Arisaema Cum Bile possesses significantly antiepileptic effects through regulation of PI3K/Akt/Nrf2 signaling pathways

Abstract

This paper intended to study the mechanism and active ingredients of ACB anti-epilepsy. The antiepileptic activity of ACB validated in PTZ kindled rats, ACB could increase the seizure latency and reduce seizure duration, attenuate spatial learning and memory deficits, improve hippocampus neuronal damage and regulate unbalanced neurotransmitters. Furthermore, network pharmacology and molecular docking analysis predicted four potential active compounds, in addition, PI3K/Akt signal pathway may be the main signal pathway of ACB anti-epilepsy. In vitro, ACB greatly increased the vitality and reduced apoptosis of PC12 cells exposed to H₂O₂. Additionally, ACB elevated Bcl-2 and downregulated C-caspase-3 and Bax proteins expression. Importantly, ACB improved the phosphorylation of PI3K and Akt in H₂O₂-stimulated PC12 cells, and stimulated the nuclear transfer of Nrf2. These findings indicated that ACB has effective on antiepileptic by activating of the PI3K/Akt/Nrf2 pathway to reduce oxidative stress and neuronal cell apoptosis.

KEYWORDS:

• Arisaema Cum Bile (ACB)
• epilepsy
• network pharmacology
• PI3K/Akt/Nrf2 signaling pathway
• apoptosis

1. Introduction

Epilepsy, a serious chronic neurological disorder, had affected more than 70 million people worldwide, and more than 10 million people in China [1–3]. Epilepsy occurrs on all ages, especially in neonates, young children and the elderly [4,5]. And the convulsions, loss of consciousness for several seconds, myoclonus and prolonged duration of muscle contraction were the main clinical manifestations. At present, over 20 kinds of antiepileptic drugs have been used to treat epileptic seizures, but drug insensitivity still exists [6–8]. Besides, long-term use of antiepileptic drugs may also cause side effects [9,10]. Consequently, more dependable antiepileptic drugs with fewer side effects are required for the clinical management of epilepsy. It is now generally accepted that the most common cause of epileptic is pathological changes in the synapses of neurones, which then lead to the imbalance of excitatory and inhibitory transmitters in the central nervous system, and finally lead to highly synchronized abnormal firing of neurones in the brain [11,12].

Several neuropharmacological studies have provided evidence that the onset of epilepsy is intricately linked to several factors such as neurotransmitters, synaptic connections [13]. Oxidative stress is well recognized as a primary mechanism underlying the development of epilepsy [14]. Numerous investigations have shown an elevation in mitochondrial oxidative stress and consequent cellular damage after prolonged convulsions.

Arisaema Cum Bile (ACB) was a traditional Chinese medicine with “Xi-Feng-Ding-Jing”, and produced by the fermentation of the roots of Arisaematis rhizoma with bile [15]. It was used for clearing heat and reducing phlegm in traditional Chinese medicine theory since the Song Dynasty [16]. In clinic, it is frequently taken with antiepileptic medications in order to lessen the severity of adverse effects such as insomnia, depression and memory impairment. In addition, pharmacological evidence shows that ACB has extensive spasticity relieving and anti-inflammatory effects [16,17]. Su et al [18] found that aqueous extracts of ACB could protect rats against febrile seizures. Chen et al [19] confirmed the antiepileptic activity effects of ACB by patch clamp technology. However, the potential molecular mechanism of the ACB’s antiepileptic effects is still lacking. Bioinformatics analysis has become increasingly popular in recent decades for screening genomic-level gene alterations. The mechanisms of traditional Chinese medicine have also been revealed through the application of network pharmacology, a holistic method that takes into account multiple components, multiple targets and multiple pathways [20,21]. Finding the targets and mechanisms of ACB for treating epilepsy was the focus of our study, which integrated network pharmacology, bioinformatics analysis and in vivo and vitro experiments.

In this study, the rat model of PTZ induced epilepsy was taken as the research object. To predict the mechanism of antiepileptic effects, network pharmacology and molecular docking were employed, and molecular biology was used to confirm the mechanism. The possible target and mechanism of action of ACB were comprehensively studied, which lays an experimental foundation for the further development of ACB. Figure 1 displays the comprehensive flowchart of the study approach.

Figure 1. Comprehensive flowchart of the research methodology in the study.

