Research Studies
Study Tracker
Urinary proteomic biomarkers of systemic fluoride exposure in children.
| Ecotoxicology and Environmental Safety | Pongsapipatana N, Gavila P, Cho SD, El-Tanani M, Islam S, Essigmann JM, Sriwattanapong K, Porntaveetus T. | 309:119643.
Children Endemic fluorosis areas Thailand Urine-F Human Study
View PDF

Abstract

Highlights

  • Elevated fluoride exposure alters urinary proteomes in schoolchildren.
  • Ten proteins identified: 6 upregulated, 4 downregulated in HF group.
  • SPP1 increased and TNC decreased, suggesting bone remodeling disruption.
  • Findings extend fluoride risks beyond dental fluorosis to systemic effects.

While optimal fluoride (F) levels support oral health, chronic exposure to high concentrations can lead to skeletal and dental fluorosis, especially in children. Emerging evidence suggests that excessive fluoride intake may disrupt systemic physiology, yet the underlying mechanisms remain poorly understood. To address this gap, we performed a comparative urinary proteomic analysis using LC-MS/MS in schoolchildren residing in a high-fluoride region, categorizing participants into high-fluoride (HF) and low-fluoride (LF) groups based on urinary fluoride excretion. Among 460 quantified proteins, ten were differentially expressed in the HF group, six upregulated (PHPT1, SPP1, COLEC12, CST4, DCHS1, LDHB) and four downregulated (CTSH, NECTIN1, TNC, KLK1). Gene Ontology enrichment highlighted associations with cell adhesion, ossification, and tissue development. Notably marked alterations in osteopontin (SPP1) and tenascin-C (TNC), key regulators of bone remodeling and dental matrix organization, suggesting disrupted focal adhesion signaling, impaired matrix integrity, and dysregulated biomineralization. Additional changes in proteins associated with enamel formation, oxidative stress, and immune regulation suggest that high fluoride exposure may broadly disrupt extracellular matrix organization and trigger inflammatory pathways. These findings demonstrate that excessive fluoride exposure induces systemic molecular disturbances in children, with implications for bone and soft tissue homeostasis. This research adds to the body of human evidence concerning fluoride’s biological impact, advocating for vigilant exposure monitoring.

Graphical Abstract

https://ars.els-cdn.com/content/image/1-s2.0-S0147651325019888-ga1_lrg.jpg

    Keywords

    Fluorides; Healthcare access; Toxicity; Proteomics; Urine; Environmental exposure

    1. Introduction

    Fluoride ion (F–) is a naturally occurring ion, with primary exposure routes including drinking water, fluoridated foods, and dental products such as toothpaste (Iamandii et al., 2024, Samaranayake et al., 2025). The World Health Organization (WHO) has established a maximum safe fluoride concentration of 1.5 ppm in drinking water (Du et al., 2022). However, more than 50 countries report fluoride levels exceeding this threshold, posing significant health risks (Du et al., 2022, Zhou et al., 2019). Similar to other emerging environmental contaminants such as micro- and nanoplastics, which have been shown to exert widespread systemic effects and raise global health concerns (Nasibova et al., 2025), excessive fluoride exposure represents an important environmental health issue requiring molecular-level investigation. Chronic overexposure to fluoride has been associated not only with dental and skeletal fluorosis but also with a broad spectrum of systemic disorders, including neurological, reproductive, hepatic, immunological, cardiovascular, and renal dysfunctions, highlighting fluoride’s potential to affect multiple biological systems (Bittencourt et al., 2023, Chhabra et al., 2025, Grandjean, 2019, Sharma et al., 2017, Xie et al., 2020). Studies indicate that childhood fluorosis can disrupt bone development, potentially leading to long-term skeletal abnormalities and increased susceptibility to fractures (Saeed et al., 2020, Zhou et al., 2019). Several studies in both humans and animals have demonstrated that excessive fluoride exposure alters the physical and chemical properties of enamel cells, bone minerals, and bone cells, while also influencing bone remodeling by enhancing osteoblast activity and delaying the mineralization of new bone (Ajrithirong et al., 2025, Godebo et al., 2020, Linghu et al., 2023, Liu et al., 2016, Yao et al., 2019).

