Intra-familial phenotype variant in hypoplastic amelogenesis imperfecta under a complex genetic component: a family report, whole-exome sequencing, and literature review

Abstract

Amelogenesis imperfecta (AI) encompasses a group of conditions characterized by abnormalities in the development or function of tooth enamel. Clinical manifestations include different forms and degrees of enamel frailty, associated with sensitivity, tooth fractures, stains, abnormal tooth morphology, missing teeth, etc. AI is genetically heterogeneous, with over 70 genes associated with autosomal dominant, autosomal recessive, X-linked, and oligogenic inheritance.

Objective  To identify genetic variants associated with AI in a single family.

Methodology  We describe the clinical findings of a family affected by AI, composed of five individuals: four affected (the father and three daughters) and one unaffected (the mother). The observed segregation pattern suggests a dominant, X-linked inheritance. Genetic variants were screened using whole-exome sequencing. The initial bioinformatic analysis was conducted using Qiagen QCI, and variants were selected based on their presence in all four affected family members and absence in the unaffected mother. Search terms included “amelogenesis imperfecta,” “tooth,” and “enamel.” Several types of software were used to classify variants according to pathogenicity.

Results  Candidate variants were identified in six genes. Three of these variants were detected in autosomal genes: NM_031889.3(ENAM):c.1726T>C (p.F576L), NM_022168.4(IFIH1):c.1764dupA, (p.A589fs*21), and NM_032383.5(HPS3):c.1897A>T (p.M633L). Three variants were detected in X-linked genes: NM_006150.5(PRICKLE3):c.8C>G (p.A3G), NM_004484.4(GPC3):c.584A>G (p.N195S), and NM_152787.5(TAB3):c.1936G>A (p.V646M). None of these variants were classified as pathogenic or likely pathogenic in AI.

Discussion  Among the identified genes, only ENAM has previously been associated with AI; however, IFIH1, PRICKLE3, and GPC3 are associated with dental/enamel development. The relatively high number of candidate genes and variants detected may reflect an oligogenic component already proposed for AI.

Conclusions  This study provides a set of new candidate genes and genetic variants for AI. Despite sharing the same variants, AI-affected family members show considerable phenotypic variant, suggesting the involvement of non-shared genetic or environmental factors.

Keywords
Amelogenesis Imperfecta; Enamel; Tooth; Gene; ENAM

Introduction

Amelogenesis imperfecta (AI) is a group of inherited defects of dental enamel formation with high genetic and clinical heterogeneity, presenting as isolated or syndromic conditions. Non-syndromic AI presents a spectrum of enamel malformations caused by mutations in more than 70 genes.¹ These mutations have been associated with disruptions in the secretion and function of enamel matrix proteins, enamel matrix proteases, cell-cell and cell-matrix adhesion proteins, and transport proteins.² Clinically, malformed enamel manifests itself in several ways: it can be abnormally thin, soft, rough, fragile, stained, or even absent. Manifestations may include tooth sensitivity, occlusion abnormalities, enamel wear and fractures, visible staining, periodontal disease resulting from poor dental hygiene, dental impactions, premature tooth loss, and aesthetic concerns.³

Given the broad spectrum of clinical presentations, non-syndromic AI is categorized based on enamel phenotype into four types (MIM:PS104500). Type I hypoplastic AI (reduced amount of enamel), Type II hypomature AI (poor enamel maturation), Type III hypocalcified AI (deficient enamel calcification), and Type IV hypoplastic-hypomature AI with taurodontism.⁴ Each type includes multiple subtypes. However, genetic heterogeneity and intra-familial variation make classification into subtypes challenging. Moreover, phenotyping of teeth in AI patients is complicated by post-eruptive changes that occur over time.²

Next-generation sequencing (NGS) in multigene panels or whole-exome sequencing (WES), combined with a wide range of bioinformatics tools, has advanced the understanding of AI.¹,⁵ In addition to the growing number of associated genes, an oligogenic inheritance pattern has been observed in several AI families.¹This study investigates a Brazilian family presenting with a highly variable clinical phenotype of the hypoplastic form of AI. Using WES, potentially associated variants were detected in six genes. This study aimed to investigate these variants using bioinformatic tools to evaluate pathogenicity mechanisms. The use of WES in dentistry is recent and this study can help understanding AI etiology.

Methodology

Ethical aspects and family recruitment

The research project was approved by the Research Ethics Committee at Universidade Federal de Minas Gerais (CAAE: 30228919.1.0000.5149). Written informed consent was obtained from all five participants (affected father, four affected daughters, and unaffected mother). Medical history, clinical evaluation, and radiological images were considered for the AI diagnosis.

Whole-exome sequencing

Peripheral venous blood samples were collected from five family members (four affected and one unaffected). Genomic DNA was isolated using the QIAamp DNA Blood Mini Kit (QIAGEN; Hilden, Rhineland, Germany). Agilent Clinical Research Exome v1 was used for exome capture and the Illumina NextSeq platform was used for sequencing. Exome FastQC data were used in the variant calling file (VCF) in a customized bioinformatic pipeline (Dragen Germline, v3.6, Illumina). Sequences were aligned to the GRCh37.75/hg19 reference genome. Additionally, data was analyzed using an in-house pipeline (details and references are available in Supplementary Material S1). All five exomes had at least 95% of targeted bases at a minimum depth of >20X. The average coverage per samples ranged from 70X to 120X. The depth of coverage for the variants selected for pathogenicity analysis was specifically controlled for PHRED≥20 (meaning that there is a chance smaller than 1 in 100 of an incorrect base call).

