Christoph Sprössler

Field-effect transistor array for monitoring electrical activity from mammalian neurons in culture

A field-effect transistor (FET) array has been fabricated and used for recording of electrical signals from neural cells. The array consists of p-channel FETs with non-metalized gates. The size of the gates of the 16 FETs are from 28 x 12 microns2 down to 10 x 4 microns2 and are arranged in a 4 x 4 matrix on 200 microns centers. For the device fabrication process we have especially focused on high sensitivity, good long-term stability in physiological conditions, and sufficient reduced signal-to-noise ratio. Special care was taken on the encapsulation technique of the device to allow surface modification based on the self-assembly technique. It can be shown that the microelectronic device surface can be modified with a synthetic peptide linked to the surface. Tailoring of the surface composition using this method allows hippocampal neurons to adhere and grow for days. More importantly, these cells develop typical electrical characteristics when cultured on this artificial surface. Using this approach neuron-FET couplings were recorded.

Ordered networks of rat hippocampal neurons attached to silicon oxide surfaces.

The control of neuronal cell position and outgrowth is of fundamental interest in the development of applications ranging from cellular biosensors to tissue engineering. We have produced rectangular networks of functional rat hippocampal neurons on silicon oxide surfaces. Attachment and network formation of neurons was guided by a geometrical grid pattern of the adhesion peptide PA22-2 which matches in sequence a part of the A-chain of laminin. PA22-2 was applied by contact printing onto the functionalised silicon oxide surface and was immobilised by hetero-bifunctional cross-linking with sulfo-GMBS. Geometric pattern matching was achieved by microcontact printing using a polydimethylsiloxane (PDMS) stamp. In this way the produced grid pattern ranged from 3 to 20 microm in line width and from 50 to 100 microm in line distances. As shown by atomic force microscopy (AFM), line widths and line distances of the peptide pattern differ less than 0.5 microm from the used PDMS stamp. The height of the layer of immobilised PA22-2 was approximately 3.5 nm implying the layer to be monomolecular. Immobilised PA22-2 was capable of binding anti-PA22-2 antibodies indicating that the function of the peptide was not compromised by immobilisation. Rat hippocampal neurons, cultured at low density in serum-free medium, were applied to the growth matrix of PA22-2-coated substrates and, within 1-3 h of culture, formed a network-like pattern that more or less matched the printed grid. Reliability and reproducibility of neuronal network formation depended on the geometry, line width and node diameter of the grid pattern. The immobilised neurons showed resting membrane potentials comparable with controls and, already after 1 day of culture, were capable of eliciting action potentials. The suitability of the immobilised neurons for the study of man-made neural networks and for multi-site recordings from a functional neuronal network is discussed.

Extended gate electrode arrays for extracellular signal recordings

Abstract We have fabricated arrays of planar gold electrodes arranged in a matrix of 8×8 with active areas ranging from 6 to 30 μm in diameter. An electronic amplification circuitry based on commercial junction field-effect transistors was used where the gold sensor fields act as extended gate electrodes (EGE) of the transistors, which leads to a new approach for long-term extracellular recording systems in vitro. The high input resistance of the amplification circuitry allows the use of small planar bare gold electrodes without further modification which therefore extends the frequency range of the measuring set-up down to the DC-level. The performance of our recording system has been tested using rat cardiac myocytes cultured directly on the device surface. The recorded signals were then compared in shape and size with recordings performed with a similar extracellular measurement set-up based directly on field-effect transistors with non-metallized gate electrodes. By simulations we could show the influence of the electrode capacitance on the time-course and on the size of the measured signals.

Long-term recording system based on field-effect transistor arrays for monitoring electrogenic cells in culture

Abstract This paper describes the use of a PC-based system for acquisition and processing of data recorded from electrical active and excitable cells cultured over microfabricated arrays of field-effect transistors. Using these recording devices, limitations of conventional recording techniques such as those associated with making long-term and multisite recordings can be overcome. This system has been tested using neonatal rat cardiac myocytes, the beating of which is correlated with simultaneous recorded intra- and extracellular voltage measurements.

