PAN Yu, WANG Liang, WANG Peng, YANG Ming-hui

Department of Traditional Chinese Medicine, Chinese PLA General Hospital, Beijing 100048, China

Keywords:

ABSTRACT Objective: To analyze the dosing pattern and mechanism of herbal compound patents for the treatment of Parkinson's disease using data mining and network pharmacology methods,and to provide ideas for the clinical use of Parkinson's disease and new drug development.Methods: The Chinese herbal medicine compound prescriptions for Parkinson's disease built up to May 31, 2022 by searching the official website of patent publication notice.An Excel table was built, and after term normalization of the included compound prescriptions,Excel and IBM SPSS Modeler 18.0 were used for data mining such as frequency statistics.We also applied network pharmacology methods to study HF drugs, using TCMSP and TCM database@Taiwan to obtain drug components and using TCMSP platform and Swiss ADME screening, collecting targets through UniProt and Swiss Target Prediction platform; obtaining PD disease from databases such as GeneCards The targets were obtained from GeneCards and other databases, drug-disease target intersections were obtained, and "drug-disease-target"networks were created using Cytoscape 3.8 software, and GO and KEGG enrichment analyses were performed using the Metascapep platform.Results: A total of 113 patented Chinese medicine recipes were included, involving 394 drugs, among which Tianma, Angelica and Bai Shao were the most commonly used drugs.A total of 442 drug targets, 4 884 disease targets and 324 drug-disease common targets were obtained.Conclusion: The study found that the medicinal properties of the patented compound Chinese medicine for Parkinson's disease are mainly warm and cold, the taste of the medicine is mainly sweet and bitter, the normalizing meridian is mainly liver, heart and spleen meridians, and the treatment method is mainly to strengthen the liver and kidney, nourish the qi and blood, and extinguish the wind and dredge the luo.The core ingredients include 4-ethoxymethylphenyl-4'-hydroxyl, palmitic acid, lignan,etc.; the main action targets are PTGS2, PTGS1, SCN5A, etc.; involved in PI3K-Akt, cAMP and other signaling pathways.The development of relevant compound can be based on clinical symptoms, appropriate tailoring, and flexible use of such drugs, in order to obtain the best therapeutic effect.

1.Introduction

Parkinson's disease (PD) is a neurodegenerative disease characterized by motor symptoms such as bradykinesia, resting tremor, and myotonia, as well as non-motor symptoms such as insomnia, depression, and constipation.By 2015, the cumulative number of Parkinson's disease cases reached 6.2 million worldwide,and with the advent of an aging society, the incidence of PD is expected to increase, with 12.9 million people expected to be affected by PD in 2040[1], posing a huge socioeconomic and medical challenge.Existing conventional treatments are based on dopaminergic or anticholinergic medications, such as levodopa,dopamine agonists, and benztropine, to compensate for dopamine levels or mimic their effects[2].However, as the main drug for PD,levodopa not only has a limited "honeymoon period", but also has side effects such as dyskinesia with long-term use.Therefore, it is necessary to find a better treatment.

Chinese medicine has rich clinical practice in the treatment of Parkinson's disease, emphasizing the holistic concept and evidencebased treatment, with multi-system and multi-target characteristics,and is safe, reliable and economical, and has been accepted by more and more patients and their families in recent years.The national patent database is a distillation and summary of long-term clinical and scientific research experience, which is an important wind vane for the development of drugs.Therefore, the author analyzes the dosing patterns of herbal compound patents for Parkinson's disease by mining the national patent database, with a view to providing reference for the clinical use and new drug development of Parkinson's disease.

2.Data and Methods

2.1 Sources of information and search strategy

"State Intellectual Property Office-China Patent Publication Notice"webpage (http://epub.sipo.gov.cn).

In the advanced search options, select "Invention Grant" and"Invention Publication", and select "Parkinson's and Chinese Medicine" in the "Text-Name" item, and "Parkinson's and Herbal Medicine" in the "Text-Abstract/Brief Description" item."Parkinson's and herbal medicine", and "Parkinson's and Chinese medicine" in the "Text-Abstract/Brief description" item."Parkinson's and herbal medicine".The time period is from the establishment of the library to May 31, 2022.

