Phenolic compounds with pancreatic lipase inhibitory activity from Korean yam (Dioscorea opposita)

Abstract

Twenty-three phenolic compounds were isolated from Dioscorea opposita by bioactivity-guided method and their inhibitory effect against pancreatic lipase was evaluated. A total of 15 isolates reduced lipase activity at IC₅₀ values of less than 50 µM and 3,3′,5-trihydroxy-2′-methoxybibenzyl showed the highest inhibition with an IC₅₀ value of 8.8 µM. This study is a first to reveal the pancreatic lipase inhibitory activity by both D. opposita and its isolated compounds.

• Dioscorea opposita
• pancreatic lipase inhibitory activity
• phenolic compounds
• 3,3′,5-trihydroxy-2′-methoxybibenzyl

Introduction

Lipid digestion is initiated in the stomach and completed in the small intestine by the action of diverse gastrointestinal lipases1. The human lipases include the pre-duodenal lingual and gastric lipase and the extra-duodenal pancreatic, hepatic, lipoprotein and endothelial lipase2,3. Of these several lipases, pancreatic lipase is a predominant lipolytic enzyme in humans responsible for the hydrolysis and absorption of 50–70% of total dietary fats in the intestinal lumen2,4. Therefore, pancreatic lipase plays a key role in preventing obesity, and the pancreatic lipase inhibition is a well-known mechanism for the determination of natural products with anti-obesity property5.

Phenolic compounds are secondary aromatic plant metabolites possessing one or more hydroxyl substituents6. Naturally occurring polyphenols have beneficial health-related properties, which are based on their strong antioxidant activities6,7. According to recent studies, many polyphenol-rich extracts are effective inhibitors of pancreatic lipase in vitro5,8. Furthermore, phenolics from plant extracts, such as flavonols and tannins, also reduced obesity in vivo together with lipase inhibition8,9. These results suggest that phenolic materials may be useful to decrease dietary fat absorption and accumulation.

Dioscorea opposita has been cultivated in China, Japan and Korea as a food and used for digestive problems with poor absorption for a long time10. In the preliminary experiment, the extract of D. opposita Thunb. (Dioscoreaceae) effectively suppressed lipase activity. Phytochemical investigations of D. opposita have revealed many chemical components such as allantoin, batatasin, diosgenin and dioscin11,12. It has also been reported to possess diverse pharmacological properties such as antioxidant, anti-inflammation and anti-diabetic effects13,14. However, the lipase inhibitory effects of D. opposita and its phenolic components have not been reported previously. This study attempts to isolate and identify the active compounds from D. opposita through bioactivity-guided fractionation and investigate their inhibitory activities against porcine pancreatic lipase in vitro.

Materials and methods

General methods

¹H- and ¹³C-NMR, COSY, HMQC, HMBC and NOESY spectral data were run on Bruker AVANCE 400 and 500 spectrometer (Bruker, Karlsruhe, Germany). EIMS and HREIMS were recorded on JEOL JMS 700 spectrometer (Tokyo, Japan). UV spectra were measured on a Perkin Elmer Lambda 25 UV/Vis spectrophotometer (California). IR spectral data were taken on a JASCO FT/IR-300E spectrophotometer (Tokyo, Japan). CPC was carried using an LLB-M high-performance centrifugal partition chromatography (Sanki Engineering, Kyoto, Japan) with a four-way switching valve operating in either descending or ascending mode together with binary Gilson 321 pump, a Gilson UV/Vis-151 detector, a Gilson FC 203B fraction collector and Rheodyne valve with a 5 mL sample loop for manual injection (Cotati, Sonoma County, CA). A Gilson HPLC system (Gilson, Fairfield, NJ) was used to isolate compounds, and was equipped with two 321 pumps, a UV/Vis-151 detector, an autosampler 234 and a fraction collector 204. Silica gel (230–400 mesh, Merck, Darmstadt, Germany) and Sephadex LH-20 gel (25–100 µm mesh, Pharmacia, Stockholm, Sweden) were used in column chromatography (CC).

