Lipocalin-2 upregulation in hypoxic murine osteocytes enhances RANKL-induced osteoclastogenesis

Abstract

Osteocytes regulate bone remodeling by interacting with osteoblasts and osteoclasts. Hypoxia influences osteocyte function and has been linked to increased osteoclastogenesis in pathological conditions such as orthodontic tooth movement (OTM) and bone metabolic diseases; however, the molecular mechanisms underlying these effects remain unclear. This study aimed to identify hypoxia-responsive genes in osteocytes and investigate their effects on osteoclastogenesis. Transcriptome analysis of murine long bone osteocyte-Y4 (MLO-Y4) osteocytes cultured under hypoxia (2% O₂) revealed that lipocalin-2 (Lcn2) was the most significantly upregulated gene. Real-time RT-PCR confirmed increased Lcn2 expression and an elevated Rankl/osteoprotegerin (Opg) ratio. Primary osteocytes were purified from DMP1-Topaz mice showed same hypoxic response. Functional analysis demonstrated that Lcn2 did not directly affect osteoclast precursors. However, it enhanced osteoclastogenesis via osteocytes in co-culture experiments. Western blot analysis demonstrated that LCN2 activated the MAPK signaling pathway in osteocytes. Furthermore, immunohistochemical analysis of hypoxic osteocytes on the compression side of OTM exhibited increased LCN2 expression. These findings suggest that LCN2 is upregulated in osteocytes under hypoxia and promotes osteoclastogenesis by increasing RANKL expression. This study provides new insights into the molecular mechanisms of bone resorption under hypoxic conditions and suggests Lcn2 as a potential therapeutic target for bone metabolic diseases.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-34575-2.

Introduction

Osteocytes are the primary cells embedded within the bone matrix and play a crucial role in maintaining bone homeostasis^1,2. They extend dendritic processes, which allow interaction with osteoblasts and osteoclasts, thereby regulating bone remodeling^3,4. When exposed to mechanical or inflammatory signals, osteocytes can alter the expression of key factors essential for osteoclast formation and bone resorption, such as RANKL, which is essential for osteoclast formation and bone resorption^5–8.

Hypoxia is a key factor in pathological conditions such as osteoporosis, rheumatoid arthritis, and cancer metastasis to bone, where it has been linked to the activation of osteoclasts and inflammation^9–11. In orthodontic tooth movement (OTM), a hypoxic environment is created within the periodontal tissues on the compression side¹², rendering exposure of osteocytes in this region to hypoxia. Recent studies suggest that hypoxia-induced osteocyte changes may increase bone resorption^12,13; however, the underlying molecular mechanisms remain unclear. The activation of hypoxia-inducible factor-1 alpha (HIF-1α) promotes the expression of several hypoxia-related genes, leading to alterations in cell metabolism and signaling pathways¹². Activated HIF-1α in osteocytes has been shown to increase the expression of RANKL and promote osteoclast formation¹⁴. However, the broader transcriptional response of osteocytes to hypoxia and its impact on bone remodeling remain poorly understood.

Lipocalin-2 (LCN2), also known as neutrophil gelatinase-associated lipocalin (NGAL), is a secreted protein with diverse functions in inflammation, CKD, energy metabolism, iron homeostasis, and tumor progression^15–17. Megalin and 24p3R, the known receptors of LCN2, are involved in intracellular signaling, inflammation, and cell survival^18,19. These receptors have been identified in osteoblasts²⁰; however, their presence in osteocytes has not been confirmed. Recent studies suggest that LCN2 plays a role in bone metabolism and osteoclast formation^21,22, as mechanical unloading has been shown to elevate LCN2 levels in osteoblasts. Moreover, osteoblast-derived LCN2 has been suggested to influence bone mass, energy metabolism, and appetite^20–22. However, whether LCN2 is regulated in osteocytes under hypoxia and how it contributes to osteoclastogenesis and bone remodeling remains unknown.

In this study, we performed transcriptome analysis to comprehensively analyze osteocyte gene expression changes under hypoxic conditions. We identified LCN2 as a significantly upregulated gene under hypoxic conditions and explored its potential function on osteoclastogenesis and bone resorption.

Results

Transcriptome analysis of MLO-Y4 cells under hypoxia and increased expression of LCN2

Principal component analysis (PCA) and heatmap analysis showed distinct gene expression patterns between the hypoxia-treated and control groups in MLO-Y4 osteocytes (Fig. 1a). Among the DEGs, 1,638 genes were significantly upregulated, while 427 were significantly downregulated under hypoxia (Fig. 1b).

Fig. 1.

Transcriptome analysis of hypoxic response in MLO-Y4 cells. (a) Principal Component Analysis (PCA): Each point represents a sample (red: hypoxia, blue: control). (b) Heatmap: Rows represent genes, columns represent samples, and colors indicate normalized expression values (red: high expression, green: low expression). The heatmap was generated using iDEP (version 0.96; http://bioinformatics.sdstate.edu/idep96/). (c) KEGG pathway analysis: The top 30 statistically significant pathways are shown. The vertical axis represents KEGG pathways, while the horizontal axis represents statistical significance (Adjusted P-Value). Dot size indicates the number of associated genes, and color represents the direction of expression change (red: upregulated, blue: downregulated). KEGG pathway information is from the Kyoto Encyclopedia of Genes and Genomes and is used with permission from Kanehisa Laboratories^51–53. (d) Volcano Plot: Each dot represents a gene. The horizontal axis represents the log₂ fold change, and the vertical axis represents statistical significance (− log₁₀ p-value). Red indicates significantly upregulated genes, while blue indicates significantly downregulated genes. (e-g) Gene Ontology (GO) Analysis: The top 30 statistically significant pathways are displayed. (e) Biological Process, (f) Molecular Function, and (g) Cellular Component. Dot size indicates the number of associated genes, and color represents the direction of expression change (red: upregulated, blue: downregulated). (h–k) Expression Changes of Specific Genes: Gene expression levels in MLO-Y4 cells under hypoxic conditions were measured using real-time RT-PCR and presented as box plots. (l) Western blot analysis of HIF-1α expression after HIF-1α siRNA knockdown in MLO-Y4 cells. (m) Quantification of HIF-1α protein levels normalized to β-actin. Band densities were measured using ImageJ software (n = 3). (n) LCN2 mRNA expression levels were evaluated by real-time RT-PCR in MLO-Y4 cells under control (normoxic), hypoxic, and hypoxia with HIF-1α knockdown conditions. Data are shown as mean ± standard deviation (SD). Statistical significance was set at *p < 0.05, **p < 0.01.

