Taxes & Accounting

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

Description
Long-term recording system based on field-effect transistor arrays for monitoring electrogenic cells in culture
Published
of 6
0
Published
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Transcript
  Biosensors & Bioelectronics 13 (1998) 613–618 Long-term recording system based on field-effect transistor arraysfor monitoring electrogenic cells in culture Christoph Spro¨ssler  a , Dirk Richter  a , Morgan Denyer  b,c , Andreas Offenha¨usser  a,c,* a  Max-Planck-Institute for Polymer Research, Ackermannweg 10, D-55128 Mainz, Germany b Centre for Cell Engineering, Joseph Black Building, University of Glasgow, Glasgow G12 8QQ, U.K. c Frontier Research Program, The Institute of Physical and Chemical Research (RIKEN), 2-1 Hirosawa, Wako-shi, Saitama 351-01, Japan Received 17 November 1997; accepted 5 March 1998 Abstract This paper describes the use of a PC-based system for acquisition and processing of data recorded from electrical active andexcitable cells cultured over microfabricated arrays of field-effect transistors. Using these recording devices, limitations of conven-tional recording techniques such as those associated with making long-term and multisite recordings can be overcome. This systemhas been tested using neonatal rat cardiac myocytes, the beating of which is correlated with simultaneous recorded intra- andextracellular voltage measurements.  ©  1998 Elsevier Science S.A. All rights reserved. Keywords:  Extracellular whole cell recording; Field-effect transistor array; Multisite recording 1. Introduction Monitoring of cell electrical activity is of interest in awide range of biological situations and is conventionallycarried out by recording the intracellular electrical poten-tial of nerve or muscle cells using either glass microelectrodes or patch clamp pipettes. Such electrodes canbe produced cheaply and relatively easily. However,they also have some limitations. The major disadvan-tages of these electrodes is that it is very difficult to rec-ord simultaneously from more than two or three sites ona small tissue or culture preparation, and impossible tomake long-term recordings due to the invasive nature of the technique.In contrast, over the past 20 years it has become poss-ible to detect and to measure rapid changes in membranepotential of neurons by voltage sensitive dyes as opticalprobes (Grinvald et al., 1981). This method allows therecording from multiple sites and therefore enables themonitoring of spatiotemporal optical parameters directlyrelated to neural activity. However, the toxicity of thesedyes on illumination makes this method unsuitable forlong-term recording. * Corresponding author. 0956-5663/98/$—see front matter  ©  1998 Elsevier Science S.A. All rights reserved.PII: S0956-5663(98)00016-5 To enable simultaneous detection of activity from alarge number of recording sites, extracellular metal elec-trodes can also be used. These record only a fraction of the transmembrane voltage. Several groups have usedmicrofabrication techniques to produce recording arraysin which the metal electrodes are embedded in the sub-stratum (Thomas et al., 1972; Gross et al., 1977). Inaddition extracellular signals of electrogenic cells can berecorded by means of a field-effect transistor (FET)(Fromherz et al., 1991; Offenha¨usser et al., 1997). In thisapproach cells are cultured directly on the chip-surfaceof an array of field effect transistors (FET). Signals arerecorded from the cell on top of the non-metallized gateof a FET. Changes in membrane voltage will then causea variation of the gate voltage which can be recorded asa difference in source-drain current.In this paper we describe a measuring system(controlled by a personal computer) for acquisition, stor-age, and analysis of electrical signals from a network of electrogenic cells. The proposed system is suitable forthe recording signals from a 4  ×  4 array of FETs. Thesignals can be monitored from up to 16 channels simul-taneously, and stored on a hard disk. In addition signalsform other devices necessary for the characterization of the functional link between cell and transducer (i.e.patch-clamp electrode) can be monitored and recordedsimultaneously with the signals recorded via the FETs.  614  C. Spro¨ ssler et al./Biosensors & Bioelectronics 13 (1998) 613–618  2. Measurement system The block diagram indicating the basic componentsof the recording system is shown in Fig. 1. It consistsof a first stage preamplifier, which converts the currentsignals of the 16 FETs on one chip into a voltage, anda second stage amplifier and control unit, which compen-sates for the offset and then amplifies the signal. At thissecond stage amplifier, additional external devices (i.e.patch-clamp amplifier) can be connected to the recordingsystem. The temperature of the FET array can be con-trolled by a heating system which is included in thesecond stage control unit. The whole recording systemis connected to a computer which is equipped with aADwin-4 SVIO 16 channel A/D-conversion boardincluding processor unit and memory (Keithley Instru-ments, Germering, Germany), for real time data rec-ording, and an 8255 IO-board (Conrad Electronic, Hir-schau, Germany), for the control of the amplifier stage.The chips