Genetic and molecular mechanisms of sarcomas

Abstract

Sarcomas are heterogeneous malignant mesenchymal tumors with diverse biological features and unique clinical characteristics, the genetic alterations of sarcomas are highly variable. With the development of sequencing technologies, efficient and practical approaches to detect gene expressions and gene variants contribute to the prediction of patient prognosis and the choice of treatment modalities. Given the rarity of sarcomas, the comprehensive transcriptomic or genomic profiles are still lacking for many subtypes. In the present thesis, by applying sequencing technology in sarcoma cohorts, combined with bioinformatics data analysis and molecular biology experiments, we have revealed new biological mechanisms dictating sarcoma behavior and provided insights for clinical applications.

In Paper I, we characterized the gene signatures related to poor prognosis, first-line treatment failure, and chemotherapy resistance in Ewing sarcoma (ES). High expression of IGF2 was associated with shorter overall survival in ES patients and promoted cell proliferation, radiation resistance, and apoptosis inhibition in vitro. The transcriptome analysis of clinical samples and cell lines uncovered an IGF-dependent signature and potentially related to stem cell-like signatures in ES.

Paper II continued to highlight the transcriptome signatures in ES. Here, we identified prognosis-related RNA-binding proteins (RBPs) and constructed an RBP-based prognostic risk model that showed stable predictive power for evaluating overall survival in clinical samples. Within the model, NSUN7 is considered an independent prognostic favorable prognostic marker, which was also validated by immunohistochemistry.

In Paper III, we discovered that TERT promoter mutations were present in 45% of patients in a cohort of 190 patients with conventional chondrosarcoma (CHS). The mutation was significantly associated with recurrence, distant metastasis, and high tumor grade. The heterogeneity of primary tumors and the altered mutational status between asynchronous metastatic lesions revealed that CHS is a multiclonal disease that progresses through branching evolution.

In Paper IV, we identified three clusters with distinct transcriptomic and genomic patterns in synovial sarcoma (SS), of which SS cluster I (SSC-I) was characterized by hyperproliferation, immune cell silencing, and poor prognosis; SSC-II was characterized by high vascularity and stromal component with the better clinical outcome; SSC-III was characterized by epithelial components with genomic complexity and checkpoint-mediated immune suppression.

Collectively, the present thesis illustrated the pathogenic mechanisms of ES, CHS, and SS through the analysis of transcriptomic and genomic data, identified prognostic biomarkers, and at the clinical application-level provided strong evidence for patient stratification, risk prediction, and personalized treatment assessment.