5,7-Dihydroxyflavone Analogues May Regulate Lipopolysaccharide-Induced Inflammatory Responses by Suppressing IκBα-Linked Akt and ERK5 Phosphorylation in RAW 264.7 Macrophages

We studied the anti-inflammatory activity of twelve 5,7-dihydroxyflavone analogues in lipopolysaccharide- (LPS-) stimulated RAW 264.7 macrophages. We found that chrysin (1) and 4′-methoxytricetin (9) showed relatively significant anti-inflammatory activity and low cytotoxicity. Moreover, 1 and 9 recovered the expression levels of iNOS and COX2, as well as those of the intracellular inflammatory mediators IL-1β and IL-6, which were upregulated by LPS stimulation. In addition, 1 and 9 actively regulated the phosphorylation of IκBα, leading to the activation of NFκB. Phosphorylation of Akt and ERK5 (upstream of NFκB) by LPS stimulation was significantly regulated by 1 and 9, as well as by BIX 02189 and LY 294002, which are phosphorylation inhibitors of ERK5 and Akt, respectively. The results suggest that compounds 1 and 9 may suppress the levels of iNOS and COX2 by regulating phosphorylation of Akt, ERK5, and IκBα and thus NFκB-related signaling pathways, resulting in anti-inflammatory effects in the cells. Because 1 and 9 showed low cytotoxicity and regulated both PGE2 and NO production caused by inflammatory responses, they may hold promise as natural anti-inflammatory agents.


Introduction
Inflammatory responses constitute a biological defense system evoked by various biological events, including pathogenic infections. The major trigger is the recognition of pathogenic components by specific receptors [i.e., Toll-like receptors (TLRs)] of innate immunity, which play crucial roles in the initial promotion of signal transduction [1]. The main function of inflammatory responses is to exclude pathogens. However, responses can induce excessive cellular and tissue injuries due to overreaction by the host immunity, resulting in systemic inflammatory response syndrome. Therefore, the regulation of excessive inflammation is important for the maintenance of homeostasis [2].
The inflammatory responses arise from the humoral and cellular immunity systems [3]. Humoral immunity is associated with inflammatory cytokines (interleukin-(IL-) 1 , IL-6, and TNF-) and other inflammatory mediators, such as prostaglandins [4], while cellular immunity consists of phagocytes such as neutrophils, monocytes, and 2 Evidence-Based Complementary and Alternative Medicine macrophages. These components have additive and/or synergistic effects that can lead to excessive inflammation [5].
Many drugs have been developed to reduce excessive inflammation in vivo. Of these, steroidal or nonsteroidal antiinflammatory drugs are generally used to eliminate acute inflammation. However, these drugs have a variety of side effects and are not suitable for the treatment of some chronic inflammatory diseases [6]. Therefore, it is important to develop other anti-inflammatory compounds with fewer side effects. Traditional medicines, such as Sino-Japanese medicines, frequently use plant extracts to treat chronic inflammatory diseases, including rheumatoid arthritis [7]. Previous studies suggest that these medicines contain active components, that is, polymethoxyflavones, which are not known to cause side effects [8].
Flavonoids are thought to be the active components of unrefined Chinese or Sino-Japanese traditional medicines used to provide anti-inflammatory effects. To date, more than 4000 flavonoids have been identified [9], and many of them are major components of the colors of flowers, fruit, and leaves [10]. Flavonoids occur as aglycones, glycosides, and methylated derivatives. The basic aglycone of flavonoids consists of a benzene ring (A) condensed with a six-membered ring (C) and a phenyl (B) group at the 2position. Flavonoids can be classified into flavonols, flavones, catechins, flavanones, anthocyanidins, and isoflavones.
Since the anti-inflammatory effects of flavonoids are generally gentle, the mechanisms of action and structureactivity relationships have not been defined. However, many reports on the anti-inflammatory effects of isolated or synthesized flavonoids have been published. Indeed, many previous studies suggest that flavonoids have a variety of useful effects such as antitumor, antibacterial, antivirus, antiinflammation, immunoregulation, and antithrombogenicity activities [11][12][13][14][15].
In the present study, we evaluated the anti-inflammatory effects of various 5,7-dihydroxyflavone analogues (12 compounds) on LPS-stimulated RAW 264.7 macrophages to confirm the anti-inflammatory structure-activity relationships of flavonoids [17].

