Slc26a2-mediated sulfate metabolism is important in tooth development

Sulfate [or sulfate ion (SO₄²⁻)] is an anion involved in a wide range of significant biological processes, including biosynthesis and metabolism of a variety of endogenous biological molecules, and is important for cell growth and development of the organism (Foster and Mueller, 2018; Soares da Costa et al., 2017). Because sulfate cannot freely pass through the plasma membranes of cells, transport mechanisms are required for the movement of sulfate into and out of mammalian cells (Alper and Sharma, 2013). Such mechanisms are also necessary for sulfate absorption from the gastrointestinal tract and re-absorption by the renal tubules (Seidler and Nikolovska, 2019). After sulfate is transported into the cytoplasm or nucleus, 3'-phosphoadenosine 5'-phosphosulfate (PAPS) is synthesized from ATP and sulfate by two enzymatic reactions – catalyzed by ATP sulfurylase and adenosine 5'-phosphosulfate kinase. After PAPS is transported to the Golgi apparatus, sulfotransferases transfer sulfate groups from PAPS to glycosaminoglycans and tyrosine. In addition, cytosolic sulfotransferases transfer sulfate groups from PAPS to steroid hormones in cytosol (Klaassen and Boles, 1997). However, part of the sulfate supply is also known to be derived from breakdown of the sulfur-containing amino acids cysteine and methionine (Elgavish and Meezan, 1991; Markovich and Aronson, 2007).

In mammals, the Slc26 gene family, which encodes anion transporters, consists of 11 genes – Slc26a1-11. The Slc26 gene family encodes transporters of a broad range of anionic substrates, including sulfate, HCO³⁻, Cl⁻, oxalate, I⁻ and formate. Slc26a1 and Slc26a2 encode the proteins SAT1 and DTDST, respectively, which are sulfate/chloride exchangers that function as cell membrane sulfate transporters and enable the intracellular transport of inorganic sulfate (Kere, 2006). In humans, variants in the SLC26A2 gene cause a spectrum of recessively inherited chondrodysplasias. Although the phenotype differs according to the type of SLC26A2 variants, the main clinical features are short stature, joint contractures, club feet, shortening of the limbs and a waddling gait (Cai et al., 2015). More recent reports also suggest that SLC26A2 variants are associated with a wide range of clinical manifestations in the craniomaxillofacial region, including large upper facial height, micrognathia, high palate, cleft palate (25-60%), tooth agenesis (30%) and microdontia (Härkönen et al., 2021). Variants in the other eight members of the SLC26 family have also been implicated in human disease. However, unlike SLC26A2 variants, variants in the other SLC26 family members do not induce any abnormalities in the skeletal system or craniofacial region.

The function of Slc26a2 in mammals has been investigated using genetic mouse models. Forlino et al. (2005) generated a DTDST knock-in mouse model harboring human variants, and the mice showed partial loss of function of the sulfate transporter. These mice exhibited a short stature, joint contracture, reduced Toluidine Blue staining of cartilage and irregular chondrocyte size (Forlino et al., 2005). Zheng et al. (2019) generated Slc26a2^−/− mice to investigate the effects of SLC26A2 deficiency on chondrodysplasia. Although patients with variants in the SLC26A2 gene reportedly show abnormalities in the craniomaxillofacial region, including dwarf teeth and congenital absence of teeth (Karlstedt et al., 1996), the role of SLC26A2 during tooth development has not been fully elucidated.

In the present study, we investigated the pattern of Slc26a2 expression in developing tooth germs, performed morphological and histological evaluations of tooth germs in Slc26a2 knockout mice, and examined the effects of Slc26a2 deficiency on enamel and dentin matrix formation using kidney-capsule grafting. This is the first study to demonstrate the pivotal role of the SLC26A2-mediated sulfate transporter during tooth development.

