1. Introduction
Rheumatoid arthritis (RA) affects approximately 0.5% to 1% of the global population and is characterized by inflammation and bone destruction in affected joints, ultimately leading to joint deformities and loss of function
[1],
[2]. According to previous reports, more than 60% of RA patients experience bone erosion within two years of disease onset, and erosive bone damage is an important determinant of long-term functional outcomes
[3]. With the continuous development of disease-modifying antirheumatic drugs (DMARDs), major success has been achieved in preventing and alleviating symptoms in RA patients. However, these drugs are often unable to reverse existing joint destruction, and even with potent anti-inflammatory treatment strategies such as the blockade of tumor necrosis factor-α (TNF-α) or interleukin-6 receptor (IL-6R), only limited signs of bone erosion repair are observed
[4],
[5]. In contrast, therapies directly targeting osteoclasts (OCs) have received relatively less attention, despite the pathological role of these cells in bone resorption driven by their excessive generation and/or increased activity
[6]. Although several therapeutic agents are available to address joint destruction by inhibiting OC-mediated bone resorption, their use is often accompanied by significant adverse effects, including stroke, atypical femoral fractures, and osteonecrosis of the jaw
[7],
[8]. These clinical challenges highlight the incomplete understanding of the mechanisms underlying RA-associated bone erosion, particularly the molecular and cellular pathways regulating osteoclastogenesis, differentiation, and bone resorption. Furthermore, shifting the therapeutic focus from traditional inflammation-centered approaches to strategies emphasizing osteocyte biology represents a promising avenue for managing OC-mediated focal joint destruction in RA patients.
N
6-methyladenosine (m
6A) methylation, the most prevalent form of RNA methylation and an epigenetic modification, plays a critical role in regulating the functions of fibroblast-like synoviocytes (FLSs) and immune cells in RA
[9],
[10],
[11]. m
6A methylation is regulated primarily by the methyltransferase-like 3 (METTL3)/METTL14/Wilms tumor 1-associated protein (WTAP) methyltransferase complex and is reversed by the demethylases fat mass and obesity-associated protein (FTO) and AlkB homolog 5 (ALKBH5). This modification is recognized by m
6A reader proteins, including members of the YT521-B homology (YTH) family (YTH domain family protein 1 (YTHDF1), YTHDF2, YTHDF3, YTH domain containing protein 1 (YTHDC1), and YTHDC2) and the insulin-like growth factor 2 messenger RNA (mRNA)-binding protein (IGF2BP) family (IGF2BP1, IGF2BP2, and IGF2BP3), and modulates RNA stability, transcription, translation, and degradation, thereby playing a pivotal role in cellular processes
[12],
[13]. Previous studies have shown that ALKBH5 expression is increased in RA FLSs, where it exacerbates synovial invasion and inflammation by modulating the m
6A methylation of cholesterol-25-hydroxylase (CH25H)
[14]. Multiple machine learning methods have shown that IGF2BP3, which can regulate the G2/M transition, promotes RA-FLS proliferation, and affects M1 macrophage polarization, suggesting that it can be used for the high-precision prediction of an RA diagnosis
[13]. Other studies have shown that METTL14, a methyltransferase, promotes FLS migration and invasion, as well as the TNF-α-induced inflammatory response
[15]. The levels of METTL14 and m
6A in the peripheral blood mononuclear cells (PBMCs) of RA patients are reduced and negatively correlated with the disease activity score according to a 28-joint count (DAS28)
[11]. m
6A modifications targeting OCs have also been gradually reported in recent years. In primary bone marrow mononuclear cells and RAW264.7 cells, METTL3 deletion inhibits the expression of osteoclast-specific genes (
Nfatc1,
c-Fos,
Ctsk, and
Dc-stamp).
METTL3 knockout increases the expression and stability of the
Atp6v0d2 mRNA through YTHDF2, resulting in the formation of larger multinucleated OCs with reduced bone-resorbing ability
[16]. In RAW264.7 cells, METTL14 was found to regulate OC-mediated bone resorption by methylating specific functional sites of NFATc1. Upon increasing the m
6A methylation level, YTHDF1 and YTHDF2 exert antagonistic effects on the posttranscriptional regulation of
Nfatc1 [17]. Recent studies have shown that m
6A modifications occur in key cell types involved in RA pathogenesis, regulating their functions and contributing to disease progression. However, the methylation processes in OCs remain underexplored, and research on m
6A regulation of OC-mediated bone erosion is largely indirect. Investigating the roles of m
6A in OC differentiation and function could provide valuable insights into the mechanisms underlying RA-associated bone erosion and facilitate the identification of novel therapeutic targets.
Combination therapy for RA is a new type of treatment method, and the synergistic effect of drugs can further increase drug efficacy, playing a complementary role
[18]. The combination of Chinese herbal monomers, the main active ingredients of traditional Chinese medicine, has potential
[19],
[20]. As a basic therapy,
Tripterygium wilfordii Hook. F. (TwHF) is the most effective and key traditional Chinese medicine for treating RA and is often used alone or in combination with methotrexate (MTX)
[21]. However, excessive use of TwHF can cause multiorgan toxicity, which hinders the widespread use of traditional Chinese medicine in clinical practice
[22],
[23]. An increasing number of studies are combining TwHF with other drugs to compensate for their shortcomings
[24]. Triptolide (TP) is one of the main active ingredients of TwHF and has been shown to regulate m
6A methylation and inhibit OC differentiation to alleviate RA
[25].
