Analysis of the Molecular Mechanism of Punicalagin in the Treatment of Alzheimer's Disease by Computer-Aided Drug Research Technology.

The objective of this work is to explore the effect and potential mechanism of Punicalagin (Pun) in managing Alzheimer's disease (AD) based on computer-aided drug technology. The following methods were used: the intersection genes of Pun and AD were retrieved from the database and subjected to PPI analysis, GO, and KEGG enrichment analyses. Preliminary verification was performed by molecular docking, molecular dynamics (MD) simulation, and combined free energy calculation. The motor coordination and balance ability, anxiety degree, spatial learning, and memory ability of mice were measured by a rotating rod fatigue instrument, elevated cross maze, and Y maze, respectively. The amyloid β protein (Aβ) in the hippocampus was examined by immunohistochemistry, and the phosphorylation of serine at position 404 of the tau protein (Tau-pS404) was examined by western blot in the mouse brain. The PPI network of Pun showed that the intersection genes were closely related and enriched in muscle cell proliferation and the response to lipopolysaccharide. Results of molecular docking, MD simulations, and MM-GBSA demonstrated that Pun was closely bound to the target protein. Pun could improve the cognitive function of AD mice, as well as reduce Aβ1-42 deposition and Tau phosphorylation in the brain (P < 0.05, P < 0.01). It can be concluded that Pun holds great promise in improving the cognitive function of AD mice. Mechanistically, Pun potentially acts on ALB, AKT1, SRC, EGFR, CASP3, and IGF-1 targets and mediates proteoglycan, lipid, and atherosclerosis in cancer, so as to reduce the accumulation of neurotoxic proteins in the brain.

Introduction

Alzheimer’s disease (AD) is a neurodegenerative disease that often manifests as cognitive impairment. Current evidence shows that the exact pathophysiology of AD is elusive and urgently requires a more safe and effective therapy. Studies on the health characteristics of natural products and their potential as a functional food or preventive medicine are mature.[1] In the Compendium of Materia Medica·Fruit·Punicalagin, an ancient Chinese medicine book in the mid-16th century states that the pomegranate skin was mainly used to manage “red and white dysentery, long dysentery and diarrhea, swelling and malice, and stomach sores”. Studies have revealed numerous pharmacological functions of the pomegranate peel extract Punicalagin (Pun), including anti-inflammatory, antioxidant, antibacterial, antitumor effects, liver protection, and prevention of cardiovascular and cerebrovascular diseases.[2]

There have been many research findings about Pun. The mechanism underlying the efficacy of Pun in managing neurodegenerative diseases is that the NF-κB inhibitor potentially improves neuroinflammation and oxidative stress in the brain, causing a neuronal loss.[3] Also, Pun can improve the lipopolysaccharide (LPS) induced-memory damage and prevent LPS-induced inflammatory protein expression.[4] As a natural inhibitor of advanced glycation end products (AGEs), Pun has been demonstrated to play a role in the prevention and treatment of degenerative diseases via antiglycosylation.[5] Moreover, Pun inhibits the activation of T cells and microglia in vitro using AD transgenic mouse models.[6] These findings provide concrete evidence on the role of Pun in alleviating nerve injury and degenerative diseases. However, the possible molecular mechanism of Pun in improving the symptoms of AD is not clear.

Combined with a computer-aided drug research technology, this study fully discussed the potential molecular mechanism of the main active components and main targets of Pun in the treatment of AD, provided a theoretical basis for drug molecular design and transformation, and provided a reference for further targeted drug experiments and clinical applications.

Results

PPI Network Construction and Core Target Analysis

From the 290 drug targets and 1194 disease targets identified via Venny 2.1, 69 core targets were obtained (Figure ). The protein interaction network showed a close relationship between the nodes in the PPI network. Also, Pun and ad intersection genes were highly correlated. The top 10 core targets include serum albumin (ALB), serine/threonine-protein kinase 1 (AKT1), non-receptor tyrosine kinase c-SRC (SRC), epidermal growth factor receptor (EGFR), caspase-3 (CASP3), insulin-like growth factor 1 (IGF-1), Matrix metalloproteinase (MMP) 9, estrogen ESR1, gelatinase A (MMP-2), and catalase (CAT) (Figure ).

