Chemical constituents from acid hydrolyzates of Panax quinquefolius total saponins and their inhibition activity to α-glycosidase and protein tyrosine phosphatase 1B.

Objective: To investigate the hypoglycemic components from the acid hydrolyzates of Panax quinquefolius total saponins, and screen the active compounds by in vitro inhibitory activities to α-glycosidase enzymes and protein tyrosine phosphatase-1B (PTP1B).
Methods: The hydrolyzates were chromatographed repeatedly over silica gel column, and the structures of the compounds were determined by means of NMR. The in vitro bioassay was performed through the inhibitory effects on α-glucosidase or/and PTP1B.
Results: Eight compounds were isolated, which identified as 20(S)-panaxadiol (1), (20S,24R)-dammarane-20,24-epoxy-3β,6α,12β,25-tetraol (2), 20(R)-dammarane-3β,12β,20,25-tetraol (3), 20(S)-dammarane-3β,6α,12β,20,25-pentol (4), 20(R)-dammarane-3β,12β,20,25-tetrahydroxy-3β-O-β-d-glucopyranoside (5), β-sitosterol (6), oleanolic acid (7) and 20(S)-protopanaxadiol (8). Compound 5 was ginseng triterpenoid isolated from the acid hydrolysates of total saponins from P. quinquefolius for the first time. In this paper, the possible in vitro inhibitory activities were investigated. Compound 5 exhibited significantly inhibitory activity against α-glucosidase, and the IC50 value [(0.22 ± 0.21) µmol/L] was about 43-fold lower than positive control. For the PTP1B inhibition assay, compound 5 indicated the strongest inhibitory effect with IC50 of (5.91 ± 0.38) µmol/L, followed by compound 4 with IC50 of (6.21 ± 0.21) µmol/L, which were all showed competitive inhibitory pattern by using a Lineweaver-Burk plot.
Conclusion: These results supported the potential application of dammaranes from acid hydrolyzates of P. quinquefolius total saponins can be used as ingredients of ancillary anti-diabetic agent or functional factor.

Introduction

Diabetes mellitus (DM) is one of the major global public health problems, which describes a group of metabolic disorders characterized by high blood glucose levels. Its prevalence is currently increasing at an alarming rate (Cho et al., 2018). Type 2 diabetes is complicated metabolic disorder arising from insulin deficiency, or/and insulin resistance (Imam, 2012). Lots of researches on development of effective hypoglycaemic methods have provided discoveries of molecular therapy targets, such as α-glucosidase and protein tyrosine phosphatase-1B (PTP1B) (Fang, Cao, Duan, Tang, & Zhao, 2014). Recently, an increased interest in bioactive natural products by scientists, as a source of potential anti-diabetic agents, has been observed.

Panax quinquefolius L., also called American ginseng, is widely distributed in North America and cultivated in China, which has been traditionally used as a tonic with prophylactic and restorative properties. And the biologically active constituents are dammarane-type glycosides (Liu et al., 2013; Wang et al., 2001), with beneficial effects reported about treatment of type 2 diabetes mellitus (Vuksan et al., 2000), anticancer (Qi, Wang, & Yuan, 2010) and immunopotentiating.

As a continuation of our work for discovering more effective components, this paper focused on the isolation of triterpenoids from the acid hydrolyzates of P. quinquefolius total saponins and their inhibitory activities of α-glucosidase and PTP1B were evaluated.

Materials and methods

General

Melting points were determined with the X-4 Electrothermal melting point apparatus (Taike, Beijing) and are uncorrected. NMR Spectra were obtained from Bruker AV-600 and ARX-400 spectrometers using TMS as internal standard. The 96-well microplate reader was from Microplate Reader (imark, BIORAD, USA). Silica gel (Qingdao Marine Chemical Co., Ltd, China) were used for TLC and column chromatography.

Plant material and reagents

The crude saponins of P. quinquefolius (> 80%) were purchased from Longpu Bio-Tech Inc., Liaoning Province, China.

