A novel phex gene variant causes non-syndromic tooth agenesis

Abstract

Objectives

For a long time, phosphate regulating endopeptidase X-linked (phex) gene variants have been considered the primary cause of hypophosphatemic rickets. Here, we recruited a pedigree with non-syndromic tooth agenesis (NSTA), where all patients exhibit solely the absence of mandibular central incisors, indicating a highly conserved clinical phenotype. This observation leads us to hypothesize a potential connection between phex variants and NSTA.

Materials and methods

Whole-exome sequencing and genetic co-segregation analysis were employed to investigate the genetic basis of NSTA in this family. To further explore this hypothesis, we used phex knockdown zebrafish models and dental pulp stem cells (DPSCs) from the patient (DPSCs-MUT) and normal control (DPSCs-CON) to assess tooth development and cellular functions.

Results

We identified a novel PHEX variant (NM_000444.5 c.1763 A > T, p.N588I) that co-segregated with the NSTA phenotype. The phex knockdown zebrafish displayed a tooth loss phenotype, which closely aligns with characteristics of NSTA. Additionally, DPSCs-MUT exhibited significantly reduced mineralization and proliferation capabilities compared to DPSCs-CON, along with increased enzymatic activity. These findings suggest that PHEX variants adversely affect DPSC function. Transcriptome sequencing analysis of the DPSCs revealed significant differences in gene expression between DPSCs-MUT and DPSCs-CON. Specifically, genes linked to the cGMP-PKG signaling pathway were abnormally expressed, implicating this pathway in the potential pathogenesis of NSTA due to PHEX variants.

Conclusions

Collectively, this study offers a foundation for further research on the relationship between PHEX variants and NSTA, which could enhance the diagnosis and treatment strategies for this condition in the future.

Clinical relevance

Our research not only advances the understanding of NSTA pathogenic mechanisms but also expands the phenotypic spectrum associated with PHEX variants, highlighting its role beyond phosphate homeostasis in the development of specific dental structures.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-025-06827-0.

Introduction

Tooth agenesis (TA) represents a common developmental anomaly characterized by the congenital absence of teeth, excluding third molars, resulting from disrupted early tooth morphogenesis. Clinically, TA is classified into non-syndromic tooth agenesis (NSTA) and syndromic tooth agenesis (STA) based on the presence or absence of extra-oral manifestations [1]. STA occurs in conjunction with additional developmental abnormalities affecting other organs or tissues, often presenting as part of genetic syndromes such as ectodermal dysplasia, Down syndrome, or cleft lip and palate disorders [2]. In contrast, NSTA is characterized by isolated tooth absence without associated systemic abnormalities, typically affecting specific teeth including permanent maxillary lateral incisors or mandibular second premolars [3].

This tooth-specific pattern reflects the spatiotemporal nature of tooth development, where later-developing teeth are more vulnerable to genetic disruptions due to extended developmental exposure and distinct gene expression programs [4, 5]. The differential susceptibility is influenced by developmental timing, morphogenetic field effects, and gene expression gradients that create tooth-type-specific vulnerability windows [6, 7].

The genetic Landscape of NSTA demonstrates remarkable heterogeneity, with causative variants identified in at least 15 genes: MSX1, PAX9, AXIN2, EDA, EDAR, EDARADD, WNT10A, WNT10B, BMP2, LRP6, LTBP3, PITX2, SMOC2, KREMEN1, and GREM2 [8, 9]. Notably, several of these genes are also associated with syndromic forms of tooth agenesis, highlighting the complex genetic networks governing dental development [10]. Gene-specific variant patterns often correlate with distinct tooth agenesis phenotypes: MSX1 variants predominantly affect premolars and molars, PAX9 variants primarily cause molar agenesis, while AXIN2 variants are associated with incisor and premolar absence [11, 12].

The phosphate regulating endopeptidase X-linked (PHEX) gene has been extensively studied in the context of X-linked hypophosphatemia (XLH), where variants lead to impaired phosphate homeostasis and bone mineralization defects [13]. Interestingly, XLH patients present with a spectrum of craniofacial anomalies beyond the classic skeletal manifestations. These include dental anomalies such as abnormal tooth number, delayed tooth eruption, enlarged pulp chambers (taurodontism), increased susceptibility to dental caries and periodontal abscesses, and defective dentin and enamel mineralization [14]. Additionally, craniofacial structural abnormalities may include delayed cranial suture closure, frontal bossing, and altered facial morphology, reflecting the broader impact of phosphate dysregulation on craniofacial development [15]. These observations suggest potential roles for PHEX in dental development beyond its established function in phosphate metabolism. However, the phenotypic variability in PHEX-related manifestations ranges from subtle dental abnormalities to complete tooth absence, which may depend not only on the nature and location of underlying variants but also on individual genetic background and epigenetic factors influencing dental development [16]. Despite these clinical observations, direct evidence linking PHEX variants specifically to NSTA remains limited.