2. Materials and methods

2.1. Materials and chemicals

Fisher Scientific (USA) supplied LC-MS-grade acetonitrile and methanol. Sigma-Aldrich (USA) supplied 98% formic acid, H₂O₂ and PVDF membrane; ACB (Batch No. 160511) was provided by BaiSheng Pharmaceutical Co., Ltd. (Sichuan, China); TransGen Biotech (Beijing, China) supplied TransSerum^® EQ Fetal Bovine Serum (FBS) (Cat. No. FS201-02); Abclone (Hunan, China) supplied the prestained protein marker, nuclear protein extraction kit, DAPI and CCK-8 kits, Everbright^® Inc (Suzhou, China) supplied JC-1 probe; Beijing 4A Biotech Co. (Beijing, China) supplied the DCFH-DA fluorescent probe, annexin V-FITC/PI kit and ECL luminescence reagent. p-PI3K(Cat. No. ab125568), Bcl-2(Cat. No. ab194583), C-caspase-3(Cat. No.ab2302) and Nrf2 (Cat. No. ab137550) were obtained from Abcam (Cambridge, MA, USA). Akt(Cat. No. sc-5298), p-Akt(Cat. No. sc-377556), PI3K(Cat. No. sc-1637), Bax(Cat. No.sc-70405), β-actin(Cat. No. sc-81178) and Lamin B1(Cat. No. sc-374015) were obtained from the Santa cruz Biotechnology (USA).

2.2. In vivo experiment

2.2.1. Kindling procedure

From SiPeiFu Biotechnology Co., Ltd. (Beijing, China), we obtained male SD rats (180g ± 10 g). The animals were given adaptive feeding for 7 days. Animals were given PTZ to cause convulsions [22,23]. Sixty rats were divided into normal, model, positive and treatment groups by a random selection process, each of which normal and model groups received normal saline, positive group treated with diazepam (0.6 mg/kg), low dose group received 0.3 g/kg/day of ACB, middle dose group received 0.6 g/kg/day and high dose group received 1.2 g/kg/day (n = 10), respectively. The drugs were formulated into a certain volume of liquid according to the dosage, and the volume of intragastrical administration was calculated based on 2 mL per 200 g. Animals were intragastrical administered with these drugs and 30 min following the administration of the pharmaceuticals, rats except normal groups were treated with PTZ (35 mg/kg, i.p.), this procedure was executed once every other day for 30 days.

2.2.2. Morris water maze test

MWM test followed kindling. 1 d before the first training trial, for a period of sixty seconds, the rats were allowed to swim in the pool without a platform. Rats conducted four trials each day with over 30-minute intervals throughout a five-day training period. In each trial, ensure their heads facing the pool wall and entered the water from one of four quadrants. Each rat could swim to find the platform in 60 s, if the rat didn’t discover the platform, manually guidance should be administered, and the escape latency was noted as 60 s. The rat was allowed to stay on the platform for 10 seconds after entering. During the sixth day of the spatial probing test, the covert platform was removed, the rats were free to swim for 60 s in the pool. Smart soft (3.0 version) was applied to capture the track and analysis the data.

2.2.3. Histological analysis

The brain tissues, collected and fixed for 48 h, were dehydrated using ethanol, paraffin-embedded, and then cut into 5 μm slices. H&E and Nissl staining were both performed.

2.2.4. ELISA

The levels of GABA(Cat. No.RXJ303082R), 5-HT(Cat. No.RXJ303050R), Glu(Cat. No.RXJ303150R), dopamine(Cat. No.RXJ302756R) and histamine(Cat. No.RXJ302030R) were detected by ELISA kits (Ruixin Co., Ltd., China). At 4°C, hippocampus samples were homogenized, centrifuged and collected the supernatant for analysis. Each ELISA kit was measured according to the instructions.

2.3. HPLC-Q-Exactive Orbitrap MS/MS assays

Thermo HPLC system (Thermo Scientific, Bremen, Germany) and Hypersil GOLD aQ C18 column (150 mm*2.1 mm, i.d.1.9 μm, Thermo Fisher Scientific, USA) were used to perform chromatographic separation. Mobile phases A was acetonitrile, mobile phases B was 0.1% formic acid in water, which the chromatographic separation was carried out at flow rate 0.3 mL/min and 30°C using a gradient elution. The gradient programme was configured to run for the following durations: 0–3 min, 10%–35% A; 3–18 min, 35%–55% A; 18–21 min, 55%–90% A; 21–25 min, 90%–95% A; 25–25.1 min, 95%–10% A; and 25.1–30 min, 10% A. Injection volume was 2 μL. An ESI source equipped Q-Exactive Orbitrap mass spectrometer was used for high-resolution MS detection and analysis. Full scan data were obtained in both positive and negative modes from m/z 50 to 1500 with a resolution of 30,000.

2.4. Network analysis

Search the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) [24] for the drug chemical components to gain their molecular information and store in MOL format, the Swiss Target Prediction tool (http://www.swisstargetprediction.ch) [25] collected ACB’s target from these MOL format. Next, GeneCards (https://www.genecards.org/) [26] was searched for epilepsy pathogenic targets using “epilepsy” as the keyword. All duplicate targets were eliminated using Microsoft Excel software. Venny (https://bioinfogp.cnb.csic.es/tools/venny/index.html) [27] carried out an analysis of the intersection targets of ACB and epilepsy. Based on STRING database version 11.5 (https://cn.string-db.org), the PPI network was created [28]. These chosen targets were uploaded into DAVID (https://david.ncifcrf. gov/) [29] to perform GO and KEGG pathway enrichment analysis, where targets were only restricted to Homo sapiens. An online tool (http://www.bioinformatics.com.cn) was used to generate a visualization of the top 20 enriched phrases. Cytoscape soft (3.8.2 version) was used to develop the network of drug-components-targets-disease-pathways.