Spatholobus suberectus (
S. suberectus) Dunn is a commonly used traditional Chinese medicine and is considered a nontoxic treatment for RA
[26]. Clinical studies have shown that the combination of
S. suberectus Dunn and TwHF has a synergistic effect
[27]. The active ingredient in
S. suberectus Dunn, medicarpin (Med), can inhibit OC differentiation
[28]. Preliminary studies suggest that Med not only inhibits OC differentiation but also increases the efficacy of TP, allowing for a reduced dosage and potentially minimizing its associated toxicity. This study aimed to explore the synergistic effects of TP and Med on RA and identify their potential m
6A methylation targets. Therefore, determining the effects and potential mechanisms of TP and Med combination therapy in the inhibition of osteoclastogenesis and differentiation to alleviate RA-associated bone erosion is clinically important and valuable.
In the present study, we analyzed the effects of TP and Med, both individually and in combination, on a collagen-induced arthritis (CIA) rat model and observed that the combined treatment significantly impacted OC formation and bone destruction. By systematically analyzing m
6A methylation targets from clinical samples alongside the molecular targets of the drugs, we identified METTL3 and YTHDF1 as critical mediators of the synergistic effects of TP and Med. Further
in vivo and
in vitro experiments confirmed that combined treatment with TP and Med targeted m
6A methylation pathways to synergistically inhibit OC formation and function (
Fig. 1). This approach provides deeper insights into the mechanisms underlying the synergistic actions of these drugs.
2. Materials and methods
2.1. Induction of arthritis and treatment
Fifty male Sprague–Dawley (SD) rats, aged 6–8 weeks with a mean weight of 180–200 g, were purchased from the National Institutes for Food and Drug Control (license number: SCXK (Beijing) 2024-0001). All the animals were maintained under specific pathogen-free environmental conditions ((25 ± 2) °C, (55 ± 5)% humidity, and a 12 h/12 h light/dark cycle) for three days before the experiments.
After acclimation, the rats were divided into six groups: the normal control (Con), model control (CIA), MTX, TP, Med, and 1/2 TP + 1/2 Med (T+M) groups. Except for those in the normal control group, the rats in the other groups were intradermally immunized with 100 μL of an emulsified mixture of bovine type II collagen (Chondrex, USA) and incomplete Freund’s adjuvant (IFA; Chondrex) at a 1:1 ratio on one side at the base of the tail on day 1. On day 7, 50 μL of the emulsified mixture was administered on the other side for a booster immunization. The experimental procedure was performed as described previously
[29]. A total of 50 rats were included in the study. After eight rats from the control group were excluded, 42 rats were subjected to the CIA model, 40 of which successfully developed arthritis, resulting in an incidence rate of 95% in the CIA model.
On day 10, the rats were randomly divided into five groups and orally administered the following agents for 21 days: the MTX group received 1.5 mg∙kg
–1 MTX (MTX tablets; Shanghai Sine Pharmaceutical Co., Ltd., China), the TP group received 18.62 μg∙kg
–1 TP (B01216; Yongjian Pharmaceutical Co., Ltd., China)
[30], the Med group received 2 mg∙kg
–1 Med (B01086), and the T+M group received 9.31 μg∙kg
–1 TP and 1 mg∙kg
–1 Med once a day. The CIA group received the solvent control of 0.9% stoke-physiological saline solution (BL158; Biosharp, China). All the experimental procedures were approved by the Ethics Committee of the Institute of Clinical Medical Sciences of China–Japan Friendship Hospital (ethical approval number: zryhyy11-23-04-02).
2.2. Arthritis evaluation
Beginning on the day of the booster immunization (day 7) and drug administration (day 10), the weight and arthritis index score of the hind limbs of the rats were recorded every three days. The severity of arthritis per hind leg was reported as the arthritic index (AI) score on a scale of 0–4 points according to conventional criteria: 0 = no change, 1 = redness or slight swelling, 2 = mild swelling, 3 = pronounced swelling, and 4 = deformity of and inability to use the limb. Rats with an AI score of 0 on day 10 were considered to have failed modeling and were excluded. The experimental procedure was performed as described previously
[31].
2.3. Micro-computed tomography (micro-CT)
Micro-CT scans of the ankle and knee joints were performed using a SKYSCAN 1174 Micro CT (Bruker, Belgium) with a resolution of 10.2 μm to evaluate the extent of bone erosion. The matching software N-Recon was then used for three dimensional (3D) image reconstruction of the ankles and knees. 3D analyses of the bone volume (BV), bone surface (BS), BS/BV, and BV/total volume (TV) were performed with CT-AN software. Finally, the images were processed with CTvox software.
2.4. Histological analysis, histology, and immunohistochemistry (IHC)
On day 31, the rats were anesthetized with isoflurane gas (RWD, China) and sacrificed. The left ankle and knee joints were isolated and fixed with 4% paraformaldehyde (Solarbio, China) for one week and then decalcified in 10% ethylenediaminetetraacetic acid (EDTA; Sinopharm Chemical, China) for two months to prepare paraffin sections. The joint slices were stained with hematoxylin and eosin (HE; Solarbio). The histological score was evaluated by assessing inflammation, cartilage damage, bone damage, inflammatory cell infiltration, synovial hyperplasia, and pannus on a scale from 0 to 4 according to the severity.
