ANO5 p.I616F variant drives florid cemento-osseous dysplasia by disrupting transmembrane domain structure

Abstract

Background

To date, genetic studies on florid cemento-osseous dysplasia (FCOD) remain limited. This study aimed to elucidate the clinical features and genetic basis of FCOD in Chinese populations, with particular emphasis on identifying pathogenic ANO5 variants and their mechanistic contributions to disease development.

Methods

Clinical data from 17 FCOD patients diagnosed between May 2015 and December 2024 at Kunming Medical University Affiliated Stomatological Hospital were retrospectively analyzed. Whole-exome sequencing was performed on a familial cohort (8 individuals). Identified variants were validated using Sanger sequencing. Pathogenicity was predicted via SIFT, PolyPhen-2, and MutationTaster, while protein structural alterations were modeled with AlphaFold 2. Histological assessment included H&E staining, Masson trichrome, RUNX2 immunohistochemistry, and TRAP staining to quantify osteoclastic activity.

Results

Among the 17 FCOD patients included in this study, 10 (58.8%) were female, with a mean age of 47.4 ± 13.2 years. Genetic analysis identified Anoctamin 5 (ANO5) variants in five cases, including a novel variant (p.I616F). Pedigree analysis confirmed that p.I616F was a heterozygous variant, and bioinformatics prediction indicated its likely pathogenicity. Structural modeling further demonstrated that this variant disrupted the α-helix of ANO5 and altered the catalytic site’s conformation. Additionally, histological evaluation revealed significantly elevated RUNX2 expression and increased osteoclast activity in the jawbone tissues of FCOD patients.

Conclusion

This study provides the first evidence that ANO5 variants (particularly p.I616F) may represent key genetic factors contributing to FCOD development in the Chinese population. The mutation likely disrupts the α-helical structure of ANO5 protein, impairing its catalytic function and potentially dysregulating bone remodeling processes in FCOD pathogenesis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-025-06717-5.

Background

Fibro-osseous lesions of the jaw constitute a group of disorders that severely impair oral function and quality of life, primarily driven by dysregulated bone metabolism and genetic factors. Florid cemento-osseous dysplasia (FCOD), a major subtype of cemento-osseous dysplasia (COD), was clearly defined in the 2022 WHO classification, underscoring the pressing need for standardized diagnosis and treatment [1–3]. However, current research on FCOD faces three critical challenges: First, reported variants in the RAS-MAPK pathway explain only 28% of cases, highlighting substantial genetic heterogeneity [4]. Second, while Anoctamin 5 (ANO5) variants are clearly linked to gnathodiaphyseal dysplasia (GDD) and familial gigantiform cementoma [5–7], their role in FCOD remains unverified. Moreover, newly identified ANO5 variants associated with FCOD lack functional validation [8]. To bridge these gaps, we conducted the first systematic analysis of the ANO5 p.I616F variant in a Chinese familial FCOD cohort using whole-exome sequencing. Additionally, we integrated AlphaFold 2 protein structure prediction with histological analysis, establishing a comprehensive framework from variant identification to functional validation. These findings provide novel insights into the potential role of ANO5 variants in bone metabolism dysregulation, which may inform future mechanistic investigations.

Materials and methods

Study population

This observational case series study analyzed patients diagnosed with FCOD at the Department of Oral and Maxillofacial Surgery, Affiliated Stomatological Hospital of Kunming Medical University, from May 2015 to December 2024. Inclusion criteria required: (1) radiographic confirmation of characteristic FCOD features on computed tomography (CT), showing multiple hyperdense jawbone masses, and (2) documented informed consent. Exclusion criteria included: (1) concurrent jawbone lesions, such as fibrous dysplasia, ossifying fibroma, or osteosarcoma, and (2) prior head/neck radiotherapy or bisphosphonate therapy, which could interfere with diagnostic accuracy. Complete cone-beam CT (CBCT) and panoramic radiographic datasets were retrieved from the hospital’s Picture Archiving and Communication System (PACS) for analysis. We concurrently collected jaw bone tissues surgically removed from orthognathic surgery patients, which were used as controls for tissue staining.