Figure 1

Protein interaction network. (A) Wenn diagram of drug disease common targets. (B) Network diagram of disease target component interaction. (C,D) Protein interaction network diagram.

Figure 2

PPI enrichment analysis. (A) Ranking of core targets based on PPI topology analysis. (B–D) PPI cluster analysis.

GO and KEGG Enrichment Analysis

The GO results demonstrated that 1421 biological processes were enriched in the intersection genes (Figure ). The anti-AD mechanism of Pun mainly involves biological processes, including muscle cell proliferation, response to LPS, response to bacterial derived molecules, phosphatidylinositol 3-kinase signal transduction and regulation, regulation of smooth muscle cell proliferation, phosphatidylinositol mediated signal transduction, and cell response to chemical stress, among others. Following the R language runs, 120 KEGG channels were obtained. The top 20 pathways were revealed (Figure ). The main signal pathways that enriched more targets included proteoglycan, lipid and atherosclerosis, prostate cancer, endocrine resistance, diabetic cardiomyopathy, relaxin signaling pathway, FoxO signaling pathway, epithelial cell signal transduction in Helicobacter pylori infection, AGE-RAGE signaling pathway, and MAPK signaling pathway in complications of H. pylori infection (Figures and 6).

Figure 3

GO enrichment analysis.

Figure 4

KEGG enrichment analysis.

Figure 5

2D (A) and 3D (B) structure of Punicalagin.

Figure 6

Molecular docking results.

Molecular Docking Results

Pun demonstrated a good binding effect with the target protein with the binding energy of less than −6 kcal/mol (Table ). A lower energy requirement for binding implies that Pun can readily bind to the target protein.

Table 1

Molecular Docking of the Core Target Proteins with Pun

target	PDB	delta G (kcal/mol)	delta Gvdw (kcal/mol)	full fitness (kcal/mol)	energy (kcal/mol)
ALB	6YG9	–10.14	–51.07	–2118.24	–8.17
AKT1	4GV1	–16.11	–52.68	–798.82	–9.09
SRC	4U5J	–16.19	–63.97	–4718.23	–9.49
EGFR	4HJO	–14.42	–46.32	–4305.31	–8.52
CASP3	1NMS	–13.40	–58.28	–1437.84	–8.26
IGF1	2DSR	–14.75	–77.03	–1540.95	–9.68

Molecular Dynamics Simulation Results

Root-mean-square deviation (RMSD) reflected the stability of the proteins and the small molecules. The larger the RMSD was, the more unstable the proteins were. The RMSD of CASP3 and Pun was small and reached equilibrium at about 2 ns (Figure Aa). The average RMSD of EGFR and Pun was less than 2.2 Å, which reached equilibrium at about 10 ns (Figure Ba). The average RMSD of AKT1 and Pun was less than 2.0 Å, which reached equilibrium at about 40 ns (Figure Ca). In addition, the RMSD of small molecules of EGFR and Pun ligand fluctuated slightly in the first 50 ns. The RMSD of small molecules of the AKT1 and Pun ligand fluctuated slightly in the first 70 ns and then decreased. It might be that small molecules, after constantly colliding with the active site in the protein pocket, found the right conformation so as to reach the balance.

Figure 7

Molecular dynamics simulation. RMSD plot during molecular dynamics simulations of (A) (a) CASP3, (B) (a) EGFR, (C) (a) AKT1, (D) (a) IGF1, (E) (a) ALB, (F) (a) SRC with Punicalagin. The interaction residues of Punicalagin with (A) (b) CASP3, (B) (b) EGFR, (C) (b) AKT1, (D) (b) IGF1, (E) (b) ALB, and (F) (b) SRC.

The average RMSD of IGF1 and Pun was less than 3.5 Å. The equilibrium was reached at about 30 ns. When the small molecules of IGF1 and Pun ligand were in the first 15–80 ns, RMSD tended to be in equilibrium, but the RMSD was smaller after that (Figure Da). The average RMSD of ALB and Pun was less than 3.8 Å (Figure Ea). The average RMSD of SRC and Pun was less than 2.9 Å (Figure Fa). It could be seen from Figure b that there were strong hydrogen bonds and hydrophobic interactions between Pun and the active site which made an important contribution to the stability of small molecules in the protein pocket.

