【作者】 江本元；

【导师】 吴一龙；

【作者基本信息】 南方医科大学 ， 肿瘤学， 2010， 博士

【摘要】 研究背景肺癌居我国恶性肿瘤死因首位,可分为非小细胞肺癌(non-small cell lung cancer, NSCLC)和小细胞肺癌(small cell lung cancer, SCLC)两大类,其中非小细胞肺癌约占80%以上。尽管肺癌的早期诊断和规范化治疗都取得了一定的进步,但只有15-25%的非小细胞肺癌有手术治疗的机会,而即使接受根治性手术治疗,仍然有50%的NSCLC患者会复发。传统的多学科综合治疗(手术,放疗和化疗)在提高非小细胞肺癌的五年生存率方面已经出现了“天花板效应”。随着近几年转化医学的出现,将肺癌基础研究的结果向临床诊疗转化,开发了一些靶向治疗药物,取得一些令人鼓舞的结果。EGFR突变及其分子构象改变与靶向抑制剂的关系即是近年来在肺癌转化研究方面突出的例子。对于EGFR敏感突变的特定人群,酪氨酸蛋白激酶抑制剂吉非替尼和厄洛替尼能明显延长晚期NSCLC患者的无复发生存期。但这些治疗手段最终都会出现耐药,导致治疗失败,归根结底是由于对NSCLC的疾病机制不清楚。非小细胞肺癌是一种异质性疾病,肺癌发生发展过程中存在多层次的基因表达调控异常,包括基因组DNA水平多种基因的体细胞遗传变异、转录水平基因表达异常及转录本剪切异常、microRNA表达调控异常,进而表现为蛋白质水平的功能异常。越来越多的分子表达调控异常与肿瘤特定的细胞周期、生长分化、细胞移动、凋亡调节等生物学过程改变相关。阐明具有驱动作用的特定基因表达调控变异机制有助于阐明肿瘤特定阶段生物学行为,并有助于针对变异靶点进行临床干预。原癌基因活化是许多恶性肿瘤包括肺癌发生的一个重要机制,正常情况下,原癌基因编码一些生长因子和受体,通过严密的调控网络,控制机体正常的生长与凋亡。多种机制包括点突变、基因融合与易位、基因扩增等都能导致原癌基因异常活化,使细胞获得超常的增殖能力和存活能力,从而促进肿瘤的发生。大量的分子遗传学研究已经显示在肺癌中主要有以下染色体片段存在扩增,包括1p (NRAS, MYCL1,PAX7、3q(EIF4G1, TP73L, PIK3CA)、5p(SKP2)、7p(EGFR)、8q(PTK2、MYC)、11q(CCND1、BIRC2、BIRC3)和20q(E2F1)。另外,如EGFR和KRAS的点突变、EML4-ALK融合基因与肺癌发生的关系也已经得到证实。与抑癌基因不同的是,癌基因发挥功能是显性作用的,即一对等位基因其中的一个基因存在活化,就能发挥恶性转化细胞的功能。但是,这些经典的遗传学并不能解释所有的现象,近几年来兴起的表观遗传学让我们对非小细胞肺癌有了进一步的认识,如p16基因甲基化与鳞癌的发生有着密切的关系,Bhutani等人还发现p16基因甲基化与早期肺癌的预后相关,这些可遗传的基因异常并不是由于DNA序列本身发生任何改变。常见的表观遗传学修饰包括甲基化,组蛋白乙酰化等,近几年来,科学家们发现了一种新的表观遗传学修饰-microRNA调控,在基因表达调控中发挥着重要的作用。微小RNA (microRNA,miRNA)分子是近年来发现具有调控蛋白翻译功能的内源性非编码RNA分子,长约22nt。最早microRNA是由Lee等在秀丽线虫体内发现Lin-4,目前在包括人类、果蝇等多个物种鉴别出数千个microRNAs(详见microma.sanger.ac.uk)。成熟的microRNA可通过互补或不完全互补的方式与mRNA相应的区域结合,从而抑制蛋白质的翻译,在转录后水平调节靶基因的表达。多种microRNA已被证实参与白血病、淋巴瘤、肺癌、乳腺癌等多种肿瘤发生。MicroRNA参与肿瘤发生发展机制复杂多样,microRNA异常表达可能上调原癌基因或下调抑癌基因,也可能因自身突变或拷贝数扩增导致调控功能异常。从microRNA水平探索肿瘤的异常翻译后调控机制正成为肿瘤分子遗传学的新的研究热点。在肺癌研究中,已发现多个microRNA分子与肺癌相关。Takamizawa等首次发现了let-7在肺癌中低表达,并且低表达的患者预后较差。Johnson等确定Ras受let-7家族的调控,其中包括线虫的RAS基因、人的KRAS、HRAS和NRAS基因,且let-7表达量与生存预后相关,体外实验也证实了过量表达let-7可抑制肺癌生长。miR-128可直接调节表皮生长因子受体(EGFR)的水平,并与靶向药物EGFR TKI (Gefitinib)疗效相关。随着研究进展,越来越多的microRNA分子被发现与肺癌发生及预后等相关。尽管肺癌相关microRNA资料越来越多,但多数microRNA异常调控的下游分子机制不清楚。如何确定microRNA调节的靶基因、并将microRNA表达变异与特定的基因表达水平变化有机的联系起来并加以验证是基因表达调控研究中新的课题,也是生命科学研究中一项繁重的任务。研究目的本研究通过高通量的芯片技术筛选非小细胞肺癌与癌旁组织中差异表达的基因,找到一种差异表达的基因BCL11A。接着用实时荧光定量PCR实验和免疫组织化学技术分别在基因水平和蛋白水平验证芯片的结果,并将BCL11A表达水平与临床病理资料进行关联分析,以探明其在非小细胞肺癌发生,发展中的作用。最后从基因拷贝数变化和microRNA调控两方面探讨其异常表达的机制。实验设计1.我们选取2003-2006年在广东省肺癌研究所接受根治性手术治疗的114例非小细胞肺癌患者和其中53例癌旁肺组织,应用基因芯片技术分析非小细胞肺癌与癌旁组织之间的基因表达差异,结果发现一种原癌基因BCL11A在非小细胞肺癌组织中的表达明显高于癌旁组织。2.为了证实这种差异,本研究应用实时荧光定量PCR和组织芯片免疫组织化学技术分别在基因水平和蛋白水平对差异表达的基因BCL11A进行验证,并将BCL11A表达水平与临床病理资料和患者预后进行关联分析,以明确BCL11A在非小细胞肺癌发生,发展中的作用。3.为了弄清BCL11A在NSCLC中上调的机理,本研究借助比较基因组杂交技术以期从基因组水平了解BCL11A基因拷贝数变异情况。4.另外,近几年发现的一类称之为微小RNA(microRNA)的非编码RNA,能够在转录后水平调节目的基因的表达,广泛参与非小细胞肺癌进程。本研究应用Agilent人类microRNA表达芯片分析37对配对的非小细胞肺癌与癌旁组织之间的microRNA表达差异,找到在非小细胞肺癌组织中下调的microRNAs。同时通过miRGen生物学软件分析,找到潜在调控BCL11A的microRNA分子集,然后与在非小细胞肺癌组织中下调的microRNA分子集进行交集运行,找到在非小细胞肺癌组织中表达下调同时又是潜在调控BCL11A的microRNA分子。5.探讨这些microRNA分子的功能,并在分子水平研究验证microRNA对靶基因BCL11A的调控。结果1.基于基因芯片数据发现BCL11A在非小细胞肺癌组织中的表达水平高于癌旁组织,平均上调倍数为3.9。2.