GWAS on retinal vasculometry phenotypes

Xiaofan Jiang; Pirro G Hysi; Anthony P Khawaja; Omar A Mahroo; Zihe Xu; Christopher J Hammond; Paul J Foster; Roshan A Welikala; Sarah A Barman; Peter H Whincup; Alicja R Rudnicka; Christopher G Owen; David P Strachan; The UK Biobank Eye and Vision Consortium

doi:10.1371/journal.pgen.1010583

. 2023 Feb 9;19(2):e1010583. doi: 10.1371/journal.pgen.1010583

GWAS on retinal vasculometry phenotypes

Xiaofan Jiang ^1,², Pirro G Hysi ^2,^3,⁴, Anthony P Khawaja ⁵, Omar A Mahroo ^1,⁴, Zihe Xu ⁴, Christopher J Hammond ^3,^4,⁶, Paul J Foster ⁵, Roshan A Welikala ⁷, Sarah A Barman ⁷, Peter H Whincup ⁸, Alicja R Rudnicka ⁸, Christopher G Owen ^8,^*, David P Strachan ⁸; The UK Biobank Eye and Vision Consortium^¶

Editor: Sudha K Iyengar⁹

¹UCL Institute of Ophthalmology, London, United Kingdom

²UCL Great Ormond Street Institute of Child Health, University College London, London, United Kingdom

³Department of Twin Research and Genetic Epidemiology, King’s College London, London, United Kingdom

⁴Section of Ophthalmology, School of Life Course Sciences, King’s College London, London, United Kingdom

⁵NIHR Biomedical Research Centre, Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, London, United Kingdom

⁶Department of Medical Biochemistry, Oslo University Hospital, Oslo, Norway

⁷Faculty of Science, Engineering and Computing, Kingston University, Penrhyn Road, Kingston upon Thames, Surrey, United Kingdom

⁸Population Health Research Institute, St George’s, University of London, London, United Kingdom

⁹Case Western Reserve University, UNITED STATES

We have read the journal’s policy and the authors of this manuscript have the following competing interests: APK has consulted for the following companies: Abbvie, Aerie, Google Health, Novartis, Reichert, Santen and Thea. CJH has consulted for Nevakar Inc.

¶ Membership of UK Biobank Eye and Vision Consortium is provided in Supporting Information file S1_Acknowledgments.

^✉

* E-mail: cowen@sgul.ac.uk

Roles

Xiaofan Jiang: Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing

Pirro G Hysi: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

Anthony P Khawaja: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Validation, Visualization, Writing – review & editing

Omar A Mahroo: Investigation, Methodology, Validation, Visualization, Writing – review & editing

Zihe Xu: Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – review & editing

Christopher J Hammond: Funding acquisition, Investigation, Methodology, Resources, Validation, Visualization, Writing – review & editing

Paul J Foster: Data curation, Funding acquisition, Investigation, Methodology, Resources, Validation, Visualization, Writing – review & editing

Roshan A Welikala: Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Validation, Visualization, Writing – review & editing

Sarah A Barman: Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing – review & editing

Peter H Whincup: Funding acquisition, Investigation, Methodology, Validation, Visualization, Writing – review & editing

Alicja R Rudnicka: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Supervision, Validation, Visualization, Writing – review & editing

Christopher G Owen: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

David P Strachan: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

Sudha K Iyengar: Editor

PMCID: PMC9910644 PMID: 36757925

Abstract

The eye is the window through which light is transmitted and visual sensory signalling originates. It is also a window through which elements of the cardiovascular and nervous systems can be directly inspected, using ophthalmoscopy or retinal imaging. Measurements of ocular parameters may therefore offer important information on the physiology and homeostasis of these two important systems. Here we report the results of a genetic characterisation of retinal vasculature. Four genome-wide association studies performed on different aspects of retinal vasculometry phenotypes, such as arteriolar and venular tortuosity and width, found significant similarities between retinal vascular characteristics and cardiometabolic health. Our analyses identified 119 different regions of association with traits of retinal vasculature, including 89 loci associated arteriolar tortuosity, the strongest of which was rs35131825 (p = 2.00×10⁻¹⁰⁸), 2 loci with arteriolar width (rs12969347, p = 3.30×10⁻⁰⁹ and rs5442, p = 1.9E-15), 17 other loci associated with venular tortuosity and 11 novel associations with venular width. Our causal inference analyses also found that factors linked to arteriolar tortuosity cause elevated diastolic blood pressure and not vice versa.

Author summary

Vessels at the back of the eye (the “retina”) can be imaged easily. This paper reports on the largest genetic study of retinal vessel shape and size characteristics so far undertaken, to the best of our knowledge. Our study is novel in using an automated artificial intelligence imaging approach to distinguish between arteries and veins, and in demonstrating more genetic associations with vessel characteristics than any previous study (119 genetic loci in all). We also show that the tortuosity of retinal arteries is the most strongly genetically determined vessel characteristic (replicated remarkable well in a separate second large dataset). In addition, using a particular type of genetic analysis (so called “Mendelian Randomization”) we show for the first time that the tortuosity of arteries in the retina is causally related to elevated diastolic blood pressure and not the other way around. This is important as it provides unique insights into the mechanism of elevated blood pressure and hypertension, providing pointers to novel therapeutic targets for future treatment.

Introduction

The retina is one of the most metabolically active tissues of the human body [1] whose physiological functions require a steady supply of oxygen and nutrients and prompt removal of metabolic waste products. The retina is supplied by branches of the ophthalmic artery, which originates from the internal carotid artery that splits into a central and several ciliary arteries, which penetrate the globe to supply the inner and outer portions of the retina. Retinal vessel calibre, tortuosity and other morphological features vary as a result of constitutive, heritable factors [2], including genes participating in angiogenesis [3, 4], but also as a function of the level of metabolic activity in the retina [5, 6], and systemic haemodynamic changes like blood pressure [7, 8] and blood viscosity [9].

Retinal vessel morphology is an important marker of cardiometabolic [10–12] and eye health [13, 14]. Previously published works have investigated the influence of genetic factors over retinal vessel calibre and tortuosity [3, 4], but, as for many other complex traits, the heritability explained by the identified genes is modest, the underlying mechanisms underlying retinal vessel morphological variations remain poorly characterized, and the specific systemic impairments that each single feature of retinal vasculature reflects, are still poorly understood.

Genome-wide association analyses provide a tool to discover unsuspected common variants having potential risk for complex diseases [15]. There were previously reported genome-wide association studies (GWAS) results on retinal vasculometry characteristics, including venular width (not arteriolar width) [3] and more recently arteriolar and venular tortuosity [4], but sample sizes were small, the number of loci limited (4 at best) [3], and findings from discovery data-sets were not fully replicated [3].

A recent paper has examined the GWAS of retinal vessel density and complexity in a large study of European participants from the UK Biobank identified several loci, but did not distinguish between arterioles and venules, which may have different genetic determinants [16]. Hence, we aim to improve our knowledge of the genetic basis of retinal vasculometry phenotypes in the UK Biobank cohort, for arterioles and venules separately, and the mechanisms that determine their shape and tortuosity, to better understand relationships between the ocular fundus vascular phenotype and systemic conditions.

Methods

Ethics statement

UK Biobank was conducted with approval from the North-West Multi-Centre Research Ethics Committee (11/NW/0382), in accordance with the principles of the Declaration of Helsinki and the Research Governance Framework for Health and Social Care. All participants gave written informed consent.

Participants

The UK Biobank is a large multi-site cohort study of UK residents ages 40 to 69 years who were registered with the National Health Service (NHS) and living up to 25 miles from a study centre. A baseline questionnaire, measurements, and biological samples were undertaken in 22 assessment centres across the UK between 2006 and 2010. All UK Biobank genotypes were obtained as described elsewhere [17].

Phenotyping and retinal imaging

Ophthalmic assessment was not part of the original baseline assessment and was introduced as an enhancement in 2009 for 6 assessment centres which were spread across the UK (Liverpool and Sheffield in North England, Birmingham in the Midlands, Swansea in Wales, and Croydon and Hounslow in Greater London). Participants completed a touch-screen self-administered questionnaire. Retinal vessel information was extracted from digital fundus photography and spectral domain OCT images. An automated system (the QUARTZ system) can distinguish anatomical features from retinal images, including the optic disc, venules, arterioles and vessel segments, and out-puts centreline coordinates, and measures vessel width and tortuosity [18–20]. QUARTZ measures were summarized using mean width in microns and tortuosity (arbitrary units) [21] weighted by the length of the vessel segment, for arterioles and venules separately for each image, averaged across both eyes to give a person-level mean for analyses. Such an approach was considered appropriate given the pervasive examination of systemic traits.

Image processing modules were all validated on a subset of 4,692 retinal images from a random sample of 2346 UK Biobank participants: modules included vessel segmentation, image quality score, optic disc detection, vessel width measurement, tortuosity measurement, arteriolar venular recognition [18–20]. The performance of the Arteriole/Venule (A/V) recognition had detection rates of up to 96% for arterioles and 98% for venules when the automated probability of arteriole or venule was set to a cut-off of 0.8. An automated assessment of image quality was also made based on the segmented vasculature [18]. The algorithm achieved a sensitivity of 95% and a specificity of 94% for the detection of inadequate images. A model eye was used to quantify the magnification characteristics of the Topcon 3D OCT-1000 Mk 2 fundus camera, allowing pixel dimensions of vessel diameter to be converted to real size [22].

Statistical analyses

GWAS. Details about the UK Biobank study including microarray genotyping and SNP imputation have been described previously [17]. The UK Biobank team performed imputation from a combined panel consisting of the Haplotype Reference Consortium [23] and UK10K reference panels. Phasing on the autosomes was carried out using a modified version of the SHAPEIT2 [24] program.

Ethnicity-stratified GWAS were performed on each of the different retinal vessel phenotypic traits. For this study, we included 52,798 research individuals of European ancestry participating in the UK Biobank. For comparative purposes, additional sets of 933 individuals of South Asian and 1,288 of African ancestry were available and analysed separately (Fig 1). These analyses were underpowered and were used for comparison purposes only, and not for any post-GWAS analyses. Mixed linear regressions, with the retinal vasculometry phenotypes as the dependent variable, the allelic dosage at each locus as the independent predictors, adjusted for age, sex, spherical equivalent and the first ten principal components were conducted using the Bolt-LMM software [25] in the European ancestry subgroup. PLINK [26] was used, after the removal of related subjects, for the analyses of non-European samples. The GWAS significance threshold was set at the customary 5×10⁻⁰⁸. Although we conducted four separate GWAS, in part because of the correlation between traits, we did not correct for multiple testing.

