Polyphenols with Anti-Inflammatory Properties: Synthesis and Biological Activity of Novel Curcumin Derivatives

Herein, we describe the synthesis and evaluation of anti-inflammatory activities of new curcumin derivatives. The thirteen curcumin derivatives were synthesized by Steglich esterification on one or both of the phenolic rings of curcumin with the aim of providing improved anti-inflammatory activity. Monofunctionalized compounds showed better bioactivity than the difunctionalized derivatives in terms of inhibiting IL-6 production, and known compound 2 presented the highest activity. Additionally, this compound showed strong activity against PGE2. Structure–activity relationship studies were carried out for both IL-6 and PGE2, and it was found that the activity of this series of compounds increases when a free hydroxyl group or aromatic ligands are present on the curcumin ring and a linker moiety is absent. Compound 2 remained the highest activity in modulating IL-6 production and showed strong activity against PGE2 synthesis.

Protecting the reactive sites of curcumin (the aromatic rings and the keto-enol region) (Figure 1) through the formation of derivatives could be an alternative strategy to improve its stability and take advantage of the benefits of this polyphenol [1, 16,17]. Previous studies have shown that the phenolic rings and β-diketone moieties in the curcumin structure suffer from degradation by oxidation and hydrolysis [5,24]. Thus, protection of the hydroxyl groups might increase the bioavailability of this compound by suppressing degradation [22,25,26]. In this sense, some research groups have incorporated amino acids [27], glucose [28], alkyl [7] and succinyl groups into the curcumin structure [5]. In 2011, Wichitnithad et al. demonstrated that succinylation of curcuminoids protects the curcumin from hydrolysis and is an effective strategy against colon cancer [5]. Hence, we added the succinyl group to curcumin structure to protect the new molecules from degradation and to improve their bioavailability. and curcumin glucuronide[1, [14][15][16][17]. Curcumin displays anti-inflammatory effects by modulating several pathways involved in the inflammatory process. This polyphenol inhibits the production of proinflammatory cytokines, such as tumor necrosis factor (TNF)-α and interleukins (ILs) 1, 2, 6, 8, and 12 [18], and regulates the activity of cyclooxygenase 2 (COX-2) [19]. Previous studies have shown that the anti-inflammatory effect induced by curcumin in vitro is achieved by regulating the activation of transcription factors such as activating protein-1 (AP1) and nuclear factor (NF) κB [18][19][20]. Curcumin inhibits NFκB by blocking IκB protein phosphorylation [18][19][20]. The curcumin-mediated downregulation of these intracellular signals leads to a reduction in the expression of cytokines [21]. It has been suggested that curcumin could be used as a nonsteroidal anti-inflammatory drug (NSAID) [22]. However, its low bioavailability due to its susceptibility to degradation in biological systems and poor solubility in water and plasma has prevented the medical use of curcumin [23].
Protecting the reactive sites of curcumin (the aromatic rings and the keto-enol region) (Figure 1) through the formation of derivatives could be an alternative strategy to improve its stability and take advantage of the benefits of this polyphenol [1, 16,17]. Previous studies have shown that the phenolic rings and β-diketone moieties in the curcumin structure suffer from degradation by oxidation and hydrolysis [5,24]. Thus, protection of the hydroxyl groups might increase the bioavailability of this compound by suppressing degradation [22,25,26]. In this sense, some research groups have incorporated amino acids [27], glucose [28], alkyl [7] and succinyl groups into the curcumin structure [5]. In 2011, Wichitnithad et al. demonstrated that succinylation of curcuminoids protects the curcumin from hydrolysis and is an effective strategy against colon cancer [5]. Hence, we added the succinyl group to curcumin structure to protect the new molecules from degradation and to improve their bioavailability. In this research, new curcumin derivatives were synthesized to establish a structureactivity relationship (SAR) between curcumin and its derivatives [19]. Specifically, we focus on protecting the hydroxyl groups of the aromatic ring of curcumin through the incorporation of the succinyl group. We evaluated how the anti-inflammatory activity of the derivatives was altered as a result of the structural changes compared with curcumin.

