【作者】 陈慧芳；

【导师】 李春岩；

【作者基本信息】 河北医科大学 ， 神经病学， 2010， 博士

【摘要】 肌萎缩侧索硬化(Amyotrophic lateral sclerosis, ALS)是一种最常见的成人发病的神经退变性疾病,其特征是脊髓、脑干及运动皮质上、下运动神经元进行性及相对选择性的退变。病人表现为进行性的瘫痪与肌肉萎缩,约半数在发病3～5年内死亡,死因多为呼吸衰竭。每年全球发病率约为(1～2)/100 000,其终生发病率约为1/600至1/2 000。90%的ALS病例为散发性,其余10%为家族性,二者的临床表现及病理特点难以区分。20%的家族性病例与细胞质中铜/锌超氧化物歧化酶(SOD1)发生的某种基因突变相关联。这种SOD1基因突变属于发挥主导作用的错义突变,致使一种氨基酸被另一种氨基酸所替代,例如甘氨酸被丙氨酸所替代(SOD1-G93A)。因为表达人SOD1突变的动物亦在成年期发病,且临床表现及病理特点与人类患者非常相似,故这种转基因动物已成为广泛采用的ALS动物模型。实际上,绝大多数的突变SOD1并未影响酶学活性,其引发的病理效应可能源于生物学功能方面的毒性获得。尽管在对ALS深入研究的基础上,学者们提出了解释SOD1突变效应的几种假说,包括氧化应激、谷氨酸毒性、大分子聚集体形成、线粒体功能失调及凋亡等假说,但是,目前人们对SOD1突变促发ALS的确切机制依然所知甚少。高效的基因组学技术为阐释复杂的疾病发病机制及其调控提供了宝贵的机遇。mRNA转录本的可变剪接是转录后调控的重要形式,它将mRNA前体中的内含子选择性移除,同时将外显子组合连接成多种形式的成熟RNA。数量有限的基因可扩展表达为非常复杂的蛋白质组,因此可变剪接在产生蛋白质组学的多态性方面发挥着重要的作用。据估算,在所有的人类基因中发生可变剪接者约占3/4。目前资料显示,90%以上的人类基因发生了可变剪接。ASTD数据库(http://www.ebi.ac.uk)资料显示57%的小鼠基因发生可变剪接。研究者越来越认识到mRNA剪接在许多重要的生理过程如发育、生理和疾病中,发挥着主导作用。Affymetrix外显子芯片迅速获得业界认可并成为同时检测基因水平和外显子水平表达的操作标准。新近的研究表明可变剪接在神经系统尤其普遍。已有报道ALS中TAR DNA-结合蛋白43(TDP-43)的异常剪接可增加聚集和细胞毒性作用。啮齿类动物中的谷氨酸转运体1(在人类称为EAAT2)以一种复杂的模式发生可变剪接,包括5′端与3′非翻译区及几个编码区异构体。尽管可变剪接在神经系统普遍存在,但在ALS中的研究甚少。本研究应用Affymetrix小鼠外显子1.0 ST芯片来检测SOD1-G93A转基因小鼠及其同窝同性别非转基因对照动物腰髓基因的差异表达与外显子的可变剪接,并通过RT-PCR方法对上述检测结果进行验证。每张芯片大约包含5.4兆个探针(1.4兆个探针组),覆盖了约1百万个已知的和预测的外显子。大约90%的外显子被4个探针所覆盖,这些针对同一个外显子设计的探针组成的集合称为探针组(probesets)。对每一个基因而言,平均的探针数是30～40个,通常分布于整个转录本序列的全长。上述外显子芯片可整合每一个已知的和预测的外显子的表达,为研究可变剪接(外显子表达分析)提供了强有力的工具,同时可对基因水平的表达进行分析。本研究首次对SOD1-G93A转基因小鼠的腰髓进行了外显子芯片分析,以确定在突变SOD1存在的情况下差异表达基因及可变剪接事件。本课题分为三部分。第一部分肌萎缩侧索硬化转基因鼠模型腰髓差异表达基因研究目的:研究在突变SOD1存在的情况下,G93A转基因小鼠腰髓组织的基因表达谱。方法:应用Affymetrix小鼠外显子1.0 ST芯片检测发病期的雌性SOD1-G93A小鼠及非转基因同性别的同窝对照小鼠腰髓组织的基因转录本,并对差异表达基因进行分层聚类分析、GO分类及生物学通路分析。结果:(1)在突变SOD1蛋白存在的情况下,有322个转录本的差异表达具有统计学意义,其中309个(占95.96%)上调,13个(占4.04%)下调。基于T检验的结果,经可视化处理生成一个火山图以对比发病组与对照组之间的差异表达基因。(2)通过分层聚类分析,我们可见各样品之间有很好的可重复性,而且发病组与对照组之间的基因可成功区分。(3)GO分类。对这些发生差异表达的基因根据其分子功能进行了分类,其中满足P值< 0.001的上述基因类别包括肝素结合、胰岛素样生长因子1结合、超氧化物引发的NADPH氧化活性、整合素结合、脂蛋白酯酶活性及化学因子活性等。从差异表达基因的生物学过程考虑,主要涉及细胞黏附、炎症反应及免疫反应。同时还涉及白细胞黏附、整合素复合体、整合素介导的信号途径以及MAPK活性等。(4)生物学通路分析。基于KEGG数据库分析,差异表达基因主要涉及如下生物学通路:细胞因子及其受体相互作用、actin细胞骨架调节、自然杀伤细胞介导的细胞毒作用、细胞黏附、补体及凝血瀑布、白细胞跨内皮迁移、细胞黏附分子及凋亡等。结论:肌萎缩侧索硬化模型SOD1-G93A转基因小鼠腰髓组织的基因表达与非转基因同性别的同窝对照小鼠相比,具有显著的统计学差异,且发生差异表达的基因可明确区分并进行相应的归类与分析。第二部分肌萎缩侧索硬化转基因鼠模型腰髓可变剪接研究目的:研究在突变SOD1存在的情况下,G93A转基因小鼠腰髓组织的可变剪接。方法:应用小鼠外显子1.0 ST芯片检测发病期的雌性SOD1-G93A小鼠及非转基因同性别的同窝对照小鼠的腰髓组织的可变剪接事件,并对可变剪接基因进行分层聚类分析、GO分类及生物学通路分析。结果: (1)在全部44 087个探针组中,有333个探针组符合质量控制标准,可用于后继的分析处理。这些探针组分属于297个发生可变剪接的基因,提示单个转录本可能发生多个可变剪接事件。我们的研究结果显示,有63个发生可变剪接的外显子源于27个转录本。在多个编码蛋白的转录本中检测到3个或3个以上出现可变剪接的外显子,提示了小鼠体内剪接调控模式的复杂性。(2)根据其相似性用聚类分析3.0软件进行了HC分析。结果显示两组间具有良好的相关性。(3)剪接基因GO分类。根据其分子学功能分类,这些发生可变剪接的基因主要涉及铁离子结合、化学因子活性、基质金属蛋白酶活性及分子活性。而从生物学过程考虑,这些基因主要与炎症反应、免疫反应、趋化作用、MAPK活性的活化及抗凋亡等相关。(4)生物学路径分析。以KEGG数据库分析为基础,发现297个可变剪接基因中的95个与已知的生物学路径相关,并且鉴定出其中具有统计学差异的通路,包括细胞因子及其受体的相互作用、MAPK信号路径、细胞黏附分子、actin细胞骨架的调控、细胞凋亡、T细胞受体信号路径及自然杀伤细胞介导的细胞毒等。结论:单个转录本可能发生多个可变剪接事件,提示了小鼠体内剪接调控模式的复杂性。发生可变剪接的基因与发生差异表达的基因相比,虽然重叠部分很少,但其GO归类与生物学通路重叠比例很高,提示相应的生物学效应受到转录表达与外显子剪接两个水平的共同调控。第三部分肌萎缩侧索硬化转基因鼠模型腰髓差异表达基因及可变剪接的验证研究目的:验证在突变SOD1存在的情况下,G93A转基因小鼠腰髓组织的差异表达基因及可变剪接事件。方法:应用RT-PCR方法对发病G93A转基因小鼠及其同窝对照的腰髓组织中感兴趣差异表达基因及可变剪接进行验证。结果:发病动物腰髓样本中CD68抗原(CD68 antigen,CD68)与脂蛋白脂酶(lipoprotein lipase,LPL)的mRNA表达水平明显增高(P < 0.001),提示在疾病的早期阶段存在炎症反应与脂类高代谢状态。我们也对ALS中程序性细胞死亡1(programmed cell death 1, PDCD1)mRNA转录本的可变剪接进行了RT-PCR验证。结果表明包含外显子2的异构体主要存在于发病动物,而不含外显子2的异构体主要存在于对照组。芯片预测发生可变剪接与不发生可变剪接的外显子均通过RT-PCR得以证实。结论:上述结果证实了外显子芯片技术的可靠性与敏感性,其可用以检测基因水平及外显子水平的差异表达。更多还原