    Fluoride has long been recognized for its beneficial role in caries prevention, however, excessive and prolonged intake is known to cause dental and skeletal fluorosis (Peckham and Awofeso, 2014, ten Cate and Buzalaf, 2019). Accordingly, numerous studies have evaluated fluoride levels in plasma, saliva, urine, hair, and nails as biomarkers of recent or cumulative exposure (Baez et al., 2014, Eskandari et al., 2023, Idowu et al., 2020, Rigo et al., 2018). These biomarkers, particularly fluoride concentrations in 24-h urine, plasma, and toenails, show strong correlations with total daily fluoride intake (TDFI) and have been widely used to estimate exposure patterns and assess the risk of fluorosis in children (Idowu et al., 2020). Urine is considered the most practical indicator of short-term exposure at the population level, as it rapidly reflects fluctuations in fluoride intake and aligns with WHO-recommended reference values for 24-h urinary fluoride excretion (Baez et al., 2014, Idowu et al., 2020). These biomarkers have been valuable for characterizing exposure patterns and establishing associations between environmental fluoride levels and internal fluoride burden. However, these methods primarily quantify fluoride concentrations and lack the sensitivity to detect early or subtle biological alterations. To address this gap, the present study employs LC-MS/MS-based urinary proteomics as a non-invasive approach capable of detecting molecular-level changes associated with chronic fluoride exposure, providing deeper insight into fluoride-induced systemic effects.

    Urinary protein detection, particularly albumin, has long been central in diagnosing renal and urinary tract disorders, using conventional methods such as dipstick tests, enzymatic assays, Jaffe’s reaction, immunoassays, turbidimetry, nephelometry, and automated urinalysis (García et al., 2025, Hodel et al., 2025, Supornsilchai et al., 2025). While albumin is the main clinical marker, urine contains thousands of lower-abundance proteins. Recent proteomic advances enable identification of more than 1000 proteins in normal urine (Joshi et al., 2024, Zhao et al., 2017), supporting its potential as a non-invasive platform for biomarker discovery, including conditions such as fluorosis (Hwang et al., 2024, Joshi et al., 2024, Li et al., 2025)

    Although the dental and skeletal manifestations of fluoride toxicity are well documented, our mechanistic understanding relies heavily on animal studies, leaving significant gaps in regarding the impact of chronic fluoride exposure affects human biological systems. Preclinical proteomic investigations have demonstrated that fluoride disrupts critical pathways involved in collagen organization, extracellular matrix (ECM) remodeling, oxidative stress, and endocrine and immune regulation (Bittencourt et al., 2023, El-Tanani et al., 2025, Miranda et al., 2022, Nagendra et al., 2022, Xie et al., 2020). Studies across various tissues indicate that medium- to high-dose fluoride exposure markedly alters the expression of proteins governing collagen metabolism, proteoglycans, and matrix metalloproteinases which are molecules essential for maintaining bone integrity and marrow function (Kobayashi et al., 2014, Kobayashi et al., 2011, Pereira et al., 2013, Wei et al., 2018). Furthermore, urinary proteomic analyses in murine models have revealed fluoride-induced changes in androgen-regulated proteins and detoxification enzymes, suggesting potential impacts on reproductive and endocrine pathways through disrupted hormone regulation and detoxification processes (Kobayashi et al., 2011). Despite these advances, human-based molecular evidence remains limited, particularly in pediatric populations where actively developing bone and dental tissues render children uniquely susceptible to systemic toxicity.

    While literature on the structural impact of fluoride is extensive, proteomic investigations into human biological fluids following chronic exposure are scarce. A recent study examining salivary proteomics in patients with dental fluorosis identified associations with proteins related to the cystic fibrosis transmembrane conductance regulator (CFTR) ion channel, suggesting a potential role for CFTR dysregulation in the pathogenesis of severe fluorosis (Gavila et al., 2024). This study also reported decreased levels of immune-related salivary proteins, indicating possible impairment of local immune defense mechanisms in advanced fluorosis. Recognizing this knowledge gap regarding the broader systemic effects of chronic high fluoride exposure, we employed a proteomic approach to investigate fluoride-induced molecular alterations in urine samples. This study focused on characterizing the urinary proteomic profiles of school-aged children (6–14 years old) residing in areas with documented elevated fluoride levels in the drinking water. By comparing the proteomic profiles of individuals with high and low fluoride exposure, as determined by urinary fluoride concentrations, this study aimed to characterize the molecular signatures and underlying pathomechanisms associated with established chronic toxicity.

    Our findings revealed alterations in proteins associated with bone remodeling, enamel formation, oxidative stress, and inflammation, suggesting that high fluoride exposure may lead to skeletal and dental abnormalities potentially through disruption of the extracellular matrix and activation of the immune response.

    2. Materials and methods

    2.1. Subject recruitment and study design

    Participants were recruited from fluoride-endemic regions in Ratchaburi Province, Thailand. Schools and community health centers in these areas helped identify potential participants aged 6–14 years, an age range corresponding to the mixed dentition period and active dental development that is highly susceptible to fluorosis (Baez., 2014). Information sessions were held to explain the study’s purpose, procedures, and potential benefits to both parents and children. Before enrollment, informed consent was obtained from parents or guardians, along with assent from the children.

    Participants were categorized into two groups based on their calculated daily urinary fluoride excretion (DUFE) in relation to the WHO guidelines (Baez., 2014) for children aged 6–14 years. The low fluoride (LF) group, serving as the control group, consisted of children with a DUFE between 0.182 to 1.090 mg/day. The high fluoride (HF) group, the study group, comprised children with a DUFE exceeding 1.090 mg/day.

