Exploring the therapeutic potential of Rutin and Morin in Type 2 Diabetes: A transcriptomics and molecular dynamics simulation for proteins

doi:10.21203/rs.3.rs-5022136/v1

Type 2 diabetes (T2D) is a chronic, multifactorial disorder characterized by hyperglycemia or hyporinsulinemia. Despite numerous previous studies, its prevalence and associated complications continue to pose a significant global health concern. This study primarily focuses on identifying and investigating potential biomarkers and underlying molecular mechanisms that substantially impact T2D progression. A human microarray dataset (GSE20966 & GSE41762) exclusively containing pancreatic beta cells was sourced from the GEO database to facilitate the following research. The analysis of differentially expressed genes (DEGs) and the construction of co-expression networks were carried out using R and Bioconductor packages. The differentially co-expressed genes were further utilized to analyse hub genes and comprehensively characterize their functional importance by STRING, DAVID and ClueGo. The analysis conducted in this study emphasises the significance of seven hub genes (CEL, CPA1, CPB1, CTRB2, CEL3B, PLA2G1B, and REG1A) primarily implicated in T2D-associated molecular pathways such as pancreatic secretion, protein digestion and absorption and fat digestion and absorption. Among seven hub markers, direct scientific evidence underscores the role of PLA2G1B as a causative factor in the development of T2D. Therefore, further MD simulation for proteins study was conducted to comprehend the possible inhibition of the gene thereby reducing the severity of T2D. The study includes molecular docking and MD Simulation where it was established that phytocompounds (Rutin and Morin) for an instant posses higher binding affinity towards PLA2G1B in comparison to the standard inhibitors n-(p-Amylcinnamoyl) anthranilic acid. Thus, overall, it can be predicted that the flavonoids RU and MO could target PLA2G1B and might serve as a focused therapeutic approach for treating patients with T2D.

Microarray

MD simulation for proteins

Phytocompounds

Type 2 Diabetes

Docking

MD Simulation

Type 2 diabetes (T2D) is a chronic metabolic disorder characterized by insulin resistance or inadequate insulin production by pancreatic β cells (Staiger et al. 2009). The prevalence of diabetes among adults has increased from 8.3% in 2010 to 10.2% in 2021, affecting 537 million people worldwide. In comparison to high (12.2%) and low-income (11.9%) countries, middle-income countries are predicted to experience the largest relative increase in diabetes prevalence between 2021 and 2045 (21.1%) (Sun et al. 2022).

This metabolic disease involves a condition where the body develops a reduced sensitivity to insulin, resulting in cells not adequately reacting to insulin's presence. Consequently, elevated blood sugar resulting from this condition damages the pancreatic beta cells within the islets of Langerhans, subsequently leading to a decrease in insulin production. These specialized cell clusters play a vital role in maintaining blood glucose balance and producing insulin, which is crucial for cellular glucose uptake (Cárdenas-Aguayo et al. 2014; Ilie et al. 2013; Kerry et al. 2020). When blood sugar levels increase, like after eating, beta cells release insulin, which helps tissues like muscles and fat cells absorb glucose, lowering blood sugar. This strict glucose control prevents issues like diabetes. Therefore, grasping pancreatic beta cell physiology is crucial for diabetes management and understanding metabolic disorders (Dludla et al. 2023).

Further transcriptomics and

MD simulation for proteins is crucial since it sheds light on the underlying molecular pathways in understanding T2D. MD simulation for proteins studies look at protein variations and post-translational modifications, while transcriptomics explores gene expression changes in diabetes (Shaikh et al. 2021). Through integrating these approaches allow us to identify potential targets for drug development, early detection, and gain more detail molecular intricacies of these disease (Brendel et al. 2022; Shin et al. 2023; Shojima and Yamauchi 2023; Sohail et al. 2022; Tiwari et al. 2023; Wei et al. 2023). The regulatory factors that govern T2D include the interconnection between metabolic factors, genetic structure, and various environmental determinants (Kerry et al. 2020), where a large proportion of people living with T2D rely on injectable insulin. Recent years have seen notable progress in diabetes management, particularly with the emergence of inhibitors targeting T2D-associated proteins and enzymes. Among these, Sodium-glucose Cotransporter-2 inhibitors, which lower blood glucose by impeding kidney glucose reabsorption (Suzuki et al. 2022). Additional options involve Glucagon-Like Peptide 1 receptor agonists, boosting insulin production and aiding weight loss. Similarly, Phospholipase A2 (PLA2) inhibitors can ameliorate T2D-related issues like elevated proinflammatory markers (IL-6) (Batsika et al. 2021; Hinnen 2017; Hui et al. 2009; Wihastuti et al. 2019). This enzyme is secreted from pancreatic acinar cells into the digestive tract where it plays a vital role in the digestion of dietary fats (Hayashi and Dennis 2023; Vasquez et al. 2018). Thus, understanding the crucial significance of these enzymes in transcriptome-based research may help in unravelling intricate molecular mechanisms and signalling pathways associated with physiological and pathological processes of T2D. In addition, inhibition of PLA2 activity may have therapeutic potential in the treatment of diabetes (Kuefner 2021).

Previous reports suggests that plant extracts and compounds such as a menthoflavone, asiaticoside, and diosgenin have been found to demonstrate the ability to inhibit PLA2 activity (Kammoun et al. 2011; Tap et al. 2018). Research has demonstrated the noteworthy T2D inhibitory activity of the flavonoids morin (MO) and rutin (RU). These natural compounds, commonly found in various fruits and vegetables, exhibit the capacity to modulate glucose homeostasis and improve insulin sensitivity (Dubey et al. 2017; Kerry et al. 2023; Kerry et al. 2022). There are scientific evidence that demonstrates the enhanced cellular glucose uptake, reduce insulin resistance, and mitigate oxidative stress, all of which are critical factors in managing T2D (Kerry et al. 2023; Prince and Kamalakkannan 2006; Vinayagam and Xu 2015). Their potential as therapeutic agents in the management of this metabolic disorder has sparked considerable interest in further investigating their mechanisms of action and potential clinical applications (Prince and Kamalakkannan 2006). However, further research is needed to fully understand the role of PLA2 in diabetes and the potential therapeutic implications.

Therefore, the present study employs an integrated bioinformatics analysis to identify potential biomarkers by examining gene expression profiles at the transcriptional and protein level. The microarray dataset from GEO was retrieved to screen out differentially expressed genes (DEGs) between T2D and normal samples. The DEGs underwent weighted co-expression network analysis (WGCNA) to uncover co-expressed genes associated with T2D by correlating gene modules with the clinical trait. Subsequently, these genes were utilized to identify and elucidate hub genes within a protein-protein interaction (PPI) network, employing Database of Annotation, Visualization, and Integration Discovery (DAVID, https://david.ncifcrf.gov/tools.jsp) and Cluego plugin to gain insights into their functional significance. A validation step on the hub genes were conducted with an external dataset in conjunction with the docking and simulation study. The current study has the potential to improve complicated molecular mechanisms involved in T2D, leading to the comprehensive understanding of the diseases. Figure 1(Graphical Abstract) depicted the current study design.

2.1. In silico Transcriptomic Analysis

2.1.1. Data Acquisition

The Microarray dataset was retrieved from the Gene Expression Omnibus of National Centre for Biotechnology Information (GEO, NCBI, https://www.ncbi.nlm.nih.gov/). To gather the dataset, certain criteria were established: the dataset exclusively consisted of "Homo Sapiens", with a sample size larger than three, individuals with disease condition Type 2 Diabetes (T2D) and Non-Diabetes (ND) and lastly the dataset should specifically be focused on the “pancreas” tissue. Microarray dataset with the accession ID GSE20966 of the GPL135 platform and GSE41762 of the GPL6244 platform met the above criteria (Marselli et al. 2010; Tang et al. 2014). As per the requirement, GSE20966 encompasses 10 T2D and 10 Non-diabetics (ND) from pancreatic tissues. Likewise, GSE41762 contains 20 T2D and 57 ND patient samples (Tang et al. 2014). The raw expression matrix of both datasets along with the series matrix files was downloaded with the help of “GEOquery” package (v2.66.0) of R Bioconductor (Davis and Meltzer 2007).

2.1.2. Data pre-processing

Initially, the raw expression matrix of GSE20966 and GSE41762 have processed with Robust Multi Array (RMA) method applied from the “affy” package (v1.76.0) for background correction followed by quantile normalization and log₂transformation (Gautier et al. 2004). Probe annotation was done by converting probe IDs into gene symbol symbols with the help of the “u133x3pcdf” package (v2.18.0). In case of duplicate gene symbols with identical probe IDs, genes with the highest mean expression were selected and added to the final list of genes. For the subsequent analysis, GSE20966 was continued as a discovery dataset and GSE41762 was chosen as a validation dataset.

2.1.3. Differentially Expressed Genes (DEGs) Analysis

Normalized expression data were utilised to suspect DEGs among 10 T2D patients and 10 ND patients (T2D vs ND). The toptable function from the “limma” package (v3.54.2) (Ritchie et al. 2015) was applied to screen DEGs. This function results in a p-value, corrected p-value, log₂fold change (FC), and additional supporting parameters. Benjamini and Hochberg (BH) are one of the standardized methods employed to acquire the corrected p-value (adjusted p-value). A primary criterion of |logFC| ≥ 0.6 and a p-value < 0.05 was defined in order to screen significant upregulated and downregulated DEGs. For the visualization, a volcano plot and hierarchically clustered heatmap were created using “EnhancedVolcano” package (v1.16.0) and “pheatmap” package (v1.0.1) of R.

