Dataset Information

Identification of novel subtypes based on ssGSEA in immune-related prognostic signature for tongue squamous cell carcinoma.

ABSTRACT:

Background

Tongue squamous cell carcinoma (TSCC) is characterized by aggressive invasion and poor prognosis. Currently, immune checkpoint inhibitors may prolong overall survival compared with conventional treatments. However, PD1/PDL1 remain inapplicable in predicting the prognosis of TSCC; thus, it is urgent to explore the genetic characteristics of TSCC.

Materials and methods

We utilized single-sample gene set enrichment analysis (ssGSEA) to classify TSCC patients from the TCGA database into clusters with different immune cell infiltrations. ESTIMATE (immune-related scores) and CIBERSORT (immune cell distribution) analyses were used to evaluate the immune landscape among clusters. GO, KEGG, and GSEA analyses were performed to analyze the different underlying molecular mechanisms in the clusters. Based on the immune characteristics, we applied the LASSO Cox regression to select hub genes and construct a prognostic risk model. Finally, we established an interactive network among these hub genes by using Cytoscape, and a pan-cancer analysis to further verify and decipher the innate function of these genes.

Results

Using ssGSEA, we constructed three functional clusters with different overall survival and immune-cell infiltration. ESTIMATE and CIBERSORT analyses revealed the different distributions of immune cells (T cells, B cells, and macrophages) with diverse immune-related scores (ESTIMATE, immune, stromal, and tumor purity scores). Moreover, pathways including those of the interferon-gamma response, hypoxia, and glycolysis of the different subtypes were investigated to elucidate their involvement in mediating the heterogeneous immune characteristics. Subsequently, after LASSO Cox regression, a signature of 15 immune-related genes was established that is more prognostically effective than the TNM stage. Furthermore, three hub genes-PGK1, GPI, and RPE-were selected using Cytoscape evaluation and verified by immunohistochemistry. PGK1, the foremost regulator, was a comprehensively profiled pan-cancer, and a PGK1-based interactive network was established.

Conclusion

Our results suggest that immune-related genes and clusters in TSCC have the potential to guide individualized treatments.

SUBMITTER: Jin Y

PROVIDER: S-EPMC8633230 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Esophageal carcinoma (EC) is a common malignant cancer worldwide. Esophageal squamous cell carcinoma (ESCC), the main type of EC, is difficult to treat because of the widespread morbidity, high fatality rates, and low quality of life caused by postoperative complications and no specific molecular target. In this study, we screened genes to establish a prognostic model for ESCC. The transcriptome expression profiles of 81 ESCC tissues and 340 normal esophageal mucosal epithelium tissues were obtained from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) cohorts. The transcriptome expression datasets of 19 esophageal squamous carcinoma cell lines were downloaded from Cancer Cell Line Encyclopedia (CCLE). The R software Limma package was used to identify 6,231 differentially expressed genes and 647 differentially expressed immune-related genes between normal and ESCC tissues. Gene functional analysis was performed using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). Weighted gene co-expression network analysis (WGCNA) was used to screen out 18 immune-related prognostic genes. We then established the prognostic and risk signature using these genes, and the patients were divided into low-risk and high-risk groups. Compared with high-risk group patients, the low-risk group patients had longer overall survival. M1 macrophages and resting dendritic cells were differentially distributed between the low-risk and high-risk groups and were related to patient survival. We also examined the functional immune cell and immune molecule levels in low-risk and high-risk group patients, with significant differences in the tumor microenvironment between the two groups. To further verify the accuracy of the prognostic risk model, we performed area under the ROC curve (AUC) analysis. The AUC value was 0.931 for the prognostic risk, which was better than the microsatellite instability (MSI) and Tumor Immune Dysfunction and Exclusion (TIDE) scores. In conclusion, we found 18 immune-related prognostic genes related to the occurrence of ESCC and established a prognostic model for predicting disease severity.

