WNT10A-SMOC2-LRP4 network affects permanent tooth development via potential tooth-bone interaction

Abstract

Background

Permanent tooth loss is a key feature of non-syndromic tooth agenesis (NSTA). However, whether this is influenced by the mandibular environment remains unclear. We report a case of simultaneous variations in the tooth agenesis-associated genes WNT10A and SMOC2 and the bone formation-associated gene LRP4, exploring the potential impact of the mandibular environment on tooth agenesis.

Methods

A 19-year-old patient with NSTA underwent clinical examination and genetic analysis using whole-exome sequencing and Sanger sequencing. We reanalyzed single-cell RNA-seq datasets from human molars and surrounding tissues and performed immunofluorescent staining on key cell subsets.

Results

The patient exhibited significant retention of primary teeth and loss of nine permanent teeth, with cone-beam computed tomography(CBCT) showing increased bone density and structural abnormalities. Genetic analysis revealed a novel variation in LRP4 (c.1831 C > T, p.R611C), and variations in SMOC2 (c.1138 C > T, p.R380C) and WNT10A (c.354T > G, p.Y118*). Protein structure analysis revealed differences in the RMSD and hydrogen bonds between mutated LRP4 and SMOC2. Additionally, LRP4-positive alveolar bone cells send Fibroblast Growth Factor (FGF) signals to WNT10A-positive and SMOC2-positive dental pulp cells.

Conclusions

Based on a case of NSTA with SMOC2 and novel LRP4 variations linked to tooth agenesis and bone development, we revealed that LRP4 + alveolar bone cells may influence tooth development through FGF signaling, thereby regulating WNT10A + and SMOC2 + dental pulp cells.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-025-06451-y.

Background

Non-syndromic tooth agenesis (NSTA) is a common genetic disorder of tooth development and is characterized by abnormalities in tooth number, size, and shape. Patients with NSTA often experience permanent tooth loss with retained deciduous teeth due to the absence of permanent tooth germs in the underlying bone [1, 2]. WNT10A and SMOC2 have been identified as tooth agenesis-associated genes that impair the tooth development process [3–8]. The surrounding alveolar and jaw bones anchor the teeth and develop along with permanent tooth replacement for 6–20 years [9]. The process of bone development, known as osteogenesis, greatly impacts tooth formation by establishing a solid foundation crucial for tooth eruption, alignment, and overall dental health. Mutations affecting mandible development also impair tooth formation [10]. For example, MSX1-/- mice exhibit tooth and mandibular abnormalities. Moreover, bone resorption is commonly observed after tooth extraction. Therefore, tooth and bone interact to guide their developmental process. Therefore, impairment of these bones may contribute to tooth agenesis. However, clinical evidence remains unclear. Previous studies have shown that the LRP4 gene plays a crucial role in bone formation, and is associated with Sclerosteosis 2. However, whether this influenced permanent tooth development in this patient remains unknown [11].

Here, we present the case of a 19-year-old female patient with NSTA and increased maxillary and mandibular bone densities. She lacked nine permanent teeth and had increased jaw bone density. Whole Exome Sequencing (WES) identified variations in LRP4, WNT10A, and SMOC2. Protein structure analysis revealed structural differences between the mutated and wild-type proteins. To explore the functions of these genes, we re-analyzed single-cell RNA sequencing (scRNA-seq) data from developing human molar and periodontal tissues and revealed that WNT10A and SMOC2 marked distinct dental pulp subsets, while LRP4 was expressed in alveolar bone cells. Cell-cell communication analysis showed a potential interaction between the two cell types, indicating the involvement of alveolar bone cells in permanent tooth development.

Materials and methods

Clinical data collection

A 19-year-old female patient was referred to the Center of Stomatology, Xiangya Hospital, Central South University (China), with a complaint of deciduous tooth retention. With informed consent from the guardian and patient, we collected the medical history, performed oral and general physical examinations, recorded oral photos and X-ray examinations, and collected peripheral venous blood samples.

Measurement of hounsfield unit (HU) and analysis of alveolar bone

CBCT imaging was performed to measure the HU values of the cancellous bone in the alveolar bone using KaVo 3D Exam Vision. The control group consisted of 10 randomly selected healthy women, aged 19–20 years, without known bone pathologies, who had previously undergone CBCT imaging for unrelated purposes at Xiangya Hospital. The mean HU values of the control group were compared with the HU values of the patient using a one-sample t-test.

