Melatonin alleviates fluoride-induced developmental neurotoxicity by restoring SIRT3/HIF-1a axis-mediated mitochondrial dysfunction and reversing energy metabolism reprogramming.

Fluoride is ubiquitously distributed in the natural environment. Groundwater fluoride concentrations in many regions, including central Australia, western North America, eastern Brazil, and extensive areas of Africa and Asia, exceed the World Health Organization’s recommended limit of 1.5 mg/L (Podgorski and Berg, 2022). Human exposure occurs primarily through the gastrointestinal tract, respiratory tract, and dermal contact (Mini Vijayan et al., 2025). While appropriate fluoride intake benefits bone mineralization and prevents dental caries, excessive exposure poses serious health risks (Zhou et al., 2023). Notably, fluoride can cross the placental and blood-brain barriers, directly targeting the developing brain, which is highly vulnerable to environmental toxicants during this critical period (Grandjean, 2019; Malin et al., 2024). Epidemiological evidence indicates that low-level fluoride exposure during early life can adversely affect children’s intelligence and cognition, and prenatal exposure is associated with reduced cognitive ability without a threshold effect observed (Veneri et al., 2023). Consistent with human data, animal experiments demonstrate that perinatal fluoride exposure diminishes learning, memory, and memory retention in offspring, accompanied by hippocampal structural alterations and disrupted neurotransmitter homeostasis (Zhang et al., 2025; Zhu et al., 2024). Fluoride-induced neurotoxicity involves multifaceted mechanisms, including oxidative stress, inflammatory responses, cytoskeletal abnormalities, autophagic flux blockade, calcium overload, signal transduction disruptions, ferroptosis, and mitochondrial impairment (Chen et al., 2023; Tang et al., 2024; Zhang et al., 2024; Zhao et al., 2024). Furthermore, fluoride inhibits various metabolic enzymes involved in the tricarboxylic acid cycle, fatty acid B-oxidation, and protein and nucleotide metabolism, thereby perturbing the balance of choline, arachidonic acid, and other vital substrates (Araujo et al., 2019; Zhu et al., 2024). Despite evidence linking fluoride to cerebral oxidative stress, dysfunction, and metabolic disturbances, the precise mechanisms driving these metabolic disruptions are not fully elucidated, necessitating further investigation into the pathogenesis of fluoride-induced developmental neurotoxicity.

The brain, representing only 2% of body weight, accounts for approximately 20% of the body’s basal energy expenditure. Mitochondria, the primary energy-producing organelles, are abundant in neuronal somata, dendrites, axons, and synaptic terminals (Cunnane et al., 2020). Neurons rely predominantly on mitochondrial oxidative phosphorylation (OXPHOS) for ATP generation, with cytoplasmic glycolysis serving as a supplementary source, collectively sustaining fundamental brain functions (Cunnane et al., 2020). Under aerobic conditions, glycolytically derived pyruvate enters mitochondria and is converted to acetyl-coenzyme A to fuel OXPHOS; under anaerobic conditions, it is reduced to lactate (Li et al., 2023). OXPHOS efficiently produces substantial ATP, supporting essential processes such as synaptic maintenance and electrical excitability, thereby facilitating neuronal maturation (Iwata et al., 2023). Conversely, OXPHOS disruption leads to oxidative stress, ATP deficiency, and impaired signaling, ultimately compromising neuronal development and cognitive function (Liu et al., 2024). Fluoride exposure has been shown to impair neuronal mitochondrial function, suppress the expression of OXPHOS complexes, and disrupt energy-metabolizing enzymes, culminating in neuronal death and neurodevelopmental deficits (Li et al., 2023; Wang et al., 2021). Mitochondrial homeostasis and efficient energy metabolism depend on the coordinated action of numerous enzymes to ensure stable ATP production. Nevertheless, the specific mechanisms through which fluoride impairs energy metabolism to induce developmental neurotoxicity remain incompletely understood.

