【作者】 高宏生；

【导师】 伍瑞昌；

【作者基本信息】 中国人民解放军军事医学科学院 ， 军事预防医学， 2008， 博士

【摘要】 矽肺是最为严重的一种职业病之一,是由于长期吸入大量含有游离二氧化硅粉尘所引起的一种不可逆转的渐进性加重的肺间质纤维化疾病。引起矽肺的主要职业有武警黄金水电部队的开矿挖掘作业,地方各行业如矿山开采、冶金工业中的原料破碎、陶瓷工业的原料加工,这些职业均可导致不同程度的肺纤维化。矽肺患病率逐年增多,造成直接和间接经济损失巨大,是危害最严重的职业病之一,而且已经成为重大社会公共卫生问题。国内外学者对矽肺纤维化的单个细胞因子作用虽已取得了一定的研究进展,但其系统网络调控机制尚未阐明。目的确定矽肺纤维化相关的生物活性介质,构建生物活性介质调控网络及其基因调控网络,以及二者整合网络,以此系统阐明矽肺纤维化的网络调控机制。方法本文主要应用系统生物学的方法构建生物活性介质调控网络。1.生物活性介质的确定。采用RevMan4.2进行meta分析,对研究的偏倚进行漏斗图、异质性、敏感性分析后进而确定相关活性介质。2.初始调控网络的构建。利用meta分析后数据,通过三次样条插值,采用微分方程模型并应用最小二乘法求解加权矩阵,构建活性介质调控网络。3.大鼠矽肺模型及实验后调控网络构建。应用气管暴露法注入二氧化硅粉尘悬液或生理盐水建立大鼠矽肺模型或对照模型。随机将Wistar大鼠分为矽尘组和对照组,每组再分为第1d,第3d,第7d,第14d,第21d,第28d共6个时间点,每个时间点分别取8只大鼠处死,收集相应血清和肺组织后系统检测相关生物活性介质。采用天狼猩红染色法结合偏振光显微镜观察肺组织胶原变化,并用图像分析系统定量分析Ⅰ、Ⅲ型胶原面积比;采用ELISA法系统测定血清中活性介质NF-κB、IL-1β、IL-10、TNF-α、INF-γ、TGF-β1、GM-CSF的含量;利用硝酸还原酶法测定NO含量。通过系统检测数据应用相关系数模型和微分方程模型进一步构建调控网络。4.干扰实验对调控网络的校正。同样应用气管暴露法建立大鼠矽肺模型,随机将32只大鼠分为TNF-α干扰组和矽尘组,每组再分为第7d和第14d两个时间点,每个时间点取8只大鼠并检测以上相同活性介质,通过比较两组之间各指标的差异校正调控网络。5.基因调控网络实验确证和扩充。造模后第14d用Trizol提取染矽尘组和对照组中肺组织总RNA,通过Illumina大鼠全基因表达芯片的检测,结合Diffscore差异分值筛选差异表达基因。通过KEGG、GO、DAVID等数据库平台,应用功能富集分析方法以及Cluster、TreeView等工具建立相关细胞因子基因聚类树,并利用差异表达基因相关系数矩阵构建基因调控网络,整合活性介质和基因调控网络,扩充和完善生物活性介质调控网络。结果1.meta分析后肺纤维化相关活性介质为TNF-α、TGF-β1、IFN-γ、MCP-1、IL-4、IL-6、IL-18、MMP-9、IL-10、IL-1β、IL-5、IL-8、IGF-1、GM-CSF、MMP-2、EGF。其中肺组织中TNF-α随时间的变化并不是直线升高,而是呈波浪性升高。肺组织、巨噬细胞、血清活性介质TNF-α各时点变化并不一致,其中巨噬细胞在第1天变化比其余两者更为明显,肺组织除第1天外各时点变化具有显著性差异,血清在各个时点变化均较肺组织表达弱一些。2.当权值阈值为2σ时,肺纤维化活性介质调控网络中重要节点为:IFN-γ、IL-10、IL-1β、IL-18、IL-6、IGF-1、GM-CSF、TGF-β1、TNF-α;非重要节点的活性介质为IL-4、IL-5、IL-8、MCP-1、MMP-2、MMP-9、EGF。当权值阈值为3σ时,肺纤维化活性介质调控网络中重要的节点变为:IFN-γ、IL-10、IL-1β、GM-CSF、TGF-β1、TNF-α;非重要节点的活性介质为IL-18、IL-6、IGF-1、IL-4、IL-5、IL-8、MCP-1、MMP-2、MMP-9、EGF。3.肺组织天狼猩红染色显微镜及图像分析仪观察表明第21天、第28天矽肺模型是成功的。实验组与对照组比较,血清中NF-κB和NO各时间点(除第21天NO外)表达具有统计学意义,动态表达比趋势较为一致。TGF-β1第21天、第28天表达增加,而IFN-γ第21天、第28天表达有降低趋势。TNF-α表达比呈先降低后增高趋势,但整体上与对照组相比没有统计学意义。IL-1β、IL-10、GM-CSF不同时间点动态表达比均大于1,前两者先升高后降低,后者呈波浪性升高。4.调控网络中度大于等于5的节点介质为IFN-γ、GM-CSF、TGF-β1、TNF-α;度小于5的节点为IL-10、IL-1β、NF-κB、NO、COLⅠ、COLⅢ。TNF-α可溶性受体干扰实验是可行的,该受体能够抑制矽尘诱导的大鼠肺组织Ⅰ、Ⅲ型胶原的增加,抑制外周血清中NO、IL-1β、TGF-β1、NF-κB的蛋白表达。原调控网络中同TNF-α相关联的介质可能还包括COL I、GM-CSF等介质。5.矽肺纤维化血清中通过功能富集分类得出相关细胞因子基因共有29个,其调控网络中“度”大于5的关键基因为CCL7、GRN、IL-18、SFTPD、SPN、SPP1、TNF-α、TNFRSF。结论矽尘诱发机体活性介质的变化通过网络状态调控肺纤维化的形成。16种生物活性介质(TGF-β1,IFN-γ,MCP-1, IL-4, IL-6,IL-18,MMP-9, TNF-α,IL-10,IL-1β,IL-5,IL-8, IGF-1, GM-CSF, MMP-2, EGF)可能参与肺纤维化的调控,且具有一定的时空变化规律。时间趋势上活性介质的含量不是呈直线性,而是呈波浪性升高或降低趋势,各时点变化并不一致。空间表现上活性介质在细胞中出现最早,血清中出现最晚;各时点在肺组织中表达最强,血清中表达较弱。肺纤维化生物活性介质调控网络中重要节点介质为:IFN-γ、IL-10、IL-1β、GM-CSF、TGF-β1、TNF-α、NF-κB、NO;非重要节点为IL-18、IL-6、IGF-1、IL-4、IL-5、IL-8、MCP-1、MMP-2、MMP-9、EGF。在肺纤维化活性介质调控网络中,NO和NF-κB是一个启动中枢,而TNF-α、TGF-β1既担当启动子又担当致纤维化调控网络的中间传递子,GM-CSF、IL-1β、IFN-γ呈现级联瀑布放大或抑制网络效应。在网络效应活性介质促进和抑制功能不平衡的条件下形成肺纤维。更多还原

