INTRODUCTION

Pierre ex A. Froehner, known as Robusta coffee, is one of five coffee species grown popularly in Vietnam1 and contributes significantly to the Vietnamese economy. Coffee beans and leaves have been investigated for chemistry and bioactivity aspects. Oestreich-Janzen’s study reported the presence of lipids, terpenoids, steroids, caffeic acid, chlorogenic acids, trigonelline, caffeine, and alkaloids in Metabolites isolated from coffee exhibit a reduced risk of Alzheimer’s3 and Parkinson’s diseases4 and anti-inflammatory5 and antioxidant activities6. In particular, the benefits of regular coffee consumption on the reduced risk of type 2 diabetes have been reported7. Previously, we reported the isolation, structural analysis and -glucosidase inhibitory activity of triterpenoids and diterpenoids from the trunk of In this article, two alkaloids and four phenolic glycosides were isolated from trunk, and their -glucosidase inhibitory ability was evaluated.

MATERIALS AND Methods

RESULTS

From the -hexane and ethyl acetate extracts prepared from the trunk of grown in Lam Dong Province, Vietnam, six compounds, including 1 (10 mg), 2 (4.2 mg), 3 (23 mg), 4 (4.6 mg), 5 (6.1 mg), and 6 (5.4 mg), were isolated by chromatographic methods. The spectroscopic data of compounds 1-4 are displayed as follow, while the data of compounds 5 and 6 are shown in Table 2. The results of the -glucosidase inhibitory assay applied to several isolated compounds are presented in Table 1.

Caffeine(1): White amorphous powder. ESI-MS: 194.8 [M+H] (calcd. for CHNO+H, 195.1). H-NMR data (CDCl, 500 MHz) (, in Hertz): 7.51 (1H, , H-8), 4.00 (3H, , 7-CH), 3.59 (3H, , 3-CH), 3.41 (3H, , 1-CH). C NMR data (CDCl, 125 MHz): δ 155.4 (C-6), 151.7 (C-2), 148.7 (C-4), 141.4 (C-8), 107.6 (C-5), 33.6 (7-CH), 29.7 (3-CH), 27.9 (1-CH). Selected HMBC correlations are depicted in Figure 1.

Uridine(2): Yellowish amorphous powder. ESI-MS: 244.7 [M+H] (calcd. for CHNO+H, 245.1). H-NMR data (CDOD, 500 MHz) (, in Hertz): 7.99 (1H, , 8.0, H-6), 5.89 (1H, , 4.5, H-1'), 5.69 (1H, , 8.0, H-5), 4.17 (1H, , 10.0, 5.5, H-2'), 4.15 (1H, , 10.0, 5.0, H-3'), 4.00 (1H, , H-4'), 3.84 (1H, , 12.0, 2.5, H-5'a), 3.73 (1H, , 12.0, 3.0, H-5' ). C NMR data (CDOD, 125 MHz): δ 166.2 (C-4), 152.5 (C-2), 142.7 (C-6), 102.7 (C-5), 90.8 (C-1'), 86.4 (C-4'), 75.7 (C-2'), 71.3 (C-3'), 62.3 (C-5'). Selected HMBC and COSY correlations are depicted in Figure 1.

3,4,5-Trimethoxyphenol (3): Yellowish amorphous powder. HR-ESI-MS: 207.0630 [M+Na] (calcd. for CHO+Na, 207.0633). H-NMR data (CDCl, 500 MHz) (, in Hertz): 6.08 (2H, , H-2/6), 3.78 (3H, , 4-OCH), 3.77 (6H, , 3/5-OCH). C-NMR data (CDCl, 125 MHz): δ 153.9 (C-3/5), 152.7 (C-1), 131.9 (C-4),93.2 (C-2/6), 61.2 (4-OCH), 56.1 (3/5-OCH). Selected HMBC correlations are depicted in Figure 1.

