A Soft Coral-Derived Compound, 11-epi-Sinulariolide Acetate Suppresses Inflammatory Response and Bone Destruction in Adjuvant-Induced Arthritis

In recent years, a significant number of metabolites with potent anti-inflammatory properties have been discovered from marine organisms, and several of these compounds are now under clinical trials. In the present study, we isolated 11-epi-sinulariolide acetate (Ya-s11), a cembrane-type compound with anti-inflammatory effects, from the Formosa soft coral Sinularia querciformis. Preliminary screening revealed that Ya-s11 significantly inhibited the expression of the proinflammatory proteins induced nitric oxide synthase and cyclooxygenase-2 in lipopolysaccharide-stimulated murine macrophages. We also examined the therapeutic effects of Ya-s11 on adjuvant-induced arthritis (AIA) in female Lewis rats, which demonstrate features similar to human rheumatoid arthritis (RA). Animal experiments revealed that Ya-s11 (subcutaneously 9 mg/kg once every 2 days from day 7 to day 28 postimmunization) significantly inhibited AIA characteristics. Moreover, Ya-s11 also attenuated protein expression of cathepsin K, matrix metalloproteinases-9 (MMP-9), tartrate-resistant acid phosphatase (TRAP), and tumor necrosis factor-α (TNF-α) in ankle tissues of AIA-rats. Based on its attenuation of the expression of proinflammatory proteins and disease progression in AIA rats, the marine-derived compound Ya-s11 may serve as a useful therapeutic agent for the treatment of RA.


Introduction
Rheumatoid arthritis (RA) is an autoimmune and chronic erosive inflammatory disease characterized by chronic edema and inflammation of the synovial tissue that lines joints. RA affects approximately 1% of the adult population worldwide, and RA patients present an average of 43% of maximum possible joint destruction after 20 years of suffering from the disease [1][2][3]. As RA progresses, the lining of the joints degenerates, leading to articular destruction and decreased joint mobility with radiological evidence of erosive damage and significantly impacting quality of life within 2 years of disease onset [4,5]. Over the past 2 decades, more effective therapeutic strategies for RA including synthetic modifying antirheumatic drugs and/or biologic agents have been developed, but they carry potential risks. Additionally, nonsteroidal medications and corticosteroids are required as adjunctive therapy [6]. Therefore, new drug development for RA remains essential.
Prior studies have indicated that inflammatory processes are critical to the development of RA [2,7,8]. Synovial inflammation causes hyperplasia of the synovial tissue, with clusters of large numbers of infiltrating cells, and a tumor-like structure called pannus invades the joint lining and destroys local articular structure [9,10]. Although the precise etiology of RA remains unclear, macrophages, neutrophils, lymphocytes, and synovial fibroblasts in hyperplastic synovial tissue have been identified as major participants in the initiation and development of RA [11,12]. These infiltrating cells can release proinflammatory cytokines such as tumor necrosis factor alpha (TNF-a) that mediate synovial inflammation and joint destruction [8,13]. In addition, proinflammatory cytokines activate synovial fibroblasts and chondrocytes and lead to upregulation of osteoclast-related proteins including cathepsin K, matrix metalloproteinase-9 (MMP-9), and tartrate-resistant acid phosphatase (TRAP), which also participate in inflammatory arthritis resulting in joint destruction [14].
Lipopolysaccharide (LPS)-challenged murine macrophages are widely used for in vitro anti-inflammatory screening of terrestrialand marine-derived natural compounds [15][16][17]. LPS activates the nuclear factor-kB pathway to massively upregulate the proinflammatory proteins inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) in murine macrophages [18]. Several studies have indicated that macrophages play a critical role in RA by expressing pro-inflammatory proteins in inflamed synovial tissue and at the cartilage-pannus junction [11,19]. They also release cytokines and chemokines that mediate inflammation and form a complex cytokine network with neutrophils, lymphocytes, and synovial fibroblasts in RA [13]. Macrophages are also important for their capacity to differentiate into osteoclasts, multinucleated giant cells, or mononuclear precursor cells. A number of previous studies have employed macrophage-related cell lines to investigate RA-related mechanisms of action and identify the anti-arthritic activity of compounds [8,[20][21][22]. Adjuvant-induced arthritis (AIA) is also widely used as an animal model for the study of clinical RA. Several potential anti-RA agents have been discovered from this model [1,8,23].
In recent years, a significant number of metabolites with potent bioactive properties have been discovered in marine organisms, and several of these compounds are now under clinical trials [24,25]. In this study, we aimed to examine an anti-inflammatory cembrane-type compound for its ability to suppress RA progression. We isolated 11-epi-sinulariolide acetate (Ya-s11), a known cembrane-type compound, from the soft coral Sinularia querciformis. We found that Ya-s11 inhibited expression of the pro-inflammatory proteins iNOS and COX-2 in LPS-induced murine RAW264.7 macrophages. We also evaluated the anti-inflammatory and anti-rheumatic effects of Ya-s11 in AIA rats.

