Luteolin reprograms macrophage polarization via the STING-TBK1 pathway to accelerate bone repair

Abstract

Macrophages play a pivotal role in bone regeneration, making their polarization a key target for immune regulation and therapeutic intervention. Modulating macrophage polarization represents a promising strategy for enhancing bone repair. Luteolin, a plant-derived flavonoid with well-documented anti-inflammatory properties, has been explored for its role in bone repair. However, its specific effects on macrophage polarization in bone repair remain unclear. This study investigates the role of luteolin in macrophage polarization and its underlying mechanisms. Our findings demonstrate that luteolin promotes M2 polarization while suppressing M1 polarization, as indicated by a reduction in the expression of pro-inflammatory markers, including IL-6 and iNOS, and an increase in the expression of anti-inflammatory factors, such as CD206, IL-10, and TGF-β. Mechanistically, luteolin inhibits STING oligomerization, thereby suppressing the STING-TBK1 pathway and mitigating downstream inflammatory responses. In vivo, in a mouse tibial bone defect model, luteolin effectively alleviates inflammation, facilitates angiogenesis, enhances collagen deposition, and improves bone density. Collectively, these findings highlight the potential of luteolin as a therapeutic agent for bone repair by modulating macrophage polarization and inhibiting STING-TBK1 signaling.

Introduction

Bone defects, characterized by compromised structural integrity and impaired regeneration, pose significant clinical challenges by disrupting biomechanical balance and sustaining inflammation [ 1, 2] . Current treatments, such as autologous bone grafting and tissue-engineered scaffolds, have shown limited success due to issues such as immune rejection, poor osteoinductive potential, and inadequate vascularization [ 3‒ 8] . Thus, it is crucial to create advanced therapies that support bone regeneration and address the limitations of current treatments.

Macrophages are key players in the innate immune system, influencing both the initiation and resolution of inflammation [9]. Found in nearly all tissues, these versatile cells are involved in infection, tissue repair, and regeneration [ 10‒ 13] . Increasingly, macrophages are recognized for their role in bone healing, where they polarize into unique phenotypes (M1/M2) to regulate the inflammatory environment [ 14, 15] .

M1 macrophages are pivotal in the initiation of inflammatory responses through the secretion of key inflammatory mediators, including interleukin-1 (IL-1), tumor necrosis factor-α (TNF-α), and IL-6 [16]. These mediators trigger localized inflammation and promote the differentiation of monocytes into osteoclasts [ 17‒ 21] , thereby accelerating bone resorption and contributing to a reduction in bone density. In contrast, M2 macrophages are involved in promoting tissue repair and regeneration through the secretion of diverse cytokines [22]. These cytokines support osteogenic differentiation, stimulate angiogenesis, and enhance extracellular matrix deposition, which collectively contribute to tissue healing and regeneration [ 23‒ 27] . Thus, modulating macrophage polarization represents a promising therapeutic approach for regulating bone homeostasis and enhancing bone regeneration, providing an effective strategy for the repair of bone tissue.

The stimulator of interferon genes (STING) is a critical adaptor protein involved in the regulation of innate immune responses and inflammation [28]. Aberrant activation of the cGAS-STING pathway initiates inflammatory cascades through downstream activation of the STING-IRF3 or NF-κB pathways, with key kinases such as TBK1 and IKK playing central roles in this process [ 29‒ 31] . Recent studies highlight the substantial involvement of the cGAS‒STING axis in the pathophysiology of various musculoskeletal disorders, including osteoarthritis (OA) [ 32, 33] . Previous study has indicated that the STING agonist DMXAA can shift macrophages from the M2 phenotype to the pro-inflammatory M1 phenotype, which is characterized by elevated levels of cytokines such as IL-6 and TNF-α [34]. In contrast, itaconate modulates the STING/NF-κB axis, promoting M2 polarization and reducing inflammation and tissue damage in OA [35]. These findings suggest that the cGAS-STING signaling pathway could serve as a therapeutic target for adjusting macrophage polarization. However, the exploration of interventions aimed at repairing bone defects through the modulation of macrophage polarization via the STING pathway remains underexplored.

