Recent studies have shown that tumor antigens, particularly tumor-specific antigens, which induce tumor-specific cytotoxic T lymphocytes (CTLs) and damage tumor cells, are a significant component of tumor vaccines. Epitope peptide vaccine has been of particular interest as it induces specific CTL in vitro and in vivo in order to kill target cells. Previously, a number of human genes that code for tumor antigens identified by autologous CTLs were isolated (1,2), and the epitopes derived from these tumor antigens were further identified to serve as targets for CTLs in the context of HLA class I molecules (3). MAGE-n is a newly identified member of the MAGE gene family which was first reported by our laboratory (Genebank, locus no. AF443295) (4), and was also found to be highly homologous to MAGE-A subfamily genes. Since HLA-A2/A24 is one of the most frequently expressed molecules in the Chinese population (5,6), it is crucial to identify the tumor antigen epitopes which are presented by HLA-A2/A24 and to induce epitope-specific CTLs against tumor cells. The present study reported a simple and efficient bioinformatics method to identify candidate HLA-A2/A24-restricted CTL epitopes from the tumor antigen MAGE-n.

MAGE-n (316 aa) was selected as the antigen of interest in this study. The amino acid sequences of the tumor antigen were obtained from the Genbank database: MSLEQRSQHCKPEEGLEARGEALGLVGAQAPATEEQEAASSSSTLVEVTLGEVPAAESPDPPQSPQGASSLPTTMNYPLWSQSYEDSSNQEEEGPSTFPDLESEFQAALSRKVAELVHFLLLKYRAREPFTKAEMLGSVIRNFQDFFPVIFSEASEYLQLVFGIEVVEVVRIGHLYILVTCLGLSYDGLLGDNQIMPKTGFLIIVLVMIAMEGGHAPEEEIWEELSVMEVYDGREHSAYGEPRKLLTQDLVQEKYLEYRQVPDSDPARYEFLWGPRALAETSYVKVLEYVIKVSARVRFFFPSLREAALREEEEGV.

The NetCTL-1.2 prediction values of all possible nonamers of a given protein sequence were added together. The peptides of each protein were selected for further analysis. The primary predicted epitopes were compared to those of previous studies (7–9). It was found that the 4 reported HLA-A2-restricted CTL epitopes, i.e., MAGE-1_278–286(KVLEYVIKV), MAGE-3_278–286(KVAELVHFL), AGE-3_271–279(FLWGPRALV) and MAGE-n_159–167(QLVFGIEVV) were eliminated from the NetCTL-1.2 prediction results. The epitopes with COMB scores <0.85 were eliminated from the prediction results. A total of 8 HLA-A2- and 9 HLA-A24-restricted CTL epitope candidates (nonamers) derived from the tumor antigen MAGE-n were predicted. The results of predicted epitopes are shown in Tables I and II, i.e., that the NetCTL1.2 Server prediction method improved prediction efficacy and accuracy.

Recently, with the innovation of immunology methodology, the molecular mechanisms of tumor antigen presentation and CD8⁺ T-cell activation have been gradually elucidated. Tumor antigens and their coding genes have been identified in a consecutive manner. The breakthrough in theoretical tumor immunity has accelerated the development of vaccines targeting tumor antigens. The induction of antigen-specific CTLs has been suggested to be highly efficacious in the prevention and treatment of various types of tumor. Induction of potent anti-tumor CTL responses results in the regression and prevention of metastasis formation, as shown in experimental model tumor systems (10,11). Thus, efforts towards the development of cancer immunotherapy have recently focused on the generation of tumor-specific T-cell responses.

Epitope prediction was recently conducted in tumor antigen specificity for CTL epitope identification in a majority of studies. Approximately 120 CTL epitopes presented by HLA-A, B and C molecules were reported (12), and various epitopes were used as peptide vaccines in animal and clinical experiments (13,14). Schirle et al reported that two new CTL epitopes of gastrointestinal tumors were identified by epitope prediction combined with acid elution methods (15). Dong et al found that MAGE-n-derived peptide QLVFGIEVV was a new HLA-A2.1-restricted CTL epitope that induces MAGE-n-specific CTLs in vitro (9). Identification of these CTL epitopes allows for clinical applications of these peptides as cancer vaccines for patients with MAGE-n⁺/HLA-A2⁺ tumors. Our previous study showed that the combination of MAGE-3 and MAGE-n epitopes induced more effective anti-tumor immune responses than either of the peptides alone (16). In addition, the peptide-specific activity was observed to be in a MHC-restricted manner. This study suggested that the combination of a number of tumor antigen-derived peptides is more efficacious as a peptide-based cancer immunotherapy.

The combination of epitope prediction, epitope reconstruction and immunological methods improves the efficacy and accuracy of CTL epitope studies (9). The identification of human tumor-rejection antigens and CTL epitopes of these antigens allows for the development of new cancer vaccines since patients are likely to benefit from immunization with an identified CTL epitope. Among the identified tumor-rejection antigens, MAGE gene products are of particular interest due to their wide expression in various types of tumor and their potential to induce tumor-specific CTL responses. A previous study showed that a vaccine with 5 common HLA-I-restricted epitopes was beneficial in 80–90% of patients (17). It was also found that 8–9 common HLA-I-restricted epitopes may exhibit benefical effects on patients worldwide (17). The identification of the HLA-A2-restricted CTL epitopes from melanoma antigen MAGE-n is crucial. In the present study, the HLA-A2/A24-restricted CTL epitopes of MAGE-n were predicted using the methods from the NetCTL1.2 Server. These candidate epitopes were selected for further immunology experiments. The results showed that the NetCTL1.2 Server prediction method improves prediction efficacy and accuracy. These nonamers, following identification through experimentation may contribute to the development of potential antigen peptide tumor vaccines.