The Genetic Architecture of Iris Colour and Surface Feature Variation in Populations of Diverse Ancestry

Melissa Edwards

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Anthropology University of Toronto

Melissa Edwards

Doctor of Philosophy

Department of Anthropology University of Toronto

2016

Abstract

Research focused on iris colour and surface features has begun to play an increasingly important role in the fields of anthropology, forensics and public health. Although both eye colour and, to a much lesser extent, iris features have been studied in populations of European ancestry, very few research groups have attempted to investigate these traits in regions outside of

Europe. This may be partly due to difficulties associated with obtaining a quantitative measurement of eye colour and a lack of standardized methodology. For my doctoral thesis, I developed a novel method of characterizing iris colour and feature variation in diverse populations. I then applied this method to a sample of individuals of East Asian, European and

South Asian ancestry. I was primarily interested in looking at the phenotypic distribution and genetic basis of these traits across all three sample sets. I found that: 1/ Quantitative methods provide an ideal means of characterizing iris colour in populations of diverse ancestry; 2/ The phenotypic distributions of iris colour and surface features are different in East Asian, European and South Asian populations; 3/ HERC2 rs12913832 controls iris pigmentation variation in both

European and South Asian populations. However, it affects eye colour differently in both regions; 4/ The variants responsible for controlling iris pigmentation in Europe are different from

those responsible for controlling iris pigmentation in East Asia; 5/ SEMA3A rs10235789 is significantly associated with Fuchs’ crypts in East Asian, European and South Asian populations.

Ultimately, my thesis is the first research work to compare iris pigmentation and structure variation in populations of diverse ancestry. The methodology that I have developed will provide other researchers with a new framework that they can use to approach the study of iris colour and feature diversity.

iii

Table of Contents Abstract ...... ii Table of Contents ...... iv List of Tables ...... vii List of Figures ...... viii Forward ...... ix Chapter 1: ...... 1 Introduction 1.1 The Evolution of Global Iris Colour Variation ...... 2 1.2 Anatomy of the Human Iris ...... 3 1.2.1 Iris Structure ...... 3 1.2.2 Iris Pigmentation ...... 8 1.3 Factors that Influence Iris Pigmentation Variation ...... 9 1.4 Genetic Basis of Iris Pigmentation ...... 11 1.5 Genetic Basis of Iris Surface Features ...... 14 1.6 Limitations in our Understanding of Iris Colour and Texture Variation ...... 16 1.6.1 Characterizing Iris Colour and Surface Feature Variation ...... 16 1.6.2 Characterizing Central Heterochromia ...... 17 1.6.3 Population Variation ...... 18 1.7 Rationale and Study Goals ...... 20 1.8 References ...... 22 Chapter 2: ...... 28 Quantitative Measures of Iris Colour Using High Resolution Photographs 2.1 Abstract ...... 29 2.2 Introduction ...... 30 2.3 Materials and Methods ...... 31 2.3.1 Recruitment ...... 31 2.3.2 Genotyping ...... 32 2.3.3 Acquisition and Processing of Iris Photographs ...... 32 2.3.4 Acquisition of Colour Measurements ...... 33 2.3.5 Statistical Analysis ...... 34 2.4 Results ...... 34

2.5 Discussion ...... 36 2.6 Tables ...... 42 2.7 Figures ...... 44 2.8 References ...... 47 Chapter 3: ...... 49 Iris Pigmentation as a Quantitative Trait: Variation in Populations of European, East Asian and South Asian Ancestry and Association with Candidate Gene Polymorphisms 3.1 Abstract ...... 50 3.2 Introduction ...... 51 3.3 Materials and Methods ...... 54 3.3.1 Sample Collection ...... 54 3.3.2 Acquisition of Photographs ...... 54 3.3.3 Acquisition of Iris Colour ...... 55 3.3.4 Marker Selection and Genotyping ...... 57 3.3.5 Statistical Analysis ...... 57 3.4 Results ...... 59 3.5 Discussion ...... 62 3.5.1 Iris Colour in European Populations ...... 63 3.5.2 Iris Colour in South Asian Populations ...... 65 3.5.3 Iris Colour in East Asian Populations ...... 67 3.5.4 Central Heterochromia ...... 69 3.5.5 A Global View of the Genetic Basis of Iris Pigmentation ...... 70 3.6 Tables ...... 73 3.7 Figures ...... 86 3.8 References ...... 89 Chapter 4: ...... 94 Analysis of Iris Surface Features in Populations of Diverse Ancestry 4.1 Abstract ...... 95 4.2 Introduction ...... 96 4.3 Materials and Methods ...... 98 4.3.1 Participant Recruitment ...... 98 4.3.2 Genotyping ...... 98 4.3.3 Acquisition and Processing of Iris Photographs ...... 99

4.3.4 Iris Analysis ...... 99 4.3.5 Characterization of Iris Structure ...... 101 4.3.6 Statistical Analysis ...... 102 4.4 Results ...... 103 4.5 Discussion ...... 106 4.5.1 Fuchs’ Crypts...... 106 4.5.2 Pigment Spots ...... 108 4.5.3 Contraction Furrows ...... 109 4.5.4 Wolfflin Nodules ...... 111 4.5.5 Conjunctival Melanosis ...... 112 4.5.6 Correlations of Iris Features ...... 113 4.5.7 Future Research ...... 113 4.6 Tables ...... 116 4.7 Figures ...... 120 4.8 References ...... 129 Chapter 5: ...... 132 Concluding Remarks 5.1 Introduction ...... 133 5.2 Summary of Findings ...... 133 5.2.1 Quantifying Iris Pigmentation ...... 133 5.2.2 Characterizing Iris Surface Features ...... 134 5.2.3 Phenotypic Distribution of Iris Colour and Central Heterochromia ...... 135 5.2.4 Phenotypic Distribution of Iris Surface Features ...... 136 5.2.5 Genetic Basis of Iris Colour ...... 136 5.2.6 Genetic Basis of Iris Surface Features ...... 138 5.3 Research Limitations ...... 139 5.4 Conclusion ...... 140 5.5 References ...... 142 Copyright Acknowledgements ...... 144

List of Tables

Table 2-1 General Descriptive Statistics ...... 42 Table 2-2 General Linear Model Analysis Results ...... 43 Table 3-1 Results of the Inter- and Intra-Rater Reliability Analysis ...... 73 Table 3-2 Summary Statistics for All 1448 Participants Included in the Study ...... 74 Table 3-3 Results of the Meta-Analysis in the European Sample (A) Before and (B) After Conditioning for the Effects of HERC2 rs12913832 ...... 75 Table 3-4 Results of the Meta-Analysis in the South Asian Sample (A) Before and (B) After Conditioning for the Effects of HERC2 rs12913832 ...... 79 Table 3-5 Results of the Meta-Analysis in the East Asian Sample (A) Before and (B) After Conditioning for the Effects of OCA2 rs1800414 ...... 83 Table 3-6 Average, Minimum and Maximum ∆E Stratified by HERC2 rs12913832 genotype ...85 Table 4-1 General Descriptive Statistics for the East Asian, European and South Asian Irises .116 Table 4-2 Results of the Genetic Ordinal Regression ...... 117 Table 4-3 Iris Surface Feature Frequencies ...... 118 Table 4-4 Distribution of Iris Surface Features ...... 119

vii

List of Figures

Figure 1-1 Surface Features of the Humans Iris ...... 5 Figure 1-2 Cross-Section of the Human Iris ...... 6 Figure 2-1 Colour Extraction Methodology and Iris Colour Coordinates ...... 44 Figure 2-2 Distribution of Irises in CIELAB Colour Space ...... 45 Figure 2-3 Visualization of Central Heterochromia ...... 46 Figure 3-1 Distribution of Irises Across the b* and a* (A), a* and L* (B), and b* and L* (C), Coordinates of CIE 1976 L*a*b* (CIELAB) Colour Space for the Second Camera Body ...... 86 Figure 3-2 Sample Iris Wedges with Corresponding Colour Space Values ...... 88 Figure 4-1 Five Surface Features Commonly Found in the Human Iris ...... 120 Figure 4-2 Boundaries of the Iris ...... 121 Figure 4-3 Iris Quadrants ...... 122 Figure 4-4 Pigment Spot Categories ...... 123 Figure 4-5 Contraction Furrow Categories ...... 124 Figure 4-6 Wolfflin Nodule Categories ...... 125 Figure 4-7 Crypt Categories...... 126 Figure 4-8 Self-Described Iris Colour, as Defined by the Fitzpatrick Phototype Scale ...... 127 Figure 4-9 Conjunctival Melanosis Categories...... 128

viii

Forward

This thesis consists of five chapters that explore the global distribution and genetic basis of iris colour and surface features in populations of East Asian, European and South Asian ancestry.

Chapter 1 provides a general introduction to the structure and colour of the human eye. It provides important information about the evolution of iris colour, the anatomy of the iris, the process of melanogenesis and the genetic basis of human eye colour and surface features. This chapter also addresses some of the limitations that are currently present in this field.

Chapter 2 introduces a quantitative method for measuring iris colour variation that was developed using a pilot sample of 205 volunteers of East Asian, European and South Asian ancestry. It also looks at the association between HERC2 rs12913832 and quantitative measures of eye colour in the European and South Asian groups. This work has been previously published in The American Journal of Physical Anthropology (Edwards et al., 2012)

Chapter 3 improves the quantitative method for measuring iris colour that was introduced in Chapter 2. It also looks at the association between 14 putative pigmentation markers and eye colour variation in 1,465 individuals of East Asian, European and South Asian descent. This work has been previously published in Pigment Cell and Melanoma Research (Edwards et al, 2015)

Chapter 4 looks at the global distribution of four surface features (Fuchs’ crypts, contraction furrows, Wolfflin nodules and pigment spots) in the human eye. It also investigates the genetic basis of these traits. This work has been previously published in Royal Society Open Science (Edwards et al., 2016)

Lastly, Chapter 5 summarizes the general conclusions of the preceding articles and highlights some persisting limitations in this field.

Chapter 1: Introduction

1 2

1.1 The Evolution of Global Iris Colour Variation The iris is a complex structure that displays extensive colour variation in human populations. Some eyes even show heterochromia, where multiple colours are found in the same (central or sectoral heterochromia) or different (binocular heterochromia) eyes. The full spectrum of iris colour variation is largely restricted to European populations, although moderate diversity can also be observed in North Africa, the Middle East and parts of South Asia. Throughout the rest of the world, however, iris colour is primarily limited to varying shades of brown. However, even these browns can show extensive diversity and range from light reddish-yellows to dark brownish blacks. The evolution of light eye colour does not appear to be a simple by-product of genetic drift or other random genetic forces. Rather, this trait appears to have been subjected to strong selection pressures. This is supported by the exceptionally high-frequency of blue eyes in some European populations, and the fact that the genetic marker (HERC2 rs12913832) which controls the difference between blue and brown eye colour shows strong signals of positive selection in both modern and ancient sample sets (Eiberg et al., 2008; Mathieson et al., 2015; Nakagome et al., 2015; Wilde et al., 2014). It has thus far been very difficult to date the origin of blue eye- colour. It was initially believed that blue eyes first evolved 6,000-10,000 years ago when humans moved into the Baltic Sea region of Europe (Eiberg et al., 2008). However, this estimate has since lengthened considerably. One recent study suggested that it is unlikely that selection on HERC2 rs12913832 began any more recently than the spread of agriculture 12,800 years ago (Nakagome et al., 2015). A second study, which looked at 230 ancient Eurasian skeletal samples, found that the blue eye colour variant had a frequency of 100% in European Hunter Gatherers dating from 7,500-8,200 years ago (Mathieson et al., 2015). In addition, the distribution HERC2 rs12913832 within other skeletal groups mirrored the regional distribution of this marker today. These findings suggest that selection on blue eyes occurred prior to this time period. The actual selective forces that drove the evolution of light eye colour in European populations are still poorly understood. Even within Europe, eye colour shows a unique distribution where the frequency of blue eyes increases with latitude (Donnelly et al., 2011). Several research groups have suggested that blue eye colour may have initially evolved as a product of frequency-dependent sexual selection. When humans penetrated the steppe-tundra, they encountered harsh climates (Frost, 2006, 2014). Food was primarily acquired by hunting

3 over long distances and this in turn led to a high male mortality rate and a low-operational sex ratio. It is possible that light eye colour provided females with a rare colour advantage that helped them procure mates. Once this trait became established in European populations, it continued to be favored as an indicator of group membership (Wilde et al., 2014). It may have also allowed males to better ascertain paternity confidence. This hypothesis is supported by an excess of HERC2 rs12913832 homozygotes in both modern and ancient populations (Wilde et al., 2014). Alternatively, iris colour may not have been selected for at all. Rather, light eyes may have hitchhiked along with some other trait that is controlled by the same genes (Donnelly et al., 2011; Wilde et al., 2014). HERC2 rs12913832 is a pleiotropic marker that has also been associated with skin pigmentation, hair colour and tanning ability (Cook et al., 2008; Han et al., 2008; Nan et al., 2009a). In addition, it has been found to interact with a number of other genes that play a role in the pigmentation pathway (Branicki et al., 2009; Pospiech et al., 2011). Thus, iris colour variation may simply be the product of selection on other traits, such as skin and/or hair colour. Some researchers have even pointed towards behavioural selection pressures. Most notably, Sturm and Larsson (2009) suggested that blue eyes may help mitigate some of the effects of seasonal affective disorder. It is unlikely that any one hypothesis can fully explain the global diversity that is observed in eye colour today. Rather, the evolution of iris pigmentation variation is likely the product of many complex interacting environmental factors (Mathieson et al., 2015; Wilde et al., 2014).

1.2 Anatomy of the Human Iris 1.2.1 Iris Structure The embryogenesis of the iris is a long and complicated process (see review in Oyster, 1999). The optical vessels first appear during the fourth embryonic week. By the sixth week, a structure known as the pupillary membrane grows in front of the lens. The different components and layers of the iris then begin to differentiate underneath this membrane. At around eight months, the centre of the pupillary membrane is absorbed, forming the pupil. The tissue that remains after resorption resembles a small and jagged circular ring on the surface of the iris. This ring, which is known as the collarette, divides the iris into two different regions: the pupillary

4 zone and the ciliary zone (Figure 1-1). The white region surrounding the iris is known as the sclera. The resulting iris is comprised of five distinct layers (Eagle, 1988; Oyster, 1999) [Figure 1-2]. The posterior-most layer is called the iris pigment epithelium (IPE). This layer is primarily protective and blocks the retina from absorbing excess light (Wilkerson et al., 1996). Attached to the IPE are two muscle layers known as the sphincter muscle and the dilator muscle (Eagle, 1988). These muscles assist in the contraction and dilation of the pupil. Lastly, the two anterior- most layers are the stromal layer and the anterior border layer (Eagle, 1988). Both of these layers contain an assortment of melanocytes, collagen fibres and fibroblasts. However, the anterior border layer is much more densely packed then the stromal layer (Eagle, 1988). The majority of iris colour and structure variation can be attributed to differences in the colour and composition of these two layers (Baranoski and Lam, 2007; Mackey et al., 2011; Rennie, 2012; Sturm and Larsson, 2009). The texture of the iris surface also shows substantial variation. This is largely due to the presence of several surface features that can affect the organization and overall stability of the eye. These features include Fuchs’ crypts, contraction furrows, Wolfflin nodules and pigment spots (Figure 1-1). Fuchs’ crypts are diamond-shaped lacunae in the anterior border and anterior stromal layers of the iris (Purtscher, 1965). They first appear at around 6 months of age and resemble deep pits covering the iris surface (Larsson and Pedersen, 2004). Although there are several different types of lacunae, Fuchs’ crypts are notable because they vary widely in size, are found in the ciliary region and most often originate at the collarette (Larsson and Pedersen, 2004; Purtscher, 1965; Sidhartha et al., 2014a). The developmental factors that control the formation of Fuchs’ crypts are not well-understood. However, it has been suggested that they may be a phylogenetic defect (Purtscher, 1965). Due to the decreasing importance of the anterior layers of the iris over the course of evolutionary history, these layers may have become rudimentary tissues in the anthropoid apes. As a result, they are more prone to acquiring Fuchs’ crypts and other forms of hypoplasia. Interestingly, the extent of Fuchs’ crypts in the iris appears to be correlated with age (Larsson and Pedersen, 2004). This suggests that secondary crypts may also form due to the pushing and pulling of the pupil over time. However, both primary and secondary crypts are ultimately a product of the overall stability and strength of the anterior layers (Purtscher, 1965).

Figure 1-1 Surface Features of the Human Iris This figure shows the surface of a typical human iris. The iris can be divided into two zones that are separated by the collarette (shown in white). The pupillary zone is the region adjacent to the pupil and the ciliary zone encompasses the rest of the iris. There are a number of textural elements that can be observed on the iris surface. Fuchs’ crypts are diamond-shaped lacunae. Wolfflin nodules are small bundles of atrophied collagen. Contraction furrows are discontinuous rings that extend around the outer edge of the ciliary zone. Pigment spots are small accumulations of melanin.

Figure 1-2 Cross-Section of the Human Iris The human iris is composed of five distinct layers. The posterior-most layer is known as the iris pigment epithelium (IPE). This is a melanin-dense layer that does not vary greatly between individuals. Attached to the IPE are two muscle layers, known as the sphincter muscle and the dilator muscle. The anterior-most layers are the stromal layer and the anterior border layer. The majority of variation in iris pigmentation and surface texture between individuals can be attributed to differences in these two layers.

Contraction furrows are thin rings that extend around the outer border of the iris. They resemble ‘wrinkles’ on the surface of the eye and are formed by the contraction and dilation of the pupil (Eagle, 1988; Larsson et al., 2011). Although contraction furrows can extend around the entire iris, each individual furrow is rarely longer than a quadrant in length (Eagle, 1988). Consequently, furrows often look like a series of discontinuous and staggered lines. Very little is known about how contraction furrows affect the overall stability and organization of the iris. However, it is interesting to note that more extensive contraction furrows have been associated with a thicker peripheral iris (Sidhartha et al., 2014a). Wolfflin nodules are round, whitish-yellow collagen deposits that are distributed along the peripheral border of the iris (Donaldson, 1961; Sturm and Larsson, 2009; Williams, 1981). They are formed from the remnants of atrophied collagen fibres in the anterior border and stromal layers. Wolfflin nodules are strongly associated with iris colour. This is exemplified particularly well in one study, which found that Wolfflin nodules had an incidence of 35% in blue-eyed individuals and only 13% in brown-eyed individuals (Donaldson, 1961). Although it was initially believed that Wolfflin nodules are not present in darker irises, they are actually just obscured by pigment in the anterior border layer (Falls, 1970). Wolfflin nodules have received much attention over the years due to their apparent similarity to an iris structure known as Brushfield spots. Brushfield spots are found in over 80% of children who have been diagnosed with Down’s syndrome (Donaldson, 1961; Falls, 1970; Kim et al., 2002; Williams, 1981). They closely resemble Wolfflin nodules in appearance. However, they tend to be slightly smaller and more variable in shape. In addition, whereas Wolfflin nodules are located near the periphery of the iris, Brushfield spots are distributed around the mid-zone area. Many irises show small, discrete areas of brown pigmentation on the iris surface. These areas are known as pigment spots. There are two different types of pigment spots that are commonly found in human eyes: nevi and freckles (Larsson et al., 2011; Rennie, 2012; Sturm and Larsson, 2009). Nevi have a prevalence of 4-6% in individuals of European ancestry (Harbour et al., 2004; Schwab et al., 2015). They are nodular in shape and distort the underlying stromal layer. In contrast, freckles have a prevalence of 48-70% in European populations and have no effect on the stromal architecture (Harbour et al., 2004; Reese, 1944). Both types of pigment spots first appear around the age of six, and gradually increase in frequency over time (Larsson and Pedersen, 2004). Although nevi and freckles affect the stromal layer of the iris in

8 different ways, they are very difficult to distinguish topographically. Consequently, throughout this work I will use the term ‘pigment spot’ to refer to both nevi and freckles. Interestingly, despite their structural differences, nevi and freckles appear to have a similar physiological origin (Harbour et al., 2004). This is because individuals with nevi are also much more likely to have freckles.

1.2.2 Iris Pigmentation Eyes become pigmented due to the presence of a polymer known as melanin. In humans, there are two types of melanin that contribute to normal colour variation: eumelanin and pheomelanin. Eumelanin is a dark brownish-black pigment that is found in high quantities in dark skin, hair and eyes (Hu et al., 2009; Prota et al., 1998; Wakamatsu et al., 2008). In contrast, pheomelanin is a reddish-yellow pigment that is found in higher ratios in light skin, hair and eyes. Apart from their colour, there are a number of biochemical properties that differentiate eumelanin and pheomelanin (Menon et al., 1987; Samokhvalov et al., 2005). Eumelanin is photoprotective and an antioxidant. In contrast, pheomelanin is photoreactive and oxidative. This means that eumelanin is more protective against the damaging effects of ultraviolet radiation. Melanin is synthesized in specialized organelles known as melanosomes, which are located in cells called melanocytes (Eagle, 1988). Melanocytes and melanosomes are both distributed throughout the three non-muscle layers of the iris. However, they take a different form in each of the layers. Melanosomes in the iris pigment epithelium are large, spherical and made up primarily of eumelanin (Imesch et al., 1996; Peles et al., 2009; Prota et al., 1998). Interestingly, the amount of melanin in the iris pigment epithelium appears to show little variation between individuals, irrespective of eye colour (Imesch et al., 1996). In contrast, the melanosomes in the anterior and stromal layers are much smaller and ovoid-shaped (Eagle, 1988; Peles et al., 2009). The synthesis of pheomelanin and eumelanin both begin with the hydroxylation of L- tyrosine to L-dopa, and the oxidation of L-dopa to L-dopaquinone (reviewed in Sturm and Frudakis, 2004). These steps are catalyzed by the enzyme tyrosinase (TYR). After these initial steps, L-dopaquinone can either be converted into cyclodopa (in the case of eumelanin production) or cysteinyldopa (in the case of pheomelanin production) depending on which of two divergent molecular pathways is activated. A membrane bound receptor, known as MC1R, is the

9 primary control switch between pheomelanin and eumelanin synthesis. When MC1R binds to agonist α-MSH (the alpha-melanocyte stimulating hormone), cAMP levels inside the melanosome increase and eumelanin production is stimulated. If MC1R binds to an antagonist known as the agouti signalling protein (ASIP), or if α-MSH is unavailable, pheomelanin synthesis begins. It has been suggested that pheomelanin is created preferentially over eumelanin when the correct compounds are available (Peles et al., 2009). Consequently, most melanin takes the form of a bipolymer with an eumelanic coat covering a pheomelanic core. Apart from the proteins and genes that are directly involved in the melanogenesis pathway there are many other genes and gene-gene interactions that play a role in the synthesis of melanin (Parra, 2007; Sturm and Frudakis, 2004). A large number of membrane transporters (i.e. OCA2, MATP/SLC25A4, and SLC24A4) are required to maintain appropriate ionic conditions and transport substrates into the melanosomes. TYRP1 is needed to stabilize the TYR protein and maintain melanosome structure and transcription factors (i.e. IRF4, MITF) regulate the expression of the different components in the pathway.

1.3 Factors that Influence Iris Pigmentation Variation The primary factor controlling iris pigmentation diversity is the amount and type of melanin in the iris. Multiple studies have found that darker coloured eyes have much higher concentrations of melanin than lighter coloured eyes (Hu et al., 2009; Prota et al., 1998; Wielgus and Sarna, 2005). The eumelanin to pheomelanin ratio also plays a role. Most eyes appear to be primarily eumelanic in nature and contain only trace amounts of pheomelanin. In fact, pheomelanin shows roughly similar concentrations in both blue and brown eyes (Wielgus and Sarna, 2005). Green eyes, however, are unique in that they contain higher concentrations of pheomelanin and very low concentrations of eumelanin (Prota et al., 1998). Although much of iris pigmentation can be explained by the melanin composition of the iris, there are other physiological, biochemical and demographic processes that must also be considered. This is particularly noticeable when looking at intermediate irises, which show melanin concentrations that do not appear to correlate very strongly with colour (Prota et al., 1998; Wielgus and Sarna, 2005). Other contributing factors may include the structure of the

10 anterior border and stromal layers, the distribution of melanosomes in the iris, the presence of iris surface features, and life stage. Interestingly, light blue eyes may contain as much as 40% more melanin than dark blue eyes (Wielgus and Sarna, 2005). There are several explanations for this finding. Blue eye colour is primarily the product of light entering the melanin-poor stromal layer of the eye and bouncing against collagen fibres (Mackey et al., 2011; Rennie, 2012; Sturm and Larsson, 2009). Thus, the thickness of the stroma, as well as the ordering of the collagen fibres, may influence the intensity of blue that is reflected back. The distribution of melanosomes across the anterior layer of the iris may also play some role (Baranoski and Lam, 2007). Computer simulations have found that brown eyes tend to be darker when most of the melanin is present in the anterior border layer (as opposed to the stromal layer). In contrast, blue eyes tend to be darker when most of the melanin is present in the stromal layer. Thus, it is very important to consider the overall structure of the iris when investigating eye colour variation. The presence or absence of surface features also influences iris pigmentation diversity. Many individuals show substantial colour variation within the eye. The most common form of this is central heterochromia, where the pupillary zone is different in colour than the ciliary zone (Bond, 1913; Gladstone, 1969; Rennie, 2012). Central heterochromia can greatly influence perceived eye colour. When I was collecting data for this thesis, a number of participants who believed they had green eyes actually had blue eyes with a brown ring surrounding the pupil. The genetic and physiological factors that control central heterochromia are still largely unknown. However, this condition may result from differences in the thickness of the ciliary and pupillary zones (Imesch et al., 1996). Alternatively, it may be controlled by genetic variants that determine where pigment is placed in the iris (Larsson et al., 2011). Although central heterochromia may be the most visible iris surface feature, the presence of Fuchs’ crypts, pigment spots and the position of the collarette have also been found to influence perceived eye colour (Mackey et al., 2011). Lastly, iris colour has been tentatively associated with demographic factors, like age and sex. A study that followed monozygotic and dizygotic twins from childhood to adulthood noted that 17% of twins showed a change in iris colour with age (Bito et al., 1997). Before the age of six, eye colour was most likely to darken. After the age of six, however, dark eyes tended to become lighter and light eyes tended to become darker. These changes appear to be genetic, as the eye colour of monozygotic twins was more likely to change in the same direction.

Interestingly, eye colour changes are also seen in older adults (Bito et al., 1997; Taylor, 1967). However, in this age group, eyes tends to become lighter. This may be the product of age-related melanin degeneration. There is some evidence that sex may also play a role in iris colour variation. Two recent research studies using Italian and Spanish sample sets found that males were more likely to have blue eyes and females were more likely to have brown/intermediate eyes (Martinez-Cadenas et al., 2013; Pietroni et al., 2014). However, this finding was not replicated in Danish or Swedish samples (Pietroni et al., 2014). Consequently, the association between eye colour and sex may be largely restricted to Mediterranean populations.

1.4 Genetic Basis of Iris Pigmentation Human eye colour has a very high (~98%) heritability (Larsson and Pedersen, 2004). This means that most of the diversity observed in this trait can be explained by variation in the human genome. At present, a large number of polymorphisms have been identified in European populations that are believed to play some role in determining eye colour. Remarkably, the vast majority of iris pigmentation diversity has been attributed to a single locus known as rs12913832 (Eiberg et al., 2008; Kayser et al., 2008; Sturm et al., 2008). This marker, which is located in intron 86 of the gene HERC2, is believed to be responsible for determining the difference between brown and blue eyes. This variant can carry either an adenine (A) or guanine (G) allele. The ‘A’ allele is present at very high frequencies in populations of non-European ancestry, and has been associated with brown iris colour. In contrast, the ‘G’ allele is found at very high frequencies in European populations and has been associated with blue iris colour. It has traditionally been believed that this marker operates in a dominant/recessive manner; individuals who are homozygous for the ‘G’ allele have blue eyes, and individuals who are heterozygous or homozygous for the ‘A’ allele have brown eyes. One research group found that individuals who had the genotype G/G only had a 1% probability of having brown eyes (Sturm et al., 2008). Likewise, individuals who had the genotype A/A had an 80% probability of having brown eyes. A similar study found that 98.3% of individuals with blue eyes were homozygous for the derived allele (Mengel-From et al., 2010). HERC2 rs12913832 is located in a highly conserved region of the genome that contains binding sites for many important regulatory proteins (Sturm et al., 2008; Visser et al., 2012). The

12 marker itself falls in the middle of a binding site for the Helicase-Like Transcription Factor (HLTF), which remodels the chromatin structure around genes. As HERC2 does not appear to play any direct role in the pigmentation pathway, it is unlikely that the gene itself is affected by the mutation. Rather, HERC2 rs12913832 may influence the regulation of the nearby pigmentation gene, OCA2. When the ancestral ‘A’ allele is present, HLTF is able to bind to this region and begin unwinding the chromatin structure. This exposes sites for other regulatory proteins that promote the transcription of OCA2. In contrast, when the derived allele is present, the binding site for HLTF is disrupted. As a result, fewer regulatory proteins are able to access this region and OCA2 transcription is greatly reduced. This hypothesis has been supported by a recent research study that observed cross-linking between HERC2 rs12913832 and OCA2 (Visser et al., 2012). Although HERC2 rs12913832 may play the primary role in controlling the difference between blue and brown iris colour, the lack of complete concordance between HERC2 and iris pigmentation suggests that many other markers may also be involved. Thus far, a number of variants that are believed to play a more subtle role in iris colour diversity have been identified in European populations. A review of these markers can be found below:

I. ASIP. This gene contains a polymorphism known as rs6058017 that results in the exchange of a ‘G’ allele for an ‘A’ allele in the 3’ UTR of exon 4. It has been suggested that this variant may have a negative impact on overall mRNA stability, causing the transcript to degrade prematurely (Kanetsky et al., 2002). The role that this marker plays in iris pigmentation is still under heavy debate. A few early pigmentation studies found that the derived ‘G’ allele was significantly associated with darker skin, eyes and hair (Bonilla et al., 2005; Kanetsky et al., 2002). However, these results have not been consistently replicated (Frudakis et al., 2003; Liu et al., 2010). II. IRF4. The marker rs12203592 has been associated with iris pigmentation variation in multiple studies (Han et al., 2008; Liu et al., 2010). The ancestral cytosine (C) allele appears to be associated with darker skin, hair and eye colour. Not much is known about how this marker functions. However, the derived thymine (T) allele has a very high frequency in Irish populations. This has led some researchers to suggest that this variant may play a role in generating the set of pigmentation characteristics (i.e. red hair, green

eyes, light skin) that is commonly observed in this region of the world (Walsh et al., 2012). III. OCA2. Although many variants have been identified in OCA2, there is only one marker that has been consistently associated with iris pigmentation variation. OCA2 rs1800407 is a non-synonymous mutation that results in the replacement of an arginine by a glutamine at position 419 of the protein. This marker is associated with the penetrance of hazel/green eye colour when HERC2 rs12913832 is either heterozygous or homozygous for the ancestral ‘A’ allele (Mengel-From et al., 2010; Sturm et al., 2008). Remarkably, although OCA2 rs1800407 has a very low frequency in all human populations, it has a relatively strong effect on iris colour. This marker is presently the second strongest predictor of iris pigmentation diversity in European populations (Walsh et al., 2012). IV. SLC24A4. The marker rs12896399 has been associated with iris pigmentation variation in multiple research studies. One study found that this variant was associated with the difference between blue and green eye colour (Sulem et al., 2007). A second found it to be positively associated with the hue and saturation of the iris (Liu et al., 2010). Although not much is known about rs12896399, this marker may fall in the middle of a transcription factor binding site (Hart et al., 2013). Interestingly, a recent study noted that this marker may also be associated with central heterochromia (Larsson et al., 2011). Therefore, rs16891982 may actually function by determining where pigment is placed in the eye. V. SLC24A5. The marker rs1426654 is presently believed to be one of the principal polymorphisms responsible for the lightening of skin pigmentation in populations of European ancestry (Lamason et al., 2005). This marker codes for a non-synonymous mutation that results in the substitution of a threonine for an alanine at position 111 of the protein. The derived allele has become largely fixed in European populations. Therefore, it is difficult to test the association between this variant and iris colour. However, recent studies in an admixed Cape Verdean population and an Italian population suggest that rs1426654 may contribute extensively to eye colour diversity (Beleza et al., 2013; Pietroni et al., 2014). VI. SLC45A2. A marker in this gene known as rs16891982 is believed to have played a large role in evolution of light skin pigmentation in European populations (Graf et al., 2005).

This is a non-synonymous polymorphism that results in the substitution of a leucine for a phenylalanine at position 374 of the protein. The derived allele has been strongly associated with light eye and hair colour in multiple pigmentation studies (Graf et al., 2005; Liu et al., 2010). VII. TYR. Two variants in this gene, rs1393350 and rs1126809, have been associated with iris pigmentation variation in multiple pigmentation studies (Liu et al., 2010; Sulem et al., 2007). Both of these markers are located very close together in the genome and are in strong linkage disequilibrium. Therefore, it is likely that they have the same effect on eye colour variation. TYR rs1126809 is most likely the causative variant, as it is a non- synonymous mutation that results in the replacement of an arginine by a glutamine at position 402 of the protein (Sulem et al., 2007). Interestingly, although neither of these variants show any signals of positive selection, they are largely restricted to European populations. VIII. TYRP1. A variant in this gene, known as rs1408799, has been significantly associated with iris colour variation in several research studies. One study found that this marker was associated with the difference between blue and non-blue eyes (Frudakis et al., 2003). A second found that it played a role in controlling the hue and saturation of the iris (Liu et al., 2010). TYRP1 rs1408799 shows strong signals of positive selection in European populations (Myles et al., 2006). However, its function is still largely unknown. IX. Other Genes. A recent genome-wide association study on quantitative iris colour variation identified several new markers that may contribute to eye colour diversity (Liu et al., 2010). These include rs7277820 in DSCR9, rs3768056 in LYST, and rs9894429 in NPLOC4. However, these markers have not yet been replicated in any additional studies. Interestingly, none of these genes appear to be directly involved in the pigmentation pathway. Rather, they may play some role in regulating the overall structure and function of the iris.

