression of Bcl2   C and Bax   D in A549 cells after exposure to the diluted serum of vaccinated mice for 24 and 48 h. This test showed the level of Bcl2 was decreased and the level of BAX was increased in both doses and in both exposure times. The results of the Humanized PDX model of lung cancer, immunized with 10 mg/mL and 100 mg/mL of the novel liver cancer peptide vaccine based on nanoliposomes Cancer size E and volume F significantly decreased in the mice immunized with 10 mg/mL and 100 mg/mL in week 3 and later. * means significant difference compared with 10 mg/ml at P  < 0.05. The sample size was n  = 6 mice per each group The result of MTT assay A and apoptosis assay B in A549 cancer cells after 24 and 48 h of exposure to diluted serum of treated mice. This test showed that the amount of cell viability decreased and the number of apoptotic cells was increased in both doses and in both exposure times. The expression of Bcl2   C and Bax   D in A549 cells after exposure to the diluted serum of vaccinated mice for 24 and 48 h. This test showed the level of Bcl2 was decreased and the level of BAX was increased in both doses and in both exposure times. The results of the Humanized PDX model of lung cancer, immunized with 10 mg/mL and 100 mg/mL of the novel liver cancer peptide vaccine based on nanoliposomes Cancer size E and volume F significantly decreased in the mice immunized with 10 mg/mL and 100 mg/mL in week 3 and later. * means significant difference compared with 10 mg/ml at P  < 0.05. The sample size was n  = 6 mice per each group Figure 6 C and D illustrate the expression levels of the Bcl2 and Bax , respectively, in the A549 lung cancer cells, after exposure to the serum of vaccinated mice for 24 and 48 h. The results indicate a decrease in Bcl2 gene expression at both doses and exposure times, while Bax gene expression increased under the same condition. In the case of the Bcl2 gene, there is a significant difference between the Bcl2 gene expression and the dose of the vaccine. Regarding the Bax gene, there is also a significant difference in the increase of Bax gene expression at two doses of vaccine. Besides, the ratio of Bax expression to Bcl2 expression was calculated. Our findings revealed that exposure treatment significantly increased the pro-death/anti-death ratio, particularly at a concentration of 100 mg/ml compared to 10 mg/ml. Figure 6 E and F show the results of the Humanized PDX model of lung cancer, immunized with 10 mg/mL and 100 mg/mL of the liver cancer peptide vaccine based on nanoliposomes. As can be seen, cancer volume has increased with increasing time in the control group. On the other hand, this parameter significantly decreased in the mice immunized with 10 mg/mL or 100 mg/mL in week 3 and later. To check immune responses, the levels of cytokines, including interferon-gamma, interleukin 10 and interleukin 6, interleukin 4, and tumor necrosis factor-alpha, were investigated by ELISA technique in all groups. The significant increase of all mentioned cytokines was seen ( P  < 0.05) in the vaccinated PDX model, when compared with PBS-injected group, irrelevant peptide- injected group, and the non-humanized mice (Supplementary 6). Importantly, we checked TME and Immune Profiling (Supplementary 7) and saw the significant increased number of CD8 + T cells, Treg cells, and CD45RA(+) CCR7 (-) cells in vaccinated mice during 5 weeks compared with PBS-injected group, irrelevant peptide- injected group, and the non-humanized mice ( P  < 0.05). Also, we found that the quantity of MDSCs and level of immune checkpoints (PD-1 and CTLA-4) was significantly decreased in vaccinated mice during 5 weeks compared with PBS-injected group, irrelevant peptide- injected group, and the non-humanized mice ( P  < 0.05). Based on histopathology data (Supplementary 8), although the level of Ki-67 and cleaved caspase-3 markers in tumor tissue was significantly decreased in vaccinated mice ( P  < 0.05), the level of these markers was not high in liver or kidney, indicating no toxicity.

The selected peptide was synthesized by Proteogenix, France. 2,3-dioleyloxy-N-[2-(sperminecarboxamido)ethyl]-N, N-dimethyl-1-propanaminium (DOSPA), 1,2-Dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE), MTT (3- (4, 5-dimethylthiazol-2-yl)−2, 5-diphenyltetrazolium bromide) were purchased from Sigma-Aldrich. RNA extraction buffer, cDNA synthesis kit, and SYBR ® Green Real-Time PCR Master Mix were purchased from Pars tous company, Iran. Cytokines ELISA kit including interferon-gamma, interleukin 10 and interleukin 6, interleukin 4, and Tumor Necrosis Factor-alpha were purchased from Bio Legend Company, USA. Mouse Antibody titration kits were purchased from Bio Legend Company, USA. DMEM cell culture media was purchased from Bio Idea Company, Iran. Fetal Bovine serum (FBS), Trypsin/EDTA, and Penicillin/Streptomycin (Pen/Strep) were purchased from Gibco, USA. Primers were designed and synthesized by SinaClon company, in Iran. The whole process of this study is summarized in the flow chart (Fig. 1 ). Fig. 1 The flow chart of the study The flow chart of the study Here, antigen selection was based on 4 parameters, including (1) Gene analysis by UALCAN ( https://ualcan.path.uab.edu/ ), (2) The co-expression network or protein-protein interactions [ 38 ] by Zs Revelen server ( https://revelen.zsservices.com/ ), (3) The correlation between selected genes expressions and Tumor purity by TIMER database ( http://timer.cistrome.org/ ), and (4) the previous studies and their importance of selected genes in lung tumor [ 39 – 41 ]. The additional details about the specific bioinformatics criteria used for target selection were considered at following sections. We utilized the UALCAN web resource ( https://ualcan.path.uab.edu/ ) to conduct a comprehensive analysis of Lung cancer OMICS data, focusing on the VEGF-A , TGF-β2 , and MAGE-A3 genes. Through this platform, we accessed a wealth of information encompassing diverse parameters, contributing to a more nuanced understanding of the role and implications of the selected genes in NSCLC. The obtained results include insights into gene expression, survival rates, pan-cancer views, and histological subtypes of the datasets. We performed bioinformatic analyses in NSCLC patients with lung adenocarcinoma (LUAD) and conducted Patient survival determination and meta-analysis to check the clinical significance of the selected genes. We also investigated the expression of target genes in several human tissues using the HPA database. For statistical analysis and visualization of gene expression in cancer, we obtained RNAseq data from the pan-cancer project in The Cancer Genome Atlas (TCGA) database. The Zs Revelen ( https://revelen.zsservices.com/ ) is an extensive online database designed to aggregate, evaluate, and integrate all publicly available information on PPI. This data is further enhanced with computational predictions of potential functions. The Zs Revelen was utilized to perform a PPI network analysis of VEGFA , TGF-β2 , and MAGE-A3 , revealing their interactions with other proteins. The resulting PPI network was then visually represented using Cytoscape software (version 3.6.0). Through comprehensive enrichment analyses focusing on biological processes, pathways, and diseases, we obtained valuable insights. Correlation between VEGFA , TGF-β2 , MAGE-A3 expressions, and Tumor purity. The TIMER ( http://cistrome.org/TIMER/ ) database encompasses TCGA samples from NSCLC cancer types and facilitates a comprehensive assessment of immune infiltration levels. Using the TIMER algorithm, we investigated the association between VEGFA , TGF-β2 , and MAGE-A3 expression and the