The active compounds found in Tridax procumbens were identified through the Indian Medicinal Plants, Phytochemistry and Therapeutics, and KNApSAcK databases (Afendi et al. 2012) (Fig. 1 for compound structures). We conducted an extensive literature search via Google Scholar and PubChem using the search term “Tridax procumbens”.

Figure 1. 

Structure of Tridax procumbens containing phytochemicals.

A GeneVenn diagram was constructed to highlight the common molecular targets between the active compound’s gene and target genes for viral and bacterial diseases. This tool helps visualize the overlap and relevance of these compounds in combating various pathogens.

We specifically used the Swiss Target Prediction server to identify the proteins with which our compounds interact. This server compares 2D and 3D chemical structures to predict the possible targets for the used compounds. Furthermore, we relied on the GeneCards and the Online Mendelian Inheritance in Man database (OMIM)(Amberger et al. 2015) to identify target proteins associated with diseases. We used the Ven Plot Diagram method to look for common genes between our compounds and target disease genes to understand potential impacts better.

To identify the targets, we mapped protein-protein interaction: Gene interaction on the STRING database: solid circles represent the genes, and the interconnected structures represent the proteins, as illustrated in Fig. 5. Interaction by lines of different colors indicates biological processes. We used Cytoscape and its CytoHubba plug-in to view the interaction network easily. In this, compounds and protein interactions are represented by nodes and edges.

We used FunRich 3.1.3 to perform functional enrichment analysis of gene ontology, which includes studying biological pathways (BPA), cellular components (CC), and biological processes (BPR). This factor assists us in determining the role of genes in various contexts. Furthermore, we explored the Kyoto Encyclopedia of Genes and Genomes pathways for our chosen common gene. A p-value of less than 0.05 was used to determine statistical significance, indicating the importance of our gene ontology findings.

To evaluate how well the top target identified through the described topological analysis interacts with 29 selected bioactive compounds. As mentioned, three-dimensional structures of these compounds were retrieved from the PubChem database and then imported into Maestro, where they underwent energy minimization. As a result, a molecular database in.sdf format was created. Finally, the 3D structure of the TNF target protein was retrieved from the RSCB using PDB ID 2AZ5 (He et al. 2005), and preparation was conducted with the help of Maestro’s protein preparation wizard (Madhavi Sastry et al. 2013). The preparation process included the removal of irrelevant molecules and water molecules. Thereafter, with the Dock protocol, we docked compounds into the protein’s ligand area to determine the compounds’ interaction with the target protein.

We ran molecular dynamics simulations of this most promising molecule to know how it would behave under hypothesized biological conditions. In other words, these simulations show how the molecule behaves in the solvent environment. We simulated an orthorhombic box with a dimension of 12 Å on each side; the volume was optimized via the buffer size method. We used TIP3P and force-filed OPLS3e from Schrodinger Inc. standards for simulating proteins and ions16 for the standard water model. To mimic the physiological environment, we added sodium chloride (NaCl) to the system in a concentration of 0.15 M; we used sodium (Na+) and chloride (Cl-) ions (Harrach and Drossel 2014; Padhi et al. 2022). These simulations were run for 150 nanoseconds based on the Desmond molecular dynamics module, resulting in around 1500 snapshots showing the model behavior in the system. These were run under the NPT ensemble at a constant temperature of 300 K and 1 bar of pressure, ensuring the system was equilibrated and in the proper position during the simulation (Yu and Dalby 2020).

A total of 309 compound targets of Tridax procumbens containing phytochemicals were obtained from the Swiss Target Prediction server. The additional targets linked to bacterial and viral infectious diseases were sourced from the Genecard and OMIM databases. As shown in Fig. 2, the Venn diagram of viral infectious disease targets and bacterial disease targets depicts the overlap between 14,388 and 14,893 targets. The three datasets have 126 common gene targets of the compounds and the selected target diseases.

Figure 2. 

A Venn plot diagram highlighting overlapping targets between the potential compounds and disease-related genes.

