Sage Journals: Discover world-class research

Abstract

Objective: Aruncus dioicus has been known by the scientific literature and folk experiences for its diverse biological activities, including anti-hyperglycemic effects. Methodology: The aerial parts of the plant collected from Quan Lan Island (Vietnam) were subjected to a methodical theory-experiment investigation for its chemical composition and biological potentials. Results: Firstly, experimental isolation and spectroscopic characterization identified the compositional compounds, ie sambunigrin (1), prunasin (2), uridine (3), and adenosine (4). Secondly, their elucidated structures were predicted with promising bio-chemo-pharmacological potentiality by different computational platforms, ie: docking simulation (docking scores < 10 kcal.mol⁻¹); quantum calculation (dipole moments < 3 Debye); QSARIS model (satisfying Lipinski's rule of five); SwissADME model (satisfying Pires’ interpretations). Finally, the compounds (1-4) were under accumulative purification and in vitro tests against diabetes-related enzymes, ie: protein tyrosine phosphatase 1B (1 with lowest IC₅₀ value 0.39 ± 0.26 μM, 2 with no activity) and α-glucosidase (1 with lowest IC₅₀ value 44.89 ± 0.93 μM, 2 with no activity); also, the inhibitory kinetics based on Lineweaver-Burk and Dixon plot experiments revealed the competitive inhibition mode of 1 against PTP1B (K_i value 0.49 μM). Conclusions: Altogether, the results obtained suggest that A. dioicus and its bioactive compounds, especially sambunigrin (1), uridine (3), and adenosine (4), could be considered as a natural source for further research and development of anti-diabetic agents.

Keywords

diabetes obesity

Introduction

Diabetes mellitus is a worldwide health concern due to chronic glucose-related metabolic disorder. The previous evidences from clinical trials in the literature indicate that the disease not only results in hyperglycemia-related conditions but also might induce further complications.^1–5 Based on the mechanism, it can be categorized into two main types: type 1 (by pancreatic β-cells deterioration) and type 2 (by insulin-metabolic abnormal activity); the latter accounts for 90–95% of cases.⁶ Despite the well-rounded knowledge, the certain cause is still highly debated among the scientific community, possibly due to genetic abnormalities, pathologic disorders, clinical conditions, or gestational failures.⁷ In effect, the condition is putting significant burden on the healthcare system.

Until now, diabetes mellitus has been treated by symptomatic relief rather than therapeutic cure. The current medicines available on the market only serve as the remedies. They only provide temporary effects against undesirable activities of certain biological enzymes, such as insulin- and glucose-based enzymes. For instance, the phosphorylation process is known to be catalyzed by protein tyrosine phosphatase 1B (PTP1B), which further stimulates the responsiveness of insulin in cells to uptake glucose.⁸ On the other side, α-glucosidase is the bio-catalyst for breaking down -(1/4) and -(1/6) bonds in starch and disaccharide molecules through hydrolysis; meanwhile, α-amylase is the major form of salivary and pancreatic amylase found in humans and other mammals, responsible for the hydrolysis of α-linked polysaccharides (eg starch and glycogen) into shorter chains (eg dextrins and maltose). Their physicochemical roles induce the postprandial spikes from digestive carbohydrates.^9,10 Given that, a solution to reduce glucose uptake can be possible; in particular, negative regulatory effects on the hyperglycemic activity or insulin signaling pathway are achievable with effective bio-inhibition against these proteins. In the commercial market, many hypoglycemic drugs have been introduced targeting glucosidases (eg Sulfonylureas, Biguanides, or Acarbose^11,12) or the insulin-related enzyme (e.g Trodusquemine¹³). However, they are known either with moderate-to-severe side effects, such as diarrhea and flatulence, or for its expensive price. Therefore, looking for alternative anti-diabetic agents, especially whose sources are ubiquitous in nature and friendly to the body, is still needed.

Goat's beard (Rau re in Vietnamese) has the scientific name of Aruncus dioicus (Walter) Fernald (synonyms A. sylvester Kostel. ex Maxim.; A. vulgaris (Maxim.) Raf. ex H. Hara), belonging to the family Rosaceae.^14,15 This species is widely distributed in temperate regions of the Northern Hemisphere from Central Europe and the Caucasus to the Himalayas, China, Korea, Japan, Russia, Korea (Ulleung-do) and North America.^16–20 It is a perennial herb up to 3 m tall with drooping flowers with numerous small, single, capsule, and sepalless clusters.²¹ Goat's beard is known as a source of food, cosmetics, and medicinal plants in traditional Korean and Chinese medicine.^22–28 However, regarding the Vietnamese A. dioicus, the knowledge on the active components and the biological activities is still poorly collected in the literature.

Evidence from preceding studies in the literature revealed the potential of the plant given its both compositional richness and bio-active diversity. Regarding the chemical bioavailability, a variety of phytochemicals have been identified in the aerial part and underground part of Goat's beard, including quercetin and kaempferol derivatives,^29–33 3,4-dicaffeoylglucopyranosides,^31,33 phenolic acids, especially hydroxycinnamic acid derivatives,^33–36 tannins,³⁷ monoterpenoids^29,32 and cyanide compounds (mainly prunasin).^29,38 Particularly, the edible young shoots were found as a rich source of polyphenols; in which, 24 polyphenols were detected in the study by Fusani et al in 2016.³³ Further, Zhao et al reported the isolation of palmitic acid, 10-nonacosanol, pentacosan-1-ol, phytol, β-sitosterol, β-sitosterol-3-O- β-D-glucopyranoside, 2,4-dihydroxycinnamic acid, and hyperoside.³² Some other works have demonstrated to present of lipophilic compounds, ie steroids, fatty acids, alcohols,³² and caffeic acid derivatives.³³ In terms of bio-active potential, experimental findings from numerous studies indicate a profound diversification, including cytotoxic, anti-inflammatory, antioxidant, anti-cancer, antibacterial, neuroprotective, and anti-diabetic activities.^{20,23,28,29,39–42} Besides, the plant is rich in saponins, which are thought to be promising for the treatment of chronic degenerative diseases. The whole plant was used as the supplemental product for detoxification, prevent of tonsillitis, and stamina boost.²⁴ Several in vivo studies revealed the reduction of lipid accumulation and the improvement of insulin resistance, observed from C57BL/6 mice fed with high-fat diet supplemented with this plant.^43,44 Also, it seemed to induced weight loss in goats⁴⁵ and alleviated streptozotocin-induced diabetes in rats.⁴¹ Therefore, A. dioicus is considered as a potential natural source for anti-hyperglycemic candidates given its biochemical characteristics and evidences of anti-diabetic effects in particular.

However, the high chemical bioavailability and biological versatility, on the downside, might pose significant challenges in terms of experimental trials; therefore, cost-friendly and time-effective intervals, preliminarily examining the chemo-bio-pharmacological suitability of the extracted phytochemicals, are necessary. Given this, in silico research is of a propitious approach, which is able to harness the computer power to compensate for equivalent lab-based wet screening. In fact, the promising candidates can be quickly selected with desirable properties by different computational tools. For instance, based on molecular mechanics, docking technique with its simple computational algorithms can be an effective method for prediction of ligand-protein interaction; the cons often refer to its neglect of bio-medium factors, eg dipole compatibility or polar sensibility. Nevertheless, the former can be complemented by the results from quantum chemical calculations; while, the latter can be solved by the incorporation of proper arguments on physicochemical properties.^46,47 Also, quantum chemical computation might provide other useful molecular properties for argument on intermolecular and chemical tendencies. Furthermore, pharmacokinetics and pharmacological properties of a chemical structure are able to be predicted by statistically regressive models, thus evaluating drug-likeness and medicinal chemistry friendliness. Altogether, these results can provide a reliable assessment of the biocompatibility and pharmaceutical suitability of numerous compounds in the cost-reduced and time-efficient manners.

In this work, A. dioicus was subject to a theory-experiment combinatory study. Firstly, lab-based experiments were carried out for its phytochemical characterization; the obtained structures were preliminarily used as the input for computational screening for biological and pharmacological potentials. Afterwards, a variety of computer-based platforms were utilized to predict the bio-chemo-pharmacological suitability and compatibility of the candidates; the arguments served as the justification for further in vitro tests. Finally, the phytochemicals were under accumulative isolation and subjected to enzymatic assays to evaluate their in-practice inhibitory activities against diabetes-related proteins, thus confirming the results from computational efforts. To better serve the readership, this report contents are structured by methodological contexts (ie experimental and computational parts) rather than following a chronological flow.