abundance of six distinct categories of immune cells, including B cells, neutrophils, myeloid dendritic cells, macrophages, CD8 + T cells and CD4 + T cells in LUAD. First, the sequence of protein targets was extracted from the NCBI database. Based on their diverse variants, the most overlapping position of the isoforms was chosen by protein Blast. Then the FASTA sequence: (Query ID belonging to VEGF: NP_001020537.2 , TGF-β: NP_001129071.1 , MAGE-A3: NP_005353.1 ) of the parts of the protein that had the most overlap was selected and their sequences were given to the ABCpred server ( https://webs.iiitd.edu.in/raghava/abcpred/ABC_submission.html ). The peptide’s antigenic power was checked by the Vaxijen server ( http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html ). Then, with the AllerTOP server ( http://ddg-pharmfac.net/AllerTOP/ ), the allergenicity of the epitope sequences was assessed. In the next step, the toxicity of epitopes was evaluated using the ToxinPred server ( http://webs.iiid.edu.in/raghava/toxinpred/design.php ), and non-toxic sequences were selected. Afterward, the VaxiTop server ( http://www.violinet.org/vaxign/vaxito/index.php ) was used to detect epitope recognition by class MHC I/II. Moreover, by using the UniProt database, the exact location of each peptide was determined on the 3D structure file of the desired protein targets. Finally, three peptide sequences were selected for MAGE-A3, VEGF-A, and TGF-β2. These peptides were merged by two KKK linkers. In the field of bioinformatics, accurately predicting protein secondary structure is vital. This structure consists of regular elements like α-helices and β-strands, as well as irregular coil regions. Our study utilized the widely-recognized PSIPRED tool, which employs neural networks to analyze PSI-BLAST output for predicting protein secondary structure ( http://bioinf.cs.ucl.ac.uk/psipred ). The physicochemical properties of the final merged peptide, such as molecular weight, isoelectric pH, hydropathic index, and estimated half-life, were carefully analyzed using the ProtParam server. In the design of epitopes, it is essential to consider the solubility of target proteins. The scaled solubility value (QuerySol) predicts solubility, with values above 0.45 indicating higher solubility than the average soluble E.coli protein from the Niwa et al. 2009 dataset, and values below 0.45 predicting lower solubility, as per the Protein-Sol database. Also, we used the C-immsim server ( http://150.146.2.1/C-IMMSIM/index.php ) to analyze and simulate the immune response to our vaccine [ 42 ]. Here, the injection profile of the vaccine was set, and the 4-week period is the minimum suggested time. All the default simulation parameters were used. The time steps of injection were specified at 1, 84, and 168. The simulation volume of the vaccine (containing no LPS) injection was set at 10,000. The final merged peptide, SVRGKGKGQKRKRKKSKKKHHMVKISGGPHISYPPKKKRLESQQTNRRKKRALD, was first synthesized by the solid phase method. Then, 1 mL PBS buffer was added to a synthesized peptide (10 mg) and completely mixed. Afterward, 0.5 mL of peptide solution and 0.5 mL of cationic lipid mixture (DOSPA: DOPE, 3:1) were vigorously mixed and then centrifuged (Eppendorf, Germany). After discarding the supernatant, 1mL of PBS buffer was added and mixed. The amount of peptide loading into nanoliposomes was determined by measuring the optical density of the supernatant at 280 nm by a NanoDrop (Thermo Fisher, USA). Furthermore, characterization tests including, DLS, Zetasizer (Malvern Panalytical,), and TEM, (Ziesis, Germany) were done to verify the size distribution and shape of nanoliposomes. The stability assay was done for nanoliposome formulations in physiological buffer (PBS) and 50% FBS during 1–4 weeks by measuring their mean size by DLS (Malvern Panalytical,) and mean Zeta potential by Zetasizer (Malvern Panalytical,). The Balb/c mice with an average age of 1 month and a weight of about 10–12 g were thoughtfully chosen from the research center of the Pasteur Institute of Iran. They were acclimated to their new environment for a week under optimal conditions, including a temperature range of 20–24 °C, a 12:12 Light-dark cycle, appropriate humidity levels, and continuous access to water and feed. They were then randomly distributed into 2 groups receiving the peptide vaccine with doses of 10 and 100 mg/ml. Here, 100 µl of vaccine were injected into the mice, and this process was repeated every seven days for three consecutive weeks. In the fourth week, the whole blood was collected from the heart tissue of vaccinated mice, and its serum was separated for further study. In the control group, mice were treated with PBS. The sample size was n  = 6 mice per each group. Anesthesia method was using isoflurane 5% and the euthanasia method was using over dose of sodium pentobarbital intravenously under the Guideline for the Care and Use of Laboratory Animals and approved by ARRIVE guidelines ( https://arriveguidelines.org ). ELISA technique was used to determine the antibody titer produced after vaccine injection. For this purpose, the wells of the 96-well plate were separately coated with the synthesized peptide at a concentration of 100 mg/ml and then covered with a 5% skim milk solution. Next, the serum serial concentrations (1:10, 1:100, 1:1000, and 1:10000) were prepared and exposed to the coated wells. Mice antibody conjugated with HRP enzyme was then added, and after incubation and washing, the color change was read. The absorbance of the samples was evaluated at a wavelength of 630 nm to determine the produced antibody titer based on the optical density of the control (cutoff). The levels of cytokines, including interferon-gamma, interleukin 10 and interleukin 6, interleukin 4, and tumor necrosis factor-alpha, in the serum of vaccinated mice were investigated by ELISA technique. First, 50 µl of diluted mouse serum (1:10) were added to the ELISA kit wells and incubated at 37 °C in a Memmert incubator (Germany) for 30 min. After washing the wells, the conjugate solution was added and incubated for another 30 min at 37 °C. In the next step, solution A and then solution B were added to the wells and the microplates were again placed in an incubator at 37 °C for 30 min. The absorbance of the wells was measured at a wavelength of 450 nm with an ELISA reader (Mindray, China). Finally, the concentrations of cytokines were determined based on the standard curve. The lung cancer cell line (A549) obtained from the Iranian stem cell bank was cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) with 10% fetal bovine serum (FBS), 1% (100 µg/ml) penicillin/streptomycin and 2 mM L-glutamine at 37 °C with 5% CO2. The cell culture medium was refreshed every three days. Upon reaching 70% density in T-25 flasks, the cells were detached from the flask using 0.25% trypsin/EDTA (1X) enzyme. They transferred to new flasks and the cell passage process was repeated three times to obtain a sufficient number of cancer cell lines. To study the effect of antibodies from vaccinated mice, approximately one hundred thousand cells were seeded in each plate after being counted using trypan blue and an Inverted Microscope (Optica, Italy). The effect of mice serum-containing antibodies on the survival of lung cancer cells (A549) was investigated using the MTT assay. The lung cancer cell line (A549) was cultured in DMEM medium and seeded into duplicate 96-well plates after reaching optimal growth. After 24 h, when the cells adhered to the plate and regained their morphology, the cells were treated with 50 µl of diluted mice serum (1:10). After 24 and 48 h, 20 µl of MTT dye solution (5 mg/mL) was added to each well and the plate was incubated for 3 h. Afterward, the supernatant was removed and 100 µl of dimethyl sulfoxide (DMSO) was added to each well. Following pipetting, the absorbance was measured at a wavelength of 570 nm using an ELISA reader. To calculate the cell viability percentage of the studied cells, the optical density was obtained from the test and control wells by the below equations. The PBS-injected group was considered a negative control. Finally, the inhibitory concentration of 50% (IC50) was calculated. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\text{C}\text{y}\text{t}\text{o}\text{t}\text{o}\text{x}\text{i}\text{c}\text{i}\text{t}\text{y}\%=\:1-\:\frac{\text{M}\text{e}\text{a}\text{n}\:\text{a}\text{b}\text{s}\text{o}\text{r}\text{b}\text{a}\text{n}\text{c}\text{e}\:\text{o}\text{f}\:\text{t}\text{e}\text{s}\text{t}\:\text{w}\text{e}\text{l}\text{l}}{\text{M}\text{e}\text{a}\text{n}\:\text{a}\text{b}\text{s}\text{o}\text{r}\text{b}\text{a}\text{n}\text{c}\text{e}\:\text{o}\text{f}\:\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}\:\text{w}\text{e}\text{l}\text{l}}\:\times\:100$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text{Viability} \%= 100-\text{Cytotoxicity}\%$$\end{document} To evaluate the number of apoptotic and necrosis cells following exposure to vaccinated mice serum, we employed flow cytometry with annexin-V and propidium iodide (PI). The cells were initially seeded in a plate and then exposed to diluted serum (1:10) from vaccinated mice for 24 and 48 h. After the incubation times, the cell suspension was separated with trypsin/EDTA, and washed in phosphate buffered saline (PBS), followed by buffer wash. Subsequently, the cells were incubated with 5 µg/ml of annexin V-FITC antibody at 4 °C for approximately 15 min in a dark environment, and then 10 µg/ml of propidium iodide dye was added. After 30 min, the percentage of apoptotic cells was then calculated using flow cytometry (BDbiosciences, USA). The PBS-injected group was considered a negative control. To extract the RNA, we utilized the RNA extraction kit (ParsTous, Iran) and meticulously followed the manufacturer’s instructions. After RNA extraction, the RNA concentration was assessed using a NanoDrop, and the OD260/OD280 ratio was quantitatively checked to ensure no contamination. To validate the quality of the extracted RNA, it was subsequently run on a 1.5% agarose gel. For cDNA synthesis, 5 micrograms of total RNA were mixed with 1 µl of DNase I enzyme, 10 µl of buffer-Mix, and 2 µl of RT enzyme 200 U/µl in an RNase-free microtube. The cDNA synthesis followed the ParsTous kit protocol, with a temperature cycle of 5 min at 25 °C, 60 min at 42 °C, and finally 5 min at 70 °C. Afterward, the obtained cDNA was stored at −20 °C and its concentration was measured using a NanoDrop before normalization. To perform Real-Time PCR with the Gentier (China), the Syber Green method was used to determine gene expression levels. The primer sequences for the Bcl2 (anti-apoptotic) and Bax (pro-apoptotic) genes, as well as the housekeeping gene GAPDH , were designed using Primer3 software and synthesized from Sinaclon company. The following primers and annealing temperatures were used for amplification are listed below; Bcl2 forward primer (F) 5´- GAGGCAGGCAGTAGTATGGTG-3ʹ and reverse primer (R) 5ʹ-AGGATAACGGAGGCTGGACA-3ʹ (60˚C); Bax : (F) 5ʹ-CCCCCGAGAGGTCTTTTT-3ʹ and (R) 5ʹ-GGAGGAAGTCCAATGTCCAG-3ʹ (60˚C); GAPDH : (F) 5ʹ-GGGAGCCAAAAGGGTCATCA-3ʹ and (R) 5ʹ-AGTGATGGCATGGACTGTGG-3ʹ (60˚C). Our procedure entailed combining 6 µl of cDNA, 0.5 µl of each forward and reverse primer, 0.5 µl of ROX dye, and 10 µl of the SYBR Green master mix. A temperature cycle was initiated with one stage at 95 °C for 5 min, followed by 40 cycles including 95 °C for 15 s, 60 °C for 30 s, and 72 °C for 1 min. Subsequently, a melting temperature curve was generated to confirm specificity and the absence of non-specific products. The test results were analyzed by the 2 ^−ΔΔCt method. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text{Fold change}=2^{-\triangle\triangle \text{Ct}}$$\end{document} where ΔΔCt= [(mCt target -mCt reference ) test sample - (mCt target -mCt reference ) control sample ] For the generation of a humanized PDX model (Fraunhofer Institute for Cell Therapy and Immunology IZI, Germany), NSCLC tissues were successfully grown in an NSG mouse. Here, lung tumors were dissected, and primary cultures were derived. They were harvested and minced into 1 mm 3 fragments. Pieces of the tumor tissue were grafted subcutaneously into humanized NSG mice using a previously described technique [ 43 , 44 ]. Tumor samples were obtained from 12 patients with lung cancer in 2024. All the patients provided signed informed consent. This study was approved by the Fraunhofer Institute for Cell Therapy and Immunology IZI, Germany. The clinical characterization of donor patients for PDX models is summarized in Supplementary 1. To establish PDX models, tumor samples were subcutaneously implanted into the flanks of NSG mice (Jackson Laboratory, Sacramento, CA, USA). To validate PDX models, they were characterized by different assays, such as Stroma content, vascularization, differentiation, take rate, PDL-1 expression, TIL expression, and doubling time (Supplementary 2). Then, mice were randomly distributed into 2 groups receiving, 10 mg/ml and 100 mg/ml of peptide vaccine. Here, 100 µl of peptide vaccine were injected every seven days for three consecutive weeks. Here, we considered irrelevant peptide and PBS control groups. The sample size was n  = 6 mice per each group. To take images and calculate tumor volume, they were killed using anesthetic and then the cancer mass was removed. The volume of the cancer was weekly determined [ 43 , 45 ]. The tumor volume was calculated as 0.5 × length × width 2 . This section of work is performed under the Guideline for the Care and Use of Laboratory Animals and approved by ARRIVE guidelines ( https://arriveguidelines.org ). To check immune responses, the levels of cytokines, including interferon-gamma, interleukin 10 and interleukin 6, interleukin 4, and tumor necrosis factor-alpha, in the serum of vaccinated mice were investigated by ELISA technique. Also, the quantification of CD8 + T cells, Tregs, MDSCs, and immune checkpoints (PD-1 and CTLA-4) and memory T-cell responses (CD45RA(+) CCR7 (-)) in humanized and non-humanized mice was done to check TME and Immune Profiling by flowcytometery. The quantification of CD8, Ki-67, cleaved caspase-3 markers in tumor, liver, and kidney in humanized and non-humanized mice at week 5 was done to assess toxicity based on histopathology data, IHC. Finally, The HLA profile of PDX models was determined (Supplementary 3). Anesthesia method was using isoflurane 5% and the euthanasia method was using over dose of sodium pentobarbital intravenously under the Guideline for the Care and Use of Laboratory Animals and approved by ARRIVE guidelines ( https://arriveguidelines.org ). The maximal tumor size/burden permitted by the ethics committee of the North Tehran Branch of Islamic Azad University. Also, the ethics committee confirmed that the maximal tumor size/burden was not exceeded. In this study, all statistical analyses were conducted using the Statistical Package for the Social Science software (SPSS v.22). Our analytical procedures encompassed calculating mean ratio (M) and Standard Derivation (SD). Additionally, we utilized two-way ANOVA with a significance level of P  ≤ 0.05 to compare the groups. It is important to note that each test was repeated six times, as there were 3 mice in each group and each test was performed twice.

