【作者】 周劲旭；

【导师】 陈菊祥；

【作者基本信息】 第二军医大学 ， 外科学， 2012， 硕士

【摘要】 脑胶质瘤是危害人类健康的中枢神经系统常见疾病，约有50%新诊断的脑肿瘤属于神经胶质起源的原发性脑肿瘤。常见的起源有星形细胞、少突胶质细胞、室管膜细胞等。恶性胶质瘤约占所有神经胶质瘤总数的70％，以起源于星形细胞的胶质母细胞瘤和间变性星形细胞瘤最为多见，预后较差。文献报道脑胶质瘤的侵袭、复发是由多种基因、蛋白以及细胞因子组成的各种不同信号传导通路之间相互调控的结果，这一过程有可能涉及到肿瘤细胞粘附、基质降解、迁移、细胞增殖等诸多环节，但具体的分子病理机制仍未清楚，有待进一步研究和探讨。目前对于难以治愈的恶性脑胶质瘤的诊治，关注的焦点正在由传统的手术、放化疗等转移到胶质瘤特征性分子靶向治疗，并取得了一定的进展。高尔基磷酸化蛋白-3（Golgi phosphoprotein-3，GOLPH3），属于高尔基体基质蛋白，是新近发现的促癌基因。已有研究证实GOLPH3在前列腺癌、结肠腺癌和非小细胞肺癌等实体性肿瘤中呈现高表达；在体外实验中已证实GOLPH3作为促癌基因可促进肿瘤细胞的增殖。目前认为GOLPH3蛋白的促癌机制主要是通过增强mTOR通路（mammalian target of rapamycin）进行，同时GOLPH3的高表达可以增强mTOR通路对雷帕霉素（mTOR通路的抑制剂）的敏感性。因此，GOLPH3的表达水平可作为mTOR通路对雷帕霉素敏感性的判定指标。文献报道脑胶质瘤中GOLPH3表达与胶质瘤存在相关性，但GOLPH3是如何参与脑胶质瘤调控的分子机制目前尚未阐明，GOLPH3与胶质瘤的预后相关性亦未有报道。本课题拟利用课题组制备的高通量组织芯片技术，同时采用RT-PCR、Western blot，以及在胶质瘤细胞系中的体外实验，在脑胶质瘤组织和细胞系中研究靶标GOLPH3基因和蛋白的表达水平与患者预后之间的关系，并初步探讨其调控脑胶质瘤恶性生物学行为的分子机制。骨唾液蛋白（Bone sialoprotein，BSP）通过调节基质金属蛋白酶-2（MMP-2），和整合素αvβ3相互作用来激活MMP信号通路，促进肿瘤组织血管生成，是新发现的促癌基因。虽有报道显示BSP在前列腺癌、乳腺癌以及胰腺癌等实体肿瘤中高表达。但国内外文献目前尚无有关BSP在胶质瘤中表达水平，其与患者预后之间关系，以及其参与胶质瘤形成过程中的分子机制等方面研究。本课题拟利用高通量组织芯片和RT-PCR，探讨在脑胶质瘤组织中BSP的表达量与胶质瘤患者预后的关系。第一部分：高通量脑胶质瘤组织芯片的制备目的：利用具备完整随访资料的脑胶质瘤标本库，制备高通量组织芯片（TMA）。研究相关靶标在脑胶质瘤组织中的表达水平，并结合临床随访数据进行回顾性分析，判定相关靶标与脑胶质瘤患者病理级别、预后之间的关系。方法：1）收集2000年1月至2010年11月在第二军医大学附属长征医院神经外科接受手术、经临床病理证实为胶质瘤的病理标本共计300例；正常脑组织16例（通过外伤减压手术取得），制备高通量脑胶质瘤组织芯片；2）课题组对300例胶质瘤患者进行详细的随访（包括年龄、性别、癫痫、颅内高压、平均肿瘤直径、切除程度、术前及术后影像学表现、术后放化疗、生存状态、复发情况等），随访时间截止至2011年9月；3）使用SPSS16.0软件进行生存分析,无进展生存期和总体生存期计算使用Kaplan-Meier方法，并用log-rank检测校正；单因素分析显示p<0.2的变量纳入多因素分析，使用Cox回归；设定p<0.05有统计学显著意义。结果：脑胶质瘤组织芯片包含有各种病理类型的300例胶质瘤组织和16例正常脑组织，共计316点。组织芯片设计构成如下：WHO I级11例：9例毛细胞星形细胞瘤，1例室管膜下巨细胞星形细胞瘤、1例室管膜下瘤；WHO II级109例：67例弥漫性星形胶质瘤，27例少突胶质细胞瘤，7例混合性少突星形细胞瘤，7例室管膜瘤，1例脉络丛乳头状瘤；WHO III级47例：33例间变星形细胞瘤，11例间变少突胶质细胞瘤，2例间变室管膜瘤，1例星形母细胞瘤；WHO IV级133例：119例胶质母细胞瘤，12例髓母细胞瘤，1例胶质肉瘤，1例幕上原始神经外胚层肿瘤。课题组通过详实的随访获得临床数据如下：300例脑胶质瘤患者中，男性占66.3%，女性占33.7%；85%的患者其发病年龄小于60岁；80.3%的患者在术前出现癫痫发作；12.7%的患者术前MRI影像中出现坏死；共有78.7%的患者接受了显微镜下全切手术；64.7%的患者在术后接受了规范化的放疗；62%的患者在术后接受了规范化的化疗。截止至2011年09月，共有284名患者获得随访，中位总体生存期（OS）为30月（95%CI：21.8-38.2月）；中位无进展生存期（PFS）为25月（95%CI：19.5-30.5月）；5年总体生存率为34%，5年无进展生存期为28%。结论：组织芯片技术稳定成熟，结果可靠。组织芯片是高效、简便和标准化的研究技术，适用于多基因、大样本的临床肿瘤生物特性研究。芯片数据结合统计学软件，可以对患者生存预后进行回顾性分析。第二部分： GOLPH3与胶质瘤的预后和恶性生物学行为调控机制的研究目的：本课题通过高通量组织芯片测定GOLPH3在胶质瘤中的表达量，分析其与脑胶质瘤患者预后之间的关系；同时利用定量RT-PCR、WB、细胞增殖、克隆及细胞周期实验在U251和U87胶质瘤细胞中验证干扰和过表达的效果。方法：1）使用GOLPH3抗体（Abgent, San Diego, CA, USA）进行免疫组化；2）RT-PCR、Western Blot在U251和U87细胞中验证RNA干扰和过表达效果；3）细胞增殖、克隆形成及细胞周期等实验研究GOLPH3在脑胶质瘤恶性生物学行为的作用。结果：在获得随访结果的284例胶质瘤标本中，免疫组化分析获得有效结果270例（14例染色脱片）。单因素分析显示GOLPH3表达量与全部胶质瘤患者总体生存期（OS）及无进展生存期（PFS）不相关（p=0.698，p=0.704）。在106例原发胶质母细胞瘤患者中有97例数据有效（9例染色脱片），免疫组化分析结果显示GOLPH3表达评分为3.73±2.13；97原发GBM患者中获得随访的有93例，其中位生存期（OS）为12月(95%CI:10.31-13.69月)，中位无进展生存期（PFS）为9月(95%CI:7.33-12.67月)。