Calculation Results of MM-GBSA

The calculation of static molecular docking and MM-GBSA will provide a good view of binding posture and binding free energy, so as to ensure that the complexes of compounds and targets have enough energy for biochemical reaction. The binding free energy calculated by MM-GBSA supports the molecular docking results (Table ).

Table 2

MM/GBSA of the Best Candidate Compounds and Protein Targets for Each Target

target	PDB	van der Waals force/(kJ/mol)	electrostatic potential energy/(kJ/mol)	polar solvation energy/(kJ/mol)	surface solvation/(kJ/mol)	binding free energy/(kJ/mol)
AKT1	4GV1	–239.23 ± 2103	–100.03 ± 15.22	218.45 ± 17.36	–60.24 ± 1.09	–188.16 ± 17.45
ALB	6YG9	–226.37 ± 15.82	–95.67 ± 19.30	227.39 ± 17.20	–56.90 ± 0.95	–169.12 ± 19.32
CASP3	1NMS	–236.12 ± 19.02	–92.29 ± 15.19	201.40 ± 21.01	–62.37 ± 1.43	–170.98 ± 14.90
IGF1	2DSR	–241.19 ± 16.23	–99.94 ± 19.90	219.06 ± 19.28	–57.86 ± 1.93	–200.37 ± 19.13
SRC	4U5J	–226.87 ± 18.20	–90.57 ± 15.28	233.02 ± 16.17	–63.05 ± 1.67	–196.44 ± 20.37
EGFR	4HJO	–241.80 ± 14.39	–110.57 ± 17.20	224.81 ± 18.72	–57.87 ± 1.05	–176.36 ± 15.29

Pun can Improve the Motor Coordination and Balance Ability of Mice

The rotating rod fatigue instrument is often used to detect brain injury, motor coordination, and fatigue of animals. Compared with the normal, the rod rotation time of the model group was significantly reduced (P < 0.01); compared with the model group, the rod rotation time of the three dose groups of Pun was significantly prolonged (P < 0.01), and the extension rates of rod rotation time were 39.50, 69.12, and 70.86%, respectively (Table ).

Table 3

Effects of Pun on the RRT and ER-RRT in mice (χ̅ ± SD, n = 12)a

groups	RRT (s)	ER-RRT (%)
normal	20.31 ± 3.47
model	10.33 ± 3.42##
Pun-L	14.41 ± 4.32**	39.50
Pun-M	17.47 ± 2.31**	69.12
Pun-H	17.65 ± 2.66**	70.86
DHT	15.26 ± 6.03**	47.73

Note: #P < 0.05 ##P < 0.01 versus normal; *P < 0.05 **P < 0.01 versus model, the same as below.

Pun can Improve the Anxiety Psychology of Mice

The elevated cross maze test is often used to evaluate the anxiety response of rodents.

Mice have a dark preference, so they tend to move in the closed arm, but they will move in the open arm due to their own exploratory behavior. Therefore, the residence time of mice in the open arm residence time (OART) was inversely proportional to the degree of anxiety, while the residence time of mice in the closed arm residence time (CART) was directly proportional to the degree of anxiety. Compared with the normal, the number of mice entering the open arm and the residence time of mice in the open arm in the model group were reduced (P < 0.05, P < 0.01); compared with the model group, the residence time of mice in the open arm increased significantly in the drug group and decreased significantly in the closed arm in the drug group (P < 0.01) (Table ). In addition, Pun had little effect on the number of mice entering the closed arm. The abovementioned results show that Pun can improve the exploratory motor behavior and autonomic motor activity of mice, and Pun has a good ability to recover anxiety. In addition, the effect of Pun low dose group (Pun-L) on the mouse Tau protein was small, and there was no significant difference. This may be because the drug dose is so small that the efficacy is low.