2.5. Molecular docking

Molecular docking was employed to further investigate the binding mode and capacity of active components in ACB to target protein in potential pathway based on the findings of network pharmacology. The PubChem database was used to collect the chemical 3D structures of the primary active substances, then uploaded into OpenBabel (3.1.1 version) for conversion to pdb format. The proteins were downloaded in pdb format from the RCSB database (https://www.rcsb.org), water molecules and the protein’s original ligand were removed using PyMOL (1.1.0 version), then adding hydrogen, building active pocket and docking using AutoDock Tools (version 1.5.7), and the results were visualized using PyMoL.

2.6. In vitro experiment

2.6.1. Cell culture

DMEM medium containing 10% FBS (v/v), penicillin (100 units/mL) and streptomycin (100 μg/mL) was used to cultivate PC12 cells, the temperature was 37°C, and the CO₂ level was 5%.

2.6.2. Determination of cell viability

PC12 cells were planted onto 96-well plates while they were in the logarithmic growth phase of their development. Different concentrations of ACB (2.5 ∼ 320 μg/mL) were pretreated for 6, 12 and 24 h, added CCK-8 solution and cultured at 37°C for an additional half an hour before being detected, then cell survival rate was calculated. Similarly, Using the CCK-8 assay, we examined ACB’s influence on PC12 cell viability after H₂O₂ (200 μM, 4 h) exposure and the presence of PI3K inhibitor LY294002.

2.6.3. Mitochondrial membrane potential (MMP) determination

PC12 cells were seeded into laser confocal dishes and 6-well plates when the fusion degree reaches 80% ∼ 90%. Once the cells had adhered to the plate’s surface completely, varying concentrations of ACB were added to the wells of the treatment group as a pretreatment. Next, 200 μM H₂O₂ was incubated in the cells for 4 h. Finally, PC12 cells were gathered and incubated with JC-1 at 37°C in the dark for 15 min. The cells’ MMP changes were detected by a confocal laser microscopy and flow cytometer.

2.6.4. Cell apoptosis determination

Annexin V-FITC kit was adopted to measure the cell apoptosis. In brief, the cells were collected and stained according to the requirements of the kit instructions, and finally, flow cytometers evaluated PC12 cells apoptosis. Similarly, we also examined the effect of ACB on apoptosis in the presence of PI3K inhibitor LY294002.

2.6.5. ROS determination

DCFH-DA probe was employed to measure the contents of ROS. After being collected, PC12 cells were stained with the DCFH-DA probe for a period of 30 min. The level of ROS in PC12 cells was detected by flow cytometer after washing three times with PBS.

2.6.6. Western blot assay

PC12 cells were treated according to the conditions under “2.6.2”. After the end of the experiment, total cell protein was extracted by RIPA extraction kit. The determination and adjustment of protein concentration were performed by the BCA protein assay kit and loading buffer (5×) respectively. Then 10%SDS-PAGE was employed to isolate the protein(30 μg), followed by blotting on the PVDF membrane and blocking with 5% skimmed milk for one hour. The PVDF membranes were incubated overnight with prediluted (dilution 1:1000) primary antibody of Bax, Bcl-2, PI3K, p-PI3K, Akt, p-Akt, C-caspase-3 and Nrf2 at 4°C, respectively. After incubation, The membranes should be washed with TBST solution for three times. Subsequently, a HPR mixed secondary antibody (dilution 1:5000) incubated the membranes for 1 h. Finally, ECL detection kit was used to stain the proteins band. Image J software (version 1.51) was applied for greyscale analysis of proteins band.

2.6.7. Immunofluorescence assay

PC12 cells were seeded into laser confocal dishes. Once the cells had adhered to the plate’s surface completely, added ACB and LY294002 for pretreatment for 24 h, then added H₂O₂ and continued to incubate for 4 h. Then washing the cells three times, the cells were fixed by adding paraformaldehyde, and then sealed with 5% BSA for 1 h. Then, p-Akt (1:300) and Akt (1:300) antibody were incubated overnight at 4 °C, after that, continued the incubation with a fluorescently assigned secondary antibody for 1 h after rinsing the cells three times with PBS. Finally, added anti-fluorescence quenching agent containing DAPI, then recorded images.

2.7. Statistical analysis

Student t-test in GraphPad Prism 9 software (GraphPad Software Inc., La Jolla, USA) was adopted to analyse data. p < 0.05 was considered the significant level. In vitro experiment for data expressed as mean ± SD (n = 3).

3. Results

3.1. ACB exerts antiepileptic profile in kindled rats

In order to evaluate the potential antiepileptic action of ACB, the latency to seizure as well as the duration of generalized seizures were observed. The results are shown in Table 1, compared with the model group, ACB at the doses of 0.6 and 1.2 g/kg could significantly increase the latency to generalized seizures and reduce the duration of generalized seizures (p < 0.01, respectively), indicating that ACB exhibits antiepileptic activity in PTZ kindled rats.