Whole-exome sequencing (WES) and and analysis of sequencing results

DNA extraction and WES were performed by Genergy Bio-Technology (Shanghai) Co., Ltd. Variants with a minor allele frequency (MAF) greater than 0.01 in public databases of East Asian populations, including Exome Aggregation Consortium (ExAC; http://exac.broadinstitute.org/), the 1000 Genomes Project (1000G; https://www.internationalgenome.org/) and Genome Aggregation Database (gnomAD; https://gnomad.broadinstitute.org/), were excluded. Synonymous variants and variations within repetitive sequences were removed, leaving only nonsynonymous single-nucleotide variants (SNVs) and insertions/deletions (indels) in protein-coding regions for further analysis. We filtered variations in genes from a list of 55 genes associated with tooth agenesis (TA) and 364 genes linked to genetic skeletal disorders [12, 13]. Pathogenicity predictions were made using tools including Protein Variation Effect Analyzer (PROVEAN, http://provean.jcvi.org/index.php), MutationTaster (https://www.mutationtaster.org/) and polymorphism phenotyping (PolyPhen-2, http://genetics.bwh.harvard.edu/pph2/). Variants were further evaluated according to the American College of Medical Genetics and Genomics (ACMG) guidelines. Variants meeting these criteria and associated with TA or increased bone mineral density were filtered for further analysis and discussion.

Sanger sequencing analysis

LRP4, WNT10A and SMOC2 fragments were sequenced by Sanger sequencing to verify the WES results. Primers of the variations were specifically designed (Table 1). The results were compared with the reference sequences for each gene (LRP4, NM_002334; WNT10A, NM_025216; SMOC2, NM_001166412) (UCSC, http://genome.ucsc.edu/) to verify the results of WES.

Table 1.

Gene variations and primer sequences

Variation	Forward primer	Reverse Primer
LRP4 (NM_002334:exon14:c.C1831T: p.R611C)	5’-ATGGATAAAAGCGCCCTACC-3’	5’-AACCTCTACCAAGGCTCCTG-3’
WNT10A (NM_025216:exon2:c.T354G: p.Y118*)	5’-TGCCAACACAGTGTGCCTAA-3’	5’-CAGCTTGAGGCAGTGGGTTA-3’
SMOC2 (NM_001166412:exon11:c.C1138T:p.R380C)	5’-TTCCTATGACATGAAGGATGTGT-3’	5’-GCAAACAGACGCCCAACAAT-3’

Prediction of the secondary and tertiary structure of proteins

The NCBI database (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) was used to obtain the original amino acid sequences of LRP4, WNT10A and SMOC2. SOPMA (https://npsa-prabi.ibcp.fr/NPSA/npsa_sopma.html) was used to analyze the differences in the secondary structures of the wild-type and mutant proteins. AlphaFold 2 (https://www.alphafold.ebi.ac=/) was used to predict the tertiary structures of the wild-type and mutated proteins. Tertiary structures of the wild-type and mutated proteins were visually compared using PyMOL (PyMOL Molecular Graphics System; DeLano Scientific). Structural similarity is often expressed as the root-mean-square deviation (RMSD) value, which measures the difference between two sets of proteins [14]. The protein structures were aligned using PyMOL, and the alignment precision was assessed using the RMSD value. Additionally, variations in hydrogen bonding between the native and mutant structures were visualized using PyMOL, highlighting the differences in the interaction patterns.

Analysis of conserved amino acid residues

The ConSurf server (https://consurf.tau.ac.il), accessed on September 12, 2024, was used to determine the evolutionary conservation of amino acid positions [15]. The FASTA sequences of LRP4, SMOC2, and WNT10A proteins were submitted to the ConSurf server. The resulting color scheme assigns conservation scores ranging from 1 to 9, where a score of 9 indicates the most conserved amino acid, whereas a score of 1 indicates a highly variable amino acid.