To determine the biological mechanisms of Tridax procumbens phytochemicals against viral and bacterial disease, we performed GO enrichment analysis on 126 potential therapeutic targets of Tridax procumbens phytochemicals used in the treatment of using the DAVID database (Fig. 3). The study encompassed various GO categories, including cellular components, molecular function, biological pathways, and biological processes. We obtained the top 10 GO terms that were visualized in the bar plot based on their probability and percentage of genes involved in their terms. The graph represents the percentage of genes associated with each term and their enrichment’s statistical significance (p-value). The GO terms for cellular components highlighted several key locations where modulated genes were significantly enriched. The cytoplasm had the highest proportion of genes (60.3%), indicating a predominant localization and potential cytoplasmic functions. The plasma membrane also showed significant enrichment (42.9%), suggesting that Tridax procumbens may influence membrane-associated activities. Additionally, nucleoplasm (11.1%) and microsomes (6.3%) showed notable enrichment, though to a lesser extent. Other components, such as the extracellular space (9.5%), endoplasmic reticulum (16.7%), and membrane (7.9%), displayed moderate gene associations, indicating diverse cellular localizations affected by Tridax procumbens (Fig. 3). The molecular function GO terms identified several critical activities modulated by the treatment. Catalytic activity (17.5%) was significantly enriched, underscoring the influence of Tridax procumbens on enzymatic functions. Notably, protein serine/threonine kinase activity (6.3%) showed strong modulation (p < 0.001), indicating significant impacts on kinase activities crucial for various signaling pathways. Transmembrane receptor protein tyrosine kinase activity (6.3%) was also significant, suggesting an effect on receptor-mediated signaling mechanisms. Other activities, such as oxidoreductase (6.3%), DNA topoisomerase (1.6%), and hydrolase (4.8%) activities, showed lesser but notable enrichments. The pathway analysis revealed several significant pathways. The TRAIL signaling pathway recorded the highest enrichment, 46.8%, which indicated a possible role of Tridax procumbens in apoptotic signaling. The glypican pathway was 45.6% enriched, thus showing the possible role in cellular communication and signaling. 45.6% enrichment was also recorded for the IFN-gamma pathway, PAR1-mediated thrombin signaling, or thrombin/protease-activated receptor pathways, showing the possible immunomodulatory and thrombotic response of Tridax procumbens. The TNF receptor signaling pathway contributed 30.1% enrichment, showing the major inflammatory responses. Other highly enriched biological processes by Tridax procumbens include cell communication, 34.1%, the most significant process, demonstrating its effects on intercellular signaling (Fig. 3). At 26.2%, signal transduction was also highly significant among cell cellular signaling mechanisms. Other processes, such as energy pathways (30.2%); metabolism (34.1%); or protein modification (0.8%), demonstrate how the metabolic and functional configurations change under treatment. The G-protein coupled receptor protein signaling pathway at 10.3% shows the possible effects on GPCR-mediated signaling. However, the hormone metabolism, 1.8%, and the metabolism of numerous neurotransmitters, 1.6%, indicate additional regulatory roles in hormonal and neurotransmitter pathways.

Figure 3. 

Gene ontology categorization and pathway enrichment bar chart.

In conclusion, the results of the gene ontology analysis provide valuable information on the dual anti-viral and anti-bacterial activity of Tridax procumbens. The overwhelming contribution of the processes that occurred in the cytoplasm and plasma membrane indicates that the action of this plant is primarily aimed at the cytoplasmic process and the functions associated with the membrane. The localization of the processes corresponds to the potential ways of the plant’s influence on the cell, likely through membrane integrity and intracellular signaling pathways.

In protein-protein interaction analysis, we performed a network-based analysis to determine the main factors involved in the dual anti-viral and anti-bacterial activity of Tridax procumbens. We analyzed which metrics could be used, such as out-degree, in-degree, clustering coefficient, average clustering coefficient, and protein-protein interaction visualization. We also viewed several statistics on the number of nodes in the network, network density, and shortest path length. The out-degree distribution is presented in the top left corner, and it resembles the in-degree distribution shown in the top right corner. Both charts depict a right-skewed pattern, meaning that most genes have fewer connections than a few genes with a high number of connections. This enables us to suppose that it has a scale-free network property with a small number of hub genes that have a lot of interactions with other genes (Fig. 4). The highest out-degree is dashed in red and equal to 60, meaning it has the most outgoing degree of interaction. The highest in-degree is also dashed in red and amounts to 60 (Fig. 4). In the scatter plot in the bottom left corner, we plot the clustering coefficient as a function of the number of neighbors. As we can see, a clustering coefficient is more likely to increase as neighbors increase. This means that genes with more connections are more prone to clustering in densely knit groups. Hence, highly connected genes might be central to some subnetworks and be vital in keeping them fully functional.

Figure 4. 

Metric analysis from protein-protein interaction analysis.