2.1.4. WGCNA network construction

The gene expression profiles of significant upregulated and downregulated DEGs were used to construct a weighted co-expression network. Outliers were detected based on Euclidian distance by hierarchical clustering of both sample conditions. To choose an appropriate soft threshold, pick Soft Threshold function of “WGCNA” package (v1.72.1) was applied to the normalized expression matrix (Zhang and Horvath 2005). In order to achieve a scale-free network, an approximate scale-free topology was calculated with soft power β=12 (R*2=0.85, slope= -1.57). An unsigned adjacency matrix was created by choosing top strongly connected genes. Later, the adjacency matrix was converted into a topological overlap matrix (TOM), and the distance between each gene and module was determined by calculating TOM dissimilarity matrix (DissTOM). The conventional hierarchical clustering approach effectively grouped all genes into modules, each consisting of at least 20 genes. A cut height of 0.25 was employed in conjunction with the dynamic tree-cut technique to merge modules made up of genes with similar expression patterns.

2.1.5. Identification of highly correlated modules

The Pearson correlation coefficient method has been employed to calculate the correlation among expression profiles with Module Eigengenes (MEs) which are termed Module Membership (MM). The above method was also applied to determine Gene Significance (GS) by calculating the correlation between each gene with T2D status as of clinical trait. Eventually, the p-value was assigned to each module-trait pair with the help of Fisher’s exact test. The value corresponds to a module-trait association close to 1 or -1 with a p-value < 0.001 is considered a highly correlated significant module.

2.1.6. Identification of hub genes network construction

To characterize hub genes within highly correlated modules associated with T2D, MM > 0.8 and GS > 0.5 were applied as cut-off criteria. Module hub genes were uploaded to the STRING (version 11.5, https://string-db.org/) tool for the construction of a PPI network with a medium interaction score of 0.40. For the visualization and hub gene identification within the PPI network, the TSV-formatted network file from the STRING was downloaded and imported into Cytoscape (Shannon et al. 2003). Subsequently, the top genes were ranked by four methods namely MCC, MNC, Betweenness and Degree with the help of the "Cytohubba" plugin (Chin et al. 2014).

2.1.7. GO/KEGG Pathways Enrichment Analysis

To thoroughly observe the module hub genes engaged in the Gene Ontology (GO) and pathways, a comprehensive enrichment analysis was established using DAVID. The functional enrichment analysis was performed based on the Gene Ontology (GO) and Kyoto Encyclopaedia of Gene and Genome (KEGG) pathway. The enriched genes that came under GO fall into three categories such as Biological Processes (BP), Cellular Component (CC), and Molecular Function (MF). The functional terms possessing p-value < 0.05 was regarded as statistically significant enriched GO term/pathways. Lastly, all the terms and pathways were visualised with Cluego plugin of the Cytoscape and “GOplot” package (v1.0.2) of R (Bindea et al. 2009; Walter, Sánchez-Cabo, and Ricote 2015).

2.2. In silico Proteomic Analysis

2.2.1. Software and program

For primary visualization and removal of impurities PyMol version 2.5.2 (DeLano Scientific LLC, Palo Alto, California, USA) was employed. For post docking visualization assembly of the docked complex Discovery Studio Visualizer version 21.1.0.20298 (Discovery Studio Biovia 2020) (Dassault Systèmes, San Diego, California, USA) were employed. In this study, for docking, AutoDockVina version 1.2.0 (The Scripps Research Institute, La Jolla, San Diego, USA) was taken as primary docking program (Trott and Olson 2010). Open BabelGUI version 2.3.1 was used for conversion of file to the required format. The ligand.PDBQT and protein.PDBQT file format and the grid box size were prepared and determined employing AutoDock Tools version 1.5.7 (ADT; Scripps Research Institute, La Jolla, San Diego, USA).

2.2.2. Preparation of ligand structure

Chemical structures of Rutin, Morin, and n-(p-Amylcinnamoyl) anthranilic acid (ACA) were downloaded in the Spatial Data File (.SDF) file format from the PubChem Compound Database (National Center for Biotechnology Information; https://pubchem.ncbi.nlm.nih.gov/). Donor-acceptor hydrogen-bonding interactions and molecular weights are important structural factors of protein targets and ligand-binding sites (Afriza et al. 2018; Lipinski 2016; Zhang and Wilkinson 2007). Babel was used to convert the 3D structures in.SDF format to.PDB format. ADT was then utilised to analyse ligand structures in requisites of Gasteiger change additions, nonpolar hydrogen combinations and rotatable bonds. Structures in the ligand.PDB format were then converted to the ligand. PDBQT format using ADT, enabling use with AutoDock4 (AD4) and AutoDockVina (Afriza et al. 2018).

2.2.3. Preparation of protein structure

The crystal structure of the human phospholipase-A2 (6Q42) was downloaded from the RCSB protein data bank (http://www.rcsb.org). All water molecules and attached salts and inhibitors by PyMol. Further, ADT software was used to produce the necessary data for AutoDockVina by assigning hydrogen polarities, addition of Kollman charge, calculating Gasteiger charges, assigning atom type to AD4 type and converting protein structures from the .PDB file format to .PDBQT file format.

2.2.4. Possible Ligand Toxicity Prediction

The potential toxicity profiles of the ligands were studied utilising the produced simple molecular-input line entry system (SMILES) code for each chemical structure. For each candidate, the toxicity profiles included hepatotoxicity, carcinogenicity, immunotoxicity, mutagenicity, cytotoxicity, toxicity class, and potential lethal dosage (LD₅₀) value using the ProTox programme was carried out (Banerjee et al. 2018a; Banerjee et al. 2018b; Drwal et al. 2014).

2.2.5. Docking Methodology

The AutoDockVina software was used for targeted molecular docking. Ligands were docked to the receptor separately using grid coordinates (grid centre) and grid boxes of varying sizes for each enzyme and receptor. The configuration file (config.txt) was prepared from the grid box file (grid.txt) where the box center (grid center) and size (npts) coordinates are present through ADT. On the basis of the AutoDockVina scoring system, negative Gibbs free energy (ΔG) scores (kcal/mol) were projected to represent ligand-binding affinities (Trott and Olson 2010). The visualization tools such as PyMOL and Discovery Studio Biovia 2022 were used to visualise the results of post-docking analyses, which revealed the dimensions and locations of binding sites, hydrogen-bond interactions, hydrophobic interactions, and bonding distances as interaction radii of <5 Å from the position of the docked ligand. Each compound was docked to the active site or binding site of the corresponding enzyme or receptor. The best and most energetically favourable conformations of each ligand were then chosen after binding poses of each ligand were visualized and their interactions with the proteins were defined.

2.2.6. Molecular Dynamics Simulations

The GROningen MAchine for Chemical Simulations (GROMACS) 2019 version was utilized to perform molecular dynamics (MD) simulations on three complexes Phospholipase_ACA, Phospholipase_MO and Phospholipase_RU complexes (Abraham et al. 2015). In the present study utilizes GROMACS molecular dynamics software package, (Version 2019), using Charmm36-mar2019 force field for the protein parameters. The complex was solvated explicitly using the TIP3P water model inside the cubic box, and its size extends 0.1 nm away from the protein on the edges of the box in each direction. The system’s overall charge was neutralized by adding a 0.1 M salt concentration (Na⁺Cl⁻). The entire system was minimized till the maximum force was less than 10 kj/mol with the maximum steps of 50,000. The system was then equilibrated for 100ps under NVT and NPT conditions with temperature coupling for two separate groups, at 300 K. Lincs algorithm is used to constrain the bonds of the hydrogen atoms (Lemak and Balabaev 1994). Berendsen thermostat and V-rescale were used to keep the temperature and pressure constant, respectively. The cutoff distances for Coulomb and van der Waals interactions were set as 1.2 nm. Particle mesh Ewald method (PME) was used to calculate the long-range electrostatic interactions.

Then, GROMACS modules were used to do the necessary analysis, and origin was used to generate the necessary charts and figures from the 200 ns of simulations run under periodic boundary conditions (PBC). The purpose of these simulations was to ensure the flexibility and stability of the complexes that were formed after the docking process. Initially, the protein and ligand topologies were obtained separately using the 'pdb2gmx' script and the CGenFF server, respectively. These topologies were then combined to create a conformation, which underwent energy minimization using the CHARMM36 force field. The protein-ligand complex was placed in a dodecahedron and solvated with TIP3P water molecules. Counter ions were introduced to neutralize the system, followed by a rapid energy minimization using the steepest descent algorithm. The equilibration process involved restrained simulations in an ensemble with a constant number of particles, volume, and temperature (NVT), followed by another ensemble with a constant number of particles, pressure, and temperature (NPT) for 100 ps. Throughout the equilibration, thermodynamic properties such as pressure, density, potential energy, and temperature were monitored to ensure proper equilibration before the production run. The particle mesh Ewald method was employed for calculating long-range electrostatic interactions. Finally, unrestrained simulations were conducted for 200 ns at 300 K and 1 bar atmospheric pressure for each system. Simulation duration, 200 nsec could provide a more comprehensive and significant exploration of the dynamic behaviour and interactions of the protein-ligand complexes under investigation (Tantar et al. 2008). The solvated and equilibrated complexes (three protein-ligand complexes) were then subjected to the production phase to generate trajectories for root mean square deviation (RMSD), root mean square fluctuation (RMSF), hydrogen bond (Hbond) graphs, radius of gyration (RG), and solvent accessible surface area (SASA) using built-in GROMACS scripts.

The analysis like RMSD, RMSF, Rg, SASA, and Number of hydrogen bonds estimation were carried out using the Gromacs utilities. The trajectory files obtained by the MD simulation was examined using the GROMACS commands gmxrms, gmxrmsf, gmx gyrate, and gmxhbond to obtain the root-mean-square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration, and hydrogen bond interaction, respectively. After the preparation of XVG files using the Gromacsbuilit-in functions, origin was employed to plot the graph and comparing the results in both ligand-free and ligand bounded forms of the target protein.