Currently, there is compelling evidence indicating that PHEX plays an autonomous and localized role in modulating the formation and quality of the extracellular matrix (ECM), thus influencing the regulation of mineralized tissue formation [17]. Notably, the mineralization-regulating SIBLING (Small Integrin-Binding Ligand N-linked Glycoprotein) protein family has emerged as substrates for PHEX enzyme activity [13]. This family comprises various ECM components such as dentin matrix protein 1 (DMP1), osteopontin (OPN), bone sialoprotein (BSP), and dentin sialophosphoprotein (DSPP), among others [18].Teeth, highly mineralized organs, result from the interaction between oral epithelial cells and mesenchymal cells, regulated by various factors. Changes in minerals, hormones, and vitamins during this process can disrupt tooth growth, development, and mineralization [19].

Here, we report a novel missense PHEX variant (c.1763 A > T, p.N588I) associated with isolated mandibular central incisor agenesis, providing the first documented case of PHEX-related NSTA and expanding our understanding of the genetic basis underlying tooth development disorders.

Materials and methods

Patients, clinical and laboratory data

The proband, a 28-year-old man with NSTA, was enrolled at the Stomatological Hospital, Southern Medical University, Guangzhou, China. We also recruited all his family members, none of whom were diagnosed with any other craniofacial anomalies or syndromes. A comprehensive maternal history was obtained, including information about pregnancy complications, medication use during pregnancy, nutritional supplementation, infections, and other environmental exposures during the proband’s intrauterine development. Data on variables such as sex, age, height, weight, bone phenotypes, dental phenotypes, and available biochemical indicators were collected at their initial visit. Peripheral blood samples were collected from the proband and their available family members. The study was conducted following the Declaration of Helsinki and national guidelines. Informed consent was obtained from all participants in accordance with the regulations of the Committee of the Stomatological Hospital, Southern Medical University (No. 2022-YW-39-002).

Mutation analyses

Genomic DNA was extracted using the classical phenol-chloroform method. Samples from selected individuals (II3, III1, and II5) underwent whole exon sequencing (WES). Exon targets were captured using the Agilent SureSelect V4 Human All Exon 50 Mb kit (Agilent Technologies, Santa Clara, California, USA) and sequenced on a HiSeq 2000 sequencing system with 100 bp paired-end reads (Illumina, San Diego, California, USA). Identified variants were filtered based on minor allele frequency (MAF) in population databases such as 1000 Genomes, dbSNP, ExAC Browser, and Human Genetic Variation Database (absent or < 0.01), as well as functional prediction (missense, nonsense, indel, or splicing). Candidate pathogenic variants and segregation within family members were determined via Sanger sequencing with polymerase chain reaction (PCR) amplification, following previously established protocols. The primers used for PCR amplification of PHEX (NM_000444.5) were as follows: Forward, 5′-ATGGTGCTATAGGAGTAATTGTCGG- 3′; Reverse, 5′-TCATCGTTGGAGCAGAGCTT-3’.

Bioinformatics analysis

To validate the conservation of amino acid substitutions across species evolution, we aligned typical protein sequences from various species using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/) to compare mutated positions with conserved domains. PolyPhen 2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), SIFT (http://sift.bii.a-star.edu.sg/), Mutation Taster (https://www.mutationtaster.org/), and Provean (http://sift.jcvi.org/index.php) were utilized to predict the variant effects of candidate variants. Additionally, three-dimensional (3D) protein structures of both wild-type (WT) and mutant (MUT) PHEX proteins were forecasted using I-TASSER (https://zhanggroup.org//I-TASSER/). The PyMOL Viewer software facilitated the generation of mutant PHEX protein models, enabling visualization of the effects of altered residues on protein structure.

