Anti-lipase activity from Polygonum multiflorum : An in vitro and in silico study

Abstract

A study on the phytochemical composition of the roots of Polygonum multiflorum collecting in Vietnam and an evaluation of its capacity to inhibit pancreatic lipase enzyme was conducted. In total, 12 metabolites (1–12) were isolated from anti-lipase activity-guided fractionation of P. multiflorum. The isolated compounds (1–12) were identified as trans-resveratrol (1), methyl protocatechuate (2), methyl gallate (3), catechin (4), epicatechin (5), myrciaphenone A (6), quercetin (7), kaempferol (8), apigenin (9), luteolin (10), tricin (11), and afzelin (12). This is the first time Compound 12 has been reported from P. multiflorum. However, 12 second metabolites (1–12) were tested in vitro to assess the anti-lipase potential. The results revealed Compound 7 exhibited the most significant lipase inhibition, with an IC₅₀ value of 0.30 ± 0.04 μM. Following closely, Compounds 6, 8, 10, and 12 displayed IC₅₀ values of 5.38 ± 0.65, 0.72 ± 0.10, 0.79 ± 0.19, and 1.22 ± 0.18 μM, respectively. For comparison, orlistat (positive control), the most common anti-obesity drug available on the market, has an IC₅₀ value of 0.91 ± 0.10 μM. Molecular docking was applied to explore the affinity and interactions between active compounds 6–8, 10, and 12 and proteins associated with obesity, namely pancreatic lipase–colipase complex (PLCC) and pancreatic lipase-related protein 1 (PLRP1). In addition, absorption, distribution, metabolism, excretion, and toxicity (ADMET) were used in predictions. The results indicate that the active compounds isolated from P. multiflorum could be further in-depth research on lipase inhibition.

Introduction

Lipases are a group of enzymes responsible for the hydrolysis of triacylglycerols (triglycerides), the primary form of fat storage in the human body. Pancreatic lipase, secreted by the pancreas, is the primary lipase responsible for the breakdown of dietary triglycerides in the small intestine. This enzyme catalyzes the hydrolysis of triglycerides into absorbable free fatty acids and monoglycerides, which can then be transported across the intestinal epithelium and taken up by the body for various physiological ¹ processes. One such approach that has garnered substantial research interest is the inhibition of pancreatic lipase, a key enzyme central to the digestion and absorption of dietary ² fats. ³ If the initial movement of triglycerides from the intestinal lumen could be blocked, it may help prevent hyperlipidemia-related diseases. The development of potent and selective pancreatic lipase inhibitors has been a focus of considerable research in obesity management. Synthetic lipase inhibitors, such as orlistat, have been approved and used clinically to treat ⁴ obesity. However, searching for natural, plant-derived compounds with lipase-inhibiting properties has gained momentum due to their potential for improved safety profiles and diverse ⁵ bioactivities. Therefore, natural inhibitors of pancreatic lipase could be a useful agent for treating ⁶ obesity.

Polygonum multiflorum (synonym Reynoutria multiflora) belongs to the Polygonaceae family and is a famous herb with a long history of medicinal applications in traditional Eastern pharmacology. It is one of the most extensively used traditional medicines across China, East Asia, and Southeast ⁷ Asia. Ancient Chinese medical records have extensively documented the numerous therapeutic benefits of P. multiflorum, such as anti-hyperlipidemic, antioxidant, anti-aging, enhancement of liver and kidney function, anti-inflammatory, anticancer, cardiovascular protection, and hepatoprotective ⁸ effects. Extensive phytochemical investigations on P. multiflorum have revealed the presence of a diverse array of secondary metabolites, including glycosides, flavonoids, fatty acids, anthraquinones, tannins, aldehydes, stilbenes, phospholipids, ketones, and alcohols, all exhibiting a wide range of ⁹ bioactivities. Most of the current studies suggest the stilbene glycosides from P. multiflorum play the most important role in the total cholesterol-lowering effects among all the plant’s chemical constituents. Other reports have also noted favorable triglyceride-reducing effects of ingredients, such as emodin and physcion from P. multiflorum.^10,11 However, there has been a notable lack of research exploring the potential anti-lipase activity of other isolated chemical constituents from P. multiflorum, such as phenolic glycosides, phenolic acids, and especially flavonoids. Compounds from natural resources could affect the activity of pancreatic ¹² lipase, ¹³ an important enzyme involved in the absorption of triglycerides from the small intestine into the enterocytes. With the existing research gap and anti-inflammatory results from our previous ¹⁴ study, a continuous investigation was undertaken. This investigation focused on the chemical constituents of P. multiflorum and their ability to inhibit pancreatic lipase, both in vitro and in silico.

