The dinitrothiophenol SN-2 reduces osteoclastogenesis and estrogen deficiency-induced bone loss by inhibiting TRAF6 auto-ubiquitination Authors: Jian Lu, Huang suo Wang, Jun Wang, Tao Fan, Jianbin Hu, Baotong Xun, Xiaodong Jia, Haoyu Feng, Haishan Guan, Yongwen Qiao, Feng Gao Published: 2026 ABSTRACT Post-menopausal osteoporosis is characterized by overactive bone resorption in response to estrogen withdrawal. As Receptor Activator of NF-κB ligand (RANKL)-induced auto-ubiquitination of TNF receptor-associated factor 6 (TRAF6) on its K124 residue promotes osteoclastogenesis, inhibiting TRAF6K124 auto-ubiquitination may be a promising therapeutic strategy for post-menopausal osteoporosis. We investigated the impact of the synthetic dinitrothiophenol SN-2 (4-(Dimethylamino)-2,6-bis[[N-[2-[(2-nitrophenyl)dithio]-ethyl]amino]methyl]pyridine) on TRAF6K124 auto-ubiquitination, osteoclast generation, and osteoporotic bone loss. In vitro, SN-2 inhibited RANKL-induced osteoclast differentiation and osteoclast marker expression in a dose-dependent manner. SN-2 also inhibited RANKL-induced F-actin ring formation and bone resorption in vitro. SN-2 suppressed RANKL-induced TRAF6K124 auto-ubiquitination (TRAF6-Ub), TRAF6's binding to Transforming Growth Factor Beta-Activated Kinase 1 (TAK1), and TAK1 phosphorylation, thereby inhibiting downstream NF-κB and c-Jun N-terminal Kinase (JNK) signaling pathways critical to osteoclast differentiation. In an ovariectomized mouse model of osteoporosis, SN-2 ameliorated estrogen-deficiency-induced bone loss, improved femoral biomechanical parameters, and inhibited osteoclastogenesis in vivo. SN-2 may show promise as a therapeutic agent in treating bone disorders characterized by RANKL signaling over-activation and excessive bone resorption. Introduction Osteoporosis, a bone disorder defined by a loss in bone mineral density (BMD), is the most common bone disorder and affects millions of individuals worldwide (Pavone et al., 2017). The World Health Organization (WHO) defines osteoporosis as a BMD of >2.5 standard deviations (SD) below the mean BMD of young adults (Nojiri et al., 2013). Clinically, this manifests as high fracture rates (particularly in the hip, spine, and distal forearm), substantial morbidity, and high medical costs (Pavone et al., 2017). The disease is far more common in post-menopausal women due to estrogen deficiency (Pavone et al., 2017). Non-pharmacological interventions – such as smoking cessation, reduced alcohol consumption, and weight-bearing activity of 30 min per day – have been shown to improve bone health; however, the most efficacious treatments for post-menopausal osteoporosis remain pharmacological (Pavone et al., 2017). Post-menopausal osteoporotic skeletal decline is linked to osteoclastogenesis in response to estrogen deficiency (D'Amelio et al., 2008; Wu et al., 2012). Osteoclasts are large multi-nucleated cells that play a vital role in bone loss and pathogenic remodeling due to their role in bone resorption (Portal Núñez et al., 2012; Loi et al., 2016). Consequently, inhibiting osteoclastogenesis has become a promising therapeutic strategy for post-menopausal osteoporosis. Differentiation of osteoclast precursors into mature osteoclasts requires two key cytokine signals: Macrophage-Colony Stimulating Factor (M-CSF), which promotes precursor survival and proliferation, and Receptor Activator of NF-κB Ligand (RANKL; encoded by the Tnfsf11 gene), which binds to its receptor RANK (encoded by the Tnfrsf11a gene) on the precursor surface to initiate osteoclast differentiation (Park et al., 2017). Current treatments, such as the RANKL antibody denosumab and bisphosphonates, are effective antiresorptives (Cummings et al., 2009; Shoback et al., 2020). However, denosumab can produce serious adverse events upon discontinuation (e.g., rapid bone loss, rebound vertebral fractures) (Tsourdi et al., 2021), while bisphosphonates are limited by poor adherence (Fatoye et al., 2019) and serious adverse events (e.g., mandibular osteonecrosis, atypical fractures) (Borromeo et al., 2015; Edwards et al., 2013). Therefore, improved oral small-molecule therapeutics for post-menopausal osteoporosis are needed. The RANKL-RANK interaction on the surface of osteoclast precursors enables RANK's recruitment of TNF Receptor-Associated Factor 6 (TRAF6), an adaptor protein that promotes osteoclastogenesis in bone marrow precursors (Kadono et al., 2005) and is necessary for functional osteoclast differentiation in vivo (Naito et al., 1999). Mechanistically, RANK's recruitment of TRAF6 enables TRAF6 auto-ubiquitination on the Ub acceptor site residue K124 (Park et al., 2017; Lamothe et al., 2007). This auto-ubiquitinated TRAF6K124 (TRAF6-Ub) complexes with Transforming Growth Factor Beta-Activated Kinase 1 (TAK1) and induces TAK1 phosphorylation, thereby activating downstream NF-κB and c-Jun N-terminal Kinase (JNK) signaling and transcription of the key osteoclastogenic transcription factors c-Fos and Nuclear Factor of Activated T cells cytoplasmic 1 (NFATc1) (Lamothe et al., 2007; Lamothe et al., 2008). As evidenced by the preliminary success of the TRAF6K124 auto-ubiquitination inhibitor C25–140 (Du et al., 2024) and the failure of the CD40-TRAF6 interaction inhibitor 6,877,002 (Marino et al., 2022) in reducing bone loss in osteoporosis mouse models, specific inhibition of TRAF6K124 auto-ubiquitination may be a promising therapeutic strategy for post-menopausal osteoporosis. TRAF6K124 auto-ubiquitination is achieved through TRAF6's RING domain and first zinc finger (ZF1) domain (Lamothe et al., 2008). Notably, the synthetic dithiol SN-1 (4-(Dimethylamino)-2,6-bis[[(2-mercaptoethyl)amino]-methyl]pyridine) (Fujita et al., 1996) has been shown to bind to TRAF6's ZF1 domain, thereby downregulating TRAF6K124 auto-ubiquitination and downstream NF-κB activation (Koga et al., 2017). However, SN-1 can be easily oxidized as it has two exposed thiol groups (Radwan et al., 2019). To overcome this limitation, Radwan et al. utilized the SN-1 dinitrothiophenol derivative SN-2 (4-(Dimethylamino)-2,6-bis[[N-[2-[(2-nitrophenyl)dithio]-ethyl]amino]methyl]pyridine) (Fujita et al., 1996), which more effectively inhibits downstream NF-κB activation with no evidence of cytotoxicity (Radwan et al., 2019). Despite SN-2's encouraging in vitro efficacy and safety profile, the impact of SN-2 on TRAF6K124 auto-ubiquitination, osteoclast generation, and osteoporotic bone loss remains completely unknown. Consequently, here we used a combination of in vitro and in vivo experiments to address this question. We found that SN-2 reduced RANKL-induced osteoclastogenesis