Hypoglycemic Effect and Experimental Validation of Scutellariae Radix based on Network Pharmacology and Molecular Docking

Abstract

Scutellariae Radix (SR) is a well-known traditional herb that has good pharmacological effects against diabetes. However, the mechanism of SR against diabetes is not clear. In this study, the ingredient–target–pathway relationship and hypoglycemic effect of SR on diabetes were explored using network pharmacology, molecular docking and an animal experiment. The targets of SR and diabetes were mined. The selected targets were studied using Gene Ontology (GO) enrichment analysis and pathway enrichment analysis. The network of active components, targets and pathways was integrated to analyze the ingredient–target–pathway relationship. Then, the correspondence between the active components and targets was verified using molecular docking. Finally, an animal experiment was used to verify the hypoglycemic effect of SR. There were 52 components and 22 targets for the hypoglycemic effect of SR. We identified 18 biological processes, 9 cellular components, 15 molecular functions and 25 signaling pathways. Molecular docking results indicated that the targets of diabetes bound strongly to the main components. The animal experiments showed that SR could significantly decrease the blood glucose level of diabetic rats (p ≤ 0.05). This study explored the potential targets and signaling pathways of SR in diabetes, and the results may help to illustrate the hypoglycemic mechanism of SR.

1. Introduction

Diabetes mellitus is a metabolic disorder caused by a decrease in insulin secretion and/or lower insulin activity [1]. The global report of the World Health Organization (WHO) shows that from 1980 to 2019, the number of diabetic patients has almost doubled to 422 million, and it is expected to increase to 693 million by 2045 [2]. Although chemotherapy and/or insulin injections could decrease the blood glucose level [3], there is no effective treatment for diabetes.

Nowadays, traditional Chinese medicine has attracted more and more attention due to its remarkable therapeutic effects and low side effects on diabetes [4,5]. As a perennial herb in Labiatae, Scutellariae Radix (SR), well known as Huang-Qin in Chinese, is derived from the dry roots of Scutellaria baicalensis Georgi. SR is a famous traditional herb in East Asia, especially in China. Recent experimental studies found that SR has good pharmacological effects against diabetes [6,7,8,9]. Zhao Lijuan et al. [6] and Xiao Suwei et al. [7] found that SR could ameliorate hyperglycemia by adjusting the gut microbiota and bile acid metabolism in diabetic rats. Sun Min Park et al. [8] revealed that the extracts of SR had an insulinotropic action in cell experiments of pancreatic islets. Fang Penghua et al. [9] studied the major active ingredients of SR (baicalin and its aglycone baicalein) on metabolic disorders and proposed a novel treatment strategy against hyperglycemia, insulin resistance and type 2 diabetes. However, the hypoglycemic components and mechanism of SR in diabetes are not clear.

Network pharmacology is widely used in drug research. It uses bioinformatics, molecular biology and database systems to study the drug–target–pathway–disease relationship [10]. Therefore, as a useful tool, it could help to further study the role of traditional Chinese medicine in the treatment of diseases [10,11]. Molecular docking is a method that is used to simulate the interaction between a ligand (active component) and a receptor (biological macromolecule or target). The binding ability between the active components and the target is studied through molecular docking technology [12]. Qiqiang Zhang et al. revealed the active components, targets and molecular mechanisms of the Gandhi capsule in treating diabetic nephropathy through network pharmacology and molecular docking technology [12]. L. Chen et al. studied the Yangxinshi tablet in the treatment of heart failure using a network pharmacology–molecular docking strategy [11].

In this study, the potential targets and signal pathways of SR in diabetes were identified and the potential underlying mechanisms were uncovered by constructing the ingredient–target–pathway network. The binding ability of ingredients and targets were analyzed via molecular docking. Furthermore, an animal experiment was used to confirm the hypoglycemic effect of SR.

2. Materials and Methods

2.1. Reagents

Streptozotocin (STZ, no. WXB8D5718V) was obtained from Sigma Aldrich Company Co. Ltd. (Saint Louis, MO, USA). Sodium citrate (no. GB/T 16493-1996), citric acid (no. GB/T 9855-2008) and ethanol (no. 64-17-5) were supplied by Tianjin Hengxing Chemical Reagent Manufacturing Co. Ltd. (Tianjin, China). Metformin hydrochloride tablets (no. H11021518) were purchased from Bei-Jing-Jing-Feng Pharmaceutical Co., Ltd. (Beijing, China). Normal saline (no. H13023201) was supplied by Shijiazhuang-Si-Yao Co., Ltd. (Shijiazhuang, China). The blood glucose meter was the product of Jiangsu Yuwell Medical Equipment Co., Ltd. (Jiangsu, China).

