Histone H3K18 lactylation: An exercise-induced epigenetic mechanism that inhibits osteoclast activity and protects against osteoporosis

Abstract

Objective

Osteoclasts, through their excessive production, are the primary cause of postmenopausal osteoporosis. However, the influence of metabolism on osteoclastogenesis remains poorly understood. This study reveals that protein lactylation plays a critical role in osteoclast differentiation.

Methods

The ovariectomized (OVX) mouse model was used to investigate osteoporosis by examining glycolysis and lactate levels during osteoclast differentiation. Bone resorption was assessed through histomorphometric analysis. The effects of elevating lactate levels were tested both endogenously via exercise and exogenously through sodium lactate (NaLac) administration. Protein lactylation, focusing on histone modifications, was analyzed, and key osteoclastogenesis genes, including cathepsin K (Ctsk), matrix metalloproteinase 9 (Mmp9), and matrix metalloproteinase 12 (Mmp12), were quantified. The enzymes responsible for lactylation and delactylation were identified through cleavage under targets and tagmentation (CUT&Tag) and RNA-Seq analyses.

Results

Our findings showed that glycolysis and lactate levels were reduced during osteoclast differentiation in the OVX model, despite increased bone resorption. Elevating lactate through exercise or sodium lactate supplementation increased protein lactylation and mitigated OVX-induced bone loss. Mechanistically, lactate enhanced histone H3 lysine-18 lactylation (H3K18la), which suppressed osteoclast differentiation by downregulating key osteoclastogenesis genes like Ctsk, Mmp9, and Mmp12. Alanyl-tRNA synthetase 1 (AARS1) was identified as the lactylation “writer” that mediates H3K18la, with sirtuin 6 (SIRT6) acting as an “eraser” in a regulatory circuit.

Conclusions

Lactate suppresses osteoclast differentiation and alleviates osteoporosis through histone H3K18 lactylation, which downregulates osteoclastogenic genes including Ctsk, Mmp9, and Mmp12. The dynamic regulation of H3K18la involves AARS1 as the lactylation “writer”and SIRT6 as an “eraser”.

The translational potential of this article

This study reveals an epigenetic mechanism by which lactate regulates osteoclast function and suggests that exercise-induced lactate elevation or lactate supplementation may represent a viable therapeutic strategy for postmenopausal osteoporosis.

Graphical abstract

Graphical abstract.Schematic illustrating the proposed mechanism: exercise increases lactate production, which promotes H3K18la deposition, suppresses osteoclast differentiation, and alleviates osteoporosis.

1. Introduction

Osteoporosis (OP) is a systemic metabolic bone disorder characterized by a reduction in bone mass and the deterioration of bone microarchitecture, resulting in decreased bone strength and an elevated risk of fractures [[1], [2], [3]]. Globally, more than 200 million people suffer from OP. The greatest burden is borne by postmenopausal women, with over half of all women aged 60 and above affected by this disease. The increased brittleness and fragility of bones associated with osteoporosis can result in pain, mobility impairments, and even death, imposing a substantial medical and personal burden on affected individuals [2,4,5]. Therefore, preventing and treating osteoporosis is essential for relieving pain, reducing fracture risk, and improving overall health outcomes.

Postmenopausal osteoporosis is characterized by a decrease in bone mass due to excessive production of osteoclasts [6]. Therefore, targeting the differentiation and function of osteoclasts is a key strategy in the treatment of osteoporosis [7]. Osteoclasts are multinucleated cells that differentiate from bone marrow-derived macrophages (BMDMs). Osteoclasts have exceptionally high energy demands during differentiation. Recent studies have demonstrated that the differentiation of osteoclasts relies on energy derived from mitochondrial oxidative phosphorylation (OXPHOS), glycolysis, and other metabolic pathways [[8], [9], [10]]. Previous studies have shown that inhibiting oxidative phosphorylation blocks the differentiation of osteoclast precursors into mature osteoclasts [11]. Some studies suggest that oxidative phosphorylation and glycolysis are coupled during osteoclast differentiation, collectively influencing its progression and maturation [9,10]. Although much is known about the signals and transcription factors that regulate osteoclast differentiation, the bioenergetics of this process is only beginning to be understood.

Lactate is a key metabolite in glycolysis and is produced by the conversion of pyruvate through lactate dehydrogenase (LDH) under anaerobic conditions or during periods of high glycolytic flux [12,13]. Although lactate was historically viewed as a simple waste product of glucose breakdown, mounting evidence now recognizes it as a crucial signaling molecule, energy substrate, and immune regulator [[14], [15], [16]]. Clinical and preclinical studies indicate a potential connection between lactate levels and bone density. Specifically, recent research indicates that postmenopausal women with low bone mineral density (BMD) exhibit lower plasma glucose and lactate levels compared to those with high BMD [17]. Correspondingly, reduced serum lactate levels have been observed in osteoporotic mouse models [18]. Lactylation, a novel form of post-translational modification (PTM), directly links lactate metabolism to cell function by acting as the driver of this modification [19]. Lactylation is the covalent linking of lactate to lysine residues on proteins, which modulates their function and is involved in epigenetics and immune response [[20], [21], [22]]. Current research indicates that lactylation may play a direct or indirect role in various pathological and physiological conditions, and has been implicated in a wide range of diseases, including sepsis, cancer, and cardiovascular diseases [[23], [24], [25]]. A recent study reported that glycolysis in endothelial cells promotes the differentiation of bone marrow mesenchymal stem cells into osteoblasts through histone lactylation to improve osteoporosis [18]. However, the function of lactylation in osteoclasts is still unclear.

The purpose of this study was to investigate the effects and mechanisms of lactylation on osteoclast differentiation in osteoporotic mice. We found that in ovariectomized (OVX) mice, glycolysis, lactate, and protein lactylation levels were reduced during osteoclast differentiation, despite enhanced osteoclast differentiation and bone resorption. Our findings suggest that histone protein lactylation may inhibit osteoclast differentiation. Both endogenous lactate (from exercise) and exogenous lactate (intraperitoneal sodium lactate) alleviated bone loss in OVX mice by enhancing histone protein lactylation. Specifically, sodium lactate treatment increased histone H3 lysine 18 lactylation (H3K18la), which suppressed osteoclast differentiation by downregulating the expression of cathepsin K (CTSK), matrix metalloproteinase 9 (MMP9), and matrix metalloproteinase 12 (MMP12). We also identified alanyl-tRNA synthetase 1 (AARS1) as the lactylation “writer”for H3K18la, and found a regulatory circuit between H3K18la and the lactylation “eraser” sirtuin 6 (SIRT6). These findings provide new insights into the role of protein lactylation in governing osteoclast differentiation, thereby deepening our understanding of the molecular mechanisms underlying osteoporosis and highlighting potential therapeutic avenues targeting metabolic and epigenetic regulation.

2. Materials and methods

2.1. Experimental animals

The animal facility is certified by the Association for Assessment and Accreditation of Laboratory Animal Care. All animal procedures conformed to the Guide for the Care and Use of Laboratory Animals [26] and were approved by the Animal Research Ethics Committee of a university-affiliated stomatology hospital ([blinded]). Female mice were used for all experiments. In this study, 8-week-old C57BL/6 female wild-type mice, free of pathogens, were randomly assigned to one of four groups: the sham-operation group (Sham), the ovariectomized group (OVX), the ovariectomized mice with sodium lactate intervention group (OVX + NaLac), and the ovariectomized mice with exercise intervention group (OVX + Exercise). Each group consisted of six mice. Mice in the OVX group underwent bilateral ovariectomy, while mice in the Sham group underwent a similar procedure in which adipose tissue adjacent to the ovaries was resected. The ovariectomy procedure was performed as previously described [18]. In brief, mice were anesthetized with isoflurane, their dorsum shaved, and the skin was disinfected with povidone-iodine by the surgical staff. Mice were then placed in sternal recumbency, and an incision was made along the mid-dorsum. The ovaries were located and carefully removed using sterilized fine tweezers, and bleeding was controlled before suturing the incisions. One week after ovariectomy, the OVX + Exercise group began treadmill training according to a previously published protocol [27], with minor modifications. The exercise regimen was conducted daily for 7 weeks, with each session lasting 30 min at a speed of 16 m per minute. The OVX + NaLac group received intraperitoneal injections of sodium L-lactate (0.5 g/kg body weight, Sigma–Aldrich, 71718) according to a previously established protocol [28,29], administered daily for 7 weeks. As a control, mice in the OVX group received intraperitoneal injections of normal saline. The four groups of mice were subjected to weekly body weight measurements, and blood samples were collected via tail vein puncture. Blood glucose and lactate concentrations were measured using a portable glucometer and lactate analyzer (M221, Eaglenos Co., Ltd). Two months after surgery, all mice were euthanized for further experiments.

2.2. Biochemical and metabolite analyses

Blood serum was obtained by submandibular (facial vein) puncture from both Sham and OVX mice (n = 6). Blood biochemical parameters were measured using an automatic biochemical analyzer (Celercare V5, MNCHIP, China). Metabolites from serum and bone marrow samples of both Sham and OVX mice were extracted, and liquid chromatography-mass spectrometry (LC-MS) analysis was performed following previously described protocols [30]. LC-MS analyses were carried out using a Vanquish UHPLC system coupled with an Orbitrap Exploris 120 mass spectrometer. The resulting dataset was imported into SIMCA 16.0.2 for multivariate analysis. Principal component analysis (PCA) was employed to visualize the distribution of samples and identify potential outliers. Orthogonal partial least squares discriminant analysis (OPLS-DA) was utilized to identify significant metabolites, with model robustness assessed through cross-validation. Metabolites with a variable importance in projection (VIP) score greater than 1 and a p-value less than 0.05 were considered significantly altered. Pathway enrichment analysis was performed using KEGG to explore the associated metabolic pathways.