Preparation of 11-epi-sinulariolide acetate
In the present study, 11-epi-sinulariolide acetate (Ya-s11, structure shown in Fig. 1) was isolated and purified from nonprotected soft coral, S. querciformis, collected from the Dongsha Islands, Taiwan in 2008. We state that no specific permissions were required for sample collection in Dongsha Islands. Moreover, we also state that the area was not privately owned or protected. The method of 11-epi-sinulariolide acetate extraction was modified from that of Lu et al. [26,27]. The soft coral specimens were minced and exhaustively extracted with 95% ethanol (8 liters). The crude extract was concentrated to an aqueous suspension and further partitioned between n-hexanes, ethyl acetate, and H 2 O. The n-hexaneethyl acetate layer was separated over normal phase silica gel by column chromatography and eluted with n-hexane, ethyl acetate, acetone, and methanol to yield 29 fractions. Fraction 11 was further eluted with n-hexanes and acetone (1:1) over normal phase silica gel to generate 5 subfractions. Subfraction 4 was further separated by reverse-phase RP18 gel elution with H 2 O and acetonitrile (2:1) by column chromatography to yield Ya-s11. The structure of Ya-s11 was confirmed by nuclear magnetic resonance spectroscopy (NMR) [28]. The purity (.98%) of Ya-s11 was determined and verified by 1H-NMR and 13C-NMR spectra (Varian Mercury Plus 400 FT-NMR at 400 MHz, Varian, CA, U SA).

In vivo study
Animals. Female Lewis rats (180-220 g) were used for the experiments and obtained from National Laboratory Animal Center, Taiwan. The rats were maintained in plexiglass cages in a temperature-controlled (2461 uC) room on a 12-h light/dark cycle and given free access to food and water. Each rat was used only once during the experiment. All drug injections were performed under isoflurane anesthesia. The use of animals accorded to the Guiding Principles in the Care and Use of Animals of the American Physiology Society and was approved by the institutional animal care and use committee of National Sun Yat-sen university. Every effort was made to minimize the number of animals used and their suffering.

Adjuvant induce arthritis (AIA) and compound treatment
The method of generating AIA rats was modified from that of Sano et al. (1992) and Turull and Queralt (2000) [31,32]. Heatkilled and lyophilized Mycobacterium butyricum was suspended in incomplete Freund's adjuvant 10 mg/ml (Sigma, St. Louis, MO, USA) on ice. Rats were immunized with an adjuvant injection of 10 mg/ml M. butyricum in incomplete Freund's adjuvant. On day 0, rats were injected intradermally at the base of the tail with 0.1 ml of adjuvant, and the development of arthritis was monitored from day 0 to day 28. Lewis rats were randomly divided into 5 groups: AIA (n = 7), AIA plus Ya-s11 (3 mg/kg) (n = 6), AIA plus Ya-s11 (9 mg/kg) (n = 6), naïve (n = 6), and Ya-s11 (9 mg/kg) treatment alone (n = 6). In the AIA plus Ya-s11 groups, rats received 10 times of Ya-s11 at 2-day intervals between days 7 and 28. All rats underwent measurement of hindpaw edema and clinical evaluation at day 0 and before every Ya-s11 injection between days 7 and 28, and then rats were sacrificed on day 28 for histopathological analysis and immunochemical staining. The extent of edema in the foot and hindpaw was measured from day 0 (baseline) to day 28 after AIA using a plethysmometer (Paw Volume Meter, Singa, Taiwan). Rats were evaluated for arthritic processes every 2 days using a macroscopic scoring system, with score 0 = no signs of arthritis, 1 = swelling and/or redness of the paw or 1 digit, 2 = two tow joints involved, 3 = more than two joints involved, 4 = severe arthritis of the entire paw and all digits [9,33,34]. The macroscopic score for each rat was calculated by adding the scores of each individual paw [9,34].

