Three Novel KIT Polymorphisms Found in Horses with White Coat Color Phenotypes

Simple Summary

While more than 50 alleles are associated with white spotting in horses, there remain many instances of white coat patterns with no genetic explanation. Of the previously identified variants, over 35 involve the proto-oncogene Receptor Tyrosine Kinase (KIT), which is critical for melanocyte survival and proliferation. This study investigated three horse groups with heritable white spotting phenotypes of no known genetic cause. After screening candidate genes previously associated with depigmentation in mammals, we identified three novel variants impacting the coding sequence of KIT. Further examination of these variants predicted that the mutations are deleterious to protein function. One novel variant was only observed in the presence of a single W35 allele, suggesting the alleles may be in complete linkage. We propose to term the novel variants W37, W38, and W39 in accordance with standard nomenclature. We report these three novel genetic variants as being likely to cause otherwise unexplained white spotting patterns in horses, and plan future work to examine this association with white spotting in larger populations. Depigmentation is a common phenotype that influences the economic value of a horse; thus, understanding the genetics behind this phenotype is valuable to horse breeders and owners selecting for specific coat color patterns.

Abstract

This paper reports three novel KIT variants likely responsible for previously unexplained white patterning phenotypes observed in three groups of horses. White spots and markings may have substantial consequences on the value and health of domesticated horses. This study aims to elucidate the genetic mechanisms underlying depigmented coat colors to aid in producing prosperous herds. Aligned whole genome sequences were manually screened to identify three polymorphisms in a family of Anglo-Arabian horses (N = 7), a family of Warmblood horses (N = 5), and a single stock-type mare with unexplained white markings. Sanger sequencing confirmed the presence of the variants, and in silico predictive programs were used to predict the functional impacts of each. We propose to term the novel variants W37, W38, and W39, respectively, per convention. The W37 polymorphism was always observed in the presence of one W35 allele, suggesting complete linkage. All three variants were predicted to alter or remove the KIT protein active domain, repressing typical protein folding and impacting pathways that upregulate pigmentation. The severe predicted impact on biological function suggests that these variants may cause increased white spotting, providing a possible explanation for the depigmentation phenotypes observed in affected individuals.

1. Introduction

While advances in genome sequencing have enabled the detection of genes that influence coat color, the genetic mechanisms behind many white spotting phenotypes in horses remain unknown. White spotting traits frequently impact the commercial and sentimental value of domesticated equids. White markings may determine whether a horse is eligible to enter a breed registry, which strongly affects the economic value of the individual. While some registries select against white markings, others are opening their studbooks to horses with white alleles in an attempt to decrease inbreeding within the population, or may positively select for white spotting phenotypes [1,2,3]. In addition, certain white mutations may lead to deleterious phenotypes, such as sterility or higher risk of deafness or blindness [4,5,6,7,8,9]. For these reasons, detecting the presence of genetic variants that cause white spotting phenotypes has become a principal interest for horse breeders aiming to produce distinctive and healthy herds.

Genetic variations associated with depigmentation (or white spotting) in skin and hair are common in domesticated mammals. In horses, more than 35 of over 50 currently reported white spotting mutations [10] are attributed to mutations in the proto-oncogene, Receptor Tyrosine Kinase (KIT), as reviewed in McFadden et al. [1]. In mammals, successful melanoblast proliferation and differentiation requires signaling by the KIT protein [11,12,13], and some of these variants are predicted to be embryonic lethal in the homozygote [14]. Given that polymorphisms disrupting function or regulation of KIT lead to white spotting in humans, mice, and horses, KIT is a likely candidate for novel white spotting alleles. Here, we report three novel polymorphisms in KIT: a frameshift insertion leading to a premature stop codon, a splice site variant, and a stop-gain variant. This study is a continuation of ongoing research which endeavors to identify the genetic basis for unexplained instances of depigmentation in horses. Building upon previous work in equine genetics, this research expands the existing record of white spotting variants on the KIT gene. The findings gathered in this study aim to explain depigmentation in horses belonging to three different breeds.