and osteoclast function in vitro. Mechanistically, SN-2 suppressed RANKL-induced TRAF6K124 auto-ubiquitination, TRAF6-TAK1 binding, and TAK1 phosphorylation, thereby inhibiting downstream signaling pathways critical to osteoclast differentiation. In vivo, SN-2 therapy ameliorated ovariectomy (OVX)-induced bone loss. Therefore, our work suggests that SN-2 reduces bone resorption and osteoclastogenesis via suppression of TRAF6K124 auto-ubiquitination. Materials and methods Ethics statement All animal experiments were approved in advance by the Animal Care and Use Committee of General Hospital of TISCO, Shanxi Medical University (Taiyuan, China). All animal experiments were conducted in compliance with the ARRIVE guidelines, the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 8023, revised 1978), and the Basic & Clinical Pharmacology & Toxicology policy for experimental and clinical studies (Tveden-Nyborg et al., 2021). Female C57BL/6 mice (5–7 weeks old) were obtained from the SLAC Laboratory (China) and were habituated to laboratory conditions (3–4 mice per cage at room temperature provided standard chow and tap water ad libitum, 12-h/12-h light/dark cycle) for one week prior to experimentation. Isolation of bone marrow-derived macrophages and osteoclast differentiation Murine bone marrow-derived macrophages (BMDMs) were generated by isolating monocyte precursors from the long bones of C57/BL6 mice (6–8 weeks old) and differentiating them into BMDMs through the use of Eagle a-modified Minimum Essential Media (a-MEM) media (Thermo Fisher Scientific) supplemented in M-CSF (30 ng/ml, PerproTech) for 3 to 4 days. Cell cultures were kept in 5% CO2 at 37 °C and the media was refreshed every 2 days. SN-2 was custom synthesized by GL Biochem Ltd. (Shanghai, China) as previously reported (Fujita et al., 1996; Radwan et al., 2019). The impact of SN-2 on BMDM viability was determined through the use of CCK-8 assays (Dojindo Molecular Technologies, China). In brief, cells were plated into a 96-well plate (8× 103/well) and incubated for 24 h before being treated with SN-2 for a period of 5 days across a broad range of concentrations (0 nM [DMSO vehicle], 1 nM, 10 nM, 100 nM, 1 μM, 2.5 μM, 5 μM, 10 μM, 20 μM, and 100 μM). Next, CCK-8 buffer (10 μl) was added to each well and left to incubate for 4 h before the optical density was measured (450 nm) using an ELX808 absorbance microplate reader (BioTex). To analyze the effects of SN-2 on osteoclast differentiation, BMDMs were plated into a 96-well plate (8× 103/well) and incubated for 5 days in osteoclastogenic media (i.e., a-MEM media supplemented in M-CSF [30 ng/ml] and RANKL [100 ng/ml, PerproTech]) in the presence of DMSO vehicle or SN-2 at the indicated concentrations. Finally, the BMDM-derived osteoclasts were paraformaldehyde (4%) fixed for 15 min, washed with PBS twice and stained with the TRAP staining kit as per the manufacturer's protocol (TRAP staining kit, MilliporeSigma). Cells were considered mature TRAP+ osteoclasts if they: (i) displayed TRAP+ staining and (ii) contained three or more nuclei. ImageJ was used to quantify the number and spread area of mature TRAP+ osteoclasts in each well. Immunofluorescence staining To assess the impact of SN-2 on the formation of F-actin rings, BMDMs were cultured in osteoclastogenic media on bone discs for 5 days. The BMDM-derived osteoclasts were then incubated for a further 48 h with DMSO vehicle (MilliporeSigma) or SN-2 (10 μM). Cells were paraformaldehyde (4%) fixed at room temperature for 15 min before being permeabilized with Triton X-100 (0.1%) in PBS for 5 min. Staining of F-actin was achieved with rhodamine-phalloidin stain (Cytoskeleton) diluted in bovine serum albumin (2%) in PBS for 1 h. A 5-min DAPI incubation was used to stain cell nuclei and samples were mounted onto glass slides using ProLong Gold media (Thermo Fisher Scientific). The Nikon A1Si confocal microscope system was used to acquire fluorescence images. ImageJ was used to quantify the number and size of F-actin. Bone resorption assays To assess the impact of SN-2 on bone resorption, BMDMs were cultured in osteoclastogenic media for 5 days. The BMDM-derived osteoclasts were then incubated overnight on bone discs. Following adherence, cells were incubated in osteoclastogenic media with DMSO vehicle or SN-2 (10 μM) for 2 days. Next, discs were treated with NaOH (2 M) and adhered cells were gently removed by brushing. The Hitachi S-3700 N electron microscope (Japan) was used to visualize each disc. ImageJ was used to quantify resorbed areas. In silico ligand-protein docking analysis In silico ligand-protein docking analysis was performed using the three-dimensional chemical structure of SN-2 (4-(Dimethylamino)-2,6-bis[[N-[2-[(2-nitrophenyl)dithio]-ethyl]amino]methyl]pyridine) obtained from the ZINC database (https://zinc.docking.org) using the unique identifier ZINC143559666 in conjunction with the X-ray diffraction-derived crystal structure of the human TRAF6-RANK protein complex accessed from the Protein Data Bank (PDB) website (https://www.rcsb.org) under the unique identifier 1LB5. The web-based docking software algorithm SwissDock (https://www.swissdock.ch/) was employed to calculate the Gibbs free energy difference (ΔG) values for all ligand-protein docking configurations. UCSF Chimera v15 was then used to visualize the most probable ligand-protein docking configuration. Generation of TRAF6-knockout macrophages with retroviral TRAF6 overexpression TRAF6-knockout (Traf6−/−) RAW264.7 macrophages and Traf6−/− BMDMs were generated via CRISPR/Cas9 employing two single-guide RNA (sgRNA) sequences targeting murine Traf6 (sgRNA1: 5’-GATGGAACTGAGACATCTCG-3′; sgRNA2: 5’-GGAGATCCAGGGCTACGATG-3′) (Tan and Kagan, 2019). For WT TRAF6 overexpression, retroviral vectors were engineered on a pMX vector backbone containing the full-length murine Traf6 cDNA sequence (NM_009424; #MR208489, Origene) cloned upstream from an IRES-GFP element with a puromycin-selectable marker as previously described (Lamothe et al., 2007). For mutant TRAF6K124R overexpression, site-directed mutagenesis of the nucleotides coding for the K124 residue (K124 → R124) on the murine Traf6 cDNA element was conducted using a QuikChange kit (Stratagene) and verified by DNA sequencing. The empty pMX vector (i.e., containing no cDNA element) was used as a negative control. Retroviral particle production and infections were performed as previously described (Lamothe et al., 2007). Briefly, GP2–293 cells were co-transfected with the aforedescribed retroviral pMX vectors and pVSV-g (Clontech) via FuGENE (Roche Applied Science). Following overnight