Isopimarane-type diterpenoids from the rhizomes of Kaempferia galanga L. and their biological activities

Abstract Fourteen isopimarane diterpenoids (1–14) were isolated from the rhizomes of Kaempferia galanga, including four new compounds (1–4). The isolated secondary metabolites were identified through analysis of spectroscopic (1 D and 2 D NMR) and mass spectrometric data, together with X-ray diffraction studies. Compounds 4–5, 7–11, and 13 showed strong antimalarial activities, with IC50 values in the range of 1.46–3.99 μg/mL. Moreover, compounds 4, 5, 8, and 12 showed cytotoxicity against KB cell line with IC50 values in the range of 6.13–38.2 μg/mL, while compounds 4, 5, and 12 showed cytotoxicity against MCF-7 cell line with IC50 values in the range of 11.75–47.4 μg/mL. Eventually, the isolated compounds were screened against six bacterial strains and Mycobacterium tuberculosis, demonstrating weak to moderate activities. Graphical Abstract

Compound 1 was isolated as a pale yellow amorphous powder. Its molecular formula was determined as C 22 H 34 O 3 via high-resolution mass spectrometry (HRESIMS) (m/z 369.2408 [M þ Na] þ ). The 13 C NMR spectroscopic data (Table S1) showed 22 carbon signals, including five methyl, seven methylene, four methine, and six quaternary carbons. The 1 H NMR spectroscopic data (Table S1) -19). The 1 D NMR spectroscopic data of compound 1 were similar to those of the known isopimarane diterpenoid, kaempulchraol C (5) , with differences arising from the presence of an acetyl group on the oxygen at C-14. The planar structure of 1 was unequivocally confirmed by detailed interpretation of two-dimensional (2 D) NMR experiments ( Figure S1). The COSY correlations ( Figure S1) showed the presence of four spin systems, H-1/H-2/H-3, H-5/H-6/H-7, H-11/H-12, and H-15/H-16. HMBC correlations ( Figure S1) from H 3 -17 to C-12, C-13, C-14, and C-15, confirmed the location of the oxygenated proton at C-14. The acetate group was located at C-14 due to correlation of H-14 with the acetate's carbonyl. Moreover, the hydroxy group was located at C-6 due to HMBC correlations observed from H-6 to C-4, C-5, C-8, and C-10. The relative configuration of 1 was assigned on the basis of 2 D NOESY experiments. The NOESY correlations ( Figure S2) between H-6a and H-5a/H-7a, and between H-14b and H 3 -17b/H-7b suggested that the orientation of the hydroxy at C-6 should be b and that the acetyl group at C-14 should be a. Compound 1 was therefore identified as 14a-acetoxy-6b-hydroxyisopimara-8(9),15-diene, and was named galangol A.
Compound 2 was obtained as a pale yellow amorphous powder. Its molecular formula was determined as C 22 H 34 O 4 on the basis of its HRESIMS (m/z 385.2352 [M þ Na] þ ) data. Comparison of the 1 H and 13 C NMR data (Table S1) of 2 with those of galangol A (1) suggested that both compounds shared a similar skeleton. Importantly, an extra oxygenated proton was located at position C-1 [d H 3.71 (1H, brt, J ¼ 3.2 Hz, H-1), d C 73.0] due to HMBC correlations ( Figure S1) observed from H 3 -20 to C-1/C-5/C-9 and C-10. The NMR data of compound 2 were also very similar to those of kaempulchraol N . Major differences arose from the presence of an extra acetyl group at C-14 (d H 2.07/d C 21.2 and 171.4). The relative configuration of C-1 was determined by the NOESY correlation ( Figure S2) appearing between H-1b and H 3 -20b, placing the hydroxy group in a-position. Compound 2 was therefore identified as 14aacetoxy-1a,6b-dihydroxyisopimara-8(9),15-diene, and was named galangol B.
Compound 3 was obtained as a pale yellow amorphous powder. Its molecular formula was determined as C 20 H 32 O 4 from its HRESIMS (m/z 359.2193 [M þ Na] þ ) data. Its 1 H and 13 C NMR data (Table S1) suggested the presence of 20 carbons, comprising four methyl, five methylene, six methine, and five quaternary carbons. The 1 H NMR spectroscopic data (Table S1)  . The 1 D NMR spectroscopic data of 3 were very similar to those of the known isopimarane diterpenoid sphaeropsidin F (Evidente et al. 2003), although the chemical shifts of several carbon signals were slightly different. Analysis of 2 D NMR (COSY, HSQC, and HMBC) data ( Figure S1) confirmed that the flat structure of compound 3 