PTP1B (human, recombinant), yeast α-glucosidase, NaVO4, acarbose, dithiothreitol (DTT), pnitrophenyl-α-d-glucopyranoside (pNPG), p-nitrophenyl phosphate (pNPP), β-mercaptoethanol were purchased from Sigma-Aldrich. Ethanol, dichloromethane, petroleum ether, acetone, HCl, Na2CO3 and other chemicals of analytical grade were from Shenyang Factory of Chemicals, China.

Extraction and isolation

According to the acid hydrolysis method reported (Bai et al., 2010), the hydrolyzates (30 g) were isolated from the total saponins of P. quinquefolius. The extract was chromatographed repeatedly over silica gel column with petroleum ether-acetone (100:0–0:100, volume percent) to provide five fractions A−E. Fraction B (2.5 g) was separated into eight fractions, frs. Ba−Bh. Compound 1 (20 mg) was afforded from Bc (petroleum ether: acetone 100:10). Compounds 2 (30 mg), 3 (20 mg) and 4 (40 mg) were afforded from Be (petroleum ether: acetone 100:15). Compounds 6 (30 mg), 7 (35 mg) and 8 (25 mg) were gained from Bg (petroleum ether: acetone 100:25). Fraction D (0.8 g) was purified on silica gel column eluted with petroleum ether-acetone to afford compound 5 (20 mg) from subfraction 8 (100:40).

Assay for α-glycosidase inhibitory activity

The α-glucosidase inhibitory activities were determined following the modified method (Tao, Zhang, Cheng, & Wang, 2013), using a 96-well plate with PNPG as the substrate. A total of 30 µL of the enzyme solution (2 U/mL α-glucosidase in potassium phosphate buffer, pH 6.8) and 20 µL of each compound (in 1% DMSO) at various concentrations were mixed, and incubated at 37.0 °C for 5 min, subsequently. The reaction was started when 150 µL of pNPG (10 mmol/L) and 800 µL of 0.1 mol/L potassium phosphate buffer were added. After incubation for 30 min, 2 mL of 1 mol/L Na2CO3 in potassium phosphate buffer was added to stop the reaction. The blank control group was added 20 µL of 1% DMSO instead of the sample solution. The amount of PNP released was quantified by the absorbance at 405 nm.

Assay for PTP1B inhibitory activity

The method of PTP1B inhibition assay was conducted as previously reported method (Xu, Cao, Yue, Zhang, & Zhao, 2018). The different concentrations (in 1% DMSO) of all compounds were tested, and NaVO4 was used as a positive control. Totally, 10 µL of test samples (5, 25, 50, 75, 150 µmol/L), 83 µL enzyme in buffer (pH7.5) and 4 µL pNPP in reaction system pH 7.5, 2 mmol/L EDTA, 100 mmol/L NaCl, 1 m mol/L LDTT and tendency for 50 mmol/L Tris–HCl were mixed at 37 °C for 30 min, and 5 µL NaOH (2 mmol/L) was added to stop the reaction. At 405 nm, dephosphorylation of pNPP generated product pNP can be measured (Xiao, Guo, Sun, & Zhao, 2017).

Kinetic of α-glucosidase and PTP1B inhibitions

Compound 5 was analyzed for inhibitory type against α-glucosidase and PTP1B. The methods were measured as previously reported by Xu et al. (2018). The reaction mixture consisted of different concentrations of pNPG (or pNPP) as a substrate in the absence or presence of compound 5 at different concentrations. Concentration of inhibitors was fixed at 0, 5, 20 and 60 µmol/L (0, 25, 50 and 75 µmol/L), while concentration of substrate was varied at 0.125, 0.2, 0.25, 0.33, 0.5, and 1.0 mmol/L (or 0.2, 0.25, 0.33, 0.5, 1.0, and 2.0 mmol/L). According to Lineweave–Burk method, 1/[S] was used as the horizontal coordinate and 1/V as the vertical coordinate to plot the inhibition kinetics curves of compound 5, and we analyzed to determine the type of inhibition.