Pedigree analysis and whole-exome sequencing

In accordance with clinical research protocols and following informed consent from patients and their families, peripheral venous blood samples (2 mL per participant) were collected for genetic analysis. Whole-exome sequencing (WES) was carried out by two specialized companies—Biolinker Technology (Kunming) Co., Ltd. and MGI Tech Co., Ltd.—utilizing the high-throughput Illumina NovaSeq 6000 platform. This sequencing covered a 12 Mb capture region with an average depth of 200× (based on the GRCh37/hg19 human reference genome). To identify genetic variants, raw sequencing data underwent variant calling (detection of SNVs and small indels) using GATK (Genome Analysis Toolkit) and SAMtools. Subsequently, ANNOVAR was employed for functional annotation, with filtering criteria applied: only exonic variants with a minor allele frequency (MAF) < 1% in population databases were retained for further evaluation. The predicted pathogenicity of these variants was then assessed using computational tools, including SIFT, PolyPhen-2, and MutationTaster. Sanger sequencing was ultimately performed to confirm segregation of candidate variants within the family.

Sanger sequencing

To identify ANO5 gene variants, we performed standard molecular genetic analyses. Genomic DNA was extracted from peripheral blood samples of the patient and available family members. PCR amplification of the target ANO5 regions was carried out using specific primers (ANO5-F: 5′-AGGTTTGGAACTATTGGGCTAA-3′; ANO5-R: 5′-TCACAGGAAAGAAAAGTAACAAAA-3′). The PCR products were purified and sequenced bidirectionally using an ABI PRISM 3730xl sequencer. The raw sequencing data were aligned against the human reference genome (GRCh37/hg19) and analyzed using Sequencher software (v5.0, Gene Codes Corporation). Putative pathogenic variants were identified by comparing the sequences with the reference dataset. Detected variants were checked for co-segregation within the family and screened against population databases (dbSNP, ExAC, gnomAD) to exclude common polymorphisms (minor allele frequency, MAF < 1%). Finally, the functional consequences of candidate variants were predicted using in silico tools, including SIFT, PolyPhen-2, and Mutation Taster. To comprehensively evaluate the novelty of this variant, we performed a multi-database analysis by systematically querying: 1) international authoritative databases including gnomAD (population frequency database), ClinVar (clinical variant database), and HGMD (Human Gene Mutation Database).

Functional prediction of ANO5 gene variants

To evaluate the pathogenic potential of ANO5 variants, we implemented a multi-tiered bioinformatics strategy. First, the functional effects of amino acid substitutions were assessed using SIFT to predict tolerance to changes, PolyPhen-2 to model structural and functional perturbations, and Mutation Taster to generate integrated pathogenicity scores based on evolutionary conservation and protein structural integrity. To elucidate mechanistic consequences, wild-type and mutant ANO5 protein structures (PDB ID: AF-P31324-F1) were computationally modeled via the AlphaFold 2 TIB server, with particular emphasis on conformational shifts in the transmembrane domains (TM3–TM6) and calcium-binding motifs. The Combined Annotation Dependent Depletion (CADD) algorithm was further employed to quantify the systemic impact of variants on protein function. Stringent inclusion criteria were applied: putative variants were retained only if classified as deleterious or pathogenic by ≥ 3 prediction tools and concurrently yielded a CADD score ≥ 20.

Histological analysis

Hematoxylin & Eosin (H&E) staining

Fixed and decalcified mandibular tissues were embedded in paraffin and sectioned at 5 μm thickness. The sections were deparaffinized in xylene, rehydrated through an alcohol gradient, and stained with Mayer’s hematoxylin for 5 min. After differentiation in acid alcohol, sections were counterstained with eosin for 30 s, dehydrated through an ascending alcohol series, cleared in xylene, and mounted with neutral balsam prior to bright-field microscopy.

