Optimization of Phage and Surface Blocker Loading for the Magnetoelastic Biosensor

This paper investigates the optimization of phage and surface blocker loading for the magnetoelastic (ME) biosensor that is being developed for real-time, direct pathogen detection on food surfaces. E2 phage with high binding affinity toward Salmonella Typhimurium was selected as a model bio-recognition element and tested along with five commercially available surface blocking reagents: BSA, StartingBlock, SuperBlock, Casein and BLOTTO (purchased from Thermo Fisher Scientific, Inc.). The effects of the concentration and loading time of these phage and surface blockers on biosensor performance were investigated. The concentrations of E2 phage and the surface blockers tested were 5×10⁹, 5×10¹⁰, 5×10¹¹ and 5×10¹² vir/mL and 1×, 0.5× and 0.25×, respectively. In addition, various loading times were tested: 30, 60 and 90 min for E2 phage and 60 and 120 min for the blockers. To visualize and measure surface coverage, E2 phage was labeled with the Alexa Fluor 488 dye, whereas the blockers were each labeled with the Texas Red dye (purchased from Thermo Fisher Scientific, Inc.). After loading steps, fluorescence micrographs of biosensor surfaces were captured, followed by image processing (gray-scale conversion, background subtraction, and thresholding) to measure the surface coverage. Concentrations and loading times that yield the largest phage coverage and avoid over-coverage of the blockers were selected. Furthermore, comparisons among the blockers were conducted to select the best blocking reagent (i.e., SuperBlock for this investigation). These optimized loading conditions were then used to construct measurement sensors (with phage) and control sensors (without phage) for direct detection of S. Typhimurium on fresh apples. Both measurement and control sensors were placed on S. Typhimurium-spiked apple surfaces (1.0×10⁶ CFU/mm²), and the sensors’ resonant frequency shifts were measured as a function of time using a surface-scanning coil detector. While the resonant frequency shifts of the control sensors remained negligible, those of the measurement sensors decreased largely as time elapsed, indicating that specific binding of S. Typhimurium on the measurement sensor had occurred. The frequency data were further analyzed statistically, and the detection was found to be realized in 15 minutes. Finally, scanning electron microscopy was used to confirm the specific binding of S. Typhimurium on the measurement sensors.