Results

Structural Elucidation

Compounds 1 and 2 were isolated as white powder and reacting positively with Grignard reagents. The ¹H- and ¹³C-NMR spectra of 1 and 2 revealed the presence of a benzene ring (C-3 to C-8), together with those of methine (C-2), nitrile (C-1), and sugar moiety (C-1′ to C-6′); these suggested 1 and 2 to be mandelonitrile glycosides.⁴⁸ The anomeric proton was detected as doublets δ_H 4.67 (1H, d, J = 7.6 Hz, H-1′) regarding compound 1 and δ_H 4.24 (1H, d, J = 7.2 Hz, H-1′) given compound 2, indicating the β-configuration in glycosidic linkage. Compounds 1 and 2 have similar structural formulae but different optical configurations at the chiral center C-2, explicitly distinguished by a notable chemical shift differences in the cyanogenic methine protons [δ_H 6.04 ppm for compound 1 and 5.91 ppm for compound 2].⁴⁹ In addition, 1 possesses a negative optical rotation value, ie −34.2 (concentration: 0.01; solvent: MeOH); while, the corresponding value for 2 is −15.6 (c 0.01, MeOH). Comparing ¹H-, ¹³C-NMR data, and optical rotation values of these compounds to those published in the literature led to the identification of 1 and 2 to be 2S-sambunigrin and 2R-prunasin, respectively.⁵⁰

Compound 3 was isolated as white powder and reacting positively with Grignard reagents. The ¹H-NMR spectrum showed one cis-olefinic group at δ_H 8.01 (1H, d, J = 8.4 Hz, H-6) and δ_H 5.72 (1H, d, J = 8.4 Hz, H-5), an anomeric proton [δ_H 5.92 (1H, d, J = 4.8 Hz, H-1′)], and the existence of a β-glycosyl. Its ¹³C-NMR spectrum displayed 9 carbons in total, which were identified as two carbonyl groups [δ_C 166.2 (C-4) and 152.5 (C-2)], two olefin carbons [δ_C 142.7 (C-6) and 102.7 (C-5)], and one ribose moiety [δ_C 90.7 (C-1′), 86.3 (C-4′), 75.7 (C-2′), 71.3 (C-3′), and 62.3 (C-5′)]. This data revealed a ribose attaching to an uracil via β-N₇-glycosidic bond. Altogether, compound 3 was characterized as uridine after matching its ¹H- and ¹³C-NMR data with the published values.⁵¹

Compound 4 was isolated as white powder and reacting positively with Grignard reagent. The ¹H- and ¹³C-NMR revealed a adenine skeleton linkage to a ribose via β-N₉-glycosidic bond.^52,53 The ¹H-NMR spectrum displayed one singlet aromatic proton at δ_H 8.32 (1H, s, H-3), an olefinic proton [δ_H 8.20 (1H, s, H-8)], one amine [δ_H 4.55 (2H, br s, 1-NH₂)], an anomeric proton [δ_H 5.98 (1H, d, J = 6.6 Hz, H-1′)], and a β-glycosyl. Its ¹³C-NMR spectrum displayed 10 carbon signals, which were characterized with four aromatic carbons [δ_C 157.6 (C-1), 153.5 (C-3), 150.1 (C-5), and 121.2 (C-6)], one olefinic carbon [δ_C 142.0 (C-8)], and one ribose moiety [δ_C 91.3 (C-1′), 88.2 (C-4′), 75.5 (C-2′), 72.7 (C-3′), 63.5 (C-5′)]. Comparing ¹H- and ¹³C-NMR data of compound 4 with those published in the literature led to the structural identification of 4 as adenosine.⁵³

Enzymatic Inhibition

Compounds (1 − 4) were evaluated in vitro for their inhibitory effects against PTP1B and α-glucosidase; the observables are summarized in Table 1. This also includes the results from kinetic investigations regarding PTP1B. The kinetics of the candidates (1, 3, and 4) and ursolic acid inhibitory effects towards PTP1B were analyzed based on a mixed-competitive mode, which had been precedingly described⁵⁴; the results reveal the corresponding inhibition types.

Table 1.

in Vitro Inhibitory Activity of Compounds (1-4) on PTP1B and α-Glucosidase Enzymes.

Compound	Protein tyrosine phosphatase 1B			α-Glucosidase
Compound	IC₅₀ (µM) a	K_i value (µM)	Inhibition type	IC₅₀ (µM) a
1	0.39 ± 0.26	0.47	Competitive	44.89 ± 0.93
2	> 50	- c	-	> 200
3	11.96 ± 0.79	13.09	Mixed-competitive	133.07 ± 0.96
4	30.87 ± 0.44	24.99	Non-competitive	183.13 ± 1.22
Ursolic acid b	3.31 ± 0.19	- c	Mixed-competitive	-
Acarbose b	-	-	-	147.23 ± 1.03

Results are expressed as IC₅₀ values (µM), determined by regression analysis and expressed as the means ± SD of three replicates.

Positive control.

Data not determined.

Overall, all the tested compounds exhibited promising inhibition against PTP1B, except for 2; in particular, the respective IC₅₀ values are 0.39 ± 0.26 (1), 11.96 ± 0.79 (3), and 30.87 ± 0.44 (4) μM. As the positive control in this experiment, ursolic acid registered IC₅₀ value 3.31 ± 0.19 µM. Firstly, 1 (sambunigrin) and 2 (prunasin) are optical isomers yet exhibit noticeably opposite potentials. Particularly, sambunigrin (1) with 2S-configuration was performed the highest inhibitory activity (IC₅₀ value of 0.39 ± 0.26 µM), almost ten-fold regarding that of the positive control (IC₅₀ value of 3.31 ± 0.19 µM); meanwhile, prunasin (2) with 2R-configuration displayed no activity (IC₅₀> 50 µM). Although 3 (uridine) and 4 (adenosine) both contain a ribose unit in their structures, uridine (3) bearing two nitrogen atoms performed an upper inhibitory effect against PTP1B enzyme activity (IC₅₀ value of 11.96 ± 0.79 µM) while adenosine (4) with five nitrogen atoms showed a moderate inhibitory effect (IC₅₀ value of 30.87 ± 0.44 µM).

Figure 1 presents the analyses for Lineweaver-Burk and Dixon experiments under linear concentrations of substrate (p-NPP) with/without the presence of compounds 1, 3, and 4; the kinetic data was analyzed using SigmaPlot (SPCC Inc., Chicago, IL). The mixed-competitive inhibition mode was identified when all the lines intersected above the x-axis and to the left of the y-axis.^55–57; in other words, uridine (3) showed mixed-inhibition mode as all the straight lines intersected on x-axis and behind the y-axis in the Lineweaver–Burk plot.⁵⁸ In case of adenosine (4), the intersection of all lines occurred at a negative point on the x-axis (1/(intensity/min) = 0), indicating a non-competitive inhibition mode. Sambunigrin (1) exhibited a competitive inhibition, characterized by the fact that all the straight lines intersect on y-axis and upper the x-axis in the Lineweaver–Burk plot⁵⁵; this indicates that sambunigrin (1) may directly bind to the active binding site of the enzyme to inhibit the enzyme activity.^57,58

Figure 1.

Lineweaver-Burk plots for the inhibition of compounds (1, 3, and 4) on the protein tyrosine phosphatase 1B-catalyzed hydrolysis of p-NPP.

Figure 2 presents the analyses of Dixon plot experiments used to determine the K_i (inhibition constant) values from the x-axis where the straight lines intersected for the inhibitors.⁵⁶ In particular, the K_i values of 1, 3, and 4 were determined to be 0.47, 13.09, and 24.99 μM, respectively. In this experimental assay, three concentrations of substrate (p-NPP: 2.0, 1.0, and 0.5 mM) and three concentrations of inhibitors (10.0, 1.0, and 0.1 μM) were used for assay.

Figure 2.

Dixon plots for active compounds (1, 3, and 4) used for determining the inhibition constant k_i. K_i values were determined from the negative x-axis value at the point of the intersection of three lines.

Regarding α-glucosidase, the in vitro assays with the compounds (1-4) and acrabose (as the control) were carried out using α-glucosidase enzyme extracted from rat intestinal tract. Acarbose exhibited a consistent inhibitory IC₅₀ value (147.23 ± 1.03 μM) against α-glucosidase with one already reported to the literature.⁵⁹ Overall, a similar pattern of the inhibitory potentials was observed, cf that from PTP1B. For instance, sambunigrin (1) exhibited potent inhibition against α-glucosidase activity with IC₅₀ value 44.89 ± 0.93 μM, almost three-fold higher than acarbose did; meanwhile, prunasin (2) did not show inhibitory effect (IC₅₀ value over 200 μM). Further, uridine (3) exhibits a similar IC₅₀ (133.07 ± 0.96 μM) to the value regarding the positive control, while adenosine (4) is considered to also possess a significant IC₅₀ value (183.13 ± 1.22 μM).

