Introduction
Acute local inflammation is a healthy immune response that protects the body from pathogens such as bacteria, viruses, fungi, and other parasites. Dendritic cells and macrophages encountering the pathogens are triggered to release cytokines and chemokines. In acute-phase inflammation, these cytokines and chemokines increase blood flow and vascular permeability along with the accumulation of fluid and leukocytes that are important for an effective defense [
1]. However, in a variety of chronic pathological conditions, pro-inflammatory cytokines are released at high levels and mediate systemic inflammatory responses. These cytokines include IL-1, IL-6, IL-12, IL-23 and TNF-α. For instance, when TNF-α is released locally and physiologically at low levels, it plays beneficial roles in protective immune responses against infectious pathogens by helping neutrophil migration to a site of infection. However, when it is systemically released at high levels, TNF-α can cause severe inflammatory diseases. Clinical administrations of neutralizing antibodies against TNF-α or IL-6 have been successful in patients with rheumatoid arthritis, inflammatory bowel diseases and psoriasis [
2-
5].
Adenosine is a key molecule that regulates numerous physiological processes by activating four G-protein-coupled adenosine receptors (ARs), A1, A2A, A2B and A3 ARs. The nature and magnitude of the effect of adenosine on the cell depend on the extracellular adenosine concentrations, receptor density and the functional characteristics of the intracellular signaling pathways [
6]. Adenosine is a potent endogenous anti-inflammatory and immunoregulatory molecule [
6,
7]. When adenosine is physiologically released from cells at sites of inflammation or tissue injury, it regulates the immune and inflammatory systems and plays a central role in wound healing by increasing angiogenesis through upregulating vascular endothelial growth factor [
6]. It is well accepted that adenosine downregulates the release of pro-inflammatory mediators primarily through A2A AR [
8,
9]. Furthermore, adenosine mediates the anti-inflammatory effects of methotrexate [
10]. Despite these beneficial properties, adenosine has an extremely short half-life because of its rapid metabolism in blood due to conversion by adenosine kinase to adenosine monophosphate (AMP) or its change to inosine by adenosine deaminase. These conversions prevent its clinical usage [
7,
11].
In this regard, the therapeutic benefit of AR-selective agonists is fully appreciated and varieties of selective agonists have been chemically synthesized. Using these synthetic adenosine agonists, clinical trials have been carried out, although some have been withdrawn due to potential side-effects or poor bioavailability [
12,
13].
In addition to the anti-inflammatory action through activation of A2A AR, the three remaining ARs possess some anti-inflammatory activity. For example, when an adenosine A1R agonist was injected 24 h prior to challenge with
E. coli, upregulation of serum pro-inflammatory cytokines and peritoneal leukocyte recruitment were inhibited, thereby reducing the severity of peritonitis [
14]. Moreover, A2B AR-deficient mice failed to induce regulatory T cells after endotoxin-induced pulmonary inflammation and underwent enhanced recruitment of pro-inflammatory effector T cells. Those results suggest that A2B AR serves a potent anti-inflammatory role through the upregulation of regulatory T cells [
15]. It has been shown that methotrexate-induced adenosine downmodulates acute inflammation by activating A2A and/or A3 ARs [
16]. These findings suggest that adenosine, as a physiological agonist, has therapeutic potential for inflammatory disorders through one or more of the ARs, if the drawbacks of short half-life could be overcome.
We have previously shown that royal jelly (RJ) contains low and high molecular weight substances (<5 kDa and >30 kDa, respectively) that can inhibit the secretion of pro-inflammatory cytokines by activated macrophages [
17]. We previously isolated MRJP3 protein as a high molecular weight anti-inflammatory substance [
17]. Using multiple steps of chromatography, we also isolated adenosine N1-Oxide (ANO), an anti-inflammatory substance from the low-molecular weight fraction of RJ. In this study, the anti-inflammatory actions of ANO were examined
in vitro and
in vivo and compared with those of adenosine, synthetic AR-selective agonists, and dipotassium glycyrrhizinate (GK2).
