【作者】 李梅；

【导师】 吕申；

【作者基本信息】 大连医科大学 ， 病理学与病理生理学， 2014， 博士

【摘要】 背景:人类基因组的破译、多种疾病全基因组关联研究(genome wide association studies,GWAS)的开展以及疾病发生分子机制研究的深入使与疾病发生、发展、药物敏感性和预后等因素密切相关的基因变异(突变或多态)逐渐得以揭示,从而使人们对一些疾病的治疗开始步入针对特异性分子变异的分子靶向治疗时代。众所周知,恶性肿瘤的发生及演进常涉及多种基因变异,而每例恶性肿瘤的驱动基因变异又各不相同。因此,治疗肿瘤方式也从同病标准化同治发展为基于特异性基因变异的同病个体化异治或异病个体化同治的分子靶向个体化治疗。研究恶性肿瘤分子特征,寻找驱动性或激活性的分子变异,从而筛选一个或几个分子靶向药物作为治疗以及后备治疗用药物,对预防耐药性产生,制定合理治疗方案具有重要意义。EGFR tyrosine kinase inhibitor(EGFR-TKI)是目前疗效肯定,应用最多的肿瘤分子靶向治疗药物。它不但可用于治疗非小细胞肺癌(non-small cell lung cancer,NSCLC),对NSCLC以外具有药物作用分子靶点的恶性肿瘤具有同样肯定的疗效。虽然国人NSCLC患者中能够被预测对EGFR-TKI敏感程度的人群比例较高,但仍有一半以上患者因缺乏特征性基因变异,对该药物的疗效无法预测。另外,一部分起初对EGFR-TKI敏感的患者也会发生获得性耐药。探讨NSCLC中与EGFR激活性突变相关的分子变化将可能有助于这些问题的解决。近来EGFR-TKI治疗NSCLC以外其他恶性肿瘤的疗效倍受人们关注。多家研究机构发现头颈部癌、胃癌、结直肠癌等恶性肿瘤也存在EGFR突变,且有该突变的肿瘤也对EGFR-TKI敏感。胶质瘤是最常见的中枢神经系统恶性肿瘤,目前尚无对其疗效确切的分子靶向药。胶质瘤较少发生转移,其典型的生物学特征是恶性增殖和复发,且该肿瘤的EGFR常因胞外区编码基因突变(点突变或第2-7外显子缺失突变)、基因异常扩增、蛋白过度表达等被异常激活,胶质瘤是否能用EGFR-TKI治疗值得探讨。分子靶向个体化治疗的关键是确定肿瘤细胞的分子靶点,而最常见的分子靶点类型是基因变异。因此,建立能够有效地检测出基因变异的技术平台是分子靶向个体化治疗得以实施的基础。鉴于肿瘤标本的特殊性,检测方法的灵敏度,特别是从混杂DNA样本检出未知变异能力显得尤为重要。High resolution melting analysis(HRMA)技术是近年发展起来的基因变异检测技术,该技术无需PCR后的电泳分离,也无需荧光标记探针,只需PCR后直接加热使双链DNA解离,通过DNA双链饱和荧光染料记录熔解曲线即可实现对整个PCR扩增片段中基因变异的检测或基因分型,具有检测灵敏、快速、闭管操作、成本低廉,且可以检测到未知和已知基因变异的优点。HRMA技术检测基因变异通过熔解曲线形状或位置(Tm值)的改变来实现,而熔解曲线形状或位置不仅受待测基因自身序列和长度、PCR扩增体系中化学因素影响,还受HRMA技术仪器平台硬件及软件等因素影响。前者可以通过实验设计加以控制,但来自仪器的影响往往难以避免。目前可用于HRMA技术的仪器众多,在检测样品方式和数量、熔解(melting)条件和时间、检测方式、荧光信号采集方式以及分析软件等诸多方面存在较大差异,这些差异是否会影响仪器的基因变异检测准确率目前尚不清楚。因此,比较分析这些检测平台各自的技术特点、建立相应的技术评价体系是能够合理选择检测平台的关键,同时也有助于发现影响HRMA基因变异检测准确率、敏感性的相关因素,使该技术得以不断完善,进而推动其临床应用步伐。快速检测是临床样品检测时必须解决的问题。对于HRMA技术,检测速度加快需要DNA熔解时升温速率增加,然而长期以来存在争议的问题是升温速率增加是否会导致熔解曲线分辨率降低,从而使检测准确率下降。因此,系统研究这些问题对HRMA技术的完善和临床推广应用具有重要意义。方法和结果:一、错配修复基因hMLHl在EGFR突变的非小细胞肺癌高表达及其临床意义收集NSCLC标本181例,应用HRMA方法检测EGFR第19、21外显子和KRAS基因第2外显子突变,应用免疫组织化学方法检测错配修复蛋白hMSH2、hMLH1和细胞增殖标志蛋白PCNA、Ki67表达。NSCLC 中,EGFR 和 KRAS 突变率分别为:36.5%和 5.5%;hMSH2、hMLH1、PCNA 和 Ki67 表达率分别为:59.6%、71.3%、88.4%和 61.9%。NSCLC中,hMLH1过表达和腺癌分别是独立与EGFR突变密切相关的因素(p=0.013,p<0.0005)。随着hMLHl表达增强,EGFR突变率显著增高(p<0.0005)。与无EGFR突变NSCLC相比,EGFR突变NSCLC的hMSH2、hMLH1和PCNA表达率都增高,但Ki67表达率显著下降(p=0.030)。二、BRAF过表达促进胶质瘤恶性增殖不依赖EGFR的激活用免疫组织化学方法检测82例石蜡包埋胶质瘤组织标本EGFR、KRAS、BRAF、PI3K、Ki67 和 phospho-EGFR 蛋白表达,以 phospho-EGFR 表达作为 EGFR活化标志。胶质瘤标本 82例中,EGFR、KRAS、BRAF、PI3K、Ki67 和 phospho-EGFR阳性表达率分别为 75.6%、63.8%、72.0%、68.1%、69.5%和 59.8%。EGFR 活化与非活化胶质瘤的临床病理特征无显著差别。EGFR表达与胶质瘤患者年龄正相关(r=0.270,p=0.014),且与其活化状态显著相关(p=0.029)。在EGFR活化胶质瘤中,EGFR、BRAF及PI3K表达均与肿瘤恶性增殖相关,而在EGFR非活化胶质瘤中,仅BRAF表达与肿瘤恶性增殖显著相关。三、HRMA法在NSCLC患者胸水标本EGFR、KRAS基因突变检测中的应用应用HRMA法检测36例晚期NSCLC患者胸水KRAS基因第2外显子和EGFR基因第19、21外显子突变,并计算突变率。胸水标本36例中KRAS第2外显子突变2例(5.6%),EGFR第19、21外显子突变共10例(27.8%)。前期应用HRMA方法检测181例肺癌组织标本,KRAS突变10例(5.5%),EGFR第19、21外显子突变共66例(36.5%)。HRMA法检测胸水标本得到的KRAS突变率及EGFR突变率与同法检测肺癌组织标本比较均无显著差别(p=1.000 和 p=0.318)。四、常见9种HRMA仪器检测基因变异的技术评价用LCGreen Plus,PCR扩增CPS1基因涵盖同一单核苷酸变异(c.3405-29A>T)3个基因型(A/A,T/T和A/T)的4个不同长度基因片段,将扩增产物作为待检测靶基因片段,用9种HRMA仪器(10种HRMA检测平台)分别进行检测(部分仪器以仪器推荐条件和0.1℃/s升温速率两种条件检测),应用仪器自带软件及公共软件分析检测结果,分别计算杂合子与纯合子的基因分型准确率。