Similarly, the average clustering coefficient on the bottom right chart decreases as neighbors increase. This means that although the highly connected nodes form a dense cluster, the overall network becomes less densely connected as it grows. It is possible that as more genes are included in the network, fewer new clusters of highly interconnected nodes are expected to form. The PPI network shows interaction between the top-ranked genes. The TNF node is the most top-ranked, with a score of 69. Other top hubs in the network include AKT1 with 64 and EGFR with 56, while SRC and ESR1 are equally ranked with a score of 55. The hub genes in the network have high connectedness points, making them at the center of the network. These hub genes are potential targets that Tridax Procumbens act upon to affect the anti-viral and anti-bacterial processes. The network statistics further expound on the structure and characteristics of the gene interaction network. The network has 125 nodes and 6 in diameter, with a density of 0.02 (Fig. 5A). The clustering coefficient, 0.275, reveals moderate clustering throughout the network. There are 3,537 shortest paths, and the characteristic path length is 2.206, suggesting a relatively shorter path between nodes, characteristic of small networks. The average number of neighbors per node is 14.624, revealing modest connectivity. TNF Tumor Necrosis Factor The high score indicates that the Tridax Procumbens act on the inflammatory response. AKT1, Protein Kinase B The high connectivity shows more centrality on the drug’s effect on the cellular system (Fig. 5B).

Figure 5. 

A. Interactions of gene targets were visualized using Cytoscape and network analysis; B. Top 10 Gene Target Interactions that was Visualized through Cytoscape.

Another is epidermal growth factor receptor (EGFR), which is highly involved in regulating cell growth, survival, proliferation, and differentiation. The high centrality of this node within the network structure indicates that Tridax procumbens may affect cell-signaling pathways related to growth and proliferation. SRC and ESR1 are similarly highly expressed genes (Wee and Wang 2017; Patnaik et al. 2022). Proto-oncogene tyrosine-protein kinase Src and estrogen receptor 1 (ESR1) participate in signaling pathways regulating cell division, survival, and differentiation (Roskoski 2015). Among other important genes whose inhibition could be targeted by the drug, I can mention PPARG, PTGS2, MMP9, GSK3B, and PARP1 (Fig. 5B). PPARG is PPAR-gamma involved in regulating the body’s energy balance and metabolism, whereas PTGS2 is normally expressed by the immune system cells in normal tissues and is induced to the intercellular space by bacterial lipopolysaccharides and cytokines (Ahmadian et al. 2013). Matrix metallopeptidase 9 (MMP9) is an enzyme from the matrix metalloproteinase family; GSK3B is a glycogen synthase kinase-3 beta involved in energy compromise between the metabolism of athletes and mental health in opioid addiction and schizophren (Pramanik et al. 2018). The final high-expressed gene is PARP1—Poly [ADP-ribose] polymerase 1, which removes loose strands of nucleotides within the DNA double helix before the individual strands are copied (Wang et al. 2019b).

The KEGG pathway analysis of the TNF signaling pathway depicts numerous critical components and interactions influenced by the tumor necrosis factor. It plays a crucial role in several biological processes, such as inflammation, immune response, and apoptosis, and is associated with several key nodes and molecular interactions. TNF binds to its receptors, TNFR1 and TNFR2, and by recruiting adaptor proteins like TRADD, TRAF2, and RIP1, triggers multiple downstream signaling events. The activation of the MAPK and the NF-kappa B signaling pathways is triggered. In the MAPK pathway, the intermediaries from the top are TAK1, MEKK1, and MEK3/6, which activate JNK and p38 MAP kinases, which in turn activate transcription factors like AP-1, modulating the expression of pro-inflammatory genes (Wang et al. 2019a) (Fig. 6). In the NF-kappa B pathway, the IKK complex is activated, leading to the phosphorylation and degradation of IκBα. This phenomenon allows NF-kappa B to translocate into the nucleus and promote transcription, thus increasing inflammatory cytokines, chemokines, and survival genes (Guo et al. 2024).

Figure 6. 

KEGG pathway analysis of TNF signaling pathway.

The TNF signaling pathway shows its bifurcation capacity to induce apoptotic or survival signals. TNF can stimulate apoptotic cell death when it agitates complex formation with FADD and caspase-8, while in contrast, it can also agitate activation of NF-kappa B, producing anti-apoptotic proteins to enhance cell survival (van Loo and Bertrand 2022). Moreover, this pathway links with other vital pathways, such as the PI3K-Akt pathway, which affects cell survival, and the JNK pathway, which affects stress responses. The pathway for TNF connects with inflammation and immune response, signaling since TNF executes a vast range of immune functions. TNF also boosts immune response by upturning inflammatory mediators such as IL-1 and IL-6, and subsequently, it upregulates infection or injury sites through the recruitment of immune cells (Alam et al. 2021) (Fig. 6). The TNF pathway is crucial as it facilitates the activation of NF-kappa B, which is necessary to express genes that participate in immune cell activation, differentiation, and survival. These genes include transmembrane proteins such as the adhesion molecules and cytokines necessary for leukocyte recruitment and activation (Webster and Vucic 2020). The pathway of TNF signaling has been studied with the aid of systems biology and molecular modeling to describe its potential in dual anti-viral and anti-bacterial agents. This signaling pathway has also been noted to promote the generation of anti-viral protein production while enabling immune cells to easily eliminate viral infection. It has been identified that the pathway enhances the phagocytic activity by promoting the responsive production of ROS and antimicrobial peptides targeting selective anti-viral infection and bacteria, allowing systems to obliterate disease-causing microorganisms such as bacteria by marepherages and neuophils (Chen et al. 2023; Liu et al. 2023).

