Bioinformatics and network pharmacology identify the therapeutic role and potential targets of diosgenin in Alzheimer disease and COVID-19

ZHANG Hua-xiong, ZHANG Ming-hui, LI Hong-yan✉

1. Department of Neurology, Xinjiang Uygur Autonomous Region People's Hospital, Urumqi 830000, China

2. Xinqikang Pharmaceutical Co, LTD, Urumqi 830000, China

Keywords:

ABSTRACT Objective: This study aims to investigate the potential targets of diosgenin for the treatment of Alzheimer's disease (AD) and Coronavirus Disease 2019 (COVID-19) through the utilization of bioinformatics, network pharmacology, and molecular docking techniques.Methods:Differential expression genes (DEGs) shared by AD and COVID-19 were enriched by bioinformatics.Additionally, regulatory networks were analyzed to identify key genes in the Transcription Factor (TF) of both diseases.The networks were visualized using Cytoscape.Utilizing the DGIdb database, an investigation was conducted to identify potential drugs capable of treating both Alzheimer's disease (AD) and COVID-19.Subsequently, a Venn diagram analysis was performed using the drugs associated with AD and COVID-19 in the CTD database, leading to the identification of diosgenin as a promising candidate for the treatment of both AD and COVID-19.SEA, SuperPred, Swiss Target Prediction and TCMSP were used to predict the target of diosgenin in the treatment of AD and COVID-19, and the target of diosgenin in the treatment of AD and COVID-19 was determined by Wayne diagram intersection analysis with the differentially expressed genes of AD and COVID-19.Their Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG)were analyzed jointly.Genomes The Protein Protein Interaction (PPI) network of these drug targets was constructed, and core targets with the highest correlation were screened out.The binding of diosgenin to these core targets was analyzed by molecular docking.Results:Through enrichment and cluster analysis, it was found that the biological processes, pathways and diseases enriched by DEGs in AD and COVID-19 were all related to inflammation and immune regulation.These common DEGs and Trust databases were used to construct AD and COVID-19 TFs regulatory networks.Diosgenin was predicted as a potential drug for the treatment of AD and COVID-19 by network pharmacology, and 36 targets of diosgenin for the treatment of AD and 27 targets for COVID-19 were revealed.The six core targets with the highest correlation were selected for molecular docking with diosgenin using CytohHubba to calculate the scores.Conclusions: This study firstly revealed that the common TFs regulatory network of AD and COVID-19, and predicted and verified diosgenin as a potential drug for the treatment of AD and COVID-19.The binding of diosgenin to the core pharmacological targets for the treatment of AD and COVID-19 was determined by molecular docking, which provides a theoretical basis for developing a new approach to clinical treatment of AD and COVID-19.

1.Introduction

In recent years, a lot of studies have found that AD is one of the most common central nervous system comorbidities of COVID-19[1], and there is a clear correlation between the two diseases[2,3].In addition to respiratory symptoms such as fever, sore throat,cough, and difficulty breathing, COVID-19 can also cause various neurological symptoms, including increase the risk of developing AD.A retrospective cohort study of 6,245,282 elderly people (age65 years) found that compared to the control group, COVID-19 patients had a significantly increased risk of developing new AD within 360 days after their first diagnosis of COVID-19 (HR: 1.69,95% CI: 1.53-1.72)[4].Although no direct evidence of COVID-19 inducing AD has been found so far, studies have shown that COVID-19 may be involved in systemic inflammation, overactivation of the renin angiotensin system, innate immune system activation,oxidative stress, direct viral infection, and direct cell lysis β Mechanisms such as cell damage , all of those can induce AD.At the same time, AD patients have a variety of pathological changes,lifestyle changes, drug interactions and other factors, such as the overexpression of viral receptor angiotensin converting enzyme 2, proinflammatory molecules, the incidence rate and mortality of COVID-19 patients with AD increase[2].This indicates that there may be a common molecular regulatory network between AD and COVID, and studying the shared molecular regulatory network of these two diseases may provide new clinical pathways for the treatment of AD and COVID.