Masson’s trichrome staining

Deparaffinized and rehydrated sections were stained sequentially with Weigert’s iron hematoxylin (10 min) for nuclei identification, Biebrich scarlet-acid fuchsin (5 min) for muscle and cytoplasm, 2.5% phosphomolybdic acid solution for differentiation (5 min), and aniline blue (5 min) for collagen fiber visualization. Following dehydration and clearing in xylene, slides were mounted with resinous medium and examined under bright-field microscopy to assess fibrotic tissue distribution.

Immunohistochemical staining

After antigen retrieval in citrate buffer at 95 °C for 20 min, sections were incubated with 3% H₂O₂ to block endogenous peroxidase activity. Non-specific binding was blocked with 5% bovine serum albumin for 30 min at room temperature. Primary antibody against RUNX2 (1:200 dilution) was applied overnight at 4 °C, followed by incubation with HRP-conjugated secondary antibody (goat anti-rabbit IgG, 1:500) for 1 h at 37 °C. Signals were developed using DAB substrate (5 min) and counterstained with modified Harris hematoxylin (1 min), with subsequent dehydration, clearing, and mounting for quantitative histomorphometric analysis.

TRAP staining for osteoclasts

Tartrate-resistant acid phosphatase (TRAP) detection was performed by incubating sections in sodium acetate buffer containing naphthol AS-MX phosphate (0.1 mg/mL), fast red violet LB salt (0.5 mg/mL), and 10 mM sodium tartrate at 37 °C for 45 min. After counterstaining with Gill’s hematoxylin (2 min), TRAP-positive multinucleated osteoclasts (≥ 3 nuclei) were identified by bright fuschia cytoplasmic staining under light microscopy at 200× magnification. Three non-overlapping fields per section were imaged for quantitative assessment of osteoclast number and activity.

Results

Clinical characteristics

This study analyzed clinical and genetic data from 17 patients (10 females, 7 males; a mean age of 47.4 ± 13.2) diagnosed between 2015 and 2024 (Table S1). The most common presentations included odontogenic pain (41.2%, n = 6), mandibular/maxillary purulent discharge (29.4%, n = 5), infected extraction sockets (5.9%, n = 1), and orofacial swelling (5.9%, n = 1), while 23.5% (n = 4) reported nonspecific symptoms. CBCT imaging identified sclerotic bone lesions in all cases, with secondary infections in 58.8% (n = 10).

In this study, we retained variants with a population allele frequency < 1% and predicted deleterious effects on protein-coding genes. Our screening identified 12 genes involved in osteogenesis and osteoclastogenesis (Table S2). In contrast to previous reports, we did not detect any pathogenic mutations in the known FCOD-associated genes (KRAS, FGFR1, BRAF, HRAS, NRAS, FGFR3, GNAS, ATG9A, MDM2, FOS, FOSB, COL1A1, TBX3, SATB2, SETD2, MAPK10, COL6A2, DCDC1, or NOTCH4) [4, 9]. Through systematic literature review and bioinformatic analysis, we ultimately identified ANO5 as the most promising candidate gene for further investigation. Genetic testing detected ANO5 variants in 29.4% (n = 5) of patients, including the variants p.I616F (n = 2), p.I466F (n = 1), and p.L322F (n = 1). Furthermore, we detected several significant non-coding variants: a 24-bp insertion (c.−162_−161insAGGAGGAGGGGAATGAGGAGGAGG) and a single nucleotide variant (c.−136G > C) in the 5’UTR region, along with six single-base deletions (including c. *100delT, c.*115delT and c.*1757delA) and four single-nucleotide substitutions (c. *496A > G, c.*1286G > T, c. *1429A > G and c.*3178C > G) in the 3’UTR region. Additionally, we observed an exon 13 duplication/deletion. These UTR variants may influence transcriptional regulation or mRNA stability, while the exon 13 structural variation and missense mutations could collectively impair protein function. The p.I616F (c.1846 A > T) variant was absent from all population databases (gnomAD, ClinVar and HGMD), confirming its status as a previously unreported mutation (Fig. 1A and Table S1).

