A bacterial biosensor based on gold nanoparticles functionalized by a hydrophobin-chimera and combined with machine learning for user-friendly detection

A chimeric protein, in which a cationic antimicrobial peptide was fused to the adhesive self-assembling moiety of a fungal class I hydrophobin, was produced and used to easily functionalize gold nanoparticles (f-AuNPs). f-AuNPs are able to bind Escherichia coli and Staphylococcus epidermidis bacteria, tested as model organisms. A reverse colorimetric assay was then set up based on the induced aggregation of the free f-AuNPs. This leads to changes in optical properties of the solution, visible even by naked eyes. This approach allows detection of 10 CFU mL-1 of E. coli within 15 min and can be applicable to detect bacterial contamination in real matrices. Notably, the colorimetric variation was combined to machine-learning methodologies inferring data from photos of the colored solutions, acquired by smartphones. This fast, highly practical and flexible system, would enable on-site measurements and, therefore, help to prevent significant contaminations.