2.3. Cell culture

Bone marrow-derived macrophages (BMMs) were prepared as described previously [31]. In brief, bone marrow cells were extracted from 6-week-old female C57BL/6 J mice. Cells were cultured in α-MEM containing 10% fetal bovine serum (FBS) and 30 ng/mL macrophage colony-stimulating factor (M-CSF) to expand BMMs. Osteoclast differentiation was induced by stimulating BMMs with 20 ng/mL M-CSF and 100 ng/mL RANKL. After 2 days of culture, early-stage osteoclasts were obtained, and mature-stage osteoclasts were generated after an additional 3 days of culture.

2.4. Scanning electron microscopy (SEM) analysis

To evaluate bone-resorbing activity, bone marrow-derived macrophages (BMMs) were seeded at a density of 1 × 10⁴ cells per well on bovine bone slices (Zhejiang JoyTech Bio Co., Ltd) in 96-well plates and stimulated with M-CSF and RANKL for 14 days to induce osteoclast formation. Following induction, cells were treated with 10% sodium hypochlorite (NaClO; Sangon, China) and removed using a cotton swab. The bone slices were then sonicated in PBS and air-dried. For scanning electron microscopy (SEM) preparation, the slices were fixed overnight at 4°C in 2.5% glutaraldehyde (Aladdin, China) in Sorensen's buffer (pH 7.4), dehydrated through graded ethanol solutions, critically point-dried, and sputter-coated with gold/palladium. Imaging was performed using a Zeiss GeminiSEM 500, and resorption pits were quantified using ImageJ software.

2.5. LDH and LD assay

Cell lysates were prepared for lactate dehydrogenase (LDH) analysis using a Lactate Dehydrogenase (LDH) Assay Kit (catalog no. A020-2, Nanjing Jiancheng, China). Sample absorbance at 450 nm was measured using a microplate spectrophotometer (BioTek, USA). Cell lysates were also prepared for lactate (LD) analysis using a Lactate (LD) Assay Kit (catalog no. A019-2-1, Nanjing Jiancheng, China). Sample absorbance at 530 nm was detected using the microplate spectrophotometer (BioTek, USA).

2.6. Protein extraction, immunoprecipitation, and Western blot

When the cells in the six-well plates reached approximately 100% confluence, protein extraction and co-immunoprecipitation were performed. Total protein was extracted following the methods described in a previous study [32]. Cells were lysed using RIPA lysis buffer (Beyotime, China) supplemented with protease inhibitors. For co-immunoprecipitation, cells were washed three times with cold PBS and then lysed with 500 μL of NP40 cell lysis buffer per well (Invitrogen) containing a 1× protease inhibitor cocktail (Sigma). After 30 min of lysis on ice, the lysates were collected and centrifuged at 14,000 g for 10 min. The supernatant was then collected for subsequent analysis. Immunoprecipitation was performed using the Dynabeads Protein G Immunoprecipitation Kit (Thermo Fisher Scientific) according to the manufacturer's instructions. The lysates were incubated with either an anti-AARS1 antibody (Proteintech, China) or an anti-H3K18la antibody (PTM BIO, China) at 4°C for 2 h. The lysate/antibody complexes were then incubated with protein G coated Dynabeads and rotated at 360° at 4°C overnight. After four washes, the bound proteins were eluted with the elution buffer and analyzed by Western blot. The protein concentrations were determined using the BCA kit (Beyotime, China). Proteins were separated by SDS-PAGE, transferred to PVDF membranes, and blocked with 5% skimmed milk. Membranes were incubated overnight at 4°C with primary antibodies against Pan-Kla (PTM-1401, PTM BIO, China), H2BK16la (PTM-1424, PTM BIO, China), H3K14la (PTM-1414, PTM BIO, China), H3K18la (PTM-1406RM, PTM BIO, China), H4K8la (PTM-1415, PTM BIO, China), H4K12la (PTM-1411, PTM BIO, China), H3 (PTM-1001, PTM BIO, China), NFATc1 (66963, Proteintech, China), CTSK (R222, Abcepta, China), MMP9 (A0289, Abclonal, China), MMP12 (22989, Proteintech, China), SIRT6 (R25725, Zenbio, China), and AARS1 (67909, Proteintech, China). According to the manufacturer's validation data, the anti-H3K18la antibody (PTM-1406RM) has been rigorously validated for specificity, with no detectable cross-reactivity against other modifications (Supplementary Fig. S1). After incubation, the membranes were probed with a horseradish peroxidase (HRP)-conjugated secondary antibody (Biosharp, China) for 1 h at room temperature. Finally, the blots were developed using a chemiluminescent reagent (Advansta, CA) and imaged with a Bio-Rad XRS + Imaging System (USA).

2.7. Cell viability CCK-8 assay

To evaluate the effect of sodium lactate (NaLac), Galloflavin, and 2-(1-benzofuran-2-yl)-N-(diphenylmethyl) quinoline-4-carboxamide (12q) treatment on osteoclast viability, a CCK-8 assay (Biosharp, China) was performed. Briefly, 1 × 10³ osteoclasts were seeded in each well of a 96-well plate and treated with various concentrations of NaLac, Galloflavin, and 12q for 24 h. Following treatment, 100 μL of complete culture medium and 10 μL of CCK-8 reagent were added to each well and incubated for 2 h. The absorbance of the samples at 450 nm was then measured using a microplate spectrophotometer (BioTek, USA).

2.8. Immunofluorescence staining

For immunofluorescence staining, BMMs were fixed in 4% paraformaldehyde and permeabilized with 0.2% Triton X-100. A four-color multiplex fluorescent immunohistochemistry kit (RS0035, ImmunoWay Biotechnology) based on the tyramide signal amplification (TSA) technique was applied according to the manufacturer's protocol and as described previously [33]. The primary antibodies used included Pan Kla (PTM-1401, PTM BIO, China), H3K18la (PTM-1406RM, PTM BIO, China), CTSK (R222, Abcepta, China), MMP9 (A0289, Abclonal, China), MMP12 (22989, Proteintech, China), SIRT6 (R25725, Zenbio, China), and AARS1 (67909, Proteintech, China). Nuclei were counterstained with DAPI in Antifade Mounting Medium, and fluorescence images were captured using a confocal microscope (Olympus Corporation, Japan).

For bone tissue immunofluorescence staining, paraffin-embedded sections were deparaffinized and rehydrated, followed by antigen retrieval using a gastric enzyme (MXB Biotechnologies, China) for 20 min. The four-color multiplex fluorescent immunohistochemistry kit (RS0035, ImmunoWay Biotechnology) based on the TSA technique was then applied according to the manufacturer's instructions and as previously described [33]. The primary antibodies used were Pan Kla (PTM-1401, PTM BIO, China), H3K18la (PTM-1406RM, PTM BIO, China), CTSK (R222, Abcepta, China), and MMP9 (A0289, Abclonal, China). Nuclei were counterstained with DAPI using Antifade Mounting Medium, and fluorescence images were acquired using a confocal microscope (Olympus Corporation, Japan).

2.9. Calcein and xylenol orange double-labeling

Mice were administered intraperitoneal injections of 25 mg/kg calcein (2.5 mg/mL in saline) 14 days before sacrifice, followed by 90 mg/kg xylenol orange (15 mg/mL in saline) 7 days before sacrifice. The vertebrae were then fixed, dehydrated, and embedded in optimal cutting temperature (OCT) compound (SAKURA, Japan). Serial 10 μm sections were cut using a freezing microtome (Leica, Germany), and fluorescence images were acquired using a Thunder microscope (Leica, Germany).

2.10. Bone histology analysis

The vertebrae were fixed, dehydrated, and embedded in optimal cutting temperature (OCT) compound (SAKURA, Japan). Serial 10 μm sections were prepared using a freezing microtome (Leica, Germany), and von Kossa staining (Servicebio, China) was performed to evaluate mineralization.

Femurs from each group were fixed in 4% paraformaldehyde (PFA) and decalcified in 10% ethylenediaminetetraacetic acid (EDTA) at 4°C for 3 weeks. The decalcified samples were embedded in paraffin and sectioned into 5-μm-thick slices. Hematoxylin and eosin (H&E) staining (Servicebio, China) and tartrate-resistant acid phosphatase (TRAP) staining (Servicebio, China) were conducted to assess histological alterations.

2.11. ELISA for IL-6 and TNF-α

Serum levels of interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) were measured using commercially available ELISA kits (Newbosco Biotechnology Co., Ltd.) according to the manufacturer's instructions.

2.12. Additional methods

Detailed methods are provided in the Supplementary Methods.