2. Materials and Methods

2.1. Horses

Three groups of horses were submitted, with photographs, to Etalon, Inc. for commercial genetic testing. The first family (Family 1) consisted of a half-Arabian, half-Thoroughbred sire and six of his offspring, belonging to the Anglo-Arabian (n = 6) and Zangersheide (n = 1) breed registries. Two horses, the sire and one offspring, were heterozygous for the Tobiano (TO) allele and possessed a copy of Dominant White 35 (W35). One offspring was homozygous for Tobiano. The other four offspring carried a copy of W35 but did not have Tobiano. All tobiano horses presented large patches of white markings, which is typical of the TO variant. The four half-siblings without the TO variant displayed white markings on the face and scattered body markings with pink patches of skin and depigmented legs. These four horses exhibited more depigmentation than expected based on their genotypes for other known white spotting alleles.

In addition to the TO variant, several members of Family 1 possessed copies of the Eden White 3 (EDXW3) allele. Found across diverse horse breeds, EDXW3 is associated with white spotting phenotypes such as depigmented faces and white legs. Further, horses with two copies of EDXW3 display an increase in white spotting phenotypes relative to individuals with just one copy of the allele [15]. Two offspring who displayed scattered white markings on their coats possessed neither TO nor EDXW3. No horses in Family 1 possessed any white spotting alleles other than those explicitly mentioned (Figure 1).

The second family (Family 2) consisted of a Warmblood mare, her dam, her half-sibling, and her two offspring. One of the offspring showed no white markings on her body while the other four family members displayed variable white markings from sabino-like (n = 1) to all-white (n = 3). The sabino horse carried a single copy of EDXW3 and no other white alleles. Of the three all-white horses, one carried a single copy of W20 and another carried a copy of EDXW3 and Dominant White 34 (W34) in-phase with W35. One all-white horse did not carry any known white alleles. None of the horses in Family 2 carried any white spotting alleles other than those explicitly mentioned, nor did they carry the Grey (G) allele (Figure 2).

A single stock-type mare displaying a nearly all-white phenotype was also investigated in this study. No relatives of this individual were available for phenotyping or genotyping. This mare exhibits some red-pigmented areas of skin and hair limited to her mane, tail, and small patches on the body. Black pigment is retained in all four legs from the fetlock to the knee. No known white spotting alleles or the Grey allele were observed in this horse.

2.2. Genotyping and Variant Screening

Genomic DNA (gDNA) was extracted from 20 to 30 roots of mane or tail hair using the Puregene Extraction Kit and following the manufacturer’s protocol (QIAGEN, Inc., Germantown, MD, USA). DNA libraries were prepared using the freshly extracted gDNA with xGen Library Preparation Kits, according to the manufacturer’s protocol (Integrated DNA Technologies, Inc., Coralville, IA, USA). Sequences of 150 bp paired-end reads from the NextSeq1000 (Illumina, Inc., San Diego, CA, USA) were aligned to the EquCab3.0 reference genome [16]. The resulting sequences had base call quality scores ≥ Q30 and read depth >40× for all regions under investigation. The binary alignment map (BAM) files produced from the sequencing run were manually screened for novel mutations using NCBI Equus caballus Annotation Release 103 gene annotations and the European Variant Archive (EVA) SNP Release 5 on the Integrated Genomics Viewer (IGV) [17,18,19]. Exonic regions of KIT, MITF, PAX3, ENDRB, SOX10, LMX1A, EDN3, HPS5, MCOLN3, SWAP70, and DOCK7 were screened for novel mutations due to the involvement of these genes in pigmentation across diverse mammalian species [1,15,20,21,22,23,24,25,26,27,28,29,30,31,32]. Novel variants were analyzed for predicted significance using the Expert Protein-Analysis System (ExPASy) Translate tool, the Simple Modular Architecture Research Tool (SMART) (version 9), the Iterative Threading ASSEmbly Refinement (I-TASSER) server, and the GENSCAN web server (version 1.0), all with default parameters [33,34,35,36]. Novel variants were confirmed through Sanger sequencing using a commercial service (Azenta Life Sciences, Inc., Burlington, MA, USA).

3. Results and Discussion

3.1. Novel KIT Insertion Identified in Family 1

After screening candidate genes implicated in white spotting phenotypes, a novel single base pair insertion was identified within the ninth exon of KIT (NM_001163866.1) at chr3:79,551,897_79,551,898insA (EquCab3.0). No other novel mutations were identified in the other genes under investigation for Family 1. The novel insertion was not found in a sample of 5334 horses of diverse breeds or in the EVA Release 5 [18]. We found that six of the seven horses from Family 1 possessed the novel insertion, which we term Dominant White 37 (W37). Sanger sequencing results confirmed the presence or absence of the novel insertion (Figure 3).