Materials and methods
Mice
BALB/c female mice, aged 8–12 weeks, were purchased from Charles River Japan (Kanagawa, Japan). All animal experiments described in this article were conducted according to the guidelines on Animal Experimentation at the R&D center of Hayashibara Co., LTD.
Reagents
RJ that had been collected from Anhui in China was used. Lipopolysaccharide (LPS) (E. coli 055:B5), adenine, adenosine, zymosan A, Wortmannin and adenosine deaminase were purchased from Sigma-Aldrich Japan (Tokyo, Japan). Adenine N1-Oxide was purchased from MP Biomedicals (Santa Ana, CA). GK2 was purchased from Maruzen Pharmaceuticals (Hiroshima, Japan). Adenosine deaminase inhibitor, erythro-9-(2-hydroxy-3-nonyl) adenine hydrochloride (EHNA) was obtained from Enzo Life Sciences (Farmingdale, NY). Pam3CSK4 was purchased from Bachem (King of Prussia, PA). Poly I:C was purchased from Calbiochem (La Jolla, CA). Murine recombinant interferon-γ (IFN-γ), human TNF-α, and monoclonal antibodies (mAb) for human TNF-α ELISA were prepared and purified in our laboratories. Mouse cytokine standards for ELISA (TNF-α, IL-6, IL-10, and IL-12) were obtained from BD Pharmingen (San Diego, CA). The following pairs of mAbs for ELISA capture and biotinylated detection were purchased from BD Pharmingen: TNF-α, G281-2626 and MP6-XT3; IL-6, MP5-20 F3 and MP5-32C11; IL-10, JES5-2A5 and SXC-1; and, IL-12 p70, 9A5 and C17.8.
A1 AR-selective agonist 2-chloro-N6-cyclopentyladenosine (CCPA), A2A AR- selective agonist 2-[p-(2-carboxyethyl)phenethylamino]-5’-N-ethylcarboxamideadenosine hydrochloride (CGS21680), A3 AR-selective agonist N6-(3-iodobenzyl)adenosine-5’-N-methyluronamide (IB-MECA), and A3 AR-selective antagonist 3-ethyl-5-benzyl-2-methyl-4-phenylethynyl-6-phenyl-1,4-(±)-dihydropyridine-3, 5-dicarboxylate (MRS1911) and adenosine were purchased from Sigma-Aldrich Japan. A1 AR-selective antagonist 8-cyclopentyl-1,3- dipropylxanthine (DPCPX), A2A AR-selective antagonist 4-(2-[7-amino-2-(furan-2-yl)-[1,2,4]triazolo[1, 5-a][1,3,5]triazin-5-ylamino]ethyl) phenol (ZM241385), A2B AR-selective antagonist N-(4-cyanophenyl)-2-[4-(2, 3, 6, 7-tetrahydro-2, 6-dioxo-1, 3-dipropyl-1H-purin-8-yl)phenoxy]-acetamide (MRS1754) were purchased from Tocris Bioscience (Bristol, UK). N-[2-[[3-(4-Bromophenyl)-2-propenyl]amino]ethyl]-5-isoquinolinesulfonamide dihydrochloride (H-89) was purchased from Merck Millipore (Darmstadt, Germany).
Fresh RJ was suspended in 200 mM Tris–HCl buffer, pH 8.0. The supernatants of the RJ suspensions were collected by centrifugation at 10,000×g for 15 min at 4°C. The low-molecular weight fraction of RJ was prepared by centrifugation (2,000×g) using an Ultrafree centrifugal filter device, with a molecular cut-off of 6 kDa. From the low-molecular weight fraction of RJ, ANO was purified to homogeneity by sequential purification on three types of reversed-phase column chromatography: a Vydac 218TP510 column (Grace, Deerfield, IL), a TSKgel ODS-80Ts column (Tosoh, Tokyo, Japan), followed by an YMC-Pack ODA-A column (YMC, Kyoto, Japan). The yield of ANO was 123 ± 25 μg per 1 g of RJ (mean ± S.D., n = 3).