2.13. Statistical analysis

Data are expressed as means ± standard deviation (SD), with “n" representing the number of independent experiments or the number of individual mouse phenotypes. Cells and mice were randomly assigned to different experimental groups. Each independent in vitro experiment included at least three technical replicates. Comparisons between two groups were performed using a two-tailed unpaired Student's t-test. For comparisons involving more than two groups, a one-way analysis of variance (ANOVA) was followed by Dunnett's post-hoc test. Statistical analyses were conducted using GraphPad Prism Version 9.0 (GraphPad Software, USA). P-values <0.05 were considered statistically significant. Statistical significance is indicated as follows: ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

3. Results

3.1. Serum and bone marrow metabolomic analyses reveal low lactate levels in OVX mice

To investigate the mechanism of osteoporosis, serum and bone marrow samples were collected from Sham and ovariectomized (OVX) mice for biochemical tests and untargeted metabolomics (Fig. 1A). The C-terminal telopeptide of type I collagen (CTX-1) level was higher, while the serum alkaline phosphatase (ALP) level was lower in OVX mice than in Sham mice. These results indicate that OVX mice exhibit an osteoporotic state characterized by increased bone resorption and reduced bone formation. Moreover, OVX mice exhibited elevated blood glucose levels and reduced lactate (Fig. 1B). According to Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, the tricarboxylic acid (TCA) cycle and 2-oxocarboxylic acid metabolism were significantly disrupted, both of which are essential for energy metabolism (Fig. 1E). Several TCA cycle-related metabolites, including α-ketoisovaleric acid, 3-hydroxydecanoic acid, 3-hydroxyisovalerylcarnitine, and itaconic acid, were significantly altered in the serum of OVX mice, as shown in the volcano plot (Fig. 1C). Organic acids and derivatives accounted for the largest proportion (21.69%) in the serum metabolome of OVX mice, suggesting a systemic disturbance in energy metabolism accompanying bone loss (Fig. 1F).

Fig. 1.

Serum and bone marrow metabolomic analyses reveal low lactate levels in OVX mice.

(A) Schematic overview of the experimental design, including the induction of osteoporosis (OVX) and sham-operated (Sham) mouse models, followed by serum and bone marrow collection for subsequent biochemical assays and untargeted metabolomic profiling. (B) Serum biochemical parameters in Sham and OVX mice (n = 6). (C) Volcano plot depicting upregulated and downregulated metabolites in the serum of Sham and OVX mice (n = 3). (D) Donut plot illustrating the overall metabolic profile of bone marrow samples from Sham and OVX mice (n = 3). (E) KEGG pathway enrichment analysis of the top 10 significantly downregulated pathways in the serum of Sham and OVX mice (n = 3). (F) Donut plot showing the overall metabolic profile of serum from Sham and OVX mice (n = 3). (G) Volcano plot displaying the upregulated and downregulated metabolites in the bone marrow of Sham and OVX mice (n = 3). (H) KEGG classification of metabolic pathways identified in the bone marrow of Sham and OVX mice (n = 3). (I) Heatmap showing relative levels of glycolysis-associated metabolites in the bone marrow of Sham and OVX mice (n = 3). Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

We found 569 distinct metabolites in our untargeted metabolomic analysis of bone marrow. The donut plot shows the distribution of chemical classifications of these metabolites. Among them, lipids and lipid-like molecules had the largest representation (24.78%), while organic acids and derivatives account for 18.81% (Fig. 1D). The volcano plot displays the results of the differential metabolite analysis between the OVX group and the Sham group. We identified a total of 2117 significantly altered metabolite features, comprising 1449 upregulated and 668 downregulated metabolites (|Fold Change| > 1 and p < 0.05) (Fig. 1G). KEGG pathway enrichment analysis revealed that metabolic pathways such as glycine, serine, and threonine metabolism; alanine, aspartate, and glutamate metabolism; mineral absorption; biosynthesis of cofactors; and biosynthesis of amino acids were significantly altered (Fig. 1H). These results suggested that there was an energy metabolism disorder in the bone marrow of OVX mice. Heatmap analysis further demonstrated a decrease in lactate levels in the glycolysis pathway (Fig. 1I). It suggested a potential reduction in glycolytic activity and a shift in cellular energy metabolism in the bone marrow of OVX mice.

3.2. Glycolysis in osteoclasts of OVX mice is reduced, leading to decreased lactate and protein lactylation, while osteoclast differentiation is enhanced

Serum and bone marrow lactate levels were reduced in OVX mice. Postmenopausal osteoporosis is characterized by a decrease in bone mass due to excessive production of osteoclasts [6]. Thus, to investigate the effect of lactate on osteoclasts, we isolated bone marrow-derived macrophages (BMMs) from Sham and OVX mice and induced osteoclast differentiation to perform subsequent experiments (Fig. 2A). The osteoclasts from OVX mice were more efficiently differentiated into tartrate-resistant acid phosphatase (TRAP)-positive multinucleated cells (Fig. 2B, C, E, F) and had stronger bone-resorptive activity (Fig. 2D–G). Western blot results showed that the protein nuclear factor of activated T cells c1 (NFATc1), a marker of osteoclast differentiation, was upregulated in early and mature osteoclasts derived from OVX mice (Fig. 2O, P, Q). These results indicated that osteoclast differentiation was enhanced in OVX mice.

Fig. 2.

Glycolysis in osteoclasts of OVX mice is reduced, leading to decreased lactate and lactylation, while osteoclast differentiation is enhanced

(A) Schematic diagram of the experimental design for isolating BMMs and inducing osteoclast differentiation. (B, E) F-actin staining and quantification of actin ring area of osteoclasts in Sham and OVX mice (n = 3). Scale bar = 200 μm. (C, F) TRAP staining and quantification of TRAP-positive multinucleated cells in Sham and OVX mice (n = 3). Scale bar = 200 μm. (D, G) Scanning electron microscopy (SEM) images and quantitative analysis of bone resorption pits in osteoclasts from Sham and OVX mice (n = 3). Scale bar = 100 μm. (H, I) Extracellular acidification rate (ECAR) profile and corresponding parameters of osteoclasts from Sham and OVX mice (n = 3). (J) Lactate dehydrogenase (LDH) activity levels in osteoclasts derived from OVX mice are lower than those in Sham mice at both the early stage (ES) and mature stage (MS) of osteoclastogenesis (n = 3). (K) Lactate levels in osteoclasts derived from OVX mice are lower than those in Sham mice at both the early stage (ES) and mature stage (MS) of osteoclastogenesis (n = 3). (L, O, P, Q) Western blot analysis of pan-Kla, H2BK16la, H3K14la, H3K18la, H4K8la, H4K12la, and NFATc1 during both the early (ES) and mature stages (MS) of osteoclast differentiation in Sham and OVX mice, with relative pan-Kla, histone lactylation marks, and NFATc1 levels normalized to histone H3 (n = 3). (R, S) ECAR profile and corresponding parameters of osteoclasts at the early stage (ES) and mature stage (MS) of differentiation (n = 3). (T) Lactate levels in osteoclasts at the early stage (ES) are lower than at the mature stage (MS) of osteoclastogenesis (n = 3). (M, N) Western blot analysis of pan-Kla during osteoclast differentiation at the early stage (ES) and mature stage (MS), with relative pan-Kla levels normalized to histone H3 (n = 3). Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

Given that lactate is produced via glycolysis [34]. we measured the extracellular acidification rate (ECAR) and found that the glycolytic rate was significantly lower in osteoclasts from OVX mice (Fig. 2H and I). Consequently, levels of lactate dehydrogenase (LDH) and lactate (LD) were lower in osteoclasts of OVX mice at both the early and mature stages of differentiation (Fig. 2J and K). In light of the apparent paradox that osteoclast differentiation was enhanced despite reduced glycolysis in OVX mice, we examined whether alternative energy pathways were upregulated to meet the increased energetic demands. Seahorse metabolic analysis revealed that osteoclasts from OVX mice exhibited significantly higher basal respiration, ATP-linked respiration, and maximal respiratory capacity compared to those from Sham mice (Supplementary Fig. S2A–B). These results indicate that oxidative phosphorylation (OXPHOS) is compensatorily upregulated in OVX osteoclasts, providing the necessary energy to sustain enhanced differentiation and resorptive activity when glycolysis is compromised.

Previous studies have shown that lactate serves as a substrate for lactylation [19]. In our study, lactylation levels in osteoclasts of OVX mice at both early and mature stages were lower than those in Sham mice (Fig. 2L). To investigate whether the decreased lactate content affected histone lactylation, we assessed the lactylation levels of several well-characterized lactylated residues in histones, including histone H2B lysine-16 lactylation (H2BK16la), histone H3 lysine-14 lactylation (H3K14la), histone H3 lysine-18 lactylation (H3K18la), histone H4 lysine-8 lactylation (H4K8la), and histone H4 lysine-12 lactylation (H4K12la). Our results showed that the lactylation levels of H2BK16la, H3K14la, and H3K18la were downregulated in osteoclasts of OVX mice during both early and mature stages (Fig. 2O, P, Q). These results demonstrated that reduced glycolytic activity and lactate levels in OVX mouse osteoclasts led to decreased histone lactylation, particularly at H2BK16la, H3K14la, and H3K18la.

Moreover, we found that there was a gradual increase in glycolysis in osteoclasts during osteoclast differentiation as measured by ECAR (Fig. 2R and S). Lactate levels, as well as lactylation, were increased during osteoclast differentiation (Fig. 2T–M, N). Therefore, we subsequently used mature stage osteoclasts to study the effect of lactylation on osteoclast differentiation.

