基于TCGA数据库的甲状腺癌相关miRNA和基因表达谱的分析及临床相关性的研究

发布时间：2018-04-04 05:34

  本文选题：甲状腺癌　切入点：miRNA　出处：《吉林大学》2016年博士论文

【摘要】：甲状腺癌是世界范围内最常见的内分泌系统恶性肿瘤。甲状腺癌最常见的病理类型是甲状腺乳头状癌(PTC),约占甲状腺癌总数的80%。尽管大多数甲状腺癌具有良好的预后,5年生存率在95%以上,但随着甲状腺癌发病率的逐年增高,促使甲状腺癌发生和发展的分子生物学事件尚不明确,遗传学病因有待进一步研究。目前术前诊断甲状腺癌的金标准是超声引导下细针穿刺细胞学检查(FNAB),但尚有20-30%的FNA结果显示为不确定的/可疑的,这部分患者需要进行诊断性手术以明确恶性或良性。分子标志物的应用将有望协助FNAB提高术前对甲状腺癌的诊断能力。micro RNA(mi RNA)是由18-22个核苷酸组成的单链RNA,通过与靶基因的非转录区结合,调控靶基因的表达,进而影响细胞生化过程和信号传导通路,引起肿瘤的发生和发展。mi RNA在肿瘤发生发展过程中具有关键作用的已达成普遍共识,然而甲状腺肿瘤中mi RNA的表达信息及其与靶基因之间的相互作用尚不明确。据此,本研究纳入了TCGA数据库中499例甲状腺乳头状癌样本和58例甲状腺癌旁对照组织样本,分析了甲状腺癌相关mi RNA和基因的表达谱,得到了38个甲状腺癌相关差异表达的mi RNA和2731个差异表达的基因。为了寻找参与甲状腺癌发生发展的mi RNA和基因,我们对2731个差异表达的基因进行KEGG Pathway聚类分析,得到了82个肿瘤信号通路相关的基因。通过靶基因预测,其中16个mi RNA和20个基因具有靶向调控关系,形成参与甲状腺癌发生发展的调控网络。为了探求具有诊断能力的分子标志物,我们通过绘制ROC曲线对甲状腺癌相关且差异表达的mi RNA和基因进行评估。在差异表达的mi RNA中,上调的mi R-34a、mi R-221、mi R-222、mi R-146b,和下调的mi R-9-1、mi R-486、mi R-451、mi R-144、mi R-9-2、mi R-195,AUC值均大于0.9,并且敏感性或特异性均大于80%,适宜作为诊断标志物;其中mi R-34a的敏感性和特异性均大于90%,是具有突出诊断潜能的分子标志物。在差异表达的基因中,上调的CDKN2B、CCND1、TGFA、ELFN1、BID、DVL1、BCL2L1、RARA、BAX、E2F1、PDGFA、RXRG、MET、LPAR5、ITGA2,和下调的BCL2、TRAF6、GLI1、KIT、LPAR1、WNT11、RARB、PRKX、FGFR2,AUC值高于0.9,并且敏感性和特异性均高于80%,适宜作为诊断标志物;其中,上调的CDKN2B、CCND1和TGFA,敏感性和特异性均高于90%,是理想的独立分子诊断标志物。为了进一步提高分子标志物对甲状腺癌的诊断能力,我们将单个的分子标志物作以组合,组合的形式为mi RNA靶基因的联合,其依据为mi RNA与靶基因通常是负向调控的关系,即若某mi RNA表达上调,其靶基因相应地表达下调;若以二者表达水平值作比,可能将mi RNA上调水平和靶基因下调水平放大,以增加鉴别效能。在筛选出的具有良好诊断能力的mi RNA和基因之间,有20对具有靶向调控关系,其中有18/20的组合AUC值比单独mi RNA或基因有所提高。特别是mi R-34a/BCL2组合,AUC值由单独的0.945和0.955提高到0.981,敏感性由单独的92.2%和88.4%提高到95%,特异性由单独的94.8%和93.1%提高到96.6%,可以认为是一个理想的肿瘤组合标志物。为了探索差异表达的mi RNA和基因是否具有侵袭特征的临床相关性,我们将差异表达的mi RNA和基因逐一进行临床相关性分析,包括腺体外浸润、淋巴结转移和TNM分期等。结果显示,在38个差异表达的mi RNA,mi R-146b、mi R-222、mi R-21、mi R-375、mi R-652与淋巴结转移、腺体外浸润和高TNM分期都相关。mi R-221、mi R-181a-1、mi R-152与淋巴结转移相关;mi R-221、mi R-34a、mi R-424、mi R-138-1、mi R-138-2、mi R-20b、mi R-152、mi R-486、hsa-mir-451、mi R-153、mi R-7-2腺体外浸润相关;mi R-34a、mi R-424、mi R-138-1、mi R-20b、mi R-486、hsa-mir-451、mi R-7-2与高TNM分期相关。在82个差异表达的基因中,MET、TGFBR1、RARA、BAX、LAMC2、BCL2、PDGFRA、BCL2L1、TCF7L1、CDKN2B、KIT、DAPK2、LPAR5、FGF11、PAX8、FGFR2、PRKACB、GNAI1、RUNX1、GNG7、TGFB1、HSP90B1、TRAF6、ITGA3、LAMB3与淋巴结转移、腺体外浸润和高TNM分期都相关。PTCH2、EGF、DVL1、ITGA2B、PRKCA、RXRG、WNT3、AXIN2、TGFA与淋巴结转移相关;DVL1、ITGA2、BID、FZD5、ITGA2B、PTCH2、T GFA、AXIN2、CDKN1A、RARB、WNT11、GLI3、RUNX1T1与高TNM分期相关。综上,本研究基于TCGA数据库的大样本生物信息数据,分析了甲状腺癌相关mi RNA和基因表达谱,并形成了甲状腺癌中mi RNA与靶基因的调控网络。部分差异表达的mi RNA和基因显示出良好的诊断潜能和临床相关性,可在临床实践中指导甲状腺癌的诊断、风险分层、制定治疗策略和评估预后等。
[Abstract]:Thyroid cancer is the most common endocrine system in the world. The most common pathological type of thyroid cancer is papillary thyroid carcinoma (PTC), accounting for the total number of thyroid cancer 80%. although most thyroid carcinoma has a good prognosis, 5 years survival rate is more than 95%, but with the increasing incidence of thyroid cancer. The thyroid cancer occurrence and development of molecular biological events is unclear, genetic etiology needs further study. The current gold standard for preoperative diagnosis of thyroid cancer is the ultrasound guided fine needle aspiration cytology (FNAB), but there is still 20-30% FNA results is uncertain / suspected, these patients require diagnostic the operation to identify benign or malignant. The application of molecular markers is expected to help FNAB RNA.Micro diagnosis for thyroid cancer to improve the preoperative (MI RNA) is composed of 18-22 nucleotides. The single stranded RNA, through the combination of untranslated region and target genes, regulating the expression of target genes, thereby affecting cellular biochemical processes and signal transduction pathways, which plays a key role in tumor occurrence and