中文摘要

第一部分 microRNA-31在β-氨基丙腈诱导的主动脉瘤/夹层小鼠主动脉中的表达

目的：主动脉瘤/夹层是十分严重的心血管疾病。目前，主动脉瘤/夹层的分子机制尚不完全清楚，基因突变无法解释主动脉瘤/夹层发病的全部问题。表观遗传修饰是指在不改变基因序列的前提下调控基因的表达，包括DNA甲基化、翻译后组蛋白修饰和非编码RNA。microRNAs（miRNAs）是非编码RNA的一种。microRNA-31（miR-31）在多种组织和细胞中广泛表达，与肿瘤、心血管病等疾病的发生和发展关系密切。我们前期通过miRNAs芯片检测了主动脉夹层患者的主动脉组织对比正常主动脉中的miRNA表达水平，发现miR-31在人主动脉夹层病变组织中表达升高，提示miR-31可能在主动脉夹层的发病机制中起作用。血管紧张素 Ⅱ（Angiotensin Ⅱ, AngⅡ）是构建小鼠主动脉瘤/夹层模型常用的方法。AngII诱导的主动脉瘤/夹层小鼠模型中，过表达miR-31-5p增加主动脉瘤/夹层发病率和最大主动脉直径。miR-31可能在主动脉夹层患者和AngⅡ诱导的主动脉瘤/夹层小鼠的发病机制中起作用。AngⅡ通过升高血压等方式诱导小鼠主动脉瘤/夹层形成。除高血压因素外，主动脉瘤/主动脉夹层的发生机制中还有很多非高血压因素，miR-31是否在非高血压因素导致的主动脉夹层/动脉瘤中同样发挥重要作用尚不完全清楚。β-氨基丙腈（β-Aminopropionitrile monofumarate, BAPN）与Ang II诱导产生主动脉瘤/夹层小鼠的机制不同，BAPN是赖氨酰氧化酶抑制剂，通过抑制胶原蛋白和弹力蛋白的交联进而破坏主动脉壁结构，不通过升压作用诱导产生小鼠主动脉瘤/夹层模型。本文探究了BAPN诱导的主动脉瘤/夹层小鼠中miR-31的表达水平。

方法：选取3周龄的C57BL/6野生型雄性小鼠（n=40），随机分为BAPN组（n=20）和对照组（n=20），BAPN组小鼠饮用含有0.4%浓度BAPN的饮用水，对照组小鼠饮用普通饮用水，实验周期为28天。造模结束后测量小鼠的主动脉直径和血压，统计小鼠生存率和主动脉瘤/夹层发生率。分离两组小鼠主动脉组织，制作石蜡切片并进行苏木素伊红染色和弹力纤维染色；提取小鼠主动脉组织总RNA，应用加尾法、茎环法逆转录和实时定量PCR对比分析小鼠主动脉组织miR-31的表达水平。

结果：存活曲线显示BAPN组小鼠生存率显著降低（45% : 100%，P<0.01）。BAPN组小鼠主动脉瘤/夹层的发生率为75%，主动脉瘤/夹层破裂死亡率为55%。与对照组相比，BAPN组小鼠升主动脉最大直径[（1.86±0.51） mm vs.（1.18±0.11）mm, P<0.01]和主动脉弓最大直径[（1.95±0.50 ）mm vs.（1.16±0.17） mm, P<0.01]显著大于对照组；两组之间收缩压[（101.20±5.26） mmHg vs.（100.60±7.56 ）mmHg, P=0.891]和舒张压[（59.40±9.94）mmHg vs.（62.00±5.76）mmHg, P=0.600]差异无统计学意义。使用加尾法逆转录和实时定量PCR检测BAPN诱导的小鼠主动脉瘤/夹层主动脉组织中miR-31的表达水平，主动脉瘤/夹层主动脉组织中miR-31的表达水平有升高趋势，但差异无统计学意义（1.11±1.41 vs. 1.00±1.26，P=0.686）。扩大样本量，使用茎环法逆转录和实时定量PCR检测小鼠主动脉组织miR-31的表达水平，BAPN诱导的主动脉瘤/夹层小鼠主动脉组织中miR-31表达水平对比对照组有升高趋势，但差异依旧无统计学意义（1.73±1.69 vs. 1.00±2.65，P=0.328）。