Dioscin is a kind of natural steroid saponin extracted from dioscoreaceae plants such as reed grass, Dioscorea nipponica, Mount Huangshan medicine, which has anti-inflammatory, antioxidant, lipid lowering, weight loss, liver protection, anti-tumor and other activities[6].This biologically active compound can act on various cell targets and affect various signaling pathways in the human body.It has good efficacy in the treatment of diseases such as cancer, gastrointestinal diseases, cardiovascular and cerebrovascular diseases, and organ toxicity[6].At present, research has confirmed that diosgenin can target infected or diseased tissues without damaging normal cells,so it can be considered as an efficient and low cytotoxic drug.This drug also meets the current demand for non-toxic, non-invasive,and effective treatment alternatives in society to a certain extent.At the same time, studies have confirmed that diosgenin can exert good therapeutic effects on AD by regulating mechanisms such as oxidative stress, inflammatory response, and cellular autophagy[7].However, the potential therapeutic effects and related mechanisms of diosgenin on COVID-19 still require extensive research and exploration.In this study, we revealed the biological functions, TF regulatory networks, and core targets of AD and COVID-19 through bioinformatics analysis.In addition, we also identified diosgenin as a potential drug for treating AD/COVID-19 using network pharmacology methods, and established its pharmacological targets for treating AD/COVID-19, revealing its potential therapeutic effect by regulating inflammation and immune pathways.

2.Research method

2.1 Differential expression genes of AD and COVID-19

Data download from GEO database (https://www.ncbi.nlm.nih.gov/geo/) 100 tissues (56 AD hippocampal tissues and 44 control hippocampal tissues) in GSE122063 (GPL570 platform), 21 tissues(16 COVID-19 serum samples and 5 control serum samples) in GSE211979 (GPL16699 platform).Data standardization involves first using the normalizeBetweenArrays function of the limma package to standardize the dataset, and then grouping it according to AD Control and COVID-19 Control for difference analysis using the limma package (set threshold adj.P<0.05&| logFC |>0.5).

2.2 Differential intersection genes between AD and COVID

For the differentially expressed genes in the AD dataset GSE122063 and the COVID dataset GSE211979, the up-regulated and downregulated genes were intersected, respectively.Draw Venn diagrams using R language’s Venn package to display intersecting genes.

2.3 Enrichment analysis of differential genes

Import the upregulation and downregulation differential intersection genes of AD/COVID into the DAVID website separately (https://david.ncifcrf.gov/) Perform GO and KEGG enrichment analysis,and in the enrichment analysis results obtained from DAVID, set a threshold PValue of less than 0.05, with a count of genes enriched in each pathway greater than 3.Filter the enrichment results of upregulated and downregulated genes, respectively.Draw a bar chart using the ggplot2 package of R language for enrichment results that meet the threshold.

2.4 Transcription Factor Target Network

From Trust Database ( https://www.grnpedia.org/trrust/ ) Download TF and its downstream target gene data.We used the Wayne diagram to select TFs and at least downstream target genes that coexist in both AD and COVID-19 DEGs as key genes in both AD and COVID-19.We used Cytoscape software to construct a TF target network of these TFs and their target genes.

2.5 Using DGIbd and CTD databases for drug prediction

To prioritize the AD/COVID drug list, we use the DGIbd database( https://dgidb.genome.wustl.edu/) Conduct key TF and target gene screening to predict therapeutic drugs for AD/COVID.Then the predicted drugs will be linked to the CTD database (http://ctdbase.org/) Download relevant drugs for treating AD/COVID for intersection analysis to obtain potential candidate drugs for treating AD/COVID.And use the ggallivian package of R language to draw a Sangi diagram, visualizing the relationship between diseases,candidate drugs, and key genes related to prediction.

2.6 Prediction of pharmacological targets of diosgenin in the treatment of AD/COVID

For the selected potential candidate drug diosgenin, in the SEA( https://sea.bkslab.org/),SuperPred(https://prediction.charite.de/index.php) ,SwissTargetPrediction(http://swisstargetprediction.ch/),TCMSP(http://tcmspw.com/tcmsp.php),Four websites for target prediction.Intersect the predicted targets with the differential genes of AD and COVID, and then draw a Venn diagram.