Docking-Based Inhibitability

The results from docking simulation can be utilized for argument on the inhibitory effects of each compound (1-4) against the protein structures (PTP1B, 3W37, and 4W93), whose most susceptible sites are highlighted in Figure 3. In this scope, the primary indicators chosen to assess inhibitory effectiveness are the total docking score (DS) values and the count of hydrogen-like bonds, which respectively represent pseudo values for Gibbs free energy of ligand-protein complex formation and their strong intermolecular bonds.

Figure 3.

Quaternary structures of protein 3W37, PTB1B and 4W93 with the approachable sites by 1–4 and the controlled drug Acarbose (D): site 1 (cyan), site 2 (yellow), site 3 (gray), site 4 (green).

The primary docking parameters are summarized in Table 2; the control drugs are ursolic acid (D1) and acarbose (D2). Regarding the natural sources with rather arbitrary structural characteristics, the average DS values should be useful since they represent the effectiveness of a multi-site inhibition rather than that of a specific site. In principle, simultaneous inhibitory forces to different protein sites might result in denaturation of enzyme shape, thus enzymatic ceases ensuing. Overall, all the candidates are considered as effective bio-inhibitors against the protein structures with some exception; particularly, 2 was proven experimentally with no bioactivity against any enzymes, yet is still predicted computationally with explicit inhibitability. This is understandable as the compound is a chiral isomer, which induces special interaction with other chiral molecules (ie protein structures in this case) in the manner that cannot be covered by the algorithms of docking technique. Therefore, the docking-based predictions might not be highly accurate regarding the optical isomers (ie 1 and 2). Given the argument, the most effective inhibitors against PTP1B are predicted by the order: D1-PTP1B (DS_average −9.5 kcal.mol⁻¹) ≈ 3-PTP1B (DS_average −9.3 kcal.mol⁻¹) ≈ 4-PTP1B (DS_average −9.4 kcal.mol⁻¹). However, 1-PTP1B (DS_average −7.7 kcal.mol⁻¹) is predicted in silico with lowest inhibitability against PTP1B yet observed in vitro with the predominant efficacy. This can be explained by the incompatibility between the two models; in fact, while the computation evaluated the inhibitability based on the mixed-competitive mode, the experiment proved that 1 was under competitive process. In terms of the ligand-3W37 inhibitory complexes, the corresponding prediction can be arranged by the order: 3-3W37 (DS_average −9.2 kcal.mol⁻¹) > D2-3W37 (DS_average −9.2 kcal.mol⁻¹) > 4-3W37 (DS_average −9.2 kcal.mol⁻¹). If also including 1-3W37 (DS_average −9.2 kcal.mol⁻¹), this perfectly resembles the results observed from the in vitro assay. In terms of 4W93, the retrievals from docking simulation are purely prediction, ie: 4-4W93 (DS_average −11.0 kcal.mol⁻¹) > 3-4W93 (DS_average −10.5 kcal.mol⁻¹) = D2-4W93 (DS_average −10.5 kcal.mol⁻¹). Generally, the results indicate that all the candidates are seemingly competent bio-inhibitors against diabetes-related protein structures; especially, the optical isomers (ie 1 and 2) need additional consideration from other theoretical models since the docking algorithm does not cover chiral interactions by its nature.

Table 2.

Ground state electronic energy and dipole moment values of 1–4 calculated by DFT at level of theory M052X/6-311++g(d,p).

Compound	Ground state electronic energy (a.u)	Dipole moment (Debye)
1	−1049.96534	3.056
2	−1049.96639	4.687
3	−911.20830	7.591
4	−963.73396	3.618

The interacting configurations are also visually rendered, presented in Figure 4; these regard to the ligand-protein duo associated with highest DS value. The descriptive specification includes hydrogen-like bonding (dashed arrow), van de Waals interaction (blurry purple), and conformational fitness (dashed contour). The 3D configurations within the site indicate that the sites are relatively tight in relation to the size of the inhibitor. This means that significant structural modification on the current ligand structures are not recommended since the ligands would be oversized. Based on the 2D maps, it appears that the ligands exhibit favorable conformational fitness with the site features indicated by the continuousness of dashed contours. In-detail bonding parameters are summarized in Tables S1 (for ligand-PTP1B complexes), S2 (for ligand-3W37 complexes), and S3 (for ligand-4W93 complexes) in the supplemental information.

Figure 4.

Visual presentation and in-pose interaction map of ligand-PTP1B, ligand-3W37, and ligand-4W93 inhibitory structures.

Quantum-Chemical Properties

The results from quantum calculation provide the preliminary view on the bio-medium compatibility and intermolecular interactability of the candidates based on ab initio insights of their chemical properties. Unlike docking, this argument concerns only the candidates (1-4), without references to any specific protein structures.

The optimized geometries of the bioactive compounds are shown in Figure 5. In general, the input structures can be easily converged without encountering geometrical constraints or abnormal bonding angles and lengths. This is often of the characteristics of natural compounds; thereby in-turn also validating their spectroscopic characterisation and structural elucidation from the experimental characterization.

Figure 5.

Geometrically optimised structures of 1–4 by DFT at level of theory M052X/6–311++g(d,p); Units: bond length (Å), bond angle (°).

The main molecular properties, ie ground state energy and dipole moment, are summarized in Table 3. Firstly, the energy represents the overall chemical activeness of the host molecule; in the interpretation, the lower value equals to the less likeliness that the molecule might react with the body chemicals before reaching the targeted protein, thus more likely that it can retain the structural characteristics and biological properties. Given this argument, 1 and 2 (ca. −1050 a.u.) are considered as more favored candidates, cf 3 (−911.21 a.u.) and 4 (−963.73 a.u.). Secondly, the dipole moment can represent the compatibility with dipole-solvent environments (eg physio-chemical media) of the host molecule as the value is the separation of positive-negative charge in a system. Given this standpoint, 3 (7.591 Debye) is the most favorable candidate; otherwise, 1, 2, and 4 (from 3.0 to 4.7 a.u) do not see a significant difference. Therefore, all the compounds are considered likely competent and highly promising for applications in physio-chemical environments in general and as bio-inhibitors in particular.

Table 3.

Prescreening Results on Inhibitability of Ligands (1-4) and Controlled Drug (D) Towards the Sites of Proteins PTP1B, 3W37, and 4W93.

P	C	Site 1		Site 2		Site 3		Site 4		Average
P	C	E	N	E	N	E	N	E	N	E
PTP1B	1	−10.2	3	−8.0	1	−7.6	1	−7.5	1	−7.7
	2	−12.9	7	−11.0	4	−10.4	2	−10.9	3	−10.8
	3	−12.2	5	−10.3	2	−9.2	2	−8.1	1	−9.3
	4	−12.7	5	−10.1	2	−8.7	1	−9.4	2	−9.4
	D1	−11.2*	2	−9.5	1	−9.1	1	−8.2	0	−9.5
3W37	1	−11.5	4	−9.4	2	−9.7	2	−8.6	1	−9.2
	2	−9.3	2	−11.2	4	−8.7	1	−8.3	1	−8.8
	3	−12.8	5	−9.9	2	−9.7	2	−8.5	1	−9.4
	4	−11.9	3	−8.6	1	−8.4	1	−7.8	0	−8.3
	D2	−10.3	2	−12.1	4	−8.2	1	−7.8	1	−8.8
4W93	1	−11.2	4	−9.9	2	−8.1	1	−8.0	1	−8.7
	2	−10.2	2	−11.8	4	−8.1	1	−8.3	1	−8.9
	3	−12.0	6	−10.7	3	−9.9	2	−10.8	3	−10.5
	4	−12.2	8	−11.0	4	−10.9	3	−11.2	3	−11.0
	D2	−10.9	2	−12.7*	5	−10.3	2	−10.2	2	−10.5

P: Protein; C: Compound; E: DS value (kcal.mol⁻¹); N: Number of hydrophilic interactions.

