To support our hypothesis the event
To support our hypothesis, the event consisting of a bacterium contacting both IDEs was studied with the help of finite-element simulations (Comsol Multiphysics®). A three-shell model reported by Bai et al.  was used to simulate the bacterium. It describes E. coli with all its main components and its electrical properties, that is outer membrane (ε=10, σ=0S/m, thickness=7nm), periplasm (ε=60, σ=3S/m, thickness=10nm), inner membrane (ε=6, σ=0S/m, thickness=7nm) and cytoplasm (ε=81, σ=0.22S/m). A cylinder with hemispherical shells having a radius of 0.25μm and a total length of 2μm was chosen as the cell geometry, which is more similar to the real cell shape than the spheroid proposed in Bai\'s work. Dielectric parameters of the medium were ε=81 and σ=3·10S/m. Hence, the impedance of the IDEs was simulated as a capacitance (double layer and native oxide capacitance) and a resistance in series (polysilicon resistance). These parameters show values of 1.7·10F/m2 and 40Ω, respectively, when obtained from fitting the equivalent circuit to the experimental impedance spectra. Figs. 2a and b confirm that at low frequencies (1kHz) currents do not penetrate the bacterium outer membrane and by contrast, they do so at higher frequencies (1MHz). That is, at high frequencies bacteria behave as a conductive particle short-circuiting electrodes. Interestingly, at 1MHz, the bacterium periplasm accounts for 72.29% of the current owing to its higher conductivity despite it only represents 8.5% of the total cross-sectional area of the bacterium. We also simulated the perturbation that would cause cell shaped particles without membranes that were entirely insulating (ε=1, σ=0S/m), or entirely conductive (ε=81, σ=0.22S/m). Likewise, the perturbation of the IDE impedance caused by an arbitrary fixed decrease in the interface capacitance of 5·10F/m2 was estimated. Simulated spectra obtained for all these cases are plotted in Fig. 3a together with one experimental measurement. From these plots it is clear that impedance increments are related to the particle insulating behavior and decrements are associated to particle conductive behavior. Bacteria exhibit a dual electrical behavior that corroborates our initial hypothesis. Nevertheless, it should be mentioned that the experimental positive peak is larger than that simulated one and starts at lower frequency values. This behavior suggests that fasudil also produce a change in the interface impedance. An explanation to this phenomenon could be that the buffer used to carry out the measurement shows low conductivity and concentration of nutrients so that it could activate a general stress response to bacteria. This response is a common bacteria behavior that allows them to persist in a wide variety of environments. One of the effects of these stress conditions could be the formation of adhesion structures, such as curli and fimbriae and the secretion of exopolysaccharides [20, 21], which would produce an increase in the IDE interface impedance and that was not taken into consideration in the simulation experiments. In order to corroborate this explanation, the relative impedance variation spectra was obtained for other related events, namely an increase of media conductivity, the addition of an adsorbed BSA layer at the surface, and the attachment of 1μm-diameter silica beads. Fig. 3 b shows the curves obtained in each case. For a conductivity increase, from 2.8·10S/m to 3.1·10S/m, the impedance decreases in the middle frequency range similar to the response obtained when conductive particles were simulated. The impedance spectrum was measured before and after BSA adsorption in the same glycine solution (3·10S/m), as above. The corresponding spectrum in Fig. 3 b shows an increase of impedance in the low frequency range, as expected. This result is in accordance with those obtained for capacitive biosensors , where a decrease in capacitance is monitored due to antibody–antigen interactions taking place at the interface. Silica beads, which are dielectric particles, were deposited at the surface of the electrodes by precipitation from a suspension in deionized water. A bead density of 8.1·105beads/mm2 on the IDE surface was estimated by bead counting using an optical microscope. The impedance spectrum was measured in the same solution as above before and after silica bead deposition. The presence of the beads on the IDE surface does not significantly alter the impedance at low frequencies. This can be explained by the fact that they only block the electrode-solution interface at a single point of contact. Therefore, if the cells were cylindrical particles like those simulated, they would only contact the IDE surface on a single point or a line, and the effect on the interface impedance would be similar to that of the silica beads. The only feasible explanation for that impedance increase in the low frequency range is that the E. coli adhered covering a significant area of the IDE surface, this being likely related to the stress conditions to which they are submitted, as explained above, and thus showing a similar effect on the spectra to that due to the adsorption of proteins.