KEGG pathway analysis was performed to identify the biological pathways influenced by hypoxia. The analysis revealed significant changes in pathways related to inflammation, metabolism, and cell death. Particularly, inflammatory pathways such as TNF signaling (adj.P = 2.3 × 10⁻³), IL-17 signaling (adj.P = 1.1 × 10⁻²), and cytokine-cytokine receptor interaction (adj.P = 1.0 × 10⁻²) were significantly enriched. Additionally, metabolic pathways, including fatty acid metabolism (adj.P = 2.3 × 10⁻⁴), biosynthesis of cofactors (adj.P = 8.4 × 10⁻⁴) and apoptosis pathway (adj.P = 1.6 × 10⁻²) were activated under hypoxia (Fig. 1c). A volcano plot analysis identified Lcn2 as the most upregulated gene in response to hypoxia (Fig. 1d; Table 1).

Table 1.

Top 20 upregulated genes and bottom 20 downregulated expressed genes in hypoxia-stimulated osteocytes.

Gene names	log2(hypoxia/control)	Gene names	log2(hypoxia/control)
Lcn2	5.657498247	Ifnb1	-5.03139778
Saa3	5.348509024	Hspb1	-4.794834403
Hp	5.091967973	Hsp25-ps1	-4.637273212
Cxcl5	4.56127558	Ifit3	-4.579191628
Cxcl2	4.325700288	Ifit3b	-4.226851493
Ccl2	4.270203398	Hspa1b	-4.214155473
Angptl4	3.99637083	Nr1d1	-4.032413927
Il6	3.991524473	Gm12902	-3.74775981
Il12b	3.750625547	Eno2	-3.702702262
Gpr84	3.572982729	Bhlhe40	-3.395421888
Csf3	3.565860988	Sik1	-3.274072262
Mmp3	3.505791142	Arl4d	-3.258021924
Sema7a	3.455065655	Chaserr	-3.206128235
Zc3h12a	3.435985302	Dnajb1	-3.161369148
Col3a1	3.308782705	Slc2a1	-3.125635175
Sdf2l1	3.263274845	Ero1a	-3.095201693
Cxcl1	3.248553642	Ccl4	-3.006684166
Id3	3.184104711	Pet117	-2.994211064
Csf2	2.993232351	Ankrd37	-2.855582559
Gm6091	2.962308717	Gsta4	-2.763850474

GO analysis provided further insights into the cellular processes influenced by hypoxia. In the Biological Process category, pathways related to metabolic regulation were significantly enriched. For example, the ncRNA metabolic process (adj.P = 2.8 × 10⁻¹¹) and ncRNA processing (adj.P = 1.2 × 10⁻¹⁰) were highly significant. Additionally, pathways related to glycoprotein metabolic process (adj.P = 5.8 × 10⁻⁹) and protein glycosylation (adj.P = 4.7 × 10⁻⁸) were enriched (Fig. 1e). In the Molecular Function category, genes involved in catalytic activity, acting on RNA (adj.P = 2.4 × 10^− 5), hexosyltransferase activity (adj.P = 5.2 × 10⁻⁴), and transferase activity were significantly affected (Fig. 1f). In the Cellular Component category, the Golgi membrane (adj.P = 1.9 × 10⁻¹²), organelle inner membrane (adj.P = 2.3 × 10⁻¹⁰), and mitochondrial inner membrane (adj.P = 2.1 × 10⁻⁹) were significantly enriched (Fig. 1g).

Next, we performed real-time RT-PCR to measure the mRNA expression of Lcn2, Rankl, and Opg. The findings confirmed that Lcn2 and Rankl expression levels were significantly increased under hypoxia, while Opg expression remained unchanged, leading to a significantly elevated Rankl/Opg ratio in hypoxia-treated MLOY4 cells (Fig. 1h–k). To further investigate the effect of HIF-1α on hypoxia-induced LCN2 expression, siRNA-mediated knockdown of HIF-1α was performed in MLO-Y4 cells. Western blot analysis showed a significant reduction in HIF-1α protein expression by siRNA, while no significant changes were observed in cells transfected with negative control (NC) siRNA (Fig. 1l,m). Full-length blots are provided in Supplementary Fig. S1. In addition, knockdown of HIF-1α in MLO-Y4 cells under hypoxia significantly reduced Lcn2 mRNA expression. (Fig. 1n).

Increased expression of LCN2 and RANKL in primary osteocytes under hypoxia

To isolate primary osteocytes, fluorescence-activated cell sorting was performed using DMP1-Topaz mice, which express a green fluorescent protein (GFP) under the control of the Dmp1 promoter (Fig. 2a). The isolated cells displayed a characteristic stellate morphology of osteocytes (Fig. 2b).

Fig. 2.

Real-time RT-PCR analysis of primary osteocytes under hypoxia. (a) Number of Topaz-positive cells isolated using fluorescence-activated cell sorting (FACS): The number of Topaz-positive cells isolated using FACS is shown. (b) Morphology of isolated Topaz-positive cells: The morphology of isolated Topaz-positive cells is shown. Scale bar: 100 μm. (c–f) Gene expression analysis using real-time RT-PCR. Expression levels of (c) Lcn2, (d) Rankl, (e) Opg, and (f) Rankl/Opg ratio in primary osteocytes under hypoxia and control conditions. Data are presented as mean ± SD. Statistical significance was set at *p < 0.05, **p < 0.01.

Real-time RT-PCR analysis also showed that primary osteocytes cultured under hypoxia significantly increased Lcn2 and Rankl mRNA expression. In contrast, Opg mRNA expression remained unchanged, significantly increasing Rankl/Opg under hypoxia (Fig. 2c–f).

LCN2 induces RANKL expression in MLO-Y4 cells

Immunofluorescence staining was performed to investigate whether MLO-Y4 cells express receptors for LCN2. The results confirmed the presence of megalin and 24p3R, both known LCN2 receptors, in MLO-Y4 cells (Fig. 3a). Next, we examined the effect of LCN2 on RANKL and OPG expression. MLO-Y4 cells were treated with rmLCN2 (Recombinant mouse Lcn-2), and gene expression levels were assessed using real-time RT-PCR. After 1 day of rmLCN2 treatment, no significant changes were observed in Rankl or Opg expression, and the Rankl/Opg ratio remained unchanged (Fig. 3b–d). However, after 3 days of rmLCN2 treatment, RANKL expression was significantly upregulated (Fig. 3e), while OPG expression remained unchanged (Fig. 3f). Consequently, the Rankl/Opg ratio was significantly increased (Fig. 3g), indicating that prolonged exposure to LCN2 enhances RANKL expression in MLO-Y4 cells.

Fig. 3.

