Tetrahydrocurcumin Derivatives Enhanced the Anti-Inflammatory Activity of Curcumin: Synthesis, Biological Evaluation, and Structure–Activity Relationship Analysis

Tetrahydrocurcumin, the most abundant curcumin transformation product in biological systems, can potentially be a new alternative therapeutic agent with improved anti-inflammatory activity and higher bioavailability than curcumin. In this article, we describe the synthesis and evaluation of the anti-inflammatory activities of tetrahydrocurcumin derivatives. Eleven tetrahydrocurcumin derivatives were synthesized via Steglich esterification on both sides of the phenolic rings of tetrahydrocurcumin with the aim of improving the anti-inflammatory activity of this compound. We showed that tetrahydrocurcumin (2) inhibited TNF-α and IL-6 production but not PGE2 production. Three tetrahydrocurcumin derivatives inhibited TNF-α production, five inhibited IL-6 production, and three inhibited PGE2 production. The structure–activity relationship analysis suggested that two factors could contribute to the biological activities of these compounds: the presence or absence of planarity and their structural differences. Among the tetrahydrocurcumin derivatives, cyclic compound 13 was the most active in terms of TNF-α production, showing even better activity than tetrahydrocurcumin. Acyclic compound 11 was the most effective in terms of IL-6 production and retained the same effect as tetrahydrocurcumin. Moreover, acyclic compound 12 was the most active in terms of PGE2 production, displaying better inhibition than tetrahydrocurcumin. A 3D-QSAR analysis suggested that the anti-inflammatory activities of tetrahydrocurcumin derivatives could be increased by adding bulky groups at the ends of compounds 2, 11, and 12.

Tetrahydrocurcumin could be a more favorable drug candidate than curcumin because it has multiple properties that curcumin does not have, such as being more stable under physiological conditions, more lipophilic, and more bioactive than curcumin.[17,18,22,23].Tetrahydrocurcumin has the same reactive sites as curcumin (phenolic rings and β-diketone moieties), suggesting that it could also be degraded in biological systems under physiological conditions [23] (Figure 1).We hypothesized that structurally modifying tetrahydrocurcumin would improve the biological activity and increase the bioavailability of this compound.This approach could be an alternative to take advantage of the benefits of curcumin while eliminating its disadvantages.We focused on protecting the phenolic rings of tetrahydrocurcumin by incorporating a succinyl group into its structure [24,25].Previous research has demonstrated that the succinylation of curcuminoids protects these molecules from hydrolysis [12].Therefore, we added a succinyl group to the tetrahydrocurcumin structure to protect it from degradation [12,24,25].In this study, we synthesized new tetrahydrocurcumin derivatives and performed a structure-activity relationship (SAR) analysis.A succinyl group was incorporated to protect the hydroxyl group, forming a diester on the aromatic ring of tetrahydrocurcumin.We evaluated the anti-inflammatory activities of the derivatives for comparison with that of tetrahydrocurcumin.A 3D-QSAR model including a group of compounds synthesized by our research group was developed, with the aim of improving the activity-structure relationship analysis and forecasting the activities of untested compounds.Tetrahydrocurcumin (2) has been isolated from Curcuma wenyujin [16].This polyphenol is also the major curcumin metabolite in biological systems [6,17].Tetrahydrocurcumin has similar biological activities to its precursor, including anti-inflammatory, antioxidant, and anticancer effects [10,[17][18][19][20][21].This polyphenolic compound provides renoprotection against several nephritic disorders by modulating inflammation and oxidative stress [19].Previous research showed that this metabolite ameliorated the inflammatory response by decreasing the expression of the cytokines TNF-α and IL-6 in a mouse model of sepsisinduced acute kidney injury (AKI), and this effect was dependent on silent information regulator sirtuin 1 (SIRT1) signaling [19].Tetrahydrocurcumin also inhibited the LPSinduced release of TNF-α and IL-6 by inhibiting IκB-α degradation [10].In a mouse model of obesity-related inflammatory skin diseases, tetrahydrocurcumin reduced the levels of TNF-α and phosphorylated p65 in the animals' skin [21].
Tetrahydrocurcumin could be a more favorable drug candidate than curcumin because it has multiple properties that curcumin does not have, such as being more stable under physiological conditions, more lipophilic, and more bioactive than curcumin [17,18,22,23].Tetrahydrocurcumin has the same reactive sites as curcumin (phenolic rings and β-diketone moieties), suggesting that it could also be degraded in biological systems under physiological conditions [23] (Figure 1).We hypothesized that structurally modifying tetrahydrocurcumin would improve the biological activity and increase the bioavailability of this compound.This approach could be an alternative to take advantage of the benefits of curcumin while eliminating its disadvantages.We focused on protecting the phenolic rings of tetrahydrocurcumin by incorporating a succinyl group into its structure [24,25].Previous research has demonstrated that the succinylation of curcuminoids protects these molecules from hydrolysis [12].Therefore, we added a succinyl group to the tetrahydrocurcumin structure to protect it from degradation [12,24,25].
In this study, we synthesized new tetrahydrocurcumin derivatives and performed a structure-activity relationship (SAR) analysis.A succinyl group was incorporated to protect the hydroxyl group, forming a diester on the aromatic ring of tetrahydrocurcumin.We evaluated the anti-inflammatory activities of the derivatives for comparison with that of tetrahydrocurcumin.A 3D-QSAR model including a group of compounds synthesized by our research group was developed, with the aim of improving the activity-structure relationship analysis and forecasting the activities of untested compounds.