SR was obtained from Inner Mongolia Province, China. The voucher specimen was deposited at Xinxiang Medical University (Xinxiang, China). About 5.0 g of SR was first decocted with 100 mL water for 1.0 h, and then 50 mL water for 0.5 h. Then, the above extracts were mixed and concentrated to 48 mL.

2.2. Chemical Components and Targets

TCMSP (https://www.tcmsp-e.com/, accessed on 1 July 2020) is an open database that can provide information about the chemical composition, chemical structure and clinical application of traditional Chinese medicines. The chemical components of SR were obtained from the TCMSP database. The chemical structures were drawn with ChemOffice software (Version 8.0, CambridgeSoft Corporation, Cambridge, MA, USA) [13]. The Batman-tcm database (http://bionet.ncpsb.org.cn/batman-tcm/, accessed on 1 July 2020) is a good bioinformatics analysis tool for studying the molecular mechanisms of traditional Chinese medicine [14]. PharmMapper (http://www.lilab-ecust.cn/pharmmapper/, accessed on 1 July 2020) is a useful pharmacophore-matching database for target identification [15]. The targets of components were obtained from the TCMSP database, Batman-tcm database and Pharmmapper database.

2.3. Selection of Targets for SR in Diabetes

Drugbank (https://go.drugbank.com/, accessed on 1 July 2020) is an online database containing drug and target information. The related targets of diabetes were screened in the DrugBank database. The potential targets of SR were compared with the targets of diabetes to confirm the targets of SR in diabetes.

2.4. Network of Targets

The String database (https://cn.string-db.org/, accessed on 1 July 2020) can analyze protein interaction information and calculate the confidence scores of all protein interactions. A score, ranging from 0 to 1, indicates confidence in the interaction between targets, and medium confidence (score ≥ 0.4) was used in this study [16]. The targets of SR in diabetes were analyzed in the Homo sapiens species mode. At last, the network of interrelationships between targets was analyzed.

2.5. Gene Ontology (GO) and KEGG Pathway Enrichment Analysis

The Metascape database (https://metascape.org/gp/index.html, accessed on 1 July 2020) was used for the target annotation and analysis. First, the targets of SR in diabetes were input into the Metascape database. Then, the parameters were set as follows: Homo sapiens for the species option, 1 for the minimum overlap and 0.05 for the p-value. Finally, the GO analysis and KEGG pathway enrichment analysis were carried out.

2.6. Network Analysis

Cytoscape software was used to analyze molecular interactions and integrate targets and biological pathways. In this study, the active ingredients, potential targets and KEGG pathways were integrated to obtain the ingredient–target–pathway relationship network.

2.7. Animal Experiment

Sprague Dawley rats (60 adult males weighing 200 to 300 g) were supplied by the Laboratory Animal Centre at Xinxiang Medical University (Xinxiang, China). The rats were raised with sufficient feed and drinking water at 25 ± 2 °C and exposed to light for 12 h a day. After one week of adaptive feeding, the rats were randomly divided into a normal group, a model group, a positive group and an SR group, as shown in

Table 1. The protocol of the animal experiment.

	Normal Group	Model Group	Positive Group	Scutellariae Radix Group
N	6	6	6	6
Diet before modeling	Standard diet	High-fat and high-sugar diet	High-fat and high-sugar diet	High-fat and high-sugar diet
Streptozotocin	/	50 mg/kg	50 mg/kg	50 mg/kg
Dosage (g/kg)	Normal saline	Normal saline	Metformin, 0.16 g/kg	Scutellariae Radix extract, 1.05 g/kg *

* The dosage “1.05 g/kg” was for raw Scutellariae Radix. Before administration, raw Scutellariae Radix was processed according to the method in Section 2.1: Reagents.

. Then, the rats were raised with different diets for one week. After 12 h starvation and enough water, the rats were administered STZ through an intraperitoneal injection. The STZ (1% m/v) was dissolved in a citric acid–sodium citrate buffer (pH 4.2–4.5). The normal group was treated with the same volume of buffer solution. If the fasting blood glucose of rats exceeded 16.6 mmol/L, it was judged that the diabetic model was successful. Then, 6 eligible rats were randomly selected from each group and administered each day via gavage for 4 weeks. The rats were monitored weekly for blood glucose, water intake and body weight. Four weeks later, the rats were anesthetized with isoflurane first, and then the necks were pressed down while its tail was sharply pulled to the rear. The pancreata of rats were collected and fixed with 4% paraformaldehyde for HE staining.