Electrophysiological development of embryonic hippocampal neurons from the rat grown on synthetic thin films

We have studied the electrophysiological properties of hippocampal neurons grown on surfaces of organic thin films formed on glass or silicon substrates and on microelectronic device surfaces in culture. Hippocampal neurons were dissociated from embryonic rats and plated on substrates chemically modified with laminin peptide in a chemically defined medium. The electrophysiological properties of the neurons were studied using patch-clamp amplifier technique. We observed that the neurons grown on these substrates develop resting membrane potentials more negative than -33 mV after 3 days in culture and are able to produce action potentials. More interestingly we found that the neurons when grown on the microelectronic surfaces develop similar electrophysiological characteristics as those on the glass surfaces. Passive electrical properties (Cm = 27 +/- 5 pF, Rm > or = 1 G omega) of the neurons studied by impedance spectroscopy did not change considerably during the first week in culture.

Model network architectures in vitro on extracellular recording systems using microcontact printing

A PDMS stamp is used to transfer a synthetic peptide in a given pattern to any suitable surface. Using this method two-dimensional neuronal model networks could be formed on glass substrates as well as on electronic devices and adjusted to the given microelectronic structure. The present work focuses on the mechanism of neurite guidance under simplified in vitro conditions, using in vitro guidance cues and outline the incorporation of these interfacial methods into microelectronic sensor devices.

Neuron-silicon junction: electrical recordings from neural cells cultured on modified microelectronic device surfaces

A field-effect transistor (FET) array has been fabricated and used for recording of electrical signals from neural cells. The array consists of p-channel FETs with non-metallized gates. The size of the gates of the 16 FETs are from 28/spl times/12 /spl mu/m/sup 2/ down to 10/spl times/4 /spl mu/m/sup 2/ and are arranged in a 4/spl times/4 matrix on 200 /spl mu/m centers. Electrical signals of neural cells can be recorded by direct coupling with the FET. In order to control the neuronal survival and growth, the microelectronic device surface is modified with a synthetic peptide linked to the surface via self-assembly techniques. It can be shown that the composition of the surface can be tuned in such a way that hippocampal neurons show good adhesion and growth for days. More importantly, these cells develop typical electrical characteristics when cultured on this artificial surface. Using this approach passive neuron FET couplings were recorded.

Preparation, Structural Characterization and Functional Coupling of Tethered Membranes to Solid Substrates

One of the remaining major challenges of current scientific “hot topics” at the boundary between physics, chemistry, biology, medicine, materials science, and mechanical and electrical engineering is the interface between the living world of biomolecules, cells and tissues and the technical world of implants, sensor surfaces or signal transducers in neuroelectronic circuits. The understanding, design, fabrication, control and modification of these “bio-interfaces” will be in the center of many scientific activities aiming at compatibilizing the two spheres in an effort to not only induce a passive mutual toleration, but rather generate an interactive network of components originating from living organisms and, e.g., microelectronic devices (Nicolini, 1996).

Planar transistor array systems for extracellular recordings

An extracellular recording system has been designed for the detection of electrical cell signals. A field-effect transistor (FET) array has been fabricated which consists of p-channel or n-channel FETs with nonmetallized gates. The size of the gates of the 16 FETs are from 16/spl times/3 /spl mu/m/sup 2/ down to 5/spl times/1 /spl mu/m/sup 2/ and are arranged in a 4/spl times/4 matrix on 200 and 100 /spl mu/m centers. On the other side extended gate electrode (EGE) arrays were used which are arranged in a 8/spl times/8 matrix, on 200 and 100 /spl mu/m centers. The gate electrodes are made from gold, titanium and silicides with diameters down to 6 /spl mu/m. The cell-device coupling has been studied using various cell types e.g. neuronal cells, cardiac myocytes, and cells from cell lines. The recorded signals will be discussed on the base of a simple point contact model where contributions from passive as well as active membrane properties are included.

Neural cell pattern formation on microelectronic device surfaces

For the study of networks of live nerve cells it is necessary to control adhesion and outgrowth of these cells and to record signals from multiple sites. We have chosen to detect the electrical signal of the neuron by direct coupling with a field-effect transistor. Our approach to control the chemical composition of the surface of such a device is to attach ultrathin polymer films to the device surface by using a novel grafting from procedure. It can be shown that the composition of the surface can be tuned in such a way that neural cells grow only in defined regions. This method allow the patterning with lithographic techniques to create cell guidance for real neural networks.