左、右结肠癌系统性多维组学差异特征及耐药标志识别

Based on the gene expression profiles, there are a substantial number of several types of prognostic signatures for colon cancer (CC). However, few have succeeded in clinical application. Our previous study found that the following three factors will have an important impact on the clinical transformation of CC prognostic signature: (1) Most of current studies analyzed CC samples with and without adjuvant chemotherapy together in the training and/or validation datasets. It will cause indistinguishable between prognostic and predictive signatures. (2) Right-sided and left-sided CC differ with prognoses and the response to adjuvant chemotherapy. It may be necessary to identify prognostic signature respectively. (3) Cellular components of tissue samples exert an important influence on the reliability of prognostic signature. In this work, based on the high-throughput multi-omics data obtained from the TCGA database, we systematically analyze the genome, transcriptome and epigenome molecular differences between right-sided and left-sided CC to reveal molecular mechanisms underlying the difference of the response to adjuvant chemotherapy between right-sided and left-sided CC. Then, based on the Laser Capture Microdissection data and numerical simulations, we confirm that the within-sample relative expression orderings (REOs) of genes are highly stable in tissue samples with different cell constituent ratios. Lastly, a predictive signature for the response to fluorouracil-based adjuvant chemotherapy which was insensitive to the cellular composition of tissue sample was developed and validated for the right-sided and left-sided CC, respectively.

基于基因表达谱，已报道了许多结肠癌的各类预后标志，但是尚没有标志能有效应用于临床。我们研究发现以下因素将对结肠癌预后标志的临床转化产生重要影响：(1)多数研究将用药和未用药样本混杂使用，导致预后与耐药标志意义不明；(2)发病位置不同的结肠癌患者存在显著的预后和耐药差异，可能有必要分别识别预后标志；(3)肿瘤组织样本中的上皮和基质细胞的比例依赖于取样部位，因此基因表达水平会受到组织取样部位的影响，存在不确定性。在本课题中，我们拟基于TCGA的多维组学数据，分析左、右结肠癌基因组、表观组与转录组的系统差异，据此揭示左、右结肠癌耐药差异的分子机制；然后，将利用激光微切割数据及仿真实验，证实基因表达值的相对大小秩序关系在细胞构成比不同的组织样本内具有高度的稳定性，并据此设计不受细胞构成比影响的预后标志识别方法；最后，我们将分别筛选稳健的II-III期左、右结肠癌以5氟为基础的术后化疗的耐药标志。

结肠癌临床指南针对II-III期患者提出了多个风险因素，其中包括微转移、分级和微卫星状态等。但是，这些因素的检测存在很大的不确定性，无法有效地预测患者预后及药效。基于基因表达谱，已报道了许多结肠癌的术后复发风险和药效标志。然而，由于基因表达水平会受到组织取样部位的影响，存在不确定性，尚没有标志能有效应用于临床。在本课题中，我们利用对扰动因素稳健的基因间定性信息，构建II-III期结肠癌的微转移标志，用于预测患者术后复发风险。接着，我们还构建了结肠癌的定性转录组分级标志，辅助临床治疗决策。然后，基于左、右结肠癌患者的分子特征差异，我们分别构建左、右结肠癌5-氟尿嘧啶药效标志，并结合TCGA的多维组学数据，揭示左、右结肠癌耐药差异的分子机制；构建预测右侧结肠癌MSI状态的定性标志；基于基因的外显子片段，识别一组在临床上易于操作的、能够准确鉴别右侧结肠癌患者突变负荷的外显子标志。最后，我们利用前期开发的个体化差异基因算法，通过收集大量结直肠癌样本，识别结直肠组织中普遍扰动的基因，并利用蛋白印迹分析证明这些基因是结肠癌重要的诊断生物标志物和治疗靶点。