Single-cell data processing

For the scRNA-seq data analyzed in this study, we downloaded raw count matrices from GSE146123 [16] and GSE171213 [17] for downstream filtering and analysis. Low-quality cells with fewer than 200 detected genes, proportion of mitochondrial gene counts per cell > 10%, and proportion of hemoglobin gene counts per cell > 10% were filtered. Genes detected in fewer than three cells were excluded. We also performed doublet detection and removal using Scrublet, with a predicted doublet score > 0.25, which was then filtered out. Next, we normalized and logarithmically transformed the data, and found highly variable genes (HVGs) using the ScanPy pipeline. A principal component analysis were performed using the HVGs. Then we corrected the batch effects using the BBKNN algorithm with batch_key set to “sample.” We used the Uniform Manifold Approximation and Projection (UMAP) to visualize the unsupervised clustering results. To identify the role of the WNT10A, SMOC2, and LRP4 genes, we subset dental pulp cells and alveolar bone cells and performed the next round of clustering using Leiden clustering. Next, we utilized CellRank2 to infer cell-cell communications and Metascape to infer Gene Ontology (GO) enrichment of the top 50 genes in each subset.

Tissue preparation and immunofluorescence (IF) staining

This study was approved by the Ethics Committee of Xiangya Hospital, Central South University (202203062). The mandible of a 28-week human embryo was obtained from a non-genetically diseased aborted embryo in the Department of Obstetrics, Central South University.

Sections were cut to 5 mm and sections of the first molar germ were selected. After dewaxing and hydration, antigen retrieval was performed using an improved citrate antigen retrieval solution (Beyotime, P0083, 1:50) in a pressure cooker. Sections were blocked with endogenous peroxidase-blocking buffer (Beyotime, P0100A) for 15 min, permeabilized in 0.5% Triton X-100 for 20 min, and blocked with goat serum working solution (ZLI-9056; Zsbio) for 20 min. Multiplex IF staining for WNT10A (rabbit, 1:100, Abcam) and SMOC2 (rabbit, 1:50, Abiowell) was performed according to the manufacturer’s instructions (AiFang Biological, 2-Color Multiple Fluorescence Kit, China). Nuclear staining and section mounting were performed using a DAPI-containing sealer (F6057; Sigma). Images were captured using a Leica DM6B microscope.

Results

A 19-year-old female patient presented with the absence of nine permanent teeth and diffuse enhanced bone mineral density of alveolar bone

The patient’s mother recalled that the patient had retained a large number of primary teeth for 6 years and denied a history of tooth extraction. Clinical examination revealed that the patient was 1.60 m tall and showed no structural abnormalities in ectoderm-derived tissues such as sweat glands or sebaceous glands (Supplementary Fig. 1a). Comprehensive oral examination and CBCT confirmed the absence of nine permanent teeth: 12, 13, 14, 22, 23, 24, 35, 42, and 45. Shortened roots of the residual deciduous teeth were also observed (Fig. 1a-c). CBCT imaging also showed a diffuse increase in the bone mineral density of the alveolar bone, leading to a loss of distinction between the cortical and cancellous bones, as well as the cancellous bone region losing its original structure (Fig. 1d). The HU values of the cancellous bone region of the alveolar bone were measured in the control group (n = 10, M = 221.20, SD = 119.170) and compared with the patient’s HU value of 933. A one-sample t-test revealed that the average HU value of the control group was significantly lower than that of the patient [t(9)=-18.89, p < 0.001]. This indicates that the patient’s HU value was considerably higher than that of the control group, suggesting a greater bone density in the patient. (Fig. 1e). In addition, radiographic examination of various parts of the body showed increased bone density in the pelvis, femur, rib, scapula, spine, skull, maxilla, and mandible (Supplementary Fig. 1b-e).

Fig. 1.

Phenotypic presentation and sanger sequencing results of the patient.(a-c) Comprehensive oral examination and CBCT confirmed that: 9 permanent teeth were missing and the residual deciduous teeth were severely worn, resulting in a reduced size. (d)CBCT imaging shows a diffuse increase in bone density, resulting in the loss of distinction between cortical and cancellous bone, with a significant elevation in the density of the cancellous bone region, which has lost its original structure(red arrowhead highlights the cancellous bone region). (e) Comparison of cancellous bone HU values between the patient and the control group. (f-h) Sanger sequencing confirms the WES results that the patient carried gene variations WNT10A c.354T>G, LRP4 c.1831C>T and SMOC2 c.1138C>T

New missense variation of LRP4 [c.1831 C > T (p.R611C)], and the SMOC2 [c.1138 C > T (p.R380C)] and WNT10A variation c.354T > G (p.Y118*) were identified