is identical to that of sphaeropsidin F, suggesting that a stereoisomer was isolated. The NOESY spectrum ( Figure S2) showed cross peaks between H-1b and H 3 -20b, H-6a and H-5a/H 3 -19a, H-7b and H-14b, as well as H-11b and H 3 -17b, establishing the relative stereochemistry of compound 3. Moreover, the coupling constant value of H-6 and H-7 (J ¼ 2.8 Hz) confirmed a trans equatorial orientation (Boonsombat et al. 2017). Therefore, compound 3 is a new diastereoisomer of sphaeropsidin F, as the absolute configurations at C-1 (S) and C-6 (S) are opposite to those of sphaeropsidin F at these positions. The absolute configuration of compound 3 was determined by comparison of its experimental ECD data with those of compound 10 (Figures S3B), and by comparison of its optical rotation ([a] 25 D ¼ À60 , c 0.1, MeOH) with that of compound 10 ([a] 25 D ¼ À23 , c 0.1, CHCl 3 ). Compound 3 was named galangol C.
Compound 4 was obtained as a pale yellow viscous oil. Its molecular formula was determined as C 20 H 32 O from its HRESIMS (m/z 311.2353 [M þ Na] þ ) data. The 13 C NMR spectroscopic data (Table S1) showed 20 carbon signals, consisting of four methyl, eight methylene, three methine, and five quaternary carbons. The 1 H NMR spectroscopic data (  (Table  S2) suggested that compound 4 is 9a-hydroxysandaracopimaradiene (Wang et al. 2012), which was reported as a hydroxylation product of sandaracopimaradiene. Analysis of the 2 D NMR (COSY, HSQC, and HMBC) data validated this hypothesis ( Figure S1). The absolute configuration of compound 4 was determined by comparison of its experimental ECD data with those of compound 7 ( Figure S4C). Although compound 4 is a known synthetic compound, it is the first report of its isolation from a natural source. Therefore, compound 4 was named galangol D.
The isolated compounds were tested for antimycobacterial activity against Mycobacterium tuberculosis. Most compounds were inactive (MIC >50 lg/mL) compared to two standards (ofloxacin MIC ¼ 0.78 lg/mL, streptomycin MIC ¼ 0.63 lg/mL).
The isolated compounds were then screened for antiplasmodial activities (Table S7). In a previous study, a preliminary biological assay on the crude extract of K. galanga indicated that the dichloromethane extract of the whole plant exhibited activity against Plasmodium falciparum, with an IC 50 value of 26.4 mg/mL (Thongnest et al. 2005). No antiplasmodial assay has been described for the known, isolated compounds, except for compound 11, which showed strong antiplasmodial activity (IC 50 ¼ 3.52 lg/mL, mefloquine IC 50 ¼ 0.011 lg/mL) (Boonsombat et al. 2017). The antiplasmodial activity for the other compounds (4-5, 7-10, and 13) were evaluated here for the first time. They significantly inhibited the growth of P. falciparum, with IC 50 values of 3.61, 3.99, 1.46, 2.79, 3.59, 3.57, and 2.93 lg/mL, respectively. The presence of isopimaranes can therefore be considered to be responsible for the antimalarial activity of this plant.
Eventually, the isolated compounds were tested for antimicrobial activities against a panel of Gram-positive (Staphylococcus aureus, Bacillus subtilis, and Bacillus cereus) and Gram-negative bacteria (Escherichia coli, Pseudomonas aeruginosa, and Shigella sonnei) (Table S8). Only compound 12 showed significant activity against S. aureus, with a MIC value of 16 lg/mL. Compounds 1, 6, and 8 showed moderate antibacterial activities with MIC values in the range of 32-64 lg/mL. Compounds 1, 3-4, 6-9, and 12 showed moderate antibacterial activities against B. cereus and B. subtilis, with MIC values in the range of 32-128 lg/mL. Moreover, compounds 1, 3-4, 6, 8-10, and 12 exhibited moderate antibacterial activities against Ps. aeruginosa, with MIC values in the range of 32-128 lg/mL. In addition, compound 4 showed a weak activity against E. coli, with a MIC value of 64 lg/mL. All compounds showed either weak (MIC value of 128 lg/mL) or no antibacterial activities against S. sonnei. The reference drugs were gentamycin (MIC ¼ 2 lg/mL against Ps. aeruginosa), vancomycin (MIC ¼ 1 lg/mL against B. cereus and S. aureus, MIC ¼ 2 lg/mL against B. subtilis), kanamycin (MIC ¼ 0.5 lg/mL against E. coli), and chloramphenicol (MIC ¼ 0.125 lg/mL against S. sonnei).