【Abstract】 Amyotrophic lateral sclerosis (ALS) is one of the most common adult-onset neurodegenerative diseases, characterized by progressive and relatively selective degeneration of the upper and lower motor neurons in the spinal cord, brainstem, and motor cortex. Patients have progressive paralysis and muscle atrophy, and half of the patients die on average within 3 to 5 years of symptom onset, mostly resulting from respiratory failure. The global incidence of ALS is about 1-2 per 100 000 each year and the lifetime risk of developing ALS is estimated to be approximately 1/600 to 1/2 000.Ninety perccent of ALS are sporadic and the remaining 10% is familial, but the two forms of ALS are clinically and pathologically indistinguishable. Of the familial cases, 20% are associated with a mutation in the cytosolic copper/zinc superoxide dismutase gene (SOD1). The SOD1 mutations are predominantly missense mutations leading to substitution of one amino acid by another, for example glycine by alanine (SOD1-G93A). Because expression of a hSOD1 mutant in animals causes an adult-onset phenotype recapitulating closely the human disease both clinically and histopathologically, these transgenic animals have become a widely used ALS model. In fact, the effects of the mutant are due to the toxic gain of function, while most of SOD1 mutants retain enzymatic activity. Yet, at the present time, the actual mechanism by which mutant SOD1 initiates ALS is poorly understood, although several hypotheses have been proposed to explain the toxic effect of mutated SOD1, on the basis of extensive research on ALS: oxidative stress, glutamate excitotoxicity, formation of high-molecular-weight aggregates, mitochondrial dysfunction and apoptosis.The availability of powerful genomics technologies provides the opportunity to unravel complex regulatory and disease mechanisms. Alternative splicing of mRNA transcripts is a major form of post-transcriptional gene regulation, during which introns in a pre-mRNA are differentially removed and exons are joined to form multiple forms of mature RNA. Alternative splicing plays an important role in generating proteomic diversity, as a relatively limited number of genes may expand into very complex proteomes. It has been estimated that about three-quarters of all human genes exhibit alternative splicing. Current data suggest that greater than 90% of all human genes undergo alternative splicing. Statistics report of ASTD database Mus musculus shows that 57% genes undergo alternative splicing. Researchers are becoming aware of the dominant role of mRNA splicing in many important physiological processes such as development, physiology and disease. Affymetrix Exon Arrays are rapidly gaining popularity and becoming a standard for both gene- and exon-level expression analysis.Recent studies have shown that alternative splicing is especially prevalent in the nervous system. Zhang et al. reported that aberrant cleavage of TAR DNA-binding protein-43 (TDP-43) enhances aggregation and cellular toxicity in ALS. The glutamate transporter 1 in rodents, or EAAT2 in humans, is alternatively spliced in a complex manner, including the use of multiple 5’ and 3’ untranslated regions and several coding variants. Though alternative splicing is prevalent in the nervous system, little is known in ALS.In the present study, we used Affymetrix GeneChip? Mouse Exon 1.0 ST Array to identify genes and exons that are differentially expressed in SOD1-G93A transgenic mice. The array contains approximately 5.4 million probes grouped into 1.4 million probe sets and interrogates more than 1 million known and predicted exons. For each gene, the median number of probes is 30 to 40, usually distributed along the entire transcript sequence. The commercial availability of specific software provides a powerful tool for the study of alternative splicing (exon-level expression analysis) and interrogation of the expression of every known