Expression of LCN2 receptors and real-time RT-PCR analysis of LCN2 treatment in MLO-Y4 cells. (a) Immunofluorescence staining of LCN2 receptors in MLO-Y4 cells. Expression of LCN2 receptors, Megalin and 24p3R, was analyzed using immunofluorescence staining. Nuclei were stained with DAPI. Scale bar: 100 μm. (b–d) Gene expression analysis after 1 day of LCN2 treatment. MLO-Y4 cells were cultured with rmLCN2 protein for 1 day, and gene expression levels of (b) Rankl, (c) Opg, and (d) Rankl/Opg ratio were analyzed using real-time RT-PCR. (e–g) Gene expression analysis after 3 days of LCN2 treatment. MLO-Y4 cells were cultured with rmLCN2 protein for 3 days, and gene expression levels of (e) Rankl, (f) Opg, and (g) Rankl/Opg ratio were analyzed using real-time RT-PCR. (h) Gene expression level of Lrp2 (Megalin) in MLO-Y4 cells after siRNA-mediated knockdown. (i) Gene expression level of Lrp2 in MLO-Y4 cells transfected with negative control (NC) siRNA. (j) Gene expression level of Slc22a17 (24p3R) in MLO-Y4 cells after siRNA-mediated knockdown. (k) Gene expression level of Slc22a17 after NC siRNA transfection. (l) Expression of Rankl was evaluated by real-time RT-PCR after rmLCN2 treatment in MLO-Y4 cells with knockdown of Megalin or 24p3R. Data are presented as mean ± SD (n = 4). Statistical significance was set at *p < 0.05. **p < 0.01.

To examine the role of LCN2 receptors in this process, siRNA-mediated knockdown of megalin and 24p3R was performed in MLO-Y4 cells. Expression of Lrp2 (megalin) and Slc22a17 (24p3R) was significantly reduced by siRNA (Fig. 3h,j), while no significant changes were observed in cells transfected with negative control (NC) siRNA (Fig. 3i,k). After 3 days of rmLCN2 treatment, RANKL expression was significantly reduced in the 24p3R-knockdown group compared to the group treated with rmLCN2 alone (Fig. 3l).

LCN2 enhances osteoclastogenesis via osteocytes

To investigate whether LCN2 directly influences osteoclast differentiation, osteoclast precursors were cultured under four conditions: M-CSF alone, M-CSF + RANKL, M-CSF + RANKL + LCN2, and M-CSF + LCN2. TRAP staining demonstrated that osteoclast formation occurred only in the M-CSF + RANKL and M-CSF + RANKL + LCN2 groups, indicating that LCN2 alone does not directly induce osteoclastogenesis (Fig. 4a,b).

Fig. 4.

LCN2 enhanced osteoclastogenesis via MLO-Y4 cells. (a) Microscopic images of osteoclasts. Osteoclast precursors were cultured under the following conditions: macrophage colony-stimulating factor (M-CSF), M-CSF (100 ng/mL) + receptor activator of nuclear factor-κB ligand (RANKL) (100 ng/mL), M-CSF + RANKL + rmLCN2 (100 ng/mL), or M-CSF + rmLCN2. Representative TRAP-stained images of osteoclasts under each condition are shown. Scale bar = 100 μm. (b) Quantification of TRAP-positive osteoclasts: The number of TRAP-positive cells with three or more nuclei was quantified. Data are presented as mean ± standard deviation (SD). (c) Osteoclast formation in the MLO-Y4 co-culture system: Osteoclast precursors were co-cultured with MLO-Y4 cells in the presence of vitamin D3 and prostaglandin E2, with or without rmLCN2. TRAP staining was used to visualize osteoclast formation. Scale bar = 100 μm. (d) Quantification of TRAP-positive osteoclasts in the co-culture system. The number of TRAP-positive osteoclasts was quantified in control and rmLCN2-treated groups. Data are presented as mean ± SD. Statistical significance was determined using the Tukey–Kramer test (n = 4), with *p < 0.05 and **p < 0.01 considered significant.

To assess the indirect effects of LCN2 on osteoclast differentiation via osteocytes, osteoclast precursors were co-cultured with MLO-Y4 cells in the presence of vitamin D3 and prostaglandin E2. TRAP staining revealed a significant increase in TRAP-positive multinucleated osteoclasts in the LCN2-treated group compared to the control, suggesting that LCN2 enhances osteoclastogenesis through its effects on osteocytes (Fig. 4c,d).

LCN2 activates ERK1/2, p38, and JNK MAPKs signaling in MLO-Y4 cells

To elucidate the signaling pathways involved in LCN2-mediated effects, we examined the activation of MAPK pathways in MLO-Y4 cells treated with rmLCN2 (Fig. 5a). Full-length blots are provided in Supplementary Fig. S2. Western blot analysis demonstrated that ERK1/2 phosphorylation (p-ERK1/2) significantly increased at 15 min post-treatment before declining (Fig. 5b,c). JNK phosphorylation (p-JNK) peaked at 5 min and rapidly decreased (Fig. 5d,e). p38 phosphorylation (p-p38) increased at 15 min and subsequently declined (Fig. 5f,g).

Fig. 5.

Effect of LCN2 on ERK1/2, p38, and JNK MAPK phosphorylation in MLO-Y4 cells. MLO-Y4 cells were incubated with 100 ng/mL of rmLCN2 for 0, 5, 15, 30, and 60 min; 0 indicates the condition without rmLCN2 treatment. (a) Western blot analysis: Cells were lysed and analyzed by western blotting using antibodies against phospho-ERK1/2, ERK1/2, phospho-p38, p38, phospho-JNK, JNK, and β-actin. (b–g) Quantification of protein phosphorylation levels. Band densities were measured using ImageJ software. (b) p-ERK1/2, (d) p-JNK, and (f) p-p38 phosphorylation levels were normalized to β-actin. (c), (e), and (g) phosphorylation levels were normalized to their respective total protein levels (n = 3). (h) RANKL mRNA relative expression as measured by real-time RT-PCR. MLO-Y4 cells were treated with rmLCN2 (100 ng/mL) and MAPK inhibitors—U0126, SB203580, and JNK inhibitor II (SP600125)—at 10 µM for 3 days (n = 4). Data are shown as mean ± SD. Statistical significance was determined using Tukey-Kramer test, with *p < 0.05 and **p < 0.01.

To further evaluate the role of MAPK pathways in LCN2-induced Rankl expression, MLO-Y4 cells were treated with rmLCN2 for 3 days in the presence of MAPK inhibitors. Inhibition of ERK1/2, JNK, and p38 MAPK activation by U0126, JNK inhibitor II (SP600125), and SB203580, significantly decreased Rankl mRNA expression compared to cells treated with LCN2 alone (Fig. 5h).

Expression of HIF-1α and LCN2 in osteocytes during OTM

To establish an OTM model, a Ni-Ti closed-coil spring was fixed between the upper anterior alveolar bone and the upper-left first molar to induce mesial movement of the first molar (Fig. 6a). The distances of tooth movement were measured on days 0, 2, and 6. After the initial displacement, tooth movement stagnated until day 6 (Fig. S3). Immunohistochemical analysis showed that HIF-1α and LCN2 expression increased significantly in osteocytes at the compression side during OTM (Fig. 6b). Expression levels peaked on day 2, with HIF-1α slightly decreasing on day 6, while LCN2 remained elevated (Fig. 6c,d).

Fig. 6.

