Mechanistic Insights on Dyspepsia Modulation by Salvia rosmarinus Through Network Pharmacology, Molecular Docking, and Molecular Dynamics

Abstract

Dyspepsia or indigestion is known to be a symptom of several gastrointestinal issues including gastroesophageal reflux disease (GERD) and gastric cancer. With the ever-increasing popularity of traditional herbal medicines, the research selected rosemary, or Salvia rosmarinus, due to its historical ethnopharmacological applications and ease of cultivation. Potential targets of molecules from S. rosmarinus are explored for the molecular pathogenesis of functional dyspepsia. Through a network pharmacology approach, it has been shown that there is a significant interaction between the disease and the plant’s compounds. The pathways involving the target genes of S. rosmarinus that are related to functional dyspepsia were revealed to be implicated in the development of certain diseases like gastric cancer and fibrosis which are both precursors to dyspepsia. Moreover, through molecular docking, the results of the pathway analyses were computationally validated indicating that the compound luteolin has the most significant interaction against dyspepsia-related genes. It effectively promotes apoptosis in cancer pathways, reducing the chances of gastric cancer carcinogenesis. To further validate these findings, molecular dynamics simulation was performed to compare the stability and binding behavior of the top-performing compound against the reference compound. The results of this study could be a possible basis in developing pharmaceuticals against gastrointestinal diseases, specifically, dyspepsia.

1. Introduction

Dyspepsia, more commonly known as indigestion, is a condition that inflicts discomfort or even pain in the upper abdomen [1]. Dyspepsia is usually associated with gastrointestinal afflictions and affects 20% of the whole populace who got diagnosed. Some examples of gastrointestinal disorders are gastroesophageal reflux disease (GERD), ulcers, and gastric cancer involving Helicobacter pylori [2]. In the Philippines, the risk factor for dyspepsia was found to be around 34% for people in general, while 26.5% to 79.9% for patients who have already experienced dyspepsia. Moreover, the reported risk factor was also found to be associated with gastric cancer [3]. Dyspepsia is used as a medical term to refer to the occurrence of stomach lining irritation caused by gastric acid [4]. Dyspepsia on its own does not necessarily require immediate attention, but due to how it affects the quality of life of several people, it is important to address it promptly to alleviate discomfort and prevent potential complications.

There are organic and non-organic causes underlying dyspepsia. Organic causes tend to be more easily identified due to their nature, which results from medical conditions and what a person consumes. If there is an imbalance in the production of gastric acid or enzymes, or if the stomach lining is irritated, it can lead to symptoms such as pain, bloating, or discomfort [5]. Gastric acid is highly impacted by the proton pumps on the surface of the stomach wall. They primarily pump hydrogen ions from the parietal cells to the stomach lumen. This affects the production of HCl, a main component of gastric acid. Proton pump inhibitors are used to control the amount of gastric acid in treating related medical conditions [6]. Non-organic classifications tend to have no specific causes and may be linked to the brain [7]. The gut–brain axis is a communication system connecting the brain to the gastrointestinal tract and vice versa and involves the microbiota in the gut [8]. Several studies indicate that depression and anxiety also contribute to dyspepsia. Frequent visits should prompt medical practitioners to also test for possible psychiatric comorbidities [9]. Dyspepsia is easily prevented through a change of diet and avoiding certain food types such as fats, alcohol, and caffeine. Antacids are also commonly used to neutralize stomach acid for relief. In terms of inhibitors, the widely used proton pump inhibitors are omeprazole and lansoprazole [10].

Omeprazole is a known proton pump inhibitor and has been used for gastrointestinal issues for decades [11]. It inhibits the proton pumps namely the hydrogen (H+)/potassium (K+) ATPase enzymes which are located at the parietal cells of the stomach lining. The role of these enzymes is to exchange g K+ ions in the stomach lumen with H+ ions from the parietal cell cytoplasm. Specifically targeting these proton pumps effectively reduces the amount of gastric acid made, lessening the irritation in the stomach [12]. It has been used for GERD and ulcers especially in the presence of dyspepsia as a symptom [13,14]. Lansoprazole has similar functions as a proton pump inhibitor assisting in the relief of gastrointestinal afflictions [15]. However, lansoprazole may have differing effects depending on the etiology of dyspepsia [16].

