Integrated Mechanisms of Polarity–Based Extracts of Cucumis melo L. Seed Kernels for Airway Smooth Muscle Relaxation via Key Signaling Pathways Based on WGCNA, In Vivo, and In Vitro Analyses

The present study aimed to determine the mechanisms responsible for calcium–mediated smooth muscle contractions in C. melo seeds. The phytochemicals of C. melo were identified and quantified with the aid of Liquid Chromatography Electrospray Ionization Tandem Mass Spectrometric (LC/ESI–MS/MS) and high–performance liquid chromatography (HPLC), and then tested in–vitro and in vivo to confirm involvement in smooth muscle relaxation. Allergic asthma gene datasets were acquired from the NCBI gene expression omnibus (GEO) and differentially expressed gene (DEG) analysis, weighted gene co–expression network analysis (WGCNA), and functional enrichment analysis were conducted. Additionally, molecular docking of key genes was carried out. Kaempferol, rutin, and quercetin are identified as phytochemical constituents of C. melo seeds. Results indicated that C. melo seeds exhibit a dose–dependent relaxant effect for potassium chloride (80 mM)– induced spastic contraction and calcium antagonistic response in calcium dose–response curves. The functional enrichment of WGCNA and DEG asthma–associated pathogenic genes showed cytokine–mediated pathways and inflammatory responses. Furthermore, CACNA1A, IL2RB, and NOS2 were identified as key genes with greater binding affinity with rutin, quercitrin, and kaempferol in molecular docking. These results show that the bronchodilator and antidiarrheal effects of C. melo were produced by altering the regulatory genes of calcium–mediated smooth muscle contraction.


Introduction
Asthma is a chronic inflammatory disorder of the airways involving airflow obstruction and underlying inflammation. Its clinical symptoms depend on allergic status, age, and environmental triggering factors [1]. It affects about 300 million people globally, including 25 million Americans [1]. Asthma is characterized by severe airway structural alterations, including epithelial lining loss, hyperplasia in goblet cells, thickening of the basement membrane of subepithelial reticular, and enlargement of airway smooth muscle mass, contributing to airflow obstruction [2]. Airway hyper-responsiveness (AHR) is another clinical feature with increased sensitivity to inhaled constrictor agonists, resulting in stronger airway constriction in response to environmental stimuli [3]. It also shows an association with the histamine cytokines, i.e., interleukin-13 (IL-13), interleukin-4 (IL-4), 5

. Effects on Isolated Rabbit Jejunum Preparation
Seed kernel extracts of C. melo were tested for their effects on spontaneous spasms in jejunal preparations. Despite the ethanol extract displaying a spasmogenic contractile response, all subsequent extracts had a spasmolytic effect and inhibited spasms evoked by potassium chloride (K + ; 80 and 25 mM). A dose-dependent inhibition of spontaneous rhythmic spasms was found at 0.3 and 1 mg/mL for n-hexane (Cm-hexane), DCM (Cm-DCM), and aqueous (Cm-aqueous) extracts of C. melo seed kernels. The EC 50 was 0.1767 mg/mL (0.1244 to 0.2636; 95% CI), 0.2490 mg/mL (0.1873 to 0.3344; 95% CI), and 4.662 mg/mL (2.125 to 16.28; 95% CI), respectively. In contrast, ethanol (Cm-ethanol) extract of C. melo seed kernels did induce a contractile response in a jejunal preparation that was reduced by pretreatment with atropine (1 µM). An effective relaxation was observed at 1 mg/mL dose with an EC 50 Table 2). In both the absence and presence of C. melo seed kernel sequential extracts, the antagonistic response of calcium ion channels was determined, and the calcium DRCs were produced in cytosolic calcium-free jejunal preparation. However, Cm-hexane suppressed calcium contractile response DRCs only at 0.3 and 1 mg/mL, whereas all other extracts significantly suppressed calcium DRCs at 3 mg/mL. Although cholinomimetic bioactive substances may affect or alter Cm-ethanol effects, calcium DRCs in atropine-pretreated tissue displays a rightward shift with suppression of DRCs during atropine pretreatment [23]. Atropine showed no effect on calcium DRCs [24]. On the other hand, verapamil was used to confirm these findings of C. melo seed kernel sequential extracts. Verapamil reduced spontaneous contractions of jejunal preparations, potassium chloride (80 and 25 mM)elicited spasms at corresponding doses of 0.1, 1, and 0.1 µM with EC 50  to 0.05968; 95% CI), 0.1194 (0.09889 to 0.1445; 95% CI), and 0.01595 µM (0.01303 to 0.01956; 95% CI). Additionally, calcium DRCs were constructed, resulting in calcium DRCs blocked at a dose of 0.3 µM equivalent to C. melo seed kernel sequential extracts, indicating an antagonistic effect towards the calcium channels ( Figure S2, Table 2).

Effect of Cm-Ethanol on Isolated Rat Ileum Preparations
The bolus doses of Cm-ethanol were added to isolated rat ileum to induce spasms. We found that Cm-ethanol elicited more-significant contractile responses than acetylcholine (0.1, 0.3, and 1 µM) at 0.3, 1, and 3 mg/mL. With the presence of atropine (1 µM), Cm-ethanol spasms were decreased to a similar extent as those of Ach-induced spasm inhibition ( Figure 1, Table 2). A Student's t-test was performed to determine significance when compared to Ach: p < 0.05, * p < 0.01, ** p < 0.001, *** p < 0.0001, **** p < 0.00001 were used to determine significance.

Effect on Isolated Rabbit Tracheal Preparations
Sequential extracts of C. melo seed kernels were tested in rabbit tracheal preparations for their bronchodilator properties. The dilatory effect of C. melo seed kernel sequential extracts was dose-dependently induced by spastic spasms of potassium chloride (80 and 25 mM) and CCh (1 µM). Cm-hexane, Cm-DCM, Cm-ethanol, and Cm-aqueous re-  4). A Student's t-test was performed to determine significance when compared to Ach: p < 0.05, * p < 0.01, ** p < 0.001, *** p < 0.0001, **** p < 0.00001 were used to determine significance.

Effect on Isolated Rabbit Tracheal Preparations
Sequential extracts of C. melo seed kernels were tested in rabbit tracheal preparations for their bronchodilator properties. The dilatory effect of C. melo seed kernel sequential extracts was dose-dependently induced by spastic spasms of potassium chloride (80 and 25 mM) and CCh (1 µM). Cm-hexane, Cm-DCM, Cm-ethanol, and Cm-aqueous relaxed spastic spasms of potassium chloride (80 mM) at doses of 3, 5, 10, and 5 mg/mL, with EC 50 Figure S3, Table 2). Carbachol DRCs were formed in the absence and presence of Cm-DCM, Cm-ethanol, and Cm-aqueous (1 and 3 mg/mL), except Cm-hexane at doses of 0.3 and 1 mg/mL on tracheal preparation, and their suppression resulted in a non-competitive shift to the right. In tracheal preparations, verapamil inhibited spastic spasms of potassium chloride (80 and 25 mM) and carbachol (1 µM) in a dose-dependent manner at respective doses of 3, 0.3, and 0.1 µM with EC 50 of 0.2564 µM (0.2111 to 0.3118; 95% CI), 0.06765 µM (0.05113 to 0.09052; 95% CI), and 0.01278 µM (0.01017 to 0.01608; 95% CI). Verapamil inhibited DRCs for carbachol at a dose of 1 µM, similar to C. melo seed kernel sequential extracts ( Figure S3, Table 2).