Molecular docking study

To investigate the interaction between compound 5 and PTP1B, the active site analysis method was carried out in molecular docking study. The crystal structure of PTP1B (PDB ID 2QBQ) was obtained from the Protein Data Bank (PDB) online. The structure optimization of compound 5 was processed by ChemBioDraw, and the docking calculation was performed using Schrodinger Suite 2009 software package installed on Dell Precision TM T5500 workstation. The operating system was Red Head Enterprise Linux server 5.0. Molecular docking analysis and mapping were performed by Discovery Studio Visualizer 3.5 and PyMOL (DeLano Scientific). Blind docking of 5 to the catalytic active site (site A, residues 214–221) of PTP1B and the secondary lipophilic non-catalytic aryl phosphate ester binding site (site B) was investigated.

Statistical analysis

Data were statistically treated by analysis of variance using SPSS 16.0 software. All data was expressed as mean ± S.D. in triplicate, P < 0.05 was considered statistically significant.

Results and discussion

Structural identification of compounds 1–8

20(S)-panaxadiol (1)

White crystals. Mp 239−241 °C. ESI-MS m/z: 459.5 [M−H]−. 1H NMR (400 MHz, C5D5N) δ: 5.32 (1H, t, J = 7.2 Hz, H-24), 3.91 (1H, m, H-12α), 3.43 (1H, m, H-3α), 1.59 (3H, s, H-28), 1.46 (3H, s, H-21), 1.27 (3H, s, H-26), 1.23 (3H, s, H-27), 1.22 (3H, s, H-18), 1.03 (3H, s, H-29), 0.99 (3H, s, H-19), 0.88 (3H, s, H-30). 13C NMR data (100 MHz, C5D5N) see Table 1.

Table 1

13C NMR (100 MHz or 150 MHz) data of compounds 1−8 in C5D5N.

NO.	1	2	3	4	5	6	7	8
1	39.6	39.4	39.6	39.1	39.2	30.1	38.7	39.4
2	28.3	28.1	28.3	27.9	28.2	31.8	27.9	28.3
3	78.1	78.4	78.0	78.2	88.8	71.7	77.8	78.0
4	39.5	40.4	39.4	40.1	39.7	41.9	39.2	39.6
5	56.4	61.9	56.4	61.6	56.4	140.9	55.6	56.4
6	18.8	67.7	18.8	67.5	18.5	121.8	18.6	18.8
7	35.3	47.5	35.3	47.3	35.2	30.0	33.0	35.2
8	40.1	41.1	40.1	41.0	40.1	31.9	39.5	40.1
9	50.3	50.5	50.5	49.9	50.4	50.8	47.9	50.5
10	37.4	39.4	37.4	39.1	37.0	37.8	37.1	37.3
11	31.4	32.4	32.2	31.2	32.2	22.7	23.5	32.1
12	70.3	71.2	70.9	70.8	70.9	37.2	122.3	71.0
13	49.9	48.4	49.3	48.1	49.3	44.0	144.6	48.6
14	51.4	52.1	50.8	51.4	50.8	56.5	41.9	51.7
15	31.3	31.7	31.5	31.9	31.5	27.7	28.1	31.4
16	25.4	25.5	26.7	26.6	26.7	27.3	23.6	26.9
17	55.0	49.4	51.8	54.5	51.8	58.3	46.4	54.8
18	17.3	17.8	16.5	17.2	16.8	20.7	41.8	16.3
19	16.5	17.2	17.4	16.8	17.4	23.1	46.3	15.9
20	76.9	86.7	73.4	73.1	73.3	36.0	30.7	72.9
21	19.7	27.0	22.9	27.0	22.9	19.4	34.0	27.1
22	35.8	32.8	44.1	36.3	44.1	33.9	33.0	35.9
23	16.5	28.8	18.8	18.9	18.8	29.9	28.6	23.0
24	36.6	85.6	45.6	45.5	45.6	46.0	16.3	126.3
25	73.1	70.4	69.7	69.4	69.7	25.7	15.3	130.7
26	33.2	27.2	30.0	29.7	30.0	12.2	17.2	25.8
27	27.4	27.7	30.3	30.0	30.2	31.7	25.9	17.6
28	28.7	31.9	28.7	31.7	26.8	21.0	180.0	28.7
29	15.9	16.5	15.9	16.3	15.9	21.0	33.0	16.4
30	16.4	18.4	16.4	17.3	16.4		23.5	17.0
glc-1′					107.0
2′					75.8
3′					78.8
4′					71.9
5′					78.4
6′					63.1