Currently, the use of herbal medicines is becoming more popular due to several factors including dissatisfaction with conventional medicines, the perceived safety of natural herbs, and its rising number of scientific validations, [17,18,19]. A study from Alsukaibi et al. presented the assessment of natural products from dates and their bioactivity through in silico analysis [20]. Relatedly, some plants have been long used as traditional medicines due to their ability to synthesize a large variety of organic molecules which could have a significant pharmacological effect, thus, making them important in the field of drug discovery [21]. Several regions in the world have a history of using additional medicines including China, India, Japan, Egypt, and the Mediterranean region [22,23]. One of the herbs that has been traced to have originated in the Mediterranean was S. rosmarinus or rosemary [24]. This plant was known to grow in mild, wet winters and dry summers [25]. Despite the conditions necessary for its growth, rosemary was still able to be cultivated in several parts of the world if it is in the right conditions [26]. The medicinal properties of S. rosmarinus, however, show potential as an anti-inflammatory agent or as an antioxidant. Additionally, it can even be used in aromatherapy [27]. S. rosmarinus was also reported to have antibacterial, anti-inflammatory, and spasmolytic properties [28]. Furthermore, a study conducted by Sereiti et al., discussed the uses of rosemary extracts for choleretic, hepatoprotective, and antitumorigenic activities [29]. Alongside, the in silico method for analyzing the potentials of phytochemicals have been recently becoming prominent. Several studies have utilized this method to identify the possible mechanisms and drug capabilities of compounds [30]. Additionally, a study from Yadav et al. utilized the same specific method of molecular docking to identify the drug-likeliness of CK2 inhibitors as a potential drug for COVID-19 [31].

To further explore the potential uses of S. rosmarinus, an in silico analysis using bioinformatics will be conducted. Moreover, this study aims to explore the bioactive compounds of S. rosmarinus as well as their potential targets, and to assess their viability in the treatment of dyspepsia through network pharmacology analysis. This study also aims to evaluate the binding affinities of the bioactive compounds from S. rosmarinus against the targets to be traced related to dyspepsia in order to understand the potential efficacy of the compounds as therapeutic agents. Nevertheless, this study aims to highlight the key pathways involved in the development of functional dyspepsia and its precursors.

2. Materials and Methods

This study was conducted in silico and was divided into several parts: starting with data mining and literature search for the bioactive compounds of S. rosmarinus up to validation using molecular docking. The step-by-step process of this study can be seen in Figure 1.

2.1. Screening of Bioactive Compounds in S. rosmarinus

The bioactive compounds of S. rosmarinus were enlisted from the findings of Sharma et al. [32]. Here, the samples were analyzed by ultra-high-performance liquid chromatography, electrospray ionization coupled with quadrupole-time of flight mass spectrometry (UHPLC-ESI-QTOF-MS). The compounds were filtered according to their physicochemical properties which were mainly and drug-likeliness using SwissADME (http://www.swissadme.ch/index.php, accessed on 22 July 2024) [33]. The bioavailability threshold was set to ≥ 0.55, while the oral bioavailability was set to OB ≥ 30%, similar to the standards set by existing studies [30]. The established cutoff for both criteria may indicate a higher probability that a greater amount of compounds from S. rosmarinus reaches the systemic circulation. This may also indicate that the desired therapeutic effects are achieved. Consequently, the thresholds reflect a balance between the absorption and distribution of the compounds [34].

2.2. Target Prediction of Bioacidtive Compounds from S. rosmarinus

The bioactive compounds that have met the criteria for absorption, distribution, metabolism, and excretion (ADME) were subjected to target prediction using SwissTargetPrediction (http://www.swisstargetprediction.ch/, accessed on 22 July 2024) [35]. This web-based server is capable of predicting the target of small molecules. The canonical smiles of each compound were obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/, accessed on 22 July 2024) [36]. Moreover, in predicting the targets of the compounds, the prediction was set to only “Homo sapiens” in order to narrow down the variation. The collected target genes were then filtered with a probability threshold ≥ 0.70.

2.3. Crossover of S. rosmarinus Targets and Dyspepsia-Related Genes

Using the GeneCard database (https://www.genecards.org/, accessed on 22 July 2024) [37], the genes related to dyspepsia were obtained. Here, the genes obtained were based of the relevance scores in the database. Identified disease-related genes were then filtered to top 25% by quantile using Python 3.12.4, and the Pandas 2.2.2 library. The top 25% results were selected to narrow down the set of genes involved and to ensure their relevance while having sufficient data. Consequently, the common genes from dyspepsia and S. rosmarinus were identified using the bioinformatics tool Venny 2.1 (https://www.biotools.fr/misc/venny, accessed on 22 July 2024) [38].

2.4. Construction of PPI Network

The protein–protein interaction network of the target genes from S. rosmarinus and dyspepsia was constructed using the STRINGdb (https://string-db.org/, accessed on 22 July 2024) [39]—a database for known protein to protein interaction. Here, the interaction was limited to “Homo sapiens”. The PPI networks were then sent to Cytoscape version 3.10.2 for further validation. The common genes were also filtered based on their maximal clique centrality (MCC), maximum neighborhood component (MNC), and degree using the Cytohubba plugin. This plugin is responsible for the identification of the important nodes in the network itself [40]. The same plugin was used to identify the core compound of S. rosmarinus based on its degree which is a metric to identify which node is the most biologically important. The analysis was set to the top five genes.