1.5 Genetic Basis of Iris Surface Features The heritability of Fuchs’ crypts (66%), contraction furrows (78%), Wolfflin nodules (78%) and pigment spots (58%) is moderately high (Larsson and Pedersen, 2004). However, at

15 present, the genetic basis of these traits is still very poorly understood. Thus far, only one research group has attempted to identify the variants in the human genome that are directly associated with iris surface feature diversity (Larsson et al., 2011; Sturm and Larsson, 2009). It appears that the markers that determine iris structure may be very different from those that determine eye colour (Larsson et al., 2011). Whereas most of the genes associated with iris pigmentation code for proteins directly involved in the pathway of melanogenesis, the markers that have been associated with iris surface features are primarily responsible for the embryogenesis and development of the eye. A variant known as rs10235789 in the gene SEMA3A has been the most strongly associated with Fuchs’ crypts (Larsson et al., 2011). This marker, which was identified in an Australian population of European ancestry, appears to explain approximately 1.5% of the overall variation in this trait. It has been suggested that the derived ‘C’ allele may negatively affect the overall stability of the anterior layers of the iris, leading to an increased frequency of crypts. The variant rs3739070 in TRAF3IP1 has been strongly associated with contraction furrows (Larsson et al., 2011). This variant, which explains about 1.7% of the variation in furrows, may be linked to the overall thickness and density of the iris. Lastly, the polymorphism rs11630290 in HERC1 has been associated with the presence of pigment spots (Larsson et al., 2011). Although the function of HERC1 is largely unknown, it may play some role in membrane transport. It has long been suggested that the Downs Syndrome Critical Region (DSCR) of the genome may hold the genetic variants responsible for Wolfflin nodules (Sturm and Larsson, 2009). This is largely because of the perceived similarity between Wolfflin nodules and Brushfield spots. Interestingly, in a recent genome-wide association study on iris colour variation, a variant known as rs7277820 in DSCR9 was significantly associated with multiple iris colour characteristics (Liu et al., 2010). It was suggested that this variant may modulate iris pigmentation by controlling the presence and extension of Wolfflin nodules in the eye. However, the relationship between DSCR9 rs7277820 and Wolfflin nodules has not yet been tested. At present, the genetic basis of iris surface features is still very poorly understood. The genetic markers that have been identified thus far explain a very small percentage of the overall variation in in these traits. Given the moderate heritability of all four iris features, there are probably many as of yet unidentified markers involved.

1.6 Limitations in our Understanding of Iris Colour and Texture Variation

Although both iris colour and structure variation have begun to receive much attention over the past decade, there are still many gaps and limitations in our understanding of these traits. The most significant of these limitations stem from difficulties associated with obtaining quantitative measurements of iris colour, the complexities of characterizing central heterochromia, and the limited population groups that have traditionally been studied.

1.6.1 Characterizing Iris Colour and Surface Feature Variation Skin colour and hair colour have conventionally been measured using a technique known as reflectance spectroscopy (reviewed in Shriver and Parra, 2000). Spectroscopy involves shining light at a particular wavelength on the skin or hair, and measuring the amount of light that is reflected back. Due to the nature of the eye, spectroscopy cannot be used on the iris. Consequently, early iris pigmentation studies chose to characterize iris colour using sets of discrete categories (i.e. blue, green, brown) [Frudakis et al., 2003; Graf et al., 2005; Mengel- From et al., 2009; Rebbeck et al., 2002; Sulem et al., 2007]. Such categorical classification methods have severe limitations. Iris colour is an innately quantitative trait. Although discrete categories may capture gross differences in iris colour (i.e. blue versus brown), they are unable to capture the full range of eye colour variation. In addition, there has been no formal consensus on how many categories should be used. Some research studies use as few as three categories, while others use as many as eight. Thus, it is not surprising that the inter- and intra-observer reliability of categorical methods is not very high (Frudakis et al., 2003; Seddon et al., 1990). Over the past decade, a number of research groups have begun to develop quantitative methods of measuring iris colour. Liu et al. (2010) designed an automated system to extract hue and saturation values from photographs of the iris. Saturation was used to represent the amount of melanin in the eye, and hue was used to represent the type. Beleza et al. (2013) extracted RGB values from digital photographs of the eye. They used the term ‘T-index’ to describe the distribution of irises across RGB colour space. Most recently, Andersen et al. (2013) automatically extracted a Pixel Index of the Eye (PIE) score from photographs of the eye. This score, which ranges from -1 to 1, is calculated by comparing the number of pixels labelled blue

17 with the number of pixels labelled brown. Such quantitative methods offer numerous advantages. For one, they are able to pick up much more subtle variation in pigmentation variation. As a result, they have been very successful at identifying several of the genetic variants that play a smaller role in controlling iris colour variation. In addition, they allow researchers to approach the study of iris colour in populations with more homogenously coloured eyes. The distribution of iris surface features is also remarkably complex. Fuchs’ crypts vary considerably in number, shape and size. Some eyes show many clear and well-demarcated crypts, whereas others show indistinct and sparsely distributed crypts. It is also not uncommon for an iris to be completely devoid of crypts, or only show small, bulbous-shaped crypts scattered around the collarette. Pigment spots can be either large or small. They differ widely in colour and can range from light yellowish-browns to dark brownish-blacks. Some pigment spots show clear boundaries, whereas others are not very well-defined. Contraction furrows may extend around the entire iris, or be restricted to particular quadrants of the eye. Some eyes show numerous strata of furrows, whereas others have only one or two. Wolfflin nodules differ between individuals in both number, size and colour. Some irises show large, nodules whereas others eyes show numerous, small nodules. In some eyes they are a pale white, whereas in other eyes they appear orange or yellow. Due to the extensive diversity inherent in iris surface features, it has traditionally been very difficult to characterize these traits. Consequently, most research groups have used categorical schema that attempt to capture the most important properties of surface feature variation. However, there has been very little standardization across these methods. Research studies have used anywhere between three and seven categories to characterize iris texture (Larsson et al., 2011; Quillen et al., 2011; Sidhartha et al., 2014a). What this means is that the results obtained from different groups are not comparable. Accordingly, the overall frequency and distribution of these traits is still poorly understood.

1.6.2 Characterizing Central Heterochromia Central heterochromia has thus far been one of the most poorly studied areas of iris research. Thus, it is not surprising that we currently know very little about the genetic basis and global distribution of this trait.

Early categorical methods tended to group irises that showed evidence of heterochromia into a broad ‘intermediate’ category (Graf et al., 2005; Rebbeck et al., 2002; Sulem et al., 2007). When heterochromia was studied on its own, it was characterized using a ‘presence’ or ‘absence’ schema (Larsson et al., 2011). However, this greatly simplifies the overall variation present in heterochromatic irises. Some irises with heterochromia show a gross colour difference between the ciliary and pupillary zones, whereas others show a much more subtle difference. Just as iris pigmentation variation is quantitative, so is heterochromatic variation. Current quantitative methods have not improved the study of central heterochromia. Although such methods are capable of picking up subtle colour differences, they tend to use a single ‘average’ value to capture eye colour. This means that an iris that is uniformly green might give the same average colour value as a blue iris with a heterochromatic ring. The difficulties associated with measuring central heterochromia in the iris have had a significant impact on our overall understanding of iris colour variation. At present, a number of forensic groups have developed algorithms that can be used to predict eye colour from small samples of DNA (Hart et al., 2013; Spichenok et al., 2011; Walsh et al., 2012). The most well- known of these algorithms is perhaps IrisPlex, which uses six SNPs (HERC2 rs12913832, SLC45A2 rs16891982, IRF4 rs12203592, OCA2 rs1800407, SLC24A4 rs12896399, TYR rs1393350) to differentiate between blue, brown or intermediate/green eye colour (Walsh et al., 2012). Although IrisPlex has been found to work well in populations with a high frequency of blue or brown eyes, it does poorly in populations with a high frequency of intermediate eyes. In a sample of 803 individuals of mixed ancestry from New York City, IrisPlex showed 209 inconclusive results and 135 errors (Pneuman et al., 2012). It is clear we need to develop more effective methods of characterizing central heterochromia.

1.6.3 Population Variation One of the biggest gaps in our current understanding of iris colour variation is the lack of attention that has been devoted to populations of non-European ancestry. Although brown eye colour may predominate in regions outside of Northern Europe, there is still substantial variation within these browns. In fact, brown iris colour can range from light reddish-yellows to dark brownish-blacks. The lack of research being conducted outside of European populations may be largely due to the categorical methods that have been traditionally use to characterize iris colour.

As such methods can only pick up gross differences in eye colour, they are largely ineffective in populations with more homogenously coloured irises. However, even with the advent of new quantitative methodologies, there has been a noticeable lack of research being performed in populations of non-European ancestry. This dearth of research is problematic for a number of reasons. Firstly, there may be many genetic variants associated with iris pigmentation variation that cannot be identified in European populations. Such variants may either be fixed in Europe, or not present at all. In addition, as the sample size of brown eyes is moderately low in Northern Europe, it may be more difficult to detect variants that play a role in controlling brown iris colour. The lack of iris pigmentation research outside of Europe also means that we have a very poor understanding of the distribution of iris colour in these regions. Iris colour in non-European populations is typically described using the blanket term ‘brown’. However, is it possible that the shade and intensity of these browns may differ from one population to the next? Is the brown iris colour found in South Asian populations the same as the brown iris colour found in East Asian populations? Are the genetic variants that control brown iris colour variation the same in all populations? These are questions that we are still unable to answer. Finally, it is very difficult to trace the evolutionary history of a trait when we only know the genetic basis of that trait in one particular region of the world. At the moment, the evolutionary factors that drove the evolution of light eye colour are poorly understood. Looking at the distribution of genetic variants associated with iris pigmentation in other populations may provide some insights into the evolution of this trait. In contrast to iris colour research, there have been some attempts to understand iris structure diversity in different populations. At present, there have been extensive studies performed in East Asian, European, Portuguese, Brazilian and Cape Verdean populations (Larsson and Pedersen, 2004; Larsson et al., 2011; Quillen et al., 2011; Sidhartha et al., 2014a, 2014a). However, all of these studies have limited their work to a particular continent. As each research group categorized iris structure variation differently, we are still not able to compare the phenotypic distribution of these traits in different continental regions of the world. There is some evidence that the distribution of the textural features of the iris may be population-specific (Qiu et al., 2005; Quillen et al., 2011). However, until different ancestral groups are compared using

20 the same methodology, we will not be able to fully understand the global distribution of these traits.

1.7 Rationale and Study Goals Developing a better understanding of iris colour and surface features will contribute greatly to the fields of forensics, anthropology and public health. At present, many forensics groups are developing pigmentation predictor algorithms. The primary purpose of these algorithms is to allow crime scene investigators to piece together the hair, skin and eye colour of unidentified individuals from small samples of their DNA. IrisPlex, which was discussed briefly in the previous section, is perhaps the most well-known of these systems (Walsh et al., 2012). However, other algorithms that rely on different sets of genetic variants have also been proposed (Hart et al., 2013; Spichenok et al., 2011). The use of iris colour predictor equations has far- reaching effects outside of the field of forensics. Anthropologists and molecular biologists are now able to piece together the pigmentation characteristics of early humans and their ancestors. For example, genetic variants in the MC1R gene have been used to show that Neanderthals most likely had red hair and pale skin (Lalueza-Fox et al., 2007). Likewise, HERC2 rs12913832, the polymorphism that controls the difference between blue and brown eyes, has been used to predict the iris colour of ancient skeletal remains (Keyser et al., 2009; Olalde et al., 2014). Such research will ultimately provide us with important information about the evolution and early migratory history of human populations. Apart from the anthropological and forensic benefits, there is some evidence that both iris colour and structure may have an influence on individual health and well-being. Many of the markers associated with iris colour variation have also been associated with an increased risk of basal cell carcinoma, squamous cell carcinoma, and/or melanoma (Gudbjartsson et al., 2008; Nan et al., 2009b). Understanding how these markers function may provide important information about the distribution and prevalence of such carcinomas in different global populations. Similarly, there appears to be an association between several ocular disorders and eye colour. Within populations, for instance, individuals with lighter coloured irises are more likely to develop age-related macular degeneration (AMD) than individuals with darker coloured irises (Frank et al., 2000; Wakamatsu et al., 2008). The surface features found on the iris may

21 also serve as important risk indicators of a number of diseases and disorders. Two recent studies in a Malaysian population living in Singapore found that increasing grades of Fuchs’ crypts are associated with a thinner iris and a wider angle closure and more extended contraction furrows are associated with a thicker peripheral iris (Sidhartha et al., 2014a, 2014b). As iris thickness and angle closure are both linked to a number of ocular disorders, the use of photographs to predict these traits may represent a cost-effective way of evaluating overall iris structure. It is clear that iris colour and structure have the potential to contribute greatly to a number of different fields. However, we cannot fully use these traits until we overcome some of the numerous limitations in these fields. How do we measure iris colour quantitatively? How do we take central heterochromia into account when using such quantitative methods? What categories best characterize iris structure across global populations? To what extent do the genes involved in iris colour and texture variation overlap in different populations? Do the surface features on the iris show the same frequencies in populations of different ancestry? These are questions that need to be answered before we can begin to fully understand the evolutionary events that led to the iris colour variation that we see in the world today. With that in mind, my research has six specific goals:

1) To develop a quantitative method of measuring iris colour from high-resolution photographs 2) To develop a quantitative method of investigating central heterochromia 3) To develop a categorical method of characterizing Fuchs’ crypts, contraction furrows, Wolfflin nodules, and pigment spots from high-resolution photographs of the iris 4) To look at the phenotypic distribution of both iris colour and surface features in populations of East Asian, European and South Asian ancestry 5) To look at the association between putative genetic markers and iris colour and texture variation in populations of East Asian, European and South Asian ancestry

In summary, this research will look at the distribution and genetic basis of iris colour and surface feature variation in three different populations. It will also provide other researchers with a methodology that they can use to look at iris colour and structure in their own sample sets. This will have many benefits for the fields of forensics, anthropology and public health.

1.8 References

Andersen, J.D., Johansen, P., Harder, S., Christoffersen, S.R., Delgado, M.C., Henriksen, S.T., Nielsen, M.M., Sørensen, E., Ullum, H., Hansen, T., et al. (2013). Genetic analyses of the human eye colours using a novel objective method for eye colour classification. Forensic Sci. Int. Genet. 7, 508–515. Baranoski, G.V.G., and Lam, M.W.Y. (2007). Qualitative assessment of undetectable melanin distribution in lightly pigmented irides. J. Biomed. Opt. 12, 1–3. Beleza, S., Johnson, N.A., Candille, S.I., Absher, D.M., Coram, M.A., Lopes, J., Campos, J., Araújo, I.I., Anderson, T.M., Vilhjálmsson, B.J., et al. (2013). Genetic architecture of skin and eye color in an African-European admixed Population. PLoS Genet 9, e1003372. Bito, L.Z., Matheny, A., Cruickshanks, K.J., Nondahl, D.M., and Carino, O.B. (1997). Eye color changes past early childhood: The Louisville Twin Study. Arch. Ophthalmol. 115, 659–663. Bond, C. (1913). On heterochromia iridis in man and animals from the genetic point of view. Mol. Gen. Genet. 9, 344–346. Bonilla, C., Boxill, L.-A., Mc Donald, S.A., Williams, T., Sylvester, N., Parra, E.J., Dios, S., Norton, H.L., Shriver, M.D., and Kittles, R.A. (2005). The 8818G allele of the agouti signaling protein (ASIP) gene is ancestral and is associated with darker skin color in African Americans. Hum. Genet. 116, 402–406. Branicki, W., Brudnik, U., and Wojas-Pelc, A. (2009). Interactions between HERC2, OCA2 and MC1R may influence human pigmentation phenotype. Ann. Hum. Genet. 73, 160–170. Cook, A.L., Chen, W., Thurber, A.E., Smit, D.J., Smith, A.G., Bladen, T.G., Brown, D.L., Duffy, D.L., Pastorino, L., Bianchi-Scarra, G., et al. (2008). Analysis of cultured human melanocytes based on polymorphisms within the SLC45A2/MATP, SLC24A5/NCKX5, and OCA2/P loci. J Invest Dermatol 129, 392–405. Donaldson, D.D. (1961). The significance of spotting of the iris in mongoloids: Brushfield’s spots. Arch. Ophthalmol. 65, 26–31. Donnelly, M.P., Paschou, P., Grigorenko, E., Gurwitz, D., Barta, C., Lu, R.-B., Zhukova, O.V., Kim, J.-J., Siniscalco, M., New, M., et al. (2011). A global view of the OCA2-HERC2 region and pigmentation. Hum. Genet. 131, 683–696. Eagle, R.C. (1988). Iris pigmentation and pigmented lesions: an ultrastructural study. Trans. Am. Ophthalmol. Soc. 86, 581–687. Eiberg, H., Troelsen, J., Nielsen, M., Mikkelsen, A., Mengel-From, J., Kjaer, K.W., and Hansen, L. (2008). Blue eye color in humans may be caused by a perfectly associated founder mutation in a regulatory element located within the HERC2 gene inhibiting OCA2 expression. Hum. Genet. 123, 177–187.

Falls, H.F. (1970). Ocular changes in mongolism. Ann. N. Y. Acad. Sci. 171, 627–636. Frank, R.N., Puklin, J.E., Stock, C., and Canter, L.A. (2000). Race, iris color, and age-related macular degeneration. Trans. Am. Ophthalmol. Soc. 98, 109–117. Frost, P. (2006). European hair and eye color: a case of frequency-dependent sexual selection? Evol. Hum. Behav. 27, 85–103. Frost, P. (2014). The puzzle of European hair, eye, and skin color. Adv. Anthropol. 4, 78–88. Frudakis, T., Thomas, M., Gaskin, Z., Venkateswarlu, K., Chandra, K.S., Ginjupalli, S., Gunturi, S., Natrajan, S., Ponnuswamy, V.K., and Ponnuswamy, K.N. (2003). Sequences associated with human iris pigmentation. Genetics 165, 2071–2083. Gladstone, R.M. (1969). Development and significance of heterochromia of the iris. Arch. Neurol. 21, 184–192. Graf, J., Hodgson, R., and Van Daal, A. (2005). Single nucleotide polymorphisms in the MATP gene are associated with normal human pigmentation variation. Hum. Mutat. 25, 278–284. Gudbjartsson, D.F., Sulem, P., Stacey, S.N., Goldstein, A.M., Rafnar, T., Sigurgeirsson, B., Benediktsdottir, K.R., Thorisdottir, K., Ragnarsson, R., and Sveinsdottir, S.G. (2008). ASIP and TYR pigmentation variants associate with cutaneous melanoma and basal cell carcinoma. Nat. Genet. 40, 886–891. Han, J., Kraft, P., Nan, H., Guo, Q., Chen, C., Qureshi, A., Hankinson, S.E., Hu, F.B., Duffy, D.L., Zhao, Z.Z., et al. (2008). A genome-wide association study identifies novel alleles associated with hair color and skin pigmentation. PloS Genet. 4, e1000074. Harbour, J., Brantley, M., Hollingsworth, H., and Gordon, M. (2004). Association between posterior uveal melanoma and iris freckles, iris naevi, and choroidal naevi. Br. J. Ophthalmol. 88, 36–38. Hart, K.L., Kimura, S.L., Mushailov, V., Budimlija, Z.M., Prinz, M., and Wurmbach, E. (2013). Improved eye-and skin-color prediction based on 8 SNPs. Croat. Med. J. 54, 248–256. Hu, D.N., Wakamatsu, K., Ito, S., and McCormick, S.A. (2009). Comparison of eumelanin and pheomelanin content between cultured uveal melanoma cells and normal uveal melanocytes. Melanoma Res. 19, 75–79. Imesch, P.D., Bindley, C.D., Khademian, Z., Ladd, B., Gangnon, R., Albert, D.M., and Wallow, I.H. (1996). Melanocytes and iris color: electron microscopic findings. Arch. Ophthalmol. 114, 443–447. Kanetsky, P.A., Swoyer, J., Panossian, S., Holmes, R., Guerry, D., and Rebbeck, T.R. (2002). A polymorphism in the agouti signaling protein gene is associated with human pigmentation. Am. J. Hum. Genet. 70, 770–775. Kayser, M., Liu, F., Janssens, A.C.J.W., Rivadeneira, F., Lao, O., Duijn, K. van, Vermeulen, M., Arp, P., Jhamai, M.M., IJcken, W.F.J. van, et al. (2008). Three genome-wide association studies

24 and a linkage analysis identify HERC2 as a human iris color gene. Am. J. Hum. Genet. 82, 411– 423. Keyser, C., Bouakaze, C., Crubézy, E., Nikolaev, V.G., Montagnon, D., Reis, T., and Ludes, B. (2009). Ancient DNA provides new insights into the history of south Siberian Kurgan people. Hum. Genet. 126, 395–410. Kim, J., Hwang, J., Kim, H., and Yu, Y. (2002). Characteristic ocular findings in Asian children with Down syndrome. Eye 16, 710–714. Lalueza-Fox, C., Römpler, H., Caramelli, D., Stäubert, C., Catalano, G., Hughes, D., Rohland, N., Pilli, E., Longo, L., and Condemi, S. (2007). A melanocortin 1 receptor allele suggests varying pigmentation among Neanderthals. Science 318, 1453–1455. Lamason, R.L., Mohideen, M.-A.P., Mest, J.R., Wong, A.C., Norton, H.L., Aros, M.C., Jurynec, M.J., Mao, X., Humphreville, V.R., and Humbert, J.E. (2005). SLC24A5, a putative cation exchanger, affects pigmentation in zebrafish and humans. Science 310, 1782–1786. Larsson, M., and Pedersen, N. (2004). Genetic correlations among texture characteristics of the human iris. Mol Vis 10, 821–831. Larsson, M., Duffy, D.L., Zhu, G., Liu, J.Z., Macgregor, S., McRae, A.F., Wright, M.J., Sturm, R.A., Mackey, D.A., Montgomery, G.W., et al. (2011). GWAS findings for human iris patterns: Associations with variants in genes that influence normal neuronal pattern development. Am. J. Hum. Genet. 89, 334–343. Liu, F., Wollstein, A., Hysi, P.G., Ankra-Badu, G.A., Spector, T.D., Park, D., Zhu, G., Larsson, M., Duffy, D.L., Montgomery, G.W., et al. (2010). Digital quantification of human eye color highlights genetic association of three new loci. PLoS Genet 6, e1000934. Mackey, D.A., Wilkinson, C.H., Kearns, L.S., and Hewitt, A.W. (2011). Classification of iris colour: review and refinement of a classification schema. Clin. Experiment. Ophthalmol. 39, 462–471. Martinez-Cadenas, C., Peña-Chilet, M., Ibarrola-Villava, M., and Ribas, G. (2013). Gender is a major factor explaining discrepancies in eye colour prediction based on HERC2/OCA2 genotype and the IrisPlex model. Forensic Sci. Int. Genet. 7, 453–460. Mathieson, I., Lazaridis, I., Rohland, N., Mallick, S., Patterson, N., Roodenberg, S.A., Harney, E., Stewardson, K., Fernandes, D., Novak, M., et al. (2015). Genome-wide patterns of selection in 230 ancient Eurasians. Nature 528, 499–503. Mengel-From, J., Wong, T.H., Morling, N., Rees, J.L., and Jackson, I.J. (2009). Genetic determinants of hair and eye colours in the Scottish and Danish populations. BMC Genet. 10, 1– 13. Mengel-From, J., Børsting, C., Sanchez, J.J., Eiberg, H., and Morling, N. (2010). Human eye colour and HERC2, OCA2 and MATP. Forensic Sci. Int. Genet. 4, 323–328.

Menon, I., Basu, P., Persad, S., Avaria, M., Felix, C., and Kalyanaraman, B. (1987). Is there any difference in the photobiological properties of melanins isolated from human blue and brown eyes? Br. J. Ophthalmol. 71, 549–552. Myles, S., Somel, M., Tang, K., Kelso, J., and Stoneking, M. (2006). Identifying genes underlying skin pigmentation differences among human populations. Hum. Genet. 120, 613–621. Nakagome, S., Alkorta-Aranburu, G., Amato, R., Howie, B., Peter, B.M., Hudson, R.R., and Di Rienzo, A. (2015). Estimating the ages of selection signals from different epochs in human history. Mol. Biol. Evol, msv256. Nan, H., Kraft, P., Qureshi, A.A., Guo, Q., Chen, C., Hankinson, S.E., Hu, F.B., Thomas, G., Hoover, R.N., Chanock, S., et al. (2009a). Genome-wide association study of tanning phenotype in a population of European ancestry. J. Invest. Dermatol. 129, 2250–2257. Nan, H., Kraft, P., Hunter, D.J., and Han, J. (2009b). Genetic variants in pigmentation genes, pigmentary phenotypes, and risk of skin cancer in Caucasians. Int. J. Cancer 125, 909–917. Olalde, I., Allentoft, M.E., Sánchez-Quinto, F., Santpere, G., Chiang, C.W., DeGiorgio, M., Prado-Martinez, J., Rodríguez, J.A., Rasmussen, S., and Quilez, J. (2014). Derived immune and ancestral pigmentation alleles in a 7,000-year-old Mesolithic European. Nature 507, 225–228. Oyster, C.W. (1999). The human eye: structure and function (Sinauer Associates). Parra, E.J. (2007). Human pigmentation variation: evolution, genetic basis, and implications for public health. Am. J. Phys. Anthropol. 134, 85–105. Peles, D.N., Hong, L., Hu, D.-N., Ito, S., Nemanich, R.J., and Simon, J.D. (2009). Human iridal stroma melanosomes of varying pheomelanin contents possess a common eumelanic outer surface. J. Phys. Chem. B 113, 11346–11351. Pietroni, C., Andersen, J.D., Johansen, P., Andersen, M.M., Harder, S., Paulsen, R., Børsting, C., and Morling, N. (2014). The effect of gender on eye colour variation in European populations and an evaluation of the IrisPlex prediction model. Forensic Sci. Int. Genet. 11, 1–6. Pneuman, A., Budimlija, Z.M., Caragine, T., Prinz, M., and Wurmbach, E. (2012). Verification of eye and skin color predictors in various populations. Leg. Med. 14, 78–83. Pospiech, E., Draus-Barini, J., Kupiec, T., Wojas-Pelc, A., and Branicki, W. (2011). Gene-gene interactions contribute to eye colour variation in humans. J Hum Genet 56, 447–455. Prota, G., Hu, D.-N., Vincensi, M.R., McCormick, S.A., and Napolitano, A. (1998). Characterization of melanins in human irides and cultured uveal melanocytes from eyes of different colors. Exp. Eye Res. 67, 293–299. Purtscher, E. (1965). On the development and morphology of iris crypts*. Acta Ophthalmol. (Copenh.) 43, 109–119. Qiu, X., Sun, Z., and Tan, T. (2005). Global texture analysis of iris images for ethnic classification. In Advances in Biometrics, (Springer), pp. 411–418.

Quillen, E.E., Guiltinan, J.S., Beleza, S., Rocha, J., Pereira, R.W., and Shriver, M.D. (2011). Iris texture traits show associations with iris color and genomic ancestry. Am. J. Hum. Biol. 23, 567– 569. Rebbeck, T.R., Kanetsky, P.A., Walker, A.H., Holmes, R., Halpern, A.C., Schuchter, L.M., Elder, D.E., and Guerry, D. (2002). P gene as an inherited biomarker of human eye color. Cancer Epidemiol. Biomarkers Prev. 11, 782–784. Reese, A.B. (1944). Pigment freckles of the iris (benign melanomas): their significance in relation to malignant melanoma of the uvea. Am. J. Ophthalmol. 27, 217–226. Rennie, I. (2012). Don’t it make my blue eyes brown: heterochromia and other abnormalities of the iris. Eye 26, 29–50. Samokhvalov, A., Hong, L., Liu, Y., Garguilo, J., Nemanich, R.J., Edwards, G.S., and Simon, J.D. (2005). Oxidation potentials of human eumelanosomes and pheomelanosomes. Photochem. Photobiol. 81, 145–148. Schwab, C., Zalaudek, I., Mayer, C., Riedl, R., Wackernagel, W., Juch, H., Aigner, B., Brunasso, A.M., Langmann, G., and Richtig, E. (2015). New insights into oculodermal nevogenesis and proposal for a new iris nevus classification. Br. J. Ophthalmol. 99, 644–649. Seddon, J.M., Sahagian, C.R., Glynn, R.J., Sperduto, R., Gragoudas, E., Group, E.D.C.-C.S., and others (1990). Evaluation of an iris color classification system. Invest Ophthalmol Vis Sci 31, 1592–1598. Shriver, M.D., and Parra, E.J. (2000). Comparison of narrow-band reflectance spectroscopy and tristimulus colorimetry for measurements of skin and hair color in persons of different biological ancestry. Am. J. Phys. Anthropol. 112, 17. Sidhartha, E., Gupta, P., Liao, J., Tham, Y.-C., Cheung, C.Y., He, M., Wong, T.Y., Aung, T., and Cheng, C.-Y. (2014a). Assessment of iris surface features and their relationship with iris thickness in Asian eyes. Ophthalmology 121, 1007–1012. Sidhartha, E., Nongpiur, M.E., Cheung, C.Y., He, M., Wong, T.Y., Aung, T., and Cheng, C.-Y. (2014b). Relationship between iris surface features and angle width in Asian eyes. Invest. Ophthalmol. Vis. Sci. 55, 8144–8148. Spichenok, O., Budimlija, Z.M., Mitchell, A.A., Jenny, A., Kovacevic, L., Marjanovic, D., Caragine, T., Prinz, M., and Wurmbach, E. (2011). Prediction of eye and skin color in diverse populations using seven SNPs. Forensic Sci. Int. Genet. 5, 472–478. Sturm, R.A., and Frudakis, T.N. (2004). Eye colour: portals into pigmentation genes and ancestry. Trends Genet. 20, 327–332. Sturm, R.A., and Larsson, M. (2009). Genetics of human iris colour and patterns. Pigment Cell Melanoma Res. 22, 544–562.

Sturm, R.A., Duffy, D.L., Zhao, Z.Z., Leite, F.P.N., Stark, M.S., Hayward, N.K., Martin, N.G., and Montgomery, G.W. (2008). A single SNP in an evolutionary conserved region within intron 86 of the HERC2 gene determines human blue-brown eye color. Am. J. Hum. Genet. 82, 424– 431. Sulem, P., Gudbjartsson, D.F., Stacey, S.N., Helgason, A., Rafnar, T., Magnusson, K.P., Manolescu, A., Karason, A., Palsson, A., Thorleifsson, G., et al. (2007). Genetic determinants of hair, eye and skin pigmentation in Europeans. Nat Genet 39, 1443–1452. Taylor, A.I. (1967). Eye colour in the Tristan da Cunha population. Hum. Biol. 39, 316–318. Visser, M., Kayser, M., and Palstra, R.-J. (2012). HERC2 rs12913832 modulates human pigmentation by attenuating chromatin-loop formation between a long-range enhancer and the OCA2 promoter. Genome Res. 22, 446–455. Wakamatsu, K., Hu, D., McCormick, S.A., and Ito, S. (2008). Characterization of melanin in human iridal and choroidal melanocytes from eyes with various colored irides. Pigment Cell Melanoma Res. 21, 97–105. Walsh, S., Wollstein, A., Liu, F., Chakravarthy, U., Rahu, M., Seland, J.H., Soubrane, G., Tomazzoli, L., Topouzis, F., and Vingerling, J.R. (2012). DNA-based eye colour prediction across Europe with the IrisPlex system. Forensic Sci. Int. Genet. 6, 330–340. Wielgus, A.R., and Sarna, T. (2005). Melanin in human irides of different color and age of donors. Pigment Cell Res. 18, 454–464. Wilde, S., Timpson, A., Kirsanow, K., Kaiser, E., Kayser, M., Unterländer, M., Hollfelder, N., Potekhina, I.D., Schier, W., Thomas, M.G., et al. (2014). Direct evidence for positive selection of skin, hair, and eye pigmentation in Europeans during the last 5,000 y. Proc. Natl. Acad. Sci. 111, 4832–4837. Wilkerson, C.L., Syed, N.A., Fisher, M.R., Robinson, N.L., and Albert, D.M. (1996). Melanocytes and iris color: light microscopic findings. Arch. Ophthalmol. 114, 437–442. Williams, R.D. (1981). Brushfield spots and Wolfflin nodules in the iris: an appraisal in handicapped children. Dev. Med. Child Neurol. 23, 646–650.

Chapter 2: Quantitative Measures of Iris Colour Using High Resolution Photographs

Author Contributions: E.J. Parra and I designed this project. A. Gozdzik carried out the participant recruitment and data collection. J. Miles created the macro that automatically cropped a square from each iris photograph. I prepared the samples for genotyping with help from K. Ross. I prepared the iris photographs, analysed the data, developed the methodology and wrote the manuscript with help from E. J Parra.

This Chapter Has Been Previously Published As: Edwards, M., Gozdzik, A., Ross, K., Miles J., and Parra, E.J. (2012) Technical note: quantitative measures of iris colour using high resolution photographs. Am. J. Phys. Anthro. 147, 141-149.

Acknowledgements: The authors thank all the individuals who participated in the study. Jon Miles is the owner of Miles Research, a company that designs and builds specialized iris cameras.