Associations are reported for “genomic regions”, which we defined as contiguous regions of markers associated at GWAS-level statistical significance separated by less than 1 million base-pairs from each-other. The Online Mendelian Inheritance In Man dataset (www.omim.org) was used to obtain information about diseases caused by rare mutation in any of the genes near our association peaks.

Replication GWAS—EPIC-Norfolk

The European Prospective Investigation into Cancer (EPIC) study is a pan-European prospective cohort study designed to investigate the aetiology of major chronic diseases [27]. EPIC-Norfolk, one of the UK arms of EPIC, recruited and examined 25,639 participants between 1993 and 1997 for the baseline examination [28]. The EPIC-Norfolk study was carried out following the principles of the Declaration of Helsinki and the Research Governance Framework for Health and Social Care. The study was approved by the Norfolk Local Research Ethics Committee (05/Q0101/191) and East Norfolk and Waveney NHS Research Governance Committee (2005EC07L). Recruitment was via general practices in the city of Norwich and the surrounding small towns and rural areas, and methods have been described in detail previously [29]. Since virtually all residents in the UK are registered with a general practitioner through the National Health Service, general practice lists serve as population registers. All participants gave written, informed consent. Ophthalmic assessment formed part of the third health examination and this has been termed the EPIC-Norfolk Eye Study [30]. In total, 8,623 participants were seen for the Eye Study between 2004 and 2011. Although de-anonymising UK Biobank participants or linking them with external datasets is not allowed, we expect no overlap between discovery and replication datasets, because the EPIC-Norfolk study participants were recruited in an area geographically distinct from where the UK Biobank participants were recruited. Digital photographs of the optic disc and macula were taken using a TRC-NW6S non-mydriatic retinal camera and IMAGEnet Telemedicine System (Topcon Corporation, Tokyo, Japan) with a 10-megapixel Nikon D80 camera (Nikon Corporation, Tokyo, Japan). Pupils were not dilated. All participants gave written, informed consent.

Genotyping was carried out using the Affymetrix UK Biobank Axiom Array (the same array as used in UK Biobank). SNP exclusion criteria included: call rate < 95%, abnormal cluster pattern on visual inspection, plate batch effect evident by significant variation in minor allele frequency, and/or Hardy-Weinberg equilibrium P < 10–7. Sample exclusion criteria included: DishQC < 0.82 (poor fluorescence signal contrast), sex discordance, sample call rate < 97%, heterozygosity outliers (calculated separately for SNPs with minor allele frequency >1% and <1%), rare allele count outlier, and impossible identity-by-descent values. We removed individuals with relatedness corresponding to third-degree relatives or closer across all genotyped participants. 99.7% of EPIC-Norfolk are of European descent and we excluded non-White participants. Imputation was carried out using the Sanger imputation service (https://imputation.sanger.ac.uk) with reference to the Haplotype Reference Consortium panel, version 1.

Linkage Disequilibrium (LD) score regression analyses

Bivariate regression intercepts were calculated to distinguish between polygenetic effects and effects of population stratification within and between GWAS. Inter-trait genetic correlation on the summary statistics from GWASs was performed using LD Hub [31] and the LD score regression program for phenotypes that were not available through the hub [32]. All the parameters used for the LDSC analyses were following the authors’ recommendations [31] for both the analyses using the LD Hub GWAS information and our own.

Tissue enrichment across multiple tissues

Multi-tissue expression enrichment analyses were performed in order to identify the expression levels in particular tissues using LD-score regression (LDSC) based procedures described elsewhere [33]. The GWAS results were compared to gene expression data assembled and made available by the authors of the LDSC program (https://storage.googleapis.com/broad-alkesgroup-public/LDSCORE/LDSC_SEG_ldscores/Multi_tissue_gene_expr_1000Gv3_ldscores.tgz).

Transcriptome-based association analyses

We applied whole-genome transcriptomic prediction models trained with reference data from version 8 of the Genotype-Tissue Expression (GTEx) project [34] to infer transcriptome variation in the GWAS study and then compute transcriptome-to-trait associations using GWAS associations’ summary statistics as input. We used the S-PrediXcan programme [35, 36] to estimate genetically regulated gene expression using whole-genome tissue-dependent prediction models, trained on GTEx version 8 reference transcriptome data and using GWAS-generated information to infer gene-based associations with retinal vasculature phenotypes. Based on the GTEx analysis described above, we applied this framework to 49 GTEx tissues and 4 retinal vasculometry traits.

Integration of genotypic and expression data

Summary-based Mendelian Randomization (SMR) integrates summary-level GWAS data with expression quantitative trait loci (eQTL) studies, to identify pleiotropic associations between a complex trait and a specific gene [37].

Due to availability constraints and the need to minimise the number of tests, the eQTL-based analyses were performed in selected tissues only. We selected tissues where the original sample size in the GTEx datasets was comparatively large. To maximise power, we conducted these analyses in subcutaneous adipose (N = 581), tibial artery (N = 584, chosen a priori in preference to coronary artery due to increased sample size), and a less specific dataset of less-tissue specific leukocytes for which larger sample sizes (N = 5,311) were available [38]. Tests based on cis-methylation information (cis-mQTL), were carried out in brain tissues [39].

Mendelian Randomisation (MR)

The MR-base package [40] was used for two sample MR analyses testing for causality of vascular changes over metabolic traits. MR-base uses instruments SNPs selected on the basis of several hundreds of GWAS summary data sets in its repository and estimates the causal impact of specific traits (exposures) on retinal vasculature phenotype outcomes. These SNPs are all significantly associated with the “exposure” trait and they are meant to be in linkage equilibrium with each other, even if they are located in relative proximity of each other. In all cases, our MR cases were two-sample; the genetic effects were obtained from our GWAS on the four retinal vasculometry traits of interest, and they were compared with effect estimates on other traits from samples that did not include any UK Biobank subjects. Studies in which the UK Biobank had contributed information were specifically removed from the analyses reported here. We are reporting results from different MR tests: the inverse variance weighted, weighted median and MR-Egger tests (both MR and intercept). The results are primarily reported with the reference of the random-effect inverse variance weighted test, but the other tests may be valuable to interpret the relationship between exposure and outcome. In particular, the MR-Egger intercept tests for unbalanced (unidirectional) horizontal pleiotropy, which is usually taken as evidence against causation.

Results

Association results

We performed four separate GWAS of arteriolar tortuosity (AT), arteriolar width (AW), venular tortuosity (VT) and venular width (VW) in a total of 52,798 participants of European ancestry from the UK Biobank. The GWAS results were all in line with polygenic architecture for all these traits, without any sign of inflation due to uncontrolled population structure (LDSC intercept in the range 1.01–1.04, Table A in S1 Table). The four retinal vasculature traits examined were genetically correlated with each-other, with the highest correlation (r_g = 0.44) observed between VW and AW (Table B in S1 Table), suggesting that these parameters estimate complementary aspects of the retinal vasculature.

The trait for which SNPs explained the largest proportion of variance (SNP heritability) was AT (h²_SNP = 0.51). We also observed some of the most significant associations between this trait and polymorphic variants across the genome. A total of 14,021 SNPs were associated at genome-wide level with AT, and they clustered around 89 different unique genomic regions (Table C and Fig A i-iv in S1 Text). The most statistically significant association was observed for polymorphic variants within the genomic sequence of the COL4A2 gene (p = 2.0x10^-108 for rs35131825). Other novel significant associations with this trait were observed for polymorphisms overlapping with the genomic sequences of genes involved in cardiovascular function, such as those within the PDE3A gene (p = 1.30x10^-64 for rs11045245). The statistically strongest association for VW was found in the region located on chromosome 6q24 (p = 1.40x10^-14 for rs12206319), within the NMBR gene. The strongest association for AW was identified with a SNP (rs5442, p = 1.9x10^-15) within GNB3 gene [41]. The strongest significant association for VT was found for a SNP (rs1136956, p = 6.0x10^-58) within the protein coding region of the ACTN4 gene. Interestingly, this region was also significantly associated with all retinal vasculometry phenotypic traits that we analysed, except for AW.

In addition to the genetic loci above, which have been previously related to vascular morphology features, we observed unique associations that have not been previously associated with vascular, or any other phenotypic trait. For example, strong association with AT was found for rs1950127 (p = 1.80×10⁻⁶⁸), located within the coding sequence of LOC101928978, a transcript of unknown function. Novel association with AW was found for one gene locus (CLUL1, p = 3.30×10⁻⁰⁹ for rs12969347). In addition, a novel association with VT was found, among others, at a locus overlapping with the genomic sequence of the DLX6 gene (p = 4.30 × 10⁻²³ for rs2948244), a transcriptor gene linked to forebrain and craniofacial development.

Replication and transethnic comparison

We sought to replicate these findings in a fully independent population sample. For this purpose, we used results obtained from the EPIC-Norfolk study, a population-based cohort whose general ethnic and demographic characteristics as well as ophthalmic assessment modalities closely matched those of the UK Biobank participants. Despite the considerably smaller sample size replication was robust for all traits; for example, SNPs associated with arterial tortuosity in the UK Biobank were associated at a Bonferroni (p < 0.05/73 available SNPs) over 11 loci (Table D in S1 Table) and there was generally a linear relationship between effect sizes estimated in the discovery and replication cohorts (Fig B in S1 Text).

We subsequently meta-analysed data from both cohorts and, as expected, we obtained associations not previously seen in the analyses of the UK Biobank data only (Table E in S1 Table). However, given that we were unable to provide further replication to these new results, in the following sections of the manuscript, we will continue to refer to results obtained from the UK Biobank only.

Next, we sought to investigate the variation in the amplitude of the association signals across the different available ancestral groups (Table F in S1 Table), by comparing results from the European panel with association data from 1,288 African and 933 South Asian UK Biobank participants. Due to sample size differentials, p-value based comparisons were underpowered. However, we observed considerable correlations between effect sizes observed in subjects of European and African (Pearson’s r = 0.33, 95% CI [0.12, 0.51], p = 0.02]) and South Asian ancestry (r = 0.46, 95% CI [0.27, 0.61], p<0.00001). Fig C in S1 Text shows the correlated relationship between different ancestry.