For IHC, the slices were subjected to antigen retrieval by immersion in sodium citrate buffer at 99 °C for 20 min, followed by an incubation with primary antibodies against METTL3 (1:1000 dilution, 15073-1-AP; Proteintech, China) and YTHDF1 (1:1000 dilution, 17479-1-AP; Proteintech) overnight at 4 °C and then with the appropriate secondary antibody. Afterward, the slices were processed with a DAB kit (PV-6000D; ZSGB-BIO, China) for color reactions.
2.5. Tartrate-resistant acid phosphatase (TRAP) and tetramethylrhodamine isothiocyanate (TRITC)–phalloidin staining
TRAP staining was performed according to the manufacturer’s instructions (387A; Sigma, USA) to evaluate OC formation. TRITC–phalloidin staining of F-actin rings was conducted according to the manufacturer’s instructions (HB221130; Yeasen, China) to assess the bone resorption capacity of OCs. The slices were scanned using a 3DHISTECH Pannoramic SCAN (Hungary) and analyzed with Case Viewer. OC numbers and areas were quantified using ImageJ.
2.6. Flow cytometry and enzyme-linked immunosorbent assay (ELISA)
Fresh popliteal spleens were isolated and carefully ground through a cell strainer. Splenocytes were further passed through a 70 μm cell strainer, followed by the lysis of erythrocytes. Antibodies against fluorescein isothiocyanate (FITC)-cluster of differentiation 4 (CD4), allophycocyanin (APC)-CD25, and phycoerythrin (PE)-forkhead box protein P3 (FoxP3) (eBioscience, USA) were used to stain regulatory T cells (Tregs), and antibodies against FoxP3 were added to the cells after fixation and permeabilization with a FoxP3/transcription factor staining buffer set (eBioscience). The cells were detected with a Beckman instrument and analyzed using FACSDiva software. Sample processing and data analysis were performed as previously described
[32].
The serum levels of TNF-α, IL-1β, IL-6, IL-10, and transforming growth factor-β (TGF-β) were examined using ELISA kits (E-EL-R0012, E-EL-R0015, E-EL-R2856, E-EL-R0016, and E-EL-0162; Elabscience, China) according to the manufacturer’s instructions.
2.7. OC preparation
Primary bone marrow cells from 6–8-week-old C57BL/6J mice (Charles River, China) were isolated. The cells were flushed from the diaphysis of the bilateral tibias and femurs and carefully passed through a 70 μm cell strainer. After the removal of erythrocytes, the cells were cultured in α-minimum essential medium (α-MEM; Gibco, USA) containing 20 ng∙mL
–1 M-CSF (Peprotech, USA) overnight. Floating cells were collected and maintained in the same medium for three days to obtain bone marrow monocytes (BMMs). For OC induction, 50 ng∙mL
–1 RANKL (Peprotech) was added to the culture medium for 4–6 d, and 10 nmol∙L
–1 TP and 10 μmol∙L
–1 Med were added to the medium. After the induction day, the culture medium was replaced with fresh medium every two days
[33].
2.8. Cell viability assay
BMMs were prepared in a 96-well plate as described above, and 0–80 nmol∙L–1 TP and 0–80 μmol∙L–1 Med were used to maintain OC differentiation for five days. Then, a 10% cell-counting kit-8 (CCK-8; GK10001; Glpbio, USA) solution was added to each well, and the plate was maintained in an incubator at 37 °C for 2 h. The absorbance at 450 nm was measured with a microplate reader (Thermo Fisher, China).
2.9. Scanning electron microscopy (SEM)
Bone resorption pits were imaged via SEM, and the bone samples were first cleaned to remove the soft tissue. The samples were fixed with 2.5% glutaraldehyde in phosphate-buffered saline (PBS) at 4 °C. Bone slices were mounted on the SEM stakes with the front side facing up using conductive tape. A thin layer of gold was plated on the front side. The samples were placed in the SEM chamber and photographed. The bone resorption pit area was analyzed using ImageJ.
2.10. Real-time quantitative polymerase chain reaction (RT-qPCR)
Total RNA was extracted with a Fast Pure Cell/Tissue Total RNA Isolation Kit (RN001; Yishan, China) according to the manufacturer’s instructions and quantified using a NanoDrop 2000 spectrophotometer, after which it was reverse transcribed into complementary DNA (cDNA) using the Evo M-MLV Reverse Transcription Reagent Premix (AG11706; Accurate Biology, China). RT-qPCR was performed on a QuantStudioTM 5 RT–PCR System using the SYBR Green Pro Taq HS premixed qPCR Kit (AG11701; Accurate Biology). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. Table S1 in Appendix A lists the primers used in this study. Relative gene expression was determined with the 2–△△Ct method.