Results and discussion

Determination of isolates from P. multiflorum

P. multiflorum methanol (MeOH) extract has been fractionated into n-hexane, dichloromethane (CH₂Cl₂), and ethyl acetate (EtOAc)-soluble fractions. However, 12 compounds (1–12) were obtained from the EtOAc fraction (Figure 1).

Figure 1.

Structure of the isolates (1–12) from P. multiflorum.

Compounds 8 and 12 were obtained as yellow powder. The ¹H NMR spectra of 8 and 12 displayed characteristic signals for six aromatic protons (H-6/H-8/H-2′/H-3′/H-5′/H-6′) of two benzene rings, along with hydroxyl groups at C-5, C-7, and C-4′. Their ¹³C NMR spectra revealed signals of oxygen-substituted carbons at C-2 and C-9, as well as two quaternary carbons at C-10 and C-1′ (Figure 1 and Supplementary material). The presence of an olefinic ((C-2) and (C-3)) and a ketone (C-4) group indicated Compounds 8 and 12 to be flavonols, ¹⁵ respectively (Figure 1 and Supplementary material). Analysis of the NMR spectra in detail, it was found that Compound 12 possessed a rhamnose moiety (δ_H 5.36 (1H, br s, H-1′′); 4.22 (1H, br s, H-2′′), 3.71 (1H, m, H-5′′), 3.33–3.29 (2H, m, H-3′′/H-4′′), 0.91 (3H, d, J = 5.2 Hz, 5′′-CH₃)/δ_C 103.6 (C-1′′), 73.3 (C-5′′), 72.2 (C-3′′), 72.1 (C-2′′), 72.0 (C-4′′), 17.8 (C-5′′)) at C-3 (Figure 1 and Supplementary material). Comparing the ¹H NMR and ¹³C NMR data of these compounds with literature data, Compounds 8 and 12 were identified as kaempferol ¹⁵ and afzelin, ¹⁶ respectively. The remaining compounds including trans-resveratrol (1), methyl protocatechuate (2), methyl gallate (3), catechin (4), epicatechin (5), myrciaphenone A (6), quercetin (7), apigenin (9), luteolin (10), and tricin (11) have been isolated from the previous study. ¹⁴

The pancreatic lipase inhibitory activity of the isolated compounds was evaluated, using 2,4-dinitrophenyl butyrate (DNPB) as a substrate. ¹⁷ Compound 7 showed the most inhibitory activity, with an IC₅₀ value of 0.30 μM, followed by Compounds 8, 10, and 12 with IC₅₀ values of 0.72, 0.79, and 1.22 μM, respectively (Table 1). Compound 6 showed moderate inhibitory activity, with an IC₅₀ value of 5.38 μM. In addition, orlistat, a positive control inhibited the pancreatic lipase with an IC₅₀ value of 0.91 μM (Table 1).

Table 1.

Anti-lipase activity of Compounds 1−12.

Compound	Anti-lipase (IC50, μM) a
1	ND b
2	ND
3	ND
4	ND
5	ND
6	5.38 ± 0.65
7	0.30 ± 0.04
8	0.72 ± 0.10
9	ND
10	0.79 ± 0.19
11	ND
12	1.22 ± 0.18
Orlistat c	0.91 ± 0.10

a

Results are represented as IC₅₀ value (μM);

b

ND: Not determined;

c

Positive control; Values are mean ± SD (n = 3).