Luteolin, a flavonoid with the chemical structure 3′,4′,5,7-tetrahydroxyflavone, is found in plants such as green pepper, celery, and perilla leaves [36]. It is known for its anticancer, antimicrobial, and anti-inflammatory properties [ 37‒ 39] . Studies indicate that luteolin prevents bone loss in Wistar rats with periodontitis by inhibiting osteoclast activity and reducing the expression of markers such as matrix metalloproteinase-9 (MMP-9) and receptor activator of nuclear factor κB ligand (RANKL). It also promotes osteoblastic activity, increasing tissue inhibitor of metalloproteinases-1 (TIMP-1), bone morphogenetic protein-2 (BMP-2), and osteoprotegerin (OPG) [40]. These findings suggest that luteolin may protect bone health by regulating both osteoclast and osteoblast functions. In addition to influencing bone cells, luteolin has been proven to regulate inflammatory factor expression in macrophages, enhancing their anti-inflammatory abilities [ 41, 42] . However, the mechanisms by which luteolin affects the polarization of bone marrow-derived macrophages (BMDMs) remain largely unexplored. This study examined the impact of luteolin on macrophage polarization and its therapeutic potential in a mouse model of unilateral tibial bone defects while exploring the relationships between these effects.

Materials and Methods

Cell culture

BMDMs were extracted from 6-week-old male C57BL/6 mice (Cavens, Changzhou, China). The mice were euthanized, and the femur and tibia were isolated. The residual connective tissue was removed from the bone surface, and the bone marrow was flushed with phosphate-buffered saline (PBS; Servicebio, Wuhan, China). The mixture was centrifuged at 180 g for 15 min, after which the red blood cells were lysed with red blood cell lysis solution (Servicebio). The samples were washed 1–2 times with PBS and cultured in α-minimal essential medium (αMEM; Gibco, Carlsbad, USA) supplemented with 10% FBS, 1% penicillin/streptomycin (P/S) (Invitrogen, Carlsbad, USA), and 25 ng/mL MCSF (Sigma-Aldrich, St Louis, USA) at 37°C with 5% CO ₂. After three days of culture, a half-medium change was performed. Adherent BMDMs were harvested for subsequent experiments.

Induction of macrophage polarization

BMDMs were plated at 2 × 10 ⁵ cells/mL in 6-well plates and divided into M1 and M2 groups. M1 macrophages were preincubated with 10 μM luteolin (MedChemExpress, Monmouth Junction, USA) for 6 h before being stimulated with 25 μg/mL DMXAA (MedChemExpress) or 1000 ng/mL lipopolysaccharide (LPS) (Sigma-Aldrich). In contrast, M2 macrophages were co-cultured with 10 μM luteolin or H-151 (MedChemExpress), together with 10 ng/mL IL-4 (PeproTech, Cranbury, USA) for a duration of 24 h.

Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) assay

RNA was extracted from the cells via TRIzol reagent (Vazyme, Nanjing, China) in accordance with the manufacturer’s instructions. Total RNA (500 ng) was transcribed into cDNA via a reverse transcription kit (#RR037A; TaKaRa, Kusatsu, Japan). The target gene expression levels were measured via RT-qPCR with SYBR Green qPCR premix reagents (#RR820B; TaKaRa). Gene expression was normalized to that of the reference gene 18S rRNA, and relative expression levels with standard errors were determined via the 2 ^–ΔΔCt method. The sequences of primers used are listed in Table 1.

Table 1 Sequences of primers used for RT‒qPCR

Gene	Primer sequence (5′→3′)
iNOS (mouse)	Sense	CTTCCGGGCAGCCTGTGAGACG
	Antisense	ATCCCCAGGTGTTCCCCAGGTAGG
IFN-β (mouse)	Sense	AGCTCCAAGAAAGGACGAACA T
	Antisense	GCCCTGTAGGTGAGGGTTGATCT
IL-6 (mouse)	Sense	GAGTCCTTCAGAGAGATACAG
	Antisense	TGGTCTTGGTCCTTAGCC
IL-1β (mouse)	Sense	CATCTTCTCAAAATTCGAGTGACAA
	Antisense	TGGGAGTAGACAAGGTACAACCC
Arg-1 (mouse)	Sense	CTCCAAGCCAAAGTCCTTAGAG
	Antisense	AGGAGCTGTCATTAGGGACATC
Fizz1 (mouse)	Sense	CCTGCTGGGATGACTGCTA
	Antisense	TGGGTTCTCCACCTCTTCAT
Ym1 (mouse)	Sense	CCCTGCTGGGATGACTGCTA
	Antisense	TGCAAGTATCTCCACTCTGGATCT
Gapdh (mouse)	Sense	GACAATTTTGGCATCGTGGA
	Antisense	ATGCAGGGATGATGTTCTGG