transfection, the medium was replaced with fresh medium. After 72 h, the viral particle-containing supernatants were collected and used to infect the Traf6−/− RAW264.7 macrophages and Traf6−/− BMDMs via polybrene (8 μg/ml). After 2 days, selection was initiated with puromycin (5 μg/ml). Then, selected cells were single-cell plated onto 96-well plates until colony formation at three weeks. CRISPR-based knock-out and retroviral overexpression were validated by immunoblotting. Quantitative real-time PCR (qPCR) The RNAiso reagent (Takara) was used to isolate total RNA. RNA (1 μg) was transcribed into cDNA (20 μl volume; Takara) before being subjected to qPCR using SYBR Premix Ex Taq II on an ABI StepOnePlus (Thermo Fisher Scientific). Primer sequences are provided in Supp. Table S1. PCR cycling conditions were as follows: 30 s at 95 °C, followed by 40 cycles of: 5 s 95 °C then 30 s at 60 °C. β-actin was used as a housekeeping gene. The 2-ΔΔCt method was used to calculate relative gene expression. Co-immunoprecipitation assays Protein co-immunoprecipitation assays were conducted as previously described (Bai et al., 2005). In brief, RAW264.7 macrophages were plated into 6-cm dishes and incubated with DMSO vehicle or SN-2 (10 μM) for 6 h. Then, macrophages were exposed to DMSO vehicle or RANKL (100 ng/ml) for 20 min, then incubated with Nonidet P-40 buffer supplemented with protease inhibitors. Lysates were centrifuged for 10 min at 12,000 rpm at 4 °C. Supernatants were incubated with an anti-TRAF6 antibody (#sc-8409, Santa Cruz Biotechnology) and protein A/G agarose beads for 4 h at 4 °C. Following incubation, protein levels for ubiquitin (Ub) and TAK1 were respectively measured with an anti-ubiquitin antibody (#3933, Cell Signaling Technology) or anti-TAK1 antibody (#sc-7967, Santa Cruz Biotechnology) via Western blotting as described below. In vitro ubiquitination assays In vitro ubiquitination assays were conducted as previously described with minor modifications (Jang et al., 2019; Chan et al., 2012). Briefly, murine Myc-DDK-tagged TRAF6 (#MR208489, Origene) or murine Myc-DDK-tagged RANK/TNFRSF11A (#MR209538, Origene) were individually expressed in 293 T cells, subjected to anti-Myc antibody immunoprecipitation (#2278, Cell Signaling Technology), and eluted using protein A/G agarose beads. Myc-DDK-TRAF6 (2 μg) was pre-incubated with Myc-DDK-RANK (0 or 2 μg) and SN-2 (0, 5, 10, or 20 μM) in a 20-μl volume of 10× reaction buffer [300 mM HEPES (pH 7.2), 50 mM MgCl2, 20 mM ATP, and 2 mM DTT] for 30 min at room temperature. The ubiquitination assays were then initiated by adding ubiquitin (59 μM), E1 (50 nM), and E2 (Ubc13/Uev1A, 850 nM) into this 10× reaction buffer. Reactions were incubated for 60 min at 30 °C, ceased by adding sample buffer, and immediately heated at 95 °C for 5 min. Proteins were immunoprecipitated with an anti-TRAF6 antibody (#sc-8409, Santa Cruz Biotechnology) for 1 h at 4 °C, and TRAF6 ubiquitination was measured via Western blotting using an anti-ubiquitin antibody (#3933, Cell Signaling Technology) as described below. Western blotting To assess the impact of SN-2 on numerous signaling pathways, RAW264.7 macrophages were plated into 6 well plates (8 × 105/well) in a-MEM for 24 h. Cells were then treated with DMSO vehicle or SN-2 (10 μM) and incubated for 6 h prior to RANKL stimulation (50 ng/ml) for 0, −5, −10, −20, −30 and − 60 min. To assess the impact of SN-2 on osteoclast differentiation, BMDMs were plated into 6-well plates (24 × 104/well) in osteoclastogenic media supplemented with either DMSO or SN-2 (10 μM) for 0, −1 or − 3 days. Protein was extracted using RIPA buffer supplemented with phosphatase and protease inhibitors. In brief, cells were incubated in RIPA for 20 min on ice, centrifuged for 10 min at 12,000 rpm and 4 °C before cell supernatants were collected. Protein (30 μg) were loaded into SDS-PAGE gels (10%, MilliporeSigma) and separated via electrophoresis. Protein was then heat transferred onto PVDF membranes before being blocked at room temperature with bovine serum albumin (5%) and Tween-20 (0.1%) in Tris-buffer saline for 1 h. Next, membranes were incubated overnight at 4 °C with the following primary antibodies against: ubiquitin (#3933), phospho-IKKα/β (Ser176/180) (#2697), phospho-NF-κB p65 (Ser536) (#3031), NF-κB p65 (#4764), and β-actin (#8457) from Cell Signaling Technology; TAK1 (#sc-7967), TRAF6 (#sc-8409), RANKL (#sc-377079), RANK (#sc-374360), IKKα/β (#sc-7607), phospho-JNK (Thr183/Tyr185) (#sc-6254), and JNK (#sc-7345) from Santa Cruz Biotechnology; and phospho-TAK1 (Thr184/Thr187) (#MA5–15073), phospho-MKK7 (Ser271/Thr275) (#MA5–28042), and MKK7 (#PA5–79627) from Invitrogen. After washing membranes with Tween-20 (0.1%) in Tris-buffer saline, membranes were incubated for 2 h at 4 °C with species-appropriate horseradish peroxidase (HRP)-conjugated secondary antibody (Cell Signaling Technology). Finally, protein bands were detected using chemiluminescence reagents (MilliporeSigma) on the Bio-Rad XRS chemiluminescence system. Murine model of OVX-induced bone loss A murine model of OVX-induced bone loss was constructed as previously described (Wu et al., 2012). Mice (6–8 weeks old) were randomly allocated into four groups by a random number generator: Sham control, OVX + vehicle (negative control), OVX + zoledronic acid (ZA, MilliporeSigma), and OVX + SN-2. For the OVX mice, at one-week post-OVX, DMSO vehicle or SN-2 (40 mg/kg) were injected intraperitoneally (i.p.) every 2 days, while ZA was subcutaneously injected once a week. Body weights were monitored over the 6-week post-OVX period. All mice were humanely euthanized at 8 weeks post-OVX. There were no observable side effects or deaths during the 8-week study period. Humerus bones were collected in order to assess markers of bone formation via qPCR. The side tibia of each mouse was obtained for micro-computed tomography (micro-CT). Bone mineral density (BMD), bone surface/bone volume (BS/BV), bone volume/tissue volume (BV/TV), cortical thickness (Ct.Th), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), and structure model index (SMI) of the spongiosa were calculated to assess tibia microstructure. All samples were paraformaldehyde (4%) fixed for 1 day, then decalcified in EDTA (10% pH 7.4) for 3 weeks. Following decalcification, samples were paraffin-embedded and sections of the proximal femur, distal femur and proximal tibia were collected for histomorphometric analysis. For biomechanical analysis, each mouse had a random side femur removed and tested for a 3-point bend as previously described (Cardelli et al., 2015). In brief, frozen femurs were thawed at room