2.11. Western blotting
Radioimmunoprecipitation assay (RIPA) buffer (PC101; Yamay, China) and protease inhibitors (GRF101; Yamay) were prepared at a 1:100 ratio for cell lysis. The lysates were further subjected to ultrasonic lysis for 5 min and centrifugation at 12 000 r∙min–1 at 4 °C for 15 min. A bicinchoninic acid (BCA) protein assay kit (ZJ101; Yamay) was used to determine the total protein concentration, and the proteins were diluted in 5× sodium dodecyl sulfate (SDS) loading buffer and heated at 95 °C for 5 min for denaturation. Each sample with an equal amount of total protein was separated on 10% SDS–polyacrylamide gels and subsequently transferred to polyvinylidene fluoride (PVDF) membranes via a wet transfer apparatus. The membranes were blocked with 5% milk for 2 h at room temperature and then incubated overnight at 4 °C with the following primary antibodies: METTL3, YTHDF1, WTAP (1:2000 dilution, 10200-1-AP; Proteintech), IGF2BP1 (1:5000 dilution, 22803-1-AP; Proteintech), IGF2BP3 (1:1000 dilution, 14642-1-AP; Proteintech), NFATC1 (1:1000 dilution, 32303; SAB, USA), c-Fos (1:1000 dilution, 66590-1-Ig; Proteintech), CTSK (1:500 dilution, sc-48353; Santa Cruz Biotechnology, USA), DC-STAMP (1:1000 dilution, MABF39-I; Millipore, USA), and ATP6V0d2 (1:1000 dilution, SAB2103220; Sigma, USA). The membranes were washed with Tris-buffered saline (TBS)-Tween and incubated with the appropriate secondary antibody. The blots were visualized via a chemiluminescence detection system.
2.12. Total m6A quantification of RNA
Colorimetric assay: Total RNA was isolated from BMMs, and the total m6A content was determined in 200 ng RNA aliquots using an EpiQuik m6A RNA Methylation Quantification Kit (P-9005; EpiGentek, USA) according to the manufacturer’s instructions.
m6A dot blot: Total RNA was extracted with TRIzol (AG21102; Accurate Biology), denatured by heating at 95 °C for 3 min, and then chilled on ice immediately. Next, RNA was spotted on Biodyne Nylon Transfer Membranes (RPN303B; Cytiva, USA) and crosslinked with 254 nm ultrasonic velocity profiler (UVP) for 2 min. The m6A levels were measured by immunoblotting using an anti-m6A antibody (ab284130; Abcam, UK).
2.13. RNA immunoprecipitation (RIP)-qPCR
RIP-qPCR was performed using a RIP Kit (BersinBio, China) according to the manufacturer’s protocol. The cells were lysed in complete RIP lysis buffer to preserve RNA–protein interactions. Anti-METTL3 or anti-YTHDF1 and immunoglobulin G (IgG) antibodies were added to the immunoprecipitation (IP) and IgG groups, respectively, and the vertical homogenizers were incubated overnight at 4 °C, after which equilibrated protein A/G magnetic beads were added to both the IP and IgG groups and incubated at 4 °C for incubation for 1 h. Then, the RNA–protein IP complexes were washed five times. Finally, RNA was extracted via phenol chloroform RNA extraction and purified for qPCR analysis. The relative enrichment was normalized to the input as follows: percentage of input (%Input) = 1/10 × 2Ct[IP] – Ct[input].
2.14. Cell transfection
BMMs were obtained according to the experimental methods described above, and the cells were plated in α-MEM supplemented with 10% fetal bovine serum (FBS) at a concentration of 1 × 106 cells per well in 6-well plates. BMMs were transfected with 20 nmol∙L–1 small interfering RNA (siRNA) using CALNPTM RNAi (D-Nano Therapeutics Co., Ltd., China) according to the manufacturer’s instructions. Briefly, the siRNA, Reagent A, and Reagent B were mixed and incubated at room temperature for 5 min, added to α-MEM, and then added to the cells. The cells were incubated for 24 h for subsequent experimental manipulation and analysis. The siRNA sequences are shown in Table S2 in Appendix A.
2.15. Drug target prediction and molecular docking
Using the online databases PubChem, Swiss Target Prediction, and GeneCards, we identified target protein information for TP and Med. This information was integrated with RA-associated genes (derived from a PCR microarray; results to be published) to analyze network interactions.
Key target proteins were obtained from the Protein Data Bank, and the molecular docking of TP or Med with these key target proteins was performed using AutoDock Vina version 1.1.2.
2.16. Human samples
We obtained a total of three joint samples from RA and osteoarthritis (OA) patients at the China–Japan Friendship Hospital. Each sample was divided into two parts. A portion of the total RNA was extracted with TRIzol according to standard protocols and subsequently subjected to total RNA m6A quantification, whereas the other part was fixed, decalcified, embedded in paraffin and subjected to IHC. Ethical approval for this study was granted by the Ethics Committee of China–Japan Friendship Hospital (ethical approval number: 2024-KY-019).
2.17. Statistical analysis
GraphPad Prism software (USA) was used for statistical analyses and graphing. The experimental data were analyzed with Student’s t test for comparisons between two groups. Comparisons between multiple treatment groups were performed via one-way analysis of variance (ANOVA). All the experiments were repeated three times. The measurement data are presented as the means ± standard errors of the mean (SEMs), and differences were considered statistically significant at P < 0.05.
3. Results
3.1. Single and combined treatment with TP and Med prevented arthritis progression and delayed disease onset in rats with CIA
We used a CIA rat model to simulate RA, with an incidence rate of 95%, confirming the validity of the model. The therapeutic potential and combined effects of TP and Med on RA were investigated in this model. On day 7 after the booster immunization, the rats developed arthritis, which was characterized by swelling of the paws and ankles (
Figs. 2(a) and
(b)). MTX, TP, and Med treatments inhibited the progression of arthritis and the half-dose combination treatment delayed the onset of the disease (
Figs. 2(a)–(c)). The histological analysis of ankle joints revealed extensive cartilage damage and joint destruction in the CIA group, characterized by a hyperplastic invasive synovium with pannus formation and inflammatory cell infiltration (
Figs. 2(d) and
(e)). The pathological features of the MTX group were alleviated. Notably, TP and Med alone and in combination mitigated bone destruction and the associated pathological alterations, and the combination was superior to monotherapy. Furthermore, the combination treatment ameliorated the weight loss associated with RA in rats with CIA (
Fig. 2(f)). In addition, the HE results and organ indices revealed that the drugs did not have toxic effects on the liver, kidney, or colon tissues of rats with CIA (
Figs. 2(g)–(i)).