2.2.7. MM/PBSA binding free energy calculation

To calculate the binding free energy of protein-ligand complexes, a combination of the Molecular Mechanic/Poisson-Boltzmann Surface Area (MM-PBSA) approach and molecular dynamics (MD) was employed. The final 20 ns of the final production run were specifically utilized for the calculation of binding free energy. The binding free energy estimation provides a comprehensive understanding of the interactions between the protein and ligand at a molecular level. It encompasses various energy contributions, including potential energy and solvation energies (both polar and non-polar). Further in case of MM/PBSA calculations the "MmPbSaStat.py" script was designed to automate the calculations and provide a convenient way to perform these types of binding free energy calculations for molecular complexes in a consistent and reproducible manner (Kumari et al. 2014). It takes input files containing molecular structures, force field parameters, and other necessary parameters to carry out the calculations and output the estimated binding free energies and their components. This script stands for the Molecular Mechanics, Poisson-Boltzmann Solvation Model, Surface Area, and Entropy Contributions. Molecular Mechanics involves classical force fields to compute the gas-phase energy of the molecular complex. This energy represents the energy changes due to molecular deformation and interaction in the absence of solvent effects. The Poisson-Boltzmann model calculates the electrostatic solvation energy based on the distribution of charges and dielectric properties of the molecules and the solvent. Surface Area estimates the change in solvation energy due to the exposure of the nonpolar surface area of the molecules to the solvent. This is calculated using a solvent accessible surface area model. Entropy Contributions include entropy contributions, such as the conformational entropy of the molecules, to provide a more accurate estimation of the binding free energy. The GROMACS 'g_mmpbsa' script (Kumari et al. 2014) was employed to compute the MM-PBSA binding free energies. The following equation was used to compute the binding energy in this method:

ΔG _binding = G _complex - (G _receptor + G _ligand)

The ΔG binding represents the total binding energy of the complex, while the binding energy of free receptor is G receptor, and that of unbounded ligand is represented by G ligand.

3.1. In silico Transcriptomic Analysis

3.1.1. DEGs screening

After normalization and pre-processing (Figure 2A-B) of 10 T2D and 10 control samples, a total of 20,994 gene symbols were obtained which were applied for DEGs analysis. Based on above DEGs screening criteria, 194 up-regulated genes and 140 down-regulated genes were obtained. The DEGs are represented in the volcano plot (Figure 2C) in which the top 10 over-expressed and top 10 under-expressed genes are highlighted. To visually depict the expression patterns of these genes among T2D and ND condition, a hierarchically clustered heatmap (Figure 2D) was plotted based on their normalized expression value.

3.1.2. WGCNA network construction and selection of key modules

The distance between each sample was estimated by clustering them with the normalized expression of 340 DEGs into Pearson’s correlation matrix. As a result, no outlier was detected (Figure 3A), and all 20 samples further proceeded to obtain a scale-free network. A soft threshold of β=12 (R2 =0.85 & slope= -1.57) was chosen as the appropriate power (Figure 3 B1-B2) to construct a captivating cluster within the dendrogram.

The top 300 highly connected genes were clustered into 18 modules by the Dynamic tree-cut method (Figure 4 A-B). The analysis reveals a total of 334 genes that are expressed differentially in T2D conditions, with 58% over-expressed and 42% being under-expressed which have been shown in Figure 3C. Genes with similar expression patterns were merged which results in eight highly correlated modules. The module-trait association and average gene significance (Figure 4D) reveal that the MEgreen (r=0.9, P=2e-06) with 31 genes, MEyellow (r=0.8, P=2e-04) with 63 genes, and MEgrey60 (r=0.7, P=2e-04) with 61 gene sets are highly correlated with the T2D condition. The subsequent modules were prioritized for the hub genes identification by evaluating their GS and MM values. The correlation among GS and MM for key modules is illustrated in scatterplot (Figure 4E-G). Following the aforementioned criteria, a select group of 60 genes emerged as prominent hub genes from three pivotal modules (MEgreen: 11 genes, MEyellow: 16 genes, and MEgrey60: 33 genes). Additional information is provided in Figure S1-S2 & Table S1.

3.1.4 Functional Enrichment results

According to the enrichment result (Table 1) from DAVID, digestion, proteolysis and neutrophil chemotaxis are the highly significant enriched BP terms. For CC, most of the genes significantly participated in extracellular space, extracellular region and extracellular exosome. MF results reveal that the serine-type endopeptidase activity and oligosaccharide binding are significant terms (Figure 5A). Moreover, the hub genes are closely linked with pancreatic secretion, protein digestion & absorption and fat digestion & absorption pathways (Figure 5B-C2).

Table 1: Functional Enrichment Analysis (Top 5 enriched terms).
Category	ID	Term	p-value	Genes symbol
BP	GO:0007586	Digestion	1.41E-06	PRSS1, CLPS, CTRB2, CTRB1, PRSS2
	GO:0006508	Proteolysis	1.95E-05	CPA2, PRSS1, CPA1, CPB1, ANPEP, CTRB2, CTRB1, CTRC, PRSS2
	GO:0030593	Neutrophil chemotaxis	8.09E-05	LGALS3, EDN3, PLA2G1B, CXCL1, SLC37A4
	GO:0061844	Antimicrobial humoral immune response mediated by antimicrobial peptide	0.000243	LGALS3, REG3A, PLA2G1B, REG1A, CXCL1
	GO:0016042	Lipid catabolic process	0.00283	PNLIPRP1, CLPS, PLA2G1B, PNLIP
CC	GO:0005615	Extracellular space	1.78E-11	REG3A, PRSS1, CPB1, PLA2G1B, REG1A, CXCL1, LGALS3, ANPEP, DPYSL3, ANXA9, PRSS2, CPA2, CPA1, EDN3, CTRB2, CTRB1, GP2, CEL, DKK3, COL1A2, TST, ALB, SERPINI2, SERPING1, PNLIP
	GO:0005576	Extracellular region	1.83E-07	PNLIPRP1, REG3A, CPA2, PRSS1, PCOLCE2, EDN3, PLA2G1B, CTRB2, CTRB1, GP2, CXCL1, CEL, DKK3, LGALS3, CLPS, COL1A2, ALB, SERPING1, CTRC, PRSS2, PNLIP
	GO:0070062	Extracellular exosome	0.000429	SPINK1, KLK1, ANXA4, REG1A, GP2, CEL, AQP1, EPN3, LGALS3, GATM, COL1A2, FXYD3, ANPEP, ALB, SERPINI2, SERPING1
	GO:0009986	Cell surface	0.006525	LGALS3, PLA2G1B, ANXA4, GP2, CEL, ANXA9, CFTR
	GO:0042589	Zymogen granule membrane	0.031045	GP2, CUZD1
MF	GO:0004252	Serine-type endopeptidase activity	0.000203	PRSS1, KLK1, CTRB2, CTRB1, CTRC, PRSS2
	GO:0070492	Oligosaccharide binding	0.00059	LGALS3, REG3A, REG1A
	GO:0004180	Carboxypeptidase activity	0.001878	CPA2, CPA1, CPB1
	GO:0004806	Triglyceride lipase activity	0.0024	PNLIPRP1, CEL, PNLIP
	GO:0004181	Metallocarboxypeptidase activity	0.003403	CPA2, CPA1, CPB1
KEGG	hsa04972	Pancreatic secretion	2.84E-14	PNLIPRP1, CPA2, PRSS1, CPA1, CPB1, PLA2G1B, CTRB2, CTRB1, CEL, PRSS2, CFTR, PNLIP
	hsa04974	Protein digestion and absorption	6.43E-08	CPA2, PRSS1, CPA1, CPB1, COL1A2, CTRB2, CTRB1, PRSS2
	hsa04975	Fat digestion and absorption	1.62E-05	PNLIPRP1, CLPS, PLA2G1B, CEL, PNLIP
	hsa00561	Glycerolipid metabolism	0.021335	PNLIPRP1, CEL, PNLIP
	hsa04976	Bile secretion	0.041584	AQP8, CFTR, AQP1

3.1.5 Candidate gene signature for T2D

The PPI network was constructed in STRING tool with 60 module hub genes with an average node degree of 3.42 and a clustering coefficient of 0.32. The network with 30 nodes and 94 edges was visualised within Cytoscape (Figure 6A), later top 15 highly interconnected genes were calculated from four algorithms of “CytoHubba” plugins (Table 2). The intersection (Figure 6B) of these four methods results in 7 commonly identified hub genes within the PPI network namely carboxypeptidase A1 (CPA1), chymotrypsin-like elastage 3B (CELA3B), carboxypeptidase B1 (CPB1), carboxy ester lipase (CEL), chymotrypsinogen B2 (CTRB2), phospholipase A2 group 1B (PLA2G1B), and regenerating family member 1 alpha (REG1A). The functional insights of the candidate genes associated with T2D are represented in Figure 6A.

Table 2: Hub gene Identification by four methods of CytoHubba.
Rank	MCC	DMNC	Degree	Betweenness
1	CTRB2	CTRB2	CTRB2	CPB1
2	CPB1	CPB1	CPB1	AQP1
3	CLPS	CLPS	CLPS	SERPING1
4	PNLIPRP1	PNLIPRP1	PNLIPRP1	NEFL
5	CPA2	CPA2	CPA2	ANXA4
6	CTRC	CTRC	CTRC	CPA1
7	CPA1	CPA1	CPA1	CEL
8	CTRB1	CTRB1	CTRB1	CTRB2
9	CEL	CEL	CEL	ANPEP
10	SPINK1	SPINK1	SPINK1	CELA3B
11	ANPEP	CELA3B	ANPEP	GP2
12	CELA3B	LGALS3	CELA3B	PLA2G1B
13	PLA2G1B	PLA2G1B	PLA2G1B	REG1A
14	ALB	REG1A	ALB	ALB
15	REG1A	CXCL1	REG1A	REG3A

3.1.6. Validation of hub genes

The diagnostic value of each hub gene was evaluated in discovery dataset (Figure 7A) as well as the validation dataset (Figure 7B). The training cohort exhibited the following area under curve (AUC) values based on the ROC curve: CPA1 (AUC=0.83), CELA3B (AUC=0.86), CEL (AUC=0.87), CTRB2 (AUC=0.8), CPB1 (AUC=0.82), PLA2G1B (AUC=0.88), REG1A (AUC=0.94). Further, the validation in GSE41762 dataset confirmed the AUC values as follows: CPA1 (AUC=0.78), CELA3B (AUC=0.84), CEL (AUC=0.76), CTRB2 (AUC=0.8), CPB1 (AUC=0.82), PLA2G1B (AUC=0.82), REG1A (AUC=0.79). The expression levels of these genes were also studied in both discovery and validation dataset (Figure 7C-D). Consequently, an exquisite outcome reveals a significant over-expression of CEL, CPA1, CPB1, CTRB2, CEL3B, PLA2G1B, and REG1A in T2D when compared to the ND state. This remarkable observation holds true not only in the discovery dataset but also in the validation dataset.

