Virtual screening of a milk peptide database for the identification of food-derived antimicrobial peptides

Scope: Milk provides a wide range of bioactive substances, such as antimicrobial peptides and proteins. Our study aimed to identify novel antimicrobial peptides naturally present in milk. Methods and results: The components of an endogenous bovine milk peptide database were virtually screened for charge, amphipathy, and predicted secondary structure. Thus, 23 of 248 screened peptides were identified as candidates for antimicrobial effects. After commercial synthesis, their antimicrobial activities were determined against Escherichia coli NEB5 alpha, E. coli ATCC25922, and Bacillus subtilis ATCC6051. In the tested concentration range (<2 mM), bacteriostatic activity of 14 peptides was detected including nine peptides inhibiting both Gram-positive and Gram-negative bacteria. The most effective fragment was TKLTEEEKN-RLNFLKKISQRYQKFALPQYLK corresponding to alpha(S2)-casein(151-181), with minimum inhibitory concentration (MIC) of 4.0 mu M against B. subtilis ATCC6051, and minimum inhibitory concentrations of 16.2 mu M against both E. coli strains. Circular dichroism spectroscopy revealed conformational changes of most active peptides in a membrane-mimic environment, transitioning from an unordered to alpha-helical structure. Conclusion: Screening of food peptide databases by prediction tools is an efficient method to identify novel antimicrobial food-derived peptides. Milk-derived antimicrobial peptides may have potential use as functional food ingredients and help to understand the molecular mechanisms of anti-infective milk effects.