ANO structure was characterized spectroscopically by
1H and
13C NMR and electrospray ionization MS (ESI-MS): ESI-MS m/z 284.15 [M + H]
+;
1H-NMR (dimethylsulphoxide-d
6, 300 MHz) δ ppm: 8.663 (s, 1H), 8.565 (s, 1H), 7.5 - 9.0 (br, 2H), 5.885 (d, J = 5.4, 1H), 5.546 (d, J = 4.8, 1H), 5.240 (d, J = 3.3, 1H), 5.059 (bs, 1H), 4.530 (bd, J = 4.8, 1H), 4.148 (bd, J = 3.0, 1H), 3.946 (bd, J = 3.6, 1H), 3.670 (m, 1H), 3.553 (m, 1H); and
13C-NMR (dimethylsulphoxide-d
6, 300 MHz) δ ppm: 143.28, 142.41, 148.32, 141.92, 118.82, 87.43, 85.49, 73.75, 70.13, 61.15. Chemical shifts in the ribose moiety of ANO were very close to those of adenosine in both
1H-NMR and
13C-NMR analyses. Chemical shifts in the adenine base moiety of ANO showed similar patterns as those of adenosine monophosphate N1-oxide (AMP N1-Oxide) reported previously [
18].
Synthesis of ANO
ANO was prepared according to the procedure described previously [
19]. Adenosine (20 g) was suspended in 1 L of acetic acid, and 100 mL of 30% hydrogen peroxide solution was added to the suspension. The mixtures were stirred for 5 days at room temperature. After decomposing excess amounts of hydrogen peroxide by adding 5 g of 5% palladium on carbon (Kawaken Fine Chemicals, Tokyo, Japan) to the mixture, the palladium on carbon was separated by filtration. The filtrate was desiccated under reduced pressure. ANO crystallized after adding ethanol to the residue and was isolated by filtration. After repeated recrystallization from methanol, 8 g of ANO with 99.1% purity was obtained and used in this study.
Cell cultures and stimulation
Murine peritoneal macrophages were elicited by intraperitoneal injection of 2 mL 4% Brewer’s thioglycollate medium (Nissui Pharmaceutical, Tokyo, Japan) into the peritoneal cavities of BALB/c mice. Peritoneal exudate cells were collected by lavage 3 to 4 days after injection. Cells were washed twice and plated in 10-cm diameter plastic dishes (Nippon Becton Dickinson, Tokyo, Japan) at a density of 1 × 108 cells/dish in 10 mL of RPMI1640 medium (Nissui Pharmaceutical) containing 10% (v/v) FBS (Life Technologies, Grand Island, NY). After 2 h incubation at 37°C in a humidified atmosphere of 5% CO2 and 95% air, non-adherent cells were removed by rinsing. RPMI1640 medium containing 10% FBS was then added to the adherent cells that were recovered by scraping (Nippon Becton Dickinson). The recovered cells were used as macrophages. The murine macrophage-like cell line, RAW264.7 and the human monocytic cell line, THP-1 were maintained in RPMI1640 medium containing 10% FBS. THP-1 cells were cultured with 1 mM sodium butyrate for 4 days before being used in the experiments.
For pro-inflammatory cytokine production, peritoneal macrophages were seeded at 5 × 104 cells per well in flat bottom 96-well microtiter plates and stimulated with LPS (1 μg/mL) with or without murine IFN-γ (muIFN-γ) (10 IU/mL) in the presence or absence of various concentrations of ANO, adenosine, adenine, adenine N1-oxide, or GK2 at 37°C for 24 h. In some experiments, peritoneal macrophages were stimulated with TLR agonists. Sodium butyrate-treated THP-1 cells (1 × 105 cells per well) were stimulated with LPS (5 μg/mL) plus human IFN-γ (huIFN-γ) (500 IU/mL) (termed LPS/huIFN-γ). In other experiments, THP-1 cells were pretreated with AR-selective antagonists 30 min before stimulation with LPS/huIFN-γ in the presence or absence of 10 μM ANO. After 24 h, the culture supernatants were collected for the measurement of cytokines and PGE2. PGE2 was measured using an ELISA kit (Amersham Pharmacia Biotech, Tokyo, Japan). The lower limits of detection were 16, 50, 25, 250, 16 and 50 pg/mL for human TNF-α, murine TNF-α, IL-6, IL-10, IL-12 and PGE2, respectively.