Table 4

Effects of Pun on the OAET, OART, CAET, CART in mice (χ̅ ± SD, n = 12)a

groups	OAET (times)	OART (s)	CAET (times)	CART (s)
normal	6.13 ± 1.33	203.51 ± 25.41	3.24 ± 0.46	89.17 ± 1.15
model	3.22 ± 1.27#	150.26 ± 10.45##	3.13 ± 0.77	113.04 ± 4.78##
Pun-L	3.14 ± 2.74	161.28 ± 18.84*	3.34 ± 1.17	88.17 ± 7.41**
Pun-M	4.68 ± 1.24*	186.74 ± 10.61**	3.06 ± 1.08	74.63 ± 4.45**
Pun-H	4.17 ± 3.02*	190.43 ± 12.41**	4.28 ± 1.35*	95.46 ± 5.61**
DHT	3.25 ± 2.77	189.71 ± 6.64**	3.46 ± 0.94	93.28 ± 11.24**

Note: #P < 0.05, P < 0.01 versus normal; *P < 0.05, **P < 0.01 versus model.

Pun Potentially Improves the Working Memory of Mice

The Y maze is mainly used to test the discrimination learning ability, working memory ability, curiosity, and exploration behavior of animals. The spatial memory impairment of mice shortened the time and distance of exploring the new arm. Compared with the normal, the spontaneous alternating response rate of mice in the Y maze decreased significantly (P < 0.05, P < 0.01), indicating that their working memory performance was impaired; compared with the model group, the total times of entering the arm, the times of entering the new arm, and the spontaneous alternating response rate of Pun in each dose group increased (P < 0.05, P < 0.01), indicating that Pun can improve this memory ability (Table ).

Table 5

Effects of Pun on the Total NAE, Number of NOA and SAR in mice (χ̅ ± SD, n = 12)a

groups	NAE (times)	NOA (times)	SAR (%)
normal	36.21 ± 2.40	18.13 ± 1.21	72 ± 0.45
model	28.21 ± 1.32#	9.86 ± 1.79##	35 ± 1.21##
Pun-L	29.34 ± 1.76	13.12 ± 0.44**	42 ± 1.09**
Pun-M	30.33 ± 0.80*	13.27 ± 1.02**	46 ± 0.87**
Pun-H	31.17 ± 1.48*	15.67 ± 1.50**	48 ± 1.66**
DHT	34.63 ± 1.73**	16.45 ± 0.28**	54 ± 1.34**

Note: #P < 0.05, P < 0.01 versus normal; *P < 0.05, **P < 0.01 versus model.

Effects of Pun Treatment on Aβ1-42 and Tau-pS404 in the Mouse Brain

AD is generally characterized by extracellular cells located in the cerebral cortex and hippocampus β Amyloid (Amyloid β protein, Aβ) deposition and intracellular hyperphosphorylation of Tau (a microtubule associated protein) to form neurofibrillary tangles and cholinergic neurotransmission defects. In the immunohistochemical picture of this experiment, Aβ1-42 is extracellular deposition protein, and an extracellular tan color is positive deposition (marked by red arrow). Aβ1-42 in the model group was significantly higher than that in the normal group (P < 0.01), while Aβ1-42 decreased significantly after Pun treatment (P < 0.05, P < 0.01) (Figure Aa). Compared with the normal group, the expression of Tau-pS404 in the model group was significantly up-regulated (P < 0.01). Compared with the model group, the expression of Tau-pS404 in low, medium, and high dose experimental groups was significantly down-regulated (P < 0.01, P < 0.05) (Figure Bb).

Figure 8

(A) Aβ1-42 deposition in hippocampus was detected by immunohistochemistry (×200), the positive deposition of Aβ1-42 is tan color and indicated by red arrow. (a) immunohistochemistry was quantified by ImageJ software. (B) Tau-pS404 expression in the brain of mice; (b) western blots were quantified by ImageJ software. (C) neurotoxic proteins (Aβ1-42 and Tau-pS404) and nine behavioral experimental indexes correlation network; (D) neurotoxic proteins (Aβ1-42 and Tau-pS404) and nine behavioral experimental indexes clustering correlation heatmap. The nine behavioral experiment indexes are derived from three behavioral experiments, including rotating rod fatigue instrument, elevated cross maze, and Y maze. Rotating rod fatigue instrument: the RRT and ER-RRT in mice; elevated cross maze: the OAET, OART, CAET, CART in mice; and Y maze experiment: total NAE, number of NOA and SAR in mice.