development of.Mi RNA in tumor development process has reached a general consensus, but the interaction between the expression of MI in thyroid tumor information RNA and its target gene is not clear. Therefore, the research into the TCGA database in 499 cases of thyroid papillary carcinoma samples and 58 cases of thyroid cancer adjacent normal tissue samples, analysis of RNA and MI for radical related thyroid cancer expression, obtained the expression of MI RNA expression in 38 thyroid cancer related differences and 2731 genes involved in MI and RNA. In order to find the genetic development of thyroid carcinoma, we analysis KEGG Pathway cluster on the expression of 2731 genes differentially, obtained 82 tumor signal Pathway related genes. The target gene prediction, with 16 mi RNA and 20 genes, which targeted regulation, regulatory networks involved in the formation of the occurrence and development of thyroid cancer. In order to explore the molecular markers with diagnostic capabilities, we assessed by MI and RNA gene expression and the difference of thyroid cancer ROC curve in the MI RNA. The differential expression, up regulation of MI R-34a, MI R-221, MI R-222, MI R-146b, MI R-9-1 MI and down-regulation of R-486, MI, R-451, MI, R-144, MI, R-9-2, MI, R-195, AUC values are greater than 0.9, and the specificity and sensitivity were higher than 80%, suitable as a diagnostic marker MI; the sensitivity and specificity of R-34a were greater than 90%, is a molecular marker with prominent diagnostic potential. The differentially expressed genes, upregulation of CDKN2B, CCND1, TGFA, ELFN1, BID, DVL1, BCL2L1, RARA, BAX, E2F1, PDGFA, RXRG, MET, LPAR5, ITGA2, and down the BCL2, TRAF6, GLI1, KIT, LPAR1, WNT11, RARB, PRKX, FGFR2, AUC higher than 0.9, and the sensitivity and specificity were higher than 80%, suitable as a diagnostic marker; wherein, upregulation of CDKN2B, CCND1 and TGFA, the sensitivity and specificity were higher than 90%, is the ideal independent molecular markers. In order to further improve the molecular markers of thyroid cancer diagnosis ability, we will be single molecular markers in combination, combination of the form of joint mi RNA target gene, which is based on MI RNA and target gene is usually negative relation to regulation of expression, if a mi RNA, the corresponding target genes the expression levels of expression in the two cut; value ratio, MI may upregulate RNA levels and lower levels of target gene amplification, in order to increase the identification efficiency. Among the screened with good diagnostic ability of MI and RNA gene, 20 of the targeted regulation relationship, including 18/20 The combined AUC value increased more than Mi alone or RNA gene. Especially mi R-34a/BCL2, the AUC value from 0.945 separate and 0.955 increased to 0.981, the sensitivity increased from 92.2% and 88.4% to 95% alone, the specificity increased by 94.8% and 93.1% to 96.6% separate, can be considered an ideal combination of tumor markers in order to explore the differences of expression. Whether the MI gene has RNA and clinical relevance of invasive features, we will mi RNA and differential gene expression by clinical correlation analysis, including glandular infiltration, lymph node metastasis and TNM stage. The results showed that MI RNA, the 38 differentially expressed mi R-146b, MI R-222. Mi R-21, MI R-375, MI R-652 and lymph node metastasis, infiltration glands and high TNM stage are.Mi R-221, MI R-181a-1, MI R-152 was correlated with lymph node metastasis; MI R-221, MI R-34a, MI R-424, MI R-138-1, MI R-138-2, MI R-20b, MI R-152 ,mi R-486,hsa-mir-451,mi R-153,mi R-7-2鑵轰綋澶栨蹈娑︾浉鍏，

本文编号：1708586