For measurement of cell proliferation, 20 μL alamarBlue dye (Trek Diagnostic Systems, OH), a redox indicator, was added to each microplate well for the last 2 to 3 h of the incubation period. Fluorescence intensity (FI) was measured at 544 nm excitation wavelength and 590 nm emission wavelength.
Stability of ANO in the presence of adenosine deaminase
The stability of ANO and adenosine in culture medium containing 10% FBS at 37°C was examined by quantifying the concentrations at the end points of the different incubation periods by reverse phase high pressure liquid chromatography (HPLC) using an ODS AQ-303 column (YMC). In some experiments, an inhibitor of adenosine deaminase, EHNA, was added together with adenosine. In separate experiments, ANO or adenosine was mixed with 6.7 U/L of adenosine deaminase enzyme in 53.3 mM potassium phosphate buffer containing 0.003% bovine serum albumin, and the mixtures were incubated for up to 60 min at 37°C. The time course of ANO or adenosine conversion was analysed by reverse phase HPLC.
LPS-induced endotoxin shock studies
BALB/c mice were intravenously administered saline or ANO just before intraperitoneal injection of LPS (18 mg/kg) (n = 6 mice in each group). In a separate experiment, ANO was administered orally 1 h before and 1 and 6 h after LPS injection (n = 8 mice in each group). Survival of the mice was monitored for 3 days after LPS injection. In a parallel experiment, blood samples were collected from the abdominal aorta 2 h after the LPS injection. Sera from mice were obtained after centrifugation for 10 min at 4°C and stored at - 80°C until cytokine measurements were performed.
RNA extraction and quantitative real-time polymerase chain reaction (PCR)
Briefly, RAW264.7 cells (1 × 10
6/mL) were stimulated with LPS (2 μg/mL) in the presence or absence of 10 μM ANO in 6-well plates, and incubated for 0.5 to 3 h at 37°C in 5% CO
2. Total RNA was extracted from RAW264.7 cells using an RNeasy Mini kit (QIAGEN, Tokyo, Japan) and DNase (QIAGEN) according to the manufacturer’s instructions. Subsequently, first-strand cDNA was synthesized using Superscript® VILO™ cDNA synthesis kit (Life Technologies, Carlsbad, CA). Specific primers for PCR analysis of
c-fos were identical to those described previously [
20]. PCR assay was carried out with the following sense and antisense primers:
c-fos [GenBank: NM_010234], CGAAGGGAACGGAATAAGAT and GCAACGCAGACTTCTCATC;
Gapdh [GenBank: NM_008084], ACCATCTTCCAGGAGCGAG and AGTGATGGCATGGACTGTGG. Synthesized cDNA was mixed with SYBR Green Master Mix (Roche, Mannheim, Germany) and different sets of gene-specific primers. Real-time PCR was performed using a LightCycler 480 system (Roche). For quantitative purposes, the expression of the
c-fos gene was normalized to a house keeping gene,
Gapdh.