Plant material

The rhizomes of D. opposita Thunb. (Dioscoreaceae) were provided by Tong Yang Moolsan Co., Ltd (Nonsan, South Korea) and identified by Prof. Gwang Jin Chang of the Korea National College of Agricultural and Fisheries. A voucher specimen (SNUPH-0822) has been deposited in the Medicinal Herb Garden, Seoul National University.

Extraction and isolation

Fresh rhizomes of D. opposita (19 kg) were sliced into small pieces and freeze-dried prior to extraction. The dried samples were extracted with 95% EtOH (20 L × 3) and evaporated under reduced pressure. The 95% EtOH extract (380 g) was suspended in distilled water (600 mL) and the suspension was partitioned with CHCl₃ (1 L), EtOAc (1 L) and n-BuOH (1 L), sequentially.

The active n-BuOH soluble fraction (50 g) was separated by silica gel CC (CH₂Cl₂:n-hexane (1:1) → CH₂Cl₂:MeOH (25:1) →CH₂Cl₂:MeOH (12.5:1) → CH₂Cl₂:MeOH (6.25:1) → CH₂Cl₂:MeOH (3:1) → CH₂Cl₂:MeOH (1:1)) to yield six fractions. Fraction 1 (9.3 g) was subjected to CPC with two-phase solvent system composed of the upper phase of n-hexane:MeOH:water (10:9:1, v/v) as a stationary phase (solvent A) and the lower phase as a mobile phase (solvent B), eluted with the mixture of solvent B and water (0–100% solvent B), to yield 7 subfractions (Frs 1-1–1-7). Fraction 1-2 (587.2 mg) was subjected to Sephadex LH-20 CC eluted with the same solvent of MeOH and 6 subfractions (Frs 1-2-1–1-2-6) were collected.

Fraction 1-2-2 was submitted to RP HPLC (YMC-pack ODS-A C18, S-5 µm, 250 × 10 mm, Kyoto, Japan; eluent, MeCN:H₂O (60:40); detection, UV at 254 nm; flow rate, 4 mL/min) to yield compounds 8 (3.5 mg, t_R 16.8 min), 10 (4.1 mg, t_R 20.5 min) and 6 (5.3 mg, t_R 27.1 min). Compounds 15 (2.4 mg, t_R 28.5 min), 5 (11.0 mg, t_R 40.2 min), 9 (6.9 mg, t_R 43.7 min), 13 (1.9 mg, t_R 52.5 min) and 20 (2.1 mg, t_R 66.0 min) were isolated from fraction 1-2-4 using Phe-Hex HPLC (YMC-pack Ph, S-5 µm, 250 × 10 mm; eluent, MeCN:H₂O (50:50); detection, UV at 280 nm; flow rate, 2 mL/min). Fraction 1-2-6 was subjected to Phe-Hex HPLC (YMC-pack Ph, S-5 µm, 250 × 10 mm; eluent, MeCN–H₂O (50:50); detection, UV at 280 nm; flow rate, 2 mL/min) to obtain compounds 14 (6.5 mg, t_R 20.2 min), 4 (10.6 mg, t_R 23.0 min), 11 (2.7 mg, t_R 27.5 min) and 16 (1.4 mg, t_R 31.0 min).

Fraction 1-4 (420 mg) was subjected to Sephadex LH-20 CC eluted with an isocratic elution of MeOH to yield six subfractions (Frs 1-4-1–1-4-6). Fraction 1-4-2 was subjected to RP HPLC (YMC-pack ODS-H80, S-4 µm, 250 × 20 mm; eluent, MeCN–H₂O (80:20); detection, UV at 254 nm; flow rate, 4 mL/min) to yield compounds 12 (11.0 mg, t_R 12.1 min) and 7 (3.2 mg, t_R 18.1 min). Fraction 1-6 (165 mg) was applied to a Sephadex LH-20 CC, using MeOH, 5 subfractions (Frs 1-6-1–1-6-5) were collected and fraction 1-6-2 yielded compound 19 (14.5 mg, t_R 19.8 min) by RP HPLC (Zorbax column Eclipse XDB-C18, S-5 µm, 250 × 9.4 mm, Agilent, CA; eluent, MeCN:H₂O (50:50); detection, UV at 210 nm; flow rate, 4 mL/min). Fraction 1-7-2 from the first Sephadex chromatography for fraction 1-7 (230 mg) was purified by RP HPLC (Zorbax column Eclipse XDB-C18, S-5 µm, 250 × 9.4 mm; detection, UV at 210 nm; flow rate, 4 mL/min) eluted with MeCN–H₂O (55:45), affording compounds 23 (2.7 mg, t_R 20.4 min) and 21 (2.6 mg, t_R 24.9 min).