Variant selection and classification

Variants were annotated using wANNOVAR (https://wannovar.wglab.org/) and QIAGEN Clinical Insight (QCI) Interpret (www.qiagen.com). A list of potentially associated genes was provided by QCI using the keywords “amelogenesis imperfect” AND/OR “teeth” AND/OR “enamel.” Criteria for variant selection and interpretation were: (i) minor allele frequency (MAF) < 0.01 in global or Latin American populations; (ii) pathogenicity classification according to ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) and VarSome (https://varsome.com/); (iii) presence in databases or literature, including PubMed (https://pubmed.ncbi.nlm.nih.gov/), OMIM (https://omim.org/), MalaCards (https://www.malacards.org/), and others; (iv) variants not previously reported in databases or literature were classified as novel; (v) functional impacts of variants were predicted using CADD Score (https://cadd.gs.washington.edu/), PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/), SIFT (https://sift.bii.a-star.edu.sg/), MutationTaster (https://www.mutationtaster.org/), AlphaFold2 (https://alphafold.ebi.ac.uk/), AlphaMissense (https://alphamissense.hegelab.org/), among others; (vi) evolutionary conservation was evaluated using sequence database searches (BLAST: https://www.ncbi.nlm.nih.gov/BLAST/) and multiple sequence alignment (COBALT: https://www.ncbi.nlm.nih.gov/tools/cobalt/re_cobalt.cgi); (vii) variants were classified according to the guidelines of the American College of Medical Genetics and Genomics (ACMG).⁵ The literature review was conducted using an advanced search in PubMed, restricted to the “Title or Abstract” fields. Search strings included (“amelogenesis imperfecta” OR “enamel” OR “teeth”) AND (“ENAM” OR “IFH1” OR “HPS3” OR “PRICKLE3” OR “GPC3” OR “TAB3”), with no date restriction. Only articles that correlated genetic variation with phenotype were included.

Copy number variation (CNV) identification and interpretation

Copy number variation (CNV) were called using Illumina’s DRAGEN (Dynamic Read Analysis for GENomics) pipeline. CNVs were annotated using QIAGEN Clinical Insight (QCI) Interpret (www.qiagen.com) and wANNOVAR (www.wannovar.wglab.org). A standard CNV filtering method was used and only high-confidence CNVs were considered. Genes located within the chromosomal regions of the high-confidence CNVs were screened against the 567 genes included in the GenoDENT panel for AI diagnoses and research.¹ Clinical interpretation of CNVs followed the guidelines of the ACMG.⁶

Results

Five members were recruited from a nonconsanguineous family (Figure 1A). Clinical history from previous generations was unavailable. The proband was a 10-year-old girl (III-4) affected by generalized hypoplastic AI (Figure 1 B1-B4), characterized by quantitative enamel defects and morphological alterations of the permanent incisors, which presented thin or absent enamel. The occlusal surface of both primary and first permanent molars were filled with composites. She also exhibits a narrow dental arch, high palatal vault, and anterior open bite. The panoramic radiograph confirmed hypoplastic AI, with enamel defects visible on unerupted premolars and second permanent molars. Enamel coverage was almost entirely absent, with only a small band present on interproximal surfaces and an absence of occlusal surfaces. The proband’s father (II-12; Figure 1 C1-C4) presented severe periodontitis, multiple missing teeth, and attrition of the incisal surfaces. The incisors and canines showed flattened, smooth occlusal surfaces, leading to a reduced crown height dentin exposure. The panoramic radiograph showed a thin layer of enamel covering the mandibular right molars and the maxillary left second molar. Severe bone loss was also visible in the molars. The proband’s 14-year-old sister (III-2) presented localized clinical features, with no alterations in tooth shape. Notable findings included horizontal hypoplastic grooves and pits in the buccal side of lateral incisors, pits on the palatal surfaces of the central incisors, thinner enamel on the lower incisors, and pits and irregularities on the occlusal surfaces of premolars and molars (Figure 1 D1-D4). The panoramic radiograph showed a thin layer of enamel covering the permanent teeth, a small enamel-free area on the incisors, and near-total absence of enamel on the occlusal surfaces of unerupted molars. The proband’s 12-year-old sister (III-3) presented with enamel defects on the labial side of the maxillary permanent lateral incisors, mandibular incisors, and canines, including horizontal hypoplastic grooves and pits. She had an overbite and wide interdental spacing (Figure 1 E1-E4). The panoramic radiograph showed a thin layer of enamel on the occlusal surfaces of premolars and molars, as well as irregular hypoplastic pits on the interproximal surfaces. The incisors and canines were the most affected teeth. The proband’s 16-year-old sister (III-1), who did not participate in the genetic study, also presented clinical features of hypoplastic AI, with severe involvement of the premolars, permanent incisors, second molars, and presence of an overbite. The proband’s mother was considered unaffected (II-13).

Figure 1
A - Family Pedigree. Clinical status is unknown for individuals I-1 and I-2. Except for II-12 and II-13, clinical information for other Generation II individuals is based on information provided by II-12. III-5 is six months old, impeding clinical evaluation; B-E: Clinical features and panoramic radiographic. B1-B4: Proband (III-4); C1-B4: Father (II-12); D1-D4: Sister (III-2); E1-D4: sister (III-3).

Figure 2 presents details of enamel defects. The proband, her sisters, and her father are tall, with long arms and legs. The proband and her sisters exhibited no other systemic changes, except for recurrent upper airway infections. The father has a history of systemic arterial hypertension, and the mother reports no health issues.

Figure 2
Details of enamel defects: 2A - In the proband (III.4), severe enamel hypoplasia with almost no enamel on all permanent teeth; 2B - In the father (II.12), absence of incisal enamel and pronounced attrition; 2C - In the 14-year-old sister (III.2), irregular, rough enamel on the occlusal third of the premolar crowns; 2D - In the 12-year-old sister (III.3), localized enamel defects on the buccal surfaces, circumscribed by normal enamel.

Exome sequencing and bioinformatics analysis

Six genetic variants were shared by the father and his three affected daughters but were absent in the unaffected mother: NM_031889.3(ENAM):c.1726T>C (p.F576L), NM_022168.4(IFIH1):c.1764dupA, (p.A589fs*21), NM_032383.5(HPS3):c.1897A>T (p.M633L), NM_006150.5(PRICKLE3):c.8C>G (p.A3G), NM_004484.4(GPC3):c.584A>G (p.N195S), and NM_152787.5(TAB3):c.1936G>A (p.V646M). All affected individuals tested were heterozygotes for these six variants. Among them, only ENAM has been associated with AI. Results of the bioinformatic analysis are shown in Figure 3.

Figure 3
Variants shared by the four AI-affected individuals investigated

A PubMed (https://pubmed.ncbi.nlm.nih.gov/) search was conducted using the following descriptors: amelogenesis imperfecta AND ENAM AND [mutation OR variant]. The search was restricted to “titles and abstracts” and retrieved 44 articles (May 15, 2024). Furthermore, articles had to describe dental manifestations of affected individuals and provide a description or identifier of the ENAM variation. Ultimately, 19 articles were selected and organized in Figure 4. For the genes IFIH1, PRICKLE3, TAB3, HPS3, and GPC3, literature searches included the terms amelogenesis imperfecta, enamel, and tooth. Due to the small number of publications identified for each gene, no additional exclusion criteria were applied.