结论：BAPN诱导的主动脉瘤/夹层小鼠对比对照组小鼠主动脉组织中miR-31的表达水平有升高趋势，但无显著差异，提示miR-31可能不参与BAPN诱导的小鼠主动脉瘤/夹层形成的分子机制。基于miR-31在AngⅡ诱导产生的小鼠主动脉瘤/夹层中可能起到的重要作用，需要扩大样本量进一步探讨BAPN诱导对比AngII诱导产生的小鼠主动脉瘤/夹层组织中miR-31表达水平的差异及意义。

关键词：主动脉瘤；主动脉夹层；microRNA；β-氨基丙腈。

第二部分 Stanford-A型急性主动脉夹层的DNA甲基化改变

目的：急性主动脉夹层（acute aortic dissection, AAD）是十分严重的心血管危急重症，死亡率高且伴有严重的并发症。Stanford-A型急性主动脉夹层占主动脉夹层病例的近75%，是最常见的主动脉夹层类型，其特点是发病突然、进展迅速、预后不良。DNA甲基化是一种主要的表观遗传机制，可由DNA甲基转移酶（DNA methyltransferases , DNMTs）介导。DNA高甲基化可降低基因启动子的转录活性导致基因沉默，而持续的低甲基化可以使基因表达上调。我们前期的研究发现，主动脉夹层样本中DNMT1和DNMT3b的蛋白表达水平显着下降，这提示DNA甲基化在主动脉夹层中的潜在作用。最近，部分研究关注了主动脉夹层中的 DNA甲基化改变，但主动脉夹层中的 DNA甲基化模式仍不完全清楚。我们设计了一个两阶段的研究，包括探索阶段和验证阶段，以探究Stanford-A型急性主动脉夹层患者主动脉组织DNA甲基化的改变。

方法：在探索阶段，使用Infinium Human Methylation 450K BeadChip检测了主动脉夹层患者主动脉组织DNA甲基化的改变，本阶段共纳入4名Stanford-A型主动脉夹层患者和4名对照组。使用京都基因和基因组百科全书（Kyoto Encyclopedia of Genes and Genomes , KEGG）途径和基因本体数据库（gene ontology, GO）分析基因富集情况。根据生物信息学分析结果选择候选基因。在验证阶段，在由16名主动脉夹层患者和7名对照组组成的验证队列中，通过焦磷酸测序验证候选基因的DNA甲基化水平。使用Western Blot验证两组主动脉组织中候选基因的蛋白表达水平。

结果：在探索阶段，共发现主动脉夹层组中的589个差异甲基化位点，包括315个低甲基化位点和274个高甲基化位点。约55.2%的差异甲基化位点位于body区域，21.8%的差异甲基化位点位于5'UTR区；在TSS200区域、5'UTR区域、1stExon区域、Body区域和3’UTR区域，低甲基化位点多于高甲基化位点。主动脉夹层组和对照组之间的589个差异甲基化位点对应197个差异甲基化基因，包括96个高甲基化基因和101个低甲基化基因。KEGG结果显示，高甲基化基因主要富集在神经活动配体-受体相互作用、谷氨酸能突触、松弛素信号通路、磷脂酶D信号通路、MAPK信号通路等，低甲基化基因主要富集在癌症中的蛋白多糖、TNF 信号通路、细胞凋亡、MAPK 信号通路等。对高甲基化基因和低甲基化基因分别进行GO分析，发现均主要富集在蛋白结合条目上。根据生物信息学分析结果，选取血管生成素-2（Angiopoietin -2,  ANGPT2）、Fas细胞表面死亡受体（Fas Cell Surface Death Receptor, Fas）、双特异性磷酸酶6（dual-specificity phosphatase 6, DUSP6）、软骨细胞Ezrin样蛋白（ARH/RhoGEF And Pleckstrin Domain Protein 1, FARP1）和胱天蛋白酶募集域蛋白6（Caspase recruitment domain protein 6, CARD6）五个候选基因，在验证阶段使用焦磷酸测序验证其CpG位点的DNA甲基化情况。验证阶段的结果显示候选基因的DNA甲基化情况与探索阶段一致，再次证实主动脉夹层患者存在DNA甲基化的差异。CpG位点cg26478401定位在Fas基因TSS区域1500bp内的位置，为主动脉夹层组显著低甲基化的位点。既往研究发现主动脉夹层和主动脉瘤患者主动脉组织Fas基因的蛋白表达增加，本研究证实主动脉夹层患者主动脉组织中Fas蛋白表达水平上调了约1.78倍（1.00±0.34 vs.1.78±0.16, P＜0.05），提示 DNA甲基化在调节基因表达和主动脉夹层发病机制中的潜在作用。

