1、 冠心病论文：脂质代谢及炎性通路基因与冠心病关联关系研究【中文摘要】冠状动脉粥样硬化性心脏病(Coronary Artery Disease or Coronary Atherosclerosis Disease, CAD),简称冠心病,是一种最常见的心脏病,严重危害人类的健康。作为一种复杂疾病,它的发生是由遗传因素和环境因素共同作用引起的。随着越来越多的易感基因被发现,针对动脉粥样硬化的发病机制形成了多种学说,目前被学者广泛接受的两种学说是脂肪浸润学说和炎症学说。本研究选取的 7 个候选易感基因中,与脂质代谢相关的 5 个基因是：LRP5 (low density lipoprotein r
2、eceptor-related protein 5)、LRP6 (low density lipoprotein receptor-related protein 6)、APOC1 (Apolipoprotein C-I)、ABCG1 (ATP-binding cassette, sub-family G (WHITE), member 1)和 PCSK9 (proprotein convertase subtilisin/kexin type 9),与炎症相关的 2 个基因是：TNFSF4 (Tumor necrosis factor superfamily, number 4)和 TNFR
3、SF4 (tumor necrosis factor receptor superfamily, member 4)。应用基于群体的病例对照研究设计,逐步采用多种统计分析方法,探讨中国汉族人群中,脂质代谢通路中的 5 个候选易感基因(LRP5、LRP6、APOC1. ABCG1、PCSK9)及炎性通路中的 2 个易感基因(TNFSF4、TNFRSF4)同冠心病的关联性。1.从山东大学齐鲁医院选取了 498 例病例及 509 例对照,对研究对象临床资料进行统计描述。2.对 LRP5 基因上 2 个 SNP (single nucleotide polymorphism)位点(rs41494349
4、、rs3736228)、LRP6 基因上 9 个 SNP 位点(rs2160525、rs2284396、rs4477532、rs2302685、rs7305037、rs12823243、rs1181333、rs11054731、rs17848270)、APOC1 基因上 1个 SNP 位点(rs4420638)、ABCG1 基因上 4 个 SNP 位点(rs4148082、rs1893590、rs1378577、rs1044317)、PCSK9 基因上3 个 SNP 位点(rs572512、rs2483205、rs2495477)、TNFSF4 基因上5 个 SNP 位点(rs1234314、
5、rs45454293、rs3850641、rs1234313、rs3861950)以及 TNFRSF4 基因上 1 个 SNP 位点(rs2298212)进行基因分型。3.针对每个 SNP 位点,利用 Armitage 趋势检验检测其与冠心病的关联性。4.针对每个基因,将从该基因上选取的 SNP 位点及混杂因素(年龄、性别、体重指数、收缩压、舒张压、总胆固醇、甘油三酯和葡萄糖 8 个指标)引入 logistic 回归模型中,以检测在排除混杂因素影响后,该基因上 SNP 位点与冠心病的关联性。5.针对选取位点数大于 1 的每个基因,利用 SAS 9.1.3 软件的 genetics 模块进行单倍
6、型分析。6.针对选取位点数大于 1 的每个基因,采用基于主成分的 logistic 回归分析来研究每个基因与冠心病的关联性。7.采用偏最小二乘路径模型(partial least square path modeling, PLS-PM)来研究这些基因与冠心病的关联性,在 R 软件中利用 PLS-PM 软件包来实现。1.对单个 SNP 位点的 Armitage 趋势检验结果显示,APOC1 基因的rs4420638 (P=0.0001), PCSK9 基因的 rs572512 (P=0.0308)以及TNFSF4 基因的 rs3861950 (P=0.0324)这三个 SNP 位点与冠心病存在
7、关联性。2.调整混杂因素的 logistic 回归分析结果显示,LRP5 基因的 rs41494349 (P=0.0372), LRP6 基因的 rs4477532 (P=0.0130)、rsl2823243 (P=0.0117)和 rs11054731 (P=0.0024), APOC1 基因的rs4420638 (P=0.0021)以及 PCSK9 基因的 rs2483205 (0.0402)这六个 SNP 位点与冠心病存在关联性。3.单体型分析结果显示,LRP6、ABCG1 和 TNFSF4 这三个基因存在在病例组和对照组中的频率差异具有统计学意义的单体型。4.基于主成分的 logist
8、ic 回归分析结果显示,TNFSF4 基因的第一主成分(PC1)与冠心病具有统计学关联性(P=0.0236)。5.对 LRP5 基因与 LRP6 基因交互作用的 PLS-PM 处理结果显示,LRP5-LRP6 路径系数在病例组和对照组的差异具有显著性(P=0.0099)。对 TNFSF4 基因与 TNFRSF4 基因交互作用的 PLS-PM处理结果显示,TNFSF4- TNFRSF4 路径系数在病例组和对照组的差异不具有显著性(P=0.4455)。对LRP5、LRP6、APOC1、ABCG1、PCSK9 这 5 个基因与血脂之间的交互作用 PLS-PM 的处理结果显示,LRP6-BL (P=0
9、.0196)以及 PCSK9-BL (P=0.0392)这两个路径系数在病例组和对照组的差异具有显著性。1.在基于主成分的 logistic 回归分析方法处理下,TNFSF4 基因同冠心病存在关联性,而 LRP5、LRP6、ABCG1、PCSK9 这 4 个基因同关心病不存在关联性。2.LRP5 基因与 LRP6 基因的交互作用与冠心病存在关联性,而 TNFSF4 基因与 TNFRSF4 基因的交互作用与冠心病不存在关联性。与脂质代谢相关的 5 个基因中(LRP5、LRP6、APOC1、ABCG1、PCSK9), LRP6 基因与 PCSK9 基因在病例组与对照组中对血脂影响具有统计学意义。【
