Protein Tyrosine Phosphatase 1B Inhibition and Glucose Uptake Potentials of Mulberrofuran G, Albanol B, and Kuwanon G from Root Bark of Morus alba L. in Insulin-Resistant HepG2 Cells: An In Vitro and In Silico Study

Type II diabetes mellitus (T2DM) is the most common form of diabetes and has become a major health problem across the world. The root bark of Morus alba L. is widely used in Traditional Chinese Medicine for treatment and management of diabetes. The aim of the present study was to evaluate the enzyme inhibitory potentials of three principle components, mulberrofuran G (1), albanol B (2), and kuwanon G (3) in M. alba root bark against diabetes, establish their enzyme kinetics, carry out a molecular docking simulation, and demonstrate the glucose uptake activity in insulin-resistant HepG2 cells. Compounds 1–3 showed potent mixed-type enzyme inhibition against protein tyrosine phosphatase 1B (PTP1B) and α-glucosidase. In particular, molecular docking simulations of 1–3 demonstrated negative binding energies in both enzymes. Moreover, 1–3 were non-toxic up to 5 µM concentration in HepG2 cells and enhanced glucose uptake significantly and decreased PTP1B expression in a dose-dependent manner in insulin-resistant HepG2 cells. Our overall results depict 1–3 from M. alba root bark as dual inhibitors of PTP1B and α-glucosidase enzymes, as well as insulin sensitizers. These active constituents in M. alba may potentially be utilized as an effective treatment for T2DM.


Introduction
Diabetes mellitus (DM) is a chronic disease that occurs when the pancreas is no longer able to produce insulin or when the body cannot take full advantage of its insulin. DM is a leading cause of cardiovascular disease, blindness, kidney failure, and lower limb amputation, since it affects the heart, blood vessels, eyes, kidneys, nerves, and teeth [1]. According to the 2015 International Diabetes Federation report, there are approximately 415 million adults with diabetes in the world, and this number will rise to 642 million by 2040. In 2015, five million deaths were attributed to DM, and it caused at least USD 673 billion in health expenditure. Therefore, DM has become a major health problem around the world. Type II DM (T2DM), characterized by resistance to insulin, is responsible for over 90% of the overall cases [2]. In the last several decades, numerous attempts have been made two enzymes, and demonstrate glucose uptake activity in insulin-resistant HepG2 cells which is worthwhile to explore the pharmacological mechanism of M. alba in type II diabetes.

Enzyme Kinetics of PTP1B and α-Glucosidase Inhibition
To explain the mode of enzymatic inhibition, kinetic analysis was performed at different concentrations of the substrates (pNPP for PTP1B and pNPG for α-glucosidase) and inhibitors.
As shown in Table 1, 1-3 showed mixed type inhibition against both PTP1B (K i values; 0.70, 1.02, and 1.98 µM, respectively) and α-glucosidase (K i values; 1.20, 0.90, and 2.51 µM, respectively) (Figures S1 and S2). Since the K i value represents the concentration needed to form an enzyme-inhibitor complex, a lower K i value may indicate more effective inhibition against the enzyme system in the development of preventive and therapeutic agents.

Molecular Docking Simulation of PTP1B Inhibition
Investigation of protein-ligand interaction geometries at the molecular level can be accomplished via molecular docking simulations, a widely used computational approach.
Simulation results are shown in Figures 2 and 3. The PTP1B-1 inhibitor complex at the allosteric site showed a −8.24 kcal/mol binding energy with three hydrogen bonds and interacting residues of Ser187, Glu276, and Tyr152. As illustrated in Figure 2D, bond distances between 1 and the interacting amino acid residues were 3.05 Å for Ser187 and Glu276 and 2.73 Å for Tyr152. In addition, hydrophobic interactions were also observed between Phe196, Pro188, Asn193, Tyr153, Ser151, Phe280, Ala189, Leu192, and Gly277.

Molecular Docking Simulation of α-Glucosidase Inhibition
Molecular docking simulations of 1-3 with α-glucosidase were performed, and the ligand-enzyme complexes of the three test compounds/or acarbose and (Z)-3-butylidenephthalide (BIP) were stably posed in the same pocket of the α-glucosidases by AutoDock 4.2 (Figures 4 and 5). Binding energies of the test compounds with interacting residues, including H-bond interacting residues and hydrophobic interacting residues, along with the number of H-bonds are listed in Table 3. With −8.65 kcal/mol binding energy, 1 showed allosteric inhibition forming four hydrogen bonds with interacting residues Ile262, Ile272, Glu296, and Leu297. As illustrated in Figure 4D, bond distances between 1 and the respective interacting amino acid residues were 2.50, 2.03, 3.05, and 2.66 Å. In addition, hydrophobic interactions were also observed between Arg263, Val266, Gly269, Arg270, Glu271, Thr290, Ser291, Ala292, and His295.