Tissue expression enrichment and exploration of functional mechanisms in vasculometry regulation

Analyses of eQTL offer insights on the genetic architecture of expression regulation. These analyses also help to annotate the exact genes among the many located within associated regions whose variations are at the origin of the association signals. Many of the GWAS associations were expressed across a wide range of tissues (Table G in S1 Table). Due to considerable tissue variability of expression patterns, we first sought to identify the tissues where our GWAS results observed in European participants were most enriched. A multi-tissue enrichment analysis for traits [33] found that, among available GTEx tissues/cell types, genes associated with retinal vessel parameters in our four GWAS analyses showed statistically significant enrichment in adipose tissues (Fig D in S1 Text), but also in artery, venous tissues, as well as smooth muscle-rich tissues such as oesophagus and myometrium tissues. Genes associated with VW diverged most from the pattern observed for the other vasculometry traits and were mostly expressed in kidney and pancreas, although the results of these analyses were not statistically significant after multiple testing correction.

We subsequently tested the hypotheses that many of our GWAS findings were exerting their effects over the phenotypes through the eQTL effects affecting the transcription of the genes. Using tissues in which we observed the highest enrichment of GWAS results (subcutaneous adipose and tibial artery) as well as white blood cells due to the abundance of information available for them, we conducted summary-based Mendelian randomization tests for loci associated with AT, which was the GWAS analysis for which we obtained most results and therefore had most power (Fig E in S1 Text; Table A in S1 Table shows AT heritability value 0.512). For the other retinal vasculature traits, the SNP instruments were of low power due to relatively few associations, meaning non-significant results in MR causality analyses. We found statistically significant (p_HEIDI ≥ 0.05) evidence that many SNPs associated with AT causally influences this phenotype through the control of the transcription process of several genes, such as the SNPs in the ACTN4 locus (Table H in S1 Table), which was associated with three out of the four retinal vessel phenotypes we analysed in our dataset.

Relationship of vasculometry and other phenotypes

To establish a relationship between the retinal vessel phenotypes that we analysed, with other human phenotypic traits or disease, we calculated the genome-wide correlation between effect sizes observed in our GWAS with those estimated for other diseases. These analyses suggest that our phenotypes share significant proportions of their genetic risk with a number of cardiometabolic traits and disease (Table I in S1 Table). Statistically significant genetic effects’ correlations were observed, notably between retinal vessel parameters and metabolic phenotypes such as the systolic and diastolic blood pressure, cardiovascular disease, fat mass, but also standing body height, etc.

We consider retinal vasculometry parameters as biomarkers of vascular changes elsewhere in the body. We further explored causality by performing two-sample MR analyses of potentially relevant traits that were significantly correlated with the retinal vasculometry phenotypes that we analysed (Table J in S1 Table). For these purposes, we used data that were available through the MR-base platform. We specifically tested the hypotheses that processes associated with retinal vasculometry variability can lead to other diseases. We found statistically significant evidence that processes underlying arteriolar tortuosity in the retina cause elevated diastolic blood pressure (p_IVW = 3.5 × 10⁻⁰⁵, Fig 2). Our analyses also suggested that changes in arteriolar tortuosity may also causally affect risk of coronary heart disease.

Fig 2 — Footnotes: Among 108 SNPs that showed independent genome-wide significant associations with AT in conditional regression analysis, 99 had published information available on their association with DBP in non-UK-Biobank studies [reference: decipher "id:ieu-b-39" and remove from y-axis label]. Single points in the graph represent co-ordinates determined by the effect of each specific SNP on DBP (in non-Biobank studies) and AT (in UK Biobank). Outlying SNPs are identified individually by rs-number. [see comment below.]. MR analyses were conducted with all 99 SNPs as instrumental variables using the conventional inverse-variance-weighted (IVW) and Egger methods. The slope of each line corresponds to the estimated MR effect for each method. Significance tests for the MR slope (Table I in S1 Table) were: p = 0.00004 (3.55E-05) by MR-IVW. p = 0.0050 (0.005017737) by MR-Egger. Using a weighted-median MR method (which is less sensitive to outliers), the corresponding p-value was p = 0.00001 (1.01E-05).

We also tested the reverse model that these phenotypes causally lead to changes in arterial tortuosity, using the limited analyses on whose traits whose summary statistics are publicly accessible (Table K in S1 Table). We did not find any evidence of reverse causation, which further reinforces the confidence in the original MR findings that identified AT as one of the causes for elevated diastolic blood pressure and a potential contributing cause to coronary heart disease.

Discussion

Based on the largest dataset of retinal vessel morphology to date, we identified 119 different loci associated with retinal vasculometry traits. We found the most loci (n = 89) associated with AT, and 2 novel loci associated with AW, 17 loci associated with VT, and 11 associations with VW. Our study used AI-enabled automated image processing techniques to derive these retinal vasculometry traits, which have been developed, validated internally and externally, and applied in a harmonised fashion across two large population-based cohorts, namely UK Biobank and EPIC-Norfolk cohorts [11, 12, 18, 20, 21]. Previous work using our AI-enabled automated retinal vessel image analysis system (QUARTZ) has described phenotypic associations of these four vasculometry indices (i.e., AW, AT, VW, VT) with conventional cardiometabolic risk factors including blood pressure (BP) [11], adiposity/BMI [12], DM and blood lipids [21]. Analysis in UK Biobank has demonstrated strong associations between VW and VT with markers of adiposity [12], and that AW, VW and AT all show strong inverse associations with blood pressure (systolic and diastolic) [11]. Hence, these microvascular characteristics could be the cause or consequence of important pathophysiological traits. We sought to clarify the direction of causality by examining their genetic determinants by instrumental variable (Mendelian randomization, MR) analysis.

Several loci have been identified by earlier, statistically less-powered studies with smaller sample sizes. A meta-analytic study of over 18,000 participants of European ancestry identified rs2194025 locus on chromosome 5q14 near the MEF2C gene as being associated with retinal arteriolar diameter [42], later updated to include over 24,000 multi-ethnic participants to find 4 loci associated with both central retinal vein and arterial width [41]. We were unable to replicate these findings, with our most significant associations for AW being on chromosomes 18 and 22, with completely different loci (Table E in S1 Table). More recent studies have focussed on examining genetic determinants of retinal vessel tortuosity. A study of just over 3000 individuals of European ancestry showed one locus associated with VT (ACTN4/CAPN12 genes), and another associated with AT (COL4A2 gene) [4]. A recent study using the UK Biobank data source in fewer participants, showed that retinal vessel density and fractal dimensions, as a measure of vessel complexity (extracted from the entire image after deep learning vessel segmentation without distinction between arterioles and venules) were associated with 7 and 13 novel loci respectively [16]. The fewer number of loci identified, suggesting a lower level of heritability, may reflect the indiscriminate identification of vessels (without being able to distinguish between arterioles and venules), and while the strong signal between overall vessel density with MEF2C was replicated by the association we observed specifically with venular width, the level of overlap with our vasculometry phenotypes was limited to 8 loci (Table L in S1 Table), and the strong signal with OCA2 was not observed.

In our study, the most statistically significant association with AT was observed for polymorphic variants within the genomic sequence of the COL4A2 gene (p = 2.0x10^-108 for rs35131825). Importantly, these associations were apparent in both the validation UK Biobank dataset (also shown in a sub-set of non-European ancestry, see Tables D and E in S1 Table), and the EPIC-Norfolk replication dataset with remarkable replication. The COL4A2 gene is implicated in angiogenesis and tumour growth suppression [43] and previous work has reported specific changes in its genomic sequence that cause familial and sporadic small vessel disease [44, 45]. Interestingly, the COL4A2 gene is located next to the COL4A1 gene on chromosome 13q34. COL4A1 is known to harbour mutations that cause familial retinal arteriolar tortuosity [46]. It is therefore unclear if COL4A1, COL4A2 or both can contribute to changes in the retinal vasculature. Other novel significant associations with this AT trait were observed for polymorphisms overlapping with the genomic sequences of genes involved in cardiovascular function, such as those within the PDE3A gene (p = 1.30x10^-64 for rs11045245), a gene previously associated with the size of the aortic root [47] and blood pressure [48], but to the best of our knowledge, not with retinal or other small vessel morphological features.

Significant associations with loci that have been previously implicated in angiogenesis and regulation of vasculature function was a feature of all four traits we examined. For example, the statistically strongest association for VW was found in the region located on chromosome 6q24 (p = 1.40x10^-14 for rs12206319), within the NMBR gene, which has previously been associated with retinal vascular calibre [3]. The strongest association for AW was identified for a SNP (rs5442, p = 1.9x10^-15) within the GNB3, a gene also associated previously with central retinal vein calibre [41], but also multiple ocular phenotypes such as refractive error [49], macular thickness [50] and corneal astigmatism [51]. The strongest significant association for VT was found for a SNP (rs1136956, p = 6.0x10^-58) within the protein coding region of the ACTN4 gene. The ACTN4 gene codes for an alpha actinin protein, a ubiquitous and multifunctional family involved in the cytoskeleton framework and cell-cell adhesion [52]. Interestingly, this region was also significantly associated with all retinal vasculometry phenotypic traits that we analysed, except for AW.

Another important novel question we sought to answer from this study was the relationship between retinal vasculometry and systemic metabolic traits. This is particularly relevant for BP, in examining whether microvascular changes (as evidenced in the retina in our study) precede or cause raised BP, or whether they result in hypertension. Genetic influences on AT were strongest with high heritability (h² value 0.5), and intermediate for VW and VT (with h² values ~0.2), offering scope for the MR approach to be used for all these three traits. The low heritability observed for AW deprived us of the opportunity to explore the systemic consequences of factors linked to AW variation. These differences suggest that these vasculometry traits might be differently genetically and phenotypically determined, with AT being by far the most heritable of the 4 traits, exhibiting associations with 89 unique genomic regions. Using Mendelian Randomization (MR), we provide strong statistical evidence (p = 3.5x10^-5) that processes leading to increased AT causally influence blood pressure.