Orthodontic tooth movement (OTM) model and the role of LCN2 in osteoclastogenesis. (a) Schematic illustration of OTM in mice. Intraoral image of a nickel-titanium (Ni-Ti) closed-coil spring positioned between the upper-left first molar and the maxillary alveolar bone beneath the incisors. A 0.1 mm stainless steel wire was used to secure the appliance to the posterior surface of the first molar and through holes drilled in the alveolar bone. The maxillary first molar was moved mesially by the applied orthodontic force. (b) Immunohistochemical analysis of HIF-1α and LCN2 expression. The expression of HIF-1α and LCN2 in osteocytes was examined during OTM. The percentage of HIF-1α- and LCN2-positive osteocytes was evaluated within a 400 × 200 μm region on the mesial periodontal ligament compression side, approximately 150 μm from the distobuccal root branch of the upper left first molar. After initiating OTM, observations were conducted on days 0, 2, and 6. Arrowheads indicate HIF-1α- or LCN2-positive osteocytes. Black arrows indicate the direction of orthodontic force. M: mesial side; D: distal side; a: alveolar bone; p: periodontal ligament; r: root. Black scale bar = 50 μm. White scale bar = 25 μm. (c, d) Quantification of HIF-1α- and LCN2-Positive Osteocytes during OTM: The percentage of HIF-1α- and LCN2-positive osteocytes was quantified at each time point. (e) TRAP-stained images of the mesial compression side of the distal buccal root of the maxillary left first molar at day 0, 2, and 6 after OTM, and at day 6 group treated with LCN2-neutralizing antibody every two days. Scale bar = 50 μm. (f) Number of TRAP-positive cells located on the alveolar bone surface on the mesial compression side. Data are presented as mean ± SD. Statistical significance was determined using the Tukey–Kramer test (n = 4), with *p < 0.05 and **p < 0.01 considered significant.

To evaluate the functional role of LCN2 in osteoclastogenesis during OTM, TRAP staining was performed on histological sections of the mesial alveolar bone surrounding the upper-left first molar. On day 6, the number of TRAP-positive cells was significantly increased on the compression side. In contrast, mice treated with an LCN2-neutralizing antibody every other day showed a significant reduction in TRAP-positive cells numbers compared to the untreated group (Fig. 6e,f).

Discussion

Osteocytes play a central role in maintaining bone homeostasis by regulating osteoblast and osteoclast activity^1,2. The expression of RANKL by osteocytes is crucial for osteoclast differentiation^5,6. Recent studies suggest that hypoxia can promote osteoclastogenesis both directly—by acting on osteoclast precursors—and indirectly—by influencing osteocytes and osteoblasts^21–26. However, the mechanisms by which hypoxic osteocytes contribute to osteoclast formation and bone remodeling remain incompletely understood. In diseases, such as osteoporosis, rheumatoid arthritis, bone metastasis, and OTM, hypoxia has been implicated in the activation of osteoclasts and inflammation^9–27. Therefore, understanding the adaptive responses of osteocytes to hypoxia holds clinical significance.

In this study, transcriptome analysis of hypoxia-exposed MLO-Y4 osteocytes revealed significant changes in expression of several genes, with Lcn2 showing a particular upregulation. LCN2 is widely known as a factor involved in inflammation, iron metabolism, and energy homeostasis^15–28. Previous studies have shown that hypoxia induces Lcn2 expression in astrocytes and cardiomyocytes^29,30. These findings are consistent with the findings that osteocytes also upregulate Lcn2 under hypoxic conditions. This suggests a broader role for Lcn2 in cellular adaptation to hypoxia.

Pathway analysis indicated the activation of inflammatory pathways, including the TNF and IL-17 signaling pathways and cytokine-cytokine receptor interaction. Osteocytes contribute to the production of cytokines and regulate osteoclastogenesis under inflammatory conditions^31–33. Consistent with these findings, our data showed that hypoxic stimulation upregulated inflammation-related cytokine genes, including Il6, Il12B, and Ccl2, and increased RANKL expression. Enhanced TNF-α signaling in osteocytes has been shown to promote RANKL expression and enhance bone resorption³⁴. This suggests that hypoxia-activated inflammatory signaling in osteocytes may induce osteoclast formation. In low-oxygen environments, ATP production via oxidative phosphorylation is reduced, leading to a shift toward glycolysis and fatty acid oxidation. Hypoxia has been reported to induce oxidative stress in adipose tissue and enhance lipolysis³⁵. Furthermore, intensified lipid metabolism may affect intracellular energy supply and bone remodeling^36,37. These findings suggest a mechanism by which osteocytes adapt to hypoxic stress. Hypoxia is also a well-established inducer of apoptosis. KEGG pathway analysis revealed activation of apoptosis-related pathways under hypoxic conditions, aligning with previous findings that apoptotic osteocytes promote osteoclast differentiation by HIF-1α signaling²⁶. Among the genes upregulated under hypoxia, Lcn2 exhibited the most significant upregulation. Lcn2 regulates inflammatory pathways and lipid metabolism^38–40 and may also play a role in apoptosis⁴¹. The concurrent upregulation of inflammatory and metabolic pathways observed in KEGG analysis suggests that LCN2 may regulate multiple processes. While the precise functions of LCN2 in osteocytes remain to be fully elucidated, our findings support its involvement in hypoxia-induced osteocyte adaptations.

We further employed Gene Ontology (GO) analysis to investigate the effects of hypoxia on osteocyte function. In the Biological Process category, significant alterations were observed in RNA metabolic pathways, including ncRNA metabolic processes and ncRNA processing. This suggests that hypoxia induces significant changes in osteocyte RNA regulation as part of adaptive responses. Recent studies indicate that HIF-1α interacts with ncRNAs to regulate metabolic adaptation and stress responses⁴². Our identification of altered ncRNA-related gene expression under hypoxia suggests a potential role for ncRNAs in osteocyte adaptation. In the Molecular Function category, significant changes were observed in catalytic activity, acting on RNA, hexosyltransferase activity, and transferase. The alterations in catalytic activity, targeting RNA suggest modifications in enzyme activity that may influence mRNA stability and translation. Changes in hexosyltransferase activity indicate potential glycosylation modifications, which could affect cell signaling and the function of membrane proteins in osteocytes. Furthermore, significant changes were detected in intracellular membrane structures such as the Golgi apparatus, organellar membranes, and mitochondrial membranes. Previous studies in tumor cells have shown that hypoxic stress induces Golgi adaptation, affecting protein secretion and modifications⁴³. A similar mechanism in osteocytes may enhance LCN2 secretion. Furthermore, osteocytes may adjust mitochondrial function under hypoxic conditions as part of their metabolic response, potentially increasing ATP production through enhanced electron transport⁴⁴. A related mechanism has been reported in osteoclasts, where hypoxia-induced ATP production promotes short-term bone resorption⁴⁵. Our findings suggest that osteocytes undergo similar mitochondrial adaptations, which may influence energy metabolism and bone remodeling.