Zebrafish studies

All animal experiments were approved by the Experimental Animai Ethics Committee of Hunter Biotechnology, Inc (License number: IACUC-2025-11672-01) and conducted according to the Southern Medical University Health Guidelines for the Care and Use of Laboratory Animals. All study methods and animal handling were in accordance with the ARRIVE guideline. Zebrafish used in this study were purchased from Hunter Biotechnology, Inc (Guangzhou, China) and bred in the Central Laboratory, Southern Medical University. Adult AB zebrafish were kept at 28.5°C under a 14 h light/10 h dark cycle. Adult fish were fed twice daily with commercial zebrafish pellets (Zeigler Bros., Inc.) supplemented with live brine shrimp (Artemia nauplii) three times per week. Embryos and larvae up to 5 days post-fertilization (dpf) were not fed, relying on their yolk sac for nutrition. From 6 dpf onwards, larvae were fed with paramecia and commercially available larval food (ZM-000, Zebrafish Management Ltd.) twice daily until they reached juvenile stage. Embryos were obtained from natural spawns. Visualization of larval teeth was achieved through alizarin red staining (ARS), following the method outlined by Wise and Stock. A single antisense morpholino oligonucleotide (MO) was custom-designed to selectively inhibit the translation of zebrafish phex mRNA (Gene Tools, LLC). The antisense MO (phex: 5’-GGCTTAGGATCTCACCTTGATGATG-3’) introduces a premature termination codon in-frame, potentially triggering nonsense-mediated decay of the transcript. Additionally, the standard control MO (5’-ATAAAAACACTGCCTTATCCTCTCA-3’) was also included in the composition. The MOs, or control MOs, were administered via injection into the egg yolks of 1-2-cell-stage Oregon AB embryos. These embryos were then allowed to develop to the desired stage for subsequent analysis. Teeth structures were observed using ARS, following previously established protocols, at 6 days postfertilization (dpf). Quantitative real-time polymerase chain reaction (qRT-PCR) was performed to confirm the knockdown effect caused by phex MO. The primers used for qRT-PCR amplification of phex (NM_001089349.2) were as follows: Forward, 5′-CCTGGACTTTGCTCGGGTTA-3′; Reverse, 5′-AACAAAGAGCCTACCAGCGG-3’. The zebrafish were anesthetized and euthanized using MS-222 (tricaine methanesulfonate). The fish were immersed in a 0.1% MS-222 solution until they were unconscious, and then euthanized by rapid decapitation.

Assessment of cell proliferation and odontoblastic differentiation

The mandibular third molars were extracted and collected from both the NSTA patient and his healthy sister. This was done after obtaining informed consent from the participants. Isolation of hDPSCs was performed as described elsewhere [20]. hDPSCs were isolated from the proband and his healthy sister to ensure identical genetic background for comparative analysis. All experiments were performed using cells between passages 3–5 to maintain cellular characteristics and ensure reproducibility. Each experimental condition was tested in triplicate to ensure technical reproducibility. The use of a familial control minimizes genetic background variations while allowing specific assessment of PHEX variant effects on cellular function. Study limitations include the use of a single familial control due to limited tissue availability and the sex difference between the proband (male) and control (female), which could potentially influence PHEX expression patterns given the X-linked nature of the gene. Future studies will incorporate additional age- and sex-matched control and stable cell line models for comprehensive functional characterization. For odontoblastic differentiation experiments, the cells were cultured in an odontogenic medium for a total of 21 days. This medium consisted of DMEM, 10% FBS, 50 µg/mL ascorbic acid (Sigma-Aldrich, St. Louis, MO, USA), 5 mM β-glycerophosphate (Sigma-Aldrich), and 10 nM dexamethasone (Sigma-Aldrich).

The impact of PHEX variant on the proliferation of hDPSCs was assessed through the Cell Counting Kit-8 (CCK-8; Dojindo, Tokyo, Japan) assay and flow cytometric analysis, following established protocols. Additionally, the influence of PHEX variant on the odontoblastic differentiation of hDPSCs was examined using Alizarin Red S (ARS) and alkaline phosphatase (ALP) staining, with methodologies described protocols (Cyagen, Guangzhou).

PHEX enzyme activity assay in hDPSCs

DPSCs from proband and healthy control were cultured and passaged routinely. Total cellular proteins were extracted, quantified using the BCA assay kit (Beyotime Biotech), and equal amounts of total protein were incubated with the substrate peptide (Abz-GFSDYK(Dnp)-OH, synthesized by Sangon Biotech) at 37 °C. Fluorescence intensity during the hydrolysis reaction at 20, 40, and 60 min was measured using a fluorescence spectrophotometer (excitation wavelength λex = 320 nm, emission wavelength λem = 405 nm) to characterize enzyme activity.

Transcriptome sequencing was conducted on hDPSCs

Primary DPSCs were obtained from proband (III1), and control cells were obtained from his age-matched sister (III2), both of whom underwent standard surgical extraction of wisdom teeth. Total RNA from cultured DPSCs was extracted using TRIzol (Invitrogen Corporation), followed by library preparation with the Trio RNA-Seq Library Preparation Kit (Nugen Technologies, USA) and sequencing on an Illumina HiSeq platform. Differentially expressed transcripts were identified using the DEGseq package in BioConductor (https://bioconductor.org/packages/release/bioc/html/DESeq2.html). Transcripts meeting criteria of |log2 fold change (FC)| ≥ 1 and p ≤ 0.05 were visualized via volcano plots and heatmap. Enrichment analysis of Kyoto Encyclopedia of Genes and Genomes (KEGG) terms was performed using the Cluster Profiler package (4.0.5) in R.