2.7 PPI

Import the differential genes of AD/COVID and the intersection genes of diosgenin targets into a String database to construct a PPI network, and import the obtained interaction pairs into Cytoscape mapping.Calculate the score of nodes using CytohHubba (due to the limited number of nodes in the PPI network, it is difficult to mine MCODE modules, so the CytohHubba score is directly calculated).Sort according to Degree and select the top 6 genes, which are hub node genes.

2.8 Enrichment analysis of drug target genes

For the intersection genes of AD/COVID differentially expressed genes and diosgenin targets, enrichment analysis was conducted using the cluster profiler package in R language.Including Gene Ontology (GO) and KEGG pathway enrichment analysis.Set the threshold p.adjust<0.05 to screen the enriched pathways, and use the dotplot function of enrichplot package to draw a bubble plot.For each enrichment, only the top ten with the lowest P-value are selected for display.

2.9 Molecular docking

Download the PBD file of pharmacological targets from the RSCBPDB database and the mol2 file of diosgenin from TCMSP.Next, use Pymol software to remove ligand molecules and water molecules from CASP7, and then use the SwissDock website (http://www.swissdock.ch/) Conduct online molecular docking.

3.Research results

3.1 DEGs Enrichment Analysis of AD and COVID-19

We screened 4471 ADDEGs and 3079 COVID-19 DEGs obtained through the Wayne graph package tool in R language, and ultimately obtained 89 shared up regulated DEGs for AD and COVID-19, and 208 shared down regulated DEGs for AD and COVID-19 (Figure 1A).GO analysis of differentially expressed genes upregulated and downregulated by AD and COVID-19 revealed that upregulated DEGs were mainly enriched in biological processes such as immune response, cell apoptosis, and inflammatory response (Figure 1B),while downregulated DEGs were mainly enriched in biological processes such as vesicle mediated transport and hydrogen ion transmembrane transport (Figure 1C).Further KEGG analysis revealed that upregulated DEGs were mainly enriched in TNF signaling pathway and NF KappaB signaling pathway IL-17 signaling pathway and other related pathways (Figure 1B), while downregulated DEGs are mainly enriched in aldosterone synthesis and secretion, gastric acid secretion, saliva secretion and other related pathways (Figure 1C).Metascape analysis found that upregulation of DEGs is mainly related to biological processes such as neutrophil degranulation, inflammatory response, and bacterial response(Figure 1D), as well as diseases such as atopic dermatitis, infection,and pneumonia (Figure 1F).It is mainly regulated by RELA,NFKB1, and STAT1 (Figure 1H).Metascape analysis found that downregulation of DEGs is mainly related to biological processes such as calcium regulation in cardiac cells, cation homeostasis in vivo, and differential regulation of intracellular protein transport in myeloid cells (Figure 1E), as well as diseases such as cerebellar ataxia, cerebellar dysarthria, and cognitive impairment (Figure 1I).

3.2 TF regulatory network and potential drug candidate screening for AD and COVID-19

In order to further reveal the TFs regulatory network in AD and COVID-19, we used the Trust database as the root data to select TFs from the DEGs in AD and COVID-19.The TFs present in both AD and COVID-19 DEGs, as well as downstream target genes present in at least one disease DEGs, were selected as key genes for AD and COVID-19.Using this as a standard, we used Wayne plot analysis to identify 6 core genes in AD/COVID-19 upregulated DEGs (Figure 2A) and 3 core genes in AD/COVID-19 downregulated DEGs(Figure 2B).Next, we constructed a visual TFs target gene network and discovered 7 other target genes regulated by TFs in the DEGs shared by AD and COVID-19.