2.2 Inclusion and exclusion criteria

Inclusion criteria: Chinese herbal compound patents for the treatment of Parkinson's disease in the national patent database;patents whose patent item names do not specify the treatment of Parkinson's disease, but whose patent applications for inventions state that they can treat Parkinson's disease; all types of internal dosage forms, including but not limited to tonics, pills and granules.Exclusion criteria: single-formula preparations with less than two Chinese herbal medicines; specific prescriptions are not specified;compound prescriptions explicitly dedicated to the treatment of Parkinson's syndrome caused by other diseases; external compresses,ointments and oils; patents of food, medicinal wine and health products are excluded.

2.3 Data specification and database creation

2.3.1 Establishment of the database

An Excel sheet was established to enter the patented Chinese medicine compound that met the inclusion criteria but did not meet the exclusion criteria.

2.3.2 Specification of the database

The drug names are standardized according to the 2020 edition of the Pharmacopoeia of the People's Republic of China [3] and the 1999 edition of the Chinese Materia Medica [4], such as "Da Yun"and "Cun Yun" are standardized as "Rou Congrong", "Gui Ban" is standardized as "Gui Jia", "Yan Hu" is standardized as "Yan Husuo","Quan Chong " is standardized as "Quan Xie", etc.Incorporate the same composition of the patent, the dosage form is different to enter one time.

2.4 Statistical analysis

The data were entered into Excel 2019, and the included drugs and their properties were analyzed by frequency; IBM SPSS Modeler 18.0 was used to analyze the association rules and cluster the highfrequency drugs.

2.5 Collection of high-frequency drug components and targets

The herbs with frequency greater than or equal to 19 in 1.4 were selected as the target of analysis, and the active ingredients of the drugs were queried with the help of the pharmacological database and analysis platform of Chinese medicine system (https://tcmsp-e.com) and the database of Chinese medicine in Taiwan (http://tcm.cmu.edu.tw/zh-tw/), and the active ingredients of the drugs were collected on the Swiss ADME platform (http://www.swissadme.ch/)and TCMSP with GI absorptionwei as high, Druglikeness greater than or equal to two as Yes and OB 30% and DL 0.18 as criteria for ingredient screening, and in Swiss Target Prediction platform(http: //www.swisstargetprediction.ch), and the Global Protein Resource Database (https://www.uniprot.org) for target prediction.

2.6 Disease target collection

Using "Parkinson disease" as the keyword, we searched the Gene Cards (https//www.genecards.org), Drug Bank (https://go.drugbank.com), OMIM database (https://omim.org) were searched for Parkinson's disease target genes, and targets with correlation values greater than the median were taken.The targets collected from each database were combined and de-duplicated.

2.7 Construction of drug-disease target network graph

The obtained drug and disease targets were made Venn diagrams to find the intersecting genes, and then the drug-disease targets were entered into Cytoscape 3.8 for drug-disease target network mapping.2.8 Drug-disease target GO and KEGG enrichment analysis Drugdisease intersection targets were imported into the metascape platform (http://metascape.org) to enable GO and KEGG pathway enrichment analysis.

3.Results

3.1 Patent Publication

After being screened by the nadir criteria, 113 patents were included, and the publication time and number of patent items of herbal compound for the treatment of Parkinson's disease from 2007 to May 2022 are shown in Figure 1.It can be seen that the research and development of herbal patent compound for the treatment of Parkinson's disease started late and has gradually received attention in recent years.

Fig 1 Patent publication of Chinese herbal compound for treating Parkinson's disease from 2007 to 2022

3.2 High frequency drugs

The 113 patents involved a total of 394 flavors of drugs, with the inclusion of single patent flavors ranging from 2 to 36 flavors.The total frequency of drugs was 1 391 times, and 22 flavors were used more than 10%, among which the frequency of the top 10 flavors were: Tianma (44 times, 38.94%), Angelica (40 times, 35.40%),Bai Shao (37 times, 32.74%), Huang Qi (37 times, 32.74%), Dan Shen (34 times, 30.09%), Rhynchophylla (32 times, 38.32%), Radix Rehmanniae (32 times, 28.32%), Chuanxiong (30 times, 26.55%),Dilong (23 times, 20.35%), and Quanscorpion (19 times, 16.81%).The table of high frequency drugs is shown in Table 1.

