Decoding bacterial identity and resistance through D-amino acid metabolic fingerprinting with a cascade colorimetric sensor array.

The rapid rise of antimicrobial resistance (AMR) constitutes a global health crisis, demanding diagnostic tools capable of simultaneous bacterial identification and rapid antibiotic susceptibility testing (AST) in a single workflow. Herein, a colorimetric sensor array was developed to address this need by profiling bacterial metabolic fingerprints via a streamlined single-substrate, multi-wavelength strategy. The strategy is built upon the inherent bacteria-specific metabolism of D-amino acids (D-AAs), coupled with the enzymatic selectivity of D-amino acid oxidase (DAAO), which catalyzes the oxidation of D-AAs to generate H2O2. Differential bacterial consumption of D-Ala directly modulates the substrate available for the subsequent DAAO-coupled cascade reaction, producing varying amounts of H2O2. Consequently, H2O2 triggers the peroxidase (POD)-like activity of Prussian blue-coated gold nanoparticle (Au@PB NP) nanozyme to produce distinct multi-wavelength optical patterns from TMB oxidation. When coupled with linear discriminant analysis (LDA), this sensing platform enables clear discrimination of bacterial species based on their metabolic fingerprints. Moreover, by assessing the antibiotic-induced inhibition of this metabolic activity, the platform accurately determines bacterial susceptibility profiles, achieving 100% concordance with the standard disk diffusion method for both reference and clinical isolates, and successfully determining minimum inhibitory concentration (MIC) values for reference strains in full agreement with Clinical Laboratory Standards Institute (CLSI) breakpoint interpretation. This work establishes an integrated, rapid, and robust metabolic sensing platform for bacterial phenotyping and AST, offering a promising tool to combat AMR.