Table 1. Antiepileptic effect of ACB in PTZ kindled rats.

Group	Seizure latency (s)	Seizure duration(s)
Model	201.33 ± 15.34	88.94 ± 10.89
Positive	411.87 ± 20.09**	32.29 ± 9.98**
0.3 g/kg	209.32 ± 11.33	84.09 ± 11.33
0.6 g/kg	341.09 ± 14.97**	54.98 ± 14.52**
1.2 g/kg	388.43 ± 18.62**	48.56 ± 12.22**

**p < 0 01 vs. the model group, data were presented as mean ± SD (n = 10).

3.2. ACB attenuates spatial learning and memory deficits

To assess how ACB affected spatial learning and memory impairments in kindled rats, the MWM test was employed. The escape latency of all groups gradually decreased during the course of five training days, rats in the model group showed longer escape latencies than those in the normal group and ACB group (1.2 g/kg) (Figure 2(A)). On the final training day, the model group had to swim farther than the other groups to get to the platform. The travelled distance to locate the platform significantly decreased after treatment with ACB (Figure 2(B–C)). On spatial probe test, compared with the model group, the ACB groups considerably increased the crossed platform times and the percent of time in the target quadrant (Figure 2(D–E)).

Figure 2. Effects of ACB on cognitive deficits in PTZ kindled rats. (A) Effects of ACB on decreasing escape latency. (B) Travelled distance in the first and fifth training day. (C) Representative traces in MWM test. (D) The number of times crossing the platform. (E) Percent of time in the target quadrant. * p < 0.05 and **p < 0.01, vs. model group.

3.3. ACB protects hippocampal neurones injury in kindled rats

Epilepsy was prone to induce neuronal cell damage, especially in the CA1 and CA3 regions of the hippocampus [30–32]. According to the staining and injury index (Figure 3(A–C)), PTZ-induced rats showed clear signs of pathology, such as disordered cell arrangement and abnormal cell shape. Fortunately, ACB therapy reversed these pathological changes. Results mentioned above indicate that hippocampal neuronal cell injury caused by epilepsy could mitigate by ACB.

Figure 3. Effects of ACB on morphology and structure of hippocampus of PTZ kindled rats. (A) H&E staining (100× and 400×). (B) Nissl staining (100x and 400x). (C) statistical charts for injury index. **p < 0.01, vs. model group.

3.4. Effects of ACB on GABA, 5-HT, Glu, dopamine and histamine in hippocampus of PTZ kindled rats

Epilepsy has been related to a neurotransmitter imbalance in the brain [33–35]. In model group, the levels of GABA, 5-HT, dopamine and histamine were significantly decreased (p < 0.01) and the Glu contents were increased (p < 0.01) when compared with normal group (Figure 4). Interestingly, ACB (0.6 and 1.2 g/kg) groups significantly reduced Glu contents (p < 0.01) and increased the levels of GABA, 5-HT, dopamine and histamine (p < 0.01). However, compared with the model group, ACB (0.3 g/kg) group could significantly reduce Glu contents (p < 0.01) and showed no significantly effective on GABA, 5-HT, dopamine and histamine (p > 0.05).

Figure 4. Effects of ACB on GABA, 5-HT, Glu, dopamine and histamine in hippocampus. **p < 0.01, vs. model group.

3.5. Identification of compounds in ACB

Within 30 min, more than 30 peaks were detected, Table 2 and Figure 5 shown the 26 main identified compounds based on ion fragmentation features, published spectrogram and literature data.

Figure 5. Full ms chromatogram on negative-ion polarity mode.

Table 2. Precursor and product ions of constituents in ACB.