Molecular electronic potential (MEP) maps of the structures are given in Figure 6, providing the information on the distribution of electrostatic potential. By convention, reddish regions have negative tendency; bluish regions have positive tendency; greenish regions hold neither of the tendencies. From the conceptual view, the distribution of the chemical activities over each molecular plane can be argued for its flexibility when in interaction with external structures, which is especially useful in the scope of this study as protein-based sites often have arbitrary and unpredictable configurations. It can be seen that all the molecules (especially 1, 2, and 3) possess rather interspersed distributions of electrostatic potential; in other words, the negative-positive tendencies are rather alternate over their molecular plane. These molecular features are thought to facilitate the physical interactions with external complex structures, such as protein-based folded conformations.

Figure 6.

Molecular electrostatic potential (MEP) formed by mapping of total density over the electrostatic potential of 1–4 by DFT at level of theory M052X/def2-TZVPP.

Physicochemical Properties

The physicochemical properties of the compounds are given in Table 4. QSARIS-derived physical properties include molecular mass (Da), polarizability (Å³), size (Å), and dispersion coefficients (logP and logS); the number of hydrogen bonds was counted from docking-based results. Overall, all the candidates well satisfy Lipinski's criteria for the argument on drug-likeness, ie: molecular mass < 300 amu; hydrogen-like donors < 5; hydrogen-like acceptors < 5; partition coefficient logP < 0. From the view of biological compatibility, 3 seems to be most promising given its lowest octane-water partition ratio (logP −2.57), followed by 4 (logP −1.98). From the view of biological interactability, the potential can be ranked by the order of polarizability, ie : 1 ≈ 2 (ca. 29 Å³) > 4 (23.6 Å³) > 3 (21.4 Å³), since the property indicates the sensitivity of a structure to external electric fields such as those created by other polarized agents (eg amino-acid-based protein structures); the unit conversion is given by Claussius-Mossotti relation: $10^{6} / 4 π ϵ_{0} [A^{2} . s^{4} . k g^{- 1}] \equiv 1 [c m^{3}]$ .⁶⁰ Therefore, the compounds are preliminarily considered compatible with the biological environments and suitable for drug-developed applications.

Table 4.

Physicochemical Properties of Studied Compounds 1–4.

Compound	Mass (amu)	Polarizability (Å³)	Size (Å)	Dispersion coefficients		Hydrogen bond* (3W37/PTP1B/4W93)
Compound	Mass (amu)	Polarizability (Å³)	Size (Å)	LogP	LogS	H-acceptor	H-donor
1	295.3	28.9	336.7	−0.93	−0.64	0/0/0	4/3/3
2	295.3	28.7	337.9	−0.93	−0.64	2/4/2	2/3/2
3	244.1	21.4	245.3	−2.57	−0.59	1/3/3	4/2/3
4	267.3	23.6	220.4	−1.98	−0.47	0/3/3	2/1/4

counted from docking results of each most effective ligand-PTP1B inhibitory complex.

Pharmacological Properties

The ADMET properties of the compounds are summarized in Table 5. The thresholds for theoretical prediction are based on Pires’ interpretations.⁶¹ Regarding absorption, all the candidates register moderate intestinal absorbability (ca. 40%) and low Caco2 permeability (logPapp < $10^{- 8}$ ); meanwhile, they are expected to not inhibit the activity of the P-glycoprotein family, thus having no effect on the extrusion of the toxins and xenobiotics out of cells. Regarding distribution, the compounds are more likely to be diffused into the plasma, rather than to be accumulated in tissues, given by its relatively low logVDss (< −0.15); in addition, they are not likely to cross the blood-brain barrier (logBB < -1) nor penetrate the central nervous system (logPS < -3). Regarding metabolism, there is no prediction on their interactions with the cytochromes P450 family (neither as the substrates nor the inhibitors), thus no effects to the liver. Regarding excretion, they are unlikely to be carried out of the body by organic cation transporter 2, thus conducive to their prolonged circulation in the body and retaining medicinal effects longer. In terms of toxicity, all the compounds are predicted to be considerably safe for medical use, ie: (i) no mutagenic potentials; (ii) no potential for fatal ventricular arrhythmia as hERG inhibitors; (ii) no hepatotoxicity (except for 3); (iv) no skin sensitization; (v) toxicity to bacterium T. Pyriformis (pIGC50 >> −0.5) yet safety to animal organisms, eg fish Flathead Minnows (LC50 >> −0.3). Therefore, all the candidates are highly recommended for pharmacological developments.

Table 5.

ADMET-based pharmacokinetics and pharmacology of the studied compounds 1–4.

Property	1	2	3	4	Unit
Absorption
Water solubility	−0.968	−0.968	−1.542	−1.57	⁽¹⁾
Caco2 permeability	0.037	0.037	−0.146	−0.184	⁽²⁾
Intestinal absorption	40.17	40.17	45.878	46.558	⁽³⁾
Skin Permeability	−2.864	−2.864	−2.795	−2.735	⁽⁴⁾
P-glycoprotein substrate	No	No	No	Yes	⁽⁵⁾
P-glycoprotein I inhibitor	No	No	No	No	⁽⁵⁾
P-glycoprotein II inhibitor	No	No	No	No	⁽⁵⁾
Distribution
VDss	−0.358	−0.358	−0.704	−0.183	⁽⁶⁾
Fraction unbound	0.591	0.591	0.681	0.83	⁽⁶⁾
BBB permeability	−0.801	−0.801	−0.899	−1.419	⁽⁷⁾
CNS permeability	−4.271	−4.271	−4.57	−4.124	⁽⁸⁾
Metabolism
CYP2D6 substrate	No	No	No	No	⁽⁵⁾
CYP3A4 substrate	No	No	No	No	⁽⁵⁾
CYP1A2 inhibitor	No	No	No	No	⁽⁵⁾
CYP2C19 inhibitor	No	No	No	No	⁽⁵⁾
CYP2C9 inhibitor	No	No	No	No	⁽⁵⁾
CYP2D6 inhibitor	No	No	No	No	⁽⁵⁾
CYP3A4 inhibitor	No	No	No	No	⁽⁵⁾
Excretion
Total Clearance	0.255	0.255	0.771	0.772	⁽⁹⁾
Renal OCT2 substrate	No	No	No	No	⁽⁵⁾
Toxicity
AMES toxicity	No	No	No	No	⁽⁵⁾
Max. tolerated dose	0.67	0.67	1.865	0.616	⁽¹⁰⁾
hERG I inhibitor	No	No	No	No	⁽⁵⁾
hERG II inhibitor	No	No	No	No	⁽⁵⁾
Oral Rat Acute Toxicity	2.249	2.249	2.055	1.779	⁽¹¹⁾
Oral Rat Chronic Toxicity	3.485	3.485	2.833	3.128	⁽¹²⁾
Hepatotoxicity	No	No	Yes	No	⁽⁵⁾
Skin Sensitisation	No	No	No	No	⁽⁵⁾
T. Pyriformis toxicity	0.285	0.285	0.285	0.285	⁽¹³⁾
Minnow toxicity	1.7	1.7	4.055	3.351	⁽¹⁴⁾

⁽¹⁾ log mol.L⁻¹; ⁽²⁾ log Papp (10^{<συπ>−6<συπ>} cm.s⁻¹); ⁽³⁾ %; ⁽⁴⁾ log Kp; ⁽⁵⁾ Yes/No; ⁽⁶⁾ log L.kg⁻¹; ⁽⁷⁾ log BB; ⁽⁸⁾ log PS.

⁽⁹⁾ log mL.min⁻¹.kg⁻¹; ⁽ ¹⁰ ⁾ log mg.kg⁻¹.day⁻¹; ⁽¹¹⁾ mol.kg⁻¹; ⁽¹²⁾ log mg.kg⁻¹_bw.day⁻¹; ⁽¹³⁾ log μg.L⁻¹; ⁽¹⁴⁾ log mM.

Discussion

Regarding the significant findings, this work preliminarily argue for the insights on the anti-diabetic potentiality of Aruncus dioicus. The isolated components, ie sambunigrin (1), prunasin (2), uridine (3), and adenosine (4), were not only successfully assessed for their enzymatic actions by the in vitro studies but also consistently validated for their activities by the in silico models, thus encouraged for more specialized research. Also, this systematic theory-experiment methodology can be considered as a new procedural approach for natural product reseach, especially for the screening of bioactive phytochemicals.

