A novel nonsense variant in SLC24A4 causing a rare form of amelogenesis imperfecta in a Pakistani family

Background Amelogenesis imperfecta (AI) is a highly heterogeneous group of hereditary developmental abnormalities which mainly affects the dental enamel during tooth development in terms of its thickness, structure, and composition. It appears both in syndromic as well as non-syndromic forms. In the affected individuals, the enamel is usually thin, soft, rough, brittle, pitted, chipped, and abraded, having reduced functional ability and aesthetics. It leads to severe complications in the patient, like early tooth loss, severe discomfort, pain, dental caries, chewing difficulties, and discoloration of teeth from yellow to yellowish-brown or creamy type. The study aimed to identify the disease-causing variant in a consanguineous family. Methods We recruited a consanguineous Pashtun family of Pakistani origin. Exome sequencing analysis was followed by Sanger sequencing to identify the pathogenic variant in this family. Results Clinical analysis revealed hypomaturation AI having generalized yellow-brown or creamy type of discoloration in affected members. We identified a novel nonsense sequence variant c.1192C > T (p.Gln398*) in exon-12 of SLC24A4 by using exome sequencing. Later, its co-segregation within the family was confirmed by Sanger sequencing. The human gene mutation database (HGMD, 2019) has a record of five pathogenic variants in SLC24A4, causing AI phenotype. Conclusion This nonsense sequence variant c.1192C > T (p.Gln398*) is the sixth disease-causing variant in SLC24A4, which extends its mutation spectrum and confirms the role of this gene in the morphogenesis of human tooth enamel. The identified variant highlights the critical role of SLC24A4 in causing a rare AI type in humans.


Background
Mature enamel is a thin outer protective layer and covers the crown of the tooth in the form of a shell [1]. Naturally, it is tough, hard, and highly mineralized translucent human tissue produced by ameloblasts and is epithelial in its origin [2]. The biochemical architecture of dental enamel is of crystals of substituted calcium hydroxyapatite (96%), and the 4% is of organic matter and water [3]. Amelogenesis is a highly intricate biomineralizing process controlled by the expression of several genes [2]. AI affects both the primary and permanent dentition with exceptionally variable severity of the disease conditions [4,5].
Various accounts of both syndromic and nonsyndromic/isolated cases of AI have been published in the literature. Depending upon the amount, structure, and composition of the dental enamel, the phenotypes of non-syndromic AI are highly variable and may be divided into hypoplastic, hypocalcified, and hypomaturation forms [3,4].
Occasionally, AI has been reported as a part of a syndrome. The most common of them include Tricho Here, we report a novel nonsense variant c.1192C > T (p.Gln398*) in exon-12 of SLC24A4 in non-syndromic AI patients in a family of Pakistani origin.

Methods
Patients recruitment, pedigree construction, and DNA extraction The recommendations of the Declarations of Helsinki were strictly followed for the approval of the study from the Research and Ethical Committee of Kohat University of Science and Technology (KUST), Khyber Pakhtunkhwa, Pakistan. Informed written consent was obtained from the affected and unaffected participants. A five generational pedigree diagram was constructed after a thorough interview of the unaffected mother (III-4). The pedigree showed an autosomal recessive mode of inheritance (Fig. 1A). Venous blood samples were collected from seven members of the family, including two patients (IV-4, IV-5) and five phenotypically unaffected individuals (III-4, IV-1, IV-3, IV-7, V-1). The extraction of genomic DNA from whole peripheral blood was performed by using the GeneJET Genomic DNA extraction Kit (Thermo-scientific, Lithuania), strictly following the manufacturer's protocol.
Exome sequencing, and validation of rare variants through DNA sequencing DNA (70 ng/μl) of an affected member (IV-4) was subjected to exome sequencing. The enrichment of DNA for the intron-exon boundaries was carried out with the SeqCap EZ human exome library v2.0 kit. The Illumina HiSeq 4000 sequencing machine via a paired-end 100-bp protocol [31] was used to run the generated libraries. The Cologne Center for Genomics (CCG) Varbank pipeline v2.26 (https://varbank.ccg.uni-koeln.de/) was used for exome data analysis. The mean coverage of the data was 77%, while at 20X and 10X, the coverage of the targeted bases was 92.6 and 96.6%, respectively. Genome Aggregation Database (gnomAD; https://gnomad.broad institute.org/) was consulted to establish the minor allele frequency (MAF; value < 0.01) of the variants. As controls, an in-house database of 511 exomes, and a dataset of 65 exomes from the Pakistani population, including 44 exomes from Punjabi, Sindhi, and Balochi patients, and 21 exomes of ethnically matched Pakhtoon patients were also consulted. The rare variants in PSPH, CHCHD2, BNC2, and SLC24A4 were selected from the exome data and were considered for the co-segregation analysis. The online prediction tools like MutationTaster, PROVEAN, SIFT, and PolyPhen2.0 were used to predict the pathogenicity of the missense variants (