Figure 4
Phenotypes described for families with ENAM variants

ENAM

ENAM (MIM*606585) is located at 4q13.3 and encodes the protein enamelin. The variant NM_031889.3(ENAM):c.1726T>C (p.Phe576Leu), rs2609428, is a single nucleotide variant (SNV) in the ninth and last ENAM exon. This variant was submitted twice to ClinVar and was classified as benign on both occasions. Ensembl Variant Effect Predictor, PolyPhen, and SIFT also predicted the variant as benign or tolerated. However, AlphaMissense classified it as likely pathogenic (pathogenicity score: 0.663), and a CADD C-score of 18.29 (CADD v1.6, GRCh38/hg38) places the variant amongst the 10% to 1% most deleterious substitutions. The variant allele frequency is 0.009 in the general population (gnomAD), 0.128 in the African/African American population (gnomAD), and 0.034 in the Brazilian population (ABraOM). BLASTP and multiple sequence alignments (COBALT) revealed that Phe576 is conserved down to marsupials. Nonetheless, Phe576 is frequently substituted by leucine in rodents and by isoleucine in carnivores. Protein 3D-structure analysis using AlphaFold revealed that Phe576 interacts with glutamic acid (Glu578) via hydrogen bonds. Phe576 is part of a compositional bias of basic and acidic residues (UniProt), spanning amino acids 571 to 591, where nine out of 21 residues are basic or acidic. This region corresponds to the conserved enamelin domain, which comprehends amino acids 213 to 1,114. No evidence for Phe576 post-transcriptional modifications has been described (UniProt). We also investigated the impact of this variant on splicing and splicing regulation. The altered thymine is part of four exonic splicing silencers (ESS) and two exonic splicing enhancers SE). However, Human Splicing Finder (http://www.umd.be/hsf/) predicted no significant effects on splicing. The variant NM_031889.3(ENAM):c.1726T>C (p.Phe576Leu) has been submitted three times to ClinVar, including once in association with AI, where it was classified as benign.

IFIH1

IFIH1 (MIM*606951) is located at 2q24.2 and encodes the protein MDA5 (Melanoma Differentiation-Associated Protein 5). The variant NM_022168.4(IFIH1):c.1764dup (p.A589fs21), corresponding to rs553669430, involves the duplication of an adenine at position 1764 within the coding sequence of IFIH1. This region contains a poly-A track, where the reference allele is (A)₈, while alternative alleles include (A)₅, (A)₇, (A)₉, and (A)₁₀. These InDel variants cause a frameshift mutation, starting after K588. ClinVar submissions have classified this variant as “uncertain significance,” “likely benign,” and “benign.” It has been linked to Aicardi-Goutières Syndrome 7 (MIM615846) and Singleton-Merten Syndrome 1 (MIM#182250). The variant allele frequency is 0.000527 in the general population (gnomAD), with higher frequencies observed in specific populations (0.0021 in East Asians and up to 0.005 in African/African American populations).

This adenine duplication causes a frameshift mutation that leads to a stop codon after 21 codons (Figure 5). Therefore, we investigated the possibility of nonsense-mediated decay (NMD). According to the 50-55bp NMD rule, NMD is highly probable, as the stop codon is located more than 50–55 bases upstream of the last exon-exon junction, potentially leading to loss of function.⁷

Figure 5
Impacts of the IFIH1 variant, NM_022168.4(IFIH1):c.1764dupT (p.Ala589fs). The mRNA sequence (NM_022168.4 or ENST00000649979.2) was translated using ExPASy Translate. The same sequence was edited to include an additional adenine at position 1764. 3A: Partial sequence of the normal cDNA and its corresponding amino acid sequence; 3B: Effect of adenine insertion at position 1764, resulting in a frameshift mutation and a stop codon 21 codons downstream. Note: Although the variant was annotated as a thymine insertion on the (+) strand, IFIH1 is on the (-) strand; therefore, according to the mRNA sequence (NM_022168.4 or ENST00000649979.2), the variant is an adenine insertion.

HPS3

HPS3 (MIM*606118) is located at 3q24 and encodes the HPS3 protein. The variant NM_032383.5(HPS3):c.1897A>T (p.M633L) has not been reported in ClinVar. Another variant affecting Met633 (Variant ID: 2235382), leads to the substitution of methionine by valine and is classified as variant of uncertain significance (VUS) in ClinVar. The frequency of the HSP3 M633L variant in gnomAD is low (f=0.000001239). BLASTP and multiple sequence alignments (COBALT) revealed that Met633 is highly conserved in mammals; however, some rodent species have isoleucine or leucine at this position. The protein sequence surrounding leucine at position 633 in rodents is well conserved, suggesting that leucine may be functional at this site. While some in silico predictors referenced by VarSome classify this variant as benign supporting (BLOSUM, DANN, SIFT, PrimateAI) or benign strong (REVEL), clinical data regarding mutations near the beginning of exon 11 are scarce, limiting conclusive classification. No specific function has been identified in the region surrounding the 633rd residue, nor is there information on how the protein interacts with its functional complex (BLOC-2). This variant was ranked in the >1% range of pathogenicity probability by CADD (C-score: 23.8, CADD v1.6, GRCh38/hg38).

PRICKLE3

PRICKLE3 (MIM*300111) is a nine-exon gene located at Xp11.23 on the antisense strand. It encodes the PRIC3 protein and has two transcripts: NM_006150.5, the complete one, and NM_001307979.2, which originates from an alternative start site and lacks the first exon. In this study, we report a missense variant, NM_006150.5(PRICKLE3):c.8C>G (p.A3G), rs1242595797. This variant was identified in ALPHA Allele Frequency (0.0007) and gnomAD (0.00003), both reported in African American populations. It was not identified in other exome or genome projects, and has not been submitted to ClinVar. In VarSome, the variant was classified as benign, although some predictors suggest pathogenicity (PrimateAI) or VUS (SIFT, M-CAP, DANN, and BLOSUM). In our analysis, AlphaMissense classified the variant as probably benign (score 0.266), while CADD predicted it as probably damaging (C-score: 21.5, CADD v1.6, GRCh38/hg38). Evolutionary conservation was accessed using BLASTP and multiple alignments (COBALT) (2024-07-16). The NM_006150.5 isoform, containing the first exon, is highly conserved and the alanine in the third position is present in almost all Eutherian. In some species, this alanine is replaced by proline, valine, serine, or threonine, but never by glycine.