    2.2. Ethical approval

    The study received approval from the Human Research Ethics Committee of the Faculty of Dentistry, Chulalongkorn University, Thailand (HREC-DCU 2021–061, approved on October 1, 2021) and adhered to the 1964 Helsinki Declaration and its subsequent amendments.

    2.3. Sample collection and fluoride quantification in urine

    Urine samples (HF group: n = 10 and LF group: n = 10) were collected during scheduled school visits to ensure a standardized and consistent sampling process. Sample collection and fluoride quantification procedures followed the validated methodology described in (Baez et al., 2014, Gavila et al., 2024). Briefly, the 24-hour urine samples were obtained from each participant using designated collection containers. On the morning of the collection, the first urine was discarded, and all subsequent urine samples were collected in wide-necked containers. The samples were then transferred into screw-capped collection bottles using a funnel. To maintain sample integrity during transportation, the bottles were stored at 4 °C. Fluoride concentrations were measured using a fluoride electrode (4-star benchtop, Orion, USA) in conjunction with total ionic strength adjustment buffer, containing cyclohexylenedinitrilotetra acetate, sodium hydroxide, sodium chloride, and acetic acid dissolved in deionized water.

    2.4. Liquid chromatography with tandem mass spectrometry (LC–MS/MS)

    Urinary protein extraction was adapted from a previous proteomic workflow (Gavila et al., 2024), with modifications to accommodate urine samples. Briefly, urine samples were centrifuged at 14,000 × g for 20 min at 4°C. Proteins were precipitated from 1 mL of the supernatant using 100 uL of 100 % trichloroacetic acid on ice for 30 min, followed by centrifugation at 20,000 × g for 20 min. After removing the supernatant, the pellet was washed with 500 uL of acetone, vortexed, and centrifuged again at 20,000 × g for 10 min. This washing step was repeated twice. The final pellet was resuspended in 300 uL of 8?M urea in 100 mM triethylammonium bicarbonate (TEAB) containing Halt protease inhibitor cocktail (Thermo Fisher Scientific, CA, USA).

    Proteomic analysis using LC-MS/MS was conducted at the Center of Excellence in Systems Biology, Faculty of Medicine, Chulalongkorn University, following their validated analytical protocol (Makjaroen et al., 2018, Techawiwattanaboon et al., 2023) with some modification. Briefly, protein samples (100 uL) were treated with 10 uL of dithiothreitol (DTT) and incubated at 37°C for 30?min. The samples were then reacted with 40 uL of 100 mM iodoacetamide (IA) in the dark at room temperature (RT) for 30 min. Subsequently, 40 uL of 100 mM DTT was added, and the mixture was incubated at RT for at least 15 min. To lower the urea concentration from 8 M to below 1 M, the samples were diluted with 100 mM TEAB. Trypsin was added at a 1:50 (w/w) ratio, and the samples were incubated at 37°C for 16 h. To stop the reaction, 100 % TFA was added to achieve a final concentration of 0.5 %, and the samples were incubated at RT for 15 min before being dried using Speed-Vac. The quantity of tryptic peptides was determined using the Pierce Quantitative Fluorometric Peptide Assay (Thermo Fisher Scientific, CA, USA). Peptides were labeled with TMT10plex™ Isobaric Reagents (Thermo Fisher Scientific, CA, USA) and fractionated using the High pH Reversed-Phase Peptide Fractionation Kit (Pierce). Fractionated peptide samples were dried and stored at -80°C.

    Before analysis, samples were resuspended in 0.1 % formic acid (FA) and analyzed using LC-MS/MS. The EASY-nLC 1000 system (Thermo Fisher Scientific, CA, USA) was connected to a Q-Exactive Orbitrap Plus mass spectrometer (Thermo Fisher Scientific, CA, USA) equipped with a 25 cm EASY-Spray C18 column (75 um internal diameter). Peptides were injected and eluted using a 90-minute gradient at a constant flow rate of 300 nL/min (0–5 % solvent B in 0 min, 5–20 % B in 60 min, 20–40 % B in 23 min, 40–95 % B in 2 min, and 95 % B in 5 min). An electrospray voltage of 2.0 kV was applied. The mass spectrometer was operated in data-dependent acquisition (DDA) mode. Precursor spectra (MS1) were acquired in the Orbitrap over an m/z range of 350–1400 at a resolution of 70,000, with an AGC target of 3 × 106 and a maximum injection time of 250 ms. From each full scan, the 10 most intense precursor ions were selected for fragmentation using a normalized collision energy (NCE) of 32. The resulting fragment ions (MS2) were analyzed in the Orbitrap at a resolution of 35,000, with an AGC target of 1 × 105 and a maximum injection time of 100 ms.