The sire and one offspring had one copy each of W37, W35, and TO, and exhibited a large white patched phenotype, which is often observed in Tobiano heterozygotes (Figure 4a,b,g). The remaining four family members were found to have one copy of W35 and one copy of the W37 insertion (Figure 4c–f). Horses with the W37 variant and without TO did not demonstrate the entirely white phenotype but displayed sabino-like markings similar to other white spotting mutations involving KIT variants. This extensively depigmented abdomen is not a typical phenotype of the W35 allele. While W35 is associated with mild white patterning phenotypes ranging from no markings to sabino-like, patches of white hair across the abdomen is a less common trait of the sabino-like horses [1,10]. The sire possessed one blue eye, which is sometimes observed in horses with the TO variant. In addition, one offspring with a single copy of the Cream (CR) allele and the W37 variant possessed two partial blue eyes. This is not an unusual phenotype for a Cream heterozygote, given that the CR allele is known to dilute hair, skin, and eye color in horses [6]. The sire and three offspring with the W37 variant also possessed the EDXW3 allele, which is associated with increased areas of white spotting, including depigmented faces, lips, and legs. Horses possessing a combination of W37 with one or more EDXW3 alleles displayed increased white spotting compared to those with only the W37 allele.

3.2. Novel KIT Splice Site Identified in Family 2

A novel splice site variant in a European Warmblood mare was identified at the +5 base pair of the exon 15 splice donor site of KIT (NM_001163866.1, chr3:79,545,867C>A, EquCab3.0). We termed this variant Dominant White 38 (W38) and confirmed its presence with Sanger sequencing (Figure 5). This splice site mutation was not found in a sample of over 5334 horses of various breeds or previously reported in the EVA Release 5 [18]. No other novel mutations were identified in the exonic regions of the other genes examined in this study, nor did any of the horses show the Grey allele.

The proband, one of her offspring, and her half-sibling displayed an entirely white phenotype (Figure 6a,c,f). Both the W38 proband and one of her offspring possess blue eyes, a less common phenotype of W and EDXW alleles (Figure 6b,d). The mare’s dam carried one copy of W38 and one copy of EDXW3, displaying a sabino-like phenotype (Figure 6e). One of the horse’s offspring did not demonstrate the all-white phenotype and did not possess the splice site variant (Figure 6g). Two of the all-white horses in Family 2 possess other KIT variants out-of-phase with the novel splice site mutation. One of these horses had one copy of W20, which was previously reported to increase white spotting in horses with other KIT variants [37]. Another all-white horse carried one copy of W38, one copy of EDXW3, and both W34 and W35, likely also demonstrating a similar phenomenon in which the presence of multiple KIT variants enhances the total white spotting phenotype. No variants previously associated with white markings were found in the third all-white horse, so it is currently unclear whether another locus is responsible for the phenotypic differences between this horse and another horse in Family 2 whose phenotype is limited to sabino-like markings (Figure 6a,e). For the W38 variant, this phenotypic diversity could also be a result of variable KIT splicing during development. KIT mutations such as W10 are reported to cause variable degrees of white spotting, even when horses share similar genotypes [1].

3.3. Novel Stop-Gain Variant in a Stock-Type Mare

We identified a novel stop-gain mutation in exon 3 of KIT, termed Dominant White 39 (W39), that exchanges an arginine residue at position 181 to a stop codon (NM_001163866.1, chr3:79,579,796G>A, EquCab3.0). The mutation was found in a single stock-type mare, and its presence was confirmed by Sanger sequencing (Figure 7).

This three-year-old mare displayed a depigmented body and face. Red pigmentation was retained in the mare’s mane, tail, and small patches on the body. Its four legs were black from the fetlock to the knee, with slight depigmentation in all four pasterns (Figure 8). No overt health defects were reported for this individual. We were unable to source photographs or DNA samples for horses related to this mare, although the owner reported that the dam displayed the same phenotype, so it is unknown how many horses carry this variant. There were no other occurrences of this mutation in a sample group of 5334 other horses of various breeds, nor has it been reported in EVA Release 5. It is likely that the stop-gain mutation is only found within this stock-type horse and her immediate relatives. No other novel mutations or white-causing mutations were found in the exonic regions screened in this study.