To understand the role of SLC26A2 in tooth development, we quantitatively analyzed the expression of sulfate transporter family genes, including Slc26a1 and Slc26a2, in developing tooth germ. We re-analyzed a public single-cell RNA sequencing dataset on isolated mice incisors at postnatal day (P)0. We identified 11 cell population clusters of odontoblasts, sub-odontoblasts, dental mesenchyme 1 and 2, ameloblasts, pre-ameloblasts, inner enamel epithelium or outer enamel epithelium, stratum intermedium or stellate reticulum, leukocytes, erythrocytes and endothelial cells (Fig. 1A; Fig. S1). We next characterized the expression of Slc26a1, Slc26a2, Slc26a6, Slc26a7, Slc26a10 and Slc26a11 in the dataset. A dot plot showed that Slc26a2 was more strongly expressed in all clusters, including ameloblast and odontoblast clusters, than other sulfate transporter family genes (Fig. 1B).

We next used in situ hybridization to confirm the expression pattern of Slc26a2 in tooth development in mice at embryonic day (E)18.5. The prominent expression of Slc26a2 was observed in odontoblasts and ameloblasts at E18.5 (Fig. 1C). Interestingly, at each developmental stage, the expression of Slc26a2 was found to be markedly higher than that of Slc26a1 (Fig. S2A). Furthermore, the expression of Slc26a2 increased with developmental stage, with craniofacial formation beginning at E9.5 and the tooth bud forming at E13.5 (Fig. S2A). At E18.5, the expression of Slc26a2 was higher than that of Slc26a1 in the upper molars (Fig. S2B). These results suggest that Slc26a2 plays a major role in the transport of sulfate ions through the cell membrane during tooth development.

To investigate the role of Slc26a2 deficiency in tooth development, we generated Slc26a2 knockout mice by targeted deletion of Slc26a2 exon 2 (Slc26a2-KO-Δexon2) using the CRISPR-Cas9 gene-editing method (Makino et al., 2016). Previous reports showed that Slc26a2 knockout (KO) mice died immediately after birth, with no respiratory movement and an overall skeletal phenotype characterized by a short neck, small chest and very short limbs (Zheng et al., 2019). Consistently, Slc26a2-KO-Δexon2 mice died immediately after birth owing to respiratory abnormalities. Whole-mount images of Slc26a2-KO-Δexon2 mice at E18.5 revealed hypoplasia of the maxilla rather than the mandible (Fig. 2A-D; Fig. S3). Micro-computed tomography (CT) of Slc26a2-KO-Δexon2 mice demonstrated a short stature, small chest and very short limbs. The long tubular bone was shorter in length and longer in diameter than that of the control mice (Fig. S4A,B). Whole-mount skeletal preparations demonstrated chondrodysplasia and reduced Alcian Blue staining of the cartilage in Slc26a2-KO-Δexon2 mice (Fig. S4C,D and Fig. S5). In the cranio-maxillofacial region, Slc26a2-KO-Δexon2 mice showed a hypoplastic maxilla, hypoplasia of nasal cartilage (Fig. 2E,F), small cranial base (Fig. 2G,H) and short ribs (Fig. 2I,J), while the mandible size remained consistent with that of control mice (Fig. S5). Additionally, we observed reduced Alcian Blue staining intensity in the nasal septum, synchondroses of the cranial base, costal cartilage and growth plate cartilage at E18.5 (Fig. 2; Fig. S5). Furthermore, at E15, Meckel's cartilage showed decreased Alcian Blue staining intensity and hypoplasia, indicating impaired cartilage formation in these regions (Fig. S6). At E18.5, mutant and control tooth germs were examined by contrast-enhanced micro-CT (Fig. 2K-R). Slc26a2-KO-Δexon2 mice showed a shorter anterior–posterior length of the upper and lower incisor, measured from the cervical loop to the incisor tip, compared to that of control mice. Similarly, the upper molar tooth germs in Slc26a2-KO-Δexon2 mice were shorter with regard to both crown proximal–central width and height compared to those of control mice. In contrast, there were no significant differences in the width or height of the lower molars between Slc26a2-KO-Δexon2 and control mice (Fig. 2S).

We performed a histological evaluation to clarify the effects of Slc26a2 deficiency on tooth development. Upper incisors and molars in Slc26a2-KO-Δexon2 mice had flattened odontoblasts and nuclei that did not show intracellular polarity (Fig. 3A-D′). The height of pre-secretory ameloblasts was lower in Slc26a2-KO-Δexon2 mice than in control mice.

