An integrated in-silico Pharmaco-BioInformatics approaches to identify synergistic effects of COVID-19 to HIV patients

Abstract

Background

With high inflammatory states from both COVID-19 and HIV conditions further result in complications. The ongoing confrontation between these two viral infections can be avoided by adopting suitable management measures.

Purpose

The aim of this study was to figure out the pharmacological mechanism behind apigenin's role in the synergetic effects of COVID-19 to the progression of HIV patients.

Method

We employed computer-aided methods to uncover similar biological targets and signaling pathways associated with COVID-19 and HIV, along with bioinformatics and network pharmacology techniques to assess the synergetic effects of apigenin on COVID-19 to the progression of HIV, as well as pharmacokinetics analysis to examine apigenin's safety in the human body.

Result

Stress-responsive, membrane receptor, and induction pathways were mostly involved in gene ontology (GO) pathways, whereas apoptosis and inflammatory pathways were significantly associated in the Kyoto encyclopedia of genes and genomes (KEGG). The top 20 hub genes were detected utilizing the shortest path ranked by degree method and protein-protein interaction (PPI), as well as molecular docking and molecular dynamics simulation were performed, revealing apigenin's strong interaction with hub proteins (MAPK3, RELA, MAPK1, EP300, and AKT1). Moreover, the pharmacokinetic features of apigenin revealed that it is an effective therapeutic agent with minimal adverse effects, for instance, hepatoxicity.

Conclusion

Synergetic effects of COVID-19 on the progression of HIV may still be a danger to global public health. Consequently, advanced solutions are required to give valid information regarding apigenin as a suitable therapeutic agent for the management of COVID-19 and HIV synergetic effects. However, the findings have yet to be confirmed in patients, suggesting more in vitro and in vivo studies.

1. Introduction

The ongoing respiratory infection COVID-19 (Coronavirus Disease 2019), an outbreak, was diagnosed in December 2019 as a recent worldwide health issue [1]. The novel coronavirus was similar to a virus named SARS-CoV and, therefore, called SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2) [2]. Though the COVID-19 fatality rate is between 2 and 3%, it is lower than the SARS fatality rate of approximately 10% [3]. They were initially discovered in Wuhan, Hubei Province, China, then spread worldwide. The origin of COVID-19 was connected to the Hunan Seafood Wholesale market epidemiologically. Inoculating the bronchoalveolar fluid from pneumonia-infected patients of unknown origin in human epithelial cells and Vero E6 and Huh7 cell lines resulted in the identification of new coronavirus SARS-CoV-2 [4]. They are positive-sense, single-stranded, enveloped RNA virus that infects humans and other animals, with genome sizes ranging from 26 to 32 kb. Coronavirus has four subfamilies, i.e., alpha, beta, gamma, and delta. The alpha and beta subfamilies have been found in mammals, whereas the gamma and delta subfamilies have been found in pigs and birds. While beta coronaviruses can cause severe illness and even death, alpha coronaviruses are asymptomatic or slightly symptomatic [5]. COVID-19 is a fast-spreading virus that has infected millions of people, with hundreds of thousands of deaths. COVID-19 symptoms can last for two weeks in an acute condition and three to six weeks in a severe illness [6]. Researchers discovered that SARS-CoV-2 hinders normal immune responses, resulting in an impaired immune system and uncontrolled inflammatory reactions in patients with acute and severe COVID-19 infection. Elevated cytokine levels, granulocyte and monocyte abnormalities, lymphocyte activation and dysfunction, lymphoma and a rise in total antibodies and immunoglobulin G (IgG) were among the clinical manifestations in a COVID-19 patient [7]. Additionally, infection with COVID-19 also causes a decrease in CD4⁺ and CD8⁺ T cells, B cells, NK cells, monocytes, eosinophils and basophils, and increases neutrophil count, neutrophil to lymphocyte ratio [8]. COVID-19 displayed a broad array of clinical manifestations, ranging from asymptomatic to symptomatic, including viral pneumonia with respiratory failure, multiorgan and systemic dysfunctions, and may eventually lead to death [9]. However, the lungs are the primary target organ for COVID-19, but it has also expanded to several other organs, i.e., the heart, kidney, brain, gut, and blood vessels [10,11]. Patients with COVID-19 generally suffer from fever, dry cough, dyspnea, headache, weakness, dizziness, diarrhea, and vomiting [12]. While about 37.9 million HIV-positive people are at higher risk of contracting COVID-19 [13], lower CD4 levels, an unsuppressed HIV RNA viral load, or the frequent use of antiretrovirals including non-nucleoside reverse transfer inhibitors (NNRTI), nucleoside reverse transfer inhibitors (NNRTI), or protease inhibitors may all increase the chance of COVID-19 infection in HIV patients [14]. HIV patients have a weakened immune system, making them more susceptible to infection by other viruses. This immunodeficiency in HIV infection makes them more prone to COVID-19 viral infection, which can cause cytokine storms by increasing the levels of interleukins such as IL-1, IL-2, IL-10, IL-17, interferons, MCP1, MIP1A and 1B, IP10 within the blood serum level [15]. Moreover, HIV prevention, HIV testing, and HIV treatment have become more difficult due to COVID-19, resulting in deteriorating HIV conditions in HIV infected patients [16]. Furthermore, HIV-positive people are more prone to COVID-19 infection because they contain abnormal serum electrolytes, as well as suffer from neutropenia, anemia, and thrombocytopenia, all of which are associated with the infection [17]. Although the world's attention is focused on the COVID-19 outbreak, millions of people are suffering from chronic illnesses such as HIV. The continuous conflict between these two viral infections and their synergetic effects must be reduced by establishing appropriate treatment approaches for COVID-19 and HIV patients.

Quercetin, Kaempferol, Myricetin, Apigenin, and Lutein are the five most common plant flavonoids out of the over 6000 distinct types [18]. Due to its low toxicity, lower molecular weight and several advantageous biological properties, apigenin, one of the main monomeric flavonoids included in a typical diet, has drawn interest among researchers [19]. Apigenin, a bioactive flavonoid found in fruits and vegetables with anti-inflammatory, anti-cancer, and antioxidant properties [20]. Apigenin has an anti-HIV-1 effect because it inhibits the actions of ribonuclease H (RNase H) and integrase [21], and it has been discovered to be a potential COVID-19 viral antagonist by particularly binding to RNA polymerase and crucial viral proteins [22]. Apigenin is a potent therapeutic agent, as evidenced later by its significant interactions with 5 hub proteins MAPK3, RELA, MAPK1, EP300, and AKT1 as well as by its pharmacokinetic propertiesthat could be utilized as a key element in treating COVID-19 and HIV patients. In this study, we will consider 75 common genes from HIV and COVID-19 virus and target apigenin as the major therapeutic element to minimize the synergetic effects of COVID-19 to the progression of HIV patients, using systems biology and bioinformatics tools. Apigenin is expected to be a useful molecular target for HIV and COVID-19 infection treatment.

2. Methods

2.1. Apigenin related gene screening

Different databases are used to detect possible targets of apigenin, such as Comparative Toxicogenomics Database (http://ctdbase.org/), DGIdb: The Drug Gene Interaction Database (https://www.dgidb.org/), PharmMapper (http://www.lilab-ecust.cn/pharmmapper/), and Swiss Target prediction (http://www. swisstarget prediction. ch/). CTD (Comparative Toxicogenomics Database) is a literature-based public resource that shows the correlations between chemicals, diseases, environmental exposure, phenotypes, and gene products [23]. DGIdb offers a user interface for exploring gene lists against a database of drug-gene interactions and genes that could be draggable [24]. Through reversed pharmacophore matching of the query substance against an in-house pharmacophore model database, the online tool pharmMapper can be used to identify therapeutic targets [25]. Utilizing Swiss Target prediction, the targets of bioactive components can be determined using a collection of 2D and 3D similarity characteristics with known ligands [26]. To find potential apigenin targets and enhance the reliability of the findings, we choose all of the genes from each database.