1年总体生存率为52%；无进展生存率为44%。Log-rank显示GOLPH3表达量低的GBM患者其中位生存期（OS）较GOLPH3表达量高的GBM患者明显延长（15月vs.10月）（p<0.05）；无进展生存期GOLPH3表达量低的GBM患者与GOLPH3表达量高的GBM患者相比较明显延长（12月vs.7月）（p<0.05）。单因素及多因素分析均显示对于原发GBM患者，GOLPH3对其预后是独立的预测因素（p<0.05）。siRNA能够在mRNA水平和蛋白水平显著降低GOLPH3的相对表达量（p<0.05）；过表达GOLPH3则使GOLPH3相对表达量明显提高（p<0.05）。在体外细胞系实验（U251和U87）中，siRNA降低GOLPH3表达水平后，胶质瘤细胞系U251和U87的增殖、克隆明显减弱（p<0.05），胶质瘤细胞生长停滞在G1期。结论：GOLPH3在原发GBM患者中表达量越高，预后越差；同时在RT-PCR及WB以及在体外胶质瘤细胞实验中的结果提示GOLPH3有可能参与到胶质瘤的发生、发展及侵袭过程中，是潜在的促癌蛋白。第三部分：BSP与胶质瘤患者的病理级别和预后的相关性研究目的：课题通过制备的高通量组织芯片测定BSP在各病理级别胶质瘤组织中的表达，分析其表达量与胶质瘤患者预后之间的关系，并在胶质瘤组织中利用RT-PCR来进行验证其表达水平。方法：1）使用BSP抗体（Abcam，Cambridge, MA，USA）进行免疫组化；2）定量RT-PCR在胶质瘤组织和正常脑组织中（包括25例低级别胶质瘤组织，30例高级别胶质瘤组织和7例正常脑组织）验证BSP表达水平的高低。结果：RT-PCR和组织芯片分别在mRNA和蛋白水平显示BSP的表达量在高级别胶质瘤组明显高于低级别胶质瘤组织和正常脑组织，和病理级别呈正相关关系(P<0.001)。入选270例胶质瘤患者中（其他30例胶质瘤组织标本和1例正常脑组织标本染色脱片），258名患者（95.6%）有随访结果，其总体中位无进展生存期为21月(95%CI,16.72-25.28月）；总体中位生存期为26月(95%CI,18.89-33.11月)；截至随访结束，28%(28/100)的低级别胶质瘤患者和87.3%(138/158)的高级别胶质瘤患者死亡。单因素及多因素分析显示BSP表达量高，则胶质瘤患者无进展生存期和总体生存期越短（PFS：HR=1.508, P=0.014；OS：HR=1.572, P=0.009）。BSP表达量低的胶质瘤患者其中位生存期和无进展中位生存期与BSP表达量高组相比，明显延长（OS：42vs.16月； PFS：30vs.13月）。结论：靶标BSP表达量在高级别胶质瘤患者组中明显高于低级别患者组和正常患者组，单因素及多因素分析显示BSP高表达水平预示着患者预后不良。本课题验证BSP对于脑胶质瘤患者的总体生存期和无进展生存期均是独立的预测因素，是诊断和潜在的治疗靶标，具有重要的临床意义和应用价值。更多还原

【Abstract】 Gliomas are the most common primary brain tumor and accounts for approximately50%of all brain tumors, and originated mostly from astrocytes, oligodentrocytes andependymal cells. As we all known, nearly70%are malignant gliomas, and most of themare originated from astrocytes, such as glioblastoma and anaplastic astrocytoma, and ofpoor prognosis. It is reported that the formation and the recurrence of glioma areinfluenced by different signaling pathways which contained several genes, proteins andcytokines. The molecular mechanisms underlying the initiation, maintenance andprogression of glioblastoma still remain largely unclarified. Recent understanding ingenomic and molecular abnormalities in glioma has shifted the treatment paradigmtowards using molecularly targeted agents and allowed more rational uses of targetedmolecular therapies.Golgi phosphoprotein3(GOLPH3), is encoded by a gene residing on humanchromosome5p13, and located in the trans-Golgi network. GOLPH3is implicated inprotein trafficking, receptor recycling and glycosylation from Golgi to plasma membrane.Recent studies have shown GOLPH3could promote cellular transformation and tumorgrowth. Specifically, GOLPH3is highly expressed in several solid tumors, includingmelanoma, colon adenocarcinoma and non-small-cell lung cancer. GOLPH3enhancesmammalian