TAB3

TAB3 (MIM*300480) is located at Xp21.2 and encodes the TGF-beta-activated kinase 1 and MAP3K7-binding protein 3 (TAB3). The variant NM_152787.5(TAB3):c.1936G>A (p.V646M), rs149043136, affects an amino acid residue present in four TAB3 isoforms. The canonical transcripts are ENST00000288422.4, encoding the protein isoform Q8N5C8-1 (ENSEMBL: https://www.ensembl.org/index.html). This variant was submitted to VarSome but not to ClinVar. In the former, it was predicted as damaging (M-CAP), VUS (seven prediction software), and benign (13 prediction software). In this study, it was predicted as tolerated by SIFT and likely pathogenic by AlphaMissense. The CADD C-score for this variant is 20.8 (CADD v1.6, GRCh38/hg38). The allele frequency for this variant is 0.00005 in the general population (gnomAD), 0.0000246 in exomes, and 0.000322 in genomes. Allele frequency in the Brazilian population was not available at ABraOM.

Splicing and protein structure analyses were conducted to evaluate the functional impacts of this variant. Human Splicing Finder predicted no significant impacts on splicing signals. However, the affected nucleotide is located within three exonic splicing enhancers and one exonic splicing silencer. BLASTP and multiple sequence alignments (COBALT) revealed that Val646 is evolutionarily conserved down to fish. According to UniProt, Val646 is part of an antibody-binding sequence, which comprehends residues 639 to 712, and no evidence for Val646 post-transcriptional modifications has been described. This is the first report of a TAB3 variant co-segregating with AI in a family.

GPC3

GPC3 (MIM*300037) is located at Xq26.2 and encodes the protein Glypican-3. The variant NM_004484.4(GPC3):c.584A>G (p.N195S), variation ID: 476659, rs745317547, is situated in the third of eight GPC3 exons. This variant was submitted once to ClinVar and VarSome, where it was classified as benign. Ensembl Variant Effect Predictor, SIFT, and AlphaMissense also predicted the variant as benign. Additionally, its CADD score is 15.2 (CADD v1.6, GRCh38/hg38), its allele frequency is 0.000009 (gnomAD), but higher—0.001044—in the Brazilian population (ABraOM).

To better clarify the functional impacts of this variant, additional analyses were conducted. BLASTP and multiple sequence alignments (COBALT) unveiled that the affected residue and its surrounding region are evolutionarily conserved. Alignments with 996 GPC3 sequences showed that N195 is conserved as far back as fish. Human Splicing Finder predicted the occurrence of a new donor splice site at chrX:133,753,933, GRCh38/hg38 (HSF mutation score 79.4; reference score 52.26) and a new acceptor splice site at chrX:133,753,941 GRCh38/hg38 (HSF mutation score 77.6; reference score 49.73), with potential effects on splicing (Figure 6).

Figure 6
Effects of the GPC3 variant (NM_004484.4(GPC3):c.584A>G (p.N195S) 6A: Partial sequence of the wild-type cDNA and its corresponding amino acid sequence. 6B: Effect of the adenine-to-guanine substitution at the 195th residue, leading to serine substitution and to a new donor splice site. 6C: The new donor splice site causes a frameshift mutation. 6D: Comparison between the canonical and variant protein sequences.

CNVs

The high-confidence CNVs of our samples are listed in Supplementary Table S1A and S1B. No CNV overlap with genes present in the GenoDENT panel.¹ Two benign CNVs (Figure 7, CNV-1, and CNV-2), with no dental relevance, were detected in individuals III-4, III-3, III-2 and II-13. Only one likely pathogenic CNV (Figure 7, CNV-3) is found having heterozygous deletion harboring DMBT1 gene, present in individuals III-4, II-12, II-13. Another likely benign CNV (Figure 7, CNV-4), harboring DEFB4A gene, is found in III-4 and II-12.

Figure 7
Potential CNVs associated with dental issues in this family

Discussion

In this study, we investigated a Brazilian family of four AI-affected individuals and one unaffected member. The familial history suggests dominant X-linked inheritance, but autosomal dominant inheritance cannot be excluded. WES identified six genetic variants shared among the four affected members and absent in the unaffected mother. Among these genes, only ENAM has been consistently associated with AI (Figure 4).

ENAM

Variants in ENAM are involved in both dominant and recessive AI and cause hypoplastic AI (Type IB – localized hypoplastic, autosomal dominant; MIM#104500).⁹,¹⁰ The major secretory products are post-translationally modified, secreted, and processed by proteases. These modifications significantly influence the structure and function of enamelin, affecting its abundance and affinity for hydroxyapatite. The identification of ENAM variants in AI-kindreds provided evidence for enamelin’s critical role in proper dental enamel formation.¹¹,¹²

Different phenotypes of hypoplastic AI, characterized by thin or absent enamel, have been associated with ENAM variants. These defects may present as horizontal row of pits, linear depressions, or large hypoplastic areas, most prominently affecting buccal surfaces of teeth involving the middle third of the enamel.¹Figure 4 presents AI phenotypes related to ENAM variants identified in 44 different families, encompassing both autosomal recessive¹³-¹⁹ and dominant inheritance patterns.⁹,¹⁴,¹⁵,¹⁷,¹⁸,²⁰-²⁸ ENAM variants showed variable expressivity, ranging from non-penetrance or localized hypoplasia to generalized, severe enamel loss. The phenotype included mild enamel defects (thin, irregular, rough localized surfaces, pits in horizontal lines) to generalized hypoplastic enamel with yellow-brown discoloration, interdental spacing, and loss of vertical occlusal height.²⁹,³⁰

In our study, the missense variation in exon 9 was shared by all four affected family members, but their phenotypes were quite different. The intra-familial phenotype variation has been described in several studies involving ENAM variants, ⁹,¹³,¹⁴,¹⁹,²⁰,²²,³¹suggesting the contribution of modifier factors—both genetic (e.g., digenic or oligogenic inheritance) and environmental.¹³,¹⁸,¹⁸,²¹,³⁰,³²Phenotype variation observed in this family is similar to those described in other families carrying ENAM variants (Figure 4).

