Sage Journals: Discover world-class research

Abstract

Objective

Magnolia coriacea is a rare species of Magnoliaceae and categorized as critically endangered, yet known for its valuable pharmaceutical properties. Phytochemical study of the plant was carried out, and the isolated compounds were subjected to different computational platforms to predict their chemical, inhibitory, physicochemical, and pharmacological properties.

Methods

M coriacea dried twig or leaf powder was partially extracted with n-hexane, ethyl acetate, and 90% methanol to produce the corresponding extracts, which were subjected to column chromatography. The isolated compounds were identified by nuclear magnetic resonance (NMR) spectroscopic and electrospray ionization mass spectrometry (ESI-MS) methods.

Results

Eight known aporphine alkaloids were isolated, (+)-glaucine N-oxide (1), N-methyl glaucine acetate (2), (+)-(S)-glaucine (3), magnoflorine (4), pontevedrine (5), O-methylatheroline (6), 7-hydroxydehydroglaucine (7), and mangochinine acetate (8). Quantum-based calculations found no abnormal structural constraints and suggested 2, 4, and 8 as the most promising inhibitors of protein structures in general with intensive nucleophilic reactive sites at their nitrogen atoms. Classical-based docking simulation agrees with this with respect to 3W37 (docking score [DS] < −12 kcal·mol⁻¹) and PTP1B (DS ca −13 kcal·mol⁻¹) structures. The biocompatibility and pharmacokinetics of the three candidates given by quantitative structure–activity relationship (QSAR) and absorption, distribution, metabolism, excretion, and toxicity (ADMET) regressions justify their potential for medicinal development. The results encourage further experimental studies of the promising M coriacea-extracted aporphine alkaloids for drug-developing purposes, especially mangochinine acetate (8).

Conclusions

Isolation and molecular docking simulation of 8 aporphine alkaloids were accomplished. QSAR and ADMET regressions suggested that three compounds (2, 4, and 8) were the most suitable for medicinal applications given both biocompatibility and pharmacokinetics based on specific prediction toward 3W37 (α-glucosidase) and PTP1B (tyrosine phosphatase 1B) structures.

Keywords

Magnoliaceae mangochinine acetate quantum chemical calculation molecular docking simulation

Introduction

Over the past decades, diabetes has become a health concern worldwide, particularly in middle-income countries. Notably, type 2 diabetes is characterized by insulin resistance, which accounts for 90% of all diabetic cases,^1,2 resulting in reduced glucose uptake and postprandial hyperglycemia.¹ Therefore, controlling the blood-based glucose level of diabetic patients is a key strategy for reducing more serious complications such as stroke and other cardiovascular diseases.^3,4

Two main strategies are commonly used to control glucose levels in blood, primarily based on the inhibition of glucose-related enzymes in the body. First, the exoenzyme α-glucosidase, extensively expressed in microbes, plants, and animal tissues, is responsible for the hydrolysis of -(1/4) and -(1/6) bonds in starch and disaccharide molecules to yield glucose.⁵ A glucose-based inhibition directly reduces the blood-glucose concentration released from digestive carbohydrates.^6,7 Given this significance, myriad research attempts have been invested to find potential medicines for the inhibition of enzyme activity. In particular, acarbose was the first approved substance (AGI) for an α-glucosidase inhibitory effect resulting in delayed absorption of glucose from carbohydrate-containing foods. However, this group of drugs is often known to relate to certain abdominal side efﬂects, eg diarrhea or ﬂatulence.^8,9 Therefore, highly effective and low-aftereffect candidates for α-glucosidase inhibition are still of interest, for which the mechanism is based upon binding to α-glucosidase thus preventing hydrolysis of the glycosidic bonds of oligo- and polysaccharides and producing the α-glucose moiety.¹⁰ Another approach is based on protein tyrosine phosphatase 1B (PTP1B) activity. Functionally, PTP1B negatively modulates insulin signaling by taking off the phosphate moiety from essential tyrosine residues on the activated insulin receptor, inducing a reduction of glucose uptake in insulin-responsive cells.¹¹ PTP1B-targeted inhibition could alter the phosphorylation that leads to the insulin-responsive cell for glucose uptake. In other words, PTP1B possesses a negative regulatory effect on the insulin signaling pathway. Given this well-established knowledge, α-glucosidase¹² and tyrosine phosphatase 1B¹³ can be considered antidiabetic targets of high potential.

Magnolia (Magnoliaceae) comprised approximately 210 species¹⁴ and is often found in regions such as East and Southeast Asia, Central America (especially Mexico), and the north of South America. Many plants from this genus have been used as medicinal folk for the treatment of headache, hypertension, fever, diarrhea, and rheumatism. Magnolia coriacea (Hung T Chang & B.L.Chen), syn. Michelia coriacea, is a rare species, categorized as critically endangered (CR), level B2ab (i,ii,iii,v),¹⁵ and endangered (EN), level B1ab (iii,v),¹⁶ in the IUCN Red List of Threatened Species. This plant is found in China and some northern high mountain provinces in Vietnam, eg Ha Giang (Quan Ba District), Cao Bang, Tuyen Quang, and Son La,¹⁷ as a timber tree, about 10–20 m high.¹⁷ Phytochemical study of M coriacea confirmed the presence of essential oil, with the main components being bicyclogermacrene (12.6%) and spathulenol (17.0%)¹⁸ and flavonol glycosides.¹⁹ As part of our ongoing research on this species, 8 known aporphine alkaloids were successfully isolated and structurally elucidated from the twigs and leaves, ie (+)-glaucine N-oxide (1), N-methyl glaucine acetate (2), (+)-(S)-glaucine (3), magnoflorine (4), pontevedrine (5), O-methylatheroline (6), 7-hydroxydehydroglaucine (7), and mangochinine acetate (8). Although aporphine alkaloids such as boldine, magnoflorine,²⁰ and liriodenine²¹ have been reported for their inhibitory potential on α-glucosidase, the M coriacea alkaloids were still unknown and deserved more in-depth consideration. Nevertheless, because of the species value and its conservation status, large-scale experimentally based investigations are neither cost-effective nor time-efficient at this stage.

Instead of relying solely on experimental trial-and-error efforts, in silico techniques have gained increasing interest from the scientific community, especially in medical science. For various purposes, a variety of computations can be utilized in order to reach the most reliable predictions. Density functional theory (DFT) calculation is a quantum-based technique for the analysis of molecular properties, thus providing first insights into the chemical tendencies of a chemical structure. Molecular docking simulation is a classical mechanics-based method to build up inhibitory behavior, particularly of ligand–protein structures, as in this work, thus in turn reflecting the overall inhibitory effectiveness. Quantitative structure–activity relationship (QSAR) modeling can give a collective correlation between theory-based structural properties and experiment-based biological activities of certain chemical families. Among QSAR-regressed retrieval, physicochemical properties can be subjected to evaluation of biocompatibility; on the other hand, the pharmacokinetics and pharmacology of a candidate can be preliminarily viewed from the output of absorption, distribution, metabolism, excretion, and toxicity (ADMET) analysis. Altogether, the most promising chemical structures (representative input for screened candidates) can be justified for experimental attempts to specify their desirable biological activities.