The physicochemical properties of natural phytocompounds and anti-diabetic drugs with PubChem IDs were recorded, including molecular weight (g/mol), number of H-bond acceptors, H-bond donors, and number of rotatable bonds, and octanol/water partition coefficient (XlogP) (Table 3). In current drug development modules, the aforementioned physicochemical characteristics are generally known as the Lipinski rule five (RO5), a common strategy for identifying the likely drug-able candidate at an early stage. However, several diabetes drugs may not meet RO5 standards, but a few phytoflavonoids do. Nonetheless, there is a mutual negotiation in both biological activity and physicochemical property during optimal medicine selection.

Table 3: Physicochemical properties of the ligands.
Physicochemical properties	Ligands
Physicochemical properties	n-(p-Amylcinnamoyl) anthranilic acid	Morin	Rutin
PubChem CID	5353376	5281670	5280805
Molecular Formula	C21H23NO3	C₁₅H₁₀O₇	C₂₇H₃₀O₁₆
Molecular Weight (g mol⁻¹)	337.4	302.23	610.5
Hydrogen-binding donors	2	5	10
Hydrogen-binding acceptors	3	7	16
Rotatable bond count	8	1	6
XLogP3	6.4	1.5	-1.3
Molar Refractivity (mol/m³)	-	74.1	137.5

Similarly for Human Pancreatic Phospholipase A2c which contributes towards the progression of diabetes was retrieved from RCSB protein data bank. The details are as follows where the PDB ID: 6Q42, classification: - Hydrolase, method used for their analysis was X-ray Diffraction with resolution of 1.90 Å, R-value free:- 0.225 and R-value work:- 0.175.

3.2.2. Possible Ligand Toxicity Predection through LD₅₀ value

Possible Ligand Toxicity for each compound was predicted calculation of LD50 value using ProTox programme. The phytocompound RU and MO had higher Predicted LD₅₀i.e. 3919 and 5000 mg/kg respectively in comparison to all the standard inhibitors accept to ACA presented in Table 4. Moreover, both RU and MO shared a Predicted Toxicity Class of five, surpassing the standard inhibitors with a lower safety level of four (Table 4).

Table 4: Details of toxicity parameters of the ligands.
Toxicity Parameters	Ligands
Toxicity Parameters	n-(p-Amylcinnamoyl) anthranilic acid (ACA)	Rutin	Morin
Predicted LD₅₀	1190 mg/kg	3919 mg/kg	5000 mg/kg
Predicted Toxicity Class	4	5	5

3.2.3. Molecular Docking Studies

Molecular docking studies of MO and RU in comparison to the standard inhibitor were predicted to possess higher ΔG values. For instant Table 5 provide the details of the Phospholipase A2 docking with MO and RU, where it can be seen that the phytocompounds had a higher ΔG values i.e. -6.4 and -7.3 kcal/mol respectively whereas the standard inhibitor ACA had an ΔG values of -5.8 kcal/mol. The details of the amino acids of enzyme involved in the interaction with MO includes ASN24, TYR25, GLY26, THR120, SER34, GLY33 of which THR120 was common among amino acids interacting with ACA. In case of RU ASN117, LYS121, THR120, ASN24, GLY30, LEU31, GLY32, CYS29 of which THR120 was the common amino acids involved in interactions with the standard inhibitor ACA represented in Figure 8.

Table 5: Docking details of Phospholipase A2 and the inhibitory molecules.
Enzyme	Inhibitor	Binding Affinity (Kcal/ mol)	Amino Acids	Interaction Type	Types of Bonds	Subtypes	Number of Bonds
Phospholipase A2	n-(p-Amylcinnamoyl) anthranilic acid (ACA)	-5.8	PHE19, CYS29, THR120	Favourable	Hydrogen Bonds	Classic (Conventional)	1
						Non-Classic (Carbon)	1
					Hydrophobic Bonds	Pi-Alkyl	1
						Pi-Sigma	2
				Unfavourable	-	-	-
	Morin	-6.4	ASN24, TYR25, GLY26, THR120, SER34, GLY33	Favourable	Hydrogen Bonds	Classic (Conventional)	3
						Non-Classic (Carbon)	2
					Hydrophobic Bonds	Pi-Sigma	1
				Unfavourable	Acceptor/Donor Clash	Unfavourable Donor-Donor	1
	Rutin	-7.3	ASN117, LYS121, THR120, ASN24, GLY30, LEU31, GLY32, CYS29	Favourable	Hydrogen Bonds	Classical	4
					Hydrophobic Bonds	-	-
				Unfavourable	Acceptor/Donor Clash	Unfavourable Donor-Donor	1
						Unfavourable Acceptor-Donor	1

3.2.4. Molecular Dynamic Simulation

3.2.4.1. Root mean square deviation (RMSD) characteristics

The mean distance between superimposed set of atoms in the peptide is measured by the root-mean-square deviation (RMSD). It examines the structural distance between atomic coordinates from the locus point. There has been a constant increase in deviation of native Phospholipase enzyme and its complexes throughout 100 nsec trajectories as shown in Figure 9A. The RMSD values of Phospholipase in complex with standard inhibitor - ACA decreases in comparison with native Phospholipase enzyme with a maximum deviation of 0.2755 nm at 47.61 nsec initially and a high deviation ~0.7 nm at 150-160 nsec, after that maintain its initial confirmation. However, the RMSD values of Phospholipase-MO conjugate gradually increases and out passes the native Phospholipase enzyme with a peak deviation of 0.2170 nm at 96.52 nsec and highest peak observed between 80-90 nsec and again it maintains its initial stability. Additionally, it has been noted that the deviation of RMSD values for the Phospholipase-RU complex exhibits a similar pattern to that of the native Phospholipase enzyme which, shows an initial deviation of 0.2245 nm at 47.5 nsec in comparison to native enzyme and maintain the stability throughout the 200 nsec simulation. Comparing the all RMSD of complexes, binding of Phospholipase-RU maintains high stability of the complex.

3.2.4.2. Root mean square fluctuation (RMSF) characteristics

The RMSF is an average measure of the fluctuations of a specific atom with respect to its initial coordinates in the protein backbone. It denotes the flexibility of the protein for given complex in each time frame. Analyzing the graph, it worth nothing that 4 peaks are observed fluctuating above 0.3 nm, where first 3 peaks are due to the presence of loop region and 4^th is the terminal region. The fluctuations for both native Phospholipase enzyme and its complexes are significant in amino acid residues ranging from 28 to 35 and 120 to 126 positions as illustrated in Figure 9B. Moreover, in case of Phospholipase with MO the maximum deviation observed at 31^th, 80^th and 121^th amino acid residue with a value of 0.55, 0.28 and 0.42 nm respectively. Additionally, the RMSF values of Phospholipase-RU complex shows a noticeable variation of 0.4594 nm at 126^th amino acid residue. Similarly, the Phospholipase-ACA complex shows a noticeable variation of 0.565 nm at 126^th amino acid residue. Overall the all the complexes showed fluctuation here, in a manner present here Enzyme -ACA<Enzyme -RU<Enzyme -MO< Native Enzyme. Concluding from analysis that there is no any significant changes observed in the binding side region.

3.2.4.3. Deviation in Radius of Gyration

The radius of gyration (Rg) is the root mean square distance between a set of atoms and their shared gravitational center. The Rg access the dynamic trajectory of flexible systems in addition to compactness of molecular systems. The Rg values increase for native Phospholipase enzyme as well as all the complexes throughout 100,000 psec trajectories as seen in Figure 9C. The Rg values of Phospholipase in complex with standard inhibitor - ACA decreases in comparison with native Phospholipase enzyme with a maximum deviation of 0.12182 nm at 29900 psec. However, in case of Phospholipase-MO complex the Rg values significantly increases and out passes the Rg values of native Phospholipase with a noticeable variation of 0.13728 nm at 15240 psec. Similarly, in the case of Phospholipase-RU complex the Rg values exhibit maximum deviation out of all complexes with a significant deviation0.12475nm at 6330 psec. There is no significant fluctuation observed and comparing with standard native enzyme the graph of all other complexes decreasing at the end of the simulation. So we can conclude that at the end of the simulation compactness of the complexes are increasing and maintain at certain point.

3.2.4.4. Solvent-accessible surface area (SASA) Characteristics

The SASA analysis of a peptide establishes whether the amino acid residues are hidden or exposed to the solvent molecule. The protein molecule's van der Waals interface serves as a virtual core of the solvent sphere, and as SASA levels rise, protein instability and relaxation occur. However, the SASA values of native Phospholipase and all the complexes exhibit significant fluctuation throughout 100 nsec as shown in Figure 9D. The complex Phospholipase-RU exhibits maximum SASA fluctuation at 15.01 – 30 nsec with a value of 91.491 nm²which further decreased to 79.292 nm²at 100 nsec. Whereas the values of Phospholipase – ACA complex peaks 88.034 nm² at 9.18 nsec which further decreased to 80.792 at 100 nsec followed by Phospholipase-MO complex whose value decreases from 87.45 nm² at 93.98 nsec to 83.197 nm² at the end of trajectories. Concluding, similar patterns are observed with radius of gyration graph, the solvent accessible surface area is decreasing for all complexes.