Histopathological examination and immunohistochemical staining
Rats were sacrificed by perfusion with ice-cold PBS and 4% paraformaldehyde on day 28 after immunization, and ankle joints were removed and fixed in 10% neutral formalin for 4 days. The ankle joints were decalcified with 12.5% ethylenediaminetetraacetic acid (EDTA) in 10% neutral formalin for 2 weeks and then sectioned on the sagittal plane through the center of samples. The specimens were dehydrate in a graded series of alcohol (Tissue-Tek, Sakura Finetek Japan Co., Ltd, Japan), embedded with paraffin, and cut into 2mm sections (Microm HM550, Microm, Waldorf, Germany for hematoxylin and eosin (H&E) and immunohistochemical staining. General and pathological changes in morphology were assessed by microscopic examination using an upright microscope for higher magnification (DM 6000B, Leica Inc., Germany) and a stereomicroscope for lower magnification (APO Z16, Leica Inc., Singapore) with a microscope digital image output system (SPOT idea 5.0 Mp Color Digital Camera, Diagnostic Instruments Inc., Sterling Height, MI., USA). To quantitatively evaluate joint destruction in the ankle, the degree of morphologic changes in each group was scored on photomicrographs of tissue sections, with score 0 = no damage, 1 = edema, 2 = presence of inflammatory cells, and 3 = cartilage and bone damage (Cuzzocrea et al., 2005). Infiltrating cells were further quantified according to the histopathological features of neutrophils, lymphocytes, macrophages, and synovial fibroblasts [35][36][37].
Ankle joint specimens were processed for immunohistochemical analysis as described in previous studies [38][39][40]. Paraffinembedded ankle joint sections were placed on slides, deparaffi-nized with xylene, and dehydrated in a graded series of alcohol, after which endogenous peroxidase activity was quenched by 30min incubation in 0.3% hydrogen peroxide.

Statistical analysis
All data are presented as mean 6 SEM. For the immunoreactivity data, each test band is shown as the integrated optical density (IOD) computed with respect to the average optical density of the corresponding control (LPS-only treatment) band. The data were analyzed using 1-way analysis of variance (ANOVA) followed by Student-Newman-Keuls post hoc test (SigmaStat 3.5 for Windows). Differences resulting in P values less than 0.05 were considered significant.
Effect of Ya-s11 on the clinical features of AIA AIA developed rapidly in rats immunized with heat-killed M. butyricum, and Figure 3 show the square of typical representative macroscopic photographs. Both AIA and AIA plus Ya-s11 (3 mg/ kg) groups demonstrated edema and erythema on the ankle joints and hindpaws ( Fig. 3B and C). Ya-s11 (9 mg/kg) significantly attenuated the AIA-induced edema and erythema of the hindpaw (Fig. 3D). Figure 3E illustrates the time-dependent increase of paw edema in immunized rats. Paw edema significantly increased to approximately 121.4-127.4% of baseline values from day 19 to day 23 in the AIA group, and the AIA plus Ya-s11 (9 mg/kg) group demonstrated a dose-dependent inhibition of paw edema compared with the AIA-only group. Quantitative analysis using a macroscopic scoring system (Fig. 3F) revealed a significant reduction in the AIA-induced upregulation of arthritis score in the AIA plus Ya-s11 (9 mg/kg) group. The AIA plus Ya-s11 (3 mg/kg) group demonstrated a slight but not significant attenuation in arthritis score (Fig. 3F). No change in paw edema or arthritis score was observed in the group treated with Ya-s11 (9 mg/kg) (Fig. 3E-F).
The effect of Ya-s11 on histological features of AIA in the ankle joint Rats were sacrificed on day 28 after immunization, and paraffin sections of ankle joints were subjected to H&E staining for histopathological analysis. Representative photographs of ankle joint sections are shown from control (Fig. 4A, E, I), AIA (Fig. 4B, F, J), AIA plus Ya-s11 (3 mg/kg) (Fig. 4C, G, K) and AIA Plus Ya-s11 (9 mg/kg) rats (Fig. 4D, H, L), respectively. Similar to previous studies [4,41], in the AIA group, synovial tissue demonstrated synovial hyperplasia, cartilage and bone erosion (Fig. 4I), and moderate to severe infiltration of immune cells into subchondral bone marrow (Fig. 4J). The AIA plus Ya-s11 (3 mg/kg) group did not show attenuation of the synovial hyperplasia or cartilage and bone erosion, but moderated bone resorption in bone marrow (Fig. 4G, K). By contrast, the AIA plus Ya-s11 (9 mg/kg) group demonstrated significantly inhibited AIA-induced joint destruction and synovial hyperplasia, cartilage and bone erosion, pannus formation, and bone resorption (Fig. 4H, J). Histological damage scores were significantly higher in the AIA group compared to the naïve group. Those of the AIA plus Ya-s11 (9 mg/kg) group were significantly lower than in the AIA and AIA plus Ya-s11 (3 mg/kg) groups. No significant differences were observed between the AIA and AIA plus Ya-s11 (3 mg/kg) groups (Fig. 4M). Effect of Ya-s11 on infiltrating cells in synovial tissue Pannus formation is accompanied by the infiltration of inflammatory cells, including lymphocytes, monocytes/macrophages, neutrophils, and synovial fibroblasts, into synovial tissue [2]. Representative photographs show synovial tissue stained with H&E from the naïve group (Fig. 5A), AIA group (Fig. 5B), AIA plus Ya-s11 (3 mg/kg) group (Fig. 5C), and AIA plus Ya-s11 (9 mg/kg) group (Fig. 5D). The synovial tissue of the naïve group demonstrated synovial fibroblasts with few immune cells (Fig. 5A), and marked upregulation of infiltrating cells was apparent in synovial tissue from the AIA group (Fig. 5B). Ya-s11 (9 mg/kg) significantly inhibited AIA-induced upregulation of infiltrating cells in synovial tissue (Fig. 5D). Neutrophils, lymphocytes, macrophages, and synovial fibroblasts all significantly increased between the naïve and AIA groups (Fig. 5E-R), and the AIA plus Ya-s11 (9 mg/kg) group demonstrated a significant decrease in the number of neutrophils, lymphocytes, and macrophages as well as inhibition of synovial fibroblast proliferation. No significant change was found in the AIA plus Ya-s11 (3 mg/kg) group compared with the AIA group.