The sirtuin (SIRT) family comprises highly conserved NAD⁺-dependent deacetylases in mammals. SIRT3, a major mitochondrial deacetylase, is a critical regulator of micochondrial homeostasis and metabolism (Hu and Wang, 2022). Beyond its role in energy metabolism, SIRT3 modulates reactive oxygen species (ROS) levels, notably by enhancing superoxide dismutase (SOD) activity to mitigate oxidative stress (Wang et al., 2022). Fluoride exposure has been reported to downregulate SIRT3 protein expression in the hippocampus, increasing acetylation of its downstream target SOD and promoting ROS accumulation (Wang et al., 2021; Wang et al., 2022). ROS stabilization elevates levels of the transcription factor hypoxia-inducible factor (HIF) (Xueqiang et al., 2025). HIF, a heterodimer of a and B subunits, acts as a master nuclear regulator of cellular homeostasis (Huang et al., 2022). HIF-1a controls the transcription of numerous genes encoding growth factors, enzymes, transcription factors, and cytokines, which are pivotal in both normal physiology and pathogenesis (López-Barneo and Simon, 2020). Critically, HIF-1a reprograms cellular energy metabolism by enhancing glycolytic flux and suppressing mitochondrial function. It upregulates glucose transporters and glycolytic enzymes, inhibits pyruvate conversion to acetyl-CoA, and promotes degradation of mitochondrial proteins like COX4–1, shifting the balance from OXPHOS to glycolysis (Bao et al., 2021). Previous studies suggest fluoride upregulates enzyme activity within the HIF-1 signaling pathway to influence cellular energy acquisition in dental fluorosis (Ba et al., 2022), and activates HIF-1a to induce the Wnt/B-catenin pathway, contributing to skeletal sclerosis in rats (Zhu et al., 2022). Based on this evidence, we hypothesize that sodium fluoride (NaF) induces developmental neurotoxicity by suppressing SIRT3 and activating HIF-1a-mediated energy metabolism reprogramming.

Melatonin (Mel), an endogenous hormone secreted by the pineal gland, readily crosses the blood-brain barrier and exhibits sedative, hypnotic, and potent antioxidant properties. It has been investigated as a neuroprotective agent in neurodegenerative disorders such as Alzheimer’s and Parkinson’s diseases, exerting its benefits through antioxidant effects, cytokine modulation, and mitochondrial protection (Bocheva et al., 2024). Critically, Mel is recognized as a specific SIRT3 agonist, with its cytoprotective actions being largely SIRT3-dependent. For instance, Mel upregulates SIRT3 to protect mitochondrial function and stimulate autophagy in models of sepsis-induced intestinal and acute kidney injury (Deng et al., 2024; Xu et al., 2021); it activates the SIRT3/TFEB pathway to restore autophagic flux in doxorubicin-induced cardiotoxicity(Ma et al., 2023); and during myoblast differentiation, it enhances mitochondrial energy metabolism and antioxidant enzyme activity via SIRT3, effects abolished by SIRT3 silencing (Ge et al., 2024). Moreover, in ischemic injury, Mel selectively rescues SIRT3 downregulation in the ipsilateral hippocampus without affecting the contralateral side or controls, underscoring SIRT3 as a central mediator of Mel’s efficacy (Mohammadi et al., 2024). However, it remains unclear whether Mel protects against fluoride-induced neurotoxicity and whether such protection depends on the SIRT3/HIF-1a axis and its regulation of energy metabolism.

Therefore, we established complementary in vivo and in vitro models to mechanistically test this hypothesis. In vivo, Sprague-Dawley (SD) rats were perinatally exposed to sodium fluoride (NaF: 10, 20, 40 mg/kg/day) with or without Mel (10 mg/kg/day) to assess cognitive outcomes and underlying molecular changes. In vitro, HT22 cells were treated with NaF (0, 20, 40, 60 mg/L) and/or Mel (20 umol/L), or transfected with HIF-1a siRNA, to delineate the precise role of the SIRT3/HIF-1a axis. This study was designed to achieve two primary objectives: (1) to investigate whether the SIRT3/HIF-1a axis mediates fluoride-induced disturbances in neuronal energy metabolism, and (2) to determine whether Mel alleviates fluoride-induced developmental neurotoxicity by targeting this specific signaling pathway to restore metabolic homeostasis. Our findings aim to elucidate a novel molecular mechanism and provide a theoretical basis for Mel as a promising preventive agent against fluoride-related neurodevelopmental disorders.