总体上,杂合子检出准确率为99.7%(n=2141),而纯合子为70.3%(n=4441)。对纯合子检出准确率单一样品扫描仪器显著好于全盘扫描仪器(分别为78.1%和61.8%,p<0.0005)。51、100、272和547bp四个基因片段的纯合子检出准确率分别为72.2%、83.1%、59.8%和65.9%(p<0.0005)。慢升温速率仪器的纯合子检出准确率并无显著提高。Log(△Tm/TmSD)可以作为仪器的质量控制参数,预测仪器的纯合子检出性能(AUC=0.863,p<0.0005)。五、HRMA升温速度对检测准确率的影响将HR-1的升温速率分别调整为0.01、0.02、0.04、0.08、0.16、0.32、0.64℃/s。在各升温速率条件下分别检测4个不同长度DNA片段的3个基因型样品(基因片段与第四部分相同),每一样品检测3次。以仪器检测温度与样品实际温度间的偏差校正原始数据,然后分析校正后熔解曲线。随着升温速率增大,所有样品熔解曲线均向右(高温方向)平移,片段<300bp杂合子和小片段(51和100bp)纯合子更容易检出。升温速率越慢,片段长度272bp的纯合子越容易检出。升温速率居中(0.8℃/s)时,片段长度547bp的杂合子最难检出,而该长度的纯合子片段最容易检出;升温速率增加或减小均会使杂合子更易检出而纯合子片段变的不容易检出。结论:错配修复基因hMLH1过表达可能成为NSCLC患者对EGFR-TK1敏感的新的分子标志。胶质瘤BRAF过表达不依赖于EGFR的活化,可能作为独立影响因素参与调节胶质瘤恶性增殖,应用针对EGFR、BRAF、PI3K多靶点的分子靶向药物治疗胶质瘤可能优于应用单一靶点的EGFR-TKI。HRMA方法可保持极高的杂合子检出准确率,仪器因素、被检测基因片段长度等因素对检出准确率影响较小;纯合子检出的准确率依据检测仪器、基因序列片段大小及基因变异类型的不同而有所不同。提高升温速率并不意味着降低HRMA方法基因变异检出的灵敏度。HRMA方法敏感、快速、操作简单、无污染,且能够实现对未知和已知突变的检测,适用于临床,尤其适用于主要以杂合性突变为主的肿瘤标本,包括肿瘤患者胸水的检测。如果待测基因片段较短,提高升温速率可以使基因变异检测变得更加快速、敏感。更多还原

【Abstract】 BACKGROUNDS:With the deciphering of the human genome,conducting of genome-wide association studies in many diseases and learning the molecular mechanism of many diseases more in-depth,it is revealed gradually that the effect of genetic variants on disease occurrence,development,drug sensitivity and prognosis.Targeted therapies,based on specific gene variation,opened a new era of disease treatment,especially cancer therapy.Carcinogenesis is a multi-steps process involving multi-genes mutation.The driver or activated mutation is various among different tumors,even among different patients with the same tumor.It is possible that the different modality of treatment is used for the same tumor and the same treatment modality for the different tumor,while not the same standardized treatment for the same tumor like traditional therapy.Therefore,it is necessary to profile tumor molecular characters and provide a specific molecular portrait of each individual in order to aid clinicians in improving diagnosis and tailoring the best therapy to each patient.EGFR tyrosine kinase inhibitor(EGFR-TKI)is one of the most widely used molecular targeted drugs.It has been used in the treatment of non-small lung cancer(NSCLC)with EGFR mutation in exon 19 and 21 with good response.Although there is high proportion of NSCLC patients with EGFR mutation in China,about half of the patients can’t be predicted the response to EGFR-TKI.Moreover some of the patients who have good response to it would obtain resistance after treatment with it for a while.It would be helpful to resolve the problems by profiling the molecular characters related to EGFR mutation in NSCLC and then finding new molecular markers.It is valuable to detect if other tumors can be treated with EGFR-TKI.EGFR mutation has been found in head and neck neoplasm,gastric cancer and colorectal cancer,etc.and the patients with EGFR mutation also had good