We validated our transcriptome findings using real-time RT-PCR, confirming a significant upregulation of LCN2 and RANKL mRNA levels in response to hypoxia, while OPG expression remained unchanged. This resulted in an increased RANKL/OPG ratio. Similarly, in primary osteocytes, hypoxic treatment led to a substantial elevation in both LCN2 and RANKL, further increasing the RANKL/OPG ratio. These findings suggest that osteocytes upregulate LCN2 expression under hypoxic conditions and enhance RANKL production, thereby modulating the RANKL/OPG ratio. Furthermore, we showed that knockdown of HIF-1α significantly reduced LCN2 expression under hypoxia in osteocytes. This indicated that LCN2 upregulation is mediated by HIF-1α signaling.

We found that MLO-Y4 cells express both megalin and 24P3R proteins, suggesting that LCN2 may function in osteocytes through receptor-mediated signaling. When LCN2 was added to MLO-Y4 cells, RANKL expression remained unchanged at day 1 but significantly increased after 3 days of stimulation, leading to a higher RANKL/OPG ratio. These findings indicate that LCN2 does not immediately induce RANKL expression in osteocytes but exerts its effects over time. Moreover, we found that knockdown of 24p3R reduced the LCN2-induced RANKL expression. In contrast, knockdown of megalin did not significantly affect RANKL expression. These results suggest that LCN2 enhances RANKL expression via signaling through 24p3R in osteocytes. This indicates that 24p3R is the principal receptor mediating LCN2 signaling in osteocytes. The use of receptor antagonists or cell-specific knockout models would further support this mechanism.

To determine whether LCN2 directly influences osteoclast formation, we treated osteoclast precursor cultures with LCN2 and assessed TRAP staining. No direct stimulation of osteoclastogenesis was observed, aligning with Rucci et al.²¹ who reported that LCN2 indirectly regulates osteoclastogenesis via osteoblasts rather than acting directly on osteoclast precursors. However, Kim et al. showed that LCN2 directly suppresses osteoclast differentiation⁴⁵. These differences are likely due to variations in the origin of the recombinant protein. We and Rucci et al. used mammalian-derived recombinant LCN2 with a verified low endotoxin level (< 1.0 EU/µg, as measured by the LAL method), which retains proper post-translational modifications. In contrast, Kim et al. used LCN2 expressed in Escherichia coli, for which endotoxin contamination levels were not reported. The absence of these modifications may alter receptor interactions and signaling, potentially explaining the differences in observed effects. In addition, LPS has been reported to suppress osteoclastogenesis in bone marrow-derived precursor cells⁴⁶. Therefore, the possible presence of endotoxins, such as LPS, may have contributed to the discrepancy.

In co-culture experiments with MLO-Y4 cells and osteoclast precursors, the addition of LCN2 significantly increased osteoclast formation. However, when LCN2 was added to osteoclast precursors alone, no increase in osteoclastogenesis was observed, suggesting that LCN2 acts indirectly through osteocytes. This effect is likely mediated by the upregulation of RANKL expression in osteocytes. Our findings align with those of Rucci et al.²¹ who demonstrated that osteoblast-derived LCN2 enhances RANKL expression and promotes osteoclastogenesis.

To investigate the signaling mechanisms involved, we examined MAPKs (ERK1/2, p38, and JNK) activation following LCN2 stimulation and found that LCN2 activated these pathways. Moreover, inhibition of these MAPKs significantly reduced LCN2-induced RANKL expression. These results indicate that MAPKs signaling mediate the transcriptional response of osteocytes to LCN2. This is consistent with prior research showing that LCN2 regulates neuroinflammation in microglia via p38 MAPK signaling in the central nervous system⁴⁷. A similar mechanism may operate in osteocytes, further supporting LCN2-mediated MAPK activation. Additionally, previous studies have shown that MAPK signaling, particularly ERK and p38 MAPK, regulates RANKL expression in osteocyte, osteoblastic cells and bone marrow stromal cells under inflammatory conditions^34,48–50. Thus, MAPK activation may also contribute to RANKL regulation in osteocytes. As inhibition of MAPK signaling only partially suppressed LCN2-induced RANKL expression, it is possible that other signaling pathways, such as STAT3 or NF-κB, may also be involved.

Our findings suggest that LCN2 expression in osteocytes is upregulated under hypoxic conditions. LCN2 secreted by osteocytes may act through paracrine and/or autocrine signaling, increasing RANKL expression via MAPK activation and subsequently promoting osteoclastogenesis.

Finally, in an OTM model, immunohistochemical analysis revealed significantly increased HIF-1α and LCN2 expression in osteocytes on the compression side. Since orthodontic force is thought to induce hypoxia in the compressed periodontal ligament⁹, osteocytes in the compression zone are likely subjected to hypoxic stress. By day 2 of OTM, we observed a significant increase in HIF-1α-positive osteocytes, confirming their hypoxic state. Furthermore, the number of LCN2-positive osteocytes significantly increased on days 2 and 6, suggesting that hypoxia-induced LCN2 expression in osteocytes contributes to osteoclast formation on the compression side during OTM. Notably, although HIF-1α expression declined by day 6, LCN2 levels remained elevated. This suggests that, while LCN2 is initially induced by HIF-1α, its sustained expression may be supported by additional regulatory mechanisms. In particular, inflammatory cytokines such as TNF-α and IL-1β have been reported to enhance LCN2 expression under hypoxic and inflammatory conditions¹⁷. LCN2 expression was mainly observed in osteocytes on the compression side and overlapped with the HIF-1α–positive region, which indicates that osteocytes are a major source of LCN2 during OTM. However, it cannot be completely excluded that other hypoxia-responsive cells such as osteoblasts or immune cells may also contribute to LCN2 production in vivo. Furthermore, Local injection of an LCN2-neutralizing antibody significantly reduced the number of TRAP-positive cells on the compression side. This suggests that LCN2 may contribute to osteoclast differentiation during orthodontic tooth movement.

This study has several limitations. First, the precise mechanisms by which LCN2 signaling, particularly via 24p3R and megalin, regulates RANKL expression remain unclear. While our data confirm that osteocyte-derived LCN2 increases RANKL, we have not directly established whether this occurs specifically through MAPK signaling. The potential involvement of other pathways, such as NF-κB or STAT3, is not fully characterized. Further studies incorporating receptor antagonists, cell-specific knockout models, and combined MAPK inhibition and gene-silencing approaches would help clarify whether these pathways are essential mediators of LCN2-induced RANKL expression. Secondly, the in vivo role of LCN2 in bone remodeling requires further investigation. This study primarily focused on in vitro experiments using MLO-Y4 cells, leaving its impact on bone resorption and remodeling at the in vivo level unresolved. Specifically, to clarify the role of LCN2 in OTM, future research should employ OTM models using LCN2 knockout mice to assess how LCN2 deficiency affects hypoxia-induced bone remodeling.

While our findings focus on orthodontic tooth movement, hypoxia is also known to play a role in many bone-destructive diseases. LCN2 upregulation in osteocytes under hypoxic conditions may likewise contribute to pathological bone loss in conditions such as rheumatoid arthritis and bone metastasis. LCN2 expression has also been detected in chronic bone disorders such as osteoporosis and osteoarthritis^16,23, although its functional effects in these conditions remain unclear. Because the molecular and time course of hypoxia differs between chronic diseases and acute OTM, a direct comparison may not be appropriate. Further studies are needed to investigate the role of LCN2 in these disease contexts.

