KIT gene mutation causes deafness and hypopigmentation in Bama miniature pigs

Waardenburg syndrome (WS) is a common syndromic hearing loss disease. A large group of patients affected by WS were found no mutations in the existed gene panel, indicating that there are still potential genes responsible for WS yet to be detected. In our previous study, we established an autosomal-dominant KIT (OMIM# 164920) mutation (c.2418T>A, p.Asp806Glu) pig pedigree which presented congenital bilateral severe sensorineural hearing loss and hypopigmentation, exact the same as human WS. Histological analysis showed nearly normal structures of the organ of Corti, stria vascularis (SV) and spiral neuron ganglions at E85. Scanning electron microscopy (SEM) exhibited that hair cells started to degenerate at E100, and totally gone at P1. Transmission electron microscope (TEM) showed disorganization of SV and disappearance of intermediate cells. The absence of endocochlear potentials also demonstrated the dysfunction of stria. Our study demonstrated that KIT mutation (c.2418T>A, p.Asp806Glu) interrupted the development of melanocytes in cochlea, which led to SV malformation and dysfunction, resulting in degeneration of hair cells and finally hearing loss. Therefore, KIT was highly supposed to be a newly found gene associated with WS and be added to the WS related gene screening panel clinically.


Introduction
More than one hundred genes have been identified to cause hereditary hearing loss to date [1]. Waardenburg syndrome (WS) is one of the most common types of syndromic hearing loss, characterized by the combination of sensorineural hearing loss and hypopigmentation of various organs, like skin and hair, with or without heterochromia iridis [2]. WS accounts for 2-5% of congenital deafness with the prevalence of 1/42,000 [2] and is classified to 4 types (WS1, OMIM# 193500; WS2, OMIM# 193510; WS3, OMIM# 148820; WS4, OMIM# 277580) according to clinical presentation [3]. It is verified as an autosomal dominantly inherited disease caused by disorders of neural crest cells. PAX3 (OMIM# 606597), MITF (OMIM# 156845), SNAI2 (OMIM# 602150), SOX10 (OMIM# 602229), EDNRB (OMIM# 131244) and EDN3 (OMIM# 131242) are the several reported genes related to WS and have been added to the screening panel in recent years [3][4][5]. However, the etiology of the relative large group of WS patients remains unknown, indicating that there are still potential genes responsible for WS yet to be detected [4,5].
It is believed that all of the WS-associated genes acted in the common pathways that regulating the developmental process of neural crest-derived melanocytes by influencing MITF functions [5]. MITF is suggested to mutually interact with KITLG-KIT signaling in melanocyte development. It has been proved that activation of KIT by binding of its ligand KITLG (OMIM #184745) on the cell surface, leads to the activation of downstream MITF, and this signaling cascades are primarily involved in the development and migration of neural crest-derived melanocytes, including intermediate cells of SV [6][7][8][9][10]. In rodents, loss-of-function mutations of MITF led to the absence of melanocytes which gave rise to hypopigmentation and hearing loss, similar to the phenotypes in cases with KIT and KITLG mutations [6,[10][11][12]. More importantly, Zazo Seco (2015) and Ogawa (2017) reported that mutations at KITLG could cause asymmetric and unilateral hearing loss and WS2 in humans, adding a new candidate to WS gene screening panel [13,14].
In addition to WS, there are other syndromes manifesting hearing loss and hypopigmentation due to melanocyte defects, like Piebaldism (OMIM# 172800). It is an autosomal dominant disorder characterized by a congenital white forelock, scattered normal pigmented and hypopigmented patches on the skin, which is caused by loss-of-function mutations in KIT, encoding the receptor of KITLG [7,13,14]. Sporadic cases of piebaldism with KIT mutations presented with congenital hearing loss, indicating that mutations at KIT and KITLG might be part of the same disorder [17,18].
KIT locates on 4q11-12 in humans, chromosome 5 in the mice, and 8p12 in pigs [19,20]. It encodes a 145-kD glycosylated transmembrane protein consisting of an extracellular domain, a transmembrane region, and a tyrosine kinase domain. The extracellular domain contains 5 Ig-like domain which functions in KITLG lignad binding, and receptor dimerization. Activation of KIT by binding of KITLG causes autophosphorylation at tyrosine residues and triggers the downstream signaling cascades [7,9,21,22]. Furthermore, several KIT mutations at tyrosine kinase domain including W(v), W(41), W(s), KIT(wads) have been identified in deaf mice and rats with decreased number of intermediate cells and malfunction of SV [23][24][25][26][27][28]. Thus, KIT is highly suspected as a potential gene underlies WS.
Animal model provides a useful tool to study the pathogenic mechanisms of deafness. Rodents are the most commonly used laboratory animals in otologic researches. However, huge differences in inner ear development, morphology and electrophysiology between human and mouse limit the application of mice as an ideal deafness model. According to our previous studies, Bama miniature pig, a natural Chinese miniature pig breed, is a more ideal mammalian model for otologic research [29][30][31]. Bama miniature pigs are characterized by their genetic stability and white coat color with large black patches around the head and buttocks. In addition to the genome, the size, composition, morphology and the function of auditory system highly assembles that of human. Pigs have a larger cochlea than rodents and are ideal for evaluating the fine anatomic changes of cochlear and audiological studies [29][30][31].
In our previous work, we performed random mutagenesis of the porcine genome by ENU treatment and successfully established a novel dominant mutant line with piebaldism phenotype. In this study, we identified c.2418T>A missense mutation in exon 17 of the KIT gene, resulting in amino acids substitution (p.Asp806Glu) in this model and named it as KIT D806E/+ Porcine model. Auditory tests, morphological analysis from E85 to P1 and Endocochlear Potential (EP) recordings were performed aiming at exploring the role of KIT functioning in auditory system. Our study firstly described the pathological changes in cochlea in the KIT D806E/+ porcine model and help us better understanding the mechanism of KIT associated hearing loss in human beings.