    2.5. Data analysis and visualization

    For proteomic analysis, the raw mass spectrometry files were searched against the human Swiss-Prot database, with a list of common protein contaminants included. Data processing was performed using Proteome Discoverer™ Software 2.1 (Thermo Fisher Scientific). Precursor and fragment ion mass tolerances were set at 10 ppm and 0.02 Da, respectively. The false discovery rate for identified peptides was set at 0.01 (Techawiwattanaboon et al., 2023). Reporter ions intensity ratios for the control group (LF) and study group (HF) were transformed to log2 and normalized by the median of each sample. Statistical significance was assessed using independent t-tests, based on a minimum of three valid log2 values per group. Proteins with P-value < 0.05 and log2(FC) greater than 0.5 or less than -0.5 were considered significantly altered in expression (Ting et al., 2009).

    Functional analysis of the identified proteins were conducted using DAVID Bioinformatics Resources (https://davidbioinformatics.nih.gov/summary.jsp) (Sherman et al., 2022) to assess Gene Ontology (GO) enrichment, examining functional classifications related to biological processes (BP), molecular functions (MF), and cellular components (CC). Pathway analysis was performed using the KEGG (Kyoto Encyclopedia of Genes and Genomes) database (https://www.kegg.jp/kegg/pathway.html) (Kanehisa et al., 2023), Reactome (https://reactome.org/) (Milacic et al., 2023), and WikiPathways (https://www.wikipathways.org/) (Agrawal et al., 2023). Additionally, protein–protein interactions (PPIs) was explored using STRING (https://string-db.org/) (Szklarczyk et al., 2023). Data visualization was performed using SRplot, generating volcano plots and clustered heatmaps to illustrate differential protein expression patterns (Tang et al., 2023). A Venn diagram was conducted by Venny 2.1.0 (https://bioinfogp.cnb.csic.es/tools/venny/index.html) (Oliveros, 2007-2015). The study was summarized in the graphical illustration in Fig. 1.

    Fig. 1

    1. Download: Download high-res image (508KB)
    2. Download: Download full-size image

    Fig. 1. Graphical illustration summarizing the study.

    To contextualize our findings and support the selection of analytical methods, a focused literature search was conducted using PubMed and Google Scholar. We used specific keywords related to fluoride exposure (e.g., fluoride toxicity,’ ‘dental fluorosis,’ ‘skeletal fluorosis’), urinary proteomics (e.g., ‘urinary proteome,’ ‘non-invasive biomarkers’), and biological pathways relevant to bone remodeling and extracellular matrix regulation (e.g., ‘SPP1,’ ‘TNC,’ ‘biomineralization proteins’). The search included studies published up to October 2025 to capture both foundational toxicology research and recent proteomic advances.

    2.6. Statistical analysis

    SPSS software version 22 (SPSS Inc., Chicago, IL, USA) and Microsoft Excel were employed for statistical analyses. Nonparametric Mann-Whitney U-test and Fisher’s Exact Test were applied to assess differences in age and sex distribution between groups, respectively. Independent t-tests were performed to evaluate statistical significance in the proteomic analysis.

    3. Results

    3.1. Participants characteristics

    The LF group, consisting of 4 males and 6 females (n = 10), exhibited a DUFE of 0.49 ± 0.21 mg/day, while the HF group, comprising 3 males and 7 females (n = 10), demonstrated a DUFE of 2.03 ± 0.50 mg/day. No significant differences were observed in age or sex distribution between the groups. However, fluoride concentrations differed significantly between the HF and LF groups (Table S1).

    3.2. Urine proteomic profiles of LF and HF groups

    Urine proteomic expressions of the LF and HF groups were identified as a total of 474 proteins. After removing contaminants and filtering out missing data, 460 proteins were retained for further analysis (Fig. 2A). A Venn diagram revealed that 396 proteins were shared between LF and HF groups, with 10 proteins significantly up-regulated and 7 proteins significantly down-regulated (Fig. 2B). Hierarchical clustering analysis of the data generated a heatmap that displayed distinct patterns of protein expression between the two groups (Fig. 2C).

    Fig. 2

    1. Download: Download high-res image (719KB)
    2. Download: Download full-size image

    Fig. 2. Data filtering, sample selection and analyses. (A) Flowchart of data filtering based on the criterion of 24-hour urinary fluoride excretion (DUFE) ranging from 0.182 to 1.090 mg F/day (LF group) and exceeding 1.090 mg F/day (HF group) for individuals aged 6–14 years; (B) Venn diagram depicting the total proteins, highlighting 17 significant proteins; (C) Cluster heatmap showing the distribution of these 17 significant proteins (p-value < 0.05). LF; Low fluoride consumption group (n = 10) and HF; High fluoride consumption group (n = 10); (D) Volcano plot indicated 17 proteins with 6 up-regulated (pink spots), 4 down-regulated (blue spots) and 7 proteins with log2FC lower than 0.5 or higher than -0.5 (gray spots). Among 396 proteins, 6 proteins (E-J) were significantly up-regulated in HF group (p-value < 0.05 and a log2FC > 0.5) and 4 proteins (K-N) were significantly down-regulated in HF group (p-value < 0.05 and a log2FC < -0.5). The graphs presented as mean ± SEM. The asterisk (*) indicates a statistically significant difference (p < 0.05).