2.2. COVID-19/HIV related gene screening

Through evaluating available transcriptome RNA-seq COVID-19 data (GSE152418) [27], COVID-19-associated targets were identified in DEGs, available at: (https://www.ncbi.nlm.nih.gov/geo/). We retrieved the DEGs by downloading the information from this research literature, while the DEGs were screened using Benjamini Hochberg modified P-values < 0.05 and logfoldchange > 1. Further research will be conducted on the genes that match the possible inclusion requirements. To look for human target genes relating to COVID-19, we also used PubChem (https://pubchem.ncbi.nlm.nih.gov/), CTD (http://ctdbase.org/), COVID-19 DisGeNET data (https://www.disgenet.org/covid/diseases/summary/), COVID-19 host gene database (https://baillielab.net/). Finally, we selected all of the genes from the databases that could be used as possible targets. In the case of HIV, to determine HIV-associated genes, DEGs were used through transcriptome analysis of HIV data (GSE116073) [28], available at (https://www.ncbi.nlm.nih.gov/geo/), where the screening requirement for DEGs was P-values < 0.05 and logfoldchange > 1. Moreover, to find genes associated with HIV, we used CTD (http://ctdbase.org/), DisGeNET (https://www.disgenet.org/), and Gene cards (https://www. genecards. org/). The Venn diagram tool (https://bioinfogp.cnb.csic.es/tools/venny/) was used to identify shared genes between apigenin and COVID-19/HIV-associated targets.

2.3. GO (gene ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway enrichment analysis

Gene Ontology is the most extensively used resource, a knowledge-based resource that deals with gene functions and provides organized, computable knowledge about gene functions and gene products [29]. The GO covers the roles of the major target of apigenin in the therapy of COVID-19 and HIV in the current research. KEGG pathway is a resource provided by the Kyoto Encyclopedia of Genes and Genomes (KEGG), which is widely used as a knowledge base to analyze biological pathways and cellular processes [30]. For GO enrichment and KEGG analysis, we searched for David's (https://david.ncifcrf.gov/tools.jsp) database, inserted the 75 core target genes, selected the official gene symbol, and Chose “Homo sapiens” as the species. Biological process (BP), cellular component (CC), and molecular function (MF) are the three ontologies used in GO enrichment analysis. The KEGG pathway analysis set the p-value to less than 0.05 (P<0.05). Weishengxin (http://www.bioinformatics.com.cn/en) was then used to depict the bar and bubble plot after obtaining the GO enrichment and KEGG analysis.

2.4. PPI network construction

PPI (protein-protein interaction) refers to a group of genes and proteins that communicate to attain biological processes, molecular function, and cellular components [11]. We searched for String (https://string-db.org/) database (version: 11.5) to construct a PPI network via uploading the core targets linked to apigenin, COVID-19, and HIV. Organisms were confined to “Homo sapiens” with a high confidence score of 0.900. To generate and analyze networks, Cytoscape v3.9.1 was adopted [31]. The number of nodes denotes proteins in the PPI network, while the number of edges represents protein interactions.

2.5. Cluster analysis and selection of Hub proteins

When it comes to finding functional modules and predicting protein activities, cluster analysis of biological networks is essential. Clustering findings must be shown to analyze the structure of biological networks. In our study, we employed the ClusterViz plugin of Cytoscape 3 for cluster analysis. Among three algorithms, we choose only MCODE (Molecular Complex Detection) to find highly interconnected proteins with an established ‘parameter’ including degree score threshold: 2, node score threshold: 0.2, and K-Core: 2 [32].

The cytohubba plugin of Cytoscape (http://apps.cytoscape.org/apps/cytohubba) is used to do network topology studies [33]. The hub nodes of networks play a significant biological role and exhibit a more vital link to disease pathophysiology. Cytohubba is an open platform used to study the essential hub targets for regulating the subnetwork. In this study, cytohubba is used to determine the top genes with the shortest path, which are considered significant regulators of the subnetwork. For hub genes identification, we searched for the STRING (https://string-db.org/) database (version: 11.5) by uploading the core targets linked to apigenin, COVID-19 and HIV. With a high confidence score of 0.900, organisms were confined to “Homo sapiens.” Then the network was sent to Cytoscape (version:3.9.1) for analysis. We applied five methods of cytoHubba plugin such as three local rank methods: degree, MNC (maximum neighborhood component), MCC (maximal clique centrality) and two global rank algorithms: betweenness, and closeness centrality. The networks were then analyzed through Cytoscape v3.8.2 version. However, the local methods graded hub proteins based on their proximity. According to the global approach, hubs were assigned a ranking based on their contact with the whole network. However, we have identified common hub proteins comparing five cytohubba methods through jvenn online server (http://jvenn.toulouse.inra.fr/app/example. html) [34].

2.6. Chord plot for GO term and KEGG pathway

The Chord plot illustrates a comprehensive linkage between the logFC of DEGs and their enriched GO terms or KEGG pathways. Weishengxin (http://www.bioinformatics.com.cn/en), a free web platform, generated the chord plot. A GO chord plot represented 12 GO terms, four from each of the BP, CC and MF ontologies. Furthermore, the chord plot of the KEGG pathway contains a total of 15 pathways. The logFC is calculated from their DEGs. The 12 GO terms, together with their hub genes and logFC, are then entered into Weishengxin, and the chord plot for GO terms is obtained. Similarly, 15 KEGG pathways with their hub genes and logFC are input into Weishengxin, creating the chord plot for KEGG pathways.

2.7. Validation of hub proteins through molecular docking

For further studies on molecular docking, the most significant hub proteins were found from the protein-protein interactions network analysis [35]. It is an area of dynamics study with many applications in structural medicine architecture, lead optimization, and biochemical routes analysis [36].

2.7.1. Ligands and proteins preparation

The PubChem database was used to get the three-dimensional structure of Apigenin, the co-crystalized structure of each protein and control inhibitors of target proteins [37]. We used abacavir as a controlled drug based on suitable pharmacokinetics and pharmacodynamics properties, food-drug interactions, experimental characteristics and toxicity analysis. Besides, Abacavir is a nucleoside reverse transcriptase inhibitor that fights viruses and is used in conjunction with other anti-retroviral to treat HIV [38]. Semi-quantitative RT-PCR results indicated that the RELA, MYC, HIF1A, and TERT genes had increased expression following treatment with various amounts of abacavir in tumor cells [39]. All information was found in the DrugBank database with accession number DB01048. Using Online Smiles Translator, smile (simplified molecular-input line-entry system) id was formatted to convert SDF file PDB [40]. Then open babel software was used to energy-minimize of selected ligand compounds.

RCSB Protein Data Bank provided the receptors protein data bank (PDB) ID, such as MAPK3 (PDB ID: 4QTB) and AKT1 (PDB ID: 5WBU). After downloading the PDB format of proteins, we have minimized energy through Swiss PDB Viewer software [41]. We have also used Biovia Discovery Studio for removing co-crystalized ligand molecules, adding hydrogen atoms and choosing protein chains.

2.7.2. Molecular docking

Molecular docking was performed using Auto Dock Vina software and utilized in all docking experiments. AutoDock Vina is one of the most popular and frequently used open-source molecular docking systems. It also has a scoring function, the ability to dock several ligands simultaneously, and a batch mode for docking a large number of ligands [42]. The 3D format of hub proteins and ligands was prepared in pdbqt format by removing hetero atoms, water molecules and cofactors using Auto Dock Vina embedded in Molecular Graphics Laboratory (MGL 1.5.6) tools and a grid box was also generated using these tools. Rectangular boxes are used by Autodock and Vina to define the binding site. The box center of the plugin may be set either by supplying precise coordinates or, more visually attractive. In this study, we used grid box at maximizing center and diameter (x, y, z for covering the maximum binding sites area) of each protein and ligand. The “exhaustiveness” parameter in AutoDock Vina was set by default option to 8 since huge grid boxes were employed. From standard Protein Data Bank file input, the LIGPLOT tool automatically builds schematic 2-D representations of protein-ligand complexes. The software is universal for any ligand and may also be used to demonstrate different protein and nucleic acid interactions [43]. Finally, the docking results with the lowest binding energy were visualized using PyMOL 2.3.2 and Discovery studio (Biovia).

2.7.3. Validation of molecular docking tools

Before performing molecular docking, we predicted the active site of each protein and also performed binding pocket analysis through Castp online server. The Computed Atlas of Protein Surface Topography (CASTp) is an online tool for identifying, defining, and quantifying certain geometric and topological features of protein structures [44]. It has shown to be a beneficial tool for a variety of studies, including signaling receptors [45], cancer therapy [46], and identification of drug mechanisms [47].