Preparation of 5,7-Dihydroxyflavone Analogues.
A series of 5,7-dihydroxyflavones (2, 3, and 5-12) was prepared by a convenient synthetic method as previously described [17]. To gain insight into the influence of hydroxylation and methylation in the B ring, we synthesized a series of 5,7dihydroxyflavones through the condensation reactions of tert-butyldimethylsilyl-(TBS-) protected acetophenone and methyl benzoates, followed by acid cyclodehydration [18].  The condensation reaction of TBS-protected acetophenone and methyl benzoates was performed using 8 equiv. of lithium bis(trimethylsilyl)amide (LiHMDS) in THF at −78 ∘ C and with a temperature increase to room temperature for 1 day produced intermediates comprising mixtures of tautomers. These intermediates were subjected to acid cyclodehydration and deprotection with 0.5% H 2 SO 4 in acetic acid at 100 ∘ C for 4.5 h. These reaction conditions resulted in the formation of 5,7-dihydroxyflavones, with the chemical structures depicted in Figure 1. Purity of all 5,7-dihydroxyflavones was 98% or more as results of measurement by HPLC.    [22]. In this study, we did not examine the negative control assays because the monoclonal antibodies used in this study showed high specificity [23].

Statistical Analysis.
The results were expressed as mean ± standard deviation (SD). The significant differences between the groups compared were determined by Steel-Dwass test.

Cytotoxicity and Downregulation of NO and PGE 2 Productions by 5,7-Dihydroxyflavone Analogues and Inhibitors.
We examined the effects of twelve 5,7-dihydroxyflavone analogues on cytotoxicity and the inhibition of nitric oxide (NO) and prostaglandin E 2 (PGE 2 ) production on LPSstimulated RAW 264.7 cells ( Figure 2). We also examined five inhibitors including NG-monomethyl-L-arginine (LN), indomethacin (IM), NS-398 (NS), BIX 02189 (BIX), and LY294002 (LY) as the reference compounds to compare the effects. Three of the 5,7-dihydroxyflavone analogues (1, 9, and 10) did not induce a significant amount of cell death even at 100 M. Compounds 1, 4, and 9 inhibited LPSinduced NO production at comparatively low concentrations. In addition, the inhibition of LPS-induced PGE 2 production by compounds 7 and 12 was relatively weak. On the other hand, the reference compounds, LN, IM, NS, BIX, and LY, did not show significant cytotoxic effects ( Figure 2). Although the half maximal inhibitory concentration (IC 50 ) of LN for NO was about 50 M, it did not inhibit the production of PGE 2 . On the contrary, two reference compounds, IM, a nonselective COX inhibitor, and NS, a specific inhibitor of COX2, downregulated the production of PGE 2 but not NO. Moreover, BIX and LY inhibited LPS-induced production of NO and PGE 2 in a dose-dependent manner. Relationships between chemical structures and biological activities could not be confirmed, but the cytotoxic effects of 1 and 9 were relatively low and the inhibitory effects on LPS-induced NO and PGE 2 production by 1 and 9 were stronger than other 5,7-dihydroxyflavone analogues. Thus, compounds 1 and 9 were selected as the representatives from twelve kinds of 5,7-Dihydroxyflavone analogues and effects of compounds 1 and 9 in RAW 264.5 cells were measured in detail hereafter.

Effects of 1, 9, and Inhibitors on Intracellular Levels of Inflammation-Related Proteins.
We also examined the effects of 1, 9, and the inhibitors (LN, IM, and NS) on the expression of inflammatory mediators, TNF-, IL-1 , and IL-6 ( Figure 3). Intracellular levels of TNF-were not affected by LPS stimulation or the addition of the compounds (Figure 3), while intracellular levels of IL-1 and IL-6 were increased by LPS stimulation. IM and NS restored the levels slightly, though these results were not significant; LN had no effect.
On the other hand, the increased intracellular levels of IL-1 and IL-6 were significantly reduced by the addition of 1 or 9 ( < 0.05). It was presumed that the downregulation of LPSinduced NO and PGE 2 was mainly caused by decreases in the expression levels of iNOS and/or COX2 (Figure 3).

Effects of 1, 9, and Inhibitors on NF B-Related Proteins.
It is suggested that LPS activates nuclear factor-kappa B (NF B) through stimulation of TLR4, resulting in expression of iNOS, COXs, and other inflammatory-related cytokines [24].   and I B were significantly induced ( < 0.01) by LPS stimulation without 1, 9, or inhibitors, while NF B phosphorylation remained the same, although the expression levels of the three proteins were not changed by LPS stimulation. The phosphorylation of IKK was not affected by LN, 1, or 9 but was significantly suppressed by IM and NS. On the other hand, the phosphorylation of I B was not affected by LN, IM, or NS but was significantly suppressed by 1 and 9. Moreover, the addition of the other compounds described in Figure 4 had no significant effect on the phosphorylation of NF B.
From these results, it was deduced that 1 and 9 or IM and NS downregulated the inflammatory reactions due to reversal of I B or IKK phosphorylation, respectively.

