Revealing the genetic basis of eyelid pigmentation in Hereford cattle

Abstract

Ocular squamous cell carcinoma and infectious keratoconjunctivitis are common ocular pathologies in Hereford cattle with considerable economic impact. Both pathologies have been associated with low eyelid pigmentation, and thus, genetic selection for higher eyelid pigmentation could reduce their incidence. The objective of the present study was to reveal the genetic basis of eyelid pigmentation in Hereford cattle. The analysis included a single-step genome-wide association study (ssGWAS) and a subsequent gene-set analysis in order to identify individual genes, genetic mechanisms, and biological pathways implicated in this trait. Data consisted of eyelid pigmentation records in 1,165 Hereford bulls and steers, visually assessed in five categories between 0% and 100%. Genotypic data for 774,660 single-nucleotide polymorphism markers were available for 886 animals with pigmentation records. Pedigree information of three generations of ancestors of animals with phenotype was considered in this study, with a total of 4,929 animals. Our analyses revealed that eyelid pigmentation is a moderately heritable trait, with heritability estimates around 0.41. The ssGWAS identified at least eight regions, located on BTA1, BTA3, BTA5, BTA14, BTA16, BTA18, BTA19, and BTA24, associated with eyelid pigmentation. These regions harbor genes that are directly implicated in melanocyte biology and skin pigmentation, such as ADCY8, PLD1, KITLG, and PRKCA. The gene-set analysis revealed several functional terms closely related to melanogenesis, such as positive regulation of melanocyte differentiation and regulation of ERK1 and ERK2 cascade. Overall, our findings provide evidence that eyelid pigmentation is a heritable trait influenced by many loci. Indeed, the ssGWAS detected several candidate genes that are directly implicated in melanocyte biology, including melanogenesis. This study contributes to a better understanding of the genetic and biological basis of eyelid pigmentation and presents novel information that could aid to design breeding strategies for reducing the incidence of ocular pathologies in cattle. Additional research on the genetic link between eyelid pigmentation and ocular pathologies is needed.

Eyelid pigmentation in Hereford cattle is a complex trait, with a moderate heritability, affected by multiple loci, including genes related to melanocyte biology, melanogenesis, and melanomagenesis.

Introduction

The two most common ocular pathologies affecting cattle production are ocular squamous cell carcinoma (SCC) and infectious keratoconjunctivitis (IBK; Tsujita and Plummer, 2010; Irby and Angelos, 2018). Both pathologies have considerable economic impact on beef production since affected animals have poorer performance and reduced productive life and are underpaid at slaughter due to low final weights and sanitary issues (Seid, 2019). Despite the relevance of SCC and IBK, little is known about the genetic basis of these two diseases in cattle. Both diseases have low heritabilities, with estimates around 0.10 for SSC (Russell et al., 1976), and between 0.00 and 0.26 for IBK (Snowder et al., 2005).

High incidence of SCC and IBK have been reported in cattle breeds with white skin in the head, such as Hereford, Simmental, and Holstein. Of particular interest, studies comparing different cattle breeds have shown that Hereford has the highest incidence of ocular pathologies (Snowder et al., 2005). Low eyelid pigmentation is considered as a predisposing factor that leads to the occurrence of SCC, ocular lesions, and subsequent IBK (Snowder et al., 2005; Angelos, 2015), particularly under increasing exposure to solar radiation (Pausch et al., 2012). Although there is no information available on the genetic correlations of eyelid pigmentation with either SCC or IBK, eyelid pigmentation is arguably a trait that can be easily recorded, and hence, it could be expected that selection of animals with pigmented eyelids would contribute to decrease the incidence of ocular pathologies in the population (Vogt et al., 1963).