LCN2 is also involved in iron transport, oxidative stress regulation, and systemic energy metabolism^16,21. Since iron availability and energy metabolism are closely related to osteoblast and osteoclast activity, LCN2 may also influence bone remodeling through these metabolic pathways.

In conclusion, our study demonstrates that osteocyte-derived LCN2, upregulated under hypoxia, may promote RANKL expression, thereby enhancing osteoclast formation (Fig. 7). These findings provide new insights into the molecular mechanisms underlying tooth movement in orthodontic therapy highlighting LCN2 as a potential therapeutic target.

Fig. 7.

Illustration of how hypoxic stress induces LCN2 in osteocytes and stimulates bone resorption. Under hypoxia conditions, osteocytes increase the expression of LCN2, which, through autocrine and paracrine signaling, enhances RANKL expression. Increased RANKL then drives the differentiation of osteoclast precursor cells into osteoclasts, thereby promoting bone resorption.

Materials and methods

Mice and reagents

This study was approved by the Institutional Animal Care and Use Committee of the Tohoku University Environmental & Safety Committee (Approval Number: 2018DnA-028-06). All procedures were conducted in accordance with the Regulations for Animal Experiments and Related Activities at Tohoku University. We also complied with ARRIVE guidelines. All mice were housed in specific pathogen-free conditions under a 12-hour light/dark cycle with ad libitum access to feed (Labo MR Stock, Nosan Corporation, Kanagawa, Japan).

Eight-week-old C57BL/6J male mice (wild-type; WT) were purchased from CLEA Japan (Tokyo, Japan). C57BL/6-Tg (Dmp1-Topaz) 1lkal/J mice were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). Mice were housed in specific pathogen-free conditions under a 12-h light/dark cycle and provided ad libitum access to food (Labo MR Stock, Nosan Corporation, Kanagawa, Japan). MLO-Y4 cells were purchased from AddexBio Technologies (San Diego, CA, USA). Recombinant mouse Lcn-2 (rmLCN2) for the in vitro experiments was purchased from R&D Systems (Minneapolis, MN, USA). Recombinant mouse macrophage colony-stimulating factor (M-CSF) was obtained from the M-CSF expressing CMG14-12 cell line. Recombinant mouse RANKL was purchased from PeproTech (Rocky Hill, NJ, USA). The following polyclonal antibodies were used: SLC22A17 (Abonova, Taipei, Taiwan), megalin (Bioss, Woburn, MA, USA), HIF1-α (GeneTex, Irvine, CA, USA), and LCN2 (Aviva Systems Biology, San Diego, CA, USA). IgG polyclonal isotype control was purchased from Abcam (Cambridge, UK). Anti-mouse LCN2/NGAL monoclonal antibody (Clone 228418, R&D Systems) was used for in vivo LCN2 neutralization. For siRNA transfection, the following reagents were used: Lipofectamine™ RNAiMAX Transfection Reagent, Silencer™ Select Negative Control No. 1 siRNA (#4390843), Silencer Select siRNA for mouse Hif1a (#4390771), Lrp2 (#AM16708), and Slc22a17 (#AM16708), all from Thermo Fisher Scientific.

Cell culture and hypoxic culture

MLO-Y4 cells (2.2 × 10⁶ cells/dish) were cultured in 10 cm culture dishes (Corning Costar, USA) in α-minimum essential medium (α-MEM) (Fujifilm, Osaka, Japan) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (P/S) at 37 °C under 5% CO₂ for 24 h. For hypoxic treatment, cells were incubated in a BIONIX hypoxic culture kit (Sugiyama-Gen, Tokyo, Japan), where the oxygen concentration was adjusted to 2% O₂ using the gas control reagent. The culture system was then sealed in a gas-barrier pouch and incubated at 37 °C with 5% CO₂ for 24 h.

Transcriptome analysis and bioinformatics analysis

Total RNA was extracted from hypoxia-treated MLO-Y4 cells using the RNeasy Mini Kit (Qiagen, Valencia, CA, USA) and treated with RNase-Free DNase Set (Qiagen) to remove residual genomic DNA. RNA quality was assessed using the RNA Integrity Number (RIN), with all samples exhibiting RIN values of ≥ 8, indicating high-quality RNA.

Poly (A)-tailed mRNA was selectively captured for library preparation while preserving strand information. Library quality was verified, confirming the absence of primer dimers and a single peak distribution with a median fragment size of approximately 300 bp, ensuring high-quality libraries. The final library concentrations ranged from 46 to 138 nM, providing sufficient material for sequencing. Unique dual indices (10 bp) were used to minimize misassignment due to index hopping. Sequencing was performed using NovaSeq 6000, and raw output files (.cbcl format) were converted to fastq format. Quality control analysis using FastQC confirmed that each sample yielded ≥ 25 million reads. The DRAGEN RNA pipeline was used to map the fastq data to the mouse reference genome, followed by quantifying transcripts per million (TPM) values at both gene and transcript levels. Differential expression analysis (n = 3 per group) was conducted using the DRAGEN Differential Expression tool. Differentially expressed genes (DEGs) were identified based on TPM values, and TPM values were used to generate an input file for iDEP. Clustering, enrichment, and pathway analyses were performed using the iDEP tool. Pathway enrichment analysis was performed using KEGG pathway definitions^51–53. Use of KEGG materials was permitted by Kanehisa Laboratories.

Preparation of primary osteocytes

We isolated osteocytes using a previously described method³⁴. Osteocytes were isolated from the calvariae of 5–6-day-old DMP1-Topaz mice. Calvariae were dissected under aseptic conditions and sequentially digested at 37 °C with agitation using a 0.2% (w/v) collagenase solution (Fujifilm) for 20 min, followed by a 5 mM ethylenediaminetetraacetic acid (EDTA) solution for 15 min. The digestion process consisted of the following sequential steps: collagenase (fraction 1), EDTA (fraction 2), collagenase (fraction 3), collagenase (fraction 4), and EDTA (fraction 5). All fractions, except fraction 1, were cultured overnight in α-MEM supplemented with 10% FBS and 1% P/S under 5% CO₂ at 37℃. Adherent cells were harvested using trypsin–EDTA (Thermo Fisher/Sigma-Aldrich, Japan) and filtered through a 40 μm nylon cell strainer (Falcon, USA). Topaz-positive osteocytes were subsequently isolated by fluorescence-activated cell sorting (FACS; FACSAria II, BD Biosciences, Franklin Lakes, NJ, USA). The isolated Topaz-positive cells were used as primary osteocytes, which were cultured in α-MEM in a 24-well plate at a density of 3.0 × 10⁴ cells/well. The plates were cultured for 24 h under normal (20% O₂) or hypoxia (2% O₂) conditions using the BIONIX hypoxic culture kit.