3.3. Sodium lactate promotes glycolysis and protein lactylation but suppresses osteoclast differentiation

To investigate the function of protein lactylation in osteoclasts, we modulated the lactylation level by adding sodium lactate (NaLac) to osteoclasts. Based on previous studies [35,36], osteoclasts were treated with different concentrations (0, 5, or 25 mM) of NaLac to explore its effect on osteoclast differentiation. To exclude the possibility that the observed effects were due to pH changes rather than lactylation, we measured the pH of the culture media in control and NaLac (25 mM)-treated groups at 0 h and 24 h. The results showed no significant differences between the two groups at either time point (Supplementary Fig. S3A). Cell viability was assessed using the Cell Counting Kit-8 (CCK-8) assay, which indicated that these two concentrations of NaLac did not significantly affect osteoclast proliferation (Fig. 3A). Notably, the glycolytic rate, as measured by the ECAR, increased with rising NaLac concentrations (Fig. 3B and C). Consistently, LDH and LD levels also increased in a dose-dependent manner (Fig. 3D and E), and treatment with NaLac augmented protein lactylation, specifically inducing a greater rise in the levels of H3K18la and H4K12la (Fig. 3L, M, N, P). To further exclude the possibility that the observed effects were due to non-specific competition between lactylation and acetylation at the same residue, we assessed H3K18ac levels following NaLac treatment. NaLac treatment did not significantly alter H3K18ac levels, indicating that the effects of NaLac are specific to H3K18la rather than reflecting general acyltransferase dysregulation (Supplementary Fig. S8A–B). Nonetheless, a higher dose of NaLac also inhibited osteoclast differentiation, as seen by the low number of TRAP-positive multinucleated cells (Fig. 3F, G, I, J) and low bone-resorptive activity (Fig. 3H–K). Western blot analysis further revealed that expression of the osteoclast marker protein cathepsin K (CTSK) decreased with increasing NaLac concentrations (Fig. 3M and N). Immunofluorescence staining confirmed these findings, showing increased pan-Kla signals and decreased CTSK expression (Fig. 3O–Q). In conclusion, our findings demonstrate that sodium lactate promotes glycolysis and protein lactylation but suppresses osteoclast differentiation in osteoclasts.

Fig. 3.

Sodium lactate promotes glycolysis and lactylation but suppresses osteoclast differentiation

(A) Cell viability of osteoclasts treated with different concentrations of sodium lactate (0, 5, or 25 mM) for 24 h, assessed using the CCK-8 assay (n = 3). (B, C) Extracellular acidification rate (ECAR) profiles and corresponding glycolytic parameters of osteoclasts cultured with 0, 5, or 25 mM sodium lactate (n = 3). (D) Lactate dehydrogenase (LDH) levels in osteoclasts cultured with different concentrations of sodium lactate (0, 5, or 25 mM) (n = 3). (E) Lactate levels in osteoclasts cultured with different concentrations of sodium lactate (0, 5, or 25 mM) (n = 3). (F, I) TRAP staining and quantification of TRAP-positive multinucleated cells cultured with different concentrations of sodium lactate (0, 5, or 25 mM) (n = 3). Scale bar = 200 μm. (G, J) F-actin staining and quantification of actin ring area of osteoclasts cultured with different concentrations of sodium lactate (0, 5, or 25 mM) (n = 3). Scale bar = 200 μm. (H, K) Scanning electron microscopy (SEM) images and quantitative analysis of bone resorption pits in osteoclasts cultured with different concentrations of sodium lactate (0, 5, or 25 mM) (n = 3). Scale bar = 100 μm. (L, P) Western blot analysis of pan-Kla levels in osteoclasts cultured with different concentrations of sodium lactate (0, 5, or 25 mM), with relative pan-Kla levels normalized to histone H3 (n = 3). (M, N) Western blot analysis of H2BK16la, H3K14la, H3K18la, H4K8la, H4K12la, and CTSK in osteoclasts cultured with different concentrations of sodium lactate (0, 5, or 25 mM), with relative histone lactylation marks and CTSK levels normalized to histone H3 (n = 3). (O, Q) Immunofluorescent staining and quantification of relative fluorescence intensity in osteoclasts cultured with different concentrations of sodium lactate (0, 5, or 25 mM) (n = 3). Scale bar = 200 μm. Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

3.4. Galloflavin partially reverses the lactylation increase and impaired osteoclast differentiation caused by NaLac treatment

To investigate whether inhibiting lactate production in osteoclasts can restore osteoclast differentiation, we compared osteoclasts treated with sodium lactate (NaLac), the lactate dehydrogenase inhibitor (Galloflavin), and a combination of both. Based on previous studies [37], osteoclasts were treated with Galloflavin (5 μM). The CCK-8 assay was used to evaluate cell viability, and it was confirmed that Galloflavin did not significantly affect osteoclast proliferation (Fig. 4A). Therefore, osteoclasts were subsequently treated in the following groups: control, NaLac (25 mM), and NaLac (25 mM) + Galloflavin (5 μM). The glycolytic rate, as measured by the ECAR, was elevated in the NaLac group compared to the control, but the addition of Galloflavin significantly reduced this glycolytic increase (Fig. 4B and C). Similarly, LDH and LD levels were markedly higher in the NaLac group, but co-treatment with Galloflavin resulted in a decrease in both LDH and LD levels (Fig. 4D and E). Furthermore, protein lactylation levels of H3K18la and H4K12la were significantly increased in the NaLac group, but these levels were significantly reduced upon Galloflavin treatment. It suggested that lactate dehydrogenase inhibition reversed the NaLac-induced enhancement of protein lactylation (Fig. 4L, M, N, P). To further explore whether the Galloflavin-mediated rescue was due to compensatory metabolic changes, we assessed mitochondrial respiration and intracellular pyruvate levels in these three groups. Seahorse analysis revealed that basal respiration, ATP-linked respiration, and maximal respiratory capacity were highest in the control group and significantly reduced in the NaLac group. In the NaLac + Galloflavin group, maximal respiratory capacity showed a slight, non-significant increase compared to the NaLac group, while basal respiration and ATP-linked respiration remained unchanged with no statistically significant difference from the NaLac group (Supplementary Fig. S4A and B). These data indicate that LDH inhibition does not substantially reverse the NaLac-induced suppression of mitochondrial oxidative phosphorylation. We also measured intracellular pyruvate levels. Control cells exhibited the lowest pyruvate levels. Both NaLac-treated and NaLac + Galloflavin-treated groups showed significantly elevated pyruvate, with the combination group displaying the highest levels (Supplementary Fig. S4C). Inhibition of LDH leads to the accumulation of pyruvate, as its conversion to lactate is blocked. Notably, despite elevated pyruvate in both NaLac and NaLac + Galloflavin groups, osteoclast differentiation was only partially restored in the combination group. The divergence between pyruvate levels and differentiation rescue supports the conclusion that the effect is mediated primarily through reduced lactylation rather than pyruvate-induced metabolic changes.

Fig. 4.

Galloflavin treatment partially reverses the lactylation increase and impaired osteoclast differentiation caused by NaLac treatment

(A) Cell viability of osteoclasts treated with Galloflavin (5 μM) for 24 h, assessed by the CCK-8 assay (n = 3). (B, C) Extracellular acidification rate (ECAR) profiles and corresponding glycolytic parameters of osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin (n = 3). (D) Lactate dehydrogenase (LDH) levels in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin (n = 3). (E) Lactate levels in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin (n = 3). (F, I) TRAP staining and quantification of TRAP-positive multinucleated cells cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin (n = 3). Scale bar = 200 μm. (G, J) F-actin staining and relative density analysis of osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin (n = 3). Scale bar = 200 μm. (H, K) Scanning electron microscopy (SEM) images and quantitative analysis of bone resorption pits in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin (n = 3). Scale bar = 100 μm. (L, P) Western blot analysis of pan-Kla levels in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin, with relative pan-Kla levels normalized to histone H3 (n = 3). (M, N) Western blot analysis of H2BK16la, H3K14la, H3K18la, H4K8la, H4K12la, and CTSK in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin, with relative histone lactylation marks and CTSK levels normalized to histone H3 (n = 3). (O, Q) Immunofluorescent staining and quantification of relative fluorescence intensity in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 5 μM Galloflavin (n = 3). Scale bar = 200 μm. Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

Notably, a significant reduction in the formation of TRAP-positive multinucleated cells, indicating impaired osteoclast differentiation, was observed in the NaLac group compared to the control. Galloflavin co-treatment partially restored this effect (Fig. 4F, G, I, J). The results from pit formation assays showed a decrease in bone-resorptive activity in the NaLac group, while co-treatment with Galloflavin led to a significant recovery of bone-resorptive activity (Fig. 4H–K). Western blot analysis revealed that the expression of the osteoclast marker protein CTSK was decreased in the NaLac group, but Galloflavin treatment restored CTSK expression (Fig. 4M and N). Immunofluorescence staining confirmed these findings, showing a decrease in pan-Kla signals and a restoration of CTSK expression in the NaLac + Galloflavin group compared with the NaLac group (Fig. 4O–Q).

In conclusion, these findings demonstrate that the lactate dehydrogenase inhibitor partially reverses the lactylation increase and impaired osteoclast differentiation caused by treatment with NaLac.