应用实时荧光定量PCR验证mRNA芯片实验,结果发现无论是比较114例非小细胞肺癌组织与42例癌旁组织还是比较42对配对组织,BCL11A mRNA在非小细胞肺癌组织中的表达水平均高于癌旁组织,差异有统计学有意义(P=0.004),证实芯片实验结果的正确性。BCL11A mRNA水平与疾病分期(P=0.040)和淋巴结状态(P=0.013)有关,疾病分期越早,淋巴结未受累的患者,BCL11A mRNA水平越高。BCL11A mRNA高表达的NSCLC患者无论是总生存期还是无病生存期均优于BCL11A低表达患者,Cox回归分析还发现BCL11A mRNA水平是预测早期NSCLC患者术后无病生存期(DFS)的独立预后因子。3.运用组织芯片免疫组织化学技术进行BCL11A蛋白表达的检测,结果发现BCL11A在113例非小细胞肺癌组织中的阳性表达率为70.8%,而在25例癌旁组织中均不表达。BCL11A蛋白在肺腺癌、鳞癌和大细胞肺癌中的阳性表达率分别为63.2%(43/68),85.7%(31/36)和66.7%(6/9)。另外一个有趣的现象是BCL11A蛋白在不同组织学类型中具有不同的亚细胞定位,在肺腺癌中主要在胞浆表达,而在鳞癌中主要在细胞核表达。接着将BCL11A表达水平与临床资料关联分析,结果发现BCL11A表达水平与疾病分期和淋巴结状态呈负相关,BCL11A在非小细胞肺癌组织中表达水平升高,但相对早期而言,BCL11A在晚期NSCLC患者中的表达水平较低,与之前的荧光定量PCR实验结果一致。进一步分析BCL11A表达水平与患者生存时,发现高表达BCL11A的患者预后优于BCL11A低表达的患者(P=0.051),而且在早期病例(IA-IIB期)中这种现象更明显(P=0.028)。Cox回归分析发现BCL11A蛋白表达能独立预测NSCLC患者的术后DFS [P=0.002, HR 0.396,95%(CI) 0.219-0.715]。另外,我们发现在早期患者中,BCL11A蛋白表达不仅能预测DFS还能预测OS,而在晚期病例中,没有观察到BCL11A蛋白表达与预后有任何相关性。对于早期NSCLC患者,无论是总生存期还是无病生存期,BCL11A高表达的患者预后均优于BCL11A低表达的患者。4.为了弄清BCL11A在非小细胞肺癌组织中高表达的内在机理,我们应用比较基因组杂交技术检测非小细胞肺癌基因组水平变异情况。实验结果发现在鳞癌中BCL11A在基因组水平明显有扩增,平均信号强度为0.345,而在腺癌和大细胞肺癌中的信号强度值分别为0.026和0.230。将BCL11A基因拷贝数变化与BCL11A mRNA水平进行相关分析,结果发现在鳞癌中BCL11A mRNA水平与BCL11A基因拷贝数呈正相关(rs=0.346,P=0.041),提示在鳞癌中BCL11A基因异常表达可能是由于基因扩增所致,但在腺癌和大细胞肺癌中BCL11A上调的机理仍不清楚。5.同时,本研究用microRNA表达芯片比较37对非小细胞肺癌与癌旁组织microRNA表达差异,结果发现多种microRNAs包括let-7家族、mir-30家族、mir-34家族和mir-1在NSCLC组织中均明显下调,接着用基于Taqman探针的实时荧光定量PCR成功验证了microRNA芯片的结果。通过生物学软件分析,成功预测出潜在调控BCL11A转录后翻译的microRNA集合,与在非小细胞肺癌组织中表达下调的microRNA集合运行交集,找到几种调控BCL11A的microRNAs,如miR-1、miR-30a、miR-103和miR-217等,其中miR-1和miR-30a在非小细胞肺癌组织中下调最明显,于是,本研究选取这两种microRNAs进一步研究,我们发现miR-30a而不是miR-1能明显转录后抑制BCL11A翻译。初步结论1.BCL11A在非小细胞肺癌组织中的表达无论是mRNA水平还是蛋白水平均明显高于癌旁组织,提示原癌基因BCL11A在非小细胞肺癌中存在异常活化。2.BCL11A蛋白特异性地表达于非小细胞肺癌组织,而在癌旁正常组织中不表达,提示BCL11A蛋白可能作为一种非小细胞肺癌特异的肿瘤标志物。3.BCL11A在非小细胞肺癌中上调存在着一定的组织特异性,在鳞癌中上调最明显。4.BCL11A mRNA水平和蛋白表达水平均与NSCLC患者疾病分期和淋巴结状态有关,疾病分期为早期(Ⅰ-Ⅱ)、淋巴结未转移的患者,BCL11A表达水平越高。5.BCL11A mRNA或BCL11A蛋白高表达的非小细胞肺癌患者无论是总生存期还是无病生存期均明显优于BCL11A低表达患者,BCL11A mRNA表达水平能独立预测早期非小细胞肺癌患者术后无病生存期。而BCL11A蛋白表达水平不仅能预测早期NSCLC患者的无病生存,还能独立预测早期NSCLC患者的术后总生存,提示BCL11A蛋白表达作为早期NSCLC术后的预后指标更有意义。6.非小细胞肺癌与癌旁组织间的microRNA表达水平存在明显差异,多种microRNAs包括miR-1和miR-30a非小细胞肺癌组织中明显下调。7.在鳞癌中,BCL11A高表达的原因可能主要是由于基因组水平扩增所致,但miR-30a等microRNAs的下调,可能是原癌基因BCL11A在NSCLC中的活化的另一种机制。课题特色和创新之处1.应用多种芯片技术在不同层面上系统分析了BCL11A在非小细胞肺癌与癌旁组织中的表达差异,成功发现一种差异表达的原癌基因BCL11A;2.成功发现一种特异性较好的非小细胞肺癌标志物BCL11A,为NSCLC提供一种潜在的诊断和治疗靶点；3.通过BCL11A,我们将microRNA表达和基因表达联系起来,并提出microRNA失活是原癌基因BCL11A在非小细胞肺癌中异常活化的机制之一。更多还原

【Abstract】 BackgroundLung cancer is the leading cause of cancer related death in China,including non small cell lung cancer(85%) and small cell lung cancer(15%) Despite much advance in the early diagnosis and standard treatment, the prognosis of NSCLC patients is still poor with 5-year survival less than 15%[25]. Only 15-25% of all NSCLC patients are candidates for surgery. Moreover, even in this selected population, approximately 50% of the patients will relapse after complete resection. Conventional multimodality therapies (surgery, radiation and chemotherapy) have reached a