Synthesis of Novel Curcumin Derivatives
The synthesis of the curcumin derivatives was carried out in two stages. In the first stage, the alkyl succinate derivative was synthetized through the interaction of an alcohol of interest with the succinic anhydride (Supplementary Figure S1). In the second stage, each alkyl succinate derivative (S1-S9) was coupled with curcumin to produce the curcumin derivative (Supplementary Figure S2). Thirteen novel curcumin derivatives (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15) were synthesized to determine their biological activity and establish a structure-activity relationship (SAR). In this research, new curcumin derivatives were synthesized to establish a structure-activity relationship (SAR) between curcumin and its derivatives [19]. Specifically, we focus on protecting the hydroxyl groups of the aromatic ring of curcumin through the incorporation of the succinyl group. We evaluated how the anti-inflammatory activity of the derivatives was altered as a result of the structural changes compared with curcumin.
Compound 1 was obtained commercially from Alfa Aesar with a 95% total curcuminoid content from turmeric rhizome. In this study, compound 1 was used as a reference to compare its activity to that of the novel curcumin derivatives. It is thought that the modification of curcumin, in addition to protecting its reactive sites (Figure 1), provides improvement in activity. We previously described a curcumin derivative, compound 2 ( Figure 2), with a prominent inhibitory effect on IL-6 production [16]. Compound 2 has a different structural modification than the novel curcumin derivatives presented herein. Thus, this compound was also used as a reference to evaluate which modifications are more relevant to the anti-inflammatory activity of curcumin derivatives.

Anti-Inflammatory Activity of the Curcumin Derivatives In Vitro
To evaluate the anti-inflammatory activity of curcumin and its derivatives (2-15), we measured the production of inflammatory mediators by murine macrophages stimulated with LPS in the presence or absence of the compounds. We previously reported a curcumin derivative 2, which preserves the effect of curcumin on IL-6 production [16]. That compound and curcumin (1) were used as references when determining activity in this study. We first evaluated the effect of a single concentration of each compound on the production of IL-6 and TNF-α. Monofunctionalized compounds 4, 6, 10, and 13-15 inhibited the production of IL-6 at 30 µM. Although the effect of compound 8 was not statistically significant, we observed a conserved trend throughout the experiments. Difunctionalized compounds 3, 7, 9, 11, and 12 did not show this activity ( Figure 3B). In our experimental conditions, none of the compounds influenced TNF-α production ( Figure 3A). Compounds 2-13 and 15 were not cytotoxic at this concentration ( Figure 3C). However, under our experimental conditions, 1 and 14 affected cell viability. All monofunctionalized compounds were selected for further experiments.
We then evaluated the effects of the monofunctionalized compounds on IL-6 production at different concentrations. All compounds, including compounds 1 and 2, inhibited the production of IL-6 in a dose-dependent manner ( Figure 4). Greater inhibitory effects were observed at 10 and 30 µM, although compounds 1 and 2 still showed stronger activity. In general, at lower concentrations (1 and 3 µM), the effects of the compounds were similar to that of compound 1. Compound 2 exhibited a statistically significant inhibitory effect even at the lowest concentration (1 µM), which was better than that of the rest of the compounds, including compound 1 (Figure 4). The effect of the compounds was not due to cellular death, as none of the compounds were cytotoxic at any of the evaluated concentrations. compounds influenced TNF-α production ( Figure 3A). Compounds 2-13 and 15 were not cytotoxic at this concentration ( Figure 3C). However, under our experimental conditions, 1 and 14 affected cell viability. All monofunctionalized compounds were selected for further experiments. We then evaluated the effects of the monofunctionalized compounds on IL-6 production at different concentrations. All compounds, including compounds 1 and 2, inhibited the production of IL-6 in a dose-dependent manner (Figure 4). Greater inhibitory effects were observed at 10 and 30 μM, although compounds 1 and 2 still showed stronger activity. In general, at lower concentrations (1 and 3 μM), the effects of the compounds were similar to In addition to IL-6, we evaluated whether the compounds could reduce the production of prostaglandin E 2 (PGE 2 ). PGE 2 is secreted by macrophages that have been exposed to inflammatory stimuli, and its synthesis is influenced by the enzyme COX-2 [29]. It has been shown that curcumin can suppress COX-2 expression and PGE 2 production in models of inflammation [30]. We decided to examine whether these curcumin derivatives, including compound 2, preserved the effect on PGE 2 production. We stimulated macrophages with LPS in the presence or absence of compounds and determined the levels of PGE 2 six hours after stimulus. All monofunctionalized compounds inhibited the production of PGE 2 in a dose-dependent manner, and the effect was statistically significant at concentrations of 10 and 30 µM ( Figure 5). Among these, only compound 15 significantly reduced PGE 2 production at a concentration of 3 µM. Compound 2 also showed an inhibitory effect on PGE 2 production at all concentrations tested. At a concentration of 1 µM, only compound 2 showed an effect ( Figure 5). that of compound 1. Compound 2 exhibited a statistically significant inhibitory effect even at the lowest concentration (1 μM), which was better than that of the rest of the compounds, including compound 1 (Figure 4). The effect of the compounds was not due to cellular death, as none of the compounds were cytotoxic at any of the evaluated concentrations. In addition to IL-6, we evaluated whether the compounds could reduce the production of prostaglandin E2 (PGE2). PGE2 is secreted by macrophages that have been exposed to inflammatory stimuli, and its synthesis is influenced by the enzyme COX-2 [29]. It has been shown that curcumin can suppress COX-2 expression and PGE2 production in models of inflammation [30]. We decided to examine whether these curcumin derivatives, including compound 2, preserved the effect on PGE2 production. We stimulated macrophages with LPS in the presence or absence of compounds and determined the levels of PGE2 six hours after stimulus. All monofunctionalized compounds inhibited the production of PGE2 in a dose-dependent manner, and the effect was statistically significant at concentrations of 10 and 30 μM ( Figure 5). Among these, only compound 15 significantly reduced PGE2 production at a concentration of 3 μM. Compound 2 also showed an inhibitory effect on PGE2 production at all concentrations tested. At a concentration of 1 μM, only compound 2 showed an effect ( Figure 5). The IC50 values representing the effects of all of the compounds are listed in Table 1 and range from 1.94 ± 0.66 μM to 10.6 ± 0.33 μM for the inhibition of IL-6 production and from 0.51 ± 0.08 μM to 5.93 ± 2.29 μM for the effect on PGE2 production.   Table 1 and range from 1.94 ± 0.66 µM to 10.6 ± 0.33 µM for the inhibition of IL-6 production and from 0.51 ± 0.08 µM to 5.93 ± 2.29 µM for the effect on PGE 2 production.