To explore the direction of causality we took advantage of the availability of powerful "instrumental variables" drawn from among genetic variants strongly predisposing to traits of interest. However, the SNPs with the largest per-allele effects on AT had little or no effect on diastolic BP (Fig 2), suggesting that there may be multiple biological pathways represented among the 99 SNPs that were independently associated with AT and selected for our MR analysis, some not implicated in DBP. It should also be noted that although we find evidence to support a causal influence of (genetically determined) AT on BP (particularly diastolic), AT is not the most important trait observationally correlating with DBP (stronger phenotypic associations with DBP are observed for AW than for AT [11]). So, if the AT-DBP association is causal, it is clearly not the whole story as far as microvasculature and systemic blood pressure is concerned. It is noteworthy, that while a recent study concluded that polygenic risk scores for hypertension and T2 diabetes might cause retinal vessel changes (specifically vessel density and fractal dimensions), unlike our study there were too few specific genetic markers to confirm the direction of causality [16].

The major strengths of our study lie in the very large sample sizes, standardised and validated trait measures [11, 12, 18, 20, 21], comprehensive genomic data including imputations, replication of GWAS results, and interpretation of our results both biologically (by use of quantitative trait locus analyses) and epidemiologically (by use of MR approaches). Although our primary analyses focused on European ancestry populations, we considered other ancestry groups in sensitivity analyses. The main limitation we encountered was the lack of genetic associations with AW and the modest number of loci associated with VW and VT, all traits with a low SNP-based heritability, which limited the statistical power of MR analyses of these venular traits. Also, while the use of person level averages in vasculometry phenotypes was suitable to examine systemic traits, it was not possible to examine ocular outcomes that could have led to between-eye vessel asymmetry (e.g., Central Retinal Vein Occlusion), and potentially heritable conditions leading to ocular media opacity (such as corneal disease and cataract), which may have impacted on retinal image quality. However, this affect is likely to be modest given that only a portion of the vascular tree needed to be imaged (as opposed to the whole image), resulting in high inclusion rates (~80% in both UK Biobank and EPIC-Norfolk cohorts). Another potential issue is the need for adjustment for covariates, given that SNPs associated with unadjusted covariates could generate spurious signals. However, this is of practical importance only when there are strong correlations with both the dependent and independent variables. Theoretically, covariate-adjustment may reduce the residual variance of the outcome and thereby increases power to detect covariate-independent genetic associations. For this reason, we have run models for traits of interest with additional adjustment for age (and age-squared to examine presence of non-linearity), diabetes mellitus and current smoking status, which reassuringly appears to have little effect (Table M in S1 Table).

In conclusion, microvascular morphology, measured objectively and systematically in retinal images, shows distinct phenotypic traits of arterial and venular width and tortuosity, each with distinct patterns of genetic association. For arterial tortuosity, supportive evidence emerges for microvascular morphology as a contributory cause, rather than a consequence, of raised blood pressure, raising the possibility of novel treatment pathways.

Supporting information

S1 Acknowledgments. UK Biobank Eye and Vision Consortium.

(DOCX)

Click here for additional data file.^{(14.2KB, docx)}

S1 Text. Supplementary Figures.

Fig A. Manhattan plots depicting the genome-wide associations for arteriolar tortuosity (i), arteriolar width (ii), venular tortuosity (iii) and venular width (iv). Fig B. The relationship between effects sizes estimated in association with retinal vessel parameters in the UK Biobank (x-axes) and those estimated in the EPIC-Norfolk cohort (y-axes). Fig C. The relationship between effects sizes estimated in association with retinal vessel parameters in the UK Biobank (x-axes) and those estimated in the Africa/Asian ancestry (y-axes). Fig D. Tissue expression enrichment analysis results, for AT (i), AW (ii), VT (ii) and VW (iv). Fig E. SMR Integration of results combined GWAS and eQTL for AT on Peripheral blood.

(DOCX)

Click here for additional data file.^{(1.5MB, docx)}

S1 Table. Supplementary Tables.

Table A. GWAS quality control metrics. Table B: Genetic intercorrelations between the four retinal vasculature parameters. Table C. Most significant associations observed for the traits of interest. Table D. Replication of GWAS results in the EPIC-Norfolk cohort. Table E. Results of the meta-analyses of the UK Biobank and EPIC-Norfolk cohorts. Table F. Known loci replicated in trans-ancestry analysis combining Asian, African and European analysis. Table G. eQTL effects of SNPs associated in the GWAS. Table H. Functional analyses of causation for SNPs associated with AT. Table I: Ldsc genetic correlation between retinal parameters and systemic traits and disease. Table J. Mendelian randomisation analyses testing causality of vascular changes over metabolic traits. Table K. Mendelian randomisation analyses testing causality of metabolic traits over vascular changes. Table L. Replication of GWAS results in a recent publication (Zekavat BS et al., 2021). Table M. Comparison of part associations observed for the traits of interest with additional covariates (age2, smoking, diabetes).

(XLSX)

Click here for additional data file.^{(217.2KB, xlsx)}

Data Availability

Data are available from UK Biobank through an application process. The process is outlined at the following website: https://www.ukbiobank.ac.uk/enable-your-research/apply-for-access. Other data sources used include the 1000 Genomes Project (available from http://www.1000genomes.org/), PLINK (https://www.cog-genomics.org/plink2/), the Genotype-Tissue Expression (GTEx) project: (http://www.gtexportal.org/home/), RegulomeDB (http://www.regulomedb.org/), UK Biobank http://www.ukbiobank.ac.uk), MRbase (https://app.mrbase.org/).