The Pun treatment decreased the phosphorylation level of Tau protein ser404 site in the brain of model mice.

The correlation analysis between neurotoxic protein index and behavioral changes showed that neurotoxic protein Aβ1-42 and Tau-pS404 were negatively correlated with RRT, open arm entry times (OAET), OART, number of arm entries (NAE), NOA, and SAR, positively correlated with CART, and had no statistical significance with closed arm entry times (CAET) and extension rate of rod rotation time (ER-RRT). The results of correlation analysis show that Aβ1-42 and Tau-pS404 can significantly destroy various behavioral abilities and reduce the cognitive level of mice. After Pun treatment, the neurotoxic protein of mice is reduced, and their various behavioral abilities are restored (Figure C,D).

Discussion

Studies have shown that Pun has a beneficial effect on neurodegenerative diseases including AD.[3−6] Pun are the main phenolic compounds in the pomegranate peel. The corresponding targets of drug molecules and the specific mechanism of action of active molecules need to be further studied. Molecular dynamics simulation can build a protein solvation model close to the human physiological state, study the relationship between dynamic experimental data established by computer and static molecular conformation, and provide dynamic structural change data that cannot be detected by experiments. It is an ideal technology for the research and development of natural products.

In this study, the main effective targets against AD were selected from the targets of Pun by a network pharmacological method, and six core genes ALB, AKT1, SRC, EGFR, CASP3, and IGF-1 were selected through drug disease common targets. The conformation and chemical state of the target bound by drugs were identified by ligand protein molecular docking. According to molecular docking results, Pun contains multiple donors and receptors. It can form strong hydrogen bond interactions with the active groups of amino acids of ALB, AKT1, SRC, EGFR, CASP3, and IGF-1 target proteins. Pun also forms hydrophobic and conjugate interactions with the hydrophobic residues in the pocket, which promotes the binding of small molecules to protein sites. These results provide support to the conclusion of network pharmacological screening. Pun matches well with protein targets and therefore, is a promising compound in terms of therapeutic activity.

In order to further investigate the interaction between small molecules and proteins, we performed molecular dynamics analysis of the complex of proteins and small molecules for 100 ns using molecular dynamics. Molecular dynamics simulation shows that the complexes formed by the binding of these six targets with Pun have good stability, and the close binding between molecules and protein active sites is mainly related to the interaction between the hydrogen bond and hydrophobicity. Finally, the free energy calculation of a ligand protein complex shows that the active component has a strong affinity with the target.

The core targets of PPI, including ALB, AKT1, SRC, EGFR, CASP3, and IGF-1, are suggested to be the key components of Pun in AD treatment. Evidence shows the potential interaction of ALB in the blood with Aβ. As such, to improve AD and aging brain lesions, Pun may influence Aβ in the circulatory system by activating ALB.[7] A cohort study demonstrated that brain AKT phosphorylation (pT308AKT1/total AKT1) is a key node of growth factors signal transduction, for example, insulin, which is associated with AD neuropathology and cognitive dysfunction.[8] More evidence has shown a role for the tyrosine phosphotransferase Fyn of SRC family nonreceptor kinases in the pathophysiological changes of AD.[9] The soluble polymer binds with high affinity to the cellular prion protein on the surface of neuronal cells, triggering a pathological cascade on the non-receptor tyrosine kinase Fyn. In this view, the therapeutic effect of Pun in AD may be related to Fyn inhibition of this signal cascade. Studies indicate that the neuroprotective function of EGFR is influenced by presenilin 1 (PSEN1) expression and that its down-regulation may be related to AD.[10] The present work suggests a neuroprotective role for Pun through the regulation of EGFR and its upstream and downstream pathways. In the AD mouse model, the pro-apoptotic factor Caspase-3 signal could tune the activation and neurotoxicity of microglia. Emerging data indicate the potential role of Pun in Caspase-3 signal regulation, exerting a neuroprotective effect on AD.[11,12] IGF signaling (IIS) regulates stress resistance and aging, which are determinants of life expectancy.[13] Reduced IIS can improve pressure resistance and prolong service life. Analysis shows that Pun may delay the toxicity of aging-related proteins in mice by regulating IIS and reducing the IGF-1 signal. These events improve AD.