In conclusion, Tridax procumbens has immense potential as a medicinal plant to modulate the TNF signaling pathway, offering therapeutic potential. The compounds derived from T. procumbens probably inhibit TNF signaling from excessively overactivating, preventing excessive inflammation and subsequent tissue damage. Some bioactive compounds from T. procumbens could enhance certain aspects of TNF signaling, allowing the body to respond effectively against pathogens. Understanding the interaction of Tridax procumbens with the TNF pathway can also guide the development of new biomimetics that could have dual anti-bacterial and anti- viral potential in a balanced manner to avoid side effects. More so, the findings show an interplay of several signaling pathways and molecular functions, which might interconnect to enable T. procumbens therapies. The network analysis suggests TNF, AKT1, EGFR, SRC, and ESR1’s centrality, the probable target, and interaction for the T. procumbens dual anti-viral and anti-bacterial potential.

The molecular docking study was conducted to examine the anti-bacterial and anti-viral action of the Tridax procumbens derivatives. The study primarily included target proteins, and TNF was the particular focus. Ten phytochemicals from Tridax procumbens were used in the study. The results confirmed that these phytochemicals had better binding affinities over others on the target genes. The reported 5,7,2,3,4-pentahydroxy-3,6-dimethoxyflavone 7-glucoside compound showed a binding affinity value close to the co-crystal ligand on the TNF gene. The docking score of the same was found to be -2.308. Cynaroside was observed to have a highly remarkable binding affinity value of -8.157 in docking, whereas the other compound 5,7,2,3,4-Pentahydroxy-3,6-dimethoxyflavone 7-glucoside also showed the highly prominent binding affinity value of -8.371. Moreover, isoquercitrin also had a significant docking value of -7.016 in the binding site (Table 1).

A more detailed picture of these interactions was obtained. Thus, three hydrogen bonds were observed for compound 5,7,2’,3’,4’-pentahydroxy-3,6-dimethoxyflavone 7-glucoside with the TNFG target gene: one bond with Gly148 and Gln149 through 4^th and 5^th OH groups on the tetrahydro-2H-pyran moiety and Leu120 with the OH group of the benzene ring (Fig. 7). Furthermore, Tyr59 forms a pi-pi hydrophobic interaction with the octahydro-2H-chromene moiety. Cynaroside forms four hydrogen bonds with the TNFG target gene: three between the 3^rd, 4^th, and 5^th OH groups of the tetrahydro-2H-pyran moiety with Tyr119 and Gly121 and one between the OH atom of the benzene ring and Tyr151 (Fig. 7). Furthermore, non-bonded hydrophobic interactions bond with Tyr59, Ala96, Leu94, Ile118, and Ile155. In contrast, the co-crystal ligand forms only one hydrogen bond between the OH group of the octahydro-2H-chromene moiety and Leu120 but also has non-bonded hydrophobic interactions with Tyr59, Ala96, Leu94, Pro117, Ile118, Tyr119, and Ile155, as well as polar interactions with Ser60, Gln61, and Ser95 (Fig. 7). The visualization of these data is presented in fig. 11 and shows the areas and types of interactions between these compounds and the TNF target gene.

Figure 7. 

2D and 3D visualization of compound-protein interactions for selected compounds (A. Image shows compound for 5,7,2’,3’,4’-Pentahydroxy-3,6-dimethoxyflavone 7-glucoside; B. Image shows compound for cynaroside, and C. Image shows compound for co-crystal ligand) with the protein.

In order to better understand the interaction of the compounds with the target protein, we generated a 150 ns molecular dynamics simulation using DESMOND software. The compounds studied here are TNF-bound with 5,7,2’,3’,4’-pentahydroxy-3,6-dimethoxyflavone 7-glucoside, TNF/cynaroside, and TNF in a co-crystal ligand. The parameters followed in this study are RMSD, root-mean-square fluctuation, and protein-compound interactions.