Fig. 1.

Genetic pedigree and radiographic findings in the FCOD family. A The pedigree chart identifies the proband (arrow) and unaffected individuals. B Clinical frontal photographs and corresponding panoramic radiographs of affected family members. Individual II-2: Suspected FCOD; Individual III−2: Affected with FCOD, 38y, presents the mutation c.1846 A > T; III−5: Affected with FCOD, 50y, presents the mutation c.1846 A > T; III−5:Healthy, 37y; IV3: Healthy, 18y; IV4: Healthy, 14y; IV5: Healthy, 8y

Family investigation

The proband (III-2), a 38-year-old female, reported persistent pain in her lower right posterior teeth for over a month, accompanied by radiating facial pain and difficulty in mastication. Clinical examination revealed severe attrition of tooth 46 extending to the gingival margin, buccal bone expansion with tenderness, and oral floor swelling causing tongue elevation. CBCT imaging identified diffuse, irregular hyperdense lesions in the mandibular apical region and narrowed or obliterated periodontal ligament spaces (Fig. 1B). These findings confirmed the calcified mature stage of FCOD. Laboratory tests indicated elevated alkaline phosphatase (122.8 IU/L), while calcium and phosphorus metabolism remained within normal limits. Postoperative pathology confirmed hypercementosis (cemental dysplasia), marked by characteristic cementum-like deposits and proliferating cementoblasts. Family investigation revealed phenotypic heterogeneity among affected relatives. The proband’s mother exhibited mandibular pain and purulent discharge, with CBCT showing localized hyperdense lesions in the right mandible—though complete edentulism precluded a definitive diagnosis. Meanwhile, the proband’s children (IV-1 and IV-2) remained asymptomatic. A more severe presentation was observed in the proband’s cousin (III-5), who displayed panoral hypercementosis complicated by secondary osteomyelitis (evidenced by sequestrum exposure and purulent discharge), consistent with florid FCOD in its terminal stage. Notably, her children (IV-6 and IV-7) showed no symptoms. The pedigree exhibited an autosomal dominant inheritance pattern, with significant variations in disease progression among carriers.

Genetic sequencing and variant analysis

WES was conducted on eight pedigree members, revealing a shared heterozygous ANO5 variant (NM_213599:exon17:c.1846 A > T, p.I616F) in three individuals (II2, III2, and III5). Sanger sequencing confirmed the variant (Fig. 2A), which has not been previously reported in public databases (ClinVar and gnomAD). Computational tools consistently predicted its pathogenicity: SIFT (score 0, damaging), PolyPhen-2 (score 1, probably damaging), and MutationTaster (score 1, disease-causing) all classified it as deleterious. Further supporting this, the MutationAssessor score (3.32) and PROVEAN value (−3.88) indicated strong functional disruption. Strikingly, the CADD_phred score (25.8) ranked this variant among the top 0.5% of harmful human genomic variants, aligning it with known pathogenic ANO5 variants like p.Cys356Tyr [7].

Fig. 2.

Molecular characterization of FCOD family members. A Sanger sequencing result of the c.1846 A > T. Het: heterozygote. B Evolutionary conservation analysis of the mutated protein domain. AOA3B5BMZ4_9TELE: Stegastes partitus (bicolor damselfsh); AOA8VOZCQ4_CHICK: Gallus gallus (chicken); ANO5_ MOUSE: Mus musculus (Mouse); F1M3M3_RAT: Rattus norvegicus (Rat); AOAAA9TV24 BOVIN: Bos taurus (Bovine); W5PD36_SHEEPO: vis aries (Sheep); AOA5F9CRL2 RABIT_Oryctolagus cuniculus (Rabbit); ANO5_HUMAN: Homo sapiens (Human); AOA8 × 9AEQ1_PIG: Sus scrofa (Pig); AOA8CONPM4_CANLF: Canis lupus familiaris (Dog); AOA2I2UVW4_FELCA: Felis catus (Cat); AOA9LOT9A1_ HORSE: Equus caballus (Horse); C-D Predicted three-dimensional protein structure highlighting variant localization. C The structure of the wild - type ANO5 protein; D The structure of the mutant ANO5 protein (p.I616F); Blue arrow: loop structure