temperature before a bending force perpendicular to the long axis was applied to the femur (1 mm/min) and then stopped once it dropped to half the maximum load. Data for the maximum bend stress was collected in accordance with the loading displacement curve. Statistical analysis All statistical analyses were performed using SPSS software. Data are represented as means ± standard deviations (SDs) from six independent biological replicates (in vitro experiments) and nine mice per cohort (in vivo experiments). Prior to hypothesis testing, distributional assumptions were assessed for each group using the Shapiro-Wilk test for normality and Levene's test for homogeneity of variances. For comparisons between two groups, Welch's t-test was used when both groups satisfied the normality assumption, and the Mann-Whitney U test was used when one or both groups violated normality. For comparisons among three or more groups, the choice of omnibus test depended on the assumption checks. When all groups were normally distributed and variances were homogeneous, a one-way ANOVA was performed followed by Tukey's Honestly Significant Difference (HSD) post-hoc test. When all groups were normally distributed but variances were unequal (Levene's test P < 0.05), Welch's ANOVA was performed followed by the Games-Howell post-hoc test. When one or more groups violated the normality assumption, the Kruskal-Wallis H test was performed followed by Dunn's test with Bonferroni correction. For experiments with two experimental factors (e.g., treatment group and time point or treatment group and genotype), a two-way ANOVA was first conducted to evaluate main effects and interactions. When a significant interaction was detected, per-condition pairwise comparisons were then performed using the appropriate two-group test (Welch's t-test or Mann-Whitney U test) as determined by the assumption checks described above. Statistical significance was set at P < 0.05 for all analyses. Results SN-2 inhibits osteoclastogenesis and mature osteoclast function in vitro Following treatment with vehicle control (VCtrl) or varying concentrations of SN-2 (i.e., 1 nM, 10 nM, 100 nM, 1 μM, 2.5 μM, 5 μM, 10 μM, 20 μM, and 100 μM) for 5 days, BMDMs showed no significant alterations in cell viability (Fig. 1A). To assess the impact of SN-2 on osteoclast generation, BMDMs were incubated with RANKL (100 ng/ml) and M-CSF (30 ng/ml) in the presence of SN-2 (0, 2.5, 5, and 10 μM). The results showed that differentiation of osteoclasts was inhibited in a dose-dependent manner by SN-2, confirmed by the reduction in TRAP-positive cells (Fig. 1B-D). Osteoclast precursors began to fuse roughly 3 days following stimulation with RANKL, whereas SN-2 (10 μM) completely inhibited osteoclast formation (Fig. 1E-G). To investigate the stage at which osteoclastogenesis was predominantly affected by SN-2, BMDMs were treated with SN-2 (10 μM) during either the early stages (day 1 to 3), or late stages (day 3 to 5), or both (day 1 to 5) of RANKL stimulation. SN-2 dramatically reduced the number of osteoclasts during early-stage osteoclastogenesis but had minimal impact on late-stage osteoclastogenesis (Fig. 1H-J).Fig. 1SN-2 inhibits osteoclastogenesis and mature osteoclast function in vitro(A) The viability of murine bone marrow-derived macrophages (BMDMs) determined via CCK-8 assays following treatment with the indicated concentrations of SN-2 for 5 days. (B—D) BMDMs cultured in M-CSF (30 ng/ml) and RANKL (100 ng/ml) with VCtrl or SN-2 at the indicated concentrations for 5 days prior to TRAP staining to enable quantification of mature TRAP+ osteoclast counts and spread area. Scale bar = 200 μm. (E-G) BMDMs were cultured in M-CSF (30 ng/ml) and RANKL (100 ng/ml) with VCtrl or SN-2 (10 μM) prior to TRAP staining and quantification of mature TRAP+ osteoclast counts and spread area at day 1, −3 and − 5. Scale bar = 200 μm. (H-J) BMDMs cultured with M-CSF (30 ng/ml) and RANKL (100 ng/ml) with VCtrl or SN-2 (10 μM) at either early-stage (days 1 to 3), late-stage (days 3 to 5) or both (days 1 to 5). Mature TRAP+ osteoclast counts and spread area were quantified through TRAP staining on day 5. Scale bar = 200 μm. (K) BMDMs were differentiated in M-CSF (30 ng/ml) and RANKL (100 ng/ml) for 5 days, followed by VCtrl or SN-2 (10 μM) treatment for 48 h prior to cell staining and confocal microscopy analysis, and (L) the size of F-actin rings was quantified. (M) BMDM-derived osteoclasts were incubated overnight on bone discs in the presence of M-CSF (30 ng/ml) and RANKL (100 ng/ml), followed by VCtrl or SN-2 (10 μM) treatment for 48 h prior to electron microscope analysis, and (N) resorption was quantified. Scale bar = 50 μm. Data reported as means ± standard deviations (SDs). n = 6 independent biological replicates. Statistical tests: (A) one-way ANOVA; (C, D) Kruskal-Wallis H test with Dunn's post-hoc (Bonferroni); (F, G) two-way ANOVA with per-condition Mann-Whitney U test or Welch's t-test; (I, J) one-way ANOVA with Tukey's HSD; (L,N) Welch's t-test.Fig. 1 SN-2 inhibits osteoclastogenesis and mature osteoclast function in vitro (A) The viability of murine bone marrow-derived macrophages (BMDMs) determined via CCK-8 assays following treatment with the indicated concentrations of SN-2 for 5 days. (B—D) BMDMs cultured in M-CSF (30 ng/ml) and RANKL (100 ng/ml) with VCtrl or SN-2 at the indicated concentrations for 5 days prior to TRAP staining to enable quantification of mature TRAP+ osteoclast counts and spread area. Scale bar = 200 μm. (E-G) BMDMs were cultured in M-CSF (30 ng/ml) and RANKL (100 ng/ml) with VCtrl or SN-2 (10 μM) prior to TRAP staining and quantification of mature TRAP+ osteoclast counts and spread area at day 1, −3 and − 5. Scale bar = 200 μm. (H-J) BMDMs cultured with M-CSF (30 ng/ml) and RANKL (100 ng/ml) with VCtrl or SN-2 (10 μM) at either early-stage (days 1 to 3), late-stage (days 3 to 5) or both (days 1 to 5). Mature TRAP+ osteoclast counts and spread area were quantified through TRAP staining on day 5. Scale bar = 200 μm. (K) BMDMs were differentiated in M-CSF (30 ng/ml) and RANKL (100 ng/ml) for 5 days, followed by VCtrl or SN-2 (10 μM) treatment for 48 h prior to cell staining and confocal microscopy analysis, and (L) the size of F-actin rings was quantified. (M) BMDM-derived osteoclasts were incubated overnight on bone discs in the presence of M-CSF (30 ng/ml) and RANKL (100 ng/ml), followed by VCtrl or SN-2 (10 μM) treatment for 48 h prior to electron microscope analysis, and (N) resorption was quantified. Scale bar = 50 μm. Data reported as means ± standard deviations (SDs). n = 6 independent biological replicates. Statistical