To further examine the effect of Slc26a2 deficiency on odontoblast differentiation, we evaluated the expression of the odontoblast marker genes dentin sialoprotein (Dspp) and dentin matrix protein 1 (Dmp1), which encode non-collagenous organic substances in dentin and show increased expression in association with odontoblast differentiation (Chen et al., 2016; Yamashiro et al., 2007). In Slc26a2-KO-Δexon2 mouse tooth germ, the expression of Dspp and Dmp1 in upper molars was significantly decreased compared to that in control tooth germ (Fig. 3E,F). The tooth phenotype in Slc26a2-KO-Δexon2 mice was much milder in lower molar tooth germs than that in upper molar tooth germs (Fig. 3G-J′). Taken together, these results demonstrate that Slc26a2 is required for odontoblast differentiation of upper molars in mice.

To gain further insight into the effect of SLC26A2 knockdown on the differentiation of dental pulp stem cells into odontoblasts, human dental pulp stem cells (hDPSCs) were used to analyze the odontoblast differentiation potential. SLC26A2 knockdown cells were generated via lentivirus-mediated delivery of shRNA, and the expression of SLC26A2 was decreased by more than 80% in sh-SLC26A2 knockdown cells (Fig. 4A). In sh-SLC26A2 knockdown cells, the expression of DSPP and DMP1 was significantly decreased compared to that in control cells (Fig. 4A). This result indicates that SLC26A2 knockdown can directly affect the differentiation of pulp stem cells into odontoblasts.

Owing to the neonatal lethality of Slc26a2-KO-Δexon2 mice, it was not possible to assess the effects of Slc26a2 deficiency on dentin and enamel matrix production in Slc26a2-KO-Δexon2 mice. We performed ex vivo organ culture of tooth germs by implanting the tooth germs under the kidney capsule of nude mice. After 4 weeks of organ culture, the implanted tooth germs were collected and analyzed by micro-CT. The sagittal sections of the micro-CT images showed hypoplasia of the tooth crown, consisting of enamel and dentin (Fig. 4B). A quantitative assessment of the dentin volume demonstrated a significant reduction in Slc26a2-KO-Δexon2 mouse tooth germ compared to control tooth germ (Fig. 4B). In addition, tooth root shortening was observed in Slc26a2-deficient tooth germs (Fig. 4B). These results support our hypothesis that Slc26a2 deficiency leads to impairment of odontoblast differentiation.

To further examine the molecular mechanisms underlying defective odontoblast differentiation in Slc26a2-deficient mice, we performed RNA-sequencing (RNA-seq) analysis of genes exhibiting differential expression in Slc26a2 knockdown primary mouse dental papilla mesenchymal cells (mDPCs). The heatmap generated through hierarchical clustering displayed distinct gene expression patterns between Slc26a2-silenced mDPCs and control mDPCs (Fig. 5A). Additionally, volcano plot analysis highlighted significantly upregulated and downregulated genes [|log2fold change (FC)|>2] in Slc26a2-silenced mDPCs (Fig. 5B). Gene Ontology (GO) analysis indicated that several biological processes, such as Ossification, Skeletal system development, Osteoblast differentiation, Biomineral tissue development and Odontogenesis, were downregulated in Slc26a2-silenced mDPCs compared to control mDPCs (Fig. 5C). Notably, odontogenesis-related genes – including Col1a1, Aspn, Dmp1, Axin2, Wnt10a, Sp6, Sp7 and Fgfr2 – were downregulated in Slc26a2-defecient mDPCs compared to control mDPCs. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis revealed that several downregulated genes were involved in the Wnt signaling pathway, which was identified as the most significantly enriched pathway (Fig. 5C; Fig. S7). The expression of genes specifically related to Wnt signaling (Axin2 and Wnt10a) and odontogenesis (Dmp1 and Dspp) is presented in Fig. 5D.