2.8. Estimation of pharmacokinetics

According to the results of the above computational analysis, apigenin may be a suitable treatment option for the co-infection of COVID-19 and HIV. Apigenin has significant physicochemical, pharmacokinetics, and drug-likeness, and no significant side effects, as per the ADMET assay, and can be evaluated as possible therapeutic targets against COVID-19 and HIV. The ADME (absorption, distribution, metabolism, and excretion) properties of apigenin, as well as its toxicity and drug-likeness are reviewed in this section. The ADME parameters (absorption, distribution, metabolism, excretion) and toxicity of apigenin were estimated using pKCSM [48], predicted on the canonical SMILES of apigenin acquired from Pubchem, and drug-likeness was assessed using the SwissADME [49] tool.

2.9. 9: Molecular dynamics (MD) simulation

Molecular dynamics (MD) simulation of protein-ligand complexes were done using the WebGRO server (https://simlab.uams.edu/) [50]. WEBGRO, a highly autonomous online platform, was used to simulate molecular dynamics on the compounds with greater binding affinities to analyze the protein-ligand interactions. PRODRG tool was used for generating Ligand topology files [51]. For the simulation GROMOS96 43a1 force field was incorporated while SPC was used as the water model, Triclinic was selected for box type and 0.15 M NaCl salt was added to neutralize the system [52]. For energy minimization Steepest Descent Algorithm was selected with 5000 steps. The temperature was set to 300 K, while the pressure was set to 1.0 bar. Leap-frog MD integrator was used to perform the simulations for 50 ns (max allowed 50 ns) with the approximate number of frame per simulation was 1000. Root mean square deviation (RMSD) was used to analyze the conformational alterations detected in the structural level integrity of docked complexes. The Root Mean Square Fluctuation (RMSF), which averages the amount of atoms present and measures the displacements or groups of atoms in relation to a reference structure. It is also used to assess the stability of a structure throughout the course of simulations [53].

3. Results

3.1. Determination of targets for apigenin, COVID-19 and HIV

Several bioinformatics analyses were performed to identify the genes linked with COVID-19, HIV and apigenin therapeutic targets were shown in Fig. 1 . Correspondingly, we searched and detected 475 apigenin-related genes, with 249, 27, 100, and 99 targets chosen from CTD, DGIdb, Swiss Target, and pharmMapper. After testing for duplicated genes, we were able to identify 408 common unique targets. We found 2166 different covid-19 related genes out of 2500 genes after eliminating duplications: 500 out of 2459 (selected from GSE152418), 500 out of 7246 (CTD), 500 out of 1843 (DisGeNET), 500 out of 654 (Pubchem), and 500 (Covid-19 Host genes database). Furthermore, after checking, HIV-duplicated genes were eliminated. There were 1218 different HIV genes out of 1508: 500 out of 26163 (GSE116073), 500 out of 1000(CTD), 129 (DisGeNET), and 379(Gene Cards). Finally, we identified 75 similar genes related to apigenin, HIV and COVID-19 co-infection using an online Venn diagram tool (https://bioinfogp.cnb.csic.es/tools/venny), while volcano plot provided significant up and down-regulated genes, both were represented in Fig. 2 . These 75 core targets can be utilized in future host factor interaction network research.

Fig. 1.

Flow diagram of network pharmacology and bioinformatics approaches on apigenin against COVID-19 and HIV.

Fig. 2.

Volcano plot of DEGs from the RNA seq datasets. (A) Volcano plot of HIV (GSE116073), (B) Volcano plot of COVID-19 (GSE152418). X-axis indicates logFC and y-axis indicates -log10 of the p-value is showing the statistical significance. (C) Venn diagram represented all candidate targets of Apigenin, COVID-19 and HIV.

3.2. Assessment of biological attributes of apigenin against COVID-19 and HIV

We carried out functional annotation of the core targets on GO enrichment and KEGG pathway assay to explore apigenin's biological characteristics and signaling pathways against COVID-19 and HIV. Biological process (BP), cellular component (CC), and molecular function (MF) are the three components of the GO (Gene Ontology) enrichment study, as shown in Fig. 3 and Table 1 . A green, orange and blue bar indicated each of the three categories of biological activity, molecular function, and cellular component. A total of 2293 GO terms were generated, among which BP contributed for 2008, CC represented 126, and MF comprised 159, and the top 10 BP, CC, and MF were chosen demonstrated in a bar diagram. In the bar plot, the height of the bar indicated fold enrichment, while the width specified GO terms. The biological process (BP) in the GO enrichment analysis includes a response to abiotic response to oxygen-containing compound, cellular response to chemical stimulus, response to the drug, response to chemical, response to an organic substance, response to lipid, response to lipopolysaccharide, apoptotic process, and response to the organic cyclic compound. Cellular component (CC) comprised cytosol, membrane raft, membrane microdomain, cytoplasmic part, membrane region, cytoplasm, mitochondrion, extracellular region part, membrane bound organelle, and membrane-enclosed lumen. Molecular function (MF) exhibited enzyme binding, heme binding, identical protein binding, tetrapyrrole binding, oxidoreductase activity, Oxidoreductase activity acting on paired donors with incorporation or reduction of molecular oxygen, NAD(P)H as one donor, and incorporation of one atom of oxygen, receptor binding, Oxidoreductase activity, acting on paired donors, with incorporation or reduction of molecular oxygen, oxygen binding, and protein homodimerization activity. These observations clarify the alterations in biological activity in the body following the delivery of apigenin in the management of COVID-19 in association with HIV.

Fig. 3.

Functional enrichment analyses of the differentially expressed genes (DEGs) in apigenin mediated COVID-19 and HIV co-infection. GO using Biological process (BP), Cellular component (CC) and Molecular function (MF). The combined score was used to sort the pathway term results (based on P-value).

Table 1.

The core targets of apigenin-treated COVID-19 and HIV were linked to the top 30 most GO pathways.

GO ID	GO Pathway	Count	%	P Value	Fold Enrichment	FDR
Gene Ontology Biological Process
GO:0009628	response to abiotic stimulus	48	64	8.49E-39	10.44400389	2.47E-35
GO:1901700	response to oxygen-containing compound	52	69.33	4.76E-37	7.941646203	6.92E-34
GO:0070887	cellular response to chemical stimulus	61	81.33	1.82E-35	5.157663646	1.68E-32
GO:0042493	response to drug	35	46.67	2.31E-35	19.06634225	1.68E-32
GO:0042221	response to chemical	69	92	7.03E-35	3.752402235	4.10E-32
GO:0010033	response to organic substance	61	81.33	7.25E-34	4.839650366	3.52E-31
GO:0033993	response to lipid	42	56	1.05E-33	10.93432558	4.39E-31
GO:0032496	response to lipopolysaccharide	31	41.33	2.34E-33	22.98242826	8.53E-31
GO:0006915	apoptotic process	52	69.33	2.74E-33	6.664254913	8.88E-31
GO:0014070	response to organic cyclic compound	42	56	8.53E-33	10.37916115	2.48E-30
Gene Ontology Cellular Component
GO:0005829	cytosol	41	54.67	9.02E-12	2.930995391	2.39E-09
GO:0045121	membrane raft	14	18.67	7.99E-11	12.28091456	6.33E-09
GO:0098857	membrane microdomain	14	18.67	8.36E-11	12.23673861	6.33E-09
GO:0044444	cytoplasmic part	61	81.33	9.55E-11	1.832161516	6.33E-09
GO:0098589	membrane region	14	18.67	1.62E-09	9.609642185	8.58E-08
GO:0005737	cytoplasm	67	89.33	1.15E-08	1.510914772	5.10E-07
GO:0005739	mitochondrion	25	33.33	2.83E-08	3.573333333	1.07E-06
GO:0044421	extracellular region part	37	49.33	1.58E-07	2.318335912	5.23E-06
GO:0043227	membrane-bounded organelle	69	92	4.10E-07	1.373255795	1.21E-05
GO:0031974	membrane-enclosed lumen	39	52	5.37E-07	2.126201481	1.42E-05
Gene ontology Molecular Function
GO:0019899	enzyme binding	37	49.33	1.53E-16	4.65509223	6.92E-14
GO:0020037	heme binding	15	20	7.99E-16	24.64379562	1.80E-13
GO:0042802	identical protein binding	32	42.67	1.89E-15	5.303799705	2.35E-13
GO:0046906	tetrapyrrole binding	15	20	2.08E-15	23.12465753	2.35E-13
GO:0016491	oxidoreductase activity	18	24	2.09E-08	5.351968296	1.89E-06
GO:0016709	incorporation of one atom of oxygen	7	9.333	3.09E-08	36.64093023	2.17E-06
GO:0005102	receptor binding	24	32	3.36E-08	3.694883721	2.17E-06
GO:0016705	reduction of molecular oxygen	10	13.33	4.42E-08	13.55903614	2.50E-06
GO:0019825	oxygen binding	7	9.333	7.87E-08	31.5112	3.70E-06
GO:0042803	protein homodimerization activity	17	22.67	8.18E-08	5.241589041	3.70E-06