GO enrichment analysis revealed that the anti-AD mechanism of Pun is mainly related to biological processes, including myocyte proliferation and LPS. Low-density lipoprotein receptor-associated protein 6 (LRP6) can inhibit the atypical Wnt signal, promoting the proliferation of arterial smooth muscle cells and vascular calcification.[14] This abolishes the cognitive and cerebrovascular defects in the AD mouse model. For AD patients presented with cognitive and cerebrovascular defects, Angiotensin IV (ang IV)[15] exerts various effects in the fight against Aβ: (i) it can alleviate the cerebral vasodilation mediated by endothelial and smooth muscle cells; (ii) it can normalize the level of Ang IV receptor (AT4R) in the hippocampus; (iii) it can increase the cell proliferation and dendritic dendrite branching in the subgranular area of the hippocampus: (iv) it can reduce oxidative stress. These data suggest a role for Pun in improving cerebrovascular defects by blocking the proliferation of vascular endothelium and/or smooth muscle cells. By doing so, it achieves the therapeutic effect against AD. Studies have demonstrated that LPS can induce neuroinflammation and pathological changes related to AD and other diseases,[16] which, in most cases, is employed as a model of AD. The entry of LPS into the intestinal wall and blood via the intestinal wall epithelium activates the inflammatory signal pathway. This consequently releases inflammatory factors into the blood and induces inflammation-mediated Aβ in the brain abnormal increase.[17] Therefore, the therapeutic effect of Pun against AD may be related to the inflammatory pathway.

KEGG enrichment analysis revealed that the main active component of Pun exerts therapeutic effects against AD by mediating proteoglycan, lipid, and atherosclerosis in cancer. Proteoglycans in cancer, including hyaluronic acid, heparan sulfate proteoglycan, chondroitin sulfate/dermatan sulfate proteoglycan, form dense lattice perineural networks (PNNs) to regulate neural activity and plasticity following a central nervous system injury.[18] Pun may act on PNNs, enhance neuronal activity and synaptic plasticity, and promote recovery from central nervous degenerative diseases. Atherosclerosis is a lipid-driven arterial immune inflammatory disease and a major risk factor for cognitive impairment.[19] Emerging data indicate that Pun may improve cognitive behavior via cardiovascular-related mechanisms.

The disorder of Aβ and Tau production and metabolism is one of the main hypotheses of AD pathogenesis. The present experiment selected Aβ1-42 and Tau-pS404 as the detection index. In addition, animal experiments used the Y maze experiment, rotating rod fatigue instrument experiment, and elevated cross maze experiment to evaluate the improvement of Pun on AD from three aspects: behavioral ability, mental state, and cognitive function.

The results showed that compared with normal, the neurotoxic protein accumulation in the brain of the model group was larger, the desire to explore in the three behavioral experiments was reduced, and the ability of work, learning, and memory was impaired. After Pun treatment, the expression of neurotoxic protein in mouse brain decreased, and the behavioral ability and cognitive x increased. Pun can improve the curiosity and exploration behavior of mice, improve the working memory ability of mice, and alleviate the pathological changes of neurodegenerative diseases in mice. Finally, we have used correlation analysis to confirm this result.

Conclusions

In conclusion, through network pharmacology, this work analyzes and predicts the active components and important targets of Pun in AD treatment. Further, the efficacy of Pun is explored through molecular docking and animal experiments. We preliminarily prove that the main active component of Pun may be through multi-target and multi-channel reduction of Aβ and Tau protein, so as to reduce nerve injury and death in the brain, and finally improve the cognitive injury of AD mice. This study proved that Pun is a bioactive component against neurotoxic protein deposition. This finding lays a foundation for further exploring the molecular mechanism of Pun improving cognitive decline.

Materials and Methods

Network Pharmacology

Drug Disease Common Target

The structure of Pun was retrieved from the PubChem database (Table ) and imported into the PharmMapper database. The target with a norm fit prediction score greater than 0 was considered the drug target. To obtain the drug target, we corrected and unified the target name in the UniProt database. The OMIM, TTD, and GeneCards databases were employed to retrieve and remove duplicates with the keyword “AD”. The disease target component network was constructed and analyzed by the Cytoscape 3.7.2 software.