Because the phenotype of upper tooth germ is more pronounced than that of lower tooth germ in Slc26a2-KO-Δexon2 mice (Fig. 3), we hypothesized that the function of Slc26a2 as a sulfate transporter would be compensated for in the lower tooth germ. To examine the sulfate uptake in the upper and lower tooth germs in Slc26a2-deficient and control mice, we quantified the amount of sulfated glycosaminoglycan (GAG), which reflects the sulfate uptake through its transporter. The total amount of sulfated GAG in the upper molar tooth germs was significantly (P<0.0001) decreased by Slc26a2 deficiency (Fig. 6A). In contrast, there was no significant difference in the total amount of sulfated GAG in the lower molar tooth germs between the Slc26a2-deficient and control mice (Fig. 6A). Although there was no significant difference in the total amount of sulfated GAG between the upper and lower molar tooth germs in the control group, the total amount of sulfated GAG was significantly decreased in the upper molars compared to that in the lower molars in the Slc26a2-deficient group.

To assess the genetic redundancy of Slc26a1, a homolog of Slc26a2, and Slc26a2, we performed micro-dissection, which enables the extraction of RNA from odontoblasts with high purity and the absolute quantification of mRNA. As a result, we found that the expression of Slc26a2 was significantly higher than that of Slc26a1 in odontoblasts from upper tooth germ (Fig. 6B). Conversely, the expression of Slc26a1 was higher than that of Slc26a2 in odontoblasts from lower tooth germ (Fig. 6B). These results suggest that Slc26a1 is predominantly expressed in lower tooth germs and may compensate for the function of Slc26a2 in Slc26a2-deficient lower tooth germs.

In the present study, we demonstrated that Slc26a2 is predominantly expressed in dental tissues, with particularly high expression in odontoblasts and ameloblasts, during tooth development (Fig. 1). Deficiency in Slc26a2 in chondrocytes reportedly disrupts cartilage growth via the attenuation of chondrocyte proliferation and induction of cell death (Zheng et al., 2019). We also confirmed that cell proliferation was decreased, and apoptosis was increased, in Slc26a2-deficient chondrocytes in vivo (Fig. S8). However, cell proliferation and apoptosis were not significantly altered in the tooth germ of Slc26a2-KO-Δexon2 mice compared to that of control mice (Fig. S9). The sulfation level was markedly higher in cartilage than in the developing tooth germ (Fig. S10). This indicated that the consumption of the sulfate anion in developing tooth germ is much lower than that in cartilage, and susceptibility to Slc26a2 deficiency may be dependent on the tissue/cell-specific requirement of sulfate transportation. Interestingly, the tooth phenotype in Slc26a2-KO-Δexon2 mice was more prominent in the upper incisors and upper molars than in the lower molars (Fig. 3). We found that the expression of Slc26a1 was higher in the lower tooth than in the upper tooth (Fig. 6B), suggesting that SLC26A1 is able to compensate for SLC26A2 in lower tooth germ. In fact, Slc26a2 deficiency did not affect the amount of sulfated GAG in the lower tooth germ. However, a significant difference was observed in the upper tooth germ of Slc26a2-KO-Δexon2 mice compared to control mice (Fig. 6A). The upper and lower jaws are derived from the first branchial arches. Dlx transcriptional factors are regionally expressed within branchial arches and are implicated in regulating jaw-specific genetic programs for proper patterning during craniofacial development (Dollé et al., 1992; Bulfone et al., 1993; Robinson and Mahon, 1994; Depew et al., 2002; Qiu et al., 1997). The Slc26a1 and Slc26a2 expression patterns might be part of the jaw-specific gene regulation machinery. Further studies will be required to clarify the mechanisms underlying the regulation of the different expression patterns of Slc26a1 and Slc26a2 in the upper tooth germ and lower tooth germ during tooth development.