Fraction 2 (15.0 g) was subjected to CPC, eluted with the same solvent mixture as referred to above, to yield seven subfractions (Frs 2-1–2-7). Fraction 2-3 was re-fractionated by CC on Sephadex LH-20 eluted with MeOH to give five fractions as Frs 2-3-1–2-3-5. Fraction 2-3-5 was successively purified by Phe-Hex HPLC (YMC-pack Ph, S-5 µm, 250 × 10 mm; eluent, MeCN:H₂O (20:80); detection, UV at 254 nm; flow rate, 2 mL/min) to afford compounds 1 (5.1 mg, t_R 25.6 min) and 3 (8.8 mg, t_R 30.4 min), respectively. Fraction 2-4 was re-chromatographed on a Sephadex LH-20 as described above. A total of five fractions (Frs 2-4-1–2-4-5) were collected and fraction 2-4-2 was submitted to RP HPLC (YMC J’sphere ODS-H80, S-4 µm, 250 × 20 mm; eluent, MeOH:H₂O (90:10); detection, UV at 203 nm; flow rate, 4 mL/min) to provide compound 18 (2.2 mg, t_R 15.0 min).

Fraction 2-6 was chromatographed on a Sephadex LH-20 as an eluent of MeOH to give five fraction groups (Frs 2-6-1–2-6-5) in order of elution. Fraction 2-6-4 gave compound 17 (2.2 mg, t_R 12.0 min) using RP HPLC (Phenomenex Gemini C18 110A, S-5 µm, 250 × 21.20 mm, Phenomenex, CA; eluent, MeCN–H₂O (85:15); detection, UV at 210 nm; flow rate, 5 mL/min). Fraction 2-7 was subjected to Sephadex LH-20 CC (MeOH) to yield six subfractions (Frs 2-7-1–2-7-6). Compounds 2 (4.2 mg, t_R 18.3 min) and 22 (6.9 mg, t_R 42.0 min) were isolated from fraction 2-7-5 by RP HPLC (YMC-pack Pro C18 RS, S-5 µm, 250 × 20 mm; eluent, MeCN:H₂O (30:70); detection, UV at 254 nm; flow rate, 4 mL/min). The purities of isolated compounds were evaluated as over 98% by HPLC method, and they were used as test samples in this study.

In vitro pancreatic lipase assay

Pancreatic lipase (type II, from porcine pancreas, Sigma-Aldrich, St. Louis, MO) activity was determined using 4-methylumbelliferyl oleate (4-MU oleate, Sigma-Aldrich) as a substrate15,16. Various concentrations of test compound (5, 10, 25, 50 and 100 µM) were dissolved in DMSO (final culture concentration 0.1%) and 4-MU oleate and lipase were diluted in 0.1 M McIlvane buffer containing 0.1 M citric acid–Na₂HPO₄ (pH 7.4). The mixed solution that composed of 100 µL of 0.1 mM 4-MU oleate, 40 µL of the above-mentioned buffer and 10 µL of test samples was prepared before adding lipase. The enzymatic reaction was started by adding 0.05 mL of 1 U/mL pancreatic lipase and incubation was carried out for 20 min at 37 °C. The amount of 4-methylumbelliferon released by the enzyme was monitored at an emission of 450 nm and excitation wavelength of 320 nm using a fluorescence spectrophotometer. The inhibition activity was calculated using the following formula: where A₀ is the absorbance of the control without inhibitor, A₁ the absorbance in the presence of the test sample and A₂ the absorbance sample blank (without porcine pancreatic lipase). IC₅₀ was determined to be the sample concentration decreasing the initial enzymatic activity by 50% of under the experimental conditions given. The value was obtained to the linear portion of the log concentration–response curves by the least square regression method.