    A volcano plot was utilized to highlight differentially expressed proteins, revealing 17 significant candidates. Among these significant proteins (Table 1), ten proteins exhibited a log2 fold change (FC) greater than 0.5 or less than -0.5. Six of these proteins were upregulated comprising 1) phosphohistidine phosphatase (PHPT1), 2) osteopontin (SPP1), 3) collectin-12 (COLEC12), 4) cystatin-S (CST4), 5) protocadherin-16 (DCHS1), and 6) L-lactate dehydrogenase B chain (LDHB) (Figs. 2D and 2E-2J). Additionally, four proteins were down-regulated: 1) pro-cathepsin H (CTSH), 2) nectin-1 (NECTIN1), 3) tenascin (TNC), and 4) kallikrein-1 (KLK1) (Figs. 2D and 2K-2N).

    Table 1. A list of 17 differentially expressed proteins found in both low and high fluoride consumption groups.

    No. Proteins Symbols UniProtKB entry High vs low fluoride consumption groups
    Change p-valuea Log2 FC FC
    1 Phosphohistidine phosphatase PHPT1 Q9NRX4 Up 0.0004 0.911 1.88
    2 Osteopontin SPP1 P10451 Up 0.006 0.572 1.49
    3 Collectin-12 COLEC12 Q5KU26 Up 0.013 0.665 1.59
    4 Cystatin-S CST4 P01036 Up 0.020 0.683 1.61
    5 Protocadherin-16 DCHS1 Q96JQ0 Up 0.022 0.513 1.43
    6 L-lactate dehydrogenase B chain LDHB P07195 Up 0.033 0.674 1.60
    7 Pro-cathepsin H CTSH P09668 Down 0.013 -0.546 0.68
    8 Nectin-1 NECTIN1 Q15223 Down 0.013 -0.561 0.68
    9 Tenascin TNC P24821 Down 0.014 -0.746 0.60
    10 Kallikrein-1 KLK1 P06870 Down 0.016 -0.562 0.68
    7 proteins with log2FC lower than 0.5 or more than -0.5
    11 Twisted gastrulation protein homolog 1 TWSG1 Q9GZX9 Up 0.014 0.276 1.21
    12 Nectin-4 NECTIN4 Q96NY8 Up 0.022 0.415 1.33
    13 Isoform 2C2A of Collagen alpha-2(VI) chain COL6A2 P12110–2 Up 0.031 0.440 1.36
    14 Immunoglobulin superfamily member 8 IGSF8 Q969P0 Up 0.044 0.260 1.20
    15 Thyroxine-binding globulin SERPINA7 P05543 Down 0.018 -0.463 0.73
    16 Transthyretin TTR P02766 Down 0.032 -0.373 0.77
    17 Cubilin CUBN O60494 Down 0.050 -0.335 0.79
    a
    Differences between groups were analyzed using an independent t-test, with significance set at p-value < 0.05. Proteins exhibiting a log2 fold-change (log2FC) of either > 0.50 or <-0.50 were considered differentially expressed; FC; Fold-change.

    3.3. Gene ontology (GO) and pathway analysis

    GO and pathway analysis were conducted on 17 significant proteins to gain a deeper understanding of their biological roles, molecular functions, and involvement in cellular pathways. BP with the most significant p-values highlightened enrichment in morphogenesis (TNC, CTSH, DCHS1, TWSG1, NECTIN1), tissue development (IGSF8, TWSG1, SPP1, TNC, CTSH, DCHS1, NECTIN1), and cell adhesion processes (COL6A2, SPP1, TNC, DCHS1, NECTIN4, NECTIN1) (Fig. 3A and Table S2). Other proteins were associated with regulation of peptidase activity (CTSH, SERPINA7, CST4) and multicellular activity (IGSF8, CUBN, TWSG1, KLK1, SPP1, TNC, CTSH, DCHS1, CST4, NECTIN1). COLEC12 and PHPT1 were associated with cell signaling and immune function. In terms of CC, 94.12 % of the analyzed proteins were localized to the extracellular region (Fig. 3B and Table S3). The MF of these genes focus on hormone binding and transport (TTR, CTSH), cell adhesion and communication (SPP1, TNC, DCHS1, NECTIN1, COLEC12, COL6A2, PHPT1), proteolysis regulation (CTSH, SERPINA7, CST4), and protein complex interactions (TTR, SPP1, CTSH, PHPT1, SERPINA7, CST4). These functions play critical roles in cellular signaling, structural integrity, and systemic regulation (e.g., thyroid hormone metabolism and immune response) (Fig. 3C and Table S4).