In Slc26a2-KO-Δexon2 mice, we observed short, flattened odontoblasts, suggesting the loss of odontoblast polarity (Fig. 3A-D′). Additionally, the expression of Dspp and Dmp1, well-defined odontoblast differentiation markers, was also decreased in Slc26a2-deficient tooth germ compared to control tooth germ in vivo (Fig. 4A). These findings highlight the critical role of Slc26a2 in odontoblast differentiation and dentin formation. Furthermore, in vitro experiments showed that odontoblastic differentiation of hDPSCs and mDPCs was substantially suppressed by Slc26a2 silencing (Fig. 4A and Fig. 5A-D), suggesting that Slc26a2 directly influences odontoblast differentiation, independent of secondary effects from the surrounding tissues or systemic sulfate insufficiency.

Owing to the postnatal lethality of Slc26a2-deficient mice, we evaluated tooth morphogenesis and dentin formation using ex vivo transplantation of tooth germ under the kidney capsule, revealing reduced dentin formation in Slc26a2-deficient tooth germ compared to control tooth germ (Fig. 4B). Ex vivo organ culture of tooth germ is often selected as an alternative method for examining tooth germs from genetically modified mice. Kidney-capsule grafting has previously been reported to provide tooth germs with an in vivo biological environment (Ono et al., 2017; Otsu et al., 2012). It is thus possible to retain the physiological features, and morphogenesis of transplanted tooth germs proceeds normally under conditions of kidney-capsule grafting.

GAGs are linear polysaccharides composed of repeating disaccharide units and are typically found as part of proteoglycans (PGs) by attaching to specific serine residues within a core protein. The sulfation of GAGs introduces negative charges at various positions on the PGs, which in turn induces a wide array of biological functions owing to the structural microheterogeneity of these molecules. Heparan sulfate (HS), a highly sulfated GAG, plays a crucial role in both odontogenesis and amelogenesis, as evidenced by previous studies (Bishop et al., 2007; Inubushi et al., 2024). Yasuda et al. (2010) demonstrated the critical role of sulfation in dental tissue development by generating mice deficient in Golgi-associated N-sulfotransferase 1 (NSDT1), an enzyme responsible for the sulfation of HS–PG glycosaminoglycan chains. These Ndst1 knockout mice exhibited hypodontia in the formation of incisors and molars, alongside abnormal differentiation and organization of odontoblasts (Yasuda et al., 2010). Further supporting the importance of sulfation, Hayano et al. (2012) showed that the removal of sulfate groups from the 6-O position of N-acetylglucosamine by extracellular glucosamine-6-sulfatases SULF1 and SULF2 was significant for odontoblast differentiation and dentin matrix production during dentinogenesis. Specifically, Sulf1/Sulf2 double-null mice displayed a thinner dentin matrix and shorter roots, accompanied by reduced Dspp mRNA expression (Hayano et al., 2012). Importantly, HS proteoglycans have also been implicated in the modulation of the canonical Wnt signaling pathway during odontogenesis (Ai et al., 2003; Hayano et al., 2012; Inubushi et al., 2023). Our study's findings of diminished odontoblast differentiation and reduced dentin matrix production in Slc26a2-KO-Δexon2 mice are consistent with the phenotypes observed in transgenic mice with altered sulfation of HS. Moreover, our RNA-seq data revealed significant downregulation of Wnt signaling pathway-related markers in Slc26a2-deficient mDPCs (Fig. 5C,D). This suggests that the attenuation of odontoblast differentiation in these cells may be, at least in part, due to the dysregulation of Wnt signaling during odontogenesis. Given that chondroitin sulfate, another GAG, is also implicated in tooth development (Ida-Yonemochi et al., 2022), it is plausible that Slc26a2 deficiency affects odontogenesis through the reduced sulfation of GAGs, particularly HS, during tooth development. These insights highlight the critical interplay between Slc26a2-dependent sulfate metabolism, GAG sulfation and the canonical Wnt signaling pathway in the proper differentiation of odontoblasts and formation of the dentin matrix.

In conclusion, our study demonstrates that Slc26a2 is a key sulfate transporter in tooth germ, and its deficiency leads to hypoplasia of the incisors and molars, particularly in the upper jaw. This is the first study to establish the critical role of SLC26A2-mediated sulfate transport in tooth development, providing insights into the mechanisms behind tooth abnormalities in patients with recessively inherited chondrodysplasias caused by SLC26A2 variants.