LCN2 protein treatment

MLO-Y4 cells (1.0 × 10⁵ cells/well) were cultured in a 12-well plate (Corning Costar) in α-MEM with 10% FBS and 1% P/S at 37 °C under 5% CO₂ for 24 h. rmLCN2 was dissolved in sterile PBS and added to the culture medium at a final concentration of 100 ng/mL. Cells were incubated with rmLCN2 for 1 or 3 days before further analysis.

siRNA transfection

MLO-Y4 cells (5 × 10⁴ cells/well) were cultured in a 24-well plate for 24 h. Transfection with siRNAs targeting Hif1a, Lrp2, or Slc22a17 was performed at a final concentration of 30 nM using Lipofectamine RNAiMAX, following the manufacturer’s protocol. After 72 h, total RNA and protein were extracted for analysis.

MAPK inhibitor experiments

To test the effect of MAPK inhibition, MLO-Y4 cells were pre-incubated in a 24-well plate with 10 µM of U0126 (InSolution™ U0126, EMD Millipore, MA, USA), a MEK1/2 inhibitor targeting ERK1/2 MAPK, 10 µM of SB 203,580 (InSolution™ SB 203580, EMD Millipore), a p38 MAPK inhibitor, and 10 µM of JNK inhibitor II (SP600125, InSolution™, EMD Millipore), an inhibitor of JNK I, II, and III, simultaneously with rmLCN2 for 3 days.

Real-time RT-PCR analysis

Total RNA was extracted from MLO-Y4 osteocytic cells using the RNeasy Mini Kit. cDNA was synthesized using SuperScript IV reverse transcriptase (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. Real-time RT-PCR was performed using TB Green Premix Ex Taq II (Takara, Shiga, Japan) on a Thermal Cycle Dice Real-Time System TP800 (Takara). The cycling conditions were initial denaturation at 95 °C for 30 s, followed by 50 cycles of 95 °C for 5 s and 60 °C for 30 s. The following gene-specific primers were used: Gapdh: 5′-GGTGGAGCCAAAAGGGTCA-3′ (forward), 5′-GGGGGCTAAGCAGTTGGT-3′ (reverse), Actb: 5′-GAAATCGTGCGTGACATCAAA-3′ (forward), 5′-TGTAGTTTCATGGATGCCACAG-3′ (reverse), Rankl: 5′-CCTGAGGCCAGCCATTT-3′ (forward), 5′-CTTGGCCCAGCCTCGAT-3′ (reverse), Opg: 5′-ATCAGAGCCTCATCACCTT-3′ (forward), 5′-CTTAGGTCCAACTACAGAGGAAC-3′ (reverse), Lcn2: 5′-CCAGTTCGCCATGGTATTTT-3′ (forward), 5′-CACACTCACCACCCATTCAG-3′ (reverse), Lrp2: 5′-TGACCTCCCCAAATGCAAGT-3′ (forward), 5′-CAGAAGAGCCATTGTCGTCC-3′ (reverse), Slc22a17: 5′-CAGCCACCTCCTAACCGCTGTG-3′ (forward), 5′-CTCCCACTAGGCTCAAAGGCTGCT-3′ (reverse). The expression levels of the target genes were analyzed using the 2^−ΔΔCt method, with Gapdh as the reference gene for LCN2-treated cells and Actb as the reference gene for hypoxia-treated cells.

Immunofluorescence staining

MLO-Y4 cells (1 × 10³ cells/well) were seeded in 96-well plates and cultured overnight at 37 °C under 5% CO₂ in α-MEM with 10% FBS and 1% P/S. Cells were then washed three times with PBS and fixed using 4% formaldehyde solution in PBS for 15 min at room temperature. After washing with PBS, cells were permeabilized with 0.5% Triton X-100 (v/v) in PBS for 15 min. After another PBS wash, cells were blocked using 3% bovine serum albumin (BSA) in PBS for 30 min at room temperature. For immunofluorescence staining, cells were incubated overnight at 4 °C with the following primary antibodies, diluted in 3% BSA in PBS: SLC22A17, megalin polyclonal antibody (1:400 dilution), (1:400 dilution), IgG polyclonal isotype control. The next day, cells were washed with PBS and incubated for 1 h at room temperature in the dark with Alexa Fluor-conjugated secondary antibodies (1:100 dilution in 3% BSA in PBS). Finally, cells were stained with 4′,6-diamidino-2-phenylindole (DAPI) for 5 min before visualizing fluorescent signals under a fluorescence microscope (Olympus IX71, Tokyo, Japan).

Preparation of osteoclast precursors

C57BL/6J mice were sacrificed, and the femora and tibiae were immediately dissected. The epiphyses were removed, and bone marrow cells were flushed out with α-MEM. The cell suspension was filtered through a 40 μm nylon cell strainer and cultured in α-MEM supplemented with 10% FBS, 1% P/S, and M-CSF (100 ng/mL) for 3 days. Adherent bone marrow macrophages (BMMs) were collected using trypsin-EDTA (Sigma-Aldrich) and used as osteoclast precursors.

Osteoclast differentiation

Osteoclast precursors were seeded in 96-well plates at 5.0 × 10⁴ cells/well and cultured under the following conditions: M-CSF (100 ng/mL) alone, M-CSF (100 ng/mL) + RANKL (100 ng/mL), M-CSF (100 ng/mL) + RANKL (100 ng/mL) + rmLCN2 (100 ng/mL), M-CSF (100 ng/mL) + rmLCN2 (100 ng/mL). The medium was changed every 2 days. On day 4, cells were fixed with 4% paraformaldehyde for 30 min, permeabilized with 0.1% Triton X-100, and stained with tartrate-resistant acid phosphatase (TRAP) staining consisting of acetate buffer (pH 5.0), naphthol AS-MX phosphate, fast red violet LB salt, and 50 mM sodium tartrate. Cells were considered osteoclasts if they were TRAP-positive and had two or more nuclei.

Co-culture of MLO-Y4 and osteoclast precursors

MLO-Y4 osteocytic cells (1.0 × 10³ cells/well) were seeded in 96-well plates and cultured for 6 h in α-MEM supplemented with 10% FBS, 1% P/S at 37 °C under 5% CO₂. Afterward, osteoclast precursors (2.0 × 10⁴ cells/well) were added to the wells simultaneously with 10⁻⁸ M 1,25-dihydroxyvitamin D₃ (Sigma-Aldrich) and 10⁻⁶ M prostaglandin E₂ (PGE₂; Sigma-Aldrich). One group was additionally treated with and without rmLCN2(100 ng/mL). The medium was changed every 2 days. On day 7, cells were fixed with 4% paraformaldehyde for 30 min, permeabilized with 0.1% Triton X-100, and TRAP staining. Cells were considered osteoclasts if they were TRAP-positive and had two or more nuclei.

