Molecular genetics of primary open-angle glaucoma and related ocular quantitative traits

Glaucoma is a leading cause of irreversible blindness worldwide. It is a group of optic neuropathies with characteristic optic disc cupping and visual field defects. Subtypes of glaucoma have different etiology, prevalence, genetic susceptibility, clinical manifestations and treatment strategies. Primary glaucoma in adults can be classified into primary open-angle glaucoma (POAG) and primary angle-closure glaucoma (PACG) based on the anterior chamber angle structure.POAG is a complex disease with multiple risk factors, such as age, intraocular pressure (IOP), high myopia, family history, and genetic factors. Understanding the genetic risk factors of POAG would improve our understanding of the disease pathways, which in return may provide new insights into novel therapeutic targets and modalities. Different methodologies have been applied to identify disease-causative and associated genes and variants for POAG, including linkage analysis, candidate gene association and re-sequencing analysis, genome-wide association study (GWAS) and next generation sequencing (NGS). So far, more than 50 candidate loci have been identified for POAG by linkage analyses and association studies. Among them, mutations in the myocilin (MYOC) gene has been found to be causative for the monogenic form of POAG. However, these genes each account for only a small proportion of POAG. Apart from the linkage loci, multiple associated loci for POAG have been identified by GWAS, such as CAV1-CAV2, TMCO11, CDKN2B-AS1, SIX1-SIX6, ABCA1, PMM2, AFAP1, GMDS, FNDC3B, TGFBR3, TXNRD2, ATXN2, and FOXC1. However, inconsistent association patterns (such as failed replication in some cohort studies and different effect directions) have been revealed for some of these genes, such as SIX1-SIX6 and CAV1-CAV2, across different study populations. Moreover, some of these genes were correlated with the endophenotypes of glaucoma, such as IOP (e.g., FOXC1) and retinal nerve fibre layer (RNFL) thickness (e.g., SIX1-SIX6), in adult subjects, while such correlations in the younger age groups, e.g., children, remained largely unknown.Therefore, in the work related to this thesis we investigated the associations of three major loci associated with POAG, i.e., SIX1-SIX6, CAV1-CAV2 and FOXC1, and explored the effects of these genes and other related genes on the ocular quantitative traits (OQTs) in children.We first investigated the association between the SIX1-SIX6 locus and POAG in Chinese and Japanese. Eleven haplotype-tagging single-nucleotide polymorphisms (SNPs) covering this locus were genotyped in 1027 POAG patients and 375 controls from a Hong Kong Chinese cohort by using TaqMan® SNP genotyping assays. Three SNPs that showed significant associations, including a novel tagging SNP rs12436579 and two reported GWAS SNPs (rs33912345 and rs10483727), were further tested in a Chinese cohort from Shantou (157 POAG and 731 controls) and a Japanese cohort from Osaka (256 POAG and 207 controls). All these three SNPs were significantly associated with POAG in the pooled samples from