response to EGFR-TKI.Glioma is the most common malignant tumor of central nervous system.The rampant proliferation and recurrence are its obvious biological characters.Despite the available of many traditional treatment options,the prognosis is often dismal.It is still unclear if EGFR-TKI can be used to treat glioma,though the activation EGFR was often found in glioma,including EGFR mutation in extracellular domain(point mutation and deletion in exon 2 to 7),amplification and over expression.Gene variance detection is the basis of molecular targeted therapy.Compared to germline mutation detection with blood samples of genetic disease,gene variance detection is harder with tumor samples.The proportion of variant DNA is unknown,because normal cells will be mixed unavoidably in the tumor samples.In addition,unknown mutation is often found in tumor samples.It is key to select a gene variance detection method which is sensitive enough and can be used to detect the unknown mutations.High resolution melting analysis(HRMA)is a new gene variance detection technique,which was first introduced in 2003.Because HRMA is a simple,high sensitivity,high speed and low contamination risk method,it is rapidly adapted used into most real-time PCR instruments as well as widely used in research and clinical diagnostics for detecting DNA variants,either known or unknown.Detection of gene variance by HRMA depends on subtle melting curve differences,including shape changes and melting temperature(Tm)shifts.The accuracy and detection sensitivity were affected by chemical factors from PCR,amplicon features such as fragment lengths,gene variance types and sequences,and hardware and software from the instruments.The effect of chemical factors and amplicon features can be controlled by the design of the assay,but the instrument factor is often hard to avoid.There are large differences among HRMA instruments,including sample format,melting rates,detection mode,data density,acquisition mode and software.It is not clear if the differences would affect gene variance detection accuracy.Therefore,it is necessary to select the proper instrument to detect the gene variance by the metric of quality control built depending on comparison the genotyping accuracy among 9 instruments.It would be also helpful to identify those factors that really matter for continued improvement of HRMA.The fast detection speed is often expected when a method used in clinical molecular diagnostics.HRMA may have fast detection speed,but many real-time PCR instrument with HRMA slow the melting rate down to obtain adequate resolution.Because it was thought for a long time that increasing the melting rate means sacrificing the detection resolution.The effect of melting rate on genotyping accuracy is controversial.It is very important to explore the problem systematically in order to make HRMA used in clinical gene variance detection better.METHODS AND