Animals
Bama miniature pigs (Department of Laboratory Animal Science, College of Basic Medicine, Army Military Medical University, Chongqing, P. R. China) were bred under a specific pathogen-free environment. The experimental procedures and animal welfare were performed in accordance with the care and use of laboratory animals and the related ethical regulations of Army Military Medical University.

Ethics approval
All animal experiments and procedures were performed under the guidelines for the Care and Use of Laboratory Animals established by the Chongqing Association for Laboratory Animal Science and approved by the Animal Ethics Committee of Army Medical University, China. Am J Transl Res 2020;12(9):5095-5107 Phenotype record, auditory brainstem response test and endocochlear potential recording Phenotype of the F2 and F3 offspring, including the coat color, iris color and sex, were recorded. Auditory brainstem responses (ABRs) (Tucker-Davis-Technology RZ6) were evaluated as previously described [30]. Animals were anesthetized (3% pentobarbital sodium, 1 ml/kg, i.m.) and placed in a sound-proof chamber. The reference electrode was fixed in the lobulus auriculae of both ears, the recording electrode subcutaneously in the cranial vertex, and the grounding electrode in the nasal tip. Stimuli intensity descended from 90 to 30 dB SPL in a step of 10 dB with sweep number of 1,024 times. When approaching the threshold, a step of 5 dB was used to define the accurate threshold. Details of endocochlear potential (EP) measurement were consistant with our previous study [30].

Genome wide linkage analysis and familybased genome wide association analysis
The mutant founder was mated with 3 wild-type females to generate 29 G2 offspring. DNA samples from all 33 Bama miniature pigs (parents and offspring) were genotyped. Genotyping was performed using porcine SNP60 Bead-Chips (Illumina, Inc., USA), which contain 62,163 SNP markers. The raw data were filtered for those SNPs with a call rate >90% and a minor allele frequency ≥0.05. Genome wide linkage analysis was performed using the Merlin tool, and family-based genome wide association analysis was performed using PLINK software packages.

Mutation identification
Mutations in the candidate gene were detected by PCR and Sanger DNA sequencing of the exons and the exon-intron boundaries. DNA samples were prepared from 4 pigs, including 2 mutants and 2 normal individuals from the mutant lineage. Primers were designed to amplify 20 exons and the exon-intron flanking regions of gene KIT (Table S1). The primers were designed using Primer Premier 5.0 software. The 50 μl PCR reactions contained 1 μl genomic DNA, 11.5 μl Mix (Vazyme, China), 35.5 μl ddH 2 O and 1 μl each primer. The reactions were incubated at 94°C for 3 min; fol-lowed by 35 cycles of 94°C for 30 sec, 56°C for 35 sec and 72°C for 50 sec; and a final extension at 72°C for 8 min. The PCR products were visualized by agarose gel electrophoresis and purified by DNA gel extraction A according to the manufacturer's instructions. The PCR products were sequenced on an ABI PRISM 3730 genetic analyzer, and the sequences were compared using the SeqMan software.