Kelampayoside A (4): White amorphous powder, HR-ESI-MS: 501.1563 [M+Na] (calcd. for CHONa, 501.1579). H NMR data (CDOD, 500 MHz) (, in Hertz): 6.46 (2H, , H-2/6), 4.97 (1H, , 2.5, H-1″), 4.79 (1H, overlapped, H-1'), 4.04 (1H, , 11.0, 2.0, H-6'a), 3.95 (1H, , 10.0, H-4″a), 3.88 (1H, , 2.5, H-2″), 3.82 (6H, , H-3/5), 3.74 (1H, , 9.5, H-4″b), 3.71 (3H, , 4-OCH), 3.61 (1H, , H-6'b), 3.56 (1H, , H-5'), 3.55 (2H, , H-5″), 3.44 (1H, , H-3'), 3.42 (1H, , H-2'), 3.32 (1H, H-4' C NMR data (CDOD, 125 MHz): δ 155.9 (C-1), 154.8 (C-3/5), 134.8 (C-4), 110.9 (C-1″), 103.2 (C-1'), 96.5 (C-2/6), 80.5 (C-3″), 78.0 (C-2″), 77.9 (C-3'), 77.0 (C-5'), 75.0 (C-2'), 74.9 (C-4″), 71.6 (C-4'), 68.7 (C-6'), 65.5 (C-5″), 61.3 (4-OCH), 56.8 (3/5-OCH). Selected HMBC and COSY correlations are depicted in Figure 1.

Polystachyol (5): Yellowish gum. HR-ESI-MS: 443.1671 [M+Na] (calcd. for CHONa, 443.1676). H-NMR data (CDOD, 500 MHz) and C-NMR data (CDOD, 125 MHz) are presented in Table 2. Selected HMBC and COSY correlations are depicted in Figure 1.

(±)-Lyoniresinol 3---D-glucopyranoside (6): Yellowish gum. HR-ESI-MS: 605.2216 [M+Na] (calcd. for CHONa, 605.2205). H-NMR data (CDOD, 500 MHz) and C-NMR data (CDOD, 125 MHz) are presented in Table 2. Selected HMBC and COSY correlations are depicted in Figure 1.

DISCUSSION

Compound 1isolated from fraction EA9 was a white amorphous powder. The H-NMR spectrum of compound 1 exhibited a singlet signal at 7.51 (1H, s, H-8) for a characteristic aromatic proton on a fully substituted aromatic system. Three singlet signals at 3.41 (3H, s, 1-CH), 3.59 (3H, s, 3-CH), and 4.00 (3H, s, 7-CH) revealed the presence of three methyl groups attached to heteroatoms. Its C-NMR displayed 8 carbons, including two carbonyl carbons resonated at 155.4 (C-6) and 151.7 (C-2), three olefinic carbons at 107.6 (C-5), 148.7 (C-4), and 141.4 (C-8), and three methyl groups at 27.9 (1-CH), 29.7 (3-CH), and 33.6 (7-CH). The skeleton of compound 1 and the positions of the three methyl groups were determined by the HMBC correlations displayed in Figure 1. Its ESI-MS showed the molecular ion peak [M+H]at 194.8 (calcd. for CHNO+H, 195.1) corresponding to the molecular weight of caffeine, an abundant compound in coffee beans. The NMR data of compound 1 were compatible with those in Sitkowski’s report11. Therefore, compound 1 was identified as caffeine.