3.2. Single and combination treatment with TP and Med abolished bone erosion and inhibited osteoclastogenesis
The hyperactivation of OCs is a key contributor to the bone resorption and joint damage observed in patients with RA
[34]. We further evaluated the effects of TP and Med alone and in combination on regulating bone erosion and osteoclastogenesis by conducting micro-CT scans on the ankle and knee joints and further performed TRAP staining on the ankle joints. Micro-CT and 3D analyses revealed that the CIA group presented severe bone erosion and joint destruction in the ankle joints (
Fig. 3(a)), whereas the MTX, TP, Med, and combination treatments significantly alleviated bone erosion, as shown by the reduced BS/BV, and the combination treatment was superior to the monotherapy (
Fig. 3(b)). TRAP staining revealed that increased numbers OCs were predominantly localized in hyperplastic synovium that invade cartilage and joints, contributing to bone erosion (
Fig. 3(c)). MTX, TP, Med, and combination treatments significantly reduced the number and area of OCs, and the results from the combination group were superior to those from the monotherapy group (
Fig. 3(d)). In addition, CIA rats presented an elevated degree of bone resorption and a reduced bone density in the knee joint (
Fig. 3(e)), whereas the MTX, TP, Med, and combination treatments significantly reduced the knee BS/BV values and increased the BV/TV values, and the combination treatment was superior to the single treatment (
Fig. 3(f)). The histological analysis of the knee joint revealed severe cartilage damage and inflammatory cell infiltration in the rats with CIA (
Fig. 3(g)), but these pathological features were alleviated after treatment with MTX, TP, Med, or their combination (
Fig. 3(h)).
3.3. TP and Med regulated the inflammatory response in rats with CIA
Tregs and inflammatory factors are important inducers of osteoclastogenesis and differentiation
[3]. Tregs maintain the dynamic homeostasis of bone tissue by secreting anti-inflammatory cytokines, such as IL-10 and TGF-β, which inhibit the proliferation and differentiation of OC precursors and prevent excessive bone resorption
[32],
[33],
[35]. Second, inflammatory factors such as IL-1β, IL-6, and TNF-α are major promoters of OC differentiation, and in the inflammatory environment of RA, these factors drive the differentiation, maturation, and functional activity of OCs, which are closely related to the occurrence of RA-related bone erosion
[36]. Further evaluation is therefore needed to determine whether the combined effects of TP and Med are reflected in Tregs and inflammatory factors. We examined the ratio of Tregs in the spleen, and the flow cytometry analysis revealed that the proportion of Tregs was reduced in the CIA group (
Fig. 4(a)), whereas MTX and Med could replenish Tregs in the spleen without a statistically significant difference. In addition, TP and the combination treatment could replenish Tregs, but the difference between the monotherapy group and the combination group was not significant (
Fig. 4(b)). The ELISA results revealed that the levels of proinflammatory factors were significantly increased and the levels of anti-inflammatory factors were significantly decreased in the CIA group, whereas MTX, TP, Med, and the combination treatment modulated the inflammatory response. Notably, the combination group was more effective in reducing IL-1β and IL-6 levels compared to the monotherapy (
Figs. 4(c) and
(d)). However, for TNF-α, IL-10, and TGF-β1 expression levels, the difference between the combination group and the monotherapy group was not significant (
Figs. 4(e)–(g)). Therefore, the effect of the combined TP and Med treatment on rats with CIA was reflected in the elimination of articular bone destruction.
3.4. Synergistic inhibition of OC differentiation and function in vitro by the combination of TP and Med
In vivo experiments revealed that the combined effect of TP and Med was reflected in the inhibition of bone erosion in CIA rats; therefore,
in vitro experiments were performed after primary bone marrow mononuclear cells were extracted and differentiated into OCs. The TRAP results revealed that TP or Med alone inhibited RANKL-induced osteoclastogenesis and reduced the number of OCs in a dose-dependent manner (
Fig. 5(a)). Notably, the combination of 10 nmol∙L
–1 TP and 10 μmol∙L
–1 Med had an optimal effect on decreasing the number of OCs (
Fig. 5(b)), with a synergistic combination index (CI) value of less than 1 (
Fig. 5(c)). Compelling evidence for the identification of functional OCs is the detection of bone resorption activity, which can be quantified by staining F-actin rings and examining bone resorption dimple pits. TRITC–phalloidin staining revealed that TP or Med alone inhibited F-actin ring formation in a dose-dependent manner (
Figs. 5(d) and
(e)). The treatment did not show a significant effect on cell survival under the tested conditions (
Figs. 5(f)–(h)). Electron microscopy revealed that TP or Med alone inhibited bone resorption pit formation, whereas the combination had a more significant effect (
Figs. 5(i) and
(j)).
