膀胱上皮肿瘤尿液差异蛋白的筛选

【作者】 吴昕；

【导师】 李汉忠；

【作者基本信息】 北京协和医学院 ， 临床医学， 2009， 博士

【摘要】 背景：膀胱肿瘤(Bladder Cancer, BC)是泌尿生殖系统的常见肿瘤,在我国的发病率逐年提高,且有年轻化的趋势。导致膀胱肿瘤的原因很复杂,目前比较认可的有：长期接触芳香胺类化学药物、膀胱的局部刺激、某些药物的滥用、盆腔放疗、吸烟等。基因层次的研究,也发现了ras、HER-2/neu、c-myc等癌基因,p53、RB1等抑癌基因都可能是膀胱肿瘤的致病原因。其最常见的初始症状是无症状性肉眼血尿,随着肿瘤的增大、增多与浸润,间歇性血尿可转为持续性血尿。目前针对膀胱肿瘤的诊断手段多种多样,但是都存在着各自的缺陷。尿液细胞学检查对于分级低的膀胱肿瘤敏感性较低,且检查结果往往对检查者的技术水平有较大的依赖性。膀胱镜检查可以在直视下直接观察肿瘤的形态、大小等,可以取活检组织,但是对设备依赖性较大,且创伤性较大。常用的影像学检查方法共同的优点是具有无创性,但对小肿瘤或是表浅肿瘤等显示不佳。而对于肿瘤标记物的寻找,更是结论丰富。NMP22、 SurVivin、UBC、端粒酶、p27蛋白、VEGF等等各种标记物都对膀胱肿瘤具有一定的诊断价值,但都在特异性或敏感性等方面存在缺陷,因此没有一个公认的标准。近年来,随着蛋白质组学技术的发展与成熟,以及其在肿瘤研究方面独特的优越性,很多学者开始通过这种途径来寻找膀胱肿瘤的标记物,并且已经发现了CRT、 UBC1-5、角蛋白、MRP8、MRP14、银屑素等多种物质在膀胱肿瘤患者体内呈现独特的表达方式,揭示出了利用蛋白质组学技术来寻找膀胱肿瘤标记物的可观前景。本课题通过利用SELDI质谱技术,寻找能够在膀胱肿瘤的早期诊断中发挥作用的标记蛋白,以期能够为临床应用提供一种可能的且具有无创性的诊断方法。目的：1.筛选膀胱肿瘤患者与对照组成员尿液样本中峰值存在着明显差异的蛋白质分子。2.利用膀胱肿瘤患者与对照组成员尿液样本中蛋白质分子峰值的差异,建立膀胱肿瘤的诊断筛选模型。方法：1.采集膀胱肿瘤患者、非肿瘤的泌尿系统其它疾病患者以及健康志愿者的晨起二次尿标本。2.采用美国BIORAD公司生产的H4疏水芯片及SELDI质谱仪测定所有尿液样本中的蛋白质分子峰值。使用Ciphergen Protein Software3.2.1显示读取的蛋白质分子峰值图示。3.使用SPSS16.0统计软件对读取的蛋白质分子峰值结果进行统计学分析。P<0.05认为存在明显统计学差异。4.使用BPS对上述所得结果进行聚类分析,建立膀胱肿瘤的诊断筛选模型。结果：1.本实验中膀胱肿瘤患者的临床特点：共有38例膀胱肿瘤患者进入实验组,最小年龄35岁,最大年龄88岁,中位年龄68.5岁,其中60岁以上的患者占了73.7%。其中男性27例,女性11例,男：女=2.5：1。术后病理明确为高级别肿瘤的有17例,低级别的有19例,未明确高低级别的有2例。2. SELDI质谱仪读取样本蛋白质分子峰值的结果：设置Ciphergen Protein Software3.2.1显示的最小分子量为1000,最大分子量为50000,信噪比≥5,蛋白最低出现比率5%,可以分析得到共214个可读点。3.使用SPSS16.0统计软件得到的统计结果：对所有的214个可读点进行正态分布检验,发现其中90个点所代表的蛋白质分子峰值在所有样本中呈现正态分布,余124个点所代表的蛋白质分子峰值在所有样本中不呈现正态分布。对所有符合正态分布的数据进行T检验分析,共发现分子量为1712.263、1896.238、2228.543、2464.589、2977.719和4792.879的6个蛋白质峰值在实验组与对照组之间存在显著差别,均符合P<0.05。对所有不符合正态分布的数据进行非参数Mann-WhitneyU检验分析,共发现分子量为1877.678、2212.175、2578.946、2847.638、3249.263、3385.764、4129.112、13304.09、15108.4、16652.69、22197.04和33354.89的12个蛋白质峰值在实验组与对照组之间存在显著差别,均符合P<0.05。4.使用BPS建立膀胱肿瘤诊断筛选模型的结果：利用上述18个峰值在实验组与对照组之间存在显著差异的蛋白,对所有的样本进行聚类分析。发现当依次选用2464.589、1896.238、1877.678、2977.719和2847.638等5个蛋白质的峰值作为分组结点建立诊断筛选模型时,可以将实验组与对照组按照最高的正确率区分开来。5.图形分析的结果：使用Ciphergen Protein Software3.2.1,选择上述诊断筛选模型中的分子量为2464.589的蛋白质作为示例,分别用线型图和凝胶模拟图观察实验组和对照组图形中的差别。发现在线型图中实验组与对照组在分子量为2464.589的蛋白点上存在峰值高低的明显差异,而凝胶模拟图中两组在相同的蛋白点上存在蛋白电泳痕迹的明显差异。结论：1.采用H4疏水芯片,通过SELDI质谱仪的技术可以很好地捕获尿液中的蛋白分子,特别是低分子量范围内的蛋白质。2.膀胱上皮肿瘤患者与对照组成员的尿液样本中的蛋白质分子峰值存在统计学上的显著差异,并且可以利用这种差异建立起膀胱上皮肿瘤的诊断筛选模型。更多还原