Regarding the experimental settings, the methanol extract from the A. dioicus leaf underwent partitioning to yield ethyl acetate and butanol fractions. Overall, the latter was preliminarily tested for overall inhibitory effect on PTP1B assay with a promising result (over 50% inhibition at concentration 30 μg.mL⁻¹), thus herein subjected for chromatographic isolation and spectroscopic characterization obtaining four alkaloidal compounds (1 − 4) in this report. Meanwhile, the former was observed with a marginal potential, thereby disregarded from the context of this report and redirected for other research interests.

Regarding the computational limitations, the computer-based platforms this work were utilized for their cost and time advantages given the purpose of wide-scale screenings. After specifying the most promising candidates, the more power-intensive technique, ie molecular dynamics simulation, is recommended for further consideration. This can monitor the real-time behavior of all the concerned ligand-protein atoms, thus deriving the inhibitory mechanism more particularly. However, it is noteworthy that the computer power required for this in-depth analysis is considerably intensive, thus not highly reasonable at screening and discovering stages.

Conclusions

This study first-time investigates the anti-diabetic potentials of glycosidic alkaloidal compounds isolated from Quan Lan Island A. dioicus aerial parts by a thoroughgoing investigation, from theoretical screening to experimental work. Firstly, experimental isolation and spectroscopic characterization identified for compounds, ie sambunigrin (1), prunasin (2), uridine (3), and adenosine (4). Secondly, all the elucidated structures were input to different computational platforms for theoretical screening, which resulted in promising bio-chemo-pharmacological potentiality, ie: docking simulation (docking scores < 10 kcal.mol⁻¹); quantum calculation (dipole moments < 3 Debye); QSARIS model (satisfying Lipinski's rule of five); SwissADME model (satisfying Pires’ interpretations). Finally, their biological activities were validated by in-practice in vitro tests against diabetes-related proteins, ie: protein tyrosine phosphatase 1B (1 with lowest IC₅₀ value 0.39 ± 0.26 μM, 2 with no activity) and α-glucosidase (1 with lowest IC₅₀₍₁₎ value 44.89 ± 0.93 μM, 2 with no activity). Besides, the inhibitory kinetics based on Lineweaver-Burk and Dixon plot experiments revealed the competitive inhibition mode of 1 against PTP1B (K_i value 0.49 μM); while, that of 3 was a mixed-competitive inhibition (K_i value 13.09 μM) and the corresponding mechanism for 4 was a non-competitive inhibition (K_i value 24.99 μM). Altogether, sambunigrin (1), uridine (3), and adenosine (4) are suggested for further attempts on the development of anti-diabetic applications.

Methodology

Experimental Material Preparation

The aerial part of A. dioicus was collected in June 2021 from Quan Lan island, Quang Ninh province, Vietnam. The sample was sent for botanical identification by Dr Nguyen Quoc Binh, and a voucher specimen (AD01QL) was sent for formal deposit at the Institute of Natural Products Chemistry (INPC), Vietnam Academy of Science and Technology (VAST).

Extraction, Isolation, and Purification

The aerial parts of A. dioicus were chopped, washed, and dried (temperature appx. 50 °C) to reach desirable moisture (ca. 12%). Afterwards, a dried sample was grounded into powder (particle size appx. 0.2 mm; quantity 2 kg), before subjected to two consecutive extractions using ethyl acetate (EtOAc) at the feed-solvent ratio 1/3 w/v (ultrasonic assistance; temperature 35 °C; duration 90 min). Subsequently, the solvent was evaporated; the remaining powder residue underwent further extraction with methanol at the feed-solvent ratio 1/5 w/v (ultrasonic assistance; temperature 40 °C; duration 120 min). The extract obtained was subsequently filtered and the solvent was removed through vacuum distillation, yielding a total methanol extract (145 g).

The total methanol extract was then transferred into a standard silica gel chromatography column (particle sizes 63-200 μm). Elution was carried out using a solvent system CH₂C₂/MeOH. Initially, elution was performed with a polarity ratio of 10/1 v/v. Afterwards, the value was gradually increased to 1/10 v/v. This multi-step process yielded 15 fractions, denoted as AD.1-AD.15.

The certain fractions were selected for purification. Firstly, fraction AD.12 was further subjected to chromatography on an RP-C18 column (column size 4.0 × 60 cm; particle size 150 μm). A stepwise gradient of MeOH-H₂O (from 1:4 to 2:1) was employed; this resulted in six subfractions denoted as AD.12-1 - AD.12-6. Subfraction AD.12-3 was continuously subjected to further purification using semi-preparative Agilent HPLC 1260 by an isocratic solvent system (25% MeOH in H₂O; duration 40 min); this yielded compounds 1 (5.2 mg, t_R= 25.5 min) and 2 (7.5 mg, t_R= 27.2 min). Similarly, fraction AD.13 was chromatographed on an RP-C18 column (column size 4.0 × 60 cm; particle size 150 μm) and subjected to a stepwise gradient of MeOH-H₂O (from 1:4 to 1:1); this resulted in eight subfractions, denoted as AD.13-1 - AD.13-8. Subfraction AD.13-4 was further applied for a reverse-phase column using the stepwise gradient solvent system of MeOH-H₂O (from 1:4 to 1:2); this conduction yielded compounds 3 (5.4 mg) and 4 (15.4 mg).

Spectroscopic Characterization

Sambunigrin (1): White powder; $[α]_{D}^{25}$ −34.2° (c 0.01, MeOH); ¹H-NMR (400 MHz, methanol-d₄) δ_H (ppm): 7.59 (2H, m, H-4/H-8), 7.45 (3H, m, H-5/H-6/H-7), 6.05 (1H, s, H-2), 4.68 (1H, d, J = 7.6 Hz, H-1′), 3.94 (1H, dd, J = 2.0, 12.0 Hz, H-6′a), 3.69 (1H, dd, J = 6.4, 12.0 Hz, H-6′b), 3.42 − 3.28 (4H, m, H-2′/H-3′/H-4′/H-5′); ¹³C-NMR (100 MHz, methanol-d₄) δ_C (ppm): 135.3 (C-3), 130.9 (C-6), 130.0 (C-4/C-8), 128.8 (C-5/C-7), 118.6 (C-1), 102.2 (C-1′), 78.7 (C-5′), 78.0 (C-3′), 74.9 (C-2′), 71.7 (C-4′), 68.6 (C-2), 62.9 (C-6′).

Prunasin (2): White powder; $[α]_{D}^{25}$ −15.6° (c 0.01, MeOH); ¹H-NMR (400 MHz, methanol-d₄) δ_H (ppm): 7.59 (2H, m, H-4/H-8), 7.45 (3H, m, H-5/H-6/H-7), 5.91 (1H, s, H-2), 4.24 (1H, d, J = 7.2 Hz, H-1′), 3.91 (1H, dd, J = 2.0, 11.6 Hz, H-6′a), 3.70 (1H, dd, J = 5.2, 11.6 Hz, H-6′b), 3.42 − 3.28 (4H, m, H-2′/H-3′H-4′/H-5′); ¹³C-NMR (100 MHz, methanol-d₄) δ_C (ppm): 135.0 (C-3), 131.2 (C-6), 130.3 (C-4, 8), 129.1 (C-5, 7), 119.6 (C-1), 102.1 (C-1′), 78.5 (C-5′), 78.0 (C-3′), 74.9 (C-2′), 71.6 (C-4′), 68.5 (C-2), 62.9 (C-6′).

Uridine (3): White powder; ¹H-NMR (600 MHz, methanol-d₄) δ_H (ppm): 8.01 (1H, d, J = 8.4 Hz, H-6), 5.92 (1H, d, J = 4.8 Hz, H-1′), 5.72 (1H, d, J = 8.4 Hz, H-5), 4.20 (1H, t, J = 4.8 Hz, H-2′), 4.17 (1H, t, J = 4.8 Hz, H-3′), 4.03 (1H, m, H-4′), 3.86 (1H, dd, J = 3.0, 12.0 Hz, H-5′a), 3.75 (1H, dd, J = 3.0, 12.0 Hz, H-5′b); ¹³C-NMR (150 MHz, methanol-d₄) δ_C (ppm): 166.2 (C-4), 152.5 (C-2), 142.7 (C-6), 102.7 (C-5), 90.7 (C-1′), 86.3 (C-4′), 75.7 (C-2′), 71.3 (C-3′), 62.3 (C-5′).