Effects of 1, 9, and Inhibitors on Intracellular Signal
Transduction-Related Kinases. Some studies on the interactions between inflammation and mitogen-activated protein kinase (MAPK), which exists in the cytoplasm, reported that the phosphorylation of both p38MAPK and Akt is associated with inflammatory reactions [25][26][27][28]. We examined the effects of compounds 1 and 9 and the inhibitors on the phosphorylation of MAPKs and Akt ( Figure 5). LPS stimulation enhanced the phosphorylation of Akt, JNK, p38MAPK, and ERK5 but had no such effect on ERK1/2. The phosphorylation of intracellular signal transduction-related kinases was not influenced by LN, IM, or NS ( Figure 5). Moreover, it was confirmed that 1 and 9 reversed the phosphorylation of Akt and ERK5 induced by LPS stimulation. From these results, it was deduced that inflammatory reactions may be depressed by 1 or 9 via reversal of the phosphorylation of Akt and ERK5 induced by LPS stimulation followed by downregulation of I B phosphorylation.

Effects of Akt or ERK5 Inhibitor on LPS-Induced Inflammatory Reaction.
Though ERK5 has the TEY array, as well as classical ERK1/2 [29], it is not activated by MAPK kinase (MEK1/2) but is specifically activated by MEK5. Previous reports showed that ERK5 is activated by hyperosmosis or oxidative stress and it is recognized as a stress responder MAPK, similar to JNK and p38MAPK [30,31]. However, because it was confirmed that ERK5 can be activated even by trophic factors, such as epidermal growth factor (EGF), nerve growth factor (NGF), and serum [32,33], it is now recognized that it has multiple functions, including those involved in inflammation [29]. Thus, to confirm the anti-inflammatory mechanisms of compounds 1 and 9, we examined the effects of LY294002 (LY), an Akt phosphorylation inhibitor, through the inhibition of PI3 kinase and BIX02188 (BIX), a specific inhibitor of ERK5 phosphorylation in LPS-stimulated RAW 264.7 cells. The results (Figure 2) showed that neither IM nor NS-COX inhibitors-inhibited the production of NO. However, they did downregulate the production of PGE 2 , resulting from LPS stimulation (Figure 2). On the other hand, LN-a competitive inhibitor of iNOS-regulated the production of NO, but not the production of PGE 2 . Compounds 1 and 9 as well as LY and BIX downregulated both NO and PGE 2 production induced by LPS stimulation in a dose-dependent manner. Among the proteins described in Figures 3-5, the effects of BIX or LY on iNOS, COX2, IkB , Akt, and ERK5, which were remarkably recovered by 1 and 9, are shown in Figure 6. BIX and LY downregulated the expression of iNOS and COX2 but there was no significant downregulation of IkB , Akt, or ERK5 phosphorylation induced by LPS stimulation. Moreover, it was suggested that the effects of 1, 9, BIX, or LY were similar because the tendency of reinforcement had been observed by using 1 or 9 with BIX or LY ( Figure 6). Furthermore, because BIX or LY are inhibitors of ERK5 or Akt phosphorylation, respectively, it was assumed that the inhibition of ERK5 or Akt phosphorylation is necessary to achieve the anti-inflammatory effects of 1 and 9.