tests: (A) one-way ANOVA; (C, D) Kruskal-Wallis H test with Dunn's post-hoc (Bonferroni); (F, G) two-way ANOVA with per-condition Mann-Whitney U test or Welch's t-test; (I, J) one-way ANOVA with Tukey's HSD; (L,N) Welch's t-test. Given SN-2's differential impact on early-stage vs. late-stage osteoclastogenesis, we examined if mature osteoclast function was impaired by SN-2. In mature osteoclasts, we assessed the impact of SN-2 on F-actin ring formation, which is crucial for osteoclast adherence (Zalli et al., 2016). Under untreated conditions, F-actin rings formed on top of bone discs (Fig. 1K, L). However, under SN-2 conditions, mature osteoclast size was significantly reduced (Fig. 1K, L). Moreover, SN-2 reduced the number of resorption pits, which was consistent with morphological alterations and cytoskeletal retraction (Fig. 1M, N). These data indicate that SN-2 attenuates osteoclastogenesis and mature osteoclast function in vitro. SN-2 downregulates osteoclast differentiation markers in vitro To validate the impact of SN-2 on the molecular markers of osteoclast differentiation, BMDMs were incubated with RANKL (100 ng/ml) and M-CSF (30 ng/ml) in the presence of SN-2 (0, 2.5, 5, and 10 μM), and gene expression levels of key osteoclast markers were measured via qPCR on day 5. SN-2 downregulated gene expression levels of the transcription factors c-Fos (Fos) and NFATc1 (Nfatc1) (Fig. 2A, B) as well as the NFATc1 target genes calcitonin receptor (Calcr), cathepsin K (Ctsk), Sarcoma virus oncogene (Src), dendritic cell-specific transmembrane protein (Dcstamp), matrix metalloproteinase-9 (Mmp9), and tartrate-resistant acid phosphatase (Acp5) in a dose-dependent manner (Fig. 2C-H).Fig. 2SN-2 downregulates osteoclast differentiation markers in vitro(A-H) Murine bone marrow-derived macrophages (BMDMs) cultured in M-CSF (30 ng/ml) and RANKL (100 ng/ml) with vehicle control (VCtrl) or SN-2 (2.5, 5, or 10 μM) for 5 days prior to RNA isolation and qPCR analysis of the indicated osteoclast differentiation markers. (I—P) BMDMs cultured in M-CSF (30 ng/ml) and RANKL (100 ng/ml) with VCtrl or SN-2 (10 μM) for 0, −1, −3, and − 5 days prior to RNA isolation and qPCR analysis of the indicated osteoclast differentiation markers. Data reported as means ± standard deviations (SDs). n = 6 independent biological replicates. Statistical tests: (A, E, H) Welch's ANOVA with Games-Howell post-hoc; (B, C, D, F, G) one-way ANOVA with Tukey's HSD; (I—P) two-way ANOVA with per-condition Welch's t-test.Fig. 2 SN-2 downregulates osteoclast differentiation markers in vitro (A-H) Murine bone marrow-derived macrophages (BMDMs) cultured in M-CSF (30 ng/ml) and RANKL (100 ng/ml) with vehicle control (VCtrl) or SN-2 (2.5, 5, or 10 μM) for 5 days prior to RNA isolation and qPCR analysis of the indicated osteoclast differentiation markers. (I—P) BMDMs cultured in M-CSF (30 ng/ml) and RANKL (100 ng/ml) with VCtrl or SN-2 (10 μM) for 0, −1, −3, and − 5 days prior to RNA isolation and qPCR analysis of the indicated osteoclast differentiation markers. Data reported as means ± standard deviations (SDs). n = 6 independent biological replicates. Statistical tests: (A, E, H) Welch's ANOVA with Games-Howell post-hoc; (B, C, D, F, G) one-way ANOVA with Tukey's HSD; (I—P) two-way ANOVA with per-condition Welch's t-test. To investigate SN-2's effects on the molecular markers of osteoclast differentiation over time, BMDMs were treated with SN-2 (10 μM) during days 1 to 5 of RANKL stimulation. SN-2 downregulated gene expression levels of c-Fos and NFATc1 (Fig. 2I, J) as well as the NFATc1 target genes (Calcr, Ctsk, Src, Dcstamp, Mmp9, Acp5) in a time-dependent manner (Fig. 2K-P). These data indicate that SN-2 downregulates the molecular markers of osteoclast differentiation in a dose-dependent and time-dependent manner in vitro. SN-2 downregulates RANKL-induced TRAF6 auto-ubiquitination, TRAF6-TAK1 binding, and TAK1 phosphorylation in vitro As SN-2's parent compound SN-1 has been shown to inhibit TRAF6K124 auto-ubiquitination by binding to TRAF6's ZF1 domain (Koga et al., 2017), we investigated whether SN-2 also inhibits TRAF6K124 auto-ubiquitination. First, we applied in silico docking analysis to model SN-2's binding to the TRAF6 protein. The docking model revealed that SN-2's nitrothiophenol sidechain may preferentially bind to the Cys349 and Thr382 residues within TRAF6's RING domain (Fig. 3A, B), which is necessary for TRAF6K124 auto-ubiquitination (Lamothe et al., 2008). In vitro, RAW264.7 macrophages were treated with RANKL and exposed to vehicle or SN-2. Protein lysates were immunoprecipitated with an anti-TRAF6 antibody, followed by immunoblotting analysis for Ub to detect TRAF6-Ub. RANKL stimulation increased TRAF6K124 auto-ubiquitination (Fig. 3C), which was abrogated by SN-2. Moreover, SN-2 displayed a dose-dependent effect in abrogating RANKL-induced TRAF6K124 auto-ubiquitination (Supp. Fig. S1). Upon TRAF6-Ub's binding to TAK1, the TRAF6 pUb chain triggers TAK1's phosphorylation at Thr184 and Thr187 (p-TAK1T184/187), enabling TAK1 to phosphorylate and activate downstream signaling cascades (Huang et al., 2025). Accordingly, RANKL stimulation increased TRAF6-TAK1 binding and TAK1T184/187 phosphorylation, which was abrogated by SN-2 (Fig. 3D, E).Fig. 3SN-2 downregulates RANKL-induced TRAF6 auto-ubiquitination, TRAF6-TAK1 binding, and TAK1 phosphorylation in vitro(A) SN-2's chemical structure displaying its 4-dimethylaminopyridine core and two nitrothiophenol sidechains. (B)In silico docking model illustrating SN-2's binding to the Cys349 and Thr382 residues within TRAF6's RING domain (estimated ΔG = −8.41 kcal/mol). (C-E) Wild-type (WT) RAW264.7 macrophages were plated into 6-well plates and treated with vehicle control (VCtrl) or SN-2 (10 μM) for 6 h prior to a 20-min RANKL stimulation (100 ng/ml). (C, D) Co-immunoprecipitation (IP)/immunoblotting (IB) assays for (C) TRAF6 auto-ubiquitination using an IP-grade anti-ubiquitin (Ub) antibody and (D) TRAF6-TAK1 binding using an IP-grade anti-TRAF6 antibody. (E) Western blotting assay for TAK1 phosphorylation. Data reported as means ± standard deviations (SDs). n = 6 independent biological replicates. Statistical tests: (C-E) one-way ANOVA with Tukey's HSD.Fig. 3 SN-2 downregulates RANKL-induced TRAF6 auto-ubiquitination, TRAF6-TAK1 binding, and TAK1 phosphorylation in vitro (A) SN-2's chemical structure displaying its 4-dimethylaminopyridine core and two nitrothiophenol sidechains. (B)In silico docking model illustrating SN-2's binding to the Cys349 and Thr382 residues within TRAF6's RING domain (estimated ΔG = −8.41 kcal/mol). (C-E) Wild-type (WT) RAW264.7 macrophages were plated into 6-well plates and treated with vehicle control (VCtrl) or SN-2 (10 μM) for 6 h prior to a 20-min RANKL stimulation (100 ng/ml). (C, D) Co-immunoprecipitation (IP)/immunoblotting (IB) assays for (C) TRAF6 auto-ubiquitination using an IP-grade anti-ubiquitin (Ub) antibody and (D) TRAF6-TAK1 binding using an IP-grade anti-TRAF6 antibody. (E) Western blotting assay for TAK1 phosphorylation. Data reported as means ± standard deviations (SDs). n = 6 independent biological replicates. Statistical tests: (C-E) one-way ANOVA with Tukey's HSD. Due to the important roles of the upstream RANKL/RANK axis in stimulating TRAF6K124 auto-ubiquitination (Lamothe et al., 2008), we also examined if SN-2 affects RANKL or RANK expression in M-CSF-stimulated BMDMs. Notably, neither RANKL expression (Supp. Fig. S2A, B) nor RANK expression (Supp. Fig. S2C, D) was affected by SN-2. This combined evidence reveals that SN-2 inhibits TRAF6K124 auto-ubiquitination, TRAF6-TAK1 binding, and TAK1 phosphorylation without impacting upstream expression of RANKL or RANK. SN-2 downregulates RANKL-induced NF-κB and JNK signaling activation in a TRAF6K124-dependent manner As TRAF6K124 auto-ubiquitination is necessary for stimulating NF-κB and JNK signaling activity via TAK1 (Lamothe et al., 2007; Lamothe et al., 2008), we hypothesized that SN-2 would inhibit RANKL-induced NF-κB and JNK signaling activity in a TRAF6K124-dependent manner. Therefore, RANKL-treated WT and Traf6−/− RAW264.7 macrophages transduced with either empty pMX control, WT TRAF6, or mutant TRAF6K124R overexpression vectors (Supp. Fig. S3A) were treated with vehicle or SN-2 and then NF-κB and JNK signaling activity were analyzed by Western blotting. In WT macrophages and Traf6−/− macrophages transduced with WT TRAF6, IKK and NF-κB p65 phosphorylation were inhibited following SN-2 treatment (Fig. 4A). Moreover, phosphorylation of MKK7 and JNK were inhibited by SN-2 treatment (Fig. 4B). Notably, SN2's inhibitory effects were not observed in TRAF6K124R-overexpressing macrophages. Collectively, these data indicate that SN-2 inhibits RANKL-induced NF-κB and JNK signaling in a TRAF6K124-dependent manner.Fig. 4SN-2 downregulates RANKL-induced downstream signaling activation in a TRAF6K124-dependent manner(A, B) WT RAW264.7 macrophages and Traf6−/− RAW264.7 macrophages infected with either empty control, WT TRAF6, or mutant TRAF6K124R overexpression vectors were treated with either vehicle or SN-2 (10 μM) for 6 h prior to RANKL (100 ng/ml) stimulation for 0, −30, or − 60 min. Protein lysates were subjected to immunoblotting and densitometric quantification of (C) IKKα/β and NF-κB p65 phosphorylation levels and (D) MKK7 and JNK phosphorylation levels. (E-J) Murine bone marrow-derived macrophages (BMDMs) cultured with M-CSF (30 ng/ml), RANKL (100 ng/ml), with vehicle or SN-2 (10 μM) for 0, −1 or − 3 days prior to RNA isolation and qPCR analysis of the indicated osteoclast differentiation markers. Data reported as means ± standard deviations (SDs). n = 6 independent biological replicates. Statistical tests: (A, D, F, G, I, J) two-way ANOVA with per-condition Welch's t-test; (B) two-way ANOVA with per-condition Welch's t-test or Mann-Whitney U test; (C, E, H) two-way ANOVA with per-condition Mann-Whitney U test.Fig. 4 SN-2 downregulates RANKL-induced downstream signaling activation in a TRAF6K124-dependent manner (A, B) WT RAW264.7 macrophages and Traf6−/− RAW264.7 macrophages infected with either empty control, WT TRAF6, or mutant TRAF6K124R overexpression vectors were treated with either vehicle or SN-2 (10 μM) for 6 h prior to RANKL (100 ng/ml) stimulation for 0, −30, or − 60 min. Protein lysates were subjected to immunoblotting and densitometric quantification of (C) IKKα/β and NF-κB p65 phosphorylation levels and (D) MKK7 and JNK phosphorylation levels. (E-J) Murine bone marrow-derived macrophages (BMDMs) cultured with M-CSF (30 ng/ml), RANKL (100 ng/ml), with vehicle or SN-2 (10 μM) for 0, −1 or − 3 days prior to RNA isolation and qPCR analysis of the indicated osteoclast differentiation markers. Data reported as means ± standard deviations (SDs). n = 6 independent biological replicates. Statistical tests: (A, D, F, G, I, J) two-way ANOVA with per-condition Welch's t-test; (B) two-way ANOVA with per-condition Welch's t-test or Mann-Whitney U test; (C, E, H) two-way ANOVA with per-condition Mann-Whitney U test. SN-2 downregulates RANKL-induced c-Fos and NFATc1 induction in a TRAF6K124-dependent manner The NF-κB and JNK signaling pathways induce transcription of Fos (encoding the osteoclastic transcription factor c-Fos), while NF-κB and c-Fos co-stimulate transcription of Nfatc1 (encoding the osteoclastic master regulon NFATc1) (Park et al., 2017). Therefore, WT BMDMs and Traf6−/− BMDMs transduced with either empty pMX control, WT TRAF6, or mutant TRAF6K124R overexpression vectors (Supp. Fig. S3B) were incubated with RANKL (100 ng/ml) and M-CSF (30 ng/ml) in the presence of SN-2 (10 μM), and gene expression levels of key osteoclast markers were measured via qPCR on day 5. In WT BMDMs and Traf6−/− BMDMs transduced with WT TRAF6, SN-2 downregulated Fos mRNA expression (Fig. 4C) and Nfatc1 mRNA levels (Fig. 4D). Moreover, SN-2 downregulated the NFATc1 target genes Calcr, Ctsk, Src, Dcstamp, Mmp9, and Acp5 (Fig. 4E-J). Notably, SN2's inhibitory effects were not observed in TRAF6K124R-overexpressing macrophages. Collectively, these data indicate that SN-2 inhibits RANKL-induced c-Fos and NFATc1 induction in a TRAF6K124-dependent manner. SN-2 ameliorates OVX-induced bone loss and inhibits osteoclastogenesis in vivo To assess SN-2's therapeutic effects in vivo, we examined the impact of SN-2 on bone loss and osteoclastogenesis in a mouse model of estrogen-deficient osteoporosis. Treatment with zoledronic acid (ZA) was used as a positive control (Shoback et al., 2020). Body weights did not differ significantly among the four cohorts over the 6-week post-OVX period (Supp. Fig. S4A). As expected, uterine weights were significantly lower in all three OVX cohorts relative to sham controls, confirming successful ovariectomy, while neither ZA nor SN-2 prevented OVX-induced uterine atrophy (Supp. Fig. S4B). Micro-CT analysis of the tibia spongiosa revealed that OVX induced pronounced deterioration of bone microarchitecture relative to sham controls across all measured parameters (Fig. 5A–I). SN-2 treatment significantly ameliorated this OVX-induced bone loss. Relative to OVX + VCtrl mice, SN-2-treated mice exhibited significantly higher BMD (838.6 ± 15.0 vs. 743.7 ± 22.6 mg/cc, P < 0.0001; Fig. 5B), BV/TV (0.040 ± 0.004 