LGG, as common primary intracranial tumor, is one of the most aggressive malignant tumors. Due to its biological heterogeneity and immunosuppressive properties, conventional treatments such as surgical resection, radiotherapy, and chemotherapy are not sufficient to combat their progression[52]. Immunotherapy is a novel and explosively growing field and the mRNA vaccines are a promising immunotherapy against tumors[53]. Tumor eradication through vaccine-induced active immunity is the ultimate goal of tumor immunotherapy[54]. The initial success of a peptide vaccine targeting IDH1 mutants in fighting IDH1 mutant gliomas[55], has greatly encouraged the development of mRNA vaccines. However, the clinical application of the mRNA vaccine in LGGs is still needed to be explored, thus, we hope our research could provide a valuable reference for the development and application of mRNA vaccine.
In the current study, we identified seven potential tumor antigens for developing mRNA vaccine by using the aberrantly expressed profile and exploring the mutational landscape of LGGs. These tumor antigens, including CREB3L2, DDR2, IRF2, NCSTN, RECQL, REST, and TGFBR1, are overexpressed, mutated, and amplified in LGGs and are promising targets for developing mRNA vaccine. Their high expression was not only correlated with shortened survival but also significantly correlated with the high infiltration of APCs, demonstrating their key role in the invasion and progression of LGGs, and the useful value as vaccine candidates. Although further experimental verification and clinical evaluation are still urgently needed, their potential value as mRNA vaccine candidates were supported in available studies. CREB3L2, a member of the CREB3 transcription factor family, is characterized in cancer by the formation of the chimeric gene FUS-CREB3L2, which is responsible for low-grade fibromyxoid sarcoma [56]. While, in malignant glioma, CREB3L2 binds directly to the ATF5 promoter, leading to ATF5 transcription and thus exerting an anti-apoptotic effect[57]. DDR2, a member of tyrosine kinase receptors (TKRs) families, is probably a potential molecular biomarker for multiple types of cancer[58]. It has been demonstrated to promote a variety of tumor cells proliferation, such as lung cancer[59], lymphoma[60], and oral squamous cell carcinoma[61] among others. In addition, its overexpression was positively correlated with poor prognosis of patients[62]. Besides, the mutation of DDR2 was also closely related to the formation in some tumors[58]. It has been shown that there is a significant positive correlation between the expression level of IRF2 and the grade of glioma, and that its overexpression enhances the invasion and migration of glioma cells, and vice versa[63]. Nicastrin (NCSTN) is one of a core subunit of γ-Secretase, which was found to be closely associated with the tumor progression of a variety of tumors[64, 65]. RECQL, namely RECQ1 or RECQL1, played critical role in ensuring chromosomal stability[66]. However, it was high expressed in transformed or a variety of cancer cells, for instance, multiple myeloma[67], hematological cancers[68], glioblastoma[69], and so on, showing that RECQL may be involved in tumorigenesis. Repressor element silencing transcription factor (REST), a well-known transcription repressor, has been reported to be an oncogene and be related to poor prognosis in glioma[70]. In addition, its upregulation was associated with larger tumor size, higher grade and worse therapeutic efficacy in glioma[71]. TGF-β receptor 1 (TGFBR1), a member of the TGF-β signaling pathway, has been increasingly shown to promote epithelial-mesenchymal transition (EMT) and the proliferation and migration of tumor cells[72]. And it was also found to be high expressed in a variety of tumors and involved in the progression and metastasis of tumor[73, 74]. More importantly, IHC staining showed that TGFBR1 is higher expressed in glioma samples and positively correlated with the co-stimulatory factors, which indicated that TGFBR1 may serve as an important tumor antigen for vaccine development.
Given the critical role of ICDs in antitumor immunotherapy[75, 76] and the fact that mRNA vaccines are beneficial for only a small subset of cancer patients. Thus, we classified LGG patients into two ICD subtypes based on the ICD gene profile to select optimal populations for mRNA vaccine. To our knowledge, this is the first article on ICD-based typing to screen for optimal vaccine recipients with LGGs. The two ICD subtypes characterized by a distinct ICD transcription profiles expression, clinical prognosis, molecular and cellular features. Compared to Icds1, Icds2 had a prolonged survival, indicating that ICD typing could be a potential prognostic biomarker for LGGs. More importantly, different molecular, and cellular features were observed between ICD subtypes, indicating that distinct mechanisms regulated the tumor immune environment between the subtypes and distinct therapeutic strategies are required for different ICD subtype patients. We analyzed the TMB and gene mutation count differences between ICD subtypes and found a significantly higher TMB, gene mutated counts, TP53 mutation rate and a lower IDH mutation rate were shown in the Icds1 subtype. As shown in the publications[77, 78], LGGs with wild-type IDH and TP53 mutation has a poor prognosis, which were consistent with our findings. In addition, high TMB has the potential to generate immunogenic neoantigens[79] and further drive APCs and immune effector cells infiltration. However, Icds1 exhibited significantly higher ICD modulators, infiltration of APCs, and immune effector cells such as natural killer cell, activated B cell, and activated CD8 + T cell, it had obviously poorer prognosis. These results indicated that immunosuppression may play a dominant role in the regulation of the tumor microenvironment in this subtype. Consistent with this hypothesis, immunosuppressive cells (for instance MDSC, and regulatory T cell), immune inhibiting functions, and immunosuppression signaling pathways were all enriched in Icds1 subtype. The immunosuppressive tumor microenvironment of Icds1 may inhibit the effective immune response elicited by mRNA vaccine. Thus, the immunologically "hot" and immunosuppressive phenotype of Icds1, which is not suitable for mRNA vaccine alone but more suitable for immune checkpoint blockade (ICB), or Icds1 were likely to benefit from the strategy with mRNA vaccine combination with ICB. Indeed, TMB is becoming a potential biomarker for predicting response to ICB[80]. In contrast, Icds2 was immunologically "cold" phenotype with lower level of immune infiltration cells, immune function, and immune-related signaling pathway. Since mRNA vaccine could elicit an effective immune response, it might be more suitable for patients with low infiltrated immune cells. Hence, Icds2 but not Icds1 were the optimal candidates for mRNA vaccine in LGGs. To validate the reliability of our ICD subtypes, we further explored the association between ICD subtypes and six reported pan-cancer immune subtypes and found consistent results that C4 (lymphocyte depleted) was mainly enriched in Icds1, while C5 (immunologically quiet) was mainly clustered in Icds2. Taken together, our ICD subtypes could not only predict survival of LGG patients, but also select optimal candidates for mRNA vaccine. Specifically, Icds2 is an immunological “cold” phenotype and is suitable for mRNA vaccine.
Considering that ICD subtype may be not stable among different patient populations and the biomarkers of ICD subtypes are the key of linkage candidate population screening, typing specificity, and mechanism research[54]. Hence, we constructed immune gene co-expression network by using WGCNA and identified six ICD subtype-associated gene modules. Among them, the yellow, brown, and turquoise modules were significantly related to overall survival. Finally, based on the threshold MM value of larger than 0.93, we identified 5 hub genes including HLA-DMB, SPI1, CASP1, ITGB2, and LAIR1 from turquoise module, which might be potential biomarkers for predicting response to mRNA vaccine.