On the low-frequency dispersion observed in dielectrophoresis spectra.

The foundation of dielectrophoresis (DEP) as a tool for biological investigation is the use of the Clausius-Mossotti (C-M) factor to model the observed behaviour of cells experiencing DEP across a frequency range. Nevertheless, it is also the case that at lower frequencies, the DEP spectrum deviates from predictions; there exists a rise in DEP polarisability, which varies in frequency and magnitude with different cell types and medium conductivities. In order to evaluate the origin of this effect, we have studied DEP spectra from five cell types (erythrocytes, platelets, neurons, HeLa cancer cells and monocytes) in several conditions including medium conductivity and cell treatment. Our results suggest the effect manifests as a low-pass dispersion whose cut-off frequency varies with membrane conductance and capacitance as determined using the DEP spectrum; the effect also varies as a logarithm of medium conductivity and Debye length. These together suggest that the values of membrane capacitance and conductance depend not only on the impedance of the membrane itself, but also of the surrounding double layer. The amplitude of the effect in different cell types compared to the C-M factor was found to correlate with the depolarisation factors for the cells' shapes, suggesting that this ratio may be useful as an indicator of cell shape for DEP modelling.

Label-free, non-contact determination of resting membrane potential using dielectrophoresis.

Measurement of cellular resting membrane potential (RMP) is important in understanding ion channels and their role in regulation of cell function across a wide range of cell types. However, methods available for the measurement of RMP (including patch clamp, microelectrodes, and potential-sensitive fluorophores) are expensive, slow, open to operator bias, and often result in cell destruction. We present non-contact, label-free membrane potential estimation which uses dielectrophoresis to determine the cytoplasm conductivity slope as a function of medium conductivity. By comparing this to patch clamp data available in the literature, we have demonstratet the accuracy of this approach using seven different cell types, including primary suspension cells (red blood cells, platelets), cultured suspension cells (THP-1), primary adherent cells (chondrocytes, human umbilical mesenchymal stem cells), and adherent (HeLa) and suspension (Jurkat) cancer cell lines. Analysis of the effect of ion channel inhibitors suggests the effects of pharmaceutical agents (TEA on HeLa; DMSO and neuraminidase on red blood cells) can also be measured. Comparison with published values of membrane potential suggest that the differences between our estimates and values recorded by patch clamp are accurate to within published margins of error. The method is low-cost, non-destructive, operator-independent and label-free, and has previously been shown to allow cells to be recovered after measurement.