In this work, evidence for the chemical composition of M coriacea extracts was collected by experimental studies, while the inhibitory potential of individual components for their diabetes-related biological assemblies was retrieved from computational analyses.

Results and Discussion

Chemistry

From the leaf and twig extracts of M coriacea, 8 aporphine alkaloids were isolated and characterized by comparison of their nuclear magnetic resonance (NMR) spectroscopic and mass spectrometry (MS) data with those in the literature: (+)-glaucine N-oxide (1),²² N-methylglaucine acetate (2),²⁰ (+)-(S)-glaucine (3),^23,24 magnoflorine (4),²⁰ pontevedrine (5),²⁵ O-methylatheroline (6),²⁶ 7-hydroxydehydroglaucine (7),²⁷ and mangochinine acetate (8).^28,29 To the best of our knowledge, these alkaloids were isolated for the first time from M coriacea. Figure 2 summarizes the structure of the alkaloids isolated and characterized from M coriacea. The data were further subjected to computational analyses for their biopotential prediction.

Figure 2.

Structures of alkaloids isolated from Magnolia coriacea.

Computational

DFT-Based Chemical Properties

From the standpoint of quantum mechanics, DFT calculation can be utilized to retrieve the optimized geometry of the bioactive structures (1–8) and their molecular electronic configurations. In turn, the former can be used to check the bonding abnormality (if any); on the other side, the latter can provide insight into the chemical activities of the host molecules.

The data for the geometrical optimization are summarized in Table 1 and visualized in Figure 3. Overall, the self-consistent convergence reached implies stability, regarding all structures. This confirms their in-nature existence. More particularly, no abnormal constraints were detected regarding bond lengths and angles in comparison to the average characteristic values, eg 1.5 Å for C–C, 1.4 Å for C=C (aromatic), 1.4 Å for C-O, 1.2 Å for C=O, and 1.4 Å for C-N; ground state electronic energies are significantly negative, ie from −29 738.27 to −782 029.95 kcal·mol⁻¹. All the structures considered to be polarized molecules are given their dipole moment; especially, the figures regarding 2 (11.387 D) and 4 (11.166 D) are of most significance having been observed by our experience in this approach. In essence, this should likely be conducive to the formation of polarized forces in ligand–protein inhibition, such as van der Waals interactions or ionic bonds.³⁰

Figure 3.

Optimized structures of 1–8 calculated by density functional theory (DFT) at level of theory M052X/6-311++g(d,p).

Table 1.

Ground State Electronic Energy and Dipole Moment Values of 1–8.

No.	Compound	Symbol	Ground state electronic energy (kcal·mol⁻¹)	Dipole moment (Debye)
1	(+)-Glaucine N-oxide	1	−782 029.95	3.922
2	N-Methyl glaucine acetate	2	−759 770.35	11.387
3	(+)-(S)-glaucine	3	−734 856.86	2.504
4	Magnoflorine	4	−710 455.93	11.166
5	Pontevedrine	5	−827 032.54	5.544
6	O-Methylatheroline	6	−32 746.54	4.440
7	7-Hydroxydehydroglaucine	7	−33 882.19	2.556
8	Mangochinine acetate	8	−29 738.27	3.889

The optimized geometries of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) quantum parameters are summarized in Table 2, and the corresponding electronic distributions are visualized in Figure 4. Except for 8, the electronic localization is rather evenly distributed in the geometrical planes. According to the theory of charge transferring, HOMO represents the intermolecular electron donation tendency, and LUMO represents the electron-accepting ability; therefore, theoretically speaking, high flexibility for intermolecular interaction can be predicted. In the case of 8, although the HOMO and LUMO are partially located on two sides of the molecular plane, the molecule itself is symmetrical. This means the charge-transferring tendencies (ie donating and accepting) in different parts of the molecule can swap their role easily.³¹ Therefore, 8 can also be considered highly flexible for external attacks. Another notice is that the electronic density is almost undistributed at the nitrogen atoms. From the quantum-mechanics standpoint, this suggests that these atoms are less likely to form ionic bonding. In terms of quantitative justification, all structures are considered to possess electronic stability having low E_HOMO values, all under −6 eV (widely accepted as stable if under −5 eV), significantly those of 2 (−9.621 eV), 4 (−10.129 eV), and 8 (−10.391 eV); in addition, their insulation-semiconduction band-gap energy (3.2 eV < ΔE_GAP < 9 eV)³² might suggest intermolecular binding capability toward enzymes and proteins. The latter is reasoned by the well-observed and explained electric conductivity of these polypeptide structures, within the mechanism of electron tunneling (superexchange theory).^33,34

Figure 4.

HOMO and LUMO of compounds 1–8 calculated by DFT at level of theory M052X/def2-TZVPP.

Table 2.

Quantum Chemical Parameters of Compounds 1–8.

Compound	E_HOMO (eV)	E_LUMO (eV)	ΔE_GAP	I = −E_HOMO	A = −E_LUMO	χ	µ	η	S = 1/η
1	−6.826	−0.154	6.672	6.826	0.154	3.490	−3.490	3.336	0.300
2	−9.621	−3.104	6.517	9.621	3.104	6.363	−6.363	3.259	0.307
3	−6.949	−0.177	6.771	6.949	0.177	3.563	−3.563	3.386	0.295
4	−10.129	−3.085	7.044	10.129	3.085	6.607	−6.607	3.522	0.284
5	−7.007	−1.806	5.201	7.007	1.806	4.407	−4.407	2.601	0.385
6	−7.356	−1.881	5.476	7.356	1.881	4.619	−4.619	2.738	0.365
7	−6.562	−0.163	6.399	6.562	0.163	3.363	−3.363	3.200	0.313
8	−10.391	−3.104	7.287	10.391	3.104	6.748	−6.748	3.644	0.274

Energy gap (ΔE_GAP = E_LUMO−E_HOMO); ionization potential I = −E_HOMO; electron affinity A = −E_LUMO; electronegativity χ = (I + A)/2; chemical potential µ = −χ = −(∂E/∂N)_ν(r); hardness η = (I-A)/2; softness S = 1/η.

Molecular electrostatic potential (MEP) maps of the structures are given in Figure 5; by convention, reddish regions represent negativity (related to electrophilic reactivity) while bluish regions represent positivity (related to nucleophilic reactivity). Noticeably, 2, 4, and 8 have a strong tendency for electrophilicity over the molecular regions (by bluish shading). Those with the most reactivity are dense around nitrogen atoms, with the highest values being ca −16 × 10⁻³ au. This confirms the conceptualization of the formation of nitrogen-based ionic bonding; extensively, the nitrogen cations are expected to be intensive nucleophilic sites despite the notable electronegativity of individual nitrogen. In contrast, although still containing a nitrogen cation in its molecule, 1 is unlikely to centralize its reactivity at the atom according to the scope of quantum mechanics. Instead, the prediction indicates that 1, 3, and 5–7 are more likely to perform balanced tendencies over the aromatic rings and moderate electrophilicity of oxygen-based species.

Figure 5.

Molecular electrostatic potential (MEP) formed by mapping of total density over the electrostatic potential of 1–8 calculated by density functional theory (DFT) at level of theory M052X/def2-TZVPP.