Taken together, based on both in vitro and in vivo experiments such as antibody titration, cytokine tests, MTT, Annexin V/PI assays, apoptotic gene expression and humanized PDX model, this study demonstrated the impact of the multi-epitope vaccine on the immune system to inhibit the survival and proliferation of lung cancer cells. It must be mentioned that this research also had some limitations that should be addressed in future studies. One of the most important limitations was the small size of statistical population for both in vitro tests and PDX models. Another limitation was the use of only one type of lung cancer cell line (A549). We propose that the efficacy of this vaccine should be evaluated on other lung cancer cell lines. Another limitation of this study was that the selected final peptides were based on limited HLA alleles and should be considered more broadly to have greater coverage and be applicable to humans. Generally, development of a multi-epitope peptide for universal use is challenging, due to MHC limitations and the wide heterogeneity in MHC alleles. Finally, it is proposed that both efficacy and side effects of the multi-epitope vaccine should be evaluated in the future clinical trials.

The design and development of multi-epitope peptides against lung cancer is a new and promising research field. Lung cancer poses a significant health threat worldwide and requires novel approaches for treatment and prevention. Anti-lung cancer peptide targets strengthen the body’s immunity against cancer. Peptide targets activate the body’s immune system, stimulating the individual’s immune response to identify, attack, and destroy tumor cells. This process results in a balanced induction of the activation of CD8 + and CD4 + effective T and B cells [ 22 ]. In this study, we used advanced computational tools to analyze protein targets. Furthermore, our experimental phase involved the synthesis of nanoliposomes containing the selected epitopes, leading to higher antibody titers and elevated cytokine levels with increased peptide target doses. We observed a significant production of IgG antibodies against the selected peptides, even at dilutions of up to 10,000. Mice treated with 100 mg/ml of the peptide vaccine exhibited notably higher antibody levels than those vaccinated with 10 mg/ml. Both groups of mice showed a significant increase in all the investigated cytokines compared to the control. Moreover, our findings illustrated that the multi-epitope peptide up-regulated Bax expression and down-regulated Bcl2 expression, leading to reduced cell viability and increased apoptotic cells. These findings were further validated through tests on humanized mice and PDX models. For bioinformatics and target selection, we used different bioinformatics tools such as UALCAN, Zs Revelen, and TIMER to select the targets MAGE-A3, TGF-β2, and VEGF-A. The recent work by Abd El-Mawgoud et al. (2025) on the synthesis and molecular docking of pyrazole derivatives as targeted inhibitors of the PI3K/AKT and JAK/STAT pathways in lung cancer provided a broader context for how molecular-targeted therapies are advancing in the treatment of lung cancer. The potential synergistic effects of our multi-epitope vaccine in modulating tumor immunity can be addressed by the importance of signaling pathways in modulating cancer progression [ 46 ]. Here, we presented the synthesis of nanoliposomes as a vehicle for delivering multi-epitope vaccine. This is an important step in the design of vaccine formulations with enhanced stability and immunogenicity. The use of lipid-based nanoparticles in cancer therapy has been well explored, with promising results in various cancers. Abd-Rabou et al. (2025) showed efficacy of silymarin-functionalized selenium nanoparticles for inflammation-related pathways (PI3K/AKT/NF-κB) in cancer treatment and this supports the potential of nanomaterials for delivering immunotherapies [ 47 ]. The increased levels of cytokines, such as IL-4, IL-6, and TNF, suggest that the vaccine triggers a broad immune response. A relevant comparison could be made with the work of Abd-Rabou et al. (2025) in their exploration of adipose tissue mesenchymal stem cells conditioned media, which showed promising anti-inflammatory and anti-oxidant effects in cancer therapy. The study also involved modulation of immune responses that could potentially correlate with the findings in this manuscript, where cytokine profiles [ 48 ]. We found the reduced A549 cell viability and enhanced apoptosis by sera of vaccinated mice. We provided valuable data suggesting that the vaccine-induced immune response leads to tumor cell death. Mansour et al. (2024) worked on novel thiouracil-fused heterocyclic compounds in breast cancer cells. This study explores apoptosis through modulation of the Bax / Bcl-2 ratio, similar to the our findings on the gene expression [ 49 ]. The in vivo results in PDX models showing a significant reduction in tumor volume. We think that further analysis of immune cell infiltration and potential off-target effects must be checked in future to provide a more comprehensive picture of the vaccine’s therapeutic potential. Abd-Rabou et al. (2023) indicated the role of copper-tin nanocomposites in inducing apoptosis in skin cancer cells and it highlights the role of nanomaterials in mediating anti-cancer immune responses [ 50 ]. We think that this vaccine has the potential for a clinical trial, especially considering the current landscape of lung cancer immunotherapies. It may provide a parallel approach to treat lung cancer [ 51 ]. The physicochemical properties of our vaccine construct were comparable with other previously in silico-designed vaccine constructs for the cancer-testis antigen BORIS [ 52 ], antigens of SARS-CoV-2 [ 42 ], transmembrane protein 31 (TMEM31) [ 53 ], synaptonemal complex protein 1 (SYCP1), and acrosin binding protein (ACRBP) antigens [ 54 ], BORIS cancer-testis antigen [ 55 ]. Importantly, it is possible to clone the vaccine construct and to design DNA or mRNA vaccines based on the merged peptide, as previously reported for SYCP1 and ACRBP epitopes [ 56 , 57 ] and BORIS cancer-testis antigen [ 55 ]. Paal F. Brunsvig and colleagues (2006) conducted a phase I/II study to investigate the safety, tolerability, and clinical response of a vaccine combining two telomerase peptides, GV1001 and HR2822, in patients with lung cancer. The result indicated that the peptides were immunogenic and safe for use in NSCLC patients, and the induction of GV1001-specific immune responses led to tumor eradication [ 58 ]. Another project was done by YOSHIYAMA et al. in 2011 on peptide vaccines for the treatment of resistant lung cancer. Four peptides were selected according to pre-existing humoral immune responses for resistant NSCLC patients. Forty-one patients with refractory NSCLC who did not respond to chemotherapy and/or targeted therapy were treated with a vaccine. The average overall survival of patients was 304 days, with a one-year survival rate of 42%. Based on these results, the study proposes a feasible peptide vaccine for the treatment of resistant NSCLC [ 59 ]. In 2014, Lee M. Krug et al. conducted research on WT1 peptide vaccines that trigger T-cell immune responses against lung cancer. The study investigated the immunogenicity of a WT1 vaccine containing four class 1 and class 2 peptides in patients with chest neoplasm expressing WT1. The vaccine induced multifunctional T-cell responses and triggered the production of CD8 + cells. The study’s result demonstrated that the multivalent peptide vaccine (WT1) stimulated immune responses in a large number of patients with chest malignancy while causing minimal toxicity [ 60 ]. In 2001, Wei et al. investigated the first DNA-based cancer vaccine to evaluate the anti-tumoral effect of VEGF on three diverse tumor models and observed that the humoral immune response against VEGF can inhibit primary tumor growth [ 61 ]. Kamstock and co-workers developed another vaccine against VEGF and studied its efficacy in dogs with soft tissue sarcoma. In their research, they combined DNA-liposome complexes with human VEGF165, and the antibody titer