All active compounds exhibit a flavonoid structure, which is essential for their inhibitory activity. The lipase inhibitory effect of flavonoids is influenced by the number and position of phenolic hydroxyl groups; a higher number generally enhances inhibitory activity. In addition, the overall size of the molecule and the positioning of hydroxyl groups within the rings significantly impact lipase inhibition.^18,19 Zhang et al. indicated that the enzyme structure becomes less stable in the presence of flavonoids, reducing substrate affinity and leading to a 75% inhibition of lipase at a concentration of 0.27 mg mL⁻¹. ²⁰ Quercetin (7), with its multiple hydroxyl (–OH) groups, exhibits the strongest inhibitory activity, attributed to its optimal combination of these groups that enhance binding affinity through hydrogen bonding. Specifically, the hydroxyl group at position 3′ (3′-OH) appears to favor effective binding to pancreatic lipase.^21,22 Conversely, the nonplanarity of some flavonoids may weaken their binding affinity. Other studies have shown that the C2=C3 double bond in the C ring of flavones increases π-conjugation, thereby maintaining the planarity of the flavonoid structure and promoting π–π interactions with lipase compared to catechin and epicatechin. ²³ Apigenin (9), in contrast, has fewer hydroxyl groups than more potent inhibitors such as quercetin, which critically affects its binding affinity. Kaempferol (8) and luteolin (10) also contain hydroxyl groups, but their varying positions yield moderate activity. Luteolin (10), with its hydroxyl group at position 3′ (3′-OH), exhibits slightly higher activity than kaempferol. These findings align with previous studies.^24,25Afzelin (12) shows reduced activity compared to quercetin, indicating that glycosylation may hinder inhibitory potency. While luteolin (10) and afzelin (12) have similar hydroxyl patterns, structural differences, such as afzelin’s additional sugar moiety, may reduce binding due to changed planarity. Disruption of molecular planarity can significantly increase the solubility and bioactivity of flavonoids.^26,27 Myrciaphenone A (6) demonstrates the weakest activity, likely due to its structural differences, including fewer hydroxyl interaction sites and a glycosylation that diminishes inhibitory activity.^28,29 The size of the compounds also influences their ability to fit into the lipase active site; quercetin’s smaller and more flexible structure allows for better accommodation compared to myrciaphenone A (6). While quercetin’s catechol structure enhances its inhibitory potential, myrciaphenone A (6) lacks such features, resulting in weaker activity. ³⁰

Previous studies have indicated that P. multiflorum extract showed relatively strong inhibition of anti-hyperlipidemic effects, and it is widely used in the prevention and treatment of hyperlipidemia in traditional Chinese medicine. ³¹ While the P. multiflorum extract as a whole showed relatively potent inhibition of pancreatic lipase, the isolated compounds catechin (4) and epicatechin (5) did not demonstrate meaningful inhibitory activity against this enzyme. ³² Other research also agrees that individual constituents, such as resveratrol (1), catechin (4), and epicatechin (5) exhibited insignificant inhibitory effects on pancreatic lipase activity (IC₅₀ > 20 μM). ³³ In contrast to our findings, scientists have screened 37 traditional Chinese medicinal herbs for anti-lipase properties and have found that quercetin (7) and luteolin (10) demonstrated insignificant inhibitory activity against pancreatic lipase. ³⁴ To the best of our knowledge, this study represents the first evaluation of the anti-lipase activity of myrciaphenone A (6), which has been found to exhibit moderate inhibition of pancreatic lipase with an IC₅₀ of 5.38 ± 0.65 μM (Table 1). In addition, myrciaphenone A (6) and its derivatives also have antibacterial and antifungal effects. ³⁵ The flavonoids are known to have many biological effects, notably anti-cancer effects. ³⁶ However, their anti-lipase effects are minimal. The lipase inhibitory effects of quercetin (7) observed in this study appear stronger than those reported in previous research. ³⁴ This suggests that the anti-lipase potency of quercetin may be context-dependent and influenced by factors, such as the experimental conditions or the specific assay system employed. In addition, apigenin (9) demonstrated good anti-lipase activity in the current investigation and is consistent with the findings from earlier studies. These collective results indicate that plants rich in apigenin may have a high potential for anti-lipase and anti-obesity therapeutic formulations.^37,38 Tricin (11) also exhibited good anti-lipase activity in the present research, with a potency comparable to previously reported results. This corroborates the potential of tricin-containing plant sources as a rich resource for identifying novel pancreatic lipase inhibitors. Overall, this study provides new insights into the anti-lipase properties of several plant-derived compounds, including the first report on the inhibitory effects of myrciaphenone A (6).