vs. 0.025 ± 0.003, P < 0.0001; Fig. 5D), cortical thickness (115.1 ± 11.2 vs. 66.0 ± 17.6 μm, P < 0.0001; Fig. 5E), trabecular thickness (0.050 ± 0.006 vs. 0.041 ± 0.005 mm, P = 0.017; Fig. 5F), and trabecular number (2.08 ± 0.18 vs. 1.46 ± 0.16 mm−1, P < 0.0001; Fig. 5G). SN-2 also significantly reduced trabecular separation (0.387 ± 0.028 vs. 0.511 ± 0.043 mm, P < 0.0001; Fig. 5H), structure model index (2.76 ± 0.08 vs. 3.07 ± 0.07, P < 0.0001; Fig. 5I), and BS/BV (64.1 ± 4.4 vs. 71.6 ± 5.0 mm−1, P = 0.012; Fig. 5C). Notably, BMD, BV/TV, Ct.Th, Tb.Th, and SMI values in SN-2-treated mice did not differ significantly from sham controls, indicating substantial preservation of bone architecture. ZA-treated mice showed a similar pattern of protection across all parameters (Fig. 5B–I). Histological analysis corroborated the micro-CT findings: hematoxylin and eosin staining of the proximal tibia, proximal femur, and distal femur demonstrated that both SN-2 and ZA preserved trabecular bone volume relative to OVX + VCtrl mice (Fig. 5J).Fig. 5SN-2 ameliorates ovariectomy (OVX)-induced bone loss and inhibits osteoclastogenesis in vivo(A) Representative micro-CT images of the four mouse cohorts (i.e., sham control, ovariectomy + vehicle control (OVX + VCtrl), ovariectomy + zoledronic acid (OVX + ZA), and ovariectomy + SN-2 (OVX + SN-2)) at eight weeks after surgery. (B—I) Micro-CT analyses of (B) bone mineral density (BMD), (C) bone surface/bone volume (BS/BV), (D) bone volume/tissue volume (BV/TV), (E) cortical thickness (Ct.Th), (F) trabecular thickness (Tb.Th), (G) trabecular number (Tb.N), (H) trabecular separation (Tb.Sp), and (I) structure model index (SMI) of the tibia spongiosa in the four mouse cohorts. Scale bar = 500 μm. (J) Hematoxylin and eosin staining of the proximal tibia, proximal femur, and distal femur in the four mouse cohorts. Scale bar = 500 μm. Representative TRAP staining images are also shown. Scale bar = 50 μm. (K) Loading-displacement curve calculated from the biomechanical assessment in the four mouse cohorts. (L) The maximum loading force in the four mouse cohorts during the biomechanical assessment. (M) Osteoclast surface per BS quantification in the four mouse cohorts. Data reported as means ± standard deviations (SDs) or medians ± interquartile ranges (IQRs). n = 9 mice per cohort. Statistical tests: (B, C, E, F, H, L, M) one-way ANOVA with Tukey's HSD; (D, G, I) Welch's ANOVA with Games-Howell post-hoc.Fig. 5 SN-2 ameliorates ovariectomy (OVX)-induced bone loss and inhibits osteoclastogenesis in vivo (A) Representative micro-CT images of the four mouse cohorts (i.e., sham control, ovariectomy + vehicle control (OVX + VCtrl), ovariectomy + zoledronic acid (OVX + ZA), and ovariectomy + SN-2 (OVX + SN-2)) at eight weeks after surgery. (B—I) Micro-CT analyses of (B) bone mineral density (BMD), (C) bone surface/bone volume (BS/BV), (D) bone volume/tissue volume (BV/TV), (E) cortical thickness (Ct.Th), (F) trabecular thickness (Tb.Th), (G) trabecular number (Tb.N), (H) trabecular separation (Tb.Sp), and (I) structure model index (SMI) of the tibia spongiosa in the four mouse cohorts. Scale bar = 500 μm. (J) Hematoxylin and eosin staining of the proximal tibia, proximal femur, and distal femur in the four mouse cohorts. Scale bar = 500 μm. Representative TRAP staining images are also shown. Scale bar = 50 μm. (K) Loading-displacement curve calculated from the biomechanical assessment in the four mouse cohorts. (L) The maximum loading force in the four mouse cohorts during the biomechanical assessment. (M) Osteoclast surface per BS quantification in the four mouse cohorts. Data reported as means ± standard deviations (SDs) or medians ± interquartile ranges (IQRs). n = 9 mice per cohort. Statistical tests: (B, C, E, F, H, L, M) one-way ANOVA with Tukey's HSD; (D, G, I) Welch's ANOVA with Games-Howell post-hoc. To assess the functional consequences of these microarchitectural changes, femoral three-point bending tests were performed. SN-2 treatment significantly improved femoral strength relative to OVX + VCtrl mice (9.66 ± 0.53 vs. 7.01 ± 0.45 N, P < 0.0001; Fig. 5K, L), restoring it to a level not significantly different from ZA-treated mice. However, SN-2-treated mice still exhibited a modest but statistically significant reduction in maximum loading force relative to sham controls (P = 0.033). Finally, to assess SN-2's effects on in vivo osteoclastogenesis, TRAP staining was performed to quantify osteoclast surface per bone surface (OC·S/BS) (Lee et al., 2019). SN-2 treatment significantly reduced OC·S/BS relative to OVX + VCtrl mice (0.20 ± 0.04 vs. 0.30 ± 0.04%, P < 0.0001; Fig. 5M), though OC·S/BS in SN-2-treated mice remained modestly elevated compared with ZA-treated mice (P = 0.0043). This decrease in osteoclast surface was consistent with the preserved bone volume observed in SN-2-treated mice. Overall, SN-2 ameliorates OVX-induced bone loss and inhibits osteoclastogenesis in vivo. Discussion RANKL signaling over-activation and excessive bone resorption are known to play a role in post-menopausal osteoporosis as well as other numerous bone disorders such as rheumatoid arthritis, Paget disease, and periodontitis (Hu et al., 2019). Consequently, RANKL signaling pathways have become an attractive target for therapeutic interventions to treat bone-related disorders (Hu et al., 2019). Therefore, here we examined the impact of the synthetic dinitrothiophenol SN-2 on RANKL-mediated osteoclastogenesis and osteoclast function. Furthermore, we assessed the impact of SN-2 on OVX-induced bone loss in a murine model of post-menopausal osteoporosis. Collectively, our data highlight SN-2 as a potential novel therapeutic candidate for treating post-menopausal osteoporosis. RANKL signaling pathways are important for differentiating these osteoclast precursors into mature osteoclasts (Park et al., 2017). Binding of RANKL to RANK results in RANK's recruitment of TRAF6, TRAF6K124 auto-ubiquitination, TRAF6-Ub complexing with TAK1 that induces TAK1 phosphorylation and TAK1-mediated activation of the kinases IKK and MKK7, and the respective downstream activation of pro-osteoclastogenic NF-κB and JNK signaling (Park et al., 2017; Lamothe et al., 2007; Huang et al., 2025). Our study revealed that SN-2 inhibits RANKL-induced TRAF6K124 auto-ubiquitination, TRAF6-TAK1 binding, and TAK1 phosphorylation, thereby inhibiting NF-κB and JNK signaling in a TRAF6K124-dependent manner. As NF-κB and JNK signaling co-stimulate transactivation of the osteoclastogenic transcription factor Fos (Park et al., 2017), we found that SN-2 also downregulated Fos expression in a