Tab 1 113 high-frequency drugs of patented Chinese herbal compound for treating Parkinson's disease( 10%)

3.3 The four gases and five tastes of drugs and their attribution

The four gases, five tastes, and meridians of the 394 drugs included in the study were analyzed.(22.37%); the cumulative frequency of attribution to meridians was 3 448, with liver (868 times, 25.17%),heart (477 times, 13.83%), spleen (466 times, 13.52%) and kidney(499 times, 13.02%) being the most frequent.See Table 2-4 and Figure 2.

Tab 2 Frequency of four natures of drugs in 113 patents of Chinese herbal compound for treating Parkinson's disease

Tab 3 Frequency of the five flavors of drugs in 113 Chinese herbal compound patents for Parkinson's disease

Tab 4 Frequency of six meridians of drugs in 113 Chinese herbal compound for treating Parkinson's disease

3.4 Analysis of association rules

Using IBM SPSS Modeler 18.0 with association rule analysis and complex network analysis, the network analysis diagram is shown in Figure 2.By mining drug combinations through Apriori in the software, setting a minimum support of 10%, a confidence level of 70%, and a maximum antecedent term of 5, 2 groups of drug pairs, 19 groups of 3-flavored drug combinations, and 6 groups of 4-flavored drug combinations were obtained.The details are shown in Table 5-7.

Fig 2 The complex network diagram of 113 Chinese herbal compound patents for treating Parkinson's disease

Tab 5 Two drug combinations of 113 Chinese herbal compound patents for treating Parkinson's disease

3.5 Cluster analysis

The first 33 drugs (frequency greater than 10 times) were analyzed in clusters using IBM SPSS Statistics 25.0.When the Pearson correlation was at 20, the drugs could be classified into 8 categories.Ⅰ: Momordica charantia, Safflower, Radix Angelicae Sinensis,Astragalus membranaceus, Rhizoma gigantea, Rhizoma Chuanxiong.Ⅱ: Scorpion, Centipede, Danshen, Dioscorea, Stipula, Tianma,Rhynchophylla, White peony, Shu Di Huang.Ⅲ : Pueraria lobata,roasted licorice.Ⅳ: Poria cocos, hyoscyamus, biliary nancellus,peony skin, radix et Rhizoma ginseng, scutellaria.Ⅴ: Cistanche,Acorus calamus.Ⅵ: Cornus officinalis, Schisandra chinensis,Tortoise shell, Glycyrrhiza glabra.Ⅶ : Lycium barbarum.Ⅷ : Chai Hu, Yuan Zhi.The genealogical chart is shown in Figure 3.

Tab 6 Three drug combinations of 113 Chinese herbal compound patents for Parkinson's disease

Tab 7 Four drug combinations of 113 Chinese herbal compound patents for Parkinson's disease

Fig 3 Cluster analysis dendrogram of high-frequency drugs of Chinese medicine for Parkinson's disease

3.6 High frequency drug active ingredient and target screening

The number of active ingredients was obtained by TCMSP and TCM database@Taiwan query of high frequency drug components,after screening by swiss ADME and TCMSP: 16 of asparagus, 2 of angelica, 41 of white peony, 13 of astragalus, 59 of salvia, 33 of Rhynchophylla, 5 of rehmannia, 6 of chuanxiong, 6 of cistanches,and the ground dragon and whole scorpion were animal drugs.Not included in the TCMSP database.The targets of the active ingredients were predicted and screened using the uniprot database and Swiss Target Prediction, and the number of targets for each HF drug was obtained as follows: 241 for TCMSP, 49 for Angelica sinensis, 41 for Paeonia lactiflora, 80 for Astragalus membranaceus,132 for Salvia miltiorrhiza, 202 for Crocus sativus, 35 for Rhizomarehmanniae, 29 for Rhizoma Chuanxiong, and 172 for Cistanches sp.The above targets were combined and the duplicates were deleted to a total of 442 targets.

