Fast detection and quantification of Escherichia coli using the base principle of the microbial fuel cell

• We adapted an MFC technique to use as a detection unit of an Escherichia coli sensor.
• High voltage output of an MFC was generated at the critical concentration of E. coli.
• We could calculate the quantity of E. coli in the sample without false-positive results.
• Detection could be made within a working day's schedule by the GUS induction.

Escherichia coli is an important microbial indicator of fecal contamination, making accurate quantitative detection of E. coli a key to ensuring public health. In this study, a microbial fuel cell (MFC) was used as a detection unit of an E. coli sensor, and specific enzymes expressed in E. coli, such as β-D-galactosidase (GAL) and β-D-glucuronidase (GUS), were exploited as biological detection elements. As substrates, 4-aminophenyl-β-D-galactopyranoside (4-APGal) were used for GAL detection, whereas 8-hydroxyquinoline glucuronide (8-HQG) and 4-nitrophenyl β-D-glucuronide (PNPG) were used for GUS detection. Once these substrates were hydrolyzed by GAL or GUS, they became electrochemically active products, which were, in turn, oxidized on the anode of the MFC reactor. The power output of the MFC reactor increased sharply when E. coli in the reactor reached the critical concentration. Accordingly, the time required to reach the highest voltage output was recorded as a detection time (DT), and a negative linear relationship was established between DT and the logarithm of the initial concentration of E. coli in the samples studied. The DTs of laboratory samples were 140 min and 560 min for initial concentrations of 1.9 × 107 CFU/mL and 42 CFU/mL at 44.5 °C. Moreover, the DTs for GUS assays were further shortened by induction with methyl β-D-glucuronide sodium salt (MetGlu). The quantitative relationship between DTs and initial E. coli concentrations established from replicate laboratory sample assays allowed estimation of the E. coli concentration in environmental samples, but with approximately 100 min of lag time. The lag time was also observed with E. coli samples that were prepared by starving cells in a laboratory.