Carbon Monoxide Inhibits Tenascin-C Mediated Inflammation via IL-10 Expression in a Septic Mouse Model

Tenascin-C (TN-C), an extracellular matrix (ECM) glycoprotein, is specifically induced upon tissue injury and infection and during septic conditions. Carbon monoxide (CO) gas is known to exert various anti-inflammatory effects in various inflammatory diseases. However, the mechanisms underlying the effect of CO on TN-C-mediated inflammation are unknown. In the present study, we found that treatment with LPS significantly enhanced TN-C expression in macrophages. CO gas, or treatment with the CO-donor compound, CORM-2, dramatically reduced LPS-induced expression of TN-C and proinflammatory cytokines while significantly increased the expression of IL-10. Treatment with TN-C siRNA significantly suppressed the effects of LPS on proinflammatory cytokines production. TN-C siRNA did not affect the CORM-2-dependent increase of IL-10 expression. In cells transfected with IL-10 siRNA, CORM-2 had no effect on the LPS-induced expression of TN-C and its downstream cytokines. These data suggest that IL-10 mediates the inhibitory effect of CO on TN-C and the downstream production of proinflammatory cytokines. Additionally, administration of CORM-2 dramatically reduced LPS-induced TN-C and proinflammatory cytokines production while expression of IL-10 was significantly increased. In conclusion, CO regulated IL-10 expression and thus inhibited TN-C-mediated inflammation in vitro and in vivo.


Introduction
Damage-associated molecular patterns (DAMPs) are endogenous molecules that can perpetuate inflammatory responses during cell stress or injury. The ECM glycoprotein TN-C, is specifically induced upon tissue injury [1,2] and infection [3,4] and upregulated in septic patients [5]. TLR4mediated TN-C expression induces cytokine production in both human and murine macrophages and in rheumatoid arthritis synovial fibroblasts [6]. Activated TLR4 induces TN-C expression in synovial fibroblasts and myeloid cells [7]. TN-C induction is crucial for the proinflammatory response in vivo [8]. Importantly, glucocorticoids can inhibit the expression of TN-C in bone marrow stromal cells and fibroblasts [9]. In addition, mice and bone marrow-derived macrophages (BMDMs) deficient in TN-C display lower production of proinflammatory cytokines such as TNFduring LPS-induced sepsis. Thus, TN-C has been recognized as a regulator of the early immune response [8].
IL-10 is a vital anti-inflammatory cytokine which is required for dampening inflammatory signals and defending the host from excessive inflammation [10]. Mice lacking IL-10 infected with bacterial pathogens display high mortality, associated with excessive inflammatory responses [10]. Low levels of IL-10 expression were associated with various inflammatory diseases such as ulcerative colitis, Crohn's disease, and asthma in humans [11,12]. The anti-inflammatory effect of IL-10 is mediated through the JAK1-STAT3 pathway which 2 Mediators of Inflammation leads to the inhibition of proinflammatory proteins such as TNF-and IL-6 [10,13]. Higher expression of IL-10 was found in BMDMs from TN-C-deficient mice while there was lower expression of proinflammatory cytokines [8], indicating an anti-inflammatory role of IL-10 in the TN-Cmediated inflammatory disease model.
Carbon monoxide (CO) is generated as an end product of the oxidative degradation of heme by the enzymatic action of heme oxygenase, which converts heme into biliverdin, free iron, and CO [14]. Anti-inflammatory effects of CO have been evident in murine models of sepsis, postoperative ileus, and organ xenotransplantation [15,16]. In addition, CO has been found to be an important regulator in the suppression of inflammatory cytokines and mediators including inducible nitric oxide synthase (iNOS), TNF-, and IL-6 [17,18] as well as induction of the anti-inflammatory cytokine IL-10 [19]. The production of endogenous CO was essential for IL-10-dependent inhibition of TNF-expression [20]. To date, there are no reports regarding the effects of CO-mediated IL-10 production on the regulation of TN-C-mediated inflammation. Therefore, in the current study, we examined the effects of CO-dependent IL-10 generation on TN-C-mediated inflammation in macrophages and in the septic mice model.