结论：Stanford-A主动脉夹层患者的主动脉组织中存在DNA甲基化的改变，并与基因表达改变有关，进而可能参与主动脉夹层发病机制。 

关键词：DNA甲基化; 表观遗传学; 急性主动脉夹层; 血管。

Abstract

Part I. Expression of microRNA -31 in the aorta tissues of β-Aminopropionitrile monofumarate-induced aortic aneurysm and dissection mice

Objective: Aortic aneurysm/dissection is a severe cardiovascular disease. At present, the molecular mechanism of aortic aneurysm/dissection is not fully understood, and genetic mutations cannot explain all the mechanisms of aortic aneurysm/dissection. Epigenetic modifications refer to the regulation of gene expression without altering the gene sequences, including DNA methylation, post-translational histone modifications, and non-coding RNAs. microRNA (miRNA) is a class of non-coding RNA. microRNA-31 (miR-31) is widely expressed in a variety of tissues and cells and is closely related to the development and progression of various diseases such as tumors and cardiovascular disease. We previously examined the miRNAs expression profiles of aorta tissues using microarray in aortic dissection patients and control group. And we found that miR-31 was up-regulated in aortic dissection, which indicated the possible role of miR-31 in the pathogenesis of aortic dissection. Angiotensin II (Ang II) is commonly used to induce aortic aneurysm/dissection models in mice. In the AngII-induced mouse model of aortic aneurysm/dissection, overexpression of miR-31-5p increased the incidence of aortic aneurysm/dissection and the maximum aortic diameter. miR-31 may play an important role in aortic dissection patients and Ang II-induced aortic aneurysm/dissection mice. Ang II induces aortic aneurysm/dissection in mice through elevating blood pressure, et al. However, there are many non-hypertensive factors in the pathogenesis of aortic aneurysm/dissection. Whether miR-31 also plays an important role in aortic aneurysm/dissection caused by non-hypertensive factors is unclear. β-Aminopropionitrile monofumarate (BAPN) and Ang II induce aortic aneurysm/dissection with different mechanisms. BAPN is a lysyl oxidase inhibitor that induces aortic aneurysm/ dissection by inhibiting the cross-linking of collagen and elastin and disrupting the structure of the aortic wall without raising the blood pressure. The present study was aimed to explore the expression level of miR-31 in the aorta tissues of BAPN-induced aortic aneurysm/dissection mice.

Methods: Three-week-old C57BL/6 wild-type male mice (n=40) were randomly divided into BAPN group (n =20) and control group (n =20). The mice in BAPN group were given the drinking water containing 0.4% BAPN for 28 days while the mice in control group were given the ordinary drinking water. The aortic diameters and blood pressures of mice were measured after model construction. The survival rates and incidences of aortic aneurysm/dissection were counted. We isolated the aorta tissues of mice. Paraffin sections were prepared and stained with hematoxylin-eosin and Verhoeff’s Van Gieson, and total RNA was extracted. Poly (T) adaptor and stem-loop reverse transcription and real-time quantitative PCR were used to detect miR-31 expression level in aorta tissues between the BAPN-induced aortic aneurysm/dissection mice and the control group.

Results: The survival curve showed that the survival rate of mice in the BAPN group was significantly decreased (45% vs. 100%，P<0.01). In the BAPN group, the incidence of aortic aneurysm/dissection was 75%, and the mortality rate due to aortic aneurysm/dissection rupture was 55%. The maximal aortic diameters of ascending aorta [(1.86±0.51) mm vs. (1.18±0.11) mm, P<0.01] and aortic arch [(1.95±0.50) mm vs. (1.16±0.17) mm, P<0.01] of BAPN group were larger than the control group. No statistical difference was observed between BAPN group and control group in systolic blood pressure [(101.20±5.26) mmHg vs. (100.60±7.56) mmHg, P=0.891] and diastolic blood pressure [(59.40±9.94) mmHg vs. (62.00±5.76) mmHg, P=0.600]. Compared with the control group, the expression level of miR-31 in aorta tissues of BAPN-induced aortic aneurysm/dissection mice had an increasing trend with no statistical difference detected by poly(T) adaptor real-time quantitative PCR (1.11±1.41 vs 1.00±1.26，P=0.686). We used stem-loop real-time quantitative PCR to detect the expression level of miR-31 in aorta tissues of the two groups in a larger sample size, and got the similar result (1.73±1.69 vs 1.00±2.65，P=0.328).