The ENAM variant observed in this family is classified as benign in both ClinVar and VarSome. This classification is primarily based on its relatively high allele frequency in the general population (0.003), as well as in some specific populations—such as the Brazilian population (ABraOM, 0.034) and the African/African American population (0.128). Although Phe576 is highly conserved amongst primates, its frequent substitution by leucine in some placentals (e.g., rodents) may suggest a reduced functional impact on enamelin. However, this assumption is contradicted: AlphaMissense predicts that only four Phe576 substitutions (isoleucine, serine, valine, and tyrosine) are benign. In contrast, three other substitutions (lysine, proline, and leucine), including the one observed in the family, are classified as likely pathogenic. Considering the hypothesis of oligogenic inheritance—which could involve relatively common variants—and the overlapping between AI and common conditions, we searched the literature for this variant in association with dental caries. Notably, this association has been investigated but yielded no significant results.⁷

IFIH1

The IFIH1 gene encodes MDA5, a protein that recognizes viral dsRNA and initiates innate immune responses. MDA5 detects viral dsRNA based on its structure and length, triggering an interferon-mediated proinflammatory response. Two conditions have been associated with IFIH1 variants: Singleton-Merten Syndrome 1 (SGMRT1, MIM#182250) and Aicardi-Goutières Syndrome 7 (AGS7, MIM#615846).

In SGMRT1, dental findings include delayed primary tooth exfoliation and permanent tooth eruption, truncated tooth root formation, early-onset periodontal disease, and severe root and alveolar bone resorption associated with dysregulated mineralization, leading to tooth loss.³³ Conversely, AGS7 is an autosomal dominant inflammatory disorder characterized by severe neurological impairment, with no specific teeth abnormalities. Most patients present in infancy with delayed psychomotor development, axial hypotonia, spasticity, and brain imaging findings such as basal ganglia calcification and cerebral atrophy. Neither condition was observed in the family reported here; however, periodontal disease and severe alveolar bone loss were observed in the proband’s father (Figure 1-C4). Still, a non-syndromic manifestation involving the same gene cannot be ruled out.

The IFIH1 variant investigated in this study probably results in loss of function. NM_022168.4(IFIH1):c.1764dupT (p.Ala589fs21) causes a frameshift mutation that would lead to a stop codon 21 codons downstream. This stop codon is expected to induce nonsense-mediated decay (NMD) according to the 50-55 bases rule.⁸ However, according to GTEx Transcript Browser, some isoforms of the canonical protein may remain functional even without the exons affected by this variant. Therefore, the impact of this frameshift may differ depending on tissue-specific expression. These findings suggest potential functional implications for this variant. Its classification in variant databases is conflicting. In ClinVar, it has been classified as benign, based on its population frequency. This relatively high frequency reflects the sum of variations that can affect the (A)₈ track, all grouped under the rs553669430. In contrast, a CADD C-score 34.0 (apud gnomAD) suggests pathogenicity. The analyses described provide additional evidence for the potential pathogenicity of this variant.

GPC3

The GPC3 gene encodes glypican 3, a member of the heparan sulfate proteoglycan family. Glypicans are attached to the exocytoplasmic surface of the plasma membrane via a covalent glycosylphosphatidylinositol (GPI) anchor. Their primary role is to modulate the signaling pathways, including WNTs, Hedgehogs, fibroblast growth factors, and bone morphogenetic proteins.

There are no reports of NM_004484.4(GPC3): c.584A>G (p.N195S), rs745317547, in association with AI (PubMed, 2024), nor are there studies specifically linking this variant to any other dental abnormalities. GPC3 has been associated with oral tumors and alterations in mandibular morphology.³⁴

The classification of this variant is complex. Its allele frequency in the Brazilian population is high (ABraOM, 0.001) compared to the global population (0.000009). This may reflect an ascertainment bias, since the number of Brazilian individuals tested is relatively small, or it may reflect a higher frequency in this population. High population frequency is one of the strongest criteria for classifying a variant as benign. We identified a possible pathogenicity mechanism involving the creation of a new donor splice site, which may lead to partial exclusion of exon 3, causing a frameshift mutation and a stop codon 12 amino acids downstream. This new donor splice site would not abolish the original one. Consequently, the cell can present two differently spliced mRNA formed from the allele bearing this variant. This may explain how a variant with relatively high population frequency could exert a deleterious effect.

HPS3

The HPS3 gene encodes a protein (HPS3 or HPS3 Biogenesis of Lysosomal Organelles Complex 2 Subunit 1, BLOC2S1) containing a potential clathrin-binding motif, consensus dileucine signals, and tyrosine-based sorting signals for targeting vesicles of lysosomal lineage. The HPS3 protein may be important for organelle biogenesis associated with melanosomes, platelet-dense granules, and lysosomes. However, the literature on this gene, its metabolic pathways, and mutational impacts is limited. Mutations in this gene are associated with Hermansky-Pudlak syndrome type 3 (HPS3, MIM614072), an autosomal recessive disorder characterized by oculocutaneous albinism and platelet storage pool deficiency, resulting in abnormal bleeding. The only mention of teeth in these studies is “bleeding by tooth brushing.” No abnormal bleeding episodes were described in this family. Eleven types of Hermansky-Pudlak syndrome have been identified, and AI or other dental abnormalities have not been described, except for periodontitis in Hermansky-Pudlak syndrome type 2. HPS3 is part of the Biogenesis of Lysosomal Organelles Complex 2 (BLOC-2), involved in a central cellular trafficking pathway, although its precise molecular function remains unknown. The lack of information on functional domains and interaction sites within the BLOC-2 complex limits further discussions on the functional implications of the variant.³⁵

The HPS3 variant, NM_032383.5(HPS3):c.1897A>T (p.Met633Leu), has been classified as benign or likely benign by in silico predictors. Another variant in the same position, NM_032383.5(HPS3):c.1897A>G (p.Met633Val), was classified as VUS in ClinVar. Currently, no phenotypes have been associated with HPS3 variants in heterozygosis. Low allele frequency and strong evolutionary conservation of Met633 suggest functional impacts. However, considering the lack of association between HPS3 (or other HPS genes) and AI, it is difficult to conclude that heterozygous HPS3 variants contribute to AI.

TAB3

TAB3 encodes a homonymous protein required for the activation of the JNK and NF-kB signaling pathways (UniProt). Variants in this gene have not yet been associated with any disorder. However, its potential oncogenic role in various cancer types is currently under investigation. We did not find functional studies linking TAB3 to AI or to important processes inherent to tooth development. Therefore, despite some evidence suggesting this TAB3 variant may be deleterious—such as low allele frequency (0.00005 in the general population), the high evolutionary conservation of Val646, the functional role of Val646 within an antibody binding sequence, and high AlphaMissense, CADD, and SIFT scores—we could not theorize how this variant could impact teeth structure.