Moreover, the enrichment analysis of the network was done using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/, accessed on 22 July 2024) [41]. This tool is capable of identifying several biological meanings given in a set of genes. Nevertheless, the analyses were limited to the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Using the same tool, the biological processes, cellular components, and molecular functions of the common genes were identified. Using Enrichr (https://maayanlab.cloud/Enrichr/enrich, accessed on 30 June 2025), the protein where the top-performing ligand was bound, was cross-checked to further analyze its GO biological processes [42].

2.5. Compound, Target, and Pathway (CTP) Network Construction

Cytoscape (version 3.10.2) [43] was then utilized to construct the CTP network. The compound to target interaction was first constructed by loading the targets that were screened by probability ≥ 0.70 into the software and was merged into the pathways network where the top scoring connected the genes were accounted. On the other hand, the top compounds from S. rosmarinus were identified using the degree algorithm from CytoHubba in the same software.

2.6. Molecular Docking of the Key Compounds from S. rosmarinus and Dyspepsia

To validate the results of the network pharmacology analyses of dyspepsia and S. rosmarinus, molecular docking was done for the bioactive compounds of the plant against the dyspepsia-related genes. The control compounds used were omeprazole and lansoprazole which are both proton pump inhibitors and has an efficient binding activity against dyspepsia-related genes [44]. They are used as a basis due to their established therapeutic effects on dyspepsia-related afflictions and not directly due to their specific mechanisms. The SDF structures of the bioactive compounds (ligands) were obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/, accessed on 22 July 2024) [40] which were translated into PDB format using the chemistry toolbox Open Babel (https://openbabel.org/index.html, accessed on 22 July 2024) [45]. On the other hand, the structures of the dyspepsia proteins (target) were obtained from the Protein Data Bank (https://www.rcsb.org/, accessed on 22 July 2024) [46]. Consequently, the structures were prepared using the AutoDock tool where ions, waters, and hetero atoms were removed. Polar hydrogens were also added into the protein. Using the same software, the binding sites of the targets were also identified, specifically, by using the grid tool. Lastly, the molecular docking was done using the AutoDock Vina software (https://vina.scripps.edu/, accessed on 22 July 2024) [47]. The visualization of the protein-ligand complexes was done using the Biovia Discovery Studio (https://www.3ds.com/products/biovia, accessed on 22 July 2024) [48]. Prior to the visualization, the complexes were analyzed using UCSF ChimeraX (https://www.cgl.ucsf.edu/chimerax/, accessed on 22 July 2024) [49], where the proteins and ligands were loaded individually. Visualizing the complexes was able to give the results regarding their molecular interactions, which is also important to elucidate their possible mechanism of actions.

It is also worth noting that the protein target models were chosen according to the best resolution possible.

Table 1. Compound targets and their respective IDs.

Target	Symbol	PDB ID	Resolution
proto-oncogene c-Kit	KIT	8PQ9	1.70 Å
transthyretin	TTR	1BM7	2.00 Å
matrix metalloproteinase-9	MMP9	1L6J	2.50 Å
polyunsaturated fatty acid 5-lipoxygenase,	ALOX5	3O8Y	2.39 Å
72 kDa type IV collagenase	MMP2	7XJO	2.00 Å
xanthine dehydrogenase/oxidase	XDH	1JR0	2.70 Å

indicates the proteins’ respective IDs, as well as their resolution from the Protein Data Bank. Subsequently, the native ligands of the models were redocked to validate their RMSD. The same docking software was utilized as proteins and bioactive compounds. The RMSD of the native co-crystallized ligand and the redocked ligand were computed by aligning both the ligands in the PyMol version 2.5.8 software [50]. The RMSD of the redocked ligands are found in

Table 2. RMSD of the redocked ligands.

Protein	Native Ligand	RMSD
1JRO	Flavin-Adenine Dinucleotide	0.526
1BM7	2-[[3-(Trifluoromethyl)phenyl]amino] Benzoic Acid	0.000
7XJO	2-[3-(2-Hydroxy-1,1-Dihydroxymethyl-Ethylamino)-Propylamino]-2-Hydroxymethyl-Propane-1,3-Diol	1.277
8PQ9	2-[3-(2-Hydroxy-1,1-Dihydroxymethyl-Ethylamino)-Propylamino]-2-Hydroxymethyl-Propane-1,3-Diol	0.001

. Here, only four out of the six protein models were redocked with the native ligands as some only contain ions.