Table 6

Database and Analysis Platform

name of database and software	web address
PubChem	https://pubchem.ncbi.nlm.nih.gov/
PharmMapper	http://www.lilab-ecust.cn/pharmmapper/
OMIM	https://omim.org/
GeneCards	https://www.genecards.org/
TTD	http://db.idrblab.net/ttd/
STRING	https://string-db.org/
Venny 2.1	https://bioinfogp.cnb.csic.es/tools/venny/
Uniprot	https://www.uniprot.org/
Cytoscape3.7.2	http://www.cytoscape.org
RCSB PDB	https://www.rcsb.org/

PPI Network Construction and Core Target Analysis

The target obtained in 2.1.1 was searched in the STRING database based on the protein species “Homo sapiens” and a threshold value of 0.4. After that, the protein interaction association map was constructed. Genes whose gene degree value exceeded the average score were identified through CytoScan 3.7.2 topological analysis. A bar graph was generated in R 3.6.0.

Pathway and Functional Enrichment Analysis

The bioconductor was used to analyze the gene biological process (GO) and genome encyclopedia (KEGG) of key genes with p-value < 0.05 and Q-value < 0.05 in R software. First, the common target was run in the R language. Next, GO analysis selected the biological process and the top 20 KEGG pathways.

Molecular Docking

The Schrodinger software was employed to construct the ligand molecular database of molecular docking. The crystal structure was downloaded from the RCSB PDB. The protein structure was imported into Maestro 11.9 platform, and the protein was prepared using Schrodinger. The receptor was pretreated, optimized, and minimized (OPLS3e force field was applied for constraint minimization).

Punicalagin has two main isomers, alpha and beta.[20] These two isomers can be quickly converted to each other under most conditions,[21] so we selected one of the alpha conformations as the initial conformation for the molecular docking. In addition, we have performed multiple docking comparisons of the two configurations before the molecular docking and found that the two configurations are basically the same in terms of docking scoring. Therefore, all subsequent calculations use the Punicalagin molecules in alpha configuration.

Molecular Dynamics Simulation

Desmond version 2020 was used for MD simulation of proteins and compounds. Here, OPLS3e was selected as the molecular force field for MD simulation, and the TIP3 water model was used to solvate the system. The energy minimization of the whole system was achieved by using the OPLS3e force field (all-atomic force field). Berendsen coupling algorithm was used to create a coupling between temperature and pressure parameters. In the later preparation of the system, 100 ns was run at a time step of 1.2 fs, and the track was recorded every 10 ps, recording a total of 10,00 frames. RMSD of backbone atoms was calculated and graphical analysis was performed to comprehend the nature of the interaction between proteins and ligands. The RMSF (root-mean-square fluctuation) of each residue was calculated to realize the major conformational changes of the residue between the initial state and the dynamic state.

Calculation of MM-GBSA

The basic principle of method MM-GBSA (Molecular Mechanics/Poisson Boltzmann (Generalized Born) Surface Area) is to calculate the difference between the binding free energies of two solvated molecules in the binding and free states, or to compare the free energies of different solvated conformations of the same molecule.

Among them, includes bond, bond angle, and dihedral angle energy; is non bond van der Waals energy; is non bond electrostatic energy; and TΔS0 is entropy contribution, which can be obtained by normal mode analysis.

Animal Experiment

Animals and Materials

Experimental Mice

Forty-eight APP/PS1 transgenic mice, SCXK (Su) 2019-0001, the Model Animal Center of Nanjing University. Twelve C57BL/6 mice, Changsha Tian Qin Biology SCXK (Xiang) 2019-0014. All mice were SPF grade, 2 months old, weighing 25 ± 3 g, half male and half female. The mice were kept under conventional feeding, constant temperature (23 ± 2 °C), constant humidity range (40–60%), alternating light and dark conditions, free drinking, and feeding. No treatment was administered to mice one week before the experiment.