3.3. Evaluation of the apigenin-mediated signaling pathway with COVID-19 and HIV

KEGG pathway enrichment was utilized to identify further the putative actions of apigenin against COVID-19 and HIV. A total of 113 pathways were enriched based on the adjusted p-value (less than 0.05; P<0.05) parameter that was substantially correlated with the target genes through KEGG enrichment analysis on 75 core targets. Among the 113 KEGG pathways, the top 30 pathways were selected and represented by a bubble plot in Fig. 4 and Table 2 . The top 30 pathways with adjusted p-values and gene enrichment included pathways in cancer, hepatitis B, TNF signaling pathway, Tuberculosis, Chagas disease (American trypanosomiasis), toxoplasmosis, influenza A, HTLV-I infection, Herpes simplex infection, osteoclast differentiation, non-alcoholic fatty liver disease (NAFLD), apoptosis, Toll-like receptor signaling pathway, T cell receptor signaling pathway, pertussis, NOD-like receptor signaling pathway, NF-kappa B signaling pathway, Leishmaniasis, HIF-1 signaling pathway, FoxO signaling pathway, small cell lung cancer, prostate cancer, Epstein-Barr virus infection, salmonella infection, legionellosis, pancreatic cancer, p53 signaling pathway, colorectal cancer, amyotrophic lateral sclerosis (ALS), shigellosis. A bubble plot displays these substantially enriched pathways. The number of related genes and the p-values determined the size and color of the nodes in the bubble plot. The shades from green to red reflect the p-values from high to low, and the size of the nodes denoted the related target genes.

Fig. 4.

KEGG pathway study of two diseases (COVID-19 and HIV) and discovery of Apigenin targets. The combined score was used to sort the pathway term results (based on P-value).

Table 2.

The core targets of apigenin-treated COVID-19 and HIV were linked to the top 30 most enriched KEGG terms.

KEGG ID	Pathway	Count	%	P Value	Fold Enrichment	FDR
hsa05161	Hepatitis B	27	36	4.76E-26	17.30969245	3.90E-24
hsa04668	TNF signaling pathway	24	32	6.42E-25	20.85071988	2.63E-23
hsa05142	Chagas disease	21	28	1.27E-20	18.77066008	3.46E-19
hsa05200	Pathways in cancer	31	41.33333	4.63E-19	7.332680008	9.49E-18
hsa04210	Apoptosis	17	22.66667	8.30E-19	25.48888405	1.36E-17
hsa05145	Toxoplasmosis	20	26.66667	1.14E-18	16.9017199	1.55E-17
hsa05152	Tuberculosis	22	29.33333	4.07E-17	11.5542831	4.77E-16
hsa04621	NOD-like receptor signaling pathway	15	20	2.48E-16	24.89985521	2.54E-15
hsa05164	Influenza A	20	26.66667	8.31E-15	10.68499534	7.57E-14
hsa04380	Osteoclast differentiation	18	24	1.43E-14	12.7730555	1.17E-13
hsa05133	Pertussis	15	20	2.11E-14	18.59189189	1.57E-13
hsa04932	Non-alcoholic fatty liver disease (NAFLD)	18	24	1.59E-13	11.08126007	1.04E-12
hsa04620	Toll-like receptor signaling pathway	16	21.33333	1.66E-13	14.03161652	1.04E-12
hsa05140	Leishmaniasis	14	18.66667	2.54E-13	18.33003426	1.49E-12
hsa05168	Herpes simplex infection	19	25.33333	2.96E-13	9.651528578	1.62E-12
hsa04660	T cell receptor signaling pathway	15	20	1.35E-12	13.94391892	6.94E-12
hsa04064	NF-kappa B signaling pathway	14	18.66667	3.93E-12	14.95899348	1.90E-11
hsa05134	Legionellosis	12	16	5.87E-12	20.65765766	2.67E-11
hsa04066	HIF-1 signaling pathway	14	18.66667	1.44E-11	13.55658784	6.22E-11
hsa05222	Small cell lung cancer	13	17.33333	5.70E-11	14.21732909	2.34E-10
hsa05014	Amyotrophic lateral sclerosis (ALS)	11	14.66667	7.03E-11	20.45108108	2.74E-10
hsa05166	HTLV-I infection	19	25.33333	7.97E-11	6.953660353	2.97E-10
hsa05215	Prostate cancer	13	17.33333	8.68E-11	13.73264742	3.10E-10
hsa05210	Colorectal cancer	11	14.66667	6.65E-10	16.49280732	2.27E-09
hsa05132	Salmonella infection	12	16	7.77E-10	13.43992185	2.55E-09
hsa04068	FoxO signaling pathway	14	18.66667	1.03E-09	9.712182332	3.26E-09
hsa05212	Pancreatic cancer	11	14.66667	1.08E-09	15.73160083	3.28E-09
hsa04115	p53 signaling pathway	11	14.66667	1.47E-09	15.26200081	4.31E-09
hsa05169	Epstein-Barr virus infection	13	17.33333	4.14E-09	9.905516172	1.17E-08
hsa05131	Shigellosis	10	13.33333	1.78E-08	14.52491554	4.86E-08

3.4. Construction of PPI network

Getting an insight into cell tissue, biological processes, and function requires a thorough knowledge of the protein protein interaction (PPI) network. Abnormal PPIs can cause various health problems because they are part of the organisms' interactomics. When two diseases share one or more protein sub-networks, they are said to be related. The current research employed the 75 common targets among apigenin, COVID-19, and HIV to create a PPI network. There are 70 nodes and 326 edges in the PPI network. According to the result of the PPI network analysis, the network density is 0.135. Increasing levels of the functional association are correlated with higher levels of network density, which is defined as the percentage of observed connections to possible connections (possible connections correspond to the number of links in a completely connected network) [54]. Network heterogeneity is 0.651, which represents a parameter that can be utilized to estimate the variability of node degrees and the variation of the network's structure, two distinct features of complex networks [55]. An average number of neighbors is 9.314, network diameter is 6, a network radius is 3, a characteristic path length is 2.555, the clustering coefficient is 0.519, and a network centralization is 0.219 were also identified. Typically, the clustering coefficient is applied to determine the frequency of node clusters throughout the PPI network [56]. One of the most useful metrics for identifying vital proteins is network centrality, which is a key indicator of important protein prediction and is based on the edge clustering coefficient. The total edge clustering coefficients of all edges that are directly associated to proteins is the measure of network centrality [57].

The combined scores of core targets were gathered from a string database in this analysis, and the core targets and combined scores were then uploaded into Cytoscape software to build the core target network. The nodes in the network indicate core targets, while the edges reflect interactions between nodes, with the thickness of the edges varying depending on the total score. The PPI network is illustrated in Fig. 5 .

Fig. 5.

PPI network visualization of the potential therapeutic targets of apigenin against COVID-19 and HIV. The network nodes depict target proteins and the edges represent protein-protein relationships.

3.5. Cluster analysis and selection of Hub proteins

Cluster analysis is one of the most widespread techniques for detecting functional modules, active modules and protein functions. It is employed to analyze significant biological networks, including PPI, metabolic, transcription factor-binding, and gene networks. By using the MCODE technique of the clusterViz plugin of Cytoscape, we were able to identify two major modules inside the PPI network. Cluster 1 had 6 nodes with 15 edges, while cluster 2 had 5 nodes with 10 edges (Fig. 6 ). Details information was provided in Table 3 .

Fig. 6.

Cluster analysis of PPI network hunig clusterviz in cytoscape. Based on K-mer scoring Figure (A) indicates cluster 1, (B) cluster 2, (C) cluster 5, (D) cluster 4 and (E) cluster 3.

Table 3.

Clustering information of PPI network through MCODE of clusterViz plugin.