Reagents and Antibodies.
Tenascin-C antibody was purchased from Cell Signaling Technology (MA, USA).actin, anti-mouse and anti-goat antibodies conjugated to horseradish peroxidase were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Lipopolysaccharide (LPS) and protease inhibitor cocktail sets were purchased from Sigma-Aldrich (St. Louis, MO, USA). Zinc protoporphyrin IX (ZnPPIX) was obtained from Frontier Scientific Inc. Recombinant mouse IL-10 proteins were purchased from R&D systems. Dulbecco's Modified Eagle Medium (DMEM), fetal bovine serum (FBS), penicillin-streptomycin, and sodium pyruvate were purchased from Invitrogen (Grand Island, NY, USA). All other chemicals were obtained from Sigma-Aldrich.

Cell
Culture. RAW 264.7 cells and peritoneal macrophages were cultured in DMEM (Invitrogen) containing 10% Fetal Bovine Serum (FBS) and 1% penicillin streptomycin at 37 ∘ C in 5% CO 2 until 75-80% confluence. For preparation of peritoneal macrophages, mice were injected intraperitoneally with 3% thioglycolate for 3 days and cells were collected for culture. Cells (5 × 10 5 /mL) were seeded in 6-well plates and incubated overnight for subsequent experiments.

Animal Model.
Seven-week-old wild type male C57BL/6 mice were pretreated with CORM-2 (30 mg/kg, i.p.) [21] or RuCl 3 (30 mg/kg, i.p.). And then, mice were injected with LPS (10 mg/kg, i.p.). After 2 hours, blood serum and liver tissues were collected and stored at −80 ∘ C for protein and RNA analysis. All experiments with mice were approved by the Animal Care Committee of the University of Ulsan, Ulsan, Korea.

Western
Blotting. Cell extracts were lysed using lysis buffer containing RIPA buffer, protease inhibitor, and phosphatase inhibitors. After lysis, protein concentration was measured by BCA assay (Pierce Biotechnology Inc., Rockford, IL, USA). Samples containing equal amounts of protein were subjected to electrophoresis and proteins were transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were blocked with 5% skim milk for 20 min and then incubated at 4 ∘ C overnight with primary antibodies, followed by secondary antibodies against TN-C andactin conjugated with horseradish peroxidase. The enhanced chemiluminescence (ECL) Western Blotting Detection System (GE Healthcare Life Sciences, Buckinghamshire, UK) was used to visualize the immunoreactive bands.

Real Time RT-PCR.
Total RNA was extracted from RAW 264.7 peritoneal macrophages/liver tissues using TRIzol reagent (Invitrogen) according to the manufacturer's instructions. In addition, cDNA was prepared by using M-MLV reverse transcriptase (Promega) and oligo (dT) 15 primer (Promega). The formulated cDNA was subjected to Real Time RT-PCR using SYBR Green qPCR Master Mix (2x) (USB products, Affymetrix) on an ABI 7500 Fast Real-Time PCR System (Applied Biosystems) for mouse GAPDH . GAPDH was used as internal loading control to normalize all PCR products.

Enzyme Linked Immunosorbent Assay (ELISA).
Macrophages on 6-well plates were incubated overnight and then pretreated with CORM-2 for 1 h followed by stimulation with LPS for 24 h. In addition, mice were administrated with CORM-2 for 2 h and then sepsis was induced by LPS injection. After 2 h, supernatants collected from various samples or blood serum collected from different mice were assayed for TNF-and IL-6 by using a mouse ELISA kit (Biolegend).

Statistical
Analysis. Statistical differences between groups were evaluated by one-way ANOVA (nonparametric) or Student's -test when multiple groups were compared. Results are expressed as the means ± SEM. Differences were considered to be significant when * < 0.05, * * < 0.01, and * * * < 0.001.

LPS Increases TN-C Expression in a Time-and Dose-Dependent
Manner. Macrophages have pro-or antiinflammatory functions depending on the type of stimuli [22]. Stimulation of macrophages with Gram-negative bacterial LPS can enhance the expression of TN-C [7]. TLR4 was involved in the induction of TN-C and subsequent cytokine synthesis in both human and murine macrophages [23] and human chondrocytes [24]. In the present study, we examined inflammatory responses in murine RAW 264.7 macrophages treated with LPS (100 ng/mL). TN-C mRNA and protein expression increased at 4 and 8 h after LPS treatment (Figures 1(a) and 1(b)), respectively. Therefore, in subsequent experiments, we measured TN-C mRNA and protein expression at 8 h. Furthermore, LPS dosedependently increased TN-C mRNA and protein expression at 8 h (Figures 1(c) and 1(d)). These results suggest that LPS induces TN-C expression in a time-and dose-dependent manner in RAW 264.7 macrophages.