Eyelid pigmentation is related to melanin pigment production in melanocytes in the skin. Although eyelid pigmentation in cattle has received less attention, skin pigmentation has been studied and characterized in humans (Rees, 2004) and domestic animals (Anderson, 1991). Several studies have been carried out to identify genomic regions and individual genes associated with pigmentation patterns in cattle (Pausch et al., 2012; Fan et al., 2014; Mészáros et al., 2015). Differences in skin pigmentation are due to differences in melanocyte activity in combination with other nongenetic factors, such as environmental and physiological conditions, e.g., exposure to ultraviolet radiation and reaction to thermal stress (Jablonski and Chaplin, 2000). Both animals and humans with higher degree of pigmentation have a greater proportion of melanin in the skin, and hence, better protection against ultraviolet radiation (Haider et al., 2014; Solano, 2014). Our previous work has shown that eyelid pigmentation in cattle is related to the expression of genes involved in melanocyte biology, inflammatory response, infectious processes, and tumoral pathways (Jara et al., 2020).

The objective of this study was to reveal the genetic basis of eyelid pigmentation in Hereford cattle. We first estimated the heritability of eyelid pigmentation. Then, we performed a single-step genome-wide association study (ssGWAS) and a subsequent gene-set analysis to detect individual genes and functional gene-sets that could explain part of the variation observed in eyelid pigmentation. The findings of this study improve our understanding of eyelid pigmentation in cattle but could also contribute to the development of effective breeding strategies for reducing the incidence of ocular pathologies.

Materials and Methods

Phenotypic, pedigree, and genomic data

Pigmentation records were collected at the Central de Pruebas de Toros de Kiyú (San José, Uruguay) between 2014 and 2018. Percentage of eyelid pigmentation in each eye was visually assessed by two technicians when animals were 382 ± 95 d old. The percentage of the eyelid area that was pigmented of each eye was categorized in five levels, namely 0%, 25%, 50%, 75%, and 100% pigmentation. A total of 1,165 Hereford animals, including bulls and steers, had pigmentation records and were used in this study.

The pedigree file was created by tracing the pedigree of animals with pigmentation records back to three generations, and included a total of 4,929 animals. In addition, 886 of the animals with pigmentation records had genotypic information. These animals were genotyped with the Illumina High-Density Bovine BeadChip with roughly 777,000 single-nucleotide polymorphism (SNPs). SNPs were removed from the SNP data set if they mapped to the sex chromosomes, were monomorphic, had a minor allele frequency less than 5% or call rates less than 90%. After data editing, a total of 591,755 SNP markers were retained for subsequent genomic analyses.

Variance component estimation

Variance components were estimated for both left and right eyelid pigmentation, and also for total eyelid pigmentation, defined as the sum of eyelid pigmentation in each eye, using the following univariate linear mixed model (model 1),

$${\displaystyle \begin{array}{l}{\displaystyle \mathbf{y}=\mathbf{X}\mathrm{\beta }+\mathbf{Z}\mathbf{u}+\mathbf{e}}\end{array}}$$

where is y the vector of pigmentation records, $\mathrm{\beta }$ is a vector that includes technician (2 levels), sex (2 levels; bulls and steers), and age as linear and quadratic covariates, and $\mathbf{u}$ is the vector of additive genetic effects. The matrices $\mathbf{X}$ and $\mathbf{Z}$ are the corresponding incidence matrices. The effects $\mathbf{u}$ and $\mathbf{e}$ were assumed to follow a normal distribution with $\mathbf{u}\sim N(0,\mathbf{A}{\sigma }_u^2)$ and $\mathbf{e}\sim N(0,\mathbf{I}{\sigma }_u^2)$, where $\mathbf{A}$ is the additive relationship matrix and $\mathbf{I}$ is the identity matrix, ${\sigma }_u^2$ is the additive genetic variance, and ${\sigma }_e^2$ is the residual variance. Note that genomic data were not included for variance component estimation.