A 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino]-2-deoxyglucose (2-NBDG) test was used to measure glucose uptake in insulin-resistant HepG2 cells. Treatment of HepG2 cells with high dose insulin induced insulin resistance as indicated by the marked reduction in glucose uptake in insulinresistant group
The relative glucose uptake percentages of 1 and 2 at 5 µM concentrations were 140% and 125%, which were higher than the positive control, rosiglitazone (relative % glucose uptake; 123% at 10 µM). Considering the relative % glucose uptake activity of rosiglitazone at 10 µM and test compounds at 5 µM in insulin-resistant HepG2 cells, our test compounds are more potent. All these results suggests that 1-3 from root bark of M. alba are able to induce sensitivity of insulin-resistant HepG2 cells towards insulin thereby increasing the relative percentage glucose uptake. Data shown represent means ± standard deviation of triplicate experiments. ### p < 0.001 indicates significant differences from the control group; * p < 0.05, ** p < 0.01 and *** p < 0.001 indicate significant differences from the 10 −6 M insulin-treated control group.

Effects on PTP1B Expression Level in Insulin-Resistant HepG2 Cells
PTP1B is a negative regulator of insulin signaling, and its high expression is implicated in insulin resistance. Therefore, in order to confirm whether these compounds from M. alba increase insulin sensitivity by inhibiting PTP1B expression, we evaluated the expression in insulin-resistant HepG2 cells by Western blot. Under normal condition, PTP1B is expressed feebly while it is intensified in insulin-resistant state. So the downregulation of PTP1B expression in insulin-resistant HepG2 cells could be the indication of insulin sensitizing property. As shown in Figure 8, all three compounds inhibited PTP1B expression in a dose-dependent manner, disclosing their insulin-sensitizing potential, which was in accordance with the glucose uptake activity of respective compounds. Data shown represent means ± standard deviation of triplicate experiments. ### p < 0.001 indicates significant differences from the control group; * p < 0.05, ** p < 0.01 and *** p < 0.001 indicate significant differences from the 10 −6 M insulin-treated control group.
The relative glucose uptake percentages of 1 and 2 at 5 µM concentrations were 140% and 125%, which were higher than the positive control, rosiglitazone (relative % glucose uptake; 123% at 10 µM). Considering the relative % glucose uptake activity of rosiglitazone at 10 µM and test compounds at 5 µM in insulin-resistant HepG2 cells, our test compounds are more potent. All these results suggests that 1-3 from root bark of M. alba are able to induce sensitivity of insulin-resistant HepG2 cells towards insulin thereby increasing the relative percentage glucose uptake.

Effects on PTP1B Expression Level in Insulin-Resistant HepG2 Cells
PTP1B is a negative regulator of insulin signaling, and its high expression is implicated in insulin resistance. Therefore, in order to confirm whether these compounds from M. alba increase insulin sensitivity by inhibiting PTP1B expression, we evaluated the expression in insulin-resistant HepG2 cells by Western blot. Under normal condition, PTP1B is expressed feebly while it is intensified in insulin-resistant state. So the downregulation of PTP1B expression in insulin-resistant HepG2 cells could be the indication of insulin sensitizing property. As shown in Figure 8, all three compounds inhibited PTP1B expression in a dose-dependent manner, disclosing their insulin-sensitizing potential, which was in accordance with the glucose uptake activity of respective compounds. Figure 7. Effects of 1-3 on insulin stimulated glucose uptake in insulin-resistant HepG2 cells as measured by 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino]-2-deoxyglucose (2-NBDG) method. Data shown represent means ± standard deviation of triplicate experiments. ### p < 0.001 indicates significant differences from the control group; * p < 0.05, ** p < 0.01 and *** p < 0.001 indicate significant differences from the 10 −6 M insulin-treated control group.