The K_i value was obtained from the Cheng–Prusoff equation17: where K_i is the dissociation constant, K_m the Michaelis constant and [S] the concentration of the substrate. The K_m constant was determined using the Lineweaver–Burk plots from the relevant Michaelis–Menten equations.

Statistical analysis

Data are expressed as the mean ± SD. The values were expressed as percent changes from the mean value of the control experiment. Statistical analyses were performed by a one-way analysis of variance using statistical package. P values less than 0.05 were considered statistically significant.

Results and discussion

The n-BuOH fraction from 95% EtOH extract of D. opposita appeared the lipase inhibition (57% at final concentration of 10 µg/mL) and was further partitioned on the basis of polarity to find the active fractions. Of the six subfractions, fractions 1 and 2 were most active against pancreatic lipase, implying that they might have active components with lipase inhibitory activity (Figure 1).

Figure 1. Effect of D. opposita on pancreatic lipase in the presence or absence of samples at the final concentration of 10 µg/mL. Inhibition of lipase by 95% EtOH extract and fractions of D. opposita (A) and the subfractions of D. opposita n-BuOH fraction (B). Each values are expressed as the % control means ± SD of triplicate experiments; *p < 0.05, **p < 0.01 and ***p < 0.001 compared with the control.

Bioassay-guided fractionation of D. opposita extract led to the isolation and identification of 23 phenolic compounds including 10 dihydrostilbenes, tristin18 (1), 2′,3,5-trihydroxybibenzyl19 (2), 3,3′,5-trihydroxy-2′-methoxybibenzyl14 (3), batatasin III20 (4), batatasin IV19 (5), 2′,4-dihydroxy-3,5-dihydroxy-4-methoxybibenzyl14 (6), 3,5-dimethoxy-2′-hydroxybibenzyl21 (7), 3,5-dimethoxybibenzyl22 (8), 3,4-dimethoxy-2′,5-dihydroxybibenzyl23 (9), batatasin V24 (10), four phenanthrenes, 3,5-dimethoxy-2,7-phenanthrenediol25 (11), batatasin I26 (12), hircinol26 (13) and 2,5-dihydroxy-7-methoxy-9,10-dihydrophenanthrene27 (14), two dibenzoxepins, 9,10-dihydro-dibenzoxepin-2,4-diol14 (15) and 9,10-dihydro-4-methoxy-dibenzoxepin-2-ol14 (16), five diarylheptanoids, p-hydroxyphenylethyl p-coumarate28 (17), p-hydroxyphenethyl trans-ferulate29 (18), (1E,4E,6E)-1,7-bis(4-hydroxyphenyl)-1,4,6-heptatrien-3-one30 (19), (4E,6E)-1,7-bis(4-hydroxyphenyl)-4,6-heptadien-3-one31 (20) and (4E,6E)-7-(4-hydroxy-3-methoxyphenyl)-1-(4-hydroxyphenyl)-4,6-heptadien-3-one32 (21), two p-hydroxyphenylethyl-p-hydroxyphenyl propenoic acids, (3R,5R)-3,5-dihydroxy-1,7-bis(4-hydroxyphenyl)-3,5-heptanediol33 (22) and (3R,5R)-1,7-bis(4-hydroxy-3-methoxyphenyl)-3,5-heptanediol33 (23) by spectroscopic methods using 2D-NMR (Figure 2).

Figure 2. Structures of phenolic compounds isolated from D. opposita n-BuOH extract.