the three cohorts (P0.5), which might imply an ethnic difference in the SNP associations.Normal-tension glaucoma (NTG) is a subtype of POAG, in which the IOP of patients falls within the normal range (10-21 mm Hg). In our previous work on the CAV1-CAV2 locus, we had confirmed the association of SNP rs4236601 and high-tension POAG (HTG) in subjects from Hong Kong, Shantou and Osaka. In the thesis work, we further validated the impact of this locus on NTG. Two SNPs (rs4236601 and rs1052990) reported in previous POAG studies and 6 haplotype-tagging SNPs (rs6466579, rs7801950, rs3779512, rs3807989, rs3801994 and rs1049337) were analysed. The significant association of rs4236601 with NTG was confirmed in the Hong Kong and Shantou Chinese cohorts (Pmeta=0.0019, OR=4.55, I²=0). In contrast, rs4236601 was monomorphic in the Osaka cohort. The association of rs1052990 was insignificant in meta-analysis combining Chinese and Japanese cohorts (Pmeta=0.81, OR=1.05; I²=64%), likely due to the opposite effects between Chinese (OR=1.26) and Japanese (OR=0.69). There findings again suggested ethnic differences in the genetic associations of POAG.The third locus investigated was the FOXC1 locus, which was identified in previous GWAS to be associated with POAG and IOP. A SNP, rs2745572, located at the intergenic region 50kb upstream of FOXC1, was significantly associated with HTG (Pnon-adjusted=0.00067 and Padjusted=0.019) rather than NTG (Pnon-adjusted=0.81 and Padjusted=0.99) in the Hong Kong cohort. Such finding was consistent with previous findings on its POAG and IOP associations. This work also illustrated some differences in the genetic association patterns between HTG and NTG.Furthermore, we investigated the genetic associations in OQTs related to POAG, i.e., IOP, peripapillary RNFL (p-RNFL) thickness, central corneal thickness (CCT) and axial length (AL). Reported SNPs associated with OQTs were mainly from studies in adults. The phenotypes in adults have been influenced greater by environmental factors (e.g., aging, reading habits, education and diseases) than in children in terms of accumulated exposure time and degree. In order to determine the effects of the genetic factors in OQT development, children subjects are good candidates for genotype-phenotype association analysis to reduce the cumulative impacts from environment. In this thesis, we tested the SNPs selected from genetic studies of POAG in a total of 3,194 children aged 5-11 years. We found two SIX1-SIX6 SNPs (rs33912345 and rs10483727) to be associated with temporal-inferior p-RNFL thickness (P=2.69×10⁻⁴ and P=3.46×10⁻⁴). Another SIX1-SIX6 SNP rs12436579 was associated with nasal-inferior p-RNFL (P=0.002). Similarities and differences in these associations between children and adults were revealed.Apart from p-RNFL thickness, we also identified significant associations of WNT7B rs10453441 with AL (P=8.03×10⁻⁶), corneal curvature (P=1.17×10⁻¹⁴) and corneal thickness (P=0.0026) in children, being consistent with those in adults. A novel association between ZNRF3 rs12321 and corneal curvature was identified (P=0.0060). These findings altogether suggested that the genetic profiles of OQTs in children are partly different from adults, and have proven the merit of genetic association studies in children.In conclusion, we confirmed a set of SNPs associated with POAG, HTG, NTG and POAG-related OQTs (p-RNFL thickness, CCT and AL), which provided potential targets for further understanding of genetic associations in POAG. In addition, our results indicated subtype-specific and ethnic specific genetic associations in POAG.