Rutoside decreases human macrophage-derived inflammatory mediators and improves clinical signs in adjuvant-induced arthritis

For in vitro experiments, 3,3',4',5,7-pentahydroxyflavone-3-rutinoside trihydrate (RU) (>97% purity powder; Acros Organics, Noisy-le-grand, France) was used after suspension in distilled water. For in vivo subcutaneous (s.c.) injections in rats, RU was suspended in saline.

Peripheral blood samples were obtained from healthy volunteers with their informed consent after the approval of this study by the institutional ethics committee. These samples were pretested for the absence of HIV or hepatitis virus infections. Peripheral blood-derived mononuclear leukocytes (PBLs) were obtained by Ficoll gradient separation, and monocytes were subsequently separated from other leukocytes by adherence to CD14 beads (Miltenyi Biotec, Paris, France). CD14^- PBLs were used for the hematotoxicity test of various RU preparations. Briefly, cells were incubated in medium alone or with 10^-6 M phytohemagglutinin-P (PHA) (5 μg/mL; Murex Biotech Ltd, Dartford, UK) in order to induce lymphocyte proliferation. RU was added at different concentrations, cells were harvested 4 days later, and apoptotic/necrotic versus total cells were counted as indicated below. CD14⁺ cells were then suspended in RPMI 1640 medium supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin, 2 mM glutamine, and 10% fetal calf serum (FCS) (all from Gibco-Europe, Paisley, Scotland). The above culture medium, chemicals, and FCS were endotoxin-free and tested for the absence of direct activation effects on human monocytes (CD23 expression and TNF-α production as activation markers). After these procedures, more than 95% of cells expressed CD14 antigen and displayed cytochemical characteristics of monocytes/macrophages [22].

Female Lewis rats (Janvier, Le Genest St Isle, France) were housed under standard laboratory conditions with free access to food and water. The temperature was kept at 22°C ± 2°C, and a 12-hour light/dark schedule was maintained. The Animal Research Committee of the Agriculture Ministry approved this investigation. All animal procedures were performed in strict accordance with the guidelines issued by the European Economic Community (directive 86/609). Adjuvant-induced arthritis (AIA) was obtained in 6-week-old animals by s.c. injection at the base of the tail of 300 μL (1.8 mg) of inactivated Mycobacterium butyricum (Difco Laboratories Inc., now part of Becton Dickinson and Company, Franklin Lakes, NJ, USA) diluted in an emulsion of 8 mL of Vaseline oil, 1 mL of polysorbate 80, and 1 mL of phosphate-buffered saline (PBS) (Laboratoires Eurobio, Courtaboeuf, France). Rats were boosted 1 week later with the same dose of antigen and observed for up to 50 days after immunization for clinical signs of chronic arthritis. Evaluation of AIA severity was performed by two independent observers with no knowledge of the treatment protocol. The severity of AIA in each paw was quantified daily by an arbitrary clinical score measurement from 0 to 2 as follows: no signs of inflammation (0), swelling alone (0.5) for each paw, immobility (0.5) for each paw, 2 being the highest score with both paws swelling and immobile. Weight evolution of the animals was measured daily. Rats were treated with five injections (every 2 days) of RU, saline, or hydrocortisone (HC) (Sigma-Aldrich, Saint Quentin Fallavier, France) as indicated in the Results section. Injections were initiated 1 day after the appearance of the first arthritic symptoms (therapeutic) or on the first day of immunization (preventive). For macrophage collection, animals were anaesthetized with ether and the peritoneal cavity was washed with 10 mL of cold PBS (pH 7.4). After centrifugation at 300 g for 10 minutes at 4°C, cells were collected, counted, and adjusted to 10⁶ cells per milliliter with culture medium. After 96-hour incubation in medium alone, cell supernatants from disease-free and AIA rats were collected for mediator quantification. To activate the production of various inflammatory mediators from rat peritoneal macrophages, cells were incubated with lipopolysaccharide (10 μg/mL; Sigma-Aldrich). RU was added simultaneously to cell activation. Cells or their supernatants were then tested for the presence of various inflammatory mediators.