28 29

2.1 Abstract Our understanding of the genetic architecture of iris colour is still limited. This is partly related to difficulties associated with obtaining quantitative measurements of eye colour. Here we introduce a new automated method for measuring iris colour using high resolution photographs. This method extracts colour measurements in the CIE 1976 L*a*b* (CIELAB) colour space from a 256 by 256 pixel square sampled from the 9:00 meridian of the iris. Colour is defined across three dimensions: L* (the lightness coordinate), a* (the red-green coordinate), and b* (the blue-yellow coordinate). We applied this method to a sample of individuals of diverse ancestry (East Asian, European and South Asian) that was genotyped for the HERC2 rs12913832 polymorphism, which is strongly associated with blue eye colour. We identified substantial variation in the CIELAB colour space, not only in the European sample, but also in the East Asian and South Asian samples. As expected, rs12913832 was significantly associated with quantitative iris colour measurements in subjects of European ancestry. However, this SNP was also strongly associated with iris colour in the South Asian sample, although there were no participants with blue irides in this sample. The usefulness of this method is not restricted only to the study of iris pigmentation. High-resolution pictures of the iris will also make it possible to study the genetic variation involved in iris textural patterns, which show substantial heritability in human populations.

2.2 Introduction Eye colour is determined by the type of melanin present and the density and distribution of melanosomes located within the melanocytes of the iris stroma (Sturm and Larsson, 2009). Iris pigmentation exhibits a variable global distribution. In most populations, eye colour is primarily limited to varying shades of brown. However, individuals of European, and to a lesser extent, North African, Middle Eastern, Central Asian, and South Asian ancestry, express a wide range of colours that include shades of brown, green, and blue. In recent years, the use of linkage analyses and genome-wide association studies has led to the identification of several of the key genes associated with iris colour variation (Frudakis et al., 2003; Duffy et al., 2007; Sulem et al., 2007; Eiberg et al., 2008; Kayser et al., 2008; Sturm et al., 2008; Branicki et al., 2009; Mengel- From et al., 2009; Liu et al., 2010). However, most of the research efforts have focused on European populations, and there have been very few studies exploring the phenotypic variation and genetic basis of iris pigmentation in populations of non-European ancestry. One of the major challenges for unraveling the genetic architecture of iris pigmentation is obtaining a quantitative measurement of eye colour. Traditional methods of measuring skin and hair pigmentation, which are based on reflectometry, cannot be used on the iris. Consequently, the majority of studies investigating iris pigmentation variation have used a limited number of discrete categories to characterize eye colour (for a recent review about iris colour classification, see Mackey et al., 2011). Although such discrete classification methods have successfully identified some of the major genes associated with iris colour, they also have a number of weaknesses. For one, they are subjective and have limited inter- and intra-observer reliability (Seddon et al., 1990; Frudakis et al., 2003). Additionally, they are unable to account for the extensive quantitative variation that is inherent in iris pigmentation. Recently, a number of research groups have developed quantitative methods for measuring iris colour. Melgosa et al. (2000) used a spectroradiometer to obtain a measurement of the combined pupil and iris in the Commission Internationale de L'Eclairage L*a*b* (CIELAB) colour space. German et al. (1998) studied drug response in irides by manually extracting a number of colour measurements from photographs taken of the human eye. These measurements included parameters in the XYZ and CIELAB colour spaces. Frudakis (2008) extracted RGB and luminosity measurements from iris photographs and condensed this information into a single ‘iris melanin index.’ Most recently, Liu et al. (2010) isolated hue and saturation values from iris

31 photographs; hue was used to represent the type of melanin, while saturation was used to represent the amount of melanin in the iris. The development of such quantitative methods is creating new opportunities to study the genetics of iris pigmentation variation in a much more objective way. In this article, we introduce a new automated method to quantify iris colour. This method extracts colour measurements in the CIELAB colour space from high-resolution photographs of the iris. CIELAB is an international standardized colour system that was designed to approximate human colour vision (CIE, 1986). The CIELAB colour space provides quantitative measurements across three different dimensions: a brightness dimension (L*), a green/red dimension (a*), and a blue/yellow dimension (b*). The CIELAB space has advantages over other colour systems because it is perceptually uniform: the distance between two colour points closely corresponds to perceptual differences in human vision. Additionally, the CIELAB is a metric space that factors in the colour temperature of the illuminant, also known as the white balance of the source illumination. Colour data from an RGB colour space does not include the illuminant temperature; this factor is however included in the conversion from RGB to CIELAB. The output of the xenon flash as used in photography is approximated as D55—that is, Daylight at 5,500 degrees Kelvin. We applied this new method to a sample of individuals of diverse ancestry (East Asia, Europe, and South Asia) that was genotyped for the HERC2 rs12913832 polymorphism, which shows a strong association with blue eye colour (Eiberg et al., 2008; Sturm et al., 2008).

2.3 Materials and Methods 2.3.1 Recruitment Study participants were recruited at the Molecular Anthropology Laboratory of the University of Toronto Mississauga (UTM) between 2007 and 2009. The majority of participants were UTM students and staff who responded to online and print advertisements that were distributed throughout the University of Toronto community. To assess geographic ancestry, each participant was asked to complete a questionnaire that inquired about the participant's parental and grandparental places of birth and native languages. For example, individuals who stated that their ancestors were from China, Japan, and Korea, were grouped as East Asian, while

32 those who reported ancestors from India and Pakistan were grouped as South Asian. Individuals who reported being of multiple ancestries were placed into a subgroup designated as ‘Other,’ and were not included in the analysis. In total, 205 subjects were included in the study. Of these, 66 were of East Asian ancestry, 72 were of European ancestry, and 67 were of South Asian ancestry. This study was approved by the University of Toronto Health Sciences Research and Ethics Board. All participants provided written informed consent.

2.3.2 Genotyping We selected the rs12913832 polymorphism in intron 86 of the HERC2 gene for genotyping. The derived ‘G’ allele of this marker has been strongly associated with blue iris colour in populations of European ancestry (Eiberg et al., 2008; Sturm et al., 2008). A blood sample was collected from each participant in a 4-mL EDTA tube, and DNA was extracted from the blood samples using the E.Z.N.A. Blood DNA Midi Kit (Omega Bio-Tek, United States). The samples were then sent to Kbioscience (United States) for genotyping. Kbioscience uses a genotyping method that combines allele-specific PCR with detection by a Fluorescence Resonance Energy Transfer (FRET) system. To evaluate genotyping quality, blind duplicates were included in the genotyping plates. The concordance rate for the genotype calls of the duplicated samples was 100%.

2.3.3 Acquisition and Processing of Iris Photographs We took a photograph of each subject's left iris using the Miles Research Professional Iris camera (Miles Research, United States). This camera consists of a Fujifilm Finepix S3 Pro 12- megapixel DSLR mounted on a Nikkor 105-mm macro lens. A coaxial biometric illuminator was used to deliver a constant and uniform source of light to each iris at 5,500 K (D55 illuminant). Camera performance in terms of colour recording fidelity was verified using the 24-patch GretagMacbeth Colour Checker and Imatest Master software (http://www.imatest.com). All photographs were taken using an aperture of f19 and exposure sensitivity (ISO) of 200. The shutter speed was set to 1/60 second and all pictures were recorded in 12-megapixel JPEG format.

Upon examination of the photographs used for the analysis of iris colour there was evidence of underexposure. As a corrective measure, an exposure normalization procedure was applied to all 205 images using a colour-neutral brightness operator, the screen-mode blending of each image with itself. This exposure compensation filter shifts the colours away from black and white so as to improve discrimination of small differences in colour. The effect is analogous to using two slide projectors illuminating the same transparency image, aligned on a screen, to create a brighter result. Mathematically, the screen blend lightening filter (blending an eight-bits- per-channel image with itself) is expressed as: c′ = 255 – ((255 − c0)2/255), where c0 represents each R, G, B channel value as an eight-bit integer from 0 to 255, and c′ is the filter result value.

2.3.4 Acquisition of Colour Measurements A single RGB measurement was extracted from each iris image using a multistep process: 1/ A macro was created in Adobe Photoshop CS5 that cropped each image to the middle-third of the image height, centred on the pupil; 2/ A second macro was developed that cropped out the regions of the image containing the sclera; 3/ A final macro was used to crop the sample down to a 256 by 256 pixel square at the ciliary zone of the iris (mid-periphery) in the 9:00 meridian of each sample; and 4/ The program ImageJ (National Institutes of Health, United States) was used to extract RGB data from each of the cropped image samples. We used the ‘average RGB’ of the 256 by 256 pixel square to characterize the iris colour. The RGB values were transformed into CIE 1976 L*a*b* (CIELAB) colour space. The CIELAB colour space is a universal standardized system that uses three dimensions to represent colour. The first dimension, L*, is used to represent brightness and can have values that range from 0 to 100, where 0 is black and 100 is white. The a* and b* dimensions correspond to differences in colour, with negative values of a* indicating green and positive values of a* indicating red, and negative values of b* indicating blue and positive values of b* indicating yellow. To transform the RGB values into the CIELAB colour space, the RGB measurements were first converted into XYZ values using standard equations found at http://www.easyrgb.com/index.php?X=MATH&H=02#text2. The XYZ values were then transformed into L*, a*, and b* measurements using standard equations found at http://www.easyrgb.com/index.php?X=MATH&H=07#text7. As the coaxial biometric

34 illuminator delivered light to each iris at a temperature of 5,500 K, the colour space conversion was based on the D55 standard illuminant. The observer was set at 2°.

2.3.5 Statistical Analysis Deviations from Hardy–Weinberg proportions in each population were evaluated using the Hardy–Weinberg exact test (see http://ihg2.helmholtz-muenchen.de/cgi-bin/hw/hwa1.pl). We used general linear model (GLM) analyses to test the association between the rs12913832 genotypes and L*, a*, b* values in the European and South Asian samples (in the East Asian sample, 64 out of 65 individuals were homozygous for the ancestral ‘A’ allele, so due to the lack of polymorphism in this sample, we did not carry out the statistical analysis). The tests were done with the statistical program SPSS (version 17.0, SPSS, 2008, United States). Associations were considered significant when P ≤ 0.05.

2.4 Results Figure 2-1 shows a graphical representation summarizing the procedure that was used to obtain measurements in the CIELAB colour space from each iris, with a representative sample including two blue, one green, and three brown irides. Figure 2-2 depicts the measurements obtained for the full sample. There is a substantial dispersion of the L*, a*, and b* values in the three-dimensional space. As expected, the dispersion is more pronounced in the European sample, which includes eye colours that ranged from light blue to dark brown. In the European sample, the L* values ranged from 17.63 to 67.08, the a* values from −6.56 to 25.52, and the b* values from −14.34 to 37.62. There was also considerable variation in the East Asian and South Asian samples. In the East Asian sample, which is comprised primarily of dark brown irides, the L* values fluctuated from 10.96 to 40.92, the a* values from −2.67 to 24.42, and the b* values from 5.73 to 11.12. In the South Asian sample, which includes a broader representation of shades of brown than the East Asian sample, the L* (ranging from 10.18 to 44.31) and a* (ranging from −2.86 to 25.42) values showed a distribution that is broadly similar to that observed in the East Asian sample. However, there was considerably more dispersion in the b* axis (0.32–31.36). When examining the CIELAB coordinates considering broad categorical classifications of iris colour, blue irides tended to have very high L* values and negative a* and b* values. Green

35 irides usually mapped between brown and blue irides, and had a high L* value, a b* value that hovered around 0, and a negative a* value. In contrast, brown irides tended to have low L* values and positive a* and b* values. Irides with lighter shades of brown had higher L*, a*, and b* values than irides with darker shades of brown (see Figure 2-1B). Overall, it is evident that classifications that use broad categories such as ‘brown,’ ‘intermediate,’ or ‘blue’ do not satisfactorily capture iris pigmentation variation. We also observed that in many individuals, particularly those of European ancestry, the pigmentation of the central pupillary zone is darker than the pigmentation of the peripheral ciliary zone (central heterochromia). To test the usefulness of our method to quantify iris colour, we evaluated the effect of the HERC2 rs12913832 polymorphism on the L*, a*, and b* values. This single nucleotide polymorphism (SNP) has been associated with iris pigmentation in previous studies (Eiberg et al., 2008; Sturm et al., 2008; Liu et al., 2010). Table 2-1 shows the genotype and allele frequencies in the East Asian, European, and South Asian samples. There were no significant deviations from Hardy–Weinberg proportions in any of the samples (Table 2-1). The frequency of the derived ‘G’ allele, which has been associated with blue iris colour in previous studies, was 71% in the European sample and 13% in the South Asian sample. In the East Asian sample, 64 out of 65 individuals were homozygous for the ancestral ‘A’ allele, so we eliminated this sample from the statistical analysis. We employed an unconstrained genetic model, which estimates separately the effects of the homozygotes for the derived ‘G’ allele and the heterozygotes, using the homozygotes for the ancestral ‘A’ allele as the reference genotype. The results of these tests for the European and South Asian sample are depicted in Table 2-2. In the European sample, rs12913832 was significantly associated with the L*, a*, and b* (P < 0.001) dimensions of colour space. In the lightness dimension, the estimates of the regression coefficient (beta) indicated that L* increased by 12.37 U in AG heterozygotes and by 23.27 U in GG homozygotes, with respect to AA homozygotes. In the green/red dimension, a* decreased by 7.19 U in AG heterozygotes and by 23.16 U in GG homozygotes, with respect to AA homozygotes. In contrast, in the blue/yellow dimension, b* decreased dramatically (22.58 U) in the GG homozygotes, but was very similar for the AG heterozygotes and the AA homozygotes. Therefore, rs12913832 seems to fit a dominant model for the blue/yellow dimension (b*) of the CIELAB colour space, with the ancestral ‘A’ allele dominant over the recessive ‘G’ allele, and a codominant model for the green/red (a*) and lightness (L*) dimensions. In the South Asian

36 sample, rs12913832 was also significantly associated with the L*, a*, and b* dimensions of the CIELAB colour space (P < 0.001). In this sample we only observed one homozygote for the derived ‘G’ allele, so it was not possible to explore in detail the effects of the three rs12913832 genotypes on the colour coordinates. However, it is interesting to note that in the South Asian sample, heterozygotes for the AG allele showed higher L*, a*, and b* values than homozygotes for the ancestral ‘A’ allele, indicating that, on average, they have lighter shades of brown (see Figure 2-1B). To verify the results, we ran the same tests on a smaller collection of irides that were photographed 1 year later, using the same camera, but with a slightly different shutter speed. This collection included photographs of 27 East Asian, 37 European, and 25 South Asian participants and all of the major eye colour phenotypes (blue, brown, green) were represented. The results are very consistent with those described above. In the European sample, the rs12913832 marker was strongly associated (P < 0.001) with the three dimensions of the CIELAB colour space, and the parameters of the unconstrained model are overly similar to those of the larger sample. L* increased by 17.45 U in AG heterozygotes and by 24.79 U in GG homozygotes, a* decreased by 11.24 U in AG heterozygotes and by 20.80 U in GG homozygotes, and b* decreased by 20.45 U in GG homozygotes with respect to AA homozygotes, but there were very small differences between AA homozygotes and AG heterozygotes in the b* dimension. In the South Asian sample, rs12913932 was significantly associated with the L* (P < 0.001) and b* (P = 0.021) dimensions, and was not significant for the a* dimension (P = 0.369). Again, heterozygote AG individuals showed, on average, higher L*, a*, and b* values than homozygote AA individuals.

2.5 Discussion We introduce a new method to measure iris colour, using a series of crops from high- resolution photographs to obtain a quantitative estimate of colour in the three coordinates of the CIELAB colour space. The CIELAB colour space is a universal colour system that was designed to approximate human vision (CIE, 1986). Iris colour variation is reported across three different planes: a brightness plane (L*) and two colour planes (a* and b*). When we plotted our set of 205 human irides in CIELAB colour space, most of the points were distributed throughout the

37 three-dimensional space in a fairly continuous manner. This confirms that iris pigmentation, like skin and hair colour, is a quantitative trait. This is evident in the three population groups studied in our analysis, which included two samples (East Asians and South Asians) that were comprised primarily of individuals with ‘brown’ irides. Treating eye colour as a categorical trait in studies aimed at characterizing the genetic basis of iris pigmentation disregards a substantial amount of the existing variation and reduces the statistical power to identify main effects and gene–gene interactions. To validate our method, we tested the association between the HERC2 rs12913832 polymorphism and our CIELAB colour space measurements in the European and South Asian samples. This particular marker is found in a highly conserved region of intron 86 of the HERC2 gene and has been strongly associated with blue/brown iris colour in previous studies in populations of European ancestry (Eiberg et al., 2008; Sturm et al., 2008). Although the exact mechanisms of action of this polymorphism have not yet been fully determined, it is purported to play a role in the regulation of OCA2. Eiberg et al. (2008) suggested that rs12913832 falls within an OCA2 silencer complex that may be disturbed or stabilized depending on which of the two rs12913832 alleles is present (Eiberg et al., 2008). In contrast, Sturm et al. (2008) have argued that this particular polymorphism may be a binding site for the regulatory protein HLTF (Sturm et al., 2008). Under this alternative model, when the ancestral ‘A’ allele is present, HLTF binds to this region and promotes the transcription of OCA2. When the derived ‘G’ allele is present, however, HLTF cannot properly bind to the DNA and OCA2 transcription is reduced. Studies in human melanocyte strains have shown that the presence of the rs12913832 ancestral ‘A’ allele is associated with increased levels of OCA2 transcript compared to the derived ‘G’ allele (Cook et al., 2009). The frequencies for the derived ‘G’ allele that we observed in our sample (East Asians <1%, Europeans = 71%, and South Asians = 13%) are in overall agreement with previously reported data. For example, the frequency of this allele in the Human Genome Diversity Project (HGDP) East Asian samples was <1%, for the European samples 51%, and for Central-South Asian samples, 15% (http://spsmart.cesga.es/). As expected, rs12913832 was significantly associated with the L*, a*, and b* planes of the CIELAB colour space in our European sample. The use of quantitative measures, instead of categorical variables, highlights some interesting aspects of the effect of rs12913832 on iris

38 colour. For the b* (blue/yellow) dimension of the CIELAB colour space, rs12913832 fits a dominant model of inheritance quite well, with the ancestral ‘A’ allele dominant over the G- derived allele. Homozygous GG individuals have lower b* values (corresponding to the blue regions of the CIELAB space) than heterozygous AG or homozygous GG individuals, which show similar b* values. However, the L* and a* dimensions appear to fit a codominant model, with AG heterozygotes showing intermediate values between AA and GG homozygotes AA. Traditionally, eye colour has been described as a trait that fits a dominant model of inheritance, with ‘brown’ dominant over ‘blue,’ although there have been many reports demonstrating inconsistencies with a dominant model of inheritance (Frudakis, 2008; Sturm and Larsson, 2009). Focusing specifically on the rs12913832 polymorphism, Eiberg et al. (2008) showed that in a sample of ∼200 Danish individuals this polymorphism was perfectly associated with blue/brown colour, strongly supporting a model in which blue eye colour is caused by homozygosity of the rs12913832 ‘G’ allele. However, Sturm et al. (2008) showed that the association of the rs12912832 GG genotype with blue eye colour is not perfect: some individuals with this genotype had brown eyes, and some heterozygous AG individuals blue eyes. These two studies used categorical classifications of iris colour based on self-report (Eiberg et al., 2008), or the rating of a trained observer with cross-validation with the individual's self-report (Sturm et al., 2008). In our sample, we also see clear evidence that, although strongly associated with blue eye colour, rs12913832 does not fit a simple dominant model of inheritance. Some homozygous GG individuals and heterozygous AG individuals occupy positions in the CIELAB space that do not overlap with the coordinates observed for most of the GG or AG genotypes. We also observed that, contrary to the expectations of a simple dominant model, the AG heterozygotes show L* and a* values that are intermediate between those characteristic of the AA and GG genotypes. This suggests that AG heterozygotes have intermediate levels of OCA2 expression, and this is reflected as perceptible differences in the L* and a* dimensions of the colour space. When we performed the same analysis on the South Asian sample, rs12913832 was also significantly associated with the L*, a*, and b* dimensions of colour space. As our South Asian sample is comprised almost entirely of irides that would have been defined as ‘brown’ using a categorical iris classification system, this shows that rs12913832 also plays a role in modulating subtle gradations in brown eye colour. In particular, AG heterozygotes showed higher L*, a*,

39 and b* values (e.g., lighter shades of brown) than AA homozygotes. Again, this strongly suggests that rs12913832 does not fit a dominant model of inheritance. We repeated the analysis in an independent sample measured one year later, and we obtained similar results. However, the parameter values (beta coefficients) obtained in this replication sample should be considered with caution due to the small sample size (e.g., in the European sample, there were only three homozygotes for the ancestral allele). It is worth noting that, in our combined iris samples, we found two individuals of South Asian ancestry that had two copies of the rs12913832 derived ‘G’ allele and brown or heterochromatic irides. Additionally, we observed that, whereas AG heterozygotes show significantly lower a* values than homozygotes for the ancestral ‘A’ allele in the European sample, in the South Asian sample AG heterozygotes have significantly higher a* values than AA homozygotes. These findings, as well as information from other studies described above, clearly indicate that eye colour is a polygenic trait with a complex genetic architecture, where the effect of some of the major loci may be modified by other polymorphisms (gene–gene interaction). A recent study in a large sample of European descent has shown that interactions between the genes HERC2 and OCA2, HERC2, and SLC24A5 and HERC2 and TYRP1 play a role in the determination of eye colour (Pospiech et al., 2011). Similarly, Liu et al. (2010) described that interactions of several polymorphisms in pigmentation genes are responsible for some of the variation of hue and saturation observed in their iris sample. In this study, we automatically extracted a 256 by 256 pixel square from the iris to estimate the average colour coordinates in the CIELAB colour space. As shown above, this method is an improvement over the categorical classifications conventionally used to study the genetics of eye colour. However, this method does not fully capture the complexity of iris colour. Central heterochromia is a common feature observed in human irides. In individuals with central heterochromia, the pupillary zone of the iris has a darker colour than the peripheral ciliary zone. Alternative methods of quantifying iris colour based on high-resolution photographs can be employed to specifically study central heterochromia. For example, a measure describing colour (e.g., hue angle, saturation, or just wavelength) can be obtained in a radial section of the iris, making it possible to compare the colour of the pupillary and ciliary zones of the iris. Figure 2-3 illustrates such a method using high-resolution photographs of individuals with and without central heterochromia. In our sample, central heterochromia was present in the three populations

40 examined. As well, it was found in irides of all of the major colour groups, although it tends to be most noticeable in individuals with lighter coloured eyes. Finally, apart from central heterochromia, there are a number of other iridial structures that can be studied using high-resolution photographs, such as nevi (hyper-pigmented spots), Wolfflin nodules (pale nodules encircling the iris that are composed primarily of collagen tissue), Fuchs' crypts (pit-like depressions near the collarette and in the ciliary zone of the iris), and contraction furrows (circular and radial folds due to iris contraction or dilation) [Larsson and Pedersen, 2004; Sturm and Larsson, 2009]. These traits show a substantial heritability in human populations (between 58 and 78%), but the genetic basis of these and other iris patterns remains largely unexplored (Larsson et al., 2003). Using a genome-wide association (GWA) strategy, Larsson et al. (2011) reported that variants within the gene SEMA3A are significantly associated with crypt frequency, polymorphisms within the gene TRAF3IP1 with contraction furrows, and variants near the pigmentation gene SLC24A4 with the presence of a peripupillary pigmentary ring. Interestingly, the genes SEMA3A and TRAFIP1 are involved in pathways that control neurogenesis, neural migration, and synaptogenesis, and this led Larsson et al. (2011) to suggest that genes involved in normal neuronal pattern development may also influence iris structures. In conclusion, we present a method to measure iris colour quantitatively using high- resolution photographs obtained with uniform illumination and show that, using this approach in individuals of diverse ancestry, it is possible to gain interesting insights about the role of the HERC2 rs12913832 polymorphism in iris pigmentation. There are still important gaps in our understanding of the genetic architecture of iris pigmentation. The application of quantitative methods for measuring eye colour opens up new avenues in pigmentation research. As these methods are able to distinguish between subtle colour gradations, it will be possible to increase the statistical power to identify main effects and gene interactions, and to study the genetics of iris pigmentation in populations in which traditional categorical systems are of limited use. Additionally, such methods will provide a means of unraveling the genetic architecture of other traits, such as central heterochromia, and a number of iris patterns that show substantial variation in human populations. Advancing our current knowledge of the genetic basis of iris colour and structure is important from multiple perspectives. For example, from a forensic perspective it will be useful to reliably predict iris colour and structure based on DNA information. From an anthropological and evolutionary perspective, it will allow researchers to gain new insights about

41 the major evolutionary events associated with the current distribution of eye colour variation, which is quite different from the geographical distribution of skin pigmentation. Of interest is to which extent natural or sexual selection, or the action of demographic factors (e.g., genetic drift, range expansions) shaped the current distribution of eye colour. Finally, from the developmental and physiological perspectives, discovering the genes associated with iris colour and structure will improve our understanding of the pathways that determine these traits.

2.6 Tables

Table 2-1 General Descriptive Statistics Observed and expected genotype frequencies, allele frequencies, and the Hardy–Weinberg exact test for HERC2 rs12913832 in the East Asian, European, and South Asian samples

Observed and Expected Genotype Hardy-Weinberg Exact Population Allele Frequencies Frequencies Test (p) A:A = 64 (64.00) A = 99% East Asian A:G = 1 (0.99) 1.000 G = 1% G:G = 0 (0.00) A:A = 9 (6.00) A = 29% European A:G = 23 (28.99) 0.087 G = 71% G:G = 38 (35.00) A:A = 50 (50.09) A = 87% South Asian A:G = 15 (14.81) 1.000 G = 13% G:G = 1 (1.09)

Table 2-2 General Linear Model Analysis Results Results of the general linear model analysis for the brightness dimension (L*), the green/red dimension (a*), and the blue/yellow dimension (b*) of the CIELAB colour space in the European and South Asian samples

Population Colour Space Dimension p-value Genotype Beta p-value AA - - L* 0.000 AG 12.368 0.000 GG 23.267 0.000 AA - - European a* 0.000 AG -7.187 0.003 GG -23.158 0.000 AA - - b* 0.000 AG 1.989 0.585 GG -22.584 0.000 AA - - L* 0.000 AG 8.358 0.000 AA - - South Asian a* 0.000 AG 2.701 0.037 AA - - b* 0.000 AG 6.788 0.000

2.4 Figures

Figure 2-1 Colour Extraction Methodology and Iris Colour Coordinates (A) An illustration of the sample 256 by 256 pixel square that was isolated from the 9:00 meridian of six irides of varying colour (two blues, one green, three browns). The average RGB of this square was determined, and transformed into L*, a*, and b* values in the CIELAB colour space. (B) The coordinates of overall iris colour of six exemplar samples in the CIELAB colour space. Each iris is graphically represented by the 256 by 256 pixel square sample region that was used to evaluate iris colour and its colour space coordinates (L*, a*, b*).

Figure 2-2 Distribution of Irises in CIELAB Colour Space (A) The distribution of the full sample of irides of participants of East Asian (red circles), South Asian (green triangles), and European (blue squares) ancestry in the three coordinates of CIELAB colour space: L* (the lightness coordinate), a* (the red-green coordinate), and b* (the blue-yellow coordinate). (B) The coordinates of the same irides for a* and b*, the chromatic components of the CIELAB colour space.

Figure 2-3 Visualization of Central Heterochromia One method of visualizing central heterochromia. The iris photographs of two heterochromatic and one homogenous irides were reduced to a width of 125 pixels. The program ImageJ (National Institutes of Health, United States) was used to extract a horizontal band (left to right) from the 9:00 meridian of each iris in RGB space and the RGB values were transformed into saturation values using standard equations found at http://www.easyrgb.com/index.php?X=MATH&H=18#text18. Irides with substantial heterochromia showed sharp changes in saturation across the width of the iris.

2.7 References Branicki, W., Brudnik, U., and Wojas-Pelc, A. (2009). Interactions between HERC2, OCA2 and MC1R may influence human pigmentation phenotype. Ann. Hum. Genet. 73, 160–170. CIE. (1986). CIE publication no. 15.2—colorimetry. Austria: Central Bureau of the Commission Internationale de L'Eclairage. Cook, A.L., Chen, W., Thurber, A.E., Smit, D.J., Smith, A.G., Bladen, T.G., Brown, D.L., Duffy, D.L., Pastorino, L., Bianchi-Scarra, G., et al. (2008). Analysis of cultured human melanocytes based on polymorphisms within the SLC45A2/MATP, SLC24A5/NCKX5, and OCA2/P loci. J Invest Dermatol 129, 392–405. Duffy, D.L., Montgomery, G.W., Chen, W., Zhao, Z.Z., Le, L., James, M.R., Hayward, N.K., Martin, N.G., and Sturm, R.A. (2007). A three–single-nucleotide polymorphism haplotype in intron 1 of OCA2 explains most human eye-color variation. Am. J. Hum. Genet. 80, 241–252. Eiberg, H., Troelsen, J., Nielsen, M., Mikkelsen, A., Mengel-From, J., Kjaer, K.W., and Hansen, L. (2008). Blue eye color in humans may be caused by a perfectly associated founder mutation in a regulatory element located within the HERC2 gene inhibiting OCA2 expression. Hum. Genet. 123, 177–187. Frudakis, T. (2010). Molecular photofitting: predicting ancestry and phenotype using DNA (Elsevier). Frudakis, T., Thomas, M., Gaskin, Z., Venkateswarlu, K., Chandra, K.S., Ginjupalli, S., Gunturi, S., Natrajan, S., Ponnuswamy, V.K., and Ponnuswamy, K.N. (2003). Sequences associated with human iris pigmentation. Genetics 165, 2071–2083. German, E.J., Hurst, M.A., Wood, D., and Gilchrist, J. (1998). A novel system for the objective classification of iris colour and its correlation with response to 1% tropicamide. Ophthalmic Physiol. Opt. 18, 103–110. Kayser, M., Liu, F., Janssens, A.C.J.W., Rivadeneira, F., Lao, O., Duijn, K. van, Vermeulen, M., Arp, P., Jhamai, M.M., Ijcken, W.F.J. van, et al. (2008). Three genome-wide association studies and a linkage analysis identify HERC2 as a human iris color gene. Am. J. Hum. Genet. 82, 411– 423. Larsson, M., and Pedersen, N. (2004). Genetic correlations among texture characteristics of the human iris. Mol Vis 10, 821–831. Larsson, M., Pedersen, N.L., and Stattin, H. (2003). Importance of genetic effects for characteristics of the human iris. Twin Res. Hum. Genet. 6, 192–200. Larsson, M., Duffy, D.L., Zhu, G., Liu, J.Z., Macgregor, S., McRae, A.F., Wright, M.J., Sturm, R.A., Mackey, D.A., Montgomery, G.W., et al. (2011). GWAS findings for human iris patterns: associations with variants in genes that influence normal neuronal pattern development. Am. J. Hum. Genet. 89, 334–343.

Liu, F., Wollstein, A., Hysi, P.G., Ankra-Badu, G.A., Spector, T.D., Park, D., Zhu, G., Larsson, M., Duffy, D.L., Montgomery, G.W., et al. (2010). Digital quantification of human eye color highlights genetic association of three new loci. PLoS Genet 6, e1000934. Mackey, D.A., Wilkinson, C.H., Kearns, L.S., and Hewitt, A.W. (2011). Classification of iris colour: review and refinement of a classification schema. Clin. Experiment. Ophthalmol. 39, 462–471. Melgosa, M., Rivas, M.J., Gómez, L., and Hita, E. (2000). Towards a colorimetric characterization of the human iris. Ophthalmic Physiol. Opt. 20, 252–260. Mengel-From, J., Wong, T.H., Morling, N., Rees, J.L., and Jackson, I.J. (2009). Genetic determinants of hair and eye colours in the Scottish and Danish populations. BMC Genet. 10, 1– 13. Pospiech, E., Draus-Barini, J., Kupiec, T., Wojas-Pelc, A., and Branicki, W. (2011). Gene-gene interactions contribute to eye colour variation in humans. J Hum Genet 56, 447–455. Seddon, J.M., Sahagian, C.R., Glynn, R.J., Sperduto, R., Gragoudas, E., Group, E.D.C.-C.S., and others (1990). Evaluation of an iris color classification system. Invest Ophthalmol Vis Sci 31, 1592–1598. Sturm, R.A., and Larsson, M. (2009). Genetics of human iris colour and patterns. Pigment Cell Melanoma Res. 22, 544–562. Sturm, R.A., Duffy, D.L., Zhao, Z.Z., Leite, F.P.N., Stark, M.S., Hayward, N.K., Martin, N.G., and Montgomery, G.W. (2008). A single SNP in an evolutionary conserved region within Intron 86 of the HERC2 gene determines human blue-brown eye color. Am. J. Hum. Genet. 82, 424– 431. Sulem, P., Gudbjartsson, D.F., Stacey, S.N., Helgason, A., Rafnar, T., Magnusson, K.P., Manolescu, A., Karason, A., Palsson, A., Thorleifsson, G., et al. (2007). Genetic determinants of hair, eye and skin pigmentation in Europeans. Nat Genet 39, 1443–1452.

Chapter 3: Iris Pigmentation as a Quantitative Trait: Variation in Populations of European, East Asian and South Asian Ancestry and Association with Candidate Gene Polymorphisms

Author Contributions: M. Edwards participated in the design of the study, collected the data, conducted the data analysis and drafted the manuscript. D. Cha developed the iris colour program and assisted with data analysis. S. Krithika, M. Johnson and G. Cook assisted with data collection and molecular laboratory work. E.J. Parra conceived the study, participated in the design of the study, coordinated the study and helped draft the manuscript.

This Chapter Has Been Previously Published As: Edwards, M., Cha, D., S. Krithika., Johnson, M., Cook, G., Parra, E.J. (2015) Iris pigmentation as a quantitative trait: variation in populations of European, East Asian and South Asian ancestry and association with candidate gene polymorphisms. Pigment Cell Melanoma Res.

Acknowledgements: We would like to thank all the individuals who participated in this study. ME was funded by a 3-yr Natural Sciences and Engineering Research Council (NSERC) CGSD award and an Ontario Graduate Scholarship (OGS). EJP was funded by an NSERC Discovery Grant.