We performed WES screening and Sanger sequencing to identify whether the retained primary teeth and lack of nine permanent teeth phenotype in this patient were caused by genetic factors. WES screening revealed that the patient carried three gene variations related to tooth agenesis and sclerosteosis. c.354T > G (p.Y118*) in exon 2 of WNT10A (NM_025216) is a nonsense variation. c.1831 C > T (p.R611C) in exon 14 of LRP4(NM_002334) and c.1138 C > T (p.R380C) in exon 11 of SMOC2 (NM_001166412) were nonsynonymous variations. The c.354T > G (p.Y118*) termination variation of WNT10A was previously reported [18]. The missense variation c.1138 C > T (p.R380C) in SMOC2 was documented in the genomeAD database but not reported in any literature. However, the nonsynonymous heterozygous variation in LRP4 was novel (Table 2). Sanger sequencing was used to confirm the WES results (Fig. 1f-h). According to ACMG guideline, the pathogenicity analyses were made for all variants (Table 2). Unfortunately, the DNA samples of her parents were not available for genetic screening as her father passed away.

Table 2.

The pathogenicity of three variations were predicted by in Silico tools

Gene	Exon	Nucleotide change	Protein change	Variation type	dbSNPs	1000 g-2015	gnomAD	Provean	Mutation-taster	Polyphen2	ACMG classification
LRP4	14	c.1831 C > T	p.R611C	nonsynonymous SNV	rs771793515	--	0	Deleteri-ous	Disease causing	Probably damaging	PM2 Uncertain significance
WNT10A	2	c.354T > G	p.Y118*	nonsense	/	--	--	/	Disease causing automatic	/	PVS1 + PM2 Likely pathogenic
SMOC2	11	c.1138 C > T	p.R380C	nonsynonymous SNV	rs148539759	--	0.0002	Deleteri-ous	Disease causing	Probably damaging	PM1 + PM2 + BP1 Uncertain significance

--: Variant was not found in 1000g2015 or gnomAD; /: Variant cannot be predicated in dbSNP, PROVEAN or Polyphen2. ACMG PVS1: Null variant (nonsense, frameshift, canonical ± 1 or 2 splice sites, initiation codon, single or multiexon deletion) in a gene where LOF is a known mechanism of disease; PM1: Located in a variational hot spot and/or critical and well-established functional domain without benign variation; PM2: Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes Project, or Exome Aggregation; BP1: Missense variant in a gene for which primarily truncating variants are known to cause disease

Analysis of conserved amino acid residues and structural prediction of mutated LRP4, SMOC2 and WNT10A proteins

Conservation analysis helps evaluate the functional importance and pathogenic potential of variations by revealing which genes or sequences remain unchanged throughout the evolutionary process. To determine the functional significance of the three identified variations, we performed a conservation analysis, with the evolutionary rate calculated in ConSurf based on the evolutionary relationship between the protein and its homologs, as well as the amino acid similarity reflected in the substitution matrix. The conservation scale of residues R611 in the LRP4 protein and Y118 in the WNT10A protein was rated as average (Fig. 2a, c), indicating that these residues were under moderate evolutionary pressure and that they may impact the specific functions or structures of the proteins. Furthermore, the R380 residue in the SMOC2 protein was highly conserved and exposed, with a score of 9 (Fig. 2b), further emphasizing its critical importance in protein function and molecular interactions.

Fig. 2.

Analysis of conserved amino acid residues and prediction of secondary structure of proteins (a-c) ConSurf server predicted conserved amino acids in three proteins (the variation site is marked with a red triangle). Based on the conservation scale, the amino acids are classified as variable, average, or conserved. (d-d’) Structural similarity of the LRP4 protein before and after the p.R611C variation. (e-e’) Structural similarity of the SMOC2 protein before and after the p.R380C variation. (f-f’) Most of the WNT10A protein structure is lost after the p.Y118* variation. Variant sites are indicated by red arrowheads. Helices are represented as blue lines, sheets are represented as purple lines, turns are represented as green lines and coils are represented as yellow lines

The function of a protein is determined by its three-dimensional structure, and abnormalities in this structure cause many genetic diseases. To evaluate the impact of the three genetic variations on their respective protein structures at different levels, we predicted both secondary and tertiary protein structures. In the secondary structure prediction, the percentages of α-helix, β-turn, extended chain, and percentage of random coil in the three mutant proteins were altered (Fig. 2d-f’ and Table 3), suggesting that the variations may influence the local folding and organization of the proteins.