therapeutic ceiling in improving the five year overall survival rate of non-small cell lung cancer (NSCLC) patients. With the emerge of translational medicine in recent years,many basic researchs had worked from bench to beside,resulting in exploring many target therapy drugs and getting challenging effects. The most representative example in lung cancer translational research is about EGFR mutation and TKI respons[2]. For NSCLC patients with sensitive mutation, TKI gefitinib and erotinib can prlong the progress free survival of advanced stage patients.[3,4].Unfortunately,these therapeutic strategies unavoidably induce drug resistance and result in therapeutic failure, ultimately due to poor understanding of the disease and its resistance to the therapy.Non-small cell lung cancer is characterized by a variable clinical course and known for its unpredictable behavior. Molecular epidemiological studies have provided evidence that this disease arised through different molecular mechanisms including hereditarily somatic variation of genome DNA, abnormal gene expression in transcriptional level and transcript splice,dysregulation of microRNA,and procedurely dysfunction of protein expression. However, these molecular alterations were not enough to explain the heterogeneity of NSCLC[5,6].More and more molecular abnormality be found associated with specific tumor cell cycle,tumor differentiation,cell migration,apotosis and other biological process[7]. Clarity specific mechanism of gene expression regulation can be helpful for understanding tumor biological activity,and for clinical interference of abnormal target.Activation of proto-oncogene act as an important mechanism in tumorigensis including lung cancer.Proto-oncogene code many growth factor and receptor maintain normal growth and apotosis via accurate regulation network in normal condition[8].Several mechanisms including point mutation[9], gene fusion and chromosome translocation[10],gene amplication[11] can result in abnormal activation of proto-oncogene, acquire stronger capacity of proliferation and survival and promote tumorigenesis. In lung cancer, amplified oncogenes are present on 1p (NRAS, MYCL1, PAX7),3q(EIF4G1, TP73L, PIK3CA),5p (SKP2),7p (EGFR),8q (PTK2, MYC), 11q (CCND1,BIRC2, BIRC3) and 20q (E2F1). Important role of EGFR, k-ras mutation and EML4-ALK fusion gene in lung cancer carcinogensis haved been validated. Different from tumor surpressor, oncogene function dominantly, can malignantly transform cell only one allele activiated. However, these molecular alterations were not enough to explain the heterogeneity of NSCLC. The concept of epigenetics offers a partial explanation of the phenomena,such as methylation of p16 gene associated with squamous cell carcinoma and prognosis of early stage NSCLC patients[14,15].These heritable changes in gene expression that are not due to any alteration in the DNA sequence[14]。