49 50

3.1 Abstract In this study, we present a new quantitative method to measure iris colour based on high resolution photographs. We applied this method to analyse iris colour variation in a sample of individuals of East Asian, European and South Asian ancestry. We show that measuring iris colour using the coordinates of the CIELAB colour space uncovers a significant amount of variation that is not captured using conventional categorical classifications, such as ‘brown’, ‘blue’ or ‘green’. We tested the association of a selected panel of polymorphisms with iris colour in each population group. Six markers showed significant associations with iris colour in the European sample, three in the South Asian sample and two in the East Asian sample. We also observed that the marker HERC2 rs12913832, which is the main determinant of ‘blue’ vs. ‘brown’ iris colour in European populations, is also significantly associated with central heterochromia in the European sample.

3.2 Introduction Iris colour in humans is a complex trait that is primarily the product of differences in the structure and organization of the eye. The typical iris consists of five layers: the iris pigment epithelium (IPE), the sphincter and dilator muscles, the stromal layer (SL) and the anterior border layer (ABL) (Eagle, 1988; Oyster, 1999; Sturm and Larsson, 2009). The IPE is derived from the neuroectoderm and contains many large, densely-packed spherical melanosomes (Eagle, 1988; Prota et al., 1998; Wilkerson et al., 1996). The amount and composition of melanin in this layer is similar in all healthy humans as the primary purpose of the IPE is to protect the retina and absorb excess light (Wilkerson et al., 1996). As a result, variation in this layer does not contribute significantly to normal iris pigmentation variation. In contrast, the melanosomes in the SL and ABL, which originate from the neural crest, are smaller, ovoid-shaped and differ widely between individuals (Imesch et al., 1996; Prota et al., 1998; Sturm and Larsson, 2009). It is presently believed that the majority of iris pigmentation variation can be directly linked to the amount and type of melanin found in these two layers, with darker eyes having more melanin and a higher eumelanin/pheomelanin ratio than lighter eyes (Peles et al., 2009; Wakamatsu et al., 2008; Wielgus and Sarna, 2005). However, there are a number of other factors that can also influence iris colour, including the depth of the SL, the organization of the extracellular components in the SL and ABL, the thickness and curvature of the cornea, the presence of structures (e.g. pigment spots, Wolfflin nodules) in the iris and the amount of melanin found in the SL relative to the ABL (Baranoski and Lam, 2007; Mackey et al., 2011; Prota et al., 1998; Wielgus and Sarna, 2005). Iris colour shows a very distinct global distribution. Across Europe, and to a lesser degree North Africa, the Middle East, and Central Asia, irises show extensive pigmentation variation and range from light blues to greens to browns (Donnelly et al., 2012; Eiberg et al., 2008). Many irises in these regions also show central heterochromia, where there is a band of colour around the pupil that differs from the rest of the eye. Throughout the rest of the world, however, iris colour appears to be much more homogenous and is primarily limited to varying shades of brown. At present, the genetic basis of iris colour has been extensively studied in populations of European ancestry. The majority of variation between blue and brown eye colour has been attributed to the marker HERC2 rs12913832, which is located in a highly conserved region of the genome that is believed to regulate the transcription of the nearby pigmentation gene, OCA2

(Eiberg et al., 2008; Kayser et al., 2008; Sturm et al., 2008). It has been suggested that this polymorphism is located in the middle of a helicase-like transcription factor (HLTF) binding site (Sturm et al., 2008; Visser et al., 2012). When the ancestral ‘A’ allele is present, HLTF (a chromatin remodelling protein) is able to recognize the sequence and begin unwinding the chromatin. This exposes sequences for other regulatory proteins and promotes the synthesis of OCA2. When the derived ‘G’ allele is present, however, there is reduced recruitment of HLTF and OCA2 transcription is limited. Although HERC2 rs12913832 may be the primary determinant of global iris colour variation, a number of other markers have also been identified that have a more subtle effect on iris pigmentation. These include polymorphisms in OCA2, SLC45A2, SLC24A4, TYRP1 and IRF4 (Graf et al., 2005; Rebbeck et al., 2002; Sturm et al., 2008; Sulem et al., 2007; Walsh et al., 2011). The identification of markers associated with iris colour in European populations has allowed several forensic science groups to develop algorithms that are capable of predicting iris colour from DNA samples obtained at crime scenes and from other unidentified persons (Ruiz et al., 2014; Spichenok et al., 2011; Walsh et al., 2011). Despite our growing understanding of the genetic basis of iris pigmentation in European populations, very few groups have attempted to look at eye colour in populations of non- European ancestry. Although brown eye colour does dominate in regions outside of Europe, there is much variation within these browns and they can range from light reddish-yellows to dark brownish-blacks. In recent years, a number of methods have been developed that allow researchers to characterize iris colour quantitatively. Liu et al. (2010) obtained hue and saturation values from iris photographs, with hue being used to represent the type of melanin and saturation being used to represent the amount of melanin in the iris. Andersen et al. (2013) developed a PIE (Pixel Index of the Eye) score that was computed using the number of pixels labelled brown and the number of pixels labelled blue in photographs of the iris. Recently, we extracted a colour measurement in CIE 1976 L*a*b* (CIELAB) colour space from a 256 by 256 pixel square isolated using high-resolution photographs of the iris (Edwards et al., 2012). These methods go beyond the traditional categorical classification systems that have been used in the past (i.e. ‘green’ ‘blue’ ‘brown’) and provide a means of studying iris colour variation in populations with more homogenously coloured irises. However, thus far, they have not yet been broadly applied to global populations.

Using quantitative methods to study the genetic basis of iris colour in populations of non- European ancestry would greatly contribute to our overall understanding of global iris colour variation. At present, we have a limited understanding of the markers responsible for intermediate iris colours. This is reflected in the fact that most eye colour prediction algorithms perform poorly and inconsistently when looking at intermediately coloured eyes (Dembinski and Picard, 2014; Pneuman et al., 2012; Yun et al., 2014). In addition, identifying novel variants associated with iris colour variation in populations of non-European ancestry would provide valuable information about the genetic basis of skin and hair colour in these regions, given the known pleiotropic effects of some genetic markers on pigmentary traits. Studying iris colour may also be of biomedical interest. Eye colour has been found to be associated with a number of ocular disorders (Mitchell et al., 1998; Wakamatsu et al., 2008). The most well-known of these is perhaps age-related macular degeneration (AMD). This disease, which primarily affects older adults, results from damage to the retina and can lead to vision loss and blindness. Age-related macular degeneration has been strongly linked to population-specific genetic effects (Frank et al., 2000). However, within populations eye colour may be one of the determinants of AMD. In European samples, it has been found that individuals with blue eyes have a higher risk of developing AMD (Frank et al., 2000; Mitchell et al., 1998). Similarly, the incidence of primary angle-closure glaucoma in East Asian populations may be associated with variation in brown iris colour. A recent study found that individuals with dark brown eyes in a Malaysian population living in Singapore had a narrower angle closure than individuals with light brown eyes (Sidhartha et al., 2014). Thus, it is not only necessary to investigate the genes responsible for the difference between blue, brown and intermediate phenotypes in European populations, but also to look at the genes that modulate brown iris pigmentation in other regions as well. In this study, we have three primary goals: 1/ To improve our quantitative method of measuring iris colour in order to better capture iris colour variation in populations of diverse ancestry; 2/ To look at the association between putative pigmentation markers and iris colour in a sample of European, East Asian, and South Asian ancestry; 3/ To look at the association between pigmentation markers and central heterochromia within the European and South Asian samples.

3.3 Materials and Methods 3.3.1 Sample Collection Between 2012 and 2014, 1465 healthy volunteers of East Asian, European and South Asian ancestry volunteered for a research study on human pigmentation variation. All participants ranged between 18 and 35 years of age and were recruited using online and print advertisements directed towards the University of Toronto student community. A personal questionnaire was administered to each participant to determine their age, sex, self-described eye colour and whether or not they had been diagnosed with any pigmentation-related diseases or disorders. Biogeographical ancestry was determined using information from the personal questionnaire, which inquired about the ancestry, place of birth and first language of each participant’s maternal and paternal grandparents. Individuals who stated that all of their grandparents originated in China, Japan, Korea or Taiwan were categorized as East Asian, and individuals who stated that all of their grandparents originated in Pakistan, India, Bangladesh or Sri Lanka were categorized as South Asian. Individuals were categorized as European if all of their grandparents originated in any country in Europe, other than Turkey. Admixed individuals who had grandparents from two different regions (i.e. East Asia and Europe) were excluded from the analysis. When information about the grandparents was not known, the self-described ancestry of both parents was used to assess biogeographical ancestry. In addition to the 1465 participants that were included in this paper, 308 additional volunteers were recruited and excluded because they could not be categorized as East Asian, European or South Asian using this criteria. This study was approved by the University of Toronto Research and Ethics Board (Protocol Reference #27015), and all participants were required to provide written informed consent.

3.3.2 Acquisition of Photographs A photograph of each participant’s right eye was taken using a Miles Research Professional Iris Camera (Miles Research, United States). This camera consists of a Fujifilm Finepix S3 Pro DSLR 12-megapixel camera body attached to a 105-mm Nikkor lens. A

55 biometric coaxial cable was used to deliver light to the iris at a constant light temperature to maintain colour and brightness fidelity and reduce the impact of ambient light. The camera body needed to be replaced after the first 552 participants and a camera body with an identical make and model was acquired. We could not adequately assess if the photographs taken with the first and second camera bodies were identical as we did not have a large enough sample of individuals whose eyes had been photographed using both cameras. Therefore, we split our sample into two groups. All photographs were taken with an ISO of 200, a shutter speed of 1/125” and an aperture of f19. Photographs were initially acquired in RAW format and later converted to JPEG format using Adobe Camera Raw in Adobe Photoshop CS5 (Adobe Systems Incorporated, United States). They were resized from 3043 x 2036 pixels to 1200 x 803 pixels to optimize the processing of iris colour. The white balance was set to flash, the contrast and blacks levels were set to zero, and all other camera defaults were preserved for each conversion.

3.3.3 Acquisition of Iris Colour One of the authors (D.C.) designed an eleven step Web-based application to characterize iris structure and acquire iris colour from photographs of the eye. The web application can be accessed at http://iris.davidcha.ca/. Accounts can be set up for interested users by request. Only the first six steps are necessary to obtain a measurement of iris colour. In the first step, the user must note whether or not the iris is obstructed to an extent that could affect colour and structure measurements. In steps 2-6, the user is asked to identify the approximate centre point of the iris, the approximate centre point of the pupil and then draw a best fit circle around the scleral, pupillary and collarette boundaries. After the scleral, pupillary and collarette boundaries are defined, the application is able to automatically extract a measurement of average eye colour. We decided to estimate iris colour from a 60˚ angle wedge taken from the left side of the iris. This region was chosen because it was least likely to be obscured by eyelids, eyelashes or reflections in our sample. To isolate this wedge, the program defines a start point located at the centre of the iris and moves to a point on the scleral boundary located 240˚ clockwise from the top of the iris. It then follows an arc that stretches for 60˚ until it reaches the point that is 300˚ from the top of the iris. This point is connected to the centre of the iris, which demarcates an isolated wedge. These steps are repeated

56 for the pupillary boundary and the collarette boundary so that the pupil can be completely excised from the wedge and the pupillary and ciliary zones can be separated. Two images are then saved in PNG format: the portion of the wedge that represents the ciliary zone and the portion of the wedge that represents the pupillary zone (Figure 3-2). After processing all iris photographs, we manually scanned the wedges for any evidence of obstruction or incorrect cropping. To obtain a measurement of average iris colour, the application counts all of the pixels located in the wedge (consisting of both the ciliary and pupillary zones) and determines a Red, Green and Blue (RGB) value for each individual pixel. The average RGB value of the entire wedge is calculated by adding up the R, G and B values for each pixel, and then dividing this value by the total number of counted pixels. We chose to describe iris colour in CIE 1976 L*a*b* (CIELAB) colour space (McLaren, 1976). CIELAB is a colour system that was designed to characterize colour across three different coordinates. The L* coordinate represents the lightness dimension and ranges from 0 to 100, with 0 being black and 100 being white. The a* and b* coordinates represent variation in colour, with negative values of a* indicating green and positive values of a* indicating red, and negative values of b* indicating blue and positive values of b* indicating yellow. To convert our RGB measurements into CIELAB colour space, the application first transformed the RGB coordinates into XYZ coordinates and then into L*, a* and b* coordinates using the equations provided by EasyRGB (Logicol Colour Technology, United States). For each conversion, the illuminant was set to D55 and the observer was set to 2 degrees. These are standard conversion settings when using flash photography. This process ultimately provides an average colour estimate for ciliary zone, the pupillary zone and the entire iris in CIELAB colour space. We were also interested in quantifying the total amount of central heterochromia in the iris. To do this, we used CIEDE2000 (ΔE), a colour metric that looks at the difference between two colours in CIELAB colour space (McLaren, 1976). The Web application used the equations provided by EasyRGB (Logicol Colour Technology, United States) to calculate the difference between the CIELAB values in the pupillary and ciliary zones. After acquiring estimates of iris colour from all 1465 irises, the Web application was used to output an Excel spreadsheet which contained the average colour of the entire iris, the average colour of the ciliary zone, and the average colour of the pupillary zone in RGB and CIELAB

57 colour space and the ΔE between the ciliary and pupillary zones. The spreadsheet also produced information on how many pixels were counted in each eye during the calculation of iris colour. All iris analyses were carried out by M.E.

3.3.4 Marker Selection and Genotyping A 2-mL saliva sample was obtained from each participant using the Oragene˖DNA (OG- 500) collection kit (DNA Genotek, Canada). All participants were instructed not to eat, drink or smoke for at least 30 min prior to obtaining the sample to ensure maximal sample purity. DNA was isolated from each sample using the protocol provided by DNA Genotek and eluted in 500 mL of TE (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) Buffer. We selected fourteen markers for genotyping that have either previously been associated with iris colour in European populations or been associated with some other pigmentation phenotype (i.e. skin colour) in East or South Asia (Eaton et al., 2015; Edwards et al., 2012; Eiberg et al., 2008; Graf et al., 2005; Kayser et al., 2008; Liu et al., 2010; Rebbeck et al., 2002; Sturm et al., 2008; Walsh et al., 2011). These markers consisted of HERC2 rs12913832, OCA2 rs1800407, SLC24A4 rs12896399, SLC45A2 rs16891982, SLC24A5 rs1426654, TYR rs1393350, IRF4 rs12203592, DSCR9 rs7277820, TYRP1 rs1408799, NPLOC4 rs9894429, LYST rs3768056, ASIP rs6058017, OCA2 rs1800414 and OCA2 rs74653330. Markers were only included for a population if the minor allele frequency (MAF) exceeded 0.05. The exception to this was OCA2 rs74653330, which has a low MAF (<0.04) in East Asia, but has been found to have a strong effect on skin pigmentation variation (Eaton et al., 2015). All DNA samples were sent to LGC Genomics (United States) for genotyping. LGC Genomics uses a KASP based genotyping method that combines allele-specific amplification with FRET (fluorescent resonance energy transfer) technology. Twenty-nine samples were sent as blind duplicates and 14 samples were sent as blanks to validate the quality of the genotyping results. The concordance rate for both blind duplicates and blanks was 100%.

3.3.5 Statistical Analysis Unless otherwise noted, all statistical analyses were carried out using PLINK (http://pngu.mgh.harvard.edu/~purcell/plink/), a genome analysis program designed to look at the association between genotype and phenotype data. We carried out the following analyses: 1/

A test of deviations from Hardy-Weinberg proportions for each marker included in the East Asian, European and South Asian sample. 2/ A linkage disequilibrium test to determine the amount of linkage between the markers. 3/ An exploratory analysis to look at the association between age and sex and the four colour space measurements (L*, a*, b* and ∆E). As no significant associations were noted between iris colour and age or sex, these variables were not included in the downstream statistical analyses. 4/ A linear regression analysis to determine the effect of each SNP on the L*, a* and b* dimensions of CIELAB colour space. In each population, we only looked at SNPs that reached a MAF of at least 0.05 (with the exception of OCA2 rs74653330 in the East Asian sample). For each SNP, we used a genotypic model, which reports a P-value and beta coefficient independently for the minor allele homozygote and the heterozygote, with respect to the major allele homozygote. We also explored whether there was evidence for deviations from an additive model of inheritance for any of the markers. Deviations were considered significant if the statistic reporting deviation from additivity (DOMDEV) yielded a P-value of < 0.05 for both cameras. Each population was divided into two groups (representing the first camera body and the second camera body) and the regression was carried out individually on each group to account for differences in colour and brightness between the two cameras. 5/ A linear regression analysis in the European and South Asian samples conditioning on the effect of HERC2 rs12913832 and a linear regression analysis in the East Asian sample conditioning on the effect of OCA2 1800414. Again, this analysis was performed separately on each camera body. 6/ A meta-analysis of the linear regression results from both camera bodies. This meta-analysis was carried out both before and after conditioning on HERC2 rs12913832 in the European and South Asian samples and OCA2 rs1800414 in the East Asian sample. For each marker, we reported the P-value of the association, the effect size (beta) and the P-value for Cochran's Q statistic (a measure of the heterogeneity between the two groups used in the meta-analysis). After Bonferroni's correction for multiple comparisons, associations were significant in the East Asian population if P < 0.00714, in the European population if P < 0.00454 and in the South Asian population if P < 0.00500. 7/ A linear regression analysis in the European and South Asian samples looking at the association between the colour difference between the pupillary and ciliary zones (∆E) and the putative pigmentation SNPs. This analysis was carried out separately on each camera body. 8/ A meta-analysis on the linear regression results for ∆E from each camera body. The P-value of the association, the effect size (beta) and

59 the P-value for Cochran's Q statistic were reported for each marker. 9/ Four months after the initial iris colour classification, intra-rater reliability measurements were carried out by M.E and inter-rater reliability measurements were carried out by D.C on 40 random irises. The intraclass correlation coefficient between the original and repeated iris colour measurements was calculated in IBM Statistics SPSS (version 20.0, SPSS Incorporated, United States) using a two- way random model with absolute agreement for the L*, a*, b* and ∆E measurements.

3.4 Results In total 474 East Asians, 624 Europeans and 367 South Asians were included in the study. Six East Asians, four Europeans and four South Asians were unable to remove their contact lenses for the photograph and were excluded from the analysis. Two participants of South Asian descent were removed from the study because they had an obstructed iris wedge. No iris obstruction could be observed in any of the other participants. A final participant of South Asian descent was excluded because the participant reported having a form of albinism that affected ocular pigmentation. Inter- and intra-rater reliability measurements were excellent for all four colour space measurements (Table 3-1). Substantial dispersion across CIE 1976 L*a*b* (CIELAB) colour space could be seen in all three populations (Figure 3-1). Blue eyes tend to show high L* values and negative a* and b* values, while brown eyes have low L* values and high a* and b* values. Due to the arc shape in which irises are distributed across CIELAB colour space, dark brown eyes have a lower L*, a* and b* values than lighter brown eyes (Figure 3-1). Green eyes tend to have a* values that are intermediate between blue and brown irises. The greatest amount of iris colour variation was found in the European sample. However, there was also extensive variation in the South and East Asian groups. The average, minimum and maximum L*, a*, b* and ∆E measurements for all 1448 participants is presented in Table 3-2. There is a noticeable difference between the two camera bodies in the L* dimension, confirming that the bodies have some variation in brightness. As a* and b* are colour dimensions that were designed to be independent from brightness, it is not surprising that there is less variation in these coordinates (McLaren, 1976). However, all three dimensions appear to show somewhat higher values in photographs taken with the first camera compared to those taken with the second camera.

In each population, we looked at the association between markers purported to play a role in pigmentation variation and the L*, a* and b* coordinates. In the European and South Asian populations, the association between each of these markers and ∆E was also examined. With the exception of OCA2 rs74653330 in the East Asian sample, we only looked at polymorphisms that had a minor allele frequency of at least 0.05. For the European population, this included HERC2 rs12913832, OCA2 rs1800407, SLC24A4 rs12896399, SLC45A2 rs16891982, TYR rs1393350, IRF4 rs12203592, DSCR9 rs7277820, TYRP1 rs1408799, NPLOC4 rs9894429, LYST rs3768056 and ASIP rs6058017. In the East Asian population, this included markers SLC24A4 rs12896399, DSCR9 rs7277820, NPLOC4 rs9894429, LYST rs3768056, ASIP rs6058017, OCA2 rs1800414 and OCA2 rs74653330. Lastly, in the South Asian population this included markers HERC2 rs12913832, SLC24A4 rs12896399, SLC45A2 rs16891982, SLC24A5 rs1426654, TYR rs1393350, IRF4 rs1126809, DSCR9 rs7277820, TYRP1 rs1408799, NPLOC4 rs9894429, LYST rs3768056 and ASIP rs6058017. Linkage disequilibrium was low (r2 < 0.1) between all pairs of markers located in the OCA2/HERC2 region. Most markers showed no deviations from Hardy- Weinberg equilibrium. However, DSCR9 rs7277820 showed minor deviations in the European (P = 0.0169) and South Asian (P = 0.0473) samples and SLC24A5 rs1426654 showed major deviations (P = 1.199 x 10-5) in the South Asian sample. Twenty-five individuals (5 East Asians, 10 Europeans and 10 South Asians) were missing genotypes for more than 20% of the polymorphisms and were excluded from the genetic analysis. As the two camera bodies appeared to show some variation in the L*, a* and b* dimensions, we chose to carry out a meta-analysis on the two data sets. A linear association analysis was first performed in each population to look at the association between each SNP and the L*, a* and b* values for each camera-body. In the European and South Asian populations, an additional analysis was carried out for the ∆E measurement. The meta-analysis function in PLINK, using a fixed effects model, was then used to determine the combined significance and effect size in each population. The results of the meta-analysis in the European population can be found in Table 3-3A. HERC2 rs12913832 was strongly associated with the L*, a* and b* dimensions of colour space, with the ancestral ‘A’ allele decreasing the L* and increasing the a* and b* coordinates. For this marker, the a* and b* dimensions showed significant deviations from additivity, indicating that these two colour dimensions are best modelled using a dominant/recessive mode of inheritance

(e.g. the ancestral ‘A’ allele is dominant over the derived ‘G’ allele). In contrast, the L* coordinate showed no significant deviations from additivity. SLC24A4 rs12896399, SLC45A2 rs16891982 and IRF4 rs12203592 were significantly associated with the a* and b* dimensions. For all three markers, the derived alleles decreased both coordinates. Although we had no individuals who were homozygous for the derived genotype in our sample, OCA2 rs1800407 was significantly associated with an increase in the b* dimensions of colour space. Apart from HERC2 rs12913832, none of the markers showed any significant deviations from an additive mode of inheritance. As HERC2 rs12913832 is known to be the primary variant responsible for controlling the difference between blue and brown eye colour, we ran the analysis again after conditioning for this marker (Table 3-3B). Most polymorphisms showed similar effects after conditioning. Interestingly, although OCA2 rs1800407 was no longer associated with the b* dimension of colour space, it was now associated with the L* and a* dimensions, with the derived allele increasing the L* coordinate and decreasing the a* coordinate. Lastly TYR rs1393350 was now associated with the L* dimension of colour space, with the derived ‘A’ allele decreasing this coordinate. In the South Asian sample, HERC2 rs12913832 was strongly associated with the L*, a* and b* dimensions of colour space (Table 3-4A). The derived ‘G’ allele was associated with an increase in the L* and b* dimensions, and a decrease in the a* dimension. HERC2 rs12913832 showed significant deviations from additivity for all three dimensions of colour space indicating that it is best modelled using a dominant/recessive mode of inheritance. SLC24A5 rs1426654 was strongly associated with the L*, a* and b* dimensions of colour space, and each copy of the ancestral allele was responsible for a decrease in all three coordinates. The only other marker associated with iris colour in the South Asian sample was LYST rs3768056, with the derived ‘G’ allele increasing both the a* and b* coordinates. Conditioning on HERC2 rs12913832 in this population had very little effect on both significance levels and effect size (Table 3-4B). In the East Asian population, only OCA2 rs1800414 and OCA2 rs74653330 were associated with iris colour variation (Table 3-5A). OCA2 rs1800414 had a significant effect on the L*, a* and b* dimensions of colour space, with each copy of the ancestral allele decreasing all three coordinates. Although no participants were homozygous for the derived allele at OCA2 rs74653330, having one copy significantly increased both the a* and b* coordinates. After

62 conditioning for OCA2 rs1800414, the effect of OCA2 rs74653330 became stronger, and this marker was now associated with all three dimensions of colour space (Table 3-5B). For this marker, the derived ‘A’ allele increases the L*, a* and b* coordinates. Neither marker showed any significant deviations from an additive mode of inheritance. The only marker that was associated with ∆E in the European sample was HERC2 rs12913832. This marker had an additive effect, with each copy of the ancestral allele decreasing the colour difference between the ciliary and pupillary zones (Table 3-3A). In the South Asian sample, both HERC2 rs12913832 and SLC24A5 rs1426654 were associated with ∆E. HERC2 rs12913832 showed significant deviations from additivity, suggesting that it is best modelled using a dominant/recessive mode of inheritance. In contrast, SLC24A5 did not show any deviation from additivity. After conditioning for the effects of HERC2 rs12913832, SLC24A5 rs1426654 was no longer associated with ∆E.

3.5 Discussion In this study, we present an improved method for measuring iris colour in diverse populations and looked at the association between 14 SNPs purported to play a role in global pigmentation variation and eye colour in a sample of East Asian, European and South Asian ancestry. CIE 1976 L*a*b* (CIELAB) colour space is an ideal system for measuring iris colour. This colour space was designed to capture perceptible differences in colour variation, which is important when looking at visible human traits that are typically distinguished by colour difference (McLaren, 1976). Each unit change in CIELAB colour space is visible to at least 50% of observers (Kuehni and Marcus, 1979). CIELAB also characterizes colour across three coordinates that closely parallel visual differences in eye colour. The L* dimension represents a brightness dimension and ranges from 0 to 100, with 0 being black and 100 being white. The a* and b* dimensions represent variation in colour, with negative values of a* indicating green and positive values of a* indicating red, and negative values of b* indicating blue and positive values of b* indicating yellow. CIELAB also has a colour metric, ∆E that can be used to quantify the difference between two points in colour space (McLaren, 1976). This allows us to investigate colour variation between different regions of the iris. Unlike previous studies, we selected a wedge to represent iris colour instead of the entire iris. We made this decision because the left

63 quadrant of the iris was least likely to be obstructed in our sample. In addition, if we chose to use the entire iris but crop out regions of obstruction, such as eyelashes and eyelids, it would bias the colour of the iris towards the pupillary region. Although several automated methods have been developed to facilitate the isolation of the iris from photographs of the eye (Liu et al., 2010; Pietroni et al., 2014), we chose to manually define the boundaries of the iris. This allowed us to separate the eye into different regions, and look at the difference in colour between the ciliary and pupillary zones. Contrary to previous studies, we did not find any association between iris colour and age or sex in the East Asian, European, or South Asian sample populations. Iris colour has been tentatively associated with sex in a small number of European populations (Pietroni et al., 2014). It has been suggested that this association may be highly population-specific. As we looked at a broad range of biogeographical ancestries across Europe and had a small number of male participants, it is not surprising that this association was absent in our sample. Similarly, the lack of an association between age and eye colour is not surprising, as we sampled from a relatively narrow age range compared to other studies that included participants that ranged from children to seniors (Bito et al., 1997).

3.5.1 Iris Colour in European Populations The European sample showed the greatest variation and spread across CIELAB colour space. This is to be expected, given that eye colour in this population ranges from blues to greens to browns. Interestingly, very few individuals of European descent fell into the region of the colour space occupied by the darkest brown irises. Instead, when individuals of European ancestry had brown eyes, they tended to be lighter. This suggests that there may be fixed genetic markers in this population that modulate the intensity of brown iris pigmentation. In recent years, a number of iris colour prediction models have been developed by forensic groups (Ruiz et al., 2014; Spichenok et al., 2011; Walsh et al., 2011, 2013). These models predict the iris colour of an individual based on their genotype at a set of established iris pigmentation markers. The best-known of these is perhaps IrisPlex, a forensically validated genetic assay and prediction model that uses 6 iris colour polymorphisms (HERC2 rs12913832, OCA2 rs1800407, SLC24A4 rs129896399, SLC45A2 rs16891982, TYR rs1393350, IRF4 rs12203592) to estimate blue, brown and intermediate eye colour from unknown DNA samples

(Walsh et al., 2011, 2013). A second system suggested that the inclusion of additional markers from the HERC2-OCA2 region could improve the prediction of intermediate iris colour phenotypes (Ruiz et al., 2014). A third system, known as 7-Plex was designed to estimate both skin and iris colour (Spichenok et al., 2011). This system uses three SNPs (HERC2 rs12913832, SLC45A2 rs16891982, IRF4 rs12203592) associated with iris and skin colour, three SNPs associated with genetic ancestry (MC1R rs885479, ASIP rs6119471 and OCA2 rs1545397) and one SNP associated with skin colour alone (SLC24A5 rs1426654) to predict brown, not brown, not blue and green eye colour. Apart from the markers that are currently incorporated into iris prediction models, several studies have identified additional polymorphisms that may have an effect on iris colour. TYRP1 rs1408799 has been associated with the difference between ‘blue and not blue’ eyes in a Northern European population and ASIP rs6058017 may play some role in modulating brown iris colour (Kanetsky et al., 2002; Sulem et al., 2008). In addition, three new potential iris colour predictors (DSCR9 rs7277820, NPLOC4 rs9894429, LYST rs3768056) were recently identified in a genome-wide association study (Liu et al., 2010). However, as of yet, none of these markers have been incorporated into any forensic algorithms. In our sample, HERC2 rs1291832 had the strongest effect on eye colour across all three dimensions of colour space. HERC2 rs12913832 has been traditionally characterized as having a dominant/recessive mode of inheritance, where heterozygotes and homozygotes for the ancestral allele have brown or intermediate eyes and homozygotes for the derived allele have blue eyes. This was largely reflected in our sample, as both the a* and b* dimensions of colour space showed significant deviations from additivity. In contrast, the effect of HERC2 rs12913832 on the L* coordinate was largely additive, with one copy of the ancestral allele increasing the brightness in the eye by 7.143 units and two copies increasing it by 12.586 units. This suggests that the inheritance of this polymorphism in European populations is more complex than traditionally modelled. Although this marker may control the difference between blue and brown eye colour, it also has more subtle effects on the overall lightness of the iris. The additive nature of HERC2 rs12913832 is supported by recent functional studies (Cook et al., 2009). Cultured melanocyte strains that were homozygous for the ancestral allele had more melanin content than homozygotes for the derived allele, while heterozygotes had intermediate amounts. Thus, it appears that HERC2 rs12913832 may have both dominant and additive effects on iris colour variation, depending on which dimensions of colour space are being studied.

In addition to HERC2 rs12913832, our study supports previous research showing that OCA2 rs1800407, SLC45A2 rs16891982, SLC24A4 rs12896399, IRF4 rs12203592 and TYR rs1393350 are the primary determinants of iris colour in European populations (Eiberg et al., 2008; Kayser et al., 2008; Liu et al., 2010; Sturm et al., 2008; Walsh et al., 2011). All of these markers were significantly associated with iris colour in at least one dimension either before or after conditioning for HERC2 rs12913832. We were not able to identify an association between iris colour variation and any of the other markers. However, after conditioning for HERC2 rs12913832, having two copies of the derived allele at TYRP1 rs1408799 showed a borderline significant association with the L* (P = 0.054, β = -1.629) and a* (P = 0.007, β = 3.091) dimensions of colour space and having one copy of the ancestral allele at ASIP rs6058017 showed a borderline significant association (P = 0.007, β = 2.942) with the b* dimension. It is possible that our sample size was too small to pick up these associations as significant.