In addition, (co)variance components were estimated for left and right eyelid pigmentation in a bivariate analysis using models that included the same effects as in the univariate analysis (model 1). The effects $\mathbf{u}$ and $\mathbf{e}$ were assumed to follow the multivariate normal distribution with mean zero and $\mathbf{V}\mathbf{a}\mathbf{r}(\mathbf{u})=\mathbf{K}\otimes \mathbf{A}$ and $\mathbf{V}\mathbf{a}\mathbf{r}(\mathbf{e})=\mathbf{R}\otimes \mathbf{I}$, where $\mathbf{A}$ is the numerator relationship matrix, $\mathbf{I}$ is the identity matrix, and $\mathbf{K}$ and $\mathbf{R}$ are additive genetic and residual covariance matrices, respectively; and $\otimes$ is the Kronecker product.

Variance and covariance components were estimated in a Bayesian framework using the software GIBBS2F90 (version 1.93; Misztal et al., 2002). A total of 500,000 iterations were performed, with the first 100,000 iterations discarded as burn-in. Every 100th sample was saved to calculate features of the posterior distributions, such as posterior means and standard deviations. This sampling approach resulted in a total of 4,000 samples that were used for post-Gibbs analysis using the software POSTGIBBSF90 (version 3.14; Misztal et al., 2002).

Heritability (h²) for left and right eyelid pigmentation and total eyelid pigmentation was calculated as $h^2={\sigma }_u^2/({\sigma }_u^2+{\sigma }_e^2)$, where ${\sigma }_u^2$ is the additive genetic variance and ${\sigma }_e^2$ is the residual variance. In addition, the genetic correlation between right and left eyelid pigmentation was estimated as $r_{LR}^G={\sigma }_{LR}/\sqrt{{\sigma }_L^2\cdot {\sigma }_R^2}$, where ${\sigma }_L^2$ and ${\sigma }_R^2$ are the additive genetic variance of left and right eyelid pigmentation, respectively, and ${\sigma }_{LR}$ is the genetic covariance.

Genomic regions and individual genes

Single-step genome-wide association study (ssGWAS) was performed to identify genomic regions and candidate genes that affect total eyelid pigmentation. The method, initially developed for performing genomic prediction and later extended for performing gene mapping, allows the inclusion of genotyped and ungenotyped animals in a single genomic analysis, based on genomic breeding values (GEBVs) that were estimated using model 1 but replacing the pedigree relationship matrix A by the matrix H, which combines pedigree and genotypic information (Aguilar et al., 2010). The inverse of H was calculated as

$${\displaystyle \begin{array}{l}{\displaystyle {\mathbf{H}}^{-1}={\mathbf{A}}^{-1}+\left(\begin{array}{cc}0& {\displaystyle 0}\\ 0& {\displaystyle {\mathbf{G}}^{-1}-{\mathbf{A}}_{22}^{-1}}\end{array}\right)}\end{array}}$$

where${\mathbf{A}}^{-1}$ is inverse of the pedigree relationship matrix, ${\mathbf{A}}_{22}^{-1}$ is the inverse of the pedigree relationship matrix of the animals with genomic data. The additive genomic relationship matrix $\mathbf{G}$ was calculated according to the first method in VanRaden (2008) with allele frequencies directly estimated from genotypes. The matrix ${\mathbf{G}}^{-1}$ is the inverse genomic relationship matrix, and has the dimension of 886 × 886, while the ${\mathbf{A}}^{-1}$ matrix has the dimension of 4,929 × 4,929.

Genomic regions associated with total eyelid pigmentation were identified based on two alternative approaches. The first approach was the amount of genetic variance explained by 2.0 Mb window of adjacent SNPs evaluated across the entire bovine genome. Given the GEBVs, the SNP effects can be estimated as $\mathbf{s}={\mathbf{D}\mathbf{M}}^{\prime }{\left[\mathbf{M}\mathbf{D}{\mathbf{M}}^{\prime }\right]}^{-1}$, a_g where s is the vector of SNP marker effects, $\mathbf{D}$ is a diagonal matrix of weights of SNPs, $\mathbf{M}$ is a matrix relating genotype of each SNP marker to observations, and ${\mathbf{a}}_{\mathbf{g}}$ is the vector of GEBVs for genotyped individuals (Wang et al., 2012). The percentage of genetic variance explained by a given 2.0 Mb region was then calculated as described by Wang et al. (2014):

$${\displaystyle \begin{array}{l}{\displaystyle \frac{var\left(u_i\right)}{{{\sigma }^2}_a}\times 100=\frac{var({\displaystyle \sum _{j=1}^B}{M}_j{s}_j)}{{{\sigma }^2}_a}}\end{array}}$$