The best-known epigenetic alterations are methylation and histone deacetylation, a striking epigenetic modification called microRNA play a important role in gene regulation[16]. MicroRNAs constitute a large family of small, approximately 21-nucleotide-long, non-coding RNAs that have emerged as key post-transcriptional regulators of gene expression in metazoans,plants and mammals. Lin4 is the first microRNA founded in nematode C. elegan, thousands of microRNAs have been identified in human, drosophila et(microrna.sanger.ac.uk).The mature microRNA inhibit protein translation via binding to specific region of mRNA by perfect or incomplete complementary. Several microRNA had been implicated in carcinogenesis including leukemia,lymphoma,lung cancer and breast cancer.Deregulated microRNAs can upregulate the oncogene expression and downregulate suppressor, result in dysfunction due to mutation or copy number variation.To explore the post-translated regulation mechanism of tumor tumorigenesis have being a research focus of tumor molecular genetics.Many microRNAs contributed to lung cancer initation and development, Takamizawa identified that let-7 was downregulated in NSCLC and correlated with poor pgognosis.Johnson found Ras family including K-Ras、H-Ras and N-Ras, were the target gene of let-7, they also identified that let-7 can repress tumor growth and associated with prognosis.MiR-128 can inhibit EGFR protein expression and LOH of mir-128 correlated with Gefitinib response.More and more microRNAs have been found associated with lung cancer carcinogenesis and prognosis,but the accurate mechanisms of downstream regulation remained unclear.To the summary, predicting and validating the target genes of microRNA, and establishing the stable relationship between microRNA and gene regulation are an important and promising work.PurposeThis study found a BCL11A gene which have different expression level between NSCLC and adjacent non-cancerous tissues by high mRNA array screening. Second,we used real-time PCR to validate the mRNA array result, and link the BCL11A mRNA level and clinicopathological factors to explore the role of BCL11A in NSCLC initation and development. Lastly, we investigated the mechanism of abnormal BCL11A expression turned to copy number and microRNA regulation.Experiment Design1.Surgical specimens were obtained with informed consent from 114 NSCLC patients (81 males and 33 females) who underwent potentially curative surgery at the Gangdong General Hospital (GGH) between 2003 and 2006. This study was approved by the Institutional Review Board (IRB) of Guangdong General Hospital. We found a ectopic BCL11A gene between NSCLC and adjacent non-cancerous tissues by mRNA array.2.To validate this different expression,we detect BCL11A expression both in mRNA and protein levels, and link BCL11A level with patient’s clinicopathological factors and prognosis.3.To investigate the mechanism of BCL11A deregulation in NSCLC,we performed aCGH to evaluate BCL11A gene copy number variance.4.A non-coding RNA called microRNA can regulate gene expression