Discussion
Many reports have shown the anti-inflammatory effects of the extracted and synthesized flavonoids [34]. However, their anti-inflammatory effects are mild [35]. In addition, the relationships between their chemical structures and biological activities are not fully understood. We studied the cytotoxicity and anti-inflammatory effects of twelve 5,7-dihydroxyflavone analogues in inflammatory cells (macrophages). We found that compounds 1 and 9 showed significant anti-inflammatory effects and low cytotoxicity. Moreover, 1 and 9 restored the expression levels of iNOS and  COX2, as well as the intracellular inflammatory mediators, IL-1 and IL-6, which were upregulated by LPS stimulation. It is suggested that the inflammatory response is composed of step-by-step reactions involving inflammatory cascades [36]. Of these cascades, inflammatory responseassociated intracellular signaling pathways involving NF B may play crucial roles in the production of inflammatory cytokines in vivo [37]. The key molecule, NF B, is located downstream of the inflammatory cascades. Another molecule, I B , binds to NF B and regulates NF B activity. Other important signaling pathways, such as those of Akt and ERK5, are located upstream of NF B and these molecules regulate the activation of NF B. Moreover, it is known that Akt regulates the expression of TNF-, NF B, and AP-1 complex [38]. Thus, regulation of these signaling pathways is important in the prevention of excessive inflammatory responses. In the present study, 1 and 9 significantly suppressed the phosphorylation of Akt and ERK5 in LPSstimulated macrophages. In addition, BIX and LY (specific inhibitors of ERK5 or Akt) showed similar effects to 1 and 9 in LPS-stimulated cells. Thus, it can be considered that compounds 1 and 9 significantly suppressed the phosphorylation  Figure 7: The suggested pathway of inflammatory reaction [16] and a tentative mechanism of anti-inflammation by 5,7-dihydroxyflavones. They significantly suppressed the phosphorylation of I B by inhibition of both ERK5 and Akt phosphorylation and inhibition of NF B activity, resulting in the inhibition of iNOS and COX2 production. TLR: toll-like receptor, Ub: ubiquitin, and P: phosphoric acid. of I B by inhibition of both ERK5 and Akt phosphorylation and inhibition of NF B activity resulting in the inhibition of iNOS and COX2 production. A tentative mechanism of anti-inflammatory response by 1 and 9 via downregulation of phosphorylation of ERK5 or Akt was summarized in Figure 7. These processes may be associated with the suppression of inflammatory responses in LPS-stimulated macrophages.
Here, we used some specific inhibitors (LN, IM, and NS) as reference compounds. First, LN is a specific inhibitor of NOS due to competitive binding at the arginine binding site in the enzyme, resulting in inhibition of NO production [39]. Our results showed that LN affected NO production alone. Thus, the effects of the compound were compatible with earlier reports [40]. Second, IM is a nonspecific inhibitor of COXs, while NS selectively inhibits COX2. Thus, both compounds act as anti-inflammatory agents due to the inhibition of PGE 2 production [41][42][43]. In this study, compounds 1 and 9 inhibited both NO and PGE 2 production, while IM and NS inhibited PGE 2 production alone ( Figure 2). Moreover, 1 and 9 inhibited I B phosphorylation, while IM and NS inhibited the phosphorylation of IKK (Figure 4). Thus, the mechanisms of the anti-inflammatory effects between our compounds and the reference compounds are distinct.
It is known that the flavonoid kaempferol inhibits inflammatory reactions via inhibition of Src, Syk, interleukin-1 receptor-associated kinase 1 (IRAK1), and IRAK4 activation, followed by inhibition of the transcriptional activity of NF B and AP-1 [38]. In addition, compound 8 used in this study and PI3K/Akt inhibitor (LY294002) both inhibited the production of TNF- [26]. Moreover, kolaviron (a tetramer of 4) inhibited inflammatory responses, including the production of IL-6, by inhibiting the activation of ERK1/2, NF B, p38MAPK, Akt, phospho-c-Jun, and JNK [44]. Among the 12 analogues, we selected 1 and 9 as candidate anti-inflammatory drugs, because they have low toxicity and strong anti-inflammatory effects on RAW 264 cell macrophages (Figure 2). The results showed that 1 and 9 effectively inhibited the phosphorylation of Akt, ERK5, and I B , resulting in the suppression of iNOS and COX2 expression and the inhibition of NO and PGE 2 . The actions of 1 and 9 in cells under LPS stimulation were similar to those of Akt and ERK5 inhibitors. Thus, we suspect that the antiinflammatory effects of 1 and 9 may be due to the suppression of Akt and/or ERK5 phosphorylation in cells. Further studies are needed to gain a better understanding of the antiinflammatory mechanisms of flavonoids and relationship between Akt or ERK5 and the inflammatory responses.

Conclusion
We found that chrysin (1) and 4 耠 -methoxytricetin (9) showed significant anti-inflammatory activity and low cytotoxicity. Moreover, 1 and 9 recovered the expression levels of iNOS and COX2, as well as those of the intracellular inflammatory mediators IL-1 and IL-6, which were upregulated by LPS stimulation. In addition, 1 and 9 actively regulated the phosphorylation of I B , leading to the activation of NF B. Phosphorylation of Akt and ERK5 (upstream of NF B) by LPS stimulation was significantly regulated by 1 and 9, as well as by BIX 02189 and LY 294002, which are phosphorylation inhibitors of ERK5 and Akt, respectively. The results suggest that compounds 1 and 9 may suppress the levels of iNOS and COX2 by regulating phosphorylation of Akt, ERK5, and I B and thus NF B-related signaling pathways, resulting in antiinflammatory effects in the cells. Because 1 and 9 showed low cytotoxicity and regulated both PGE 2 and NO production caused by inflammatory responses, they may hold promise as natural anti-inflammatory agents.