target of rapamycin (mTOR) signaling in human cancer cells, and modulatescellular response to mTOR inhibitors. mTOR, a serine/threonine protein kinase and ‘targetof rapamycin’, acts as a primary regulator of protein synthesis and cell growth, and functions as a key integrator of the receptor tyrosine kinases and phosphatidyl-inositol-3kinases (RTK-PI3K) pathway, which in turn is the most frequently hyperactivated pathwayin glioma. Recent study has shown that in mRNA and protein level, GOLPH3expressionin human glioma is associated with the pathological grade, and no expression in normalbrain. However, the relationship between GOLPH3and the development and proliferationand invasion of glioma was still unknown. The aim of this study is to clarify the relationshipbetween the expression of GOLPH3glioma tissues and the outcome of glioma patients bytissue micorarray; and also clarify the exact role for GOPH3in the oncogenic process ofglioma cells by knocking down or overexpression its expression level in U251and U87celllines.Bone sialoprotein (BSP), a specific regulator of matrix metalloproteinase-2(MMP-2),is highly expressed in numerous cancers such as prostate cancer, breast cancer, andpancreatic cancer and may contribute to their invasive potential. BSP may also act as apro-angiogenic factor to promote angiogenesis through binding to integrin αvβ3. However,there has been no study on the role in the the development and proliferation and invasionof glioma and the expression of BSP in gliomas. In the current study, we sought toexamine the expression of BSP in human glioma tissue and normal brain tissues byimmunohistochemistry using tissue microarray and find out the correlation between BSPexpression and prognosis of glioma patients.Part I: Construct the glioma tissue microarrayPurpose: By using of paraffin wax of glioma specimens to construct the tissue microarray (TMA), and the method of immunohistochemistry, we studied special targets inthe glioma tissues to find out the relationship between the expressions of the target and thepathological grade and the prognosis of glioma patients.Method:1）All300glioma tissue specimens were obtained from glioma patients whounderwent surgical treatment at the Department of Neurosurgery, Changzheng Hospitalbetween January2000and November2010. The16normal brain tissues were obtainedfrom the patients who need the decompression surgery;2）The follow-up was carried out inall patients, and the survival time was censored in September2011.4）Cumulative survivalwas calculated by the Kaplan-Meier method and analyzed by log-rank test. Univariate andmultivariate analyses were