As global freshwater supplies become impacted, water quality is an increasingly critical concern. Human activity has introduced toxins, particularly heavy metals, into many natural water supplies across the globe. Bacteria that grow in environments with high toxin levels, occurring either naturally or as a result of pollution, have evolved pathways to mitigate their effects. We use these native pathways and synthetic biology techniques to engineer bacteria that produce fluorescent responses in the presence of various toxins.
Our process combines this genetic engineering with microfluidic technology and mathematical classification to develop a highly sensitive and specific platform for sensing contaminants in water. We use microfluidics to integrate many independent sensor strains within a self-contained platform that can continuously monitor water supplies for contamination. Our device consists of an array of cell trapping units connected in parallel, each supporting the long-term culturing of an individual cell strain. We then use machine learning models to infer the relationships between the fluorescent sensor responses and the presence of a toxin at a specific concentration. The flexibility of our approach provides a framework for future expansion and customization.