Fig 1 DEGs enrichment analysis of AD and COVID-19

Next, we found downstream target genes of the core TF genes mentioned above in the shared DEGs of AD and COVID-19.These target genes were intersected to obtain the intersection target genes.The intersection of target genes and core TF constitutes the core gene.Among them, there are 13 key genes (6 TF+7 downstream target gene mRNA) in the upregulated DEGs shared by AD and COVID-19: CEBPB, BCL6, SPI1, NFKBIA, BCL3, TNFAIP3,ADM, S100A9, GADD45G, TNFAIP6, RNASE2, FAS, MMP9.There are three key genes in the downregulated DEGs shared by AD and COVID-19: BCL11A, STAT4, and BDP1.Visualize the interaction network between key genes using Cytoscape (Figure 2F).Introducing key genes into DGIdb predicted drugs, it was found that upregulation of key genes predicted 51 drugs, and downregulation of key genes predicted 3 drugs.Furthermore, Venn plot analysis was performed on 4129 drugs predicted by the CTD website for the treatment of AD and 6939 drugs for the treatment of COVID-19.It was found that there were 20 intersecting drugs such as aspirin, diosgenin, and curcumin between the upregulated key gene prediction drugs and the treatment of AD/COVID-19 (Figure 2D).However, there is no intersection between downregulating key gene prediction drugs and treating AD/COVID-19 drugs (Figure 2E).The Sangi diagram reveals the correlation between diseases, drugs, and targets (Figure 2C).

3.3 Diosgenin therapy for AD targets and molecular docking

Through SEA and SuperPred databases, it was predicted that diosgenin has 379 action targets.Intersection analysis with AD DEGs revealed 36 intersecting genes, indicating that it is a pharmacological target for diosgenin in treating AD (Figure 3A).GO analysis of 36 genes showed that diosgenin can affect the regulation of intracellular calcium ion levels, response to oxygen concentration,and response to hypoxia in cells (Figure 3B).In addition, KEGG analysis showed that these 36 intersecting genes are closely related to pathways such as neural active ligand receptor interaction, PI3K Akt signaling pathway, and extracellular trap formation in neutrophils(Figure 3E).Import 36 intersecting genes into the String database to construct a PPI network, obtain and visualize them using Cytoscape(Figure 3C).At the same time, calculate the scores of nodes based on CytohHubba, and sort them according to Degree.Select the first 6 genes as key node genes, namely CC Chemokine receptor 1 (CCR1),CXC chemokine receptor 4 (CXCR4) Hypoxanthine phosphoribose transferase 4 (HPRT4), integrin β 1 (Integratin Beta 1, ITGB1), Toll like receptors 4 (TLR4), Toll like receptors 8 (TLR8) (Figure 3D).Then, through the SwissDock website (http://www.swissdock.ch/ )Through online molecular docking, it was determined that diosgenin can bind with CCR1, CXCR4, HPRT4, ITGB1, TLR4, and TLR8(Figure 3H-M).

Fig 2 TF regulatory network and potential drug candidate screening for AD and COVID-19

3.4 Diosgenin therapy for COVID-19 targets and molecular docking

Intersection analysis was conducted between 379 action targets of diosgenin and COVID-19 DEGs, and 27 intersection genes were found.It is considered that this is the pharmacological target of diosgenin for treating COVID-19 (Figure 4A).GO analysis of 27 genes showed that diosgenin can affect biological processes such as adenosine cyclase regulating G protein coupled receptor signaling pathway, negative regulation of catabolism, and lymphocyte differentiation (Figure 4B).In addition, KEGG analysis showed that these 27 intersecting genes are mainly associated with neural active ligand receptor interactions, sphingolipids signaling pathways,AMPK signaling pathways, insulin signaling pathways, and other signaling pathways (Figure 4E).Import 27 intersecting genes into the String database to construct a PPI network, obtain and visualize them using Cytoscape (Figure 4C).At the same time,calculate the scores of nodes based on CytohHubba, sort them based on Degree, and select the first 6 genes as key node genes,namely B cell lymphoma 6 (BCL6), cannabinoid receptor 2 (CNR2)Phosphoinosine 3-kinase regulatory subunit 1 (PIK3R1), Mechanical Target of Rapamycin (mTOR), Heat Shock Protein 90 kDA , Class B, member 1 (Heat shock protein 90 kDA alpha, class B, member 1, HSP90AB1), SIPR2 (Figure 4D).Then, through the SwissDock website ( http://www.swissdock.ch/ ) Through online molecular docking, it was determined that diosgenin can bind to BCL6, CNR2,PIK3R1, MTOR, and HSP90AB1 (Figure 4H-M).