10、英文摘要】BackgroundCoronary atherosclerotic disease (CAD), as the most common type of heart disease, is very dangerous for human health. As a kind of complex disease, CAD results from the interaction of a number of susceptibility genes and environmental factors. With the discovery of more and more susce
11、ptibility genes, there form many kinds of hypothesis about the pathogenesis of CAD. At present, fatty infiltration hypothesis and inflammatory hypothesis are most accepted by researchers.In my research,7 candidate susceptibility genes were selected. LRP5 (low density lipoprotein receptor-related pro
12、tein 5), LRP6 (low density lipoprotein receptor-related protein 6), APOC1 (Apolipoprotein C-I), ABCG1 (ATP-binding cassette, sub-family G, member 1) and PCSK9 (proprotein convertase subtilisin/kexin type 9), these five genes were related with lipid metabolism. TNFSF4 (Tumor necrosis factor superfami
13、ly, number 4) and TNFRSF4 (tumor necrosis factor receptor superfamily, member 4), these two genes were related with inflammation.By applying population-based case-control study design, we used many kinds of statistical methods to explore the association between CAD and 5 candidate susceptibility gen
14、es (LRP5, LRP6, APOC1, ABCG1, PCSK9) in lipid metabolism pathway and the association between CAD and 2 candidate susceptibility genes (TNFSF4, TNFRSF4) in inflammatory pathway in Chinese Han population.Methods1. We collected 498 cases and 509 controls from Qilu Hospital of Shandong University and co
15、nducted statistical description of the clinical data of them.2.2 SNPs of LRP5 (rs41494349, rs3736228),9 SNPs of LRP6 (rs2160525, rs2284396, rs4477532, rs2302685, rs7305037, rs12823243, rs1181333, rs11054731, rs17848270),1 SNP of APOC1 (rs4420638),4 SNPs of ABCG1 (rs4148082, rs1893590, rs1378577, rs1
16、044317),3 SNPs of PCSK9 (rs572512, rs2483205, rs2495477),5 SNPs of TNFSF4 (rs1234314, rs45454293, rs3850641, rs1234313, rs3861950) and 1 SNP of TNFRSF4 (rs2298212) were genotyped.3. We used Armitage trend test to detect the association between every single SNP and CAD.4. For every single gene, we pu
17、t the SNPs selected from the gene and the confounding factors (sex, age, Body Mass Index, Systolic Blood Pressure, Diastolic Blood Pressure, Total Cholesterol, Triglyceride, Glucose) into logistic regression model. After eliminating the impact of confounding factors, we detected the association betw
18、een every single SNP and CAD.5. For genes with more than 1 selected SNP, we used Genetics module of SAS 9.1.3 software to do haplotype analysis.6. For genes with more than 1 selected SNP, we used principal component-based logistic analysis to detect the association between every gene and CAD.7. We a
19、pplied partial least squares path model to study the association between these 7 genes and CAD. This was done by using PLS-PM software package of R software.Results 1. The results of Armitage trend test suggested that, rs4420638 of APOC1 (P=0.0001), rs572512 of PCSK9 (P=0.0308) and rs3861950 of TNFS