Celloidin embedding-hematoxylin-eosin Stain, scanning electron microscopy and transmission electron microscopy
The celloidin embedding-hematoxylin-eosin (CE-HE) stained cochlea sections, scanning electron microscopy (SEM) and transmission electron microscopy (TEM) samples were made as our previous study [30,31]. The Bama miniature pigs were sacrificed in accordance with the Care and Use of laboratory animals, and the cochlea were dissected from the temporal bones and post-fixed in 4% paraformaldehyde or 2.5% glutaraldehyde at room temperature overnight. Cochlea were washed in 1% phosphate-buffered saline (PBS) and decalcified in 10% EDTA for approximately one month.
For the histological analysis, the cochlea were dehydrated using graded ethanol and dipped in graded celloidin. The cochlea were sectioned (15 μm) using a freezing microtome (Leica CM1900) and stained with hematoxylin and eosin. The sections were visualized under a Leica DMI3000 microscope. For the SEM analysis, decalcified cochlea was post-fixed for 2 h in 1% osmium, dehydrated in graded ethanol (50%-100%), and dried in a critical point dryer (HCP-2, Hitachi) using liquid CO 2 . Fixed sections were then coated using a sputter coater and examined under a scanning electron microscope. For the TEM analysis, after dehydrated in graded ethanol, cochlear tissues were then embedded in Epon resin and sectioned on an Reichart Ultracut E ultramicrotome (Boeckeler Instruments, Tucson, NM, USA). Ultrathin sections were mounted on formvar-coated slot grids, stained with lead citrate and uranyl acetate and examined via transmission electron microscope.

HGVS variant nomenclature
The Human Genome Variation Society recommendations were used to standardize the nomenclature of KIT mutant variants. RefSeq sequence NM_001044525.1 was used for the coding region transcript. And NP_001037990.1 was used for the amino acid sequence resulting from translation of this transcript.

Statistical analysis methods
The differences were analyzed using Student's t-test for assessment of more than two groups on GraphPad Prism software. P<0.05 were considered as statistical significance.

Hypopigmentation and deafness of KIT D806E/+ pigs
The wild-type Bama miniature pigs exhbits white coat color with large black patches around the head and buttocks ( Figure 1A). The mutant pigs showed pigmentation, including systemic white hair with small dark spots on the head and buttocks but normal irises ( Figure  1B). For auditory tests, Click and 1-32 kHz tonepip auditory brainstem response (ABR) were conducted both in the wild-type and mutant pigs. The wild-type pigs had an ABR waveforms similar to humans and its hearing thresholds were at 20 dB SPL in Click, and 20-60 dB SPL in 1-32 kHz tone-pips ( Figure 1C, 1E). No identifiable waveforms could not be evoked in the mutant pigs at 90 dB SPL in all frequencies, demonstrating profound sensorineural hearing loss ( Figure 1D, 1E). Moreover, the hypopigmentation feature was co-segregated with hearing loss.

Localization of the mutation to the KIT gene on chromosome 8
A gene mapping analysis was performed on 16 mutants and 17 controls using illumina 60K chips. In total, 61,565 single nucleotide polymorphisms (SNPs) had a call rate >95% and were annotated genomic positions in the Sscrofa10.2 assembly of the Sus scrofa genome. Parametric linkage analysis was performed, and the results indicated that the most significantly associated SNPs localized to a 31 Mb to 48 Mb region of chromosome 8 and the KIT gene (chr8: 43550231-43601377), which is associated with coat color in pigs, is also located on this segment (Figure 2A, 2B). Therefore, KIT was selected as a candidate gene responsible for the deafness and hypopigmentation observed in the mutant lineage.
In order to screen the causative mutation in the KIT gene, all 20 exons and the exon-intron boundaries of KIT were PCR amplified and sequenced. PCR primers were listed in Table  S1. A point mutation (T>A) was identified at CDS 2418 bp site on exon 17 which are co-segregated with the mutant phenotype, resulting in a conserved substitution of aspartate by glutamate at amino acid 806 of the protein (p. D806E) ( Figure 2C, 2D). The mutation corre-sponds to position 2517 in exon 17 of the human KIT gene and is located within the tyrosine kinase domain where numbers of mutation were reported with piebaldism. These results strongly suggest that the KIT p.D806E mutation is linked to the deafness and hypopigmentation observed in the dominant inheritance lineage.

Cochlear morphology in KIT D806E/+ pigs
Firstly, we compared the CE-HE images of cochlea of KIT D806E/+ and KIT +/+ pigs at P1. Figure  3A-D showed the fine cochlear structure of KIT +/+ pigs, including the organ of Corti, Reissner's membrane (RM), basilar membrane (BM), spiral ganglion neurons (SGNs) and SV. Similar to humans, the cochlea consists of three scala, including the scala vestibule (SVE), the scala media (SM) and the scala tympani (ST). The organ of Corti was lying on BM, and tectorial membrane (TM) was contacting with inner hair cells (IHCs) and outer hair cells (OHCs). However, the stereocilia of hair cells could not be observed clearly in CE-HE sections. No identifiable disformalties of cochlear fine structures were oberseved in KIT D806E/+ pigs, no thinner SVs nor underdeveloped hair cells ( Figure 3E-H).