In addition, the mRNA and protein levels of OC specificity-related indicators were examined, and the results revealed that TP and Med decreased the mRNA and protein levels of
Nfatc1,
c-Fos,
Dc-stamp,
Ctsk, and
Atp6v0d2. Notably, the combination group exhibited a stronger inhibitory effect on the transcription of
Nfatc1,
Atp6v0d2,
Dc-stamp, and
Ctsk, as well as on the protein translation of c-Fos, DC-STAMP, and CTSK, compared to the monotherapy group (
Fig. 6). Thus, the
in vitro results also showed that TP and Med exerted inhibitory effects on OC differentiation, bone resorption, and F-actin ring formation.
3.5. TP and Med regulated METTL3/YTHDF1 expression in OCs
We collected knee samples from patients undergoing clinical knee arthroplasty and extracted RNA, which was assayed via colorimetric and dot blot assays. The results revealed that the m
6A content was higher in the RA patients than in the OA patients (
Figs. 7(a) and
(b)). The above results indicate that m
6A methylation is highly upregulated during RA-related bone destruction, which may contribute to the identification of drug targets with good clinical value. The pathogenesis of RA-related bone destruction is associated with RNA m
6A methylation. Moreover, we compared the methylation of PBMCs in the peripheral blood of RA patients and healthy controls using a PCR microarray and detected a notable difference in m
6A methylase levels (article to be published). Additional comparisons of m
6A methylation enzyme transcript levels in joint tissues from RA and OA patients revealed marked variations in
Mettl3,
Ythdf1,
Igf2bp2,
Ythdc1, and
Ythdc2 expression (
Fig. 7(c)). Building on these findings, we further investigated the m
6A methylation targets of TP and Med. By querying online databases for the molecular targets of TP and Med, we analyzed their interactions with the m
6A methylation enzymes that exhibited significant changes in mRNA in the PCR microarrays. This analysis identified strong network associations between the targets of TP and Med and RA-related m
6A methylation enzymes, particularly METTL3, YTHDF1, ALKBH5, and FTO. The genes of these proteins may serve as novel therapeutic targets for TP and Med (
Fig. 7(d)).
During the induction of OC differentiation
in vitro, the total amount of m
6A was elevated, whereas TP or Med alone or in combination inhibited the total amount of m
6A (
Figs. 8(a) and
(b)). Therefore, m
6A is important in RANKL-induced osteoclastogenesis. An analysis of the mRNA levels of the most common m
6A “writers,” “erasers,” and “readers” revealed that RANKL increased the levels of
Mettl3,
Wtap,
Igf2bp1,
Igf2bp3, and
Ythdf1 in the OCs, TP and Med decreased
Mettl3,
Igf2bp3, and
Ythdf1 levels, and the combination treatment decreased
Mettl3,
Wtap,
Igf2bp3, and
Ythdf1 levels. Notably,
Mettl3 and
Ythdf1 expression were inhibited to a greater extent in the combined group than in the TP or Med groups (
Fig. 8(c)). Next, the protein levels of the m
6A methylation-modifying enzymes of interest were examined, and the results revealed that TP and Med inhibited METTL3 and YTHDF1 expression, and a significant difference between the combination treatment group and the groups treated with TP or Med alone was observed, which was almost consistent with the transcript levels (
Figs. 8(d) and
(e)). IHC and TRAP staining were performed after the decalcification of joint tissues from the RA patients. Our results revealed that METTL3 and YTHDF1 expression were significantly increased in TRAP
+ cells (
Figs. 8(f) and
(g)).
In vivo, simultaneous labeling of OCs for METTL3 and YTHDF1 via IHC along with TRAP staining revealed that TP could decrease the level of METTL3 in OCs and that Med could decrease the level of YTHDF1 in OCs (
Figs. 8(h)–(j)). The molecular docking results revealed that TP can interact with METTL3, with a binding energy of –5.90 kcal∙mol
–1 (
Fig. 8(k), and that Med can interact with YTHDF1, with a binding energy of –6.92 kcal∙mol
–1 (
Fig. 8(l)).
3.6. Knockdown of Mettl3 or Ythdf1 inhibits OC differentiation and bone resorption
Next, we transfected si-
Mettl3 or si-
Ythdf1 into BMMs for further study to investigate the relationship between the “writer” METTL3 or “reader” YTHDF1 and osteoclastogenesis. The mRNA and protein expression of
Mettl3 or
Ythdf1 in BMMs was significantly decreased after si-
Mettl3 (Figs. S1(a)–(c) in Appendix A) or si-
Ythdf1 (Figs. S1(d)–(f) in Appendix A) transfection, suggesting that the knockdown was effective. TRAP and F-actin ring staining revealed a decrease in the number and area of OCs after si-
Mettl3 (
Figs. 9(a)–(c)) or si-
Ythdf1 (
Figs. 9(d)–(f)) transfection. In addition, the degree of F-actin fusion was reduced, indicating that the cytoskeletal structure was affected and that OC formation was inhibited. The area of the F-actin ring was reduced, indicating that OCs had a decreased erosive capacity, further supporting the critical roles of METTL3 and YTHDF1 in OC differentiation and bone resorption. Knockdown of
Mettl3 resulted in the loss of further decreases in the OC number, area, and cytoskeletal structure induced by TP; similarly, the knockdown of
Ythdf1 resulted in the loss of further inhibition of OC production and function by Med. RT-qPCR revealed that si-
Mettl3 (
Fig. 9(g)) or si-
Ythdf1 (
Fig. 9(h)) significantly decreased
Nfatc1,
c-Fos,
Dc-stamp,
Atp6v0d2, and
Ctsk mRNA expression, further confirming that knockdown of
Mettl3 or
Ythdf1 affects the expression of OC-associated genes. Western blot analysis further confirmed that the expression levels of OC-associated proteins were significantly decreased after si-
Mettl3 (
Figs. 9(i) and
(j)) or si-
Ythdf1 (
Figs. 9(k) and
(l)) transfection, which was consistent with the RT-qPCR results, suggesting that YTHDF1 and METTL3 affect both the transcription and translation of these genes. The inhibitory effects of TP and Med on OC-associated factors were not further enhanced by the knockdown of
Mettl3 and
Ythdf1. The results of the RIP experiments revealed that METTL3 (
Fig. 9(m)) and YTHDF1 (
Fig. 9(n)) interacted with OC-associated genes (
Nfatc1,
c-Fos,
Dc-stamp,
Atp6v0d2, and
Ctsk), suggesting that these two proteins may regulate their expression by directly binding to these mRNAs. These experimental results indicated that knockdown of
Mettl3 or
Ythdf1 could reduce the number and area of OCs by decreasing the expression of OC-related genes, suggesting that METTL3 and YTHDF1 share a common regulatory network in osteoclastogenesis and OC function and that they may have a cooperative and complementary relationship in the OC regulatory system. Moreover, TP and Med act through METTL3 and YTHDF1, respectively, with similar effects on the inhibition of osteoclastogenesis and OC function, suggesting that these two drugs synergistically regulate osteoclastogenesis and OC function at different levels through their specific m
6A modification pathways.