Adenosine (4): White powder; ¹H-NMR (600 MHz, methanol-d₄) δ_H (ppm): 8.32 (1H, s, H-3), 8.20 (1H, s, H-8), 5.98 (1H, d, J = 6.6 Hz, H-1′), 4.76 (1H, dd, J = 4.8, 6.0 Hz, H-2′), 4.55 (2H, br s, 1-NH₂), 4.34 (1H, dd, J = 3.0, 4.8 Hz, H-3′), 4.18 (1H, q, J = 3.0 Hz, H-4′), 3.90 (1H, dd, J = 2.4, 12.6 Hz, H-5′a), 3.77 (1H, dd, J = 3.0, 12.6 Hz, H-5′b); ¹³C-NMR (150 MHz, methanol-d₄) δ_C (ppm): 157.6 (C-1), 153.5 (C-3), 150.1 (C-5), 142.0 (C-8), 121.2 (C-6), 91.3 (C-1′), 88.2 (C-4′), 75.5 (C-2′), 72.7 (C-3′), 63.5 (C-5′).

The chemical formulae of the characterized compounds (1-4) are presented in Figure 7. This information was subjected to computational screening for the prediction of their biological potentiality and physicochemical suitability before to experimental tests for in-practice activities.

Figure 7.

Chemical structure of isolated compounds (1 − 4) from A. dioicus.

PTP1B in Vitro Assay

The inhibitory activity of isolated compounds (1-4) against the enzyme PTP1B was determined based on an in vitro enzymatic assay previously described.⁶² The methodological principle is due to the hydrolysis of p-nitrophenyl phosphate (p-NPP) by the enzyme, resulting in the production of p-nitrophenyl; as the result, p-nitrophenyl produced after the phosphatase reaction has a yellow solution, whose density is directly proportional to the activity of the enzyme PTP1B and measurable by the intensity of optical absorbance at 405 nm. Each well (final volume 110 μL) was loaded with p-NPP (2 mM) and PTP1B enzyme (0.05-0.1 µg) in a buffer solution containing citrate (50 mM; pH 6.0), NaCl (0.1 M), ethylenediaminetetraacetic acid (EDTA - 1 mM), and dithiothreitol (DTT - 1 mM); subsequently, each test compound was added (or not, in case of blank sample). The mixtures were then incubated (temperature 37 °C; duration 10 min), followed by the addition of p-NPP (50 μL); after a further incubation (temperature 20 °C; 20 min), the reaction was stopped at the addition of NaOH (10 M). The quantity of p-nitrophenyl produced during the phosphatase reaction was determined by measuring the absorbance (wavelength 405 nm) using Infinite F50 spectrophotometer (Tecan, Switzerland). The hydrolysis of p-NPP without the presence of PTP1B was corrected by measuring the decrease in absorbance intensity at 405 nm. All experiments were in triplicate.

The reaction kinetics (inhibition mode) was studied using the Lineweaver and Burk (double-reciprocal) plot method (GraphPad PRISM version 5, USA). The experiments were conducted using the similar procedure used in the PTP1B assay at various concentrations of compounds 1, 3, and 4 and p-NPP (0.1, 1, 2 mM). After incubation (temperature 37 °C; 30 min), each sample was subjected to the measurement of optical absorbance (wavelength 405 nm). Using a standard curve of absorbance-molar (of hydrolyzed p-NPP), the reaction velocities were calculated; in essence, the Lineweaver-Burk plot represents a reciprocal of the enzymatic velocity (V) against the reciprocal of the substrate concentration (1/V vs 1/[p-NPP]). The Michaelis constant (K_m) and maximum velocity (V_max) of PTP1B can be determined using the Lineweaver-Burk plot. The inhibition constant (K_i) of compounds 1, 3, and 4 were calculated using the following expression:

\frac{1}{{V^{'}}_{m a x}} = (1 + \frac{[I]}{K_{i}}) V_{m a x}

where,

V_{m a x}^{'}

is the value of V_max in the presence of a concentration [I] of inhibitor.

α-Glucosidase in Vitro Assay

The assay to assess the inhibitory activity of the α-glucosidase enzyme was conducted utilizing a 96-well microplate. The reaction components include p-nitrophenyl α-D-glucopyranoside (2.5 mM), phosphate buffer (100 mM; pH 6.8), α-glucosidase (0.2 U/mL), and the test samples (10 µL) at different concentrations; the total volume of each reaction mixture was 200 µL. In negative control lines, the test sample was replaced with a phosphate buffer (100 mM); while, acarbose was used as a positive control. The mixtures were incubated (temperature 37 °C; duration 30 min); afterwards, the reaction was stopped at the addition of Na₂CO₃ solutions (0.1 M). The absorbance of each product mixture was measured (wavelength 405 nm) using Infinite F50 spectrophotometer (Tecan, Switzerland). All tests are in triplicate. The percentage inhibition of α-glucosidase activity (%I) was calculated using the following expression:

% I = \frac{A_{C} - A_{T}}{A_{C} - A_{0}} \times 100

where,

$% I$ : Percentage of α-glucosidase enzymatic activity inhibited.

$A_{C}$ : Absorbance of the control sample (without test sample).

$A_{T}$ : Absorbance of the test sample .

$A_{0}$ : Absorbance of the blank sample (containing only buffer, no α-glucosidase enzyme).

The IC₅₀ values were calculated based on the log(concentration of test sample) and % inhibition curve, which represent the concentration of the test samples that inhibits 50% of α-glucosidase enzyme activity.

Statistical Analysis

The data were expressed as the mean (± standard deviation) from a minimum of three independent experiments (each performed in triplicate assays), and determined through regression analysis. SigmaPlot (version 11.0) was used for the statistical analysis.

Computational Input Preparation

The data is from experimental findings in this study and the existing literature serves as the input for the computational screening. Figure 7 presents the chemical formulae of potential compounds (1-4) elucidated from spectroscopic characterization, which were input as the ligands for docking simulation and as chemical structures for other molecular analyses. Figure 8 presents the biological assemblies of tyrosine phosphatase 1B (PTP1B; ID: UniProtKB-A0A0U1XP67), α-glucosidase (3W37; DOI: 10.2210/pdb3W37/pdb), and α-amylase (4W93; DOI: https://doi.org/10.2210/pdb4W93/pdb), which were input as the targeted proteins for docking simulation. The structures of ursolic acid (D1) for the insulin-related protein and acarbose (D2) for the glucose-related proteins were used as the references.

Figure 8.

Biological assemblies of PTP1B, 3W37, 4W93, and structural formula of controlled drugs (ursolic acid - D1 and acarbose - D2).

Docking Simulation

A typical procedure of molecular docking simulation (by MOE 2012.10⁶³) follows three steps, ie: (i) Input preparation (configuration: protein active range 4.5 Å, ligand charge-assigning using Gasteiger-Huckel method); (ii) Docking simulation (configuration: retaining poses 10; solutions per iteration 1000; solutions per fragmentation 200); (iii) Re-docking iteration (threshold: root-mean-square deviation (RMSD) values < 2 Å). Given theoretical interpretation, the inhibitory effectiveness of a ligand towards the targeted protein structure can be primarily evaluated by docking score (DS) energy of the associated inhibitory system, which represents pseudo-Gibbs free energy (formed by hydrophilic binding and hydrophobic interaction); also, RMSD values and number of hydrogen-like interactions can be considered for arguments on bio-conformational rigidity and binding strength, respectively. In addition, MOE can provide visual rendering for ligand-protein interaction maps (2D) and in-pose arrangements (3D).

Quantum Calculation

Molecular chemical properties of the investigated structures were given by density functional theory (DFT) calculation using Gaussian 09 without symmetry constraints.⁶⁴ Level of theory M052X/6–311++G(d,p) and basis set def2-TZVPP⁶⁵ were selected. The converged geometries were checked for the structural global minimum on the potential energy surface (PES) by vibrational frequencies. The frozen-core approximation for non-valence-shell electrons was applied. The resolution-of-identity (RI) approximation was set. The frontier orbital analysis was carried out by NBO 5.1 at the level of theory M052X/def2-TZVPP.⁶⁶

Physicochemical Analysis

A combination model was used to predict the drug-like properties of the phytochemicals. The parameters were the physical properties retrieved from QSARIS⁶⁷ (based on Gasteiger–Marsili method⁶⁸). The references were from Lipinski's rule of five,⁶⁹ which provides the theoretical criteria for a well membrane-permeable candidate, ie: molecular mass < 500 Da; hydrogen-bond donors ≤ 5; hydrogen-bond acceptors ≤ 10; logP < + 5.^70,71

ADMET Analysis

A combination model was used to predict the pharmacological potentiality of the compounds. The parameters were ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties retrieved from SwissADME (Swiss Institute of Bioinformatics; http://www.swissadme.ch/; 30^th October 2023). The references were from Pires’ theoretical interpretations.⁶¹