… 3. Results

… 4. Discussion

Our study delineates a novel molecular pathway through which fluoride induces developmental neurotoxicity: suppression of SIRT3 leads to mitochondrial dysfunction and ROS accumulation, which stabilizes and activates HIF-1a, driving a reprogramming of neuronal energy metabolism from OXPHOS to glycolysis and resulting in insufficient ATP production. ATP, the core energy currency of neuronal activity, is essential for maintaining synaptic transmission, ion homeostasis, and neuronal plasticity—all critical processes for normal brain development (Iwata et al., 2023). This identifies ATP depletion as a key downstream effector of the SIRT3/HIF-1a axis in fluoride-induced neurotoxicity. Crucially, we demonstrate that intervention with Mel reverses this metabolic dysregulation and alleviates the associated neurotoxic effects. These findings not only establish the central role of the SIRT3/HIF-1a axis in mediating fluoride-induced energy disruption but also reveal the mechanism by which Mel confers neuroprotection by targeting this specific pathway.

Extensive epidemiological and experimental evidence has established fluoride as a significant developmental neurotoxin. Consistent with this, our study found that perinatal NaF exposure induced marked deficits in spatial learning and memory in offspring rats, as evidenced in the MWM test by prolonged escape latencies, reduced swimming speed, fewer platform crossings, and decreased time and distance in the target quadrant. These results align with previous experimental reports (Du et al., 2024; Zhang et al., 2025) and, more importantly, are supported by epidemiological studies linking early-life fluoride exposure to diminished intelligence, cognitive function, and perceptual reasoning in children (Singh et al., 2025; Veneri et al., 2023). Our data thus reinforce the consensus on the developmental neurotoxicity of fluoride.

The developing brain is exceptionally reliant on a continuous and efficient energy supply to support neuronal maturation and network homeostasis. Our data indicate that NaF directly assaults mitochondrial integrity, inducing ultrastructural damage (cristae fragmentation, vacuolization), oxidative stress (elevated ROS), and functional failure (loss of MMP), findings consistent with prior reports (Wang et al., 2021; Xin et al., 2023). The core of mitochondrial energy metabolism lies in OXPHOS, and mitochondrial respiratory chain Complex I (CI) is a key initiator of the electron transport chain. We found that NaF significantly inhibited the protein expression of NDUFS1, a core subunit of CI, which suggests that mitochondrial OXPHOS may be impaired. The Seahorse energy metabolism assay revealed that NaF exposure reduced basal respiration but increased glycolytic reserve in neuronal cells: the decrease in basal respiration directly reflects impaired mitochondrial OXPHOS capacity, which is the primary cause of insufficient energy production; the increase in glycolytic reserve indicates that neuronal cells enhance glycolysis as a compensatory response to mitochondrial dysfunction, thereby strengthening their ability to sustain energy supply under metabolic stress. This adaptive metabolic shift may represent a compensatory mechanism to counteract impaired mitochondrial energy production, although it is insufficient to fully prevent ATP depletion and subsequent neuronal damage under prolonged NaF exposure. In line with these metabolic changes, NaF exposure decreased the OCR and increased the ECAR in neuronal cells, indicating that OXPHOS was inhibited while glycolysis was compensatorily increased. Consistent with this, the upregulated protein expression of glycolytic rate-limiting enzymes (PFKFB3, PKM2) and lactate dehydrogenase (LDHA), as well as the increased levels of pyruvate and lactate, support the activation of glycolysis. However, since glycolysis (producing 2 ATP molecules from 1 glucose molecule) is far less efficient at generating energy than OXPHOS (producing approximately 33 ATP molecules from 1 glucose molecule) (Mookerjee et al., 2017), this ultimately leads to a significant decrease in intracellular ATP levels. As mentioned above, insufficient ATP production caused by inefficient compensatory glycolysis serves as a key link between metabolic reprogramming and neuronal dysfunction, which may underlie the cognitive impairments in our in vivo experiments. Studies have shown that NaF inhibits CI activity, downregulates genes involved in OXPHOS, and significantly reduces ATP levels in the hippocampus (Wang et al., 2021; Wang et al., 2022; Xin et al., 2023); In contrast, NaF exposure compensatorily upregulates glycolysis (Shobudani et al., 2024). These findings are consistent with the results of the present study. When exposed to NaF, the shift in neural cell energy metabolism from OXPHOS to glycolysis may represent a form of cellular self-preservation. This is because when mitochondrial function is impaired, electron leakage occurs in OXPHOS, leading to a further increase in ROS production and thereby exacerbating neuronal damage (Zhao et al., 2019). Interestingly, Li et al. (2023) found that NaF inhibits glycolysis levels in the mouse hippocampus, reduces ATP content, and leads to neurological damage and impaired spatial learning and memory. This partially contradicts our findings. Such discrepancies may stem from differences in animal models, exposure doses, and duration. Additionally, their study did not employ in vitro experiments to measure OXPHOS and its dynamic interactions with glycolysis. As noted by Araujo et al. (2019), the effects of NaF on energy metabolism are dynamic and jointly determined by exposure dose and duration. This further supports that divergent outcomes may arise from different experimental designs. In summary, our study comprehensively characterized the dynamic shift of the entire energy metabolism pathway and verified these findings using both in vitro and in vivo experiments. This provides a more complete mechanistic explanation for fluoride-induced developmental neurotoxicity and highlights the novelty of our work.