Effect of Extracts on Castor Oil-Induced Diarrhea
A dose-dependent inhibitory effect was found for castor oil-induced diarrhea after the pretreatment of C. melo seed kernel sequential extracts ( Figure 2). C. melo seed kernel sequential extracts protect the animals from defecation. The percentage of protection of Cm-hexane, Cm-DCM, Cm-ethanol, and Cm-aqueous at a dose of 150 mg/kg was 55.28 ± 2.5, 45.22 ± 5.9, 35.36 ± 2.3, and 71.91 ± 4.1% and at a dose of 300 mg/kg was 78.33 ± 3.05, Figure 2. Gastrointestinal activities of in vivo activities of C. melo seed kernel sequential extracts. (A). Antiperistalsis, (B). antidiarrheal activity, and (C). GI fluid accumulation activity of verapamil, loperamide, and C. melo seed kernel sequential extracts. Data are expressed as the mean ± S.D. (n = 4). A Student's t-test (pairwise) was performed to determine significance when compared to control: p < 0.05, * p < 0.01, ** p < 0.001, *** p < 0.0001, **** p < 0.00001 were used to determine significance.

Effect of Extracts on Intestinal Fluid Accumulation
Results from this investigation ( Figure 2) indicated that, in contrast to the control group, castor oil treatment significantly increased gastrointestinal intestinal fluid accumulation (164.41 ± 2.6 g). On the other hand, C. melo seed kernel sequential extracts caused a dose-dependent significant (p < 0.001 vs. castor oil group) decrease in fluid in the intestine, with fluid weight 92.20 ± 4.4, 104.33 ± 6.6, 114.01 ± 3.7, and 91.96 ± 2.2 g, respectively, at the dose of 150, and 69.39 ± 5.10, 74.32 ± 3.40, 77.72 ± 3.40, and 65.77 ± 4.60 g at the dose of 300 mg/kg in comparison to loperamide (68.4 ± 4.3 g) and verapamil (73.65 ± 3.30 g).
2.6. WGCNA and DEG Studies 2.6.1. Data Preprocessing and Standardization GSE41649 and GSE15823, two genes associated with asthma phenotype, were obtained from the GEO database. GSE41649 included four healthy controls and four allergic asthmatic participants. Control and Asthmatic research groups expressed a total of 22,283 genes in GSE14649. A total of 12,625 gene expressions were obtained from GSE15823, including three research groups (control, asthmatic, and post-inhaled corticosteroids (ICSs)). Following unsupervised analysis, pairwise correlation was used for scaling the data. For further analysis, 11,564 and 7265 gene expressions were retrieved for GSE41649 and GSE15823, respectively, after removing genes with low expression using the coefficient of variance ( Figure 3).

DEG Identification and Enrichment Analyses
We analyzed GSE41649 and GSE15823 using Limma R for DEGs, identifying 202 and 223 important genes, respectively. Among the genes included in GSE41649, 109 exhibited upregulation, and 93 displayed downregulation ( Figure 4A). The gene expression levels of 129 genes were downregulated, while 94 genes were upregulated in GSE15823. Following the identification of the genes with the most significant differences between GSE41649 and GSE15823, hierarchical clustering was conducted ( Figure 4B).
Cluster-profiler R was also used to compare down-and upregulated genes in both datasets to determine GO and KEGG enrichment. For GSE41649 and GSE15283, Figure S5 shows the most significant GO BP and KEGG terms. GO BP terms for GSE41649 were calcium-mediated signaling, protein serine/threonine kinase activity, and inflammation regulation. In contrast, KEGG terms included inflammatory mediators of TRP channels, cholinergic synapse signaling, and smooth muscle contraction of vascular vessels. Active stimulation of glutamatergic, dopaminergic, and serotonergic synapses also induced inflammation, bronchoconstriction, and blood vessel contraction. GSE15283 up-and downregulated genes were significantly enriched in GO BP related to calcium ion transmembrane transport, membrane depolarization, and second messenger signaling. The KEGG pathways associated with up-and downregulation include GnRH secretion, serotonergic synapses, cholinergic synapses, glutamatergic synapses, and the JAK-STAT signaling pathway.

DEG Identification and Enrichment Analyses
We analyzed GSE41649 and GSE15823 using Limma R for DEGs, identifying 202 and 223 important genes, respectively. Among the genes included in GSE41649, 109 exhibited upregulation, and 93 displayed downregulation ( Figure 4A). The gene expression levels of 129 genes were downregulated, while 94 genes were upregulated in GSE15823. Following the identification of the genes with the most significant differences between GSE41649 and GSE15823, hierarchical clustering was conducted ( Figure 4B).   Cluster-profiler R was also used to compare down-and upregulated genes in both datasets to determine GO and KEGG enrichment. For GSE41649 and GSE15283, Figure S5 shows the most significant GO BP and KEGG terms. GO BP terms for GSE41649 were calcium-mediated signaling, protein serine/threonine kinase activity, and inflammation regulation. In contrast, KEGG terms included inflammatory mediators of TRP channels, cholinergic synapse signaling, and smooth muscle contraction of vascular vessels. Active genes for GSE15823 for WGCNA analysis. The cut height for GSE41649 and GSE15823 was set at 4.9 and 9.5, respectively, to omit the most noticeable outliers ( Figure 5A). A gene expression matrix was created when GSE41649 and GSE15823 samples were standardized, background-corrected, and polymerized. To build a gene co-expression network, we chose a soft threshold of β = 16 to verify the attributes of the scale-free network when the fitting coefficient R2 approached 0.90 ( Figure 5B).   A clustering dendrogram can be drawn by merging modules with high similarity using dynamic tree cutting and the hclust function. A clustering dendrogram ( Figure 6A) was drawn by combining highly similar modules with dynamic tree cutting and hclust. A total of 27 gene modules were found for GSE41649, while 9 were for GSE15823.