Protein structure prediction and functional analysis

To assess the functional consequences of the ANO5 p.I616F variant, we generated structural models of wild-type and mutant proteins using AlphaFold 2 TIB, leveraging the gene’s high evolutionary conservation (Fig. 2B). Structural analysis revealed that this variant may disrupt a critical α-helical domain, markedly reducing protein stability—a conclusion corroborated by molecular dynamics simulations demonstrating widespread structural destabilization. Closer inspection of the variant site uncovered profound spatial rearrangement of the active site. While the wild-type protein maintained a highly conserved helical architecture around its active site, the mutant showed displaced helices featuring localized coiling and distortion. These structural perturbations likely compromise protein function through three key mechanisms: (1) diminished thermal stability, (2) increased susceptibility to proteasomal degradation, and (3) impaired calcium-binding site integrity. Collectively, these alterations may contribute to aberrant membrane localization (Fig. 2C).

Histological analysis

Histopathological evaluation of FCOD lesions revealed distinct structural alterations compared to healthy controls. Hematoxylin and eosin (H&E) staining demonstrated replacement of normal bone with dense fibrous tissue interspersed with irregular cementum-like deposits (Fig. 3A), while Masson’s trichrome staining confirmed excessive collagenous matrix deposition (Fig. 3B). Immunohistochemical analysis showed marked upregulation of RUNX2, the master osteogenic regulator (Fig. 3C), concurrent with increased TRAP-positive osteoclasts (Fig. 3D). These findings indicate a dysregulated bone remodeling process characterized by paradoxical co-activation of both osteoblastic (RUNX2-mediated) and osteoclastic activities in FCOD pathogenesis [3].

Fig. 3.

Histopathological evaluation. A Hematoxylin and eosin (H&E) stained sections. Blue arrow: cementoid; B Masson’s trichrome connective tissue staining; C Immunohistochemical (IHC) detection of RUNX2; D Tartrate-resistant acid phosphatase (TRAP) activity assessment (Scale bar = 100 μm, 100×)

Discussion

Pathogenic mechanism and clinical diagnostic value of the ANO5-p.I616F variant in FCOD

The human ANO5 gene, located at chromosome 11p14.3, comprises 23 exons that encode ANO5 (TMEM16E), a 913-amino-acid transmembrane protein. This protein functions both as a calcium-activated chloride channel (CaCC) and as a phospholipid scramblase (PLS) [10]. GDD represents a rare autosomal inherited skeletal disorder manifesting complex osseous abnormalities, such as fibro-osseous jaw lesions, compromised bone integrity, long bone diaphyseal sclerosis, and recurrent pathological fractures. This condition demonstrates strong genotype-phenotype correlation with ANO5 gain-of-function mutations [11, 12]. Importantly, pathogenic heterozygous ANO5 variants are also documented in familial gigantiform cementoma [7], thereby reinforcing ANO5’s pathogenic role in skeletal dysplastic disorders. Furthermore, the ANO5 p.C356W missense mutation has been reported in familial florid cemento-osseous dysplasia, implicating ANO5 pathogenic mechanisms in FCOD etiology [8]. Collectively, these findings were integrated into our discussion to elucidate the pathophysiological relationship between ANO5 dysfunction and maxillofacial developmental pathologies.

In this study, we identified the novel ANO5-p.I616F missense variant in patients with FCOD. Previous studies have established that GDD, a rare autosomal dominant disorder resulting from ANO5 gain-of-function variants [13]. Due to the limited sample size in this study, we cannot definitively determine its inheritance pattern. Regarding III-5 (whose parents were clinically unaffected), we propose two potential explanations: (1) Parental gonadal mosaicism: A special genetic phenomenon where a gene mutation exists in one parent’s germ cells (sperm or oocytes) but is not detected in their somatic cells (e.g., blood, skin). This leads to offspring presenting with “apparently de novo” genetic disorders while parental genetic testing appears normal [14, 15]. (2) Undocumented familial variant: The mutation may represent a multi-generationally transmitted but previously unrecognized familial variant.