Overall, the results reveal the electronic behavior of the molecules, which can be used to deduce their behavior when interacting with other molecular structures.

Docking-Based Inhibition

From the standpoint of classical mechanics, molecular docking simulation can be utilized to examine the ligand–protein static behavior, thus predicting the inhibitory effectiveness of a compound toward its targeted protein structure. The docking score (DS) values and hydrophilic bonding numbers are the most important parameters; in principle, the former is the value of overall Gibbs free energy, and the latter is considered strong intermolecular bonding. Also, root-mean-square deviation (RMSD) (the average backbone-atom distances) and van der Waals interactions (hydrophobic forces) can be considered to certain extents.

In each ligand–protein duo, only the most stable intermolecular structure (in bold) is selected for more in-depth discussion since it most likely corresponds to the dominant product in reality. The susceptible sites of the biological assemblies (3W37 and PTP1B) to the ligands (1–8) are highlighted in Figure 6, and their corresponding screening data are summarized in Table 3. Regarding ligand-3W37, site 2 appears to be the most vulnerable to the ligands (1–8); in terms of ligand-PTP1B, there is variation over sites 1–3. Preliminarily, the ranks are rather consistent with the quantum-based analyses as 2, 4, and 8 are likely to show elevated effectiveness in comparison to the others.

Figure 6.

Quaternary structures of 3W37 and PTP1B with approachable sites by 1–8 and D: site 1 (cyan), site 2 (yellow), site 3 (gray), site 4 (blue), and site 5 (orange).

Table 3.

Prescreening Results on Inhibitory Effect of 1–8 and Controlled Drug Acarbose (D) Toward the Sites of 3W37 and PTP1B.

P	L	Site 1		Site 2		Site 3		Site 4		Site 5
P	L	E	N	E	N	E	N	E	N	E	N
3W37	1	−8.0	1	−11.6	2	−8.9	1	−7.1	0	−8.3	1
	2	−10.1	1	−11.3	2	−8.3	1	−7.9	1	−8.2	1
	3	−6.7	0	−10.6	2	−7.2	1	−7.7	1	−6.0	0
	4	−7.4	1	−12.3	3	−7.0	1	−10.5	2	−7.8	1
	5	−9.0	1	−11.8	3	−7.6	1	−6.9	0	−7.8	1
	6	−6.0	0	−10.7	2	−6.3	0	−8.2	1	−7.6	1
	7	−7.2	1	−11.2	2	−6.6	1	−7.4	1	−7.0	1
	8	−10.9	2	−12.9	4	−8.7	1	−9.9	2	−10.5	2
	D	−11.0	2	−13.0	5	−10.8	2	−11.9	3	−10.6	2
PTP1B	1	−11.0	3	−9.3	2	−7.0	1	−6.9	1	−7.2	1
	2	−10.7	2	−10.0	2	−12.9	5	−7.5	1	−9.6	2
	3	−9.8	1	−11.6	3	−8.8	1	−10.3	2		2
	4	−13.1	5	−10.9	2	−8.9	1	−9.2	2	−9.7	2
	5	−6.0	0	−10.8	2	−7.3	1	−6.3	0	−6.5	0
	6	−9.1	1	−11.3	3	−8.6	1	−8.8	1	−10.1	2
	7	−7.4	1	−12.2	3	−8.1	1	−7.0	0	−8.2	1
	8	−10.9	2	−11.4	3	−13.4	5	−10.4	2	−8.4	1
	D	−7.9	1	−12.8	4	−8.3	2	−7.0	1	−9.1	2

P, protein; L, ligand; E, DS value (kcal·mol⁻¹); N, number of hydrophilic interactions.

The selected data are given in Table 4 (ligand-3W37) and Table 5 (ligand-PTP1B). According to the theoretical interpretation as given, the most effective ligand-3W37 inhibitory structures are 8-3W37 (DS −12.9 kcal·mol⁻¹; RMSD 1.29 Å) > 4-3W37 (DS −12.3 kcal·mol⁻¹; RMSD 1.37 Å); the significance is defined as DS < −12 kcal·mol⁻¹, which is thought to correspond with the ability to induce effective inhibitory effects. In fact, our preceding works correlated this range with assay-based IC₅₀ values < 50 μM (control drug IC₅₀ ca 300 μM),^35,36 which were considered effective inhibition against the enzymatic activity of α-glucosidase. In terms of ligand-PTP1B inhibitory structures, the corresponding order is 8-PTP1B (DS −13.4 kcal·mol⁻¹; RMSD 1.75 Å) > 4-PTP1B (DS −13.1 kcal·mol⁻¹; RMSD 1.49 Å) > 2-PTP1B (DS −12.9 kcal·mol⁻¹; RMSD 1.17 Å); the significance is defined by comparison to the D-PTP1B complex (ie DS −12.8 kcal·mol⁻¹; RMSD 1.50 Å). This relies purely on theoretical reference since there has been no work on a direct correlation between docking-based inhibitability (PTP1B) and assay-based inhibition (tyrosine phosphatase 1B) in the existing literature. In summary, all the figures (those of 2, 4, and 8) are comparable to those of acarbose (ie D-3W37 and D-PTP1B), the commercial medicine often prescribed for diabetic patients. Therefore, 2, 4, and 8 are expected to also be effective inhibitors against the protein in further experimental tests, thus deserving efforts for sustainable conservation, in-mass isolation, and in vitro investigations (against either α-glucosidase or tyrosine phosphatase 1B).

Table 4.

Molecular Docking Simulation Results for Ligand-3W37 Inhibitory Complexes.