with the FET arrays were fabricated usingconventional integrated circuit technology. The chipswere mounted on a 28 DIL chip carrier (NTK, Ratingen,Germany) and partially encapsulated to form an elec-tronic culture dish. This allows a fast exchange of theFET array. The details of the fabrication process aredescribed elsewhere (Offenha¨usser et al., 1997). Fig.2(a) shows such an electronic culture dish mounted onthe ‘zero-force’-socket of the preamplifier system. Fordemonstration purpose the upper lid of the preamplifiershousing was removed. Signals recorded with the FETarray usually have amplitudes ranging from 200 to 1000  V and are mainly superimposed with thermal noise of the transistor. The preamplifier stage was designed in away that the noise of the preamplifier should be below10–20   V which is sufficient for the requirements of future chip generations (thermal noise below 50   V).This was mainly achieved by a reduction of the overallsize of the preamplifier and by building the housing of the preamplifier stage from aluminum to reduce any Fig. 1. Block diagram showing the recording system consisting of the amplifier unit, the patch-clamp amplifier, the impedance analysatorand the data acquisition unit.Fig. 2. Photographs showing the ‘preamplifier’ unit (upper and lowerlids were removed). (a) ‘Zero-force’-socket with integrated heating platand mounted electronic culture dish. (b) The operation amplifiers arewere mounted in as single die directly on the printed-circuit board. possible electromagnetic interference. The reduction of size was mainly accomplished by a hybrid technology,i.e. the operation amplifiers were mounted as single dieon the printed-circuit board. Fig. 2(b) shows a photogra-phy of the amplifying electronic of the preamplifier stage(to allow this view the lower lid of the housing wasremoved).Fig. 3 shows a schematic of the current-to-voltageconverter (preamplifier), the compensation and theamplifier stage. The measurement system has beendesigned for 16 channels, i.e. every single transistor of the FET array can be addressed and recorded individu-ally. The first stage (current-to-voltage conversion)requires low offset current (typ. 30 fA), good CMRR(Common Mode Rejection Ratio, typ. 112 dB) for thedifferential stage, low-noise components in the fre-quency range between 10 and 5000 Hz and high-inputimpedance. Therefore an OPA128JD amplifier (Burr–Brown, Tucson, U.S.A.) with low input current noise [3fA (p–p)] has been chosen to meet the requirements. Thecurrent–voltage conversion was set to  U   =  10k   ·  I   (seeFig. 3). In order to be compatible with standard electro-  615 C. Spro¨ ssler et al./Biosensors & Bioelectronics 13 (1998) 613–618  Fig. 3. Block diagram of the amplification unit. The amplifier includes the eighteen 16-bit DA-converters for power supply and the compen-sation voltages. physiological recording systems (e.g. patch-clampamplifier) the reference electrode (Ag/AgCl-electrode)was set to ground. Therefore, to apply the gate-source( V  GS ) and the drain-source ( V  DS ) voltage to the FET-array the common source connector of the transistor wasset to  −  V  GS  and the positive input of the operationamplifier was set to  V  DS  −  V  GS  (see for comparisonFig. 3).The second stage of the system has been designed tocompensate the offset signals and then further amplifythese signals. This is necessary, because whilst recordingelectrical cell signals, the FETs are driven into the steep-est range of the transfer characteristic (maximum trans-conductance  g m ) leading to a large offset current [seeFig. 4(a)]. Therefore, in a first step the offset current iscompensated [see Fig. 4(b)] using a DAC707JP 16 bitDA-converter (Burr–Brown, Tucson, U.S.A.) which wascontrolled by the computer. Using this method offsetcompensation was carried out in    4 s for all 16 FETs.In a second step the signal is amplified by a gain of 100with an OP-07 amplifier (Analog Devices, Norwood,U.S.A.) [see Fig. 4(c)]. At this level a high-pass filterwith a cutoff frequency of 3 kHz was realized using acapacity of 780 pF in the feedback loop of the amplifier.Taking advantage of the compact design and the inte-gration of all components in one unit the overall noiseof the whole amplification system was reduced to 20  V(p–p). The noise spectrum (see Fig. 5) of the amplifiershows no significant interfered frequencies.A heating plate was integrated into the socket of thepreamplifier stage. This allows the temperature of theelectronic culture dish to be controlled within a range of RT to 45 ° C (input power maximum 2 W) [cf. Fig. 2(a)].To avoid additional electromagnetic interference an ana-log control consisting of a PG1.0910.1 PT100-sensorelement (JUMO, Fulda, Germany) and a PI-controller(maximum 6.8 V) is used (long-term stability   0.1 ° C).In addition an external 500-0420-002 PT100-sensorelement (Klara Ru¨senberg, Gachenbach, Germany)could be placed in the culture dish. 3. Data acquisition The data acquisition process was controlled by thecomputer: application of the voltages to the FETs, com-pensation of the signal offsets, and additional tasks werecontrolled by the 48  +  6 digital IO channels. The signalswere acquired using the analog-to-digital conversionboard with 16 single-ended input channels with 12-bitresolution, and the gain