Multi-scale bioinformatics analysis and molecular dynamics simulations revealed that this variant induces conformational changes in the ANO5 calcium ion channel protein, resulting in a distinct pathological phenotype marked by concurrent osteogenic hyperactivity and osteoclast proliferation. These findings offer crucial insights for molecular diagnosis and targeted therapeutic strategies. Current FCOD diagnosis heavily depends on imaging features, yet radiological overlap with periapical inflammation often leads to misdiagnosis [16, 17]. The molecular marker identified here enhances diagnostic precision, enabling clearer differentiation of FCOD from other fibrous - osseous lesions, such as ossifying fibroma (CDC73 variants) [18–20], fibrous dysplasia (GNAS variants) [21], and gnathodiaphyseal dysplasia [6, 22–25]. Unlike the ANO5 p.C356Y variant, commonly associated with familial gigantiform cementoma [7, 26], p.I616F localizes near transmembrane domains (TM6–TM7) and introduces a rigid aromatic ring, likely impairing conformational dynamics in this region. This structural perturbation may disrupt ion channel function or lipid-scrambling activity, leading to a more severe loss of function. In contrast, p.C356Y, located in the extracellular domain, exerts a milder effect on core protein activity. Additionally, variants in the RAS-MAPK pathway—frequently reported in Western populations [4]—were absent in our cohort, further underscoring ethnic genetic heterogeneity in FCOD. Based on these results, we propose prioritizing ANO5 variant screening for cases with characteristic radiological features and integrating genetic testing into diagnostic protocols for early-onset (< 50 years) or familial FCOD.

ANO5 p.I616F variant potentially disrupts bone homeostasis via calcium signaling and altered Osteogenesis-Osteoclastogenesis balance

The ANO5 p.I616F variant disrupts bone homeostasis through multiple interconnected mechanisms, primarily via calcium signaling dysregulation and altered osteogenesis-osteoclastogenesis balance. As demonstrated by structural modeling (AlphaFold 2), this variant induces α-helix disruption and conformational changes at the active site, likely impairing ANO5-mediated Ca²⁺ influx and intracellular calcium oscillation dynamics [5, 27, 28]. This calcium signaling dysfunction triggers bidirectional activation of bone metabolic pathways: while aberrant WNT/β-catenin pathway stimulation promotes excessive bone formation, concurrent upregulation of the RANKL-NFATc1 cascade enhances bone resorption [27, 29]. As both processes become dysregulated, the normal bone remodeling cycle fails to maintain homeostasis. Supporting this finding, our study revealed elevated RUNX2 expression (indicating increased osteogenesis) alongside heightened TRAP staining (demonstrating osteoclast hyperactivity) in patient tissues - observations consistent with previous murine models [30, 31]. Additionally, ANO5’s lipid scramblase activity may contribute to bone homeostasis through membrane phospholipid metabolism regulation [22, 32], suggesting that the p.I616F variant could disrupt this secondary function. While these mechanisms collectively explain FCOD pathology observed in patients, further validation using patient-derived cellular models remains essential for comprehensive functional characterization.

ANO5 variants in FCOD: distinct variant spectrum and controversial osteogenic regulation mechanisms

This study demonstrates that ANO5 variants linked to bone metabolic disorders exhibit a unique variant spectrum in FCOD patients. Unlike the p.C356Y variant prevalent in familial gigantiform cementoma [7, 26], the identified p.I616F variant localizes to the transmembrane domain, with AlphaFold 2 predicting substantial conformational changes in channel gating. FCOD patients also present delayed disease onset, suggesting site-specific mutational effects drive phenotypic divergence. Mechanistically, while ANO5 variants consistently impair osteoclast activity [11, 27, 29, 30, 33, 34], their osteogenic role remains debated—some studies associate deficiency with enhanced differentiation [5, 11, 24, 30, 35, 36], while others report suppression [27]. Our histology showed concurrent RUNX2 upregulation and increased osteoclast activity in FCOD, with discrepancies potentially arising from: (1) variant-dependent channel dysfunction (e.g., p.C356Y increases vs. p.I616F decreases calcium permeability); (2) model variability (in vitro vs. in vivo systems); or (3) sampling timepoint differences.