All animal experiments were performed in strict accordance with the guidelines of the Animal Care and Use Committee of the Osaka University Graduate School of Dentistry, Osaka, Japan. The protocol was approved by the Committee on the Ethics of Animal Experiments of Osaka University Graduate School of Dentistry. Welfare guidelines and procedures were performed with the approval of the Osaka University Graduate School of Dentistry Animal Committee (approval number: 3745, 29-033-0).

Pronuclear stage embryos from C57BL6/J mice were purchased from ARK Resource (Kumamoto, Japan). Recombinant Cas9 protein, crRNA and tracrRNA were obtained from Integrated DNA Technology. For the generation of Slc26a2-KO-Δexon2 mice, we used the Technique for Animal Knockout System by Electroporation (TAKE), as previously described (Inubushi et al., 2022). Embryos were washed twice with Opti-MEM solution (Gibco) and aligned in the electrode gap filled with 50 μl Cas9/gRNA (crRNA-tracrRNA complex)/single-stranded oligodeoxynucleotide (200/100/100 ng/μl) mixture. The intact embryos were subjected to electroporation using poring (225 V) and transfer (20 V) pulses. After electroporation, embryos were returned to KSOM Mouse Embryo Media (Millipore Sigma) at 37°C. Genome-edited two-cell embryos were transferred to pseudopregnant ICR female mice oviducts, and genomic DNA from newborn mice was analyzed by PCR. The sequence of gRNA was as follows: left, 5′-AGTCTGAGACCGGTCATGGC-3′; right, 5′-ACAATGAGCTCGACCGGAAT-3′. In all experiments, Slc26a2^wild/wild offspring were used as controls. Genotyping of mice and embryos was performed by PCR with the specific primers listed in Table S1. Similarly to previously reported Slc26a2 KO mouse models, the Slc26a2-KO-Δexon2 model utilizes CRISPR/Cas9 technology to target exon 3, with the insertion of a stop codon 137 bases downstream of the start codon. This modification results in a truncated protein that lacks both the transmembrane domain and critical regions necessary for enzymatic activity. Mice were fed ad libitum on solid feed and sterile water irradiated with ultraviolet light. The environmental conditions of the animal facility were maintained at constant temperature and humidity and kept under a 12 h light–dark cycle (08:00-20:00 as the light period).

Maxillary and mandibular tissues from control and Slc26a2-KO-Δexon2 mice at P0 were collected and fixed in 4% paraformaldehyde. A mild decalcifier, Osteosoft (Sigma-Aldrich), was used for tooth decalcification. Sagittal sections of paraffin-embedded mandible were prepared and used for Hematoxylin and Eosin (H&E) staining and in situ hybridization, as previously described (Sarper et al., 2018). The digoxigenin-labeled RNA probes used in this study were prepared using a DIG RNA Labeling Kit (Roche), according to the manufacturer's protocol, using each cDNA clone as the template. The probes were synthesized from fragments of Slc26a2 (Allen Institute for Brain Science) and were amplified with T7 and SP6 adaptor primers through PCR. After hybridization, the expression patterns for each mRNA were detected and visualized according to their immunoreactivity with anti-digoxigenin alkaline phosphatase-conjugated Fab fragments (Roche). A minimum of three embryos of each specimen type were examined per probe.

Frozen sections of 12 µm thickness were used for immunostaining and incubated with M.O.M (Mouse on Mouse) Blocking Reagent (Vector Laboratories), 5% goat serum/PBS and 0.1% sodium citrate buffer. Immunofluorescence staining was performed overnight at 4°C on 15 μm sections using polyclonal rabbit anti-Ki67 (anti-MKI67; 1:400; ab15580, Abcam). Sections were then counterstained with 4′,6-diamidino-2-phenylindole (DAPI; 1:500; Dojindo) and mounted using fluorescent mounting medium (Dako). At least three embryos were used for each genotype for each analysis. Apoptotic cells were identified using an in situ cell death detection kit (11684795910, Roche) according to the manufacturer's instructions.