No.	Compound name	tR(min)	Molecular formula	[M-H]^-	MS/MS (m/z)	Reference
1	L-Malic acid	1.19	C4H6O5	133.0132	115.0024;71.0126;89.0231;107.0351	[67]
2	Xanthine*	1.23	C5H4N4O2	151.0250	108.0191;79.9560	[68]
3	Isocitric acid	1.23	C6H8O7	191.0189	111.0074;87.0074;85.0281;105.6934	[67]
4	Citric acid	1.58	C6H8O7	191.0189	111.0076;87.0075;85.0282;105.6936	[67]
5	Phenylalanine*	1.74	C9H11NO2	164.0709	147.0441;163.8393;72.0078	[68]
6	Vicenin 2*	3.43	C27H30O15	593.1517	383.0775;353.0662;297.0759;473.1091	[69]
7	Schaftoside	3.72	C26H28O14	563.1406	545.1303;503.1201;473.1092;443.0985	[70]
8	Luteolin-7-O-glucoside	3.96	C21H20O11	447.0936	327.0510;285.0411;241.0490;227.0351	[71]
9	Isoschaftoside	3.85	C26H28O14	563.1404	545.1303;503.1203;473.1092;443.0984	[70]
10	Taurocholic acid*	5.58	C26H45NO7S	514.2836	496.2757;430.2970;371.2572;79.9561	[16]
11	Daidzein	5.85	C15H10O4	253.0506	224.0469;209.0598;135.0076;91.0172	[72]
12	Taurohyocholic acid	5.95	C26H45NO7S	514.2836	496.2739;430.2990;371.2571;79.9561	[16]
13	Genistein	6.19	C15H10O5	269.0456	224.0481;133.0280;159.0442;91.0171	[73]
14	Apigenin*	6.68	C15H10O5	269.0455	251.6099;201.0549;225.01281;207.6033	[68]
15	Glycocholic acid*	7.47	C26H43NO6	464.3015	446.2900;418.2985;371.2628;74.0234	[16]
16	Taurochenodeoxycholic acid	7.48	C26H45NO6S	498.2889	480.2757;432.3115;414.3013;386.3071	[16]
17	Taurohyodeoxycholic acid*	7.93	C26H45NO6S	498.2891	480.2815;355.2646;414.2994;79.9560	[16]
18	Glycohyocholic acid	8.34	C26H43NO6	464.3014	446.2914;418.2968;400.2853;74.0234	[16]
19	Glycochenodeoxycholic acid	8.66	C26H43NO5	448.3067	430.2945;404.3174;386.3069;74.0234	[16]
20	Hyocholic acid	10.30	C24H40O5	407.2801	389.2701;371.2588;369.2448;341.2486	[16]
21	Cholic acid*	11.15	C24H40O5	407.2802	389.2704;343.2644;289.2174;271.2072	[16]
22	Glycohyodeoxycholic acid*	11.81	C26H43NO5	448.3067	430.2939;404.3176;386.3069;74.0234	[16]
23	Chenodeoxycholic acid	12.52	C24H40O4	391.2854	373.2771;355.2661;329.2875;273.2211	[16]
24	Hyodeoxycholic acid	16.50	C24H40O4	391.2853	373.2768;355.2661;329.2876;273.2211	[16]
25	Deoxycholic acid	17.01	C24H40O4	391.2854	343.2641;373.2769;355.2644;329.2851	[16]
26	Hexadecanoic acid	21.49	C16H32O3	271.1916	225.2223;155.1429;100.9330	[68]

*Compounds confirmed by Reference standards.

3.6. Potential targets of ACB against epilepsy

ACB compound related targets were retrieved from the Swiss Target Prediction database, a total of 398 targets for 26 constituents were found. Then GeneCards were employed to screen the authorized epileptic therapeutic targets, a total of 6023 therapeutic targets were gathered after redundancy was removed. Finally, 221 possible shared targets relevant to both epilepsy and ACB were screened using a Venn diagram for additional investigation (Figure 6(A)). In the PPI network, there were 221 nodes and 2012 edges (Figure 6(B)). Once the data were imported into Cytoscape, 50 core objectives were screened according to degree.

Figure 6. Prediction of antiepileptic effects of ACB by network pharmacology. (A) The Venn diagram of targets of ACB and epilepsy. (B) Protein–protein interaction network (PPI). (C) Biological process of GO functions enrichment. (D) Molecular function of GO functions enrichment. (E) Cellular component of GO functions enrichment. (F) KEGG pathways analysis. (G) Drug-components-targets-disease-pathways network diagram.

3.7. GO and KEGG enrichment analysis

GO enrichment analysis could be evaluated for functional prediction. The results of the GO investigation shows that 1673 GO terms were enriched, including 1514 biological processes, 56 cellular Components and 103 molecular functions. From the three categories shown in the bubble charts in Figure 6(C–E), the top 20 highly enriched terms were extracted. In the biological process (BP) category, positive regulation of cell adhesion (GO:0045785), regulation of cell–cell adhesion (GO:0022407), peptidyl-tyrosine modification (GO:0018212) and the biological processes of peptidyl-tyrosine phosphorylation (GO:0018108) were significantly enriched. In the cellular components (CC) category include membrane microdomain (GO:0098857), the membrane raft (GO:0045121), glutamatergic synapse (GO:0098978), etc. In the molecular functions (MF) category include protein serine/threonine kinase activity (GO:0004674), phosphatase binding (GO:0019902) and growth factor receptor binding (GO:0070851), etc.

ACB’s relationship to epilepsy was further clarified through using KEGG enrichment analysis. Figure 6(F) displayed the top 20 signaling pathways as a bubble chart, analysis of the signaling pathways showed that the PI3K-Akt pathway was substantially enriched.

3.8. Drug-components-targets-disease-pathways network analysis

Figure 6(G) depicted the network of drug, components, targets, disease and pathways. The network diagram which had 98 nodes and 482 edges constructed showed that there were intricate relationships between various chemicals and targets. According to the result of topological parameters of drug component disease target pathway network analysis (Table 3), glycohyodeoxycholic acid, glycocholic acid, taurohyodeoxycholic acid and luteolin-7-O-glucoside may be the key ingredients of ACB in the treatment of epilepsy as filtered by degree value. The most essential signaling pathway in treating epilepsy is PI3K/Akt pathway.