Study limitations and future directions

While our study identifies ANO5 p.I616F as a novel FCOD-associated variant, several limitations must be acknowledged. First, the phenotypic manifestation may be age-dependent, especially in the younger Generation IV individuals of this study. Second, although structural modeling and previous studies indicate disrupted calcium signaling, definitive mechanistic validation requires electrophysiological studies in CRISPR-edited cells and animal models demonstrating bone remodeling defects. Finally, the restricted sample size (n = 17) limits detailed phenotype-genotype correlation analyses. To overcome these limitations, future research should focus on three primary objectives: (1) establishing patient-derived induced pluripotent stem cell models to investigate variant-induced cellular changes; (2) developing humanized mouse models for therapeutic target validation; and (3) expanding mutational distribution analyses to diverse ethnic populations through international multicenter collaborations.

Conclusions

This study identifies ANO5 variants (notably p.I616F) as potential contributors to FCOD pathogenesis in Chinese individuals. Computational analyses indicate the mutation may alter ANO5’s α-helical conformation, possibly compromising its catalytic activity and implicating ANO5 dysfunction in FCOD-related bone metabolism abnormalities.

Abbreviations

FCOD

Florid cemento-osseous dysplasia

COD

Cemento-osseous dysplasia

GDD

Gnathodiaphyseal dysplasia

CBCT

Complete cone-beam CT

WES

Whole-exome sequencing

MAF

minor allele frequency

CADD

Combined Annotation Dependent Depletion

H&E

Hematoxylin & Eosin

TRAP

Tartrate-resistant acid phosphatase

Authors’ contributions

Lanlan Zhang conceptualized the study, performed data analysis, and drafted the original manuscript. Wenxi Jin, Yunfa Qin, and Rongju Xie conducted comprehensive pedigree analyses and genetic investigations. Yonghui Zhang, Yemei Qian, Jingyi Li, Yu Liu, and Yan’an Shi were responsible for patient recruitment, clinical evaluations, and data acquisition. Weihong Wang provided project supervision, funding acquisition, manuscript revision, and final approval. Lanlan Zhang, Wenxi Jin, and Yunfa Qin should be considered equal contributors. All authors participated in study design, data interpretation, and critical manuscript review. Each contributor approved the final version and agreed to public data disclosure in accordance with ethical guidelines.

Funding

This study was financially supported by the National Natural Science Foundation of China (No. 82160321), Yunnan Province Xing Dian Talent (No. XDYC-MY-2022-0052) and Kunming Medical University First-Class Discipline Team Construction Project (No. 2024XKTDPY09) and Yunnan Provincial Science and Technology Department–Kunming Medical University Joint Research Initiative (No.202501AY070001-016).

Data availability

WES data that support the findings of this study have been deposited in the Genome Sequence Archive (Genomics, Proteomics & Bioinformatics 2021) in National Genomics Data Center (Nucleic Acids Res 2022), China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences with accession number HRA011607.The datasets used or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was performed in line with the principles of the Declaration of Helsinki. The Institutional Review Board of the Affiliated Stomatological Hospital, Kunming Medical University, granted a waiver for obtaining informed consent regarding the retrospective analysis of patient imaging data. For peripheral blood collection and pedigree investigations, written informed consent was obtained from all participating individuals. Approval was granted by the Committee of Affiliated Stomatological Hospital of Kunming Medical University (Ethics Approval No. KYKQ2024MEC0088).

Consent for publication

Written consent was obtained from all patients for publication of clinical data and imaging materials.

Competing interests

The authors declare no competing interests.