Western immunoblot analysis of cell signaling molecules
RAW264.7 cells were stimulated with LPS (2 μg/mL) in the presence or absence of various concentrations of ANO for 30 min at 37°C. Cellular proteins were prepared in lysis buffer (20 mM Tris–HCl, pH 7.5 containing 0.15 M NaCl, 1 mM EDTA, 0.1% SDS, 1% Triton X-100, 0.5% sodium deoxycholate and inhibitors of proteases and phosphatases). Nuclei and cell debris were removed by centrifugation at 500 g for 10 min. Cell lysates (20 μg) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (Multigel 10/20; COSMO Bio, Tokyo, Japan) under reducing conditions, and transferred to polyvinylidene difluoride (PVDF) membranes (Merck Millipore Japan, Tokyo, Japan) by electrophoretic transfer. PVDF membranes were then blocked with a solution containing 10% Block Ace (Dainippon Pharmaceutical, Osaka, Japan) for 0.5 h, and incubated with anti-c-Fos mAb (9 F6; Cell Signaling Technology, Danvers, MA) for 1 h. After washing the primary mAb, membranes were washed three times with Tris-buffered saline containing 0.05% Tween 20, and then incubated with a 1:1000 dilution of secondary Ab conjugated to horseradish peroxidase (Dako) for 1 h. The membranes were then washed three times and reaction products visualized using the enhanced chemiluminescence Western blot system (Amersham). The same membrane was stripped and reprobed with anti-c-Jun mAb (60A8; Cell Signaling Technology) and anti-β-actin (ACTBD11B7; Santa Cruz Biotechnology, Santa Cruz, CA), and then subjected to Western blotting analysis as described above.
Statistical analysis
Statistical analysis of the data was performed by Dunnett’s test as appropriate. Survival differences were evaluated with the log-rank tests using Kaplan-Meier survival curves. P-values < 0.05 were considered statistically significant.
Discussion
In this study, we have shown that ANO is contained in RJ, which has been widely consumed as a dietary supplement and shows anti-inflammatory actions
in vitro and
in vivo [
17,
28]. RJ contains AMP and its N1-oxide, AMP N1-Oxide, both of which suppressed proliferation of rat pheochromocytoma PC12 cells and stimulated expression of neurofilament M [
18]. It was also reported that RJ contains adenosine [
29]. Using two-dimensional gel electrophoresis, we previously showed that glucose oxidase is present in RJ [
30]. It is well known that glucose oxidase in honey oxidizes glucose to gluconic acid and hydrogen peroxide, the latter of which acts as an anti-microbial agent [
31]. Therefore, it seems likely that ANO and AMP N1-Oxide in RJ are oxidative products of adenosine and AMP, respectively, by hydrogen peroxide generated by glucose oxidase.
In peritoneal macrophages, LPS/muIFN-γ upregulated the production of both TNF-α and IL-6. This process was significantly inhibited by adenine at concentrations ≥10 times of adenosine (Figure
2A and B), although the mechanism of action is unclear. Adenine N1-oxide also significantly reduced pro-inflammatory cytokines. However, the inhibitory action of adenine N1-oxide was comparable or inferior to that of adenine. This outcome contrasts with the superiority of ANO to adenosine in anti-inflammatory functions. These results suggest that the ribose moiety plays an essential role for manifestation of potent inhibitory activity in ANO.
ANO inhibited IL-6 secretion by peritoneal macrophages stimulated by TLR 1/2, TLR 3 and TLR 4 agonists (Figure
1). The effect of AR-selective agonists on cytokine production by human mononuclear cells depends on the TLR subtype [
32]. For instance, CGS 21680 inhibited TLR 4-mediated TNF-α release but potentiated TLR 3- and TLR 5-mediated IL-6 release. In our study, we observed that the effects of AR-selective agonists on cytokine production were different depending on the pro-inflammatory cytokines released even under the same stimulatory conditions. Furthermore, functional differences between species in the anti-inflammatory actions of these agonists were observed. Both CCPA and CGS21680 significantly inhibited TNF-α secretion but potentiated IL-6 and IL-12 secretion by peritoneal macrophages stimulated with LPS/muIFN-γ, while these two agonists inhibited both TNF-α and IL-6 secretion by LPS/huIFN-γ-stimulated THP-1 cells (Figure
5). However, ANO efficiently inhibited all of those pro-inflammatory cytokines under the same stimulatory conditions. These results suggest that the anti-inflammatory effects of ANO are not restricted to a particular pro-inflammatory cytokine or to a situation in which a specific pathogen was involved.