Table 3. Topological parameters of drug component disease target pathway network analysis.

No	Description	Degree	Neighbourhood Connectivity	Type
1	Epilepsy	50	9.120	disease
2	Arisaema Cum Bile	26	7.115	drug
3	PIK3CA	24	13.875	gene
4	PIK3R1	22	14.091	gene
5	AKT1	22	14.091	gene
6	MAPK3	22	14.091	gene
7	GRB2	20	14.300	gene
8	PRKCA	19	14.895	gene
9	MAPK14	19	14.105	gene
10	hsa04151:PI3K-Akt signaling pathway	19	14.053	pathway
11	MTOR	18	14.222	gene
12	EGFR	18	14.278	gene
13	Glycohyodeoxycholic acid	16	14.000	mol
14	GSK3B	15	15.533	gene
15	Glycocholic acid	15	10.800	mol
16	hsa01521:EGFR tyrosine kinase inhibitor resistance	15	16.067	pathway
17	hsa05205:Proteoglycans in cancer	15	16.267	pathway
18	hsa05417:Lipid and atherosclerosis	15	13.933	pathway
19	hsa05163:Human cytomegalovirus infection	15	17.000	pathway
20	MDM2	13	14.538	gene
21	Taurohyodeoxycholic acid	13	14.462	mol
22	Luteolin-7-O-glucoside	13	12.154	mol

3.9. Molecular docking analysis

According to the result of drug-components-targets-disease-pathways network analysis, The four potential active compounds and target proteins in PI3K/Akt signal pathway (PI3K and AKT1) were reserved for molecular docking. As the results showed in Figure 7 and Table 4, the target proteins had lower docking score with compounds, which the docking scores of glycohyodeoxycholic acid bound to 4EJN (AKT1), glycocholic acid and taurohyodeoxycholic acid bound to 4ANW (PI3K) were lower than −8 Kcal/mol.

Figure 7. The binding modes between targets and active compounds in ACB. The number of A-H is the same as the number in Table 4.

Table 4. Docking simulation for active molecular and targets of ACB.

No.	Molecular name	Targets	PDB ID	Residue involved in H bonding	H-bond length(Å)	Docking score(kcal/mol)
A	Glycocholic acid	PI3K	4ANW	CYS357;LEU423;LYS425	1.7;1.7;2.1	−8.07
B	Glycohyodeoxycholic acid	PI3K	4ANW	GLN391;PHE497;LYS501;LEU567	2.8;1.8;2.0;1.9,2.2	−6.94
C	Luteolin-7-O-glucoside	PI3K	4ANW	GLN396;ARG397;LYS419;ASP422;GLY599	2.6,2.4;2.0;1.9;2.1,2.2;2.5	−4.58
D	Taurohyodeoxycholic acid	PI3K	4ANW	CYS357;LYS425	2.6,2.1;2.2	−8.00
E	Glycocholic acid	AKT1	4EJN	SER396;ALA50	2.0;2.4,2.3,1.8	−6.87
F	Glycohyodeoxycholic acid	AKT1	4EJN	GLN203;LYS268;THR82;ASP292	1.9;2.0;2.0;1.9,2.1	−8.37
G	Luteolin-7-O-glucoside	AKT1	4EJN	ASN53;SER205;THR211;THR291;ASP292	2.3;2.7;2.0;2.0;1.7,1.9	−7.69
H	Taurohyodeoxycholic acid	AKT1	4EJN	THR82;THR211;LYS268;GLU267;GLN203	2.8;2.5,2.0;2.6;1.7;2.0	−7.83

3.10. ACB protects PC12 cells from H₂O₂ damage

Firstly, we determined how ACB affected PC12 cells viability at varying concentrations (Figure 8(A)). The PC12 cells survival rate in the 6 and 12 h intervention experiments were above 90% when the ACB concentration was less than 20 μg/mL. However, when 20 μg/mL ACB was intervened for 24 h, the viability of PC12 cells was about 85%. Finally, we chose 2.5, 5 and 10 μg/mL of ACB as the working concentration of subsequent experiments. At the same time, we set the pretreatment time of ACB as 24 h. Referring to the experiment of Zhang et al [36], we chose 200 μM H₂O₂ intervention for 4 h as the modelling method of PC12 cells. Pretreatment with 5 and 10 μg/mL ACB could significantly increase the cells survival rate (p < 0.05, p < 0.01, respectively). Figure 8(B,C) indicates that ACB might prevent PC12 cells from H₂O₂-induced injury.

Figure 8. Protective effects of ACB on H₂O₂ induced PC12 cells. (A) Effects of various concentrations and treatment times of ACB on the viability of PC12 cells. (B) Effects of ACB (2.5, 5 and 10 μg/mL) on the viability of H₂O₂ induced PC12 cells. (C) The represented cell morphology of H₂O₂ induced PC12 cells with ACB (×100). *p < 0.05 and **p < 0.01 vs. model group.