Effects on PTP1B Expression Level in Insulin-Resistant HepG2 Cells
PTP1B is a negative regulator of insulin signaling, and its high expression is implicated in insulin resistance. Therefore, in order to confirm whether these compounds from M. alba increase insulin sensitivity by inhibiting PTP1B expression, we evaluated the expression in insulin-resistant HepG2 cells by Western blot. Under normal condition, PTP1B is expressed feebly while it is intensified in insulin-resistant state. So the downregulation of PTP1B expression in insulin-resistant HepG2 cells could be the indication of insulin sensitizing property. As shown in Figure 8, all three compounds inhibited PTP1B expression in a dose-dependent manner, disclosing their insulin-sensitizing potential, which was in accordance with the glucose uptake activity of respective compounds.  Western blotting was performed and protein band intensities were quantified by densitometric analysis. Upper panels display representative blots. Equal protein loading was ensured and normalized against β-actin levels. Values are the mean ± standard deviation of three independent experiments; ### p < 0.001 indicates significant differences from the control group; * p < 0.05, ** p < 0.01 and *** p < 0.001 indicate significant differences from the 10 −6 M insulin-treated control group. At 2.5 and 5 µM concentration, 1 diminished the PTP1B expression almost completely. While 2 showed the expression that was comparable to the normal group at 5 µM concentration. Although 3 could not normalize the condition completely, it downregulated the expression towards normal in a concentration dependent manner. Overall result of PTP1B downregulation by 1-3 in insulin-resistant HepG2 cells coincide with their PTP1B enzyme inhibition results.