Where ${\sigma }_a^2$ is the additive genetic variance, $u_i$ is the genetic value of the ith genomic region under consideration, B is the total number of adjacent SNPs within 2.0 Mb region, $M_j$ is the genotype code of jth marker, $s_j$ is the marker effect of the jth SNP within the ith region. In this study, all SNPs were equally weighted. The second approach was the identification of relevant SNP markers based on their P-values (Aguilar et al., 2019). All these calculations were performed using the program POSTGSF90 (Aguilar et al., 2014; 2019).

Gene-set analysis

Gene-set analysis, also known as overrepresentation or enrichment analysis, is a powerful tool to reveal biological pathways and molecular mechanisms underlying complex phenotypes. The gene-set analysis consisted basically of three steps (Han and Peñagaricano, 2016): first, the assignment of SNP markers to annotated genes; then, the assignment of genes to functional gene-sets; and finally, the association between each functional term and the phenotype of interest.

The assignment of SNPs to bovine genes was performed using the Bioconductor R package biomaRt (version 2.48.3) based on the UMD3.1 bovine genome assembly (Durinck et al., 2009). SNPs were assigned to genes if they were located within the genomic sequence of the gene or at most 5 kb either upstream or downstream of the gene. Putative genes affecting eyelid pigmentation were defined as those genes that contained at least one significant SNP (P-value ≤ 0.05). Different gene databases, including GO, KEGG, MeSH, Reactome, InterPro, and Molecular Signatures Database (MsigDB), were used to define functional sets of genes. Finally, the identification of gene-sets significantly associated with eyelid pigmentation was evaluated using a Fisher’s exact test (Peñagaricano et al., 2013). The gene-set analysis was performed using the R package EnrichKit, developed by Lihe Liu and Francisco Peñagaricano, available at https://github.com/liulihe954/EnrichKit.

Results and discussion

Variance component estimation

Variance components for left and right eyelid pigmentation were estimated using a classical best linear unbiased prediction (BLUP) analysis. Results show that both pigmentation traits are moderately heritable, with heritability estimates of $h_L^2$= 0.26 ± 0.06 and $h_R^2$= 0.31 ± 0.06, for left and right eyelid pigmentation, respectively (Table 1). We also evaluated the genetic correlation between left and right eyelid pigmentation using a bivariate BLUP analysis. As expected, these two traits are highly correlated ($r_{LR}^G$ = 0.94 ± 0.06; Table 1). Now, given its practical simplicity, we decided to use total eyelid pigmentation, defined as the sum of the eyelid pigmentation in both eyes, as the trait of interest. Total eyelid pigmentation captures the variation in both eyes, with a larger number of classes and a more continuous distribution. This trait is moderately heritable with a heritability estimate equal to 0.41 ± 0.12 (Table 1). This heritability estimate was lower than previous studies also performed in Hereford cattle (0.55, Vogt et al., 1963) and human (0.83, Clark et al., 1981), probably due to important difference in the methodology available decades ago. Vogt et al. (1963) estimated the heritability using paternal half-sib analysis of variance, while Clark et al. (1981) estimated the heritability using data from pair of twins and the mean squares method. On the other hand, our heritability estimate is very similar to the value reported by Reimann et al. (2018) working on Hereford and Braford. These authors estimated the heritability by BLUP (only pedigree information) and by single-step genomic BLUP (pedigree and genotypes), resulting both in heritability estimates of 0.46.