MMP is a sensitive indicator of cellular injury, and its reduction was viewed as an early diagnostic event of apoptosis [37,38]. JC-1 is a fluorescent probe that has been found widely used as an appropriate tool for detecting MMP ΔΨm, with a high or low MMP, JC-1 emits red and green fluorescence respectively [39–41]. The effect of ACB on the MMP of PC12 cells induced by H₂O₂ was evaluated by observing the fluorescence changes of JC-1 by confocal laser microscopy and flow cytometer. The results are shown in Figure 9(A), after 4 h of exposure to 200 μM H₂O₂, the green fluorescence of PC12 cells enhanced significantly. Fortunately, pretreatment ACB with different concentrations could alleviate this situation. In addition, Figure 9(B) showed that the levels of JC-1 monomers in the cells were significantly reduced after 5 and 10 μg/mL of ACB pretreatment compared with the model group, indicating that ACB pretreatment could significantly improve cell injury caused by H₂O₂ (p < 0.05, p < 0.01, respectively).

Figure 9. Effects of ACB on the ΔΨm in PC12 cells. (A) MMP assay by confocal laser microscopy of PC12 cells. (B) MMP assay by flow cytometer. *p < 0.05 and **p < 0.01 vs. model group.

Flow cytometer was employed to measure cell apoptosis and thereby assess protective effect of ACB on PC12 cells. The results are shown in Figure 10(A), the apoptosis rate of cells increased significantly after H₂O₂ intervention. ACB pretreatment can significantly reduce H₂O₂ induced apoptosis compared with the model group. Figure 10(B) demonstrates that H₂O₂ intervention may considerably upregulate the expression of the apoptotic proteins C-caspase 3 and Bax in cells and downregulate the expression of the anti-apoptotic protein Bcl-2. However, ACB pretreatment (5 and 10 μg/mL) could reverse the expression of these proteins in PC12 cells (p < 0.05, p < 0.01, respectively). According to the aforementioned findings, ACB therapy could increase cell MMP, decrease the H₂O₂ injury to PC12 cells, and have a protective effect by suppressing apoptosis.

Figure 10. Antiapoptotic effects of ACB on H₂O₂ induced PC12 cells. (A) Apoptotic assay by flow cytometer. (B) Effects of ACB on proteins expression of Bcl-2, Bax and C-caspase-3. *p < 0.05 and **p < 0.01 vs. model group.

3.11. ACB protects PC12 cells from H₂O₂ injury by activating PI3K/Akt/Nrf2 pathway

The antiepileptic activity of ACB in vitro by establishing H₂O₂-stimulated PC12 cells model, and found that ACB can significantly protect PC12 cells from H₂O₂ injury via reducing cell apoptosis. Based on our previous work in network pharmacology, the PI3K/Akt pathway may contribute to ACB’s therapeutic effects on epilepsy. It is often recognized that the PI3K/Akt pathway is closely connected to the process of apoptosis [42–45], plenty of research has confirmed that stimulating the PI3K/Akt pathway is an effective way to treat epilepsy [46–49]. The results were shown in Figure 11(A), H₂O₂ may inhibit the PI3K/Akt pathway in PC12 cells as it dramatically lowered the phosphorylation levels of PI3K and Akt in those cells following treatment. Interestingly, the phosphorylation levels of PI3K and Akt in PC12 cells pretreated with ACB were significantly increased, suggesting that ACB can activate the PI3K/Akt pathway.

Figure 11. (A) Effects of ACB on protein expressions of PI3K, p-PI3K, Akt and p-Akt in H₂O₂ induced PC12 cells. (B) Effects of ACB on ROS levels in H₂O₂-induced PC12 cells. (C) Effects of ACB on protein expressions of Nrf2(C) and Nrf2(N) in H₂O₂ induced PC12 cells. *p < 0.05 and **p < 0.01 vs. model group.

Cell damage caused by H₂O₂ was closely related to oxidative stress. Figure 11(B) showed the results that H₂O₂ intervention could significantly increase the level of ROS in the cells, while ACB (5 and 10 μg/mL) pretreatment could significantly inhibit the production of ROS (p < 0.05). Nrf2 has attracted much attention as one of the important antioxidant targets in cells. Phosphorylated Akt may facilitate Nrf2 entrance into the nucleus, thereby increasing the antioxidant capacity of cells. Therefore, the expression of Nrf2 was detected in cells, as we expected, Nrf2 nuclear translocation was decreased following H₂O₂ stimulation compared with normal group. (Figure 11(C)). However, H₂O₂-stimulated Nrf2 nuclear translocation may be increased by ACB pretreatment.

PI3K inhibitor LY294002 was adopted in the experiment to clarify whether ACB protects PC12 cells from oxidative injury by activating the PI3K/Akt pathway. Figure 12(A–C) showed that LY294002 could counteract the protective effect of ACB on PC12 cells and increase cell apoptosis. Further, LY294002 also could inhibit the phosphorylation of Akt in ACB-intervened PC12 cells by immunofluorescence experiments (Figure 12(D)). In summary, our current results prove that ACB pretreatment can significantly protect PC12 cells from oxidative injury, and the mechanism was related to the activation of PI3K/Akt/Nrf2 pathway.