【Abstract】 Silicosis, one of most severe occupation diseases, is an gradually aggravated and irreversiblely pulmonary interstitial fibrosis disease, induced by long-term inhalation of free silica powder dust, such as the dust of mining excavation in the gold and hydropower armed police force, civil mine exploitation, raw material quassation in the procedure of metallurgy and ceramics, which can cause momentous lose of economy directly and indirectly, and induce substantially aggravated problems in the fields of society and public health. Up to now, the study of the single cytokine effectiveness of silicosis fibrosis had got some achievements, but the mechanism of systematic regulatory network is still not being demonstrated.Objective To ascertain related bioactivity mediums of silica-reduced pulmonary fibrosis in rats, to construct the regulator network of the material bioactivity and its genes as well as both integration, to evaluate network regulatory mechanism of silicosis fibrosis systematically.Method The regulatory network of the bioactivity mediums was constructed by the method of systems biology in this paper. 1. Identification of the bioactivity mediums. The mediums of biological activity were identified by the meta-analysis software RevMan 4.2. The investigative biases were also evaluated by analytical methods such as funnel plot, heterology and susceptibility. 2. Construction of initial regulatory network with meta-analysis database. The potential regulatory network pathway was constructed and analyzed by meta–analysis, cubic spline interpolation, differential model and weighting matrix obtained with least square method. 3. Silicosis model in rats and construction of initial regulatory network with animal experiment database. The rats models in silicosis or control group were established with intratracheal infusion of silica dust suspension or physiological saline by trachea exposure method. Wistar rats were divided randomly into two groups including control group and silicosis group. At each time point such as day 1, 3, 7, 14, 21, and 28 after establishment of the rat model, eight rats from each group were sacrificed. After the serum and pulmonary tissue of the model were required, the related bioactivity mediums were detected systematically. The abnormal syntheses of collagenⅠandⅢin pulmonary tissue with Sirius red staining were detected by polarization microscopy. The area percentages of collagensⅠandⅢwere quantitatively analyzed by image analysis systems. The concentration of NF-κB, IL-1β, IL-10, TNF-α, INF-γ, TGF-β1, and GM-CSF protein in serum was measured by ELISA method. And NO content was determined by nitric acid reeducates method. The regulatory network was constructed with correlation coefficient and differential equation model by systematical detection database. 4. Revise of regulatory network by intervention experiment. Silicosis model in rats was established by the same trachea exposure method. 