Synthesis of the Novel Tetrahydrocurcumin Derivatives
The tetrahydrocurcumin derivatives were synthesized in two steps.First, each alkyl succinate derivative was synthesized from the alcohol of interest with succinic anhydride (Figure 2).In our recently published study [6], compounds S1, S3, and S5-S9 (Figure S2) were synthesized on a large scale, and they were used here for coupling with tetrahydrocurcumin (2).Moreover, compounds S2, S4, and S10-S11 were synthesized using our previously described methodology [5].Compounds S1-S11 were synthesized via succinylation reactions (Figure 2).

Synthesis of the Novel Tetrahydrocurcumin Derivatives
The tetrahydrocurcumin derivatives were synthesized in two steps.First, each alkyl succinate derivative was synthesized from the alcohol of interest with succinic anhydride (Figure 2).In our recently published study [6], compounds S1, S3, and S5-S9 (Figure S2) were synthesized on a large scale, and they were used here for coupling with tetrahydrocurcumin (2).Moreover, compounds S2, S4, and S10-S11 were synthesized using our previously described methodology [5].Compounds S1-S11 were synthesized via succinylation reactions (Figure 2).In the second step, each alkyl succinate derivative (S1-S11) was coupled with tetrahydrocurcumin (2) to produce the corresponding tetrahydrocurcumin derivative (Figure 3).Compounds 3-13 were synthesized via Steglich esterification.Eleven novel tetrahydrocurcumin derivatives (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13) were synthesized to determine their biological activities and establish the structure-activity relationship (SAR).The NMR signals of the protons of succinate are around 2.7 ppm and 2.9 ppm, while the carbon signals are around 28 ppm and 29 ppm.However, the signal proton of C-H of the tetrahydrocurcumin derivative is around 5.4 ppm, and the carbon signal is around 99 ppm, pointing to the presence of the Z-enol region, in line with previously observed values for other curcumin derivatives [26].In the second step, each alkyl succinate derivative (S1-S11) was coupled with tetrahydrocurcumin (2) to produce the corresponding tetrahydrocurcumin derivative (Figure 3).Compounds 3-13 were synthesized via Steglich esterification.Eleven novel tetrahydrocurcumin derivatives (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13) were synthesized to determine their biological activities and establish the structure-activity relationship (SAR).The NMR signals of the protons of succinate are around 2.7 ppm and 2.9 ppm, while the carbon signals are around 28 ppm and 29 ppm.However, the signal proton of C-H of the tetrahydrocurcumin derivative is around 5.4 ppm, and the carbon signal is around 99 ppm, pointing to the presence of the Z-enol region, in line with previously observed values for other curcumin derivatives [26].

