Antimicrobial peptides selected through deep learning can be utilized for the control of Pseudomonas fluorescens in...

Area of Science:

• Food science
• Microbiology
• Bioinformatics

Background:

• Psychrotrophic Pseudomonas species contaminate milk, leading to dairy product spoilage.
• Antimicrobial peptides (AMPs) offer a safe and effective alternative to conventional preservatives.
• Current AMP screening methods are inefficient and resource-intensive.

Purpose of the Study:

• To develop a deep learning framework for predicting AMPs against Pseudomonas species.
• To identify and synthesize novel AMP candidates.
• To evaluate the efficacy of identified AMPs in inhibiting Pseudomonas growth and preserving milk.

Main Methods:

• An LSTM-based deep learning model was developed for predictive identification of AMPs.
• Computational screening identified four candidate AMPs.
• Candidates were chemically synthesized and tested for inhibitory activity against Pseudomonas fluorescens W3.
• Mechanism of action was investigated using NPN and PI probe labeling.
• Effectiveness in a contaminated milk model was assessed by measuring microbial biomass and protease activity.

Main Results:

• The deep learning framework successfully predicted AMPs with inhibitory activity.
• Three out of four synthesized AMPs showed significant inhibition against Pseudomonas fluorescens W3.
• The peptide WY-1 disrupted bacterial membrane integrity, leading to cell death.
• WY-1 significantly reduced microbial load and protease activity in contaminated milk.

Conclusions:

• Deep learning provides an efficient approach for identifying novel AMPs.
• WY-1 demonstrates potent antimicrobial activity and potential for dairy preservation.
• This study offers a promising strategy for combating milk spoilage caused by Pseudomonas species.