Supplemental Material

sj-docx-1-npx-10.1177_1934578X241271648 - Supplemental material for Inhibitory activities of Aruncus dioicus alkaloidal glycosides against protein tyrosine phosphatase 1B and α-glucosidase: A methodical theory-experiment investigation

Supplemental material, sj-docx-1-npx-10.1177_1934578X241271648 for Inhibitory activities of Aruncus dioicus alkaloidal glycosides against protein tyrosine phosphatase 1B and α-glucosidase: A methodical theory-experiment investigation by Phi-Hung Nguyen, Thanh Q. Bui, Thi-Tuyen Tran, Thi-Thuc Bui, Thi-Thuy Do, Dao-Cuong To, Manh Hung Tran, Phan Tu Quy, Nguyen Quang Co, Nguyen Vinh Phu and Nguyen Thi Ai Nhung in Natural Product Communications

Footnotes

Acknowledgements

Nguyen Thi Ai Nhung acknowledges the partial support of Hue University under the Core Research Program [Grant Number: NCTB.DHH.2024.04]

Authorship Contributions

All authors contributed to the study conception and design. All authors read and approved the final manuscript. Specific roles are summarized in the follow-up table.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical Approval

Ethical Approval is not applicable for this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Vietnam Academy of Science and Technology with project [Grant number: VAST04.05/22-23].

ORCID iDs

Dao-Cuong To

Nguyen Thi Ai Nhung

Supporting Information

In-detail spectra for experimental characterization provided in supporting information.

Statement of Human and Animal Rights

This article does not contain any studies with human or animal subjects.

Statement of Informed Consent

There are no human subjects in this article and informed consent is not applicable.

Supplemental Material

Supplemental material for this article is available online.

References

Abraira

Colwell

Nuttall

, et al. Veterans affairs cooperative study on glycemic control and complications in type II diabetes (VA CSDM): results of the feasibility trial. Diabetes Care. 1995;18(8):1113-1123.

Ohkubo

Kishikawa

Araki

, et al. Intensive insulin therapy prevents the progression of diabetic microvascular complications in Japanese patients with non-insulin-dependent diabetes mellitus: a randomized prospective 6-year study. Diabetes Res Clin Pract. 1995;28(2):103-117.

Klein

BEK

Moss

Cruickshanks

. Relationship of hyperglycemia to the long-term incidence and progression of diabetic retinopathy. Arch Intern Med. 1994;154(19):2169-2178.

Ding

Tanaka

Zhang

. Risk factors contributing to type 2 diabetes and recent advances in the treatment and prevention. Int J Med Sci. 2014;11(11):1185.

Sandholm

Forsblom

. Microvasc Dis Diabetes. John Wiley & Sons, Inc.; 2020.

Henning

. Type-2 diabetes mellitus and cardiovascular disease. Future Cardiol. 2018;14(6):491-509.

Draznin

Aroda

Bakris

, et al. 2. Classification and diagnosis of diabetes: standards of medical care in diabetes-2022. Diabetes Care. 2022;45(Supplement_1):S17-S38.

Vieira

MNN

Lyra e Silva

Ferreira

De Felice

. Protein tyrosine phosphatase 1B (PTP1B): a potential target for Alzheimer’s therapy? Front Aging Neurosci. 2017;9(7):Article ID 28197094.

Holman

Cull

Turner

. A randomized double-blind trial of acarbose in type 2 diabetes shows improved glycemic control over 3 years (UK prospective diabetes study 44). Diabetes Care. 1999;22(6):960-964.

10.

Lebovitz

. Alpha-glucosidase inhibitors. Endocrinol Metab Clin North Am. 1997;26(3):539-551.

11.

Özil

Emirik

Etlik

Ülker

Kahveci

. A simple and efficient synthesis of novel inhibitors of alpha-glucosidase based on benzimidazole skeleton and molecular docking studies. Bioorg Chem. 2016;68:226-235.

12.

Nikookar

Mohammadi-Khanaposhtani

Imanparast

, et al. Design, synthesis and in vitro α-glucosidase inhibition of novel dihydropyrano [3, 2-c] quinoline derivatives as potential anti-diabetic agents. Bioorg Chem. 2018;77:280-286.

13.

Cho

. Protein tyrosine phosphatase 1B (PTP1B) and obesity. Vitam Horm. 2013;91:405-424.

14.

Tutin

. Flora Europaea: Plantaginaceae to Compositae (and Rubiaceae). Cambridge university press; 1964;4.

15.

Potter

Eriksson

Evans

, et al. Phylogeny and classification of Rosaceae. Plant Syst Evol. 2007;266:5-43.

16.

Ohwi

. Flora of Japan. Smithson Inst; 1965;1067.

17.

Cherepanov

. Vascular Plants of Russia and Adjacent States (the Former USSR). Cambridge University Press; 1995.

18.

Lee

. Flora of Korea. National Institute of Biological Resources; 1996.

19.

Šet̕ekauri

Shetekauri

Jacoby

. Gebirgsflora & Bäume Des Kaukasus. Verlag nicht ermittelbar; 2009.

20.

Dorjsembe

Joo

Nho

Ham

Kim

. Aruncus dioicus var. kamtschaticus extract ameliorates psoriasis-like skin inflammation via akt/mTOR and JAK2/STAT3 signaling pathways in a murine model. Nutrients. 2022;14(23):5094.

21.

Suh

Min

Park

. The complete chloroplast genome of Aruncus dioicus var. kamtschaticus (Rosaceae). Mitochondrial DNA Part B. 2021;6(3):1256-1258.

22.

Ahn

Kim

Auh

Choe

. In vitro α-glucosidase and pancreatic lipase inhibitory activities and antioxidants of Samnamul (Aruncus dioicus) during rehydration and cooking. Food Sci Biotechnol. 2014;23:1287-1293.

23.

Kim

Yun

Jeong

Baek

. Hypoglycemic and hypolipidemic effects of samnamul (shoot of Aruncus dioicus var. kamtschaticus Hara) in mice fed a high-fat/high-sucrose diet. Food Sci Biotechnol. 2018;27:1467-1473.

24.

Rural Development Administration. Wild Plants in Korea. Iljinsa Publishing Co.; 1991.

25.

Chung

Choi

Moon

. Maintaining quality of wild vegetables (Aster glehni and Aruncus dioicus var. kamtschaticus) from Ulleungdo (Island), Korea by modified atmosphere packaging. Food Sci Biotechnol. 2010;19:193-199.

26.

Kim

, et al. Antioxidative and antimicrobial activities of Aruncus dioicus var. kamtschaticus Hara extracts. J Korean Soc Food Sci Nutr. 2011;40(1):47–55.

27.

Youn

Shin

, et al. Antioxidant and anti-wrinkling effects of Aruncus dioicus var. kamtschaticus extract. Korean J Food Preserv. 2012;19(3):393-399.

28.

Zhang

Kim

. DNA Damage protection and anti-inflammatory activity of different solvent fractions from Aruncus dioicus var. kamtschaticus. Korean J Plant Resour. 2014;27(6):714-719.

29.

Jeong

Kim

Min

Woo

. Monoterpenoids from the aerial parts of Aruncus dioicus var. kamtschaticus and their antioxidant and cytotoxic activities. Bioorg Med Chem Lett. 2011;21(11):3252-3256.

30.

Bohm

Bhat

. Flavonoids of Astilbe and Rodgersia compared to Aruncus. Biochem Syst Ecol. 1985;13(4):437-440.

31.

Schwaiger

Zeilner

Ellmerer

Antal

Rollinger

Stuppner

. Bioactivity-guided isolation of acetylcholinesterase inhibiting constituents of the flowers of Bride’s Feathers (Aruncus dioicus). Planta Med. 2009;75:Article ID PA18.

32.

Zhao

Jeong

Min

Kim

Woo

. Cytotoxic and anti-oxidant constituents from the aerial parts of Aruncus dioicus var. kamtschaticus. Nat Prod Sci. 2013;19(1):66-70.

33.

Fusani

Piwowarski

Zidorn

Kiss

Scartezzini

Granica

. Seasonal variation in secondary metabolites of edible shoots of Buck’s beard [Aruncus dioicus (walter) fernald (Rosaceae)]. Food Chem. 2016;202:23-30.

34.

Kulesh

Krasovskaya

Maksimov

. Phenolic compounds of Aruncus dioicus and Adenocaulon adhaerescens. Chem Nat Compd. 1986;22(4):475-476.

35.