3.5.2 Iris Colour in South Asian Populations The South Asian sample showed considerably less diversity across CIELAB colour space than the European sample, and the vast majority of participants in this sample had eyes that would be traditionally described as brown. However, there were still a small number of participants that fell into the intermediate region of colour space. At present, very little research has been devoted to the study of iris colour in populations of South Asian ancestry. HERC2 rs12913832 is present at low frequencies in South Asia (Edwards et al., 2012). In this population, this polymorphism appears to modulate variation in brown iris colour. In a recent study on brown-eyed individuals of South Asian ancestry, one copy of the derived ‘G’ allele was found to significantly increase the L*, a* and b* dimensions of CIELAB colour space (Edwards et al., 2012). However, there was only one homozygote for the derived allele in that study. Apart from HERC2 rs12913832, the role that other iris colour markers play in South Asia has not yet been tested. Both OCA2 rs1800407 and IRF4 rs12203592 have very low minor allele frequencies (MAF) in this population and are unlikely to contribute to normal pigmentation variation. In our South Asian sample, we were able to replicate the association between HERC2 rs12913832 and iris colour variation. This marker showed significant deviations from additivity across all three dimensions of colour space. Having two copies of the derived allele increased L*

66 values by 17.870 and decreased a* values by 10.856 relative to the ancestral homozygotes. These effects are similar to what was observed in the European sample. However, there is a clear difference in the effect of HERC2 rs12913832 in the b* dimension of colour space in Europeans and South Asians. In Europeans, the derived homozygote is associated with a very strong reduction in the b* values with respect to the ancestral homozygote (~20 units). In contrast, in the South Asian sample, the derived homozygote has no significant effect on b* values. In fact, in this sample, derived homozygotes have slightly higher b* values than the ancestral homozygote. In our South Asian sample, there were only 7 individuals homozygous for the derived HERC2 rs12913832 allele. These individuals had an average b* value between 8 and 9 (depending on the camera), indicative of intermediate colour irises, in contrast to the average b* values observed in individuals homozygous for the derived allele in Europe, which is around -7, within the blue region of the colour space. Additionally, the South Asian individuals homozygous for the ancestral allele have much lower b* values (average 7-8, depending on the camera) than the European ancestral homozygotes (average 13-15, depending on the camera). This is primarily due to the fact that the South Asian individuals have darker brown irises than the Europeans, and dark brown colours have lower b* values than light brown colours. The difference in the effect of the HERC2 rs12913832 polymorphism in Europe versus South Asia strongly suggests that there are other polymorphisms modifying the effect of this marker in both populations. Our hypothesis is that the effect of HERC2 rs12913832 may be modified by other variants that are common in Europe, but not in South Asia. Exploring this would require much larger samples including substantial numbers of derived homozygotes in South Asia, to explore potential interactions of HERC2 rs12913832 with other pigmentation variants. The only other marker that had a strong effect on iris pigmentation in the South Asian sample was SLC24A5 rs1426654. SLC24A5 rs1426654 is a non-synonymous polymorphism (A111T) that results in the substitution of a guanine to an adenine in exon three of SLC24A5. This marker shows strong signals of positive selection in Europe and South Asia, and has been associated with skin pigmentation variation in South Asians and also in admixed individuals of African-European descent (Lamason et al., 2005; Stokowski et al., 2007). Functional analyses have found that homozygotes for the ancestral ‘G’ allele have 2.2 fold higher melanin content and 1.7 higher TYR activity than homozygotes for the derived ‘A’ allele (Cook et al., 2009). This is not the first study to suggest that this marker may play some role in iris pigmentation

67 variation. Beleza et al. (2013) found that this marker was strongly associated with iris colour in an admixed population of African-European ancestry. In our study, SLC24A5 rs1426654 was one of the main determinants of iris colour in South Asians. This marker showed an additive mode of inheritance for all three dimensions of colour space, with one copy of the ancestral allele decreasing the L*, a* and b* values by 2.337, 2.118 and 3.512 respectively and a second copy decreasing the values by 3.673, 3.905 and 5.882 relative to the homozygous-derived genotype. This suggests that the ancestral ‘G’ allele is responsible for darkening brown iris colour in this population. It is important to note that in our South Asian sample, SLC24A5 rs1426654 showed significant deviations from Hardy-Weinberg proportions. This is not surprising as there is well- documented population stratification in South Asia and this marker is under strong global selection (Izagirre et al., 2006; McEvoy et al., 2006; Stokowski et al., 2007). Although it is possible that the association between this marker and iris colour is a secondary effect resulting from population stratification, it is not likely. SLC24A5 rs1426654 has well-established functional effects on the synthesis of melanin and was the only marker that showed major deviations from Hardy-Weinberg proportions in this group. Although SLC24A5 rs1426654 and HERC2 rs12913832 were the primary determinants of iris colour in the South Asian sample, LYST rs3768056 also showed a much smaller association with the a* and b* dimensions of CIELAB colour space, with the derived ‘G’ allele increasing both coordinates. The association between LYST rs3768056 and iris pigmentation in the South Asian population is interesting, as this is one of the new putative pigmentation markers that was suggested for the European population (Liu et al., 2010). This marker may be a stronger predictor of iris colour in populations of non-European ancestry. In addition, having one copy of the derived ‘T’ allele at TYRP1 rs1408799 showed a borderline significant effect on the a* and b* dimensions of colour space, with the derived ‘C’ allele increasing both coordinates.

3.5.3 Iris Colour in East Asian Populations Iris colour showed the most limited distribution in the East Asian sample. Blue iris colour was completely absent in this group, and only a small number of eyes had an intermediate or green phenotype. The majority of irises fell into the region of colour space associated with various shades of brown. Average iris colour was comparable to the South Asian sample.

Very little is known about pigmentation phenotypes in East Asia. Light skin pigmentation appears to have evolved independently in Europe and East Asia, and there is very little overlap in the markers responsible for pigmentation diversity between these two populations (Eaton et al., 2015; Edwards et al., 2010; McEvoy et al., 2006; Norton et al., 2007). The majority of polymorphisms associated with light skin pigmentation in Europe and South Asia are absent in East Asia. In addition, the only putative iris pigmentation markers that have a MAF allele frequency greater than 0.05 are SLC24A4 rs129896399, DSCR9 rs7277820, NPLOC4 rs9894429, LYST 3768056 and ASIP rs6058017. Thus, the genetic basis of iris pigmentation in East Asian populations is likely very different than in other groups. The strongest determinants of iris pigmentation diversity in our sample were OCA2 rs1800414 and OCA2 rs74653330. OCA2 rs1800414 had an additive effect on all three dimensions of CIELAB colour space. One copy of the ancestral ‘A’ allele decreased the value of L*, a* and b* by 1.460, 1.318 and 1.757 respectively, and two copies of the ancestral ‘A’ allele decreased the value by 2.793, 2.831 and 4.024. This suggests that the ancestral allele for this marker plays a role in darkening iris colour. Although we did not have any participants who were homozygous for the derived ‘A’ allele for OCA2 rs74653330, having one copy of the derived allele increased both the a* and b* values. After conditioning for OCA2 rs1800414, the effect of OCA2 rs74653330 became even stronger, with heterozygotes having L*, a* and b* values that were 2.1509, 2.6713, 3.9180 higher respectively than the homozygous ancestral genotype. OCA2 rs1800414 is a non-synonymous polymorphism (His615Arg) that is present in very high frequencies in East Asia (Donnelly et al., 2012; Yuasa et al., 2007). The derived ‘G’ allele is strongly associated with lower skin melanin levels, with each copy of the ‘G’ allele decreasing the skin melanin index by approximately 0.9 units (Edwards et al., 2010). OCA2 rs74653330 is another non-synonymous (Ala481Thr) polymorphism that is largely restricted to East Asia (Eaton et al., 2015). The derived ‘T’ allele has a low frequency in this region. Each copy of the derived ‘T’ allele has been found to reduce the skin melanin index by approximately 1.9 units (Eaton et al., 2015). The OCA2 gene appears to be under strong selective pressure in East Asia (Donnelly et al., 2012; Eaton et al., 2015; Edwards et al., 2010; Lao et al., 2007). Given the role that the derived allele plays in lightening skin pigmentation for both of these markers, it is likely that their influence on iris colour is only secondary.

3.5.4 Central Heterochromia Central heterochromia is a trait that is prevalent in populations with lighter coloured irises (Larsson et al., 2011). It usually takes the form of a blue/green iris with a ring of darker pigment around the pupil. When central heterochromia is present in the iris, the darker colour is typically restricted to the pupillary zone of the iris, although it occasionally extends into the ciliary zone. To characterize the magnitude of central heterochromia in the iris, we looked at the difference in iris colour between the ciliary and pupillary zones. We used the ∆E metric to quantify this trait, as it was designed to capture perceptible differences in colour across CIELAB colour space (Kuehni and Marcus, 1979). In our sample, the degree of central heterochromia was largely population-dependent, with the European group showing the highest prevalence of this trait (Table 3-2). There was some heterochromia in the South Asian sample, however most irises had very low ∆E values. The East Asian group showed the least colour difference. As heterochromia was largely absent in our East Asian sample, we only looked at the association between the putative pigmentation markers and heterochromia in the European and South Asian samples. The only polymorphism that was associated with central heterochromia in the European sample was HERC2 rs12913832. The effect of this marker appeared to be additive, with one copy of the ancestral allele decreasing the colour difference by 1.220, and two copies of the ancestral allele decreasing the colour difference by 3.828 relative to the homozygous derived genotype. HERC2 rs12913832 was also associated with central heterochromia in the South Asian population. However, in this group, it appears to be better modelled using a dominant/recessive mode of inheritance, with one copy of the derived allele increasing the colour difference by 1.305 and two copies increasing the colour difference by 7.520 relative to the homozygous ancestral genotype. It is interesting to note that heterozygotes for this marker in the European and South Asian population show very different amounts of colour difference (Table 3-6). In the European sample, the average ∆E ranged between 6.668 and 8.128 when HERC2 rs12913832 is in the heterozygous state. In the South Asian sample, however, the average ∆E only ranged between 3.634 and 3.722. In contrast, when this marker is found in the derived homozygous state both groups showed closer amounts of colour difference. It is well known that HERC2 rs12913832 can have an effect on intermediate phenotypes. (Spichenok et al., 2011; Walsh et al., 2011, 2012). In particular, individuals who are heterozygous for this

70 marker commonly have iris colours that range from blues, to greens to browns. However, it is likely that there are other polymorphisms interacting with HERC2 rs12913832 to produce variation in central heterochromia. This is particularly evident when looking at the average ∆E in heterozygotes for this marker in the European and South Asian groups. Although heterozygotes in the European sample show substantial amounts of central heterochromia, this is not the case in the South Asian sample. Therefore, there are likely other polymorphisms modulating the effect of this marker in both of these groups. Larsson et al. (2011) identified an association between SLC24A4 rs12896399 and heterochromatic variation in a genome-wide association study performed in a sample of European ancestry living in Australia. However, we were not able to replicate this finding in our study. This may be because they looked at both the extent and spread of central heterochromia, and we only looked at colour difference. Interestingly, we did find a borderline significant effect between ∆E and TYR rs1393350 after conditioning for HERC2 rs12913832 in the European sample, with the derived ‘A’ allele increasing the amount of heterochromia in the eye. TYR rs1393350 has been previously associated with the difference between blue and green eyes (Sulem et al., 2008). Likewise, in the South Asian sample, SLC24A5 rs1426654 was associated with ∆E before conditioning for HERC2 rs12913832. However, after conditioning, this marker was no longer significant. There are a number of structural differences between the pupillary and ciliary zones, including the thickness and opacity of these regions (Mackey et al., 2011; Oyster, 1999). It is possible that markers associated with the structure of the iris may also be determinants of central heterochromia.

3.5.5 A Global View of the Genetic Basis of Iris Pigmentation Although the evolution and genetic basis of iris colour in European populations has been well-studied over the past decade, very few research groups have attempted to explore the genetic basis of iris pigmentation variation in populations of non-European ancestry. Although European populations may show the greatest diversity in eye colour phenotypes, iris colour variation extends far past the difference between blue and brown. This is especially evident when looking at the distribution of iris pigmentation in East and South Asian samples across CIE 1976 L*a*b (CIELAB) colour space. Although the majority of the irises in these populations

71 would traditionally be described as brown, these browns showed much diversity and ranged from very light to very dark in colour. Thus, studying iris colour in populations with more homogenously coloured irises will allow us to approach the study of iris colour from a number of new directions and better understand the global diversity in this trait. It is interesting to note that the markers associated with iris colour variation in all three groups have largely been associated with other pigmentary traits. SLC45A2 rs16891982, SLC24A5 rs1426654, OCA2 rs1800414 and OCA2 rs7465330 are responsible for major differences in global skin pigmentation diversity (Eaton et al., 2015; Edwards et al., 2010; Graf et al., 2005; Lamason et al., 2005; Stokowski et al., 2007). In addition, SLC24A4 rs12896399 and HERC2 rs12913832 have been associated with hair colour variation and IRF4 rs12203592 has been associated with both hair and skin pigmentation (Han et al., 2008; Sulem et al., 2007). TYR rs1126809, which is in strong linkage disequilibrium with TYR rs1393350, has also been associated with both skin and hair colour (Nan et al., 2009; Sulem et al., 2008). Thus, it is very likely that many of the iris pigmentation markers were selected for because of the effect that they had on other pigmentary characteristics, and not for their effect on iris colour variation. HERC2 rs12913832, which shows evidence of strong positive selection in regions where blue eyes dominate, may be the exception (Donnelly et al., 2012). It is also interesting to note that the markers which modulate iris colour variation in each region appear to be very different. The six markers commonly used as predictors in European populations were all significant in our European sample. However, only HERC2 rs12913832 was associated with eye colour in the South Asian sample and none of the markers were associated with eye colour in East Asia. Rather, the markers that were significantly associated with iris colour variation in the East and South Asian populations were largely polymorphisms that played a role in skin and hair pigmentation variation in those regions. These include SLC24A5 rs1426654 in the South Asian sample, and OCA2 rs1800414 and OCA2 rs7465330 in the East Asian sample. Continuing to develop a better understanding of the global distribution of iris colour variation will have a number of advantages. At present, the markers associated with intermediate eye colours are largely unknown. As a result, forensic eye colour predictor models have been found to perform poorly when applied to populations with a large proportion of green or intermediate eyes (Dembinski and Picard, 2014; Pneuman et al., 2012; Yun et al., 2014). As the

72 genetic basis of iris colour appears to be highly population-specific, it is critical to study eye colour in populations outside of Europe to get a better understanding of the genetic architecture of this trait. Determining the variants responsible for variation in brown iris colour may allow future iris predictor systems to distinguish between dark and light brown eyes in populations with more homogenous irises. This would broaden their use beyond European populations. Lastly, given that the markers associated with iris pigmentation appear to be strongly associated with skin and hair pigmentation variation, identifying novel variants associated with iris colour variation in populations of non-European ancestry may provide valuable information about the genetic basis of skin and hair colour in these regions. The development of quantitative methods of measuring iris colour has opened up many doors in pigmentation research. However, these methods have not yet been widely applied to populations of non-European ancestry. In this study, we present a new method to estimate iris colour and heterochromia based on high-resolution photographs. This method has been implemented in a Web application that can be accessed at http://iris.davidcha.ca/. Accounts can be set up for interested users by request. Using this quantitative approach, we identified one novel variant associated with iris colour in a South Asian sample and two novel variants in an East Asian sample. We suggest that future research should apply such quantitative methods to other global populations, such as African-American and Hispanic groups. In addition, as the genetic basis of central heterochromia and intermediate irises continues to be poorly understood, we suggest that using quantitative methods that allow researchers to divide the iris into separate regions may provide a new approach for studying these traits.

3.6 Tables

Table 3-1 Results of the Intra- and Inter-Rater Reliability Analysis We report the intracorrelation coefficient for a two-way mixed model for both the intra-rater and inter- rater measurements.

Intra-Rater Intracorrelation Coefficient Inter-Rater Intracorrelation Coefficient L* 0.999 0.999

a* 0.999 0.997

b* 0.997 0.992

∆E 0.985 0.985

Table 3-2 Summary Statistics for All 1448 Participants Included in the Study We show the sex, average, minimum and maximum age, L*, a*, b* and ∆E values of all 1448 participants included in the analysis (excluded individuals have been omitted from this table) for each of the two camera bodies. Participants are divided both by broad ancestry (East Asian, European and South Asian) and regional ancestry. European participants are divided into regions using the United Nations geoscheme for Europe (http://unstats.un.org/unsd/methods/m49/m49regin.htm). East Asian and South Asian participants are divided by country of origin.

Regional Camera Total Average Age Males Average L* Average a* Average b* Average ∆E Ancestry Body Number (Min, Max) (Females) (Min, Max) (Min, Max) (Min, Max) (Min, Max) 1 87 21 (18,34) 26 (61) 26.914 (19.525,36.683) 13.674 (8.832,20.391) 8.158 (0.982,17.891) 2.476 (0.638,4.748) China 2 274 22 (18,35) 78 (196) 22.182 (14.239,36.669) 12.9 (7.299,20.010) 7.289 (0.3,18.203) 2.109 (0.063,5.705) 1 2 21 (19,23) 1 (1) 27.372 (24.384,30.36) 14.311 (13.202,15.419) 7.329 (6.544,8.115) 1.762 (1.605,1.918) Japan 2 6 24 (18,29) 1 (5) 21.846 (15.43,26.468) 13.563 (9.845,15.087) 8.714 (5.543,11.411) 2.006 (0.657,2.788) East 1 21 24 (18,29) 10 (11) 27.055 (22.65,38.532) 13.732 (8.399,17.410) 8.728 (0.932,16.177) 2.251 (0.399,6.519) Korea Asian 2 52 22 (18,31) 24 (28) 21.034 (15.814,28.17) 12.56 (6.594,18.147) 7.136 (0.098,18.372) 2.203 (0.523,5.199) 1 2 21 (20,22) 1 (1) 27.08 (23.85,30.31) 12.524 (10.539,14.509) 6.531 (3.354,9.708) 2.607 (2.046,3.168) Taiwan 2 7 20 (19,23) 2 (5) 19.798 (16.502,23.522) 10.961 (7.709,13.119) 4.635 (1.36,6.767) 1.881 (1.329,2.868) 1 4 21 (19,25) 0 (4) 27.43 (24.298,30.477) 13.437 (10.063,16.724) 7.911 (3.51,12.166) 2.992 (2.311,3.861) Other 2 13 21 (18,26) 5 (8) 20.44 (14.685,26.547) 11.949 (7.921,16.724) 6 (-0.664,11.902) 1.999 (0.748,3.203) Eastern 1 53 21 (18,35) 24 (29) 45.002 (30.834,62.23) 1.055 (-8.444,18.733) 2.252 (-15.797,20.44) 9.277 (1.114,16.684) Europe 2 67 23 (18,35) 25 (42) 38.229 (18.899,53.004) 1.663 (-7.734,19.962) 1.795 (-17.65,21.611) 7.671 (0.683,16.153) Northern 1 33 23 (19,35) 11 (22) 45.279 (30.857,57.16) 0.43 (-8.225,16.027) 1.828 (-18.61,21.473) 9.129 (3.508,17.378) Europe 2 74 24 (18,35) 33 (41) 37.97 (25.269,49.584) -1.5 (-6.961,17.543) -3.406 (-15.832,22.095) 6.832 (1.702,14.721) Southern 1 47 21 (18,27) 16 (31) 38.821 (26.084,55.444) 10.894 (-6.601,22.009) 12.512 (-14.18,22.418) 6.451 (0.796,14.674) European Europe 2 52 22 (18,35) 22 (30) 33.813 (21.294,52.467) 8.942 (-6.713,21.318) 10.045 (-12.828,19.776) 6.4 (0.657,15.402) Western 1 3 20 (18,23) 0 (3) 47.903 (41.213,52.606) -3.663 (-6.977,-1.149) 2.576 (-1.76,4.923) 12.418 (10.731,13.312) Europe 2 5 24 (19,32) 2 (3) 35.911 (28.341,43.317) 4.705 (-5.882,13.278) 5.376 (-17.356,13.636) 6.456 (1.515,10.712) 1 99 22 (18,31) 36 (63) 44.622 (27.588,61.609) 1.611 (-8.035,21.032) 1.911 (-18.957,24.087) 8.447 (1.343,19.264) Other 2 187 23 (18,35) 76 (110) 38.191 (19.674,52.872) 0.436 (-7.315,19.87) 0.328 (-18.475,22.067) 7.239 (0.509,16.921) 1 5 24 (18,33) 1 (4) 27.792 (22.245,30.599) 12.977 (7.632,15.868) 6.099 (0.481,9.061) 3.257 (2.339,4.588) Bangladesh 2 12 21 (18,32) 5 (7) 25.107 (19.351,31.789) 13.419 (11.086,16.25) 8.679 (3.468,17.146) 2.663 (0.416,8.207) 1 93 21 (18,32) 32 (61) 29.56 (19.918,50.209) 13.81 (-0.912,22.671) 8.564 (-1.732,20.147) 2.807 (0.908,10.391) India 2 88 21 (18,35) 38 (50) 25.105 (15.533,37.93) 14.16 (6.26,21.256) 8.134 (-1.427,19.367) 2.117 (0.156,7.782) South 1 48 20 (18,24) 15 (33) 30.401 (21.266,50.995) 14.665 (0.429,25.925) 9.595 (-1.2,26.796) 3.204 (0.451,13.486) Pakistan Asian 2 36 20 (18,31) 9 (27) 27.53 (18.311,48.769) 14.028 (1.643,19.802) 9.148 (-1.955,18.111) 2.99 (0.561,11.616) 1 23 20 (18,21) 5 (18) 27.698 (18.31,42.544) 12.049 (7.101,18.796) 6.001 (-0.198,15.982) 2.343 (0.247,7.118) Sri Lanka 2 26 20 (18,26) 5 (21) 23.467 (16.704,40.457) 11.5 (1.746,21.599) 4.772 (-1.493,17.479) 2.317 (0.926,9.633) 1 19 20 (18,23) 6 (13) 32.627 (25.853,47.392) 13.747 (-2.94,18.521) 9.735 (2.136,21.973) 3.437 (0.378,10.773) Other 2 10 20 (18,27) 2 (8) 24.451 (18.644,30.629) 14.169 (10.034,19.247) 8.212 (3.967,16.983) 2.577 (1.003,5.219)

Table 3-3 Results of the Meta-Analysis in the European Sample (A) Before and (B) After Conditioning for the Effects of HERC2 rs12913832 We report the P-value of the association, the effect size (beta) and the P-value for Cochran's Q statistic for each additional copy of the minor allele relative to the reference genotype. After the Bonferroni correction for multiple comparisons, associations were significant if P < 0.00454. Reference genotypes marked with an asterisk (*) represent the homozygous ancestral state. Significant associations are bolded and italicized.

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) (A) GA 3.907 x 10-54 -7.143 0.041 L* AA 1.023 x 10-56 -12.586 0.023 GA 4.641 x 10-216 12.971 0.031 a* AA 5.203 x 10-137 17.985 0.039 HERC2 rs12913832 GG GA 1.872 x 10-223 18.343 0.071 b* AA 1.863 x 10-96 20.659 0.271 GA 1.098 x 10-05 -1.236 0.443 ∆E AA 3.667 x 10-15 -3.846 0.088 GA 0.796 0.221 0.252 L* AA - - - GA 0.150 1.542 0.748 a* AA - - - OCA2 rs1800407 GG* GA 0.003 4.297 0.887 b* AA - - - GA 0.852 -0.080 0.431 ∆E AA - - - GT 0.066 1.121 0.556 L* TT 0.325 0.789 0.648 GT 0.004 -2.202 0.759 a* TT 0.080 -1.759 0.596 SLC24A4 rs12896399 GG* GT 7.191 x 10-04 -3.446 0.470 b* TT 1.758 x 10-04 -5.018 0.707 GT 0.949 -0.019 0.718 ∆E TT 0.071 -0.725 0.060 GC 0.059 -1.610 0.085 L* CC 0.230 -3.818 0.026 GC 9.655 x 10-07 5.139 0.170 a* CC 0.050 7.714 0.047 SLC45A2 rs16891982 GG GC 2.069 x 10-06 6.740 0.666 b* CC 0.244 6.100 0.116 GC 0.150 -0.621 0.659 ∆E CC 0.209 -2.039 0.422 GA 0.543 -0.349 0.022 L* AA 0.029 2.604 0.077 GA 0.443 -0.556 0.356 a* AA 0.261 -1.694 0.725 TYR rs1393350 GG* GA 0.077 -1.721 0.308 b* AA 0.261 -2.258 0.883 GA 0.896 -0.038 0.247 ∆E AA 0.019 1.404 0.026

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) CT 0.206 0.802 0.791 L* TT 0.033 3.381 0.921 CT 0.008 -2.077 0.473 a* TT 0.003 -5.917 0.464 IRF4 rs12203592 CC* CT 0.013 -2.624 0.167 b* TT 1.072 x 10-04 -10.058 0.523 CT 0.041 0.652 0.967 ∆E TT 0.704 -0.305 0.634 CT 0.790 -0.156 0.238 L* TT 0.041 -1.857 0.409 CT 0.245 0.858 0.425 a* TT 0.007 3.091 0.890 TYRP1 rs1408799 CC CT 0.248 1.151 0.739 b* TT 0.083 2.669 0.447 CT 0.154 -0.420 0.381 ∆E TT 0.096 -0.763 0.134 CT 0.109 -1.035 0.305 L* TT 0.686 -0.321 0.359 CT 0.890 0.112 0.623 a* TT 0.117 -1.564 0.518 NPLOC4 rs9894429 CC* CT 0.845 -0.213 0.252 b* TT 0.040 -2.755 0.549 CT 0.489 0.225 0.245 ∆E TT 0.167 0.554 0.530 AG 0.151 -0.842 0.762 L* GG 0.438 1.010 0.015 AG 0.097 1.226 0.354 a* GG 0.814 0.387 0.142 LYST rs3768056 AA* AG 0.056 1.890 0.594 b* GG 0.904 -0.267 0.095 AG 0.165 -0.408 0.462 ∆E GG 0.875 -0.103 0.006 AG 0.185 -0.906 0.296 L* GG 0.670 -0.951 0.391 AG 0.047 1.695 0.448 a* GG 0.393 2.396 0.480 ASIP rs6058017 AA AG 0.012 2.887 0.536 b* GG 0.309 3.789 0.248 AG 0.391 -0.294 0.023 ∆E GG 0.065 2.079 0.551 (B) GA 1.767 x 10-07 3.552 0.498 L* AA - - - GA 2.203 x 10-09 -3.835 0.872 a* AA - - - OCA2 rs1800407 GG* GA 0.013 -2.389 0.624 b* AA - - - GA 0.082 0.736 0.772 ∆E AA - - -

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) GT 0.111 0.909 0.200 L* TT 0.457 0.555 0.520 GT 0.007 -1.902 0.311 a* TT 0.115 -1.453 0.464 SLC24A4 rs12896399 GG* GT 0.001 -3.205 0.201 b* TT 2.134 x 10-04 -4.725 0.617 GT 0.691 -0.118 0.439 ∆E TT 0.039 -0.803 0.039 GC 0.687 -0.327 0.176 L* CC 0.876 -0.458 0.000 GC 2.278 x 10-04 3.675 0.340 a* CC 0.258 4.105 0.001 SLC45A2 rs16891982 GG GC 9.493 x 10-05 5.458 0.968 b* CC 0.602 2.625 0.017 GC 0.647 -0.197 0.946 ∆E CC 0.542 -0.961 0.106 GA 0.907 -0.062 0.005 L* AA 0.004 3.185 0.109 GA 0.214 -0.828 0.192 a* AA 0.086 -2.367 0.999 TYR rs1393350 GG* GA 0.034 -1.974 0.201 b* AA 0.127 -2.914 0.876 GA 0.845 0.055 0.158 ∆E AA 0.005 1.614 0.033 CT 0.332 0.575 0.862 L* TT 0.027 3.220 0.688 CT 0.016 -1.745 0.561 a* TT 0.001 -5.809 0.617 IRF4 rs12203592 CC* CT 0.024 -2.282 0.651 b* TT 5.917 x 10-05 -9.916 0.997 CT 0.058 0.588 0.951 ∆E TT 0.628 -0.375 0.487 CT 0.839 -0.111 0.506 L* TT 0.054 -1.629 0.363 CT 0.254 0.776 0.820 a* TT 0.008 2.811 0.864 TYRP1 rs1408799 CC CT 0.289 1.011 0.892 b* TT 0.106 2.372 0.405 CT 0.149 -0.415 0.613 ∆E TT 0.124 -0.684 0.111 CT 0.124 -0.925 0.179 L* TT 0.698 -0.287 0.291 CT 0.844 -0.147 0.795 a* TT 0.081 -1.608 0.465 NPLOC4 rs9894429 CC* CT 0.637 -0.494 0.349 b* TT 0.029 -2.808 0.539 CT 0.399 0.267 0.194 ∆E TT 0.155 0.553 0.477

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) AG 0.166 -0.755 0.909 L* GG 0.589 0.655 0.006 AG 0.098 1.122 0.395 a* GG 0.636 0.712 0.094 LYST rs3768056 AA* AG 0.061 1.773 0.631 b* GG 0.985 0.040 0.071 AG 0.198 -0.367 0.328 ∆E GG 0.718 -0.228 0.004 AG 0.143 -0.923 0.298 L* GG 0.964 0.093 0.404 AG 0.027 1.737 0.479 a* GG 0.682 1.053 0.528 ASIP rs6058017 AA AG 0.007 2.942 0.589 b* GG 0.491 2.444 0.277 AG 0.380 -0.290 0.020 ∆E GG 0.024 2.444 0.560

Table 3-4 Results of the Meta-Analysis in the South Asian Sample (A) Before and (B) After Conditioning for the Effects of HERC2 rs12913832 We report the P-value of the association, the effect size (beta) and the P-value for Cochran's Q statistic for each additional copy of the minor allele relative to the reference genotype. After the Bonferroni correction for multiple comparisons, associations were significant if P < 0.00500. Reference genotypes marked with an asterisk (*) represent the homozygous ancestral state. Significant associations are bolded and italicized.

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) (A) AG 5.042 x 10-14 4.652 0.530 L* GG 7.535 x 10-26 17.867 0.879 AG 7.571 x 10-04 1.601 0.885 a* GG 9.703 x 10-17 -10.856 0.897 HERC2 rs12913832 AA* AG 7.073 x 10-11 4.165 0.669 b* GG 0.390 1.510 0.770 AG 8.583 x 10-10 1.305 0.074 ∆E GG 6.443 x 10-38 7.520 0.916 AG 8.901 x 10-04 -2.337 0.948 L* GG 0.002 -3.673 0.689 AG 1.400 x 10-05 -2.118 0.710 a* GG 1.538 x 10-06 -3.905 0.956 SLC24A5 rs1426654 AA AG 6.703 x 10-09 -3.512 0.789 b* GG 5.271 x 10-09 -5.881 0.921 AG 0.002 -0.775 0.527 ∆E GG 0.261 -0.476 0.379 GT 0.807 0.147 0.414 L* TT 0.793 -0.299 0.871 GT 0.147 -0.615 0.133 a* TT 0.343 -0.763 0.186 SLC24A4 GG* rs12896399 GT 0.172 -0.741 0.200 b* TT 0.523 -0.657 0.470 GT 0.325 0.210 0.874 ∆E TT 0.210 0.507 0.542 CG 0.943 -0.050 0.094 L* GG 0.607 -1.597 0.427 CG 0.409 -0.408 0.225 a* GG 0.701 0.849 0.431 SLC45A2 rs16891982 CC* CG 0.964 -0.028 0.114 b* GG 0.398 2.385 0.365 CG 0.609 0.127 0.759 ∆E GG 0.964 -0.050 0.639 GA 0.534 0.510 0.321 L* AA 0.907 -0.280 0.241 GA 0.715 -0.213 0.807 a* AA 0.437 1.315 0.121 TYR rs1393350 GG* GA 0.752 0.233 0.135 b* AA 0.572 1.217 0.170 GA 0.448 0.222 0.174 ∆E AA 0.317 -0.850 0.962

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) TC 0.116 0.946 0.170 L* CC 0.795 0.252 0.857 TC 0.009 1.106 0.226 a* CC 0.014 1.674 0.280 TYRP1 rs1408799 TT* TC 0.024 1.224 0.606 b* CC 0.007 2.366 0.671 TC 0.109 -0.344 0.027 ∆E CC 0.509 -0.229 0.940 TC 0.615 -0.315 0.382 L* CC 0.802 -0.213 0.891 TC 0.297 -0.464 0.412 a* CC 0.636 -0.284 0.877 NPLOC4 rs9894429 TT TC 0.148 -0.819 0.696 b* CC 0.163 -1.068 0.717 TC 0.198 0.283 0.097 ∆E CC 0.839 -0.061 0.359 AG 0.091 1.012 0.933 L* GG 0.292 1.145 0.603 AG 0.002 1.308 0.475 a* GG 0.046 1.544 0.168 LYST rs3768056 AA* AG 0.002 1.696 0.808 b* GG 0.064 1.836 0.241 AG 0.730 -0.074 0.655 ∆E GG 0.331 -0.380 0.630 GA 0.961 -0.031 0.844 L* AA 0.628 -0.450 0.174 GA 0.321 0.441 0.228 a* AA 0.276 -0.714 0.114 DSCR9 rs7277820 GG* GA 0.356 0.524 0.271 b* AA 0.507 -0.556 0.050 GA 0.311 -0.228 0.907 ∆E AA 0.556 -0.195 0.988 AG 0.204 0.797 0.895 L* GG 0.763 -0.356 0.655 AG 0.472 0.318 0.027 a* GG 0.661 -0.367 0.523 ASIP rs6058017 AA AG 0.248 0.655 0.171 b* GG 0.733 0.366 0.505 AG 0.406 0.185 0.168 ∆E GG 0.750 0.134 0.349 (B) AG 0.003 -1.918 0.979 L* GG 0.002 -3.255 0.637 AG 1.651 x 10-08 -2.461 0.759 a* GG 4.900 x 10-09 -4.233 0.997 SLC24A5 rs1426654 AA AG 5.237 x 10-09 -3.560 0.823 b* GG 4.738 x 10-09 -5.922 0.902 AG 0.007 -0.586 0.499 ∆E GG 0.417 -0.292 0.283

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) GT 0.896 0.071 0.093 L* TT 0.450 -0.778 0.329 GT 0.157 -0.554 0.337 a* TT 0.547 -0.447 0.471 SLC24A4 GG* rs12896399 GT 0.167 -0.757 0.196 b* TT 0.501 -0.699 0.461 GT 0.362 0.166 0.261 ∆E TT 0.399 0.291 0.782 CG 0.386 -0.542 0.019 L* GG 0.633 -1.320 0.381 CG 0.766 -0.135 0.328 a* GG 0.744 0.662 0.389 SLC45A2 rs16891982 CC* CG 0.903 -0.078 0.097 b* GG 0.404 2.365 0.370 CG 0.781 -0.059 0.357 ∆E GG 0.957 0.051 0.572 GA 0.492 0.507 0.568 L* AA 0.965 0.094 0.198 GA 0.653 -0.240 0.474 a* AA 0.494 1.059 0.089 TYR rs1393350 GG* GA 0.788 0.200 0.151 b* AA 0.577 1.204 0.174 GA 0.358 0.228 0.347 ∆E AA 0.331 -0.701 0.939 TC 0.025 1.215 0.382 L* CC 0.407 0.719 0.729 TC 0.021 0.905 0.405 a* CC 0.032 1.343 0.269 TYRP1 rs1408799 TT* TC 0.028 1.208 0.680 b* CC 0.007 2.362 0.643 TC 0.225 -0.224 0.069 ∆E CC 0.926 -0.027 0.945 TC 0.601 -0.296 0.682 L* CC 0.864 -0.130 0.751 TC 0.290 -0.432 0.630 a* CC 0.532 -0.343 0.986 NPLOC4 rs9894429 TT TC 0.177 -0.771 0.623 b* CC 0.165 -1.066 0.715 TC 0.135 0.280 0.205 ∆E CC 0.930 -0.022 0.383 AG 0.041 1.093 0.828 L* GG 0.369 0.870 0.231 AG 0.001 1.299 0.319 a* GG 0.014 1.732 0.325 LYST rs3768056 AA* AG 0.002 1.744 0.743 b* GG 0.069 1.808 0.227 AG 0.813 -0.042 0.713 ∆E GG 0.137 -0.486 0.848

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) GA 0.948 -0.037 0.658 L* AA 0.264 -0.932 0.062 GA 0.287 0.437 0.484 a* AA 0.485 -0.421 0.173 DSCR9 rs7277820 GG* GA 0.378 0.506 0.273 b* AA 0.469 -0.611 0.056 GA 0.204 -0.242 0.337 ∆E AA 0.144 -0.41 0.635 AG 0.168 0.778 0.307 L* GG 0.429 -0.840 0.882 AG 0.476 0.291 0.089 a* GG 0.884 -0.112 0.871 ASIP rs6058017 AA AG 0.278 0.621 0.145 b* GG 0.790 0.288 0.549 AG 0.315 0.191 0.580 ∆E GG 0.885 -0.052 0.759

Table 3-5 Results of the Meta-Analysis in the East Asian Sample (A) Before and (B) After Conditioning for the Effects of OCA2 rs1800414 We report the P-value of the association, the effect size (beta) and the P-value for Cochran's Q statistic for each additional copy of the minor allele relative to the reference genotype. After the Bonferroni correction for multiple comparisons, associations were significant if P < 0.00714. Reference genotypes marked with an asterisk (*) represent the homozygous ancestral state. Significant associations are bolded and italicized.