showed a dramatic increase following immunization with the protein vaccine [ 62 ]. Kaumaya et al. (2010) developed a peptide vaccine that consisted of synthetic VEGF peptides as antigens and the measles virus T-cell epitope fused to protein (MVF) as an adjuvant. After evaluating the effect of this peptide vaccine in inhibiting the VEGFR2 signaling pathway [ 63 ], Wang and colleagues used this vaccine in a mouse model of ovarian cancer and detected high antibody titers against the peptides in primary tumor models, which was consistent with the inhibition of angiogenesis. Given the time limitation for inhibiting the possibility of metastasis, it seems that anti-angiogenic vaccination strategies, such as immunization against VEGF, may be promising after early-stage diagnosis [ 64 ]. Another clinical trial was performed on MAGE-A3, with patients divided into two groups: one receiving recombinant MAGE-A3 protein + adjuvant and the other receiving a placebo. The study underscored a 27% improvement in progression and disease-free survival in patients who received the vaccine. Although this investigation was not conclusive, the results are promising enough to warrant a phase III clinical study [ 65 ]. A MAGRIT study (phase III) was conducted by Vansteenkiste et al. in 2007 on patients with resected lungs and positive MAGE-A3. This trial assessed the recombinant MAGE-A3 vaccine combined with the AS15 immunostimulator. The comparison of the two groups receiving MAGE-A3 and placebo in this trial illustrated that the MAGE-A3 vaccine did not lead to increased patient survival without the presence of disease symptoms in melanoma cancer. The expression of MAGE-A3 antigen in the comparison groups was not significant, therefore, the study failed to reach its endpoint. Several factors may explain this outcome, including the potential need for stronger adjuvants to enhance the immune response and an effective delivery method to prevent probable vaccine degradation. Moreover, the patient selection process was insufficient, as it relied solely on MAGE-A3 expression without considering the overall immune status of patients. This approach is problematic because MAGE-A3-positive patients can exhibit varying immune responses, and not all cancer cells express MAGE-A3 consistently. Low levels of antigen expression complicate the immune system’s ability to effectively target these cells, leading to a diminished response. Additionally, the MAGE-A3 vaccine treatment protocol consisted of repeated injections over 27 months, which may have been inadequate for maintaining and enhancing the immune response over the long term. It’s possible that the golden period for a robust immune response was missed, or that the vaccine was not administered at the optimal dosage. Furthermore, genetic alterations in the tumor or diminished expression of MAGE-A3 antigens could allow cancer cells to evade the immune response. Since tumors can alter their antigen expression, concentrating solely on one specific antigen may have contributed to vaccine resistance. If the tumor stopped expressing MAGE-A3, the vaccine would consequently have lost its effectiveness. The MAGE-A3 vaccine was administered after surgery, when the tumor burden was either low or absent, potentially limiting immune system activation. The MAGE-A3 protein consists of 314 amino acids, many of which are allergenic, non-antigenic, or toxic, highlighting the necessity for careful peptide selection. As an intracellular protein, MAGE-A3 can evade detection by immune cells, which may contribute to the limited effectiveness observed in some vaccines aimed at it [ 5 , 39 , 66 ]. Although certain studies indicate that MAGE-A3-based vaccines have not produced significant results [ 39 , 67 ], this antigen remains vital in the context of lung cancer. A failed clinical trial does not necessarily indicate that the antigen is ineffective; often, it reflects the use of an unsuitable platform. Recent research has shown that T cells engineered with TCR targeting MAGE-A3 exhibited a positive response, suggesting that the immune system can effectively engage with MAGE-A3 when properly stimulated [ 68 ]. Furthermore, advancements in the selection of targeted peptides and the development of multi-epitope vaccines have significantly enhanced immunogenicity and therapeutic potential. By utilizing multiple epitopes, there is an increased chance that at least one will be expressed in patients, thereby minimizing the risk of tumors escaping the immune response. We hypothesized that perhaps combining this important antigen with two other important antigens (VEGF-A and TGF-β2) would be effective. In other words, it can be claimed that combining these three antigens and stimulating the immune system leads to better outcomes. We think that this designed vaccine effectively blocks the three important proteins in lung cancer. We think VEGF-A/TGF-β2 targeting alters TME immunosuppression (e.g., FasL expression and endothelial cell interactions), and this leads to cancer immunotherapy. Kristof Coppens and colleagues were another group in 2014 that searched for vaccination therapy for lung cancer. They found that the MAGE-A3 peptide vaccine effectively cured patients in the early stages (phase II) of cancer [ 69 ]. Phase II investigation about the Belagenpumatucel-L vaccine, which is a whole-cell vaccine that expresses the TGF-β2 antisense strand and reduces the expression of TGF-β2, was searched for NSCLC patients with stage II, IIIA, IIIB, and IV. A 15% relative response was observed in advanced-stage patients. The probability of one or two years surviving for patients receiving the low-dose vaccine was 20–39%, and for higher doses was 52–68%, respectively [ 24 , 70 ]. Based on these results, a phase III clinical trial was initiated to investigate overall survival improvement for 532 patients with T3N2-IIIA, IIIB, and IV stages who were treated with belagenpumatucel-L or placebo for about 4 to 17 weeks after the end of first-line chemotherapy. Although the phase III clinical trial did not reach its survival endpoint and the use of belagenpumatucel-L did not result in an improvement in overall survival compared to the placebo group, a preplanned subgroup analysis revealed that patients who received radiotherapy before vaccination showed a significant improvement in their overall survival [ 71 ]. In contrast to the above studies, the multi-epitope peptide in our research was carefully selected through a comprehensive evaluation of its properties using algorithms available in biological databases and immunogenic structures. Hence, as expected from bioinformatic analysis, the selected peptides were chosen based on their ability to stimulate immune cells and their potential to interact with human HLA alleles, ensuring they are antigenic and possess the best epitopes [ 72 , 73 ]. In our research, we designed a 54-mer peptide based on the evaluation in UniProt and other in-silico databases. This peptide was included from three conserved 16-amino acid segments of VEGF-A/TGF-β2/MAGE-A3, which were linked by two linkers of three lysines joined to each other. Multi-epitope peptide tends to significantly increase the possibility of targeting tumor cells that have lost certain TAA epitopes while preserving others. As a result, these peptides offer increased flexibility in dealing with antigen/epitope loss, leading to a tumor-specific immune response and potentially reducing vaccine resistance. To effectively protect peptides, we have incorporated them into a nanoliposome carrier. This nanoliposomal vaccine enhances immune responses by safeguarding peptides and improving their uptake by antigen-presenting cells (APCs). Besides, prior research showed that pH-sensitive liposomes formulated from DOPE promote the escape of antigens from endosomes, significantly boosting cross-presentation through the cytosolic pathway. Their larger surface area improves peptide efficacy by increasing solubility and bioactivity, and controlled release. According to studies, nanoliposomes are transforming biomedical applications due to their stability, which allows for precise cell targeting while protecting healthy tissues. Research