Conclusion: The expression level of miR-31 in aorta tissues of BAPN-induced aortic aneurysm/ dissection mice showed an increasing trend with no statistical difference compared with the control group, suggesting that miR-31 may not be involved in mechanisms of BAPN-induced aortic aneurysm/dissection. Based on the possible important role of miR-31 in aortic aneurysm/dissection mice induced by AngII, the expression level of miR-31 in aorta tissues of BAPN-induced and AngII-induced aortic aneurysm/dissection mice needs to be compared and explored in a larger sample size.

Key Words: Aortic aneurysm; Aortic dissection; microRNA; β-Aminopropionitrile monofumarate

Part II. DNA Methylation Alternation in Stanford-A Acute Aortic Dissection

Objectives: Acute aortic dissection (AAD) is a severe cardiovascular emergency with high mortality and serious complications. Stanford-A AAD, which accounts for almost 75% of AAD cases, is the most frequent type of AAD characterized by sudden onset, rapid progression, and poor prognosis. DNA methylation is a major epigenetic mechanism, which can be mediated by DNA methyltransferases (DNMTs). The hypermethylation of DNA could reduce the transcription activity of gene promotor, which can lead to gene silencing. And sustained hypomethylation of DNA could lead to upregulation of gene expression. It was found in our previous study that the protein expression levels of DNMT1 and DNMT3b decreased significantly in AAD samples, which indicated the potential role of DNA methylation in AAD. Most recently, several studies had paid attention to the role of DNA methylation in AAD. However, the DNA methylation patterns in AAD remain not completely clear. A two-stage study including the discovery stage and replication stage was designed to explore the panorama of DNA methylation in aorta tissues of Stanford-A acute aortic dissection.

Methods: In discovery stage, DNA methylation patterns were screened using Infinium Human Methylation 450K BeadChip in the aorta tissues from 4 patients with Stanford-A AAD and 4 controls. Gene enrichment was analyzed by Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and gene ontology (GO). The candidate genes were selected based on the results of bioinformatics analyses. In replication stage, DNA methylation levels of candidate genes were validated by pyrosequencing in the replication cohort including 16 patients with AAD and 7 controls. Protein expression level of candidate gene was assessed by Western blot in human aorta tissues of two groups.

Results: In discovery stage, a total of 589 differentially methylated positions including 315 hypomethylated and 274 hypermethylated positions were found in AAD group. About 55.2% of the differentially methylated positions were in body region and 21.8% in 5'UTR. More hypomethylated probes than the hypermethylated probes were found in TSS200, 5'UTR, 1stExon, body and 3'UTR. A total of 197 differentially methylated genes were found associated with 589 differentially methylated positions, including 96 hypermethylated and 101 hypomethylated genes. KEGG analysis demonstrated that hypermethylated genes were enriched in neuroactive ligand-receptor interaction, glutamatergic synapse, relaxin signaling pathway, phospholipase D signaling pathway and MAPK signaling pathway, et al. The hypomethylated genes were enriched in proteoglycans in cancer, TNF signaling pathway, apoptosis and MAPK signaling pathway, et al. Protein binding was the most significantly enriched GO term enriched by hypermethylated and hypomethylated genes respectively. Based on our bioinformatics analyses, a total of 5 CpG dinucleotides covering 5 genes including Angiopoietin -2 (ANGPT2), Fas Cell Surface Death Receptor (Fas), dual-specificity phosphatase 6 (DUSP6), ARH/RhoGEF And Pleckstrin Domain Protein 1 (FARP1), and Caspase recruitment domain protein 6 (CARD6) were selected for pyrosequencing validation in the replication stage. The results in replication cohort were consistent with those in discovery stage, validating the differential DNA methylation in AAD. The CpG site cg26478401 mapped to within 1500bp of the TSS of Fas was significantly hypomethylated in AAD group. In the previous studies, the protein expression of Fas was increased in aortic dissection and aortic aneurysm aorta tissues, which was validated increased by 1.78 times in AAD group in our research (1.00±0.34 vs.1.78±0.16, P＜0.05), suggesting the potential role of DNA methylation in regulating gene expression and the pathogenesis of AAD.

Conclusion: DNA methylation was markedly changed in the aorta tissues of Stanford-A AAD and associated with gene expression, involved in AAD pathogenesis.

Key words: DNA Methylation; Epigenetics; Stanford-A Acute Aortic Dissection; Vascular.