Correlation between Modules and Clinical Traits
The correlation results of GSE41649 and GSE15283 are illustrated in Figure 7, which shows the association between clinical features and module characteristics. According to the heatmap, modules that are positively correlated for control in GSE41649 were green-yellow (cor = 0.54; p-value = 0.17), pink (cor = 0.39; p-value = 0.34), midnight-blue (cor = 0.38; p-value = 0.35), and tan (cor = 0.37; p-value = 0.37). Conversely, the following modules showed GSE41649 had a positive correlation for asthma: yellow (cor = 0.85; p-value = 0.01), cyan (cor = 0.84; p-value = 0.01), light yellow (cor = 0.66; p-value = 0.07), and orange (cor = 0.59; p-value = 0.12). grey (cor = 0.48; p-value = 0.12), and red (cor = 0.41; p-value = 0.19). In contrast, the modules in GSE15283 positively correlated for the post-ICS-treated asthmatic group were brown (cor = 0.59; p-value 0.04), yellow (cor = 0.52; p-value = 0.08), and green (cor = 0.02; p-value = 0.94). 2.6.5. Hub-Gene Detection and Functional Pathway Enrichment Figure 8 illustrates the correlation between modules in the heatmap of eigengene adjacency. In GSE41649, the modules with positive correlation were cyan, yellow, light yellow, and magenta; in GSE15823, the modules were blue, yellow, pink, and black ( Figure  8A). Figure 8B illustrates the relationship between these four modules and gene expression. Mostly, genes have a positive correlation association. The modules in GSE15283 positively correlated for control were pink (cor = 0.66; p-value = 0.02), blue (cor = 0.62; p-value = 0.03), and black (cor = 0.47; p-value = 0.12), whereas the negatively significant modules were blue (cor = −0.28; p-value = 0.32), grey (cor = −0.21; p-value = 0.46), and green (cor = −0.08; p-value = 0.78). The modules in GSE15283 positively correlated for asthma were turquoise (cor = 0.77; p-value = 0.003), grey (cor = 0.48; p-value = 0.12), and red (cor = 0.41; p-value = 0.19). In contrast, the modules in GSE15283 positively correlated for the post-ICS-treated asthmatic group were brown (cor = 0.59; p-value 0.04), yellow (cor = 0.52; p-value = 0.08), and green (cor = 0.02; p-value = 0.94). Figure 8 illustrates the correlation between modules in the heatmap of eigengene adjacency. In GSE41649, the modules with positive correlation were cyan, yellow, light yellow, and magenta; in GSE15823, the modules were blue, yellow, pink, and black ( Figure 8A). Figure 8B illustrates the relationship between these four modules and gene expression. All modules were utilized to identify hub genes highly linked to asthma pathogenesis to analyze the relationship between gene significance (GS) and module membership (MM). We then calculated correlations between MM and GS and analyzed the data ( Figure  S6, Figure 9). It was determined that the most correlated modules for GSE41649 were cyan, yellow, light yellow, and blue with 98, 329, 57, and 535 hub genes, respectively. GSE15283 had four correlated modules: blue, yellow, pink, and black. These modules contained 429, 236, 63, and 78 hub genes. Functional pathway enrichment was conducted using Cluster Profiler R on both datasets of hub genes. The results of the functional enrichment of both datasets are depicted in Figure S7. All modules were utilized to identify hub genes highly linked to asthma pathogenesis to analyze the relationship between gene significance (GS) and module membership (MM). We then calculated correlations between MM and GS and analyzed the data ( Figures S6 and 9). It was determined that the most correlated modules for GSE41649 were cyan, yellow, light yellow, and blue with 98, 329, 57, and 535 hub genes, respectively. GSE15283 had four correlated modules: blue, yellow, pink, and black. These modules contained 429, 236, 63, and 78 hub genes. Functional pathway enrichment was conducted using Cluster Profiler R on both datasets of hub genes. The results of the functional enrichment of both datasets are depicted in Figure S7. The blue module of GSE41649 was enriched in the following GO biological terms: heart morphogenesis, arachidonic acid, and eicosanoid metabolism, and regulation of voltage-gated calcium channel signaling; for cyan, enriched terms were modulation of phosphatidylinositol 3 kinase signaling, and protein kinase B signaling; for light yellow, enriched terms were the carbohydrate catabolic process, cytosolic transport, and ADP metabolism; for yellow, enriched terms were apoptotic process-related negative modulation of cysteine-type endopeptidase activities and mitochondrial ATP synthesis coupled electron transport. The blue module of GSE41649 enriched for KEGG pathways was the glutathione metabolic pathway, calcium signaling pathway, IL-17 signaling pathway. For cyan, it was arginine biosynthesis. For light-yellow, it was the cholinergic synapse, HIF-1 signaling pathway, and ErbB signaling pathway. For the yellow module, it was vascular smooth muscle contraction, Rap1 signaling, estrogen signaling, and TNF signaling.

Hub-Gene Detection and Functional Pathway Enrichment
In GSE15283, the black and pink modules were enriched in domains concerning cellular compartments (CC) and molecular functions (MF) gene ontologies. The GO BP terms for the blue module of GSE15283 were immune cell activation, positive regulation of cytokine production, interleukin-12 production, and positive regulation of inflammatory response. The KEGG pathways for the blue module of GSE15283 include cytotoxicity induced by natural killer cells, Th1 and Th2 cell differentiation, primary immunodeficiency, the chemokine signaling pathway, and asthma.

GSEA Analysis of Common Genes
We performed GSEA analysis on the common genes of DEG lists and the top modules of both datasets to identify related gene expressions based on the DEG list. In GSE41624, genes were identified as related to GO BP stress response, calcium homeosta- The blue module of GSE41649 was enriched in the following GO biological terms: heart morphogenesis, arachidonic acid, and eicosanoid metabolism, and regulation of voltage-gated calcium channel signaling; for cyan, enriched terms were modulation of phosphatidylinositol 3 kinase signaling, and protein kinase B signaling; for light yellow, enriched terms were the carbohydrate catabolic process, cytosolic transport, and ADP metabolism; for yellow, enriched terms were apoptotic process-related negative modulation of cysteine-type endopeptidase activities and mitochondrial ATP synthesis coupled electron transport. The blue module of GSE41649 enriched for KEGG pathways was the glutathione metabolic pathway, calcium signaling pathway, IL-17 signaling pathway. For cyan, it was arginine biosynthesis. For light-yellow, it was the cholinergic synapse, HIF-1 signaling pathway, and ErbB signaling pathway. For the yellow module, it was vascular smooth muscle contraction, Rap1 signaling, estrogen signaling, and TNF signaling.
In GSE15283, the black and pink modules were enriched in domains concerning cellular compartments (CC) and molecular functions (MF) gene ontologies. The GO BP terms for the blue module of GSE15283 were immune cell activation, positive regulation of cytokine production, interleukin-12 production, and positive regulation of inflammatory response. The KEGG pathways for the blue module of GSE15283 include cytotoxicity induced by natural killer cells, Th1 and Th2 cell differentiation, primary immunodeficiency, the chemokine signaling pathway, and asthma.

GSEA Analysis of Common Genes
We performed GSEA analysis on the common genes of DEG lists and the top modules of both datasets to identify related gene expressions based on the DEG list. In GSE41624, genes were identified as related to GO BP stress response, calcium homeostasis, calcium-mediated signaling, MAPK, phosphatidylinositol 3 kinase, and positive regulation of ERK1 and ERK2. The GSEA KEGG included chemokine signaling, the chemical MAPK signaling pathway, Ras signaling pathway, Rap1 signaling pathway, calcium-mediated signaling pathway, and PI3K-Akt signaling pathway ( Figure S8).
In GSE15823, genes were identified as being related to GO BP: positive modulation of TNF production, the ERK1, and ERK2 cascade are regulated, stress response is mediated by G protein-coupled receptor signaling, actin polymerization or depolymerization is regulated, calcium ion transport is regulated, and apoptosis signals are regulated. As a result of KEGG GSEA, we identified interactions of cytokine-cytokine receptors, Rap1 and HIF-1 signaling pathways, cholinergic and serotonergic synapses, and calcium signaling pathways ( Figure S9).

Identification of Smooth Muscle Contraction Pathways
We used cross-functional pathway screening to enhance findings from both datasets related to calcium-mediated signaling and smooth muscle contraction. Using the hclust function, genes and functional terms obtained from a particular sample were examined for expression and clustering in various DEG and WGCNA modules. In Figure 10A, GO biological terms are expressed in various groups of GSE41649 and GSE15283. Several key mechanisms are involved in calcium-release channels, calcium ion transport, calciummediated signaling, actomyosin structure organization, calcium ion release, import into cell cytosol, and positive modulation of calcium-mediated signaling.

