Clematichinenoside AR was isolated from the roots of C. manshurica Rupr. as described previously (24), and the purity (>98%) was determined by high performance liquid chromatography (HPLC) analysis. The structure of C-AR was showed in Figure S1 in Data Sheet in Supplementary Material.

Complete Freund’s adjuvant (CFA) (#7001), incomplete Freund’s adjuvant (IFA) (#7002), and bovine type II collagen (#20021) were obtained from Chondrex, Inc. (Redmond, WA, USA). Recombinant Human TGF-β1 (PeproTech Inc., USA) and murine IL-1β (PeproTech Inc., USA) were dissolved in citric acid as a stock solution and then further diluted with PBS containing 5% trehalose before use. Digoxin and PX-478 were obtained from NRCCRM (Beijing, China) and Taizhou Rico Biological Technology Co., Ltd. (Taizhou, China), respectively. These agents were dissolved in dimethyl sulfoxide (DMSO) as a stock solution and the final working concentration of DMSO was <0.1% (v/v). Dimethyl malonate (V900555), dimethyl succinate (V900547), and aminooxyacetate (AOA, C13408) were provided by Sigma (St. Louis, MO, USA). The following items were purchased from the cited commercial sources: anti-IL-1β antibody (MAB201) (R&D, USA); anti-NLRP3 (NBP2-12446) and Novus Biologicals (Littleton, CO, USA); anti-α-SMA (ab32575), anti-cleaved caspase-1 (ab1872), anti-HIF-1α (ab51608), anti-TGF-β1 (ab64715), Goat Anti-Rabbit IgG H&L (Alexa Fluor 488) (ab181448), and Abcam (Cambridge, MA, USA); anti-GAPDH (AP0063), Goat Anti-Rabbit IgG (H+L) HRP (BS13278), and Bioworld Technology (St. Paul, MN, USA).

Female Wistar rats (130–140 g, n = 50) were supplied by the Laboratory Animal Center of Yangzhou University (Yangzhou, China). Rats were housed under standard laboratory condition of constant temperature (23 ± 1°C) and 55 ± 5% relative humidity with a 12-h alternate light/dark cycle, fed with standard diet and tap water. Animal welfare and all experimental protocols were approved by the Animal Ethics Committee of China Pharmaceutical University.

To prepare CIA in the rat, on day 0, bovine type II collagen was emulsified in CFA (1:1, v/v) and then administrated intradermally into the base of rat tail, as the primary immunization. On day 7, rats were given a booster injection of bovine type II collagen emulsified in IFA (1:1, v/v), as the secondary immunization. After that, rats were administrated with C-AR (50 mg/kg, by gavage) or digoxin (100 μg/kg, by intraperitoneal injection) once a day for 2 weeks. The choice of dosage value for C-AR was based on the published reports (22, 23). Normal and model rats were orally given an equal volume of distilled water in parallel.

The clinical symptoms were recorded every 3 days since the secondary immunization, with arthritis index scores according to the following criteria: score 0, no swelling or inflammation; score 1, slight swelling and/or redness of paw; score 2, swelling in two or more joints; score 3, pronounced swelling with more than two joints; and score 4, severe swelling with joint rigidity. Meanwhile, paw swelling and body weight were recorded during the course, of which paw swelling was expressed as the volume of increase with respect to day 0 volume.

Rat synovial fibroblasts were obtained as previously described (25). In brief, the isolated synovium was minced and digested with 4 mg/ml collagenase II (Sigma, USA) for 2 h at 37°C and then cultured in RPMI Medium 1640 until the cells outgrowing from the explants. Passages 3–6 of the synovial fibroblasts were used for the experiments.

To specifically suppress HIF-1α expression, NIH-3T3 cells (a cell line of embryo fibroblasts) were grown to 60% confluence and then transfected with small interfering RNA (siRNA) duplexes specific for HIF-1α (sc-35562, Santa Cruz Biotechnology) or control siRNA (sc-37007) by siRNA transfection reagent (sc-29528). After transfection, cells were cultured in Dulbecco’s Minimum Essential Medium (DMEM) for 24–48 h before experiments.

Total RNA from rat synovium was extracted with TRIzol Reagent (Beyotime Biotechnology Corporation, China), and cDNA was synthesized by an All-In-One-RT MasterMix synthesis kit (ABM, Canada), following the manufacturer’s instructions. cDNA was amplified by quantitative real-time PCR with the SYBR Green I PCR kit (BIO-RAD, USA) and the CFX96™ Realtime System (BIO-RAD, USA). Primer sequences are listed in Table S1 in Data Sheet in Supplementary Material. The mRNA level of individual genes was normalized to β-actin and calculated by 2^−ΔΔCt data analysis method (26).

Synovial tissue of RA was lysed in ice-cold RIPA buffer with Tissue Lyser, and then centrifuged at 13,000 g for 15 min at 4°C. The supernatant was collected. The contents of lactate and succinate and the activity of succinate dehydrogenase (SDH) in the supernatant were assayed with the commercial Kits, while IL-1β was measured with ELISA Kit (Neobioscience Technology Co., China). For the assay in cultured cells, confluent synovial fibroblasts or transfected NIH-3T3 cells were pretreated with indicated agents and then stimulated with TGF-β1 (10 ng/ml) or dimethyl succinate (5 mM) for 48 or 8 h. After incubation, lactate and IL-1β in the supernatant and succinate and SDH activity in the lysed cells were assayed as mentioned above.