Materials

Punicalagin (P0349, Pure Biological); Donepezil hydrochloride tablets (DHT) (h20010723, Chongqing Zhi En pharmaceutical industry); Primary antibody: anti-Aβ1-42 Rabbit anti-mice (BA0762, sigma, USA); Anti-Tau (ser S404) Rabbit anti-mice (EPR2605, Abcam, UK); Horseradish enzyme-labeled Goat anti-rabbit IgG (H + L) (ZB-2301, ZSGB-BIO, USA).

Grouping and Sampling

48 APP/PS1 mice were randomly divided into model groups and low, medium, and high dose experimental groups, with 12 mice in each group; another 12 C57BL/6 mice were taken as the normal control group. The dosage was set as described previously:[22] the normal control group and model group were given sterile water by gavage; the high, medium and low dose groups (Pun-H, Pun-M, and Pun-L) were given Pun 50, 25, and 12.5 mg·kg–1, respectively; the positive control group was given DHT 3 mg·kg–1, all the mice were treated by gavage once a day for 45 days. After completing the behavioral experiment, the eyeballs were removed to draw blood. The mice were killed by dislocating the cervical vertebrae. The whole-brain of mice was divided into two parts from the coronal plane, one part was fixed with 4% formaldehyde, and the other part was freshly preserved and protein was extracted. All animal experiment procedures strictly followed the protocol approved by the ethics committee of Wannan Medical College (YJS-2020-10-006).

Rotating Rod Fatigue Instrument

The mice were put into the mouse rotating rod fatigue instrument for experiment. The rotating speed was 30 rpm and each time was limited to 1 min. The time from the start of rotation of the rotating rod to leaving the rotating rod was taken as the rotating rod time (RRT) of the mouse, and test 0 (clockwise rotation mode) was used in the formal experiment. Mice in each group were trained twice before each experiment, with an interval of 10 min, using TRAIN (training mode). Before each experiment, it needs to be cleaned to remove the traces left by the previous mouse.

Elevated Cross Maze

The maze consists of two open arms and two closed arms. This experiment examines the anxiety state of animals by using the contradiction between the fear of the open environment and the exploratory nature of the new environment. Animals with high anxiety levels tended to spend more time in the closed arm than animals with low anxiety levels. In the experiment, each mouse was put into a maze from the center lattice facing the closed arm, and its activity was recorded for 5 min. Before each experiment, it needs to be cleaned to remove the traces left by the previous mouse.

Y Maze Experiment

Before each test,[10] the mouse was allowed a 5 min free exploration learning time. The blocking arm stuck during free exploration was opened at a 3 h interval. During the formal test, the mouse was placed in the center of the Y-shaped equipment with its back to the baffle, recording the activities of the mouse in each arm within 5 min. The equipment was cleaned at the end of each mouse test. Spontaneous alternation rate (%) = number of spontaneous alternations/(total number of entering the arm – 2) × 100%, entering three arms continuously is a spontaneous alternating reaction.

Immunohistochemistry

Hippocampal was detected through immunohistochemistry analysis of the expression of Aβ1-42. Put the slice in H2O2 and rinse and add 5% BSA, drop a rabbit anti-mouse primary antibody Aβ1-42 (1:150), 4 °C overnight, add biotinylated sheep anti-rabbit IgG (1:150) and rabbit anti-goat IgG (1:100) after rinsing, elute at room temperature for 60 min, and DAB color. Immunohistochemical analysis was performed using ImageJ software. By measuring the integrated option density (IOD) value and area value of each image, we calculate mean density = IOD/area, which reflects the unit area concentration of the target protein. Finally, take the average value of the mean density of the five random regions of each sample, that is, the value of this sample.

Western Blot

Fresh brain tissue was taken, the protein was extracted and quantified, and then SDS-PAGE electrophoresis was carried out. Then, the membrane was transferred and blocked, the primary antibody and secondary antibody were incubated, chemiluminescence and development were observed, the gray scale of the target protein and internal reference protein by ImageJ software was scanned, and the protein was analyzed semi-quantitatively to obtain the relative expression of the protein.

Statistical Processing

SPSS 18.0 software was employed for all statistical analyses. Measurement data were expressed in χ̅ ± s. One way ANOVA was used. LSD-t test was applied for comparison between groups.