Table 1.

Relevant genetic parameters associated with eyelid pigmentation in Hereford cattle.

Parameters	Estimates
${\sigma }_u^2$	2116 (661) [956; 3419]
${\sigma }_e^2$	2982 (620) [1745; 4115]
$h^2$	0.41 (0.12) [0.17; 0.63]
$h_L^2$	0.26 (0.06) [0.04; 0.49]
$h_R^2$	0.31 (0.06) [0.08; 0.54]
$r_{LR}^G$	0.94 (0.06) [0.81; 1]

Additive genetic variance $({\sigma }_u^2)$, residual variance $({\sigma }_e^2)$ and heritability $(h^2)$ of total eyelid pigmentation, defined as the sum of eyelid pigmentation in both eyes.

Heritability of the left $(h_L^2)$ and right $(h_R^2)$ eyelid pigmentation, respectively.

Genetic correlation ($r_{LR}^G)$ between right and left eyelid pigmentation.

Posterior standard deviations are in parenthesis, while highest posterior density intervals (95%) are shown in brackets.

Overall, the magnitude of the heritability indicates that it would be possible to increase the content of pigments in the eyelids of Hereford cattle by genetic selection, and thus contribute to reducing the predisposition to ocular pathologies. Note that direct selection against ocular pathologies is complicated to implement given that these complex phenotypes are difficult to collect and lowly heritable (0.10 for SSC, Russell et al., 1976; 0.00 to 0.26 for IBK, Snowder et al., 2005). Interestingly, our results suggest that eyelid pigmentation is more heritable than either SCC or IBK, which reinforces the value of eyelid pigmentation as selection criteria to reduce the incidence of ocular diseases in the Hereford population. Nevertheless, the magnitude of the correlated response depends on the genetic correlation between eyelid pigmentation and the ocular pathologies, which requires further investigations.

Gene mapping

Two complementary approaches, proportion of genetic variance explained and P-values, were used to identify genomic regions and putative genes associated with eyelid pigmentation. On the one hand, the ssGWAS method allows to identify genomic regions that explain a given amount of genetic variance. The magnitude of the genetic variance depends on the marker effects and the allelic frequencies. On the other hand, the ssGWAS method also allows to formally evaluate the significance of the association (using a statistical test) between each marker and the phenotype of interest. The significance of the test depends on the estimated marker effect and the standard error of the estimate. Here, these two approaches yielded similar results (Figure 1) and we defined as candidate regions and genes those with proportion of variance explained ≥0.5% and nominal P-value $\le$0.05 (Table 1).

Figure 1.

Genomic analysis of total eyelid pigmentation. (A) Genomic regions implicated in the pigmentation of eyelids based on the amount of genetic variance explained by 2.0 Mb windows. (B) Genomic regions implicated in the pigmentation of eyelids based on P-values.

The ssGWAS identified sizeable peaks on BTA1, BTA5, BTA14, BTA16, and BTA19 associated with pigmentation in the eyelids. These genomic regions harbor putative genes, such as PLD1 (BTA1, 95.37 to 97.37 Mb, P-value = 0.03), KITLG (BTA5, 16.51 to 18.51 Mb, P-value = 0.02), ADCY8 (BTA14, 10.11 to 12.11 Mb, P-value = 8.5 × 10⁻⁰⁵), and PRKCA (BTA19, 63.50 to 63.58 Mb, P-value = 0.002), that are all implicated in melanogenesis and skin pigmentation (Table 2). For instance, gene ADCY8 encodes a membrane-bound enzyme that catalyzes the formation of cyclic AMP from ATP, and high levels of cyclic AMP trigger the expression of several key genes, including TYR, TYRP1, and DCT, that are directly implicated in the quality and quantity of melanin production (Takeda et al., 2007; Brenner and Hearing, 2008). Genes TYR, TYRP1, and DCT showed upregulated expression in pigmented eyelid samples in cattle (Jara et al., 2020). Both genes PLD1 and PRKCA play key roles in the regulation of melanogenesis (D’Mello et al., 2016). Gene KITLG is involved in melanocyte development, melanocyte proliferation and melanin distribution. It has been shown that mutations in KITLG cause different pigmentation disorders (Amyere et al., 2011), and of special interest, KITLG has been suggested as a candidate gene affecting UV-protective eye area pigmentation in cattle (Pausch et al., 2012). Overall, our analysis identified several genes with known roles in melanogenesis and skin pigmentation as putative genes affecting eyelid pigmentation in cattle. Notably, some of these candidate genes, such as ADCY8 and KITLG, participate in multiple signaling pathways that activate melanogenesis, including the cAMP-PKA-CREB and the RAS/MAPK pathways (Dessinioti et al., 2011; D’Mello et al., 2016).