Discussion
Curcumin derivatives were prepared by first considering forming succinate and then esterifying curcumin.
When macrophages are activated, they release a wide variety of mediators, including proinflammatory cytokines such as TNF-α, IL-6, and prostaglandins (such as PGE 2 ). These mediators are usually used as markers of an inflammatory response. We determined the anti-inflammatory effects of curcumin and its derivatives in macrophages stimulated by LPS, an inducer of inflammation. An anti-inflammatory effect was observed after treatment with the monofunctionalized curcumin derivatives with statistically significant dose-dependent inhibition of IL-6 production.
A structure-activity relationship (SAR) evaluation indicated that the difunctionalized compounds lost the anti-inflammatory activity of curcumin, while this property was maintained with the monofunctionalized compounds. Compounds 6, 8, and 10 showed the best activity among the monofunctionalized, with IC 50 values of 4.21 ± 0.73, 1.94 ± 0.66, and 3.60 ± 0.21, respectively. Curcumin has activity comparable to those previously mentioned. Compound 2 exhibited the best activity with an IC 50 of 3.59 ± 0.27. The compounds that were less active were 4, 5, 13, and 15. The structural difference between compound 2 and this new set of curcumin derivatives is the succinate linker between the curcumin structure and the ligand (benzyl alcohol). The addition of this linker to the new derivatives decreased the anti-inflammatory activity compared to compound 2. These results indicate that the overall length of the derivative structure might influence activity [16].
The ligands of compounds 2, 6, 8, and 10 all contain aromatic rings in the benzylic position. Among the compounds with two aromatic rings attaching to the benzylic position, 6 and 10 showed the greater effect on IL-6 production. However, when a single aromatic ring was attached to this position (5, 15), the activity decreased [16].
A common feature of compounds 4, 13, and 15 is an aliphatic six-membered ring, and these compounds showed a lower activity than curcumin, indicating that the six-membered ring attached directly to the keto group of the ligand induces a loss in activity. However, when a six-membered aromatic ring is attached at a benzylic position (8), the activity of the compound increases. These results suggest that there are three key points to increasing anti-inflammatory activity: (i) the phenolic ring of the unesterified curcumin derivatives can form hydrogen bonds with another molecule, affecting the activity of these compounds; (ii) a π-π interaction with the curcumin ligand is possible since compounds 2, 6, 8, and 10 showed high anti-inflammatory activity against IL-6, while compounds 4, 13, and 15 had low activity. (iii) The length of the curcumin ligand can influence its activity, as observed with compound 2 compared to compounds 6, 8, and 10, in which the succinate linker in the latter compounds had an increased length and decreased activity compared to compound 2.
Several studies have shown that nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit LPS-stimulated PGE 2 production [31]. Because curcumin is considered an NSAID, we evaluated the effect of this polyphenol and its derivatives on PGE 2 production. A significant reduction in PGE 2 secretion was observed, suggesting that these curcumin derivatives have a broad effect on the production of proinflammatory mediators associated with the LPS response.
For PGE 2 , structure-activity relationships similar to like those found with IL-6 were observed, as a free hydroxyl group on the aromatic ring of curcumin enhanced activity. Additionally, the length of the molecule remained an influential factor when the aromatic rings in the benzylic position were present. The curcumin derivatives inhibited the production of IL-6 and PGE 2 without affecting TNF-α secretion. NSAIDs suppress the synthesis of PGE 2 by a mechanism that involves regulation of COX-2 activity [31]. Since curcumin derivatives presented herein affect the production of PGE 2 , further studies are necessary to elucidate a mechanism engaged in this effect.
These results lead us to infer that the protection of both sides of the aromatic ring results in the inactivity, seemingly because of degradation. However, protection of one of the rings maintains the activity, and, depending on the type of the substituent, it can enhance it to the point of exceeding the activity of the unprotected molecule, contradicting the above conclusion. Thus, it remains unclear whether it suffers degradation in the unprotected ring. Given the uncertainty, we plan evaluate the mechanism of action of the monofunctionalized compounds in the near future and continue our search for protective groups (succinyl or ether) that are more favorable for prevention of the degradation of the curcumin derivative. At the moment, it can be seen that curcumin has in vitro activity when it is protected on only one side, and that the ether group is the most active.