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) (A) GA 1.434 x 10-05 -1.460 0.503 L* AA 5.875 x 10-09 -2.793 0.902 GA 5.187 x 10-08 -1.319 0.493 OCA2 rs1800414 GG a* AA 1.704 x 10-16 -2.831 0.926 GA 1.137 x 10-07 -1.757 0.510 b* AA 1.308 x 10-17 -4.024 0.783 GA 0.069 1.201 0.177 L* AA - - - GA 4.310 x 10-04 1.700 0.153 OCA2 rs74653330 GG* a* AA - - - GA 1.141 x 10-04 2.556 0.063 b* AA - - - GT 0.273 -0.375 0.432 L* TT 0.902 -0.067 0.267 GT 0.088 -0.429 0.263 SLC24A4 GG* a* rs12896399 TT 0.161 -0.562 0.190 GT 0.027 -0.761 0.129 b* TT 0.018 -1.294 0.057 GA 0.186 0.477 0.296 L* AA 0.251 0.521 0.499 GA 0.140 0.395 0.984 DSCR9 rs7277820 GG* a* AA 0.087 0.577 0.468 GA 0.042 0.745 0.734 b* AA 0.024 1.041 0.609 TC 0.634 -0.161 0.506 L* CC 0.844 0.170 0.089 TC 0.782 -0.069 0.648 NPLOC4 rs9894429 TT a* CC 0.990 0.008 0.154 TC 0.442 -0.265 0.739 b* CC 0.590 -0.474 0.337 AG 0.306 0.359 0.323 L* GG 0.379 0.879 0.001 AG 0.168 0.364 0.471 LYST rs3768056 AA* a* GG 0.703 0.288 0.748 AG 0.161 0.505 0.388 b* GG 0.397 0.873 0.032

Reference Colour Significance Heterogeneity Gene Marker Genotype Beta Genotype Metric (P) (P) AG 0.338 0.332 0.796 L* GG 0.434 0.582 0.081 AG 0.670 0.109 0.754 ASIP rs6058017 AA a* GG 0.725 0.195 0.093 AG 0.921 0.035 0.682 b* GG 0.707 0.283 0.003 (B) GA 8.781 x 10-04 2.151 0.298 L* AA - - - GA 3.071 x 10-09 2.671 0.310 OCA2 rs74653330 GG* a* AA - - - GA 1.847 x 10-10 3.918 0.143 b* AA - - - GT 0.304 -0.347 0.315 L* TT 0.984 -0.011 0.373 GT 0.085 -0.420 0.138 SLC24A4 GG* a* rs12896399 TT 0.189 -0.509 0.229 GT 0.023 -0.754 0.052 b* TT 0.021 -1.213 0.112 GA 0.275 0.388 0.321 L* AA 0.495 0.306 0.468 GA 0.255 0.294 0.888 DSCR9 rs7277820 GG* a* AA 0.282 0.352 0.416 GA 0.090 0.599 0.808 b* AA 0.107 0.718 0.567 TC 0.712 -0.123 0.274 L* CC 0.796 0.220 0.132 TC 0.812 -0.058 0.259 NPLOC4 rs9894429 TT a* CC 0.908 0.071 0.255 TC 0.451 -0.250 0.726 b* CC 0.645 -0.388 0.525 AG 0.293 0.363 0.298 L* GG 0.445 0.749 0.002 AG 0.146 0.369 0.410 LYST rs3768056 AA* a* GG 0.819 0.167 0.814 AG 0.138 0.510 0.324 b* GG 0.474 0.706 0.032 AG 0.491 0.235 0.802 L* GG 0.480 0.519 0.117 AG 0.971 0.009 0.751 ASIP rs6058017 AA a* GG 0.816 0.124 0.157 AG 0.744 -0.110 0.681 b* GG 0.805 0.178 0.006

Table 3-6 Average, Minimum and Maximum ∆E Stratified by HERC2 rs12913832 Genotype We report the average, minimum and maximum ∆E values for the three possible HERC2 rs12913832 genotypes for both the European and South Asian populations. A * is used to represent the homozygous ancestral state. The East Asian sample was monomorphic for this marker, and was not included in the table. Note the substantial difference in average ∆E between the heterozygous European and South Asian samples.

HERC2 genotype Camera Average ∆E Minimum ∆E Maximum ∆E Camera 1 4.806 0.823 11.503 AA* Camera 2 4.578 0.509 9.936 Camera 1 8.128 1.394 16.527 European AG Camera 2 6.668 0.683 16.153 Camera 1 9.669 1.515 19.264 GG Camera 2 7.751 1.605 16.921 Camera 1 2.589 0.247 9.433 AA Camera 2 2.015 0.156 5.655

South Camera 1 3.634 0.957 10.391 AG Asian Camera 2 3.722 0.511 9.918 Camera 1 10.043 5.328 13.486 GG Camera 2 9.598 7.545 11.616

3.7 Figures

Figure 3-1 The Distribution of Irises Across the b* and a* (A), a* and L* (B), and b* and L* (C), Coordinates of CIE 1976 L*a*b* (CIELAB) Colour Space for the Second Camera Body East Asian participants are represented by triangles, European participants are represented by squares, and South Asian participants are represented by circles. The colour inside of each shape is equivalent to the average iris colour of the participant's wedge in CIELAB colour space. Ample dispersion across colour space can be observed in all three populations. Both cameras show very similar distributions across colour space. However, the average L*, a* and b* values from the first camera are higher than those from the second camera.

Figure 3-2 Sample Iris Wedges with Corresponding Colour Space Values A wedge with a 60-degree angle is cropped from the left side of the iris based on the scleral and pupillary boundaries that the user manually defines during iris categorization. To obtain a measurement of average iris colour, the application determines the red, green and blue (RGB) value of each pixel located in the wedge and divides the sum of the RGB values by the total number of counted pixels. The RGB values are then transformed into CIE 1976 L*a*b* (CIELAB) colour space using an illuminant of D55 and an observer angle of 2 degrees. The Web application also divides the iris into a ciliary and pupillary zone based on the user-defined collarette. In the image below, we show the pupillary and ciliary wedges for six irises of differing colour. The L*, a*, b* and ∆E values are written to the right of each iris.

3.8 References Andersen, J.D., Johansen, P., Harder, S., Christoffersen, S.R., Delgado, M.C., Henriksen, S.T., Nielsen, M.M., Sørensen, E., Ullum, H., Hansen, T., et al. (2013). Genetic analyses of the human eye colours using a novel objective method for eye colour classification. Forensic Sci. Int. Genet. 7, 508–515. Baranoski, G.V.G., and Lam, M.W.Y. (2007). Qualitative assessment of undetectable melanin distribution in lightly pigmented irides. J. Biomed. Opt. 12, 1–3. Beleza, S., Johnson, N.A., Candille, S.I., Absher, D.M., Coram, M.A., Lopes, J., Campos, J., Araújo, I.I., Anderson, T.M., and Vilhjálmsson, B.J. (2013). Genetic architecture of skin and eye color in an African-European admixed population. PLoS Genet 9, e1003372. Bito, L.Z., Matheny, A., Cruickshanks, K.J., Nondahl, D.M., and Carino, O.B. (1997). Eye color changes past early childhood. The Louisville Twin Study. Arch. Ophthalmol. 115, 659–663. Cook, A.L., Chen, W., Thurber, A.E., Smit, D.J., Smith, A.G., Bladen, T.G., Brown, D.L., Duffy, D.L., Pastorino, L., and Bianchi-Scarra, G. (2009). Analysis of cultured human melanocytes based on polymorphisms within the SLC45A2/MATP, SLC24A5/NCKX5, and OCA2/P loci. J. Invest. Dermatol. 129, 392–405. Dembinski, G.M., and Picard, C.J. (2014). Evaluation of the IrisPlex DNA-based eye color prediction assay in a United States population. Forensic Sci. Int. Genet. 9, 111–117. Donnelly, M.P., Paschou, P., Grigorenko, E., Gurwitz, D., Barta, C., Lu, R.-B., Zhukova, O.V., Kim, J.-J., Siniscalco, M., and New, M. (2012). A global view of the OCA2-HERC2 region and pigmentation. Hum. Genet. 131, 683–696. Eagle, R.C. (1988). Iris pigmentation and pigmented lesions: an ultrastructural study. Trans. Am. Ophthalmol. Soc. 86, 581–687. Eaton, K., Edwards, M., Krithika, S., Cook, G., Norton, H., and Parra, E.J. (2015). Association study confirms the role of two OCA2 polymorphisms in normal skin pigmentation variation in East Asian populations. Am. J. Hum. Biol. 520–525. Edwards, M., Bigham, A., Tan, J., Li, S., Gozdzik, A., Ross, K., Jin, L., and Parra, E.J. (2010). Association of the OCA2 polymorphism His615Arg with melanin content in east Asian populations: further evidence of convergent evolution of skin pigmentation. PLoS Genet 6, e1000867. Edwards, M., Gozdzik, A., Ross, K., Miles, J., and Parra, E.J. (2012). Technical note: quantitative measures of iris color using high resolution photographs. Am. J. Phys. Anthropol. 147, 141–149. Eiberg, H., Troelsen, J., Nielsen, M., Mikkelsen, A., Mengel-From, J., Kjaer, K.W., and Hansen, L. (2008). Blue eye color in humans may be caused by a perfectly associated founder mutation in

90 a regulatory element located within the HERC2 gene inhibiting OCA2 expression. Hum. Genet. 123, 177–187. Frank, R.N., Puklin, J.E., Stock, C., and Canter, L.A. (2000). Race, iris color, and age-related macular degeneration. Trans. Am. Ophthalmol. Soc. 98, 109–117. Graf, J., Hodgson, R., and Van Daal, A. (2005). Single nucleotide polymorphisms in the MATP gene are associated with normal human pigmentation variation. Hum. Mutat. 25, 278–284. Han, J., Kraft, P., Nan, H., Guo, Q., Chen, C., Qureshi, A., Hankinson, S.E., Hu, F.B., Duffy, D.L., and Zhao, Z.Z. (2008). A genome-wide association study identifies novel alleles associated with hair color and skin pigmentation. PLoS Genet, e1000074. Imesch, P.D., Bindley, C.D., Khademian, Z., Ladd, B., Gangnon, R., Albert, D.M., and Wallow, I.H. (1996). Melanocytes and iris color. Electron microscopic findings. Arch. Ophthalmol. 114, 443–447. Izagirre, N., García, I., Junquera, C., de la Rúa, C., and Alonso, S. (2006). A scan for signatures of positive selection in candidate loci for skin pigmentation in humans. Mol. Biol. Evol. 23, 1697–1706. Kanetsky, P.A., Swoyer, J., Panossian, S., Holmes, R., Guerry, D., and Rebbeck, T.R. (2002). A polymorphism in the agouti signaling protein gene is associated with human pigmentation. Am. J. Hum. Genet. 70, 770–775. Kayser, M., Liu, F., Janssens, A.C.J., Rivadeneira, F., Lao, O., van Duijn, K., Vermeulen, M., Arp, P., Jhamai, M.M., and van IJcken, W.F. (2008). Three genome-wide association studies and a linkage analysis identify HERC2 as a human iris color gene. Am. J. Hum. Genet. 82, 411–423. Kuehni, R.G., and Marcus, R.T. (1979). An experiment in visual scaling of small color differences*. Colour Res. Appl. 4, 83–91. Lamason, R.L., Mohideen, M.-A.P., Mest, J.R., Wong, A.C., Norton, H.L., Aros, M.C., Jurynec, M.J., Mao, X., Humphreville, V.R., and Humbert, J.E. (2005). SLC24A5, a putative cation exchanger, affects pigmentation in zebrafish and humans. Science 310, 1782–1786. Lao, O., De Gruijter, J., van Duijn, K., Navarro, A., and Kayser, M. (2007). Signatures of positive selection in genes associated with human skin pigmentation as revealed from analyses of single nucleotide polymorphisms. Ann. Hum. Genet. 71, 354–369. Larsson, M., Duffy, D.L., Zhu, G., Liu, J.Z., Macgregor, S., McRae, A.F., Wright, M.J., Sturm, R.A., Mackey, D.A., Montgomery, G.W., et al. (2011). GWAS findings for human iris patterns: associations with variants in genes that influence normal neuronal pattern development. Am. J. Hum. Genet. 89, 334–343. Liu, F., Wollstein, A., Hysi, P.G., Ankra-Badu, G.A., Spector, T.D., Park, D., Zhu, G., Larsson, M., Duffy, D.L., Montgomery, G.W., et al. (2010). Digital quantification of human eye color highlights genetic association of three new loci. PLoS Genet 6, e1000934.

Mackey, D.A., Wilkinson, C.H., Kearns, L.S., and Hewitt, A.W. (2011). Classification of iris colour: review and refinement of a classification schema. Clin. Experiment. Ophthalmol. 39, 462–471. McEvoy, B., Beleza, S., and Shriver, M.D. (2006). The genetic architecture of normal variation in human pigmentation: an evolutionary perspective and model. Hum. Mol. Genet. 15, R176– R181. McLaren, K. (1976). XIII—The development of the CIE 1976 (L* a* b*) uniform colour space and colour‐difference formula. J. Soc. Dye. Colour. 92, 338–341. Mitchell, P., Smith, W., and Wang, J.J. (1998). Iris color, skin sun sensitivity, and age-related maculopathy. The Blue Mountains Eye Study. Ophthalmology 105, 1359–1363. Nan, H., Kraft, P., Hunter, D.J., and Han, J. (2009). Genetic variants in pigmentation genes, pigmentary phenotypes, and risk of skin cancer in Caucasians. Int. J. Cancer 125, 909–917. Norton, H.L., Kittles, R.A., Parra, E., McKeigue, P., Mao, X., Cheng, K., Canfield, V.A., Bradley, D.G., McEvoy, B., and Shriver, M.D. (2007). Genetic evidence for the convergent evolution of light skin in Europeans and East Asians. Mol. Biol. Evol. 24, 710–722. Oyster, C.W. (1999). The human eye: structure and function (Sunderland, Massachusetts: Sinauer Associates, Inc.). Peles, D.N., Hong, L., Hu, D.-N., Ito, S., Nemanich, R.J., and Simon, J.D. (2009). Human iridal stroma melanosomes of varying pheomelanin contents possess a common eumelanic outer surface. J. Phys. Chem. B 113, 11346–11351. Pietroni, C., Andersen, J.D., Johansen, P., Andersen, M.M., Harder, S., Paulsen, R., Børsting, C., and Morling, N. (2014). The effect of gender on eye colour variation in European populations and an evaluation of the IrisPlex prediction model. Forensic Sci. Int. Genet. 11, 1–6. Pneuman, A., Budimlija, Z.M., Caragine, T., Prinz, M., and Wurmbach, E. (2012). Verification of eye and skin color predictors in various populations. Leg. Med. 14, 78–83. Prota, G., Hu, D.-N., Vincensi, M.R., McCormick, S.A., and Napolitano, A. (1998). Characterization of melanins in human irides and cultured uveal melanocytes from eyes of different colors. Exp. Eye Res. 67, 293–299. Rebbeck, T.R., Kanetsky, P.A., Walker, A.H., Holmes, R., Halpern, A.C., Schuchter, L.M., Elder, D.E., and Guerry, D. (2002). P gene as an inherited biomarker of human eye color. Cancer Epidemiol. Biomarkers Prev. 11, 782–784. Ruiz, Y., Phillips, C., Gomez-Tato, A., Alvarez-Dios, J., Casares de Cal, M., Cruz, R., Maroñas, O., Söchtig, J., Fondevila, M., Rodriguez-Cid, M.J., et al. (2014). Further development of forensic eye color predictive tests. Forensic Sci. Int. Genet. 7, 28–40.

Sidhartha, E., Nongpiur, M.E., Cheung, C.Y., He, M., Wong, T.Y., Aung, T., and Cheng, C.-Y. (2014). Relationship between iris surface features and angle width in Asian eyes. Invest. Ophthalmol. Vis. Sci. 55, 8144–8148. Spichenok, O., Budimlija, Z.M., Mitchell, A.A., Jenny, A., Kovacevic, L., Marjanovic, D., Caragine, T., Prinz, M., and Wurmbach, E. (2011). Prediction of eye and skin color in diverse populations using seven SNPs. Forensic Sci. Int. Genet. 5, 472–478. Stokowski, R.P., Pant, P.K., Dadd, T., Fereday, A., Hinds, D.A., Jarman, C., Filsell, W., Ginger, R.S., Green, M.R., and van der Ouderaa, F.J. (2007). A genomewide association study of skin pigmentation in a South Asian population. Am. J. Hum. Genet. 81, 1119–1132. Sturm, R.A., and Larsson, M. (2009). Genetics of human iris colour and patterns. Pigment Cell Melanoma Res. 22, 544–562. Sturm, R.A., Duffy, D.L., Zhao, Z.Z., Leite, F.P.N., Stark, M.S., Hayward, N.K., Martin, N.G., and Montgomery, G.W. (2008). A single SNP in an evolutionary conserved region within intron 86 of the HERC2 gene determines human blue-brown eye color. Am. J. Hum. Genet. 82, 424– 431. Sulem, P., Gudbjartsson, D.F., Stacey, S.N., Helgason, A., Rafnar, T., Magnusson, K.P., Manolescu, A., Karason, A., Palsson, A., and Thorleifsson, G. (2007). Genetic determinants of hair, eye and skin pigmentation in Europeans. Nat. Genet. 39, 1443–1452. Sulem, P., Gudbjartsson, D.F., Stacey, S.N., Helgason, A., Rafnar, T., Jakobsdottir, M., Steinberg, S., Gudjonsson, S.A., Palsson, A., and Thorleifsson, G. (2008). Two newly identified genetic determinants of pigmentation in Europeans. Nat. Genet. 40, 835–837. Visser, M., Kayser, M., and Palstra, R.-J. (2012). HERC2 rs12913832 modulates human pigmentation by attenuating chromatin-loop formation between a long-range enhancer and the OCA2 promoter. Genome Res. 22, 446–455. Wakamatsu, K., Hu, D.-N., McCormick, S.A., and Ito, S. (2008). Characterization of melanin in human iridal and choroidal melanocytes from eyes with various colored irides. Pigment Cell Melanoma Res. 21, 97–105. Walsh, S., Liu, F., Ballantyne, K.N., von Oven, M., Lao, O., and Kayser, M. (2011). IrisPlex: a sensitive DNA tool for accurate prediction of blue and brown eye colour in the absence of ancestry information. Forensic Sci. Int. Genet. 5, 170–180. Walsh, S., Wollstein, A., Liu, F., Chakravarthy, U., Rahu, M., Seland, J.H., Soubrane, G., Tomazzoli, L., Topouzis, F., and Vingerling, J.R. (2012). DNA-based eye colour prediction across Europe with the IrisPlex system. Forensic Sci. Int. Genet. 6, 330–340. Walsh, S., Liu, F., Wollstein, A., Kovatsi, L., Ralf, A., Kosiniak-Kamysz, A., Branicki, W., and Kayser, M. (2013). The HIrisPlex system for simultaneous prediction of hair and eye colour from DNA. Forensic Sci. Int. Genet. 7, 98–115.

Wielgus, A.R., and Sarna, T. (2005). Melanin in human irides of different color and age of donors. Pigment Cell Res. 18, 454–464. Wilkerson, C.L., Syed, N.A., Fisher, M.R., Robinson, N.L., Wallow, I.H., and Albert, D.M. (1996). Melanocytes and iris color. Light microscopic findings. Arch. Ophthalmol. 114, 437– 442. Yuasa, I., Umetsu, K., Harihara, S., Kido, A., Miyoshi, A., Saitou, N., Dashnyam, B., Jin, F., Lucotte, G., and Chattopadhyay, P. (2007). Distribution of two Asian-related coding SNPs in the MC1R and OCA2 genes. Biochem. Genet. 45, 535–542. Yun, L., Gu, Y., Rajeevan, H., and Kidd, K.K. (2014). Application of six IrisPlex SNPs and comparison of two eye color prediction systems in diverse Eurasia populations. Int. J. Legal Med. 128, 447–453.

Chapter 4: Analysis of Iris Surface Features in Populations of Diverse Ancestry

Author Contributions: M. Edwards participated in the design of the study, collected the data, conducted the data analysis and drafted the manuscript. D. Cha developed the iris structure program and assisted with data analysis. S. Krithika and M. Johnson assisted with data collection and molecular laboratory work. E.J. Parra conceived the study, participated in the design of the study, coordinated the study and helped draft the manuscript.

This Chapter Has Been Previously Published As: Edwards, M., Cha, D., S. Krithika., Johnson, M., Parra, E.J. (2016) Analysis of iris surface features in populations of diverse ancestry. R. Soc. Open Sci. 3, 150424.

94 95

4.1 Abstract There are many textural elements that can be found in the human eye, including Fuchs’ crypts, Wolfflin nodules, pigment spots, contraction furrows and conjunctival melanosis. Although iris surface features have been well-studied in populations of European ancestry, the worldwide distribution of these traits is poorly understood. In this paper, we develop a new method of characterizing iris features from photographs of the iris. We then apply this method to a diverse sample of East Asian, European and South Asian ancestry. All five iris features showed significant differences in frequency between the three populations, indicating that iris features are largely population dependent. Although none of the features were correlated with each other in the East and South Asian groups, Fuchs’ crypts were significantly correlated with contraction furrows and pigment spots and contraction furrows were significantly associated with pigment spots in the European group. The genetic marker SEMA3A rs10235789 was significantly associated with Fuchs’ crypt grade in the European, East Asian and South Asian samples and a borderline association between TRAF3IP1 rs3739070 and contraction furrow grade was found in the European sample. The study of iris surface features in diverse populations may provide valuable information of forensic, biomedical, and ophthalmological interest

4.2 Introduction The human iris is a complex tissue consisting of many different regions and strata. A healthy human eye typically has five different layers. The posterior-most layer is called the iris pigment epithelium (IPE). This layer is tightly packed with cuboidal melanin-rich melanocytes in all healthy individuals and does not contribute significantly to variation in iris colour or structure (Eagle, 1988; Peles et al., 2009; Wilkerson et al., 1996). Just above the IPE are two muscle layers, known as the sphincter muscle and the dilator muscle (Eagle, 1988). The two layers that are responsible for most of the variation between individuals are the anterior border layer and the stromal layer. The anterior border layer is made up primarily of fibroblasts and melanocytes (Eagle, 1988; Imesch et al., 1996; Oyster, 1999). By contrast, the stroma is a loose mesh of collagen fibres, melanocytes, fibroblasts and clump cells. The surface of the eye can also be divided into two regions: the pupillary zone and ciliary zone (Oyster, 1999). These regions are bounded by a ring of tissue known as the collarette, which is a product of the reabsorption of the pupillary membrane during development. There are differences in thickness between these two zones, which leads to variation in colour and structure (Oyster, 1999). There are many textural elements that can be found in the healthy human eye. These include Fuchs’ crypts, Wolfflin nodules, pigment spots, contraction furrows and conjunctival melanosis (Figure 4-1). Fuchs’ crypts are diamond-shaped lacunae in the anterior border layer of the iris, which first arise during the reabsorption of the pupillary membrane (Purtscher, 1965). Wolfflin nodules are small bundles of collagen that accumulate along the outer edge of the iris (Donaldson, 1961; Williams, 1981). Pigment spots are small regions of hyper-pigmentation in the anterior-border layer. They may be superficial (freckles) or distort the underlying stromal layer (nevi) [Eagle, 1988; Harbour et al., 2004; Rennie, 2012]. Lastly, contraction furrows are folds that fall in rings around the outer edge of the iris (Eagle, 1988). They are believed to be the product of the dilation and the contraction of the pupil. Some irises may also show conjunctival melanosis, which is pigment spotting that can be found on the sclera surrounding the iris (Damato and Coupland, 2008). This tends to be more common in populations with darker irises. Considerable research has been devoted to iris pigmentation variation (Eiberg et al., 2008; Liu et al., 2010; Sturm et al., 2008). However, very few studies have attempted to look at global variation in iris surface features. Although the functional consequences of these features remain largely unknown, they have become a topic of significant forensic, biomedical, and

97 ophthalmological interest. From a forensics perspective, a number of markers have been identified over the past 10 years that are capable of predicting pigmentation characteristics, such as hair, skin and iris colour, from crime scene DNA samples (Walsh et al., 2011). It has been suggested that some of the iris surface features, such as Wolfflin nodules and pigment spots, may have an influence on the perception of overall iris colour (Liu et al., 2010; Mackey et al., 2011). Therefore, a better understanding of the genetic basis of iris surface features may lead to improved eye colour predictor algorithms. Photographs of the iris may also represent a cost- effective alternative to more expensive ophthalmological procedures. Sidhartha et al. (2014a, 2014b) recently characterized contraction furrow and Fuchs’ crypt grade in a Malaysian population living in Singapore. They found that both features were associated with iris thickness and/or the degree of iris angle closure. Thus, it may be possible to predict patient’s disease risk from iridial surface features. Lastly, although very little is known about the pathology of iris texture, there is some evidence that these features may influence individual health and well- being. In particular, nevi and pigment spots may be a risk marker for uveal melanoma (Holly et al., 1990; Horn et al., 1994). As the textural elements in the iris have the potential to be of interest to many different disciplines, it has become imperative to develop a better understanding of the worldwide frequency and genetic basis of these traits. At present, iris features have been primarily studied in populations of European ancestry (Larsson and Pedersen, 2004; Larsson et al., 2011; Sturm and Larsson, 2009). Very few studies have focused on non-European populations, and these studies have indicated that there are differences in the distribution of iris features between major population groups (Qiu et al., 2005; Quillen et al., 2011). When iris texture was examined in Portuguese, Cape Verdean, and Brazilian populations, for example, increasing European biogeographical ancestry was significantly associated with a greater number of pigment spots, Fuchs’ crypts, and contraction furrows (Quillen et al., 2011). This paper has four primary goals: 1/ To develop a method for describing iris surface features in populations of diverse ancestry; 2/ To characterize global differences in iris features across European, East Asian and South Asian populations; 3/ To look at correlations between the structural elements within each of the populations; 4/ To look at the association between genetic markers that have been associated with iris surface features in European populations and our iris feature measurements.

4.3 Materials and Methods 4.3.1 Participant Recruitment Between 2012 and 2014, 1773 healthy volunteers of East Asian, European and South Asian ancestry were recruited at the University of Toronto to participate in a study looking at global pigmentation diversity. All participants ranged in age between 18 and 35 years, and were recruited using online and print advertisements that targeted members of the University of Toronto student community. A personal questionnaire asking about each participant’s maternal and paternal grandparents was administered in order to assess geographical ancestry. Individuals who stated that all of their grandparents came from Japan, Korea, China or Taiwan were categorized as East Asian, and individuals who stated that their grandparents originated in Bangladesh, Pakistan, Sri Lanka or India were categorized as South Asian. Participants were defined as European if their grandparents came from any region in Europe, other than Turkey. Only individuals who stated that all relatives came from the same general geographical region (i.e. East Asia) were included in the analysis. Admixed participants who came from two different regions (i.e. East Asian and Europe) were eliminated. When information about the grandparents was unknown, the self-reported ancestry of both parents was used to assess geographical ancestry. In total, our study included 623 participants of European ancestry, 475 participants of East Asian ancestry and 367 participants of South Asian ancestry. The remaining 308 participants were excluded from the analysis. In addition to determining ancestry, the personal questionnaire was also used to ensure that each participant was healthy, and had not been previously diagnosed with any ocular pigmentation-related disorders. Lastly, participants were asked to provide a self-assessment of their iris colour using the Fitzpatrick Phototype Scale (Fitzpatrick, 1975).

4.3.2 Genotyping A 2-mL saliva sample was taken from each participant using the Oragene˖DNA (OG- 500) collection kit (DNA Genotek, Canada). All participants were instructed not to eat, drink or smoke for at least 30 min prior to their appointment in order to ensure maximal sample purity. DNA was then isolated from each sample using the protocol provided by the manufacturer.

We selected four markers for genotyping that have either been directly associated with, or are purported to be associated with, iris texture in European populations (Larsson et al., 2011; Liu et al., 2010). These include TRAF3IP1 rs3739070 (contraction furrows), SEMA3A rs10235789 (crypts), DSCR9 rs7277820 (Wolfflin nodules) and HERC1 rs11630290 (pigment spots). All DNA samples were sent to LGC Genomics (United States) for genotyping. LGC Genomics uses a KASP based genotyping method that combines allele-specific amplification with FRET (fluorescent resonance energy transfer) technology. Twenty-nine samples were included as blind duplicates and 14 samples were included as blanks in order to check the quality of the genotyping results. The concordance rate for both blind duplicates and blanks was 100%.

4.3.3 Acquisition and Processing of Iris Photographs A photograph of each participant’s right iris was acquired using a Miles Research Professional Iris Camera (Miles Research, United States). This camera consists of a FujiFilm Finepix S3 Pro 12-Megapixel DSLR mounted on a 105-mm Nikkor macro lens. All photographs were taken in RAW format with an aperture of f19, a shutter speed of 1/125” and an ISO of 200. A biometric coaxial cable was used to deliver light to the iris at a constant temperature in order to maintain colour and brightness fidelity and reduce the impact of ambient light. A Krypton 2.33 V light bulb was used as a focusing light, which allowed participants’ pupils to become adjusted to a standardized light source. A chin rest and camera mount was used to ensure that each photograph was taken from a standard distance and angle. Photographs were converted to JPEG format and resized from 3043 by 2036 to 1200 by 803 pixels using Adobe Camera Raw in Adobe Photoshop CS5 (Adobe Systems Incorporated, United States). The white balance was set to flash, the contrast and blacks levels were set to zero, and the default values were maintained for all remaining settings.

4.3.4 Iris Analysis One of the authors (D.C.) designed a web-based application in order to accurately and reliably characterize iris texture. The web application can be accessed at http://iris.davidcha.ca/. Accounts can be set up for interested users by request. This program consists of 11 steps, which take approximately 2 min to complete for each iris. In the first step, the user is required to note

100 whether or not the iris is significantly obstructed by eyelids, eyelashes or reflections (to the extent that any of the five structures of interest cannot be accurately characterized). Steps 2 to 6 have the user identify the approximate centre point of the iris, the approximate centre point of the pupil, and then draw a best fit circle around the scleral boundary, the collarette, and the pupillary boundary. This allows the program to separate the iris into a ciliary zone and pupillary zone (Figure 4-2), and divide the iris into four different quadrants based on the centre of the iris (Figure 4-3). In Step 7, the user must click on all of the pigment spots present in the iris. The iris is magnified by 1.5× times for this step, in order to more easily distinguish between pigment spots and other iridial textural elements. For each iris, the program records the number of pigment spots, as well as the quadrant in which each pigment spot is found. In step 8, the user notes the presence and extension of contraction furrows. A rotating line, fixed at the centre point of the iris, is used to help the user determine whether or not the furrows cover more than 180˚ of the iris. In order to facilitate the identification of furrows, the iris is magnified by 1.5× for this step. The user also notes in which quadrants, if any, the furrows are found. Wolfflin nodules are characterized in a manner similar to contraction furrows. In step 9, a rotating line is fixed at the centre point of the iris, and the user must determine whether nodules are present, and if so, whether or not they extend around more than 180˚ of the iris. As with contraction furrows, the user must also note in which quadrants the Wolfflin nodules can be found. In step 10, the user clicks on the outermost edge of all of the crypts found in the iris. Crypts that originate from the collarette are characterized as either ‘small’ or ‘large’. Large crypts are defined as those that extend from the collarette into more than 50% of the ciliary zone. Small crypts are defined as those that do not extend into more than 50% of the ciliary zone. This is calculated automatically by the program. The program also records the quadrant in which each crypt is found. Crypts that do not originate from the collarette are manually defined as small. The iris is magnified by 1.5× for this step in order to increase the visibility of crypts in darker irises. For the last step, the user must note whether or not there are any pigment rings or pigment spotting on the visible scleral region of the eye. The iris is obscured for this step, and only the sclera is visible.

101

After analysis of all 1465 irises, the program was used to output an Excel spreadsheet which contained the category for each of the five surface features, the quadrants in which these features were found and the diameter of the iris in pixels. Detailed information about the position of each of the crypts and pigment spots identified could also be retrieved. All initial iris categorizations were carried out by M.E.