3.4. Predicted Functional Impacts of Novel Variants

We estimated the functional impact of W37 on the KIT protein using the ExPASy Translate Tool, SMART, and I-TASSER [33,34,35]. The insertion causes a frameshift mutation leading to premature termination of translation. The resulting W37 protein contains only 504 amino acids compared to the chain of 972 amino acids found in the wild-type protein [17,38]. The SMART-generated prediction of the protein domain architecture demonstrates that the W37 protein lacks the active kinase domain found in wild-type KIT (Figure 9a,b), suggesting that protein function is severely impacted. Since the active kinase domain is downstream of the stop-gain mutation, it would not be translated [39]. The resulting protein would likely lose its selectivity for tyrosine and its enzymatic activity, rendering it incapable of performing signal transduction [11,12,13]. Furthermore, superimposed protein structure predicted by I-TASSER software reveals that the W37 KIT protein is significantly more linear in structure compared to wild-type KIT. Possessing no alpha helices and fewer pleated beta sheets, the variant represses typical protein folding (Figure 10a,b). The W37 mutation is therefore likely causative due to its extreme predicted impact on protein function. This is similar to the predicted functional impact observed with the Dominant White (W) variants W1, W3, and W31. Like W37, these variants result from the introduction of a premature stop codon to the translated region of the KIT protein and are predicted to critically impact KIT receptor function [1].

The functional impact of the W38 splice site mutation was predicted using the ExPASy Translate Tool, GENSCAN, SMART, and I-TASSER. Mutations at the +5 donor splice site may result in one or more skipped exons [40,41]. The predicted change in KIT structure due to W38 could also permit similar instances of exon skipping [36]. The mutant mature mRNA transcript would be 248 amino acids short of the normal 972 residues. The altered protein would feature just 724 residues with only the first 709 aligning to wild-type KIT [33,36,42]. The SMART server predicted that the splice variant would convert the active tyrosine kinase domain to a general kinase [34] (Figure 9a,c). This would decrease specificity for tyrosine and likely disrupt pathways that upregulate pigmentation [11,12,13]. Furthermore, the predicted catalytic domain shows an alteration in the primary, secondary, and tertiary structure relative to the wild-type, suggesting that protein function is impacted [35] (Figure 10a,c). Similarly to W7, W8, W11, and W13, this variant would likely inhibit translation of critical domains, leading to impaired melanocyte development and an increase in white spotting [1].

Of all three novel variants, the most severe predicted impact on protein structure is observed in the case of W39. Protein architecture predicted by SMART reveals that the stop-gain variant truncates the encoding protein after a single immunoglobulin (IG) domain, removing the entire active kinase domain (Figure 9a,d). In addition, I-TASSER software demonstrates that W39 KIT is the most linear of all the superimposed protein structures (Figure 10d). The lack of structural complexity observed in the I-TASSER server prediction suggests the ability of the altered protein to perform its biological function is significantly impaired [35]. This mutation is most similar to W12, a small deletion of five bases in exon 3, resulting in the truncation of the encoding protein. The W12 proband exhibited approximately 50% depigmentation of the body, four white stockings, and a white blaze covering the muzzle [43]. By contrast, the W39 proband displays nearly full depigmentation of body and face. Some red and black pigment is retained in the mane, tail, legs, and small patches on the body.

A summary of the genotypes, phenotypes, and predicted effects for novel variants for all individuals examined in this study is provided in

Table 1. Individual genotypes, associated phenotypes, and predicted functional impact for novel variants W37, W38, and W39.

