Sesquiterpenoids and Their Anti-Inflammatory Activity: Evaluation of Ainsliaea yunnanensis

Four new sesquiterpenoids (1–4) and six known sesquiterpenoids (5–10), were isolated from the EtOAc phase of the ethanolic extract of Ainsliaea yunnanensis. Their structures were established by spectroscopic methods, including 1-D, 2-D NMR and HPLC-MS. All compounds were tested for their anti-inflammatory effect by the inhibition of the activity of NLRP3 inflammasome by blocking the self-slicing of pro-caspase-1, which is induced by nigericin, then the secretion of mature IL-1β, mediated by caspase-1, was suppressed. Unfortunately none of the compounds showed an anti-inflammatory effect.


Introduction
The genus Ainsliaea belongs to the family Compositae and comprises around 70 species distributed primarily in southeastern Asia [1]. More than 20 species are used in medicine to treat colds, coughs and asthma, rheumatism, numbness and pain, injury, blood circulatory disorders and hemostasis, enteritis and dysentery, laryngitis, and urinary and gynecological diseases [2]. Modern pharmacological studies have revealed varied biological activities, such as anti-inflammatory [3] and antitumor effects [4]. Ainsliaea yunnanensis is a widely used species in this genus, and is known as "zhui feng jian", "yan mai ling" and "bone arrow" in folk, and is also the common folk herb used for treating various disorders [5].

Results
The EtOH extract of Ainsliaea yunnanensis was partitioned between water and petroleum ether, EtOAc, and normal butanol. The EtOAc phase was concentrated under a vacuum (<50 °C ) and then separated repeatedly by the various chromatographic separation and purification methods to obtain compounds 1-10 ( Figure 1, the detailed data are in the Supplementary Materials).

Structural Elucidation of New Compounds
Compound 1 was obtained as a white amorphous powder. The HR-ESIMS at m/z 269.1382 [M + H] + (calcd. 269.1382) indicated the molecular formula of 1 as C14H20O5. The 1 H-NMR spectrum of 1 in CD3OD displayed signals for one singlet methyl group (δH 0.76), two oxygenated methines (δH 3.79 (1H, dd, J = 5.0, 10.0 Hz); 4.14 (1H, t, J = 10.0 Hz)) and two exocyclic olefinic signals (δH 5.67, 6.23). The 13 C-NMR spectrum showed 14 carbon signals, which were classified as one methyl, five methylenes (one olefinic methylene), four methines (two oxygenated methylenes), and four nonprotonated carbons (one carbonyl, one carboxyl and one olefinic carbon) based on HSQC spectra ( Table 1). All spectroscopic data above in combination with the five degrees of unsaturation required by the molecular formula suggested that 1 was a dual ring containing one double bond, one carbonyl group and one carboxyl group. The 1 H-NMR and 13 C-NMR signals in 1 were very similar to those of the known compound ainsliatone A (5) [13]. However, the molecular formula established by using HR-

Results
The EtOH extract of Ainsliaea yunnanensis was partitioned between water and petroleum ether, EtOAc, and normal butanol. The EtOAc phase was concentrated under a vacuum (<50 • C) and then separated repeatedly by the various chromatographic separation and purification methods to obtain compounds 1-10 ( Figure 1, the detailed data are in the Supplementary Materials).

Structural Elucidation of New Compounds
Compound 1 was obtained as a white amorphous powder. The HR-ESIMS at m/z 269.1382 [M + H] + (calcd. 269.1382) indicated the molecular formula of 1 as C 14 H 20 O 5 . The 1 H-NMR spectrum of 1 in CD 3 OD displayed signals for one singlet methyl group (δ H 0.76), two oxygenated methines (δ H 3.79 (1H, dd, J = 5.0, 10.0 Hz); 4.14 (1H, t, J = 10.0 Hz)) and two exocyclic olefinic signals (δ H 5.67, 6.23). The 13 C-NMR spectrum showed 14 carbon signals, which were classified as one methyl, five methylenes (one olefinic methylene), four methines (two oxygenated methylenes), and four nonprotonated carbons (one carbonyl, one carboxyl and one olefinic carbon) based on HSQC spectra (Table 1). All spectroscopic data above in combination with the five degrees of unsaturation required by the molecular formula suggested that 1 was a dual ring containing one double bond, one carbonyl group and one carboxyl group. The 1 H-NMR and 13 C-NMR signals in 1 were very similar to those of the known compound ainsliatone A (5) [13]. However, the molecular formula established by using HR-ESIMS in compound 1 (C 14 H 20 O 5 ) and 5 (C 14 H 18 O 4 ) are very different. The structure of 1 was finally characterized by the careful analysis of its 2D-NMR spectroscopic data including 1 H-1 H COSY and HMBC (Figure 2). Two structural fragments, as shown with bold lines in Figure 2 (C-1 through C-2 to C-3 and C-5 through C6 up to C-9), were first established by the correlations observed in the 1 H-1 H COSY spectrum. The connectivity of the two structural fragments, one dual ring and the other functional groups were mainly achieved by the analysis of the HMBC spectrum ( Figure 2). A very long-range HMBC correlation from H-5 to C-1, C-3, C-4, C-6, C-7, C-9, and C-10, along with their chemical shifts, not only indicated the structure of the dual ring but also the position of the two hydroxyl groups. The other long correlation from CH 3 -14 to C-1, C-5, C-9 and C-10 located the CH 3 -14 at C-10. In addition, an important correlation from H-7 to C-11, C-12, and C-13 confirmed the structure of the side chain. The planar structure of 1 was, therefore, determined as ainsliatone A acid ( Figure 1).
The relative stereochemistry of 1 was elucidated by the analysis of its NOESY data and compared with that of the known compound ainsliatone A [13] as shown in Figure 2. NOE correlations between H-1 and H-5, and H-5 and H-7 were observed. Thus H-1, H-5, and H-7 adopted the same orientation and were arbitrarily designated as the α-orientation. The large coupling constant between H-5 and H-6 (J H-6/H-5 = 10.0 Hz) suggested a trans configuration and therefore H-6 should be in the β-orientation. The correlation between H-6 and CH 3 -14 suggested the β-orientation of the methyl at C-14. Accordingly, the structure of 1 was established as ainsliatone A acid.
Molecules 2019, 24, x; doi: FOR PEER REVIEW www.mdpi.com/journal/molecules Two structural fragments, as shown with bold lines in Figure 2 (C-1 through C-2 to C-3 and C-5 through C6 up to C-9), were first established by the correlations observed in the 1 H-1 H COSY spectrum. The connectivity of the two structural fragments, one dual ring and the other functional groups were mainly achieved by the analysis of the HMBC spectrum (Figure 2). A very long-range HMBC correlation from H-5 to C-1, C-3, C-4, C-6, C-7, C-9, and C-10, along with their chemical shifts, not only indicated the structure of the dual ring but also the position of the two hydroxyl groups. The other long correlation from CH3-14 to C-1, C-5, C-9 and C-10 located the CH3-14 at C-10. In addition, an important correlation from H-7 to C-11, C-12, and C-13 confirmed the structure of the side chain. The planar structure of 1 was, therefore, determined as ainsliatone A acid (Figure 1).
The relative stereochemistry of 1 was elucidated by the analysis of its NOESY data and compared with that of the known compound ainsliatone A [13] as shown in Figure 2. NOE correlations between H-1 and H-5, and H-5 and H-7 were observed. Thus H-1, H-5, and H-7 adopted the same orientation and were arbitrarily designated as the α-orientation. The large coupling constant between H-5 and H-6 (JH-6/H-5 = 10.0 Hz) suggested a trans configuration and therefore H-6 should be in the β-orientation. The correlation between H-6 and CH3-14 suggested the β-orientation of the methyl at C-14. Accordingly, the structure of 1 was established as ainsliatone A acid. Compound 2 was obtained as a white amorphous powder. The HR-ESIMS at m/z 435.1985 [M + Na] + (calcd. 435.1989) indicated the molecular formula of 2 as C21H32O8. The 1 H-NMR spectrum of 1 in CD3OD displayed the signals for two singlet methyl groups (δH 0.85, 1.58), one cyclic olefinic signal (δH 5.31), two exocyclic olefinic signals (δH 5.58, 6.14) and a signal attributable to the anomeric proton of the glucosyl moiety (δH 4.30). The 13 C-NMR spectrum showed 21 carbon signals, which were classified as four olefinic carbons (δC 120.6, 123.0, 136.2, 148.0), one carboxyl group (δC 170.9), two methyl groups (δC 10.9, 21.0), one oxygenated methylenes (δC 82.2) and one glucose moiety (δC 101.5, 78.2, 77.8, 75.2, 71.9, 63.0) based on HSQC spectra ( Table 1). All spectroscopic data above in combination with six degrees of unsaturation required by the molecular formula suggested that 2 was a dual ring containing two double bonds, one carboxyl group and one glucose. The 1 H-NMR and 13 C-NMR signals in 2 were very similar to eudesmane-type sesquiterpene glycoside [14]. The structure of 2 was finally characterized by the careful analysis of its 2D-NMR spectroscopic data including 1 H-1 H COSY and HMBC (Figure 1).
Two structural fragments as shown with bold lines in Figure 3 (C-2 through C3 to C-4 and C-6 through C7 up to C-10) were first established by the correlations observed in the 1 H-1 H COSY spectrum. The connectivity of the two structural fragments, one dual ring and the other functional groups were mainly achieved by the analysis of the HMBC spectrum (Figure 3). Long-range HMBC correlations from CH3-15 to C-4, C-5, C-6 and C-10, CH3-14 to C-1, C-2 and C-10 combining their chemical shifts not only indicated the structure of the dual ring but also located the CH3-14 at C-1, CH3-15 at C-5, respectively. In addition, two important correlations from H-4 to Glu-1 and C-6, confirmed the position of the glucose, and from H-7 to C-11, C-12, and C-13 confirmed the structure of the side chain. The planar structure of 2 was, therefore, determined as 4-O-(D-glucopyranosyloxyl) -eudesma-1,11(13)-dien-12-oic acid (Figure 1). Compound 2 was obtained as a white amorphous powder. The HR-ESIMS at m/z 435.1985 [M + Na] + (calcd. 435.1989) indicated the molecular formula of 2 as C 21 H 32 O 8 . The 1 H-NMR spectrum of 1 in CD 3 OD displayed the signals for two singlet methyl groups (δ H 0.85, 1.58), one cyclic olefinic signal (δ H 5.31), two exocyclic olefinic signals (δ H 5.58, 6.14) and a signal attributable to the anomeric proton of the glucosyl moiety (δ H 4.30). The 13 C-NMR spectrum showed 21 carbon signals, which were classified as four olefinic carbons (δ C 120.6, 123.0, 136.2, 148.0), one carboxyl group (δ C 170.9), two methyl groups (δ C 10.9, 21.0), one oxygenated methylenes (δ C 82.2) and one glucose moiety (δ C 101.5, 78.2, 77.8, 75.2, 71.9, 63.0) based on HSQC spectra (Table 1). All spectroscopic data above in combination with six degrees of unsaturation required by the molecular formula suggested that 2 was a dual ring containing two double bonds, one carboxyl group and one glucose. The 1 H-NMR and 13 C-NMR signals in 2 were very similar to eudesmane-type sesquiterpene glycoside [14]. The structure of 2 was finally characterized by the careful analysis of its 2D-NMR spectroscopic data including 1 H-1 H COSY and HMBC (Figure 1).
Two structural fragments as shown with bold lines in Figure 3 (C-2 through C3 to C-4 and C-6 through C7 up to C-10) were first established by the correlations observed in the 1 H-1 H COSY spectrum. The connectivity of the two structural fragments, one dual ring and the other functional groups were mainly achieved by the analysis of the HMBC spectrum ( Figure 3). Long-range HMBC correlations from CH 3 -15 to C-4, C-5, C-6 and C-10, CH 3 -14 to C-1, C-2 and C-10 combining their chemical shifts not only indicated the structure of the dual ring but also located the CH 3 -14 at C-1, CH 3 -15 at C-5, respectively. In addition, two important correlations from H-4 to Glu-1 and C-6, confirmed the position of the glucose, and from H-7 to C-11, C-12, and C-13 confirmed the structure of the side chain. The planar structure of 2 was, therefore, determined as 4-O-(d-glucopyranosyloxyl) -eudesma-1,11(13)-dien-12-oic acid (Figure 1).
The relative stereochemistry of 2 was elucidated by the analysis of its NOESY data and compared with the known literature [14] as shown in Figure 3. NOE correlations between H-4 and H-10, and H-10 and H-7 were observed. Thus H-4, H-7, and H-10 adopted the same orientation and were arbitrarily designated as the α-orientation. Acid hydrolysis of 2 afforded D-glucose which was determined by GC-MS of methanolysate and silylated derivatives, and the characteristic coupling constant of the anomeric proton (J = 8.0 Hz) indicated that it was a β-d-glucoside. Accordingly, the structure of 2 was established as 4β-O-(β-d-glucopyranosyloxyl)-eudesma-1,11(13)-dien-12-oic acid, and named alatoside M.
Molecules 2019, 24, x FOR PEER REVIEW 2 of 11 The relative stereochemistry of 2 was elucidated by the analysis of its NOESY data and compared with the known literature [14] as shown in Figure 3. NOE correlations between H-4 and H-10, and H-10 and H-7 were observed. Thus H-4, H-7, and H-10 adopted the same orientation and were arbitrarily designated as the α-orientation. Acid hydrolysis of 2 afforded D-glucose which was determined by GC-MS of methanolysate and silylated derivatives, and the characteristic coupling constant of the anomeric proton (J = 8.0 Hz) indicated that it was a β-D-glucoside. Accordingly, the structure of 2 was established as 4β-O-(β-D-glucopyranosyloxyl)-eudesma-1,11(13)-dien-12-oic acid, and named alatoside M.  13 C-NMR signals in 3 were very similar to those of compound 2 except that 3 had one additional double methyl group (δH 1.13) and a lack of two exocyclic olefinic signals (δH 5.58, 6.14), two olefinic carbons (123.0, 148.0) and one carboxyl group (170.9) ( Table 1). Acid hydrolysis of 3 afforded D-glucose, which was determined by GC-MS of methanolysate and silylated derivatives, and the characteristic coupling constant of the anomeric proton (J = 7.5 Hz) indicated that it was a β-D-glucoside. So the structure of 3 was elucidated as 4β-O-(β-D-glucopyranosyloxyl)-eudesma-7-methyl-1-olefin, and named alatoside N (Figure 1).  Table 1). All spectroscopic data above in combination with seven degrees of unsaturation required by the molecular formula suggested that 4 was three rings containing two double bonds, one carboxyl group and one glucose. The 1 H-NMR and 13 C-NMR signals in 4 were very similar to compound 9, belonging to guaiane-type sesquiterpene glycoside [15]. However, the quantity of the olefinic carbons in 4 (four olefinic carbons) and 9 (six olefinic carbons) is very different. In the 1 H-NMR spectrum of 4, the signals of the exocyclic methylene protons at C-15 were replaced by methyl. An important correlation from CH3-15 to C-3, C-4 and C-5 confirmed the position of the methyl. Acid hydrolysis of 4 afforded D-glucose, which was determined by GC-MS of methanolysate and silylated derivatives, and the characteristic coupling constant of the anomeric proton (J = 7.5 Hz) indicated that it was a β-D-glucoside. The structure of 4 was finally characterized by the careful comparison to spectroscopic data of compound 4 and 9 including 1 H-1 H COSY and HMBC (Figure 1), which was named 3-O-β-D-glucopyranosyloxyl -1α, 5α, 7α-H-10(14), 11(13)-dien-12, 6α-guaiactone (4β, 15-dihydrozaluzaninC) (Figure 1). . The 13 C-NMR spectrum showed 21 carbon signals, which were classified as four olefinic carbons (δ C 112.7, 119.5, 140.1, 149.5), one carboxyl group (δ C 169.6), one methyl group (δ C 18.1), two oxygenated methylenes (δ C 86.1, 86.2) and one glucose moiety (δ C 104.1, 76.8, 76.8, 73.6, 70.2, 61.2) based on HSQC spectra (Table 1). All spectroscopic data above in combination with seven degrees of unsaturation required by the molecular formula suggested that 4 was three rings containing two double bonds, one carboxyl group and one glucose. The 1 H-NMR and 13 C-NMR signals in 4 were very similar to compound 9, belonging to guaiane-type sesquiterpene glycoside [15]. However, the quantity of the olefinic carbons in 4 (four olefinic carbons) and 9 (six olefinic carbons) is very different. In the 1 H-NMR spectrum of 4, the signals of the exocyclic methylene protons at C-15 were replaced by methyl. An important correlation from CH 3 -15 to C-3, C-4 and C-5 confirmed the position of the methyl. Acid hydrolysis of 4 afforded d-glucose, which was determined by GC-MS of methanolysate and silylated derivatives, and the characteristic coupling constant of the anomeric proton (J = 7.5 Hz) indicated that it was a β-d-glucoside. The structure of 4 was finally characterized by the careful comparison to spectroscopic data of compound 4 and 9 including 1 H-1 H COSY and HMBC (Figure 1), which was named 3-O-β-d-glucopyranosyloxyl -1α, 5α, 7α-H-10(14), 11(13)-dien-12, 6α-guaiactone (4β, 15-dihydrozaluzaninC) (Figure 1).

Anti-Inflammatory Activity Assay of Compounds
Compounds 1-10 were tested for their anti-inflammatory effect through the inhibition of the activity of the NLRP3 inflammasome by blocking the self-slicing of pro-caspase-1, which was induced by nigericin, then the secretion of mature IL-1β, mediated by caspase-1, was suppressed. Unfortunately none of the compounds showed any anti-inflammatory effect (Figure 4).

Anti-Inflammatory Activity Assay of Compounds
Compounds 1-10 were tested for their anti-inflammatory effect through the inhibition of the activity of the NLRP3 inflammasome by blocking the self-slicing of pro-caspase-1, which was induced by nigericin, then the secretion of mature IL-1β, mediated by caspase-1, was suppressed. Unfortunately none of the compounds showed any anti-inflammatory effect (Figure 4). Figure 4. LPS-primed iBMDMs treated with 10 μm/L of compounds 1-10, and then stimulated with nigericin. The activity of caspase-1 was analyzed in the supernatant of BMDMs by the Caspase-Glo ® 1 Inflammasome Assay. RLU, recombinant luciferase, which is proportional to caspase-1 activity.

Discussion
As the results showed, none of the compounds showed any anti-inflammatory effect. We believe that the sesquiterpenes are not responsible for anti-inflammatory activity. A key question is whether the fact that the proportion of compounds 1-10 in the ethyl acetate phase was very small had an effect, and whether other potentially active substances were in the remaining material. The second possibility is the non-specificity of the bioassays used. We speculate that perhaps this method is not suitable for this group of compounds. Next, we will use at least one alternative method (an antiinflammatory effect through the inhibition of the activity of NF-κB by blocking the nuclear translocation of p65) to determine anti-inflammatory activity. Following this, we will be able to state with confidence that it is not sesquiterpenes that are responsible for the anti-inflammatory activity, and that other substances should be sought in this fraction (maybe the triterpene fraction) for their anti-inflammatory effect.

Discussion
As the results showed, none of the compounds showed any anti-inflammatory effect. We believe that the sesquiterpenes are not responsible for anti-inflammatory activity. A key question is whether the fact that the proportion of compounds 1-10 in the ethyl acetate phase was very small had an effect, and whether other potentially active substances were in the remaining material. The second possibility is the non-specificity of the bioassays used. We speculate that perhaps this method is not suitable for this group of compounds. Next, we will use at least one alternative method (an anti-inflammatory effect through the inhibition of the activity of NF-κB by blocking the nuclear translocation of p65) to determine anti-inflammatory activity. Following this, we will be able to state with confidence that it is not sesquiterpenes that are responsible for the anti-inflammatory activity, and that other substances should be sought in this fraction (maybe the triterpene fraction) for their anti-inflammatory effect.

Experimental Material
The whole herb of Ainsliaea yunnanensis was collected from Chuxiong city, Yunan province, China. It was identified by Lu Jin Mei, Associate Researcher at the Kunming Institute of Botany, Chinese Academy of Sciences, Yunnan, China. The specimen had been deposited at Beijing Union University, and the Beijing Key Laboratory of Bioactive Substances and Functional Foods, Beijing, China.