Ligand–protein			Hydrogen bond						van der Waals interaction
Name	DS	RMSD	L	P		T	D	E	van der Waals interaction
1-3W37	−11.6	1.76	6-ring	N	Arg 670	π-cation	4.12	−0.9	Glu 301, Ile 672, Arg 676, Leu 663, Asp 666, Gly 791, Thr 790, Glu 792, Arg 814, Phe 816
1-3W37	−11.6	1.76	6-ring	C	Arg 699	π-H	4.11	−0.7
2-3W37	−11.3	1.69	N	O	Asp 666	Ionic	3.55	−1.7	Leu 663, Glu 792, Arg 699, Thr 790, Glu 301, Ile 672, Arg 676
2-3W37	−11.3	1.69	6-ring	N	Arg 670	π-cation	3.98	−1.1
3-3W37	−10.6	1.24	N	N	Arg 676	H-acceptor	3.25	−4.8	Arg 699, Leu 663, Ile 672, Asp 666, Gly 791, Glu 301, Thr 299, Thr 790, Glu 792
3-3W37	−10.6	1.24	6-ring	N	Arg 670	π-cation	4.02	−0.7
4-3W37	−12.3	1.37	C	O	Glu 792	H-donor	3.03	−1.5	Val 760, Gly 698, Asn 758, Arg 670, Phe 680, Thr 662, Thr 681, Glu 301, Thr 790, Asp 666, Leu 663, Gly 791, Ile 759, Tyr 659, Ile 672
			O	N	Arg 676	H-acceptor	2.71	−0.6
			6-ring	N	Arg 699	π-cation	3.37	−0.6
5-3W37	−11.8	1.83	O	N	Arg 676	H-acceptor	2.84	−1.4	Phe 816, Thr 790, Gly 791, Phe 815, Ile 759, Glu 792, Leu 663, Gly 698, Tyr 659, Glu 301, Arg 670, Ile 672, Asp 666, Arg 814
			O	N	Arg 676	H-acceptor	2.93	−1.1
			O	N	Arg 699	H-acceptor	3.31	−1.7
6-3W37	−10.7	1.49	N	N	Arg 676	H-acceptor	3.26	−3.4	Thr 681, Glu 301, Ile 672, Asp 666, Gly 791, Leu 663, Glu 792, Arg 814, Arg 699
6-3W37	−10.7	1.49	6-ring	N	Arg 670	π-cation	3.53	−0.8
7-3W37	−11.2	1.09	O	O	Asp 666	H-donor	3.14	−1.2	Arg 814, Arg 699, Thr 790, Val 760, Leu 663, Ile 759, Gly 791, Ile 672, Arg 676, Thr 662, Tyr 659, Glu 792
7-3W37	−11.2	1.09	N	N	Arg 670	H-acceptor	3.09	−9.1
8-3W37	−12.9	1.29	C	O	Asp 666	H-donor	2.76	−1.3	Arg 670, Thr 662, Leu 663, Gly 791, Ile 759, Tyr 659, Val 760, Arg 676, Arg 699, Glu 301, Ile 672
			C	O	Asp 666	H-donor	3.09	−0.9
			N	O	Asp 666	Ionic	3.03	−4.3
			6-ring	C	Glu 792	π-H	4.24	−0.8
D-3W37	−13.0	1.17	O	O	Glu 792	H-donor	3.20	−0.7	Asp 666, Arg 670, Thr 299, Pro 683, Glu 301, Phe 680, Arg 814, Thr 681, Gly 698, Leu 663, Gly 700, Asn 758, Thr 790, Tyr 659, Val 760, Gly 791
			O	O	Ile 759	H-donor	2.77	−2.1
			O	N	Arg 699	H-acceptor	2.77	−4.4
			O	N	Arg 699	H-acceptor	3.23	−1.7
			O	N	Arg 676	H-acceptor	2.99	−0.6

DS, docking score energy (kcal·mol⁻¹); RMSD, root-mean-square deviation (Å); L, ligand; P, protein; T, type; D, distance (Å); E, energy (kcal·mol⁻¹).

Table 5.

Molecular Docking Simulation Results for Ligand-PTP1B Inhibitory Complexes.

Ligand–protein			Hydrogen bond						van der Waals interaction
Name	DS	RMSD	L	P		T	D	E	van der Waals interaction
1-PTP1B	−11.0	1.54	O	C	Ser 216	H-acceptor	3.56	−0.6	Lys 120, Tyr 46, Arg 221, Asp 181, Val 49, Ile 219, Gln 262, Asp 48
			O	N	Ala 217	H-acceptor	2.97	−0.9
			C	6-ring	Phe 182	H-π	4.11	−0.6
2-PTP1B	−12.9	1.17	C	O	Glu 252	H-donor	3.41	−1.2	Met 74, Glu 76, Lys 255, Ala 77, Lys 248, Arg 238, Val 249, Leu 234
			C	O	Glu 75	H-donor	3.07	−1.0
			C	O	Glu 75	H-donor	3.13	−1.4
			N	O	Glu 75	Ionic	3.60	−1.5
			N	O	Glu 252	Ionic	3.91	−0.7
3-PTP1B	−11.6	1.47	O	N	Lys 73	H-acceptor	3.01	−1.3	Arg 79, Met 74, Glu 75, Gln 78, Ser 205, Ser 203, Val 211, Gly 209, Leu 204, His 208, Pro 210
			6-ring	C	Ser 80	π-H	4.06	−0.6
			6-ring	C	Pro 206	π-H	4.42	−0.6
4-PTP1B	−13.1	1.49	C	O	Asp 48	H-donor	3.07	−1.2	Val 49, Tyr 46, Ser 216, Phe 182, Ala 217, Cys 215, Lys 120, Gly 220, Asp 181, Ile 219, Gln 262
			C	O	Asp 48	H-donor	3.03	−1.1
			O	N	Arg 221	H-acceptor	2.85	−1.9
			N	O	Asp 48	Ionic	3.48	−2.0
			N	O	Asp 48	Ionic	3.18	−3.4
5-PTP1B	−10.8	0.99	O	N	Lys 237	H-acceptor	3.37	−1.1	Val 211, Leu 204, Ala 77, Pro 210, Gln 78, Ser 80, His 208, Ser 203, Ser 205, Gly 209, Pro 206
5-PTP1B	−10.8	0.99	6-ring	C	Arg 79	π-H	3.64	−0.6
6-PTP1B	−11.3	1.20	6-ring	C	Arg 79	π-H	3.69	−0.6	Pro 210, Gly 209, Val 211, Ser 205, Pro 206, His 208, Gln 78, Leu 204, Lys 73
			6-ring	N	Ser 80	π-H	4.24	−1.2
			6-ring	N	Ser 80	π-H	4.10	−1.2
7-PTP1B	−12.2	1.44	O	O	Leu 204	H-donor	3.12	−1.9	Val 211, His 208, Gly 209, Ser 205, Pro 210, Pro 206, Ser 80, Lys 237, Ala 77, Lys 73
			O	N	Arg 79	H-acceptor	3.01	−0.7
			6-ring	C	Gln 78	π-H	3.73	−0.7
8-PTP1B	−13.4	1.75	O	O	Glu 252	H-donor	2.85	−2.0	Arg 238, Ser 243, Val 244, Val 249, Lys 255
			C	O	Glu 76	H-donor	3.32	−1.0
			O	N	Asp 245	H-acceptor	3.09	−0.6
			N	O	Glu 76	Ionic	3.91	−0.7
			6-ring	N	Lys 248	π-cation	3.55	−0.7
D-PTP1B	−12.8	1.50	O	O	Asp 236	H-donor	2.98	−3.8	Pro 206, Ala 77, Ser 80, Arg 79, Leu 233
			O	N	Lys 73	H-acceptor	3.12	−1.6
			O	N	Lys 237	H-acceptor	3.27	−2.2
			N	N	Gln 78	H-acceptor	3.44	−0.5

DS, docking score energy (kcal·mol⁻¹); RMSD, root-mean-square deviation (Å); L, ligand; P, protein; T, type; D, distance (Å); E, energy (kcal·mol⁻¹).

The corresponding visual projections (3D in-pose morphology and 2D interaction map) are presented in Figure 7 (ligand-3W37) and Figure 8 (ligand-PTP1B). From the scope of classical mechanics, 2, 4, and 8 create the most stable ligand–protein complexes in comparison to the others (1, 3, and 5–7), as expected by quantum-based analyses based on electronic configurations (natural bond orbital [NBO]) and chemical potential intensity (MEP). Regarding sites of reactivity, the nitrogen atoms of 2, 4, and 8 are also expected to exhibit exclusive nucleophilic reactivity via ionic bonding with the amino acid residues Asp 666 (3W37), Glu 75, Glu 252, Asp 48, and Glu 76 (PTP1B); meanwhile, the corresponding one (N⁺) of 1 does not contribute to the formation of either 1-3W37 or 1-PTP1B. On the other hand, 1, 3, and 5–7 tend to interact with the targeted proteins based on weaker hydrophilic forms (hydrogen bonds) and hydrophobic forces (van der Waals interactions). Finally, the sites are large and open, compared to the inhibitors. Also, the continuity of contours indicates that the inhibitors fit well into the surface inside the sites. This suggests room for further functionalization/modification on these ligands. For instance, one can consider the improvement of the nucleophilic reactivity of the nitrogen-based regions to a higher degree.