Western blot analysis

Cells were harvested and then lysed in RIPA buffer (Beyotime, Nanjing, China) supplemented with protease inhibitors. The cell lysates were cleared by centrifugation at 12,000 × g for 10 min at 4 °C. Total-protein samples were separated using SDS-PAGE in a 10% gel and then transferred to a PVDF membrane (Amersham, Little Chalfont, UK) at 200 mA for 2–3 h. The membranes were blocked for 1 h with 5% skim milk and incubated overnight at 4 °C with primary antibodies: anti-PHEX (1:1000; Biorbyt), anti-DSPP (1:500; Affinity), anti-DMP1 (1:500; Affinity), and anti-GAPDH (1:3000; Affinity) antibodies. The next day, the membranes were incubated for 1 h at 37 °C with the corresponding secondary antibodies (Proteintech, China), and the immunoreactive proteins were visualized with the ECL Kit (Beyotime Biotech, Shanghai, China) according to the manufacturer’s instructions. For Western blot analysis, each experiment was repeated three times with consistent results, and representative blots are shown in the figures.

Statistical analysis

All experiments were performed with at least three independent biological replicates (n ≥ 3). Statistical analysis Results are expressed as the mean ± SEM. Statistical analyses were conducted using GraphPad Prism (version 9, GraphPad Software). For normally distributed datasets, unpaired Student’s t-test was used for two-group comparisons, with one-way ANOVA followed by post hoc analysis used for multiple group comparisons where applicable. Specific test details and sample sizes are provided in each figure legend. A significance level of p < 0.05 was used.

Results

Clinical characterization of patients with NSTA

We identified a large Chinese family with autosomal dominant non-syndromic tooth agenesis (NSTA) based on intraoral diagnostic features. Family 1 comprises 14 individuals spanning three generations, as depicted in the pedigree (Fig. 1A). Intraoral clinical examination revealed that all affected individuals presented with congenital absence of mandibular central incisors without compensatory tooth movement (Fig. 1C). Periodontal examination showed healthy gingival tissues with no signs of periodontitis or significant alveolar bone loss in the affected regions. The remaining dentition exhibited normal morphology and positioning. Notably, Individual I1 has multiple missing teeth due to age, and the proband III1 has undergone orthodontic treatment. The clinical phenotype of NSTA patients in Family 1 is typically highly conservative. The sporadic case (Family 2) presented similar intraoral features with healthy periodontal status (Fig. 1B and D). Laboratory examination data revealed no abnormalities in phosphate, calcium, and other indicators related to hypophosphatemia in the patient (Table 1).

Fig. 1.

Pedigree and panoramic radiograph of family 1 (A) and family 2 (B). Males are marked as squares and females as circles. An arrow indicates the proband, and the black symbols indicate the affected individuals. (C-D) Panoramic radiograph of the patients shows all individuals had a congenital absence of mandibular central incisors. An asterisk marks the congenital missing tooth. Notably, Individual I1 has multiple missing teeth due to age, and the proband III1 has undergone orthodontic treatment

Table 1.

Laboratory examination data of the patients

Subject	III1	II3	II1 (Family1)	II1 (sporadic case)	Reference range
Sex	male	male	female	male	≥ 130
Height (cm)	175	173	162	165
Bow legs	No	No	No	No	-
Serum phosphate (mmol/L)	1.29	1.31	1.30	1.24	0.70 ~ 1.40
Serum calcium (mmol/L)	2.30	2.21	2.33	2.31	2.20 ~ 2.60
Alkaline phosphatase (U/L)	119	121	129	101	30 ~ 130
Parathyroid hormone (ng/L)	42.4	43.9	54.5	40.0	15 ~ 70
25(OH)vitamin D3 (µg/L)	65	67	88	72	30 ~ 100
Urinary phosphate (mmol/24 h)	25.7	32.0	31.2	26.3	7.0 ~ 42.0
Urinary calcium (mmol/24 h)	4.9	4.9	5.6	4.7	2.5 ~ 7.5
FGF-23 (pg/mL)	41.2	48.0	33.6	39.7	33.9 ~ 51.8