Cluster	Score (density)	Nodes	Edges	Gene name
C1	6	6	15	LCX, SRC, RELA, MAPK3, MAPK8 and MAPK3
C2	5	5	10	CYP3A4, CYP2C9 CYP2C19, CYP2E1 and CYP1A2
C3	4.917	25	59	IFNG, FOS, IL10, JUN, CCL2, IL6, EP300, NFKBIA, PPARG, CHUK, IKBKB, APP, NFKB1, ICAM1, APAF1, CASP7, CYCS, CASP3, CASP9, BAX, CDKN1A, CCNA2, CCNB1, PCNA, and CDK1
C4	3.8	11	19	AKT1, BCL2L1, MAPK1, BCL2, NOS2, SOD2, SOD1, CXCL2, F2, MOL1 and CAT
C5	3	3	3	DDIT3, HSPA5 and ATF6

Hub genes are strongly associated with a large extent of association with other genes in the network, and changes in their expression can affect the majority of the network [58]. Remarkably, five out of ten hub proteins, including MAPK1, MAPK3, RELA, AKT1 and EP300 were common in all methods and represented in Fig. 7 . While only SRC was found in 4 methods except MNC and TNF in 4 methods except MCC. MAPK8, NFKB1 and JUN were found in the closeness, degree, mcc and mnc algorithm. However, these hub proteins can serve as therapeutic targets in COVID-19 and HIV with the therapeutic agent Apigenin. Finally, we discovered the top 5 targets of apigenin against COVID-19 and HIV, including MAPK1, MAPK3, RELA, AKT1 and EP300. Additionally, Table 4 includes information on the topological characteristics of hub proteins through MCC, MNC, degree, betweenness, closeness centrality, stress and clustering coefficient parameter analysis. With the therapeutic agent apigenin, these hub genes can be used as therapeutic targets against COVID-19 and HIV.

Fig. 7.

Network diagram of the core targets of apigenin against COVID-19 and HIV by using five cytoHubba algorithm including (A) Betweenness, (B) Closeness, (C) Degree, (D) MCC, and (E) MNC. Besides, Venn diagram (F) shows five common hub proteins.

Table 4.

Hub proteins identified from protein-protein interaction network analysis and provided information on topological characteristics of most 10 hub proteins.

Hub protein	Description	MCC	MNC	Degree	Closeness	Betweenness	Stress	Clustering Coefficient
SRC	proto-oncogene c-Src	4980	19	21	43.12	495.59	2586	0.31
NFKB1	nuclear factor-kappa B	4164	18	18	41.12	68.81	662	0.49
TNF	Tumor Necrosis Factor	2350	24	24	44.5	601.35	3844	0.32
AKT1	AKT Serine/Threonine Kinase 1	476	23	23	44.12	594.21	3270	0.24
MAPK1	Mitogen-Activated Protein Kinase 1	3073	18	19	41.45	250.48	1558	0.40
JUN	Jun Proto-Oncogene	3426	21	21	42.62	203.77	1794	0.36
EP300	E1A binding protein p300	1566	20	20	42.12	397.53	2636	0.30
HSP90AA1	Heat shock protein alpha 90-kDa	72	10	12	38.28	259.19	1274	0.29
MAPK3	Mitogen-activated protein kinase 3	5496	19	19	41.78	127.18	1228	0.46
RELA	REL-associated protein	7201	22	23	43.78	228.27	1480	0.34

3.6. Chord plot for GO term and KEGG pathway

Entities are grouped radially as geometric chords in a typical chord plot, with their interactions visualized by arcs linking them. The size of the connecting arc represents the importance of relationships, which are assessed in this study by the support values. When the item set had more than two categories, many bridges of the same width were drawn. Different arc colors are used to distinguish among groups of data. As shown in Fig. 8 , the top 12 GO terms, four from each BP, CC, and MF category, were directly associated with apigenin's core targets against COVID-19 and HIV. The top 12 GO terms in the chord plot include response to abiotic stress, cytosol, enzyme binding, response to oxygen-containing compound, membrane raft, heme binding, cellular response to chemical stimulus, membrane microdomain, identical protein binding, response to the drug, cytoplasmic part, and tetrapyrrole binding. The top 15 KEGG pathway related to the core targets of apigenin against COVID-19 and HIV, as shown in Fig. 9 , involves-hepatitis B, apoptosis, influenza A, Toll-like receptor signaling pathway, TNF signaling pathway, toxoplasmosis, osteoclast differentiation, leishmaniasis, Chagas diseases (American trypanosomiasis), tuberculosis, pertussis, herpes simplex infection, pathways in cancer, NOD-like receptor signaling pathway, non-alcoholic fatty liver disease (NAFLD). Finally, we have used Cytoscape V3.8 software to create an interaction network diagram for “Apigenin-KEGG-Gene Ontology Hub Proteins-COVID19-HIV” based on network pharmacology/system biology and represented in Fig. 10 .

Fig. 8.

Gene Ontology Chord plot for significant pathways and potential core targets of apigenin mediated HIV and COVID-19 co-infection.

Fig. 9.

KEGG Chord plot for significant pathways and potential core targets of apigenin mediated HIV and COVID-19 co-infection.

Fig. 10.

Integrated visualization of interaction network of target-disease-function-pathway in apigenin-treated HIV and COVID-19.

3.7. Results of Molecular docking

The in-silico technique is a standard drug design prediction and confirmation tool. This approach has several advantages, including being economical, time-saving, and minimizing the separation of inactive compounds [59]. The capacity of apigenin to attach to 5 tested target proteins was predicted using a molecular docking study and represented in Table 5 . The validation of the docking tools in terms of the RMSD of the co-crystalized ligand in each complex, we have found only 3 complexes that are co-crystalized with unique ligands (CID: 9543416 for 1PME, CID: 23727984 for 3BIY and CID: 24866313 for 4QTB). The docking score in this study indicates the strength of the chemical-protein binding activity: more stable binding conformation and molecular interaction with a lower docking score. Among the five examined target proteins, MAPK3 (−9.1 kcal/mol), AKT1 (−8.9 kcal/mol) and MAPK1 (−8.1 kcal/mol) were shown higher binding affinity above −7.0 kcal/mol compared to control (abacavir) drug and co-crystalized ligands.

Table 5.

Docking score of Apigenin against selected 5 hub proteins of COVID-19/HIV.

Protein (PDB)	Co-crystalized ligand Docking Score	Control Docking Score (kcal/mol)	Apigenin Binding Energy (kcal/mol)	Apigenin Hydrogen bond interaction		Apegenin Hydrophobic bond interaction
				No. of bonds	Residues involved	No. of bonds	Residues involved
MAPK1 (1PME)	−6.0	−4.8	−8.3	3	LYS54, ASP167, LYS54	9	VAL39, ALA52, LYS54, LYS54, ILE84, ILE31, VAL39, ALA52, LEU156
EP300 (3BIY)	−5.8	−5.3	−6.6	1	ARG1312	6	ARG1305, ARG1305,
							GLU1477, PHE1301,
							ASN1304, ARG1305
RELA (3GUT)	Not found	−5.0	−7.4	2	GLU225, TYR100	3	ASP94, PHE239, LYS93
MAPK3 (4QTB)	−6.5	−5.6	−9.1	2	ASN171, CYS183	6	VAL56, TYR53, CYS183, ILE48, ALA69, LEU173
AKT1 (5WBU)	Not found	−6.0	−8.9	3	ARG1945, ALA2420, THR1908	4	ASP1912, THR1908,
							TYR2423, ALA2420

3.8. Protein-ligand binding interaction analysis

As seen in Fig. 11 , apigenin binds to pocket number 1 of each protein in suitable molecular area. Apigenin mainly interacted with MAPK3 protein via two hydrogen bonds ASN171 and CYS183, while 6 hydrophobic bonds were formed using VAL56, TYR53, CYS183, ILE48, ALA69 and LEU173 amino acids. MAPK1 mainly interacted via LYS54, ASP167 and LYS54 residues through hydrogen bond formation, whereas 9 residues were involved hydrophobically including VAL39, ALA52, LYS54 and ILE84 etc. EP300 protein can interact with one hydrogen bond ARG1312 and 6 hydrophobic bonds like ARG1305, GLU1477, PHE1301, and ASN1304 residues etc. In addition, RELA can connect two amino acids GLU225 and TYR100 by forming hydrogen bonds, while 3 amino acids ASP94, PHE239 and LYS93 interacted via hydrophobic bonds formation. Besides, AKT1 also formed 3 hydrogen bonds (ARG1945, ALA2420, and THR1908) and 4 hydrophobic bonds (ASP1912, THR1908, TYR2423, and ALA2420 residues).