The results in Table 1, structure–activity relationship, and previous studies^34,37,38 have shown that Compounds 6–8, 10, and 12 have the potential to inhibit lipase. However, the use of molecular docking to explore the affinity and interactions between active compounds (6–8, 10, and 12) and proteins associated with obesity has not been studied. Therefore, these compounds (6–8, 10, and 12) and proteins, namely pancreatic lipase–colipase complex (PLCC) and pancreatic lipase-related protein 1 (PLRP1) were chosen for molecular docking studies. The PLCC is an essential enzyme system involved in the digestion and absorption of dietary lipids (fats and oils). ³⁹ PLRP1 is a pancreatic lipase gene family member. This complex, secreted by the pancreas, plays a critical role in normal digestion. Disruptions in the production or activity of the lipase–colipase complex can lead to the malabsorption of lipids and related digestive or nutritional problems.^40,41 In this study, these complexes are the targets for the docking simulation. The best-docked poses of the ligands (Compounds 6−8, 10, and 12) with pancreatic lipase were observed. These findings are displayed in Tables 2 and 3, which outline the interactions of these compounds with the protein. In addition, the 3D representations of the docked positive reference, orlistat, are shown in Figures 1 and 2.

Table 2.

Molecular docking results of the PLCC.

Ligand	Compound	Best binding energy (kcal mol−1)	H-bond interactions	Hydrophobic interactions
1	Myrciaphenone A (6)	−7.479	Gly76, His263	Phe215
2	Quercetin (7)	−9.726	Phe215, Arg256	Tyr114, Phe215, Ala260, His263, Leu264
3	Kaempferol (8)	−9.456	Ser152	Tyr114, Ala178, Ala260, His263, Leu264
4	Luteolin (10)	−9.525		Tyr114, Phe215, Ala260, His263
5	Afzelin (12)	−8.430	Asp79, Arg256, Ala259	Phe77, Ala259
6	Orlistat	−6.760	Gly76, Phe 77, Tyr114, Ser152, Pro180, His263	Phe77, Ile78, Tyr114, Phe215, Trp252, Ala259, His263

Table 3.

Molecular docking results of the PLRP1.

Ligand	Compound	Best binding energy (kcal mol−1)	H-bond interactions	Hydrophobic interactions
1	Myrciaphenone A (6)	−6.469	Asp24, Leu70, Leu72, Ser73	Leu71
2	Quercetin (7)	−7.244	Leu70, Ser73, Gln149	Arg55, Leu71, Leu72
3	Kaempferol (8)	−7.235	Leu70, Gln149	Arg55, Leu71, Leu72
4	Luteolin (10)	−7.357	Gly272, Phe276	Ile97, Lys99
5	Afzelin (12)	−7.281	Tyr59, Asn65, Ser103, Tyr106	Phe67, Val105, Cys122, Lys126
6	Orlistat	−5.376	Glu23, Asp24, Gln146, Gln149	Glu23, Asp24, Pro49, Arg55, Ile69, Leu71, Leu72

Figure 2.

Binding of compounds on PLCC.

To assess the affinity of the ligands toward the PLCC and PLRP1 targets, the binding energies of the ligands (Compounds 6−8, 10, and 12) were calculated. In the PLCC simulations, Compounds 7, 8, and 10 were predicted to interact with the PLCC at similar binding energy values of −9.726, −9.456, and −9.525 kcal mol⁻¹, respectively (Figure 2). However, the binding sites and the hydrogen bonding and hydrophobic interactions exhibited little similarity among these compounds, as shown in Table 2.