3.5. RNA-seq analysis reveals that osteoclast differentiation and oxidative phosphorylation are involved in osteoclast responses to lactate stimulation

To identify potential genes involved in osteoclasts following lactylation, RNA sequencing (RNA-seq) was performed to detect differentially expressed genes after NaLac treatment in osteoclasts (Fig. 5A). Volcano plot analysis revealed 1143 upregulated genes and 1801 downregulated genes in NaLac treatment osteoclasts (Fig. 5B). KEGG pathway analysis of these differentially expressed genes indicated significant enrichment in pathways associated with “Osteoclast differentiation” and “Oxidative phosphorylation” (Fig. 5C). Additionally, RNA sequencing was conducted to identify differentially expressed genes in osteoclasts from Sham and OVX mice (Fig. 5D). Volcano plot analysis identified 1617 upregulated genes and 189 downregulated genes (Fig. 5E). KEGG pathway analysis of these differentially expressed genes indicated significant enrichment in pathways associated with “Osteoclast differentiation” (Fig. 5F). These data confirm that osteoclast differentiation and oxidative phosphorylation are involved in osteoclast responses to lactate stimulation.

Fig. 5.

Analysis of differentially expressed genes from RNA-seq

(A) Schematic diagram of the experimental design for RNA-seq analysis of NaLac-treated osteoclasts. (B) Volcano plot showing significantly upregulated and downregulated mRNAs in osteoclasts cultured with 0 or 25 mM sodium lactate. (C) KEGG pathway enrichment analysis of osteoclasts cultured with 0 or 25 mM sodium lactate. (D) Schematic diagram of the experimental design for RNA-seq analysis of osteoclasts from Sham and OVX mice. (E) Volcano plot showing significantly upregulated and downregulated mRNAs in osteoclasts from Sham and OVX mice. (F) KEGG pathway enrichment analysis of osteoclasts from Sham and OVX mice. Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

3.6. Identification of potential downstream targets of H3K18la by genome-wide CUT&Tag analysis

Histone lactylation plays a crucial role in regulating the transcription of target genes [38]. In our study, H3K18la level changed significantly in both osteoclasts of OVX mice and those treated with NaLac. To identify candidate genes regulated by H3K18la in osteoclasts, we employed genome-wide cleavage under targets and tagmentation (CUT&Tag) analysis. This method was recently developed to study protein-DNA interactions [39]. CUT&Tag analysis was performed using antibodies against H3K18la (Fig. 6A). Pearson correlation analysis of H3K18la between replicates from the control and NaLac-treated groups revealed a high correlation coefficient of over 0.91 (Fig. 6B). Volcano plot analysis revealed 2219 genes with increased H3K18la enrichment and 610 genes with decreased H3K18la enrichment (Fig. 6C). Analysis using deepTools revealed that H3K18la peaks were enriched in osteoclasts (Fig. 6D and E). Decreased H3K18la occupancy was primarily observed in promoter regions (38.2%) and distal intergenic regions (30%) (Fig. 6F), whereas increased H3K18la was predominantly located in distal intergenic regions (48.6%) and introns (39.4%) (Fig. 6G). KEGG pathway analysis of the target genes associated with the H3K18la binding peaks showed significant enrichment in pathways associated with “Osteoclast differentiation” (Fig. 6L). By combining CUT&Tag and RNA-sequencing data, we identified 59 potential target genes for H3K18la in osteoclasts, including Cathepsin K (Ctsk), matrix metalloproteinase 9 (Mmp9), matrix metalloproteinase 12 (Mmp12), and sirtuin 6 (Sirt6) (Fig. 6H). Notably, CUT&Tag analysis revealed decreased H3K18la levels at the promoters of these genes in NaLac-treated osteoclasts (Fig. 6I). ChIP-qPCR analysis confirmed that H3K18la is enriched at Ctsk, Mmp9, Mmp12, and Sirt6 promoter regions, and NaLac treatment reduced enrichment (Fig. 6J). To confirm this, quantitative real-time polymerase chain reaction (RT-qPCR) analysis revealed significantly decreased mRNA levels of Ctsk, Mmp9, Mmp12, and Sirt6 in NaLac-treated osteoclasts (Fig. 6K). Ctsk, Mmp9, and Mmp12 are key molecules involved in osteoclast differentiation [40,41]. In summary, our findings demonstrate that H3K18la regulates osteoclast differentiation by modulating the expression of target genes.

Fig. 6.

Identification of downstream targets of H3K18la by genome-wide CUT&Tag analysis

(A) Schematic representation of the cell collection and sequencing process. (B) Correlation matrix showing the Pearson correlation coefficients of H3K18la between replicates of osteoclasts cultured with 0 or 25 mM sodium lactate. (C) Volcano plot depicting significantly upregulated and downregulated mRNAs in osteoclasts cultured with 0 or 25 mM sodium lactate. (D, E) Binding density of H3K18la visualized using deepTools: heatmaps showing CUT&Tag counts at different H3K18la binding peaks in osteoclasts between the control and treated groups. (F, G) Pie charts illustrating the genomic distribution of regions with decreased (F) and increased (G) H3K18la signals in osteoclasts between the control and treated groups. (H) Bioinformatics analysis conducted to identify the downstream targets of H3K18la in osteoclasts. (I) IGV tracks for Ctsk, Nfatc1, Mmp9, Mmp12, the lactylation eraser Sirt6, and the lactylation writer Aars1, derived from CUT&Tag analysis. (J) ChIP- qPCR analysis of the indicated promoters was performed using antibodies against H3K18la in osteoclasts cultured with 0 or 25 mM sodium lactate. (K) qPCR analysis of Ctsk, Mmp9, Mmp12 and Sirt6 mRNA expression in osteoclasts cultured with 0 or 25 mM sodium lactate. (L) KEGG pathway enrichment analysis of osteoclasts cultured with 0 or 25 mM sodium lactate. Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

3.7. SIRT6 acts as an H3K18la eraser to inhibit H3K18la and promote osteoclast differentiation

SIRT6 is a nicotinamide adenine dinucleotide (NAD+) dependent deacetylase and participates in histone modification, DNA repair, cell-cycle regulation, and apoptosis [42]. Studies have found that traditional deacetylases can catalyze the removal of Kla [43]. In recent studies, SIRT6 exhibited effective delactylase activity against histone H3 lysine-9 lactylation (H3K9la) and H3K18la in vitro reconstructed nucleosomes [44,45]. To decipher the structural basis of SIRT6 recognition and binding to H3K18la, we used molecular docking and found that the SIRT6 protein could bind to the unmodified H3C1 (H3K18) peptide, with a binding energy of −125.57 kcal/mol; intriguingly, it could also bind to the lactylated H3C1 (H3K18la) peptide with a binding energy of −119.31 kcal/mol (Fig. 7A). This suggested that while SIRT6 could interact with both the unmodified and lactylated H3K18 peptides, the binding affinity was slightly reduced upon lactylation, indicating potential alterations in SIRT6 recognition due to the modification. To confirm the functional impact of SIRT6 as an eraser of H3K18la, we activated SIRT6 in NaLac-treated osteoclasts with the selective activator 2-(1-benzofuran-2-yl)-N-(diphenylmethyl) quinoline-4-carboxamide (12q). Based on previous studies [46], osteoclasts were treated with 12q (25 μM). The CCK-8 assay was conducted to assess cell viability, which showed that 12q did not significantly affect osteoclast proliferation (Fig. 7B). Consequently, osteoclasts were treated with the following groups: control, NaLac (25 mM), and NaLac (25 mM) + 12q (25 μM). Western blotting showed that 12q treatment significantly alleviated the NaLac-induced decrease in SIRT6 expression. Total protein lactylation (pan-Kla) and H3K18la levels were significantly elevated in the NaLac group, while both were drastically downregulated upon 12q treatment (Fig. 7I, J, K). Thus, enhancing SIRT6 expression may inhibit H3K18la levels. Osteoclast differentiation was severely impaired in the NaLac group as evidenced by the formation of TRAP-positive multinucleated cells, which was partially restored by 12q co-treatment (Fig. 7C, D, F, G). As evaluated through pit formation assays, the NaLac group exhibited less bone-resorptive activity; however, co-treatment with 12q significantly restored bone-resorptive activity (Fig. 7E–H). Western blotting showed that CTSK and MMP9 were reduced in the NaLac group but were restored by 12q treatment (Fig. 7J and K). Immunofluorescence staining confirmed these findings, showing a reduction in H3K18la signals and restoration of MMP12 and SIRT6 expression in the NaLac +12q group (Fig. 7L and M).

Fig. 7.

Addition of a SIRT6 agonist to sodium lactate solution reduces H3K18la in osteoclasts and enhances osteoclast differentiation

(A) Molecular docking analysis of the interaction between SIRT6 and the unmodified (H3K18) or lactylated (H3K18la) H3 peptide. (B) Cell viability of osteoclasts treated with the SIRT6 agonist 12q (25 μM) for 24 h, assessed by the CCK-8 assay (n = 3). (C, F) TRAP staining and quantification of osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 25 μM 12q (n = 3). Scale bar = 200 μm. (D, G) F-actin staining and relative density analysis of osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 25 μM 12q (n = 3). Scale bar = 200 μm. (E, H) Scanning electron microscopy (SEM) images and quantitative analysis of bone resorption pits in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 25 μM 12q (n = 3). Scale bar = 100 μm. (I, J, K) Western blot analysis of pan-Kla, H3K18la, SIRT6, CTSK, and MMP9 levels in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 25 μM 12q, with relative pan-Kla, H3K18la, SIRT6, CTSK, and MMP9 levels normalized to histone H3 (n = 3). (L, M) Immunofluorescent staining and quantification of relative fluorescence intensity in osteoclasts cultured with 0, 25 mM sodium lactate, or 25 mM sodium lactate + 25 μM 12q (n = 3). Scale bar = 200 μm. Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

To further validate the role of SIRT6 as an eraser of H3K18la, we performed loss-of-function experiments using siRNA-mediated knockdown. Efficient knockdown of Sirt6 was confirmed by RT-qPCR and Western blot assays (Supplementary Fig. S5A–C). Notably, SIRT6 knockdown significantly increased H3K18la levels in osteoclasts, while H3K18ac showed no detectable changes under the same conditions (Supplementary Fig. S5B and C), suggesting that SIRT6 preferentially regulates lactylation at this residue. Consistent with elevated H3K18la, the expression of osteoclast marker protein MMP9 was markedly reduced following SIRT6 knockdown (Supplementary Fig. S5B and C). RT-qPCR analysis further confirmed that the mRNA levels of Mmp9, Mmp12, and Ctsk were significantly decreased upon SIRT6 silencing (Supplementary Fig. S5A). Functional assays demonstrated that Sirt6 knockdown impaired osteoclast differentiation, as evidenced by reduced formation of TRAP-positive multinucleated cells (Supplementary Fig. S5D and G), disrupted F-actin ring structure (Supplementary Fig. S5E and H), and decreased bone-resorptive activity (Supplementary Fig. S5F and I).

