Electric field-induced effects on neuronal cell biology accompanying dielectrophoretic trapping.

Trapping neuronal cells may aid in the creation of the cultured neuron probe. The aim of the development of this probe is the creation of the interface between neuronal cells or tissue in a (human) body and electrodes that can be used to stimulate nerves in the body by an external electrical signal in a very selective way. In this way, functions that were (partially) lost due to nervous system injury or disease may be restored. First, a direct contact between cultured neurons and electrodes is created. This is realized using a microelectrode array (MEA) which can be fabricated using standard photolithographic and etching methods. Section 1 gives an overview of the human nervous system, methods for functional recovery focused on the cultured neuron probe, and the prerequisites for culturing neurons on a microelectrode array. An important aspect in the selective stimulation of neuronal cells is the positioning of cells or a small group of cells on top of each of the electrode sites of the MEA. One of the most efficient methods for trapping neuronal cells is to make use of di-electrophoresis (DEP). Dielectrophoretic forces are created when (polarizable) cells are located in nonuniform electric fields. Depending on the electrical properties of the cells and the suspending medium, the DEP force directs the cells towards the regions of high field strength (positive dielectrophoresis; PDEP) or towards regions of minimal field intensities (negative dielectrophoresis; NDEP). Since neurons require a physiological medium with a sufficient concentration of Na+, the medium conductivity is rather high (~ 1.6 S/m). The result is that negative dielectrophoretic forces are created over the entire frequency range. With the use of a planar quadrupole electrode sturcture negative forces are directed so that in the center of this structure cell can be collected. The process of trapping cortical rat neurons is described in Sect. 2 theoretically and experimentally. Medium and cell properties are frequency-dependent due to relaxation processes, which have a direct influence on the strength of the dielectrophorectic force. On the other hand, the nonideal material properties of the gold electrodes and glass substrate largely determine the electric field strength created inside the medium. Especially, the electrode-medium interface results in a significant loss of the imput signal at lower frequencies (< 1 MHz), and thus a reduction of the electric field strength inside the medium. Furthermore, due to the high medium conductivity, the electric field causes Joule heating. Local temperature rises result in local gradients in fluid density, which induces fluid flow. The electrode-medium interface and induced fluid flow are theoretically investigated with the use of modeling techniques such as finite elements modeling. Experimental and theoretical results agreed with each other on the occurrence of the effects described in this section. For the creation of the cultured neuron probe, preservation of cell viability during the trapping process is a prerequisite. Cell viability of dielectrophoretically trapped neurons has to be investigated. The membrane potential induced by the external field plays a crucial role in preservation of cell viability. The membrane can effectively be represented by a capaticance in parallel woth a low conductance; with increasing frequency and /or decreasing field strength the induced membrane potential decreases. At high induced membrane potentials ths representation for the membrane is no longer valid. At this point membrane breakdown occurs and the normally insulating membrane becomes conductive and permeable. The creation of electropores has been proposed in literature to be the cause of this high permeability state. Pores may grow or many small pores may be created which eventually may lead to membrane rupture, and thus cell death. Membrane breakdown may be reversible, but a chemical imbalance created during the high permeability state may still exist after the resealing of the membrane. This may cause cell death after several hours or even days after field application. Section 3 gives a detailed description of membrane breakdown. Since many investigations on electroporation of lipid bilayers and cell membranes are based on uniform electric fields, a finite element model is used to investigate induced membrane potentials in the nonuniform field created by the quadropole electrode structure. Modeling results are presented in cmbination with the results of breakdown experiments using four frequencies in the range from 100 kHz to 1MHz. Radomly positioned neuronals cells were exposed to stepwise increasing electric field strengths. The field strength at which membrane rupture occurred gives an indication of the maximum induced membrane potential. Due to the nonuniformity of the electric field, cell collapse was expected to be position-dependent. However, at 100 kHz cells collapsed at a break down level of about 0.4 V, in contradistinction to findings at higher frequencies where more variation in breakdown levels were found. Model simulations were able to explain the experimental results. For examining whether the neuronal cells trapped by dielectrophoresis were still viable after the trapping process, the frequency range was divided into two ranges. First, a high frequency (14 MHz) and a rather low signal amplitude (3 Vpp) were used to trap cells. At this high frequency the field-induced membrane potential is small according to the theoretical model, and therefore no real damage is expected. The experimental analysis included the investigation of the growth of the neurons, number and length of the processes (dendrites and axons), and the number of outgrowing (~ viable) versus nonoutgrowing (~ nonviable) neural cells. The experimental results agreed with the expectation. The effect of the use of driving signals with lower frequencies and/or higher amplitudes on cell viability was investigated using a staining method as described in the second part of Sect. 4. Survival chances are not directly linked to the estimated maximum induced membrane potential. The frequency of the dield plays an important role, decreasing frequency lowering the chance of survival. A lower frequency limit of 100 kHz is preferable at field strengths less than 80 k V/m, while with increasing field strength this limit shifts towards higher frequencies. The theoretical and experimental results presented in this review form the inception of the development of new electrode structures for trapping neuronal cells on top of each of the electrodes of the MEA. New ways to investigate cell properties and the phenomenon of electroporation using electrokinetic methods were developed that can be exploited in future research linking cell biology to technology.