2.7. Molecular Dynamics Simulation

The molecular dynamics simulation was initiated through the preparation of the system using BioBB Python module v5.0.0 2024.2 [51]. It should be noted that the simulation workflow closely follows the protocol from a previously published study by Chen et al. [52]. A comparative dynamics simulation was done for the top-performing complex against the top performing control of the same protein which were KIT-luteolin and KIT-lansoprazole. The .pdb files of both complexes from the docking simulation were subjected to analysis, whereas the ligands and the proteins were isolated. Consequently, GAFF forcefield was applied to the ligands, while amber99sb-ildn forcefield was applied to the proteins. After the production of protein and ligan topologies, the complexes were remade and solvated using space as the water model. An octahedron 0.8 nm away from the system was then used as the solvent box, with Na and Cl ions acting as neutralizers for any protein charges present. Furthermore, the energy minimization was set to 5000 steps. The last step of the preparation involved the canonical ensemble (NVT) and isothermal-isobaric ensemble (NPT), having 5000 steps each at 10 ps.

Moreover, the simulation was conducted using Gromacs 2024.4 at 100 ns (50,000,000 steps). This software allows for the high-performance simulation of biomolecular systems such as proteins, lipids, and nucleic acids under physical forces [53]. The statistical analysis of the structural stability was done using the Kolmogorov-Smirnov test [54]. Here, the parameters that were considered include RMSD, radius of gyration, RMSF, and total number of hydrogen bonds. Lastly, the molecular dynamics simulation was validated through Cliff’s Delta analysis [55].

3. Results

3.1. Bioactive Compounds from S. rosmarinus

The screening of the 50 bioactive compounds from S. rosmarinus revealed only 10 compounds with the bioavailability score ≥ 0.55 and oral bioavailability ≥ 30%. Moreover, 8 out of the 10 compounds have also satisfied the Lipinski’s rule of five, indicating that these 8 compounds could possibly have good absorption occurrence due to the ratio of hydrogen bond donors to acceptors, molecular weight greater than 500, and a calculated partition coefficient greater than 5 [56]. The compounds were also screened according to other drug-likeliness scales including Ghose, Veber, Egan, and Muegge and was narrowed down to six compounds only which were luteolin, isorhamnetin, hesperitin, diosmetin, pectolinarigenin, and genkwanin. The following information on the compounds are found in

Table 3. Bioactive compounds of S. rosmarinus.

Compound	GI Absorption	Lipinski #Violations	Bioavailability	OB%
Quinic acid	Low	0	0.56	63.53
Caffeic acid	High	0	0.56	54.97
p-Coumaric acid	Low	3	0.85	43.29
Luteolin	High	0	0.55	36.16
Isorhamnetin	High	0	0.55	49.6
Hesperetin	High	0	0.55	70.31
Diosmetin	High	0	0.55	31.14
Pectolinarigenin	High	0	0.55	41.17
Genkwanin	High	0	0.55	37.13
Betulinic acid	Low	1	0.85	55.38

.

3.2. Predicted Dyspepsia-Related Genes and Target Genes of S. rosmarinus

Using the GeneCard database, there were a total 2832 dyspepsia-related genes obtained and was limited to 602 genes after getting the top 25% by quantile. On the other hand, the target genes of S. rosmarinus had a total of 1000 which was narrowed down to 63 after getting the targets with probability ≥ 0.70. After the crossover, it was found that 11 common genes were found: proto-oncogene c-Kit, transthyretin, matrix metalloproteinase-9, polyunsaturated fatty acid 5-lipoxygenase, 72 kDa type IV collagenase, xanthine dehydrogenase/oxidase, broad substrate specificity ATP-binding cassette transporter, glycogen synthase kinase-3 beta, poly [ADP-ribose] polymerase 1, ADP-ribosyl cyclase/cyclic ADP-ribose hydrolase 1, and multidrug resistance-associated protein 1. The crossover of genes was done through a venn diagram and is visualized in Figure 2. Moreover, the known functions of each traced proteins can be seen in

Table 4. Common genes of S. rosmarinus and dyspepsia.