Table 3.

Secondary structure analysis of mutated proteins

	Alpha helix	Extended strand	Beta turn	Random coil
LRP4(Wild-type)	3.78%	26.67%	4.72%	64.83%
LRP4(p.R611C)	5.14%	26.88%	4.04%	63.94%
WNT10A(Wild-type)	33.57%	8.63%	1.68%	56.12%
WNT10A (p.Y118*)	44.44%	0.85%	2.56%	52.14%
SMOC2(Wild-type)	20.40%	5.38%	0.90%	73.32%
SMOC2(p.R380C)	21.08%	4.48%	0.90%	73.54%

The α-helix, β-turn, extended strand and percentage of random coil of the mutant proteins were changed

In the tertiary structure prediction, RMSD values of 3.142 nm for LRP4, 18.041 nm for SMOC2, and 20.818 nm for the WNT10A mutant and wild-type proteins (Fig. 3a-c) indicated structural variations among the three groups of proteins before and after variation. The RMSD values suggest that the variations may cause substantial rearrangements in the tertiary structure of the proteins, which could have considerable implications for their stability and function. In addition, the result of the c.354T > G (p.Y118*) variation in the WNT10A gene, which was identified in the patient, showed that the mutant WNT10A protein lost most of its protein structure compared with the wild-type (Fig. 3c-c’’). The contribution of hydrogen bonds to protein stability is significant [19]. Variations can disrupt hydrogen bonds and cause local or global structural instabilities. In the LRP4(p.R611C) and SMOC2(p.R380C) proteins, hydrogen bonds were lost after the variation (Fig. 3d, e). The results of the protein tertiary structure and hydrogen bonding prediction suggested that these variations disrupt hydrogen-bonding interactions and protein stability and function.

Fig. 3.

Prediction of tertiary structure of proteins and hydrogen bonding results (a-a’’) Tertiary structure predictions of wild-type LRP4 and LRP4 p.R611C. (b-b’’) Tertiary structure prediction of wild-type SMOC2 and SMOC2 p.R380C. (c-c’’) Tertiary structure prediction of wild-type WNT10A and WNT10A p.Y118*. (d-f) Hydrogen bonding results of LRP4, SMOC2, and WNT10A proteins

Construction of a human permanent molar and surrounding microenvironment development cell landscape at single-cell resolution

To investigate the correlation of WNT10A, SMOC2, and LRP4 with human permanent molar development, we first compiled scRNA-seq data of the developing human third molar mesenchyme and periodontal tissues from two published datasets (Fig. 4a). Using these two datasets, we can infer how the two crucial tooth agenesis (TA)-associated genes are correlated and how the surrounding alveolar bone affects tooth development. These data included four periodontal and two developing human third molar samples (Fig. 4b). After stringent quality control, batch correction, and unsupervised clustering, we identified nine main cell types: alveolar bone cells, pulp cells, odontoblasts, endothelial cells, pericytes, T cells, B cells, classical monocytes, and nonclassical monocytes (Fig. 4c). We showed that all cell types were derived from the six samples except pulp cells, which were limited by the sampling location, indicating that there was no obvious batch effect that may undermine downstream analysis among those samples (Fig. 4d). We found that the nine main cell types expressed classical marker genes (Fig. 4e), including SOX9, S100A13, CD14, FCGR3B, RGS5, CDH5, ASPN, SPP1, CD3D, and JCHAIN.

Fig. 4.

Construction of a human developing permanent molar and surrounding microenvironment cell landscape at single-cell resolution. (a-b) scRNA-seq data covered four periodontal samples and two developing human third molar samples is from two published datasets. (c) After stringent quality control, nine main cell types were identified. (d) All cell types derived from those six samples except pulp cells which are limited by the sampling location, indicating that there is no significant batch effect that may undermine downstream analyses among those samples. (e) The nine main cell types expressed their classical marker genes, including SOX9, S100A13, CD14, FCGR3B, RGS5, CDH5, ASPN, SPP1, CD3D, and JCHAIN

LRP4-positive alveolar bone cell subset regulates WNT10A-positive and SMOC2-positive dental pulp cell subsets