Compound 2 was isolated as a yellowish amorphous powder from fraction EA14. Downfield, the H-NMR spectrum of 2 displayed two olefinic methine signals coupling each other at δ 7.99 (, 8.0 Hz, H-6) and δ 5.69 (, 8.0 Hz, H-5). Four oxymethine signals resonated at δ 5.89 (, 4.5 Hz, H-1'), 4.17 (, 10.0, 5.5 Hz, H-2'), 4.15 (, 10.0, 5.0 Hz, H-3') and 4.00 (, H-4') along with two oxymethylene signals at δ 3.84 (, 12.0, 2.5 Hz, H-5'a) and 3.73 (, 12.0, 3.0 Hz, H-5'b) were features of a pentose. The C-NMR spectrum of compound 2 showed nine carbons, including two carbonyl groups at δ 166.2 (C-4) and 152.5 (C-2), two olefinic methines at δ 142.7 (C-6) and 102.7 (C-5), four oxymethines at δ 90.8 (C-1'), 86.4 (C-4'), 75.7 (C-2') and 71.3 (C-3'), and an oxymethylene at δ 62.3 (C-5'). The signals at δ 5.89 (, 4.5 Hz, H-1') and δ 90.8 (C-1'), which were characteristic of an anomer position, indicated the presence of a sugar unit. This fact was supported by the COSY correlations of H-1'–H-2'–H-3'–H-4'–H-5'. The HMBC cross-peaks displayed in Figure 1 allowed us to determine a ribose unit for compound 2. Two olefinic methines belonging to the same ring with two carbonyl groups were demonstrated using COSY correlations between δ 7.99 (H-6) and 5.69 (H-5) and HMBC correlations from H-6 (δ 7.99) to C-2 (δ 152.5) and C-4 (δ 166.2) and from H-5 (δ 5.69) to C-4 (δ 166.2). In addition, based on the HMBC cross-peaks from H-1' (δ 5.89) to C-6 (δ 142.7) and C-2 (δ 152.5), the ribose was determined to attach the aglycon at C-1'. The H-NMR and C-NMR data of compound 2 were compatible with those of uridine12. The ESI–MS displayed the pseudomolecular ion peak at 244.7 [M+H] which was appropriate for uridine (calcd. for CHNO+H, 245.1). The above information supported the identification of compound 2 as uridine.

Compound 3 was isolated from fraction H5 as a yellowish amorphous powder. Its HR-ESI-MS showed a sodiated molecular peak at 207.0630 [M+Na] appropriate for a calculated molecular weight of CHONa of 207.0633. Therefore, the molecular formula of compound 3 was assigned to CHO. Its H-NMR spectrum exhibited a singlet at 6.08 (, H-2/6) for aromatic protons. In addition, two singlets resonated at δ 3.78 (, 4-OCH) and 3.77 (, 3/5-OCH), suggesting that the aromatic ring bears three methoxy groups. The C-NMR spectrum of 3 displayed seven signals, including 2 aromatic methines at δ 93.2 (C-2/6), 4 oxygenated aromatic carbons at δ 153.9 (C-3/5), 152.7 (C-1), 131.9 (C-4), and 3 methoxy groups at δ 61.2 (4-OCH), 56.1 (3/5-OCH).The locations of methoxy and hydroxyl groups were determined by HMBC correlations, which are displayed in Figure 1. The 1D-NMR data of compound 3 had good compatibility with 3,4,5-trimethoxyphenol13. Therefore, the structure of compound 3 was elucidated as 3,4,5-trimethoxyphenol.

Compound 4 was isolated as a white amorphous powder from fraction EA13. Its molecular formula was identified as CHO using HR-ESI-MS with a quasi-molecular ion peak [M+Na] at / 501.1563 (calc. for CHO+Na, 501.1579). The H-NMR and C-NMR spectra of compound 4 were similar to those of compound 3 in the presence of signals indicating a tetrasubstituted aromatic ring, except for the addition of two sugar moieties. The presence of two sugar units was demonstrated by two anomer protons at δ 4.79 (H-1') and 4.97 (, 2.5 Hz, H-1″) and oxygenated protons in the zone of 3.0-4.0 ppm. The signal at δ 4.79 (H-1') overlapping with the residual water signal was assigned using the HSQC spectrum. Additionally, a quaternary carbon at δ 134.8 (C-4) was explored by the HMBC spectrum. The analysis of the C-NMR spectrum aligned with the H-NMR and HSQC spectra along with the COSY spectrum allowed the assignment of six oxygenated carbons at δ 103.2 (C-1'), 75.0 (C-2'), 77.9 (C-3'), 71.6 (C-4'), 77.0 (C-5'), 68.7 (C-6') belonging to a hexose moiety and five remaining oxygenated carbons at δ 110.9 (C-1″),78.0 (C-2″), 80.5 (C-3″), 74.9 (C-4″) and δ 65.5 (C-5″) belonging to a pentose unit. The attachment of two sugar units to the aromatic ring was assigned by the HMBC correlations (Figure 1). From the above analysis, compound 4 was predicted to be a phenolic diglycoside. Based on a previous report14, the NMR data of compound 4 showed good compatibility with 3,4,5-trimethoxyphenyl 1---D-apiofuranosyl(1→6)-glucopyranoside or kelampayoside A and was elucidated as this structure.