    Fig. 3

    1. Download: Download high-res image (784KB)
    2. Download: Download full-size image

    Fig. 3. Functional annotation of 17 significant proteins. (A) Biological processes; (B) Cellular compartment; (C) Molecular function; (D) Reactome pathway; (E) WIKI pathway and (F) KEGG pathway.

    Pathway analysis was performed using three databases: Reactome, KEGG and WIKI pathway. The results indicated that these pathways were strongly associated with tissue repair and development, structural integrity, cell adhesion and intracellular signaling. Key pathways included integrin cell surface interactions, extracellular matrix organization, Nectin/Necl trans heterodimerization, ECM-receptor interaction, focal adhesion, mTOR signaling, and PI3K Akt signaling (Fig. 3D-3F and Table S5). These pathways shared common proteins, including COL6A2, SPP1, TNC, which are crucial for maintaining tissue architecture, mediating growth signals, and responding to environmental stimuli.

    3.4. Protein-protein interaction analysis

    Based on BP analysis conducted using DAVID, four key processes, including tissue development, ossification, cell adhesion, and response to vitamin D were identified as critical for the growth, repair, and overall health of bones and teeth. A Venn diagram illustrated the proteins associated with these processes, including COL6A2, SPP1, TNC, DCHS1, NECTIN4, NECTIN1, TWSG1, IGSF8, and CTSH (Fig. 4A). The analysis revealed that SPP1 and TNC are involved in all four processes, highlighting the central role of these proteins in processes related to bones and tooth health. Subsequently, SPP1 and TNC were further examined using STRING to explore potential interactions and functional relationships.

    Fig. 4

    1. Download: Download high-res image (876KB)
    2. Download: Download full-size image

    Fig. 4. The Venn diagram and STRING analysis highlighted two proteins, SPP1 (Osteopontin) and TNC (Tenascin). (A) Venn diagram of proteins involved in four biological processes: cell adhesion, ossification, tissue development, and response to vitamin D; (B) STRING analysis of SPP1 (Osteopontin) and TNC (Tenascin). The core proteins (SPP1 and TNC) were highlighted with dark blue circles. A group of proteins involved in Biomineralization was highlighted with red circles; (C) STRING analysis of proteins involved in Biomineralization process. The core proteins, SPP1 (Osteopontin), FAM20A (Pseudokinase) and FAM20C (Extracellular serine/threonine protein kinase), were highlighted with dark blue circles.

    The expanded STRING analysis of the two proteins, SPP1 and TNC, revealed their critical roles in adhesion and biomineralization (Fig. 4B). Interactions with integrins (ITGB5, ITGB6, ITGB8, ITGA8, ITGA9, ITGA10, ITGA11), components of the focal adhesion: PI3K-Akt-mTOR signaling pathway, were identified. Moreover, SPP1 was shown to interact with FAM20A and FAM20C, key players in biomineralization. The interactions between SPP1, FAM20A, FAM20C, CEMP1, and integrins were further implicated in enamel mineralization and amelogenesis imperfecta. To gain deeper insight into biomineralization, a STRING analysis was performed focusing on SPP1, FAM20A, and FAM20C as core proteins (Fig. 4C). This analysis uncovered additional interactions with other extracellular matrix proteins such as dentin matrix acidic phosphoprotein 1 (DMP1) and matrix extracellular phosphoglycoprotein (MEPE), as well as Fibroblast growth factor 23 (FGF23), and structural enamel matrix proteins (EMPs) like amelogenin (AMELX), ameloblastin (AMBN), amelotin (AMTN), and enamelin (ENAM). These findings underscore the pivotal roles of SPP1 and TNC in dental health, particularly in processes related to enamel formation and mineralization.

    4. Discussion

    Urinary proteomic analysis of school-aged children residing in high-fluoride areas revealed significant alterations in protein expression, with six upregulated proteins (PHPT1, SPP1, COLEC12, CST4, DCHS1, and LDHB) and four downregulated proteins (CTSH, NECTIN1, TNC, and KLK1). These differentially expressed proteins suggest potential molecular mechanisms contributing to fluoride toxicity, particularly concerning bone and tooth development, extracellular matrix remodeling, and immune responses. Considering the established association between chronic high fluoride exposure and skeletal/dental fluorosis, alongside emerging evidence of systemic inflammation, these protein changes warrant further investigation as potential biomarkers and therapeutic targets in fluoride-induced pathology.

