【作者】 李敏；

【导师】 程书钧； 高燕宁；

【作者基本信息】 中国协和医科大学 ， 肿瘤学， 2008， 博士

【摘要】 研究背景目前,肺癌已经成为我国死亡率最高的恶性肿瘤。肺癌的早期诊断是保证临床治疗成功和提高生存率的关键一步。外周血血浆中含有全身组织来源的蛋白,机体器官的病理改变可表现为血浆蛋白在结构和数量上的改变。这类特征性的变化有可能作为微创伤性实时监测的指标,对于疾病(包括肿瘤)的诊断和治疗均具有十分重要的意义。研究目的本项研究旨在发现鉴定外周血血浆中与人非小细胞肺癌发生发展密切相关的标志性蛋白,远期目标为建立肺癌相关的的血浆蛋白联合检测模型,以辅助临床诊断及疗效监测。实验设计采用本实验室前期工作构建的体液中肿瘤相关蛋白研究体系,针对9例非小细胞肺癌病人的肿瘤及对照组织样品建立原代培养,收集其条件培养基(conditioned medium, CM)。随后采用液相色谱和线性离子阱质谱(Linear iontrap,LTQ)对条件培养基样品的总蛋白进行分离和全谱鉴定。继而采用BRBarraytools、ArraySVG等软件,以无标记定量蛋白质组学方法产生的蛋白质谱数(spectral count)为定量参数分析质谱数据。分别通过比较蛋白质组学和系统生物学类比分析两种策略进行游离蛋白数据的信息挖掘,以此获得重要的肺癌相关血浆标志蛋白；然后分别以ELISA、Western blot和免疫组化方法在血浆和组织样品中进行验证。实验结果通过质谱鉴定共得到987个高可信度的蛋白,初步建立了肺癌相关游离蛋白数据库；生物功能分类提示这些蛋白在15个生物学过程中富集,其中包括蛋白结合(protein binding)、水解酶活性(hydrolase activity)、钙离子结合(calcium ion binding)、细胞骨架结合(cytoskeletal protein binding)等过程。经微阵列显著性分析(Significance analysis of microarrays, SAM)运算得到143个差异表达蛋白,其中78个在肿瘤组织培养基中丰度升高(包括目前临床上用于辅助肺癌诊断的血清肿瘤标志物Cyfra21-1和SCC,65个在肿瘤组织培养基中丰度降低。Gene Ontology富集分析结果显示,在肿瘤组织培养基中丰度升高的蛋白主要参与细胞的氧化还原反应及糖酵解等途径,其中糖酵解途径中6个关键性的催化酶(ALDOC、TPI1、ENO1、ENO2、GAPDH、PGK1和PK)在肿瘤组织培养基中均显著上调。针对差异蛋白的Network分析显示,ENO1属于“连线”较多的蛋白之一。根据以上结果,选择ENO1、ALCAM、CLU和DCN分别采用ELISA和免疫组化方法对血浆和组织样品进行了验证。结果显示,在组织水平,相对于配对癌旁肺组织,ENO1在肿瘤中的表达显著升高,ALCAM、CLU、DCN的表达显著降低,其中细胞膜中ALCAM的表达与肺腺癌的临床分期负相关；肺癌病人血浆中ENO1的蛋白浓度显著高于良性肺部疾患者和健康献血者,而肺癌病人血浆中ALCAM的蛋白浓度则较良性肺部疾患者和健康献血者显著降低。此外,针对从上述建立原代培养的9例肺癌组织中的2例以上条件培养基样品鉴定到的511个高可信度蛋白,使用GSEA-P软件分析后发现,与人足月分娩胎盘RNA表达谱比对,来自肺癌条件培养基的197个蛋白在侵袭性较强的中期胎盘母胎界面表达谱中有显著的富集。进而挑选了其中的MMP-1和LGALS3分别进行血浆和组织水平的验证。结果显示,肿瘤组织及血浆中MMP-1的水平随着肺癌进展均显著升高,并可预示肺癌病人的不良预后；肿瘤细胞LGALS3的表达与非小细胞肺癌的淋巴结转移正相关,而肿瘤组织间质LGALS3的表达与肺鳞癌的淋巴结转移正相关。从而证实这种富集方法得到的蛋白与肿瘤的侵袭转移密切相关。结论无标记定量蛋白质组学分析方法对鉴定体液中恶性肿瘤相关的蛋白标志物具有潜在应用价值。胚胎植入与肿瘤侵袭转移生物学行为有很大的相似性,发育生物学提供的信息有助于诠释肿瘤的侵袭转移——因而,从系统生物学的角度将类似的生物学行为以及不同分子水平的分析结合将给肿瘤研究带来新的启示。应用以上两种策略鉴定到的血浆蛋白标志物将为建立肺癌诊断的联合检测模型奠定基础。更多还原