Synthesis of the Novel Tetrahydrocurcumin Derivatives
The tetrahydrocurcumin derivatives were synthesized in two steps.First, each alkyl succinate derivative was synthesized from the alcohol of interest with succinic anhydride (Figure 2).In our recently published study [6], compounds S1, S3, and S5-S9 (Figure S2) were synthesized on a large scale, and they were used here for coupling with tetrahydrocurcumin (2).Moreover, compounds S2, S4, and S10-S11 were synthesized using our previously described methodology [5].Compounds S1-S11 were synthesized via succinylation reactions (Figure 2).In the second step, each alkyl succinate derivative (S1-S11) was coupled with tetrahydrocurcumin (2) to produce the corresponding tetrahydrocurcumin derivative (Figure 3).Compounds 3-13 were synthesized via Steglich esterification.Eleven novel tetrahydrocurcumin derivatives (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13) were synthesized to determine their biological activities and establish the structure-activity relationship (SAR).The NMR signals of the protons of succinate are around 2.7 ppm and 2.9 ppm, while the carbon signals are around 28 ppm and 29 ppm.However, the signal proton of C-H of the tetrahydrocurcumin derivative is around 5.4 ppm, and the carbon signal is around 99 ppm, pointing to the presence of the Z-enol region, in line with previously observed values for other curcumin derivatives [26].