    The disruption of proteins related to bone remodeling and enamel formation in response to fluoride exposure is evident in our findings. The upregulation of SPP1, a key regulator of bone remodeling, osteoclast differentiation, proliferation, adhesion, immune response, and inflammation, suggests the promotion of osteoclast differentiation and inhibition of osteoclast apoptosis, leading to excessive bone resorption and bone loss (Sapthanakorn et al., 2025, Si et al., 2020). This finding is consistent with fluoride-induced osteosclerosis, as supported by previous studies (Zhu et al., 2022). Furthermore, fluoride exposure has been shown to reduce bone quality and increase fragility due to disrupted osteoblast and osteoclast activity (Du et al., 2022, Simon et al., 2014). The overexpression of tissue factor pathway inhibitor-2 (TFPI-2) and subsequent upregulation of DCHS1, a member of the cadherin superfamily involved in cellular adhesion and tissue integrity, has been shown to promote ectopic calcification of human renal interstitial fibroblasts (Liu et al., 2024). DCHS1 interacts with FAT4 to regulate skeletal morphogenesis and osteoblast differentiation (Crespo-Enriquez et al., 2019, Mao et al., 2016). Thus, the upregulation of DCHS1 may further indicate structural remodeling in response to fluoride toxicity.

    NECTIN1, which plays a role in cell adhesion and tissue integrity, mediates Ca² -independent cell–cell adhesion, epithelial morphogenesis, and organogenesis (Yoshida et al., 2010). Reduced NECTIN1 expression in ameloblasts could impair enamel formation (Barron et al., 2008). The downregulation of NECTIN1 may therefore lead to fluoride-induced disruption in epithelial and mesenchymal cell communication, resulting in defective enamel formation. Additionally the downregulation of TNC suggests a weakened regenerative response, as TNC is involved in extracellular matrix remodeling and tissue repair (Midwood et al., 2016). The decrease in TNC expression impairs osteoblast differentiation, collagen synthesis, fibroblast motility, and matrix remodeling, which may contribute to bone and tooth abnormalities (Hasegawa et al., 2020, Morgan et al., 2011). Fluoride exposure impairs collagen synthesis and bone matrix integrity (Gupta et al., 2016), decreased TNC expression may reflect impaired bone healing and increased susceptibility to skeletal deformities.

    Fluoride toxicity has been linked to metabolic dysregulation, particularly alterations in glycolysis and oxidative stress responses (Agalakova and Gusev, 2012). The upregulation of LDHB suggests a shift in cellular metabolism, potentially as a compensatory response to fluoride-induced oxidative stress. Children exposed to fluoride levels over 2.0 mg/L have elevated LDH levels, indicating mitochondrial dysfunction and increased lactate production (Xiong et al., 2007). This metabolic shift may exacerbate tissue damage, as excessive lactate accumulation could amplify inflammatory responses and impair cellular function. Furthermore, the overexpression of COLEC12, a pattern recognition receptor involved in innate immune responses, was positively correlated with immune cell infiltration, similar to our findings in the HF group (Sun et al., 2023). The increased expression of COLEC12 in HF group may dysregulate the immune response, contributing to systemic inflammation and exacerbating the adverse effects of fluoride toxicity.

    CST4 is a critical regulator of protease activity in the saliva and oral mucosa. It modulates inflammatory responses, protects oral tissues from excessive proteolysis, and maintains extracellular matrix homeostasis (Wang et al., 2019). Elevated CST4 expression has been associated with inflammatory conditions, such as periodontitis (Zhang and Zhan, 2023). In our proteomic findings, CST4 was found to be upregulated, and may influence the inflammatory responses. PHPT1, an enzyme responsible for dephosphorylating histidine residues, plays a role in various signaling pathways, including ion transport, cell migration, and immune responses (Ek et al., 2015, Ning et al., 2024). Upregulation of PHPT1 could activate the extracellular signal-regulated kinase/mitogen-activated protein kinase (ERK/MAPK) pathway, promoting cell proliferation and tumorigenesis (Ning et al., 2024). The induction of PHPT1 by fluoride may disrupt cellular signaling, leading to increased cell proliferation, oxidative stress, and inflammation.

    In contrast, the downregulation of CTSH and KLK1 suggests disruptions in proteolytic activity and inflammatory regulation. Cathepsins are essential for immune cell activation and extracellular matrix degradation, and their reduced expression may impair tissue remodeling, prolong inflammation and delaying wound healing (Jevnikar et al., 2013). Similarly, KLK1 is a key regulator of vascular homeostasis and inflammatory signaling (Alexander-Curtis et al., 2019). Thereby, fluoride exposure may disrupt the kallikrein-kinin system, leading to increased oxidative stress and inflammatory damage.