WES identifies variants in phex underlying NSTA

We obtained peripheral blood samples from family members (III1, II3, II5 in Family 1) for WES. Through bioinformatics analysis, we screened for homozygous, compound heterozygous variants, and X-linked gene variants, with a variant frequency of less than 1% in the patients. This analysis involved referencing variant databases such as ExAC, gnomAD, 1000 Genomes, dbSNP, and ESP6500. We scrutinized for various variant types including missense variants, frameshift variants, nonsense variants, variants at splice sites, and others (Fig. 2A). WES failed to detect above known NSTA-related genes, indicating the possibility of potential novel pathogenic genes associated with NSTA. Results revealed a missense variant in the PHEX gene (NM_000444.5 c.1763 A > T, p.N588I), which is consistent with an X-linked dominant inheritance pattern and segregated within the patient’s family (Fig. 2B). This variant site was verified in a sporadic case of NSTA, where the patient also exhibited the absence of mandibular central incisor teeth, adds another layer of intrigue to this finding. The missense variant in the PHEX gene was located in the 17th exon of the PHEX genomic sequence (Fig. 2C), situated at a crucial position within the PHEX domain. This variant entailed the substitution of nucleotide A by T at position 1763 in PHEX, leading to a change in the mRNA transcribed from the gene. Consequently, the acidic asparagine (Asn, abbreviated as N) at position 588 of the translated PHEX was replaced by nonpolar, hydrophobic isoleucine (Ile, abbreviated as I). This variant is predicted to significantly alter the structure of the translated PHEX protein, thereby affecting PHEX enzyme activity and severely impacting its function.

Fig. 2.

A-B Whole exome sequencing and Sanger sequencing revealed a novel PHEX variant (NM_000444, c.1763 A > T, N588I) identified in members with NSTA. C Location of the identified variant (marked in red) in schematic diagram of PHEX

Using the ClusterX online tool for the conservation analysis of the selected variant c.1763 A > T, results revealed that the amino acid at this position is highly conserved across multiple species (Fig. 3A). High-resolution melting curve analysis for this variant in population screening showed that c.1763 A > T was not found among 200 normal individuals (Fig. 3B). Predictions of the harmfulness of this variant using Polyphen2, SIFT, Mutation Taster, and Provean software all demonstrate it as a deleterious variant (Table 2). SOPMA showed that the N588I variant leads to changes in the protein’s secondary structure, including α-helix, β-sheet, irregular coils, and extended chains. I-TASSER showed changes in tertiary structures of both wild-type and mutant PHEX proteins, with alterations concentrated around the variant site (Fig. 3C-D).

Fig. 3.

A Conservation analysis of the variation site among different species. B High-resolution melting (HRM) analysis of exon 17 of PHEX in 200 DNA specimens from random samples. C-D Protein prediction of the secondary and tertiary structure of PHEX

Table 2.

Predictions of the harmfulness of the c.1763 a > t variant

Variants	AA change	PolyPhen2	SIFT	Mutation Taster	Provean
c.1763 A > T	N588I	Damaging (0.939)	Damaging (0.000)	Disease-causing (0.999)	Deleterious (−4.411)

Phex knockdown zebrafish exhibit tooth loss phenotype

We used RT-PCR to analyze phex gene expression in zebrafish, revealing expression from 1 to 6 days post fertilization (dpf), peaking at 3 dpf (Fig. 4A). To suppress phex expression and translation, morpholinos (MOs) were designed and synthesized by GeneTools in the United States. After microinjection, we monitored tooth development in zebrafish eggs. Morpholino concentrations used were MO1: 2.1 µg/µl and MO2: 4.2 µg/µl. qRT-PCR at 3 dpf demonstrated effective suppression of phex gene expression by both MO1 and MO2 (Fig. 4B). MO-mediated knockdown of phex in zebrafish embryos achieved approximately 45% reduction in phex mRNA expression compared to control embryos (Fig. 4B). Among 120 injected embryos, we observed a dose-dependent phenotypic spectrum: 67% (80/120) of embryos showed the primary phenotype of tooth absence at 5 days post-fertilization (dpf), while 23% (28/120) exhibited additional skeletal abnormalities including spinal curvature and mild body axis defects (Fig. 4D). However, although MO2 showed higher efficacy, it also led to an increased rate of malformations and mortality (Fig. 4C). Consequently, we selected MO1 for subsequent injections due to its lower adverse effects. At 6 dpf, wild-type zebrafish exhibited a total of six teeth, three on each side. Following injection into zebrafish eggs of control groups and MO1, specimens were fixed in paraformaldehyde after 6 days, dehydrated, stained with Alizarin Red, and cleared. Results revealed that the control group had all six teeth, whereas suppression of phex expression resulted in tooth loss phenotypes (Fig. 4D).

Fig. 4.