RESULTS:1.Expression of the mismatch repair gene hMLH1 is enhanced in NSCLC with EGFR mutationsThe expression of MMR proteins hMSH2 and hMLH1,and the proliferation markers PCNA and Ki67 were measured by immunohistochemistry in 181 NSCLCs.EGFR and KRAS mutations were identified by high resolution melting analysis.In 181 NSCLCs,the frequencies of EGFR and KRAS mutation were 36.5%and 5.5%;the frequencies of hMSH2、hMLH1、PCNA and Ki67 positive expression were 59.6%,71.3%,88.4%and 61.9%,respectively.Stronger hMLH1 expression correlated to a higher frequency of EGFR mutations in exon 19 and 21(p<0.0005).Overexpression of hMLHl and the adenocarcinoma subtype were both independent factors that related to EGFR mutations in NSCLCs(p=0.013 and p<0.0005).The expression of hMLHl,hMSH2 and PCNA increased,while Ki67 expression significantly decreased(p=0.030)in NSCLCs with EGFR mutations.2.BRAF overexpression inducing the rampant proliferation in glioma independent on EGFR activationEGFR,KRAS,BRAF,PI3K,phospho-EGFR and Ki67 expressions were detected with immunohistochemistry in 82 glioma specimens.The positive expression of phospho-EGFR was used as the marker of EGFR activation,or vice versa.In 82 glioma,the frequencies of phospho-EGFR、EGFR、KRAS、BRAF、PI3K and Ki67 positive expression were 59.8%、75.6%、63.8%、72.0%、68.1%and 69.5%.The expression of EGFR positively correlated with patients’ age.There were no significant differences in clinicopathological characteristics between gliomas with and without EGFR activation.EGFR overexpression was significantly related to its activation.In gliomas with EGFR activation,overexpressions of EGFR,BRAF and PI3K were each significantly related to proliferation respectively.However,in gliomas without EGFR activation,only BRAF overexpression was significantly related to proliferation of tumors cells.3.The feasibility of HRMA to detect EGFR,KRAS mutations in pleural effusion samples of NSCLC patientsThe mutations of KRAS gene in exon 2 and EGFR gene in exon 19,21 were detected in 36 pleural effusion samples with HRMA.The difference of EGFR and KRAS mutation frequencies between pleural effusion samples and cancer tissue samples from patients with NSCLC were analyzed with Chi square.There were 2 cases(5.6%)with KRAS mutations and 10 cases(27.8%)with EGFR mutations found in 36 pleural effusion samples with HRMA.The frequencies of KRAS and EGFR mutation,detected in 181 NSCLC tissues before,were 5.5%and 36.5%,respectively.There was no significant difference in either KRAS or EGFR between pleural effusion samples and cancer tissue samples(p=1.000 and p=0.318).4.Genotyping accuracy in nine HRMA instrumentsThree genotype of the single nucleotide variance(c.3405-29A>T)of CPS 1 was amplified by PCR in the presence of LCGreen Plus in four PCR product lengths.After blinding and genotype randomization,samples were melted in 9 instruments(10 instrument configurations)under conditions recommended by