Funding Statement

The retinal vasculometry work was supported by the Medical Research Council Population and Systems Medicine Board (MR/L02005X/1) and British Heart Foundation (PG/15/101/31889) (PJF,SAB,PHW,ARR,CGO,DPS). PJF has received additional support from the Richard Desmond Charitable Trust (via Fight for Sight) and the Department for Health through the award made by the National Institute for Health Research to Moorfields Eye Hospital and the UCL Institute of Ophthalmology for a Biomedical Research Centre. The views expressed in this article are those of the authors and not necessarily those of the Department for Health. APK is funded by a UK Research and Innovation Future Leaders Fellowship (MR/T040912/1) and an Alcon Young Investigator Award. XJ is funded by Moorfields Eye Charity. OAM is funded by the Wellcome Trust (206619/Z/17/Z). EPIC-Norfolk funding: The EPIC-Norfolk study (DOI 10.22025/2019.10.105.00004) has received funding from the Medical Research Council (MR/N003284/1 and MC-UU_12015/1) and Cancer Research UK (C864/A14136). The clinic for EPIC-Norfolk 3HC was funded by Research into Aging, now known as Age UK (Grant Ref: 262). The genetics work in the EPIC-Norfolk study was funded by the Medical Research Council (MC_PC_13048). We are grateful to all the participants who have been part of the project and to the many members of the study teams at the University of Cambridge who have enabled this research. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Wong-Riley MT. Energy metabolism of the visual system. Eye Brain. 2010;2:99–116. doi: 10.2147/EB.S9078 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Sun C, Zhu G, Wong TY, Hewitt AW, Ruddle JB, Hodgson L, et al. Quantitative genetic analysis of the retinal vascular caliber: the Australian Twins Eye Study. Hypertension. 2009;54(4):788–95. doi: 10.1161/HYPERTENSIONAHA.109.132902 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Ikram MK, Sim X, Jensen RA, Cotch MF, Hewitt AW, Ikram MA, et al. Four novel Loci (19q13, 6q24, 12q24, and 5q14) influence the microcirculation in vivo. PLoS Genet. 2010;6(10):e1001184. doi: 10.1371/journal.pgen.1001184 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Veluchamy A, Ballerini L, Vitart V, Schraut KE, Kirin M, Campbell H, et al. Novel Genetic Locus Influencing Retinal Venular Tortuosity Is Also Associated With Risk of Coronary Artery Disease. Arterioscler Thromb Vasc Biol. 2019;39(12):2542–52. doi: 10.1161/ATVBAHA.119.312552 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Joyal JS, Sun Y, Gantner ML, Shao Z, Evans LP, Saba N, et al. Retinal lipid and glucose metabolism dictates angiogenesis through the lipid sensor Ffar1. Nat Med. 2016;22(4):439–45. doi: 10.1038/nm.4059 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Sun Y, Lin Z, Liu CH, Gong Y, Liegl R, Fredrick TW, et al. Inflammatory signals from photoreceptor modulate pathological retinal angiogenesis via c-Fos. J Exp Med. 2017;214(6):1753–67. doi: 10.1084/jem.20161645 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Cheung CY, Zheng Y, Hsu W, Lee ML, Lau QP, Mitchell P, et al. Retinal vascular tortuosity, blood pressure, and cardiovascular risk factors. Ophthalmology. 2011;118(5):812–8. doi: 10.1016/j.ophtha.2010.08.045 [DOI] [PubMed] [Google Scholar]
8.Koch E, Rosenbaum D, Brolly A, Sahel JA, Chaumet-Riffaud P, Girerd X, et al. Morphometric analysis of small arteries in the human retina using adaptive optics imaging: relationship with blood pressure and focal vascular changes. J Hypertens. 2014;32(4):890–8. doi: 10.1097/HJH.0000000000000095 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Luo T, Gast TJ, Vermeer TJ, Burns SA. Retinal Vascular Branching in Healthy and Diabetic Subjects. Invest Ophthalmol Vis Sci. 2017;58(5):2685–94. doi: 10.1167/iovs.17-21653 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Ikram MK, Ong YT, Cheung CY, Wong TY. Retinal vascular caliber measurements: clinical significance, current knowledge and future perspectives. Ophthalmologica. 2013;229(3):125–36. doi: 10.1159/000342158 [DOI] [PubMed] [Google Scholar]
11.Tapp RJ, Owen CG, Barman SA, Welikala RA, Foster PJ, Whincup PH, et al. Associations of Retinal Microvascular Diameters and Tortuosity With Blood Pressure and Arterial Stiffness. Hypertension. 2019;74(6):1383–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Tapp RJ, Owen CG, Barman SA, Welikala RA, Foster PJ, Whincup PH, et al. Retinal Vascular Tortuosity and Diameter Associations with Adiposity and Components of Body Composition. Obesity. 2020;28(9):1750–60. doi: 10.1002/oby.22885 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Newman A, Andrew N, Casson R. Review of the association between retinal microvascular characteristics and eye disease. Clin Exp Ophthalmol. 2018;46(5):531–52. doi: 10.1111/ceo.13119 [DOI] [PubMed] [Google Scholar]
14.Rudnicka AR, Owen CG, Welikala RA, Barman SA, Whincup PH, Strachan DP, et al. Retinal Vasculometry Associations With Glaucoma: Findings From the European Prospective Investigation of Cancer-Norfolk Eye Study. Am J Ophthalmol. 2020;220:140–51. doi: 10.1016/j.ajo.2020.07.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Manolio TA, Brooks LD, Collins FS. A HapMap harvest of insights into the genetics of common disease. J Clin Invest. 2008;118(5):1590–605. doi: 10.1172/JCI34772 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Zekavat SM, Raghu VK, Trinder M, Ye Y, Koyama S, Honigberg MC, et al. Deep Learning of the Retina Enables Phenome- and Genome-Wide Analyses of the Microvasculature. Circulation. 2022;145(2):134–50. doi: 10.1161/CIRCULATIONAHA.121.057709 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK Biobank resource with deep phenotyping and genomic data. Nature. 2018;562(7726):203–9. doi: 10.1038/s41586-018-0579-z [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Welikala RA, Foster PJ, Whincup PH, Rudnicka AR, Owen CG, Strachan DP, et al. Automated arteriole and venule classification using deep learning for retinal images from the UK Biobank cohort. Computers in Biology and Medicine. 2017;90:23–32. doi: 10.1016/j.compbiomed.2017.09.005 [DOI] [PubMed] [Google Scholar]
19.Welikala RA, Fraz MM, Foster PJ, Whincup PH, Rudnicka AR, Owen CG, et al. Automated retinal image quality assessment on the UK Biobank dataset for epidemiological studies. Computers in Biology and Medicine. 2016;71:67–76. doi: 10.1016/j.compbiomed.2016.01.027 [DOI] [PubMed] [Google Scholar]
20.Welikala RA, Fraz MM, Habib MM, Daniel-Tong S, Yates M, Fosters PJ, et al. Automated Quantification of Retinal Vessel Morphometry in the UK Biobank Cohort. Int Conf Imag Proc. 2017. [Google Scholar]
21.Owen CG, Rudnicka AR, Welikala RA, Fraz MM, Barman SA, Luben R, et al. Retinal Vasculometry Associations with Cardiometabolic Risk Factors in the European Prospective Investigation of Cancer-Norfolk Study. Ophthalmology. 2019;126(1):96–106. doi: 10.1016/j.ophtha.2018.07.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Bennett AG, Rudnicka AR, Edgar DF. Improvements on Littmann’s method of determining the size of retinal features by fundus photography. Graefes Arch Clin Exp Ophthalmol. 1994;232(6):361–7. doi: 10.1007/BF00175988 [DOI] [PubMed] [Google Scholar]
23.Loh PR, Danecek P, Palamara PF, Fuchsberger C, Y AR, H KF, et al. Reference-based phasing using the Haplotype Reference Consortium panel. Nat Genet. 2016;48(11):1443–8. doi: 10.1038/ng.3679 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Delaneau O, Marchini J, Zagury JF. A linear complexity phasing method for thousands of genomes. Nat Methods. 2011;9(2):179–81. doi: 10.1038/nmeth.1785 [DOI] [PubMed] [Google Scholar]
25.Loh PR, Tucker G, Bulik-Sullivan BK, Vilhjálmsson BJ, Finucane HK, Salem RM, et al. Efficient Bayesian mixed-model analysis increases association power in large cohorts. Nat Genet. 2015;47(3):284–90. doi: 10.1038/ng.3190 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81(3):559–75. doi: 10.1086/519795 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Riboli E, Kaaks R. The EPIC Project: rationale and study design. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol. 1997;26 Suppl 1:S6–14. doi: 10.1093/ije/26.suppl_1.s6 [DOI] [PubMed] [Google Scholar]
28.Day N, Oakes S, Luben R, Khaw KT, Bingham S, Welch A, et al. EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br J Cancer. 1999;80 Suppl 1:95–103. [PubMed] [Google Scholar]
29.Hayat SA, Luben R, Keevil VL, Moore S, Dalzell N, Bhaniani A, et al. Cohort profile: A prospective cohort study of objective physical and cognitive capability and visual health in an ageing population of men and women in Norfolk (EPIC-Norfolk 3). Int J Epidemiol. 2014;43(4):1063–72. doi: 10.1093/ije/dyt086 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Khawaja AP, Chan MP, Hayat S, Broadway DC, Luben R, Garway-Heath DF, et al. The EPIC-Norfolk Eye Study: rationale, methods and a cross-sectional analysis of visual impairment in a population-based cohort. BMJ Open. 2013;3(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Zheng J, Erzurumluoglu AM, Elsworth BL, Kemp JP, Howe L, Haycock PC, et al. LD Hub: a centralized database and web interface to perform LD score regression that maximizes the potential of summary level GWAS data for SNP heritability and genetic correlation analysis. Bioinformatics. 2017;33(2):272–9. doi: 10.1093/bioinformatics/btw613 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Bulik-Sullivan BK, Loh PR, Finucane HK, Ripke S, Yang J, Patterson N, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47(3):291–5. doi: 10.1038/ng.3211 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Finucane HK, Reshef YA, Anttila V, Slowikowski K, Gusev A, Byrnes A, et al. Heritability enrichment of specifically expressed genes identifies disease-relevant tissues and cell types. Nat Genet. 2018;50(4):621–9. doi: 10.1038/s41588-018-0081-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.The Genotype-Tissue Expression (GTEx) project. Nat Genet. 2013;45(6):580–5. doi: 10.1038/ng.2653 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Gamazon ER, Wheeler HE, Shah KP, Mozaffari SV, Aquino-Michaels K, Carroll RJ, et al. A gene-based association method for mapping traits using reference transcriptome data. Nat Genet. 2015;47(9):1091–8. doi: 10.1038/ng.3367 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Barbeira AN, Dickinson SP, Bonazzola R, Zheng J, Wheeler HE, Torres JM, et al. Exploring the phenotypic consequences of tissue specific gene expression variation inferred from GWAS summary statistics. Nat Commun. 2018;9(1):1825. doi: 10.1038/s41467-018-03621-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Zhu Z, Zhang F, Hu H, Bakshi A, Robinson MR, Powell JE, et al. Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet. 2016;48(5):481–7. doi: 10.1038/ng.3538 [DOI] [PubMed] [Google Scholar]
38.Westra HJ, Peters MJ, Esko T, Yaghootkar H, Schurmann C, Kettunen J, et al. Systematic identification of trans eQTLs as putative drivers of known disease associations. Nat Genet. 2013;45(10):1238–43. doi: 10.1038/ng.2756 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Qi T, Wu Y, Zeng J, Zhang F, Xue A, Jiang L, et al. Identifying gene targets for brain-related traits using transcriptomic and methylomic data from blood. Nat Commun. 2018;9(1):2282. doi: 10.1038/s41467-018-04558-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7. doi: 10.7554/eLife.34408 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Jensen RA, Sim X, Smith AV, Li X, Jakobsdottir J, Cheng CY, et al. Novel Genetic Loci Associated With Retinal Microvascular Diameter. Circ Cardiovasc Genet. 2016;9(1):45–54. doi: 10.1161/CIRCGENETICS.115.001142 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Sim X, Jensen RA, Ikram MK, Cotch MF, Li XH, MacGregor S, et al. Genetic Loci for Retinal Arteriolar Microcirculation. Plos One. 2013;8(6). doi: 10.1371/journal.pone.0065804 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Kuo DS, Labelle-Dumais C, Gould DB. COL4A1 and COL4A2 mutations and disease: insights into pathogenic mechanisms and potential therapeutic targets. Hum Mol Genet. 2012;21(R1):R97–110. doi: 10.1093/hmg/dds346 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Verbeek E, Meuwissen ME, Verheijen FW, Govaert PP, Licht DJ, Kuo DS, et al. COL4A2 mutation associated with familial porencephaly and small-vessel disease. Eur J Hum Genet. 2012;20(8):844–51. doi: 10.1038/ejhg.2012.20 [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Rannikmae K, Davies G, Thomson PA, Bevan S, Devan WJ, Falcone GJ, et al. Common variation in COL4A1/COL4A2 is associated with sporadic cerebral small vessel disease. Neurology. 2015;84(9):918–26. doi: 10.1212/WNL.0000000000001309 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Zenteno JC, Crespi J, Buentello-Volante B, Buil JA, Bassaganyas F, Vela-Segarra JI, et al. Next generation sequencing uncovers a missense mutation in COL4A1 as the cause of familial retinal arteriolar tortuosity. Graefes Arch Clin Exp Ophthalmol. 2014;252(11):1789–94. doi: 10.1007/s00417-014-2800-6 [DOI] [PubMed] [Google Scholar]
47.Vasan RS, Glazer NL, Felix JF, Lieb W, Wild PS, Felix SB, et al. Genetic variants associated with cardiac structure and function: a meta-analysis and replication of genome-wide association data. JAMA. 2009;302(2):168–78. doi: 10.1001/jama.2009.978-a [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Maass PG, Aydin A, Luft FC, Schächterle C, Weise A, Stricker S, et al. PDE3A mutations cause autosomal dominant hypertension with brachydactyly. Nat Genet. 2015;47(6):647–53. doi: 10.1038/ng.3302 [DOI] [PubMed] [Google Scholar]
49.Hysi PG, Choquet H, Khawaja AP, Wojciechowski R, Tedja MS, Yin J, et al. Meta-analysis of 542,934 subjects of European ancestry identifies new genes and mechanisms predisposing to refractive error and myopia. Nature Genetics. 2020;52(4):401–7. doi: 10.1038/s41588-020-0599-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Gao XR, Huang H, Kim H. Genome-wide association analyses identify 139 loci associated with macular thickness in the UK Biobank cohort. Hum Mol Genet. 2019;28(7):1162–72. doi: 10.1093/hmg/ddy422 [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Shah RL, Guggenheim JA, Eye UKB, Vision C. Genome-wide association studies for corneal and refractive astigmatism in UK Biobank demonstrate a shared role for myopia susceptibility loci. Hum Genet. 2018;137(11–12):881–96. doi: 10.1007/s00439-018-1942-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Sjoblom B, Salmazo A, Djinovic-Carugo K. Alpha-actinin structure and regulation. Cell Mol Life Sci. 2008;65(17):2688–701. doi: 10.1007/s00018-008-8080-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS Genet. doi: 10.1371/journal.pgen.1010583.r001

Decision Letter 0

Hua Tang, Sudha K Iyengar

30 Aug 2022

Dear Dr Owen,

Thank you very much for submitting your Research Article entitled 'GWAS on retinal vasculometry phenotypes' to PLOS Genetics.