The KEGG pathway enrichment analysis indicated that the HIF-1 signaling pathway is associated with NaF-induced neurotoxicity. This study also found elevated HIF-1a expression in the nervous system following NaF exposure. As a key transcription factor regulating glycolytic enzyme expression, HIF-1a can directly control pyruvate production by upregulating glycolytic enzymes (Arias et al., 2025). It can also accelerate pyruvate conversion to lactate by increasing LDHA expression, but this reduces pyruvate conversion to acetyl-CoA and its contribution to OXPHOS (Arias et al., 2025). Previous studies have shown that NaF can affect cellular energy acquisition by upregulating the activity of enzymes enriched in the HIF-1 signaling pathway, thereby leading to dental fluorosis (Ba et al., 2022). Additionally, NaF can activate the HIF-1a signaling pathway, which in turn triggers the activation of the Wnt/B-catenin signaling pathway and autophagy, resulting in osteosclerosis in rats (Zhu et al., 2022). In acute ischemic stroke, HIF-1a activation induces glucose metabolism reprogramming in neurons and astrocytes, enhancing glycolysis (Madai et al., 2024). Exposure to a-synuclein activates the AKT-mTOR-HIF-1a signaling pathway in microglia, driving a metabolic shift in microglia from highly efficient OXPHOS to rapidly energy-supplying glycolysis to adapt to early pathological stress (Lu et al., 2022). However, following long-term exposure, the AKT-mTOR-HIF-1a pathway in microglia exhibits dysfunction, ultimately leading to energy metabolism defects (Lu et al., 2022). Baik et al.(Baik et al., 2019) demonstrated that primary microglia exposed to AB undergo metabolic reprogramming from OXPHOS to glycolysis via the mTOR-HIF-1a pathway, but prolonged AB exposure induces metabolic defects that impair microglial function. Inhibition of the mTOR-HIF-1a pathway reverses this shift, restoring OXPHOS while limiting glycolytic activity (Yang et al., 2023). Under simulated hypoxia, C3aR-deficient microglia exhibit reduced HHIF-1a levels accompanied by downregulation of its glycolytic target genes (Gedam et al., 2023). In the present study, knockdown of HIF-1a also induced identical changes in NaF-exposed neuronal cells, which confirms that HIF-1a activation is a critical step in NaF-induced energy metabolism disruption. We further observed that NaF exposure reduced the protein expression of SIRT3 in the hippocampus and HT22 cells. This finding is consistent with the results reported by Wang et al. (2021). Decreased SIRT3 expression leads to intracellular ROS accumulation, and increased ROS can suppress the degradation pathway of HIF-1a, thereby promoting its stabilization and accumulation (Wang et al., 2022; Xueqiang et al., 2025). This causal relationship further confirms that SIRT3 acts as an upstream negative regulator of HIF-1a in fluoride-induced developmental neurotoxicity, and the SIRT3/ROS/HIF-1a axis constitutes a complete signaling cascade that mediates metabolic reprogramming and ATP depletion.