This study was approved by the Ethics Committee of Xinxiang Medical University (no. XXLL-2020055) and followed the internationally accepted principles for laboratory animal use and care, such as the European Community guidelines (EEC Directive of 1986; 86/609/EEC) and Guidelines for the Euthanasia of Animals (2020).

2.8. Molecular Docking Analysis

The chembio3D (ChemBio3D, Ultra 14.0) function of ChemOffice software was used to convert the chemical structure of the ingredient into a three-dimensional format. The three-dimensional protein structure of the target was obtained from the PDB database (http://www.rcsb.org/pdb/home/home.do, accessed on 1 July 2020). The target protein was routinely processed (hydrogenation, dehydration, charging, etc.) by using Autodock tools software (version 1.5.6) (La Jolla, CA, USA). Then, the structures of the protein and ingredient were converted into pdqpt format files. Autodock tools software was used to set the molecular docking parameters, and Autodock Vina software (version 1.1.2) was used for molecular docking. Finally, PyMOL software (version 1.7.2.1) (Schrödinger, Inc., New York, NY, USA) was used to analyze the docking results.

2.9. Statistical Analysis

In this study, all data were analyzed in the form of mean ± standard deviation using the IBM SPSS Statistics 26 software (Armonk, NY, USA). In the statistical comparison, p < 0.05 indicated there was statistical significance.

3. Results and Discussion

3.1. Hypoglycemic Ingredients and Targets

A total of 140 active ingredients with 408 related targets were obtained. According to the intersection analysis of the diabetic targets and the SR targets, 22 targets were obtained, as shown in

Table 2. The potential hypoglycemic targets of Scutellariae Radix.

Target Gene	Targets	Uniprot ID	Number of Ingredients
HDAC2	Histone deacetylase 2	Q92769	24
PTGS1	Prostaglandin G/H synthase 1	P23219	17
PPARG	Peroxisome proliferator-activated receptor gamma	P37231	12
PPARD	Peroxisome proliferator-activated receptor delta	Q03181	10
PRKAB1	5′-AMP-activated protein kinase subunit beta-1	Q9Y478	9
ACSL4	Long-chain-fatty-acid–CoA ligase 4	O60488	5
GAA	Lysosomal alpha-glucosidase	P10253	3
MGAM	Maltase-glucoamylase, intestinal	O43451	3
CFTR	Cystic fibrosis transmembrane conductance regulator	P13569	3
SLC6A3	Sodium-dependent dopamine transporter	Q01959	2
AGTR1	Type-1 angiotensin II receptor	P30556	1
CYP2B6	Cytochrome P450 2B6	P20813	1
CYP2C8	Cytochrome P450 2C8	P10632	1
KCNJ1	ATP-sensitive inward rectifier potassium channel 11	Q14654	1
KCNJ8	ATP-sensitive inward rectifier potassium channel 8	Q15842	1
LPL	Lipoprotein lipase	P06858	1
RXRA	Retinoic acid receptor RXR-alpha	P19793	1
RXRB	Retinoic acid receptor RXR-beta	P28702	1
RXRG	Retinoic acid receptor RXR-gamma	P48443	1
SLC22A1	Solute carrier family 22 member 1	O15245	1
SLC6A2	Sodium-dependent noradrenaline transporter	P23975	1
SLC6A4	Sodium-dependent serotonin transporter	P31645	1

. There were 52 SR components related to these 22 targets (as listed in Table S1), and 15 components (listed in

Table 3. The chemical information of active ingredients for Scutellariae Radix.

No.	Ingredient	Molecular Formula	CAS No.	Number of Targets
1	EIC	C18H32O2	60-33-3	7
2	1,1,6-trimethyl-2H-naphthalene	C13H16	30364-38-6	5
3	Baicalin	C21H18O11	21967-41-9	4
4	Baicalein	C15H10O5	491-67-8	4
5	Methyl linolelaidate	C19H34O2	2566-97-4	4
6	Methyl palmitelaidate	C17H32O2	3913-63-1	4
7	Methyl (Z)-cinnamate	C10H10O2	/	4
8	Methyl icos-11-enoate	C21H40O2	2390-09-2	4
9	Tyrosol	C8H10O2	501-94-0	3
10	Palmitic acid	C16H32O2	57-10-3	2
11	Sitoside	C35H60O6	474-58-8	2
12	5-(2-hydroxyethyl)-2-methoxyphenol	C9H12O3	50602-41-0	2
13	P-coumaric acid	C9H8O3	501-98-4	2
14	Hyacinthin	C8H8O	122-78-1	2
15	WLN	C7H6O	100-52-7	2

) had more than two targets.