A Phex expression at different stages post-fertilization in zebrafish (n = 20 embryos per time point, 3 independent experiments). B Validation of Phex expression inhibition efficiency (n = 15 embryos per group, 3 independent experiments). C Statistics depicting mortality and malformation rates at varying concentrations of MO (n = 120 embryos total, pooled from 3 independent experiments with n = 40 per experiment). D Microinjection of zebrafish eggs and observation of tooth development at 6 days post-fertilization (dpf) via Alizarin Red staining (n = 30 embryos per group, 3 independent experiments). Scale bars: 250 μm (embryo images), 100 μm (tooth staining images). Data represent mean ± SEM. Statistical analysis performed using Student’s t-test for pairwise comparisons and one-way ANOVA for multiple group comparisons. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001 vs. control group

Enhanced PHEX activity disrupts DPSC mineralization and proliferation capabilities

Our experimental findings demonstrate a dynamic upregulation of PHEX during the mineralization induction of human DPSCs (Fig. 5A-B), indicating a significant association between PHEX and DPSCs’ mineralization. Moreover, we cultured primary DPSCs from extracted wisdom teeth of NSTA patients, using DPSCs from their healthy sister as control, to investigate changes in their biological characteristics. The results revealed reduced mineralization and proliferation abilities in DPSCs from NSTA patients (Fig. 5C-E). These findings demonstrate that PHEX variants cause deregulation of DPSCs’ proliferation and mineralization, though specific mechanisms warrant further investigation.

Fig. 5.

The qRT-PCR A and Western blot B results demonstrate upregulation of PHEX expression during mineralization induction in DPSCs (qRT-PCR: n = 3 per time point, 4 independent experiments; Western blot: n = 3 per time point, 3 independent experiments). C The results of ARS and ALP staining reveal reduced mineralization capability in mutant cells (ARS: n = 6 per group, 4 independent experiments; ALP: n = 6 per group, 3 independent experiments). Cell cycle analysis D and CCK-8 assay E demonstrate reduced cell proliferation capability after mutation (Cell cycle: n = 3 per group, 3 independent experiments; CCK-8: n = 6 per group, 4 independent experiments). F PHEX enzyme activity assay showed enhanced enzymatic activity in the MUT group (n = 4 per group, 3 independent experiments). G Western blot results demonstrate that after mutation, the expression of PHEX enzyme substrates DMP1 and OPN is downregulated, while the expression of PHEX itself remains unchanged (n = 3 per group, 3 independent experiments). Data represent mean ± SEM. Statistical comparisons performed using Student’s t-test for two-group comparisons and one-way ANOVA followed by Tukey’s post hoc test for multiple comparisons. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001 vs. control group (DPSCs-WT)

In this study, a PHEX variant replaced the hydrophilic amino acid asparagine (Asn) at position 588 with the hydrophobic amino acid isoleucine (Ile), which may alter enzyme function. Our findings corroborate this notion, demonstrating an increase in PHEX enzyme activity following the variant (Fig. 5F). Additionally, we conducted a assessment of the protein expression levels of PHEX enzyme substrates, DMP1 and DSPP, which revealed diminished expression in the variant group compared to the control group (Fig. 5G). Based on these observations and the evaluation of PHEX enzyme activity, our data demonstrate that the PHEX point variant at position 588, substituting a hydrophilic amino acid with a hydrophobic one, amplifies PHEX enzyme activity, consequently augmenting the hydrolysis of its substrates. This, in turn, leads to a downregulation of substrate protein expression, thereby impeding cell proliferation and mineralization.

Analysis of transcriptome sequencing

We conducted transcriptome sequencing on DPSCs from individuals with PHEX variants and healthy control to study its impact on tooth development. The RNA-Seq analysis identified 114 differential transcripts (|log2FC|≥ 1 and p < 0.05), comprising 52 downregulated and 62 upregulated genes (Fig. 6A-B). Gene expression profiles and changes relative to control are summarized in Supplementary Table 1, showing average counts per million mapped reads (CPM) for each gene, as well as expression FC and statistical significance false discovery rate (FDR). A part of such genes, including FAM20A, COL14A1, and FBLN2, was down-regulated in DPSCs compared with control; whereas another part was up-regulated, suggesting that the special NSTA in our study was attributed to a corresponding part of mineralization factors, including cGMP-PKG signaling pathway. The enriched GO-BP pathways included biomineral tissue development, biomineralization, cell proliferation, collagen-containing ECM, and ECM structural constituent, among others (Fig. 6C-D).

Fig. 6.

A Volcano plot of differentially expressed genes (n = 3 technical replicates per group, single RNA-seq experiment). Green dots represent down-regulated genes, and red dots represent up-regulated genes (|log2FC|≥ 1 and adjusted p ≤ 0.05). B Heatmap showing the expression levels of transcripts in samples from the DPSCs-MUT and DPSCs-WT (n = 3 technical replicates per group). Red color refers to upregulation, and blue color refers to downregulation of gene transcription. C KEGG pathway analysis ranked the top 10 KEGG pathways (|log2FC|≥ 1 and adjusted p ≤ 0.05, based on 3 technical replicates per group). D Go results of biological process, cellular component, and molecular function (enrichment analysis based on differentially expressed genes with adjusted p ≤ 0.05, using 3 technical replicates per group). RNA sequencing was performed using Illumina platform with > 20 million clean reads per sample. Differential expression analysis was conducted using DESeq2 with Benjamini-Hochberg correction for multiple testing. Statistical significance: ^*adjusted p < 0.05, ^**adjusted p < 0.01, ^***adjusted p < 0.001

Discussion

This study identifies and characterizes the first PHEX gene variant specifically associated with non-syndromic tooth agenesis, expanding the known genetic spectrum of NSTA. The novel c.1763 A > T (p.N588I) variant presents with a distinctive phenotype of isolated mandibular central incisor agenesis without the typical skeletal or systemic features of X-linked hypophosphatemia.