Compound 5 was isolated as a yellowish gum from fraction EA10. The HR-ESI-MS spectrum exhibited a sodiated molecular ion peak [M+Na]at 443.1671 (calcd. for CHO+Na, 443.1676) to identify the molecular formula of compound 5 as CHO with 9 degrees of unsaturation. Its H-NMR displayed two singlet signals at δ 6.59 (, H-8) and δ 6.38 (, H-2'/6'), indicating the presence of a 1,2,3,4,5-pentasubstituted ring and a 1,3,4,5-tetrasubstituted ring. In addition, three singlet signals at δ 3.38 (, 5–OCH), 3.74 (, 3'/5'–OCH) and 3.86 (, 7–OCH) were assigned to four methoxy groups. The signals at δ 3.50 (, 5.5 Hz, H-3α), 3.59 (, 11.0, 5.0 Hz; H-2αa) and 3.48 (, 10.5, 7.0 Hz; H-2αb) demonstrated the presence of two oxymethylene . The signals at δ 2.70 (, 15.0, 4.5 Hz, H-1a), 2.57 (, 15.0, 11.5 Hz, H-1b), 1.63 (m, H-2), 4.31 (, 5.5 Hz, H-4) and 1.97 (m, H-3) were characteristic of an aliphatic fragment. The C-NMR along with H-NMR and HSQC indicated that compound 5 had 22 carbons, including two oxymethylenes at δ 64.2 (C-3α) and 66.8 (C-2α), four methoxy groups at δ 60.2 (5–OCH), 56.6 (7–OCH), and δ 56.8 (3'/5'–OCH), a methylene at δ 33.6 (C-1), three methine at δ 49.5 (C-3), 42.3 (C-4), and 40.9 (C-2), and 12 aromatic carbons. The COSY spectrum supported the identification of two associated proton spin systems of H-1–H-2–H-2α and H-4–H-3–H-3α connecting each other at C-2–C-3, which are depicted in Figure 1. From the above information, compound 5 was predicted to have two benzene rings connected via a side chain with the feature of a tetrahydronaphthalene lignan or aryl-tetralin-type lignan. The HMBC cross-peaks from the aromatic protons and the methoxy protons to the aromatic carbons supported the identification of two aromatic ring systems as 4-hydroxy-3,5-dimethoxyphenyl moieties. The attachment of two aromatic rings to the side chain was established by HMBC correlations, which are displayed in Figure 1. The HMBC correlations from H-1 and H-2 to C-9, from H-3 to C-10, and from H2'/6' to C-4 indicated the connection between the two rings to the side chain. The combination of the above NMR analysis and comparison of NMR data with a previous report15 demonstrated the structure of compound 5 as polystachyol.

Mixture 6 was isolated from fraction EA13. The molecular formula of 6 was established as CHO using HR-ESI-MS with a sodiated molecular ion peak [M+Na] at / 605.2216 (calcd. as 605.2205). The H-NMR and C-NMR spectra of 6 were similar to those of compound 5 except for an additional set of glucopyranosyl signals in the zone 3.0-4.0 ppm according to 62-105 ppm, along with anomeric protons at 4.29 (, 8.0 Hz, H-1″, 6a) and 4.14 (, 8.0 Hz; H-1″, 6) and two acetal carbons at 104.8 and 104.3. Using the HMBC experiment, the location of the glucopyranosyl moiety was determined at C-3α (Figure 1). Based on a previous publication, 6 had good compatibility with (+)-lyoniresinol 3-D-glucopyranoside and (−)-lyoniresinol 3-D-glucopyranoside.16 Therefore, 6 was identified as a mixture of (±)-lyoniresinol 3---D-glucopyranoside.

Five compounds, 1, 2, 4, 5, and 6, were evaluated for their -glucosidase inhibitory capacity. The activities of compounds 2 (IC = 202.40 M), 4 (182.87 M), and 5 (139.51 M) were better than that of the positive control, acarbose (209.80 M), while compound 1 was weaker than acarbose and mixture 6 was inactive for -glucosidase inhibition.

