- 學位論文
探討氧化壓力基因對情緒相關表徵和社會心理特質之多效性
Investigating the pleiotropic effects of oxidative stress genes on mood related traits and psychosocial variables
摘要
氧化壓力為體內活性氧物質和抗氧化劑之間無法達成平衡時所產生的現象，此現象與多種不良健康狀況有關，包含癌症、神經退化性疾病、心血管疾病等。過去研究表明，在情緒疾患的患者中觀察到大腦存在氧化壓力的現象，因此氧化壓力在情緒疾患的病理機制中越來越受到關注，而情緒疾患患者通常伴隨著情緒變化、焦慮、睡眠困擾、自殺行為等症狀，由於情緒疾患的多種臨床症狀，診斷存在高度異質性。除了臨床症狀之外，社會心理特質也會增加情緒疾患的風險。然而並非所有過去的研究都顯示氧化壓力基因多型性與情緒疾患之間存在顯著關聯，且鮮少有研究檢查它們對與情緒疾患相關的異質性表現的影響。因此本研究目的為探討氧化壓力基因和情緒相關表徵與社會心理特質的相關性，以更好了解哪些特定表徵可能參與氧化壓力路徑，另外我們也評估氧化壓力基因對歐洲和亞洲人群的影響是否有差異。 本研究使用英國人體生物資料庫，總共納入425,214位受試者來自歐洲族群，平均年齡56.81歲(標準差=7.97)，其中53.90%為女性。表型由5個類別組成，總共包含183個表徵，當中，情緒類別包含82個表徵，焦慮類別有37個表徵，睡眠類別7個表徵，自殺類別19個表徵，以及社會心理類別38個表徵。在基因數據中，總共有10,882個單核苷酸變異(SNP)予以保留在往後的分析，這些SNP位在40個氧化壓力基因。我們使用累加遺傳模型執行羅吉斯和線性回歸模型，模型皆校正性別、年齡、第一至第十主成分，以確定哪些表徵與氧化壓力基因相關。此外我們檢驗表型相關性以檢查和氧化壓力基因顯著相關的表徵之間是否存在高度相關。另一方面，我們使用英國人體生物資料庫中的亞洲人群 (N=6,675) 來評估相關性結果是否在不同族群中存在差異。使用R 版本4.0.2和PLINK 1.9進行統計分析。 我們確定了40個與氧化壓力相關的基因，且將其分為兩類。我們的結果顯示8個基因(GPX1, GPX5, PRDX1, CCS, SOD1, NOX4, NOXA1, NOS1)對於情緒相關表徵與社會心理特質具有顯著的影響。GPX1、GPX5、NOS1和NOXA1基因當中包含9個SNPs (rs11710434, rs111761769, rs6774721, rs1389529156, rs146918648, rs139410431, rs139582054, rs1875140, rs112535944)可能存在多效性。我們發現在183個表徵中有18個在Bonferroni校正後顯示出與氧化壓力基因顯著相關，而這些顯著的關聯都沒有在亞洲人群中複製。另一方面，我們觀察到與氧化壓力基因相關的表徵之間存在中低度相關。 我們找到的SNP中發現了潛在的調控功能，當中有六個是潛在的甲基化數量性狀基因座(mQTL)，可能會影響血液中的DNA甲基化，而其中五個也是可能的表現數量性狀基因座(eQTL)，與人腦或其他組織中的表現量相關。本研究提供了4個氧化壓力基因(包含9個SNP)對歐洲人群的情緒、焦慮、睡眠和社會心理特質的多效性的結果，此外我們發現氧化壓力基因對情緒疾患相關特徵的影響可能存在種族差異。我們的研究提供了情緒疾患的症狀表現異質性可能存在著共同氧化壓力路徑的新見解，未來值得進一步研究調查情緒疾患中和氧化壓力相關的生物學路徑。
並列摘要
Background Oxidative stress, imbalances between reactive oxygen species (ROS) and antioxidants, was associated with a wide range of adverse health effects, including cancers, neurodegenerative disorders, and cardiovascular diseases. Oxidative stress was commonly observed in patients with mood disorders, which was gaining attention in the pathogenesis of mood disorders. Moreover, mood disorders were also frequently accompanied by sleep disturbances, anxious symptoms and suicidal behaviors. Due to a variety of symptoms of mood disorders, there were present highly heterogeneity of diagnosis. In addition to clinical symptoms, psychosocial factors contributed to increasing the risk of mood disorders. Not all previous studies showed significant associations between oxidative stress gene polymorphisms and mood disorders, and few studies examined the effects of oxidative stress genes on the heterogeneous presentation in relation to mood disorders. We aimed to investigate the effects of oxidative stress genes on mood related traits and psychosocial variables to better understand which traits related to oxidative stress pathway. We also aimed to evaluate the effects of oxidative stress genes across European and Asian populations. Materials and Methods In the current study, we used the data from 425,214 individuals with European ethnic background who participated in the UK Biobank. The average age of participants was 56.81 years (SD=7.97) and 53.90% were female. The phenotypes consisted of 5 domains, containing 183 traits in total. Across 5 domains, there were 82 traits in mood domain, 37 traits in anxiety domain, 7 traits in sleep domain, 19 traits in suicide domain, 38 traits in psychosocial and well-being domain. In the genotypic data, we obtained 10,882 variants in 40 oxidative stress genes which were retained after quality control procedures. We performed logistic and linear regression models with additive genetic model to identify which traits were associated with oxidative stress genes. Furthermore, we performed phenotypic correlation to check if there were strong correlations among traits significantly correlated with oxidative stress genes. On the other hand, we used Asian population in UK Biobank (N=6,675) to evaluate whether associated results in the European samples could be replicated across ethnic populations. Statistical analyses were conducted using R version 4.0.2 and PLINK 1.9. Results We identified 40 genes linked to oxidative stress, which was classified into two categories. Our results showed the significant effects of 8 genes (GPX1, GPX5, PRDX1, CCS, SOD1, NOX4, NOXA1, NOS1) on mood related traits and psychosocial variables. Further, GPX1, GPX5, NOS1, and NOXA1 genes containing 9 SNPs (rs11710434, rs111761769, rs6774721, rs1389529156, rs146918648, rs139410431, rs139582054, rs1875140, rs112535944) had pleiotropic effects. We found that 18 out of 183 traits in the European samples showed significant associations with oxidative stress genes after Bonferroni correction. However, all significant associations in the European were not replicated in the Asian population. On the other hand, we observed moderate to low correlations among traits which were associated with oxidative stress genes. Conclusion We found potential regulatory functions of the identified markers. Six of the them were potential methylation quantitative trait loci (mQTL) and might affect DNA methylation in blood, and five of them were found to be expression quantitative trait loci (eQTL) associating with the expression in human brain or other tissues. In conclusion, this study provided the information of the pleiotropic effects of four oxidative stress genes (GPX1, GPX5, NOS1, NOXA1) containing 9 SNPs on specific mood, anxiety, sleep, and psychosocial variables in European population. Our study provided a new insight into the heterogeneous symptoms presentations of mood disorders sharing oxidative stress pathway. It warrants further study to investigate the biological pathways in relation to oxidative stress in mood disorders.