基于药物诱导的肿瘤耐药细胞系识别临床相关的耐药基因

发布时间：2018-11-20 09:10

【摘要】：耐药性的出现是目前癌症治疗中存在的一大难题。为了分析癌细胞的耐药机制及筛选耐药相关基因,一种最常用的实验策略是通过持续增加药物浓度处理亲本细胞系来产生耐药细胞株,并检测相应的基因表达谱数据。大量的研究通过筛选药物诱导的耐药细胞系与亲本细胞系间的差异表达的基因(Differentially expressed genes,DEGs)来识别耐药相关基因及生物学通路。但是基于耐药细胞系模型筛选的耐药相关基因常难以应用于指导临床用药。药物诱导的耐药细胞系与亲本细胞系间的差异表达的基因可能主要反映细胞系对药物处理产生的应答变化,而不一定与细胞系耐药相关。因此,有必要通过临床组织样本来评价基于细胞系耐药模型筛选的耐药基因标志的临床相关性。此外,由于同一种细胞系无生物学差异,因此研究者通常只检测亲本与耐药细胞系各少数几个(如2-3个)技术重复样本。对此类小样本数据,研究者普遍按基因在两类细胞中平均表达值之间的倍数变化(Fold Change,FC)排秩筛选差异表达基因。显然,在对照组中高表达的基因难以有较大的倍数增加,因此难以被FC排秩方法判断为差异高表达;而在对照组中表达量低的基因则容易受检测偏差的影响而随机产生较大的倍数变化,导致出现较多的假阳性结果。针对FC方法存在的偏倚问题,已有研究工作者提出了根据基因在两类样本中的平均表达差值(Average Difference,AD)大小排秩及加权平均差异排秩的方法,以识别那些在两类样本间表达水平差值较大而改变倍数不大的基因。因此,有必要评价AD方法用于识别小样本细胞系实验研究的意义。本文将接受基于氟尿嘧啶(5-FU)与奥沙利铂(L-OHP)联合化疗的临床非响应-响应结肠癌患者的化疗前取样的组织样本间的差异基因定义为临床耐药相关基因(clinically relevant drug resistance genes,记为CRG5-FU/L-OHP),并以CRG5-FU/L-OHP基因为参照,用于评价细胞耐药基因的临床相关性。我们利用结肠癌细胞系HCT116与其经氟尿嘧啶、奥沙利铂分别诱导的耐药细胞系的基因表达谱数据,证实:(1)耐药细胞系与亲本细胞系之间的差异表达基因主要反映药物处理亲本细胞系后引起的变化,其大多数可能并不与耐药相关;(2)反映亲本及耐药细胞系对药物短暂处理(如24h)诱导的差异表达基因与CRG5-FU/L-OHP基因显著吻合,并且与化疗后取样的临床非响应-响应组织样本间的差异基因显著吻合;(3)不同于FC排秩方法,采用AD排秩方法可以识别出在两类细胞中表达丰度高的差异基因,它们与CRG5-FU/L-OHP基因显著一致。本文提出了新的基于耐药细胞系模型识别肿瘤耐药标志的实验设计与数据分析策略。
[Abstract]:The emergence of drug resistance is a major problem in cancer treatment. In order to analyze the mechanism of drug resistance of cancer cells and to screen drug-resistance related genes, one of the most commonly used experimental strategies is to produce drug-resistant cell lines by continuously increasing drug concentration to produce drug-resistant cell lines, and to detect the corresponding gene expression profile data. A large number of studies have been conducted to identify drug-resistance-related genes and biological pathways by screening differentially expressed genes (Differentially expressed genes,DEGs between drug-induced drug resistant cell lines and parental cell lines. However, drug resistance related genes based on drug resistance cell line model are difficult to be used to guide clinical drug use. The differentially expressed genes between drug-resistant cell lines and parent cell lines may reflect the response of drug resistant cell lines to drug treatment, but may not be related to drug resistance of cell lines. Therefore, it is necessary to evaluate the clinical correlation of drug resistance gene markers based on cell line resistance model through clinical tissue samples. In addition, because there is no biological difference in the same cell line, researchers usually detect only a few (e.g. 2-3) technical repeat samples of each parent and drug-resistant cell line. For such small sample data, the researchers generally ranked the multiple variation (Fold Change,FC) between the average expression values of genes in the two types of cells to screen differentially expressed genes. It is obvious that the high expression genes in the control group are difficult to increase by a large multiple, so it is difficult to be judged as differential high expression by the FC rank method. However, the genes with low expression in the control group were susceptible to the influence of the detection deviation, resulting in a large number of random changes, resulting in more false positive results. In view of the bias problem of FC method, some researchers have proposed a method to rank the genes according to the average expression difference (Average Difference,AD) size and weighted average difference in the two kinds of samples. In order to identify those between the two types of samples between the level of expression difference and the change is not multiple genes. Therefore, it is necessary to evaluate the significance of AD method in identifying small cell lines. In this study, the difference gene between tissue samples taken before chemotherapy in patients with nonresponsive colon cancer was defined as clinical drug-resistance correlation based on combination chemotherapy of fluorouracil (5-FU) and oxaliplatin (L-OHP). Gene (clinically relevant drug resistance genes, To evaluate the clinical association of cell resistance genes with CRG5-FU/L-OHP as reference. We used the gene expression profile data of colon cancer cell line HCT116 and its drug-resistant cell lines induced by fluorouracil and oxaliplatin respectively. The results showed that: (1) the differentially expressed genes between drug-resistant cell lines and parental cell lines mainly reflect the changes caused by drug treatment of parent cell lines, and most of them may not be related to drug resistance; (2) the differentially expressed genes induced by transient treatment (such as 24 h) by their parents and drug resistant cell lines were significantly consistent with those of CRG5-FU/L-OHP gene. And the difference gene between the clinical non-response-response tissue samples sampled after chemotherapy was significantly consistent. (3) different from the FC method, AD rank method can identify the differentially expressed genes in the two types of cells, which are significantly consistent with the CRG5-FU/L-OHP gene. In this paper, a new experimental design and data analysis strategy for the identification of tumor resistance markers based on drug resistant cell line model is proposed.
【学位授予单位】：福建医科大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：R730.5

【相似文献】