We performed laser microdissection as previously reported (Sarper et al., 2018). Dissected heads were freshly mounted in Tissue-Tek O.C.T. Compound (Sakura Finetek, Japan) and immediately frozen. The tissue was then sectioned serially at a thickness of 20 µm using a cryostat (CM 1950, Leica). Sections were mounted on film-coated slides; whole sections were obtained consecutively from the anterior palate at E18.5 and stained with H&E. Odontoblasts from maxillary and mandibular molar were collected in tubes and separated from the sample sections with a manual laser capture microdissection system (LMD6500, Leica). Tissues were serially sectioned at −20°C with a thickness of 25 μm using a cryostat (CM 1950, Leica).

The protocol for RNA extraction and quantitative PCR (qPCR) analysis was as reported previously (Inubushi et al., 2012). Total RNA was extracted from dissected tissue using an RNeasy Mini Kit (Qiagen), following the manufacturer's protocol, with purity and quantity assessed by a Nanodrop spectrophotometer (Thermo Fisher Scientific). The extracted RNA was reverse transcribed to cDNA using an oligo dt with reverse transcriptase (Takara Bio). For real-time PCR, aliquots of total cDNA were amplified using Fast SYBR Green PCR Master Mix (Applied Biosystems) or Fast TaqMan Fast Universal PCR Master Mix (Applied Biosystems). Data were acquired and analyzed with a Step One Real-Time PCR System using Step One Software, Version 2.1 (Applied Biosystems). The PCR products were quantified using Gapdh as the reference gene. The mouse primers used in this study have been previously described (Inubushi et al., 2023). Other primers and probes are listed in Table S2. Each experiment was performed in triplicate.

RNA libraries were prepared using a TruSeq RNA Library Preparation Kit according to the manufacturer's protocol. Libraries were amplified by PCR and purified using AMPure XP beads. RNA-seq was performed using a sequencing system (Novaseq 6000, Illumina). The biological significance of differentially expressed genes was explored by volcano plot and GO enrichment analysis using DESeq2.

Control and Slc26a2-KO E18.5 upper molar tooth germ were implanted under the renal capsule of 8-week-old BALB/cSlc-nu/nu mice. Four weeks later, the implanted tooth embryos were harvested, and micro-CT was performed. Sagittal slice images (50 µm) were taken with VG Studio and the image processing software The roots were cut in ImageJ, and the volume of the crown dentin was measured.

A Blyscan sulfated glycosaminoglycan assay kit (Biocolor, Carrickfergus, UK) quantitatively measured sulfated proteoglycans and GAGs in biological samples. The assay was performed according to the manufacturer's instructions. Briefly, 4×10⁵ cells were plated in a T25 flask, and, 48 h later, the cells were treated with 1 μg/ml DS-500 or 50 mM NaClO₃. Twenty-four hours after treatment, the medium was aspirated, and the cells were rinsed with PBS. The cells were lysed in papain extraction reagent added to the cell monolayer for 3 h at 65°C. Total cell extract containing total GAGs was harvested, and samples were centrifuged at 10,000 g for 10 min. A total of 100 μl of the supernatant was used for the assay.

Mice were fixed in 95% ethanol overnight at room temperature. They were then left in acetone overnight at room temperature and incubated overnight in a cartilage staining solution containing 0.03% (w/v) Alcian Blue, 80% ethanol and 20% acetic acid. The first rinse was performed with several changes of 70% ethanol. To improve visibility of cartilage morphology, washings were terminated before cartilage was completely de-stained. Ossified tissue was stained with an Alizarin Red solution containing 0.005% (w/v) Alizarin Red in 1% (w/v) KOH for 4 h at room temperature. Samples were placed in a 50% glycerol solution containing 1% (w/v) KOH to remove excess staining.