20、F4 (P=0.0324) these 3 SNPs were significantly associated with CAD.2. The results of logistic analysis adjusting confounding factors suggested that, rs41494349 of LRP5 (P=0.0372), rs4477532 of LRP6 (P=0.0130), rsl2823243 of LRP6 (P=0.0117), rs 11054731 of LRP6 (P=0.0024), rs4420638 of APOC1 (P=0.0021
21、 and rs2483205 of PCSK9 (P=0.0402) these 6 SNPs were significantly associated with CAD.3. The results of haplotype analysis suggested that, there existed haplotypes which were significantly different between case and control in LRP6, ABCG1 and TNFSF4.4. The results of principal component-based logis
22、tic analysis suggested that, the first principal component of TNFSF4 has statistical significance (P=0.0236).5. The results of applying PLS-PM to detect the interaction between LRP5 and LRP6 suggested that, the path coefficient was significant between case and control (P=0.0099). The results of appl
23、ying PLS-PM to detect the interaction between TNFSF4 and TNFRSF4 suggested that, the path coefficient was not significant between case and control (P=0.4455). The results of applying PLS-PM to detect the interactions between LRP5, LRP6, APOC1, ABCG1, PCSK9 this 5 genes and blood lipid (BL) suggested
24、 that, the path coefficients of LRP6-BL (P=0.0196) and PCSK9-BL (P=0.0392) were significant between case and control.Conclusions1. When applying principal component-based logistic analysis, TNFSF4 is associated with CAD, but LRP5, LRP6, ABCG1 and PCSK9 these 4 genes are not associated with CAD.2. Th
25、e interaction between LRP5 and LRP6 is associated with CAD. The interaction between TNFSF4 and TNFRSF4 is not associated with CAD. In the 5 genes (LRP5, LRP6, APOC1, ABCG1, PCSK9) which are connected with lipid metabolism, LRP6 and PCSK9 are significantly associated with blood lipid between case and
26、 control.【关键词】冠心病 主成分分析 logistic 回归分析 偏最小二乘路径模型【英文关键词】Coronary atherosclerotic disease principal component analysis logistic regression analysis partial least squares path model【目录】脂质代谢及炎性通路基因与冠心病关联关系研究 中文摘要 8-11 ABSTRACT 11-14 符号说明 15-16 前言 16-17 第一章 单个基因与冠心病关联关系的研究 17-38 前言 17-19 材料与方法 19-22 1 研究对
27、象 19-20 2 SNP 等位基因分型 20 3 统计分析方法 20-22 结果 22-36 1 协变量的统计描述 22-23 2 单位点 Armitage 趋势检验 23-26 3 调整混杂因素的 logistic 回归分析 26-30 4 单体型分析 30-33 5 基于主成分的 logistic 回归分析 33-36 讨论 36-38 结论 38 第二章 多个基因与冠心病关联关系的研究 38-54 前言 38-39 材料与方法 39-45 1 研究对象 39 2 SNP等位基因分型 39-40 3 统计分析方法:PLS-PM 的基本原理 40-45 结果 45-51 1 模型一 PLS-PM 处理结果 45-47 2 模型二 PLS-PM 处理结果 47-49 3 模型三 PLS-PM 处理结果 49-51 讨论 51-53 结论 53-54 附录 54-70 附录 1 LRP6 基因单体型分析结果表 54-63 附录 2 软件 SAS 9.1.3 中程序 63-66 附录 3 软件 R 中程序 66-70 1 模型一 PLS-PM 程序 66-67 2 模型二 PLS-PM 程序 67-68 3 模型三 PLS-PM 程序 68-70 参考文献 70-75 致谢 75-76 攻读学位期间发表的学术论文 76-77 学位论文评阅及答辩情况表 77