4.3.5 Characterization of Iris Structure We developed iris feature categories that could capture variation in diverse populations with different frequencies of iris colour. These were primarily based on categorization systems that had been developed in prior studies (Larsson and Pedersen, 2004; Larsson et al., 2011; Sidhartha et al., 2014a, 2014b; Sturm and Larsson, 2009). However, they were modified to account for the pigmentation diversity present in our sample set. Pigment spots were initially measured using 4 grades: 1, no pigment spots; 2, between one and two pigment spots; 3, between three and five pigment spots; 4, more than five pigment spots. However, grades 3 and 4 were later collapsed into a single grade to account for the lower frequency of pigment spots in the East and South Asian groups compared to the European group (Figure 4-4). As it is not possible to reliably differentiate between pigment spots and nevi in photographs, we chose pigment spot categories that attempted to capture the overall magnitude of spotting in the iris. Both contraction furrows and Wolfflin nodules were measured using a grading system that took into account the extension of these structures around the iris. Contraction furrows were measured across three grades: 1, no contraction furrows; 2, contraction furrows that extend less than 180° around the iris and 3, contraction furrows that extend more than 180° around the iris (Figure 4-5). Wolfflin nodules were also measured using three grades: 1, no Wolfflin nodules; 2, Wolfflin nodules that extend less than 180° around the iris and 3, Wolfflin nodules that extend more than 180° around the iris (Figure 4-6). Crypts were measured across a four grade system that attempted to capture the size and overall grade of crypts in the iris: 1, no crypts; 2, only small crypts centred around the collarette; 3, at least one large crypt located in fewer than three quadrants of the iris and 4, at least three large crypts located in three or more quadrants of the iris (Figure 4-7). Self-described iris colour was measured using the categories developed for the Fitzpatrick Phototype Scale (Figure 4-8) [Fitzpatrick, 1975]. These categories emphasize the intensity, rather than the shade, of the eye: 1, light blue, green or grey; 2, blue, green or grey; 3, hazel or light brown; 4, dark

102 brown and 5, brownish black. Lastly, conjunctival melanosis was characterized using a presence or absence schema (Figure 4-9).

4.3.6 Statistical Analysis Unless otherwise noted, statistical analyses were carried out using IBM Statistics SPSS (v. 20.0, SPSS Incorporated, United States). Correlations between the ordinal iris feature categories and iris colour were tested using Goodman and Kruskal’s gamma statistic, a measure of rank correlation typically used for ordinal traits. We report both the G-value and p-value for each of the correlations. Surface features were considered to be significantly correlated with each other or with iris colour if p<0.05. The relationships between iris feature and sex, age and iris width (the diameter of the best fit circle around the outer border of the iris) were examined using ordinal regression. The assumptions of proportional odds and goodness of fit were tested. Differences in iris feature frequency between the three sample sets were tested using the chi- square test. For features that were significantly different, additional pairwise comparisons between populations were conducted using an independent samples t-test with a Bonferonni correction for multiple comparisons. After correction, differences between samples were significant if p<0.0167. Differences in the width of the iris between the three populations were tested using a one-way ANOVA. Normality was tested using Q–Q plots and the Levene Statistic was checked before running the analysis. Associations between the four polymorphisms of interest and the iris feature categories were evaluated using ordinal regression in each population, including other relevant covariates in the analyses (e.g. other iris structures, iris width or sex). The assumptions of proportional odds and goodness of fit were tested before running the analysis. Genotype deviations from Hardy– Weinberg proportions were evaluated using the Court Lab Calculator (Court, 2008). Four months after the initial iris classification, intra-rater reliability was assessed on 40 irises by M.E. and inter-rater reliability was assessed on 40 irises by D.C. using the linear weighted kappa statistic. This was done using the web application provided by GraphPad software (GraphPad, United States).

103

4.4 Results A total of 475 East Asian, 623 European and 367 South Asian irises were evaluated using the iris structure program (Table 4-1). Of these, 14 irises (eight East Asian, four European and two South Asian) were judged to be too obscured or blurry to accurately characterize, and were excluded from the analysis. One additional participant of South Asian ancestry with a pigmentation-related ocular disorder (albinism) was also removed from the study. None of the remaining participants reported ocular disorders. However, as medical history was self-reported, it is possible that additional participants with ocular disorders may have been incorporated into the study. The intra- and inter-rater reliability was assessed using the kappa statistic (for additional information, see Material and Methods section) [Viera and Garrett, 2005]. The intra- and inter-rater reliabilities were good for all five structures (intra-rater kappa values; crypt κ=0.881, furrow κ=0.835, nodule κ=0.884, pigment spot κ=0.889, melanosis κ=0.827; inter-rater kappa values; crypt κ=0.809, furrow κ=0.857, nodule κ=1.000, pigment spot κ=0.898, melanosis κ=0.867). The width of the iris was significantly different across the three sample sets (F=200.161, p<0.001), with East Asians having the smallest iris widths (mean=376.72 pixels), followed by South Asians (mean=384.92) and then Europeans (mean=394.45 pixels). Within the European sample set, there was a significant negative correlation between the grade of Fuchs’ crypts and the extension of contraction furrows (G=−0.474, p<0.001) and a significant, but weaker, negative correlation between the grade of Fuchs’ crypts and the number of pigment spots (G=−0.238, p<0.001). There was also a significant, positive correlation between the extension of furrows and the number of pigment spots (G=0.218, p=0.007). No significant correlations between the five iridial structures were found in either the South or East Asian sample sets. In the European sample set, self-reported darker iris colour showed a significant positive correlation with the extension of contraction furrows (G=0.461, p<0.001) and a significant negative correlation with Wolfflin nodules (G=−0.409, p<0.001). In the South Asian sample set, in contrast with what was observed in the European sample set, darker iris colour showed a significant negative correlation with the extension of contraction furrows (G=−0.233, p=0.041). Sex was significantly associated with crypt grade in the East Asian (Nagelkerke R2=0.013, p=0.019), European (Nagelkerke R2=0.022, p<0.001) and South Asian (Nagelkerke

104

R2=0.037, p<0.001) samples. In all three groups, males had a significantly higher crypt grade than females (East Asian male/female OR=1.54, European male/female OR=1.72, South Asian male/female OR=2.07). Sex was not associated with any of the other four iridial structures. Iris diameter was significantly associated with a higher number of pigment spots in the European (Nagelkerke R2=0.012, p=0.009) and South Asian sample sets (Nagelkerke R2=0.018, p=0.041). Iris diameter was also significantly associated with more extended Wolfflin nodules in Europeans (Nagelkerke R2=0.021, p=0.001). No associations were found between age and iris structure in any of the three groups. All five structures examined showed significant differences in frequency between the three sample sets (Fuchs’ crypts: χ2=67.388, p<0.001; contraction furrows: χ2=186.819, p<0.001; pigment spots: χ2=260.587, p<0.001; Wolfflin nodules: χ2=350.627, p<0.001; conjunctival melanosis: χ2=273.177, p<0.001). Additional pairwise independent sample T-tests with a Bonferonni correction were run in order to determine the source of these differences. Europeans had a significantly higher grade of Fuchs’ crypts (T=8.333, p<0.001), more extended contraction furrows (T=10.802, p<0.001), more pigment spots (T=14.686, p<0.001) and more extended Wolfflin nodules (T=18.028, p<0.001) than individuals of East Asian ancestry and a significantly higher grade of crypts (T=2.961, p=0.003), greater number of pigment spots (T=15.571, p<0.000) and more extended Wolfflin nodules (T=16.345, p<0.001) than individuals of South Asian ancestry. The South Asian and East Asian samples had a higher proportion of individuals with conjunctival melanosis than the European sample (South Asian: T=15.954, p<0.001; East Asian: T=6.674, p<0.001). There was no significant difference in contraction furrow extension between the European and South Asian groups (T=−0.306, p=0.759). Individuals of South Asian ancestry had a significantly higher grade of crypts (T=4.231, p<0.001), more extended contraction furrows (T=10.474, p<0.001) and a higher frequency of conjunctival melanosis (T=6.674, p<0.001) than individuals of East Asian ancestry. However, there was no significant difference in the number of pigment spots (T=−1.527, p=0.127) and the presence of Wolfflin nodules (T=2.372, p=0.018) between these two groups. Self-described iris colour was significantly different (χ2=892.674, p<0.001) across all three samples. Individuals of East Asian ancestry had significantly darker eyes than individuals of South Asian (T=4.327, p<0.001) and European (T=40.776, p<0.001) ancestry and individuals of South Asian ancestry had significantly darker eyes (T=32.724, p<0.001) than individuals of European ancestry.

105

Genotype proportions were in agreement with the expected Hardy–Weinberg proportions, with the exception of moderate deviations observed for DSCR9 rs7277820 in the European (χ2=5.373, p=0.020) and South Asian (χ2=3.930, p=0.047) sample sets and HERC1 rs11630290 in the South Asian sample set (χ2=5.224, p=0.022). We looked at the association between SEMA3A rs10235789 and Fuchs’ crypts, TRAF3IP1 rs3739070 and contraction furrows, DSCR9 rs7277820 and Wolfflin nodules, and HERC1 rs11630290 and pigment spots. In the statistical tests, we included as covariates the variables that were significantly associated with each iris structure in our exploratory analyses (e.g. other iris structures, iris width, self-reported iris colour and sex). Fuchs’ crypt grade was significantly associated with SEMA3A rs10235789 in all three sample sets (Table 4-2). In the European sample set, having one copy of the derived ‘C’ allele significantly increased the odds of having a higher grade of crypts by 1.507 (p=0.023) and having two copies of the derived ‘C’ allele significantly increased the odds of having a higher grade of crypts by 2.203 (p<0.001). In the South Asian sample, having one copy of the derived ‘C’ allele significantly increased the odds of having a higher grade of crypts by 2.206 (p<0.001) and having two copies of the derived ‘C’ allele significantly increased the odds of having a higher grade of crypts by 2.721 (p=0.011). In the East Asian sample set, having one copy of the derived ‘C’ allele significantly increased the odds of having a higher grade of crypts by 1.679 (p=0.018). As there were fewer than five individuals with two copies of the derived ‘C’ allele, they were eliminated from this analysis. The association between contraction furrow grade and TRAF3IP1 rs3739070 was borderline significant in the European group but not in the East and South Asian sample sets (Table 4-2). Having two copies of the derived ‘A’ allele increased the odds of having more extended contraction furrows by 6.385 (p=0.051). However, having only one copy of the ‘A’ allele did not significantly (p=0.333) increase the odds. Both pigment spot grade and Wolfflin nodule grade were not significantly associated with their respective markers in any of the three groups examined.

106

4.5 Discussion The use of a program to characterize iris surface features has a number of advantages over traditional qualitative assessments of iris photographs. For one, it allows for the storage and retrieval of a substantial amount of information that would not be easy to obtain using descriptive methods. This includes the size of the iris, the position in which each pigment spot and crypt is found, and the distribution of traits across the different quadrants of the iris. In addition, the program displays irises in a random order to limit bias, can handle large amounts of data and is relatively fast. It is also simple to set up replicate users for intra- and inter-reliability estimates. The inter- and intra-rater reliability estimates were both good for all five iris structures examined, which underscores the ability of the program to return consistent and repeatable results. The iris structure categories that we developed were designed to account for the extensive iris texture variation that is inherent in global populations. Previous attempts to study iris structure have primarily been targeted at single, more homogenous populations (Larsson and Pedersen, 2004; Larsson et al., 2011; Sidhartha et al., 2014a, 2014b; Sturm and Larsson, 2009). In this paper, we have attempted to expand on these methods and develop categories that can capture iris structure variation in global populations with different iris characteristics.

4.5.1 Fuchs’ Crypts Substantial amounts of hypoplasia can develop in the iris following the absorption of the pupillary membrane. The optic vessels first appear around the fourth week of fetal life, and the anterior layer begins to grow in front of the lens around the second embryonic month (Eagle, 1988; Eriksson et al., 1965; Oyster, 1999). This ultimately forms the iridopupillary membrane, which covers the entire iris. By the sixth month, this membrane begins to atrophy and is gradually reabsorbed until the pupil is completely free of mesodermal tissue. It has been suggested that the anterior layer that remains is a rudimentary tissue that has no major function in humans (Purtscher, 1965). As a result, there is a sparsity of tissue in this layer. Most irises have a moderate amount of hypoplasia around the collarette. However, in some eyes, this hypoplasia can extend widely into the ciliary zone. This leads to the formation of diamond-shaped lacunae known as Fuchs’ crypts. These lacunae are primarily believed to be a phylogenetic defect that

107 results from the decreasing importance of the anterior layer over evolutionary history (Purtscher, 1965). However, secondary lacunae can also gradually form due to the pushing and pulling effects of the pupil on the anterior layer when it dilates and contracts. Thus, it is not surprising that age has been associated with a higher number of crypts in prior studies (Larsson and Pedersen, 2004). For Fuchs’ crypts, we attempted to develop a categorical system that could capture the total amount of hypoplasia in the iris. Irises that contained very little hypoplasia, or hypoplasia centred solely along the collarette, were placed into the first two categories. Irises that had a greater amount of hypoplasia that extended into at least 50% of the ciliary zone were placed in the latter two categories. In all three sample sets, the largest number of participants fell into the second category (small crypts around the collarette in at least three quadrants of the iris). However, individuals of European descent had a greater probability of having a higher crypt grade than individuals of South Asian descent, and individuals of South Asian ancestry had a greater probability of having a higher crypt grade than individuals of East Asian descent (Table 4-3). Among East Asians, only 29.8% of participants had crypts that extended into the ciliary zone, compared with 52.9% of Europeans and 43.4% of South Asians. There may be several explanations for this finding. Individuals of East Asian ancestry have significantly darker irises than those of South Asian ancestry or European ancestry. It is possible that the presence of greater amounts of melanin in the anterior stromal layer may make the iris less prone to the development of both primary and secondary crypts. However, this is unlikely as the frequency of crypts appears to be independent of iris colour within populations (Larsson and Pedersen, 2004). Iris width may be another explanation, as we noted that individuals of East Asian ancestry had significantly smaller irises than those of South Asian and European ancestry. However, we were unable to find an association between iris width and crypt grade within populations. As the frequency of Fuchs’ crypts is believed to be associated with the overall stability of the anterior border layer, it is possible that there are other genetic and developmental population differences that affect the stability of this layer. In all three groups, sex was weakly associated with Fuchs’ crypts. Most notably, males consistently had greater odds of having a higher crypt grade than females. We were not able to replicate an association between Fuchs’ crypts and age (Larsson and Pedersen, 2004). However,

108 this was most likely because of the relatively narrow age range used in our study, compared with prior studies where participants ranged in age from children to seniors. SEMA3A rs10235789 is one marker that appears to be associated with variation in Fuchs’ crypts (Larsson et al., 2011). This marker, which is believed to play a role in the initial development of the pupillary membrane, has previously been associated with approximately 1.5% of the variation in Fuchs’ crypts in an Australian sample of European ancestry. When we looked at the association between SEMA3A rs10235789 and Fuchs’ crypt grade in our European, East Asian, and South Asian samples, we found significant associations in all three groups. The effects of SEMA3A rs10235789 appear to be additive, with two copies of the derived ‘C’ allele having a greater effect than one copy. The geographical distribution of the SEMA3A rs10235789 polymorphism may help to explain some of the differences observed in the frequency of crypts in the three groups. The derived ‘C’ allele is found at a very high frequency in our European sample (0.48), while it is found at much lower frequencies in the East Asian (0.08) and South Asian (0.28) groups. Therefore, we would expect crypts to be found at a lower frequency in the latter groups. Given the role of SEMA3A rs10235789 in the absorption of the pupillary membrane, it is possible that there are population differences in the stability of the anterior border layer after pupil reabsorption in the different population groups.

4.5.2 Pigment Spots Pigment spots are discrete areas of dark pigmentation that are found on the anterior border layer of the iris (Rennie, 2012). There are many different types of spots, the most common of which are iris freckles and nevi. Both variants of pigment spots are similar topographically, but are very different ultrastructurally (Eagle, 1988; Harbour et al., 2004). Freckles, which range from light to dark in colour, are found in 50-60% of healthy adults (Reese, 1944; Rennie, 2012). They lie flat on the surface of the iris, and do not distort the iris stroma. By contrast nevi are found only in 4-6% of adults, and look like well-demarcated nodular lesions (Rennie, 2012). They tend to be more common on the lower half of the iris and may increase in size over time (Horn et al., 1994). Unlike freckles, nevi do distort the underlying stromal layer. Although older individuals have a predisposition for both types of pigment spots, they may be found in all ages (Reese, 1944; Sturm and Larsson, 2009).

109

Pigment spotting was significantly more common in individuals of European ancestry than in individuals of East or South Asian ancestry (Table 4-3). Over half of our European sample (57.9%) showed some degree of iridial spotting, compared with only 22.1% of East Asians and 16.7% of South Asians. This finding is not surprising. As pigment spots are discrete regions of brown or black pigment, we would expect to see fewer spots in populations with a predisposition for darker iris colours given that regions of hyper-melanin can be difficult to distinguish in darker eyes. Interestingly, in all three groups, pigment spots were most likely to be found in the lower temporal quadrant of the iris (Table 4-4). Several prior studies have noted that pigment spots preferentially appear on the lower half of the iris (Schwab et al., 2014). This is because sun exposure appears to be a risk factor for the development of iridial spots, and the lower half of the iris is least protected by the eyelids. However, we only noted this preference on the lower temporal quadrant, not on the lower nasal quadrant. In fact, the second most common quadrant for pigment spotting was the upper temporal quadrant. It is possible that the upper temporal quadrant may be less protected by the eyelids than the lower nasal quadrant, leading to this discrepancy. We did not find a correlation between pigment spotting and sex or age in any of the three groups. However, we did note an association between iris width and pigment spot grade in the European and South Asian sample sets. Given the well-established link between sun exposure and iris pigment spots, it may simply be the case that a wider iris has more contact with the sun than a smaller iris. We were also not able to identify an association between HERC1 rs11630290 and pigment spots in any of the three samples. This is not surprising, as HERC1 rs11630290 was the weakest association reported in prior studies. It is possible that our sample size was too small to replicate this association.

4.5.3 Contraction Furrows Contraction furrows, which are produced by the contraction and dilation of the pupil, are deep depressions that lie around the outer periphery of the iris (Eagle, 1988; Reese, 1944). Although each contraction furrow rarely extends more than an iris quadrant in length, there are typically many furrows staggered around the eye. It has been suggested that the overall thickness and density of the iris play an important role in their formation and overall appearance (Larsson and Pedersen, 2004; Larsson et al., 2011). Although prior studies have used several different

110 systems for describing contraction furrows, we were primarily interested in looking at their overall extension around the iris. This is because furrow extension is easier to quantify than furrow depth or number of furrows when looking at irises of varying shades. In all three samples, there were very few individuals with the lowest grade of contraction furrows (15.2% of East Asians, 6.6% of Europeans, 3.8% of South Asians). Individuals of South Asian and European ancestry had similar distributions of contraction furrow grade (Table 4-3). However, participants of East Asian ancestry had a significantly lower furrow grade than either South Asians or Europeans. In several prior studies, contraction furrow grade has been associated with both darker irises and a thicker peripheral iris within populations (Larsson and Pedersen, 2004; Quillen et al., 2011; Sidhartha et al., 2014a). Thus, it is unexpected to see an overall lower grade of contraction furrows in the East Asian sample, given that this group has the darkest self-described eye colour and the fact that East Asian individuals have also been found to have a thicker peripheral iris than individuals of European ancestry (Lee et al., 2013). This suggests that some factor other than iris colour or iris depth is responsible for generating differences in contraction furrow grade between populations. One explanation may be iris width and iris area. There is some evidence that a higher overall iris area (irrespective of iris depth) may be associated with a higher grade of contraction furrows (Sidhartha et al., 2014a). In our sample, individuals of East Asian ancestry had a significantly smaller iris width than individuals of European or South Asian ancestry. Likewise, prior studies have found that individuals of East Asian ancestry have an overall smaller iris area than individuals of other population groups (Albert et al., 2003). Although, we were unable to find a significant association between contraction furrow grade and iris width in any of the population groups, this association was close to the established significance value of p<0.05 in the East Asian (p=0.096) and European (p=0.073) groups. TRAF3IP1 rs3739070 has been associated with approximately 1.7% of the variation in contraction furrows in Australian individuals of European ancestry (Larsson et al., 2011). It has been suggested that this marker may play a role in determining the overall thickness and density of the iris. We were able to detect a borderline significant association (p=0.051) between TRAF3IP1 rs3739070 and contraction furrows in our European sample. The derived allele appears to have a recessive effect, with one copy of the derived allele not increasing the odds of having more extended contraction furrows. We estimated that having two copies of the derived

111

‘A’ allele increases the odds by 6.385. However, it is important to note that we only had five individuals who were homozygous for the ancestral allele. In both East and South Asians, the ancestral TRAF3IP1 rs3739070 allele also has a very low frequency. No participants of South Asian descent, and only one individual of East Asian descent, were homozygous for the ancestral ‘A’ allele. In addition, the frequency of heterozygotes was also very low in both populations. Thus, it is not surprising that we were unable to find an association between TRAF3IP1 and furrow grade in either group. It is likely that markers other than TRAF3IP1 rs3739070 play a role in the development of contraction furrows in populations of non-European ancestry. East Asian populations may be an ideal group for studying the genetic basis of contraction furrows. Unlike the European and South Asian samples, where there is little trait variation and the vast majority of individuals have highly extended furrows, participants of East Asian ancestry were more likely to fall into the lower furrow grades.

4.5.4 Wolfflin Nodules Wolfflin nodules are small (0.1-0.2 mm) circular lesions that are distributed uniformly along the outer border of the ciliary zone (Sturm and Larsson, 2009; Williams, 1981). They primarily comprise of atrophied collagen from the anterior border and stromal layers. Wolfflin nodules are highly associated with iris colour, and are much more common in light-eyed individuals than in dark-eyed individuals. For example, in one study of 123 brown-eyed East Asian children, Wolfflin nodules were absent from all participants (Kim et al., 2002). Much attention has been devoted to Wolfflin nodules during the past 50 years due to their structural similarity to Brushfield spots, an iris feature found in 85-90% of Down’s syndrome patients (Skeller and Øster, 1951; Williams, 1981). Although the genetic basis of Brushfield spots remains largely unknown, they closely resemble Wolfflin nodules and also appear to be restricted to individuals with light-coloured eyes. However, they tend to be larger, less uniform, and located closer to the mid-zone of the iris. We decided to characterize Wolfflin nodules based on their overall extension around the iris. Individuals who had Wolfflin nodules that extended more than 180° around the iris were put into the highest grade, and individuals who had Wolfflin nodules that extended less than 180° were placed in the second grade. Individuals with no discernable Wolfflin nodules were placed in the lowest grade. Interestingly, very few individuals fell into the second category and participants

112 were more likely to have no Wolfflin nodules, or Wolfflin nodules that extended around the entire iris (Table 4-3). Unsurprisingly, Wolfflin nodules were significantly more common in the European sample, with 25.8% of this group having Wolfflin nodules that extended more than 180˚ around the iris, compared to 0% of East Asians and 0.8% of South Asians. Given the strong association between Wolfflin nodules and iris colour, we would expect to see fewer Wolfflin nodules in the primarily brown-eyed East and South Asian populations. It has been suggested that both Wolfflin nodules and Brushfield spots are not actually absent in brown irises, but merely obscured by the abundance of melanin particles in the anterior border layer (Falls, 1970). Therefore, it is possible that methods other than colour photography may be necessary to accurately characterize Wolfflin nodules in populations of diverse ancestry. Interestingly, we did find an association between larger iris width and more extended Wolfflin nodules in the European population. As Wolfflin nodules are composed of atrophied collagen from the stromal layer, it is possible that larger irises may be more prone to the accumulation of nodules in the anterior stromal layer. In a recent genome-wide association study on iris colour in a population of Danish ancestry, DSCR9 rs7277820 was found to be associated with iris colour variation (Liu et al., 2010). It was suggested that this marker may actually be correlated with Wolfflin nodules, due to its presence in the Downs’ Syndrome Critical Region. When we looked at the association between DSCR9 rs7277820 and Wolfflin nodules in the European sample, we were not able to find a significant relationship.

4.5.5 Conjunctival Melanosis Finally, the last trait that we looked at was conjunctival melanosis. In some irises, there are areas of pigment spotting on the scleral region surrounding the iris (Damato and Coupland, 2008). These spots can comprise either of discrete regions, or rings surrounding the iris. Conjunctival melanosis has not yet been widely explored. However, it does appear to be found more commonly in populations with darker irises. This was largely reflected in our sample. Here, almost half of our South Asian participants showed some form of conjunctival melanosis, compared to only 23.3% of East Asians and 2.1% of Europeans (Table 4-3). It is important to note, however, that these results may be an underestimate given that the upper and lower portions of the sclera were not fully visible in our sample. Interestingly, the presence of

113 conjunctival melanosis does not appear to be entirely linked to iris colour, as the distribution of this trait is very different among the East and South Asian participants.

4.5.6 Correlations of Iris Features Correlations between iris features and sex, age and iris width were examined in all three population groups. Within the European population we were able to replicate a number of the correlations that have been identified in prior studies. These include: 1/ A higher grade of crypts is correlated with less extended contraction furrows (Larsson and Pedersen, 2004); 2/ A higher grade of crypts is correlated with fewer pigment spots (Quillen et al., 2011); 3/ More extended contraction furrows are correlated with a greater amount of pigment spotting (Larsson and Pedersen, 2004). We were not initially able to replicate the association between crypt grade and Wolfflin nodule grade (Larsson and Pedersen, 2004). However, we were interested in determining if this was just because of the higher frequency of brown-eyed Europeans in our sample, in comparison with the previous studies reporting the correlation. When we restricted our sample to individuals who self-described their iris colour as either ‘light blue, green or grey’ or ‘blue, green or grey’, a higher grade of crypts (p<0.044, G=0.142) and more extended contraction furrows (p=0.001, G=0.343) were both correlated with more extended Wolfflin nodules. This emphasizes the effects that iris colour can have on the study of global Wolfflin nodule variation. In contrast with the significant correlations observed in the European sample, none of the five traits were significantly correlated in either the East or South Asian populations. There are several potential explanations for this: 1/ Iris colour may be obscuring some of the associations; 2/ There may be population differences responsible for the lack of correlations; or 3/ Our ability to identify significant correlations in the East Asian and South Asian samples was hampered due to smaller sample sizes.

4.5.7 Future Research Several prior studies have suggested that the distribution of iris features may be population- dependent (Qiu et al., 2005; Quillen et al., 2011). In this paper, we looked at the global distribution of Fuchs’ crypts, Wolfflin nodules, contraction furrows, pigment spots and conjunctival melanosis in participants of East Asian, European and South Asian ancestry. We found that all five traits showed significant differences in frequency across the three groups. We

114 also showed that SEMA3A rs10235789 is significantly associated with crypts not only in the European sample, but also in the East Asian and South Asian samples. By contrast, TRAF3IP1 rs3739070 is only correlated with contraction furrows in Europeans. Lastly, we were able to replicate all of the iris feature correlations that had been identified in prior studies in our European sample. However, we were not able to identify any correlations in the East and South Asian groups. Future research will be necessary to explore the frequency of these features in other populations, such as African and Hispanic groups. There are still many gaps in our understanding of iris surface features. At present, very little is known about the genetic basis and global distribution of Wolfflin nodules. This is primarily because it is difficult to study this trait in populations that are primarily composed of individuals with brown eyes. While it was initially suggested that these collagen bundles were absent in darker irises, it has since been established that they are most likely masked when there are large amounts of melanin particles in the anterior stromal layer. At present, the only way to accurately characterize Wolfflin nodules is through magnification (Falls, 1970). Pigment spots are similarly very difficult to study. Most notably, we do not have a way of confidently differentiating between pigment freckles and nevi from photographs of the iris. Nevi and freckles both appear to have a different relationship with the development of ocular disorders such as uveal melanoma (Holly et al., 1990; Horn et al., 1994). Therefore, it would be extremely useful to develop ways of distinguishing these features in photographs. We suggest that infrared photography may provide a potential solution to these issues. As infrared light is neither absorbed nor reflected by melanin, it may provide a way of visualizing traits, such as Wolfflin nodules, that lie below the melanin granules in the anterior border layer. In addition, as the primary difference between nevi and freckles is whether or not the underlying stromal layer is distorted, infrared photography may provide a means of identifying the spots that have an effect on this layer. The field of iris recognition already relies heavily on the use of infrared photography. However, it also amalgamates all of the structures in the iris into a single code (Acharya and Kasprzycki, 2010). Looking at infrared photographs from a qualitative perspective may provide valuable information about the distribution of both pigment spots and Wolfflin nodules in global populations. Iris features are beginning to play an increasingly important role in many different fields. However, there are still many gaps in our knowledge of these traits. Future research will be

115 necessary in order to determine the functional differences in these traits in global populations, as well as the effects that these traits may have on population-specific ocular diseases and disorders.

116

4.6 Tables

Table 4-1 General Descriptive Statistics for the East Asian, European and South Asian Irises

Population Irises Included Females Males Average Age Average Iris Width East Asian 467 320 147 21.61 376.72 European 619 376 243 22.65 394.45 South Asian 364 246 118 20.67 384.92

117

Table 4-2 Results of the Genetic Ordinal Regression. Significance values and odds ratios for each genotype are presented for each population. Genotypes labelled with an ‘a’ refer to the ancestral genotype.

Polymorphism Population Genotypes Significance Odd's Ratios CC - - East Asian CT 0.018 1.679 TT a a CC < 0.001 2.203 rs10235789 European CT 0.023 1.507 (Fuchs' Crypts) TT a a CC 0.011 2.721 South Asian CT < 0.001 2.206 TT a a AA - - East Asian CA 0.956 0.980 CC a a AA 0.051 6.385 rs3739070 European CA 0.333 2.601 (Contraction Furrows) CC a a AA - - South Asian CA 0.168 0.360 CC a a TT - - East Asian CT 0.173 0.241 CC a a TT 0.187 1.685 rs11630290 European CT 0.998 1.000 (Pigment Spots) CC a a TT - - South Asian CT 0.136 0.536 CC a a AA 0.375 0.758 rs7277820 European GA 0.345 0.794 (Wolfflin Nodules) GG a a

118

Table 4-3 Iris Surface Feature Frequencies The percentage (and number) of individuals with each category of Fuchs’ crypts, contraction furrows pigment spots, Wolfflin nodules, melanosis and iris colour in all three populations.

Trait Category East Asian European South Asian 0 20.3% (95) 9.2% (57) 15.1% (55)

Fuchs’ 1 49.9% (233) 38.0% (235) 41.5% (151) Crypts 2 19.7% (92) 30.9% (191) 25.0% (91) 3 10.1% (47) 22.0% (136) 18.4% (67) 0 15.2% (71) 6.6% (41) 3.8% (14) Contraction 1 37.7% (176) 11.1% (69) 15.4% (56) Furrows 2 47.1% (220) 82.2% (509) 80.8% (294) 0 77.9% (364) 42.1% (260) 83.2% (303) Pigment 1 20.6% (96) 38.4% (238) 14.8% (54) Spots 2 1.5% (7) 19.5% (121) 1.9% (7) 0 100% (467) 62.7% (388) 98.6% (359) Wolfflin 1 0% (0) 12.0% (74) 0.5% (2) Nodules 2 0% (0) 25.4% (157) 0.8% (3) 0 76.7% (358) 97.9% (606) 54.9% (200) Melanosis 1 23.3% (109) 2.1% (13) 45.1% (164) 0 0% (0) 11.8% (73) 0% (0) 1 0% (0) 45.1% (279) 1.6% (6) Eye 2 4.5% (21) 23.1% (143) 8.2% (30) Colour 3 45.4% (212) 19.1% (118) 51.9% (189) 4 50.1% (234) 1.0% (6) 38.2% (139)

119

Table 4-4 Distribution of Iris Surface Features The percentage (and number) of irises with large crypts, contraction furrows, pigment spots, and Wolfflin nodules in each of the four quadrants of the iris

Trait Population Quadrant 1 Quadrant 2 Quadrant 3 Quadrant 4 East Asian 10.9% (51) 15.8% (74) 14.3% (67) 18.4% (86) Large Crypts European 31.5% (195) 35.5% (220) 25.8% (160) 29.2% (181) South Asian 23.6% (86) 29.9% (109) 22.5% (82) 22.3% (81) East Asian 26.8% (125) 70.4% (329) 82.4% (385) 61.0% (285) Furrows European 82.1% (508) 84.8% (525) 89.2% (552) 84.3% (522) South Asian 70.6% (257) 91.2% (332) 95.1% (346) 84.1% (306) East Asian 2.8% (13) 5.1% (24) 10.3% (48) 8.1% (38) Pigment Spots European 17.3% (107) 18.6% (115) 31.7% (196) 31.2% (193) South Asian 2.7% (10) 4.9% (18) 6.3% (23) 6.6% (24) East Asian 0% (0) 0%(0) 0% (0) 0% (0) Nodules European 23.7% (147) 28.6% (177) 34.9% (216) 31.2% (193) South Asian 0.8% (3) 1.1% (4) 1.4% (5) 1.4% (5)

120

4.7 Figures

Figure 4-1 Five Surface Features Commonly Found in the Human Iris The arrows are pointing to the features in images b, c and e. Fuchs’ crypts (a) are lacunae in the anterior border of the iris which arise during resorption of the pupillary membrane. They may be either large or small and closely resemble windows. Four sample crypts are outlined in the image below. Wolfflin nodules (b) are small bundles of collagen that are the consequence of atrophy in the stromal layer of the iris. Pigment spots (c) are discrete areas of pigmentation that can be observed on the surface of the iris. Spots that distort the stromal layer are referred to as nevi and spots that do not distort the stromal layer are referred to as freckles. Contraction furrows (d) are rings that extend around the outer border of the iris. They closely resemble wrinkles and are the product of the contraction and dilation of the pupil. Furrows are typically discontinuous and staggered across the iris. In the image below, the black line follows the path of the furrows around the eye. Conjunctival melanosis (e) is spotting that can be observed on the scleral region surrounding the iris. It is usually benign, and is found more commonly in some ancestries than in others.

121

Figure 4-2 Boundaries of the Iris In steps 2-6 of the web-based application, the user draws a best fit circle around the pupillary ruff (purple circle), collarette (blue circle) and iris (red circle). This allows the program to identify the pupillary zone (bounded by the blue and purple circles) and the ciliary zone (bounded by the blue and red circles)

122

Figure 4-3 Iris Quadrants The program divides the iris into four different quadrants based on the user-defined centre of the iris. Quadrant 1 is the upper nasal quadrant, quadrant 2 is the lower nasal quadrant, quadrant 3 is the lower temporal quadrant, and quadrant 4 is the upper temporal quadrant.

