Substrate Arrays of Microelectrodes for in Vitro Electrophysiology

Substrate arrays of microelectrodes (MEAs) are microfabricated devices that can be electrochemically coupled to portions of brain tissue maintained in vitro (i.e., either cultures of dissociated neurons or brain tissue slices), because the electrically excitable cellular components in these experimental preparations are comparable in size with the MEA-engineered microstructures (see Figs. 1 and 2). These arrays are usually made of up to 100 independent microelectrodes, each with a diameter of a few tens of micrometers, embedded in a planar biocompatible substrate and spatially distributed over an area of a few square millimeters (see Figs. 3 and 4) (1). MEAs are employed as the bottom of an ad hoc electrophysiological recording chamber, or as a kind of Petri dish for cell cultures, in the standard settings of in vitro cellular electrophysiology. Thanks to appropriate electronic instrumentation, such as low-noise multichannel preamplifiers and stimulus isolators, it is possible to extracellularly and noninvasively detect and stimulate the distributed electrophysiological activity of a neuronal population in vitro, either acutely coupled on top of the MEAs (see Fig. 1, right panel), or growth on it during several weeks and cultured under sterile physiological conditions (see Fig. 2). The advantages of the MEAs can be better appreciated when compared with the current state-of-the-art networklevel electrophysiological approaches in vitro. These consist of simultaneous intracellular/patch-clamp access to the membrane voltage from visually identified single neurons, on manual or semiautomatic placement of several glass pipette electrodes (e.g., 5–10) in close contact with the somatic membranes (2). However, the maximal number of electrodes, thus of cells, is severely restricted by space constraints below the stage of an upright microscope and by the availability of very expensive (e.g., piezoelectric) multi-micromanipulator units. More importantly, such an approach is highly invasive and requires the experimenter to possess considerable skill for a quick and precise micromanipulation. Although superior in terms of the quality of the electrophysiological recordings, patch-clamp electrodes dialyze the intracellular compartments with the pipette solution, altering the physiological cytosolic concentrations, and they always irreversibly damage the neuronal membrane, which makes it extremely difficult to keep stable and reliable recording conditions for more than a few hours, posing serious limits to the experimental issues that can be investigated. On the contrary, MEAs offer the unique opportunity to monitor and stimulate the temporal electrochemical activity of a neuronal network with a much higher spatial resolution, noninvasively and over a longer time horizon (i.e., up to several months with cultured neurons). Although only an extracellular detection of neuronal spiking activity is usually possible [but see (3)], MEAs further opened the way to the long-term study and identification of the impact of metabolic and homeostatic neuronal mechanisms, affecting and modulating the spatial patterns of the network electrical activity, as well as of the development and plasticity of synaptic connections.