Organic Fusion of Molecular Simulation and Wet-Lab Validation: A Promising High-Throughput Strategy for Screening Bioactive Food Peptides

Abstract

Peptides derived from protein sources in food exhibit a diverse array of biological activities. The screening, preparation, and functional investigation of bioactive peptides have become a focal area of research. This review summarizes the status of peptide activity mining, including the latest research progress in protein sources, peptide functions, and processing conditions. It critically evaluates the limitations of current bioactive peptide screening methods, including the drawbacks of traditional methods and molecular simulations. The potential of using molecular simulation for the virtual screening of potentially bioactive peptides is summarized. This includes virtual enzymatic digestion, molecular docking, simulation of non-thermal processing technologies, and the construction of organelle/cell models. The driving role of artificial intelligence in molecular simulation is also discussed. In addition, the structural information, mechanism, and structural analysis technique of action of the popular target proteins of foodborne bioactive peptides are summarized to provide a better reference for virtual-reality combinations.

Keywords: artificial intelligence; bioactive peptide; food protein; molecular simulation; non-thermal technology; receptor protein.

Figure 1

Research progress of food source peptides in the past 5 years. Different colors indicate different clusters, and the top 10 most researched directions or fields are listed according to keywords and references. Figure made by CiteSpace (6.3. R1).

Figure 2

Partial distribution of species producing different functional bioactive peptides. Aquatic products do not distinguish between seawater and fresh water. Figure made by BioRender (agreement number: WL28ICF78U).

Figure 3

The mechanism or distribution of action of protein receptors. (A) Keap1-NrF2 oxidative stress pathway involved in keap1; (B) MYD888 and NF-κB in which TLR4 is involved; (C) distribution of 5-HT receptors in human organs; (D) renin–angiotensin system and kallikrein–kinin system in which ACE participates; (E) melanin biosynthesis involving tyrosinase [59]; (F) part of purine dehydrogenase metabolic pathways involved in xanthine dehydrogenase/oxidase. Figure made by BioRender (agreement number: SZ28ICEP4K).

Figure 4

Overview of virtual enzymolysis, molecular docking, molecular dynamics simulation, construction of non-thermal processing techniques, and virtual cell models. (A) Common enzymes used in virtual enzymatic hydrolysis and their cleavage sites on amino acid sequences. (B) The mutual verification of traditional experimental and molecular simulation for bioactive peptide mining. (C) The construction of non-thermal processing techniques in molecular simulation, taking ultrasonic processing as an example (reproduced with permission from Fu et al., Journal of Physical Chemistry Letters; published by American Chemical Society 2015) [78]. (D) The development process of simulated organelles: from the past (lipid bilayer) to the future (AI organs) (some pictures obtained from [3,79,80]. Reproduced with permission from Shamloo et al., Journal of Magnetism and Magnetic Materials; published by Elsevier 2016 [79]. Figure made by BioRender (agreement number: IN28NA81P9).

Figure 5

Comparison between molecular simulation and traditional method of bioactive peptide mining. Figure made by BioRender (agreement number: BU28ICEZO6).