Unravelling the mechanisms of food-derived antidiabetic peptides: Insights from cell-based models

Food-derived antidiabetic peptides (ADPs) have emerged as promising natural interventions for the management of diabetes mellitus. Cell-based assays are essential for evaluating their antidiabetic properties, facilitating both functional screening and mechanistic elucidation. While conventional monocultures have delineated key signalling pathways, including the IRS-1/PI3K/Akt and AMP-activated protein kinase (AMPK) axes, recent methodological advances, such as three-dimensional (3D) culture and co-culture platforms, and clustered regularly interspaced short palindromic repeats (CRISPR)-based gene editing, offer more physiologically relevant models to characterise peptide activity. This review synthesises current evidence from widely employed cell lines, such as HepG2, Caco-2, and INS-1, and critically examines technological innovations that address existing limitations in peptide validation. Furthermore, we evaluate the integration of omics technologies and machine learning with cellular platforms to deepen the mechanistic understanding and enhance predictive accuracy of ADPs. By identifying critical research gaps, including the need to integrate in vitro -to- in vivo extrapolation (IVIVE) and physiologically based pharmacokinetic (PBPK) modelling, this review provides a structured framework for incorporating advanced in vitro tools into the rational development of peptide-based therapeutics, nutraceuticals, and functional foods.