3.2. Relationship between Targets

The relationship between these 22 targets was obtained through the transformation using the String database. As shown in Figure 1, the long-chain acyl CoA synthase 4 (ACSL4), cytochrome P450 2B6 (CYP2B6), lipoprotein lipase (LPL), maltase-glucoamylase (MGAM), peroxisome proliferator-activated receptor D (PPARD), peroxisome proliferator-activated receptor G (PPARG) and prostaglandin G/H synthase 1 (PTGS1, cyclooxygenase 1, COX1) were closely connected.

Peroxisome proliferator-activated receptor (PPAR) belongs to the nuclear receptor superfamily of regulatory factors, and its function is mainly related to the regulation of genes on glucose and lipid metabolism [17]. Studies found that PPARA, PPARG and PPAR signal transduction pathways and the adipocytokine signal pathway were upregulated to prevent hyperglycemia, and PPARG receptors regulated lipid uptake and lipogenesis through gene transcriptional activation [18]. On the other hand, PPAR could mediate the induced transcription of the LPL gene [19]. LPL is an important part of the synthesis of triglycerides and it plays a key role in lipoprotein and cholesterol metabolism [20].

MGAM can hydrolyze linear α-1,4-linked oligosaccharide substrates and plays an essential role in the production of human intracavity glucose; therefore, it is an effective drug target for the treatment of diabetes and obesity [21,22].

The targets of methyl palmitelaidate, 1,1,6-trimethyl-1,2-dihydronaphthalene and palmitic acid were PTGS1, CYP2B6 and CYP2C8, respectively. These three targets were related to the cyclooxygenase pathway and arachidonic acid epoxygenase activity, and prostaglandin H synthase (PGHS) was a rate-limiting enzyme. Studies showed that the PGHS-2 inhibitors could prevent streptozotocin (STZ)-induced diabetes [23]. Studies found that the expression of prostaglandin synthase 2 (COX2) in diabetic patients was significantly higher than that in healthy controls, and the COX-2 signal may lead to high inflammatory levels and apoptosis in diabetes [24,25].

3.3. Molecular Docking Results

As shown in Table 2 and Figure 1, four targets (PTGS1, PPARG, MGAM and GAA) were selected for molecular docking with the main active components of SR. It is generally believed that the smaller the binding free energy (affinity, unit: kcal·mol⁻¹) between a molecule and a protein, the more tightly and firmly the two would bind. A binding energy smaller than −5.0 kcal·mol⁻¹ means the active component and protein have good binding activity, and smaller than −7.0 kcal·mol⁻¹ means strong binding activity. The docking results were shown in Figure 2. As shown in Figure 2A, baicalin docked to MGAM through glutamine (GLN) and glycine (GLY) with −8.4 kcal·mol⁻¹ of intermolecular binding energy. Figure 2B indicates the sitoside was bound to GAA targets via glutamate (GLU) and threonine (THR) with −7.8 kcal·mol⁻¹. As shown in Figure 2C, baicalein was bound to the GAA target via aspartic acid (ASP) and GLY with −7.8 kcal·mol⁻¹. Figure 2D shows 4-hydroxycinnamic acid docked to PPARG targets through serine (SER) with −6.2 kcal·mol⁻¹. The molecular docking results illustrated that SR had good pharmacological effects against diabetes.

3.4. GO Analysis, KEGG Analysis and Ingredient–Target–Pathway Network

As shown in Figure 3A and Tables S2–S4, 15 molecular functions were identified (p ≤ 0.01), including monoamine transmembrane transporter activity, nuclear receptor activity, maltose alpha-glucosidase activity and arachidonic acid epoxygenase activity. As shown in Figure 3B and Table S3, nine cellular components were obtained (p ≤ 0.01), including RNA polymerase II transcription regulator complex, presynaptic membrane, ion channel complex and apical plasma membrane. As shown in Figure 3C and Table S2, 18 biological processes were identified (p ≤ 0.01), including a fatty acid metabolic process, maltose metabolic process and positive regulation of lipid localization. As shown in Figure 3D and Table S5, 25 KEGG pathways were obtained (p ≤ 0.05), including the PPAR signaling pathway, arachidonic acid metabolism, galactose metabolism, and starch and sucrose metabolism. Combining the results of the KEGG analysis, molecular docking and relationship between targets, the mechanism of SR in diabetes may be through regulating the metabolism of sugar and resisting the oxidation function.

