New functional signatures for understanding melanoma biology from tumor cell lineage-specific analysis [single channel experiments]. Homo sapiens

The identification of molecular signatures specific to distinct tumor types is urgently required for the design of therapies overcoming treatment resistance. It remains unclear whether such signatures are shared among tumors and corresponding cell lines, a key issue given the use of cell lines for drug development. We developed SCA (similarity core analysis) an unsupervised computational framework for extracting the core molecular features common to tumors and corresponding cell lines. We applied this algorithm to mRNA and miRNA expression data from various sources, comparing melanoma cell lines with melanoma metastases as an example. The signature obtained was associated with phenotypic characteristics in vitro and the core genes CAPN3 and TRIM63 were identified as new actors in melanoma cell migration and invasion. About 90% of the genes in the melanoma signature are potentially regulated by an intrinsic network of transcription factors known to regulate neural development (TFAP2A, DLX2, ALX1, MITF, PAX3, SOX10, LEF1 and GAS7) and three miRNAs (211-5p, 221-3p, 10a-5p). The SCA signature effectively discriminated between two subpopulations of melanoma patients with significant difference in overall survival. Furthermore, it classified MEKi and BRAFi resistant and sensitive melanoma cell lines. The SCA algorithm is potentially applicable to any tumor cell type. Overall design: 21 melanoma cell lines are analyzed in single color

精准识别不同肿瘤类型特异性分子标志物（molecular signatures），是开发克服治疗耐药性疗法的迫切需求。目前尚不清楚此类分子标志物是否可在肿瘤及其对应细胞系间共享，鉴于细胞系广泛应用于药物开发，这一问题至关重要。我们开发了SCA（相似性核心分析，similarity core analysis）——一种用于提取肿瘤与对应细胞系共有核心分子特征的无监督计算框架。我们将该算法应用于多来源的mRNA与miRNA表达数据，并以黑色素瘤细胞系与黑色素瘤转移灶的比较为例进行验证。所获得的分子标志物与体外表型特征相关，且核心基因CAPN3与TRIM63被鉴定为参与黑色素瘤细胞迁移与侵袭的新作用因子。该黑色素瘤标志物中约90%的基因，可能受调控神经发育的内在转录因子网络（TFAP2A、DLX2、ALX1、MITF、PAX3、SOX10、LEF1及GAS7）以及三种miRNA（211-5p、221-3p、10a-5p）调控。SCA标志物可有效区分总生存期存在显著差异的两类黑色素瘤患者亚群。此外，该标志物可有效区分对MEKi及BRAFi分别耐药与敏感的黑色素瘤细胞系。SCA算法理论上可应用于任意肿瘤细胞类型。整体实验设计：采用单色法分析21株黑色素瘤细胞系。