Type 2 diabetes mellitus (T2DM), a metabolic disorder characterized by insulin resistance and impaired glucose tolerance, accounts for 90% of global diabetes cases, with projections suggesting that this number will reach 592 million by 2035 [1]. T2DM is associated with severe complications including retinopathy, nephropathy, and cardiovascular diseases. Current therapeutic strategies rely heavily on α-glucosidase inhibitors (e.g., acarbose, voglibose) to retard carbohydrate digestion. However, these agents frequently elicit gastrointestinal adverse effects, which compromise long-term patient adherence [2]. Consequently, the development of natural glucosidase inhibitors has gained increasing importance for T2DM management, owing to their favorable safety profiles and efficacy [3], [4].

Bioactive peptides exhibit distinct advantages for T2DM prevention and management, attributed to their low toxicity, high specificity, and multifunctional properties. These peptides modulate blood glucose via two key mechanisms: competitive binding to the active sites of α-glucosidase to suppress carbohydrate hydrolysis, and amelioration of insulin resistance through modulation of the insulin signaling pathway. Liu et al. [5] showed via enzyme kinetic analyses that the 1–3 kDa rice bran peptide fraction (IC50 = 1.17 mg/mL), generated by alkaline and dispase-catalyzed hydrolysis, exerted a competitive inhibitory mechanism against α-glucosidase.

The distinct structural and functional characteristics of animal and plant-derived peptides limit their individual applicability. While plant peptides such as those from soy offer benefits in availability, cost, and safety, animal-based peptides like casein provide complete amino acid profiles and high bioavailability, despite associated allergenic potential and sourcing limitations [6]. This study developed a composite system combining soy and casein peptides. Through screening and optimization, a formulation with enhanced bioactivity was obtained. This synergistic approach improves functional properties, including α-glucosidase inhibition, while mitigating adverse effects.

Virtual screening anchored by molecular docking has become a crucial foundation for exploring bioactive peptides [7], [8]. Molecular docking is an essential component that reduces false positives and enhances hit identification efficacy, as demonstrated by Tondar et al. [9], [10]. The Caco-2 cell monolayer model with its biomimetic intestinal epithelial traits remains indispensable for investigating nutrient absorption and metabolism [11]. Animal oral glucose tolerance tests (OGTT) provide vital evidence for hypoglycemic efficacy [12], yet direct extrapolation to humans is challenging. Notably, most existing studies on food-derived peptides are confined to in vitro or animal models, lacking systematic validation across hierarchical scales. Human trials remain scarce due to high costs, ethical constraints, and metabolic variability. In contrast, this work integrates virtual screening with molecular docking, in vitro Caco-2 cell assays, in vivo murine OGTT and human postprandial blood glucose tests to conduct a comprehensive, multi-tiered evaluation of dual-protein composite peptides, an integrated approach that has rarely been reported previously.

This study aimed to determine the optimal preparation conditions for dual-protein peptides and investigate their underlying mechanisms. Various protein sources and protease preparations were screened, followed by ultrafiltration fractionation and characterization via liquid chromatography-tandem mass spectrometry (LC-MS/MS) combined with amino acid composition analysis. Furthermore, molecular docking and virtual screening were employed to elucidate the potential interaction mechanisms and binding sites of the composite dual-source peptides with α-glucosidase. The in vivo and in vitro activity mechanism were comprehensively verified using the Caco-2 cell monolayer model, OGTT and the human postprandial blood glucose test. This multi-hierarchical validation framework integrates evidence from in vitro cellular models, in vivo animal experiments, and human physiological contexts to validate biological effects across scales. It thereby not only enhances the reliability of bioactive peptide discovery and provides a solid foundation for optimizing evidence-based, non-pharmacological diabetes management strategies, but also establishes a direct translational path for developing functional food ingredients.