4.Discussion

AD is the most common type of dementia, with a complex pathogenesis and great difficulty in treatment.Currently, several widely recognized drugs for treating AD, such as donepezil hydrochloride and galantamine, have only played a very limited role in alleviating cognitive function, mental disorders, delaying disease progression, and cannot cure AD.However, in recent years, studies have found that AD is accompanied by various diseases during its occurrence and development, such as inflammatory diseases,mental disorders, circulatory system diseases, and so on[12].Because studies have found that intervention in these comorbidities can also have certain preventive and therapeutic effects on AD, and these comorbidities are also considered variable risk factors for AD.Therefore, we speculate that analyzing and speculating on the shared etiology molecular network of AD and its comorbidities may provide new therapeutic ideas for the treatment or prevention of AD[12].

Fig 3 Diosgenin treatment of AD targets and molecular docking

Fig 4 Diosgenin therapy for COVID-19 targets and molecular docking

COVID-19 is a novel infectious disease caused by SARS-CoV-2.Since its discovery in December 2019, COVID-19 has swept the world, causing countless losses and casualties.Epidemiological evidence suggests that advanced age and complications are the biggest risk factors for poor prognosis in COVID-19 patients[13],which are also the main risk factors for AD.There are various signs indicating that there may be an intrinsic connection between COVID-19 and AD,.A study has found that one-third of COVID-19 patients admitted to the hospital due to acute respiratory distress syndrome have cognitive impairment upon discharge[14].Another observational study found that two COVID-19 patients experienced prolonged infection and rapidly developed dementia after 5-10 months of follow-up[15].Recently, a comprehensive molecular survey showed that patients who died from COVID-19 experienced extensive inflammation and degeneration in the brain, including those who did not experience neurological symptoms.These studies also reported a certain overlap between AD marker genes and genes upregulated in COVID-19 infection[16].This is consistent with the bioinformatics analysis results of this study.It also indicates that the shared pathogenic pathway between AD and COVID-19 is mainly related to inflammation and immune activity.At present,patients with these two diseases are widely present worldwide, and the vicious cycle of exacerbating each other and promoting disease progression greatly increases the difficulty of treatment and social burden[3].Therefore, it is urgent to develop an effective treatment plan that can simultaneously treat AD and COVID-19 based on the shared pathogenic pathways of AD.

4.1 Analysis of Common Pathogenic Molecular Networks between AD and COVID-19

Based on bioinformatics analysis, we revealed the regulatory network of AD/COVID-19TF and found that six core genes play a key role in the pathogenesis of AD and COVID-19, including the B-cellCLL/lymphoma6 gene (BCL6) and nuclear factor Kappa B 1a (NuclearFactorKappa B 1A, NFkBIA, Tumor Necrosis Factor Alpha induced Protein3 (TNFAIP3), Adrenomedullin (ADM),FAS, Matrix metalloproteinase 9 (MMP9).BCL6 is a transcription inhibitor belonging to the BTB/POZ transcription factor family[17].And BCL6 is the target of miR-155, and previous studies have shown that miR-155 increases after swallowing SARS-CoV2 spike protein in cells[18].In addition, the content of BCL6 in the AD brain was significantly reduced compared to the control group [19].It can be seen that BCL6 may play an important role in the pathogenesis of AD/COVID-19.In addition, BCL6 can inhibit hundreds of target genes by binding to them, and then recruit several different chromatin modification assisted inhibitory complexes to suppress these genes,thereby mediating their effects.The structural characterization of BCL6 co inhibitory complex also confirms that BCL6 may be a promising drug target[17].NF- κ The B signaling pathway is one of the most widely recognized intracellular signaling pathways in inflammatory response, and NFKBIA is a key regulatory factor in inflammatory response to AD and SARS-CoV-2 infections[20,21].TNFAIP3 is a cytoplasmic zinc finger protein initially identified as NF- κ Double inhibitor of B activation.At the same time, TNFAIP3 can downregulate the proliferation of advanced microglia, which is also believed to play an important role in the pathogenesis of AD.ADM can participate in inflammatory response by reducing ROS production[23] and directly promoting the activation of type II intrinsic lymphocytes in vivo and in vitro, and is an important mediator in regulating inflammation[24].Fas (CD95) is a cell surface receptor of the tumor necrosis factor superfamily, and its expression is significantly increased on CD4+T and CD8+T cells in COVID-19 patients[25].Some studies have shown that serum soluble Fas in AD patients is positively correlated with age, disease course, and MMSE score[26].MMP9 is a glycoprotein with type IV collagenase activity,present in neutrophils, lymphocytes, and dendritic cells, involved in angiogenesis and the production of inflammatory cytokines[27,28].Studies have also confirmed that MMP9 is a drug target for treating AD and COVID-19[29,30].

