A digital platform for One-Pot signal enhanced foodborne pathogen detection based on mesophilic argonaute-driven polydisperse microdroplet reactors and machine learning

Foodborne pathogens endanger human health. Therefore, sensitive and accurate detection assays are key for their prevention and control. Microdroplet digital nucleic acid platforms can achieve single-molecule sensitivity in the presence of specialized instrumentation for generating homogenized monodisperse droplets, expensive thermal cyclers, high temperature-resistant reagents, and time-consuming multistep operations in the lab. Argonaute (Ago), an emerging next-generation gene editing tool, can be used for the detection of pathogen genetic information. Current reports mainly adopt the high temperature-dependent thermophilic Ago in stepwise nucleic acid detection, which hinders its applications in ultrasensitive digital detection. Here, we expressed and purified the mesophilic Clostridium butyricum Argonaute (CbAgo) and developed a polydisperse microdroplet reactors-assist asymmetric recombinase polymerase amplification driven CbAgo report digital detection platform (DRAGON). It realized the signal-enhanced nucleic acid amplification and CbAgo report compatibility in novel low-cost and ultrafast-prepared polydisperse microdroplet reactors. The microdroplet reactor area and fluorescence intensity were combined to establish a dual parameter-based random forest machine learning model, which achieved high sensitivity (1 CFU/mL) and efficiency (<45 min) for detecting pathogens. This is the first time that mesophilic Ago was used to achieve true sensitive one-step and one-pot low-temperature digital nucleic acid detection. Our system accomplished multiplex detection in one-step and one-pot with a single nuclease which is impossible for CRISPR/Cas system and thermophilic Ago-based stepwise platforms. DRAGON represents the next-generation platform for pathogen detection based on gene editing tools and intelligent digital nucleic acid detection system.