Figure 7.

Visual presentation and in-pose interaction map of compounds 1–8 and D and ligand-3W37.

Figure 8.

Visual presentation and in-pose interaction map of compounds 1–8 and D and ligand-PTP1B.

The quantum-classical consistency seen should be of great interest. The former is almost purely about the wavelike behavior of molecular electrons, while the latter is established based on the assumption that each atom in a molecule is a rigid particle and assigned with arbitrary coefficients representing its individual physical properties. However, more observations based on this approach are still needed in order to reach a solid conclusion.

QSAR-Based Physicochemical Properties

The physicochemical properties of input structures (1–8) are summarized in Table 6. In detail, hydrogen bond counts were referenced from the docking-based results; others were given by QSAR-based computation; druglike evaluation followed Lipinski's rule of 5. Overall, all inhibition-promising candidates (2, 4, and 8) satisfy the criteria: molecular mass < 500 amu (from 387.5 to 429.8 amu), logP < +5 (from 1.98 to 3.96), hydrogen bond–accepting counts < 5, and hydrogen bond–donating counts < 10. In summary, from Lipinski's perspective, 2 (N-methyl glaucine acetate), 4 (magnoflorine), and 8 (mangochinine acetate) might possess physicochemical properties conducive to their biocompatibility, in general, and to drug-developing applications, in particular.

Table 6.

Physicochemical Properties of Compounds 1–8 and D.

Compound	Mass (amu)	Polarizability (Å³)	Size (Å)	Dispersion coefficients		Hydrogen bond (3W37/PTP1B)
Compound	Mass (amu)	Polarizability (Å³)	Size (Å)	LogP	LogS	H-donor	H-acceptor	H-π
1	372.9	34.8	450.6	3.37	−3.97	0/0	0/2	1/1
2	429.8	49.5	489.3	3.96	−4.01	0/3	0/0	0/0
3	355.6	32.1	390.4	2.49	−3.15	0/0	1/1	0/2
4	343.1	31.6	430.1	1.98	−2.95	1/2	1/1	0/0
5	380.9	41.0	487.3	2.69	−4.03	0/0	3/1	0/1
6	351.7	39.4	448.5	2.87	−4.12	1/0	1/0	0/3
7	370.4	42.9	489.3	3.43	−4.27	1/1	1/1	0/1
8	387.5	43.6	490.1	3.95	−4.11	2/2	0/1	1/0
D	645.9	60.7	614.1	2.74	−0.93	2/1	3/3	0/0

ADMET-Based Pharmacokinetics and Pharmacology

The ADMET properties of the candidates (1–8) are predicted in Table 7. This model is based on statistical regression without external referencing; hence, the results obtained can be treated as a preliminary probability-based assessment. Since 2, 4, and 8 have demonstrated their inhibitory and biocompatible potential, they received special consideration for ADMET-based discussion rather than the others (1, 3, and 5–7). Firstly, all the M coriacea aporphine alkaloids isolated (1–8) appear to possess effective intestinal absorbance with a percentage of >90% regarding any compounds; particularly, those of 2, 4, and 8 are 97.3%, 96.2%, and 93.5%, respectively. Nevertheless, they register relatively low Caco-2 permeability values, ca 1.0–1.2 × 10⁻⁶ cm.s⁻¹. The model predicts that the compounds are likely to play the roles of P-glycoprotein substrates and inhibitors, thus affecting the extrusion of toxins and xenobiotics out of cells. In some aspects, these properties can be exploited for specific therapeutic advantages. Secondly, 2, 4, and 8 are more likely to accumulate in tissue rather than in plasma given their log VDss values > 0.45, yet unlikely to readily cross the blood–brain barrier (logBB < 0.2). Thirdly, 2 and 8 can be metabolized by cytochrome P450 CYP3A4 (oxidized in the liver) or excreted by organic cation transporter 2. Regarding potential toxicity, all candidates (2, 4, and 8) are considered safe: no mutagenic (AMES toxicity), fatal ventricular arrhythmia, hepatotoxic, and skin sensitization potentials. The values regarding Tetrahymena pyriformis (a protozoan) and minnow (a species of temperate freshwater fish) register on the thresholds recommended, thus requiring actual observations from experimental tests if the compounds are selected for further development. Overall, this preliminary assessment of M coriacea aporphine alkaloids suggests that 2 (N-methyl glaucine acetate), 4 (magnoflorine), and 8 (mangochinine acetate), especially the last, are likely to possess the appropriate pharmacokinetics and pharmacology for drug-developing efforts.

Table 7.

ADMET-Based Pharmacokinetics and Pharmacology of Compounds 1-8 and D.