Western blot analysis

MLO-Y4 cells were cultured in α-MEM supplemented with 10% FBS, 1% P/S at 37 °C under 5% CO₂ overnight. Cells were then serum-starved in 2% FBS for 4 h, followed by 1% FBS for 2 h. After starvation, cells were treated with rmLCN2 (100 ng/mL) for 0, 5, 15, 30, or 60 min. Control wells (0 min) received no LCN2. Cells were lysed using radioimmunoprecipitation assay (RIPA) buffer (Millipore, Burlington, MA, USA) supplemented with 1% protease and phosphatase inhibitors (Thermo Fisher Scientific, Waltham, MA, USA) for 20 min on ice and insoluble material was removed by centrifugation. Protein samples were mixed 3:1 with Laemmli sample buffer (Bio-Rad, CA, USA) containing β-mercaptoethanol (Bio-Rad) and denatured at 95 °C for 5 min. Equal amounts of protein were separated using Mini-PROTEAN TGX Precast Gels (Bio-Rad) and transferred onto 0.2 μm PVDF membranes using the Trans-Blot Turbo Transfer System (Bio-Rad). Membranes were blocked in BlockAce (DS Pharma Biomedical, Osaka, Japan) at room temperature for 1–2 h, followed by overnight incubation at 4 °C with the following primary antibodies: Phospho-p38 MAPK (Thr180/Tyr182) rabbit monoclonal antibody (Cell Signaling Technology, Danvers, MA, USA, 1:3000 dilution). p38 MAPK rabbit monoclonal antibody (Cell Signaling Technology, 1:3000 dilution). Phospho-ERK1/2 (Thr202/Tyr204) rabbit monoclonal antibody (Cell Signaling Technology, 1:3000 dilution). ERK1/2 MAPK rabbit monoclonal antibody (Cell Signaling Technology, 1:3000 dilution). Phospho-SAPK/JNK (Thr183/Tyr185) rabbit polyclonal antibody (Cell Signaling Technology, 1:3000 dilution). SAPK/JNK MAPK rabbit polyclonal antibody (Cell Signaling Technology, 1:3000 dilution). β-actin mouse monoclonal antibody (Sigma-Aldrich, 1:5000 dilution). After washing with tris-buffered saline with Tween-20 (TBS-T), membranes were incubated with HRP-conjugated anti-rabbit IgG (Cell Signaling Technology, 1:5000 dilution) or anti-mouse IgG (Cytiva, Tokyo, Japan, 1:10,000 dilution) for 1 h at room temperature. Signals were detected using an enhanced chemiluminescence (ECL) system (SuperSignal West Femto Maximum Sensitivity Substrate, Thermo Fisher Scientific, IL, USA). Band intensity was quantified using ImageJ software (NIH, USA).

To validate siRNA-mediated Hif1a knockdown, western blotting was performed using MLO-Y4 cells transfected with Hif1a or control siRNA for 72 h. The primary antibody was anti-HIF-1α (GeneTex, 1:1000), and the secondary antibody was HRP-conjugated anti-rabbit IgG (1:5000). Signal detection was performed as described above.

Experimental tooth movement

OTM was performed as previously described^54–57. Male WT mice (8–12-week-old) (n = 4) were anesthetized with an intraperitoneal injection of a mixture of medetomidine, midazolam, and butorphanol prior to appliance placement. A nickel-titanium (Ni-Ti) closed-coil spring (TOMY SEIKO Co. Ltd., Tokyo, Japan) was attached between the upper incisor and left first molar. The appliance was secured using a stainless-steel wire (0.01 mm diameter), which was fixed to a hole drilled in the upper anterior alveolar bone and tied to the first molar. The first molar was moved mesially with a force of 10 g. Each group contained four mice, and OTM was carried out for 0, 2, and 6 days. After tooth movement was completed, the mice were anesthetized via intraperitoneal injection of medetomidine, midazolam, and butorphanol, subsequently euthanized by cervical dislocation, and then the Ni-Ti closed-coil spring was removed. To measure tooth movement, stereoscopic microscopy (VH-7000; Keyence, Osaka, Japan) was used to determine the shortest distance between the distal marginal ridge of the first molar and the mesial marginal ridge of the second molar. For in vivo LCN2 neutralization, 20 µg of anti-mouse LCN2/NGAL monoclonal antibody was locally injected every other day adjacent to the upper-left first molar.

Histological preparation and immunohistochemistry

Calvariae and maxillae were harvested and fixed overnight in 4% paraformaldehyde at 4 °C. Samples were then decalcified in 14% EDTA at 4 °C for 3 days (calvariae) or 1 month (maxillae). Following dehydration, the specimens were embedded in paraffin and sectioned: Coronal Sect. (5 μm thick) for calvariae and horizontal Sect. (4 μm thick) for maxillae. Maxillae sections were obtained approximately 150 μm from the root branch of the upper-left first molar. For immunohistochemistry, paraffin sections were deparaffinized, rehydrated, and treated with 3% hydrogen peroxide (H₂O₂) for 15 min to block endogenous peroxidase activity. Sections were then blocked with 5% skim milk for 30 min at 37 °C and incubated overnight at 4 °C with HIF-1α (1:100 dilution), LCN2 (1:50 dilution) polyclonal antibodies and IgG polyclonal isotype control. After washing, sections were processed using the VECTASTAIN Elite ABC Kit (PK-6105, Vector Laboratories Inc., Burlingame, CA, USA) and developed with 3,3′-diaminobenzidine (DAB, Vector Laboratories). Hematoxylin was used for counterstaining. The percentage of HIF-1α- and LCN2-positive osteocytes was quantified within a region parallel to the long axis of the root, starting from the periodontal ligament. TRAP staining was also performed to visualize osteoclasts, using a staining solution containing acetate buffer (pH 5.0), naphthol AS-MX phosphate, fast red violet LB salt, and 50 mM sodium tartrate.

Statistical analysis

Statistical analyses were performed using JMP Pro 17 software (JMP Statistical Discovery, Cary, NC, USA). For comparisons between the two groups, Student’s t-test was used. For multiple group comparisons, ANOVA followed by the Tukey–Kramer post hoc test was applied. Statistical significance was set at p < 0.05.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

K.N., H.K., and F.O. contributed to designing this study. K.N., H.K., F.O., J.R., A.M., J.M., A.L., Z.F., and K.M. performed the experiments. K.N., H.K., and F.O. analyzed the data and confirmed the results. K.N., H.K., and F.O. drafted the manuscript. H.K. supervised the project. All authors approved the final version.

Funding

This work was supported by JST SPRING, Grant Number JPMJSP2114. JSPS KAKENHI Grant Numbers JP21K10178 and JP22K17244 from the Japan Society for the Promotion of Science.

Data availability

The RNA-seq datasets generated and analyzed during this study are available in the NCBI Gene Expression Omnibus (GEO) under accession number GSE298138. All other data supporting the findings of this study are included in the article and its Supplementary Information files.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval

All animal experiments were performed in accordance with the ARRIVE guidelines and were approved by the Regulations for Animal Experiments and Related Activities at Tohoku University (2018DnA-028-06).