Table 2.

Candidate regions and genes associated with eyelid pigmentation in Hereford cattle

Chromosome	Position	Variance (%)	P-value	Candidate genes
BTA14	10.11 to 12.11	1.69	8.5 × 10−05	otoconin 90, ADCY8, GSDMC
BTA18	24.86 to 26.86	0.7	0.001	NUP93, NLRC5, SETD6, CX3CL1
BTA24	8.69 to 10.69	0.67	0.002	TMX3
BTA1	95.37 to 97.37	0.58	0.03	PLD1
BTA16	61.27 to 63.27	0.57	0.003	CEP350
BTA6	73.14 to 75.14	0.56	0.02	IGFBP7
BTA5	16.51 to 18.51	0.54	0.02	KITLG
BTA19	61.98 to 63.98	0.50	0.002	PRKCA

Of special interest, the ssGWAS identified several candidate genes that are directly implicated in the development of melanoma, including otoconin 90 and GSDMC (BTA14, 10.11 to 12.11 Mb, P-value = 8.5 × 10⁻⁰⁵), NUP93, NLRC5, SETD6, and CX3CL1 (BTA18, 24.86 to 26.86 Mb, P-value = 0.001), TMX3 (BTA24, 8.69 to 10.69, P-value = 0.002), CEP350 (BTA16, 61.27 to 63.27 Mb, P-value = 0.003), and IGFBP7 (BTA6, 73.14 to 75.14 Mb, P-value = 0.02; Table 2). For instance, the expression of otoconin 90 is associated with different types of cancer, especially skin melanoma (Pearlman et al., 2019). Gene GSDMC, a member of the gasdermin family, is upregulated in metastatic melanoma, and is involved in the course of tumorigenesis and melanoma progression (Watabe et al., 2001). It was also identified as one of the genes affecting UV-protective eye area pigmentation in cattle by Pausch et al. (2012). Gene NUP93 mediates the repression of gene HOXA, and aberrant expression levels of HOXA are correlated with different cancers, including cutaneous melanoma (Maeda et al., 2005). Genes NLRC5 and CEP350 stimulate the protective antitumor immunity in melanoma (Mann et al., 2015; Chelbi and Guarda, 2016; Rodriguez et al., 2016). Indeed, elevated expression of NLRC5 is associated with decreased melanoma growth in murine models and prolonged survival of humans with skin cancer (Kim et al., 2020). Gene SETD6 is upregulated in melanocytes of dark-skinned individuals (López López et al., 2015), as well as in melanoma cancer cell lines (Mukherjee et al., 2017). Gene TMX3 is involved in the stage and severity of the melanoma (Zhang et al., 2019). Gene IGFBP7 mediates senescence in melanocytes and suppresses melanoma growth in vivo by inducing apoptosis (Wajapeyee et al., 2008). In summary, our findings provide evidence that genes implicated in skin cancer development explain at least part of the differences observed in eyelid pigmentation in cattle. Both candidate genes involved in melanogenesis, as well as candidate genes involved in melanomagenesis, are promising targets for future research aimed to identify functional mutations affecting eyelid pigmentation in cattle.