post-transcriptionally and contribute lung cancer tumorigenesis. This study analysized the different microRNA expression of 37 paired NSCLC and adjacent non-cancerous tissues via Agilent human microRNA expression array. We found several microRNAs which are downregulated in NSCLC, on the other hand, we search potential microRNAs that regulated BCL11A protein and conduct intersection set.5. Exploring the fuction of these microRNAs and validate the relationship between microRNAs and BCL11A gene.Result1.BCL11A gene was upregulated in NSCLC based on Affymetrix GeneChip(?) platform,the average folds were 3.9.2.To validate the up-regulated BCL11A expression in NSCLC, Taqman(?) real time qRT-PCR assays were used to detect the same set of total RNA samples used in Affymetrix GeneChip(?) tests. Again, BCL11A mRNA levels were found significantly different between 114 NSCLC tissues and 42 adjacent normal tissues, with up- regulation in cancer tissues (P=0.004). In the 42 pairs of cancer versus non-tumor adjacent tissues, BCL11A expression at mRNA level was also up-regulated in cancer tissues(P=0.024). BCL11A mRNA have a minus relationship with disease stage (P=0.040) and lymph node status(P=0.013). Patients with an early disease stage and no lymph node involvement have a higher BCL11A mRNA expression. Furthermore, high BCL11A mRNA levels correlate with better overall survival and longer disease free survival.Cox analysis result demonstrated that BCL11A mRNA were an independent prognosis factor of desease free survival for early stage NSCLC patients.3.Expression of BCL11A at the protein level was investigated in 113 NSCLC tissues and 25 non-cancerous adjacent lung tissues by immunohistochemistry. BCL11A protein staining was not found in the 25 adjacent normal lung tissues.80 out of 113 NSCLC patients had positive staining for BCL11A protein. The positive expression rates in adenocarcinoma, squamous carcinoma and large cell carcinoma were 63.2% (43/68),85.7%(31/36) and 66.7%(6/9) respectively. Interestingly, BCL11A localized predominantly in nucleus of squamous cell carcinoma, however, in adenocarcinoma, the immunohistochemical staining was detectable predominantly in cytoplasm of cancer cells. Subsquently,we linked BCL11A protein level with clinicopathological factors and found that BCL11A protein levels BCL11A mRNA have a minus relationship with disease stage (P=0.029) and lymph node status(P=0.029). Patients with an early disease stage and no lymph node involvement have a higher BCL11A protein expression. These results coincide with previous mRNA results.Similarily, high BCL11A protein levels correlated with better overall survival and longer disease free survival. Patients with higer BCL11A protein expression had longer overall survival(P=0.051),escepically for early stage patients(P=0.028). Cox analysis result demonstrated that BCL11A protein levels were an independent prognosis factor of desease free survival for NSCLC patients [P=0.002, HR 0.396,95%(CI) 