Western blot analysis

Tibial tissues or cells were lysed with RIPA buffer (Beyotime, Shanghai, China). Proteins were equally loaded and separated via sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), followed by transfer to a PVDF membrane (Millipore, Billerica, USA). The membrane was blocked with 5% skim milk in TBST (Beyotime) for 1 h and then incubated overnight at 4°C with primary antibodies (CST, Danvers, USA), including anti-TBK1/NAK (#D1B4; 1:1000), phospho-TBK1/NAK (#D52C2; 1:1000), NF-κB p65 (#D14E12; 1:1000), phospho-NF-κB p65 (#93H1; 1:1000), IRF-3 (#D83B9; 1:1000), phospho-IRF-3 (#4D4G; 1:1000), STING (#D2P2F; 1:1000), phospho-STING (#D8F4W; 1:1000), COL2a (#ER1906-48; 1:1000), COL1a1 (#E8F4L; 1:1000), and GAPDH (#D16H11; 1:1000). Following TBST washing, the membrane was incubated for 1 h with fluorescence-conjugated secondary antibodies (#5151; CST), and images were subsequently captured using the ChemiDocTM MP Imaging System (Bio-Rad Laboratories, Hercules, USA).

Cell immunofluorescence staining

BMDMs were plated at 4 × 10 ⁵ cells per well in 6-well plates and polarized toward the M1 or M2 phenotype. The cells were subsequently exposed to 10 μM luteolin. Following treatment, the cells were fixed with 4% paraformaldehyde (PFA) and permeabilized with 0.05% Triton X-100 for 1–2 min. To block nonspecific binding, the cells were incubated with 5% bovine serum albumin (Solarbio, Beijing, China) in PBS for 1 h. Primary antibodies targeting CD86 (#14-0862-82; 1:500, Invitrogen) or CD206 (#E6T5J; 1:200, CST) were then applied, and the cells were incubated overnight at 4°C. The next day, the cells were washed with PBS and incubated with secondary antibodies (#A-11008 or #A-11012; Invitrogen) for 1 h. The nuclei were stained with DAPI, and images were captured via a fluorescence microscope (LSM900; Zeiss Microscopy, Oberkochen, Germany).

Animal models and treatment

All animal experiments were reviewed and approved by the Ethics Committee of Shanghai University (No. ECSHU 2023-034). The environmental parameters of the animal room were strictly controlled to the following standards: room temperature 22°C–25°C, relative humidity (55% ± 10%), and a 12-h light/dark cycle. C57BL/6 mice were randomly assigned to four groups: the Sham, monocortical tibial defect (MTD), low-dose luteolin (LUT-L, 1 mg/kg), and high-dose luteolin (LUT-H, 5 mg/kg) groups. Under isoflurane anesthesia, the skin and muscle covering the tibia were carefully incised to expose the bone. A bone defect model, 1.2 mm in diameter and 1 mm in depth, was created on the unilateral tibia via a high-speed drill. After surgery, luteolin was administered via intraperitoneal injection every two days.

Microcomputed tomography (micro-CT)

After 7 and 10 days of intravenous luteolin injection, the mice were euthanized under general anesthesia, and their tibias were removed and preserved in 4% PFA. The morphology of the reconstructed tibias was assessed via a micro-CT scanner (Skyscan 1172; Bruker, Billerica, USA). Micro-CT analysis was used to quantify parameters such as the relative new bone volume to total tissue volume (BV/TV) and the trabecular number (Tb.N), along with other pertinent bone metrics.

Transcriptome analysis of tibial tissue

Tibial samples from the Sham, MTD, and LUT-H groups were collected for transcriptomics sequencing. Total RNA was extracted using Trizol reagent, and the concentration, purity, and integrity of the RNA were assessed using a Nanodrop 2000 spectrophotometer and agarose gel electrophoresis. The enriched mRNA was reverse-transcribed into cDNA, and sequencing libraries were prepared for high-throughput RNA sequencing on an Illumina platform. Differential expression analysis was conducted using DESeq2, with genes identified as significantly differentially expressed when P < 0.05 and log2 |Fold Change| > 1.5. Functional enrichment analysis, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotations, was performed using Goatools and KOBAS to identify biological processes, molecular functions, and metabolic pathways associated with differentially expressed genes. Data visualization was carried out using volcano plots, heatmaps, and pathway enrichment analysis to provide a comprehensive overview of the results.