The manuscript was fully evaluated at the editorial level and by independent peer reviewers. The reviewers appreciated the attention to an important problem, but raised some substantial concerns about the current manuscript. Based on the reviews, we will not be able to accept this version of the manuscript, but we would be willing to review a much-revised version. We cannot, of course, promise publication at that time.

Should you decide to revise the manuscript for further consideration here, your revisions should address the specific points made by each reviewer. We will also require a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript.

If you decide to revise the manuscript for further consideration at PLOS Genetics, please aim to resubmit within the next 60 days, unless it will take extra time to address the concerns of the reviewers, in which case we would appreciate an expected resubmission date by email to plosgenetics@plos.org.

If present, accompanying reviewer attachments are included with this email; please notify the journal office if any appear to be missing. They will also be available for download from the link below. You can use this link to log into the system when you are ready to submit a revised version, having first consulted our Submission Checklist.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please be aware that our data availability policy requires that all numerical data underlying graphs or summary statistics are included with the submission, and you will need to provide this upon resubmission if not already present. In addition, we do not permit the inclusion of phrases such as "data not shown" or "unpublished results" in manuscripts. All points should be backed up by data provided with the submission.

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

PLOS has incorporated Similarity Check, powered by iThenticate, into its journal-wide submission system in order to screen submitted content for originality before publication. Each PLOS journal undertakes screening on a proportion of submitted articles. You will be contacted if needed following the screening process.

To resubmit, use the link below and 'Revise Submission' in the 'Submissions Needing Revision' folder.

[LINK]

We are sorry that we cannot be more positive about your manuscript at this stage. Please do not hesitate to contact us if you have any concerns or questions.

Yours sincerely,

Sudha K Iyengar

Academic Editor

PLOS Genetics

Hua Tang

Section Editor

PLOS Genetics

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: The manuscript by Jiang et al describes genome-wide analyses of retinal vascular parameters with several bioinformatic analyses designed to identify phenotypic overlap. The manuscript is largely well-written and fairly clear, however there are some elements which would benefit from additional clarification. Broadly speaking, it would add quite a bit to the manuscript to discuss the results in the context of gene regulation across your various tissues in addition to phenotypic causation. Other comments include:

1. Is there overlap between EPIC-Norfolk and the UKB participants included in this study? Given the similar catchment areas it would be best to include a statement of independent samples.

2. Were the parameters similar for the LDSC and LD Hub analyses? Specifically LDSC recommends filtration to HapMap SNPs, and use of stringent info score thresholds, were identical thresholds applied in both approaches?

3. Which version of GTEx was used for your S-PrediXcan analysis? 44 tissues would suggest use of v6 rather than the newer (and larger) v7 or v8, and if so ought to be stated. Also where are the results from these analyses? They're never discussed or presented.

4. The tissue enrichment methods should be described in bit more detail, specifically indications of the versions of databases used, etc. as the 200+ tissues interrogated are not listed in more detail. The results indicates GTEx tissues were used for enrichment, but that doesn't add up - there are not 200+ tissues available in GTEx, and myometrium isn't one of them. Please clarify.

5. I am puzzled by the selection of tibial artery rather than coronary artery as one of the tissues, given that coronary artery was a top signal in two of the four analyses.

6. The manhattan plots show many significant signals, far more than are mentioned in the case of arteriolar width. If filters were applied, that needs to be reflected in what is plotted and described more thoroughly. If not, then the results are not described properly and there would be more than enough datapoints to enable comparison of effect sizes between EPIC-Norfolk and the UKB.

7. This isn't critical, but it would be nice to see figures analogous to Supplementary Figure 2 for the cross-ancestry analyses in addition to the supplementary table. This information should also be mentioned in the discussion paragraphs of particular gene results.

8. Line 278: hereditability = heritability

9. The cross-disease genetic correlation analyses seem highly unfocused. The numbers of phenotypes need to be clarified, preferably restricted to those with a scientific rationale ("Distance between home and job workplace"). Also ST9 lists all populations as UKBB, of which your discovery analyses is a subset, this is somewhat concerning.

10. Previously identified results from reference 42 are mentioned in the second paragraph of the discussion, were these among the top hits in this manuscript, and/or were these SNPs evaluated systematically? Noting the replication of these loci would be useful.

Reviewer #2: The authors perform a large GWAS using four QUARTZ-derived fundus features as quantitative traits in 52,798 UK biobank individuals. While this is not the first GWAS on vascular features of the UK biobank fundus photos, this is an important topic of investigation given the link to systemic disease. The research methods in this paper are sound. I do think there needs to be substantial additional consideration to the analysis that I will discuss below. I have some general comments followed by more specific page/line comments:

Image processing: I do not have experience with QUARTZ but it seems reasonable. However, I think that the description of imaging processing needs more detail (I looked in the supplemental but do not see any further detail):

• The authors do not describe how they handle each eye – is one eye used? Both eyes? If both, how do they account for correlation? How do they handle significant differences between the two eyes and the possible confounding etiologies?

• How do the authors correct for image quality? This will degrade with heritable things like corneal disease and cataract, and I assume degradation of the image will systematically affect the 4 measurements.

• How about correcting for retinal pathology in general? Are these all patients without a diagnosis of certain major categories of ocular disease? For example: someone with a history of retinal laser (for diabetes, a retinal tear, a retinal detachment, etc) will have different vasculometry measurements due to the ablation of the retina in the area of laser. Traits leading to the need for laser are heritable.

The GWAS itself: the methods are sound and use well-established analysis pipelines. The authors correct for age, sex, spherical equivalent and the first ten principal components. My main concerns revolve around confounders and clear delineation about findings here that imply a novel pathway as separate from one that is already well understood by ophthalmologists. To be specific:

• Smoking status is not controlled for in this paper – it should be added as a covariate, as it has definite secondary effects on the vasculature throughout the body.

• retinal vein occlusion (or impending vein occlusion), one of the more common secondary reasons for vascular tortuosity, is not mentioned in this paper. A sign of this would be a difference in tortuosity between the two eyes (in the case of CRVO and assuming it wasn’t bilateral) or a sectoral change within the eye (BRVO). Can the authors account for this in any way (there are likely several ways of doing this)?

• Systemic hypertension, cardiovascular disease, diabetes and (to a certain degree) hyperlipidemia are other well known clinical associations with vascular tortuosity on exam. In a way, it is not novel to find SNPs that co-associate with VT and these traits. Yet, there is no analysis here that controls for this. One could argue that controlling for these traits is not necessarily desired, because they are in the causal pathway, but it would be interesting to have a secondary analysis that controls for blood pressure and diabetes (for example) and reports SNPs that associate independent of this confounder that is already understood as associated with vascular tortuosity.

• It is standard to correct for age-squared to model its effect and account more accurately for any non-linearity

• Uncontrolled diabetes and hypertension (amongst other things) can lead to vascular narrowing and pruning and eventual non-perfusion of the peripheral retina. I assume the time with respect to disease onset will have an effect on the vasculometry measurements. Can the authors discuss how they handled this (or why they decided not to control for it)?

• I don’t believe genotyping array is added as a covariate and it likely should be if it varied in this large sample size

More specific comments:

Page 3 line 68: This description of the retinal vasculature is not correct. The long and short ciliary artery split off BEFORE penetrating the sclera. Together, the long posterior ciliary artery and the short posterior ciliary arteries feed the choroid (long and short), iris (long) and ciliary body (long).

Page 3 line 73: Retinal vein occlusion (or impending vein occlusion) should have been considered in this manuscript as a cause for venous tortuosity, especially if unilateral.

Page 3 line 74: Cite Zekavat et al. Circulation 2022 as a paper that is already published using the UK biobank to study retinal vasculature. Please alter statements about novelty and explain the distinction between VT/VD and the traits in the Zekavat paper. Also cite this work: https://www.medrxiv.org/content/10.1101/2020.06.25.20139725v6 which is a pre-print of a paper that performs a GWAS on vascular tortuosity in the UK Biobank. Of note this work is still in pre-print only, however.

Page 6 lin 223: COL4A1 (next to COL4A2) is the gene thought to be causal of the medialian disorders FRAT and HANAC. I’m wondering why FRAT is not mentioned in this paper (HANAC is less well known but also thought to be caused by COL4A1 mutation). FRAT is a well known genetic cause of vascular tortuosity. It’s interesting that COL4A2, the neighboring gene, comes up – perhaps the authors can comment on this – are we wrong about COL4A1 being the cause of FRAT? Or is the authors’ signal actually tagging COL4A1? How are COL4A1 and COL4A2 related intracellularly?

Page 7 line 247: Great that the authors replicate their work in this paper. However, the text should be more specific. They report a “generally linear relationship” between effect sizes – please report the specific correlation with confidence interval and p-value, which should be strong based on the supplementary figure.

Page 7 line 258: report p-value and confidence interval for correlations with other ancestries

Page 7 line 280: can the authors be more statistically specific – what is the exact p-value, how many SNPs/eQTLs were tested, what is the resulting Bonferroni cutoff.

Page 8 line 293: The discussing of two-sample MR analysis to assess causality is really interesting. Yet, certain clinical observations make it hard to reconcile that retinal arteriolar tortuosity causes elevated diastolic blood pressure. A mechanical elevation in systemic blood pressure via microvasculature is more believable in the kidneys, for example, where there is a large enough tissue mass to actually affect systemic blood pressure. But the retinal arteries are massively smaller in quantity. Can the authors please provide more details as to how well powered they were to decide about causality direction? Could there be an unmeasured confounder here? For example, in the situation of RVO (2% prevalence) or impending RVO, we know that vascular tortuosity comes second, with elevated blood pressure being in the causal pathway. On longitudinal patient exam, the vessels are normal, then there is uncontrolled hypertension which leads to thickening of arteriolar walls and subsequent compression of the vein, resulting in venous tortuosity with associated RVO or impending RVO. To me, this seems like the opposite direct as is described by the authors.

**********

Have all data underlying the figures and results presented in the manuscript been provided?

Large-scale datasets should be made available via a public repository as described in the PLOS Genetics data availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.