32 rats were divided into two groups randomly, i.e. TNF-αintervention group and silica group, which were observed at the 7th day and 14th day after establishment of the animal model, respectively. Eight rats from each group at each time point were sacrificed, and the same bioactivity mediums as above-mentioned were detected systematically. The regulatory network was reconstructed and revised by comparing the difference of each bioactivity mediums in the two groups. 5. The experiment ascertainment and expansion of the gene regulatory network. The total RNA of pulmonary tissue in silicosis and control group was abstracted by Trizol method at the 14th day after model construction. The different express genes were screened with diffscore value by illumina gene expression microarray chip. The cluster of the related cytokines gene was analysised by the software including Cluster and TreeView and by the database platform such as KEGG, GO and DAVID. Then the gene regulatory network was established by correlation coefficient matrix and was also integrated with regulatory network of bioactivity medium for the purpose of expanding and improving the network.Result 1. The related bioactivity mediums of silica-reduced lung fibrosis were ascertained to be TNF-α,TGF-β1,IFN-γ,MCP-1, IL-4, IL-6,IL-18,MMP-9, IL-10,IL-1β,IL-5,IL-8, IGF-1, GM-CSF, MMP-2 and EGF by meta-analysis. Of the mediums, TNF-αcontent in lung tissue increased not on a straight line but on a wave line in a time-dependent manner. The medium level wasn’t coincident in lung tissue, macrophage and serum. Specially, the medium content in macrophage was more than the others’at the 1st day after model constructing, and the medium of pulmonary tissue in silica group was more significant statistically than that in control group except on the 1st day, and the mediums content in the serum was lower than that in the lung tissues. 2. The mediums, IFN-γ,IL-10,IL-1β,IL-18,IL-6,IGF-1, GM-CSF, TGF-β1 and TNF-αwere the key nodal points in regulatory network of the pulmonary fibrosis, but IL-4, IL-5,IL-8,MCP-1, MMP-2, MMP-9 and EGF were not in the regulatory, if threshold value in weighting matrix exceeded absolute value 2σ. The mediums, IFN-γ,IL-10,IL-1β, GM-CSF, TGF-β1 and TNF-αwere the key nodal points in regulatory network of the pulmonary fibrosis, but IL-18, IL-6,IGF-1, IL-4, IL-5, IL-8,MCP-1, MMP-2, MMP-9 and EGF were not in the regulatory, if threshold value in weighting matrix exceed absolute value 3σ. 3. The lung tissue with sirius red staining were observed by microscopy and image analysator, with the result that the rat model of silicosis was established successfully at day 21, 28. Compared with that in the control group, the content of NF-κB and NO in serum increased significantly at all the time points in silicosis group except for NO at day 21, and dynamic state express ratio tendency of both mediums was coincident. At the 21st day and 28th day, the express ratio of TGF-β1 in silicosis group was higher, but that of IFN-γwas lower