Immunofluorescence staining

Tibial tissues from the Sham, MTD, LUT-L, and LUT-H groups were fixed in 4% PFA, followed by graded ethanol dehydration (75% ethanol for 4 h, 85% ethanol for 2 h, 90% ethanol for 2 h, 95% ethanol for 1 h, and absolute ethanol for 30 min). Tissues were then paraffin-embedded and sectioned at a thickness of 4 μm. After deparaffinization, antigen retrieval was performed using trypsin digestion. Sections were blocked with 10% goat serum and incubated with primary antibodies (Abcam, Cambridge, UK) targeting CD31 (Cat #ab222783; 1:100), F4/80 (Cat #ab300421; 1:100) or antibodies (Invitrogen) targeting CD86 (Cat #14-0862-82; 1:500) overnight at 4°C. Following PBS washes, sections were incubated with fluorophore-conjugated secondary antibodies [Goat anti-rabbit IgG (H+L) (Alexa Fluor™ 488, Cat # A-11008; 1:1000) or Goat anti-rabbit IgG (H+L) (Alexa Fluor™ 594, Cat # A-11012; 1:1000)] for 2 h at room temperature. After additional PBS washes, sections were mounted with a DAPI-containing mounting medium. Images were captured using a fluorescence microscope (LSM900; Zeiss Microscopy).

Hematoxylin-eosin and Masson staining

Tibial tissues from the Sham, TMD, LUT-L, and LUT-H groups were harvested, fixed in 4% paraformaldehyde, and subsequently decalcified in 12% EDTA solution for one month. Following decalcification, the samples were processed for paraffin embedding and sectioned into 4–7 μm. Sections were stained with hematoxylin and eosin (H&E) (Solarbio) and Masson trichrome stain (Leagene, Beijing, China) for histological analysis according to standard procedures.

ELISA analysis

Blood samples were obtained from the Sham, MTD, LUT-L, and LUT-H groups. The samples were subsequently centrifuged at 1123 g for 15 min at 4°C. The serum was analyzed with ELISA kits (Multisciences, Hangzhou, China) for IL-1β, IL-6, IL-10, TGF-β, and TNF-α according to the manufacturer’s instructions.

Liver and renal function detection

Whole blood samples were obtained from the Sham, MTD, LUT-L, and LUT-H groups. The samples were incubated at room temperature for 30 min before being centrifuged at 1123 g for 15 min at 4°C. Measurements of alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and creatinine (Cr) levels were conducted using corresponding kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instructions.

Statistical analysis

The results are presented as the mean ± standard deviations (SDs). To evaluate the significance of group differences, one-way ANOVA was conducted, followed by post hoc tests or t tests. Animal survival curves were analyzed via the log-rank (Mantel-Cox) test. A P value less than 0.05 was considered statistically significant.

Results

Luteolin suppresses M1 macrophage polarization through the STING-TBK1 signaling pathway

To evaluate the effect of luteolin on M1 macrophage polarization, BMDMs were pretreated with 10 μM luteolin and activated with DMXAA, a mouse STING-specific agonist ( Figure 1A). Luteolin at 10 μM was selected on the basis of prior studies demonstrating optimal bioactivity and minimal cytotoxicity [ 43, 44] . Western blot analysis was performed to quantify the expression of inducible nitric oxide synthase (iNOS), a key marker of M1 polarization. The results showed that DMXAA treatment significantly increased iNOS expression in BMDMs, an effect that was notably reduced by luteolin ( Figure 1B,C). RT-qPCR analysis demonstrated that luteolin counteracted the DMXAA-induced upregulation of M1 macrophage-associated genes, including iNOS, IL-1β, and IL-6 in BMDMs ( Figure 1D). Given that CD86 is a vital marker for M1 polarization in macrophages [45], we performed immunofluorescence staining to assess alterations in CD86 expression. The findings revealed a notable increase in CD86 expression in BMDMs after LPS stimulation compared with that in the control group. However, this upregulation was effectively inhibited by luteolin ( Figure 1F,G). These results suggest that luteolin effectively prevents macrophage polarization to the M1 phenotype.

Figure 1 .

Luteolin suppresses the polarization of macrophages toward the M1 subtype

(A) Flow chart of M1 macrophages. Bone marrow cells were extracted from C57BL/6 mice and cultured into BMDMs. BMDMs were pretreated with luteolin and subsequently stimulated with DMXAA. (B) BMDMs were treated with or without luteolin or DMXAA, and the expression of iNOS was detected via western blot analysis. (C) Quantitative analysis of the ratio of the gray values of iNOS and GAPDH. (D,E) The mRNA levels of iNOS,IL-1β,IL-6 (D), and IFN-β (E) were detected by RT-qPCR after treatment with luteolin and DMXAA. The results are shown as the mean ± SD; ****P < 0.0001 vs the DMSO group. (F) Fluorescence staining for CD86 in BMDMs after treatment with luteolin or LPS. Green represents CD86, and blue represents the cell nucleus. Scale bar: 20 μm. (G) Analysis of the mean fluorescence intensity (MFI) of CD86. (H) STING dimerization was analyzed by immunoblotting. (I–M) Western blot analysis was performed to determine the protein levels of (I) p-STING, STING, p-TBK1, TBK1, p-p65, p65, p-IRF3, and IRF3. (J–M) Quantitative analysis was conducted by Image J. The data are presented as the mean ± SD of triplicate experiments. **P < 0.01, and ****P < 0.0001 vs the DMXAA group.