Cytotoxicity Assay
After the removal of supernatants, 100 µL of MTT (0.5 mg/mL) dissolved in RPMI were added to each well and cells were incubated overnight at 37 • C. The supernatants were removed and formazan crystals were dissolved in 100 µL of 0.04 M HCl in isopropanol. The color was analyzed at 570 nm using an ELISA plate reader. The percent of viable cells was calculated using the formula: % viability: [(OD sample) × 100%]/(OD control). The non-stimulated cells, cultured in medium plus 10% FCS and 0.5% DMSO, represented 100% viability.

Statistical Analysis
Results were analyzed using the statistical software package GraphPad Prism 5. Data are presented as means ± S.E.M. Statistical analysis was performed by Student's t-test. A significant difference between groups was considered when p < 0.05. The half maximal inhibitory concentration (IC 50 ) was calculated by adjusting a sigmoidal dose-response curve following GraphPad Prism5 procedure.

Conclusions
In this study, we synthesized 13 new curcumin derivatives and evaluated their effect on the production of inflammatory mediators such as TNF-α, IL-6 and PGE 2 . A common feature among these derivatives was a succinate linker that was coupled to curcumin via an esterification reaction. The alkoxide group attached to the connector varied, including aliphatic substituents to aromatic rings, where the former included acyclic structures and cyclic rings, and the latter had an increase in the presence of aromatic rings. We added the succinyl group to curcumin structure to protect the new derivatives from degradation and determine how this group affect the anti-inflammatory response. Our results indicate that a free hydroxyl group on the curcumin ring, the absence of a linker, and aromatic ligands all increase the anti-inflammatory activity of this series of compounds. In this investigation, we analyzed two different kinds of curcumin derivatives, the ether group (2) and succinyl group (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15), where compound 2 continues to have the best anti-inflammatory effects of the series that we have studied. Further studies will be needed to describe the mechanism of action of this compound and to evaluate the stability and bioavailability in in vivo systems.