Binding strategies for capturing and growing Escherichia coli on surfaces of biosensing devices.

Antibiotic resistant bacteria have become a threat to world health. An advanced method of detection, based on matrix assisted laser desorption ionization time-of-flight mass spectroscopy can identify bacteria relatively rapidly, but it is not suitable to measure bacterial antibiotic resistance. Biosensors may be able to detect resistance by monitoring growth after capture on sensor surfaces but this option has not been addressed adequately. We have evaluated the growth of Escherichia coli after capture in 96 well microplates and observed that growth/capture efficiency was relatively similar for antibody-based techniques, but non-specific capture varied considerably. We confirm that neutravidin binds E. coli non-specifically, which limited its use with biotinylated antibodies or aptamers. Centrifugation enhanced bacterial growth/capture considerably, indicating that procedures enhancing the interaction between bacteria and surface-bound antibody have the potential to improve growth efficiency. Capture and growth required larger numbers of bacteria than capture and detection on biosensor surfaces. Previously, we reported that the minimum concentration of live E. coli required for initiating growth on a GaAs/AlGaAs biosensor was ~ 105 CFU/mL (Nazemi et al., 2018), and we speculated that this could be related to the poisonous effect of Ga- and As-ions released during dark corrosion of the biosensor, however in the present report we observed that the same minimum concentration of E. coli was required for growth in an ELISA plate. Thus, we argue that this limitation was related rather to bacterial inhibition by the capture antibodies. Indeed, antibodies at titres designed to capture bacteria inhibited bacterial growth when the bacteria were added to growth medium at titres less than 105 CFU/mL, indicating that antibodies may be responsible for the higher limits of sensitivity due to their potential to restrict bacterial growth. However, we did not observe E. coli release after 6 h following the capture indicating that these bacteria did not degrade antibodies.