General experimental procedures
Optical rotations were measured with a JASCODIP-1000 digital polarimeter (JASCO Inc., Japan). UV and CD spectra were recorded using a JASCO J-810 apparatus. IR analyses were performed using a Bruker Tenser 27 spectrophotometer (Bruker, Germany). NMR spectra were recorded on a Varian Mercury Plus 400 spectrometer (Varian Inc., U.S.A.) using CDCl 3 and CD 3 OD as solvents. The residual peaks of these solvents were used as internal references. The HRESITOFMS were carried out on a Bruker micrOTOF mass spectrometer (Brucker, Germany). Column chromatography was carried out on MERCK silica gel 60 (230-400 mesh) (Merck, Darmstadt, Germany). Thin-layer chromatography was carried out with pre-coated MERCK silica gel 60 PF254 (Merck, Darmstadt, Germany); the spots were visualized under UV light (254 and 365 nm) and further stained by spraying p-anisaldehyde and then heated until charred. Unless otherwise noted, all chemicals were obtained from commercially available sources and were used without further purification.

Plant material
The rhizomes of K. galanga were obtained from Mueang District, Khon Kaen Province, Thailand, in June 2018. The plant was identified by Prof. Dr. Pranom Chantaranothai, Faculty of Science, Khon Kaen University. A voucher specimen (R. LekphromKKU042) was deposited at the herbarium of the Faculty of Science, Khon Kaen University.

Extraction and isolation
The air-dried rhizomes of K. galanga (2.1 kg) were ground to powder and then extracted successively with n-hexane, EtOAc, and MeOH to yield hexane (48.7 g), EtOAc (30.1 g), and MeOH (28.6 g) extracts, respectively.
The crude hexane extract (9.0 g) was chromatographed over silica gel column chromatography (CC), eluting first with pure n-hexane, then with a gradient of n-hexane:EtOAc, and finishing with a gradient of EtOAc:MeOH, in increasing polarity, to give ten fractions, HF 1 -HF 10 . Fraction HF 4 was subjected to silica gel FCC, gradually eluting with n-hexane:EtOAc to give seven fractions, HF 4.1 -HF 4.7 . Subfractions HF 4.4 and HF 4.6 provided compound 7 (1.0 g) as a white powder and compound 8 (225.9 mg) as a yellow solid. Fraction HF 5 was subjected to silica gel CC and gradually eluted with a gradient of n-hexane:EtOAc to give five subfractions, HF 5.1 -HF 5.5 . Fraction HF 8 was subjected to silica gel flash column chromatography (FCC), eluting with a gradient system of n-hexane:EtOAc to give three subtractions, HF 8.1 -HF 8.3 . Subfraction HF 8.1 was chromatographed over silica gel FCC, using a gradient system of n-hexane:EtOAc to give six fractions, HF 8.1.1 -HF 8.1.6 . Fractions HF 8.1.3 and HF 8.1.4 afforded compounds 14 (1.18 g), 10 (20.4 mg), and 11 (69.4 mg) as white amorphous solids. Fraction HF 9 was subjected to silica gel CC and gradually eluted with a gradient of n-hexane:EtOAc to give six subfractions, HF 9.1 -HF 9.6 . Subfractions HF 9.2 and HF 9.5 provided compound 4 (10.0 mg) as a pale yellow viscous oil and compound 3 (20.0 mg) as a pale yellow amorphous powder. Subfraction HF 9.3 was chromatographed over silica gel FCC, eluting with a gradient system of n-hexane:EtOAc to give compound 13 (24.3 mg) as a white powder. Fraction HF 10 was separated by CC, eluting with a gradient system of n-hexane:EtOAc to give five subfractions, HF 10.1 -HF 10.5 . Subfraction HF 10.4 was further separated by CC using n-hexane:EtOAc to obtain compound 6 (20.5 mg) as a white powder.
The EtOAc extract (30.0 g) was separated over silica gel FCC, eluting with gradient systems of n-hexane:EtOAc and then EtOAc:MeOH to give 7 fractions, EF 1 -EF 7 . Fraction EF 4 was subjected to silica gel FCC, gradually eluting with n-hexane:EtOAc to give seven subfractions, EF 4.1 -EF 4.7 . Subfractions EF 4.2 and EF 4.4 provided compound 9 (27.6 mg) as a pale yellow solid. Fraction EF 5 was separated by CC, eluting with a gradient system of CH 2 Cl 2 :EtOAc to give eight subfractions, EF 5.1 -EF 5.8 . Subfraction MF 5.4 was further purified by preparative TLC using CH 2 Cl 2 -EtOAc (95:5) as developing eluent to yield compound 1 (23.2 mg) as a pale yellow solid. Subfraction EF 5.5 was further chromatographed over silica gel, eluting with CH 2 Cl 2 :EtOAc to afford compound 5 (112.3 mg) as a yellow solid. Fraction EF 7 was purified using a gradient system of CH 2 Cl 2 :EtOAc to give two subfractions, FE 7.1 -EF 7.2 . Subfraction EF 7.2 was further purified by CC over silica gel with CH 2 Cl 2 :EtOAc to yield compound 2 (69.0 mg) as a pale yellow amorphous powder. Fraction EF 9 was subjected to silica gel CC and gradually eluted with a gradient of CH 2 Cl 2 :EtOAc to give six subfractions, EF 9.1 -EF 9.6 . Subfraction HF 9.4 provided compound 12 (6.9 mg) as a pale yellow solid and an additional amount of compound 13 (8.0 mg) as a white powder.