Figure 12. Effects of ACB with PI3K inhibitor LY294002 on the cell viability (A) and cell apoptosis (B) of H₂O₂ induced PC12 cells. (C) Apoptotic assay by flow cytometer of PC12 cells. (D) Immunofluorescence images of ACB with PI3K inhibitor LY294002 on protein expression of Akt/p-Akt in H₂O₂ induced PC12 cells. *p < 0.05 and **p < 0.01 vs. model group.

4. Discussion

More and more evidence confirmed the effectiveness of traditional Chinese medicine in preventing and treating various diseases [50–52]. This study aims to provide a theoretical basis for the antiepileptic activity of ACB by confirming the active components, predicting the molecular mechanism and verifying the experiment.

Epilepsy is a complex neurological disease, it is also typically accompanied by cognitive impairments. Mounting evidence suggests that epilepsy will damage nerve cells especially in CA1 and CA3 regions of hippocampus, and may cause imbalance of neurotransmitters in brain [30–32]. Maladaptive changes in neurotransmitter levels make nerves excitable abnormally resulting in more severe seizure activity [53,54]. According to certain studies, PTZ kindling was widely used to create a chronic epilepsy model [55–58]. This provides an important reference for our research, the results indicated that ACB has potential antiepileptic activity in epileptic animals ACB possessed significant antiepileptic effects mainly acted on improving the pathological changes in hippocampal CA1 and CA3 regions and regulating neurotransmitters of PTZ-induced epileptic rats, but the detailed molecular mechanism was yet unclear.

The chemical constituents of ACB were studied by HPLC-MS/MS. Twenty-six chemical constituents were identified by database and literature, among which glycohyodeoxycholic acid, glycocholic acid, taurohyodeoxycholic acid and luteolin-7-O-glucoside may be the main active constituents of ACB in anti-epilepsy. PI3K/Akt signal pathway enriched by network pharmacology may be the main signal pathway of ACB antiepileptic activities. The PI3K/Akt pathway has been identified as having an antioxidant function in both central and peripheral neurones. Akt, a kinase that phosphorylates serine and threonine residues, plays a crucial role in mediating signaling started by PI3K. The activation of PI3K leads to the phosphorylation of Akt (p-Akt), which serves as a reliable marker for PI3K activity [59]. It exerts its neuroprotective effects by modulating the expression levels of caspase-3, Bcl-2, Bax and other relevant factors [60,61]. There is a well accepted understanding in the scientific community that oxidative stress has the potential to trigger neuronal death. It has been shown that Caspase family members play a crucial role in the process of cell apoptosis, Caspase-3 is well recognized as a biomarker for the detection of cellular apoptosis [62]. Nevertheless, Bax has the capability to impede the activity of antiapoptotic Bcl-2 proteins [63]. Furthermore, Bcl-2 has been acknowledged for its advantageous properties such as promoting cell survival, reducing oxidative stress, inhibiting programmed cell death and safeguarding cellular integrity [63]. In addition, PI3K/Akt pathway is recognized as a cellular defence mechanism against H₂O₂-induced cell damage [32,36], Nrf2 is a downstream protein of PI3K/Akt pathway, under oxidative stress, it is shown that Nrf2 dissociates from Keap 1 and then binds to the antioxidant response element (ARE), facilitating its translocation into the nucleus. This event triggers the transcription process, leading to the production of a diverse range of antioxidative enzymes inside the cytoplasm.

Consequently, we concentrated even more on the probable mechanisms of ACB antiepileptic effects. Neuronal cell pathology and physiology are studied extensively using the PC12 cell line as a model [36,64]. The highly harmful hydroxyl radicals produced when H₂O₂ crosses the cell membrane and attacks biological molecules lead to cellular destruction, apoptosis and necrosis [65,66]. Therefore, in our study, H₂O₂-induce oxidative injury of PC12 cells was adopted to exploit molecular mechanism. Pretreatment with ACB could correct the reduction in cell viability induced by H₂O₂, and ACB might upregulate the phosphorylation of PI3K and Akt and levels of Bcl-2 while downregulate levels of C-caspase-3 and Bax. In all, ACB plays an antiepileptic role by regulating PI3K/Akt/Nrf2 signal pathway to prevent neuronal apoptosis.

5. Conclusion

In summary, the experiment confirmed that ACB has antiepileptic activity. The main active ingredient may be glycohyodeoxycholic acid, glycocholic acid, taurohyodeoxycholic acid and luteolin-7-O-glucoside, and its activity may be attributed to oxidative stress damage reduction, suppression of neuronal apoptosis and activation of the PI3K/Akt/Nrf2 signal pathway. These results provide a foundation for further biological investigation of ACB as an epilepsy treatment drug.

Disclosure statement

No potential conflict of interest was reported by the author(s).