Identification of Smooth Muscle Contraction Pathways
We used cross-functional pathway screening to enhance findings from both datasets related to calcium-mediated signaling and smooth muscle contraction. Using the hclust function, genes and functional terms obtained from a particular sample were examined for expression and clustering in various DEG and WGCNA modules. In Figure 10A, GO biological terms are expressed in various groups of GSE41649 and GSE15283. Several key mechanisms are involved in calcium-release channels, calcium ion transport, calciummediated signaling, actomyosin structure organization, calcium ion release, import into cell cytosol, and positive modulation of calcium-mediated signaling.

Identification of Potentially Active Genes for Bioactive Compounds and Disease
A threshold score of more than 50 or a probability greater than 0.5 was used to extract the most likely target profiles for every bioactive molecule from the TCMSP, Swiss-target prediction, and Drugbank databases. A total of 417 genes were identified for C. melo seed kernel sequential extracts after removing duplicates. Respiratory-tract-disease genes with the key terms "asthma", "bronchitis", and "coughing" were retrieved from PubMed, DisGeNET, GeneCards, and OMIM databases. The target genes were rectified using Uniprot and then validated with the VarElect tool. These targets were used to identify key genes ( Figure 11).

Identification of Potentially Active Genes for Bioactive Compounds and Disease
A threshold score of more than 50 or a probability greater than 0.5 was used to extract the most likely target profiles for every bioactive molecule from the TCMSP, Swiss-target prediction, and Drugbank databases. A total of 417 genes were identified for C. melo seed kernel sequential extracts after removing duplicates. Respiratory-tract-disease genes with the key terms "asthma", "bronchitis", and "coughing" were retrieved from PubMed, DisGeNET, GeneCards, and OMIM databases. The target genes were rectified using Uniprot and then validated with the VarElect tool. These targets were used to identify key genes ( Figure 11).

Identification of Key Genes
The DEG list, bioactive compounds target genes, disease-associated genes, and the genes of the top four modules were intersected for comparison, and the common or overlapping genes were assumed as key genes for network construction ( Figure 11A). Results showed that GSE41649 has eight overlapping genes, including CACNA1A, PIK3R1, NOS2, PTGDR2, ALOX15, IGF1R, ABCC1, and IL2RB, whereas GSE15283 has four overlapping genes: IL2RB, CACNA1A, NOS2, and TBXAS1. The common genes in GSE41649 and GSE1482 were CACNA1A, IL2RB, and NOS2 and were considered as the most common key genes for both datasets. Genes from both datasets were pooled and considered for network pharmacology investigations.

Identification of Key Genes
The DEG list, bioactive compounds target genes, disease-associated genes, and the genes of the top four modules were intersected for comparison, and the common or overlapping genes were assumed as key genes for network construction ( Figure 11A). Results showed that GSE41649 has eight overlapping genes, including CACNA1A, PIK3R1, NOS2, PTGDR2, ALOX15, IGF1R, ABCC1, and IL2RB, whereas GSE15283 has four overlapping genes: IL2RB, CACNA1A, NOS2, and TBXAS1. The common genes in GSE41649 and GSE1482 were CACNA1A, IL2RB, and NOS2 and were considered as the most common key genes for both datasets. Genes from both datasets were pooled and considered for

Physicochemical Characteristics
The physicochemical properties of IL2RB, NOS2, and VGCAC1C were determined using ProtParam. These physicochemical properties of each protein are outlined in Table 3.

Validation of Homology Modeling
A BLAST search was conducted to identify the template of IL2RB, NOS2, and VG-CAC1C for homology models ( Figure 12). IL2RB had 61 percent identity, 39 percent gaps, and 1781 overall scores on PDB template 5M5E; NOS2 had 56 percent identity, 44 percent gaps, and 8238 overall scores on PDB template 1NSI; calcium ion channels had 71 percent identity and 29 percent gaps on PDB template 7UHF. Ramachandran plots were used to verify the accuracy of the protein model with amino acid distributions in phi and psi angles ( Figure 12A). As a result of Ramachandran plot analysis, the amino acid residues in IL2RB, NOS2, and VGCAC1C were distributed in favorable areas in 85.9%, 89.6%, and 88.8%, respectively. In comparison, 0.5%, 0.0, and 0.3% were found in forbidden or disallowed areas. To assess the model quality, we calculated the z-score of the protein structures based on the energy of separation between native and misfolded proteins. Based on the results, IL2RB, NOS2, and VGCAC1C protein models were found to have z-scores of 4-3.33, −9.47, and 7 ( Figure 12B). A model of high sequence similarity may be helpful for molecular docking studies if alignment is adequately performed.

Molecular Docking
Molecular docking calculations are helpful for estimating the position of a ligand within the binding site of a target protein. Using physical energy factors (such as solvation energy), docking models can be more accurate [23,25,26]. A docking study was conducted using the models IL2RB, NOS2, and VGCAC1C to evaluate the efficacy of bioactive compounds on target genes. Luteolin binds most strongly to NOS2 (∆G = −47.68 kcal/mol) and IL2RB (∆G = −38.97 kcal/mol), those that have the most negative Gibbs free energy (∆G) values in comparison to other compounds. Quercetin strongly binds to VGCAC1C (∆G = −40.37 kcal/mol) (Tables 4 and S5, Figure S10). dues in IL2RB, NOS2, and VGCAC1C were distributed in favorable areas in 85.9%, 89.6%, and 88.8%, respectively. In comparison, 0.5%, 0.0, and 0.3% were found in forbidden or disallowed areas. To assess the model quality, we calculated the z-score of the protein structures based on the energy of separation between native and misfolded proteins. Based on the results, IL2RB, NOS2, and VGCAC1C protein models were found to have zscores of 4-3.33, −9.47, and 7 ( Figure 12B). A model of high sequence similarity may be helpful for molecular docking studies if alignment is adequately performed.

Molecular Docking
Molecular docking calculations are helpful for estimating the position of a ligand within the binding site of a target protein. Using physical energy factors (such as solvation energy), docking models can be more accurate [23,25,26]. A docking study was conducted using the models IL2RB, NOS2, and VGCAC1C to evaluate the efficacy of bioactive compounds on target genes. Luteolin binds most strongly to NOS2 (∆G = −47.68 kcal/mol) and IL2RB (∆G = −38.97 kcal/mol), those that have the most negative Gibbs free energy (ΔG) values in comparison to other compounds. Quercetin strongly binds to VGCAC1C (∆G = −40.37 kcal/mol) ( Table 4, Table S5, Figure S10).
Based on the results, some ligands exhibit a strong binding affinity for the receptor, as indicated by their highest negative Gibbs free energy (∆G). In order to achieve equilibrium at constant pressure and temperature, a system must have a negative ∆G change and have reached equilibrium. ∆G is believed to be responsible for the stability of a proteinligand combination depending upon the strength of the interactions between the proteins Based on the results, some ligands exhibit a strong binding affinity for the receptor, as indicated by their highest negative Gibbs free energy (∆G). In order to achieve equilibrium at constant pressure and temperature, a system must have a negative ∆G change and have reached equilibrium. ∆G is believed to be responsible for the stability of a protein-ligand combination depending upon the strength of the interactions between the proteins and the ligands. Alternately, it may be defined as the affinity of the ligand for its acceptor [27]. The inhibitory constant (Ki) is the concentration of inhibitor required to cut the reaction's maximal rate in half [28]. Therefore, it helps determine whether an inhibitor is effective. In general, Ki determines how much substance is required to block a receptor's activity. Luteolin had predicted logKi values of −17.48 and −13.70 µM, respectively, for NOS2 and IL2RB. Quercetin had predicted logKi values of −14.30 µM for VGCAC1C.
The Coulomb binding energy, covalent binding energy, lipophilic binding energy, generalized born electrostatic solvation energy, and others are shown in Table S5.