Bioassays
3.4.1. Cytotoxicity assays Cytotoxicity assays against human epidermoid carcinoma (KB) and human breast cancer (MFC-7) cell lines were performed employing the colorimetric method described by Skehan (Skehan et al. 1990). The reference substances were ellipticine, tamoxifen, and doxorubicin.

Antimalarial activity assay
Antiplasmodial activity was evaluated against the parasite Plasmodium falciparum (K1, multidrug-resistant strain), using the method of Trager and Jensen (Trager and Jensen 1976). Quantitative assessment of antimalarial activity in vitro was determined by means of the microculture radioisotope technique, based upon the method described by Desjardins et al. (Desjardin et al. 1979). The inhibitory concentration (IC 50 ) represents the concentration which causes 50% reduction in parasite growth as indicated by the in vitro incorporation of [3H]-hypoxanthine by P. falciparum. The standard compound was mefloquine.

Antibacterial assay
The minimum inhibitory concentrations (MICs) were determined by the dilution method as described in the M100-S24 (Clinical and Laboratory Standards Institute 2014). The tests were performed in triplicate. Kanamycin, gentamicine, vancomycin, and chloramphenicol were used as positive control drugs.

Antimycobacterial assay
The antimycobacterial activity was assessed against Mycobacterium tuberculosis H37Ra using the Microplate Alamar Blue Assay (MABA) (Collins and Franzblau 1997). Ofloxacin and streptomycin were used as the reference compounds.

Conclusions
Our phytochemical investigation of the n-hexane and ethyl acetate extracts from the rhizomes of K. galanga led to the isolation of four new compounds, which were named galangols A-D (1-4), and ten known compounds (5-14). Their structures were elucidated by analysis of their spectroscopic data and the absolute configurations of 3-4 were determined by comparison of their experimental ECD data with that of compound 7. The absolute configuration of compound 7 was unambiguously determined by X-ray diffraction of its (S)-naproxen ester. From bioassays, significant antimalarial effect of several isopimarane diterpenoids was demonstrated, as they strongly inhibited the growth of P. falciparum. Moreover, antibacterial and antimycobacterial activities were investigated. However, the isolated compounds showed either moderate, weak (MIC values in the range of 32-128 lg/mL), or no antibacterial and antimycobacterial activities against the selected strains, with the exception of compound 12 on S. Aureus (MIC ¼ 16 lg/mL). Anti-proliferative activities were also investigated on KB and MCF-7 cell lines. Several isolated compounds showed promising anti-proliferative activities.

Disclosure statement
No potential conflict of interest was reported by the authors.