【Abstract】 Background:Lung cancer has become the leading cause of cancer death in China. Detection of lung cancer at early stage is critical for successful clinical therapy and increased survival rate. The blood contains a treasure trove of biomarkers that could reflect the ongoing physiological state of "all tissues" including tumor tissues, allowing for early detection, monitoring therapeutic efficacy, or understanding the biology for the disease.Purpose:To develop a free protein profile associated with non-small cell lung cancers (NSCLCs); and to search for the biomarkers in the circulating plasma, which could be used for early detection and/or monitoring progression of NSCLCs.Experiment design:Based on the novel method established in this laboratory previously which can be used to detect cancer-related free proteins released into the blood, primary organ cultures were generated from the tumor tissue and matched adjacent normal bronchial epithelium derived from 9 cases of NSCLC. The conditional media were collected from the primary cultures and subsequently analyzed using a liquid chromatography coupled with linear ion trap for the proteome. Two strategies of bioinformatic analysis were adopted to explore the protein profiles dysregulated in the lung cancer. One was concerning on the identification of differential free proteins with label-free quantitative proteomic technique. The other one was emphasized the enrichment analysis of the free proteins derived from lung cancers comparing with the expression pattern of invasive placental cells. The resulting proteins were then validated in human samples of plasma and tumor tissue with ELISA, Western blot and immunohistochemical analysis, respectively.Results:In total,987 unique proteins were identified with high confidence from the conditional media from both the lung cancer and the control tissues. Quantitative protein differences between the paired samples were determined by comparing the spectral counts (label-free). A total of 143 proteins were found to exhibit at least 1.5-fold difference in protein level, including 78 proteins upregulated and 65 downregulated in the tumors. Gene Ontology term enrichment analysis revealed that the predominant functional classes of these proteins were related to the oxidoreductase activity, glycolysis, etc., and the subcellular location of them was mainly membrane-bound. Interestingly, six glycolytic enzymes were upregulated concordantly, including ENO1. Network analysis showed that ENO1 was one of the "nodes" regulatory proteins. According to the statistical analysis together with the gene annotation information, ENO1, ALCAM, CLU, DCN were validated in tissue and plasma samples. The elevated ENO1 expression and repression of CLU and DCN were confirmed in 87.5% (14/16),66.7% (8/12), and 62.5% (10/16) of lung squamous cell carcinoma tumor tissues, respectively, by Western blot analysis. Moreover, the expression level of membrane protein ALCAM was correlated with the clinical staging of lung adenocarcinoma. Consistent with the expression in tissue, the plasma level of ENO1 was significantly higher in lung cancer patients than that in patients with nonmalignant diseases in lung and healthy controls, while the ALCAM plasma level was significantly lower compared to the control subjects.There are striking similarities present between the behavior of invasive placental cells and that of invasive cancer cells. The comparison to the under-controlled developmental life phenomena may hint a simplified tumor metastasis-related pattern. A total of 197 proteins with high confidence (at least twice) identified from the conditional media of lung cancer tissues showed significant enrichment in the more invasive middle term human placentation maternal-fetal interface with Gene Set Enrichment Analysis. Among them, the MMP-1 and LGALS3 were validated in the clinical samples from lung cancer patients. The result showed that the high levels of MMP-1 in both tumor tissue and plasma were correlated with lung cancer progression and poor prognosis. The intracellular expression of LGALS3 was higher in the primary tumors with lymph node metastases than that without metastases (P<0.001). Furthermore, the LGALS3 expression in the stroma was associated with lymph node invasion of SCC in lung.Conclusions:Label-free quantification method could be a feasible and effective strategy to search potential biomarkers of cancer in the simplified conditional media system. The study from a view of systems biology combined with similar biological process at different molecular level may likewise lead to a better understanding for invasion and metastasis of cancer cells. The differentially expressed proteins reported in this study would make promise for the successful protein profile and biomarkers for the non-small cell lung cancers.更多还原