than one in the control group. TNF-αprotein level decreased in earlier period and increased in the later period, but had no statistical significance during the whole experiment in both comparisons. The dynamic express ratio of the content of IL-1β, IL-10, and GM-CSF exceed 1 at various time point. Of the total mediums, the tendency of the former two protein level was elevated in fibrosis prophase and degraded in fibrosis anaphase, but the tendency of the later GM-CSF was raised by waved method. 4. The mediums in regulatory network in which the degree weren’t less than 5 had IFN-γ, GM-CSF, TGF-β1, TNF-α, and the mediums of which the degree was less than 5 had IL-10, IL-1β, NF-κB, NO, COLⅠ, COLⅢ. The soluble TNF-αreceptor which was feasible to the intervention of pulmonary fibrosis could suppress formation of collagenⅠandⅢin lung tissues, and play role of anti-fibrosis through peripheral serum mediums such as NO, NF-κB, TGF-β1, and IL-1βin cytokine network in silicotic process. In the above regulatory network, the mediums in the pathway related to TNF-αshould include COL I, GM-CSF. 5. 29 genes related to cytokines function categorization in serum were divided to the regulatory network in silicosis fibrosis. The key genes with degree exceeding 5 in regulatory network were CCL7, GRN, IL-18, SFTPD, SPN, SPP1, TNF-αand TNFRSF.Conclusion The pulmonary fibrosis was regulated by the network method of silica-induced changes of bioactivity mediums. The regulatory factors which involve in pulmonary fibrosis consist of 16 bioactive mediums including TGF-β1,IFN-γ,MCP-1, IL-4, IL-6,IL-18,MMP-9, TNF-α,IL-10,IL-1β,IL-5,IL-8, IGF-1, GM-CSF, MMP-2, EGF with space-time transmutation rule. In the fibrosis process, the mediums’levels which weren’t consistent in various time points increased or decreased not by straight line method, but by wave line method. In the various specimens related pumomary fibrosis, the detection of the mediums’variation in the related cells was the earliest than that in the others, but the one in serums was the latest than that in the others. The mediums levels in lung tissue were the maximum, but the one in serums was the minimum. The mediums, IFN-γ, IL-10, IL-1β, GM-CSF, NF-κB, NO, TGF-β1 and TNF-α, were the key nodal points in regulatory network of the pulmonary fibrosis, but IL-18, IL-6,IGF-1, IL-4, IL-5, IL-8,MCP-1, MMP-2, MMP-9 and EGF weren’t in the regulatory. NO and NF-κB were a prime center in the regulatory network of pulmonary fibrosis, and TNF-α, TGF-β1 not only acted as the promoter, but also held the post of intermediate transferring factor in regulatory network which caused lung fibrosis. The changed content of GM-CSF,IL-1β,IFN-γamplified and inhibited cascade network effective, respectively. Then network effective of these bioactivity mediums resulted in lung fibrosis under the condition of imbalance of promoting agent and suppressing agent.更多还原