Fig. 11.

Molecular docking diagram showed that the binding capability of apigenin against COVID-19 and HIV were presented as 3D diagrams and 2D diagrams, respectively MAPK1 (1pme), EP300 (3biy), RELA (3gut), MAPK3 (4qtb) and AKT1 (5wbu).

3.9. Pharmacokinetics prediction

The ADMET analysis revealed that apigenin does not violet Lipinski's five roles (Lipinski, Ghose, Veber, Egan, Muegge) with a high bioavailability score, i.e., 0.55. The primary stage of pharmacokinetics is absorption, which was assessed using pKCSM. Apigenin is poorly soluble in water (QPlogPo/w = −3.329 mol/L), readily absorbed by the intestine (93.25%), and permeable to Caco2 and skin. Apigenin has the ability to act as a P-glycoprotein substrate but not as a p-glycoprotein inhibitor. Furthermore, distribution via pKCSM demonstrated that apigenin crosses the blood-brain barrier (BBB) (QPlogBB = −0.734) and enters the central nervous system (CNS) (−2.061). The VDss (steady-state volume of distribution) of apigenin is 0.822 L/kg, while a small portion (0.147) of apigenin remains unbound. The metabolism of apigenin is largely mediated by human cytochrome P450 (CYP) present in the liver. We found that apigenin is a non-substrate for CYP2D6 and CYP3A4, as well as an inhibitor for CYP1A2 and CYP2C19 but not for CYP2C9, CYP2D6, and CYP3A4. In terms of excretion, apigenin had a total clearance (logCLtot) of 0.566 ml/min/kg, and no renal OCT2 substrate was found. According to toxicity testing, apigenin does not have AMES toxicity, hepatoxicity, or skin sensitization. However, apigenin has slight toxicity to T. pyriformis (0.38 μg/L) and minnow toxicity (2.432 Mm) and is not an hERG (ether go-go related gene) inhibitor. In humans, apigenin has a maximum tolerated dose of 0.328 mg/kg/day. Moreover, apigenin produces acute (LD50 = 2.45 mol/kg) and chronic toxicity (LOAEL = 2.298 mg/kgbw/day) in rats. The summary of pharmacokinetics characteristics was provided in Table 6 .

Table 6.

Pharmacokinetics and toxicity characteristic of apigenin.

Pharmacokinetics features		Remarks
Drug likeness and Medicinal properties	Lipinski	Yes
	Ghose	Yes
	Veber	Yes
	Egan	Yes
	Muegge	Yes
	Bioavailability score	0.55
	PAINS	0 alert
	Leadlikeness	Yes
Absorption	Water solubility	−3.329
	Caco2 permeability	1.007
	Intestinal absorption (human)	93.25
	Skin permeability	−2.735
	P-glycoprotein substrate	Yes
	P-glycoprotein I inhibitor	No
	P-glycoprotein II inhibitor	No
Distribution	VDss (human)	0.822
	Fraction unbound (human)	0.147
	BBB permeability	−0.734
	CNS permeability	−2.061
Metabolism	CYP2D6 substrate	No
	CYP3A4 substrate	No
	CYP1A2 inhibitor	Yes
	CYP2C19 inhibitor	Yes
	CYP2C9 inhibitor	No
	CYP2D6 inhibitor	No
	CYP3A4 inhibitor	No
Excretion	Total clearance	0.566
	Renal OCT2 substrate	No
Toxicity	AMES toxicity	No
	Max. tolerated dose (human)	0.328
	hERG I inhibitor	No
	hERG II inhibitor	No
	Oral Rat Acute Toxicity (LD50)	2.45
	Oral Rat Chronic Toxicity (LOAEL)	2.298
	Hepatotoxicity	No
	Skin Sensitization	No
	T.Pyriformis toxicity	0.38
	Minnow toxicity	2.432

3.10. Interpretation of molecular dynamics simulation

Root mean square deviation (RMSD) measures the typical distance between the atoms of two molecules that are superimposed and assesses the stability of a certain system. The 1PME receptor complexation with Apigenin was examined using the RMSD plot (Fig. 12 A), which revealed equilibria between 25 and 40 ns with a range of 0.40–0.52 Å. 3BIY receptor shows equilibria between 8 to 16 ns and 20–42 ns with the range of 0.23–0.27 Å and 0.25 to o.35 Å range (Fig. 12B). In the 5WBU receptor, we did not attain an average distance between the backbone atoms of starting structure (reference structure). Instead of a stable structure, it shows some great fluctuations (Fig. 12C) throughout the simulation time. On the other hand, in the 4QTB receptor, we found equilibria between 30 and 50 ns with a range of 0.25–0.35 Å (Fig. 12D).

Fig. 12.

The Root mean square deviation (RMSD) plot of Apigenin-receptor complexes. (A) MAPK1, (B) EP300, (C) AKT1 and (D) MAPK3.

Root mean square fluctuation (RMSF) computes fluctuations (standard deviation) of atomic positions of each amino acid (residues) in the trajectory. A protein structure with a high RMSF value is flexible, has loops, or has weak molecular bonds, whereas one with a low value is stable and has secondary structures like sheets and helices. The 5WBU receptor has revealed a higher RMSF value (about 0.9 nm in 2300 residues) which indicates the structure is flexible, has loops and weak molecular bonds and great fluctuation (Fig. 13 C) than other receptor-complexes.

Fig. 13.

The Root mean square fluctuation (RMSF) plot of Apigenin-receptor complexes. (A) MAPK1, (B) EP300, (C) AKT1 and (D) MAPK3.

The surface area of a protein-ligand complex that interacts with solvent molecules directly is known as the Solvent Accessible Surface Area (SASA). Higher levels signify relative expansion, while lower values suggest greater stability. Fig. 14 C represents the higher levels of SASA value (area) between 470 and 520 nm² and then decreasing the value with equilibria between 20 ns and 48 ns

Fig. 14.

The Solvent Accessible Surface area (SASA) plot of Apigenin-receptor complexes. (A) MAPK1, (B) EP300, (C) AKT1 and (D) MAPK3.

The hydrogen bonds in critical in figuring out how specifically ligands bind. Due to their impact on drug metabolism and absorption, they are seen as a crucial factor in drug design. Fig. 15 A and D indicate higher number of four hydrogen bonds average with two bonds. 3BIY and 5WBU (Fig. 15B and C) complexes represent maximum three hydrogen bonds and average with 1.5 bonds.

Fig. 15.

Hydrogen bonding pattern plot of Apigenin-receptor complexes. (A) MAPK1, (B) EP300, (C) AKT1 and (D) MAPK3.

The radius of gyration (Rg) value reveals the stability of the complex, which is connected to the structure's compactness. The higher Rg value in Fig. 16 C was 3.5 nm which indicates multiple fluctuations and then gradually decreased to reach in 3.33 nm compare to other complexes.

Fig. 16.

The Radius of Gyration plot of Apigenin-receptor complexes. (A) MAPK1, (B) EP300, (C) AKT1 and (D) MAPK3.

