Metabolomic profiling, computational and experimental validation of sunflower seeds as therapeutics against type-2diabetes mellitus

Abstract

Type 2 diabetes mellitus (T2DM) is a chronic metabolic disorder characterized by impaired glucose metabolism due to insufficient insulin secretion or insulin resistance. This global health crisis is projected to affect an estimated 7079 individuals per 100,000 by 2030. While medications like metformin are effective, accessibility and affordability are issues consistent with low-income populations alongside potential side effects like hypoglycaemia, nausea and gastrointestinal issues that have limited their use in clinical practice. More importantly, uncontrolled T2DM can lead to serious complications like retinopathy, nephropathy, neuropathy, and delayed wound healing. Therefore, this prompts the search for alternative management options that are safer, easily accessible, affordable and with minimal side effects. Plants and their products are becoming increasingly important due to their relative ease of accessibility, affordability and potential health benefits. Sunflower seed, a popular dietary snack, has rich nutritional profile and has found significant health benefits as an antiinflammatory, antioxidant, anticancer, antimicrobial, and antidiabetic agent. While the antidiabetic potential of sunflower seeds has been explored, there remains a lack of understanding on its mechanism of action. This study addressed this knowledge gap by establishing the comprehensive metabolite profiles and investigating the antidiabetic efficacy of sunflower seed extracts through a two-pronged approach: targeted enzyme inhibition and network pharmacology analyses complemented with experimental validation in vitro. Metabolomic profiling of six cultivars of sunflower seeds commonly consumed in South Africa, namely, AGSUN 8251, 5270, 5101 CLP, 5103 CLP, 5106 CLP and 5108 CLP was performed using Liquid chromatography – mass spectrometry (LC-MS) and Gas chromatography – mass spectrometry (GC-MS) techniques. A total of 94 metabolites were identified, with LC-MS analysis revealing 44 phenolic compounds across the six cultivars with a minor variance of 39.7%, while GC-MS analysis revealed the presence of volatile compounds such as organic acids, alkanes, alcohols, terpenes, heterocyclic compounds and hydrocarbons in all the cultivars in similar abundance. Noteworthily, 84 of the 94 metabolites profiled passed Lipinski’s rule of five and were selected for further analysis. For the enzyme inhibition study, molecular docking analysis was initially used to screen the profiled metabolites against the key enzymes [α-amylase (AAMY), α-glucosidase (AGLU), aldose reductase (AR), sorbitol dehydrogenase (SDH), dipeptidyl peptidase 4 (DPP-4) and protein tyrosine phosphatase 1B (PTP1B)] implicated in T2DM pathogenesis and its secondary complications. The top-ranked metabolites against each enzyme were further subjected to molecular dynamics (MD) simulation to identify putative leads with the strongest binding affinity, and unperturbed structural integrity through evaluation of their stability, compactness and intermolecular interactions. This aspect of the study identified sonchuside I (SON I) - AAMY (–47.26 kcal/mol), sacranoside A (SAC A) - α-glucosidase (–40.10 kcal/mol), pelatoside A (PLT) - AR (–58.84 kcal/mol), sacranoside A (SAC A) - SDH (–48.03 kcal/mol), 4α,6S,7α)-6α-[6-O-(4-Hydroxybenzoyl)-β-D-glucopyranosyloxy]-7βmethyloctahydrocyclopenta[c]pyran-1-one) (PYR) -DPP-4 (–37.93 kcal/mol) and chlorogenic acid (CGA)-PTP1B (–24.32 kcal/mol) as potential lead inhibitors of the respective enzyme relative to their respective reference standards. This was further supported by their improved thermodynamic properties and favourable post-dynamic simulation parameters such as improved stability and compactness of their resulting complexes. These observations are suggestive of multiple mechanisms by which sunflower seed may exert its antidiabetic effects such as anti-hyperglycaemia (α-amylase and α-glucosidase), prevention and management of diabetic complications (AR and SDH), increasing insulin signalling (DPP-4) and sensitivity (PTP1B) by the respective putative leads. For network pharmacology analysis, the filtered sunflower seed metabolites were used to create a gene-compound library that was subsequently used to identify genes commonly associated with both the metabolites and T2DM. Thereafter, Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed to identify the most significantly enriched pathways with key target genes for molecular docking and MD simulations to identify lead metabolites. Finally, the antidiabetic activity of sunflower seed extracts and the findings from the network pharmacology analysis were validated using insulin-resistant HepG2 cells where glucose consumption assay and gene expression analysis were performed. The network pharmacology analysis revealed a total of 87 genes common to sunflower seeds metabolites and T2DM, whereas KEGG enrichment analysis highlighted 35 signalling pathways potentially influenced by the metabolites. Of these, the Peroxisome proliferator-activated receptor (PPAR) signalling pathway and its hub receptors, Matrix metalloproteinase-1(MMPI) and peroxisome proliferator-activated receptor alpha (PPAR) were selected as the most significant. These receptors interacted mostly with the identified metabolites, with CGA (– 43.74 kcal/mol), GPA (–41.62 kcal/mol), and CFG (–45.36 kcal/mol) having lower binding free energy than both reference standards, rosiglitazone (ROS) and metformin (MET) against MMP1 after 100 000 ps MD simulation. In contrast, ROS (–46.98 kcal/mol) had better affinity against PPARA compared to the top-hits derived from sunflower seeds. However, against both genes, the top-hits had significant thermodynamic stability, flexibility, and compactness, which are attributable to their bond interactions and molecular orbital properties. These findings are suggestive of the essential role of the top-hits in the antidiabetic potential of sunflower seeds through activation of the PPAR signalling pathway and most especially MMP1. In this regard, the modulation of MMP1 and PPARA genes by the identified metabolites of sunflower seeds may enhance insulin sensitivity and glucose homeostasis in the management of T2DM. Finally, the in vitro validation using insulin-resistant HepG2 cells revealed cultivar-specific effects on cell viability, with each cultivar having a unique optimal concentration. Overall, all cultivars demonstrated the ability to stimulate glucose consumption, suggesting their potential antihyperglycemic activity. Among the cultivars, AGSUN 5103 CLP (14.4 mmol/L), 8251 (14.6 mmol/L), and 5101 CLP (13.7 mmol/L) exhibited the most pronounced glucose lowering action compared to the untreated cells (23.3 mmol/L) after 24 h, highlighting their promising antidiabetic effects. These three cultivars also modulate the PPAR signalling pathway, as evidenced by the upregulation of MMP1 and PPARA expression. Specifically, AGSUN 5101 CLP emerged as a particularly promising candidate based on its superior glucose lowering potential and higher fold increase expression of MMP1 (1.88) and PPARA (4.59) compared to the effect observed with the untreated cells (1.00). In conclusion, this study provides compelling evidence for the antidiabetic potential of sunflower seeds. The observed effects on enzyme inhibition, activation of the PPAR signalling pathway, and stimulation of glucose uptake in HepG2 cells suggest a multifaceted approach by the seeds in regulating blood sugar levels. The identification of cultivar-specific effects and promising lead compounds warrants further investigation to explore the therapeutic potential of sunflower seeds in managing T2DM.