WNT10A and SMOC2 are the dominant TA genes, whereas LRP4 is associated with bone development. To further analyze the relationships among these three genes, we subset developing dental pulp and alveolar bone cells and performed unsupervised Leiden clustering, which divided these cells into four clusters (Fig. 5a, b). We showed that the pulp cells could be divided into two subsets, namely clusters 0 and 1 (Fig. 5b, c). These two pulp subsets were marked by WNT10A and SMOC2, indicating their different roles in permanent molar development (Fig. 5c, d). Using IF staining on developing human permanent tooth sections, we examined whether WNT10A and SMOC2 were expressed in developing dental pulp cells and different pulp subsets (Fig. 5e). We showed that WNT10A and SMOC2 were expressed in different mesenchymal cells in the human tooth germ (Fig. 5f, g), verifying the scRNA-seq analysis results. To specify the functional roles of these cell subtypes in human permanent tooth development, we first performed Gene Ontology (GO) enrichment analysis based on the top 50 genes in each cell subset using Metascape. We found that WNT10A + pulp cells were significantly enriched in processes such as matrix formation, blood vessel development, and cell proliferation, indicating their critical roles in tissue development and repair (Fig. 6a). SMOC2 + pulp cells were enriched in hormone response, Transforming Growth Factor-beta (TGF-β) signaling, and blood vessel migration, suggesting their involvement in vascular and immune regulation (Fig. 6b). Alveolar bone cells were enriched in extracellular matrix organization, response to growth factors, platelet-derived growth factor signaling, and ossification, highlighting their importance in bone remodeling and microenvironment maintenance (Fig. 6c). These results indicate that the crucial roles of these three cell subsets in the human developing permanent tooth germ and alveolar bone cells in regulating the microenvironment of the human tooth germ, thereby influencing permanent tooth development. To further analyze cell-cell communication among WNT10A + pulp cells, SMOC2 + pulp cells, and LRP4 + alveolar bone cells, we performed CellChat analysis. We found that LRP4 + alveolar bone can act as a fibroblast growth factor (FGF)-signal sender to regulate WNT10A + pulp cells (Fig. 6e, f), revealing the functional role of the alveolar bone in permanent tooth development. The above results indicate the crucial role of the TA-associated genes WNT10A and SMOC2 in permanent tooth development and further reveal that LRP4 + alveolar bone cells affect TA phenotypes by regulating dental pulp cells via FGF signaling.

Fig. 5.

Identification of dental pulp cell clusters and immunofluorescence-based validation (a) Developing dental pulp cells and alveolar bone cells were subsetted. (b) Unsupervised Leiden clustering divided these cells into four clusters and pulp cells can be divided into two subsets: cluster 0 and 1. (c-d) These two pulp subsets are marked by WNT10A and SMOC2 separately, indicating their different roles in permanent molar development. (e-g) IF staining on developing human permanent tooth sections confirmed that WNT10A(green) and SMOC2(red) are expressed in developing dental pulp cells and in different pulp subsets(the white boxes highlight areas of high expression, displayed under high magnification). The odontoblast layers were marked with white arrows

Fig. 6.

The GO functional enrichment analysis and the cell-cell communications (a-c) GO functional enrichment analysis of WNT10A + pulp cells, SMOC2 + pulp cells, and alveolar bone cells. (d) WNT10A + pulp cells contribute the most ligand-receptor pairs in the developing molar and surrounding microenvironment, indicating their central regulatory role. (e) LRP4 + alveolar bone can become an FGF-signaling sender to regulate WNT10A + pulp cells, revealing the functional role of alveolar bone in permanent tooth development

Discussion

TA refers to the loss or defect of dentition, often presenting as permanent tooth loss, with a prevalence rate higher than that of primary tooth loss [20]. TA includes different phenotypes, ranging from hypodontia (deficiency of fewer than six teeth) to oligodontia (deficiency of six or more teeth) and anodontia (all teeth missing) [21]. The prevalence of congenital TA in different populations ranges from 2.6 to 11.3% [22]. Oligodontia has been reported to affect 0.08–0.16% of children [23]. Genetic factors are primary contributors to congenital TA [24]. Therefore, the identification of novel TA variation sites would aid in the development of specific congenital TA diagnosis and treatment options. In this study, we identified an unreported NSTA-associated variation in SMOC2 (c.1138 C > T, p.R380C) and a novel variation in LRP4 (c.1831 C > T, p.R611C) that affect bone development.