We then explored how luteolin affects the cGAS-STING pathway. As previously mentioned, the recognition of aberrant cytosolic DNA by cGAS potently activates the cGAS-STING signaling pathway, ultimately leading to the production of IFN-β [29]. In this study, luteolin attenuated the DMXAA-induced upregulation of IFN-β mRNA ( Figure 1E). Subsequently, BMDMs were pretreated with luteolin, followed by DMXAA stimulation, and STING oligomerization was assessed by immunoblotting. The results indicated that DMXAA stimulation led to substantial dimerization and oligomerization of STING in BMDMs, whereas luteolin pretreatment notably diminished these effects ( Figure 1H). Compared with unstimulated BMDMs, BMDMs treated with DMXAA exhibited significantly elevated levels of phosphorylated STING, TBK1, and IRF3 ( Figure 1I–M). Luteolin treatment significantly reduced the activation of the STING-TBK1 pathway, thereby inhibiting the polarization of BMDMs toward the M1 phenotype.

Luteolin promotes macrophage polarization to the M2 phenotype

To determine whether luteolin promotes the polarization of BMDMs toward the anti-inflammatory M2 phenotype, BMDMs were treated with luteolin, H-151 (a STING inhibitor), or IL-4 for 24 h ( Figure 2A). Figure 2B shows that treatment with either luteolin or H-151 increased the protein levels of Arg-1, a marker of M2 macrophages, compared with those in the control group. Moreover, cotreatment with luteolin and IL-4 resulted in a dramatic increase in Arg-1 expression, as assessed by western blot analysis ( Figure 2B,C). Similar results were observed at the mRNA level. The combination of luteolin and IL-4 also significantly increased the expression of M2-related genes, including Arg-1, Fizz1, and Ym1 ( Figure 2D). Further immunofluorescence analysis revealed a significant increase in the expression of CD206, an M2 macrophage marker, following IL-4 stimulation compared with that in the control group ( Figure 2E). Pretreatment with luteolin further intensified the IL-4-induced increase in CD206 expression, a trend supported by the MFI data ( Figure 2F). We next investigated the role of STING agonists in M2 polarization. Western blot analysis showed that DMXAA treatment did not affect Arg-1 expression ( Figure 2G,H). RT-qPCR analysis revealed that DMXAA reversed the increase in the levels of M2 macrophage marker genes, including Arg-1, Fizz1, and Ym1 in BMDMs treated with luteolin ( Figure 2I). These findings suggest that luteolin promotes M2 polarization of macrophages in vitro.

Figure 2 .

Luteolin promotes M2 macrophage polarization

(A) Schematic illustration showing the process by which M2 macrophages are induced from BMDMs. (B) BMDMs were co-cultured with either H-151 or luteolin. After 24 h of incubation, western blot analysis was performed to determine the protein levels of Arg-1. (C) Western blot images were quantified using ImageJ grayscale analysis and are presented relative to the level of GAPDH. (D) RT-qPCR was conducted to detect the mRNA levels of Arg-1,Fizz1, and Ym1. (E) Immunofluorescence images of classical macrophage markers (CD206 for the M2 type) in different groups. Scale bar: 20 μm. Red fluorescence, CD206; blue fluorescence, nuclei stained with DAPI. (F) Statistical analysis of the MFI based on the fluorescence images. (G) BMDMs were treated with or without luteolin or DMXAA, and the expression of Arg-1 was detected by western blot analysis. (H) Western blot images were quantified via ImageJ grayscale analysis and are presented as the relative Arg-1 level to the GAPDH level. (I) RT-qPCR analysis of the relative mRNA expression of Arg-1,Fizz1, and Ym1 in DMSO-, DMXAA- and luteolin-treated BMDMs. The data are presented as the mean ± SD. Statistical significance is denoted as follows: **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Luteolin accelerates bone healing