Gene-set analysis

A subset of 1,344 bovine genes were considered as significantly associated with eyelid pigmentation, given that these genes were flagged by at least one significant SNP (P-value ≤ 0.05; Figure 1B). Figure 2 shows the most relevant gene-sets associated with eyelid pigmentation. The most relevant terms were involved in at least three processes, namely positive regulation of melanocyte differentiation, protein kinase activity, and regulation of ERK1 and ERK2 cascade (Supplementary Table S1).

Figure 2.

Functional terms and pathways significantly enriched with genes associated with total eyelid pigmentation. Six gene annotation databases were analyzed: GO, Kyoto Encyclopedia of Genes and Genomes (KEEG), MeSH, Reactome, InterPro and MsigDB. The y-axis displays the names of the significant terms and pathways. The black dots represent the significance of enrichment (−log₁₀P-value, Fisher’s exact test), and the x-axis represents the percentage of significant genes in each functional term.

Some of the enriched terms are directly involved in skin pigmentation, including positive regulation of melanocyte differentiation (GO: 0045636), regulation of ERK1 and ERK2 cascade (GO: 0070372), positive regulation of Ras protein signal transduction (GO: 0046579), protein kinase domain (GO: IPR000719), and protein kinase activity (GO: 0004672; Figure 2). Gene KITLG is involved in melanin distribution, melanocyte proliferation and melanocyte development. Mutations in KITLG gene cause different pigmentation disorders (Amyere et al., 2011) and have been suggested as a candidate gene affecting UV-protective eye area pigmentation in cattle (Pausch et al., 2012). Gene ADAMTS20 belongs to the family of metalloproteases, and it has been shown that point mutations in ADAMTS20 gene cause pigment defect in mouse (Rao et al., 2003). Indeed, mutations in ADMTS20 cause an increase in the death of pigment cells, thus generating a decrease in the number of pigment cells (Silver et al., 2008). In addition, Adamts20 mutant mice have disrupted the function of KIT and/or KITLG, proteins that regulate pigment cell development (Silver et al., 2008). Finally, two enriched terms are directly involved in the via of the melanogenesis, namely regulation of ERK1 and ERK2 cascade (GO: 0070372; Herraiz et al., 2011) and positive regulation of Ras protein signal transduction (GO: 0046579; D’Mello et al., 2016).

Conclusions

Our study provides a comprehensive overview of the genetic and biological basis of eyelid pigmentation in Hereford cattle. First, our results indicate that eyelid pigmentation is a heritable trait in cattle, and the magnitude of the heritability estimate suggests an important scope for increasing eyelid pigmentation by genetic selection. Given the relationship between ocular pathologies and pigmentation reported in the literature, direct selection for eyelid pigmentation could reduce the incidence of ocular diseases in the Hereford population, including ocular squamous cell carcinoma and infectious keratoconjunctivitis. Further research is needed to estimate the genetic correlation between eyelid pigmentation and the ocular pathologies and quantify the expected correlated genetic response. Second, the ssGWAS revealed that eyelid pigmentation is a complex trait affected by multiple loci, including genes related to melanocyte biology, melanogenesis, skin pigmentation, and development of melanoma. Third, the overrepresentation analysis provided further evidence that biological processes such as melanocyte differentiation explain part of the observed variation in eyelid pigmentation. Overall, this study provides a better understanding of the genetics underlying eyelid pigmentation in Hereford and could contribute to the development of more effective breeding strategies for reducing the incidence of ocular pathologies in cattle by genomic selection.

Glossary

Abbreviations

BLUP

best linear unbiased prediction

GO

Gene Ontology

GEBV

genomic breeding value

IBK

infectious keratoconjunctivitis

SCC

squamous cell carcinoma

SNP

single-nucleotide polymorphism

ssGWAS

single-step genome-wide association study

Conflict of Interest Statement

The authors declare no real or perceived conflicts of interest.