Property	1	2	3	4	5	6	7	8	D	Unit
Absorption
Water solubility	−5.064	−3.445	−3.174	−3.37	−5.053	−4.283	−4.524	−2.985	−1.482	Numeric (log mol·L⁻¹)
Caco-2 permeability	1.034	0.783	1.153	1.291	1.399	1.219	1.078	1.258	−0.481	Numeric (log Papp in 10⁻⁶ cm·s⁻¹)
Intestinal absorption (human)	99.439	97.312	94.376	96.232	100	100	97.672	93.501	4.172	Numeric (% absorbed)
Skin permeability	−2.726	−2.565	−2.813	−2.716	−2.721	−2.74	−2.741	−2.945	−2.735	Numeric (log Kp)
P-glycoprotein substrate	No	No	Yes	Yes	No	No	No	Yes	Yes	Categorical (yes/no)
P-glycoprotein I inhibitor	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	No	Categorical (yes/no)
P-glycoprotein II inhibitor	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	No	Categorical (yes/no)
Distribution
VDss (human)	0.303	0.74	1.028	1.234	−0.275	−0.013	0.08	0.439	−0.836	Numeric (log L·kg⁻¹)
Fraction unbound (human)	0.024	0.285	0.319	0.243	0.076	0.162	0.102	0.309	0.505	Numeric (log L·kg⁻¹)
BBB permeability	−0.2	0.076	0.078	0.023	−0.877	−0.899	−0.485	0.01	−1.717	Numeric (log BB)
CNS permeability	−2.167	−1.652	−1.674	−2.139	−3.014	−3.262	−2.209	−2.228	−6.438	Numeric (log PS)
Metabolism
CYP2D6 substrate	No	No	No	No	No	No	No	No	No	Categorical (yes/no)
CYP3A4 substrate	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	No	Categorical (yes/no)
CYP1A2 inhibitor	Yes	Yes	Yes	Yes	Yes	Yes	Yes	No	No	Categorical (yes/no)
CYP2C19 inhibitor	Yes	No	No	No	Yes	Yes	Yes	Yes	No	Categorical (yes/no)
CYP2C9 inhibitor	Yes	No	No	No	Yes	No	Yes	No	No	Categorical (yes/no)
CYP2D6 inhibitor	No	Yes	Yes	No	No	No	No	No	No	Categorical (yes/no)
CYP3A4 inhibitor	No	No	No	No	Yes	Yes	Yes	Yes	No	Categorical (yes/no)
Excretion
Total clearance	0.523	1.133	0.996	1.185	0.414	0.601	0.542	1.098	0.428	Numeric (log mL·min⁻¹·kg⁻¹)
Renal OCT2 substrate	No	Yes	Yes	No	No	No	No	Yes	No	Categorical (yes/no)
Toxicity
AMES toxicity	Yes	No	No	No	Yes	Yes	Yes	No	No	Categorical (yes/no)
Max tolerated dose (human)	0.142	0.295	0.101	−0.538	0.241	0.415	0.358	−0.179	0.435	Numeric (log mg·kg⁻¹·day⁻¹)
hERG I inhibitor	No	No	No	No	No	No	No	No	No	Categorical (yes/no)
hERG II inhibitor	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Categorical (yes/no)
Oral rat acute toxicity (LD50)	2.893	3.402	3.749	2.696	2.899	3.148	3.069	2.815	2.449	Numeric (mol/kg)
Oral rat chronic toxicity (LOAEL)	1.44	1.588	1.558	1.718	0.993	1.542	1.166	1.724	5.319	Numeric (log mg·kg⁻¹_bw·day⁻¹)
Hepatotoxicity	No	No	No	No	No	Yes	No	No	No	Categorical (yes/no)
Skin sensitization	No	No	No	No	No	No	No	No	No	Categorical (yes/no)
T pyriformis toxicity	0.823	0.301	0.301	0.289	0.305	0.31	0.314	0.285	0.285	Numeric (log µg·L⁻¹)
Minnow toxicity	−0.207	−0.293	0.097	−0.073	0.424	1.09	−0.212	−0.815	16.823	Numeric (log mM)

Abbreviations: BBB, blood–brain barrier; CNS, central nervous system; LOAEL, lowest observed adverse effect level; ADMET, absorption, distribution, metabolism, excretion, and toxicity.

Material and Methods

General

General experimental procedures are given in the supplementary file.

Extraction and Isolation

M coriacea dried twig powder (1.7 kg) was extracted with n-hexane, ethyl acetate, and 90% methanol (3 × 5 L each solvent, room temperature, overnight). After low-pressure concentration of the corresponding extracts, the yields obtained were n-hexane (5.5 g), ethyl acetate (28.6 g), and methanol (50 g). The methanol extract (MCR3, 50 g) was fractionated by column chromatography using gradient elution with CH₂Cl₂/MeOH (100/0 to 50/50, v/v), then with CH₂Cl₂/MeOH/H₂O (50/50/0.1% to 0/90/10, 0/80/20, and 0/50/50 v/v), and finally MeOH/H₂O/AcOH (50/50/0.1% to 0/100/0.1%) to provide 33 fractions (MCR3.1→MCR3.33). Fraction MCR3.12 (50.5 mg), after MeOH-based recrystallization, gave 5 (15.3 mg) as an orange powder. Fraction MCR3.18 (2.58 g) was first passed through a RP-18 column (acetone/H₂O 6/4), followed by repeated purification by column chromatography (n-hexane/CH₂Cl₂/MeOH/NH₃ 10/90/0.1%/0.1%) to provide 6 (2.6 mg) and 3 (4.3 mg). Column chromatography of fraction MCR3.24 (2.15 g) (n-hexane/CH₂Cl₂/MeOH/NH₃ 10/90/0.1%/0.1%) provided 8 subfractions (MCR3.24.1→MCR3.24.8). Subfraction MCR3.24.6 (70.5 mg), after Sephadex LH-20 column chromatography, gave 1 (3.4 mg). Fraction MCR3.33 (2.78 g), by silica-gel chromatography and elution first with CH₂Cl₂/MeOH (60/40 to 10/90), then with MeOH/H₂O/AcOH (50/50/0.1%), afforded 11 subfractions (MCR3.33.1→MCR.3.33.11). Subfraction MCR3.33.7 (722 mg), after purification by silica-gel column chromatography (MeOH/NH₃ 100/0.1%), followed by Sephadex LH-20 column chromatography (MeOH/H₂O 90/10), yielded 4 (3.7 mg). Subfraction MCR3.33.10 (130 mg) was fractionated using a Sephadex LH-20 column (MeOH 100%) to yield 2 (56.6 mg). M coriacea dried leaf powder (1.1 kg) was extracted using a similar procedure as that described for the twigs. The extraction provided n-hexane (6.5 g), ethyl acetate (69.2 g), and methanol (35.7 g) extracts. The ethyl acetate extract (MCLE, 69 g) was fractionated by silica-gel column chromatography eluting with n-hexane/EtOAc (100/0, 90/10, and 0/100), and then with EtOAc/MeOH (90/10, 50/50 to 0/100) to give 19 fractions (MCLE1→MCLE19). Fraction MCLE15 (2.5 g), after repeated silica-gel column chromatography (CH₂Cl₂/MeOH 98/2, 90/1, 80/20, to 0/100), yielded compound 7 (4.5 mg). From the methanol extract (MCRLM, 35 g) Diaion column chromatography (H₂O/MeOH 100/0, 80/20, 20/80, 0/100) provided 16 fractions (MCRLM1→MCRLM16). Fraction MCRLM6 (8.0 g), by silica-gel column chromatography (acetone/MeOH/AcOH 10/90/0.1%), gave 9 subfractions (MCRLM6.1→MCRLM6.9). Purification by silica-gel column chromatography (n-hexane/CH₂Cl₂/n-BuOH/AcOH 1/1/1/1/0.1%) of subfraction MCRLM6.9 yielded 8 (47.3 mg).

Computation

Quantum Chemical Calculation

Molecular geometrical and electronic structures (of 1–8) were studied using DFT performed on Gaussian 09 without symmetry constraints;³⁷ GaussView 05 was responsible for visualization. Functional M052X and 6–311++G(d,p) basis sets³⁸ determined optimized geometries, potential energy surface (PES), and MEP. Frozen-core approximation of non-valence-shell electrons was based on the basis set def2-TZVPP,³⁹ resulting in single-point energies. Resolution-of-identity (RI) approximation was applied to each iteration. M052X/def2-TZVPP⁴⁰ was also used for frontier orbital analysis using NBO 5.1. In principle, the HOMO (E_HOMO) energy and the LUMO (E_LUMO) can be assigned to the tendencies of electron donors and acceptors, respectively. Energy gap ΔE = E_LUMO–E_HOMO provides molecular electronic excitability. These values can also be related to ionization potential (I = -E_HOMO) and electron affinity (A = -E_LUMO), which can be used for evaluation of frontier molecular orbital parameters, ie electronegativity (χ = (I + A)/2), hardness (η = (I-A)/2), and softness (S = 1/η). The interpretations accord to Koopman's theorem.⁴¹

Molecular Docking Simulation

Ligand–protein interactability was predicted by a docking-based technique using MOE 2015.10, typically, following four main steps.^42,43,44

Step 1. Input preparation.