Clinical and radiological investigations
For clinical and radiological investigations, a 35-years old patient (IV-4) was referred to Khyber Medical College of Dentistry, Peshawar, Khyber Pakhtunkhwa, Pakistan. His major complaints were yellow-brown staining, eating, and chewing difficulties of all the teeth   Fig. 1B-i,ii,iii). The patient presented no complications of other body organs during the clinical evaluation. The Orthopantomogramm (OPG) of this patient showed a thin (hypoplastic) mandible with missing posterior teeth on the right side and carious molars with a periapical infection on the left side. The maxilla showed impacted canine in the right premolar region with a missing molar and spacing among the dentition on the right side of the arch. Additionally, the teeth showed generalized horizontal bone loss, more prominent around the maxillary molars. OPG also showed the presence of a thin layer of enamel, especially in the region of molars of the upper jaw. Furthermore, enamel appeared to have higher radio-density compared to the dentin. Moreover, the dentin appeared normal and distinct from the enamel (Fig. 1B-iv).
Patient IV-5, the 27-year old brother of patient IV-4, presented with creamy discoloration and attrition of the frontal maxillary teeth while dental caries in the mandibular premolars and molars ( Fig. 1B-v,vi). p.Arg133Ser appears in 20 alleles out of 282,490 alleles (none homozygous) and c.2860G > A; p.A954T is found in 4 alleles (one is homozygous) out of 246,026 alleles. These variants are not identified in the in-house database of 511 exomes and 65 exomes of Pakistani patients with diverse phenotypes other than AI. The pathogenicity predictions of the variants in PSPH, CHCHD2, and BNC2 by four online prediction algorithms are described in Table 1.

Discussion
Five functionally different types of K + -dependent Na + / Ca + 2 exchangers (NCKX1-5) have been characterized in humans [32,33]. NCKXs are bidirectional membrane transporters; for example, NCKX4 transports an intracellular Ca + 2 and a K + ion in exchange for four extracellular Na + ions [34]. Each NCKX protein has a unique role in various biochemical pathways governing the vision, olfaction, and skin pigmentation [35]. During the maturation stage of tooth development, SLC24A4 (NCKX4) is involved in the active transport of Ca + 2 ions from ameloblasts into the enamel matrix. Genetic alterations in SLC24A4 in the human genome and its knockout mice Slc24a4 −/− lead to the development of indisposed calcified enamel [36]. Clinical findings of Slc24a4 −/− mice signify the essential role of this protein in enamel development [25].
SLC24A4 (OMIM 609840) encodes a protein of 622 amino acids, called solute carrier family 24 member 4 (SLC24A4), which is one of the members of K +dependent Na + /Ca 2+ exchanger family (SLC24A), comprising a total of five members. It has been mapped to the chromosome 14q32 [33,36]. SLC24A4 has various transcripts (NM_153646, NM_153647, NM_153648) resulting from alternative splicing and the longest isoform (NM_153646) contains 17 coding exons. SLC24A4 is highly expressed in many types of tissues, such as aorta, brain, lungs, and thymus gland [34]. In the case of developing dentine, SLC24A4 is expressed in ameloblasts, and it borders to the membrane in contact with the developing enamel [37]. The predicted structure for full-length SLC24A4 protein consists of 11 transmembrane helices having two highly conserved transmembrane clusters (consisting of 5 transmembrane helices) linked together by an intracellular (cytoplasmic) loop. The Na + /Ca 2+ exchanger domains are composed of these transmembrane pockets. Each domain contains a hydrophobic and highly conserved region of 30-40 residues called alpha-1 (139-179 amino acids), and alpha-2 (495-526 amino acids) repeats, respectively, which form ion-binding regions after undergoing highly intricate interactions with each other [25,38].
We have identified a novel nonsense variant (c.1192C > T; p.Gln398*) in exon-12 of SLC24A4 by using exome sequencing. This unusual genetic alteration is expected to lead to the loss of function of SLC24A4 protein either by nonsense-mediated decay (NMD) or by the production of a truncated protein lacking the Cterminus. Since this nonsense variant introduces a premature stop codon at the position 398 in the cytoplasmic loop between the alpha-1 and alpha-2 repeats; hence the loss of remaining 225 amino acids (containing the alpha-2 repeat) is predicted. The two Na + /Ca 2+ exchanger domains (alpha-1 and alpha-2 repeats) are crucial for the smooth transport of ions, which verifies the exceptional role of SLC24A4 during amelogenesis. The absence of one of the two Na + /Ca 2+ exchanger domains, in this case, alpha-2-repeat only will ultimately render the protein nonfunctional and causes amelogenesis imperfecta, hypomaturation type AI2A5 (OMIM: 615887) phenotype [25].
To date, a total of five pathogenic variants causing AI have been identified in the SLC24A4, including three missense variants, one nonsense variant, and a gross deletion (Fig. 2B). Parry et al. in 2013 screened 15 Pakistani families and identified two homozygous variants in SLC24A4, including a missense c.1495A > T (p.Ser499Cys), and a nonsense variant c.1015C > T (p.Arg339*) in two consanguineous families. They performed Sanger sequencing of 37 AI patients of different ethnicities and suggested that pathogenic sequence variants in SLC24A4 are a rare cause of AI in general, but might be a frequent cause of AI in Pakistani population [25]. Researches on three consanguineous Turkish families have revealed two homozygous missense pathogenic variants c.437C > T; (p.Ala146Val), c.1317 T > G (p.Leu436Arg) and a 10 kb (10,042 bp) homozygous deletion, comprising of exons 15, 16 and most of the exon-17 (Chr14: 92,957,680-92,967,722del) [36,39,40]. During a comparison of AI phenotypes caused by SLC24A4 variants in patients reported so far in the

Conclusion
The present study aimed to perform a clinical and molecular evaluation of an autosomal recessive Pakistani family. We have identified the sixth disease-causing variant in SLC24A4 (Fig. 2B), which extends its mutation spectrum and confirms the role of this gene in the morphogenesis of human tooth enamel.