At present, the differences in the response of the two agonists (CCPA and CGS21680) to mouse and human monocytes/macrophages remain unclear. It has been shown that the signaling pathways used by the A2A receptor may vary with the cellular background and the signaling machinery that the cell possesses [
33]. Furthermore, species differences in the order of potency of A2A AR agonists have been reported [
34]. These findings may provide insight for understanding the differential responses observed in our study.
We examined the role of ARs in ANO-mediated inhibition of TNF-α secretion by LPS/huIFN-γ-stimulated THP-1 cells. We found evidence for the involvement of all four ARs, although the extent of recoveries of TNF-α inhibition was different among the four AR-selective antagonists. Although still controversial, there is evidence that the four ARs can be effective in the treatment of inflammatory disorders, including ischemia-reperfusion injuries, endotoxin-induced injuries, arthritis and colitis. The protective effects of A1 AR have been demonstrated in models of ischemia reperfusion- and endotoxin-induced lung injury [
35,
36]. Due to its pre-dominant expression in immune cells and immunoregulatory actions, A2A AR has been shown to have high potential in the treatment of ischemia-, immune-, and inflammation-induced tissue injury [
8,
9]. Although it remains controversial whether A2B AR is pro-inflammatory or anti-inflammatory, recent studies using A2B AR-deficient mice showed that A2B AR played therapeutic roles in endotoxin-induced lung injury [
15,
37]. A3 AR activation inhibited of TNF-α release by endotoxin-stimulated monocyte/macrophage lineage cells [
38] in accordance with our findings using IB-MECA.
A common feature of the above results suggests that activation of the four ARs might exert anti-inflammatory effects in TLR 4-mediated responses. In fact, CCPA, CGS21680 and IB-MECA significantly inhibited TNF-α production by LPS/IFN-γ-stimulated peritoneal macrophages and THP-1 cells (Figure
5). A2B AR-selective agonist with an adenosine-like structure was not available in our study. It is therefore not surprising that ANO-mediated inhibition of TNF-α release was slightly but significantly recovered by the four AR-selective antagonists. However, since the activation of A2B AR and A3 AR elicits a pro-inflammatory response in non-monocyte/macrophage lineage cells [
39,
40], this issue has to be taken into account for clinical application of ANO.
As expected from the
in vitro results, intravenous ANO administration significantly reduced the lethality of LPS-induced endotoxin shock (Figure
7A). In marked contrast and as reported previously [
41], adenosine failed to protect mice against endotoxin-induced mortality in our study, which is probably due to its rapid metabolism
in vivo (data not shown). Reduced lethality in mice was also observed when ANO was administered orally before and after LPS injection (Figure
7B), suggesting that ANO could exert its anti-inflammatory effects systemically through oral administration.
In the LPS-induced endotoxin shock model, serum levels of pro-inflammatory cytokines (TNF-α, IL-6 and IL-12) were reduced but the levels of anti-inflammatory cytokine were upregulated by intravenous administrations of ANO (Figure
8). In a manner similar to that observed by intravenous administration, oral administration of ANO (200 mg/kg) 1 h before intraperitoneal injection of LPS (18 mg/kg) downregulated serum levels of TNF-α and IL-6 and upregulated IL-10 levels in serum when determined 2 h after LPS injection (data not shown). Pro-inflammatory cytokines TNF-α, IL-1 and IL-12, which are derived from monocyte/macrophages, play a key role in the pathogenesis of endotoxin shock [
25]. In turn, anti-inflammatory cytokine IL-10 plays a protective role by inhibiting pro-inflammatory cytokines in LPS-induced endotoxin shock [
42]. Therefore, it is most likely that the decrease in pro-inflammatory cytokines and upregulation of IL-10 explain the reduced lethality achieved by ANO injection.