and predicted exon. At the same time, exon arrays provide robust gene-level expression analysis.For the first time, we have identified not only genes that are altered in expression in response to mutant SOD1, but also genes that undergo alternative splicing. The study was focused on the lumbar spinal cord which is the most affected tissue in the SOD1-G93A transgenic mice. Our study includes three parts.Part I Differential expression of genes in lumbar spinal cord of an amyotrophic lateral sclerosis mouse modelObjective: The study was aimed to identify the transcription profiles of the lumbar spinal cord removed from SOD1-G93A mice at the onset of the disease in the presence of the mutant SOD1 protein.Methods: Using the mouse exon 1.0 ST, we analyzed the transcription profiles of the lumbar spinal cord removed from three female SOD1-G93A mice at the onset of the disease and three female non-transgenic littermates as the same age. Hierarchical cluster (HC) analysis, the gene ontology (GO) hierarchy analysis and a pathway-based analysis were then carried out on the differentially expressed genes.Results: (1) The analyses identified a total of 322 transcripts that were statistically differential in the presence of the mutant SOD1 protein: 309 (95.96%) transcripts were up-regulated and 13 (4.04%) transcripts were down-regulated. Based on T-test of the results, we generated a valcano plot to compare visually the differentially expressed genes between the onset group and the control group. (2) Cluster analysis 3.0 software was used for hierarchical cluster (HC) analysis. There is a good reproducibility between samples, and the genes can be successfully distinguish between the onset group and the control group. (3) To reveal global functional differences, the gene ontology (GO) hierarchy analysis was then carried out on the differentially expressed genes. These differentially expressed genes were classified into molecular function with a P value cutoff of less than 0.001, including heparin binding, insulin-like growth factor 1 binding, superoxide-generating NADPH oxidase activity, integrin binding, lipoprotein lipase activity and chemokine activity. Taking into account the biological process of the differentially expressed genes, they were critically involved in cell adhesion, inflammatory response and immune response. Leukocyte adhesion, integrin complex, integrin-mediated signaling pathway and activation of MAPK activity were also identified. (4) We performed a pathway-based analysis of the differentially expressed genes. The differentially expressed genes based on the KEGG database analysis were involved in some pathways, such as cytokine-cytokine receptor interaction, regulation of actin cytoskeleton, natural killer cell mediated cytotoxicity, focal adhesion, complement and coagulation cascades, leukocyte transendothelial migration, cell adhesion molecules and apoptosis.Conclusion: There is significant difference between the gene expression of SOD1-G93A transgenic mice and that of their littermate control. In addition, the differentially expressed genes can be defined clearly, clarified and analyzed.PartⅡAlternative splicing of genes in lumbar spinal cord of an amyotrophic lateral sclerosis mouse modelObjective: The study was aimed to identify alternatively spliced genes of the lumbar spinal cord removed from SOD1-G93A mice at the onset of the disease in the presence of the mutant SOD1 protein.Methods: Using the mouse exon 1.0 ST, we analyzed alternatively spliced genes of the lumbar spinal cord removed from three female SOD1-G93A mice at the onset of the disease and three female non-transgenic littermates as the same age. Hierarchical cluster (HC) analysis, the gene ontology (GO) hierarchy analysis and a pathway-based analysis were then carried out on these alternatively spliced genes.Results: (1) Three hundred and thirty three probesets (44 087 probesets in all) passed our quality control criteria and were used for subsequent analyses. The three hundred and thirty three probesets belonged to 297 alternatively spliced genes, reflecting that multiple alternative splicing events might occur in a single transcript. Our result showed that 63 alternatively spliced exons were derived from 27 transcripts. Three or more alternatively spliced exons were detected individually in some transcripts, indicating a complicated pattern of splicing regulation of these transcripts in the mice. (2) HC analysis was carried out according to similarity by using cluster analysis 3.0 software. The result showed that a good correlation between two groups. (3) The GO hierarchy analysis was then carried out on the alternatively spliced genes. These alternatively spliced genes are involved in iron ion binding, chemokine activity, metalloendopeptidase activity or motor activity, according to their molecular function categories. Taking into account biological process, the genes are predominately associated with inflammatory response, immune response, chemotaxis, activation of MAPK activity and anti-apoptosis. (4) Applying the same bioinformatics analyses that were used for the differential gene expression, we found that 95 of 297 alternatively spliced genes were involved in known pathways, based on the KEGG database analysis. Significant pathways were identified, including cytokine-cytokine receptor interaction, MAPK signaling pathway, cell adhesion molecules, regulation of actin cytoskeleton, apoptosis, T cell receptor signaling pathway and natural killer cell mediated cytotoxicity.Conclusions: Two or more alternative splicing events may take place in a single transcript, which indicates the complexity of alternative splicing model. Though the overlapped part of alternative spliced genes and differently expressed ones is quite small, the GO attribution and biological pathways of the two are both much similar, which suggests the corresponding biological effects are co-regulated at the two different levels.Part III Validation of differentially expressed and alternatively spliced genes in lumbar spinal cord of an amyotrophic lateral sclerosis mouse modelObjective: The study was aimed to validate differentially expressed and alternatively spliced genes identified by exon array in the presence of the mutant SOD1 protein.Methods: Semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) was used to validate differentially expressed and alternatively spliced genes of the lumbar spinal cord removed from SOD1-G93A mice at the onset of the disease..Results: The mRNA levels of CD68 and lipoprotein lipase (LPL) were significantly increased in the lumbar spinal cords of SOD1-G93A transgenic mice compared to the controls (P < 0.001), suggesting that there existed inflammatory responses and a hypermetabolic state of lipid at the onset stage. We also carried out RT-PCR validation of differential splicing in programmed cell death 1 (PDCD1) mRNA transcription in ALS. The results showed that the exon2-skipped isoform predominated in control samples, while the exon2-included isoform predominated in disease samples. Both exon predicted to have alternative splicing and exon predicted not to have alternative splicing were confirmed by RT-PCR.Conclusions: The results we obtained indicate that the exon array technology is reliable and sensitive enough to detect differential expression at both the gene level and exon level.更多还原