Discussion
T2DM is the most common form of DM. It occurs in quick succession and is characterized by inadequacy in insulin secretion and resistance to insulin in its target organs. Several hypoglycemic agents with different mechanisms of action, such as increasing production of insulin, decreasing production of hepatic glucose, limiting absorption of postprandial glucose, and inhibiting gluconeogenesis, have been developed. However, periodic diminishment in efficacy of single and even combined drugs leads to the administration of insulin [31].
The association of insulin resistance with obesity and T2DM and the fact that insulin receptor signaling is mediated by tyrosine (Tyr) phosphorylation have generated great interest in the homeostasis of tyrosine phosphorylation. Among several implicated Tyr phosphatases, PTP1B is the most prominent. Therefore, targeting PTP1B and developing novel PTP1B inhibitors could lead to an effective therapeutic model against diabetes via insulin sensitization, without the weight gain associated with thiazolidinediones [32]. However, use of available synthetic PTP1B inhibitors has been limited due to their dose-dependent side effects and failure in Phase II clinical trials. In addition, natural molecules with improved pharmacodynamics/pharmacokinetic parameters can have advantages over synthetic molecules. The other approach to treating diabetes, particularly non-insulin-dependent diabetes (postprandial hyperglycemia), is to retard glucose absorption via inhibition of the carbohydrate hydrolyzing enzyme α-glucosidase. α-Glucosidase is the principle enzyme in the catalyzation of the final step of carbohydrate metabolism. Inhibiting this enzyme retards the formation of D-glucose from dietary carbohydrates, delays glucose absorption/reduces postprandial plasma glucose level, and finally suppresses postprandial hyperglycemia. The existing clinical α-glucosidase inhibitors (acarbose, miglitol, and voglibose) are associated with gastrointestinal side effects [33]. Therefore, discovery of novel phytochemicals with potent PTP1B and α-glucosidase inhibition with no or fewer side effects is urgent.
Morus alba L. is a well-known TCM herb with a decades-long history of medicinal value. Various parts of this plant (fruits, leaves, branches, and root bark) have been used for medicinal purposes. As part of our ongoing work to discover novel phytochemicals against diabetes, we evaluated the inhibitory potentials of 1-3 that we isolated from root bark of this plant [34] against PTP1B and α-glucosidase. All compounds showed potent inhibition against both enzymes and were superior to reference drugs. Further, in order to validate our enzyme inhibition result, we performed enzyme kinetics. Enzyme kinetic analyses performed with various concentrations of substrates and inhibitors showed mixed-type inhibition against PTP1B and α-glucosidase, implying that these compounds bind to the active site as well as an additional binding site of the respective enzymes. The findings of this study demonstrate that 1-3 might act as preventive or therapeutic agents for type II diabetes by inhibiting PTP1B and α-glucosidase enzyme activity. Because lower K i values indicate tighter binding to the enzyme, the low K i values of these compounds signify that they could be highly effective against PTP1B and α-glucosidase.
Structure-based drug design is one of various vital computational approaches found to be effective in identification of hits for in vitro testing. Molecular docking is an extensively used method of virtually screening small molecules for identification and optimization. Additionally, by using docking score functions, one can accurately predict the binding affinities between ligands and receptors and understand the mechanism behind binding sites interactions. Structural and kinetic analyses have identified two important amino acids, Cys215 in the PTP signature motif (P-loop) and Asp181 in the WPD loop, as the primary active site residues that attack substrates. Once the active sites are blocked by the WPD loop upon inhibitor/substrate binding, dephosphorylation of the phosphocenter of the substrate by Cys215 occurs via nucleophilic attack. Similarly, Arg47, Lys120, and Val49 demonstrate electrostatic, hydrophobic, and H-bonding interactions [35]. In addition, Puius et al. reported the existence of second catalytic site ("site B") in which Arg24 and Arg254 were the important residues including H-bond interactions with Met258 and Gln262, and van der Waals contact with Ile219, Asp48, and Val49 [36]. So, to confirm the inhibition mode of enzyme inhibition, we further conducted molecular docking studies and validated our results using reference ligands (compound 2 as allosteric and compound 23 as catalytic inhibitors).
Molecular docking studies on PTP1B inhibition revealed that inhibition by the reference allosteric inhibitor, compound 2, occurred via Asn193 and Glu276. 1 showed interaction with Glu276 while, 2 and 3 interacted with Asn193 for allosteric inhibition. In addition, for catalytic inhibition, the mode of interaction was through Arg24, Arg254, Asp48, and Tyr46 along with additional amino acid residues compared with the reference inhibitor, compound 23 (Table 2). Since low binding energy indicates high binding affinity, 1-3 showed preference for allosteric inhibition. The enzyme kinetic study revealed 1-3 as mixed inhibitors; however, the molecular docking study was more specific and concluded the inhibition to be mixed but tending toward allosteric mode. Similarly, for α-glucosidase, 1-3 showed low binding energies for catalytic inhibition compared to allosteric inhibition (Table 3), which indicates that these compounds bind preferentially to catalytic sites. This means the test compound mode of inhibition for α-glucosidase was mixed but tending toward catalytic inhibition.
To further investigate the anti-diabetic mechanisms of the test compounds, we evaluated the effect of these compounds on insulin sensitivity in normal versus insulin-resistant HepG2 cells. The insulin-resistant HepG2 cell model is widely used in the field of drug discovery for evaluating anti-insulin resistance bioactivity from plant resources. PTP1B is a well-known negative regulator of the insulin signaling pathway, and its expression is elevated in insulin-resistant conditions [37]. The biological effect of insulin is initiated with insulin binding to the α-subunit of IR and activating the intrinsic tyrosine kinase activity of the β-subunit of the receptor. The activation of IR leads to tyrosine phosphorylation of IRS. Thus phosphorylated IRS then activates PI3K, which subsequently phosphorylates Ser/Thr kinase Akt. Beside this, activation of Akt promotes the translocation of intracellular GLUT4 to plasma membrane and lead to glucose transport. However, in case of insulin resistance state, cell become unresponsive to insulin as a result of which insulin signaling cascade get dysregulated. Studies have shown that overexpression of PTP1B inhibits IR and IRS phosphorylation and leads to insulin resistance [38,39]. In a recent report, PTP1B inhibitor increased the tyrosine phosphorylation of IR, and activated the downstream molecules of insulin signaling such as IRS-1, Akt, and Erk1/2 [40,41]. Similarly, a study on fumosorinone as a novel PTP1B inhibitor showed decreased PTP1B expression, that led to increased phosphorylation of IRβ, IRS-2, Akt, GSK3β and Erk1/2 in insulin-resistant HepG2 cells along with the increased phosphorylation of IRβ, IRS-2 and Akt in diabetic KKAy mice [42]. Increased glucose uptake in insulin-resistant HepG2 cells treated with sample compared to the untreated insulin-resistant/normal group confirms improvement of insulin sensitivity. In the 2-NBDG assay of present study, 1-3 increased insulin-stimulated glucose uptake significantly in a dose-dependent manner. Similarly, in Western blot analysis, all the test compounds suppressed PTP1B expression in a concentration-dependent manner. PTP1B was feebly expressed in the normal group but intensified in the insulin-resistant group. However, its expression declined in the rosiglitazone-and sample-treated groups. Though PTP1B inhibition is recognized as an effective way to improve insulin sensitivity, better understanding of the PTP1B-regulatory mechanism is still lacking. Dysregulation of miRNA-122 contributes to hepatic insulin resistance through PTP1B induction. So, recovering the expression of miRNA-122 down regulate PTP1B and abolish hepatic insulin resistance [43]. In in vitro and in cells, PTP1B recognizes the activated insulin receptor as a substrate and preferentially dephosphorylates tyrosine residue in the activation loop of the β-subunit of insulin receptor. Additionally, IRS-1 is also a potential substrate of PTP1B. The pathology of hyperinsulinemia-induced insulin resistance is associated with impaired Ser/Tyr phosphorylation of IRS-1 and IRS-2 that impairs their interactions with cytoplasmic domain of insulin receptors and deregulate the insulin signaling. As a result, in insulin-resistant condition, PTP1B is highly expressed. However, treatment with PTP1B inhibitors regulate the Ser/Tyr phosphorylation of IRS-1 and IRS-2 thereby down regulating PTP1B expression. Therefore, improvement of insulin sensitivity of high insulin-induced insulin-resistant HepG2 by 1-3 might be via regulation of Ser/Tyr phosphorylation of IRS-1 and IRS-2 in insulin-signaling cascade. Besides, activation of Akt and translocation of intracellular GLUT4 to plasma membrane might also be the underlying mechanism of enhanced glucose uptake by these compounds.
Our results in overall reveal that 1-3 from M. alba and rosiglitazone improve insulin sensitivity by suppressing PTP1B expression. In conclusion, Morus alba root bark contains a familiar TCM herb that has been widely used for various biological disorders, especially T2DM. Notably, our study provided systematic and scientific evidence supporting its pharmacological and therapeutic effects in diabetes. Overall results showed the usefulness of 1-3 from M. alba root bark in the management of T2DM through an increase in the sensitivity of insulin-resistant cells to insulin and an enhancement in glucose uptake via PTP1B down-regulation. However, possible involvement of 1-3 in other mechanisms for improvement of diabetic conditions should be studied further.