All the isolates obtained in this investigation were evaluated for their ability decreasing lipase activity against the porcine pancreatic lipase. Fifteen compounds from the active subfractions of D. opposita extract revealed in vitro lipase inhibitions at the IC₅₀ values of less than 50 µM (Table 1). In particular, 3,3′,5-trihydroxy-2′-methoxybibenzyl (3) and (4E,6E)-1,7-bis(4-hydroxyphenyl)-4,6-heptadien-3-one (20) effectively reduced lipase activity with the IC₅₀ of less than 10 µM and dose-dependently in the concentration range 5–100 µM (Figure 3). Of the 15 active constituents, 8 contain dihydrostilbene skeleton, more commonly known as bibenzyl, and 3,3′,5-trihydroxy-2′-methoxybibenzyl showed the highest inhibition activity with an IC₅₀ value of 8.77 ± 2.03 µM (p < 0.05) and K_i value of 1.82 ± 0.43 µM. These results clearly suggest that phenolic compounds with stilbenoid are considered to play important roles in the lipase inhibition of the D. opposita extract.

Figure 3. A dose-dependent inhibition of lipase activity by 3,3′,5-trihydroxy-2′-methoxybibenzyl (A), (4E,6E)-1,7-bis(4-hydroxyphenyl)-4,6-heptadien-3-one (B) and Orlistat (C). Values are means of triplicate assays ± standard error.

Table 1. Inhibitory effects of compounds isolated from D. opposita against porcine pancreatic lipase†.

Compounds	Ki‡ (µm)	IC50§ (µm)
1	2.75 ± 0.14	13.50 ± 0.65**
2	6.45 ± 1.32	29.20 ± 6.26**
3	1.82 ± 0.43	8.77 ± 2.03*
4	6.79 ± 0.40	32.32 ± 1.91***
5	6.28 ± 0.68	30.14 ± 3.25***
6	5.06 ± 0.24	25.36 ± 1.14**
7	9.28 ± 0.26	44.11 ± 1.23***
10	9.66 ± 0.45	45.70 ± 2.14***
11	2.67 ± 0.30	12.72 ± 1.41*
13	8.57 ± 0.41	40.61 ± 1.93***
14	4.96 ± 1.34	25.09 ± 6.35**
17	7.49 ± 0.77	34.81 ± 3.64***
19	3.34 ± 0.86	16.36 ± 4.07*
20	1.92 ± 0.16	9.12 ± 0.76*
21	2.17 ± 0.74	11.08 ± 3.50*
Orlistat¶	0.26 ± 0.01	1.24 ± 0.05

†Compounds 8, 9, 12, 15, 16, 18, 22 and 23 were considered to be inactive

‡The K_i values were calculated from Cheng–Prusoff equation and K_m constant was determined using the Lineweaver–Burk plot.

§Each values were expressed as the means ± SD of triplicate experiments.

¶Positive control.

*p < 0.05, **p < 0.01 and ***p < 0.001 compared with the control.

Compounds 1–6 with a hydroxyl group of C-3 were more potent than 7–10 which have the methoxy group at the same position. Especially, the stilbenoids possessing 3,5-dihydroxybibenzyl moiety showed higher inhibitory potencies than the others. The lipase inhibitory activities of stilbenoids depend on the presence of the hydroxyl group in the C-3 position. The structural difference also influences the inhibitory effect of diarylheptanoids on pancreatic lipase. Three diarylheptanoids (19–21) with the carbonyl group reduced pancreatic lipase activity at low concentrations but C-3 hydroxylation (22 and 23) had no effect on the lipase inhibition. This indicates that the carbonyl group at C-3 contributed to the inhibitory activity of diarylheptanoids against pancreatic lipase.

To date, there were a few attempts to reveal the pancreatic lipase inhibition for Dioscorea species and saponin-type constituents such as dioscin and diosgenin were found to be lipase inhibitors as Dioscorea secondary metabolites16. There has been no previous research associated with the beneficial effects of phenolics from the D. opposita extract against pancreatic lipase. Current data suggested that the phenolic constituents from D. opposita were effective in inhibiting pancreatic lipase in vitro. Based on the results of all isolates, dihydrostilbene, phenanthrene and diarylheptanoid structures were deemed to be responsible for lipase inhibition. Particularly, 3,3′,5-trihydroxy-2′-methoxybibenzyl including dihydrostilbene moiety exhibited the most potent inhibition as a new lipase inhibitor.

Declaration of interest

The authors declare no conflict of interest.