In conclusion, these results demonstrate that SIRT6 knockdown increases H3K18la levels, suppresses osteoclast marker gene expression, and impairs osteoclast differentiation and bone-resorptive activity. Together with our activator studies using 12q, these findings support that SIRT6 acts as an eraser of H3K18la, promoting osteoclast differentiation by regulating H3K18la levels. Of note, through CUT&Tag sequencing, we identified Sirt6 as a direct target gene of H3K18la, and ChIP-qPCR confirmed that NaLac treatment reduced H3K18la enrichment at the Sirt6 promoter (Fig. 6J–K), which correlated with decreased Sirt6 mRNA and SIRT6 protein expression (Figs. 6K and 7I-J). Thus, what we observe is a transcriptional regulatory circuit: H3K18la promotes Sirt6 transcription, and SIRT6 in turn removes H3K18la.

3.8. AARS1 acts as a lactylation writer to promote H3K18la and inhibit osteoclast differentiation

To identify the regulators of H3K18 lactylation, we analyzed the RNA-seq data from the NaLac-treated group and found Aars1 to be upregulated. Recent studies revealed that alanyltRNA synthetase 1 (AARS1) was a lactylation writer. AARS1 exhibits a strong binding affinity for L-lactate and catalyzes ATP-dependent lactylation of lysine residues [[47], [48], [49]]. Based on these findings, we hypothesized that AARS1 serves as the writer for H3K18la and contributes to inhibiting osteoclast differentiation. To investigate the interaction between H3K18la and AARS1, we used molecular docking and found that the AARS1 protein could bind to the unmodified H3C1 (H3K18) peptide, with a binding energy of −57.58 kcal/mol; intriguingly, it could also bind to the lactylated H3C1 (H3K18la) peptide with a binding energy of −77.09 kcal/mol (Fig. 8A). And we performed Co-immunoprecipitation (Co-IP) assays. Immunoprecipitation with an anti-H3K18la antibody followed by Western blotting with an anti-AARS1 antibody confirmed the interaction between AARS1 and H3K18la in NaLac-treated osteoclasts (Fig. 8B). We continued to verify the efficiency of siRNA-mediated knockdown of Aars1 with qPCR and Western blot assays (Fig. 8C–K, L). We examined the levels of H3K18la in Aars1-silenced cells to assess whether AARS1 transferred the lactyl group from lactyl-CoA to histone H3K18. Notably, knockdown of Aars1 reduced H3K18la lactylation in NaLac-treated osteoclasts (Fig. 8K and L). Additionally, the depletion of Aars1 further rescued osteoclast differentiation, as shown by the TRAP-positive multinucleated cell formation (Fig. 8E, F, H, I). And the pit formation assays indicated that siAars1 restored the bone-resorptive activity of NaLac-treated osteoclasts (Fig. 8G–J). Furthermore, qPCR showed that treating siAars1 with NaLac resulted in the restoration of expression of the osteoclast marker genes Ctsk, Mmp9, Mmp12, and Nfatc1 in NaLac-treated osteoclasts (Fig. 8D). Western blotting confirmed that the expression of the osteoclast marker protein MMP12 was also restored following siAars1 treatment in NaLac-treated osteoclasts (Fig. 8K and L). Immunofluorescence staining confirmed these findings, demonstrating a reduction in both H3K18la and AARS1 signals and restoration of MMP12 expression in the siAars1 group (Fig. 8M and N).

Fig. 8.

AARS1 knockdown reduces H3K18la in osteoclasts and enhances osteoclast differentiation

(A) Molecular docking analysis of the interaction between AARS1 and the unmodified (H3K18) or lactylated (H3K18la) H3 peptide. (B) Western blot analysis of whole cell lysates and immunoprecipitates using anti-AARS1 and anti-H3K18la antibodies in osteoclasts cultured with 25 mM sodium lactate, showing that AARS1 interacts with H3K18la. (C) qRT-PCR validation of Aars1 knockdown efficiency in osteoclasts transfected with siNC or siAars1 (n = 3). (D)siAars1 rescued the inhibited mRNA expression of Ctsk, Mmp9, Mmp12, and Nfatc1 in osteoclasts cultured with 25 mM sodium lactate (n = 3). (E, H) TRAP staining and quantification of osteoclasts cultured with siNC or siAars1 (n = 3). Scale bar = 200 μm. (F, I) F-actin staining and relative density analysis of osteoclasts cultured with siNC or siAars1 (n = 3). Scale bar = 200 μm. (G, J) Scanning electron microscopy (SEM) images and quantitative analysis of bone resorption pits in osteoclasts cultured with siNC or siAars1 (n = 3). Scale bar = 100 μm. (K, L) Western blot analysis of H3K18la, AARS1, and MMP12 levels in osteoclasts cultured with siNC or siAars1, with relative levels of H3K18la, AARS1, and MMP12 normalized to histone H3 (n = 3). (M, N) Immunofluorescent staining and quantification of relative fluorescence intensity in osteoclasts cultured with siNC or siAars1 (n = 3). Scale bar = 200 μm. Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

Conversely, we overexpressed AARS1 in NaLac-treated osteoclasts to assess whether enhanced AARS1 activity could phenocopy the effects of lactate stimulation. Overexpression efficiency was confirmed by RT-qPCR (Supplementary Fig. S6A). Western blot analysis revealed that AARS1 overexpression significantly increased H3K18la levels, while suppressing the expression of osteoclast marker proteins CTSK and MMP12 (Supplementary Fig. S6B and C). Consistent with these molecular changes, AARS1 overexpression impaired osteoclast differentiation, as evidenced by reduced formation of TRAP-positive multinucleated cells (Supplementary Fig. S6D and F) and disrupted F-actin ring structure (Supplementary Fig. S6E and G).

Taken together, these loss- and gain-of-function experiments demonstrate that AARS1 acts as a lactylation writer that facilitates H3K18la deposition, thereby inhibiting osteoclast differentiation.

3.9. Endogenous lactate (exercise) and exogenous lactate (intraperitoneal sodium lactate injection) reverse osteoporosis through lactylation in mice

Using the ovariectomized (OVX) model, the effects of endogenous lactate (exercise) and exogenous lactate (intraperitoneal sodium lactate injection) on osteoporotic mice were evaluated in vivo. OVX mice were injected intraperitoneally with sodium lactate or saline once daily for 7 weeks. Another group of OVX mice exercised daily for 7 weeks (Fig. 9A). Body weight was monitored in Sham, OVX + saline (OVX Control), OVX + sodium lactate (OVX NaLac), and OVX + exercise (OVX Exercise) groups. The mice of the OVX Control group exhibited a rapid increase in body weight following estrogen depletion, as expected. Both the OVX NaLac and OVX Exercise groups showed a slower rate of weight gain compared with the OVX Control group. Notably, the body weight of the OVX NaLac group was nearly comparable to that of the Sham group (Fig. 9B). Oral glucose tolerance test (OGTT) was performed before euthanasia. The dynamics of blood glucose levels among the four groups showed similar patterns with no significant differences (Fig. 9C). There was no significant difference in the calculated area under the curve (AUC) values among the four groups (Fig. 9D). Weekly measurements of tail vein blood lactate revealed that the OVX NaLac group consistently maintained higher lactate levels than the other groups. In comparison, the OVX Control group had lower lactate levels. Starting in week 3, lactate levels in the OVX Exercise group started getting closer to the sham group (Fig. 9E). To assess whether chronic sodium lactate administration affected systemic pH, we measured the pH of serum samples from all four groups. There were no statistically significant differences in serum pH among the Sham, OVX Control, OVX NaLac, and OVX Exercise groups, indicating that the chronic lactate intervention did not cause a persistent systemic pH shift (Supplementary Fig. S3B). Together, these findings demonstrated that both endogenous and exogenous lactate intake impact overall lactate levels in the body without altering systemic pH homeostasis.

Fig. 9.