PRICKLE3

The PRICKLE3 gene plays multiple crucial roles in cellular processes. It is primarily associated with the planar cell polarity (PCP) pathway and is therefore essential for epithelial cell polarization during morphogenetic events (GeneCards). Additionally, the PRICKLE3 protein contributes to the organization and function of the centrosome/basal body, and has been proposed to regulate ciliogenesis mediated by Wtip, a LIM domain protein.³⁶ PRICKLE3 also functions as a component of the C-Jun N-terminal kinase (JNK) complex in the odontoblastic layer. This kinase is integral to the JNK pathway, which regulates processes such as adhesion, migration, and the formation of focal adhesion complexes in human dental papilla cells.³⁷ Moreover, PRICKLE3 participates in the WNT pathway, which regulates many cellular processes. Mutations in PRICKLE3 have been reported in Leber’s hereditary optic neuropathy (MIM535000), a mitochondrial DNA disorder. Furthermore, core PCP proteins have been shown to contribute to amelogenesis, a crucial process for enamel formation. Immunofluorescent staining trials in a rat model have revealed PRICKLE3 expression from the basement membrane of the outer enamel epithelium in the proliferation zone through to the maturation zone in the middle inner enamel secretion zone.³⁸

AI is a complex phenotype, characterized by a wide range of types and subtypes, an expanding list of associated genes, significant intra-familiar phenotype variation, and frequent overlap with post-eruptive conditions such as caries, enamel loss, and teeth loss. Autosomal dominant, autosomal recessive, and X-linked forms have all been described. Recent studies have also identified an oligogenic component in some families. In this study, analyzing WES with Qiagen’s QCI revealed candidate variants in six genes. Only ENAM has been previously associated with AI. IFIH1, PRICKLE3, and GPC3 have been associated with tooth or enamel development . The AI segregation in this family is suggestive of dominant X-linked inheritance, and three X-chromosomal candidate genes were identified: HPS3, TAB3, and PRICKLE3.

CNVs

A CNV involving the DMBT1 gene has been reported in association with enamel formation and mucosal defense.³⁹ Additionally, a promoter polymorphism in the DEFB4A gene is associated with chronic periodontitis.⁴⁰ Although the CNVs identified do not provide a strong causal explanation of AI in this Brazilian family, they may contribute to the oligogenic component of AI in this family.

Attributing a definitive value to these findings is challenging. Every individual carries many potentially damaging variants, but cellular and biological mechanisms often compensate for or bypass their effects. The high frequency of affected individuals in this family suggests a monogenic mechanism; however, the results described here do not allow us to identify a single “major effect gene,” nor to classify the remaining variants as modifiers. Thus, we cannot determine whether the inheritance pattern is monogenic or oligogenic. Rather, we propose a set of new candidate genes. Conclusions drawn from this study are subject to certain limitations. We report a relatively small family, and a larger number of affected individuals would be necessary to conduct a segregation analysis—a powerful method for variant classification. A possible workaround for this limitation is to publish these findings in the hope that other cases or families may be described in the future. Another limitation lies in the use of WES, which primarily captures variants outside the exons and intron borders are mostly not ascertained. Although Sanger sequencing is sometimes used to validate WES findings, it deemed unnecessary in this study, as the variant emerged independently in the WES data of all four affected patients.

This study also highlights the complexity of variant classification according to potential pathogenicity. Most classification systems and databases are dedicated to rare diseases; as such, a variant allele frequency > 0.001 is typically considered strong evidence of benign status. However, this criterion may be misleading when the phenotype is heterogeneous and overlaps with common characteristics. Evolutionary conservation is also complex due to alternative splicing. Considering these challenges, we reevaluated the six variants carefully, and our interpretations are presented here with caution. The pronounced AI clinical variability in this family suggests the involvement of additional genetic and/or environmental factors. The variants described here may act as causal modifiers, and functional studies will be essential to clarify their relevance. The use of multigene panels and WES in dentistry is relatively recent.¹,⁵,²⁵ This study illustrates the complexity associated with incorporating high-throughput methodologies into clinical research. Investigating a greater number of patients and families using WES will contribute to a better understanding of AI and its underlying genetic, molecular, and biochemical mechanisms—paving the way for new therapeutic approaches.

Conclusions

Six genetic variants were identified in this study; however, none could individually explain the phenotype. Therefore, we were unable to classify this family under any of the currently described monogenic mechanisms associated with AI.

Although oligogenic inheritance has been described for AI, the variants identified in this family do not allow a definitive conclusion of such mechanisms.

This study provided new candidate genes (IFIH1, GPC3, HPS3, TAB3, and PRICKLE3), along with their respective genetic variants, which may help clarify genetic mechanisms underlying AI.