(20S,24R)-dammarane-20,24-epoxy-3β,6α,12β,25- tetraol (2)

White crystals. Mp 199−201 °C. ESI-MS m/z: 493.4 [M + H]+. 1H NMR (400 MHz, C5D5N) δ: 4.40 (1H, m, H-6β), 3.74 (1H, m, H-12α), 3.53 (1H, m, H-3α), 1.98 (3H, s, H-28), 1.51 (3H, s, H-21), 1.46 (3H, s, H-26), 1.33 (6H, s, H-18, H-27), 1.09 (3H, s, H-29), 1.00 (3H, s, H-19), 0.95 (3H, s, H-30). 13C NMR data (100 MHz, C5D5N) see Table 1.

20(R)-dammarane-3β,12β,20,25-tetraol (3)

White crystals. Mp 251−253 °C. ESI-MS m/z: 477.7 [M − H]−. 1H NMR (400 MHz, C5D5N) δ: 3.94 (1H, m, H-12α), 3.43 (1H, m, H-3α), 1.44 (9H, s, H-26, 27, 28), 1.07 (3H, s, H-21), 1.05 (3H, s, H-18), 1.04 (3H, s, H-29), 0.93 (3H, s, H-19), 0.94 (3H, s, H-30). 13C NMR data (100 MHz, C5D5N) see Table 1.

20(S)-dammarane-3β,6α,12β,20,25-pentol (4)

White crystals. Mp 247−249 °C. ESI-MS m/z: 493.6 [M − H]−. 1H NMR (400 MHz, C5D5N) δ: 4.41 (1H, m, H-6β), 3.94 (1H, m, H-12α), 3.53 (1H, m, H-3α), 1.99 (3H, s, H-28), 1.45 (3H, s, H-21), 1.42 (3H, s, H-26), 1.39 (3H, s, H-27), 1.38 (3H, s, H-18), 1.17 (3H, s, H-29), 1.02 (3H, s, H-19), 0.96 (3H, s, H-30). 13C NMR data (150 MHz, C5D5N) see Table 1.

20(R)-dammarane-3β,12β,20,25-tetrahydroxy-3β-O-β-d-glucopyranoside (5)

White amorphous powder, Libermann-Burchard and Molish reactions were positive. Mp 213−215 °C. ESI-MS m/z: 639.9 [M − H]−. 1H NMR (400 MHz, C5D5N) δ: 3.93 (1H, s, H-12α), 3.38 (1H, s, H-3α), 1.99 (3H, s, H-28), 1.42 (3H, s, H-27), 1.41 (3H, s, H-26), 1.32 (3H, s, H-21), 1.01 (3H, s, H-29), 1.00 (3H, s, H-18), 0.95 (3H, s, H-19), 0.82 (3H, s, H-30). 13C NMR data (150 MHz, C5D5N) see Table 1.

β-Sitosterol (6)

White amorphous powder. Mp 139−141 °C. 13C NMR data (100 MHz, C5D5N) see Table 1.