Maxilla and mandible were collected from control and Slc26a2-KO-Δexon2 mice at P0. These tissues were fixed in 4% paraformaldehyde overnight. Embryos were placed in 70% ethanol for 1 day and 100% ethanol for 3 days, then in 100% ethanol with xylene (1:3) for 1 h. The embryos were then placed in 100%, 90%, 80% and 70% ethanol for 30 min each, before being placed in 70% ethanol containing 1% phosphotungstic acid for 1 week for contrast. Both maxilla and mandible were then scanned by micro-CT (R_mCT2, Rigaku) at 90 KV, 200 µA, microfocus 5 µm/voxel size. Volume Graphics (VGstudio) MAX 2.2 software was used for reconstruction of three-dimensional images. Measurements of the maxilla, incisors and molars were performed using VGstudio MAX 2.2 software.

A public single-cell RNA-seq dataset, GSE146855, of mouse incisors was downloaded from the Gene Expression Omnibus database to reanalyze the expression profile in odontoblast differentiation. The data were analyzed using a Seurat (version 4.0.5) package with R studio. A total of 6260 cells were reanalyzed. After normalization, scaling and principal component analysis of the data, cells were clustered using FindNeighbors (dims=1:6) followed by FindClusters (resolution=0.35). The RunUMAP function was used to visualize the cell clusters. Differential expression and cell identification were performed using FindAllMarkers (min.pct=0.25, logFC.threshold=0.25) with the Wilcoxon rank sum test. The top 5 differentially expressed features (cluster biomarkers) and cluster biomarkers are shown in Fig. S1. Visualization of Slc26a1, Slc26a2, Slc26a6, Slc26a7, Slc26a10 and Slc26a11 gene expression with a dot plot was generated with Seurat function DotPlot.

hDPSCs isolated from human adult third molars (Lonza) were cultured in Dulbecco's modified Eagle medium (Wako) containing 20% fetal bovine serum (Invitrogen) and 1% penicillin/streptomycin (Sigma-Aldrich), which was designated as growth medium (GM). To induce odontoblast differentiation, hDPSCs were cultured in odontoblast differentiation medium (EM) consisting of alpha Modified Eagle's Medium (Wako) containing 20% fetal bovine serum (Invitrogen) supplemented with 10 nM dexamethasone, 10 mM β-glycerophosphate and 50 μg/ml vitamin C (Sigma-Aldrich). To knock down SLC26A2 expression in hDPSCs, we used lentivirus-mediated shRNA transduction. Lentivirus particles expressing an shRNA that is validated to deplete human SLC26A2 (Mission shRNA, TRC Clone ID TRCN8607, MilliporeSigma) and control lentivirus particles expressing an shRNA that does not target any known genes (Mission shRNA, SHC005, MilliporeSigma) were purchased from MilliporeSigma. Lentivirus particles were added to hDPSCs cultured in growth medium supplemented with 5 μg/ml polybrene and cultured for 2 days. Cells transduced with lentiviral shRNAs were selected and maintained in the presence of 10 μg/ml puromycin.

Frozen frontal sections of the maxillary from control mice at P0 were stained with Alcian Blue for 30 min and rinsed with tap water for 5 min, followed by two changes of distilled water. The sections were dehydrated with ethanol, washed with xylene and mounted with Dako Fluorescent Mounting Medium (Agilent Technologies).

Statistical methods were not used to predetermine sample size. Statistical analyses were performed with GraphPad Prism 8. Unpaired two-tailed Student's t-tests and two-way ANOVA were used under the assumption of normal distribution and observance of similar variance. P<0.05 was considered significant. Bonferroni post hoc analysis was performed where applicable. Values are expressed as mean±s.d. For all quantitative experiments performed in this study, statistical analyses were conducted. Variances between groups were similar, and the data were symmetrically distributed. Data shown are representative images; each analysis was performed on at least three mice per genotype. Immunostaining was performed at least in triplicate. For other experiments, the numbers of biological replicates, animals or cells are indicated in the text. No randomization was used to allocate experimental units. There were no exclusion or inclusion criteria; all mutant mice were allocated to each experiment and there were no excluded animals. Confounders were not controlled in the experiments. For the quantitative measurements and statistical analysis, genotype information was masked.

We thank Ms Yuki Okamoto and Mayumi Yoshimoto for the excellent care and maintenance of our mouse colony and for valuable assistance in the histological, molecular and protein work.