Anti-Inflammatory Activities of the Tetrahydrocurcumin Derivatives In Vitro
The production of inflammatory mediators induced by LPS in macrophages treated with the compounds was measured to evaluate the anti-inflammatory activities of tetrahydrocurcumin and its derivatives.The effect of a single concentration (30 µM) of each compound on the secretion of TNF-α and IL-6 was evaluated in murine macrophages stimulated with LPS.Tetrahydrocurcumin (2) and derivatives 7-9 and 12-13 inhibited the production of TNF-α (Figure 4A).All compounds, except compound 3, inhibited the production of IL-6 (Figure 4B).Tetrahydrocurcumin derivatives 6, 7, and 9 were cytotoxic (Figure 4C) and were excluded from further analysis.Tetrahydrocurcumin was used as a reference to evaluate which modifications had a greater influence on the anti-inflammatory activity.
The production of inflammatory mediators induced by LPS in macrophages treated with the compounds was measured to evaluate the anti-inflammatory activities of tetrahydrocurcumin and its derivatives.The effect of a single concentration (30 μM) of each compound on the secretion of TNF-α and IL-6 was evaluated in murine macrophages stimulated with LPS.Tetrahydrocurcumin (2) and derivatives 7-9 and 12-13 inhibited the production of TNF-α (Figure 4A).All compounds, except compound 3, inhibited the production of IL-6 (Figure 4B).Tetrahydrocurcumin derivatives 6, 7, and 9 were cytotoxic (Figure 4C) and were excluded from further analysis.Tetrahydrocurcumin was used as a reference to evaluate which modifications had a greater influence on the anti-inflammatory activity.We then evaluated the effects of different concentrations of the selected compounds on TNF-α and IL-6 production (Figure 5).Tetrahydrocurcumin and derivatives 12-13 inhibited the production of TNF-α at all concentrations tested (Figure 5A).Moreover, compounds 4 and 8 inhibited TNF-α at the highest concentration (30 µM), while compound 11 showed an effect only when used at a concentration of 3 µM.Otherwise, compounds 4, 5, 8, 10, 12, and 13 inhibited the production of IL-6 at 30 µM, whereas tetrahydrocurcumin and derivatives 4, 8, and 10-12 inhibited IL-6 at 10 µM (Figure 5B).At a concentration of 3 µM, tetrahydrocurcumin and derivatives 4 and 10-12 also exhibited this effect.At the lowest concentration (1 µM), only tetrahydrocurcumin and derivatives 11 and 12 inhibited the production of this cytokine.compounds 4 and 8 inhibited TNF-α at the highest concentration (30 μM), while compound 11 showed an effect only when used at a concentration of 3 μM.Otherwise, compounds 4, 5, 8, 10, 12, and 13 inhibited the production of IL-6 at 30 μM, whereas tetrahydrocurcumin and derivatives 4, 8, and 10-12 inhibited IL-6 at 10 μM (Figure 5B).At a concentration of 3 μM, tetrahydrocurcumin and derivatives 4 and 10-12 also exhibited this effect.At the lowest concentration (1 μM), only tetrahydrocurcumin and derivatives 11 and 12 inhibited the production of this cytokine.PGE2 is a proinflammatory mediator produced by macrophages in response to LPS, and its synthesis is regulated by the enzyme cyclooxygenase (COX)-2 [27].We evaluated the effect of tetrahydrocurcumin and its previously selected derivatives on the production of PGE2 (Figure 6).The levels of PGE2 were determined in the supernatants of macrophages stimulated with LPS with or without compound treatment (30 μM).Our results showed that compounds 3, 11, and 12 significantly reduced PGE2 production.PGE 2 is a proinflammatory mediator produced by macrophages in response to LPS, and its synthesis is regulated by the enzyme cyclooxygenase (COX)-2 [27].We evaluated the effect of tetrahydrocurcumin and its previously selected derivatives on the production of PGE 2 (Figure 6).The levels of PGE 2 were determined in the supernatants of macrophages stimulated with LPS with or without compound treatment (30 µM).Our results showed that compounds 3, 11, and 12 significantly reduced PGE 2 production.Table 1 shows the IC50 values of the compounds in terms of their effects on TNF-α (ranging from 0.18 ± 0.18 μM to 3.21 ± 4.52 μM) and IL-6 (ranging from 0.17 ± 0.20 μM to 9.13 ± 5.90 μM).Table 1 shows the IC 50 values of the compounds in terms of their effects on TNF-α (ranging from 0.18 ± 0.18 µM to 3.21 ± 4.52 µM) and IL-6 (ranging from 0.17 ± 0.20 µM to 9.13 ± 5.90 µM).Tetrahydrocurcumin was used as a template for the alignment of the compounds present in the database because it is one of the compounds with the best anti-inflammatory activity values.Figure 7a illustrates the alignment of all the compounds.The results obtained from the PLS analysis are summarized in Table 2.For the CoMFA model, first with the model with STD, the leave-one-out cross-validated r 2 value (q 2 ) obtained was 0.424, and the non-cross-validated conventional r 2 value was 0.973.Second, after applying region focusing (RF), the leave-one-out cross-validated r 2 value (q 2 ) obtained was 0.597, and the non-cross-validated conventional r 2 value was 0.987.
For the following explanations and discussions, the RF model will be taken as a reference.Figure 7b shows the linear scattering plots of experimental versus predicted activity for this model, with an r 2 of 0.987 for the model, which is quite good, and a q 2 of 0.597, with four being the optimal number of components.In both cases, the results indicate that the model is dominated by the steric component (65% for the SDT model and The results obtained from the PLS analysis are summarized in Table 2.For the CoMFA model, first with the model with STD, the leave-one-out cross-validated r 2 value (q 2 ) obtained was 0.424, and the non-cross-validated conventional r 2 value was 0.973.Second, after applying region focusing (RF), the leave-one-out cross-validated r 2 value (q 2 ) obtained was 0.597, and the non-cross-validated conventional r 2 value was 0.987.For the following explanations and discussions, the RF model will be taken as a reference.Figure 7b shows the linear scattering plots of experimental versus predicted activity for this model, with an r 2 of 0.987 for the model, which is quite good, and a q 2 of 0.597, with four being the optimal number of components.In both cases, the results indicate that the model is dominated by the steric component (65% for the SDT model and 57% for the RF model).
Tables 3 and S2 in the supporting information show the experimental pIC 50 values versus those predicted by the model and their respective residuals for the compounds in the training set and for the test set, respectively.