The maxilla, mandible, and teeth all develop from the first branchial arch, and defects in one can affect others [25]. It has been suggested that spatial impediments in the developing bone may be an important factor during tooth morphogenesis [26]. For example, variations in the Osx gene, which is associated with bone formation, undermine tooth development in mice [27]. However, the mechanism by which defects in the mandibular bone mass influence tooth development in humans remains unclear. The LRP4 gene plays a crucial role in bone development, and variations in LRP4 can lead to sclerosteosis 2, characterized by increased bone density [28]. Carriers of LRP4 variations exhibit higher bone density in the mandibular region. However, to date, no carriers of congenital tooth loss caused by LRP4 variations have been reported. WNT10A and SMOC2 gene variations have been reported to cause NSTA [6, 8]. Previous studies have shown that variations in WNT10A and SMOC2 can lead to TA, in which permanent incisors are often preserved. Additionally, some patients with these variations may retain their primary teeth, whereas others may not. In this study, we present the case of a 19-year-old female patient with NSTA and increased jawbone density. We found that the missing permanent teeth were mostly incisors and canines, with most primary teeth retained. To unravel why this patient showed most primary tooth retention and permanent tooth loss, we used WES screening to identify the causative genetic factors and re-analyzed the human permanent teeth single-cell RNA-seq dataset to investigate the possible relationship between genetic factors and clinical phenotypes. We showed that the patient not only harbored TA-associated genes but also presented an LRP4 variation. After analyzing the human permanent teeth single-cell RNA-seq dataset, we revealed that LRP4-expressing alveolar bone cells could regulate WNT10A-expressing and SMOC2-expressing dental pulp cell subsets via FGF signaling, indicating that impacted alveolar bone mass may have led to the retention of most primary teeth in this patient.

The FGF pathway plays a critical role in various stages of tooth development. Signaling molecules from the FGF family, such as FGF8, FGF9, and FGF17, initiate tooth formation during early development through epithelial expression [29, 30]. Specifically, FGF8 regulates PAX9 expression, which is essential for the optimal positioning and initiation of molar development [31, 32]. Research has also shown that multiple FGFs can compensate for individual signal deficiencies that affect the thickening of dental layers [33]. FGF3 and FGF10 are crucial during the transition to the bud stage, primarily influencing the development of the tooth crown and cusps through the FGFR2IIIb pathway [34, 35]. Additionally, FGF signaling is involved in root development, with FGF18 regulating the condition of the periodontal mesenchyme to promote root elongation [36]. FGF2 plays a significant role in enhancing the growth and accumulation of human dental pulp cells [37]. Overall, the FGF pathway ensures the normal progression of tooth development from the early stages to maturity by regulating cell proliferation, differentiation, and epithelial-mesenchymal interactions. In this study, we found that the LRP4 + alveolar bone cells may regulate the function of SMOC2 + and WNT10A + dental pulp cells through the FGF pathway, and variations in these genes may lead to dysregulation of this signaling pathway, impacting tooth formation and development.

WNT10A is a crucial ligand activating canonical signaling pathway whose transcription factors are expressed in early tooth development and the osteogenesis process [38, 39]. As a result, variations in WNT10A disrupt normal tooth formation and bone formation. Using cell-cell communication analysis we showed that WNT10A-expressing dental pulp cells have the strongest interaction with the alveolar bone, indicating its crucial role in regulating the surrounding bone environment. Additionally, SMOC2—a member of the BM-40 family—appears to function through a different set of signaling interactions and extracellular matrix influences. Its variation is frequently associated with the loss of replacement permanent teeth. We showed that SMOC2 is expressed in a different dental pulp cell subset. The heterogeneity of dental pulps distinguished by WNT10A and SMOC2 indicate the different mechanism of which they impact tooth development. Therefore, our findings offer a clearer framework for understanding the individual and complementary roles of different NSTA genes.