performed by stepwise forward Cox regression model (P<0.2was considered as the inclusion criterion for factors that could be added into multivariateanalysis). All statistical analyses were performed with the SPSS16.0software (SPSS Inc,Chicago, IL), P<0.05was considered as statistically significantly different.Results: The tissue microarray contains: WHO grade I:9pilocytic astrocytomas,1Subependymal giant cell astrocytoma,1Subependymoma; WHO grade II:67Astrocytomas,27oligodendrocytomas,7oligoastrocytomas,7ependymomas,1papillomachoroideum; WHO grade III:33anaplastic astrocytomas,11anaplasticoligodendrocytomas,2anaplastic ependymomas,1astroblastoma; WHO grade IV:119glioblastomas,12medulloblastomas,1gliosarcoma,1PNET. By follow-up, we found thatin all300glioma patients,66.3%patients were male;85%patients showed younger agethan60years; the elipsy were observed in80.3%patients; necrosis on MRI was observed in12.7%patients.78.7%patients underwent total resection.62%patients receivedadjuvant chemotherapy and64.7%patients underwent adjuvant radiotherapy. We obtained284results of follow-up: the median overall survival was30months (95%CI：21.8-38.2months); The median progression-free survival was25months (95%CI:19.5-30.5months).The5years OS was34%, and the5years PFS is28%.Conclusion: Tissue microarray is an analysis tool for large sample of tumor. We canobtained the result of prognosis of patients by retrospectively analyzes combining withstatistical tools. Part Ⅱ: Study the relationship between the GOLPH3’sexpression and the prognosis of glioma patients and the role ofGOLPH3in glioma cells in vitroPurpose: GOLPH3, as an important protein in mTOR signaling, is overexpressed inand correlates with the pathological grade of glioma. In the current study, we examined theexpression of GOLPH3in gliomas with tissue microarray and correlated the measure to thepatient outcome.Method:1) GOLPH3expression in tumor tissue from300glioma patients wasexamined with tissue microarray and immunohistochemistry. Potential effects of GOLPH3on tumor growth were also examined in representative cell lines (U251and U87) by downregulating GOLPH3with RNA interference.Results: In all300glioma samples, we obtained284results of follow-up. In284glioma samples, we adopted270glioma samples for analyzing (the other14tissue dotswere lost from the TMA slide). In270glioma patients, the univariate analysis showed nosignificance between the expression of GOLPH3and the prognosis of glioma patients(p=0.698，p=0.704). In all106GBM samples, we adopted the97promary GBM samplesfor analyzing (the other9tissue dots were lost from the TMA slide). Tissue microarrayanalysis revealed