Tab 8 Network node characteristic parameters of main active ingredients of high-frequency drugs

3.7 PD disease target collection and Venn diagramming

The target genes of Parkinson's disease were searched through Gene Cards, Drug Bank, and OMIM databases, and the targets collected from the three databases were combined and de-duplicated,resulting in 4 884 targets.

The 442 targets obtained for HF drugs were plotted against the 4 884 targets for PD in a Venn diagram, yielding a total of 324 drugdisease targets.See Figure 4.

Fig 4 Venny diagram of drug - disease intersection targets

3.8 Drug-disease-target network modeling

The drug names, ingredients, and common targets with diseases were imported into Cytoscape 3.8 to obtain the "drug-disease-target"network diagram (Figure 5).There are 466 nodes and 2 352 edges.The diamond-shaped nodes in the middle rectangle represent the drug-disease shared target genes, the hexagonal nodes around the ring represent the drug components, the circular nodes in the middle of the ring represent the drugs, and the hexagonal nodes in the bottom horizontal row represent the drug shared components, and the edges represent the component-target association.It can be seen that high-frequency drug therapy for Parkinson's disease has multilevel, multi-target and multi-pathway characteristics.

The network analyzer was used to analyze the network diagram,and the top five were arranged in descending order by Degree value to determine the core components and core targets.The main active ingredients of the national patent compound HF drug for Parkinson's disease are 4-Ethoxymethylphenyl-4'-hydroxybenzylether,hexadecanoic acid, luteolin luteolin), 7-O-methylisomucronulatol,4,4-Dihydroxydiphenyl methane (4,4'-Dihydroxydiphenyl methane).The main targets of action are PTGS2, PTGS1, SCN5A,HSP90AB1, CHRM1.The characteristic parameters of the main active ingredients and targets of the national patent compound HF drugs for the treatment of Parkinson's disease are shown in Table 8 and Table 9.

Fig 5 Drug - disease - target network

Tab 9 Network node characteristic parameters of main targets of high-frequency drugs

3.9 GO and KEGG enrichment analysis

The drug-disease shared targets were imported into Metascape, GO,KEGG enrichment analysis was performed, and the top 20 with the smallest LogP values were taken for visualization (Figure 6).It can be seen that the main pathways of HF drugs in the national patent compound for Parkinson's disease contain cancer pathway, PI3KAkt, neuroactive ligand-receptor interaction, cellular senescence,cAMP, MAPK, dopaminergic synapses, Parkinson's disease, cGMPPKG, etc.High frequency drug therapy for Parkinson's disease contains mainly inflammatory responses, positive regulation of cell death, response to oxygen levels, response to extracellular stimuli,synaptic signaling, involving membrane rafts, dendrites, receptor complexes, transferase complexes, transfer of phosphorus-containing groups and other cellular components, neurotransmitter receptor activity, neurotransmitter receptor activity involved in the regulation of postsynaptic membrane potential, NF-κB binding and other molecular functions.

Fig 6 Enrichment analysis of GO and KEGG of high-frequency drugs in national patent compound for Parkinson's disease

4.Discussion

Parkinson's disease belongs to the category of "tremors" in Chinese medicine.The study and analysis of the patent of Chinese medicine compound prescriptions for the treatment of Parkinson's disease,including 113 compound prescriptions, a total of 394 Chinese medicines, through frequency analysis, we found that the high frequency drugs are Tianma, Angelica, Bai Shao, Astragalus, etc.Tianma enters the liver meridian, which contains active ingredients such as asparagine, p-hydroxybenzyl alcohol and palisade A.[5],pharmacological studies have shown that Tianma extract treats neurological diseases through mammalian target of rapamycin(mTOR) and other pathways, with the effects of scavenging reactive oxygen species, attenuating abnormal protein deposition, promoting neurological recovery and protecting neurons[6].Angelica sinensis is a blood tonic and blood activator.Pharmacological studies have shown that angelica can inhibit cellular senescence, increase antioxidant active enzymes, etc., and has the effect of regulating neurotransmitters and protecting nerve cells[7,8].Bai Shao pacifies the liver and suppresses Yang.Pharmacological studies have shown that Bai Shao can reduce apoptosis, protect neuronal cells, regulate function and antidepressant, etc[9].Astragalus polysaccharide can achieve anti-aging and memory improvement by enhancing the body's antioxidant capacity and protecting nerve cells[10].Therefore,from both traditional medicine and modern medical research, HF drugs can better treat Parkinson's disease.