Discussion
Since ancient times, herbal treatment as supplemental medicine or dietary regimens for treating various complex illnesses has been widely acknowledged as a safe and effective therapeutic. Cucumis melo L. seed kernels are an effective traditional healer and can treat various illnesses [29][30][31][32]. The objective of the current study was to look into the possible therapeutic uses of C. melo seed kernels for treating chronic disorders. This study investigated the motility of intestinal, tracheal, and urinary longitudinal smooth muscle strips in vitro and in vivo to determine how C. melo seed kernel sequential extracts regulate smooth muscle contractions [33]. We found that extracts of C. melo seed kernels modulate intracellular calcium homeostasis in smooth muscle contractions, thereby having anti-asthmatic effects.
Asthma is associated with increased smooth muscle mass due to hyperplasia and proliferation of smooth muscle, alterations in the extracellular matrix (ECM), and decreased apoptosis in ASM cells [34]. Th2 cells secreting IL-4, IL-5, and IL-13 play a key role in the inflammatory process in allergic asthma, smooth muscle spasms, and mucus production [35]. During pulmonary inflammation, IκB kinase phosphorylated NF-κB, producing iNOS and COX-2, which are inflammatory mediators [35]. As a result, NF-κB was partly responsible for the remodeling and inflammation of the ASM. In addition to signaling muscle contraction and metabolism, intracellular calcium contributes to cell motility, cell proliferation, protein release, and cell differentiation [34]. The development of asthmatic ASM phenotype and hyperresponsiveness has long been hypothesized to be associated with abnormalities in calcium homeostasis in ASM. In studies, rutin and quercetin significantly reduce eosinophilia and inflammation caused by cigarette smoke, while ovalbumin causes allergic asthma. Both rutin and quercetin reduced IgE levels, Th17 levels, iNOS levels, and cytokines, while stimulating IFN-γ levels [36,37].
These target genes directly or indirectly have a pivotal role in smooth muscle contraction with calcium-mediated signaling and smooth muscle contraction. In molecular docking, rutin and quercetin had a binding affinity with the L-type voltage-gated calcium ion channel (genes: CACNA1C, CACNA1D) and phosphoinositide phospholipase C (gene: PLCB, PLCE), myosin light chain kinase (genes: MYLK, MYL9), and calcium/calmodulin kinase (CAMK2B). Hence, it can be assumed that the strong antispasmodic activity of C. melo seed kernels was due to the strong binding affinity of quercetin and rutin towards target proteins, i.e., L-type voltage-gated calcium ion channel and phosphoinositide phospholipase C, myosin light chain kinase, and calcium/calmodulin kinase. It blockaded the signal transduction responsible for contraction. These proposed GSEA, GO, and KEGG pathways were verified with in vitro and in vivo experimentation. Both experiments were performed on the jejunum, trachea, and urinary bladder for the calcium signal transduction pathway on K + (80 mM) and calcium concentration-response curves.
We investigated the impact of calcium ions to determine how C. melo seed kernels extract alleviated muscle tension. A well-known physiological process, "excitation-contraction", is essential for maintaining intracellular calcium homeostasis, which is essential for relaxed and contracted smooth muscles [38]. Calcium entering the cytosol from extracellular fluid or calcium released from the endoplasmic reticulum causes the intestinal smooth muscles to contract spontaneously when calcium levels rise. Voltage-dependent calcium ion channels are vital in controlling intestinal motility [39]. C. melo seed kernel sequential extracts were tested for antispasmodic activity at isolated jejunum tissue contractions in a tissue organ bath. Results revealed that C. melo seed kernel sequential extracts ease smooth muscle spasms by reducing contractions and inhibiting calcium ions influx into the cell [40], except Cm-ethanol, which, in atropinzed jejunal preparation, relaxed the spontaneous tissue. The smooth muscle relaxant potential of C. melo seed kernel sequential extracts was further examined against Potassium chloride (80 and 25 mM) spastic contractions to ascertain whether it interfered with calcium ion channel inhibition or potassium ion channel activation. The potassium chloride (80 mM) triggered an intense depolarization of the membrane action potential of the cells, associated with increases in calcium inward current that produced a prolonged contractile response [41]. C. melo seed kernel sequential extracts repolarized this intense depolarization of membrane action potential to relax intestinal smooth muscle and obstruct the influx of calcium current. It was predicted that the following chronicle events would be interrupted [38,42]: (1) the concentration of cytosolic calcium ions was reduced due to a reduction in calcium inward current; (2) As there was insufficient interaction with regulatory protein phosphokinase C, ae calcium-calmodulin complex was not formed; (3) in the absence of calcium-calmodulin, myosin light chain kinase (MLCK) failed to activate, which resulted in a reduction in myosin light chain (MLC) phosphorylation, and; (4) this reduction decreased the availability of phosphorylated MLCs, resulting in the loss of myosin and actin filament crosslinking and the absence of a contractile response [43]. The pretreatment of jejunum tissue with C. melo seed kernel sequential extracts suppressed calcium DRCs at 1 and 3 mg/mL and induced a rightward shift comparable to verapamil [44], indicating that C. melo seed kernel sequential extracts behaved as calcium ion channel blockers and repolarized membrane action potential [41]. Amira et al. (2008) reported that six flavonoids (apigenin, genistein, quercetin, rutin, naringenin, and catechin) had an antispasmodic effect (at dose 30 µM) by regulating the gastric tone of the stomach. Thus, the antispasmodic and antidiarrheal activity of C. melo seed kernel sequential extracts may be attributed to the abundant presence of quercetin, epicatechin, rutin, and apigenin.
We investigated the antispasmodic properties of C. melo by inducing action potential depolarization in tracheal and urinary bladder tissue preparations in response to carbachol (1 µM) and potassium chloride (80 mM)-induced contractions [45]. C. melo seed kernel sequential extracts demonstrated dose-dependent relaxation of carbachol (1 µM) and potassium chloride (80 mM) elicited contraction, sustaining a repolarization state. Similar results for induced contractions were seen in C. melo seed kernel extract on the jejunal tissue preparation. The muscarinic M3 receptor activity of C. melo seed kernel sequential extracts on tracheal and urinary bladder preparations was established by the EC 50 value of carbachol (1 µM). It is known that muscarinic agonists (Ach, carbachol) cause muscular contractions by triggering the inflow of calcium ions into cells or intracellular cellular calcium release. Thus, we assumed that C. melo seed kernel sequential extracts reduced carbachol (1 µM)-induced contraction may be mediated through the calcium signaling pathway. The stimulation of the M 3 muscarinic receptor produced secondary regulatory proteins trisphosphate (IP3) and diacylglycerol (DAG), which are involved in the intracellular calcium-calmodulin complex. Therefore, muscarinic receptors share the same calcium-mediated signaling pathway for smooth muscle contractile response [41,46]-the pretreatment of C. melo seed kernel sequential extracts on tracheal preparations caused a rightward shift in carbachol DRCs, such as the verapamil validated hypothesis [47]. In addition, pretreatment with C. melo seed kernel sequential extracts on the urine bladder resulted in a rightward shift and calcium DRCs inhibition, which may indicate the function of calcium in the therapeutic action of sequential extracts of C. melo seed kernels. Therefore, C. melo seed kernel sequential extracts exhibited calcium-mediated signaling and smooth muscle relaxation.
Chan et al. [48] reported the anti-asthmatic activity of quercetin and rutin. This antimuscarinic response was produced due to the blockade of phosphoinositide phospholipase C (PLC), an active member in calcium mediates signaling. As previously shown in molecular docking, PLC showed a strong binding affinity with rutin (−37.75 kcal/mol) and quercetin (26.15 kcal/mol). Phosphoinositide phospholipase C, a key enzyme present under cell members, promotes signal transduction of M 3 muscarinic receptor to induce contraction in smooth muscles. The activation of the M 3 muscarinic receptor stimulates PLC, which hydrolyzes phosphatidylinositol 4,5-bisphosphate into two secondary messengers: inositol 1,4,5-trisphosphate (IP 3 ) and diacylglycerol (DAG). IP 3 stimulates inositol 1,4,5-trisphosphate receptors (IP 3 R) on the sarcoplasmic reticulum to release calcium ions, increasing cytosolic calcium levels. DAG, along with calcium, activates a regulatory protein kinase C (PKC, gene: PRKCA), through which phosphorylation of calmodulin occurs to form a calcium/calmodulin complex. This calcium/calmodulin complex activates another myosin light chain kinase (MLCK) that causes phosphorylation of myosin light chains (MLCs), phosphorylated MLCs, and actin, which form an interaction network to produce a contractile response [38,39,41]. Hence, C. melo seed kernel sequential extracts exert their bronchodilator and dysuria action by decreasing cytosolic calcium release from the sarcoplasmic reticulum and blockading the signal transduction of the muscarinic receptor pathway of contractile response.
Diarrhea is the abnormal expulsion of low-consistency stool due to disturbance in electrolytes and water transport in the intestine. Castor oil induces electrolyte and water transport changes in the intestine to cause diarrhea and increase peristaltic movements [49]. Our present study studied antidiarrheal, antiperistalsis, and fluid intestinal accumulation activities of C. melo seed kernel sequential extracts. The antispasmodic properties of C. melo seed kernel sequential extracts also inhibit the movement of charcoal meal at 150 and 300 mg/kg, producing a substantial decrease in peristalsis, diarrhea, and electrolyte imbalance. Castor oil contains ricinoleic acid [50,51], an active component that may enhance intestinal fluid and electrolytes [52] when the mucosa of the intestinal tract breaks it down. Subsequently, significant contractions occurred in the transverse colon and distal colon. C. melo seed kernel sequential extracts inhibited fluid secretion and castor oil-induced diarrhea [53] in animals. C. melo seed kernel sequential extracts exert antiperistaltic, antisecretory, and antidiarrheal effects mediated through calcium channel inhibition mechanisms, similar to loperamide and verapamil [54,55].