Monocytes/macrophages were activated through the physiological CD23 pathway [22]. PBL-derived adherent cells must be preactivated to express surface CD23 as described [22] following their incubation with various cytokines, including recombinant human IL-4 (50 ng/mL), during 24 hours at 37°C. After washing, cells were tested for their surface CD23 expression (>80% CD23⁺) and then incubated in the presence of crosslinking CD23-MAb (clones 135, IgG1k, 20 μg/mL) during 48 to 96 hours or of IgE and anti-IgE as described [22, 23]. Cells may then be analyzed for their NOx (nitrite/nitrate) content (see below) and cell supernatants for the presence of various inflammatory mediators. To detect cell apoptosis, externalization of inner membrane phosphatidylserine and DNA content was investigated by flow cytometry using a fluorescein-conjugated annexin V and propidium iodide kit (Immunotech, Marseille, France).

After cultures, total RNA was extracted using Trizol (Invitrogen, Cergy Pontoise, France) and was subsequently purified on RNeasy columns (Qiagen, Hilden, Germany). Synthesis of biotin-labelled cRNA, purification, and hybridization (6 μg) to custom array membranes were performed according to the manufacturer's recommendations (OHS-011; SuperArray Bioscience Corporation, Frederick, MD, USA). For detailed gene content and housekeeping controls, see reference [24]. After local background subtraction, average signal intensity from duplicate spots was compared with values obtained for housekeeping genes using Alpha Imager HP automatic image capture software (Alpha-Innotec, San Leandro, CA, USA). For each gene, modulation was defined as the relative expression value for stimulated versus control sample.

For intracellular NO measurements, cells (>10⁵) were incubated with 10 μM DAF-FM-DA (4-amino-5-methylamino-2',7'-difluoro-fluorescein diacetate; Molecular Probes, now part of Invitrogen) for 1 hour at 37°C in 1 mL of culture medium. Cells were then washed in PBS and incubated in 0.5 mL of PBS for 30 minutes at 37°C. Intracellular NO content was then investigated by flow cytometry. Cell supernatants (48 to 72 hours) were assayed for the stable end product of NO, NO₂^- using the Griess reaction modified as detailed elsewhere [25]. The inhibitory analog of L-arginine, N(6)-(1-iminoethyl)-L-lysine/2HCl (L-NIL) (Coger SA, Paris, France) [26], was used to inhibit inducible nitric oxide synthase (iNOS)-mediated NO generation at a concentration of 1 mM. To detect human cytokine levels, we have used a human multi-assay Th1/Th2 II plex kit (Bender MedSystems, Vienna, Austria) and flow cytometry. Inflammatory mediators in rat sera or cell supernatants were quantified using appropriate enzyme-linked immunosorbent assay kits in accordance with the manufacturer's recommendations for prostaglandin E₂ (PGE₂) (R&D Systems Europe, Lille, France), monocyte chemoattractant protein-1 (MCP-1) (Tebu, Le Perray-en Yveline, France), TNF-α, and IL-1β (Biosource, Montrouge, France).

Comparisons were assessed using the Fisher exact test for proportions and the Mann-Whitney U test for quantitative values. A p value of less than 0.05 was considered significant. For some rat in vitro experiments, results were analyzed and compared using the Student t test for paired data.

In vitro, RU has generally been shown to display its activities at concentrations of 30 to 200 μM [19]. Prior to RU use, we first tested its effect on human normal or PHA-mediated cycling PBLs in vitro. We had no significant increase of the number of apoptotic (annexin⁺) or necrotic (propidium iodide⁺) cells after 96 hours of cell incubation in the presence of 1 to 200 μM RU (range from -5% to +7% of cells). RU was then used at concentrations of less than or equal to 100 μM in the following experiments. After their separation, human monocyte-derived macrophages were activated in the presence of 100 μM RU. The transcription of inflammatory genes was then analyzed using a macrophage-specific macroarray. Compared with resting cells, activated macrophages acquire a significant expression of 20 new mRNAs encoding various inflammatory mediators, chemotactic factors, and their receptors (Figure 1). In addition, we observed a significant increase in mRNA expression for genes encoding IL-1β, IL-8, TNF-α, TNF-R1, and macrophage migration inhibitory factor (MIF) (>110%; p < 0.05), known for their critical role during inflammatory response [2, 27]. Treatment of macrophages with RU inhibited the expression of most of the above genes (19/20 were totally suppressed), including IL-1β, TNF-α, TNF-R1, and many chemoattracting factors, or decreased the expression of genes such as MIF and IL-8 (p < 0.05) (Figure 1). Surprisingly, the transcription of the gene encoding IL-10, known as Th2 type cytokine, was significantly increased (>230%; p < 0.05). These findings indicate that RU is a potent inhibitor of the transcription of proinflammatory genes in human macrophages with the exception of that encoding IL-10.