Discussion
This study employed LPS-induced RAW264.7 murine macrophage cells and AIA as in vitro and in vivo models, respectively, to assess the anti-inflammatory and anti-arthritic effects of Ya-s11. Marine-derived Ya-s11 significantly down-regulated expression of the proinflammatory proteins iNOS and COX-2 in LPS-induced RAW 264.7 murine macrophage cells. Administration of Ya-s11 also significantly inhibited AIA-induced paw edema and the upregulation of arthritis score in a dose-dependent manner.
Histopathological and immunohistochemical examination further demonstrated that AIA-induced histological features in the ankle joint and the osteoclast-related proteins, cathepsin K, MMP-9, TRAP, and TNF-awere upregulated in ankle joint tissue in the AIA group. Systemic injection of Ya-s11 (9 mg/kg) not only attenuated AIA-induced pathological changes in the ankle joint, but also significantly reduced the osteoclast-related protein expression.

Effect of Ya-s11 anti-inflammatory activity in vitro and in vivo
RA is a synovial inflammatory disease characterized by proliferative synovial fibroblasts and infiltrating cells [2]. Previous studies have highlighted the important role played by macrophages in the process of RA [11,19]. Macrophages mediate synovial inflammation in RA by forming complex cytokine networks with neutrophils, lymphocytes, and synovial fibroblasts and are also critical to osteoclast differentiation [1,2,13,23,41]. Ya-s11 was able to downregulate iNOS and COX-2 protein expression in LPS-stimulated macrophage RAW 264.7 cells (Fig. 2), a well-established in vitro model for assessing the antiinflammatory activity of compounds. In the in vivo study, AIA rats demonstrated a significantly increased number of macrophages in ankle joint synovial tissue as well as AIA-induced increased of neutrophils and lymphocytes with fibroblast proliferation (Fig. 5). Although subcutaneous injection of 3 mg/ml Ya-s11 did not significantly decrease the number of infiltrating cells in synovial tissue, synovial hyperplasia was reduced in this treatment group (Fig. 4). Treatment with 9 mg/ml Ya-s11 inhibited synovial inflammation with reduced cell infiltration in AIA-rats.

AIA-induced joint destruction with osteoclast-related protein expression
Many studies have clearly illustrated the mediation of joint inflammation, pannus formation, and invasion of infiltrating cells into cartilage and bone by continuous release of osteoclast-related proteins [43][44][45]. TNF-a is a cytokine produced by macrophages that also plays an important role in RA joint destruction and may mediate MMP-9 and cathepsin K expression in RA by upregulating  the transcription factor c-Fos/AP-1 [10,11]. MMP-9 and cathepsin K in turn play important roles in osteoclastogenesis and osteoclastic activity [38,40,44] and are expressed by leukocytes, synovial fibroblasts, chondrocytes, and osteoclasts [39,40,42,43,46,47]. Our immunohistochemical analyses revealed AIA-induced MMP-9 and cathepsin K expression in synovial tissue, cartilage, and bone marrow (Fig. 6-7). TRAP, which is considered a marker of osteoclasts and plays important roles in osteoclast activity [4], was also observed in bone marrow (Fig. 7). Previous studies have implicated MMP-9, cathepsin K, and TRAP in the bone resorption process in bone marrow [42,48], and our histopathological assessment of AIA rats also indicated significant bone resorption in the bone marrow (Fig. 4). Hence, the continuous formation of these destructive enzymes in joints affected by RA, with the increase of infiltrating cells, pannus formation, and cartilage and bone erosion and resorption, leads to the development of severe arthritis with joint edema [40,44]. Accordingly, in the present study, the AIA group demonstrated significant differences in foot and paw edema and clinical evaluation between day 11 and day 28, with edema and erythema of the ankle joint apparent on typical representative macroscopic photographs (Fig. 3).