The absence of classical XLH features in our patients raises important questions about genotype-phenotype correlations in PHEX-related disorders. Previous studies have documented significant phenotypic heterogeneity in PHEX variants, with some patients presenting mild skeletal involvement despite truncating variants, while others with missense variants show severe manifestations. Goljanek-Whysall et al. (2018) identified a novel PHEX variant (c.2158G > T; p.Ala720Ser) in an adult female with hypophosphatemic osteomalacia without childhood rickets, demonstrating that PHEX mutations can manifest with delayed onset and without severe skeletal abnormalities in early life [21]. Similarly, Zheng et al. (2020) found no consistent genotype-phenotype correlation in XLHR, with even truncating variants sometimes resulting in mild skeletal involvement, highlighting the heterogeneous clinical presentation of PHEX variants [22]. The tissue-specific expression patterns and potential compensatory mechanisms may explain why our patients developed isolated dental abnormalities without systemic phosphate metabolism dysfunction.

However, several key features distinguish our patients from reported cases of mild XLHR. Most importantly, the biochemical profile in our patients shows completely normal phosphate homeostasis, with normal serum phosphate levels, normal FGF23 concentrations (within reference ranges), and absence of renal phosphate wasting. Patient II1 demonstrated notably lower FGF23 levels (30% reduction compared to patient II3) alongside elevated vitamin D and PTH levels, though all parameters remained within normal ranges. This contrasts sharply with even the mildest reported XLHR cases, which invariably demonstrate some degree of hypophosphatemia or elevated FGF23 levels. For instance, the patient described by Goljanek-Whysall et al. presented with hypophosphatemic osteomalacia and raised serum FGF23 concentrations, representing delayed-onset but otherwise typical XLHR biochemistry [21].

Our transcriptomic analysis revealed differential expression of genes crucial for dental development and mineralization. The enrichment of cGMP-PKG signaling pathway, along with altered expression of FAM20A, COL14A1, and FBLN2, provides mechanistic insights into PHEX-mediated tooth development. FAM20A encodes a pseudokinase essential for enamel formation, while COL14A1 and FBLN2 are involved in extracellular matrix organization and cell adhesion processes critical for tooth morphogenesis [23]. The downregulation of these genes in mutant cells suggests that PHEX dysfunction disrupts multiple pathways converging on dental tissue formation and mineralization.

The functional studies demonstrated reduced mineralization capacity and proliferation in dental pulp cells harboring the PHEX variant, consistent with impaired odontoblast differentiation and dentin formation. These findings align with previous observations of defective mineralization in XLH dental pulp cell cultures, though our patients lacked the systemic manifestations typically associated with XLH [24].

Importantly, this study addresses why PHEX has not previously been recognized among genes causing tooth agenesis. The rarity of isolated dental phenotypes in PHEX variants, combined with the predominant focus on skeletal manifestations in XLH research, may have obscured the connection between PHEX variants and NSTA. Additionally, the mild biochemical abnormalities in our patients might not trigger routine phosphate metabolism screening, potentially leading to underdiagnosis of PHEX-related dental anomalies [25].

The absence of systemic calcium and phosphate metabolism abnormalities in patients with the PHEX c.1763 A > T (p.N588I) mutation, despite its location within the zinc finger catalytic domain, can be explained by the variant’s unique functional consequences and tissue-specific effects. Unlike classical PHEX loss-of-function variants that cause X-linked hypophosphatemic rickets (XLH), our functional studies revealed that the N588I substitution enhances rather than diminishes PHEX enzymatic activity [26, 27]. This gain-of-function effect suggests that the variant alters enzyme regulation or substrate specificity rather than abolishing catalytic function entirely. Previous studies have demonstrated that different PHEX variants can have varying effects on enzyme activity, with some variants affecting protein stability, subcellular localization, or substrate binding affinity rather than catalytic capacity per se [28, 29].