Transcriptomic data led us to investigate the effect of RU on cytokine production at the protein level. Quantification of various cytokines in macrophage cell supernatants indicates that the addition of RU to human activated macrophages significantly (p < 0.02) decreased the concentrations of TNF-α, IL-1β, and IL-6 (Figure 2a), three critical proinflammatory cytokines. We failed to detect significant modulation of IL-10 levels in contrast to mRNA data (Figures 1b and 2a).

In addition to the above cytokines, inflammatory macrophages produce various non-proteic mediators, including iNOS-dependent NO [22]. Upon their activation, human macrophages produce NO, detected at the intracellular level by specific probe (DF-FM) and at the extracellular level through the quantification of nitrites, the final metabolites of NO in cell supernatants. Data in Figure 2b show that RU partly inhibited NO generation in a dose-dependent manner at both the intracellular and extracellular levels. Generation of NO from human macrophages was also reversed by L-NIL (Figure 2b), a specific inhibitor of iNOS [26], suggesting the ability of RU to reverse iNOS-mediated NO production.

In vitro observations in human cells led us to investigate the therapeutic anti-inflammatory effect of RU in vivo in rat AIA, an experimental model with many clinical and histopathological features of chronic human RA [28]. RU formulations and doses used in this work were based upon various in vivo studies [29] and our preliminary analysis showing the absence of an apparent toxic effect of up to 3,000 mg/kg total doses (data not shown). AIA Lewis rats thus were treated with RU suspension at 133 mg/kg s.c. doses (one dose every 2 days × 5). As therapeutic positive control, rats were treated with HC at 150 mg/kg total dose [30]. We found that administering five doses of RU, beginning the day after the appearance of the first arthritic symptoms, significantly improved the clinical course, including arthritic scores and weight progression of AIA rats as compared with the control groups (Figure 3a; p < 0.0001). The effect of RU was superior to that of HC in inhibiting arthritic scores (p < 0.001) and remained stable after treatment (Figure 3a). Figure 3b shows the external aspect of inflamed pads in RU-treated rats compared with untreated rats. No obvious toxicity was observed resulting from the treatment in the rats (for example, the weight of the treated normal rats was close to untreated controls). Hence, these experiments indicate that the RU effectively ameliorates AIA.

In a second set of experiments, we tested the ability of RU to prevent AIA establishment in rats. We injected RU every 2 days for a total of five injections, beginning on the day of adjuvant injection to the rats and before the appearance of any arthritic signs. Notably, we found that pretreatment with RU could significantly reduce the severity of arthritis and growth delay observed in control groups (Figure 3c; p < 0.0002). Hence, as shown in Figure 3d, RU treatment was effective either at disease onset or prevention of severe disease establishment, whereas no such activity was obtained with controls.

To confirm the anti-inflammatory effects of RU and the role of macrophages, we tested the levels of selected mediators in the sera from AIA animals following their treatment with or without RU. Data (Figure 4a) show that RU significantly reduced the levels of TNF-α, IL-1β, and MCP-1 in animal sera (p < 0.02), whereas the level of PGE₂, an indirect marker of COX-2 (cyclooxygenase-2) activity, was not affected by this treatment. These data reinforced human findings as they revealed that RU decreased the level of inflammatory cytokines, critical for RA pathogenesis in vivo. Although known as a major source of the above mediators [2], the role of macrophages during RU treatment required more direct ex vivo analysis. Peritoneal macrophages were then isolated from RU-treated or untreated arthritic rats and promptly incubated in medium alone. After 48 hours of incubation, cell supernatants were analyzed for the levels of TNF-α and nitrites as inflammatory markers. Our results (Figure 4b) clearly show that increased inflammatory macrophage-derived TNF-α and NO in AIA rats were attenuated after treatment with RU.

Finally, we tested the ability of RU to reduce rat macrophage inflammatory responses in vitro. Peritoneal macrophages were isolated from healthy rats and activated without various doses of RU. Our data show (Figure 4d) that the levels of NO, as indicated by the concentration of nitrites in cell supernatants, decreased after the addition of RU in a dose-dependent manner (p < 0.006). As in human cells, L-NIL had a similar inhibitory effect compared with RU, suggesting its ability to downregulate iNOS-mediated NO generation in macrophages. Inhibition reached a plateau at 50 μM RU, a concentration that was subsequently used to investigate the RU effect on other mediators. As in in vivo findings, RU decreased the level of inflammatory cytokines (TNF-α, IL-1β, and MCP-1) generated from activated rat macrophages (Figure 4c). As in ex vivo data, the production of PGE₂ was not modified after the addition of similar RU dilution.