3.2.4.5. Change in hydrogen bonds

The stability of a ligand-protein complex depends critically on the hydrogen bond formation that takes place during the simulation. Although there are multiple gaps in hydrogen bond formation for all the complexes of Phospholipase enzyme but the Phospholipase - ACA complex have higher intervals where number of hydrogen bond is zero as during 8-16 nsec in Figure 9E. Furthermore, Phospholipase-RU exhibits maximum number of hydrogen bonds in comparison to all complexes which is marked by 9 hydrogen bonds at 0.46 nsec. However, in case of Phospholipase-MO complex maximum number of hydrogen bonds observed is 5 at 73.86 nsec. Similarly in case of Phospholipase - ACA complex the maximum number of hydrogen bonds in is 4 which peak at 0.08 nsec. Concluding the result Phospholipase – ACA complex retain 2 hydrogen bond while Phospholipase- RU complex and Phospholipase Morin complex retain 3 hydrogen bond throughout the 200 nsec simulation.

3.2.4.6. MM-PBSA binding free energy

To perform MM-PBSA binding free energy calculation the Python script MmPbSaStat.py from the “g_mmpbsa” package was taken into consideration for protein ligand complexes as listed in Table 6. This script computes the average free binding energy as well as its standard deviation/error based on the output files obtained from g_mmpbsa. The binding energy represents the energy released during the formation of bonds or the interaction between a protein and a ligand. A lower binding energy indicates a stronger binding between the ligand and protein. The final binding energy is a combination of van der Waals, electrostatic, polar solvation, and solvent-accessible surface area (SASA) energies. Among all the selected molecules, the bioactive compound MO exhibited the lowest binding free energy (-69.828 kJ/mol). RU and ACA had the second and third lowest binding free energies of -15.648 kJ/mol and -2.639 kJ/mol, respectively.

Table 6. Binding free energy analysis of the protein-ligand complexes expressed in kJ/mol.
Energy (kJ/mol)	Phospholipase_ACA	Phospholipase_Morin	Phospholipase_Rutin
Van der Waal Energy	-14.894 ± 36.114	-132.362 ± 13.114	-96.961 ± 21.612
Electrostatic Energy	-2.347 ± 8.100	-30.182 ± 12.729	-27.489 ± 13.945
Polar Salvation Energy	16.511 ± 35.172	105.401 ± 18.545	121.621 ± 35.660
SASA Energy	-1.910 ± 4.705	-12.685 ± 0.987	-12.818 ± 1.762
Binding Energy	-2.639 ± 34.036	-69.828 ± 12.971	-15.648 ± 37.999

Type 2 diabetes is the most common metabolic disorder suffered by 90% of individuals globally. Numerous diagnosis studies implicated the factor that increases the risk of T2D and advanced prevention strategies, still the disease is more widespread around the world. Therefore, highly effective diagnostic and therapeutic biomarkers discovery is a matter of utmost urgency. Extensive bioinformatics analysis was implicated to identify potential biomarkers that have a possible role in T2DM. The microarray dataset GSE20966 has been chosen in order to determine the pattern of gene expression in T2D circumstances. Normalization was the first step in adjusting the level of hybridized gene intensities and enhancing the accuracy of measured expression (Quackenbush 2002).

WGCNA is a gene screening approach that focuses on identifying biomarkers in relation to clinicopathological characteristics. This has been notably utilized in the analysis of various types of genomic data, including microarray, RNA-Seq, epigenetic data, and brain image data analysis, among others (Kadarmideen et al. 2011; Mumford et al. 2010; Voineagu et al. 2011). In WGCNA, genes exhibiting similar expression patterns are grouped together into modules and subsequently, the correlations between these modules and clinical features are computed. Study on correlation between modules and clinical traits can simplify the relationships between modules and hub genes. In a similar study, Rezaei et al. (2022) identified ITGAX, CCL14, ADHFE1 and HOXB13 as novel biomarkers for gastric cancer. Similarly, Hu et al. (2022) constructed a co-expression network that identified S100A6 as a promising immune-related biomarker with potential therapeutic implications for individuals diagnosed with both pancreatic cancer and T2D. In this study, the same method was employed to analyse clinical features such as T2D and ND using DEGs that were upregulated and downregulated. From WGCNA and PPI network results illustrated in Figs. 4 & 6, seven genes (CEL, CPA1, CPB1, CTRB2, CEL3B, PLA2G1B, and REG1A) were identified as hub genes that are highly associated with T2D pancreatic islets of beta cells.

The findings indicate that in both cohorts, the AUC values were above 0.7, indicating a strong diagnostic power and suggesting that these hub genes have the potential to serve as biomarkers. The expression levels of the hub genes were further investigated in the external dataset GSE41762 (Fig. 7D), confirming a significant increase in their expression in individuals with T2D compared to those with ND. After confirming their diagnostic value and examining their expression, CEL, CPA1, CPB1, CTRB2, CEL3B, PLA2G1B, and REG1A show promise as potential biomarkers for the patients with T2D. This analysis delved into the intricate biological processes and molecular pathways associated with the biomarkers, illuminating their vital roles in the broader biological context of T2D (Table 1). Mutation in the digestive enzyme, CEL contributes to maturity onset diabetes of the young (nearly 1–2% of all diabetes cases), type 8 (MODY8) (Wu et al. 2023). The genes CPA1 and CPB1 are the most expressed genes in acinar cells of pancreas. However, impaired secretion or mutation-induced misfolding results in ER stress. These developments lead to chronic pancreatitis that further contributes diabetes (Sarkar et al. 2023; Tamura et al. 2018). Further the gene CELA3B, an exocrine gene responsible for cleaving proteins after alanine residues, and additionally it has also been found to be overexpressed in T2D acinar cells. Moreover, CELA3B protein accumulates in the islet milieu and impairs cell homeostasis in T2D (Basile et al. 2023). Another gene named REG1A gene whose product levels if elevated in the blood can be a sign for β-cell apoptosis, it is also present in various types of diabetes, including Type 1 Diabetes, MODY, and T2D. It has been evidenced that patient with T2D have upregulated Reg levels both at initial diagnosis and after the emergence of long-term complications (Chen et al. 2019).

Richmond et al. (2001) reported that Pla2g1b is responsible for hydrolysing phospholipids in the intestinal lumen, enabling cholesterol absorption in the mouse model, this process contributes to postprandial glucose intolerance and hyperglycemia through the absorbed lysophospholipid. Methyl indoxam-driven Pla2g1b inhibition diminishes obesity and diabetes in mice, indicating promise in addressing human intestinal PLA2G1B to alleviate hypercaloric diet consequences, benefiting individuals with excessive caloric intake (Hui et al. 2009). According to Labonté et al. (2006), Postprandial hyperglycemia decrease in Pla2g1b-inactivated mice as compared to the active Pla2g1b mice suggests improved glucose metabolism and potential early T2D mitigation. The over-expression of PLA2G1B plays a critical role in diet induced in transgenic mice (Cash et al. 2011). As per a recent study in high-fat and high-cholesterol diet-fed mice, PLA2G1B inhibition not only inhibited diet-induced hyperglycemia and hyperlipidaemia, but also reversed and normalised plasma glucose and lipid levels after diabetes onset (Cash et al. 2018).

As PLA2G1B emerges as a prominent gene with persuasive scientific evidence directly linking its involvement as a triggering element in the progression of T2D therefore the protein product of the gene became the concern of the present investigation. Studies have been conducted using in silico analysis to investigate the potential of various phytochemicals as inhibitors in diabetic condition. However, no previous studies directly addressed the potential side effects of RU and MO in diabetic conditions. There are very few papers provide evidence on the IC50 value of MO in diabetic conditions. For instance Zeng et al., (2016) found that MO inhibited α-glucosidase in a reversible mixed-type manner with an IC50 value of (4.48 ± 0.04) µM. However, there were hardly in direct evidence that reported the IC50 value of RU in diabetic conditions.

Past investigations indicate that both MO and RU are effective in the treatment of management of T2DM and some of its complications, such as T2D, diabetic neuropathy, diabetic retinopathy, and diabetic liver damage. Most recently, a study found that MO administration decreased oxidative stress and regulated the altered activities of carbohydrate metabolizing enzymes in the liver of diabetic rats (Shali et al. 2022). These conclusions were in line with another study that found that MO supplementation decreased oxidative stress and inflammation and increased neurotrophic support in the diabetic brain (Ola et al. 2015; Ola et al. 2014). In a recent publication by Jiang et al., (2020) reported that, the bioactive compound MO supplementation showed a noteworthy decline in elevated serum glucose level and notably dropped the concentration of pro-inflammatory players like TNF-α, IL-1β, and VEGF in the tissue homogenate of the retina. This phytocompound further have been proven to improved motor and sensory nerve conduction velocities (MNCV and SNCV), nerve blood flow (NBF) and reduced sensorimotor alterations in diabetic models (Bachewal et al. 2018). Further, MO administration attenuated diabetic-induced osteopenia while it was established that zinc-MO complex possesses antidiabetic, antidyslipidemic, and antioxidant potentials in HFD-fed STZ induced diabetic rats (Abuohashish et al. 2013; Sendrayaperumal et al. 2014). Similarly, in case of RU, an investigation found that the phytocompound could improved body weight, reduced plasma glucose, and improved antioxidant status in high fat diet (HFD) + streptozotocin (STZ) induced T2D rats (Niture et al. 2014). Previously Kamalakkannan and Prince (2006) found that RU exhibited antihyperglycemic and antioxidant activity in STZ-induced diabetic rats. Reports also suggests that this compound could led to significant amelioration of hyperglycemia, hyperlipidemia, and oxidative stress in diabetic rats (Ahmed et al. 2010). A similar study focussing on cardiac myopathies in diabetes condition concluded that RU could attenuated cardiac remodeling and left ventricular and myocardial dysfunction caused by STZ-induced diabetes mellitus (Guimaraes et al. 2015). Concurrently, another evaluation revealed that the same phytocompound could decrease fasting blood sugar, systolic and diastolic blood pressure, and serum urea and creatinine levels in diabetic patients (Sattanathan et al. 2011). However, none of the papers directly address the safety profile of MO and RU administration in diabetic conditions. As per present research these phytocompounds (MO and RU) could effectively bind the enzyme with a binding affinity higher than that of the standard inhibitor ACA. This implies that it may be necessary to conduct experimental assessments both in vitro & in vivo to confirm the inhibitory effects of MO and RU.