Reviewer #1: Yes

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

PLoS Genet. 2023 Feb 9;19(2):e1010583. doi: 10.1371/journal.pgen.1010583.r002

Author response to Decision Letter 0

16 Nov 2022

Attachment

Submitted filename: PGENETICS-D-22-00757_Response to reviewers (141122).docx

Click here for additional data file.^{(45KB, docx)}

PLoS Genet. doi: 10.1371/journal.pgen.1010583.r003

Decision Letter 1

Hua Tang, Sudha K Iyengar

20 Dec 2022

Dear Dr Owen,

We are pleased to inform you that your manuscript entitled "GWAS on retinal vasculometry phenotypes" has been editorially accepted for publication in PLOS Genetics. Congratulations!

Before your submission can be formally accepted and sent to production you will need to complete our formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Please note: the accept date on your published article will reflect the date of this provisional acceptance, but your manuscript will not be scheduled for publication until the required changes have been made.

Once your paper is formally accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you’ve already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosgenetics@plos.org.

In the meantime, please log into Editorial Manager at https://www.editorialmanager.com/pgenetics/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production and billing process. Note that PLOS requires an ORCID iD for all corresponding authors. Therefore, please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager.

If you have a press-related query, or would like to know about making your underlying data available (as you will be aware, this is required for publication), please see the end of this email. If your institution or institutions have a press office, please notify them about your upcoming article at this point, to enable them to help maximise its impact. Inform journal staff as soon as possible if you are preparing a press release for your article and need a publication date.

Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Genetics!

Yours sincerely,

Sudha K Iyengar

Academic Editor

PLOS Genetics

Hua Tang

Section Editor

PLOS Genetics

www.plosgenetics.org

Twitter: @PLOSGenetics

----------------------------------------------------

Comments from the reviewers (if applicable):

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: The authors have responded satisfactorily to my previous comments.

Reviewer #2: Thank you to the authors for your detailed responses. All my questions were answered.

**********

Have all data underlying the figures and results presented in the manuscript been provided?

Reviewer #1: Yes

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

----------------------------------------------------

Data Deposition

If you have submitted a Research Article or Front Matter that has associated data that are not suitable for deposition in a subject-specific public repository (such as GenBank or ArrayExpress), one way to make that data available is to deposit it in the Dryad Digital Repository. As you may recall, we ask all authors to agree to make data available; this is one way to achieve that. A full list of recommended repositories can be found on our website.

The following link will take you to the Dryad record for your article, so you won't have to re‐enter its bibliographic information, and can upload your files directly:

http://datadryad.org/submit?journalID=pgenetics&manu=PGENETICS-D-22-00757R1

More information about depositing data in Dryad is available at http://www.datadryad.org/depositing. If you experience any difficulties in submitting your data, please contact help@datadryad.org for support.

Additionally, please be aware that our data availability policy requires that all numerical data underlying display items are included with the submission, and you will need to provide this before we can formally accept your manuscript, if not already present.

----------------------------------------------------

Press Queries

If you or your institution will be preparing press materials for this manuscript, or if you need to know your paper's publication date for media purposes, please inform the journal staff as soon as possible so that your submission can be scheduled accordingly. Your manuscript will remain under a strict press embargo until the publication date and time. This means an early version of your manuscript will not be published ahead of your final version. PLOS Genetics may also choose to issue a press release for your article. If there's anything the journal should know or you'd like more information, please get in touch via plosgenetics@plos.org.

PLoS Genet. doi: 10.1371/journal.pgen.1010583.r004

Acceptance letter

Hua Tang, Sudha K Iyengar

19 Jan 2023

PGENETICS-D-22-00757R1

GWAS on retinal vasculometry phenotypes

Dear Dr Owen,

We are pleased to inform you that your manuscript entitled "GWAS on retinal vasculometry phenotypes" has been formally accepted for publication in PLOS Genetics! Your manuscript is now with our production department and you will be notified of the publication date in due course.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript.

Soon after your final files are uploaded, unless you have opted out or your manuscript is a front-matter piece, the early version of your manuscript will be published online. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting PLOS Genetics and open-access publishing. We are looking forward to publishing your work!

With kind regards,

Marianna Bach

PLOS Genetics

On behalf of:

The PLOS Genetics Team

Carlyle House, Carlyle Road, Cambridge CB4 3DN | United Kingdom

plosgenetics@plos.org | +44 (0) 1223-442823

plosgenetics.org | Twitter: @PLOSGenetics

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Acknowledgments. UK Biobank Eye and Vision Consortium.

(DOCX)

Click here for additional data file.^{(14.2KB, docx)}

S1 Text. Supplementary Figures.

(DOCX)

Click here for additional data file.^{(1.5MB, docx)}

S1 Table. Supplementary Tables.

(XLSX)

Click here for additional data file.^{(217.2KB, xlsx)}

Attachment

Submitted filename: PGENETICS-D-22-00757_Response to reviewers (141122).docx

Click here for additional data file.^{(45KB, docx)}

Data Availability Statement

[pgen.1010583.ref001] 1.Wong-Riley MT. Energy metabolism of the visual system. Eye Brain. 2010;2:99–116. doi: 10.2147/EB.S9078 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref002] 2.Sun C, Zhu G, Wong TY, Hewitt AW, Ruddle JB, Hodgson L, et al. Quantitative genetic analysis of the retinal vascular caliber: the Australian Twins Eye Study. Hypertension. 2009;54(4):788–95. doi: 10.1161/HYPERTENSIONAHA.109.132902 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref003] 3.Ikram MK, Sim X, Jensen RA, Cotch MF, Hewitt AW, Ikram MA, et al. Four novel Loci (19q13, 6q24, 12q24, and 5q14) influence the microcirculation in vivo. PLoS Genet. 2010;6(10):e1001184. doi: 10.1371/journal.pgen.1001184 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref004] 4.Veluchamy A, Ballerini L, Vitart V, Schraut KE, Kirin M, Campbell H, et al. Novel Genetic Locus Influencing Retinal Venular Tortuosity Is Also Associated With Risk of Coronary Artery Disease. Arterioscler Thromb Vasc Biol. 2019;39(12):2542–52. doi: 10.1161/ATVBAHA.119.312552 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref005] 5.Joyal JS, Sun Y, Gantner ML, Shao Z, Evans LP, Saba N, et al. Retinal lipid and glucose metabolism dictates angiogenesis through the lipid sensor Ffar1. Nat Med. 2016;22(4):439–45. doi: 10.1038/nm.4059 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref006] 6.Sun Y, Lin Z, Liu CH, Gong Y, Liegl R, Fredrick TW, et al. Inflammatory signals from photoreceptor modulate pathological retinal angiogenesis via c-Fos. J Exp Med. 2017;214(6):1753–67. doi: 10.1084/jem.20161645 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref007] 7.Cheung CY, Zheng Y, Hsu W, Lee ML, Lau QP, Mitchell P, et al. Retinal vascular tortuosity, blood pressure, and cardiovascular risk factors. Ophthalmology. 2011;118(5):812–8. doi: 10.1016/j.ophtha.2010.08.045 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref008] 8.Koch E, Rosenbaum D, Brolly A, Sahel JA, Chaumet-Riffaud P, Girerd X, et al. Morphometric analysis of small arteries in the human retina using adaptive optics imaging: relationship with blood pressure and focal vascular changes. J Hypertens. 2014;32(4):890–8. doi: 10.1097/HJH.0000000000000095 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref009] 9.Luo T, Gast TJ, Vermeer TJ, Burns SA. Retinal Vascular Branching in Healthy and Diabetic Subjects. Invest Ophthalmol Vis Sci. 2017;58(5):2685–94. doi: 10.1167/iovs.17-21653 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref010] 10.Ikram MK, Ong YT, Cheung CY, Wong TY. Retinal vascular caliber measurements: clinical significance, current knowledge and future perspectives. Ophthalmologica. 2013;229(3):125–36. doi: 10.1159/000342158 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref011] 11.Tapp RJ, Owen CG, Barman SA, Welikala RA, Foster PJ, Whincup PH, et al. Associations of Retinal Microvascular Diameters and Tortuosity With Blood Pressure and Arterial Stiffness. Hypertension. 2019;74(6):1383–90. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref012] 12.Tapp RJ, Owen CG, Barman SA, Welikala RA, Foster PJ, Whincup PH, et al. Retinal Vascular Tortuosity and Diameter Associations with Adiposity and Components of Body Composition. Obesity. 2020;28(9):1750–60. doi: 10.1002/oby.22885 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref013] 13.Newman A, Andrew N, Casson R. Review of the association between retinal microvascular characteristics and eye disease. Clin Exp Ophthalmol. 2018;46(5):531–52. doi: 10.1111/ceo.13119 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref014] 14.Rudnicka AR, Owen CG, Welikala RA, Barman SA, Whincup PH, Strachan DP, et al. Retinal Vasculometry Associations With Glaucoma: Findings From the European Prospective Investigation of Cancer-Norfolk Eye Study. Am J Ophthalmol. 2020;220:140–51. doi: 10.1016/j.ajo.2020.07.027 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref015] 15.Manolio TA, Brooks LD, Collins FS. A HapMap harvest of insights into the genetics of common disease. J Clin Invest. 2008;118(5):1590–605. doi: 10.1172/JCI34772 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref016] 16.Zekavat SM, Raghu VK, Trinder M, Ye Y, Koyama S, Honigberg MC, et al. Deep Learning of the Retina Enables Phenome- and Genome-Wide Analyses of the Microvasculature. Circulation. 2022;145(2):134–50. doi: 10.1161/CIRCULATIONAHA.121.057709 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref017] 17.Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK Biobank resource with deep phenotyping and genomic data. Nature. 2018;562(7726):203–9. doi: 10.1038/s41586-018-0579-z [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref018] 18.Welikala RA, Foster PJ, Whincup PH, Rudnicka AR, Owen CG, Strachan DP, et al. Automated arteriole and venule classification using deep learning for retinal images from the UK Biobank cohort. Computers in Biology and Medicine. 2017;90:23–32. doi: 10.1016/j.compbiomed.2017.09.005 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref019] 19.Welikala RA, Fraz MM, Foster PJ, Whincup PH, Rudnicka AR, Owen CG, et al. Automated retinal image quality assessment on the UK Biobank dataset for epidemiological studies. Computers in Biology and Medicine. 2016;71:67–76. doi: 10.1016/j.compbiomed.2016.01.027 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref020] 20.Welikala RA, Fraz MM, Habib MM, Daniel-Tong S, Yates M, Fosters PJ, et al. Automated Quantification of Retinal Vessel Morphometry in the UK Biobank Cohort. Int Conf Imag Proc. 2017. [Google Scholar]