Gene Symbol	Common Name	Known Functions	Reference
KIT	proto-oncogene c-Kit	Involved in several cell types’ proliferation, differentiation, migration, and apoptosis.	NCBI KIT proto-oncogene [57]
TTR	transthyretin	Transports thyroid hormone thyroxine and retinol- binding protein bound to retinol.	A Narrative Review of the Role of Transthyretin in Health and Disease [58]
MMP9	matrix metalloproteinase-9	Degradation of matrix and non-matrix proteins in the extracellular environment of the cell	Structure and function of matrix metalloproteinases and TIMPs [59]
ALOX5	polyunsaturated fatty acid 5-lipoxygenase,	Catalysis of leukotrines biosynthesis from arachidonic acid	5-Lipoxygenase: mechanisms of regulation [60]
MMP2	72 kDa type IV collagenase	Extracellular matrix degradation, angiogenesis, tissue repair, tumor invasion, inflammation, atherosclerotic plaque rupture	UniProt MMP2 [61]
XDH	xanthine dehydrogenase/oxidase	Catalyzes oxidation of hypoxanthine to xanthine and xanthine to uric acid	Redox signaling in cardiovascular health and disease [62]
ABCG2	broad substrate specificity ATP-binding cassette transporter	Involved in porphyrin homeostasis, protoporphyrin export mediator from mitochondria to cytosol and cytosol to extracellular matrix, and heme cellular export. Also involves possible H. pylori proliferation in gut microbiota.	Selective substrate uptake: The role of ATP-binding cassette (ABC) importers in pathogenesis [63]
GSK-3-β	glycogen synthase kinase-3 beta	An enzyme responsible for converting glucose to glycogen. Over-expression shows possibilities of gastrointestinal cancers which commonly have dyspepsia as a symptom.	Glycogen synthase kinase 3 beta: can it be a target for oral cancer [64]
PARP1	poly [ADP-ribose] polymerase 1	An enzyme involved in cellular processes such as DNA repair and cell death. Inhibition shows risk in dyspepsia as a side effect.	Combination of talaporfin photodynamic therapy and Poly (ADP-Ribose) polymerase (PARP) inhibitor in gastric cancer [65,66]
CD38	ADP-ribosyl cyclase/cyclic ADP-ribose hydrolase 1	Possible role in the immunology of stomach possibly H. pylori proliferation in gut microbiota.	CD38: a new paradigm in lymphocyte activation and signal transduction [67]
ABCC1	multidrug resistance-associated protein 1	an ATP-binding cassette transport protein which is an efflux pump uses ATP hydrolysis in exporting structurally diverse antineoplastxic agents in human cancers. Also involves possible H. pylori proliferation in gut microbiota.	Small-molecule inhibitors of multidrug resistance-associated protein 1 and related processes: A historic approach and recent advances [68]

.

3.3. PPI Network and Pathway Enrichment Analysis

The PPI network constructed on the STRING database had a total of 11 nodes and 20 edges compared to the expected number of edges of 5. On the other hand, the average node degree was 3.64, while the average clustering coefficient was 0.715. The PPI enrichment p-value was $5\times {10}^{-7}$, thus, indicating a higher protein interaction than what was expected. This also means that the proteins in the network are functionally related. On the other hand, since there were only 11 genes the application of different algorithms: MCC, MNC, and degree to filter out the hub genes was not necessary. The hub genes considered was that of the original 11 genes and can be seen in Figure 3.

3.4. Pathway and Process Enrichment Analysis

The enrichment analysis revealed 30 biological processes, 5 cellular components, and 17 molecular functions, and 13 KEGG pathways. Figure 4 shows that the biological processes are found to be concentrated on several immune responses to different triggers, as well as positive and negative regulations of several processes which indicate that S. rosmarinus can possibly control the triggers of functional dyspepsia, whether by up-regulation or down regulation of those that are involved in the development of the disease. On the other hand, the cellular components traced include most extracellular components as well as the nuclear envelope which may indicate that S. rosmarinus targets those that are outside of the cell components. Lastly, the molecular functions involved with the genes were very diverse—varying from transportation activities, enzymatic activities, and binding activities. Five of the 13 KEGG pathways traced from S. rosmarinus were found to be related in the functional dyspepsia. All of the pathways showed strong significance on the genes with a range of p-values from 0.025 to 0.075. A bar graph to visualize the terms and KEGG pathways can be seen in Figure 5. A cross-check analysis was conducted for KIT and its contribution to biological processes. Here, KIT was found to be highly significant in the Erythropoietin pathways, smooth muscle differentiation, dendritic and mast cells, sensory mechanotransduction, and megakaryocyte development. The biological processes where KIT is enriched are displayed in Figure 6.

3.5. Compound, Target, and Pathway Network

The CTP network revealed the necessary interactions within the bioactive compounds and the filtered targets. The compound and target network that was constructed can be seen in Figure 7. After the merging of the compound-target network with the pathway network, it was found that there were a total of 64 nodes and 96 edges, while the average number of neigbors was 3.082. Based on degree of association, it was revealed that the compounds with most significant interaction were luteolin, caffeic acid, p-coumaric acid, isorhamnetin, betulinic acid, hesperitin, pectolinarigenin, and diosmetin with degrees of 31, 14, 13, 6, 4, 1, and 1, respectively. These molecules are considered to be the core compounds of the plant as it indicates a greater interaction among others. It can also be suggested that these compounds will have a positive interaction with the genes related to dyspepsia.

