In vitro and ex vivo antitubercular activity of diarylheptanoids from the rhizomes of Alpinia officinarum Hance

Abstract Phytochemical investigation of methanol extract of the rhizomes of Alpinia officinarum Hance afforded four known diarylheptanoids 1,7-diphenylhept-4-en-3-one (1), 5-hydroxy-1,7-diphenyl-3-heptanone (2), 5-hydroxy-7-(4″-hydroxy-3″-methoxyphenyl)-1-phenyl-3-heptanone (3), and 7-(4″-hydroxy-3″-methoxyphenyl)-1-phenyl heptan-3-one (4).The acetate derivative of (4), 7-(4″-actetate-3″-methoxy phenyl)-1-phenyl heptan-3-one (5), was prepared. These diarylheptanoids exhibited promising in vitro and ex vivo antitubercular activity for the first time against dormant Mycobacterium tuberculosis H37Ra with the IC50 values between 0.34–47.69 and 0.13–22.91 μM, respectively. All compounds showed comparable activity against Mycobacterium bovis BCG (dormant phage) and did not show any activity against two gram + ve and two gram –ve bacterial strains. These compounds were also weakly cytotoxic up to 300 μM against three human cancer cell lines THP-1, Panc-1 and A549.

Crude methanol extract of rhizomes was tested against Mycobacterium tuberculosis (MTB). The promising results from this testing prompted us to further fractionate and isolate the active compounds (1-4) and preparation of acetate derivative (5) of compound (4) (Figure 1). We have presented here the results of bioassay-guided isolation, purification, structure determination by 1-D, 2-D NMR spectroscopy and screening of these five compounds for their antimycobacterial, antibacterial and cytotoxic activity.
The compound (4) (Kiuchi et al. 1992) was obtained as a pale yellow liquid, and the complete spectral data are reported here for the first time. Molecular formula C 20 H 24 O 3 of (4) was determined from MS, 1 H and 13 C NMR data. Its IR spectrum showed absorption at 3600 cm −1 (-OH), 1707 cm −1 (>C=O). Analysis of its 1 H NMR data indicated the presence of six methylenes on heptanoid chain; five aromatic protons showed resonance at δ 7.15-7.27, and other three aromatic methines were observed at δ 6.74, 6.90 and 6.71 towards upfield region due to the substitution of hydroxyl and methoxyl groups. The 13 C NMR data showed similar substitution pattern on the aromatic ring as in (3) which proved the diarylheptanoid type of skeleton, only change was at C-4, 5, 6 positions. The molecular ion peak was observed at 312 [M + ] in GC-MS. On the basis of above spectral data, chemical structure of (4) was established as 7-(4″-hydroxy-3″-methoxyphenyl)-1-phenyl heptan-3-one. The position of the hydroxyl group on aromatic ring was fixed by preparing the acetyl derivative of (4) and analysis of its spectral data.
Acetylation of (4): compound (4, 6 mg) was dissolved in 0.2 mL pyridine and 0.2 mL acetic anhydride, and the mixture was kept overnight at room temperature. Then, the mixture was poured in ice water and extracted with ethyl acetate. The ethyl acetate layer was washed with water 3-4 times and dried over sodium sulphate. The solvent was evaporated under  vacuum on a rotary evaporator, and the acetate product was purified by preparative TLC using ethyl acetate: hexane (1:9) as solvent gradient to give compound (5). Compound (5, 5 mg) was obtained as a yellow liquid and gave molecular ion peak at 354[M + ] in GC-MS which along with 1 H and 13 C NMR data suggested the molecular formula C 22 H 26 O 4 . The major change was observed in IR spectrum where the acetyl group showed absorption at 1762 cm −1 (ester > C=O) instead of the hydroxyl group. The 1 H NMR spectrum of (5) showed signals for 1,3,4-trisubstituted phenyl protons, and 13 C NMR study showed the presence of two benzene rings. The 4″-acetoxy and 3″-methoxy substitution pattern of one of the benzene rings was demonstrated by HMBC and HSQC spectra. Additionally, seven carbons of heptanone were also observed at δc 29.79, 44.32, 210, 42.79, 23.41, 30.91 and 35.72. The location of carbonyl group at C-3 was determined by the analysis of 1 H-1 H COSY and HMBC spectra. In the COSY experiment ( Figure S1), the (H-1) protons at δ 2.92 were coupled with methylene protons at δ 2.74 (H-2), whereas (H-4) protons appeared at δ 2.40 showed the coupling with the protons at δ 1.58 (H-5), and the methylene protons at δ 1.58 (H-6) coupled with the methylene protons at δ 2.56 (H-7). The methylene protons at δ 2.74 (H-2) and δ 2.40 (H-4) showed long-range correlations with the resonance at δ 210.0 (C-3) in the HMBC spectrum ( Figure S1) The HMBC spectrum further showed the correlations of H-1 with C-1′, C-2′, C-6′, C-2 and C-3, of H-2 with C-1′, H-5 with C-7, of H-6 with C-4,7, of H-7 with C-5, 6, 1″, 2″, 6″ of OCH 3 with C-3″ and of H-5″ with C-3″, 4″. Based on the above spectral data (1-D and 2-D), the chemical structure of (5) was established as 7-(4″-actetate-3″-methoxy phenyl)-1-phenyl heptan-3-one.
The crude extracts and compounds (1-5) were evaluated for antimycobacterial activity against MTB. Compounds (1) and (2) Table 1). The in vitro biological testing of (1) against MTB (IC 90 = 4.96 μM), Mycobacterium bovis BCG (IC 90 = 4.16 μM) dormant cells (Singh et al. 2011) and ex vivo testing (Sarkar & Sarkar 2012) using infected model of macrophages THP-1 monocytic cell line (IC 90 = 3.52 μM) showed potential inhibition as seen in Table 1. Despite having higher antimycobacterial activity among compounds (1-5), (1) exhibited lower potencies compared with that of rifampicin and isoniazid, standard anti-TB drugs. Compound (2) inhibited the growth of dormant as well as active phage of H37Ra by ex vivo (IC 90 /IC 50 value of <9 μM < 0.4 μM). Compounds (3-5) were less active against the same strain H37Ra. It may be noted here that the efficiency of inhibiting MTB by crude extract observed in MTB was less as compared to those observed with pure diarylheptanoids.
The selectivity of (1-5) was assayed for their cytotoxic effects in three different cell lines THP-1, PANC-1, A549 using MTT assay (Ciapetti et al. 1993) (Table S5). The IC 90 > 300 μM values of these compounds indicate that the compounds are effective and specific inhibitors against dormant MTB.

Conclusion
Bioassay-guided fractionation, isolation of compounds from the rhizomes of A. officinarum and derivatisation has provided five antitubercular leads possessing promising inhibitory action against MTB for the first time. These compounds were weakly cytotoxic and inactive against bacterial strains. It thus demonstrates the hidden potential of diarylheptanoids specifically as antimycobacterial leads.