TRAF6K124-dependent manner. The NF-κB and c-Fos transcription factors bind to the promoter of the osteoclast master regulon Nfatc1 to induce NFATc1 auto-amplification (Patel et al., 2015). NFATc1 promotes the transcription of several key genes associated with osteoclast differentiation, including Calcr, Ctsk, Mmp9, and Acp5 (Hansen et al., 2024; Charles et al., 2012). Consistent with this model, we showed that expression of Nfatc1 and NFATc1 target genes were suppressed by SN-2 in a TRAF6K124-dependent manner. Our findings are consistent with previous research showing downregulation of NF-κB and JNK signaling as well as c-Fos and NFATc1 expression through inhibiting TRAF6 (Hu et al., 2019). Consistent with our work, Lamothe et al. found that mutating the major Ub acceptor lysine residue necessary for TRAF6 auto-ubiquitination K124 abolishes TRAF6-mediated TAK1 activation, NF-κB activation, and spontaneous osteoclast differentiation (Lamothe et al., 2007; Lamothe et al., 2008). However, in direct contrast to Lamothe et al., Walsh et al. showed that TRAF6 auto-ubiquitination is dispensable for TAK1 activation, NF-κB activation, and osteoclastogenesis using a total lysine-deficient TRAF6 mutant (i.e., TRAF6 K32-518R) (Walsh et al., 2008). It should be noted that Walsh et al. retrovirally overexpressed the lysine-deficient TRAF6 mutants, yielding mutant TRAF6 expression levels that were much higher than physiological TRAF6 levels (Walsh et al., 2008). Moreover, as Walsh et al. concedes, the total lysine-deficient TRAF6 mutation may have abrogated TRAF6's ability to undergo other lysine-dependent post-translational modifications (PTMs) (Walsh et al., 2008), such as acetylation, methylation, sumoylation, or K48 ubiquitination (which marks TRAF6 for proteasomal degradation (Takayanagi et al., 2000)). As TRAF6 overexpression can promote unregulated activation of downstream signaling pathways during osteoclastogenesis (Franzoso et al., 1997; Ishida et al., 1996), the aberrantly high mutant TRAF6 expression levels or the absence of regulatory lysine-dependent PTMs on mutant TRAF6 may have driven Walsh et al.'s conclusion that TRAF6 auto-ubiquitination is dispensable for osteoclastogenesis. In addition, Walsh et al.'s work was performed in primary murine bone marrow-derived macrophages (BMMs), while Lamothe et al.'s work was performed in the murine macrophage RAW264.7 cell line and spleen-derived monocytes. Thus, the use of differing cell lines between these two studies may have also contributed to this discrepancy in findings. Further research is needed to better elucidate this issue. In the clinic, a common presentation for patients with post-menopausal osteoporosis is fragility fractures. Bisphosphonates are effective in preventing fragility fractures (Ghirardi et al., 2014) but suffer from poor adherence (Fatoye et al., 2019) and serious adverse events (e.g., mandibular osteonecrosis, atypical fractures) (Borromeo et al., 2015; Edwards et al., 2013). Therefore, novel candidate therapies with a superior safety profile are still needed (Li et al., 2011). Here, we used a well-established OVX-induced mouse model of osteoporosis to mimic human post-menopausal osteoporosis. In support of our in vitro findings, SN-2 treatment significantly ameliorated OVX-induced bone loss and inhibited osteoclastogenesis in vivo. Specifically, SN-2 restored BMD, BV/TV, Ct.Th, Tb.Th, Tb.N, and SMI to levels that did not differ significantly from sham controls, while significantly reducing OVX-induced increases in Tb.Sp and BS/BV. SN-2 also improved femoral biomechanical strength, with maximum loading force recovering to 9.66 ± 0.53 N compared with 7.01 ± 0.45 N in OVX + VCtrl mice (P < 0.0001). Consistent with these structural and functional improvements, TRAP staining revealed that SN-2 significantly reduced osteoclast surface coverage (OC·S/BS) relative to OVX + VCtrl mice (P < 0.0001). Notably, SN-2's effects on bone microarchitecture and biomechanical strength were broadly comparable to those of the positive control ZA. Our findings are consistent with previous research showing improvements in OVX-induced bone loss in mice through inhibiting TRAF6 (Hu et al., 2019). It should be noted that we observed no evidence of SN-2-induced cytotoxicity in vitro or in murine bone tissue in vivo. While global Traf6 knockout produces low-sweating ectodermal dysplasia, osteoporosis, and early death in mice, compounds that inhibit TRAF6K124 auto-ubiquitination, such as C25–140, do not show strong adverse effects (Brenke et al., 2018). Moreover, although downstream NF-κB inhibition can produce side effects such as inflammatory tissue damage, T/B-cell immunity dysregulation, and increased hepatic and skin cancer risk, these side effects have been largely attributed to JNK hyperactivation in NF-κB-suppressed cells (Liu et al., 2014). Therefore, it is possible that SN-2's dual inhibition of NF-κB and JNK signaling may reduce or eliminate any potential adverse effects arising from NF-κB inhibition. Future studies should investigate the impact (if any) of SN-2 therapy on non-bone tissues. In conclusion, the synthetic dinitrothiophenol SN-2 inhibits osteoclast differentiation and function in vitro and in vivo. Mechanistically, SN-2 suppresses RANKL-induced TRAF6K124 auto-ubiquitination and TAK1 activation, thereby inhibiting downstream signaling pathways critical to osteoclast differentiation. On this basis, SN-2 may show promise as a therapeutic agent in treating bone disorders characterized by RANKL signaling over-activation and excessive bone resorption. CRediT authorship contribution statement Jian Lu: Writing – review & editing, Writing – original draft, Conceptualization. Huang suo Wang: Writing – review & editing, Writing – original draft, Validation, Investigation, Formal analysis. Jun Wang: Supervision, Methodology, Conceptualization. Tao Fan: Validation, Investigation. Jianbin Hu: Validation, Investigation. Baotong Xun: Investigation. Xiaodong Jia: Visualization, Formal analysis. Haoyu Feng: Formal analysis. Haishan Guan: Formal analysis. Yongwen Qiao: Writing – review & editing, Writing – original draft. Feng Gao: Conceptualization. Ethics statement All animal experiments were approved in advance by the Animal Care and Use Committee of General Hospital of TISCO, Shanxi Medical University (Taiyuan, China). All animal experiments were conducted in compliance with the ARRIVE guidelines, the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 8023, revised 1978), and the Basic & Clinical Pharmacology & Toxicology policy for experimental and clinical studies. 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.