For protein analysis, synovial fibroblasts or transfected NIH-3T3 cells were lysed in ice-cold RIPA buffer and incubated in an ice bath for 45 min. Protein was obtained by centrifugation at 12,000 g for 20 min at 4°C and quantified with Bicinchoninic Acid Protein Assay Kit (Beyotime Biotechnology Corporation, China). Equal amount of protein was separated on polyacrylamide SDS gels and transferred to polyvinylidene difluoride (PVDF) membranes by electrophoresis, following incubation with the specific primary antibodies and HRP-conjugated secondary antibody. The values of band intensities were detected by enhanced chemiluminescence (ECL) and densitometry was analyzed quantitatively by Image-Pro Plus 6.0 (IPP 6.0) software.

Synovial tissue were fixed with 4% (w/v) paraformaldehyde, embedded in paraffin, and then sectioned at 5-μm thickness. The sections were dewaxed using xylene and dehydrated in a gradient of alcohol. For the histological evaluation, the tissue slices were stained with hematoxylin and eosin (H&E) or Masson’s trichrome staining, according to standard protocols. For immunohistochemistry analysis, the sections were subjected to 3% hydrogen peroxide to quench the endogenous peroxidase activity, and then incubated with specific primary antibodies and HRP-conjugated secondary antibodies. Images were observed and captured with an optical microscope (Olympus, Japan).

To investigate synovial hypoxia, rats were injected with hypoxyprobe (pimonidazole HCL) (Hypoxyprobe™-1 Plus Kit, Burlington, VT, USA) at a dosage of 60 mg/kg for 45 min prior to sacrifice. Immunoperoxidase staining for hypoxyprobe binding in a formalin-fixed paraffin embedded synovium section using 1:100 dilution of FITC-MAb1 and 1:100 dilution of horseradish peroxidase conjugated rabbit anti-FITC IgG. The image was observed by inverted fluorescence microscope (Nikon ECLIPSE Ti-s).

For cellular hypoxia, synovial fibroblasts were detected after adding hypoxyprobe (20 μM) and treated with C-AR or PX-478, then, exposed to hypoxia (1% O₂, 4 h) or TGF-β (10 ng/ml, 48 h). For immunofluorescence assay, synovial fibroblasts were cultured on coverslips, pretreated with C-AR, and then stimulated with TGF-β1 (10 ng/ml, 48 h) or dimethyl succinate (5 mM, 8 h). After incubation, synovial fibroblasts were fixed with 4% paraformaldehyde in PBS for 15 min, and then blocked with 5% BSA containing 0.1% Triton X-100 for 60 min. Following incubation with primary antibodies and secondary antibodies, cells were visualized under a confocal scanning microscope (Zeiss LSM 700).

All of the experimental results are expressed as the mean ± SD (standard deviation), and each experiment was performed a minimum of four times. Data were analyzed by two-tailed t-test, or one-way ANOVA test with Student–Newman–Keuls test for comparison of two groups. A value of p < 0.05 was considered to be statistically significant.

Collagen challenge induced arthritis in rats, as evidenced by red swelling in the paw and functional impairment (Figures 1A,B). Administration of C-AR (50 mg/kg) and HIF-1α inhibitor digoxin (100 μg/kg) reduced swollen ratio and ameliorated the clinical symptoms indicated the reduced arthritis index scores (Figures 1A,B). Histological examination showed that C-AR and digoxin treatment ameliorated synovial hyperplasia and congestion with a reduction in the infiltration of inflammatory cells in the synovial tissue (Figure 1C). In addition, C-AR treatment improved body weight gain in the course of arthritis (Figure S2 in Data Sheet in Supplementary Material). These results well demonstrated the beneficial effects of C-AR on the amelioration of arthritis in rats.

Meanwhile, we examined fibrosis in the synovial tissues. Immunohistochemistry staining showed TGF-β1 and α-SMA overexpression in the synovial tissue, indicating fibroblast activation, whereas these alternations were prevented by C-AR and digoxin treatment (Figures 1D,E). Masson trichrome stain sections showed an attenuated fibrosis in C-AR or digoxin-treated CIA rats (Figure 1F). Coincidentally, C-AR treatment normalized gene expressions for Col1α, Col3α, and LOX in synovial tissue (Figure 1G). These results showed that C-AR prevented fibrosis in synovial tissue. Digoxin, a HIF-1α inhibitor, effectively reduced protein or gene expressions of TGF-β1, α-SMA, and ECM, indicative of the potential involvement of HIF-1α.