There are various other researches where the possibility of inhibiting the enzyme PLA2 is proven by phytocompounds there is hardly any evidence regarding the inhibitory activity of MO and RU specifically. For instant, Tap et al., (2018) performed in silico and in vitro analysis of the Bromelain for its PLA2G1B inhibitory activity. As per their research it can be evidenced that the phytocompound could effectively bind and inhibit the enzyme. Similarly, another study, demonstrated that ethanol extract of Newbouldia laevis stem bark could inhibit PLA2G1B both in silico and in vitro. Through their analysis they found that photocompounds like Apigenin, Catechin, Isorhamnetin, Kaempferol, Quercetin, Naringin and Luteolin are mostly present in N. laevis stem bark ethanol extract therefore might have a role in inhibiting PLA2 (Omeje et al. 2023). While these in silico studies provide promising evidence for the potential of phytochemicals as inhibitors of PLA2G1B, further experimental research is needed to confirm these findings and investigate the safety and efficacy of these compounds for treating diseases associated with elevated PLA2G1B activity.

Further, MD simulation analysis also provided similar evidence where the complexes of Phospholipase with ACA, MO and RU were explored for their confirmatory binding affinities. The analysis validated that these phytocompounds have significant effect on protein structure and stability by forming hydrogen bonds in comparison to the standard inhibitor ACA. For MD simulation CHARMM36 force field was selected for protein-ligand complexes under investigation, previous studies have reported successful applications of the CHARMM36 force field in studying phospholipase A2 (PLA2) enzymes and their interactions with ligands (Mouchlis et al. 2016). Additionally, the force field has been used to study other proteins involved in T2D-related molecular pathways, demonstrating its relevance to our research area (Huang et al. 2017; Xu and Huang 2021). Moreover, the force field incorporates parameters for a wide range of biomolecules, including lipids, nucleic acids, and carbohydrates, which makes it well-suited for studying complex biological systems and their interactions (Xu and Huang 2021). There are recent investigations that highlights its applicability to the presently studied protein-ligand complexes and T2D-related molecular interactions (Syaifie et al. 2022). There has been significant alteration of protein structure during 200 nsec trajectories that arises due to change in flexibility of amino acid residues in the structure. Furthermore, the binding of phytocompounds – MO and RU subsequently affect compactness of the enzyme resulting in exposing of amino acid residues of phospholipase to the solvent molecule.

Through combining these different components, MM/PBSA attempts to capture both the energetic and solvation effects involved in molecular binding that is previously shown in Table 6. It can be seen that the phytocompound-enzyme complex presents higher Van der Waals energy which is an indication of stronger non-covalent Van der Waals attraction force. This can have significant scientific implications in various fields, including drug design, structural biology, and our understanding of molecular interactions in biological systems. Therefore it is a fundamental parameter that helps to assess and predict the stability and strength of these interactions (Bitencourt-Ferreira et al. 2019a). Higher electrostatic energy between complexes is an indication of stronger electrostatic interactions, which helps to understand and predict the stability, specificity, and functionality of molecular interactions in biological and chemical systems (Bitencourt-Ferreira et al. 2019b). Similarly higher polar salvation energy in the context of a ligand-receptor complex reflects the energetics of solvation and structural changes associated with the binding process. It provides valuable information for understanding and optimizing ligand-receptor interactions, in heterogeneous biological processes (Kar et al. 2013). Higher SASA energy in a ligand-receptor complex indicative of enzyme structures those are unfolded or more extended. Higher SASA energy is an indicative of larger SASA value of phytocompound-enzyme complex that makes it more likely to interact with solvent molecules (Ashraf et al. 2022). Understanding this energy component is essential because it sheds light on the thermodynamics, kinetics, and structural aspects of ligand-receptor interactions. It is particularly useful in various context more importantly studies relating to molecular recognition processes. Finally, higher binding energy in between phytocompound-enzyme complex reflects the strength, specificity, and stability of the molecular interaction. It is a key parameter in drug design, structural biology, and the study of molecular recognition processes in biology. Understanding and optimizing binding energies are fundamental to advancing our knowledge in biomedical applications specifically in the search of therapeutic molecules and inhibitors (Meng et al. 2011; Pantsar and Poso 2018).

In short, it can be said that T2D remains a significant global health concern, characterized by hyperglycemia or hypoinsulinemia. Despite numerous previous studies, its prevalence and associated complications continue to pose challenges. This study aimed to identify potential biomarkers and investigate the underlying molecular mechanisms influencing the progression of T2D. Through analyzing a specific human microarray dataset containing pancreatic beta cells, the study identified DEGs and constructed co-expression networks using R and Bioconductor packages. Furthermore, the hub genes exhibit an intimate association with the intricate orchestration of pancreatic secretion, the complex processes of protein digestion, and the sublime absorption of fats, all of which are intricately linked to T2D. Among these genes, PLA2G1B emerged as a particularly relevant factor in T2D development, supported by scientific evidence. Consequently, further proteomic studies were conducted to examine the possible inhibition of PLA2G1B as a means to decrease the severity of T2D. Molecular docking and MD simulation techniques were employed to assess the inhibitory effects of phytocompounds MO and RU on PLA2G1B. The results indicated that MO and RU exhibited more efficient binding affinity towards PLA2G1B compared to standard inhibitor ACA. These findings indicate that the flavonoids i.e. MO and RU have the potential to be utilized supplementary therapeutic strategies for combating T2D. In general, this study provided insights into the molecular mechanisms that underlie T2D and identified important hub genes that orchestrate a role in the advancement of the disease. The investigation of PLA2G1B and the subsequent analysis of its higher binding affinity by MO and RU provide valuable insights for potential therapeutic strategies in the management of T2D. Further research in this direction may contribute to the development of novel treatments that address the complex nature of T2D and improve patient outcomes. Despite of the thorough computational exploration conducted at both the transcriptomics and MD simulation for proteins, the research does have certain constraints, particularly the absence of comprehensive validation via experimentation conducted in in vitro and in vivo. The primary emphasis of the paper was centred on revealing the potential utility of transcriptomics and MD simulation for proteins in sorting out the intricacies of T2D. This knowledge could eventually be harnessed to formulate innovative combined therapeutic and diagnostic approaches for addressing the disease in the times ahead.

DEGs

Differentially Expressed Genes

T2D

Type 2 Diabetes

MO

Morin

RU

Rutin

WGCNA

Weighted Co-Expression Network Analysis

PPI

Protein-Protein Interaction

DAVID

Database of Annotation

Visualization

and Integration Discovery

GEO

Gene Expression Omnibus

ND

Non-Diabetes

RMA

Robust Multi Array

FC

Fold Change

BH

Benjamini and Hochberg

TOM

Topological Overlap Matrix

DissTOM

TOM Dissimilarity Matrix

MEs

Module Eigengenes

MM

Module Membership

GS

Gene Significance

GO

Gene Ontology

BP

Biological Processes

CC

Cellular Component

MF

Molecular Function

KEGG

Kyoto Encyclopaedia of Gene and Genome

ROC

Receiver-Operating Characteristic Curve

ADT

AutoDock Tools

ACA

n-(p-Amylcinnamoyl) anthranilic acid

SDF

Spatial Data File

AD4

AutoDock4

SMILES

Simple Molecular-Input Line Entry System

LD50

Lethal Dosage

GROMACS

GROningen MAchine for Chemical Simulations

MD

Molecular Dynamics

NVT

Constant Number of Particles

Volume

And Temperature

NPT

Constant Number of Particles

Pressure

And Temperature

RMSD

Root Mean Square Deviation

RMSF

Root Mean Square Fluctuation

Hbond

Hydrogen Bond

RG

Radius of Gyration

SASA

Solvent Accessible Surface Area

MM-PBSA

Molecular Mechanic/Poisson-Boltzmann Surface Area

CPA1

Carboxypeptidase A1

CELA3B

Chymotrypsin-Like Elastage 3B

CPB1

Carboxypeptidase B1

CEL

Carboxy Ester Lipase

CTRB2

Chymotrypsinogen B2

PLA2G1B

Phospholipase A2 Group 1B

REG1A

Regenerating Family Member 1 Alpha

AUC

Area Under Curve

XlogP

Octanol/Water Partition Coefficient

RO5

Lipinski Rule Five

Acknowledgments

All the authors are grateful to their respective institutions for their support. R.G.K. would like to thank the Department of Science & Technology, Government of Odisha, India, for the Biju Patnaik Research Fellowship (Ref. No. ST-BT-MISC-0008-2018-1933/ST), for pursuing Ph.D. in Biotechnology (with registration ID: - 05-Biotechnology-2017-18), at Department of Biotechnology, Utkal University, Bhubaneswar, Odisha, India. We would also like to thank the Centre of Excellence in Integrated Omics and Computational Biology, Department of Biotechnology, Government of India, for providing infrastructural support to carry out the research work smoothly. G.D and J.KP. are grateful to Dongguk University, Republic of Korea for support.

Consent to publish statement

All authors consent to publish this manuscript in this esteemed journal.