Discussion
Macrophages are cells of the innate immune system that are equipped with a variety of membrane receptors and intracellular machinery that can respond to harmful stimuli.These cells secrete a variety of mediators, including the cytokines TNF-α and IL-6 and prostaglandins, such as PGE 2 , when exposed to bacterial components.We evaluated the effect of tetrahydrocurcumin (2) and tetrahydrocurcumin derivatives (3-13) on the production of TNF-α, IL-6, and PGE 2 by macrophages stimulated by the bacterial endotoxin LPS.Eleven tetrahydrocurcumin derivatives (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13) were synthesized by first forming succinate and then esterifying tetrahydrocurcumin.The anti-inflammatory effects of the new tetrahydrocurcumin derivatives were analyzed to identify more active compounds with better structural stability in biological systems.Several tetrahydrocurcumin derivatives showed anti-inflammatory effects, with significant inhibition of TNF-α, IL-6, and PGE 2 production.
Tetrahydrocurcumin was active in terms of both TNF-α and IL-6 production but not PGE 2 production.The structural difference between curcumin and tetrahydrocurcumin is that the former has conjugated double bonds that give this molecule a planar structure, whereas the latter molecule is the hydrogenated form of curcumin, which has free rotation around these bonds.The structural difference between tetrahydrocurcumin and this new set of eleven derivatives is the incorporation of a succinyl group with different substituents (acyclic, cyclic, acyclic aromatic, and cyclic aromatic).
In terms of TNF-α production, tetrahydrocurcumin showed the best activity, with an IC 50 value of 0.18 ± 0.18 µM, followed by compounds 8, 12, and 13, with IC 50 values of 3.21 ± 4.52 µM, 0.70 ± 0.10 µM, and 0.35 ± 0.047 µM, respectively (Table 1).Tetrahydrocurcumin and its difunctionalized derivatives (8, 12, and 13) were active in terms of TNF-α production.The difunctionalized derivatives are more attractive because both sides of the phenolic ring of tetrahydrocurcumin are protected.Although the activities of these new derivatives were lower than that of tetrahydrocurcumin, the protection they provide to both sides of the phenolic ring is relevant for increasing their bioavailability in biological systems.Compound 13, with a bulky substituent, was the most active tetrahydrocurcumin derivative.However, including this same substituent in the difunctionalized curcumin derivative yielded a compound with no effect on TNF-α production [25].
In terms of PGE 2 production, acyclic compounds 3, 11, and 12 showed the highest activity, followed by aromatic compound 4 and cyclic compound 13.We found that compound 12 was more biologically active than tetrahydrocurcumin in terms of PGE 2 production.
Further studies are necessary to evaluate the mechanism of action of the difunctionalized compounds and continue the search for protective groups (with or without succinyl moieties) that are more favorable for improving biological stability and preventing degradation.
The cross-validated q 2 is an important criterion for measuring the internal predictive ability of the 3D-QSAR model.The metrics obtained in the 3D-QSAR model were as follows: the cross-validated q 2 is 0.597 with an ONC value of 4, the non-cross-validated r 2 is 0.987, the SEE value is 0.1077, and the F value is 204.088(Table 2).The validation metrics for the generated model are within the accepted parameters q 2 > 0.5 and r 2 > 0.6, which indicates that the model has predictive potential [28].The validity of the 3D-QSAR model is confirmed by the good predictions of the activity of the compounds belonging to the test set, with residues in six of the seven compounds less than unity (Table S2).
The contributions of the steric and electrostatic fields were 57.3% and 42.7%, respectively; these results indicate that the model is slightly dominated by the steric component.The contour maps for compound 11 are shown in Figure 8 to explain the relationship between the structures and activities of the study compounds.green and yellow display the favorable and unfavorable regions.The red color means that the electronegative groups were beneficial for activity, while the blue color means that the electropositive groups were favorable.
To better explain the generated contour maps, compound 11 will be divided into two sides, A and B, as shown at the top of Figure 8. Around the aromatic rings of compound 11 (pIC50 = 6.77) on both sides of the compound and the methyl of the methoxy group on the side, there is a green contour map indicating that the bulky substituents in this part of the compound lead to higher activity.A yellow contour map is observed around the oxygen atom of the methoxy groups present on both sides of compound 11, indicating that the bulky substituents lead to a decrease in activity.The same behavior is observed in compounds 2 and 12 (Figures S4 and S5).
The blue contour near the position shows that the substitution of the electron-donating group will increase the activity of the compound.For compound 11, a blue contour is observed in the region close to the hydrogens of the methoxy group on the A side of the compound, indicating that the increase in the positive charge of the substitute can help improve activity.The same behavior is observed in compound 1 (Figure S6).The red contours around regions indicate that the addition of an electronegative group can improve the activity.Small red regions can be observed near the hydroxyl groups present on both sides of the compound; therefore, increasing the electronegativity of the substituents in these areas could improve biological activity.