Novel Variant	Relationship of Individual to Proband	Genotype	Associated Phenotype	Predicted Functional Impact
W37	Proband (Sire) (Figure 4a,b)	A/a E/e TO/n EDXW3/n W35/n W37/n	Near all-white with small patches of black pigment on face; one blue eye	Truncation of KIT protein prior to active kinase domain inhibits signal transduction processes Typical protein folding is repressed; W37 KIT is more linear in structure than wild-type KIT, featuring no alpha helices and fewer pleated beta sheets
	Offspring 1 (Figure 4c)	A/A e/e CR/n EDXW3/EDXW3 W35/n W37/n	All-white; pink skin
	Offspring 2 (Figure 4d,e)	A/A e/e CR/n EDXW3/n W35/n W37/n	All-white; pink skin; two partial blue eyes
	Offspring 3 (Figure 4f)	A/A E/e W35/n W37/n	Bay with partial bald face and sabino-like depigmentation observed on belly and hind legs
	Offspring 4 (Figure 4g)	A/a e/e W35/n W37/n	Full bald face with extensive sabino-like depigmentation observed on body and legs with red pigment on upper body, mane, and tail
	Offspring 5 (Figure 4h)	A/A e/e TO/n EDXW3/EDXW3 W35/n W37/n	Near all-white with small patches of red pigmentation on body and face
	Offspring 6 (Figure 4i)	A/a E/E TO/TO	Bay with large white patches on body, typical of Tobiano
W38	Proband (Figure 6a,b)	A/A E/e W38/n	All-white; pink skin with small pigmented spots on skin of body and face; at least one blue eye	Truncated KIT protein with conversion of active tyrosine kinase domain to a general kinase; decreased specificity for tyrosine disrupts pathways upregulating pigmentation W38 KIT lacks structural complexity of wild-type protein, with fewer alpha helices and only a number of pleated beta sheets
	Offspring 1 (Figure 6c,d)	A/A E/e W20/n W38/n	All-white; pink skin; at least one blue eye
	Dam (Figure 6e)	A/a E/e EDXW3/n W38/n	Sabino-like with patches of white on body, high white socks, and wide blaze
	Half-sibling (Figure 6f)	A/A e/e EDXW3/n W34/n W35/n W38/n	All-white; pink skin
	Offspring 2 (Figure 6g)	A/A E/e CR/n	Diluted coat; minimal white marking on muzzle
W39	Proband (Figure 8)	A/a e/e W39/n	Roan-like with red hair mixed in with the white hair on body, mane, and tail; Full bald face with small spots of red pigment; black pigment below knees and hocks	KIT protein is truncated after just one IG domain and active domain is lost, impairing biological function W39 KIT is the most linear of all superimposed proteins, with no alpha helices and few pleated beta sheets

. While the populations of horses bearing these phenotypes is small, small sample size is a common attribute of studies that investigate spontaneous and rare alleles in diverse species, including horses and humans. Spontaneous equine W alleles are frequently observed in only the founding individual and its descendants [1]. Furthermore, in a database of 5334 other horses spanning diverse breeds, there have been no additional occurrences of these three novel variants. To analyze the impact of the polymorphisms on KIT protein function, this study deferred to guidelines for the interpretation of sequence variants set by the American College of Medical Genetics and Genomics (ACMG). In accordance with ACMG recommendations, multiple in silico predictive programs were utilized as evidence supporting disrupted gene function [44].

Future work could expand the database population to test for the presence of these polymorphisms in a broader and more global sample. Due to the small number of births within these families, further investigation is required to conclude embryonic lethality for each of the three variants. Additional investigation of the combined effects and interactions between the novel alleles and previously identified depigmentation variants is warranted.

4. Conclusions

Three novel equine KIT variants were identified and based on the predicted functional impacts, each is a strong candidate to explain the observed white spotting phenotypes. A novel insertion, designated W37, was identified in Family 1. Horses carrying this frameshift mutation exhibited varying degrees of white spotting. Every horse with W37 had one copy of W35, suggesting the alleles are likely in-phase. A novel splice site mutation, termed W38, was discovered in Family 2 horses showing a completely white or sabino-like phenotype. The only offspring without this mutation did not display a sabino-like or all-white phenotype. A stop-gain mutation in exon 3 of KIT termed W39 was discovered in a single mare who demonstrated a nearly all-white phenotype. The W37, W38, and W39 alleles all result in truncated transcripts predicted to alter or remove the kinase domain of KIT. Therefore, all novel variants are likely to cause the white spotting patterns observed in the three horse groups under investigation.

Although homozygosity was not observed for any mutation, too few births have occurred to conclude embryonic lethality for any of the three novel alleles. Nevertheless, nonviable homozygosity may be likely due to observed consequences of other similar KIT polymorphisms. These variants were not observed in other breeds and were not found in other horses in our sample set of 5334 horses. Future undertakings would expand the number of tested individuals across diverse horse breeds to a global dataset enabling a more comprehensive analysis of the association of the three variants to white patterning phenotypes.