Abraham MJ, Murtola T, Schulz R, Páll S, Smith JC, Hess B, et al. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX. 2015 Sep 1;1–2:19–25.
Abuohashish HM, Al-Rejaie SS, Al-Hosaini KA, Parmar MY, Ahmed MM. Alleviating effects of morin against experimentally-induced diabetic osteopenia. Diabetol. Metab. Syndr. 2013 Dec;5(1):5.
Afriza D, Suriyah WH, Ichwan SJA. In silico analysis of molecular interactions between the anti-apoptotic protein survivin and dentatin, nordentatin, and quercetin. J. Phys. Conf. Ser. IOP Publishing; 2018 Aug;1073(3):032001.
Ahmed O, Moneim AA, Yazid IA, Mahmoud A. Antihyperglycemic, antihyperlipidemic and antioxidant effects and the probable mechanisms of action of Ruta graveolens infusion and rutin in nicotinamide-streptozotocin-induced diabetic rats. 2010 [cited 2023 Aug 22]. Available from: https://www.semanticscholar.org/paper/Antihyperglycemic%2C-antihyperlipidemic-and-effects-Ahmed-Moneim/94a9f2784a123d6717b458b97f1f592134717d6c
Ashraf N, Asari A, Yousaf N, Ahmad M, Ahmed M, Faisal A, et al. Combined 3D-QSAR, molecular docking and dynamics simulations studies to model and design TTK inhibitors. Front. Chem. 2022 Nov 2;10:1003816.
Bachewal P, Gundu C, Yerra VG, Kalvala AK, Areti A, Kumar A. Morin exerts neuroprotection via attenuation of ROS induced oxidative damage and neuroinflammation in experimental diabetic neuropathy: pharmacological effects of morin in diabetic neuropathy. BioFactors. 2018 Mar;44(2):109–22.
Banerjee P, Dehnbostel FO, Preissner R. Prediction is a balancing act: Importance of sampling methods to balance sensitivity and specificity of predictive models based on imbalanced chemical data sets. Front. Chem. 2018a Aug 28;6:362.
Banerjee P, Eckert AO, Schrey AK, Preissner R. ProTox-II: A webserver for the prediction of toxicity of chemicals. Nucleic Acids Res. 2018b Jul 2;46(Web Server issue):W257–63.
Basile G, Vetere A, Hu J, Ijaduola O, Zhang Y, Liu K-C, et al. Excess pancreatic elastase alters acinar-β cell communication by impairing the mechano-signaling and the PAR2 pathways. Cell Metab. 2023 Jul 11;35(7):1242-1260.e9.
Batsika CS, Gerogiannopoulou A-DD, Mantzourani C, Vasilakaki S, Kokotos G. The design and discovery of phospholipase A2 inhibitors for the treatment of inflammatory diseases. Expert Opin. Drug Discov. Taylor & Francis; 2021 Nov 2;16(11):1287–305.
Bitencourt-Ferreira G, Veit-Acosta M, de Azevedo WF. Van der Waals potential in protein complexes. Methods Mol. Biol. Clifton NJ. 2019a;2053:79–91.
Bitencourt-Ferreira G, Veit-Acosta M, de Azevedo WF. Electrostatic energy in protein-ligand complexes. Methods Mol. Biol. Clifton NJ. 2019b;2053:67–77.
Brendel M, Su C, Bai Z, Zhang H, Elemento O, Wang F. Application of Deep Learning on Single-cell RNA Sequencing Data Analysis: A Review. Genomics Proteomics Bioinformatics. 2022 Oct;20(5):814–35.
Cárdenas-Aguayo M del C, Silva-Lucero M del C, Cortes-Ortiz M, Jiménez-Ramos B, Gómez-Virgilio L, Ramírez-Rodríguez G, et al. Physiological role of amyloid beta in neural cells: The cellular trophic activity. In: Heinbockel T, editor. Neurochemistry [Internet]. London, SW7 2QJ, UK: IntechOpen; 2014 [cited 2020 Dec 16]. p. 404. Available from: https://www.intechopen.com/books/neurochemistry/physiological-role-of-amyloid-beta-in-neural-cells-the-cellular-trophic-activity
Cash JG, Konaniah E, Hegde N, Kuhel DG, Watanabe M, Romick-Rosendale L, et al. Therapeutic reduction of lysophospholipids in the digestive tract recapitulates the metabolic benefits of bariatric surgery and promotes diabetes remission. Mol. Metab. 2018 Oct 1;16:55–64.
Cash JG, Kuhel DG, Goodin C, Hui DY. Pancreatic acinar cell-specific over-expression of group 1B phospholipase A2 exacerbates diet-induced obesity and insulin resistance in mice. Int. J. Obes. 2005. 2011 Jun;35(6):877–81.
Chen Z, Downing S, Tzanakakis ES. Four Decades After the Discovery of Regenerating Islet-Derived (Reg) Proteins: Current Understanding and Challenges. Front. Cell Dev. Biol. 2019;7:235.
Dludla PV, Mabhida SE, Ziqubu K, Nkambule BB, Mazibuko-Mbeje SE, Hanser S, et al. Pancreatic β-cell dysfunction in type 2 diabetes: Implications of inflammation and oxidative stress. World J. Diabetes. 2023 Mar 15;14(3):130–46.
Drwal MN, Banerjee P, Dunkel M, Wettig MR, Preissner R. ProTox: A web server for the in silico prediction of rodent oral toxicity. Nucleic Acids Res. 2014 Jul 1;42(Web Server issue):W53–8.
Dubey S, Ganeshpurkar A, Ganeshpurkar A, Bansal D, Dubey N. Glycolytic enzyme inhibitory and antiglycation potential of rutin. Future J. Pharm. Sci. 2017 Dec 1;3(2):158–62.
Guimaraes JFC, Muzio BP, Rosa CM, Nascimento AF, Sugizaki MM, Fernandes AAH, et al. Rutin administration attenuates myocardial dysfunction in diabetic rats. Cardiovasc. Diabetol. 2015 Dec;14(1):90.
Hayashi D, Dennis EA. Molecular basis of unique specificity and regulation of group VIA calcium-independent phospholipase A2 (PNPLA9) and its role in neurodegenerative diseases. Pharmacol. Ther. 2023 May;245:108395.
Hinnen D. Glucagon-Like Peptide 1 Receptor Agonists for Type 2 Diabetes. Diabetes Spectr. 2017 Aug 1;30(3):202–10.
Huang J, Rauscher S, Nawrocki G, Ran T, Feig M, de Groot BL, et al. CHARMM36m: An improved force field for folded and intrinsically disordered proteins. Nat. Methods. 2017 Jan;14(1):71–3.
Hui D, Cope M, Labonté E, Chang H-T, Shao J, Goka E, et al. The phospholipase A2 inhibitor methyl indoxam suppresses diet-induced obesity and glucose intolerance in mice. Br. J. Pharmacol. 2009 Aug;157(7):1263–9.
Ilie I, Ilie R, Mocan T, Tabaran F, Iancu C, Mocan L. Nicotinamide-functionalized multiwalled carbon nanotubes increase insulin production in pancreatic beta cells via MIF pathway. Int. J. Nanomedicine. 2013;8:3345–53.
Jiang B, Geng Q, Li T, Mohammad Firdous S, Zhou X. Morin attenuates STZ-induced diabetic retinopathy in experimental animals. Saudi J. Biol. Sci. 2020 Aug 1;27(8):2139–42.
Kadarmideen HN, Watson-Haigh NS, Andronicos NM. Systems biology of ovine intestinal parasite resistance: disease gene modules and biomarkers. Mol. Biosyst. 2011 Jan;7(1):235–46.
Kamalakkannan N, Prince PSM. Antihyperglycaemic and antioxidant effect of rutin, a polyphenolic flavonoid, in streptozotocin-induced diabetic wistar rats. Basic Clin. Pharmacol. Toxicol. 2006 Jan;98(1):97–103.
Kammoun M, Miladi S, Ali YB, Damak M, Gargouri Y, Bezzine S. In vitro study of the PLA2 inhibition and antioxidant activities of Aloe vera leaf skin extracts. Lipids Health Dis. 2011 Feb 11;10:30.
Kar P, Lipowsky R, Knecht V. Importance of polar solvation and configurational entropy for design of antiretroviral drugs targeting HIV-1 protease. J. Phys. Chem. B. 2013 May 16;117(19):5793–805.
Kerry RG, Mahapatra GP, Maurya GK, Patra S, Mahari S, Das G, et al. Molecular prospect of type-2 diabetes: Nanotechnology based diagnostics and therapeutic intervention. Rev. Endocr. Metab. Disord. [Internet]. 2020 Oct 14 [cited 2020 Nov 19]; Available from: https://doi.org/10.1007/s11154-020-09606-0
Kerry RG, Mohapatra P, Jena AB, Panigrahi B, Pradhan KC, Khatua BR, et al. Biosynthesis of rutin trihydrate loaded silica nanoparticles and investigation of its antioxidant, antidiabetic and cytotoxic potentials. J. Inorg. Organomet. Polym. Mater. [Internet]. 2022 Mar 7 [cited 2022 Mar 21]; Available from: https://doi.org/10.1007/s10904-022-02269-1
Kerry RG, Singh KR, Mahari S, Jena AB, Panigrahi B, Pradhan KC, et al. Bioactive potential of morin loaded mesoporous silica nanoparticles: A nobel and efficient antioxidant, antidiabetic and biocompatible abilities in in-silico, in-vitro, and in-vivo models. OpenNano. 2023 Mar 1;10:100126.
Kuefner MS. Secretory Phospholipase A2s in Insulin Resistance and Metabolism. Front. Endocrinol. 2021 Aug 27;12:732726.
Kumari R, Kumar R, Lynn A. g_mmpbsa—A GROMACS tool for high-throughput MM-PBSA calculations. J. Chem. Inf. Model. American Chemical Society; 2014 Jul 28;54(7):1951–62.
Labonté ED, Kirby RJ, Schildmeyer NM, Cannon AM, Huggins KW, Hui DY. Group 1B Phospholipase A2–Mediated Lysophospholipid Absorption Directly Contributes to Postprandial Hyperglycemia. Diabetes. 2006 Apr;55(4):935–41.
Lemak AS, Balabaev NK. On the berendsen thermostat. Mol. Simul. Taylor & Francis; 1994 Sep 1;13(3):177–87.
Lipinski CA. Rule of five in 2015 and beyond: Target and ligand structural limitations, ligand chemistry structure and drug discovery project decisions. Adv. Drug Deliv. Rev. 2016 Jun 1;101:34–41.