CO Inhibits LPS-Induced TN-C Expression and Proinflammatory Cytokines
Production. The anti-inflammatory, antiapoptotic, and cytoprotective properties of CO are well known [25]. Furthermore, it has been reported that the generation of endogenous CO was necessary for IL-10-dependent inhibition of TNF-expression [20]. CO can be generated pharmacologically from CO-releasing molecules (CORMs), which consist of a heavy metal such as ruthenium surrounded by carbonyl groups [26,27]. In brain endothelial cells, the LPS-induced activation of inflammatory signals such as NF-B (p65), COX-2 expression, and PGE2 production was inhibited by CORM-2 pretreatment [28]. In the present study, we investigated the effects of CO on TN-C-mediated inflammation. RAW 264.7 cells were pretreated with CORM-2 and then stimulated with LPS. We found that CORM-2 significantly and dose-dependently suppressed the expression of LPS-stimulated TN-C expression (Figure 2(a)). In addition, pretreatment with CORM-2 significantly reduced the mRNA and protein levels of proinflammatory cytokines such as TNF-and IL-6 ( Figures 2(b), 2(c), and 2(d)), respectively. To confirm the effects of CO on TN-C-meditated inflammation, cells were pretreated with or without CORM-2 or RuCl 3 (negative control for CORM-2) and then stimulated with or without LPS. Interestingly, CORM-2 significantly downregulated LPS-induced TN-C as well as proinflammatory cytokines expression whereas RuCl 3 did not have any effect (Figures 2(e), 2(f), 2(g), and 2(h)). To confirm the effects of CO, we pretreated the cells with CO gas and then incubated them with LPS. Consistently, CO gas dramatically reduced LPS-stimulated expression of TN-C ( Figure 2(i)), as well as TNF- (Figure 2(j)) and IL-6 ( Figure 2(k)).
To further confirm the effects of CO on LPS-induced TN-C expression and proinflammatory cytokines, mouse peritoneal macrophages were pretreated with CORM-2 at various concentrations and incubated with LPS. We found that CORM-2 dramatically decreased LPS-induced TN-C ( Figure 3(a)) and its downstream cytokines (Figure 3(b)). In contrast, RuCl 3 did not reduce the expression of TN-C and proinflammatory cytokines (Figures 3(c) and 3(d)). These results suggest that CO can suppress LPS-mediated inflammatory signals in vitro.

CO/HO-1 Inhibits TN-C-Mediated Inflammation via IL-10
Induction. Low doses of CO suppressed inflammatory responses in a murine model of sepsis through inhibition of inflammatory cytokines production [18,29] as well as increased LPS-induced expression of the anti-inflammatory cytokine IL-10 in various cell types [19,29]. In addition, mice deficient with TN-C displayed lower levels of TNFand downstream cytokine production in LPS-treated septic mice and bone marrow-derived macrophages (BMDMs) [8]. In our study, we investigated the effects of CO on TN-C-induced proinflammatory cytokines expression and the expression of anti-inflammatory IL-10 in RAW 264.7 and peritoneal macrophages. CORM-2 significantly and dose-dependently induced the expression of IL-10 in LPSstimulated macrophages (Figures 4(a) and 4(b)). Furthermore, treatment with CO gas significantly increased levels of IL-10 ( Figure 4 4(g), and 4(h)), whereas it had no effect on IL-10 expression (Figure 4(i)), suggesting that IL-10 is regulated independently of TN-C and its downstream cytokines. To confirm the function of CO-induced IL-10 on TN-C-mediated inflammation, macrophages were transfected with IL-10 siRNA and treated with CORM-2 prior to LPS-stimulation. The efficiency of IL-10 siRNA transfection was shown in Figure 4(j). We found that IL-10 siRNA reversed the inhibitory effect of CORM-2 on TN-C expression and inflammatory cytokines production in LPS-stimulated macrophages relative to control siRNA (Figures 4(k), 4(l), and 4(m)). Pretreatment of recombinant IL-10 with or without LPS stimulation showed the same efficiency of CORM-2 to significantly decrease TN-C expression (Figure 4(n)). Also, HO-1 increases IL-10 production [30]. According to Inoue and colleagues [30], overexpressions of HO-1 provide the cytoprotection via the mediation of IL-10 production. Thus, we examined the expression of antiinflammatory gene HO-1 under these conditions. Interestingly, we found that CORM-2 significantly increased the level of HO-1 expression (Figure 4(o)), whereas RuCl 3 did not have any effect on HO-1 expression (Figure 4(o)) and conversely decreased the expression levels of TN-C. The inhibition of HO activity using ZnPPIX, however, did not reverse the effects of CORM-2 on TN-C expression (Figure 4(p)) indicating that CORM-2 mediated suppression of TN-C is independent of HO activity in LPS-stimulated RAW 264.7 macrophages. Based on these results, we conclude that COinduced IL-10 inhibits TN-C-mediated inflammation.