Sources for simulation input: data for biological assemblies of α-glucosidase and tyrosine phosphatase 1B (PTP1B) were retrieved from Worldwide Protein Data Bank (Entry ID: PDB-3W37; DOI: 10.2210/pdb3W37/pdb) and UniProtKB (Entry ID: UniProtKB-A0A0U1XP67), respectively. Chemical formulae (1–8) of ligands were given by structural elucidation; the docking references were the formula of control drug acarbose (D).

Protein-active region specification: active range (between ligand and amino acids) was defined within 4.5 Å. Experiment-based attached residues, if any, were deleted. Configuration: strength-5000 tether–receptor method; refinement 0.0001 kcal·mol⁻¹ Å⁻¹; format *.pdb.

Ligand optimization: structural abnormalities (if any) of the ligands were checked. Configuration: energy-minima Conj Grad method; termination 0.0001 kcal·mol⁻¹; allowed interactions 1000; empirical charge-assigning Gasteiger–Huckel method.

Step 2. Docking simulation.

Ligand–protein inhibitory interaction was simulated. Configuration: selected poses 10; iterative solutions 1000; fragmental solutions 200; format *.sdf.

Step 3. Benchmarking simulation.

The ligand molecules were removed from the converged ligand–protein structures and then docked back. Validity conditions: (1) all RMSD values < 1.5 Å and (2) no significant differences (between corresponding systems) in RMSD value.

Step 4. Parameter interpretation.

In principle, DS value represents a pseudo-Gibbs free energy, thus serving as the primary indicator for the stability of a ligand–protein structure or, in other words, its inhibitory effectiveness.

Figure 1 presents the referenced structures (α-glucosidase and tyrosine phosphatase 1B) and the control drug (acarbose). Those for the ligands are given in Figure 2 (after the discussion for structural elucidation).

Figure 1.

Crystal structures of (A) α-glucosidase (PDB-3W37; DOI: 10.2210/pdb3W37/pdb), (B) tyrosine phosphatase 1B (PTP1B; ID: UniProtKB-A0A0U1XP67), and (C) structural formula of controlled drug, acarbose (D).

QSAR-Based Analysis

Druglikeness properties of phytochemicals were predicted by reference of QSAR-derived physical properties (based on the Gasteiger–Marsili method⁴⁵) to Lipinski's rule of 5:⁴⁶ the former output molecular mass (Da), polarizability (Å³) and size (Å), and dispersion coefficients (logP and logS); on the other side, the rule set criteria for a good membrane-permeable candidate, ie (1) molecular mass < 500 Da, (2) hydrogen bond donors ≤ 5, (3) hydrogen bond acceptors ≤ 10, and (4) logP < +5.^47,48 As a result, oral pharmacological suitability, in particular, and biological compatibility, in general, can be drawn out.

ADMET-Based Analysis

ADMET properties of the input structures (1–8) were obtained from a web-based regressive model SwissADME (http://www.swissadme.ch/; October 27, 2022). The system has been developed and maintained by the Swiss Institute of Bioinformatics. The theoretical interpretations were based on the work by Pires et al,⁴⁹ powered for public reference by the University of Melbourne and University of Cambridge (http://biosig.unimelb.edu.au/pkcsm/theory; October 27, 2022). Overall, the pharmacokinetics and pharmacology of the candidates can be predicted.

Conclusions

This study reports for the first time the isolation of aporphine alkaloids from M coriacea and examines their diabetes-related potential based on computational models. The leaf and twig extracts contain 8 aporphine alkaloids: (1) (+)-glaucine N-oxide, (2) N-methyl glaucine acetate, (3) (+)-(S)-glaucine, (4) magnoflorine, (5) pontevedrine, (6) O-methylatheroline, (7) 7-hydroxydehydroglaucine, and (8) mangochinine acetate. Based on quantum mechanics (DFT calculations), geometry optimization found no abnormal structural constraints, and molecular electronic configurations suggest 2, 4, and 8 as the most promising inhibitors against protein structures in general with intensive nucleophilic reactive sites at their nitrogen atoms. Based on classical mechanics, molecular docking simulation verified this, specifically toward 3W37 (α-glucosidase) and PTP1B (tyrosine phosphatase 1B) structures. QSAR and ADMET regressions suggest that all three compounds are suitable for medical applications given both the predicted biocompatibility and pharmacokinetics. Altogether, the results collected encourage further experimental attempts on 2 (N-methyl glaucine acetate), 4 (magnoflorine), and 8 (mangochinine acetate), especially the last, for drug-developing efforts.

Supplemental Material

sj-docx-1-npx-10.1177_1934578X231176926 - Supplemental material for In Silico Investigation of Aporphine Alkaloids Isolated From Magnolia coriacea Against α-Glucosidase and Tyrosine Phosphatase 1B

Supplemental material, sj-docx-1-npx-10.1177_1934578X231176926 for In Silico Investigation of Aporphine Alkaloids Isolated From Magnolia coriacea Against α-Glucosidase and Tyrosine Phosphatase 1B by Pham Thi Ninh, Thanh Q. Bui, Nguyen Thi Dung, Tran Van Sung, Tran Thi Phuong Thao, Chu Thi Thu Ha, Bui Van Thanh, Tran Van Chien, Phan Tu Quy, Nguyen Thanh Triet, Nguyen Minh Thai and Nguyen Thi Ai Nhung in Natural Product Communications

Footnotes

Declaration of Conflicting Interests

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

Ethical Approval

Our institution does not require ethical approval for reporting individual cases or case series.

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 (grant number VAST04.09/20–21). This research was funded by Hue University (grant number DHH2022-01-198). The authors also acknowledge the partial support of Hue University under the Core Research Program, Grant No. NCM.DHH.2020.04.

Informed Consent

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

Statement of Human and Animal Rights

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

ORCID iDs

Tran Van Chien

Nguyen Thanh Triet

Nguyen Thi Ai Nhung

Supplemental Material

Supplemental material for this article is available online.

References

Kahn

. The relative contributions of insulin resistance and beta-cell dysfunction to the pathophysiology of type 2 diabetes. Diabetologia. 2003;46:3‐19.

Zeng

Yin

Wang

Yang

Liu

Hong

. Synthesis of water soluble pentacyclic dihydroxyterpene carboxylic acid derivatives coupled amino acids and their inhibition activities on α-glucosidase. Bioorg Chem. 2019;86:277‐287.

Proença

Freitas

Ribeiro

, et al. α-Glucosidase inhibition by flavonoids: an in vitro and in silico structure–activity relationship study. J Enzyme Inhib Med Chem. 2017;32:1216‐1228.

Cohen

Goedert

. GSK3 inhibitors: development and therapeutic potential. Nat Rev Drug Discov. 2004;3:479‐487.

Kimura

Lee

, et al. Two potent competitive inhibitors discriminating α-glucosidase family I from family II. Carbohydr Res. 2004;339:1035‐1040.

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;22:960‐964.

Lebovitz

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

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.

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.

10.