To assess the effect of luteolin on bone repair in vivo, a MTD model was established in mice ( Figure 3A). The mice received intraperitoneal injections of 1 mg/kg (LUT-L) or 5 mg/kg (LUT-H) luteolin. Tibial samples were collected on days 7 and 10 and analyzed via micro-CT ( Figure 3B). The results revealed the formation of new bone tissue at the defect edges in both the LUT-L and LUT-H groups by day 7. Compared with the MTD group, the LUT-H group presented significantly more new bone formation at the defect center on days 7 and 10. Quantitative analysis of the bone parameters revealed an increase in the BV/TV, Tb.Th, and Tb.N, accompanied by a reduction in the Tb.Sp, in the LUT-H group relative to the MTD group on day 7. These trends were even more pronounced on day 10 ( Figure 3C). Histological evaluations with H&E and Masson staining revealed a significant increase in regenerated bone tissue in both the LUT-L and LUT-H groups compared with the MTD group on day 10 ( Figure 3D,E), underscoring the potential of luteolin to enhance bone regeneration.

Figure 3 .

In vivo bone regeneration evaluation of luteolin in the MTD model

(A) Schematic illustration of the MTD model and its therapeutic process. (B) Representative 3D-reconstructed images obtained from micro-CT scans of tibias after intraperitoneal injection of luteolin for 7 and 10 days. (C) BV/TV, Tb.Th, Tb.N, and Tb.Sp obtained from the micro-CT results. The results are shown as the mean ± SD (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. (D,E) Histological analysis of tibia sections, including H&E staining (D) and Masson staining (E), on days 7 and 10. Scale bar: 250 μm for H&E staining; Scale bar: 250 μm for Masson staining.

Transcriptome analysis of the effects of luteolin on bone healing

A transcriptomic analysis was carried out to elucidate the possible mechanisms by which luteolin promotes bone repair. Principal Component Analysis (PCA) revealed clear distinctions in the transcript profiles among the Sham, MTD, and LUT-H groups ( Supplementary Figure S1A–C). Volcano plots indicated that, compared with the sham group, the MTD group exhibited significant upregulation of 240 DEGs and downregulation of 50 DEGs ( Figure 4A). Compared with those in the MTD group, 33 genes were upregulated, and 88 genes were downregulated in the LUT-H group ( Figure 4D). The volcano plot further highlighted collagen as a significantly upregulated gene in the MTD group. This increase, in combination with the expression of the MMP gene family, suggests a dynamic process of matrix degradation and regeneration, pointing to an active stage of inflammation and repair ( Figure 4B). These findings were corroborated by gene set enrichment analysis (GSEA; Figure 4G). Additionally, the analysis revealed that genes associated with immune responses, such as Oas and Ifit, were significantly downregulated in the LUT-H group compared with the MTD group ( Figure 4E). GO enrichment analysis of the DEGs revealed that luteolin treatment in MTD mice influenced key pathways, such as immune system processes and type I interferon signaling ( Figure 4C,F). Through Venn diagram analysis, 8 key DEGs were identified between the groups, including C3ar1, Sirpb1a, Igkv8-21, Angptl2, Acta2, Lrrc15, Tagln, and Ankrd2. These genes were further validated through RT-qPCR in each group of mice ( Figure 4H). These findings suggest that luteolin promotes bone repair through the modulation of relevant immune signaling pathways.

Figure 4 .

Transcriptome analysis of the role of luteolin in promoting bone repair

(A) Volcano plot of upregulated genes and downregulated genes in the sham vs. MTD; upregulated DEGs: red dots; downregulated DEGs: blue dots. (B) Heatmap of the DEGs in the MTD group compared with those in the Sham group. (C) GO enrichment analysis of DEGs in the sham vs. MTD groups. (D) Volcano plot of upregulated genes and downregulated genes in the MTD vs. luteolin. (E) Heatmap of the DEGs in the MTD vs. luteolin, with the color bar on the right indicating expression levels, red indicating high expression and gray indicating low expression. (F) GO enrichment analysis of DEGs in the MTD vs. luteolin. (G) GSEA plot of DEGs in the sham vs. MTD groups. (H) Common DEGs in the sham vs. MTD and MTD vs. luteolin. RT-qPCR analysis of the mRNA expression levels of representative genes. The data are presented as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Luteolin promotes M2 polarization in the bone tissues of MTD model mice

We next assessed the serum levels of IL-β, IL-6, TNF-α, and IL-10 in the Sham, MTD, LUT-L, and LUT-H groups. The findings indicated a notable decrease in pro-inflammatory cytokines IL-1β, IL-6, and TNF-α and a significant increase in the anti-inflammatory cytokine IL-10 in both the LUT-L and LUT-H groups on day 7 ( Figure 5A). Immunofluorescence staining confirmed a decrease in the number of M1 macrophages (CD86 ⁺/F4/80 ⁺ cells) and an increase in the number of M2 macrophages (CD206 ⁺/F4/80 ⁺ cells) in the luteolin-treated groups compared with those in the MTD group ( Figure 5B). These results collectively suggest that luteolin suppresses M1 macrophage polarization and promotes M2 polarization, thereby contributing to the enhancement of bone repair.