of each channel could be indi-vidually set to 1, 2, 4, 8 by software control.The software for the control of the measurement andthe data acquisition was written using TestPoint(Keithley, Germering, Germany) and the real-time pro-cessing was written using ADBasic (Ja¨ger Messtechnik,Lorsch, Germany). Firstly recording channels wereselected and the gains were set as required. After apply-ing the gate-source ( V  GS ) and the drain-source ( V  DS ) volt-age to the FET array the internal channels were compen-sated automatically. Using this method a signal shift dueto the drift of an FET can be promptly balanced. Inaddition, verification of signals recorded by the FETswas achieved by selecting a channel through which sig-nals from an external source (such as a patch-clampamplifier) could be processed. Two modes of data acqui-sition were used: continuous recording and recording fora fixed time. In the continuous mode the data were con-  616  C. Spro¨ ssler et al./Biosensors & Bioelectronics 13 (1998) 613–618  Fig. 4. Principle of the compensation and amplification. The signaloverlaid with a large offset (a) is compensated (b) and amplified (c).Fig. 5. Noise spectrum of the whole amplification unit. The spectrumshows no significant frequency interference. tinuously sampled and monitored. Experiments with atime resolution of up to 10 kHz can be done for about80 s and are stored on the memory of the AD-board.They can be selectively monitored and stored to harddisk. In addition a simple signal processing of the datais possible by using software Fast Fourier Transform-ation and bandpass filter. For long-term experiments theprocessor on the AD-board can be used to control thedata in order to detect spike events. 4. Results In order to evaluate the performance of the measuringsystem electrical signals from rat cardiac muscle cellswere recorded. The cardiac myocytes were prepared fol-lowing a technique which has been described in detailelsewhere (Bhatti et al., 1989). Briefly the hearts wereremoved from 1 to 3 day old rats, finely minced anddissociated and plated in serum containing medium. TheFET array were cleaned, coated with fibronectin forabout 1 h, washed and then plated with 80   l of a 5  × 10 5 to 9  ×  10 5 cells/ml cardiac myocyte suspension.After plating the devices were incubated at 37 ° C in ahumidified atmosphere. Under these conditions, within2–3 days of plating, a confluent monolayer of cells exhi-biting spontaneous rhythmic activity developed. Theelectrical activity of these cells was studied starting onthe 3rd day after plating.Figs. 6 and 7 summarizes the results of our studies onsingle and multi-site recordings conducted in normal airusing the culture medium as recording solution. Fig. 6shows typical recordings from cardiac muscle cells per-formed simultaneously with an intracellular microelec-trode (upper trace) and an FET (lower trace). Intracellu-lar recordings were made from cells grown several   maway from the recording site of the FET. Details aboutthe interpretation of the recorded extracellular signals arepublished elsewhere (Spro¨ssler et al., 1998). Fig. 7shows spontaneous extracellular activity from a mono-layer of cardiac muscle cells recorded from four differentFETs. From the time delay between the recordings of the action potentials at the various sites we could esti-  617 C. Spro¨ ssler et al./Biosensors & Bioelectronics 13 (1998) 613–618  Fig. 6. Intracellular (with microelectrode) and extracellular (with FET) recording of cardiomyocites. The cell layer showed a regular spontaneousbeating. The measured source-drain current is assigned to a gate voltage using the transconductance  g m  of the FET. The signals at the gate electroderecorded with the FET were in the range of 500   V.Fig. 7. The FET-array allows cell-recording from many recording sites simultaneously. The graph shows the time delayed current signals of fourFETs at different locations on the array. mate the srcin, the direction and the velocity (ca 0.2m/s) of the burst pattern assuming an isotropic spreadingof the excitation between the cells. 5. Conclusion A system for extracellular recording with an FETarray has been described. Care was taken to achieve alow-noise amplifying system. In order to perform long-term studies without affecting the cells a temperaturecontrol unit was integrated. The performance and thefunctionality of the recording system was demonstratedby monitoring the electrical activity of rat cardiac myo-cytes in culture. The system will allow the simultaneousrecording from up to 16 channels with a time resolutionof 2.5 kHz. In a future development of our recordingsoftware data preprocessing can be done to enable spikedetection and exclusion of insignificant data. Acknowledgements We thank Professor Dr W. Knoll (Max-Planck-Insti-tute for Polymer Research and FRP, RIKEN), ProfessorA.S.G. Curtis (University of Glasgow) and Dr S.T. Brit-land (University of Bradford) for their support and sti-mulating discussions. Special thanks to Mr. Schuster andMr. Mu¨ller (Max-Planck Institute for Polymer Research)for their work on the electronic setup. M. Lacher and DrT. Zetterer (all from the Institute fu¨r Mikrotechik,Mainz) are acknowledged for their helpful discussions
Search
Similar documents
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x