Oleanolic acid (7)

White amorphous powder. Mp 308−310 °C. ESI-MS m/z: 457.7 [M + H]+. 1H NMR (400 MHz, C5D5N) δ: 5.28 (1H, brs, H-12), 3.22 (1H, dd, J = 10.2, 4.5 Hz, H-3), 1.13 (3H, s, H-29), 0.99 (3H, s, H-23), 0.93 (3H, s, H-27), 0.91 (3H, s, H-30), 0.90 (3H, s, H-26),0.77 (3H, s, H-24), 0.75 (3H, s, H-25). 13C NMR data (100 MHz, C5D5N) see Table 1.

20(S)-protopanaxadiol (8)

White amorphous powder. Mp 237−239 °C. ESI-MS m/z: 459.5 [M − H]−. 1H NMR (400 MHz, C5D5N) δ: 5.33 (1H, t, J = 7.2 Hz, H-24), 3.92 (1H, m, H-12α), 3.44 (1H, m, H-3α), 1.60 (3H, s, H-28), 1.59 (3H, s, H-21), 1.37 (3H, s, H-26), 1.17 (3H, s, H-27), 0.98 (3H, s, H-18), 0.95 (3H, s, H-29), 0.88 (3H, s, H-19), 0.83 (3H, s, H-30). 13C NMR data (100 MHz, C5D5N) see Table 1.

The structures of eight known compounds (1−8) were shown in Fig. 1. They were white amorphous powder, and identified mainly by their corresponding NMR spectra (Table 1). The data obtained were compared with those reported in the literatures. The compounds were characterized as 20(S)-panaxadiol (1) (Cao, Fu, & Zhao, 2013), (20S,24R)-dammarane-20,24-epoxy- 3β,6α,12β,25-tetraol (2) (Wang, Bi, & Xu, 2013), 20(R)-dammarane-3β,12β,20,25-tetraol (3) (Ma & Yang, 2015), 20(R)-protopanaxadiol (4) (Ma & Yang, 2015), 20(R)-dammarane-3β,12β,20,25- tetrahydroxy-3β-O-β--glucopyranoside (5) (Ma & Yang, 2015), β-sitosterol (6), oleanolic acid (7) (Seebacher, Simic, Weis, Saf, & Kunert, 2003), 20(S)-protopanaxadiol (8) (Ma & Yang, 2015).

Fig. 1

Chemical structures of isolated compounds 1–8.

α-Glucosidase inhibition assay

α-Glucosidase and α-amylaseare are important enzymes in starch decomposition and intestinal absorption (Meng et al., 2016). The in vitro inhibitions against α-glucosidase of eight compounds were analyzed. As shown in Fig. 2A, seven samples displayed good inhibitory effects in a dose dependent manner. Among them, compounds 5 and 4 showed significant activities with IC50 values of 0.22, and 6.26 µmol/L, respectively, compared with IC50 (9.49 µmol/L) of acarbose in Table 2. The inhibitory activities (5: 86.93%; 4: 67.82%, at 100 µmol/L). Compounds 1, 3, 7, and 8 displayed moderate inhibition with IC50 values of 22.72, 15.42, 41.04, and 12.49 µmol/L, respectively. Compound 6 showed weak inhibitions at IC50 value of 69.41 µmol/L.

Fig. 2

Hypoglycemic activities of isolated compounds 1–8 (mean ± SD, n = 3). A: Inhibitory activities against a-glucosidase. B: Inhibitory activities against PTP1B. C: Lineweave–Burk double reciprocal curve of compound 5 against a-glucosidase. D: Lineweave–Burk double reciprocal curve of compound 5 against PTP1B. *P < 0.05, denotes a significant effect, compared to acarbose or Na3VO4.

Table 2

Inhibitory activities of compounds 1−8.