Most tooth agenesis patients exhibit lower craniofacial bone density. For example, loss-of-function mutations of TA-associated genes WNT1, LRP5, and BMP4 can cause osteoporosis [40–42]. Interestingly, this patient showed increased bone density. NSTA variations (WNT10A and SMOC2) identified in this patient can have the capability to impair the bone, indicating that there may be some other variation contributing to this phenotype. We showed that the patient presented LRP4 variation with impaired protein structures. LRP4 can interact with sclerostin, a key extracellular factor of bone remodeling. Lrp4-null mutation or specific mutation in osteoblast-lineage cells in mice showed an increased cortical and trabecular bone mass, elevating bone formation [43], indicating that the LRP4 variation may contribute to the patient’s phenotype.

One limitation of this research is that the DNA samples were unavailable from the parents, limiting our ability to perform family segregation analysis. we acknowledge the importance of such data in confirming the pathogenicity of these variations. To mitigate this limitation, we conducted in silico analyses, including PROVEAN, MutationTaster and PolyPhen-2 predictions, all of which supported the pathogenic potential of the identified variants. Additionally, as tooth follicle cells are more proximal to the surrounding bones, tooth follicles seemed prioritized over dental pulp cells to study tooth-bone communication. However, as scRNA-seq data of human fetal tooth development is limited and there is no tooth follicle cells in the dataset that we used, we analysed the cell-cell communication between dental pulp cells and aveolar bones. We showed that alveolar bone cells can regulate dental pulp cells via FGF signaling. Previous studies showed that mandible bone cells can secrete signaling molecules and microRNAs to regulate tooth development [44, 45], corresponding to our findings. In the future, identifying the concise relationships among tooth follicle cells, dental pulp cells and surrounding bones will help unravel how individual NSTA-associated variations and Bone-associated variations collaborated impact tooth and mandible development.

Conclusion

This study suggests a link between maxillofacial and permanent tooth development by identifying a SMOC2 variation (c.1138 C > T, p.R380C) linked to NSTA and a novel LRP4 variation (c.1831 C > T, p.R611C) associated with impaired bone development. This provides valuable insights into human craniofacial development and indicates that tooth-bone communication may play a role in permanent tooth development, with its disruption potentially contributing to the NSTA phenotype.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

NSTA

Non-syndromic tooth agenesis

CBCT

Cone-beam computed tomography

WES

Whole-exome sequencing

scRNA-seq

Single-cell RNA sequencing

MAF

Minor allele frequency

ExAC

Exome Aggregation Consortium

1000G

1000 Genomes Project

gnomAD

Genome Aggregation Database

SNV

Single-nucleotide variant

ACMG

American College of Medical Genetics and Genomics

PolyPhen-2

Polymorphism phenotyping

PROVEAN

Protein Variation Effect Analyzer

PCR

Polymerase chain reaction

RMSD

Root-mean-square deviation

HVG

Highly variable genes

IF

Immunofluorescence

HU

Hounsfield Unit

FGF

Fibroblast growth factor

TA

Tooth agenesis

Author contributions

YQ.C., XC.L. and SY.M. wrote the main manuscript text and prepared Figs. 2, 3, 4, 5 and 6. YQ.C., XC.L. and SY.M. contributed equally to the manuscript. J.S. prepared Figs. 1 and 6. CQ.Z.,Y.W. and YT.L. performed data analysis and wrote sections of the manuscript. QP.G. and XS.W. supervised the study and edited the manuscript. All authors commented on previous versions of the manuscript and approved the submitted version.All authors reviewed the manuscript.

Funding

Qingping Gao was supported by a grant from the Department of Science and Technology of Hunan Province (2022SK2049). Xiaoshan Wu was supported by grants from the National Natural Science Foundation of China (82370917) and the Science and Technology Innovation Program of Hunan Province (2024RC1024). Xuechun Li was supported by grants from the China Postdoctoral Science Foundation (2023M733941) and the Natural Science Foundation of Hunan Province (2023JJ40960). Jing Sun was supported by grants from the Natural Science Foundation of Hunan Province (2023JJ40988) and the Changsha Municipal Natural Science Foundation (kq2208394).

Data availability

The datasets analysed during the current study are available in the ClinVar repository (SUB14871290,SUB14871279,SUB14869404).

Declarations

Ethics approval and consent to participate

This research has been independently reviewed and approved by the Medical Ethic Committee of Xiangya Hospital, Central South University (Approval Number: 202203062).

Consent for publication

The patient reported in this manuscript provided written informed consent for the publication of the case details.

Competing interests

The authors declare no competing interests.