For compound cynaroside, the RMSD values of the C-alpha atoms within the protein complex stabilized at about 80 nanoseconds and became stable at 0.8 Å throughout the rest of the simulation (Fig. 8A). Compound 5,7,2,3,4-Pentahydroxy-3,6-dimethoxyflavone 7-glucoside first took about 80 nanoseconds to equilibrate, remaining steady in the binding pocket until the end of 150 nanoseconds. It started stabilizing after 50 nanoseconds when the values fell below 1.2 Å, indicating a solid bond to the protein site. On the other hand, the RMSF values of 5,7,2,3,4-Pentahydroxy-3,6-dimethoxyflavone 7-glucoside and cynaroside indicated few fluctuations, but those present were minor and within the range of 3.0 Å. The low values at the binding site revealed that the protein had a stable interaction with cynaroside and 5,7,2,3,4-Pentahydroxy- 3,6-dimethoxyflavone 7-glucoside. The secondary structure of the protein indicated that the alpha-helices and beta-strands constituted about 17.88% of the protein. The high proportion of these structures in the protein contributed to its stability and function (Fig. 8B).

Figure 8. 

A. The hit compound cynaroside targeted the TNF protein complex using molecular dynamics (MD) simulation evaluation. The top left image shows the root mean square deviation (RMSD), with the RMSD of proteins in blue and the compound in red. The up-right image presents a heat map analysis for protein-ligand complexes. The middle-left image represents individual amino acids’ root mean square fluctuation (RMSF) within proteins. The bottom image has a two-dimensional (2D) interaction diagram that maps interactions between the compound and the protein; B. The hit compound 5,7,2,3,4-Pentahydroxy-3,6-dimethoxyflavone 7-glucoside targeted the TNF protein complex using molecular dynamics (MD) simulation evaluation. The top left image shows the root mean square deviation (RMSD), with the RMSD of proteins in blue and the compound in red. The up-right image presents a heat map analysis for protein-ligand complexes. The middle-left image represents individual amino acids’ root mean square fluctuation (RMSF) within proteins. The bottom image has a two-dimensional (2D) interaction diagram that maps interactions between the compound and the protein; C. Co-crystal ligand analysis of the TNF protein complex through molecular dynamics (MD) simulation. In the top left picture, the root mean square deviation (RMSD) is blue for the protein and red for the ligand. The top right image shows a heatmap analysis of the protein-ligand complex. The middle left figure demonstrates root mean square fluctuation (RMSF) for amino acids in proteins individually. The lower image is a two-dimensional (2D) interaction diagram illustrating interactions between ligands and proteins.

Regarding interactions, 5,7,2,3,4-Pentahydroxy-3,6-dimethoxyflavone 7-glucoside formed three hydrogen bonds with TNF protein residues, including Tyr119, Gly121, and Tyr151. The role in hydrophobic interactions involved residues such as Tyr59, Ala96, Leu94, Ile118, and Ile155 (Fig. 8B). As for cynaroside, its simulation demonstrated the stability of RMSD values after five nanoseconds and continued until 80 nanoseconds. The values were stable at around 0.3 Å. In this case, the cynaroside compound was quickly stabilized within 80 nanoseconds. Next, it experienced fluctuations within 2.7 Å but stabilized at 1.2 Å after 130 nanoseconds. RMSF value for this C-compound was minimal, and enhancement in the protein loop and terminal shows 3.0 Å. The lesser RMSF at the binding site was between the compound and protein. The secondary structure analysis shows that alpha helixes and beta strands are counted in the 17.88% structure of the protein (Fig. 8A).

The interaction analysis for cynaroside showed three hydrogen bonds with residues Tyr119, Gly121, and Ser60, with the latter interacting through a water bridge with the OH group of the octahydro-2H-chromene moiety, as shown in figures (Fig. 8A). The co-crystal ligand stabilized in RMSD values after around 30 nanoseconds and maintained around 2.4 Å, as illustrated in figure. The ligand equilibrated in around ten nanoseconds and stayed stable in the binding pocket until 50 nanoseconds. After 50 nanoseconds, it fluctuated but remained within 2.4 Å until the end of the simulation. The RMSF values for the co-crystal ligand indicated minimal fluctuation in the protein’s loop and terminal regions within 3.0 Å, which was good, and the lower RMSF values in the binding site showed stable interaction. The secondary structure analysis showed that the alpha helix and beta strand were 17.95% of the protein. The interaction analysis for the co-crystal ligand showed one hydrogen bond with Tyr151, and the OH group of the octahydro-2H-chromene moiety contributed 30%, as shown in figure. The histogram plot in figure showed different types of interaction in the co-crystal ligand, such as hydrogen bonding, water bridges, and hydrophobic interaction. (Fig. 8C).