Determination of the Viability of Escherichia coli at the Single Organism Level by Electrorotation

If a bioparticle, such as a cell, bacterium, or virus, is exposed to an external electric field, it becomes electrically polarized. This polarization takes the form of electrical charges that are created on its external surface, as well as at the interfaces defining its internal structure. These field-induced charges lend to the bioparticle the properties of a large, electric dipole moment. The polarity and temporal orientation of this dipole, with respect to the applied field, depends on the dielectric properties of the suspending medium and the bioparticle (1) . We used microelectrodes to create various forms of localized electric fields having frequencies covering the range from ∼100 Hz to 10 MHz. There are three principal types of AC fields: (a) a stationary, nonuniform field (1)(2)(3) that creates a dielectrophoretic force; (b) a traveling wave (1)(2)(3) that creates a dielectrophoretic force and a traveling wave force; and (c) a rotating field (4)(5) that creates rotational torque. The effects on the suspended bioparticles are, respectively, localized lateral motion, long-range lateral motion, and rotation (references are to the background theory and technology). In the work reported here, rotating electric fields are used to investigate the binding of microorganisms to 6-μm diameter beads. In particular, we wished to determine whether electrorotation measurements are capable of detecting the viability of single Escherichia coli microorganisms attached to latex beads. E. coli (K12) was obtained from the National Collection of Industrial and Marine Bacteria, Ltd., Aberdeen. Carboxylated latex beads of major axis 6 μm were obtained from Bangs Laboratories, and the polyclonal E. coli antibody was from Biogenesis. E. coli antibody was covalently attached to the latex beads via a two-step, carbodimer-activated process (6) . The beads and E. coli were suspended in solutions of 170 mmol/L mannitol in 15 mmol/L phosphate-buffered saline solution, and the conductivity was adjusted to 0.5 ± 0.05 mS/m using KCl. The antibody-coated beads were suspended at a working concentration of 18 ± 4 × 109/L. E. coli attachment was achieved by introducing the microorganisms into the bead suspension at a concentration of approximately double that of the bead concentration. The E. coli concentration was chosen for the speed of attachment between it and the bead. The suspension of beads and E. coli were left for no longer than 30 min before experimentation, by which time a sufficient number of E. coli had attached to the surface of the bead. Some of the E. coli were rendered nonviable by suspending them in Virkon disinfectant (Antec International) to minimize damage to their surface antigens. For the electrorotation experiments, the rotating electric field was generated by energizing four electrodes with sinusoidal voltages of 90° phase separation. The microelectrodes were fabricated on glass microscope slides using photolithography (7)(8)(9) , and the working space between the electrodes was 400 μm. Rotation was monitored using a microscope and video recorder. As shown in Fig. 1 , throughout the frequency range monitored (100 Hz to 5 MHz) the antibody-coated beads were observed to rotate in the same sense as the applied rotating field. This is termed co-field rotation. The spectrum is characterized by two co-field rotation peaks, a low-frequency peak centered around 200 Hz and a high-frequency peak centered around 400 kHz. Antibody-coated latex beads therefore exhibit a distinct electrorotation spectrum. This can be used to monitor the efficiency of the antibody-attachment process. Electrorotation spectra obtained on addition of one viable (▴) and one nonviable (▵) E. coli to the surface of separate antibody-coated beads (○). Fig. 1 also shows that the electrorotation spectrum of an antibody-coated bead is markedly altered when either a single viable or nonviable E. coli particle becomes attached to it. In particular, an anti-field rotation peak, centered around 40 kHz, emerges for a latex bead with one viable E. coli attached, and the high-frequency peak shifts from 400 kHz up to 1.5 MHz. On attachment of a single nonviable E. coli, an anti-field rotation peak does not appear, but both co-field rotation peaks are markedly reduced, with the high-frequency peak shifting from 400 kHz up to 800 kHz. The electrorotation spectra of beads with more than a single attached E. coli were also investigated. An anti-field peak emerges for the attachment of two nonviable E. coli, and because of this, the clearest indication of the attachment of either a viable or nonviable E. coli is at the single organism level. With further addition of either viable or nonviable E. coli, the anti-field peak grows in prominence. The further accrual of E. coli on the bead surface also affects the “cross-over frequencies” at which the sense of rotation reverses. The lower “cross-over frequency” decreases, whereas the higher one increases, and this trend has been observed for the attachment of up to seven viable and four nonviable E. coli particles. This work clearly demonstrates the sensitivity to be achieved using the electrorotation assay. The attachment of just one E. coli particle to the surface of a bead produces a marked response. Furthermore, the technique provides a rapid method for determining the viability of a single microorganism. To understand the various observed transformations in the electrorotation spectra, mathematical modelling was performed with MatLab®. The particles were represented using the so-called multishell model (7)(8) . For the case where a clean bead is coated with an antibody, analysis of a single shell model indicates that the change in the electrorotation spectrum arises as a result of a decrease in the effective conductivity of the surface of the bead. This result is probably related to the fact that the uncoated latex particles have carboxyl surface functional groups, which result in their surface being able to support a relatively high surface conductivity associated with the percolation of counterions. Binding of the antibody probably yields a reduction of net negative surface charge. To analyze the effects arising from the attachment of E. coli to the bead, a multishell model was required ((7)(8)). Mathematical analysis implied that the results, in general, could also be explained by a reduction in the effective conductivity of the bead/antibody surface. For viable E. coli, this is mainly associated with the highly resistive nature of its intact cytoplasmic membrane. This effect is less marked when nonviable bacteria became attached, because one result of cell death is that the membrane looses its function as a selectively permeable membrane and becomes porous (conductive) to ions in general. The rate of attachment of E. coli to the beads, and hence the rate of change of their electrorotation spectra, was found to depend on the concentration of the bacteria in the suspending fluid. Therefore, by monitoring the changes of electrorotation of a number of beads suspended in a solution containing E. coli, it should in principle be possible to assay both the concentration and relative viability of the E. coli. The analysis could be accomplished by using a recently developed computer-aided technique for the automatic measurement of electrorotation (9) . This technique could also be used to determine the sensitivity of antimicrobial agents toward viable microorganisms attached to the beads, in the same way as electrorotation measurements have been used to determine the toxicity of biocides toward biofilms grown on bead surfaces (8) . Current work is directed toward exploiting these observed changes in the electrokinetic behavior of bead plus E. coli complexes in the selective dielectrophoretic manipulation and characterization of bioparticles at low concentrations (10) . The electrorotation assay is a sensitive technique that can be used for the rapid detection and viability determination of microorganisms, such as E. coli, at the single organism level. Thus, in principle, by interacting a large number of antibody-coated beds into a solution containing a small concentration of microorganisms, it should be possible to assay the number density and viability of these microorganisms without the requirement for culturing. Institute of Molecular and Biomolecular Electronics, University of Wales, Bangor LL57 1UT, UK This work has been supported by an EPSRC-CASE studentship sponsored by Severn Trent Water, Ltd. We thank Stephen Bone, Julian Burt, Adrian Parton, Andy Goater, John Kerslake, and John Tame for technical advice.