The tooth-specific phenotype observed with this variant likely reflects the distinct roles of PHEX in dental versus skeletal development. During odontogenesis, PHEX is highly expressed in odontoblasts and ameloblasts, where it regulates extracellular matrix mineralization through pathways that may be mechanistically different from its role in renal phosphate handling and bone metabolism [30, 31]. The enhanced enzymatic activity resulting from the N588I mutation may specifically disrupt the precise balance of extracellular matrix proteins required for proper tooth formation, while the pathways governing systemic phosphate homeostasis remain intact. This tissue-specific functional divergence is supported by the differential expression patterns and substrate preferences of PHEX in various tissues, suggesting that the same enzyme can have distinct physiological roles depending on the cellular context and local protein environment [32, 33].

While our study design using the proband’s sister as a familial control provides optimal genetic background matching, we acknowledge the potential influence of sex on PHEX expression and related pathways, as recently demonstrated by [34]. The X-linked nature of the PHEX gene introduces complexity regarding sex-specific expression patterns, where female carriers may show variable expression due to X-chromosome inactivation, while hemizygous males typically exhibit more consistent expression levels. However, several factors support the validity of our comparative analysis despite the sex difference between our proband and control. First, the biochemical profiles of both individuals show normal phosphate homeostasis (serum phosphate, FGF23, and urinary phosphate within normal ranges), suggesting that any potential sex-related differences in PHEX expression do not translate to clinically significant biochemical variations in our specific case. Second, our focus on the functional consequences of a specific missense variant (c.1763 A > T, p.N588I) rather than general expression differences allows us to assess mutation-specific effects that would be consistent regardless of baseline expression variations. Third, the consistent tooth agenesis phenotype observed across affected family members of both sexes supports that this specific variant effect transcends sex-related expression differences.

Our findings suggest that PHEX should be considered in the genetic evaluation of patients with specific patterns of tooth agenesis, particularly when involving mandibular incisors. The identification of PHEX as a novel NSTA gene not only expands the genetic landscape of tooth agenesis but also provides new insights into the multifaceted roles of phosphate metabolism in dental development.

Future studies examining larger cohorts of NSTA patients and investigating the functional consequences of different PHEX variants will be essential for establishing comprehensive genotype-phenotype correlations and developing targeted therapeutic approaches for PHEX-related dental anomalies.

Conclusion

We report the first documented case of a PHEX gene variant causing isolated non-syndromic tooth agenesis, specifically affecting mandibular central incisors. This novel variant (c.1763 A > T, p.N588I) expands the genetic spectrum of NSTA and reveals previously unrecognized roles for PHEX in dental development. The absence of typical XLH features suggests tissue-specific effects and potential compensatory mechanisms. These findings highlight the importance of considering PHEX variants in genetic screening for tooth agenesis and provide new insights into the complex molecular networks governing dental development.

Abbreviations

NSTA

Non-syndromic tooth agenesis

PHEX

Phosphate regulating endopeptidase X-linked

DPSCs

Dental pulp stem cells

XLH

X-linked hypophosphatemia

WES

Whole-exome sequencing

MAF

Minor allele frequency

dbSNP

Database of Single Nucleotide Polymorphisms

PCR

Polymerase chain reaction

qRT

PCR-Quantitative real-time polymerase chain reaction

ARS

Alizarin red staining

dpf

Days post fertilization

Authors’ contributions

Thg study was conceived and designed by P.Y., H.B., and W.H. who contributed equally. Clinical data was collected and gene function analysis were performed by P.Y., T.S., H.B., L.T., and P.Y. conducted the statistical analysis and writted manuscript. M.D. and X.F. supervised the project. All authors reviewed the manuscript.

Funding

This study was supported by funding from the Guangdong Basic and Applied Basic Research Foundation and National Natural Science Foundation of China. This work was supported by Guangdong Basic and Applied Basic Research Foundation (2023A1515110129), Science research cultivation program of stomatological hospital, Southern medical university (PY2022007), National Natural Science Foundation of China (81970930, 82201133); National Key S&T Special Projects (2021YFC100530, 2022YFC2703303), Guangzhou municipal science and technology project (SL2023A04J02637), 2022A1515110535,Guangdong Basic and Applied Basic Research Foundation (2022A1515110535) and 2023SZX013,Maoming City Special Science and Technology Fund Project, 2023SZX013.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The Ethics Committee of Southern Medical University (No.2022-YW-39-002) approved all studies. The procedures used in this study adhere to the tenets of the Declaration of Helsinki. All participants provided written informed consent. All animal experiments were approved by the Experimental Animai Ethics Committee of Hunter Biotechnology, Inc (License number: IACUC-2025-11672-01) and conducted according to the Southern Medical University Health Guidelines for the Care and Use of Laboratory Animals. All study methods and animal handling were in accordance with the ARRIVE guideline.

Consent for publication

A written consent for publication was obtained from the patients to publish all clinical details and any accompanying images.

Competing interests

The authors declare no competing interests.