4. Discussion

It has been found that HIV-positive individuals have a greater chance of SARS-CoV-2 infection and fatalities than people who do not have HIV [17]. The world's attention has shifted to the propagation and implication of COVID-19 and the synergetic effects of COVID-19 on the progression of HIV. Even though the COVID-19 infectious virus is new, infections have risen rapidly over time. Apigenin has been discovered to be a possible coronavirus inhibitor in a recent study [35], although the mechanism of action against COVID-19 and HIV is still not clear. In this study, we utilized computer-aided approaches to identify common biological targets and signaling pathways linked to COVID-19 and HIV, as well as bioinformatics and network pharmacology approaches of apigenin on COVID-19 and HIV. Additionally, we aim to target apigenin to identify the synergetic effects in COVID-19 to the progression of HIV. To explore the pharmacological mechanism of apigenin, gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) were analyzed. The top 12 GO enrichment pathways were primarily engaged in stress-responsive, membrane receptor and induction pathways. The top 15 KEGG pathways, on the other hand, mainly participated in apoptosis and inflammation pathways. We can therefore speculate that apigenin, by influencing these pathways, could aid in the treatment of COVID-19 and HIV. PPI and hub genes were also assessed, and the top 20 (i.e., MAPK3, RELA, MAPK1, EP300, and AKT1) hub genes were found with the shortest path ranked by degree method. According to our findings, the GO enrichment pathway represents abiotic stress response, cytosol, enzyme binding, membrane raft, and heme binding. When COVID-19 cells are disrupted by pathogens, cytotoxic compounds, or abiotic stresses, overexpression of HSP70 suppresses caspase stimulation and stressed mitochondria, preventing neutrophil deposition in the lungs [36]. Elevated cytosolic concentrations of pro-inflammatory metabolic regulators are responsible for serious symptoms of COVID-19 [37], although reducing HIV peptide stability in cytosol limits epitope presentation and CTL identification, demonstrating an immunological escape mechanism [60]. SARS-CoV-2 penetrates the host cell by enzyme binding, attaching to Angiotensin-converting enzyme 2 (ACE2), a vital controller of the RAAS (renin-angiotensin-aldosterone system). However, COVID-19 alters ACE/ACE2 balance and RAAS activation, eventually contributing to COVID-19 progression [61], whereas angiotensin-converting enzyme inhibitors provide anti-inflammatory advantages for HIV infection management [62]. Both SARS-CoV-2 and HIV attach to solid regions of the plasma membrane, such as the membrane lipid raft, which are also crucial for activation, endocytosis, vesicle transport, apoptosis, and cell-to-cell transmission, as well as being crucial mediators of immunological and inflammatory reactions [[63], [64], [65]]. COVID19 infection is defined by disruption of the lipid membrane raft, and lipid raft therapeutics may be very beneficial in controlling COVID-19 infections [64]. The viral protein possesses an equal binding affinity to the human heme-binding protein, which binds to free heme and reaches the liver, causing COVID-19 disease progression through cytotoxic, procoagulant, vasculotoxic, and pro-inflammatory consequences, as well as stimulation of the complement system. Improved protection against free heme mitigates the intensity of infection in COVID-19 individuals [66,67]. Conversely, hypersensitivity in HIV patients is caused by cell-free heme-dependent redox signaling, which triggers ER stress as well as a reduction in leukocyte endorphin [68]. Hepatitis B, apoptosis, Toll-like receptor signaling pathway and TNF-signaling pathway are largely described by KEGG pathways. Immunosuppressive medications like IL-6 receptor antagonists (tocilizumab, siltuximab), IL-1 receptor antagonists (anakinra), and high-dose corticosteroids are used to minimize immune-mediated multiorgan damage in chronic COVID-19 individuals regulating cytokine storm. Still, they increase the possibility of Hepatitis B virus (HBV) reactivation through upregulating HBV DNA levels [69]. In contrast, HIV infection raises the chances of chronicity after HBV exposure by reducing CD4 lymphocytes, resulting in severe hepatic injury and liver deterioration, either through induction of fibrogenic mechanisms or via simplifying the negative effects of opportunistic infection or drugs used for prophylaxis or therapies [70,71]. TNF is producing T cells undergone apoptosis as a result of consecutive induction of ATM, p38MAPK, and p53, as well as upregulation of apoptosis-related proteins such as CASP8, TNFSF14, HGF, and TGFB1, resulting in T cell reduction and an impaired immune system that promotes COVID-19 severity and HIV pathogenesis [[72], [73], [74], [75]]. The activation of toll-like receptor (TLR) pathways contributes to the release of pro-inflammatory cytokines such as interleukin 1 (IL1), IL6, and tumor necrosis factor (TNF), as well as type 1 interferon, which are critical in the identification of viral particles and stimulation of the innate immune system. Numerous TLRs, such as TLR2, TLR3, TLR4, TLR6, TLR7, TLR8, and TLR9 are possibly relevant in COVID-19 infection and might be a target for regulating the infection in the initial phases of the disease [76], while TLR-pathogen associations may perform an intermediate function in controlling HIV-associated disease. Infectious agents promote proviral transcription by cytokine-mediated stimulation of the HIV long terminal repeat (LTR) via NF–B activation, which could lead to LTR-dependent signaling. Pathogen-associated molecular patterns (PAMP)-TLR association might result in MyD88-dependent induction of HIV transcription by NF–B or MAP kinase-dependent AP-1, in contrast to indirect cytokine-mediated HIV development [77]. TNF (Tumor necrosis factors) is significant in COVID-19 pathogenesis and are a powerful activator of HIV proviral transcription and mature virus generation [60,64]. TNF plays a key role in triggering the inflammatory cascade by synchronizing cellular recruitment via chemokines and cell adhesion molecules, resulting in lower levels of circulating pro-inflammatory cytokines.

Such as IL-1, TNF, and IL-6, which help to promote the development of COVID-19 [78]. Furthermore, TNF provokes various genes, together with the HIV long terminal repeat, through a signal transduction pathway that causes nuclear translocation of NF–B, which is involved in improving the progression of HIV [79]. During SARS-CoV-2 infection, mitogen-activated protein kinases (MAPK) are vital for cytokine production, and rising pro inflammatory cytokine production via these pathways impair airway epithelial cells and alveolar tissues, leading to reduced ventilation, acute lung injury, and acute respiratory distress syndrome (ARDS). MAPKs affect immune cell activity through reacting to external stimuli such as viral infections, stress, and inflammatory cytokines [80], as well as the MAPK signal may directly modulate HIV-1 replication. Moreover, MAPK antagonists prevent HIV-1 infection by blocking the transcription factors that control HIV-1 LTR promoter activity [81]. Capase-3, Bax, and Bcl-2 levels were restored after apigenin strongly suppressed the activation/phosphorylation of MAPKs, reduced inflammatory cytokines (TNF-α, IL-1β), and decreased oxidative stress via upregulating antioxidant enzymes [82,83]. EP300, also known as p300 [84], is a cellular acetyltransferase implicated ineffective viral integration that modulates chromatin conformation through histone acetylation, as well as acetylates and regulates the activity of viral protein integrase, resulting in increased integrase affinity for DNA and enhancing the protein's DNA strand transfer activity in both COVID-19 and HIV [85,86]. According to Y. Zhang et al., Apigenin inhibits the transcription of EP300/CBP-associated factor (PCAF) in NPC (nasopharyngeal carcinoma) cells when administered in vitro [87]. RelA/p65, which is essential in cytokine storms and promotes COVID-19 disease progression, is rapidly hyperactivated in COVID-19 patients, and the linkage between p65 and glucocorticoid receptor (GR) is crucial for target gene transcriptional regulation. RELA is a subunit of the transcription factor NF-kB, implying that the NF-kB signaling pathway is the most important therapeutic pathway for HSBD in the treatment of severe cases of COVID-19 [[88], [89], [90]]. In addition, RelA transcriptional activity is regulated by Tat; p65 expression in conjunction with Tat improves p65 binding to HIV-1 LTR, which corresponds with increased HIV-1 promoter activity [91,92]. Apigenin influences gene transcription by reducing the acetylation of RELA/p65 protein, which is crucial for cell survival, growth, and apoptosis [93,94]. Apoptosis inhibitor AKT phosphorylates the forkhead transcription factor (FKHR) family, glycogen synthase kinase 3 h (GSK 3 h), caspase-9, and Bad, causing these proteins to inactivate and unable to initiate pro-apoptotic pathways, hence supporting cell survival [79,95]. In the advanced stage of COVID-19 patients, AKT inhibition may reduce pathological inflammation, cytokine storm, fibroproliferation, and platelet activation associated with COVID-19 [70]. Similarly, HIV-1 infected macrophages have a better survival pattern as the AKT pathway is well-known for cell survival [96]. Apigenin, in particular, prevented AKT1 phosphorylation through the PDK-1 pathway [97]. Previous studies have shown that apigenin prevented the invasion and migration of colorectal cancer by preventing the phosphorylation of AKT protein [98].