suggests that nanoliposomal peptide vaccines generally outperform traditional options in eliciting antitumor responses [ 74 , 75 ]. Our results exemplify that the designed nanoliposomal peptide can activate the immune system by targeting antigens in various locations within the cells, including extracellular, intracellular, and cell surfaces. Innate immune cells, such as dendritic cells (DC) and macrophages, are believed to detect the extracellular and cell surface peptides, especially VEGF-A and TGF-β, triggering their activation. This immune response results in localized inflammation, leading to certain cell destruction and the release of internal cell contents, such as the MAGE-A3 peptide. This cascade not only amplifies local inflammation but also enhances immune responses. Our findings align with previous studies on nanoliposomal vaccines, underscoring the potential of our multi-epitope peptide to activate the immune system [ 74 , 76 ]. Comparing our results with related research emphasizes significant advancements in multi-epitope peptide treatments. There are several hypotheses concerning the mechanism of action of our vaccine. The foremost hypothesis posits that the vaccine stimulates the immune system, resulting in the production of antibodies that identify and block the protein markers: TGFβ, VEGFA, and MAGE-A3. By obstructing these targets, the vaccine blocks the markers present in the tumor microenvironment and also has the potential to inhibit angiogenesis. The second mechanism hypothesis posits that CD8 + cells recognize protein targets through T cell surface receptors, thereby inducing cellular immunity. Given that MAGE-A3 is located in the cell nucleus, VEGF is found on the surface, and TGF-β is found in the extracellular space, it is likely that this mechanism plays an important role. However, in the case of MAGE-A3, it seems that both antibodies and CD8 + cells can effectively penetrate the depths of the cell. However, for the other two protein targets located outside the cell and there is a strong hypothesis that antibodies or CD8 + immune cells are integral to the destruction of tumor cells. It is essential to recognize that although MAGE-A3 is categorized as an intracellular marker, its release during the death of cancer cells allows it to find MAGE-A3. This process enables the antibodies to find it and act effectively, leading to inflammation in the tumor microenvironment. Another possible mechanism underlying the effectiveness of the therapeutic vaccine is its ability to elicit a strong inflammatory response within the tumor microenvironment. This inflammatory reaction, driven by the secretion of inflammatory cytokines, serves to enhance the innate immune system, thereby fortifying its collaboration with the acquired immune system in combating cancer and facilitating a significant reduction of tumor size. However, further researches are required to determine the mechanism of action of the therapeutic vaccine. We think that after the treatment of a patient with this vaccine, it will be a preventive vaccine. If neoplastic masses or malignant tissues arise again, the activation of B and T memory cells may occur. A brief comparison can be made with the findings of Zahedipour et al. [ 74 ], who emphasize the critical role of angiogenesis in various malignancies, noting that cancer cells commonly overexpress VEGF-R2. While their study focused on melanoma, our research specifically addressed lung cancer. The Zahedipour team developed a vaccine targeting vascular endothelial cells using the VEGF-R2 receptor epitope peptide. They conducted an in silico analysis to select three potent peptides, each consisting of 9 amino acids, which exhibited the highest binding affinity for mouse MHC I (H-2Db, H-2Kb) and human HLA-A*02:01. The vaccines were administered subcutaneously at ten-day intervals, with each dose containing 5 µmol of vaccine per mouse. In contrast, our study employs three 16-amino acid peptides: VEGF-A, TGF-β, and MAGE-A3, which showed the highest binding affinities for both MHCI and MHCII. These peptides were merged and administered to two groups of mice at doses of 10 and 100 mg/ml, with results compared to a control group. Unlike Zahedipour et al., who evaluated individual peptides at a uniform dose, we examined the effects of varying peptide concentrations on immune stimulation. The VEGFR 2-derived peptides and those examined in our study exhibit low immunogenicity and favorable tolerance by the immune system, representing a significant advancement for their application in a nanoliposome-based formulation. We achieved an impressive encapsulation efficacy of 98% and zeta potential of 30 mV, compared to the 70% and − 17 mV reported by Zahedipour et al. Additionally, we enhanced the immune response by incorporating an alum adjuvant into the nanoliposome carrier. Both studies observed significant increases in IFNγ and IL-4 levels following vaccination, along with a reduction in tumor size. While Zahedipour et al. reported increased survival times for tumor-bearing mice, our study did not assess the survival rates. Although both studies focused on the stimulation of immune cells, Zahedipour et al. did not include qPCR and gene expression profiling. Their findings indicate that Lip-V1 significantly outperformed the V2 and V3 peptide vaccines in terms of immune system stimulation, tumor growth inhibition, and lifespan extension. Only V1 functioned as a strong immunogenic epitope, effectively inducing an immune response and enabling tumor control, while V2 and V3 demonstrated inadequate immunogenicity and failed to induce a robust antitumor response. Unlike our study, where the peptides demonstrated a synergistic effect in activating a broader of MHC molecules and immune cells, V2 and V3 underperformed, likely due to challenges related to lower immunogenicity, instability, processing in antigen-presenting cells (APCs), and their capacity to respond to the antigenic diversity of cancer cells [ 74 ]. A thorough and detailed description of each peptide is discussed below. The targets under our study have also been extensively researched in recent years due to their critical role and importance. It is known that VEGF-A increases the expression of PD-1 and other suppressive checkpoints like CTLA-4 by being localized on the surface of T cells. Also, it can repress T cell function by regulating the levels of Fas ligand (Fas L), which is regulated by VEGF-A in the tumor microenvironment (TME). In human tumors, the expression of FasL in endothelial cells leads to the loss of CD8 + T cells. Moreover, it directly regulates the amount of Treg in the TME by binding to VEGF-R2, which in turn stimulates the proliferation of Treg cells and enhances the action of immune suppressors. Furthermore, by binding to NRP-1, VEGF-A induces Treg cell infiltration in the TME. Therefore, VEGF-A may indirectly augment the invasive and metastatic potential of cancer cells through an induced increase in vascular permeability. Additionally, it can directly bind to hematopoietic cells containing the CD34 + marker and activate nuclear factor κB (NF-κB), inhibiting transcription in these cells to prevent dendritic cell (DC) differentiation and maturation, allowing cancer cells to escape immune surveillance. Moreover, IL-4-activated TAMs promote tumor cell malignancy by releasing various cytokines such as EGF and VEGF-A. This growth factor activates the MEK/ERK signaling pathways to enhance cell proliferation in NSCLC. VEGF-R activation can stimulate downstream MAPK/Akt signaling pathways, which play key roles in tumor and endothelial cell function, including proliferation, survival, and motility. The Akt signaling pathway positively regulates mTOR, which is associated with poor prognosis and tumor recurrence [ 77 ]. Further, high expression of cyclooxygenase 2 (COX-2) in NSCLC increases EGFR signaling through the production of prostaglandin E2 (PGE2), leading to higher expression of VEGF and augmented angiogenic potential. The EGFR pathway is frequently dysregulated in NSCLC and at some points converges with the VEGF pathway, playing an inductive role in angiogenesis. EGFR activation