Que

Yan

Zhou

Gao

. Two new phenylpropanoid glycosides from the roots of Aruncus sylvester. J Asian Nat Prod Res. 2014;16(2):158-162.

36.

Nguyen

Zhao

, et al. Bioactive constituents from the n-butanolic fraction of Aruncus dioicus var. kamtschaticus. Nat Prod Sci. 2014;20(4):274-280.

37.

Oszmianski

Wojdylo

Lamer-Zarawska

Swiader

. Antioxidant tannins from Rosaceae plant roots. Food Chem. 2007;100(2):579-583.

38.

Fikenscher

Hegnauer

Ruijgrok

HWL

. Die Verbreitung der Blausäure bei den Cormophyten. Planta Med. 1981;41(04):313-327.

39.

Han

Lee

. Attenuation of brain injury by water extract of goat’s-beard (Aruncus dioicus) and its ethyl acetate fraction in a rat model of ischemia-reperfusion. J Food Sci Nutr. 2011;16(3):217-223.

40.

Granica

Fusani

Stanisławska

, et al. Monoterpenoids from the traditional North Italian vegetable Aruncus dioicus (Walter) Fernald var. vulgaris (maxim.) H. Hara (Rosaceae). Food Chem. 2017;221:1851-1859.

41.

Shin

Lee

Woo

Kim

. Effect of ethanol extracts of goat’s beard on streptozotocin induced diabetic symptoms and oxidative stress in rats. J East Asian Soc Diet Life. 2008;18(6):939-948.

42.

Lee

. Total polyphenol contents and antioxidant activities of methanol extracts from vegetables produced in Ullung island. Korean J Food Sci Technol. 2005;37(2):233-240.

43.

Kim

Effect of Duduk (Codonopsis Lanceolatae Radix) on the Quality of Beef Hamburger Patties. Korean J Community Living Sci. 2018;29(4):507-519.

44.

Park

Kang

Kim

, et al. Effect of Aruncus dioicus var. kamtschaticus extract on neurodegeneration improvement: ameliorating role in cognitive disorder caused by high-fat diet induced obesity. Nutrients. 2019;11(6):1319.

45.

Smith

. A case of poisoning by Tragopogon pratense, or goat’s beard. Lancet. 1911;177(4569):794-795.

46.

Thao

TTP

Bui

Quy

, et al. Isolation, semi-synthesis, docking-based prediction, and bioassay-based activity of Dolichandrone spathacea iridoids: new catalpol derivatives as glucosidase inhibitors. RSC Adv. 2021;11(20):11959-11975.

47.

Thao

TTP

Bui

Hai

NTT

, et al. Newly synthesised oxime and lactone derivatives from Dipterocarpus alatus dipterocarpol as anti-diabetic inhibitors: experimental bioassay-based evidence and theoretical computation-based prediction. RSC Adv. 2021;11(57):35765-35782.

48.

Reitnauer

. Mandelonitrile-glycosides in cancer research and cancer therapy. The amygdalin problem. Arzneimittelforschung. 1972;22(8):1347-1361.

49.

Towers

GHN

McInnes

Neish

. The absolute configurations of the phenolic cyanogenetic glucosides taxiphyllin and dhurrin. Tetrahedron. 1964;20(1):71-77.

50.

Seigler

Pauli

Nahrstedt

Leen

. Cyanogenic allosides and glucosides from Passiflora edulis and Carica papaya. Phytochemistry. 2002;60(8):873-882.

51.

Jiang

. Chemical constituents from fruits of Xanthium sibiricum. Chinese Tradit Herb Drugs. 2017;24:47-51.

52.

Kun

Jae

Sam

. Isolation of adenosine from the rhizomes of Polygonatum sibidcum. Arch Pharm Res. 1991;14(2):193-194.

53.

Fuentes

Castro

Astudillo

, et al. Bioassay-guided isolation and HPLC determination of bioactive compound that relate to the antiplatelet activity (adhesion, secretion, and aggregation) from Solanum lycopersicum. Evidence-Based Complement Altern Med. 2012;2012(1):Article ID 147031.

54.

Hoang

Tran

Pham

Huynh

Nguyen

. PTP1B Inhibitory flavonoids from orthosiphon stamineus benth. And their growth inhibition on human breast cancer cells. Nat Prod Commun. 2020;15(1):Article ID 1934578X19899517.

55.

Cornish-Bowden

Eisenthal

. Statistical considerations in the estimation of enzyme kinetic parameters by the direct linear plot and other methods. Biochem J. 1974;139(3):721-730.

56.

Dixon

. The determination of enzyme inhibitor constants. Biochem J. 1953;55(1):170.

57.

Nguyen

Han

, et al. Selaginellin and biflavonoids as protein tyrosine phosphatase 1B inhibitors from Selaginella tamariscina and their glucose uptake stimulatory effects. Bioorg Med Chem. 2015;23(13):3730-3737.

58.

Jeong

Nguyen

Zhao

, et al. Chemical constituents of Euonymus alatus (Thunb.) Sieb. and their PTP1B and α-glucosidase inhibitory activities. Phyther Res. 2015;29(10):1540-1548.

59.

Bui

Nhung

NTA

, et al. On the inhibitability of natural products isolated from Tetradium ruticarpum towards tyrosine phosphatase 1B (PTP1B) and α-glucosidase (3W37): an in vitro and in silico study. Molecules. 2021;26(12):3691.

60.

Feynman

. In: Gottlieb

ed., The Feynman Lectures on Physics - Volume II. Millenium. Basic Books; 2010:11.3.

61.

Pires

DEV

Blundell

Ascher

. pkCSM: predicting small-molecule pharmacokinetic and toxicity properties using graph-based signatures. J Med Chem. 2015;58(9):4066-4072.

62.

Bui

Nhung

NTA

63.

Molecular Operating Environment (MOE), 2022.02 Chemical Computing Group ULC, 910-1010 Sherbrooke St. W., Montreal, QC H3A 2R7, Canada, 2024.

64.

Frisch

Trucks

Schlegel

, et al. Gaussian 09, Revision A.02.

65.

Zhao

Schultz

Truhlar

. Design of density functionals by combining the method of constraint satisfaction with parametrization for thermochemistry, thermochemical kinetics, and noncovalent interactions. J Chem Theory Comput. 2006;2(2):364-382.

66.

Reed

Weinstock

Weinhold

. Natural population analysis. J Chem Phys. 1985;83(2):735-746.

67.

Cameron

. Computer software reviews. J Am Chem Soc. 2001;123(35):8644-8645.

68.

Gasteiger

Marsili

. Iterative partial equalization of orbital electronegativity-a rapid access to atomic charges. Tetrahedron. 1980;36(22):3219-3228.

69.

Lipinski

Lombardo

Dominy

Feeney

. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev. 1997;23:3-25.

70.

Ahsan

Samy

Khalilullah

, et al. Molecular properties prediction and synthesis of novel 1,3,4-oxadiazole analogues as potent antimicrobial and antitubercular agents. Bioorganic Med Chem Lett. 2011;21(24):7246-7250.

71.

Mazumdera

Chakraborty

Sena

Vadrab

Dec

Ravi

. Synthesis and biological evaluation of some novel quinoxalinyl triazole derivatives. Der Pharma Chem. 2009;1(2):188-198.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

8.14 MB

Inhibitory activities of Aruncus dioicus alkaloidal glycosides against protein tyrosine phosphatase 1B and α -glucosidase: A methodical theory-experiment investigation

Abstract

Keywords

Introduction

Results

Structural Elucidation

Enzymatic Inhibition

Docking-Based Inhibitability

Quantum-Chemical Properties

Physicochemical Properties

Pharmacological Properties

Discussion

Conclusions

Methodology

Experimental Material Preparation

Extraction, Isolation, and Purification

Spectroscopic Characterization

PTP1B in Vitro Assay

α-Glucosidase in Vitro Assay

Statistical Analysis

Computational Input Preparation

Docking Simulation

Quantum Calculation

Physicochemical Analysis

ADMET Analysis

Supplemental Material

sj-docx-1-npx-10.1177_1934578X241271648 - Supplemental material for Inhibitory activities of Aruncus dioicus alkaloidal glycosides against protein tyrosine phosphatase 1B and α-glucosidase: A methodical theory-experiment investigation

Footnotes

Acknowledgements

Authorship Contributions

Declaration of Conflicting Interests

Ethical Approval

Funding

ORCID iDs

Supporting Information

Statement of Human and Animal Rights

Statement of Informed Consent

Supplemental Material

References

Supplementary Material