Efficient Capture of Pathogens in Liquid Streams By Phage-Based Biomolecular Filtering

This paper presents a filtration method to capture, concentrate, and isolate small quantities of bacterial pathogens in large volumes of liquid streams. The filter used in this work is a multi-layered planar arrangement of phage-coated, strip-shaped magnetic particles (4 mm × 0.8 mm × 0.03 mm) through which a liquid of interest flows. These magnetic particles are held and arranged by a magnetic force produced from solenoid coils. This “phage filter” is designed to capture specific bacterial pathogens through phage-based biomolecular recognition and allow non-specific targets to pass. An advantage of the phage filter is to eliminate a common clogging issue, which often occurs in conventional bead filters. The phage filter consists of a support frame, solenoid coils coupled to the support frame, and a plurality of phage-coated magnetic particles. The support frame is fabricated from a soft magnetic material. The solenoid is configured to generate a defined magnetic field. The phage-coated magnetic particles, magnetically coupled to the support frame, are arranged in a planar array, positioned within the opening of the support frame. Each magnetic particle is fixed at one end and can swing freely at the other end, allowing large, non-specific debris to pass. ANSYS-Maxwell magnetic simulations were run to obtain the magnetic flux vectors and field strength around the filter with varying electric currents and numbers of, coil layers and turns. The results were compared to experimental results produced using iron powders and a gaussmeter, followed by the selection of the optimal filter design. Then, a series of experiments were conducted where liquid solutions containing different concentrations of Salmonella were passed through the filter, and the capture efficiency was calculated by plate counting. The biomolecular recognition filter can be combined with standard detection methods, such as qPCR, to detect and identify pathogens rapidly.