Accompanied with fibrosis, hypoxia also occurred in synovial tissue, indicated by increased staining of hypoxia adducts with pimonidazole (Figure 2A). C-AR and digoxin treatment attenuated hypoxia and then inhibited HIF-1α protein accumulation in the synovium of CIA rats (Figures 2A,B). Meanwhile, HIF-1α gene expression was also reduced in the synovium (Figure S3 in Data Sheet in Supplementary Material). Similarly, C-AR treatment effectively attenuated hypoxia and reduced HIF-1α accumulation at concentrations ranging from 0.1 to 10 μM in synovial fibroblasts exposed to 1% O₂ (Figures 2C,D). HIF-1α inhibitor PX-478 showed a similar regulatory tendency as C-AR. TGF-β1 stimulation increased HIF-1α induction with increased pimonidazole staining in synovial fibroblasts, indicating that TGF-β1-induced HIF-1α induction in a manner of low oxygen tension-dependent (Figures 2E,F). C-AR as well as PX-478 prevented TGF-β1-induced low oxygen tension in fibroblasts, and thereby inhibited HIF-1α induction with reduced lactate accumulation (Figure S3 in Data Sheet in Supplementary Material). Meanwhile, C-AR treatment reduced lactate accumulation in the synovial tissue and fibroblasts (Figure S4 in Data Sheet in Supplementary Material). These results in vivo and in vitro demonstrated that C-AR reduced HIF-1α accumulation by preventing low oxygen tension.

Accompanied with local hypoxia, succinate accumulated in the synovium of CIA rats (Figure 3A). C-AR and digoxin treatment reduced succinate accumulation with inhibition of SDH activity (Figures 3A,B). Consistent with the regulation in vivo, C-AR treatment inhibited SDH activity and reduced succinate accumulation in a concentration-dependent manner in synovial fibroblasts when exposed to TGF-β1 treatment (Figures 3C,D). In CAC, SDH converts succinate to fumarate, inhibition of SDH should increase succinate accumulation. However, we found that SDH inhibitor dimethyl malonate inhibited TGF-β1-induced succinate production in fibroblasts. In view of the increased SDH activity, this result suggested that SDH might act in reverse for succinate formation. As an inhibitor of malate/aspartate shuttle (MAS), AOA reduced succinate accumulation with inhibited SDH activity, suggesting that MAS pathway might provide substrate for the reversal succinate production (Figure 3E).

We observed increased IL-1β production accompanied with NLRP3 and cleaved caspase-1 overexpression in synovial tissue (Figures 4A–C). C-AR and digoxin administration attenuated NLRP3 and cleaved caspase-1 expression, and then reduced IL-1β production (Figures 4A–C), which was indicative of the role in suppressing NLRP3 inflammasome activation. As succinate is documented to promote IL-1β in macrophages, we investigated the possible implication of succinate in synovial IL-1β production. As shown in Figure 4D, SDH inhibitor dimethyl malonate inhibited TGF-β1-induced IL-1β production, indicating the involvement of succinate. Dimethyl succinate is a cell permeable derivative of succinate readily taken up by cells, where it is then amplified thereby increasing succinate level. Similar to TGF-β1, succinate also increased IL-1β production, despite less extent than TGF-β1 (Figure 4D). In line with the action, we observed succinate enhanced NLRP3 induction in a concentration-dependent way (Figure 4E). These results suggested that TGF-β1-induced NLRP3 inflammasome activation through succinate, at least in part. HIF-1α inhibitor PX-478 diminished the enhanced effect of succinate on IL-1β production and NLRP3 expression, suggesting the potential involvement of HIF-1α in NLRP3 inflammasome activation (Figures 4D,E).

In synovial fibroblasts, succinate increased HIF-1α production (Figure 5A). HIF-1α was distributed in the cytosol, and in response to succinate stimulation, HIF-1α accumulation was accompanied by its translocation to the nucleus (Figure 5B). To directly address the functional interaction between HIF-1α and NLRP3 inflammasome activation, we investigated NLRP3 expression and IL-1β secretion in NIH-3T3 cells, a cell line of an embryo fibroblast, by transfecting them with siRNA of HIF-1α. Knockdown of HIF-1α diminished the enhancing effects of succinate on NLRP3 expression and IL-1β secretion (Figures 5C,D), indicating that HIF-1α induction was required for succinate-associated NLRP3 inflammasome activation.

To know the contribution of succinate and NLRP3 inflammasome activation to synovial fibroblasts activation, we observed TGF-β1 and α-SMA protein expression in fibroblasts exposed to dimethyl succinate. Succinate challenge increased TGF-β1 and α-SMA expression, whereas these increased expressions were reduced by neutralizing IL-1β with anti-IL-1β antibody treatment (Figures 6A,B), suggesting the potential role of IL-1β in succinate-mediated fibroblasts activation. For further confirmation, we observed the expression of TGF-β1 and α-SMA in fibroblasts to IL-1β stimulation. As shown in Figures 6C,D, incubation of fibroblasts with IL-1β resulted in TGF-β1 and α-SMA proteins accumulation, and these alternations were reversed by pretreatment with C-AR. Together with above-mentioned results, they indicated that succinate worked as a metabolic signaling, amplifying TGF-β1 action in fibroblasts through NLRP3 inflammasome activation. C-AR reduced succinate accumulation and thereby blocked the forward cycle toward sustained fibroblast activation (Figure 6E).