The association rule analysis yielded common drug combinations for the treatment of Parkinson's disease, involving Tianma, Angelica,Rhynchophylla, Di Long, Bai Shao, Chuanxiong, Shu Di Huang,Dan Shen, Tao Ren and Huang Qi, which were more uniform with high frequency drugs.Among them, the drug combinations with the highest confidence and support were Tianma-Chostrongylus-Astragalus and Tianma-Chostrongylus, respectively.Tianma and Crotalus are the main constituent drugs of Tianma-Chostrongylus Drink, which is a representative formula for treating hyperactivity of liver yang and upward disturbance of liver wind, and can treat tremor, insomnia and vertigo.Modern research has shown that the Tianma-Choxidium pair can reduce oxidative stress in nerve cells and inhibit apoptosis, thus achieving neuroprotective effects[11].Radix Angelicae Sinensis, Radix Paeoniae Alba and Rhizoma Ligustici Chuanxiong are the constituent drugs of Angelicae Sinensis and Paeoniae Herb San.Studies have shown that Angelicae Sinensis and Paeoniae Herb San can inhibit neuroinflammatory response and reduce neuronal apoptosis, thus exerting neuroprotective effects[12,13].

The 32 HF drugs were classified into 8 categories by cluster analysis, and the drug classification was based on benefiting Qi and nourishing blood, calming the liver and quenching the wind,dispelling wind and phlegm, nourishing Yin and tonifying the kidney and activating blood, nourishing the heart and tranquilizing the mind, and relieving liver and depression.This corresponds to the pathological mechanisms of Parkinson's disease such as deficiency of Qi and Blood, internal movement of liver and wind,interlocking of phlegm and stagnation, kidney deficiency and blood stasis, and Yin deficiency and wind movement[14-16].In addition, it also provides reference for the treatment of non-motor symptoms in PD patients, such as insomnia due to deficiency of heart and blood, spleen and stomach disharmony, and evil qi disturbing the heart, constipation due to qi deficiency, yin deficiency, and blood deficiency, depression due to liver qi stagnation and heart blood failure, cognitive impairment due to phlegm stasis and marrow failure, and urinary system disorders due to kidney qi deficiency and kidney yang deficiency[16-18], all of which can seek treatment ideas from the HF drug classification.The main active ingredients of HF drugs are 4-ethoxymethylphenyl-4'-hydroxyl, palmitic acid,lignocaine and so on.The main targets of action are PTGS2, PTGS1,SCN5A, etc.

Based on the Chinese herbal compound formulas for Parkinson's disease in the national patent database, this study has developed new ideas for the development of Chinese herbal compound formulas for Parkinson's disease and the clinical tailoring of the drugs.However, because most of the patents did not specify the applicable evidence type of drugs, this study has some limitations in the dialectical treatment.The subsequent study will combine the empirical prescriptions used in the treatment of Parkinson's disease by classical Chinese medicine and famous veteran Chinese medicine practitioners to conduct a more comprehensive study and analysis.

Author's contributions:

Pan Yu proposed the research idea, designed the research protocol,conducted data retrieval, screening, entry and completed database creation, conducted data mining and web pharmacology analysis,and completed the paper writing.Wang Liang performed data entry and proofreading for the second times, provided guidance on network pharmacology methods, and initial revision of the paper.Wang Peng proofread the data, provided guidance on data mining methods, and participated in the revision of the thesis.Yang Minghui supervise the thesis writing, provide guidance on the thesis writing, and correct the thesis revision.

All authors declare no conflict of interest.