123

Figure 4-4 Pigment Spot Categories Category 1: no pigment spots. Category 2: between one and two pigment spots. Category 3: more than two pigment spots. Any well-demarcated lesion that ranged in colour from tan to dark brown was considered a pigment spot. The black arrows illustrate example pigment spots.

124

Figure 4-5 Contraction Furrow Categories Category 1: no contraction furrows. Category 2: contraction furrows that extend less than 180° around the iris. Category 3: contraction furrows that extend more than 180° around the iris. The black lines follow the extension of contraction furrows around the iris.

125

Figure 4-6 Wolfflin Nodule Categories Category 1: no Wolfflin nodules. Category 2: Wolfflin nodules that extend less than 180° around the iris. Category 3: Wolfflin nodules that extend more than 180° around the iris. Wolfflin nodules were defined as well-demarcated lesions that were white to orange in colour. The black arrows are pointing at example Wolfflin nodules.

126

Figure 4-7 Crypt Categories Category 1: no crypts. Category 2: only small crypts centred around the collarette. Category 3: at least one large crypt located in fewer than three quadrants of the iris. Category 4: at least three large crypts located in three or more quadrants of the iris. Large crypts are defined as those that extend from the collarette into more than 50% of the ciliary zone. The web application used the scleral and iris boundaries to automatically distinguish between large crypts and small crypts. Examples of crypts are illustrated in the images below. Not all crypts are labelled in each image. Crypts that would be defined as large are bounded in green and crypts that would be defined as small are bounded in yellow.

127

Figure 4-8 Self-Described Iris Colour, as Defined by the Fitzpatrick Phototype Scale The Fitzpatrick scale characterizes iris colour across five categories: 1/ light blue, green or grey, 2/ blue, green or grey, 3/ hazel or light brown, 4/ dark brown, and 5/ brownish black. The photographs below represent a self-described example from each of the five categories.

128

Figure 4-9 Conjunctival Melanosis Categories Conjunctival melanosis (black arrows) was measured using a presence or absence schema. Any iris that showed spotting on the sclera was characterized as having conjunctival melanosis. This spotting typically took the form of a ring around the iris, or as isolated spots on the sclera.

129

4.8 References Acharya, L., and Kasprzycki, T. (2010). Biometrics and government (Library of Parliament). Albert, D.M., Green, R.W., Zimbric, M.L., Lo, C., Gangnon, R.E., Hope, K.L., and Gleiser, J. (2003). Iris melanocyte numbers in Asian, African American, and Caucasian irides. Trans. Am. Ophthalmol. Soc. 101, 217–221. Court M (2008). Court lab calculator. Damato, B., and Coupland, S.E. (2008). Conjunctival melanoma and melanosis: a reappraisal of terminology, classification and staging. Clin. Experiment. Ophthalmol. 36, 786–795. Donaldson DD (1961). The significance of spotting of the iris in mongoloids: Brushfield’s spots. Arch. Ophthalmol. 65, 26–31. Eagle, R.C. (1988). Iris pigmentation and pigmented lesions: an ultrastructural study. Trans. Am. Ophthalmol. Soc. 86, 581–687. Eiberg, H., Troelsen, J., Nielsen, M., Mikkelsen, A., Mengel-From, J., Kjaer, K.W., and Hansen, L. (2008). Blue eye color in humans may be caused by a perfectly associated founder mutation in a regulatory element located within the HERC2 gene inhibiting OCA2 expression. Hum. Genet. 123, 177–187. Eriksson, A.W., Fellman, J., Nieminen, H., and Forsius, H. (1965). Influence of age on the position and size of the iris frill and the pupil. Acta Ophthalmol. (Copenh.) 43, 629–641. Falls, H.F. (1970). Ocular changes in mongolism. Ann. N. Y. Acad. Sci. 171, 627–636. Fitzpatrick, T.B. (1975). Soleil et peau. J Med Esthet 2, 33–34. Harbour, J.W., Brantley, M.A., Hollingsworth, H., and Gordon, M. (2004). Association between posterior uveal melanoma and iris freckles, iris naevi, and choroidal naevi. Br. J. Ophthalmol. 88, 36–38. Holly, E.A., Aston, D.A., Char, D.H., Kristiansen, J.J., and Ahn, D.K. (1990). Uveal melanoma in relation to ultraviolet light exposure and host factors. Cancer Res. 50, 5773–5777. Horn, E.P., Hartge, P., Shields, J.A., and Tucker, M.A. (1994). Sunlight and risk of uveal melanoma. J. Natl. Cancer Inst. 86, 1476–1478. Imesch, P.D., Bindley, C.D., Khademian, Z., Ladd, B., Gangnon, R., Albert, D.M., and Wallow, I.H. (1996). Melanocytes and iris color. Electron microscopic findings. Arch. Ophthalmol. 114, 443–447. Kim, J.H., Hwang, J.-M., Kim, H.J., and Yu, Y.S. (2002). Characteristic ocular findings in Asian children with Down syndrome. Eye Lond. Engl. 16, 710–714. Larsson, M., and Pedersen, N.L. (2004). Genetic correlations among texture characteristics in the human iris. Mol. Vis. 10, 821–831.

130

Larsson, M., Duffy, D.L., Zhu, G., Liu, J.Z., Macgregor, S., McRae, A.F., Wright, M.J., Sturm, R.A., Mackey, D.A., Montgomery, G.W., et al. (2011). GWAS findings for human iris patterns: associations with variants in genes that influence normal neuronal pattern development. Am. J. Hum. Genet. 89, 334–343. Lee, R.Y., Huang, G., Porco, T.C., Chen, Y.-C., He, M., and Lin, S.C. (2013). Differences in iris thickness among African Americans, Caucasian Americans, Hispanic Americans, Chinese Americans, and Filipino-Americans. J. Glaucoma 22. Liu, F., Wollstein, A., Hysi, P.G., Ankra-Badu, G.A., Spector, T.D., Park, D., Zhu, G., Larsson, M., Duffy, D.L., Montgomery, G.W., et al. (2010). Digital quantification of human eye color highlights genetic association of three new loci. PLoS Genet 6, e1000934. Mackey, D.A., Wilkinson, C.H., Kearns, L.S., and Hewitt, A.W. (2011). Classification of iris colour: review and refinement of a classification schema. Clin. Experiment. Ophthalmol. 39, 462–471. Oyster, C.W. (1999). The human eye: structure and function (Sunderland, Massachusetts: Sinauer Associates, Inc.). Peles, D.N., Hong, L., Hu, D.-N., Ito, S., Nemanich, R.J., and Simon, J.D. (2009). Human iridal stroma melanosomes of varying pheomelanin contents possess a common eumelanic outer surface. J. Phys. Chem. B 113, 11346–11351. Purtscher, E. (1965). On the development and morphology of iris crypts*. Acta Ophthalmol. (Copenh.) 43, 109–119. Qiu, X., Sun, Z., and Tan, T. (2005). Global texture analysis of iris images for ethnic classification. Adv. Biom. 3832, 411–418. Quillen, E.E., Guiltinan, J.S., Beleza, S., Rocha, J., Pereira, R.W., and Shriver, M.D. (2011). Iris texture traits show associations with iris color and genomic ancestry. Am. J. Hum. Biol. 23, 567– 569. Reese, A.B. (1944). Pigment freckles of the iris (benign melanomas): Their significance in relation to malignant melanoma of the uvea ⋆. Am. J. Ophthalmol. 27, 217–226. Rennie, I.G. (2012). Don’t it make my blue eyes brown: heterochromia and other abnormalities of the iris. Eye 26, 29–50. Schwab, C., Zalaudek, I., Mayer, C., Riedl, R., Wackernagel, W., Juch, H., Aigner, B., Brunasso, A.M., Langmann, G., and Richtig, E. (2014). New insights into oculodermal nevogenesis and proposal for a new iris nevus classification. Br. J. Ophthalmol. 99, 644-649. Sidhartha, E., Gupta, P., Liao, J., Tham, Y.-C., Cheung, C.Y., He, M., Wong, T.Y., Aung, T., and Cheng, C.-Y. (2014a). Assessment of iris surface features and their relationship with iris thickness in Asian eyes. Ophthalmology 121, 1007–1012.

131

Sidhartha, E., Nongpiur, M.E., Cheung, C.Y., He, M., Wong, T.Y., Aung, T., and Cheng, C.-Y. (2014b). Relationship between iris surface features and angle width in Asian eyes. Invest. Ophthalmol. Vis. Sci. 55, 8144–8148. Skeller, E., and Øster, J. (1951). Eye symptoms in mongolism. Acta Ophthalmol. (Copenh.) 29, 149–161. Sturm, R.A., and Larsson, M. (2009). Genetics of human iris colour and patterns. Pigment Cell Melanoma Res. 22, 544–562. Sturm, R.A., Duffy, D.L., Zhao, Z.Z., Leite, F.P.N., Stark, M.S., Hayward, N.K., Martin, N.G., and Montgomery, G.W. (2008). A single SNP in an evolutionary conserved region within intron 86 of the HERC2 gene determines human blue-brown eye color. Am. J. Hum. Genet. 82, 424– 431. Viera, A.J., and Garrett, J.M. (2005). Understanding interobserver agreement: the kappa statistic. Fam Med 37, 360–363. Walsh, S., Liu, F., Ballantyne, K.N., von Oven, M., Lao, O., and Kayser, M. (2011). IrisPlex: a sensitive DNA tool for accurate prediction of blue and brown eye colour in the absence of ancestry information. Forensic Sci. Int. Genet. 5, 170–180. Wilkerson, C.L., Syed, N.A., Fisher, M.R., Robinson, N.L., Wallow, I.H., and Albert, D.M. (1996). Melanocytes and iris color. Light microscopic findings. Arch. Ophthalmol. 114, 437– 442. Williams, R.D. (1981). Brushfield spots and Wolfflin nodules in the iris: an appraisal in handicapped children. Dev. Med. Child Neurol. 23, 646–650.

Chapter 5: Concluding Remarks

132 133

5.1 Introduction The overall goal of my research was to investigate the global distribution and genetic basis of iris colour and structure in populations of diverse ancestry. In order to accomplish this, I acquired high-resolution iris photographs from participants of East Asian, European and South Asian ancestry who lived in the Greater Toronto Area. For my first paper (Chapter 2), I used a pilot sample of 205 participants in order to develop my methodology. For my second and third papers (Chapters 3-4), I carried out a full research study on 1,465 participants. I was specifically interested in answering the following research questions: 1/ How can I quantify iris colour and central heterochromia in populations of diverse ancestry? 2/ How can I characterize iris surface features in populations of diverse ancestry? 3/ How does the phenotypic distribution of iris colour and surface features differ in populations of East Asian, European and South Asian ancestry? 4/ What genetic variants are responsible for iris colour and structure variation in European, East Asian and South Asian populations?

5.2 Summary of Findings 5.2.1 Quantifying Iris Pigmentation Obtaining a measurement of iris colour has traditionally been a very difficult endeavor. Categorical methods of characterizing iris colour are unable to account for the extensive continuous variation observed in this trait. In contrast, quantitative methods ignore regions of heterochromia in the eye. A major aim of my research was to develop a quantitative method of measuring eye colour that could also account for central heterochromia. In the first study of my thesis (Chapter 2), I attempted to extract iris colour from a 256 by 256 pixel square in the left quadrant of the eye. A collaborator created a macro in Adobe Photoshop CS5 that carried out this process automatically. The average colour measurement of this square was then recorded in CIE 1976 L*a*b* (CIELAB) colour space. CIELAB is a universal colour system that closely approximates human vision (McLaren, 1976). This means that most people should be able to detect a one unit change in colour space (Kuehni and Marcus, 1979). This is very advantageous for a trait like eye colour, which has traditionally been graded visually. Although this initial quantitative method was quite successful in capturing phenotypic variation, it had considerable limitations. Other research groups correctly noted that this method

134 isolated only a very small portion of the iris (Andersen et al., 2013). This means that if a pigmentation abnormality, such as a large pigment spot, is found in that 256 by 256 pixel square, the overall colour of that iris might be grossly misrepresented. In addition, this method provided no easy way to isolate or study central heterochromia. For my second study (Chapter 3), I refined my iris colour quantification methodology to account for these limitations. A web application was designed by a second collaborator that was able to extract a 60˚ wedge from the left quadrant of the iris. Although this new method is not automatic, it delivers a much deeper and comprehensive estimation of iris colour. As the user is required to define the scleral, pupillary and collarette boundaries, the program is able to obtain individual colour measurements from both the pupillary and ciliary zones. This means that the amount of central heterochromia in the iris, as well as the overall colour of the iris, can be quantified. I continued to use CIELAB colour space to quantify iris colour, as this system proved to be very effective in the early stages of my research. This new methodology is powerful, accurate and allows researchers to approach iris pigmentation variation from an entirely new direction.

5.2.2 Characterizing Iris Surface Features Although diversity in iris surface features has been studied within several continental population groups, there has not yet been an attempt to look at variation in these traits across different continental groups. As a result, most of the methods that have been used to characterize iris structure were designed with a single sample set in mind. In the third research study of my thesis (Chapter 4), I attempted to develop a method of grading iris surface features that could be applied to diverse populations. At present, the most effective way of characterizing iris structure is categorically. Both Wolfflin nodules and contraction furrows were graded based on their presence and extension around the iris. Fuchs’ crypts were graded based on the amount and extent of hypoplasia in the iris. Finally, pigment spots were graded based on their overall count in the eye. The same web application that was used to characterize iris colour was modified to also describe iris texture. This web application consists of eleven steps that have the user manually note the presence and extent of each of the surface features in the iris. The program then automatically assigns a grade for each of the features. Interestingly, while investigating iris texture variation I noticed that many eyes showed spots or rings of

135

pigment on the sclera. This condition, which is known as conjunctival melanosis, has not yet been widely characterized. Therefore, I chose to also incorporate the ‘presence or absence’ of this trait into the web application. This categorical method of measuring iris surface features shows consistent and reliable results. In addition, the frequency of each of the traits described in my study mirrors previously reported frequencies (Donaldson, 1961; Harbour et al., 2004; Sidhartha et al., 2014a, 2014b). A standardized method of measuring iris surface features in diverse populations should allow researchers to better compare the distribution of these traits across different ancestral groups.

5.2.3 Phenotypic Distribution of Iris Colour and Central Heterochromia It is well-established that the greatest range of iris pigmentation diversity is found in populations of European ancestry. However, very little is known about the distribution of iris colour variation in populations outside of Europe. In the first and second papers of my thesis (Chapters 2 and 3), I investigated the distribution of this trait in East Asian, European and South Asian populations. Unsurprisingly, the European sample showed the greatest range of colour variation across CIELAB colour space. However, extensive diversity could also be observed in the South and East Asian samples, with the South Asian sample showing somewhat more dispersion than the East Asian sample. The distribution of all three populations across CIELAB colour space confirms that eye colour is a continuous trait that needs to be approached quantitatively. Interestingly, brown eyes in the European sample appear to be much lighter than brown eyes in the East and South Asian samples. This suggests that there may be fixed population markers that control the intensity of brown eye colour in these different groups. The distribution of central heterochromia also showed clear population differences. The European sample displayed the greatest amounts of central heterochromia, followed by the South Asian and East Asian samples. Given the low colour difference between the pupillary and ciliary zones in populations outside of Europe, central heterochromia appears to be a trait found almost exclusively in lighter coloured irises. Populations with darker irises do not show very much variation between the different regions of the eye.

136

5.2.4 Phenotypic Distribution of Iris Surface Features In the third paper of my thesis (Chapter 4), I looked at the phenotypic distribution of Fuchs’ crypts, contraction furrows, Wolfflin nodules and pigment spots in populations of East Asian, European and South Asian ancestry. Interestingly, the global distribution of all four iris features appears to be largely dependent on ancestry. Fuchs’ crypts, pigment spots and Wolfflin nodules had a significantly higher grade in the European sample compared to the other two groups. Only contraction furrows showed an equally high grade in both the European and South Asian samples. In turn, the South Asian sample had a higher grade of Fuchs’ crypts and contraction furrows than the East Asian sample. However, both of these groups showed a similar degree of pigment spotting and Wolfflin nodules. The population-specific differences in pigment spots and Wolfflin nodules are most likely explained by differing eye colour frequencies. As both of these traits are more prominent in lighter irises, it makes sense that they would be most common in individuals of European ancestry. However, it is unlikely that the global differences in contraction furrow and crypt grade can be attributed to iris colour alone. Fuchs’ crypts were not associated with iris colour in any of the three groups. Similarly, the darkly pigmented East Asian population showed the lowest grade of contraction furrows. However, within the European population, contraction furrow extension was positively associated with darker irises. These results suggest that the distribution of iris surface features is population dependent. This may help explain why so many ocular diseases and disorders show ancestry-dependent distributions.

5.2.5 Genetic Basis of Iris Colour At present, the genetic basis of iris pigmentation is fairly well understood in populations of European ancestry. The majority of variation between blue and brown eye colour can be attributed a to a single marker, rs12913832, located in intron 86 of the gene HERC2 (Eiberg et al., 2008; Kayser et al., 2008; Sturm et al., 2008). However, other polymorphisms in OCA2, SLC24A4, SLC45A2, TYR, IRF4, TYRP1, ASIP, DSCR9, NPLOC4 and LYST may play a more subtle role as well (Beleza et al., 2013; Frudakis et al., 2003; Kanetsky et al., 2002; Liu et al., 2010; Sulem et al., 2007). At present, very few studies have attempted to examine iris colour diversity in populations of non-European ancestry.

137

In my first and second publications (Chapters 2 & 3), I looked at the association between putative iris pigmentation markers and eye colour in participants of East Asian, European and South Asian ancestry. My first publication consisted of a small-scale pilot study that looked specifically at HERC rs12913832. My second publication included HERC rs12913832 and 13 additional markers. In both studies, HERC rs12913832 was significantly associated with iris pigmentation variation in the European and South Asian sample sets. However, this marker had a very different effect in each population. This is exemplified particularly well in the second publication by individuals who were homozygous for the derived allele. European homozygotes had b* values that fell into the ‘blue’ region of colour space. In contrast, South Asian homozygotes had b* values that fell into the ‘green/intermediate’ region of colour space. This suggests that HERC2 rs12913832 does not just code directly for the difference between blue and brown eye colour. Rather, there must be other variants modifying its effect that are different in both populations. In the European sample, five other markers were also associated with iris pigmentation diversity: OCA2 rs1800407, SLC45A2 rs16891982, SLC24A4 rs12896399, IRF4 rs12203592 and TYR rs1393350. Apart from HERC2 rs12913832, eye colour in the South Asian sample set was largely determined by SLC24A5 rs1426654. This marker had an additive effect on iris colour, with each copy of the derived allele producing lighter brown eyes. SLC24A5 rs1426654 is an interesting polymorphism because it is one of the principal markers associated with the lightening of skin pigmentation in European populations (Lamason et al., 2005). It has also been found to contribute substantially to skin pigmentation diversity in South Asia (Stokowski et al., 2007). Thus, it is unlikely that it was selected for due to its effect on iris pigmentation. A second marker, LYST rs3768056, was also associated with eye colour in the South Asian sample. However, it had a much smaller effect. The lone determinants of eye colour in the East Asian sample were OCA2 rs1800414 and OCA2 rs74653330. The derived allele at both markers was associated with lighter iris pigmentation. Although not much is known about pigmentation variation in East Asia, both markers show a strong association with skin pigmentation in this region, and it is believed that both of these markers may have played a role in the evolution of light skin pigmentation in East Asia (Eaton et al., 2015; Edwards et al., 2010). Thus, their association with iris colour is likely secondary to their effect on skin colour.

138

My first and second research papers ultimately showed that the genetic basis of iris pigmentation is, at least to some extent, population-specific. With the exception of HERC2 rs12913832, the markers associated with iris pigmentation variation were different in all three groups. In the South Asian sample, this may be largely related to allele frequency differences. The derived allele frequency at SLC24A4 rs12896399 (European: 0.402, South Asian: 0.278), SLC45A2 rs16891982 (European: 0.931, South Asian: 0.1134) and TYR rs1393350 (European: 0.259, South Asian: 0.083) is much lower in the South Asian group compared to the European group. Thus, it is possible that the sample size in the South Asian group was simply too small to pick up these associations. In contrast, the markers controlling iris colour in the East Asian sample appear to be entirely independent from those observed in the European and South Asian samples. HERC2 rs12913832, OCA2 rs1800407, SLC45A2 rs16891982, IRF4 rs12203592 and TYR rs1393350 are all either entirely absent or found at negligible (< 0.010) frequencies in the East Asian sample. It is interesting to note that the genes that determine iris pigmentation diversity in all three regions are largely those that have been associated with other pigmentary characteristics, such as skin colour. This may provide some support to the hypothesis arguing that these markers were initially selected for due to the influence that they had on other pigmentary phenotypes.

5.2.6 Genetic Basis of Iris Surface Features In the third paper of my thesis (Chapter 4), I looked at the association between Fuchs’ crypts, contraction furrows, Wolfflin nodules and pigment spots and a set of genetic variants that had previously been identified in an Australian population of European ancestry (Larsson et al., 2011). Fuchs’ crypts were significantly associated with SEMA3A rs10235789 in the East Asian, European and South Asian samples. In all three groups, each additional copy of the derived allele was associated with a higher grade of crypts. I was also able to detect a borderline association between TRAF3IP1 rs3739070 and contraction furrow grade in the European sample. However, these results were not replicated in the East or South Asian groups. The ancestral TRAF3IP1 allele has a very low frequency in all three groups. However, it was found at a higher frequency in the European group (0.08) compared to the East (0.04) and South Asian (0.03) groups. It is possible that a larger sample size would be necessary to pick up this association.

139

Although HERC1 rs10235789 has been tentatively associated with the number of nevi in the eye, I was unable to detect a relationship between this marker and pigment spots in my sample set. Likewise, DSCR9 rs7277820 was not associated with Wolfflin nodules in any of the three ancestral groups. Even though iris surface features show a moderately high heritability, SEMA3A rs10235789 and TRAF3IP1 rs3739070 explain very little of the total variation in Fuchs’ crypts and contraction furrows. Given that all four iris features showed very different distributions in my three study populations, it will be necessary to investigate the genetic basis of these traits using denser panels of genetic markers, and a diverse representation of populations in order to identify other contributing markers.

5.3 Research Limitations The findings published in my three research papers represent an important contribution to the fields of iris structure and colour research. However, there are several limitations which must be addressed. These limitations include standardization across different sample sets, the number of populations studied, and difficulties associated with characterizing colour-dependent iris features The web application that was developed for this thesis represents a novel approach to the study of iris colour and structure variation. Unlike other methods, which generate an average measurement of eye colour and ignore other features like heterochromia, this application allows users to obtain a separate colour measurement from the ciliary and pupillary zones. In addition, because the web application does not require the eye to be a certain size or located in a particular place in the photograph, it can be applied to a broad set of iris photographs. Despite these advantages, this method still does not allow for the comparison of iris photographs taken from different instruments and cameras. This is largely because different instruments do not have the same exposure and colour settings. The next step in iris pigmentation research will be to find a way to standardize the ways in which photographs of the eye are taken across different research groups. Perhaps one solution would be to place a colour card in each photograph. This colour card could then be used to standardize colour and brightness variation in different sample sets. However, implementing such a method would be challenging. An alternative tactic to combine

140 the results of different studies would be to use meta-analytic approaches in which a global effect can be calculated based on the effect and variance of the individual studies. Another limitation of my research was the number of populations that I chose to study. The papers presented in this thesis are the first to look at iris colour in East and South Asian populations. They are also the first to compare iris features across different continental groups. However, there are many populations that still need to be studied in order to fully understand the global diversity inherent in these traits. The Greater Toronto Area was an ideal setting for the recruitment of East Asian, European and South Asian volunteers. However, the pilot study indicated that I would not be able to recruit a sufficient number of individuals of African, Hispanic and Middle Eastern ancestry. Future research will be needed in order to obtain a larger sample size with a better ancestry representation. Lastly, we need to develop better ways of measuring colour-dependent iris features. Although high-resolution photographs of the iris do a wonderful job of capturing Fuchs’ crypts and contraction furrows, the same cannot be said for Wolfflin nodules and pigment spots. Wolfflin nodules are often impossible to see in darker coloured eyes (Falls, 1970). This means that their overall frequency is greatly underestimated in populations where brown eyes predominate. Likewise, ‘pigment spots’ is a category that is used to encompass two types of pigment aberrations: nevi and freckles. Both of these structures affect the iris in different ways. Although pigment spots lay flat on the anterior border layer, nevi ultimately distort the underlying stromal layer (Eagle, 1988; Harbour et al., 2004). It is presently not possible to distinguish between these two types of spots using photographs of the iris. I believe that the use of near infrared photography may be a solution to both of these problems. Infrared light is neither absorbed nor reflected by melanin. Therefore, its use may allow for the visualization of structures that lie below the iris surface. In the case of pigment spots, it may increase the visibility of nevi and decrease the visibility of pigment spots due to their differing effects on the stromal layer.

5.4 Conclusions

Ultimately, this thesis has made the following findings:

141

1) Quantitative methods of measuring eye colour represent an ideal means of investigating iris colour variation in populations of diverse ancestry. Quantitative methods can also be used to successfully quantify heterochromatic variation in the iris. 2) The phenotypic distribution of iris colour shows large population differences. Individuals of European ancestry show the greatest amounts of colour diversity. However, substantial variation can also be observed in East and South Asia. Central heterochromia is largely restricted to lighter coloured irises. 3) Fuchs’ crypts, contraction furrows, Wolfflin nodules and pigment spots show very different distributions in populations of East Asian, European and South Asian ancestry. 4) The marker HERC2 rs12913832 is a major determinant of iris colour variation in individuals of both European and South Asian ancestry. However, it has different effects in each group. 5) The markers associated with iris pigmentation variation in East and South Asia are, to some extent, different from the markers associated with iris pigmentation variation in Europe. OCA2 rs1800414 and OCA2 rs74653330 are the primary determinants of iris colour in the East Asian sample. HERC2 rs12913832, SLC24A5 rs1426654 and LYST rs3768056 are the primary determinants in the South Asian sample. 6) Many of the markers associated with iris pigmentation variation in all three groups have been associated with other pigmentary characteristics, like skin and hair colour. 7) SEMA3A rs10235789 shows an association with Fuchs’ crypt grade in the East Asian, European and South Asian samples. TRAF3IP1 rs3739070 shows a borderline association in the European sample.

Iris colour and structure research has the potential to contribute greatly to the fields of anthropology, forensics, molecular biology and public health. In this thesis, I have shown that both eye colour and surface features are highly complex traits that show considerable inter- population variation. Thus, it is important that research groups branch out and begin to consider the genetic basis and global distribution of both traits in populations outside of Europe. Although the study of these topics is still very much ongoing, this thesis has provided a framework which can be used by other researchers in the same field.

142

5.5 References Andersen, J.D., Johansen, P., Harder, S., Christoffersen, S.R., Delgado, M.C., Henriksen, S.T., Nielsen, M.M., Sørensen, E., Ullum, H., Hansen, T., et al. (2013). Genetic analyses of the human eye colours using a novel objective method for eye colour classification. Forensic Sci. Int. Genet. 7, 508–515. Beleza, S., Johnson, N.A., Candille, S.I., Absher, D.M., Coram, M.A., Lopes, J., Campos, J., Araújo, I.I., Anderson, T.M., Vilhjálmsson, B.J., et al. (2013). Genetic architecture of skin and eye color in an African-European admixed population. PLoS Genet 9, e1003372. Donaldson, D.D. (1961). The significance of spotting of the iris in mongoloids: Brushfield’s spots. Arch. Ophthalmol. 65, 26–31. Eagle, R.C. (1988). Iris pigmentation and pigmented lesions: an ultrastructural study. Trans. Am. Ophthalmol. Soc. 86, 581–687. Eaton, K., Edwards, M., Krithika, S., Cook, G., Norton, H., and Parra, E.J. (2015). Association study confirms the role of two OCA2 polymorphisms in normal skin pigmentation variation in East Asian populations. Am. J. Hum. Biol. 27, 520–525. Edwards, M., Bigham, A., Tan, J., Li, S., Gozdzik, A., Ross, K., Jin, L., and Parra, E.J. (2010). Association of the OCA2 polymorphism His615Arg with melanin content in east Asian populations: further evidence of convergent evolution of skin pigmentation. PLoS Genet 6, e1000867. Eiberg, H., Troelsen, J., Nielsen, M., Mikkelsen, A., Mengel-From, J., Kjaer, K.W., and Hansen, L. (2008). Blue eye color in humans may be caused by a perfectly associated founder mutation in a regulatory element located within the HERC2 gene inhibiting OCA2 expression. Hum. Genet. 123, 177–187. Falls, H.F. (1970). Ocular changes in mongolism. Ann. N. Y. Acad. Sci. 171, 627–636. Frudakis, T., Thomas, M., Gaskin, Z., Venkateswarlu, K., Chandra, K.S., Ginjupalli, S., Gunturi, S., Natrajan, S., Ponnuswamy, V.K., and Ponnuswamy, K.N. (2003). Sequences associated with human iris pigmentation. Genetics 165, 2071–2083. Harbour, J., Brantley, M., Hollingsworth, H., and Gordon, M. (2004). Association between posterior uveal melanoma and iris freckles, iris naevi, and choroidal naevi. Br. J. Ophthalmol. 88, 36–38. Kanetsky, P.A., Swoyer, J., Panossian, S., Holmes, R., Guerry, D., and Rebbeck, T.R. (2002). A polymorphism in the agouti signaling protein gene is associated with human pigmentation. Am. J. Hum. Genet. 70, 770–775. Kayser, M., Liu, F., Janssens, A.C.J.W., Rivadeneira, F., Lao, O., Duijn, K. van, Vermeulen, M., Arp, P., Jhamai, M.M., IJcken, W.F.J. van, et al. (2008). Three genome-wide association studies

143 and a linkage analysis identify HERC2 as a human iris color gene. Am. J. Hum. Genet. 82, 411– 423. Kuehni, R.G., and Marcus, R.T. (1979). An experiment in visual scaling of small color differences*. Color Res. Appl. 4, 83–91. Lamason, R.L., Mohideen, M.-A.P., Mest, J.R., Wong, A.C., Norton, H.L., Aros, M.C., Jurynec, M.J., Mao, X., Humphreville, V.R., and Humbert, J.E. (2005). SLC24A5, a putative cation exchanger, affects pigmentation in zebrafish and humans. Science 310, 1782–1786. Larsson, M., Duffy, D.L., Zhu, G., Liu, J.Z., Macgregor, S., McRae, A.F., Wright, M.J., Sturm, R.A., Mackey, D.A., Montgomery, G.W., et al. (2011). GWAS findings for human iris patterns: associations with variants in genes that influence normal neuronal pattern development. Am. J. Hum. Genet. 89, 334–343. Liu, F., Wollstein, A., Hysi, P.G., Ankra-Badu, G.A., Spector, T.D., Park, D., Zhu, G., Larsson, M., Duffy, D.L., Montgomery, G.W., et al. (2010). Digital quantification of human eye color highlights genetic association of three new loci. PLoS Genet 6, e1000934. McLaren, K. (1976). XIII—The development of the CIE 1976 (L* a* b*) uniform colour space and colour‐difference formula. J. Soc. Dye. Colour. 92, 338–341. Sidhartha, E., Gupta, P., Liao, J., Tham, Y.-C., Cheung, C.Y., He, M., Wong, T.Y., Aung, T., and Cheng, C.-Y. (2014a). Assessment of iris surface features and their relationship with iris thickness in Asian eyes. Ophthalmology 121, 1007–1012. Sidhartha, E., Nongpiur, M.E., Cheung, C.Y., He, M., Wong, T.Y., Aung, T., and Cheng, C.-Y. (2014b). Relationship between iris surface features and angle width in Asian eyes. Invest. Ophthalmol. Vis. Sci. 55, 8144–8148. Stokowski, R.P., Pant, P.K., Dadd, T., Fereday, A., Hinds, D.A., Jarman, C., Filsell, W., Ginger, R.S., Green, M.R., and van der Ouderaa, F.J. (2007). A genomewide association study of skin pigmentation in a South Asian population. Am. J. Hum. Genet. 81, 1119–1132. Sturm, R.A., Duffy, D.L., Zhao, Z.Z., Leite, F.P.N., Stark, M.S., Hayward, N.K., Martin, N.G., and Montgomery, G.W. (2008). A single SNP in an evolutionary conserved region within intron 86 of the HERC2 gene determines human blue-brown eye color. Am. J. Hum. Genet. 82, 424– 431. Sulem, P., Gudbjartsson, D.F., Stacey, S.N., Helgason, A., Rafnar, T., Magnusson, K.P., Manolescu, A., Karason, A., Palsson, A., Thorleifsson, G., et al. (2007). Genetic determinants of hair, eye and skin pigmentation in Europeans. Nat Genet 39, 1443–1452.

144

Chapter 1 has been previously published as: Edwards, M., Gozdzik, A., Ross, K., Miles J., and Parra, E.J. (2012) Technical note: quantitative measures of iris colour using high resolution photographs. Am. J. Phys. Anthro. 147, 141-149.

Permission has been acquired from John Wiley and Sons for publication in this thesis (License Number: 3802541113357)

Chapter 2 has been previously published as: Edwards, M., Cha, D., S. Krithika., Johnson, M., Cook, G., Parra, E.J. (2015) Iris pigmentation as a quantitative trait: variation in populations of European, East Asian and South Asian ancestry and association with candidate gene polymorphisms. Pigment Cell Melanoma Res.

This article was published as open access. No copyright permissions were required.

Chapter 3 has been previously published as: Edwards, M., Cha, D., S. Krithika., Johnson, M., Parra, E.J. (2016) Analysis of iris surface features in populations of diverse ancestry. R. Soc. Open Sci. 3, 150424.

This article was published as open access. No copyright permissions were required.