Abnormal hair cell morphology in KIT D806E/+ pigs
In order to inspect the pathological changes of hair cells through cochlear maturation, we seriesly observed the dynamic hair cells changes at E85, E100 and P1 via SEM. The structures of porcine cochlea were similar to human beings as one row of inner hair cells (IHCs) located in the inner most row toward the central region of the cochlea and three rows of outer hair cells (OHCs) located toward the peripheral region. IHCs and OHCs were separated by the head plate of inner pillar cells. The hair bundles of IHCs are arranged in a line. While the hair bundles of OHCs are positioned in a V-shaped pattern. The structures of hair cells at E85 were well developed ( Figure 4A-C) and stayed the same through E100 ( Figure  5A-C) and P1 ( Figure 6A-C) in wild types. For the mutants, no significant abnormality was observed at E85 ( Figure 4D-F). The stereocilia bundles began to degenerated in the basal and middle turns at E100 (Figure 5D-F). Then, pathological changes aggravated from E100 to P1, as stereocilia bundles could not be observed on the cuticular plates of all turns by P1 ( Figure  6D-F).

Disorganized structures and impaired function of stria vascularis in KIT D806E/+ pigs
To investigate the pathological changes of SV in KIT D806E/+ pigs, we examined the morphology of the SV by TEM and its function by recording EP. The TEM images revealed ultrastructural differences between the KIT +/+ and KIT D806E/+ cochlea. For the wild-types, the SV consisted of three layers: the marginal cell (MC) layer, the intermediate cell (IC) layer and the basal cell (BC) layer. MCs were in the side of the scala media and BC of the spiral ligament. The MCs connected with each other via tight junctions and formed a barrier. Elongated processes of MCs interacted with the ICs and capillaries ( Figure 7A-C). In KIT D806E/+ , ICs were visibly reduced in mutants and three layers of SV were disorganized ( Figure 7D-F). EP of the middle turn of wild types was around 60 mV ( Figure 8A) and was reduced to nearly 0-2 mV in KIT D806E/+ pigs, which were statistically different (P<0.0001) ( Figure 8B).