In addition to the SIRT3/ROS/HIF-1a axis identified in this study, oxidative stress may also serve as an independent pathway involved in fluoride-induced developmental neurotoxicity. NaF-induced mitochondrial dysfunction leads to excessive ROS production, which in turn causes DNA damage, lipid peroxidation, and protein oxidation in neurons, thereby impairing neuronal structure and function (Duann and Lin, 2017; Wang et al., 2021). Moreover, oxidative stress can further inhibit mitochondrial OXPHOS activity, thereby amplifying energy metabolism disorders and ATP depletion (Fan et al., 2025). Although our study confirmed that the SIRT3/ROS/HIF-1a axis is the core pathway, the independent role of oxidative stress should not be overlooked, and its crosstalk with the SIRT3/HIF-1a axis may be a key direction for in-depth research on fluoride neurotoxicity. Based on the above speculation, we used Mel—a potent agonist of SIRT3 (Ge et al., 2024; Mohammadi et al., 2024). Mel intervention effectively upregulated SIRT3 expression, reduced ROS levels, and inhibited the protein expression of HIF-1a. More importantly, Mel successfully reprogrammed the metabolic pattern of NaF-exposed cells from glycolysis back to efficient OXPHOS, which was manifested by the upregulation of NDUFS1, downregulation of glycolytic enzymes and metabolites, and recovery of ATP levels. ATP depletion was identified as a key downstream effector linking this signaling axis to neuronal functional damage. This is highly consistent with the evidence that Mel can reverse the Warburg effect (a metabolic shift from mitochondrial OXPHOS to cytoplasmic glycolysis for ATP production) in pathological neurons (Reiter et al., 2024). Our results indicate that Mel can reprogram energy metabolism through the SIRT3/HIF-1a signaling axis, thereby alleviating learning and memory impairments in offspring rats caused by perinatal NaF exposure.

… 5. Conclusion

In conclusion, our study elucidates a novel mechanism for NaF-induced developmental neurotoxicity, wherein NaF impairs neuronal energy metabolism via the SIRT3/ROS/HIF-1a axis, leading to ATP depletion and subsequent cognitive deficits. Importantly, we demonstrate that Mel exerts potent neuroprotective effects by targeting this axis to reverse the pathological metabolic reprogramming. These findings not only advance our understanding of the molecular etiology of fluorosis but also highlight Mel as a promising prophylactic agent and the SIRT3/HIF-1a pathway as a potential therapeutic target for mitigating the neurodevelopmental risks associated with environmental fluoride exposure.

However, this study has certain limitations that need to be acknowledged: first, our research focused only on hippocampal neurons, a key region for learning and memory, whereas fluoride-induced developmental neurotoxicity may affect multiple brain regions with distinct molecular mechanisms; second, we only verified the core role of the SIRT3/ROS/HIF-1a axis, while its crosstalk with other pathways remains to be explored. Future research can extend the scope to other brain regions, clarify crosstalk among multiple signaling pathways, and establish more human-relevant models, thereby providing a more comprehensive theoretical basis for the prevention and treatment of fluoride-induced developmental neurotoxicity.

CRediT authorship contribution statement

Meng Zhang: Visualization, Data curation. Yajie Li: Investigation. Yongkang Liang: Supervision. Jingjing Zhang: Conceptualization. Qiang Niu: Writing – review & editing, Funding acquisition. Runjiang Ma: Writing – original draft, Methodology. Chun Wang: Validation, Software. Wenqi Qin: Visualization, Investigation. Chulin Yan: Supervision.