Preparation of Extract
In the summer of 2018, a botanist from Bahauddin Zakariya University, Multan (60000), Pakistan, verified and validated Cucumis melo L seed kernels and fruit obtained from locally framed fruits. A voucher specimen (Sp. Pl. 1011-1753) was placed at the herbarium of the Institute of Biology and Applied Sciences, Bahauddin Zakariya University, Multan (60000), Pakistan. The inner portion of the seed kernels was manually separated from the husk and processed into a coarse powder using an herbal grinder. Then, it was processed in a Soxhlet system using n-hexane, dichloromethane (DCM), ethanol, and water at their boiling points for 24 ± 2 h. Solvents were evaporated under reduced pressure at 35 ± 2 • C through rotary evaporation to produce yellowish-brown greasy extracts. The yields of Cucumis melo seed kernel sequential extracts, i.e., Cm-hexane, Cm-DCM, Cm-ethanol, and Cm-aqueous, were 41. 60, 9.31, 16.96, and 21.20 g, respectively, with an overall yield of 64%. Dried extracts thus obtained were stored at −20 • C in an amber glass jar and were later screened for their pharmacological properties. On the day of the experiment, distilled water or normal saline was used to dissolve and dilute the extract to the required concentration. Cm-hexane and Cm-DCM were dissolved in Tween-80 and diluted in normal saline or distilled water.

Chemicals
All analytical-grade solvents and chemicals used in this experiment were provided by Sigma Compounds Co., St. Louis, MO, USA.

Phytochemical Analysis of C. melo Seed Kernel Sequential Extracts
LC/ESI-MS/MS studies were conducted for the determination, characterization, and quantification of the constituents present in polarity-based sequential extracts of C. melo, following parameters and thresholds mentioned in one of our earlier reports [23,24].

Preparation of Samples
A 1.0 mL amount of 100 percent methanol (MeOH) was mixed with sequential extracts of C. melo (100 mg) to prepare samples for chromatographic analysis. We centrifuged the mixture for 14 ± 2 min at 14,000 rpm and then filtered the supernatant with a 0.22 µm syringe filter.

LC/ESI-MS/MS Analysis
C. melo seed kernel sequential extracts were investigated using LC/ESI-MS/MS to determine their phytoconstituents [56]. The sample was injected directly into an ESI probe for negative and positive ion modes. The mobile phases for Cm-ethanol and Cm-aqueous were made of Solvent-A, which contains 0.1% formic acid (FA) in MeOH, and Solvent-B, which contains 0.1% F.A. in acetonitrile (ACN) and water (20:80), respectively. The mobile phases for Cm-hexane and Cm-DCM were made of Solvent-A, which contains 0.1% FA in MeOH, and Solvent-B contains 0.1% FA in ACN, respectively. Other optimal chromatographic and mass spectroscopic parameters are detailed in earlier reports [23,24].

RP-HPLC Quantification and Method Validation
We performed HPLC analysis to confirm and quantify the phytoconstituents of C. melo seed kernel sequential extracts [24,57]. The mobile phases for Cm-ethanol and Cm-aqueous were made of Solvent-A, which contains 0.1% trifluoroacetic acid (TFA) in MeOH, and Solvent-B contains 0.1%canA with ACN and water (20:80), respectively. On the other hand, the mobile phases for Cm-DCM were made of Solvent-A, which contains 0.1% TFA in MeOH, and Solvent-B contains 0.1% TFA and ACN, respectively. Other optimal chromatographic and analytical method validation processes are detailed in our previous publications [23,24].

Ethical Committee Provision
The animal ethical committee of the Department of Pharmacology permitted the experimental methods for animal studies to vide E.C./04PhDL/S2018. The research was conducted following the Commission on Laboratory Animal Resources [58]. We used male and female albino rabbits (1.26-1.65 kg), Sprague-Dawley rats (155-180 g), and albino mice (17-26 g) throughout this study. Animals were housed in controlled conditions with a standard prescribed diet. For in vitro and in vivo testing, rats and mice were dislocated at the cervical spine, and rabbits were sacrificed with a sharp knife.