[pgen.1010583.ref021] 21.Owen CG, Rudnicka AR, Welikala RA, Fraz MM, Barman SA, Luben R, et al. Retinal Vasculometry Associations with Cardiometabolic Risk Factors in the European Prospective Investigation of Cancer-Norfolk Study. Ophthalmology. 2019;126(1):96–106. doi: 10.1016/j.ophtha.2018.07.022 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref022] 22.Bennett AG, Rudnicka AR, Edgar DF. Improvements on Littmann’s method of determining the size of retinal features by fundus photography. Graefes Arch Clin Exp Ophthalmol. 1994;232(6):361–7. doi: 10.1007/BF00175988 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref023] 23.Loh PR, Danecek P, Palamara PF, Fuchsberger C, Y AR, H KF, et al. Reference-based phasing using the Haplotype Reference Consortium panel. Nat Genet. 2016;48(11):1443–8. doi: 10.1038/ng.3679 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref024] 24.Delaneau O, Marchini J, Zagury JF. A linear complexity phasing method for thousands of genomes. Nat Methods. 2011;9(2):179–81. doi: 10.1038/nmeth.1785 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref025] 25.Loh PR, Tucker G, Bulik-Sullivan BK, Vilhjálmsson BJ, Finucane HK, Salem RM, et al. Efficient Bayesian mixed-model analysis increases association power in large cohorts. Nat Genet. 2015;47(3):284–90. doi: 10.1038/ng.3190 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref026] 26.Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81(3):559–75. doi: 10.1086/519795 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref027] 27.Riboli E, Kaaks R. The EPIC Project: rationale and study design. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol. 1997;26 Suppl 1:S6–14. doi: 10.1093/ije/26.suppl_1.s6 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref028] 28.Day N, Oakes S, Luben R, Khaw KT, Bingham S, Welch A, et al. EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br J Cancer. 1999;80 Suppl 1:95–103. [PubMed] [Google Scholar]

[pgen.1010583.ref029] 29.Hayat SA, Luben R, Keevil VL, Moore S, Dalzell N, Bhaniani A, et al. Cohort profile: A prospective cohort study of objective physical and cognitive capability and visual health in an ageing population of men and women in Norfolk (EPIC-Norfolk 3). Int J Epidemiol. 2014;43(4):1063–72. doi: 10.1093/ije/dyt086 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref030] 30.Khawaja AP, Chan MP, Hayat S, Broadway DC, Luben R, Garway-Heath DF, et al. The EPIC-Norfolk Eye Study: rationale, methods and a cross-sectional analysis of visual impairment in a population-based cohort. BMJ Open. 2013;3(3). [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref031] 31.Zheng J, Erzurumluoglu AM, Elsworth BL, Kemp JP, Howe L, Haycock PC, et al. LD Hub: a centralized database and web interface to perform LD score regression that maximizes the potential of summary level GWAS data for SNP heritability and genetic correlation analysis. Bioinformatics. 2017;33(2):272–9. doi: 10.1093/bioinformatics/btw613 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref032] 32.Bulik-Sullivan BK, Loh PR, Finucane HK, Ripke S, Yang J, Patterson N, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47(3):291–5. doi: 10.1038/ng.3211 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref033] 33.Finucane HK, Reshef YA, Anttila V, Slowikowski K, Gusev A, Byrnes A, et al. Heritability enrichment of specifically expressed genes identifies disease-relevant tissues and cell types. Nat Genet. 2018;50(4):621–9. doi: 10.1038/s41588-018-0081-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref034] 34.The Genotype-Tissue Expression (GTEx) project. Nat Genet. 2013;45(6):580–5. doi: 10.1038/ng.2653 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref035] 35.Gamazon ER, Wheeler HE, Shah KP, Mozaffari SV, Aquino-Michaels K, Carroll RJ, et al. A gene-based association method for mapping traits using reference transcriptome data. Nat Genet. 2015;47(9):1091–8. doi: 10.1038/ng.3367 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref036] 36.Barbeira AN, Dickinson SP, Bonazzola R, Zheng J, Wheeler HE, Torres JM, et al. Exploring the phenotypic consequences of tissue specific gene expression variation inferred from GWAS summary statistics. Nat Commun. 2018;9(1):1825. doi: 10.1038/s41467-018-03621-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref037] 37.Zhu Z, Zhang F, Hu H, Bakshi A, Robinson MR, Powell JE, et al. Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet. 2016;48(5):481–7. doi: 10.1038/ng.3538 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref038] 38.Westra HJ, Peters MJ, Esko T, Yaghootkar H, Schurmann C, Kettunen J, et al. Systematic identification of trans eQTLs as putative drivers of known disease associations. Nat Genet. 2013;45(10):1238–43. doi: 10.1038/ng.2756 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref039] 39.Qi T, Wu Y, Zeng J, Zhang F, Xue A, Jiang L, et al. Identifying gene targets for brain-related traits using transcriptomic and methylomic data from blood. Nat Commun. 2018;9(1):2282. doi: 10.1038/s41467-018-04558-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref040] 40.Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7. doi: 10.7554/eLife.34408 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref041] 41.Jensen RA, Sim X, Smith AV, Li X, Jakobsdottir J, Cheng CY, et al. Novel Genetic Loci Associated With Retinal Microvascular Diameter. Circ Cardiovasc Genet. 2016;9(1):45–54. doi: 10.1161/CIRCGENETICS.115.001142 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref042] 42.Sim X, Jensen RA, Ikram MK, Cotch MF, Li XH, MacGregor S, et al. Genetic Loci for Retinal Arteriolar Microcirculation. Plos One. 2013;8(6). doi: 10.1371/journal.pone.0065804 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref043] 43.Kuo DS, Labelle-Dumais C, Gould DB. COL4A1 and COL4A2 mutations and disease: insights into pathogenic mechanisms and potential therapeutic targets. Hum Mol Genet. 2012;21(R1):R97–110. doi: 10.1093/hmg/dds346 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref044] 44.Verbeek E, Meuwissen ME, Verheijen FW, Govaert PP, Licht DJ, Kuo DS, et al. COL4A2 mutation associated with familial porencephaly and small-vessel disease. Eur J Hum Genet. 2012;20(8):844–51. doi: 10.1038/ejhg.2012.20 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref045] 45.Rannikmae K, Davies G, Thomson PA, Bevan S, Devan WJ, Falcone GJ, et al. Common variation in COL4A1/COL4A2 is associated with sporadic cerebral small vessel disease. Neurology. 2015;84(9):918–26. doi: 10.1212/WNL.0000000000001309 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref046] 46.Zenteno JC, Crespi J, Buentello-Volante B, Buil JA, Bassaganyas F, Vela-Segarra JI, et al. Next generation sequencing uncovers a missense mutation in COL4A1 as the cause of familial retinal arteriolar tortuosity. Graefes Arch Clin Exp Ophthalmol. 2014;252(11):1789–94. doi: 10.1007/s00417-014-2800-6 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref047] 47.Vasan RS, Glazer NL, Felix JF, Lieb W, Wild PS, Felix SB, et al. Genetic variants associated with cardiac structure and function: a meta-analysis and replication of genome-wide association data. JAMA. 2009;302(2):168–78. doi: 10.1001/jama.2009.978-a [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref048] 48.Maass PG, Aydin A, Luft FC, Schächterle C, Weise A, Stricker S, et al. PDE3A mutations cause autosomal dominant hypertension with brachydactyly. Nat Genet. 2015;47(6):647–53. doi: 10.1038/ng.3302 [DOI] [PubMed] [Google Scholar]

[pgen.1010583.ref049] 49.Hysi PG, Choquet H, Khawaja AP, Wojciechowski R, Tedja MS, Yin J, et al. Meta-analysis of 542,934 subjects of European ancestry identifies new genes and mechanisms predisposing to refractive error and myopia. Nature Genetics. 2020;52(4):401–7. doi: 10.1038/s41588-020-0599-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref050] 50.Gao XR, Huang H, Kim H. Genome-wide association analyses identify 139 loci associated with macular thickness in the UK Biobank cohort. Hum Mol Genet. 2019;28(7):1162–72. doi: 10.1093/hmg/ddy422 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref051] 51.Shah RL, Guggenheim JA, Eye UKB, Vision C. Genome-wide association studies for corneal and refractive astigmatism in UK Biobank demonstrate a shared role for myopia susceptibility loci. Hum Genet. 2018;137(11–12):881–96. doi: 10.1007/s00439-018-1942-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgen.1010583.ref052] 52.Sjoblom B, Salmazo A, Djinovic-Carugo K. Alpha-actinin structure and regulation. Cell Mol Life Sci. 2008;65(17):2688–701. doi: 10.1007/s00018-008-8080-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

GWAS on retinal vasculometry phenotypes

Xiaofan Jiang

Pirro G Hysi

Anthony P Khawaja

Omar A Mahroo

Zihe Xu

Christopher J Hammond

Paul J Foster

Roshan A Welikala

Sarah A Barman

Peter H Whincup

Alicja R Rudnicka

Christopher G Owen

David P Strachan

Roles

Abstract

Author summary

Introduction

Methods

Ethics statement

Participants

Phenotyping and retinal imaging

Statistical analyses

Fig 1. Particpant flow diagram.

Replication GWAS—EPIC-Norfolk

Linkage Disequilibrium (LD) score regression analyses

Tissue enrichment across multiple tissues

Transcriptome-based association analyses

Integration of genotypic and expression data

Mendelian Randomisation (MR)

Results

Association results

Replication and transethnic comparison

Tissue expression enrichment and exploration of functional mechanisms in vasculometry regulation

Relationship of vasculometry and other phenotypes

Fig 2. Results of Mendelian Randomization analyses for arteriolar tortuosity (AT) as a causal influence on diastolic blood pressure (DBP).

Discussion

Supporting information

Data Availability

Funding Statement

References

Decision Letter 0

Hua Tang

Sudha K Iyengar

Roles

Author response to Decision Letter 0

Decision Letter 1

Hua Tang

Sudha K Iyengar

Roles

Acceptance letter

Hua Tang

Sudha K Iyengar

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases