Phytochemical Investigation and Anti-Inflammatory Activity of the Leaves of Machilus japonica var. kusanoi

In a series of anti-inflammatory screenings of lauraceous plants, the methanolic extract of the leaves of Machilus japonica var. kusanoi (Hayata) J.C. Liao showed potent inhibition on both superoxide anion generation and elastase release in human neutrophils. Bioassay-guided fractionation of the leaves of M. japonica var. kusanoi led to the isolation of twenty compounds, including six new butanolides, machinolides A–F (1–6), and fourteen known compounds (7–20). Their structures were characterized by 1D and 2D NMR, UV, IR, CD, and MS data. The absolute configuration of the new compounds were unambiguously confirmed by single-crystal X-ray diffraction analyses (1, 2, and 3) and Mosher’s method (4, 5, and 6). In addition, lignans, (+)-eudesmin (11), (+)-methylpiperitol (12), (+)-pinoresinol (13), and (+)-galbelgin (16) exhibited inhibitory effects on N-formyl-methionyl-leucyl-phenylalanine/cytochalasin B (fMLP/CB)-induced superoxide anion generation in human neutrophils with IC50 values of 8.71 ± 0.74 μM, 2.23 ± 0.92 μM, 6.81 ± 1.07 μM, and 7.15 ± 2.26 μM, respectively. The results revealed the anti-inflammatory potentials of Formosan Machilus japonica var. kusanoi.


Introduction
Neutrophils play an important role in the human body against infections [1]. In response to immune stimulation, activated neutrophils generate a series of cytotoxic substances, such as the superoxide anion (O 2 •− ), a precursor of other ROS, granule proteases, and bioactive lipids. The superoxide anion is known to cause damage to cells and tissues, stimulate macrophages, and trigger a cascade of inflammatory pathways [2]. Neutrophil elastase is one of the serine proteases stored in large amounts in neutrophil granules and is involved in the nonoxidative pathway of the intracellular
Molecules 2020, 25, x FOR PEER REVIEW 3 of 14 enhancement spectroscopy (NOESY) spectrum (H-2 showed correlation with H-3, H-4, and no  correlation with H-5; H-3 showed correlation with H-4 and no correlation with H-5) confirmed that   H-2, H-3, and H-4 were in the same phase ( Figure 3). However, a remaining hydroxy group ( C 71.9) was located at a position of the alkyl chain which cannot be determined by NMR spectrum. Finally, the location of the remaining hydroxy group and the absolute configuration of 1 was further confirmed by single-crystal X-ray diffraction ( Figure 4). The results proved that the stereochemistry of 1 should be shown as 2R,3S,4S,11R-form in the Oak Ridge thermal ellipsoid plot program (ORTEP) diagram. Thus, compound 1 was elucidated and named machinolide A.             Figure 3). The absolute configuration of 2 was confirmed by single-crystal X-ray diffraction and assigned as 2R,3S,4S,11S-form ( Figure 4). According to the above data, the structure of 2 was determined and named machinolide B.
Compound 3 was yielded as colorless needles and assigned the molecular formula C15H28O4 through analysis of its HRESIMS data (m/z 273.20656 [M + H] + (calcd. for 273.20658). All the spectra of 3 were similar to those of 1. However, electron ionization mass spectra (EIMS) showed the different fragments between 3 (m/z 215 (56), 186 (37)) and 1 (m/z 229 (39), 200 (24)), which suggests that the position of the hydroxy group in the alkyl chain was different. The hydroxy group of 3 was located at C-12 and the absolute configuration of 3 was assigned as 2R,3S,4S,12R-form, which were both determined by single-crystal X-ray diffraction ( Figure 4). Therefore, compound 3 was named machinolide C, and its structure was further confirmed by COSY and HMBC experiments ( Figure 2). Compound 4 was obtained as a colorless oil. The ESIMS analysis of 4 showed the [M+H] + ion at m/z 287, in agreement with the molecular formula of C15H26O5, as confirmed by HRESIMS. Compound 4 had similar IR and 1 H-NMR spectra to those of 3, except for the presence of a ketone group at C-11 ( 212.6) in the 13 C-NMR spectrum ( Table 2). The HMBC correlation between H-9, H-10, H-12/C-11, and H-12/C-11, C-13, C-14 supported the position of the ketone group and hydroxy group at C-11 and C-12, respectively ( Figure 2). The planar structure of 4 was decided. The CD spectrum of 4 showed a negative cotton effect at 219.5 nm, which was similar to malleastrumolide A [10]. Thus, the absolute configuration of C-2 was determined as R-form. The NOESY correlations between H-2/H-3, H-2/H-4, and H-3/H-4 confirmed that H-2, H-3, and H-4 were in the same phase ( Figure 3). Hence, the absolute configuration of 4 was determined as 2R,3S,4S-form. Based on the 13 C-NMR-based empirical rules, the chemical shifts of C-3 and C-4 in 4 were similar to those of 2R,3S,4S-form compounds in the literature [11]. According to these two pieces of evidence, the absolute configuration of C-2, C-3, and C-4 in 4 was established to be 2R,3S,4S-form. The absolute configuration of C-12 was determined by Mosher's method [12]. Based on the Δ values of the (S)-MTPA and (R)-MTPA esters in chloroform-d1, the absolute configuration of C-12 was established as  (24)), which suggests that the position of the hydroxy group in the alkyl chain was different. The hydroxy group of 3 was located at C-12 and the absolute configuration of 3 was assigned as 2R,3S,4S,12R-form, which were both determined by single-crystal X-ray diffraction ( Figure 4). Therefore, compound 3 was named machinolide C, and its structure was further confirmed by COSY and HMBC experiments ( Figure 2).
Compound 4 was obtained as a colorless oil. The ESIMS analysis of 4 showed the [M+H] + ion at m/z 287, in agreement with the molecular formula of C 15 H 26 O 5 , as confirmed by HRESIMS. Compound 4 had similar IR and 1 H-NMR spectra to those of 3, except for the presence of a ketone group at C-11 (δ 212.6) in the 13 C-NMR spectrum ( Table 2). The HMBC correlation between H-9, H-10, H-12/C-11, and H-12/C-11, C-13, C-14 supported the position of the ketone group and hydroxy group at C-11 and C-12, respectively ( Figure 2). The planar structure of 4 was decided. The CD spectrum of 4 showed a negative cotton effect at 219.5 nm, which was similar to malleastrumolide A [10]. Thus, the absolute configuration of C-2 was determined as R-form. The NOESY correlations between H-2/H-3, H-2/H-4, and H-3/H-4 confirmed that H-2, H-3, and H-4 were in the same phase ( Figure 3). Hence, the absolute configuration of 4 was determined as 2R,3S,4S-form. Based on the 13 C-NMR-based empirical rules, the chemical shifts of C-3 and C-4 in 4 were similar to those of 2R,3S,4S-form compounds in the literature [11]. According to these two pieces of evidence, the absolute configuration of C-2, C-3, and C-4 in 4 was established to be 2R,3S,4S-form. The absolute configuration of C-12 was determined by Mosher's method [12]. Based on the ∆δ values of the (S)-MTPA and (R)-MTPA esters in chloroform-d 1 , the absolute configuration of C-12 was established as S-form ( Figure 5). Accordingly, the absolute configuration of 4 was defined as 2R,3S,4S,12S. The structure of 4 was confirmed and named machinolide D. S-form ( Figure 5). Accordingly, the absolute configuration of 4 was defined as 2R,3S,4S,12S. The structure of 4 was confirmed and named machinolide D.   (Figure 2). The CD spectrum (a negative cotton effect at 217.5 nm) and NOESY correlation (Figure 3) of 5 were also similar to 4. Moreover, in accordance with the 13 C-NMR-based empirical rules [11], the chemical shifts of C-3 and C-4 in 5 were similar to those of 2R,3S,4S-form compounds in the previous data [11], showing that the absolute configuration of 6 was 2R,3S,4S-form. The absolute configuration of C-12   (Figure 2). The CD spectrum (a negative cotton effect at 217.5 nm) and NOESY correlation (Figure 3) of 5 were also similar to 4. Moreover, in accordance with the 13 C-NMR-based empirical rules [11], the chemical shifts of C-3 and C-4 in 5 were similar to those of 2R,3S,4S-form compounds in the previous data [11], showing that the absolute configuration of 6 was 2R,3S,4S-form. The absolute configuration of C-12  (18), and clovane-2α,9β-diol (19) [24], and one steroid: β-sitosterol (20) [25].

Discussion
Inflammation is triggered by infection or tissue injury. In our series of anti-inflammatory screenings of lauraceous plants, the leaves of M. japonica var. kusanoi stand out as a research candidate. Focusing on the anti-inflammatory activity results in this paper, the lignans, (+)-eudesmin (11), (+)-methylpiperitol (12), (+)-pinoresinol (13), and (+)-galbelgin (16) exhibited inhibitory activities on superoxide anion generation. (+)-Methylpiperitol (12) showed better anti-inflammatory activity than (+)-eudesmin (11), suggesting the methylenedioxy group may enhance the anti-inflammatory activity. (+)-Methylpiperitol (12) exhibited similar anti-inflammatory activity as (+)-pinoresinol (13), indicating the replacement of the methoxy group may not influence anti-inflammatory activity. The results suggested that the furofuran-type lignan containing a methylenedioxy group showed the best anti-inflammatory activity in this study. More importantly, this is the first report on the anti-inflammatory activity of M. japonica var. kusanoi.
Although the potency of the lignans exhibiting anti-inflammatory activity in this study was similar to bioactive lignans described in the literature [42], it is worth noting that most of the lignans with anti-inflammatory activity in this study have not been reported previously. Besides, there are no anti-inflammatory medicines act via inhibiting superoxide anion and neutrophil elastase. The research shows some lead compounds and will help develop novel anti-inflammatory drugs.

Superoxide Anion and Elastase Release Assays
The ability of the test compounds to modulate superoxide anion generation and elastase release by neutrophils was evaluated according to the studies published by co-author Professor Tsong-Long Hwang [2,43]. The superoxide generation assay was based on the reduction of ferricytochrome c by superoxide dismutase (SOD). Elastase substrate (methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide, 100 µM; Merck) was used to detect elastase release. Elastase level was detected at OD405 nm using a spectrophotometer. PI3K inhibitor LY29002 served as a positive control for the neutrophil assays. All assays were repeated at least three times. Results are presented as mean ± standard error of the mean (SEM). The Student's t-test was used to compare the test compound with a DMSO (0.1%) control. A probability of less than 0.05 was considered significant.

Conclusions
Six new butanolides, machinolides A-F (1-6), together with 14 known compounds, were obtained from the leaves of M. japonica var. kusanoi. The absolute configurations of these new compounds were assigned by their CD spectrum, single-crystal X-ray diffraction analyses, and Mosher's method. Hence, absolute configurations of all new compounds were determined as 2R,3S,4S-form in a furan ring, and the chiral center in the side chain group was R-form in 1 and 3, and S-form in 2, 4, 5, and 6. Besides, butanolides and lignans were major skeletons in this study. Bioactivity results indicated that lignans could reduce superoxide anion generation in fMLP/CB-stimulated human neutrophils, and the anti-inflammatory activities of those compounds were as potent as compounds in the literature [42]. Furthermore, the structure-activity relationship (SAR) discussion of anti-inflammatory activity compounds indicated that furofuran lignan with methylenedioxy was the most active structure. To our knowledge, this is the first report on anti-inflammatory activity from the leaves of M. japonica var. kusanoi and the results are helpful to patients with inflammation-related disease.