References

• ¹ - Bloch-Zupan A, Rey T, Jimenez-Armijo A, Kawczynski M, Kharouf N, O-Rare consortium, Dure-Molla M, et al. Amelogenesis imperfecta: next-generation sequencing sheds light on Witkop's classification. Front Physiol. 2023;9(14):1130175. doi:10.3389/fphys.2023.1130175
  » https://doi.org/10.3389/fphys.2023.1130175
• ² - Smith CE, PoulterJA, Antanaviciute A, Kirkham J, Brookes SJ, Inglehearn CF, et al. Amelogenesis imperfecta; genes, proteins and pathways. Front Physiol. 2017;26(8):435. doi: 10.3389/fphys.2017.00435
  » https://doi.org/10.3389/fphys.2017.00435
• ³ - Broutin A, Bidi-Lebihan AK, Canceill T, Vaysse F, Bloch-Zupan A, Bailleul-Forestier I, et al. Association between malocclusions and amelogenesis imperfecta genotype and phenotype: a systematic review. Int Orthod. 2023;21(4):100789. doi: 10.1016/j.ortho.2023.100789
  » https://doi.org/10.1016/j.ortho.2023.100789
• ⁴ - Witkop CJ. Amelogenesis imperfecta, dentinogenesis imperfecta and dentin dysplasia revisited: problems in classification. J Oral Pathol. 1988;17(9-10):547-53. doi: 10.1111/j.1600-0714.1988.tb01332.x
  » https://doi.org/10.1111/j.1600-0714.1988.tb01332.x
• ⁵ - Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-24. doi: 10.1038/gim.2015.30
  » https://doi.org/10.1038/gim.2015.30
• ⁶ - Riggs ER, Andersen EF, Cherry AM, Kantarci S, Kearney H, Patel A, et al. Technical standards for the interpretation and reporting of constitutional copy-number variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics (ACMG) and the Clinical Genome Resource (ClinGen). Genet Med. 2020;22(2):245-57. doi: 10.1038/s41436-019-0686-8. Erratum in: Genet Med. 2021;23(11):2230. doi: 10.1038/s41436-021-01150-9.7.
  » https://doi.org/10.1038/s41436-019-0686-8
• ⁷ - Reza Khami M, Asgari S, Valizadeh S, Karami J, Rezaei A, Rezaei N. AMELX and ENAM Polymorphisms and Dental Caries. Int J Dent. 2022;31:2022:8501179. doi: 10.1155/2022/8501179
  » https://doi.org/10.1155/2022/8501179
• ⁸ - Nagy E, Maquat LE. A rule for termination-codon position within intron-containing genes: when nonsense affects RNA abundance. Trends Biochem Sci. 1998;23(6):198-9. doi: 10.1016/s0968-0004(98)01208-0
  » https://doi.org/10.1016/s0968-0004(98)01208-0
• ⁹ - Mårdh CK, Backman B, Holmgren G, Hu JC-C, Simmer JP, Forsman-Semb K. A nonsense mutation in the enamelin gene causes local hypoplastic autosomal dominant amelogenesis imperfecta (AIH2). Hum Mol Genet. 2002;11(9):1069-74. doi: 10.1093/hmg/11.9.1069
  » https://doi.org/10.1093/hmg/11.9.1069
• ¹⁰ - Kim JW, Simmer JP, Lin BP, Seymen F, Bartlett JD, Hu JC. Mutational analysis of candidate genes in 24 amelogenesis imperfecta families. Eur J Oral Sci. 2006;114 (Suppl 1):3-12; discussion 39-41, 379. doi: 10.1111/j.1600-0722.2006.00278.x
  » https://doi.org/10.1111/j.1600-0722.2006.00278.x
• ¹¹ - Hu JC, Yamakoshi Y. Enamelin and autosomal-dominant amelogenesis imperfecta. Crit Rev Oral Biol Med. 2003;14(6):387-98. doi:10.1177/154411130301400602
• ¹² - Shore RC, Bäckman B, Elcock C, Brook AH, Brookes SJ, Kirkham J. The structure and composition of deciduous enamel affected by local hypoplastic autosomal dominant amelogenesis imperfecta resulting from an ENAM mutation. Cells Tissues Organs. 2010;191(4):301-6. doi: 10.1159/000258703
  » https://doi.org/10.1159/000258703
• ¹³ - Hart TC, Hart PS, Gorry MC, Michalec MD, Ryu OH, Uygur C, et al. Novel ENAM mutation responsible for autosomal recessive amelogenesis imperfecta and localised enamel defects. J Med Genet. 2003;40(12):900-6. doi: 10.1136/jmg.40.12.900
  » https://doi.org/10.1136/jmg.40.12.900
• ¹⁴ - Kang HY, Seymen F, Lee SK, Yildirim M, Tuna EB, Patir A, et al. Candidate gene strategy reveals ENAM mutations. J Dent Res. 2009;88(3):266-9. doi: 10.1177/0022034509333180
  » https://doi.org/10.1177/0022034509333180
• ¹⁵ - Dashash M, Bazrafshani MR, Poulton K, Jaber S, Naeem E, Blinkhorn AS. Enamelin/ameloblastin gene polymorphisms in autosomal amelogenesis imperfecta among Syrian families. J Investig Clin Dent. 2011;2(1):16-22. doi: 10.1111/j.2041-1626.2010.00038.x
  » https://doi.org/10.1111/j.2041-1626.2010.00038.x
• ¹⁶ - Wang X, Zhao Y, Yang Y, Qin M. Novel ENAM and LAMB3 mutations in Chinese families with hypoplastic amelogenesis imperfecta. PLoS One. 2015;13;10(3):e0116514. doi: 10.1371/journal.pone.0116514
  » https://doi.org/10.1371/journal.pone.0116514
• ¹⁷ - Koruyucu M, Kang J, Kim YJ, Seymen F, Kasimoglu Y, Lee ZH, et al. Hypoplastic AI with highly variable expressivity caused by ENAM mutations. J Dent Res. 2018;97(9):1064-9. doi: 10.1177/0022034518763152
  » https://doi.org/10.1177/0022034518763152
• ¹⁸ - Zhang H, Hu Y, Seymen F, Koruyucu M, Kasimoglu Y, Wang SK, et al. ENAM mutations and digenic inheritance. Mol Genet Genomic Med. 2019;7(10):e00928. doi: 10.1002/mgg3.928
  » https://doi.org/10.1002/mgg3.928
• ¹⁹ - Yu S, Zhang C, Zhu C, Quan J, Liu D, Wang X, et al. A novel ENAM mutation causes hypoplastic amelogenesis imperfecta. Oral Dis. 2022;28(6):1610-9. doi: 10.1111/odi.13877
  » https://doi.org/10.1111/odi.13877
• ²⁰ - Hart PS, Michalec MD, Seow WK, Hart TC, Wright JT. Identification of the enamelin (g.8344delG) mutation in a new kindred and presentation of a standardized ENAM nomenclature. Arch Oral Biol. 2003;48(8):589-96. doi: 10.1016/s0003-9969(03)00114-6
  » https://doi.org/10.1016/s0003-9969(03)00114-6
• ²¹ - Gutierrez SJ, Chaves M, Torres DM, Briceño I. Identification of a novel mutation in the enamalin gene in a family with autosomal-dominant amelogenesis imperfecta. Arch Oral Biol. 2007;52(5):503-6. doi: 10.1016/j.archoralbio.2006.09.014