Figure 5 .

The impact of luteolin on macrophage polarization in MTD-treated mice

(A) The expression levels of IL-1β, IL-6, TNF-α, and IL-10 in the serum of each group were determined via ELISA kits and analyzed via histograms. The results are shown as the mean ± SD (n = 6). **P < 0.01, ***P < 0.001, and ****P < 0.0001 vs the MTD group. (B) Representative immunofluorescence staining of F4/80 (green) and CD86 (red) in the left images and of F4/80 (green) and CD206 (red) in the right images. Scale bar: 50 μm.

Luteolin promotes vascular differentiation and collagen generation

Since collagen formation and vascularization are essential for tissue healing and remodeling, the impact of luteolin on these processes was examined. The ELISA results showed that the LUT-L and LUT-H groups exhibited significantly enhanced serum TGF-β levels compared to the MTD group on day 10 ( Figure 6A). Western blot analysis of bone tissue confirmed an increase in the expression levels of COL1a1 and COL2a in both the LUT-L and LUT-H groups compared with those in the MTD group ( Figure 6B,C), which was consistent with the transcriptomic analysis findings. To investigate the potential of luteolin to improve vascularization in the MTD model, double fluorescence staining for CD31 and α-SMA was conducted on days 7 and 10 postluteolin injection. These markers, which represent endothelial and smooth muscle cells, are crucial elements of blood vessels [46]. Compared with those in the MTD group, the LUT-L and LUT-H groups exhibited increased areas of CD31 and α-SMA immunofluorescence, indicating enhanced vascular formation ( Figure 6D). The fluorescence intensities of these angiogenic markers were notably higher in the LUT-H group than in the other groups on day 10. These findings suggest that luteolin accelerates bone repair by promoting collagen synthesis and vascular differentiation.

Figure 6 .

In vivo evaluation of collagen formation and vascularization by luteolin using the MTD model

(A) TGF-β concentrations in the serum of each group were detected by ELISA (n = 6). The data are expressed as the mean ± SD of three independent experiments. (B) COL1a1 and COL2a protein expression was detected via western blot analysis. (C) Quantification of bands of COL1a1 and COL2a with GAPDH. **P < 0.01, ***P < 0.001, and ****P < 0.0001. (D) Immunofluorescence staining of CD31 (red) and α-SMA (green) expression in the MTD, LUT-L, and LUT-H groups. The cell nuclei were stained with DAPI (blue). Scale bar: 50 μm.

Evaluation of the safety of luteolin in mice in vivo

The safety of luteolin in mice was assessed by assessing systemic toxicity, with daily recordings of body weights indicating a gradual weight increase among the different groups. Luteolin treatment did not significantly affect the body weights of the mice compared with those of the sham group ( Figure 7A). Liver function markers (ALT, AST) and kidney function parameters (BUN, Cr) were examined. After treatment with luteolin, all parameters were verified to be normal ( Figure 7B,C). Furthermore, H&E staining showed that there were no obvious pathological changes in major organs (livers or kidneys) between the sham and luteolin-treated groups ( Figure 7D), indicating that luteolin did not cause notable adverse effects.

Figure 7 .

In vivo safety evaluation of luteolin

(A) Body weights of the mice in the different groups. (B,C) Quantification of liver functional indicators (ALT and AST) and kidney function parameters (BUN and Cr) in the serum from different groups. (D) H&E staining of liver and kidney tissues from the Sham, MTD, LUT-L, and LUT-H groups. Scale bar: 200 μm.

Discussion

Bone defects are common clinical challenges that require effective therapeutic strategies. The healing process begins with an inflammatory response, during which immune cells are recruited to the injury site, along with the release of various signaling molecules from the circulatory system [47]. As key components of the innate immune system, macrophages play crucial roles in bone tissue repair, with moderate inflammation promoting healing and regeneration [48]. However, excessive or prolonged inflammation can cause tissue damage, hindering the repair process and delaying bone recovery.