Discussion
In this study, we described the phenotype and the morphological changes from embryonic to postnatal stages found in the porcine KIT D806E/+ model. The CE-HE images of KIT mutant cochlea was roughly normal, no disrupted hair cells nor disorganized SVs was found. SEM revealed that OHCs and IHCs began to degenerate since E100, and stereocilia bundles were totally disappeared at P1. Further TEM studies revealed reduced ICs and disorganized SV structure. EP at middle turn of KIT D806E/+ cochlea was 0-2 mV compared to 60 mV in wild types. Our results demonstrated that the hearing loss in this KIT D806E/+ porcine model was primarily caused by the malformation and dysfunction of the SVs and subsequent progressive degeneration of sensory epithelium through inner ear maturation. Our findings of this KIT D806E/+ miniature pigs are similar to our previous reports about MITF-M mutant Rongchang porcine model [32,33].
It is known that hearing loss in WS is caused by disorders of melanocytes in the middle layer of the SV called intermediate cells, like abnormal proliferation, survival, migration, or differentiation of neural crest cells [4,5,34]. Various studies have proved that the histopathological changes of WS were primarily reduced number of intermediate cells in the SV both in patients and animal models, then subsequently degeneration of organ of Corti as its secondary alterations [4,34]. In the KIT D806E/+ porcine cochlea, the reduction but not absence of intermediate cells in the SV was revealed and decreased EP was recorded. As the progressive apoptosis of hair bundles began at E100, the Reissner's membrane and cochlear duct did not shown obvious changes after birth. Therefore, the pathological changes of KIT D806E/+ pigs are not so severe as MITF-M mutant pigs [32,33].
Mice are mostly used as hereditary deafness model to date. However, the biological differences between mice and humans limit the use of these models in further therapy studies. Bama miniature pigs could be used as a useful model bridge the gap as the high similarity of the genetic and physiological characteristics and their inner ear structural with humans [29][30][31]. In mice and rats, mutations at KIT, also known as W locus, resulted in depigmentation of the skin and hair as well as deafness. Many mutant strains including W(s), W(v), KIT(wads) mice and W(s) rats had been demonstrated to interrupt development and migration of neural crest-derived melanocytes, resulting in decreased number of intermediate cells and malformation of SV which failed to generate normal EP and subsequently caused hearing loss [23][24][25][26][27][28]. These pathological changes are in accordance with KIT D806E/+ pigs, indicating they might be part of the same disorder. However, unlike other KIT mutant rodents that cochleasaccular retrogressions were mainly observed at postnatal period, our porcine models and humans shared a similar process that cochlear disorders began at embryonic stages [17,18,[23][24][25][26][27][28]. Likewise, studies on KIT mutation cats and canines showed similar phenotypes as human, but the pathological changes had not been revealed [35][36][37].
As for the condition in human beings, KIT is a causitive gene of piebaldism, an autosomal  dominant disorder that interupts neural crestderived melanocyte development and migration resulting in congenital white forelock and hypopigmented patches of skin where melanocytes are absent [15,16,38,39]. In sporadic cases, piebaldism has been found to be associated with congenital hearing loss, the same manifestations as WS [17,18,40]. Hulten et al. (1987) described a Pakistani case of presumed homozygous piebaldism. The proband manifested the congenital absence of pigmentation in the eyes, skin and hair, deafness, hypotonia and delayed developmental milestones with his parents both being affected by piebaldism. However the disease locus had not been studied [40]. Kilsby  (2019) reported a family with expanded syndrome of piebaldism with p.Ala608Asp induced unilateral deafness in only one patient. The above reported cases with KIT mutations manifested hypopigmentation associated with congenital deafness, overlapping the features of WS, and were analogous to our KIT D806E/+ porcine model (align to p.Asp810Glu in human) and all above mutations located in tyrosine kinase domain coding region of KIT (Figure 9).
KIT plays a role, partly overlapping MITF, in cell survival, proliferation, differentiation and not unexpected piebaldism occasionally associated with deafness, since intermediate cells of SV are neural crest-derived. The similar mechanism was confirmed in MITF associated WS of both human and animal models [6, 8-11, 33, 46].
Previous reports indicate that ion channels in the SV facilitate K + recycling in the cochlea, and help maintain a high level of K + in the endolymph, which is the fundation of sensory hair cell mechano-transduction. K + circulation in cochlea consists of following processes: K + flowing from the endolymph into the perilymph via the sensory hair cells, K + being uptaken from the perilymph by fibrocytes of the spiral ligament, K + funneling into the basal and intermediate cells of the SV via gap junctions, K + outflowing from the intermediate cells to the intrastrial fluid, and K + being secreted into the endolymph by the marginal cells of the SV. We speculated that the KIT D806E/+ mutation disrupted the balance of K + homeostasis because of the lack of intermediate cells in the SV. As a result, EP was nearly completely absent because of the lack of K + in the scala media. Moreover, hair cell transduction was inhibited which further impacted the depolarization of hair cells and release of neurotransmitters, ultimately resulting in deafness [47,48].
Furthermore, the absence of intermediate cells of SV would lead to accumulated K + in intrastrial fluid, resulting in disorders of strial capillaries. This would cause increase of endothelial permeability, and subsequently result in series of changes in ischemia and hypoxia of inner ear [27] which led to the degeneration of welldeveloped hair cells as we found in E100 KIT D806E/+ cohclea. migration of neural crest cells, which are the precursors of melanocytes, hematopoietic stem cells, germ cells, mast cells and the interstitial cells of Cajal [9,21,[41][42][43][44]. In humans, gain-of-function mutations in the KIT gene have been associated with many neoplastic diseases. While loss-of-function KIT mutations are associated with piebaldism, due to abnormal development of neural crest-derived melanocytes [11,23,24]. It is However, the molecular mechanisms by which KIT mutations causing hereditary deafness remain unknown. KITLG was a newly identified gene associated with WS [13,14]. Reports have shown that the binding of KIT with its ligand KITLG would activate downstream MITF via the Ras-Erk signaling pathway, the PI3 kinase survival pathway and miRNAs [8,24,[48][49][50]. Other studies also showed that MITF was necessary for KIT expression in melanoblasts [7,23]. It was reported that nearly all WS-associated genes would influence MITF expression and function by viarous pathways [5,6,10,12]. As a key role of KITLG/KIT-MITF signaling in neural crest-derived melanocyte development, KIT could be a candidate for WS [6,9,21,45]. Therefore, we speculated that the KIT D806E/+ mutation in Bama miniature pigs would affect MITF expression, which led to pigmentation defects and deafness. Additional analysis of the MITF gene and KITLG/KIT-MITF signaling pahtway is underway to better understand its role in the KIT D806E/+ deafness model.