Marselli L, Thorne J, Dahiya S, Sgroi DC, Sharma A, Bonner-Weir S, et al. Gene expression profiles of Beta-cell enriched tissue obtained by laser capture microdissection from subjects with type 2 diabetes. PloS One. 2010 Jul 13;5(7):e11499.
Meng X-Y, Zhang H-X, Mezei M, Cui M. Molecular docking: A powerful approach for structure-based drug discovery. Curr. Comput. Aided Drug Des. 2011 Jun 1;7(2):146–57.
Mouchlis VD, Limnios D, Kokotou MG, Barbayianni E, Kokotos G, McCammon JA, et al. Development of potent and selective inhibitors for group VIA calcium-independent phospholipase A2 guided by molecular dynamics and structure-activity relationships. J. Med. Chem. 2016 May 12;59(9):4403–14.
Mumford JA, Horvath S, Oldham MC, Langfelder P, Geschwind DH, Poldrack RA. Detecting network modules in fMRI time series: A weighted network analysis approach. NeuroImage. 2010 Oct 1;52(4):1465–76.
Niture NT, Ansari AA, Naik SR. Anti-hyperglycemic activity of rutin in streptozotocin-induced diabetic rats: an effect mediated through cytokines, antioxidants and lipid biomarkers. Indian J. Exp. Biol. 2014 Jul;52(7):720–7.
Ola MS, Ahmed MM, Ahmad R, Abuohashish HM, Al-Rejaie SS, Alhomida AS. Neuroprotective Effects of Rutin in Streptozotocin-Induced Diabetic Rat Retina. J. Mol. Neurosci. 2015 Jun 1;56(2):440–8.
Ola MS, Aleisa AM, Al-Rejaie SS, Abuohashish HM, Parmar MY, Alhomida AS, et al. Flavonoid, morin inhibits oxidative stress, inflammation and enhances neurotrophic support in the brain of streptozotocin-induced diabetic rats. Neurol. Sci. 2014 Jul;35(7):1003–8.
Omeje KO, Ezema BO, Onaebi CN, Onoyima SC, Ezeorba TPC, Eze SOO. HPLC fingerprint of flavonoids, enzyme inhibition and antioxidant activity of Newbouldia laevis stem-bark: an in vitro and in silico study. Future J. Pharm. Sci. 2023 Apr 24;9(1):36.
Pantsar T, Poso A. Binding affinity via docking: Fact and fiction. Mol. J. Synth. Chem. Nat. Prod. Chem. 2018 Jul 30;23(8):1899.
Prince PSM, Kamalakkannan N. Rutin improves glucose homeostasis in streptozotocin diabetic tissues by altering glycolytic and gluconeogenic enzymes. J. Biochem. Mol. Toxicol. 2006;20(2):96–102.
Quackenbush J. Microarray data normalization and transformation. Nat. Genet. Nature Publishing Group; 2002 Dec;32(4):496–501.
Rezaei Z, Ranjbaran J, Safarpour H, Nomiri S, Salmani F, Chamani E, et al. Identification of early diagnostic biomarkers via WGCNA in gastric cancer. Biomed. Pharmacother. 2022 Jan 1;145:112477.
Richmond BL, Boileau AC, Zheng S, Huggins KW, Granholm NA, Tso P, et al. Compensatory phospholipid digestion is required for cholesterol absorption in pancreatic phospholipase A2–Deficient mice. Gastroenterology. 2001 Apr 1;120(5):1193–202.
Sarkar MS, Mia MM, Amin MA, Hossain MS, Islam MZ. Bioinformatics and network biology approach to identifying type 2 diabetes genes and pathways that influence the progression of breast cancer. Heliyon. 2023 May;9(5):e16151.
Sattanathan K, Dhanapal CK, Umarani R, Manavalan R. Beneficial health effects of rutin supplementation in patients with diabetes mellitus. J. Appl. Pharm. Sci. 2011 Oct 30;:(Issue):227–31.
Sendrayaperumal V, Iyyam Pillai S, Subramanian S. Design, synthesis and characterization of zinc–morin, a metal flavonol complex and evaluation of its antidiabetic potential in HFD–STZ induced type 2 diabetes in rats. Chem. Biol. Interact. 2014 Aug;219:9–17.
Shaikh F, shah T, Madkhali NAB, Gaber A, Alsanie WF, Ali S, et al. Frequency distribution and association of Fat-mass and obesity (FTO) gene SNP rs-9939609 variant with Diabetes Mellitus Type-II population of Hyderabad, Sindh, Pakistan. Saudi J. Biol. Sci. 2021 Aug 1;28(8):4183–90.
Shali KS, Soumya NPP, Mondal S, Mini S. Hepatoprotective effect of morin via regulating the oxidative stress and carbohydrate metabolism in STZ induced diabetic rats. Bioact. Compd. Health Dis. 2022 Mar 16;5(3):53.
Shin A, Connolly S, Kabytaev K. Protein glycation in diabetes mellitus. Adv. Clin. Chem. [Internet]. Elsevier; 2023 [cited 2023 May 12]. p. 101–56. Available from: https://linkinghub.elsevier.com/retrieve/pii/S0065242322000993
Shojima N, Yamauchi T. Progress in genetics of type 2 diabetes and diabetic complications. J. Diabetes Investig. 2023 Apr;14(4):503–15.
Sohail MU, Mashood F, Oberbach A, Chennakkandathil S, Schmidt F. The role of pathogens in diabetes pathogenesis and the potential of immunoproteomics as a diagnostic and prognostic tool. Front. Microbiol. 2022 Nov 14;13:1042362.
Staiger H, Machicao F, Fritsche A, Häring H-U. Pathomechanisms of Type 2 Diabetes Genes. Endocr. Rev. 2009 Oct 1;30(6):557–85.
Sun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Duncan BB, et al. IDF Diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res. Clin. Pract. 2022 Jan;183:109119.
Suzuki Y, Kaneko H, Okada A, Itoh H, Matsuoka S, Fujiu K, et al. Comparison of cardiovascular outcomes between SGLT2 inhibitors in diabetes mellitus. Cardiovasc. Diabetol. 2022 Dec;21(1):67.
Syaifie PH, Harisna AH, Nasution MAF, Arda AG, Nugroho DW, Jauhar MM, et al. Computational study of Asian propolis compounds as potential anti-type 2 diabetes mellitus agents by using inverse virtual screening with the DIA-DB web server, tanimoto similarity analysis, and molecular dynamic simulation. Mol. Basel Switz. 2022 Jun 21;27(13):3972.
Tamura K, Yu J, Hata T, Suenaga M, Shindo K, Abe T, et al. Mutations in the pancreatic secretory enzymes CPA1 and CPB1 are associated with pancreatic cancer. Proc. Natl. Acad. Sci. U. S. A. 2018 May 1;115(18):4767–72.
Tang Y, Axelsson AS, Spégel P, Andersson LE, Mulder H, Groop LC, et al. Genotype-based treatment of type 2 diabetes with an α2A-adrenergic receptor antagonist. Sci. Transl. Med. 2014 Oct 8;6(257):257ra139.
Tantar A-A, Conilleau S, Parent B, Melab N, Brillet L, Roy S, et al. Docking and biomolecular simulations on computer grids: Status and trends. Curr. Comput. Aided Drug Des. 2008;4(3):235–49.
Tap FM, Abd Majid FA, Ismail HF, Wong TS, Shameli K, Miyake M, et al. In silico and in vitro study of the bromelain-phytochemical complex inhibition of phospholipase A2 (PLA2). Mol. Basel Switz. 2018 Jan 19;23(1):73.
Tiwari A, Rathor P, Trivedi P, Ch R. Multi-Omics Reveal Interplay between Circadian Dysfunction and Type2 Diabetes. Biology. 2023 Feb 14;12(2):301.
Trott O, Olson AJ. AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 2010 Jan 30;31(2):455–61.
Vasquez AM, Mouchlis VD, Dennis EA. Review of four major distinct types of human phospholipase A2. Adv. Biol. Regul. 2018 Jan;67:212–8.
Vinayagam R, Xu B. Antidiabetic properties of dietary flavonoids: a cellular mechanism review. Nutr. Metab. 2015 Dec 23;12(1):60.
Voineagu I, Wang X, Johnston P, Lowe JK, Tian Y, Horvath S, et al. Transcriptomic Analysis of Autistic Brain Reveals Convergent Molecular Pathology. Nature. 2011 May 25;474(7351):380–4.
Wei L, Han Y, Tu C. Molecular Pathways of Diabetic Kidney Disease Inferred from Proteomics. Diabetes Metab. Syndr. Obes. 2023 Jan;Volume 16:117–28.
Wihastuti TA, Aini FN, Tjahjono CT, Sulfia YH, Sholichah Z, Heriansyah T. <p>Lp-PLA₂ Selective Inhibitor (Darapladib) Effect In Lowering The Expression Level Of IL-1B And IL-6 In The Renal At Type 2 Diabetes Mellitus</p>. Vasc. Health Risk Manag. Dove Press; 2019 Nov 1;15:503–8.
Wu H, Shu M, Liu C, Zhao W, Li Q, Song Y, et al. Identification and characterization of novel carboxyl ester lipase gene variants in patients with different subtypes of diabetes. BMJ Open Diabetes Res. Care. BMJ Specialist Journals; 2023 Jan 1;11(1):e003127.
Xu Y, Huang J. Validating the CHARMM36m protein force field with LJ-PME reveals altered hydrogen bonding dynamics under elevated pressures. Commun. Chem. 2021 Jun 28;4(1):99.
Zeng L, Zhang G, Liao Y, Gong D. Inhibitory mechanism of morin on α-glucosidase and its anti-glycation properties. Food Funct. 2016;7(9):3953–63.
Zhang M-Q, Wilkinson B. Drug discovery beyond the “rule-of-five.” Curr. Opin. Biotechnol. 2007 Dec;18(6):478–88.

No competing interests reported.

Exploring the therapeutic potential of Rutin and Morin in Type 2 Diabetes: A transcriptomics and molecular dynamics simulation for proteins

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and Methods

3. Results

4. Discussion

5. Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1