4. Discussion
Patients with RA experience joint erosion shortly after the onset of the disease, and joint bone destruction is evident in the pre-RA period, even before notable arthritis develops
[37]. Current medical achievements have improved inflammation, but many patients fail to recover from bone destruction
[38]. Targeting OCs is a very promising strategy for the treatment of RA, but the currently available drugs are limited, and several serious side effects have been observed in the clinic
[23]. The ability of the natural components in herbal medicine to inhibit OC formation and function has attracted increasing attention
[39],
[40]. Therefore, the identification of the mechanisms and regulatory programs involved in OC formation is needed. In this study, we observed a synergistic effect of the combination treatment with TP and Med on rats with CIA, and the combination treatment was superior to the single agents in inhibiting bone erosion. Similarly,
in vitro experiments revealed a synergistic effect of TP and Med on OC differentiation.
Despite advances in clinical management, the treatment of RA has always been a challenge. Given the inadequate therapeutic effect and toxic side effects of individual drugs, combination drugs are gradually becoming a clinical trend
[41]. Therefore, the development of combinations with desirable efficacy for the treatment of RA and the avoidance of adverse effects has medical and economic importance
[42],
[43]. TwHF is a well-recognized drug in the guidelines for the treatment of RA, and its main component, TP, has toxic side effects on the liver, kidneys, and reproductive organs
[44],
[45],
[46]. The active compounds in
S. suberectus Dunn have been reported to exert synergistic effects when combined with other drugs
[47]. In this study, we investigated the therapeutic potential of combining low doses of TP and Med for RA treatment to maximize efficacy while minimizing toxicity. Our results showed that the administration of TP (18.62 μg∙kg
–1) and Med (2 mg∙kg
–1) in combination significantly inhibited disease progression in CIA rats, outperforming the individual drugs. Notably, the combination exhibited a remarkable ability to suppress bone erosion and increase OC counts at these low doses, offering advantages over high-dose monotherapies. Strikingly, the efficacy of the combination was comparable to that of the positive control MTX, a first-line drug for RA management. The low-dose combination of TP and Med represents an innovative therapeutic approach that addresses the major limitation of high-dose monotherapy—systemic toxicity—while enhancing the targeted anti-bone destruction effects. This advantage is particularly relevant in clinical scenarios where long-term use of high-dose TP is contraindicated due to its cumulative toxicity. Additionally, the synergistic interaction between TP and Med suggests the possibility of further dose optimization, which could allow for tailored therapies based on patient-specific needs and tolerability.
In recent years, increasing evidence has shown that epigenetic regulation, including DNA methylation and histone modifications, has a profound impact on RA development
[48],
[49]. Our previous study focused on methylation and revealed that m
6A methylation levels are increased in the synovium and PBMCs of RA patients
[14],
[25],
[50]. However, OCs are the core cells involved in bone metabolism and are responsible for bone resorption and remodeling. The role of m
6A, a major RNA epigenetic regulator, in RA-related bone erosion and osteoclastogenesis has been less studied. During osteoclastogenesis, the level of intracellular methylation is increased after RANKL stimulation, which promotes osteoclastogenesis and pathological bone resorption
[17],
[51]. In contrast, the inhibition of methylation suppressed both the number and function of OCs. Our experimental results suggest that the m
6A modification is not only important in the regulation of gene expression but is also a key regulatory mechanism for OC differentiation and functional maintenance
[52]. Therefore, the m
6A mechanism that regulates osteoclastogenesis and activity may provide a new therapeutic target for diseases such as osteoporosis and RA. An analysis of differentially methylated genes in RA and the targets of TP and Med revealed that genes such as
Mettl3,
Ythdf1, and
Alkbh5 may serve as novel drug targets
[53]. We hypothesize that the joint protective effects of TP and Med on joint erosion may be related to varying levels of m
6A methylation. The results showed that TP inhibits METTL3, reducing the m
6A methylation of mRNA, whereas Med blocks the translation of methylated mRNA by inhibiting YTHDF1. Thus, the coadministration of these two drugs can simultaneously interfere with OC production and function through multiple mechanisms of posttranscriptional regulation, leading to significant decreases in OC number, area, and bone resorption capacity. The discovery of this synergistic effect further supports the importance of the m
6A modification pathway in the regulation of osteoclastogenesis and suggests that cotargeting through different regulators can produce similar biological effects. Moreover, the therapeutic effect can be increased by cointerfering with different m
6A regulators at different levels at different stages. These findings provide new possibilities for the treatment of bone destruction in individuals with RA.