Synthesis
Chemical reagents were received from commercial sources (Tedia (Fairfield, OH, USA), Fischer (Ottawa, ON, Canada), AppliChem (Darmstadt, Germany), and Sigma Aldrich (St. Louis, MO, USA)).Curcumin was obtained commercially from Alfa Aesar with 95% total curcuminoid content from the turmeric rhizome.Tetrahydrocurcumin was obtained commercially from Sigma-Aldrich with 96% purity, as determined by HPLC.All reactions were conducted in borosilicate glass tubes (20 mL or 16 mL) fitted with screw caps and magnetically stirred under an argon atmosphere.Preparative thin-layer chromatography (prep-TLC) was carried out on hard-layer silica-gel-coated green and yellow display the favorable and unfavorable regions.The red color means that the electronegative groups were beneficial for activity, while the blue color means that the electropositive groups were favorable.
To better explain the generated contour maps, compound 11 will be divided into two sides, A and B, as shown at the top of Figure 8. Around the aromatic rings of compound 11 (pIC 50 = 6.77) on both sides of the compound and the methyl of the methoxy group on the side, there is a green contour map indicating that the bulky substituents in this part of the compound lead to higher activity.A yellow contour map is observed around the oxygen atom of the methoxy groups present on both sides of compound 11, indicating that the bulky substituents lead to a decrease in activity.The same behavior is observed in compounds 2 and 12 (Figures S4 and S5).
The blue contour near the position shows that the substitution of the electron-donating group will increase the activity of the compound.For compound 11, a blue contour is observed in the region close to the hydrogens of the methoxy group on the A side of the compound, indicating that the increase in the positive charge of the substitute can help improve activity.The same behavior is observed in compound 1 (Figure S6).The red contours around regions indicate that the addition of an electronegative group can improve the activity.Small red regions can be observed near the hydroxyl groups present on both sides of the compound; therefore, increasing the electronegativity of the substituents in these areas could improve biological activity.
1 H and 13 C NMR spectra were recorded at 500 MHz ( 1 H) and 126 MHz ( 13 C) on a Bruker AVANCE III 500 instrument in CDCl 3 , DMSO-d6, or other specified deuterated solvents with and without tetramethylsilane (TMS) as an internal standard at 25 • C unless specified otherwise.Chemical shifts (δ) are reported in parts per million (ppm) from tetramethylsilane ( 1 H and 13 C).Coupling constants (J) are given in Hz.Proton multiplicity is assigned using the following abbreviations: singlet (s), doublet (d), triplet (t), quartet (q), quintet (quint.),septet (sept.), and multiplet (m); additionally, the abbreviation for broad (br) is used.Infrared measurements were carried out neat on a Bruker Vector 22 FT-IR spectrometer fitted with a Specac diamond attenuated total reflectance (ATR) module.MS analyses were carried out on a Waters Xevo TQD spectrometer with an electrospray ionization (ESI) ion source (Waters Corporation, Milford, MA, USA).HRMS analyses were carried out on a maXis plus ESI-Q-TOF mass spectrometer (Bruker Daltonics, Billerica, MA, USA).The detailed synthetic procedures and spectral characterizations are described below.

Cytotoxicity Assay
After removing the supernatants, 100 µL of 0.5 mg/mL MTT in RPMI was added to each well.Cells were incubated overnight at 37 • C, and formazan crystals were dissolved in 100 µL of 0.04 M HCl in isopropanol.Then, the absorbance was measured at 570 nm using an ELISA plate reader.Cell viability was calculated using the following formula: % viability: [(OD sample) × 100%]/(OD control).Nonstimulated cells represented 100% viability.

Statistical Analysis
Results from the cytokine measurements and cytotoxicity assays were analyzed using the statistical software package GraphPad Prism 9.4.0.All data are presented as the mean ± S.E.M. Statistical analysis was performed by Student's t-test.The differences between samples were considered significant when p < 0.05.The 50% inhibitory concentration (IC 50 ) was calculated by adjusting a sigmoidal dose-response curve following the standard procedure in GraphPad Prism 5.

Three-Dimensional QSAR Model
Twenty-three compounds (Figures 3 and 7) synthesized by our research group with their respective biological activities, reported as IC 50 (Table S1), were used for the development of the 3D-QSAR model.Nine of these compounds are presented in the current work, and fourteen have been reported in previous works [24,25].The 3D structures of all compounds were constructed by using the Gaussview program [29], and their geometry was optimized by using the density functional theory (DFT) with the B3LYP functional [30] and the 6-31G basis set by using the Gaussian program [31].
The optimized structures of the 23 compounds were aligned using compound 2 as a template in the Sybyl X (version 2.1.1,Tripos Inc. (Certara, L.P. St. Louis, MO, USA.2014) program.The total compounds were divided into two sets: the training set consisting of 16 compounds (70% of total compounds) and the test set containing 7 compounds (30% of total compounds) for validating the reliability of the model.In the 3D-QSAR study, the anti-inflammatory activities of the studied compounds were expressed as IC 50 values and converted into pIC 50 values.The comparative molecular field analysis (CoMFA) descriptor was obtained by using the QSAR tool implemented in Sybyl X 2.1.1 [32].The structures of derivatives used in the training and test sets and their anti-inflammatory activities (IC 50 values) are shown in Table S1.
A partial least-squares (PLS) approach [33] was used to derive the 3D-QSAR models, in which the CoMFA descriptors were used as independent variables, and the experimental pIC 50 values were used as dependent variables.Cross-validation with the leaveone-out (LOO) option in the SAMPLS program [34] was applied to obtain the optimal number of components to be used in the final analysis.The q 2 (cross-validated r 2 ), r 2 (non-cross-validated r 2 ), Spress (cross-validated standard error of prediction), and F values were computed.

Conclusions
In this research, we synthesized 11 new tetrahydrocurcumin derivatives and determined their effects on the production of the inflammatory mediators TNF-α, IL-6, and PGE 2 .A common feature among these derivatives was that a succinyl group was attached to tetrahydrocurcumin via a Steglich esterification reaction.The alkoxide group coupled to the connector moiety varied and was composed of acyclic, cyclic, acyclic aromatic, or cyclic aromatic groups.We added a succinyl group to the tetrahydrocurcumin structure to protect the new derivatives from degradation and determine how this group affects the anti-inflammatory response.These difunctionalized tetrahydrocurcumin derivatives showed biological activity against TNF-α, IL-6, and PGE 2 production.An analysis of the contour maps suggests that the anti-inflammatory activity of the newly derived compounds could be increased by adding bulky groups at the ends of compounds 2, 11, and 12.These compounds thus provide key components to build future anti-inflammatory molecules.