Poivre

Duez

. Biological activity and toxicity of the Chinese herb Magnolia officinalis Rehder & E. Wilson (Houpo) and its constituents. J Zhejiang Univ B. 2017;18:194‐214.

11.

Vieira

MNN

Lyra e Silva

Ferreira

De Felice

. Protein tyrosine phosphatase 1B (PTP1B): a potential target for Alzheimer’s therapy? Front Aging Neurosci. 2017;9:28197094. doi:10.3389/fnagi.2017.00007

12.

Amina

Marianne

Ahmed

Lina

. A review of alpha-glucosidase inhibitors from plants as potential candidates for the treatment of type-2 diabetes. Phytochem Rev. 2022;21:1049‐1079. doi:10.1007/s11101-021-09773-1

13.

Sukhbir

Ajmer

Rupanshi

, et al. Recent updates on development of protein-tyrosine phosphatase 1B inhibitors for treatment of diabetes, obesity and related disorders. Bioorg Chem. 2022;121:105626. doi:10.1016/j.bioorg.2022.105626

14.

Lee

. Use of magnolia (Magnolia grandiflora) seeds in medicine, and possible mechanisms of action, in: Nuts Seeds Heal. Dis Prev. Elsevier. 2011:727‐732.

15.

Cicuzza

Newton

Oldfield

. The red list of Magnoliaceae. IUCN: International Union for Conservation of Nature; 2007.

16.

IUCN. Global Tree Specialist Group, 2014. Magnolia coriacea. The IUCN Red List of Threatened Species 2014:e.T39016A2885659.

17.

Ngan

Hiep

Thanh

. Some species of genera Michelia L. at Bat Dai Son Natural Reserve, Quan Ba District, Ha Giang Province. VNU J Sci Nat Sci Technol. 2014;30:61‐70.

18.

CTT

Van Thanh

Thuy

DTT

. Chemical composition and antimicrobial activity of the essential oil from leaves of Magnolia coriacea (Hung T. Chang & BL Chen) figlar growing in Vietnam. Acad J Biol. 2020;42:135‐144. doi:10.15625/2615-9023/v42n3.14739

19.

Ninh

Dung

Van Loc

CTT

Thao

TTP

Van Chien

. Phytochemistry of the aerial parts of Magnolia coriacea collected in Ha Giang, Viet Nam. Vietnam J Chem. 2022;60:667‐673.

20.

Patel

Mishra

. Magnoflorine from Tinospora cordifolia stem inhibits α-glucosidase and is antiglycemic in rats. J Funct Foods. 2012;4:79‐86.

21.

Suthiphasilp

Maneerat

Rujanapun

, et al. α-Glucosidase inhibitory and nitric oxide production inhibitory activities of alkaloids isolated from a twig extract of Polyalthia cinnamomea. Bioorg Med Chem. 2020;28:115462.

22.

Mukusheva

Nurmaganbetov

Ismagulova

, et al. Synthesis and phagocytosis-stimulating activity of harmine and glaucine N-oxides. Pharm Chem J. 2011;45:458‐460.

23.

Chiu

Chou

Chen

Wang

Chen

. Bio-functional constituents from the stems of Liriodendron tulipifera. Molecules. 2012;17:4357‐4372.

24.

Menezes

LRA

Costa

CODS

da C Rodrigues

ACB

, et al. Cytotoxic alkaloids from the stem of Xylopia laevigata. Molecules. 2016;21:890.

25.

Blanco

Castedo

Cortes

Villaverde

. Alkaloids from Sarcocapnos saetabensis. Phytochemistry. 1991;30:2071‐2074.

26.

Harrigan

Gunatilaka

AAL

Kingston

DGI

Chan

Johnson

. Isolation of bioactive and other oxoaporphine alkaloids from two annonaceous plants, Xylopia aethiopica and Miliusa cf. banacea. J Nat Prod. 1994;57:68‐73.

27.

Chang

Wei

Teng

. Antiplatelet aggregation constituents from Annona purpurea. J Nat Prod. 1998;61:1457‐1461.

28.

Lee

Chen

Huang

Chen

. Two dibenzopyrrocoline alkaloids from Litsea cubeba. J Nat Prod. 1996;59:80‐82.

29.

Qiu

Liu

Zhao

Xia

Farnsworth

Fong

HHS

. (1r,2S)-(−)-Mangochinine, a naturally-occurring dibenzopyrrocoline alkaloid from Manglietia chingii. Tetrahedron Lett. 1998;39:4167‐4170.

30.

Rad

Ardjmand

Esfahani

Khodashenas

. DFT calculations towards the geometry optimization, electronic structure, infrared spectroscopy and UV–vis analyses of Favipiravir adsorption on the first-row transition metals doped fullerenes;a new strategy for COVID-19 therapy. Spectrochim Acta Part A Mol Biomol Spectrosc. 2021;247:119082.

31.

Omar

. Indistinguishable particles in quantum mechanics: an introduction. Contemp. Phys. 2005;46:437‐448.

32.

Rosenberg

. Electrical conductivity of proteins. Nature. 1962;193:364‐365.

33.

Kharkyanen

Petrov

Ukrainskii

. Donor-acceptor model of electron transfer through proteins. J Theor Biol. 1978;73:29‐50.

34.

Cordes

Giese

. Electron transfer in peptides and proteins. Chem Soc Rev. 2009;38:892‐901.

35.

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;11:35765‐35782. doi:10.1039/D1RA04461C

36.

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;11:11959‐11975. doi:10.1039/D1RA00441G

37.

Frisch

Trucks

Schlegel

, et al. Gaussian 09, Gaussian Inc. 2009.

38.

Markovi

Dimitri

. Mechanistic study of the structure – activity relationship for the free radical scavenging activity of baicalein. J Mol Model. 2011;17:2575‐2584. doi:10.1007/s00894-010-0942-y

39.

Weigend

Ahlrichs

. Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to rn: design and assessment of accuracy. Phys Chem Chem Phys. 2005;7:3297‐3305.

40.

Reed

Weinstock

Weinhold

. Natural population analysis. J Chem Phys. 1985;83:735‐746.

41.

Koopmans

. Über die Zuordnung von Wellenfunktionen und Eigenwerten zu den Einzelnen Elektronen Eines Atoms. Physica. 1934;1:104‐113.

42.

Tarasova

Poroikov

Veselovsky

. Molecular docking studies of HIV-1 resistance to reverse transcriptase inhibitors: mini-review. Molecules. 2018;23:11‐13.

43.

Thai

Tran

NVK

Nguyen

TTH

Tran

. Computational assay of Zanamivir binding affinity with original and mutant influenza neuraminidase 9 using molecular docking. J Theor Biol. 2015;385:31‐39.

44.

Du Ngo

Tran

Thai

. Computational predictive models for P-glycoprotein inhibition of in-house chalcone derivatives and drug-bank compounds. Mol Divers. 2016;20:945‐961.

45.

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.

46.

Gasteiger

Marsili

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

47.

Ahsan

Samy

Khalilullah

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

48.

Mazumdera

Chakraborty

Sena

Vadrab

Dec

Ravi

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

49.

Pires

DEV

Blundell

Ascher

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

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

2.50 MB