Caffeine (1) is the main alkaloid in coffee. According to Herawati's research, caffeine is not a key bioactive component for antioxidant, anti--glucosidase, and antiglycation activities.17 Phenolic acids, specifically 5--caffeoylquinic acid, 4--caffeoylquinic acid, and 3--caffeoylquinic acid, are major contributors to antioxidant activity as well as prospective inhibitors against -glucosidase. The paper by Duangjai also mentioned caffeine's ineffective glucosidase inhibition but highly potent amylase inhibitory activity.18 The IC value of compound 1 in our investigation was higher than that of acarbose. It makes sense that caffeine has a limited anti--glucosidase activity.

In the Phoopha report, uridine (2) had no effect on -glucosidase inhibition.19 The IC value of compound 2 was the lowest among those of the isolated compounds in our study.

Kelampayoside A (4) has been evaluated at 1.0 × 10 M and found to be inactive for human tumor cytotoxicity, anti-inflammatory activity, antioxidant activity, anti-HIV activity, and neuroprotective activity, but it has inhibitory activity against protein tyrosine phosphatase 1B.20 To the best of our knowledge, there has not yet been any report on an -glucosidase inhibitory assay applied to kelampayoside A (4). Our investigation found that compound 4 (IC= 182.87 M) inhibited -glucosidase better than acarbose (IC= 209.80 M).

Thao reported that the sucrase inhibitory experiment revealed a considerable -glucosidase inhibitory activity for polystachyol (5).21 In our investigation utilizing a colorimetric reaction, compound 5 (IC= 139.51 M) also exhibited -glucosidase inhibitory activity better than acarbose (IC= 209.80 M).

Research from Jong-Anurakkun revealed that (-)-lyoniresinol 3---D-glucopyranoside exhibited inhibitory activity against both sucrase and maltase with IC values of 1.95 and 1.43 mM, respectively, while its enantiomer was inactive.22 In our investigation, a mixture of (+)-lyoniresinol 3---D-glucopyranoside and (-)-lyoniresinol 3---D-glucopyranoside (6) was isolated and was unable to separate. It was shown to be inactive against -glucosidase inhibition.

CONCLUSION

From the -hexane and ethyl acetate extract of trunk, five compounds and a mixture were isolated, including two alkaloid (caffeine and uridine), a phenolic compound (3,4,5-trimethoxyphenol) and three phenolic glycosides (kelampayoside A, polystachyol, and (±)-lyoniresinol 3---D-glucopyranoside). Based on NMR and MS data analysis as well as comparison to published values, their chemical structures were elucidated. To the best of our knowledge, all the isolated compounds were investigated from the genus for the first time with the exception of 1. Compounds 2, 4 and 5 were discovered to be potent -glucosidase inhibitors that are stronger than acarbose, a popular commercial medicine.

ABBREVIATIONS

CNMR: Carbon-13 Nuclear Magnetic Resonance

calcd.: calculated

CC: Chromatographic column

COSY: Homonuclear COrrelation Spectroscopy

: doublet :doublet of doublets : multiplet : singlet

HMBC: Heteronuclear Multiple Bond Correlation

HSQC: Heteronuclear single quantum coherence

HR-ESI-MS: High Resolution-ElectroSpray Ionization-Mass Spectrometry

HNMR: Proton Nuclear Magnetic Resonance

IC: The half maximal inhibitory concentration

RP: reversed-phase

COMPETING INTEREST

The authors declare no competing financial interest.

AUTHORS’ CONTRIBUTION

Vo Thi Nga and Nguyen Thi Anh Tuyet contributed to designing the experiments, acquiring the data, interpreting the data and giving final approval of the manuscript to be submitted. Hoang Thi Hien, Nguyen Vu Nhat Ha, Le Thi Minh Phuong, Nguyen Thi Thucarried out the experiments, interpreted NMR and MS data and searched the bibliography.

Corresponding author: Dr. Nguyen Thi Anh Tuyet, Department of Chemistry, Ho Chi Minh City University of Education, Ho Chi Minh City. Email: tuyetnta@hcmue.edu.vn

ACKNOWLEDGMENTS

Thanks are extended to Ho Chi Minh City University of Technology and Education, who supported the facilities for this research.