The binding energies for the ligands (Compounds 6−8, 10, and 12) were also calculated for the PLRP1 target (Figure 3 and Table 3). In the PLRP1 simulations, Compounds 7 and 8 shared the same hydrophobic interactions with Arg55, Leu71, and Leu72, with close binding energies of −7.244 and −7.235 kcal mol⁻¹, respectively. Compound 7 (quercetin) interacted with the PLRP1 receptor through three hydrogen bonds, located at Leu70, Ser73, and Gln149. Compound 8 (kaempferol) interacted with the PLRP1 receptor through two hydrogen bonds, located at Leu70, Gln149, and without the interaction at Ser73. Compound 7 formed three hydrogen bonds with Leu70, Ser73, and Gln149, and three hydrophobic interactions with Arg55, Leu71, and Leu72. Compound 8 formed two hydrogen bonds with Leu70 and Gln149, and three hydrophobic interactions at Arg55, Leu71, and Leu72. Both Compounds 7 and 8 have a similar number of hydrophobic interactions formed with the PLRP1 receptor, each having three hydrophobic interactions.

Figure 3.

Binding of compounds on PLRP1.

The docking simulations of the studied ligands with the PLCC have indicated the potential of Compounds 6−8, 10, and 12 as pancreatic lipase inhibitors. This is evidenced by the strong interactions between the ligands (Compounds 6−8, 10, and 12) and the residues in the receptor’s active site, as outlined in Tables 2 and 3.

“Lipinski’s rule of five” was used to determine the druggability of active compounds.^42–44 In this study, all five studied compounds (6−8, 10, and 12) have molecular masses smaller than 500 daltons (Table 4). The log p-values are also smaller than 5, the number of hydrogen bond donors does not exceed 5, and the number of hydrogen bond acceptors of these compounds does not exceed 10 (Table 4). However, afzelin (12) has 7 hydrogen bond donors and 11 hydrogen bond acceptors, which is not the optimal profile for a drug-like compound (Table 4).

Table 4.

Physicochemical properties of P. multiflorum components and orlistat analyzed with SwissADME.

Ligand	Compound	MW (g moL−1)	Log P	nHBD	nHBA	TPSA	MR	Lipinski violation	Log S	nRotB
1	Myrciaphenone A (6)	330.29	−0.91	6	9	156.91	74.83	1	−1.21	4
2	Quercetin (7)	302.24	1.23	5	7	131.36	78.04	0	−3.16	1
3	Kaempferol (8)	286.24	1.58	4	6	111.13	76.01	0	−3.31	1
4	Luteolin (10)	286.24	1.73	4	6	111.13	76.01	0	−3.71	1
5	Afzelin (12)	448.38	0.22	7	11	190.28	109.00	2	−3.33	3
6	Orlistat	495.75	7.05	1	5	81.70	145.36	1	−7.60	24

MW: molecular weight; log P: log of octanol/water partition coefficient; nHBD: number of hydrogen bond donor(s); nHBA: number of hydrogen bond acceptor(s); TPSA: total polar surface area; MR: molar refractivity; log S: log of solubility; nRotB: number of rotatable bond(s).

In addition to the Lipinski parameters, other physicochemical properties were also investigated to assess the drug-like potential of these compounds. For good oral bioavailability and intestinal absorption, the number of rotatable bonds and the total polar surface area (TPSA) value should not exceed 10 and 140 Å, respectively.^43–45 The results in Table 4 showed good physicochemical properties of the studied compounds (6−8, 10, and 12). Specifically, the aqueous solubility (log S) values were −3.16, −3.31, and −3.71 for Compounds 7, 8, and 10, respectively, indicating they are all moderately soluble. The log Kp (log of skin permeability) values for the three compounds (7, 8, and 10) were −7.05, −6.70, and −6.25, respectively (Table 5).

Table 5.

ADME predictions of investigated compounds.

Ligand	Compound	Log Kp (cm s−1)	GI Abs	BBB per	Inhibitor Interaction
					P-gp substrate	CYP1A2 inhibitor	CYP2C19 inhibitor	CYP2C9 inhibitor	CYP2D6 inhibitor	CYP3A4 inhibitor
1	Myrciaphenone A (6)	−9.00	Low	No	No	No	No	No	No	No
2	Quercetin (7)	−7.05	High	No	No	Yes	No	No	Yes	Yes
3	Kaempferol (8)	−6.70	High	No	No	Yes	No	No	Yes	Yes
4	Luteolin (10)	−6.25	High	No	No	Yes	No	No	Yes	Yes
5	Afzelin (12)	−8.42	Low	No	No	No	No	No	No	No
6	Orlistat	−2.25	Low	No	Yes	No	No	Yes	No	No

Log K_p: log of skin permeability; GI Abs: Gastrointestinal absorption; BBB Per: Blood–brain barrier permeability; P-gp: P-glycoprotein; CYP: cytochrome-P.