Isolated Tissue Experimentation for Smooth Muscle Contraction
For in vitro investigations, we followed the protocols of Saqib and Janbaz [41] and Wahid et al. [57]. We recorded the physiological response of the tissue with an isotonic and isometric transducer.

Isolated Rabbit Jejunum Preparations
Jejunum portions were dissected and placed in a tissue organ containing Tyrode's solution continuously recirculated with the carbogen gas at 37 ± 0.5 • C to keep them alive and ready for use under preload tension of 1 g. Before administering the C. melo seed kernel sequential extracts, the jejunum tissue was equilibrated for 25 ± 5 min. Test samples (extracts) were applied to the bath solution containing the jejunum tissue. The decrease in spontaneous rhythmic contractions for each tested sample was measured with repeated flushes of Tyrode solution every 10 min [47]. As reported previously, the calcium-mediated signaling pathways are involved in smooth contraction by triggering the calcium ion channel closing or the potassium channel opening [38]. The potassium chloride (80 mM) initiated the closing of the calcium ion channel, and the potassium ion channel was opened with potassium chloride (25 mM) [41,56]. Potassium chloride (80 mM) depolarized the cells and induced a calcium ion influx, which may alter smooth muscle contraction. The extract may be categorized as a calcium channel blocker if it inhibits these contractions [43]-the doses of C. melo seed kernel sequential extracts were successively administered to establish an inhibitory effect.
In addition, the calcium-antagonistic properties of C. melo seed kernel sequential extracts were examined. The preparation of jejunal tissue was pretreated with potassium chloride (80 mM), subsequently incubated in an EDTA Tyrode solution for 30 ± 7 min, and incubated for 55 ± 7 min in calcium-free Tyrode solution that is rich with potassium, to reduce or diminish intracellular calcium concentrations. Following incubation, doses of calcium were gradually added to produce superimposable control calcium dose-response curves (DRCs). Calcium DRCs were produced in Cm-hexane, Cm-DCM, Cm-ethanol, and Cm-aqueous extract pretreated jejunal preparations [43].

Isolated Rabbit Tracheal Preparations
We dissected the rabbit tracheal tissue, removed the adherent substances, formed the 2 to 3 mm wide rings, and each ring was incised longitudinally to produce sandwich smooth muscle tracheal strip preparation. Each preparation was placed in a Krebs bufferfilled tissue organ bath (pH 7.4) continuously circulated with carbogen at 37 ± 0.5 • C, under tension of 1 g. Before administering C. melo seed kernel sequential extracts, the tracheal preparation was stabilized for 55 ± 10 min. The bronchodilator activity of C. melo seed kernel sequential extracts was studied with potassium chloride (80 and 25 mM) and CCh (1 µM). A dose-dependent inhibitory response was established by cumulative administration of Cm-hexane, Cm-DCM, Cm-ethanol, and Cm-aqueous extracts. The preparation of tracheal tissue was treated with CCh (1 µM), then incubated for 55 ± 10 min in normal Kreb's solution. Following incubation, different concentrations of carbachol were gradually administered to the tissue preparation to form superimposable control carbachol dose-response curves (DRCs). The tracheal preparation was incubated with C. melo seed kernel sequential extracts for 55 ±10 min, and then carbachol DRCs were produced. To ascertain the presence of muscarinic receptor antagonism action, these carbachol DRCs of C. melo seed kernel sequential extracts were compared to control carbachol DRCs [43].

Isolated Urinary Bladder Preparations
We dissected the rabbit urinary bladder tissue, removed the adherent substances, and formed 2 to 3 cm wide strips. Each preparation was placed in a Krebs buffer-filled tissue organ bath continuously gassed with carbogen at 37 ± 0.5 • C, under tension of 1 g. Before administering the test drug, the tissue preparation was equilibrated for 55 ± 5 min. The dysuria activity of C. melo seed kernel sequential extracts was studied with potassium chloride (80 and 25 mM) and CCh (1 µM). A dose-dependent inhibitory response was established by cumulative administration of Cm-hexane, Cm-DCM, Cm-ethanol, and Cm-aqueous extracts. In addition, calcium antagonistic properties of C. melo seed kernel sequential extracts were examined. The preparation of bladder tissue was pretreated with potassium chloride (80 mM), subsequently incubated in an EDTA Tyrode solution for 30 ± 7 min, and incubated for 55 ± 7 min in calcium-free Tyrode solution that is rich with potassium to reduce or diminish intracellular calcium concentrations. Following incubation, doses of calcium were gradually added to produce superimposable control calcium dose-response curves (DRCs). Calcium DRCs were produced in Cm-hexane, Cm-DCM, Cm-ethanol, and Cm-aqueous extract pretreated bladder preparations [43]. 4.6. In Vivo Experimentation 4.6.1. Evaluation of Maximum Tolerated Dose C. melo seed kernel sequential extracts were tested to their maximum tolerated dose. Rats were orally given C. melo seed sequential extracts at doses of 50, 100, 150, 200, and 300 mg/kg every day for 28 days following the OECD guidelines, whereas the control rats were given saline. During the 28-day study, death rates, body weights, behavioral changes, and clinical signs of discomfort were recorded [59].

Protocol and Design
We randomly divided the mice, of either sex, into six groups with five animals each. Before the trial, water was freely available to all groups, but the food was withdrawn for 10 ± 3 h.
Group I was orally fed with normal saline (10 mL/kg). Group II was orally fed with castor oil or charcoal meal.
Group III and IV were orally fed with 10 mg/kg of verapamil and loperamide. Groups V and VI were orally fed with 150 and 300 mg/kg of Cm-hexane. Groups VII and VIII were orally fed with 150 and 300 mg/kg of Cm-DCM. Groups IX and X were orally fed with 150 and 300 mg/kg of Cm-ethanol. Groups XI and XII were orally fed with 150 and 300 mg/kg of Cm-aqueous.

Charcoal Meal GI Transit Test
In the antiperistalsis activity [41] of C. melo seed kernel sequential extracts, all animals received the charcoal meal orally after 15 min except Group I. Animals were sacrificed after 30 min, and their small intestines were excised to estimate the distance traveled by charcoal. The percentage of the peristaltic index was calculated [24].

Castor Oil-Induced Diarrhea
Castor oil-induced diarrheal activity [57] of C. melo seed kernel sequential extracts; all animals received the castor oil orally after 30 min except Group I. After 6-7 h, wet fecal spots were observed on white paper. The percentage of protection for defecation was calculated [24].

Castor Oil-Induced Intestinal Fluid Accumulation
In the antisecretory activity [44] of C. melo seed kernel sequential extracts, mice in all groups received the castor oil orally after 15 min except for Group I. Animals were sacrificed after 30 min and ligated pylorus and caecum ends of the small intestine before being removed. The weight of the small intestine with and without intestinal fluid was monitored. The amount of fluid accumulated (g) was determined by weighing the fluid accumulated in the intestine [24]. Asthma trait datasets GSE41649 and GSE15283 were acquired using the NCBI GEO (https://www.ncbi.nlm.nih.gov/geo/ (accessed on 22 June 2022)). The coefficient of variation was calculated after normalizing both datasets and eliminating the expression of genes with low expression levels.