Macrophage polarization is crucial for regulating inflammation, with the M1 and M2 phenotypes playing distinct and complementary roles. M1 macrophages are pro-inflammatory and promote inflammation through the secretion of cytokines such as iNOS and TNF-α, which can impede tissue repair [49]. In contrast, M2 macrophages are anti-inflammatory, secreting cytokines such as IL-10 that facilitate the resolution of inflammation and promote tissue repair [50]. The balance between M1 and M2 polarization is essential for controlling inflammation and ensuring effective tissue repair.

Luteolin, a flavonoid prevalent in various medicinal herbs and vegetables, is known for its potent anti-inflammatory and antioxidant effects [37]. However, the molecular basis of its role in macrophage polarization remains insufficiently defined. In this study, we established classical M1 and M2 macrophage models by stimulating BMDMs with LPS and IL-4, respectively, and confirmed the expression of typical pro- and anti-inflammatory markers. Luteolin treatment of M1-polarized macrophages significantly suppressed pro-inflammatory cytokines (iNOS, IL-1β, IL-6, TNF-α, and CD86) while increasing the expression of anti-inflammatory mediators (Arg-1, IL-10, and CD206), suggesting that luteolin facilitates a phenotypic shift from the M1 phenotype to the M2 phenotype, potentially promoting tissue repair.

Although the exact molecular targets of luteolin remain to be fully elucidated, accumulating evidence suggests that it may modulate several key inflammatory pathways. Previous studies have shown that luteolin inhibits the TLR4/TRAF6/NF-κB axis, thereby attenuating inflammatory signaling [ 51, 52] . It also interferes with NLRP3 inflammasome activation by blocking apoptosis-associated speck-like protein containing a CARD (ASC) oligomerization and caspase-1 cleavage [53], which aligns with the reduced IL-1β expression observed in our study. Notably, our findings implicate the STING-TBK1 axis in luteolin-mediated effects, suggesting that this pathway may serve as a direct or indirect target. Despite its therapeutic promise, the low bioavailability of luteolin, owing to its poor stability and inefficient absorption, remains a major limitation [54]. The development of optimized delivery systems or pharmaceutical formulations is critical for enhancing the stability and clinical utility of these materials in treating inflammatory diseases.

STING serves as a crucial mediator, influencing a range of autoimmune, autoinflammatory, and degenerative diseases [ 55‒ 60] . As a result, small-molecule inhibitors targeting STING have emerged as promising therapies for inflammatory diseases. Previous studies have demonstrated that STING expression is upregulated in macrophages undergoing M1 polarization. STING agonists can repolarize M2 macrophages toward an M1-like phenotype [61]. However, the role of STING in regulating macrophage polarization during bone healing remains largely unexplored. Our results show that DMXAA induces the polarization of BMDMs toward the M1 phenotype, as indicated by a significant increase in CD86 and iNOS expression, along with elevated levels of pro-inflammatory cytokines, including IL-6 and TNF-α. Notably, pre-treatment with luteolin effectively attenuates these changes, suggesting that luteolin modulates macrophage polarization.

Recent research has highlighted the importance of TBK1, a member of the IKK family, in the STING signaling pathway. TBK1 dimerizes with STING, leading to its activation. This activation triggers the phosphorylation of IκB proteins, resulting in their degradation and subsequent activation of the NF-κB pathway [ 62, 63] . In this study, we assessed the effects of luteolin on the STING pathway in DMXAA-stimulated BMDMs. Our findings show that DMXAA induces STING pathway activation, which is marked by increased pSTING, pTBK1, and pIRF3 expression. Luteolin treatment significantly reduced the expression of these proteins, indicating its role in preventing STING pathway activation. To further understand how luteolin inhibits STING pathway activation, we examined STING oligomerization, a key step in the pathway. Our study demonstrates that luteolin markedly decreases STING oligomerization, leading to the inhibition of the STING-TBK1 pathway activation.

In summary, our study found that luteolin treatment in the MTD model promotes bone repair by enhancing bone volume, density, and regeneration, as well as increasing the formation of CD31 ⁺/α-SMA ⁺ blood vessels, which contributes to both bone and blood vessel formation.

Supplementary Data

Supplementary data are available at Acta Biochimica et Biophysica Sinica online.

COMPETING INTERESTS

The authors declare that they have no conflict of interest.

Funding Statement

This work was supported by the grants from the National Natural Science Foundation of China (No. 81981340417) and the Central Guidance on Local Science and Technology Development Fund of Sichuan Province (No. 2023ZYD0062).