3.6. Molecular Docking of Core Compounds of S. rosmarinus Against Dyspepsia-Related Proteins

The pathways and network elucidation of S. rosmarinus with functional dyspepsia revealed the top interacting compounds within the gene set and were found to be luteolin, caffeic acid, p-coumaric acid, isorhamnetin, betulinic acid, hesperitin, pectolinarigenin, and diosmetin—indicating that these are the core compounds from the plant. Here, the complete compound, target, and pathway network is visualized in Figure 8. Out of the 8 interacting compounds, five have satisfied the drug-likeliness screening with 0 violations from all the criteria which were Lipinski, Ghose, Veber, Egan, and Muegge. Similarly, these compounds were found to have high GI absorption and have passed the screening for bioavailability and oral bioavailability. The following compounds were luteolin, isorhamnetin, hesperitin, and pectolinarigenin, and diosmentin. In relation, an additional bioactive compound, genkwanin, satisfied the said criteria and drug-likeliness. These compounds were analyzed according to their binding affinity against the hub genes.

The results of the molecular docking showed that the compound with the highest binding affinity was luteolin and proto-oncogene c-Kit, with a docking score of −9.1 kcal/mol. Compared to the two controls, omeprazole and lansoprazole, luteolin exhibited a significantly higher docking score. The complete docking scores for all the protein candidates are visualized through a heatmap and can be seen in Figure 9. Moreover, the result of the docking indicates that luteolin may bind to the proteins in the pathways of dyspepsia, specifically, that of KIT. Another point of interest are the intermolecular forces acting upon the complexes. As seen in Figure 10, luteolin and KIT exhibited unfavorable acceptor-acceptor interaction, which nevertheless demonstrated key stabilizing interactions such as van der Waals, conventional hydrogen bond, and carbon hydrogen bond. These interactions suggest that the ligand retains the potential to bind within the binding sites despite the structural limitations. The models and diagram for the remaining complexes are found in Appendix A.

3.7. Molecular Dynamics Simulation Results

A comparative analysis on molecular dynamics of the KIT-luteolin (test) complex against the KIT-lansoprazole (control) complex was done to verify the nature of stability of the top-binding compound at 100 ns simulation. The results of the simulation revealed that there is a slight difference between the control complex and the test complex in terms of their stability. As seen in Figure 11 the KIT-lansoprazole complex exhibited a slightly less stable RMSD of 0.24 nm, while KIT-luteolin complex exhibited a slightly more stable RMSD of 0.22 after 20 ns. The protein-ligand residue fluctuations analysis (RMSF) on the other hand, revealed that the both complexes have localized flexibility, particularly at the N-terminal region for both alpha carbon (Cα) atoms only, and all backbone atoms considered. However, the KIT-lansoprazole complex showed a higher peak RMSF value as seen in Figure 9 and Figure 10, indicating that luteolin binding conferred higher restrains. Equally important is the radius of gyration analysis in which both complexes exhibited a closely related compactness where the KIT-lansoprazole yielded an average radius of gyration equal to 19.96 nm, while the KIT-luteolin yielded an average radius of gyration equal to 19.56 nm. Another factor is to take into account is the hydrogen bonding patterns. Throughout the 100 ns simulation, the KIT-luteolin complex consistently maintained a significantly higher number of hydrogen bonds, with an average of 2.26 and a maximum of 6 hydrogen bonds. Conversely, the KIT-lansoprazole complex formed virtually no stable hydrogen bonds, with an average of 0.00 and not exceeding a maximum of 1. This underscores the superiority of luteolin’s binding affinity and stability which is also facilitated by multiple and sustained intermolecular forces within the KIT binding pockets. All things considered, the molecular dynamics data point to Luteolin being a more favorable ligand for KIT, as compared to lansoprazole. The comparison of alpha carbons and all backbone residues for the KIT and lansoprazole, as well as KIT and luteolin complexes are visualized in Figure 12 and Figure 13, respectively.

It is also worth mentioning that Cliff’s Delta was applied to furhter evaluate the results of the comparative molecular dynamics between the test complext and control complex. The three key parameters were RMSD, radius of gyration, and hydrogen bonds. As seen in

Table 5. Cliff’s delta analysis summary.

Metric	Cliffs_Delta	Interpretation
RMSD (nm)	0.284337	small
Radius of Gyration (nm)	0.9473320343396936	large
H-Bonds	−0.99791	large

, the RMSD values were not large enough to imply a significant difference in te overall backbone stability between the two complexes. Despite this result, the radius of gyration and the hydrogen binds shows a strong favor to KIT-luteolin complex, with an enhanced overall compactness and a considerable higher amount of hydrogen bonds. This indicates that luteolin binding may lead to a greater conformational stability.