by EGF or TGF-α can lead to increased VEGF production, which has been associated with tumor metastasis and a worse disease prognosis [ 78 ]. Recognizing the importance of the VEGF-A factor in tumor cell survival and angiogenesis, we have chosen this factor as one of the protein targets to be inhibited and blocked by antibodies and the immune system, to use anti-angiogenic agents. When VEGF is blocked, tumor blood vessels are reduced, leading to induced tumor tissue hypoxia. This, in turn, activates hypoxia-inducible factor (HIF-1α) and produces cytokines to activate CD8 + T cells. Therefore, it is hypothesized that injecting a few doses of the VEGF target may lead to an increase in the number of immune cells, boost the penetration of mature TCD8 + cells, and attenuate the number of immunosuppressive cells in tumors. In addition, the direct effect of this peptide target will lead to a decline in the sprouting and formation of blood vessels and will destroy tumor vessels. TGF-β is another protein target that plays vital roles in multiple cellular functions and acts as an oncogene by promoting the survival, invasion, and metastasis of cancer cells. In the context of cancer, the TGF-β signaling pathway frequently assumes complex roles in the transition from non-cancerous conditions to malignancies. Additionally, this pathway significantly influences stromal cells and the immune system, potentially aiding tumors to evade immune-mediated elimination [ 37 ]. When it is produced and released by inflammatory cells (such as macrophages) or cancer cells in the tumor microenvironment, it causes the creation of a malignant microenvironment. As a result, through specific signaling pathways, it leads to drug resistance [ 79 ]. Overexpression of TGF-β causes resistance to anti-angiogenic therapies by promoting tumor vasculature and tolerance to immune checkpoint blockade. TGF-β inhibits the differentiation of Th1 and Th2 cells and causes the expansion and differentiation of Treg cells. Thus, the induction of Tregs can be an important immunosuppressive activity of TGF-β. It also induces Foxp3, a key regulator of Tregs in primary T cells [ 80 ]. TGF-β can convert host macrophages into suppressors of CD4 + T cell proliferation. In recent years, it has been revealed that CD4 + and CD25 + regulatory T cells can provide an essential source of TGF-β, which is responsible for attenuating tumor antigen-primed CTLs. Furthermore, TGF-β leads to inhibiting the production of effective cytokines such as IL-2, IFN-γ, and IL-4. It also inhibits macrophage activation, such as the induction of nitric oxide synthase and matrix metalloproteinase (MMP-12) through the Smad3 pathway [ 81 ]. It was reported that cisplatin reduced the level of TGF-β protein in surgically induced endometriosis. Theoretically, the cancer-specific antigen is essential for optimizing immune therapies. According to pan-cancer analysis, MAGE-A3 exhibits high expression in specific cancers, particularly in cases of lung cancer [ 82 ]. In 2019, Zhou et al. developed a vector using genetically attenuated Plasmodium sporozoite (GAS) to express human MAGE-A3 protein by the CRISPR-Cas9 system, to investigate the outcome of CD8 + T cells response induction against lung cancer in the A549 cell line compared with the GAS vector without MAGE-A3 protein. Their research revealed the significant advantages of GAS as a vector for triggering MAGE-A3-specific CD8 + T cell response against the tumor antigen GAS/MAGE-A3. This response led to a notable delay in the growth of the A549 cell line in mice and resulted in the reduction of tumor size [ 83 ]. These results suggest potential improvements in lung cancer therapy by targeting MAGE-A3 to stimulate specific T and B cell responses, impede tumor cell growth, and induce apoptosis [ 84 ]. One of the limitations of our project was the small size of the statistical population and the unfortunate loss of some mice during the peptide dose administration. Within one month of the peptide injection into Balb/c mice, we experienced the death of five mice, possibly due to inflammation caused by the peptide injection. This inflammation response could be attributed to the secretion of growth factors, cytokines, proteases, and extracellular matrix-modifying enzymes from cell populations. Subsequent autopsies revealed inflammation in all the organs of the affected mice. Hence, a smaller amount of peptide target was used for the remaining mice. Secondly, the population studied in this research was Balb/c mice, but due to the price and difficult maintenance conditions of other experimental models, such as rabbits or monkeys, the study on them was omitted. Further, we only examined the expression level of two genes associated with apoptosis, but to confirm our study, it is proposed to investigate the expression levels of checkpoint inhibitors and caspase genes. The ELISA test may have inconsistent results due to technical errors or incorrect sampling. It is suggested to confirm the results using complementary techniques such as Western Blot. Another limitation of the TAA-derived target design is that not only Th1 cells but also peptide-specific Th2 cells or Treg cells, which normally do not contribute to TAA-specific CTLs, might be induced after vaccination. However, this challenge can be addressed with the use of adjuvants. According to Xuedan He et al.‘s 2018 paper, it was found that alum can boost Th2 immune responses instead of strong Th1 reactions, which Th2 is more related to cellular responses and is favorable for most cancer vaccines [ 85 ]. Another limitation of peptide usage is the determined restriction of the peptide to a specific HLA type, which requires vaccine personalization and patient selection accordingly. Product heterogeneity is possible because peptide-based targets are designed based on various amino acid sequences. Therefore, in vivo, instability can lead to unpredictable biodistribution profiles and modulate therapeutic effects. Development of a multi-epitope peptide for universal use is challenging, due to MHC limitations and the wide heterogeneity in MHC alleles in the human population [ 22 ]. Nevertheless, the development of bioinformatic data to predict HLA alleles with the most overlap and focusing on the most common MHCs, as well as designing peptides capable of binding to more than one MHC allele, are possible solutions to this issue. While humanized mouse HLA exhibits similarities to human HLA, notable differences remain across various human populations and among different racial groups. Moving forward, it’s crucial to design peptides that encompass a broader range of alleles and to analyze a wider array of populations to achieve a more comprehensive average. The discrepancies between animal and human alleles must not be overlooked, as they can lead to tissue rejection rather than the intended therapeutic effect. Any observed results may well reflect this rejection issue. Therefore, to attain more reliable and impactful outcomes, it is essential to invest in innovative models that more closely resemble human HLA, despite the potentially high costs associated with such approaches. Tools like NetMHCpan and IEDB can facilitate the prediction of peptide binding to numerous human HLA alleles, allowing for the design of peptides that reflect diversity. Furthermore, leveraging databases such as the Allele Frequency Net Database (AFND) can inform vaccine design based on global HLA distribution, thereby reducing concerns related to rejection. Only through this rigorous and informed methodology can we effectively advance the field of immunotherapy. Although this study proved the efficacy of the multi-epitope vaccine in vitro and in vivo, there is still a long way to go, and further steps need to be taken. In the next step, both the efficacy and side effects of this vaccine should be checked on animal models close to humans, such as chimpanzees. It is also necessary to design a clinical trial following ethical codes and examine the above-mentioned parameters in lung cancer patients. The important limitation is the financial support of the project, which must be addressed by pharmaceutical companies.