4.2 Network pharmacology analysis to identify potential candidate drugs that can simultaneously treat AD and COVID-19

Then, we submitted the key TF and its targets to the DGIdb database for predicting drugs for AD and COVID-19, and identified 20 drugs with overlapping predicted drugs, AD treatment drugs,and COVID-19 treatment drugs.However, during the screening process, we found that celecoxib, indomethacin, and others belong to antipyretic, analgesic, and anti-inflammatory drugs; Aspirin and atorvastatin belong to the secondary prevention drugs of cerebrovascular diseases; Progesterone and testosterone belong to hormones; Cisplatin, gefitinib, daunomycin, fenveramide,methotrexate, and fluorouracil are anti-tumor drugs; Vitamin D3 belongs to the category of drugs that promote the absorption of calcium and phosphorus; Villisone belongs to the category of positive inotropic drugs; Lithium preparations and paroxetine belong to the category of antipsychotic drugs; The above drugs are very limited in their applicability to the population and have many side effects,making them unsuitable for large-scale promotion.However, we found that diosgenin, wedelia lactone, and curcumin are biologically active drugs with high safety and wide range of effects, suitable for various populations.Due to the relatively limited research on the treatment of inflammation and dementia with trilobalide and curcumin, and insufficient research evidence, in this study, we chose diosgenin as a potential candidate drug for the treatment of AD and COVID-19 for further research.

Diosgenin is a steroidal saponin that can target both NFKBIA and FAS[31,32], and has various anti-inflammatory, immune regulatory,and lipid-lowering effects[33].In recent years, studies have confirmed that diosgenin is safe, low toxic, and has shown beneficial effects on various diseases, including Alzheimer’s disease[6,7].Studies have shown that diosgenin can exhibit good anti AD effects by regulating oxidative stress and inflammatory response[7], indicating that diosgenin may also alleviate inflammation and oxidative response of COVID-19.We used network pharmacology methods to analyze the detailed anti AD/COVID-19 mechanism of diosgenin.The 13 core genes in the TF regulatory network have been identified as potential pharmacological targets of diosgenin against AD/COVID-19.Next, we conducted GO/KEGG analysis on the medicinal targets of diosgenin and the cross genes of AD and COVID-19DEGs,respectively.It was found that diosgenin exerts therapeutic effects on AD and COVID-19 by affecting the homeostasis and inflammatory response of the internal environment.

4.3 Molecular docking analysis of pharmacological targets of diosgenin in the treatment of AD and COVID-19

Finally, we revealed the direct targets of diosgenin on AD and COVID-19 through molecular docking analysis.Diosgenin may improve AD symptoms through CCR1, CXCR4, HPRT4, ITGB1,TLR4, TLR8, all of which belong to genes related to inflammation and immune activity.CCR1 is a widely studied G-protein coupled receptor target that is expressed on various types of white blood cells.It is associated with triggering and exacerbating inflammation, and is therefore considered a good target for autoimmune and inflammatory treatment applications[34].Studies have confirmed that the number of CCR1 positive plaques like structures in the hippocampus and olfactory cortex is highly correlated with dementia status, and is an early specific marker of AD.CXCR4 is a highly conserved 7-fold transmembrane G protein coupled receptor.Research has shown that the CXCL12/CXCR4 signaling pathway is involved in the pathological process of AD, indicating that CXCR4 may serve as a target for early prevention of mild cognitive impairment and the diagnosis and treatment of AD.No correlation has been found between HPRT4 and AD.Further experiments are needed to verify this.ITGB1 is a superfamily of integrins.In recent years, studies have found that ITGB1 is significantly elevated in astrocyte specific extracellular vesicles enriched in total extracellular vesicles of AD,and is associated with the brain β- There is a correlation between amyloid protein and Tau protein load[37].TLRs have been proven to be involved in the pathological process of Alzheimer’s disease β The production and inflammatory response of TLR4[38], targeting TLR4, is also believed to improve the symptoms of AD by inhibiting microglial cell activation and reducing cytokine production[39].However, the impact of TLR8 on AD is still unclear and further research is needed to confirm.