【Abstract】 Background:Bladder cancer (BC) is the most common urinary tumor which has a gradually increased incidence year by year. What’s more, it is not infrequent to see some young people suffering from the disease. The causes which lead to bladder cancer are complex. There are some of them have been recognized and studied:long-term exposure to aromatic amines chemicals, local stimulation of bladder, the abuse of certain drugs, pelvic radiotherapy, smoking and so on. At the same time, some oncogene like ras, HER-2/neu, c-myc and some suppressor gene like p53, RBI have been discovered in the research of gene-level. All of them could be the probable causes of bladder cancer. The most common initial symptom of bladder cancer is usually asymptomatic gross hematuria which could become from intermittent to continuity because of the increase and infiltration of tumor. At present, a wide range of diagnose ways of bladder cancer are used, but none of them is perfect. Urine cytology test couldn’t find out low-level tumor and its result is usually highly dependent on the technical level of the examiner. Cystoscope could not only observe the shape and size of tumor directly, but also be an easy way to take biopsy. But it couldn’t be done without complete equipment. Meanwhile the cystoscope brings too much pain to the patients. Several common imaging methods share the same advantage which is non-invasive, but none of them could find small or flat tumor easily. In the search of tumor markers, NMP22, SurVivin, UBC, telomerase, p27, VEGF and several others were found to be useful in the diagnose of bladder cancer. But there is still no public accepted standard.With the development and maturity of proteomics technology and its unique advantages in the study of tumor, some researchers have already found their new ways to pick out the tumor markers of bladder cancer. CRT, UBC1-5, keratin, MRP8, MRP14and some other markers which have relationship with bladder cancer have already been found. It is said that there would surely be a bright future in the search of BC tumor markers with the help of proteomics technology. In this study, SELDI technology was used to find some markers which could be useful in the early stage diagnose of bladder cancer. And the new way was supposed to be non-invasive.Objective:1. To find out the special protein peaks in the urine of bladder cancer patients or control group members. 2. To build a diagnosis screening model of bladder cancer with the differences between the protein peaks of bladder cancer patients and control group members.Methods:1. Second morning urine samples of the patients with bladder cancer, other non-tumor diseases of urinary system and healthy people were collected.2. H4hydrophobic chips and SELDI machine which were produced by BIORAD company of America were used to determinate the protein molecular peaks of all urine samples. Ciphergen Protein Software3.2.1was used to display all of the captured protein peaks.3. Laboratory data was analyzed by SPSS16.0software, and P<0.05was considered to indicate statistical significance.4. BPS was used to do cluster analysis work and build the diagnosis screening model of bladder cancer.Results:1. Clinical features of bladder cancer patients:38bladder cancer patients were recruited. The range of age was from35to88. The median age of patients was68.5years, with73.7%patients were older than60years. The number of males was27, with the females’11. The high-level tumor group included17patients, while low-level group included19ones. There were2patients’ pathological levels were uncertain.2. The result provided by SELDI machine:set the minimum protein molecular weight which could be displayed to1000, the maximal to50000, S/N≥5, the min peak rate to5%. There were214peaks could be found.3. The result analyzed by SPSS16.0software:after the test of normality,90peaks were classified as normal distribution group, while124other peaks as non-normal distribution. T-test picked out6protein peaks from the normal distribution group, which molecular weight were1712.263,1896.238,2228.543,2464.589,2977.719and4792.879. Mann-Whitney U analysis picked out12protein peaks from the non-normal distribution group, which molecular weight were1877.678,2212.175,2578.946,2847.638,3249.263,3385.764,4129.112,13304.09,15108.4,16652.69,22197.04and33354.89. All of the18protein peaks were considered to indicate statistical significance with P<0.05.4. The result of diagnosis screening model provided by BPS:when2464.589,1896.238,1877.678,2977.719and2847.638were used as select nodes to build the diagnosis screening model, the cancer group and control group could be distinguished best.5. The result of figure analysis:the protein M=2464.589was used as an example. Both the linear and gel figures could show visible differences between cancer and control groups.Conclusions:1. H4hydrophobic chip is quite useful to capture the protein molecule in the urine samples, especially the ones with low molecular weight.2. The differences of protein peaks between cancer and control groups are considered to indicate statistical significance and could be used to build a diagnosis screening model of bladder cancer.更多还原