high GOLPH3expression in40patients (40/97,41.2%) and lowGOLPH3expression in the remaining57patients (57/97,58.8%).We obtained93GBMpatients’ follow-up. In93GBM patients, the median overall survival (OS) was12months(95%CI:10.31-13.69months). The median progression-free survival (PFS) was10months(95%CI:7.33-12.67months). The one-year OS and PFS rate was52%and44%,respectively. Log-rank showed that patients with low GOLPH3expression hadsignificantly longer median OS (15vs.10months in patients with high GOLPH3expression) and median PFS (12vs.7months). Univariate and multivariate analysisindicated that GOLPH3was an independent prognostic factor for OS and PFS. In the invitro experiments, GOLPH3downregulation with siRNA suppressed the proliferation,clonogenic growth and cell cycle in cultured cell lines.Conclusion: High level of GOLPH3expression in GBM tissue is associated withpoor outcome, and GOLPH3may be as a potential oncogene. Part Ⅲ: Study the relationship between the expression of BSPand the tumor garde and the prognosis of glioma patientsPurpose： Investigate the expression and prognostic value of bone sialoprotein (BSP)in glioma patients.Method: We determined the expression of BSP in normal brain and glioma samplesby immunohistochemistry using tissue microarray and real-time RT-PCR.Results: Both BSP mRNA and protein levels were significantly elevated inhigh-grade glioma tissues compared with those of normal brain and low-grade gliomatissues, and BSP expression positively correlated with tumor grade (P<0.001). In all300glioma samples, we adopted270glioma samples for analyzing (the other tissue dots werelost from the TMA slide). And in270glioma patients, we obtained258patients’ follow-up.In258glioma patients, the median overall survival (OS) was26months (95%CI:18.89-33.11months). The median progression-free survival (PFS) was21months (95%CI:16.72-25.28months). Death was found in28%LGG patients and87.3%HGG patientsduring the follow-up. Univariate and multivariate analysis showed high BSP expressionwas an independent prognostic factor for a shorter progression-free survival (PFS) andoverall survival (OS) in glioma patients [hazard ratio (HR)=1.508, P=0.014andHR=1.572, P=0.009, respectively]. Patients with low BSP expression had a significantlylonger median OS and PFS than those with high BSP expression (OS:42versus16months;PFS:30versus13months). Additionally, high-grade glioma patients who had a worseprognosis were more likely to exhibit high BSP expression (HR=2.158, PFS; HR=2.100, OS).Conclusion: High BSP expression occurs in a significant subset of high-grade gliomapatients and predicts a poorer outcome.更多还原