In summary, the analysis of the drug-like properties of the studied compounds indicates that Compounds 7, 8, and 10 exhibit favorable characteristics, with no major violations of Lipinski’s rule of five or other key physicochemical parameters. This suggests their potential for further development as pancreatic lipase inhibitors. ⁴⁶ The DL-AOT prediction server ⁴⁷ was employed to investigate the acute oral toxicity of the studied compounds. Myrciaphenone A (6) and quercetin (7) were categorized in the “Warning” group with LD₅₀ values (mg kg⁻¹) of 3.61 and 2.79, respectively, while the remaining compounds, including the reference one of orlistat, were classified as “Caution” (Table 6). This points to the potential of kaempferol (8), luteolin (10), and afzelin (12) as drug candidates to treat obesity.

Table 6.

Toxicity of investigated compounds predicted by DL-AOT Prediction Server.

Ligand	Compound	LD50 (mg kg−1)	Toxicity
1	Myrciaphenone A (6)	3.61	Warning
2	Quercetin (7)	2.79	Warning
3	Kaempferol (8)	2.98	Caution
4	Luteolin (10)	3.17	Caution
5	Afzelin (12)	3.02	Caution
6	Orlistat	4.00	Caution

Conclusion

Overall, 12 metabolites, namely trans-resveratrol (1), methyl protocatechuate (2), methyl gallate (3), catechin (4), epicatechin (5), myrciaphenone A (6), quercetin (7), kaempferol (8), apigenin (9), luteolin (10), tricin (11), and afzelin (12) were isolated from the roots of Vietnamese P. multiflorum. Their anti-lipase activity was evaluated using DNPB as a substrate. Afzelin (12) was isolated from P. multiflorum for the first time. Quercetin (7) showed the strongest inhibitory activity (IC₅₀ = 0.30 μM), followed by kaempferol (8), luteolin (10), and afzelin (12) with IC₅₀ values of 0.72, 0.79, and 1.22 μM, respectively. Compound 6 showed moderate inhibitory activity, with an IC₅₀ value of 5.38 μM. Molecular docking results revealed Compounds 6−8, 10, and 12 as inhibitors of PLCC and PLRP1 receptors with the binding-free energy −7.479, −9.726, −9.456, −9.525, and −8.430 kcal mol⁻¹, and −6.469, −7.244, −7.235, −7.357, and −7.281 kcal mol⁻¹, respectively. The findings are significant in identifying the phytochemical composition and expanding the knowledge of the bioactivities of the chemical ingredients of P. multiflorum roots. These results indicate the potential of the plant’s chemical constituents for use in pharmaceutical applications.

Experimental

Chemistry

General experimental procedure, plant material, extraction and isolation, and physiochemical and NMR data of isolated compounds (1−12), please see Supplementary materials file.

Biology

Anti-lipase activity. Lipase activity was measured using 2,4-dinitrophenyl butyrate (DNPB) as a substrate. ¹⁷ Porcine pancreatic lipase stock solutions (1 mg mL⁻¹) were prepared in 0.1 mM potassium phosphate buffer (pH 6.0) and then stored at −20°C. To determine lipase inhibitory activity, isolated compounds (at different concentrations) were preincubated with the enzyme for 1 h in potassium phosphate buffer (0.1 mM, pH 7.2, combined with 0.6 mL/100 mL Tween 80) at 30°C before assaying the enzyme activity. ⁴⁸ The reaction was then started by adding 0.1 mL 25 mM DNPB, all in a final volume of 5.0 mL. After incubation at 30°C for 5 min, the amount of 2,4-dinitrophenol released in the reaction was measured at 360 nm. The activity of the negative controls was also checked with and without inhibitor. The inhibitory activity (I) was calculated according to the following formula: ⁴⁹

$$I(\%)=\left(1-\frac{B-b}{A-a}\right)\times 100$$

where A is the enzyme activity without inhibitor, a is the negative control without inhibitor, B is the activity of the enzyme with inhibitor, and b is the negative control with inhibitor.