Bispecific Metabolic Monitoring Platform for Bacterial Identification and Antibiotic Susceptibility Testing

Prompt and reliable bacterial identification and antibiotic susceptibility testing are vital for combating bacterial infections and drug resistance. Herein, we designed a bispecific metabolic monitoring platform that targets enzyme-catalyzed biochemical reactions for bacterial identification and antibiotic susceptibility testing. Specifically, we designed two kinds of coreshell-structured persistent luminescence nanoparticles with surface-confined red and green persistent luminescence, respectively. The persistent luminescence nanoparticles were functionalized with energy acceptors that can be specifically cleaved by bacterial enzymes. The surface-confined persistent luminescence amplified the Förster resonance energy transfer (FRET) efficacy from the nanoparticles to the surface energy acceptors, even though the diameter of the nanoparticles exceeded the critical size of FRET, which improved the sensitivity of bacterial enzyme monitoring. Due to the differentiated expression and secretion of enzymes, different species of bacteria produced discrepant red and green persistent luminescence after incubation with the persistent luminescence nanoprobes. Machine learning models were trained by the characteristic persistent luminescence patterns of bacteria for unknown bacterial identification. Prompt bacteria identification was realized, and the overall accuracy reached 100%. Moreover, the machine learning model could identify the active and inactive states of bacteria treated with antibiotics, which provided a prompt and convenient method to determine whether the bacteria were susceptible to the antibiotics. This study provides a robust method to monitor bacterial metabolism and offers a promising strategy for infection treatment, bacterial communication monitoring, and pathogenicity investigation.