4. Discussion

Functional dyspepsia is a type of digestive disorder which can be persistent and will cause discomfort in the upper stomach without any cause [69]. Although the mechanism of development of dyspepsia is yet to be known until now, several studies have related the central factors that might trigger the occurrence of functional dyspepsia including stress and anxiety and is a precursor to gastroesophageal reflux diseases, peptic ulcer, gastric cancer, and gastroparesis [70,71]. Several drugs have been developed to treat dyspepsia and its related diseases which are mainly proton-pump inhibitors. Some of these include omeprazole and lansoprazole and are commercially available [72]. These drugs can both provide short reliefs to gastric symptoms of the stomach and can possibly pose side effects which is not efficient to persistent dyspepsia. With the rising opportunities for herbal medicines through several scientific validations, the molecules from S. rosmarinus have been analyzed for their possible drug-likeliness and interactions against functional dyspepsia.

The network pharmacology results revealed that the bioactive compounds of S. rosmarinus can interact with dyspepsia-related genes. Through the compound-target network, it was found that the compound with the highest degree of interaction was luteolin. Consequently, the molecular docking revealed that among all protein-ligand complexes, luteolin showed the most activity with a good binding affinity against most of the dyspepsia-related genes. Several compounds including isorhamnetin, hesperitin, diosmetin, pectolinarigenin, and genkwanin have also exhibited plausible interactions with spontaneous binding affinities against the disease, indicating that the herb can possibly treat dyspepsia. KIT-luteolin complex indicates the highest affinity even surpassing omeprazole and lansoprazole. This gives the possibility of further testing into possible applications for treating dyspepsia. Despite the presence of the unfavorable acceptor-acceptor bond, it is still worth highlighting the consistency of the other intermolecular forces including van de Waals, conventional hydrogen bond, and carbon hydrogen bond. Another point of interest is the molecular dynamics simulation conducted to validate the results of the molecular docking. Here, KIT-Luteolin complex was analyzed in comparison to KIT-Lansoprazole complex which is the control with highest binding affinity to KIT. The simulation revealed that Luteolin is a more favorable ligand, as opposed to Lansoprazole. The bioactive compound exhibited a more consistent hydrogen bonding, and more compact protein-ligand complex. The molecular dynamics simulation was then validated by Cliff’s Delta analysis where a small positive effect in RMSD, a highly negative effect in terms of radius of gyration, and a substantial amount of hydrogen bonds. These results further substantiate the potential of Luteolin as a ligand compared to Lansoprazole.

Luteolin, isorhamnetin, hesperetin, diosmetin, pectolinarigenin, and genkwanin are flavonoids which were studied to have potential therapeutic effects. Several studies discussed the various pharmacological activities of luteolin. One of which is its capability to produce anti-inflammatory effects through the inhibition of nuclear factor-kꞵ (NF-kꞵ) to reduce interleukin-8 mRNA expression in human intestinal Caco-2 cells, oxidative stress inhibition to activate the nuclear factor erythroid 2-related factor 2 (Nrf2) pathway, and regulating intestinal flora [73,74,75]. Moreover, isorhamnetin was found to exhibit various biological activities including anti-inflammatory and anti-gastric cancer effects [76]. In a study by Ricardo et. al., isorhamnetin was found to exhibit inhibitory effects on Tumor Necrosis Factor-ɑ (TNF-ɑ) and interleukin-6 (IL-6) [77]. Similarly, lipopolysaccharide (LPS)-induced models for inflammation studies revealed that hesperetin showed anti-inflammatory effects during treatments. A study by Yang et. al., presented that hesperetin metabolites are capable of downregulating LPS-induced prostaglandin E2 (PGE2) which is an important inflammatory product of COX-2 activity in smooth muscle cells. Additionally, hesperetin metabolites also exhibited suppression of NF-kꞵ in the same model for raw cells [78]. This indicates that hesperetin might also act as an inhibitor for the dyspepsia-related genes. Research indicates the use of diosmetin for treatment shows 90% reduction of growth due to inhibiting H. pylori by interfering with its biological functions [79]. In ethnopharmacology, pectolinarigenin has been known to have inhibitory effects which effectively treat various human ailments such as ulcers, stomach aches, and several more [78,80]. Genkwanin shows antagonist properties in inflammation which could possibly inhibit dyspepsia providing relief from the discomfort and pain [81,82]. The majority of the mechanisms of the six compounds were found to be mostly inhibition in terms of other diseases. This suggests that these compounds can possibly have inhibitory effects against dyspepsia-related genes in relieving intestinal discomfort.