Effect of Ya-s11 on AIA-induced joint destruction
We demonstrated that Ya-s11 exerts a therapeutic effect on joint destruction in a rat model of AIA. The therapeutic efficacy of Ya-s11 was not limited to general anti-inflammatory effects, but included substantial inhibition of cartilage and bone destruction compared to AIA rats, as well as inhibition of osteoclast-related protein expression. The AIA plus Ya-s11 (9 mg/kg) group demonstrated inhibition of MMP-9 and cathepsin K expression in the synovial tissue and bone marrow (Fig. 6) as well as inhibited MMP-9 protein expression in articular cartilage. Although cathepsin K expression was not inhibited in chondrocytes, the histological features of the joint display did not display significant morphologic changes in the AIA+Ya2s11 (9 mg/kg) group. TRAP expression in the bone marrow was also inhibited by treatment with Ya-s11 (9 mg/kg). Immunohistochemical analysis further demonstrated an increase in TNF-a in the synovial tissue of the AIA group and its reduction by treatment with Ya-s11 (9 mg/kg). Thus, osteoclast-related protein expression and synovial inflammation were inhibited by Ya-s11, which also demonstrated a dose-dependent effect on foot and paw edema and the clinical evaluation of arthritis and delayed the onset of arthritis. Rats in the Ya-s11 (9 mg/kg) group demonstrated only approximately 10% increase of foot and paw edema compared to baseline, with typical representative macroscopic photographs of the paw implying that edema was not apparent in the photographs, only erythema. In summary, Ya-s11 demonstrates anti-RA activity, with reduced expression of osteoclast activityrelated proteins TNF-a, MMP-9, cathepsin K, and TRAP and effective reduction of the clinical features of AIA.
Ya-s11 as a potential anti-rheumatoid arthritis agent for drug development The present study demonstrated the efficacy of Ya-s11 as a potential anti-inflammatory compound. We also illustrated its anti-RA activity in AIA model rats, in which Ya-s11 inhibited TNF-a cathepsin K, TRAP, and MMP-9 expression and decreased the major features of RA pathogenesis. Ya-s11 is a cembrane-type natural compound of marine origin originally isolated from the Red Sea [28,49], which we isolated from the soft coral Sinularia querciformis. A cembrane-type compound was first isolated from Sinularia querciformis in 1985 [50], and since then 8 types of cembrane-type compound with anti-inflammatory activity have been isolated from this species by Lu et al. [26,27]. However, Ya-s11 can also be isolated and purified from the same genus of soft coral, Sinularia flexibilis, which can be cultured in a culture tank [29], and an increasing number of anti-inflammatory compounds have also been isolated from Sinularia sp. [29,[51][52][53][54]. The main structure of Ya-s11 is a cembranolide analogue, the chemical skeleton of which differs from that of steroids, which may further highlight the potential of Ya-s11 as a useful therapeutic agent for rheumatic diseases and other inflammatory disease.

Conclusions
In this study, we isolated and purified Ya-s11 from the soft coral Sinularia querciformis. In vitro study revealed that Ya-s11 significantly inhibited the expression of the proinflammatory proteins iNOS and COX-2 in LPS-challenged murine macrophages cell model. In vivo study revealed that Ya-s11 significantly reduced AIA characteristics. Moreover, using histological analysis, we had found that Ya-s11 also improved the histopathologic features of RA. Immunohistochemical analysis showed that Ya-s11 also attenuated protein expression of cathepsin K, MMP-9, TRAP, and TNF-a in ankle tissues of AIA-rats. We concluded that Ya-s11 ameliorated the infiltration of inflammatory cells and bone destruction and downregulated the expression of osteoclast-related proteins in the ankle tissue of AIA rats. Hence, the soft coralderived compound Ya-s11 may serve as a useful therapeutic agent for the treatment of RA.