The active ingredients, potential targets and KEGG pathways were integrated to establish the ingredient–target–pathway relationship diagram. As shown in Figure 4, there were 52 chemical components, 22 targets and 25 pathways for SR in diabetes. Through network pharmacology and molecular docking, it was found that baicalin, sitoside, baicalein and 4-hydroxycinnamic acid of SR could exert hypoglycemic effects through PPAR, MGAM, PTGS1 and so on.

3.5. Animal Experiment

3.5.1. Blood Glucose

As listed in

Table 4. The levels of blood glucose, body weight and drinking water before and after modeling (mean ± SD).

	Group	Normal Group	Model Group	Positive Group	Scutellariae Radix Group
N		6	6	6	6
Blood glucose (mmol/L)	Before modeling	5.2 ± 0.3	4.7 ± 0.9	5.1 ± 0.5	5.1 ± 0.4
	After modeling	4.9 ± 1.0	23.1 ± 2.4 #	20.6 ± 3.0 #	22.7 ± 4.4 #
	After 4 weeks of treatment	4.7 ± 0.9 *	23.4 ± 6.2	6.2 ± 0.8 *	6.2 ± 0.7 *
Body weight change (g)	After 4 weeks of treatment	37.5 ± 7.7 * ↑	66.3 ± 15.9 ↓	54.5 ± 12.5 *↓	45.3 ± 8.8 * ↓
Drinking water (mL)	Before modeling	19.9 ± 2.0	20.1 ± 2.0	20.4 ± 1.4	19.9 ± 1.3
	After 4 weeks of treatment	54.7 ± 3.1 *	195.5 ± 30.2	198.5 ± 14.0	144.3 ± 24.3 *

For the blood glucose levels before and after modeling, a significant difference between them is shown as # (p < 0.05); after four weeks of treatment, a significant difference compared with the model group is marked as * (p < 0.05); ↑: weight gain; ↓: weight loss.

, the blood glucose of rats, except the rats of the normal group, was more than 16.6 mmol/L; therefore, the diabetic rats were successfully induced. After 4 weeks of treatment, the blood glucose levels of the SR and positive control groups were significantly decreased compared with the rats in the model group (p < 0.05), and no significant difference was found between the SR and positive control groups (p > 0.05). Therefore, the results indicated that SR had a good hypoglycemic effect on diabetes.

3.5.2. Body Weight and Water Drunk

The weights and water drunk of the rats in each group were recorded every week. As shown in Table 4, only the rats of the normal group gained weight, while all the rats of the other three groups lost weight. However, the weight loss level of the SR group and positive control group was significantly smaller than the model group (p < 0.05). The water intake of the diabetic rats was significantly higher than the normal group (p < 0.05). Meanwhile, the amount of water drunk by the SR group was lower than that of the model group (p < 0.05). Therefore, SR had some beneficial effects for diabetic rats in delaying weight loss and reducing water intake.

3.5.3. Effect of SR on the Pancreas in Diabetic Rats

After 4 weeks, the pancreatic tissue was taken out for HE staining. As shown in Figure 5, the islet of normal rats (Figure 5A) was quasi-circular and the cells were closely arranged with a clear nucleus. However, the islet of the diabetic group rats (Figure 5B) changed obviously with an irregular shape; at the same time, the number of islet cells decreased and some islet cells were even completely atrophied with biased nuclei. For the rats of the positive (Figure 5C) and SR groups (Figure 5D), the volume of the pancreas and the number of pancreatic cells were improved, and the cells were arranged neatly with a clear boundary.

4. Conclusions

In this study, 52 SR components, 22 potential targets and 25 KEGG pathways were identified using network pharmacology. The results of molecular docking increased the reliability of SR in the treatment of diabetes. Finally, an animal experiment showed that SR could significantly decrease the blood glucose level of diabetic rats (p ≤ 0.05). In a future study, the hypoglycemic ingredients of an SR extract will be separated and identified. In addition, the targets and signaling pathways will be studied through a cell test and an animal experiment. In summary, further study will be carried out to confirm the ingredient–target–pathway relationship of SR on diabetes. This study explored the potential targets and pathways of SR in diabetes, and the results may help to illustrate the hypoglycemic mechanisms of SR.