Molecular docking analysis also revealed that diosgenin may affect COVID-19 through BCL6, CNR2, PIK3R1, MTOR, and HSP90AB1.BCL6 is a transcription inhibitory factor necessary for the development of follicular T helper cells (Tfh) to support the formation of germinal centers[40].Previous literature has shown that defective BCL-6+Tfh cell generation and humoral immune induction disorders in early COVID-19 disease are associated with limited persistence of antibody responses in COVID-19[41].In recent years,studies have found that CNR2 can affect the regulatory effect of endogenous cannabinoids on the immune system, and variations in CNR2 may have an impact on the severity of COVID-19[42].PIK3R1 is considered a tumor suppressor factor, and in a preclinical research literature, PIK3R1 has also been identified as a pharmacological target for ferulic acid treatment of COVID-19[43].MTOR is a serine/threonine kinase, which is the main regulatory factor of cell metabolism.mTOR is a serine/threonine kinase, which is the main regulatory factor of cell metabolism.It can respond to multiple signals to regulate cell growth and proliferation, and its signaling pathway is not regulated in many human diseases.MTOR also plays a crucial role in regulating autophagy[44], and current computer, in vitro, and in vivo data unanimously support the role of the PI3K/Akt/mTOR pathway in COVID-19.It is recommended to use specific inhibitors of this pathway, which may improve the course of COVID-19 by downregulating a comprehensive mechanism of excessive inflammatory response, cellular protection, and antiviral effects[45].HSP90 plays an important role in various viral infections,and relevant bioinformatics analysis has found that the key signaling pathway against COVID-19 is the estrogen signaling pathway associated with AKT1, HSP90AB1, and BCL2 targets[46].Therefore,we speculate that diosgenin can regulate the inflammatory pathway and immune cell activity to exert therapeutic effects on AD and COVID-19 by targeting these pharmacological targets.In this study, we revealed for the first time through bioinformatics analysis the TF regulatory network shared by AD and COVID-19, as well as a potential drug that can simultaneously treat both diseases -diosgenin.In addition, we also discovered therapeutic targets of diosgenin against AD/COVID-19 through network pharmacology,theoretically proving that the use of diosgenin can simultaneously treat AD and COVID-19.For non elderly patients with COVID-19,diosgenin can not only improve the clinical symptoms of COVID-19, but also act on targets such as CCR1, CXCR4, HPRT4, ITGB1,TLR4, TLR8, etc., reducing the risk of developing AD in this group of people in the later stage, and even have a certain therapeutic effect on preclinical AD in this group, delaying the onset of AD.And for AD patients without COVID-19, diosgenin can also act on targets such as BCL6, CNR2, PIK3R1, MTOR, HSP90AB1 while treating AD, reducing the risk of COVID-19 infection.Finally, for AD patients who have already suffered from COVID-19, diosgenin can also treat both diseases simultaneously, break the vicious cycle of their mutual aggravation, and improve their prognosis.It is an effective strategy for treating AD/COVID-19.In summary, this study provides a theoretical basis for exploring new clinical pathways for the treatment of AD and COVID-19.Further basic experiments and clinical studies are needed to validate it in the future.We will further explore the molecular mechanisms related to the treatment of AD and COVID-19 with diosgenin, laying a theoretical foundation for the development of anti AD and COVID-19 related products with diosgenin.

All authors in this article do not have any conflicts of interest

Authors’ Contribution

Li Hongyan is responsible for project design and article revision;Zhang Huaxiong is responsible for data collection, organization,analysis, and article writing; Zhang Minghui is responsible for organizing images and writing articles.