0.219-0.715]. Moreover, we found that BCL11A protein levels can not only predict disease free survival but also predict overall survival in early stage NSCLC patients,however, there were not correlation between BCL11A protein and prognosis in advanced stage patients.5.To investigate the mechanism of BCL11A deregulation in NSCLC,we performed aCGH to evaluate BCL11A gene copy number variance. The result indicated that BCL11A gene ampilified frequently in squamous cell carcinoma,but not in adenocarcinoma or large cell carcinoma,the normalized intensity of probes in squamous,adenocarcinoma and large cell carcinoma were 0.345,0.026 and 0.230 respectively. We also associated BCL11A gene CNV with BCL11A mRNA level, the result suggested BCL11A mRNA level had a positive correlation with BCLl 1A CNV in squamous cell carcinoma (rs=0.346, P=0.041).These data suggest that BCL11A gene overexpression in squamous cell carcinoma may result from genomic CNV, but the mechanism for adenocarcinoma and large cell carcinoma remain unclear.6. Meanwhile, this study compared microRNAs profiles between 37 NSCLC tissue and adjacent non-cancerous tissues by Agilent microRNA array. The result demonstrated many microRNAs including let-7 family,mir-30 cluster,mir-34 family and mir-1 were downregulated in NSCLC. Subsequently, we validated this result by Taqman real-time PCR successfully. We predicted the microRNA clusters which can potentially regulate BCL11A gene translation,and conduct intersection with microRNA clusters downregulated in NSCLC, several microRNAs can meet the requirement including miR-l,miR-30a,miR-103 and miR-217. Among these microRNAs, miR-1 and miR-30a were selected to further analysis, the result demonstrated that miR-30a but not miR-1 can suppress BCL11A gene translation post-transcriptionally.Primary Conclusion1. BCL11A is upregulated both in mRNA and protein levels,indicating that proto-oncogene BCL11A is activiated in NSCLC.2.BCL11A expressed exclusively in NSCLC tissues but not in adjacent non-cancerous tissues,suggesting that BCL11A might be a special tumor marker.3.The deregulation of BCL11A in NSCLC have a significant histological difference,BCL11A levels in squamous carcinoma are much higher than other types. 4.BCL11A associasted with disease stage and lymph node stage both in mRNA and protein levels.5.Patients with high BCL11A expression both in mRNA or protein levels have a better prognosis BCL11A mRNA is a independent prognosis for disease free survival, but BCL11A Protein can not only predict disease free survival but also predict overall survival.Which suggested that BCL11A can be a better prognosis factor than BCL11AmRNA.6.MicroRNAs differently expressed between NSCLC and adjacent non-cancerous tissues, many microRNAs are downregulated in NSCLC tissues.7.BCL11A overexpression in squamous cell carcinoma mostly result from BCL11A gene copy number gain,however, hsa-miR-30a inactivation might be another mechanism for activation of proto-oncogene BCL11A.更多还原