Non-small cell lung cancer encompasses adenocarcinoma, squamous cell carcinoma, and large cell carcinoma [ 1 , 2 ]. Lung cancer is a critical health concern, ranking the highest in mortality among 36 types of cancer and standing as the second most commonly diagnosed cancer worldwide [ 3 – 5 ]. Despite localized declines in morbidity and mortality, the global incidence rate of lung cancer continues to rise, with more than 2 million new cases each year [ 6 ]. Notably, a significant proportion of these cases occur in non-smokers and women [ 2 ]. It is crucial to note that standard treatments like chemotherapy and radiation therapy often fail against lung cancer [ 7 , 8 ]. Current early detection methods, such as chest radiography and sputum cytology, have limited effectiveness, resulting in late-stage diagnoses due to the absence of clinical symptoms and effective screening programs [ 9 ]. A striking 65.33% of lung cancer cases were already in the advanced stage at the time of diagnosis, highlighting the urgent need for more effective treatment options and improved prognoses for NSCLC [ 10 ]. Consequently, specific and effective screening tests are imperative for individuals over 50 years old [ 11 , 12 ]. In the past decade, much progress has been made in the treatment of NSCLC, such as targeted therapies, immunotherapy, and cancer vaccines [ 4 ]. Combined treatment, including vaccine-mediated therapy, is under investigation [ 13 ]. Although vaccines utilize the immune system to target and kill specific tumor cells, this approach has been disappointing [ 1 , 7 ]. However, over the past decade, a greater understanding of the immune system and antigens expressed by tumors, along with advances in immune adjuvants and developed delivery systems, has led to progress in the use of immunotherapy, including vaccination, to target lung cancer [ 1 , 14 , 15 ]. Personalized cancer vaccines represent a cutting-edge approach to treat a wide range of cancers. By activating the immune system to target specific targets, offers hope for improved outcomes for individuals grappling with this challenging disease [ 16 ]. Previously, researchers have provided different novel approaches in diagnosis, prediction, and treatment of solid lung cancer [ 17 ], such as circle-map profiling of extrachromosomal circular DNA [ 18 ], EV-derived non-coding RNA [ 19 ], and celery-derived extracellular vesicles [ 20 ]. Today, most cancer vaccine researches are focused on targeting tumor-associated antigens (TAA) [ 21 , 22 ]. The first therapeutic cancer vaccine approved by the FDA is sipuleucel-T. Sipuleucel-T [ 23 ]. Cancer vaccines are protein-based, peptide-based, cell-based, DNA or RNA-based, and viral or bacteria-based [ 23 – 25 ]. More than 70 targets have been identified as TAAs [ 24 , 26 ]. Targeting these antigens provides a logical strategy to preferentially eliminate tumor cells. Among them, Melanoma-associated antigen 3 (MAGE-A3) [ 27 ], Vascular endothelial growth factor A (VEGF-A) [ 28 ], and transforming growth factor beta (TGF-β) [ 29 ] exhibit high expression in lung cancer. These targets serve as crucial biomarkers for evaluating tumor aggressiveness. Understanding these markers can provide valuable insights into tumor behavior, patient prognosis, and potential therapeutic targets in cancers [ 30 ]. Testicular cancer antigens, such as MAGE-A3, are tumor-specific antigens encoded by cancer germline genes. It is highly prevalent in 35–55% of NSCLC tissues and has been highly expressed in many tumors, including melanoma, non-small cell lung cancer, and hematological malignancies [ 27 , 31 ]. Another target is VEGF-A, which is known as a key regulator of tumor vessel growth (neoangiogenesis) and leads to tumor-induced immune suppression. VEGF-A enhances the proliferation of immunosuppressive cells and limits the recruitment of T cells into tumors, while also increasing the exhaustion of T cells by up-regulating inhibitory receptors. In cancer patients, the production of VEGF-A by the tumor leads to the “angiogenic switch” required for tumor growth and metastasis [ 32 , 33 ]. Another target is TGF-β which is also known as a tumor marker correlated with lung cancer. TGF-β growth factor is a multifunctional polypeptide that is considered the most potent immunosuppressive factor, as it can reduce the proliferation, activation, and differentiation of immune cells [ 32 , 34 ]. In the early stages of cancer, TGF-β exhibits tumor-suppressive effects by inhibiting cell cycle progression, promoting apoptosis, and repressing the growth of normal and premalignant epithelial cells. Conversely, in the advanced stage, the situation is reversed [ 35 , 36 ]. The ongoing development of immune checkpoint inhibitors in combination with anti-TGF-β antibodies emphasizes the potential of this approach in reshaping cancer treatment [ 37 ]. The purpose of this research was to determine linear epitopes for MAGE-A3, TGF-β2, and VEGF-A and then to determine the allergenicity, and toxicity of selected and merged epitopes. After their laboratory synthesis, nanoliposomes containing selected epitopes were characterized. After the injection of synthesized nanoliposomes into Balb/c mice, the antibody titer was determined and the serum level of cytokines was detected. Moreover, the ability of produced antibodies to induce apoptosis in cancer cells was also evaluated. Finally, the effect of a multi-epitope peptide on cancer tumor growth in the Patient-derived xenograft (PDX) model was investigated.

Supplementary Material 1: Supplementary 1. The clinical characterization of donor patients for PDX models. Supplementary 2. The basic characterization of PDX models used in this study. Supplementary 3. The of PDX models HLA profiles. Supplementary 4. The results of the C-ImmSim server. (a) Antigen (Ag) count along with antibody titers with specific subclasses, (b) B cells population, (c) Cytokines responses, (d) Th (helper) cells population, (e) Tc (cytotoxic) cells population, (f) Tc cells population per state. Supplementary 5. The stability test for nanoliposome formulation in physiological buffer (PBS) and 50% FBS during 1-4 weeks by measuring their mean size and Zeta potential. Supplementary 6. The mean cytokine levels in controls and vaccinated PDX models to check immune system response. * Indicates significant difference when compared with PBS-injected group, irrelevant peptide- injected group, and the non-humanized mice at the same time (P<0.05, n=6). Supplementary 7. The quantification of CD8+ T cells, Tregs, MDSCs, and immune checkpoints (PD-1 and CTLA-4) and memory T-cell responses (CD45RA(+) CCR7 (-)) in humanized and non-humanized mice to check TME and Immune Profiling. * Indicates significant difference when compared with irrelevant peptide- injected group and the non-humanized mice at the same time (P<0.05, n=6). Supplementary 8. The quantification of CD8, Ki-67, cleaved caspase-3 in tumor, liver, and kidney in humanized and non-humanized mice at week 5 to assess toxicity based on histopathology data. Supplementary Material 1: Supplementary 1. The clinical characterization of donor patients for PDX models. Supplementary 2. The basic characterization of PDX models used in this study. Supplementary 3. The of PDX models HLA profiles. Supplementary 4. The results of the C-ImmSim server. (a) Antigen (Ag) count along with antibody titers with specific subclasses, (b) B cells population, (c) Cytokines responses, (d) Th (helper) cells population, (e) Tc (cytotoxic) cells population, (f) Tc cells population per state. Supplementary 5. The stability test for nanoliposome formulation in physiological buffer (PBS) and 50% FBS during 1-4 weeks by measuring their mean size and Zeta potential. Supplementary 6. The mean cytokine levels in controls and vaccinated PDX models to check immune system response. * Indicates significant difference when compared with PBS-injected group, irrelevant peptide- injected group, and the non-humanized mice at the same time (P<0.05, n=6). Supplementary 7. The quantification of CD8+ T cells, Tregs, MDSCs, and immune checkpoints (PD-1 and CTLA-4) and memory T-cell responses (CD45RA(+) CCR7 (-)) in humanized and non-humanized mice to check TME and Immune Profiling. * Indicates significant difference when compared with irrelevant peptide- injected group and the non-humanized mice at the same time (P<0.05, n=6). Supplementary 8. The quantification of CD8, Ki-67, cleaved caspase-3 in tumor, liver, and kidney in humanized and non-humanized mice at week 5 to assess toxicity based on histopathology data.