the manufacturer.For each configuration and PCR product length,32 to 96 samples(depending on batch size)were analyzed with both commercial and custom software.The accuracy of heterozygote detection and homozygote were assessed.Overall,the heterozygote accuracy was 99.7%(n=2141),while homozygote accuracy was 70.3%(n=4441).Instruments with single sample detection as opposed to full plate imaging better distinguished homozygotes(78.1%and 61.8%,respectively,p<0.0005).PCR products of 51,100,272 and 547 bp had accuracies of 72.2%,83.1%,59.8%and 65.9%,respectively(p<0.0005).Slower melting rates did not improve accuracy.The index,Log(△Tm/Tm SD),was a good predictor of an assay’s ability to distinguish homozygotes(AUC=0.863,p<0.0005).5.The effect of melting rate to genotyping accuracyThree genotype of the single nucleotide variance(c.3405-29A>T)of CPS1 was amplified by PCR in the presence of LCGreen Plus in four PCR product lengths.All samples in triplication were melted in HR-1 with the ramp rate 0.01、0.02、0.04、0.08、0.16、0.32、0.64℃/s,separately.The corrected coefficient was calculated with the difference between the instrument temperature and sample temperature at each ramp rate.All raw data were corrected with the coefficient before analysis.When the ramp rate increased,melting curves of all samples translated to the right(high temperature).Heterozygotes(<300 bp)and homozygotes(51 and 100 bp)were easier to detected with the ramp rate increasing,while the homozygote(272 bp)was easier to be distinguished with the ramp rate decreasing.For 547 bp gragment,heterozygote was the most difficult to detect in the middle ramp rate(0.8℃/s),it was easier with the ramp rate increasing or decreasing.On the contrary,547 bp homozygotes were the easiest to be distinguished in the middle ramp rate,it was more difficult with the ramp rate increasing or decreasing.CONCLUSIONS:Overexpression of hMLHl could be a new molecular marker to predict the response to EGFR-TKIs in NSCLCs.BRAF overexpression would be an independent factor inducing rampant proliferation in gliomas independent on EGFR activation;the molecular target drug with multi-targets of EGFR,BRAF,PI3K could be more effective in treating glioma than EGFR-TKI.HRMA detects heterozygotes with excellent accuracy,but homozygote accuracy is dependent on instrumentation,PCR product size as well as melting temperature difference,and variation within homozygotes.Fast ramp rate is helpful to increase the sensitivity of genotyping of shrot amplicons,either heterozygote or homozygote.HRMA is a simple,sensitive,low contamination risk,high speed method of gene variance(both known and unknown)detection.It is suitable for gene variance detection in clinic,especially for tumor samples including pleural effusion,which is mainly with heterozygous mutations.It is possible to make the detection more fast and sensitive by increasing the ramp rate when the amplicon is not too long.更多还原