  » https://doi.org/10.1016/j.archoralbio.2006.09.014
• ²² - Chan HC, Mai L, Oikonomopoulou A, Chan HL, Richardson AS, Wang SK, et al. Altered enamelin phosphorylation site causes amelogenesis imperfecta. J Dent Res. 2010;89(7):695-9. doi:10.1177/0022034510365662
• ²³ - Lindemeyer RG, Gibson CW, Wright TJ. Amelogenesis imperfecta due to a mutation of the enamelin gene: clinical case with genotype-phenotype correlations. Pediatr Dent. 2010;32(1):56-60.
• ²⁴ - Song YL, Wang CN, Zhang CZ, Yang K, Bian Z. Molecular characterization of amelogenesis imperfecta in Chinese patients. Cel Tissues Organs. 2012;196(3):271-9. doi: 10.1159/000334210
  » https://doi.org/10.1159/000334210
• ²⁵ - Simmer SG, Estrella NM, Milkovich RN, Hu JC. Autosomal dominant amelogenesis imperfecta associated with ENAM frameshift mutation p.Asn36Ilefs56. Clin Genet. 2012;83(2):195-7. doi: 10.1111/j.1399-0004.2012.01887.x
  » https://doi.org/10.1111/j.1399-0004.2012.01887.x
• ²⁶ - Seymen F, Lee KE, Koruyucu M, Gencay K, Bayram M, Tuna EB, et al. ENAM mutations with incomplete penetrance. J Dent Res. 2014;93(10):988-92. doi: 10.1177/0022034514548222
  » https://doi.org/10.1177/0022034514548222
• ²⁷ - Kim YJ, Lee Y, Zhang H, Wright JT, Simmer JP, Hu JC, et al. Translational attenuation by an Intron retention in the 5' UTR of ENAM causes amelogenesis imperfecta. Biomedicines. 2021;9(5):456. doi: 10.3390/biomedicines9050456
  » https://doi.org/10.3390/biomedicines9050456
• ²⁸ - Zhang Z, Zou X, Feng L, Huang Y, Chen F, Sun K, et al. Splicing mutations in AMELX and ENAM cause amelogenesis imperfecta. BMC Oral Health. 2023;23(1):893. doi: 10.1186/s12903-023-03508-8
  » https://doi.org/10.1186/s12903-023-03508-8
• ²⁹ - Pavlic A, Petelin M, Battelino T. Phenotype and enamel ultrastructure characteristics in patients with ENAM gene mutations g.13185-13186insAG and 8344delG. Arch Oral Biol. 2007;52(3):209-17. doi: 10.1016/j.archoralbio.2006.10.010
  » https://doi.org/10.1016/j.archoralbio.2006.10.010
• ³⁰ - Pavlic A, Battelino T, Trebusak Podkrajsek K, Ovsenik M. Craniofacial characteristics and genotypes of amelogenesis imperfecta patients. Eur J Orthod. 2011;33(3):325-31. doi: 10.1093/ejo/cjq089
  » https://doi.org/10.1093/ejo/cjq089
• ³¹ - Urzúa OB, Ortega PA, Rodríguez ML, Morales BI. Analisis genético, clínico y molecular de una familia afectada con una malformación del esmalte dental [Genetic, clinical and molecular analysis of a family affected by amelogenesis imperfecta]. Rev Méd Chile. 2005;133(11):1331-40. Spanish. doi: 10.4067/s0034-98872005001100009
  » https://doi.org/10.4067/s0034-98872005001100009
• ³² - Wang S-K, Zhang H, Wang Y-L, Seymen F, Koruyucu M, Simmer JP, et al. Phenotypic variability in LAMA3-associated amelogenesis imperfecta. Oral Dis. 2023;29(8):3514-24. doi: 10.1111/odi.14425
  » https://doi.org/10.1111/odi.14425
• ³³ - Rutsch F, MacDougall M, Lu C, Buers I, Mamaeva O, Nitschke Y, et al. A Specific IFIH1 gain-of-function mutation causes singleton-merten syndrome. Am J Hum Genet. 2015;96(2):275-82. doi: 10.1016/j.ajhg.2014.12.014
  » https://doi.org/10.1016/j.ajhg.2014.12.014
• ³⁴ - Mian M, Ranjitkar S, Townsend GC, Anderson PJ. Alterations in mandibular morphology associated with glypican 1 and glypican 3 gene mutations. Orthod Craniofac Res. 2017;20(3):183-7. doi: 10.1111/ocr.12170
  » https://doi.org/10.1111/ocr.12170
• ³⁵ - Thomason PA, Corbyn R, Lilla S, Sumpton D, Gilbey T, Insall RH. Biogenesis of lysosome-related organelles complex-2 is an evolutionarily ancient proto-coatmer complex. Curr Biol. 2024;34(15):3564-81.e6. doi: 10.1016/j.cub.2024.06.08
  » https://doi.org/10.1016/j.cub.2024.06.08
• ³⁶ - Chu C-W, Ossipova O, Ioannou A, Sokol SY. Prickle3 synergizes with Wtip to regulate basal body organization and cilia growth. Sci Rep. 2016;6:24104. doi: 10.1038/srep24104
  » https://doi.org/10.1038/srep24104
• ³⁷ - Luo J, Tan X, Ye L, Wang C. C-Jun N-terminal kinase (JNK) pathway activation is essential for dental papilla cells polarization. PLoS One. 2021;16(3):e0233944. doi: 10.1371/journal.pone.0233944
  » https://doi.org/10.1371/journal.pone.0233944
• ³⁸ - Nishikawa S, Kawamoto T. Localization of core planar cell polarity proteins, prickles, in ameloblasts of rat incisors: possible regulation of enamel rod decussation. Acta Histochem Cytochem. 2015;48(2):37-45. doi: 10.1267/ahc.1404639
  » https://doi.org/10.1267/ahc.1404639
• ³⁹ - Polley S, Louzada S, Forni D, Sironi M, Balaskas T, Hains DS, et al. Evolution of the rapidly mutating human salivary agglutinin gene (DMBT1) and population subsistence strategy. Proc Natl Acad Sci U S A. 2015;112(16):5105-10. doi: 10.1073/pnas.1416531112
  » https://doi.org/10.1073/pnas.1416531112
• ⁴⁰ - Kurt-Bayrakdar S, Ozturk A, Kara N. DEFB4A promoter polymorphism is associated with chronic periodontitis: a case-control study. Genet Test Mol Biomarkers. 2020;24(3):113-9. doi: 10.1089/gtmb.2019.0218
  » https://doi.org/10.1089/gtmb.2019.0218

• Data availability:
  The datasets generated during during the current study are available in the SciELO Data Repository - https://doi.org/10.48331/SCIELODATA.CTQBYB.

Edited by

• Editor:
  Ana Carolina Magalhães

• Associate Editor:
  Joel Santiago Junior

Data availability

The datasets generated during during the current study are available in the SciELO Data Repository - https://doi.org/10.48331/SCIELODATA.CTQBYB.

Publication Dates

• Publication in this collection
  01 Sept 2025
• Date of issue
  2025

History

• Received
  10 Nov 2024
• 
  09 June 2025
• Accepted
  05 July 2025