Protein Tyrosine Phosphate 1B (PTP1B) Inhibitory Assay
PTP1B inhibitory activity was evaluated using pNPP [44]. Recombinant PTP1B enzyme (0.5 units diluted in PTP1B reaction buffer) was added to a plate with or without sample. The plate was pre-incubated at 37 • C for 10 min, and then substrate (2 mM pNPP) was added. Following incubation at 37 • C for 15 min, the enzymatic reaction was terminated by addition of 10 M NaOH. The absorbance was measured at 405 nm using a microplate spectrophotometer (Molecular Devices, Sunnyvale, CA, USA). Ursolic acid was used as a reference compound.

α-Glucosidase Inhibitory Assay
The enzyme inhibition study was carried out spectrophotometrically using the procedure previously reported [45]. The α-glucosidase activity was determined by measuring release of pNPG at 405 nm using a microplate spectrophotometer (Molecular Devices). Acarbose was used as a reference compound.

Kinetic Study Against PTP1B and α-Glucosidase
To determine the modes of enzyme inhibition, enzymatic inhibition of the test samples was evaluated by monitoring the effects of different concentrations of the substrates (0.5, 1, or 2 mM pNPP for PTP1B and 0.4, 0.8 or 1.6 mM pNPG for α-glucosidase) in the Dixon plots (single reciprocal plots). Lineweaver-Burk plots for inhibition of PTP1B were obtained in the presence of various concentrations of the test compounds (0, 0.18, 0.44 and 1.78 µM for 1; 0, 0.27, 0.54, and 1.07 µM for 2; 0, 0.87, 1.44, and 2.89 µM for 3). Similarly, for the inhibition of α-glucosidase, text concentrations used were 0, 0.71, 1.07, and 1.78 µM for 1; 0, 0.36, 0.9, and 1.79 µM for 2 and 0, 0.72, 2.88, and 5.77 µM for 3. The enzymatic procedure consisted of the same aforementioned assay method. The inhibition constants (K i ) were determined via interpretation of the Dixon plots, where the value of the x-axis implies −K i [46,47].

Cell Culture, MTT Assay and Insulin Resistance Induction
HepG2 (human hepatocarcinoma) cells, purchased from the American Type Culture Collection (HB-8065; Manassas, VA, USA), were maintained in 10% FBS MEM at 37 • C in a humidified atmosphere with 5% CO 2 . Medium was changed every 48 h. Cytotoxicity was evaluated using the MTT assay [50]. Following the method previously described [42], insulin-resistant HepG2 cell model was established. All the other experiment conditions and procedures were similar to those reported in our previous paper [51].

Glucose Uptake Assay
The glucose uptake rate in insulin-resistant HepG2 cells was assessed using the fluorescent D-glucose analog 2-NBDG. Once the induction of insulin resistance was accomplished, cell solutions in the wells were substituted with different concentrations of 1-3 or a positive control, rosiglitazone (10 µM), prepared in serum-free MEM (SFMEM) in respective groups. After 24 h incubation with samples, the media in the wells was replaced with 100 nM insulin diluted in SFMEM and incubated for 30 min. Cells were then incubated with 40 µM 2-NBDG for 30 min and washed with ice-cold PBS. The fluorescence intensity of 2-NBDG was measured on a fluorescence microplate reader (FL ×800, Bio-Tek Instruments, Inc., Winooski, VT, USA) at 485 nm excitation and 528 nm emission wavelengths. Six replicate wells were established, and each experiment was repeated three times.