Interestingly, TP and Med may target METTL3 and YTHDF1, as si-
Mettl3- and si-
Ythdf1-treated OCs exhibit highly similar phenotypes, suggesting their potential involvement in a shared regulatory pathway. We hypothesize that METTL3 and YTHDF1 form a “methylation–translation” axis during osteoclastogenesis. METTL3 first introduces m
6A modifications to the mRNAs of OC-related genes, and YTHDF1 then binds these modified mRNAs to facilitate their translation
[54],
[55]. This interaction ensures the rapid and efficient expression of critical OC-related genes. Consequently, the knockdown of
Mettl3 reduces m
6A modification levels, whereas the knockdown of
Ythdf1 inhibits the translation of modified mRNAs, ultimately leading to decreased expression of key genes, including
Nfatc1,
c-Fos,
Dc-stamp,
Atp6v0d2, and
Ctsk. NFATc1 is a master regulator of OC differentiation, whereas c-Fos plays a key role in RANKL-induced osteoclastogenesis
[56].
Dc-stamp [57] and
Atp6v0d2 [58] are genes essential for OC fusion and function, whereas
Ctsk encodes a protein responsible for bone matrix degradation
[59]. Based on these findings, we hypothesized that this functionally cooperative and complementary relationship reveals a novel mechanism of m
6A modification in osteoclastogenesis, which also complements the synergistic effects of the combination of TP and Med. Previous studies have shown that METTL3 regulates the area and number of OCs formed
[16]. The attenuation of the m
6A modification via
Mettl3 knockdown and overexpression of ALKBH5 reduced the stability and function of Circ_0008542 and impaired OC activity
[60]. Elevated levels of methylation during osteoblastic differentiation and posttranscriptional regulation of NFATc1 by YTHDF1 and YTHDF2 have antagonistic effects
[17]. YTHDF1 can drive osteolytic cancer metastasis by reading m
6A sites in the
Ezh2 and
Cdh11 mRNAs and promoting their translation
[61]. METTL3 mediates the methylation of the
Icam2 mRNA in synovial fibroblasts, thereby affecting their proliferation, migration, and invasion
[62]. Additionally, METTL3 promotes FLS activation and inflammatory responses
[63]. Studies have shown that METTL3 could serve as a potential therapeutic target for autoimmune diseases, as its deletion in regulatory Tregs leads to impaired suppressive function and severe autoimmune disorders
[64]. Moreover, YTHDF1 may regulate inflammatory responses by influencing the differentiation of Tregs and T helper cells
[65]. Therefore, METTL3 and YTHDF1 may be potential pathways through which TP and Med modulate inflammatory responses in CIA rats. Our results indicate that both METTL3 and YTHDF1 bind to the mRNAs of these genes. Knockdown of
Mettl3 and
Ythdf1 significantly downregulates the expression of these genes and proteins, disrupting the m
6A modification and reducing osteoclastogenesis. This disruption may explain the similar phenotypic effects observed with
Mettl3 or
Ythdf1 knockdown, emphasizing the critical role of the m
6A modification in posttranscriptional regulation during osteoclastogenesis. Furthermore, these findings suggest that METTL3 and YTHDF1 can serve as therapeutic targets to inhibit osteoclastogenesis, thereby preventing RA-associated bone destruction. The combination of TP and Med, which act as synergistic inhibitors of METTL3 and YTHDF1, represents a safe and innovative therapeutic approach with significant potential for clinical translation.
5. Conclusions
OC overactivation is a core pathological feature of RA. This study shows for the first time that TP and Med have combined advantages in the treatment of RA, especially bone destruction. Furthermore, the combination of TP and Med significantly increased the inhibitory effect on osteoclastogenesis and OC function by targeting different key factors in the m6A modification pathway (METTL3 and YTHDF1) and interfering with the expression of OC-related genes through different pathways. Moreover, the m6A modification is a key regulatory mechanism of OC differentiation, and METTL3 and YTHDF1 play important positive regulatory roles in this process. The above evidence provides a new molecular target for RA, a new target for the active ingredients of traditional Chinese medicine, and a theoretical basis for the development of combination drug strategies.
CRediT authorship contribution statement
Yi Jiao: Writing – original draft, Data curation, Conceptualization. Zhaoran Wang: Validation, Software. Wenya Diao: Validation, Software. Qishun Geng: Validation, Software. Xing Wang: Methodology. Xiaoxue Cao: Methodology. Tong Shi: Formal analysis. Jiahe Xu: Formal analysis. Lu Zhao: Visualization, Investigation. Zihan Wang: Visualization, Investigation. Tiantian Deng: Visualization, Investigation. Lei Yang: Resources. Tingting Deng: Supervision. Cheng Xiao: Supervision, Project administration.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This study was supported by the National Natural Science Foundation of China (U22A20374).
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2025.03.014.
PDF
(6033KB)