    Furthermore, our STRING analysis revealed interactions between SPP1, TNC, and adhesion molecules, particularly in focal adhesion via the PI3K-Akt-mTOR pathway. Both proteins are associated with FAM20A, an activator of FAM20C, which plays a crucial role in biomineralization by phosphorylating matrix proteins that aid enamel deposition (Sriwattanapong et al., 2024, Sriwattanapong et al., 2024, Xu et al., 2021). FAM20C knockout mice exhibited reduced bone mineralization, an expansion of hypertrophic cartilage, upregulation of osteoclast differentiation genes, and downregulation of osteogenesis-related genes, including Mepe, Col2, Osx, Ocn, and Alp (Jiang et al., 2024). Proteomic analysis of gingival fibroblasts from patients with enamel renal syndrome due to FAM20A mutations revealed alterations in extracellular matrix organization, collagen fibril assembly, and biomineralization, which similar to our findings. Additionally, this study identified changes in TNC protein expression in patients (Simancas Escorcia et al., 2021). Furthermore, SPP1, FAM20A, and FAM20C interact with extracellular matrix proteins such as DMP1, MEPE, and FGF23, which regulate phosphate metabolism and bone function. DMP1 and MEPE are involved in bone mineralization (Donmez et al., 2022, Staines et al., 2012). These findings highlight the interconnected roles of TNC, SPP1, FAM20A, and FAM20C in regulating mineralization and modulating extracellular matrix proteins. The interactions among these proteins suggest a complex regulatory system governing phosphate homeostasis and bone mineralization, which may be disrupted by high fluoride exposure. Elevated fluoride levels downregulate TNC, defecting focal adhesion and potentially interacting with FAM20A and FAM20C, thereby influencing extracellular matrix remodeling and osteoblast function. These disruptions may impair bone repair and remodeling. Furthermore, excessive SPP1 upregulation could stimulate osteoclast activity, leading to increased bone resorption and loss (Fig. 5).

    Fig. 5

    1. Download: Download high-res image (608KB)
    2. Download: Download full-size image

    Fig. 5. The illustration depicts the impact of elevated fluoride exposure on SPP1 (Osteopontin) and TNC (Tenascin) expression, highlighting its effects on focal adhesion and biomineralization.

    This study focused on children within the dental development age range in high-fluoride areas, representing the critical period most susceptible of tooth formation to fluoride exposure. All samples were collected under standardized conditions to ensure consistent exposure assessment. However, we recognize that the sample size is limited, and therefore the findings should be interpreted as preliminary. In addition, although a stringent FDR threshold was applied during protein identification, multiple-testing correction was not applied during differential expression analysis due to the exploratory design and small cohort size; larger studies will be needed to confirm the robustness of these protein-level differences using stricter statistical thresholds. SPP1, TNC, and other differentially expressed proteins identified here are promising candidates for future validation studies in larger cohorts. Larger studies will be important to validate these results and further clarify their clinical implications.
    In conclusion, our findings suggest potential pathomechanisms by which chronic high fluoride exposure may contribute to systemic, skeletal, and dental abnormalities. This involves the dysregulation of proteins implicated in biomineralization, immune modulation, and metabolic processes. This study offers novel molecular insights into the disruption of bone and enamel formation, the potential promotion of inflammatory responses, and possible contributions to skeletal deformities following high fluoride exposure. These results underscore the need for deeper investigation into the systemic consequences of chronic fluoride exposure and for identifying potential therapeutic strategies to mitigate its adverse effects on bone and dental health, particularly in vulnerable populations. At the same time, our findings demonstrate the biomedical value of urinary proteomics as a non-invasive approach for detecting early molecular disturbances caused by high fluoride intake. The observed alterations, especially in proteins related to biomineralization, extracellular matrix regulation, and inflammatory pathways, may function as early indicators of skeletal or dental impairment before clinical symptoms emerge. Together, these insights highlight the promise of urinary proteomic markers to inform future diagnostic, monitoring, and preventive strategies aimed at reducing the burden of chronic fluoride toxicity.

    CRediT authorship contribution statement

    Nawapan Pongsapipatana: Writing – review & editing, Writing – original draft, Visualization, Validation, Formal analysis, Conceptualization. Patcharaporn Gavila: Writing – review & editing, Investigation. Sung-Dae Cho: Writing – review & editing. Mohamed El-Tanani: Writing – review & editing. Syed Arman Rabbani: Writing – review & editing. Sofiqul Islam: Writing – review & editing. John M. Essigmann: Writing – review & editing. Kanokwan Sriwattanapong: Writing – review & editing, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization. Thantrira Porntaveetus: Writing – review & editing, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization.

    Declaration of Competing Interest

    The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Prof. Dr. Thantrira Porntaveetus reports financial support was provided by Health Systems Research Institute, Thailand. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

    Acknowledgements

    NP was supported by the Faculty of Dentistry (grant number 2320042000), Chulalongkorn University. KS was supported by the Faculty of Dentistry (DRF 68_006), the Ratchadaphiseksomphot Endowment Fund, Chulalongkorn University (grants for development of new faculty staff; DNS_67_042_3200_002), and Thailand Science Research and Innovation Fund Chulalongkorn University (HEA_FF_69_051_3200_004). TP was supported by Thailand Science Research and Innovation Fund Chulalongkorn University (HEA_FF_69_036_3200_003).

    Appendix A. Supplementary material

    Data availability

    Data is provided within the manuscript or supplementary information files.

    References