Differentially Expressed Genes
In this study, differentially expressed genes (DEGs) from GSE41649 and GSE15823 were evaluated using the Limma package (v3.52) of R in the context of the following contrasts: bronchial asthma vs. control for GSE1484, asthma vs. control, and post-ICS vs. asthma for GSE15283 with log2 fold-change thresholds of ≥0.2 and 1, and p-value ≥ 0.05. The data were plotted using the R package ggplot2 (v3.3). In the following step, we determined the top 15 DEG genes and separated them into groups based on the hclust method.

Weighted Correlation Network Analysis
Identification of modules: We used the R package WGCNA (v1.70) to analyze GSE41649 and GSE15823 datasets for co-expression networks. The expressions of all genes were grouped using hclust to eliminate outliers. A Pearson test was used to create an adjacency matrix, and a paired gene expression similarity matrix was generated. The scale-free gene co-expression topological approach was used to evaluate the adjacency matrix at the lowest possible soft-threshold values. We constructed dissimilarity TOM (dissTOM) and topological overlap matrices (TOM) to assess gene expression profiles. A minimum module size of 30 was calculated to identify the co-expressed modules of genes. The modules with high similarity scores were combined at a threshold of 0.25.
Construction of module-trait relationships: We identified modules with a strong correlation to clinical characteristics following Module Eigengene (ME). Gene expression for each module was associated with its primary component as an eigenvector. An eigengene is the average gene expression pattern representing a module's central element.
Detection of hub genes: We analyzed gene significance scores (GS) to identify asthmaassociated gene expressions. Each gene expression was assigned a module membership (MM) to determine the correlation between ME and gene expression. The most significant genes in the modules are strongly associated with GS and MM, suggesting they are strongly associated with clinical characteristics. In addition, they can be used to identify module hub genes as well as create network diagrams.

Functional Annotation and Pathway Enrichment
The Cluster Profiler (v4.4.4) software was used to analyze functional gene ontology annotation and pathway enrichment from WGCNA modules and DEG genes of each dataset with an FDR ≤ 0.05 threshold.

Gene Set Enrichment Analysis (GSEA)
In both datasets, GSEA was performed to determine the biological significance of significantly overlapping gene expressions between the WGCNA and DEG genes. Gene expression levels were divided into low and high categories according to their expression levels. GSEA analysis of KEGG and GO biological processes was conducted with Cluster profiler software with an FDR ≤ 0.05 threshold.

Detection of Key Genes
We identified key genes based on the intersection of genes of bioactive substances, disease, WGCNA, and DEGs using the VennDiagram (v1.7.3) R package. The relationship between bioactive chemicals, illness, WGCNA, and DEGs was represented through a Venn diagram.

Construction of PPI, Bioactive Compounds, and Functional Enrichment Networks
Cytoscape 3.8.0 was used to construct and evaluate the bioactive compounds and disease-associated target genes (C-T-D), protein-protein interaction (PPI), bioactive compounds, disease-associated target proteins, and pathway (C-T-P) networks. PPI was constructed string plugin of Cytoscape.

Protein Homology Modeling
We used the Uniprot database to obtain high-quality protein sequences (https:// www.uniprot.org/, accessed on 27 August 2021). Uniprot protein sequences were blasted against one or more PDB templates to identify a three-dimensional homology model (Maestro v11.8, Schrodinger suite 2018-4). In brief, we initially evaluated the Uniprot sequence in the Extasy ProtParam tool (http://www.web.expasy.org/protparam, accessed on 30 August 2021) for physical and chemical characteristics [23,60]. Homology modeling involves the following steps: (1) import of Uniprot sequence, (2) BLAST the sequence in the PDB Database, (3) sequence alignment, and (4) model building. After modeling the protein, the model was tested and submitted for a protein reliability report. The Prime Refine-Loops and Prime Minimize modules were applied to refine the constructed model if protein reliability did not meet expectations. A Ramachandran plot using PROCHECK and SAVES v6 (https://saves.mbi.ucla.edu/, accessed on 30 August 2021) was used to validate the minimized protein model [61,62].

Molecular Docking
The various modules of Maestro and BIOVIA Discovery Studio 2021 were used to carry out molecular docking according to the previously mentioned parameters and thresholds [23]. Two-dimensional ligands from NCBI PubChem (https://pubchem.ncbi.nlm. nih.gov, accessed on 28 August 2021) were optimized, minimized, and ionized using the LigPrep module. The protein preparation module was then employed for assigning hydrogen bonds, minimization of het-states, reducing zero-order, disulfide bond formation, ionization, and optimization of 3D structures of the protein model. Furthermore, structural gaps were filled with the help of the prime tool, and het groups were protonated at a pH of (7.0 ± 2.0) using the Epik tool. At a pH of 7.0, PROPKA was employed for the optimal arrangement of hydrogen bonds in protein models, and OPLS3e was used for constraint energy minimization. To facilitate ligand interaction, the sitemap module was utilized to identify the coordinates for the protein binding pockets and the cubic grid box of the receptor grid generation module. The molecular docking was performed with Glide's extra precision using prepared ligands, proteins, and receptor grid files. The Epik tool was utilized to apply penalties to the docking score. The Prime MM-GBSA module calculated the binding energies of Glide ligand-protein complexes. As reported in earlier studies, the predicated inhibition constant (Ki) was calculated [61,62].

Software
Physiological responses of tissues were recorded using LabChart Pro 7. Molecular docking was carried out using Maestro v11.8 (Schrodinger suite 2018-4) and BIOVIA Discovery Studio 2021. The statistical analysis was conducted by GraphPad Prism 8 and RStudio (version 2021.02.3), and the network was constructed using Cytoscape 3.8.0.

Conclusions
In summary, LC/ESI-MS/MS and HPLC analyses showed the presence of stigmasterol, rutin, kaempferol, quercetin, apigenin, and luteolin in C. melo seed kernel sequential extracts. In addition, an integrated approach of DEGs and WGCNA of asthmatic patients was employed to find the probable regulatory gene network involved in calcium-mediated smooth muscles of the extracts. This investigation suggests that C. melo seed kernel sequential extracts exert antispasmodic, antiperistaltic, antidiarrheal, dysuric, and anti-asthmatic activities by modulating contractile response via a calcium-mediated signaling pathway to repolarize the membrane action potential of a smooth muscle cell. However, the presence of stigmasterol, rutin, kaempferol, quercetin, apigenin, and luteolin in C. melo seed kernel sequential extracts may be responsible for the bioactivities of the C. melo sequential extracts. However, more research is needed to understand and corroborate the putative modes of action and pathways responsible for the medicinal benefits of C. melo seed kernels.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ph15121522/s1, Figure S1: HPLC DAD-UV/Vis chromatograms of C. melo seed extracts at different wavelengths and standard addition validation chromatograms of bioactive compounds; Figure S2. In jejunal tissue preparations, C. melo kernel sequential extracts and verapamil inhibit smooth muscle contraction; Figure S3. A. In tracheal tissue preparations, C. melo kernel sequential extracts and verapamil inhibit smooth muscle contraction; Figure S4. A. In urinary bladder tissue preparations, C. melo kernel sequential extracts and verapamil inhibit smooth muscle contraction; Figure S5. Functional enrichments of up-regulated and down-regulated DEGs of datasets. Figure S6 Identification of potential WGCNA hub-genes; Figure S7. The functional enrichment of top four WGCNA modules of GSE41649 and GSE15283; Figure S8. GSEA of GSE41649 for GO and KEGG functional enrichment; Figure S9. GSEA of GSE15823 for GO and KEGG functional