The difference in the doses of ANO between
in vitro and
in vivo studies requires discussion. IB-MECA prevents lethality in endotoxemic mice when injected at 0.5 mg/kg 30 min before administration of a lethal dose of LPS [
43]. In our study, a dose of 135 mg/kg of ANO was necessary to significantly prolong survival in endotoxemic mice, although ANO was comparable to or more potent than IB-MECA in suppressing the release of pro-inflammatory cytokines
in vitro (Figure
5). ANO was refractory to adenosine-mediated conversion to inosine and therefore was stable in mouse serum (Figure
4 and data not shown). However, serum levels of ANO gradually decreased when ANO was incubated
in vitro in blood taken from normal mice (data not shown). When ANO was added to the blood taken from mice injected with adenosine transport blocker dipyridamole, the decrease in serum levels of ANO was significantly inhibited (data not shown). These results suggest that ANO might be taken up through an adenosine transporter. The uptake of ANO through an adenosine transporter might be one reason why high-dose ANO was necessary in
in vivo experiments. Structural modification of ANO to prevent its uptake by an adenosine transporter might decrease the effective dose
in vivo. Studies are ongoing to address this issue.
Although ANO and adenosine are structurally similar, ANO’s capacity to inhibit pro-inflammatory cytokine production was much more potent than that of adenosine (Figure
2). Refractoriness of ANO to adenosine deaminase might be a plausible explanation for the potent anti-inflammatory activity of ANO (Figure
4B). However, anti-inflammatory activity of adenosine in the presence of EHNA was still inferior to that of ANO (Figure
4C and D), suggesting the existence of another functional difference between ANO and adenosine.
The intracellular signaling pathways induced by ANO and adenosine seem to differ. In LPS-stimulated peritoneal macrophages, ANO-mediated inhibition of TNF-α production was not reversed by H-89, whereas adenosine-mediated inhibition of TNF-α was completely recovered by H-89 (Figure
9). In accordance with our results, the non-selective AR agonist adenosine-5’-N-ethylcarboxamide (NECA) inhibited TNF-α production by LPS-stimulated murine macrophages via a signaling pathway that was independent of PKA and exchange protein activated by cAMP (Epac) [
44].
Another cAMP-mediated suppressive mechanism of LPS-induced pro-inflammatory cytokine production involves the transcription factor c-Fos. c-Fos protein, which is upregulated by cAMP and stabilized following phosphorylation by LPS-activated Ikkβ, physically binds to the p65 subunit of NF-κB. Through this binding, the recruitment of p65:p65 homodimer to the
TNF-α promoter region is reduced, resulting in the suppression of TNF-α production [
27]. Induction of
c-fos mRNA was partially suppressed by the phosphoinositide-3 kinase (PI3K) inhibitor Wortmannin but not by the PI3K inhibitor LY294002, suggesting a role of a Wortmannin-sensitive kinase in the induction of
c-fos mRNA [
27]. The role of c-Fos as an anti-inflammatory transcription factor was reported previously [
45]. Macrophages collected from Fos
−/− mice showed significantly enhanced production of pro-inflammatory cytokines compared with macrophages from wild-type control mice [
45]. In our study, ANO-mediated inhibition of TNF-α production by LPS-stimulated macrophages was partially recovered by Wartmannin (Figure
9). Furthermore, we observed upregulation of
c-fos mRNA and c-Fos protein expression (Figure
10). These results suggest that up-regulation of c-Fos is, at least in part, responsible for ANO-mediated suppression of TNF-α production.
However, it should be noted that cAMP-mediated suppression of IL-6 release was not complete and that recruitment of p65 to the IL-6 promoter region was not reduced but rather enhanced by cAMP treatment [
27], suggesting the presence of another mechanism for ANO-induced suppression of IL-6 secretion. Experiments are ongoing to address this issue.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
KK: determination of anti-inflammatory effects of ANO in vitro, animal experiments, drafting of the manuscript, interpretation and statistical analysis of the data. EO: stability test, animal experiments and Western blotting. OS: purification of ANO from RJ and HPLC analysis. HK: preparation and structural analysis of ANO. TK: animal experiments. NA: RT-PCR analysis. TH: interpretation and statistical analysis of the data. TK: preparation and structural analysis of ANO. TN: preparation of ANO and interpretation of the data. SF: conception of the study and interpretation of the data. All authors read and approved the final manuscript.