The KEGG pathway analysis revealed that the predicted targets of S. rosmarinus are involved in several pathways. One of which is the metabolic pathway which is responsible for producing metabolites from breaking down of molecules [83]. In the context of functional dyspepsia, several metabolic pathways were found to be altered in an individual with functional dyspepsia. These metabolic pathways were said to be involved with energy metabolism, amino acid, gut microbiota, and host-co metabolism [84]. Furthermore, the gastrointestinal microbiota is related to the metabolic process in the human body, whereas, intestinal contents along with endogenous mucus secreted in the intestinal epithelium are used to produce metabolites including short-chain fatty acids (ScFAs), cholic acid, choline metabolites, phenols, lipids, and carbohydrates which can either be harmful or beneficial to the human body [85]. Thus, alterations in the pathway can contribute to the development of functional dyspepsia. Interestingly, this pathway exhibited a relationship with incredibly high number of genes with the lowest p-value compared to other pathways that were traced. There were also a plausible number of genes related to the nitrogen metabolism, most of which were carbonic anhydrases. This enzyme was known to form carbonic acid which can neutralize intracellular hydroxyl ions from luminal acid secretion [86]. Another pathway that was traced was the IL-17 pathway which is responsible in many inflammatory diseases [87]. Although the p-value of the genes in the pathway was the highest, however, a low p-value still suggests significance. The same pathway was also found to be associated with H. pylori infection where dyspepsia is once again a precursor by triggering gastrointestinal inflammation [88,89]. Furthermore, increased levels of IL-17 were also correlated with the symptoms of some gastrointestinal disorders including dyspepsia [90]. Nevertheless, all pathways that were traced were still significantly related to S. rosmarinus with smaller p-value, indicating that they are statistically significant despite the small count of genes involved [91].

Among the 13 pathways that were traced, five pathways were found related to dyspepsia-related genes. The top three pathways with lowest p-values were diabetic cardiomyopathy, pathways in cancer, and antifolate resistance. The diabetic cardiomyopathy pathway involves the matrix metalloproteinase enzyme which is involved in the growth factor beta (TGF-β) synthesis which is a master regulator of fibrosis. In relation, several studies suggested that the pathogenesis of dyspepsia can be found in the gastric mucosa in chronic gastritis and fibrosis [92]. Additionally, poly [ADP-ribose] polymerase 1 is involved in the inhibition of glyceraldehyde 3-phosphate dehydrogenase which is known for its primarily known for its role in glycolysis. This phenomenon suggests that GADPH might play a role in the regulatory functions of dyspepsia.

Moreover, the proto-oncogene receptor tyrosine kinase (KIT) gene was found to be involved in the MAPK signaling pathway. This pathway is responsible for the downregulation of Glial Cell Line-Derived Neutrophic Factor (GDNF)—resulting to the overexpression of microRNA-211 the expression of which in dyspepsia was found to be significantly increased [93]. The matrix metalloproteinases were found to be related to the sustained angiogenesis, which is closely associated with chronic inflammation due to H. pylori infection. Here, the upregulation of pro-angiogenic factors can lead to changes in the gastric mucosa, which can lead to dyspepsia [94]. The enrichment analysis of KIT through gene ontology biological processes also provides a valuable insight into functional dyspepsia. Among the most notable pathways are those associated with rythropoietin signaling, smooth muscle cell differentiation, immune cell activation, and mechanosensory perception which are all interconnected in the disruption of gut motility, immune regulation, and visceral sensitivity.

5. Conclusions

In light of these findings, it can be concluded that the bioactive compounds of S. rosmarinus can potentially treat functional dyspepsia. Results have shown that luteolin, isorhamnetin, hesperitin, diosmetin, pectolinarigenin, and genkwanin might pose possible pose possible effects on the dyspepsia-related proteins: KIT, TTR, MMP9, ALOX5, MMP2, and XDH. The pathway analysis revealed five pathways related to both the bioactive compounds of S. rosmarinus and dyspepsia-related genes. The top three pathways were found to be diabetic cardiomyopathy, pathways in cancer, and antifolate resistance. These path-ways were also found to be involved in the upregulation and downregulation of sever-al proteins involved in the pathogenesis of dyspepsia. The simulations also confirmed that luteolin that exhibited low binding energy requirement through molecular docking, validated by the strong interaction to KIT through molecular dynamics. Luteolin outperformed the controls, omeprazole and lansoprazole in most metrics, suggesting that it has promising potential as a natural therapeutic agent for treating functional dyspepsia. Substantiated by the bioavailability and oral bioavailability of the compound, luteolin may pose less risk and side effects than that of the existing drugs. While this study focused on the six bioactive compounds of S. rosmarinus, ABCG2, GSK-3-β, PARP1, CD38, and ABCC1 were also identified but not subjected to in-depth analysis. Nevertheless, this does not diminish the value of the present findings, rather, can be subjected to future studies to explore the full therapeutic potential of S. rosmarinus in functional dyspepsia.