Figure 4 .
Figure 4. Anti-inflammatory activities of the tetrahydrocurcumin derivatives.Macrophages from the peritoneal cavities of C57BL/6 mice were treated with 30 μM of each compound.After 1 h, the cells were stimulated with 10 ng/mL LPS.The supernatants were harvested after 6 h of stimulus, and the levels of TNF-α (A) and IL-6 (B) were determined by ELISA.(C) After supernatant collection, cell viability was evaluated by using the MTT assay.The results are presented as the mean ± S.E.M. from two experiments performed in triplicate.*, p < 0.05; **, p < 0.01; ***, p < 0.001 compared to LPS stimulus alone (black bar).C, negative control.

Figure 4 .
Figure 4. Anti-inflammatory activities of the tetrahydrocurcumin derivatives.Macrophages from the peritoneal cavities of C57BL/6 mice were treated with 30 µM of each compound.After 1 h, the cells were stimulated with 10 ng/mL LPS.The supernatants were harvested after 6 h of stimulus, and the levels of TNF-α (A) and IL-6 (B) were determined by ELISA.(C) After supernatant collection, cell viability was evaluated by using the MTT assay.The results are presented as the mean ± S.E.M. from two experiments performed in triplicate.*, p < 0.05; **, p < 0.01; ***, p < 0.001 compared to LPS stimulus alone (black bar).C, negative control.

Figure 6 .
Figure 6.Tetrahydrocurcumin derivatives inhibit the production of PGE2 in macrophages stimulated with LPS.Macrophages from the peritoneal cavities of C57BL/6 mice were treated with 30 μM of each compound and then stimulated with 10 ng/mL LPS.Supernatants were collected after 6 h of stimulus, and levels of PGE2 were determined by ELISA.The results are presented as the mean ± S.E.M. from two experiments performed in triplicate.**, p < 0.01; ***, p < 0.001 compared to LPS stimulus alone (black bar).C, negative control.

Figure 6 .
Figure 6.Tetrahydrocurcumin derivatives inhibit the production of PGE 2 in macrophages stimulated with LPS.Macrophages from the peritoneal cavities of C57BL/6 mice were treated with 30 µM of each compound and then stimulated with 10 ng/mL LPS.Supernatants were collected after 6 h of stimulus, and levels of PGE 2 were determined by ELISA.The results are presented as the mean ± S.E.M. from two experiments performed in triplicate.**, p < 0.01; ***, p < 0.001 compared to LPS stimulus alone (black bar).C, negative control.

Figure 7 .
Figure 7. (a) Structural alignment used to obtain 3D-QSAR model.(b) Experimental pIC50 versus predicted pIC50 values for the 16 compounds in the training set.

Figure 7 .
Figure 7. (a) Structural alignment used to obtain 3D-QSAR model.(b) Experimental pIC 50 versus predicted pIC 50 values for the 16 compounds in the training set.

Figure 8 .
Figure 8. CoMFA contour maps are based on compound 11 as the template.Steric contour maps:green and yellow display the favorable and unfavorable regions.The red color means that the electronegative groups were beneficial for activity, while the blue color means that the electropositive groups were favorable.

Figure 8 .
Figure 8. CoMFA contour maps are based on compound 11 as the template.Steric contour maps:green and yellow display the favorable and unfavorable regions.The red color means that the electronegative groups were beneficial for activity, while the blue color means that the electropositive groups were favorable.

Table 1 .
Anti-inflammatory activities of the novel tetrahydrocurcumin derivatives.

Table 1 .
Anti-inflammatory activities of the novel tetrahydrocurcumin derivatives.
Values represent the average IC 50 values from two independent experiments performed in triplicate ± S.D.

Table 2 .
Statistical data for CoMFA models.

Table 3 .
Experimental and predicted activities (pIC 50 ) of the training set compounds and residual values.