Endogenous lactate (exercise) or exogenous lactate (intraperitoneal sodium lactate injection) reverses osteoporosis in mice

(A) Schematic diagram of the in vivo experimental design. Ovariectomized (OVX) mice were treated with daily intraperitoneal injections of sodium lactate or saline (0.5 g/kg body weight per day) or subjected to 30 min of treadmill exercise per day for 7 weeks. (B) Body weight changes of mice during the experimental period (n = 6). (C, D) Oral glucose tolerance test (OGTT) and the total area under the curve (AUC) of blood glucose levels (n = 6). (E) Serum lactate levels measured weekly after drug administration or exercise. (F) Representative micro-CT images and three-dimensional reconstructions of the proximal tibias from the Sham, OVX control, OVX NaLac, and OVX Exercise groups. Scale bar = 1 mm. (G–J) Quantitative micro-CT analysis of relative bone volume (BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular separation (Tb.Sp) in the region of interest (ROI) of the proximal tibias (n = 6). (K) Representative images of calcein (green) and xylenol orange (red) double labeling showing newly formed bone in vertebrae at 7-day intervals. Scale bar = 400 μm. (L, M) Quantification of the mineral apposition rate (MAR) and bone formation rate (BFR/BS) in vertebrae (n = 6). (N) Von Kossa staining of vertebrae showing mineralized areas. Scale bar = 1 mm. (O, P) Quantification of the calcification area fraction (CAF) and mineralizing surface per bone surface (MS/BS) in vertebrae (n = 6). Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

Micro-computed tomography (Micro-CT) analysis and three-dimensional (3D) reconstruction were performed to assess the effects of endogenous and exogenous lactate on bone mass (Fig. 9F). Micro-CT analysis showed that bone volume/total volume (BV/TV), trabecular thickness (Tb.Th), and trabecular bone number (Tb.N) in OVX NaLac and OVX Exercise groups were significantly higher than those in the OVX Control group. However, trabecular bone separation (Tb.Sp) was much lower in OVX NaLac and OVX Exercise groups than in the OVX Control group (Fig. 9G–J). These results indicated that both the OVX NaLac and OVX Exercise groups exhibited significantly greater bone mass and denser trabecular structures compared with the OVX Control group. Consistently, bone histomorphometric analysis showed that the OVX NaLac and OVX Exercise groups displayed higher bone formation rate (BFR/BS) and mineral apposition rate (MAR) based on calcein and xylenol orange labeling (Fig. 9K–M). Moreover, von Kossa staining revealed a higher calcification area fraction (CAF) and mineralizing surface per bone surface (MS/BS) in the OVX NaLac and OVX Exercise groups compared with the OVX Control group (Fig. 9N–P). These data suggest that both endogenous and exogenous lactate alleviate osteoporosis-induced suppression of bone formation and mineralization.

Hematoxylin and eosin (H&E) staining of femoral sections demonstrated that both the OVX NaLac and OVX Exercise groups exhibited markedly improved bone microarchitecture compared with the OVX Control group (Fig. 10A and B). Tartrate-resistant acid phosphatase (TRAP) staining showed fewer osteoclasts (N.Oc) and a smaller osteoclast surface area (Oc.S) in the OVX NaLac and OVX Exercise groups relative to the OVX Control group (Fig. 10C–E). Furthermore, in vivo immunofluorescence staining confirmed that OVX NaLac and OVX Exercise groups exhibited decreased expression of CTSK and MMP9 and increased pan-Kla and H3K18la levels in comparison to OVX Control mice (Fig. 10F–I). Collectively, these results demonstrate that both endogenous lactate (from exercise) and exogenous lactate (via intraperitoneal sodium lactate injection) attenuate osteoporosis-induced bone loss by promoting protein lactylation in mice.

Fig. 10.

Endogenous lactate (exercise) or exogenous lactate (intraperitoneal sodium lactate injection) reverses osteoporosis-induced bone loss through lactylation in mice

(A) Representative H&E staining images of distal femoral trabeculae. Scale bar = 1 mm. (B) Quantification of trabecular bone area (TB, %Area) in H&E-stained sections. TB (%Area) was calculated as the percentage of trabecular bone area relative to the total bone area within the region of interest (ROI) (n = 6). (C) Representative TRAP staining images of distal femoral trabeculae. Scale bar = 0.5 mm. (D, E) Quantification of osteoclast parameters in TRAP-stained sections. N.Oc/BS represents the number of osteoclasts per bone surface, and Oc.S/BS (%) indicates the percentage of bone surface covered by TRAP-positive osteoclasts (n = 6). (F, G) Representative immunofluorescence staining of pan-Kla (red), CTSK (green), and DAPI (blue) in femoral sections, and quantification of relative fluorescence intensity (n = 6). Scale bar = 0.1 mm. (H, I) Representative immunofluorescence staining of H3K18la (red), MMP9 (green), and DAPI (blue) in femoral sections, and quantification of relative fluorescence intensity (n = 6). Scale bar = 0.1 mm. Data are presented as mean ± SEM. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant.

To evaluate the in vivo biosafety and anti-inflammatory effects of our interventions, we performed histological and serological analyses in the four experimental groups of mice. Hematoxylin and eosin (H&E) staining was conducted on major organs (heart, liver, spleen, and lung) collected after the 7-week intervention period. Histopathological examination revealed no obvious morphological abnormalities or signs of tissue damage, inflammation, necrosis, or fibrosis in any of the four groups. The overall tissue architecture and cellular morphology appeared comparable across the Sham, OVX Control, OVX NaLac and OVX Exercise groups (Supplementary Fig. S7A), indicating that neither ovariectomy itself nor the long-term exercise training or sodium lactate administration induced detectable organ toxicity at the histological level. Furthermore, to assess systemic inflammatory status, serum levels of the pro-inflammatory cytokines IL-6 and TNF-α were quantified by enzyme-linked immunosorbent assay (ELISA). Compared with the Sham group, OVX mice exhibited significantly elevated serum concentrations of both IL-6 and TNF-α, consistent with the chronic low-grade inflammation commonly observed in estrogen-deficient states [50]. Notably, both intervention groups showed marked alleviation of this inflammatory response (Supplementary Fig. S7B and C). Collectively, these results demonstrate that long-term exercise training and sodium lactate supplementation exert favorable systemic anti-inflammatory effects in the OVX-induced postmenopausal osteoporosis model without causing discernible histopathological damage to major organs. Both interventions therefore appear to possess good in vivo biocompatibility and safety profiles, supporting their potential as non-pharmacological or adjunctive strategies for the management of postmenopausal osteoporosis and associated inflammatory burden.

4. Discussion

Osteoporosis is a common systemic bone disease characterized by reduced bone mass and destruction of bone microarchitecture. It is caused by an imbalance between bone resorption by osteoclasts and bone formation by osteoblasts, with osteoclast overactivation serving as a central driver of disease progression [5]. In this study, we provide a novel epigenetic perspective on osteoclast differentiation by revealing a previously unrecognized role of histone lactylation. We identify AARS1 and SIRT6 as the “writer” and “eraser” of H3K18la, respectively, and demonstrate that both endogenous (exercise-induced) and exogenous (intraperitoneal sodium lactate) lactate attenuate ovariectomy (OVX)-induced bone loss by promoting histone lactylation.

Our metabolomic analysis revealed that OVX mice exhibit systemic and bone marrow-specific metabolic reprogramming characterized by decreased glycolytic flux and reduced lactate levels, along with compensatory increases in fatty acid and branched-chain amino acid oxidation. Previous studies have reported insulin resistance and hyperlactatemia in estrogen-deficient models, but those observations were largely obtained under high-fat diet or metabolic stress conditions [51,52]. In contrast, our mice were maintained on standard chow, and blood samples were collected under basal (non-fasted, resting) conditions. Oral glucose tolerance tests revealed no significant differences among the groups, indicating preserved glucose tolerance [53]. Notably, our finding of reduced basal serum lactate is consistent with clinical reports showing lower plasma lactate in postmenopausal women with low bone mineral density [17] and with preclinical studies documenting reduced serum lactate in osteoporotic mouse models [18]. Thus, the divergent metabolic phenotypes likely reflect differences in diet, metabolic state at measurement, and the specific estrogen deficiency model used [51,52,54].

In osteoclasts from OVX mice, we observed significantly enhanced differentiation and bone resorption, but with markedly reduced glycolytic activity, lactate production, and protein lactylation. The apparent observation that osteoclast differentiation is enhanced despite suppressed glycolysis can be explained by compensatory upregulation of oxidative phosphorylation (OXPHOS) to meet the increased energetic demands, consistent with literature identifying OXPHOS as a primary energy source for osteoclast function [8]. However, while enhanced OXPHOS provides the necessary energy supply, it does not directly explain the altered differentiation program. Notably, we found that reduced glycolytic flux in OVX osteoclasts was accompanied by decreased protein lactylation, a modification that links cellular metabolism to gene expression [19]. This uncoupling between enhanced differentiation and reduced lactylation raises the possibility that lactylation may play a regulatory role in osteoclastogenesis, consistent with previous studies showing that impairing glycolysis does not disrupt osteoclast formation [55]. Thus, while metabolic compensation maintains energetic homeostasis, the concurrent reduction in protein lactylation may function as a discrete regulatory mechanism modulating osteoclast differentiation in osteoporosis.