青光眼是全球範圍內導致不可逆失明的主要原因。它被定義為一組具有特徵性神經盤凹陷擴大和視野缺損的疾病。青光眼的亞型具有不同的病因、患病率、遺傳易感性、臨床表現和治療策略。 根據前房角解剖結構的不同，成人原發性青光眼可分為原發性開角型青光眼（POAG）和原發性閉角型青光眼（PACG）。POAG是一種複雜的疾病，具有多種風險因素，例如年齡眼壓（IOP）、高度近視、家族病史和遺傳因素。了解POAG的遺傳危險因素將增進我們對疾病途徑的了解，這可能為新的治療靶點和方式提供新的方向。目前，應用于鑑定POAG的致病基因和相關基因及變體的方法包括連鎖分析、候選基因關聯和重測序分析、全基因組關聯研究（GWAS）和下一代測序（NGS）。到目前為止，已有發現50多個與POAG相關的候選基因位點。其中，已在myocilin (MYOC), optineurin (OPTN) and WD repeat-domain 36 (WDR36) 基因中發現作為POAG單基因形式誘因的突變。但是，這些基因各自僅佔POAG風險的一小部分。除連鎖基因位點外，GWAS還發現了一組與POAG相關的位點，例如CAV1-CAV2，TMCO11，CDKN2B-AS1，SIX1-SIX6，ABCA1，PMM2，AFAP1，GMDS， FNDC3B，TGFBR3，TXNRD2，ATXN2和FOXC1。但是，在不同人群的研究中，某些位點（例如SIX1-SIX6和CAV1-CAV2）存在不一致的關聯模式。此外，在成年受試者中，有一些基因與青光眼的內表型相關，例如FOXC1和IOP、SIX1-SIX6和视盘周围視網膜神經纖維層（p-RNFL）厚度。而這種關聯在低年齡組中（例如孩子）是很大程度上仍存於未知。因此，在此論文報告的工作中，我們已調查了與POAG相關的三個主要位點SIX1-SIX6、CAV1-CAV2和FOXC1，並探討了這些基因和其他相關基因對眼部定量特徵的影響（ OQTs）。我們首先在中國和日本的人群中進行了SIX1-SIX6位點與POAG風險的關聯研究。使用TaqMan®SNP基因分型分析方法，對來自香港中國隊列的1027名POAG患者和375名對照進行11個單倍型標籤單核苷酸多態性（SNP）分型分析。在香港隊列中顯示出顯著關聯的三個SNP，包括一個新標記rs12436579和兩個先前來自GWAS的SNP（rs33912345和rs10483727），在汕頭中國隊列（157名POAG患者和731名對照）和來自大阪的日本隊列（256名患者和207名對照）中進行了進一步測試。在三個隊列的所有樣本中，這三個SNP均與POAG顯著相關（P0.5），這可能暗示SNP關聯中存在種族差異。正常血壓青光眼（NTG）是POAG的一種亞型，患者的IOP處於正常範圍（10-21 mmHg）。在我們先前對CAV1-CAV2位點的研究中，已經證實SNP rs4236601與香港中國人，汕頭中國人和大阪日本人的高壓POAG（HTG）相關。在現階段研究中，我們進一步驗證了該位點對NTG的影響。分析了先前HTG研究中報告的兩個SNP（rs4236601和rs1052990）和6個單倍型標記SNP（rs6466579，rs7801950，rs3779512，rs3807989，rs3801994和rs1049337）。 rs4236601與NTG的顯著相關性已在香港和汕頭中國人群中得到證實（Pmeta=0.0019，OR=4.55，I²=0）。相反，rs4236601在大阪隊列中是單態的。 rs1052990的相關性在結合中國和日本隊列的薈萃分析並不顯著（Pmeta=0.81，OR=1.05； I²=64％），而作用方向在中國人群（OR=1.26）和日本人群（OR = 0.69）之間相反。研究結果再次表明了POAG基因關聯中存在種族差異。在本研究中的第三個位點位於FOXC1附近。該位點在先前的GWAS中被發現與POAG和IOP相關。 SNP rs2745572位於FOXC1上游50kb的基因間區域，與香港人群中的HTG（校正前P=0.00067，校正后P=0.019）存在顯著相關，而不是NTG（校正前P=0.81，校正后P= 0.99） 。 這一發現與先前關於其POAG和IOP關聯研究中的結果相似。 這項工作還說明了HTG和NTG之間的遺傳關聯模式存在一些差異。此外，我們調查了POAG相關OQT的遺傳關聯，即IOP，p-RNFL厚度，中央角膜厚度（CCT）和軸長（AL）。已知的OQT相關SNP主要來自成年隊列。就累積的暴露時間和程度而言，成人的表型受環境因素（例如，年齡，閱讀習慣，教育和疾病）的影響比兒童更大。為了確定遺傳因素在OQT發育中的作用，應進行兒童受試者的基因型-表型關聯分析，以減少環境的累積影響。在本文中，我們測試了從POAG遺傳研究中選擇的SNP，共對3194名5-11歲的小學生進行了測試。我們發現兩個SIX1-SIX6 SNP（rs33912345和rs10483727）與顳下p-RNFL厚度（P=2.69×10⁻⁴和P=3.46×10⁻⁴）相關。另一個SIX1-SIX6 SNP，即rs12436579，與鼻下p-RNFL相關（P=0.002）。揭示了兒童與成人在這些關聯上的異同。除p-RNFL厚度外，我們還發現WNT7B rs10453441與兒童的AL（P=8.03×10⁻⁶），角膜曲率（P=1.17×10⁻¹⁴）和角膜厚度（P=0.0026）有顯著相關性。這關聯與成年人一致。 同時， ZNRF3 基因中的rs12321和角膜曲率之間的新型關聯也被發現（P=0.0060）。 這些發現完全表明，兒童OQTs的遺傳特徵與成年人有部分不同，並且已經證明了兒童進行遺傳關聯研究的優點。總之，我們確認了一組與POAG、HTG、NTG和POAG相關OQT（即p-RNFL厚度、CCT和AL）相關的SNP，為進一步了解POAG中的遺傳關聯提供了潛在的靶標。 此外，我們的結果還表明了POAG中與遺傳相關的亞型特異性和種族特異性。Lu, Shiyao.Ph.D. Chinese University of Hong Kong 2020.Includes bibliographical references (leaves 221-277).Abstracts also in Chinese.Title from PDF title page (viewed on October 14, 2022)