CO Inhibits TN-C-Mediated Inflammation In Vivo in a Septic Mice
Model. Sepsis, a systemic inflammatory response, results from excessive production of proinflammatory cytokines by LPS stimulation [31]. In addition, proinflammatory cytokines such as TNF-, IL-1 , and IL-6 have been found at higher levels in septic patients [32,33]. Administration of LPS in mice revealed that TN-C expression is necessary for proinflammatory signaling [8]. Furthermore, application of exogenous CO inhibits LPS-induced production of TNF-while it increases IL-10 production in vitro     (20 M) or RuCl 3 (20 M) for 1 h and stimulated with or without LPS (100 ng/mL) for 8 h. RuCl 3 was used as a negative control for CORM-2. TN-C, TNF-, and IL-6 mRNA levels were detected by Real Time RT-PCR. Data represents mean ± SEM, * * < 0.001 as compared with control; # < 0.05, ## < 0.001, and ns, nonsignificant, as compared with the cells exposed to LPS alone. and in vivo [29]. However, there are no reports regarding the effects of CO-mediated IL-10 production in relation to the regulation of TN-C and inflammation in a septic mouse model. In our study, to examine the in vivo effects of CO using CORM-2 on LPS-induced endotoxemia and TN-Cmediated inflammatory cytokines expression, we pretreated mice with CORM-2 (30 mg/kg, i.p.) or RuCl 3 (30 mg/kg, i.p.) for 2 h and LPS (10 mg/kg, i.p.) for 2 h. Interestingly, CORM-2 significantly decreased TN-C ( Figure 5(a)), TNF-and IL-6 mRNA expression ( Figure 5(b)), and protein secretion ( Figure 5(c)) and simultaneously increased IL-10 expression ( Figure 5(d)) in liver tissue from LPS-induced endotoxemic mice. Also, the levels of IL-10 ( Figure 5(e)) were increased and reversely TN-C levels ( Figure 5(f)) were decreased in the serum of mice treated with CORM-2. Therefore, the results from in vivo experiments suggest that CO inhibited TN-C and its downstream inflammatory cytokines whereby IL-10 expression was upregulated in a septic mice model. A scheme is provided to illustrate each of these results ( Figure 6).