Project description:BackgroundEarly-stage lung squamous cell carcinoma (LUSC) progression is accompanied by changes in immune microenvironments and the expression of immune-related genes (IRGs). Identifying innate IRGs associated with prognosis may improve treatment and reveal new immunotherapeutic targets.MethodsGene expression profiles and clinical data of early-stage LUSC patients were obtained from the Gene Expression Omnibus and The Cancer Genome Atlas databases and IRGs from the InnateDB database. Univariate and multivariate Cox regression and LASSO regression analyses were performed to identify an innate IRG signature model prognostic in patients with early-stage LUSC. The predictive ability of this model was assessed by time-dependent receiver operator characteristic curve analysis, with the independence of the model-determined risk score assessed by univariate and multivariate Cox regression analyses. Overall survival (OS) in early-stage LUSC patients was assessed using a nomogram and decision curve analysis (DCA). Functional and biological pathways were determined by gene set enrichment analysis, and differences in biological functions and immune microenvironments between the high- and low-risk groups were assessed by ESTIMATE and the CIBERSORT algorithm.ResultsA signature involving six IRGs (SREBF2, GP2, BMX, NR1H4, DDX41, and GOPC) was prognostic of OS. Samples were divided into high- and low-risk groups based on median risk scores. OS was significantly shorter in the high-risk than in the low-risk group in the training (P < 0.001), GEO validation (P = 0.00021) and TCGA validation (P = 0.034) cohorts. Multivariate Cox regression analysis showed that risk score was an independent risk factor for OS, with the combination of risk score and T stage being optimally predictive of clinical benefit. GSEA, ESTIMATE, and the CIBERSORT algorithm showed that immune cell infiltration was higher and immune-related pathways were more strongly expressed in the low-risk group.ConclusionA signature that includes these six innate IRGs may predict prognosis in patients with early-stage LUSC.

Project description:BackgroundGlioblastoma multiforme (GBM) is extensively genetically and transcriptionally heterogeneous, which poses challenges for classification and management. Long noncoding RNAs (lncRNAs) play a critical role in the development and progression of GBM, especially in tumor-associated immune processes. Therefore, it is necessary to develop an immune-related lncRNAs (irlncRNAs) signature.MethodsUnivariate and multivariate Cox regression analyses were utilized to construct a prognostic model. GBM-specific CeRNA and PPI network was constructed to predict lncRNAs targets and evaluate the interactions of immune mRNAs translated proteins. GO and KEGG pathway analyses were used to show the biological functions and pathways of CeRNA network-related immunity genes. Consensus Cluster Plus analysis was used for GBM gene clustering. Then, we evaluated GBM subtype-specific prognostic values, clinical characteristics, genes and pathways, immune infiltration access single cell RNA-seq data, and chemotherapeutics efficacy. The hub genes were finally validated.ResultsA total of 17 prognostically related irlncRNAs were screened to build a prognostic model signature based on six key irlncRNAs. Based on GBM-specific CeRNAs and enrichment analysis, PLAU was predicted as a target of lncRNA-H19 and mainly enriched in the malignant related pathways. GBM subtype-A displayed the most favorable prognosis, high proportion of genes (IDH1, ATRX, and EGFR) mutation, chemoradiotherapy, and low risk and was characterized by low expression of four high-risk lncRNAs (H19, HOTAIRM1, AGAP2-AS1, and AC002456.1) and one mRNA KRT8. GSs with poor survival were mainly infiltrated by mesenchymal stem cells (MSCs) and astrocyte, and were more sensitive to gefitinib and roscovitine. Among GSs, three hub genes KRT8, NGFR, and TCEA3, were screened and validated to potentially play feasible oncogenic roles in GBM.ConclusionConstruction of lncRNAs risk model and identification of GBM subtypes based on 17 irlncRNAs, which suggesting that irlncRNAs had the promising potential for clinical immunotherapy of GBM.