Histone lysine lactylation is a recently identified epigenetic modification that links cellular metabolic state to transcriptional regulation [19]. We initially observed that exogenous sodium lactate (NaLac) increased global H3K18la levels but paradoxically suppressed osteoclast differentiation. This apparent contradiction arises because the effect of lactylation on gene expression is inherently gene-specific: lactylation can activate some genes while repressing others, depending on the local chromatin context. Our CUT&Tag data revealed that NaLac treatment specifically reduces H3K18la occupancy at the promoters of key osteoclastogenic genes (Ctsk, Mmp9, Mmp12), which correlates with reduced mRNA expression of these genes. The observation that lower H3K18la occupancy is associated with lower transcription supports the role of H3K18la as an activation-associated mark at these promoters [56]. Importantly, similar gene-specific regulatory effects of lactylation have been reported in other systems. For example, Du et al. found that lactate increased global histone lactylation but decreased its occupancy at the Tyrp1 promoter, leading to reduced Tyrp1 expression [57]. Likewise, Hayashi et al. observed that histone lactylation predominantly localizes to transcriptionally repressive meiotic chromatin in Drosophila [58]. To address potential concerns about confounding factors, we performed control experiments showing that no pH difference was detected in the culture media and that H3K18ac levels remained unchanged after NaLac treatment. These results confirm that the observed effects are specifically attributable to lactylation rather than pH artifacts or non-specific competition with acetylation. Together, these results demonstrate that H3K18la negatively regulates osteoclastogenesis through gene-specific occupancy changes.

Similar to other post-translational modifications, lactylation is dynamically regulated by specific “writer” and “eraser” enzymes [59]. Here, we identify alanyl-tRNA synthetase 1 (AARS1) as a key writer for H3K18la in osteoclasts. AARS1 has recently been recognized as a lactate sensor that translocates to the nucleus and functions as a lactyltransferase [48,49,60]. Although AARS1 is canonically a cytoplasmic aminoacyl-tRNA synthetase, multiple independent studies have demonstrated that it can translocate into the nucleus in response to elevated lactate levels [48,49]. This nuclear translocation is mediated, at least in part, by a conserved nuclear localization signal (NLS) and its interaction with the importin KPNA4 [61]. Consistent with these reports, our data confirm that AARS1 translocates to the nucleus and functions as a lactyltransferase for H3K18la. Thus, AARS1 links metabolic cues (lactate availability) to epigenetic regulation (histone lactylation) and ultimately to osteoclast differentiation programs.

In addition to AARS1, we identified SIRT6 as a key regulator of H3K18la in osteoclasts. SIRT6 is a well-established multifunctional regulator of bone homeostasis, with context-dependent roles in both osteoblasts and osteoclasts [42]. Beyond its well-known deacetylase functions [61,62], emerging evidence has identified SIRT6 as a bona fide delactylase that efficiently removes lactylation from H3K9 and H3K18 [44,45]. Using both gain-of-function (12q activation) and loss-of-function (siRNA knockdown) approaches, we demonstrated that SIRT6 directly modulates H3K18la levels and osteoclast differentiation. Notably, through CUT&Tag and ChIP-qPCR, we found that Sirt6 is itself a direct target gene of H3K18la; NaLac treatment reduced H3K18la occupancy at the Sirt6 promoter, leading to decreased Sirt6 mRNA and SIRT6 protein. This establishes a transcriptional regulatory circuit in which H3K18la promotes Sirt6 transcription and SIRT6 in turn removes H3K18la. Of note, the reduced H3K18la occupancy at the Sirt6 promoter correlates with decreased Sirt6 expression, which is consistent with H3K18la functioning as an activation-associated mark [56]. Beyond its delactylase function, SIRT6 may also influence osteoclast differentiation through its well-established deacetylase activity, and these two mechanisms likely operate in parallel. For instance, SIRT6 has been reported to suppress NF-κB signaling via H3K9ac deacetylation and to regulate glycolysis through HIF-1α [62] pathways that may operate in parallel with H3K18la regulation in the context of estrogen deficiency. Additionally, SIRT6 interacts with Blimp1 to repress anti-osteoclastogenic genes such as Mafb [63], representing another layer of transcriptional control that may intersect with lactylation-mediated epigenetic changes. Thus, SIRT6 likely orchestrates osteoclast differentiation through multiple regulatory inputs, including epigenetic delactylation, transcriptional deacetylation, and protein-protein interactions—a model that warrants further investigation.

Exercise increases circulating lactate levels, and lactate produced by skeletal muscle can be transported to bone, where it contributes to metabolic support [[64], [65], [66], [67], [68]]. In this study, both endogenous (exercise) and exogenous (intraperitoneal sodium lactate) lactate effectively reversed OVX-induced bone loss. Notably, exogenous sodium lactate alone recapitulated the bone-protective effects of exercise in the absence of mechanical loading, strongly supporting that lactate-driven lactylation is a key molecular mediator of exercise-induced bone protection. Furthermore, lactate may operate through multiple complementary pathways to synergistically maintain bone homeostasis. A recent study has shown that lactate can signal through the GPR81 receptor to promote bone formation [69], suggesting that both receptor-mediated signaling and epigenetic regulation via histone lactylation may contribute to the beneficial effects of exercise on bone health. These mechanisms are not mutually exclusive; instead, these pathways likely operate in a coordinated manner, whereby GPR81 mediates rapid signaling events and protein lactylation drives sustained epigenetic reprogramming to collectively orchestrate the multifaceted effects of lactate on bone metabolism. Together, these findings demonstrate that lactate supplementation and exercise share a common epigenetic mechanism, positioning lactylation as a novel therapeutic target for osteoporosis. Furthermore, lactate and lactylation may serve as biomarkers for tailoring exercise regimens to prevent and treat osteoporosis. For example, resistance training, impact-loading activities, and balance exercises, which are known to be effective in combating osteoporosis, may be tailored to stimulate lactate production and enhance lactylation [[70], [71], [72]]. This personalized approach could be further refined by considering individual differences in lactate metabolism, ensuring that the exercise intensity and duration are optimized for each patient.

Several limitations of this study should be noted. Firstly, the upstream signaling pathways linking glycolytic metabolism to histone lactylation remain unclear. Secondly, our study focused primarily on osteoclasts; the effects of lactylation on osteoblasts and osteocytes require further investigation. Thirdly, the relative contributions of SIRT6's delactylase versus deacetylase activities in osteoclasts remain to be fully delineated. Fourthly, while exogenous lactate recapitulated the bone-protective effects of exercise, we cannot exclude additive or independent effects of mechanotransduction or myokines. Finally, although we demonstrated the therapeutic potential of lactate, the optimal dosage, administration route, and long-term safety of lactate-based interventions require systematic evaluation in future studies. Future studies using conditional genetic models or catalytically inactive SIRT6 mutants will be essential to definitively establish causality and therapeutic potential.

5. Conclusion

In summary, our study reveals that lactate regulates osteoclast differentiation and bone metabolism through histone lactylation. Specifically, we demonstrate that sodium lactate increases H3K18la levels in osteoclasts, thereby downregulating the expression of key genes such as Ctsk, Mmp9, and Mmp12, and ultimately inhibiting osteoclast differentiation. The dynamic regulation of this modification involves AARS1 as the lactyl-transferase (“writer”) and SIRT6 as the delactylase (“eraser”), forming a transcriptional regulatory circuit that governs osteoclast gene expression programs. Furthermore, both endogenous (exercise) and exogenous lactate interventions effectively mitigate osteoporosis in ovariectomized (OVX) mice, highlighting lactate-based metabolic strategies as a promising therapeutic avenue. By elucidating the role of lactylation in bone metabolism, our research suggests that lactate from either endogenous (exercise) or exogenous (supplementation) sources can treat osteoporosis through its dual effects on energy metabolism and epigenetic regulation. This novel insight may offer new avenues for targeting osteoclasts to improve bone health.

Author contributions

Zhenru He: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Yun Wu: Writing – review & editing, Writing – original draft, Visualization, Validation, Resources, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Yuan You: Writing – review & editing, Writing – original draft, Methodology, Data curation, Conceptualization. Ting Li: Writing – review & editing, Formal analysis, Conceptualization. Yilin Liao: Writing – review & editing, Formal analysis, Conceptualization. Jingqiu Chen: Writing – review & editing, Formal analysis, Conceptualization. Yuting Wang: Writing – review & editing, Conceptualization. Yue Sheng: Writing – review & editing, Conceptualization. Yaoyu Zhao: Writing – review & editing, Conceptualization. Weng Wan Yue Ye: Writing – review & editing, Formal analysis, Conceptualization. Mengjie Yin: Writing – review & editing, Visualization, Conceptualization. Peiqi Zhang: Writing – review & editing, Conceptualization. Ji Li: Writing – review & editing, Conceptualization. Huilin Tang: Writing – review & editing, Visualization, Conceptualization. Jingyang Lou: Writing – review & editing, Visualization, Conceptualization. Xiazhou Fu: Writing – review & editing, Visualization, Conceptualization. Xiaohong Yang: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Yaoting Ji: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Ethical statement

All animal procedures conformed to the Guide for the Care and Use of Laboratory Animals [26] and were approved by the Animal Research Ethics Committee of Wuhan University School of Stomatology (S07924110 B).

Declaration of generative AI in scientific writing

This manuscript was prepared with the assistance of Grammarly (Grammarly, Inc.), a writing aid tool used solely for grammar refinement, spelling correction, and language polishing. No generative AI technologies were used to generate or develop the scientific content, hypotheses, interpretations, or conclusions presented in this work. All intellectual content, data analysis, and manuscript drafts were produced by the authors, who have reviewed and approved the final version and take full responsibility for its accuracy and integrity.

Funding sources

This work was supported by the Noncommunicable Chronic Diseases-National Science and Technology Major Project [grant number 2025ZD0550400], National Natural Science Foundation of China [grant number 82372463, 82172493, 82301033], Shenzhen Science and technology Program (JCYJ20220531093814033).

Declaration of interest statement

The authors declare no conflict of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.