Compounds	IC50 /(µmol·L−1)
	α-Glucosidase	PTP1B
1	22.72 ± 0.79	27.23 ± 2.36
2	> 100	23.63 ± 5.09
3	15.42 ± 0.87	10.39 ± 0.21
4	6.26 ± 1.59	6.21 ± 0.21
5	0.22 ± 0.21	5.91 ± 0.38
6	69.41 ± 0.03	> 100
7	41.04 ± 0.34	18.99 ± 1.46
8	12.49 ± 1.22	13.38 ± 0.88
Acarbose	9.49 ± 0.18	−
Na3VO4	−	26.20 ± 1.78

PTP1B inhibition assay

PTP1B plays a major role in modulating both insulin sensitivity and fuel metabolism, establishing it as a potential therapeutic target in the treatment of type 2 diabetes and obesity (Elchebly et al., 1999). The roles of compounds 1−8 against PTP1B were assessed using enzyme activity assay. It was indicated that 5 exhibited the highest inhibition effect (IC50 5.91 µmol/L), its inhibitory activity was 81.76% at 150 µmol/L. All the samples displayed good inhibitory effects in a dose-dependent manner (Fig. 2B). In Table 2, we also observed compounds 1, 6 and 7 showed significant activities with IC50 values of 10.77, 15.02, and 20.65 µmol/L, respectively, compared with IC50 (26.20 µmol/L) of Na3VO4. Compounds 2 and 3 displayed moderate inhibition at IC50 values of 37.46 and 49.12 µmol/L, respectively. 4 showed little inhibitions at IC50 value of 62.19 µmol/L.

Enzyme kinetic assay

The enzyme kinetics studies revealed that compound 5 inhibited α-glucosidase and PTP1B with characteristics typical of competitive inhibitors (Figs. 2C and D). It showed that, with increasing concentration of 5, K increased significantly, whereas the V remained unaltered, which indicated they were competitive inhibitions for α-glucosidase and PTP1B.

Docking results of compound 5 and PTP1B

Molecular docking simulation of compound 5 to PTP1B space was discussed. The 12-OH and 20-OH interacted with Ala217 and Tyr46, respectively (Fig. 3). The 3′-OH and 6′-OH in glucose formed hydrogen bonds with Arg254 and Ser28. And the molecule was also by π–π interaction with Val49, Phe182, Gly259, Val49 and Tyr46 of PTP1B. The inhibitor could bind to amino acid residues near the catalytic center, which may be the reason for its good PTP1B inhibition, thus achieving hypoglycemic activity. This result indicates that glucoside may be the better skeleton than the aglycone for the inhibition of PTP1B. The findings also confirmed that the number and substitutional position of OH in the skeleton were important in the PTP1B inhibition activity (Xiao et al., 2017).

Fig. 3

Three-dimensional structure of compound 5 and PTP1B (A) and interaction between compound 5 and PTP1B (B).

By comparing the test compounds, it was found that the sequence of inhibitory activities were as following: 5 (IC50 = 5.91 µmol/L) > 4 (IC50 = 6.21 µmol/L) > 3 (IC50 = 10.39 µmol/L) > 8 (IC50 = 13.38 µmol/L) > 7 (IC50 = 18.99 µmol/L) > 2 (IC50 = 23.63 µmol/L) > 1 (IC50 = 27.23 µmol/L) > 6 (IC50 > 100 µmol/L), which suggested that the compound with more OH had better PTP1B inhibition activity.

Conclusion

In this work, we isolated eight compounds and identified their structures. To evaluate their hypoglycemic activity, the inhibitory activities were tested against α-glucosidase and PTP1B in vitro, which are directly related to the absorption or metabolism of glucose. Compound 5 is ginseng triterpenoid isolated from the acid hydrolysates of total saponins from P. quinquefolius for the first time. Further research demonstrated that 5 had significant inhibition activities. Based on the results, the discovery of active dammarane-type triterpenoids may provide an opportunity to develop P. quinquefolius as a potential source for type 2 diabetes treatment.

Declaration of Competing Interest

The authors declare no conflict of interests.