Molecular docking is a useful method for discovering new medicines or improving existing drugs. For this purpose, docking technology confirmed that five key targets are indeed bound to apigenin. Apigenin-target docking scores were less than −6.5 kcal/mol based on the above-mentioned molecular docking results, indicating that apigenin may have strong binding activities to the receptor proteins, including MAPK1 and MAPK3, AKT1, RELA and EP300. Zhu YW et al. also identified most potential hub proteins including RELA, MAPK1 and MAPK3 through integrating network pharmacological methods on severe COVID-19 disease [90]. MAPK is a serine-threonine kinase that regulates cell development, differentiation, and cellular responses to cytokines and stress [99]. MAPK phosphorylates a variety of apoptotic regulatory proteins, and inhibiting MAPK1 activity reduces cellular growth and survival [100]. MAPK1 integrated into HIV virions may have a similar function in the early stages of cellular infection [101]. Vitamin C may potentially influence several key targets including MAPK1 in the SARS-CoV-2 virus infection [102], and it has been revealed that the anti-pneumonia activity of vitamin A is linked to selective inhibition of the MAPK signaling pathway [103]. Van Dijk D et al. found a MAPK3 protein by HIV-1 virus-host protein interaction network analysis which is known to have a role in a number of crucial cellular functions [104]. The MAPK3 protein is implicated in the inflammatory response to lung damage; according to Di Paola et al. studies indicate that inhibiting MAPK3/MAPK1 can lower the levels of proinflammatory cytokines including TNF- and IL-1 in lung injury [105]. Apigenin preserved.

Mitochondrial function by blocking ROS-induced p38 mitogen-activated protein kinases (p38-MAPK) and regulated apoptotic pathways [106]. Chugh P et al. and Gupta AK et al. findings suggest that Akt inhibitors might be a potential therapeutic for preventing the formation of long-lived HIV-1 infected storage [96,107]. Targeting AKT for COVID-19 appears to be a potential strategy because pharmacological inhibition of AKT has been shown to reduce angiotensin converting enzyme 2 (ACE2) production, a receptor for viral entrance into lung cells [108]. Apigenin has been shown to impact the PI3K/AKT/mTOR pathway, preventing AKT phosphorylation by disrupting the PI3K protein's ATP binding domain [109]. The AKT1 protein controls several cellular activities involved in lung cancer development and progression. By targeting the AKT1 protein, apigenin inhibited proliferation, migration, and invasion in the A549 human lung cancer cell line [110].

A flavonoid derivative with three hydroxyl substituents, i.e., apigenin, is produced as a secondary metabolite from several different plant species, including onions, maize, tea, rice, oranges, parsley, and so on, and has anti-oxidant, anti-inflammatory, and anti-cancer properties, as well as some molecular targets participating in inflammation [111]. Apigenin dramatically reduced the inflammatory and allergic responses of RAW264.7 and RBL cells, suggesting that it might be used as a preventive and therapeutic treatment for immune-related disorders [112]. Apigenin has attracted attention in recent decades because of its health-promoting benefits on cancer and inflammation. It has been found as a possible COVID-19 viral antagonist by specifically attaching to RNA polymerase and important viral proteins, notably the primary protease and the spike-RBD [22]. According to an in vitro study published by C. Sanna et al. apigenin possesses an anti-HIV-1 effect since it blocks the functions of ribonuclease H (RNase H) and integrase (IN) [21]. According to all information obtained now, apigenin appears to defend against COVID-19 synergetic effects on the progression of HIV. Apigenin has attracted the attention of scientists because of its low toxicity and various positive bioactivities. Apigenin is neither mutagenic nor hemolytic, according to the results of the Ames and hemolytic tests. K. Banerjee et al. mentioned that apigenin appears non-toxic in mammalian systems and is safe for intravenous administration [113]. As per the previous study, apigenin demonstrated the best absorption profile compared to the other flavonoids, with an approximate 30% human intestinal absorption (HIA) rate [114]. Before apigenin enters the systemic circulation and the liver, the gastrointestinal tract is crucial to its metabolism and conjugation [115]. The distribution and bioavailability of circulating flavonoids may be greatly influenced by the flavonoid extract made of apigenin glycosides, which was found in the blood, kidneys, intestine, liver, urine, or feces and distributed well in the tissues [18]. Apigenin may be rapidly metabolized by the liver, which would lower the body's effective therapeutic concentration. Phase I Enzymes were discovered to be involved in the metabolism of apigenin in the rat liver when NADPH, P450, or FMO were present [18]. It is a positive indicator that dietary apigenin is available for digestion by gut bacteria when it is excreted through the feces after oral consumption [116]. The flavonoids under consideration are not fetal, as per estimations of their acute toxicity, but they can still be grouped into toxic classes that are detrimental [114]. The oral availability of apigenin does not violate Lipinski's rule of five, indicating that the compounds have drug-like molecular composition [114]. In high-throughput screens, false positive results are produced by the promiscuous chemical class known as pan-assay interference compounds (PAINS). Even though PAINS are frequently described as reactive, nonspecific chemicals, new research has demonstrated that PAINS operate in noncovalent ligand–target interactions [117]. Utilizing the Protein-Ligand Interaction Profiler (PLIP) with default settings, the noncovalent interactions of each protein-ligand complex in the crystal structures were examined. When a ligand binds to a protein target, PLIP can distinguish between different forms of noncovalent interactions [118]. The activity of these PAINS classes is likely influenced mostly by noncovalent interactions. According to PLIP rule, apigenin can bind to the target proteins via maximum non-bond interactions. PAINS are frequently excluded from chemical libraries as non-druglike molecules. Lipinski's rule of five is also used to exclude substances that don't fit the description of a druglike substance [119]. In our study, apigenin showed zero violation of Lipinski's rule of five identified though SwissADME online server. In addition, the molecular weight of the PAINS determined their size. Apigenin showed a larger size of MW 270.24 (170 ≥ MW < 500) which indicates the classes of PAINS compounds.

In this study, we adopted network pharmacology and molecular docking techniques to evaluate the fundamental mechanism of apigenin for managing the synergetic effects of COVID-19 and HIV co-infection. Although the process by which apigenin exerts its therapeutic effects against COVID-19 to the progression of HIV is unknown, it is thought that apigenin's antioxidant and anti-inflammatory activities mediate them. Finally, to validate the compound of network pharmacology, pharmacokinetics prediction, including absorption, distribution, metabolism, elimination, and toxicity, was used to detect apigenin's effects in the human body. Apigenin's fast absorption in the intestine and delayed excretion enhances its bioavailability, making it a potential therapeutic agent [18]. We found apigenin shows better pharmacokinetic.

Properties with minor adverse effects like toxicity to the liver [120]; this backed up the network pharmacology findings and the molecular docking experiment. The complexes of these substances were then submitted to molecular dynamics simulations, and the outcomes were compared to those of the apo form of apigenin and 4 receptors complex. By analyzing the RMSD, RMSF, Hydrogen bond, Rg, and SASA, the complexes were verified, and the lead phytochemical complexes were discovered to be stable throughout the simulations. Additional research may be done to clarify the molecule's function, and apigenin may be a good option for pharmaceutical development. Therefore, effective experimental research is needed to be carried out to support and expand the findings to evaluate the potential therapeutic targets and apigenin may reduce the synergetic effects in HIV and COVID-19 infection progression.

5. Conclusion

Notably in nations with high HIV incidence, to reduce the transmission of COVID-19, it is important to focus on prevention strategies. People living with HIV are more susceptible to the COVID-19 pandemic's direct and indirect impacts. To control illness and reduce the synergetic effects of COVID-19 on the progression of HIV, effective and reliable therapeutic agents are required immediately. Depending on bioinformatics and network pharmacology approaches, apigenin tends to mitigate the synergetic effects of COVID-19 on the development of HIV and hence it might be used therapeutically to manage COVID-19 and HIV. The current findings suggest that apigenin may aid in the treatment of COVID-19 through controlling stress response, membrane receptor, and induction pathways, as well as apoptosis and inflammatory pathways. Moreover, the therapeutic significance of apigenin-treated COVID-19 and HIV was detected via molecular docking and molecular dynamics simulation studies. Apigenin's pharmacokinetic properties were investigated further, revealing a safe and effective therapeutic agent with a few adverse effects like hepatoxicity. Despite the shortcomings, the results proposed a treatment strategy for COVID-19 and HIV, establishing the basis for future clinical trials.

Availability of data and material

The datasets used or analyzed during the current study are available from the corresponding author on reasonable request.

Author contribution statement

Md. Arju Hossain, Md Habibur Rahman and Habiba Sultana: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper. Md Imran Hasan, Md Sohel and Saiful Islam Rayhan: Analyzed and interpreted the data; Wrote the paper. Asif Ahsan and Pratul Dipta Somadder: Interpreted the data; Wrote the manuscript. Md Habibur Rahman and Mohammad Ali Moni: Designed the experiments; Analyzed and interpreted the data; Wrote the paper and Supervised the whole project.

Declaration of competing interest

The authors declare no conflict of interest.