Discussion
TN-C is unique in its distinct pattern of expression. Upon tissue injury TN-C is transiently expressed, whereas its expression is reduced after the tissue is repaired [1]. Moreover, persistent TN-C expression occurs during chronic inflammation [34]. In addition, TN-C is absent in most healthy adult tissues [2] whereas high levels are found during infection and in patients with sepsis. However, TN-C is expressed at sites of inflammation regardless of the location or type of causative  Data represents mean ± SEM, * < 0.05, and * * < 0.001 as compared with control; # < 0.05, ## < 0.001, and ns, nonsignificant, as compared with the cells exposed to LPS alone (normal condition or TN-C/IL-10 siRNA, compared separately).  Figure 5: CO inhibits TN-C-mediated inflammation a septic mouse model. Wild type seven-week-old male C57BL/6 mice were pretreated with CORM-2 (30 mg/kg, i.p.) or RuCl 3 (30 mg/kg, i.p.) for 2 h and then mice were injected with LPS (10 mg/kg, i.p.) for 2 h. Liver tissues were analyzed for mRNA levels of TN-C (a) and TNF-and IL-6 (b) by Real Time RT-PCR. Blood serum was analyzed for protein expression of TNF-, IL-6 (c). IL-10 mRNA (d) by Real Time RT-PCR and IL-10 protein (e) by ELISA were measured. The protein levels of TN-C from serum were measured by Western blotting (f). Loading control was detected by ponceau staining. The relative signal intensity of bands was determined and standardized using ImageJ software. Data represents mean ± SEM, * < 0.05, and * * < 0.001 when compared with control; # < 0.05, ## < 0.001, and ns, nonsignificant, as compared with the cells exposed to LPS alone. insult, indicating its capability to participate in the global inflammatory response.
TN-C can increase the synthesis of cytokines in human chondrocytes [24] and myeloid cells [7] in a TLR4-dependent manner and also activate murine myeloid cells [35,36]. Additionally, TN-C expression is transiently induced by LPS in innate immune cells in a NF-B-dependent manner [7,37] and its dysregulation is observed in both autoimmune and inflammatory diseases, such as sepsis [38]. In the present study, we found that LPS significantly increased TN-C and proinflammatory cytokines production in macrophages as well as in a mouse model. Therefore, understanding which compounds can inhibit TLR-mediated TN-C expression and cytokines production may refine strategies to manipulate excessive inflammation.
Recently, researchers reported that CO gas can exert beneficial effects in various cell and animal models. CO plays an important role in preventing apoptosis in several cell types such as endothelial cells [39], fibroblasts [40], and pancreatic -cells [41] and inhibits the proliferation of smooth muscle cells [42], thus preventing atherosclerotic lesions. In animal models, CO reduced graft rejection [43] and lung inflammation [44].
CORM compounds provide a reliable source of CO that can mimic CO gas in many biological functions [26,45]. Therefore, CORMs represent important tools to understand the biological significance of CO in physiology and disease. CORM-2 was the first compound used to deliver CO in biological systems in a controlled manner [27]. In the current study, to examine the effects of CO, we pretreated macrophages and mice with CORM-2 in an LPS-stimulated inflammation model. Interestingly, CORM-2 significantly decreased LPS-induced TN-C and proinflammatory cytokines production in vitro and in vivo. Similarly, pretreatment with CO gas significantly decreased LPSstimulated TN-C and cytokines production in macrophages, supporting the direct effects of CO on TN-C-mediated inflammation. In addition, RuCl 3, a negative control for CORM-2, did not affect LPS-mediated TN-C, TNF-, and IL-6 expression in macrophages or in septic mice. These results confirm that CO inhibits LPS-mediated TN-C and proinflammatory cytokines production.
The anti-inflammatory cytokine IL-10 plays a crucial role in dampening Toll-like receptor (TLR) signaling-induced proinflammatory genes. Interestingly, CO was found to increase the levels of anti-inflammatory, IL-10 [19], while the levels of proinflammatory cytokines were decreased in several in vitro systems [18,29]. Additionally, CORM-2 was also found to regulate inflammatory responses through decreasing IL-1 expression and increasing IL-10 expression [46]. In a sepsis model, CO-mediated activation of the MKK3/p38 MAPK signaling pathway was involved in the induction of IL-10 [29]. In our investigation, we determined that the effects of CO significantly increased IL-10 expression under LPS-stimulated conditions while RuCl 3 had no effect in vitro or in vivo. Furthermore, IL-10 siRNA significantly reversed the effects of CORM-2 on TN-C and proinflammatory cytokines production whereas TN-C siRNA significantly decreased proinflammatory gene expression levels without having an effect on CO-mediated IL-10 expression. This evidence suggests that CO-mediated IL-10 expression was involved in inhibition of TN-C-mediated inflammation.
In summary, we identified that CO-induced IL-10 was involved in the inhibition of TLR4 signaling-dependent TN-C expression and thus inhibited the inflammatory response in vitro and in vivo. This study describes a novel CO-dependent IL-10 signaling pathway responsible for the inhibition of TN-C-driven inflammation and potentially provides the rationale for novel therapeutic strategies for the treatment of inflammatory diseases.